052119380 x

This page intentionally left blank

Case Studies inNeuroanesthesia andNeurocritical Care

Case Studies inNeuroanesthesia andNeurocritical CareEdited byGeorge A. MashourUniversity of Michigan

Ehab FaragCleveland Clinic

cambridge university pressCambridge, New York, Melbourne, Madrid, Cape Town, Singapore,Sao Paulo, Delhi, Dubai, Tokyo, Mexico City

Cambridge University PressThe Edinburg Building, Cambridge CB2 8RU, UK

Published in the United States of America byCambridge University Press, New York

www.cambridge.orgInformation on this title: www.cambridge.org/9780521193801

c© Cambridge University Press 2011

This publication is in copyright. Subject to statutory exception andto the provisions of relevant collective licensing agreements, noreproduction of any part maytake place without the written permission of Cambridge UniversityPress.

First published 2011

Printed in the United Kingdomat the University Press, Cambridge

A catalog record for this publication is available from theBritish Library

Library of Congress Cataloging in Publication dataCase studies in neuroanesthesia and neurocritical care / edited byGeorge A. Mashour, Ehab Farag.p. ; cm.

Includes bibliographical references and index.ISBN 978-0-521-19380-1 (pbk.)1. Anesthesia in neurology – Case studies. 2. Nervous system –Surgery – Case studies. 3. Neurological intensive care – Casestudies. I. Mashour, George A. (George Alexander), 1969–II. Farag, Ehab.[DNLM: 1. Anesthesia – methods – Case Reports.2. Neurosurgical Procedures – methods – Case Reports.3. Critical Care – methods – Case Reports. 4. Nervous SystemDiseases – surgery – Case Reports.5. Perioperative Care – methods – Case Reports. WL 368]RD87.3.N47C37 2011617.9′6748 – dc22 2010041507

ISBN 978-0-521-19380-1 Paperback

Cambridge University Press has no responsibility for thepersistence or accuracy of URLs for external or third-party internetwebsites referred to in this publication, and does not guarantee thatany content on such websites is, or will remain, accurate orappropriate.

Every effort has been made in preparing this book to provideaccurate and up-to-date information which is in accord withaccepted standards and practice at the time of publication.Although case histories are drawn from actual cases, every efforthas been made to disguise the identities of the individuals involved.Nevertheless, the authors, editors and publishers can make nowarranties that the information contained herein is totally freefrom error, not least because clinical standards are constantlychanging through research and regulation. The authors, editors andpublishers therefore disclaim all liability for direct or consequentialdamages resulting from the use of material contained in this book.Readers are strongly advised to pay careful attention to informationprovided by the manufacturer of any drugs or equipment that theyplan to use.

Contents

List of contributors xPreface xvii

Neuroanesthesia

Part I – CraniotomySupratentorial craniotomy

1. Preoperative evaluation 3Paul Smythe and Mohit Rastogi

2. Evaluation and anestheticmanagement of elevated intracranialpressure 7Andrew Bielaczyc and Paul Smythe

3. Traumatic brain injury 11Adam Brown and Paul S. Moor

4. Postoperative seizure 15Allen Keebler

5. Painmanagement for craniotomies 17Paul E. Hilliard and Chad M. Brummett

Posterior fossa craniotomy

6. Preoperative evaluation 21George A. Mashour

7. Air embolism 24Jennifer Vance and Sheron Beltran

8. Delayed emergence after posteriorfossa surgery 27Leif Saager and Alparslan Turan

9. Trigeminal neuralgia 30Basem Abdelmalak and Joseph Abdelmalak

10. Trigeminocardiac reflex 33Heather Hervey-Jumper andChristopher R. Turner

11. Sitting craniotomy 36Michael D. Maile and George A. Mashour

12. Cerebellar hemorrhage 38Gene H. Barnett

Part II – Vascular proceduresAneurysm clipping

13. Preoperative evaluation 41Milad Sharifpour and Paul S. Moor

14. Intracranial aneurysm clipping withintraoperative rupture 45Paul S. Moor

15. Awake fiberoptic intubation 48Heather Hervey-Jumper and ChristopherR. Turner

16. Patient with coronary artery stent 51Richard Bowers and George A. Mashour

17. Deep hypothermic circulatory arrestfor intracranial aneurysm clipping 55Matt Giles and George A. Mashour

18. Neuroprotection during surgical clipligation of cerebral aneurysms 59Neeraj Chaudhary, Joseph J. Gemmete,B. GregoryThompson and Aditya S.Pandey

19. Dexmedetomidine and nitrousoxide for cerebral aneurysmclipping 63Michael J. Claybon andGeorge A. Mashour

20. Anaphylaxis associated withindocyanine green administration forintraoperative fluorescenceangiography 66Marnie B. Welch and Laurel E. Moore

v

Contents

Aneurysm coiling

21. Subarachnoid hemorrhage duringaneurysm coiling 70Khoi D. Than, Anthony C. Wang, NeerajChaudhary, Joseph J. Gemmete, Aditya S.Pandey and B. GregoryThompson

22. Cardiac abnormalities aftersubarachnoid hemorrhage 73William R. Stetler, Jr. and George A.Mashour

23. Postoperative retroperitonealbleed 77Stacy Ritzman

24. Angiography in the patient withkidney failure 80Jerome O’Hara and Mauricio Perilla

Arteriovenous malformation

25. Postoperative normal perfusionpressure breakthrough 83Juan P. Cata and Andrea Kurz

Carotid endarterectomy

26. Preoperative evaluation 85Maged Argalious

27. Monitoringmodalities 88Todd Nelson and Paul Picton

28. Neurologic decline after carotidsurgery 91Phil Gillen

29. Postoperative hematoma and airwaycompromise after carotidendarterectomy 94Maged Argalious

30. Postoperative stroke after carotidendarterectomy 96J. Javier Provencio

31. Postoperative myocardial infarction 99Maged Argalious

Part III – Functional neurosurgery32. Preoperative evaluation for deep

brain stimulator surgery 103Milind Deogaonkar

33. Airway crisis during deep brainstimulator placement 106Ehab Farag

34. Postoperative management ofParkinson’s medications 108Milind Deogaonkar

35. Epilepsy surgery: intraoperativeseizure 110Oren Sagher and Shawn L. Hervey-Jumper

36. Awake craniotomy and intraoperativeneurologic decline 113Oren Sagher and Shawn L. Hervey-Jumper

37. Epilepsy surgery and awakecraniotomy 117Sumeet Vadera and William Bingaman

Part IV – Spine surgerySpinal cord injury

38. Acute surgery: spinal and neurogenicshock 121Ruairi Moulding and Scott T. McCardle

39. Returning patient with autonomichyperreflexia 125Sherif S. Zaky

Complex spine surgery

40. Preoperative evaluation 128Julie McClelland and Ellen Janke

41. Loss of evoked potentials 133Thomas Didier and Ellen Janke

42. Effects of anesthesia onintraoperative neurophysiologicmonitoring 137Uma Menon and Dileep R. Nair

vi

Contents

43. Neurofibromatosis type 1 and spinaldeformity 140George A. Mashour

44. Major vascular complication duringspine surgery 143Zeyd Ebrahim and Jia Lin

45. Complex spine surgery for a Jehovah’sWitness 146Matthew Martin and Vijay Tarnal

46. Diplopia following spine surgery 149Alaa A. Abd-Elsayed and Ehab Farag

47. Postoperative visual loss in spinepatients 150Goodarz Golmirzaie and Laurel E.Moore

48. Prone cardiopulmonary resuscitation 154Justin Upp and Sheron Beltran

49. Open spine stabilization withpolymethylmethacrylateaugmentation 157Mariel Manlapaz and Greta Jo

50. Extubating the trachea afterprolonged prone surgery 160Vijay Tarnal and Robert Scott Kriss

51. Perioperative peripheral nerveinjury 163Lisa Grilly, George A. Mashour andChad M. Brummett

Cervical spine and airway issues

52. Unstable cervical spine 165David L. Adams and David W. Healy

53. Cervical spine limitations 168David L. Adams and David W. Healy

54. Rheumatoid disease 172David L. Adams and David W. Healy

Part V – Neuroendocrine surgery55. Preoperative evaluation 177

Nicholas F. Marko and Robert J. Weil

56. Acromegaly and gigantism 182D. John Doyle

57. Perioperative diabetes insipidus 184Juan P. Cata and Ehab Farag

Part VI – Pediatric neuroanesthesia58. Craniotomy 187

William Bingaman and Marco Maurtua

59. Ventriculoperitoneal shunt 191Ruairi Moulding and Peter Stiles

60. Craniosynostosis repair 195Vera Borzova and Julie Niezgoda

61. Scoliosis 199George N. Youssef

62. Hemispherectomy for treatment ofintractable epilepsy in an infantwith congenital antithrombin IIIdeficiency 203Rami Karroum and Alina Bodas

63. Neurosurgical procedures for pediatricpatients with cardiac malformations 207Stephen J. Kimatian, Kenneth Saliba andErin S. Williams

64. Neuroprotection during pediatriccardiac anesthesia 210Stephen J. Kimatian, Erin S. Williams andKenneth Saliba

Part VII – Neurologic sequelae inother patient populationsPregnancy

65. Pregnant patient with aneurysm 215Karen K. Wilkins and Alexandra S.Bullough

66. Anesthetic management of pregnantpatients with brain tumors 218Alaa A. Abd-Elsayed and Ehab Farag

67. Eclamptic seizures 221Negmeldeen F. Mamoun

vii

Contents

68. Postpartum headache 227Alexandra S. Bullough

Miscellaneous

69. Increased intracranial pressure withacute liver failure 230Ryen D. Fons and Paul Picton

70. Permissive hypertension in a patientwith vonWillebrand’s disease and apreexisting ventriculoperitonealshunt 233Miguel Cruz, Maged Guirguis and WolfH. Stapelfeldt

71. Perioperative acute ischemic stroke ingeneral surgical procedures 238Milad Sharifpour and George A.Mashour

72. Neurologic complications followingcardiothoracic surgery 241Donn Marciniak and Colleen G. Koch

73. Anesthetic management forsubdural hematoma evacuation in apatient with a left ventricular assistdevice 244Marcos Gomes, Hesham Elsharkawy,Endrit Bala and Ehab Farag

Neurocritical care

Part VIII – General topics inneurocritical care

74. Hypotension 251Edward Noguera

75. Mechanical ventilation 253Piyush Mathur and Vikram Dhawan

76. Mechanical ventilation for acutelung injury in the neurosurgicalpatient 257James M. Blum

77. Change inmental status 260Edward Manno

78. Therapeutic hypothermia using anendovascular approach in theneurocritical care patient 262Anupa Deogaonkar and Andrea Kurz

79. Therapeutic hypothermia aftercardiac arrest 265James W. Jones and Piyush Mathur

Part IX – Subarachnoid hemorrhage80. Cerebral vasospasm 269

Edward Manno

81. Ventriculoperitoneal shuntdependence 272Vivek Sabharwal and Asma Zakaria

82. Ventriculostomy infection 275Samuel A. Irefin

83. Sodium abnormalities in neurocriticalcare 277William R. Stetler, Jr. and George A. Mashour

Part X – Stroke84. Initial management 281

Lauryn R. Rochlen

85. Increased intracranial pressure at48 hours poststroke 284Lauryn R. Rochlen

Part XI – Intraparenchymalhemorrhage

86. Hypertensive intracerebralhemorrhage 287James F. Burke and Teresa L. Jacobs

87. Intracerebral hemorrhage andanticoagulation 291Eric E. Adelman and Teresa L. Jacobs

88. Cerebral amyloid angiopathy-relatedintracranial hemorrhage 295Lesli E. Skolarus and Teresa L. Jacobs

viii

Contents

Part XII – Traumatic brain injury89. Traumatic brain injury 299

Venkatakrishna Rajajee

90. Elevated intracranial pressure 304Richard Bowers and George A. Mashour

91. Succinylcholine in the patient withincreased intracranial pressure 309Andrew Zura

Part XIII – Seizures92. Pharmacologic management of status

epilepticus 311Alaa A. Abd-Elsayed, George K.Istaphanous and Ehab Farag

93. Nonconvulsive status epilepticus 314Sheron Beltran

94. Rhabdomyolysis 317Mauricio Perilla and Jerome O’Hara

Part XIV – Neuromuscular disease95. Myasthenic crisis 321

Wael Ali Sakr Esa

96. Guillain-Barre syndrome 324Anupa Deogaonkar andEhab Farag

Part XV – End-of-life issues97. Conducting a family meeting to

decide withdrawal of care 327Marc J. Popovich

98. Brain death 329Venkatakrishna Rajajee

Index 333

ix

Contributors

Basem AbdelmalakOutcomes ResearchCleveland ClinicCleveland, OHUSA

Joseph AbdelmalakOutcomes ResearchCleveland ClinicCleveland, OHUSA

Alaa A. Abd-ElsayedOutcomes ResearchCleveland ClinicCleveland, OHUSA

David L. AdamsDepartment of AnesthesiologyUniversity of Michigan Medical SchoolAnn Arbor, MIUSA

Eric E. AdelmanDepartment of NeurologyUniversity of Michigan Medical SchoolAnn Arbor, MIUSA

Maged ArgaliousAnesthesia InstituteCleveland ClinicCleveland, OHUSA

Endrit BalaOutcomes ResearchCleveland ClinicCleveland, OHUSA

Gene H. BarnettTaussig Cancer CenterCleveland ClinicCleveland, OHUSA

Sheron BeltranDepartment of AnesthesiologyUniversity of Michigan Medical SchoolAnn Arbor, MIUSA

Andrew BielaczycDepartment of AnesthesiologyUniversity of Michigan Medical SchoolAnn Arbor, MIUSA

William BingamanEpilepsy CenterCleveland ClinicCleveland, OHUSA

James M. BlumDepartment of AnesthesiologyUniversity of Michigan Medical SchoolAnn Arbor, MIUSA

Alina BodasAnesthesia InstituteCleveland ClinicCleveland, OHUSA

Vera BorzovaAnesthesia InstituteCleveland ClinicCleveland, OHUSA

x

List of contributors

Richard BowersDepartment of AnesthesiologyUniversity of Michigan Medical SchoolAnn Arbor, MIUSA

Adam BrownDepartment of AnesthesiologyUniversity of Michigan Medical SchoolAnn Arbor, MIUSA

ChadM. BrummettDepartment of AnesthesiologyUniversity of Michigan Medical SchoolAnn Arbor, MIUSA

Alexandra S. BulloughDepartment of AnesthesiologyUniversity of Michigan Medical SchoolAnn Arbor, MIUSA

James F. BurkeDepartment of NeurologyUniversity of Michigan Medical SchoolAnn Arbor, MIUSA

Juan P. CataAnesthesia InstituteCleveland ClinicCleveland, OHUSA

Neeraj ChaudharyDepartment of RadiologyUniversity of MichiganAnn Arbor, MIUSA

Michael J. ClaybonDepartment of AnesthesiologyUniversity of Michigan Medical SchoolAnn Arbor, MIUSA

Miguel CruzAnesthesia InstituteCleveland ClinicCleveland, OHUSA

Milind DeogaonkarCenter for Neurological RestorationCleveland ClinicCleveland, OHUSA

Vikram DhawanDepartment of NeurologyUniversity Hospitals – Case Medical CenterCleveland, OHUSA

Thomas DidierDepartment of AnesthesiologyUniversity of Michigan Medical SchoolAnn Arbor, MIUSA

D. John DoyleAnesthesia InstituteCleveland ClinicCleveland, OHUSA

Zeyd EbrahimAnesthesia InstituteCleveland ClinicCleveland, OHUSA

Hesham ElsharkawyCleveland ClinicCleveland, OHUSA

Wael Ali Sakr EsaAnesthesia InstituteCleveland ClinicCleveland, OHUSA

Ehab FaragAnesthesia InstituteCleveland ClinicCleveland, OHUSA

Ryen D. FonsDepartment of AnesthesiologyUniversity of Michigan Medical SchoolAnn Arbor, MIUSA

xi


Joseph J. GemmeteDepartment of RadiologyUniversity of Michigan Medical SchoolAnn Arbor, MIUSA

Matt GilesDepartment of AnesthesiologyUniversity of Michigan Medical SchoolAnn Arbor, MIUSA

Phil GillenDepartment of AnesthesiologyUniversity of Michigan Medical SchoolAnn Arbor, MIUSA

Goodarz GolmirzaieDepartment of AnesthesiologyUniversity of Michigan Medical SchoolAnn Arbor, MIUSA

Marcos GomesCleveland ClinicCleveland, OHUSA

Lisa GrillyDepartment of AnesthesiologyUniversity of Michigan Medical SchoolAnn Arbor, MIUSA

Maged GuirguisAnesthesia InstituteCleveland ClinicCleveland, OHUSA

David W. HealyDepartment of AnesthesiologyUniversity of Michigan Medical SchoolAnn Arbor, MIUSA

Heather Hervey-JumperDepartment of AnesthesiologyUniversity of Michigan Medical SchoolAnn Arbor, MIUSA

Shawn L. Hervey-JumperDepartment of NeurosurgeryUniversity of Michigan Medical SchoolAnn Arbor, MIUSA

Paul E. HilliardDepartment of AnesthesiologyUniversity of Michigan Medical SchoolAnn Arbor, MIUSA

Samuel A. IrefinAnesthesia InstituteCleveland ClinicCleveland, OHUSA

George K. IstaphanousCleveland ClinicCleveland, OHUSA

Teresa L. JacobsDepartment of NeurosurgeryUniversity of Michigan Medical SchoolAnn Arbor, MIUSA

Ellen JankeDepartment of AnesthesiologyUniversity of Michigan Medical SchoolAnn Arbor, MIUSA

Greta JoAnesthesia InstituteCleveland ClinicCleveland, OHUSA

James W. JonesCleveland ClinicCleveland, OHUSA

Rami KarroumAnesthesia InstituteCleveland ClinicCleveland, OHUSA

xii


Allen KeeblerAnesthesia InstituteCleveland ClinicCleveland, OHUSA

Stephen J. KimatianAnesthesia InstituteCleveland ClinicCleveland, OHUSA

Colleen G. KochAnesthesia InstituteCleveland ClinicCleveland, OHUSA

Robert Scott KrissDepartment of AnesthesiologyUniversity of Michigan Medical SchoolAnn Arbor, MIUSA

Andrea KurzOutcomes ResearchCleveland ClinicCleveland, OHUSA

Jia LinAnesthesia InstituteCleveland ClinicCleveland, OHUSA

Michael D. MaileDepartment of AnesthesiologyUniversity of Michigan Medical SchoolAnn Arbor, MIUSA

Negmeldeen F. MamounAnesthesia InstituteCleveland ClinicCleveland, OHUSA

Mariel ManlapazAnesthesia InstituteCleveland ClinicCleveland, OHUSA

Edward MannoCerebrovascular Center, NICUCleveland ClinicCleveland, OHUSA

DonnMarciniakCleveland ClinicCleveland, OHUSA

Piyush MathurCleveland ClinicCleveland, OHUSA

Nicholas F. MarkoBrain Tumor and Neuro-OncologyCleveland ClinicCleveland, OHUSA

MatthewMartinDepartment of AnesthesiologyUniversity of Michigan Medical SchoolAnn Arbor, MIUSA

George A. MashourDepartment of Anesthesiology and Departmentof NeurosurgeryUniversity of Michigan Medical SchoolAnn Arbor, MIUSA

Marco MaurtuaAnesthesia InstituteCleveland ClinicCleveland, OHUSA

Scott T. McCardleDepartment of AnesthesiologyUniversity of Michigan Medical SchoolAnn Arbor, MIUSA

Julie McClellandDepartment of AnesthesiologyUniversity of Michigan Medical SchoolAnn Arbor, MIUSA

xiii


UmaMenonEpilepsy Center, Neurological InstituteCleveland ClinicCleveland, OHUSA

Paul S. MoorDepartment of AnesthesiologyUniversity of Michigan Medical SchoolAnn Arbor, MIUSA

Laurel E. MooreDepartment of AnesthesiologyUniversity of Michigan Medical SchoolAnn Arbor, MIUSA

Ruairi MouldingDepartment of AnesthesiologyUniversity of Michigan Medical SchoolAnn Arbor, MIUSA

Dileep R. NairEpilepsy Center, Neurological InstituteCleveland ClinicCleveland, OHUSA

Todd NelsonDepartment of AnesthesiologyUniversity of Michigan Medical SchoolAnn Arbor, MIUSA

Julie NiezgodaAnesthesia InstituteCleveland ClinicCleveland, OHUSA

Edward NogueraGeneral Surgical ICUCleveland ClinicCleveland, OHUSA

Jerome O’HaraAnesthesia InstituteCleveland ClinicCleveland, OHUSA

Aditya S. PandeyDepartment of NeurosurgeryUniversity of Michigan Medical SchoolAnn Arbor, MIUSA

Mauricio PerillaAnesthesia InstituteCleveland ClinicCleveland, OHUSA

Paul PictonDepartment of AnesthesiologyUniversity of Michigan Medical SchoolAnn Arbor, MIUSA

Marc J. PopovichDepartment of General AnesthesiologyCleveland ClinicCleveland, OHUSA

J. Javier ProvencioCerebrovascular Center, NICUCleveland ClinicCleveland, OHUSA

Venkatakrishna RajajeeDepartment of NeurosurgeryUniversity of Michigan Medical SchoolAnn Arbor, MIUSA

Mohit RastogiDepartment of AnesthesiologyUniversity of Michigan Medical SchoolAnn Arbor, MIUSA

xiv


Stacy RitzmanAnesthesia InstituteCleveland ClinicCleveland, OHUSA

Lauryn R. RochlenDepartment of AnesthesiologyUniversity of Michigan Medical SchoolAnn Arbor, MIUSA

Leif SaagerAnesthesia InstituteCleveland ClinicCleveland, OHUSA

Vivek SabharwalGeneral Surgical ICUCleveland ClinicCleveland, OHUSA

Oren SagherDepartment of NeurosurgeryUniversity of Michigan Medical SchoolAnn Arbor, MIUSA

Kenneth SalibaAnesthesia InstituteCleveland ClinicCleveland, OHUSA

Milad SharifpourDepartment of AnesthesiologyUniversity of Michigan Medical SchoolAnn Arbor, MIUSA

Lesli E. SkolarusDepartment of NeurologyUniversity of Michigan Medical SchoolAnn Arbor, MIUSA

Paul SmytheDepartment of AnesthesiologyUniversity of Michigan Medical SchoolAnn Arbor, MIUSA

Wolf H. StapelfeldtAnesthesia InstituteCleveland ClinicCleveland, OHUSA

William R. Stetler, Jr.Department of NeurosurgeryUniversity of Michigan Medical SchoolAnn Arbor, MIUSA

Peter StilesDepartment of AnesthesiologyUniversity of Michigan Medical SchoolAnn Arbor, MIUSA

Vijay TarnalDepartment of AnesthesiologyUniversity of Michigan Medical SchoolAnn Arbor, MIUSA

Khoi D. ThanDepartment of NeurosurgeryUniversity of Michigan Medical SchoolAnn Arbor, MIUSA

B. Gregory ThompsonDepartment of NeurosurgeryUniversity of Michigan Medical SchoolAnn Arbor, MIUSA

Alparslan TuranAnesthesia InstituteCleveland ClinicCleveland, OHUSA

xv


Christopher R. TurnerDepartment of AnesthesiologyUniversity of Michigan Medical SchoolAnn Arbor, MIUSA

Justin UppDepartment of AnesthesiologyUniversity of Michigan Medical SchoolAnn Arbor, MIUSA

Sumeet VaderaCleveland ClinicCleveland, OHUSA

Jennifer VanceDepartment of AnesthesiologyUniversity of Michigan Medical SchoolAnn Arbor, MIUSA

Anthony C. WangDepartment of NeurosurgeryUniversity of Michigan Medical SchoolAnn Arbor, MIUSA

Robert J. WeilBrain Tumor and Neuro-OncologyCleveland ClinicCleveland, OHUSA

Marnie B. WelchDepartment of AnesthesiologyUniversity of Michigan Medical SchoolAnn Arbor, MIUSA

Karen K. WilkinsDepartment of AnesthesiologyUniversity of Michigan Medical SchoolAnn Arbor, MIUSA

Erin S. WilliamsAnesthesia InstituteCleveland ClinicCleveland, OHUSA

George N. YoussefAnesthesia InstituteCleveland ClinicCleveland, OHUSA

Asma ZakariaNeurointensive Care UnitUniversity of Texas Health Science Centerat HoustonHouston, TXUSA

Sherif S. ZakyAnesthesia InstituteCleveland ClinicCleveland, OHUSA

Andrew ZuraAnesthesia InstituteCleveland ClinicCleveland, OHUSA

xvi

Preface

Neuroanesthesia and neurocritical care encompassthe perioperative management of patients sufferingfrom neurologic disease. There have been numeroustextbooks on this subject that are generally organizedaccording to basic physiology and pharmacology, neu-rosurgical procedure, postoperative neurocritical care,and the various techniques andmonitoring modalitiesparticular to the field. What has not been available isa comprehensive, case-based approach to the periop-erative care of the neurologic patient that allows foran integrated discussion of management principles.This was the motivation behind Case Studies in Neuro-anesthesia and Neurocritical Care, which is organizedaround compelling and realistic clinical scenariosfollowed by a concise discussion that is practicallyoriented but nonetheless informed by the currentliterature. Each brief but high-yield chapter highlightsbasic principles of neuroanesthesia and neurocritical

care (e.g., management of elevated intracranial pres-sure), as well as classic complications (e.g., venousair embolism or the intraoperative rupture of ananeurysm). Tables, neuroimaging and other visualaids are abundant. Case Studies facilitates clinicalpreparation, scholarly review, and board examinationstudy. Although written primarily for residents andfellows, the book will be a helpful aid for attendingphysicians, medical students, nurses and ancillarystaff.Case Studies also serves as a quick reference guidefor virtually any neuroanesthetic procedure, which isespecially helpful for junior residents or student nurseanesthetists that are rotating through neuroanesthesia.The educational philosophy is consistent with the“problem-based learning” approach implemented innumerous medical schools and we hope it will serve asa valuable complement to more traditional textbooksin neuroanesthesia and neurocritical care.

xvii

Section

INeuroanesthesia

Part I Craniotomy. Supratentorial craniotomyCase

1 Preoperative evaluationPaul Smythe and Mohit Rastogi

Supratentorial craniotomy is a common neurosurgicalprocedure, which may be emergent (e.g., evacuationof an expanding hematoma) or routine (e.g., sched-uled resection of a mass lesion). Although the corepreoperative evaluation is similar to that of the non-neurosurgical population, there are special considera-tions for patients with neurologic disease.

Case descriptionThe patient was a 61-year-old female who presentedwith new-onset seizures, which were preceded by sev-eral weeks of bilateral frontal headaches described as“dull and achy.” The headaches had become nearlycontinuous and were more recently accompanied byvertigo and nausea. The morning of admission, shehad a witnessed generalized seizure. The seizure ter-minated and she presented to the emergency roomwhere she was treated with benzodiazepines and sup-portive care; reversible physiologic and pharmacologicderangements were ruled out. Computed tomography(CT) and magnetic resonance imaging (MRI) wereperformed, which revealed a large left temporal fossamass suggestive of a meningioma. She was admittedto the neurosurgical intensive care unit, where dexam-ethasone and a loading dose of phenytoin were admin-istered. The patient was scheduled for tumor excisionvia stealth-guided craniotomy.

Prior to the headaches and seizure, the patient wasotherwise healthy except for a history of asthma. Sheused an albuterol inhaler several times each week andhad been taking ibuprofen for her headaches. On themorning of surgery, the patient was visibly anxious,but was alert and oriented to person, place, and time.Her cranial nerve examination was normal, pupilswere equal and reactive to light, she demonstrated nor-mal strength bilaterally and showed no pronator drift.Heart sounds were normal, but wheezing was appre-ciated on auscultation. Her blood pressure during theexamination was 175/92, considerably elevated from

her usual baseline. Electrocardiogram and laboratoryvalues were normal, including a therapeutic phenytoinlevel.

The patient’s anxiety, increased blood pressure, andbronchospasm were addressed immediately.The anes-thesiologist had a reassuring conversation with thepatient and midazolam was administered, followed byan albuterol breathing treatment. Oxygen was deliv-ered through a nasal cannula and she was monitoredcontinuously with a blood pressure cuff, pulse oxime-ter, and by direct visualization to assure she did notexperience somnolence or respiratory depression dueto the benzodiazepine.

Anesthetic concerns for this patient included:(1) vigilance for any signs or symptoms of increasedintracranial pressure (ICP), (2) keeping the patient ascomfortable as possible during an emotionally difficultpreoperative period, (3) history of asthma, (4) recenthistory of seizures and consequent phenytoin use,(5) maintaining proper fluid status during a procedurethat could include fluid shifts both planned (diuret-ics) and unplanned (bleeding), and (6), the needfor a timely, hemodynamically stable, and thoroughawakening.

Once in the operating room induction of anes-thesia was initiated with fentanyl, sodium pentothaland vecuronium; great care was taken to maintain thepatient’s blood pressure within 20% of baseline. Thetrachea was intubated with a 7.5 mm LITA (Laryn-gotracheal Instillation of Topical Anesthesia) endotra-cheal tube. An arterial catheter and two large-boreintravenous catheters were placed. General anesthe-sia was maintained with isoflurane and a total ofapproximately 10 mcg/kg of fentanyl was adminis-tered throughout the case for analgesia. The surgeonsreported swelling of the brain and so the patient washyperventilated to a PaCO2 of 30 mmHg. Mannitolwas administered to reduce brain volume and dex-amethasone was given for the prevention of cerebraledema.There was an estimated blood loss of 1300 mL,

3

I. Craniotomy. Supratentorial craniotomy

and the patient was resuscitated with 2 liters of crystal-loid, 500 mL of 5% albumin and one unit packed redblood cells. As surgical closure began, isoflurane wasdiscontinued and infusions of propofol and remifen-tanil were initiated. After the surgical drapes wereremoved, the LITA tube was dosed with 4% lido-caine and albuterol was delivered via the endotrachealtube. The patient emerged within several minutes ofthe head dressing being applied, without bucking orcoughing. She was neurologically intact.

DiscussionAswith other complex cases, the patient presenting forsupratentorial craniotomy is best served by an orga-nized and systematic approach. Below is one strategyfor the preoperative evaluation of these neurosurgicalpatients.

1. Stratify the caseIt is first important to distinguish between the “elec-tive” or scheduled resection of a supratentorial masslesion and the patient presenting for emergent decom-pression due to intracranial hypertension. The formersituation allows time for a full evaluation process [1]while the latter demands immediate attention to the“ABCs” of airway, breathing, and circulation, as well asmanagement of increased intracranial pressure (ICP;discussed in the next chapter) [2].

2. Assess the patientEven during the introductory process, the anesthesi-ologist should be assessing the patient’s mental statusand gross neurologic function. Are they alert or som-nolent? Do they have any obvious facial asymmetry?Do they respond appropriately and with intact lan-guage skills? If the patient is interactive, obtain any his-tory possible. If not, assess for ABCs.

Next, assess the vital signs – significant hyper-tension and bradycardia could be a Cushing’s reflexrevealing poor intracranial elastance. The physicalexamination that follows should focus on neurologicstatus, including a cranial nerve examination, assess-ment of the pupils and their reflexes, and motorstrength. It is imperative to establish a preoperativeneurologic baseline to which the postoperative statuscan be compared.

3. Review the chartFor a non-emergent case, there is often valuableinformation that can be derived from the surgicalevaluation and that can help guide the preoperativeanesthetic assessment. Any available neuroimagingstudies should also be reviewed, with particular atten-tion to the size and location of the lesion as well assigns of increased ICP such as midline shift, efface-ment of sulci and gyri, loss of gray–white differentia-tion, and herniation. Laboratory values should also bereviewedwith particular attention to hematocrit, coag-ulation studies (especially in patients with hemorrhageor hematoma), sodium irregularities (common in neu-rosurgical patients), and glucose (which, when high,can exacerbate neural injury) [3]. Cardiac studies suchas electrocardiogram or echocardiogram can provideimportant information, especially if the patient hassuffered a stroke [3]. Review allmedications, especiallyany anticonvulsants that could affect drugmetabolism.

4. Develop a planBased on (1) your stratification of the case, and (2)your assessment of the patient, develop an anestheticplan that addresses any medical issues that might bemanifested in the perioperative course of a craniotomy.

Preparation: In general, an arterial catheter andlarge-bore intravenous lines are recommended forthese cases, but especially in those involving vascularlesions and tumorswith high potential for hemorrhage(such as meningiomas). Invasive blood pressure mon-itoring allows beat-to-beat readings and the ability toregularly assess pH status and PaCO2 levels with arte-rial blood gases. Depending on the co-morbidities andvenous access, a central line may be required.The sub-clavian approach ismore appropriate for the neurosur-gical patient, as an internal jugular catheter can poten-tially compromise venous return from the brain andthus increase ICP.

Ensure that commonly used agents such as manni-tol, dexamethasone, and furosemide are available. Dis-cuss anticonvulsant administration with the neurosur-gical team. Phenytoin is often administered, but fos-phenytoin is preferable because of decreased cardio-vascular side effects. In either case, the drug should beadministered slowly, as the goal is prophylaxis ratherthan acute termination of seizures. An anti-emeticstrategy should also be established [4].

Premedication: Neurologic disease and neurosur-gical intervention is a highly stressful event for

4

Case 1. Preoperative evaluation

the patient and the family. The need for compas-sion and reassurance cannot be overemphasized.Pharmacologic support can be attained using anynumber of drugs – it must be noted, however, that anysedative can cause respiratory depression. The patientmust be monitored continuously, since a decrease inventilation can cause an increase in CO2 and there-fore an increase in ICP via cerebral vasodilation. Seda-tion may also unmask or exacerbate focal neurologicdeficits [5].

Anesthetic regimen: The choice of anestheticsdepends on the state of intracranial elastance. Induc-tion with propofol or sodium pentothal is appropri-ate, with attention to cardiovascular status. Ketamineis typically avoided, as it increases cerebral metabolicrate. In terms of maintenance, inhalational anesthet-ics may be used if the patient has normal intracranialelastance [6]. If the patient has compromised elastanceand intracranial hypertension, then these anestheticsshould be avoided. Although inhalational anesthet-ics reduce cerebral metabolic rate, they dilate cerebralblood vessels (nitrous � desflurane � isoflurane �sevoflurane), which increases cerebral blood volumeand thus ICP. Inhalational agents can also disrupt cere-bral autoregulation, which normally maintains con-stant cerebral blood flow (50 mL/100 g tissue/min) inthe face of changing mean arterial pressures. The clas-sically taught range of autoregulation is amean arterialpressure of 50–150 mmHg, but the lower limit may beas high as 70 mmHg. Chronic hypertension can shiftthe autoregulatory curve to the right, which meansthat higher pressures are needed to stay in the autoreg-ulatory range. The intravenous agents propofol andpentothal decrease cerebral metabolic rate but do notdilate cerebral blood vessels – their use is therefore rec-ommended in patientswith poor intracranial elastanceand they also have minimal effects on autoregulation.

Analgesia: Appropriate pain control is imperativefor patient comfort and a smooth emergence [7]. Pre-operative scalp block may help reduce intraopera-tive opioid requirements and postoperative pain (seeCase 5). Major stimulating events include intubation,cranial pinning, incision, removal of bone flap, anddurotomy. An early focus on analgesia during the caseis important such that opioid-associated sedation andhypoventilation are not a problem during emergence.

Emergence: All physiologic parameters should beoptimized prior to emergence. There are two mainstrategies for achieving an emergence that occursdirectly after undraping and does so without the

patient undergoing hemodynamic responses thatmight increase the chance of intracranial hemorrhage.The first is by using, or least switching to, short-actinganesthetics and analgesics. For example, isofluranecan be turned off 1–1.5 hours prior to wake-up andreplaced with remifentanil, propofol, or both. Withthis strategy, less remifentanil or propofol is requiredto achieve general anesthesia since the offset of isoflu-rane anesthesia is relatively slow and the stimulus ofclosing is minor. Remifentanil, like most narcotics,tends to soothe the airway and is especially helpful.Another method of facilitating a smooth wake-up is todecrease the stimulus of the endotracheal tube, whichcan be accomplished with the use of a LITA tube. TheLITA allows the easy administration of lidocaine bothabove and below the cuff, anesthetizing the airway andthereby minimizing the irritation during wake-up. Ifa LITA is not available, one can place lidocaine downthe endotracheal tube or into the posterior pharynx.This strategy may be especially important in patientswith a reactive airway, as the patient described above.Careful planning to minimize physiologic and phar-macologic confounds in the immediate postoperativeperiod is essential for an effective neurosurgical evalu-ation [8].

ConclusionIn conclusion, supratentorial craniotomy is a commoncase in neurosurgery during which both the neuro-surgeon and anesthesiologist are modulating the sameorgan. Thoughtful planning and clear communicationbetween the teams is required for optimal patient care.

References1. A. Schiavi, A. Papangelou,M.Mirski. Preoperative

preparation of the surgical patient with neurologicdisease. Anesthesiol Clin 2009; 27: 779–86.

2. L. Rangel-Castillo, S. Gopinath, C. S. Robertson.Management of intracranial hypertension. Neurol Clin2008; 26: 521–41.

3. K. Lieb,M. Selim. Preoperative evaluation of patientswith neurological disease. Semin Neurol 2008; 28:603–10.

4. L. H. Eberhart, A. M. Morin, P. Kranke et al.Prevention and control of postoperative nausea andvomiting in post-craniotomy patients. Best Pract ResClin Anaesthesiol 2007; 21: 575–93.

5. G. D.Thal,M. D. Szabo,M. Lopez-Bresnahan et al.Exacerbation or unmasking of focal neurologic deficitsby sedatives. Anesthesiology 1996; 85: 21–5.

5


6. K. Engelhard, C. Werner. Inhalational or intravenousanesthetics for craniotomies? Pro inhalational. CurrOpin Anaesthesiol 2006; 19: 504–8.

7. A. Gottschalk, L. C. Berkow, R. D. Stevens et al.Prospective evaluation of pain and analgesic use

following major elective intracranial surgery.J Neurosurg 2007; 106: 210–16.

8. N. Fabregas, N. Bruder. Recovery and neurologicalevaluation. Best Pract Res Clin Anaesthesiol 2007; 21:431–47.

6


2 Evaluation and anesthetic management ofelevated intracranial pressureAndrew Bielaczyc and Paul Smythe

Understanding the appropriate interventions to man-age increased intracranial pressure (ICP) is an essentialskill of a neuroanesthesiologist. Intracranial pressureis of paramount importance because the cranial vaultis nondistensible and within it are contained threenoncompressible substances: brain, blood, and cere-brospinal fluid (CSF).

Case descriptionThe patient was a 75-year-old male who presentedfor emergent subdural hematoma evacuation. Duringa witnessed slip and fall accident with no associatedloss of consciousness, the patient suffered a left tem-poral contusion. Shortly after the accident he becameconfused and lethargic. Noncontrast head computedtomography (CT) revealed an acute right temporalsubdural hematoma with 5 mm of horizontal shift.Past medical history was significant for hypertensionand atrial fibrillation; current medications includedcoumadin, metoprolol, and lisinopril. Neurologicallythe patient was oriented to person only and moved allfour extremities equally with full strength. No cranialnerve deficits were appreciated. The patient’s bloodpressure upon presentation was 155/85.

The immediate anesthetic goal was to minimizethe rise in ICP while at the same time maintain-ing adequate cerebral perfusion pressure until theneurosurgeons could provide definitive treatment. Inorder to better monitor arterial pressure, a radialartery catheter was placed. Laboratory tests includ-ing type and screen, coagulation values, and completeblood count were drawn. After placement of stan-dardmonitoring and preoxygenation, general anesthe-sia was induced with propofol and care was taken toprevent precipitous increases in blood pressure dur-ing both laryngoscopy and placement of the head ina Mayfield frame. The patient was then positionedsupine with the head of the bed elevated 30 degrees.A midline head alignment and unobstructed cer-

vical venous drainage were ensured. Based on thepatient’s baseline blood pressure and estimated ICP,a mean arterial pressure (MAP) �80 mmHg wasmaintained in order to maintain cerebral perfusionpressure (CPP) �60 mmHg. Prior to the dura beingopened, mannitol and furosemide were administeredto decrease parenchymal volume. Definitive correc-tion of intracranial hypertension was achieved withhematoma evacuation.

DiscussionThe cranium is a nonexpandable bony structure witha fixed volume and therefore with only limited meansof compensating for increased ICP. If pressure withinthat nonexpandable structure is allowed to rise theresult is life-threatening neural injury. Normal ICP inthe adult is �20 mmHg with a high degree of normalphysiologic variability. A sustained ICP �20 mmHgshould be considered pathologic and treatment shouldbe initiated. Intracranial pressure along with MAPdetermine cerebral perfusion based on the equa-tion CPP = MAP – ICP. Therefore, when managingpatients with elevated ICP, adequate MAP must beachieved in order tomaintainCPP�60mmHg. Belowthis threshold cerebral ischemia can occur.

The determinants of ICP are the contents of thecranium and include brain parenchyma (neurons,glia), blood (arterial and venous circulation), and fluid(interstitial and CSF). Increases in one of these vol-umes must be offset by decreases in the other con-stituent volumes or ICP will increase [1]. The mag-nitude of change in ICP is directly related to theintracranial elastance. As the volume of blood, fluid, orbrain expands, physiologic mechanisms of compensa-tion such as displacement of CSF out of the cranium orvasoconstriction result in relatively small increases inICP. As intracranial elastance decreases, these mech-anisms are eventually outstripped and further smallincreases in volume result in large increases in ICP

7


Pressure

elastancereducedelastance

Volume

poor elastance

Figure 2.1. Intracranial elastance curve.

(Figure 2.1) [1]. A direct relationship exists betweencerebral blood flow (CBF) and ICP, such that elevatedICP can be modified by decreases in CBF so long ascerebral perfusion is not compromised. If insufficientCBF results in brain ischemia, the associated cerebraledema will increase ICP, further compromising CBFand setting in motion a dangerous downward spiral.

Increased ICP is associated with many pathophys-iologic states, each with the common mechanism ofincreasing the intracranial volume of CSF, blood, orbrain tissue (Table 2.1). Increased tissue volume iscaused by space-occupying lesions such as tumors,either primary or metastatic. Cerebral malignanciesalso increase ICP by obstructing absorption and out-flow of CSF and producing cerebral edema. Cerebraledema is classified as cytogenic or vasogenic in ori-gin. Cytogenic cerebral edema develops as the resultof inadequate intracellular energy supplies resultingin failure of membrane integrity, increased cellularswelling and ultimately cell death. Vasogenic cerebraledema by comparison is the result of increased vas-cular permeability and the movement of protein-richfluid into the cerebral interstitial space.

Increased intracranial blood volume elevates ICPand can present in varied forms. Subarachnoid hem-orrhage is frequently caused by head trauma, disrup-tion of aneurysms, and arteriovenous malformations.Blood in the subarachnoid space or in the ventriclesimpedes CSF outflow and results in ICP elevation.Intracranial hematomas, blood-filled space-occupyinglesions, are classified by their location relative to themeningeal layers. Subdural hematomas are more fre-quently seen in elderly patients due to natural cerebralatrophy and the resulting fragility of cerebral bridgingveins. Acute epidural hematomas are frequently the

Table 2.1. Common causes of elevated intracranial pressure.

Substrate Pathophysiology

Parenchyma MalignancyPrimaryMetastatic

Cerebral edemaCytogenicVasogenic

EclampsiaFulminant hepatic failure

Blood Subarachnoid hemorrhageIntraparenchymal hemorrhageSubdural hematomaEpidural hematomaVasodilatationVenous outflow obstruction

Cerebrospinal fluid HydrocephalusCSF outflow obstruction

result of severe head trauma and produce a character-istic lenticular-shaped hemorrhage on CT scan.

Mental status can range from drowsiness to comadepending on ICP and the extent of brain tissue dis-placement. As ICP increases, signs and symptomsinclude headache, nausea, vomiting, and papilledema.Bradycardia, hypertension with widened pulse pres-sure, and Cheyne–Stokes breathing are referred to asCushing’s Triad, an ominous sign of decompensation.When intracranial elastance is low, further increases inICP can result in brain tissue displacement and herni-ation of brain parenchyma across meningeal barriers,which heralds serious brain injury or death.

Intracranial pressure can be monitored in sev-eral ways. As described above, the clinical examina-tion can indicate relative changes in ICP and there-fore impact management. In addition, neuroimagingcan be used to estimate intracranial elastance basedon the degree of sulcal effacement and ventricular col-lapse. Invasive methods of ICP monitoring are rou-tine; correct interpretation and management is criti-cal. Ventricular catheters, parenchymal monitors, andsubarachnoid bolts can all be placed at the bedsideby trained neurosurgeons and neurointensivists. Ven-tricular catheters, which are considered the gold stan-dard for ICP monitoring, can be used to drain CSFand therefore modulate ICP [1, 2]. Transduction ofthe ICP reveals a waveform composed of a larger P1,and smaller P2 deflections. This relationship becomesinverted as intracranial compliance decreases. The“A” wave or plateau wave is a sign of intracranialhypertension.

8

Case 2. Evaluation and anesthetic management of elevated intracranial pressure

Numerous therapeutic maneuvers exist for low-ering ICP, each with the common mechanism ofdecreasing the volume of one or more intracranialcomponents. It cannot be overemphasized that alltreatments of ICP carry risks and so each must bemonitored closely. A rapid method of decreasingICP is to raise the head as much as possible; thisis not always possible, especially during surgery.Decreasing the size of a brain mass is not a job forthe anesthesiologist; that can only be accomplishedsurgically or, if the mass is inoperable, perhaps byradiation or chemotherapy. Decreasing the amountof CSF can be accomplished using a ventriculostomy.Interstitial edema can be treated with corticosteroidsbut this is not an effective acute treatment. Reductionof cytogenic cerebral edema can be achieved withthe administration of osmotic or loop diuretics.Hyperosmotic agents like mannitol increase plasmaosmolarity and draw water out of tissues includingacross an intact blood–brain barrier. Mannitol isinfused slowly at a dose of 0.25–1 g/kg and may tran-siently increase ICP due to increased cerebral bloodvolume. Maximal reduction in ICP is evident after10–15 minutes and remains effective for two hours.Hypertonic saline, commonly used in 3% and 7.5%solution, has proved effective in decreasing ICP. How-ever, no evidence of improved neurologic outcomesor survival has been shown [3]. The loop diureticfurosemide is used to decrease ICP by systemicdiuresis and decreasing CSF production. Furosemidedoes not increase intravascular volume and there-fore is a better choice in patients with impairedleft ventricular function. Intravascular volume andelectrolyte balance should be monitored closelyduring the administration of furosemide and diuretictherapy. Hyperventilation reduces CBF 1–2mL/100 g/min for each 1 mmHg reduction in PaCO2 andtherefore reduces ICP. However, reducing end-tidalCO2 (ETCO2) below 30 mmHg is not recommendeddue to compromise of CBF and the development ofcerebral ischemia. As discussed above, MAP, CBF,and ICP are intertwined and therefore stable hemo-dynamics are critical to preventing further increasesin ICP. Ultimately the pathologic process underlyingintracranial hypertension must be corrected.

Anesthetic management for patients with elevatedICP begins in the preoperative setting. Preopera-tive anxiolysis should consist of a calm environmentfree of unnecessary distractions and, when indicated,medication. However, administration of preopera-

tive sedatives may cloud the neurologic examinationand should be considered carefully in patients withdecreased intracranial elastance as modest hypercar-bia resulting from hypoventilation may dangerouslyincrease ICP via cerebral vasodilation. Arterial andcentral venous catheters may need to be placed afterinduction of anesthesia to avoid additional patientstress.

The operating room should be a calm environmentbefore induction, free from music, loud conversation,or undue distractions. After application of standardmonitoring and thorough preoxygenation, anesthesiacan be induced with a combination of agents. Suc-cinylcholine can be used if a rapid sequence induc-tion is required but the resulting defasciculations canproduce a transient increase in ICP. Adequate depthof anesthesia should be achieved prior to trachealintubation and a bolus of sedative hypnotic or lido-caine can be administered before laryngoscopy. Afterplacement of the endotracheal tube and confirmationof ETCO2, the patient can be modestly hyperventi-lated. The arterial/end-tidal CO2 gradient should beinvestigated and an ETCO2 of 30–35 mmHg targeted.The use of positive end-expiratory pressure to main-tain oxygenation must be applied carefully. Increasedintrathoracic pressures, transmitted through compli-ant lungs, can impede cerebral venous drainage andfurther increase ICP.

In order to assure a stable operative field, thepatient’s head may be placed in Mayfield pins. Priorto the application of pins, a bolus of sedative and ornarcotic should be administered to blunt the responseto this profound stimulation. Positioning for poten-tially lengthy neurosurgical procedures is of criticalimportance and should be carried out with the cooper-ation of members of the operative team. Patients withelevated ICP should be maintained in a neutral headposition with slight reverse Trendelenburg bed tilt toensure cerebral venous drainage [4].

Maintenance of anesthesia can be achieved witha variety of agents tailored to the patient’s medicalhistory and the presence of neurologic monitoring.All combinations should have the common traits ofadequate depth, stable hemodynamics and loweredICP. Volatile anesthetics are potent cerebral vasodila-tors and progressively abolish the cerebral autoregu-latory curve at higher doses. However, volatile agentsdo decrease cerebral metabolic rate of oxygen andat modest doses do not produce significant eleva-tions of ICP. By contrast, propofol and thiopental

9


both reduce cerebral metabolism and decrease cere-bral blood volume by vasoconstriction [5]. This com-bination produces a reduction in ICP. The use ofneuromuscular blockers may also facilitate the man-agement of ICP by preventing patient bucking orstraining.

ConclusionIn conclusion, increased ICP can be a life-threateningcondition, the definitive treatment of which is often inthe hands of the neurosurgeon. It is the job of the anes-thesiologist to effectively manage increased ICP untilsuch treatment can be accomplished.

References1. L. A. Steiner, P. J. Andrews. Monitoring the injured

brain: ICP and CBF. Br J Anaesth 2006; 97: 26–38.2. M. Smith. Monitoring intracranial pressure in

traumatic brain injury. Anesth Analg 2008; 106: 240–8.3. G. F. Strandvik. Hypertonic saline in critical care: a

review of the literature and guidelines for use inhypotensive states and raised intracranial pressure.Anaesthesia 2009; 64: 990–1003.

4. I. Ng, J. Lim,H. B. Wong. Effects of head posture oncerebral hemodynamics: its influences on intracranialpressure, cerebral perfusion pressure, and cerebraloxygenation. Neurosurgery 2004; 54: 593–7.

5. L. T. Dunn. Raised intracranial pressure. J NeurolNeurosurg Psychiatry 2002, 73: i23–7.

10


3 Traumatic brain injuryAdam Brown and Paul S. Moor

Traumatic brain injury (TBI) is a leading cause ofdeath and permanent disability in the western worldand imparts a significant social burden. According torecent US statistics, of the reported annual TBI cases,approximately 50 000 patients die, 80 000 have perma-nent disability, and 235 000 are admitted to hospital[1]. With respect to mortality, 50% of patients who diefrom TBI do so within 2 hours of injury, and approx-imately 30% of patients admitted to a hospital with aGlasgow Coma Scale (GCS, Table 3.1) �13 will ulti-mately die.

The injury spectrum in TBI encompasses not onlythe initial insult, but includes the cascade of sys-temic responses and pathophysiology that occur afterthe focal or diffuse brain injury. Anesthesiologists areoften involved throughout the care of TBI patients. It isthrough our effective physiologic manipulation of andpharmacologic intervention in elevated intracranialpressure (ICP), decreased cerebral perfusion pressure(CPP), secondary ischemic brain injury, andmetabolicderangements, that we can positively influence patientoutcome.

Case descriptionThe patient was a 49-year-old male motorcyclist whowas evaluated in the emergency room after a motorvehicle collisionwith an articulated truck (“Semi”). Hewas found prone 100 feet from the expressway withbilateral upper limb deformities, coarse respirations,and a GCS of 3. Endotracheal intubation was per-formed at the scene and transportation to our facilitywas complicated by five episodes of ventricular tachy-cardia, each requiring direct current cardioversion tosinus rhythm.

On arrival at the emergency room, primary sur-vey revealed a normothermic intubated male witha patent size 7.0 mm endotracheal tube and cervi-cal collar in place; breath sounds and chest excur-sion were equal bilaterally. Pulse oximetry revealed

saturations of 100% on 15 liters oxygen, via non-rebreather mask. Five-lead EKG monitoring showedsinus tachycardia of 110 beats per minute withoutectopy. Neurologically the patient was moving all fourlimbs spontaneously and purposefully, prior to anes-thesia for computed tomography (CT) scan. Stan-dard monitoring was applied and large bore intra-venous access was obtained in two limbs; radial arterycannulation afforded invasive arterial blood pressuremonitoring, as well as assessment of adequate gasexchange, balanced electrolytes and blood glucose.Thepatient was stabilized, sedated, and transferred forCT scanning.

Ongoing monitoring of pupillary size and reactiv-ity was also maintained prior to direct ICP measure-ment. In the paralyzed and sedated patient, such clin-ical signs were indicative of tentorial herniation.

Noncontrast head CT revealed a right-sided epidu-ral hematoma (Figure 3.1). Preliminary evaluation ofthe CT revealed no fractures or malalignments ofthe cervical spine. The preliminary review of the CTof the chest, abdomen, and pelvis revealed evidenceof numerous, but stable, pelvic fractures and a righthumeral fracture.

Around the time of CT scanning, the patient’s rightpupil became dilated and unreactive (6 mm unreac-tive right vs 3 mm reactive left). He was subsequentlytransported to the operating room for immediate cran-iotomy, evacuation of the extradural hematoma, andventriculostomy insertion. The primary concern ofthe anesthesia team was poor intracranial elastance,increased ICP and potentially inadequate cerebral per-fusion in the face of hemorrhagic hypovolemia. Theteam was also focused on the prevention of secondarybrain injury [2, 3].This required vigilance for and pre-vention of hypotension, hypoxia, hypercarbia, hyper-thermia, and hyperglycemia, as these derangementscontribute to enhanced parenchymal injury. A furtherconcern, heightened by the mechanism of injury andpolytrauma, was the suspicion of an accompanying

11


Table 3.1. Glasgow Coma Score.

Score Infant<1 year Child 1–4 years Age 4–adult

Eyes4 Open spontaneously Open spontaneously Open spontaneously3 To voice To voice To voice2 To pain To pain To pain1 No response No response No response

Verbal5 “Coos, babbles” “Oriented, speaks, interacts, social” Oriented and alert4 “Irritable cry, consolable” “Confused speech, disoriented, consolable” Disoriented3 Cries persistently to pain “Inappropriate words, inconsolable” Nonsensical speech2 Moans to pain “Incomprehensible, agitated” “Moans, unintelligible”1 No response No response No response

Motor6 “Normal, spontaneous movement” “Normal, spontaneous movement” Follows commands5 Withdraws to touch Localizes pain Localizes pain4 Withdraws to pain Withdraws to pain Withdraws to pain3 Decorticate flexion Decorticate flexion Decorticate flexion2 Decerebrate extension Decerebrate extension Decerebrate extension1 No response No response No response

Figure 3.1. Head computed tomography demonstrating rightepidural hematoma.

cervical spinal cord injury without radiologic abnor-mality and possibly unstable neck.

On arrival in the operating room, it was notedthat despite adequate oxygen saturation by SpO2, theendotracheal tube had migrated down the right mainbronchus; this was corrected swiftly. Mannitol 1 g/kgwas administered to acutely treat his increased ICP andtentorial herniation.Despite an apparently normal cer-vical CT scan, an occult injury could not be excluded

in our patient due to the mechanism of injury and thepresence of an injury above the clavicle. As a result cer-vical immobilization continued throughout his courseof care. Central venous access was obtained and afterfinal confirmation of intended surgery, surgical anes-thesia was carefully induced with propofol and fen-tanyl. Anesthesia wasmaintainedwith a propofol infu-sion, as well as vecuronium for paralysis. The goal forCPP was �60 mmHg. Initial arterial blood gas anal-ysis demonstrated a metabolic acidemia and hyper-glycemia (186 mg/dL). Compensatory efforts werethus made by increasing minute ventilation and aninsulin infusion was started with a target blood glu-cose level of 120 mg/dL. Patient temperature was ini-tially 34.2 ◦C rising through the procedure to 36.3 ◦Cwith forced air warming. On subsequent lavage of theepidural hematoma, bleeding from an epidural arterybecame evident and two units of packed red bloodcells were transfused in response to a hematocrit of22%.The remainder of the surgery was uneventful andthe patient was transferred to a neurosurgical intensivecare unit, still intubated and sedated.

DiscussionTraumatic brain injuries can be diffuse or focal, andresult from multiple traumatic mechanisms. Diffusebrain injury includes diffuse axonal injury, hypoxicbrain injury, brain swelling, and brain hemorrhages.Focal brain injury includes contusions, avulsions,

12

Case 3. Traumatic brain injury

hematomas, hemorrhages, infarctions, and infections.All of the above can lead to brain edema and all canincrease ICP.

The severity of the initial injury, as defined by theGCS, is predictive of eventual outcome. For example,an admission GCS of 9–15 carries a low risk of devel-oping intracranial hypertension. However, the rela-tionship between GCS and survival is nonlinear, witha high rate of mortality for GCS 3–7, and a decline inmortality between GCS 8–15 [4].

Both military and civilian data support the corre-lation between initial (field) GCS and eventual out-come, despite them seeming disparate patient pop-ulations and mechanisms of injury. Mortality ofAmerican military service members with severe headtrauma is 65% for GCS from 3–5, and 10% for GCSfrom 6–8. Of the survivors, progression to an indepen-dent living status is �10% for GCS from 3–5, and 60%for GCS from 6–8.

On completion of the ATLS primary survey theclinician’s focus should remain on the provision ofhomeostasis for salvageable brain tissue. Such actionsminimize secondary hypoxic brain damage and areessential in all phases of the patient’s care. For ourpatient this began in the prehospital arena, with endo-tracheal intubation at the scene, rapid establishmentof intravenous access, and fluid loading to resuscitatethe hypovolemic polytrauma patient. Efforts shouldbe directed at the treatment of increased ICP and themaintenance of adequate CPP (�60 mmHg), to driveperfusion to salvageable brain parenchyma and pre-vent secondary brain injury. Traumatic brain injuryis often associated with polytrauma, which makesmaintaining an adequate CPP difficult, especially inthe face of concomitant traumatic hypovolemia andhemorrhage.

Efforts to re-establish and maintain normotensionshould be aggressive in head injury. It is well doc-umented that just a single episode of systolic bloodpressure �90 mmHg has a direct negative effect onpatient outcome after TBI. Hypoxic cerebral damage,a common postmortem finding in TBI, is associatedwith arterial hypoxemia, decreasedmean arterial pres-sure, or cerebral hypoperfusion, occurring as a con-sequence of shock, intracranial hypertension, or cere-bral vasospasm. The vulnerability of TBI patients tocerebral ischemia (defined as inadequate cerebral oxy-gen delivery), hypotension, and reduced CPP mostlikely occurs due to many factors, including regionsof brain tissue with precariously low cerebral blood

flow and defective autoregulatory ability and under-scores the need formaintenance of CPP. Such concernsover CPP maintenance should influence the choice ofinduction agent, a subject that has long been debated.The requirement for a cardiostable induction is clear;current opinion suggests the consideration of agentslike ketamine, that have long been discouraged as aresult of a transient rise in ICP. It is argued that theeffects on ICP may well be attenuated by the sub-sequent cardiostability and reversal of hypoxia andhypercarbia [5].

The maintenance of oxygenation (maintain SpO2�90%), normocarbia or mild hypocarbia, normoth-ermia (maintain temperature of 37 ◦C) and normo-glycemia are vital actions. An increase in body (andthus brain) temperature is associated with an increasein metabolism, blood flow, and oxygen utilization,which can exacerbate potential brain ischemia. Cur-rent evidence for the role of hypothermia in TBI isinconclusive.Overall if hypothermia is initiatedwithin60 minutes of injury and maintained for 48 hours, itappears to confer some benefit on neurologic outcomein severe TBI patients.

As the brain is an obligate glucose user, hyper-glycemia is associated with an increase in cerebralmetabolism. Due to the decreased CBF subsequentto polytrauma and shock, hyperglycemia can resultin increased anaerobic metabolism, changes in pH,and worse outcome [3]. As a result, glycemic con-trol with insulin is often indicated. Additionally, all ofthe above pathologic processes interact synergistically,which ultimately can produce a greater deficit thancould be attributed to the primary central nervous sys-tem injury alone.

In our patient, the epidural hematoma repre-sented a neurosurgical emergency for hemorrhagecontrol. Unchecked hematoma expansion led to ele-vated ICP and clinical signs such as an ipsilateraldilated pupil (due to herniation with dysfunction ofcranial nerve III). These events could have led tofurther brain herniation and death in our patientwithout immediate pharmacologic intervention ordecompression.

Given the mechanisms of injury that result inTBI, it is not surprising that spinal cord injuriesoften accompany TBI. With GCS�8, a high index ofsuspicion for concomitant cervical spine injury mustbe maintained. Appropriate spinal stabilization tech-niques should be taken to help improve the chances ofneurologic recovery.

13


ConclusionThemanagement of TBI patients requires swift physi-ologic and/or surgical manipulation in order to reducethe impact of primary injury on the remainder of thebrain. Regardless of the apparent prognosis, resusci-tation should proceed aggressively, keeping in mindthe vulnerability of the acutely injured CNS to sys-temic insults. Definitive airway control, while main-taining a neutral cervical spine (due to the need forsuspected concomitant spinal cord injury) should beinstituted immediately. Hemodynamic resuscitationmust be accomplished promptly, to maintain nor-mal perfusion and normal CPP and because cerebralischemia or intracranial hypertension may result frominadequately treated shock. Although a definitivelyeffective neuroprotective therapy in CNS traumaremains elusive, through the skilled interactions ofprehospital, emergency department, anesthesiology,and surgical personnel, the lives of many criticallyinjured individuals can be saved and their neurologicfunction preserved.

References1. N. Badjatia, N. Carney, T. J. Crocco et al. Guidelines

for prehospital management of traumatic braininjury. 2nd edition. Prehosp Emerg Care 2008; 12:S1–52.

2. I. K. Moppett. Traumatic brain injury: assessment,resuscitation and early management. Br J Anaesth2007; 99: 18–31.

3. Brain Trauma Foundation. American Association ofNeurological Surgeons; Congress of NeurologicalSurgeons; Joint Section on Neurotrauma and CriticalCare, AANS/CNS. Guidelines for the management ofsevere traumatic brain injury. J Neurotrauma 2007; 24Suppl. 1: S1–95.

4. P. Udekwu, S. Kromhout-Schiro, S. Vaslef et al.Glasgow Coma Scale score, mortality, and functionaloutcome in head-injured patients. J Trauma 2004; 56:1084–9.

5. C. Morris, A. Perris, J. Klein et al. Anaesthesia inhaemodynamically compromised emergency patients:does ketamine represent the best choice of inductionagent? Anaesthesia 2009; 64: 532–9.

14


4 Postoperative seizureAllen Keebler

Postoperative seizures are a relatively common occur-rence after surgery. They are very common in theintensive care unit, even in patients without a primaryneurologic diagnosis.The incidence has been reportedto be as high as 12% in this setting. In the post-craniotomy patient the causes are more likely relatedto the procedure itself, while in the general critical-carepopulationwithout a primary neurologic diagnosis themost common cause is a metabolic disturbance [1].

Case descriptionA 56-year-old female presented for resection of a 2 cmby 2.5 cm mass in the right temporal lobe. She wasbrought to the emergency room after her first seizurethree days prior, when the mass was discovered andshe was scheduled for a craniotomy. She was started onphenytoin for seizure prophylaxis.Magnetic resonanceimaging demonstrated the lesion in question and amild decrease in ventricular size with no evidence ofmidline shift.Her pastmedical history included hyper-tension, which was well controlled with an angiotensinreceptor blocker. Past surgical history showed a hys-terectomy for dysfunctional uterine bleeding at age 46that was uneventful. Laboratory testing showed only amild anemia.

The patient was brought to the operating roomfor a scheduled resection of the tumor under generalendotracheal anesthesia. After an uneventful induc-tion with propofol and endotracheal intubation facil-itated with rocuronium, anesthesia was maintainedwith isoflurane and fentanyl boluses. An arterialcatheter and two large-bore intravenous catheters wereinserted. The initial dissection was unremarkable, butduring resection of the tumor the surgical team haddifficulty with hemostasis. The acute blood loss wasestimated at about 1.5 liters for which 2 units of packedred blood cells and 500 mL albumin 5% were infusedfor a final hematocrit of 28%. After an uneventful clo-sure neuromuscular blockade was reversed and emer-

gence was begun with 100 mg lidocaine administeredthrough the endotracheal tube. Despite a gentle emer-gence the patient began to cough on the endotrachealtube and became hypertensive. Nitroglycerin boluseswere given, the tube was removed and the patientnormalized. The initial neurologic examination wasfavorable.

The patient was then transported to the postanes-thesia care unit on 2 L nasal cannula where her neu-rologic exam was still unremarkable. After 20 minutesin the unit she had a grand mal seizure. Midazolamwas administered and the seizure was terminated, butshe was now lethargic and combative. Over the next10 minutes her blood pressure began to rise and shebecame bradycardic with a heart rate of around 48.

DiscussionThecauses of postoperative seizure aremany.Themostcommon causes seen in the postoperative period arelisted in Table 4.1. Regardless of the cause, seizuresneed to be treated rapidly as they represent anacute imbalance between cerebral oxygen supply anddemand and if uncorrected can lead to irreversibleneuronal damage. Benzodiazepines, barbiturates, andpropofol are all acceptable choices for seizure lysis [2].

The key to successful treatment is early seizure lysisas well as correct diagnosis and treatment of the under-lying cause. The initial differential must include themost life-threatening possibilities. Adequate oxygena-tion and ventilation must first be confirmed. Bleedingwithin the cranial vault must be ruled out if the neu-rologic exam worsens or vital signs are ominous (i.e.,Cushing reflex). In this case, hypertension and brady-cardia represented worrisome signs of an uncompen-sated increase in intracranial pressure. This is a crit-ical time because often there is a reflex reaction totreat the hypertension, whichwould be a gravemistakein this situation. The brain is an organ that autoreg-ulates blood flow. In this case the brain requires the

15


Table 4.1. Causes of seizures in postoperative patients.

Physiologicfactors

Hypoxia, hypercarbia, ischemia

Metabolicdisorders

Eclampsia, sodium imbalance, phosphateimbalance, calcium imbalance, renaldysfunction, hepatic dysfunction

Medications Antibiotics, antipsychotics, local anesthetics,cocaine, amphetamines, phencylidine

Drugwithdrawal

Barbiturates, benzodiazepines, alcohol, opioids

Infection Febrile seizures, abscess, encephalitis

Traumatichead injury

Hemorrhage or contusion

Surgicalinjury

Cortical irritation from craniotomy orendovascular cranial intervention

higher blood pressure to continue to perfuse because itis being opposed by the increase in the pressure withinthe cranial vault. Lowering this patient’s blood pres-suremay result in brain ischemia and irreversible braindamage. If bleeding is indeed believed to be the issuesuch as in this case, surgical decompression with orwithout imaging can be employed. If imaging is used,computed tomography scan is the modality of choiceas it can be obtained and performed quickly.

If imaging does not confirm the diagnosis ofhemorrhage, other etiologies must be considered.Electrolytes should be measured. In cases of mas-sive transfusion citrate toxicity may occur and causehypocalcemia and seizures. Citrate is added to storedblood to bind calcium and prevent coagulation of thestored blood. If enough blood is given, serum calciumlevels can fall low enough to cause a seizure. Sodiumimbalance may also be present and serum/urinesodium levels are helpful. Should prophylactic anti-epileptics have been given in the operating room?Depending on the dose and route of administrationof the antiepileptic given, therapeutic levels may nothave been achieved by the time of operation.This maysimply be cortical irritation from the surgery becauseof nontherapeutic levels of the antiepileptic. A thor-

ough medication review and discussion with the sur-gical team will help you decide on the prophylaxis tobe given.

In this case, the cause was found to be a subduralhematoma at the surgical site and the patient returnedto the operating room for decompression.Thepatient’semergence was less than desirable during the initialoperation. Could we do anything different from ananesthesia perspective to make this emergence better?While administration of lidocaine intratracheallyor intravenously may be of use, it may also lowerthe seizure threshold. Some advocate a remifentanilinfusion be used intraoperatively as it can provide asmooth emergence and because of its short half lifea quick return to neurologic baseline. One can alsoconsider extubating the patient under deep anesthesiaand emerging the patient without an entotrachealtube in place. This technique is not without risk. Therisks include aspiration, laryngospasm if the patient isnot deeply anesthetized, inability to ventilate and theneed to reintubate along with its undesirable hemo-dynamic changes. These risks make this techniqueone to be considered only when the benefits trulyoutweigh the risks. These techniques can be used inan attempt to make the emergence smooth. However,there is no guarantee and emergence can still beunpredictable.

ConclusionPostoperative seizures can have physiologic, pharma-cologic, andpathologic causes. Immediate terminationof the seizure should be followed by a rapid assessmentof reversible physiologic causes and then amore exten-sive differential to identify the underlying source.

References1. M. A. Mirski. Seizures and status epilepticus in the

critically ill. Critical Care Clinics 2008; 24: 115–47.2. T. P. Bleck. Intensive care unit management of status

epilepticus. Epilepsia 2007; 48: 59–60.

16


5 Painmanagement for craniotomiesPaul E. Hilliard and Chad M. Brummett

The perioperative pain management for craniotomiescan be extremely challenging. Patients often presentwith other significant co-morbidities that can affecttheir anesthetic management and complicate post-operative pain therapy. While it is essential to keeppatients comfortable and hemodynamically stable, itis often equally important that they be responsive topostoperative neurologic examination. In this chap-ter, we present a common clinical scenario and offeroptions for perioperative pain management.

Case descriptionA 52-year-old female American Society of Anesthesi-ologists class 3 patient presented for clipping of a cere-bral aneurysm. She was morbidly obese (body massindex= 36) and carried diagnoses of obstructive sleepapnea and refractory gastroesophageal reflux disease.She had a history of postoperative nausea and vomit-ing requiring hospitalization after an elective surgicalprocedure. In addition, she had hypertension, with abaseline blood pressure of 140/80 and described short-ness of breath after climbing one flight of stairs.

The primary concerns for the anesthesiology teamin this case were (1) airwaymanagement in amorbidlyobese patient with severe gastroesophageal reflux dis-ease, (2) maintaining hemodynamic stability duringlaryngoscopy and incision in this patient with a cere-bral aneurysm, and (3) pain management in a patientwith sleep apnea, obesity and a history of postoperativenausea and vomiting.

In preparation for the stimulation associated withintubation, a radial arterial line was placed prior toinduction.A rapid sequence intubationwas performedwith a bolus of esmolol prior to laryngoscopy toprevent hemodynamic derangement. After a smoothinduction, an additional dose of fentanyl was givenprior to performing a scalp nerve blockade (describedbelow). Anesthesia was maintained with dexmedeto-midine, propofol, and sufentanil infusions. The nerve

block was completed uneventfully and without hemo-dynamic compromise and adequate time was allowedfor the block to take effect. The patient maintaineda stable blood pressure during insertion of Mayfieldpins and had no response to incision or calvarial con-tact. The case was completed without incident, mul-tiple anti-emetics were administered, and the tracheawas extubated in the operating room.The patient com-pleted a normal neurologic examination. She had anuneventful recovery in the hospital with no nausea,and her pain was well controlled with minimal opioidmedication.

DiscussionThe case described above is a common example of thecomplexity frequently associated with neurosurgicalpatients. The combined regimen provided for analge-sia and hemodynamic control, while allowing for anadequate neurologic examination. In addition, opioidswere limited, thereby decreasing the risk of postoper-ative nausea and vomiting.

Intravenous analgesicsOpioids are a key component of intraoperative andpostoperative pain management for craniotomies.Fentanyl is the most commonly used analgesic dueto its rapid onset of action and hemodynamic sta-bility. Morphine can cause histamine release, whichcan lead to venodilation and subsequent hypotension.Remifentanil has become a very popular option due toits ultra-short half life and ease of titration; however,longer-acting opioids must be given after the infusionis stopped. Sufentanil is more potent than fentanyl andhas a duration of action between that of remifentaniland fentanyl.

Dexmedetomidine is an alpha-2-adrenoceptoragonist that is frequently used in neurosurgical anes-thesia [1]. Whereas dexmedetomidine will not sufficeas a sole anesthetic, it provides sedation, anxiolysis,

17


Zygomaticotemporal n.Supraorbital n.

Auriculotemporal n.

Supratrochlear n.

Greater occipital n.

Lesser occipital n.

Figure 5.1. Innervation of the scalp.Blockade of the superficial nervesinnervating the scalp can be easilyaccomplished with a small volume oflocal anesthetic using anatomiclandmarks.

and analgesia with minimal respiratory depression.The sedation associated with dexmedetomidine hasbeen termed “cooperative sedation.” The calm andcooperative nature of patients on dexmedetomidinecan be ideal for postoperative neurologic examination.Although it has not been demonstrated in humans,there are multiple laboratory studies demonstratingthat it is also neuroprotective in animal models ofcerebral and spinal cord ischemia.

Regional anesthesiaIn 1918 Harvey Cushing used local infiltration forcraniotomy in head-injury patients. He also mappedthe sensory distribution of the trigeminal nerveand coined the term “regional anesthesia.”φ Since

that time the “scalp block” has continued to evolve.Approximately 60–80% of patients experience mod-erate to severe pain after craniotomy [2]. Blockadeof the scalp is an effective, simple regional anesthetictechnique that significantly improves patient comfortafter craniotomy [3]. The use of regional anesthetictechniques for pain control is generally desirablewhen compared with opioid-based therapy, which canbe associated with nausea and vomiting; such sideeffects are particularly problematic in neurosurgicalpatients. Additionally, a scalp block after induction ofanesthesia and prior to placement of Mayfield headpins has been shown to significantly blunt hemo-dynamic response to pinning [4]. This is of obviousbenefit in patients with an intracranial aneurysm,vascular anomaly or mass effect causing increased

18

Case 5. Pain management for craniotomies

Supratrochlear n.

Supraorbital n

Zygomaticotemporal n.

Auriculo-temporal n.

Lesser occipital n.

Greater occipital n.

Third occipital n.

Figure 5.1. (cont.)

intracranial pressure. While regional anesthesia inthis situation would appear to be an attractive option,it may not have significant impact on postoperativeopioid requirements for pain control. Recently, a smallprospective analysis of perioperative skull block-ade failed to demonstrate significant benefit whencompared with patients treated with a remifentanilinfusion, which was titrated to blunt hemodynamicresponse during pin insertion [5].

Scalp blockade is generally performed after induc-tion of anesthesia and preferably prior to noxious stim-ulation (placement of head in pins, incision, periostial-dural contact) and/or at the end of surgery, priorto awakening. This blockade provides anesthesia tothe peripheral nerves that innervate the scalp. Theseinclude the greater and lesser occipital nerves, thesupraorbital and supratrochlear nerves (from cranialnerveV, first division), the zygomaticotemporal nerves(from cranial nerve V, second division), the auricu-lotemporal nerves (from cranial nerve V, third divi-

sion), and the greater auricular nerves that arise fromthe superficial cervical plexus (Figure 5.1).

Blockade of the supraorbital and supratrochlearnerves are accomplished with 2 mL of local anestheticsolution as the nerves emerge from the orbit usinga 23- or 25-gauge needle introduced above the eye-brow perpendicular to the skin. The auriculotempo-ral nerves are blocked with 5 mL of anesthetic injected1.5 cm anterior to the ear at the level of the tragus byintroducing the needle perpendicular to the skin andinfiltrating along the deep and superficial fascia. Thepostauricular branches of the greater auricular nerveare blocked with 2 mL of local anesthetic between theskin and bone located 1.5 cm posterior to the ear atthe level of the tragus. The greater and lesser occipi-tal nerves can be blocked with 5 mL of local anestheticintroduced by infiltration technique along the superiornuchal line, approximately halfway between the occip-ital protruberance and mastoid process. The greateroccipital nerve lies in a groove, the occipital notch,

19


which can be palpated just lateral to the external occip-ital protuberance. Mild pressure should be applied atthe site of injection for approximately 1 minute todecrease the risk of a hematoma.

Long-acting local anesthetics, such as bupivacaineor ropivacaine, should be used to maximize the dura-tion of analgesia. If the speed of onset is important, asis needed to blunt hemodynamic response to head pin-ning, lidocaine can be added to the solution. Unlikeother regional anesthesia techniques, motor stimula-tion is not required to complete the block. Therefore,in long cases, the nerve blocks can be repeated aftersurgery to maximize the duration of analgesia andimprove the immediate postoperative course.

Regional anesthesia has a limited but importantrole in anesthetic management of the neurosurgicalpatient. These blocks are relatively simple to performand may offer significant advantages, such as man-agement of an awake patient, blunted hemodynamicresponse to pin insertion, and improved postoperativepain control with less need for opioid analgesics. Theblocks described above are limited to superficial struc-tures and are associated with a low rate of complica-tions. The neuroanesthesiologist must weigh the riskand benefit of each block, but they are clearly a valu-able asset.

ConclusionIn conclusion, craniotomies are often challengingcases with a number of important anesthetic consid-

erations. A combination of intravenous analgesics andregional anesthesia can provide excellent pain reliefand decrease the wide hemodynamic changes that canaccompany anesthesia and surgery.

References1. A. Bekker,M. K. Sturaitis. Dexmedetomidine for

neurological surgery. Neurosurgery 2005; 57: 1–10.2. G. De Benedittis, A. Lorenzetti,M.Migliore et al.

Postoperative pain in neurosurgery: a pilot study inbrain surgery. Neurosurgery 1996; 38: 466–9.

3. A. Nguyen, F. Girard,D. Boudreault et al. Scalp nerveblocks decrease the severity of pain after craniotomy.Anesth Analg 2001; 93: 1272–6.

4. M. L. Pinosky, R. L. Fishman, S. T. Reeves et al.Theeffect of bupivacaine skull block on the hemodynamicresponse to craniotomy. Anesth Analg 1996; 83: 1256–61.

5. F. M. Gazoni, N. Pouratian, E. C. Nemergut. Effect ofropivacaine skull block on perioperative outcomesin patients with supratentorial brain tumors andcomparison with remifentanil: a pilot study.J Neurosurg 2008; 109: 44–9.

Endnote

φ http://www.scalpblock.com/, 2007, Irene P. Osborne, MD, Lastaccessed December 5, 2009.

20

Part I Craniotomy. Posterior fossa craniotomyCase

6 Preoperative evaluationGeorge A. Mashour

The posterior fossa is an intracranial compartmentthat houses the cerebellum and the brainstem. Masslesions and increased intracranial pressure (ICP) inthis area can have profound consequences for neuro-logic, cardiac, and respiratory functions (Table 6.1).

Case descriptionThe patient was a 54-year-old female who presentedto her primary-care physician with complaints ofheadaches, progressive hearing loss and episodes ofaspiration that were increasing in frequency. Thepatient was referred to a neurologist. Computedtomography and magnetic resonance imaging wereperformed, revealing bilateral acoustic neuromas. Adiagnosis of neurofibromatosis type 2 was made anda posterior fossa craniotomy was scheduled. Thepatient’s husband reported that her voice was hoarseand she had become increasingly somnolent in the pre-ceding 24 hours. Physical examination revealed a som-nolent but arousable patient who had difficulty hear-ing. The patient reported headaches when asked toextend her neck for an airway evaluation.

Immediate concerns for the anesthetic teamincluded (1) a depressed level of consciousness,(2) the potential for increased ICP, and (3) aspirationon induction. Intravenous and arterial line accesswas obtained and the patient was transported tothe operating room. The head of the bed was ele-vated to ensure adequate venous drainage and nopremedication was given in order to avoid hypoven-tilation. A rapid sequence induction with propofoland succinylcholine was achieved without difficulty.Further large-bore intravenous access was obtainedand sevoflurane was used for maintenance anesthesia.A remifentanil infusion was initiated because of theneed to avoid neuromuscular blockade for cranialnerve monitoring.

With the exception of several episodes of transientvagal asystole, the case proceeded without difficulty.

Table 6.1. Diseases of the posterior fossa.

Extra-axial Intra-axial

Schwannoma Medulloblastoma

Nerve sheath tumors Cerebellar astrocytoma

Clival tumors Brainstem glioma

Arachnoid cysts Ependymoma

Glomus jugulare tumors Choroid plexus papilloma/Carcinoma

Epidermoid cysts Dermoid tumors

Epidermoid tumors Hemangioblastoma

Meningioma Metastatic disease

Fatty lesions Astrocytoma

Leptomeningeal diseases Lymphoma

Infection/inflammatory

Vascular

Metabolic

At the end of the case the sevoflurane and remifen-tanil were discontinued and no long-acting opioid hadbeen administered for several hours. However, thepatient failed to emerge. Irregular respiratory patternswere observed and the patient was not following com-mands. Laboratory values from 15 minutes prior tothe end of the case revealed normal glucose and elec-trolytes. An arterial blood gas sample was analyzed,which demonstrated normal pH, CO2, and O2. Afterruling out physiologic and pharmacologic causes, anemergent computed tomography scan was performeddemonstrating hemorrhage in the posterior fossa. Thepatient was brought back to the operating room for re-exploration and hemostasis.

DiscussionNeurosurgical procedures involving the posteriorfossa can be challenging for both the surgical and anes-thetic teams. It is imperative to understand the vital

21

I. Craniotomy. Posterior fossa craniotomy

Figure 6.1. Structures of thecerebellopontine angle.

structures housed within this small compartment inorder to best evaluate the patient in the preoperativesetting.

MedullaIn addition to the ascending and descending fiber bun-dles that traverse themedulla, there are important car-diac and respiratory centers. Compression or dysfunc-tion of the medulla can therefore have catastrophicconsequences.

PonsThe pons contains several critical nuclei that serve toarouse the cerebral cortex [1] including the noradren-ergic locus ceruleus, the serotonergic dorsal raphe,and the cholinergic laterodorsal and pedunculopon-tine tegmentum. Activity of these and other arousalcenters is necessary for consciousness.Thus, compres-sion or dysfunction of the pons can be associatedwith decreased levels of consciousness, as seen in ourpatient preoperatively.

Fourth ventricleThe fourth ventricle connects the spinal and cere-bral subarachnoid compartments, across which cere-brospinal fluid normally communicates. Mass lesionsin the posterior fossa may therefore lead to a noncom-municating hydrocephalus resulting in increased ICP.

Cerebellopontine angleLesions of the cerebellopontine angle (such as acous-tic neuromas, which technically should be referred toas vestibular schwannomas) can compress the cranialnerves exiting from the brainstem (Figure 6.1). Oneof the most important of these nerves from the per-spective of the anesthesiologist is the vagus or cranialnerve X.The significance is 2-fold. First, intraoperativecompression or traction on the vagus can cause asys-tole, which is best terminated by cessation of surgeryuntil the heart rate returns. The other point of sig-nificance relates to the superior and recurrent laryn-geal nerves, which arise from the vagus. Since theseinnervate the glottis, dysfunction can result in diffi-culties with control of the airway, leading to chronic

22


aspiration as well as aspiration pneumonia [2].Impaired gag reflex during preoperative evaluationindicates dysfunction of cranial nerves IX or X.

CerebellumDisorders of the cerebellum will manifest as ataxia,dysmetria, intention tremor, and wide-based gait.

VasculatureAneurysms and arteriovenous malformations are alsofound in the posterior fossa; cerebellar infarctions andhematomas may occur.

In addition to the general preoperative evalua-tion, attention to level of consciousness, cardiorespira-tory abnormalities, signs of intracranial hypertension,and airway control are of paramount importance withlesions in the posterior fossa. Complications of pos-terior fossa procedures include bleeding from venoussinuses, vagal asystole, venous air embolism, seizures,and delayed emergence [3]. Care should be taken to (1)obtain large-bore peripheral access or central venousaccess if necessary, (2) obtain arterial line for beat-to-beat blood pressure monitoring, (3) use appropriatemonitors for air emboli, depending on the positioning,and (4) develop a strategy for prompt emergence fromanesthesia such that any pathology can be rapidly iden-tified. Brainstem auditory evoked potentials and facialnerve monitoring are common during posterior fossaprocedures. Unlike somatosensory or motor evokedpotentials, brainstem potentials are remarkably resis-tant to the effects of inhalational anesthetics [4]. As

such, the only important modulation of the anestheticplan relates to neuromuscular blockade, which shouldbe avoided if the facial nerve is being monitored. Opi-oid infusions such as the remifentanil used in our casecan aid in maintaining a quiescent surgical field.

ConclusionIn conclusion, the posterior fossa contains impor-tant neural structures that control essential functionssuch as cardiac and respiratory modulation, corticalarousal, and airway control. Preoperative assessmentand preparation of these patients should focus on theconsequences of posterior fossa abnormalities as wellas the potentially life-threatening complications thatmay occur intraoperatively.

References1. N. P. Franks. General anaesthesia: from molecular

targets to neuronal pathways of sleep and arousal.Nat Rev Neurosci 2008; 9: 370–86.

2. S. N. Shenoy, A. Raja. Acute aspiration pneumoniadue to bulbar palsy: an initial manifestation ofposterior fossa convexity meningioma. J NeurolNeurosurg Psychiatry 2005; 76; 296–8.

3. G. P. Rath, P. K. Bithal, A. Chaturvedi et al.Complications related to positioning in posterior fossacraniectomy. J Clin Neurosci 2007; 14: 520–5.

4. M. Banoub, J. E. Tetzlaff, A. Schubert. Pharmacologicand physiologic influences affecting sensory evokedpotentials: implications for perioperative monitoring.Anesthesiology 2003; 99: 716–37.

23


7 Air embolismJennifer Vance and Sheron Beltran

A venous air embolism (VAE) is a potentially life-threatening event caused by air in the vascular system.The entrainment of air from an operative site into thevenous vasculature produces a wide array of systemiceffects. The incidence is difficult to estimate due to thefact that many cases are subclinical. In addition, detec-tion is dependent upon the sensitivity of the equip-ment used during the procedure.

Case descriptionThe patient was a 37-year-old female with a right-sided acoustic neuroma presenting for a suboccipitalapproach to tumor resection. Her past medical historywas positive for obstructive sleep apnea treated witha continuous positive airway pressure device, obesitywith a body mass index of 35, and a history of a dif-ficult airway with a previous Cormack–Lehane GradeIV view on laryngoscopy. Neurophysiologic monitor-ing of brainstem auditory evoked responses and cra-nial nerve function was planned.

The primary anesthetic concerns were (1) manage-ment of the known difficult laryngoscopy, (2) anes-thetic maintenance consistent with successful neuro-physiologic monitoring, and (3) prompt emergenceto enable rapid assessment of any pathologic eventsin the cerebellopontine angle. In preparation for theawake fiberoptic intubation, glycopyrrolate was givenas the antisialogogue. This was followed by nebu-lized and atomized 4% lidocaine for airway topical-ization. A remifentanil infusion provided sedationfor the procedure. After confirmation of endotrachealintubation, maintenance anesthesia was achieved bya remifentanil infusion, isoflurane and nitrous oxideeach at 0.5 minimum alveolar concentration. Thepatient was placed in the lateral park bench positionand the case proceeded without complication until2 hours after surgical incision when the anesthesiateam noticed a simultaneous fall in end-tidal car-bon dioxide (ETCO2), oxygen saturation, and blood

pressure. The oxygen saturation dropped from 99%to a nadir of 68%, the ETCO2 dropped from 30 to18 mmHg, and the blood pressure fell from 110/60to 75/40. Extra assistance was immediately called tothe room, the patient was placed on 100% oxygen,and after the surgeons were notified of probable VAEthey proceeded to flood the surgical field with normalsaline. The patient was placed in the head down posi-tion and a valsalva maneuver was delivered. Phenyle-phrine was bolused for blood pressure support. Theendotracheal tube was visualized and its positionappeared unchanged. The patient remained easy toventilatewith no suspicion of airway obstruction.Withthe above maneuvers, the patient’s condition rapidlyimproved and returned to baseline within 6 minutes.As the patient stabilized, the decisionwasmade to pro-ceed with a precordial Doppler in place for the rest ofthe case. The surgery continued without further com-plication and the trachea was uneventfully extubatedat the end of the operation; the patient was neurologi-cally intact.

DiscussionVenous air embolism was historically most often asso-ciated with craniotomies performed in the sittingposition. It has been estimated that VAE occurs inup to 40% of sitting posterior fossa craniotomies,although the majority of these do not result in hemo-dynamic compromise [1]. A combination of fac-tors results in the increased risk for air entrain-ment during neurosurgical procedures. Unlike mostof the venous system, the dural sinuses and bridg-ing epidural veins are noncollapsible structures thatremain open to the surrounding atmosphere. Theheight of the surgical site above the right side of theheart also contributes to the likelihood of venous airembolism.

Despite this, venous air emboli have also beendescribed with high frequency during neurosurgical

24

Case 7. Air embolism

operations performed in the supine, lateral, or pronepositions, lumbar laminectomies, hip replacementsurgeries, laparoscopic surgeries, and Cesarean sec-tions, among others [2]. Morbidity andmortality asso-ciated with VAE are dependent upon the rate of airentrainment, the volume of air entrained, and theposition of the patient. The lethal volume of air isunknown, however, injections of 100–300 mL havebeen reported to be fatal [2]. In large air emboli(5mL/kg), there can be complete right ventricular out-flow tract obstruction, which rapidly leads to right-sided heart failure and cardiovascular collapse. Inmore moderate-sized air emboli, there can be par-tial outflow tract obstruction leading to decreased car-diac output, hypotension, myocardial ischemia and, ifuntreated, death [3].

Clinical presentation depends on the severity of theair embolus. The use of nitrous oxide does not worsenthe clinical severity of a VAE provided its use is dis-continued upon detection of an air embolus [4]. Car-diovascular changes include tachyarrhythmias, evi-dence of right heart strain on EKG, and hypotensionsecondary to decreased cardiac output. Pulmonaryartery pressures increase due to decreased cardiac out-put and spasm; central venous pressures increase dueto right heart strain. Pulmonary changes depend onwhether the patient is awake or under anesthesia.Awake patients will complain of dyspnea, coughing,and a sense of impending doom. In ventilated patients,one will see decreases in ETCO2, arterial oxygen sat-uration, arterial oxygen tension, and hypercapnia [3].Neurologic changes consist of altered mental statuseither due to cerebral hypoperfusion, cardiovascularcollapse, or from direct cerebral air embolus through apatent foramen ovale.

There are several monitors that are capable ofdetecting venous air emboli. The most sensitive istransesophageal echocardiography (TEE). It is able todetect 0.02 mL/kg of air administered by bolus injec-tion [2]. It also allows for detection of paradoxical arte-rial emboli via a patent foramen ovale. The presenceof TEE also enables direct visualization of air aspira-tion through a central catheter if a VAE should occur.The downside is that it is expensive, invasive, andrequires both expertise and constant surveillance [2].Positioning aTEEduring a sitting craniotomy inwhichthe head is flexed can also prove difficult and thereis little information on the safety of maintaining theTEE probe in this position for a prolonged period oftime.

Table 7.1. Monitors used to detect venous air emboli.

MonitorSensitivity(ml/kg) Comment

Transesophagealechocardiography

0.02 Can identify paradoxicalembolus; invasive, requirestraining

PrecordialDoppler

0.05 Most sensitive noninvasivemonitor; difficult to use inobese patients and inlateral position

Pulmonary arterycatheter

0.25 Increased pulmonaryartery pressure correlateswith amount of airentrained; invasive

End-tidal N2 0.5 Affected by use of nitrousoxide; change occurs30–90 s before end-tidalCO2 change; notuniversally available

End-tidal CO2 0.5 Standard monitor; lesssensitive than above,nonspecific for VAE

Oxygen saturation low Late clinical sign requiressevere physiologicdisturbance

Direct observation low Look for back bleedingfrom bone or veins; nophysiologic data

Themost sensitive noninvasive monitor is the pre-cordial Doppler ultrasound which can detect as lit-tle as 0.25 mL (0.05 mL/kg) of bolused air [3]. It ispositioned along the right sternal border or betweenthe right scapula and the spine. With small amountsof air entrainment a “washing machine” sound canbe appreciated. With larger amounts of air, a classic“mill wheel” murmur can be heard. Pulmonary arterycatheter, end-tidal nitrogen, ETCO2, pulse oximetry,and direct observation can also be used to detect VAE(Table 7.1).

Upon detection of a VAE, several interventionscan help prevent further air entrainment thus mini-mizing catastrophic hemodynamic changes. First, fur-ther air entrainment is prevented by notification ofthe surgeon to flood the field with normal saline.The operative site should also be lowered below thelevel of the right atrium. If the surgical procedureinvolves the cranium, then temporary occlusion ofthe bilateral jugular veins can be employed to preventfurther entrainment, increase venous pressure, andpromote retrograde flow. At the same time, the anes-thesia provider should switch to 100% oxygen asnitrous oxide will increase the size of the air embolus.

25


Fluids should be administered judiciously in order toavoid further right heart strain. A valsalva maneu-ver can be performed to limit air entrainment. If aright atrial catheter is in place, aspiration of air canbe attempted. However, there are no data to supportemergent insertion of a venous catheter. As cardiovas-cular collapse can ensue, hemodynamic support withvasopressors and inotropes must be readily employed.Onemust also be prepared to initiate cardiopulmonaryresuscitation, which may require patient reposition-ing. The application of positive end expiratory pres-sure is controversial in this scenario, as it may increasethe risk for paradoxical air embolism or worsen car-diovascular disturbance; however, it may be beneficialfor prevention of a VAE [3].

ConclusionIn conclusion, VAE is a potentially life-threateningevent that can occur not only in neurosurgical

procedures, but also in orthopedic, laparoscopic, andpelvic surgeries.Monitors for high-risk cases should bechosen depending on the expertise of the anesthesiol-ogist, the surgery being performed, and the position ofthe patient.

References1. A. R. Fathi, P. Eshtehardi, B. Meier. Patent foramen

ovale and neurosurgery in sitting position: asystematic review. Br J Anaesth 2009; 102: 588–96.

2. S. C. Palmon, L. E. Moore, J. Lundberg et al. Venousair embolism: a review. J Clin Anesth 1997; 9: 251–57.

3. M. A. Mirski, A. V. Lele, L. Fitzsimmons et al.Diagnosis and treatment of vascular air embolism.Anesthesiology 2007; 106: 164–77.

4. T. J. Losasso, S. Black,D. A. Muzzi et al. Detectionand hemodynamic consequences of venous airembolism: does nitrous oxide make a difference?Anesthesiology 1992; 77: 148–52.

26


8 Delayed emergence after posteriorfossa surgeryLeif Saager and Alparslan Turan

Immediate emergence after neurosurgery is desirableto facilitate neurologic examination and early iden-tification of complications. Awakening is determinedby many factors including preoperative status, type ofsurgery, and intraoperative events.

Case descriptionA 58-year-old female with a body mass index of 32complained about gradual hearing loss, increasing fre-quency of headaches and vertigo and subsequently wasdiagnosed with an acoustic neuroma. Her past med-ical history included type 2 diabetes, hypertension,seasonal allergies and 15 pack-years of smoking. Pre-operative medications included metformin, enalapril,ibuprofen, atorvastatin, and diphenhydramine.

Preoperative magnetic resonance imaging wasconsistent with a 1.8-cm large left acoustic neuromaand the patient was scheduled for elective microsurgi-cal tumor resection using a posterior fossa approach.

Anesthesia was induced with 150 mcg fentanyl,200 mg propofol and 50 mg rocuronium and anesthe-sia was maintained with sevoflurane and 0.1–0.3 mcg/kg/min remifentanil. Throughout the surgery meanblood pressure was maintained between 70–85mmHgand heart rate between 65–90 beats per minute. After5 hours of surgery, estimated blood loss was minimal.A total of 1650 mL crystalloid and 500 mL colloidfluid were given. At closure of dura, 6 mg of morphinewas administered for analgesia and consequently thepatient was turned supine and all anesthetic agents dis-continued.

During the following 30 minutes the patient failedto emerge from anesthesia. During this period onlyinsufficient spontaneous breathing patterns and noresponse to painful stimuli were observed. Heart rateand blood pressure were within preoperative range.Core temperature was recorded as 36.2 ◦C despiteintraoperative warming with forced heated air.The leftpupil was 4 mm in diameter and minimally reactive to

light, the right pupil was 2mm in diameter and reactedbriskly.

After checking for residual neuromuscular blockand determining a normal train-of-four response anarterial blood gas sample was sent to the laboratory.Results returned within normal ranges except for ablood glucose of 180mg/dL.Adose of 0.1mgnaloxonewas given to antagonize opioid effects but no improve-ment in neurologic status was observed. An emer-gent computed tomography (CT) scan was obtainedshowing no hematoma, ischemia, pneumocephalus orbrain-stem compression. A CT-angiogram was per-formed,which demonstrated an occludingmid-basilarartery clot. After consultation with radiologists, neu-rosurgeons, and family members a decision for intra-arterial local application of alteplase was made andpartial recanalization achieved. The patient was trans-ferred to the neurologic intensive care unit and keptonmechanical ventilation.Onemonth postoperativelythe patient was discharged to a nursing care facilitywith ataxia, and swallowing difficulties.

DiscussionIn order to facilitate early detection of neurologiccomplications, anesthesiologists usually aim for anearly emergence following intracranial surgery. How-ever, some clinical conditions may require prolongedsedation. Hypothermia, hypertension, and coagula-tion disorders as well as prolonged surgery or intra-operative brain swelling might be reasons for delayingemergence.

Delayed awakening from anesthesia is defined asa failure to regain consciousness within 20–30 min-utes after the end of surgery.Themost common causesof delayed postoperative emergence include residualdrug effects, respiratory failure, metabolic derange-ments, and neurologic complications (Table 8.1). Earlyrecognition of these postoperative complications isextremely important since most of them requireimmediate surgical or medical attention.

27


Table 8.1. Common causes for delayed emergence afterneurosurgery.

SystemicResidual drug effectsOverdose, potentiation of drugs, prolonged neuromuscular

blockadeCentral anticholinergic syndromeRespiratory failureHypoxia, hypercapnia

Metabolic and endocrine derangementsHypo-, hyperglycemia, electrolyte imbalance,

hypothyroidismHypothermiaHypotension

NeurologicCerebral hypoxiaHemorrhageThrombosis, embolismSeizureCerebral swellingPneumocephalus

Neuromuscular blockade is commonly used tobetter allow for ventilation and improved operat-ing conditions. Residual neuromuscular blockade hasbeen identified as a significant postoperative problemincreasing morbidity and mortality [1]. Risk factorslike advanced age and hepatic or renal dysfunctiondecrease clearance of muscle relaxants and thus pro-long their effects. The duration of action and recov-ery time of muscle relaxants are significantly increasedby hypothermia during anesthesia, mainly because ofreduced elimination rate [2]. In this case hypothermiawas identified as a possible risk factor and neuromus-cular monitoring was used to exclude residual effectsof muscle relaxant.

Opioids, in addition to analgesic action, can alsohave a strong sedative effect in susceptible patients.Remifentanil has an extremely short context-sensitivehalf life that allows for excellent analgesic control andrapid emergence. In order to avoid an analgesic gapat the end of surgery, postoperative pain therapy hasto be initiated before discontinuation of remifentanil.The patient received morphine for translational anal-gesia, which might have contributed to sedation anddelayed emergence.Naloxonewas administered to ruleout a possible opioid effect.

With conclusion of surgery the patient is weanedoff controlled mechanical ventilation and transferredto a spontaneous breathing regimen before extuba-tion. Ineffective spontaneous respiration results inhypoxia and hypercapnia. Arterial carbon dioxide lev-els exceeding 80 mmHg are associated with signifi-cantly impaired consciousness. This usually resolves

without neurologic deficit as arterial PaCO2 declines.In the present case arterial blood gas analysis revealedno abnormality.

Hypoglycemia, hyponatremia, and hypo-osmo-lality can cause delayed emergence from anesthe-sia. An underlying metabolic disorder, side effects ofcurrent medications or consequences of the surgicalprocedure may be responsible for these electrolytederangements. In patients with type 2 diabetes melli-tus frequent measurements of blood glucose levels arenecessary to prevent potentially devastating complica-tions of undetected hypoglycemia. Metformin, com-monly used to treat diabetes mellitus, can cause lacticacidosis and hypothyroidism, both delaying recoveryfrom anesthesia.

After excluding common systemic and anesthesia-related causes for prolonged awakening, a brief neuro-logic assessment should be performed to further dif-ferentiate cerebral complications.

Cerebral swelling can occur intraoperatively orpreexist. In the context of posterior fossa surgerythis is of particular concern. Extensive intraoperativemanipulation close to the brainstem can lead to nega-tive effects on the respiratory drive. Rapid emergencemight increase risk and not be the optimal choice forthese patients.

Pneumocephalus, the presence of air in the cere-bral cavity, is a common complication of posteriorfossa surgery and may occur in up to 78% of cases [3].Surgical decompression, decreased brain volume fol-lowing mannitol administration and hyperventilationare contributing factors. Positioning during surgeryhas been associated with increased incidences for sit-ting, park-bench, and prone positions. Tension pneu-mocephalus always represents a neurosurgical emer-gency requiring immediate surgical intervention.

Intracranial thrombotic events as well as hemor-rhage can lead to rapid deterioration of the patientand cause failure to emerge after anesthesia. All ofthe abovementioned cerebral complications requireradiologic imaging to diagnose and develop a treat-ment strategy. Due to the immediate nature of thecomplications and possibly fatal outcomes an emer-gent CT is considered the gold standard. The CTscan of the presented patient revealed none of theobvious complications of posterior fossa surgery.Computed tomography scans have a low sensitiv-ity for early ischemia detection and interpretation isimpaired by bony structures. Spiral CT–angiography,magnetic resonance imaging, or magnetic resonance

28

Case 8. Delayed emergence after posterior fossa surgery

angiography are more helpful in identifying occludedvessels.

A CT–angiogram led to the definite diagnosis inour patient. Basilar artery occlusion is a rare compli-cation following neurosurgery and is associated withpoor prognosis. The mechanisms are due to embolismor results of artherothrombosis. Abnormal level ofconsciousness and pupillary abnormalities were typ-ical signs of basilar artery thrombosis demonstratedby our patient. Others would include hemi- or quadri-paresis, coma, bulbar and pseudo-bulbar signs. Thesemight be difficult to evaluate in the immediate periop-erative setting.

A variety of treatment options are available, includ-ing anticoagulation with heparin, surgical thrombec-tomy, transluminal angioplasty and stenting, intra-arterial or systemic thrombolysis. The immediatepostoperative period limited the therapeutic possibil-ities in our patient. In a multidisciplinary approachintra-arterial thrombolysis was considered the bestbalance between aggressive treatment to reverse theocclusion and the high risk of hemorrhage.

ConclusionEarly recovery after neurosurgical procedures is pre-ferred in order to facilitate neurologic assessment anddiagnosis of adverse outcomes. Multiple factors con-tribute to a delayed emergence from anesthesia and asystematic approach to rule out all possible causes isnecessary.

References1. M. S. Arbous, A. E. Meursing, J. W. van Kleef et al.

Impact of anesthesia management characteristics onsevere morbidity and mortality. Anesthesiology 2005;102: 257–68.

2. T. Heier, J. Caldwell. Impact of hypothermia on theresponse to neuromuscular blocking drugs.Anesthesiology 2006; 104: 1070–80.

3. T. J. K. Toung, R. W. McPherson, R. T. Donham et al.Pneumocephalus: effects of patient position and theincidence and location of aerosol after posterior fossaand upper cervical cord surgery. Anesth Analg 1986;65: 65–70.

29


9 Trigeminal neuralgiaBasem Abdelmalak and Joseph Abdelmalak

Trigeminal neuralgia (also known as tic douloureux) ischaracterized by unilateral, severe, paroxysmal burstsof neuropathic facial pain, which causes severe disabil-ity and is limited to the distribution of one or more ofthe divisions of the trigeminal nerve. It is not homo-geneous in presentation. It can be aching, throbbing,lancinating, or burning pain. Trigeminal neuralgia ismultifactorial in etiology: it may be due to vascu-lar compression, tumor, demyelination, or idiopathiccauses. The treatment options are either medical (car-bamazepine, oxacarbazepine, lamotrigine, neurontin,clonazepam, pregabalin, phenytoin, and baclofen) orsurgical [1]. Surgical management includes micro-surgical exploration/decompression [2], percutaneousprocedures (rhizotomies) on the Gasserian ganglion(radiofrequency lesioning, glycerol injection, ballooncompression), radiosurgery (“gamma knife” proce-dure), and neurectomy [1, 3]. General anesthesia istypically the technique of choice. Bradycardia and evenasystole as a result of the trigeminocardiac reflex cancomplicate these procedures [4, 5].

Case descriptionThe patient was an 88-year-old, right-handed femalewith a history of trigeminal neuralgia affecting the sec-ond division of the right trigeminal nerve (V2). Shewas treated with a gamma knife procedure, with verygood results for 2 years. However, her pain recurred. Itwas severe, 9/10 on pain visual analog scale, and “elec-trical shock” in quality. She had some improvementwith pregabalin to a pain score of 5/10. The patientwas admitted for a percutaneous balloon compressionprocedure.

The patient’s past medical history was also signifi-cant for essential hypertension, congestive heart fail-ure, and coronary artery disease, status post singleartery coronary artery bypass graft and aortic valvereplacement 15 years ago, as well as abdominal aor-

tic aneurysm repair around the same time. She had a40 pack-year tobacco history but quit 6 years ago.

On physical examination, her blood pressure was155/95 mmHg, pulse was 74. Her height and weightwere 144.8 cm and 69 kg with a body mass indexof 33. The patient’s medication included: allopuri-nol, colchicine, furosemide, pregabalin, coumadin,tramadol, lisinopril, hydrochlorothiazide, and meto-prolol.

After standard monitors were applied and twointravenous lines were inserted, general anesthesia wasinduced using propofol and fentanyl. Sevoflurane inair and oxygen was utilized for maintenance of anes-thesia.The head was positioned for satisfactory stereo-tactic localization utilizing fluoroscopy.The cheek wassterilized and draped in the usual fashion. About2.5 cm lateral to the labial crease, the skin was punc-tured. Through this puncture, a 20-gauge spinal nee-dle was then advanced to the region of the foramenovale using biplanar fluoroscopic guidance. The fora-men was then punctured. A 14-gauge introducer nee-dle was then advanced alongside this and it too punc-tured the foramen ovale. The 20-gauge needle wasthen removed and a 4-French Fogarty-type catheterwas passed through the needle. The needle was with-drawn from the foramen, then the balloon inflatedwith 0.75 mL of contrast for 60 seconds, monitoredby fluoroscopy. During inflation of the balloon, thepatient became bradycardic and hypotensive. She wastreated successfully with 0.6 mg glycopyrrolate givenintravenously. The balloon was then deflated andthe needle and catheter withdrawn. The case endeduneventfully.

DiscussionThe etiology of trigeminal neuralgia is poorlyunderstood, but common theories include vascularcompression, tumor, demyelination, or idiopathic

30

Case 9. Trigeminal neuralgia

causes [3]. If the pain is refractory to medical treat-ment, the patient may try trigeminal nerve block orone of the following surgical procedures:

1. Microsurgical exploration/decompression. Thesurgery is done under general anesthesia, with anincision usually performed behind the ear. Anopening is made in the skull about the size of aquarter. After opening the dura, the cerebellum isretracted and, using microsurgery, the trigeminalnerve inspected. If a vessel is found, it is dissectedfree of the trigeminal nerve and a Teflon sponge isplaced to keep the vessel from recompressing thenerve. The wound is then closed.

2. Percutaneous procedures. These proceduresinclude chemical, thermal, radiofrequencylesioning and the balloon compression procedure[1]. This is usually done under general anesthesiawhen the Gasserian ganglion is accessedpercutaneously. General anesthesia is preferred asthese procedures require a completely motionlesspatient often accomplished by adequate musclerelaxation, to facilitate the intervention andminimize injury to surrounding structures. Thisprocedure is described above.

3. The gamma knife procedure. Topicalization isaccomplished by injecting local anesthetics, atfour sites on the scalp. A reference frame is thensecured to the head at those sites. Some pressure iscommon for the first 5–10 minutes. Measurementsare taken and then the patient is sent for magneticresonance imaging and computed tomographyscans. These are transferred to the planningcomputer to define the region to be treated. Ifthere is a vessel compressing the nerve, theprocedure may be aborted. A plan for focusedradiation to the trigeminal nerve is created by theneurosurgeon. The radiation oncologist reviewsthe plan and assigns a dose of radiation. Thepatient lies down on the treatment couch. Thehead holder is used as a positioning device thatwill direct with great accuracy, up to 192 beams ofradiation to converge at the point picked on thecomputer. The patient is usually monitored by3 video cameras and 2 intercom systems. Thepatient then slides into the device up to theirlower chest for about 25–30 minutes and thenexits. When the treatment is completed, thepatient is taken to the recovery room where the

Table 9.1. A summary of the trigeminocardiac reflex pathway,as well as the triggering events, clinical effects, possiblepredisposing factors, prevention, and treatment.

Afferent Trigeminal nerve

Efferent Vagus nerve

Triggers 1. Surgical manipulations of trigeminalnerve or its sensory branches(stimulation of the ophthalmic branch→ oculocardiac reflex)

2. Surgical division of the dorsal sensoryroots of the trigeminal nerve

3. Percutaneous microcompression of thetrigeminal ganglion

Clinical effects Lasts for as long as the trigger is applied,will cease on removing the trigger

Negative chronotropic and ionotropicresponses:

BradycardiaHypotensionAsystole

Ventricular fibrillationApneaGastric hypermotility

Prevention andpreparation

1. May consider local anesthetic infiltrationin anticipation of the trigeminal nervestimulation/manipulation

2. Gentle manipulation of the trigeminalnerve and its sensory branches

3. A noninvasive and temporarypacemaker may be prepared inadvance for high-risk cardiac patients

4. May pretreat with glycopyrrolate inbradycardic patients

5. Tachycardia is not protective

Postulatedpredisposingfactors

HypercapneaHypoxemiaLight anesthesiaNarcoticsPreoperative beta-blockersPreoperative calcium channel blockersHigh resting vagal tone

Treatment 1. Stop the trigger2. Spontaneous recovery3. Mild bradycardia → glycopyrrolate ±ephedrine

4. Severe cases → atropine ±epinephrine

frame is removed and bandages applied. Thepatient is observed, typically between 30–120minutes, then discharged.

In the above case, the patient had a gamma knifeprocedure in the past. Upon recurrence of symptoms,balloon compression procedure was considered. Dur-ing inflation of the balloon, the anesthesiologist shouldbe ready with atropine or glycopyrrolate tomanage the

31


bradycardia and/or asystole that is reported to occurfrequently at this stage of surgery [5]. Sure enoughwe did encounter bradycardia and hypotension; it wasof moderate degree and was treated successfully with0.6 mg glycopyrrolate IV bolus. It should be notedthat it is not recommended to give any more narcoticsafter induction as it may predispose the patient to atrigeminocardiac reflex (see Table 9.1) and can alsohinder evaluating the success of surgery by observingpostoperative pain relief.

The trigeminocardiac reflex is well described withsuch procedures and is described in the next case.The noxious stimulus on the trigeminal nerve duringneedle insertion and balloon inflation is thought toinduce the reflex.This reflex can also be activatedwith-out surgery – the mere pain of trigeminal neuralgiahas reportedly resulted in bradycardia that was severeenough to cause syncope and in other instances degen-erated to complete cardiac arrest [4, 5]. Events such as

these demonstrate the potentially debilitatingnature oftrigeminal neuralgias.

References1. S. Prasad, S. Galetta. Trigeminal neuralgia: historical

notes and current concepts. Neurologist 2009; 15:87–94.

2. M. Sindou, J. M. Leston, C. Le guerinel et al.Treatment of trigeminal neuralgia with microvasculardecompression. Neurochirurgie 2009; 55: 185–96.

3. M. Obermann, Z. Katsarava. Update on trigeminalneuralgia. Expert Rev Neurother 2009; 9: 323–9.

4. R. Campbell,D. Podrigo, L. Cheung. Asystole andbradycardia during maxillofacial surgery. Anesth Prog1994; 41: 13–16.

5. S. T. Cha, J. B. Eby, J. T. Katzen et al.Trigeminocardiac reflex: a unique case of recurrentasystole during bilateral trigeminal sensory rootrhizotomy. J Craniomaxillofac Surg 2002; 30: 108–11.

32


10 Trigeminocardiac reflexHeather Hervey-Jumper and Christopher R. Turner

Bradycardia and even asystole may occur suddenlyduring posterior fossa surgery and requires immedi-ate evaluation and treatment in order to prevent poten-tial ischemia andmajor neurologic complications.Oneimportant mechanism for this clinical presentation isthe trigeminocardiac reflex (TCR).

Case descriptionA 53-year-old female with a history of progressiveheadaches and syncopal episodes was found to havea right-sided tentorial mass consistent with a falcinemeningioma. She underwent a surgical resection ofthe mass via a right-sided occipitotemporal cran-iotomy in the left lateral decubitus position with herhead rotated toward the floor. The pedicle of themeningioma was attached to the falx. During dissec-tion of the mass from its pedicle, the patient experi-enced asystole that resolved upon cessation of falcinemanipulation. Three subsequent individual and dis-crete mechanical stimulations of the falx reproducedsimilar but less severe hypotensive and bradycardicresponses that resolved with cessation of stimulation.After administration of glycopyrrolate the bradycar-dia with falcine traction was no longer evident. Thebalance of the surgery and the patient’s recovery wereuneventful.

Discussion

PathophysiologyThe TCR is typically described as bradycardia andhypotension seen with mechanical stimulation of thedistribution of the trigeminal nerve (cranial nerve(CN) V) [1]. The most common manifestation of theTCR is bradycardia, but other dysrhythmias includingjunctional rhythms, ventricular ectopy, AV blockade,asystole, and ventricular tachycardia can occur. Thisreflex is classically known as the “oculocardiac reflex”because of its prevalence in ophthalmologic surgery

Figure 10.1. The diagram shows the pathway of atrigeminocardiac reflex originating from the posterior falx.Mechanical stimulation of the posterior falx causes activation of thenervus tentorii [1]. Signal travels through V1 [2] via the Gasserianganglion and the trigeminal nerve [3] into the spinal trigeminalganglion [4]. Short internuncial fibers [5] connect to the motornucleus of the vagus [6]. Stimulation of this pathway results in reflexbradycardia.

[2]. The oculocardiac reflex is typically triggered byocular or extraocular stimulation,which is transmittedvia the ophthalmic branch (V1) of CN V to Gasserian(trigeminal) ganglion where V1 joins fibers from themaxillary (V2) and mandibular (V3) branches of CNV. From there the signal is transmitted to the trigem-inal sensory nuclei located in the pons and medullaalong the floor of the fourth ventricle (Figure 10.1).

33


The signal is then relayed via short internuncial nervefibers in the reticular formation to the immediatelyadjacent vagal nuclei (in the medulla). The efferentlimb is the vagus nerve, which when activated causesparasympathetic stimulation of the sinoatrial noderesulting in bradycardia and subsequent hypotension.Most but not all of this pathway (fromV1 to the sinoa-trial node) appears to be cholinergically mediated.

While classically described as resulting from thepathway described above, in fact stimulation of any ofthe branches of CN V can result in the initiation ofthe TCR.The reflex has been documented with maxil-lary,mandibular, zygomatic, and petrosal bonemanip-ulation, rhizolysis of the Gasserian ganglion, tumorresection in the cerebellopontine angle, irrigation ofthe temporomandibular joint, and endoscopic browlift surgery.

Penfield and McNaughton [3] in the early twenti-eth century using silver nitrate preparations of humanand rhesus monkey dura demonstrated extensiveinnervation of the meninges by all three divisions ofCN V. As examples, branches of V1 form the ethmoidnerves that follow the small branches of the anteriormeningeal artery. The meningeus medius nerve is abranch of V2 and runs with the anterior branch ofthe middle meningeal artery to innervate the middlefossa dura. The nervus spinosus, a branch of V3, fol-lows the middle meningeal artery outside the craniumand through the foramen spinosum to innervate themiddle fossa dura. The tentorial nerves arise from theintracranial portions of V1 and course into the duraof the parieto-occipital region and the posterior thirdof the falx, where there is a converging and bilaterallyoverlapping innervation at its midpoint. The posteriorthird of the falx seems to be the most sensitive regionfor the TCR. In the case described, it is likely the fal-cine distribution of V1 sensory fibers that resulted inthe TCR [1].

The TCR has been reported to occur in 8–18% ofcases involving surgical resection of lesions in the ante-rior and middle cranial base. In these cases however,none of the patients show postoperative neurologicdeficits. Though the TCR does not appear to resultin many major complications, there is evidence thatan intraoperative TCR may be associated with a sig-nificantly worse postoperative hearing function aftervestibular schwannoma surgery, and there is a higheroccurrence of ipsilateral tinnitus postoperatively inpatients that had intraoperative TCR than those thatdid not [4].

TreatmentWhen stimulation of the falx (or other structure)results in the TCR, cessation of the surgical manip-ulation in that area is the first step in correcting thehemodynamic instability. Often this alone terminatesthe reflex and allows restoration of hemodynamic sta-bility [5]. If not, atropine or glycopyrrolate are bothusually effective at blocking the cardiovascular effectsof vagal stimulation [6]. Atropine has a faster onsetand more pronounced heart rate effect and shouldbe used if cessation of surgical stimulation does notimmediately terminate the reflex. Advanced cardiaclife-support measures should be instituted as condi-tionswarrant, keeping inmind that patients in this sce-nario must usually be taken out of pins and returnedto the supine position for effective CPR. This requiressome time, but CPR must be instituted as rapidly aspossible if the bradycardia is preventing effective cir-culation of the resuscitationmedications or if hypoten-sion lasts long enough to pose a risk of neurologicischemia. External or internal cardiac pacing may beeffective in particularly refractory cases of TCR.

While the TCR fatigues with repeated stimulationit nevertheless carries significant risk for recurrence ifleft untreated. Glycopyrrolate and atropine both blockrecurrence of the TCR by blocking the effects of vagalstimulation; glycopyrrolate may be advantageous iftime permits in that the resulting tachycardia is lesssevere. Adequate depth of anesthesia (using volatileagents as opposed to opiates) serves to blunt the sever-ity of the TCR as does correcting physiologic derange-ments such as hypoxia, hypercarbia and acidosis. Ingeneral, routine prophylactic anticholinergic use is notrecommended as refractory tachydysrhythmias maydevelop.

ConclusionIn conclusion, TCR most commonly manifests asbradycardia and hypotension in response to mechani-cal stimulation of any of the branches of the trigeminalnerve. Trigeminocardiac reflex episodes that resolvedo not appear to directly cause longstanding neuro-logic injury but some postoperative hearing deficitsappear to correlate with its occurrence. Ultimately itis vigilant attention to the depth of anesthesia and thecorrection of physiologic derangements with continu-ous communication between the anesthesia and surgi-cal teams during resection that are paramount inman-aging surgical procedures when TCR appears.

34

Case 10. Trigeminocardiac reflex

References1. D. F. Bauer, A. Youkilis, C. Schenck et al.The falcine

trigeminocardiac reflex: case report and review of theliterature. Surg Neurol 2005; 63: 143–8.

2. M. Ostachowicz, J. Burau, A. Sadkiewicz. Theoculocardiac reflex during surgical correction ofstrabismus in local anesthesia. Ophthalmologica 1965;150: 259–62.

3. W. Penfield, F. McNaughton. Dural headache andinnervation of the dura mater. Arch Neurol Psychiatr1940; 44: 43–75.

4. B. Schaller, J. F. Cornelius,H. Prabhakar et al.Thetrigemino-cardiac reflex: an update of the currentknowledge. J Neurosurg Anesthesiol 2009; 21: 187–95.

5. A. Koerbel, A. Gharabaghi, A. Samii et al.Trigeminocardiac reflex during skull base surgery:mechanism and management. Acta Neurochir 2005;147: 727–33.

6. E. F. Meyers, S. A. Tomeldan. Glycopyrrolatecompared with atropine in prevention of theoculocardiac reflex during eye-muscle surgery.Anesthesiology 1979; 51: 350–2.

35


11 Sitting craniotomyMichael D. Maile and George A. Mashour

Sitting craniotomies pose a unique set of problemsfor perioperative care of the neurosurgical patient.Although there are benefits to neurosurgery in thisposition, a number of potentially catastrophic compli-cations may also result.

Case descriptionA 20-year-old morbidly obese male presented forresection of a pineocytoma via a supracerebellarapproach in the sitting position.The patient had a longhistory of headaches that were recently increasing inseverity; over the past year, he had also been experi-encing visual loss with his headaches. The patient pre-sented to the emergency department after having aseizurewith loss of consciousness. A computed tomog-raphy scan demonstrated dilated third and lateral ven-tricles concerning for obstructive hydrocephalus. Thisfinding was further investigated by magnetic reso-nance imaging, which showed a mass lesion in thepineal region. Subsequent open, stereotactic biopsymade the microscopic diagnosis of pineocytoma. Thepatient elected to proceed with resection of the tumor,for which the sitting position would be employed.

After induction of general anesthesia and place-ment of an endotracheal tube, a right-side internaljugular multi-orifice catheter and right radial arterialline were placed. The patient’s head was secured inpins. Care was taken to pad all pressure points. Beforethe head was secured in place, a transesophageal echo(TEE) probe was inserted and an examination wasperformed that ruled out patent foramen ovale. Afterthe patient was in his final position, a precordialDoppler monitor was secured to his chest. PrecordialDoppler wasmonitored continuously and the TEEwasused intermittently. Maintenance anesthesia consistedof isoflurane, fentanyl, and vecuronium. Somatosen-sory evoked potentials were monitored continuouslythroughout the case.The surgery was completed with-out complication, the patient emerged from anesthesia

quickly and the trachea was extubated in the operatingroom. Postoperative examination did not reveal anynew deficits. The patient was then transferred to theneurosurgical intensive care unit for further recovery.He was discharged from the hospital on postoperativeday 4.

DiscussionThe sitting position for craniotomies offers severalsurgical advantages. Exposure to posterior cervicaland posterior fossa structures is improved. Cere-bral venous drainage and cerebrospinal fluid (CSF)drainage is enhanced by the effects of gravity, produc-ing a less “tense” brain and allowing for better surgicalexposure. Pooling of blood in the surgical field is min-imized in the sitting position, potentially improvingoperating conditions and reducing blood loss. It hasalso been suggested that access to the airway and chestwill allow easier patient resuscitation should their con-dition deteriorate [1].

The surgical advantages of the sitting position areoffset by several issues for the surgical and anesthesiateams [2, 3]. The risk of venous air embolism (VAE)is increased because the surgical field is significantlyhigher than the level of the heart. Because of thisrisk, precordial Doppler and/or TEE are often used tomonitor for VAE; a multi-orifice central line may beplaced to aspirate a VAE should one occur.

Hemodynamic instability may complicate surgeryin the sitting position. In addition to the usual vasodi-latation and myocardial depression caused by manyanesthetic agents, the sitting position leads to poolingin the lower extremities, whichmay further exacerbateintraoperative hypotension. The decrease in cerebralperfusion caused by placing the head above the heartmay make hypotension more detrimental. In order tominimize intraoperative hypotension, the patient canbe positioned slowly by incrementally elevating theback and head. Volume loading can decrease the effect

36

Case 11. Sitting craniotomy

of pooling of blood in the lower extremities. Vaso-pressors and inotropes should be available to maintainorgan perfusion throughout the surgery.

The risk of pneumocephalus may be increased inthe sitting position compared with other positions [4].The drainage of CSF and venous blood out of thecranial vault by gravity provides more room for airto enter epidural or dural spaces. This can be com-pounded by othermaneuvers that decreaseCSF, blood,or brain volume such as CSF drainage, administrationof diuretics, hyperventilation, or resection of a masslesion. The use of nitrous oxide is generally avoideddue to the risk of VAE and tension pneumocephalus.

