Artificial Organs

New Techniques in Surgery Series

Nadey Hakim (Ed)

Artificial Organs

1 3

Editor

Nadey HakimRenal and Transplant ServicesHammersmith HospitalLondon, UK

ISBN 978-1-84882-281-8 e-ISBN 978-1-84882-283-2DOI 10.1007/978-1-84882-283-2

British Library Cataloguing in Publication DataA catalogue record for this book is available from the British Library

Library of Congress Control Number: 2009920379

# Springer-Verlag London Limited 2009Apart from any fair dealing for the purposes of research or private study, or criticism or review, as permittedunder the Copyright, Designs and Patents Act 1988, this publication may only be reproduced, stored ortransmitted, in any form or by any means, with the prior permission in writing of the publishers, or in the caseof reprographic reproduction in accordance with the terms of licenses issued by the Copyright Licensing Agency.Enquiries concerning reproduction outside those terms should be sent to the publishers.The use of registered names, trademarks, etc., in this publication does not imply, even in the absence of a specificstatement, that such names are exempt from the relevant laws and regulations and therefore free for general use.Product liability: The publisher can give no guarantee for information about drug dosage and application thereofcontained in this book. In every individual case the respective user must check its accuracy by consulting otherpharmaceutical literature.

Printed on acid-free paper

Springer ScienceþBusiness Mediaspringer.com

Foreword

Artificial organs have come a long way since the first dialysis machine, the rotatingartificial kidney, was invented in 1944 by Willem Johan Kolff, who is known as the‘‘father of artificial organs’’. At that time he met stiff resistance from his hospitalsuperiors but his persistence paid off and a million saved lives have been attributedto his first invention. He was indeed the first to mix medicine and engineering. Anartificial organ is any machine, device, or other material that is used to replace thefunctions of a faulty or missing organ or other parts of the human body. Some bodyparts are more of a challenge than others. The heart has one purpose and it is topump blood; however, the liver has biochemical and physiological functions whichare difficult to simulate. The implantation of an artificial organ is critical because ofthe patient life dependency on the artificial organ itself. The treatment of choice isorgan transplantation, however, transplant candidates face a long waiting time andmany die while on the waiting list. In addition there are patients who are excludedfrom transplantation because of age or presence of other diseases. This book pre-sents an overview of the current state of knowledge of artificial organs including theliver, pancreas, kidney, heart, cochlea, skin, stem cells, composite tissue allograft,and sphincters. It is designed for students interested in the field of organ replacementand bioartificial organs and it promotes an understanding of the designs andmaterials required for successful implants. It fulfills a useful place in the medicalliterature. The editors have gathered together experts in a variety of different fieldsboth from the experimental and clinical aspects. Research in the field of artificialorgans will lead to a closer relationship between science and technology and providesa stepping stone for the future.

Professor the Lord Darzi of Denham KBE

v

Contents

1 Management of Multiorgan Failure After Artificial Organ ImplantationMichael Devile, Parind Patel and Carlos MH Gomez . . . . . . . . . . . . . . . . . 1

2 Artificial Circulatory SupportJohn Mulholland. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21

3 The Artificial KidneyChristopher Kirwan and Andrew Frankel . . . . . . . . . . . . . . . . . . . . . . . . . . 39

4 Liver SubstitutionSambit Sen and Roger Williams . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 57

5 Glucose Sensors and Insulin Pumps: Prospects for an Artificial PancreasMartin Press . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 77

6 From Basic Wound Healing to Modern Skin EngineeringL.C. Andersson, H.C. Nettelblad and G. Kratz. . . . . . . . . . . . . . . . . . . . . . . 93

7 Artificial SphinctersAustin Obichere and Ibnauf Suliman . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 107

8 Cochlear ImplantGeorge Fayad and Behrad Elmiyeh . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 133

9 Stem Cells and Organ ReplacementNatasa Levicar, Ioannis Dimarakis, Catherine Flores,Evangelia I Prodromidi, Myrtle Y Gordon and Nagy A Habib. . . . . . . . . . 137

10 Composite Tissue Transplantation: A Stage Between SurgicalReconstruction and CloningEarl R. Owen and Nadey S. Hakim . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 165

Index . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 179

vii

Contributors

L C Andersson MD Dr med PhDConsultant Plastic SurgeonAnelca ClinicSpecialist Clinic for Aesthetic

and Reconstructive Plastic SurgeryLondonUK

Michael Devile BSc MBBS MRCP(UK)Specialist Trainee Registrar

in AnaesthesiaOxford DeaneryOxfordshireUK

loannis Dimarakis MRCSDepartment of Surgical Oncology

and TechnologyFaculty of MedicineImperial College LondonLondonUK

Mr B Elmiyeh MBBS MRCS DO-HNSBasildon HospitalBasildon, EssexUK

Mr G Fayad MD FRCS FICSBasildon HospitalNethermayneBasildonEssexUK

Catherine Flores PhDDepartment of HaematologyFaculty of MedicineImperial College LondonLondonUK

Andrew Frankel MBBS MD BSc FRConsultant NephrologistWest London Renal and Transplant

CentreImperial College Healthcare NHS TrustHammersmith HospitalLondonUK

Carlos MH Gomez LMS FRCA MDEDICM

Consultant in Intensive Care Medicineand Anaesthesia

St Mary’s HospitalImperial NHS TrustLondonUK

Myrtle Y Gordon PhD DScDepartment of HaematologyFaculty of MedicineImperial College LondonLondonUK

Nagy A Habib ChM FRCSDepartment of Surgical Oncology


Nadey Hakim KCJSJ, MD, PhD, FRCS,FRCSI, FACS, FICS(Hon)

Max Thorek Professor of SurgeryWest London Renal and Transplant

CentreImperial College Healthcare NHS TrustLondonUK

ix

Christopher Kirwan MDSpecialist Registrar Renal MedicineWest London Renal and Transplant

CentreHammersmith Hospitals NHS TrustLondonUK

G Kratz MDDepartment of Hand SurgeryPlastic Surgery and BurnsUniversity Hospital LinkopingLinkopingSweden

Natasa Levicar PhDDepartment of Surgical Oncology


John Mulholland BEng BSc MScClinical Perfusion ScientistLondon Perfusion ScienceWestminsterLondonUK

HC Nettelblad M.D. PhDAssociate ProfessorAnelca ClinicSpecialist Clinic for Aesthetic

and Reconstructive Plastic SurgeryLondonUK

Mr Austin Obichere MBBS MD FRCSFRCS (Gem)

Consultant Colorectal SurgeonDepartment of SurgeryUniversity College London HospitalLondonUK

Professor Earl R Owen AO MBBS(Uni of Syd) MD(Lyon) DScFRACS FRCS PICS FRCSE

Specialist in Microsurgery, Handand Infertility Surgery

North Sydney, NSWAustralia

Parind Patel BSc MBBS DMS FRCAEDICM

Consultant in Intensive CareMedicine

Hammersmith HospitalImperial NHS TrustLondonUK

Dr Martin Press MA MSc FRCPConsultant EndocrinologistRoyal Free HospitalLondonUK

Evangelia I Prodromidi PhD MScDepartment of Renal MedicineFaculty of MedicineImperial College LondonLondonUK

Sambit Sen MD MRCPDepartment of GastroenterologyLuton and Dunstable HospitalLewsey RoadLuton, BedfordshireUK

Mr Ibnauf Suliman BSc(Hons) BMMRCS(Eng)

Specialist RegistrarSouth East ThamesLondon DeaneryLondonUK

Roger Williams CBE MD FRCPFRCS FRCPE FRACP FMedSciFRCPI(Hon) FACP(Hon)

Professor of Hepatology and ConsultantPhysician

The Institute of HepatologyUniversity College LondonLondonUK

x Contributors

1Management of Multiorgan Failure After ArtificialOrgan Implantation

Michael Devile, Parind Patel and Carlos MH Gomez

Cardiovascular Failure

The requirements of hemodynamic support inpatients with artificial organs follow the sameprinciples as for any other patient. Ultimatelythe need for such support will arise if the patientis in shock. The key disturbance in shock is theinability of the tissues and cells to obtain andutilize enough oxygen for their aerobic respira-tory needs, ultimately resulting in cell death.

Classification

Shock is classified into four types based on theunderlying pathophysiological mechanism:hypovolemic, cardiogenic, obstructive, and dis-tributive. (Table 1-1). The patient with artificialorgans will have a higher risk of all of the abovemechanisms at various stages of the peri andpostoperative process and extra vigilance istherefore necessary. The management of shockshould nevertheless follow basic critical careprinciples: re-establishing perfusion to vitalorgans while treating the underlying condition.

Hypovolemic Shock

This can often present as an emergency (due tomassive hemorrhage for instance) or in a moreindolent manner leading to slow organ hypoper-fusion (due to a mismatch of fluid input andoutput). Initial resuscitation should consist ofrestoring intravascular volume, but moreimportantly reversing the cause (bleeding

vessels, coagulopathy, reversing gastrointestinallosses). The choice of fluid should be based onthe fluid type that has been lost. For example,blood and blood products should replace bloodand crystalloid should be used for vomiting anddehydration. For mild to moderate blood losssimilar to that expected in a routine renal trans-plant crystalloid (2–3 times the lost bloodvolume) or colloid (1–2 times the lost bloodvolume) is usually appropriate.

The debate of crystalloids versus colloid hasnot been resolved; there exists numerous con-flicting trials, meta-analysis, and opinions. Thecrystalloids of choice in resuscitation are normalsaline or lactated Ringer’s solution because theirosmolality is similar to that of the intravascularvolume. In large-volume resuscitation, however,excessive normal saline infusion may producehyperchloremic metabolic acidosis. As thevolume of distribution is much larger for crys-talloids than for colloids, resuscitation withcrystalloids requires more fluid to achieve thesame end points and results in more edema.Crystalloids are less expensive. The currentchoice of colloids includes hydroxyethyl starch,gelatins, and albumin all of which offer moreefficient intravascular volume expansion. Theadministration of hydroxyethyl starch mayincrease the risk of acute renal failure, althoughnewer preparations appear to be less proble-matic. Previous concerns of albumin resuscita-tion leading to increased mortality has recentlybeen dispelled by the SAFE study indicating thatalbumin administration was safe and as effectiveas crystalloid. With any fluid there is inevitabledilution of blood (red blood cells) and coagula-tion factors which needs to be considered.

N. Hakim (ed.), Artificial Organs, New Techniques in Surgery Series 4, DOI 10.1007/978-1-84882-283-2_1,� Springer-Verlag London Limited 2009

1

Whatever the fluid, the therapy should betitrated to re-establish normal blood pressure,pulse, cardiac output, and end-organ perfusion.

Cardiogenic Shock

Cardiogenic shock essentially refers to the fail-ing pump. The blood flow to the vital organs andperipheries is decreased due to an intrinsicdefect in cardiac function. This can be due topathology affecting the heart muscle, rhythm, orthe valves.

The most common cause, both generally andwithin transplant medicine and artificial organs,is myocardial ischemia. As in all types of shock,treating must be the priority: in cardiogenic shockthis is achieved by reversing the ischemia (withangioplasty, thrombolysis, or emergency coron-ary artery bypass grafting), correcting the dys-rhythmias, or simply supporting the circulationwhile allowing time for the stunned myocardiumto recover while preserving perfusion. The early

introduction of such measures has resulted in asignificant improvement in ischemia induced car-diogenic shock and mortality.

In terms of hemodynamic support, variousinotropic agents exist (Table 1-2) which act viamodulation of the sympathetic nervous systemeither as direct agonists (dobutamine, dopexa-mine, dopamine, adrenaline) or indirectly byinhibiting phosphodiesterase III (e.g., milri-none). Both mechanisms result in an increasein cAMP, an intracellular modulator of myocar-dial contractility and arteriolar vasodilation.Often the inotropes, because of the arterial vaso-dilatory effects or the concurrent systemicinflammatory response syndrome that mayoccur with cell death, are combined with vaso-pressors to maintain overall perfusion pressurein general and coronary artery perfusion pres-sure in particular. Although the use of theseagents is thought of as essential to the care of ahemodynamically unstable patient, there is greatvariability within countries and centers as to theoptimum agent. It is also notable that the role of

Table 1-1. Causes of Shock

Type of shock Signs Hemodynamic variables Causes relevant to artificial organs

Hypovolemic Weak pulse -CVP/PAOP Massive hemorrhageNarrow pulse pressure -CO Gastrointestinal fluid lossPoor peripheral perfusion "SVR Diuresis

Cardiogenic Pulmonary/peripheral edema "CVP/PAOP Acute coronary syndromeGallop rhythm, arrhythmias, & murmurs -CO Valvular heart disease/dysrhythmiasPoor peripheral perfusion "SVR Cardiomyopathy

Obstructive Distended neck veins "CVP/PAOP Tension pneumothoraxPoor peripheral perfusion -CO Cardiac tamponadeRespiratory failure "SVR Pulmonary embolism

Distributive Pyrexia -CVP/PAOP Sepsis/SIRSWarm and vasodilated peripheries "CO AnaphylaxisEdema -SVR Acute liver failure

Table 1-2. Vasoactive Drugs in Sepsis and the Usual Hemodynamic Responses

Drug Dose Principal mechanism Cardiac output Blood pressure SVR

Dobutamine 2–20 mcg/kg/min Beta 1 ++ + +Dopamine 2–10 mcg/kg/min Dopamine!Beta 1!Alpha ++ + +Dopexamine 0.25–4 mcg/kg/min Dopamine + Beta 2 ++ +/– –Epinephrine 0.01–1 mcg/kg/min Beta 1+ Beta 2 + Alpha ++ + +Milrinone 0.3–0.7 mcg/kg/min Phosphodiesterase inhibitor + +/– –Norepinephrine 0.01–1 mcg/kg/min Alpha > beta 1, beta 2 + ++ ++Phenylephrine 0.2–2.5 mcg/kg/min Alpha + ++ ++Vasopressin 0.4–0.10 U/min V1 receptor + + ++

2 Artificial Organs

sympathetic nervous system stimulation is ofquestionable benefit in the failing heart wherethe underlying cause of failure has not beenaddressed.

Recently Levosimendan, an inotropic agentthat acts independently of the sympathetic ner-vous system by sensitizing the myocardial con-tractile proteins to calcium has been shown toimprove hemodynamics and cardiac output inpatients with cardiogenic shock following ACSas well as after coronary bypass surgery. Thismay prove useful as a pharmacological bridgeduring myocardial stunning, possibly with fewer(sympathetically mediated) detrimental effects.

Combined with revascularization, mechanicalsupport with IABP counterpulsation improvesoutcome in patients with cardiogenic shock fol-lowing acute coronary syndrome. Specific to car-diogenic shock, the use of ventricular assistdevices may transiently improve hemodynamicsbut this has not yet been translated into mortalitybenefits and presently is reserved as a salvagestrategy in centers where it exists and other mea-sures have failed. The same can be said of extra-corporeal membrane oxygenation, although bothof these techniques are more useful followingcardiac transplantation.

It must also be remembered that hypovolemiamay also play a significant part in cardiogenicshock, requiring the careful administration offluid to improve cardiac output via the Frank–Starling principle. Simultaneous monitoring forfluid overload via central venous, pulmonaryartery pressures, and signs of pulmonary edemamust be undertaken.

Distributive Shock

Distributive shock occurs when peripheral vas-cular dilatation causes a fall in systemic vascularresistance. As a result cardiac output is oftenincreased but the perfusion of many vital organsis compromised because the blood pressure istoo low and the body loses its ability to distri-bute blood properly

The most common cause is septic shock, buta similar picture can be seen in anaphylaxis andacute adrenal insufficiency. In addition a signif-icant organ reperfusion induced systemicinflammatory response syndrome (SIRS) imme-diately follows transplantation which can mimicdistributive shock. However, newer immuno-suppressive agents have dramatically reduced

the rates of acute graft rejection over the lastdecade but instead have exacerbated the problemof post-transplant infections. Although clinicalfeatures include warm peripheries boundingpulses, the significant alterations in the microcir-culation results in tissue and organ dysfunction.The longer the distributive shock remains thegreater the organ failure and mortality. As a resultof the persistently high mortality from septicshock, the Surviving Sepsis Guidelines havebeen acknowledged and implemented across theworld. Incorporated within these are the essentialthemes of early and appropriate administrationof antibiotics and early goal directed shock resus-citation. In the initial stage, because of capillaryleakage, hypovolemia contributes significantly tothe shock. Hence fluid resuscitation should beinitiated as in hypovolemic shock. Simulta-neously, vasopressor therapy is required to main-tain perfusion in the face of life-threateninghypotension, even when hypovolemia has notyet been resolved. Below a certain mean arterialpressure, autoregulation in various vascularbeds can be lost and perfusion can becomelinearly dependent on pressure. A mean arterialpressure of 65 mmHg is generally accepted as atarget to preserve tissue perfusion. However, inthose who have hypertension, as in many renaltransplant patients, autoregulation will occur ata higher point and hence greater MAP will beneeded.

Fluids and vasopressors are targeted tohemodynamic variables along with assessmentand trends of regional and global perfusion suchas urine output, conscious level, acidemia, andblood lactate concentrations.

As shown in Table 1-2 many vasopressors areavailable. Norepinephrine may have advantagesover epinephrine in causing less compromise tothe splanchnic circulation and lactatemia. Dopa-mine may cause unwanted tachycardia andarrhythmias and provides no advantage in pre-venting renal failure. Concerns also exist aboutthe immunosuppressive effects of dopamine.Vasopressin may be an alternative to norepi-nephrine or in patients with norepinephrineresistance, although again there is little data tosupport its routine use and certainly in higherdoses cardiac, digital, and splanchnic ischemiahave been reported.

To complicate matters, there is significantcardiac dysfunction directly related to sepsisthat may require combination therapy with ino-tropic agents as for cardiogenic shock.

Management of Multiorgan Failure After Artificial Organ Implantation 3

Finally, steroids remain controversial: theyprobably have significant effects in reversingshock, but long-term side effects such as neuro-pathy and recurrent infections are thought to beat least in part responsible for the observed lackof mortality benefit.

Obstructive Shock

The treatment of obstructive shock is relativelysimple in principle although may prove to bedifficult in practice.

Relief of the obstruction is achieved by

• Urgent pericardiocentesis in cardiac tampo-nade. This is a relatively likely complicationfollowing cardiac transplant, but less likely inother solid organs although anticoagulationor coagulopathy that may occur can lead totamponade.

• Needle thoracostomy followed by formalchest drain in tension pneumothorax. Thismay occur perioperatively secondary to dia-phragmatic damage or as a complication ofcentral venous catheterization or positivepressure ventilation.

• Thrombolysis, pulmonary embolectomy, orsurgical removal of a massive pulmonaryembolism should be considered in any post-transplant patient presenting with suddendyspnoea and hypoxia. The condition alongwith fat embolism is more frequent followinglung transplant.

Maintaining a high preload may help withobstructive shock, with fluid resuscitationimproving the patient’s cardiac output andhypotension temporarily while buying time fordefinitive intervention.

Conclusion

Shock is one of the most frequent situationsencountered in critical care patients. Post-transplant and/or artificial organ insertionshock is associated with poor outcome unlessrapidly treated and reversed. In any shockedpatient, hemodynamic therapy should be tar-geted to achievable goals. This includes

• Find and treat the cause: this is the mostdifficult and yet crucial aspect as the diagnosisof the underlying problem may be difficult to

establish. Without addressing the ischemia orhemorrhage the remaining targets would bepointless.

• Establish adequate intravascular volumeresuscitation: this is relevant to hypovolemicbut also to other forms of shock in order tooptimize cardiovascular function.

• Establish an adequate blood pressure: thismay require the use of vasopressors and ino-tropes, often in combination.

• Ensure organ function by targeting the abovetherapy to maintain and preserve function ofall systems and prevent the situation cascad-ing into multi-organ failure.

The challenge is to reverse the patient’s deterior-ating condition rapidly to achieve an overallsuccessful outcome.

Respiratory Failure

Respiration is the complex process involving ven-tilation, pulmonary gas exchange, oxygen deliv-ery by the circulation, and oxygen utilization bythe tissues for the production of cellular high-energy phosphate. By convention, respiratoryfailure is used in a clinical context to mean failureof ventilation and/or pulmonary gas exchange.

Respiratory support in a transplant patientmay be required immediately postoperativelyor at a later occasion for indications similar tothat of the general population.

The principle of the management of respira-tory failure should begin primarily with thediagnosis and treatment of the underlyingpathology, often simultaneously with respira-tory support.

Definition

Type 1This is defined as hypoxemia without hyper-

capnia, indeed the CO2 level may be normal orlow. It is typically caused by a ventilation/perfu-sion mismatch; the air flowing in and out of thelungs is not matched with the flow of blood tothe lungs.

This type is caused by conditions that affectoxygenation like:

• Parenchymal disease

• Diseases of vasculature and shunts.

4 Artificial Organs

Type 2This is defined as build up of carbon dioxide

from energy utilization. The underlying causesinclude

• Reduced breathing effort (in the fatiguedpatient)

• Increased resistance to breathing (such as inasthma)

• A decrease in the area of the lung available forgas exchange (such as in emphysema).

There are of course situations in which bothmechanisms coexist in varying degrees such asasthmatics who develop chest infections.

Causes of Respiratory Failure

Many of the causes (Table 1-3) have specifictreatments. Respiratory support may becomenecessary due to either continued deteriorationof the primary cause or development of acutelung injury. Table 1-4 outlines some factorsknown to precipitate lung injury.

Mortality from acute respiratory failure isapproximately 40%, but rises significantly ifassociated with failure of other organs.

Management of Respiratory Failure

Treating the underlying cause must be the prior-ity. Ventilation, either non-invasive or invasive,merely supports the respiratory function of the

lungs while they are recovering from the pri-mary pathology.

Initial supportive management may consist ofsupplemental oxygen progressing to non-invasiveventilation. The effectiveness of NIV will dependon the underlying condition and is now consideredas a first-line intervention in patients with COPDexacerbation. In acute cardiogenic pulmonaryedema, NIV either as continuous positive airwaypressure (CPAP)or as Bi-level positive airway pres-sure (BiPAP) will improve gas exchange, reducerespiratory rate, and the need for intubation. Itsuse in other causes of respiratory failure (pneumo-nia, asthma, acute lung disease, and neuromuscu-lar disease) are less well defined , but at the veryleast may provide time to prepare for intubation ormore definitive treatment.

The indications for intubation and mechan-ical support are also difficult to describe as theydepend not only on the degree of hypoxia and/orhypercarbia but also on the clinical observa-tions. These include trends in respiratory rate,use of accessory muscles, ability to cough andtalk, and changes in mental state.

Once ventilated, the emphasis should continuewith treatment of the predisposing cause of respira-tory failure, as well as avoidance of ventilatorinduced lung injury and ventilator-associated pneu-monia. The causes of respiratory failure (Table 1-3)may also lead to acute lung injury directly or as aresult of prolonged mechanical ventilation.

Acute Lung Injury

Definition

Acute lung injury and its more severe form, acuterespiratory distress syndrome (ARDS) are aninflammatory response of the lung. The insult

Table 1-3. Causes of Respiratory Failure

Pulmonary dysfunctionAsthmaChronic obstructive pulmonary diseasePneumoniaPneumothoraxHemothoraxAcute lung injury (ALI) or acute respiratory distress syndrome (ARDS)Cystic fibrosis

Cardiac dysfunctionPulmonary edemaValve pathology

OtherFatigue due to prolonged tachypnoea in metabolic acidosisIntoxication with drugs (i.e., morphine, benzodiazepines)

suppresses respiration.ComaNeuromuscular disease

Table 1-4. Precipitating Conditions Relevant to TransplantMedicine

Direct Indirect

Pneumonia Massive transfusionLung contusion SepsisAspiration of gastric contents PancreatitisFat embolism Cardiopulmonary bypassToxic inhalation BurnsPulmonary surgery Bone marrow transplantReperfusion injury Drugs and toxins


may be direct (pulmonary) or indirect (non-pulmonary) and the result is characterized by:

Acute onsetBilateral infiltrates on chest radiographClinical or measurable absence of left atrial

hypertensionHypoxemia: Pa O2/FiO2 of <40 kPa for ALI

and < 26.7 kPa for ARDS

Epidemiology

The reported incidence of ALI is variable. A recentpopulation-based study from Washington showedan incidence of acute lung injury of 78.9 per 100,000person-years.. The mortality does appear to bedeclining. In the late 1980s, the reported mortalitywas around 60–70% (1995). The KCLIP studyshows the in-hospital mortality rate for ALI of38.5% and ARDS of 41%. The incidence of acutelung injury increased with age from 16 per 100,000person-years for those 15 through 19 years of age to306 per 100,000 person-years for those 75 through84 years of age. Mortality increased with age from24% for patients 15 through 19 years of age to 60%for patients 85 years of age or older. It is estimatedthat there are 190,600 cases of acute lung injury inthe United States which are associated with 74,500deaths and 3.6 million hospital days.

There is no data regarding the incidence ofALI or ARDS following artificial organ implan-tation, although it seems likely the incidence isgreater as patients are more susceptible to manyof the causes, in particular infection and sepsis.

Pathogenesis

The initial insult to the lungs results in an inflam-matory response which causes alveolar depositionof fibrin and collagen, pulmonary microthrombi,and acute destruction of the alveolo-capillarymembrane leading to increased permeability andwidespread inflammatory alveolar exudate.28

There may follow a fibrotic response character-ized by repair and proliferation of fibroblasts andtype II pneumocytes.

The consequences of such changes arealveolar edema and collapse, reduced lungcompliance, increased venous admixture, andhypoxemia. There is also pulmonary vasocon-striction, which increases alveolar dead space,exacerbates the ventilation-perfusion mismatchand may lead to pulmonary hypertension and

increased right ventricular work with conse-quent failure.

Management

Continuous positive airways pressure (CPAP)may be of benefit in mild cases, however, mostpatients will require early intubation andmechanical ventilation. Indications includehypoxemic or hypercarbic respiratory failure,acidemia, exhaustion, and reduced or alteredconsciousness. Profound sedation is usuallyrequired for ventilation as struggling or cough-ing can cause loss of recruited lung and worsen-ing oxygenation. Paralysis may be necessary ifsedation alone does not settle the patient.

The aim of ventilation is to improve oxygena-tion without causing further damage to the lungs.Difficulties arise as some alveoli are normal andopen while others are stiff and collapsed. It istherefore necessary to try to open the collapsedalveoli without damaging the normal areas. Themain causes of ventilator-induced lung damageare over-distension, alveolar collapse, capillaryleak as well as cyclical opening and closing.

Lung Protective Ventilation

‘‘Lung-protective’’ ventilation strategies usesmaller tidal volumes and pressures in anattempt to protect the lung from both overdis-tension (VILI – ‘‘Ventilator Induced LungInjury’’) and dampen or prevent the release ofinflammatory mediators that may drive theinflammatory response. If respiratory acidosisdevelops through the use of low tidal volumes,this is often tolerated (‘‘permissive hypercap-nia’’). The ARDSnet conducted a multicenter,prospective, randomized, controlled trial from1996 to 1999 to determine whether the use of lowtidal volumes would improve clinical outcomesin ARDS. The trial compared ‘‘traditional venti-lation’’ (tidal volume of 12 ml/kg of predictedbody weight and a plateau airway pressure of<50 cm H2O) with ‘‘low tidal volume’’ ventila-tion (tidal volume of 6 ml/kg of predicted bodyweight and a plateau pressure of <30 cm H2O).The lower tidal volume group displayed a sig-nificantly reduced mortality rate, 31 versus 40%.The debate continues as to whether the low tidalvolumes were beneficial or the high tidal

6 Artificial Organs

volumes were harmful. Nevertheless it appearscommon practice to ventilate patients on lowertidal volumes (<10 mls/kg).