Orthopedic, dermatologic, and peripheral nerveinjuries have been reported in the sitting positionand are likely due to pressures exerted on dependentregions of the body. Care should be taken when posi-tioning the patient and liberal padding of all pres-sure points should be performed. Swelling of thetongue and oropharynx can occur during sitting cran-iotomies. This may be due to neck flexion obstruct-ing venous and lymphatic drainage. Placement of aTEE probe may also contribute to this complication;smaller diameter probes may be utilized if available.If swelling is significant, extubation may have to bedelayed after completion of the surgery. Common per-oneal nerve injuries have been reported and are likelydue to nerve compression and/or stretching of thenerve secondary to flexion of the thigh. Recurrentlaryngeal nerve palsies have also been reported for sit-ting craniotomies and are likely also related to com-pression by the TEE probe and endotracheal tube.

A rare but catastrophic complication of the sittingposition is quadriplegia [5]. Several factors have beenpostulated to contribute to this complication. Flex-ion of the neck may compress the cervical spine andresult in ischemia. A decrease in intraoperative bloodpressure will also decrease perfusion, especially with

the head being elevated above the level of the heart.Monitoring evoked potentials may allow for earlydetection and intervention with spinal ischemia.

ConclusionOverall, the benefits of the sitting position may out-weigh the risk if patients are carefully selected andintraoperativemanagement is conductedwith care [6].It has been suggested that a ventriculo-atrial shunt,pulmonary hypertension, a patent foramen ovale, andsymptomatic cerebral ischemia may be absolute con-traindications to this procedure. Performing surgeryin the sitting position for patients with uncontrolledhypertension, significant chronic obstructive airwaydisease, or at the extremes of age should also be donewith caution. However, there are few data to supportappropriate patient selection.

References1. T. Gale, K. Leslie. Anaesthesia for neurosurgery in the

sitting position. J Clin Neurosci 2004; 11: 693–6.2. J. Matjasko, P. Petrozza,M. Cohen et al. Anesthesia

and surgery in the seated position: analysis of 554cases. Neurosurgery 1985; 17: 695–702.

3. M. Standefer, J. W. Bay, R. Trusso. The sittingposition in neurosurgery: a retrospective analysis of488 cases. Neurosurgery 1984; 14: 649–58.

4. T. Sloan. The incidence, volume, absorption, andtiming of supratentorial pneumocephalus duringposterior fossa neurosurgery conducted in thesitting position. J Neurosurg Anesthesiol 2010; 22:59–66.

5. J. M. Porter, C. Pidgeon, A. J. Cunningham. Thesitting position in neurosurgery: a critical appraisal.Br J Anaesth 1999; 82: 117–28.

6. G. P. Rath, P. K. Bithal, A. Chaturvedi et al.Complications related to positioning in posterior fossacraniectomy. J Clin Neurosci 2007; 14: 520–5.

37


12 Cerebellar hemorrhageGene H. Barnett

Stroke can be either ischemic (due to an interruptionof blood and oxygen to the brain) or hemorrhagic(due to bleeding into or around the brain). Here weexplore the management issues surrounding a hemor-rhagic stroke of the cerebellum – one of the most com-mon sites for intracerebral hemorrhage, and onewhereproper management can have a profound impact onoutcome.

Case descriptionThe patient was a 75-year-old female with a historyof hypertension and end-stage renal disease requir-ing dialysis. Shortly after breakfast, she developed asevere occipital headache. She gradually developeddif-ficulty walking and called 911. By noon, she was trans-ported to a nearby emergency department. Upon pre-sentation she was awake, alert and appeared to beneurologically intact (although gait was not tested).She denied use of antiplatelet or anticoagulationagents. She was afebrile, but with a blood pressure of190/110 mmHg, respiratory rate of 20 breaths perminute and pulse of 90 beats per minute. A computedtomography (CT) scan of the brain revealed a 2.5-cm,midline cerebellar hemorrhage, eccentric (to the left)without hydrocephalus. She was treated with labetaloland hydralazine to decrease systolic blood pressure to�160 mmHg. By 1:00 pm, a request for transfer toa tertiary care facility with neurosurgical service wasmade. While waiting for air transport, the attendingphysician noted that the patient was developing swal-lowing difficulty and respiratory distress. The patientwas then sedated and pharmacologically paralyzed,and the trachea was intubated.

Upon arrival at our facility at 2:15 pm, sheremained sedated and paralyzed with mechanical ven-tilatory support. Laboratory studies from the transfer-ring hospital indicated normal prothrombin and acti-vated partial thromboplastin times, as well as platelet

counts. A repeat CT scan showed enlargement of theclot to 3.5 cm in maximum dimension, yet withoutevidence of obstructive hydrocephalus (Figure 12.1).After telephone consultation with the family, the deci-sion was made to take the patient to surgery on anemergent basis.

In the operating room she was positioned pronewith the head flexed to optimize exposure to the pos-terior fossa. After clipping hair and allowing access fora Frazier burr hole (in the event that an emergencyventriculostomy was required), a midline incision wasmade from just above the inion to the mid-cervicalarea. The paraspinal muscles were split and dissectedoff the occiput and upper cervical spine. Then, a gen-erous craniectomy was made, along with removal ofthe posterior arch of C1. By 3:15 pm the dura wasopened with the cerebellum spontaneously deliver-ing itself due to increased posterior fossa pressure.The cerebellar fissure was microsurgically split and itbecame apparent that the clot was in the left cerebellarhemisphere, herniating to the right, giving it a midlineappearance on imaging. The hemisphere was enteredand the clot evacuated.As is commonwith clot evacua-tion, multiple points of bleeding were identified.Thesewere secured, the dura loosely patched with a dermalallograft material, and a titanium mesh cranioplastymade. After the woundwas closed, she was returned tothe intensive care unit, still intubated and sedated, andmonitored with periodic reduction of sedation withneurologic examinations, as well as intermittent CTimaging.

Examination consistently revealed appropriate,symmetric limb movements and limited cranial nerveexams. Computed tomography scans showed satisfac-tory decompression of the posterior fossa and absenceof hydrocephalus. The following day, sedation waswithdrawn, she underwent dialysis, and was extubated(Figure 12.2). After several days each in the neuro-logic intensive care unit, step-down unit and regular

38

Case 12. Cerebellar hemorrhage

Figure 12.1. Preoperative computedtomography – (a) showing cerebellar clotand (b) absence of hydrocephalus.

Figure 12.2. Postoperative computed tomography – showingresolution of cerebellar clot.

neurologic unit, she was transferred to a rehabilitationfacility. At the time of hospital discharge she wasneurologically intact, except for complaints of‘dizziness’ and imbalance when trying to walk.

DiscussionIntracerebral hemorrhage is most commonly associ-ated with chronic hypertension, amyloid angiopathy,anticoagulation, trauma or underlying pathology suchas tumor or vascular malformation [1]. In this case,chronic hypertensionwas the suspected cause – prefer-entially affecting the deep hemispheric nuclei or cere-bellum.When over-anticoagulation (usually related tocoumadin) is the causative factor, then reversal withfresh frozen plasma, vitamin K, and/or factor VII isdesirable before surgery (or to determine if it willthwart clot progression), although the time for suchcorrection may lead to neurologic worsening.

The role of surgical intervention in supratento-rial intracerebral hemorrhage is controversial, butmost evidence suggests that patients with small (i.e.,�40 mL) or large (�100 mL) clots have limited orno benefit from surgical intervention. The role, if any,of minimally invasive approaches to deep clots usingthrombolysis remains undefined [2]. Many patientswith deep clots may decompress their intraparenchy-mal clot with rupture into the ventricular system.Thiseventmay turn out to be a life-saving blessing, or resultin a cascade of further problems related to obstructiveor communicating hydrocephalus.

The posterior fossa has substantially less com-partmental volume than the supratentorial space.Here, stable clots �2 cm in maximum cross-sectionaldimension typically do not warrant surgical inter-vention, unless exhibiting radiographic progression,

39


neurologic symptoms, signs of brainstem compres-sion, or hydrocephalus. However, cerebellar clots≥2.5cm almost always require surgical intervention [3]. Incases where symptoms appear to be due to obstructivehydrocephalus, initial (or total) intervention may beplacement of an external ventricular drainage device.In this case, however, the patient had evidence ofbrainstem dysfunction (swallowing and respiratorydysfunction) without overt hydrocephalus (along withenlarging clot �2.5 cm in size) so emergent surgerywas warranted.

Although the midline posterior fossa can beaccessed with the patient in either a sitting or proneposition, the latter is usually preferred in urgent oremergent situations. As ventricular obstruction mayoccur when the patient is positioned, prepared oropened, allowing access for an emergency externalventricular drainage device is desirable in preparingand draping the patient [4]. Surgery itself is rela-tively straightforward. As noted above, often multi-ple sources of bleeding are encountered at the time ofclot decompression making it difficult to determine adefinitive source of the original hemorrhage. At times,an unexpected source of bleeding such as vascularmal-formation, aneurysm, or tumor may be encountered.The surgeon will need to use his or her experience todetermine how tomanage such findingswhen encoun-tered. Appropriate blood products should, ideally, beavailable to deal with such unexpected findings, yetnot unduly delay emergent surgery. Life-saving clotdecompression and hemostasis generally take prece-dence over vascular anomaly correction or maximaltumor resection. Staging a second, definitive proce-dure under more controlled, studied conditions (i.e.,with complete imaging, navigation, etc.) is perfectlyappropriate under such circumstances. The primarygoal of surgery is to save the patient’s best-quality life,if possible.

Closure should consider whether some additionalcompartmental volume is desirable. In the supraten-torial space, this extra capacity may be accomplishedby partial lobectomy or hemicraniectomy. Infraten-torially, leaving the dura open or a capacious dural

graft and craniectomy are means to that end. Here, weplaced a graft with excess capacity, and used a titaniummesh cranioplasty that provided both external protec-tion, alongwith some extra intracranial capacity as thiswas an onlay, as opposed to a full thickness cranialrepair.

In the end, an expeditious approach to treating thispatient led to a good, but not great early recovery –however, this early result needs to be compared withthe alternative of virtually certain severe disability ordeath.

ConclusionSimilar to ischemic stroke, “time is brain” for manyhemorrhagic stroke patients [5]. In deciding if or whento perform surgery, the neurosurgeon must weigh thelikelihood of the surgical benefit for a good recov-ery versus a futile intervention or sustaining a new,severely disabled life. In this case, many factors alignedto provide the optimal outcome for this patient.

AcknowledgmentsTheauthorwishes to thank the assistance ofMs. Chris-tine Moore for her assistance with the preparation ofthis manuscript.

References1. A. I. Qureshi, S. Tuhrim, J. P. Broderick et al.

Spontaneous intracerebral hemorrhage. N Engl J Med2001; 344: 1450–60.

2. D.W. Miller, G. H. Barnett,D.W. Kormos et al.Stereotactically guided thrombolysis of deep cerebralhemorrhage: preliminary results. Cleve Clin J Med1993; 60: 321–4.

3. S. Kobayashi, A. Sato, Y. Kageyama et al. Treatmentof hypertensive cerebellar hemorrhage – surgical orconservative management? Neurosurgery 1994; 34:246–50.

4. E. S. Connolly, G. M. McKhann, J. Huang, T. F.Choudhri. Fundamentals of Operative Techniques inNeurosurgery. New York: Thieme, 2002; 628.

5. J. L. Saver. Time is brain-quantified. Stroke 2006; 37:263–6.

40

Part II Vascular procedures. Aneurysm clippingCase

13 Preoperative evaluationMilad Sharifpour and Paul S. Moor

Rupture of an intracranial aneurysm is the lead-ing cause of nontraumatic subarachnoid hemorrhage(SAH) and accounts for 80% of the cases, with a highrate of death and complications [1]. Subarachnoidhemorrhage accounts for 2–5% of all new strokes andaffects approximately 30 000 persons each year in theUSA [2]. The most typical manifestation is suddenonset, severe headache that is described as “the worstheadache of my life,” with or without nausea, vomit-ing, and focal neurologic deficits.

Case descriptionThe patient was a 45-year-old female who had pre-sented to an outside hospital with a 1-month historyof progressive right-sided facial and body numbnessthat had worsened acutely over the week prior to heradmission. She was on home coumadin therapy fora history of upper extremity deep venous thrombo-sis. At the time of presentation she did not mani-fest any focal neurologic deficits and was awake, alert,and able to follow commands. Computed tomogra-phy (CT) scans with and without contrast demon-strated intraparenchymal hemorrhage centered in themidbrain in addition to periventricular white matterdisease. Five days after the initial presentation, thepatient became acutely somnolent. The trachea wasintubated and she underwent a noncontrast CT scan ofthe head. In addition to the preexisting intraparenchy-mal hemorrhage, a new SAH with extensive cisternalblood and fourth ventricular compressionwas demon-strated. Based on the clinical exam this was classi-fied as a Hunt–Hess grade III SAH (Table 13.1). Aventriculostomy catheter was placed and the patient’sneurologic status improved subsequently. A radialartery catheter was replaced and large-bore intra-venous access was obtained. At that time she was ableto open her eyes spontaneously and answer yes and noappropriately to questions. She subsequently under-went a diagnostic angiogram that demonstrated a basi-

Table 13.1. Hunt–Hess classification.

Grade Characteristics

I Asymptomatic or minimal headache and slightnuchal rigidity

II Moderate to severe headache, nuchal rigidity, noneurologic deficit other than cranial nerve palsy

III Drowsiness, confusion, mild focal deficit

IV Stupor, moderate to severe hemiparesis, possiblyearly decerebrate rigidity, vegetative disturbances

V Deep coma, decerebrate rigidity, moribund

lar trunk aneurysm. The trachea remained intubatedfor airway protection during transfer under propo-fol sedation to our neurosurgical intensive care unit(ICU), for physiologic optimization before undergo-ingmicrovascular clipping.The patient’s systolic bloodpressure was maintained �140 mmHg and her meanarterial pressure was kept �80 mmHg in the contextof an unsecured aneurysm. The external ventriculardrain was set to drain at 20 cmH2O above the level ofthe tragus.The patient wasmaintained normovolemic,using 0.9%normal saline. Cerebral vasospasmprophy-laxis was initiated with nimodipine and magnesiumsulfate. Levetiracetam was initiated for seizure pro-phylaxis. Sequential compressive devices were used fordeep venous thrombosis prophylaxis as the patient wasnot suitable for anticoagulation. The patient under-went definitive correction of the aneurysm the follow-ing day.

DiscussionAneurysmal SAH is a neurologic emergency, resultingfrom blood extravasation into the subarachnoid spacenormally filled with cerebrospinal fluid (CSF), thatrequires complex treatment andmonitoring. Fifteen to25% of patients die before reaching medical care andthere is an approximately 50% mortality rate associ-ated with SAH. Of those who survive, up to one thirdrequire life-long care and asmany as 46%have residual

41

II. Vascular procedures. Aneurysm clipping

cognitive deficits affecting their functional status andquality of life [3]. Subarachnoid hemorrhage is morecommon in women than inmen (3:2 ratio) and it is 2×more common among African–Americans. It happensmore frequently in persons �40 years old with peakrupture rates between 50 and 60 years. The main riskfactors for SAH include hypertension, tobacco smok-ing, first-degree relatives with SAH, and heavy alco-hol consumption. Furthermore, aneurysms are asso-ciated with heritable connective tissue disorders suchas polycystic kidney disease, Ehlers–Danlos syndrome(type IV), Marfans syndrome, pseudoxanthoma elas-ticum, and fibromuscular dysplasia. Risk of rupturevaries with size and location of the aneurysm. Themost common sites of ruptured aneurysms are (1)anterior cerebral artery, (2) posterior communicatingartery, and (3) middle cerebral artery. The most com-mon manifestations include sudden onset of severeheadache, with nausea, vomiting, neck pain, photo-phobia, loss of consciousness, and prolonged coma.Meningismus, focal neurologic deficits, third and sixthcranial nerve palsies, and altered level of consciousnesscan be present on physical examination.

The initial management of a patient with SAHshould focus on (1) airway management dependingon neurologic function, (2) avoiding hypertensionthat can lead to rupture of unsecured aneurysms orrebleeding, (3) maintenance of adequate cerebral per-fusion pressure (CPP) to minimize the risk of cere-bral ischemia in recently insulted brain with alteredautoregulation, (4) determining the degree of neuro-logic severity, as well as serial assessment of neuro-logic function using the Hunt–Hess scale (Table 13.1),(5) detection of rebleeding or hydrocephalus, and(6) provision of seizure prophylaxis. Prevention, detec-tion, and treatment of cerebral vasospasm are impera-tive in patients presenting between post-rupture day4–14, the risk period for vasospasm and delayedischemic neurologic deficit.

A noncontrast head CT scan should be the firstdiagnostic study performed in patients presentingwithSAH symptoms as this detects more than 95% of SAHsif performed within the first 24 hours. The presenceof xanthochromia on lumbar puncture helps diag-nose SAH in patients with strong suspicion of SAHand negative or equivocal head CT. Meanwhile, cere-bral angiography is the gold standard for detectionof intracranial aneurysms and should be performedpromptly once initial stabilization is achieved to facil-itate early repair of the ruptured aneurysms.

Since rebleeding is associated with increased mor-tality and the risk of rebleeding is highest immedi-ately after SAH, early intervention within the first 72hours is recommended for patients with neurologicgrades I–III. Definitive management of aneurysmalSAH includes endovascular coiling or microvascu-lar clipping. According to the International Subarach-noid Aneurysm Trial, in patients who are equally suit-able for endovascular coiling or surgical clipping, atone year, endovascular coiling is associated with sig-nificantly lower dependency or death compared withsurgical clipping [4]. However, endovascular coilingis associated with a higher rate of late rebleeding.Other factors, including patient’s age and other medi-cal co-morbidities, aneurysmal properties such as sizeand neck, accessibility, and relationship to adjacentvessels should be considered while choosing the mostappropriate treatment option.

In addition to a complete preoperative history andphysical examination, evaluation of a patientwith SAHshould focus on assessing the presence and extent ofthe commonly associated physiologic derangements.Early complications, most often occurring within thefirst 72 hours, include (1) rebleeding, hydrocephalus,and elevated intracranial pressure (ICP), (2) ele-vated blood pressure in response to pain, anxiety,and sympathetic activation, (3) cardiac abnormali-ties including electrocardiographic changes, arrhyth-mias, myocardial injury, and stunned myocardiumin response to SAH-associated catecholamine release,and (4) pulmonary complications including cardio-genic and neurogenic pulmonary edema, aspiration,and pneumonia. Late complications include (1) cere-bral vasospasm, (2) intravascular volume contrac-tion, (3) hyponatremia, and (4) shunt dependence.Hyperglycemia, anemia, fever, and hyperthermia areother complications that frequently occur in SAHpatients and are associated with increased mortalityand poor functional outcome (Table 13.2). In addi-tion, systemic hypotension may occur in responseto nimodipine therapy for vasospasm prophylaxis.Such derangements require thorough preoperativeoptimization, to help minimize their impact duringanesthesia.

Patients with a ruptured aneurysm are at the high-est risk of rebleeding within the first 24 hours fol-lowing the initial rupture and this risk decreases overthe next few days. Pain, anxiety, and SAH-associatedcatecholamine release can lead to elevated bloodpressure, which increases the risk of rebleeding in

42


Table 13.2. Systematic approach to subarachnoidhemorrhage.

System Comments

Nervoussystem

↑ ICP, hydrocephalus: can be managed w/external ventricular drain

Rebleeding; risk ↓ by early coiling or clipping.Blood pressure control with IV betablockers for unsecured aneurysms

Cerebral vasospasm: prophylaxis withnimodipine and magnesium sulfate. Tx:Permissive HTN and hypervolemia.Intra-arterial vasodilator injection andtransluminal angioplasty for refractory cases

Seizure: consider anticonvulsant therapydepending on degree of bleed

Cardiovascular Electrocardiographic abnormalities: themajority are benign and require notreatment

Arrhythmias: �5% are life threatening andrequire management according to advancecardiac life support guidelines

Stunned myocardium: useful to evaluateventricular function with transthoracicechocardiogram. May requireinotropic/pressor support, diuresis

Respiratory Cardiogenic or neurogenic edema, aspiration,pneumonia: Tx: ventilatory support, diuresisfor pulmonary edema, antibiotics andinfection management for pneumonia

Airway compromise in patients with alteredmental status: prompt intubation

Endocrine Hyponatremia secondary to cerebral saltwasting or syndrome of inappropriateantidiuretic hormone; Tx: salt replacementwith hypertonic saline

Hypomagnesemia, hypokalemia,hypocalcemia

Hyperglycemia is associated with worseoutcomes; maintain normoglycemia withsliding scale insulin

Volume status Volume contraction secondary to decreasedintake and cerebral salt wasting; maintainnormovolemia with 0.9% normal saline

Prophylaxis/prevention

Gastrointestinal: proton pump inhibitorsVenous thrombosis: compression devices, as

anticoagulation not deemed suitable forpatients with a recent intracranial bleedand unsecured aneurysm

Infectious disease: antibiotic prophylaxis forventricular drain. Protocols for preventionof ventilator-associated pneumonia

ICP = intracranial pressure; Tx = therapy; HTN = hypertension.

unsecured aneurysms. Analgesics and antihyperten-sive infusions are frequently required to achieveadequate blood pressure control. Beta-blockers andlabetalol are preferred since they do not induce cere-bral vasodilation and do not further increase ICP.

Alternatively, nicardipine can be used to achieveblood pressure control. However, in the presence ofhydrocephalus, aggressive management of blood pres-sure should be addressed only after hydrocephalusis treated. Twenty to 30% of the patients experi-ence hydrocephalus within the first three days post-rupture; incidence is increased in higher grade SAH asa result of ventricular blood load, foraminal obstruc-tion, increased CSF viscosity, and impaired CSF flow.Elevated ICP as a result of communicating hydro-cephalus can be managed with an extraventriculardrain.

Electrocardiographic abnormalities such as peakedT-waves, QT prolongation, and S-T segment elevationor depression are present in up to 70% of the SAHpatients and these are often benign and do not requirespecific treatment. Rhythm disturbances are also com-mon and up to 5% of the patients experience life-threatening arrhythmias, which should be managedaccording to advanced cardiac life-support guidelines.Stunned myocardium is the most severe cardiac com-plication associated with SAH. It can severely impairleft ventricular function and decrease cardiac out-put and mean arterial pressure, resulting in decreasedCPP and increased risk of ischemic cerebral insult aswell as cardiogenic pulmonary edema. Managementrequires inotropic support, diuresis, and ventila-tory support. Elevated cardiac enzymes and troponinare also frequently encountered laboratory values inSAH patients and may be associated with worseprognosis.

Patients with neurologic insults are at increasedrisk for pulmonary complications such as aspiration,pneumonia, and pulmonary edema, and may requiremechanical ventilation. It is important to normalizeand maintain PaO2, PaCO2, and pH in order to pre-vent increased ICP.

Patients are at risk for cerebral vasospasm frompost-rupture days 4–14. About two-thirds of patientsdemonstrate angiographic evidence of vasospasm andup to 50% develop symptomatic vasospasm with clin-ical symptoms of cerebral ischemia, such as newonset focal neurologic deficits and altered mental sta-tus. The amount of blood in the subarachnoid spaceseen on the initial CT scan is the strongest predic-tor of vasospasm. Transcranial Doppler ultrasonogra-phy is the most frequently used noninvasive methodto detect vasospasm while angiography remains thegold standard. Once clinical cerebral vasospasm isdetected, medical therapy with induced hypertension,

43


hypervolemia, and hemodilution (“triple-H” ther-apy) is often initiated. While induced hypertensionand hypervolemia are effective and well establishedin the treatment of cerebral vasospasm, the utilityof hemodilution remains controversial. Transluminalangioplasty and intra-arterial infusion of vasodila-tors should be reserved for patients whose vasospasmremains refractory to medical therapy.

Patients with SAH are at increased risk of intravas-cular volume contraction in the setting of cerebralsalt wasting syndrome and decreased oral intake. Vol-ume status should be frequently monitored and ade-quate volume replacement with 0.9% normal saline isrequired to prevent volume contraction.

Hyponatremia occurs in up to one-third of SAHpatients and is secondary to syndrome of inappro-priate antidiuretic hormone secretion, or cerebral saltwasting. Both syndrome of inappropriate antidiuretichormone secretion and salt wasting can be managedwith sodium and replacement with intravenous hyper-tonic saline. Fluid restriction is often avoided becauseof cerebral vasospasm risk. Other common electrolyteabnormalities that should be suspected and treatedinclude hypokalemia and hypocalcemia.

While 10% of patients are at risk of seizuresfollowing SAH, the need for the routine adminis-tration of anticonvulsants remains unclear [2]. Thecerebral metabolic rate increases in a seizing brain;seizure prophylaxis therefore helps prevent furtherischemic injury in susceptible patients. Fever andhyperglycemia are also common in SAH patients andare associated with increased mortality and poor out-

come, therefore therapy should be directed towardsmaintaining normothermia and normoglycemia [3].

ConclusionIn conclusion, SAH is a complex disease with highmorbidity and mortality – management of patientswith SAH requires a multisystem approach. Patientspresent for elective clipping of an unrupturedaneurysm or emergent surgery following SAH. Prog-nosis depends on the severity of the initial bleed,prompt securing of the aneurysm, and the incidenceand the extent of neurologic and systemic complica-tions associated with SAH. Thorough assessment ofthe patient, effective organ support, and correction ofpathophysiology are vital prior to leaving the ICU forwhat may be a challenging case in the operating room.

References1. J. I. Suarez, R. W. Tarr,W. R. Selman. Aneurysmal

subarachnoid hemorrhage. N Engl J Med 2006; 354:387–96.

2. M. N. Diringer. Management of aneurysmalsubarachnoid hemorrhage. Crit Care Med 2009; 37:432–40.

3. M. Smith. Intensive care management of patients withsubarachnoid haemorrhage. Curr Opin Anaesthesiol2007; 20: 400–7.

4. A. Molyneux, R. Kerr, I. Stratton et al. InternationalSubarachnoid Aneurysm Trial (ISAT) of neurosurgicalclipping versus endovascular coiling in 2143 patientswith ruptured intracranial aneurysms: a randomisedtrial. Lancet 2002; 360: 1267–74.

44


14 Intracranial aneurysm clipping withintraoperative rupturePaul S. Moor

Cerebral aneurysm surgery is challenging for bothsurgeons and anesthesiologists, requiring close col-laboration among all operating room team members.Although endovascular treatment for aneurysmalsubarachnoid hemorrhage (SAH) has increased,operative management by clipping is likely to remainfor the more distal, wide-necked aneurysms, or forthose where coiling has failed. The process, however,has a risk of intraoperative rupture, which remainsa potentially catastrophic event, adversely affectingoutcome [1].

Case descriptionThe patient was a 56-year-old African–Americanfemale smoker who presented with sudden onset offrontal headache, vomiting, and neck stiffness. Non-contrast head computed tomography (CT) revealed aright-sided SAH with minimal intraventricular bloodand no obvious hydrocephalus. Her clinical conditionrepresented a Hunt–Hess grade II SAH and she wasadmitted to the neurosurgical intensive care unit forobservation. Cerebral angiography on day 1 demon-strated a large, wide-necked, anterior communicat-ing artery aneurysm; craniotomy and clipping werearranged for later that day. Electrophysiologic moni-toring with somatosensory evoked potentials and elec-troencephalography was arranged.

Anesthetic history and physical examination wereobtained. Her past medical history was significant forcontrolled hypertension, treated with amlodipine, andobesity (body mass index = 35). Airway assessmentwas predictive of an easy intubation, the patient’s car-diopulmonary status was normal and SAH inducedsympathetic discharge did not appear to be signifi-cant. Nimodipine andmagnesium vasospasm prophy-laxis were commenced. Laboratory test values wereall within normal limits, troponins were negative andcross-match for two units was performed.

The primary concerns of the anesthesiology teamwere (1) precise control of the aneurysmal transmu-ral pressure gradient, (2) preservation of cerebral per-fusion pressure (CPP) and oxygen delivery to thebrain parenchyma to minimize risk of ischemia inrecently insulted brain, (3) avoidance of large swingsin intracranial pressure (ICP), (4) provision of optimalconditions for surgical exposure and minimal brainretraction, (5) anticipation and preparation for intra-operative aneurysm rupture and rapid blood loss, and(6) swift emergence from anesthesia to facilitate post-operative neurologic assessment.

A 14 g intravenous cannula was inserted in the pre-operative holding area, prior to obtaining radial arte-rial access under midazolam anxiolysis. The patientwas transferred to the operating room and furtherstandard monitoring was applied. After preoperativeverification, general anesthesia was induced carefullywith lidocaine, fentanyl, and propofol. Paralysis wasachieved with vecuronium. Direct laryngoscopy wasperformed after an esmolol bolus to ensure that sys-tolic blood pressure remained �150 mmHg. The tra-chea was intubated and two further large-bore intra-venous cannulae were placed. Further propofol wasadministered prior to application of the Mayfield skullclamp and end-tidal isoflurane was maintained at�1.0 minimum alveolar concentration, in addition toremifentanil and vecuronium infusions. Neurophysio-logic monitoring was of satisfactory quality. Normov-olemia was maintained with normal saline and albu-min, with a systolic pressure variation between 6–8.On raising of the bone flap, the surgeons reported ade-quate brain relaxation as a result of moderate hyper-ventilation (PaCO2 30 mmHg) and 0.5 g/kg mannitol.

Anesthesia was unremarkable until the surgeonswere preparing to place the temporary clip andmicrodissect the aneurysmal neck from the parentartery. Aneurysmal rupture and overwhelming hem-orrhage occurred into the surgical site. Assistance wassummoned and inspired oxygen increased to 100%.

45


Thiopental (3 mg/kg) was administered to induceburst suppression and lower blood pressure. A sec-ond suction device was utilized by the surgeons; sys-tolic blood pressure of 60 mmHg and burst suppres-sion was maintained with thiopental boluses. A bolusdose of adenosine induced transient circulatory arrestand allowed proximal artery identification and tempo-rary clip placement.

DiscussionAnesthesia for patients with SAH is challenging. Themaintenance of an adequate mean arterial blood pres-sure, and hence CPP, during the induction of anes-thesia is key to prevent ischemic secondary injury.Aneurysmal rebleeding is possible if blood pressuresurges from anesthetic and surgical stimuli that arenot mitigated. The provision of an adequate CPP,while ensuring the transmural pressure gradient doesnot increase the risk of rebleeding, epitomizes thechallenge.

Cerebral perfusion pressure and transmural pres-sure are essentially influenced by the same vari-ables and are equal to the mean arterial pressureminus ICP [2]. The subtle difference between bothvalues is that CPP represents the pressure drivingcerebral blood flow and transmural pressure is thepressure acting across the aneurysmal wall, whichis also influenced by local factors like clot organi-zation, aneurysm size, and Laplace’s law. The key isto keep CPP physiological while minimizing trans-mural pressure, as any increase in blood pressure orreduction in ICP increases the likelihood of aneurys-mal rupture and rebleeding. Transmural pressure is atrisk of increasing during direct laryngoscopy, appli-cation of Mayfield pins, incision and raising the boneflap.

Aneurysm rupture during laryngoscopy is anuncommon but life-threatening complication, whichshould be suspected if severe hypertension and brady-cardia develop. Urgent CT and medical managementof intracranial hypertension should commence, whilefuture surgical options are being considered.

Historically, induced hypotension was advocatedduring aneurysm dissection due to the high risk ofaneurysmal rupture. With the advent of temporaryclipping, deliberate hypotension is now reserved forintraoperative rupture only. These cases require nor-motension as baseline. Amean arterial pressure of 60–80 mmHg is adequate.

Brain relaxation is essential to improve surgicalaccess, reduce brain retraction, and assist in the clipapplication; it becomes more important in the highclinical grade SAH when intracranial complianceis reduced. This process demands a careful balancebetween the beneficial effect of brain volume reduc-tion and the risk of brain ischemia. Methods toreduce the volume of intracranial contents such ashyperventilation, hyperosmolar agents and possiblecerebrospinal drainage should be discussed with thesurgeons as they may adversely affect transmuralpressure. With mild hyperventilation to a PaCO2 of30 mmHg, use of mannitol osmotherapy 0.25–1 g/kgand/or furosemide, adequate relaxation can often beachieved.

Temperature control throughout aneurysmsurgery has been controversial. Previously, hypother-mia was induced to neuroprotect during surgery. TheIntraoperative Hypothermia for Aneurysm SurgeryTrial [3] found no benefit with induced hypothermia,although passive cooling is acceptable as long as it doesnot prevent an appropriate extubation. Hyperthermiashould be avoided as it increases cerebral blood flowand oxygen requirements. As it is not uncommonfor patients to be well insulated by surgical drapes,vigilance is required.

Temporary arterial occlusion of the aneurysm’sparent vessel is now an integral part of cerebralaneurysm surgery, preventing aneurysmal rupturewhile aiding dissection, clipping, and vessel recon-struction [4]. The safe duration is considered to bein the range of 10–15 minutes. Although not sup-ported by randomized controlled trials, many centersuse neurophysiologicmonitoring to assist in the detec-tion of regional ischemia associated with clip place-ment. Pre-occlusion burst suppression can be achievedwith a slow bolus of thiopental 3–4 mg/kg and subse-quent infusion of 3–4 mg/kg/hour, titrating to effectthroughout occlusion. Vasopressor infusion can alsobe administered during temporary clipping to offsetbarbiturate administration and to improve cerebralcollateral blood flow.

Significant intraoperative aneurysmal rupture canbe catastrophic as the brain receives approximately15% of the cardiac output and accounts for approxi-mately 20% of basal oxygen consumption. Intraoper-ative rupture can be defined as aneurysmal bleedingthat interrupts the normal sequence of the aneurysmmicrosurgery [1] and is considered significant if theresultant hemorrhage is not cleared by a double

46

Case 14. Intracranial aneurysm clipping with intraoperative rupture

Call for help & PRBC

100% Oxygen

Decrease BP and CMRO2(Propofol or Sodium Thiopental do both )

Interruption of bleeding(Long enough to establish surgical control )

Ipsilateral carotid artery compression Circulatory arrest with adenosine

Resuscitation to restore circulating volume

Figure 14.1. Intraoperative ruptureimmediate action drill. BP, blood pressure;CMRO2, cerebral metabolic rate; PRBC,packed red blood cells.

suction technique.The incidence of significant ruptureis believed to be 3–4% [5], the majority occurring attime of aneurysm manipulation and dissection.

The immediate action drill for this potentiallycatastrophic event aims to allow surgical applicationof a temporary clip (Figure 14.1). The goals are toreduce cerebral metabolic rate, while decreasing theblood pressure to allow for proximal control of theaneurysm. The proximity and routine use of thiopen-tal makes it a readily available agent for both purposes,but other drugs can be used. Associated hypotension isoften sufficient to negate the requirement for a specificantihypertensive agent. Ipsilateral carotid compres-sion and transient circulatory arrest with adenosine12mgmay be utilized. On application of the clip, stan-dard resuscitation of the intravascular compartment isessential.

ConclusionIn conclusion, intraoperative management of SAHis high-risk anesthesia with the potential for severeconsequences. Constant vigilance regarding hemody-

namic control and preparedness for the possibilityof intraoperative aneurysmal rupture are essential forgood outcomes.

References1. J. P. Chandler, C. C. Getch,H. H. Batjer.

Intraoperative aneurysm rupture and complicationavoidance. Neurosurg Clin N Am 1998; 9: 861–8.

2. H. J. Priebe. Aneurysmal subarachnoid haemorrhageand the anaesthetist. Br J Anaesth 2007; 99: 102–18.

3. M.M. Todd, B. J. Hindman,W. R. Clark et al.Mildintraoperative hypothermia during surgery forintracranial aneurysm. N Engl J Med 2005; 352:135–45.

4. C. L. Taylor,W. R. Selman, S. P. Kiefer et al.Temporary vessel occlusion during intracranialaneurysm repair. Neurosurgery 1996; 39: 893–905;discussion 905–6.

5. T. J. Leipzig, J. Morgan, T. G. Horner et al. Analysis ofintraoperative rupture in the surgical treatment of1694 saccular aneurysms. Neurosurgery 2005; 56:455–68; discussion 455–68.

47


15 Awake fiberoptic intubationHeather Hervey-Jumper and Christopher R. Turner

Patients who require fiberoptic-guided endotrachealintubation for the clipping or coiling of an intracra-nial aneurysm pose particular challenges for the safecompletion of both procedures.The following case dis-cussion highlights the considerations for awake endo-tracheal intubation in the patient with an unsecuredaneurysm.

Case descriptionA 56-year-old female with a poorly documented his-tory of “difficult intubation” presented for electiveclipping of a middle cerebral artery aneurysm. Herpast medical history was also significant for asthmaand hypertension; prior treatment of her asthma withalbuterol induced atrial fibrillation. Airway exami-nation in the preoperative area revealed a Mallam-pati Class III oropharynx, mouth opening �3 cm,thyromental distance �6 cm, good mandibular sub-luxation, thick neck, slight limitation of the cervicalspine and normal dentition. Given her further his-tory of severe gastroesophageal reflux and hiatal her-nia, the anesthesia team was concerned about therisk of aspiration with prolonged mask ventilation inthe event that direct or indirect laryngoscopy proveddifficult. A calm and clear discussion ensued withthe patient, who subsequently agreed to an awakeintubation. After placement of an intravenous andradial artery catheters, the patient was pretreatedwith glycopyrrolate as an antisialogogue, as well asranitidine and sodium citrate. Twenty minutes later,nebulized lidocaine was started and the patient wasthen brought to the operating room. Standard mon-itors were applied and a low-dose remifentanil infu-sion (0.01 mcg/kg/min) was started. A “swish-and-swallow” technique of viscous lidocaine was per-formed, followed by atomized lidocaine sprayed in theposterior pharynx and directed at the vocal cords. Vig-ilance with respect to the blood pressure was main-tained consistently. An oral airway was placed and

a flexible fiberoptic scope was passed into the pos-terior pharynx. The epiglottis then vocal cords werevisualized – further lidocaine was sprayed onto thecords through an epidural catheter fed through thesuction port of the scope. After a 30-second pause,the fiberoptic scope was inserted through the cordsand into the trachea. Labetolol was administered fora systolic blood pressure of 165 mmHg, remifentanilwas increased, and the endotracheal tube was passedafter a pause and resolution of the hypertension. Gen-eral anesthesia was then induced with propofol. Aftera brief period of bronchospasm that resolved with anincrease of the inhalational anesthetic concentration,the case proceeded uneventfully and the aneurysmwassecured.The patient was fully awake for tracheal extu-bation at the end of the case and recovered well.

DiscussionThe indications for awake fiberoptic intubation(AFOI) in this case are essentially the same as for anydifficult airway: concern for the ability to visualizethe glottic opening via direct laryngoscopy combinedwith concern for the ability to mask ventilate. In apatient with an intracranial aneurysm, the additionalconcerns are ensuring pristine hemodynamic controlin order tominimize the risk of aneurysmal rupture orre-rupture, which carries a significant risk of mortalityor major morbidity [1], while maintaining sufficientblood pressure to ensure cerebral perfusion. Hemo-dynamic derangements may arise from two sources:the hemodynamic response of a poorly preparedpatient to the stimulation of an awake intubationthat is possibly premature, and the hemodynamicresponse to hypoxia and/or hypercarbia as a conse-quence of a failed or mismanaged awake intubationattempt. Hypertension is frequently a preexistingcondition in patients with cerebral aneurysms, whichis usually worsened with anxiety or the stimulationof an awake intubation. Thus, the ability to rapidly

48

Case 15. Awake fiberoptic intubation

recognize and respond to changes in heart rate andblood pressure is critical to the safe completion of anawake intubation in a patient at risk for aneurysmalrupture. While an arterial catheter should be placedwith the same concerns as apply to the intubation, i.e.,hemodynamic control, it is easier to do so with min-imal stimulation to the patient than it is to performan AFOI.

Proper preparation of a patient for AFOI occursin four overlapping steps. Psychological preparationandmanagement of expectations should not be under-estimated or underemphasized. It is critical that thepatient understands the procedure about to occur,appreciates that they will be properly sedated and top-icalized, and knows that they have the ability to pausethe procedure if they are not tolerating it. Drying ofthe mucous membranes of the airway can be accom-plished by the early administration of glycopyrrolate –again, the patient should be made aware of the conse-quent increase in heart rate and drying of the mouth.An antisialogogue serves both to increase the effective-ness of subsequent topical anesthetics and to decreaseairway secretions that can obscure the fiberoptic view.Next comes the administration of topical anesthesiain order to anesthetize the glossopharyngeal inner-vation of the posterior pharynx, as well as the supe-rior and recurrent laryngeal nerve branches of thevagus that provide sensory innervation to the supe-rior and inferior aspects of the glottis, respectively.This can be done with a swish-and-swallow technique,staged administration of topical lidocaine via progres-sively deeper use of an atomizer, or administrationof lidocaine via a hand-held nebulizer. Use of airwaynerve blocks (glossopharyngeal, superior laryngeal,and transtracheal) is also a viable option, althoughthey are more stimulating for the patient and prob-ably should be reserved for situations where a needfor rapid intubation is relatively more important thanthe desire for the least stimulating topicalization pos-sible.While the airway is being topicalized, the patientshould also be judiciously sedated. In general we havefound that single-agent techniques are more effectivethan combining agents; in particular benzodiazepinesaugment amnesia but do little if anything to blunt thestimulation of an awake intubation. Combined agentsalso often act synergistically to suppress respiratorydrive. With enough time, intravenous fentanyl can beeffectively and safely titrated to an endpoint of a nar-cotized patient who is still breathing spontaneouslyand is oriented and cooperative. If time is short many

providers find that using remifentanil has advantagesover other agents in its rapid onset and recovery. Thisopioid may provide the ideal narcotization needed tofacilitate intubation and the speedy recovery neededif airway compromise becomes an issue. Dexmedeto-midine is an alpha-2 adrenergic agonist with signif-icant sedative and analgesic properties. Dexmedeto-midine, while slow in onset and more expensive thanother agents, has the unique characteristics of pro-viding sedation and analgesia while preserving res-piratory drive and orientation. In the setting of anunsecured aneurysm, its sympatholytic effects are alsobeneficial. As further adjuncts, adrenergic blockadewith esmolol or labetalol may also be useful to ensureblood pressure control without further respiratorydepression, and one of these should be immediatelyavailable whenever an AFOI is attempted in a patientat risk for aneurysmal rupture. Supplemental oxy-gen via nasal cannula often increases the safety ofthe procedure but does little to lengthen the time tofrank desaturation in the face of significant respiratorydepression.

As the AFOI attempt is begun, one provider shouldbe designated whose major responsibility is to mon-itor the patient’s blood pressure. The blood pressureshould not be allowed to rise above the preoperativebaseline (or in our institution, a maximum systolicpressure of about 150 mmHg) and generally should bekept 20% below baseline in order to provide a mar-gin of safety. If the patient is not tolerating the proce-dure by demonstrating a hypertensive response, it isnecessary to pause to administer supplemental seda-tion and/or topicalization to prevent rapid or uncon-trolled changes in blood pressure. Deep topicalizationcan often be supplemented via a catheter threadedthrough the suction port of the fiberoptic scope [2].With this technique, local anesthetic can be directedat the tissue that is inadequately topicalized (often theglottic opening or the trachea itself), but it is neces-sary to remember that the onset of topical anesthe-sia requires at least 30–60 seconds. After the fiberopticscope enters the trachea, the endotracheal tube shouldbe advanced gently, again avoiding unnecessary reac-tion to the procedure by the patient. Stimulation of thecarina by either the fiberoptic scope or the endotra-cheal tube should be assiduously avoided as the carinais commonly inadequately anesthetized and stimula-tion of the carina causes an intense reaction in thepatient. Once the intubation is complete the patientshould tolerate the presence of the endotracheal tube

49


enough to permit an abbreviated postintubation neu-rologic examination.

If there is a high probability for successful maskventilation, it is often better to intubate the tracheain a patient with a difficult airway after the induc-tion of anesthesia. This may be particularly attractivein patients whose cooperation with an awake intu-bation is questionable, such as patients with men-tal status changes (Hunt–Hess Grade III or higher).There is a variety of techniques applicable to this sce-nario ranging from lightwands to rigid fiberoptics(e.g., Bullard, Glidescope, or C-Mac) to asleep flexi-ble fiberoptic intubation. However, keep in mind thatan asleep fiberoptic intubation is not necessarily lessstimulating than asleep direct laryngoscopy. Fiberop-tic intubation may affect the hemodynamic responseor catecholamine levels asmuch as direct laryngoscopywhen performed on an anesthetized patient [3]. Thus,even though a fiberoptic intubationmay be performedafter the induction of anesthesia, it does not obviatethe need for vigilance with respect to hemodynamiccontrol.

Intubation via an intubating laryngealmask airway(e.g., Fastrach) or standard laryngeal mask airway andAintree catheter assisted by a fiberoptic scope are alsooptions, but should not be attempted if there is concernof not being able to mask ventilate the patient.

ConclusionIn conclusion, AFOI has the potential to triggerhypertension, tachycardia and, potentially, hypoxia

or hypercarbia. These all pose an increased risk ofrupture or re-rupture in aneursymal disease. Pre-venting the hemodynamic response to intubation inthe awake patient should thus be with a multimodalapproach of psychological preparation, suppression ofthe sympathetic reaction to instrumentation with opi-oids and topicalization, and the immediate availabilityof vasoactive medications to maintain a blood pres-sure within the target range to avoid rupture of theaneurysm. Hypercarbia or hypoxia during an awakefiberoptic intubation are frequently due to loss of res-piratory efforts in a narcotized patient. Newer seda-tion regimens such as dexmedetomidine can providesedation and analgesia while maintaining respiratorydrive and reducing hypertension. In the presence ofcerebrovascular disease where patient cooperation isquestionable, asleep intubations utilizing an appropri-ate airway device may be the preferred option.

References1. L. Elijovich, R. T. Higashida,M. T. Lawton et al.

Predictors and outcomes of intraprocedural rupture inpatients treated for ruptured intracranial aneurysms.Stroke 2008: 39: 1501–6.

2. F. S. Xue,H. P. Liu, N. He et al. Spray-as-you-goairway topical anesthesia in patients with a difficultairway: a randomized, double-blind comparison of 2%and 4% lidocaine. Anesth Analg 2009: 108: 536–43.

3. M. Barak, A. Ziser, A. Greenberg et al.Hemodynamicand catecholamine response to tracheal intubation:direct laryngoscopy compared to fiberoptic intubation.J Clin Anesth 2003: 15: 132–6.

50


16 Patient with coronary artery stentRichard Bowers and George A. Mashour

The increasing use of coronary artery angioplasty withdeployment of stents for treatment of coronary arterydisease poses several dilemmas for perioperative man-agement. These conflicting requirements are mani-festedmost acutely in themanagement of patients withneurovascular disease.

Case descriptionThe patient was a 51-year-old female with a past medi-cal history of ischemic heart disease, hypertension, andundifferentiated autoimmune disease with interstitiallung involvement. She presented with a 5-month his-tory of episodic headache and retro-orbital pain withassociated nausea and vomiting. Radiologic imagingrevealed a saccular aneurysm of the posterior com-municating artery. A decision was made for subse-quent admission and aneurysm clipping on the basisof aneurysm morphology and the patient’s age. Acardiac catheterization 7 months prior to the admis-sion had revealed 80% stenosis of the right coronaryartery. A paclitaxel-eluting stent was deployed to treatthe lesion and the patient was taking regular aspirinand clopidogrel as advised by her cardiologist. Thesemedications had been discontinued 7 days prior tosurgery.

Anesthesia was uneventful and the patient wastransferred to the neurologic intensive care unit post-operatively. Antiplatelet therapy was reintroduced at24 hours post-surgery. No arrhythmia, troponin rise,or ECG changes were recorded.

DiscussionThepresence of coronary stents in patients undergoingneurosurgical procedures warrants specific considera-tion prior to anesthesia. It is necessary to balance therisks of stent thrombosis, and the subsequent risk ofmyocardial infarction, arrhythmia, or cardiac arrest,against the risks of hemorrhage during or after a neu-

rosurgical procedure. There are several factors to beconsidered including:

1. Stent type.2. Duration since cardiac intervention.3. Antithrombotic regimen.4. Urgency of neurosurgical procedure.

Other risk factors for stent thrombosis do exist, forinstance diabetesmellitus, renal failure, and stent type,although the manipulation of these factors in the peri-operative period to reduce the risk of stent thrombosisis limited [1].

There are two available stent types. Drug-elutingstents (DES) are now the commonest type used incoronary artery stenting and are estimated to repre-sent more than 80% of stents deployed in the USA,although this rate has fallen somewhat over concernsfor long-term risk of stent thrombosis [2]. Two sub-types of this stent type are available – sirolimus-elutingandpaclitaxel-eluting.Thegoal of such agents is to pre-vent neointimal hyperplasia and occlusion. However,both of these agents also prevent neoendothelializa-tion of the implanted stent over a variable period.Onceall the drug is eluted from the stent, some degree ofendothelialization may occur, although this has beenreported as variable and incomplete [3]. The alterna-tive is the bare metal stent (BMS). As the name sug-gests, this stent does not contain any antimitotic prop-erties, so that, once implanted, it is covered by a layerof vascular endothelium.

Both stent varieties carry the risks of thrombo-sis and restenosis. Original coronary interventionsusing BMS carried an initial high risk of thrombo-sis, which ameliorated over the 6 weeks followingstent deployment. This risk follows the natural pro-gression of healing in the arterial wall and neoen-dothelialization of the stent. Thereafter the majorrisks encountered with BMS are restenosis, whichmay then require further treatment [1]. Drug-eluting

51


stents (DES) were developed to avoid this late risk ofrestenosis.

Thrombosis risk continues until a stent is endothe-lialized. Consequently, antithrombotic strategies havebeen developed to reduce the risk of stent thrombosis.Current guidelines by the American College of Cardi-ology/American Heart Association (ACC/AHA) rec-ommend the use of dual antiplatelet therapy. Aspirin isrecommended at high doses for up to 6 months post-stent deployment, followed by lifelong aspirin therapyat a lower dose. Use of clopidogrel is recommendedfor aminimumof 1month following BMS deployment[2]. Thrombotic risk is less than 1% if the appropriateregimen is adhered to [1].There is a risk of late throm-bosis due to the presence of antimitotic agents on thepolymer coating of DES, particularly at times whenantiplatelet therapy has been discontinued. The dis-continuation of clopidogrel therapy within 6 monthsof the placement of a DES has been found to be thestrongest independent risk factor for stent thrombosiswithin the subsequent 14 days [4].The same study alsofound that the lack of clopidogrel therapy, though notnecessarily acute discontinuation of therapy, beyond6 months was also a risk factor for stent thrombo-sis. The ACC/AHA guidelines recommend 12 monthsclopidogrel therapy for all types of DES [2]. Mor-tality estimates for stent thrombosis range from 19to 45% [5].

The risks of hemorrhage during or followingintracranial neurosurgical procedures include exces-sive blood loss, neurologic deficits and disability, andrisks related to blood transfusion. Small amountsof intracranial blood loss may result in significantchanges of intracranial pressure and consequent neu-ronal insult. The use of antiplatelet agents may appre-ciably increase these risks. To date, no trial hasbeen performed that demonstrates an excessive bleed-ing risk from maintenance of antiplatelet drugs dur-ing a neurosurgical procedure. There is evidencethat aspirin and nonsteroidal anti-inflammatory drugscause excessive hemorrhage in other surgical proce-dures, and while many of these studies measure differ-ent endpoints, the overall suggestion is that antiplateletagents increase blood loss and hemorrhagic complica-tions, and that combinations of drugs further increasethese problems.

Standard practice in most centers is complete dis-continuation of anticoagulation to remove any excesshemorrhage risk. This abrupt withdrawal results in arebound phenomenon of increased platelet aggrega-

tion, which may compound the thrombotic risk posedfrom the surgical stress response.

Consequently, the clinician must balance the risksof major cardiac events on the one hand with neuro-logic sequelae on the other. The context in which thepatient is seen will have a bearing on the approachtaken tomanaging anticoagulation, and all optionswillneed consideration among surgery, cardiology, andanesthesiology staff to minimize risks during the peri-operative period. There are three scenarios that mayarise:

1. The patient without a coronary stent but withconcerning symptomatology.

2. The patient with coronary artery stents takingdual antiplatelet therapy.

3. The patient with coronary stents taking aspirinalone.

Scenario 1 – Patient withno coronary stentsA patient without previously documented coronarydisease, but concerning cardiac symptoms, present-ing for elective surgery should be referred for assess-ment by a cardiologist. The indications for preopera-tive coronary revascularization are limited, however,and an improvement in mortality from interventionhas only been demonstrated for patients with signifi-cant left main stem disease treated prior to major vas-cular surgery [6]. This evidence may not be directlyapplicable to neurosurgical patients. The ACC/AHAguidelines recommend avoiding noncardiac surgery fol-lowing stent deployment for a minimum of 1 month inthe case of BMS and 1 year if using a DES [6]. Thisdelay has been correlated with a reduction in majoradverse cardiac events (MACE) [6]. In both cases, itis recommended that aspirin be continued through-out the perioperative period. It is unlikely that surgeryfor an intracranial aneurysm could be delayed for ayear or aspirin continued in the perioperative period.It has been suggested that patients may instead havea BMS with the associated shortened period of anti-coagulation cover, or a balloon angioplasty that mayavoid the risks associated with use of a stent [1]. Thechoiceswould seem to be to (1) perform surgery beforecoronary intervention, (2) perform angioplasty with-out stenting and revisit the coronary artery diseasepostoperatively, or (3) perform angioplasty with a baremetal stent allowing at least 30 days, but preferably

52

Case 16. Patient with coronary artery stent

90 days, before routine surgery on the basis ofACC/AHA guidelines [1, 3, 6].

Scenario 2 – Patient on dualantiplatelet therapyThis is a more frequent presentation: a patient pre-senting for surgery with coronary stents in situ andcurrently taking anticoagulation. Attention during thepreoperative visit must be paid regarding the risk fac-tors given earlier, although a decision as to whento proceed to surgery is usually based on the timeelapsed since coronary stenting and the urgency of theneurosurgical procedure. The risks of coronary stentthrombosis do ameliorate with time as noted above,but the recommendations for antiplatelet therapy willstill apply. Patientswhoprematurely discontinue clopi-dogrel are almost 60 times more likely to suffer astent thrombosis [5].The associatedmortality rate was45% [5]. Other studies corroborate this evidence, sug-gesting that discontinuation of thienopyridine therapywithin the first 6 months after DES implantation wasthe strongest predictor for stent thrombosis with a haz-ard ratio of 13 [1, 3]. Aspirin withdrawal would alsobe necessary for intracranial surgery and this seems tocompound the risks of stent thrombosis [3].

Scenario 3 – Patient on singleantiplatelet therapyThe risk associated with discontinuation over 1 yearfrom coronary stenting has not been defined in alarge series; however, an observational study foundthat stent thrombosis at a mean duration of 15 monthssince deployment represented approximately 20% ofall coronary events following aspirin withdrawal [3].An odds ratio of 90 for the risk of myocardial infarc-tion following acute aspirin withdrawal in patientswith coronary stents has been reported [3, 5].The aver-age time from aspirin withdrawal to thrombotic eventis 10 days [5].There continue to be case reports of verylate stent thrombosis during times of antiplatelet with-drawal and therefore the ACC/AHA guidelines stateaspirin use should continue indefinitely [2].

Alternatives to completeanticoagulation discontinuationBridging therapies have been suggested, includingwithholding clopidogrel and covering the periopera-

tive period with intravenous heparin or glycoproteinIIb/IIIa inhibitors, converting to short-acting cyclo-oxygenase inhibitors (such as flubiprofen) prior tosurgery, or preoperative platelet transfusions [1, 3].Most of these suggestions continue perioperativeaspirin administration.There is unfortunately a lack ofevidence and data to estimate the risks of continuingany form of antiplatelet therapy in the neurosurgicalperioperative period. Studies in this area have exam-ined major general, vascular, or cardiac surgery, andfind an increase in blood loss with increased trans-fusion but which does not appear to affect mortality[1]. However, expert opinion unanimously agrees thatantiplatelet therapy should bewithheld in closed-spacesurgery [1, 3].

ConclusionIn conclusion, coronary artery stents present a chal-lenging dilemma for the perioperative care of neu-rosurgical patients. There is currently an irresolvableconflict between the risks of withholding and continu-ing antiplatelet agents in the perioperative period.Thisconflict is most acute in closed-space surgeries such asintracranial procedures.

A delay of surgical intervention for as long aspossible will allow coronary stent thrombosis risk todecrease, but not disappear entirely. Indeed, no periodfollowing coronary artery stenting appears to be ableto be described as risk-free. Rapid percutaneous inter-vention is required in the event of stent thrombosis andaccess to coronary intervention laboratories is there-fore a consideration. Attempts must be made to min-imize the period of antiplatelet withdrawal and earlyreintroduction postoperatively should be considered.Please see Case 31 for further discussion.

References1. G. M. Howard-Alpe, J. de Bono, L. Hudsmith et al.

Coronary artery stents and non-cardiac surgery. Br JAnaesth 2007; 98: 560–74.

2. S. B. King, S. C. Smith, J. W. Hirshfeld et al. 2007focused update of the ACC/AHA/SCAI guidelineupdate for percutaneous coronary intervention: areport of the American College of Cardiology/American Heart Association Task Force on Practiceguidelines. J Am Coll Cardiol 2008; 51: 172–209.

3. L. T. Newsome, R. S. Weller, J. C. Gerancher et al.Coronary artery stents II: Perioperative considerationsand management. Anesth Analg 2008; 107: 570–90.

53


4. J. W. vanWerkum, A. A. Heestermans, A. C. Zomeret al. Predictors of coronary stent thrombosis: theDutch Stent Thrombosis Registry. J Am Coll Cardiol2009; 53: 1399–409.

5. P. G. Chassot, A. Delabays,D. R. Spahn. Perioperativeantiplatelet therapy: the case for continuing therapy in

patients at risk of myocardial infarction. Br J Anaesth2007; 99: 316–28.

6. L. A. Fleisher. Cardiac risk stratification fornoncardiac surgery: update from the AmericanCollege of Cardiology/American Heart Association2007 guidelines. Cleve Clin J Med 2009; 76: S9–15.

54


17 Deep hypothermic circulatory arrest forintracranial aneurysm clippingMatt Giles and George A. Mashour

Over the last 15 years, advances in the fields ofinterventional neuroradiology and neurosurgicaltechniques have changed the way in which manyintracranial aneurysms are managed. Many tertiarycenters now rely on interventional neuroradiolo-gists for coiling procedures of simple intracranialaneurysms, versus a more traditional and invasiveneurosurgical approach. However, there still remainsa subset of patients with large and/or complexintracranial aneurysms who are not candidates forneuroradiologic coiling procedures and for whomconventional neurosurgical approaches still carryexceptionally high risks of morbidity and mortality.The following is a case presentation and discussionof the application of deep hypothermic circulatoryarrest (DHCA) for patients undergoing large and/orcomplex intracranial aneurysm clipping and repair.

Case descriptionA 62-year-old female presented to the operating roomwith a diagnosis of two large intracranial aneurysms.The first was a 17-mm aneurysm located at the junc-tion of two temporal branches of the left posteriorcommunicating artery (PCA), and the second wasa 19-mm aneurysm located at the junction of theleft posterior medial choroidal artery and the PCA.The patient was hemodynamically stable, awake, alert,and had presented to the emergency room earlierthat day with vague complaints of severe headache,blurred vision, and difficulty with ambulation. Uponcomputed tomography-angiography (CTA) of thehead, the neuroradiologist concluded that the twoaneurysms were too large and complex in fusiformanatomy to allow for safe and effective coiling. Neu-rosurgery was consulted and concluded that an intra-cranial approach with clipping would likely provesuccessful. However, given the size and anatomy, cir-culatory arrest was necessary for adequate collapse ofthe aneurysms in order to facilitate surgical dissec-

tion andmanipulation. After discussing the risks, ben-efits, and alternatives with the patient, and obtaininginformed consent, a decision to perform the surgeryunder DHCA was made.

The patient was brought to the operating room,noninvasive monitors placed, and under local anes-thesia a right radial arterial line was established. Thepatient was positioned supine, preoxygenated, and atitrated, mixed intravenous induction with midazo-lam, fentanyl, and propofol was conducted.Thepatientwas easily bag/mask ventilated, and tracheal intuba-tion was facilitated with rocuronium. Ventilation wasadjusted to maintain an end-tidal CO2 in the 32–36 mmHg range. Sevoflurane was titrated to ade-quately maintain anesthesia and hemodynamic stabil-ity. Two large-bore 16-gauge peripheral intravenouscatheters were started, and a 9-French Cordis wasplaced in the patient’s right subclavian vein for cen-tral access, through which a Swan–Ganz pulmonaryartery catheter was placed. A Foley bladder catheterwas inserted, and a transesophageal echocardiography(TEE) probe was gently placed. A brief TEE exam-ination was performed. The patient was then posi-tioned in the right lateral decubitus position with thelower torso gently turned to near-neutral supine posi-tion and supported with padding to facilitate surgi-cal cannulation of the left femoral artery and vein forcardiopulmonary bypass (CPB). Temperature probeswere placed in the nasopharynx, axilla, and bladder.Craniotomy with surgical dissection and isolation ofthe aneurysms commenced.

After common femoral artery and vein cannu-lation, positioning of the venous cannula for CPBwas verified by TEE to be at the level of the rightatrium. Upon near completion of aneurysm dissec-tion, adequate CPB flow was initiated and hemo-dynamic stability was achieved. Systemic hypother-mia commenced per CPB protocol until ventricularfibrillation occurred. At that point, potassium chlo-ride was administered via the right atrial port of the

55


pulmonary artery catheter to facilitate asystole. Sys-temic cooling continued to a core body and corticaltemperature of 15 ◦C, at which point the CPB circu-lation was arrested, blood was drained via the venouscannula until vascular relaxation was reported by theneurosurgeon, and complete aneurysm reconstruc-tion/clipping ensued. A total deep hypothermic cir-culatory arrest time of 32 minutes was documented.Upon clipping of both aneurysms, CPB was slowlyreestablished and both extracorporeal flow and patientmean arterial pressure were brought to pre-CPB lev-els. Separation from CPB and surgical closure wereuneventful.

DiscussionThe rationale behind the technique of DHCA stemsfrom the significant advantage the surgeon has onceblood flow has stopped circulating to the aneurysm. Inparticular, by temporarily eliminating blood flow intoan aneurysm, the vascular anomaly is converted froman otherwise hard, pulsating mass into a much morepliable, collapsed vascular sac [1].

The technique of DHCA is not without significantrisk and should be considered an option of last resortfor large or complex intracranial vascular anomalies.In general, aneurysms located in the anterior circula-tion aremost often amenable to neurocoiling via inter-ventional radiology, or more traditional neurosurgicalclipping with a straightforward intracranial approach[1, 2]. Typically, larger aneurysms (10–25mm)or com-plex aneurysms (fusiform, branching, etc.) predomi-nantly occur within the posterior circulation, such asin the above case. Due in part to the size or complexnature of these aneurysms, conventional aneurysmcoiling and even traditional neurosurgical clipping areboth extremely difficult and with higher risk of rup-ture. Often these complex vascular anomalies can-not be collapsed easily intraoperatively because of thebreadth of the neck, the irregularity/complexity of thenetwork of arterial branches at the base, the presenceof thrombus within the lumen, or extensive artheromawithin the vessel walls [1, 2]. Cessation of the patient’sbloodflowallows for optimization of surgical exposureand provides an opportunity for clipping and recon-struction that would otherwise prove extremely diffi-cult.When one or both proximal and distal arteries areinaccessible for complex aneurysm repair, DHCAmaybe the only safe and effectivemeans of vascular control[1–4].

As discussed previously, there exists a subset ofpatients requiring extensive repair of large and/orcomplex intracranial vascular anomalies usingDHCA.The circulatory arrest component of this techniqueaffords the surgeon the advantage of a collapsed,bloodless aneurysm with which to work. However,the scientific rationale for the application of deephypothermic temperatures during the cessation ofcirculation rests primarily upon the observed effectthat a temperature-mediated reduction of metabolismoccurs.Whole body and cerebral oxygen consumptionduring induced hypothermia decreases the metabolicrate for oxygen by a factor of 2 to 2.5 for every10 ◦C reduction in temperature below normothermia(defined as 37 ◦C) [5–8]. These results are consis-tent in both in vivo and in vitro models, which sug-gest that the rate of cellular metabolism is directlyproportional to temperature. The benefit of induc-ing hypothermia in the patient undergoing circulatoryarrest, therefore, is observed as both a neuroprotectiveand systemic organ-protective effect. By decreasingthe patient’s temperature to 15–18 ◦C (common dur-ing DHCA), the systemic basal metabolic rates in allorgans (including the brain) have decreased dramat-ically [5–8]. Hence by decreasing the basal metabolicrates for cellular substrates and oxygen utilization inall tissues, hypothermia helps exert a protective effectfrom ischemia while circulation is temporarily halted[9, 10]. Other studies have noted that the reduction inoxygen supply during deephypothermic low-flowCPBis associated with preferential increases in vital organperfusion (e.g., to the brain) and increased extractionof oxygen [8–10].

The protective effect of hypothermia during low-flow and no-flow circulatory states has also beendemonstrated by studies looking at reductions in intra-cellular markers of ischemia. It has been well estab-lished that during episodes of brain ischemia, sev-eral excitatory amino acids (glutamate, aspartate, etc.)are released in relation to deprivation of cellular sub-strates (oxygen, glucose, etc.) indicating anaerobicmetabolism and ischemia. However, hypothermia hasbeen shown to significantly decrease the release ofexcitatory amino acids and other markers of focal andglobal ischemia [5–10].

Reducing a patient’s temperature during surgerycarries several risks, which have a wide range ofimplications on multiple organ systems. Moreover,the duration of the safe period of DHCA has notclearly been delineated. A number of studies aimed at

56

Case 17. Deep hypothermic circulatory arrest

determining a “safe DHCA time” have been con-ducted and suggest that an arrest period up to 40 min-utes is tolerated without evidence of long-term neuro-logic sequlae [1–4]. But by no measure is neurologicischemia and injury the only risk of DHCA. By induc-ing hypothermia, the patient is at increased risk forcoagulation dysfunction and bleeding, which can becompounded in light of heparinization during (andseparation from) cardiopulmonary bypass. Hypother-mia also places the patient at increased risk of cardiacdysrhythmias, such as ventricular tachycardia and fib-rillation. Recurrent ventricular tachycardia and fibril-lation are commonly observed during rewarming andcan persist into the postoperative period. Also associ-ated with hypothermia is the increased risk of post-operative wound infection, which can significantlyincrease morbidity and mortality. Especially impor-tant to neurosurgical patients is the finding that cere-bral autoregulation is lost during, and immediatelyafter, extremes of temperature. In this setting, cerebralperfusion therefore becomes highly dependent uponthe conduct of extracorporeal perfusion and post-bypass hemodynamic stability [1, 6–9].

In approaching the anesthetic management of apatient scheduled to undergo DHCA for aneurysmclipping, the goals of providing safe and effective anes-thesia are no different than in any other patient under-going general anesthesia, with the exception of a fewadded objectives. The additional goals of this partic-ular anesthetic should include the prevention of cere-bral injury, preserving overall hemodynamic stabilitybefore, during, and after CPB, and assisting the sur-geon/CPB team in all efforts with special attention toblood loss/replacement, coagulation status/correction,and neurologic monitoring.With respect to the choiceof monitors, all standard American Society of Anes-thesiologists monitors should be employed, as wellas an arterial catheter for blood pressure monitor-ing. Temperature probes in the bladder and nasophar-ynx, in addition to cerebral core temperature moni-toring are critical to guiding systemic cooling effortsonce CPB has been initiated. Multiple modalities fortemperaturemonitoring are recommended for reliablemeasurements. The degree of induced hypothermiamust be carefully controlled. In particular, slowly cool-ing the patient, maintaining a constant hypothermia,and finally slowly decooling the patient are key com-ponents to the success of DHCA [1–4].

As DHCA is becoming more widely employedas a technique for repairing complex aneurysms, the

use of TEE for cardiac monitoring has proven veryuseful. Specifically, TEE allows for early intraopera-tive evaluation of overall cardiac function, includingestimation of ejection fraction, assessment for wallmotion abnormalities, valvular dysfunctions, as wellas assessment of ventricular volume status and con-tractility. Transesophageal echocardiography is alsovery helpful with verifying the proper placement ofthe venous bypass cannulae into the right atriumwhile preparing for CPB [1]. Furthermore, assess-ing cardiac function immediately prior to separationof CPB, and after complete separation from CPB,can be invaluable at guiding the intraoperative man-agement. In addition to TEE, a pulmonary arterycatheter can be helpful in guiding the anestheticmanagement.

ConclusionOverall, the application of DHCA has been success-fully employed as a viable method of providing opti-mized surgical access to and control of large or com-plex intracranial aneurysms. However, DHCA is asso-ciated with unique risks and potentially devastatingcomplications. The margin for error is small, and suc-cess depends upon an experienced and knowledgeableteam. The safe practice and management of DHCArequires an extensive understanding of cardiac andneurosurgical anesthetic practice, CPB, as well as care-ful consideration and proper planning.

References1. W. L. Young,M. T. Lawton,D. K. Gupta et al.

Anesthetic management of deep hypothermiccirculatory arrest for cerebral aneurysm clipping.Anesthesiology 2002; 96: 497–503.

2. W. L. Young. Cerebral aneurysms: current anaestheticmanagement and future horizons. Can J Anaesth 1998;45: R17–31.

3. A. Levati, C. Tommasino,M. P. Moretti et al. Giantintracranial aneurysms treated with deep hypothermiaand circulatory arrest. J Neurosurg Anesthesiol 2007;19: 25–30.

4. M. G. Massad, F. T. Charbel, R. Chaer et al. Closedchest hypothermic circulatory arrest for complexintracranial aneurysms. AnnThorac Surg 2001; 71:1900–4.

5. L. Berntman, F. A. Welsh, J. R. Harp. Cerebralprotective effect of low-grade hypothermia.Anesthesiology 1981; 55: 495–8.

57


6. J. Hartung, J. E. Cottrell. Mild hypothermia andcerebral metabolism. J Neurosurg Anesthesiol 1994; 6:1–3.

7. J. Hartung, J. E. Cottrell. In response to: effects ofhypothermia on cerebral metabolic rate for oxygen.J Neurosurg Anesthesiol 1994; 6: 22.

8. L. S. Fox, E. H. Blackstone, J. W. Kirklin et al.Relationship of whole body oxygen consumptionto perfusion flow rate during hypothermiccardiopulmonary bypass. J Thorac Cardiovasc Surg1982; 83: 239–48.

9. L. S. Fox, E. H. Blackstone, J. W. Kirklin et al.Relationship of brain blood flow and oxygenconsumption to perfusion flow rate during profoundlyhypothermic cardiopulmonary bypass. Anexperimental study. J Thorac Cardiovasc Surg 1984; 87:658–64.

10. R. Busto,M. Y. Globus,W. D. Dietrich et al. Effect ofmild hypothermia on ischemia-induced release ofneurotransmitters and free fatty acids in rat brain.Stroke 1989; 20: 904–10.

58


18 Neuroprotection during surgical clip ligationof cerebral aneurysmsNeeraj Chaudhary, Joseph J. Gemmete, B. Gregory Thompson and Aditya S. Pandey

Subarachnoid hemorrhage (SAH) from a rupturedcerebral aneurysm is associated with significant mor-bidity and mortality. The endovascular treatment ofruptured cerebral aneurysms has become well estab-lished over the last decade as the preferred treat-ment modality [1]. However, there are no conclusivedata yet regarding the best management of unrup-tured aneurysms. Not every aneurysm configurationand location is feasible for endovascular treatmentdue to technical limitations.Microsurgical clip ligation(MCL) is the treatment of choice for those aneurysmsthat cannot be safely treated from an endovascularapproach. Here we discuss the different strategies ofneuroprotection applied to prevent ischemic damageto the brain during clip ligation of cerebral aneurysms.

Case descriptionA 44-year-old male presented to our hospital withacute SAH; diagnostic cerebral angiogram was per-formed emergently that demonstrated an anteriorcommunicating artery (ACOM) aneurysm. Two addi-tional aneurysms were also identified at the origin ofleft superior hypophyseal artery and right posteriorcommunicating artery. The ACOM artery aneurysmwas deemed to be the cause of the SAH based onits morphology and the distribution of blood in thesubarachnoid space. Endovascular embolization of theACOM aneurysm was technically not feasible due tothe complex shape and wide neck. Given these find-ings, MCL of the aneurysm via an open craniotomywas deemed the best option for the patient.

Thepatient was taken to the operating room,wheregeneral anesthesia was induced and endotracheal intu-bation established. The anesthesia was maintained bytotal intravenous anesthesia in order to minimize theuse of inhalational anesthetic agents that can increasecerebral blood volume and intracranial pressure.The patient’s blood pressure was tightly controlled toprevent any hypertensive surges during intubation

and pinning that could possibly cause rebleedingfrom the aneurysm. A right frontal craniotomy wasperformed followed by a meticulous durotomy andbone drilling of the sphenoid ridge to gain accessto the ACOM aneurysm. Prophylactically, mannitoland furosemide were administered before the duralincision. The right and left A1 and A2 segments of theanterior cerebral artery and the ACOM complex wereidentified.

During further exposure of the aneurysm, as theright frontal lobe was retracted there was rupture ofthe aneurysm with hemorrhage from its contralateralaspect. The patient was put into “burst suppression”with a bolus dose of propofol, which also decreasedthe blood pressure. Temporary clips were applied toboth A1 segments which controlled the hemorrhageto some extent; however, there was minimal retro-grade bleeding from the A2 segments. The aneurysmwas then decompressed. Subsequently, a permanentfenestrated clip was placed across the neck of theACOM aneurysm, maintaining patency of the A2division intimately related to the superior aspect ofthe aneurysm. This secured the aneurysm, whichwas confirmed upon removal of the temporary clips.Burst suppression was reversed after approximately13 minutes of temporary clip occlusion.

The surgical field was then irrigated with salineantibiotic solution and the craniotomy closed withmeticulous hemostasis. The patient was transferred tothe neurosurgical intensive care unit in a stable condi-tion. On the second postoperative day the trachea wasextubated.

DiscussionAn interdisciplinary approach to the managementof cerebrovascular disease is necessary for optimalpatient outcome. The role of the neuroanethesiologistin supporting the cerebrovascular physiology to over-come the stress from SAH is crucial. Subarachnoid

59


hemorrhage is a devastating disease that is accom-panied by a marked stress response with increasedplasma concentration of catecholamines. Electrolyteimbalances are also prominent, which have both cere-bral and systemic effects. Hence, the American Societyof Anesthesiologists class in this group of patients isoften IV or V [2].

Currently, application of temporary clips to theparent vessel duringMCL is increasingly utilized.Thisis used either to gain control over unforeseen rup-ture of the aneurysm before clip ligation or as partof a planned strategy to achieve optimal positioningof the definitive aneurysm clip. In either case anes-thetic interventions can offer cerebral parenchymalprotection during these periods of temporary cessa-tion of blood supply. The interventions are based onthe duration of occlusion of blood flow. If the tempo-rary clip application is going to be for less than 60–120seconds then no anesthetic intervention is necessary.However, if this duration of occlusion to blood flowis exceeded, the following anesthetic interventions canbe undertaken:� The inspiratory concentration of oxygen is

increased to 100%.� Propofol or barbiturates are administered

intravenously to achieve “burst suppression” onelectroencephalography. This offers cerebralprotection by reducing cerebral metabolism to aminimum, which in turn reduces oxygenconsumption.

� Additional doses of phenylephrine to increasearterial blood pressure to more than 20% ofbaseline may improve retrograde flow to the brainparenchyma temporarily deprived of blood flowbut can prove potentially dangerous. It mayinduce bleeding in the surgical field and thereforemake temporary clip removal more prolongedand difficult [3].

� In situations where the temporary clippingexceeds 10 minutes, the role of induced mild tomoderate hypothermia has been questioned in therecent reports of the IHAST trial secondary dataanalysis [4].

Another method that has been reported to have somebenefit in case of an unforeseen hemorrhage duringclip ligation of a cerebral aneurysm is the use of adeno-sine to induce temporary cardiac arrest [5]. Nussbaumet al. in their series of 10 cases have demonstrated

the use of 12 mg adenosine injected intravenously toinduce cardiac arrest for approximately 10 seconds[5]. During this period of circulatory arrest, blood issuctioned off the surgical field and a temporary clipis applied to gain hemostasis as the normal rhythmreturns. Although adenosine is not in itself neuropro-tective, it can afford an opportunity to achieve rapidproximal control of a ruptured aneurysm.

There has been a recent explosion of interest inthe mechanisms of cerebral ischemia, with over 1000experimental papers and over 400 clinical articlesappearing within the past 6 years. These studies, inturn, are the outcome of three decades of investigativework to define the multiple mechanisms and media-tors of ischemic brain injury, which constitute poten-tial targets of neuroprotection [6]. Apart from tempo-rary clip occlusion there are other insults to the brainduring MCL such as tissue retraction and intraopera-tive hemorrhage as discussed above. The IHAST trialwas the first prospective multicenter study to assesspatient outcome where hypothermia was applied as aneuroprotective adjunct [7]. The IHAST investigatorsconcluded that there was no significant improvementin neurologic outcome in patients with good gradeSAH who underwent a craniotomy. An inconclusivetrend towards some benefit in themale population andin patients with craniotomy 8–14 days after SAH wasshown on secondary analysis of the data. The result ofthis study is in complete discordance with the resultsof preclinical studies in the animal population [8]. Fur-ther analysis of the IHAST cohort revealed that therewas no significant added protection offered not only bythe induced hypothermia but also by the supplementalpharmacologic protective agents [4].

Another nonpharmacologic modulation that hasdemonstrated some benefit in animal studies isischemic preconditioning. Feng et al. recently havedemonstrated the neuroprotective effects of hypoxicpreconditioning in neonatal rat brain [9]. They con-clude that the underlying mechanism of neuropro-tection is mediated during the period of reperfusionfollowing hypoxic insult and is via an increase inexpression of vascular endothelial growth factor A,which in turn decreases the apoptotic cascade. Theupregulation of adenosine receptors in ischemic pre-conditioning in animal models again has been pro-posed as a potential neuroprotective agent [10]. Anes-thetic preconditioning with volatile agents is also anarea of active investigation.

60

Case 18. Neuroprotection during surgical clip ligation of cerebral aneurysms

There are several other pharmacologic agents thathave been assessed for neuroprotective propertiesbased on a growing understanding of the neuronalischemic injury pathway. There is a clearer under-standing from basic science research of the role ofthe N-methyl-D-aspartate (NMDA) glutamate recep-tor in this pathway. Hardingham, in his recent reviewon this receptor, opines that hypo- or hyper-NMDAstimulation can set off a neurodestructive cascade[11]. Although implicated in animal stroke models,NMDA receptor blockade has not shown any clini-cal benefit in acute ischemic stroke due to poor tol-erance and efficacy [11]. Another study in an animalmodel has implicated downregulation and internaliza-tion of NMDA receptors in the regions remote and infact contralateral to the area of infarct, which in turncauses cognitive impairment in ischemic stroke with-out actual cell death [12]. A detailed discussion of therelevant pathways is beyond the scope of this chapter.Competitive inhibition of the glycine site of theNMDA receptor by the general anesthetic gas xenonhas been implicated in its neuroprotective role [13]. Anin vivomodel of rat hippocampus subjected to oxygen-glucose deprivation and then exposed to xenonshowed significant neuronal protection for a period of3–4 hours from the ischemic insult, which was alsoshown to be inhibited by increased glycine concentra-tions. A clinical trial to learnmore about the neuropro-tective role of xenon is being formulated. The NMDAreceptor blocker nitrous oxide has also been impli-cated to potentially play a role in the ischemic path-way. However, the IHAST group has demonstrated nodetrimental effect of the application of nitrous oxide inpatients undergoing temporary clip occlusion duringMCL [14].They show that although the risk of delayedischemic neurologic deficits is increased in this group,the overall neurologic outcome is not affected.

Erythropoietin (Epo) is another agent that isincreasingly being studied for its neuroprotectiveproperties. It is well established at a molecularlevel that Epo in general offers cytoprotection viadecreased oxidative stress and lipid peroxidation. Inthe brain its potential mechanism of neuroprotectionhas been studied recently in an in vitro study [15].The study demonstrated that the pathway of neuro-protection is mediated by an increase in concentra-tion of glutathione (an antioxidant) that results froman upregulation of the cystine glutamate exchanger(system Xc).

ConclusionAlthough there are promising revelations regardingneuroprotective pathways and how these can be mod-ulated in cellular and animal models, there is currentlyno definite benefit shown in human subjects. Therole and efficacy of the cerebral protective measuressuggested to prevent ischemic brain damage fromtemporary clip occlusion during MCL are not clearlyunderstood. Larger multicenter studies and moretranslational research are needed to demonstrate theefficacy of currently employed cerebral protectivemeasures. The next decade will hopefully provide aclearer analysis of the pathways of ischemic neuronalcell death, which in turn will help to determine moreeffective neuroprotective options.

References1. A. J. Molyneux, R. S. Kerr, J. Birks et al.; ISAT

Collaborators. Risk of recurrent subarachnoidhaemorrhage, death, or dependence and standardisedmortality ratios after clipping or coiling of anintracranial aneurysm in the InternationalSubarachnoid Aneurysm Trial (ISAT): long-termfollow-up. Lancet Neurol. 2009; 8(5): 427–33. Epub2009 Mar 28.

2. P. H. Manninen, A.W. Gelb. The anestheticmanagement of patients during posterior fossaaneurysm surgery. In C. G. Drake, S. J. Peerless, J. A.Hernesniemi (eds.), Surgery of VertebrobasilarAneurysms. London, Ontario Experience on 1767Patients. Vienna: Springer-Verlag; 1996,pp. 280–4.

3. T. Randell,M. Niemela, J. Kytta et al. Principles ofneuroanesthesia in aneurysmal subarachnoidhemorrhage: the Helsinki experience. Surg Neurol2006; 66: 382–8.

4. B. J. Hindman, E. O. Bayman,W. K. Pfisterer et al.No association between intraoperative hypothermiaand supplemental protective drug and neurologicoutcomes in patients undergoing temporary clippingcerebral aneurysm surgery: findings from theIntra-operative Hypothermia for Aneurysm SurgeryTrial. Anesthesiology 2010; 112: 86–101.

5. E. S. Nussbaum, L. A. Sebring, I. Ostanny et al.Transient cardiac standstill induced by adenosine inthe management of intraoperative aneurysmal rupture:technical case report. Neurosurgery 2000; 47: 240–3.

6. A. Tuttolomondo, R. Di Sciacca,D. Di Raimondoet al.Neuron protection as a therapeutic target in acuteischemic stroke. Curr Top Med Chem. 2009; 9(14):1317–34.

61


7. M.M. Todd, B. J. Hindman,W. R. Clarke,J. C. Torner. Mild intraoperative hypothermia duringsurgery for intracranial aneurysm. N Engl J Med 2005;352: 135–45.

8. G. L. Clifton, J. Y. Jiang, B. G. Lyeth et al.Markedprotection by moderate hypothermia afterexperimental traumatic brain injury. J Cereb BloodFlow Metab 1991; 11: 114–21.

9. Y. Feng, P. G. Rhodes, A. Bhatt. Hypoxicpreconditioning provides neuroprotection andincreases vascular endothelial growth factor A,preserves phosphorylation of Akt-Ser-473 anddiminishes caspase-3 activity in neonatal rathypoxic-ischemic model. Brain Res 2010; 1325: 1–9.

10. R. L. Williams-Karnesky,M. P. Stenzel-Poore.Adenosine and stroke: Maximising the therapeuticpotential of adenosine as a prophylactic and acuteneuroprotectant. Curr Neuropharmacol 2009; 7:217–27.

11. G. E. Hardingham. Coupling of NMDA receptor toneuroprotective and neurodestructive events.

Colworth Medal Lecture. Biochem. Soc. Trans. 2009; 37:1147–60.

12. J. Dhawan,H. Benveniste,M. Nawrocky, S. D. Smith.Transient focal ischemia results in persistent andwidespread neuroinflammation and loss of glutamateNMDA receptors. Neuroimage 2010; 51: 599–605.

13. P. Banks, N. P. Franks, R. Dickinson. Competitiveinhibition at the glycine site of the n-methyl-d-aspartate receptor mediates xenon neuroprotectionagainst hypoxia–ischemia. Anesthesiology 2010; 112:614–22.

14. J. J. Pasternak,D. G. McGregor,W. L. Lanier et al.Effect of nitrous oxide use on long-term neurologicand neuropsychological outcome in patients whoreceived temporary proximal artery occlusion duringcerebral aneurysm clipping surgery. Anesthesiology2009; 110: 563–73.

15. B. Sims,M. Clarke,W. Njah, E. S. Hopkins,H.Sonthiemer. Erythropietin-induced neuroprotectionrequires cystine glutamate exchanger activity. BrainRes 2010; 1321: 88–95.

62


19 Dexmedetomidine and nitrous oxide forcerebral aneurysm clippingMichael J. Claybon and George A. Mashour

Dexmedetomidine and nitrous oxide are commonlyused as adjuvants to volatile agents or other intra-venous anesthetics – the following case describes theseagents in combination with each other for anestheticmaintenance of a cerebral aneurysm clipping.

Case descriptionThe patient was a 58-year-old female with multi-ple intact unsecured aneurysms, who presented forclipping of one paraclinoid aneurysm. She had ahistory of hypertension, obesity, and delayed emer-gence from general anesthesia.The anesthetic goals forthis patient included (1) maintaining hemodynamicstability throughout the entire perioperative period,(2) maintaining the ability to monitor neurophysio-logic signals, and (3) prompt emergence and neuro-logic evaluation.

General anesthesiawas inducedwith propofol, fen-tanyl, and vecuronium. The patient was then main-tained with dexmedetomidine at levels of 0.2–0.7mcg/kg/min and nitrous oxide 60–70% in oxygen. Fentanylboluses were given to total 500 mcg (7 mcg/kg) duringand shortly after induction.

Hemodynamic stability was achieved both intra-operatively and postoperatively, preserving cerebralperfusion pressure and avoiding hypertension(Figure 19.1). Somatosensory evoked potentials werestable and without interference by anesthetics. Thedexmedetomidine infusion was discontinued duringsurgical closure approximately 40 minutes prior tothe reversal of muscle relaxant. The patient was tran-sitioned to 100% oxygen after completion of surgeryand removal of head pins. She was spontaneouslyventilating and following commands within four min-utes. Extubation was smooth and delayed emergencewas avoided. Postoperatively the patient was hemo-dynamically stable, remained in the neurosurgicalintensive care unit until postoperative day 1 and wasdischarged home on postoperative day 3. During her

postoperative interview, she was specifically askedabout intraoperative awareness and she reported norecall of surgical events.

DiscussionSince its public introduction by Horace Wells in 1845,nitrous oxide has been in continuous use. Clinicianscontinue to debate the use of nitrous oxide, per-haps most avidly within the field of neuroanesthe-sia. While a drug that offers analgesia, sedation, andrapid emergence seems optimal for neurologic surgery,some properties of nitrous oxide promote concern.These include increases in cerebral metabolic rate,cerebral blood flow, intracranial pressure, and abilityto increase the volume of pneumocephalus. McGregoret al. published a study inwhich data from the Intraop-erative Hypothermia for Aneurysm Surgery Trial wereanalyzed. Of the 1000 patients studied, 373 receivednitrous oxide with no detrimental effect on long-termgross neurologic or neuropsychologic function [1].Thismulticenter study had significant variability in theuse of nitrous oxide, with some centers using it exten-sively and others seldom if at all. However, in light ofthe fact that much of the concern for nitrous oxide isbased on in vitro data and animal studies, McGregoret al.’s analysis may be helpful in supporting thoseanesthesiologists who employ nitrous oxide for neu-rovascular surgery.

Dexmedetomidine was introduced far morerecently. This selective alpha-2-receptor agonist wasinitially used in mechanically ventilated critically illpatients before it found use in the operating room.Dexmedetomidine is unique in its ability to producesedation, anxiolysis and analgesia with little respira-tory depression. It offers an interesting advantage inthat patients may be sedated with dexmedetomidinebut remain arousable and cooperative. The centralaction of dexmedetomidine at presynaptic neuronsin the brainstem’s locus ceruleus produces sedation,

63


Figure 19.1. Intraoperative andrecovery room vital signs.

while postsynaptic receptors in the spinal cord seemto be important in mediating analgesia. Additionalbenefits involving hemodynamics and “cooperativesedation” have been cited as advantages in usingdexmedetomidine for monitored anesthesia care andas an adjunct in general anesthesia. Common use ofdexmedetomidine involves a loading dose of 1 mcg/kgover 10 minutes followed by a continuous intravenousinfusion between 0.2 and 0.7 mcg/kg/hour.

Dexmedetomidine decreases anesthetic require-ments substantiallywith both opioid-sparing andmin-imum alveolar concentration (MAC)-sparing quali-ties. It is associated with reductions inMAC of inhaledanesthetics of 50–90% [2]; additionally, dexmedetomi-dine has been shown to reduce opioid requirements by30–50% [3].

The benefits of dexmedetomidine as an anestheticadjunct extend beyond sedation and analgesia. Sym-patholytic effects blunt hypertension and tachycardiaduring surgery and anesthesia. Dexmedetomidine hasbeen shown to reduce oxygen consumption peri-operatively [2]. The hemodynamic stability associ-ated with dexmedetomidine may also be cardio-protective. Alpha-2-agonists have been shown todecrease mortality after various types of surgical pro-cedures; significant benefit has been noted in vascularsurgery with a decrease in mortality and myocardialinfarction [4].

It becomes especially important when discussingan anesthetic agent in neurosurgery to examine itsneurophysiologic profile. Dexmedetomidine has beenshown to decrease cerebral blood flow (CBF) inhealthy human subjects [2]. Animal studies have sim-ilarly found a reduction in CBF, however, these stud-ies did not find concomitant decreases in cerebralmetabolic rate (CMRO2) [3]. The concern for ade-quacy of cerebral oxygenation based on these animalstudies was addressed in a human study of six healthyvolunteers. In this study, both CBF and CMRO2were found to decrease in a dose-dependent man-

ner, and the CBF/CMRO2 ratio was maintained [5].This study correlates with data supporting dexmedeto-midine having no detrimental effect on local braintissue oxygenation in patients undergoing cerebrovas-cular surgery, even under conditions of cerebrovas-cular compromise and hyperventilation [3]. Animalstudies have demonstrated decreases in intracranialpressure with dexmedetomidine; the limited humandata available suggest that lumbar cerebral fluid pres-sures are unchanged intraoperatively [2]. An appro-priate neuroanesthetic must also preserve neurophys-iologic monitoring. Dexmedetomidine has been stud-ied as an adjunct in spinal surgery and was not foundto change either somatosensory evoked potentials ormotor evoked potentials by any clinically significantamount [6].

Another favorable characteristic of dexmedetomi-dine is the potential for a neuroprotective effect. Cere-bral ischemia is associated with increased brain cat-echolamine levels, and decreases in sympathetic tonehave been shown to improve neurologic outcome[7]. Dexmedetomidine decreases norepinephrine lev-els in the brain and thus may be protective. Studiesin animals have demonstrated that dexmedetomidineimproves neuronal survival after transient global orfocal cerebral ischemia [3, 7]. Additionally, animal andin vitro studies suggest that dexmedetomidine reducesglutamate release [3]. Reduction of this excitatory neu-rotransmitter is potentially neuroprotective.

Aside from neurovascular surgery, dexmedetomi-dine may be beneficial in other realms of neuroanes-thesia. In various cases performed under monitoredanesthesia care, the “cooperative sedation” availedby dexmedetomidine is especially valuable. Carotidendarterectomies, craniotomies, and implantation ofdeep brain stimulators have been performed success-fully using dexmedetomidine as a sedative. In thesecases, dexmedetomidine provides sympatholysis andfacilitates detailed neurologic evaluation [3]. In a caseseries, Mack et al. reported that dexmedetomidine

64

Case 19. Dexmedetomidine and nitrous oxide for cerebral aneurysm clipping

appeared to be a useful sedative for awake cran-iotomies when sophisticated neurologic testing wasrequired [8]. Similar pharmacologic benefits may alsobe helpful in sedating neurocritical-care patients.

The most prominent side effects of dexmedetomi-dine involve hemodynamic changes including hypo-tension and bradycardia. Hypotension may predomi-nate when central alpha-2a-receptor activity promotesvasodilation, an effect compounded by hypovolemia.Hypertension, however, has been attributed to periph-eral alpha-2b-receptors, which may overwhelm thecentral effects after peak concentrations are reachedinitially. Blood pressure changes may have a tem-poral relationship to the use of a loading dose [2].Bradycardia is most commonly found in youngerpatients with high levels of vagal tone; dexmedeto-midine is not recommended for patients with heartblock [3].

Nitrous oxide and dexmedetomidine are mostcommonly adjuncts to volatile or intravenous anes-thetics. In this case, their combination provided acomplete anesthetic while maintaining hemodynamicstability, enabling neurophysiologic monitoring, andfacilitating a prompt emergence in a patient with ahistory of delayed recovery from general anesthesia.Although we do not recommend this combination forgeneral use, it may be appropriate in certain clinicalsituations.

References1. D. G. McGregor,W. L. Lanier, J. J. Pasternak et al.

Effect of nitrous oxide on neurologic andneuropsychological function after intracranialaneurysm surgery. Anesthesiology 2008; 108: 568–79.

2. A. T. Gerlach, J. F. Dasta. Dexmedetomidine: anupdated review. Ann Pharmacother 2007; 41: 245–52.

3. A. Bekker,M. K. Sturaitis. Dexmedetomidine forneurological surgery. Neurosurgery 2005; 57: 1–10.

4. D. N. Wijeysundera, J. S. Naik,W. S. Beattie. Alpha-2adrenergic agonists to prevent perioperativecardiovascular complications: a meta-analysis. Am JMed 2003; 114: 742–52.

5. J. C. Drummond, A. V. Dao,D. M. Roth et al. Effectof dexmedetomidine on cerebral blood flow velocity,cerebral metabolic rate, and carbon dioxide responsein normal humans. Anesthesiology 2008; 108: 225–32.

6. E. Bala,D. I. Sessler,D. R. Nair et al.Motor andsomatosensory evoked potentials are well maintainedin patients given dexmedetomidine during spinesurgery. Anesthesiology 2008; 109: 417–25.

7. J. Kuhmonen, J. Pokorny, R. Miettinen et al.Neuroprotective effects of dexmedetomidine in thegerbil hippocampus after transient global ischemia.Anesthesiology 1997; 87: 371–7.

8. P. F. Mack, K. Perrine, E. Kobylarz et al.Dexmedetomidine and neurocognitive testing inawake craniotomy. J Neurosurg Anesthesiol 2004; 16:20–5.

65


20 Anaphylaxis associated with indocyaninegreen administration for intraoperativefluorescence angiographyMarnie B. Welch and Laurel E. Moore

Indocyanine green (ICG; Cardio-green) is a tricar-bocyanine organic dye that has diverse clinical usesincluding cardiac dye-dilution studies, liver func-tion and blood flow determination, and ophthalmicangiography [1]. Fluorescence angiography with ICGdye is being increasingly utilized for the intraopera-tive visualization of the cerebral vasculature in patientsundergoing neurosurgical procedures..

Case descriptionThe patient was a 67-year-old American Societyof Anesthesiologists Class III female scheduled toundergo elective left pterional craniotomy for clippingof intracranial aneurysms. In the course of evalua-tion for other issues, the patient was found to havethree intracranial aneurysms: a complex middle cere-bral artery (MCA) bifurcation aneurysm (10 mm ×11 mm), a small anterior communicating aneurysm,and a small distal MCA aneurysm. Her past med-ical history included extensive tobacco use, hyper-lipidemia, well-controlled hypertension and knowncarotid atherosclerotic disease. A recent cardiac stresstest was negative for ischemia. She had no known drugallergies.

General anesthesia was induced and maintainedusing a balanced technique. She remained hemody-namically stable throughout the case, but requiredsmall doses of ephedrine intermittently to keep herwithin 20% of her baseline mean arterial pressure of65 mmHg. Four hours into the case, the large com-plexMCA aneurysmwas successfully clipped and ICGangiography was performed to confirm patency ofthe parent and branch vessels, as well as the exclu-sion of the aneurysm from the cerebral circulation.Within 2 minutes of intravenous administration ofICG 12.5mg (0.22mg/kg), the patient’s blood pressuredropped from 132/60 to 58/32 mmHg (Figure 20.1).

Thepatient’s pulse, oxygen saturation, peak inspiratorypressures, and end-tidal CO2 remained unchanged.This hypotension was poorly responsive to rou-tine measures including crystalloid, ephedrine, andphenylephrine – anaphylaxis was presumed. Smalldoses of epinephrine (10 mcg) and diphenhydraminewere administered with gradual return to a normalblood pressure. Hives became evident on her legs andarms, but no bronchospasm was observed. The neu-rosurgeons were informed of the reaction and theremaining two aneurysms were clipped promptly andwithout ICG angiography. The patient was neurologi-cally intact upon emergence and proceeded to have anuneventful recovery. She had no awareness during theperiod of significant hypotension, despite the decreasein anesthetic concentration because of her cardiovas-cular status. Her family was informed of her reac-tion to ICG dye. Mast cell tryptase levels were drawnapproximately 20 minutes after ICG administrationand returnedmarkedly elevated at 74.6 ng/ml (normal0.0–11.4 ng/ml). Cardiac enzymes were negative.

DiscussionIndocyanine green is a water-soluble dye that containsiodine as a contaminant (�5%). It is stored as a pow-der and is dissolved in the accompanying aqueous sol-vent in preparation for intravenous injection. It hasa peak spectral absorption at 800–810 nm when dis-solved in blood. It is stable in blood and plasma, is notbound to plasma proteins, and is not metabolized [2].It is excreted by the liver and secreted unchanged inthe bile; potentially harmful effects on the kidneys areminimal. Because of these unique physical characteris-tics and its negligible renal, peripheral, lung, and cere-brospinal fluid uptake, ICG has multiple clinical uses,including cerebral angiography.Dosing varies depend-ing on its indication. The manufacturer states that

66

Case 20. Anaphylaxis associated with indocyanine green administration

ICG dye administrated16:05

Figure 20.1. Intraoperative course from patient CC. Indocyanine green dye administrated at 16:05 with subsequent immediate, severehypotension, with eventual recovery. (Centricity CIS Copyright GE Medical Systems, 2002.)

dosing should not exceed 2mg/kg. However, ICG dos-ing in adult case reports has ranged from 0.5 mg/kg to5 mg/kg, or approximately 25–75 mg. Given itsreported low incidence of serious adverse reactionsand increasing indications in the neurosurgical liter-ature, its clinical use is expected to expand.

Since its introduction in the 1950s, adverse reac-tions to ICG dye have been reported, although thereare currently no reports in the neurosurgical or anes-thesiology literature. Four adverse events involving theuse of ICG were noted in the ophthalmology litera-ture in or before 1978, with an estimated 240 000 pro-cedures performed during that same time period [1].A 1994 review of 1226 ophthalmology patients receiv-ing ICG dye revealed one (0.05%) patient with a severereaction [3]; a similar incidence was reported in Japanthat same year in a review of 2820 ophthalmologypatients [4]. Mild (0.15%) or moderate (0.2%) adversereactions such as nausea, vomiting, sneezing, and itch-ing were more common [3].Themanufacturer reportsreactions to be one in every 42 000 doses [1]. In the car-diology literature, a higher incidence (4.3%; 4 out of 93patients) developed reactions in cardiac output deter-mination studies [5]; this was attributed to a higherdosage. In comparison, adverse reactions to radiocon-trast media appear higher than with ICG dye [3].

Adverse reactions to ICG dye vary both in systeminvolvement and severity. In one striking case report[1] the patient developed asphyxiation, tachycardia,and hypotension during a cardiac catheterization andcould not be resuscitated.The cause of death at autopsy

was determined to be anaphylaxis, with multiple con-tributing patient factors including coronary occlu-sion and liver failure. A second fatal case report byNanikawa et al. was attributed to laryngospasm [6].In general, cardiopulmonary resuscitation is success-ful for ICG adverse reactions, but few data exist onthe subject. Additional reactions documented in casereports include: urticaria, erythema, pruritis, hypoten-sion, nausea, extreme dyspnea, pulmonary congestion,and laryngospasm. Our anesthetized patient experi-enced only two signs: severe hypotension and hives.Reactions to ICG dye usually develop quickly, as illus-trated in our case. Patients have also received repeateddoses of ICG over time, with worsening of symptomswith each administration [2]. To our knowledge, ourpatient had no prior exposure to ICG dye.

Treatment in case reports has included intravenouscrystalloid and colloids, airway management if neces-sary, corticosteroids, epinephrine, diphenhydramine,beta-agonist nebulizers, and theophylline. The man-agement of our patient followed established treatmentof anaphylaxis; of note, epinephrine proved to bemuchmore effective than other vasopressors.

Risk factors for developing adverse reactions toICG dye are unknown. It was initially proposedthat patients with iodine sensitivity were susceptiblebecause of the solubilizing iodine component of thepharmaceutical product [1], but this has been refutedby a large case series [2]. The manufacturer doesstate, however, that a history of allergy to iodine is acontraindication to ICG administration, because the

67


product itself contains iodine. Seventy-five percent ofuremic patients were found to have adverse reactionsin one evaluation [7]. Iseki noted that higher levelsof eosinophils in chronic hemodialysis patients werefound pre-ICG administration in patients both withand without reactions; the author then proposed thisas the reason patients with end-stage renal diseaseseem to be at risk for adverse reactions to ICG [5].It has been previously suggested that atopic individu-als as well as a higher dose of ICG may predispose toadverse reactions, but a review of 17 cases by Benya donot support these as risk factors [2]. Our patient hadnone of these potential risk factors.

Both anaphylactoid and nonallergic reactions havebeen proposed as possible mechanisms for ICG dyereactions. Nonallergic (or anaphylactoid) reactionsare triggered by mast cell degranulation and are notimmune mediated. Drugs associated with nonallergicreactions include radiocontrast agents and opiates. Asthe reaction is non-immunologic, prior exposure toa drug is not necessary for the development of thereaction. Anaphylaxis, in contrast, is a true immuno-logic reaction involving IgE release defining it as atype I hypersensitivity reaction.The twodifferent typesof reactions are treated similarly; although true ana-phylaxis is generally a more significant reaction. Acase series with an unusually high incidence of ICGreactions occurred with higher doses of the medica-tion [8]. These observations led the author to pro-pose a dose-dependentmechanism thatmight supporta nonallergic mechanism. But this proposed relation-ship was determined by a small number of cases andthus it is difficult to confidently ascertain a significantincrease in incidence with a higher ICG dose. A caseseries by Michie et al. attempted to demonstrate thatadverse reactions to ICG are nonallergic in nature [7].The author states that the uremic patients in his caseseries have a decreased immunologic responsivenessand thus it is unlikely the high incidence of adversereactions in this population could be due to anaphy-laxis. However, patients do experience a prior sensiti-zation to ICG dye with escalating symptoms, support-ing anaphylaxis [2]. Patients, including ours, have hadserious symptoms from ICG, but this does not conclu-sively support whether the reaction is nonallergic oranaphylactoid in nature.

Laboratory tests can assist in confirming the natureof hypotension under anesthesia. Our patient had amarkedly elevated mast tryptase level. Tryptase is aneutral protease that is found almost exclusively in

mast cells. When mast cells are activated during ana-phylaxis, tryptase is released, along with other medi-ators such as histamine, for up to six hours afterthe reaction. Increased concentrations of mast celltryptase are a highly sensitive indicator of anaphy-lactic reactions during anesthesia [9]. Their presencefavors an IgE-mediated cause, based on a reviewof 416 specimens in anesthetic cases and confirma-tion with intradermal testing, radioimmunoassay, orboth. However, an elevated mast cell tryptase doesnot always distinguish between an anaphylactic andnonallergic reaction. Tryptase elevation is usually lesspronounced in nonallergic reactions, suggesting animmune-mediated reaction for our patient.

ConclusionAwareness of adverse reactions associated with ICGdye is imperative given its increasing use in neu-rosurgery. Anesthesiologists need to understand thepharmacokinetics, incidence, and type of adverse reac-tions, as well as the relative and absolute contraindi-cations to ICG usage. Administration of the dye topatients with a history of prior reaction to ICG, end-stage renal disease, and allergy to iodine (which thepreparation of ICG contains) should be avoided. Aswith other radiocontrast agents, an appropriate resus-citative plan and equipment must be available.

References1. T. R. Carski, B. J. Staller, G. Hepner et al. Adverse

reactions after administration of indocyanine green.J AmMed Assoc 1978; 240: 635.

2. R. Benya, J. Quintana, B. Brundage. Adversereactions to indocyanine green: a case report andreview of the literature. Cathet Cardiovasc Diagn 1989;17: 231–3.

3. M. Hope-Ross, L. A. Yannuzzi, E. S. Gragoudaset al. Adverse reactions due to indocyanine green.Ophthalmology 1994; 101: 529–33.

4. A. Obana, T. Miki, K. Hayashi et al. Survey ofcomplications of indocyanine green angiography inJapan. Am J Ophthalmol 1994; 118: 749–53.

5. K. Iseki, K. Onoyama, S. Fujimi et al. Shock causedby indocyanine-green dye in chronic hemodialysispatients. Clin Nephrol 1980; 14: 210.

6. R. Nanikawa, T. Hayashi, Y. Hashimoto et al. A caseof fatal shock induced by indocyanine green (ICG)test. Jpn J Legal Med 1978; 32: 209–14.

68

Case 20. Anaphylaxis associated with indocyanine green administration

7. D. D. Michie,D. G. Wombolt, R. F. Caretta et al.Adverse reactions associated with the administrationof a tricarbocyanine dye (Cardio-Green) to uremicpatients. J Allergy Clin Immunol 1971; 48: 235–9.

8. R. Speich, B. Saesseli, U. Hoffmann et al.Anaphylactoid reactions after indocyanine-green

administration. Ann Intern Med 1988; 109: 345–6.

9. M.M. Fisher, B. A. Baldo. Mast cell tryptase inanaesthetic anaphylactoid reations. Br J Anaesth 1998;80: 26–9.

69

Part II Vascular procedures. Aneurysm coilingCase

21 Subarachnoid hemorrhage duringaneurysm coilingKhoi D. Than, Anthony C. Wang, Neeraj Chaudhary, Joseph J. Gemmete, Aditya S. Pandey andB. Gregory Thompson

Intraprocedural rupture of an intracranial aneurysmis a potentially catastrophic complication, but can beeven more devastating in a heparinized patient inan offsite interventional suite. Understanding how torespond to such an event is essential to good patientoutcome.

Case descriptionA 67-year old woman with hypertension presented tothe emergency department with the sudden onset ofthe worst headache of her life. Upon initial neurologicexamination, she was drowsy but arousable, and wasdisoriented to place. The remainder of her examina-tion was intact. A noncontrast head computed tomog-raphy scan was obtained, which showed expected dif-fuse subarachnoid hemorrhage (SAH) and a smallhypo-attenuating spherical abnormality in the basi-lar artery tip. Small temporal horns were visible (anearly sign of hydrocephalus) and the remainder of herventricular system appeared slightly enlarged. Afterthe scan’s completion, the patient appeared increas-ingly lethargic, so a right frontal external ventriculardrain was placed and set to drain at 20 cm above thetragus. The opening intracranial (ICP) pressure was25 mmHg.

Given the patient’s presentation of a Hunt–Hessgrade 3 SAH, a diagnostic cerebral angiogram wasobtained, with the potential to perform endovascularcoil embolization should an aneurysm be discovered.The patient was brought to the interventional neurora-diology suite, where an arterial line and further intra-venous access was obtained and general anesthesiawas induced. Neurophysiologic monitoring was per-formed throughout the case. An angiogram showed a 7mm, wide-domed, small-necked saccular aneurysm atthe basilar tip. It was decided to attempt coil emboliza-tion to treat the aneurysm.

While the last coil was being injected into theaneurysm, the patient acutely became hypertensive.The ICP measurement from the ventriculostomy pre-cipitously increased to 40 mm Hg. An injection ofcontrast dye demonstrated extravasation from theaneurysm.

DiscussionThe advent of endovascular coil embolization in theearly 1990s represented a breakthrough in the care ofpatients with intracranial aneurysms. Particularly foraneurysms in locations difficult to access by conven-tional surgical approaches,manypatients nowhave theoption of endovascular treatment via a 2 mm incisionin the groin.This technique is frequently employed as aprimary modality for definitive treatment of both rup-tured and unruptured aneurysms.

As with open clip ligation, aneurysm rupture dur-ing endovascular embolization is a dreaded complica-tion [1–4]. The incidence of intraprocedural ruptureis between 2–4%, with associated mortality as high as40% (but generally 10–15%). Although the incidenceof rupture is slightly lower than in open operative pro-cedures, the associated mortality is higher given that,with coiling, the head is closed and direct access to thesource of bleeding for proximal and distal control isunavailable.

The etiologies of intraprocedural rupture includeelevation in blood pressure, increased aneurysmalintraluminal pressure secondary to contrast injection,mechanical perforation by the operator, and diversionof blood flow by inserted coils to the weak areas ofthe aneurysm wall. Intraprocedural rupture can occurboth in elective procedures, as well as in the treat-ment of previously-ruptured aneurysms. Risk factorsfor intraprocedural rupture include (1) recent rup-ture, given that the aneurysmwall is alreadyweakened,

70

Case 21. Subarachnoid hemorrhage during aneurysm coiling

(2) smaller aneurysms, given less room for error whenmanipulating guide-wires or coils, and (3) the pres-ence of a “daughter” aneurysm (a smaller associatedaneurysm).

To minimize morbidity and mortality, there arecertain steps that need to be undertaken by the anes-thesiologist in cases of intraprocedural aneurysm rup-ture.The following discussion assumes that the patientis already anesthetized and the trachea intubated. Gen-eral anesthesia is beneficial, as this allows for perfectimmobilization during coil injection and stent place-ment, and significantly decreases the risk of microvas-cular perforation once the small intracerebral vas-culature is catheterized. In addition, seizure oftenoccurs as a result of intraprocedural aneurysm rup-ture and thus airway control could potentially becompromised.

With “airway” and “breathing” already controlledand the FiO2 increased to 100%, the next step in a caseof intraprocedural rupture relates to “circulation” ormanagement of blood pressure. Aneurysmal ruptureis typically accompanied by a rise in blood pressure,which can be a cause of the rupture as well as an effectmediated by the Cushing reflex. Reduction in bloodpressure minimizes the amount and severity of resul-tant subarachnoid hemorrhage, decreases the likeli-hood of further re-rupture, and aids the intervention-alist in obtaining control. Blood pressure control willalso assist in the management of intracranial pressure.In general, a mean arterial pressure (MAP) between60 and 80 mm Hg will help to accomplish these goals.Agents such as esmolol, labetolol, nitroglycerin, nitro-prusside, and nifedipine can be used. Adenosine mayalso be used to induce a transient cessation of pul-satile blood flow to the aneurysm. Another priority atthis time is to lower the cerebral metabolic rate andthus minimize the brain’s need for blood flow. Thiscan be achieved with the use of thiopental or propo-fol, which will both decrease systemic blood pressure,as well.

Endovascular procedures always involve periodicinjections of heparin to prevent thromboembolicevents during the procedure. Clearly, this anticoagu-lation will prove problematic in the event of intrapro-cedural aneurysm rupture, as any bleeding will bemore difficult to control. Immediately after obtainingan ideal blood pressure, attention should be turned toreversing anticoagulation. Reversal of heparin usingprotamine sulfate should be performed immediately,administered at a dose of 1 mg for every 100 units

of heparin given. Given that patients often eitherpresent with, or are prescribed pre-procedural plateletinhibition with aspirin and/or clopidogrel, considera-tion should be given to platelet transfusion. In manyinstitutions, platelets are kept in the interventionalsuite throughout the procedure in case of such anevent.

After maximizing oxygen delivery, reducing bloodpressure, decreasing cerebral metabolic rate andreversing both anticoagulation and platelet inhibi-tion, the team’s next focus should be management ofincreased ICP. Immediate hyperventilation to a goalPaCO2 around 30 mmHg is a physiologic mecha-nism to acutely reduce ICP. The first-line pharmaceu-tical agent in these situations is typically mannitol, anosmotic diuretic administered at a dose of 0.25-1g/kg.Management of external ventricular drainage is adirect means by which ICP can be controlled. How-ever, when adjusting the rate of ventricular drainage,one must take care not to decompress the brain toorapidly, as this can result in aneurysmal re-rupture(recall that transmural pressure gradient = MAP –ICP, so lowering the ICP increases the pressure acrossthe wall). With unsecured ruptured aneurysms, a ven-triculostomy catheter should not drain at a level lowerthan 15–20 cm H2O above the tragus.

When subarachnoid hemorrhage occurs, seizuresmust also be prevented. Thiopental can be admin-istered for neuroprotection and anti-seizure prophy-laxis. Phenytoin (dose= 20mg/kg) is perhaps themostcommon anticonvulsant agent, however, this must beinfused at a rate no faster than 1mg/kg/min (slower forolder patients and those with cardiac disease). Rapidinfusion of phenytoin (and its accompanying solvent,polyethylene glycol) can result in severe hypotensionand even cardiovascular collapse. For this reason, thewater soluble fosphenytoin is preferred in cases ofintraprocedural rupture, as it can be infused morerapidly. Levetiracetam is a newer agent growing inpopularity due to its ease of administration and morebenign side-effect profile.

While the anesthesiologist is stabilizing thepatient’s respiratory, cardiovascular, sedative, andneurophysiologic issues, as well as managing intracra-nial hypertension, the attention of the neurosurgeonand/or interventional neuroradiologist is typicallyfocused on securing the aneurysm, as this is thedefinitive measure to control further SAH. As asalvage measure, liquid embolization administered toseal the vascular defect can be employed.

71

II. Vascular procedures. Aneurysm coiling

Inevitably, some cases that begin in the interven-tional neuroradiology suite will be unsuccessful inprotecting the aneurysm. Such cases complicated byrupture will commonly need to be brought emer-gently to an operating room for open clipping of theaneurysm. The anesthesiology team must thereforealways be prepared for expeditious transport to theoperating room in the event that an endovascular pro-cedure fails.They must also be ready to rapidly lightenanesthesia if an accurate neurologic examination isrequired.

By following the steps outlined above – controlof blood pressure, reduction of cerebral metabolism,reversal of anticoagulation and platelet inhibition,control of ICP, prevention of seizures, and timelytransport, if necessary – the neuroanesthesiologistcan be prepared for such cases of intraprocedural

aneurysm rupture and greatly improve the chances fora good patient outcome.

References1. A. Doerfler, I. Wanke, T. Egelhof et al. Aneurysmal

rupture during embolization with Guglielmidetachable coils: causes, management, and outcome.AJNR Am J Neuroradiol 2001; 22: 1825–32.

2. E. Levy, C. J. Koebbe,M. B. Horowitz et al. Rupture ofintracranial aneurysms during endovascular coiling:management and outcomes. Neurosurgery 2001; 49:807–13.

3. H. J. Priebe. Aneurysmal subarachnoid hemorrhageand the anaesthetist. Br J Anaesth 2007; 99: 102–18.

4. M. K. Varma, K. Price, V. Jayakrishnan et al.Anaesthetic considerations for interventionalneuroradiology. Br J Anaesth 2007; 99: 75–85.

72


22 Cardiac abnormalities after subarachnoidhemorrhageWilliam R. Stetler, Jr. and George A. Mashour

The intimate connection between the brain and hearthas long been speculated to affect neurologic disease.This relationship is perhapsmost clearly demonstratedin the significant cardiac dysfunction that may occurfollowing subarachnoid hemorrhage (SAH). Commoncardiac abnormalities include arrhythmias, diastolicand systolic dysfunction, and subendocardial infarc-tion. Understanding the neurocardiac axis is essentialto the clinician managing the patient with cardiac dys-function following SAH.

Case descriptionThe patient was a 54-year-old male with no signifi-cant past medical history who suddenly developed a“thunderclap headache” while playing basketball andthen collapsed onto the court. EmergencyMedical Ser-vice was alerted, the trachea was intubated, and hewas taken directly to the emergency department wherecomputed tomography (CT) of the head revealed sig-nificant SAH. Neurosurgery was consulted and a ven-triculostomy catheter was placed emergently. A CTangiography of the brain revealed a 15-mm anteriorcommunicating artery aneurysm and the patient wassubsequently taken to the angiography suite wherethe aneurysm was occluded by coil embolization. Thepatient was taken postoperatively to the neurosurgi-cal intensive care unit (ICU), intubated and on a nore-pinephrine infusion, where he was noted to exhibitflexor posturing.

On admission the patient’s cardiac troponin I(cTnI) was �0.10 ng/mL (normal �0.20 ng/mL), butafter the procedure it was noted to be 0.96 ng/mL andan electrocardiogram (ECG) was performed show-ing significant S-T segment changes (Figure 22.1).Overnight the patient had increasing oxygen require-ments and chest radiograph revealed diffuse pul-monary edema bilaterally. An uptrend of the cTnIcontinued to a peak value of 7.5 ng/mL on post-operative day 3. Falling blood pressures requiring

increasing vasopressor requirements and difficultywith oxygenation secondary to worsening pulmonaryedema continued during this time. A transtho-racic echocardiogram (TTE) was performed thatshowed an estimated left ventricular ejection fraction(LVEF) of 22% with significant anterior wall motionabnormality.

On postoperative day 12 a repeat TTE wasobtained, this time revealing an estimated LVEF of47% with only minimal anterior wall motion abnor-mality. He underwent an uneventful ventriculoperi-toneal shunt placement and after 2 more weeks in theneurosurgical ICU was transferred to the general careward, following commands in all four extremities.

DiscussionAnatomically, the heart is innervated through theintermediolateral gray column in the spinal cord andthe parasympathetic ganglia from the vagus nerve.These are in turn served by many nuclei within themedulla that are connected to cortex on many levels.Likewise, afferent fibers from the heart ascend to thenucleus tractus solitarius (NTS) and the dorsal vagalnucleus, which indirectly send fibers back to the ven-trolateral medulla [1].

However, the connection between brain and heartwas theorized long before the anatomy was under-stood.The idea of suddenunexplained death anddeathfrom fright, has been thought to arise from a con-nection between brain and heart in which a neuralinsult causes the heart to stop. Examples include deathduring grief, natural catastrophes, alcohol withdrawal,andmany other periods of significant emotional stress.It is thought that the emotional or neurologic insultis so severe that the heart ceases to function, pos-sibly as a result of a neurally induced arrhythmia.In fact there are well described ECG changes notedwith neurologic disease,most notably arrhythmias andrepolarization changes (S-T and T wave changes), as

73


Figure 22.1. Electrocardiogram of a patient with subarachnoid hemorrhage and no history of coronary artery disease. Note the S-T segmentdepressions in lead V5.

were seen in our case. Our patient developed S-Tchanges that mimicked coronary ischemia as well asrepolarization changes seen in themany T-wave inver-sions. These changes can often leave the heart vul-nerable to the development of ventricular tachycar-dia/fibrillation [1–3]. These ECG abnormalities oftenimprove with time [2] and after any circumstance thatdisconnects the heart from the brain, including hearttransplantation or brain death, offering further evi-dence towards the direct connection between heartand brain [3].

Although any neurologic insult may affect theheart, SAH is perhaps the prototype and has beenrecognized to cause considerable cardiac dysfunc-tion, including ECG changes and elevation of serumcardiac enzymes indicative of myocardial cell death.The effect on the heart following SAH significantlyincreases both morbidity and mortality, as seen inour case above with both increasing vasopressor andoxygen requirements following left ventricular dys-function. Increased levels of cTnI following neurologicinsults were found to result in higher in-hospital, all-cause mortality. With respect to SAH in particular,in-hospital mortality is nearly four times as high andpatients have more severe disability at discharge witha significant elevation of cTnI [4].

It has further been shown that the degree of neuro-logic injury as based on the Hunt–Hess grading sys-tem following SAH correlates with the elevation ofcTnI. Mild elevations of cTnI may be related to signsof diastolic dysfunction such as pulmonary conges-tion seen on chest radiograph. As cTnI increases it ismore likely that the patient will develop signs of leftventricular systolic dysfunction and a decline in LVEFobserved on echocardiogram. In our patient with ahigh grade hemorrhage, the cTnI rose rapidly and cor-related with his significant ventricular failure. Fortu-nately, cardiac dysfunction following SAH is usuallytransient in nature. Repeat echocardiogram at 5–10days post-bleed improves in nearly 75% of patientswho initially demonstrated wall motion abnormalitiesfollowing SAH [5], as was seen in our patient on hisrepeat preoperative echocardiogram several days fol-lowing the bleed.

Frequently hypothesized mechanisms of neuro-genic cardiac dysfunction include a systemic elevationof circulating catecholamines, elevated levels of serumsteroids in the setting of stress, and a direct nervoussystem stimulation of the heart. Experimental infu-sion of catecholamines has been shown to produceECG changes and histopathologic changes consistentwith neurocardiac disease. Similar pathology is seen in

74

Case 22. Cardiac abnormalities after subarachnoid hemorrhage

animal models after infusion of steroids and exposureto environmental stressors and in humans followingsignificant emotional stress producing what resemblesan acute Takotsubo cardiomyopathy. Finally, directstimulation of the hypothalami has been shown tocause distinctive pathologic changes in cardiac mus-cles less than 2–4 micrometers away from nerve fiberendings, implicating a direct neural insult to the heart[3].

The unifying link to the above proposed mech-anisms of neurogenic cardiac disease is sympatheticnervous system overactivity. It is proposed that follow-ing a neurologic insult like SAH, the brain causes a“sympathetic storm” in which there is a local increasein neurally released catecholamines, systemic eleva-tion of catecholamines from adrenal production, andan endogenous increase in the production of steroids[3]. The rise in catecholamines causes a depletion ofadenosine triphosphate, which contributes to the fail-ure of cardiac calcium channels. This results in aninflux of calcium that produces microscopic contrac-tion bands and causes failure of mitochondrial perme-ability, which leads to cell death [2].

The histopathology of neurogenic cardiac lesionsis distinct from the coagulation necrosis observed fol-lowing myocardial infarction. Lesions following neu-rologic disease (such as SAH) are characterized bycontraction band necrosis in which myocytes die in acontracted state following calcium influx. This lesionis totally separate from the coagulation necrosis seenfollowing traditional infarction [3], pointing towardsan entirely different mechanism of injury following aneurologic insult than those following local cardiacdisease. In fact, patients who have undergone cardiaccatheterization during the period of elevated cardiacenzymes following SAH have been shown to have nocoronary artery disease, redirecting the cause of car-diac dysfunction to a nonischemic etiology such asincreased sympathetic tone [2].

Although it was originally believed that cardiacmuscle was controlled predominantly by the sympa-thetic nervous system, it is increasingly accepted thatthere is a large influence of parasympathetic inner-vation to myocardium via the vagus nerve [1, 2].Thus, even as sympathetic activity is implicated as thelikely culprit surrounding neurocardiogenic pathol-ogy following SAH, more attention has recently beendirected towards parasympathetic dysfunction as animportant contributor to cardiac injury [2, 6]. Inde-pendently of neurologic disease, higher parasym-

pathetic tone has been shown to be an indepen-dent positive predictor for reduced all-cause mortal-ity. Furthermore, stimulation of the parasympatheticnervous system has been shown to cause reducedinflammation. This parasympathetic mediated anti-inflammatory response is directed through cholinergicreceptors onmacrophagesmaking the response imme-diate, as opposed to the systemic anti-inflammatoryresponse mitigated by cytokines and glucocorticoids.Following the sympathetic stimulation caused by SAH,macrophages release inflammatory cytokines, whichactivate afferent vagus fibers via the NTS. Efferentvagus fibers then reflexively stimulate the choliner-gic receptors on macrophages to decrease release ofinflammatory cytokines while at the same time fibersfrom NTS to hypothalamus cause release of adreno-corticotropin hormone to increase glucocorticoid pro-duction [6]. Dysfunction in this system following SAHresults in unregulated inflammatory changes, and isthought to contribute to cardiac pathology.

Thus, it seems that the most likely mechanismof cardiac abnormalities following SAH is both acatecholamine release after brain injury as well asparasympathetic dysfunction. Catecholamine releasedepletes adenosine triphosphate, causes increases inpermeability of calcium that leads to a calciuminflux, and eventually leads to myocardial necro-sis. At the same time, parasympathetic dysfunctionleads to unchecked inflammation that worsens adeno-sine triphosphate depletion and further contributesto myocardial cell death. Additionally, both sym-pathetic overactivity and parasympathetic dysfunc-tion result in a pro-arrhythmogenic state as well thatworsens ECG changes associated with myocardialnecrosis [2].