Fluid Strategy

The management of acutely unwell patients isfaced with many dilemmas and ALI is no excep-tion. ALI is characterized by increased pulmonaryvascular permeability and some would argue thatfluid restriction should decrease alveolar lungedema and improve oxygenation and ventilation.However, this may in turn render the patienthypovolemic, thus reducing cardiac output,oxygen delivery, and organ perfusion. Recentevidence showed restrictive fluid strategy wasassociated with improved gas exchange and ven-tilator free days without any detriment to otherorgans, although without any impact on mortal-ity. Guiding fluid therapy with the use of a pul-monary artery flotation catheter has also shownno benefit in a heterogeneous group of criticallyill patients. A simple titration of the leastamount of fluid required to maintain generalperfusion appears to be the favored strategy atpresent in ALI patients, particularly with otherorgan failures.

Positive End Expiratory Pressure

The issue of positive end expiratory pressure(PEEP) is also a source of varying results withlarge studies showing no benefit but othersreporting large beneficial effects on survival andlung physiology. These differences are, in addi-tion to nuances of trial design, probably alsoreflective of the difficulty in separating PEEPfrom other ventilatory interventions. There isgood evidence to suggest that certain patientswith ALI/ARDS have more recruitable lungs andthat these certainly are more responsive to PEEP.Current best practice is to set PEEP somewhereabove the lower inflection point of the pressur-e–volume curve and to then perform a PEEP trialin which PEEP is increased so long as there is nocorresponding increase in plateau pressure.

PEEP has been applied in ARDS and acutelung injury in an attempt to improve oxygena-tion and prevent lung shear-stress injury asso-ciated with the cyclical opening and closingof collapsed alveoli (‘‘atelectrauma’’). In theface of ARDS, most clinicians would apply a

PEEP of 5–10 cm H2O. However, it is importantto balance the beneficial effect of PEEP on arter-ial oxygenation and its adverse effects such ascardiovascular compromise and increased air-way pressures and over-distension. The ARDS-net group performed a randomized, controlledtrial to determine whether higher levels of PEEPwould improve clinical outcomes in patientswith ARDS. All patients received mechanicalventilation with a tidal-volume goal of 6 ml/kgof predicted body weight and an end-inspiratoryplateau-pressure limit of 30 cm H2O. Mean PEEPwas 8.3 cm H2O in the lower group and 13.2 cmH2O in the higher group. There were no signifi-cant differences between the groups in the mor-tality rate or number of ventilator-free days,ICU-free days, or organ-failure free days

Corticosteroids

Corticosteroids have no role in the acute man-agement of ARDS. However, a small trial in only25 patients (Meduri) suggested a survival benefitif high-dose methylprednisolone is given as atreatment for the later ‘‘fibroproliferative’’phase. Recently the ARDS Network performeda large multi-center randomized trial to repro-duce the previous findings. Although, overall,60-day mortality was similar between the twogroups, methylprednisolone was associatedwith significantly increased 60- and 180-daymortality rates among patients enrolled at least14 days after the onset of ARDS. Methylpredni-solone therapy increased the number of ventila-tor-free and shock-free days during the first 28days, in association with an improvement inoxygenation, with fewer days of vasopressortherapy. When compared with placebo, methyl-prednisolone did not increase the rate of infec-tious complications but was associated with ahigher rate of neuromuscular weakness. Despitethe improvement in cardiopulmonary physiol-ogy, the results of this study do not support theroutine use of methylprednisolone for persistentARDS. In addition, starting methylprednisolonetherapy more than 2 weeks after the onset ofARDS may increase the risk of death.

Inhaled Pulmonary Vasodilators

Inhaled nitric oxide can selectively vasodilatethe pulmonary vascular and reduce pulmonary


hypertension, decrease shunting, and improvegas exchange in ARDS. There are no systemiceffects because nitric oxide is scavenged rapidlyby hemoglobin. However, numerous large clin-ical trials have failed to demonstrate that theimprovements in oxygenation seen in patientstreated with nitric oxide translate into improvedoutcome. Current expert opinion suggests thatnitric oxide should not be used routinely but bereserved for patients in whom adequate oxyge-nation cannot be achieved by lung protectivemechanical ventilation and prone positioning.

Inhaled prostacyclin vasodilates the pulmon-ary bed as effectively as nitric oxide but does notresult in the same improvements in oxygena-tion. Again mortality benefits have not beenproven.

Prone Ventilation

The benefits of prone ventilation are explainedby an improvement in ventilation perfusionmatching, recruitment of atelectasis, and thedelivery of more homogenous ventilation to allparts of the lung. Two large randomized trialshave compared the effects of prone positioningon mortality from ARDS. In both trials, therewas a significant improvement in oxygenationwhen patients were turned prone. However,this did not translate into improved clinicaloutcome. The rates of displacement of endotra-cheal tubes, vascular catheters, and thoracotomytubes were similar in the two groups in one trialbut increased in the prone group in the othertrial. Consensus appears to be that prone posi-tioning is only useful in severe ARDS withdependent disease and refractory hypoxemia.

High-Frequency Oscillatory Ventilation

High-frequency oscillatory ventilation (HFOV)is, in theory, a ‘‘lung-protective’’ mode of venti-lation. HFOV provides efficient gas exchange byusing very low tidal volumes and high respira-tory rates. The application of a constant meanairway pressure in HFOV allows the mainte-nance of alveolar recruitment while avoidinglow end-expiratory pressure and high peak pres-sures. The mean airway pressure generatedby HFOV is usually higher than that generatedby conventional ventilation (tidal volumes of10 ml/kg). A randomized controlled comparing

HFOV with a conventional ventilation strategyin 148 adults with ARDS confirmed that HFOV iseffective and safe and not associated with sig-nificant hemodynamic effects. However, therewas no significant difference in mortalitybetween the groups. One of the limitations ofthis (and almost all of the other older studies ofventilation strategy) was that HFOV was notcompared with the current gold standard low*-tidal volume approach used by the ARDSnettrial.

Extracorporeal Support

Extracorporeal membrane oxygenation (ECMO)proposes to provide the gas exchange via veno-venous bypass while maintaining the lungs atrest. The main advantage is that the lungs arenot exposed to injurious ventilatory strategies.Again there have been no advantages demon-strated in randomized clinical trials thus far andthe technique is only contemplated in refractoryhypoxemia in specialist units. Retrospective sin-gle center series reported 67% rate of patientsweaned of ECMO with a 52% hospital survival(Hemmila. Ann Surg. 04;240:595–607). Resultsof a prospective controlled study (CESAR Trial)are awaited. Preliminary results suggest a rela-tive risk (survival or severe disability) in favor ofECMO in adults with severe acute respiratoryfailure of 0.69. Modifying the technique topumpless extracorporeal carbon dioxide whileemploying a fully protective lung ventilationstrategy may be of benefit where hypercapnia isthe major manifestation of ARDS.

In summary, there is good evidence to sup-port the following ventilation strategy: low tidalvolumes of 6 ml/kg and plateau pressures nothigher than 30 cm H2O, appropriate PEEP asdiscussed, target arterial oxygen saturations of90%, pH above 7.2 and judicious fluidadministration.

Disadvantages of Ventilation

Although mechanical ventilation is regardedas life-saving support, there are many potentialcomplications including pneumothorax, airwayinjury, alveolar damage, and ventilator-associatedpneumonia, among others. To tolerate endotra-cheal intubation patients need to be sedatedwhich in turn has many unwanted effects such

8 Artificial Organs

as hypotension and gastro-intestinal stasis. Posi-tive pressure ventilation reduces venous returnand increases afterload of the right ventriclewhich in turn may fail and lead to cardiovascularcollapse. Patients should be considered for wean-ing from ventilation at the earliest opportunity

Non-invasive Ventilation (NIV)

This is the use of positive pressure ventilationthrough specially designed face masks ratherthan endotracheal tubes. The two most widelyused ventilatory modes are continuous positiveairway pressure (CPAP) and bilevel positive air-way pressure (BiPAP). It is in general less effectivein delivery of pressure ventilation than invasive(endotracheal) ventilation but requires less or nosedation and can be administered in properlyequipped units outside intensive care.

It is potentially advantageous in cardiogenicpulmonary edema, a condition in which manyclinicians find it useful to undertake a’’ trial ofCPAP’’ accompanied by conventional pharma-cological afterload reduction and diuresis.Recently, a meta-analysis pointed to a reductionin the need for endotracheal intubation andmortality in this group of patients.

Some clinicians advocate non-invasive venti-lation as a useful tool to facilitate early extuba-tion. Indeed a randomized controlled trial inpredominantly medical patients showedreduced length of ventilation, intensive carestay, tracheostomy requirement, and moralityin patients receiving NIV.

NIV is widely established as an importantcomponent of the treatment of acute exacerba-tions in chronic obstructive pulmonary disease.Evidence from pooled randomized controlledtrials points to improvements in mortality,length of stay, and intubation rate with earlyintroduction of NIV in addition to usual medicaltreatment. Many groups also use NIV in thetreatment of obesity hypoventilation andobstructive sleep apnea syndromes.

NIV in acute lung injury is neither favored bymany clinician nor supported by evidence.

Weaning

Withdrawal from mechanical ventilation shouldbe considered at the earliest opportunity, butonly if the patient can support there own

ventilation and oxygenation. There are severalobjective parameters to look for when consider-ing withdrawal, but there is no specific criteriathat generalizes to all patients. The objectivecriteria include

• Adequate oxygenation (e.g., 8 kPa on FIO2 >0.4; PEEP <5–10 cm H2O; PO2/FIO2

>300 mmHg (40 kPa))

• Stable cardiovascular system (e.g., HR <140;stable BP; no (or minimal) pressors orinotropes

• No significant respiratory or metabolic acidosis

• Adequate conscious level (e.g., arousable,GCS >13, obeying commands, no continuoussedative infusions)

• Adequate cough to clear respiratory secretions

• The acute disease process that led to thepatient needing mechanical ventilation mustalso have been reversed or resolved.

Finally, weaning is a complex and disputed areaof critical care medicine in which, however, thereis evidence of benefit from a multidisciplinaryapproach, consistency and simplicity, aggressivetreatment of infection and heart failure, nutrition,daily weaning trials, and persistence.

Conclusion

The causes of respiratory failure are multiple.Each will have its own specific treatment, butadvances in both non-invasive and invasive ven-tilatory support have had a significant impact onthe outcome of such a life-threatening condi-tion. The most notable benefit is time for thediagnosis and treatment of the underlying con-dition to occur.. ALI or ARDS as a cause ofrespiratory failure or consequence of the under-lying illness and the need for mechanical venti-lation also has a dramatic impact on outcome.Despite greater understanding of the cellularand molecular mechanisms, specific therapiesdo not exist. However, much has been learntfrom salvage therapies and the importance ofprotective ventilation strategies.

Acute Renal Failure

Epidemiology

Acute renal failure (ARF) is a common compli-cation following major surgery. After major


surgery, incidences can range from 10% foruncomplicated non-cardiac solid organ trans-plants and artificial organs to 50% in cardiactransplantation and extracorporeal circulatorysupport.

Its presence carries a high mortality – rangingfrom 15% with simple pre-renal azotemia togreater than 60% in those requiring renal repla-cement therapy (RRT).

The most common etiology of ARF in thecritically ill transplant patient is in the context ofmulti-organ failure. The initial insult is usuallypre-renal, with renal hypoperfusion occurringfrom either vasodilation in severe sepsis or hypo-volaemia during the peri and immediate post-operative period, e.g., intra-operative hemorrhageor marked extracellular fluid shifts.

Progression from this initially reversible risein serum creatinine and urea concentrationsleads to acute tubular necrosis (ATN). Pre-renal azotemia and ischemic ATN constitute acontinuum of the same pathophysiological pro-cesses, and together account for 75% of the casesof acute renal failure. Intrinsic damage fromrenal disease accounts for around 30% of casesin the post-transplant patient (Table 1-5) and ismost likely due to chemical/pharmacologicalinjury. Foremost among these are radiocontrastmedia, nephrotoxic antibiotics, immunosup-pressives, and non-steroidal anti-inflammatorydrugs.

Obstructive renal failure is less common(10% approximate incidence) in critically illpatients and may follow complications asso-ciated with abdominal surgery.

Establishing the underlying cause of acuterenal dysfunction in this group of patients isvitally important for the rapid initiation of effec-tive treatment, maximizing the prospect ofreversing any acute renal disease. It is alsoimportant for prognostic reasons as mortalityin acute tubular necrosis can be as high as65%, compared with 5–20% for those patientswith ARF as part of a multisystem disease.

Risk Factors

The most commonly identified risk factors forthe development of renal failure are impairedpreoperative renal function. Transplant and arti-ficial organ patients also suffer from other factorsknown to be associated with renal failure, such ashypertension, hyperlipidemia, atherosclerosis,and diabetes. Patients with pre-existing renaldysfunction are at highest risk of developingpostoperative renal failure. Renal dysfunction isparticularly common in patients awaiting livertransplantation (and/or receiving artificial liversupport) and is well documented as an importantdeterminant of outcome. In these patients, quan-tification of the severity of renal function iscomplicated by the impact of hepatorenal dys-function. This is usually present to some degreeby the time these patients require transplanta-tion, even when it has not progressed to the fullblown hepatorenal syndrome (HRS). None of thestudies of long-term renal function in thesepatients has factored in the potential impact ofpost-transplant HRS from poor liver graft func-tion and only a few have segregated patientswith recurrent hepatitis C or hepatitis B, whomay additionally have hepatitis-associatedglomerulonephritis.

The measurement of serum creatinine, how-ever, seems to have a limited ability to identifypatients with preoperative renal dysfunctionbecause of its variation with age, sex, and musclemass. Calculated creatinine clearance as analternative measure of renal function is probablymuch better at estimating renal reserve, therebyidentifying those at greater risk of needing post-operative renal replacement therapy.

Postoperative renal failure requiring (RRT)seems to increase markedly when creatinine

Table 1.5. Causes of Acute Renal Failure in the Critically ill PostTransplant Surgery Population [After Cosgrove et al. 2007 Ann R CollSurg Engl. 89:22–29]

Pre-renal• Hypotension; sepsis• Hypovolemia: dehydration, blood loss, acute pancreatitis• Reduced cardiac output: cardiac tamponade• Intra-abdominal compartment syndromeRenal• Acute tubular necrosis• Drugs: aminoglycosides, non-steroidal anti-inflammatory drugs,

radiographic contrast medium• Toxins: endotoxins• Hepatorenal syndrome• Intra-abdominal compartment syndrome• Pigment nephropathy (myoglobinuria): drug-induced

rhabdomyolysisPost-renal• Ureteral obstruction: stones, surgical,• Bladder dysfunction: surgical• Urethral obstruction: traumatic/prostatic hypertrophy

10 Artificial Organs

clearance decreases below 60 ml/min, even ifthe serum creatinine lies in the normal range.Thus, the incorporation of this threshold intothe preoperative assessment may help to identifyhigh-risk patients for the early institution ofaggressive renal protective therapies, such asvolume expansion, use of vasopressors, and/orspecific pharmacological therapies.

Age-related predisposition to acute renal fail-ure arises due to a combination of reduced renalreserve, susceptibility to volume depletion, anda reduced ability of the kidney to conserve saltand concentrate urine. Body mass index (BMI)has also been identified as a preoperative pre-dictor of ARF, independent of associated comor-bidities such as diabetes and hypertension.

Pathophysiology

Pre-renal Dysfunction and ATN:Renovascular Hemodynamics

Pre-renal disease arises due to either a true hypo-volaemia (e.g., acute haemorrhage) or a drop inthe effective circulating volume (e.g., fall in car-diac output, widespread systemic vasodilation, orintrarenal vasoconstriction). It arises because,although the normally high renal blood flowexceeds the organ’s usual oxygen requirements,only 10% of this passes to the renal medulla,where the PaO2 is only 1.2–2.2 kPa. These verymetabolically active medullary cells are thereforemarkedly susceptible to underperfusion and con-sequent cellular hypoxia.

Autoregulation

Initially the kidney responds by maintaining GFRand renal blood flow to a relatively constant level,through autoregulation. Net filtration pressureand renal perfusion pressure are preservedthrough an angiotensin II-mediated constrictionof the efferent arteriole, accompanied by an intra-luminal myogenic compensatory vasodilation ofthe afferent pre-glomerular arteriole, principallythrough the paracrine actions of prostaglandins.The net result is maintenance of a near constantrenal blood flow within a wide range of meanarterial pressures, from around 60 to 160 mmHg.

Drugs that interfere with this autoregulatoryprocess include antagonists of the renin angiotensin

system and cyclo-oxygenase inhibitors such as non-steroidal anti-inflammatory drugs. Another impor-tant group of drugs that cause acute constriction ofthe afferent arteriole after their first postoperativedose is calcineurin inhibitor immunosuppressives(CNI), principally ciclosporin and tacrolimus.These agents cause an acute reduction in renalperfusion by interfering with autoregulation, theeffects of which are particularly apparent in situa-tions of hypovolaemia or reduced cardiac output.This is particularly relevant in the immediate post-operative period, where the altered physiologicalmilieu depends on this autoregulatory response.

Postoperative sepsis also directly and specifi-cally interferes with renal hemodynamicsthrough the release of pro-inflammatory cyto-kines. These interfere with autoregulationthrough stimulation or antagonism of variousmediators, resulting in glomerular endothelialdysfunction and consequent aggravated renalischemia.

Despite this, pre-renal dysfunction is oftenreversible if the factors causing the renal hypoper-fusion are corrected. The severity and duration ofrenal ischemia which causes fixed post-ischemicATN is not well established and variable givendifferent patient characteristics, but there isusually an interval that defines the transitionfrom functional pre-renal to post-ischemic ARF.During this early functional stage, vigilant recog-nition and aggressive optimization of renal bloodflow, through volume expansion and/or cautioussystemic vasoconstriction, can restore renal func-tion to normal.

Histological Changes

Histologically, ischemia-induced renal failure inanimals includes loss of the proximal tubularbrush border, blebbing of apical membranes,and cellular and mitochondrial swelling. Withadvanced injury, tubular cells detach from thebasement membrane and contribute to intra-luminal aggregation of cells and proteins leadingto tubular obstruction.

It is not clear though, whether this modelapplies to the critically ill. It has been postulatedthat in sepsis, cellular malfunction arises throughimmune-mediated cellular apoptosis. However,post-mortem evidence does not support this,nor do situations where recovery of renal func-tion ensues within days, far quicker than thatrequired for regeneration of renal tubular cells.


An alternative explanation may be that insepsis cellular energetics are altered due toimpaired mitochondrial respiration – a failureof oxygen utilization rather than delivery. Thismodel fits well with the relative paucity of his-tological tubular changes in septic patients andmay account for the occasionally rapid restora-tion of renal function seen in some critically illpatients.

Renal Optimization After Artificial OrganImplantation

The importance of maintaining adequate hydra-tion and renal perfusion cannot be overstated.The most effective prevention of ARF is stabili-zation of systemic hemodynamics, optimizationof cardiac output and blood pressure and theavoidance of specific risk factors such asnephrotoxic drugs – ensuring meticulous main-tenance of immunosuppressive and antibioticlevels. When possible, a CNI should be withhelduntil renal function has improved, particularlyin patients with renal failure from HRS or con-gestive cardiac failure.

Obstruction to the lower urinary tract shouldalways be excluded and, in postoperative abdom-inal surgical patients, intra-abdominal hyperten-sion must be considered a contributing factor.

Early recognition and effective managementof any acute precipitant can minimize its sever-ity, allowing natural reversal of the pathophy-siology. This may include control of surgicalbleeding, radiological, or surgical drainage of aseptic focus, relief of ureteric and/or urethralobstruction and abdominal decompression totreat abdominal compartment syndrome.

Indeed, the evolution of clinical ATN hasbeen divided, rather arbitrarily, into four stages:(1) initiation, (2) extension, (3) maintenance,and (4) recovery. It is suggested that most ofthe preventative interventions during ATNshould be active during the extension phase.

During the extension and maintenancephases of established ATN, a variety of interven-tions and drugs have been used either to preventinjury or to hasten recovery of ischemic and/ortoxic ATN (e.g., endothelin receptor blockers,growth factors adenosine and adenosine antago-nists, nitric oxide synthase inhibitors). Althoughsome of these interventions are effective in alter-ing the course of experimental ATN, all but afew have shown clinical benefit. Unfortunately

overall treatments remain largely supportive,with the exception of N-acetylcysteine in radio-contrast-induced renal failure and activatedprotein C in sepsis. Supportive stabilization ofgeneral hemodynamics and optimization of car-diac output are best achieved in critical care withvolume expansion, usually followed byvasopressors.

Volume Expansion

Initially volume expansion aimed to achievesupranormal cardiac index values comparableto those shown by survivors and normal valuesfor mixed venous oxygen saturation. Morerecently, however, less severe organ dysfunctionhas been demonstrated with early institution oftreatment to increase central venous oxygensaturation to 70% or higher as well as avoidanceof unnecessarily high increases in cardiac work,especially in individuals observed to havedecreased cardiopulmonary reserve.

The importance of volume therapy andhydration may also help to prevent radiocon-trast media and amphotericin-induced ATN, aswell as drug-induced intrarenal tubular precipi-tation of crystals following high doses of metho-trexate, sulfonamides, or acyclovir.

Vasopressors

When appropriate volume expansion fails torestore adequate mean arterial pressures, vaso-pressors are indicated. Although there is concernthat these may cause intrarenal vasoconstrictionand counter any beneficial effects of increasedblood pressure, vasopressor administration doesimprove renal blood flow, especially in patientswith sepsis characterized by systemic vasodila-tion and impaired autoregulation.

Diuretics

Loop diuretics are commonly used in volumecontrol and boast the theoretical benefit of redu-cing medullary oxygen demand and protectingagainst ischemic injury. Beyond experimentalmodels, however, renal protection has neverbeen clinically demonstrated. They do not accel-erate renal recovery, reduce the need for dialy-sis, or decrease mortality and although some


studies have suggested an increase in mortality,it is likely that they are of no harm or benefit butmight be useful in the management of fluid over-load. In some patients, they may convert oliguricto non-oliguric renal failure, the latter beingassociated with a better outcome. It is likely inthis situation that patients responding to loopdiuretics are characterized by a less severe formof renal failure, rather than a direct beneficialeffect of diuretic therapy – offering merely prog-nostic information rather than definitive treat-ment. Their use, however, should not delayinitiation of renal replacement therapy by pro-ducing a ‘‘false’’ diuresis in cases of firmly estab-lished dialysis-dependent renal failure.

Renal Replacement Therapy

Physiological Functions of the Kidney andClinical Measurement

Many renal functions are shared with otherorgans (e.g., acid–base balance with the lung,calcium regulation with the parathyroid glands,hemoglobin control with the hemopoietic sys-tem) and therefore can be monitored. In con-trast, other functions are exclusive to the kidneybut not routinely measured in the critically ill.There are only two functions routinely mea-sured that are ‘‘unique’’ to the kidney andwhich are considered clinically important: pro-duction of urine and the excretion of water solu-ble waste products of metabolism. Thus, despiteseveral existing definitions of ARF, which tendto focus on individual factors such as biochem-istry, pre-existing renal impairment, nephro-toxic drugs, and pathophysiology, clinicianshave focused on the two aspects of renal func-tion that aim to encompass all these commonelements: serum creatinine and urine output.

Assessment of Renal Function – the RIFLECriteria

Urine output and serum creatinine are virtuallyunique to the kidney and are easily measured.Other biochemical criteria such as raised urea,hyperkalaemia, and acid–base disturbance aremore open to influence from extra-renal factors,e.g., gastrointestinal, respiratory, and endocrinedisturbances. Limitations of using the serumcreatinine rather than creatinine clearance in

identifying patients with pre-existing renal dis-ease have already been discussed, as no singlecreatinine value corresponds to a given GFR inall patients. It is, however, the change in creati-nine that is useful in determining the presence ofARF and determining the exact GFR is rarelynecessary for clinical purposes. Instead it isimportant to determine whether renal functionis stable or changing and this can usually bedetermined by monitoring serum creatininealone.

Urine output is far less specific for renal fail-ure, except when severely reduced or absent;severe ARF can exist despite normal urine out-put – i.e., non-oliguric ARF. It can also varydepending on factors extrinsic to the kidney,such as post-renal tract obstruction and the useof loop diuretics which, as mentioned above,may convert a non-oliguric ARF to an oliguricrenal failure, at the expense of intravasculardepletion unless this possibility is recognizedand treated.

With this in mind, the RIFLE system hasformed a new consensus classification from thesecond International Consensus Conference ofthe Acute Dialysis Quality Initiative (Table 1-6).

RIFLE stands for Risk of renal dysfunction,Injury to the kidney, Failure of kidney function,Loss of kidney function, and End stage renaldisease.

This is a multilevel classification that helpsto stratify and define ARF, where patients canfulfil the criteria through changes in serum crea-tinine, urine output, or both. It permits the

Table 1-6. The RIFLE Classification for Acute Renal Failure [AfterBellomo et al. Acute Renal Failure – Definition, Outcome Measures,Animal Models, Fluid Therapy and Information Technology Needs:the Second International Consensus Conference of the Acute DialysisQuality Initiative (ADQI) Group. Crit Care. 2004;8:R202–12]

GFR CriteriaUrine outputcriteria

Risk Increased creatinine� 1.5 frombaseline

25% reduction in GFR

<0.5 ml/kg/hover 6 h

Injury Increased creatinine � 2 frombaseline



Failure Increased creatinine � 3 frombaseline



Loss Persistent ARF. Complete loss of kidney function for>4weeks

ESRD End-stage kidney stage (> 3 months)


incorporation of a wide spectrum of disease thatis readily applicable across a wide patient group.

Indications for Renal Replacement Therapy

Common renal and non-renal indications forrenal replacement therapy (RRT) after artificialorgan implantation are listed in Table 1-7. In thesetting of oliguria, fluid overload, deranged elec-trolytes and acidaemia RRT simplifies fluid andnutritional management and its early use mayimprove outcome.

However, establishing an extracorporeal circuitis not without risks (bleeding, hypotension,hypothermia, inflammatory response, line infec-tions); in addition, it should not be regarded sim-ply as replacement of all kidney functions. Indeed,other than removal of metabolic waste products,hydrogen ions, fluid balance, and other functionsof the kidney are not replaced. ARF requiring RRTin intensive care patients carries a mortality of40–70% compared with around 10% in similarcritically ill patients not requiring RRT. Thisincreased mortality is in fact due to non-renalcauses such as bleeding, sepsis, and respiratoryfailure, suggesting that normal kidney functionhas additional systemic effects that are notreplaced by RRT, the pathological consequencesof which are shown in Table 1-8. Thus, ARF carries

an independent risk of death, and patients may bedying ‘‘of ’’ renal failure rather than ‘‘with’’ renalfailure. The kidney is thought to play a crucial rolein inflammation and the evolution of distal organinjury particularly in the context of postoperativesepsis and multi-organ dysfunction (Table 1-9).

Hence if ARF develops, a vicious cycle ofcytokine release, dysfunction of distal organs,failure of urinary cytokine clearance, and furtherdistal organ dysfunction ensues. As mentioned,often there exists a window of opportunitybetween the potentially reversible pre-renalphase and fully established ATN. In thosepatients at risk of developing ARF either pre-operatively or postoperatively, preventativemeasures should be initiated early to avoid pro-gression of renal dysfunction and its associatedpoor outcome.

Taking into account the hazards and prog-nostic implications of RRT, and the possiblebenefits of its early commencement, there isstill no agreed consensus for the optimal pointat which to commence filtration. It is hoped,however, that application of the RIFLE criteriamay in future aid to the development of specificguidelines.