ConclusionIn conclusion, there are significant cardiac abnormali-ties observed following SAH that vary depending uponthe grade of SAH, but correlate with the degree ofelevation of cTnI. These effects are likely mitigatedthrough sympathetic and parasympathetic dysfunc-tion that results from global cerebral dysfunction fol-lowing SAH.

References1. A. M. Davis and B. H. Natelson. Brain-heart

interactions: The neurocardiology of arrhythmia and

75


sudden cardiac death. Tex Heart Inst J 1993; 20:158–69.

2. H. A. Mashaly and J. J. Provencio. Inflammation as alink between brain injury and heart damage: themodel of subarachnoid hemorrhage. Cleve Clin J Med2008; 75: S26–30.

3. M. A. Samuels. The brain-heart connection.Circulation 2007; 116: 77–84.

4. R. Sandhu,W. S. Aronow, A. Rajdev et al. Relationof cardiac troponin I levels with in-hospital mortalityin patients with ischemic stroke, intracerebral

hemorrhage, and subarachnoid hemorrhage. Am JCardiol 2008; 102: 632–4.

5. M. Tanabe, E. A. Crago,M. S. Suffoletto et al.Relation of elevation in cardiac troponin I to clinicalseverity, cardiac dysfunction, and pulmonarycongestion in patients with subarachnoid hemorrhage.Am J Cardiol 2008; 102: 1545–50.

6. M. H. Shishehbor, C. Alves, V. Rajagopal.Inflammation: implications for understanding theheart-brain connection. Cleve Clin J Med 2007; 74:S37–41.

76


23 Postoperative retroperitoneal bleedStacy Ritzman

Retroperitoneal hematomas are more commonlydescribed after patients have had procedures withinthe retroperitoneum (kidney or lumbar spine) ormay even arise spontaneously in an anticoagulatedpatient. Iatrogenic injuries leading to retroperitonealhematomas are becoming more common due to theincreasing popularity of interventional procedures.

Case descriptionThis case is a 71-year-old male with a past medicalhistory of anterior communicating artery (ACOM)aneurysm status post coiling 5 years previously whopresented to the emergency room with new onsetheadache, stiff neck, and slight confusion. Ahead com-puted tomography (CT) showed a left frontal lobarintracranial hemorrhage. His past medical history wasalso significant for left bundle branch block, coronaryartery disease status post percutaneous transluminalcoronary angioplasty and bare metal stent to left ante-rior descending artery two years previously, andhyper-tension. His blood pressure was 184/96 and heart ratewas 88 upon arrival.

He was taken to the interventional radiology suitewhere a pre-induction arterial line was placed; anes-thesia was induced and the trachea intubated unevent-fully.The systolic blood pressurewas closely controlledwith nitroglycerin and labetalol to maintain values�120 mmHg. After access was gained through the leftcommon femoral artery, a cerebral angiogram showeda 2 mm× 3 mm recurrence of a previously coiledACOM aneurysm at its neck. A stent was deployedacross the ACOM and the aneurysm was successfullycoiled. The patient was loaded with 600 mg of clopi-dogrel and the groin access site was closed with theStarClose device (a clip that closes the femoral arteryaccess site in a purse-string fashion). The patient didsustain a 20-second bout of supraventricular tachycar-dia intraoperatively, which resolved with intravenousesmolol but the remainder of his intraoperative

course was uneventful. The trachea was extubated atthe end of the procedure and he was neurologicallyintact. He was then transported to the neurocriticalcare unit.

Shortly after arrival at the intensive care unit (ICU),the patient became hypotensive (systolic pressure in80s) and tachycardic (heart rate 110) and complainedof nausea. He was beta-blocked with metoprolol dueto his cardiac history to treat his tachycardia. Whiletreating his blood pressure with crystalloids andphenylephrine, baseline laboratory values and cardiacenzymes were sent. Concomitantly, a stat bedside2-D echocardiogram was done to rule out a cardiacetiology for the hypotension. It was essentially normalwithout wall motion abnormalities. The patient thenbegan to complain of abdominal pain and it wasnoted on exam that his left abdomen was swollenand tense. He was taken emergently to the radiol-ogy suite where an abdominal CT revealed a largeleft-sided retroperitoneal hematoma (Figure 23.1).The patient was transfused three units of packedred blood cells (PRBC) to treat a hematocrit of 21.He was closely monitored in the ICU with serialhematocrits but required six additional units ofPRBC. He continued to be hypotensive, requiringa norepinephrine infusion. The decision was thenmade to take him emergently to the operating roomto repair the iliac artery and evacuate the hematoma.Anesthetic concerns at this time were: (1) inducinganesthesia in an acutely hypovolemic patient, (2) thepotential hemodynamic collapse once the tamponadeeffect on the retroperitoneal space was relieved,and (3) the potential for massive blood loss andtransfusion.

A central line was placed pre-induction sincehe had poor peripheral access. He was transfusedtwo units of PRBC and then was induced with eto-midate. He became profoundly hypotensive duringinduction requiring small boluses of epinephrine. Hewas maintained on isoflurane and resuscitation was

77


Figure 23.1. Computed tomography image of retroperitonealhematoma. Arrow pointing to hematoma.

continued with administration of blood and crystal-loid. After incision, his BP dropped again requiringboluses of epinephrine but the surgeon was able toquickly gain proximal control of the bleeding vessel.The posterior wall of the iliac artery had been lacer-ated and the source of the bleeding was localized andcontrolled.Thehematocrit stabilized, but he continuedto require norepinephrine to maintain his blood pres-sure. He was further resuscitated and monitored andwas extubated the following day. He was discharged tohome a week later.

DiscussionRetroperitoneal hematoma is a rare clinical entity butis becoming increasingly more common due to thehigher number of interventional procedures beingdone. It may develop after a femoral artery punctureregardless of whether a device was used to close thevessel at the end of the procedure. It is caused by thepuncture of the posterior wall of the femoral or iliacartery during cannulation. This complication may bemore likely to occur in female patients (smaller ves-sel size), a higher femoral artery puncture site (unableto compress above the inguinal ligament), and uncon-trolled hypertension [1].

The clinical manifestations of retroperitonealhematoma are vague and thus the clinician musthave a high index of suspicion to make the diag-

nosis. Initially patients may have very subtle signssuch as relative hypotension and mild tachycardiathat resolves transiently with fluid administration.They may complain of back, lower abdominal orgroin pain and may develop swelling of the groin orabdomen. Eventually they will manifest a drop inhematocrit. Our patient’s symptoms and signs wereinitially attributed to a cardiac etiology due to his pastmedical history and the inability to assess S-T segmentchanges with his coexisting left bundle branch block.There was also concern that the intraoperative runof supraventricular tachycardia may have been dueto cardiac ischemia. The correct diagnosis was notmade however until his retroperitoneal hematomahad expanded, causing abdominal pain and swelling.

Sometimes if the hematoma develops near orwithin the iliopsoas muscle, patients will present witha femoral neuropathy.This will manifest as severe painin the affected groin or hip with radiation to the ante-rior thigh and may eventually progress to paresthe-sia in the antero-medial thigh and leg paresis [1, 2]. Ifiliopsoas muscle spasm occurs due to the hematoma,the hip will be maintained in a flexed and externallyrotated position and any attempt to straighten the hipwill cause pain.

Cutaneous signs such as Grey–Turner (ecchymosisin the flank region) or Cullen (ecchymosis in the peri-umbilical region) signs are characteristically associ-ated with retroperitoneal hematoma. These, however,are late findings and will not aid in making the initialdiagnosis.

Abdominal compartment syndrome may alsodevelop with resultant impaired respiratory, cardio-vascular (impaired venous return and increased sys-temic vascular resistance leading to decreased cardiacoutput), and renal function (increased renal vascu-lar resistance and compression of renal veins andureters). This may necessitate abdominal decompres-sion through a laparotomy [2].

Diagnosis of retroperitoneal hematoma is madeeither via CT or angiography. Ultrasound of theabdomen and pelvis is usually not sensitive enoughto diagnose retroperitoneal hematoma as blood in theretroperitoneum will often pass into the abdominal orpelvic cavity, confusing the diagnosis [1]. Computedtomography, however, is highly sensitive in diagnosisand visualization of extravasating contrast will locatethe source of bleeding.

Management of retroperitoneal hematomadepends on the clinical status of the patient. All

78

Case 23. Postoperative retroperitoneal bleed

patients should be monitored in an ICU with fluidresuscitation, blood transfusion, and correction ofcoagulopathy if it exists. If the patient is hemodynam-ically stable, without evidence of ongoing bleeding,conservative management should be tried with theabove measures. However, if the patient is unstableor has ongoing bleeding, endovascular therapy withstent-grafting across the injured vessel is a treatmentoption if interventional radiology is available. If not,surgery can be undertaken realizing that it may releasethe effect of the tamponade and initially make mattersworse.This occurred transiently in our patient but waspromptly treated with epinephrine and proximal con-trol of the bleeding vessel by the surgeon. Surgery mayalso be indicated to decompress the retroperitonealspace if nerve or ureteral compression exists [3].

ConclusionRetroperitoneal hematoma is an uncommon clinicalentity that may be encountered more frequently as

iatrogenic injuries occur during interventional pro-cedures. Since signs and symptoms are initially non-specific, it requires a high index of suspicion onthe part of the clinician for prompt diagnosis andtreatment.

References1. Y. C. Chan, J. P. Morales, J. F. Reidy, P. R. Taylor.

Management of spontaneous and iatrogenicretroperitoneal haemorrhage: conservativemanagement, endovascular intervention or opensurgery? Int J Clin Pract 2008; 62: 1604–13.

2. C. Gonzalez, S. Penado, L. Liata et al.The clinicalspectrum of retroperitoneal hematoma inanticoagulated patients.Medicine 2003; 82(4):257–62.

3. S.I. Daliakopoulos, A. Bairaktaris,D. Papadimitriouet al. Gigantic retroperitoneal hematoma as acomplication of anticoagulation therapy with heparinin therapeutic doses: a case report. J Med Case Reports2008; 2: 162.

79


24 Angiography in the patient withkidney failureJerome O’Hara and Mauricio Perilla

IntroductionContrast-induced nephropathy (CIN) from iodinecontrast media during radiologic procedures is one ofthe most common causes of acute kidney injury. Thisrisk increases in patients with chronic renal insuffi-ciency. Balancing efforts to control blood pressure andprevent further worsening of renal function duringcerebral artery coiling after subarachnoid hemorrhagecan be challenging.

Case descriptionA 65-year-old, 88-kg female presented for cerebralangiography and planned coiling of a large middlecerebral artery aneurysm. During the physical exam-ination in the emergency department, the patientwas lethargic and complained of a headache. Bloodpressure was 205/95 mmHg and oxygen saturationwas 98% on supplemental nasal cannula oxygen. Ini-tial management included a noncontrast computedtomography (CT) scan, which diagnosed extensivesubarachnoid hemorrhage suggestive of aneurysmalrupture.The patient was admitted to the neurosurgicalintensive care unit for observation and invasive arterialblood pressure monitoring. The patient’s medical his-tory included chronic renal insufficiency (admissionserum creatinine = 1.8 mg/dL) secondary to non-insulin dependent diabetes and uncontrolled hyper-tension. On the day of the planned coiling serum cre-atinine was 2.3 mg/dL, serum glucose was 91 mg/dL,and arterial pressure was 130/70 mmHg.

In preparation for the procedure the patient wasmaintained on a normal saline intravenous infusionsupplemented with a sodium bicarbonate solutioninfusion during and after the procedure. This was ini-tiated in an attempt to prevent further decline in renalfunction when additional contrast dye exposure wasadministered during angiography. General anesthesiawith endotracheal intubation, central venous pressure

monitoring, and careful blood pressure control wasplanned for the aneurysm coiling procedure.

DiscussionAnesthetic concerns included (1) aggressive arte-rial blood pressure management to prevent fur-ther intracranial hemorrhage prior to completinganeurysm coiling, (2) maintenance of euglycemia, and(3) managing intravascular volume within the bal-ance of providing adequate renal blood flow with-out risking volume overload. With this patient havingchronic renal insufficiency and uncontrolled hyper-tension prior to the procedure, renal autoregulatorymechanisms tomaintain optimal renal perfusion likelyrequired higher systemic blood pressure. This poses agreat challenge to balance adequate renal perfusion yetmaintain lower blood pressure to avoid an extension ofcerebral hemorrhage until coiling is completed. Cen-tral venous pressure and urine output can only act as aguide in assessing renal perfusion, but less so if diuret-ics have been administered.

Contrast-induced nephropathy from the adminis-tration of iodinated contrast media remains a poten-tially serious complication associated with angio-graphic procedures. The many therapeutic effortsstudied to ameliorate CIN have met little success overthe last two decades and the incidence remains at∼12% [1]. In patients without risk factors for CIN,the incidence is lower (�5%), but as high as 50% inpatients with multiple risk factors. It has been wellestablished that patients who develop CIN have anincreased risk of morbidity, prolonged hospitalization,and mortality [2].

Risk factors for CIN include chronic renal insuf-ficiency, diabetes mellitus, advanced age, congestiveheart failure, nephrotoxic drug use, metabolic syn-drome, hyperuricemia, hypovolemia, and large vol-umes of contrast media exposure [1, 2]. The most

80

Case 24. Angiography in the patient with kidney failure

important risk factor for CIN remains chronic renalinsufficiency.

With more angiographic procedures being per-formed in the setting of cerebral aneurysm coil-ing, cerebral and coronary artery angioplasty/stenting,anti-arrhythmic cardiac ablations, and major vascularstenting, the ability to minimize the incidence of CINcarries a major impact on both patient outcomes andmedical costs. This realization has led to many clini-cal studies in an attempt to determine if there is a “bestpractice” standard in the setting of angiography to pre-vent CIN.

Etiology and diagnosis of contrast-inducednephropathyThe pathophysiology of CIN is suspected to berelated to contrast media-triggered vasoconstrictionand development of oxidative stress leading to accu-mulation of reactive oxygen species [3]. Contrast-induced nephropathy is defined as an absolute increasein baseline serum creatinine concentration of at least0.5 mg/dL or with a relative increase of 25% in the set-ting of chronic renal insufficiency. There must be noalternate etiology and the event must occur within 48–72 hours after contrast exposure [4].

Potential therapeutic interventionsIn the setting of potential exposure to contrast dye dur-ing angiography, the focus should be the identifica-tion of patients at risk and subsequent prevention ofCIN rather than treatment after exposure has alreadyoccurred. Consensus exists that proper intravascularhydration is the most important first-line therapy forprevention of CIN. The electrolyte solution of choiceis 0.9% (vs 0.45%) saline [5]. The choice of the iodi-nated contrast dye is very important and should eitherbe an iso-osmolar or low-osmolar contrast media.

Systemically administered vasodilators, suchas dopamine agonists (dopamine and fenoldopam),adenosine antagonists, prostaglandins, and endothelinantagonists have presented disappointing conclusionswhen studied in an attempt to lessen the incidenceof CIN [6]. It remains inconclusive if theophylline,ascorbic acid, or simvastatin provide any advantage inthis clinical setting [1]. Furosemide has been reportedto increase the risk of developing CIN [7]. Multiplestudies evaluating N-acetylcysteine and bicarbonate to

prevent CIN have concluded these therapies “appear”to lessen the incidence of CIN, but are not conclusive.

Numerous prophylactic strategies have been stud-ied, which have allowed several meta-analyses to bepublished on different single therapies in CIN. Theconclusion of many of the meta-analyses reflect onthe heterogeneous nature of the studies, publicationbias, inability to absolutely recommend a therapy, andsuggest that further adequately powered studies areneeded even if it appeared a benefit was derived.

Alkalinization of renal tubular fluid with bicarbon-ate may prevent the development of CIN. Renal injurymay be reduced by slowing the free radical produc-tion in the renal medulla [2]. In ameta-analysis, Meieret al. [3] reported that the use of sodium bicarbonatereduced the incidence of CIN, but revealed no differ-ence in the need for renal replacement therapy ormor-tality. Two recently published randomized controlledstudies suggest hydration with normal saline was asefficacious as using sodiumbicarbonate for the preven-tion of CIN [8, 9]. Hemodialysis has been reported toreduce the degree of CIN in patients with chronic renalinsufficiency after angiographic procedures [10]. Thisinvasive therapeutic modality reflects additional med-ical costs and potential morbidity, but may be of ben-efit in this specific patient population to lessen furtherrenal injury.

ConclusionWe presented a case that necessitated continued peri-operative assessment of renal function during cere-bral aneurysm coiling in a patient with chronic renalinsufficiency.The co-morbidities of this patient placedher at high risk for CIN, which could lead to furtherimpaired renal function and a negative outcome. Thebest plan for preventing acute renal failure secondaryto CIN includes hydration with normal saline, intra-venous sodium bicarbonate infusion prior to the pro-cedure, minimization of dye exposure, and considera-tion of postprocedure hemodialysis.

References1. D. Reddan,M. Laville, V. D. Garovic. Contrast-

induced nephropathy and its prevention: what do wereally know from evidence-based findings? J Nephrol2009; 22: 333–51.

2. M. Kanbay, A. Covic, S. G. Coca et al. Sodiumbicarbonate for the prevention of contrast-induced

81


nephropathy: a meta-analysis of 17 randomized trials.Int Urol Nephrol 2009; 41: 617–27.

3. P. Meier,D. T. Ko, A. Tamura et al. Sodiumbicarbonate-based hydration prevents contrast-induced nephropathy: a meta-analysis. BMCMed2009; 7: 23.

4. S. K. Morcos,H. S.Thomsen, J. A. Webb. Contrast-media-induced nephrotoxicity: a consensus report.Contrast Media Safety Committee, European Societyof Urogenital Radiology (ESUR). Eur Radiol 1999; 9:1602–13.

5. H. S. Trivedi,H. Moore, S. Nasr et al. A randomizedprospective trial to assess the role of saline hydrationon the development of contrast nephrotoxicity.Nephron 2003; 93: C29–34.

6. A. M. Kelly, B. Dwamena, P. Cronin et al.Meta-analysis: effectiveness of drugs for preventingcontrast-induced nephropathy. Ann Intern Med 2008;148: 284–94.

7. M. Rudnick,H. Feldman. Contrast-inducednephropathy: what are the true clinical consequences?Clin J Am Soc Nephrol 2008; 3: 263–72.

8. A. M. From, B. J. Bartholmai, A.W.Williams et al.Sodium bicarbonate is associated with an increasedincidence of contrast nephropathy: a retrospectivecohort study of 7977 patients at Mayo clinic. Clin J AmSoc Nephrol 2008; 3: 10–18.

9. P. Schmidt,D. Pang,D. Nykamp et al.N-acetylcysteine and sodium bicarbonate versusN-acetylcysteine and standard hydration for theprevention of radiocontrast-induced nephropathyfollowing coronary angiography. Ann Pharmacother2007; 41: 46–50.

10. P. T. Lee, K. J. Chou, C. P Liu et al. Renal protectionfor coronary angiography in advanced renal failurepatients by prophylactic hemodialysis. A randomizedcontrolled trial. J Am Coll Cardiol 2007; 50: 1015–20.

82

Part II Vascular procedures. Arteriovenous malformationCase

25 Postoperative normal perfusion pressurebreakthroughJuan P. Cata and Andrea Kurz

Altered flow dynamics after the surgical correction ofvascular malformations create the risk for perfusionabnormalities in the postoperative period. Vigilancefor edema or hemorrhage due to this complication isimperative in the postoperative setting.

Case descriptionA 37-year-old female with a 2-month history ofgeneralized tonic-clonic seizures underwent a suc-cessful left craniotomy for clipping and resectionof a 4 cm (in maximum diameter) arteriovenousmalformation (AVM) located in the left parieto-occipital lobe. Six hours after surgery the patientpresented with sudden and progressive right-sidedmotor weakness and rapid deterioration in mentalstatus. She subsequently required emergent trachealre-intubation due to hypercapnic respiratory failureand inability to protect her airway. Emergent com-puted tomography (CT) scan of the brain showedmassive cerebral edema, as well as enlarged vascularenhancement suggesting hyperperfusion and a smallintracerebral hemorrhage. A single photon emis-sion computed tomography (SPECT) and cerebralangiography showed marked hyperperfusion with aregional cerebral blood flow of 73 mL/100 g/min andleft posterior cerebral artery dilatation, respectively.Once admitted to the intensive care unit the patientwas placed in a barbiturate coma, started on fluidrestriction and aggressive systemic blood pressurecontrol. Four days after surgery, a CT brain scanshowed significant reduction in brain swelling and nosigns of cerebral hyperperfusion. On postoperativeday five, the barbiturate coma was stopped, thepatient was successfully extubated and her neurologicstatus gradually improved. One month after surgery,examination demonstrated no neurologic deficitand a cerebral angiography showed normalizationof flows and diameter of the left posterior cerebralartery.

DiscussionNormal perfusion pressure breakthrough (NPPB) isa major cause of morbidity and mortality followingAVM surgery. It can also be present in the postop-erative period of carotid endarterectomies and Galenvein malformations. The actual incidence of this vas-cular phenomenon is not well known, but Young et al.found an incidence of 2.6% [1]. Risk factors for NPPBinclude a large-sized AVM, large-diameter feeders andan AVM located in a border zone. The NPPB eti-ology is unclear. The brain tissue surrounding theAVM is supplied by blood vessels showing impairedautoregulation, consequently a rapid resumption of“normal” blood flow after AVM occlusion leads tohyperemia and vascular leak. Also, prolonged hypop-erfusion results in the development of new bloodvessels in the brain tissue surrounding the AVM.The sudden occlusion of the AVM after clamping orembolization resulting in an abrupt increase in per-fusion through the newly formed blood vessels maybe responsible for capillary leak and brain edema(Figure 25.1). Another hypothesis suggests that NPPBmight be associated with impaired autonomic perivas-cular innervation in vascular beds proximal and dis-tant (contralateral) to the AVM, which can explain thefact that global CBF may be increased after resectionof the AVM [1]. “Occlusive hyperemia” [2] has alsobeen suggested as a hypothesis in which postoperativeflow restriction and spontaneous thrombosis of thevenous system associated with the AVM causes occlu-sive hyperemia, which in turn is responsible for capil-lary dysfunction and rupture of the blood–brain bar-rier [3]. Abnormal vasomotor reactivity has also beendemonstrated in peri-AVM areas of the brain [4].

The clinical manifestation of NPPB ranges fromvague symptoms such as nausea, vomiting, andheadache to more severe problems such as aphasia,hemiparesis, rapid deterioration of mental status, orseizures. Postoperative neuroimaging studies (CTscan, SPECT, and magnetic resonance imaging)

83

II. Vascular procedures. Arteriovenous malformation

Figure 25.1. The picture depicts the vascular changes before(upper panel) and after (lower panel) arteriovenous malformation(AVM) feeder occlusion. It can be noticed that after occlusion, theblood flow in collateral circulation to the AVM increases and as aconsequence hyperperfusion, tissue edema, and bleeding occur.

may demonstrate localized or generalized brainedema, hyperemia, and focal ormultifocal intracranialhemorrhage. Transcranial cerebral Doppler may showvasomotor paralysis represented by decreased flowvelocities in cerebral arteries without response to theadministration of acetazolamide or during carbondioxide challenge. Cerebral angiography may showstagnation in the feeders of the resected AVM anddelayed circulation. Preoperative measures such asstaged feeders and nidus embolization of large-size,high-grade AVM have been proposed to preventNPPB.

Aggressive intraoperative control of systemicblood pressure and intracranial pressure, as well asmeticulous hemostasis of feeders �1 mm, is recom-mended. It is unknown if the use of volatile anestheticagents or total intravenous agents have an effect onthe occurrence of NPPB. More aggressive measuressuch as carotid artery clamping and hypothermia have

been proposed to prevent NPPB. It is noteworthy thatnone of the proposed preventive modalities are basedon well-conducted clinical studies.

In the postoperative period, the use of low to nor-mal systemic blood pressure, reduced fluid intake,anticonvulsants anddiuretics are indicated in theman-agement of patients at risk for or with mild forms ofNPPB. Barbiturate coma, hyperventilation, and inva-sive intracranial pressure monitoring may be usedin those patients with generalized brain edema andincreased intracranial pressure [5]. Again, none ofthese therapeutic measures is supported by random-ized controlled trials.

Vasomotor paralysis may last for several days toweeks after surgery. Thus, neuroimaging evidence ofnormal pattern of cerebral vasoreactivity along withneurologic status improvement may warrant barbitu-rate withdrawal and careful liberalization of the bloodpressure control.

ConclusionNormal perfusion pressure breakthrough is a poten-tially catastrophic event afterAVMsurgery. Anesthesiaproviders should strive for tight perioperative bloodpressure control and should be vigilant for signs ofpostoperative neurologic deterioration.

References1. W. L. Young, A. Kader, E. Ornstein et al. Cerebral

hyperemia after arteriovenous malformation resectionis related to “breakthrough” complications but not tofeeding artery pressure. The Columbia UniversityArteriovenous Malformation Study Project.Neurosurgery 1996; 38: 1085–93.

2. N. R. al-Rodhan, T. M. Sundt Jr, D. G. Piepgras et al.Occlusive hyperemia: a theory for the hemodynamiccomplications following resection of intracerebralarteriovenous malformations. J Neurosurg 1993; 78:167–75.

3. J. Hai,Q. Lin, S. T. Li et al. Chronic cerebralhypoperfusion and reperfusion injury of restorationof normal perfusion pressure contributes to theneuropathological changes in rat brain. Brain Res MolBrain Res 2004; 126: 137–45.

4. H. H. Batjer,M. D. Devous Sr.The use ofacetazolamide-enhanced regional cerebral bloodflow measurement to predict risk to arteriovenousmalformation patients. Neurosurgery 1992; 31:213–17.

5. D. Chyatte. Normal pressure perfusion breakthroughafter resection of arteriovenous malformation. J StrokeCerebrovasc Dis 1997; 6: 130–6.

84

Part II Vascular procedures. Carotid endarterectomyCase

26 Preoperative evaluationMaged Argalious

Stroke remains the third leading cause of death over-all, and the second leading cause of death for womenin the USA [1]. It has been estimated that 103 000carotid endarterectomies (CEA) were done in the USAin 2005 [1]. Among patients undergoingCEA, the inci-dence of coronary artery disease is around 35% [2],and 25% of patients undergoing CEA have concomi-tant diabetes mellitus [2], highlighting the importanceof preoperative evaluation and optimization in thispatient population.

Case descriptionA 76-year-old male was scheduled for a right CEAunder general anesthesia after an episode of transientmonocular blindness (amaurosis fugax) prompted aduplex ultrasound of the carotid vessels. This showeda �90% occlusion of the right internal carotid arteryas well as a 50–79% stenosis of the left internal carotidartery [3]. His medical history included hypertension,coronary artery disease with a documented non S-Tsegment elevationmyocardial infarction (NSTEMI) 15months prior, which required a percutaneous coro-nary intervention (PCI) with a cypher drug-elutingstent to the right coronary artery. He was able to walkaround the house until a year ago when osteoarthri-tis of his knees markedly reduced his activities of dailyliving. Hismedications includedmetoprolol, amlodip-ine, atorvastatin, aspirin, and clopidogrel.

At the time of his NSTEMI, a 2-D echocardio-gram revealed basal and mid-inferior and inferoseptalhypokinesis with a left ventricular ejection fraction of42%. Preoperative laboratory work-up was significantfor a hematocrit of 33 and a serum creatinine of 1.6mg/dL. The vascular surgeon consulted the anesthesiateam for preoperative evaluation.

DiscussionCarotid endarterectomy is considered an intermedi-ate risk surgery. The classification of cardiac risk in

noncardiac surgery is based on the incidence of car-diac death and nonfatal myocardial infarction (1–5%in intermediate risk surgeries).

The goal of preoperative preparation for noncar-diac surgery is to:

1. Assess the need for emergent or urgent surgerythat cannot be postponed even if the risk ofcoronary artery disease is high.

2. Exclude active cardiac conditions that requirepreoperative evaluation and treatment beforenoncardiac surgery (see Table 26.1).

3. Assess functional capacity since asymptomaticpatients with good functional capacity (four ormore metabolic equivalents) can proceed withtheir planned surgery and rarely require furthernoninvasive testing since their management willunlikely be changed based on the results of testing.

4. Assess the presence of clinical risk factorsaccording to the Lee’s revised cardiac risk index(see Table 26.2) in patients with low or unknownfunctional capacity to determine the need forfurther cardiac testing especially in patients withthree or more clinical risk factors and thoseundergoing vascular surgery (high incidence ofunderlying coronary artery disease). Theconsensus is to only perform further cardiactesting (noninvasive: i.e., functional stress test;invasive: i.e., cardiac catheterizations) if the resultsof those tests are likely to change management.

While the American College of Cardiology/AmericanHeart Association (ACC/AHA) offers an algorithm forevaluating cardiovascular risk [4], there are some clini-cal scenarios with inconclusive data to determine “beststrategy” and more than one plausible option can befollowed. This patient had no active cardiac condi-tions, had an unknown functional capacity and hadthree clinical risk factors (history of heart disease, cere-brovascular disease, and renal insufficiency) and wasundergoing an intermediate risk vascular surgery. The

85

II. Vascular procedures. Carotid endarterectomy

Table 26.1. Active cardiac conditions for which the patientshould undergo evaluation and treatment before noncardiacsurgery (Class I, Level of Evidence: B).

Condition Examples

Unstable coronarysyndromes

Unstable or severe angina∗ (CCS class IIIor IV)†

Recent MI‡

DecompensatedHF (NYHAfunctional classIV; worsening ornew-onset HF)

Significantarrhythmias

High-grade atrioventricular blockMobitz II atrioventricular blockThird-degree atrioventricular heart blockSymptomatic ventricular arrhythmiasSupraventricular arrhythmias (including

atrial fibrillation) with uncontrolledventricular rate (HR greater than 100beats per minute at rest)

Symptomatic bradycardiaNewly recognized ventricular

tachycardia

Severe valvulardisease

Severe aortic stenosis (mean pressuregradient �40 mmHg, aortic valve arealess than 1.0 cm2, or symptomatic)

Symptomatic mitral stenosis(progressive dyspnea on exertion,exertional presyncope, or HF)

∗ According to L. Campeau [8].† May include “stable” angina in patients who are unusuallysedentary.‡ The American College of Cardiology National Database Librarydefines recent MI as more than 7 days but less than or equal to1 month (within 30 days).CCS indicates Canadian Cardiovascular Society; HF, heart failure;HR, heart rate, MI, myocardial infarction; NYHA, New York HeartAssociation.Reprinted with permission from L. A. Fleisher, J. A. Beckman, K. A.Brown et al. ACC/AHA 2007 [1].

Table 26.2. Clinical risk factors [7].

History of heart disease

History of compensated or prior heart failure

History of cerebrovascular disease

Diabetes mellitus

Renal insufficiency

first option is to proceedwith the planned surgerywithheart rate control and no further testing.This would bebased on the following rationale:� This patient had an episode of amaurosis fugax

and the right CEA could be considered urgent andnot totally elective.

� Even if functional noninvasive stress testingshowed reversibly ischemic myocardium, PCIwith balloon angioplasty would requirepostponement of the CEA for 2–4 weeks to allow

healing of the vessel injury, while PCI with a baremetal stent (BMS) or a drug eluting stent (DES)may require postponing surgery for 4 weeks and12 months respectively especially if the surgeondoes not want to operate while patients arereceiving dual antiplatelet therapy (aspirin andclopidogrel) [5]. The surgeon’s rationale forstopping clopidogrel preoperatively was that anypostoperative bleeding into a closed space (neck)could result in catastrophic airway compromise.

The second option was to proceed with noninvasivecardiac testing with the following rationale:� Patients undergoing vascular surgery have a high

incidence of concomitant coronary artery diseaseand cardiac causes are the most common causes ofmorbidity and mortality after CEA.

� Studies document the beneficial effects ofclopidogrel combined with aspirin in reducingcerebral emboli in patients undergoing CEA [6].Therefore, even if this patient required a PCI witha BMS or DES, surgery should proceed with dualantiplatelet therapy.

ConclusionThe role of the anesthesiologist as a perioperativeconsultant requires a thorough knowledge of currentguidelines, a deep understanding of perioperative riskassociated with various noncardiac surgeries in orderto follow an evidence-based approach to perioperativemanagement.

References1. W. Rosamond, K. Flegal, K. Furie et al. Heart disease

and stroke statistics – 2008 update: a report from theAmerican Heart Association Statistics Committee andStroke Statistics Subcommittee. Circulation 2008; 117:e25–146.

2. GALA Trial Collaborative Group, S. C. Lewis, C. P.Warlow. General anesthesia versus local anesthesia forcarotid surgery (GALA): a multicentre, randomizedcontrolled trial. Lancet 2008; 372: 2132–42.

3. W. S. Moore,H. J. Barnett,H. G. Beebe et al.Guidelines for carotid endarterectomy. Amultidisciplinary consensus statement for the Ad HocCommittee, American Heart Association. Circulation1995; 91: 566–79.

4. L. A. Fleisher, J. A. Beckman, K. A. Brown.ACC/AHA 2007 Guidelines on PerioperativeCardiovascular Evaluation and Care for NoncardiacSurgery. Executive Summary: A Report of the

86


American College of Cardiology/American HeartAssociation Task Force on Practice Guidelines(Writing Committee to Revise the 2002 Guidelines onPerioperative Cardiovascular Evaluation forNoncardiac Surgery): Developed in CollaborationWith the American Society of Echocardiography,American Society of Nuclear Cardiology, HeartRhythm Society, Society of CardiovascularAnesthesiologists, Society for CardiovascularAngiography and Interventions, Society for VascularMedicine and Biology, and Society for VascularSurgery. Circulation 2007; 116: 1971–96.

5. American Society of Anesthesiologists Committeeon Standards and Practice Parameters. Practice alertfor the perioperative management of patients with

coronary artery stents. A report by the AmericanSociety of Anesthesiologists Committee on Standardsand Practice Parameters. Anesthesiology 2009; 110:22–3.

6. D. A. Payne, C. I. Jones, P. D. Hayes. Beneficial effectsof clopidogrel combined with aspirin in reducingcerebral emboli in patients undergoing carotidendarterectomy. Circulation 2004; 109: 1476–81.

7. T. H. Lee, E. R. Marcantonio, C. M. Mangione et al.Derivation and prospective validation of a simpleindex for prediction of cardiac risk of majornoncardiac surgery. Circulation 1999; 100: 1043–9.

8. L. Campeau. (Letter): Grading of angina pectoris.Circulation 1976; 54: 522–3.

87


27 MonitoringmodalitiesTodd Nelson and Paul Picton

Carotid endarterectomy (CEA) has been proven toreduce the incidence of embolic stroke in symptomaticpatients with �70% internal carotid artery stenosis[1]. The perioperative risk of stroke is significant [2],the pathogenesis of which include emboli, hemor-rhage, and hypoperfusion. Intraoperative neurologicmonitoring has been the subject of intense researchfor many years, the goal being to accurately identifyintraoperative cerebral ischemia, and predict whichpatients may benefit from intraoperative shunting.

Case descriptionThe patient was a 73-year-old male scheduled for aright CEA under general anesthesia. He presentedwith recurrent transient right-sided visual loss andhad a past medical history significant for well-controlled hypertension, ischemic heart disease, andobesity. Carotid Doppler revealed 79% stenosis ofhis right internal carotid artery and 30% stenosis ofhis left.

He was premedicated with midazolam and a 20-gauge catheter was placed in his left radial arteryfor blood pressuremonitoring. Following preoxygena-tion, anesthesia was induced with fentanyl and propo-fol; muscle relaxation was achieved with vecuronium.Anesthesia was maintained with isoflourane, oxygen,air, and remifentanil. Phenylephrine was titrated byinfusion to maintain a stable arterial blood pressurethatwas increased by nomore than 25%above baselineduring cross-clamping. The mean stump pressure was64 mmHg; shunting was not used. The carotid cross-clamp time was 35 minutes. Following cerebral reper-fusion, an intraoperative angiogram was performedand was unremarkable.

The patient experienced delayed emergence anddisplayed signs of a left hemiparesis. Once fully awakehe was safely extubated and was sent immediatelyfor imaging. A diffusion-weighted magnetic reso-

nance imaging confirmed the presence of a right-sidedischemic stroke in the middle cerebral artery (MCA)territory.

DiscussionAwake neurologic monitoring is considered the goldstandard for evaluation of intraoperative neurologicstatus during CEA. The development of new neu-rologic symptoms is an indication of ischemia andnecessitates shunt placement. For patients undergo-ing CEA under general anesthesia a number of moni-toring modalities exist (Table 27.1): monitors of cere-bral hemodynamics, monitors of cerebral oxygenationand metabolism, and monitors of electrophysiologicparameters [3].

Monitors of cerebral hemodynamicsTranscranial Doppler sonography can measure bloodflow in theMCA, but can be difficult to localize in a sig-nificant proportion of patients. It demonstrates peri-operative embolic showers and decrements in bloodflow during cross-clamping. Carotid artery stumppressure provides an indication of the degree of perfu-sion from the contralateral circulation via the circle ofWillis. The pressure in the carotid stump is measuredimmediately following placement of the carotid cross-clamp.Although invasive, it is technically easy. Ameanpressure of �50 mmHg is commonly regarded as anindication of sufficient collateral flow, though a provencutoff value remains under investigation.

Monitors of cerebral oxygenationand metabolismRegional cerebral oximetry measured using near-infrared spectroscopy is convenient and noninva-sive. Variability in pad placement can lead to unre-liable regional oximetry readings. High inter- and

88

Case 27. Monitoring modalities

Table 27.1. Comparison of cerebral monitoring modalities for CEA under general anesthesia.

Cerebral monitor Advantages Limitations

Cerebral hemodynamicsTranscranial Doppler

Stump pressure

Allows direct evaluation of blood flow in MCAPossible to visualize embolic showers

Simple, inexpensiveNo additional equipment or personnel needed

MCA sometimes difficult to localizeOnly provides information regarding MCA territoryHigh likelihood of intraoperative probe dislocationNo consensus on appropriate threshold value forischemia

One time measurement

Cerebral oxygenation andmetabolismRegional cerebral oximetry

Jugular bulb monitoring

NoninvasiveEasy, rapid to applyContinuous measure of cerebral oxygenationCan give information regarding cerebral O2

demand versus supply

High inter- and intrasubject variabilityUnclear contribution of non-brain blood sourcesAbsolute values less helpful than trendsProvides only a global value of oxygenation,nonspecific

Invasive, requires some skill

Electrophysiologic monitorsEEG

SSEP

Direct measure of brain electrical activity

Ability to identify lesions in the sensorypathways

Requires fewer electrodes than EEG

ExpensiveRequires constant presence of a trained technicianOverly sensitiveExpensiveRequires constant presence of a trained technicianExtra-cerebral lesions in sensory pathway caninfluence readings

MCA, middle cerebral artery; EEG, electroencephalography; SSEP, somatosensory evoked potentials.

intra-individual variability exists. Trends are probablyof greater value than absolute readings.

Jugular bulbmonitoring provides a global measureof oxygenation in the ipsilateral hemisphere duringcross-clamping: a rostrally placed oximeter is used tocontinuously measure jugular venous oxygen satura-tion. Alternatively, a jugular venous catheter is placedand periodic samples are drawn simultaneously witharterial line samples to compare pertinent values. Itis an invasive technique that cannot identify specificareas of ischemia.

Electrophysiologic monitorsElectroencephalography (EEG) directly detects elec-trical potentials within the brain, providing a measureof brain activity. Cerebral blood flow is closely relatedto cerebral electrical activity and EEG thereby pro-vides an indication of blood flow. It is very sensitive forischemia in the brain and has a better ability to local-ize ischemic areas. Electroencephalography has beenused for many years worldwide and, therefore, mostclinicians are familiar with it as a monitoring tech-

nique for CEA. Unfortunately, EEG can be overly sen-sitive for ischemia. It is costly, cumbersome, and thepresence of a trained technician is required throughoutsurgery.

With somatosensory evoked potentials, peripheralnerves (i.e.,median and posterior tibial) are stimulatedand the evoked potentials monitored via electrodesplaced on the scalp. Lesions are diagnosed withinsensory pathways. Like EEG, somatosensory evokedpotentials are expensive and require the presence of atrained technician.

The GALA trial, a multicenter, randomized con-trolled trial which enrolled over 3000 patients, evalu-ated the difference in outcomes between general anes-thesia and regional anesthesia for CEA [4]. This studyshowed no significant difference in stroke rates or 30-day mortality between CEA performed under generalanesthesia versus those performed using local anes-thesia alone. Multiple studies have been performed toattempt to determine the effectiveness of each modal-ity used perioperatively in conjunction with gen-eral anesthesia. Despite extensive research, no singlemethod has been proven to be more effective than any

89


other in identifying ischemia, nor has any been provento reduce the risk of stroke [5].

The challenge for all monitoring modalities is toidentify thresholds for ischemia and stroke. As valuesare chosen that provide increased sensitivity to brainischemia, the results become less specific.This leads tothe placement of unnecessary shunts, which exposesthe patient to the risk of stroke associated with shuntplacement.The converse is also true.No value has beenidentified for anymonitoring device that is both highlysensitive and highly specific.

ConclusionIn conclusion, a frequent basic neurologic examina-tion on an awake patient remains the gold standardfor neurologic monitoring during CEA. However, nodifference in outcomes between general anesthesia andregional anesthesia has been proven. None of themon-itoring modalities commonly used for CEA undergeneral anesthesia have been shown to either reli-ably identify or prevent cerebral ischemia or stroke,nor predict which patients may benefit from shuntplacement.

References1. North American Symptomatic Carotid

Endarterectomy Trial Collaborators. Beneficial effectof carotid endarterectomy in symptomatic patientswith high-grade carotid stenosis. N Engl J Med 1991;325: 445–53.

2. J. M. Findlay, B. E. Marchak,D. M. Pelz et al. Carotidendarterectomy: a review. Can J Neurol Sci 2004; 31:22–36.

3. S. Moritz, P. Kasprzak,M. Arlt et al. Accuracy ofcerebral monitoring in detecting cerebral ischemiaduring carotid endarterectomy: a comparison oftranscranial Doppler sonography, near-infraredspectroscopy, stump pressure, and somatosensoryevoked potentials. Anesthesiology 2007; 107: 563–9.

4. GALA Trial Collaborative Group. Generalanaesthesia versus local anaesthesia for carotid surgery(GALA): a multicentre, randomized controlled trial.Lancet 2008; 372: 2132–42.

5. R. Bond, K. Rerkasem, C. Counsell et al. Routineor selective carotid artery shunting for carotidendarterectomy (and different methods of monitoringin selective shunting). Cochrane Database Syst Rev2002; 2: CD000190.

90


28 Neurologic decline after carotid surgeryPhil Gillen

Carotid endarterectomy (CEA) is a surgical proce-dure that typically involves the removal of atheroma-tous plaque and tunica intima from the lumen of astenosed extracranial segment of the internal carotidartery (ICA), via an open approach. Due to themarkedchanges in cerebral perfusion, there is a spectrum ofneurologic sequelae directly attributable to this pro-cedure, ranging from intraoperative stroke and deathtomore subtle perioperative neurocognitive deteriora-tion.This decline is becoming increasingly recognizedas a consequence of CEA.

Case descriptionThe patient was a 68-year-old male with demonstratedocclusive cerebrovascular disease (�70% stenosis ofright carotid artery) with a history of a right-ICA ter-ritory transient ischemic attack (TIA) 3 months pre-viously. He was scheduled for a right CEA under gen-eral anesthesia. His past medical history also includeda myocardial infarction 5 years prior, along with mildperipheral vascular disease.

The primary anesthetic goals were maintenanceof the patient’s normal blood pressure, preventionof tachycardia and subsequent myocardial ischemia,analgesia, and prevention of coughing on emergence.A right radial arterial catheter was inserted preopera-tively. Premedication involved a small bolus of benzo-diazepine and the patient was carefully induced withsmall boluses of propofol and a concurrent infusionof remifentanil; neuromuscular blockadewas achievedwith vecuronium. Care was taken to blunt the sym-pathetic response to laryngoscopy and endotrachealintubation was achieved. The patient’s anesthesia wasmaintained with isoflurane, oxygen and air, along withremifentanil. The procedure proceeded uneventfullywith good control of blood pressure (within 20% ofbaseline) and smooth emergence was achieved withlow-dose remifentanil. Prior to removal of the endo-tracheal tube, gross neurologic function was examinedalong with observation of the patient’s pupils.

On arrival at the recovery ward, the patient – whileable to demonstrate gross neurologic movement – waspersistently disorientated in time, place, and person.Blood gas analysis including blood sugar and elec-trolyte levels were normal with adequate gas exchange.The patient’s blood pressure was maintained at slightlyelevated than normal throughout.

Following careful observation and after ruling outphysiologic and pharmacologic causes, a computedtomography scan of the patient’s head was performed,which proved unremarkable. He was admitted to theintensive care unit for further management and obser-vation. The patient’s cognition gradually improvedover the following 3–4 days and he was discharged tothe floor and eventually home. Follow-up at 6 weeksrevealed normal gross neurologic function but thepatient still complained of slight but persistent deficitsin concentration.

DiscussionThe recognized indications for CEA are numerous, butusually involve the presence of 70% or greater stenosisand/or the presence of ipsilateral TIAor stroke.Duringthe dissection of the narrowed vessel, the carotid arteryis clamped to facilitate dissection and reduce bloodloss.This clamping significantly reduces the ipsilateralhemisphere’s blood flow and has the potential to causefurther cerebral hypoperfusion. Upon clamp releasethere is a demonstrable hyperperfusion on the affectedside. The proposed mechanisms for neurologic dam-age are the hypo/hyperperfusion states, as well ascerebral ischemia combined with microembolic phe-nomena in patients who are likely to already have acompromised cerebral blood flow.

Carotid endarterectomy may be associated withboth gross and subtle neurologic damage. While thiscase demonstrated no gross neurologic deficit andnormal postoperative neuroimaging findings, therewas a clear decline in postoperative neurocognition.

91


However, a comprehensive preoperative neurocogni-tive assessement was not performed.

Predicting neurologic declinePostoperative cognitive dysfunction is a well recog-nized phenomenon following many surgical proce-dures performed under general anesthesia. Its pres-ence is associated with preexisting cognitive deficitalong with cardiopulmonary and pharmacologic fac-tors. However, there is evidence that CEA itself is arisk factor for intraoperative neurologic damage.Thereis demonstrable neurocognitive decline in approxi-mately 25% of CEA patients following postoperativeday one neuropsychometric tests as compared with apaired cohort undergoing a similar anesthetic regimenincluding the use of benzodiazepines. Unfortunately,there has also been some debate and little consensusas to which psychometric tests are most appropriate.A recent review article [1] highlighted the many con-founding variables in postoperative cognitive testing,which may explain why this area of research continuesto be contentious.

Examining the risk factors for intraoperative CEAneurologic decline, one study found increased ageto predict neurocognitive dysfunction. Of note, theauthors found that other risk factors (including thoseassociated with cerebrovascular disease such as smok-ing, diabetes, hypertension) and surgical parameters(including previous contralateral CEA, operative side,duration of surgery, cross-clamp time, and the need forshunt placement) did not appear to correlatewith post-operative day 1 cognitive deterioration [2].

Preventing neurologic declineA number of interventions have been examined todetermine their efficacy in the early demonstrationand subsequent prevention of neurologic damage.

Blood pressure: Blood pressure control has longbeen recognized as crucial during and after CEAto prevent critical hypo- or hyperperfusion states. Itis for this reason that intra-arterial blood pressureis routinely measured and controlled perioperatively,including the careful administration of balanced anes-thesia, analgesia, and vasoactive drugs.

Surgery under local anesthesia: The performanceof CEA under local anesthesia has the potential todemonstrate immediate neurologic decline. However,the gross neurologic tests performed on awake patientsduring a CEA may be unable to detect anything more

than a large neurocognitive deficit. Recent studies havefound there to be no significant difference in risk ofstroke or death, hospital stay, or postoperative qualityof life between local or general anesthetic techniques.However, further subgroup analysis demonstrated animproved neurologic outcome in surgery performedunder local anesthesia [3].

Shunt placement: Onemethod to reduce ipsilateralhypoperfusion is the use of a shunt, inserted proxi-mal to the clamp, to allow for improved blood flow. Asabove [2], this bore no correlation to postoperative day1 cognitive decline. Shunt placement is not universal inCEA surgery. Indeed its use may even be predictive ofpoor neurologic outcome as shunt usage is more likelyin patients with more pronounced cerebrovasculardisease.

Electrophysiologic analysis: The efficacy of a num-ber of electrophysiologic parameters in the predictionof neurologic deficit has been assessed. These haveincluded transcranial Doppler sonography (TCD),near-infrared spectroscopy (NIRS), stump pressure(SP) measurement, and somatosensory evoked poten-tials (SSEP).

Transcranial Doppler sonography analysis of arte-rial blood flow (usually the ipsilateral middle cere-bral artery) has demonstrated that high numbers ofmicroembolic signals, along with indicators of inap-propriate intraoperative hypo/hyperperfusion corre-late with postoperative ischemic events. Upon find-ings indicative of intraoperative cerebral ischemia,the administration of agents including platelet gly-coprotein IIB/IIIA receptor antagonists or Dextran-40 in the immediate postoperative phase has shownan improved mortality. However, the efficacy of thismodality in the prevention of embolic events has beenquestioned along with its ability to predict postop-erative cognitive dysfunction. Its routine use in CEAremains contentious.

Near-infrared spectroscopy cerebral oximetry hasbeen shown to be efficacious in those patients inwhom TCD analysis is technically more difficult toperform in the perioperative course, due to the lackof a temporal bone window. Its values have alsobeen shown to correlate to EEG changes indicative ofischemia.

One study found these parameters (TCD, NIRS,SP) to be equally efficacious in determining ischemia,but with SSEP being less accurate [4]. Analysis ofelectrophysiologic parameters has shown electroen-cephalography to be comparable with SSEP analysis

92

Case 28. Neurologic decline after carotid surgery

in determining cerebral ischemia although its use isassociated with a high false positive rate. Investigationinto the use of Bispectral index monitoring has notbeen shown to be efficacious in determining neuro-logic decline in awake patients undergoing CEA.

Enzyme markers of neurologic damage: The pres-ence and quantity of intracellular neuronal enzymeshas been predicted to correspond to the degree of neu-ronal damage. For example, elevated levels of S100betaprotein (S100B) have been shown to be a prognosticindicator of stroke size and outcome when measured2–3 days post-event whereas immediate rises maynot be associated with neurologic decline. In patientsundergoing cardiopulmonary bypass, elevated S100Bcorrelated with significant impaired cognition onlylater than 5 hours postoperatively.

Elevated serum S100B in CEA, but not neuron-specific enolase, has been associated with a subclinicalneurologic decline. However, this has been questionedby further investigators [5]. Furthermore, the use ofquantitative enzyme analysis via a jugular-venous bulbfailed to link enzyme rise with neurocognitive decline.There is much debate as to the impact of the time ofmeasurement of these enzymes. For example, S100Bhas a half life of only 2 hours.

ConclusionIn conclusion, the perioperative management ofpatients undergoing CEA, along with attempts to pre-

vent its associated neurologic decline, continues tobe a challenge to clinicians. Further investigation iswarranted to ensure the best neurologic outcomes inpatients undergoing this common procedure.

References1. P. De Rango, V. Caso,D. Leys et al.The role of carotid

artery stenting and carotid endarterectomy incognitive performance: a systematic review. Stroke2008; 39: 3116–27.

2. J. Mocco,D. A. Wilson, R. J. Komotar et al.Predictors of neurocognitive decline after carotidendarterectomy. Neurosurgery 2006; 58: 844–50.

3. C. F. Weber,H. Friedl,M. Hueppe et al. Impact ofgeneral versus local anesthesia on early postoperativecognitive dysfunction following carotidendarterectomy: GALA Study Subgroup Analysis.World J Surg 2009; 33: 1526–32.

4. S. Moritz, P. Kasprzak,M. Arlt et al. Accuracy ofcerebral monitoring in detecting cerebral ischemiaduring carotid endarterectomy: a comparison oftranscranial Doppler sonography, near-infraredspectroscopy, stump pressure, and somatosensoryevoked potentials. Anesthesiology 2007; 107:563–9.

5. L. S. Rasmussen,M. Christiansen, J. Johnsen et al.Subtle brain damage cannot be detected by measuringneuron-specific enolase and S-100beta protein aftercarotid endarterectomy. J Cardiothorac Vasc Anesth 14:166–70.

93


29 Postoperative hematoma and airwaycompromise after carotid endarterectomyMaged Argalious

In the North American Symptomatic Carotid Endar-terectomy Trial (NASCET), the incidence of postoper-ative wound hematomawas 5.5% [1].While themajor-ity of cases is the result of venous oozing and onlyrequire temporary external compression while rever-sal of anticoagulation takes effect, prompt recognitionand management of an expanding neck hematomareduces the risk of progression to airway compromise,a potentially fatal complication.

Case descriptionA 76-year-old female arrived in the postanesthesiacare unit (PACU) after an uneventful left carotidendarterectomy (CEA) for a severe left internal carotidartery stenosis (99%). Her medical history includedstable coronary artery disease, chronic atrial fibrilla-tion, hypertension and chronic obstructive pulmonarydisease (COPD) with an 80 pack-year smoking his-tory. Her surgical history included a remote cholecys-tectomy and cervical spine fusion. Her medicationsincluded aspirin, atorvastatin, amlodipine, metopro-lol, hydrochlorothiazide as well as supplemental potas-sium. She had been taking coumadin (for atrial fibril-lation), which was switched to an intravenous infusionof heparin on admission to the hospital. Her lab workshowed a prothrombin time/InternationalNormalizedRatio (PT/INR) of 13.2 seconds and 1.2 respectively.Her partial thromboplastin time (PTT) was main-tained at 40–60 seconds. Apart from a serum potas-sium of 3.3 mEq/L, her laboratory values were withinnormal limits.

On emergence from anesthesia, she developedsevere hypertension (190/115) requiring intravenousnitroglycerin boluses as well as intravenous labetalol.Her blood pressure on arrival to the PACUwas 150/80mmHg. Thirty minutes after, the PACU nurse noticedsome swelling at the incision site. Immediately there-after, the patient started to complain of shortness ofbreath despite increased supplemental oxygen. The

staff surgeon and anesthesiologist were called to thebedside.

DiscussionAirway management is often challenging in the PACUor critical care unit. Factors including surgery nearthe airway, intraoperative airway instrumentation ormanipulation, previous neck dissection or radiation[2], large volumes of intraoperative fluids, and resid-ual anesthetic effects contribute to these difficulties[3]. Postoperative bleeding after CEA is particularlyhazardous because bleeding into a closed space canquickly result in an expanding neck hematoma thatcan cause impingement on laryngeal structures andairway compromise. Even patients considered as hav-ing an ‘‘easy airway’’ in the operating room can poseairway challenges in the PACU or critical care unit [4].In patients recovering from neck surgery who developrespiratory insufficiency, the possibility of an expand-ing neck hematoma must be considered. Other pos-sible causes of dyspnea in this patient are outlined inTable 29.1.

If the neck hematoma is visible but is not causingrespiratory distress, pressure should be applied to thesurgical site to avoid further expansion. In addition tonotification of the surgeon, control of postoperativehypertension by infusions of short-acting medications(nitroglycerin, esmolol) is important in reducing thebleeding at the site of the fresh vascular anastomo-sis. In addition, confirmation of reversal of anticoag-ulation is important. If these measures are successfuland the neck hematoma is small in size, close obser-vation of the patient in a critical care environment aswell as marking of the boundaries of the hematomato ensure early identification of further expansion isimportant.

In cases of progressive expansion of the neckhematoma, even in the absence of airway compro-mise, awake intubation (possibly fiberoptic-guided

94

Case 29. Postoperative hematoma and airway compromise

Table 29.1. Common causes of dyspnea after carotidendarterectomy (CEA) not related to expanding hematomas.

1. CardiacMyocardial ischemiaAcute postoperative hypertension with resultantpulmonary edema

Atrial arrhythmia with rapid ventricular response

2. PulmonaryRight to left shunting most commonly due to atelectasisExacerbation of COPD

3. NerveRecurrent laryngeal nerve injury (especially in the setting ofprior contralateral CEA or neck surgery with bilateralrecurrent laryngeal paralysis)

Phrenic nerve paresis after cervical plexus block (especiallyin patients with severe COPD)

Carotid body denervation (more common after bilateralCEA)

COPD, chronic obstructive pulmonary disease.

according to the American Society of Anesthesiolo-gists difficult airway algorithm) [5] may be prudent,followed by surgical exploration of the wound anddrainage of the hematoma. In some cases, airwayedema persists despite drainage of the hematoma; itis advisable to maintain endotracheal intubation andpostoperative sedation after confirming the absence ofneurologic deficits.

In most instances of progressively expanding neckhematomas, airway obstruction ensues quickly asa result of encroachment and distortion of airwayanatomy. If emergent intubation attempts are unsuc-cessful, the decision to proceed with a surgical air-way (emergency cricothyroidotomy or tracheostomy)depends on the ability (vs inability) to ventilate thepatient with a face mask or laryngeal mask airway.If ventilation is unsuccessful or becomes inadequatedespite drainage of the neck hematoma, invasive air-way access should proceed. Table 29.2 summarizesthe management steps in cases of expanding neckhematoma.

References1. Beneficial effect of carotid endarterectomy in

symptomatic patients with high grade carotidstenosis. North American Symptomatic CarotidEndarterectomy Trial Collaborators. N Engl J Med1991; 325: 445–53.

2. C. M. Burkle,M. T. Walsh, S. G. Pryor et al. Severepostextubation laryngeal obstruction: The role of prior

Table 29.2. Management of postoperative neck hematoma.

Apply pressure to bleeding site

Notify surgery and anesthesia team (call for help)

Consider reversal of residual anticoagulation

Tight blood pressure control

Outcome:No further hematoma expansionCommunication with surgical teamMark the boundaries of the hematoma for early identification

of further expansionClose observation and extended monitoring (8–12 hours) in

a critical care environment

Continuous expansion of hematomawith no airwaycompromiseAwake (fiberoptic) intubation either in the PACU or after

immediate transfer to the operating room aftertopical anesthesia to the airway followed by generalanesthesia for exploration of wound and drainage of neckhematoma

Assess neurologic status at the end of the caseConsider maintaining endotracheal intubation

postoperatively until resolution of reactionary airwayedema

Expansion of neck hematomawith rapidly progressiveairway compromise (dyspnea, stridor, airwayobstruction)Emergent intubation (ASA algorithm)Cannot intubate/can ventilate: using face mask, oral or nasal

airways, laryngeal mask airway:Consider immediate surgical drainage of the neck

hematoma followed by further attempts to secure theairway

Cannot intubate/cannot ventilate:Surgical airway (emergent cricothyroidotomy, percutaneous

or surgical tracheostomy)Evacuation of hematoma and wound exploration,

neurologic assessmentMaintain airway secured postoperatively

PACU, postanesthesia care unit.

neck dissection and radiation. Anesth Analg 2006; 102:322–5.

3. G. S. Murphy, J. W. Szokol, J. H. Marymont et al.Residual neuromuscular blockade and criticalrespiratory events in the postanesthesia care unit.Anesth Analg 2008; 107: 130–7.

4. M. Argalious. PACU Emergencies ASA RefresherCourses in Anesthesiology 2009. 37; 1–12.

5. American Society of Anesthesiologists Task Force onManagement of the Difficult Airway. Practiceguidelines for management of the difficult airway:An updated report by the American Society ofAnesthesiologists Task Force on Management of theDifficult Airway. Anesthesiology 2003; 98: 1269–77.

95


30 Postoperative stroke after carotidendarterectomyJ. Javier Provencio

Perioperative stroke can be a devastating complica-tion of surgery and has a particularly high incidencein associationwith vascular procedures such as carotidendarterectomy (CEA).

Case descriptionThe patient was a 75-year-old right-handed male witha history of a previous stroke 1 month before admis-sion, which manifested as left upper extremity weak-ness that steadily improved over about a week. His ini-tial evaluation at the hospital revealed a 60–79% steno-sis of the right internal carotid artery. He returned tothe hospital for an elective CEA. He was a �100 packyear smoker of cigarettes but had recently “cut down”to 10 cigarettes per day since his original stroke. Inaddition, he had a history of hypertension, hypothy-roidism, and dyslipidemia. His medication regimenconsisted of aspirin, clopidogrel, levothyroxine, anHMG-CoA reductase inhibitor, and three antihyper-tensivemedications (a beta-blocker, a calcium channelblocker, and a peripherally acting alpha-blocker). Inaddition, he was taking hydrocodone-acetaminophenfor low back pain. Interestingly, he was allergic tolobster, which had prompted physicians to adminis-ter diphenhydramine prior to dye-containing radiol-ogy tests.

On physical exam, the patient was a well-appearingman with no obvious signs of stroke. His blood pres-sure was high at 186/90mmHg and he was tachycardicat 92 beats per minute. He had bilateral carotid bruitson auscultation and mild weakness of his left lowerextremity but was able to hold the leg up against grav-ity for 10 seconds. He had no facial droop or weaknessin the arm on the left side. He was noted to have visualfield deficit in the left visual field.The rest of his exam-ination was unremarkable.

Upon emergence from anesthesia, the patient wasnoted to be hemiplegic on the left side with a promi-nent facial droop. His National Institutes of Health

Stroke Scale score was noted as 17 putting him in ahigh-risk group for poor outcome. Computed tomog-raphy (CT) scan showed no evidence of intracerebralhemorrhage but a CT angiogram suggested a deficitof perfusion in the right hemisphere suggesting anoccluded artery. It was felt by the surgeon and thestroke physician that intravenous tissue plasminogenactivator (tPa) was too risky due to the recent surgery.The patient was taken for arterial thrombolysis witha mechanical clot retrieval device. Postprocedure, hewas awake with slightly improved arm and leg strengthon the left side. He continued to improve through hishospital stay but ultimately had multiple strokes in theright hemisphere (Figure 30.1).

DiscussionThe risk of stroke in patients with symptomatic orasymptomatic carotid stenosis is quite high. A num-ber of studies have shown that the risk of subse-quent stroke is decreased in correctly selected patientswho have CEA. Despite this, stroke is a commonperioperative complication, ranging from 0 to 8.4%[1–3]. Treatment of patients postoperatively after CEAwho experience stroke can be a challenging situationfor the clinician. On the one hand, they are in thehospital and usually come to the attention of medi-cal staff very early (which is the most important goalof stroke therapy). On the other hand, the fact thatthey have had a recent surgery limits the interven-tions that can be offered to a patient with an acutestroke.

There are three causes of acute lateralized deficitsafter CEA: one is a physiologic abnormality and twoare ischemic consequences.Hyperperfusion syndromeis loss of autoregulation of the intracerebral carotidsystem on the side of the surgery, which leads to lossof control of blood flow. When the arterial blood flowoverwhelms venous return, blood pools in the capil-lary beds resulting in ischemia and hemorrhage. The

96

Case 30. Postoperative stroke after carotid endarterectomy

Figure 30.1. Diffusion weightedmagnetic resonance images. The leftimage shows an area of stroke noticed afteremergence from anesthesia. The area isrelatively small considering the symptomsincluded complete hemiparesis. The rightimage shows the massive stroke thepatient had after recovery from his originalepisode.

treatment is unclear but systemic hypotension to con-trol the blood flow is currently the accepted standard[4].

The two common causes of ischemic stroke afterCEA are acute occlusion of the carotid artery at thesurgical site and embolization of clot to the distal cere-bral vasculature [5]. The standard treatment of acuteocclusion of the carotid artery includes emergent sur-gical revascularization of the carotid followed by intra-venous anticoagulant administration to prevent clotreformation at the surgical site [6]. In patients with sig-nificant areas of infarcted brain, there is a risk of reper-fusion cerebral edema, venous congestion and hem-orrhage. In addition, anticoagulation in patients withlarge stroke volumes may increase the risk of hemor-rhagic conversion of a bland stroke. Obviously, surgeryis not indicated for patients who have distal emboli.

There are very few prospective data on acuteintervention for stroke in patients who have distalembolism after CEA. There are a few reports at scien-tific meetings of systemic administration of recombi-nant tPa for patients after carotid endarterectomy withmixed results. It is unclear from these reports whetherthis is an effective or safe intervention.

More recently, intravascular interventional tech-niques to revascularize occluded intracerebral arter-ies has become a tempting option to open acute dis-tal embolic strokes. Interventional procedures includea number of devices and drugs (including anticoagu-lants, antiplatelets, and lytic agents) used individuallyor in combination [7]. The use of more devices anddrugs in combination increases the risk of bleeding butalso increases the probability of successful recanaliza-tion. There are no prospective trials to support using

this paradigm but it is unlikely that there will ever bea trial with sufficient power to answer the questionof efficacy. Anecdotal evidence suggests some patientsbenefit from this approach. There is no good evidenceto inform the risk of bleeding.

In addition to strategies to revascularize theblood vessel, there are a number of interventionsaimed at preventing subsequent stroke. They includeaspirin, clopidogrel, the combination of extended-release dipyridimole and aspirin. In addition, there isevidence thatHMG-CoA reductase inhibitors (statins)improve outcome after stroke. Fortunately, mostpatients with carotid endarterectomy take aspirin andstatins as part of the treatment of their carotid stenosis.

ConclusionFinally, there is an argument to be made for urgentneurologic consultation after stroke from CEA. Neu-rologists can quickly stratify patients for eligibility foracute intervention. In addition, improvement in strokeoutcome has been shown to correlate with early andaggressive physical therapy and rehabilitation. This isan important and potentially overlooked aspect of carein patients with stroke after any surgery. A neuro-logic consultationmaymake aggressive evaluation andrehabilitation more likely.

References1. M. R. Mayberg, S. E. Wilson, F. Yatsu et al. Carotid

endarterectomy and prevention of cerebral ischemiain symptomatic carotid stenosis. Veterans AffairsCooperative Studies Program 309 Trialist Group. J AmMed Assoc 1991; 266: 3289–94.

97


2. L. G. Ludington, G. I. Kafrouni,M. H. Peterson et al.Clinical review of 106 consecutive carotidendarterectomies. Int Surg 1976; 61: 155–9.

3. Endarterectomy for asymptomatic carotid arterystenosis. Executive Committee for the AsymptomaticCarotid Atherosclerosis Study. J AmMed Assoc 1995;273: 1421–8.

4. T. Karapanayiotides, R. Meuli, G. Devuyst et al.Postcarotid endarterectomy hyperperfusion orreperfusion syndrome. Stroke 2005; 36: 21–6.

5. J. B Towne, V. M. Bernhard. Neurologic deficitfollowing carotid endarterectomy. Surg Gynecol Obstet1982; 154: 849–52.

6. J. P. Berthet, C. H. Marty-Ane, E. Picard et al. Acutecarotid artery thrombosis: description of 12 surgicallytreated cases. Ann Vasc Surg 2005; 19: 11–18.

7. S. H. Kim, A. I. Qureshi, E. I. Levy et al. Emergencystent placement for symptomatic acute carotid arteryocclusion after endarterectomy. Case report.J Neurosurg 2004; 101: 151–3.

98


31 Postoperativemyocardial infarctionMaged Argalious

Patients with atherosclerotic carotid disease have ahigh incidence of concomitant coronary artery disease[1]. In theGALA study, amulticenter randomized con-trolled trial to evaluate general anesthesia (GA) ver-sus local anesthesia (LA) for carotid surgery, the pri-mary outcome of stroke, myocardial infarction, anddeath occurred in 4.7% of cases [2]. Since the risk ofmajor adverse cardiac effects after carotid endarterec-tomy (CEA) is between 1 and 5%, CEA is consideredan intermediate risk surgery [3].

Case descriptionA 59-year-old male presented for a right CEA underregional anesthesia (superficial cervical plexus block).He had a history of hypertension, hyperlipidemia,and peripheral vascular disease. He was a currentsmoker and refused to quit smoking. He had a tran-sient ischemic attack 5 weeks earlier and underwenta preoperative cardiac evaluation prior to his carotidsurgery, including a dobutamine stress echocardio-graphy for evaluation of orthopnea. This showed areduced left ventricular function of 40% and regionalwall motion abnormalities in the anterolateral wall ofthe left ventricle (LV). A cardiac catheterization wasfollowed by percutaneous coronary intervention withtwo bare metal stents to the left anterior descend-ing and obtuse marginal vessels. The patient wasinstructed to hold his clopidogrel 5 days prior to hisscheduled carotid surgery and to continue his aspirinperioperatively [4]. The rest of the patient’s medi-cations included lisinopril, carvedilol, and rosuvas-tatin. After a stable intraoperative course, the patientwas transferred to the recovery room in a stablecondition.

One hour later, however, he developed crushingsubsternal chest pain followed by acute hemodynamicdecompensation (hypotension, bradycardia). His elec-trocardiogram showed marked S-T segment elevationin the anterior chest leads.

Table 31.1. Medical management of patients withS-T segment elevation myocardial infarction (STEMI).

Oxygen

Aspirin

Statin drugs

Nitrates

Beta blockers

Narcotic analgesics (morphine)

ACE inhibitors/angiotensin receptor blockers

Management of hyperglycemia (insulin)

DiscussionPatientswith S-T segment elevationmyocardial infarc-tion (STEMI) of sufficient size have a reduction inLV function resulting in a reduced stroke volume,reduced systemic blood pressure, and a consequentreduction in coronary perfusion pressure. The reduc-tion in stroke volume results in an increase in LVfillingpressures, causing LV dilation and increasing LV enddiastolic pressures. This leads to worsening coronaryperfusion pressure, ultimately resulting in cardiogenicshock unless this vicious circle is reversed by reper-fusion (fibrinolytic therapy or invasive percutaneousintervention) [5]. In STEMI of smaller size, the unaf-fected portion of the ventricle becomes hyperkinetic inorder to sustain overall ventricular function.

Surgical patients with postoperative STEMI arepoor candidates for fibrinolytic therapy [6]. Initialmedical management options (Table 31.1) are usuallylimited in patients with STEMI in the setting of car-diogenic shock, because many of the recommendedmedications in STEMI are contraindicated for fear ofworsening the hypoperfusion state (e.g., beta block-ers, nitrates, ace inhibitors). In this patient, AdvancedCardiac Life Support guidelines (including the use ofinotropes) should be followed to restore hemodynamicstability. In STEMI secondary to stent thrombosis asin this case, the definitive treatment is percutaneous

99


Figure 31.1. Proposed approach for patients with a previous percutaneous coronary intervention (PCI) who need subsequent surgery (leftpanel of the figure) and for patients who require nonsurgical revascularization before noncardiac surgery. ACS, acute coronary syndrome; AMI,acute myocardial infarction. Reprinted with permission from S. G. DeHert. Preoperative cardiovascular assessment in noncardiac surgery: anupdate. European Journal of Anaesthesiology 2009; 26(6): 449–57.

coronary intervention to re-establish blood flow to thethrombosed stent [6]. Surgical procedures in patientswith prior percutaneous coronary intervention (PCI)should therefore be performed in institutions where24-hour interventional cardiology is available.

In patients undergoing preoperative cardiac eval-uation prior to noncardiac surgery, the decision toproceed with PCI has to be discussed among themanagement team (surgeon, anesthesiologist, and car-diologist) [5]. Patients with a recent transient ischemicattack have a higher incidence of subsequent strokewithin the first year and CEA should not be consid-ered totally elective. Therefore, if noncardiac surgerycannot be postponed until the minimum period ofdual antiplatelet therapy after PCI is achieved (Fig-ure 31.1), alternative management options should beconsidered including performing the CEA on dualantiplatelet therapy, bridging therapy with GIIb/IIIareceptor antagonists (e.g., tirofiban or eptifibatide),maximizing medical management before CEA fol-lowed by postoperative management of coronaryartery disease, or pursuing a combined approach (CEA

and coronary artery bypass grafting) [7]. Prematurediscontinuation of antiplatelet therapy has been showntomarkedly increase the incidence of stent thrombosisand should not be done for elective surgery [8].

Regardless of the revascularization technique,aspirin should be a lifelong therapy since its abrupt dis-continuation results in a rebound increase in inflam-matory prothrombotic state, adding to the prothrom-botic state caused by surgical intervention [6].

References1. N. R. Hertzer, J. R. Young, E. G. Beven et al.

Coronary angiography in 506 patients withextracranial cerebrovascular disease. Arch InternMed 1985; 145: 849–52.

2. GALA Trial Collaborative Group, S. C. Lewis,C. P. Warlow et al. General anesthesia versus localanesthesia for carotid surgery (GALA): a multicentre,randomized controlled trial. Lancet 2008; 372: 2132–42.

3. L. A. Fleisher, J. A. Beckman, K. A. Brown et al. 2009ACCF/AHA Focused Update on Perioperative Beta

100

Case 31. Postoperative myocardial infarction

Blockade Incorporated Into the ACC/AHA 2007Guidelines on Perioperative Cardiovascular Evaluationand Care for Noncardiac Surgery. A Report of theAmerican College of Cardiology Foundation/American Heart Association Task Force on PracticeGuidelines. Circulation 2009; 120: e169–276.

4. C. L. Grines, R. O. Bonow,D. E. Casey, Jr et al.AHA/ACC/SCAI/ACS/ADA Science Advisory.Prevention of premature discontinuation of dualantiplatelet therapy in patients with coronary arterystents. A science advisory from the American HeartAssociation, American College of Cardiology, Societyfor Cardiovascular Angiography and Interventions,American College of Surgeons, and American DentalAssociation, with representation from the AmericanCollege of Physicians. Circulation 2007; 115: 813–18.

5. F. G. Kushner,M. Hand, S. C. Smith, Jr et al. FocusedUpdates: ACC/AHA Guidelines for the Managementof Patients with ST-Elevation Myocardial Infarction

and Guidelines on Percutaneous CoronaryIntervention. A Report of the American College ofCardiology Foundation /American Heart AssociationTask Force on Practice Guidelines. Circulation 2009;120; 2271–306.

6. L. T. Newsome, R. S. Weller, J. C. Gerancher et al.Coronary artery stents. II. Perioperative considerationsand management. Anesth Analg 2008; 107: 570–90.

7. E. S. Brilakis, S. Banerjee, P. B. Berger. The risk ofdrug-eluting stent thrombosis with noncardiacsurgery. Curr Cardiol Rep 2007; 9: 406–11.

8. American Society of Anesthesiologists Committeeon Standards and Practice Parameters. Practice alertfor the perioperative management of patients withcoronary artery stents. A Report by the AmericanSociety of Anesthesiologists Committee on Standardsand Practice Parameters. Anesthesiology 2009; 110:22–3.

101

Part III Functional neurosurgeryCase

32 Preoperative evaluation for deep brainstimulator surgeryMilind Deogaonkar

Deep brain stimulation (DBS) is now a routine ther-apeutic option for patients with Parkinson’s disease,essential tremors (ET), and dystonia. The success ofDBS surgery depends on proper patient selection,proper placement of the DBS electrode in the intendednucleus and proper programming. In view of this, thepreoperative evaluation of the DBS patient assumesadded importance.

Case descriptionA 65-year-old male was seen in our outpatient clinicto be evaluated for DBS surgery. He had symptomsof Parkinson’s disease for 15 years, primarily motorfluctuation, tremors, and dyskinesia. The patienthad an improvement of 60% on his motor unifiedParkinson’s disease rating scale when compared inthe off and on state. The symptoms improved signif-icantly with levodopa though the improvement wasshort-lasting. In addition, he had long-standing butstable coronary artery disease for which he was takingaspirin. He also had cervical spine fusion in the pastand was asymptomatic except for a reduced range ofmotion. He had significant anxiety on his neuropsy-chological testing but no significant cognitive decline.The patient was deemed to be a good candidate forstaged bilateral subthalamic nucleus DBS surgery. Hewas told to stop his aspirin 10 days in advance. Ananxiety counselor was arranged in the operating room.Preoperative cervical magnetic resonance imagingwas done to rule out any residual cervical stenosis orcompression. The patient underwent a unilateral DBSfollowed by the second side DBS 3 weeks later. Hethen underwent bilateral implantable pulse generators2 weeks after the second side. An awake intubationwas performed and intraoperative electrophysiologicmonitoring was employed during the surgery. Thesurgeries were uneventful and he had good therapeuticbenefits.

DiscussionThis case highlights the following important preoper-ative evaluation issues in DBS patients:1. General surgical patient evaluation.2. Disease-specific evaluation.3. Neuropsychological evaluation.4. Evaluation of associated medical conditions.5. Airway evaluation.

General surgical patient evaluationIn general, patients must be able to tolerate the vari-ous components of surgery and have the social sup-port structure to comply with the demands of surgeryand postoperative care. For those undergoing DBSsurgery, both the patient and family members needto have a detailed understanding of reasonable out-comes, potential complications, and the multiple stepsinvolved in the preoperative assessments, surgery,perioperative management, and follow-up care. Thepatient needs to be cooperative with follow-up pro-gramming and adjustment of medications in the out-patient setting. Additionally, the patient and fam-ily need to have realistic expectations about surgicaloutcome.

Disease-specific evaluationParkinson’s diseaseIn general, surgery is most likely to benefit symptomsaffecting the extremities versus axial symptoms suchas posture, balance, gait, and speech. Surgical candi-dates typically have severe tremors; “Off” medication-related rigidity, freezing, dystonia, and bradykinesia;“On” medication-related dyskinesias and significantlydisabling “On–Off” medication motor fluctuations.One of the most important predictors of neurosur-gical treatment response is the patient’s responseto levodopa. Patients who demonstrate a significantimprovement in motor symptoms during “Off” versus

103

III. Functional neurosurgery

“On” levodopamedication state aremost likely to ben-efit from surgery.

TremorIn general, patients with distal tremor, either postu-ral, intention/action, or resting tremor can be con-trolled with surgery. In contrast, the more proxi-mal tremors are the most difficult to treat surgically[1–3]. Head, neck, and lower extremity tremors arealso more difficult to treat than upper extremitytremors. Tremors involving the head/neck and axialregions usually require bilateral surgery.

DystoniaPatients who are refractory to all the conservativemea-sures, includingmedication trials and botulinum toxininjections, are potential candidates. Primary general-ized dystonia [4–6], as well as patients with idiopathiccervical dystonia, can obtain the best motor benefitswith bilateral internal globus pallidus DBS. Juvenile-onset idiopathic dystonia with age of onset �5 yearsand without multiple orthopedic deformities also hasa good response to surgery.

Neuropsychological evaluationNeuropsychological assessment is recommended aspart of the preoperative assessment to determinecandidacy for DBS. The neuropsychological assess-ment should include assessment of cognition, neu-ropsychiatric symptoms, social support, and goals forsurgery. Patients with severe cognitive dysfunction ordementia on neuropsychological examination shouldbe excluded from surgical intervention. Patients withmild cognitive impairment or a frontal dysexecutivesyndromemay still undergo surgery but should receiveextra counseling along with their family regardingthe potential for increased risk of cognitive impair-ment and confusion post-surgery. Psychiatric condi-tions such as anxiety, depression, and mania must beidentified andmedically optimized preoperatively andan intraoperative anxiety counselor can be arranged.Neurosurgical intervention in patients with a delu-sional psychosis or severe personality disorder, suchas borderline personality disorder, is generally notrecommended.

Evaluation of associated medical conditionsPatients should be in stable overall health with respectto cardiac, pulmonary, and systemic conditions such

Figure 32.1. Patient in a Leksell frame on the operating table,illustrating the difficulty of accessing the airway with a frame inplace.

as hypertension, diabetes, and cancer. Patients whorequire antiplatelet medications or coumadin mustbe able to tolerate complete withdrawal from thesemedications prior to surgery. Consultation withother medical specialists (e.g., cardiologists) may berequired prior to proceeding with surgery in somepatients.

Airway evaluationAccess to the airway is difficult once the frame is inplace (Figure 32.1). A detailed evaluation of the air-way is therefore essential. An awake intubation can beused in patients with cervical spine instability or severecervical spine stenosis; intraoperative somatosensoryevoked potential monitoring is useful for monitoringthe integrity of the cervical spine cord during theprocedure.

ConclusionDeep brain stimulation surgery for Parkinson’s diseaserequires a systematic approach to preoperative assess-ment. Understanding the medical and neuropsycho-logical considerations of Parkinson’s disease is essen-tial for proper perioperative care.

104

Case 32. Preoperative evaluation for deep brain stimulation surgery

References1. C. Berk, J. Carr,M. Sinden et al.Thalamic deep brain

stimulation for the treatment of tremor due to multiplesclerosis: a prospective study of tremor and quality oflife. J Neurosurg 2002; 97: 815–20.

2. G. Deuschl, P. Bain. Deep brain stimulation fortremor [correction of trauma]: patient selection andevaluation.Mov Disord 2002; 17: S102–11.

3. M. Kitagawa, J. Murata, S. Kikuchi et al. Deep brainstimulation of subthalamic area for severe proximaltremor. Neurology 2000; 55: 114–16.

4. M. Krause,W. Fogel,M. Kloss et al. Pallidalstimulation for dystonia. Neurosurgery 2004; 55:1361–8.

5. A. Kupsch, S. Klaffke, A. A. Kuhn et al.The effectsof frequency in pallidal deep brain stimulationfor primary dystonia. J Neurol 2003; 250: 1201–5.

6. J. Y. Lee,M. Deogaonkar, A. Rezai. Deep brainstimulation of globus pallidus internus fordystonia. Parkinsonism Relat Disord 2007; 13:261–5.

105


33 Airway crisis during deep brainstimulator placementEhab Farag

Deep brain stimulation (DBS) is the ultimate ther-apy for motor disease disorders like Parkinson’s dis-ease (PD) and dystonia.The anestheticmanagement ofDBS is a challenging process. The patient needs to besedated and comfortable but awake enough to conductthe proper neurologic examination and neurophysi-ologic recording for deep brain electrode insertion.Deep brain stimulation procedures are not withoutcomplications; here we describe one of the most com-mon complications encountered during DBS.

Case descriptionThe patient was a 70-year-old male with a past med-ical history of hypertension, hyperlipidemia, obstruc-tive sleep apnea, and PD. The patient was scheduledfor unilateral DBS electrode insertion in the subthala-mic nucleus under sedation. After applying standardASA monitors, sedation was started using propofol.Propofol was titrated between 50 and 100mcg/kg/minto maintain required sedation. Systolic blood pres-sure was maintained ≤140 during the procedure withboluses of labetalol. Oxygen was administered vianasal canula at a rate of 3–4 L/min to maintain oxy-gen saturation ≥95%. The patient tolerated the inser-tion of the DBS electrode and identification of the sub-thalamic nucleus. During the closure phase of the pro-cedure the sedation was restarted. The patient becameagitated; boluses of propofol were given to deepen thelevel of sedation. The patient became apneic with arapid decrease of oxygen saturation to 40% and a con-sequent decrease of the heart rate to 30 beats/min.Attempts to ventilate the patient using a face maskfailed.The arrest codewas called to the operating roomdue to the inability to intubate the patient. While hishead was fixed in the DBS frame, a laryngeal maskairway (LMA) was inserted. Insertion of the LMAenabled successful ventilation and resuscitation. Theclosure of the procedure was accomplished under gen-eral anesthesia using sevoflurane via the LMA. At the

end of the procedure, the patient awoke neurologicallyintact.The patient was discharged the following day ina stable condition.

DiscussionDeep brain stimulation is considered the final therapyfor medically refractory PD as well as other chronicneurologic disorders such as dystonia, depression, andchronic pain syndromes. The DBS procedure requiresfixation of the patient’s head to the stereotactic appa-ratus for accurate electrode placement. This can leadto substantial patient discomfort. Patient selection andpreparation for the procedure are very important inorder to minimize the perioperative complications.Aspirin and coumadin should be stopped. Blood pres-sure should be well controlled during the procedure toavoid the development of hematomas (subdural, sub-arachnoid, intraventricular, and intracrebral). Over-all, the rate of hemorrhage has been estimated at 3–5% per patient [1]. The cessation of anti-Parkinsonianmedications before the surgery (called “Off” period)can be very unpleasant, particularly for those patientswith severe pain, dystonia, or depression. Benzodi-azepines and long-acting narcotics should be avoidedduring the procedure to facilitate the proper electrodeplacement.The use of propofol for sedation during theprocedure is not without complications. Propofol caninduce respiratory depression and airway obstruction,aswas seen in our case.Headfixation during the proce-dure usually renders endotracheal intubation very dif-ficult. Placement of an LMA is the ideal method formanaging the upper airway in emergency situationsduring DBS procedures.

We have noted that some patients receiving propo-fol experienced sneezing [1], which has already beendocumented by its manufacturer. Sneezing can bevery troublesome to the patient and interferes withphysiologic mapping during the DBS. In addition,sneezing may cause a sudden increase in intracranial

106

Case 33. Airway crisis during deep brain stimulator placement

pressure that may lead to intracranial hemorrhage.Wehave noted that age ≥64 is the single most importantindependent factor for complications during DBS [1].Properties of dexmedetomidinemake it well suited forsedation during DBS procedures. It has little effect onthe patient’s motor symptoms, respiratory functions,and upper airway. Dexmedetomidine creates an envi-ronment in which the patient feels comfortable andrelaxed during the procedure [2–4].

ConclusionDeep brain stimulation represents a significant anes-thetic challenge for the anesthesiologist. Vigilance and

awareness of the procedure’s complications are the keyelements for successful anesthetic management.

References1. R. Khatib, Z. Ehrahim, A. Rezai et al. Perioperative

events during DBS : the experience at ClevelandClinic. J Neurosurg Anesthesiol 2008; 20: 36–40.

2. I. Rozet, S. Muangman,M. S. Vavilala et al. Clinicalexperience with dexmedetomidine for implantation ofdeep brain stimulators in Parkinson’s disease. AnesthAnalg 2006; 103: 1224–8.

3. E. Farag. Dexmedetomidine in the neurointensivecare. Discovery Medicine 2010; 9: 42–45.

4. C. Trombetta, A. Deogaonkar,M. Deogaonkar et al.Delayed awakening in dystonia patients undergoingdeep brain stimulation. J Clin Neurosci 2010; 17: 865–8.

107


34 Postoperative management ofParkinson’s medicationsMilind Deogaonkar

Deep brain stimulation (DBS) for Parkinson’s dis-ease (PD) has now become a routine therapeuticoption. The success of surgery depends on properpatient selection, proper placement of electrodes,and postoperative programming. The managementof anti-Parkinsonian medication in the postoperativeperiod is affected by interaction of other perioper-ative medications with PD and the effects of DBSsurgery.

Case descriptionA 62-year-old female with a 15-year history of PD wasscheduled to have bilateral subthalamic nucleus (STN)DBS implantation. Her main symptoms were bilateralresting tremors, bradykinesia, rigidity, motor fluctua-tions, and dyskinesias. She had a unified Parkinson’sdisease rating scale motor score of 52, which improvedto 26 after levodopa. She was on 1200 mg levodopa aday in divided doses and 4mg ropinorole every day. Ofnote, she had a history of gastroesophageal reflux dis-ease. Her anti-Parkinsonianmedications were stoppedas per the protocol the previous night. She underwentbilateral stereotactic STN DBS implants without anyproblems (Figure 34.1). Following the surgery she wasobserved to be very rigid and confused and did notopen her eyes but was following commands. Her braincomputed tomography scan was normal with someresidual pneumocephalus. After she was given her reg-ular medications the rigidity improved but she devel-oped severe dyskinesias. After reducing the dose of herlevodopa by half, the dyskinesias improved. Duringthe night she developed symptomsof reflux andnauseaand was given metoclopromide and some haloperidolfor her agitation. Following this, she went into acuteParkinsonian crisis with severe rigidity and dystonicspasm. Her symptoms gradually improved once thesemedications were stopped. She was discharged after astay of 3 days in the hospital.

DiscussionThis case highlights the following points:

1. Importance of restarting PD medications soonafter surgery.

2. Interactions between the STN DBS and levodopa.3. Effect of postoperative neurologic changes and

medication.4. Interactions of other medication and

Parkinsonism.

Importance of restarting PD medicationssoon after surgeryParkinson’s disease patients are off their routine PDmedications for at least 12 hours before surgery.The reason for this medication stoppage is to aid inmicroelectrode recording and macrostimulation dur-ing surgery to confirm the proper placement of theDBS lead [1]. After completion of DBS surgery, whichusually lasts anywhere between 4 to 8 hours, thepatients need to be given the medications as soon aspossible. Delay in restarting the medication can resultin worsening of PD symptoms and severe dyskinesiasand dystonias.

Interactions between the STN DBSand levodopaPlacement of DBS in the STN itself exerts an anti-Parkinsonian effect immediately after surgery. Thiscan act synergistically with the anti-PD medicationsand can produce severe peak-dose dyskinesias. It isimportant to watch the patient closely and titratethe preoperative anti-PD medications in patientswho have a significant micro-subthalamotomy effect[2, 3].

108

Case 34. Postoperative management of Parkinson’s medications

Figure 34.1. A 3-D representation of the microelectroderecording tracts and DBS lead implantation in STN.

Effect of postoperative neurologic changesand medicationThe most common neuropsychiatric side effect in theimmediate postoperative period following STN DBSis transient confusion with an incidence between 1%and 36% [4, 5]. Evidence of greater neuropsycholog-ical deficits prior to surgery is significantly associ-ated with increased confusion following surgery [4].Eyelid apraxia is also a commonly seen transient sideeffect of bilateral STN DBS [5]. This is transient anddoes not need any specific treatment but definitelyaffects the neurologic assessment in the postoperativeperiod.

Interactions of other medication andParkinsonismAvoid any medications that have an antidopaminer-gic effect like haloperidol and metoclopramide andthat can worsen the symptoms of PD patients or causeacute Parkinsonian crisis. The best antinausea medi-cation for PD patients in the postoperative period isondansetron; lorazepam in small dosesworks verywellfor confusion.

In summary, STN DBS has been demonstrated tobe effective in alleviating the symptoms of medicallyrefractory PD across multiple reports in the litera-ture [5]. These results were confirmed by prospec-tive series with double-blinded assessments and werelargely sustained at 5-year follow-ups. Proper post-operative management can result in outstandingoutcomes.

References1. A. R. Rezai, B. H. Kopell, R. E. Gross et al.Deep brain

stimulation for Parkinson’s disease: surgical issues.Mov Disord 2006; 21: S197–218.

2. P. Krack, V. Fraix, A. Mendes et al. Postoperativemanagement of subthalamic nucleus stimulation forParkinson’s disease.Mov Disord 2002; 17: S188–97.

3. J. G. Nutt, S. L. Rufener, J. H. Carter et al. Interactionsbetween deep brain stimulation and levodopa inParkinson’s disease. Neurology 2001; 57: 1835–42.

4. J. G. Pilitsis, A. R. Rezai, N. M. Boulis et al. Apreliminary study of transient confusional statesfollowing bilateral subthalamic stimulation forParkinson’s disease. Stereotact Funct Neurosurg 2005;83: 67–70.

5. A. R. Rezai, A. G. Machado,M. Deogaonkar et al.Surgery for movement disorders. Neurosurgery 2008;62: 809–38.

109


35 Epilepsy surgery: intraoperative seizureOren Sagher and Shawn L. Hervey-Jumper

Epilepsy affects 5–10/1000 people in North Americaand is the secondmost common cause ofmental healthdisability among young adults [1]. Despite advancesin anti-epileptic drugs, 20–40% of all patients withepilepsy are refractory to medical management. Forthese patients, epilepsy surgery is an underutilizedoption.

Case descriptionThe patient was a 20-year-old healthy male with ahistory of seizure disorder who was found to have aright frontal brain tumor. Although he had multipleevents over a 6-month period characterized by briefloss of awareness, he now presented with a general-ized seizure suffered while swimming. He was pulledout of the pool and after a brief post-ictal period,recovered to his neurologic baseline. He denied weak-ness, morning nausea or vomiting, speech changes,or headaches. Head computed tomography showed aposterior right frontal lesion, which on brain mag-netic resonance imaging was non-enhancing and asso-ciated with minimal vasogenic edema. The lesion waslocated within the precentral gyrus and functionalbrain imaging confirmed its proximity to the pri-mary motor cortex with the corticospinal tract fibersdirectly posterior to the mass lesion (Figure 35.1). Thepatient was admitted for a right frontal craniotomywith awake motor mapping and image-guided tumorresection.

The primary concerns of the anesthesiology teamwere: (1) management of intraoperative pain, par-ticularly during Mayfield cranial pin placement, (2)management of the airway, (3) the potential for nau-sea and vomiting, and (4) intraoperative seizures dur-ing cortical stimulation. The patient was brought tothe operating room and brief sedation was inducedusing a propofol and alfentanil infusion.Hewas placedin a Mayfield head-holder and his head coordinateswere registered using the surgical navigational com-

puter. A scalp block was performed using a combi-nation of 1% lidocaine and 0.25% bupivacaine with1:200 000 units epinephrine. A skin incision was madeand craniotomy flap turned. Using image-guidance,the location of the tumor was identified. The propo-fol infusion was stopped for cortical mapping usingelectrocorticography and cortical stimulation. Polar-ity changes by somatosensory evoked potentials werefound two gyri behind the tumor, suggesting thatthe tumor was anterior to the primary motor cortex.Electrical stimulation induced complex movements ofthe upper extremity at a low threshold, suggesting pre-motor cortex activation. Several small seizures wereinduced by stimulation, characterized by vocalization,facial twitching, and dystonic hand movements.Theseseizures were brief, and were controlled with coldsaline irrigation of the brain surface. The patient’sneurologic status was repeatedly assessed throughoutthe procedure and his pain was well controlled. Theoperation was completed uneventfully, and the patientretained normal speech andmuscle strength at the endof the procedure.

DiscussionEpilepsy surgery is a well-established treatment optionfor patients with seizures originating from a singleresectable focus. While temporal lobe resections arethe most common surgical treatment for epilepsyworldwide, lesional resections have also shown favor-able results, with 66% of patients seizure-free for atleast 2 years [2]. Awake craniotomies are widely usedin epilepsy surgery to facilitate intraoperative electro-corticography and cortical mapping, allowing preciseidentification of areas of the brain that control motorfunction and speech.

Prior planning and discussion are needed betweensurgeon and anesthesiologist for optimal patientpositioning and room set-up. Communication isparamount, as the surgical team should be able to hear

110

Case 35. Epilepsy surgery: intraoperative seizure

Figure 35.1. Functional magnetic resonance imaging with tractography shows corticospinal tract fibers running medially and posterior tothe mass lesion.

both the patient and anesthesiologist clearly through-out the procedure. Lines and surgical drapes shouldbe placed in such a way that allows for both access toand visibility of the patient. Furthermore, care shouldbe taken early on in the procedure to ensure thatthe patient is comfortable, while allowing the patient’shead to be positioned appropriately for surgicalexposure.

Anesthetic complications and pitfalls associatedwith awake craniotomies for epilepsy surgery havebeen reviewed by Archer et al. and Skucas et al. [3, 4].After reviewing 354 and 332 consecutive cases, respec-tively, they noted complications such as intraopera-tive seizures (3–16%), nausea and vomiting (0.9–8%),local anesthetic toxicity (0–2%), brain swelling (0.6–1%), and the need for conversion to general anesthesiadue to airway and ventilation complications (1.8–2%)(Table 35.1).

The use of long-acting local anesthetic for nerveblocks helps avoid many problems associated withawake craniotomies. Nerve blocks should target theauriculotemporal, zygomaticotemporal, supraorbital,supratrochlear, lesser occipital, and greater occipitalnerves on each side of the head, allowing for a com-plete ring of scalp analgesia. As the systemic absorp-tion of long-acting local anesthetics, such as bupiva-caine, affects the cardiovascular system and centralnervous system, it is important to communicate tothe surgeon at the beginning of the procedure thedosing limits of local anesthetics used. Toxic bloodlevels depress cardiac conduction resulting in arte-riovenous blockage, ventricular dysrhythmias (oftenrefractory to treatment), and arrest. Properly dosedand distributed, scalp blocks make intraoperative air-way obstruction andnauseamuch less likely during theprocedure, and are highly encouraged.

111


Table 35.1. Anesthetic considerations for awake craniotomiesin epilepsy surgery.

Problems Causes

Intraoperative pain Inadequate analgesia causing painmay occur at time of rigid head-pinfixation, during temporalis muscledissection, with traction on the duraclose to the middle meningeal artery,and with traction of intracerebralblood vessels

Airway obstruction May occur as the result of overzealoussedation to overcome pain whenanalgesia is suboptimal

Nausea and vomiting Can be caused by surgicalmanipulation during dural opening,temporal lobe or amygdalamanipulation, meningeal vesselhandling, inadequate analgesia, orhypovolemia

Seizure May occur as a result of decreasedanticonvulsant levels, local anesthetictoxicity, or during cortical stimulation

Modified from Costello, Cormack, 2004 [7].

Direct cortical stimulation mapping has beenshown to result in simple partial seizures in 5–20% ofcases [5]. The propensity for intraoperative seizures islikely heightened in patients with a history of epilepsy.It is therefore of paramount importance for patients tobe properly dosed with their anticonvulsants prior tosurgery. It is advisable, for example, to allow patientsaccess to their scheduled doses of anticonvulsants(with a sip of water) prior to surgery. In addition, itis useful for the anesthesiologist to know the patient’spreoperative anticonvulsant blood levels (if the med-ications taken have an established efficacy range) assubtherapeutic drug levels greatly increase the prob-ability of intraoperative seizures.

In the past, intraoperative seizures have beentreated with intravenous benzodiazepines, barbitu-rates, and anticonvulsant boluses. The difficulty inrelying on pharmacologic treatment of intraoper-ative seizures is that these medications may haveunintended and often deleterious side effects. Ben-zodiazepines, for example, stop seizures effectivelybut often induce significant sedation that interfereswith the mapping process. Phenytoin may also causeunintended problems, particularly when it is infusedrapidly. Intravenous phenytoin is preparedwith propy-lene glycol; therefore, its rapid intravenous administra-tion can result in refractory hypotension, cardiac dys-rhythmia, and even death [6]. Modern-day treatment

of intraoperative seizures relies on the application oficed saline or Ringer’s solution to the brain. This tech-nique was initially described by Sartorius in 1998 [5]and has been quickly adopted by most centers, as itis able to rapidly and reliably terminate these seizuresand eliminate the need for intravenously administeredshort-acting barbiturates or other anticonvulsantmed-ications. The principal limitation of iced saline irriga-tion is related to occasional patient discomfort elicitedby dural cooling.

ConclusionIn conclusion, epilepsy is a complex disease thatimposes great disability on those affected. Epilepsysurgery is an underutilized treatment option and anawake craniotomy is sometimes warranted to allowprecise cortical mapping. Although there are severalpitfalls for which anesthesiologists should be aware,close monitoring, team communication, the safe andjudicious use of long-acting local analgesia andpromptmanagement of intraoperative seizures make for safeperioperative care.

References1. S. Wiebe,W. T. Blume, J. P. Girvin et al. A

randomized, controlled trial of surgery fortemporal-lobe epilepsy. N Engl J Med 2001; 345:311–18.

2. J. Engel Jr. Surgery for seizures. N Engl J Med 1996;334: 647–52.

3. D. P. Archer, J. M. McKenna, L. Morin et al.Conscious-sedation analgesia during craniotomy forintractable epilepsy: a review of 354 consecutive cases.Can J Anaesth 1988; 35: 338–44.

4. A. P. Skucas, A. A. Artru. Anesthetic complications ofawake craniotomies for epilepsy surgery. Anesth Analg2006; 102: 882–7.

5. C. J. Sartorius,M. S. Berger. Rapid termination ofintraoperative stimulation-evoked seizures withapplication of cold Ringer’s lactate to the cortex.Technical note. J Neurosurg 1998; 88: 349–51.

6. B. A. Boucher, C. A. Feler, J. C. Dean et al.The safety,tolerability, and pharmacokinetics of fosphenytoinafter intramuscular and intravenous administration inneurosurgery patients. Pharmacotherapy 1996; 16:638–45.

7. T. G. Costello, J. R. Cormack. Anaesthesia for awakecraniotomy: a modern approach. J Clin Neurosci 2004;11: 16–19.

112


36 Awake craniotomy and intraoperativeneurologic declineOren Sagher and Shawn L. Hervey-Jumper

Awake craniotomy is routinely used in patients under-going epilepsy surgery or surgery on eloquent areasof brain. The term “awake” may be considered a mis-nomer, as it includes a combination of local anesthe-sia, moderate sedation and analgesia, and the asleep–awake–asleep technique. This approach allows forintraoperative electrocorticography and cortical map-ping, enabling identification of areas of the braincontrolling motor function and speech. Awake cran-iotomy allows for optimal lesion resection with min-imal postoperative neurologic dysfunction. However,there are associated complications that can contributeto both intraoperative and postoperative neurologicdecline.

Case descriptionThe patient was an otherwise healthy 27-year-oldright-handed male with a history of psychotic depres-sion who worked as a baggage handler at a local air-line. He presented with a 6-month history of dailymorning headaches after suffering a complex partialseizure. On brain magnetic resonance imaging, he wasfound to have a right frontal enhancing lesion. Givenhis normal motor examination, the patient initiallyelected to postpone surgical treatment. He was startedon antiepileptic medications with close neuroimagingsurveillance. After 6 months of follow-up, the lesionhad increased in size. Because of his young age, normalpreoperative motor examination, and tumor locationwithin the posterior frontal lobe adjacent to the pri-mary motor cortex, an awake craniotomy with image-guided resection and motor mapping was deemed tobe the safest approach.

The primary concerns of the anesthesiology teamwere (1) preoperative airway assessment and manage-ment in the event of intraoperative airway obstruction,(2) intraoperative pain management, (3) managementof intraoperative nausea and vomiting, and (4) closemonitoring for signs of seizure or neurologic decline.

The patient was brought to the operating room andbrief sedation was induced using a propofol and alfen-tanil infusion. His head was fixed in a Mayfield head-holder and he was placed in the lateral position withthe left side up. Head coordinates were then registeredusing the surgical navigational computer. Incisionaland scalp blocks were performed using a combinationof 1% lidocaine and 0.25% bupivacaine with 1:200 000units epinephrine. Location of the tumor was mappedout on the scalp and the craniotomy flap was turned.Propofol infusion was stopped to accomplish corti-calmapping by electrocorticography and cortical stim-ulation. A combination of direct visual inspectionand image-guidance was used to localize the gyrus inwhich the tumor was located. The underlying brainwas hyperemic, aswell as swollen due to the underlyingtumor.Themotor region was then mapped with corti-cospinal fibers localized directly behind the region ofthe tumor. Resection of the tumor was planned usinga subpial approach at medial and posterior margins.The patient’s speech and motor status were repeat-edly assessed.While removing tumor-infiltrated whitematter posteriorly, the patient was noted to have sig-nificantly slowed responses of his left arm and leg,with an approximate 25% reduction in his preoperativestrength. Resection was halted and the degree of resec-tion assessed by stereotactic navigation. There was asmall amount of residual tumor, but the decision wasmade to halt any further resection in order to avoidharm to nearby fibers passing from the primary motorcortex.

DiscussionModern use of awake craniotomies began withthe introduction of propofol and subsequentlydexmedetomidine. These anesthetic agents facilitateintraoperative functional cortical mapping for indi-cations that include surgery for intractable epilepsy,resection of lesions involving eloquent areas of cortex,

113


as well as avoidance of general anesthesia for patientsunable to tolerate it. Awake craniotomies allowsurgical resection of tumors previously consideredinoperable and allow for maximum resection whilemonitoring for neurologic damage. While the onlyabsolute contraindication to the performance ofawake craniotomy is patient refusal, relative con-traindications include patient psychiatric illness,prone positioning, limited patient cooperation (due tocognitive deficits, emotional lability, diminished levelof consciousness, or inability to overcome languagebarriers), significant preexisting neurologic damagethat would preclude the ability to map or monitorneurologic function, or medical comorbidities thatprevent the patient from lying still for long periods oftime.

The decision to offer an awake craniotomy is madejointly by both the anesthesiologist and neurosurgeon.A preoperative anesthesia clearance appointment isrecommended, during which potential contraindica-tions should be considered. Careful airway assessmentfor predictors of difficult intubation is important in theevent that there becomes a need for emergent generalanesthesia. Preoperatively, anesthesiologists shouldalso consider the degree of intracranial pressure,inquire regarding the type, frequency, and previoustherapy for seizures, and assess the patient’s overalldegree of anxiety.

Most centers advise against the preoperative use ofsedative medications on the day of surgery. Benzodi-azepines have unpredictable effects on patient behav-ior and impair cortical mapping. Clonidine has beenused in the past; however, the more highly selectivealpha-2-adrenoreceptor agonist, dexmedetomidine, isnow widely used to provide sedation, analgesia, andhemodynamic stability with minimal effect on corticalmapping and respiratory drive [1].

Optimal patient positioning in awake craniotomiesensures patient comfort, safety, and allows favorableconditions for functional mapping and lesion resec-tion. The operating room should be set up to allowclear communication between members of the patientcare team, visualization of the operative field andpatient, and access to the patient for incident manage-ment [2].

There are three basic categories of anesthetic tech-nique used in awake craniotomies: (1) local anesthe-sia used in isolation or in combination with sedationor periods of general anesthesia, (2) deep sedation,and (3) asleep–awake–asleep anesthesia, where gen-

Figure 36.1. Diagram showing six sites for administering localanesthesia via bilateral block in awake craniotomy. Adapted fromPiccioni, 2008 [8].

eral anesthesia is given followed by intraoperativeawakening for cortical mapping and possibly lesionresection. Local anesthesia is best applied by bilat-eral block of the six scalp nerves: auriculotempo-ral, zygomaticotemporal, supraorbital, supratrochlear,lesser occipital, and greater occipital (Figure 36.1).The most commonly used local anesthetic includes acombination of bupivacaine and lidocaine (often with1:200 000 units epinephrine, and occasionally sodiumbicarbonate). Some studies suggest the use of levobupi-vacaine rather than bupivacaine because of its reducedpeak plasma concentrations and a lower toxicity pro-file [2]. Ropivacaine is also an option.

Both sedation and asleep–awake–asleep tech-niques require that the anesthesiologist provide ade-quate sedation, hemodynamic control, and anesthe-sia while ensuring patient cooperation and alertnessfor intraoperative speech and motor testing. There arethree anesthetic techniques that have beenused to offersedation: (1) neurolept anesthesia, (2) propofol-basedanesthesia, and (3) dexmedetomidine-based anesthe-sia. Neurolept anesthesia involves the use of droperi-dol combined with a lipid soluble opiate (fentanyl,sufentanil, or alfentanil) and was the anesthetic tech-nique used for early awake craniotomies [3]. Thistechnique has largely fallen out of favor due to poordrug titratability, long duration of action, and cardiaceffects. Neurolept anesthesia has been largely replacedby propofol-based techniques. Propofol is the mostcommonly used agent for awake craniotomies due toits easy titration, short duration of action, amnesicproperties, and anti-emetic effects. Termination ofpropofol infusion 15 minutes before cortical map-ping is typically sufficient. Herrick et al. comparedneurolept and propofol-based anesthesia and noted

114

Case 36. Awake craniotomy and intraoperative neurologic decline

Table 36.1. Comparison of complications in awake craniotomy studies.

Skucasa Archerb Gignacc Herrickd Dankse Huncke f Berkenstadg Blanshardh Sarangi

Airway problems 2 − − 5/0 0/0 0 4 0.4 7/0/0

Hypoxemia 2 − 0/20/10 − − − − − −Hypertension 11 − − − 0/0 − 4 − 0/11/0

Hypotension 56 − − − − − 0 0.8 0/6/21

Tachycardia 14 − − 10/35 − − 0 − −Bradycardia 0.3 − − − − − 0 − −Seizures 3 16 10/30/10 0/41 0/20 0 8 8 0/6/0

Nausea/vomiting 0.9 8 50/30/70 10/18 − 0 0 0.8 −Poor cooperation 2 2 20/0/10 − − 0 4 − 0/3/5

Brain swelling 0.6 1 − − 0/0 0 0 0 −Local anesthetic toxicity 0 2 − − − − − − 0/0/0

Adapted from Skucas, 2006 [5].Values expressed as percentages.(–), not reported.a Retrospective review of 332 cases using unsecured airway and propofol infusion.b Retrospective review of 354 cases using unsecured airways and fentanyl with droperidol.c Prospective series of 30 cases, 3 groups with unsecured airways; fentanyl vs sufentanil vs alfentanil.d Prospective series of 37 cases, 2 groups with unsecured airways; propofol vs fentanyl + droperidol.e Prospective series of 21 cases, 2 groupswith unsecured airways; midazolam+ sufentanil+ fentanyl vs propofol+midazolam+ sufentanil+ fentanyl.f Prospective series of 10 cases, 1 group with airways secured by endotracheal intubation; volatile anesthetic+ nitrous oxide+/− propofol+/− alfentanil.g Prospective series of 25 cases, 1 group with unsecured airways; propofol+ remifentanil + clonidine.h Retrospective review of 241 cases, unsecured airways.i Retrospective review of 99 cases, 3 groups; unsecured airways propofol + fentanyl +/− midazolam +/− droperidol vs airways securedwith LMA, propofol + fentanyl vs airways secured with LMA propofol + remifentanil.

increased intraoperative seizures and pain with neu-rolept anesthesia but higher incidence of transient res-piratory depression in patients receiving propofol [4].Propofol is generally used in combination with theopiates fentanyl or remifentanil. Recently, the highlyselective alpha-2-adrenoreceptor agonist dexmedeto-midine has gained popularity in awake craniotomiesdue to its ability to offer analgesia with easy arousalwhile causing minimal respiratory effects [1].

Despite improved anesthetic and operative tech-niques, neurologic changes can occur intraoperativelydue to any number of known complications associ-ated with awake craniotomies (Table 36.1). Repeatedneurologic examinations during the resection allowfor early detection of deficit caused by surgical traumato the region of function or to fibers emanating fromit. The surgical team can then make a determina-tion about whether to halt resection or to proceeddespite the deficit. A minor deficit is often tolerated,given its propensity to improve postoperatively. Atthat stage, it is of vital importance for the anesthe-

siologist to remain cognizant of the sedative proper-ties of the anesthetic technique used, as sedation canoften exacerbate the apparent focal neurologic deficit.During resection, some surgeons prefer to supplementmonitoring with mapping of the white matter fibers.This stimulation mapping technique is similar to cor-tical mapping, and allows the surgeon to gain a bet-ter understanding of the proximity of important whitematter tracts. Patients with preexisting motor deficitscan be challenging tomap andmonitor, since it ismoredifficult to detect a change in the degree of deficit thanit is to detect the presence of a new deficit. In general,the greater the presurgical deficit, themore difficult themonitoring is likely to be.

Other systemic complications of awake craniotomymay interfere with both mapping and monitoring.For example, respiratory complications resulting fromairway obstructions can cause intraoperative hypox-emia and hypercapnia. Hypercapnia can result in brainswelling (0–1% of cases), therebymaking surgical con-ditions difficult [5]. Respiratory complications occur

115


in 0–7% of cases and are highest in obese patientsand those receiving propofol-based techniques. Intra-operative seizures occur most commonly during cor-tical stimulation and can be controlled with coldRinger’s lactate irrigation [6]. The incidence is higherin patients receiving neurolept techniques (up to 41%)compared with propofol-based techniques (0–8%) [5].Severe intraoperative pain makes for a very unpleas-ant patient experience and can limit neurologic assess-ment. Some institutions have adopted the use ofremifentanil as the opiate of choice because it can berapidly titrated. Local anesthetics at supratherapeu-tic doses or delivered directly intravenous can causeboth cardiotoxicity and neurotoxicity. Although sys-temic toxicity is rare (�2%), one should recognize theearly signs of this complication and take efforts to limitits occurrence [5]. Poor patient cooperation is seenin 0–20% of cases and is often associated with inad-equate anesthetic [5]; this can often be managed afteraddressing the underlying cause. Finally, the need toconvert to a general anesthetic does happen and hasbeen reported in several large series to occur in 0.4–2% of cases [3, 7].

ConclusionIn conclusion, all patients who might benefit fromawake neurologic testing during surgery should beconsidered for an awake craniotomy. Careful patientselection and preoperative consideration of potentialcontraindications, the use of scalp blocks, improvedanesthetic agents, and clear communication among

members of the patient’s care teamwillminimizemanypotential complications and improve patient outcomeand satisfaction.

References1. A. Y. Bekker, B. Kaufman,H. Samir et al.The use of

dexmedetomidine infusion for awake craniotomy.Anesth Analg 2001; 92: 1251–3.

2. T. G. Costello, J. R. Cormack. Anaesthesia for awakecraniotomy: a modern approach. J Clin Neurosci 2004;11: 16–19.

3. D. P. Archer, J. M. McKenna, L. Morin et al.Conscious-sedation analgesia during craniotomy forintractable epilepsy: a review of 354 consecutive cases.Can J Anaesth 1988; 35: 338–44.

4. I. A. Herrick, R. A. Craen, A.W. Gelb et al. Propofolsedation during awake craniotomy for seizures:patient-controlled administration versus neuroleptanalgesia. Anesth Analg 1997; 84: 1285–91.

5. A. P. Skucas, A. A. Artru. Anesthetic complications ofawake craniotomies for epilepsy surgery. Anesth Analg2006; 102: 882–7.

6. C. J. Sartorius,M. S. Berger. Rapid termination ofintraoperative stimulation-evoked seizures withapplication of cold Ringer’s lactate to the cortex.Technical note. J Neurosurg 1998; 88: 349–51.

7. H. J. Blanshard, F. Chung, P. H. Manninen et al.Awake craniotomy for removal of intracranial tumor:considerations for early discharge. Anesth Analg 2001;92: 89–94.

8. F. Piccioni,M. Fanzio. Management of anesthesia inawake craniotomy.Minerva Anestesiol 2008; 74:393–408.

116


37 Epilepsy surgery and awake craniotomySumeet Vadera and William Bingaman

Awake craniotomies require a great deal of coopera-tion among all members of the surgical team, espe-cially the surgeon and the anesthesiologist. It is impor-tant to be aware of the many obstacles that can beencountered during these procedures and how to suc-cessfully avoid complications. Here we describe a stan-dard awake craniotomy case, the pitfalls that wereencountered, and their effective management.

Case descriptionThe patient was a 60-year-old right-handed male witha WHO Grade II astrocytoma diagnosed in 1991 andtreated with gross total resection at that time. Thepatient presented with simple partial seizures of theright upper extremity and was noted to have persistentseizures despite an adequate trial of anticonvulsanttherapy. Preoperative evaluation included magneticresonance imaging (MRI) of the brain and video elec-troencephalography (vEEG). The MRI demonstratedan enhancing lesion in the left posterior frontal lobesuspicious for recurrent tumor (Figure 37.1a) andvEEG was concordant with seizure activity arisingfrom this region. Surgical decision making involvedplans to resect the lesion and area of cortex responsi-ble for the seizures. As this involved functional senso-rimotor cortex, the plan was for “awake” craniotomywith local anesthetic and intravenous sedation.

After the patient was transferred to the operat-ing room table, he was turned to the lateral decubitusposition with the left side positioned up. The patient’shead was turned gently and placed on a foam donut(90 degrees perpendicular to the floor). The patientwas allowed to become comfortable on the table andthen sedated for application of a Mayfield head holder(Codman Inc) and insertion of an arterial line andbladder catheter. The head fixator was applied duringintravenous sedation using amixture of dexmedetomi-dine and propofol, and only after adequate local anes-thesia at the pin sites was achieved. The head holder

was utilized to provide a reference for the stereotacticnavigation system and was not fixed to the operatingtable during the procedure in order to prevent injuryto the patient as he awakened. All pressure points werepadded and double checked by all members of theoperating room team.

The lesionwas localized using the frameless stereo-tactic navigation system and a circumferential fieldblock was performed using local anesthesia injection.Once an adequate level of sedation was achieved, theprevious incision was opened by sharp dissectionand the bone flap elevated using a high-speed aircranitome. The dura was opened after injecting localanesthetic along the middle meningeal vessels usinga 25-gauge tuberculin syringe. The lesion was againlocalized with the frameless stereotactic system andvisual inspection. At this point, intravenous sedationwas discontinued, and the patient was awakened forserial neurologic exams and cortical mapping. Duringthis 10–20 minute window, electrocorticography(ECoG) and somatosensory evoked potential identifi-cation of the rolandic fissure can be accomplished. Asthe patient awakens, the surgical team must be carefulto anticipate and prevent patient head movement toavoid potential brain injury.

Somatosensory evokedpotentials (SSEPs)were ini-tially done using stimulation of the median nervefor central sulcus mapping. This mapping procedurewas correlated with the expected position based onpatient anatomy and stereotactic navigation. Intraop-erative bipolar electrical stimulating electrodes werealso used to define functional tissue including theperi-rolandic region and motor speech areas. Dur-ing stimulation, current is gradually increased at thesite tested until a positive motor or language responseis seen or a maximum current of 15 milliamperes isreached. If no response is seen, the next site to be stim-ulated is chosen and the protocol repeated.Monitoringof the electroencephalogram using cortical subduralelectrodes is performed during stimulation to detect

117


(a)

(b)

Figures 37.1a and 37.1b. Pre- and postoperative gadolinium-enhanced magnetic resonance imaging (MRI) of the brain in apatient with left frontal astrocytoma. (a) Preoperative coronal MRIwith contrast demonstrating enhancing lesion in left frontal lobejust anterior to motor strip. (b) Postoperative contrast enhancedaxial MRI demonstrating resection of lesion.

cortical afterdischarge potentials and seizures. If theseoccur, stimulation is stopped and ice cold saline irriga-tion is placed on the brain. Additionally, intravenousantiepileptic medications can be administered.

Finally, ECoG was performed to define the inter-ictal irritative zone (that region of brain irritated bythe lesion). In this case, an area just anterior to theprecentral gyrus and adjacent to the lesion was notedon ECoG to demonstrate continuous spiking. At thispoint resection of the lesion and adjacent irritatedcortex was begun with real time serial neurologicexamination of the patient. This is the real advantageof “awake” craniotomy and maximizes the chances ofresecting the entire lesion while avoiding neurologicdeficit. Resection of the lesion proceeded with bipolarcoagulation, subpial aspiration, and micro instru-ments. All pial vascular structures were preservedin order to avoid ischemic injury to adjacent func-tional cortex. During the resection, serial neurologicexams were conducted by a neurologist to ensurethat functional cortex was not being injured. Grosstotal resection of the tumor was adequately achieved(Figure 37.1b) and the lesion was sent for frozen andpermanent pathology. At this point, the patient wassedated and the wound was closed in the normalanatomic layers.

Upon awakening, the patient was noted to be athis neurologic baseline with no new deficits noted onexam.

DiscussionThis case demonstrates some common adversitiesfaced in epilepsy surgery during “awake” craniotomy.The first obstacle encountered during the procedurewas the onset of simple partial seizures during awakecortical stimulation. Studies have shown that admin-istration of ice-cold Ringer’s lactate solution directlyto the cortical surface is a rapid method to terminatecortical irritation and seizure onset [1]. Barbituratesand antiepileptic medications can also be used in thesecases to assist with seizure cessation.

Electrocorticography was important in this caseto localize the interictal irritative zone and delineatethis region from areas of functional cortex. Whencontinuous repetitive interictal spiking is measured,it is important to remove it to achieve an acceptableseizure-free outcome [2].

118

Case 37. Epilepsy surgery and awake craniotomy

Functional mapping of motor cortex is widely con-sidered the “gold-standard” technique to identify elo-quent cortical areas. When combined with “awake”craniotomy and frequent neurologic examinations,the risks of new postoperative neurologic deficits aregreatly decreased [3]. Even when these safety mea-sures are employed, there is still the risk of tempo-rary or permanent neurologic deficits arising post-operatively. These risks increase as the proximity ofthe “lesion” to eloquent cortex increases [4]. Althoughgross total resection was achieved in this case, sometumors that are located within eloquent cortex may beunresectable and a more limited debulking of tumorfor tissue diagnosis and reduction of mass effect isperformed [5]. Sedation in awake craniotomy usuallyis best performed using a mixture of dexmedetomi-dine and propofol. In order to allow the patient to beawake enough for neurologic testing, we usually givethe patient minimal narcotics and benzodiazepines.The other benefit of giving minimal amounts of nar-cotics during the procedure is to avoid the depressanteffect on the respiratory system with resultant hyper-carbia and dangerous increase of intracranial pres-sure. The addition of dexmedetomidine to propofoldecreases the amount of propofol needed for seda-tion and allows the maintenance of spontaneous res-piration. The other benefit of dexmedetomidine isits inhibitory effect on hypercarbia-induced cerebral

vasodilation and consequently intracranial hyperten-sion. Patient education and a thorough discussionof the risks and benefits of such a procedure areimportant prior to surgical intervention being offeredbecause of the potential complications that can beencountered during this procedure.

References1. C. J. Sartorius,M. S. Berger. Rapid termination of

intraoperative stimulation-evoked seizures withapplication of cold Ringer’s lactate to the cortex.Technical note. J Neurosurg 1998; 88: 349–51.

2. A. Palmini, A. Gambardella, F. Andermann et al.Intrinsic epileptogenicity of human dysplastic cortexas suggested by corticography and surgical results. AnnNeurol 1995; 37: 476–87.

3. A. R. Walsh, R. H. Schmidt,H. T. Marsh. Corticalmapping and resection under local anaesthetic as anaid to surgery of low and intermediate grade gliomas.Br J Neurosurg 1990; 4: 485–91.

4. F. B. Meyer, L. M. Bates, S. J. Goerss et al. Awakecraniotomy for aggressive resection of primarygliomas located in eloquent brain.Mayo Clin Proc2001; 76: 677–87.

5. R. C. Rostomily,M. S. Berger, G. A. Ojemann et al.Postoperative deficits and functional recoveryfollowing removal of tumors involving the dominanthemisphere supplementary motor area. J Neurosurg1991; 75: 62–8.

119

Part IV Spine surgery. Spinal cord injuryCase

38 Acute surgery: spinal and neurogenic shockRuairi Moulding and Scott T. McCardle

Spinal cord injury (SCI) is a devastating and life-threatening condition with an incidence of 27–81cases/million/year in theUSA, which equates to 10 000new patients per year [1, 2]. There are two stagesto the development of SCI. The primary injury leadsto disruption of the neuronal tissue of the spinalcord via a variety of mechanisms including directtrauma, bony impingement or compression, mechan-ical stretching or interruption of the blood supply.These initial insults to the integrity of the spinal cordare compounded by the secondary injury. This sec-ondary injury is a culmination of an intense inflam-matory response leading to edema, cell death and mayresult in further nonreversible damage to the spinalcord.

Both spinal shock and neurogenic shock can resultfrom a SCI. They are distinct entities. Spinal shockoccurs with the loss of autonomic and reflex activityof the spinal cord caudal to the injury, resulting in thecharacteristic flaccid paralysis. This flaccidity, lastingup to 4 weeks, may lead eventually to spasticity withthe loss of the descending inhibitory controls. Neu-rogenic shock is the phenomenon whereby patientswith SCIs lose the sympathetic control of cardiovas-cular functions with only unopposed parasympatheticcardiac responses remaining.

Case descriptionThe patient was a 26-year-old male with no signifi-cant medical history who presented to the emergencydepartment by ambulance approximately 2 hours afterdiving 12 feet into a shallow pond in a rock quarry.Thepatient presented with a loss of sensation and motorcontrol from the neck down. On examination he wasin a cervical collar with sand bags for immobilization.He was in severe respiratory distress with a respira-tory rate of 36 and his blood oxygen saturation (SaO2)was 90% while breathing high-flow oxygen via a non-rebreathing mask. He had thready distal pulses with

poor capillary refill, cold peripheries and a central tem-perature of 36.5 ◦C. His noninvasive blood pressureon admission was 74/51 with a heart rate of 52 beatsper minute. Computed tomography andmagnetic res-onance imaging showed complete cord transection atthe C5–C6 level.

The patient was scheduled for an immediate pos-terior cervical decompression and stabilization by theneurosurgical service. The principal anesthetic con-cerns were (1) need for immediate tracheal intuba-tion for respiratory insufficiency, (2) management ofthe airway in a patient with an unstable cervical injurywith a full stomach, (3) prevention of secondary SCI,(4) the potential for labile hemodynamics secondary tothe sympathectomy and lack of vasomotor tone belowthe level of the injury, (5) treatment of any concomi-tant injuries, and (6) prone positioning with an unsta-ble cervical spine injury.