Introduction to Renal Replacement Therapy

Renal replacement therapy (RRT) describes abroad spectrum of treatments that ultimatelyaims to take over fluid control, acid–base bal-ance, and metabolic waste product removal.Treatments involve establishing an extracorpor-eal circuit via an artery or vein, directing bloodflow through a filtration system that removeswaste solutes, water and other molecules beforebeing fed back to the patient. This filtration

Table 1-7. Indications for Continuous Renal Replacement Therapy

Renal Possible non-renal

Uremia SepsisHyperkalemia Multiple organ dysfunction syndromeMetabolic acidosis Pulmonary edemaFluid overload Raised intracranial pressure

Temperature control

Table 1-8. The Pathological Consequences of Acute Renal Failure

Cardiovascular Hyperdynamic circulation, cardiomyopathy,pericarditis

Pulmonary Pulmonary edema, alveolitis, pneumonia,pulmonary hemorrhage

Gastrointestinal Motility impairment, erosions, ulcerations,hemorrhage, pancreatitis, colitis

Neuromuscular Neuropathy, myopathy, encephalopathyImmunological Impaired humoral and cellular immunityHematological Anemia, thrombocytopenia, bleeding diathesisMetabolic Insulin resistance, hyperlipidemia, increased

protein catabolism, depletion of antioxidants

Table 1-9. Pathophysiological and Immunological Effects of AcuteRenal Failure, and Renal Replacement Therapy

The injured kidney: a pro-inflammatory mediator• Increased release/impaired catabolism and elimination of

cytokines• Activation of immunocompetent cells• Release of factors promoting distal organ injuryFactors mediated by renal replacement therapies• Hemodynamic factors• Loss of nutrients (amino acids/antioxidants)• Induction of an inflammatory reaction


system may be driven by a pressure gradient(hemofiltration) or by a concentration gradient(hemodialysis).

Vascular Access

Most current RRT is veno-venous via a large boredouble lumen central venous cannula, with bloodflow through the circuit achieved by a peristalticroller pump. Insertion sites are largely deter-mined by the presence of other central venouscatheters already in situ. Subclavian lines havehad stenosis rates quoted of up to 90%, which ispotentially problematic if formation of an arterio-venous fistula is being considered for long-termrenal replacement. The internal jugular permitsgreater patient mobility but is associated withgreater immediate complications. The femoralroute is useful for short-term renal replacementbut is associated with increased rate of infections.

Anticoagulation

All patients requiring renal replacement therapyneed suitable anticoagulation unless there issignificant coagulopathy or thrombocytopenia.This is usually in the form of unfractionatedheparin, delivered via the extracorporeal circuitrather than systemically to the patient, and isadministered upstream of the filter. Prostacyclinis useful in cases of heparin-induced thrombo-cytopenia and other agents employed includelow molecular weight heparin, sodium citrate,and nafamostat mesilate.

Adverse Effects of RRT

The most common complications of RRT arerelated to the extracorporeal circuit, vascularaccess, and the consequences of filtration.These are summarized in Table 1-10.

Principle of Hemofiltration

Ultrafiltration is the passage of fluid underpressure across a semi-permeable membranewhere solutes are carried along with the fluidby solvent drag (convection). Synthetic bio-compatible membranes are made of polyacry-lonitrile, polysulfone, polyamide, or cellulosetriacetate. These cause minimal complement orleucocyte activation. Fluid balance is maintainedby the simultaneous re-infusion of a sterile crys-talloid replacement fluid containing essentialplasma electrolytes. Replacement fluid can beadministered before the filter (pre-dilution) or,more usually, after the filter (post-dilution). Pre-dilution may be useful when high filtration ratesare required (10 l/day) or when filter clotting is aproblem as it reduces blood viscosity and subse-quent clotting in the circuit. However, it maydecrease the efficiency of the system as theblood being filtered contains a lower concentra-tion of metabolic waste products.

Hemofiltration is very effective in the removalof fluid and middle-sized molecules. The latterhas been suggested as an advantageous feature inthe treatment of sepsis, as most of the pro-inflam-matory cytokines may be theoretically removedby this technique.

Dialysis is the removal of solutes by diffusionacross a semi-permeable membrane down a con-centration gradient. The dialysate fluid is of lowosmolaltity, and blood and dialysate fluid arecirculated in a ‘‘counter-current’’ fashion. How-ever, unlike hemofiltration, no replacementfluid is re-infused. Treatment is intermittent,with large volumes of fluid being dialyzedin a period of a few hours. Blood flow isusually 150–300 ml/min, with dialysis flows at1000–2000 ml/h.

Continuous Venovenous Hemodiafiltration(CVVHDF)

This is the most popular mode in the ICU andprovides solute removal by diffusion and con-vection simultaneously. It offers high-volumeultrafiltration using replacement fluid, withadditional dialysate fluid being passed throughthe hemofilter in a counter-current fashion toblood flow. This improves the efficiency andsolute clearance rate, removing both small andmiddle molecules.

Table 1-10. The Common Complications of RRT

Extracorporealcircuit

Hypothermia, complement activation,hypotension, anemia

Vascular access Air embolism, clotting and hemorrhage,infection

Consequences offiltration

Electrolyte loss (# Na+, #Ca2+#Mg2+#K+),arrhythmias, hypovolaemia, metabolicalkalosis, removal of therapeutic drugs


Optimal Treatment Modalities for RRT

Continuous renal replacement therapy constitu-tes a diverse range of treatments offering variousways of providing RRT that may influence effi-cacy and safety. These can be separated intointermittent or continuous modalities.

Intermittent hemodialysis, although widelyused in the United States, generates huge osmo-tic changes, fluid shifts, and consequent hypo-tension – often to the detriment of renal bloodflow, potentially worsening renal ischemia. Tocompensate, an increase in vasopressor andfluid administration may be required, and thelatter can result in greater fluid accumulationthan before treatment.

Most physicians intuitively adopt continuoushaemofiltration which causes less intravascularvolume changes as time is allowed for fluid tore-equilibrate between body compartments. Thisis likely to be relevant in the postoperative trans-plant patient, in particular those with vasodilatoryshock and capillary leak where sudden, largevolume fluid movement between compart-ments can create life-threatening haemody-namic changes (e.g., sepsis, acute pancreatitis,and hepatic failure). There is, however, limitedevidence of the perceived superiority of contin-uous versus intermittent techniques. It is likelythat, in the absence of haemodynamic instabil-ity, the two modalities are probably equivalent.

Extracorporeal Inflammatory MediatorRemoval

One perceived advantage of haemofiltration inthe critically ill is clearance of middle-sizedmolecules within the range of the various pro-inflammatory cytokines involved in sepsis andthe systemic inflammatory response syndrome,not uncommonly seen in postoperative trans-plant and artificial organ implantation patients.

This theory, however, rests heavily on theassumption that cytokines can be effectivelycleared and that the non-specific removal ofsuch mediators is actually beneficial to thepatients. Despite numerous studies, there is lim-ited evidence to support this. Some cytokinesare removed by haemofilters –albeit in smallamounts – largely through adsorption, but theefficiency of this process wanes rapidly, early onin the filter’s life span. Given the high turnoverof endogenous inflammatory mediators and

required high filtration rates to make mediatorremoval significant, its efficacy is questionable.Nevertheless the theory of middle moleculeclearance carries significant weight in the treat-ment of sepsis and ARF and may have a role toplay in the future.

Optimal Dose of Haemofiltration

Higher treatment doses of RRT have been asso-ciated with improved survival. In particular cri-tically ill patients with acute renal failure hadbetter outcome at 15 days post-RRT when trea-ted with filtration rates of 35 ml/kg/h whencompared with 20 ml/kg/h. Interestingly, a simi-lar study looking at early initiation of highvolume haemofiltration (HVHF), comparedwith the lower dose found no significant differ-ence in mortality. Preliminary experience ofearly initiation of RRT and HVHF is promising,but prospective randomized comparisonsbetween different timing of RRT seem warrantedbefore one can definitively be recommendedover the other.

Choice of Anticoagulation

In the absence of coagulopathy, unfractionatedheparin given either intravenously or adminis-tered via the extracorporeal circuit is the mostcommon choice of anticoagulation. Only citratehas been shown to be more effective than heparinin terms of filter life and bleeding complications,with some studies showing low molecular weightheparins being associated with an increased riskof hemorrhage, whereas others demonstrate itssafe use and longer filter life span.

Table 1-11. Current Best Practice Guidelines of Renal ReplacementTherapy in Critical Care

Mode of RRT CVVHDF/CVVHDOnset (early/late) EarlyType of membrane

(synthetic/traditional)

Synthetic biocompatible membranes(polyacrylonitrile, polysulfone,polyamide)

Dose of filtration (35/20 ml/kg/h)

�35 ml/kg/h

Anticoagulation Unfractionated heparin or sodiumcitrate

Vascular Access Internal jugular or *femoral*short-term use only


Current Best Practice Guidelines

In summary, there are many modes, treatments,and doses of renal replacement therapy avail-able. Although current evidence is limited,there are still certain aspects that show morepromising preliminary outcomes than others.However, the diversity of worldwide RRT prac-tice is not aligned with best evidence throughoutthe many different centers, which may contri-bute to additional morbidity/mortalityobserved. The current recommended best prac-tice guidelines are summarized in Table 1-11.

References

1. Weil MH, Shubin H. Proposed reclassification of shockstates with special reference to distributive defects. AdvExp Med Biol. 1971; 23: 13–23

2. Choi PTL, Yip G, Quinonez LG, Cook DJ. Crystalloidsvs. colloids in fluid resuscitation: A systematic review.Crit Care Med. 1999;27:200–210

3. Cook D, Guyatt G. Colloid use for fluid resuscitation:Evidence and spin. Ann Intern Med. 2001;135:205–208

4. Schierhout G, Roberts I. Fluid resuscitation with colloidor crystalloid solutions in critically ill patients: Asyste-matic review of randomized trials. BMJ. 1998;316:961–964

5. Cook D, Guyatt G. Colloid use for fluid resuscitation:Evidence and spin. Ann Intern Med. 2001;135:205–208

6. Schortgen F, Lacherade JC, Bruneel F, Cattaneo I, HemeryF, Lemaire F, Brochard L. Effects of hydroxyethyl starchand gelatin on renal function in severe sepsis: a multi-centre randomised study. Lancet. 2001;357:911–916

7. Sakr Y, Payen D, Reinhart K, Sipmann FS, Zavala E,Bewley J, Marx G, Vincent JL. Effects of hydroxyethylstarch administration on renal function in critically illpatients. Br J Anaesth. 2007;98:216–224

8. Finfer S, Bellomo R, Boyce N, French J, Myburgh J,Norton R. SAFE Study Investigators. A comparison ofalbumin and saline for fluid resuscitation in the inten-sive care unit. N Engl J Med. 2004;350:2247–2256

9. Fang Jing MD a, Mensah George A MD a, AldermanMichael H MD b, Croft Janet B PhD a. Trends in acutemyocardial infarction complicated by cardiogenicshock, 1979–2003, United States. Am Heart J. December2006;152(6):1035–1041.

10. Hochman JS, Sleeper LA, Webb JG, et al. Early revascu-larization in acute myocardial infarction complicated bycardiogenic shock. SHOCK investigators. should weemergently revascularize occluded coronaries for car-diogenic shock. N Engl J Med. 1999;341:625–634.

11. Thackray S, Easthaugh J, Freemantle N, Cleland JG. Theeffectiveness and relative effectiveness of intravenousinotropic drugs acting through the adrenergic pathwayin patients with heart failure-a meta-regression analysis.Eur J Heart Fail. 2002;4:515–529.

12. Sakr Y, Reinhart K, Vincent JL, et al. Does dopamineadministration in shock influence outcome? Results ofthe Sepsis Occurrence in Acutely Ill Patients (SOAP)Study. Crit Care Med. 2006;34:589–597

13. Kohsaka S, Menon V, Lowe AM, et al. Systemic inflam-matory response syndrome after acute myocardialinfarction complicated by cardiogenic shock. ArchIntern Med. 2005;165:1643–1650.

14. Garcia-Gonzalez MJ, Dominguez-Rodriguez A, Ferrer-Hita JJ. Utility of levosimendan, a new calcium sensitiz-ing agent, in the treatment of cardiogenic shock due tomyocardial stunning in patients with ST-elevation myo-cardial infarction: a series of cases. J Clin Pharmacol.2005;45:704–708.

15. Ellger BM, Zahn PK, Van Aken HK, et al. Levosimendan:a promising treatment for myocardial stunning? Anaes-thesia. 2006;61:61–63.

16. Santa-Cruz RA, Cohen MG, Ohman EM. Aortic counter-pulsation: a review of the hemodynamic effects and indi-cations for use. Catheter Cardiovasc Interv. 2006;67:68–77.

17. Thiele H, Sick P, Boudriot E, et al. Randomized compar-ison of intra-aortic balloon support with a percutaneousleft ventricular assist device in patients with revascular-ized acute myocardial infarction complicated by cardio-genic shock. Eur Heart J. 2005;26:1276–1283.

18. Chen YS, Yu HY, Huang SC, et al. Experience and resultof extracorporeal membrane oxygenation in treatingfulminant myocarditis with shock: what mechanical sup-port should be considered first? J Heart Lung Transplant.2005;24:81–87

19. Dharnidharka VR, Stablein DM, Harmon WE.Post-Transplant Infections Now Exceed Acute Rejection asCause for Hospitalization: A Report of the NAPRTCS.Am J Transplant. 2004;4(3):384–389

20. Sakr Y, Dubois MJ, De Backer D, Creteur J Vincent JL.Persistent microcirculatory alterations are associatedwith organ failure and death in patients with septicshock. Crit Care Med. 2004;32(9)1825–31

21. LeDoux D, Astiz ME, Carpati CM, et al: Effects of perfu-sion pressure on tissue perfusion in septic shock. CritCare Med. 2000;28:2729–2732

22. Dellinger PR, Levy MM, Carlet JM et al. for the Interna-tional Surviving Sepsis Campaign Guidelines CommitteeSurviving Sepsis Campaign: International guidelines formanagement of severe sepsis and septic shock: 2008. CritCare Med. January 2008;36(1):296–327.

23. Regnier B, Rapin M, Gory G, et al. Haemodynamic effectsof dopamine in septic shock. Intensive Care Med.1977;3:47–53

24. Bellomo R, Chapman M, Finfer S, et al. Low-dose dopa-mine in patients with early renal dysfunction: A placebo-controlled randomised trial. Australian and New ZealandIntensive Care Society (ANZICS) Clinical Trials Group.Lancet. 2000;356:2139–2143

25. Russell J. Hemodynamic support of sepsis. Vasopressinversus norepinephrine for septic shock. Program andabstracts of the Society of Critical Care Medicine 36thCritical Care Congress. Orlando, Florida; February17–21, 2007.

26. Dunser MW, Mayr AJ, Tura A, et al. Ischemic skin lesionsas a complication of continuous vasopressin infusion incatecholamine-resistant vasodilatory shock: Incidenceand risk factors. Crit Care Med. 2003;31:1394–1398.

27. Rabuel C, Mebazza A. Septic Shock: a heart story sincethe 1960s . Intensive Care Med. 2006;32(6):799–807.

28. Annane D, Sebille V, Charpentier C, et al. Effect of treat-ment with low doses of hydrocortisone and fludrocorti-sone on mortality in patients with septic shock. JAMA.2002;288:862–871


29. Sprung CL, Annane D, Briegel J, et al: Corticosteroidtherapy of septic shock (CORTICUS). Abstr. Am RevRespir Crit Care Med. 2007;175:A507

30. Sin DD, McAlister FA, Man SF, Anthonisen NR. Con-temporary management of chronic obstructive pulmon-ary disease: scientific review. JAMA. 2003;290:2301–12.

31. Masip J, Roque M, Sanchez B, Fernandez R, Subirana M,Exposito JA. Noninvasive ventilation in acute pulmon-ary edema: systematic review and meta-analysis. JAMA.294:3124–3130

32. The Acute Respiratory Distress Syndrome Network.Ventilation with lower tidal volumes as compared withtraditional tidal volumes for acute lung injury and theacute respiratory distress syndrome. N Engl J Med.2000;342:1301–1308.

33. Ferguson ND, Meade MO, Esteban A et al. Influence ofRandomized Trials on usual Clinical Practice. The Inter-national Study of Mechanical Ventilation. Proc Am ThorSoc. 2006;3:A831

34. National Heart, Lung, and Blood Institute Acute Respira-tory Distress Syndrome (ARDS) Clinical Trials Network.Comparison of two fluid-management strategies in acutelung injury. N Engl J Med. 2006;354:2564–75.

35. National Heart, Lung, and Blood Institute Acute Respira-tory Distress Syndrome (ARDS) Clinical Trials Network.Pulmonary-artery versus central venous catheter toguide treatment of acute lung injury. N Engl J Med.2006;354:2213–24.

36. Derdak S, Mehta S, Stewart TE, et al. High-frequency oscil-latory ventilation for acute respiratory distress syndromein adults. Am J Respir Crit Care Med. 2002;166:801–808.

37. The National Heart, Lung and Blood Institute ARDS Clin-ical Trials Network. Higher versus lower positive end-expiratory pressure in patients with the acute respiratorydistress syndrome. N Engl J Med. 2004;351:327–336.

38. Griffiths MJ, Evans TW. Inhaled nitric oxide therapy inadults. N Engl J Med. 2005;353:2683–95.

39. AdhikariNK, Burns KE, Freidrich JO, Granton JT, CookDJ, MeadeMO. Effect of nitric oxide on oxygenation andmortality in acute lung injury: systematic review andmeta analysis. BMJ. 2007;334:779–82.

40. Gattinoni L, Tognoni G, Pesenti A, Taccone P, MascheroniD, Labarta V, Malacrida R, Di Giulio P, Fumagalli R, PelosiP, Brazzi L, Latini R, Prone-Supine Study Group. Effect ofprone positioning on the survival of patients with acuterespiratory failure. N Engl J Med. 2001;345: 568-573.

41. Guerin C, Gaillard S, Lemasson S, Ayzac L, Girard R, BeuretP, Palmier B, Le QV, Sirodot M, Rosselli S, Cadiergue V,Sainty JM, Barbe P, Combourieu E, Debatty D, RouffineauJ, Ezingeard E, Millet O, Guelon D, Rodriguez L, Martin O,Renault A, Sibille JP, Kaidomar M. Effects of systematicprone positioning in hypoxemic acute respiratory failure: arandomized controlled trial. JAMA. 2004;292:2379–2387.

42. Bein T. Weber F. Philipp A, et al. A new pumpless extra-corporeal interventional lung assist in critical hypoxe-mia/hypercapnia. Crit Care Med. 2006;34:1372–77.

43. Meduri GU, Headley AS, Golden E, Carson SJ, UmbergerRA, Kelso T, Tolley EA. Effect of prolonged methylpred-nisolone therapy in unresolving acute respiratory dis-tress syndrome: a randomized controlled trial. JAMA.1998;280:159–165.

44. Steinberg KP, Hudson LD, Goodman RB, Hough CL,Lanken PN, Hyzy R, Thompson BT, Ancukiewicz M.National Heart, Lung, and Blood Institute Acute Respira-tory Distress Syndrome (ARDS) Clinical Trials Network.

Efficacy and safety of corticosteroids for persistent acuterespiratory distress syndrome. N Engl J Med. 2006;354:1671–1684.

45. Winck JC, Azevedo L, Costa-Pereira A, et al. Efficacy andsafety of non-invasive ventilation in the treatment ofacute cardiogenic pulmonary oedema – a systematicreview and meta-analysis. Crit Care. 2006;10:R69.

46. Ferrer M, Esquinas A, Arancibia F et al. Noninvasiveventilation during persistent weaning failure. A rando-mized controlled trial. Am J Respir Crit Care Med.2003;168:70–76.

47. Ram FSF, Picot J, Lightowler J, et al. Non-invasive positivepressure ventilation for treatment of respiratory failuredue to exacerbations of chronic obstructive pulmonarydisease. Cochrane Database Syst Rev. 2004;3:CD004104.

48. Rana S, Jenad H, Gay P et al. Failure of non-invasiveventilation in patients with acute lung injury: observa-tional cohort study. Crit Care. 2006;10:R79.

49. MacIntyre NR, Cook DJ, Ely EW Jr, et al. Evidence-basedguidelines for weaning and discontinuing ventilatory sup-port: a collective task force facilitated by the AmericanCollege of Chest Physicians; the American Association forRespiratory Care; and the American College of CriticalCare Medicine. Chest. 2001;120(6Suppl):375S–395S

50. Boles JM, Bion J, Connors A, et al. Weaning frommechanical ventilation. Eur Respir J. 2007;29:1033–56

51. Parikh CR, Sandmaier BM, Rainer F, Blume SK, Sahebi F,Maloney DG, Maris MB, Nieto Y, Edelstein CL, SchrierRW, McSweeney PJ. Acute Renal Failure after Nonmye-loablative Hematopoietic Cell Transplantation Am SocNephrol. 2004;15:1868–1876.

52. Carmichael P, carmichael AR. acute renal failure in thesurgical setting. ANZ J Surg. 2003;73(3):144–153.

53. Lameire N, Van Beisen W, Van holder R. Acute renalfailure. Lancet. 2005;365:417–430.

54. Thakar CV, Arrigain S, Worley S, Yared JP, Paganini EP.A clinical score to predict acute renal failure after cardiacsurgery. J Am Soc Nephrol. 2005;16:162–168.

55. Wijeysundera DN, Karkouti K, Beattie WS, Rao V,Ivanov J. Improving the identification of patients atrisk of postoperative renal failure after cardiac surgery.Anesthesiology. 2006;104:65–72.

56. Van den Berghe G, Wouters P, Weekers F, et al. Intensiveinsulin therapy in the critically ill patients. N Engl J Med.2001;345:1359–1367.

57. Wan L, Bellomo R, Giantomasso DD, Ronco C. Thepathogenesis of septic acute renal failure. Curr OpinCrit Care. 2003;9:496–502.

58. Warren HS, Suffredini AF, Eichacker PQ, Munford RS.Risks and benefits of activated protein C treatment forsevere sepsis. N Engl J Med. 2002;347:1027–1030.

59. Krismer AC, Wenzel V, Mayr VD, Voelckel WG, Stroh-menger HU, Lurie K, Lindner KH. Arginine vasopressinduring cardiopulmonary resuscitation and vasodilatoryshock: Current experience and future perspectives. CurrOpin Crit Care. 2001;7:157–169.

60. De Vriese S. Prevention and Treatment of Acute RenalFailure in Sepsis. J Am Soc Nephrol. 2003;14:792–805.

61. Lameire NH, De Vriese An S, Vanholder R. Preventionand nondialytic treatment of acute renal Failure CurrOpin Crit Care. 2003;9(6):481–490.

62. Knaus WA, Wagner DP, Draper EA. The APACHE IIIprognostic system. Risk predicton of hospital mortalityfor critically ill hospitalised adults. Chest. 1991;1100:1619–1636.


63. Bellomo R, Ronco C, Kellum JA, Mehta RL, Palevsky P. andthe ADQI workgroup. Acute renal failure – definition,outcome measures, animal models, fluid therapy and infor-mation technology needs: the Second International Con-sensus Conference of the Acute Dialysis Quality Initiative(ADQI) Group. Crit Care. 2004;8:R202–12.

64. Gettings LG, Reynolds HN, Scalea T, Outcome in post-traumatic acute renal failure when continuous renalreplacement therapy is applied early versus late. Inten-sive Care Med. 1999;25:805–13.

65. Wilkinson AH, Cohen DJ. Renal Failure in the Recipientsof Nonrenal Solid Organ Transplants J Am Soc Nephrol.1999;10:1136–1144.

66. Metnitz PGH, Krenn CG, Steltzer H, Lang T, Ploder J,Lenz K, Le Gall JR, Druml W. Effect of acute renalfailure requiring renal replacement therapy on outcomein critically ill patients. Critical Care Med. 2002;30:2051–2058.

67. Levy EM, Viscoli CM, Horwitz RI. The effect of acuterenal failure on mortality: a cohort analysis. JAMA.1996;275(19):1489–1494.

68. Petroni KC, Cohen NH. Continuous Renal ReplacementTherapy: Anaesthetic Implications Anesth Analg. 2002;94:1288–1297.

69. Myers BD, Moran SM. Hemodynamically mediated acuterenal failure. N Engl J Med. 1986;314:97–105.

70. Van Bommel EFH, Leunissen KML, Weimar W. Contin-uous renal replacement therapy for critically ill patients:an update. J Intensive Care Med. 1994;9:265–80.

71. Druml W. Metabolic aspects of continuous renal repla-cement therapies. Kidney Int. 1999;56:S56–61.

72. Ronco C, Bellomo R, Homel P, Brendolan A, Dan M,Piccinni P, La Greca G. Effects of different doses incontinuous veno-venous haemofiltration on outcomesof acute renal failure: a prospective randomised trial.Lancet. 2000;356:26–30.

73. Bouman CS, Oudemans-Van Straaten HM, TijssenJG, Zandstra DF, Kesecioglu J. Effects of earlyhigh-volume continuous venovenous hemofiltrationon survival and recovery of renal function in inten-sive care patients with acute renal failure: A pro-spective, randomized trial. Crit Care Med. 2002;30:2205–2211.

74. Reeves JH, Cumming AR, Gallagher L, O’Brien JL,Santamaria JD. A controlled trial of low molecular-weight heparin (dalteparin) versus unfractionatedheparin as anticoagulant during continuous venove-nous hemodialysis with filtration. Crit Care Med.1999;27: 2224–2228.

75. Langenecker SA, Felfernig M, Werba A, Mueller CM,Chiari A, Zimpfer M. Anticoagulation with prostacyclinand heparin during continuous venovenous hemofiltra-tion. Crit Care Med. 1994;22: 1774–1781.

76. Vargas Hein O, von Heymann C, Lipps M, Ziemer S,Ronco C, Neumayer HH, Morgera S, Welte M, Kox WJ,Spies C. Hirudin versus heparin for anticoagulation incontinuous renal replacement therapy. Intensive CareMed. 2001;27:673–679.


2Artificial Circulatory Support

John Mulholland

Abstract Extracorporeal cardiopulmonary bypass(CPB) is the most common type of artificialcirculatory support. The evolution of cardiac sur-gery is inextricably linked with the success ofcardiopulmonary bypass. It facilitates surgeryboth on the surface and within the chambers ofthe heart providing the function of the heart andlungs, giving the blood momentum, and perform-ing gas exchange, respectively. This allows theheart and lungs to be isolated from the patient’ssystemic circulation. CPB was first used byJohn Gibbon on the May 6, 1953, to close anatrial septal defect in 18-year-old Cecilia Bavo-lek. Since then improvements in technology,management and understanding of CPB havesignificantly contributed to a reduction in patientmorbidity and mortality. It has also enabled a shifttoward an older more complicated patientpopulation.

This chapter examines the cannulation sitesnecessary for CPB, their importance, and theoptions available. The in-depth description ofthe extracorporeal circuit focuses on the mate-rials used and the properties of the variouscomponents. The non-physiological aspects ofthese components are highlighted and theimportance of full patient anticoagulation forextracorporeal support is explained. The rolesof the various associated CPB techniques arediscussed, including myocardial protectionduring cross-clamping of the aorta and deephypothermic arrest. The chapter also gives aninsight into many of the acute and long-termcomplications of CPB and finally gives back-ground information on the support devicesallied with cardiac surgery, from intra-aorticballoon counterpulsation to the componentsof long-term extracorporeal membrane oxyge-nation (ECMO).

History

The key events in cardiopulmonary bypass historyare summarised in Table 2-1. In 1881, Von Schro-der developed the method of bubbling air throughvenous blood thereby aerating the blood. The fol-lowing year he introduced the first bubble oxyge-nator. It was 4 years before Von Frey and Gruberintroduced the first film oxygenator but all of thiswas without the ability to control the anticoagula-tion of the blood. This was made possible in 1916when McLean isolated heparin. In 1928, Dale andSchuster working at the National Institute forMedical research in Hampstead, England, devel-oped a double perfusion pump intended to carryout whole body perfusion. It did not, but was,however, adopted by Dr. John Gibbon in his firstheart–lung machine prototype in 1931.

Gibbon from the Massachusetts General Hos-pital in Boston conceived the idea of the heart–lung machine for extracorporeal circulation toremove pulmonary emboli from moribundpatients. It took him 4 years before the firstsuccessful application of the heart–lungmachine for extracorporeal circulation – onhis cat! The cat survived but it was not until1939 that the second generation Gibbon heart–lung machine was revealed. It abandoned theDale–Schuster pumps and incorporated thepump Dr. Michael DeBakey had invented in1934. Strangely enough it took 10 years beforeGibbon utilized protamine to reverse the antic-oagulation effects of sodium heparin. During histenure at Massachusetts General, Gibbon metThomas Watson, chairman of InternationalBusiness Machines (IBM) Corporation. Watson,who was fascinated by Gibbon’s research, pro-mised to help and by 1949 IBM developed


21

the Gibbon Model 1 heart–lung machine. It con-sisted of DeBakey Pumps and film oxygenator.For all Gibbons research in the area it was actu-ally Dr. Clarence Dennis who performed the firsthuman open heart surgery cases involving extra-corporeal circulation in 1951. The patient didnot survive and on June 19 the same year IBMdelivered the Gibbon Model 2 heart–lung toJefferson Medical College Hospital Philadelphia.It was not until 1953 when Dr. Frank F Allbrittendesigned the left ventricular vent to solveintracardiac air complications that Gibbon

performed the first successful application ofextracorporeal circulation in a human. It wason the May 6, 1953, at the Jefferson MedicalCollege Hospital, Philadelphia, that Gibbon suc-cessfully used the heart–lung machine he hadconceived over 22 years previous to providetotal heart and lung support for 18-year-oldCecilia Bavolek while he closed a hole in her atrialseptum. Since these improvements in technology(see Figure 2-1), management and understandingof cardiopulmonary bypass have significantlycontributed to a reduction in patient morbid-ity and mortality. It has also enabled a shifttoward an older more complicated patientpopulation and opened up the possibilitiesof longer-term support devices.

Open Heart Surgery

Access to the heart is achieved by mediansternotomy and a longitudinal incision in thepericardium. Following cannulation and connec-tion to the CPB circuit (see following sections),blood flow to the heart is stopped by applying across-clamp to the ascending aorta. While thisfacilitates operation on a bloodless field it alsostops the blood supply to the myocardium (heartmuscle) a situation known as ischemia. The myo-cardium is protected during this ischemic periodusing one of the methods outlined in the sectionon myocardial protection. The required cardiacoperation can now be performed.1

Coronary Artery Bypass Grafting (CABG)

This involves the use of a suitable conduit tobypass a blockage or multiple blockages in thecoronary arteries (see Figure 2-2). The conduit

Table 2-1. Key Events in Cardiopulmonary Bypass History

1628 William Harvey, St. Bartholomew’s Hospital, London, presents his theory of the circulatory system. Describes the function of the heart,arteries, and veins. Considered to be one of the greatest advances in medicine

1881 Von Schroder developed the method of bubbling air through venous blood thereby aerating the blood1916 McLean isolated heparin making controlled anticoagulation possible1931 Dr. John H. Gibbon, Jr., Massachusetts General Hospital, Boston, conceives the idea of the heart–lung machine for extracorporeal

circulation1935 May 10, Dr. John H. Gibbon, Jr., first successful application of the heart–lung machine for extracorporeal circulation in an animal (cat)1949 Dr. John H. Gibbon, Jr. uses protamine to reverse the anticoagulation effects of sodium heparin1953 Dr. John H. Gibbon, Jr., Jefferson Medical College Hospital, Philadelphia. First successful application of extracorporeal circulation in a

human, an 18-year-old female with an atrial septal defect

Figure 2-1. Evolution of cardiopulmonary bypass.


re-establishes a full blood supply to the area ofthe myocardium previously compromised bythe blockage.

Conduits

• The internal thoracic artery – considered to bethe most successful conduit in terms of long-term patency and freedom from adverse car-diac events. It is dissected from the undersideof the chest wall and has its own blood supply.

• The long saphenous vein – this is the mostcommonly used conduit and is harvestedfrom the medial aspect of the lower limb.Once harvested, side branches are ligatedand the vein is used ensuring correct orienta-tion with regards to valves in the vein and theintended blood flow of the bypass graft.

• The radial artery – this is a free arterial graftthat is harvested from the non-dominant

upper limb. Its use and patency rates haveincreased over the last decade.

• The short saphenous vein – usually used if theabove conduits are not available but thepatency rates are not as good.

• Others – other conduits that have been usedover the years include the gastroepiploic andinferior epigastric arteries, bovine mammaryarteries, and synthetic conduits. All have rela-tively poor patency rates.

Once the cross-clamp is in place a suitablepoint in the coronary artery (post blockage) isidentified and the distal anastomosis is per-formed. A longitudinal incision is made in theartery and the anastomosis is made with a singlecontinuous polypropylene suture. Once all ofthe distal anastomosis are finished the cross-clamp is removed and the top end of the conduitis attached to the aorta using a partial clamptherefore completing the bypass graft.1

Saphenous Vein

Blockage

Left Internal Mammary Artery

Figure 2-2. Schematic of the Coron-ary Artery Bypass Operation (CABG).

Artificial Circulatory Support 23

Valve Operations

Once the cross-clamp is in place the relevantchamber of the heart is opened to provide accessto the valve. If the valve cannot be repaired(commonly only certain problems with themitral valve are repaired), the native valve isremoved leaving a clean, defined muscularvalve annulus. The prosthetic valve is thenfixed in position using sutures that pass throughboth the muscular annulus and the prostheticvalve sewing ring. Vents are used to keep theoperating field (heart chambers) bloodlessand the heart must be de-aired before the cham-ber is closed and the cross-clamp is removed.A schematic of the aortic valve replacementoperation is shown in Figure 2-3

Other Cardiac Operations

These include closure of holes in the heart,removal of growths or debris in the heart cham-bers, and replacement/repair of the aorta. Whilethe operative technique varies the basics of theoperation remain the same.

Cannulation for Cardiopulmonary Bypass

Once the patient is anesthetized, the chest isopened via medium sternotomy and the pericar-dium is lifted to make the heart accessible forcannulation. The objective of cannulation is tofacilitate the removal and subsequent return ofthe patient’s entire circulating blood volume,this process will be repeated approximatelyevery 30 s. On the arterial side, the preferredcannulation sites are the aorta, the femoralartery, and sometimes the auxiliary artery.

Isolated right atrial cannulation can be achievedusing a wire basket known as a Ross Basket. A two-stage venous cannulae is thought to give betterdrainage as its tip sits in the inferior vena cavawhile a second set of holes are positioned in theright atrium. Bi-caval cannulation is used whenaccess to the chambers on the right side of theheart is required. This involves separate cannulaefused to one pipe via a y-connector. The femoralvein is another right-sided cannulation option.

Good cannulation on the arterial side canreduce pressure drops and therefore blooddamage, as well as damage to the aorta. On thevenous side, good cannulation ensures an empty

Open Aorta

Aortic ValveAnnulus

Left Ventricle

Artificial Heart Valve

Figure 2-3. Schematic of an Aortic Valve Replacement (AVR).


heart, which is easier to operate on and reducesthe hemodilution caused by fluid addition tomaintain a circulating volume.

The ideal cannulae possess a good balancebetween the lowest possible pressure drop (seeFigure 2-4) and a small cross sectional area. Asmall cross-sectional area, known as the Frenchsize, means that smaller holes are required toinsert the cannula and that they are less intrusivein the operating field. A low-pressure drop on the

arterial side means that higher blood flow rates(artificial cardiac output) can be achieved redu-cing the risk of blood damage. On the venous side,a low pressure drop will increase the capacity toremove blood from the patient. One of the keydesign criteria for cannulae is a thin wall. Thethinner the wall, the greater the inner diameterand lower the pressure drop to French size ratio.How thin the wall can be is limited by the require-ment for structural strength. Figure 2-5 shows how

Two stage venous cannula Dispersion tip arterial cannula

Figure 2-4. Pressure loss vs. flow, for example, high tech venous and arterial cannulae.

Figure 2-5. The influence of pressuredrop on wall thickness.


the wall thickness can effect the pressure drop intwo cannulae of the same french size (24fr).

Once cannulation is completed the perfusionteam now has access for the removal and supplyof blood, the patient can be connected to theCPB circuit. Once the patient’s heart and lungsupport is being provided by the artificial cir-cuit, ventilation of the patient can be stopped.

Summary of the Short-Term CPB Circuit

Figure 2-6 shows the basic components of theCPB circuit. Venous blood drains via gravityfrom the right atrium through a 0.5 in. polyvi-nyl chloride (PVC) pipe into the venous reser-voir (1,1a). The venous reservoir serves twopurposes, the first of which is to filter theblood. This is done in the central column ofthe reservoir and consists of a porous plasticfoam and a polypropylene woven screen (1b).

The porous foam presents the blood with a tor-turous fluid path, removing particulate emboli.The woven polypropylene sheet is designed toremove gaseous microemboli. The combinationof the two provides filtration to 40 microns. Thesecond role of the venous reservoir is to act as acapacitance chamber to manage the acutevolume shifts experienced during CPB as a resultof heart manipulation, shunts, hypothermia anddrugs.

The filtered venous blood is then drawn fromthe reservoir (1c) and pumped toward the heatexchanger and oxygenator block (3). The twotypes of arterial pump commonly used are apartially occlusive peristaltic roller pump (2a) ora non-occlusive constrained vortex centrifugalpump (2b). The arterial pump is essentially ananalogue of the ventricles, providing the bloodwith momentum. The cardiac output suppliedby this pump is calculated using the patients sur-face area multiplied by a cardiac index ranging

1

1a

1c

1b

2a

4

3a

3

2b

3b

Figure 2-6. Short-term cardiopul-monary bypass circuit.


from 1.8 to 2.4 L/min/m2, 2depending on patienttemperature. Blood then passes to heat exchanger(3) passing over concertinated plastic-coatedaluminium or polypropylene membrane. Thismembrane separates the blood from tempera-ture-controlled water (controlled by separateheater chiller unit) on the other side (3a). Thisenables control of the blood temperature andtherefore patient temperature. Blood then passesto the oxygenator or gas exchange device (3). Thisconsists of a porous polypropylene membranearranged in hollow fibers (3b). The tiny holes inthe membrane create a virtual blood gas interfacebut do not allow the passage of fluid. This phe-nomenon combined with a large surface area(similar concept to the lungs) ensures the effi-cient addition of oxygen and removal of carbondioxide. The last component of a standard CPBcircuit prior to the aorta is a 40-micron arterialline screen filter (4). This may remove furthermicroemboli and definitely acts as a gross airbubble trap, reducing the risk of patient mortalityor morbidity from this source. As with many ofthe components in the CPB circuit the flow paththrough this device is from top to bottom, a keycharacteristic for trapping air.

A Pump to Substitute for the Heart

To successfully substitute for the heart, a pumpmust meet a number of criteria. It must be cap-able of delivering up to 7 L/min, be sterile, have alow prime volume and be gentle on the blood’scomponents. Furthermore, it must be affordableand utterly reliable.

The fluid dynamics of blood is complicatedby its non-Newtonian character: its viscosityvaries with its velocity. Red blood cells changeshape becoming more elongated as their velocityincrease becoming less viscous. Other notablefactors affecting viscosity are temperature –cooled blood is more viscous; and hematocrit –viscosity falls when the concentration of redblood cells falls.

Within the pump, and also throughout thecircuit, areas of extreme pressure must be engi-neered out. Flow of fluid that becomes turbulentproduce shear stresses that damage cells, andareas of negative pressure have the ability todrag gas out of solution and thus entrain gasinto blood.

The arguments about the advantages of deli-vering a pulsatile flow are unresolved; Taylor

(1989) links improved cerebral perfusion withthe delivery of pulsatile flow in prolonged CPB,however, advantages have been difficult to proveand the technique is rarely employed duringCPB. The gaining success of pumps providinglong-term support is forcing the issue back on tothe agenda.

For surgery either a roller pump or the cen-trifugal pump is utilized to deliver flow duringCPB. Outside of the surgery support may also beneeded for significantly longer periods of timeand other forms of pump might be employed.The longer-term pumps are less frequently usedand expensive, with differing cost to benefitratios.

The Roller Pump

The roller pump (shown in Figure 2-7) is anocclusive, positive displacement pump with aperistaltic action. It delivers a fixed quantity offluid with each revolution of the rollers in theraceway. The consoles give relatively little flex-ibility in positioning and the tubing inevitably

Figure 2-7. The roller pump (peristaltic pump).


suffers from spallation, tubing wear due to themotion of the pumps rollers, shortening thelength of their clinical application. Tubing iseither PVC or silicone. The roller pump isused for the majority of extracorporeal supportinternationally being both cheap and highlyreliable.

The Centrifugal Pump

Figure 2-8 shows the centrifugal pump (or con-strained vortex pump) which is characterized byfluid entering along the axis of its impeller andthen emitting in a direction radial (or tangential)to the impellers axis. It is a kinetic pump thatadds energy by fluid accelerating past the rotatingimpeller. It is both pre- and post-load dependentgiving it sensitivity to the fluctuating resistance ofthe patient’s vasculature.

The early suggestions that centrifugal pumpsreduce heparin consumption and reduce blooddamage have not been realized but it is acceptedthat it will not pump gross air as readily as theroller pump, which is an important safety feature.This is explained by equation F¼mr2, where ‘F’is the force, ‘m’ is the mass and ‘r’ is the radius; airhas no mass and therefore will receive no force ormomentum in a centrifugal pump.

A magnetic coupling between the impellerand the drive unit provides sterility and a

disposable pump head. The centrifugal pumprequires little fluid to prime and can be posi-tioned without compromise, allowing the possi-bility of a shortened circuit. Their Achilles heelis cost: the disposable kit for a roller pump is alength of tubing, whereas for a centrifugal pumpit is a relatively sophisticated pump head.

A new generation of centrifugal pumps haveentered clinical practice, devoid of bearings andseals. A magnetic field is utilized not only tocouple the impeller and the drive unit but alsoto levitate the impellers within the pump head.Their impressive longevity has earned them pro-longed licenses for clinical use.

Replicating the Lungs

The need to not only pump blood but also per-form gas exchange in the process brought manychallenges. The act of oxygenation and carbondioxide removal left the blood further damagedand containing multiple gaseous microemboli.The breakthrough of membrane oxygenatorsresolved a number of the issues generated by thedirect gas interface oxygenators. The AmericanAssociation of Medical Implants (AAMI) sets aminimum time of 8 hours for an oxygenator toperform on bypass, a target achieved by micro-porous membranes. For prolonged support, truemembranes are better suited.

Direct Gas Interface Oxygenators

The direct gas interface oxygenators includescreen, rotating disc, and bubble oxygenatorsall of which have the gas (an oxygen/air mix) indirect contact with the fluid (venous blood). Thescreen oxygenator shown in Figure 2-9 had 365working parts and provided oxygen by tricklingblood over many vertical sheets. The disc oxyge-nator shown in Figure 2-10 used discs rotating ina trough of the patient’s blood to expose a thinlayer of blood to atmospheric oxygen. Bubbleoxygenators, a technique first introduced in1882 by Von Shroder are now rarely, if ever,used in hospitals in Europe and North America.Oxygen bubbles of 3–4 mm in diameter arebubbled through the blood, creating the blood-oxygen interface. To remove these bubbles, theblood is then passed through a de-foamingchamber, a bubble trap, and a filter.Figure 2-8. Centrifugal pump (constrained vortex).


These suffer from three weaknesses: first, theconsiderable quantity of microbubbles remain-ing in the blood postoxygenation; second, thelimited amount of gas exchange; and third, theextensive blood–air interface.

Membrane Oxygenators

Microporous Membranes

The widely used hollow fiber microporous oxy-genators (Figure 2-11) were first mass producedin the early 1980s. The membrane is created bybeing drawn longitudinally, with the pores beingproduced by inducing small ruptures in the tubewall. This makes for a net-like structure withalternating unexpanded regions and smallpores.

Sheaves of microporous polypropylenefibers, with diameters between 100 and 200 mm,connect to an inlet and outlet within a cylindri-cal housing. The gas passes inside the fibers andthe blood around the outside. Laminar flow pro-duces a boundary effect that limits the percen-tage of red blood cells exposed to the membrane.To counter this non-laminar flow is encouragedby manipulating the flow paths through theoxygenator.

Like the lungs the membrane oxygenator pre-sents the blood with a large surface area in orderto increase efficiency (commonly between 1.8and 2.4 m2). This has resulted in very efficientgas exchange, probably over efficient and there-fore future design should move toward reducingthe non-physiological surface area of this com-ponents and the associated problems.

Seepage through the membrane’s microporesis halted primarily due to hydrophilic nature ofthe polypropylene membrane. Second the blood,on contact with the membrane, deposits a coat-ing of proteins and platelets on the membrane

Figure 2-9. Screen oxygenator with 365 working parts.

Figure 2-10. Disc oxygenator with a3-day turn around.


surface inhibiting plasma loss and third the sur-face tension of the blood counters seepage.

True Membranes

A pure membrane consists of silicone rubbersheets coiled in a cylindrical fashion. The sili-cone is non-porous making for a complete bar-rier between gas and blood. There is a greatlyreduced risk of fluid seeping to the gas side orconversely gas being entrained into the blood.Though a pure membrane benefits from non-plasma leakage it also stops large moleculesfrom diffusing through the membrane. Thiscan stop the anesthetic gas isoflourane fromentering the blood supply. The membranes areexpensive to manufacture and high volume ofprime is required. However, its longevity makesit an ideal membrane for long-term support,

known as extracorporeal membrane oxygena-tion, discussed later.

Heat Exchangers

Like the oxygenator, the increase in surface areaincreases efficiency. For this reason most heatexchangers are concertinated. Common materi-als are stainless steel, aluminum (both of whichcan be plastic coated), or polypropolene. Wateris pumped at a controlled temperature around10 L/min on one side of this heat exchangermembrane and this dictates the temperature ofthe blood on the other side of the membrane andtherefore patient temperature. As mentionedearlier many patients are cooled at the start ofCPB (sometimes as low as 128C for major aorticwork) and then rewarmed to normothermia(37–37.58C). Arterial blood temperature shouldnever exceed 37.58C and the gradient betweenthe water and the blood temperature should notexceed 108C. This reduces the risk of cerebralinjury, enzyme, and protein breakdown fromthis source.

Arterial Line Filters

The arterial line filter is the last filtration point onthe CPB circuit before the blood is returned to thepatient. It is designed to remove any microemboliabove 40 mm (the most common ALF size). Whilethis level of filtration has already occurred in thefilters of the venous reservoir, the ALF will catchany microemboli (gaseous or particulate) gener-ated from that point onward. This filtration isimportant because the diameters of the vesselsin the patient’s arterial network progressivelydecrease in size distally to the heart. As largebubble travel down this network, it will breakinto smaller bubbles that possess a greater surfacetension. There is a critical point when the bubblesize has decreased to a point where the surfacetension becomes greater than the force that canbe applied on it by the pressure of blood. It nowacts as a plug and is treated as a foreign body,instigating an inflammatory response. Conse-quently the tissue downstream of the blockagebecomes ischemic and eventually cell deathoccurs. If this occurs in the brain neurological,impairment is the result (see complications). Fil-ter size should not venture much smaller than 40

Figure 2-11. Microporous membrane oxygenator (positioned undervenous reservoir system).


mm due to concerns of filtering out and becomingblocked with blood components.3

Protecting the Myocardium

The ideal cardiac operating field is a blood free,stationary heart. This is achieved by positioninga large cross-clamp on the aorta above the cor-onary arteries and before the aortic cannulafrom the CPB circuit. This stops blood flow inthe coronary arteries and therefore the myocar-dium must be protected. The two principleforms of myocardial protection are cross-clamp fibrillation and cardioplegia.

Cross-Clamp Fibrillation

Cross-clamp fibrillation involves passing acurrent into the heart to induce controlledventricular fibrillation. The fibrillating or ‘nearlystill’ heart can then be operated on. Additionalmyocardial protection is provided by the sys-temic cooling of the patient to 328C, reducingthe metabolic demands of the myocardium. Thistechnique only allows the blood supply to theheart to be stopped for up to 15 min and is there-fore only suitable for CABG operations. Whilegradual revascularization and ischemic precondi-tioning are advantageous, the drawbacks are mul-tiple, potentailly damaging applications of thecross-clamp to the aorta.

Cardioplegia

Cardioplegia involves infusing a hyperkalemicsolution into the isolated coronary system caus-ing diastolic arrest. The ‘completely still’ hearthas therefore a very small metabolic require-ment. Cardioplegia is given as one large initialdose and subsequent maintenance doses. Thiscan prolong the cross-clamp or ischemic timefor longer than 1 hour if required. The hyperka-lemic solution can be given mixed with thepatient’s own blood or as a crystalloid solutionand either warm or cold.

Whole Body Hypothermia

This technique is actually whole body protection(all organs) rather than just myocardial protection.

It is mainly used for major aortic work andinvolves the patient being cooled as low as 128C.At this temperature, the patient’s metabolic rate isextremely low and the patient can be drained of alltheir circulating blood to give the ideal blood free-operating field. Surgery can then be performed onthe head and neck vessels and around the arch ofthe aorta. It is widely accepted that at this tem-perature circulatory arrest is possible for 20 minwithout having to use any isolated organ perfusion(e.g., cerebral, visceral).4

Monitoring During CPB

Venous Reservoir Level

The volume of blood in the venous reservoir isinfluenced by two variables: the inlet flowrate ofblood from the right atrium via the venous lineand the outlet flowrate of the CPB pump (rolleror centrifugal). The later is relatively constantand controlled by the perfusionist, while theinlet flowrate is dictated by the patient’s sys-temic vascular resistance, the position of theheart, surgical, and anesthetic intervention. Tomaintain a level in the venous reservoir the twovariables must be the same, the responsibility ofthe perfusionist. If this does not happen the reser-voir level will either increase (not immediatelydangerous) or drop, a situation resulting in airentrainment into the CPB circuit. For an averagepatient, a reservoir level of 400 ml could disap-pear in just over 4 s with gross air entering thepatient 12 s later. The issue surrounding gross airembolism cannot be overstated.

Activated Clotting Time

The importance of ensuring anticoagulationcannot be understated. There are many emer-gency situations that arise during CPB that thecardiac team can overcome. A patient’s bloodclotting during CPB is not a recoverable situa-tion. Full anticoagulation is required to preventclotted blood blocking the CPB circuit. Heparinis an anticoagulant that is ideal for this pur-pose: it has an affect that is easily measurableby activated clotted time (ACT), well tolerated,is relatively cheap and is easily reversible withprotamine. Heparin binds to anti-thrombin III;


300-units/kg body weight is given at the anes-thetic end as well as some in the CPB circuitprime. The level of anticoagulation is measuredusing an activated clotting time. This measuresthe time to clot formation of a known quantityof blood accelerated by a known amount ofactivator (kaolin and/or celite). The ACT-monitoring equipment shown in Figure 2-12uses a cuvette system (which reduces uservariability) to provide an accelerated real-timeactivated clotting result. Having an early real-time result is especially important to confirmadequate anticoagulation in an emergencysituation. The threshold level for satisfactoryanticoagulation varies from unit to unit butthree times the patient’s baseline ACT is anaccepted minimum. The ACT is monitoringprior to heparin, post-heparin (pre-bypass), atleast every 30 min during CPB and finally fol-lowing. As well as avoiding a lack of anticoa-gulation, ensuring good anticoagulation willhelp protect the patients clotting factors andreduce the risk of small cell aggregates formingwhich can potentially block smaller capillaries.Once the patient has been disconnected fromthe CPB circuit that heparin is reversed usingprotamine. Protamine should be administeredslowly as it has a number of adverse effects,including anaphylaxis, vasodilatation, pulmon-ary hypotension, and complement activation.5

Arterial Pump Flow

Arterial pump flow or cardiac output duringCPB must be maintained between a range thatensures the patients metabolic demands at met

without increasing blood damage and microem-boli delivery. Extracorporeal support facilitates avery acute control over the patients cardiac out-put. The patients cardiac output (and thereforeblood pressure) can be reduced to almost zerofor short periods of time. Prolonged under-perfusion can result in organ damage, forexample, the kidneys, the liver, the gut, and thebrain. As a guide the patients target cardiacoutput is calculated using the body surfacearea multiplied by a cardiac index of 2.4L/min/m2. This cardiac index can be reducedas the patient’s temperature and therefore meta-bolic demand is reduced. The monitoring anddecisions with regards to pump flow are intrin-sically linked to arterial pressure and circulatingblood volume. Most pumps used in cardiac cen-ters worldwide deliver continuous not pulsatileflow.

Gas Flows in Flowmeter

Like pump flow, gas exchange, principally theoxygen and carbon dioxide levels of the bloodmust be maintained between limits. These mustensure that patients metabolic demands aremet without exceeding the oxygen content ofthe blood or removing to much carbon dioxide.The efficiency of the gas exchange devicesmakes high O2 levels (risk of cell damage andgas coming out of solution) and low CO2 levels(interference with cerebral auto regulation) areal threat. These parameters are checked atleast every 15 min via a blood gas analyzer(see Figure 2-13 showing examples of goodand poor blood chemistry). In addition, somecardiac units have in-line blood gas monitoringwhich monitors the blood chemistry every 5 susing spectrophotometry.

Arterial Pressure

Arterial pressure or perfusion pressure is influ-enced by both blood flow and SVR. The pump iscontrolled as previously described, while theSVR is controlled pharmacologically usingvasoconstrictors and vasodilators. A meanpressure of 70 mmHg is widely accepted as atarget arterial pressure as the research dictatesthat this pressure provides adequate perfusionfor the major organs. There is room to movebelow this, specifically for the brain which

Figure 2-12. Modern ACT monitoring.


autoregulates but not below 50 mmHg for anylength of time. A pressure above a mean of70 mmHg can make the surgery more difficult,especially with regards to the visualizing the oper-ating field. There is still not a definitive answerwith regards to the effect on the organs of non-pulsatile flow during CPB (see Figure 2-14). Themain questions surround not only organ function

but the ability to generate pulsatile flow in thepatient’s circulation and the blood damage it causes.

Venous Pressure

The central venous pressure (CVP) gives a directindication of the amount of blood in the heart.This is important to ensure that we do not over-fill the heart when we are gradually terminatingCPB and asking the heart to do its own workagain. It also provides advance warning whenblood is not draining properly from the heartalerting the perfusionist to an impending reduc-tion in venous reservoir level. For most cardiacoperations, the CVP will be zero or negative dueto the gravity drainage. The right atrium is notdesigned to deal with negative pressures andwhile this clearly upset the patient’s oncotic bal-ance, some of the other potential non-physiological effects are still not clearly defined.

Urine Output

Urine output is one of the key markers for kid-ney function but also plays a role in the patient’s

Hb too low – O2 carrying capacity, oncotic pressure of the blood

Acidotic pH due to high CO2, low O2

and an overall deficiency in meeting the patient’s metabolic demands

Haematocrit (% of cells in blood) confirming the low Hb but also a low level of white cells (infection risk) and platelets (bleeding risk)

Low bicarbonate levels, negative base defecit confirming the metabolic deficiency and a low arterial % as a result of reduced O2 carrying capacity (Hb)

Good Blood Chemistry Poor Blood Chemistry

Figure 2-13. Labeled examples of good and poor blood chemistry during CPB.