The patient was evaluated in the emergency roomfor other associated injuries and high-dose methyl-prednisolone was started. Prior to induction, atropinewas given for vagolysis in preparation for the responseto succinylcholine and concern for extreme bradycar-dia or asystole due to impairment of sympathetic car-dioaccelerators. The potential for a difficult intuba-tion was anticipated and a rapid sequence inductionwas performed using succinylcholine, inline stabiliza-tion, and cricoid pressure with video-assisted laryn-goscopy. A grade I view was seen at laryngoscopy,an endotracheal tube was inserted, and the positionchecked.

Arterial and central venous catheters were insertedas well as a urinary catheter with monitoring of blad-der and nasal temperatures. Large-bore intravenous(IV) access was obtained and fluid resuscitation insti-tuted. The hemodynamic goal during the anestheticwas to prevent secondary SCI. First-line treatmentof hypotension was with IV fluid, both crystalloidand colloid. Norepinephrine was administered toobtain a targeted mean arterial pressure of 85 mmHg.

121

IV. Spine surgery. Spinal cord injury

Figure 38.1. Cervical spine X-ray showing a fracture-dislocation ofC5–C6 with risk of severe spinal cord compromise.

Figure 38.3. Computed tomography scan of compressionfractures of C5 and C6 with severe canal stenosis. These findings areconsistent with a hyperflexion type injury with retropulsed fracturefragments.

Figure 38.2. Computed tomography scan of C6 vertebral bodycompression fracture with fragments extending into the canal.

Positioning the patient prone required “log-rolling” inorder to maintain spinal integrity and reduce furtherneurologic damage.

Maintenance of anesthesia included propofol andremifentanil infusions, in order to facilitate spinal cordmonitoring with somatosensory and motor evokedpotentials. Nitrous oxide was avoided due to the possi-bility of occult intracranial or thoracic injuries, whereexpansion of entrapped air could lead to devastatinghemodynamic and neurologic complications.The casewas completedwithminimal blood loss and the patientwas taken to the intensive care unit sedated, ventilated,and with a norepinephrine infusion for blood pressurecontrol. Emergence and neurologic examination werenot attempted due to hemodynamic instability.

DiscussionSpinal cord injuries occur most often in young adultswith a 4:1 male to female preponderance. Nearly 80%of these injuries are related to either motor vehi-cle accidents or traumatic falls. The cervical spineis the most common level of injury and levels C1–2 and C5–7 are the most vulnerable (Figures 38.1–38.3) [3]. Traditionally, the leading cause of mortal-ity in both the acute and chronic SCI patient has beendue to respiratory and renal complications [4]. How-ever, there is growing recognition that cardiovascular

122

Case 38. Acute surgery: spinal and neurogenic shock

complications may contribute significantly to earlyand late mortality [1].

Complete SCIs above the T6 level result in neuro-genic shock related to sympathectomy and unopposedvagal tone below the level of the injury [4]. There isa direct link between severity of SCI at the cervicalor high-thoracic level and severity of cardiovascularsymptoms.Neurogenic shock can last for up to 5weekspost-injury [1]. Incomplete SCIs and those below T6may or may not develop cardiovascular compromise.

The characteristic components of neurogenicshock are hypotension (systolic blood pressure below90 mmHg) and bradycardia. Although bradycardiais most common, other rhythm abnormalities mayoccur due to autonomic instability. Atrioventricularblocks, supraventricular tachycardias, ventriculartachycardias, and even cardiac arrest may all occurin the context of a severe SCI [1]. The bradycardiaseen with SCI is due to the unopposed vagal toneand is amenable to treatment with vagolytics such asatropine, or transcutaneous pacing.

Sympathetic tone is an important factor in car-diovascular homeostasis and SCI patients may havereduced or absent control of coronary blood flow, car-diac contractility, and heart rate. Furlan and Fehlings[1] identified five elements of the autonomic cir-cuits as potentially contributing factors: (1) disrup-tion of the descending cardiovascular (or vasomotor)pathways; (2) morphological changes in the cardiacand vasomotor sympathetic preganglionic neurons;(3) sprouting and the potential formation of inappro-priate synapses with spinal interneurons; (4) abnormalspinal afferents, and (5) development of altered sym-pathetic neurovascular transmission and smoothmus-cle responsiveness. As only the first element (disrupteddescending pathways) may be present in the acute set-ting, this is the most likely contributing factor in thedevelopment of neurogenic shock.

A patient with neurogenic shock presents multi-ple clinical dilemmas for the anesthetic team.The careof patients with such devastating neurologic injuries ismost appropriately handled at a Level I Trauma Cen-ter [2, 5]. However, the initial treatment and resus-citation is often started at a remote location. In allinstances it is recommended to adhere to AdvancedTrauma Life Support (ATLS R©) guidelines [2]. Spinalcord injuries must be evaluated like all traumas thatpresent to the emergency room, with stabilizationof the patient and examination for any accompany-ing injuries. Patients with cervical spine injuries may

have other spinal injuries remote from the primarysite.

The timing of decompression and stabilizationof spinal cord injuries may vary but surgery within24 hours can improve neurologic outcomes [1, 5].Surgery for spinal cord injury patients is not limited toalleviating cord compression or cervical instability butmay be for other non-spinal injuries such as fracturerepair or tracheotomy.The presence of hypotension inthe context of a SCI should not preclude the physicianfrom looking for alternative causes of the hypotension.Associated injuries such as pelvic, long bone, head, andvisceral injuries are all potential causes of hypotensionin the trauma patient.

High-dose methylprednisolone (30 mg/kg startedwithin 8 hours of injury, and given over a 45-minuteperiod, followed by 5.4 mg/kg/hour over subsequent23 hours) has been shown in some spinal cord injuriesto improve neurologic outcomes (but not necessar-ily mortality or quality of life) by the National AcuteSpinal Cord Injury Study (NASCIS I, II, and III)[3]. However, these studies remain controversial, withhigher rates of infection and gastrointestinal bleedingin the steroid treatment group.

Difficult intubation must be anticipated secondaryto the need for cervical immobilization, the need forrapid sequence induction, use of cricoid pressure, pre-vertebral swelling secondary to hematoma, blood ordebris in the airway, and distortion of the airwayfrom maxillofacial trauma [3]. Alternative intubat-ing devices including intubating laryngeal mask air-ways, video-assisted laryngoscopy, fiberoptic scope,and cricothyroidotomy equipment should be imme-diately available in the event that direct laryngoscopyproves unsuccessful.

It is important to maintain thermal homeostasisbecause hypothermia may worsen secondary injury.Care should be taken with warming SCI patients.Sympathectomy-induced vasodilatation may lead tohypothermia, however, lack of sweating below the levelof injury may lead to iatrogenic hyperthermia fromwarming devices.

Intra-arterial blood pressuremanagement is highlyrecommended to ensure adequate perfusion pressurefor the spinal cord with a recommended mean arte-rial pressure of 85 mmHg [4]. It is important tonote that secondary injury from hypoperfusion andhypoxia may extend the spinal cord damage cranially.Therefore, early recognition and aggressive treat-ment of neurogenic shock is recommended. Although

123


intravenous fluids should be used initially, vasopres-sors may be needed to maintain adequate spinalcord perfusion pressure. Indeed SCI patients may beat greater risk of pulmonary edema through over-resuscitation and care should always be taken touse appropriate volumes of intravenous fluid. Centralvenous accessmay be required for resuscitation, evalu-ation of filling status and for infusions of vasopressors.

ConclusionIn conclusion, spinal cord injury is a devastating, life-threatening condition that produces a number of phys-iologic and anatomical derangements that must beacutely managed by the anesthetic team. The anes-thetic goals should focus on establishing an airway(surgical, if necessary), and close hemodynamic andrespiratory monitoring to maximize spinal cord per-fusion to prevent secondary SCI. Furthermore, main-tenance of an anesthetic that is amenable to a balanceof appropriate surgical, neurophysiologic monitor-ing, and anesthetic considerations is paramount. The

postoperative care of these patients might be extensiverequiring multiple further anesthetics. Anesthesiolo-gists must be familiar with the unique long-term com-plications of SCI such as spasticity, autonomic hyper-reflexia, and chronic ventilator support that may alteranesthetic management.

References1. J. C. Furlan,M. G. Fehlings. Cardiovascular

complications after acute spinal cord injury:pathophysiology, diagnosis, and management.Neurosurg Focus 2008; 25: E13.

2. I. Miko, R. Gould, S. Wolf et al. Acute spinal cordinjury. Int Anesthesiol Clin 2009; 47: 37–54.

3. P. Veale, J. Lamb. Anaesthesia and acute spinal cordinjury. CEACCP 2002; 2: 139–43.

4. M. Denton, J. McKinlay. Cervical cord injury andcritical care. CEACCP 2009; 9: 82–6.

5. L. A. Wuermser, C. H. Ho, A. E. Chiodo et al. Spinalcord injury medicine. 2. Acute care management oftraumatic and nontraumatic injury. Arch Phys MedRehabil 2007; 88: S55–61.

124

Part IV Spine surgery. Spinal cord injuryCase

39 Returning patient with autonomichyperreflexiaSherif S. Zaky

Patients with spinal cord injuries often return for sub-sequent surgeries. Loss of descending spinal cord inhi-bition and neuronal alterations of the cord itself maybe associated with exaggerated sympathetic responsesto what normally would be innocuous stimuli.

Case descriptionThis patient was a 44-year-old male who sustained amotor vehicle accident 6 years previously that resultedin paraplegia secondary to T6 spinal cord injury. Thepatient also had a history of hepatitis C virus, throm-bocytopenia, vascular insufficiency, and asthma. Thepatient developed gangrene of his left big toe extend-ing to the medial side of his foot.

The patient presented to the operating room fordebridement of his left foot with possible mid-tarsalamputation. He did not have sensation below T6, buthe responded to painful stimulation below that levelwith jerky movements that would prevent proceedingwith the surgery. The surgeon was concerned aboutthe jerky movements and the possibility of autonomichyperreflexia if the procedure was done under con-scious sedation or local anesthesia, as was initiallyplanned.

Although this could have been done under gen-eral anesthesia, the patient and anesthesia team pre-ferred to do it under regional given the patient co-morbidities such as reactive airway disease. However,neuraxial blockade was avoided in this patient dueto the thrombocytopenia. Regional peripheral nerveblock was therefore deemed more appropriate, so thesciatic nerve was blocked under ultrasound guid-ance in the popliteal fossa. Next, the saphenous nervewas blocked in the Sartorius canal, also with ultra-sound guidance. The surgery was performed underthe block and light sedation using midazolam withoutany complications or significant hemodynamic pertur-bations. The patient was discharged home the sameday.

DiscussionSpinal cord injuries can be caused by either traumatic(e.g., motor vehicle accident, falls, water or skiing acci-dents) or nontraumatic causes (e.g., vascular events orneoplasms). Most of these patients survive to returnfor elective surgeries, most commonly urological andorthopedic procedures. After the initial injury, thesepatients go through different phases [1].

Initial consequencesThe first response to spinal cord injury can be a briefbut severe increase in sympathetic discharge leadingto transient increase in afterload (on the order of min-utes) that sometimes causes left ventricular failure,subendocardial infarction, and/or pulmonary edema.Following the acute response, neurogenic shock canoccur in several days to 6–8 weeks after the initialinjury and is characterized by hypotension, bradycar-dia, and unopposed vagal reflexes. The predominantparasympathetic tone can lead to severe bradycardia –even asystole – especially with tracheal intubation andsuctioning [2].

The loss of sympathetic discharge leads todecreased preload and hypotension. Myocardialdysfunction may also play a role in the hypotension.

Delayed consequencesAfter the initial phase of spinal shock, changes inthe neuronal connections occur in the spinal cordbelow the level of injury. Presynaptic boutons multi-ply and form chaotic, inappropriate reflexes. Interneu-rons excited by the afferent inputs synapse withpreganglionic sympathetic neurons in the intermedi-olateral gray column. This will result in a widespreadinappropriate sympathetic response, which lacks theusual descending inhibition from higher centers,leading to profound vasoconstriction [3–6]. Auto-nomic hyperreflexia is characterized by hypertension,headaches, pallor, or flushing above the level of injury

125


and profuse sweating with sensory stimulation toareas below the level of the injury. Other symptomsinclude nausea, penile erection, pupillary changes, andHorner’s syndrome.The reported incidence is variabledue to lack of consensus on the definition of autonomichyperreflexia, but it is thought to occur about 60–85%of the time after spinal cord injury at or above the T6level [3, 7].

Anesthetic goals for patients withautonomic hyperreflexiaThe main goal of anesthesia is to blunt or preventthis reflex. This can be achieved with either gen-eral or regional anesthesia. General anesthesia canbe employed while maintaining adequate depth usingpotent volatile anesthetic, narcotics, and systemic sym-patholytics to decrease the afferent sensory discharge.This usually requires a deeper level of anesthesia andfrequently this can cause further hypotension andhemodynamic instability. The exact level or endpointfor anesthetic titration is not clear. Succinylcholineshould be avoided from 2 days up to 1 year after theoriginal injury due to increased risk of hyperkalemiasecondary to denervation hypersensitivity.

Regional anesthesia, on the other hand, is preferredbymany anesthesiologists as itmore reliably blocks theafferent sensory signal before it reaches higher centers.This can be achieved with neuraxial (spinal or epidu-ral anesthesia) or peripheral nerve blocks. In general,spinal anesthesia is more reliable than epidural anes-thesia in preventing autonomic hyperreflexia. How-ever, it might be impossible to determine the levelof the block and the dose–response characteristicsmight not be the same in spinal cord injury patients[8, 9]. Peripheral nerve blocks using nerve stimula-tion might not give the sensory or the motor responsethat would be expected in normal patients, which canmake it technically difficult. Epinephrine containinglocal anesthetic solutions should be avoided due tohigher sensitivity of spinal cord injury patients to cat-echolamines.

Anesthetic management of the patient witha history of spinal cord injuryBefore induction of anesthesia, obtaining large-boreintravenous access is recommended, followed by fluidpreloading as these patients have reduced blood

volume. Typically, 500–1000 mL of crystalloid is givento reduce the risk of hypotension associated withinduction of general or neuraxial anesthesia [10].Extra caution should be taken during positioning aspressure sores are more likely due to poor positioning.Anticholinergic drugs are commonly used, especiallyduring the phase of neurogenic shock to avoid the riskof asystole due to unopposed parasympathetic tone. Itis important to note that the risk of hypothermia issignificantly higher due to impaired thermoregulationbelow the level of injury. The use of warming devices,especially surface warming, is essential.

Although gastric emptying is delayed in spinalcord injury patients, no significant increased riskof aspiration has been established; routine rapidsequence induction for the spinal cord injury patient iscontroversial.

If autonomic hyperreflexia occurs, it is typicallybrief and self-limiting. Initial management includescessation of the stimulus and an increase in anes-thetic depth. Labetalol, nifedipine, and nitroglycerinare the most commonly used drugs when pharmaco-logic management is required. Other medications thatcan be used include propranolol, esmolol, and mida-zolam [10]. Magnesium sulfate has also been success-fully used formanagement of autonomic hyperreflexiain the chronic spinal cord injury patient [11].

ConclusionThe number of patients with autonomic hyperreflexiareturning for various surgeries is increasing due toimproved medical management of urinary tract andrespiratory tract complications in patients with spinalcord injury [1]. Perioperative management of thesepatients requires knowledge of the risks associatedwith this phenomenon as well as the pathophysiology.Whether general or regional anesthesia is used, therewill be specific challenges facing the anesthesiologist.Adequate preparation is usually the key for successfulmanagement.

References1. P. R. Hambly, B. Martin. Anesthesia for chronic spinal

cord lesions. Anesthesia 1998; 53: 273–89.2. H. L. Frankel, C. J. Mathias, J. M. L. Spalding.

Mechanisms of reflex cardiac arrest in quadriplegicpatients. Lancet 1975; 2: 1183–5.

126

Case 39. Returning patient with autonomic hyperreflexia

3. N. B. Kurnick. Autonomic hyperreflexia and itscontrol in patients with spinal cord lesions. AnnInternal Med 1956; 44: 678–86.

4. A. Fraser, J. Edmonds-Seal. Spinal cord injuries: areview of problems facing the anesthetist. Anesthesia1982; 37: 1084–98.

5. A. V. Krassioukov, L. C. Weaver. Reflex andmorphological changes in spinal preganglionicneurons after cord injury in rats. Clin Exp Hypertens1995; 17: 361–73.

6. A. V. Krassioukov, L. C. Weaver. Episodichypertension due to autonomic dysreflexia in acuteand chronic spinal cord injured rats. Am J Physiol1995; 268: 2077–83.

7. R. Lindan, E. Joiner, A. A. Freehafer et al. Incidenceand clinical features of autonomic dysreflexia in

patients with spinal cord injury. Paraplegia 1980; 18:285–92.

8. J. A. Stirt, A. Marco, K. A. Conklin. Obstetricanesthesia for a quadriplegic patient with autonomichyperreflexia. Anesthesiology 1979; 51: 560–2.

9. P. G. Loubser,W. H. Donovan. Diagnostic spinalanesthesia in chronic spinal cord injury pain.Paraplegia 1991; 29: 25–36.

10. J. Goy. Spinal injuries. In Loach A, ed. OrthopaedicAnaesthesia. London: Edward Arnold, 1994;145–57.

11. N. A. Jones, S. D. Jones. Management of lifethreatening autonomic hyperreflexia usingmagnesium sulfate in a patient with a high spinal cordinjury in the intensive care unit. Br J Anesthesia 2002;88: 434–8.

127

Part IV Spine surgery. Complex spine surgeryCase

40 Preoperative evaluationJulie McClelland and Ellen Janke

Indications for spinal surgery include congenitaldefects, tumors, infection, hematomas, trauma, arte-riovenous malformations, herniated discs, and degen-erative disease [1]. Operative treatment ranges fromminimally invasive microsurgery to open proceduresinvolving the entire spine.The preoperative evaluationprovides the opportunity for assessment of patient,surgical, and anesthetic risks in order to formulate anappropriate anesthetic plan.

Case descriptionThe patient was an otherwise healthy 27-year-oldfemale, who presented with new onset of bilateralupper extremity weakness and sensory changes in thesetting of 6 months of neck and upper back pain.Magnetic resonance imaging demonstrated a destruc-tive paraspinal lesion involving the posterior elementsof C7, T1, and T2 with pathologic fractures involv-ing the right C7 lamina and T1–T2 spinous pro-cesses. Pathology from computed tomography-guidedbiopsy revealed hemangioma. The patient was admit-ted for C6–T2 laminectomy and C5–T4 posteriorspinal fusion with endovascular embolization of thelesion preceding the case. Electrophysiologic moni-toring with somatosensory (SSEP) and motor evokedpotentials (MEP) was planned.

Review of the patient’s drug history revealedchronic use of combined acetaminophen 500 mg andhydrocodone 5 mg, 6–8 tablets per day for pain con-trol. Physical examination was notable for bilateral4/5 strength of wrist flexion, extension and grip, withdecreased sensation of both arms in the ulnar andradial nerve distributions. Cervical spine range ofmotion was limited due to pain; airway examinationwas otherwise normal. Laboratory tests included com-plete blood count, prothrombin and partial thrombo-plastin time, and type and crossmatch for four unitsof packed red blood cells. Two large-bore peripheralintravenous cannulae were inserted. A radial arterial

line was established for continuous blood pressuremonitoring and blood sampling.

Important anesthetic considerations in this caseincluded (1) the airway (possible difficult intubation),(2) a prolonged procedure in the prone position, (3) ananesthetic technique allowing for optimal electrophys-iologic monitoring and prompt emergence, (4) fluidmanagement with regards to potential for large bloodloss (despite embolization of mass), (5) adequate post-operative pain control, and (6) the potential for post-operative visual loss.

Anesthetic induction with fentanyl and propo-fol was followed by neuromuscular blockade withsuccinylcholine. The trachea was intubated witha 7.0 mm LITA (Laryngotracheal Instillation ofTopical Anesthesia) endotracheal tube, using theBullard laryngoscope. Anesthesia was maintainedwith isoflurane �0.5 minimal alveolar concentration(MAC), as well as propofol and sufentanil infusions.After turning the patient prone, SSEPs and MEPswere unchanged from baseline. Operative time was9.5 hours with 2 liters of blood loss. Fluid replacementconsisted of 3 liters of lactated Ringer’s and 2 litersof albumin 5%. The patient’s blood pressure wasmaintained within 30% of her baseline values, usingvolume resuscitation and vasopressor support. Thecase was completed uneventfully. Emergence wasprompt and smooth; neurologic examination demon-strated a small but clear improvement of bilateral gripstrength, exam was otherwise unchanged from thatpreoperatively.

DiscussionCo-morbidities associated with spinal disease rangefrom acute traumatic spinal column instability orspinal shock to decreased cardiopulmonary func-tion from chronic thoracic cage deformity [2]. Dis-ease states and their anesthetic implications for spinesurgery are summarized in Table 40.1.

128


Table 40.1. Anesthetic considerations for common diseases related to the spine.

Disease Anesthetic consideration

Anklyosing spondylitis Restricted movement of the sacroiliac joints and spineVascular inflammation that may lead to aortitis and aortic insufficiencyPulmonary fibrosisPoor chest wall compliance

Carotid or vertebrobasilar insufficiency Avoid extremes of cervical spine ROM when positioningMaintain SBP within 20% of baselineMaintain MAPs within patient-specific norms of autoregulation

Cervical spine ROM restriction (includinginstability)

Fiberoptic intubation, awake or asleepOther intubation adjuncts: Bullard larygoscope, intubating LMA, lightwand, Glidescope,C-MAC, etc.

Muscle weakness (including dystrophies,bedridden or wheelchair-bound)

Exaggerated hyperkalemic response to succinlycholineMay be resistant to nondepolarizing NMBDMay have compromised respiratory function; medical treatment should be optimized

Oncologic disease Often not isolated to spine, may affect bone marrow, brain, heart, kidney, liver, or lungsChemotherapeutic and radiation changes to organ systems

Rheumatoid arthritis Increased difficulty for intubation (including cervical ROM, mouth opening and jawprotrusion)

Atlantoaxial instabilityChronic steroid use, may need stress dosePolypharmacy with rare drugs

Scoliosis Restrictive lung disease → pulmonary HTN → RHF → cor pulmonale10–25% patients have associated neuromuscular diseases and congenital anomaliesincluding congenital heart disease

Vital capacity reliable prognostic indicator of perioperative reserveVital capacity �40% of predicted, patient likely to need prolonged ventilator support

Trauma Acute phase:sequelae of other injuriesspinal or neurogenic shockassume cervical spine instability

Post-acute phase T6 injuries and above:autonomic hyperreflexiapoikilothermia (inability to regulate temperature)

HTN, hypertension; NMBD, neuromuscular blocker drug; LMA, laryngeal mask airway; MAP, mean arterial pressure; RHF, right heart failure;ROM, range of motion; SBP, systolic blood pressure.

General principles – preoperative historyand physical examinationA comprehensive history and physical examination isa crucial component of preparation for complex spinesurgery. Particular attention should be paid to thepatient’s baseline neurologic status and airway exami-nation as well as to the disease process leading to theneed for operative intervention. An accurate drug his-tory is essential.

Cardiopulmonary function may be compromisedin patients due to a primary disease process; sec-ondary to restrictive chest wall disease from scolio-sis, ankylosing spondylitis, or rheumatoid arthritis; orfrom conditions resulting in muscle weakness. Com-plex spine surgerymay be considered intermediate risk

for cardiac events. Preoperative electrocardiogramandchest radiograph should be obtained if indicated bythe patient’s age and co-morbidities; however, if nor-mal they do not exclude significant cardiopulmonarypathology. If cardiovascular disease is suspected or isotherwise difficult to evaluate due to exercise limita-tions, a chemically induced stress test may be con-sidered. Pulmonary function testing may be usefulto distinguish cardiac from pulmonary disease in thepresence of symptoms such as dyspnea, especially if athoracotomy is required (see below). An arterial bloodgas obtained prior to anesthesia may be helpful. Pro-longed ventilation postoperatively can be anticipatedin some cases and should be discussed preoperativelywith the patient.

129

IV. Spine surgery. Complex spine surgery

The presence of renal and hepatic disease mayinfluence drug dosing and other aspects of anestheticmanagement; if indicated by history and physicalexamination, evaluation of serum electrolytes, renalfunction, and hepatic function may help establishseverity. The presence of diabetes has implicationsfor multiple organ systems and the need for invasivemonitoring.

In cases where considerable blood loss is possible,a hematocrit, platelet count, coagulation studies andtype and screen are reasonable. Type and screen ortype and cross may be indicated (see below, section onblood loss).

In all cases,medical treatment should be optimizedpreoperatively.

Airway/cervical spine instabilityThe airway examination may be notable for limitedcervical spine range of motion as well as instabilityincreasing the risk of spinal cord injury. In the casedescribed above, neurosurgical evaluation deemed thecervical spine to be stable. However, the Bullard laryn-goscope was used for intubation because it facilitatesvisualization without the need for excessive motionof an unstable spine. Other adjuncts for intubation inpatients with a difficult airway from spine disease arelisted in Table 40.1.

PositioningTheoperative approach for complex spine surgerymaybe anterior, posterior, or lateral, alone or in combina-tion. The anterior approach may be through the neckor the abdomen. For the latter, a general surgeon maybe involved and a bowel prep required.

The prone position is associated with numerouscomplications, including decreased venous return, cer-vical spine injury, endotracheal tube malposition orobstruction, direct pressure injuries, visceral ischemia,macroglossia and oropharyngeal swelling, peripheralnerve injury, postoperative visual loss, and ischemicinjury from arterial occlusion [3]. Use of the Mayfieldhead frame allows free access to the head and face;its application is highly stimulating and needs to beanticipated. Skeletal deformity may lead to difficultiesin achieving optimal prone positioning. Postoperativevisual loss is a rare complication and is discussed else-where in this volume. Strategies for avoiding compli-cations as a result of prone positioning can be found inTable 40.2.

Table 40.2. Strategies for avoiding complications in the proneposition.

Complication Strategy

ETT obstruction/malposition

Reinforced tube

Direct pressure injury Mayfield head frameEnsure eyes, chin, and nose are free

from contact with any surface;check vigilantly

Pad all pressure pointsPlace breasts neutrally or mediallyKeep male genitalia free of

compression from bolsters orthighs

Increased thoracic/abdominal pressure

Allow for chest excursion and freeabdominal movement by using chestand thigh rolls and/or special mattressfor OR table

Peripheral nerve injury Avoid positions known to cause painor paresthesias when patient isawake

Pad axillary and ulnar neurovascularbundles

Arms at sides when turning supine toprone

Consider use of SSEP and MEP tomonitor for brachial plexus ischemia

Shoulder abduction �90 degreesand elbows placed in flexion

Swelling/dependentedema of the tongueand oropharynx

Judicious use of crystalloids for fluidreplacement

Check for ETT cuff leak prior toextubation

Vascular occlusion Avoid extremes of cervical range ofmotion

Watch for signs of jugular venousoutflow obstruction

ETT, endotracheal tube; LITA tube, laryngotracheal instillation oftopical anesthesia; MEP, motor evoked potential; OR, operatingroom; SSEP, somatosensory evoked potential.

The lateral position may be used for resection oftumors located anterior to the spinal cord. A thora-cotomy may be required, with use of a double-lumenendotracheal tube to facilitate surgical exposure. Theability of the patient to tolerate one-lung ventilationshould be ascertainedpreoperatively. Skeletal deformi-ties may be important factors in positioning when thelateral approach is used.

Anesthetic techniqueAnesthetic agents used for complex spine surgerymust be compatible with neurophysiologicmonitoringof central nervous system function (SSEP, MEP,

130


electromyography) and permit rapid emergence fortimely postoperative neurologic assessment. Motorevoked potential monitoringmay preclude neuromus-cular blockade; the presence of an endotracheal tube inunparalyzed patients can be hazardous. Blunting air-way reflexes with local anesthetics topically or withappropriate nerve blocks, careful titration of opiates,and ensuring an adequate depth of anesthesia areessential.

In addition to standard monitors, spine surg-eries often warrant continuous direct blood pressuremonitoring with an arterial catheter. Maintenance ofmean arterial pressures within the patient’s presumedautoregulatory range potentially helps avoid spinalcord ischemia from hypotension and optimizes hemo-dynamic conditions for electrophysiologic monitor-ing. Direct arterial monitoring also allows for assess-ment of systolic pressure variation to help guideresuscitation efforts, and permits serial blood sam-pling for determination of hematocrit, plasma glu-cose, and arterial blood gas analysis. Preoperatively,parameters for intraoperative hemodynamic variablesshould be established and a blood transfusion triggerdetermined.

Fluid management/blood lossThepotential for blood loss duringmajor spine surgeryis significant, especially for multi-level and/or repeatprocedures. In these cases, red blood cells shouldbe cross-matched and made available for use in theoperating room. Central venous catheterization maybe indicated if massive transfusion is anticipated orif adequate peripheral access cannot be obtained. Incases not involving malignancy or infection a “cellsaver” device may be used to scavenge and rein-fuse the patient’s blood intraoperatively, in an effortto reduce the volume of banked blood transfused.Induced hypotension and hemodilution are strategiesthatmay be incorporated into the anesthetic techniquetominimize blood loss, but the patientsmust be appro-priately selected. For example, an elderly patient withdegenerative disease and chronic hypertension mayrequire higher systemic arterial pressures to ensureadequate spinal cord perfusion and thus is not likely tobe a good candidate for deliberate hypotension. Otherconditions precluding induced hypotension includemyelopathy (in which spinal cord ischemia is ongoing)major cardiovascular disease and renal insufficiency(when hypoperfusion may worsen function). Antifib-

rinolytic agents such as tranexemic acid and alpha-aminocaproic acid may also be used to reduce bloodloss intraoperatively. Finally, as in this case, preopera-tive embolization of a tumor or other vascular lesionmay be used to minimize the likelihood of excessivesurgical blood loss.

Prone positioning can result in edema of the air-way and other dependent tissues. Minimizing intra-venous infusion of crystalloid solution and use of col-loid and blood may offset this problem because of thelatter’s relatively longer intravascular lifespan. Con-cerns about adverse effects on coagulation of hydrox-yethyl starch solution (Hetastarch) limit its utility asa volume expander in the setting of complex spinesurgery.

Perioperative pain controlMany patients with spine disease experience associ-ated chronic pain. Polypharmacy and complex reg-imens are common, emphasizing the importance ofa thorough drug history during preoperative evalua-tion. Nonsteroidal anti-inflammatory drugs may havebeen discontinued preoperatively for surgical reasons,contributing to acutely increased analgesic require-ments. A good understanding of multimodal anal-gesia along with an appreciation for opiate toler-ance helps to ensure appropriate perioperative painmanagement. Alpha-2-adrenergic receptor agonists(clonidine, dexmedetomidine)may be helpful intraop-eratively and postoperatively. Consultation of a painspecialist can be invaluable in complex patients. It isimperative to discuss with patients reasonable expec-tations of pain control after surgery and to exploreoptions such as epidural catheters placed underdirect visualization by the surgeon for postopera-tive patient analgesia, and patient-controlled analgesicpumps [4].

ConclusionComplex spine surgery presents unique challengesfor the anesthesiologist with regard to airway man-agement, positioning, monitoring, fluid management,and pain control. In an effort to achieve a successfuloutcome, preoperative evaluation should be thoroughand consists of careful assessment of the risks associ-ated with patient pathophysiology, anesthetic require-ments, and the surgery itself.

131


References1. E. Ornstein, R. Berko. Anesthesia techniques in

complex spine surgery. Neurosurg Clin N Am 2006; 17:191–203.

2. D. A. Raw, J. K. Beattie, J. M. Hunter. Anaesthesia forspinal surgery in adults. Br J Anaesth 2003; 91:886–904.

3. P. H. Petrozza. Major spine surgery. Anesthesiol ClinNorth America 2002; 20: 405–15, vii.

4. J. P. Cata, E. M. Noguera, E. Parke et al.Patient-controlled epidural analgesia (PCEA) forpostoperative pain control after lumbar spinesurgery. J Neurosurg Anesthesiol 2008; 20: 256–60.

132


41 Loss of evoked potentialsThomas Didier and Ellen Janke

Perioperative neurophysiologic monitoring is a use-ful tool for measuring the functional integrity ofthe central nervous system (CNS) during proceduresthat put this system at risk. Understanding the typesof monitoring that are used – and how to addressperioperative changes – is essential for delivering aneffective and safe anesthetic that allows diagnosis ofischemia/hypoxia before irreversible damage occurssuch that surgical and anesthetic management can beoptimized.

Case descriptionThe patient was a 56-year-old male, American Societyof Anesthesiologists Class III, scheduled to undergoT9–T12 laminectomy and microsurgical correction ofa T10–T12 dural ateriovenous fistula. He presentedwith 6 months of progressively decreased strengthand sensation of his lower extremities. He under-went spinal angiography during which a complexmul-tilevel arteriovenous fistula was visualized that wasnot amenable to embolization. The patient was there-fore scheduled for operative repair. His past medi-cal history was significant for chronic poorly con-trolled hypertension and traumatic brain injury withresultant expressive and receptive aphasia. Due tothe risk of ischemic damage to the spinal cord andnerve roots, neurophysiologic monitoring with a gen-eral anesthetic was planned. Central nervous systemfunction was monitored using somatosensory evokedpotentials (SSEP), electromyography (EMG) and tran-scranial motor evoked potentials (MEP). Followinguneventful induction, anesthesia was maintained with�0.5 minimum alveolar concentration of sevoflurane,a propofol infusion (70 mcg/kg/min) and a remifen-tanil infusion (0.2 mcg/kg/min). An hour and a halfafter surgical incision but before dural opening therewere bilateral decreases in SSEPs of the lower extremi-ties.This was followed by decreasedMEP signals fromthe left foot 15 minutes later. Shortly after the open-

ing of the dura, MEP signals were also lost at theright foot. Vasopressors were administered and thevolatile anesthetic was discontinued, with compen-satory increases in the propofol infusion rate.The tem-perature was confirmed as within normal limits.Therewas no appreciated surgical cause for the loss of sig-nals. The signals did not return to baseline amplitudefor the remainder of the operation. The patient’s post-operative course was significant for profound weak-ness in the lower extremities on postoperative day1. On postoperative day 2 magnetic resonance imag-ing showed abnormal cord signal from T6–T10. Thepatient’s physical exam improved from postoperativedays 3–6 and he was discharged with an examinationunchanged from admission.

DiscussionThe neurophysiologic monitors used in this case weremonitors of CNS function, and can be described asfollows:

Sensory evoked potentialsSensory evoked potentials (SEPs) are measured elec-trophysiologic responses to somatosensory, visual, orauditory stimulation. They are obtained by stimu-lating a sensory system and recording the resultingresponse along a neural pathway. In this way the peri-operative team can monitor the integrity of the sen-sory pathway continuously throughout operations thatplace the pathway at risk.Themost commonly utilizedSEP is somatosensory. Somatosensory evoked poten-tials (SSEPs) represent reproducible electrical activityof cortical and subcortical structures in response toa peripheral nerve stimulus. This provides monitor-ing along the dorsal column–medial lemniscus path-way, which transmits vibration, proprioception, andlight touch. Stimulation is initiated at a peripheralnerve site, commonly the median, ulnar, posterior

133


Figure 41.1. Somatosensory evokedpotential (SSEP) signals. Example of theloss of SSEP signals of a posterior tibialnerve signal. The gray tracing is the initialsignal and the black lines represent thesignals with decreased amplitude andprolonged latency.

tibial, or peroneal nerve. Nerve impulses from thesesites transmit signal to the dorsal horn ganglia at theirrespective root levels.The signal then propagates alongthe dorsal column to themedulla, where fibers synapseand decussate. The signal ascends along the mediallemniscus to the thalamus, the primary somatosensorycortex, and finally association cortices.The integrity ofthe signal can be monitored anywhere along the trans-mitted pathway. Signals monitored subcortically aremore resistant to anesthetic influence [1]. Somatosen-sory evoked potentials are most commonly utilizedfor surgery involving the spinal cord and spinal col-umn, especially in decompressive laminectomies andscoliosis repair. Somatosensory evoked potentials havealso been utilized in carotid vascular surgery, aorticaneurysm repair, intracranial aneurysm repair, andintracranial mass lesion removal. The waveform ofSEPs is described by latency, or time from stimulationto onset of peak of a response, and amplitude, the volt-age of the response. A clinically significant change inSEPs is considered a 50% decrease in amplitude or a10% increase in latency (Figure 41.1) [2].

There are a few important anesthetic considera-tions to understand in order to provide an adequateanesthetic that optimizes SSEP signals. Optimizedbaseline signals should be obtained under anesthe-sia but before surgical intervention occurs in orderto best monitor for nerve damage during surgery. Ifsignals are inadequate the anesthetic plan should betailored to improve them before structures are vul-nerable to surgical injury. Notably, if the patient hasexisting impaired spinal cord function, the baselinesignals may be suboptimal. Both inhaled and intra-venous anesthetics can impact the quality of SSEP sig-nals obtained but intravenous agents have less effectthan equipotent doses of inhaled agents. All inhaledagents increase latency and decrease amplitude of thesignal with the exception of nitrous oxide, the effectof which is variable depending on the other anes-

thetics used. Somatosensory evoked potential signalsare strongest when the total minimum alveolar con-centration of the inhaled anesthetic is �1. Thiopen-tal and propofol cause dose-dependent increases inlatency and decreased amplitude that do not pre-clude intraoperative monitoring. In contrast, etomi-date and ketamine both increase amplitude. Opioidscause an increase in latency and decrease in amplitudebut these effects are not clinically significant. Alpha-2-adrenoreceptor agonists like dexmedetomidine haveminimal effects on SSEPs. When using a balancedanesthetic technique the effects of multiple drugs areusually additive. Paralysis may be helpful in minimiz-ing artifact and should be instituted if consistent withthe other monitoring and surgical goals. In addition tothe anesthetics utilized, hypothermia increases latencyand has unpredictable effects on amplitude that do notpreclude monitoring. Hypotension that extends belowthe autoregulatory threshold decreases amplitude buthas little effect on latency [1].

Motor evoked potentialsDespite the usefulness of SSEPs, one of the shortfalls ofthismonitoringmodality is the inability tomonitor thetracts of the anterior spinal circulation. Motor evokedpotentials allow surveillance of the corticospinal tractand therefore motor function.The primarymotor cor-tex is stimulated eithermagnetically or electrically.Thesignal is then conducted to the brainstem where itdecussates at the level of the pyramids in the medulla.The signal then travels down the corticospinal tractand exits the spinal cord through the ventral nerveroot. After traveling through the peripheral nerves thesignal is transmitted to the muscles causing contrac-tion. There are various ways to measure responses tomotor stimulation but the most common is the “all-or-none” criterion (Figure 41.2). In this method a clin-ically significant event is described as total loss of the

134

Case 41. Loss of evoked potentials

Figure 41.2. Motor evoked potential(MEP) signals. Example of decreasedMEP signals. These were stimulatedtranscranially and monitored at the foot.The top tracings are the initial MEPs. Thebottom portion is the MEP signal afteran acute change. Note the loss of theinflection/deflection points in the middleof the tracing.

MEP signal. Since this method may not detect subtledeficits, other methods are being developed, thoughnone has become standardized.

Prior to the use of MEPs, the only way to checkmotor function during surgery was to wake up thepatient and ask him or her to move extremities. Thisis known as the “wake up test” and is still the goldstandard for checking intraoperative motor function.Motor evoked potentials are usually induced tran-scranially in the intraoperative setting. Electrodes areplaced on the scalp, directly over the motor cortexor directly on the spinal cord and then stimulated,inducing muscular contraction that corresponds tothe area stimulated. Corresponding waveforms arethen observed in the spinal cord, nerves, or mus-cles stimulated. Motor evoked potentials can also bestimulated magnetically, though this is impractical forintraoperative use. Motor stimulation requires muchhigher levels of stimulation than SSEPs and is moni-tored intermittently, unlike SSEPs, which can bemoni-tored continuously. Additionally, MEPs are more tech-nically challenging to obtain compared with SSEPs.Motor evoked potentials are commonly used in spinalcord/column surgery [2].

Motor evoked potentials are extremely sensitiveto inhaled anesthetics. All inhaled agents induce dra-matic decreases inMEP signal. Total intravenous anes-thesia provides conditions that optimize MEP sig-nals. Ketamine, opiates, propofol, and etomidate seemto have the least effect on MEPs. Dexmedetomidine,an alpha-2-adrenoreceptor agonist, causes no signif-icant signal change in clinically used doses. Para-lytic agents may be used but a twitch height of T1in a train-of-four should be maintained at 30% ofthe control height [3]. Since this is clinically difficultif muscles are being monitored, paralytics are oftenavoided. As with SSEPs, it is vitally important tomain-tain a steady anesthetic level during the course of theoperation. Contraindications to using MEPs includepatients with deep brain stimulators and cochlearimplants. Also, it is important that bilateral molar

bite blocks be placed to avoid tongue lacerations orhematomas.

ElectromyographyElectromyography consists of monitoring muscleactivity in response to either spontaneous or activenerve stimulation. Electrodes are placed either on theskin or directly in the muscle. Nerve propagation andtherefore muscle activity is compromised with nervestretching, compression, or pulling.This is manifestedby continuous, repetitive EMG firing, called trains.Additionally, irregularity in the EMG can correlatewith surgical proximity to the nerve and may man-ifest as spikes or bursts. With active nerve stimula-tion a peripheral nerve is stimulated and the mus-cular response is monitored. This is especially usefulfor surgical dissection to distinguish nerves from sur-rounding tissue. Surgeries in which EMG is commonlyemployed include posterior fossa craniotomies, spinalcord/column surgery and peripheral nerve surgery.Anesthetic considerations during EMG are primar-ily related to paralysis. Optimal conditions for EMGare created when no paralytics are used. If paralyticsmust be administered, three or four twitches should bepresent at all times. Additionally, conditions that affectthe neuromuscular junction may limit the usefulnessof EMG. These disorders include myasthenia gravis,botulinum toxin effects and muscular dystrophy.

Intraoperative changesCareful monitoring and quick response to changesin neurophysiologic signals during spine surgery canprevent long-term morbidity. Loss of signals for evenshort periods represents an increased risk of post-operative deficits. Therefore, prompt attention shouldbe given to any change detected. Thorough com-munication among the anesthesiologist, surgeon, andneurophysiologist is essential throughout the opera-tion. Intraoperative changes that can cause a changein neuromonitoring signals include a change in the

135


Table 41.1. Response to the loss of evoked potentials.

Prompt communication with surgeon and neurophysiologist

Confirm that anesthetic is not a factorDecrease depth of anesthesia if necessary

Optimize oxygen deliveryIncrease mean arterial pressureConfirm adequate hemoglobinAdminister 100% FiO2

Rule out hypothermia

Consider small bolus of ketamine or etomidate to increasesignal amplitude

Gold standard – wake up test

anesthetic level, temperature change (hypothermia),decreased oxygen delivery from systemic hypotensionor regional hypoperfusion, electrical noise in the oper-ating room, surgical ligation or ischemia. When thereis an acute change in neuromonitoring signals theanesthesiologist should ensure that anesthetic factorsare not the inciting cause (Table 41.1). First, ensurethat the anesthetic level has been consistent. Thenoptimize oxygen delivery by increasing mean arterialpressure, confirm an adequate hemoglobin level, and

increase inspiredO2 to 100%. Next, rule out hypother-mia. Finally, consider a small bolus of etomidate orketamine to increase signal amplitude. Throughoutthis procedure continuing communication with thesurgeon is of the utmost importance. In the immedi-ate postoperative period a brief neurologic exam com-pleted by the anesthetic team should be documentedin the anesthetic record. In these ways morbidity andmortality in complex spine surgery using neurophysi-ologic monitoring can be reduced.

References1. M. Banoub, J. E. Tetzlaff, A. Schubert. Pharmacologic

and physiologic influences affecting sensory evokedpotentials: implications for perioperative monitoring.Anesthesiology 2003; 99: 716–37.

2. A. A. Gonzalez,D. Jeyanandarajan, C. Hansen et al.Intraoperative neurophysiological monitoring duringspine surgery: a review. Neurosurg Focus 2009; 27: E6.

3. A. C. Wang, K. D.Than, A. B. Etarne et al. Impact ofanesthesia on transcranial motor evoked potentialmonitoring during spine surgery: a review of theliterature. Neurosurg Focus 2009; 27: E7.

136


42 Effects of anesthesia on intraoperativeneurophysiologic monitoringUma Menon and Dileep R. Nair

The primary objective in intraoperative neurophysio-logic monitoring is to identify and prevent the devel-opment of new neurologic deficits or worsening ofa preexisting neurologic injury to a patient who isundergoing surgery. Careful use of anesthetic regi-mens and knowledge of the effects of anesthesia onneurophysiologic modalities are critical for successfuluse of monitoring.

Case descriptionA 24-year-old right-handed female diagnosed withneurofibromatosis type I at age 3 years, presented withprogressive right upper extremity weakness, new onsetleft upper extremity weakness and difficulty walking.She also complained of bilateral upper extremitynumbness and perineal numbness. She was knownto have scoliosis and multiple neurofibromas in thelumbar spine as well as cervical spine with previoussurgical removal and gamma knife treatments. Priormagnetic resonance imaging of the cervical spinedemonstrated bilateral neurofibromas at the level ofC2 with no encroachment of the spinal canal. Alsonoted was a large (�3 cm) dumbbell neurofibromawith no cord compression. A repeat magnetic res-onance imaging of the cervical spine at the time ofcurrent evaluation showed a large extramedullarymass centered within the right C2–3 and C3–4 neuralforamina with extension into the central canal andmoderate to severe mass effect on the cervical cord(Figure 42.1). She was scheduled for laminectomyat C2–3, C3–4 levels and removal of the right-sidedextradural neurofibroma. The agents used for anes-thesia were fentanyl, sevoflurane, nitrous oxide,and remifentanil. Somatosensory evoked potentials(SSEPs) and transcranial motor evoked potentials(MEPs) were recorded intraoperatively.

At baseline the right-sided SSEPs and MEPs weredecreased in comparison to the left. Following theadministration of a fentanyl bolus there was bilateraldecrease in amplitude of all responses. However, due

Figure 42.1. Demonstrates the right-sided extraduralneurofibroma.

to the asymmetry present at baseline, it appeared thatthe anesthetic effect was unilateral. There was no sur-gical explanation for the change noted. Lowering ofthe sevoflurane levels and addition of nitrous oxideand remifentanil infusion resulted in return of theresponses (Figure 42.2).

DiscussionIntraoperative neurophysiologic monitoring is aneffective way of preventing injuries during neurosur-gical procedures. However, the result of the monitor-ing depends on various factors, including temperature,blood flow, intracranial pressure, and especially theanesthetic agents that are used. Typically the effectsof anesthesia and changes in other body parameters

137


Figure 42.2. Shows an apparent asymmetric change in both right-sided motor evoked potentials (first dorsal interosseous muscle,gastrocnemius, and adductor hallucis longus) as well as posterior tibial and median nerve sensory evoked potentials. These findings areexplained by the presence of a baseline decreased amplitude on right-sided responses from the extradural neurofibroma. The apparent lossof responses on the right was associated with the administration of fentanyl bolus producing a bilateral decrease in amplitudes but in theface of the preexisting asymmetry the change appears as a unilateral loss of responses. No clear intraoperative surgical manipulation couldexplain these findings.

on neurophysiologic tests are bilateral unlike the find-ings of this case. The changes produced appeared uni-lateral only because of a preexisting lesion. Bilateralchanges especially in tests used as a control for themonitoring (such as median nerve SEPs in thora-columbar level surgery) should suggest the possibil-ity of effects of anesthetics or other body parameters.However, bilateral changes do not exclude the possi-bility of a significant change from surgery, but unilat-eral changes typically suggest that the effect is not anes-thetic (with the exception highlighted in this case).Different evoked potentials have different sensitivi-ties to anesthesia (visual evoked potentials � SSEPs �brainstem auditory evoked potentials (BAEPs)). The

effect of anesthetic agents on the recordings increaseswith the number of synapses in the pathway beingmonitored. This is related to the effect of anestheticagents on synaptic transmission and axonal conduc-tion. This is why cortical SSEPs are more affected byanesthesia than are the subcortical responses.

The following is a brief overview of the variousanesthetic agents and their effects on neurophysiologicmonitoring [1].

Inhalational agentsUse of halogenated agents causes a dose-related reduc-tion of amplitude and prolonged latency of SEPs

138

Case 42. Effects of anesthesia on intraoperative monitoring

whereas BAEPs are minimally changed, except at highdoses when they are lost. Transcranial electrical MEPs(particularly single pulse) on the other hand can becompletely abolished with doses of 0.5 minimum alve-olar concentration (MAC). Nitrous oxide at equipo-tent doses to the halogenated agents can also cause adose-related reduction in amplitude and prolongationof latency of cortical SEPs. Additionally, the effects ofnitrous oxide are also dependent on the other agentsused concurrently.

Intravenous anesthetic agentsIntravenous anesthetic agents include barbiturates,benzodiazepines, etomidate, propofol, and ketamine.In general barbiturates do not produce much changeof SEPs. Thiopental, which is commonly used forinduction, can produce transient reduction in ampli-tude and prolonged latency of cortical SEPs. However,MEPs can be completely abolished with barbituratesand are to be avoided if such monitoring is required.With single doses of midazolam, a commonly usedbenzodiazepine, there is minimal effect on evokedpotentials, but with continuous infusions it can reducethe latency and amplitude of SEPs.TheMEP responses,however, are significantly reduced with the use ofmidazolam. Etomidate is a unique anesthetic agent inthat it causes enhancement of SEP andMEP responses.It may be an agent to consider in a difficult setting, butwith awareness of the idiosyncratic reaction of adrenalsuppression and increased mortality with prolongedinfusion. Opioids used for total intravenous anes-thesia include fentanyl, remifentanil, sufentanil, andalfentanil. These intravenous opioid agents typicallyhave little effect on evoked potentials even at very highdoses making them quite useful during intraoperativemonitoring. The current case represents an excep-tion. Ketamine is another agent with some uniqueproperties in that it can enhance the cortical evokedpotential responses. Dexmedetomidine is a selectivealpha-2-adrenoreceptor agonist, which is used as anintravenous adjunct to any of the general anestheticsand which does not cause a significant change in SEPor MEP recordings up to doses of 0.6 ng/mL [2].

Neuromuscular blocking agentsUse of neuromuscular blocking agents can affect theability to record myogenic responses from a varietyof monitoring modalities such as transcranial MEPs,free-running electromyography, direct nerve stimula-tion evoked motor responses, pedicle screw stimula-tion, and cortical stimulation motor mapping. On theother hand these agents can improve the recordings ofSEPs andMEPs by reducing noise produced bymuscleactivity.

ConclusionAlmost all of the anesthetic agents can cause depres-sion of the evoked potentials if given at sufficientlyhigh doses and therefore a suitable combination ofanesthetic agents should be chosen in discussion withthe surgeon, anesthetist, and the monitoring team.Consideration should be given that the agent(s) cho-sen have the least effect on the evoked potentialswith the least possible stable dose without causing anypatient discomfort. A team approach with good com-munication between the neurophysiologist, anesthe-siologist, and surgeon is very important to the suc-cess of neurophysiologic monitoring and its clinicalutility [3, 4].

References1. A. C. Wang, K. D.Than, A. B. Etame et al. Impact of

anesthesia on transcranial electric motor evokedpotential monitoring during spine surgery: a reviewof the literature. Neurosurg Focus 2009; 27: E7.

2. E. Bala,D. I. Sessler,D. R. Nair et al. Motor andsomatosensory evoked potentials are well maintainedin patients given dexmedetomidine during spinesurgery. Anesthesiology 2008; 109: 417–25.

3. M. L. James. Anesthetic considerations. In Husain A.M., ed. A Practical Approach to NeurophysiologicMonitoring. 1st edn. New York: Demos MedicalPublishing, 2008.

4. T. B. Sloan, E. J. Heyer. Anesthesia for intraoperativemonitoring of the spinal cord. J Clin Neurophys 2002;19: 430–43.

139


43 Neurofibromatosis type 1 andspinal deformityGeorge A. Mashour

Neurofibromatosis type 1 (NF1) is one of the mostcommon genetic disorders of the nervous system,affecting approximately 1 in 3000 individuals. Themost typical manifestations are multiple neurofibro-mas and melanogenic abnormalities (both of neuralcrest origin), but the phenotypemay also include skele-tal abnormalities.

Case descriptionThe patient was a 39-year-old female with a historyof NF1 and four previous spine surgeries, the last ofwhich occurred at the age of 16. She presented witha history of progressive difficulty walking, as well asnumbness, tingling, burning, and spasm in her lowerextremities. Imaging revealed a kyphotic deformityand severe angulation at the junction of the cervicaland thoracic spine (Figure 43.1); she was also diag-nosed with a pseudomeningocele. The patient wasadmitted for correction of the deformity and her cer-vical spine was stabilized with a halo device. Elec-trophysiologic monitoring with somatosensory andmotor evoked potentials was planned.

The primary concerns of the anesthesiology teamwere (1) management of the airway in the halo,(2) the potential for significant blood loss given thehistory of multiple spine surgeries, and (3) anestheticmaintenance consistent with effective neurophysio-logic monitoring. In preparation for an awake fiberop-tic intubation, glycopyrrolate was administered as anantisialogogue; subsequently, the patient’s heart rateincreased to the 160s. Tachycardia was treated withesmolol and a dexmedetomidine infusion was startedfor both sedation and sympatholysis. Nebulized andatomized lidocaine was administered for airway topi-calization, with additional boluses of fentanyl for seda-tion. During the awake fiberoptic intubation, systolicblood pressures rose to �200 mmHg. After confirm-ing endotracheal intubation, a central venous catheterwas placed in the femoral vein and balanced anesthe-

Figure 43.1. Spinal deformity associated with neurofibromatosistype 1 (NF1). Note the marked angulation of the spine at thecervicothoracic junction.

sia was maintained with dexmedetomidine, sufentanilinfusion, and isoflurane at 0.5 minimum alveolar con-centration. The case was completed uneventfully withapproximately 1 liter estimated blood loss; emergencewas smooth and the patient was neurologically intact.

DiscussionNeurofibromatosis type 1 (also known as von Reck-linghausen’s disease) is an autosomal dominant dis-order with nearly complete penetrance and highlyvariable expressivity. It results from a mutation ofthe NF1 gene on chromosome 17, which encodes thetumor suppressor neurofibromin. Although in thiscase the patient’smother had a diagnosis of NF1, spon-taneous mutation of the gene without family historyoccurs in approximately 50% of patients. Melanogenicabnormalities such as cafe-au-lait macules, freckling,

140

Case 43. Neurofibromatosis type 1 and spinal deformity

and hyperpigmentation occur earlier in life, whiletumor formation usually manifests in puberty. Neu-rofibromatosis type 2 is a clinically and genetically dis-tinct entity. Neurofibromatosis type 2 ismore rare thanNF1, with an incidence of approximately 1 in 40 000.Neurofibromatosis type 2 also has an autosomal dom-inant mode of transmission and is characterized byvestibular schwannomas (often referred to as acousticneuromas), as well as meningiomas.

Patients with NF1 may present for surgical pro-cedures involving peripheral nervous system tumors(neurofibromas, malignant peripheral nerve sheathtumors), central nervous system tumors (benign opticgliomas, astrocytomas), scoliosis or other skeletalabnormalities, and a host of other disorders. Becausethe phenotype is so highly variable, a thorough andsystematic approach to the patient with NF1 is essen-tial (Table 43.1) [1].

The airway was of particular concern in thispatient. In addition to the obvious difficulties pre-sented by cervical spine immobilization, neurofibro-mas may also develop in the trachea and respiratorytree [2, 3]. Patients with tracheobronchial tumors maybe asymptomatic for many years andmay present withnormal chest radiographs. Recent history of dyspnea,cough, dysphagia, dysarthria, stridor, or change ofvoice should be especially concerning in a patient withNF1. The need for an awake fiberoptic endotrachealintubation because of the halo, in conjunction withthe rare but real possibility of airway tumors, led us tochoose dexmedetomidine as a mode of sedation thatwould help preserve respiratory drive. Furthermore,dexmedetomidine has also been demonstrated to be auseful adjunct for spine surgery that does not signifi-cantly interfere with neurophysiologic recording [4].

The marked increases in heart rate and bloodpressure during the preparation and performanceof fiberoptic intubation were initially concerning.The vagotonic effects of fentanyl, alpha-2 agonistsympatholytic properties of dexmedetomidine, andwhat appeared to be adequate airway topicalizationwith lidocaine did not attenuate the hemodynamicresponse. Approximately 6% of patients withNF1 havehypertension, but there is also a well-known associa-tion with pheochromocytomas (another neural crestderivative). Refractory hypertension in a patient withNF1 should therefore prompt consideration and diag-nostic evaluation of pheochromocytoma. In the cur-rent patient, the blood pressure stabilized after the air-way was secured.

Table 43.1. Anesthetic considerations of neurofibromatosistype 1 (NF1).

System Comments

Airway Neurofibromas of the tongue, pharynx,or larynx may interfere with trachealintubation

Suspicion is raised by history ofdysphagia, dysarthria, stridor, orchange of voice

Respiratory system Intrapulmonary neurofibroma,pulmonary fibrosis may producecough and dyspnea

Right ventricular failure may be presentScoliosis or kyphosis may compromise

lung function

Cardiovascularsystem

Raised arterial pressure usually essentialhypertension but considerpheochromocytoma or renal arterystenosis

Hypertrophic cardiomyopathy mayoccur

Mediastinal tumors may result insuperior vena caval obstruction

Nervous system A variety of peripheral nerve tumors andintracranial tumors are common

Increased incidence of epilepsy andlearning disorders

Cerebrovascular disease may co-existHydrocephalusCognitive impairment

Gastrointestinaltract

Intestinal tumors may present with pain,gastrointestinal hemorrhage orperforation

Carcinoid tumors occur in duodenumand may result in jaundice andcarcinoid syndrome

Genitourinarysystem

Neurofibromas may causeureteric/urethral obstruction

Musculoskeletalsystem

Vertebral deformities or spinal cordtumors may make spinal/extraduraltechniques difficult

Pseudoarthroses

Modified with permission from Hirsch et al., 2001 [1].

ConclusionIn conclusion, NF1 patients are complex and presentfor a variety of neurosurgical interventions. Given themultisystem involvement and variable phenotype, asystematic approach to theNF1 patient is necessary forsafe perioperative care.

References1. N. P. Hirsch, A. Murphy, J. J. Radcliffe.

Neurofibromatosis: clinical presentations and

141


anaesthetic implications. Br J Anaesth 2001; 86:555–64.

2. K. L. Irion, T. D. Gasparetto, E. Marchiori et al.Neurofibromatosis type 1 with tracheobronchialneurofibromas: case report with emphasis ontomographic findings. J Thorac Imaging 2008; 23:194–6.

3. S. S. Moorthy, S. Radpour, E. C. Weisberger.Anesthetic management of a patient with trachealneurofibroma. J Clin Anesth 2005; 17: 290–2.

4. E. Bala,D. I. Sessler,D. R. Nair et al.Motor andsomatosensory evoked potentials are well maintainedin patients given dexmedetomidine during spinesurgery. Anesthesiology 2008; 109: 417–25.

142


44 Major vascular complication duringspine surgeryZeyd Ebrahim and Jia Lin

The incidence of vascular injury during major spinesurgery is low. However, when this catastrophe occurs,as outlined in this case, it presents a major challenge tothe anesthesia team. Anesthesiologists involved in thecare of patients undergoing complicated spine surgeryshould be cognitive of this infrequent but serious com-plication [1, 2].

Case descriptionA 75-year-old female was scheduled for removal ofinstrumentation at L4–S1 and re-exploration of a pre-vious posterior lumbar inter-body fusion. Her cur-rent symptoms included lower extremityweakness andpain. She had a history of well-controlled hyperten-sion, hyperlipidemia, anxiety, depression, tobacco use(quit 25 years prior), morbid obesity (bodymass index36.1), lumbar stenosis and spondylolisthesis, and wasstatus post arthrodesis and segmental instrumentation(about 6months prior). Her functional class was at thelevel of FC III.

Therewas no significant abnormality seen on phys-ical examination. The laboratory values revealed nor-mal basic metabolic panel, complete blood count, andcoagulation tests; her electrocardiogram and chest X-ray were unremarkable. Magnetic resonance imag-ing showed lumbar stenosis and spondylolisthesis.Stress echocardiogram was negative for myocardialischemia, revealed normal left and right heart functionand normal valve structures.

The patient was taken to the operating room witha peripheral intravenous catheter in situ. StandardASA monitors were applied. Anesthesia was inducedwith fentanyl, propofol, and rocuronium. After induc-tion, an arterial line and an additional 16G peripheralcatheter were placed.The patient was positioned proneon the Jackson table with all pressure points paddedand the abdomen hanging unimpeded. Maintenanceof anesthesia was achieved with 1.0 minimum alveo-lar concentration (MAC) of isoflurane, and 40% oxy-

gen air mixture and IV continuous infusion of sufen-tanil. Rocuroniumwas administered as needed follow-ing neuromuscular blockade monitoring.

The first arterial blood gas after the start of the pro-cedure revealed a pH of 7.44; PaCO2: 30 (end-tidalCO2 (ETCO2) between 26 and 30); PaO2: 210 (FiO2: 0.43); HCO3: 20; Base excess:−3 and Hct: 34.The first 5hours of the procedurewere uneventful and the patientwas stable. The surgeon confirmed for the anesthesiateam that the old hardware had been removed and thathe was performing an extended dissection and evacu-ating the disk space.

A sudden drop in blood pressure, oxygen satura-tion, bispectral index and ETCO2 were noticed. Thearterial line wave form was almost flat. It was notedthat the blood pressure dropped from 100s/60smmHgto systolic pressure of 30s; ETCO2 dropped to�10 mmHg.

The breath sounds were clear, bilateral, and equal.The ventilator did not show any change in the peakairway pressure of 27 cmH2O and the arterial linewas flushed and aspirated to ensure patency. However,upon inquiry the surgeon confirmed excessive bloodloss.

After the initial evaluation and examination, theinspiratory FiO2 was changed to 1.0; phenylephrine1000mcg in divided doseswas given. Fluid blouses andblood were transfused but with little effect on the lowblood pressure. Help was called to the operating roomand epinephrine 1 mg was given which increased theblood pressure to 100s/60s and heart rate increased to120s. However, the effect was temporary.

While the patient was being resuscitated, the sur-geon quickly closed the incision and the patient wasturned to the supine position. A left subclavian centralaccess was quickly established and a vascular surgeonwas called for an evaluation of an intra-abdominalvascular injury. A large amount of bright red bloodin the abdomen was noted. The aorta was then crossclamped at the supraceliac level. Further exploration

143


Figure 44.1. Iatrogenic injury to iliac artery during L5–S1 lumbar spine surgery. Reprinted with permission. Illustration Copyright C© 2010Nucleus Medical Art, All rights reserved. www.nucleusinc.com.

of the abdomen revealed a severely calcified aorta anda posterior wall injury at the very proximal left com-mon iliac artery just at the level of aortic bifurcation.The aorta was reconstructed with a bifurcated graft.

The intraoperative course was also complicated bysignificant coagulopathy from massive blood loss andtransfusion. Total estimated blood loss was approxi-mately 10 liters. Intraoperative partial thromboplas-tin time (PTT) was 150.8 and the platelet count was29K.

The patient was transferred to the intensive careunit (ICU) with stable vital signs, reasonable urineoutput, and improving coagulopathy. The initial lab-oratory values in the ICU were: PTT: 39.4, Interna-tionalNormalized ratio (INR): 1.4, fibrinogen 104, anda platelet count of 194. Coagulopathy was further cor-rected overnight.

The postoperative course was complicated bynonoliguric renal failure, pneumonia, and urinarytract infection. Following an ICU stay and prolongedhospital course, the patient was discharged to a reha-bilitation facility on postoperative day 29.

DiscussionIatrogenic major vascular injury during lumbar spineoperations is a rare and potentially fatal complication.The incidence is about 1–5/10 000; mortality is about38–65% in cases of arterial laceration. Risk of injurycan be explained by the close anatomical relation-ship between the retroperitoneal vessels and the verte-bral column (Figure 44.1). Vessels that can be injuredinclude the left common iliac artery and vein, abdom-inal aorta and superior rectal artery.

Variable clinical manifestation is one of the mainreasons for the apparent low incidence of vascularinjury, which is thought to be underestimated. Thediagnosis can be delayed or overlooked if the acutehemorrhage is retroperitoneal. With formation of apseudoaneurysm or arteriovenous fistula, diagnosiscan be delayed for weeks or even years.

Therefore, a high degree of suspicion of vascularinjury should always be maintained with unexplainedhypotension. In addition, in these cases there is no cor-relation between the real blood loss and that viewed inthe surgical field. As always, communication between

144

Case 44. Major vascular complication during spine surgery

the anesthesia and surgical team is extremely impor-tant. Also, this case highlights the importance of hav-ing the blood products available before starting anycomplex spine surgery [3, 4, 5].

The role of central venous monitoring is alwaysdebated in the context of major spine surgery. How-ever, central venous pressure readings in the proneposition may not reflect accurate data and large boreintravascular access and invasive blood pressure mon-itoring are probably more important in the hemody-namic management of these cases.

ConclusionIn conclusion, anesthesia for complex spine surgeryrequires invasive monitoring, large-bore intravenousaccess, and awareness of the potential for disaster.

References1. E. F. Reilly, N. S. Weger, S. P. Stawicki. Vascular

injury during spinal surgery. OPUS 12 Scientist 2008;2: 7–10.

2. B. Duz,M. Kaplan, C. Gunay et al. Iliocavalarteriovenous fistula following lumbar disc surgery:endovascular treatment with a stent-graft. TurkNeurosurg 2008; 18: 245–8.

3. U. Nilsonne, A. Hakelius. On vascular injury inlumbar disc surgery. Acta Orthop Scandinav 1965; 35:329–37.

4. W. R. Smythe, J. P. Carpenter. Upper abdominalaortic injury during spinal surgery. J Vasc Surg 1997;25: 774–7.

5. J. Inamasu, B. H. Guiot. Vascular injury andcomplication in neurosurgical spine surgery. ActaNeurochir 2006; 148: 375–87.

145


45 Complex spine surgery fora Jehovah’s WitnessMatthew Martin and Vijay Tarnal

Jehovah’s Witness (JW) is an evangelical Christiandenomination best known to physicians for beliefsregarding the refusal of all blood product transfusions.Based on their interpretation of the bible, Jehovah’sWitnesses believe that acceptance of any blood once ithas left the body is “in violation of Gods Law” [1].Thisis a challenge from both a clinical and ethical stand-point during complex spine surgeries.

Case descriptionThe patient was a 54-year-old female with a history ofobstructive sleep apnea, hyperlipidemia, and scoliosiswho presented for T10–L5 posterior spinal fusion. Shepreviously had undergone a posterior lumbar fusion,but recently experienced progressive back and hip painas well as new-onset difficulty with walking. Imagingof the patient’s thoracolumbar spine showed signifi-cant progression of her scoliosis, cephalad to her previ-ous fusion (Figure 45.1).The procedure was scheduledfor 7.5 hours with somatosensory and motor evokedpotentials. A detailed preoperative discussion focusedon the anestheticmanagement, the potential for signif-icant blood loss, and the patient’s wishes should pro-found anemia occur.

The primary concerns for the anesthesiologist were(1) the potential for significant intraoperative bloodloss in a Jehovah’s Witness, especially given the sizeand duration of the procedure, and (2) a balancedanesthetic technique in the setting of somatosen-sory and motor evoked potential monitoring. Afteruneventful induction and placement of an endotra-cheal tube, dexmedetomidine and sufentanil infusionswere started. Using a SAFESETTM reservoir monitor-ing kit, a radial arterial catheter was placed for hemo-dynamic monitoring and arterial blood gas analysis.Since the patient had poor venous access, a right inter-nal jugular central venous catheter was placed. Thepatient was placed in a Mayfield headrest and posi-tioned prone, with appropriate precautions to prevent

Figure 45.1. Scoliosis of thoracolumbar spine, cephalad to thepatient’s previous fusion.

pressure on the inferior vena cava. Balanced anesthesiawas maintained using dexmedetomidine, sufentanil,and a propofol infusion; systolic blood pressure wasmaintained within 20% of the patient’s baseline. Thepatient’s preoperative hematocrit was 42%. Based onthe clinical status and blood loss, arterial blood gasanalyses were done using the SAFESETTM reservoir,which meant no additional blood loss due to frequentsampling. The arterial blood gas analysis was used notonly to assess respiratory and metabolic state but alsothe hematocrit.

The antifibrinolytic drug aminocaproic acid wasinfused during the case to reduce blood loss. The casewas completed without complications with an esti-mated blood loss of 2200 mL. At the completion of thecase, the patient’s final intraoperative hematocrit was27%. After a smooth emergence and tracheal extuba-tion, the patient was transferred to the recovery roomhemodynamically stable and neurologically intact.

146

Case 45. Complex spine surgery for a Jehovah’s Witness

DiscussionThe Jehovah’s Witness faith was founded in the latenineteenth century out of the controversy stemmingfrom the interpretation of the bible by Catholic andProtestant denominations. It was not until 1945 that itwas determined that blood transfusion should be for-bidden because it violated God’s law. Complex spineprocedures on the Jehovah’s Witness patient pose asignificant challenge for the anesthesiologist. Bloodconservation must be planned for and discussed withthe patient and surgical team preoperatively. Severalstrategies have been described in the literature in thelast 50 years including: proper patient positioning,autologous blood donation and transfusion, normo-volemic hemodilution, artificial blood products, andblood salvage [2]. Other options include pharmaco-logic therapies such as desmopressin, aminocaproicacid, and recombinant factor VII. It must be stressedthat different Jehovah’s Witnesses may follow differ-ent beliefs – it is therefore crucial to discuss and doc-ument the patient’s wishes thoroughly prior to anyprocedure.

One mainstay of spine surgery is proper patientpositioning. By leaving the abdomen free in the proneposition, there is decreased inferior vena cava pressureand thus decreased venous plexus filling around thespinal cord. There is also a theoretical benefit ofdecreased vertebral venous pressure and reducedintraoperative blood loss from decreased bonebleeding [3].

Preoperative autologous donation utilizes thepatient’s own blood, which is usually harvested priorto the operation, allowing the patient to intrinsicallyreplace the lost blood. This has been utilized forprocedures with high potential for blood loss, but hasshown only modest benefits. The technique is alsoassociated with many of the same risks of allogenicblood transfusion, including infection. Some Jehovah’sWitnesses will accept this technique, but most will notaccept any blood (even their own) once it has left thebody.

Acute normovolemic hemodilution utilizes thesame principles as above, but only shows very modest(if any) benefit in blood conservation. This techniquerequires the anesthesiologist to remove the patient’sblood and replace the volume with crystalloid or col-loid. The concept is that the hemoglobin content ofthe “lost blood” would be less if the blood were dilute.The removed whole blood would then be re-infused

at the completion of the case. Again, most Jehovah’sWitnesses would oppose this treatment, as their bloodwould leave the body.

Perioperative blood salvage employs strategies oftaking the patient’s own blood and re-infusing it afterwashing and purifying. This “cell-saving” techniquehas been used in some Jehovah’s Witnesses as the cir-cuit can be maintained in-line with the patient’s circu-lation. The cell-salvage technique is now widely usedfor procedures in which significant blood loss is antici-pated.The risks, although low, include infection, coag-ulopathy, and air embolism.The benefits of cell salvagefar outweigh the risks.

Another option is the use of pharmacologic ther-apy to reduce blood loss. Desmopressin can be used,which is thought to increase serum levels of vonWille-brand factor, which forms a complex with Factor VIIIand increases coagulation. While the potential forreduction in blood loss during spine surgery exists,there have been conflicting results throughout the lit-erature. Aminocaproic acid is a lysine analog thatworks as an antifibrinolytic and has been shown toreduce blood loss during complex spine procedures.With the removal of Aprotinin from the market in2007, aminocaproic acid has been widely used formany cardiac and complex spine procedures.

Erythropoietin is a blood-stimulating hormonenormally synthesized in the kidney that stimulates redblood cell production in the bone marrow. It is usuallyaccepted by Jehovah’s Witness patients and has beenshown useful in blood conservation [4]. This ther-apy is usually started weeks to months in anticipationof major surgery. Elevated hemoglobin levels can beadvantageous preoperatively, but increased cardiovas-cular complications have been described in patientsusing erythropoietin with hemoglobin levels greaterthan 13 g/dL.

Finally, the use of “hypotensive anesthesia” toreduce blood loss can be an effective blood conserva-tion strategy in the Jehovah’s Witness population.Thishas been described for many years as a way to reduceblood loss by up to 40% during complex spine pro-cedures [5]. Although this technique has been widelyaccepted as an effective technique for reduction inblood loss, there are potential risks associated withhypotension in the prone position. One potential com-plication, which has seen a recent surge of attention,is postoperative visual loss. The etiology can be theresult of ischemic optic neuropathy, of which both ane-mia and hypotension may be contributing factors [6].

147


The risk is significant enough such that the Ameri-can Society of Anesthesiologists recommend that anypatient undergoing a “lengthy” spine surgery in theprone position should be warned about this complica-tion. One guideline is to keep the systolic blood pres-suremaintained within 20% of the patient’s baseline. Itis prudent to exclude patients from induced hypoten-sive anesthesia if they have significant cardiovascularco-morbidities such as hypertension, cerebrovasculardisease, and coronary artery disease.

ConclusionIn conclusion, while complex spine procedures in theJehovah’sWitness patient may be challenging, success-ful outcomes can be achieved. With active preopera-tive assessment, communication, blood conservationstrategies, proper positioning and meticulous surgicaland anesthetic technique, complex spine surgeries arepossible in the Jehovah’s Witness population.

References1. TheWatchtower, September 15, 1961, p. 558.2. T. R. Kuklo, B. D. Owens,D.W. Polly Jr. Perioperative

blood and blood product management for spinaldeformity surgery. Spine J 2003; 3: 388–93.

3. T. C. Lee, L. C. Yang,H. J. Chen. Effect of patientposition and hypotensive anesthesia on inferior venacaval pressure. Spine 1998; 23: 941–7.

4. T. K. Rosengart, R. E. Helm, J. Klemperer et al.Combined aprotinin and erythropoietin use for bloodconservation: results with Jehovah’s Witnesses. AnnThorac Surg 1994; 58: 1397–403.

5. J. W. Brodsky, J. H. Dickson,W. D. Erwin et al.Hypotensive anesthesia for scoliosis surgery inJehovah’s Witnesses. Spine 1991; 16: 304–6.

6. L. A. Lee, S. Roth, K. L. Posner et al.The AmericanSociety of Anesthesiologists Postoperative Visual LossRegistry: analysis of 93 spine surgery cases withpostoperative visual loss. Anesthesiology 2006; 105:652–9.

148


46 Diplopia following spine surgeryAlaa A. Abd-Elsayed and Ehab Farag

Visual disturbances are known to occur after spinesurgery. Although postoperative visual loss will be dis-cussed in the next case, here we describe a case of post-operative diplopia.

Case descriptionA34-year-oldwomanpresented to the clinicwith post-laminectomy kyphosis and subsequently underwentposterior fusion and fixation in the prone position.Thepatient received 7000 mL of crystalloids and 1465 mLof colloids during surgery.The patient developed facialedema and reported diplopia on the first postopera-tive day. Examination and history revealed no obviouscause for the developed diplopia, which resolved com-pletely after 2 days without treatment.

DiscussionAbducens nerve palsy, which causes diplopia withoutany other neurologic signs, is reported to be the mostcommon cranial nerve palsy [1]. Abducens nerve palsywas reported to be a complication of different proce-dures such as lumbar puncture, shunt surgery, max-illary osteotomy, cranial trauma, and skull traction[2–6]. The sixth, ninth, tenth, and twelfth cranialnerves have a vertical or oblique course in the craniummaking them the most vulnerable nerves to stretchinjuries with the abducens nerve being themost at riskof injury [2]. Barsoum et al. reported a case of appar-ent abducens nerve palsy after only 5 lb of traction atlumbar spinal surgery [2]. The cause of the palsy wasthought to be stretching of the cranial nerve by skulltraction.

Abducens nerve palsy has also been recognized asa complication of lumbar puncture [7]; the reason for

this was attributed to caudal displacement and neu-rotoxic spinal arachnoiditis. All the causes of diplopiathat were reported in the literature were investigated inour case but none seemed to apply. We speculate thatthe diplopia in our patientwas caused by fluid overloadduring surgery, as the patient received enough volumeto cause facial edema and thus possibly abducens nervestretch.

ConclusionProne spine cases are associated with postoperativevisual disturbances. The present case suggests thatdiplopia in this context may be a self-limiting processdue to sixth nerve stretch after fluid overload.

References1. C.Wilkins,G. D.MacEwen. Cranial nerve injury from

halo traction. Clin Orthop Relat Res 1977; 126: 106–10.2. W. K. Barsoum, J. Mayerson, G. R. Bell. Cranial nerve

palsy as a complication of operative traction. Spine1999; 24: 585–6.

3. M. J. Botte, T. P. Byrne, R. A. Abrams et al.Haloskeletal fixation: techniques of application andprevention of complications. J Am Acad Orthop Surg1996; 4: 44–53.

4. A. R. Hodgson. Halo-pelvic traction in scoliosis. Isr JMed Sci 1973; 9: 767–70.

5. K. K. Jain. Aberrant roots of the abducent nerve.J Neurosurg 1964; 21: 349–51.

6. E. A. Miller, P. J. Savino, N. J. Schatz. Bilateralsixth-nerve palsy. A rare complication of water-solublecontrast myelography. Arch Ophthal 1982; 100: 603–4.

7. G. L. Clifton, E. R. Miller, S. C. Choi et al. Lack ofeffect of induction of hypothermia after acute braininjury. N Engl J Med 2001; 344: 556–63.

149


47 Postoperative visual loss in spine patientsGoodarz Golmirzaie and Laurel E. Moore

Postoperative visual loss (POVL) is a rare but catas-trophic complication of spine surgery. The extremelylow incidence has made its study and prevention achallenge for neuroanesthesiologists.

Case descriptionThe patient was a 62-year-old female who presentedfor a revision L4–5 foraminotomy and L4–S1 trans-verse lumbar interbody fusion. Her past medical his-tory was significant for obesity (body mass index =39), longstanding hypertension, depression, and sys-temic lupus erythematosus. The patient had a his-tory of recurrent deep vein thrombosis and pulmonaryembolus, thought to be secondary to a hypercoagula-ble state related to lupus.The patient’s activity was lim-ited secondary to chronic low back pain. Her preoper-ative laboratory values were normal with a hematocritof 40%; vital signs included a baseline blood pressureof 140/85.

The 6-hour surgery was uneventful. Systolic bloodpressure was kept in the range of 90–95 mmHg for�75%of the procedure.The lowest systolic blood pres-sure recorded was 65. The estimated blood loss was1500 mL and the hematocrit nadir was 25%. Fluidsincluded 1000 mL of 5% albumin and 3000 mL of lac-tated Ringer’s solution. Urine output for the case was450 mL. She was recovered in the postanesthesia careunit and subsequently transferred to the neurosurgi-cal intensive care unit for routine observation. Thatevening she reported that she could not see out of herleft eye.

DiscussionTo understand POVL it is necessary to understandthe anatomy of the optic nerve and its vascular sup-ply (Figure 47.1). The optic nerve can be dividedinto four portions, including the intracranial (opticchiasm to the optic canal within the lesser sphe-noid wing), the intracanalicular (within the optic

canal), the posterior or intraorbital (optic foramento the lamina cribrosa) and the anterior or intraocu-lar (from the lamina cribrosa to the optic disk). Thelamina cribrosa is a perforated membrane overlyingthe posterior scleral foramen through which the opticnerve, as well as central retinal artery and vein, enterthe eye.

The retina and optic nerve are supplied by branchesof the ophthalmic artery [1]. Once the ophthalmicartery passes through the optic foramen it branchesinto several vessels including the central retinal arteryand a series of posterior ciliary arteries. The intraor-bital optic nerve is supplied by a pial plexus whichin turn is supplied by branches of the central retinalartery. The most anterior portion of the optic nerveis supplied primarily by short posterior ciliary arter-ies. The posterior aspect of the optic nerve is suppliedby end-vessels that are comprised of easily compressedcentripetal pial vessels, thus placing this region of theoptic nerve at particular risk for ischemia.

Mechanism of injuryThere are multiple causes of POVL, including corti-cal infarction, direct injuries to the eye and ischemicinjuries to the retina and optic nerve. The mostcommon permanent injuries are ischemic in natureincluding central retinal artery occlusion (CRAO)and ischemic optic neuropathy (ION). Ischemic opticneuropathy can be further subdivided into poste-rior ischemic optic neuropathy (PION) and anteriorischemic optic neuropathy (AION). While CRAO andAION have been historically associated with cardiacsurgery, PION has more recently attracted attentionin the medical literature and lay press because of itsperceived increase in frequency following complexspine surgery [2]. However, the current literature sug-gests that the incidence for spine cases is dropping.According to Shen et al. the incidence of POVL afterspine cases is approximately 3/10 000, which is lower

150

Case 47. Postoperative visual loss in spine patients

Circle ofZinn-Haller

Short PosteriorCillary A.

SubarachnoidSpace

Pial Branchesfrom Int. Carotid A.

PerpendicuiarBranches of Pial A.

Hypophyseal Branchesof Int. Carotid A.

Dura

Sciera

Central Retinal A.

Short PosteriorCillary A.

Optic Disc

Ophthalmic A.

Optic CanalArea of P.I.O.N.

Area of A.I.O.N.

Figure 47.1. Anatomy and vascular supply of the optic nerve. Reproduced with permission from Neurosurg Focus 2007; 23 (5); E15.

than previously published incidences of approximately2/1000 [3, 4].

Central retinal artery occlusionCentral retinal artery occlusion generally presentswith painless monocular visual loss following emer-gence from anesthesia. In cardiac surgery, the cause ofCRAO is generally thought to be embolic [1]. When itoccurs during spine surgery, it ismost commonly asso-ciated with direct compression of the globe, thus theterm “headrest syndrome” [4]. Intraocular pressurehas been demonstrated to increase in prone patients[4]. If this increase in pressure is further augmentedby direct compression of the globe by improper headposition, the intraocular pressure may exceed cen-tral retinal artery pressure and retinal ischemia mayresult. In the American Society of Anesthesiologists(ASA) visual loss registry, 10/93 cases of POVL fol-lowing spine surgery were thought to be secondary toCRAO [4]. This subgroup of patients was less likelyto have been positioned in a Mayfield head holderand 9/10 had signs of injury to structures around theeye [4]. Significant visual recovery is rare followingCRAO [4].

Ischemic optic neuropathy (anteriorand posterior)Anterior ischemic optic neuropathy is generally asso-ciatedwith cardiac surgery [1]. In the last several years,PION has become increasingly recognized after com-

plex surgical spine procedures [3, 5]. Whether theoptic nerve injury is anterior or posterior to the lam-ina cribrosa, patients present with painless binocularor (less commonly) monocular visual loss. The sever-ity of the visual loss can range from visual field deficitsto complete loss of light perception. Recognition of theproblem postoperatively can be immediate or delayedby several days. Unlike CRAO there is no evidence ofexternal eye injury.There is usually a reduced or absentpupillary light reflex [2]. Anterior ischemic optic neu-ropathy and PION can be differentiated based on fun-doscopic examination. The initial fundoscopic exam-ination in AION shows an edematous disc, whereaswith PION the initial fundoscopic exam is generallynormal and subsequently deteriorates [1].

Intuitively the etiology of ION would appear to berelated to pre-existing vascular disease and reducedoxygen delivery to the optic nerve. However, the eti-ology is probably more complex than this. There arereports of ION occurring in patients with normal per-fusion pressures and normal hematocrits [4]. Therehave also been reports of ION occurring in children[3, 4]. This would suggest that there is a certain pop-ulation of patients that is at greater risk of developingION, either based on the vascular supply, the anatom-ical structure of the eye, or deficiencies in autoregula-tion of blood flow to the optic nerve. Supporting this,patients with ION often have bilateral disease [4]. Inthe ASA registry, 66% of patients had bilateral visualloss [4]. In most patients with POVL secondary toION, there was no evidence of injury to other vascularbeds (heart, liver, etc.), suggesting that the intraorbital

151


optic nerve may be particularly vulnerable to ischemicinjury.

Anatomically there appear to be patients identi-fied as having a “disc at risk” for AION.These patientshave an unusually small disc and small opening inBruch’s membrane, resulting in crowding of the opticnerve fibers as they pass through this restricted space[6]. In the setting of ischemia this creates a viciouscycle in which early edema of the optic nerve pro-duces greater compression of the nerve and its vas-cular supply, which in turn further reduces oxygendelivery [6].

Similarly, another possible mechanism for PIONmight be a form of orbital “compartment syndrome.”The intraorbital optic nerve is enclosed by dura andthus has little room for expansion. Should there beswelling of the posterior optic nerve related to increas-ing venous pressure in the prone position or from largevolumes of crystalloid, the edema could result in com-pression of sensitive vascular structures.

Finally, there are several case reports of sildenafilandAION in adultmales with erectile dysfunction andin children treated for pulmonary hypertension [4].

Risk factors for postoperative visual lossBecause of the rare incidence of POVL, prospectivestudies evaluating risk factors are limited. There aremany seemingly intuitive risk factors for POVL thatare described in the literature including prone posi-tioning, prolonged procedures, hypotension, anemia,and the need for blood transfusion [4]. Premorbid sys-temic diseases such as diabetes, hypertension, periph-eral vascular occlusive disease, and hypercoagulablestates have also been correlated with POVL [4, 5]. Ina population-based study by Patil et al., it was demon-strated that hypotension, anemia, and peripheral vas-cular disease were potential risk factors for POVLwithodds ratios of 10, 4.3, and 6.4 respectively [5]. Of thesemany potential risk factors, however, two in particularare consistently supported: prolonged surgery (morethan 6 hours) and large blood loss (more than 1 liter)[4]. In the ASA POVL registry, 94% of spine patientshad surgeries that were longer than 6 hours and 82%had operations that had an estimated blood loss of oneliter or more [4].

Prevention of postoperative visual lossSince the etiology of POVL remains poorly under-stood, prevention remains difficult. However, there are

some clinical precautions that may reduce the possi-bility of POVL based on current evidence. The firstis ensuring that the eye is free of pressure while thepatient is prone. Careful initial positioning with subse-quent eye checks (and documentation) every 15 min-utes should be routine for all prone cases. Further-more, whenever possible the head (and thus orbit)should be positioned at or above the level of the heartto optimize venous drainage and improve perfusionpressure to the retina and optic nerve.

In addition to these precautions, the following arerecommendations summarized in the ASA practiceadvisory [7] developed in 2005 by a task force com-prised of anesthesiologists, spine surgeons, and neuro-opthalmologists.These do not constitute practice stan-dards but are recommendations developed after anextensive literature review and survey of practicingphysicians.

1. Anemia, atherosclerotic disease, and obesity maybe associated with POVL, but at present theycannot be considered predisposing factors.

2. Factors that classify patients as “high risk” includeprocedures that are prolonged (greater than6.5 hours), associated with large blood loss, andperformed in the prone position.

3. The use of an arterial line is recommended forpatients with chronic hypertension, but the taskforce could not make recommendations withregards to induced hypotension based on theevidence they had at the time.

4. There was no consensus on a “minimal acceptablehemoglobin.”

5. A combination of crystalloid and colloid isrecommended to replace significant blood loss.

6. While neuroanesthesiologists as a group felt thatthe extensive use of alpha-agonists decreasedoxygen delivery to the optic nerve, no consensuswas achieved on this point.

7. Staging long procedures was one strategydiscussed.

8. No proven treatment exists for POVL but thepanel agreed that anemia should be treated,hypotension corrected, and oxygen administeredto patients with suspected POVL. All patientsshould have an ophthalmologic consultation.

9. Finally, practitioners should consider includingdiscussion of POVL in the informed consentprocess.

152

Case 47. Postoperative visual loss in spine patients

Management of postoperative visual lossOur patient with POVL had two major risk factors:she had an estimated blood loss of 1.5 liters and shehad a prolonged surgery. Furthermore, she was rela-tively hypotensive for prolonged periods. In this par-ticular case the patient recovered most of her visionprior to being discharged from the hospital. In generalonly 50% of patients with ION have significant visualrecovery [4].

Unfortunately there is no proven treatment forPOVL. In the setting of acute POVL the correctionof anemia and hypotension should be considered. Anophthalmologic consultation may be helpful in deter-mining the cause of POVL and a formal fundoscopicexamination should be performed and documented. Itis reasonable to obtain magnetic resonance imaging torule out an unusual etiology for the blindness such ascortical infarction or pituitary apoplexy. Administra-tion of acetazolamide (to reduce intraocular pressure)and CO2 inhalation (to increased blood flow to theretina and optic nerve) have not been shown to changeoutcome in POVL.

ConclusionIn conclusion, POVL is a rare complication with anapproximate incidence of 3/10 000 for spine proce-dures and 9/10 000 to 5/100 for cardiac surgeries [3, 4].When it does occur it is a devastating event forthe patient. While the incidence of CRAO may bedecreased with careful attention to the orbit, ION isless clear in its etiology and thus more difficult to pre-vent. Many risk factors have been proposed yet ourunderstanding of the etiology of ION remains inade-

quate. Clearly there are both patient and perioperativefactors involved. Until we have a better understand-ing of these risk factors, careful attention to the eyes,staged procedures, vigilance with regard to intraocu-lar pressure (venous drainage) and the optimizationof oxygen-carrying capacity are the best preventativemeasures available.

References1. E. L. Williams,W.M. Hart, R. Tempelhoff.

Postoperative ischemic optic neuropathy. Anesth Analg1995; 80: 1018–29.

2. L. M. Buono, R. Foroozan. Perioperative posteriorischemic optic neuropathy: review of the literature.Surv Ophthalmol 2005; 50: 15–26.

3. Y. Shen,M. Drum, S. Roth. The prevalence ofperioperative visual loss in the United States: a 10-yearstudy from 1996 to 2005 of spinal, orthopedic, cardiac,and general surgery. Anesth Analg 2009; 109: 1534–45.

4. L. A. Lee, S. Roth, K. Posner et al. The AmericanSociety of Anesthesiologists Postoperative Visual LossRegistry: analysis of 93 spine surgery cases withpostoperative visual loss. Anesthesiology 2006; 105:652–9.

5. C. G. Patil, E. M. Lad, S. P. Lad et al. Visual loss afterspine surgery: a population-based study. Spine 2008;33: 1491–6.

6. S. S. Hayreh. Ischemic optic neuropathy. Prog RetinEye Res 2009; 28: 34–62.

7. American Society of Anesthesiologists Task Forceon Perioperative Blindness. Practice advisory forperioperative visual loss associated with spine surgery:a report by the American Society of AnesthesiologistsTask Force on Perioperative Blindness. Anesthesiology2006; 104: 1319–28.

153


48 Prone cardiopulmonary resuscitationJustin Upp and Sheron Beltran

Neurosurgical procedures are very rarely performed ina straightforward supine position. As such, there aresignificant challenges associatedwith themanagementof cardiopulmonary arrests that are not found in othersurgeries.

Case descriptionA 69-year-old female with a history of renal cell car-cinoma developed new back pain and radiculopa-thy. Magnetic resonance imaging of the lumbar spinerevealed a pathologic lumbar compression fracturewith tumor encroachment upon the neural foramenbilaterally.Thepatient presented to the operating roomfor a complex anteroposterior corpectomy 24 hoursafter embolization of the mass. Induction of gen-eral endotracheal anesthesia with thiopental and suc-cinylcholine was uneventful. Maintenance anesthesiaconsisted of a low concentration of sevoflurane withremifentanil and propofol infusions to facilitate motorand somatosensory evoked potentials. Adequate intra-venous access and an arterial line were obtained beforethe patient was positioned prone for the posterior por-tion of the case. This initial stage of the procedurewas complicated by 1900 mL of blood loss for whichthe patient was appropriately resuscitated. After com-pletion, the patient was placed in a modified lateralposition for retroperitoneal exposure of the anteriorlumbar spine. This portion of the case was compli-cated by aortic injury and brisk blood loss. A vascu-lar surgical team was summoned to repair the aorta,but despite aggressive resuscitation with crystalloid,albumin, fresh frozen plasma, platelets, and 29 units ofpacked red blood cells, the patient developed pulselesselectrical activity (PEA) arrest. Resuscitation effortscontinuedwhile thewoundwas packed and the patientwas repositioned supine to facilitate external cardiaccompressions. The wound continued to bleed duringthe unsuccessful resuscitation effort.

DiscussionThis case was one of two at our institution in whichexposure to the surgical site of bleeding was poorlyaccessible due to the need to perform cardiopul-monary resuscitation (CPR) in the supine position [1].Although this case was performed in the lateral posi-tion, it sparked an ongoing dialog and exploration intothe possibility of prone CPR during complex spineprocedures.

In the operative setting, cardiac arrest necessitatingCPR is an event that fortunately occurs infrequently.When it does occur, however, patients undergoingneurosurgery may be in a variety of surgical positionswhich are suboptimal for the performance of adequatechest compressions. Repositioning a prone patient tothe supine position often requires additional person-nel, equipment (bed or stretcher) andmaymake surgi-cal exposure to a bleeding vessel inaccessible to repair.Additionally, it requires time to reposition, which iscrucial during an arrest. Throughout the 1960s therewas a significant amount of research into cardiopul-monary support and resuscitation eventually lead-ing to the adoption of CPR by the American HeartAssociation (AHA), and culminating in Basic LifeSupport (BLS) and Advanced Cardiac Life Support(ACLS).

There has been strong evidence supporting the useof CPR in the supine position, however, a paucity ofevidence exists regarding its application in other posi-tions. Publications by McNeil in 1989 were the firstdescriptions that effective CPR could be conducted inthe prone position [2]. His proposal of prone CPRwas originally written to address commonly encoun-tered issues with CPR outside the hospital, includ-ing a hesitancy of bystanders to administer mouth-to-mouth ventilation as well as aspiration risks. Althoughthese risks may not be applicable in the operative the-ater during spine surgery with controlled ventilationand a secure airway, the concept of CPR in the prone

154

Case 48. Prone cardiopulmonary resuscitation

position does offer an alternative to conventionalCPR.

The obvious goal of closed chest compressions,whether supine or prone, is forward blood flow fromthe heart and thoracic cavity into the systemic circula-tion.However, themechanismbywhich this successfulblood flow occurs is less obvious. There are two pro-posed physiologic models of CPR: the cardiac pumpmechanism and the thoracic pump mechanism [2].

The cardiac pump mechanism is the traditionalmodel explaining blood flow during CPR. Directexternal compression of the sternum causes directmechanical compression of the heart. This increasesintraventricular pressure and, when combined withatrioventricular valve closure, leads to an artificialcardiac output.These findings have recently been con-firmed through transesophageal echocardiography byHigano [3]. During CPR there was echocardiographicevidence of ventricular compression, atrioventricularvalve closure, and regurgitation through the mitralvalve all indicating a greater increase in ventricularpressure over atrial pressure. After compressionceases, rebound of the thorax transmits negativepressure resulting in venous return, atrioventricularvalve opening, and diastolic coronary filling.

The thoracic pump model differs from the cardiacpump mechanism in that direct compression of thecardiac chambers is not thought to be necessary togenerate forward blood flow. The thorax as a wholeacts as a pump when external compression increasesintrathoracic pressure. The more rigid aorta and arte-rial structures resist collapse, but are compressed caus-ing increased arterial pressure. Venous structures ofthe thorax collapse under thoracic pressure prevent-ing retrograde blood flow. It is also believed that retro-grade blood flow is prevented by valves in the subcla-vian and jugular veins. After compression of the thoraxis released, negative intrathoracic pressure promotesvenous return similar to that described in the cardiacpump model [4].

There is no consensus on the optimal way to per-form prone CPR, as the AHA only recommends proneCPR when the supine position is inaccessible. ProneCPR is not currently taught in BLS or ACLS courses.Prone CPR has been described in multiple ways, butthe evidence mostly consists of case reports and othersmall nonrandomized studies. The method of con-ducting compressions varies according to several fac-tors including patient’s age, number of rescuers per-forming CPR, presence and/or location of surgical

incision, the type of bed or operating table that thepatient rests on, and the presence of a definitive airway.

In infants, prone CPR can be performed much thesame way as supine CPR, with both hands wrappedaround the thorax. Rather than bilateral thumbs com-pressing the sternum, the rescuer would compressthe thoracic spine. Compressions can also be accom-plished by compressing of the thoracic spine with thefingers of one hand while providing counter pressurewith the other. For larger children and adults thereare also multiple ways of performing compressions.Compressions have been performed with and withoutcounter pressure. Counter pressure can be accom-plished by using another rescuer to support the ster-num, or a firm surface with either a saline or sand-bag beneath the patient’s sternum. Compressions havebeendescribed to be effective in themidline two-thirdsup the patient’s spine between the scapulae. Compres-sions can also be performed at the same level with bothhands placed laterally if there is a mid-line incision.

Despite prone CPR first being described in the lit-erature over 30 years ago, there still remains a paucityof evidence supporting its efficacy. There is no pub-lished trial to date addressing prone CPR outcomes,however, a few small trials exist supporting the physi-ologic feasibility of prone CPR. Mazer et al. conducteda study on six patients in the intensive care unit (ICU)after they suffered circulatory arrest and failed ACLSafter 30 minutes. Patients underwent supine CPR for15 minutes followed by 15 minutes of prone CPR.Themeans of the systolic, diastolic, andmean arterial pres-sures were recorded during the 30 minutes of supineand prone CPR. After the patients were enrolled inthe trial no other ACLS drugs were given other than1 mg of intravenous epinephrine every 3 minutes. Toconduct prone CPR, patients in this trial were placedon a CPR board with counter pressure provided by asandbag placed under the sternum. The rescuers per-formed compressions in themidline fromT7–T10 ver-tebral bodies at a rate of 60–100 compressions/minute.During proneCPR they found a statistically significantincrease in mean systolic blood pressure as well as cal-culatedmean arterial pressure.Meandiastolic pressurewas also increased, however, this finding was not sta-tistically significant [5].

Wei et al. conducted a similar study on 11 ICUpatients. After failing ACLS and expiring, both proneand supine CPR were performed for 1 minute each.Systolic and diastolic blood pressures were recordedby invasive measurement. Prone CPR was found to

155


yield higher mean systolic and diastolic blood pres-sures than those obtained by standard supineCPRwithresults reaching statistical significance [6].

ConclusionIn conclusion, current BLS and ACLS guidelines makerecommendations with the purpose of achieving effec-tive resuscitation over a broad variety of locations andscenarios. These recommendations do not currentlyinclude prone CPR. The operating room differs fromother resuscitative settings, as it is a more controlledenvironment. The patient will already have a defini-tive airway and intravenous access established, therebyeliminating potentially the largest drawbacks of proneCPR: the hindrance of airway and intravenous catheteracquisition.

During resuscitation it is well established that earlydefibrillation is associated with better outcomes, andeffective and continuous CPR aids in successful defib-rillation. In this respect prone CPR may have a theo-retical advantage over supine CPR in prone surgeries,as the time from initiation of ACLS to external car-diac compression may be reduced. However, it mustbe mentioned that no data exist on prone defibrilla-tion success rate or improved outcomes. Current datado support the hemodynamic effectiveness of prone

CPR and, possibly, a more significant finding was thatnone of the hemodynamic parameters were less thanthe averages during supine CPR [5, 6]. Intraoperativescenarios in which the patient is in the prone position,as in cases of spinal surgery, are unique settings forwhich prone CPRmay be well-suited as a resuscitationtechnique.

References1. S. L. Beltran, G. A. Mashour. Unsuccessful

cardiopulmonary resuscitation during neurosurgicalprocedures: is the supine position always optimal?Anesthesiology 2008; 108: 163–4.

2. E. L. McNeil. Re-evaluation of cardiopulmonaryresuscitation. Resuscitation 1989; 18: 1–5.

3. S. T. Higano, J. K. Oh,G. A. Ewy et al.Themechanismof blood flow during closed chest cardiac massage inhumans: transesophageal echocardiographicobservations.Mayo Clin Proc 1990; 65: 1432–40.

4. C. F. Babbs. New versus old theories of blood flowduring CPR. Crit Care Med 1980; 8: 191–5.

5. S. P. Mazer,M.Weisfeldt, D et al. Reverse CPR: Apilot study of CPR in the prone position. Resuscitation2003; 57: 279–85.

6. J. Wei,D. Tung, S. H. Sue et al. Cardiopulmonaryresuscitation in prone position: a simplified methodfor outpatients. J Chin Med Assoc 2006; 69: 202–6.

156


49 Open spine stabilization withpolymethylmethacrylate augmentationMariel Manlapaz and Greta Jo

Open spine stabilizationwith polymethylmethacrylate(PMMA) augmentation procedures requires signifi-cant attention during anesthetic management due tothe complication of PMMA embolization. We presenta case of hemodynamic instability due to embolizationduring surgery as well as its management.

Case descriptionA 54-year-old male with a T12 burst fracture pre-sented for a second stage posterior instrumentation ofT9–L4. His past medical history included a T12 cor-pectomy and T11–L1 instrumentation by a transtho-racic approach, osteoporosis, deep vein thrombosis,and mild left ventricular dysfunction.

After induction, the trachea was intubated with a7.5 mm endotracheal tube and maintained on isoflu-rane and sufentanil infusion. An arterial line, centralline and 14-gauge intravenous catheter were placed.Two hours after incision, the patient was found to havefrequent premature ventricular contractions associ-ated with a drop in blood pressure. At that time,the surgeons were augmenting T9, L3, and L4 withPMMA. The premature ventricular contractions werenot sustained and the blood pressure responded to100mcg boluses of phenylephrine. Meanwhile, severalblood samples were sent and the fluids were opened.Differential diagnoses considered included transienteffects of PMMA, bone/marrow embolism, cementmonomer toxicity, and anaphylactoid reaction.

Fifteen minutes later, the patient developed sus-tained ventricular tachycardia. The blood pressuredropped along with his oxygen saturation and end-tidal CO2. The patient was moved from the proneto the supine position and given epinephrine, amio-darone, and vasopressin. He was then maintained onnorepinephrine, epinephrine, and amiodarone infu-sions. Myocardial ischemia, pulmonary embolism(PE) from deep venous thrombosis or PMMA, andanaphylactoid/anaphylaxis reaction were considered.

An intraoperative transesophageal echocardiogramwas performed, showing global hypokinesis of theright and left ventricle, patent foramen ovale, andthrombus in the pulmonary artery.

Vascular medicine recommended a heparin bolusand infusion, cardiology agreed with the diagnosis ofPE, and cardiothoracic surgery recommended that thepatient proceed with an embolectomy. Repeat TEE inthe cardiothoracic operating room did not show a PEand the plans for embolectomy were aborted. Fur-ther work-up included a cardiac catheterization show-ing non-stenotic coronary vessels, moderate right andleft ventricular dysfunction, no PE in the main pul-monary artery or first and second branches. An infe-rior vena cava filter was then placed. A computedtomography scan of the chest showed high attenuationembolized material, presumably PMMA in the supe-rior and inferior vena cavae, right cardiac chambers,and the right pulmonary artery. A repeat transthoracicechocardiogram showed a 2 cm right ventricular massabove the tricuspid valve and 3+ tricuspid regurgi-tation. Because the patient had persistent respiratoryproblems, the decisionwasmade for surgical interven-tion: pulmonary embolectomy, closure of the patentforamen ovale, and tricuspid valve repair. Upon open-ing the cardiac chambers, PMMA was found travers-ing the right ventricle, right atrium and pulmonaryartery (Figures 49.1–49.3). At a follow-up visit, thepatient was back to his baseline cardiac and mentalstatus.

DiscussionThis patient had osteoporosis, a systemic skele-tal disease characterized by low bone mass andmicroarchitectural deterioration of bone tissue, with aconsequent increase in bone fragility and susceptibil-ity to fractures. Such people are susceptible to painfulvertebral body compression fractures. Treatmentoptions include medical management, open surgical

157


Figure 49.1. Intracardiac polymethylmethacrylate in the rightatrium, right ventricle, and pulmonary artery.

stabilization, and percutaneous vertebral augmenta-tion using vertebroplasty or kyphoplasty.

Polymethylmethacrylate is used in vertebroplastyand kyphoplasty and can be used in open surgical sta-bilization. Vertebroplasty restores stability through apercutaneous bone cement injection via high-pressurefilling. Kyphoplasty is another percutaneous proce-dure, restoring vertebral body height. Kyphoplastyinvolves low-pressure filling by first producing a voidwith a balloon and then creating an internal cast. Afundamental difference between the two is that in ver-tebroplasty, filling is done until the cement leaks, mak-ing the volume and pattern unpredictable. For kypho-plasty, filling is done until the cavity is full, making thevolume and pattern more predictable [1, 2].

While there are multiple reports in the litera-ture regarding cement leaks with vertebroplasty andkyphoplasty, reported cases of cement leakage dur-ing open surgical stabilization are rare. The rate ofcement leaks during percutaneous intervention is high

Figure 49.2. Intracardiac polymethylmethacrylate in the rightatrium, right ventricle, and pulmonary artery.

and is seen in up to 73% of vertebral bodies treated,with venous leaks reported in up to 24% of verte-bral bodies treated. With leaking comes the risk ofcement embolism. In a study by Choe et al. [1], cementPE was seen in 4.6% of 65 patients who had percu-taneous vertebroplasty or kyphoplasty [1]. Causes ofcement embolism include insufficient polymerizationof PMMAat the time of injection, poor needle positionwith respect to the basivertebral vein, or overfilling ofthe vertebral body.This allows migration of the embo-lus into the inferior vena cava and the venous system.Anticoagulation prevents thrombus formation on theemboli, but it does not reduce right ventricular after-load nor improve the resulting ventilation/perfusiondefect. Generally, 4–11 mL of cement per level is rec-ommended. Less volume should be given for multi-ple levels. For this patient, 15 mL of PMMA was used[3, 4].

An option to avoid such complications is to con-currently use a venogram to visualize venous outflow

Figure 49.3. Length ofpolymethylmethacrylate embolus.

158

Case 49. Open spine stabilization with polymethylmethacrylate

from the vertebra at the time of injection. A venogramcan be performed through a bone needle and requiresinjection of a small volume of dilute, low osmo-lar nonionic contrast agent. Furthermore, biplanefluoroscopy during PMMA application can improvevisualization and the surgeon can stop injection ifcement exudes [5].

ConclusionThe incidence of cement leakage is high with percu-taneous vertebral augmentation. The case presentedhere is PMMA venous leakage and PE during opensurgical stabilization. Cardiopulmonary presentationis either immediate or delayed and can be catastrophic.Consider a chest X-ray, echocardiogram (transtho-racic or transesophageal), and computed tomog-raphy scan as diagnostic tools. If strongly symp-tomatic, consult cardiothoracic surgery for possibleembolectomy.

References1. D. H. Choe, E. M. Marom, K. Ahrar et al. Pulmonary

embolism of polymethyl methacrylate duringpercutaneous vertebroplasty and kyphoplasty. Am JRoentgenol 2004; 183: 1097–102.

2. D. R. Fourney,D. F. Schomer, R. Nader et al.Percutaneous vertebroplasty and kyphoplasty forpainful vertebral body fractures in cancer patients.J Neurosurgery 2003; 98: 21–30.

3. J. Hodler,D. Peck, L. A. Gilula et al.Midtermoutcome after vertebroplasty: predictive value oftechnical and patient-related factors. Radiology 2003;227: 662–8.

4. C. Vasconcelos, P. Gailloud, N. J. Beauchamp et al.Is percutaneous vertebroplasty without pretreatmentvenography safe? Evaluation of 205 consecutivesprocedures. AJNR 2002; 23: 913–17.

5. F. M. Phillips, F. ToddWetzel, I. Lieberman et al. Anin vivo comparison of the potential for extravertebralcement leak after vertebroplasty and kyphoplasty.Spine 2002; 27: 2173–8.

159


50 Extubating the trachea after prolongedprone surgeryVijay Tarnal and Robert Scott Kriss

Respiratory complications after tracheal extubationare three times more common than complicationsoccurring during tracheal intubation [1]. Airway dif-ficulties after complex spine surgeries in prolongedprone position can cause catastrophic complicationsincluding severe hypoxia and death.This case will dis-cuss the importance of an appropriate extubation strat-egy in light of the known postoperative complicationsof prolonged prone positioning.

Case descriptionThe patient was a 22-year-old male with severe scolio-sis scheduled for a T3–ilium fusion with vertebral col-umn resections. The case was scheduled for 9 hours.He presented with a history of cerebral palsy compli-cated by significant cognitive deficits, seizures, and dif-fuse spasticity. His spinal curvature was so severe andpainful that he was unable to sit in his wheelchair andwas required to lie flat at all times. Imaging revealed asevere levoconvex rotary scoliosis of the thoracolum-bar spine, approximately 103 degrees from T3–L4,and associated rib cage deformity. The patient hadan uncomplicated left hip surgery 5 years previously.Electrophysiologic monitoring with somatosensoryand motor evoked potentials was planned.

Given the severity of the scoliosis, the primary con-cerns of the anesthesiology team were (1) difficulty inpositioning supine for laryngoscopy and intubation,(2) respiratory difficulties, (3) difficult prone position-ing, (4) the potential for significant blood loss giventhe complexity of the planned procedure, (5) a bal-anced anesthetic technique consistent with effectiveneurophysiologic monitoring, and (6) difficulty withextubation following the prolonged prone surgery.For preoxygenation and induction of anesthesia, thepatientwas positioned on the operating table in a semi-upright position and pressure areas protected withfoam pads. With loss of consciousness, the patient was

mask ventilated while the operating table was repo-sitioned for laryngoscopy and intubation. Followingan uncomplicated tracheal intubation, balanced anes-thetic technique using dexmedetomidine, sufentanil,and desflurane at 0.4 minimum alveolar concentra-tion (MAC) was maintained. The case was completedin 11 hours without significant surgical intraoperativecomplications. The estimated blood loss was approx-imately 28% of the patient’s total blood volume andhe required four units of packed red blood cells inorder to achieve a target hematocrit between 24 and27%. The sufentanil infusion was stopped 90 minutesbefore the case was completed. The patient was posi-tioned supine and a “cuff-leak test” was done, whichwas positive for leak around the endotracheal tube.Because of the long duration of the procedure in proneposition involving significant blood loss, as well aspreoperative mental status, a delayed extubation wasplanned. The patient was transported to the recoveryroom and started on pressure support ventilation.Thetrachea was subsequently extubated over a Cook Air-way Exchange Catheter 2 hours after arrival in thepostanesthesia care unit. The patient tolerated the air-way exchange catheter, andmaintainedhis ownairway,with a SaO2 �94% on 8 L/min of humidified oxygen.The repeat arterial blood gas analysis following extuba-tion was within normal limits. Analgesia was adminis-tered using intravenous morphine, titrated to effect.

DiscussionProne positioning during anesthesia is required to pro-vide operative access for a wide variety of surgical pro-cedures. This position is not only associated with pre-dictable changes in physiology but also a number ofcomplications [2].The airway complications andman-agement strategies are discussed in this chapter. For apatient to be positioned prone there is a potential riskof airway loss and for this reason, a definitive airway

160

Case 50. Extubating the trachea after prolonged prone surgery

is established using an endotracheal tube. The com-plications that result from this position are typicallydue to application of either direct or indirect pressureto dependent parts of the body resulting in significantinjuries to the upper airway.

Problems with extubation1. Prone position can result in extensive facial

edema, macroglossia, including the swelling ofsoft palate, pharynx, and arytenoids. Excessiveflexion of the head and endotracheal tubeobstructs the lingual and pharyngeal venousdrainage. This can result in postoperative upperairway edema. Long operative time, blood loss,and administration of large volumes of crystalloidcan make the facial and upper airway edemaprogressively worse [2, 3].

2. Although rare, the prone position can result instretching and compression of the salivary ductsleading to painful swelling of the submandibularglands especially if the head is rotated [4]. Thisdistorts the anatomy and the swelling of the softtissue of the floor including the tongue can lead topartial or complete obstruction of the pharynx.

3. Tracheal extubation can also lead to an increase insystemic and intracranial pressures. Coughing,laryngospasm, and upper airway obstruction canoccur after extubation.

Tracheal re-intubation poses significant challengescompared with the initial, elective intubations. Thepatients are more likely to be hypoxic, hypercap-nic, and combative. Furthermore, oral secretions andvomit can make the glottic view more difficult. Tomanage such difficult airway scenarios, the anesthe-siologist should have strategies for alternative airwaytechniques.TheAmerican Society ofAnesthesiologistsTask Force onManagement of theDifficult Airway rec-ommends that the anesthesiologist should have a pre-formulated strategy for extubation of a difficult air-way [5]. The preformulated extubation strategy willdepend, in part, on the surgery, the condition of thepatient, and the skills and preferences of the anesthesi-ologist. A safe extubation strategy should aim to min-imize risks and discomfort, maintain adequate oxy-genation, ventilation and if the need arises, re-intubatewithout any complications. In complex spine surgeries,the main advantage of immediate recovery and extu-

bation is to allow earlier neurologic examination andestablishment of a baseline to guide further exami-nation. This also comes with early extubation risksof hypoxemia, hypercarbia, and compromised respira-tory and hemodynamic status.

Airway assessment prior to extubationNumerous attempts have been made to recognize andassess tracheal and oropharyngeal edema prior toextubation. The leak test and visual inspection of air-way swelling are the most common risk assessmenttests for extubation.1. The “cuff-leak test” is performed by deflating the

tracheal tube balloon, occluding the tracheal tubeand assessing the air movement around theendotracheal tube [6]. Although not specific forpredicting stridor, it appears that no or low leakvolumes of air are reliably associated with anincreased risk of upper airway obstruction. Insuch situations a decision to delay the extubationis prudent.

2. Flexible fiberoptic endoscopic exam should beused to directly assess edema and other causes ofupper airway obstruction if any uncertainty existswith a cuff-leak test.

3. The spirometer on modern anesthesia machinescan also be used to quantify leakage. It iscalculated by the difference in the inspired andexpired volumes.

Strategies for difficult extubation1. Tracheal tube exchange catheter. If there is any

appreciable risk of post-extubation airwayobstruction, a tracheal tube exchange cathetershould be used. In the event of airway obstruction,re-intubation can be achieved by introducing anendotracheal tube over the exchange catheter.

2. Laryngeal mask airway, flexible fiberopticbronchoscope and Aintree catheter. The laryngealmask airway has also been used with limitedsuccess. A flexible bronchoscope using thelaryngeal mask airway and Aintree intubationcatheter can also be used. There is an increasedrisk of bleeding due to upper airway edemamaking the technique difficult and challenging.

3. Deferred extubation in the recovery room orintensive care unit should be planned in the event

161


Table 50.1. Criteria for extubation following complex spinesurgery in prone position.

Defer extubation Consider extubation

Inability to open eyes and notobeying commands

Awake and obeyingcommands

Agitated or combative Regular spontaneousbreathing

Poor respiratory efforts

O2 saturation �94% on highflow O2

O2 saturation �94% on highflow O2

Hypercarbic (PaCO2 �50 mmHg)

Normocapnic (PaCO2 �30 mmHg �50 mmHg)

Hemodynamically unstable Hemodynamically stable

Hypothermic (�36 ◦C) Normothermic

Neuromuscular blockadecompletely reversed (TOF �90%, sustained head lift andstrong hand grip)

Operating time �10 hours Operating time �10 hours

Blood transfusion �4 units Blood transfusion of �4 units

Evidence of facial edema andmacroglossia

Negative “cuff-leak test” Positive “cuff-leak test”

Evidence of pharyngeal andlaryngeal edema on flexiblefiberoptic bronchoscopy

No evidence of pharyngeal andlaryngeal edema on flexiblefiberoptic bronchoscopy

that the patient does not meet extubation criterialisted in Table 50.1. Extubation should beperformed in a location where a full range ofequipment, appropriate personnel, and all

preparations (including surgical airwayequipment) are available.

ConclusionIn conclusion, given the numerous complicationsunique to patients undergoing complex spine surgeriesin the prone position, a systematic approach to extuba-tion should begin as early as possible to optimize safeperioperative care.

References1. T. Asai, K. Koga, R. S. Vaughan. Respiratory

complications associated with tracheal intubation andextubation. Br J Anaesth 1998; 80: 767–75.

2. H. Edgcombe, K. Carter, S. Yarrow. Anaesthesia inthe prone position. Br J Anaesth 2008; 100: 165–83.

3. B. Kwon, J. U. Yoo, C. G. Furey et al. Risk factors fordelayed extubation after single-stage, multi-levelanterior cervical decompression and posterior fusion.J Spinal Disord Tech 2006; 19: 389–93.

4. P. Hans, J. Demoitie, L. Collignon et al. Acutebilateral submandibular swelling following surgery inprone position. Eur J Anaesthesiol 2006; 23: 83–4.

5. American Society of Anesthesiologists Task Forceon Management of the Difficult Airway. Practiceguidelines for management of the difficult airway:an updated report by the American Society ofAnesthesiologists Task Force on Management of theDifficult Airway. Anesthesiology 2003; 98: 1269–77.

6. R. J. Adderley, G. C. Mullins. When to extubate thecroup patient: the “leak” test. Can J Anaesth 1987; 34:304–6.

162


51 Perioperative peripheral nerve injuryLisa Grilly, George A. Mashour and Chad M. Brummett

Neurosurgical procedures are often performed withpatients in the prone, lateral, and other non-supinepositions. This creates the risk of perioperative neuro-logic deficit due to peripheral nerve injury (PNI).

Case descriptionA55-year-old female presented for scoliosis correctionwith posterior instrumentation at T5–L5.The case wasperformed under general anesthesia with neurophys-iologic monitoring of somatosensory evoked poten-tials, motor evoked potentials, and electromyogra-phy. After induction and endotracheal intubation, thepatient was positioned prone on a Jackson frame withher head in aMayfield holder. Bilateral upper extremi-ties were padded and positioned above the head. Base-line monitoring was established and anesthetic main-tenance consisted of 50% nitrous oxide and isoflurane(minimum alveolar concentration 0.3–0.4), as well assufentanil and dexmedetomidine infusions. Duringthe surgery, there were episodes of diminished motorevoked potentials of the right upper extremity, whichreversed with repositioning and additional padding ofthe arm. During periods of acute blood loss there wasalso concern about hypotension affecting themonitor-ing, which was discussed among the surgery, anesthe-sia, and neurophysiology monitoring teams. The caselasted a total of 9 hours and estimated blood loss was2 liters. The patient received one unit of packed redblood cells and one unit of fresh frozen plasma. At theend of the case the tracheawas extubatedwithout com-plication. Right upper extremity weakness was foundimmediately after emergence.The initial postoperativeneurologic examination showed 2/5 right hand grip,2/5 right wrist extension, 1/5 right biceps, 1/5 right tri-ceps, and 3/5 right deltoid strength. The patient waspresumed to have a right brachial plexus injury andwas started on dexamethasone. Neurology and physi-cal therapy were consulted after admission to the neu-rosurgical intensive care unit.The patient’s right upper

extremity strength gradually improved to a maximumof 4/5 throughout by postoperative day 8, and she wasdischarged that day to an acute rehabilitation center.

DiscussionPerioperative peripheral nerve injury can be a dev-astating complication, resulting in motor dysfunctionand the sequelae of lowermotor neuron disease. Nerveinjuries comprise approximately 15% of claims in themost recent American Society of Anesthesiologistsclosed claims analysis [1], a proportion unchangedfrom a prior study [2]. The lack of improvement overapproximately a 10-year period indicates the need fora better understanding of both the etiology and pre-vention of PNI. Mechanisms for peripheral neuropa-thy include compression, stretch, ischemia, metabolicabnormalities, and direct trauma. The “double crush”hypothesis of Upton and McComas suggests that thenerve is subject to two insults that lead to neuropathy,the first insult rendering it less tolerant of the second[3]. In the surgicalmilieu, this would imply that preop-erative co-morbidities might make peripheral nervesmore vulnerable to perioperative adverse events thatmight otherwise be tolerated.

A recent study by Welch et al. demonstrated thatdiabetes mellitus, hypertension, and tobacco use werepreexisting patient factors that were associated withPNI [4]. All three of these risk factors may compro-mise microvascular circulation, rendering peripheralnerves more susceptible to modes of perioperativeinjury associated with ischemia. General and epidu-ral anesthesia were also found to be associated withPNI, in contrast to monitored anesthesia care, spinalanesthesia, and peripheral nerve block [4]. What maybe most striking with respect to the present case isthat, among approximately 20 surgical services, neuro-surgery had one of the strongest associations with PNI.It is important to note that the nerve injuries reportedin this investigation were not due to a surgical or

163


preexisting neurologic etiology, but were likely relatedto positioning and patient co-morbidities. Neuro-surgery was associated with a 2.7-fold risk of PNI (95%confidence interval 1.4–5.1); only cardiac surgery wascomparable with a hazard ratio of 2.8 (95% confidenceinterval 1.3–5.7).

ConclusionThepresent case and the recent study byWelch et al. [4]demonstrate that neurosurgical procedures are asso-ciated with significant risk for PNI. These injuriesare distressing for patients and may mask or exacer-bate underlying pathology related to the central ner-vous system. Vigilance with respect to positioning andappropriate padding of pressure points is critical, espe-cially in patients with diabetes mellitus, hypertension,and a history of tobacco use. Neurophysiologic moni-toringmay aid in the intraoperative detection of injury

and should be taken seriously. Evidence of PNI shouldprompt an evaluation by neurology, as well as theinvolvement of physical and occupational therapy.

References1. F. W. Cheney, K. B. Domino, R. A. Caplan, K. L.

Posner. Nerve injury associated with anesthesia: aclosed claims analysis. Anesthesiology 1999; 90:1062–9.

2. D. A. Kroll, R. A. Caplan, K. Posner, R. J. Ward, F.W.Cheney. Nerve injury associated with anesthesia.Anesthesiology 1990; 73: 202–7.

3. A. R. Upton, A. J. McComas. The double crush innerve entrapment syndromes. Lancet 1973; 18:359–62.

4. M. B. Welch, C. M. Brummett, T. D. Welch et al.Perioperative peripheral nerve injuries: a retrospectivestudy of 380,680 cases during a 10-year period at asingle institution. Anesthesiology 2009; 111: 490–7.

164

Part IV Spine surgery. Cervical spine and airway issuesCase

52 Unstable cervical spineDavid L. Adams and David W. Healy

Each year in the USA there are approximately 10 000new cases of spinal injury and the cervical spine ismostcommonly affected. Young adult males have the high-est incidence of injury, with motor vehicle accidentsaccounting for 50% of cases; violent assault, falls, andsporting injuries account for the remainder.

The incidence of cervical spine injury in blunttrauma is approximately 2–4%. The incidence isincreased if there is an associated head injury with aGlasgow Coma Scale (GCS) score of �8 or if there isa focal neurologic abnormality. Up to 10% of patientswith a cervical spine injury will suffer a neurologicdeterioration while in the hospital.The term secondaryinjury refers to a deterioration or new injury fol-lowing the initial injury. The mechanism of injuryresponsible for the deterioration is often unclear. Sec-ondary neurologic injuries occurring in the periop-erative period are important as they are potentiallyavoidable.

Case descriptionA27-year-oldmalewas anunrestrained front-seat pas-senger in a motor vehicle accident. He was extractedfrom the vehicle and immobilized on a spinal board,with rigid cervical collar, sandbags and tape. On eval-uation in the emergency room he was found to havehead and facial injuries with bleeding from the noseand mouth. His GCS score was 12 (M5, V4, E3).Therewas bruising to the anterior chest wall and abdominaldistention with a heart rate of 125 beats per minuteand a blood pressure of 90/55 mmHg. Due to hemo-dynamic instability and a deteriorating level of con-sciousness, urgent airway control was deemed neces-sary. The anterior portion of the cervical collar wasremoved, manual in-line stabilization (MILS) main-tained and cricoid pressure was applied. The tracheawas intubated following a rapid sequence inductionwith propofol and succinylcholine. In an attempt tomaintain cardiovascular stability he received intra-

Figure 52.1. Injured cervical spine with a C6 fracture andspinal cord contusion revealed by magnetic resonance imaging(T2 weighted).

venous crystalloid and a bolus dose of phenylephrinearound the time of anesthetic induction. Definitiveimaging of the head and spine was deferred until afterthe patient had undergone urgent surgical manage-ment of his abdominal injuries. On the first postop-erative day he was suitably stable to undergo imagingandwas found to have a fracture at C6 with spinal cordcontusion revealed by magnetic resonance imaging(Figure 52.1).