180 mmHg

280 mmHg

Figure 2-14. The arterial Tycos gauge showing operating thresholds.


total fluid balance as well as influencing manyother CPB decisions. These include appropriatefluid use, whether the patient requires higherperfusion pressures or flows, whether pharma-cological intervention is required, and thepotential need for an artificial hemoconcentra-tor. Hemoconcentration provides a similarfunction to dialysis but fluid removal is basedon the principle of convective solute transportacross a semipermeable membrane comparedto dialysis, where the driving force is soluteosmotic pressure. With hemoconcentrationaccess is more straightforward as the patientis on extracorporeal support and therefore hasa faster filtration rate.

‘Tycos’ Pressure

The Tycos pressure is a direct measurement ofthe pressure at the tip of the aortic cannula. Itprovides information about the positioning ofthe cannula throughout the case as well as anyobstructions (for example, aortic cross-clamp).This area of interaction between the CPB circuitand the patient’s aorta is extremely non-physiological and has a knock on effect for thefluid dynamics throughout the circulation. Thestarting point during normal circulation is theaortic valve annulus (about 21 mm in diameter)with an axial exit flowrate of around 5 L/min(0.24 m/s). CPB changes this starting point to atube (about 8 mm) angled off the aortas axisstill with a flowrate of around 5 L/min but witha 7-fold increase blood velocity of 1.66 m/s.While the knock on effect with regards to thechange in fluid dynamics in the organs isunclear, the positional aspect of the cannulamust be considered. It is important to avoidblood jetting at 1.66 m/s against the wall of theaorta, especially a weak aorta or one heavilycalcified due to athero-sclerosis.

ECG

With the heart isolated from the circulation bythe cross-clamp the cardiac team have a verydifferent aim than other medical teams in termsof target ECG, namely the maintenance of VF(cross-clamp fibrillation) or the maintenance ofasystole. Once the cross-clamp is removed andblood supply to the myocardium is resumed,sinus rhythm is the goal. If sinus rhythm is not

forthcoming prior to the cessation of CPB, thenepicardial pacing wires and temporary pacingwill be used in the post-operative period. Anexample each of the ECG and arterial pressuretraces during CPB is shown in Figure 2-15.

Arterial and Venous Saturations

These are monitored to ensure that the extra-corporeal circuit is not only adding oxygen butmeeting the patient’s metabolic demands. Arter-ial saturations above 98% and venous satura-tions above 70% are the thresholds used.

Arterial and Venous Blood Temperature

These are monitored closely to ensure good glo-bal cooling and rewarming of the patient with-out exposing patients to the temperate risksdiscussed in the section on heat exchangers.

CPB Safety Features

All of the monitored parameters have eitheroperator or mechanical safety devices. Thesesafety features vary from an audible alarm tostopping the arterial pump. The mechanicalalarm are detailed below:

b)

a)

Figure 2-15. ECG and Arterial Blood Pressure on CPB: (a) x-clamp offand (b) x-clamp on.


• Venous reservoir level – level and/or bubblealarm stops arterial pump (see Figure 2-16)

• Arterial pump flow – audible alarm warns ofhigh/low flows

• Gas flows – audible alarm

• Arterial (Tycos pressure) – pressure alarmstops arterial pump

• Arterial and venous saturation – audible alarm

• Arterial and venous temperature – audible alarm

The operator or clinical perfusionist is theoverall safety device as well as ensuring the extra-corporeal support is as physiological as possible.Clinical perfusionists are trained to have a uniqueblend of in-depth cardiac-specific medical knowl-edge and advanced technical skills. Accreditationvaries across the world but always involves a bal-ance between clinical and theoretical training.

Non-physiological Aspectsof Cardiopulmonary Bypass

It is clear that extracorporeal support exposesboth the patient and the patient’s blood to forcesand an environment not experienced in normal

physiological. It is a testament to the bodiesresilience that it can withstand such an insult.Acute non-physiology can result in mortalityand moderate non-physiology can result inmorbidity.

• Anticoagulation

• Exposure of the blood to air

• Exposure of the blood to shear stresses

• Exposure of the blood to non-physiologicalpressures

• Exposure of the aortic wall to non-physiologicalpressures and shear stresses

• Exposure of the organs to non-pulsatileperfusion

• Hemodilution

• Exposure of the right atrium to negativepressure

• Generation of gaseous and particulatemicroemboli

Complications of the Non-physiologicalAspects of Cardiopulmonary Bypass

The in-hospital mortality varies depending onthe operation, for example, CABG – 2–3%, iso-lated valve surgery – 4.3%, and combined CABGand valve surgery – 7.6%. Many of the non-physiological aspects of CPB contributed to thefollowing post-operative problems.

Bleeding

The complex alterations of hemostasis lead toplatelet dysfunction and increased fibrinolysis.These alterations can result in the requirementfor donor blood products like red blood cells,platelets, and fresh plasma, all of which areassociated with risk. About 1–3% of patientsrequire re-exploration for persistent bleeding.

Neurological Events

As many as 75% of patients suffer subtle, rever-sible neurological deficits post-operatively. Strokeas a result of a macroembolism, intracerebralhemorrhage, or thrombotic occlusion is seen in<1% of patients below 70 years and 5% of patientsabove 70 years. Cerebral complications are asso-ciated with poor long-term outcome and increasedmortality and morbidity.6

Figure 2-16. Level sensor that uses capacitance to detect level falling– will stop arterial pump.


Kidney Dysfunction

Kidney or renal dysfunction is associated withhemodilution, hypothermia, and endocrineeffects during CPB. About 1–5% of cardiac sur-gical patients require acute renal failure andrequire dialysis.7

Low Cardiac Output

This can present itself in several forms, namelylow blood pressure, bradycardia or tachycardia,acidosis, and reduced urine output. CPB cancontribute to many of the underlying reasonsfor low cardiac output, for example, inadequatecirculating volume, myocardial injury, abnor-mal ECG, and electrolyte imbalance. It is impor-tant that the underlying reason is identified andthe low cardiac output is avoided/reversed.8

Pulmonary Complications

Most patients suffer a mild reduction in lungfunction post-operatively. Many of the non-physiological aspects of CPB contribute tothis. The main CPB contributor to major pul-monary dysfunction (about 2% of patients) isthe inflammatory response caused by exposureto non-physiological surfaces and air. Theinflammatory response created by the non-physiological surface still occurs even thoughthere have been advances in surface biocompat-ibility over the last 5 years. Many of the moderncircuits have a coating on all surfaces to increasebiocompatibility. Some coatings attempt tomimic endothelium (phosphorylcholine), whileothers are coated with heparin (carmeda). Goodanticoagulation with systemic heparin offerssome protection against post-operative inflam-matory response and associated morbidity but itacts near the end of the clotting cascade andtherefore does not suppress the activation of theupstream protease cascades (bleeding issues).9

ECG Complications

The most common ECG irregularity is atrialfibrillation (reported in 25–30% of CABG patientson the second or third post-operative day). Fromthe CPB point of view, this can be a result ofpoor myocardial protection, magnesium, and/or

potassium imbalance. If AF persists the risk ofmicroemboli generation (clot) should bereduced pharmacologically or via electro-cardioversion.

Long-Term Circulatory Support

Patients in cardiogenic shock or who fail to weanfrom bypass may require post-CPB support as abridge to recovery or transplant. Commonlyused circulatory support strategies are intra-aortic balloon (IAB) counterpulsation and ven-tricular assist devices (VAD).

IAB counterpulsation involves a catheter witha 34 or 40 cc balloon at one end being insertedinto the descending aorta via the femoral artery.The balloon sits just below the head and neckvessels and above the renal arteries. It uses theECG as a trigger to inflate during diastole (theperiod of aortic valve closure during the cardiaccycle – dicrotic notch to the onset of the nextcycle) and deflate during systole. This not onlyincreases the diastolic pressure and thereforeblood supply to the myocardium during infla-tion but also reduces the pressure the heart mustwork against by deflating. IAB counterpulsationcan only provide up to 1.5 L/min of cardiacoutput support. IAB placement and the influ-ence on the cardiac cycle are shown in Figure2-17. Figure 2-17b shows three cardiac cyclesand the influence of the IAB on both the secondand the third cycles. The first cycle is shown forreference.

VAD technology can be applied to the left,right, or to both sides of the heart. It is moreinvasive than an IAB as relevant cannulationsites are required to remove and return blood.The extracorporeal circuit incorporates a centri-fugal pump that can supply up to 6 L/min ofcardiac output support (for a period of months),i.e., the heart does not need to contribute any ofthe work itself. They are used as a bridge torecovery or a bridge to transplant. Some of thelonger-term VAD’s use pneumatic or hydraulicchamber technology in an attempt to providemore physiological support. Finally extracorpor-eal membrane oxygenation (ECMO) or extracor-poreal carbon dioxide removal (ECO2R) incorpo-rate a gas exchange device as well as a pump. Thesuccess rate of these techniques is considerablyhigher in pediatrics than adult surgery. The basicprinciples of these circuits are the same as theshorter-term circuits, only the materials differ


slightly. The key to success of long-term extra-corporeal support is not found in the technol-ogy of the circuits, but the clinical/medicalmanagement of the patients.

References

1. Edmunds LH. Adult Cardiac Surgery. New York:McGraw-Hill Education; Jan 1997. ISBN-10: 0070189633,ISBN-13: 978-0070189638.

2. Ward J. Oxygen delivery and demand. Surgery.2006;24(10):354–60.

3. Gravlee GP, Davis RF, Utley JR, Kurusz M. CardiopulmonaryBypass: Principles and Practice, 2nd ed. Philadelphia:Lippincott/Williams & Wilkins; May 2000.

4. Cardiopulmonary Bypass: Principles and practise, 2nd ed.Gravlee, GP., Lippincot/Williams&Wilkins, Philadelphia;2000.

5. Shannon M. Anticoagulation. Surgery. 2007;25(4):150–4.

6. Kenneth G. Shann, Donald S. Likosky, John M. Murkin,et al. An evidence-based review of the practice of cardio-pulmonary bypass in adults: A focus on neurologic injury,glycemic control, hemodilution, and the inflammatoryresponse. J Thorac Cardiovasc Surg. 2006;132:283–90.

7. Delbridge MS, Raftery A. Access for dialysis. Surgery.2004;22(11):277–83.

8. Cardiovascular Physiology, 8th ed. Berne, RM. CV Mosby,St. Louis; 2000.

9. Casthely PA, Bregman D. Cardiopulmonary Bypass:Physiology, Related Complications, and Pharmacology.Mount Kisco, NY: Blackwell/Futura; November 1991.ISBN-10: 0879933968, ISBN-13: 978-0879933968.

Figure 2-17. (a) IAB placement in the descending aorta and (b) influence of IAB on the cardiac cycle.


3The Artificial Kidney

Christopher Kirwan and Andrew Frankel

Introduction

Approximately 100 individuals, per millionpopulation, per year, reach end-stage renal fail-ure in the United Kingdom, with many morewho have used hemodialysis for short-termtreatment of acute renal failure or as a bridgeto renal transplantation. An increase in preva-lence of conditions such as diabetes means thatin the future more and more people will bediagnosed with renal failure and will requirerenal replacement therapy.1

Most of the patients who receive hemodialy-sis would die without it and owe their lives to theartificial kidney. This chapter reviews the his-tory of and the progression of the technologyassociated with the artificial kidney and endingwith consideration of possible future technicaladvances.

The Kidney

The known functions of the kidney can be sepa-rated into four main categories: excretory, reg-ulatory, endocrine, and metabolic (Table 3-1).The principle role is the elimination of wastematerial and the regulation of the volume andcomposition of body fluid.

The endocrine and metabolic functions of thekidney can be replacedwithmedications. Recom-binant erythropoietin (r-epo) in combinationwith iron supplementation can help to controlanemia. Calcium homeostasis is impairedbecause a diseased kidney cannot perform thesecond hydroxylation of vitamin D to its activecomponent calcitriol or 1,25 di-hydroxycholecal-ciferol. Calcitriol can be replaced by tablets.

The challenge for physicians and engineerswas to create a system that could replace theexcretory and regulatory role of the failingkidney.

The kidney receives approximately 25% ofcardiac output or 1300 ml of blood every min-ute. The functioning unit of the kidney is calleda nephron. There are approximately one mil-lion nephrons in each kidney, and eachnephron is made up of a filtering unit (theglomerulus) and its associated tubule, whichregulates the filtrate produced from the glo-merulus (Figure 3-1).2

Within the glomerulus, a hydrostatic pres-sure gradient of approximately 10 mmHg pro-vides the driving force for the ultrafiltration ofa virtually protein and fat-free fluid across theglomerular capillary wall into the renal tubule.The total ultrafiltration is more commonlyknown as the glomerular filtration rate (GFR)and in a healthy person this equates to theultrafiltration of approximately 180 L of watera day.

The selective secretion and re-absorption ofwater, essential electrolytes, glucose, andamino acids occur as the filtrate moves alongthe renal tubule. Virtually all of the potassium,bicarbonate amino acids, and glucose are reab-sorbed in the proximal renal tubule along with60–80% of filtered water. Further water andsodium chloride are reabsorbed more distallywith fine tuning of salt water balance occurringin the distal tubule and the collecting ductunder the influence of aldosterone and anti-diuretic hormone. This leaves approximately 2L of urine a day. As renal failure progresses,the GFR falls and eventually dialysis is requiredto replace the excretory and regulatoryfunctions.


39

The Artificial Kidney

In order to appreciate how the artificial kidney isdeveloped, it is necessary to understand thebasic components required (Table 3-2).

In this article, by concentrating on each com-ponent individually, we can build up a picture ofhow a successful artificial kidney has evolvedover the years, allowing the expansion of hemo-dialysis into a recognized and readily availabletreatment.

The Semipermeable Membrane

An understanding of osmosis, diffusion, and semi-permeable membranes began to emerge in themid-1800s. There are two ways solutes can bemoved through a membrane: diffusion and con-vection (Figure 3-2). Diffusion of solutes from ahigh concentration to a low concentration even-tually leaves an equal concentration on both sidesof the membrane. Convection pulls solution andsolutes across a membrane. Thomas Graham, aprofessor of chemistry in Glasgow, first coinedthe phrase dialysis in 1854 when he described themovement of various types of solutes through amembrane forced by osmotic pressures.3 Grahamused ox bladders for his first membranes but laterused vegetable parchment coated with albumin. Ayear later Adolf Fick described two equationswhich quantified that the flow ofmass by diffusion(i.e., the flux) across a plane was proportional tothe concentration gradient of the diffusant acrossthat plane.4 The membrane, or plane, he used inhis studies was colloidin.4 Colloidin is syrup-likeliquid made from cellulose, nitric acid, alcohol,and ether and when it dries it forms a porous film.

In 1913, John Abel, a renowned pharmacolo-gist at Johns Hopkins University USA, publishedan article describing hemodialysis in animalsusing a semipermeable membrane made of col-loidin.5 This was the first ‘‘Artificial Kidney’’.

Soon the first reports of dialysis in humansappeared. George Haas a German physician,dialyzed six patients over a colloidin membranebetween 1924 and 1925, but none of them sur-vived.6–8 Two further patients were dialyzedover colloidin in 1928 when Haas experimentedwith heparin as an anticoagulant, but hedescribed the results as disappointing, citingtechnical difficulties and opposition fromcolleagues.9

In 1923, Heinrich Necheles tried a preparedperitoneum (Gold-beater’s skin), often from asheep or an ox, as a membrane when dialyzinguremic dogs,10 but this technique did notprogress.

It became clear that a membrane had to berobust, easily sterilized without damage to thematerial or alterations in its properties and havea long shelf life. Colloidin performed badly in allof these areas but conveniently an alternativewas already in use in the food and packagingindustry.

In 1839, cellulose was the name given tothe compound that ‘‘fills the cells that makeup wood’’.11 It was first purified from woodin 1885 and subsequently cellulose acetatewas synthesized around 1895. It is thoughtthat as early as 1910 it was available insheet form under the name cellophane. Cello-phane was used for laboratory studies of dia-lysis in sheet form by Freda Wilson of theUniversity of British Columbia in 1927 andshe showed how easy it was to sterilize.12

This versatile and cheap product was alreadybeing made into tubing for the manufactureof sausages by the Visking Company of Chi-cago. It was tough, did not burst under mod-erate pressures and even in its commercialform was relatively free of microscopicholes. Andrus used the sausage skin inlaboratory dialysis experiments and it provedto have excellent diffusion characteristics.13

During the 1930s, many papers (reviewed byFagette14) were published on the physical anddialysis characteristics of various forms ofcellulose membranes.

In 1937, a New York hematologist, WilliamThalhimer, began to use cellophane tubing 2 cmwide and 30 cm long, in an artificial kidneysimilar to the one Abel constructed, to dialyzedogs. Along with his work into heparin, thisprovided the vital realization that commerciallyavailable cellophane could be used for in vivohemodialysis.15,16

Table 3-1. The Functions of the Kidney

Excretory - excretion of waste products and drugsRegulatory - control of body fluid volume and its compositionEndocrine - production of erythropoietin, renin and

prostaglandinsMetabolic - metabolism of vitamin D


Cellulose acetate or cellophane was univer-sally used as the membrane in dialysis machinesup until around the 1960s. There were frequentleaks and a number of reports of ruptured

cellophane membranes which constantly fuelledthe search for other materials. Frederick Kiil wasthe first to use a new membrane called cupro-phan. Cuprophan membranes had a greater

Figure 3-1. A diagram of the glomerulus and the nephron.

The Artificial Kidney 41

porosity for solutes and water than other mate-rials. Initially it was only available as flat sheetsbut by 1966 there was standardized industrialproduction of these dialysis membranes, and in1969, the success had stimulated the productionof the cuprophan tubular membrane.

Concurrently, 1964 saw the production of thenext major step in the development of the artifi-cial kidney, the hollow fiber dialyzers. RichardStewart began to explore themedical applicationsof cellulose acetate hollow fibers.17,18 These ultra-thin fibers were approximately 200 mm in dia-meter and 5 mm thick. The fiber wall representedthe dialysis membrane. The artificial kidney thathe designed contained 11000 fibers18 and theversion that was used in the clinical trial in 1967is very similar to current designs. He also demon-strated that substances could be selectively

removed from the blood as well as excess water.The capillary sized fibers allowed the productionof a dialyzer with a much larger surface areafor blood to membrane contact while using rela-tively small amounts of extra corporeal blood(Figure 3-3). This resulted in increased efficiencyof dialysis.

Hollow fiber dialyzers began a revolution inthe way dialysis was considered and offered.The change in membrane type and size runsparallel with the reduction in size of the sup-port structure for the artificial kidney, develop-ment of new polymers for the fibers, and aclearer understanding of the physiologicaleffects and consequences of dialysis at a micro-scale level.

The human glomerular basement membranehas a higher permeability and larger ultrafiltra-tion coefficient than the classical cellulosemembranes. In the late 1970s, the polymer poly-sulfone was introduced as the next generationmembrane and its benefits of large sieving coef-ficients for inulin and ß2 microglobulin werereported a few years later along with improvedbiocompatibility.19

Acetate, diacetate, Hemophan1, and Helix-one1, as well as the original polysulfone, are justa few of the many different polymers that arenow available for membranes in the artificialkidney. They all work in roughly the same way

Table 3-2. Basic Components of an Artificial Kidney

• Semipermeable membrane separating blood and dialyzate• Membrane support structure• Blood compartment-Anticoagulation-Access to the blood stream

• Dialyzate compartment-Dialyzate composition-Water

Figure 3-2. A diagram illustrating thedifference between diffusion andconvection.


with varying degrees of efficiency. The modernHelixone1membrane (FreseniusMedical Care),for example, is a polysulfone-based high fluxmembrane in which the porosity of the innerlayer is finely controlled at the nanoscale bynanotechnology. Nanotechnology is an atomic-ally precise, functional machine system that isdeveloped on the scale of the nanometer (1 nm =1/1,000,000,000 m). In more general terms it isany technology related to features of nanometerscale such as thin films, fine particles, chemicalsynthesis, and advanced microlithography.20

The Helixone1 membrane is produced using anano-controlled spinning procedure, whichproduces a significant effect on the structure ofthe skin layer of the membrane at the nanoscalelevel.20 The result is an increased number ofmembrane pores where the spectrum of porediameters is narrowed and concentrated aroundthe desired values.21 This enables the increasedremoval of medium sized molecules such as ß2-microglobulin, with virtually no albumin leak atall.22 The membranes also span a range of

surface areas from 0.7 to 1.8 m2. They can belabeled low flux or high flux depending onwhether they are specific for the removal of ß2-microglobulin (high flux).

Therefore the history of the artificial kidney,from an experiment to a widely used therapy,closely follows the development of the dialysismembranes and the ability to manufacture themin reproducible conditions on an industrial scale.

Membrane Support Structure

The support structure allows the membranes toperform the functional role of dialysis betweenthe patients’ blood and the dialyzate. In orderto make dialysis a reality, the technology relat-ing to the support structure for the membraneshad to progress at the same rate as the mem-branes. In 1913, Able constructed the first dia-lysis machine it was extremely basic and wasthe basis for the machine that Haas used onhumans in 1928 (Figure 3-4).

Figure 3-3. A diagram of a capillary fiber membrane with its support structure and a photo of a membrane that has been cross sectioned.


In 1943, William Kolff, a Dutch physician con-structed the rotating drum dialyzer which was thefirst artificial kidney to be successfully used reg-ularly on humans.23 This device resembled adrum barrel made of slats with open spacesbetween the slats. Approximately 30–40 m ofcellulose acetate tubing was wrapped around a

small drum which was then partially immersedinto a 100 L bath of dialyzate (Figure 3-5).

The patient’s heart and own blood pressureforced blood into the cellulose acetate tubing. Asthe drum turned, it propelled the blood from oneend of the tubing to the other. This allowed thediffusion of molecules from blood to dialyzateand vice versa. The blood was then collected in aglass cylinder with an open nipple at the lowerend which was connected by rubber tubing tothe patient’s venous access. By alternating low-ering and raising the cylinder, blood was col-lected and drained back into the patient’s vein.

The procedure required a large volume ofblood circulation outside the body during dialy-sis and required priming with at least 2 units oftransfused blood.

Kolff gave away several of the rotating drumkidneys to promote the concept of dialysis.Some of the first published reports of patientstreated with this artificial kidney came from theRoyal Postgraduate Medical School at Hammer-smith Hospital, London.24

The innovative Canadian surgeon GordonMurray (1894–1976) is often forgotten when

Figure 3-4. The first artificial kidney used on humans developed by Haas in 1928.

Figure 3-5. A picture of the rotating drum dialyzer invented by Kolff.(Cannot find copyright owner.)


discussing the history of dialysis. He is creditedwith having performed the first successful hemo-dialysis in humans in North America but neitherhe nor Kolff were aware of each other’s workduring the mid-1940s when wartime hamperedcommunication. Murray’s extensive investiga-tions and experience in the use of heparin invascular surgery laid the groundwork for theuse of this anticoagulant with the artificial kid-ney. In 1945, he first designed a coil dialyzer inwhich cellophane tubing was wound about a steelframe. This machine, however, was confined to abasement after only a few treatments due to gen-eral lack of enthusiasm. In 1952, with resurgencein interest, he developed his second-generationapparatus. This was a plate dialyzer constructedwith the help of Walter Roschlau and indepen-dent of Leonard Skeggs and Jack Leonards whoare mentioned later. In all, he performed dialysison only 11 patients with presumed acute renalfailure, 5 of whom survived. He felt this was arelative failure and his treatments were dismissedby his colleagues. Murray finally lost interestin the artificial kidney when he discovered hisidea had been marketed independently by aGerman researcher, Erwin Halstrup, who hadworked with him for a while. The two machineswere locked away and only resurfaced after hisdeath.25,26

Until Nils Alwall produced his dialyzer in 1946there was no controlled way to remove fluid fromthe patients’ blood. This process, known as ultra-filtration, was achieved by using pressure tosqueeze water from the plasma through the dia-lyzer membrane. In previous machines pressurefrom the cardiovascular system alone was nothigh enough to instigate ultrafiltration but hadit been, the membranes would not have beenstrong enough to withstand it. The Alwall Dialy-zer was similar to the rotating drum. Approxi-mately 11 m of cellulose acetate tubing waswrapped around a stationary, vertical drummade of a metal screen. The membranes, in thissituation, could withstand higher pressuresbecause of their positioning between the metalplates of the screen. The membranes wereenclosed in a tightly closed cylinder so that pres-sure adjustments to the dialyzate, in addition tothe pressure coming from the cardiovascular sys-tem, could be tailored to control ultrafiltration.

In parallel to the increased availability of dia-lysis membranes two design variations began toemerge during the 1950s: (1) modifications ofthe original rotating drum dialyzer and (2) the

twin coil dialyzer and the new flat platedialyzers.

Rather than passing blood through membra-nous tubes, plate dialyzers directed the flow ofdialysis solution and blood through alternatinglayers of membranous material.

In 1948, Skeggs and Leonards developed thefirst parallel flow artificial kidney.27 The artifi-cial kidney was designed to have a low resistanceto blood flow and to have an adjustable surfacearea. Two sheets of membrane are sandwichedbetween two rubber pads in order to reduce theblood volume and to ensure uniform distribu-tion of blood across the membrane. Using multi-ple layers added to its efficiency. The device hada very low resistance to blood flow and it couldbe used without a blood pump.

A siphon on the effluent of the dialyzing fluidallowed water to be removed from the blood inthe artificial kidney. This is the first reference tonegative pressure dialysis. This is different toultrafiltration because negative pressure fromthe dialyzate is used to draw fluid off thehuman plasma rather than positive pressure inthe blood stream to push fluid out of the plasma.

Meanwhile, Kolff sent his blueprints to GeorgeThorn at the Peter Brigham Hospital in Boston,USA. During the Korean War in the 1950s, theKolff–Brigham dialyzer, a modification of theoriginal rotating drum dialyzer, was used totreat injured American soldiers (Figure 3-6).The membrane surface area could be adjustedby increasing or decreasing the number ofwraps of tubing and a hood was added to bettercontrol the blood temperature.28

Von Garrelts had constructed a dialyzer in1948 by wrapping a cellulose acetate membranearound a core. The layers of membrane wereseparated by rods. It was very bulky and weighedover 100 pounds and thus not used.

In 1952, Inouye and Engelberg took thisconcept and miniaturized it by wrapping thecellulose acetate tubing around a beaker andseparating the layers with fiberglass mesh. Heplaced this "coil" in a pressure cooker in order toenclose it and control the temperature. Therewere openings in the pot for the dialyzing fluid.With the use of a vacuum on the dialyzate leav-ing the pot, he was able to draw the excess waterout of the patient’s blood. A blood pump, assist-ing the patients’ blood pressure, was required toovercome resistance within the device.29 Bloodpumps are now an integral part of moderndialyzers.


This machine had to be abandoned, butusing some of the concepts of the pressurecooker kidney and the principles of the Alwallmachine, the Kolff–Travenol disposable twincoil dialyzer emerged in 1952. This device con-sisted of a twin coil seated in a big cylindricalsteel tank that could hold about 100 L of water.The twin coil was pre-assembled and sterilized,but it required a liter of blood to prime it andhad a high compliance to pressure changes. Hotwater and cold water were mixed together,when making the dialyzate, to get the correcttemperature as the block heater was only able tomaintain a given temperature. It also necessi-tated a blood perfusion pump and a significantarteriovenous pressure differential across themembranes developed, which resulted in anuncontrolled ultrafiltration. There were noalarms, temperature or pressure monitors, full

function checks, blood leak detectors, or airembolism detectors, all of which are standardon today’s machines. It was, however, robustand kept in use in Edinburgh until the late1980s!30

Concurrently, in 1952, Guarino and Guarinodeveloped an artificial kidney which wouldreduce the amount of blood outside of thebody and to eliminate the need for pumpingthe blood through the device. In reverse to pre-vious designs the dialyzing fluid was directedinside the cellulose acetate tubing and theblood cascaded down over the membrane hav-ing entered the device from the top. The metaltubing inside themembrane gave structural sup-port. The artificial kidney had a very low bloodvolume but it had limited use because there wasconcern regarding the possibility of the dialyz-ing fluid leaking into the blood.

Figure 3-6. A diagram representing the Bringham-modified Kolff kidney, better known as the Kolff Brigham dialyzer.


In 1960, Fredrik Kiil provided the definitiveplate dialyzer31 which was still used in someclinics until the 1990s. Three or more ‘‘Kiil’’boards were used with two sheets of, the newlyavailable, cuprophan membrane sandwichedbetween each pair of boards. The grooves inthe boards directed the blood between thelayers of membrane and the dialyzate outsidethe membrane envelope in opposite directionsfrom one end of the boards to the other. Thepriming volume was less than 300 ml, but oftenthe patients would have to construct the devicethemselves. This was complex and time con-suming confining the patient for an extra4 hours for any individual 10 hour dialysis ses-sion (Figure 3-7). As improvements in access tothe patients’ blood stream improved no bloodpump was needed to drive this device. Excessfluid was removed by the use of negative pres-sure on the dialyzate effluent line.

As the dialysis program expanded this type ofdevice was used for overnight, unattendedhemodialysis that was pioneered by BeldingScribner32 and his group in Seattle, USA. Atotally monitored dialyzate system was required

to automatically mix the dialyzate and to controltemperature and conductivity. This delivery sys-tem was developed by Albert Babb, while at theUniversity of Washington.

Following on from the knowledge gained withthe Kiil dialyzer, Babb designed and developedthe Milton-Roy Model A, built by the Milton RoyCompany in St. Petersburg, Florida in 1964. It wasalso designed to perform nocturnal home hemo-dialysis. It wasmade in a wooden veneer to have afurniture appearance for the home. It featuredautomatic hot water disinfection (up to 908c),automatic alarm checks, solid-state (diode)logic, and acoustic tile inside to reduce noise.

This instrument became the first of a series ofnegative pressure single patient systems, whichwas modified to use hollow fiber membranes, cul-minating in the hemodiaysis machines seen today.

The introduction of hollow fiber dialyzers,during the 1960s, marked a change in dialysismachines from their initial development bydedicated artisans, who were engineers and phy-sicians, to highly sophisticated machines andcomponents that rely on large production seriesand standardized manufacturing processes. The

Figure 3-7. Modified two layer Kiil dialyzer during hemodialysis (1964).


artificial kidney is now not just the componentthat ‘‘filters the blood’’ but also includes bloodaccess system, smooth and biocompatible bloodpathways and circuit design, standardized thick-ness and porosity of membranes with accuratepump, alarm, and failure systems. A myriad ofmachines are now produced by a number ofdifferent companies (Figure 3-8).

Blood Component

Anticoagulation

Preventing the blood from clotting once it hadentered the extracorporeal circuit of a dialysismachine was one of the first major barriers in thedevelopment of the artificial kidney. George Haasinitially used hiruidin, which is found in the salivaof leeches, as an anticoagulant for blood. Hiruidinwas not practical for human use as it caused aller-gic reactions and was difficult to purify.

Jay Maclean was a medical student who dis-covered heparin in 1915 while studying pro-thrombotic agents.33 He isolated a substancefrom the liver that seemed to prevent coagula-tion but his pharmacology professor JamesHowell controversially omitted his name fromthe two papers that first used the termheparin.34,35 Howell eventually realized thatheparin was more readily available from othertissues such as lung or intestine and not a phos-pholipid as originally thought, by this time,however, the name had already stuck! After pur-ification and standardization, heparin becamereadily available in 1937. William Thalhimer, asmentioned above, was the first person to useheparin in exchange transfusion for the allevia-tion of uremia16 and then for hemodialysis15 innephrectomized dogs. With this much moreeffective and less toxic anticoagulant, Kolffcould now develop hemodialysis as a realistictreatment modality.

Access to the Blood Stream

Intermittent chronic hemodialysis also requiresregular, reusable access to the patients’ blood inorder for it to pass through the artificial kidney.Modern hemodialysis processes 70–110 L ofblood in a 4–5-hour period, which requireseasy and reusable access to the blood stream.When Kolff first designed his machine the

methods for extracting blood and returning itto the body rapidly exhausted the patients’arteries and veins. This meant that initially

Figure 3-8. A ‘state of the art’ Fresenius 5008, one of the mostadvanced dialysis machines available today.


dialysis was only suitable for acute renal failureand the chronic health of the patient relied onthe recovery of renal function. A move towardchronic renal replacement therapy could onlyoccur with safe, reliable, and reusable access tothe patients’ blood supply.

Developments began in 1960 with the Scrib-ner shunt36 (Figure 3-9).37 Permanent cannulaewere placed in the vein and an artery of theforearm of a patient. When they were not beingused for dialysis, they were linked together witha Teflon tube to allow blood to flow across.

In 1966, Brescia and colleagues created thearteriovenous fistula (AVF) by joining an arteryin the arm with a nearby vein38 (Figure 3-10).39

The vein, which is normally part of a low-pres-sure system, becomes arterialized, swells, anddevelops a thickened wall. Needles can easilybe placed into the fistula, which is just underthe surface of the skin, to allow repeated access.The AVF remains the gold standard of perma-nent dialysis access today.

Semipermanent access such as the Perm-Cath1 catheter and the Tesio1 catheter arevenous cannulae which are tunneled under theskin from their point of entry to a large centralvein. These are often placed under fluoroscopyby physicians or radiologists and provide accessto the blood stream in patients’ who cannot havean AVF or when their AVF has failed. The mainadvantage of these forms of access is they can beused immediately. The main disadvantagesinclude repeated infections and limited life span.

As another alternative, the LifeSite1 Haemo-dialysis Access System combines unique valvetechnology with a cannulae connection systemto the blood stream. The LifeSite1 system con-sists of an implantable valve and cannulae. Thevalve is implanted subcutaneously and is

attached to a patient’s blood vessel via the venouscannulae and can be used immediately. The valveis made of medical grade materials that includetitanium alloy, stainless steel, and silicone elasto-mers. The cannulaematerial used in the LifeSite1

System is similar to other venous catheters. Thevalve has a dimpled entrance in the center of itsdomed top to accept a standard 14-gauge fistulaneedle (Figure 3-11).40 Insertion of the fistulaneedle opens the valve and allows blood flow.When the fistula needle is withdrawn, the valvecloses and blood flow stops. The dialysis needle isinserted into the same site each treatment result-ing in the formation of a fibrous tract or ‘‘button-hole’’. This access system is designed to eliminatethe disfigurement associated with AV fistulae andgrafts, as well as, the lifestyle impairments asso-ciated with protruding permanent catheters.

Dialyzate

Composition

The dialyzate is the chemical bath that is used todraw waste products out of the blood. Diffusionmoves waste products in the blood across themembrane into the dialyzate compartment,where they are carried out of the body. At thesame time, electrolytes and other chemicals in thedialyzate solution cross the membrane into theblood compartment. The purified, chemicallybalanced blood is then returned to the body.

Since the inception of a chronic dialysis pro-gramme, there has been a realization that thecomposition of the dialyzate is as important asall the other components of the artificial kidney.In 1913, when Abel and Rowntree first trieddialysis using animal blood the dialyzate was a

Figure 3-9. A Scribner shunt inpatients forearm.


solution of normal saline and potassium.Gordon Murray, in his few attempts at dialysisrealized that he had to use a dialyzate thatcontained physiological similarities to plasmaand settled for Ringers solution.26 For manyyears, dialyzate was mixed by hand inlarge vats that the various membranes passedthrough. As the membranes and their supportstructures improved, it could be mixed in a largecommunal vat and actively passed across a num-ber of membranes at once.

Now dialyzate solutions are mixed to variousprescriptions by the dialysis machines and thebuffer added after mixing with a sterile watersupply. The concentrations of sodium, potas-sium, and calcium are carefully chosen in thedialyzate to match the patients need. Along withthe buffer these make up the most importantfeatures of the dialyzate. We know that the cor-rect prescription of dialyzate can improve thecardiovascular stability of patients receivinghemodialysis treatment.41

Another important role of the dialyzate is tocorrect the patients’ metabolic acidosis. This isdone by the diffusive influx of the buffer supply,mainly bicarbonate, rather than through hydro-gen ion clearance. Because of precipitation withcalcium andmagnesium and the risk of bacterialcontamination, bicarbonate was abandoned andreplaced by acetate during the first two decadesof dialysis therapy. The key advantages were its

equimolar conversion to bicarbonate, a bacter-iostatic effect, and low cost. However, acetate-induced side-effects have been reported in alarge number of studies during the 1980s dueto the limitations in hepatic acetate metabolismin some patients.42 In the last two decades, tech-nical improvements, avoiding carbonate saltprecipitation, and the development of highflux/high efficiency dialysis with worsened acet-ate-induced side-effects have gradually led to thereintroduction and generalization of bicarbo-nate as the preferred dialyzate buffer.

Direction of dialyzate flow as well as its com-position is important. The Skeggs LeonardsPlate Dialyzer was the first to push bloodthrough the blood compartment in one direc-tion and suction or vacuum pressure pulls thedialyzate through the dialyzate compartment inthe opposite direction.27 This countercurrentsystem is the most efficient mechanism of ultra-filtration and works to drain excess fluid out ofthe bloodstream and into the dialyzate and isnow standard practice.

Water

Water has an enormous role in the delivery ofsafe, good dialysis. In modern machines, thedialyzate is pumped over the membrane atapproximately 800 ml per minute.

Figure 3-10. A left brachial arterio-venous fistula (AVF).


For a unit that dialyses 90 patients a day,spread over three separate shifts, approximately17,000 L of water will pass across the dialysismembranes and patients’ blood in that period. Itis therefore imperative that the water is free ofinfection and as pure as possible. This is a mainfocus within dialysis units, who have their ownpurification systems with national standardsand strict quality control systems that have tobe adhered to. In the 1970s, as dialysis becamean established form of therapy, nephrologistsrealized the importance of water through a seriesof clinical incidents.

Sporadic outbreaks of encephalopathy inseveral dialysis units around the countrywere associated with aluminium toxicity.43

Aluminium hydroxide, along with filtration,is used to improve the taste of ‘‘peaty’’ waterand subsequently throughout the country, thedrinking water aluminium concentration var-ied. This water was then used to mix thebuffer solution of the dialyzate. To overcomethis problem mains water is now treatedusing softeners, reverse osmosis, or ionexchange to remove the aluminium.44 Waterand plasma aluminium content is monitoredroutinely as part of the quality control sys-tems and ‘‘dialysis dementia’’ in the UnitedKinhdom is a thing of the past.

Similarly hemolytic anemia was identifiedas a problem in hemodialysis patients in the1970s.45,46 Chloramine is formed when ammo-nia is added to water that contains free chlor-ine. This process is used by water companies todisinfect drinking water. Chloramine lasts along time in water to more effectively removepathogens such as bacteria and virus’. Waterthat contains chloramine is safe for people todrink, bathe, and cook in because the digestiveprocess neutralizes it. Chloramine can, how-ever, easily harm patients if it enters theblood stream during the dialysis process caus-ing hemolytic anemia. Recently, recombinanterythropoetin resistance has been recognizedas an insidious presentation of chloraminetoxicity.47

Chloramine continues to cause problems indialysis centers48–50 and often it is because itsconcentration in the water tends to varythroughout the year. Levels are increased bywater companies when bacterial contaminationincreases and the carbon filters, which easilyremove chloramine, subsequently becomeinsufficient.

The Future

Renal transplantation is the preferred treat-ment for many people with chronic renalfailure; however, due to lack of donororgans and the unsuitability of a significantproportion of patients, hemodialysis willcontinue to be a common therapy. In theUnited Kingdom, hemodialysis commonlyoccurs in 4-hour sessions, three times aweek and despite the progress that hasbeen made in the last 50 years, hemodialysisis widely recognized as unphysiologic51 andcontinues to be associated with high mor-bidity and mortality.

Since the invention of hollow fiber dialyzersthe major areas of development have been newmaterials with enhancement of the delivery indialysis using better machines cumulating inimproved efficiency. We now provide safer andbetter tolerated dialysis but without the decreasein mortality we would desire.52,53

An ideal device for renal replacement therapywould mimic the function of the natural kidney(Table 3-3). It would be continuously operating,it would remove solute with a molecular weightspectrum, it would be flexible and remove waterand solutes on an individual patient need, itwould be wearable or implantable, and it mustbe biocompatible.

Nissenson and collegues54 recently pub-lished an article about their vision for thefuture of the artificial kidney. In order toachieve the ambitious specifications theyhave pursued nanotechnology in the devel-opment of their artificial kidney, the HumanNephron Filter (HNF). As mentioned above,the hollow fiber membranes are alreadyintroducing aspects of this technology inartificial kidneys.

Table 3-3. Criteria for the Ideal Artificial Kidney

• Continuous• Remove solutes with a wide range of molecular weights• Function relative to individual patient needs• Portable/wearable/implantable• Biocompatible• Light weight• Low cost

• Safe• Reliable


The HNF consists of two membranes oper-ating in series within one device. The G mem-brane mimics the function of the glomerulus,using convection to generate plasma ultra fil-trate containing all solutes approaching themolecular weight of albumin. The T membranemimics the functions of the renal tubule, selec-tively reclaiming designated solutes to main-tain body homeostasis, again by convection(Figure 3-12)54

Although only in developmental stagesthe authors have an animal prototype andmany of the breakthrough aspects such asthe chemistry, the pores, and the membranesare almost in place.

Progress continues to be made with livingmembranes. The ultimate goal of this approachis to replace the metabolic activity of the kidneyto improve biological homeostasis, includingimmunologic, cardiovascular, and mineralmetabolism. This all depends on the ability toisolate and expand in culture tubule cells fromadult kidneys55 and then incorporate themonto a standard hemofiltration cartridge.56 Invitro and ex vivo animal experiments56–58 haveallowed an initial trial of the ‘‘renal tubule cellassist device’’ (RAD) in critically ill humansreceiving continuous veno-venous hemofiltra-tion in the ICU with encouraging results59 lead-ing to a phase II trial.

Figure 3-11. The LifeSite1 Haemodialysis Access System.


Conclusions

The development artificial kidney has pro-gressed rapidly since its application to humanswhile the need for hemodialysis has continued togrow. The availability and variety are such thatthe nephrologist can tailor dialysis to the phy-siological need of the patient by using differentcombinations of membranes, dialyzate solu-tions, and pump speeds. There is, however, stillprogress to bemade inmorbidity, mortality, andquality of life for hemodialysis patients but thefuture of the artificial kidney looks exciting asnew technologies emerge.

References

1. Ansell D, Feest T, Tomson C, Will E, Williams AJ. UKRenal Registry Report 2005.

2. Johnson R, Feehally J. Comprehensive Clinical Nephrol-ogy. London: Elsevier; 2003.

3. Graham T. The Balkerian Lecture: On Osmotic Force.Phil Trans R Soc Lond. 1854;144:177–228.

4. Fick A. Ueber Diffusion. Annalen derPhysik un Chemie.1855;170:59–86.

5. Abel J, Rowntree L, Turner B. On the removal of diffu-sible substances from the circulating blood of livinganimals by means of dialysis. J Pharmacol Exp Ther.1912;5:275–283.

6. Haas G. Dialysieren des stromendes Blutes am Leben-den. Klin Wochenschr. 1925;2:1888.

7. Haas G. Versuche der Blutauswaschung am Lebendenmit Hilfe der Dialyse. Klin Wochenschr. 1925;4:13–14.

8. Haas G. Ueber Versuche der Blutauswaschung amLebenden mit Hilfe der Dialyse. Archiv Exp Path Phar-makol. 1926;106:158–172.

9. Haas G. Die Methoden der Blutauswaschung. Aberhal-den’s Handb Biol Arbeitsmethoden Abt V teil.1935;8:717–754.

10. George C.R.P. Hirudin, heparin and Heinrich Neceles.Nephrology. 1998;4:225–228.

11. ‘En effet, il y a dans les bois le tissu primitif, isomere avecamidon, que nous applerons cellulose, et de plus unematiere qui en remplit les cellules, et qui constitue lamatiere ligneuse veritable’. [The committee was formedof Brogniart, Pelouze and Dumas, and reported to theAcademie on 14th January 1839: C R Acad Sci Paris1839;VIII: 51. It appears that the author of the word glucose,and presumably also cellulose, was Dumas, the rappor-teur of the committee.]. In: eds. Book ‘En effet, il y a dansles bois le tissu primitif, isomere avec amidon, que nousapplerons cellulose, et de plus unematiere qui en remplitles cellules, et qui constitue la matiere ligneuse veritable’.[The committee was formed of Brogniart, Pelouze andDumas, and reported to the Academie on 14th January1839: C R Acad Sci Paris1839; VIII: 51. It appears that theauthor of the word glucose, and presumably also cellu-lose, was Dumas, the rapporteur of the committee.].

12. Wilson F.L. Experiments with cellophane as a sterilisabledialysablemembrane.Arch Pathol LabMed. 1927;127:239.

13. Andrus F.C. Use of Visking sausage casing for ultrafil-tration. Proc Soc Exp Biol Med. 1929;27:127–128.

14. Fagette P. Hemodialysis 1912–1945: no medical technol-ogy: before its time: part I. ASAIO J. 1999;45:238–249.

15. Thalhimer W, Solandt DY, Best CH. Experimentalexchange transfusion using purified heparin. Lancet.1938;2:554–557.

16. Thalhimer W. Experimental exchange transfusions forreducing azotemia: Use of artificial kidney for this pur-pose. Proc Soc Exp Biol Med. 1938;37:641–643.

17. Stewart RD, Cerny JC, Mahon HI. The Capillary "Kid-ney": Preliminary Report. Med Bull (Ann Arbor).1964;30:116–118.

Figure 3-12. A diagram showing how the human nephron filter utilizes two membranes to mimic the renal tubule.


18. Stewart R, Cerny J, Mahon H. An artificial kidney madefrom capillary fibres. Invest Urol. 1966;5:614–624.

19. Streicher E, Schneider H. Polysulphone membranemimicking human glomerular basement membrane.Lancet. 1983;2:1136.

20. Drexler KE. Nanosystems molecular machinery, manu-facturing and computation. Chichester: Wiley; 1992.

21. Bowry SK. Nano-controlled membrane spinning tech-nology: regulation of pore size, distribution andmorphology of a new polysulfone dialysis membrane.Contrib Nephrol. 2002;137:85–94.

22. Bowry SK. Dialysis membranes today. Int J Artif Organs.2002;25:447–460.

23. Kloff WJ, Berk HTJ, ter Welle M, van der Ley AJW, vanDijk EC, van Noordwijk J. The artificial Kidney: a dialy-ser with a great area. Acta Med Scand. 1944;117:121–134.

24. Bywaters EGL, Joekes AM. Artificial kidney: its applica-tion to treatment of traumatic anuria. Proc Roy Soc Med.1948;41:420–426.

25. ClarkeW. A Canadian giant: Dr. GordonMurray and theartificial kidney. CMAJ. 1987;137:246–248.

26. McKeller S. Gordon Murray and the artificial kidney inCanada. Nephrol Dial Transplant. 1999;14:2766–2770.

27. Skeggs LT, Jr., Leonards JR. Studies on an ArtificialKidney: I. Preliminary Results With a New Type of Con-tinuous Dialyzer. Science. 1948;108:212–213.

28. Murphy WP, Jr., Swan RC, Jr., Walter CW, Weller JM,Merrill JP. Use of an artificial kidney. III. Current pro-cedures in clinical hemodialysis. J Lab ClinMed. 1952;40:436–444.

29. Inouye WY, Engelberg J. A simplified artificial dialyzerand ultrafilter. Surg Forum. 1953;4:438–442.

30. Wong D, Lambie AT, Winney RJ, Turner AN. Haemo-dialysis in Edinburgh 1957–1987. J R Coll PhysiciansEdinb. 2002;32:114–121.

31. Kiil F. Development of a parallel-flow artificial kidney inplastics. Acta Chir Scand Suppl Suppl. 1960;253:142–150.

32. Scribner BH, Buri R, Caner JE, Hegstrom R, Burnell JM.The treatment of chronic uremia by means of intermit-tent hemodialysis: a preliminary report. Trans Am SocArtif Intern Organs. 1960;6:114–122.

33. Maclean J. The discovery of heparin. Circulation.1959;19:75–78.

34. Howell WH. Heparin, an anticoagulant. Am J Physiol.1923;63:434–435.

35. Howell WH, Holt E. Two new factors in blood coagula-tion: Heparin and pro-antithrombin. Am J Physiol.1918;47:328–341.

36. Quinton W, Dillard D, Scribner BH. Cannulation ofblood vessels for prolonged hemodialysis. Trans AmSoc Artif Intern Organs. 1960;6:104–113.

37. Scribner BH. Lasker Clinical Medicine Research Award.Medical dilemmas: the old is new. Nat Med.2002;8:1066–1067.

38. Brescia MJ, Cimino JE, Appel K, Hurwich BJ. Chronichemodialysis using venipuncture and a surgically createdarteriovenous fistula. N Engl J Med. 1966;275:1089–1092.

39. Johnson R, Feehally J. Comprehensive Clinical Nephrol-ogy. London: Elsevier; 2003.

40. Schwab SJ, Weiss MA, Rushton F, Ross JP, Jackson J,Kapoian T, Yegge J, Rosenblatt M, Reese WJ, Soundar-arajan R, Work J, Ross J, Stainken B, Pedan A, MoranJA. Multicenter clinical trial results with the LifeSitehemodialysis access system. Kidney Int. 2002;62:1026–1033.

41. Locatelli F, Covic A, Chazot C, Leunissen K, Luno J,Yaqoob M. Optimal composition of the dialysate, withemphasis on its influence on blood pressure. NephrolDial Transplant. 2004;19:785–796.

42. Vinay P, Cardoso M, Tejedor A, Prud’homme M,Levelillee M, Vinet B, Courteau M, Gougoux A,Rengel M, Lapierre L, et al. Acetate metabolism duringhemodialysis: metabolic considerations. Am J Nephrol.1987;7:337–354.

43. Alfrey AC, LeGendre GR, KaehnyWD. The dialysis ence-phalopathy syndrome. Possible aluminum intoxication.N Engl J Med. 1976;294:184–188.

44. Petrie JJ, Fleming R, McKinnon P, Winney RJ, Cowie J.The use of ion exchange to remove aluminum fromwater used in hemodialysis. Am J Kidney Dis.1984;4:69–74.

45. Botella J, Traver JA, Sanz-Guajardo D, Torres MT, San-juan I, Zabala P. Chloramines, an aggravating factor inthe anemia of patients on regular dialysis treatment. ProcEur Dial Transplant Assoc. 1977;14:192–199.

46. Eaton JW, Kolpin CF, Swofford HS, Kjellstrand CM,Jacob HS. Chlorinated urban water: a cause of dialysis-induced hemolytic anemia. Science. 1973;181:463–464.

47. Fluck S, McKane W, Cairns T, Fairchild V, Lawrence A,Lee J, Murray D, Polpitiye M, Palmer A, Taube D.Chloramine-induced haemolysis presenting as erythro-poietin resistance. Nephrol Dial Transplant.1999;14:1687–1691.

48. Calderaro RV, Heller L. [Outbreak of hemolytic reac-tions associated with chlorine and chloramine residualsin hemodialysis water]. Rev Saude Publica.2001;35:481–486.

49. Pyo HJ, Kwon YJ, Wee KS, Kwon SY, Lee CH, Kim S, LeeJS, Cho SH, Cha CW. An outbreak of Heinz body positivehemolytic anemia in chronic hemodialysis patients. Kor-ean J Intern Med. 1993;8:93–98.

50. Tipple MA, Shusterman N, Bland LA, McCarthy MA,Favero MS, Arduino MJ, Reid MH, Jarvis WR. Ill-ness in hemodialysis patients after exposure tochloramine contaminated dialysate. ASAIO Trans.1991;37:588–591.

51. Kjellstrand CM, Evans RL, Petersen RJ, Shideman JR,von Hartitzsch B, Buselmeier TJ. The "unphysiology" ofdialysis: a major cause of dialysis side effects?Kidney Int.1975;Suppl:30–34.

52. Ronco C, Ghezzi PM, La Greca G. The role of technologyin hemodialysis. J Nephrol. 1999;12(Suppl 2):S68–81.

53. Woffindin C, Hoenich NA. Hemodialyzer performance:a review of the trends over the past two decades. ArtifOrgans. 1995;19:1113–1119.

54. Nissenson AR, Ronco C, Pergamit G, Edelstein M, WattsR. Continuously functioning artificial nephron system:the promise of nanotechnology. Hemodial Int.2005;9:210–217.

55. Humes HD, Krauss JC, Cieslinski DA, Funke AJ. Tubu-logenesis from isolated single cells of adult mammaliankidney: clonal analysis with a recombinant retrovirus.Am J Physiol. 1996;271:F42–49.

56. Humes HD. Tissue engineering of a bioartificial kidney:a universal donor organ. Transplant Proc. 1996;28:2032–2035.

57. Humes HD, Buffington DA, MacKay SM, Funke AJ,Weitzel WF. Replacement of renal function in uremicanimals with a tissue-engineered kidney.Nat Biotechnol.1999;17:451–455.


58. Humes HD, Fissell WH, Weitzel WF, Buffington DA,Westover AJ, MacKay SM, Gutierrez JMMetabolic repla-cement of kidney function in uremic animals with abioartificial kidney containing human cells. Am J KidneyDis. 2002;39:1078–1087.

59. HumesHD,WeitzelWF, Bartlett RH, Swaniker FC, Paga-nini EP, Luderer JR, Sobota J. Initial clinical results of thebioartificial kidney containing human cells in ICUpatients with acute renal failure. Kidney Int.2004;66:1578–1588.


4Liver Substitution

Sambit Sen and Roger Williams

Introduction

Liver failure, whether occurring de novo withoutpre-existing liver disease (acute liver failure,ALF) or as an acute episode of decompensationsuperimposed on a chronic liver disorder(acute-on-chronic liver failure, ACLF), carries ahigh mortality. The lack of the detoxification,metabolic, and regulatory functions of the liverleads to life-threatening complications, includ-ing kidney failure, hepatic encephalopathy (HE),cerebral edema, severe hypotension, and sus-ceptibility to infections culminating in multi-organ failure.1,2 The only established therapyfor patients of ALF who fail to improve withsupportive management is liver transplantation(LTx), but currently one-third of these patientsdie while waiting for a transplant and the organshortage is increasing.3 Living-related donortransplantation has proved encouraging,4 yet asuitable donor is found in time in only 30% ofcases (of living-related donor transplantation).5

Many patients of ACLF are not eligible for LTx,and current management focuses on treatmentof the individual organ dysfunction. However,liver failure, whether of the acute or acute-on-chronic variety, is to some extent reversible. Asupportive therapy which can tide over the acuteperiod of crisis (and act as a bridge to livertransplantation in cases of ALF) can possiblybe life-saving,6 and the search for an effectiveliver support system continues. Essentially, twotypes are under development: artificial devices,based on the removal of albumin-bound toxins,and bioartificial devices, using hepatocytes toperform the functions of the failing liver.