DiscussionThe key aspects of anesthesia care for this patientwere (1) rapid definitive airway control, (2) protec-tion of neural structures from secondary injury, and(3) adequate intravenous access and fluid resuscita-tion. Endotracheal intubation with controlled ventila-tion is a common aspect of the initial care of patients

165

IV. Spine surgery. Cervical spine and airway issues

with severe trauma. This may be essential for air-way protection and ventilation following neurologicdeficit, or be required for the operative managementof other injuries. The anesthesia provider must main-tain the mechanical integrity of the spinal cord bylimiting neck movement, as well as ensure adequatespinal cord perfusion by avoiding hypotension andsubsequent tissue hypoxia. The immobilization of thespine in trauma patients until injuries have eitherbeen excluded or definitively treated remains a cor-nerstone of modern trauma care. The introduction ofspinal immobilization techniques in pre-hospital carehas dramatically reduced the incidence of secondaryspinal cord injury.Themost effective technique is witha backboard, rigid collar, sandbags, and tape. Thesetechniques reduce cervical spine motion to approxi-mately 5% of the normal range.

It has been demonstrated that basic airwaymaneu-vers such as chin lift, jaw thrust, and placement ofboth oral and nasal airways cause movement of cervi-cal spine segments. This is seen in both the intact andinjured cervical spine and is not completely preventedby the application of cervical collars.The use of immo-bilization techniques reduces but does not eliminatespinal movement associated with basic airway inter-ventions. Even mask ventilation may lead to cervicalspine displacement.The application of cricoid pressurehas not been shown to cause significant movementof cervical spine segments in either stable or unstablemodels.

In injured, as well as uninjured patients, directlaryngoscopy causes significant motion of cervicalspinal segments, particularly in the upper cervicalspine. Manual in-line stabilization has been foundto be the most effective method of limiting cervicalspine distraction during direct laryngoscopy, thoughit does not eliminate movement altogether [1]. Histor-ically, othermethods have been attempted but result inincreased movement at laryngoscopy.

While MILS limits the motion of injured spinalsegments, it can worsen the view during directlaryngoscopy. However, this view is better than thatobtained in those patients immobilized with collar,sandbag, and tape [2]. The use of a collar significantlylimits neck extension and mouth opening, which areessential for optimizing the view obtained at laryn-goscopy. During direct laryngoscopy one may removethe anterior part of the cervical collar while maintain-ing MILS. The consequences of failed airway controland resultant hypoxia can be devastating, and if an

adequate view of the glottis cannot be obtained, it isconsidered permissible to relaxmanual stabilization tothe degree required to intubate the trachea. The useof adjuncts such as the gum elastic bougie should beencouraged in such circumstances to minimize neckmovement. The use of both straight and curved laryn-goscope blades causes a comparable amount of spinemovement and there is no evidence to favor one bladeover the other.

In an attempt to optimize the view at laryngoscopywhile maintaining MILS, alternative airway instru-ments have been evaluated. The intubating stylets(e.g., Bonfils, Lightwand) and indirect video laryn-goscopes (e.g., Bullard, Glidescope) have been stud-ied in patients in spinal immobilization. These tech-niques allow a view of the glottis that does not relyupon neck extension and the alignment of the airwayaxes. Additionally, it has been suggested that less forcemay be required during the procedure. The use of theBullard, in particular, has been shown to result in lesscervical movement that direct laryngoscopy [3], butthe other methods (notably the Glidescope) cause thesame degree, or greater, of cervical distraction. In gen-eral, these techniques have been found to take longerthan direct laryngoscopy even in practiced hands.Theincreased time to intubation can be attributed to theincreased complexity of passing an endotracheal tubeout of the line of sight. The use of extraglottic air-way devices, such as the laryngeal mask airway causeless cervical spinemovement than direct laryngoscopy.However, their placement may specifically displace adestabilized upper cervical segment. The clinical rele-vance of this remains uncertain.

Flexible fiberoptic laryngoscopy causes the leastmovement and angulation of the spine [4], and in theelective situation has become the method of choice.A higher degree of skill is required when comparedwith other techniques, and even in experienced handsthe time taken for intubation is increased. In NorthAmerica anesthesiologists have expressed a preferencefor this technique in patients with suspected spinalinjuries, although some of the same practitioners ques-tion their own competence with the technique. Even inalert, cooperative patients, oxygen desaturation duringattempts at awake fiberoptic intubation iswell reportedand is clearly an undesirable event in the setting ofacute neurologic injury. Furthermore, blood or otherdebris in the airway can make the use of the fiberop-tic laryngoscope extremely difficult. In this setting,the lightwand is beneficial and is associated with less

166

Case 52. Unstable cervical spine

Table 52.1. Key management points in a patient withsuspected cervical spine injury.

Cervical spine injury occurs in 2–4% of victims of blunt trauma,and is particularly associated with head injury. This incidence isincreased if the Glasgow Coma Scale is �8 or there is a focalneurologic deficit

“Basic” airway techniques such as chin lift, jaw thrust and maskventilation can also cause cervical displacement

The presence of a cervical collar limits mouth opening andconsistently worsens the view obtained at laryngoscopy. Theanterior part of the collar may be removed to facilitate directlaryngoscopy

The use of manual in-line stabilization limits, but does notcompletely prevent, the movement of injured spinal segmentsduring airway interventions, but is still the technique of choicecompared with the other methods

No single intubation technique has been shown to be superiorto any others. The available evidence suggests a favorable safetyprofile with all techniques. Achieving rapid airway control whileminimizing hypoxia and hypotension should be primary goals

cervical spinemobility than direct laryngoscopy or theGlidescope.

The degree of movement and angulation in spinalsegments with any of these intubation techniques is ofuncertain clinical significance. While flexible fiberop-tic laryngoscopy causes the least amount ofmovement,it is not always the most appropriate technique. In theemergency room setting, timely placement of an endo-tracheal tube with prevention of hypoxia is vital andthe technique should be chosen with consideration of

the clinical state of the patient, as well as the skill andexperience of the laryngoscopist.

The evidence does not support the use of any onetechnique over another [5]. All types of laryngoscopyhave a favorable safety profile and the incidence ofproven neurologic deterioration attributable to airwaymanagement is extremely low. Table 52.1 summarizeskey points in themanagement of patients with cervicalspine injury.

References1. M. C. Gerling,D. P. Davis, R. S. Hamilton et al.

Effects of cervical spine immobilization technique andlaryngoscope blade selection on an unstable cervicalspine in a cadaver model of intubation. Ann EmergMed 2000; 36: 293–300.

2. K. J. Heath. The effect on laryngoscopy of differentcervical spine immobilization techniques. Anaesthesia1994; 49: 843–5.

3. R. H. Hastings, A. C. Vigil, R. Hanna et al. Cervicalspine movement during laryngoscopy with theBullard, Macintosh and Miller laryngoscopes.Anesthesiology 1995; 82: 859–69.

4. J. Brimacombe, C. Keller, K. H. Kunzel et al. Cervicalspine motion during airway management; acinefluoroscopic study of the posteriorly destabilizedthird cervical vertebrae in human cadavers. AnesthAnalg 2000; 91: 1274–8.

5. E. T. Crosby. Airway management in adults aftercervical spine trauma. Anesthesiology 2006; 104:1293–318.

167


53 Cervical spine limitationsDavid L. Adams and David W. Healy

Cervical spine mobility is central to the conventionalsafemanagement of the airway.The term cervical spinelimitations (CSL) refers to a range of limitations to neckmobility including limited extension, flexion or both,and known or suspected unstable cervical spine con-ditions requiring immobilization with a cervical col-lar or halo. Limitations of cervical spine mobility arecommon in patients presenting for neurosurgery, butare not confined to this group; the incidence acrossall adults at a tertiary care center in the USA hasbeen estimated at approximately 8% [1]. Congeni-tal causes include Klippel–Feil syndrome (character-ized by fusion of two or more cervical vertebrae) andGoldenhar syndrome (a type of mandibular hypopla-sia associated with underdeveloped vertebrae, oftenin the neck). Acquired causes are mainly degenera-tive diseases (osteoarthritis, degenerative disc disease),inflammatory processes (rheumatoid arthritis, anky-losing spondylitis), trauma, and prior surgical fusion(see Table 53.1).

Case descriptionA 68-year-old male with severe ankylosing spondyli-tis sustained a fracture through the C6 vertebral bodyfollowing a fall. Despite an initial period of immobi-lization in a halo jacket, he experienced persistent painand deformity, and was therefore scheduled for a C3–T4 posterior instrumentation and vertebral columnresection. In addition to the ankylosing spondylitis,the patient’s medical history was significant for hyper-tension, type 2 diabetes with peripheral sensory neu-ropathy, peripheral vascular disease, and symptomaticgastroesophageal reflux disease. Physical examinationshowed mild symmetrical sensory neuropathy in bothfeet, but nomotor deficit. Airway examination demon-strated a marked cervical kyphosis (“chin-on-chest-deformity”) with severe limitation of both flexionand extension. Although the patient was edentulous,mouth opening was limited, the Mallampati class was

Table 53.1. Causes of cervical spine limitations.

Congenital

Klippel–Feil syndrome

Goldenhar syndrome

Diffuse idiopathic hyperostosis (DISH)

Acquired

Osteoarthritis

Degenerative disc disease

Fracture/trauma

Surgical fusion

Scoliosis

Rheumatoid arthritis

Ankylosing spondylitis

Figure 53.1. Sagittal plane computed tomography image of thecervical spine showing a transverse fracture through C6 and typicalappearances of a “bamboo spine.”

IV, and there was limited mandibular protrusion. Thethyromental distance was 3 cm. A computed tomogra-phy (CT) image is shown in Figure 53.1.

In view of the potential difficulties with airwaymanagement and presence of symptomatic reflux dis-ease, an awake fiberoptic intubation technique waschosen. A peripheral venous line and a radial arterial

168

Case 53. Cervical spine limitations

catheter were placed under local anesthesia and gly-copyrrolate 200 mcg was administered intravenously.Topicalization of the airway with 4% lidocaine wasthen started in preparation for awake fiberoptic intu-bation. The patient was connected to standard mon-itors, including invasive blood pressure, and oxy-gen was administered at 2 liters per minute throughnasal canulae. Once adequate airway anesthesia hadbeen achieved, a remifentanil infusion was started at0.01 mcg/kg/minute, and oral fiberoptic intuba-tion was commenced. The remifentanil infusion wasadjusted to amaximumof 0.05mcg/kg/minute, guidedby the arterial blood pressure, respiratory rate, andthe degree of sedation. Supplementary airway anes-thesia was provided in a “spray-as-you-go” technique.Intubation of the trachea and placement of a size 7.0cuffed tracheal tube were uneventful with no episodesof desaturation and no significant deviation in theblood pressure from baseline. The patient was awakeand comfortable and a gross assessment of neurologicfunction was carried out before the induction of gen-eral anesthesia.

The patient was positioned in the prone positionwith care taken to avoid cervical spine extension and topreserve the alignment of the cervico-thoracic spine,to the extent that was possible given the underlyingdeformity. The blood pressure was maintained at pre-induction values at all times. At the end of surgery, thepatient once again demonstrated neurologic integrity;however, extubation was deferred until the followingmorning, to allow for airway edema to resolve and aleak around the tube cuff to develop.Therewere no air-way sequelae following extubation.

DiscussionAnkylosing spondylitis is a seronegative spondy-loarthropathy strongly associated with the HLA-B27 genotype. Recurrent painful inflammatory flaresinvolve the axial skeleton and sacroiliac joints, leadingto progressive bony fusion of adjacent vertebral seg-ments causing the characteristic fixed kyphotic “bam-boo spine.” Ankylosing spondylitis also affects otherparts of the musculoskeletal system and has asso-ciations with diseases in many organ systems (seeTable 53.2).

The rigidity of the fused spine makes it vulnerableto damage and fractures can occur following appar-ently minor injuries. Hyperextension of the cervicalspine is commonly associatedwith fractures at the C5–

Table 53.2. Common clinical features and diseaseassociations of ankylosing spondylitis.

Musculoskeletal system

Spinal fusion and rigidity (“bamboo spine”)

Atlanto-axial instability

Sacroiliitis

Peripheral arthritis (hips, knees, temperomandibular,costovertebral, and costochondral joints)

Achilles tendonitis

Plantar fasciitis

Other systems

Neurologic – Spinal cord and nerve root compression,vertebrobasilar insufficiency

Cardiovascular – Aortic valve insufficiency and cardiacconduction defects

Pulmonary – Upper lobe fibrosis

Gastrointestinal – Inflammatory bowel disease

Occular – Anterior uveitis

Skin – Psoriasis

Genitourinary – Prostatitis

6 level. Common causes include falls and motor vehi-cle accidents. Iatrogenic causes also should be consid-ered; poor patient positioning in the operating roomcan easily cause neurologic damage in these vulnera-ble patients.

The purpose of the preoperative airway assessmentis to predict, as accurately as possible, the degree ofdifficulty that will be encountered during mask venti-lation and tracheal intubation. The failure to intubatecombined with failure to adequately ventilate is rightlyfeared by anesthesiologists. It should be rememberedthat prolonged or forceful attempts to perform directlaryngoscopy and intubation may cause trauma toeither the airway or to the cervical spine and cord, par-ticularly in a patient with preexisting cervical spinepathology.

The assessment of the range of cervical spinemotion is a routine part of preoperative airway eval-uation. Many patients with CSL will have a reducedrange of extension at the occipito-atlanto-axial (OAA)complex, which is vital for maximal mouth opening.Therefore, these patients may demonstrate a reducedinterincisor distance, an increased Mallampati score,and reduced mandibular protrusion. The Mallampaticlassification (with the head in the neutral position,and tongue maximally protruded) can be modified innormal patients to include an assessment of mouth

169


opening and tongue protrusion at maximal cranio-cervical extension (Extended Mallampati Score) [2].The Extended Mallampati Score increases the positivepredictive value of the examination and reduces falsepositives. In normal patients the facilitation of mouthopening afforded by extension at the OAA complexcan decrease the modified Mallampati score; this maynot be seen in patients with CSL.

The incidence of difficult or impossible mask ven-tilation in patients with CSL when compared withnormal controls is increased. A Mallampati classi-fication of III or IV, limited jaw protrusion and athyromental distance of less than 6 cm are featuresthat can be found in association with CSL, and havebeen found to be predictive of difficult or impossiblemask ventilation [3]. The incidences of difficult laryn-goscopy [4] and difficult intubation [5] are more com-mon in patients with CSL. It has also been demon-strated that significant increases in the incidences ofboth difficult laryngoscopy and intubation are foundin CSL due in part to the difficulty in positioning thepatient.

Patients with a known or potentially unstablecervical spine and those in cervical collars or halotraction will generally be immobilized in the neu-tral position. Intubation difficulties in these patientsare principally related to the lack of extension atthe OAA complex. In patients that do not requireimmobilization in the neutral position, the conven-tional positions for direct laryngoscopy are with thehead extended at the OAA complex, or with the headextended at the OAA complex and the cervical spineflexed at the thoraco-cervical junction (the “sniffingposition”). The sniffing position gives rise to a greaterdegree of extension at the OAA complex [6], and inpatients with limited cervical spine extension (and inthe obese), this has been shown to be particularlyadvantageous [7].

When difficulty with direct laryngoscopy is antici-pated, a number of options are available to the anes-thesiologist. There is an increasing role for noveldirect and indirect video-assisted airway manage-ment devices in enabling the practitioner to “lookaround the corner” to facilitate tracheal tube place-ment. The common incidence of difficult mask ven-tilation in patients with CSL should lead to cautionwhen considering these techniques, as their use usu-ally requires induction of general anesthesia. Thereare some descriptions of their use in unanesthetized,

spontaneously breathing patients – but these are casereports and very small case series.

Awake fiberoptic intubation is the most populartechnique for definitive airway control in patients withan unstable or limited cervical spine. This methodwas performed in this case as it involved an elec-tive procedure, which gave adequate time for carefulpreparation. Even in an emergent situation themethodof awake fiberoptic intubation should be consideredon a risk–benefit basis. Awake fiberoptic intubationalso allows the assessment of gross neurologic functionafter tracheal tube placement. It is therefore desirableto use a technique which allows adequate patient com-fort and airway control, while avoiding over sedationduring the intubation process. The technique requirespatient cooperation and adequate airway topicaliza-tion. Our patient was pre-treated with glycopyrrolate,which reduces secretions and dries the mucosa of theupper airway. This enhances the speed and effective-ness of topical lidocaine, and has the added benefitof preventing a bradycardia which may occur withthe subsequent use of remifentanil. There are numer-ous described techniques for the topicalization of theairway and the conduct of awake fiberoptic intuba-tion. These are technical skills that are best learned atthe bedside under the instruction of an experiencedpractitioner.

Remifentanil was used as an adjuvant during theawake fiberoptic intubation. It is rapidly titratable andsuppresses the urge to cough and gag, while verbalcontact with the patient can be maintained. Infusionshould be started at very low rates and titrated gently toeffect. A loading dose should be avoided as it can causeapnea, so infusions should start several minutes beforeairway instrumentation to allow the achievement ofsteady-state plasma drug levels. Rapidly increasing therate of infusion can cause over-sedation and subse-quent apnea and desaturation. It is therefore not rec-ommended. Unanticipated hypotension, hypoxia andover-sedation are adverse events that are well reported,and the presence of two practitioners at the time ofintubation should be encouraged. While one anesthe-siologist conducts the fiberoptic intubation, the sec-ond can administer sedation while monitoring thepatient’s level of consciousness, blood pressure, andoxygenation.

The patient presented in this case demonstratedseveral of the features that predict difficulty in air-way management. Furthermore, the “chin-on-chest-

170

Case 53. Cervical spine limitations

deformity” would have prevented the formation of asurgical airway. Awake flexible fiberoptic intubation isconsidered to be the gold standard in this challengingpatient group.

References1. G. A. Mashour,M. L. Stallmer, S. Kheterpal et al.

Predictors of difficult intubation in patients withcervical spine limitations. J Neurosurg Anesthesiol2008; 20: 110–15.

2. G. A. Mashour,W. S. Sandberg. Craniocervicalextension improves the specificity and predictive valueof the Mallampati airway evaluation. Anesth Analg2006; 103: 1256–9.

3. S. Kheterpal, R. Han, K. K. Tremper et al.Incidence and predictors of difficult and impossible

mask ventilation. Anesthesiology 2006; 105:885–91.

4. I. Calder, J. Calder,H. A. Crockard. Difficult directlaryngoscopy in patients with cervical spine disease.Anaesthesia 1995; 50: 756–63.

5. K. Karkouti,D. K. Rose,D.Wigglesworth et al.Predicting difficult intubation: a multivariableanalysis. Can J Anaesth 2000; 47: 730–9.

6. I. Takenaka, K. Aoyama, T. Iwagaki et al.The sniffingposition provides greater occipito-atlanto-axialangulation than simple head extension: a radiologicalstudy. Can J Anaesth 2007; 54: 129–33.

7. F. Adnet, C. Baillard, S. W. Borron et al. Randomizedstudy comparing the “sniffing position” with simplehead extension for laryngoscopic view in electivesurgery patients. Anesthesiology 2001; 95: 836–41.

171


54 Rheumatoid diseaseDavid L. Adams and David W. Healy

Rheumatoid arthritis (RA) is a progressive symmet-rical, deforming inflammatory polyarthropathy withnumerous extra-articular features. It is a common con-dition, with prevalence in the USA of approximately1%, affecting females three times more commonlythan males. The skeletal effects of RA are character-ized by an inflammatory synovitis with progressivedestruction of cartilage. Over time, ankylosis of jointsoccurs, causing stiffness and reduced range of motion.The small joints of the hands and feet are most com-monly involved; however, any joint can be affected.The cervical spine complications of RA can lead toinstability and limitation of motion, and surgery iscommonly required for stabilization and correctionof deformity. The cervical instability and limitationof movement (including prior surgical stabilization)must be carefully considered before anesthetizing thispatient group.

Case descriptionA 62-year-old female with seropositive rheumatoidarthritis presented for anterior cervical discectomyand fusion for long-term progressive pain and upperextremity myelopathy. The patient described worsen-ing intermittent tingling and increasing, symmetri-cal, arm weakness. Airway examination in the neutralposition revealed reduced mouth opening (�3 cm), aMallampati class III examination, and severely limitedneck extension. Blood pressure was 140/90mmHg andbody weight was 60 kg.

Preoperative imaging consisted of plain radio-graphs and magnetic resonance imaging of the cervi-cal spine. These demonstrated spinal stenosis at C5/6due to a posterior disc bulge, and erosive changes atthe occipito-atlanto-axial (OAA) complex, typical ofsevere rheumatoid disease.

In view of the degree of cervical spine pathology,an awake fiberoptic intubation was planned. This wasperformed uneventfully and a neurologic examination

after tracheal tube placement showed that no dete-rioration had occurred. General anesthesia was theninduced.

Thepatientwas positioned supine and the headwassupported on a Mayfield ring. Great care was takento ensure that the neck was in a neutral position.The intraoperative course was uneventful with main-tenance of blood pressure at or around pre-inductionvalues.

At the end of the case, the patient was awakenedbut remained intubated to allow neurologic status tobe re-evaluated before extubation. This was satisfac-tory and the airway was then assessed for the pres-ence of swelling and edema with the endotrachealtube cuff deflated. No significant leak was detectedso the decision was made to leave the trachea intu-bated and sedated and transferred to the neurosurgicalintensive care unit. Twelve hours later the trachea wasuneventfully extubated after the airway swelling hadimproved.

DiscussionThis case illustrates several important issues that mustbe considered when identifying the airway manage-ment options for patients with cervical spine pathol-ogy. A careful history and examination are vitalto establish the range of symptom-free neck move-ment and elucidate symptoms and signs of possi-ble nerve impingement or spinal cord compression.Closely allied to this is a formal examination of theairway.

The effect of RA on the cervical spine is typifiedby a combination of structural instability and stiffness[1]. The instability is due to progressive erosion ofbone and ligaments, and the stiffness due to chronicinflammation, thickening, and calcification. Thiscase exhibited both features. In addition to limitingneck extension and causing potential instability, RAmay disrupt both the temporomandibular joints and

172

Case 54. Rheumatoid disease

Figure 54.1. Diagram representing the radiological changesassociated with degradation of the transverse ligament andresulting atlantoaxial subluxation.

laryngeal cartilage articulations. Pain on masticationor mouth opening may indicate temporomandibularjoint disease. A history of increasingly hoarse voicemay suggest disease affecting the arytenoid cartilages.Along with the cervical stiffening and bony degener-ation associated with RA, atlantoaxial instability is ofparticular concern. Ligamentous degradation aroundthe OAA complex, particularly of the transverseligament, can lead to atlantoaxial subluxation. Thetransverse ligament of the axis links the lateral massesof the atlas and loops around the odontoid processholding it firmly in contact with the anterior archof the atlas. With degradation of this ligament theodontoid process becomes free to move, bringing theaxis with it resulting in atlantoaxial subluxation. Thedisplaced odontoid process can compress the spinalcord or medulla and damage the vertebral arteries.

It is well known that atlantoaxial instability canbe completely asymptomatic, but there is little con-sensus regarding mandatory cervical spine imagingfor patients with RA undergoing surgery. A reason-able approach is to obtain flexion and extension viewsof the cervical spine preoperatively for all patientswith neurologic symptoms and for those taking regu-

Figure 54.2. Lateral C spine X-ray demonstrating atlantoaxialsubluxation.

lar steroids or disease-modifying antirheumatic drugs(e.g., methotrexate). The normal distance between theanterior edge of the odontoid process and the posteriorborder of the anterior arch of the atlas should be lessthan 3 mm; any increase is pathological (Figure 54.1).Imaging was obtained preoperatively in this case. Thebest way to identify a widening of the atlanto-odontoidspace is on a lateral cervical X-ray taken in flexion(Figure 54.2); alternatively it can be imaged by com-puted tomography scans (Figure 54.3) ormagnetic res-onance imaging.

When considering the airway options for a patientwith disease at the OAA complex an appreciation ofthe importance of this area in overall airway man-agement is essential. Extension at the OAA complexis important for both basic airway management suchas “chin lift” and mask ventilation as well as to facili-tate direct laryngoscopy. Neck extension may be lim-ited by disease or iatrogenic externally applied stabi-lization (e.g.,manual in-line stabilization).The efficacyof the sniffing position, to facilitate the view at directlaryngoscopy, is thought to be due to its ability tomax-imize the degree of extension at the OAA complex.Adequate mouth opening is important for successfulairway management, for which neck extension isessential [2].

173


Figure 54.3. Computed tomography demonstrating atlantoaxialsubluxation.

Rheumatoid arthritis can also cause pathologicalchanges in the lower cervical spine. During cervicalspine extension, there can be posterior disc protrusionand buckling of the ligamentumflavum.These changesmay cause cord impingement in the neutral position.Therefore, any additional extension (and to a lesserextent, flexion) may increase spinal cord compression.Cord injurymay be produced bymechanical deforma-tion, ischemia or both.

To prevent damage to neurologic structures it isdesirable to maintain the cervical spine in a neutralposition. This may be different for each patient andcare should be taken to enlist the patient’s coopera-tion in finding the most comfortable neck position.Some practitioners advocate positioning the patientfor surgery before the induction of general anesthesia,though this may present difficulties, particularly if thepatient is to be in the prone position. The prone posi-tion often causes a degree of cervical spine extension,which should be minimized as much as possible.

Awake fiberoptic intubation is the most popu-lar technique for definitive airway control in patientswith an unstable or limited cervical spine [3]. Thismethod is associated with the least amount of cervi-cal spine movement, and in our patient, extension wascompletely avoided at the OAA complex. Preopera-tive assessment of this patient demonstrated limitedmouth opening, and aMallampati class III assessment,

strongly indicative of limited extension at the OAAcomplex. In very severe temporomandibular joint dis-ease it may be impossible to pass an oral endotrachealtube and the nasal route should instead be used. Directlaryngoscopy in our patient could have either causedcord injury at the site of an unstable OAA complex, orfailure to achieve an adequate view of the glottis dueto a severe limitation of spinal extension. A role hasbeen suggested for indirect video laryngoscopy in thisclinical setting, but its efficacy is unproven beyond animprovement in glottic view [4].

This patient was awakened at the end of the pro-cedure, and a further neurologic examination per-formed. The examination was comparable with thatperformed before induction of anesthesia, though itshould be appreciated that this can only ever be a crudeassessment of neurologic function. Most practition-ers would wait until this gross assessment of neuro-logic function had been completed before extubatingthe trachea, as a worsening of function may precipi-tate the need for urgent imaging or re-operation.At thepreoperative visit patients should be warned that theymay awake with a “breathing tube” in place and askedto perform simple tasks on command (e.g., “Squeezehands” etc.). They should be reassured that they willbe comfortable and relaxed and the tube will soon beremoved.

The trachea in this case, however, was not extu-bated at the end of the procedure. This was due tothe absence of a leak around the deflated tracheal tubecuff, suggesting airway edema and soft tissue swelling.This should always be considered at the end of cervicalspine surgery and cases in which the patient has beenpositioned in the prone or head-down position [5].Such patients should remain intubated and re-sedatedif necessary and periodically re-evaluated for evidenceof resolution of the swelling.The head-up position andthe use of corticosteroidsmay accelerate the resolutionof edema.

ConclusionThis case typifies some of the challenges encounteredby the anesthesiologist when managing the airway ina patient with rheumatoid arthritis and limited cervi-cal spine motion. Careful preoperative evaluation andplanning are essential; adequate time and personnelshould be allocated to securing the airway in a con-trolled and safemanner. Finally, it should be noted thatextubation is a high-risk time for these patients and

174

Case 54. Rheumatoid disease

that a plan should be made to ensure this is performedsafely and at the appropriate time.

References1. A. MacArthur, S. Kleimann. Rheumatoid cervical

joint disease – a challenge to the anaesthetist. Can JAnaesth 1993; 40: 154–9.

2. I. Calder, J. Calder,H. A. Crockard. Difficult directlaryngoscopy in patients with cervical spine disease.Anaesthesia 1995; 50: 756–63.

3. E. T. Crosby. Considerations for airway managementfor cervical spine surgery in adults. Anesthesiol Clin2007; 25: 511–33.

4. A. D.Watts, A.W. Gelb,D. B. Bach et al. Comparisonof Bullard and Macintosh laryngoscopes forendotracheal intubation of patients with a potentialcervical spine injury. Anesthesiology 1997; 87: 1335–42.

5. H. C. Sagi,W. Beutler, E. Carroll et al. Airwaycomplications associated with surgery on the anteriorcervical spine. Spine 2002; 27: 949–53.

175

Part V Neuroendocrine surgeryCase

55 Preoperative evaluationNicholas F. Marko and Robert J. Weil

Preoperative evaluation of patients presenting fortransphenoidal resection of pituitary tumors is a verycomplex process, requiring careful assessment of thepatient’s symptoms and the proper preoperative lab-oratory tests. The pituitary tumor can be a secretorytumor or a nonsecretory tumor. The following caseswill highlight the proper preoperative evaluation fordifferent types of pituitary tumors.

Case description 1History: A 32-year-old male was referred for evalua-tion of a possible neuroendocrine disorder.Thepatienthad a 7-year history of hyperlipidemiamanaged with astatinmedication aswell as hypertension that had beencontrolled with a diuretic and an ACE-inhibitor. Hehad requiredmultiple increases in the dosages of thesemedications over the past several years. The patientreported no acute symptoms and had no new com-plaints. On careful questioning he reported that hehad limited exercise tolerance (he could only walk2–3 blocks without resting) and had daily “arthritis inmy back, knees and shoulders.” He felt that this wasbecause he had always been “a big man,” and he notedthat he has had limited success with attempts at weightcontrol with diet and exercise in the past. He knowshe should exercise more because he has been told heis a “borderline diabetic.” He denied headache, recent,unintentional weight changes, heat or cold intolerance,polyuria, polydypsia, or galactorrhea. His medical his-tory was remarkable for hypertension and hyperlipi-demia, as well as sleep apnea since age 25 for whichhe used nightly continuous positive airway pressure(CPAP) therapy. His only surgical history was bilateralcarpal tunnel release 2 years ago, “because I type all dayat work.” His medications included hydrochloroth-iazide 25mgQD, lisinopril 20mgQD, and atorvastatin40 mg QD. He was a nonsmoker.

Examination: He was a 32-year-old man in no dis-tress, although he appearedmildly diaphoretic. Hewas

178 cm tall and weighed 95.5 kg (BMI 30). He wasneurologically intact, including no visual field deficits,no cranial neuropathies, and no other focal neurologicdeficits. His skin appeared thickened and his facial fea-tures were coarse. Scars from his carpal tunnel surgerywere noted bilaterally, and his radial pulses were diffi-cult to palpate because his hands and wrists were large.The remainder of his physical examwas unremarkableand was notable for the absence of galactorrhea, buf-falo hump, or abdominal striae.

Radiology: Gadolinium-enhanced magnetic reso-nance imaging (MRI) demonstrated a 7-mm, well-circumscribed area of diminished enhancementwithin an otherwise-normal anterior pituitary gland.There was no compression of the optic apparatus andno hydrocephalus. The remainder of the MRI wasunremarkable.

Laboratory evaluation: Laboratory tests wereobtained as shown in Table 55.1. Normal values areindicated in parentheses.

Diagnosis: Acromegaly – growth hormone secret-ing pituitary microadenoma.

Discussion: Acromegaly is the clinical syndromeassociated with abnormally elevated growth hormone(GH) levels. It is a slowly progressive disease, sothere is often a period of years between the onsetof symptoms and the final diagnosis. Patients withacromegaly classically exhibit coarse facial features,acral enlargement, and thickened skin [1]. Thesechanges give rise to several classic associated findings,including carpal tunnel syndrome (often bilateral)and obstructive sleep apnea. Joint arthropathies andcalcific spinal disc disease are also common in patientswith acromegaly. Cardiovascular manifestationsinclude hypertension, cardiomyopathy, ventricularhypertrophy, arrhythmias, and cerebrovascular dis-ease, while endocrinologic findings include diabetesand hyperparathyroidism. Patients are also proneto malignancy, particularly those of the colon andpancreas. Because of these effects of GH excess on

177

V. Neuroendocrine surgery

Table 55.1. Laboratory results for patient in case description 1.

CBC: Normal FSH: 2 mU/mL (0–10 mU/mL)

BMP: Fasting glucose 112 mg/dL (<100 mg/dL) LH: 1 mU/mL (<7 mU/mL)

Prolactin: 15 ng/µL (2–17 ng/µL) GH: 2.8 ng/mL (0–3 ng/mL)

TSH: 0.750 µU/mL (0.500–5.500 µU/mL) IGF-1: 937 ng/mL (117–329 ng/mL)

Free T4: 1.0 ng/dL (0.7–1.8 ng/dL) Total cortisol, 07:00 17 mcg/dL (3.4–26.9 mcg/dL)

Low-dose (1 mcg) ACTH (cortrosyn) stimulation test: t0 = 17 mcg/dL, t30 min = 29 mcg/dL, t60 min = 29 mcg/dL (�18 mcg/dL within60 minutes after cortrosyn administration).

Oral glucose tolerance test

Baseline (t0) 30min (t30) 60min (t60) 90 min (t90) 120min (t120) 150 min (t150)

GH (ng/mL) 2.8 3.2 3.6 3.1 2.8 2.6

Glucose (mg/dL) 112 198 186 172 161 152

(Normal: GH �1.0 by t150; Glucose �100 by t150.)

multiple organ systems, patients with acromegaly havesignificantly increased mortality versus the generalpopulation [1]. Acromegalics have a standardizedmortality ratio of 1.72 (95% CI: 1.62–1.83), a 32%increased risk for all-cause mortality, and a death rateof 50% by the age of 60 in untreated patients. However,with successful treatment resulting in sustained GHlevels �2.5 ng/mL and normalization of insulin-like growth factor 1 (IGF-1) levels, the standardizedmortality ratio returns to 1.1 (95% CI: 0.9–1.4).

The classic clinical syndrome of acromegaly is GHhypersecretion, but GH levels alone are unreliable indiagnosing hypersecretory states because GH secre-tion is pulsatile and may vary throughout the day. Thesingle, best screening test is the level of IGF-1, which isinduced by GH, has a longer half life, and maintains amore constant serum concentration over time.Normallevels of IGF-1 are age and sex dependent, so resultsmust be interpreted in this context. While elevation ofIGF-1 relative to appropriately matched control levelsis highly suggestive of GH hypersecretion, the defini-tive test remains the oral glucose tolerance test. In thistest, baseline glucose and GH levels are measured justprior to administration of a 75 g oral glucose load toa fasting patient. Growth hormone and glucose lev-els are then assayed at 30 minute intervals for the next150–180 minutes. The oral glucose load should resultin suppression of GH levels to �1 ng/mL by the end ofthe test, and failure to suppress to this level definitivelydiagnosesGHhypersecretion.Monitoring glucose lev-els during the same test can diagnose impaired glucose

tolerance, which is said to be present if the glucose levelremains elevated above 140mg/dL at the conclusion ofthe test.

Surgical resection is the treatment of choice forGH-secreting adenomas because of the significantimprovement in morbidity and survival that resultsfrom rapid and durable reduction of GH and nor-malization of IGF-1 levels. Transphenoidal resec-tion results in normalization of IGF-1 levels in 75–95% of patients with microadenomas and 40–68%of patients with macroadenomas. Medical manage-ment with somatostatin receptor ligands (octreotide,lanreotide), GH receptor agonists (pegvisomant), ordopamine agonists (cabergoline) can normalize IGF-1 levels in 60–90% of patients but are costly andare unlikely to achieve definitive cure, requiring life-long administration [2]. Medical management strate-gies are useful for preoperative reduction in tumorsize, for up-front or postoperative control in patientswith tumors not amenable to gross total resection,in patients without symptoms of mass effect await-ing definitive treatment, in patients with surgical con-traindications, or in preparation for or as an adjunct toradiation therapy or radiosurgery [1, 2].

Case description 2History: A 36-year-old female was referred by her der-matologist. She had been followed for several yearsfor acne that had been poorly controlled with stan-dard medical management. Over the past year she had

178


noticed the appearance of purple abdominal striae andmultiple ecchymoses on her arms and legs. When shereported these findings to her dermatologist she wasreferred to the neuroendocrine clinic for evaluation.She stated that she was generally healthy except fordepression and anxiety, which caused difficulty sleep-ing. On detailed questioning she reported an increasedappetite over the past 6 months with 15–20 poundsof unintentional weight gain. She had also had irregu-lar menstrual cycles over a similar time period, whichshe attributed to stress and to her depression. She hadno acute complaints. Her medications included dulox-etine 60 mg QD for her depression and anxiety. Hermedical and surgical histories were otherwise unre-markable. She drank 2–3 beers per day and smoked2 packs of cigarettes per week. She had never beenpregnant.

Examination: She was a 36-year-old woman in nodistress. She was 163 cm tall and weighed 79.5 kg(BMI 30). She had normal visual fields and acuity.Her cranial nerves were intact. She had 4/5 deltoidstrength bilaterally with some muscle atrophy but hadno other neurologic deficits. Her face was rounded,and she had mild hirsutism and prominent supraclav-icular fat deposits. Her abdominal exam was remark-able for the presence of red stria.The remainder of herphysical exam was unremarkable and was notable forthe absence of galactorrhea and acral enlargement.

Radiology: Non-enhanced magnetic resonanceimaging demonstrated a normal sellar region witha pituitary gland of normal size and configuration.With gadolinium contrast, there was a 3 mm area ofabnormal intensity to the right of midline within theanterior pituitary gland. No other lesions were seen(Figure 55.1).

Laboratory evaluation: Laboratory tests wereobtained as shown in Table 55.2. Normal values areindicated in parentheses.

Diagnosis: Cushing’s disease – pituitary microade-noma.

Discussion: Cushing’s syndrome (CS) is the clinicalmanifestation of cortisol excess, and Cushing’s disease(CD) specifically describes cortisol excess caused byan ACTH-secreting pituitary adenoma. There are sev-eral clinical hallmarks of CS. First is progressive obe-sity, including generalizedweight gain, truncal obesity,and fat deposits in the face (“moon face”) and dorsalcervical and supraclavicular regions (“buffalo hump”).Next are dermatologic manifestations, including acne,skin hyperpigmentation, abdominal striae, easy bruis-

Figure 55.1. Coronal gadolinium-enhanced magnetic resonanceimage of the sella. The arrow indicates a small area of decreasedenhancement that suggests a possible pituitary microadenoma.

ability, skin atrophy, and frequent fungal infections[3]. Third are signs of androgen excess, including hir-sutism, virilization, oily skin, and irregular menstru-ation in women. Other systemic symptoms, includingdiastolic hypertension, osteopenia, and proximalmus-cle wasting are observed. Finally, psychiatric symp-toms including anxiety, depression, and emotionallability are frequently seen in patients with CS [3]. Likeacromegaly, the many systemic effects of CS lead toincreased mortality, which may be as high as 4–5-foldthat of the general population.

Hypercortisolemia can result from unregulatedproduction of cortisol by the adrenal glands or througha normal adrenal response to central or ectopic eleva-tions of ACTH. The case above represents an exampleof the challenges associated with diagnosing CS andidentifying the source of cortisol excess. The first stepis to identify elevated cortisol levels. In this patient,the 00:00 and 07:00 random cortisol levels were atthe high end of normal, and the 24-hour urine freecortisol was mildly elevated. While this suggests cor-tisol excess, it leaves open the possibility of hyper-cortisolism associated with other medical conditions(termed “pseudo-Cushing’s”), including physical oremotional stress, infection, depression, or alcoholism,all of which are suggested by her history. Her failure to

179

V. Neuroendocrine surgery

Table 55.2. Laboratory test result for patient in case description 2.

CBC, BMP: WBC 13 K/µL (3 K/µL) LH: 6 mU/mL (�12 mU/mL)

Urine � GH: 1.2 ng/mL (0–3 ng/mL)

Prolactin: 3 ng/µL (2 ng/µL) IGF-1: 201 ng/mL (117–329 ng/mL)

TSH: 4.000 µU/mL (0.500 µU/mL) Total cortisol, 07:00: 25 µcg/dL (3.4–26.9 mcg/dL)

Free T4: 1.2 ng/dL (0.7 ng/dL) Total cortisol, 00:00: 5 mcg/dL (�5 mcg/dL)

FSH: 6 mU/mL (1 mU/mL) ACTH: 50 pg/mL (8–20 pg/mL)

24-hour urine free cortisol (UFC): 65 mcg/24h (�45 mcg/24h).Low-dose dexamethasone (1mg) suppression test: 3.2 mcg/dL (�1.8 mcg/dL).High-dose dexamethasone (8mg) suppression test: 4.2 mcg/dL (�1 mcg/dL).

Corticotropin Releasing Hormone (CRH) stimulation test

15 minutesbefore CRH(t−15)

5minutesbefore CRH(t−5)

CRHDose(t0)

15minutesafter CRH(t15)

30minutesAfter CRH(t30)

45 minutesAfter CRH(t45)

60minutesAfter CRH(t60)

Cortisol (mcg/dL)[% over baseline]

25 [0%] 26 [4%] 25 [0%] 24 [−4%] 30 [20%] 31 [24%] 28 [12%]

ACTH (pg/mL)[% over baseline]

52 [2%] 50 [−2%] 51 [0%] 66 [29%] 72 [41%] 70 [37%] 63 [24%]

Inferior petrosal sinus sampling (IPSS)

ACTH (pg/mL)

Petrosal Sinus

ACTH Ratio

Peripheral Right Left R/P L/P R/L

CRH (−10 min) 37 212 36 5.7 0.97 5.9

CRH (−5 min) 41 451 43 11.0 1.0 10.5

CRH (+2 min) 34 1214 44 35.7 1.3 27.6

CRH (+5 min) 63 2855 92 45.3 1.5 31.0

CRH (+10 min) 80 90 112 1.1 1.4 0.8

CRH (+15 min) 82 152 136 1.9 1.7 1.1

adequately suppress on the low-dose dexamethasonesuppression test, however, makes a definitive diagno-sis of true hypercortisolemia [4]. This patient’s MRIsuggests a potential microadenoma; however, even anormal pituitaryMRI does not rule out an occult pitu-itary lesion. The next step is therefore to use physio-logic testing to determine if the excess cortisol is beingproduced de novo from an adrenal lesion or repre-sents a normal adrenal response to abnormally ele-vated ACTH. In this patient, the laboratory finding

of elevated serum ACTH narrows the differential toa pituitary or an ectopic source of ACTH. The high-dose dexamethasone suppression test is used to makethis distinction, because pituitary adenomas produc-ingACTHwill generally be suppressedwith high dosesof dexamethasone while ectopic sources will not. Cor-tisol levels �1 mcg/dL reflect normal physiology or apituitary source, levels �5 mcg/dL suggest an ectopicsource, and levels from 1–5 mcg/dL are suggestive of apituitary etiology but are considered equivocal [4, 5].

180


This patient demonstrated an equivocal response,which argues toward a pituitary source but is notdefinitive. In this circumstance additional testing isrequired. The cortrosyn releasing hormone (CRH)stimulation test measures serum ACTH and cortisollevels in response to intravenous administration of100mcg of ovoidCRH.ACTH-secreting pituitary ade-nomas will respond to CRH stimulation by increas-ing ACTH and cortisol production, while ectopic (andadrenal) sources will demonstrate no response. Thepatient is considered to respond if ACTH increasesto 35–50% over baseline by 30 minutes or if cortisolincreases to 20–50% over baseline by 60 minutes. Inthis patient the maximum ACTH increase was 41%over baseline (at 30 minutes) and the maximum cor-tisol increase was 24% (at 45 minutes). These find-ings also suggest a pituitary source, but the valuesare borderline. The final, definitive test involves CRHstimulation followed by direct sampling of ACTH lev-els in blood collected from the bilateral inferior pet-rosal sinuses (IPS), the central cerebral venous struc-tures into which the pituitary vascular plexus drains.Comparing the IPS ACTH levels to those of ACTHin the peripheral venous system is used to computea central-to-peripheral ACTH gradient. A gradientvalue greater than 2-fold before stimulation or greaterthan 3-fold within 10 minutes after stimulation is con-sistent with a pituitary etiology for the hypercorti-solemia [4, 5]. This patient demonstrates a baselineright IPS:peripheral ratio of 5.7:1 at baseline and 45.3:1at 5 minutes. This is diagnostic for a pituitary eti-ology of CS or CD. Of note, these findings do notnecessarily localize the lesion to the right side of thepituitary gland.

The treatment of CD involves surgical resectionof the pituitary adenoma, because removal of thelesion and rapid normalization of the serum corti-sol improves survival in these patients. In patients

with occult microadenomas, surgery often involvesa transsphenoidal approach and surgical explorationof the entire gland. If abnormal-appearing tissue isidentified it is resected and sent to pathology for defini-tive identification of the adenoma. When an ACTH-secreting tumor is identified and completely removed,85–90% of patients will achieve sustained remission.If no grossly abnormal tissue is apparent, hemihy-pophysectomy (guided by the IPS data) may be con-sidered. If no adenoma is identified on final pathologyand/or if cortisol levels fail to normalize postopera-tively, early surgical re-exploration and possibly, com-plete hypophysectomy may be required. Such patientswill experience panhypopituitarism and will requirelifelong hormone supplementation, but they will gainthe survival benefits associated with control of theirhypercortisolemia [6].

References1. A. Ben-Shlomo, S. Melmed. Acromegaly. Endocrinol

Metab Clin N Amer 2008; 37: 101–22.2. S. Melmed, A. Colao, A. Barkan et al.The pituitary

Society and the European NeuroendocrineAssociation. Guidelines for acromegaly management:an update. J Clin Endocrinol Metab 2009; 95: 1509–17.

3. J. Newell-Price, X. Bertagna, A. B. Grossman et al.Cushing’s syndrome. Lancet 2006; 367: 1605–17.

4. B. M. Biller, A. B. Grossman, P. M. Stewart et al.Treatment of adrenocorticotropin-dependentCushing’s syndrome: a consensus statement. J ClinEndocrinol Metab 2008; 93: 2454–62.

5. L. K. Nieman, B. M. Biller, J. W. Findling et al.Thediagnosis of Cushing’s syndrome: An EndocrineSociety Clinical Practice Guideline. J Clin EndocrinolMetab 2008; 93: 1526–40.

6. M. B. Elamin,M. H. Murad, R. Mullan et al.Accuracy of diagnostic tests for Cushing’s syndrome: asystematic review and metaanalyses. J Clin EndocrinolMetab 2008; 93: 1553–1562.

181


56 Acromegaly and gigantismD. John Doyle

An excess of growth hormone (GH) from a pituitaryadenoma can result in gigantism and acromegaly; neu-rosurgical intervention is often required. Airwayman-agement is a prime consideration in the anestheticmanagement of these patients [1–3].

Case descriptionA36-year-old femaleweighed 115 kg andwas 218.5 cmtall. She was scheduled for a transsphenoidal resec-tion of the pituitary adenoma that had caused hergigantism. She also had obvious acromegalic features.The upper airway evaluation showed a large protrud-ing tongue with grade IV Mallampati score. Due tothe anticipated difficult airway management, awakefiberoptic intubation was performed after local anes-thetic topicalization of the upper airway. The fiberop-tic laryngoscopy proceeded with difficulty given thepronounced macroglossia and redundant tissue sur-rounding the glottis. Several hypertensive episodeswere treated with labetolol. After securing the upperairway, the patient had general anesthesia and anuneventful surgery.

DiscussionThe term “acromegaly” comes from the Greek wordsfor “extremities” (acro) and “enlargement” (megaly).Giants have been known from biblical times. Theacromegalic patient suffers from an excess of GH,usually from a pituitary adenoma derived from soma-totroph cells. If the condition occurs prior to closureof the epiphyseal growth plates, gigantism may occur.Once the growth plates have fused in adolescence, thepatient may take on acromegalic features.

The goals of treatment are to normalize GH pro-duction while avoiding adverse effects on other pitu-itary functions (adrenocorticotropic hormone, folliclestimulating hormone, luteinizing hormone, prolactin,thyroid stimulating hormone, etc.), to relieve pressureon nerves and other structures exerted by the tumor,

and to reverse the symptoms of acromegaly. Treatmentoptions include surgical removal of the tumor, drugtherapy (bromocriptine, octreotide), and radiation tothe pituitary. If the tumor has not invaded surroundingbrain tissue, surgical removal of the tumor, often viaa transsphenoidal approach, is usually the first choice.Radiation therapy is usually reserved for patients whohave tumor remaining after surgery. Drug therapy canbe useful in nonpituitary causes of acromegaly as wellas to shrink large tumors before surgery.

From the viewpoint of airway management in theacromegalic patient, several concerns exist: (1) thetongue may be very enlarged; (2) redundant foldsof tissue may be present in the oropharynx; (3) theepiglottis is often enlarged; and (4) laryngeal steno-sis occurs more frequently compared with the gen-eral population.These factors maymake laryngoscopyand intubation considerably more difficult and alsoincrease the likelihood of airway obstruction dur-ing anesthetic induction and recovery. Acromegalicpatients are also prone to obstructive sleep apnea.In the case of gigantism, several additional potentialproblems should also be considered: (1) possible needfor an extra long operating table; (2) possible needfor an extra large laryngoscope; (3) endotracheal tubesmay need to be placed deeper than usual; and (4) anextra large face mask may be needed as in our case.

A number of decades ago, if direct laryngoscopyor blind intubationwas unsuccessful, tracheotomywasoccasionally required to manage the airway in thesepatients. However, the development of the fiberscopehas greatly reduced the need for this. Still, it shouldbe emphasized that acromegalic patients can be agreat airway challenge. In a study by Schmitt et al.the authors found that laryngoscopy was difficult(grade III view) in over one quarter (26%) of acrome-galic patients [3]. Consequently, special attentionshould be paid to the availability of alternative air-waymanagement techniques, such as the use of airwayadjuncts, fiberoptic intubation, video laryngoscopy,

182

Case 56. Acromegaly and gigantism

etc. Awake intubation may be needed in some cases[4, 5].

References1. S. Z. Hassan, G. J. Matz, A. M. Lawrence et al.

Laryngeal stenosis in acromegaly: a possible cause ofairway difficulties associated with anesthesia. AnesthAnalg 1976; 55: 57–60.

2. J. P. Southwick, J. Katz. Unusual airway difficulties inthe acromegalic patient – indications fortracheostomy. Anesthesiology 1979; 51: 72–3.

3. H. Schmitt,M. Buchfelder,M. Radespiel-Troger et al.Difficult intubation in acromegalic patients: incidenceand predictability. Anesthesiology 2000; 93:110–14.

4. P. A. Seidman,W. A. Kofke, R. Policare et al.Anaesthetic complications of acromegaly. Br J Anaesth2000; 84: 179–82.

5. E. C. Nemergut, Z. Zuo. Airway management inpatients with pituitary disease: a review of 746patients. J Neurosurg Anesthesiol 2006; 18:73–7.

183


57 Perioperative diabetes insipidusJuan P. Cata and Ehab Farag

Sodium disturbances are common in patients present-ing with neurologic disease. However, postoperativedevelopment of sodium dysregulation may be seenafter transphenoidal surgery.

Case descriptionA 45-year-old male with a 1-year history of right tem-poral anopsia and impotence underwent a success-ful transphenoidal resection of a 21 mm (in maximaldiameter) suprasellar mass (meningioma) under gen-eral anesthesia. Intraoperative management was unre-markablewith an estimated blood loss of 125mL.Aftersurgery and successful extubation of the trachea, theneurologic exam was unremarkable and visual fieldswere grossly intact. He was then transferred to thepostanesthesia recovery unit and 8 hours later trans-ferred to the general medical ward. On postoperativeday 1, he started complaining of thirst, his urinaryoutput increased to 300 mL/hour and reached a totalvolume of 7.8 liters on postoperative day 1. A lab-oratory analysis was remarkable for urinary specificgravity of 1.000 and serum glucose of 161 g/dL. Dueto the clinical suspicion of central diabetes insipidus(DI), he was treated conservatively with liberal accessto oral intake of water and intravenous fluids. Onpostoperative day 2, he developed signs of dehydra-tion and was hypernatremic, therefore 0.1 mg of oraldesmopressin was given after which his urinary outputdecreased to 90mL/hour and the urinary specific grav-ity increased to 1.010. On postoperative day 3, he wasasymptomatic, the serum sodium level was 138mEq/Land serum osmolality was 300 mOsm/kg. The urinaryoutput remained stable. He was discharged on postop-erative day 4.

DiscussionDisorders of sodium after transphenoidal surgeryoccur due to the irritation or destruction of the pitu-

itary stalk [1]. The two main disorders are: (1) syn-drome of inappropriate antidiuretic hormone secre-tion, and (2) central DI.The first is caused by inappro-priate secretion of vasopressin and the second, by thelack of release of the hormone. It is necessary to differ-entiate central DI (caused by insufficient production ofvasopressin) from nephrogenic DI, which is caused byan impaired response of the kidneys to vasopressin. Inboth, there is a failure to concentrate urine [2].

Perioperative central DI is a common finding inpatients undergoing pituitary surgery. Preoperative DIcan be part of a panhypopituitarism syndrome inpatients with large pituitary prolactinomas or non-prolactinomas [3]. However, central DI develops mostcommonly in the postoperative setting. The incidenceof transient postoperative central DI ranges from 5%to 80% of those undergoing pituitary surgery [3–5].However, fewer patients develop permanent central DI[3]. A third form of postoperative central DI called“triphasic” has also been described. The first phaseis characterized by thirst and polyuria and is presentwithin 24 to 72 hours after surgery. The second phasedevelops 7–10 days after surgery and is characterizedby antidiuresis. Two to 3 weeks after surgery, polyuriareturns and marks the onset of the third phase. Pre-dictive factors of central DI after pituitary surgery areresection of a Rathke’s cleft cyst and intraoperativecerebrospinal fluid leak. However, conflicting resultsexist between the incidence of central DI and pituitaryadenoma size [5].

The clinical manifestation of central DI rangesfrom mild thirst to significant polyuria, polydipsia,dehydration and, if severe, hypotension. Commonly,the onset of signs and symptoms of DI starts on post-operative day 1 to 2 and may last for 1 or 2 weeks.A urine output �250 mL/hour for 2 consecutivehours is a clinical indicator of DI. However, DI isdiagnosed when all the following criteria are met:polyuria of �2.5 L, a day with concomitant poly-dipsia, serum Na �140 mmol/L, spontaneous urine

184

Case 57. Perioperative diabetes insipidus

specific gravity �1.005, and glycemia �180 mg/dL[4].

The clinical management of patients with cen-tral DI after pituitary surgery is initially conserva-tive. Most patients undergoing pituitary surgery areawake and cooperative after surgery, which facili-tates the treatment of central DI. Monitoring of dailyweights, fluid balance, and the oral intake of waterare fundamental during the initial management ofthese patients. The use of exogenous vasopressin isindicated in patients with significant urinary outputand hypernatremia. Vasopressin can be administeredintravenously, intranasally, subcutaneously, and orally.Several authors recommend a daily oral dose of 0.1 mgof desmopressin as the first choice. Serum sodiumhas to be followed during vasopressin administra-tion because the excessive retention of free water canlead to hyponatremia. See Case 83 for further dis-cussion of sodium abnormalities in the neurosurgicalpopulation.

References1. I. Ciric, A. Ragin, C. Baumgartner et al.

Complications of transsphenoidal surgery: results of anational survey, review of the literature, and personalexperience. Neurosurgery 1997; 40: 225–36.

2. G. L. Robertson. Diabetes insipidus. Endocrinol MetabClin North Am 1995; 24: 549–72.

3. N. Fatemi, J. R. Dusick, C. Mattozo et al. Pituitaryhormonal loss and recovery after transsphenoidaladenoma removal. Neurosurgery 2008; 63: 709–18.

4. R. A. Kristof,M. Rother, G. Neuloh et al. Incidence,clinical manifestations, and course of water andelectrolyte metabolism disturbances followingtranssphenoidal pituitary adenoma surgery: aprospective observational study. J Neurosurg 2009;111: 555–62.

5. D. G. Sigounas, J. L. Sharpless,D. M. Cheng et al.Predictors and incidence of central diabetes insipidusafter endoscopic pituitary surgery. Neurosurgery 2008;62: 71–8.

185

Part VI Pediatric neuroanesthesiaCase

58 CraniotomyWilliam Bingaman and Marco Maurtua

Pediatric surgical patients present special anestheticchallenges including induction without intravenous(IV) access, a higher incidence of airway complications(laryngospasm and bronchospasm), and a greater inci-dence of hemodynamic instability due to surgicalblood loss [1].

Brain tumors are the second most common causeof cancer in pediatric patients [2]. Between 5 and 60%of these central nervous system tumors are infratento-rial and include medulloblastoma, low-grade cerebel-lar astrocytoma, ependymoma, glioma, and low-gradebrainstem astrocytoma. Supratentorial tumors are thesecond most common location and include low-gradeastrocytoma, malignant and mixed glioma, ependy-moma, ganglioglioma, oligodendroglioma, choroidplexus tumor, and meningioma [3].

Case descriptionAn8-year-old child (120 cm tall and 24.6 kg) presentedwith partial complex seizures characterized by staringspells accompanied by oral and manual automatisms.These persisted despite adequate trials of three dif-ferent anticonvulsant medications. Preoperative eval-uation (magnetic resonance imaging) demonstrateda non-enhancing heterogeneous lesion in the mesialaspect of the right temporal lobe consistent with low-grade glial neoplasm. Right tailored temporal lobec-tomy was scheduled to obtain tissue for diagnosis,resect the lesion, and eliminate the seizures.

After the preoperative evaluation the patient waspremedicated with 10mg oral midazolam. Once in theoperating room standard ASA monitors were appliedand an inhalation induction with sevoflurane tookplace. Intravenous access was secured followed by tra-cheal intubation. A radial arterial line was also placedprior to cranial fixation with the three-point headholder. Prior to fixating the head, the patient was givena bolus of intravenous propofol to blunt the hyperten-sive response elicited by head pinning.

Maintenance of anesthesia was achieved withisoflurane 0.6%, remifentanil infusion at 0.1 mcg/kg/min and nitrous oxide 70%. Muscle relaxation wasobtained with intermittent rocuronium administra-tion that in this case was more frequent secondary tothe hepatic enzyme induction produced by anticon-vulsant medications, which were further available inthe operating room if needed. During the skin closure,fentanyl 25 mcg was administered to provide analge-sia in the immediate postoperative period. The surgi-cal procedure was performed without complicationswith 200mL estimated blood loss. A neurologic assess-ment was performed after extubation and it showedno neurologic deficits. Subsequently the patient wastransported to the pediatric intensive care unit wherehis recovery was satisfactory. The pathology wasganglioglioma.

Discussion

Preoperative assessmentIn elective cases the nothing by mouth (or Nil PerOs) guideline currently followed is the “2–4-6–8”rule, which indicates that clear fluid intake is allowed2 hours before surgery, breast milk 4 hours beforesurgery, formula 6 hours before surgery, and solid food8 hours before surgery [4].

A complete history should be performed focusingon the past medical history, family history, and pastexperiences with anesthesia if applicable. Co-morbidconditions should be accounted for includingmaternalcomplications during pregnancy, delivery and historyof prematurity and its complications. A system-basedreview including screening for cardiac, pulmonary,renal, hepatic, and neurologic disease should be per-formed. The past surgical and anesthetic history ofthe patient is relevant to identify prior anestheticcomplications such as difficulties with airwaymanage-ment, previous anesthetic drug reactions, and the type

187

VI. Pediatric neuroanesthesia

of surgical procedures performed in the past. Familyhistory of complications with anesthesia is aimed atidentification of diseases and conditions that can betransmitted genetically such as malignant hyperther-mia, Duchenne’s muscular dystrophy, etc. to preventpatient exposure to triggering agents and depolariz-ingmuscle relaxants. Allergies should be identified andmedications reviewed; precautions should be insti-tuted for conditions such as latex allergy. Anticonvul-sant medications should be continued and adminis-tered the day of surgery to prevent intraoperative andpostoperative seizures.

During the patient’s clinical assessment, oneshould start with the patient’s airway examination,including Mallampati score, thyromental distance,degree of cervical spine extension, presence ofcraniofacial abnormalities such as craniosynosto-sis or syndromes such as Crouzon’s, Pierre Robin,Goldenhar, etc. that generally are accompanied bymicrognathia, an independent predictive factor fordifficult ventilation and/or intubation. The clinicalassessment should continue by examining the patient’spulmonary status, especially in patients with gen-eralized seizures or mental retardation to diagnosepreoperatively the presence of pulmonary aspiration.The clinician should be alert to signs of dehydration,especially if the patient had a history of intermittentnausea and vomiting due to increased intracranialpressure (ICP). A thorough physical examinationshould follow. A neurologic examination shouldinclude documentation of abnormal neurologic devel-opment, presence of cognitive delay, level of alertnessand orientation. A motor and sensory exam shouldalso be performed. Any signs or symptoms of raisedICP need to be documented and discussed with thesurgical team. In newborns the presence of lethargyor irritability may indicate an increase in ICP.

Cell blood counts may be altered due to theuse of carbamazepine or valproic acid. Electrolyteabnormalities due to vomiting secondary to increasedICP may be found and should be corrected priorto the surgical procedure. In pediatric patients thehead is larger in proportion to the body surface areacompared with the adult patient. This is an impor-tant fact that leads to a greater blood loss and thirdspacing losses that can make the pediatric patienthemodynamically unstable intraoperatively. Closemonitoring of blood loss and coagulation parametersduring surgery is crucial to avoid potentially seriouscomplications. Therefore blood should be avail-

able before starting a craniotomy in newborns andinfants.

Premedication: Prior to bringing the patient tothe operating room, patients older than 1 year ofage should be premedicated with oral midazolam(0.5 mg/kg) or nasal midazolam (0.1 to 0.2 mg/kg). Ifincreased ICP is present or suspected, premedicationshould be avoided to prevent a further increase in ICPdue to sedation and hypercarbia.

Intraoperative managementInduction of anesthesia: Inhalation induction can beperformed in the childwithout IV access andwho doesnot have clinical signs of increased ICP. If increasedICP is present and the patient has been vomiting, anintravenous line should be placed and volume replace-ment initiated to treat dehydration. Clinical signs ofhypovolemia, such as decrease in skin turgor, sunkenfontanelles, crying without tears, low urine output,and lethargy, imply a decrease in intravascular vol-ume that should be corrected prior to induction ofanesthesia.

Placement of invasive monitors should take placeafter induction of anesthesia, including an arterial lineto monitor acute changes in blood pressure due toblood loss or bradycardia seen after stimulation of thedura or other central nervous system structures andalso to accurately determine PaCO2 levels. Bladdercatheterization is mandatory to monitor urine outputthat is frequently increased due to the administrationof hyperosmotic therapy. Intravenous access can beaccomplished with the placement of two large-boreperipheral IVs. If peripheral IV access cannot beachieved, placement of a central venous line shouldbe performed. Finally, if the procedure will increasethe risk for air embolism, a precordial Dopplerultrasound should be utilized for early detection ofair embolism especially in surgeries involving brainvenous sinuses.

Maintenance of anesthesia can be achieved bythe administration of inhalation anesthetics, such assevoflurane, isoflurane, and desflurane. It is impor-tant to remember that these agents produce dose-dependent cerebral vasodilation that is noticeable dur-ing surgery and a decrease in cerebral metabolic rate(CMRO2) producing an uncoupling effect. Nitrousoxide can be used during pediatric neurosurgical pro-cedures, however avoidance of this agent is indicatedwithin 1 month after a previous craniotomy since

188

Case 58. Craniotomy

pneumocephalus may have not yet been resolved.Nitrous oxide is also contraindicated in patients withpneumothorax, or after trauma of the chest andshould be used cautiously or completely avoided inthose procedures where there is an increased risk forair embolism. Similarly to the described fluorinatedinhalation anesthetics, nitrous oxide produces a dose-dependent cerebral vasodilation, but the difference isthat it increases CMRO2.

Narcotic infusions in conjunction with inhalationanesthetics are currently used in pediatric patientsfor neuroanesthesia maintenance. Remifentanil is anarcotic approximately 100 times more potent thanmorphine. It is useful in neuroanesthesia because ofits short context-sensitive half life due to its uniquemetabolismbynonspecific plasma and tissue esterases.It has been proven to be an ideal drug in pedi-atric neuroanesthesia, allowing for a faster wake upand reliable postoperative neurologic exam. Remifen-tanil’s side effects include bradycardia, hypotension,and decreased cardiac output. A general recommen-dation is to avoid remifentanil boluses and if given asan infusion the starting dose should be low andmay beincreased only if the hemodynamic parameters are sta-ble. Chest rigidity can also be seen as a complication ofany intravenous narcotic and should be considered ifincreased inspiratory pressures are suddenly detectedand a right main-stem intubation has been ruled out.Other narcotic infusions are also used in pediatric neu-roanesthesiology, such as fentanyl infusions. Due to itshigh lipid solubility, fentanyl is stored in adipose tis-sue; the longer the infusion is delivered the higher theplasma concentration will get and the greater the timerequired for the patient to return to therapeutic plasmalevels. Therefore, it is imperative to stop the fentanylinfusion at least 1 hour before emergence from anes-thesia so a thorough immediate postoperative neuro-logic examination is achievable. A common advan-tage of all narcotic infusions is that they potentiateinhalation anesthetics, decreasing their requirementand allowing a decrease in their plasma concentration.In a prospective randomized clinical trial performed inour institution, we showed that an infusion of remifen-tanil at 0.13 mcg/kg/min adds 50% minimum alve-olar concentration in adult patients when combinedwith 0.6% isoflurane [5]. This decrease in inhalationanesthetic requirements leads to a decreased cerebralvasodilation, that as we mentioned is dose dependent,offering the neurosurgeon a more relaxed, less edema-tous surgical field.

Keeping the patient immobilized during the surgi-cal procedure is essential because of the risk of headdislodgement from the head holder, injury to the brainfrom patient movement during the surgical proce-dure, and cervical spine injury from body motionwith a fixed cranium. Nondepolarizing muscle relax-ants should be used in pediatric patients undergoingelective procedures to avoid hyperkalemic responsesseen in children with undiagnosed myopathies afterthe administration of succinylcholine. Patients takingantiepileptic medications have an increased activity ofthe cytochrome P450 system, which produces rapidmetabolism of muscle relaxant drugs and decreasestheir usual half life. Therefore close monitoring ofmuscle relaxationwith amuscle twitchmonitor shouldbe routine. Neurosurgeons are often concerned withbrain swelling during surgery. Useful techniques tocombat this include hyperventilation therapy, hyper-osmotic therapy, head elevation, and reduction ofinhaled anesthetics. Hyperventilation can be per-formed in children to decrease brain volume by pro-ducing central nervous system vasoconstriction. Oneneeds to be cognizant about the theoretical risk ofproducing cerebral ischemia at PaCO2 levels below25 mmHg. Hyperosmolar therapy is useful to reducecerebral swelling. Mannitol can be administered ina dose of 0.25–1.0 g/kg body weight. Mannitol hasa biphasic effect after its intravenous administration;first there is an increase in intravascular volume due toa transient increase in intravascular osmotic effect thatdrives fluids from the interstitial space to the intravas-cular space, followed by a second effect of osmoticdiuresis producing a decrease in intravascular volumeby increasing urine output.

ConclusionApproaching the care of a child with a central ner-vous system lesion requires a systematic approach thattakes into consideration the principles of both pedi-atric anesthesiology and neuroanesthesiology.

References1. G. Olson, B. Hallen. Laryngospasm during

anaesthesia. A computer aided incidence study in136 929 patients. Acta Anaesthesiol Scand 1984: 28:567–75.

2. M. D. Walker. Diagnosis and treatment of braintumors. Pediatr Clin North Am 1976; 23: 131–46.

189


3. I. F. Pollack. Brain tumors in children. N Engl J Med1994; 331: 1500–7.

4. L. R. Ferrari, F. M. Rooney,M. A. Rockoff.Preoperative fasting practices inn pediatrics.Anesthesiology 1999; 90: 978–80.

5. M. A. Maurtua, A. Deogaonkar,M. H. Bakriet al. Dosing of remifentanil to prevent movementduring craniotomy in the absence of neuromuscularblockade. J Neurosurg Anesthesiol 2008; 20:221–5.

190


59 Ventriculoperitoneal shuntRuairi Moulding and Peter Stiles

Hydrocephalus is a common childhood disorder thatmay present with a spectrum of symptoms fromnonspecific nausea and fatigue to life-threateningintracranial hypertension. Timely surgical correctionis mandatory in emergent cases in order to reduceintracranial pressure (ICP) and decrease any sec-ondary injury, which may result from interstitial cere-bral edema.

A ventriculoperitoneal (VP) shunt – a series ofcatheters with a unidirectional valve to divert cere-brospinal fluid (CSF) from the brain by draining itinto the peritoneum – may be implanted from birthonwards as a definitive surgical correction for hydro-cephalus. Shunt failure, however, is common. A largecohort study demonstrated that 81% of shunt patientsexperience at least one episode of blockage in the first12 years following implantation. Further, as many as40%of thosewill fail within 1 year of revision [1].Oncea shunt has failed the subsequent revision also has areduced survival time.

Case descriptionA 15-month-old female with a history significant forstenosis of the aqueduct of Sylvius, epilepsy, and VPshunt placement as an infant, presented to the emer-gency department for evaluation. She was obtundedand had a 1-day history of lethargy, nausea, and vom-iting. Although she had walked unaided since the ageof 14 months, she had difficulty taking more than twosteps. Her mother also noted that disinterest and fre-quent crying developed over a 1-week interval. Phys-ical examination revealed an obtunded 15-month-old female with large head and bulging fontanelle.Pupils were equal and reactive to light. Brain com-puted tomography (CT) showed dilated ventricles(Figure 59.1) when compared with previous images,but a shunt series of plain films failed to reveal local-izable abnormality. Laboratory values were unremark-able. The patient’s temperature was 37.0 ◦C. Lumbar

Figure 59.1. Computed tomography of severely dilated ventriclesin an infant, with compressed and thinned cerebral hemispheres atthe periphery. The tip of the ventricular shunt is evident, inserted viaa left frontal approach.

puncturewas not performed given likely increased ICPand resulting risk of herniation. She was diagnosedwith recurrent hydrocephalus due to shunt malfunc-tion at an unknown location along the apparatus andpossible shunt infection. The patient required imme-diate surgical revision of her VP shunt.

Although the patient was lethargic on presenta-tion, every effort was made not to further stimulatethe child. An intravenous catheter was placed and flu-ids given. In view of the patient’s history of vomitingand her altered mental state, no sedation was givenand a modified rapid sequence induction with fen-tanyl, propofol, and rocuronium was performed. Fen-tanyl was used to obtund the laryngeal reflex in order

191


to avoid increased ICP. Endotracheal intubation viadirect laryngoscopy was confirmed and general anes-thesiawasmaintainedwith intravenous anesthesia andintermittent opioids. A loading dose of phenytoin forseizure prophylaxis was administered at 18 mg/kg [2].A Foley catheter was placed to monitor urine out-put. Intraoperative CSF sample yielded no white cells.The surgical team localized the VP shunt obstructionto the peritoneal portion by accessing the apparatus’valve and noting free flowing aspiration, but resistanceto injection. The case was without complication andtook 2 hours to complete with minimal blood loss anduneventful emergence. The patient was transferred tothe pediatric intensive care unit for overnight obser-vation then transferred to the general pediatric wardthe following day. The patient recovered neurologi-cally, regaining the ability to walk after 2 days.

DiscussionAny child with increased ICP presents a variety of con-cerns for the anesthesiology team. First, the patient isdrowsy and has been vomiting. This mandates a rapidsequence induction together with attention to anyelectrolyte abnormalities and dehydration. Second, apatient with increased ICP requires a high degree ofvigilance as any further increase in ICP and thereforereduction in cerebral perfusion pressure (CPP) maycause irreparable damage. Third, if the shunt revisionis successful, the anesthesiology and surgical teamsmustmaintain excellent communication to ensure thatICP is not lowered too quickly as this may result infurther intracranial pathology. Finally, the VP shuntpatient population is at risk for seizures and properprophylaxis must be administered.

Hydrocephalus is the accumulation of cere-brospinal fluid (CSF) within the brain. It may beacquired or congenital, communicating or noncom-municating. Communicating hydrocephalus meansthat CSF can still flow between ventricles but flow isblocked as it exits the ventricles. Noncommunicatinghydrocephalus, such as in the above patient withaqueductal stenosis, is also known as obstructivehydrocephalus. This occurs when the flow of CSF isblocked along one of the interconnecting passagesbetween the ventricles. The aqueduct of Sylvius, alsocalled the aqueduct of the midbrain, is a canal thatcommunicates between the third and fourth ventricles.

Hydrocephalus can present as a neurosurgicalemergency requiring quick action to reduce the pres-

sure. Although medical management may alleviatesome increases in ICP, definitive treatment is surgi-cal. In the above scenario, the patient exhibited aninsidious progression of symptoms.However, the pres-ence of nausea and vomiting is a concerning symp-tom and warrants closer scrutiny as it may be a signof acute decompensation. Surgical management willdepend on the cause of the ICP elevation. If there isno evidence of an infectious etiology, then a simplemechanical obstruction is likely and the shunt does notneed to be externalized.

Elevation of ICP can result in catastrophic con-sequences for neural tissue and may result in pooroutcomes for patients if prompt interventions are notinstituted. Normal values for ICP are age dependent,although no consensus exists. Elevations in ICP can bechronic or acute depending on etiology. Occasionallya chronically elevated ICP can change rapidly to life-threatening acute intracranial hypertension. In anypatient with elevated ICP, it is important to control anyfurther increase to reduce the risk of secondary dam-age from interstitial cerebral edema as well as stoppingthe progression from uncontrolled intracranial hyper-tension to a herniation syndrome and death.

Control of ICP requires avoidance of hyper-carbia, hypoxemia, acidemia, hyperthermia, cerebraledema, coughing, airway obstruction, obstruction tothe venous drainage of the head, and high metabolicdemand states such as seizures. Medical interven-tions that reduce ICP include head elevation to facil-itate venous drainage, hyperventilation (after intuba-tion), and hyperosmolar therapy. These measures mayneed to be employed in patients prior to definitivesurgery.

Cerebral perfusion pressure (CPP) is the pressuregradient driving cerebral blood flow. Cerebral perfu-sion pressure is derived from mean arterial pressure(MAP) minus ICP:

CPP = MAP − ICP

If central venous pressure (CVP) is higher thanICP then it directly affects CPP. Reduction in CPPdecreases the supply of metabolic substrate andoxygen to the brain, potentially resulting in ischemicdamage. It is therefore especially important to main-tain a MAP appropriate to adequate CPP in thesetting of increased ICP. Mean arterial pressure maybe increased using intravenous fluid resuscitationor vasopressors. It is worth noting that excess intra-venous fluid may be detrimental, potentially causing

192

Case 59. Ventriculoperitoneal shunt

Figure 59.2. X-ray of infant with blocked VP shunt showing thepathway from the neck, over the right hemithorax, with redundanttubing in the abdomen. No evidence of kinking or tube disruption isevident.

tissue edema. Therefore, extreme care should alwaysbe taken when instituting intravenous fluids in anypatients with raised ICP.

Ventriculoperitoneal shunts can fail to performtheir intended function for a variety of reasons. Themost common etiology of shunt failure in first-timerevisions is obstruction, accounting for 66.7% of allinstances [3]. Ventriculoperitoneal shunt blockagemay be caused by occlusion of the ventricular catheterwith choroid plexus, blood, or even brain tissue. Alter-natively, the peritoneal catheter may be occluded byclot or scar tissue. Finally, the unidirectional valve,responsible for regulating flow, may malfunction andobstruct CSF drainage (Figure 59.2). Other, less com-mon causes of shunt failure include infection (19%),loculated ventricles (7.7%), and overdrainage (5.7%)[3].

Patients undergoing emergent correction of acutehydrocephalus require a high degree of vigilance. Inva-sive intra-arterial blood pressuremeasurement is oftenunnecessary unless the patient is unstable or has sig-nificant co-morbidities. The use of central venouscatheters is rarely indicated. Duration of surgery maybe difficult to judge, therefore it is prudent to assumeall such patients will need urinary catheters and atten-tion to positioning.

Increased ICP can lead to nausea and vomiting.If a patient presents with profuse vomiting, he or shemay display changes in biochemistry and suffer dehy-dration. These fluid and electrolyte abnormalities mayneed to be corrected perioperatively. Nausea and vom-iting also exposes the patient to risk of aspiration dur-ing induction of anesthesia and therefore mandates arapid securing of the patient’s airway with an endotra-cheal tube [4].

Although perioperative anticonvulsants are notalways indicated, elevated ICP predisposes to seizuresand the associated increase in cerebral metabolicdemand may be catastrophic. Additionally, if a patientis already on seizure prophylaxis they may have sub-therapeutic plasma levels due to the nausea and vom-iting and therefore further intravenous dosing maybe required. It should be noted that some anticonvul-sant medications are hepatic enzyme inducers, mean-ing that the metabolism and clearance of some anes-thetic drugs is increased. Perioperative antibiotics maybe indicated as prophylaxis. Occasionally the surgeonmay ask for antibiotics to be withheld until a CSF sam-ple has been taken for analysis if an infectious etiologyof shunt failure is suspected.

Some patients with indwelling VP shunts maypresent for non-neurologic abdominal surgery, forexample, appendectomy. There is no consensus onwhat measures should be taken in these instancesalthough case series suggest that if an abdominal sur-gical site is infected then the VP shunt apparatus mayneed to be externalized temporarily until that localinfection has cleared. These case series suggest VPshunts may fail if subjected to infectious assaults [5].

The hydrocephalus exhibited by patients requir-ing VP shunts may be part of a syndrome or diseaseprocess and a thorough history and physical exami-nation may elicit signs and symptoms of other organdysfunction.

Hydrocephalus is a common childhood disorderand surgical correction utilizes a shunt system to divertexcess CSF out of the brain. If the drainage system

193


obstructs then the patient may suffer from recurrenceof their hydrocephalus with accompanying symptomsand signs. If timely surgical intervention is not under-taken, intracranial hypertension can lead tomore seri-ous consequences such as herniation syndrome anddeath. The obtunded child with acute hydrocephalusrequires careful preoperative assessment andmonitor-ing in order to formulate an appropriate anestheticplan to avoid further rises in ICP.

References1. C. Sainte-Rose, J. H. Piatt,D. Renier et al.Mechanical

complications in shunts. Pediatr Neurosurg 1991; 17:2–9.

2. P. Marik, K. Chen, J. Varon et al.Management ofincreased intracranial pressure: a review for clinicians.J Emerg Med 1999; 17: 711–19.

3. S. Tuli, J. Drake, J. Lawless et al. Risk factors forrepeated cerebrospinal shunt failures in pediatricpatients with hydrocephalus. J Neurosurg 2000; 92:31–8.

4. R. K. Hamid, P. Newfield. Pediatric neuroanesthesia.Hydrocephalus. Anesthesiol Clin North America 2001;19: 207–18.

5. G. Li, S. Dutta. Perioperative management ofventriculoperitoneal shunts during abdominalsurgery. Surg Neurol 2008; 70: 492–5.

194


60 Craniosynostosis repairVera Borzova and Julie Niezgoda

Craniosynostosis is a common multifactorial congen-ital anomaly affecting approximately 1 in 2000 births.It is the result of premature fusion of cranial sutureswith the development of skull deformity that may neg-atively affect future brain development. Surgical pro-cedures for craniofacial malformations are one of themost challenging cases in pediatric anesthesia.The dif-ficulties arise in part due to the well recognized andfrequently unavoidable complications of this type ofsurgery. Major considerations with regard to the man-agement of these patients are (1) patient size: early sur-gical correction of craniosynostosis is recommendedto avoid development of intracranial hypertension dueto restricted skull growth and to potentially improveneurocognitive function of the brain; (2) managementof the airway in patients with mid-facial hypoplasia;(3) extensive blood loss that coincides with the time ofthe physiologic nadir of the hematocrit in the new-born; (4) potential for development of coagulopathyrelated tomassive blood transfusion; (5) risk of venousair embolism; (6) prolonged surgery and hypother-mia; and (7) possibility of raised intracranial pres-sure when more than one suture is involved, or in theolder child.

Case descriptionA 3-month-old male who had an uncomplicated termdelivery presented for repair of craniosynostosis.His past medical history was significant for Crouzonsyndrome (see Table 60.1). Magnetic resonanceimaging (MRI) of the brain revealed fusion of thecoronal sutures andmid-face hypoplasia. Preoperativelaboratory evaluation was remarkable for a hematocritof 30%.

Premedication was avoided and a peripheral intra-venous catheter was started because of the youngage, presence of mid-facial hypoplasia, and con-cern regarding potential problems with ventilationand intubation. The infant received glycopyrrolate

as an antisialagogue and sympathomimetic effect inthe event of bradycardia with airway manipulation.Propofol was titrated tomaintain spontaneous respira-tion during induction.The airway was secured nasallywith the assistance of a flexible fiberoptic broncho-scope. In addition to standard ASA monitors, an arte-rial line and central line were placed to assist withblood pressure monitoring as well as fluid and elec-trolyte management.

Typed and crossed packed red blood cells werepresent in the operating room at the start of theskin incision. Anesthesia was maintained with isoflu-rane, remifentanil and vecuronium infusion. Esti-mated blood loss was equal to 50% of the total bloodvolume and the infant was transfused to maintain ahematocrit �27%. Upon completion of the surgerythe infant was warm and hemodynamically stable.Thetrachea was left intubated due to concern about fluidshifts and airway edema in the presence of difficultairway. He was transferred to pediatric intensive carefor recovery where he was extubated successfully a fewhours later.

DiscussionCraniosynostosis is a congenital deformity of the skullrelated to premature closure of one or more cranialsutures. Craniosynostosis can be a primary anomaly(80%) or can be associated with a syndrome, mostcommonly Apert, Crouzon, and Pfeiffer. Isolated cran-iosynostosis typically affects only one suture, usuallysagittal, which is also called scaphocephaly. Syndromiccraniosynostosis often affects two or more sutures andcan have other congenital anomalies, most noticeablymid-facial hypoplasia and obstructive sleep apnea,as in patients with Apert and Crouzon syndromes.Despite some uncertainty about the impact of earlycorrection of craniosynostosis on future brain growthand cognitive development, it is current practice toperform surgery in early infancy.

195


Table 60.1. Most common syndromes associated with craniosynostosis.

Syndrome Airway Cerebral Cardiac Musculoskeletal

Apert Maxillary hypoplasia, narrowpalate, +/− cleft palate, difficultairway

Craniosynostosis, hydrocephalus CHD syndactyly

Crouzon Maxillary hypoplasia with inverted,V-shaped palate, large tongue,difficult airway

Craniosynostosis, shallow orbitswith proptosis

Pfeiffer Choanal atresia, laryngo-, tracheo-,and bronchomalacia

Craniosynostosis, hydrocephalus,Arnold–Chiari malformation

Syndactyly, broad thumbs andtoes

CHD, Congenital heart disease.

In addition to routine preoperative evaluationincluding history and physical examination with par-ticular attention to airway and any other associatedanomalies, laboratory evaluation includes hematocrit,platelet count, coagulation profile and type/cross forpacked red blood cells equal to one estimated bloodvolume.The potential need for postoperative mechan-ical ventilation, intensive care unit admission andblood transfusion should be discussed with parents.

Intraoperativemonitoring should include standardASA monitors along with arterial line for preciseblood pressure monitoring and frequent blood sam-pling. Intravenous access sufficient for rapid bloodand fluid replacement is mandatory. Blood should bechecked and available prior to skin incision. Two large-bore intravenous catheters (22–20 gauge in the infant)are often adequate but central venous access may bedesired if large volume blood loss is anticipated asin a multi-suture correction or the risk of venous airembolism is high. Routine use of precordial Doppleris recommended for early detection of air emboli. Abladder catheter will allow monitoring of sufficienturine output.

Most infants with primary craniosynostosis arehealthy and their airway can be secured with directlaryngoscopy in an anesthetized state after inhalationinduction. However, all necessary equipment to dealwith the difficult airway should be immediately avail-able especially in an infant with syndromic craniosyn-ostosis. Spontaneous ventilation should bemaintainedwhenever there is a concern about the likelihood ofsuccessful ventilation or intubation until the airway issecured. Meticulous attention should be paid to theendotracheal tube during positioning of the patient forsurgery. Occasionally nasal or oral endotracheal tubesmay be sutured or wired to avoid displacement duringextreme neck manipulation. In the patient with a tra-

cheotomy, the stoma is intubatedwith an armored tubeand sutured in place.

The small size of infants presenting for craniosyn-ostosis repair translates into low absolute blood vol-umes, such that a small amount of blood loss canresult in significant hemodynamic compromise. Car-diac arrests were described due to sudden massiveblood loss from sinus or extradural venous tears. Slowblood loss from bone edges and scalp incisions is oftendifficult to quantify secondary to a significant por-tion being absorbed into surgical drapes, gowns, andthe plastic reservoir that is used to drain irrigationand blood from the surgical field. Because evaluationof blood loss is difficult intraoperatively, the anesthe-siologist must pay meticulous attention to the mon-itoring of intravascular status via the arterial pres-sure, central venous pressure, urine output, as wellas serial determination of acid-base status, hemat-ocrit, platelet count and coagulation profile. The anes-thesiologist and the surgeon need to be in ongoingcommunication.

Probably the most challenging part of the anes-thetic management of craniosynostosis repair is thesignificant blood loss and frequent rate of bloodproduct transfusion [1]. Various attempts have beenmade to reduce the need for allogenic blood includ-ing preoperative administration of erythropoietin andiron, autologous blood donations, acute normov-olemic hemodilution, intraoperative and postopera-tive blood salvaging, use of less invasive surgical pro-cedures, and combinations of the above. To date thereis no “ideal” technique that can guarantee freedom ofallogenic blood transfusion. Different techniques can,however, play a role in decreasing and in some patientseliminating the need for blood transfusion.

Local infiltration of the skin with epinephrine isused by surgeons to decrease blood loss. However,

196

Case 60. Craniosynostosis repair

most of the blood loss is from the periosteum andbone. Human recombinant erythropoietin given at adose of 600 U/kg subcutaneously for 3 consecutiveweeks in combination with supplemental iron signifi-cantly increased preoperative hematocrit from ameanof 24% to a mean of 38% in 41 infants with a meanage of 5.7 months [2]. The same authors showed thatcombination of erythropoietin with intraoperative useof cell saver resulted in lower transfusion rates (5%vs 100% control) and less amount of blood transfusedthan in the control group (0.05 pediatric units vs 1.74pediatric units). In both groups hypervolemic hemod-ilution using an additional 30–50 mL/kg of colloid orcrystalloid solution was used as needed [2]. Althoughcell saver appears to be an attractive option, its usedoes not completely eliminate the need for transfusion,since the transfusion is often started early in the pro-cedure to maintain hemodynamic stability, before sal-vaged blood is available. It can, however, decrease useof allogenic blood postoperatively.

Preoperative autologous blood donation has beendescribed but has been considered to be impracticalespecially in children less than 3 years of age due tosmaller blood volumes and intolerance to repeatedvascular access for the donation process. Acute nor-movolemic hemodilution is another technique wherewhole blood is exchanged for an equal amount of col-loid or crystalloid solution tomaintain normovolemia.As a result the volume of blood lost during the surgerystays the same but the amount of red blood cell masslost is less due to a lower hematocrit of the blood [3].This technique may be beneficial in infants with rareblood types.

Induced hypotension was proposed as a meansto decrease intraoperative blood loss. It has notgained acceptance for craniosynostosis repair. Rea-sons cited include compromised cerebral perfusionespecially in the presence of elevated intracranialpressure, lowering mean arterial and central venouspressure increases the risk for air emboli and thepotential for added hemodynamic instability dur-ing rapid blood loss [4]. There are no guidelinesthat specifically address blood transfusion practice inchildren.

The American Society of Anesthesiologists TaskForce on Blood Component Therapy excluded infantsand children in their practice guidelines in 1996.The British Committee for Standards in Hematol-ogy Transfusion [5] recommended the following forneonates and older children:

� Cytomegalovirus-negative blood should be usedduring first year of life.

� All components should be leukocyte reduced.� A screen filter should be used for transfusion of all

blood components.� Old packed red blood cells (greater than 2–3

weeks) should be avoided whenever possible.

The amount of blood givenwill depend on the clin-ical circumstances (estimated blood loss, presence ofhemodynamic instability despite maintenance of euv-olemia with crystalloid or colloid solution, extent ofthe surgery, presence of lactic acidosis as a marker ofcompromised tissue perfusion). One can calculate theamount of blood to be transfused using desired, actualhemoglobin (Hb) and patient’s weight: Hb desired(g/dL) – Hb actual (g/dL) × weight (kg) × 3. Sincepostoperative oozing is common and can lead to sig-nificant blood loss, one may consider transfusing thewhole unit to which the patient has been exposed inanticipation of postoperative blood loss in order tominimize exposure to additional units.

A more practical option is to have 20 mL/kgaliquots separated in order to extend the expirationtime from4 to 24 hours after it has been spiked.Hyper-kalemic cardiac arrest was described in infants whoreceived large amounts of blood rapidly due to highconcentration of potassium in stored blood. Calciumhomeostasis has to be maintained especially whenlarge volumes of citrate-containing blood productsare given. Acquired coagulopathy is rarely a problemunless blood loss has reached greater than 1.5 timesor more of the estimated blood volume. Platelets, freshfrozen plasma, and cryoprecipitate may be necessaryin the presence of microvascular bleeding or whenlaboratory studies confirm coagulopathy due to com-ponent deficiency. Keep in mind that activated par-tial thromboplastin time is often prolonged in the first6 months of age at baseline.

Venous air embolism has been reported duringcraniofacial surgery. A high index of suspicion shouldbe maintained particularly during any phase of acuteblood loss with reduction in central venous pressure.Hydration and keeping up with ongoing blood lossis critical to limit the gradient between the surgicalfield and the right atrium. Although transesophagealechocardiography is the most sensitive monitor fordetection of emboli, its use is limited in small infants.As previously stated, the precordial Doppler, continu-ous monitoring of end-tidal CO2 and nitrogen, central

197


venous and arterial pressure waveforms may be usefulin detecting significant venous air emboli.

Children with complex craniosynostosis have ahigher incidence of increased intracranial pressurethan those with simple craniosynostosis. Cranial vaultreconstruction increases intracranial capacity andreduces intracranial pressure. Techniques to dec-rease elevated intracranial pressure intraoperativelyinclude hyperventilation to PaCO2 of 25–30 mmHg,cerebral dehydration with mannitol 0.25–1 g/kg andfurosemide 0.5–1 mg/kg, dexamethasone 0.5 mg/kg,reverse Trendelenburg position to improve venousdrainage, avoidance of extreme flexion or extension ofthe neck and occasionally placement of subarachnoiddrain.

Maintaining normothermia is very important dur-ing surgery. Large surface-to-weight ratio, increasedmetabolic rate, lack of significant body fat for insu-lation as well as large surface area of the head thatis exposed to heat loss, coupled with the need forinfusion of large volumes of intravenous fluids andblood products, place infants at risk for hypothermia.Techniques for maintaining the patient’s temperatureinclude prewarming the operating room, warming flu-ids, and using convective air warmers.

ConclusionCraniosynostosis repair presents a number of chal-lenges to the anesthesiologist: (1) small size of thepatients; (2) significant and often unavoidable blood

loss; (3) need for intraoperative transfusion of bloodproducts; (4) risk of venous air emboli, especiallyduring hypotensive episodes; (4) tendency to develophypothermia; and (5) associated anomalies includingairway problems and obstructive sleep apnea. All ofthese potential complications call for careful preopera-tive and intraoperative planning, meticulous attentionto intravascular volume status and hemodynamic sta-bility as well as maintenance of normothermia. Antic-ipation of the above difficulties along with the treat-ment plan should be discussed with the surgeon andany anesthesia provider participating in the manage-ment of the patient.

References1. G. Tuncbilek, I. Vargel, A. Erdem et al. Blood loss and

transfusion rates during repair of craniofacialdeformities. J Craniofac Surg 2005; 16: 59–62.

2. K. Krajewski, R. Ashley, N. Pung et al. Successfulblood conservation during craniosynostotic correctionwith dual therapy using procrit and cell saver.J Craniofac Surg 2008; 19: 101–5.

3. C. Di Rocco, G. Tamburrini,D. Pietrini. Bloodsparing in craniosynostosis surgery. Semin PediatrNeurol 2004; 11: 278–87.

4. J. Koh,H. Gries. Perioperative management ofpediatric patients with craniosynostosis. AnesthesiolClin 2007; 25: 465–81.

5. B. E. Gibson, A. Todd, I. Roberts. Transfusionguidelines for neonates and older children. Br JHaematol 2004; 12: 433–53.

198


61 ScoliosisGeorge N. Youssef

Scoliosis is a complex deformity of the spine with lat-eral curvature and rotation of the thoracolumbar ver-tebrae leading to rib cage deformity.There are differentcauses for scoliosis (Table 61.1). Adolescent idiopathicscoliosis (AIS) is the commonest form of scoliosis andit affects 1–3% of children aged 10–16 years [1].

Measuring Cobb’s angle has been the gold standardfor quantification of the severity of scoliosis. A diagno-sis ismadewhen the curvature is 10% ormore.There isa female predominance when considering curves �30degrees, with some authors estimating the female tomale ratio to be 8:1. Clinicians and patients base theirtreatment decisions on the risk of curve progression.For curves �50 degrees at skeletal maturity progres-sion usually ceases. For curves �60 degrees progres-sion is commonand can compromise pulmonary func-tion; curves �90 degrees are usually associated withsignificant reduction in pulmonary function.

The goals for surgical treatment are to prevent pro-gression, improve alignment and balance, and to avoidnegative outcomes of the natural history of the diseasewithout introducing iatrogenic complications.

Case descriptionA 16-year-old female presented for posterior spinalinstrumentation and fusion from T4–L2. She had ahistory of idiopathic scoliosis that started when shewas 11 years old. The scoliosis curve was estimated at60 degrees – she was otherwise healthy and extremelyscared of needles. She was 56 kg with a hematocrit of40; the patient fainted during the blood draw for thepreoperative blood work.

Risks, benefits, and alternatives of the anesthetictogether with the personnel involved were discussedwith the patient and her family.The plan for inductionof anesthesia included invasive monitors, blood con-servation, neurophysiologic monitoring with the pos-sibility of performing a “wake up test.” The possibilityof postoperative facial swelling and the remote chance

Table 61.1. Etiologic classification of scoliosis.

Idiopathic Infantile: less than 4 yearsJuvenile: 4–10 yearsAdolescent: older than 10 years until

skeletal maturity

Congenital Failure of formation, e.g., hemivertebraeFailure of segmentationMixed

Neuromuscular Cerebral palsyMuscular dystrophyMyelomeningoceleSpinal muscular atrophyFriedreich’s ataxiaCharcot Marie Tooth disease

Vertebral disease TumorInfection or metabolic bone disease

Spinal corddisease

TumorSyringomyelia

Disease associated NeurofibromatosisMarfan syndromeConnective tissue disorder, e.g.,Ehlers–Danlos syndrome

of postoperative mechanical ventilation were also dis-cussed. The patient was premedicated with 20 mg oralmidazolam 30 minutes before the surgery.

After the application of ASA standard moni-tors in the operating room inhalational induction ofanesthesia using nitrous oxide/oxygen and sevoflu-rane was performed. Two large-bore peripheral intra-venous lines were started and 2 mcg/kg of fentanyl +0.5 mg/kg of rocuronium were given to facilitate intu-bation. After the endotracheal tube was secured, anasogastric tube was placed and taped to the nose,an esophageal temperature probe/stethoscope wasinserted and a soft bite block was placed. An arte-rial line was placed and isovolemic hemodilution wasstarted where two units of blood were obtained fromthe arterial line and replaced with 5% albumin at aratio of 1:1.

Meanwhile, the neurophysiology technicianworked on the placement of the monitoring leads/

199


needles and the operating room nurse catheterized thepatient’s bladder. The surgical team was then calledto the operating room to help with positioning thepatient on the surgical table. After proper positioninga forced air warming blanket was placed on thelower body before the patient was draped for surgery.Baseline somatosensory evoked potentials (SSEPs)and motor evoked potentials (MEPs) were obtainedand anesthetic was maintained with �0.5 minimumalveolar concentration (MAC) of isoflurane with50% nitrous oxide in oxygen; a remifentanil infusionwas titrated to keep mean arterial pressure between60–75 mmHg.

During the procedure periodic inspection of thepositioning of the patient’s face and arms was per-formed; blood work including arterial blood gases,electrolytes, hemoglobin, hematocrit, and lactate wereobtained. Lactated Ringer’s solution was used for fluidmaintenance and replacement in addition to periodiccell saver blood. The patient continued to be hemody-namically stable despite the continuous ooze of bloodthroughout the procedure. By the time all the pedi-cle screws and rods were placed and the surgeon wasready for spine distraction, the hematocrit was 20 andthe mean arterial pressure (MAP) was 59 mmHg. Wetransfused the patient with her own blood in additionto cell saver blood, both of which brought the MAP to75 mmHg and the hematocrit to 26.

Somatosensory evoked potentials and MEPs wereintact throughout the procedure. Before wound clo-sure, two epidural catheters were placed by the surgeonfor postoperative pain management. The patient wasgiven 20 mcg/kg of hydromorphone IV with prepara-tion forwake up andother anestheticswere tapered off.After the dressing was applied, the patient was turnedto supine position on the hospital bed. On waking upthe patient was asked to wiggle her toes and bend herlegs before the endotracheal tube was pulled out andepidural catheters were loaded with 0.2 mL/kg of 0.2%ropivacaine with epinephrine 1:200 000 divided overthe two catheters.

DiscussionThe most common form of scoliosis encountered isadolescent idiopathic scoliosis followed by neuromus-cular scoliosis and their management can be quite dif-ferent (Table 61.2). Despite being healthy otherwise,AIS patients pose a number of different challengesto the anesthesiologist in all stages of their manage-

ment. Psychological preparation of the patients andtheir families is of the utmost importance, as well asthorough preoperative teaching and assurance to alle-viate their anxiety. It is crucial to assess the develop-mental level of each patient and to individualize theanesthetic plan according to each patient’s needs. Pre-operative sedation with oral midazolam was chosenwith our case as was inhalational induction given thepatient’s high anxiety level and her needle phobia. Formoremature patients, intravenous inductionwould bepreferred. Isovolemic hemodilution was performed inaddition to cell salvage to help avoid allogenic bloodtransfusion. The option of preoperative blood dona-tion should be considered and discussed with thepatient before scheduling the surgery, although this isfraught with its own set of problems.

Great care has to be given to patient positioningby the surgical and anesthesia teams. Care has to begiven to the positions of the arms which are normallyflexed at the elbow and parallel with the head to avoidinjury to the brachial plexus. All the pressure pointshave to be well padded. The abdomen has to be free toavoid increased intra-abdominal pressure, which canlead to increased surgical bleeding. Eyes have to beclear with no compression from the padding to avoidincreased intraocular pressure leading to postopera-tive visual loss. Visual loss after spine surgery is a rarebut devastating complication with incidence of 0.28%after surgery for scoliosis repair. Pediatric patients areat increased risk for development of nonischemic opticneuropathy with hypertension being a risk factor [2].Slight head-up reverse Trendelenburg position is usu-ally helpful to decrease facial edema by the end of theprocedure and might decrease the chances of opticneuropathy.

The main determinant of anesthetic maintenancechoice is compatibility with neurophysiologic moni-toring. Usually both SSEPs and MEPs are monitored.The plan has to be agreed upon by the anesthesiol-ogist and the neurophysiologist. We chose low-doseisoflurane/nitrous oxide plus remifentanil infusion.Remifentanil is the easiest narcotic to titrate and pro-vides the fastest wake up in the event a wake up testis needed. However, there are many anesthetic alter-natives. Propofol is shown to better preserve corticalSSEP and provide a deeper level of hypnosis comparedwith low-dose isoflurane/N2O or low-dose isoflu-rane alone but with the risk of delayed emergence.Sevoflurane produces a faster decrease and recoveryof SSEP amplitude as well as a better conscious state

200

Case 61. Scoliosis

Table 61.2. Clinical consideration for anesthetic management of idiopathic versus neuromuscular scoliosis.

Idiopathic Neuromuscular

General healthHeartLungs

Usually healthyUsually normalUsually normal – unless severe curvature

Usually associated with other disease states andmalnutrition

Possible cardiomyopathy or cor pulmonalePossible severe restrictive lung disease

Preoperative testing CBC, coagulation studies, Blood type andcross match

PFT only if severe curve or Reactive airwaydisease

As idiopathic + Comprehensive metabolic panel +albumin, CXR, EKG, echocardiography, PFT ifpatient is cooperative

Induction Usually intravenous + propofol occasionallyinhalation if scared of needles

Usually inhalational unless risk of MH

Lines/invasive monitoring Usually 2 PIVs + arterial line Central venous access +arterial line +PIV

Neurophysiologic monitoring SSEP+MEP SSEP+MEP unless seizure disorder

Anesthetic maintenance Narcotic infusion + propofol infusion orlow-dose inhalational agent

Narcotic infusion + low-dose inhalational agentor TIVA with dexmedetomidine or ketamine ifassociated with cardiomyopathy or risk of MH

Estimated blood loss (EBL) Less EBL per segmentBlood conservation techniques more

successful

Higher EBLAllogenic blood usually neededAminocaproic acid shown to decrease EBL

Number of segments to befused

Usually less than 12 Commonly thoracic and lumbar vertebrae intothe pelvic

Emergence Usually extubated by the end Higher possibility of post operative mechanicalventilation

ICU stay Short, usually one night Longer, usually a few days

Postoperative painmanagement

IV PCA or PCEA Narcotic infusion or epidural infusion

Risk of infection Minimal Higher rate of infection

ICU, intensive care unit; CBC, complete blood count; PFT, pulmonary function tests; PIV, peripheral intravenous line; SSEP, somatosensoryevoked potential; MEP, motor evoked potential; IV PCA, intravenous patient controlled analgesia; PCEA, patient controlled epidural analge-sia; CXR, chest X-ray; EKG, electrocardiogram; MH, malignant hyperthermia; TIVA, total intravenous anesthesia.

on emergence than propofol. The use of dexmedeto-midine infusion was shown to reduce propofol infu-sion requirements with potential to facilitate fasterawakening compared with propofol alone [3]. Musclerelaxants are either avoided completely or an infusionis titrated to keep two twitches on the train-of-fourmonitor – this will have the benefit of decreasing thebackgroundnoise andmight improve the quality of theSSEP.

At the time of spine distraction care has to be givento optimize the patient’s hemodynamics and oxygencarrying capacity to prevent ischemia to the cord.Some clinicians try to keep the MAP above 90 mmHgduring this critical part of the procedure.

Epidural analgesia was chosen because it wasproven to provide better analgesia and allows forquicker return to consumption of solid foodwith fewer

side effects [4]. Other options include: (1) intrathe-cal morphine, which provides excellent pain controlbut is limited to �24 hours; (2) intravenous nar-cotic infusion, which is associatedwith increased seda-tion, respiratory depression, delayed return of bowelfunction, and increased risk of nausea and vomiting;(3) intravenous patient controlled analgesia providesgood pain control during the day with fewer sideeffects but with increased pain whenever the patientdrifts to sleep; and (4) intravenous patient-controlledanalgesia with a basal narcotic infusion provides betterpain control but with increased side effects.

Despite modern technology, scoliosis still carries asmall but grave risk of mortality and morbidity. Thekey for an uneventful anesthetic is proper planningand knowledge of potential complications in order toavoid them.

201


References1. J. A. Janicki, B. Alman. Scoliosis: review of diagnosis

and treatment. Paediatr Child Health 2007; 12: 771–6.2. C. G. Patil, E. M. Lad, S. P. Lad et al. Visual loss after

spine surgery. Spine 2008; 33: 1491–6.3. N. E. Ngwenyama, J. Anderson,D. G.

Hoernschemeyer et al. Effects of dexmedetomidine on

propofol and remifentanil infusion rates duringtotal intravenous anesthesia for spine surgeryin adolescents. Paediatr Anaesth 2008; 18:1190–5.

4. T. A. Milbrandt,M. Singhal, C. Minter et al. Acomparison of three methods of pain control forposterior spinal fusions in adolescent idiopathicscoliosis. Spine 2009; 34: 1499–503.

202


62 Hemispherectomy for treatment ofintractable epilepsy in an infant withcongenital antithrombin III deficiencyRami Karroum and Alina Bodas

Perinatal cerebral artery occlusion is responsible forischemic cerebral infarction and can lead to brain cav-itation and gliosis. The etiology of perinatal ischemicaccidents remains uncertain in most cases. Only in aminority of cases can a causative factor be identified,but one cause is congenital antithrombin III (ATIII)deficiency. The territory of the middle cerebral arteryis most frequently involved and epilepsy can be anassociated co-morbidity. Most of these epilepsy casescan be managed medically, but 6–7% are refractory[1]. Patients who are refractory to medical manage-ment can be candidates for surgical treatment such asanatomical or functional hemispherectomy.

Case descriptionThe patient was a 10-month-old, 7.5 kg male who wasborn at 36 weeks gestation. Shortly after birth he wasnoted to have a left-sided hemiparesis. Magnetic reso-nance imaging of the brain showed a large rightmiddlecerebral artery infarction. Awork-up revealed congen-ital ATIII deficiency. The patient subsequently devel-oped seizures refractory tomedical treatment and pre-sented for a right functional hemispherectomy.

The primary concerns of the anesthesiology teamwere (1) perioperative management of congenitalATIII deficiency including the administration ofpooled ATIII from human plasma (thrombate III);(2) the risk of using invasive hemodynamic monitorsin the setting of increased risk of thromboembolicevents secondary to congenital ATIII deficiency; and(3) complications associated with hemispherectomy,especially the increased risk of bleeding.

Following his perinatal thrombotic event, anti-coagulation therapy was initiated and consisted of8 mg subcutaneous enoxaparin, injected twice daily.Twenty-four hours prior to surgery, enoxaparin washeld. A baseline ATIII activity was found to be at 50%

of normal and the patient received 400 IU throm-bate III. Antithrombin III activity drawn 20 minutesafter the treatment was 136%. Our target range ofATIII activity in the perioperative period was set tobe between 80–120% of normal. Given the risk ofblood loss and large volume shifts that can occur dur-ing functional hemispherectomy,wedecided that inva-sive hemodynamic monitors were necessary duringthe operation. Prior to the start of surgery, an ATIIIlevel was drawn.

After inhalational induction, two peripheral intra-venous lines were secured and the trachea was intu-bated. Anesthesia was maintained using balancedtechnique of isoflurane in oxygen and nitrous oxidecombined with remifentanil infusion. Rocuroniumwas used to provide muscle relaxation. A left arterialcatheter was placed. A double lumen right internaljugular central venous catheter was placed using ultra-sound guidance. Both lines were frequently flushedintraoperatively and all lumens remained patentintraoperatively.

We closely monitored for signs and symptomsof thromboembolism including limb or face dis-coloration, swelling, or change in temperature ofextremities.

Antithrombin III level drawn in the immediateperioperative period was at 86% activity. In consulta-tion with the hematologist, we gave a dose of throm-bate III, 125 units over 20 minutes. There were nocomplications associated with infusion.The operationwas otherwise uneventful; therewas nomajor bleedingor thrombotic event. The trachea was extubated suc-cessfully in the operating room and the patient subse-quently transferred to the pediatric intensive care unitfor further management.

Postoperatively, ATIII levels were checked twicedaily and infusions of thrombate III were dosedaccordingly. The patient’s arterial catheter was

203


removed a few hours after the end of the surgerybecause blood could no longer be aspirated from thecatheter. There was no sign of thromboembolism inthe left upper extremity. The central venous catheterremained in place until postoperative day 5. Bythe time of removal both lumens had clotted. Anultrasound of the upper extremity did not revealany evidence of thrombosis. On postoperative day 6,thrombate III infusion was stopped and ATIII levelsfollowed. Once the levels dropped to the 80% range,enoxaparin subcutaneous injection was restarted at adose of 1 mg/kg twice daily. The patient did not sufferany thrombotic or hemorrhagic events throughout hishospital stay.

DiscussionEarly surgery for intractable epilepsy is recommendedas it has been shown to improve functional outcomes.Anatomic hemispherectomy consists of the resectionof the frontal, parietal and occipital cortices, completetemporal lobectomy and insular resection. Periop-erative complications associated with this procedureinclude significant changes in systemic and pulmonaryvascular resistance, arrhythmias, cardiac arrest, neu-rogenic pulmonary edema, seizures, cerebral edema,massive blood loss, and coagulopathy. Modificationof this procedure led to introduction of functionalhemispherectomy. This procedure leaves the anteriorfrontal and posterior occipital lobes intact while a tem-poral lobectomy and central resection are performed.The remaining cortex is completely disconnected byresecting the corpus callosum. In this manner, thefrontal and occipital lobes remain, but are functionallydisconnected from the brain. This procedure is asso-ciated with reduced hemorrhage and other complica-tions. Patients undergoing hemispherectomy are usu-ally on chronic anticonvulsant therapy. This leads tohepatic enzyme induction and consequently opioidsandmuscle relaxants aremore rapidlymetabolized [2].Thus, more frequent or larger dosing of these medica-tions is often required.

Antithrombin III is a serine protease inhibitor thatplays an important role in modulating coagulationby inhibiting factor IIa (thrombin), factor Xa andto a lesser extent factors IX, XI, XII (Figure 62.1).Antithrombin III is produced by the liver and hasnormal plasma concentrations of 112–140 mcg/mL(equivalent to approximately 100 IU/dL of ATIII activ-ity) with a half life of 2–3 days [3]. Congenital ATIII

deficiency is themost clinically important of the inher-ited thrombophilias resulting in thrombosis in themajority of those affected. The challenge in managingthese patients is preventing potentially life-threateningthrombosis, while minimizing the equally signifi-cant risk of hemorrhage associated with long-termanticoagulation.

Patients with congenital ATIII deficiency havearound 50% normal levels of antithrombin activity.Interestingly, childhood is a period that carries thelowest thrombotic risk. Increasing age and presenceof hypercoagulable states such as pregnancy, surgery,or immobility raise the risk of thrombosis markedly.Even rarer is the instance of arterial thromboticevents, with venous thrombotic events being muchmore prevalent. Our patient represents a very rarecase in that one of his thrombotic events that hedeveloped in the perinatal period was arterial inorigin. While primary prophylaxis with anticoagulanttherapy is generally not recommended in asymp-tomatic ATIII deficiency [4], our patient requiredlow molecular weight heparin therapy perioperativelygiven his serious thromboembolic complications. Inaddition, our patient received thrombate III in thepre-, intra-, and postoperative periods. Despite theabsence of a randomized controlled trial to provebenefit to this treatment and despite the inherent riskassociated with the use of pooled plasma products,its use has been considered appropriate in patientsundergoing procedures with a high risk of venousthromboembolic events. Antithrombin III could becontinued for few days until it is safe to administertherapeutic doses of anticoagulants. Side effects ofATIII include: hypersensitivity, dizziness, chest tight-ness, dyspnea, or nausea.The risk of viral transmissionsuch as human immunodeficiency virus, hepatitis Cvirus, hepatitis B virus, or other infectious agents isvery rare.

Potency of thrombate III is expressed in interna-tional units (units) as tested against activity of theWorld Health Organization reference standard. Oneunit is approximately equivalent to the amount ofATIII (mg) in 1 mL of pooled human plasma fromhealthy donors. The initial loading dose of thrombateIII is calculated by the following equation:

Initial dose (units) = [(Goal ATIII level %− baseline ATIII level %)× body weight (kg)] /1.4.

204

Case 62. Hemispherectomy for intractable epilepsy

Figure 62.1. The anticoagulant action ofantithrombin III.

The formula is based on the expected incremen-tal in vivo increase of antithrombin III concentra-tions above baseline values of 1.4% for each unit/kgadministered. Thrombate III is given as an infusionover 10–20 minutes. Strict aseptic technique is rec-ommended during reconstitution since the drug con-tains no preservative. It should be diluted with sterilewater and used within 3 hours after reconstitution.The maintenance dosage is calculated by determin-ing the preinfusion (trough) and peak postinfu-sion ATIII concentrations. Additional doses of ATIIIare administered at appropriate intervals (e.g., every24 hours) until peak and trough concentrations aremaintained within the therapeutic range of 80–120%of normal. Therapy with thrombate III is usually con-

tinued for 2–8 days following thromboembolism orsurgical or obstetric procedures, depending on theclinical situation. It is not known whether supra-physiologic concentrations (e.g., 150–200%of normal)increase bleeding risk in patientswith congenital ATIIIdeficiency [5].

Care of these patients not only requires anintegrated team approach with close collaborationamong the anesthesiologist, hematologist, surgeon,and intensivist, but an acute sense of vigilance inmonitoring for signs of thromboembolic events –particularly in relation to foreign material such asindwelling catheters. Invasive venous and arterialcatheters were used safely in this patient and were onlykept in place for as long as absolutely necessary.

205


References1. P. Uvebrant. Hemiplegic cerebral palsy. Aetiology

and outcome. Acta Paediatr Scand Suppl 1988; 345:1–100.

2. S. G. Soriano, L. J. Sullivan, K. Venkatakrishnan et al.Pharmacokinetics and pharmacodynamics ofvecuronium in children receiving phenytoin orcarbamazepine for chronic anticonvulsant therapy. BrJ Anaesth 2001; 86: 223–9.

3. P. S. Maclean, R. C. Tait. Hereditary and acquiredantithrombin deficiency. Drugs 2007; 67: 1429–40.

4. Haemostasis andThrombosis Task Force, BritishCommittee for Standards in Haematology.Investigation and management of heritablethrombophilia. Br J Haematol 2001; 114: 512–28.

5. Talecris Biotherapeutics.Thrombate R© III(antithrombin III) injection prescribing information.Research Triangle Park, NC; 2006.

206


63 Neurosurgical procedures for pediatricpatients with cardiac malformationsStephen J. Kimatian, Kenneth Saliba and Erin S. Williams

In the USA, approximately 30 000 infants are bornannually with congenital heart disease (CHD) andcurrently there are an estimated 750 000–1 000 000children and adults with CHD who may present fornoncardiac surgery [1, 2]. With advances in medi-cal and surgical care, patients are surviving complexcardiac palliative procedures and living well beyondadolescence into adulthood. This underscores impor-tant issues regarding the safe and proper anestheticmanagement of the CHD patient having noncar-diac surgery. Since the presence of CHD increasesthe risk for noncardiac surgery, the anesthesiologistmust understand the pathophysiology of the cardiaclesion and how their anesthetic management affectsthe patient’s cardiovascular system.

In order to minimize complications, the anes-thesiologist must tailor their anesthetic managementaccording to the age of the patient, type of lesion,extent of corrective procedure, and presence of othercongenital anomalies [3].

Case descriptionThe patient was a 7-year-old female with a historyof hypoplastic left heart syndrome (HLHS) who wasstatus post Fontan completion operation 3 years agoand who presented for resection of a right tempo-ral lobe mass. She was premedicated with oral mida-zolam 0.5 mg/kg in the preoperative suite and thenbrought to the operating room, where she under-went an uneventful inhalation induction. Two large-bore peripheral intravenous catheters were placed, aswell as an arterial line. General anesthesia was main-tained with isoflurane and fentanyl, with neuromuscu-lar blockade achieved with rocuronium.

The primary concerns of the anesthesiology teamincluded the avoidance of increased pulmonary vas-cular resistance (PVR), the maintenance of adequatesystemic oxygenation, the avoidance of increasedintracranial pressure, and themaintenance of adequate

cerebral perfusion pressure. Pulmonary vascular resis-tance is critical after a Fontan procedure, as bloodflows directly from the right atrium to the pulmonaryarteries – increased PVR can lead to cyanosis.

After removal of the bone flap the surgeon statedthat the dura was tight, so the anesthesiologist hyper-ventilated the patient to an end-tidal CO2 of 32. Thesurgery proceeded uneventfully and the tumor wasresected. The patient remained hemodynamically sta-ble throughout the case and the trachea was extubatedin the operating room at the end of the procedure.

DiscussionWhen managing a CHD patient for noncardiacsurgery it is imperative to address the specific anatomyand physiology following any palliative procedures.Also, one must appreciate the presence of intracar-diac or systemic-to-pulmonary artery shunts as wellas how manipulations of PVR and systemic vascularresistance (SVR) affect these shunts. Lastly, the pres-ence of dysrhythmias, polycythemia, and hyperviscos-ity should be identified, along with the observance ofsubacute bacterial endocarditis (SBE) prophylaxis [4].In this case, the patient had undergone a total cavopul-monary anastamosis following the Fontan procedure,without significant residual shunt. Thus, no furthersurgical intervention was required and arterial satura-tions were typically 94–98%.

Preoperative managementThe history and physical examination are extremelyimportant in evaluating a patient with CHD. Duringthe physical examination, the clinician should assessthe airway, heart, lungs, and extremities. The extremi-ties can be assessed for ease of vascular access and thepresence of adequate pulses, clubbing, or edema. It isimperative that the patient be euvolemicwith adequateperipheral perfusion and hydration prior to the induc-tion of general anesthesia. This can also be assessed by

207


evaluating the patient on physical examination. Signssuch as tachycardia, lowbloodpressure, decreased skinturgor, dry mucous membranes, and decreased capil-lary refill all indicate dehydration and hypovolemia;therefore intravenous access should be obtained andthe patient adequately resuscitated before proceedingwith induction. If possible, the intake of clear liq-uids should be encouraged up to 2 hours before theplanned institution of general anesthesia in the postFontan population to optimize intravascular volumestatus [5].

Evaluation of the most recent electrocardio-gram, echocardiogram, and cardiac catheterizationis paramount. Ideally, the patient would have had acardiologist consultation within the prior 6 months,or as part of their preoperative evaluation [6].

The presence of a fenestration in the Fontan con-duit also carries the risk for paradoxical embolism;therefore, strict bubble precautionsmust be adhered tothroughout the perioperative infusion of intravenousfluids or medications.

The history and physical examination is essentialin determining the presence of congestive heart fail-ure. Krane developed 10 rules that can help the anes-thesiologist when examining a child with CHD.Theseinclude:

1. You cannot auscultate a crying child.2. Examine the heart first.3. Most systolic murmurs will be benign.4. All pan-systolic murmurs are pathological.5. All diastolic murmurs are pathological.6. All murmurs that radiate are pathological.7. The louder the murmur the smaller the shunt.8. The softer the murmur the larger the shunt, thesicker the child.

9. Children develop biventricular failure.10. Obtain blood pressure readings in upper and

lower extremities [7].

It is also important for the anesthesiologist tounderstand how to manage preload, PVR, SVR, heartrate, and contractility in patients with CHD.

For example, in patients with tetralogy of Fallot,it is imperative to maintain preload while decreasingheart rate and contractility. Pulmonary vascularresistance should also be decreased to minimize rightto left shunting while SVR remains increased. If theanesthesiologist is presented with a patient that hasundergone a Fontan, the physiologic goals for this

Table 63.1. Effect of intracardiac shunts on intravenous andinhalation induction.

Shunt typeIntravenousinduction

Inhalationinduction

Left to right Slower Faster orunchanged

Right to left Faster Slower

patient would include maintaining preload and avoid-ing anything that could increase the PVR in order tomaximize pulmonary circulation. Systemic vascularresistance should be maintained and no changes arewarranted with heart rate and contractility [3].

Since patients with CHD will possibly have under-gone multiple surgeries, they can have a great dealof preoperative anxiety. Therefore, premedication canbe administered unless the patient is hemodynam-ically unstable, symptomatic from mass effect, orlethargic. Typically, midazolam is given either orally,intranasally, or intravenously.

Intraoperative managementThe presence of intracardiac shunts can affect thespeed of induction. The anesthesiologist should beknowledgeable of these changes (see Table 63.1).For example, a patient with left-to-right shunt willhave a quicker inhalation induction because there ismore blood passing through the lungs and therebymore blood being delivered to the brain, thus lead-ing to unconsciousness sooner. Patients with right-to-left shunts have slower inhalation inductions becauseblood is diverted away from the lungs [3]. With right-to-left shunts, however, intravenous induction agentsbypass the lungs and go directly to the systemic circu-lation – this makes the possibility of overdose muchhigher. Thus, the clinician must appropriately adjusthis or her dosages.

Once induction is underway, the anesthesiologistmust vigilantly prevent any increases in intracranialpressure.This can be done bymaintaining an adequatedepth of anesthesia during laryngoscopy, as well asduring the most stimulating portions of the plannedoperative procedure, such as skull pin placement, anddural reflection, as well as the remainder of the surgicalprocedure. During the maintenance phase of anesthe-sia, volatile anesthetics along with intermediate actingneuromuscular blockers, such as rocuronium, and fen-tanyl are acceptable. Oxygen and air mixtures would

208

Case 63. Neurosurgical procedures and cardiac malformations

be optimal since nitrous oxide can cause expansion ofvenous air emboli, which could lead to acute increasesin PVR and catastrophic cardiovascular collapse. Inaddition, patients with Fontan anatomywill often havea fenestration between the venous and arterial sys-tems at the level of the atrial baffle setting the stagefor right-to-left shunting in the presence of increasedPVR and emboli to the arterial circulation. Moni-tors such as capnography and precordial Doppler arevital in diagnosing a venous air embolism. The trans-esophageal echocardiogram is themost sensitivemon-itor for detection of air emboli, but its availability, size,and encroachment near the surgical field, plus the req-uisite expertise, may limit its feasibility.

Patients with Fontan circulation are dependentupon central venous pressure for pulmonary bloodflow. Blood must flow passively through the lungs tofill the heart, making the management of pulmonaryblood flow an important aspect of care for thesepatients. It is imperative that preload, oxygenation,and ventilation be maintained to avoid catastrophicelevations in PVR. Other etiologies of increased PVRsuch as stress, hypothermia, acidosis, and atelectasismust be avoided.

Given the need to maintain adequate cerebral oxy-genation, cerebral perfusion, and to avoid intracranialhypertension, an arterial line must supplement stan-dard ASAmonitors, and will allow real-time measure-ment of blood pressure, as well as providing the abil-ity to follow PaO2, PaCO2, and electrolytes. End-tidalCO2 to PaCO2 differences can also be followed asmea-surement of relative pulmonary dead space and anindirect indicator ofWest Lung Zone 1 and pulmonaryblood flow.

With regard to SBE or infective endocarditis pro-phylaxis, the American Heart Association has guide-lines regarding the administration of appropriateantimicrobial therapy to susceptible patients. Subacutebacterial endocarditis can develop after noncardiacsurgery in any mucosal penetrating procedure thatpermits bacteria to seed on prosthetic endocardium[3]. The previous guidelines recommended routineSBE prophylaxis for all patients at risk undergoingdental, gastrointestinal, or genitourinary procedures;however, this is no longer the case. Prophylaxis is nowrecommended only for patients considered at highestrisk of acquiring SBE, such as:

1. Patients with prosthetic heart valves or withprosthetic material used in valve repair.

2. History of previous endocarditis.3. Specific CHD (unrepaired cyanotic CHD,

completely repaired CHD with prosthetic materialduring the past 6 months, repaired CHD withresidual defect).

4. Cardiac transplant patients with cardiac valvulardisease (American Heart Association: www.heart.org/HEARTORG/).

Postoperative managementIn patients status post Fontan procedures, sponta-neous ventilation allows for better pulmonary bloodflow than does positive-pressure ventilation. Thus, ifthe operation has been otherwise uneventful and thepatient is hemodynamically stable and normothermic,extubation of the trachea in the operating room shouldbe considered. Reversal of neuromuscular blockademust first take place. Postoperative pain managementcan include carefully titrated opioids such as fen-tanyl, hydromorphone, or morphine with special careto avoid respiratory compromise that can result inhypercarbia, hypoxia, atelectasis, and increased PVR.Early postoperative emergence and extubation willalso facilitate postoperative neurologic examination.

References1. R. R. Clancy. Neuroprotection in infant heart surgery.

Clin Perinatol 2008; 35: 809–21.2. R. Sumpelmann,W. Osthaus. The pediatric cardiac

patient presenting for non-cardiac surgery. Curr OpinAnaesthesiol 2007; 20: 216–20.

3. R. Mohindra,D. Beebe, K. Belani. Anestheticmanagement of patients with congenital heart diseasepresenting for non-cardiac surgery. Ann Card Anaesth2002; 5: 15–24.

4. M. Cannesson,M. Earing, V. Collange et al.Anesthesia for noncardiac surgery in adults withcongenital heart disease. Anesthesiology 2009; 111:432–40.

5. C. D. McCLain, F. X. McGowan, P. G. Kovatsis.Laparoscopic surgery in a patient with fontanphysiology. Anesth Analg 2006; 103: 856–8.

6. L. Diaz, S. Hall. Anesthesia for non-cardiac surgeryand magnetic resonance imaging. In Andropoulos D.B., Stayer S. A., Russell I. A., eds. Anesthesia forCongenital Heart Disease. Malden, MA: BlackwellFutura, 2005; 427–52.

7. E. Krane. Anesthesia in children with congenital heartdisease. Pediatric Anesthesia and Pain Management;2–15. Available from http://pedsanesthesia.stanford.edu/downloads/guideline-chd.pdf.

209


64 Neuroprotection during pediatriccardiac anesthesiaStephen J. Kimatian, Erin S. Williams and Kenneth Saliba

In the USA, approximately 30 000 infants are bornannually with congenital heart disease (CHD) [1]. Ofthese 30 000 infants, about 11 000 will require someform of surgical correction [1]. Such procedures placethese infants at risk for morbidity and mortality asso-ciated with cardiac surgery. However, with advances inmedicine and surgical technique patients are surviv-ing these intricate procedures, leading us to focus ourattention toward issues such as optimizing the neuro-logic outcomes of patients who have undergone con-genital heart surgery [1].

The overall incidence of neurologic injury asso-ciated with cardiac surgery in children is 2–25%[2]. These insults can stem from preoperative brainanomalies, perioperative hypoxemia and low cardiacoutput states, as well as sequelae from cardiopul-monary bypass and deep hypothermic circulatoryarrest. It has also been found that approximately halfof school-aged survivors of infant heart surgery receivesome sort of special education assistance [1]. In a studyof 171 children after repair of d-transposition of thegreat arteries (d-TGA), these children hadmore devel-opmental delay, speech problems, learning disorders,and attention problems than controls at the same age[3]. With such prevalence and such an effect on psy-chosocial development, it is important that anesthesi-ologists monitor and protect the pediatric neurologicsystem.

Case descriptionThe patient was a 3-day-old neonate with a historysignificant for hypoplastic left heart syndrome. Thepatient was born at 37 weeks gestation and was cur-rently on a prostaglandin infusion tomaintain patencyof the ductus arteriosis. As expected in a patient withcomplete mixing of oxygenated and deoxygenatedblood in a single ventricle, pulse oximetry showed sat-urations in the low to mid 80s.

The patient was scheduled for a Norwood proce-dure.The Norwood procedure would result in the cre-ation of a single “neo aorta” to provide systemic bloodflow, with pulmonary blood flow from a surgically cre-ated shunt between the systemic and pulmonary arte-rial systems (Blalock Taussig Shunt). Such extensivereconstruction of the aortic arch requires a period oflow flow or no flow while on cardiac bypass, placingthe child at significant risk for neurologic injury.

In addition to standard ASA monitors, intraop-erative monitoring included an arterial line, centralvenous line, and near infrared spectroscopy (NIRS).The patient was induced intravenously with keta-mine 2 mg/kg, pancuronium 0.1 mg/kg and fentanyl2 mcg/kg. The focus of the anesthesiology team wasmaintaining a balance between pulmonary and sys-temic blood flow (Qp:Qs) that would ensure ade-quate cardiac output and oxygen content such that endorganperfusion andoxygenationwasmaintained.Thischild’s single ventricle morphology resulted in bloodflowing to the lungs or body proportional to the rela-tive resistance of the two respective systems.Maintain-ing homeostasis would require active manipulationof pulmonary, systemic, and cerebral vascular resis-tance.The surgery was approximately 5 hours in dura-tion requiring 40 minutes of deep hypothermic circu-latory arrest (DHCA) during which time the patientwas cooled to 18 ◦C and cerebral perfusion was main-tained via selected perfusion of the innominate artery.The patient remained stable throughout the case andthe trachea was extubated 3 hours after surgery with-out any gross neurologic sequelae.

DiscussionWhen considering the etiology of negative neurologicoutcomes after cardiac bypass surgery, the adult liter-ature cites primarily embolic causes, while the pedi-atric literature cites a number of different possibili-ties [2]. The neurologic sequelae in infants who have

210

Case 64. Neuroprotection during pediatric cardiac anesthesia

undergone cardiac surgery include seizures, stroke,and choreoathetosis [1]. During DHCA the pedi-atric patient endures a planned period of low, orno, cerebral blood flow. This insult causes changesvery similar to the changes seen in hypoxic ischemicencephalopathy [1].

In the past, it was thought that infants who under-went cardiac surgery had normal brains and that anyinjury that occurred was simply secondary to surgery.However, magnetic resonance imaging of the brainsof neonates with CHD has shown a significant inci-dence of anatomical components that appear prema-ture compared with term neonates without CHD. In astudy by Williams et al., preoperative magnetic reso-nance imaging showed white matter injury in �15%of neonates and cerebral atrophy changes in one-third of children with CHD [3]. Additionally, thereare increased amounts of periventricular leukomala-cia, also known as white matter neuropathology, aswell as microcephaly, hypotonia, and delayed brainmaturation [3].

Cerebral injury after cardiac surgery continuesto be a significant source of morbidity, and thereare numerous pharmacologic and nonpharmacologicmodalities that have been implemented to preventsuch injury. During the course of cardiac bypasssurgery involving DHCA, the patient is actively cooledvia the cardiopulmonary bypass (CPB) circuit to a coretemperature of 18 ◦C prior to the initiation of the low-or no-flow period. Once the repair is complete, thepatient is slowly rewarmed prior to being weaned fromCPB. Various techniques have been employed to min-imize or modify the period of DHCA as prolongedDHCA with no flow (�30 minutes) has been associ-ated with a higher incidence of neurologic complica-tions [3]. One method is regional low-flow perfusion.This allows the brain andupper body to be perfused viaa GoreTex graft connected to the innominate artery.The graft allows for CPB blood flow to the right com-mon carotid and right vertebral artery to directly per-fuse the right side of the brain. Assuming an intactcircle of Willis, the left side of the brain is also indi-rectly perfused.Animal studies have shownbetter neu-rologic outcomes and less apoptosis in piglets exposedto regional low-flow perfusion compared with thosesubjected to DHCA [3]. Other perfusionmethods thatmay lead to better neurologic outcome include inter-mittent cerebral perfusion and low-flow CBP.

Hematocrit while on bypass affects the oxygen car-rying capacity to end organs as well as the viscosity and

flow dynamics in capillary beds. Recent studies havesuggested that higher hematocrits in the range of 25–30 result in better neurologic outcomes than the hema-tocrits in the low tomid 20s that had been traditionallyused [4].

Neurologic monitoring should be used duringsurgery for congenital cardiac lesions, however, it isimportant to note that many of these modalities rep-resent indirect measurements of neurologic functionor cerebral oxygen status and are not without limi-tations. The electroencephalogram (EEG), by provid-ing a measure of cerebral electrical activity, can pro-vide an indirect measure of cerebral metabolic oxy-gen demand as well as display characteristic changes inthe presence of cerebral hypoxia. Inducing an isoelec-tric EEG via thermal or pharmacologic interventionprior to the initiation ofDHCAor lowflowcanprovideassurance that cerebral oxygen demand is at its low-est prior to the anticipated insult. Using intraopera-tive EEG does require dedicated personnel with exper-tise to place the electrodes and interpret thewaveformsduring the surgery. Intraoperative monitoring is fur-ther complicated by many sources of artifact not seenin normal EEGs including distortion by anesthetics,electrical surgical equipment, CPB and patient tem-perature [3]. Austin et al. found that EEG changeswere associated with only 5% of adverse neurologicoutcomes and that aberrant readings did not consis-tently correlate with negative postoperative outcomes[5]. For these reasons the routine use of EEG in pedi-atric cardiac surgerywas not recommended. ProcessedEEG, such as the BIS monitor, has been studied as ameans of judging depth of anesthesia but has not beenshown to correlate with cerebral injury in children.

Near infrared spectroscopy is a noninvasivemethod for monitoring cerebral oxygenation. Ituses light at wavelengths 700–1000 micrometers ina manner similar to pulse oximetry, but without adiscernible pulse to distinguish arterial from venousflow, NIRS provides an indication of regional tissuesaturation. A monitor that emits near infrared light isplaced on the forehead of the patient, which penetratesan approximately 10 mL banana-shaped region oftissue in the frontal cortex. The only Food and DrugAdministration-approved instrument for measuringcerebral oxygen saturation specifically uses the termregional saturation index (rSO2i). In order to interpretthe rSO2i, it is assumed that 75% of cerebral bloodvolume is venous while the remaining 25% is arterial.For pediatric patients there are pediatric-specific

211


probes (4–40 kg), which account for anatomicaldifferences such as a thinner skull and smaller radiusof curvature. The rSO2i value indicates a relativeratio of oxyhemoglobin to total hemoglobin thatresults in a numerical score reported as a raw score of15–95. These values are not considered actual percentsaturation of oxygen but rather a relative value frombaseline. This means that proper interpretation of theNIRS readings requires a baseline measurement atthe start of the case, preferably prior to the inductionof anesthesia. Adult studies have suggested thatdecreases of �20% from baseline could be associatedwith neurologic injury [3].

Although it is difficult to make direct applicationof animal studies to humans, there may be some rel-evance to low rSO2i and cell dysfunction, cell death,and subsequently negative neurologic outcome. Pigletmodels that usedNIRSwith different variables showedthat the nadir of rSO2i occurred earlier with lowerhematocrits and higher temperatures. In addition,lower rSO2i readings were seen with the use of alpha-stat pH management when compared with pH-statmanagement. It has been suggested that this differ-ence seen with pH-stat management results from bet-ter cerebral blood flow and more effective cooling ofthe brain [3]. Near infrared spectroscopy has also beenshown to be effective in detecting cerebral desatu-ration with superior vena cava occlusion in a pigletmodel. Extrapolation of these studies to the neona-tal population has been suggested for superior venacava cannulation for bypass, which can result in partialocclusion [3].

One study of 26 infants and children undergoingbypass and DHCA found that the three patients withlow rSO2i had acute postoperative neurologic events,which included seizures in one and prolonged comain the other two patients [2]. With such findings, itappears that NIRS may be beneficial in identifyingpotential neurologic injury. There are, however, limi-tations to NIRS such as the small area that is assessedfor cerebral desaturation, the variation in rSO2i preop-eratively (depending on the cardiac anatomy), and theneed for more prospective outcome studies.

The transcranial Doppler allows real time assess-ment of cerebral blood flow velocity and emboli dur-ing cardiac surgery. The most common blood vesselevaluated is the middle cerebral artery. Typically, a2 MHz probe is placed above the zygoma and ante-rior to the tragus of the ear. This position allows visu-alization of the bifurcation of the middle and anterior

cerebral artery in which there is antegrade (toward thetransducer with the middle cerebral artery) and ret-rograde flow (away from transducer with the anteriorcerebral artery) [2]. The anterior fontanelle is an alter-native site that can be used in infants [2]. In additionto showing flow velocity and emboli, case reports havealso shown that transcranial Doppler can detect cere-bral steal in the presence of systemic to pulmonaryshunts [6].

Transcranial Doppler has several technical andlogistic issues that must be considered. Placement ofthe probe can be technically difficult and maintain-ing probe position in the operative environment canbe challenging. In addition, artifact can also producefalse “hits” and requires a trained eye to distinguishemboli from artifact. It is also important to remem-ber that cerebral blood flow velocity depends uponthe size of the cerebral vasculature, and cerebral bloodflow is affected by cerebral vascular resistance. Thus,the clinician must consider anesthetic and physiologicchanges that affect cerebral vascular resistance wheninterpreting changes in flow velocity. While this canbe a very useful tool, the technical difficulties coupledwith the expertise required for accurate interpretationmakes regular use impractical in many institutions.As the technology improves, it has the potential ofbeing a useful adjunct for neurologicmonitoring in thefuture.

ConclusionWith so many patients surviving very complex con-genital heart surgeries, the anesthesiologist shouldlead theway in preventing, assessing, and treating neu-rologic injury in order to obtain the highest quality oflife possible for these children. While intraoperativeneurologic monitoring for pediatric cardiac surgery isstill a relatively new science, it shows a great deal ofpromise for future application in minimizing cerebralinjury and optimizing neurodevelopmental outcome.

References1. R. R. Clancy. Neuroprotection in infant heart surgery.

Clin Perinatol 2008; 35: 809–21.2. D. B. Andropoulos, S. A. Stayer, L. K. Diaz et al.

Neurologic monitoring for congenital heart surgery.Anesth Analg 2004; 99: 1365–75.

3. G. D. Williams, C. Ramamoorthy. Brain monitoringand protection during pediatric cardiac surgery. SeminCardiothorac Vasc Anesth 2007; 11: 23–33.

212

Case 64. Neuroprotection during pediatric cardiac anesthesia

4. S. J. Kimatian, K. J. Saliba, X. Soler et al. Theinfluence of neurophysiologic monitoring on themanagement of pediatric cardiopulmonary bypass.ASAIO 2008; 54: 467–9.

5. E. H. Austin 3rd,H. L. Edmonds Jr, S. M. Auden et al.Benefit of neurophysiologic monitoring for pediatric

cardiac surgery. J Thorac Cardiovasc Surg 1997; 114:707–15, 717; discussion 715–16.

6. S. J. Kimatian, J. L. Myers, S. K. Johnson et al.Transcranial Doppler-revealed retrograde cerebralartery flow during Norwood operation. ASAIO 2006;52: 608–10.

213

Part VII Neurologic sequelae in other patient populations. PregnancyCase

65 Pregnant patient with aneurysmKaren K. Wilkins and Alexandra S. Bullough

Approximately 1–2% of pregnant women undergononobstetric surgery during their pregnancy.The chal-lenge in these cases is that the anesthesia care providermust take into consideration both mother and fetus.Neurosurgery during pregnancy is rare and as a resultthere are few evidence-based recommendations in theliterature to provide guidance. Concerns are oftenraised about the perinatal implications of anestheticexposure. An understanding of maternal physiologyand a multidisciplinary approach are imperative toensure a successful outcome.

Case descriptionA 37-year-old G9 P6, ASA 3, Jehovah’s Witness femalewith multiple hematologic co-morbidities presentedat 18 weeks gestation with a 1-year history of peri-oral and periocular twitching, memory lapses and arecent sensory loss and painful paresthesias affect-ing the right side of her body. Magnetic resonanceangiography demonstrated a 7-mmsaccular aneurysminvolving the anterior communicating artery. Her co-morbidities included von Willebrand’s disease type I,alpha thalassemia trait and antiphospholipid antibodysyndrome. After amultidisciplinary discussion involv-ing neurosurgery, obstetrics, and hematology it wasdecided to proceed with intracranial aneurysm clip-ping via craniotomy at 18 weeks gestation.

Primary concerns for the anesthesiology teamwere (1) risk of significant blood loss given her mul-tiple hematologic dyscrasias and Jehovah’s Witnessstatus; (2) maintenance of hemodynamic stability –hypotension would be detrimental to placental perfu-sion and perioperative hypertension risks aneurysmrupture; (3) maternal aspiration; and (4) fetal survivaland wellbeing.

On the day of surgery, Humate-P was admin-istered preoperatively per hematology recommenda-tions. Humate-P (lyophilized concentrate of FactorVIII and von Willebrand’s Factor) is given perioper-

atively to patients with von Willebrand’s disease topromote platelet aggregation and adhesion to dam-aged vascular endothelium. Fetal heart tones wereconfirmed by the obstetricians. A private discussionbetween the anesthesiologist and patient regardingblood replacement therapy also took place.The patientagreed to receive a transfusion of packed red bloodcells if more than one unit was required to maintainmaternal and fetal safety. She also agreed to cell saverblood replacement if the systemmaintained continuitywith her circulation. The patient had a 16 g peripheralintravenous catheter and a radial arterial line placedpreoperatively.

A smooth intravenous rapid sequence inductionwith cricoid pressure was performed using lidocaine,fentanyl, propofol, and succinylcholine. After anes-thetic induction, a subclavian central line was placedfor potential administration of vasopressors or fluidresuscitation. Esmolol was utilized during cranialpin placement. Hemodynamic stability was achievedthroughout the case using phenylephrine and remifen-tanil infusions. The patient received two units ofpacked red blood cells due to increasing phenylephrinerequirements and a hematocrit of 17 prior to manipu-lation of the aneurysm.The remainder of the case pro-ceededwithout incident and the trachea was extubatedpostoperatively. In the recovery room, the fetal hearttones were confirmed to be in the 150s. Her postop-erative course was uneventful and she was transferredout of the intensive care unit on postoperative day2 and discharged home on day 5 without neurologicsequelae.

DiscussionIntracranial aneurysms have an estimated rupturerate of approximately 20 in 100 000 pregnancies withmost cases occurring between the 30th week of preg-nancy and 6 weeks postpartum [1–3]. An increasein aneurysm rupture risk is most likely due to the

215

VII. Neurologic sequelae in other patient populations. Pregnancy

Table 65.1. Physiologic changes in pregnancy.

System Physiologic changes Comments

Cardiovascular ↑ Cardiac output↑ Blood volume↑ Resting heart rate↓ Systemic vascular resistance↓ Blood pressure (second trimester)Dilutional anemiaAortocaval compression after 16–20 weeks↑ Risk of thromboembolic disease

Maintain systolic blood pressure �100 mmHgLeft uterine displacement after 16–20 weeks

Pulmonary ↑ Respiratory rate↑ Tidal volume↑ Minute ventilation↓ Functional residual capacityRespiratory alkalosisMucosal engorgement/edemaPotentially difficult airway

PreoxygenationOptimal “sniffing position” or rampingAlternative intubation plan (intubating laryngeal maskairway, Glidescope, fiberoptic scope)

Gastrointestinal ↓ Gastric motility↓ Esophageal sphincter tone

Nonparticulate antacidRapid sequence induction

Central nervous system ↓ Minimum alveolar concentration↓ Local anesthetic requirements

changes that occur in cardiac physiology during preg-nancy (Table 65.1), which comprise a 50% increase incardiac output and blood volume expansion as wellas a hormonal softening of vascular connective tissue[1, 3]. Maternal mortality associated with a rupturedaneurysm is approximately 35% [1, 2, 4].

In this case, in addition to the usual concerns ofa neurosurgical patient undergoing aneurysm clip-ping such as intraoperative aneurysmal rupture andcentral nervous system ischemic episodes, the par-turient has issues that are specific and unique topregnancy. Concerns are often raised about the peri-natal implications of anesthetic exposure. In general,propofol, morphine, local anesthetics, muscle relax-ants, and inhalational agents are considered safe inpregnancy. Antiseizure medications known to be ter-atogenic include carbamazepine, phenytoin, and val-proic acid and should be avoided in pregnant neu-rosurgical patients. It is recommended that electivenonobstetric surgery in the pregnant patient be per-formed during the second trimester. Surgery duringthe first trimester, the time of organogenesis and rapidgrowth, is associated with an increased risk of ter-atogenesis and intrauterine fetal demise. Surgery per-formed in the third trimester is associated with anincreased risk of premature labor. In this patient, theobstetricians and neurosurgeons agreed that 18 weeksgestation was an optimal time to perform maternalsurgery.

Vascular neurosurgery may be associated withrapid blood loss and hemodynamic instability, thuslarge-bore intravenous access and invasive monitor-ing are warranted. In addition to the typical invasivelines andmonitors utilized for neurosurgery, onemustconsider fetal monitoring for pregnant patients. Fetalviability is generally accepted to be 24 weeks gesta-tion. Prior to 24 weeks gestation, fetal monitoring istypically limited to a pre/postoperative assessment ofheart tones. After 24 weeks, severe changes in base-line fetal heart tones may be predictive of neonatalmortality and loss of fetal heart rate variability maybe associated with hypoxia, sedative medications, andfetal sleep. Beyond 24 weeks, the decision to monitorfetal heart tones intraoperatively is both institution-and case-dependent. If a decision is made to monitorfetal heart tones intraoperatively, it is recommendedto establish a preoperative management plan after apatient, surgery, anesthesiology, and neonatology con-sensus has been reached on how to respond to unfa-vorable changes in fetal heart tones and whether aCesarean section is desired or even feasible. If severefetal bradycardia is detected the anesthesiologist mustensure adequate maternal blood pressure, ventilation,and oxygenation.

Special attention is required during induction ofanesthesia in a pregnant patient. A rapid sequenceinduction with cricoid pressure is recommended from16 weeks gestation and left uterine displacement is

216

Case 65. Pregnant patient with aneurysm

required after 16–20 weeks gestation to avoid aorto-caval compression [3]. Due to generalized mucosaland vascular engorgement as well as potential airwayedema, intubation may prove to be more challeng-ing in the pregnant patient. It is advisable to have asmaller diameter endotracheal tube available and beprepared for a potentially difficult airway. When con-sidering respiratory physiologic changes in pregnancy,pregnant women have a decreased functional residualcapacity and increased oxygen consumption and willdesaturate rapidly during periods of hypoventilationor apnea, such as seen during anesthetic induction.Preoxygenation is therefore of critical importance.

Hemodynamically, aneurysm clipping in a preg-nant patient carries two major risks – hypertensionand hypotension. All patients are at particular risk ofhypertension and subsequent aneurysm rupture dur-ing direct laryngoscopy, placement of the head in pins,incision, and during removal of the bone flap. Thehypertensive responsemay be attenuated with esmolol(potential for transient fetal bradycardia), labetolol,lidocaine, nitroglycerin or nitroprusside (long-termuse is associated with fetal cyanide toxicity). Hypoten-sion in the pregnant patient (systolic BP�100mmHg)causes a decrease in uteroplacental blood flow and fetalwellbeing. Contrary to past teaching, phenylephrine isnow considered the vasopressor of choice for pregnantpatients due to evidence that it improves neonatal acid-base status [5].

In this case, hemodynamic stability was achievedusing a multimodal approach. An adequate depthof anesthesia was assured prior to endotrachealintubation through the use of fentanyl, lidocaine,and propofol. Postinduction hypotension due tothe administration of Humate-P was treated withphenylephrine. The hypertensive response to cranialpin placement was anticipated and preemptivelytreated with esmolol boluses. Hemodynamic stabilitywas maintained throughout this case with carefullytitrated phenylephrine and remifentanil infusions.

Acceptable methods that assist with reducingmaternal intracranial pressure include: slight head-up position, decreased tidal volumes, avoidance ofbucking on the endotracheal tube, and moderatehyperventilation (PaCO2 28–30 mmHg). The normal

parturient PaCO2 is 30–32 mmHg (respiratory alka-losis) due to increased ventilation during pregnancy.Intraoperatively, it is important to avoid severe hyper-ventilation (PaCO2 �28 mmHg) to prevent a left-ward shift of the maternal oxyhemoglobin dissocia-tion curve impairing transfer of oxygen across theplacenta [6]. In this case, the patient’s PaCO2 waskept between 30–34 mmHg. Mannitol administra-tion in the parturient to decrease intracranial pres-sure is controversial. Intravenous mannitol use mayresult in fetal accumulation leading to fetal hyperos-molality, decreased fetal lung fluid production and uri-nary blood flow, decreased plasma Na+ concentra-tion, and fetal dehydration. However, mannitol dosesof 0.25–0.5 g/kg have been reported in some caseswithout ill effects [1]. This patient received 0.3 g/kg ofmannitol.

ConclusionIn conclusion, neurosurgery in a pregnant patient israre and requires a thorough understanding of thephysiologic changes of pregnancy and the associatedconcomitant anesthetic risks to bothmother and fetus.A multidisciplinary approach is essential to optimizeoutcomes for both mother and fetus.

References1. L. P. Wang,M. J. Paech. Neuroanesthesia for the

pregnant woman. Anesth Analg 2008; 107: 193–200.2. R. Qaiser, P. Black. Neurosurgery in pregnancy.

Semin Neurol 2007; 27: 476–81.3. D. H. Chestnut, L. S. Polley, L. C. Tsen et al.

Chestnut’s Obstetric Anesthesia: Principles and Practice.Philadelphia, PA: Mosby Elsevier Press, 2009.

4. A. M. Bader,D. Acker. Neurologic and musculardisease. InDatta S., ed. Anesthetic and ObstetricManagement of High-risk Pregnancy. New York, NY:Springer-Verlag, 2004; 133–42.

5. D.W. Cooper,M. Carpenter, P. Mowbray et al. Fetaland maternal effects of phenylephrine and ephedrineduring spinal anesthesia for caesarean delivery.Anesthesiology 2002; 97: 1582–90.

6. K. M. Kuczkowski. Nonobstetric surgery duringpregnancy: what are the risks of anesthesia? ObstetGynecol Surv 2004; 59: 52–6.

217


66 Anesthetic management of pregnantpatients with brain tumorsAlaa A. Abd-Elsayed and Ehab Farag

Physiologic changes during pregnancy may result inthe development or growth of nervous system tumors.Such cases at the interface of neuroanesthesia andobstetric anesthesia pose unique challenges.

Case descriptionA 36-year-old female presented at 36 weeks gestationwith headache, nausea, and vomiting. The patient alsoreported that she experienced seizures at her 34thweekof gestation and was prescribed lorazepam as a treat-ment. Computed tomography (CT) scan of the headrevealed the presence of a mass in the left frontal lobemeasuring 5 × 4 cm. Given the signs of intracranialhypertension, a decision was made to perform urgentconcomitant Cesarean section and craniotomy to saveboth themother and the fetus. Standardmonitors wereapplied; large-bore intravenous access and an arte-rial line were also placed. Back-up emergency airwayequipment was prepared in the event of difficult endo-tracheal intubation. A rapid sequence induction wasperformed uneventfully using fentanyl, propofol, androcuronium and anesthetic maintenance was achievedusing desflurane and remifentanil. The Cesarean sec-tion was performed and the baby was delivered withnormal Apgar scores. Craniotomy then proceeded.After resection of the tumor, pathological examinationof the mass revealed anaplastic glioma. The estimatedblood loss during surgery was 100 mL; the patientreceived a total of 1800 mL intravenous fluids duringsurgery.

DiscussionBrain tumors tend to become larger during preg-nancy due to fluid retention, increased blood vol-ume, and hormonal changes. Therefore they may bediagnosed earlier than in the nonparturient. Thereare currently no guidelines for the management ofintracranial tumors in the pregnantwoman.A possiblealgorithm to follow is shown in Figure 66.1 [1].

Medical managementCorticosteroids are often used to reduce cerebraledema.These are safe in pregnancy and have the addi-tional advantage of promoting fetal lung maturity.

First and early second trimestersDuring this time, the fetus is remote fromviability and,as the hemodynamic changes in the pregnant womanhave not peaked, the risks of intraoperative hemor-rhage are not so significant. If the patient is stable, ges-tational advancement may be permitted into the earlysecond trimester, where surgical management of thetumor can be undertaken. Furthermore, radiotherapy,radiosurgery and image-guided surgery during gesta-tion beyond the first trimester may also be options.If the patient is unstable, urgent neurosurgery isindicated.

Late second and third trimestersMaternal intravascular volume peaks at the end of thesecond trimester and tumor resection risks significanthemorrhage: delay of surgery until term is preferred.In stable patients, gestational advancement can be per-mitted, with close observation of themother and fetus.In a patient with worsening symptoms, radiotherapymay be an option to delaying surgery. In an unsta-ble patient with impending herniation, delivery of thebaby by Cesarean section under general anesthesia,followed immediately by surgical decompression, maybe necessary.

Term gestationAt term, delivery can be expedited. In a stable patient,induction of vaginal delivery is an option. A shortenedsecond stage can be achieved with epidural anesthe-sia; although it should be used with care if intracranialpressure is elevated. Most authors advocate Cesareansection only for accepted obstetric and fetal indica-tions, as this procedure does not seem to provide any

218

Case 66. Anesthetic management of pregnant patients with brain tumors

Symptomatic brain tumor

1st and early 2nd trimester

Stable

Permitgestationaladvance-

ment

Permitgestationaladvance-

ment

Unstable Stable Progressiveneurology

Late 2nd and early 3rd trimester Term

Unstable Stable Unstable

Possibleradio-

therapy

Early 2ndtrimester

neurosurgery,adjuvant

radiotherapy

Neuro-surgery

Vaginal deliverypossible

Vaginaldeliverypossible

Cesareansection and

neurosurgery

Cesareansection and

neurosurgery

Figure 66.1. Algorithm for management of brain tumors in the pregnant woman.

definitive advantage over vaginal delivery in protect-ing from increased intracranial pressure. In the unsta-ble patient, as above, Cesarean section under generalanesthesia, followed immediately by surgical decom-pression, is advised.

If an asymptomatic patient is discovered to havea brain tumor during pregnancy, options include a“watch andwait” approach in the face of possible acuteworsening or initiating the above measures. Manni-tol and hypocapnia were avoided in our case to pre-vent fetal dehydration and cerebral ischemia/hypoxia,respectively. Perioperative vigilance for pulmonaryembolism is essential due to the combined risk ofhypercoagulable state/thromboembolism from preg-nancy, craniotomy, and brain tumor.

Previous reports suggest that general anesthe-sia is safe to use in parturients with intracranialtumors. Tracheal intubation is very important as itallows maternal hyperventilation thereby controllingraised intracranial pressure. Patients should be pre-medicated with a nonparticulate antacid and raniti-dine to protect against the sequelae of vomiting andaspiration.

Propofol was used in our case without producingany serious side effects. It is still controversial if it issafe to use propofol in such cases or not. We chosedesflurane because of its rapid onset and titratibil-ity. Although sevoflurane is one of the most preva-lent volatile anesthetics, there are no data regardingneuronal structure or neurocognitive function aftersevoflurane administration in humans or animals.However, sevoflurane can lead to electroencephalo-graphic abnormalities and seizures.

Remifentanil was used in our case without anyadverse neonatal effects. This can be explained bythe fact that it has a unique metabolism by plasmaand tissue esterases and a context-sensitive half lifeof 3–4 min, independent of the duration of infusion[2]. Opioid properties of remifentanil allow both con-trol of the intraoperative stress response and a morerapid recovery than with other commonly used opi-oids. Because of its metabolism and short durationof action, remifentanil can be considered to be safeand effective for general anesthesia for emergencyCesarean section in patients with neurologic riskfactors.

219


Clinically relevant concentrations of remifentanilinduce rapid, persistent increases in N-methyl-D-aspartate (NMDA) responses. A NMDA-receptorblockade during a critical stage in brain developmentleads to depression of neuronal activity, which initiatesthe apoptotic cell death cascade in immature neurons[3]. So, remifentanil potentially prevents this processand has a neuroprotective effect in the fetal brain.

Nitrous oxide was not used in our case. Thereare no human trials examining the effects of nitrousoxide in young children on neuronal structure andneurocognitive performance, but nitrous oxide isknown to inhibit methionine synthase. Case studiesin neonates after exposure to nitrous oxide in uteroduring the third trimester of pregnancy or duringCesarean delivery indicated at least transient neuro-logic sequelae [4].

Oxytocin and tocolytics were not used in ourcase although oxytocin has been used in patientswith intracranial tumors without any adverse effects.Although osmotic diuresis with mannitol is routinelyused to decrease brain bulk and intracranial pressure,we did not use them in our case as mannitol has beenshown to cross the placenta, and it may accumulate inthe fetus, leading to changes in fetal osmolality, volumeand the concentrations of various electrolytes.

Dexamethasone was used in our case to decreasecerebral edema. Its acute use may be safe for thefetus but chronic use of corticosteroids may result infetal adrenal suppression and fetal hypo-adrenalism,particularly during the third trimester. It is believedthat the administration of high-dose steroids for atleast 24–48 hours would facilitate fetal lung develop-ment for premature delivery of the fetus. Our patientreceived lorazepam for the treatment of her seizures;eclampsia must also be investigated with any presen-tation of seizures in the parturient.

Mechanisms of anesthesia-induced neurotoxicityand selectivity of anesthesia-induced neurodegener-ation are actively being investigated. It has been sug-gested that anesthesia-induced gamma-aminobutyricacid type A (GABA-A) receptor activation andNMDA-receptor blockade during a critical stage inbrain development lead to depression of neuronal

activity, which initiates the apoptotic cell death cas-cade in immature neurons [5]. Several adjuvants, suchas estradiol, pilocarpine, melatonin, and dexmedeto-midine, have been identified in animal studies toameliorate anesthesia induced neurodegeneration[3]. Repeated evidence for clinical doses of isofluraneleading to a dramatic increase in neuronal apoptoticcell death in animal models raises serious concernsfor anesthesia practice. The neurodegenerative effectsof etomidate, desflurane, and sevoflurane have yet tobe closely studied, whereas there is evidence from onestudy that the rarely used anesthetic, xenon, in clinicaldoses does not have neurodegenerative effects andmay be neuroprotective [6].

ConclusionManagement of brain tumors in pregnant women ismainly reliant on case reports and inherited wisdom.Therefore, close communication among the neurosur-geon, neuroanesthesiologist, obstetrician, and patientis of paramount importance.

References1. K. S. Tewari, F. Cappuccini, T. Asrat et al. Obstetric

emergencies precipitated by malignant brain tumors.Am J Obstet Gynecol 2000; 182: 1215–21.

2. T. Loop,H. J. Priebe. Recovery after anesthesia withremifentanil combined with propofol, desflurane, orsevoflurane for otorhinolaryngeal surgery. AnesthAnalg 2000; 91: 123–9.

3. V. Jevtovic-Todorovic. General anesthetics and thedeveloping brain: friends or foes? J NeurosurgAnesthesiol 2005; 17: 204–6.

4. K. Eishima. The effects of obstetric conditions onneonatal behaviour in Japanese infants. Early HumDev 1992; 28: 253–63.

5. T. Gerstner, S. Demirakca, T. Demiracka et al.Psychomotorische Entwicklung nach neonatalerPhenobarbitaltherapie.Monatsschr Kinderheilkd 2005;153: 1174–81.

6. A. Fredriksson, T. Archer,H. Alm et al.Neurofunctional deficits and potentiated apoptosis byneonatal NMDA antagonist administration. BehavBrain Res 2004; 153: 367–76.

220


67 Eclamptic seizuresNegmeldeen F. Mamoun

Eclampsia refers to the occurrence of one or moregeneralized convulsions and/or coma in the setting ofpreeclampsia, and in the absence of other neurologicconditions. It occurs with an incidence of 5 cases per10 000 live births in developed countries. It usuallydevelops after 20 weeks of gestation and just over one-third of cases occur at term, usually developing intra-partum or within 48 hours of delivery.

Case descriptionThe patient was a 23-year-old G4 P1 female who was33 weeks pregnant with a history of poor prenatalcare and preeclampsia during previous pregnancies.She presented with worsening headaches, generalizededema, and blood pressure of 195/115 mmHg. Thepatient was given a loading dose of 6 gmagnesium sul-fate (MgSO4), followed by infusion of 2 g/hour for pro-phylaxis of imminent eclampsia. Blood pressure wasmanaged with labetalol infusion of 1 mg/min, in addi-tion to intermittent doses of hydralazine. Blood testsrevealed mild anemia, normal electrolytes and creati-nine, platelet count of 95 000/�L, partial thromboplas-tin time (PTT) of 36, International Normalized Ratio(INR) of 1.3, and elevated transaminases.The plan wasto manage the patient expectantly, but she was givenbetamethasone to promote fetal lung maturity in casepreterm labor occurred.

The patient had a brief episode of generalizedtonic-clonic convulsions a few hours later; she wasgiven supplemental O2, placed in left uterine dis-placement position, and given a bolus of 2 g MgSO4.This episode was associated with fetal bradycardia thatlasted for a few minutes, but loss of beat-to-beat vari-ability persisted, which along with poorly controlledblood pressure, urged the obstetrician to proceed withCesarean section.

The anesthesiologist favored spinal over generalanesthesia despite mild coagulopathy; however, whilepositioning the patient on the OR table, she had

a second episode of generalized convulsions. Rapidsequence induction with cricoid pressure was per-formed using thiopental and succinylcholine, whichterminated the seizurewithin seconds.The tracheawasintubated with the aid of a Glidescope (video laryn-goscopy), after a failed first attempt of intubation usingdirect laryngoscopy. Anesthesia was maintained with0.5 minimum alveolar concentration (MAC) of isoflu-rane and a low birth weight baby was delivered withApgar scores of 7 and 9 at 1 and 5minutes respectively.

The patient had a brief episode of subtle facial andlimb twitches about 20 minutes after induction, rais-ing concerns that she might be experiencing anotherepisode of eclamptic seizures. Due to the recurrentnature of her convulsive episodes, continuous elec-troencephalogram (EEG) monitoring was urgentlyrequested. Emergence was associated with anotherepisode of generalized convulsions, which was evi-dent both clinically and on EEG (Figure 67.1a–d).This episode was quickly controlled by ventilatingthe patient with 2 MAC of isoflurane until isoelec-tric silence was achieved. Propofol infusion was thenstarted, aiming to replace isoflurane while maintain-ing isoelectric silence.

The patient was transported to the intensive careunit on a propofol infusion; multiple fluid bolusesand intermittent phenylephrine infusion were usedto support her blood pressure. The propofol infusionwas discontinued 24 hours later, with no evidence ofseizure activity, and the trachea was extubated a fewhours thereafter with no problems. MgSO4 infusionwas discontinued 48 hours postpartum due to clinicalimprovement. The patient did not have any more con-vulsive episodes and was not maintained on any long-term antiseizure medication.

DiscussionThe exact cause of seizures in eclampsia is not known.Two hypotheses have been proposed: (1) cerebral

221


(a)

(b)

Figure 67.1a–d. Progression of an eclamptic seizure during emergence.

222

Case 67. Eclamptic seizures

(c)

(d)

Figure 67.1a–d. (cont.)

223


overregulation in response to high blood pressureresults in vasospasm of cerebral arteries, localizedischemia, and intracellular edema; (2) loss of autoregu-lation of cerebral blood flow in response to high bloodpressure results in hyperperfusion, and vasogenicedema.

Eclampsia is a clinical diagnosis, presenting withgeneralized tonic-clonic convulsions that are usuallyself-limiting, lasting �3–4 minutes (average of 60–75 seconds), butmay be recurrent. Premonitory symp-toms include persistent frontal or occipital headache,blurred vision, photophobia, altered mental status,right upper quadrant or epigastric pain. Maternalcomplications include abruptio placentae, dissemi-nated intravascular coagulopathy, acute renal failure,liver rupture, intracerebral hemorrhage, cardiorespi-ratory arrest, aspiration pneumonitis, and acute pul-monary edema.

Eclamptic seizures are usually associated with fetalbradycardia, but may also be associated with compen-satory fetal tachycardia and loss of beat-to-beat vari-ability, or transient late decelerations. This does notnecessitate emergent delivery; stabilizing the mothercan help the fetus recover from the effects of mater-nal hypoxia and hypercarbia. If the fetal heart ratetracing remains non-reassuring for more than 10–15minutes despitematernal and fetal resuscitative efforts,emergent delivery should be considered. Continuousmaternal–fetal monitoring is indicated intrapartum toidentify worsening hypertension, deteriorating hep-atic, renal, cardiopulmonary, or hematologic function,and uteroplacental insufficiency.

Eclamptic seizures are clinically and electroen-cephalographically indistinguishable from other gen-eralized tonic-clonic seizures. Other etiologies are par-ticularly important in pregnant women with atypicaleclampsia, such as patients who seize before 20 weeksof gestation, or patients with focal neurologic deficits.Differential diagnosis includes:� Stroke.� Hypertensive encephalopathy.� Infection (meningitis, encephalitis).� Idiopathic epilepsy.� Cerebral vasculitis.� Space-occupying lesions (brain tumor, abscess).� Metabolic disorders (hypoglycemia,

hyponatremia).� Use of illicit drugs (methamphetamine, cocaine).� Reversible posterior leukoencephalopathy

syndrome (RPLS).

In addition to the management principles that applyto other seizures with different etiologies such as pre-vention of hypoxia, trauma, and recurrent seizures,management of eclamptic seizures includes controlof severe hypertension if present, and evaluation forprompt delivery.

General principlesMaintenance of airway patency and prevention ofaspiration should be the first priority. SupplementalO2 should be administered, and the patient shouldbe placed in a left uterine displacement position toimprove uteroplacental perfusion. A bed with raised,padded side rails provides protection from trauma.

Treatment of hypertensionAntihypertensive therapy is recommended for sus-tained diastolic blood pressures ≥110 mmHg or sys-tolic blood pressures≥160mmHg to preventmaternalcomplications. Although clinical trials have not ade-quately addressed the question of how aggressively tolower a preeclamptic patient’s blood pressure, expertsconsider systolic blood pressure of 140–160 mmHgand diastolic blood pressure of 90–110 mmHg to bea reasonable goal; their rationale is to avoid potentialreduction in either uteroplacental blood flow or cere-bral perfusion pressure [1].

The most frequently used antihypertensives arelabetalol and hydralazine; labetalol is preferredbecause it is associated with less maternal hypoten-sion. Nifedipine, nicardipine, � methyl dopa, anddiazoxide are less frequently used. It is not recom-mended to use calcium channel blockers and MgSO4concurrently due to their synergistic depressive effectson cardiac function.

Treatment of convulsionsNoneclamptic status epilepticus is traditionallytreated with four main categories of drugs: benzo-diazepines, phenytoin, barbiturates, and propofol.Benzodiazepines are considered the first-line treat-ment because they control seizures quickly, withlorazepam as the first-line drug. Lorazepam is moreeffective than diazepam in termination of seizures,and has a duration of action of 4–6 hours comparedwith 20–30 minutes after a single dose of diazepam[2]. Phenytoin or fosphenytoin are usually used as

224

Case 67. Eclamptic seizures

an adjunct to a benzodiazepine; they are effective inpreventing recurrence for extended periods of time.Propofol, barbiturates (pentobarbital and thiopental),and continuous infusion of midazolam are usuallyreserved for refractory status epilepticus. Generalanesthesia with isoflurane or other inhalational agentsmay be temporarily effective in stopping seizures, butis used only in extreme circumstances due to logisticalproblems.

In eclamptic seizures, obstetricians usually favorMgSO4 as the drug of choice, whereas neurologiststend to favor other traditional anticonvulsants suchas lorazepam. There is strong evidence that MgSO4is more effective than other anticonvulsants in reduc-ing the risk of recurrent seizures in eclamptic women.MgSO4 halves the risk of eclampsia, and probablyreduces the risk of maternal mortality [3]. The Ameri-canCollege ofObstetricians andGynecologists recom-mends its use in women with severe preeclampsia, andthe World Health Organization, the Federation Inter-nationale de Gynecologie et d’Obstetrique, the Inter-national Society for the Study ofHypertension in Preg-nancy advocate its use in the prevention and treatmentof eclampsia.

The mechanism of action of MgSO4 as an anti-convulsant is not clearly understood. It is likely mul-tifactorial including both vascular and neurologicmechanisms such as vasodilatation of the cerebral vas-culature, inhibition of platelet aggregation, preventionof calcium ion entry into ischemic cells, or its role as aN-methyl-D-aspartate receptor antagonist [4].

A recommended loading dose of 4–6 g MgSO4 isgiven intravenously, usually followedby amaintenanceinfusion of 2 g/hour. The maintenance dose shouldbe adjusted in renal insufficiency, and is given only ifa patellar reflex is present, respiratory rate �12/min,and average urine output of �25 mL/hour. Recurrentseizures occurring in patients on maintenance MgSO4therapy can be treated with an additional bolus of2 gMgSO4. If two such boluses do not control seizures,othermedications should be used such as lorazepamordiazepam.

Eclamptic seizures are typically self limiting, andEEG monitoring is usually not required for manage-ment, but due to our patient’s recurrent convulsiveepisodes despite therapy, there were concerns that shemight be at a higher risk of developing cerebral hemor-rhage, which is the major cause of maternal mortalityin this patient population. Continuous EEG monitor-ing was not done solely for diagnostic purposes, but

mainly to ensure that the patient was not having recur-rent subclinical seizures postpartum.

Anticonvulsants are usually administered for24–48 hours postpartum. The optimal duration oftherapy has not been determined, but therapy is usu-ally discontinued in women who are clearly improvingclinically as evidenced by absence of symptoms (noheadache, visual disturbances, or epigastric pain), andsigns (sustained blood pressure �150/100 mmHg, orspontaneous diuresis �100 ml/hour for �2 hours)[5].

DeliveryThedefinitive treatment of eclampsia is delivery, whichreduces the risk of maternal morbidity and mortality.Maternal end-organ damage and nonreassuring testsof fetal wellbeing are indications for delivery at anygestational age. Antenatal corticosteroids (betametha-sone) should be administered to women less than34 weeks of gestation to promote fetal lung maturity.

General anesthesia in parturients is typicallymain-tained with �1 MAC of inhalational agents due todecreased anesthetic requirements during pregnancy.This relatively lower concentration of isoflurane wasprobably not enough to suppress one of our patient’srecurrent seizures; its clinical presentation with sub-tle twitches may be due to residual muscle relaxation.Prolongation of neuromuscular blockade is expectedin patients receiving MgSO4 therapy, which potenti-ates the effect of both depolarizing and nondepolariz-ing muscle relaxants.

Neuraxial anesthesia (spinal, epidural, or com-bined spinal epidural) is the anesthetic technique ofchoice in patients with severe preeclampsia in theabsence of coagulopathy. Hypotension is a major con-cern with neuraxial anesthesia, as those patients usu-ally have depleted intravascular volume despite totalbody fluid overload. Major concerns with generalanesthesia include: (1) increased risk of difficult intu-bation secondary to airway edema; difficult intubationshould be anticipated, and equipment formanagementof a difficult airway including emergent cricothyroido-tomy should be readily available; (2) exacerbation ofa poorly controlled blood pressure with intubation;(3) increased risk of aspiration; aspiration prophylaxis(bicitra andmetoclopramide) should be administered,and the airway is managed either awake or after rapidsequence induction.

225


ConclusionIn conclusion, eclampsia is associated with increasedrisk of maternal and fetal morbidity and mortal-ity. Aggressive attempts should be made to con-trol seizures and hypertension. MgSO4 is consideredthe drug of choice for prevention and treatment ofeclampsia.

References1. B. M. Sibai. Diagnosis, prevention, and management

of eclampsia. Obstet Gynecol 2005; 105: 402–10.

2. K. Prasad, P. R. Krishnan, K. Al-Roomi et al.Anticonvulsant therapy for status epilepticus. Br J ClinPharmacol 2007; 63: 640–7.

3. D. Altman, G. Carroli, L. Duley et al. Do women withpreeclampsia, and their babies, benefit frommagnesium sulphate? The Magpie Trial: a randomisedplacebo-controlled trial. Lancet 2002; 359: 1877–90.

4. A. G. Euser,M. J. Cipolla. Magnesium sulfate for thetreatment of eclampsia: a brief review. Stroke 2009; 40:1169–75.

5. C. M. Isler, P. S. Barrilleaux, B. K. Rinehart et al.Postpartum seizure prophylaxis: using maternalclinical parameters to guide therapy. Obstet Gynecol2003; 101: 66–9.

226


68 Postpartum headacheAlexandra S. Bullough

Headache is a common complaint in the postpartumperiod. Most cases of postpartum headache (PPH)are attributed to dural puncture, migraine, pneumo-cephalus, or nonspecific causes, but all differentialdiagnoses need to be considered [1, 2].

Spontaneous internal carotid artery dissection(ICAD) is a rare cause of PPH. An arterial dissectioninvolves an intimal layer tear and the resulting intra-mural hematomamay compress and distort the lumenresulting in local stenosis or thrombosis. The conse-quent hypoperfusion or subsequent distal emboliza-tion may lead to an ischemic stroke which is alreadya recognized risk in the puerperium [3].

Case descriptionA healthy 36-year-old G3 P1 female with no knownhistory of vascular disease or connective tissue disor-der was admitted at full term in spontaneous labor.A lumbar labor epidural catheter was placed withoutcomplications. Five hours after successful neuraxialanalgesia the patient delivered a healthy 3.8 kg infantwith only 20 minutes of expulsive effort. The epidu-ral catheter was removed with the tip intact and thewoman was discharged home the next day.

Four days postpartum the woman contacted theobstetric triage desk by telephone complaining of apersistent frontal headache radiating to the back ofher neck with increased pain on the left side. Therewas no definite postural component to the headacheand the patient did not complain of fever, photopho-bia, nausea, or any other focal neurologic symptoms.As she had received epidural analgesia, she was reas-sured and told to increase her oral fluid uptake, remainsupine, take regular simple analgesia and drink caf-feinated beverages for a suspected postdural punctureheadache.

On postpartum day 8, the woman presented toobstetric triage with a severe global headache. Both

anesthesiology and neurology teams were consulted.There was no postural component to the headacheand a postdural puncture headache was dismissed asa diagnosis.

Neurologic examination was normal. The patientunderwent magnetic resonance imaging (MRI) toexclude venous sinus thrombosis. The MRI revealednormal sinuses and no evidence of an intracranialmass, hemorrhage, or dural enhancement. However,an absent flow void pattern was noted in the leftinternal carotid artery. The MRI report also com-mented upon the presence of abnormal bilateral ver-tebral arteries, which was suggestive of fibromus-cular dysplasia but not conclusive without furtherinvestigations.

Further investigation of the left internal carotidartery with magnetic resonance angiography (MRA)revealed a narrowing of the left carotid artery 2 cmdistal to the bifurcation, consistent with a left ICAD.Follow-up MRI scanning revealed a healing but nottotal resolution of left internal artery integrity. Thepatient was commenced on anticoagulation ther-apy comprising low molecular weight heparin andcoumadin. While on anticoagulation therapy, thepatient complained of right-sided visual blurring andright-hand numbness that resolved spontaneouslyafter 5 minutes. She remained in the hospital for a fur-ther 3 days during which time her headache localizedto the occipital region and radiated to her neck.

Coumadin therapy was continued for 8 monthsuntil the left ICAD resolved, whereupon a long-termaspirin regimen was prescribed. An unpleasant com-plication of her dissection was the subsequent devel-opment of frequent complex migrainous headaches,which persisted for several days and involved right-hand and leg numbness as well as occasional visualdisturbances. The patient was placed on divalproexsodium which did provide some migraine painrelief.

227


DiscussionPregnancy and puerperium increase the risk for focalischemic cerebrovascular events. The hypercoaguablestate in pregnancy and the immediate puerperiummost certainly contributes to this risk. Some 60–80%of ischemic strokes occurring in pregnancy and puer-periumare believed to be due to arterial occlusions andapproximately 40% of these occlusions occur up to amonth later in the postpartum period [3].

Spontaneous ICAD is a rare cause of postpartumheadache; the mean age of patients with ICAD is40–46 years [4]. In most cases no specific etiologyis found, but ICAD has been associated with minortrauma to the head or neck, as well as connectivetissue disorders such as Ehlers–Danlos syndrome orMarfan’s syndrome. Other trigger factors may include:“headbanging,” prolonged telephone usage, chiroprac-ticmanipulations, whiplash, childbirth, and even turn-ing one’s head to the same side during breastfeeding.In most cases no specific etiology is found. An arterialintimal wall tear may also be associated with strain-ing andValsalvamaneuver during vaginal delivery. It isunclear whether the association between puerperiumand arterial dissection is causative or coincidental [3].

Extracranial ICAD usually presents as a headache,cervical pain, Horner’s syndrome, or pulsatile tinni-tus without cerebral ischemia. An ICAD headache istypically ipsilateral and hemicranial [5]. In this casethe patient initially complained of a frontal headache,which is usually associated with a dural puncture.All parturients who receive neuraxial labor anesthesiaand who later complain of a headache are usuallytreated for postdural puncture headache, but this rep-resents only one possible diagnosis for postpartumheadache (Table 68.1). In this instance, postduralpuncture headache was discounted early as there wasno postural element associated with the headache orany history of traumatic epidural placement.

The neurology team wanted to exclude venoussinus thrombosis and therefore rapidly requestedMRI,MRA, and magnetic resonance venography (MRV)investigations. Magnetic resonance imaging can visu-alize morphological details, while MRA and MRVreflect intraluminal bloodflow [5].Duplex ultrasonog-raphy may also be used to detect extracranial ICAD,but use is limited in intracranial internal carotidinvolvement and high cervical occlusions. A skilledoperator is paramount for ultrasound investigation ofICAD.

Table 68.1. Differential diagnoses of postpartum headache.

Causes Description of headache

Postdural puncture Fronto-occipital, throbbing innature, postural-relieved whensupine

Nonspecific coincidentalheadache

Migraine Generalized or unilateralthrobbing pain, visualdisturbances, nausea andvomiting, photophobia, may lastminutes or days

Preeclampsia/eclampsia Hypertension, proteinuria, seizure

Pregnancy-inducedhypertension

Throbbing, visual disturbances

Pneumocephalus afterepidural

Frontal, worsening headache, dulland persistent

Meningitis Sudden onset, severe andpersistent +/− fever, stiff neck,nausea and vomiting, behavioralchanges, altered level ofconsciousness

Cerebral vein thrombosis Intermittent, diffuse and“pounding,” other signs mayinclude papilledema, intracranialhypertension, focal neurologicdeficits, seizures, and altered levelof consciousness

Nontraumaticintracranial hemorrhage

Intracerebral Severe headache

Subdural Mild to severe, localized orgeneralized

Subarachnoid “Worst headache of my life,” allintracranial bleeds may beassociated with nausea, vomiting,neurologic sequelae and loss ofconsciousness

Cervicocephalic arterialdissection

Variable presentation, mild tosevere pain, ipsilateral or bilateral,transient ischemic attack, stroke,Horner’s syndrome, visualdisturbance due to ophthalmicartery occlusion

Cerebral vasculitis Constant headache +/− transientischemic attack, tinnitus, visualdisturbance

Reversible cerebralvasoconstriction

“Thunderclap” headache +/−confusion, visual disturbances,seizures

Cerebral tumor No unique characteristics,headache depends on locationand size of tumor

228

Case 68. Postpartum headache

Initial MRI findings reported that the left inter-nal carotid artery lacked a normal flow-void patternthroughout the petrous and cavernous portion on theleft side which could represent thrombosis or dis-section. On a subsequent MRA an abnormality wasnoted to begin after the level of carotid bifurcation andextend up to the cavernous segment.

Usual angiographic findings of extracranial ICADinclude a string sign, double lumen, as well as irreg-ular and tapered stenosis, which usually begins 2–3cm above the bifurcation as occurred in this case andextends to the base of the skull [4].

Fortunately MRV images demonstrated no evi-dence of dural venous sinus thrombosis. Afinal follow-upMRI scan 8months later showed a persistent asym-metrical small left internal carotid artery, unchangedsince the previous studies several months earlier.

Treatment usually comprises immediate anticoag-ulation to prevent cerebral ischemia which is usuallydelayed yet may appear up to 1month later. A study byLucas et al. [6] proposes that most cerebral ischemicevents after ICAD are embolic rather than hemor-rhagic in origin. The optimal guideline duration ofanticoagulant therapy is yet to be established but 3–6 months has been recommended. Duration is usu-ally determined by the recanalization of the artery asdemonstrated on MRI or duplex follow-up [4].

Surgical intervention in ICAD is only requiredwhen anticoagulant therapy does not prevent progres-sive cerebral ischemic events.

Medicationwas later prescribed to assist withman-agement of debilitating migraines. Divalproex sodiumis teratogenic and should not be prescribed for womenof child-bearing age. The patient stated she wantedno further children. Teratogenic effects are severe andinclude anencephaly and spina bifida.

ConclusionIn conclusion, postpartum headache caused by ICADis a rare yet treatable condition with a favorable prog-nosis when recognized. Noninvasive MRI scanning isthe investigation of choice due to high sensitivity andspecificity in detecting an ICAD. Therapy comprisesanticoagulation until recanalization of the internalcarotid artery is achieved followed by an antiplateletregimen, the duration of which is determined by thepatient risk profile. Complications such as ICAD serveas a reminder to consider a broad differential for theevaluation of postpartum headache, as catastrophicneurologic sequelae may occur [2].

References1. J. Waidelich, A. S. Bullough, J. M. Mhyre. Internal

carotid artery dissection: an unusual cause ofpostpartum headache. Int J Obstet Anesth 2008; 17:61–5.

2. G. A. Mashour, L. H. Schwamm, L. Leffert.Intracranial subdural hematomas and cerebralherniation after labor epidural with no evidence ofdural puncture. Anesthesiology 2006; 104: 610–12.

3. A. P. Gasecki,H. Kwiecinski, P. A. Lyrer et al.Dissections after childbirth. J Neurol 1999; 246:712–15.

4. J. Bogousslavy, P. A. Despland, F. Regli. Spontaneouscarotid dissection with acute stroke. Arch Neurol 1987;44: 137–40.

5. M. Zetterling, C. Carlstrom, P. Konrad. Internalcarotid artery dissection Acta Neurol Scand 2000; 101:1–7.

6. C. Lucas, T. Moulin,D. Deplanque et al. Strokepatterns of internal carotid artery dissection in40 patients. Stroke 1998; 29: 2646–8.

229

Part VII Neurologic sequelae in other patient populations. MiscellaneousCase

69 Increased intracranial pressurewith acute liver failureRyen D. Fons and Paul Picton

Acute liver failure (ALF) is a rare clinical syn-drome characterized by coagulopathy and hepaticencephalopathy.There aremultiple etiologies, the clin-ical course is variable, and the mortality rate is high.Despite recent advances in the development of hep-atic assist devices, the only therapy with proven sur-vival benefit is liver transplantation. According to theUS ALF Study Group database, approximately 45% ofadult patients with ALF recover with medical treat-ment alone, 25% receive liver transplantation, andthe remaining 30% die without liver transplantation[1]. Multisystem organ failure secondary to sepsis andcerebral herniation secondary to increased intracra-nial pressure (ICP) are the leading causes of death.The decision to place an ICP monitor to facilitategoal-directed management of intracranial hyperten-sion remains controversial.

Case descriptionThe patient was a 36-year-old female with newlydiagnosed depression who ingested approximately�100 tablets (500 mg each) of acetaminophen ina suicide attempt. She presented to a local hospitallater that evening with severe nausea, vomiting,and abdominal pain. There was no other significantpast medical history. Laboratory evaluation revealedaspartate aminotransferase (AST) of 218 IU/L, alanineaminotransferase (ALT) of 288 IU/L, InternationalNormalized Ratio (INR) of 1.6, and creatinine of1.7 mg/dL. Oral N-acetylcysteine and vitamin Kwere administered. Over the next 24 hours, her ASTincreased to 2078 IU/L, ALT to 1979 IU/L, withan INR of 2.6, prothrombin time (PT) 31.5, totalbilirubin of 2.9 and creatinine of 1.5. She developedlethargy and personality changes consistent withgrade 2 hepatic encephalopathy (Table 69.1) and wastransferred to a tertiary care center for intensive caremanagement and liver transplant evaluation.

One day after patient transfer (third day postacetaminophen ingestion), her AST was 6120, ALT

6233, INR 3.7, total bilirubin 4.8, creatinine 3.2, andplatelets of 64 K/mm3. Her mental status deterioratedto somnolence and profound disorientation (grade 3hepatic encephalopathy) and the patient was listed asUnited Network for Organ Sharing (UNOS) status IA.An ICP monitor was not placed in this patient due tothe risk of intracranial hemorrhage.

The patient underwent orthotopic liver transplan-tation (OLT) 4 days following initial acetaminophenoverdose. Intraoperatively she received 2 units packedred blood cells, 7 units fresh frozen plasma (FFP), and10 units of platelets. She remained in the intensive careunit for 5 days and was discharged home on postoper-ative day 13 without neurologic complications.

DiscussionAcute liver failure is a syndrome defined as the devel-opment of coagulopathy (INR ≥1.5) and hepaticencephalopathy within 26 weeks of the onset of jaun-dice in a patient without a prior history of liver dis-ease. Acute liver failure is further subdivided based onthe jaundice-to-encephalopathy interval with hypera-cute liver failure occurring ≤7 days, ALF between 8and 28 days, and subacute liver failure between 28 daysand 26 weeks (Table 69.2). While the overall mortal-ity of ALF without liver transplantation is �50%, therate of spontaneous recovery from acute liver failureis inversely proportional to the length of the jaundice-to-encephalopathy interval. The rate of spontaneousrecovery is 80–90% for hyperacute liver failure, 50–60% for acute liver failure, and 15–20% for subacuteliver failure (Table 69.2).

Common causes of acute liver failureDrug-induced hepatotoxicity remains the leadingcause of ALF in the USA with acetaminophen over-dose constituting approximately 50% of all cases [1].Other medications commonly implicated includeisoniazid, sulfonamides, phenytoin, disulfiram,

230

Case 69. Increased intracranial pressure with acute liver failure

Table 69.1. West Haven criteria for grading of hepatic encephalopathy and rate ofspontaneous recovery of hepatic function.

Grade Signs and symptomsRate of spontaneousrecovery

Grade 0 No signs or symptoms

Grade 1 Trivial lack of awarenessEuphoria or anxietyDecreased attention spanImpaired performance of addition

Grade 2 Lethargy or apathyMinimal disorientation to time or placeSubtle personality changeInappropriate behaviorImpaired performance of subtraction

Grade 3 Somnolence, but responsive to verbal stimuliConfusionGross disorientation

40–50%

Grade 4 Coma (unresponsive to verbal or noxiousstimuli)

�20%

65–70%

Table 69.2. Acute liver failure classification.

Acute liverfailuresubcategory

Onset ofjaundice toencephalopathy

Spontaneoussurvival%

Incidence ofclinically significantcerebral edema Common causes

Hyperacute 0–7 days 80–90% 24% Acetaminophen overdose,hepatitis A, hepatic ischemia

Acute 8–28 days 50–60% 23% Hepatitis B, drug reaction

Subacute 28 days to 26 weeks 15–20% 9% Drug reaction, indeterminate

troglitazone, propylthiouracil, bromfenac, and herbalsupplements [1]. Common viral causes include hep-atitis A and B, while infrequent viral causes includehepatitis C and E, cytomegalovirus, herpes-simplexvirus and Epstein–Barr virus. Other less frequentcauses include autoimmune hepatitis, “shock liver”(ischemic hepatitis), acute Wilson’s disease, acuteBudd–Chiari syndrome, lymphoma, and acute fattyliver of pregnancy. In approximately 15% of adultpatients, the cause remains indeterminate.

Neurologic manifestations of acute liverfailure: encephalopathy, cerebral edema,and increased intracranial pressureBy definition, ALF requires the presence of hep-atic encephalopathy. The prognosis for spontaneousrecovery of liver function decreases with increasingseverity of encephalopathy (Table 69.1). The exact

etiology of hepatic encephalopathy remains unclear,but increased levels of toxins (ammonia, mercap-tans, serotonin, c-aminobutyric acid, endogenous ben-zodiazepines, and tryptophan), altered neurotrans-mitter levels, glutamatergic receptor activation andchanges in GABAergic tone have been proposed [2].In contrast to chronic liver failure where encephalopa-thy is common but cerebral edema is rare, ALF isoften accompanied by cerebral edema, especially whenacetaminophenoverdose is the cause (26%of patients).Themechanismof cerebral edema ismultifactorial andincludes increased blood flow from disruption of cere-bral autoregulation, astrocyte swelling, and inflam-mation. Because the cranium is noncompliant, dra-matic increases in ICPmay result from small increasesin brain volume once compensatory mechanismshave been exhausted. Normal ICP is �15 mmHg foradults. An ICP �20 mmHg is defined as intracranialhypertension. Prolonged increases of ICP �40 mmHgand/or cerebral perfusion pressure �40 mmHg for

231

VII. Neurologic sequelae in other patient populations. Miscellaneous

�2 hours are associated with cerebral herniation andpoor prognosis for neurologic recovery following livertransplantation [3].

CoagulopathyIn addition to hepatic encephalopathy, ALF is definedby the presence of coagulopathy. The coagulopathyof ALF is multifactorial, including impaired synthe-sis of clotting factors and fibrinogen, increased periph-eral consumption, and thrombocytopenia. Correctionof elevated INR, thrombocytopenia, hypofibrinogen-emia, or specific clotting factors is not routinely rec-ommended due to the low rate (∼10%) of spontaneousand clinically significant hemorrhage in patients diag-nosed with ALF [3]. Additionally, correction of coag-ulopathy may mask the signs of spontaneous recoveryand lead to unnecessary liver transplantation. Empiricadministration of vitamin K is recommended for allpatients with ALF. If clinically significant bleedingdevelops or an invasive procedure (e.g., ICP monitor)is planned, correction of INR to ∼1.5, platelet countof �50 000/mm3, and fibrinogen to �100 mg/dL isrecommended [3]. If FFP does not correct the INRor the patient is unable to tolerate the volume load,recombinant factor VIIa (40 mcg/kg) may be consid-ered immediately prior to the planned procedure [4].Additionally, isovolumetric plasmapheresis may alsobe effective [5].

Intracranial pressure monitor placementIn the presence of coagulopathy associated with ALF,placement of an ICPmonitor to facilitate goal-directedmanagement of intracranial hypertension remainscontroversial.TheUSAcute Liver Failure Study Groupcurrently does not recommend routine placement ofICP monitors in all patients with ALF since survivalat 30 days was similar in patients with and withoutICP monitoring who underwent OLT [3]. However,they do recommend considering an ICP monitor forall patients with grade 3 or 4 hepatic encephalopathyfor whom OLT is planned [3].

Four types of ICP monitors are used in clini-cal practice: epidural, subdural bolt, intraparenchy-mal, and intraventricular. Ventriculostomy addition-ally allows for removal of cerebrospinal fluid to treatincreased ICP. The rate of hemorrhagic complicationsfrom placement of an ICP monitor is approximately10% in ALF despite reversal of coagulopathy [4]. The

risk of intracranial hemorrhage varies depending onthe site of monitor placement (epidural 4%, subdu-ral 20%, and intraparenchymal/intraventricular 22%).The risk of fatal hemorrhage for the various monitor-ing choices is 1%, 5%, and 4%, respectively [5]. TheUS ALF Study Group does not recommend intraven-tricular ICP monitor placement due to the high riskof hemorrhagic complications [3]. In general, epidu-ral catheters have the lowest complication rate, buttend to overestimate ICP. Subdural bolts are currentlythe most commonly placed device in the USA for ICPmanagement in patients diagnosed with ALF [4].

ConclusionIn conclusion, ALF is rare but carries a high mortality.The leading cause within the USA is acetaminophenoverdose. The associated encephalopathy seen withpatients diagnosed with ALF is commonly caused bycerebral edema. The combination of coagulopathy,increased ICP, and grade 3 or 4 hepatic encephalopa-thy presents a difficult and controversial dilemma forthe critical care physician. While placement of anICP monitor to facilitate goal-directed managementof intracranial hypertension may insure that cerebralperfusion is preserved, the risks of intracranial hemor-rhage (∼10%) from ICPmonitor placement and bloodproduct administration must be considered. This isespecially true given the lack of definitive evidence ofa benefit in patient mortality.

References1. W.M. Lee, R. H. Squires Jr., S. L. Nyberg et al. Acute

liver failure: Summary of a workshop. Hepatology2008; 47: 1401–15.

2. J. G. O’Grady. Acute liver failure. Postgrad Med J 2005;81: 148–54.

3. R. T. Stravitz, A. H. Kramer, T. Davern et al.Intensive care of patients with acute liver failure:recommendations of the U.S. Acute Liver Failure StudyGroup. Crit Care Med 2007; 35: 2498–508.

4. J. Vaquero, R. J. Fontana, A. M. Larson et al.Complications and use of intracranial pressuremonitoring in patients with acute liver failure andsevere encephalopathy. Liver Transpl 2005; 11: 1581–9.

5. R. T. Stravitz. Critical management decisions inpatients with acute liver failure. Chest 2008; 134:1092–102.

232


70 Permissive hypertension in a patient withvonWillebrand’s disease and a preexistingventriculoperitoneal shuntMiguel Cruz, Maged Guirguis and Wolf H. Stapelfeldt

Hypertension is not usually considered acceptable inpatients with abnormal coagulation, including vonWillebrand’s disease. The present case describes asituation in which significant permissive hyperten-sion may have prevented an otherwise likely adverseoutcome.

Case descriptionA 56-year-old male reporting a remote history ofstroke with residual left-sided weakness, prior radia-tion therapy of a pontine brain tumor, a seizure dis-order, ventriculoperitoneal (VP) shunt, and historyof von Willebrand’s disease underwent robot-assistedlaparoscopic radical prostatectomy for prostate cancer.A previous laparoscopic cholecystectomy performedelsewhere had been complicated by significant post-operative bleeding, requiring surgical drainage as wellas multiple blood product transfusions. Prophylac-tic pretreatment with intravenous desmopressin wascompleted 1 hour before the current surgery. Afterplacement of routine monitors, general anesthesia wasinduced and invasive blood pressure monitoring ini-tiated via a radial arterial catheter. The patient wasplaced into a steep Trendelenburg position. Immedi-ately following CO2 insufflation into the peritonealcavity, a dramatic and refractory blood pressure risewas noted with mean pressure exceeding 130 mmHgdespite adequate analgesia (including supplementaldoses of 0.8 mg hydromorphone) and sufficient anes-thetic depth based on end-tidal anesthetic concentra-tion (Figure 70.1, arrow 1), suggesting some underly-ing cause other than inadequate anesthesia. Becauseof the team’s concern about increased risk of cere-bral hemorrhage the possibility was considered toabort the surgical procedure. Release of the intra-abdominal insufflation pressure caused mean arterialpressure (MAP) to rapidly normalize into the 80–90 mmHg range (Figure 70.1, arrow 2). Based on this

observation it was reasoned that the patient’s bloodpressure increase might have represented a compen-satory response to the increased intracranial pres-sure occurring during abdominal inflation, conceiv-ably brought about by a malfunctioning valve of theVP shunt, transmitting increased abdominal pressureto the cerebrospinal fluid (CSF) and brain. A lateralskull X-ray was performed to rule out the possibility ofpressurized gas having passed into the cerebral ventri-cles along such a conceivable path. In the absence ofany evidence (Figure 70.2) the team resolved to cau-tiously proceed with the prostatectomy while havingcerebral blood flow continuously monitored by tran-scranial Doppler. Immediately upon renewed infla-tion of the peritoneal cavity, the patient’s systemicblood pressure again rose precipitously (Figure 70.1,arrow 3). Doppler measurements confirmed that cere-bral bloodflowwas adequatelymaintainedunder theseconditions with mean velocities ranging between 34–46 cm/second (Figure 70.3). The decision was made tocomplete the surgical procedure while keeping peri-toneal insufflation pressure at a minimally acceptablelevel (around 8 mmHg) (Figure 70.1, arrow 4). For thebalance of the case the patient’s mean arterial bloodpressure remained within the 80–100 mmHg range.Transfusion of platelets became necessary to controlexcessive and persistent bleeding from the prostatebed. After completion of surgery the patient was per-mitted to awaken from anesthesia and to be tracheallyextubated after exhibiting the ability to sustain head liftand follow commands. There was no evidence for anyadverse event or new neurologic deficit resulting fromeither cerebral ischemia or hemorrhage, which mightconceivably have occurred during the above episodes.After the patient’s return to the operating room thatsame evening to undergo an exploratory laparotomyfor continued postoperative intra-abdominal hemor-rhage from the resection site, the patient was ulti-mately discharged from the hospital without furthercomplication.

233


Figure 70.1. Intraoperative vital signs, with arrows indicating changes in intra-abdominal pressure caused by inflation (up) and deflation(down).

Figure 70.2. Lateral radiograph of theskull, without evidence of intraventricularair.

234

Case 70. Permissive hypertension with vonWillebrand’s disease

Figure 70.3. Transcranial Doppler recordings, showing adequate cerebral blood flow signal tracings.

DiscussionThe presence of a VP shunt is not considered an abso-lute contraindication for laparoscopic surgery pro-vided the shunt and its valvular mechanism are fullyfunctional [1]. However, increases in intracranial pres-sure (ICP) are well known to occur during laparoscopyin VP shunt patients with less than normal cerebralcompliance [2]. While there was no evidence preoper-atively to suggest the presence of VP shunt malfunc-tion or of decreased cerebral compliance, the rapiddevelopment of a reversible and reproducible bloodpressure increase following the institution of the pneu-moperitoneum suggested that the VP shunt valvu-lar function was likely faulty, allowing the increase inintra-abdominal pressure during CO2 insufflation to

be transmitted to the central nervous system. This, inturn, presumably caused a compensatory increase inmean arterial pressure in order to maintain sufficientcerebral perfusion pressure and blood flow. Absenceof CO2 gas pockets in the cerebral ventricles on thelateral skull X-ray did not rule out the possibility ofvalvular incontinence of the VP shunt since the patientwas in a rather steep Trendelenburg position, presum-ably causing any pressurized gas within the shunt tobe kept floating above a contiguous CSF fluid columnthat would have precluded any further tracking of gaswhile still effectively transmitting increased pressureto the cerebral ventricles. As soon as the rising ICPwould have exceeded venous pressure it would have

235


effectively limited cerebral perfusion pressure (CPP =MAP– ICP), creating a (compensatory) need forMAPto increase proportionally in order to maintain CPP.Cushing in 1901 described the classical triad ofphysiologic responses to intracranial hypertension,consisting of apnea, increased blood pressure, andbradycardia. More recent observations during neu-roendoscopic procedures under general anesthesiadescribe the rapid development of hypertension with-out concomitant bradycardia in response to increasedICP [3]. Based on the same physiologic rationale theanesthesia team considered it absolutely imperative toallow blood pressure to rise (“permissive hyperten-sion”) during periods of increased peritoneal pressurerather than trying to normalize blood pressure. How-ever, the question arose if this would be safe in thepresent setting of abnormal coagulation due to vonWillebrand’s disease.

Von Willebrand’s disease is known to carry anincreased risk of bleeding, including that of cerebralhemorrhage [4, 5]. Is this risk increased in the pres-ence of hypertension? It could be argued that the riskof bleeding was undoubtedly increased in areas notaffected by increased pressure during abdominal gasinsufflation. However, transmural pressure (the pres-sure gradient across the blood vessel wall between thevessel lumen and surrounding tissue) and the relatedrisk of hemorrhagewould not have been elevated to thesame extent in all those tissues experiencing increasedambient tissue pressure from abdominal gas insuffla-tion, including the surgical site itself and, indirectly,in this case, brain tissue, thus mitigating an otherwiseincreased risk of hemorrhage at these sites. On theother hand, the actual risk of hemorrhage at the sur-gical site was certainly higher in the presence of com-pensatory hypertension than it would have been oth-erwise, thus likely contributing tomore severe surgicalbleeding and, ultimately, the need for platelet trans-fusion. Yet, given the risk of cerebral ischemia (con-sequent to the rise in ICP) it was most likely prefer-able to accept an increased likelihood of abdominalhemorrhage at the surgical site rather than risking a(more likely ischemic rather than hemorrhagic) insultin the CNS. This anesthetic assessment and manage-ment plan was thoroughly discussed with the surgicalteam as part of the decision to proceed with surgery.The question arose as to whether the compensatoryblood pressure increase occurring during abdominalinsufflation would, in fact, be sufficient to maintainadequate cerebral perfusion. To address this critical

question, continuous transcranialDopplermonitoringwas instituted in an ad hoc fashion, demonstrating themaintenance of adequate cerebral blood flow duringperiods of abdominal inflation. Utility of this technol-ogy has been previously advocated in a similar clinicalsetting [6].

While unrecognized shunt malfunction was mostlikely already present preoperatively, it could possi-bly have developed de novo during the laparoscopy[7]. Because of these various possibilities of shunt fail-ure, it has been suggested that VP shunts should beexternalized prior to laparoscopy as a precautionarymeasure [8]. Another approach might have been toattempt to laparoscopically clamp off the distal endof the VP shunt tubing, followed by cutting the dis-tal end at the conclusion of the procedure to reestab-lish CSF drainage. Problems with the latter approachmight have included technical difficulties locating theend of the shunt tubing in the upper abdomen with-out having to completely disengage the robot, as wellas the risk of gradual formation of a hydrocephalusthatmight conceivably have ensued over the prolongedsurgical course typical for robot-assisted laparoscopicprostatectomy. Ultimately, it is most important to con-sider this entire range of possible complications andmanagement options in VP shunt patients undergoinglaparoscopic procedures, which require coordination,advanced planning, and continued close cooperationbetween the surgical and anesthesia teams if poten-tially serious adverse sequelae are to be averted.

References1. S. V. Jackman, J. D. Weingart, S. G. Docimo.

Laparoscopic surgery in patients withventriculoperitoneal shunts: safety and monitoring.J Urol 2000; 164: 1352–4.

2. R. G. Uzzo,M. Bilsky,D. T. Minimberg et al.Laparoscopic surgery in children withventriculoperitoneal shunts: effect ofpneumoperitoneum on intracranial pressure –preliminary experience. Urology 1997; 49: 753–7.

3. A. F. Kalmar, J. Van Aken, J. Caemaert et al. Value ofCushing reflex as warning sign for brain ischaemiaduring neuroendoscopy. Br J Anaesth 2005; 94: 791–9.

4. W. S. Almaani, A. S. Awidi. Spontaneous intracranialhemorrhage secondary to von Willebrand’s disease.Surg Neurol 1986; 26: 457–60.

5. R. Nakau,M. Nomura, S. Kida et al. Subarachnoidhemorrhage associated with vonWillebrand’s disease –case report. N Neurol Med Chir 2005; 45: 631–4.

236

Case 70. Permissive hypertension with vonWillebrand’s disease

6. J. Ravaoherisoa, P. Meyer, R. Afriat et al.Laparoscopic surgery in a patient with ventriculoperi-toneal shunt: monitoring of shunt function withtranscranial Doppler. Br J Anaesth 2004; 92: 434–7.

7. J. J. Baskin, A. G. Vishteh,D. E. Wesche et al.Ventriculoperitoneal shunt failure as a complication

of laparoscopic surgery. JSLS 1998; 2:177–80.

8. J. A. Brown,M. D. Medlock,D. M. Dahl.Ventriculoperitoneal shunt externalization duringlaparoscopic prostatectomy. Urology 2004; 63:1183–5.

237


71 Perioperative acute ischemic strokein general surgical proceduresMilad Sharifpour and George A. Mashour

Perioperative acute ischemic stroke (AIS) is a fearedcomplication of surgery that is associated withincreased in-hospital mortality, length of hospital stay,disability, and discharge to long-term care facilities[1].The risk for AIS depends on a patient’s preexistingco-morbidities, as well as the type and complexityof the planned surgical procedure. The incidence isreported to range from almost 1.0% in general surgi-cal procedures to 4.5% in cardiovascular procedures[2, 3].

Case descriptionThe patient was a 70-year-old female with a past med-ical history of hypertension, hypercholesterolemia,mitral stenosis, chronic atrial fibrillation (requir-ing anticoagulation therapy with coumadin), diabetesmellitus, myocardial infarction, and morbid obesity.She underwent an elective left total hip arthroplastybecause of severe osteoarthritis. General anesthesiawith endotracheal intubation was planned and thepatient’s coumadin was discontinued 5 days prior tosurgery to allow the prothrombin time to return tonormal. The intraoperative course was complicatedby episodes of hypotension, which were treated withphenylephrine. The patient also had an episode ofatrial fibrillation, which was treated with incrementalboluses of esmolol. The remainder of the surgery wasuneventful and the case was completed with approx-imately 500 mL of estimated blood loss. Emergencewas smooth and the patient was neurologically intact.However, 3 hours after the procedure she had an acuteonset of delirium with a new left-sided hemiparesis. Anoncontrast computed tomography scan of the headdemonstrated loss of gray and white matter differen-tiation in the insular region, as well as a hyperden-sity in the right middle cerebral artery distribution,consistent with AIS. Since treatment with intravenoustissue plasminogen activator (tPa) is contraindicated

after major surgical procedures, the patient was takento the interventional radiology suite and endovascularmechanical thrombolysis was performed.

DiscussionPerioperative AIS is a serious complication of surgery.The majority of perioperative strokes are embolicand associated with systemic atherosclerosis. Availabledata indicate that approximately 45% of strokes takeplace during the first day after surgery while the restoccur after recovery from anesthesia, from the sec-ond postoperative day onward [1]. A recent large-scalestudy found an incidence of 0.7% after hemicolectomy,0.2% after hip arthroplasty, and 0.6% after lobectomyor segmental lung resection [2]. Even though the riskof stroke during the perioperative period is seeminglylow, it is a significant source of increasedmorbidity andmortality. Therefore, it is important to identify modi-fiable risk factors for perioperative stroke and evalu-ate each patient’s risk–benefit ratio prior to surgery inorder to optimize care. The main risk factors for peri-operative stroke include (1) female sex, (2) advancedage, (3) atrial fibrillation, (4) cardiac valvular disease,(5) congestive heart failure, (6) history of previoustransient ischemic attack (TIA) or stroke, (7) renaldisease, (8) diabetes mellitus, (9) hypertension, and(10) general anesthesia [1, 2, 4]. While female sex isprotective against stroke in the general population, itis associated with a higher risk of stroke during theperioperative period [1, 2]. Advanced age is also asso-ciated with increased risk of perioperative stroke andBateman et al. found that the incidence of perioper-ative AIS after hemicolectomy, total hip arthroplasty,and lobectomy was higher in patients �65 years oldcompared with that in patients �65 years old [2].

Atrial fibrillation is a significant and potentiallymodifiable risk factor and was identified as the mostcommon co-morbidity associated with stroke after

238

Case 71. Perioperative acute ischemic stroke in general surgery

general surgery in a recent study of perioperativeischemic stroke [2]. It can lead to ischemic strokeby increasing the risk of cardioembolic events orby causing cerebral hypoperfusion in patients whodevelop rapid ventricular rate and hypotension.There-fore, patients should be closely monitored for peri-operative arrhythmias. Electrolyte abnormalities andshifts in the intravascular volume should be cor-rected aggressively, and hyperadrenergic states andpulmonary complications should be treated promptlysince they predispose patients to atrial fibrillation.

Prophylactic beta-adrenergic blocker therapy hasbeen shown to decrease the incidence of atrial fib-rillation after cardiovascular procedures. The POISEtrial, however, examined patients undergoing noncar-diac surgery and found a lower risk of atrial fibril-lation but a higher risk of perioperative stroke andoverall mortality associated with �-adrenergic blockerprophylaxis [5]. Currently the neuroprotective role ofprophylactic beta-adrenergic blockade in noncardiacsurgery remains unclear andmore studies are requiredto better define this role.

Valvular heart diseases also increase the risk ofperioperative stroke. Diseased or mechanical valvesincrease the risk of cardioembolic events, or can lead todecreased cardiac output and cerebral hypoperfusion.Furthermore, certain valvular diseases such as mitralstenosis increase the risk of atrial fibrillation and con-sequently the risk of perioperative stroke. Left ven-tricular dysfunction is another potentially modifiablerisk factor of perioperative stroke. Preoperative evalu-ation with echocardiography may be required in orderto assess left ventricular function and the presence ofintracardiac emboli.

Proper anticoagulation is essential in order todecrease the risk of embolic stroke associated withatrial fibrillation and valvular heart disease. How-ever, the potentially protective effect of anticoagulationshould be weighed carefully against the risk of bleed-ing during the perioperative period. This is particu-larly important in patientswith preexisting atrial fibril-lation, in whom abrupt discontinuation of coumadintherapy leads to increased risk of perioperative stroke[1, 2, 4].

When oral anticoagulation must be withheld inanticipation of surgery, it is recommended that thetime period during which anticoagulation is beingwithheld should be minimized and bridge therapywith heparin, as well as postoperative anticoagulation

with coumadin, should be started as early as possi-ble. This is of paramount importance in patients withincreased risk of thromboembolism, namely thosewith chronic atrial fibrillation, valvular heart diseaseormechanical valves, or thosewith left ventricular dys-function.

Patients with a history of TIA or stroke are at asignificantly higher risk of perioperative stroke. Thus,anesthesiologists must specifically inquire about his-tory of such events and carefully assess neurologic sta-tus during the preoperative evaluation and documentany deficits. Furthermore, they should investigate andpossibly treat the causes of TIA and stroke prior tosurgery.

Renal disease, diabetes mellitus, and hypertensionare among the known risk factors of stroke in the gen-eral population and are also associated with increasedrisk of perioperative stroke. Renal disease predisposespatients to accelerated atherosclerosis and dialysis-dependent patients are at increased risk of hypoten-sion and fluid shifts. Intraoperative and postopera-tive hyperglycemia is associated with increased riskof atrial fibrillation, stroke, and death. It is suggestedto maintain blood glucose below 140 mg/dL to min-imize the risk of these events. The optimal bloodpressure during surgery is undetermined. However, ithas been reported that deviations more than 20% or20 mmHg from the preoperative baseline blood pres-sure for prolonged periods leads to increased periop-erative complications.

ConclusionDespite advances in surgical technique and periopera-tivemonitoring, stroke remains a significant complica-tion of general surgical procedures with considerablemorbidity and mortality. Preoperative evaluation ofpatients should focus on identifying and correctingpotentially modifiable risk factors to reduce the risk ofthis devastating complication.

References1. M. Selim. Perioperative stroke. N Engl J Med 2007;

356: 706–13.2. B. T. Bateman,H. C. Schumacher, S. Wang, S. Shaefi,

M. F. Berman. Perioperative acute ischemic stroke innoncardiac and nonvascular surgery. Anesthesiology2009; 110: 231–8.

239


3. J. Bucerius, J. F. Gummert,M. A. Borger et al. Strokeafter cardiac surgery: a risk factor analysis of 16,184consecutive adult patients. AnnThorac Surg 2003; 75:472–8.

4. V. Szeder,M. T. Torbey. Prevention and treatment ofperioperative stroke. Neurologist 2008; 14: 30–6.

5. P. J. Devereaux,H. Yang, S. Yusuf et al. Effects ofextended-release metoprolol succinate in patientsundergoing noncardiac surgery (POISE trial): arandomized controlled trial. Lancet 2008; 371:1839–47.

240


72 Neurologic complications followingcardiothoracic surgeryDonn Marciniak and Colleen G. Koch

Up to 79% of patients undergoing cardiac surgery uti-lizing cardiopulmonary bypass will demonstrate somedegree of neuropsychiatric dysfunction and up to 5.4%will manifest perioperative stroke [1]. These statisticsvary greatly in the literature and are influenced by pro-cedure type and co-morbidities, but reflect a commonproblem in patients undergoing cardiac surgery. Peri-operative neurologic disturbances are a major cause ofmorbidity andmortality in the cardiac surgery popula-tion and numerousmethods and techniques have beendeveloped to help minimize these occurrences.

Case descriptionThe patient was a 34-year-old male seen in consulta-tion for treatment of an ascending aortic aneurysmand fungal endocarditis. He had a mitral valve repairand an ascending aortic replacement with an aor-tic valve homograft 10 years prior to this admission.Approximately 6 months prior to the current admis-sion, the patient underwent cardiac surgery for aortichomograft failure. On this most recent admission, thepatient presented with fever, sternal wound drainage,and right leg pain. Computed tomography (CT) scanrevealed a false aneurysm of the distal anastomosis ofthe prior aortic graft to the aorta. Distal pulses in theright foot were diminished and he was diagnosed witha fungal thromboembolus to the right tibial artery.Thepatient underwent a third reoperation with ascendinganeurysm repair, with a distal ascending aortic graftand a cryopreserved homograft under deep hypother-mic circulatory arrest. Candida albicans was culturedfrom the graft material and the patient was placed onappropriate antibiotic therapy.

On postoperative day 1 the patient developedseizures and no demonstrable movement of his upperextremities. A 24-hour electroencephalogram (EEG)recorded seizures arising from the left hemisphere andthe occipital region. On discontinuation of propofol,the EEG demonstrated status epilepticus arising from

both hemispheres. A neurology consult attributed hisstatus epilepticus to suspected anoxic brain injury sec-ondary to intraoperative events. The seizure activ-ity resolved with appropriate therapy. A CT of thebrain revealed findings suggestive of diffuse cere-bral edema related to a global anoxic event and asmall acute anterior cerebral artery distribution cor-tical infarct without mass effect. Magnetic resonanceimaging revealed diffuse signal changes worrisome forglobal hypoxic injury. A follow-up CT of the brain3 days later reported a stable infarct, ischemia in thebifrontal lobes, and a punctate hyperdensity in theoccipital lobes. The patient required tracheostomy forrespiratory insufficiency, but was eventually weanedfrom ventilatory support. He demonstrated slow clin-ical improvement in neurologic activity, was eventu-ally less confused, and able to reply to commands; withphysical and occupational therapy, he slowly regainedupper extremity motor strength. He was discharged torehabilitation approximately 1 month after surgery.

DiscussionOur patient presented with both focal and global neu-rologic complications related to cardiac surgery. Fun-gal thromboemboli, air emboli, and circulatory arrestwere thought to contribute to his motor deficit, diffusehypoxic injury, altered sensorium, and seizure activ-ity. Risk factors for neurologic complications follow-ing cardiac surgery in general include patient char-acteristics such as age, hypertension, and operativefactors. Roach and colleagues reported the incidenceof adverse cerebral outcomes to be 6.1% following car-diac surgery [1]. Neurologic injury was divided intotype I, which constituted focal injury, stupor, or comaand type II, which constituted deterioration in intellec-tual function, memory deficit, or seizures. Predictorsof type I outcomes were proximal aortic atherosclero-sis, a history of neurologic disease, and advanced age.Predictors for type II outcomes were advanced age,

241


hypertension, pulmonary disease and excessive alco-hol use. Those who experienced adverse cerebral out-comes had higher in-hospital mortality and longerduration of hospitalization [1]. Because of variable def-initions of what comprises a neurocognitive deficitand differences in measuring techniques, there is greatvariance of rates reported in the literature.

Focal ischemia during cardiopulmonary bypassis often a consequence of an embolic phenomenonfrom gaseous or atheromatous debris. In this case,the patient was also at risk for infectious fungalemboli. Various methods to identify and protectagainst regional neurologic injury are in use todaybut many are of questionable efficacy. Epiaortic ultra-sound attempts to identify atheromatous regions ofthe aorta that can be avoided during the proce-dure, but data regarding its efficacy are mixed andembolic load may be unaffected [2]. This techniquecan evaluate transesophageal echocardiogram (TEE)‘blind spots’ in zones three and four of the aorta andthus help guide the sites of cannulation and clamp-ing. Surgical palpation of the ascending aorta is apoor indicator of the presence of atheroma. Changesin the site of cannulation may reduce the risk ofatheromatous embolization initially, but the risk of asand-blasting effect still exists. Improvements in thecardiopulmonary bypass machine, such as the incor-poration of a 25 �m aortic inflow-line filter, the useof membrane oxygenators, and cardiotomy suction fil-ters, are used in an attempt to reduce the embolicload delivered to the patient. Careful de-airing of thecardiac chambers with the help of TEE guidance isoften employed to decrease the gaseous embolic load.Avoiding nitrous oxide will also likely reduce the sizeof gaseous emboli. Carotid artery compression duringperiods of expected embolic showering is not likely tobe helpful.

Since our patient required surgery on the aorticarch, interruption of brain perfusion with circulatoryarrest was necessary. Fleck and colleagues reportedduration of circulatory arrest, in particular arresttimes �30 minutes, as the most important determi-nant of postoperative temporary neurologic dysfunc-tion (TND) defined as confusion, delirium, and agi-tation [3]. They reported an 18% incidence of TNDand a 21% incidence of combined temporary and per-manent neurologic deficit following surgery. Further-more, TND led to impaired functional recovery dueto impaired finemotor function and short-termmem-ory loss [3]. Intraoperative factors utilized in our case

to mitigate risk of neurologic injury included use ofretrograde cerebral perfusion, hypothermia, and TEEguided de-airing. Others have reported the prevalenceof TND following circulatory arrest as between 16%and 43% and dependent on the type of cerebral pro-tective strategies used [4, 5].

We used hypothermia as a technique for globalcerebral protection during circulatory arrest. Ourpatient was cooled to 18 ◦C to decrease neuronalmetabolic rate and therefore oxygen demand. Tem-perature was measured with a nasopharyngeal probeplaced before heparinizationwith the thought that thissite correlates reasonablywell with jugular venous bulbtemperature and is a good surrogate for more invasivelocations. Regional temperature differencesmay occurduring cooling and many centers pack the patient’shead in ice to compensate for this, but there are scantdata to support this maneuver and care must be takento avoid injury to the skin and eyes. Rewarming mustbe carefully controlled. If this process is too rapid, anovershootmay occur and an increased risk of neuronaldamage exists through mechanisms such as widen-ing of ischemic penumbra, accelerated free radical for-mation, and increased acidosis [6]. Limiting arterialinflow temperature to 37 ◦C is one possible maneuverto avoid this complication. Sodium thiopental admin-istration is often used during cooling tominimize neu-ronal damage by establishing an isoelectric state, alongwith a Bispectral Index value of zero, thereby furtherreducing metabolic demand. If thiopental is adminis-tered, it should be given 3–5minutes before circulatoryarrest.

We used retrograde cerebral perfusion as a brain-protective technique where the superior vena cava wascannulated and cooled; oxygenated blood flowed fromthe venous to arterial networks. The premise is thatnutrient-rich blood will reach the brain, toxic metabo-lites will be washed out, and regional warming willbe diminished. Additionally, air, atheroma, and in thiscase, fungal debris may have also been washed out.Antegrade cerebral perfusion may also be employedwhere blood flows in an anatomic manner, with thecarotid or innominate arteries cannulated. There isongoing debate as to whether one technique offers bet-ter outcomes than the other.

Hyperglycemia worsens both global and regionalneurologic injuries in cardiac surgery. Maintenance ofeuglycemia is desirable in all cardiac surgical patientsand more so in patients undergoing circulatory arrestconsidering their increased incidence of neurologic

242

Case 72. Neurologic complications following cardiothoracic surgery

insult [7]. Whether one is to allow tight or permis-sive glucose control is unclear, as hypoglycemiamay bemore deleterious than hyperglycemia during circula-tory arrest. Ultimately, blood glucosemust be carefullyregulated and it is likely a good practice to avoid intro-ducing unnecessary exogenous sources of glucose.

Near infrared spectroscopy (NIRS) is oftenemployed during circulatory arrest to allow assess-ment of regional oxygen saturation. Since thistechnology is highly regional in nature, it is spec-ulative to use the data to assess global cerebraloxygenation. Regardless, its use is well-establishedin pediatric cardiac surgery and less so in adultcirculatory arrest procedures [8]. Observing a dropin the NIRS may lead one to increase oxygen deliveryby introducing antegrade or retrograde cerebralperfusion (presuming it is not in place), increasingcerebral flow pressures, or administering red bloodcells. None of these approaches is benign in natureand great consideration must be used in allowingNIRS data to guide clinical management.

References1. G.W. Roach,M. Kanchuger, C. M. Mangano et al.

Adverse cerebral outcomes after coronary bypasssurgery. Multicenter Study of Perioperative IschemiaResearch Group and the Ischemia Research andEducation Foundation Investigators. N Engl J Med1996; 335: 1857–63.

2. G. Djaiani,M. Ali,M. A. Borger et al. Epiaorticscanning modifies planned intraoperative surgicalmanagement but not cerebral embolic load duringcoronary artery bypass surgery. Anesth Analg 2008;106: 1611–18.

3. T. M. Fleck,M. Czerny,D. Hutschala et al. Theincidence of transient neurologic dysfunction afterascending aortic replacement with circulatory arrest.AnnThorac Surg 2003; 76: 1198–202.

4. E. Apostolakis, E. N. Koletsis, F. Dedeilias et al.Antegrade versus retrograde cerebral perfusion inrelation to postoperative complications followingaortic arch surgery for acute aortic dissection type A. JCard Surg 2008; 23: 480–7.

5. A. Zierer,M. R. Moon, S. J. Melby et al. Impact ofperfusion strategy on neurologic recovery in acutetype A aortic dissection. AnnThorac Surg 2007; 83:2122–8.

6. A. M. Grigore, C. F. Murray, F. Ramakrishna et al.A core review of temperature regimens andneuroprotection during cardiopulmonary bypass:does rewarming rate matter? Anesth Analg 2009; 109:1741–51.

7. A. K. Lipshutz,M. A. Gropper. Perioperativeglycemic control: an evidence-based review.Anesthesiology 2009; 110: 408–21.

8. M. C. Taillefer, A. Y. Denault. Cerebral near-infraredspectroscopy in adult heart surgery: systematicreview of its clinical efficacy. Can J Anaesth 2005; 52:79–87.

243


73 Anesthetic management for subduralhematoma evacuation in a patient with aleft ventricular assist deviceMarcos Gomes, Hesham Elsharkawy, Endrit Bala and Ehab Farag

Developments in ventricular assist devices (VADs) andthe limited supply of donor hearts for transplantationhave made the former an important method of treat-ment for patients with end-stage heart failure. As theuse and life expectancy on a VAD increases, the prob-ability increases that the anesthesiologist will provideanesthesia for noncardiac surgery in such patients.Here we present the case of a patient who underwenttwo surgical procedures for subdural hematoma evac-uation while on left VAD (LVAD) support. Anestheticmanagement and potential problems such as coagula-tion status and hemodynamic stability in patients withan LVAD are presented and discussed.

Case descriptionThe patient was a 72-year-old male who presented foran emergent subdural hematoma decompression afterfalling at home the night before and becoming unre-sponsive that morning. A computed tomography (CT)scan of the head done in the emergency departmentshowed a large left subdural hematoma with approxi-mately 1.7 cmmidline shift, as well as evidence of her-niation. He had a past medical history significant fornonischemic cardiomyopathy, congestive heart failure,with an ejection fraction of 5%, status post pacemaker-defibrillator implantation 10 months earlier, type 2diabetesmellitus, chronic kidney disease (baseline cre-atinine 1.6), and an LVAD that had been implanted 4months earlier, which improved his ejection fractionto 20%. The patient’s LVAD was a model HeartMate R©

II, and required the patient to be continuously anti-coagulated on coumadin. Upon arrival in the oper-ating room and application of standard ASA moni-tors, an arterial line was placed and the patient hadhis defibrillator turned off for the procedure. The tra-chea had already been intubated in the emergencydepartment for airway protection; therefore, induc-tionwas performedwith sevoflurane inhalation as well

as intravenous rocuronium and fentanyl. A centralline and two extra large-bore peripheral intravenouscatheters were inserted afterwards. During the proce-dure, mean arterial pressures were maintained withnorepinephrine infusion with a mean arterial pres-sure goal of ∼80 mmHg. One unit of fresh frozenplasma and five units of platelets were given intra-operatively. Coagulopathy had also been treated inthe emergency department with fresh frozen plasmaand platelets, improving the International NormalizedRatio (INR) from 2.4 to 1.2. Two liters of crystal-loids were also given throughout the procedure andthe estimated blood loss was 200 mL. The patientwas transferred to the intensive care unit, intubated,and sedated. The following day, a control CT scan ofthe head showed a new hemorrhagic contusion withsurround edema in the posterotemporal lobe with aresidual 2.6 cm epidural hematoma, and a secondemergency evacuation was performed. For that proce-dure, a similar anesthetic technique was used, with theexception of using remifentanil infusion for analgesiainstead of intravenous fentanyl. During the procedurethe LVAD team was present to assist the managementof LVAD.The surgery took place uneventfully and thepatient returned to the cardiovascular intensive careunit. On postoperative day number 12, a tracheostomyand gastric tube placement were performed. Unfor-tunately, throughout the whole postoperative period,the patient never recovered from a neurologic pointof view, and the family decided to change his statusto comfort care without resuscitation. On postopera-tive day number 17, the LVAD was turned off and thepatient passed away.

DiscussionAccording to theAmericanHeartAssociation, approx-imately 500 000 new cases of congestive heart failureare diagnosed every year, affecting 4.6 million Ameri-cans [1]. It is estimated that 250 000–500 000 patients

244

Case 73. Anesthetic management in a patient with a ventricular assist device

in the USA, and approximately 2.2 million world-wide, are currently in the terminal phase of heart fail-ure and are refractory to maximized medical therapy.Each year approximately 30 000 patients are listed forcardiac transplantation, but only 3500 are performed,making it evident that there is a chronic shortage ofdonors [2]. Having said that, it is inevitable that wewill watch an increasing number of patients with ter-minal congestive heart failure present to our hospitalsin search of an alternative treatment.

The Randomized Evaluation of Mechanical Assis-tance for the Treatment of Congestive Heart failure(REMATCH) trial [3, 4], as well as other early clin-ical experiences [4], has shown that an implantableLVAD as a destination therapy prolongs and enhancesthe quality of life in heart failure patients [5]. Thishemodynamic restoration therapy has increased the 1-and 2-year survival rates compared with pharmaco-logic interventions alone. There is also evidence thatNewYorkHeart AssociationClass IV patients improveto Class I or II post-LVAD treatment.

As the number of patients chronically supportedwith long-term implantable devices grows, generalsurgical problems that are commonly seen in otherhospitalized patients are becomingmore common andwill eventually lead to an increase in the number ofpatients with LVADs coming in for noncardiac electiveor emergency surgery. Anesthesiologists are requiredto have knowledge of the function of these devicesand to understand its implications during anesthesiafor noncardiac surgery. There are currently differentmodels on the market. However, we are going to focusour discussion on the one used in our patient, theHeartMate R© II. This device is smaller, more durable,and promotes continuous flow, which is why it hasbeen used more often recently.

The HeartMate R© II consists of an internal bloodpump with a percutaneous lead that connects thepump to an external system driver and power source.The pump has an implant volume of 63 mL and gen-erates up to 10 L/min of flow at a mean pressure of100 mmHg. It drains blood from the left ventricle intoa mechanical pump, which then ejects the blood via aconduit that links to the ascending aorta. The pump isdriven by a small electric motor and the rotary actionof the single moving part is responsible for its effect.The LVAD is normally powered electrically and canbe connected to two rechargeable batteries worn in awaist pack, which could last for up to 3 hours. Theblood flow generated by the LVAD is nonpulsatile and

for this reason these patients require anticoagulation.Low postoperative mortality rate, low incidence ofadverse events, favorable low thrombogenicity and lowthromboembolic risk make the HeartMate R© II LVADan ideal device to be used as a bridge to transplantationand destination therapy as well [6]. A third indicationfor LVADs is the patient with potentially reversible car-diac ischemia resulting in severe cardiogenic shock. Inthis scenario, the LVAD can serve tomaintain systemiccirculation while the myocardium is allowed time torecover [1].

Four aspects have to be considered in the manage-ment of patients with LVADs presenting for noncar-diac surgery: (1) LVAD specialists, (2) power supplyand electromagnetic interference, (3) hemodynamics,and (4) anticoagulation [5].

The first priority for the anesthesiologist caringfor a patient with an LVAD is to identify the “LVADteam” in the institution. A specialized team of health-care professionals thatmay include cardiothoracic sur-geons, nurses, engineers, and cardiopulmonary perfu-sionists is usually responsible for the management ofLVAD patients in the vast majority of medical centers.This team is an indispensable and valuable resourcefor information regarding the details of LVAD man-agement.

Securing a reliable power supply to assure contin-uous operation of a mechanical assist device is thenext consideration for the LVADpatient presenting forsurgery, since the LVAD has a limited battery capacity(rechargeable batteries). An alternating current sourcein the operating room may be safe and more reli-able [1]. The potential for electromagnetic interfer-ence with LVAD function by external defibrillationor electrocautery should be recognized. As a result,device settings and connections may require adjust-ment, but it must be emphasized that such maneuversshould only be done with consultation or supervi-sion of the institution’s LVAD team.The manufacturerrecommends that the electric model be set to func-tion in the fixed-rate mode as opposed to the “fill-to-empty” (auto) mode during surgical procedures inwhich the use of electrocautery is anticipated. Bipolarcautery should be used if possible, since current flowsonly between the tips of the bipolar instrument, butbipolar cautery is impractical for many surgical proce-dures because it ismuch less powerful thanmonopolarcautery [7].

The pumping mechanism of the LVAD dependson both preload and afterload. These devices do not

245


obey Starling’s law with respect to stroke volumeor stroke work, and can only pump the deliveredvolume and therefore inadequate filling leads toinadequate flow. Optimal function is consequentlyachieved with increased intravascular volume anddecreased vascular resistance. Maintenance of ade-quate preload is thus critically important. Directdecreases in pump flow occur when preload declinesas a consequence of decreased venous return sec-ondary to increased venous capacitance (includingdrug-induced venodilatation), alterations in bodyposition that reduce venous return (e.g., lateral decu-bitus or reverse Trendelenburg position), inadequateadministration of intravenous fluids, or uncontrolledsurgical bleeding. Conventional inhalational andintravenous anesthetic techniques were well toleratedin these patients [8]. All commercially available assistdevices exhibit sensitivity to changes in afterload. Asa result, hypertension should be specifically avoidedbecause emptying of the LVAD is reduced by increasesin arterial pressure. Incomplete LVAD ejection notonly decreases forward flow but also promotes bloodstasis within the device and increases the risk ofthrombus formation, even in the presence of systemicanticoagulation. Systemic responses to laryngoscopyand surgical stimulus should be attenuated andavoidance of hypertension should be the primary aimin the perioperative management of these patients.Hemodynamically significant arrhythmias should betreated appropriately with pharmacologic or electricalmeans. External chest compression should be avoidedbecause of the risk of cannula dislodgement [9]. Theseobjectives may be achieved by assurance of adequateanesthetic depth using volatile anesthetics in combina-tion with an opioid or by the judicious administrationof arterial vasodilators (e.g., sodium nitroprusside,fenoldopam) to treat increases in arterial pressure.The drugs should be added cautiously, paying carefulattention to resultant increased venous capacitanceand decreased venous return. In the absence of hyper-tension, most cases of low LVAD flow can be correctedby volume expansion. The fluid, inotropic, andvasopressor requirements do not appear to be signif-icantly different than those required in other patientsundergoing similar procedures [8]. Right ventricledysfunction must also be considered and negativeinotropic drugs (e.g., volatile anesthetics, adrenocep-tor antagonists, calcium channel blockers), as well asfactors that increase pulmonary vascular resistance(e.g., hypoxemia, hypercarbia, acidosis), should be

avoided [7]. In such cases, a positive inodilator (e.g.,milrinone) or a selective pulmonary vasodilator(e.g., inhaled nitric oxide) may be required [7]. Inour case, we managed hypotension throughout thecase with norepinephrine, aiming at a target systolicblood pressure of 90 mmHg. Norepinephrine exertsits effects on alpha-1 and beta-1 adrenoreceptors,therefore it works as a vasopressor and an inotrope,usually with doses ranging from 0.05–1 mcg/kg/min.It is indicated for severe systemic hypotension, mostlyrelated to profound vasodilatation, and it should betapered off as the hemodynamics stabilize in orderto prevent prolonged peripheral vasoconstrictionresulting in limited perfusion to end organs [10].

Management of anticoagulant therapy is anothermajor issue that requires attention in the periopera-tive care of the LVAD patient. Experience with surgeryunder anticoagulant conditions such as the ones foundin patients with VADs is scarce. These patients requirepreoperative admission and conversion to heparinanticoagulation, which is stopped in the immediatepreoperative period. Most procedures can be delayeduntil coagulation parameters are optimized; however,in emergency operations, in which oral anticoagu-lants (coumadin) have not been stopped, transfusionwith fresh frozen plasma is required to acutely correctclotting parameters [1]. Transfusion of fresh frozenplasma and platelet concentrates can be started withliberal amounts at any time and will certainly relievebleeding. Postponing resumption of full anticoagu-lation is advisable because it may reduce bleedingcomplications without increasing the risk for throm-boembolism. On the other hand, in many institutions,treatment with heparin is indeed resumed to lowerthe risk of thromboembolism, when risk of bleedingis diminished [7].

ConclusionThe anesthetic management of patients with an LVADwho present for elective or emergent surgery repre-sents a unique challenge for the anesthesiologist. Adetailed understanding of the operation of the device,as well as the factors that affect its normal func-tioning are essential in order to promote success-ful surgical outcomes. A multidisciplinary approach,hemodynamic stability, and anticoagulation status arecrucial issues for safe perioperative care in VADpatients.

246

Case 73. Anesthetic management in a patient with a ventricular assist device

References1. A. E. Eckhauser,W. V. Melvin, K.W. Sharp.

Management of general surgical problems in patientswith left ventricular assist devices. Am Surg 2006; 72:158–61.

2. A. Garatti, G. Bruschi, T. Colombo et al. Noncardiacsurgical procedures in patient supported withlong-term implantable left ventricular assist device.Am J Surg 2009; 197; 710–14.

3. E. A Rose, A. C. Gelijns, A. J. Moskowitz et al.Long-term mechanical left ventricular assistance forend-stage heart failure. N Engl J Med 2001; 345:1435–43.

4. O. H. Frazier, C. Gemmato, T. J. Myers et al. Initialclinical experience with the HeartMate II axial-flowleft ventricular assist device. Tex Heart Inst J 2007; 34:275–81.

5. V. Kartha,W. Gomez, B. Wu et al. Laparoscopiccholecystectomy in a patient with an implantable leftventricular assist device. Br J Anaesth 2008; 100: 652–5.

6. R. John, F. Kamdar, K. Liao et al. Improved survivaland decreasing incidence of adverse events with theHeartMate II left ventricular assist device asbridge-to-transplant therapy. AnnThorac Surg 2008;86: 1227–34.

7. A. C. Nicolosi, P. S. Pagel. Perioperativeconsiderations in the patient with a left ventricularassist device. Anesthesiology 2003; 98: 565–70.

8. D. J. Goldstein, S. L. Mullis, E. S. Delphin et al.Noncardiac surgery in long-term implantable leftventricular assist-device recipients. Ann Surg 1995;222: 203–7.

9. S. T. Webb, V. Patil, A. Vuylsteke. Anaesthesia fornon-cardiac surgery in patient with Becker’smuscular dystrophy supported with a left ventricularassist device. Eur J Anaesthesiol 2007; 24:640–2.

10. N. C. Dang, Y. Naka. Perioperative pharmacotherapyin patients with left ventricular assist devices. DrugsAging 2004; 21: 993–1012.

247

Section

IINeurocritical care

Part VIII General topics in neurocritical careCase

74 HypotensionEdward Noguera

Hypotension is one of the most common findingsin the intensive care unit (ICU) patient and requiresprompt attention in order to avoid poor clinical out-comes.

Case descriptionA 42-year-old morbidly obese female (body massindex = 38) was transferred to the ICU after beingdiagnosed with a subarachnoid hemorrhage due to aruptured aneurysm of the right middle cerebral artery.Her pastmedical history was relevant for hypertrophiccardiomyopathy with an ejection fraction of 30%,hypertension, and Crohn’s disease. Her medicationsincluded digoxin, carvedilol, enalapril, and pred-nisone. Upon arrival to the ICU she was comatose andintubated. Her heart rate was 142 beats per minuteand irregular, mean arterial blood pressure (MAP)was 40 mmHg, temperature was 39.6 ◦C, and pulseoximeter saturation was 93%. She had an 18-gaugeperipheral line, 20-gauge radial artery line and a Foleycatheter. A resident physician in the emergency roomattempted central line placement before transportto the ICU, without success. After ICU admission apulmonary artery catheter (PAC) was inserted viaright internal jugular vein using ultrasound guidance.The first reading of her pulmonary capillary occlusionpressure (PCOP)was 10mmHgandher cardiac outputwas 3 liter/min. Transthoracic echocardiogram (TTE)was ordered to assess her myocardial function due toher past medical history, the potential for ventriculardysfunction due to the subarachnoid hemorrhageand to rule out other causes for hypotension-liketamponade. The patient’s hypotensive conditionwas corrected by fluid boluses to increase her fill-ing pressure, guided by PCOP and by starting thepatient on norepinephrine infusion to increase hercardiac output and maintain her perfusion pressureas well.

DiscussionClassically, blood pressure is determined by the prod-uct of cardiac output and systemic vascular resistance(BP=CO× SVR). Cardiac output itself is determinedby several interrelated factors: mainly preload, pumpfunction, and afterload. Systemic vascular resistanceis determined by a complex relationship of bloodviscosity, vessel length, and vessel radius. In approach-ing the patient with hypotension, a focused historyand physical examination should be performed, withattention to the “ABCs” of airway, breathing, andcirculation. The goal is to identify an immediatecause of hypotension based on physiologic principles.However, many cases such as the one described abovecan have multiple causes based on clinical history. Forexample, our patient could have been hypotensive dueto preexisting cardiomyopathy, ventricular dysfunc-tion due to subarachnoid hemorrhage, hypovolemia,or sepsis. In these situations, a systematic approachis essential and tools such as the PAC or echocardio-gram can be helpful. There is, of course, controversyregarding the value of the PAC in the ICU setting [1].

PreloadDisorders of preload relate primarily to hypovolemia,as can occur with inadequate oral intake, inadequatefluid administration, hemorrhage, excessive urine out-put, or insensible loss. In this situation physical exam-ination might reveal poor skin turgor or dry mucousmembranes; abnormal vital signs could include ortho-static hypotension. Pulmonary artery catheter datawould demonstrate low PCOP and low cardiac out-put, with possibly increased SVR as compensation.Although a measured pressure, PCOP is used as a sur-rogate for left ventricular end-diastolic volume. In apatient with an arterial line and endotracheal tube,high systolic pressure variation might be seen due tothe effects of positive pressure ventilation on venous

251

VIII. General topics in neurocritical care

return [2]. The treatment of choice in this situationis volume (crystalloid, colloid, blood), with temporiz-ing measures including vasopressors such as phenyle-phrine (alpha-1 agonist) and norepinephrine (alpha-1and beta-1 agonist).

Pump functionVentricular failure syndromes present with hypoten-sion due to inadequate ejection of blood. In thiscase patients may have a history of coronary arterydisease or heart failure. Physical examination signsmight include peripheral edema, jugular venous dis-tension, or crackles upon auscultation. Either TTE ortransesophageal echocardiogram are very good toolsto assess myocardial contractility and valvular func-tion. The presence of wall motion abnormality inan echocardiogram is considered the earliest sign ofmyocardial ischemia. Pulmonary artery catheter datawould include a low cardiac output but, unlike hypo-volemia, a high PCOP (since blood is not gettingejected). Vasopressors used for treatment in this sit-uation include dobutamine and dopamine (for beta-1adrenergic effects), as well as milrinone (a phosphodi-esterase inhibitor).

Dysrhythmias are also a cause of hypotension.Electrical cardioversion, defibrillation or pacing capa-bilities are indicated according to the type. Car-diac dysrhythmias can be potentiated or inducedby myocardial ischemia, hypoxia, acidosis, hypercar-bia, electrolyte disturbances and mechanical irrita-tion from intravascular devices in proximity to themyocardium.

Other causes of pump dysfunction that are exter-nal to the heart include pneumothorax and tampon-ade. Placement of a chest tube is indicated if a pneu-mothorax is suspected or documented; if this is notpossible, needle decompression can be performed bythe anesthesiologist. Our patient had an unsuccessfulattempt of central line placement and therefore wasat risk of pneumothorax. Cardiac tamponade can beanother cause of hypotension, especially in trauma set-tings, and can be manifested by severe hypotension,distant heart sounds, tachycardia, jugular venous dis-tension, and pulsus paradoxus. If one suspects cardiactamponade, TTE should be ordered and a pericardio-centesis is indicated [3]. A PAC evaluation would showequalization of pressures.

AfterloadAfterload can be affected by processes such as sep-sis, anaphylaxis, or (in part) neurogenic shock. In theseptic patient, hypotension is thought to be causedby inflammatory mediators involved in the immuneand humoral response to disseminated infection. Theincreased production of cytokines can cause severeperipheral vasodilation and severe hypotension. Pul-monary artery catheter numbers would reveal a poten-tially high cardiac output (unless there is myocardialstunning) and a very low SVR. Treatment involveseradication of the underlying infection and vaso-pressors of choice would be phenylephrine, nore-pinephrine, and vasopressin. Relative adrenal insuf-ficiency is commonly seen in septic patients, whichcould be another cause of hypotension during sepsis.Steroids are indicated in septic shock once adequatefluid resuscitation is provided and the patient is vaso-pressor dependent [4]. A recent underpowered trialdid not demonstrate that steroids are deleterious inseptic shock patients [5].

ConclusionThere are multiple causes of hypotension in the ICUpatient. Basic principles of physiology help with thedifferential diagnosis of hypotension. Promptmanage-ment of hypotension often requires invasive monitor-ing, fluid resuscitation, and the use of vasopressor orinotropic therapy.

References1. M. Hadian,M. R. Pinsky. Evidence-based review of

the use of the pulmonary artery catheter: impact dataand complications. Crit Care 2006; 10 Suppl 3: S8.

2. M. R. Minsky,D. Payen. Functional hemodynamicmonitoring. Crit Care 2005; 9: 566–72.

3. D. Spodick. Acute cardiac tamponade. N Engl J Med2003; 349: 684–90.

4. D. Annane, V. Sebille, C. Charpentier et al. Effect oftreatment with low doses of hydrocortisone andfludrocortisone in patients with septic shock. J AmMed Assoc 2002; 288: 862–71.

5. C. Sprung,D. Annane,D. Keh et al.Hydrocortisonetherapy for patients with septic shock. N Engl J Med2008; 358:111–24.

252


75 Mechanical ventilationPiyush Mathur and Vikram Dhawan

The need for mechanical ventilation in patients withacute or chronic neurologic disorders is not uncom-mon. Patients with neurologic disease are prone tohypoventilation, hypoxia, aspiration, atelectasis, andlung collapse. Decreased gag and cough reflex putsthese patients at high risk for aspiration. Inabilityto clear orotracheal secretions is common with poormental status.

Case descriptionA 55-year-old female presented to the emergencydepartment with sudden onset of severe headacheand deterioration in her level of consciousness. HerGlasgow Coma Scale on arrival to the emergencydepartment was 5 and the trachea was thereforeemergently intubated. After intubation, bilious fluidwas suctioned out of the endotracheal tube. Mechan-ical ventilation was initiated with initial settings ofsynchronized intermittent mandatory ventilation,fraction of inspired oxygen (FiO2) 50%, respiratoryrate of 12/minute, positive end expiratory pressure(PEEP) 5 mmHg, pressure control 15 cmH2O, andpressure support of 10 cmH2O. Chest X-ray revealed aright middle lobe infiltrate. Arterial blood gas (ABG)was pH 7.39, PaCO2 40, PaO2 156. A computedtomography scan of the brain showed diffuse sub-arachnoid blood in the right hemisphere. The patientwas taken to the angiography suite and an aneurysmwas coiled successfully with no intraoperative compli-cations. On arrival to the neurointensive care unit, thepatient had worsening oxygenation and ventilation.A repeat ABG revealed pH 7.22, PaCO2 61, PaO265 on 100% FiO2. A pulmonary artery catheter thatwas placed intraoperatively showed pulmonary arterywedge pressure of 8 mmHg. Repeat chest X-rayshowed development of bilateral infiltrates consistentwith acute respiratory distress syndrome. Mechanicalventilation mode was switched to pressure con-trol ventilation and PEEP incrementally increased to

15 cmH2Owith improvement in both oxygenation andventilation.

After enhancing the patient’s urine output withdiuretics, her oxygenation improved. She was subse-quently weaned to pressure support ventilation withfurther improvement in both mental and respiratorystatus. The trachea was extubated by postoperativeday 5. She developed worsening respiratory insuffi-ciency 6 hours postextubation. She was found to havepoor cough and gag reflexes and was unable to clearher secretions. A chest X-ray revealed a collapsed rightlung. She was nasotracheally suctioned and placedon a noninvasive continuous positive airway pressuremachine. Her mental status continued to decline anda decision was made to reintubate. Upon postintu-bation, both her mental status and respiratory statusgradually improved. Failure to wean the patient frommechanical ventilation led to a tracheostomy by post-operative day 10. She was subsequently discharged to along-termacute care facility for ventilatorweaning andrehabilitation.

Discussion

Modes of mechanical ventilationMultiple ventilation modes and nomenclatureschemes are in existence. No one mode of ventilationhas been proven to be superior. Mechanism of ven-tilation is characterized by the mode of ventilationdelivery and type of breath sequencing. The controlmode is the variable that is set on the ventilatoras the mechanism of delivery of ventilation (i.e.,pressure or volume). If the pressure mode is selectedthen the ventilator will generate breaths at that setpressure level and tidal volume can vary. If the volumemode is selected then the ventilator will deliver theset tidal volume and the pressure generated may vary[1].

253


Table 75.1. Classification of modes of ventilation based on mechanism of ventilation delivery and breath sequencing.

Mode Control setting Variable parameter Breath sequencing

Pressure control Pressure Volume Mandatory ventilator breaths

Volume control ventilation Volume Pressure Mandatory ventilator breaths

Assist control Pressure/volume Pressure/volume Mandatory ventilator breaths

Synchronized intermittent mandatoryventilation

Pressure/volume Pressure/volume Intermittent mandatory ventilatorbreaths+support spontaneous breaths

Pressure support ventilation/continuouspositive airway pressure

Pressure Volume Support spontaneous breaths

Three types of breath sequences are possible(Table 75.1):

1. Continuous mandatory: all breaths are deliveredby the ventilator, e.g., pressure control ventilationor controlled mechanical ventilation.

2. Intermittent mandatory: some mandatory breathsare delivered by the ventilator but the patient canalso breathe spontaneously in between themandatory breaths.

3. Continuous spontaneous: all breaths arespontaneously generated by the patient (e.g.,pressure support ventilation).

In Assist Control mode, a preset tidal volumebreath is delivered, which is either triggered by thepatient or as a mandatory breath when not triggeredin a specified time. Although this mode can decreasethe work of breathing, it can also lead to high minuteventilation. Trigger refers to the mechanism that setsthe ventilator to initiate inspiration.The trigger can betime, pressure, or flow. Ventilator cycling refers to themechanism by which the ventilator switches from theinspiratory to the expiratory cycle. It can be flow, time,or volume cycled (Table 75.2).

Ventilator settingsThe following parameters are required to be set by theoperator of the ventilator on initiation of mechanicalventilation.

Mode: Determines the mechanism of delivery of ventilation.

Control: Sets the pressure or volume limits formandatory ven-tilator breaths.

Support: As the name suggests, determines the pressure levelof a spontaneously generated but supported breath.

Table 75.2. Alternative modes of mechanical ventilation.

Adaptive pressure control

Adaptive support ventilation

Proportional assist ventilation

Airway pressure-release ventilation and biphasic positiveairway pressure

High-frequency oscillatory ventilation

Inverse ratio ventilation

Prone position ventilation

Tidal volume: Volume of inspiratory breath delivered. Usuallyset at 6–10 mL/kg of ideal body weight; 6–8 mL/kg tidalvolume has been shown to have better outcomes in acuterespiratory distress syndrome patients [2].

FiO2: Allows delivery of oxygen at different fractions duringinspiratory breaths.

PEEP: Positive end expiratory pressure is valuable inmaintain-ing alveolar patency at end of expiration.

Respiratory rate: Rate of delivery of mandatory inspiratorybreaths.

I:E ratio: The ratio of duration of inspiratory breath to expira-tory breath. Normally set at 1:2, it can be changed to pro-long the duration of either breath component depending onthe disease process. The ratio can be reversed in the man-agement of severe hypoxia.

Sensitivity: Level of negative pressure or flow required to trig-ger a ventilator breath. Usually, pressure is set at −1 to −2cmH2O and flow at 2–4 liters.

Ventilator weaningWeaning from mechanical ventilation should beattempted as soon as possible. Patients who are on

254

Case 75. Mechanical ventilation

mechanical ventilation should meet the following cri-teria before ventilator weaning is undertaken [3].� The underlying cause should have been reversed

or under control.� The patient should be able to initiate adequate

respiratory breaths on his or her own. Respiratoryrate �25, tidal volume �5 mL/kg, vital capacity�10 mL/kg, and negative inspiratory force�25 cm H2O are useful guidelines.

� The patient should have adequate oxygenationand ventilation on minimal ventilator support.That includes PaO2/FiO2 ratio �150–200, PEEPof ≤5–8 cmH2O, FiO2 ≤0.4 and an acceptable pH,PaCO2, and PaO2.

� The patient should be hemodynamically stable.The patient should not have any acute illnesswhich might lead to failure of extubation.

� Mental status should be good enough to be able toprotect airway and participate in maintenance ofbronchopulmonary hygiene.