However, one has to first consider whetherliver support can be expected to work and pro-vide worthwhile clinical benefit in the setting of

liver failure in man. A recent review looks at thehistory of the development of liver support,elegantly tracing the quest for the elusive idealliver support system.7 Early studies demon-strated that cross-circulation of the patientwith baboons or healthy human subjects led torecovery from coma.8,9 Similarly, extracorporealanimal liver perfusion was found to be of helpin supporting ALF patients,10,11 as was singleplasma exchange therapy.12 Amongst the manyother therapies tried, hemoperfusion over char-coal and, more recently, high-volume plasma-pheresis13 (which showed reduction of plasmabilirubin and arterial ammonia,14 with improve-ment of peripheral hemodynamics15 and cere-bral blood flow16) have all been the subject ofencouraging results, although with charcoalhemoperfusion this could not be confirmed ina controlled trial. Most substances are non-selectively removed by these various techniques,including some which are important physiolo-gically (growth factors, clotting factors),17,18 andwith charcoal hemoperfusion, the filters causingactivation of leucocytes with consequent cyto-kine release may have predisposed to bleedingand disseminated intravascular coagulation.19

There also exists a vast amount of publishedliterature relating to the testing of bioartificialliver (BAL) devices in in vitro and animal mod-els.19–23 Thus studies with rat models of liverfailure have shown improvement of parameterslike serum bilirubin, albumin, and clottingfactor production. A carefully controlled studywith the Academic Medical Center-BAL(Amsterdam) in pigs with hepatic ischemia-induced ALF showed improvement of bloodammonia and bilirubin levels, with prolongationof survival (longest surviving animal kept aliveup to 63 h).24 A subsequent controlled study


57

performed in anhepatic pigs with the samedevice25 showed that those treated with BALcontaining autologous hepatocytes survivedlonger (mean 65 h) compared to animals nottreated with BAL (46 h) or treated with the BALdevice without incorporating hepatocytes (43 h).The treatment was also associated with a signifi-cant decrease of arterial ammonia. What isperhaps surprising and certainly worth com-menting upon is that in the long history ofdeveloping devices for liver support, despitebeneficial effects in various animal models(large and small) of ALF, most of the deviceshave remained unproven when used subse-quently in patients with ALF.

Functions Required of a LiverSupport System

Liver failure leads to impairment of all aspects ofhepatic function, including detoxifying and syn-thetic functions, as well as bile excretion. Alteredhepatic metabolism is probably central to this.26

For example, impaired gluconeogenesis and gly-cogenolysis predisposes to hypoglycemia. Albu-min synthesis is hampered, as is the productionof clotting factors. The impaired urea cycle leadsto accumulation of ammonia in the body, and themetabolism of protein breakdown products isalso impaired. This leads to the accumulation ofa whole variety of toxins in the body, most ofwhich, apart from a few like ammonia and lactate,are albumin-bound in the plasma (Table 4-1).This means that they cannot be adequately

removed by conventional techniques of hemofil-tration/hemodialysis. The importance of thesetoxins may be related to their role in the devel-opment of associated end-organ failures, espe-cially affecting the brain, the circulation, and thekidneys, which considerably worsen the outcomein both ALF and ACLF.

The observation that these end-organ dys-functions can be reversed following LTx has ledto the hypothesis that the hepatic dysfunction iscentral to their pathogenesis, with the accumu-lation of toxins as the immediately responsiblemechanism.1 A liver support system which canremove these toxins could potentially prevent orreverse end-organ failure, and significantlyimprove outcome in liver failure, particularly ifthese toxins also cause further damage to theliver.27 Clearance of these toxins from the bodycould potentially interrupt this vicious cycleleading to further hepatic function deteriorationas well as end-organ failure. A system favorablymodulating humoral and molecular mechan-isms of liver regeneration (with reduction ofplasma levels of regeneration-inhibitory sub-stances like transforming growth factor (TGF)b) is ideally desirable,28 as would one whichcould reasonably duplicate the role of the liverin the biotransformation and metabolism of var-ious drugs and intermediate metabolites. Whilereplacing lost biosynthetic functions wouldalso be important, most substances producedby the liver can be given exogenously by intra-venous infusion, and this particular aspect ofliver synthetic function may not be of primaryconcern.29

Assessment of Current LiverSupport Systems

The several liver support systems that are cur-rently under investigation are considered underthe two main categories of artificial and bioarti-ficial. Despite natural concerns to try every pos-sible therapy in an individual patient suddenlystruck down by liver failure, especially ALF, thevalue of such devices can only be properlyassessed in the setting of a controlled clinicaltrial. Such trials need to be adequately poweredin the number of cases, with strict criteria forentry and with survival as the primary end point.However, the critical clinical condition of liverfailure patients as well as the rarity of ALF gen-erally makes performing such a study difficult.

Table 4-1. Some of the Endogenous Albumin-Bound Toxins, WhichAccumulate in Liver Failure, Which Can Be Potentially Removed byAlbumin Dialysis or Fractionated Plasma Separation and Adsorption

Aromatic amino acidsBile acidsBilirubinDigoxin-like substancesEndogenous benzodiazepinesIndolsMercaptansMiddle- and short-chain fatty acidsNitric oxidePhenolsProstacyclinsTryptophan


Patients with different degrees of severity of thesyndrome are often grouped together withintrials, and intensive care regimes are difficultto standardize. Furthermore LTx, when indi-cated, is likely to be performed at differenttimes in the clinical course of patients and willhugely influence survival. Thus surrogate mar-kers, rather than survival, are often used inassessing the efficacy of therapy. All these needto be kept in mind when interpreting the resultsof clinical trials of liver support systems. Thereason why ALF figured prominently in pre-vious studies was because it was consideredthat the chances of survival were so low thatany possible therapy could be justified. Therecent trials on artificial systems have been inthe much more common condition of ACLF,which is the major cause for liver failure and isresponsible for increasing numbers of hospitaladmissions worldwide. Therefore, one wouldexpect ACLF to be the main clinical indicationwhere liver support would be required.

Bioartificial Systems

Design of Devices

By utilizing viable hepatocytes these systemsshould reproduce the synthetic, detoxifying aswell as excretory functions of the liver. Unfortu-nately human liver cells, which would be the bestto use, are difficult to obtain and are difficult togrow in culture, becoming phenotypically unstableand rapidly losing liver-specific gene expression.The minimum quantity of cells required is notknown, but based on the experience of hepaticresection in humans,30 around 150–450 g of cells(or 1010 hepatocytes), providing the function of10–30% of the normal liver mass, is required tosupport the failing liver19,29 (though some experi-mental data in animal models have suggestedthat smaller quantities of cells may be enough toenhance spontaneous recovery, possibly by redu-cing plasma levels of regeneration-inhibitory sub-stances like TGF-b31). Most of the current devicesare based on the use of hepatocytes from otherspecies, particularly pigs (as in the Demetriou BALdevice HepatAssist (Circe Biomedical, Lexington,MA), using 6–10� 109 cells, or 30–50 g).23 A greatadvantage of porcine hepatocytes, unlike humanliver cells, is that they can be satisfactorily cryo-preserved, with cell isolation at a convenient timefollowed by storage at a clinical site prior to use,

thereby avoiding the costs and contaminationrisks of long-term hepatocyte culture.32 Cryopre-servation is associated with caspase activation andapoptosis, which can be prevented by adding cas-pase inhibitors (ZVAD-fmk).33 Another approachhas been with genetically engineered human hepa-tocytes to produce cells with the desired functionaland survival capabilities. Thus, the C3A hepato-cyte line, a sub-clone of the ubiquitous HepG2hepatoblastoma cell line, has been used in onesystem (4 � 1010, or 200 g in each bioreactorcartridge of the Sussman device (extracorporealliver assist device (ELAD, Vitagen, La Jolla,CA)).22 Another immortalized human hepatocytecell line under investigation is HHY41, whichretains many liver-specific functions, proteinsynthesis, gluconeogenesis, and cytochrome P450activity and is particularly resistant to acetami-nophen.34–36

At the heart of the design for a BAL is thebioreactor (Figure 4-1). The simplest type, andone used most commonly, consists of a columncontaining hollow fiber capillaries throughwhich flows the patient’s plasma. In the extra-capillary space lie the hepatocytes, either alone22

or attached to microcarrier beads.23 Plasma canbe separated, warmed, and oxygenated in thesecondary circuit before being perfused throughthe bioreactor capillaries. Free exchange ofmolecules can occur between plasma/blood andhepatocytes in the bioreactor across a mem-brane with a cut-off selected to allow themovement of most toxins as well as transportproteins like albumin (66 kDa molecular weight),while preventing passage of immunoglobulins(100–900 kDa), complements (>200 kDa), orviruses and cells. Most groups have used cut-offs between 50 and 150 kDa.2,19 The hepatocytesextract oxygen and nutrients and detoxify toxinsfrom the plasma, and their metabolites are simul-taneously passed back into the plasma. TheDemetriou system also incorporates two charcoalcolumns in the circuit prior to the bioreactor forremoval of toxins which could damage or impairthe function of the pig hepatocytes.

The more sophisticated system developed byGerlach et al.37 (Modular Extracorporeal LiverSupport (MELS, Hybrid Organ, Berlin, Germany))uses three sets of capillary tubes – one to provideoxygenation, and two to carry inflowing and out-flowing plasma. The hepatocytes (400–600 g ofprimary human hepatocytes from explanted liversfound unsuitable for LTx) remain in the extraca-pillary space. A detoxification module allows

Liver Substitution 59

single-pass albumin dialysis to be performed, andcontinuous veno-venous hemodiafiltration can beincluded. The Academic Medical Center-BAL(AMC-BAL, Hep-Art, Amsterdam, The Nether-lands) developed by Chamuleau et al.38 incorpo-rates a spirally wound polyester matrix sheet thatincludes an integral hollow fiber compartment foroxygenation and uses 220 � 106 porcine hepato-cytes (Figure 4-2). In addition to these hollowfiber-based bioreactors, some others have trieddesigns based on ‘‘flat plates and monolayers,’’‘‘perfused beds/scaffolds,’’ and ‘‘encapsulationand suspension.’’19 A porcine hepatocyte-basedBAL using a ‘‘radial-flow’’ bioreactor (RFB) hasbeen developed in Italy39 and has undergonephase I clinical trials.

Clinical Trials

The first clinical use of a BAL device, comprisinga kidney dialysis unit loaded with cryopreservedrabbit hepatocytes, was in 1987 to treat a single

patient with ALF. The serum bilirubindecreased, neurological status improved, andthe patient was ultimately discharged from hos-pital.40 Of the many clinical studies of the differ-ent devices since then, only those relevant tocurrently available devices are included in thefollowing assessment. The numerous difficultiesassociated with performing such trials in thisextremely sick group of patients have alreadybeen referred to.

Porcine Hepatocyte-Based BAL

One of the first trials using this system (1996),which was developed by Demetriou and collea-gues at the Cedars-Sinai Medical Center, USA,described 12 ALF patients, all of whom could bebridged over to LTx.41 Following treatment last-ing for 21–96 h, reversal of decerebration wasobserved in all 11 patients who were in grade 4encephalopathy, with significantly improved

Figure 4-1. Schematic showing the structure of bioreactors in Sussman’s extracorporeal liver assist device (ELAD) (top) and Demetriou’s HepatAssistbioartificial liver (BAL) (bottom).


intracranial pressure (from a mean of 18.2 to8.5 mmHg) and cerebral perfusion pressure.While improvement of the neurological state wasmost prominent, some important biochemicalchanges were noted as well. Serum ammonia(155.6–121.6 mmol/l) and bilirubin (21.6–18.2 mg/dl) improved, although the prothrombin time didnot. Interestingly, two patients who underwenttotal hepatectomy to reduce the necrotic tissuemass were successfully maintained on BAL untilthey were transplanted. The results were less pro-mising in eight ACLF patients, six of whom diedwithin 14 days of the last BAL treatment. In anotherreport by the same group (1997)42 data were given on18 ALF patients treated with BAL, of whom 16 couldbe successfully bridged over to transplantation, while1 patient recovered sufficiently not to need it.

A more recent French study with BAL (using5 � 109 porcine hepatocytes in each bioreactor)in 10 ALF patients43 again reported changespredominantly in the neurological state, withimprovement of Glasgow Coma Score in 6(mean of 6.5–9.6). Serum bilirubin level was

reduced (486.6–347 mmol/l), without any otherchanges in liver function tests. Arterial ammoniadid not improve significantly, and there was noevidence of improved protein synthesis. On thecontrary, serum albumin worsened significantly(27.8–22.7 g/l) after the first session of BAL treat-ment. Five patients had bleeding complications,probably explained by the significant reductionof both platelets (194.7–140 � 109/l) and fibrino-gen (0.96–0.47 g/l) after the first session. Sixpatients had transient episodes of hemodynamicinstability, although ultimately all patients couldbe bridged over to transplantation, and eightwere alive and well at 18 months.

The device was further developed to make useof cryopreserved cells and include a larger num-ber of hepatocytes (HepatAssist, CirceBiomedical), and has been evaluated in a largemulti-center randomized controlled trial in 171patients (fulminant hepatic failure – 147, pri-mary graft non-function – 24), conducted inthe United States and Europe.44 While someimprovements in intracranial pressure and con-sciousness level were seen, there was littleevidence of improved synthetic function. Mostdisappointingly, a survival advantage was evi-dent only in the sub-group with acetaminophenetiology (n¼39) (BAL – 70% vs controls – 37%).Thirty-day survival in the entire study popula-tion was 62% for controls vs 71% for BAL-treated patients, while among fulminant liverfailure patients alone, it was 59% vs 73%, respec-tively (p¼0.1). However, this primary end pointwas confounded by the major impact of LTx;54% of the entire study population was trans-planted. Thirty-day survival with LTx (n¼90)was 84% (BAL – 89% vs controls – 80%), whilethat without LTx (n¼81) was 46% (BAL – 51% vscontrols – 40%). A covariate time-dependentproportional hazard model with time-to-deathas end point was therefore employed to accountfor the impact of LTx and other factors predic-tive of survival (including etiology, stage ofencephalopathy) and showed a 47% reductionin mortality favoring BAL treatment (p¼0.03) inthe fulminant hepatic failure group (n¼147).

Hepatoblastoma-Based ExtracorporealLiver Assist Device (ELAD)

The ELAD device developed by Sussman andcolleagues was tried in one early randomizedcontrolled study in London45; 24 ALF patients

Figure 4-2. Schematic of Chamuleau’s AMC-bioartificial liver.38 (A)polysulfon dialysis housing; (B) three-dimensional non-woven polyestermatrix for high-density hepatocyte culture; (C) polypropylene hollowfiber membranes for oxygen supply and CO2 removal; (D) hollow fibersact as spacers between layers of the 3D matrix, creating numerouschannels for the uniform flow and distribution of medium; (E) endcapsvia which bioreactor is perfused with culture gas; (F) side-ports viawhich medium is perfused through the extra-fiber bioreactor space andis in direct contact with hepatocytes.


were enrolled, comprising 17 patients not ful-filling criteria for transplantation (predictedsurvival 50%) – group-I, and 7 patients fulfillingcriteria (predicted survival <10%) – group-II.The device proved safe, without causing coagu-lopathy or hemodynamic instability to any sig-nificant extent. Arterial ammonia fell marginallyin the ELAD group after 6 h treatment, and therise in serum bilirubin over the period of thetrial was more pronounced in the controls, butnot significantly so. Worsening of encephalopa-thy was less in the ELAD-treated patients (3/12,25%) compared to controls (7/12, 58%). How-ever, a clear survival advantage was not shown.Survival in group-I was 7/9 (78%) with ELADand 6/8 (75%) for controls, the much higherfigure than anticipated making it difficult toshow a survival advantage. There were only asmall number of patients in group-II, and oneeach of three treated and four control patientssurvived. Interestingly, a significant improve-ment in galactose elimination capacity wasdemonstrated during the 6 h of ELAD treatment(compared to a decrease in controls), along witha fourfold increase of growth hormone levels(which decreased in the controls). Transforminggrowth factor levels did not change.

Since then, the device has undergone mod-ifications, with plasma, rather than wholeblood, being used for perfusion of the bior-eactor hollow fibers, increase of the hepato-cyte volume to 600–800 g, and introduction ofan oxygenation system (VitaGen Inc). Theresults of a randomized controlled phase Itrial in patients with fulminant hepatic failurewere reported (still in abstract form46).Patients were stratified into those listed forLTx (n=19) and those not listed (n¼5). Ofthe 19 patients listed for LTx, 12 receivedELAD therapy; 11/12 (92%) ELAD patientswent on to receive LTx, while only 3/7 (43%)controls were transplanted (p<0.05); 10/12(83%) ELAD patients also achieved the pri-mary end point of 30-day survival, comparedto 3/7 (43%) controls (p¼0.12). The deviceappeared to be safe and, though the studywas not powered to look at outcome, showeda significant advantage for patients receiving aLTx in the ELAD group and a trend towardimproved survival for those patients listed forLTx. On this basis, a phase II randomizedcontrolled trial has been started in the UnitedStates.

Other Devices

Based on the encouraging results of carefullycontrolled studies of the AMC-BAL (Amster-dam) in large animal models, a phase I clinicaltrial with the device was carried out in Italy.47

Seven ALF patients with grade 3–4 encephalo-pathy, listed for LTx, were treated for 8–35 h.Neurological improvement was observed in allpatients. Serum bilirubin decreased by a mean of31% (range 3–62%), as did arterial ammonia(44%, range 9–66%). The only adverse effectobserved was transient hypotension in twopatients immediately after starting the device.One of the patients improved sufficiently not torequire LTx, while the remaining six weretransplanted.

Early experience with the Modular Extracor-poreal Liver Support (MELS) (Berlin) in eightpatients has been reported.48 Two patients eachwith ALF and ACLF were successfully bridgedover to LTx, and one of two others with ACLFwho were not eligible for transplant survived.Two patients with primary non-function oftheir liver grafts were bridged over to re-trans-plant. Overall treatment time varied between 7and 144 h. No adverse events were noted. Neuro-logical improvement, as well as some improve-ment of renal function and coagulation status,was observed. The Berlin group has also beenworking on a hybrid liver support system withextracorporeal plasma separation and perfusionthrough a bioreactor using 500 g of primary por-cine hepatocytes from specific pathogen-freepigs. Results in a group of seven ALF patients(with grade 2–4 encephalopathy) who were listedfor LTx have recently been published.49 Aftertreatment for 8–46 h, all seven could be success-fully bridged to LTx. Interestingly, ammoniauptake and urea synthesis by the bioreactorwere observed, along with some influence onglucose metabolism. Glucose uptake by the bior-eactor increased when the plasma glucose rose,while it was released when the plasma levelsdropped. These fascinating results have led to aphase I trial being carried out, where eightpatients with ALF were bridged to LTx usingthis system.50 The treatment was found to besafe and well-tolerated. Thrombocytopenia wasthe only significant side effect noted. Screening ofpatients’ sera for antibodies specific for porcineendogenous retroviruses (PERVs) showed noreactivity.


Some of the other bioartificial systems whichhave undergone phase I clinical trials are the‘‘radial-flow’’ bioreactor (RFB, Italy)39 device,the TECA-HALSS (China), the HBAL (China),and the BLSS (Excorp Medical, Minneapolis,MN) systems.51

Safety Concerns

Any extracorporeal blood purification systemhas potential complications associated with itsuse. These include problems related to catheter-ization and anti-coagulation, thrombocytopeniaand coagulopathies, susceptibility to infectionsand cardiac dysrhythmias, as well as hemody-namic, metabolic, and hypothermia-related con-sequences of the volume of blood circulatingthrough the extracorporeal circuit.2,19,52

In addition, there are specific safety concernsregarding the use of the hepatocyte component,namely immune reactions to foreign antigens,xenozoonosis, and escape of tumorigenic cellsinto the patient. Antibodies directed againstporcine antigens are detectable in the sera ofpatients treated with pig hepatocyte-containingBAL.53 However, high titers are not generatedfor 1 (IgM) to 3 (IgG) weeks, indicating that thiswould become a problem clinically only withrepetitive treatments. A similar response hasbeen observed following ex vivo cross-speciesliver perfusion, with, in one case, an anaphylac-tic reaction being seen after 16 such perfusionsover 2.5 months.54 Where the aim is to providelong-term liver support, a human hepatocyte-based BAL (or an artificial device) will need tobe used.

The issue regarding xenozoonosis is more com-plicated. PERV is ubiquitous among bred pigs,and transmission to humans via BAL has been apersistent fear. So strong have been the objectionsraised that trials using these systems, while per-mitted in the United States, remain on hold in theUnited Kingdom and most parts of Europe. PERVDNA and RNA have been detected in the super-natant of pig hepatocyte culture systems.55 In vitrostudies have found that these viruses can infecthuman cell lines.55–60 However, PERV transmis-sion to humans was not demonstrable in in vivostudies.61–63 A recent study also could not find anyevidence of PERV infection, using reverse tran-scriptase polymerase chain reaction, 6 monthsafter BAL treatment.43 The low membrane cut-offprobably filters off any viable virus particles which

might be present. Nyberg et al. have shown that apore size of 5 nm (corresponding to a cut-off of50–100 kDa) prevents PERV infection in humancells for 7 days59 (while treatments are meant tolast for not more than 1 day). Although the threatof PERV will probably not translate into reality,there might, quite conceivably, be other unknownviruses posing a greater threat.

The escape of tumorigenic cells from thehuman hepatoblastoma cell line is a potentialhazard in the ELAD system. However, theaddition of downstream cell filters to removeimmortalized cells from the circulating fluid isgenerally regarded as being an adequate safetymeasure against seeding.

Artificial Extracorporeal Systems

The safety concerns and the high costs asso-ciated with hepatocyte-containing devices haveled to a renewed interest in artificial liver sup-port devices. The main difference in the newersystems is an increased selectivity of the detox-ifying capacity, based on the removal of albu-min-bound toxins.

The Importance of Albumin

The mechanisms underlying the development ofthe multi-organ dysfunction of liver failure are, asyet, poorly understood. The toxin hypothesis’’implicates a variety of toxins which accumulate asa result of impaired hepatic metabolism/detoxifica-tion. Ammonia, protein breakdown products(aromatic amino acids, tryptophan, indole, mercap-tan, phenol), and endogenous benzodiazepines,among others, are implicated in the developmentof hepatic encephalopathy. Nitric oxide (NO) andprostanoids are believed to be important in thepathogenesis of circulatory and renal dysfunction.Pro-inflammatory cytokines probably have wide-ranging influences, and oxidative stress has effectsranging from increased capillary permeability tomodulating cell death.6 However, the vast majorityof these toxins (except possibly ammonia) arewater-insoluble and albumin-bound (Table 4-1),and conventional renal replacement therapy cannoteffectively remove them.

Intravenous albumin administration is ofbenefit in the treatment of patients with cirrho-sis,64–66 and improves survival in those withspontaneous bacterial peritonitis67 or with


hepatorenal syndrome,68, 69 but the benefitsexceed what could have been expected if it wasacting simply as a volume expander. Indeed, it hasbeen suggested that albumin is an importantmolecule involved in detoxification and bindsvarious substances70 and is perhaps more impor-tant in liver diseases than was previouslythought.71 This is the basis for the use of albuminas a binding and scavenging molecule in devicesbased on albumin dialysis or fractionated plasmaseparation and adsorption (FPSA), and thereforethe trial of such devices in patients with liverfailure.

Design of Devices

Molecular Adsorbents Recirculating System(MARS)

The first of the newer artificial liver supportdevices was based on the use of albumin dialysiswith a membrane having a sufficiently small poresize. This makes the system specific for albumin-bound substances, which form the majority of thetoxins accumulating in liver failure72 (Table 4-1).At the same time, larger molecules (immunoglo-bulins, growth factors) which might be physiolo-gically important are prevented from crossingover, thus eliminating one of the main problemswith the earlier devices.17,18

MARS (Gambro AB, Lund, Sweden),73,74

which has been developed over the last decade,is currently under extensive investigation andclinical trial. This uses a hollow fiber dialysismodule where the patient’s blood is dialyzedacross an albumin-impregnated polysulfonemembrane (with a cut-off of 50 kDa), whilemaintaining a constant flow of 20% albumin asdialysate in the extracapillary compartment(Figure 4-3). The premise is that toxins boundto albumin in the patient’s blood will detach andbind to the binding sites on the membrane, asalbumin, when attached to polymers, have ahigher affinity for albumin-bound toxins.75

These then pass on to the albumin in the dialysate(where albumin is present at a concentration(200 g/l) five to seven times that in the plasma).The dialysate, carrying a quantity of toxins, isthen cleansed by perfusing over activated char-coal and anion-exchange resin. These take upmost of the albumin-bound substances. Water-soluble toxins are removed by passage through ahemodialysis/hemofiltration module, which is

run in conjunction with the albumin dialysismodule. The dialysate is thus regenerated andonce more capable of taking up more toxinsfrom the blood. This recirculation component ofthe MARS system, using a fixed volume (600 ml)of albumin, is considerably more cost-effectivethan a single-pass albumin dialysis systemwould be.

Early in vitro studies showed that this methodresulted in effective removal of unconjugatedbilirubin, drugs with a high protein-bindingratio (sulfobromophthalein, theophylline), and aprotein-bound toxin (phenol).73 Further studiesdemonstrated that the system effectively removedstrongly albumin-bound toxins like unconju-gated bilirubin or free fatty acids from plasmaand blood in vivo as well.76 A subsequent study74

looked at the in vivo effect of MARS on

(a) the amino acid profile,(b) (i) physiologically important high molecu-

lar weight proteins (albumin, �-1-glycopro-tein, �-1-antitrypsine, �-2-macroglobulin,transferrin), (ii) hormone-binding proteins(thyroxine-binding globulin [TBG]), and(iii) hormone systems (thyroxine and thyr-oid-stimulating hormone [TSH]) (to evalu-ate the relative selectivity of the albumin-mediated detoxification), and

(c) albumin-bound toxic substances (bilirubin,bile acids, free fatty acids, tryptophan).

The results showed that there was an improve-ment of the amino acid profile, with relativeclearance of the aromatic amino acids (whichaccumulate in hepatic insufficiency) and animproved ratio of branched chain amino acidsto aromatic amino acids. There was no signifi-cant removal of the physiologically importantproteins or TBG, nor was there a significantchange of the thyroid hormone profile. Therewas, however, a significant removal of all thealbumin-bound toxins, which was most effectivefor fatty acids, followed by bile acids, trypto-phan, and bilirubin.

Single-Pass Albumin Dialysis (SPAD)

The newly developed SPAD system (FreseniusMedical Care AG, Bad Homburg, Germany)(Figure 4-4) dialyses blood/plasma against a4.4% solution of albumin, which is disposed ofafter a single pass. A standard renal replacementtherapy machine is used without any additional


perfusion pump system, making the equipmentrequired simpler. This fact, and the use of consid-erably more diluted albumin as the dialysate (4.4%,as opposed to 20% in case of MARS), offsets thecost of not recirculating the dialysate. Continuousveno-venous hemodiafiltration can be undertakenin conjunction as well. A point to note is that inMARS dialysis, the albumin dialysate is pre-cleansed by passing through the adsorbent col-umns for �30 min prior to starting treatment,potentially freeing up more binding sites on thealbumin molecule. This does not occur in SPAD,which, at least theoretically, is a disadvantage.

Prometheus

The fractionated plasma separation and adsorp-tion (FPSA)77 system, introduced in 1999,removes albumin-bound toxins by fractionationof the plasma with the subsequent detoxification

of the native albumin by adsorption. It uses analbumin-permeable membrane with a cut-off of250 kDa. Albumin and possibly other plasmaproteins with their bound toxins cross the mem-brane and pass through special adsorbers (oneor two columns in series in the secondary cir-cuit, containing a neutral resin adsorber and ananion exchanger) that remove the toxins. Thecleansed albumin is returned to the plasma. Inthe commercially produced Prometheus system(Fresenius Medical Care AG, Bad Homburg,Germany)78 (Figure 4-5) the FPSA method iscombined with high-flux hemodialysis (of theblood directly, as opposed to the MARS system,where hemodialysis/filtration of the albumindialysate is performed).