Some patients may not meet all the criteria andmay still be eligible for a weaning trial.

The Spontaneous Breathing Trial (SBT) hasdemonstrated efficacy in rapid discontinuation ofmechanical ventilation. The trial consists of thepatient on minimal ventilator support or a T-piecespontaneously breathing. The SBT is administered for30–120 minutes. The patient is evaluated at the endof the SBT based upon the patient’s stable respiratory,hemodynamic, and mental status. Multiple trials inone day have not shown any benefit.

The rapid shallow breathing index (RSBI) is theratio of respiratory rate divided by tidal volume. AnRSBI �105 is used as a weaning parameter which hasbeen found to predict successful discontinuation ofmechanical ventilation.

A cuff-leak test (air leak around the cuff) can beperformed in patients who have been intubated for aprolonged period to determine if they have swollenupper airways.

Protocol-driven weaning ensures that weaning isattempted every day if possible. It has shown fasterventilator discontinuation rates when compared withthe standard physician-driven weaning.

Patients who fail to wean off the ventilator despitebest efforts will need a tracheostomy. The timing oftracheostomy is still controversial, but guidelines andmeta-analysis suggest performing early tracheostomyin patients who are anticipated to be on mechanical

ventilation for �21 days. Tracheostomy decreases theduration ofmechanical ventilator use and shortens thelength of stay in the intensive care unit. It is morecomfortable for the patient and decreases the work ofbreathing and sedation requirements.

Noninvasive ventilationNoninvasive mechanical ventilation is delivery of ven-tilation without endotracheal intubation. Either a facemask or a nasal device is used to deliver ventilation.It is suited for patients who have pulmonary insuffi-ciency that is temporary, not very severe, and with thepatient breathing spontaneously.The patient should beawake and able to protect his or her airway. It is notsuited for patients with a high risk of aspiration or withongoing hemodynamic instability.

In continuous positive airway pressure mode, aset constant airway pressure is maintained throughoutinspiratory and expiratory cycles. With bi-level posi-tive airway pressure, different levels of airway pressuresettings can be set for inspiration and expiration.

Complications of mechanical ventilationMechanical ventilation has many complications suchas barotrauma, volutrauma pneumothorax, ventilatorassociated pneumonia, acute lung injury, hypotension,and decreased cardiac output. Plateau pressure shouldbemaintained at≤30 cmH2O.Auto-PEEP is the devel-opment of positive pressure at end expiration due toincomplete exhalation. This leads to subsequent buildup of pressure which can lead to hemodynamic effectssuch as hypotension by decreasing cardiac output.Simply increasing the expiratory timemight help elim-inate auto-PEEP.Oxygen toxicity characterizedmainlyby either resorption atelectasis or acute lung injury isseen with use of high FiO2 (�0.6) for prolonged peri-ods of time.

Ventilator associated pneumonia is a commoncause of morbidity and mortality in the intensive careunit. It can be decreased by keeping the head of thebed �30 degrees, administering gastrointestinal ulcerprophylaxis, maintaining oropharyngeal hygiene, anddecreasing the number of days on the ventilator.

Mechanical ventilation in patients withneurologic disordersPatients with neurologic illness often require intu-bation and mechanical ventilation secondary to

255


decreased levels of consciousness, impaired airwayprotection, neuromuscular weakness, or pulmonarycomplications. Large ischemic strokes, intracranialhemorrhages, and subarachnoid hemorrhage aresome examples in which consciousness is depressed,while conditions such as Guillain-Barre syndrome andmyasthenia gravis lead to neuromuscular weakness. AGlasgow Coma Scale ≤8 is usually taken as a cut-offto secure the airway, although in some patients with alower score successful extubation has been performed.The decision to intubate the trachea in patientswith neuromuscular weakness should be based onboth subjective and objective assessments. Poorinspiratory effort, dyspnea, diaphoresis, tachycardia,use of accessory muscles, vital capacity �15 mL/kg,maximal inspiratory pressure �−30 cmH2O andnocturnal desaturation predict impending respiratoryfailure. Acute ischemic stroke patients with intactbrain stem reflexes may do well in the long term andmechanical ventilation as a life-savingmeasure shouldbe instituted. Hypercarbia increases intracranialpressure and can be deleterious in patients with braininjury. Mechanical ventilation can be very useful indecreasing PaCO2 acutely for the short term andthus lowering the intracranial pressure. Patients withintracranial hypertension usually tolerate high levelsof PEEP (even up to 20 mmHg), without any changein their intracranial pressure or cerebral perfusionpressure [4].

ConclusionMechanical ventilation is required frequently forpatients with neurologic disorders for airway pro-tection, pulmonary insufficiency, or management ofintracranial pressure.No onemode ofmechanical ven-tilation has proven to be superior in these patients.Mechanical ventilation strategies require optimizingoxygenation and ventilation with respect to the partic-ular neurologic disorder.

References1. R. L. Chatburn. Classification of ventilator modes:

update and proposal for implementation. Respir Care2007; 52: 301–23.

2. Ventilation with lower tidal volumes as compared withtraditional tidal volumes for acute lung injury and theacute respiratory distress syndrome. The AcuteRespiratory Distress Syndrome Network. N Engl J Med2000; 342: 1301–8.

3. N. R. MacIntyre,D. J. Cook, E. W. Ely, Jr et al.Evidence-based guidelines for weaning anddiscontinuing ventilatory support: a collective taskforce facilitated by the American College of ChestPhysicians; the American Association for RespiratoryCare; and the American College of Critical CareMedicine. Chest 2001; 120: 375S-95S.

4. E. Muench, C. Bauhuf,H. Roth et al. Effects ofpositive end-expiratory pressure on regional cerebralblood flow, intracranial pressure, and brain tissueoxygenation. Crit Care Med 2005; 33: 2367–72.

256


76 Mechanical ventilation for acute lung injuryin the neurosurgical patientJames M. Blum

Acute lung injury/acute respiratory distress syndrome(ALI/ARDS) is a common problem faced by patientsin the intensive care unit (ICU). The etiology of ALIis multifactorial and depends on the clinical situation;frequently ALI is the manifestation of bilateral pneu-monia, transfusion reactions, or aspiration. The def-inition of ALI stems from the American–Europeanconsensus conference, which states that three crite-ria must be met to designate a patient as having ALI[1]. Although survival of ALI patients has improvedgreatly over the past 30 years, many of the modali-ties used to treat the syndrome have potential conse-quences in the neurosurgical population.

Case descriptionThepatient was a 26-year-old female admitted for con-fusion, continuing headache, nausea, and vomiting at10:00 pm. Noncontrast computed tomography of thehead foundmass effect consistent with possible tumor,minimal midline shift, left ventricular compression,and no intracranial blood. She was treated with anti-emetics and admitted to a step-down unit for contin-uing observation while awaiting further neuroimag-ing; dexamethasonewas administered and the patient’sheadwas elevated in order to avoid further increases inintracranial pressure.Three hours after admission, thepatient had a witnessed aspiration event and her tra-chea was rapidly intubated. During endotracheal intu-bation, the patient was noted to have copious amountsof particulate matter on the vocal cords. After intu-bation, the patient continued to have oxygen satura-tions in the range of 83–86% on 100% fraction ofinspired oxygen (FiO2). Aggressive endotracheal suc-tioning was performed and the patient’s oxygen satu-ration increased to the range of 91–93%. The patientwas transferred to the ICU, where therapeutic bron-choscopy was performed. Large particulate plugs werefound and retrieved from tertiary bronchioles in bothlungs. After bronchoscopy, the patient’s oxygen satura-

tion further improved and she was taken for magneticresonance imaging, which revealed a large mass inthe temporal lobe consistent with a resectable tumor.The patient was scheduled for craniotomy and tumorresection as the first case in the morning.

Upon arrival for transport the anesthesia teamnoted that the oxygen saturation was 93% on 50%FiO2. Arterial blood gas (ABG) analysis revealed aPaO2 of 65 mmHg, PaCO2 of 35 mmHg, and pH of7.43. The patient was transported to the operatingroomuneventfully; however, after the cranial vault wasopened, the patient was noted to have oxygen satura-tions in the low 90s despite having an FiO2 of 80%. Asthe case continued, oxygenation continued to deterio-rate with saturations into the mid 80s on 100% FiO2,despite stable hemodynamics. Further ABG analysisrevealed a PaO2 of 55, PaCO2 of 33, and pH of 7.34.The surgeon was informed of the ongoing hypoxia.

The primary concerns of the anesthesiology teamwere (1) management of potentially life-threateninghypoxia, (2) potential consequences of positive endexpiratory pressure (PEEP) on cerebral hemodynam-ics and venous drainage, and (3) the potential effects oflung-protective ventilation on CO2 and thus intracra-nial pressure. To address the hypoxia, the patient wasstarted on propofol and remifentanil infusions and anICU ventilator was employed. A recruitment maneu-ver of 40 cmH2O for 40 seconds was performed, theventilator parameters were changed to pressure con-trol ventilation with a delta P of 15, and PEEP wasincreased to 15 cmH2O with an I:E ratio of 3:1. Thesemeasures resulted in improvement of oxygen satura-tions into the high 90s on 100% FiO2. Arterial bloodgas revealed a PaO2 of 130, PaCO2 of 49, and pHof 7.31. The patient’s respiratory rate was increasedto 20 breaths/min and tidal volumes remained thesame at about 350 mL. The surgeon placed a brain tis-sue oxygenation probe to assist with continuing man-agement postoperatively. The case was completed andchest radiography revealed bilateral fluffy infiltrates.

257


The patient returned to the ICU where she was man-aged using similar ventilator settings. Slowly, her FiO2requirement began to wean, her PEEP was reduced,and the trachea was extubated on postoperativeday 9.

DiscussionAcute lung injury is defined by the American–European consensus conference as acute hypoxia witha PaO2/FiO2 (P/F) ratio �300 (�200 for ARDS), bilat-eral infiltrates on chest radiograph, and absence ofclinical signs of left atrial hypertension [1]. Mortalityfrom ARDS overall is accepted to be around 30% cur-rently; however, most patients that expire with ARDSdo not die from hypoxia but rather sepsis or othercauses [2, 3].

There are two different etiological categories ofALI: direct lung injury and indirect lung injury. Directlung injury tends to include pneumonia and aspirationalong with inhalational injury and pulmonary contu-sions. Indirect injury etiology includes sepsis, trauma,blood transfusions, and pancreatitis. Acute lung injurytends to occur with an initial phase that lasts 3–5 days,a subacute phase that lasts 5–7 days, and then a pos-sible chronic phase that may result in long-term lunginjury and pulmonary fibrosis.

Currently, the only well-supported treatment ofARDS is low tidal volume ventilation as prescribedby the ARMA trial conducted by ARDSNet [2]. Inthis trial, patients were randomized to receive either12 mL/kg/predicted body weight (PBW) with maxi-mal plateau pressures of 50 cmH2O or 6 mL/kg/PBW.In the 6 mL/kg/PBW group if plateau pressuresexceeded 30 cmH2O, the volumes were reduced to 4mL/kg/PBW. This trial used levels of PEEP not com-monly seen in the operating room in order to enhanceoxygenation. The ALVEOLI trial showed that evenhigher levels of PEEP and “open lung” ventilation wasassociatedwith fewer ventilator days although nomor-tality benefit [2].

The basic settings on a ventilator include themode,respiratory rate, FiO2, PEEP, and I:E ratio. From thisbasic point, one can then select either volume controlor pressure control settings. In volume control, oneselects a set tidal volume to be delivered to the patient.For example, one could select a 400 mL tidal vol-ume. In pressure control, one generally selects PEEPand a drive pressure to generate the peak inspiratorypressure (PIP). For example, one may set PEEP at

Table 76.1. Predicted body weight (PBW) calculations(www.ardsnet.org).

Males PBW (kg) = 50 + 2.3 (height (in) – 60)

Females PBW (kg) = 45.5 + 2.3 (height (in) – 60)

5 cmH2O and a drive pressure of 20 cmH2O whichwill result in peak pressures of 25 cmH2O. Ventilatorsettings that directly affect oxygenation are the FiO2and the combination of PEEP, PIP, and I:E ratio, allof which contribute to the mean airway pressure. Byincreasing PEEP to 10 and PIP to 30, the mean air-way pressure, with an I:E ratio of 1:2 (common onmost OR ventilators) is 30/3+ (10× 2)/3 or 17. Byusing extended inverse ratio ventilation, with an I:Eratio of 3:1, the mean airway pressure increases to25 ((30× 3/4)+ (10/4)). The advantage of these set-tings is that the patient is exposed to higher meanairway pressures without profoundly increased PIP,which would present increased risk of barotrauma.However, one must be careful in this situation that thepatient has sufficiently compliant lungs and a sufficientamount of time to exhale in order to prevent profoundauto-PEEP.

In general, the use of 10mL/kg tidal volumes is stillespoused in the anesthesia community as appropriateventilation. Typically, anesthesiologists attempt to tar-get an end-tidal CO2 of around 35 mmHg to attemptto prevent spontaneous respiration during anesthesiaandmaintain normal pH. To address hypoxia, it is rea-sonable to provide increased amounts of PEEP. In theARDSNet trials, PEEP levels �20 cmH2O were notuncommon. Another key component of the ARDSNetventilation strategy is the use of low tidal volume ven-tilation based on 6 mL/kg/PBW.The PBW calculationis shown in Table 76.1. One should note that the calcu-lation varies based on the individual’s sex.

The use of PEEP and low tidal volume ventila-tion in the neurosurgical population is problematic,as a key component of ventilator management in thispopulation is appropriate CO2 removal. Low tidal vol-ume ventilation is frequently associatedwith increasedPaCO2, which results in increased blood flow to thebrain and potential increases in intracranial pressureand consequent ischemia to the brain parenchyma.Furthermore, an increase in intrathoracic pressure byPEEP and/or increased mean airway pressure reducesvenous return and potentially increases intracranialpressure to an even greater extent. These concernsled to the exclusion of neurosurgical and intracranialpathology patients from the initial ARDSNet studies.

258

Case 76. Mechanical ventilation for acute lung injury

Despite these concerns, it is possible to safelyimprove oxygenation and potentially reduce mortal-ity using ARDSNet settings and open lung ventilationin the critically ill neurosurgical patient. Wolf et al.demonstrated the safety of the open lung approach in2002 usingmeanPEEP levels of 15 cmH2O [4].Despiteusing this relatively high level of PEEP, ICP declinedover 24 hours despite amoderate increase in PaCO2. In2005, Wolf et al. used brain tissue oxygenation probesto determine the oxygen delivery to brain tissue at riskin 13 neurosurgical patients [5]. In this study, patientsreceived recruitment maneuvers at a level of 30–40cmH2O for 40 seconds and increased PEEP. Cerebraloxygenation improved using this technique, demon-strating overall improved oxygen delivery to the braintissue in question.

ConclusionThemanagement of the hypoxic neurosurgical patientis challenging and requires the balancing of mul-tiple organ systems needs. No one technique isappropriate in every patient, as some patients maydevelop increased intracranial pressure with evenminimal changes in their ventilation. Despite these

challenges, good outcomes are possible with attentivemanagement.

References1. G. R. Bernard, A. Artigas, K. L. Brigham et al. Report

of the American-European Consensus conference onacute respiratory distress syndrome: definitions,mechanisms, relevant outcomes, and clinical trialcoordination. Consensus Committee. J Crit Care 1994;9: 72–81.

2. Ventilation with lower tidal volumes as compared withtraditional tidal volumes for acute lung injury and theacute respiratory distress syndrome. The AcuteRespiratory Distress Syndrome Network. N Engl J Med2000; 342: 1301–8.

3. R. Brower, P. Lanken, N. MacIntyre et al. Higherversus lower positive end-expiratory pressures inpatients with the acute respiratory distress syndrome.N Engl J Med 2004; 351: 327–36.

4. S. Wolf, L. Schurer,H. A. Trost et al. The safety of theopen lung approach in neurosurgical patients. ActaNeurochir Suppl 2002; 81: 99–101.

5. S. Wolf,D. V. Plev,H. A. Trost et al. Open lungventilation in neurosurgery: an update on brain tissueoxygenation. Acta Neurochir Suppl 2005; 95: 103–5.

259


77 Change inmental statusEdward Manno

Encephalopathies are commonly encountered in theintensive care unit (ICU) andportendworse outcomes.The etiology of these encephalopathies is incompletelyunderstood and treatment is oftentimes based on thepresumed underlying mechanisms.

Case descriptionA 54-year-old man with a history of alcohol abuseand cirrhosis was admitted to the neurologic ICU afterdrainage of a large right-sided subdural hematoma.On arrival the patient was awake and alert withouta focal neurologic deficit. His postoperative coursewas uneventful until postoperative day 3 when hebecame progressively obtunded and stuporous, witha low-grade fever. A repeat head computed tomogra-phy (CT) revealed no recurrence of his hematoma. Anelectroencephalogram (EEG) was diffusely slow with-out epileptiform activity. His liver function tests andserum ammonia levels were quite elevated.The patientwas therefore given a flumazenil challenge under EEGmonitoring and he became immediately responsive.His mental status returned to baseline after treatmentwith lactulose and neomycin.

DiscussionConsciousness is a product of an individual’s degreeof arousal (or wakefulness) and awareness. Arousal isregulated through a series of interconnecting neuronalcircuits in the reticular activating system (RAS). TheRAS consists of three groups of neuronal cell bod-ies (norepinephrine, serotonin, dopamine) that orig-inate throughout the brainstem and project to vari-ous regions in the diencephalon and telencephalon [1].Awareness or subjective experience is thought to bemaintained diffusely through the cerebral cortex andits interconnections with the thalamus and basal gan-glia structures. A change in consciousness requires aprocess that affects the RAS or bilateral cerebral hemi-spheres or both [1].

The above patient hasmany possible etiologies thatcould underlie his change in mental status. The mostobvious would be a recurrence of his space-occupyinglesion that could lead to distortion of the brainstemthus directly affecting the RAS. Subclinical seizuressecondary to alcohol withdrawal is a possibility. Bothof these appear unlikely given the CT evidence of lackof a structural cause and for lack of EEG evidence forseizure activity (although aprevious seizure could havebeen missed). A low-grade fever raises the possibilityof an underlying infection and a possible early septicencephalopathy. However, his elevated liver functiontests and response to treatment argue strongly that thesource of his mental status decline was secondary tohepatic failure. Alteration in consciousness is one ofthe most common issues found in all ICUs. Due to themany potential causes, a systematic approach to theevaluation of the patient is mandated.

The initial evaluation should include a thoroughhistory and physical. Does the patient have any under-lying medical issue that could lead to deterioration inmental status? Is there a history of trauma, substanceabuse, depression with previous overdose attempts? Isthere a previous history of seizures? What were thecircumstances under which the patient was admitted?The history may need to be obtained (or verified) froma witness, emergency medical services, friends, orfamily.

A general physical examination should search forevidence of trauma or intoxication. Meningismusshould be evaluated. A fundoscopic examination mayreveal papilledema. The breath may suggest intoxica-tion or ketoacidosis. Subtle findings such as nystag-mus, eyelid, finger or lip twitching may indicate sub-clinical seizure activity. Vital signs can also providediagnostic clues as to the etiology of changes inmentalstatus. Elevated temperatures can suggest an underly-ing infection and hypothermia is commonly seen aftercold exposure, Addison’s disease, hypoglycemia, and/or pituitary disease. Hypotension and hypertension

260

Case 77. Change in mental status

can also impact a patient’s mental status. Mean arte-rial blood pressures below and above the level of apatient’s cerebral autoregulation can lead to cerebralhypoperfusion or the development of cerebral edema,respectively.

A complete laboratory evaluation is mandatoryand should include serum electrolytes, liver and thy-roid function tests, arterial blood gases, completeblood count with differential, toxicology evaluations,and a sedimentation rate. Infectious sources as well asvasculitic processes should also be evaluated.Drug lev-els should be evaluated and sedation limited as muchas possible. A very common cause of decreased men-tal status is poor hepatic clearance of sedative drugs. Aroutine chest X-ray should be obtained to evaluate forpneumonia or other acute processes. Additional bodyimaging may also be required.

The neurologic examination often will provideclues as to whether the alteration of consciousnessis due to a structural or metabolic source. Obvi-ously, neuroimaging is crucial in this evaluation [2].The clinical situation should suggest whether lumbarpuncture or neurophysiologic monitoring is needed.Common encephalopathies encountered in the ICUsinclude septic and hepatic encephalopathies.The exactmechanism by which sepsis impairs consciousnessis not known. It is speculated that the release ofcytokines and interleukins during the early inflam-matory response can cause direct blood–brain bar-rier damage. This can subsequently alter the internalmilieu of the brain, affecting neuronal transmission[3].

The etiology of hepatic encephalopathies has beendiscussed extensively, with ammonia levels and thedevelopment of false neurotransmitters claimed tobe the source of the encephalopathy. More recent

evidence has suggested that the development ofendogenous benzodiazepines may contribute to theencephalopathy. However, this is not the completeanswer since only about 40% of patients will respondto intravenously administered flumazenil. Electroen-cephalographic monitoring during administration offlumazenil can be used to determine if an occa-sional subclinical seizure can be detected. Inflam-matory mediators have also been implicated in theetiology of hepatic encephalopathy [4].

ConclusionChange in mental status and encephalopathy are com-mon occurrences in the ICU. Ruling out physiologic,pharmacologic, and neurologic etiologies requires athorough history, careful physical examination, andthe appropriate use of laboratory and imaging tests.Treatment should be tailored to the underlying etiol-ogy of the encephalopathy.

References1. F. Plum, J. B. Posner. The pathologic physiology of

signs and symptoms of coma. In Plum F., Posner J. B.,eds.The Diagnosis of Stupor and Coma, 3rd edition.Philadelphia, PA: F.A. Davis Company, 1982; 1–86.

2. G. B. Young. Initial assessment and management ofthe patient with impaired alertness. In Young G. B.,Ropper A. H., Bolton C. F., eds. Coma and ImpairedConsciousness. A Clinical Perspective. New York, NY:McGraw-Hill, 1998; 79–118.

3. J. Mahler, G. B. Young. Septic encephalopathy.Intensive Care Med 1993; 38: 177–87.

4. V. Sundaram,O. S. Shaikh. Hepatic encephalopathy.Pathophysiology and emerging therapies.Med ClinNorth Am 2009; 93: 819–36.

261


78 Therapeutic hypothermia using anendovascular approach in the neurocriticalcare patientAnupa Deogaonkar and Andrea Kurz

Therapeutic hypothermia for cerebral protection wasfirst used in the 1950s [1] and its use has waxed andwaned since then. Previous trials in brain trauma [2],ischemic stroke [3], and stroke due to subarachnoidhemorrhage [4] have examined the neuroprotectiveeffect of mild to moderate hypothermia. The Intra-operative Hypothermia for Aneurysm Surgery Trial(IHAST) was a well-matched, prospective, interna-tional, multicenter, randomized study of 1001 patientswith good-grade subarachnoid hemorrhage (SAH).This trial showed that there was no difference in out-come between the mild intraoperative hypothermicand normothermic groups [5]. However, in the follow-ing case we decided to use hypothermia as a neuropro-tectivemethod due to a high-grade SAHand persistentunconsciousness after cardiac arrest.

Case descriptionA56-year-old female patient was admitted to the neur-ocritical care unit after being resuscitated by the Emer-gency Medical Services team that was called by thefamily when the patient arrested at home. On arrival,the patient’s blood pressure was 210/120 mmHg, heartrate was 86/min, body temperature was 35.9 ◦C, Glas-gow Coma Scale score was 3. The patient’s pupils weremiotic and unreactive to light. The patient arrivedintubated and mechanically ventilated. An electrocar-diogram showed diffuse elevation of the ST segment,and serum cardiac markers were mildly elevated,including creatine-kinase and troponin T. A com-puted tomography scan of the brain showed diffuseSAH. The Hunt–Hess Scale score was IV with Fisher’sscore of 3. A ventriculostomy was performed thatshowed an elevated opening pressure. Approximately20mL of cerebrospinal fluid (CSF) were removed toreduce the intracranial pressure (ICP). The CSF wasxanthrochromic on inspection with numerous redblood cells. ICPwasmeasured using an intraparenchy-mal sensor (Camino, USA) that showed an ICP of

�40 mmHg. After obtaining informed consent fromthe patient’s family, mild hypothermia therapy wasinitiated using the Reprieve Endovascular Temper-ature Management Systems (Radiant Medical, CA).This consists of a triple-lobed, helically wound, heat-exchange balloon catheter that was placed in the infe-rior vena cava through the femoral vein via a 10 Frenchfemoral introducer sheath and a microprocessor-driven controller. The distal tip of the catheter was atthe level of the diaphragm and was connected to apump that circulated cold saline. The target core tem-perature of 33 ◦C was achieved in 1 hour. Midazolamand vecuronium were continuously used for sedationand muscle relaxation, respectively. A Foley tempera-ture catheter was used to monitor body core tempera-ture. Cerebral angiography 1 day after the admissionrevealed an intracranial aneurysm. The patient wasthen taken for endovascular coiling on the sameday. Hypothermia was maintained during the proce-dure and for up to 2 days after the procedure. Thepatient was gradually decooled to 37.5 ◦C at a rate of0.2 ◦C/hour. Shivering was suppressed using a forced-air warming blanket, oral buspirone, and intravenousmeperidine.Thepatient slowly regained consciousnessover the next 2 days and, after being treated in a reha-bilitation program for another 3 months, the patientwas discharged home. Modified Rankin Score at thetime of discharge from the rehabilitation center was 4.

DiscussionCooling a patient to mild or moderate hypothermiais usually performed by conductive (liquid-circulatingwater mattress), convective (forced air cooling via fullbody blankets or air beds) surface cooling, cold infu-sions, gastric lavage, passive cooling by leaving theanesthetized patient uncovered in a cool environment(e.g., operating room, intensive care unit), or througha combination of these methods. Studies using con-vective and/or conductive techniques to cool subjects

262

Case 78. Therapeutic hypothermia using an endovascular approach

presentingwith traumatic brain injury, stroke, and car-diac arrest found that the time required to cool subjectsto the desired temperature was often 6–8 hours andthat the cooling process was not very precise. A morerecent approach to controlling temperature in crit-ically ill patients uses intravascular cooling devices.The intravascular cooling device connects to an exter-nal cooling system.The system circulates temperature-controlled sterile saline through heat exchangersmounted on the distal end of the catheter.The patient’sblood is gently cooled as it passes over the bal-loons. Similar to cooling blanket equipment, the sys-tem responds to probes measuring the patient’s risingtemperature and adjusts the temperature of the sterilesaline flowing within the catheter. Endovascular cool-ing techniques seem to be superior for rapid induc-tion of hypothermia and for maintenance of stabletemperature as compared with surface cooling tech-niques. Furthermore, during slow rewarming, bodycore temperature is comparatively easier to controlwith endovascular cooling/warming.

The majority of therapeutic hypothermia trials forbrain protection have involved surface-cooling tech-niques that require mechanical ventilation in intu-bated and paralyzed patients. Therapeutic hypother-mia using the endovascular technique has distinctadvantages, such as rapid cooling in an awake patientwith intact shivering response and without using neu-romuscular blockade, stability of temperature duringtreatment, and controlled decooling [6]. Currently,two endovascular heat-exchange catheters are avail-able for use: Celsius Control System (Innercool, Inc,San Diego, CA) and Cool Line System (Alsius Inc,Irvine, CA). The RapidBlueTM system by InnercoolInc. automatically cools or warms the patient as nec-essary to maintain the desired body temperature. Itincludes a programmable console with an enhanceduser interface and can be used with InnerCool’s Stan-dard and AccutrolTM catheters to quickly modulatepatient temperature in association with surgery orother medical procedures. Similarly, the new Ther-mogard XPTM temperature management system andnew QuattroTM catheter by Alsius Inc. together teamup to deliver more power and control for both cool-ing and warming applications. Heat exchange occursbetween cooled saline that passes through the heatexchange portion of the catheter in a coil with largesurface area for heat exchange and the blood that flowsover the outer surface of the catheter. This closed-circuit circulation of a chilled solution through a flex-

ible metallic heat-exchange element is connected to aprogrammable external heat exchange bath and recir-culates the saline returning from the catheter. Venousblood is thus cooled on its way back to the heart asit passes the cooled element. An intravascular ther-mometer built in the catheter provides feedback to pre-cisely cool core blood to a target temperature.

These devices are generally placed in the femoralvein and are associated with a low rate of complica-tions. Endovascular cooling allows rapid cooling withtight control of target temperature, minimal shivering,and the possibility of avoiding paralysis of the patient.There are other techniques involving internal coolingmethods using infusion of cold fluids for the induc-tion of mild-to-moderate hypothermia. The rates ofinduction are variable but are otherwise considered tobe rapid. Cold fluid infusion with concomitant use ofcooling blankets has also been shown to be efficacious.Typically, the infusion volume is 30 mL/kg of normalsaline or lactated Ringer solution.

All anesthetics so far tested markedly decrease theshivering threshold as well as the maximum inten-sity of shivering. Reduction in the cold responses isespecially important if hypothermia is induced fortherapeutic reasons. Thus, induction of therapeutichypothermia in awake patients is complicated by theneed to overcome arteriovenous shunt vasoconstric-tion and shivering, and to do so without provokingextreme thermal discomfort. The search continuesfor a drug or drug combination that sufficientlyimpairs thermoregulatory defenses without simul-taneously producing unacceptable toxicity. Drugssuch as meperidine, dexmedetomidine, clonidine,nefopam, and buspirone alone, as well as in variouscombinations, reduce the shivering threshold and thuscomplement external and internal cooling.

ConclusionTherapeutic hypothermia has been shown to improveoutcome in patients after cardiopulmonary resuscita-tion and might prove helpful for other circumstancesin which a compromise of neurologic function isexpected. External (convective or conductive) as wellas internal cooling can be used for induction of thera-peutic hypothermia. However, internal (endovascular)cooling induction of hypothermia is more rapid witha more precisely controlled maintenance and decool-ing phase. If certain aspects of its use such as timing,rate of rewarming, and selective perfusion strategies

263


are optimized, the neuroprotective effects can likely beenhanced.

References1. W. G. Bigelow, J. C. Callaghan, J. A. Hopps. General

hypothermia for experimental intracardiac surgery;the use of electrophrenic respirations, an artificialpacemaker for cardiac standstill, and radio-frequencyrewarming in general hypothermia. Trans Meet AmSurg Assoc Am Surg Assoc 1950; 68: 211–219.

2. D.W. Marion, L. E. Penrod, S. F. Kelsey et al.Treatment of traumatic brain injury with moderatehypothermia. N Engl J Med 1997; 336: 540–546.

3. M. A. De Georgia,D.W. Krieger, A. Abou-Cheblet al. Cooling for Acute Ischemic Brain Damage

(COOL AID): a feasibility trial of endovascularcooling. Neurology 2004; 63: 312–317.

4. B. J. Hindman,M.M. Todd, A.W. Gelb et al.Mildhypothermia as a protective therapy duringintracranial aneurysm surgery: a randomizedprospective pilot trial. Neurosurgery 1999; 44:23–32.

5. M.M. Todd, B. J. Hindman,W. R. Clarke et al.Mildintraoperative hypothermia during surgery forintracranial aneurysm. N Engl J Med 2005; 325:135–145.

6. E. Keller,H. G. Imhof, S. Gasser et al. Endovascularcooling with heat exchange catheters: a new method toinduce and maintain hypothermia. Intens Care Med2003; 29: 939–943.

264


79 Therapeutic hypothermia aftercardiac arrestJames W. Jones and Piyush Mathur

Each year in the USA, approximately 450 000 menand women suffer cardiac arrest due to various causes,most commonly myocardial injury secondary to coro-nary artery disease. The vast majority of the victimsdo not survive to hospital admission, and more thantwo-thirds of all survivors suffermajor neurologic dys-function ranging from disabling cognitive deficits topermanent coma.Themajor factors in predicting neu-rologic dysfunction secondary to cardiac arrest involvethe extent of brain insult as a function of time to returnof circulation. Even with early resuscitation, neuro-logic outcomes have remained poor. Cerebral ischemiaand reperfusion trigger multiple metabolic cascadesthat ultimately result in permanent neuronal loss. Theuse of induced hypothermia has been studied as a wayto combat neurologic injury for nearly five decades.

Case descriptionA 37-year-old female with a history of chronic backpain and depression arrived at the emergency depart-ment following a witnessed cardiac arrest 1 week afterbeginning risperidone therapy. Bystanders at the sceneperformed cardiopulmonary resuscitation for 12 min-utes until paramedics arrived, at which time the tra-chea was promptly intubated and 1 mg epinephrinewas administered intravenously. Her cardiac rhythmon the defibrillator monitor was ventricular fibrilla-tion. Following two biphasic shocks at 200 Joules, shewas noted to have a spontaneous return of circulationby carotid artery palpation. Subsequent blood pressurecuff readings consistently demonstrated mean arte-rial pressures between 70 and 95 mmHg. However,her neurologic exam upon arrival at the emergencydepartment was remarkable for a GlasgowComa Scalescore of 6, including bilateral withdrawal from painand twitching of the right eye andmouth. Urgent headcomputed tomography scan demonstrated early signsof global ischemia consistent with anoxic brain injury.Propofol and atracurium infusions were started and

an automated intravascular cooling device was placedinto the femoral vein and a target temperature of 33 ◦Cwas obtained within 6 hours of arrest. Hypokalemiaand hyperglycemia were corrected. After 24 hours, thepatient was decooled slowly to 36.0 ◦C over the sub-sequent 12 hours. Propofol and atracurium infusionswere interrupted and the patient was noted to be mov-ing all extremities and following simple commands.She was extubated on day 3 and discharged on day 7. A6-month follow-up visit with her neurologist demon-strated minimal cognitive deficits, mostly with wordrecall.

DiscussionFollowing Dr. Peter Safar’s 1965 publication,Manage-ment of the Comatose Patient, in which he proposedthe benefits of therapeutic hypothermia in comatosepatients following restoration of blood flow, multi-ple subsequent studies have confirmed these find-ings and demonstrated a clear improvement in thosepatients who are resuscitated quickly and cooled early.The mechanism by which cooling improves neuro-logic function is not clear, and the benefits of cool-ing demonstrated with anoxic type injuries have notbeen replicated for other types of brain injury includ-ing traumatic, or ischemic and hemorrhagic stroke [1].Suggested mechanisms include a reduction in cerebraloxygen consumption especially in lowflowareas, slow-ing of damaging enzymatic reactions, maintenance ofthe lipoprotein fluidity of the neurons andblood–brainbarrier, suppression of free radicals, reduced intracel-lular acidosis, and inhibition of the synthesis, release,and uptake of neurotransmitters such as glutamate anddopamine [2, 3].

Today, the use of therapeutic hypothermia iswidely accepted as the standard of care for preservingneurologic function following cardiac arrest. Goalsfor therapy include attaining a core temperature of32–34 ◦C within 8 hours of the return of spontaneous

265


Table 79.1. Phases of therapeutic hypothermia.

PrecoolingRestore circulationNeurologic assessmentIntubation, sedation, and paralysis as neededPlacement of cooling device

Initiation of coolingTarget 32–34 ◦C within 8 hoursManagement of metabolic abnormalities

MaintenanceStabilize core temperature at 32–34 ◦CKeep mean arterial pressure �65 mmHgMonitor for tissue injuryManagement of metabolic abnormalities

DecoolingGradual (0.2–0.3 ◦C /hour) restoration of normothermiaManagement of hemodynamics (fluids and vasopressors)Management of metabolic abnormalitiesDiscontinuation of sedation and paralysisReassess neurologic status and consider extubation

circulation (Table 79.1). This should be maintainedfor 12–24 hours [1, 3]. Multiple small studies haveexamined the efficacy of various cooling modalitiesbut none has proven superior. Also, data suggest apotential benefit in attaining target core temperaturesearlier via infusion of cold intravenous fluids (2 litersnormal saline at 4 ◦C) both during and after return ofcirculation, however this remains controversial as flu-ids may increase the risk of pulmonary edema in thosewith cardiac dysfunction. Current research focuses onthe use of intranasal catheters, which deliver a coolingmixture of gas to localize cooling to the brain as well asdecrease time to achieve target brain temperatures.

Evidence for coolingTwo landmark randomized clinical trials publishedin 2002 demonstrate a clear benefit in the therapeuticuse of mild hypothermia following cardiac arrest[1, 3]. Both studies randomized patients to receiveeither hypothermia or maintenance of normothermia.The Hypothermia after Cardiac Arrest Study Group(HACA) found that 55% of the hypothermic grouphad a favorable neurologic outcome (e.g., ability to liveindependently) compared with 39% in the normoth-ermic control group [1]. This was echoed by Bernardet al., which found a 49% vs 29% favorable neurologicoutcome [3]. Furthermore, the HACA Group showeddecreased 6-month mortality for the hypothermicgroup (41%) vs the normothermic group (55%)[1]. In light of this evidence, the American HeartAssociation, as well as the International Liaison

Committee on Resuscitation, currently recommendtherapeutic hypothermia in their published resuscita-tion guidelines [4]. A 2009 Cochrane review focusingon neurologic outcomes, survival, and adverse eventsfurther supports these guidelines as the “best medicalpractice” [5].

Choice of cooling deviceTherapeutic hypothermia is simple to perform.Numerous cooling devices are available on the markettoday, which utilize the principles of conduction,convection, and evaporation [6]. Devices such as coolwater or air-circulating blankets, gel-coated adhesivepads, blood-cooling central venous catheters, and ice-packs effectively cool via conduction. Less commonly,misting in combination with a fan effectively uses con-vection and evaporation to cool the patient. No singledevice has been proven superior in terms of patientoutcomes, but devices with feedback temperaturecontrols meet target temperatures sooner and moreeffectively maintain this temperature throughout thecooling and decooling phase.

Initiation of coolingCooling should be performed in all postcardiac arrestpatients regardless of documented dysrhythmia, butsupportive data are strongest for patients who are postventricular fibrillation [7]. Initiation of cooling shouldideally begin within 8 hours following return of spon-taneous circulation in all patients who do not fol-low verbal commands, are free from life-threateningbleeding or infection, and are not at risk for immi-nent cardiopulmonary collapse. Sedation, paralysis,and mechanical ventilation are generally required toprotect the airway and prevent shivering. The opti-mal time to achieve target hypothermic temperaturesis unknown –most major studies target a core temper-ature of 32–34 ◦C for 12–24 hours [1, 3]. Electrolytesshould be checked frequently during the initiation andmaintenance of cooling to monitor for hypokalemiaand hyperglycemia; potassium and insulin should besupplemented.

The negative physiologic effects of cooling includedecreased cardiac output and coagulopathy as wellas complications from the cooling device, specificallyhypothermic tissue injury from surface cooling andinfection from intravascular catheters. Such compli-cations are more common with cooling below 32 ◦C.Positive physiologic effects ofmild cooling, in addition

266

Case 79. Therapeutic hypothermia after cardiac arrest

to long-term neurologic benefits, include antiseizureactivity and improved response to electrical defibril-lation in the setting of re-arrest.

DecoolingDecooling is the most hemodynamically unstablephase (Table 79.1) and is defined by the decreasedremoval of patient-generated heat until normother-mia is reached. This is better tolerated in contrastto “rewarming,” in which heat from an externalenergy source is added to the patient. Cooling deviceswith feedback temperature control allow the patientto achieve normothermia gradually, preventing thepatient’s intrinsic thermoregulatorymechanisms fromgenerating a deleterious rebound hyperthermia, whichcan occur for up to 72 hours. Gradual decooling at0.2 to 0.33 ◦C per hour until normothermia is reachedhelps diminish hemodynamic instability as well as theoccurrence of “post-resuscitation syndrome,” charac-terized by systemic inflammation, vasodilation, andhypotension [6]. Patients may require fluid adminis-tration and vasopressor support during the decool-ing phase, as well as continued sedation and paralysisto prevent shivering. Hypokalemia and hyperglycemiamay persist during this phase and should be corrected.

ConclusionVictims of cardiac arrest are at risk for neurologicinjury. Institution of therapeutic hypothermia within8 hours of return of circulation, targeting a goal of32–34 ◦C for 12–24 hours, has been shown to improveneurologic outcomes and mortality [1, 3].Therapeutichypothermia has been shown to be relatively safe andeffective, and should be considered in the treatmentof comatose patients following cardiac arrest. Multi-ple devices are available to assist with active cooling,none of which has been proven over others to improvepatient outcomes. However, automated devices with

feedback temperature controls have improved attain-ment of target temperatures and reduced the hemody-namic instability associated with uncontrolled decool-ing. Electrolyte abnormalities are common and shouldbe corrected.

AcknowledgmentsAppreciation to Jessica E. Bollinger, Critical Care Phar-macist, Cleveland Clinic.

References1. Hypothermia After Cardiac Arrest Study Group.

Mild therapeutic hypothermia to improve theneurologic outcome after cardiac arrest. N Engl J Med2002; 346: 549–53.

2. L. A. McIntyre,D. A. Fergusson, P. C. Hebert et al.Prolonged therapeutic hypothermia after traumaticbrain injury in adults: a systematic review. J AmMedAssoc 2003; 289: 2992–9.

3. S. A. Bernard, T. W. Gray,M. D. Buist et al.Treatment of comatose survivors of out-of-hospitalcardiac arrest with induced hypothermia. N Engl J Med2002; 346: 557–63.

4. J. P. Nolan, P. T. Morely, T. L. Hock et al. Therapeutichypothermia after cardiac arrest. An advisorystatement by the Advancement Life support TaskForce of the International Liaison committee onResuscitation. Resuscitation 2003; 57: 231–5.

5. J. Arrich,M. Holzer,H. Herkner et al. Hypothermiafor neuroprotection in adults after cardiopulmonaryresuscitation. Cochrane Database Syst Rev 2009; 4:CD004128.

6. D. B. Sedar, T. E. Van Der Kloor. Methods of cooling:practical aspects of therapeutic temperaturemanagement. Crit Care Med 2009; 37: S211–22.

7. M. Oddo, V. Ribordy, F. Feihl et al. Early predictorsof outcome in comatose survivors of ventricularfibrillation and non-ventricular fibrillation cardiacarrest treated with hypothermia: a prospective study.Crit Care Med 2008; 36: 2296–301.

267

Part IX Subarachnoid hemorrhageCase

80 Cerebral vasospasmEdward Manno

Subarachnoid hemorrhage (SAH) represents bleedinginto the subarachnoid space, most commonly from aruptured cerebral aneurysm. There are approximately30 000 patients with SAH per year in the USA makingit a commonly encountered disease in the neurologicintensive care unit (ICU).

Case descriptionA 52-year-old right-handed female with a medicalhistory of smoking and hypertension developed theabrupt onset of a severe bifrontal headache. She wastaken to a local emergency department where a headcomputed tomography (CT) revealed a Fisher 3 SAHwith moderate hydrocephalus. Her initial exam wasstuporous but arousable with intact cranial nerveresponses. Shewas able to follow commandswith stim-ulation. An external ventricular drain was placed withimprovement in her mental status. Cerebral angiogra-phy revealed an 8 mm anterior communicating arteryaneurysm that was treated with coil embolization.Thepatient was then transferred to the neurologic ICU.Intravenous magnesium and oral nimodipine werestarted for cerebral vasospasm prophylaxis.

Five days posthemorrhage she developed a wors-ening of her headache. Her serum sodium decreasedfrom 143 mOsm/L to 132 mOsm/L.The following dayshe developed progressive right-sided weakness andword-finding difficulties. Repeat cerebral angiographyrevealed severe vessel narrowing of the leftmiddle andanterior cerebral arteries (Figure 80.1). She was givenaggressive volume replacementwith normal saline andher mean arterial blood pressure was increased to110 mmHg with intravenous phenylephrine. She hadimprovement but not complete resolution of her signsand symptoms. The decision was made to proceedwith angioplasty of the narrowed cerebral vessels. Hersymptoms resolved and she was returned to the neu-rologic ICU. She made a good recovery and she wasdischarged home 2 weeks later.

Figure 80.1. Cerebral angiography 5 days after a patient had asubarachnoid hemorrhage. The image reveals severe vesselnarrowing of both the middle and anterior cerebral arteries.

DiscussionAneurysmal SAH is a disease with a devastating natu-ral history. Epidemiologically, it is a disease of middle-aged women with a 3:2 female to male predominancein the 40–60 year age group. Major risk factors forthe development and rupture of cerebral aneurysmsinclude hypertension and cigarette smoking.

Two major interventions over the past 20 yearshave had a significant impact on the treatment ofSAH. The development of the operative microscopeand interventional procedures has allowed for the earlytreatment of cerebral aneurysms to prevent rebleed-ing. Rebleeding occurs within the first fewweeks of theinitial hemorrhage in one-third of patients with SAHand acutely worsens outcomes.The other interventionis the development of critical care strategies based onan improved understanding of the cerebrovascular andhemodynamic changes that occur after SAH.

269

IX. Subarachnoid hemorrhage

Figure 80.2. Serial computed tomographic images of subarachnoid hemorrhages (SAH). The left image is a Fisher 2 SAH with thin layeredblood in the basal cisterns. The center image is a Fisher 3 SAH with thick clots in the cisterns. The image on the right is categorized as a Fisher4 SAH with no or little blood in the cisterns but some intraventricular blood.

Patients are categorized into clinical and radiologi-cal groups.TheHunt–Hess scale is a subjective clinicalgrouping of patients ranging from 1 (asymptomatic)to 5 (moribund) [1]. The World Federation of Neuro-logical Surgeons scale incorporates the GlasgowComaScale into a similar 1–5 scale in an attempt to increasethe objectivity and reliability of the score. Long-termpatient outcome correlates best with the initial presen-tation of the patient [2].

The risk of the development of cerebral vasospasmis estimated by the amount of subarachnoid blood inthe basal cisterns determined on a 24–48 hour CTscan. Fisher described the original grouping of patientswith group 1 having no subarachnoid blood on the ini-tial scan, group 2 having thin layered blood in the basalcisterns, group 3 having thick clots in the cisterns, andgroup 4 having no blood or thin layered blood in thecisterns with either additional intracerebral or intra-ventricular blood (Figure 80.2) [3]. Patients with thickclots were found to be at very high risk for devel-oping cerebral vasospasm with subsequent neurologicdeficits.

Cerebral vasospasm is a self-limited vasculopathythat develops 4–14 days after SAH. Pathologically, thebasal cerebral arteries exhibit a T-cell infiltrate, colla-gen remodeling and smooth muscle proliferation. Theprocess is initiated by an unidentified metabolic prod-uct of oxyhemoglobin and leads to the vessels becom-

ing stiff and narrowed. Complications can be signif-icant as severe vessel narrowing can lead to strokeand death. Specific neurologic signs and symptoms aredependent upon the involved vessels [4].

Treatment of patients with SAH, once an aneurysmis secured, includes generous use of intravenousisotonic fluids. A cerebrally induced salt wastingnephropathy can develop in some patients, which maybe related to the release of the B-type of natriureticpeptide. In some instances hypertonic intravenousfluids may be necessary to avoid hyponatremia andvolume depletion [4].

Cerebral autoregulation is lost as cerebralvasospasm develops. Decreases in cerebral bloodflow secondary to vessel narrowing can be attenu-ated and reversed with the application of inducedhypertension. Clinical studies have suggested that amajority of neurologic deficits can be reversed withthe application of the above hemodynamic techniques[5]. Cerebral angioplasty and direct vasodilator appli-cation is reserved for patients that do not respond tohemodynamic augmentation.

ConclusionThe application of interventions for vasospasm maybe guided by noninvasive measures of vessel narrow-ing and cerebral blood flow. Transcranial Doppler

270

Case 80. Cerebral vasospasm

ultrasound is a noninvasive device used to measureblood flow velocities in the basal cerebral arteries.Rising flow velocities may portend vessel narrowing.Subsequent measures of cerebral blood flow may beneeded to verify decreases in blood flow requiringtreatment [4]. Further work is required for a bet-ter understanding of cerebral vasospasm, the leadingcause of delayed cerebral ischemia after SAH.

References1. W. E. Hunt, R. M. Hess. Surgical risk as related to time

of intervention in the repair of intracranial aneurysms.J Neurosurg 1968; 1: 14–20.

2. Anonymous. Report of World Federation ofNeurological Surgeons Committee on a universalsubarachnoid hemorrhage grading scale. J Neurosurg1988; 68: 985–6.

3. J. P. Kistler, R. M. Crowell, K. R. Davis. The relationof cerebral vasospasm to the extent and location ofsubarachnoid blood visualized by CT scan: aprospective study. Neurology 1983; 33: 424–36.

4. E. M. Manno. Subarachnoid hemorrhage. Neurol Clin2004; 22: 347–66.

5. N. F. Kassell, S. J. Peerless,Q. J. Durwaed et al.Treatment of ischemic deficits from vasospasm withintravascular volume expansion and induced arterialhypertension. Neurosurgery 1982; 11: 337–43.

271


81 Ventriculoperitoneal shunt dependenceVivek Sabharwal and Asma Zakaria

Chronic hydrocephalus as a sequela of subarachnoidhemorrhage is a complication that neurosurgeons bat-tle with every day. These patients often have a difficultrecovery period, requiring shunt revisions due to mal-function and infections.

Case descriptionA 49-year-old female presented to the hospital withfever and altered mental status. The patient had a pastmedical history significant for traumatic subarachnoidhemorrhage and subdural hematoma after a motorvehicle collision 2 months prior. The patient devel-oped hydrocephalus after the incident, which requiredplacement of an external ventricular drain (EVD).The patient continued to have high drain output sev-eral days after her injury, which mandated placementof a ventriculoperitoneal shunt (VPS). The patientwas transferred to an inpatient rehabilitation facilitywhere she was noted to be more somnolent and inter-mittently febrile over the last several days. Physicalexamination revealed a fluctuant area under the cran-iotomy scar. Computed tomography (CT) scan of thehead revealed scalp tissue edema and a fluid collec-tion under the bone flap (Figure 81.1). The patientwas taken to the operating room and the infectedbone flap and VPS removed. The patient was extu-bated postoperatively and returned to the neurosur-gical intensive care unit for observation. Six hourslater, the patient became progressively more obtundedand had difficulty protecting her airway; her cranialskin flap was noted to be tight and bulging. A CTscan of the head revealed acute hydrocephalus, withventriculomegaly and hypodensity in the surround-ingwhitematter representing transependymal translo-cation of cerebrospinal fluid (CSF) (Figure 81.2). AnEVD was inserted emergently and 20 mL of CSF weredrained. The drain was then left open at 10 cmH2O.The patient’s examination returned to baseline a fewhours later.

Figure 81.1. There is a fluid collection over the right parietal lobe.The bright spot in the left lateral ventricle is the tip of theventriculoperitoneal shunt.

DiscussionThe incidence of hydrocephalus after aneurysmal sub-arachnoid hemorrhage (SAH) has been reported torange from 6–67% in the literature [1].This may occuracutely (day 0–3), subacutely (day 4–13) or chroni-cally (�14 days) after the bleed and can be obstruc-tive or nonobstructive in nature. Several pathophysi-ologic mechanisms for this impediment to CSF flowhave been postulated, but no clear etiology has beenidentified. Most theories seem to suggest a role in thealteration of CSF flow dynamics within the ventricularsystem.

Cerebrospinal fluid is produced by active secre-tion from the cerebral arterial blood. It is generated ata constant rate, largely from the choroid plexus, andcirculates through the ventricular system, exiting into

272

Case 81. Ventriculoperitoneal shunt dependence

Figure 81.2. The ventricles are grossly enlarged upon removal ofthe ventriculoperitoneal shunt with hypodensities around thefrontal poles suggesting transependymal flow of cerebrospinal fluidfrom elevated pressures.

the subarachnoid space via the foramina of Luschkaand Magendie. It is then propelled upwards towardsthe superior sagittal sinus where most of it is reab-sorbed. SomeCSFflows down towards the lumbar sub-arachnoid space and is reabsorbed through the spinalvenous plexus. The rate of absorption is dependent onthe pressure gradient between the subarachnoid spaceand venous system. Absorption ceases if venous pres-sure exceeds the intracranial pressure. The presenceof blood or adhesions within the ventricular systemor subarachnoid space can result in acute obstructivehydrocephalus. Alternatively, a nonobstructive patterncan be seen when the arachnoid granulations drainingCSF into the sagittal sinus are plugged with posthem-orrhagic or postinflammatory debris.The resistance toCSF outflow through the arachnoid granulations maybe increased 3-fold after SAH and may not return tophysiologic parameters until 40–50 days post-SAH [1].In some cases scarring may occur resulting in chronichydrocephalus.

The diagnosis of acute hydrocephalus is madebased on CT scan evidence of SAH or intraventricularblood, with or without the presence of enlarged ven-tricles, as well as a declining mental status. An EVDis emergently placed to relieve the elevated intracra-nial pressure and is subsequently rapidly or gradually

weaned off based on the patient’s progress. Chronichydrocephalus, defined as shunt-dependent hydro-cephalus, affects between 6–37% of patients with SAH[2]. Acute hydrocephalus at presentation is the biggestpredictor of the development of shunt-dependenthydrocephalus followed by the need for ventilatorsupport on admission [3]. Other predictors of shuntdependency include: higher Hunt–Hess grade, higherFisher grade, advanced age, female sex, larger thirdventricular and bicaudate diameter, posterior circula-tion location of the ruptured aneurysm, endovascu-lar treatment, high CSF protein levels, and prolongedduration of EVD [2, 4].

Patients who require an EVD for a longer dura-tion, or have multiple EVD revisions are more likelyto develop chronic hydrocephalus [3]. This may be areflection of the persistent failure of return to nor-mal CSF dynamics in these patients. Multiple EVDrevisions can be predictive of high CSF protein frominflammation or infection, or a high clot burden, bothof which clog the draining tubes and are indepen-dent predictors of chronic hydrocephalus. Alterna-tively, numerous passes with the EVD can increase therisk of infection or intracerebral hemorrhage resultingin a prolonged need for ventricular drainage.

As the popularity of endovascular treatment ofruptured cerebral aneurysms has grown, there hasbeen speculation that this treatment modality resultsin a higher incidence of shunt-dependent hydro-cephalus. However, most studies comparing the twotreatment modalities are retrospective and have notmatched the patient groups for grade of SAH [2, 5].Aggressive irrigation and early clot evacuation inpatients undergoing microsurgery may be responsi-ble for this trend. In addition, some surgeons rou-tinely perform fenestration of the lamina termi-nalis during aneurysm clipping to facilitate CSF flow.While this is helpful in patients with ventricularobstruction, it has not been found to be beneficialin nonobstructive forms of hydrocephalus and hasnot been shown to reduce the incidence of chronichydrocephalus.

Age is a strong predictor for shunt-dependenthydrocephalus with a 2% per year of age increasein risk [3]. This may be due to an expanded sub-arachnoid space and an ability to accommodatemore blood, causing greater CSF flow disturbances,increased meningeal fibrosis in the elderly resulting inimpaired absorption and decreased ventricular com-pliance causing early symptomatic hydrocephalus.

273


Although necessary for the treatment of chronichydrocephalus, VPS are fraught with complications.Patients with shunt-dependent hydrocephalus havelonger length of stays in the ICU. Approximately4% of insertions are associated with intracerebralhemorrhage and 8–10% become infected – requiringremoval, placement of a temporary EVD and intra-venous antibiotics before the shunt can be replaced.The 1- and 5-year rates of shunt longevity are 57%and 37%, respectively, requiring a shunt revision. Inone series, almost one-third of the patients undergo-ing shunt placement required a revision procedure and43% of these required a second revision [3]. Almosthalf of the shunt failures occurred within 14 days post-operatively and 90% failed within the first 60 days.Shunt failure after 6 months of insertion was exceed-ingly rare. This may be because normal CSF dynam-ics are restored by this time making shunt failure lessnoticeable clinically.

ConclusionShunt-dependent hydrocephalus is a multifactorialdisease and many researchers have tried to formulatea mathematical model to predict its occurrence [2, 4].This would allow for early shunt placement in patientsat risk and reduce hospital lengths of stay. Unfortu-nately, no such scale exists at this time and patientsmust undergo a prolonged EVD wean before they canbe deemed shunt-dependent. The fact that shunt fail-ure rates are minimal after 6 months of insertion sug-

gests that this is a transient phenomenon and somesurgeons propose a more conservative wean or seriallumbar drainage to avoid the placement of and mor-bidity associated with ventriculoperitoneal shunts.What is known is that untreated hydrocephalus canresult in significant disability and cognitive decline,which may become irreversible if not relieved in atimely manner.

References1. A. R. Dehdashti, B. Rilliet,D. A. Rufenacht et al.

Shunt dependent hydrocephalus after rupture ofintracranial aneurysms: a prospective study of theinfluence of treatment modality. J Neurosurg 2004;101: 402–7.

2. Z. Dorai, L. S. Hynan, T. A. Kopitnik et al. Factorsrelated to hydrocephalus after aneurysmalsubarachnoid hemorrhage. Neurosurgery 2003; 52:763–71.

3. C. J. O’Kelly, A. V. Kulkarni, P. C. Austin et al.Shunt-dependent hydrocephalus after aneurysmalsubarachnoid hemorrhage: incidence, predictors, andrevision rates. J Neurosurg 2009; 111: 1029–35.

4. M. Chan, A. Alaraj,M. Calderon et al. Prediction ofventriculo-peritoneal shunt dependency in patientswith aneurysmal subarachnoid hemorrhage.J Neurosurg 2009; 110: 44–9.

5. P. Varelas, A. Helms, G. Sinson et al. Clipping orcoiling of ruptured cerebral aneurysms andshunt-dependent hydrocephalus. Neurocrit Care 2006;4: 223–8.

274


82 Ventriculostomy infectionSamuel A. Irefin

Ventriculostomies and external ventricular drainagedevices are common in neurosurgical and neurocrit-ical care practice. The potential for infection of thislow-flow system is significant and the consequencescan be severe.

Case descriptionA 45-year-old male was admitted to the neuro-surgical intensive care unit after an endovasculartreatment with complete coil occlusion of basilarartery aneurysm. Twenty-four hours after admission,a computed tomography scan revealed acute obstruc-tive hydrocephalus and an external ventriculostomycatheter was placed. On postoperative day 5 thepatient developed a fever and an increase in inflam-matory biomarkers. Analysis of cerebrospinal fluid(CSF) demonstrated a 10-fold increase in cell index;white blood cells 1526, protein 70, glucose 77.Multire-sistant coagulase-negative Staphylococci were isolatedfrom CSF culture. Vancomycin therapy was initiatedintravenously.

DiscussionVentriculostomy catheters are vital in caring for neu-rosurgical patients. They provide continuous intracra-nial pressure monitoring and external CSF drainage.First introduced in 1875, they did not gain wide accep-tance until the 1960s, when Lundberg et al. refinedthe technique and demonstrated its usefulness forbedside analysis [1]. Ventriculostomy catheters arecommonly used to allow invasive monitoring of ele-vated intracranial pressure (ICP) secondary to acutehydrocephalus in various neurosurgical disorders suchas severe head trauma, subarachnoid hemorrhage,intracranial hypertension, or intraventricular hemor-rhage. The primary aim of these catheters is to detectelevated ICP and thereby guide medical or surgicaltherapies to maintain an adequate cerebral perfusionpressure. Since they are foreign bodies, ventriculitis

related to intraventricular drainage systems is a com-mon complication. It is often difficult to make a diag-nosis of ventriculitis with a ventriculostomy catheterin place. There are often signs of cerebral infectionbecause of underlying illness. Some patients may beasymptomatic or show only minor signs of infectionsuch as fever, anemia, or leucocytosis [2]. Diagno-sis of catheter-related infection can be made accord-ing to criteria advocated by Mayhall et al. [3]. Theseinclude positive CSF culture obtained from the ven-tricular catheter or from CSF drawn via lumbar punc-ture. In addition, CSF pleocytosis, low glucose level, orhigh protein level may indicate CSF infections [4].

The reported incidence of ventriculostomy-relatedcatheter infection varies between 0% and 45% ofpatients depending on technique of insertion andmanagement of the catheter [5]. As common withthe use of percutaneous catheters, Gram-positiveinfections traditionally have been predominant inpatients who present with ventriculitis. However,Gram-negative infections have been reported in ven-triculostomy catheter use and have resulted in highermortality rates [6].The risk factors for catheter-relatedventriculitis can be categorized into three groups:(1) patient characteristics and the underlying mecha-nism of injury, (2) events that break the integrity of aclosed system, and (3) environmental influences [7].As far as patient characteristics are concerned, nei-ther age, sex, nor race increase the risk of developingcatheter infections [6].

As a result of potential pitfalls in diagnosis andsubsequent delayed initiation of appropriate antimi-crobial therapy contributing to morbidity or mor-tality, prevention of catheter-related ventriculitis isof paramount importance. Special emphasis shouldbe placed on avoiding modifiable risk factors [8].Review of the literature does not support prophylac-tic exchange of catheters in reducing the incidence ofcatheter-related ventriculitis [2]. However, the dura-tion of the ventriculostomy catheter is a significant risk

275


factor in developing catheter-related infection. There-fore, efforts must be made to identify clinically rele-vant CSF infections and suspected or confirmed infec-tions must be differentiated from contamination andcatheter colonization. The administration of prophy-lactic antibiotics decreases the incidence of CSF infec-tions at the expense of predisposing the patient toinfection by more resistant organisms when infectionsdo occur [4]. It is prudent to administer prophylac-tic antibiotics before the catheter insertion to pro-tect against the skin flora contamination of the woundsite. Consideration may be given to Gram-negativecoverage in selected patients who may require pro-longed hospitalizations.These agents may be providedfor the entire duration of catheter use. The selectedantibioticsmust penetrate the blood–brain barrier andknowledge of local resistance patterns is of paramountimportance.

ConclusionVentriculostomy-related infections remain a seriouscomplication of intraventricular catheter use. Causesof these infections are multifactorial in nature. Closemonitoring and rigorous cathetermaintenance remainthe mainstay of catheter care. Assessment of riskfactors is crucial to determine changes in micro-bial infection and to identify ways to prevent futurecomplications.

References1. N. Lundberg,H. Troupp,H. Lorin. Continuous

recording of ventricular-fluid pressure in patients withsevere traumatic brain injury: a preliminary report.J Neurosurg 1965; 22: 581–90.

2. Y. Arabi, Z. A. Memish,H. H. Balkhy et al.Ventriculostomy-associated infections: incidenceand risk factors. Am J Infect Control 2005; 33:137–43.

3. C. G. Mayhall, N. H. Archer, V. A. Lamb et al.Ventriculostomy-related infections. A prospectiveepidemiologic study. N Engl J Med 1984; 310:553–9.

4. K. E. Lyke,O. O. Obasanjo,M. A. Williams et al.Ventriculitis complicating use of intraventricularcatheters in adult neurosurgical patients. Clin InfectDis 2001; 33: 2028–33.

5. K. L. Holloway, T. Barnes, S. Choi et al.Ventriculostomy infections: the effect of monitoringduration and catheter exchange in 584 patients.J Neurosurg 1996; 85: 419–24.

6. F. J. Buckwold, R. Hand, R. R. Hansebout. Hospital-acquired bacterial meningitis in neurosurgicalpatients. J Neurosurg 1977; 46: 494–500.

7. J. K. Ohrstrom, J. K. Skou, T. Ejlertsen et al. Infectedventriculostomy: bacteriology and treatment. ActaNeurochir (Wien) 1989; 100: 67–9.

8. A. M. Korinek,M. Reina, A. L. Boch et al. Preventionof external ventricular drain-related ventriculitis. ActaNeurochir (Wien) 2005; 147: 39–45.

276


83 Sodium abnormalities in neurocritical careWilliam R. Stetler and George A. Mashour

Disorders of sodium and water balance are frequentcomplications encountered in the care of critically illneurologic patients. Common dysnatremias includehyponatremia caused by syndrome of inappropri-ate antidiuretic hormone (SIADH) and cerebral saltwasting syndrome (CSWS), as well as hypernatremiacaused by central diabetes insipidus (CDI). Properdiagnosis of the cause of each sodium abnormality iscritical, as treatment varies widely among etiologies.

Case descriptionThe patient was a 45-year-old female with a 30 pack-year history of tobacco use who had the “worstheadache of [her] life” while at dinner and then imme-diately slumped over in her chair. Family alerted emer-gency medical services; the patient was unresponsiveso the trachea was intubated and she was taken tothe emergency department. A computed tomography(CT) scan of the head revealed a large amount of sub-arachnoid hemorrhage (SAH) with ventricular exten-sion. Neurosurgery was consulted and a ventricu-lostomy catheter was placed with opening pressure of36 cmH2O above the tragus. The following morningthe patient was taken to the angiography suite where abasilar apex aneurysmwas found and embolized usingendovascular coils. The patient was then transportedto the neurosurgical intensive care unit and stabilized.On neurologic examination she was noted to brisklylocalize with all four extremities.

On admission the patient’s serum sodium was144 mmol/L. By postoperative day 1 her sodiumdropped to 135, and by postoperative day 2 to 129. Aurine osmolality was found to be 405 mOsm/kg andthe urine sodium was 45 mmol/L. Urine output wasnoted to be 12 liters over 48 hours and total intakehad been only 6 liters and central venous pressuredropped from 10 mmHg on admission to 4 mmHg.The neurocritical care team pursued fluid rehydrationwith normal saline, however the sodium continued

to drop to 123 mmol/L. Therefore, on postoperativeday 4, 3% NaCl was begun at 50 mL/hour. By post-operative day 5 the patient’s serum sodium rose to131 mmol/L, but she was now noted to be extensorposturing with her left upper and lower extremities.A noncontrast CT scan of the head did not revealappreciable change, but CT angiography revealed dif-fuse right anterior circulation vasospasm. Fluids wereincreased and the patient was taken emergently toangiography where intra-arterial administration ofnicardipine and selective angioplasty of the right ante-rior circulation was employed to achieve angiographicresolution of vasospasm.

Over the following 3 weeks the patient’s serumsodium stabilized to approximately 135 mmol/L on anoral regimen of NaCl tablets, and she was dischargedfrom the hospital to a rehabilitation facility followingcommands on her right side and localizing to noxiousstimuli on her left. At her 2-year follow-up the patientwalked into the clinic with the use of a cane.

DiscussionAs the major extracellular cation, sodium is one ofthe most important osmotic agents in the body. Dis-orders of sodium and water balance are commonamong the critical care patient and are especially com-mon among neurocritical care patients. Hyponatremiais one of the most common electrolyte disorders inthe neurologic patient and is frequently encounteredin patients following SAH, traumatic brain injury,and many neurosurgical procedures. Patients withhyponatremia are often thought to suffer from cere-bral edema [1], however, other complications such asmental status changes, seizures, vasospasm, and evendeath occur more frequently following hyponatremia.Hyponatremia has also been shown to be an indepen-dent risk factor for all-cause morbidity and mortal-ity, even when controlling for the original pathology.Furthermore, correction of serum sodium may helpimprove mortality [2].

277


Table 83.1. Comparison of common dysnatremias inneurocritical care.

SIADH CSW CDI

Hyponatremiavshypernatremia

Hyponatremia Hyponatremia Hypernatremia

Euvolemia vshypovolemia

Euvolemia Hypovolemia Hypovolemia

Serumosmolarity

�285mOsm/L

�285mOsm/L

�300mOsm/L

Urineosmolarity

�200mOsm/L

�200mOsm/L

�350mOsm/L

Urine sodium ↑ ↑ ↓SIADH, syndrome of inappropriate antidiuretic hormone secre-tion; CSWS, cerebral salt wasting syndrome; CDI, central diabetesinsipidus.

Despite its well-accepted complications, the etiol-ogy of hyponatremia in SAH is still debated with themost common causes cited as SIADH and CSWS. InSIADH, there is an inappropriate or excessive releaseof antidiuretic hormone, which acts on the distal renaltubule to increase water permeability and thereforecreate a concentrated urine and dilutional, euvolemichyponatremia.Thus, SIADH is a problem ofwater bal-ance. In CSWS, there is a renal loss of sodium (andconsequently a loss of water), leading to a hypovolemichyponatremia [1, 3]. Cerebral salt wasting syndromeis therefore a problem of sodium balance. The exactpathophysiology of CSWS is not known. However, itis hypothesized that the increase in sympathetic tonefollowing SAH may increase blood pressure, increasesystemic vascular resistance/venous capacitance vas-culature tone, and possibly cause direct natriureticeffects on the kidney, all contributing to increasednaturesis [4]. There is some evidence that releaseof natriuretic peptides following SAH (most notablyatrial (A-type) natriuretic peptide and brain (B-type)natriuretic peptide) could cause either a direct or indi-rect renal sodium loss [5]. Nevertheless, the circulatinglevels of these peptides have not been shown to corre-late with natriuresis [2, 5].

The distinction between the two pathologies isimportant as the treatment is typically opposite(Table 83.1). In SIADH, hyponatremia is controlledwith fluid restriction, while CSWS is treated withfluid replacement. In fact, the key distinguishing fea-ture between the two is volume status of the patient.The hyponatremia in SIADH is euvolemic or hyperv-olemic hyponatremia, whereas in CSWS it is a hypov-

olemic hyponatremia. Many laboratory investigationsshow similar trends in CSWS and SIADH, includ-ing serum osmolarity �285 mOsm/L, urine osmolar-ity �200 mOsm/L, urine sodium �25 mmol/L. How-ever, there is often an associated fall in central venouspressure and pulmonary capillary wedge pressure inCSWS. Physical exam should focus on skin turgor,jugular venous distention, and mucous membranes tohelp determine volume status. Additionally, an eleva-tion in hematocrit may be observed from the concen-trating effects of CSWS. Secondary to the need forhypervolemic therapy in SAH, it is imperative thatthe two etiologies be distinguished [1, 2]. Hypona-tremia alone has been shown to increase vasospasmrisk and cerebral infarction in SAHpatients [2].There-fore, aggressive fluid resuscitation even before serumsodium drops appreciably may not only treat, but off-set the drop of serum sodium associated with CSWS[4]. Because of the risk of vasospasm in SAH patients,SIADH is often not treated with fluid restriction in theneurocritical care setting.

In this case, the patient most likely developedCSWS as her urine output increased while her cen-tral venous pressure dropped precipitously. Unfortu-nately the patient went on to develop vasospasm inher right anterior circulation, an event more com-mon in hyponatremic patients with SAH (especially inpatients with hyponatremia caused by CSWS) [2].

Treatment of hyponatremia depends upon patientsymptomatology. Mild cases of SIADH traditionallyrespond to fluid restriction; use of diuretics and deme-clocycline, which has been shown to block ADHreceptors, may augment fluid restriction. For moresevere symptoms, hypertonic (3% NaCl) saline shouldbe considered, especially in patients who may nottolerate diuresis or fluid restriction because of therisk for vasospasm. Rapid correction of hyponatremia(�12 mmol/L/day) should be avoided to help preventpontine myelinolysis, especially in cases of chronichyponatremia.The foundation for treatment ofCSW isfluid resuscitation. Isotonic saline is often sufficient tocorrect both the volume deficit as well as the hypona-tremia in CSW.Hypertonic salinemay be used inmorerefractory cases. Additionally, mineralocorticoid ther-apy such as fludrocortisone may help treat the sodiumloss [2]. In this case, 3% NaCl was used to treat thehyponatremia because conventional aggressive fluidresuscitation failed. Furthermore, aggressive correc-tion of the patient’s serum sodium was warrantedbecause the patient developed cerebral vasospasm [2].

278

Case 83. Sodium abnormalities in neurocritical care

Hypernatremia is often defined as a serum sodium�145 mmol/L, and usually represents a loss oftotal body water. Common causes of hypernatremiainclude reduced intake of water, extrarenal waterloss, intrarenal water loss, mineralocorticoid excess,and iatrogenic. Central diabetes insipidus representsintrarenal loss of water and is the most commoncause of hypernatremia in the neurologic patient.As opposed to its counterpart nephrogenic diabetesinsipidus, inwhich the kidneys are insensitive to antid-iuretic hormone (ADH), in CDI there is a failure ofrelease of ADH from the hypothalamopituitary axis.This is most commonly seen in patients after pitu-itary surgery, following traumatic brain injury (whichcan shear the pituitary stalk) or SAH, and in patientsthat are brain dead [3]. Antidiuretic hormone acts onthe distal tubule and medullary collecting ducts toincrease the permeability of water and therefore helpreabsorb water through protein channels [1]. In theabsence of ADH water is not reabsorbed but ratherlost in a dilute urine resulting in hypovolemic hyper-natremia. Serum osmolarity is often �305 mOsm/Lwhile urine osmolarity is �350 mOsm/L (and urinespecific gravity is often �1.005) in CDI (Table 83.1).

Treatment of CDI depends upon both severity ofsymptoms and level of consciousness of the patient. Ifthe patient is conscious then increasing oral intake tobalance urine output is ideal. If the patient is unableto drink orally, free water may be administered viaa nasogastric tube. However, if the patient is unableto maintain a balanced intake and output or serumsodium continues to increase despite all efforts toincrease free water intake, medical therapy may be

considered. Intravenous or intranasal administrationof vasopressin (1-deamino-8-D-arginine or DDAVP)may be administered and titrated to decrease urineoutput to �200 mL/hour in the short term [2].

ConclusionIn conclusion, dysnatremia is a common occurrenceamong critically ill neurologic patients. In particu-lar, hyponatremia from either CSW or SIADH mayincrease the rate of complications and, in the caseof SAH, predispose patients to cerebral vasospasm.Therefore, prompt recognition of hyponatremia andinstitution of appropriate, aggressive therapy is war-ranted to avoid complications [2].

References1. M. R. Harrigan. Cerebral salt wasting syndrome.

Crit Care Clin 2001; 17: 125–38.2. M. Rahman,W. A. Friedman. Hyponatremia in

neurosurgical patients: clinical guidelinesdevelopment. Neurosurgery 2009; 65: 925–36.

3. M. Tisdall,M. Crocker, J. Watkiss et al. Disturbancesof sodium in critically ill adult neurologic patients: aclinical review. J Neurosurg Anesthesiol 2006; 18:57–63.

4. S. Singh,D. Bohn, A. P. Carlotti et al. Cerebral saltwasting: truths, fallacies, therories, and challenges.Crit Care Med 2002; 30: 2575–9.

5. G. Audibert, G. Steinmann, N. de Talance et al.Endocrine response after severe subarachnoidhemorrhage related to sodium and blood volumeregulation. Anesth Analg 2009; 108: 1922–8.

279

Part X StrokeCase

84 Initial managementLauryn R. Rochlen

Stroke is classified into two major types [1]. Ischemicstrokes represent 80% of cerebral vascular events andoccur as a result of thrombosis, embolism, or hypoper-fusion. The remainder are hemorrhagic in nature, dueto either intracerebral or subarachnoid hemorrhage.This chapter will focus on the initial management ofpatients with ischemic stroke.

Case descriptionThe patient was a 54-year-old female with a past med-ical history significant for hypertension and bilat-eral carotid stenosis. She presented to the hospitalcomplaining of fatigue and pallor, and was found tohave a hemoglobin of 4 g/dL due to a lower gas-trointestinal bleed. She was receiving a packed redblood cell transfusion when she had acute onset ofdysarthria, left hemiparesis, right gaze deviation, andleft-sided neglect. Immediate evaluation revealed anintact airway, appropriate ventilatory pattern and oxy-genation, and a blood pressure of 144/69. Neuro-logic exam was consistent with a National Institutes ofHealth (NIH) Stroke Scale (NIHSS) of 22. Labs werecollected. A head computed tomography (CT) scanwas significant for evidence of a large right middlecerebral artery (MCA) infarct. Given the recent his-tory of lower gastrointestinal bleed, this patient wasnot a candidate to receive recombinant tissue plas-minogen activator (rtPA). The patient was transferredto an institution capable of performing intra-arterialthrombolysis.

Upon arrival at the tertiary center, exam wasunchanged except for an NIHSS of 16. She was trans-ferred to the interventional radiology suite for intra-arterial thrombolysis. Following successful dissolutionof a thrombus in the right MCA, she was broughtto the neurosurgical intensive care unit for furthermonitoring.

DiscussionThe main goals of the immediate evaluation of apatient with an acute neurologic change are to deter-mine whether the patient’s symptoms are due to astroke and to establish potential contraindications forthrombolysis. Emergency evaluation of a patient withpossible ischemic stroke should initially focus on rapidassessment and stabilization of the airway, breath-ing, and circulation. This is quickly followed by thesecondary assessment of neurologic deficits and co-morbidities. Pertinent history is essential in orderto determine time of symptom onset and potentialcauses in order to provide for early secondary pre-vention and options for intervention [2]. The NIHSS(Table 84.1) is particularly useful for quantifying thedegree of deficit, facilitating communication betweenpractitioners, identifying possible location of vesselocclusion, providing early prognosis, and identifyingpatient eligibility for certain interventions. Diagnostictests, including electrolytes, glucose, complete bloodcount, and coagulation profiles should be performedto rule out conditions that imitate acute stroke. How-ever, as time is critical in these situations, thrombolytictherapy should not be delayed unless there is high sus-picion for a hypocoagulable state.

Early diagnosis of acute ischemic stroke is facili-tatedwith thewidespread use of neuroimaging. Anon-contrast enhanced CT scan of the head identifies mostcases of intracranial hemorrhage and aides in diagnos-ing nonvascular causes of symptoms, but is insensitivein detecting small, acute infarctions. Signs of severeischemic stroke seen on CT, such as loss of gray–whitematter differentiation and sulcal effacement, are asso-ciated with poorer outcomes and may correlate with ahigher risk of hemorrhagic transformation followingthrombolysis. Magnetic resonance imaging (MRI) ismore accurate for identifying ischemic lesions, as wellas distinguishing between acute and chronic infarcts.Due to cost limitations, limited availability, and patient

281

X. Stroke

Table 84.1. The National Institutes of Health Stroke Scale [2].

Testeditem Title Responses and scores

1A Level ofconsciousness

0 – alert1 – drowsy2 – obtunded3 – coma/unresponsive

1B Orientationquestions (2)

0 – answers both correctly1 – answers one correctly2 – answers neither correctly

1C Response tocommands (2)

0 – performs both taskscorrectly

1 – performs one taskcorrectly

2 – performs neither

2 Gaze 0 – normal horizontalmovements

1 – partial gaze palsy2 – complete gaze palsy

3 Visual fields 0 – no visual field defect1 – partial hemianopia2 – complete hemianopia3 – bilateral hemianopia

4 Facial movement 0 – normal1 – minor facial weakness2 – partial facial weakness3 – complete unilateral palsy

5 Motor function (arm)a. leftb. Right

0 – no drift1 – drift before 5 seconds2 – falls before 5 seconds3 – no effort against gravity4 – no movement

6 Motor function (leg)a. leftb. Right

0 – no drift1 – drift before 5 seconds2 – falls before 5 seconds3 – no effort against gravity4 – no movement

7 Limb ataxia 0 – no ataxia1 – ataxia in 1 limb2 – ataxia in 2 limbs

8 Sensory 0 – no sensory loss1 – mild sensory loss2 – severe sensory loss

9 Language 0 – normal1 – mild aphasia2 – severe aphasia3 – mute or global aphasia

10 Articulation 0 – normal1 – mild dysarthria2 – severe dysarthria

11 Extinction orinattention

0 – absent1 – mild (loss 1 sensory

modality)2 – severe (loss 2 modalities)

contraindications with MRI, CT scan remains thestandard mode of imaging and will provide sufficientinformation needed to proceed with decisions regard-ing treatment.

Once the diagnosis of acute ischemic stroke isestablished, focus should shift on management strate-gies to prevent secondary neurologic injury and deter-mine treatment options.The main tenets of secondaryinjury prevention are to maintain adequate oxygena-tion, cellular homeostasis, and appropriate cerebralperfusion. Patients with a considerable extent of neu-ronal damage are at risk for airway obstruction,hypoventilation, aspiration pneumonia, and atelecta-sis, which can all result in desaturation and decreasedoxygen delivery to already compromised neurons.Need for tracheal intubation should be evaluatedin every patient. Avoiding hyperthermia, hypo- andhyperglycemia are also important for maintainingoptimal neuronal function.

In the immediate poststroke period, blood pres-sure is often elevated, which may represent a pro-tective response attributed to the Cushing reflex [3].Higher blood pressure is beneficial for maintain-ing perfusion to the pressure-dependent peri-infarctpenumbra. However, hypertension will increase therisk of hemorrhagic transformation of the infarctedarea. Arguments for lowering blood pressure includeprevention of hemorrhagic conversion, minimizingcerebral edema and preventing secondary ischemicevents due to hypoperfusion of the at-risk penumbra.Data regarding blood pressure management are cur-rently inconclusive, but recommendations are avail-able. Patients who are not candidates for rtPA shouldnot receive medications to lower blood pressure untilsystolic pressure is�220mmHgor diastolic pressure is�120 mmHg. Treatment goal is to lower blood pres-sure by 15% during the first 24 hours of stroke onset.For those patients who are eligible to receive rtPA,blood pressure should be lowered to a systolic pressure�185 mmHg and a diastolic pressure �110 mmHgbefore thrombolysis is initiated.

Intravenous thrombolytic therapy with rtPA hasbeen shown to improve outcomes following ischemicstroke [4]. Risk of adverse events following rtPAadministration, such as intracranial or systemic hem-orrhage, correlate directly with time since onset ofneurologic symptoms. Current guidelines limit its useto within 3 hours of symptom onset. Other contraindi-cations include evidence of intracranial or systemic

282

Case 84. Initial management

hemorrhage, stroke or myocardial infarction in theprevious 3 months, major surgery in the previous14 days, elevated blood pressure, current anticoag-ulation or hypocoagulable state and patient/familyrefusal of treatment. No other IV thrombolytic agentis approved for use in the USA.

Intra-arterial thrombolysis may be an alternativefor patients unable to receive IV rtPA.With IA throm-bolysis, a high concentration of thrombolytic agentis deposited directly into the thrombus. Intra-arterialthrombolysis is presently limited to patients who havepresented �6 hours from symptom onset due to MCAocclusion, are otherwise not candidates to receivertPA, and are at a treatment center with qualified inter-ventionalists and immediate access to cerebral angiog-raphy.

ConclusionTreatment of patients with acute ischemic stroke canbe challenging. Initial goals are to prevent secondaryneurologic injury and determine eligibility for throm-bolysis. As new data on management of ischemic

stroke continue to emerge, guidelines will be updatedwith the latest evidence-based recommendations.

References1. P. Amarenco, J. Bogousslavsky, L. R. Caplan et al.

Classification of stroke subtypes. Cerebrovasc Dis 2009;27: 493–501.

2. H. P. Adams Jr, G. del Zoppo,M. J. Alberts et al.Guidelines for the early management of adults withischemic stroke: a guideline from the American HeartAssociation/American Stroke Association StrokeCouncil, Clinical Cardiology Council, CardiovascularRadiology and Intervention Council, and theAtherosclerotic Peripheral Vascular Disease andQuality of Care Outcomes in ResearchInterdisciplinary Working Groups: The AmericanAcademy of Neurology affirms the value of thisguideline as an educational tool for neurologists.Circulation 2007; 115: e478–534.

3. E. Cumbler, J. Glasheen. Management of bloodpressure after acute ischemic stroke: an evidence-basedguide for the hospitalist. J Hosp Med 2007; 2: 261–7.

4. Tissue plasminogen activator for acute ischemicstroke. The National Institute of NeurologicalDisorders and Stroke rt-PA Stroke Study Group.N Engl J Med 1995; 333: 1581–7.

283

Part X StrokeCase

85 Increased intracranial pressure at48 hours poststrokeLauryn R. Rochlen

Intracranial pressure (ICP) reflects the cumulativepressures of the intracranial contents – brain tissue,cerebral blood volume (CBV), and cerebrospinal fluid(CSF). Normal values for ICP are �20 mmHg [1].The primary concern when ICP is elevated is thatcerebral perfusion pressure (CPP) will be reduced,resulting in ischemia and irreversible cellular damage.Patients with intracranial pathology have decreasedintracranial elastance and are at risk for exaggeratedincreases in ICP with small changes in intracranialvolume.

Case descriptionPlease refer to Case 84 for the patient’s originalpresentation.

A 54-year-old female with a past medical his-tory significant for hypertension and bilateral carotidartery stenosis presented with an acute right mid-dle cerebral artery (MCA) ischemic infarct. She wasnot eligible to receive intravenous recombinant tissueplasminogen activator (rtPA), but did undergo intra-arterial rtPA administration under fluoroscopic guid-ance by an interventional radiologist.

During the first postoperative day, the patientremained hemodynamically stable with an intact air-way. Neurologically she was alert and oriented, withhemiparesis in the left upper and lower extremity.A follow-up noncontrast head computed tomography(CT) scan revealedminor cerebral edemawith a 9-mmmidline shift, no evidence of hydrocephalus, and smallareas of hemorrhagic conversion.

Over the course of the second postoperative day,the patient’s neurologic examination deteriorated andshe became increasingly somnolent and aphasic. Man-nitol 0.5 g/kg was administered intravenously, and ahead CT was obtained. Results from the head CTshowed worsened mass effect on the right side withincreased midline shift. The trachea was intubated forairway protection and hyperventilated to a PaCO2 of

Figure 85.1. Head computed tomography scan following righthemicraniectomy.

30 mmHg. The decision was made to proceed to theoperating room for emergent decompressive hemi-craniectomy (Figure 85.1).

DiscussionWhen ICP becomes elevated following intracranialtrauma of any cause, global cerebral blood flow (CBF)becomes compromised to levels that cannot supportbrain metabolism, ultimately resulting in neurologicchanges and irreversible damage. The mechanism ofincreased ICP can be attributed to cytotoxic and vaso-genic edema [2]. Cytotoxic edema refers to the shiftof water from extracellular to intracellular compart-ments. Cellular swelling and electrolyte imbalanceseventually result in cellular death. Vasogenic edema

284

Case 85. Increased intracranial pressure at 48 hours poststroke

refers to the phenomenon of water shifting fromintravascular to extravascular compartments due to abreakdownof the blood–brain barrier. Cerebral edematypically occurs during the first 24–48 hours follow-ing injury, and peaks approximately 3–5 days after anischemic insult [3]. Patients with large hemisphericinfarctions are at increased risk for malignant cerebralswelling [4]. In addition to reduction of CPP, increasedICP leads to tissue shifts that can eventually result inherniation and further irreversible damage.

An integrated approach is suggested for manage-ment of elevated ICP, beginning with conservativemaneuvers prior to proceeding to more aggressiveinterventions. As with any acute change in patient sta-tus, it is necessary to re-evaluate airway, breathing, andcirculation (the “ABCs”). Avoidance of hypoxemia andhypotension are essential to prevent further neuronaldestruction. Intubation may be required both for air-way protection andmaintenance of adequate oxygena-tion. In the setting of increased ICP, it will be necessaryto increase themean arterial pressure in order tomain-tain CPP.

A few simple maneuvers can be attempted initially.Elevating the head and neck �30 degrees and avoid-ing jugular venous compression enhance intracranialvenous outflow and allow gravity to assist in reductionof CBV. If the trachea is intubated, manual hyperven-tilation will quickly lower CO2 levels resulting in cere-bral vasoconstriction and decreased cerebral bloodvolume and ICP. The CSF is able to regulate pH andwill therefore return to normal pHwithin 12–36hours,limiting the beneficial effects of hyperventilation [5].

The next step in treatment of increased ICP is toinstitute hyperosmolar therapy. The two most com-mon agents used are mannitol, an osmotic diuretic,and 3% sodium chloride. Increasing the serumosmolality creates an osmotic gradient across theblood–brain barrier and therefore efficacy depends onan intact blood–brain barrier. Free water will follow itsconcentration gradient, moving from the edematousintracellular environment to the more concentratedintravascular compartment, thereby improvingintracranial elastance. Another potential advantageseen with mannitol is that it reduces blood viscosity,which improves CBF and leads to vasoconstriction.The initial dose for mannitol is 0.25–1 g/kg. Treatmentshould be targeted to a goal serum osmolality of≤320 mOsm/kg. Systemic complications of hyper-osmolar therapy include congestive heart failure dueto initial intravascular volume expansion, electrolyte

disturbances, rebound cerebral edema, and acutetubular necrosis [3].

High-dose intravenous steroids reduce the per-meability of the blood–brain barrier and have beeneffective in reducing edema surrounding intracranialtumors. However, the use of steroids in patients withedema due to ischemic stroke has not been shown toresult in any improvement in outcome.

If the above measures are ineffective, more aggres-sive options should be considered. Decompressivesurgery involves removal of part of the cranium toallow space for the swollen brain tissue to expand.Goals of decompression are to prevent herniation,increase CPP to areas that are still viable, and pre-serve CBF. Hemicraniectomy has been shown to beespecially effective in patients with malignant MCAinfarcts who have failed conventional therapies [2].Exact timing of when to proceed with decompressionhas not been established.

Induction of moderate hypothermia (32–34 ◦C)will reduce cerebral metabolic rate and ICP. Exper-imentally, hypothermia has been shown to decreasethe volume of infarcted area. Prospective studies areneeded to determine the overall effect of hypothermiaonpatient outcomes.Themore important conceptmaybe to avoid hyperthermia, which causes an increase incerebral metabolic rate.

Pharmacologic coma is reserved for the patient inwhom all other interventions have been unsuccessful.By reducing cerebral metabolic rate, pharmacologiccoma reduces cellular oxygen requirements and there-fore decreases CBF, CBV, and ICP. Maximum reduc-tions are thought to occur when a burst-suppressionpattern is evident on continuous electroencephalo-graphic monitoring. Agents used include pentobarbi-tal, thiopental, propofol, and benzodiazepines. Whilepharmacologic coma has been shown to reduce ICP,there are currently no proven benefits on overall mor-bidity and mortality.

Pharmacologic coma and hypothermia are poten-tially associated with more extensive care and seque-lae. Both usually require mechanical ventilation, mus-cle paralysis, and use of vasopressors. High levels ofsedation and paralysis make it difficult to follow clin-ical examinations. Patients are also at higher risk foradverse events such as pneumonia and coagulopathy,as well as cardiovascular and electrolyte disturbances.

The interventions discussed above are not with-out adverse effects. The risk of intervention mustbe balanced against the risk of prolonged, untreated

285

X. Stroke

increases in ICP. Unfortunately, current guidelines donot have specific recommendations regarding when toinitiate treatment – decisions must be specific for theindividual patient’s clinical status.

References1. M. Czosnyka, J. D. Pickard. Monitoring and

interpretation of intracranial pressure. J NeurolNeurosurg Psychiatry 2004; 75: 813–21.

2. C. Ayata, A. H. Ropper. Ischaemic brain oedema.J Clin Neurosci 2002; 9: 113–24.

3. V. Singh. Critical care assessment and management ofacute ischemic stroke. J Vasc Interv Radiol 2004; 15:S21–7.

4. M. Kohrmann, S. Schwab. Hemicraniectomy formalignant middle cerebral artery infarction. Curr OpinCrit Care 2009; 15: 125–30.

5. N. Stocchetti, A. I. Maas, A. Chieregato et al.Hyperventilation in head injury: a review. Chest 2005;127: 1812–27.

286

Part XI Intraparenchymal hemorrhageCase

86 Hypertensive intracerebral hemorrhageJames F. Burke and Teresa L. Jacobs

Intracerebral hemorrhage (ICH) accounts for 10–15%of all strokes and is associated with significant mortal-ity [1]. Hypertension is, by far, themost important riskfactor.

Case descriptionThe patient was a 45-year-old male with a history ofsevere hypertension on a multidrug regimen for bloodpressure control who alerted his spouse shortly afteracutely developing an acute onset headache associatedwith severe nausea. The patient’s spouse noted thatthe patient’s gait was grossly unsteady and immedi-ately activated emergency medical services. On arrivalin the emergency room, the patient had a blood pres-sure of 210/105. The patient was awake, but mini-mally somnolent, requiring that questions and com-mands be repeated in order to respond appropriately.On examination, the patient was oriented to person,place, and time and able to participate with a neu-rologic examination, albeit with some difficulty. Hisspeech was moderately dysarthric. On cranial nerveevaluation, he was noted to have spontaneous left-beating nystagmus with saccadic pursuit movements.No facial asymmetry was detected and no significantmotor or sensory deficits were noted. The patient haddifficulty maintaining seated balance with a tendencyto sway back and forth unless supported. The patienthadmild dysmetria on finger-nose-finger testing bilat-erally. Reflexes were symmetric throughout with anupgoing plantar response bilaterally.

Noncontrast head computed tomography (CT)scan (Figure 86.1) revealed a midline cerebellar hem-orrhage with maximal diameter of 4 cm, perpendic-ular diameter of 3 cm and was visualized on 6 sliceswith a 5.0 mm slice thickness (hemorrhage volumeestimated at 4 × 3 × (6 × 0.5)/3 = 12 mL) In theemergency room, the patient’s blood pressure spon-taneously decreased to 180/105 and he subsequentlyreceived two small boluses of labetalol (10mg and then

20 mg) which transiently lowered his blood pressureto 150/90. However, his blood pressure subsequentlyincreased and he was started on a nicardipine infusion(starting dose 5 mg/hour) which was titrated to a goalmean arterial pressure (MAP) of 110.

The patient was admitted to the neurologic inten-sive care unit (ICU) where the nicardipine infusionwas continued and the patient’s neurologic status wasclosely monitored. The patient failed a bedside swal-low evaluation and a Dobhoff tube was inserted. Hewas clinically stable for almost 24 hours before his levelof alertness declined further. At this time, concernemerged that the patient was not protecting his airway,and he underwent endotracheal intubation with initi-ation of mechanical ventilation. Head CT was emer-gently repeated and demonstrated a stable hemorrhagewith the interval development of significant edema,fourth ventricular compression, and hydrocephalus.A ventriculostomy was performed at bedside and anexternal ventricular drain (EVD) was inserted. Overthe next several hours, the patient’s level of alertnessimproved to his admission baseline.

The patient remained hospitalized in the neuro-logic ICU for the next 3 days as oral antihypertensiveswere initiated and he was progressively weaned fromthe ventilator, nicardipine drip, and EVD. The patientwas transferred to the general care ward where he wasevaluated by physical, occupational, and speech thera-pists. His swallowing function progressively improvedand he was progressively transitioned to an oral diet.After demonstrating the ability to maintain his caloricintake orally he was transferred to the inpatient reha-bilitation service where therapy was focused on gaittraining. He was ultimately discharged home.

DiscussionHypertensive ICH results from rupture of smallpenetrating arteries leading to intraparenchymalhematoma formation. Ongoing brain injury is

287

XI. Intraparenchymal hemorrhage

Figure 86.1. A noncontrast headcomputed tomography scan in the axialplane showing a large cerebellarhemorrhage with mass effect upon thefourth ventricle.

mediated by a host of pathologic processes includingmass effect, edema, inflammation, and direct toxicityof blood products [2]. Intracerebral hemorrhage mostcommonly occurs in subcortical regions predomi-nantly perfused by small penetrating arteries – basalganglia, pons, thalamus, cerebellum. For ICH outsidethese regions (i.e., lobar hemorrhage), a broaderdifferential diagnosis should be considered, includ-ing vascular malformation, primary brain tumor,metastasis, venous infarction, or amyloid angiopathy.

Hematoma location largely determines the pre-senting symptoms. Similar to ischemic stroke, ICHtypically presents with acute onset focal neurologicsymptoms. Deterioration in the level of conscious-ness and headache are more common in ICH thanin ischemic stroke with alteration in level of con-sciousness being particularly common with ICH inthe posterior fossa. Despite these differences, it isoften difficult to distinguish ICH from ischemic strokeon clinical grounds alone and consequently emergentneuroimaging with CT or magnetic resonance imag-ing is essential. While hypertension is the most sig-nificant risk factor for ICH, male sex, age, smoking,

and excessive alcohol consumption also constitute riskfactors.

For most patients, deficits are maximal at onset.However, early neurologic deterioration occurs inapproximately one-quarter of patients [3]. Hematomaexpansion is a common mechanism associatedwith early deterioration. Elevated blood pressureshave been associated with hematoma expansionand a recent pilot trial demonstrated a decrease inhematoma size with early, intensive blood pressurelowering [4]. However, aggressive blood pressurelowering may potentially cause clinical deteriorationeither through failure to perfuse the ischemic peri-hematomal region or through decreasing cerebralperfusion pressure (CPP) in the context of elevatedintracranial pressure (ICP). Consequently, currentguidelines do not strongly recommend for or againstaggressive blood pressure management, insteadurging selection of blood pressure targets on the basisof the totality of clinical circumstances including thepotential presence of elevated ICP and the patient’sbaseline blood pressure [5]. Factor VIIa has beenshown to decrease rate of hematoma growth, but has

288

Case 86. Hypertensive intracerebral hemorrhage

Table 86.1. Most common potential agents for blood pressurecontrol along with typical initial dosages.

Initial agentsLabetalolHydralazine

5–10 mg intravenous (IV), may repeat5–10 mg IV, may repeat

Continuous infusionsNicardipineClevidipineNitroprussideEsmolol

5 mg/hour IV drip, titrate to effect1–2 mg/hour IV drip, titrate to effect0.5–1 mcg/kg/min, titrate to effectas per local routine

not shown benefit on survival or functional outcomes[6].

Medical managementSupportive care, close neurologic monitoring andattention to potential complications form the basisof ICH management. In contrast, early surgical man-agement in the form of hematoma evacuation is notsuperior to conservative management in unselectedpatients, but may be beneficial in certain subsets ofpatients – for example, those with superficial lobarhemorrhages [7]. For subtentorial hemorrhages, pos-terior fossa decompression and clot evacuation are rea-sonablemeasures in the patient undergoingneurologicdecline. Standard conservative measures in ICH con-sist of blood pressure (see Table 86.1) and ICPmanage-ment in combination with control of hyperglycemiaand hyperpyrexia, attention to commonmedical com-plications such as deep venous thrombosis and con-sideration of antiepileptic therapy. Seizures are notuncommon after ICH – occurring in 8.1% of patientsby 30 days and most commonly in patients withlobar hemorrhages [8]. There is general consensusthat patients with seizures should be treated withantiepileptics and uncertainty about the role of pro-phylactic antiepileptics [5].

Close monitoring of neurologic status in an ICUis essential in ICH and for patients with elevated ICPinvasive monitoring may be beneficial. Elevated ICP iscommon after ICH as a consequence of variable com-binations of mass effect, edema, hydrocephalus, andintraventricular hemorrhage. No specific ICPmanage-ment strategies have been shown to be preferentiallybeneficial in this context. Most commonly a gradedincrease in invasiveness is utilized starting with lessinvasive measures (elevating the head of the bed, ade-quately managing pain) and progressively escalatingthe degree of invasiveness and risk (hyperventilation,

osmolar therapy, cerebrospinal fluid drainage, barbitu-rate coma) [5].

PrognosisIntracerebral hemorrhage is associatedwith significantmortality – between 35–50% of patients die within1 month of presentation [1]. A number of variablespredict mortality – Glasgow Coma Scale at presenta-tion, age, infratentorial ICH, presence of intraventric-ular hemorrhage, and ICH volume – these variablescan be combined into a single prognostic index, theICH score, which strongly predicts 30-day mortality[9]. Of patients who survive their acute presentation,only about half are functionally independent 6monthsafter presentation.

ConclusionIn summary, hypertensive ICH is a relatively commonreason for neurologic ICU admission and is associatedwith significantmorbidity andmortality. Conservativemanagement is the core principle with close attentionto preventing and managing neurologic and medicalcomplications.There are a number of significant unan-swered questions in the care of these patients, includ-ing blood pressure management.

References1. V. L. Feigin, C. M. Lawes,D. A. Bennett et al. Stroke

epidemiology: a review of population-based studies ofincidence, prevalence, and case-fatality in the late 20thcentury. Lancet Neurol 2003; 2: 43–53.

2. G. Xi, R. F. Keep, J. T. Hoff. Mechanisms of braininjury after intracerebral haemorrhage. Lancet Neurol2006; 5: 53–63.

3. R. Leira, A. Davalos, Y. Silva et al. Early neurologicdeterioration in intracerebral hemorrhage: predictorsand associated factors. Neurology 2004; 63: 461–7.

4. C. S. Anderson, Y. Huang, J. G. Wang et al. Intensiveblood pressure reduction in acute cerebralhaemorrhage trial (INTERACT): a randomised pilottrial. Lancet Neurol 2008; 7: 391–9.

5. J. Broderick, S. Connolly, E. Feldmann et al.Guidelines for the management of spontaneousintracerebral hemorrhage in adults: 2007 update: aguideline from the American Heart Association/American Stroke Association Stroke Council, HighBlood Pressure Research Council, and the Quality ofCare and Outcomes in Research InterdisciplinaryWorking Group. Stroke 2007; 38: 2001–23.

289


6. S. A. Mayer,N. C. Brun, K. Begtrup et al. Efficacy andsafety of recombinant activated factor VII for acuteintracerebral hemorrhage. N Engl J Med 2008; 358:2127–37.

7. A. D. Mendelow, B. A. Gregson,H. M. Fernandeset al. Early surgery versus initial conservativetreatment in patients with spontaneous supratentorialintracerebral haematomas in the International Surgical

Trial in Intracerebral Haemorrhage (STICH): arandomised trial. Lancet 2005; 365: 387–97.

8. S. Passero, R. Rocchi, S. Rossi et al. Seizures afterspontaneous supratentorial intracerebral hemorrhage.Epilepsia 2002; 43: 1175–80.

9. J. C. Hemphill,D. C. Bonovich, L. Besmertis et al.The ICH score: a simple, reliable grading scale forintracerebral hemorrhage. Stroke 2001; 32: 891–7.

290


87 Intracerebral hemorrhage andanticoagulationEric E. Adelman and Teresa L. Jacobs

Anticoagulation is used to prevent and treat throm-boembolic disease. While anticoagulation is effective,adverse events from the treatment can be devastating.One of the most serious complications of anticoagula-tion is intracerebral hemorrhage (ICH).

Case descriptionThe patient was a 66-year-old male with a history ofprior ischemic stroke, atrial fibrillation (for which hetook coumadin), hypertension, and type 2 diabetes,who developed the acute onset of right arm and legweakness. He also had difficulty speaking and his wifenoted a right facial droop. He presented to the emer-gency room with a systolic blood pressure of 225and right hemiparesis. A noncontrast head computedtomography (CT) scan was performed that showed ahemorrhage in the posterior limb of the left internalcapsule (Figure 87.1A). His blood counts were nor-mal and his International Normalized Ratio (INR)was 2.7. He was treated with vitamin K 10 mg intra-venously (IV), fresh frozen plasma (FFP) 1000 mL IV,andwas admitted to the neurologic intensive care unit.His clinical condition deteriorated during the FFPinfusion and a repeat noncontrast head CT, 53 min-utes after the initial scan, showed growth of the hem-orrhage with extension into the left lateral ventricle(Figure 87.1B). His blood pressure was controlled andhis ICH remained stable on serial imaging. Ten daysafter his ICH, coumadin was restarted, and he was dis-charged to a subacute rehabilitation facility.

Discussion

AnticoagulantsIn the USA, vitamin K antagonists, such as coumadin,are themost commonly used oral anticoagulants. Vita-min K antagonists, which inhibit vitamin K dependentcofactors (II, VII, IX, and X), can be difficult to usebecause they have a narrow therapeutic window, mul-

tiple drug and food interactions, and genetic factorsimpact dosing [1].

Coumadin’s anticoagulation intensity is measuredwith the INR.This measure standardizes values acrossdifferent laboratories, though there is still somevariability [1]. The goal INR varies depending onthe indication for anticoagulation and the individualpatient. Typically, for nonvalvular atrial fibrillation thegoal is 2–3.When the INR is supratherapeutic withoutbleeding, it can be managed with vitamin K admin-istration. However, in the setting of bleeding, moreintensive reversal is required and FFP, prothrombincomplex concentrate (PCC), or recombinant factorVIIa (rFVIIa) are used.

Unfractionated heparin and low molecular weightheparins (LMWHs) are the most commonly usedparenteral anticoagulants, though fondaparinux anddirect thrombin inhibitors (hirudin and argatroban)are also used [2]. All of the parenteral anticoag-ulants, except the direct thrombin inhibitors, bindto antithrombin and exert downstream anticoagulanteffects.

Unfractionated heparin’s anticoagulant activity ismonitored using the activated partial thromboplas-tin time (aPTT or PTT). The activated clotting time(ACT), is used primarily during procedures and surg-eries. There is no uniform PTT goal, but 1.5–2.5 timesthe upper limit of normal is often used. There aremultiple nomograms available to help guide dosing.The anticoagulant effects of unfractionated heparin arereversed by protamine.

Lowmolecular weight heparins bind to antithrom-bin but also act directly against factor Xa. They havea more predictable anticoagulant effect than heparin.Monitoring of anticoagulation intensity for LMWHscan be done by measuring anti-Xa levels, but this isof uncertain clinical benefit and only done in specialcircumstances [2].There are no agents that completelyreverse the anticoagulant effects of LMWHs, thoughprotamine has limited efficacy. Fondaparinux is a

291


Figure 87.1. A. Intracranial hemorrhagein the posterior limb of the left internalcapsule. B. Repeat imaging, 53 minuteslater shows growth of the hematomawith extension in the left lateral ventricle.

synthetic pentasaccharide that binds to the heparinbinding site on antithrombin. There is no specificreversal agent.

The direct thrombin inhibitors, often used in thesetting of heparin-induced thrombocytopenia, bindto thrombin leading to downstream anticoagulanteffects.These agents increase the PTT, but the responseis nonlinear so it is not useful for monitoring [2].Direct thrombin inhibitors also increase the INR,which can cause difficulties when transitioning from adirect thrombin inhibitor to coumadin. Some authorsrecommend using factor X levels instead of INR formonitoring [2]. Although there are no agents thatspecifically reverse the direct thrombin inhibitors,some have used rFVIIa.

Intracerebral hemorrhages andanticoagulationWhile ICH are not common in anticoagulatedpatients, the mortality is high at about 60% [3]. As theindications for anticoagulation have expanded andas the population has aged, there are more patientstaking anticoagulants.

The pathophysiology of anticoagulant-associatedICH is not entirely clear, but it is thought that micro-bleeds, which would be asymptomatic in a patientwithout a coagulopathy, become clinically significantin anticoagulated patients [3]. Risk factors for ICH

in anticoagulated patients are similar to patients notreceiving anticoagulation, but increasing anticoagula-tion intensity also contributes.

Most hemorrhages occur when patients’ antico-agulation is in the therapeutic range [4]. An INRof 3.5–3.9 increases the odds of hemorrhage by 3.6;conversely, an INR of �2 does not offer protectionagainst intracerebral hemorrhage and is less effectivein preventing thromboembolism [4].The combinationof coumadin and aspirin likely increases the risk ofICH [5], but there are also data arguing against anassociation [6].

Hematoma size and growth of the hematomaare predictors of poor outcome [7]. In compari-son to unanticoagulated patients, the anticoagulatedpatient typically has a larger initial hematoma with anincreased chance of expansion over a longer period oftime [7, 8].

Reversal of anticoagulationThe mainstay of treatment for anticoagulant-associated ICH is correction of the coagulopathyto reduce the potential for hematoma expansion(Table 87.1). There are multiple agents that reversecoumadin and their uses are guided by expert opinion,rather than comparative data. Agents that are usedto reverse coumadin include: vitamin K, FFP, PCC,and rFVIIa. As mentioned previously, heparin andLMWH are reversed with protamine. There are no

292

Case 87. Intracerebral hemorrhage and anticoagulation

Table 87.1. Reversal of anticoagulation.

Coumadin 1. Vitamin K 10 mg intravenously2. Fresh frozen plasma 15–20 mL/kgConsider:3. Prothrombin complex concentrate4. Recombinant activated factor VII

15–90 mcg/kg

Intravenous heparin 1. Protamine 1 mg per 100 U of heparin

LMWH 1. Protamine 1 mg per 100 anti-Xa unitsof LMWH (1 mg of enoxaparin is∼100anti-Xa units), if the LMWH was givenwithin 8 hours

LMWH, low molecular weight heparins.Adapted from references [2, 9].

agents that directly correct the coagulopathy inducedby direct thrombin inhibitors or fondaparinux.

Vitamin K reverses the coagulopathy caused bycoumadin. Vitamin K can be given by a variety ofroutes, though the IV formulation is recommended forlife-threatening bleeding, such as an ICH [9]. Othersfeel that the IV preparation may cause severe anaphy-laxis, and thus recommend a subcutaneous route. Therecommended dose is 10 mg and it can be hours, up to24 hours, before it effectively corrects the coagulopa-thy [10].

While waiting for vitamin K to take effect, manypatients are treated with FFP. Fresh frozen plasma isa collection of donor plasma that contains all clottingfactors. The cost is low and it is readily available –though it must be unfrozen and may need to be bloodtype matched [10]. Fresh frozen plasma is given in adose of 15–20 mL/kg [9]. The precise concentration ofclotting factors varies between batches of FFP leadingto unpredictable responses [9]. A large volume of fluidmay need to be infused and, thus, there is a potentialto precipitate congestive heart failure [9, 10]. Addition-ally, there are concerns about transmission of infectionand transfusion-related lung injury with FFP [10].

Though not commonly used in the USA, PCC isused internationally to correct coagulopathy due tocoumadin. Prothrombin complex concentrates are aheterogenous group of products that contain factor IXand variable amounts of factors II, VII, and X [10].They correct the INRwithinminutes – with a minimalamount of volume infused.There are concerns becauseof thrombotic complications and different PCC prod-ucts contain differing amount of coagulation factors[10].

While there has not been a randomized con-trol trial for patients with coumadin-associated ICH,

rFVIIa is used in this setting [10], often with the com-bination of vitamin K and FFP. Recombinant factorVIIa acts by binding to tissue factor and activatingthrombin. Concerns regarding rFVIIa include: cor-recting the INR without reversing the coagulopathy(i.e., the lab value is normal, but patient remains coag-ulopathic), thrombotic complications, and high costs[10].

Finally, the anticoagulation effects of heparin arereversed by protamine. Heparin has a half life of60–90 minutes and the dose of protamine needs tobe adjusted to account for heparin’s metabolism. Ifheparin was immediately given, 1 mg of protaminereverses 100 U of heparin [9].The dose of protamine isproportionally decreased as time has elapsed since thelast heparin dose [9]. Protamine dose not fully reverseLMWH, but it can still be given as a one-time doseof 1 mg per 100 anti-Xa units (1 mg of enoxaparinequals 100 anti-Xa units) [2]. Protamine is given intra-venously by slow injection (5mg/min) as faster admin-istration rates can cause hypotension [9].

Anticoagulation in patients with anintracerebral hemorrhageAfter an ICH, the initial indication for anticoagulationremains and the balance between prevention of throm-boembolism and further ICH can be difficult. Patientswith ICH also suffer complications such as deep veinthrombosis and pulmonary embolus that necessitateanticoagulation.

There are no high-quality prospective data to guiderestarting anticoagulation in these patients. Much ofthe decisionmaking depends on the underlying reasonfor anticoagulation. For instance, the risk of throm-boembolism is greater for patients with mechanicalheart valves than in patients with atrial fibrillationwithout a prior stroke or vascular risk factors. Thepresence of microbleeds on T2∗ imaging may increasethe risk of future hemorrhages [9]. Ultimately, thedecision to restart anticoagulation is one that shouldbe made in collaboration with the patient. If anticoag-ulation is to be restarted it is typically done 1–2 weeksafter the initial ICH after serial neuroimaging hasshown stability of the hematoma.

ConclusionsIn conclusion, anticoagulation is effective in treatingand preventing thromboembolic disease, but there is

293


a risk of ICH. Most anticoagulation agents can bereversed. Correction of the coagulopathy should occurquickly to prevent hematoma growth, a predictor ofpoor outcome.

References1. J. Ansell, J. Hirsh, E. Hylek et al. Pharmacology and

management of the vitamin K antagonists: AmericanCollege of Chest Physicians Evidence-Based ClinicalPractice Guidelines. Chest 2008; 133: 160S–98S.

2. J. Hirsh, K. A. Bauer,M. B. Donati et al. Parenteralanticoagulants: American College of Chest PhysiciansEvidence-Based Clinical Practice Guidelines. Chest2008; 133: 141S–59S.

3. R. G. Hart, B. S. Boop,D. C. Anderson. Oralanticoagulants and intracranial hemorrhage: factsand hypotheses. Stroke 1995; 26: 1471–7.

4. M. C. Fang, Y. Chang, E. M. Hyleck et al. Advancedage, anticoagulation intensity, and risk for intracranialhemorrhage among patients taking warfarin for atrialfibrillation. Ann Intern Med 2004; 141: 745–52.

5. R. G. Hart, S. B. Tonarelli, L. A. Pearce. Avoidingcentral nervous system bleeding during

antithrombotic therapy: recent data and ideas.Stroke 2005; 36: 1588–93.

6. J. Rosand,M. H. Eckman, K. A. Knudsen et al. Theeffect of warfarin and intensity of anticoagulation onoutcome of intracerebral hemorrhage. Arch Intern Med2004; 164: 800–4.

7. J. J. Filbotte, N. Hagan, J. O’Donnell et al. Warfarin,hematoma expansion, and outcome of intracerebralhemorrhage. Neurology 2004; 63: 1059–64.

8. M. L. Flaherty,H Tao,M. Heverbusch et al. Warfarinuse leads to larger intracerebral hematomas. Neurology2008; 71: 1084–9.

9. J. Broderick, S. Connolly, E. Feldmann et al.Guidelines for the management of spontaneousintracerebral hemorrhage in adults: 2007 update: aguideline from the American Heart Association/American Stroke Association Stroke Council, HighBlood Pressure Research Council, and the Quality ofCare and Outcomes in Research InterdisciplinaryWorking Group. Stroke 2007; 38: 2001–23.

10. J. N. Goldstein, J. Rosand, L. H. Schwamm. Warfarinreversal in anticoagulant-associated intracerebralhemorrhage. Neurocrit Care 2008; 9: 277–83.

294


88 Cerebral amyloid angiopathy-relatedintracranial hemorrhageLesli E. Skolarus and Teresa L. Jacobs

Intracerebral hemorrhage (ICH) accounts for approxi-mately 10% of the almost 800 000 strokes that occur intheUSA annually.Themost common risk factors asso-ciatedwith ICHare hypertension and cerebral amyloidangiopathy (CAA). While hemorrhages located in thethalamus, putamen, globus pallidus, pons and cerebel-lum are often attributed to hypertension, lobar hemor-rhages are thought to be caused by CAA.

Case descriptionAn 80-year-old male presented to the emergencydepartment after being found fallen by his wife. Hisblood pressure was slightly elevated at 150/90 mmHgand his pulse was 70 beats/minute. On examination,his level of consciousness was diminished and herequired constant noxious stimulation to remain alert.He had a decreased right nasolabial fold, a weak coughand gag reflex as well as a notable right hemipare-sis. His trachea was intubated for airway protection.A noncontrast head computed tomography (CT) scanrevealed a 3.4 × 3.0 cm area of hyperdensity in hisleft parietal lobe with no evidence of intraventricularextension of the hematoma. His coagulation panel andplatelets were within normal limits. He was admittedto the neurologic intensive care unit (ICU) for closeneurologic monitoring and ventilatory support. Man-agement of presumed increased intracranial pressure(ICP) was initiated with head of the bed placed at30 degrees to facilitate venous drainage.

On hospital day 2, the patient appeared more som-nolent and a repeat head CT revealed increased masseffect on the right lateral ventricle. Mannitol 20% wasgiven at 0.25 g/kg and serum osmolarity was checkedfor the next 48 hours.

On hospital day 4, the patient’s level of awarenesshad improved and repeat head CT revealed decreas-ing edema; the trachea was extubated. Magnetic reso-nance imaging (MRI) was performed, showing multi-ple areas of microhemorrhage and macrohemorrhage

on gradient echo as well as the acute left parietal hem-orrhage (Figure 88.1).The following day, he was stableand transferred to a general medical ward to continuevigorous rehabilitation.

DiscussionCerebral amyloid angiopathy is the most commoncause of spontaneous lobar ICH. It is characterizedby the deposition of beta-amyloid peptide into themedia and adventitia of small arteries and capillar-ies. The beta-amyloid peptide is toxic to the vascularsmooth muscle cells leading to damage to the bloodvessel wall and consequent hemorrhage. Not surpris-ingly, advancing age is the strongest risk factor forCAA. In addition, the presence of APOE2 and APOE4alleles,which result in increased beta-amyloid peptidedeposition and degeneration of vessel walls, are alsorisk factors for CAA [1].

Presentation and diagnosisThe presentation of CAA-related ICH is comparableto all lobar hemorrhages. Patients present with acuteonset of headache, decreased level of consciousness,focal neurologic signs, especially cortical signs suchas aphasia, neglect, or seizures. Currently, diagnosisis based on the Boston Criteria utilizing clinical data,autopsy, surgical pathology, or MRI (Table 88.1) [2].The criteria rely on detecting the late manifestations ofCAA-related vascular damage such as hemorrhage andmicrobleeding rather than the vascular amyloid itself.Recently, two potential biomarkers have been identi-fied that detect the presence of the diseasewithout hav-ing to rely on the presence ofmicrohemorrhages, a lateconsequence of the disease. Pittsburgh Compound B,a beta-amyloid-binding compound usedwith positronemission tomography imaging, distinguished patientswith CAA from healthy volunteers and patients withAlzheimer’s disease via localization of the compound’s

295


Figure 88.1. A magnetic resonance imaging, axial T2∗ gradient echo image showing multiple areas of lobar hemorrhage.

retention [3]. In addition, cerebral spinal fluid analy-sis of amyloid beta 40 and amyloid beta 42 also distin-guished CAA from healthy controls and Alzheimer’sdisease patients [4].

Medical managementAs with all critically ill patients, assessment and sta-bilization of the airway, breathing, and circulation arethe essential first steps. This is followed by rapid diag-nosis. Because it is difficult to clinically distinguishICH from ischemic stroke, emergent diagnostic imag-ing with either a head CT or MRI is recommended[5]. If available, admission to a neurologic or neuro-surgical ICU is preferred as it has been associated withdecreased mortality [5].

Medical treatment of CAA-related ICH is basedon control of the hemorrhage, management of blood

pressure, management of elevated ICP, and treatmentof seizures, fevers, and hyperglycemia. The FactorSeven forAcuteHemorrhagic Stroke (FAST) trial com-pared two doses of recombinant factor VII (rFVIIa)to placebo within 4 hours of onset of spontaneousICH. While both doses of rFVIIa resulted in thereduction of hematoma volume, the small hematomavolume reductions did not result in improved out-come following ICH [6]. Unfortunately, at this timethere are no effective treatments to reduce hematomaexpansion.

The data on blood pressure management inacute ICH are inconclusive. Currently underway arethe Antihypertensive Treatment in Acute CerebralHemorrhage (ATACH) and Intensive Blood PressureReduction in Acute Cerebral Hemorrhage (INTER-ACT) trials evaluating blood pressure management inthe setting of acute ICH. Until the results of these

296

Case 88. Cerebral amyloid angiopathy-related intracranial hemorrhage

Table 88.1. Boston Criteria for diagnosis of cerebral amyloidangiopathy-related hemorrhage.

Definite CAA Postmortem examination showslobar cortical, or corticosubcorticalhemorrhage ICH, severe CAA and noother diagnostic lesion

Probable CAAwith supportingpathology

Clinical data and pathologic tissue(evacuated hematoma or corticalbiopsy) showing some lobar, cortical,or corticosubcortical hemorrhage,some degree of CAA and absence ofother diagnostic lesion

Probable CAA Clinical data and MRI or CT withmultiple hemorrhages restricted tolobar, cortical, or corticosubcorticalregions (cerebellar hemorrhageallowed), age ≥55 years and no othercause of hemorrhage

Possible CAA Clinical data and MRI or CT with singlelobar, cortical, or corticosubcorticalhemorrhage, age ≥55 years and noother cause of hemorrhage

CAA, cerebral amyloid angiopathy; MRI, magnetic resonanceimaging; CT, computed tomography.

trials are published, we recommend mean arterialpressures of 70–110mmHg based on our clinical expe-rience. In our neurologic ICU, this is achieved viaintravenous labetalol as needed or an intravenous infu-sion of nicardipine.

With regards to ICP management in the setting ofICH, no randomized clinical trial has demonstratedthe efficacy of monitoring intracerebral pressure orcerebral perfusion pressure [5].We recommend a step-wise approach starting with the patient’s head midlineand elevated to 30◦ in order to improve venous out-flow, followed by osmotic therapy with either manni-tol or hypertonic saline. As a temporizing measure forthe acute patient in danger of herniation, hyperventi-lation with target CO2 of 30–35 mmHg is an effectivemethod for rapid reduction of ICP. Finally, for patientswith refractory intracranial hypertension, barbituratecoma is initiated with electroencephalography moni-toring in order to titrate the barbiturate dosing to burstsuppression.

Clinical or electrographic seizures should betreated with intravenous antiepileptic. Euglycemiashould be targeted for all patients and may requireinsulin. Finally, fevers should be aggressively con-trolled with acetaminophen and other measures suchas a cooling blanket if needed.

Surgical managementThe International Surgical Trial in Intracerebral Hem-orrhage (STICH), published in 2005, was a multi-center randomized clinical trial comparing surgerywithin 96 hours of ICH onset to best medical treat-ment for patients with spontaneous supratentorialICH. Patients were enrolled if their treating neurosur-geon felt clinical equipoise between surgery and med-ical management. Overall, no difference in functionaloutcome or mortality were found [7]. Subgroup anal-ysis revealed that surgery may be beneficial in treat-ing patients with lobar hematomas within 1 cm of thesurface of the brain [7]. Presently a randomized trial,Surgical Trial in Intracerebral Hemorrhage II (STICHII), is underway evaluating this limited population. Atthe present time, surgery is reserved for a highly selectgroup of patients at the discretion of the treating neu-rosurgeon and neurointensivist.

PrognosisUnfortunately, the prognosis is poor for many ICHpatients. Limitations in life-sustaining care may con-found our current prognostication system for ICHpatients [5]. In our opinion, prognosis after ICH maybe clouded by the self-fulfilling prophecy of with-drawal of care. We encourage aggressive full supportat least for the first 24 hours after acute ICH.

Risk of recurrent hemorrhageFollowing the initial CAA-related ICH, the presenceof additional areas of microhemorrhage on the base-lineMRIGradient Echo sequence, new areas ofmicro-hemorrhage on subsequent scans and presence of theAPOE2 or APOE4 alleles result in increased rates ofrecurrent hemorrhage [8, 9]. The risk of recurrenthemorrhage is important to take into account becausedecision models for initiation of anticoagulation afterICH are based on estimates of this risk. For exam-ple, when clinicians contemplate restarting anticoag-ulation in a patient with CAA-related ICH and atrialfibrillation, an accurate assessment of the risk of recur-rent hemorrhage is required to weigh the risks andbenefits of anticoagulation.

ConclusionThe management of CAA-related ICH is complexand close neurologic monitoring is essential. Withthe development of new potential biomarkers for the

297


disease, we may begin to explore therapeutic optionsbefore patients develop ICH.

References1. S. M. Greenberg, J. P. Vonsattel, A. Z. Segal et al.

Association of apolipoprotein E epsilon2 andvasculopathy in cerebral amyloid angiopathy.Neurology 1998; 50: 961–5.

2. K. A. Knudsen, J. Rosand,D. Karluk et al. Clinicaldiagnosis of cerebral amyloid angiopathy: validation ofthe Boston criteria. Neurology 2001; 56: 537–9.

3. K. A. Johnson,M. Gregas, J. A. Becker et al. Imagingof amyloid burden and distribution in cerebralamyloid angiopathy. Ann Neurol 2007; 62: 229–34.

4. M.M. Verbeek, B. P. Kremer,M. O. Rikkert et al.Cerebrospinal fluid amyloid beta(40) is decreased incerebral amyloid angiopathy. Ann Neurol 2009; 66:245–9.

5. J. Broderick, S. Connolly, E. Feldmann et al.Guidelines for the management of spontaneousintracerebral hemorrhage in adults: 2007 update:a guideline from the American Heart Association/

American Stroke Association Stroke Council, HighBlood Pressure Research Council, and the Quality ofCare and Outcomes in Research InterdisciplinaryWorking Group. Stroke 2007; 38: 2001–23.

6. S. A. Mayer,N. C. Brun, K. Begtrup et al. Efficacy andsafety of recombinant activated factor VII for acuteintracerebral hemorrhage. N Engl J Med 2008; 358:2127–37.

7. A. D. Mendelow, B. A. Gregson,H. M. Fernandeset al. Early surgery versus initial conservativetreatment in patients with spontaneous supratentorialintracerebral haematomas in the International SurgicalTrial in Intracerebral Haemorrhage (STICH): arandomised trial. Lancet 2005; 365: 387–97.

8. S. M. Greenberg, J. A. Eng,M. Ning et al.Hemorrhage burden predicts recurrent intracerebralhemorrhage after lobar hemorrhage. Stroke 2004; 35:1415–20.

9. H. C. O’Donnell, J. Rosand, K. A. Knudsen et al.Apolipoprotein E genotype and the risk of recurrentlobar intracerebral hemorrhage. N Engl J Med 2000;342: 240–5.

298

Part XII Traumatic brain injuryCase

89 Traumatic brain injuryVenkatakrishna Rajajee

Traumatic brain injury (TBI) is the leading cause ofdeath and disability in children and adults from ages1 to 44 [1]. The primary goal of management of thepatient with severe TBI (Glasgow Coma Scale (GCS)3–8) in the intensive care unit (ICU) is to preventsecondary brain injury.

Case descriptionA 26-year-old female was brought to the emergencyroom after a motor vehicle collision. In the emer-gency room, her GCS was 6 (E1M3V2). Her pupilswere 3 mm on the left, unreactive, and 6 mm on theright, reactive. Following endotracheal intubation thepatient was given mannitol 1 g/kg and a noncontrastcomputed tomography (CT) scan of the brain wasurgently obtained. This revealed a right temporopari-etal epidural hematoma 3.5 cm in greatest diameterwith a 1.6 cm shift of the midline (Figure 89.1).

The patient was taken to the operating room (OR)immediately. Under general anesthesia, a craniotomywas performed and the epidural clot evacuated. Thecraniotomy bone flap was replaced. A right-sidedexternal ventricular drain (EVD) was placed. A post-operative CT revealed an effective decompression anda CT of the cervical spine revealed no bony abnormal-ity. The patient was admitted to the ICU.

In the ICU, her GCS was 6T, with an intracranialpressure (ICP) of 6 mmHg. She was mechanically ven-tilated with the Assist Control mode. She was givena phenytoin loading dose followed by a maintenancedose for seizure prophylaxis. By the next morning,she was seen to withdraw with the right side and fre-quently “buck” the ventilator. At this time herGCSwas6T (E1M4V1T), with an ICP of 28 mmHg. Propofoland fentanyl infusions were started. Following initia-tion of the sedative infusion the mean arterial pres-sure dropped to 60mmHg and the ICP was 18mmHg,resulting in a cerebral perfusion pressure (CPP) of60− 18= 42 mmHg. A norepinephrine infusion was

started to maintain the CPP �60 mmHg. A naso-gastric feeding tube was placed and feeding wasadvanced to goal daily caloric requirement over thenext 48 hours. Heparin 5000 U subcutaneously threetimes daily was started 24 hours postoperatively.Through the next 48 hours her ICP increased to24 mmHg despite the sedation and analgesia.The ven-tricular catheter was opened to drain on several occa-sions for an ICP �20 cm; 3–5 mL of cerebrospinalfluid (CSF) was drained each time, followed by cessa-tion of drainage. Eventually, on postoperative day 3,the ICP was 30 mmHg despite intermittent CSFdrainage and sedation. She was given a 0.25 g/kg intra-venous push of mannitol and her ICP decreased to12 mmHg within 30 minutes. Repeat CT scansrevealed no new hematoma. The set ventilator ratewas adjusted to maintain the PaCO2 30–35 mmHg.Over the next 24 hours, bolus doses of mannitol wererequired every 4 hours for ICP �20 mmHg, whilemaintaining a serum osmolality 300–320 mOsm/kg.On postoperative day 4 the ICP was 30 mmHg despitethe use of osmotherapy. The patient was taken backto the OR and the craniotomy flap removed, the duraopened and a duroplasty performed (Figure 89.2).Postoperatively, the ICP was 6 mmHg and remained�20 mmHg for the duration of the patient’s ICU stay.No further use ofmannitol was required.On postoper-ative day 7, the patient was following commands. Shehad 5/5 strength on the right side and 2/5 strength onthe left.The trachea was extubated successfully and theventricular catheter removed. She was transferred outof the ICU 24 hours later.

DiscussionThe guidelines for the management of the patientwith severe TBI published by the Brain Trauma Foun-dation (BTF) [2] form the basis for the followingdiscussion. Management of TBI often begins with adecision to perform endotracheal intubation in the

299

XII. Traumatic brain injury

Figure 89.1. Right-sidedtemporoparietal epidural hematoma,3.5 cm maximum thickness, 1.6 cmmidline shift.

Figure 89.2. Right craniotomy bone flap removed with effectivedecompression.

emergency room. A widely used rule of thumb is“eight-intubate,” with intubation performed when theGCS is 8 or less. Other factors such as intoxication andthe risk of aspiration may need to be considered evenif the GCS is greater than 8. Hypotension (a systolicblood pressure less than 90 mmHg) has most consis-tently been associated with poor neurologic outcome

following severe TBI. Hypoxia (SpO2 �90%) is alsoassociated with poor outcome. Early aggressive resus-citation of low blood pressure and poor oxygenationis therefore of critical importance. Clinical signs oftranstentorial herniation (dilated pupil) warrant theimmediate use of a bolus of mannitol IV.The next stepin management is the removal of specific mass lesions,most commonly subdural, extradural, and intracere-bral hematomas (contusions) guided by urgent non-contrast head CT. Guidelines for the surgical manage-ment of TBI and removal of mass lesions have alsobeen published by the BTF (Table 89.1) [3].The epidu-ral hematoma in the case discussed was �2 cm ingreatest thickness and resulted in �0.5 cm of midlineshift in an individual with GCS ≤8; surgical evacu-ation is therefore urgently indicated. A decision wasthen made to place an ICP monitor, as monitoringof intracranial pressure is necessary in patients withTBI likely to have or develop intracranial hyperten-sion (ICP �20 mmHg). This consists of patients withGCS ≤8 with an abnormal head CT as well as patientswith GCS ≤8 and a normal head CT who are olderthan 40, hypotensive (SBP �90 mmHg) or have pos-turing on neurologic examination (Table 89.2). Thispatient with GCS 6 prior to transport to the operat-ing room therefore had an EVD placed. The advan-tage of using an EVD, or ventriculostomy catheter, is

300


Table 89.1. Indications for surgical evacuation of traumatic mass lesions [3].

Epidural hematomaAn epidural hematoma (EDH) greater than 30 cm3 should be surgically evacuated regardless of the patient’s GCS score

An EDH less than 30 cm3 and with less than a 15-mm thickness and with less than a 5-mmmidline shift in patients with a GCS scoregreater than 8 without focal deficit can be managed nonoperatively with serial CT scanning and close neurologic observation in aneurosurgical center

Subdural hematomaAn acute subdural hematoma (SDH) with a thickness greater than 10 mm or a midline shift greater than 5 mm on CT scan should besurgically evacuated, regardless of the patient’s GCS score

All patients with acute SDH in coma (GCS score less than 9) should undergo ICP monitoring

A comatose patient (GCS score less than 9) with an SDH less than 10-mm thick and a midline shift less than 5 mm should undergo surgicalevacuation of the lesion if the GCS score decreased between the time of injury and hospital admission by 2 or more points on the GCSand/or the patient presents with asymmetric or fixed and dilated pupils and/or the ICP exceeds 20 mmHg

Traumatic parenchymal lesion (contusion)Patients with parenchymal mass lesions and signs of progressive neurologic deterioration referable to the lesion, medically refractoryintracranial hypertension, or signs of mass effect on CT scan should be treated operatively

Patients with GCS scores of 6 to 8 with frontal or temporal contusions greater than 20 cm3 in volume with midline shift of at least 5 mmand/or cisternal compression on CT scan, and patients with any lesion greater than 50 cm3 in volume should be treated operatively.

Patients with parenchymal mass lesions who do not show evidence for neurologic compromise, have controlled ICP, and no significantsigns of mass effect on CT scan may be managed nonoperatively with intensive monitoring and serial imaging

Decompressive procedures, including subtemporal decompression, temporal lobectomy, and hemispheric decompressive craniectomy,are treatment options for patients with refractory intracranial hypertension and diffuse parenchymal injury with clinical and radiographicevidence for impending transtentorial herniation

Table 89.2. Indications for intracranial pressure monitoring insevere TBI [2].

GCS 3–8 plus abnormal CT

or

GCS 3–8 plus

Normal CT and any two of the following

Age � 40

Systolic blood pressure �90

Unilateral or bilateral posturing

that the catheter can also be used as a first-line ther-apeutic tool to reduce ICP (via drainage of CSF). Themajor alternative is an intraparenchymal ICP monitor(such as the Codman and Camino monitors). Whilethese catheters cannot drain CSF, they are technicallyeasier to place and carry a significantly lower risk ofhemorrhage and infection. Anticonvulsant use is indi-cated for prophylaxis against early (�7 days) posttrau-matic seizures, however, there is no value in primaryprophylaxis beyond that period. Early seizures do notseem to correlate with poor long-term neurologic out-come. Once in the ICU, a decision often needs to bemade about continued use of a semi-rigid collar for

immobilization of the cervical spine. In the patientwith severe TBI in whom a valid clinical examinationto clear the cervical spine cannot be performed, a non-contrast CT of the cervical spine is generally adequateto clear the cervical spine and remove the collar [4].Some physicians prefer to continue use of the semi-rigid collar until an adequate clinical examination, aflexion-extension X-ray or magnetic resonance imageof the cervical spine can be obtained to exclude liga-mentous and other soft tissue injury.

Sedation and analgesia is of critical importance inthe TBI patient, not only to permit safe and effec-tive mechanical ventilation and avoid unintended tra-cheal extubation or line removal but also to adequatelycontrol ICP. In the case discussed, a combination ofpropofol and fentanyl was used. Propofol has theadvantage of having a short half life to permit neu-rologic examination as required and is a very effec-tive sedative. Caution should be employed with long-term use, as propofol infusion syndrome may occur.Analgesia is important, particularly in the postoper-ative TBI patient. The goal is to maintain the ICP�20 mmHg and the CPP �60 mmHg, since therisk of secondary neurologic injury is considered tobe significant outside these values. A vasopressor

301


infusion was therefore warranted to maintain aCPP �60 mmHg. Maintaining the CPP �70 mmHgis no longer recommended, since the risk of acute res-piratory distress syndrome appears to be greater whenmeasures are taken to attain this goal, without a corre-sponding benefit in mortality or neurologic outcome.In the case discussed, when the patient’s ICP remainedconsistently �20 mmHg, the first step taken was todrain some CSF through the EVD. In general, CSFdrainage is performed on an intermittent, as-neededbasis for 5–10 minutes for ICP �20 mmHg, with thecatheter monitoring pressures at baseline. Early ini-tiation of nutritional support is recommended, withenteral feeding the preferred route and the total dailycalorific requirement goal attained within 7 days, usu-ally sooner (within 72 hours for the patient discussed).Deep venous thrombosis prophylaxis is necessary inall patients with severe TBI in the ICU. A combinationof immediate use of a sequential compression deviceand subcutaneous heparin (either unfractionated orlow molecular weight) after 24–48 hours is safe andeffective even in patients admitted with intracranialhematomas.

If the ICP cannot be controlled with the use ofappropriate sedation/analgesia and intermittent CSFdrainage, then as-needed osmotherapy is used.Manni-tol and hypertonic saline are both reasonable optionsfor this purpose. Mannitol often begins to act within10–30 minutes with a clinical effect that lasts for2–6 hours.The use ofmannitol carries the risk of dehy-dration, renal injury, and hypotension. This shouldbe judiciously avoided with the use of fluid resus-citation, usually crystalloid. While using mannitolthe serum osmolality is typically not pushed beyond310–320 mOsm/kg, to avoid the risk of renal injury.Prolonged or continuous use of mannitol may alsoresult in rebound cerebral edema and intracranialhypertension. Hypertonic saline does not cause dehy-dration or hypotension but may result in fluid over-load. Osmotherapy in general should be used only forepisodes of elevated ICP or, prior to the placement ofan ICP monitor, for clinical signs of herniation.

Hyperventilation to PaCO2 �30–35 mmHg israpidly effective in reducing ICP but may result incerebral ischemia and worsened neurologic outcomes.The fall in ICP with hyperventilation is transient andrebound intracranial hypertension may result froma too-rapid increase in the PaCO2 following initialhyperventilation. Hyperventilation to �30–35 mmHgshould generally be avoided in TBI and, if used,

should be done so in conjunction with ancillary brainoxygenation monitoring to detect ischemia. There isno role for high-dose corticosteroid therapy in themanagement of severe TBI. Mild hypothermia (to32–34 ◦C) lowers ICP and, on the basis of pooleddata, may result in improved GlasgowOutcome Scaleswith an improvement in mortality when used for�48 hours. In the absence of conclusive evidence ofbenefit from large randomized controlled trials, how-ever, the routine use of hypothermia in all patientswith severe TBI is not recommended by the BTF.Similarly, although the use of jugular bulb oximetryand brain tissue oxygenation monitors to detect brainhypoperfusion/ischemia shows significant promise, inthe absence of conclusive outcome data, the routineuse of these techniques in severe TBI is not specificallyrecommended by the BTF.

For the patient who develops intracranial hyper-tension refractory to the above measures, the availableoptions includemild hypothermia, barbiturate (pento-barbital or thiopental) infusion titrated to electroen-cephalogram burst suppression, and decompressivecraniectomy. Increasingly, decompressive craniectomyis being used in patients with severe TBI and intracra-nial hypertension. In the case discussed, the originalcraniotomy flap was removed with subsequent excel-lent control of ICP followed by neurologic improve-ment. Frequently, the bone flap is not replaced at thetime of initial hematoma evacuation when significantedema and intracranial hypertension is anticipated.The durotomy (with duroplasty) plays an importantrole in reduction of ICP. While this technique is veryeffective in controlling ICP, outcome data from largerandomized controlled trials are not yet available [5].Performing a decompressive craniectomy commits thepatient to a subsequent cranioplasty to replace thebone flap and carries with it the risks of acute hem-orrhage, extra-axial fluid accumulation and infection,among others. Once the patient’s neurologic conditionimproves, tracheal extubation must not be delayed,with consideration given primarily to the patient’s abil-ity to cough and clear secretions spontaneously ratherthan a specificGCS number. Performing tracheostomyin TBI patients prior to 7 days will decrease the num-ber of days on mechanical ventilation but may notreduce the risk of ventilator-associated pneumonia oralter mortality. The management of patients with TBIis often challenging but rewarding, with a significantnumber of patients achieving a good long-term neu-rologic outcome.

302


References1. J. A. Langlois,W. Rutland-Brown,M.M.Wald. The

epidemiology and impact of traumatic brain injury:a brief overview. J Head Trauma Rehabil 2006; 21:375–8.

2. Brain Trauma Foundation. American Association ofNeurological Surgeons; Congress of NeurologicalSurgeons; Joint Section on Neurotrauma and CriticalCare, AANS/CNS. Guidelines for the management ofsevere traumatic brain injury. J Neurotrauma 2007; 24Suppl. 1: S1–95.

3. M. R. Bullock, R. Chesnut, J. Ghajar et al. SurgicalManagement of Traumatic Brain Injury Author Group.Guidelines for the surgical management of traumaticbrain injury. Neurosurgery 2006; 58: S2.

4. N. D. Tomycz, B. G. Chew, Y. F. Chang et al. MRI isunnecessary to clear the cervical spine in obtunded/comatose trauma patients: the four-year experience ofa level I trauma center. J Trauma 2008; 64: 1258–63.

5. J. Sahuquillo, F. Arikan. Decompressive craniectomyfor the treatment of refractory high intracranialpressure in traumatic brain injury. Cochrane DatabaseSyst Rev 2006; 25: CD003983.

303


90 Elevated intracranial pressureRichard Bowers and George A. Mashour

Intracranial pressure (ICP) is an important parameterguiding the management of patients in the intensivecare unit (ICU) with intracranial pathology. Promptand effective interventions in the event of elevated ICPare necessary to reduce ischemic neuronal loss andimprove long-term neurologic outcomes.

Case descriptionA 57-year-old male presented following a fall downa flight of stone stairs. He had consumed a signifi-cant quantity of alcohol prior to the fall and was laterfound unresponsive. Paramedics secured his cervicalspine and the trachea was intubated at the scene. Fol-lowing transfer to hospital he underwent a noncon-trast head computed tomography (CT) scan, whichrevealed a thin left-sided subdural hemorrhage anddiffuse petechial parenchymal hemorrhage. The sub-dural hemorrhage was evacuated and a right frontalintraventricular ICP catheter was inserted under gen-eral anesthetic. The patient was subsequently trans-ferred to the ICU where his ICP was measured con-tinuously. Initial treatment of the patient consistedof infusions of sedatives and ventilation to a targetPaCO2 of 35 mmHg, but persistent spikes of his ICPwere recorded. Acute treatment of these ICP increasesconsisted of an intravenous bolus of mannitol andadditional boluses of sedation. Treatment was esca-lated with the commencement of an infusion of neu-romuscular blocking agents, which improved ICP con-trol. In addition to these treatments, the patient had anasogastric feeding tube inserted, blood glucose waskept �150 mg/dL, full-length lower limb compressionstockings fitted, and normothermia was ensured.

He began to demonstrate poor intracranial elas-tance approximately 12 hours later, with frequent ICPspikes that persisted. These spikes were treated withhypertonic saline and a repeat head CT found grosscerebral edema with evidence of hemorrhage. Treat-ment was escalated further to include active cooling to

a core temperature not lower than 34 ◦C. Cooling wasachieved by the use of a large-bore femoral vein cool-ing catheter. Shivering was prevented by the continuedadministration of neuromuscular blocking drugs.

Further episodes of elevated ICP occurred oncethe core temperature target had been achieved and itwas therefore decided to institute a barbiturate infu-sion. The patient’s electroencephalogram was moni-tored during barbiturate therapy, allowing dose titra-tion of the barbiturates to achieve burst suppression.Additional neuroimaging revealed no further changes.

Therapy continued in this manner for several dayswith little sustained improvement of the patient’s ICP,although acute elevations were less common.

Hypernatremia had become an increasingly diffi-cult problem over the period in the ICU. Laboratorytests had begun to show evidence of acute renal failureafter 8 days due to the combination of critical illness,hyperosmotic treatments, and an incipient respiratorytract infection. A decision was made with the patient’sfamily to withdraw active treatment after 10 days andhe died shortly thereafter.

Discussion

Why treatBrain injuries frommultiple etiologies consist of a pri-mary insult that may be surgically amenable, as wellas subsequent secondary insults that are the focus ofintensive care management. The loss of neurons fromsecondary insults may be significantly greater thanthat of the primary insult. Appropriate and timelymanagement may allow survival of these neurons and,hopefully, improvement in a patient’s neurologic out-come and functional status.

Who to treatThere are many causes of increased ICP, both intracra-nial and extracranial (see Table 90.1). Treatment of

304

Case 90. Elevated intracranial pressure

Table 90.1. Causes of intracranial hypertension.

Mass effectsTumorHemorrhage – extradural, subdural, subarachnoidIntracerebral abscess

VascularHypercarbia, hypoxemiaHyperemiaPyrexiaCerebral venous sinus thrombosis

Excess cerebrospinal fluidObstructive – Arnold–Chiari malformation, meningocele,

meningitis, tumors, hemorrhageExcess production – choroid plexus tumor, subarachnoid

hemorrhageFailed resorption – meningitis, subarachnoid hemorrhage

EdemaDiffuse axonal injuryInfection – encephalitis, meningoencephalitis

MiscellaneousAcute liver failureIdiopathic intracranial hypertension

the underlying cause may allow rapid resolution ofintracranial hypertension, although most will initiatea cascade cycle of increasing ICP, reducing cerebralperfusion, further intracerebral inflammation, edema,and ICP rises.

Why intracranial pressureIntracranial pressure has become central to the crit-ical care management of neurologic patients. Thereasons for this approach are several-fold, but primar-ily include theMonro–Kellie hypothesis of intracranialelastance, the relative ease of ICPmeasurement, quan-tifiable neurologic change based on ICPmanipulation,such as the recovery of pupil reactivity following ICPreduction, and the association between maintenanceof cerebral perfusion pressure (CPP) and improvedpatient outcomes, which relies on ICP measurement[1]. The Monroe–Kellie hypothesis suggests that sincethree noncompressible components (brain, blood, andCSF) are housed within a nondistensible vault (thecranium); changing the volume of one component ofnecessity means that either (1) the volume of anothermust change or (2) ICP must increase.

There are caveats to the use of ICP measurementthat must be borne in mind. First, ICP is not uniformthroughout the brain and the measured ICP maynot fully reflect the area of intracranial injury. Othersources of inaccuracy may include catheter position,

catheter type, or a separate issue such as clotting ofmanometer tubing. There are also direct risks asso-ciated with the use of monitoring catheters includ-ing infection (approximate rate 10%) and hemorrhage[2, 3].

When to treatIntracranial pressure is considered abnormal once itrises above 20 mmHg [1, 4]. Once this thresholdhas been passed, a methodical and thorough assess-ment of the patient should be instituted to (1) searchfor any reversible causes of the new deterioration,(2) maximize cerebral perfusion and oxygenation, and(3) reduce the volume of intracranial contents.

What to treatControl of intracranial pressure in the ICUdepends onthe manipulation of a variety of factors with the aim ofminimizing secondary insults. These factors involve:� optimizing cerebral blood flow and oxygenation� reducing cerebral oxygen consumption� reducing volume of intracranial contents

including:� blood – arterial and venous� cerebrospinal fluid� brain parenchyma – edema, mass lesions.

How to treat 1 – initial managementInitial management of raised ICP is straightforward,and may appear to consist of little more than “routine”ICUmanagement. The following are, however, impor-tant points.

Optimizing cerebral blood flow and oxygenationNormotension and cerebral perfusion pressure. Mainte-nance of CPP is essential. Frequency and duration ofhypotension are associated with increasing mortalityand morbidity in neurologically impaired patients [1].Cerebral perfusion pressure is indirectly dependent onICP:

CPP = MeanArterial Pressure − ICP

ACPP of�60mmHg is associatedwith a poor out-come and it is recommended that CPP be maintained�60mmHg at all times [1].This target is best achievedthrough both fluid resuscitation and vasopressor use.

305


There may be some advantage to the use of hyper-tonic saline in patients with traumatic brain injury asit significantly improved blood pressure when com-pared with resuscitation with normal saline [1] andit may also reduce cerebral edema. Failure to main-tain adequate CPP despite fluid resuscitation shouldprompt use of vasopressor support as swiftly as possi-ble. Norepinephrine is preferable to other vasopressorsdue to the relative cerebrovascular sparing of alpha-adrenergic receptors and its predictability of action.An escalating dose of norepinephrine causes concernand should prompt investigation for causes of intravas-cular fluid loss or poor cardiac output. Patients withnontraumatic etiologymay display systemic hyperten-sion as a reflex to maintain cerebral perfusion. Thisshould not be treated until CPP can be calculated, i.e.,after ICP monitoring is instituted.

Oxygenation and ventilation. All patients treated fortraumatic brain injury should have a PaO2 of greaterthan 11 kPa (approx 80 mmHg) [1]. It is reasonableto extend this goal to all patients treated for intracra-nial hypertension as hypoxiamay exacerbate ICP rises.The use of positive end expiratory pressure (PEEP)to accomplish this goal is not contraindicated and,in general, improved cerebral oxygenation will givebetter control of ICP and should outweigh any con-cerns regarding reduced intracranial venous returnfrom excessive intrathoracic pressure.

Cerebral blood vessels vasodilate as PaCO2 rises,reaching maximal dilatation at approximately 90–100mmHg (12–13 kPa) and ICP will consequently rise.Therefore, PaCO2 should be controlled to approxi-mately 35 mmHg (4.5 kPa). Further reductions ofPaCO2 should be avoided as vasoconstriction mayinduce ischemia, particularly in the early stages follow-ing traumatic head injury when cerebral blood flowis reduced [1, 4]; the long-term outcome is worsenedwith chronic hyperventilation [2]. The effect of hyper-ventilation is time-limited to cerebrospinal fluid pHbuffering and so may only be of use as a temporizingmeasure while other strategies are being employed.

Reducing cerebral oxygen consumption and demandSedation and analgesia should be provided to preventagitation and distress, thus reducing cerebral oxygenconsumption. The selection of drugs depends mainlyon the expected duration of sedation and any poten-tial side effects such as hypotension.There is no intrin-sic advantage of one drug over another for adequate

sedation provided it is performedwell. A bolus of seda-tion in the context of a longer-term infusion may alsobe used to acutely reduce ICP for ICU interventionse.g., tracheal suctioning, but caution should be paid toavoiding hypotension in this setting.

Avoidance of hyperthermia is also essential.Hyperthermia causes vasodilatation and increasesmetabolic rate [2]. Regular acetaminophen should beprovided to all patients and cooling blankets used ifnecessary to restore normothermia.

Hyperglycemia has been shown to have a detrimen-tal effect in other types of critical-care patients andshould be generally avoided; however, hypoglycemiamust absolutely be avoided due to the reliance of neu-ronal metabolism on glucose.

Reducing volume of intracranial contentsMass lesions of the brain such as tumors or abscessesmay be surgically evacuated with subsequent improve-ment of intracranial elastance. These processes alsoinitiate edema formation, which may subsequentlyworsen ICP.

Cerebrospinal fluid may be drained via an intra-ventricular catheter to provide control for ICP. Thisprocess requires patent ventricles to allow aspirationor passive drainage of CSF. No benefit is present onceventricles are effaced [2].

How to treat 2 – subsequent managementFailure to control ICP to �20 mmHg with the abovemanagement requires further treatment. It is impor-tant to consider follow-up brain imaging in order toensure that no further surgically amenable lesions havedeveloped.

Maintenance of cerebral perfusion and oxygenationA trial of elevated CPP may be attempted to exam-ine whether ICP control is improved. There is how-ever growing evidence that excessively elevated CPPis also associated with adverse cardiorespiratory andintracranial effects [1, 4]. Once a CPP is achieved thatis adequate for cerebral perfusion and avoidance ofischemia, no benefit exists for further elevation.

Reducing cerebral oxygen consumption and demandSeizures may be both a consequence and cause ofraised ICP. Seizures produce significant increases incerebral blood flow, cerebral anaerobic metabolism,cytotoxic edema, hypercarbia, and hypoxia, all ofwhich contribute to ICP increases. Seizure prophylaxis

306

Case 90. Elevated intracranial pressure

Figure 90.1. Algorithm for intracranial pressure and cerebral perfusion pressure management. Reproduced with permission from Helmyet al. Br J Anaesth 2007; 99: 32–42 [4].

307


is recommended for the first week following traumaticbrain injury [1, 4]. A diagnosis of occult seizures mustalways be considered in the event of an acute ICP riseand treated aggressively.

Reducing volume of intracranial contentsOsmotic diuretics produce a shift of water from theintracellular and interstitial spaces into the intravas-cular space. An intact blood–brain barrier is requiredfor this process. Mannitol can be used in a dose of0.25–1 g/kg and may reduce ICP for up to 6 hours[4]. The initial effects of mannitol include plasmavolume expansion and improved small vessel flowcharacteristics. Cerebral blood flow increases and con-sequently vasoconstriction ensues if cerebral autoreg-ulation is intact. Thereafter, its effects as an osmoticagent predominate. Chronic usage has been associatedwith rebound cerebral edema due to mannitol accu-mulation in the interstitium and intravascular dehy-dration. Hypertonic saline is an alternative that willalso provide intravascular replacement [2, 4]. Mainte-nance of high serum osmolarity will also help reducecellular swelling. Use of 0.9% sodium chloride as intra-venous fluid increases serum osmolarity, though itis associated with some electrolyte abnormalities [4].Osmolarity should be kept between 310–320mOsm/L,and following treatment should be returned slowly tobaseline to guard against rebound cerebral edema.

How to treat 3 – final optionsThe following options may offer control of ICP whenall the above have failed. A sequential progressionthrough these steps is suggested.� Neuromuscular paralysis may be useful if high

doses of sedation fail to control ICP particularlyduring ICU interventions such as suctioning andturning. Long-term use should be avoided,especially the steroid-based drugs, due to the risksof neuropathy and myopathy.

� The use of active hypothermia is morecontroversial, but it has shown some benefit forthe control of ICP in otherwise refractoryintracranial hypertension following traumaticbrain injury [4]. It can be undertaken by icepacking of the head, groin, and axillae, coveringthe patient with cold wet blankets, or withintravascular cooling catheters. Hypothermia hassignificant side effects and core temperatureshould not be lower than 34 ◦C.

� Barbiturate coma involves intravenous infusion ofbarbiturates to achieve burst suppression (periodsof low-voltage activity interspersed by shorterperiods of higher amplitude complexes) onelectroencephalogram. While in general ICP ismore likely to be controlled, outcomes followingbarbiturate coma are not improved. Side effectsare significant including hypotension, electrolyteabnormalities, infection and hepatorenaldysfunction [2]. It should only be considered inthe event of failure of all other treatment.

� Decompressive craniectomy may be effective inthe reduction of ICP when refractory to medicalmanagement [2]. It does offer the prospect ofimproved ICP control by effectively removing theconstraints to intracranial volume. However, thismay be at the cost of further intracranialcomplications and the overall benefit of theprocedure is not yet known. Studies arecomparing this treatment with barbiturate comafor outcomes in refractory intracranialhypertension.

Most management for ICP is standardized by use ofprotocols. An example of such a protocol is shown inFigure 90.1.

The management of intracranial hypertension inthe ICU requires a multifaceted stepwise approach,with particular care and attention to avoidable compli-cations such as hypoxia, hypercarbia, and hypotensionthat may dramatically worsen outcome. Rapid controlof ICP and resumption of adequate CPP are the focusof a management strategy, which requires a systematicapproach.

References1. Brain Trauma Foundation. American Association of

Neurological Surgeons; Congress of NeurologicalSurgeons; Joint Section on Neurotrauma and CriticalCare, AANS/CNS. Guidelines for the management ofsevere traumatic brain injury. J Neurotrauma 2007; 24Suppl. 1: S1–95.

2. L. Rangel-Castillo, S. Gopinath, C. Robertson.Management of intrcranial hypertension. Neurol Clin2008; 26: 521–41.

3. O. L. Cremer. Does ICP monitoring make a differencein neurocritical care? Eur J Anaesthesiol Suppl 2008;42: 87–93.

4. A. Helmy,M. Vizcaychipi, A. K. Gupta. Traumaticbrain injury: intensive care management. Br J Anaesth2007; 99: 32–42.

308


91 Succinylcholine in the patient withincreased intracranial pressureAndrew Zura

The use of succinylcholine in a patient with increasedintracranial pressure is still a controversial issue. Inthis case we highlight the benefits and the relativecontraindications for using succinylcholine in thesepatients.

Case descriptionThe patient was a 79-year-old female who fell andhit her head on the floor, after leaning back in herchair while eating dinner. Her mental status had pro-gressively deteriorated since that time and she wasbrought to the emergency department (ED) by ambu-lance approximately 1 hour after the event.

Her past medical history included hypertensionand atrial fibrillation, for which she was on coumadin.Incidentally, her international normalized ratio waschecked by her primary care physician the day priorto the incident and found to be 2.2. On physical exam-ination she appeared stuporous and did not answerany questions, but she reacted to painful stimuli.Vital signs were heart rate 42 beats per minute, bloodpressure 200/110 mmHg, respirations 30 breaths perminute and temperature 37.0 ◦C. A computed tomog-raphy scan in the ED showed a massive subduralhematoma with a 2-cm midline shift. The anesthesi-ology team had been called to the ED to emergentlyintubate this patient prior to taking her to theoperating room.The endotracheal intubation was per-formed using sodium pentothal and succinylcholine.After end-tidal CO2 and bilateral breath sounds wereconfirmed, the resident asked the anesthesiologyattending whether another agent such as rocuroniumshould have been used due to the patient’s increasedintracranial pressure (ICP).

DiscussionThe question “Does succinylcholine increase intracra-nial pressure in patients with neurologic injury?” hasconcerned clinicians since it was first described in

the 1950s [1]. When researchers tried to replicate theinitial study they found that they obtained differentresults depending on the species studied and anes-thetic technique used. Animal studies in cats and dogshave demonstrated reliable increases in ICP with suc-cinylcholine [2, 3], but a study in monkeys failed todemonstrate this effect [4].

Two studies have demonstrated that succinyl-choline will increase ICP in humans and that thiseffect can be attenuated by a “defasciculating” dose ofa nondepolarizing muscle relaxant [5, 6]. More recentstudies, also done in human patients with neurologicinjury, have failed to demonstrate the increase in ICP.They concluded that succinylcholine in itself does notincrease ICP or cerebral blood flow velocity [7, 8]. It isimportant to note that they also mentioned that otherfactors such as light anesthesia and tracheal intubation,which are often found together with the administra-tion of succinylcholine, can and do significantly con-tribute to increases in ICP in patients with neurologicinjury [7].

ConclusionAt this time, the data remain equivocal. There hasnever been a randomized controlled trial conducted toanswer this question conclusively. Until more defini-tive recommendations are made, the clinician mustcontinue to be aware of the risks and benefits of thedrugs administered when deciding how to best treatthe patient.

References1. M. Halldin, A.Wahlin. Effect of succinylcholine on

the intraspinal fluid pressure.Acta Anaesthesiol Scand1959; 3: 155–61.

2. J. E. Cottrell, J. Hartung, J. P. Giffin et al. Intracranialand hemodynamic changes after succinylcholineadministration in cats. Anesth Analg 1983; 62: 1006–9.

309


3. W. L. Lanier, P. A. Iaizzo, J. H. Milde. Cerebralfunction and muscle afferent activity followingintravenous succinylcholine in dogs anesthetized withhalothane: the effects of pretreatment with adefasciculating dose of pancuronium. Anesthesiology1989; 71: 87–95.

4. J. D. Haigh, E. M. Nemoto, A. M. DeWolf et al.Comparison of the effects of succinylcholine andatracurium on intracranial pressure in monkeys withintracranial hypertension. Can Anaesth Soc J 1986; 33:421–6.

5. M. D. Minton, K. Grosslight, J. A. Stirt et al. Increasesin intracranial pressure from succinylcholine:prevention by prior nondepolarizing blockade.Anesthesiology 1986; 65: 165–9.

6. J. A. Stirt, K. R. Grosslight, R. F. Bedford et al.“Defasciculation” with metocurine preventssuccinylcholine-induced increases in intracranialpressure. Anesthesiology 1987; 67: 50–3.

7. W. D. Kovarik, T. S. Mayberg, A. M. Lam et al.Succinylcholine does not change intracranialpressure, cerebral blood flow velocity, or theelectroencephalogram in patients withneurologic injury. Anesth Analg 1994; 78:469–73.

8. M.M. Brown,M. J. Parr, A. R. Manara. Theeffect of suxamethonium on intracranial pressureand cerebral perfusion pressure in patients withsevere head injuries following blunt trauma. Eur JAnaesth 1996; 13: 474–7.

310

Part XIII SeizuresCase

92 Pharmacologic management ofstatus epilepticusAlaa A. Abd-Elsayed, George K. Istaphanous and Ehab Farag

The management of status epilepticus is both chal-lenging and costly. The annual incidence is estimatedto be about 78 per 100 000 population in the USA[1]. The designation of status epilepticus may varyamongst clinicians, but it generally describes a clus-ter of frequent clinical seizures or a single unremit-ting seizure that lasts longer than 5–10 minuteswithout return to baseline or interictal state [2]. His-torically, the International League against Epilepsy in1981 defined status epilepticus as a seizure that per-sists for a sufficient length of time or is repeated fre-quently without recovery between attacks [3]. Theduration of what is accepted as status epilepticus hasvaried according to different classifications. It has beenchanged from 30 min to 20 min, according to theguidelines of the Epilepsy Foundation of America’sWorking Group on Status Epilepticus. The VeteransAffairs Status Epilepticus Cooperation Study stipu-lated 10 min and, most recently, the duration of only5 min has been proposed. Most seizures cease within aminute or two and if the seizure is prolonged beyond afew minutes, it is unlikely to be self-remitting [4]. Thelack of a specific duration of the seizures has made thisdefinition difficult to use.

Status epilepticus historically has been catego-rized into two main groups, convulsive and non-convulsive. Convulsive status epilepticus can be fur-ther classified into (a) tonic–clonic status epilepticus,(b) tonic status epilepticus, (c) clonic status epilep-ticus, and (d) myoclonic status epilepticus. Noncon-vulsive status epilepticus refers to a continuous ornear-continuous generalized electrical seizure activ-ity lasting for at least 30 min, but without convul-sions (see Case 93). Nonconvulsive status epilepti-cus can be accompanied by abnormal mental status,unresponsiveness, ocular motor abnormalities, andpersistent subclinical seizures. Nonconvulsive statusepilepticus has long been divided into two main cat-egories: absence status epilepticus and complex par-

tial status epilepticus. Here we present the pharma-cologic management of a case of convulsive statusepilepticus.

Case descriptionA 12-year-old right-handedmale with a history of cor-tical dysplasia, 3–4 tonic–clonic seizures a week andlanguage/cognitive regression presented for placementof seizure-focus-localization grids via craniotomy.After cessation of antiepileptic drugs and a success-ful procedure, he was admitted to the intensive careunit (ICU) for monitoring. The patient experienced atonic–clonic seizure on postoperative day 2 that con-tinued despite administration of two doses of intra-venous (IV) lorazepam 0.2 mg/kg (10 minutes apart)and fosphenytoin 20 mg/kg. On examination thepatient was found to be unresponsive with increasedmuscle tone. Laboratory workup was grossly normalwith nometabolic aberrations. Following intubation ofthe airway, pentobarbital 5 mg/kg IV bolus was givenand then infused at a rate of 1 mg/kg/hour (generalanesthesia) until burst-suppression was achieved, asassessed by electroencephalogram (EEG).

Computed tomography (CT) scans did not revealany intracranial hemorrhage, masses, or midline shift.The patient was monitored by a continuous EEG toassist with pharmacologic management. The patienthad a resection of the seizure focus 1 week later andhewas discharged home in stable neurologic condition1 week after the procedure.

DiscussionStatus epilepticus is a life-threatening conditionthat requires immediate identification and treat-ment. Common causes of seizures include, butare not limited to, discontinuation of antiepilepticdrugs, acute or longstanding brain injury, hemor-rhage or infections, alcohol or drug withdrawals

311

XIII. Seizures

(benzodiazepines, barbiturates, baclofen), metabolicabnormalities (hypo/hyperglycemia, uremia, hypona-tremia, hypocalcemia, hypomagnesemia) or drugoverdose (penicillin G, theophylline, flumazenil,lidocaine, bupivacaine). Publication of the VeteransAffairs Cooperative Trial in 1998 [5] and the SanFrancisco Emergency Medical Services Study in 2001[6] allowed for an evidence-based approach to thechoice of the first-line agent to be used in terminatingstatus epilepticus; current evidence indicates thatlorazepam administered intravenously in out-of-hospital settings by paramedics or in emergencyroom settings is superior to diazepam and phenytoinin abolishing status epilepticus in a short period oftime.

Fosphenytoin is second-line therapy in statusepilepticus and offers several advantages over pheny-toin. First, it can be infused with different standardintravenous solutions, whereas phenytoin can only bemixed with non-dextrose fluids. Fosphenytoin can beinfused at a faster rate with decreased risk of cardiacarrhythmia and hypotension, especially in the elderly,and may be given intramuscularly.

Intravenous valproate is another therapeuticoption for those patients with cardiorespiratoryimpairment and myoclonic status epilepticus. It isa nonsedative drug with an attractive safety profileand it is easy to use. Levetiracetam is a relatively newantiepileptic drug that has a potentially important rolein the management of refractory status epilepticus.It is available as an intravenous preparation that canbe infused rapidly without the necessity of ventilatorassistance.

Topiramate is also another new antiepileptic drugwith a mechanism of action that is different from thatof the benzodiazepines and other first-line antiepilep-tic drugs. It can be administered by nasogastric tube indoses ranging from 300–1600 mg/day for the manage-ment of refractory status epilepticus.

Patientswith refractory status epilepticuswhohavenot responded to the first-line treatment will requireadmission to the ICU for more aggressive manage-ment. Tracheal intubation and controlled mechanicalventilationwith the use ofmuscle relaxants are the firststeps in intensive caremanagement of refractory statusepilepticus. It is necessary in this setting to use con-tinuous EEG monitoring in order to guide infusionsof phenobarbital, midazolam, propofol, or pentobar-bital; arterial and central venous access may also be

required. The EEG endpoint of therapy is the burst-suppression pattern, which needs to be continued for12 hours after the last seizure. Infusion of the sedativeagent can be reduced every 3 hours with EEGmonitor-ing and if there is no EEG evidence of seizures, then thepatient can be weaned off the ventilator [7].

Phenobarbital is found to be effective in patientswho have failed lorazepam or fosphenytoin, withor without valproate. Propofol is an alkyl phenolwith a global central nervous system depressanteffect. It directly activates the gamma-aminobutyricacid (GABA) receptors [8], modulates calcium influxthrough slow calcium ion channels, and has antiox-idant activity. The pharmacokinetics and favorableadverse effect profile make propofol an excellent drugto treat refractory status epilepticus. The two mainadvantages of propofol are a rapid onset and shortduration of action. Propofol is a highly lipophilic agentwith a large volume of distribution that leads to itsrapid uptake and elimination from the central nervoussystem. The main disadvantage of prolonged propo-fol infusion in the ICU is the development of propo-fol infusion syndrome. The main presentation of thissyndrome is unexplained metabolic acidosis.

Midazolam is a fast-acting, water-soluble benzo-diazepine with a half life of 4–6 hours. It acts bybinding to GABA-A receptors. Midazolam is an alter-native to propofol. It is typically started after secur-ing endotracheal intubation and ventilator assistance.Inhalational anesthetics offer an alternative approachto the treatment of refractory status epilepticus. Isoflu-rane and desflurane are the two agents that have beentried in the treatment of refractory status epilepticusbecause of the safety associated with their long-termadministration.

Surgical intervention can be considered as a lastresort in patients who failed medical treatment andhave a lesion amenable to surgery. Managementof the underlying causes of status epilepticus suchas noncompliance with antiepileptic drug therapy,acute infections, high fever, hypoglycemia, electrolyteimbalance, organ dysfunction, drug intoxication, poi-soning, alcohol withdrawal, excess use of alcohol,stroke, trauma, and hypertensive encephalopathy isessential. Rhabdomyolysis can be a consequence ofconvulsive status epilepticus; the clinician should alsobe vigilant for the development of pneumonia inpatients who are maintained in burst suppression forprolonged periods of time.

312

Case 92. Pharmacologic management of status epilepticus

ConclusionThemanagement of status epilepticus is both challeng-ing and costly. Prognosis depends both on the under-lying condition and expeditious treatment. Thereforea thorough neurologic examination is critical, alongwith EEG to guide the management.

There are different strategies to treat status epilep-ticus, such as (1) lorazepam 0.02–0.03 mg/kg intra-venously repeated as needed up to 0.1 mg/kg as a first-line agent, followed by (2) fosphenytoin or phenytoin20mg/kg, then (3) propofol or phenobarbital infusionsto achieve burst suppression on the EEG.

References1. R. J. DeLorenzo, J. M. Pellock, A. R. Towne et al.

Epidemiology of status epilepticus. J Clin Neurophysiol1995; 12: 316.

2. H. Gastaut. A propos d’ une classificationsymptomatologique des etats de mal epileptiques. InGastaut H, Roger J., Lob H., eds. Les etats de malepileptiques. Paris: Masson, 1967; 1–8.

3. Proposal for revised clinical andelectroencephalographic classification of epileptic

seizures: from the Commission on Classification andTerminology of the International League againstEpilepsy. Epilepsia 1981; 22: 489–501.

4. W. H.Theodore, R. J. Porter, P. Albert et al. Thesecondarily generalized tonic-clonic seizure: avideotape analysis. Neurology 1994; 44:1403–7.

5. D. M. Treiman, P. D. Meyers, N. Y. Walton. Acomparison of four treatments for generalizedconvulsive status epilepticus: Veterans Affairs StatusEpilepticus Cooperative Study Group. N Engl J Med1998; 339: 792–8.

6. B. K. Alldredge, A. M. Gelb, S. M. Isaacs et al. Acomparison of lorazepam, diazepam, and placebo forthe treatment of out-of-hospital status epilepticus. NEngl J Med 2001; 345: 631–7.

7. Indian Epilepsy Society. Indian Guidelines for theManagement of Epilepsy: GEMIND. Available from:http://www.epilepsyindia.org/gemind-main.asp [citedin 2008].

8. M. Hara, Y. Kai, Y. Ikemoto. Propofol activatesGABA-A receptor-chloride ionophore complex indissociated hippocampal pyramidal neurons of the rat.Anesthesiology 1993; 79: 781–8.

313


93 Nonconvulsive status epilepticusSheron Beltran

Nonconvulsive status epilepticus (NCSE) is an under-recognized cause of mental status decline. As impliedby the name, it represents prolonged seizure activityor repeated seizures without a return to baseline men-tal status, in the absence of the classic motor symp-toms commonly associatedwith seizure activity. Itmaybe the sole cause of altered cognition or may be cou-pledwith an underlying central nervous system abnor-mality resulting in seizure activity. The dysregulationof excitatory central nervous system (CNS) pathwaysresults in suppression or impairment of normal con-sciousness. Diagnosis can prove difficult, and may noteven be considered, as many clinicians are unaware ofthe existence of NCSE. Given the wide spectrum ofclinical presentations, a high index of suspicion mustbe maintained for prompt diagnosis and treatment.

Case descriptionA 55-year-old male with a history of shunted hydro-cephalus, seizure disorder, and hypothyroidismpresented to the emergency department with com-plaints of nausea, vomiting, headache, and confusionpersisting for several days. He had been seizure-freefor several years on levetiracetam and lamotrigine.Over the preceding days he had been disoriented tothe point of being lost a few blocks from his homeand, according to family members, he had difficultiescommunicating with them on several occasions. Onneurologic examination he was alert and oriented onlyto person and place. His speech was intact, however,he was found to be repetitive and to perseverate duringconversation. On cranial nerve exam he had somedifficulty with upgaze. Imaging revealed a dysmorphicventricular system with enlarged fourth, third, andlateral ventricles (Figure 93.1). Additionally, therewas a space-occupying mass deemed likely to be anarachnoid cyst.

The patient was then taken to the operating roomfor an uneventful revision of a right frontal ventricu-loperitoneal shunt. During the immediate postopera-

Figure 93.1. Enlarged and dysmorphic ventricular system.

tive period, the patient was alert and conversant. Hehad no focal neurologic findings and seemed to beimproving. However, during an overnight neurologicexamination he was noted to be hemiparetic and unre-sponsive. Imaging revealed a large frontoparietal sub-dural hematoma along with intraventricular hemor-rhage. The hematoma was subsequently evacuated inthe operating room.

Over the next 2 days the patient remained stablyhemiparetic, but his mental status waxed and waned.He continued to have periods of disorientation andspeech disturbances. He had difficulty following direc-tions and was unable to complete multistep tasks. Hisrepeat imaging studies were unchanged, he was hemo-dynamically stable without electrolyte abnormalities,and there was no witnessed seizure activity. Hecontinued on his pre-admission anticonvulsant agents;however, the possibility of NCSE was entertained.An electroencephalogram (EEG) was performedshowing diffuse slowing of activity suggestive of

314

Case 93. Nonconvulsive status epilepticus

encephalopathy with intermittent rhythmic activityrepresenting possible seizures. The patient was placedon continuous EEG monitoring and lorazepam wasadministered. Fosphenytoin was also added to hisantiseizure regimen. The seizure activity subsequentlystopped after the addition of fosphenytoin and thepatient’s neurologic status stabilized.

By postoperative day 6, the patient remained hemi-paretic but had full return of his cognitive status tobaseline. On postoperative day 7, he had a generalizedtonic–clonic seizure and was treated with lorazepamand a loading dose of fosphenytoin, as his free pheny-toin level was subtherapeutic. His seizure activity wasrapidly terminated after administration of the benzo-diazepine. The patient was thereafter seizure-free andwas able to participate in rehabilitation.

DiscussionWhen seizure activity presents with classic motorsymptoms and loss of consciousness, there is often lit-tle debate as to the diagnosis. In this case, however,seizure activity presented subtly with waxing and wan-ing of mental status. Diagnosing NCSE may be elusivegiven the variety of possible presenting features. Non-convulsive status epilepticusmay present with nonspe-cific neurologic deficits such as impaired attention, dif-ficulties with complex tasks, or simply as a headache[1]. Nonconvulsive status epilepticus may also be pre-ceded by motor activity with prolonged postictal con-fusion.At the opposite end of the spectrum,NCSEmaybe the cause of coma [2]. The diagnostic criteria forNCSE are not universally accepted; it has been sug-gested that an altered state of consciousness coupledwith EEG evidence of seizure activity are sufficient fordiagnosis, however, some believe that a response toantiepileptic agents is necessary to prove the diagnosis.

There is no universally accepted classificationscheme for NCSE, although it may be subdividedbased on clinical features or EEG characteristics. Elec-troencephalogram patterns may be generalized, as inabsence or petit mal status epilepticus, or focal asin complex partial status epilepticus (CPSE). Focalseizures may progress to secondary generalization,further complicating the classification system.

It is difficult to determine whether NCSE, in and ofitself, causes permanent neurologic deficits. Absenceseizures are generally considered benign, whereasCPSE may lead to prolonged impairment. Addition-ally, seizure activity is often a symptom of underlying

CNS disease, and ongoing seizures may lead to sec-ondary injury and prolonged cognitive deficits. Thus,NCSE is an important entity to recognize and treat.Nonconvulsive status epilepticus often responds to ini-tial intervention strategies, but it may prove refractoryand the benefits of treating a condition in which theoutcome is controversial must be weighed against therisk of more aggressive treatment.

Initial measures should include management ofunderlying CNS disorders and metabolic derange-ments, as well as administration of anticonvulsantagents. Continuous EEGmonitoring during treatmentallows for ongoing evaluation of drug administration.Benzodiazepines are considered the first-line treat-ment for NCSE and may lead to a rapid terminationof seizure activity. However, resistance to benzodi-azepines and barbiturates may occur in patients withprolonged status epilepticus necessitating secondarytreatment. Care must be taken as benzodiazepinetreatment may result in respiratory depression.

Antiepileptic agents such as phenytoin, valproate,and levetiracetam have all been used in the treatmentof NCSE. Intravenous phenytoin has been associatedwith hypotension, respiratory depression, and cardiacarrhythmias. Fosphenytoin is a water-soluble prodrugthat is better tolerated than phenytoin. Valproate hasbeen shown to be more effective than phenytoin incontrolling seizures in generalized convulsive statusepilepticus with no serious side effects. Case reportshave suggested that valproate is also effective in the set-ting of NCSE; however, it has also been implicated asa cause of NCSE [3]. Levetiracetam has been reportedto be effective in a small case series without significantside effects [4].

Ketamine, a glutamate receptor (NMDA) antag-onist, has been shown to be effective at terminat-ing seizure activity in animals, particularly in refrac-tory cases. This may be explained by evidence thatas a seizure progresses to status, receptor traffickingoccurs as gamma-aminobutyric acid (GABA) recep-tors become internalized and NMDA receptors arerecruited to the synaptic membrane. Thus, agentsacting on GABA receptors may become ineffectivewhile NMDA antagonists may terminate the seizureactivity [5].

General anesthesia has been used in refractorycases of NCSE; however, hemodynamic support maybe required to maintain adequate perfusion pres-sure. Performing general anesthesia in an intensivecare setting is often logistically difficult requiring

315

XIII. Seizures

specialized equipment and personnel; mortality maybe significant.

ConclusionIn our case, NCSE presented subtly with waxing andwaning of mental status in a patient with under-lying central nervous system pathology (subduralhematoma) and a known seizure disorder. A highlevel of suspicion for NCSE prompted evaluation andan early diagnosis was made. Treatment with ben-zodiazepines and fosphenytoin was initiated whilemonitoring continuous EEG activity to evaluate thepatient’s response to therapy. The patient’s neurologicstatus improved and EEG evidence of seizure activ-ity resolved possibly sparing the patient further pro-longed cognitive deficits.

References1. M. Ghofrani, F. Mahvelati,H. Tonekaboni et al.

Headache as a sole manifestation in nonconvulsivestatus epilepticus. J Child Neurol 2007 May; 22: 660–2.

2. A. R. Towne, E. J. Waterhouse, J. G. Boggs et al.Prevalence of nonconvulsive status epilepticus incomatose patients. Neurology 2000; 54: 340–5.

3. S. K. Velioglu, S. Gazioglu. Non-convulsive statusepilepticus secondary to valproic acid-inducedhyperammonemic encephalopathy. Acta Neurol Scand2007; 116: 128–32.

4. S. Rupprecht, K. Franke, S. Fitzek et al. Levetiracetamas a treatment option in non-convulsive statusepilepticus. Epilepsy Res 2007; 73: 238–44.

5. D. J. Borris, E. H. Bertram, J. Kapur. Ketaminecontrols prolonged status epilepticus. Epilepsy Res2000; 42: 117–22.

316


94 RhabdomyolysisMauricio Perilla and Jerome O’Hara

Although not a disease of the nervous system itself,rhabdomyolysis can significantly complicate the careof patients presenting with neurologic disorders. Earlydiagnosis and aggressive management are key toavoiding major adverse events such as renal failure.

Case descriptionA 53-year-old patient was scheduled for L2–S1 spinerevision and new instrumentation. The patient wasadmitted the day before surgery after being foundunconscious on the floor at home. At hospitaladmission, the patient was sleepy but arousable andcomplained of severe back pain. Magnetic resonanceimaging of the spine showed L3–L4 compression fac-ture with spinal stenosis. The patient had a historyof seizure disorder, postlaminectomy syndrome, dia-betes, morbid obesity, and drug abuse. After 8 hoursof uneventful surgery in prone position, the tra-chea was extubated and the patient was transferredto the postanesthesia care unit in stable condition.On postoperative day 1, the patient reported gen-eralized muscle pain (chest and back), nausea, andvomiting. The patient was hypotensive, tachycardic,dehydrated, oliguric, and “dark tea-colored” urine wasnoted.

The patient’s initial differential diagnosis includedbut was not restricted to: hypovolemia, sepsis, rhab-domyolysis, acute coronary syndrome, allergic reac-tion, hemolytic reaction, acute renal failure, dissem-inated intravascular coagulation, thromboembolism,diabetic ketoacidosis, hyperosmolar hyperglycemia,nonketotic coma, drug withdrawal/toxicity, hyperna-tremia, hypokalemia, hypophosphatemia, myoglobin-uria, sickle cell crisis.

The following abnormal laboratory values werefound: urea 98 mg/dL, creatinine 3.1 mg/dL, potas-sium 6.2 mmol/L and lactic acid 4.1 mmol/L. Creatinekinase (CK) was reported to be 16 000 U/L (normal30–220 U/L), MB isoenzyme �0.1%. Urinalysis was

positive for hemoglobin but no red blood cells. Elec-trocardiogram showed a sinus tachycardia and a chestX-ray showed left basilar atelectasis.

DiscussionRhabdomyolysis is a potentially life-threatening clini-cal syndrome caused by damage of the muscle fibers,which results in necrosis and disintegration of thestriated muscle. The initial injury triggers the releaseof muscular cell constituents (electrolytes, myoglobin,and creatine kinase) into the extracellular fluid and thebloodstream. Extracellular calcium activates sustainedcontraction of myofibers with increase of adenosinetriphosphate consumption and further depletion.

The escalation of proteases and phospholipaseactivity and the discharge of vasoactivemolecules withproduction of free radicals magnify the damage [1].Cell leakage increases the interstitial fluid with therise of compartmental pressures, producing arterial,venous occlusion, and nerve compression (compart-ment syndrome). Massive release of myoglobin to thecirculation is filtered through the glomerular base-ment membrane. As water is reabsorbed the concen-tration ofmyoglobin rises promoting precipitation andcast formation. Tubular flowdeclineswith dehydrationand renal vasoconstriction. Degradation of intratubu-lar myoglobin results in the release of free iron, whichcatalyzes free radical production aggravating ischemicinjury. Uric acid contributes to tubular obstruction bydecreasing urine pH and favoring myoglobin precipi-tation and uric acid casts formation [2].

Rhabdomyolysis is most common in situationsof extreme muscle activity like seizures, trauma,crush injuries, prolonged lying down on the ground(patient unconscious for several hours), burns, drugs,toxins, hyperthermia, infections, and inflammatoryor ischemic processes. Metabolic, autoimmune, andgenetic causes also have been identified [3]. Some clin-ical features include fever, myalgia, stiffness, weakness,

317

XIII. Seizures

nausea, vomiting, confusion, delirium, and coma. Thephysical examination should focus on areas of mus-cle trauma, limb weakness, or just myalgias. Hypoten-sion and dehydration are common due to extensivecellular lysis. Clinical signs and symptoms might notbe evident and mild forms can be underdiagnosed.Acute renal failure is a common complication thatmight be prevented with an early diagnosis. Othercomplications include cardiac arrhythmias and cardiacarrest due to hyperkalemia, hepatic dysfunction, anddiffuse intravascular coagulation. Compartment syn-drome can both be a cause or a consequence of rhab-domyolysis. In this case the patient was lying on thefloor unconscious for several hours; his further historyof seizures suggested an etiology of the initial fall anda risk factor for rhabdomyolysis.

DiagnosisMyoglobin in plasma: After an extensive release, myo-globin overcomes plasma globulin’s binding capacityand is filtrated to the renal tubules. Visible myoglobin-uria is seen when its excretion exceeds 250 mcg/ml.The short life of myoglobin in plasma (3–6 hours) lim-its its diagnostic capacity.

Creatine kinase (CK): Increased levels five timesnormal are strongly suggestive of rhabdomyolysis.Creatine kinase levels should be tested after any evi-dence of extensive muscle injury. Creatine kinase lev-els peak in 24–36 hours [4] and can reach levels upto 100 000 U/L. Sustained levels are associated withcompartment syndrome and relate to the severity ofthe muscle damage. Typical isoenzymes from skele-tal muscle CK-MM can be differentiated from cardiacenzymes CK-MB.

Urine dipstick: This test characteristically showsa large amount of blood but the orthotoluidine por-tion of the dipstick is unable to distinguishmyoglobin-uria from hematuria or hemoglobinuria. Urine sedi-ment may show minimal or total absence of red bloodcells. Casts are often present and immunochemistry isrequired for myoglobin identification.

Electrolyte abnormalities: Hyperkalemia, hypocal-cemia, hyperphosphatemia, hypoalbuminemia, andhyperuricemia are frequently present.

Toxicology screen: Ethanol, cocaine, and otherillicit drug utilization have been associated with thissyndrome and a toxicology screen is recommendedwith the initial assessment.

ManagementThe treatment of this condition should be to (1) rapidlyrecognize the initial trigger, (2) restore intravascu-lar volume, (3) prevent acute renal failure, and(4) monitor the development of common compli-cations. Patients should be treated in an intensivecare unit where fluid balance can be continuouslyevaluated. Early fluid resuscitation includes initialadministration of high rates of normal saline (1.5 litersper hour) in order to maintain urine output around300 mL/hour [5]. Invasive hemodynamic monitoringmay be required in patients with limited cardiovas-cular reserve or previous renal impairment. Urinealkalinization prevents myoglobin protein bindingand cast precipitation; a urine pH �6.5 promotesmyoglobin cast elimination. Mannitol is an osmoticdiuretic that produces renal vasodilation and possiblyacts as a free radical scavenger. Mannitol increasesurine flow andmay prevent tubular obstruction; its usein rhabdomyolysis is still controversial. Loop diuret-ics should not be used because they acidify urine.Renal replacement therapy should be considered inpatients who develop oliguric acute tubular necrosis.Hyperkalemia is a common complication that shouldbe treated promptly due to its fatal arrhythmogeniceffect when it is associated with hypocalcemia.Hypocalcemia should be corrected in symptomaticpatients; hypercalcemia is commonly present dur-ing the recovery phase of rhabdomyolysis. Muscleswelling and pain should be regularly evaluated andfasciotomies may be needed in case of compartmentsyndrome. Succinylcholine is contraindicated due toa life-threatening potassium elevation. Disseminatedintravascular coagulation and liver failure should besuspected in rhabdomyolysis cases.

ConclusionRhabdomyolysis can be associated with life-threatening complications in neurologic patients.Rapid diagnosis, aggressive hydration, and correc-tion of electrolyte abnormalities are key features ofmanagement.

References1. F. G. O’Connor, P. A. Deuster. Rhabdomyolysis. In

Goldman L., Ausiello D. A., eds. Cecil Medicine,23rd edition. Philadelphia, PA: Saunders Elsevier,2008; 798–802.

318

Case 94. Rhabdomyolysis

2. R. Vanholder,M. S. Sever, E. Erek et al.Rhabdomyolysis. J Am Soc Nephrol 2000; 11: 1553–61.

3. J. D.Warren, P. C. Blumbergs, P. D.Thompson. Rhab-domyolysis: a review.Muscle Nerve 2002; 25: 332–47.

4. W. Tan, B. C. Herzlich, R. Funaro et al.Rhabdomyolysis and myoglobinuric acute renal

failure associated with classic heat stroke. South Med J1995; 88: 1065–8.

5. M. S. Slater, R. J. Mullins. Rhabdomyolysis andmyoglobinuric renal failure in trauma and surgicalpatients: a review. J Am Coll Surg 1998; 186:693–716.

319

Part XIV Neuromuscular diseaseCase

95 Myasthenic crisisWael Ali Sakr Esa

Myasthenia gravis (MG) is an autoimmune diseasewith antibodies directed against the nicotinic acetyl-choline receptor or other muscle membrane proteins.Myasthenic crisis is usually associated with infectionand is characterized by respiratory failure requiringinvasive or noninvasive mechanical ventilation [1].

Case descriptionA 46-year-old African–American female with a 9-yearhistory of MG developed a progressive bilateral lowerextremity numbness and pain from a bulging lumbardisc. The patient was admitted for a two level lumbardiscectomy. Her daily medications were pyridostig-mine 60 mg every 8 hours, levothyroxine 175 mcgdaily, esmoperazole 20 mg daily. Her past medi-cal history included hypothyroidism and gastroe-sophageal reflux disease. She had onemyasthenic crisis5 years ago associated with a pulmonary infection thatrequired intubation and mechanical ventilation.

In the operating room, standard ASA monitorswere placed on the patient and a left radial arteriallinewas placed. Remifentanil infusionwas then startedat 0.05 mcg/kg/minute. Anesthesia was induced withpropofol and lidocaine, then an endotracheal tubewas placed without the use of any muscle relaxant.Anesthesia was maintained with inhalational anesthe-sia, hydromorphone 1 mg and remifentanil infusion0.05–0.3mcg/kg/minute. Nomuscle relaxant was usedduring thewhole 2-hour surgery.At the endof the case,inhalational anesthesia and remifentanil infusion werestopped but the patient remained weak and was notachieving adequate tidal volume.

The patient was transferred to the intensive careunit (ICU), intubated, and sedated with a propofolinfusion. A 10 mg intravenous dose of edrophoniumwas given and 60mg pyridostigminewas administeredthrough the nasogastric tube with partial improve-ment in motor power and tidal volume. Plasmaphere-sis was performed and then the patient was weaned

successfully from the mechanical ventilation after sheregained her motor strength on the second day afterthe surgery.

DiscussionThe clinical hallmark of MG is skeletal muscleweakness. Eighty-five percent of patients with MGhave identifiable anti-acetylcholine receptor antibod-ies. Anti-acetylcholine receptor antibodies damage thepostsynaptic muscle membrane via a complement-mediated reaction, causing an increased degradationand decreased formation of acetylcholine receptors.Many myasthenic patients also have antinuclear andantithyroid antibodies and other autoimmune diseasessuch as systemic lupus erythematosus, rheumatoidarthritis, pernicious anemia, and thyroiditis. Myas-thenic crisis is defined as respiratory failure requiringmechanical ventilation in the MG patient. Myastheniccrisis is a common life-threatening complication thatoccurs in approximately 15–20% of patients with MGduring their lifetime [2].

Myasthenic crises are often precipitated by pul-monary infections and result in respiratory failurerequiring mechanical ventilation. Potential cardiacmanifestations of MG include focal myocarditis, atrialfibrillation, atrioventricular conduction delay, and leftventricular diastolic dysfunction. Many conditionssuch as viral infections, stress, surgery, and extremeheat may exacerbate the symptoms of myasthenia.TheOsserman staging system is based on the severity of thedisease. Type I: Ocular signs and symptoms only; TypeII A: generalized muscle weakness; Type II B: gener-alized moderate weakness and/or bulbar dysfunction;Type III: acute fulminant presentation and/or respi-ratory dysfunction; and Type IV: severe, generalizedmyasthenia [3].

The primary concern in anesthesia for the patientwith myasthenia gravis is the use of muscle relaxants.Patients tend to be resistant to succinylcholine and

321

XIV. Neuromuscular disease

Table 95.1. Comparison of Myasthenic syndrome and myasthenia gravis.

Myasthenic syndrome Myasthenia gravis

Manifestations Proximal limb weakness (arms �legs)Exercise improves strengthReflexes absent or decreasedMyalgia common

Extraocular, bulbar, and facial muscle weaknessFatigue with exerciseReflexes normalNo myalgia

Response to muscle relaxants Sensitive to succinylcholine and nondepolarizingmuscle relaxants

No improvement with acetylcholinesteraseinhibitors

Resistant to succinylcholine. Sensitive tonondepolarizing muscle relaxants

Improvement with acetylcholinesteraseinhibitors

Gender Male � female Female �male

Pathogenesis IgG autoantibodies directed against prejunctionalvoltage-sensitive calcium channels

Associated with small cell carcinoma of the lung,sarcoidosis

Destruction of postsynaptic acetylcholinereceptors, usually by autoimmune antibodies

Associated with thymoma

Treatment Aminopyridines that increase presynaptic releaseof acetylcholinesterase

Acetylcholinesterase inhibitors, steroids,thymectomy, immunosuppressants

have increased sensitivity to nondepolarizing relax-ants, necessitating caution with their use. As in ourcase, an anesthetic technique that eliminated the useof a muscle relaxant is preferred. Conditions suitablefor tracheal intubation can be obtainedwith a standardinduction technique using propofol and the inhala-tion of volatile halogenated agents such as isofluraneor sevoflurane with the addition of an opioid infu-sion such as remifentanil. Sevoflurane, isoflurane, anddesflurane depress neuromuscular transmission andmay provide enoughmuscle relaxation so that trachealintubation can be performed without neuromuscularblocking drugs after an adequate depth of anesthesiahas been attained. Adjuvant drugs that blunt responsesto laryngoscopy such as propofol, opioids, and lido-caine may be useful as seen in our case.

In this case, remifentanil infusion was maintainedduring the surgery to avoid the use of any musclerelaxants. Acetylcholinesterase inhibitors such as pyri-dostigmine and edrophonium enhance neuromuscu-lar function by preventing synaptic degradation ofacetylcholine and thereby increasing the amount ofacetylcholine available in the synapse. Excess pyri-dostigmine may cause a cholinergic crisis, in whichweakness and muscarinic signs (bradycardia, miosis,salivation, diarrhea) also develop from excess ace-tycholine. Conversely, too little pyridostigmine maycause a myasthenic crisis, in which weakness alsoresults from too little acetylcholine. To distinguisha cholinergic crisis from a myasthenic crisis, mus-carinic signs that are consistent with a cholinergic cri-sis should be considered. In addition, the adminis-

tration of 10 mg edrophonium intravenously resultsin persistent weakness if the patient is experiencing acholinergic crisis, but improvement in strength if thepatient is experiencing a myasthenic crisis. The otherimportant diagnostic point is to differentiate betweenmyasthenic syndrome andMG as shown in Table 95.1.

In the approach to the patient with myastheniccrisis, (1) the diagnosis of MG should be confirmed;(2) respiratory failure should be evaluated and treatedin the ICU, while potential precipitating factorsare identified and managed; (3) immunomodulatorytreatment should be initiated; and (4) complicationsshould be avoided or managed promptly. The gen-eral criteria for intubation are vital capacity of �15–20 mL/kg, peak inspiratory pressure �−40 cmH2O,peak expiratory force of �40 cmH2O, and evidenceof respiratory muscle fatigue, hypercapnia, or hypoxia[4]. Plasma exchangemay bemore effective than intra-venous immunoglobulin in the treatment of myas-thenic crisis involving respiratory failure. In the acutesetting, the role of immunosuppression and intra-venous/intramuscular pyridostigmine and the neweragents such as tacrolimus remains limited and at timescontroversial [5]. Since the advent of these immuneinterventions and improved methods of mechanicalventilation in specialized ICUs, mortality rates in MGand myasthenic crisis have substantially decreased.Also, patients in myasthenic crisis should undergocareful cardiacmonitoring and equipment for externalcardiac pacing or the insertion of a temporary pace-maker should be provided at all times for the criticallyill patient.

322

Case 95. Myasthenic crisis

ConclusionIn conclusion, the recent improvements in prognosisfor a patient withmyasthenic crisis can be attributed toimproved respiratory care during the crisis, admissionto an ICU, substitution of endotracheal intubation fortracheostomy, and improvements in monitoring, tra-cheal toilet, and antibiotics.

References1. J. Palace, A. Vincent,D. Becson. Myasthenia gravis:

diagnostic and management dilemmas. Curr OpinNeurol 2001; 14: 583.

2. A. Alshekhlee, J. D. Miles, B. Katirji et al. Incidenceand mortality rates of myasthenia gravis andmyasthenic crisis in US hospitals. Neurology 2009; 72:1548–54.

3. S. F. Dierdorf, J. S. Walton. Anesthesia for patientswith rare and coexisting disease. In Barash P. G.,Cullen B. F., Stoelting R. K., eds. Clinical Anesthesia,5th edition. Philadelphia, PA: Lippincott Williams &Wilkins, 2006; 507–8.

4. J. Senerviratne, J. Mandrekar, E. F. Wijdicks et al.Predictors of extubation failure in myasthenic crisis.Arch Neurol 2008; 65: 929–33.

5. S. Ahmed, J. F. Kirmani,N. Janjua et al. An update onmyasthenic crisis. Neurology 2005; 7: 129–41.

323

Part XIV Neuromuscular diseaseCase

96 Guillain-Barre syndromeAnupa Deogaonkar and Ehab Farag

Guillian-Barre syndrome (GBS) is an acute inflamma-tory demyelinating polyradiculoneuropathy affectingapproximately 1–2/100 000 each year. Typical clinicalmanifestations include progressive ascending motorweakness, areflexia and autonomic instability, withsevere cases progressing to respiratory failure. Usuallypatients have a history of infection 1–3 weeks prior toonset of the disease.

Case descriptionA 50-year-old, 75 kg, 166 cm male ASA physical sta-tus 2Ewas scheduled to undergo emergent exploratorylaparotomy. He had a history of tingling, numbnessand progressive ascending weakness in both handsand feet for 2 days. Two weeks previously, he hadseen his primary-care physician for a viral illnesswith fever, skin rash, and arthralgias. On admis-sion, the neurologic examination revealed bilateralfacial palsy, quadriparesis (distal more than proxi-mal), and areflexia. A lumbar puncture performedon admission revealed albuminocytologic dissociationand increased cerebrospinal fluid protein content com-pared with white cells. Nerve conduction studies indi-cated GBS. The patient was started on immunoglobu-lin therapy.On arrival to the preoperative holding area,intravenous metoclopramide, glycopyrrolate 0.2 mgand hydrocortisone were administered. A preinduc-tion radial arterial catheter was placed on the right sideand the patient was transferred to the operating tablewith full body warming blanket. After preoxygena-tion with 100% oxygen for 3 minutes, rapid sequenceinduction with cricoid pressure was performed. Aftersecuring the endotracheal tube, a nasogastric tube andesophageal stethoscope were placed. Isoflurane in 50%oxygen and air was used for maintenance. Intraoper-atively, blood pressure was labile and phenylephrineinfusion was used to maintain systolic blood pressurewithin 20% of the patient’s baseline. The patient wastransported to the intensive care unit (ICU) intubated

and mechanically ventilated. Immunoglobulin ther-apy was continued in the postoperative period andthe patient was weaned to extubate on postoperativeday 2.He developed deep venous thrombosis of the leftlower extremity on postoperative day 9, which was fur-ther complicated by pulmonary embolism.Hewas dis-charged home on postoperative day 18. His neurologicexamination after 6 months was normal.

DiscussionThere are no specific guidelines available for anes-thetic management for patients presenting with GBS;multidisciplinary care is necessary in the perioper-ative period to avoid fatal complications. Preoper-ative evaluation including history, physical exami-nation, electrocardiogram and laboratory data suchas cerebrospinal fluid and nerve conduction stud-ies will help identify patients who will require post-operative mechanical ventilation. Stress-dose steroidsshould be administered to all patients who have beenon steroids to reduce risk of cardiovascular collapsedue to inhibition of endogenous cortisol production.Autonomic instability can cause severe hypotension/hypertension intraoperatively necessitating the use ofinvasive intra-arterial blood pressuremonitoring. Suc-cinylcholine should not be used in order to avoidhyperkalemic response and cardiac arrest [1]. Non-depolarizing agents with a minimal cardiovascularside-effect profile are considered safe. Sudden changesin positioning intraoperatively should be avoided toprevent fluctuations of blood pressure. Alpha andbeta agonist and antagonists medications should bedrawn and kept ready to use. These patients arebelieved to be more sensitive to local anestheticsand overall the use of regional anesthesia is con-troversial. Profound sympathectomy due to regionalanesthesia in the setting of autonomic instability isnot desirable and may lead to cardiac arrest [2].Worsening of neurologic symptoms may occur after

324

Case 96. Guillain-Barre syndrome

epidural anesthesia for labor in pregnant patients pre-senting with GBS [3] or due to direct nerve rootdamage during regional technique [4]. Opioids maybe used to treat pain in the perioperative period inthe ICUs [5]. In nonambulatory patients, subcuta-neous heparin and compression stockings should beused to prevent deep vein thrombosis. Plasmaphere-sis and immunoglobulins may have to be contin-ued in the perioperative period besides conservativemanagement.

ConclusionEarly preoperative assessment, adequate monitoring,appropriate selection of anesthetic technique and post-operative intensive care treatment can lead to unevent-ful anesthetic management in GBS patients.

References1. J. M. Feldman. Cardiac arrest after succinylcholine

administration in a pregnant patient recovered fromGuillain-Barre syndrome. Anesthesiology 1990; 72:942–4.

2. A. Perel, A. Reches, J. T. Davidson. Anaesthesia inthe Guillian-Barre syndrome. A case report andrecommendations. Anaesthesia 1977; 32: 257–60.

3. S. Wiertlewski, A. Magot, S. Drapier et al.Worseningof neurologic symptoms after epidural anesthesia forlabor in a Guillain-Barre patient. Anesth Analg 2004;98: 825–7.

4. I. Steiner, Z. Argov, C. Cahan et al. Guillain-Barresyndrome after epidural anesthesia: direct nerve rootdamage may trigger disease. Neurology 1985; 35:1473–5.

5. D. S. Johnson,M. J. Dunn. Remifentanil for pain dueto Guillain-Barre syndrome. Anaesthesia 2008; 63:676–7.

325

Part XV End-of-life issuesCase

97 Conducting a family meeting to decidewithdrawal of careMarc J. Popovich

Conducting a family meeting to decide withdrawalof care can be a very difficult process. Both experi-ence and understanding is necessary for effective com-munication with the patient’s family. The followingcase discussion elucidates one approach to this delicatesituation.

Case descriptionA 79-year-old male was found unconscious in hishome by his wife. Upon arrival at the home, emer-gency medical services began full cardiopulmonaryresuscitation including intubation of the airway.Therewas return of spontaneous circulation after about5 minutes of resuscitation. The patient was taken tothe emergency department, where a head computedtomography (CT) scan demonstrated a large strokein the left middle cerebral artery territory. Cardiacenzymes were positive for acutemyocardial infarction.The patient was breathing spontaneously, but was oth-erwise unresponsive. He was admitted to the inten-sive care unit and over the course of the next 12 hoursbecame progressively febrile, hypotensive, and olig-uric. He remained unresponsive and required ongoingresuscitation and vasopressor support. The patient’swife and two sons had been informed of the patient’sstroke and myocardial infarction, but were now at thebedside requesting an update on the patient’s condi-tion and prognosis.

The intensivist explained to the family that ameaningful outcome (independent existence at home)would be extremely unlikely given the large stroke,myocardial infarction, and multisystem organ failure.The intensivist then asked the family about how thepatient would feel about being dependent on nursingcare in an extended care facility, perhaps indefinitely.When presented with this prognosis, it was clear to thefamily that such a treatment course would be incon-sistent with the patient’s previously stated wishes. Theintensivist offered the option of providing care primar-

ily directed towards comfort, including extubation.After a discussion, the family agreed to this treatmentcourse. One hour after extubation and withdrawal ofvasopressor support, the patient expired peacefully inthe presence of his family.

DiscussionOne of the more difficult situations a critical carephysician faces is a discussion regarding withdrawalof care or limitation of support with family mem-bers whose loved one has developedmedical problemsfrom which he or she clearly cannot recover. This dis-cussion is meant to aid in the management of the sub-set of patients who do not meet brain death criteria,but who are clearly moribund, have developed irre-versible organ failures, or where the medical opinionis that the chance ofmeaningful recovery is nil, despiteextraordinary efforts. Family discussions for these sit-uations require a great deal of forethought, as well asexperience in answering questions that typically arise.Entering into discussions with families for these typesof problems without prior consideration and expertisecan be fraught with difficulty, negative emotions, andmay lead to prolongation of suffering of an otherwisedying patient.

The first thing that the intensivist must establishbefore engaging in a withdrawal-of-life-support dis-cussion with a family is that the medical opinion isunanimous. In critical care, it is very common to havea multitude of physicians or consultants who may beasked to offer their opinion to family members dur-ing the patient’s clinical course. When there is clearevidence for need of a withdrawal-of-support discus-sion, the intensivist should review the situation withthe physicians involved tomake sure their opinions areconsistent, prior to engaging the family. It is particu-larly important, if the patient has undergone a surgi-cal procedure, that the surgeon is in agreement. Enter-ing into a family discussion without unanimity of the

327

XV. End-of-life issues

medical and surgical opinions will almost assuredly becounterproductive [1].

Second, since frequently in these situations physi-cians are reluctant to declare that the outcome of deathis 100% guaranteed, it is important for the intensivistto be prepared to describe the most realistic outcomescenario in real-world terminology. The intensivistshould additionally prepare to frame the discussion interms of how the patient would feel about, for exam-ple, indefinite ventilator dependency, dialysis, chroniclong-term care placement, constant nursing attentionincluding bathing and toileting, or the unlikelihood ofregaining functional independence.

Third, the intensivist should not fear using open,honest, and definitive words regarding the dying pro-cess.What’smore, the intensivist should be prepared tooffer a care plan individualized to the primary goal ofensuring the patient’s comfort. For example, the inten-sivist may decide that comfort relative to the patient’s

respiratory status will be impossible if extubated. Inthat situation, maintaining ventilator support (albeitwith minimal settings) may be necessary.

Most families appreciate an open, honest approachwhen discussing withdrawal of life supportive mea-sures. It is important that the intensivist feel comfort-able engaging in these types of discussions, becausealthough difficult, they are a necessary component ofan intensivist’s practice [2, 3].

References1. M. D. Siegel. End-of-life decision making in the ICU.

Clin Chest Med 2009; 30: 181–94.2. D.Wiegand. In their own time: the family experience

during the process of withdrawal of life-sustainingtherapy. J Pall Med 2008; 11: 1115–21.

3. H. M. Delisser. A practical approach to the family thatexpects a miracle. Chest 2009; 135: 1643–7.

328

Part XV End-of-life issuesCase

98 Brain deathVenkatakrishna Rajajee

Unfortunately, catastrophic events such as stroke orbrain trauma can lead to the cessation of neural func-tion. Understanding the precise diagnosis of braindeath is therefore of importance to the neurointen-sivist.

Case descriptionA 21-year-old male was brought to the neurosurgi-cal intensive care unit (ICU) following evacuationof a large right-sided traumatic subdural hematoma.Immediately prior to evacuation his Glasgow ComaScale (GCS) was 3T with absent pupillary and cornealreflexes but with intact cough and spontaneous respi-ratory effort. His vital signs appeared stable – bloodpressure (BP) 126/80mmHg, heart rate 99, respiratoryrate 16 and oxygen saturation 100% on Assist Controlventilation. The intracranial pressure was 18 mmHg.His GCS was 3T. His pupils were both 6 mm andunreactive. There was no cough, gag, corneal, or ocu-locephalic reflex. The anesthesiologist reported thatthe last use of neuromuscular blockade (4 mg vecuro-nium) was 45 minutes prior to admission. Brain deathevaluation was therefore deferred. Four hours later atrain-of-four evaluation revealed four twitches. TheGCS remained 3T and the intracranial pressure was57 mmHg. The patient’s family was counseled that hisprognosis was very poor and that a brain death eval-uation was in progress. They were counseled by therepresentative of the regional organ donation agency,who was notified of the patient’s condition at the timeof his arrival in the neurosurgical ICU. The parentsasked that the patient’s organs be donated and agreedwith completion of the brain death evaluation ratherthan choose immediate withdrawal of care. The firstbrain death evaluationwas performed, by a senior neu-rosurgical resident physician. The patient’s tempera-ture was 35 ◦C, so a warming blanket was applied.The sodium was 150 meq/L and the potassium was

3.2 meq/L; other laboratory values were normal.On clinical examination, the corneal, pupillary, gag,cough, and oculocephalic (“doll’s-eye” – when thebrainstem is intact in the comatose patient, the eyesdo not turn with the head when it is passively turnedfrom side to side) reflexes were absent. Ice-water injec-tion into the ears (following an otoscopic examina-tion) resulted in no eye movement (vestibulo-ocularreflex). On applying pain to the toes, dorsiflexionof the foot was noted. The resident physician docu-mented his evaluation as revealing “no signs of brain-stem function at this time.” The patient’s blood pres-sure (BP) then dropped to 70/40mmHg and heart rateincreased to 130, following output of 3 liters of dilute-looking urine over a 2-hour period. Normal salinewas administered with a pressure-bag and 3 litersinfused before the BP returned to 100/60 mmHg. Theurine specific gravity was 1.002, and a 2 mcg intra-venous therapy dose of desmopressin was adminis-tered. Repeat doses of desmopressin were used as nec-essary and a norepinephrine infusion was started. Abedside nuclear medicine cerebral perfusion scan wasobtained, which revealed no evidence of blood flowin the brain. A second brain death evaluation wasthen performed by an attending neurologist. Brain-stem reflexes were again absent and an apnea test wasperformed. The patient was preoxygenated at FiO2100% for 15 minutes and a baseline arterial bloodgas (ABG) test obtained. The PaCO2 was 40 mmHg.The ventilator was disconnected and a tracheal can-nula passed through the endotracheal tube, connectedto 6 L/minute flow of oxygen. The patient’s chest wasuncovered to observe for respiratory excursions andvital signs were closely monitored. After 10 minutes,an ABGwas drawn and the ventilator connected back.The repeat PaCO2 was 75 mmHg, with no evidenceof respiratory function. Brain death was declared; thepatient’s lungs, kidneys, and liver were successfullytransplanted.

329

XIII. End-of-life issues

DiscussionBrain death is the irreversible cessation of all brainfunction, including brainstem function. Brain deathis widely, although not universally, considered equiva-lent to cardiovascular death. In the USA, brain death isconsidered to represent legal death, although differentstates have very different criteria for the determinationof brain death. It is important to be aware of the specificcriteria used in one’s institution, which are generallyguided by local laws. The most common primary neu-rologic diseases that result in brain death are traumaticbrain injury and subarachnoid hemorrhage. Severalother conditions including hypoxic-ischemic injuryand fulminant hepatic failure can result in brain death.Although criteria vary, certain basic requirementsare fundamental to its determination. The AmericanAcademy of Neurology (AAN) has published practiceparameters for the diagnosis of brain death [1]. Certainprerequisites must be met before the evaluation pro-ceeds. First, there must be clear clinical or neuroimag-ing evidence of a central nervous system catastrophecompatible with the diagnosis of brain death – in thiscase the known traumatic brain injury with computedtomography evidence of intracranial hemorrhage andherniation. Metabolic confounders that might mimicbrain death must be excluded with a basic metabolicand electrolyte panel. Although the sodium level wasmildly elevated and the potassium level was low in thiscase, neither of these values was likely to confound thedetermination of brain death. Significant hypothermiashould be excluded (temperature above 32 ◦C in theAAN summary statement). No confounding medica-tions or drugs should be in the patient’s system. At thetime of this patient’s arrival in the neurosurgical ICU,brain death evaluation was postponed because he hadreceived a dose of vecuronium prior to transfer. Theevaluation commenced only after an adequate period(several half lives) was allowed for drug eliminationand train-of-four testing was performed.

The three cardinal features of brain death are coma,absence of brainstem reflexes, and apnea. A clini-cal examination is performed to confirm the patient’scomatose state and to evaluate brainstem reflexes.Thisincludes the pupillary reflex, corneal reflex, jaw jerk,oculocephalic reflex, gag reflex, and cough reflex. Thevestibulo-ocular reflex is also tested, with injection of50 mL iced water in each ear at least 5 minutes apart(after otoscopic examination to confirm patency of theexternal auditory canal and exclude tympanic injury).

No ocular movement should be seen with injection. Itis important to note that while movements originat-ing from the brain (including decerebrate posturing,decorticate posturing, and seizures) are incompatiblewith the diagnosis of brain death, both spontaneousand evoked movements originating often from spinalreflexes may be observed [2]. These include deep ten-don reflexes, triple flexion, Babinski reflex, and neckand upper body tonic reflexes. Hence, the finding ofdorsiflexion of the toes in this patient did not excludethe diagnosis of brain death.

Apnea testing has certain prerequisites – theAAN summary statement specifies a core temperature�36.5 ◦C, a systolic BP ≥90 mmHg, euvolemia,PaCO2 ≥40 mmHg and preoxygenation (usually withFiO2 100% for 10–15 minutes) preferably to PaO2�200 mmHg. Oxygenation is maintained followingventilator disconnection, typically with a tracheal can-nula with O2 flow at 6 L/minute. The next critical stepis to observe closely for any respiratory movement. Arepeat blood gas after 10 minutes that demonstrates aPaCO2 of 60 mmHg (or a rise of 20 mmHg) is con-sidered evidence of adequate stimulus for respiration.If no respiratory movement was seen with an ade-quate PaCO2 stimulus, the test is positive for apnea.If the patient demonstrates hemodynamic instabil-ity or oxygen desaturation prior to completion of the10 minute period, an arterial blood gas is drawn andthe test terminated. If the blood gas analysis confirmsan adequate CO2 stimulus with no observed respira-tory effort, the test is positive despite an observationperiod of �10 minutes.

A waiting period followed by a second evaluationis often required by institutional protocols, althoughseveral institutions do not consider this essential whenclear evidence of devastating brain injury is present.The AAN practice parameter recommends a 6-hourwaiting period for the adult patient. Longer intervalsare recommended for children, particularly neonates.With regards to this patient, the institutional proto-col specified a 12-hour waiting period, with the optionof eliminating this waiting period with the use ofconfirmatory testing. A confirmatory test – a nuclearmedicine scan – was therefore performed.

Confirmatory testing is generally not a require-ment for the diagnosis of brain death, although thereis considerable variation in international and state lawwith regards to the role of such testing and the type oftesting accepted. Confirmatory testing is useful whena full clinical evaluation cannot be performed, such as

330

Case 98. Brain death

Figure 98.1. Technitium-99 nuclear medicine scan reveals a “hollow skull” with no isotope uptake in brain.

in the patient with severe facial and orbital injury andpatients in coma from long-acting medications suchas pentobarbital. Also, as in the case discussed, specificinstitutional protocols may allow for reduction orelimination of the waiting period when confirmatorytesting is done. Confirmatory testing can be doneusing electroencephalography, cerebral angiography,nuclear medicine (Technitium-99) scanning, tran-scranial Doppler (TCD) and somatosensory evokedpotential testing. Specific stringent electroencephalo-graphic criteria exist to confirm “electrocerebralsilence” in brain death [3]. Electroencephalography

is prone to interference from medication use and toartifact that mimics brain activity. Cerebral angiog-raphy must demonstrate the complete absence ofintracranial blood flow. Somatosensory evoked poten-tial testing also must meet specific brain death testingcriteria and must demonstrate bilateral absence of theN20–P22 responses with stimulation of the mediannerve. Transcranial Doppler testing is complicated bythe fact that 10–20% of patients do not have an acous-tic window to permit TCD evaluation. Initial absenceof detectable Doppler signal (a common consequenceof true brain death) is therefore not an acceptable

331

XIII. End-of-life issues

diagnostic criterion. Transcranial Doppler maydemonstrate an absence of flow where flow was previ-ously demonstrated or may demonstrate short systolicpeaks with absent or reversed diastolic flow. The bed-side Technitium-99 nuclear medicine scan, performedin this case, demonstrates complete absence of isotopeuptake by the brain (Figure 98.1 – “hollow skull”) inthe brain-dead patient [4].

Institutional protocols may also specify the indi-viduals qualified to perform a brain death evaluation.Some institutions require that a neurologist or neuro-surgeon perform the evaluation, others permit nursesto perform the evaluation, followed by physiciancertification. Other common requirements includethat at least one of the evaluations be performed by anattending physician and that at least one evaluationbe performed by a physician who is not part of thetreating team. Members of the organ transplant teamare generally not permitted to perform the brain deathevaluation. In the case discussed, the first exam wasperformed by a neurosurgical resident (from the treat-ing team) and the second by a neurology attendingphysician.

The brain-dead patient usually, but not always,develops diabetes insipidus, caused by injury tothe hypothalamic–pituitary axis. This manifests bypolyuria anddilute urine (specific gravity�1.004) and,as seen in this case, can very rapidly lead to hypo-volemia and hypotension. Diabetes inspidus is man-aged with aggressive volume replacement and use ofeither intermittent desmopressin injections or a vaso-pressin infusion. Loss of brainstem vasomotor tone inthe brain-dead patient often necessitates the use of avasopressor infusion.

The key impetus to the accurate and timely diagno-sis of brain death is the critical shortage of organs avail-able for transplantation. Neurosurgical ICUs shouldhave a policy of early notification of the local organdonation agency, when a critically ill patient is admit-

ted, prior even to initiation of the brain death eval-uation. Typically, the representative of the organdonation agency then counsels the family. Donationfollowing brain death is considered preferable to dona-tion after cardiac death (performed when organ dona-tion is planned without declaration of brain death),which generally limits the organs available for trans-plantation. A sensitive, respectful, and supportiveapproach to communication with the family is abso-lutely essential. Every reasonable accommodation offamily wishes must be made following the declara-tion of brain death, including waiting for out-of-townfamily members prior to discontinuation of mechan-ical ventilation. Should a family refuse to accept thediagnosis of death, every effort is usually made togain their understanding and acceptance, despite therebeing legal precedent for discontinuation of mechan-ical ventilation from a brain-dead individual againstfamily wishes [5].

References1. Practice parameters for determining brain death in

adults (summary statement). The Quality StandardsSubcommittee of the American Academy ofNeurology. Neurology 1995; 45: 1012–14.

2. G. Saposnik, J. A. Bueri, J. Maurino et al.Spontaneous and reflex movements in brain death.Neurology 2000; 54: 221–3.

3. Guideline three: minimum technical standards forEEG recording in suspected cerebral death. AmericanElectroencephalographic Society. J Clin Neurophysiol1994; 11: 10–13.

4. H.Wieler, K. Marohl, K. P. Kaiser et al. Tc-99mHMPAO cerebral scintigraphy. A reliable, noninvasivemethod for determination of brain death. Clin NuclMed 1993; 18: 104–9.

5. R. E. Cranford. Discontinuation of ventilation afterbrain death. Policy should be balanced with concernfor the family. Br Med J 1999; 318: 1754–5.

332

Index

abdominal pain 77abducens nerve palsy 149ACE inhibitorsSTEMI 99

acetaminophen 297overdose 230

acidosis 34, 246acoustic neuroma 21, 22, 24, 27, 141acromegaly 177–8, 182–3activated clotting time 291activated partial thromboplastin time

291acute lung injurymechanical ventilation 257–9

acute respiratory distress syndrome257–9

treatment 258adenosine 46, 60, 71adenosine antagonists 81adenosine triphosphate. See ATPadvanced agestroke risk 238

Advanced Cardiac Life Support 154Advanced Trauma Life Support 123afterload 245, 251, 252Aintree catheter 50air embolism 36, 197clinical presentation 25detection of 25lethal volume 25posterior fossa 24–6

airwayaccess 104assessment prior to extubation

161cervical spine issues 165–7rheumatoid arthritis 172–5

cervical spine limitations 168–71compromise 94–5neurofibromatosis 141spinal surgery 130

airway crisisdeep brain stimulation 106–7

albumin 4albuterol 3, 4alfentanil 110, 114, 139allopurinol 30alteplase 27Alzheimer’s disease 295amaurosis fugax 85, 86aminocaproic acid 146amlodipine 45, 85, 94analgesia 5

craniotomy 17–20intravenous 17–18spinal surgery 131

analgesicsblood pressure control 43

anaphylaxis 68indocyanin green 66–8

anaplastic glioma 218anesthesia 5

aneurysm clipping 45–7DHCA 57

autonomic hyperreflexia 126awake craniotomy 112effect on intraoperative

neurolophysiologic monitoring137–9

emergence from 5delayed 27–9

hypotensive 147induction 3, 9children 188pregnant patients 216

inhalational 138intravenous 139left ventricular assist device

244–6maintenance 9neuraxial 225neurofibromatosis 141pediatric. See pediatric

neuroanesthesia

pregnancy 218–20prone positioning 160raised ICP 7–10regional 18–20. See regional

anesthesiaspinal cord injury 126spinal surgery 129, 130status epilepticus 315

anesthetic preconditioning 60anestheticsinhalational 5minimum alveolar concentration

64neurotoxicity 220side effects 273

aneurysm 23, 40rupture 41intraoperative 45–7, 70

aneurysm clipping 41–4anesthesia 45–7awake fiberoptic intubation 48–50coronary artery stent 51–3deep hypothermic circulatory arrest

55–7dexmedetomidine 63–5neuroprotection 59–61nitrous oxide 63–5pregnancy 215–17

aneurysm coiling 77–9renal failure 80–1subarachnoid hemorrhage 70–2

angiographyrenal failure 80–1

angioplastytransluminal 44

angiotensin receptor blockersSTEMI 99

ankylosing spondylitis 129, 168,169

clinical features 169anterior communicating artery

aneurysm 59antibioticsprophylactic 276

333

Index

anticoagulationalternatives to discontinuation 53intracerebral hemorrhage induced

by 291–4reversal of 292, 293

anticonvulsants 44see also individual drugs

antidiuretic hormone 279antihypertensive drugs 289Antihypertensive Treatment in Acute

Cerebral Hemorrhage(ATACH) trial 296

antiplatelet therapydual 53single 53

antisialogogues 49Apert syndrome 195, 196aprotinin 147aqueduct of Sylvius 192argatroban 291arousal 260arrhythmias 73arteriovenous malformations 23, 83ascorbic acid 81asleep–awake–asleep techniques 114aspirin 51, 52, 53, 71, 85, 86, 94, 96, 97,

99, 106hemorrhage 52STEMI 99

asthma 3astrocytoma 117atelectasis 209atlantoaxial instability 173atlanto-odontoid space 173atorvastatin 27, 85, 94, 177ATP 75atracurium 265atrial fibrillation 238atropine 34, 121auriculotemporal nerve 19autonomic hyperreflexia 125–6anesthesia 126symptoms 125

awake craniotomyanesthesia 112complications 115, 117–19epilepsy surgery 117–19

intraoperative neurologic decline113–16

awake fiberoptic intubation 48–50,170, 174

preparation 49

baclofen. Seeballoon angioplasty 52bamboo spine 169barbiturate coma 308barbiturates 60, 139status epilepticus 224

bare metal stents 51, 86barotrauma 255, 258Basic Life Support 154basilar artery occlusion 29benzodiazepines 139pharmacologic coma 285side effects 114status epilepticus 224

beta-blockers 43STEMI 99

betamethasone 221, 225Blalock Taussig shunt 210blood conservation 146–8blood gases 91blood glucose 27blood pressurenormalization of 13

blood pressure control 92blood salvage 147blood–brain barrier 285body temperature 13brachial plexus injury 163bradycardia 4, 15, 33, 65trigeminocardiac reflex 33

brainswelling 3

brain death 329–32hollow skull 331, 332tests for 329

brain injuryreasons to treat 304

brain relaxation 46brain tumorspregnant patients 218–20see also individual tumor types

brainstem auditory evoked potentials23, 138

bromocriptine 182bronchospasm 187Budd–Chiari syndrome 231buffalo hump 179Bullard laryngoscope 130bupivacaine 20, 110, 111, 114, 116burst suppression 59, 60reversal 59

buspirone 274

cafe-au-lait macules 140calcium channel blockers 224carbamazepine 30teratogenesis 216

carbon dioxidearterial 28end tidal 24

cardiac abnormalitiespost-subarachnoid hemorrhage

73–5cardiac arrest 154temporary 60therapeutic hypothermia 267

cardiac catheterization 208cardiac malformationschildren 207–9intraoperative management 208postoperative management 209preoperative evaluation 207

cardiac transplantation 245cardiogenic shock 99cardiopulmonary bypass 211, 241focal ischemia 242

cardiopulmonary resuscitationinfants 155prone 154–6

cardiothoracic surgery 241–3carotid endarterectomyairway compromise 94–5monitoring 88–90neurologic decline 91–3postoperative hematoma 94–5postoperative myocardial infarction

99–100preoperative evaluation 85–6stroke after 96–7

carvedilol 99, 251

334

Index

case stratification 4catheter-related infection 275–6Celsius Control System 273cement embolism 158central line 4central retinal artery occlusion 150,

151central venous access 124central venous pressure 192, 209cerebellar hemorrhage 38–40cerebellopontine angle 22, 24lesions of 22

cerebellumdisorders of 23

cerebral amyloid angiopathy 295–8Boston criteria 297

cerebral aneurysm 17cerebral angiography 269cerebral autoregulation 5cerebral blood flow 64, 284cerebral edema 3, 8, 218, 285cytogenic 8vasogenic 8

cerebral hemodynamics 88cerebral ischemia 37, 43, 60, 64cerebral metabolism 64, 88cerebral oximetry 92cerebral oxygenation 88cerebral palsy 160cerebral perfusion pressure 7, 11, 12,

42, 45, 192, 288, 305cerebral salt wasting syndrome 44, 277cerebral vasospasm 43, 269–71cerebral venous drainage 36cerebrospinal fluid drainage 36, 306cervical spineairway issues 165–7cervical spine limitations 168–71rheumatoid arthritis 172–5

unstable 165–7cervical spine injurymanagement 167

cervical spine limitations 168–71causes of 168

chartsreview of 4

Cheyne–Stokes breathing 8children

cardiac malformations 207–9induction of anesthesia 188neuroanesthesia. See pediatric

neuroanesthesianitrous oxide effects 220ventriculoperitoneal shunt 191–4

chronic obstructive pulmonary disease94

citrate toxicity 16clevedipine 289clonazepam 30clonidine 114clopidogrel 51, 52, 53, 71, 85, 86, 96,

97, 99coagulopathy 232Cobb’s angle 199cognitive dysfunction 91–3

neurological markers 93prediction of 92prevention 92

colchicine 30colloids 67coma

pharmacologic 285common peroneal nerve injury 37compartment syndrome 317, 318

abdominal 78orbital 152

computed tomography 3, 7, 11, 27, 41,73, 77, 191, 281, 284, 287, 288,309

epidural hematoma 12computed tomography angiography 28congenital heart disease 210congestive heart failure 244consciousness 260continuous mandatory ventilation

254continuous positive airway pressure

253, 255continuous spontaneous ventilation

254contraction band necrosis 75contrast-induced nephropathy 80–1

etiology and diagnosis 81management 81risk factors 80

controlled mechanical ventilation254

Cool Line System 273cooperative sedation 64coronary artery stentaneurysm clipping 51–3

corticosteroids 67, 218corticotropin releasing hormone

stimulation test 180coumadin 30, 39, 41, 94, 106, 227, 244,

291, 292cranial nerve blockade 19craniectomydecompressive 308

cranioplastytitanium mesh 40

craniosynostosis repair 195–8blood loss 196

craniotomyanalgesia 17–20awake. See awake craniotomypediatric 187–9posterior fossa 33–4postoperative seizures 15–16sitting 24, 36–7supratentorialpreoperative evaluation 1–5traumatic brain injury 11–14

cranium 7creatine kinase 318cricothyroidotomy 95Crouzon syndrome 188, 195,

196crystalloid 4, 67, 77CT. See computed tomographycuff-leak test 160, 161, 255Cullen sign 78Cushing’s reflex 4, 15, 282Cushing’s syndrome 178–81Cushing’s Triad 8cyclo-oxygenase inhibitors 53cystine glutamate exchanger 61

decadron 257deep brain stimulationairway crisis 106–7medication management 108–9preoperative evaluation 103–4

335

Index

deep hypothermic circulatory arrest210

aneurysm clipping 55–7safe period 56

deep venous thrombosis 41, 293prophylaxis 302

defibrillation 156degenerative disc disease 168desflurane 5, 160, 188, 219, 220,

322desmopressin 147, 233, 329dexamethasone 3, 4, 198, 220dexamethasone suppression test 180dexmedetomidine 17, 49, 50, 107, 113,

115, 117, 119, 139, 141, 146,160, 163, 201, 220

aneurysm clipping 63–5neuroprotective effect 64side effects 65

DHCA. See deep hypothermiccirculatory arrest

diabetes insipidus 184–5, 277, 279brain dead patients 332treatment 279

diabetes mellitus 51digoxin 251diphenhydramine 27, 66, 67diplopiapost-spinal surgery 149

dipyridimole 97divalproex sodium 227teratogenesis 229

doll’s-eye reflex 329dopamine 81, 265Doppler ultrasound 25precordial 36transcranial 43

droperidol 114drug-eluting stents 86. See stents,

drug-elutingduloxetine 179dural graft 40dyskinesia 108dyslipidemia 96dyspneacauses 95

dystonia 103deep brain stimulation 104

echocardiography 4, 208transesophageal 25, 55, 57, 73, 157transthoracic 251

eclampsia 221–6differential diagnosis 224maternal complications 224

edrophonium 322Ehlers–Danlos syndrome 42, 228electrocardiography 4, 73abnormalities 43

electrocautery 245electrocorticography 118electroencephalography 89, 117, 211electrolytes 16electromyography 135, 163embolectomy 157enalapril 27, 251encephalopathy 260–1endocarditisprophylaxis 209

end-of-life issuesbrain death 329–32withdrawal of care 327–8

endothelin antagonists 81endovascular coil embolizationsubarachnoid hemorrhage during

70–2enoxaparin 203, 293ephedrine 66epidural analgesia 201epidural anesthesiaperipheral nerve injury 163

epidural hematoma 12, 13, 301epilepsyawake craniotomy 117–19hemispherectomy 203–5intraoperative seizures 110–12

see also status epilepticusepinephrine 66, 67, 77, 79, 126, 143,

196eptifibatide 100erythropoietin 61, 147esmolol 71, 94, 215, 217, 289autonomic hyperreflexia 126

esmoperazole 321estradiol 220etomidate 77, 139, 220

evoked potentials 163extubation 160–2airway assessment 161deferred 161difficult 161problems of 161

Factor Seven for Acute HemorrhagicStroke (FAST) trial 296

factor VIIa, recombinant 291, 293Fallot’s tetralogy 208fenoldopam 81, 246fentanyl 3, 12, 15, 17, 27, 36, 45, 49, 55,

63, 114, 115, 128, 139, 141, 199,208, 209, 210, 215, 218, 299

distribution 189fetal heart rate 224fetal hypo-adrenalism 220fiberoptic intubationasleep 49awake 48–50, 170, 174

fibromuscular dysplasia 42flaccid paralysis 121flubiprofen 53fluid managementspinal surgery 131

flumazenil challenge 260fluorescence angiographyanaphylactic reaction 66–8indocyanin green 66–8

fondaparinux 291Fontan procedure 207foramen spinosum 34fosphenytoin 4, 71, 311, 315status epilepticus 224, 312, 315

fourth ventricle 22Frazier burr hole 38fresh frozen plasma 293furosemide 4, 7, 9, 30, 46, 81, 198

GALA trial 89gamma knife 30, 31Gasserian ganglion 33gastric emptying 126gastroesophageal reflux disease 17general anesthesia. See anesthesiagiganticism 182–3

336

Index

Glasgow Coma Scale 11, 12, 165, 253,256, 265, 270, 289, 299,329

“eight-intubate” rule 300and severity of injury 13

glutamate 265glutathione 61glycoprotein IIb/IIIa inhibitors 53glycopyrrolate 24, 34, 48, 49, 170, 195,

324Goldenhar syndrome 168, 188Grey–Turner sign 78growth hormoneexcess 182

Guillain-Barre syndrome 256,324–5

haloperidol 108headache 3postdural puncture 227postpartum 227–9thunderclap 73

headrest syndrome. See central retinalartery occlusion

HeartMate II 245hematocrit 4, 78, 211hematomapostoperative 94–5management 95

hemicraniectomy 40, 285hemispherectomy 203–5heparin 29, 53half-life 293low molecular weight. See low

molecular weight heparinreversal 71unfractionated 291

hepatic encephalopathy 230, 231West Haven Criteria 231

hirudin 291Horner’s syndrome 126, 228Humate-P 215Hunt–Hess Scale 270, 273hydralazine 38, 221, 224, 289hydrocephalus 36, 38, 43, 70, 269communicating 192ventriculoperitoneal shunt 191–4,

272–4hydrochlorothiazide 30, 94, 177

hydrocortisone 324hydromorphone 200, 209, 321hypercapnia 115hypercarbia 34, 50, 209, 246, 256hyperglycemia 13, 44, 242, 265, 306hyperkalemia 318hyperlipidemia 66, 177hypernatremia 304hyperperfusion syndrome 96hyperphosphatemia 318hypertension 4, 15, 17, 63, 66, 70,

96intracerebral hemorrhage 287–9permissive 233–6treatment. See antihypertensive

drugshyperthermia

avoidance of 306hypertonic saline 308hyperuricemia 318hyperventilation 9, 71, 302hypoalbuminemia 318hypocalcemia 16, 44, 318hypoglycemia 28, 306hypokalemia 44, 265hyponatremia 28, 44, 277

hypovolemic 278hypo-osmolality 28hypoplastic left heart syndrome 207,

210hypotension 24, 121, 251–2, 285

induced 46pregnant patients 217spinal cord injury 125trigeminocardiac reflex 33

hypotensive anesthesia 147hypothermia 13, 28, 56, 60, 123, 242,

285and wound infection 57induced 46, 308choice of cooling device 266decooling 267evidence for 266initiation of cooling 266post-cardiac arrest 265–7

intravascular cooling devices 273risks of 56

Hypothermia after Cardiac ArrestStudy Group 266

hypothyroidism 28, 96hypotonia 211hypoxemia 246, 285hypoxia 34, 50, 209

ibuprofen 3, 27ICP. See intracranial pressureIHAST trial 60indocyanin greenanaphylaxis 66–8properties 66

infectionventriculostomy 275–6

inferior petrosal sinus sampling 180inhalational anesthesia 138inotropes 37insulin 13Intensive Blood Pressure Reduction in

Acute Cerebral Hemorrhage(INTERACT) trial 296

intermittent mandatory ventilation 254internal carotid artery dissection 227International Normalized Ratio 291International Subarachnoid Aneurysm

Trial 42International Surgical Trial in

Intracerebral Hemorrhage(STICH) 297

intra-arterial thrombolysis 283intracerebral hemorrhage 15, 38–40anticoagulation causing 291–4cerebral amyloid angiopathy 295–8hypertensive 287–9medical management 289prognosis 289, 297

intracranial compliance curve 8intracranial hypertension 256, 302, 305intracranial pressure 7control of 192determinants of 7importance in management 305monitor placement 232

intracranial pressure, raised 3, 7–10,12, 188, 193, 304–8

acute liver failure 230–2causes 8communicating hydrocephalus 43craniosynostosis 198final treatment options 308

337

Index

intracranial pressure, raised (cont.)initial management 305later management 306reduction of 9, 71signs of 4stroke 284–6succinylcholine 309symptoms 15what to treat 305when to treat 305who to treat 304

Intraoperative Hypothermia forAneurysm Surgery Trial 46,63

intraoperative neurophysiologicmonitoring 137–9

intraparenchymal hemorrhage 41intravascular cooling devices 273intravenous anesthesia 139iodineallergy 67

ischemic optic neuropathy 150, 151ischemic preconditioning 60isoflurane 3, 5, 15, 24, 36, 45, 128, 163,

187, 188, 195, 203, 221, 322isoniazidliver failure 230

Jehovah’s Witnesses 146–8, 215jugular bulb monitoring 89

ketamine 5, 13, 139, 210status epilepticus 315

Klippel–Feil syndrome 168kyphoplasty 158kyphosis 168

labetalol 38, 43, 48, 71, 77, 106, 217,221, 224, 287, 289

autonomic hyperreflexia 126lactic acidosis 28lamotrigine 30, 314laryngeal mask airway 106, 161, 166intubating 50

laryngoscopy 45aneurysm rupture during 46

laryngospasm 187left ventricular assist device 244–6leukomalacia, periventricular 211

levetiracetam 41, 71, 314status epilepticus 312, 315

levodopaand deep brain stimulation 108

levothyroxine 96, 321lidocaine 4, 5, 9, 15, 16, 20, 24, 45, 48,

49, 110, 116, 169, 215, 217lisinopril 30, 99, 177LITA 3, 5liver failure, acutecauses 230classification 231coagulopathy 232neurologic symptoms 231raised ICP 230–2

liver transplantation 230–2lobectomypartial 40

local anesthesia 92awake craniotomy 114

local instillation of topical anesthesia.See LITA

lorazepam 218, 220, 311, 315status epilepticus 312

low molecular weight heparin 291low tidal volume ventilation 258lumbar puncture 272

macrophages 75magnesium 45magnesium sulfate 41, 221, 224autonomic hyperreflexia 126eclampsia 225

magnetic resonance angiography 227magnetic resonance imaging 3, 15, 27,

117, 195, 227, 281, 288, 295Mallampati score 169mannitol 3, 4, 7, 9, 12, 71, 189, 198,

284, 302, 308, 318pregnant patients 217, 220

manual in-line stabilization 165, 166,173

Marfan’s syndrome 42, 228Mayfield frame 7, 9Mayfield pins 17, 18, 46Mayfield skull clamp 45mechanical ventilation 253–6acute lung injury 257–9

complications 255modes of 253neurologic disorders 255noninvasive 255ventilator settings 254weaning from 254

medulladysfunction of 22

melatonin 220meningesinnervation 34

meningeus medius nerve 34meningioma 3falcine 33

mental status changes 260–1meperidine 274metaclopromide 108metformin 27, 28methotrexate 173methyldopa 224methylprednisolone 121, 123metoclopramide 324metoprolol 30, 77, 85, 94microcephaly 211microsurgical clip ligation 59midazolam 3, 15, 55, 88, 187, 199,

200autonomic hyperreflexia 126status epilepticus 310

middle cerebral artery aneurysm 48milrinone 246minimum alveolar concentration 64Monro–Kellie hypothesis of

intracranial compliance 305moon face 179morphine 17, 27, 28, 209motor evoked potentials 134–5, 137,

163, 200anesthetic sensitivity 135loss of 133–6

MRI. Seemagnetic resonance imagingmuscle relaxantsresistance to 322

myasthenia gravis 256, 321–3Osserman staging system 321

myasthenic crisis 321–3myasthenic syndrome 322

338

Index

myocardial infarctionnon-ST segment elevation. See

NSTEMIpostoperative 99–100ST segment elevation. See STEMI

myoglobin 317, 318myoglobinuria 318

N-acetylcysteine 81naloxone 27, 28National Acute Spinal Cord Injury

Study 123natriuretic peptides 278nausea and vomitingpostoperative 17

near-infrared spectroscopy 211, 243nefopam 274nerve block 17neuraxial anesthesia 225neuroendocrine surgerypreoperative evaluation 177–81

neurofibromatosis 137anesthesia 141spinal deformity 140–1

neurogenic shock 121–4neurologic declineintraoperative 113–16

neuromuscular blockade 28residual 28

neuromuscular blocking agents139

neurontin 30neuroprotection 210–12aneurysm clipping 59–61dexmedetomidine 64thiopental 71

neuropsychological assessment 104nicardipine 224, 277, 287, 289, 297nidus embolization 84nifedipine 71, 224autonomic hyperreflexia 126

NIH Stroke Scale 281, 282nimodipine 41, 45hypotension induced by 42

nitratesSTEMI 99

nitroglycerin 15, 71, 77, 94, 217autonomic hyperreflexia 126

nitroprusside 71, 217, 246, 289nitrous oxide 5, 24, 25, 37, 61, 122, 139,

163, 187, 188, 199, 203adverse effects 63aneurysm clipping 63–5effects in young children 220expansion of air emboli 209

NMDA receptor 61N-methyl-D-aspartate. See NMDAnon-ST segment elevation myocardial

infarction. See NSTEMInon-steroidal anti-inflammatory drugs

hemorrhage 52norepinephrine 252normal perfusion pressure

breakthrough 83–4normocarbia 13normoglycemia 13normothermia 13North American Symptomatic Carotid

Endarterectomy Trial(NASCET) 94

Norwood procedure 210NSTEMI 85nucleus tractus solitarius 73

obesity 17, 24, 36, 45, 63obstructive sleep apnea 17, 24occipital nerves 19occipito-atlanto-axial complex

172occlusive hyperemia 83octreotide 182oculocardiac reflex 33off-period 106opioids 17

sedative effects 28optic nerve 150osmotic diuretics 308osteoarthritis 168osteoporosis 157oxacarbazepine 30oxygen

regional saturation index. Seeregional saturation index

toxicity 255oxygen saturation 24

oxygen therapy 3oxygenation 13, 15oxytocin 220

paclitaxelstents eluting 51

PaCO2 306pain management. See analgesiapancuronium 210PaO2 306parasympathetic mediated

anti-inflammatory response75

parenchymal mass lesions 301Parkinson’s disease 103deep brain stimulation 103medication management 108–9

patent foramen ovale 37patient assessment 4peak inspiratory pressure 258pediatric cardiac surgeryneuroprotection 210–12

pediatric neuroanesthesiacardiac malformations 207–9craniosynostosis repair 195–8craniotomy 187–9neuroprotection 210–12scoliosis 199–201ventriculoperitoneal shunt 191–4

pedunculopontine tegmentum 22pentobarbital 309pharmacologic coma 285

percutaneous coronary intervention85

perinatal ischemic accidents 203perioperative stroke 238–9peripheral nerve injuryperioperative 163–4

Pfeiffer syndrome 195, 196phenobarbitalstatus epilepticus 310

phenylephrine 24, 60, 66, 77, 88, 165,215, 252, 324

phenytoin 3, 4, 15, 30, 71liver failure 230seizure prophylaxis 192, 299side effects 112status epilepticus 224, 315teratogenesis 216

339

Index

Pierre Robin syndrome 188pilocarpine 220pineocytoma 36Pittsburgh Compound B, 295plasmapheresis 325platelet aggregation 52platelet transfusion 71PMMAvenous leakage 157–9

pneumocephalus 28, 37pneumoniaventilator associated 255

pneumothorax 252volutrauma 255

polycystic kidney disease 42polymethylmethacrylate. See PMMApolytrauma 13pons 22positive end expiratory pressure 253,

257postdural puncture 227posterior fossaair embolism 24–6clots in 39craniotomy 33–4diseases of 21

posterior fossa surgerydelayed emergence from anesthesia

27–9preoperative evaluation 21–3

posterior ischemic optic neuropathy150

postoperative periodnausea and vomiting 17seizures 15–16

postoperative visual loss 150–3causes 150management 153prevention 152risk factors 152

postpartum headache 227–9causes 228

potassium chloride 55predicted body weight 258prednisone 251preeclampsia 221–6pregabalin 30pregnancy

anesthesia 218–20aneurysm clipping 215–17brain tumors 218–20eclampsia 221–6first/second trimesters 218late second/third trimesters 218physiological changes 216postpartum headache 227–9preeclampsia 221–6seizures 221–6teratogenic drugs 216term gestation 218

preload 245, 251premedication 4, 188see also individual drugs

preoperative evaluationcarotid endarterectomy 85–6deep brain stimulation 103–4neuroendocrine surgery 177–81posterior fossa surgery 21–3spinal surgery 128–31supratentorial craniotomy 1–5

pressure control ventilation 254pressure support ventilation 254prone position 160anesthesia problems 161

propofol 4, 5, 12, 15, 17, 21, 27, 41, 45,55, 59, 60, 63, 91, 106, 110, 113,117, 119, 122, 128, 139, 146,154, 165, 200, 215, 218, 219,221, 257, 265, 299, 321

advantages of 301pharmacologic coma 285side effects 106status epilepticus 224, 310

propranololautonomic hyperreflexia 126

prostaglandins 81protamine sulfate 71, 291prothrombin complex concentrate 291pseudoxanthoma elasticum 42pulmonary artery catheter 251pulmonary capillary occlusion

pressure 251pulmonary embolism 293pulmonary hypertension 37pulmonary vascular resistance 207, 208pulse oximetry 3pulseless electrical activity 154pump function 251, 252pyridostigmine 321, 322

radial artery catheter 7, 11, 17, 41, 45,55

Randomized Evaluation of MechanicalAssistance for the Treatment ofCongestive Heart failure(REMATCH) trial 245

ranitidine 48rapid shallow breathing index 255RapidBlueTM system 273reactive airway disease 125rebleeding 42, 46recombinant tissue plasminogen

activator. See tissueplasminogen activator

recurrent laryngeal nervepalsy 37

regional anesthesiaautonomic hyperreflexia 126

regional saturation index 211low 212

remifentanil 4, 5, 16, 17, 21, 24, 27, 45,48, 88, 91, 115, 122, 133, 139,154, 169, 170, 187, 195, 200,203, 215, 219, 257, 321, 322

half life 28side effects 189

renal failure 51angiography 80–1

reticular activating system 260retrograde cerebral perfusion 242retroperitoneal hematoma 77, 78diagnosis 78

rhabdomyolysis 312, 317–18diagnosis 318management 318

rheumatoid arthritis 129, 168airway issues 172–5

risperidone 265rocuronium 15, 27, 187, 199, 203, 208,

218ropivacaine 20, 200rosuvastatin 99

scalp block 18, 110scoliosis 129, 160, 163, 199–201anesthetic management 201etiologic classification 199

seizurescauses 311

340

Index

intracerebral hemorrhage 289intraoperative 110–12post-cardiothoracic surgery 241postoperative 15–16causes 16

pregnancy 221–6prophylaxis 41, 71, 193raised intracranial pressure 306supratentorial craniotomy 3treatment 15

sevoflurane 5, 21, 27, 106, 154, 188,199, 200, 219, 220, 244, 322

shock liver 231SIADH 44, 277, 278simvastatin 81single photon emission computed

tomography. See SPECTsirolimusstents eluting 51

sisseminated intravascular coagulation318

sitting craniotomy 24, 36–7complicationsnerve injuries 37quadriplegia 37

sitting position 28sleep apnea 177sniffing position 170sodium abnormalities 277–9sodium balance 278sodium bicarbonate 81sodium citrate 48sodium pentothal 3, 5somatosensory evoked potentials 36,

89, 117, 133–4, 137, 200loss of 133–6

SPECT 83spinal cord injury 13, 121–4anesthesia 126

spinal injuryautonomic hyperreflexia 125–6cervical spine 165–7

spinal shock 121–4, 125spinal stabilization, openpolymethylmethacrylate

augmentation 157–9spinal surgeryacute injury 121–4airway 130anesthesia 129, 130

blood loss 131complex 128–31, 133–6, 154–6,

160–2diplopia after 149extubation 160–2fluid management 131Jehovah’s Witnesses 146–8loss of evoked potentials 133–6pain control 131positioning 130preoperative evaluation 128–31prone cardiopulmonary

resuscitation 154–6vascular complications 143–5visual loss 150–3

spontaneous breathing trial 255ST segment elevation myocardial

infarction. See STEMIstaged feeders 84StarClose device 77Starling’s law 246statins

STEMI 99status epilepticus 221–6

complex partial 315convulsive 309management 311–13nonconvulsive 311, 314–16post-cardiothoracic surgery 241treatment 224

STEMI 99medical management 99

stents 29bare metal 51, 86coronary artery 51–3drug-eluting 51, 86paclitaxel 51sirolimus 51

thrombosis 51stress response 60stroke 4, 38, 85

hemorrhagic 38–40, 281initial management 281–3ischemic 281, 285perioperative 238–9perioperative risk 88post-carotid endarterectomy 96–7raised intracranial pressure 284–6

subarachnoid hemorrhage 8, 80, 251,277

aneurysm clipping 41–4, 59–61aneurysm rupture 41cardiac abnormalities 73–5cerebral vasospasm 269–71

during aneurysm coiling 70–2Hunt–Hess classification 41, 74hydrocephalus after 272management 42risk factors 42systematic approach 43

subdural hematoma 7, 8, 244–6, 301and postoperative seizures 16

subendocardial infarction 73succinylcholine 9, 21, 128, 165, 215,

221, 318hyperkalemic response to 324raised intracranial pressure 312resistance to 321

sufentanil 17, 114, 139, 146, 160, 163sulfonamidesliver failure 230

supraorbital nerve 19supratentorial craniotomypreoperative evaluation 1–5traumatic brain injury 11–14

supratrochlear nerve 19Swan–Ganz pulmonary artery catheter

55sympathetic storm 75syndrome of inappropriate antidiuretic

hormone secretion. See SIADHsystemic vascular resistance 207

tachycardia 50Takotsubo cardiomyopathy 75tamponade 252temperomandibular joint disease 173tentorial herniation 11teratogenic drugs 216theophylline 67, 81therapeutic hypothermia 272–4choice of cooling device 266decooling 267evidence for 266initiation of cooling 266post-cardiac arrest 265–7

thermoregulation 126thiopental 46, 47, 71, 139, 221pharmacologic coma 285

thrombate III 203dose calculation 204

thromboplastin III deficiency 203–5thunderclap headache 73

341

Index

tic douloureux. See trigeminalneuralgia

tirofiban 100tissue plasminogen activator 281, 282,

284titanium mesh cranioplasty 40tocolytics 220topiramatestatus epilepticus 310

tracheal tube exchange catheter 161tracheostomy 95, 255tramadol 30transcranial Doppler 212transcranial motor evoked potentials

133transesophageal echocardiography 25,

36, 55, 57, 73, 157transient ischemic attacks 100transthoracic echocardiography 251traumatic brain injury 11–14, 299–302treatment plan 4–5tremor 103deep brain stimulation 104

trigeminal nerve 33trigeminal nerve block 31trigeminal neuralgia 30–2gamma knife 31microsurgical

exploration/decompression 31percutaneous procedures 31

trigeminocardiac reflex 30, 31, 32,33–4

triple-H therapy 44troponins 45tryptase 68tumor resection 40

valproic acidstatus epilepticus 312, 315teratogenesis 216

Valsalva maneuver 26valvular heart disease 239vascular malformation 40vascular proceduresaneurysm clipping. See aneurysm

clippinganeurysm coiling. See aneurysm

coilingcarotid endarterectomy. See carotid

endarterectomyvascular surgeryvascular complications

143–5vasomotor paralysis 84vasopressin 252vasopressors 37vecuronium 3, 12, 36, 45, 63, 88, 91,

195, 330venous air embolism. See air

embolismventricular fibrillation 57, 74

ventricular tachycardia 57, 74ventriculo-atrial shunt 37ventriculoperitoneal shunt

233–6children 191–4dependence on 272–4

ventriculostomy 38infection 275–6

vertebroplasty 158vestibular schwannoma 34vitamin K 293vitamin K antagonists 291volume replacement 44volutrauma pneumothorax 255von Recklinghausen’s disease. See

neurofibromatosisvon Willebrand’s disease 215, 233–6

water balance 278white matter neuropathology

211Wilson’s disease 231withdrawal of care 327–8wound infection 57

xanthochromia 42xenon 61, 220

zygomaticotemporal nerve 19

342

052119380 x

Documents

Transcript of 052119380 x