It is worthwhile noting some salient differencesbetween FPSA and albumin dialysis. No albuminsolution is required as a dialysate in FPSA. Thiswould probably result in lower running costs. Thepatient’s own albumin is separated from the

Figure 4-3. The MARS circuit. Inset: The structure of the MARS membrane. Toxins pass from plasma albumin onto the binding site on themembrane, and then to the albumin in the dialysate.


plasma and then detoxified in contact with adsor-bent columns. This means that FPSA uses a farless selective membrane than albumin dialysis(MARS or SPAD). Loss of the patient’s albumin

into the circuit is a more real risk in the formercompared to the latter. This also means that usingpotent though unselective adsorbers such as acti-vated charcoal in FPSA (in contact with the

Figure 4-5. Schematic of the Pro-metheus circuit, utilizing the principleof fractionated plasma separation andadsorption (FPSA). The plasma is frac-tionated across an albumin-permeablemembrane, with the subsequentdetoxification of the native albuminby adsorption of bound toxins acrossone or two adsorber columns (contain-ing a neutral resin adsorber and ananion exchanger) in the secondary cir-cuit. The cleansed albumin is returnedto the plasma. The reconstituted bloodsubsequently undergoes high-fluxhemodialysis before being returned tothe patient.

Figure 4-4. Schematic of the single-pass albumin dialysis (SPAD) system.Plasma from a reservoir is dialyzedagainst a 4.4% solution of albumin,which is disposed of after a singlepass. A standard renal replacementtherapy machine is used to run thecircuit.


patient’s own plasma fraction, and therefore risk-ing the possible removal of biologically usefulmolecules) would be far more hazardous thantheir use in albumin dialysis (where contactwould be only with the albumin solution of thedialysate). This probably explains why charcoal,used in the earlier version of FPSA, was replacedby a neutral resin adsorber, while it continues tobe used safely in MARS. Finally, while heparincontinues to be used as the anti-coagulant ofchoice in MARS and SPAD, it seems to haveresulted in clotting problems with FPSA,78 wherecitrate is being looked at as an alternative.79

Clinical Trials

MARS

Acute-on-chronic Liver Failure (ACLF). In theinitial clinical trials MARS has predominantlybeen evaluated in the context of ACLF, eventhough most studies were small and uncontrolled.Improvement of hyperbilirubinemia, HE, circula-tory, and renal functions have been observed.80–82

The improvement of HE has been associated witha reduction of serum ammonia levels, decrease ofintracranial pressures, and increase of cerebralperfusion pressures.80–83 Circulatory changesfollowing MARS treatment have manifested inthe form of increased mean arterial pressure andsystemic vascular resistance and decreased cardiacoutput.80,81,84,85 A recent study demonstrated clini-cally significant, acute reduction of portal pressurein severe alcoholic hepatitis.86

The largest series of patients with ACLF(n=26), with intrahepatic cholestasis (bilirubinlevel > 20 mg/dL), treated with MARS has beenreported from Rostock.87 The series included 10patients with a United Network Organ Sharing(UNOS) status 2b, all of whom survived, and 16patients with a UNOS status 2a, of whom 7survived. Another study on patients with severeacute alcoholic hepatitis treated with MARS(n=8)88 showed improvement of 3-month pre-dicted mortality (pre-MARS: 76%, post-MARS:27%), with 50% of patients (4/8) still surviving at3 months.

The first randomized trial of MARS evaluated13 ACLF patients with type-I hepatorenal syn-drome who were treated with either MARS (n=8)or standard medical therapy including hemodia-filtration (n=5).85 The mortality rate was 100% inthe group receiving hemodiafiltration at day 7

compared with 62.5% in the MARS group at day7 and 75% at day 30, respectively (p< 0.01). Meansurvival was longer in the MARS group, which wasaccompanied by a significant decrease in serumbilirubin and creatinine, and increase in serumsodium and prothrombin activity. MAP at theend of treatment was significantly greater in theMARS group. Although urine output did notincrease significantly in the MARS group, 4 ofthe 8 patients showed an increase compared withnone of the control group.

A subsequent randomized controlled trial,performed in two centers (Rostock and Essen),included 24 patients with ACLF with markedhyperbilirubinemia (serum bilirubin>20 mg/dl[340 mmol/L]) who were randomized to receivestandard medical therapy alone (n=12) or MARSin addition (n=12).89 The primary end point ofbilirubin <15 mg/dL for three consecutive dayswas reached in 5 of 12 MARS patients and in 2 of12 control patients. Compared to controls, bilir-ubin, bile acids, and creatinine decreased andMAP and HE improved in the MARS group.Most importantly, albumin dialysis was asso-ciated with a significant improvement in 30-daysurvival (11/12 vs 6/11 in controls).

An FDA-controlled multi-center randomizedtrial (not yet published) on the role of MARS inadvanced hepatic encephalopathy (grade III/IV)has recently been completed in 70 cirrhoticpatients in the United States and Europe(reported only in abstract form).90 MARS treat-ment (daily for 6 h over five consecutive days)significantly improved encephalopathy com-pared to standard medical therapy. The numberof patients dying during the 180 days of follow-up and the number having LTx, however, werenot significantly different in the two groups.Resolving coma within 5 days was shown tohave a strong impact on the 2-week transplant-free survival: 81% compared with 34% in thoseremaining in coma (p<0.001).91 A Europeanmulti-center randomized controlled trial evalu-ating the role of MARS in ACLF with mortality asthe primary end point is awaiting completion.

Acute Liver Failure. In the context of ALF, nocontrolled studies have been performed as yet.Outcome data are available from three centers –Rome, Helsinki, and Gothenburg. Novelli et al.92

from Rome have treated nine cases of fulminanthepatic failure. Three patients survived withoutrequiring transplantation. The remaining sixwere transplanted, of whom four survived,


while two died due to sepsis. The authors haveextended the series to 24, in whom they reportimprovement of serum bilirubin, INR, andammonia as well as neurological status (thoughoutcome is not described).93,94 Isoniemi95 (Hel-sinki) has reported 45 cases of ALF managedwith MARS; 80% of the patients survived,which is a strikingly high proportion. Nativeliver recovered in 23 cases, while 13 patientswere transplanted successfully. Hemodynamicand neurological improvements were noted fol-lowing MARS therapy in most cases. Whilethese results are quite encouraging, especiallyin those with a toxic etiology, it is to be notedthat at least in some of these cases MARS wasstarted before the development of encephalo-pathy, and in the absence of matched controlssuch improvement is difficult to interpret. Fell-din et al. (Gothenburg) describe 10 patients ofALF treated with MARS, of whom 7 survived.Beneficial effect was most evident in those whoreceived 5 or more sessions of treatment (4/5survivors).96,97

Among other studies, one has describedimprovement in encephalopathy, a reduction ofcerebral edema and intracranial pressure, withincrease of cerebral perfusion pressure in threeALF patients treated with MARS.98 A recent smallrandomized controlled study in patients withhyperacute liver failure found that a single ses-sion of MARS treatment (n=8) improved sys-temic hemodynamics (mean arterial pressure,systemic vascular resistance, and cardiac output)compared to controls (n=5), who had only beenmechanically cooled to match the MARS group.99

A phase I trial from the United States evaluated aslightly modified system using continuous albu-min dialysis with continuous hemodiafiltrationin nine ALF patients (UNOS status 1: 5; status2a: 4).100 Of those with status 1, one recoverednative hepatic function while three were bridgedto transplantation.

Two recent studies have reported the use ofMARS in ALF related to hepatitis B. A Chineseseries describes ten such patients, of whom onewas successfully bridged to LTx and three werealive at 3-month follow-up.101 Hemodynamic,neurological, and biochemical improvementswere noted. An Italian series describes MARStreatment in five patients of hepatitis B-relatedALF in lymphoma patients undergoing che-motherapy.102 Three were alive at 1-year fol-low-up, while MARS was thought to have beencommenced too late in the two who died.

A recent report describes the interim ana-lysis on the first 27 patients included in arandomized controlled single-center trialevaluating MARS in hypoxic liver failure dueto cardiogenic shock following cardiac sur-gery.103 Seven out of 14 patients (50%) in theMARS group survived, compared to 4 out of 13(32%) in the non-MARS group. No significantdifference in survival was found. A full reportis awaited.

There are also individual case reports of theuse of MARS (n=4)104–106 or albumin dialysis(n=1)107 to treat Wilson’s disease with ALF(grade 3–4 encephalopathy, oligo-anuria, coagu-lopathy), all of whom were successfully bridgedto transplantation after a period ranging from 9to 59 days. MARS has also been used to treat ALFdue to paracetamol overdose (n=3)104,108,109 andmushroom intoxication (n=3),104,110,111 where allsurvived without requiring transplantation, whilea case of acute Budd–Chiari syndrome died inspite of MARS and transplantation.104 A series offive MARS-treated paracetamol overdosepatients report survival in four, with one patientdying of brain edema.112

Other Indications. Results with MARS in thetreatment of primary graft dysfunction follow-ing liver transplantation are limited. Onestudy113 reported six such patients (primarynon-function: n=4, delayed non-function:n=2), where MARS treatment led to a recoveryin five cases, eliminating the need for re-trans-plantation. The sixth patient died. Anotherstudy104 reported two patients of liver failurefollowing transplantation, one due to primarynon-function and the other due to severe graftdysfunction, both of whom were bridged tore-transplantation with MARS. The formerdied following sepsis, while the latter survived.Data available from four other centers describeeight cases of primary graft dysfunctiontreated with MARS, of whom three recoveredand four were successfully bridged to re-transplantation.114

There have been reports of the use of MARSfor the treatment of liver failure following hepa-tic resection (n=7),104,115–117 but only one suchpatient survived.104 There have also been anec-dotal reports of the use of MARS in progressiveintrahepatic cholestasis due to chronic graft-versus-host disease (n=1),118 and in patientswith heart failure, complicated by liver failure,awaiting heart transplantation.119


Newer Indications. Over the last 3 years, twonew clinical indications for the use of MARStherapy appear to be emerging. Intrahepaticcholestasis with intractable pruritus can be anextremely debilitating condition, severelyimpairing the patient’s quality of life. Therapeu-tic options are limited for cases which fail torespond to currently available medications. Inthis setting, MARS has given a rapid, significant,and reasonably well-sustained response.120–122

Mullhaupt et al. described rapid resolution ofpruritus following a single session of MARS ina patient with primary biliary cirrhosis, but withrecurrence of symptoms within 5 days.121 Maciaet al. described a similar improvement withMARS in two patients with primary biliary cir-rhosis and one patient with post-transplantationbiliary stenosis, with the effect persisting for 3–8weeks.123 In one of these patients, out-patientfollow-up with a MARS session every fewmonths (after the initial therapy) was successfulin achieving long-term improvement. Doria etal. reported three patients with hepatitis C-virus-related intractable pruritus, treated withseven sessions of MARS, where an improvementpersisting for several months was noted.122 Bell-mann et al. recently described seven patientswith post-transplantation intractable prurituswho were treated with three consecutive ses-sions of MARS.124 Six patients responded, andthe improvement was sustained for >3 monthsin three of them. The last two studies alsodemonstrated an associated reduction of serumbile acids, implicating this as a potentialmechanism of the response. However, removalof plasma opioids such as met-enkephalins125

may be responsible as well.The other emerging indication for MARS

therapy is in the setting of overdose/toxicitywith protein-bound, especially albumin-bound,substances which are often non-dialyzable. Wehave reported the rapid clinical improvementfollowing a single MARS session of a patientwith phenytoin toxicity, associated with a linearrapid reduction of both total and free plasmaphenytoin levels.126 This observation was takenfurther in a subsequent study in an animalmodel of ALF, where we determined whetherthis detoxifying capacity was restricted to onlyalbumin-bound substances or could be general-ized to substances bound to other proteins aswell.127 The results demonstrated that fentanyl(85% bound to �-1-acid glycoprotein in theplasma) was removed as efficiently by MARS as

midazolam (97% bound to albumin). These dataprovide the rationale to explore the role ofMARS in the treatment of toxicities with a vari-ety of protein-bound non-dialyzable substances.

Pathophysiological Basis. The pathophysiologi-cal basis of the development of the clinicalmanifestations of ACLF is only poorly under-stood. Similarly, very little is known about themechanisms underlying the observed clinicalimprovements following MARS therapy. Theimprovement in HE may be due to the associatedsignificant reduction of serum ammonialevels,80–82 as a result of direct removal by theMARS system.128 Another possible explanationmight be a favorable alteration of the aminoacids profile, with relative clearance of aromaticamino acids, as had been observed in the early invivo studies.74 A reduction of plasma nitrates andnitrites following MARS, probably due to directremoval by the system,128 may contribute to theimprovement of the circulatory status. Whetherproduction of nitric oxide is altered as well fol-lowing therapy is as yet not known. This systemichemodynamic improvement may, at least in part,be responsible for the associated improvement ofrenal function. An alteration in the profile ofinflammatory mediators and cytokines, eitherby removal by the MARS circuit or by alteredproduction due to an ‘‘improved’’ internal envir-onment following blood purification by MARS,might conceivably have a role to play. Prelimin-ary data from one study failed to show anychange of plasma levels of tumor necrosis factor(TNF)-� or its receptors following MARS treat-ment, in spite of their detection in the dialysate infairly considerable amounts, indicating removalof these substances across the membrane.128

SPAD

In vitro studies suggest that its detoxifying capa-city is similar to or even greater than (especiallywith regard to bilirubin and ammonia clearance)that of MARS.129 However, would these resultshold up in vivo too? As of now, there are only afew reports of its clinical use. One was a case offulminant Wilson’s disease, where it was foundto efficiently clear bilirubin and copper, bothprotein-bound, from the plasma.107 Subse-quently, three cirrhotic patients were treatedlong term, resulting in two successful transplan-tations and one patient dying from sepsis after


140 days.130 There is another case report of asuccessful short-term bridging to LTx fromBerlin.131

Prometheus

Since Prometheus was first used successfully in ayoung patient with ALF and multi-organ failuredue to ecstasy and cocaine intoxication (whowas deemed ineligible for LTx),132 more clinicaldata have emerged. The results of Prometheustreatment in 11 patients with ACLF and accom-panying renal failure have been published.78

Improvement of serum levels of conjugatedbilirubin, bile acids, ammonia, cholinesterase,creatinine, urea, and blood pH occurred. Adrop in blood pressure in two patients anduncontrolled bleeding in one patient were theadverse events noted. Another randomizedcross-over study compared alternating treat-ments with MARS and Prometheus in eightpatients with ACLF. Reduction ratios of bothbilirubin and urea were more with Prometheus.Their safety profiles were found to be compar-able.133 In another study, the removal capacityand selectivity of the Prometheus system wasevaluated in a clinical trial on nine patients ofACLF.134 Levels of endogenous toxins (ureanitrogen, creatinine, total bilirubin, and bileacids) were found to decrease significantly. Asubsequent report by the same group reportedtoxin removal in ACLF by Prometheus (n=9) tobe more effective compared to that by MARS(n=9).135 Prospective controlled trials areplanned for the future.

Safety Profile

The safety of the MARS device has beenevaluated through its use in over 3000patients worldwide. The MARS Registry,which is maintained by the University ofRostock, contains data on about 500 patientstreated with this device.136,137 In general, thetreatment is well tolerated and the only con-sistent adverse finding with the use of MARSis thrombocytopenia. Critical analysis of thedata from the Registry in patients with ACLFsuggests that its use should be contraindi-cated in those with established disseminatedintravascular coagulation (DIC) or in thosepatients with ‘‘incipient’’ DIC characterized

by progressive thrombocytopenia (< 50 �109/L) and/or coagulopathy (INR > 2.3).Another factor found in the interim analysisof the first multi-center European clinicaltrial in ACLF patients is the adverse influenceon outcome of already established renalsupport.

The initial pilot trial looking at the safety ofthe Prometheus system found it to be generallysafe and well tolerated.78 A drop in blood pres-sure in two patients and uncontrolled bleedingin one patient were the adverse events noted. Ingeneral, a significant drop in mean arterialpressure was noted during Prometheus treat-ment, which returned to the pre-treatmentlevels by the following day. This might berelated to the larger extracorporeal blood andplasma volume for Prometheus (because of theadditional plasma volume in the secondary cir-cuit) compared to either MARS or SPAD. Areversible increase of white cells (though noother signs of systemic inflammation), possibledue to transient leukocyte activation by themembrane, was noted too. However, no caseof thrombocytopenia (as has sometimes beenseen with MARS therapy) was observed.Finally, heparin, which remains the anti-coagu-lant of choice in MARS and SPAD, seems tohave resulted in clotting problems with Pro-metheus,78 where citrate is being looked at asan alternative.79

Conclusion

Most of the benefits observed with liver supportin the setting of ALF have been in the form ofsuccessful bridging to transplantation, ratherthan recovery of the native liver. The latter hasbeen relatively more often achieved only in stu-dies on patients with ACLF. While most studieswith hybrid bioartificial systems have demon-strated their immediate safety, concerns persist,particularly regarding the acquisition of retro-viral infections when porcine hepatocyte mod-ules are used. The complexity of the circuits andthe high costs involved inevitably will be a limit-ing factor to their use. Artificial systems are sim-pler, safer, and cheaper. The data supportingtheir use in ACLF are encouraging. However,only when a carefully conducted controlled clin-ical trial has shown proven benefit, includingsurvival, can we be certain of its value, whateverthe efficacy of toxin removal.


The difficulties in showing such a benefit witha controlled trial in ALF have already beenreferred to. LTx has a major influence on survivalin the most severe group of cases, and the timingof it is inevitably going to differ from case to caseand from center to center depending on the avail-ability of organs. One group of patients that couldbe utilized for future trials in ALF would bepatients who are early in the course of their dis-ease, in whom subsequent improvement or dete-rioration can be correlated with the requirementfor LTx as the primary end point. The maindifficulty with such a trial is in setting the selec-tion criteria for inclusion so that mortality is stillsufficiently high in the control group for a benefitto be shown. Another category of ALF patientswho could be studied are those who fulfill criteriafor LTx, but who because of co-morbidity orother reasons are considered unsuitable for it.ALF from paracetamol hepatotoxicity has a dif-ferent clinical course with lower requirement forLTx and higher incidence of spontaneous recov-ery than ALF due to other etiologies, and prob-ably merits a separate and specifically designedtrial, taking into account the percentage of casesthat do go downhill rapidly. Whatever the inclu-sion criteria and design of the trial in ALF, thepatients must be stratified for etiology, age, andrapidity of onset of encephalopathy.

As regards the setting of ACLF, well-designedcontrolled trials (especially with MARS) withsurvival as the primary end point have alreadybeen set up. The main point to note is that quitelarge numbers of patients (about 150–200)would need to be included for the trials to beadequately powered. This implies that theselogistically complex studies need to be carriedout in multiple centers while at the same timemaintaining the basic homogeneity of the trial –which in practice can be more daunting thanthey would appear in theory.

As for the devices themselves, what of thefuture? While the search for an effective liversupport system remains the Holy Grail forresearchers in this area, the value of the presentsystems, whether hybrid bioartificial or entirelyartificial, is still far from certain. What stillremains a puzzle with the bioartificial devicesis the lack of major discernible synthetic effectfrom the biological component. Further work onwhat underlies the improvement in conscious-ness levels with BAL, as well as of the variousbeneficial effects reported with MARS, is clearlyessential. The shortage of human hepatocytes

for use in BAL could be solved by the use ofstem cells, especially hemopoietic stem cells,induced to differentiate into mature functionalhepatocytes.138 Reversible immortalization oftumorigenic human cell lines using the Cre-Loxgenetic engineering system, with excision of theimmortalizing gene, thus removing the threat oftumorigenesis,139 is another approach which hasalready been tested experimentally. The viabilityand functional capacity of hepatocytes138 hasbeen shown to be improved by co-culture withnon-hepatocytic/endothelial cells, and their pre-sence would also help through the scavengingfunctions of the endothelial cells.140 As to theimprovement in the design of bioreactors, betterprovision for the functions of bile excretion couldbe of benefit. Interestingly, it has recently beenreported that bile duct-like structures form incultures of bile epithelial cells.141,142 But all thesepossible developments will add even further tothe complexity (and costs) of the hybrid bioarti-ficial devices. As to improvements for the artifi-cial devices like MARS, additional adsorptioncolumns for removal of specific toxins, like cop-per for acute Wilson’s disease, and modificationto the membrane and its albumin impregnationso as to decrease quantities of albumin requiredand increase transport efficiency would be worth-while. SPAD and FPSA are variations of thesedevices which are already being tested. Which ofthese systems will ultimately stand the test oftime? Only the future will tell.

References

1. Sen S, Williams R, Jalan R. The pathophysiological basisof acute-on-chronic liver failure. Liver. 2002;22(Suppl2):5–13.

2. Stockmann HB, IJzermans JN. Prospects for the tempor-ary treatment of acute liver failure. Eur J GastroenterolHepatol. 2002;14:195–203.

3. www.optn.org.4. Miwa S, Hashikura Y, Mita A, Kubota T, Chisuwa H,

Nakazawa Y, Ikegami T, Terada M, Miyagawa S, KawasakiS. Living-related liver transplantation for patients withfulminant and subfulminant hepatic failure. Hepatology.1999;30:1521–6.

5. Baker A, Dhawan A, Devlin J, Mieli-Vergani G, O’ Grady J,Williams R, Rela M, Heaton N. Assessment of potentialdonors for living related liver transplantation. Br J Surg.1999;86:200–5.

6. Sen S, Williams R, Jalan R. The pathophysiological basisof acute-on-chronic liver failure. Liver. 2002;22(Suppl2):5–13.

7. Williams R. The elusive goal of liver support–quest for theHoly Grail. Clin Med. 2006;6:482–7.


8. Abouna GM. Cross-circulation between man andbaboon in hepatic coma. Lancet. 1968;2:729–30.

9. Motin J, Bouletreau P, Petit P, Latarjet J. [Hepatic assis-tance by means of inter-human cross circulation. Apro-pos of 6 cases]. Minerva Chir. 1974;29:989–93.

10. Abouna GM, Garry R, Hull C, Kirkley J, Walder DN. Pig-liver perfusion in hepatic coma. Lancet. 1968;2:509–10.

11. Ranek L, Hansen RI, Hilden M, Ramsoe K, Schmidt A,Winkler K, Tygstrup N. Pig liver perfusion in the treat-ment of acute hepatic failure. Scand J Gastroenterol.1971;9:161–9.

12. Graw RG, Jr., Buckner CD, Eisel R. Plasma exchangetransfusion for hepatic coma. New technic Transfusion.1970;10:26–32.

13. Larsen FS, Hansen BA, Jorgensen LG, Secher NH,Kirkegaard P, Tygstrup N. High-volume plasmapheresisand acute liver transplantation in fulminant hepatic fail-ure. Transplant Proc. 1994;26:1788.

14. Clemmesen JO, Kondrup J, Nielsen LB, Larsen FS, Ott P.Effects of high-volume plasmapheresis on ammonia,urea, and amino acids in patients with acute liver failure.Am J Gastroenterol. 2001;96:1217–23.

15. Clemmesen JO, Larsen FS, Ejlersen E, Schiodt FV, OttP, Hansen BA. Haemodynamic changes after high-volume plasmapheresis in patients with chronic andacute liver failure. Eur J Gastroenterol Hepatol.1997;9:55–60.

16. Larsen FS, Hansen BA, Ejlersen E, Secher NH,Clemmesen JO, Tygstrup N, Knudsen GM. Cerebralblood flow, oxygen metabolism and transcranial Dop-pler sonography during high-volume plasmapheresis infulminant hepatic failure. Eur J Gastroenterol Hepatol.1996;8:261–5.

17. O’Grady JG, Gimson AE, O’Brien CJ, Pucknell A, HughesRD, Williams R. Controlled trials of charcoal hemoper-fusion and prognostic factors in fulminant hepatic fail-ure. Gastroenterology. 1988;94:1186–92.

18. McGuire BM, Sielaff TD, Nyberg SL, Hu MY, Cerra FB,Bloomer JR. Review of support systems used in themanagement of fulminant hepatic failure. Dig Dis.1995;13:379–88.

19. Allen JW, Hassanein T, Bhatia SN. Advances in bioarti-ficial liver devices. Hepatology. 2001;34:447–55.

20. Jauregui HO, Mullon CJ, Trenkler D, Naik S, SantanginiH, Press P, Muller TE, Solomon BA. In vivo evaluation ofa hollow fiber liver assist device. Hepatology. 1995;21:460–9.

21. Nyberg SL, Peshwa MV, Payne WD, Hu WS, Cerra FB.Evolution of the bioartificial liver: the need for rando-mized clinical trials. Am J Surg. 1993;166:512–21.

22. Sussman NL, Chong MG, Koussayer T, He DE, Shang TA,Whisennand HH, Kelly JH. Reversal of fulminant hepaticfailure using an extracorporeal liver assist device. Hepa-tology. 1992;16:60–5.

23. Rozga J, Williams F, Ro MS, Neuzil DF, Giorgio TD,Backfisch G, Moscioni AD, Hakim R, Demetriou AA.Development of a bioartificial liver: properties and func-tion of a hollow-fiber module inoculated with liver cells.Hepatology. 1993;17:258–65.

24. Flendrig LM, Calise F, Di Florio E, Mancini A, CerielloA, Santaniello W, Mezza E, Sicoli F, Belleza G, Bracco A,Cozzolino S, Scala D, Mazzone M, Fattore M, GonzalesE, Chamuleau RA. Significantly improved survival timein pigs with complete liver ischemia treated with a novelbioartificial liver. Int J Artif Organs. 1999;22:701–9.

25. Sosef MN, Abrahamse LS, van de Kerkhove MP,Hartman R, Chamuleau RA, van Gulik TM. Assessmentof the AMC-bioartificial liver in the anhepatic pig. Trans-plantation. 2002;73:204–9.

26. Arai K, Lee K, Berthiaume F, Tompkins RG, YarmushML. Intrahepatic amino acid and glucose metabolism ina D-galactosamine-induced rat liver failure model.Hepatology. 2001;34:360–71.

27. Hughes RD, Cochrane AM, Thomson AD, Murray-LyonIM, Williams R. The cytotoxicity of plasma from patientswith acute hepatic failure to isolated rabbit hepatocytes.Br J Exp Pathol. 1976;57:348–53.

28. Suh KS, Lilja H, Kamohara Y, Eguchi S, Arkadopoulos N,Neuman T, Demetriou AA, Rozga J. Bioartificial liver treat-ment in rats with fulminant hepatic failure: effect on DNA-binding activity of liver-enriched and growth-associatedtranscription factors. J Surg Res. 1999;85:243–50.

29. Strain AJ, Neuberger JM. A bioartificial liver–state of theart. Science. 2002;295:1005–9.

30. Nagasue N, Yukaya H, Ogawa Y, Kohno H, Nakamura T.Human liver regeneration after major hepatic resection.A study of normal liver and livers with chronic hepatitisand cirrhosis. Ann Surg. 1987;206:30–9.

31. Arkadopoulos N, Lilja H, Suh KS, Demetriou AA,Rozga J. Intrasplenic transplantation of allogeneic hepa-tocytes prolongs survival in anhepatic rats. Hepatology.1998;328:1365–70.

32. Wu L, Sun J, Wang L, Wang C, Woodman K, KoutalistrasN, Horvat M, Sheil AG. Cryopreservation of primaryporcine hepatocytes for use in bioartificial liver supportsystems. Transplant Proc. 2000;32:2271–2.

33. Yagi T, Hardin JA, Valenzuela YM, Miyoshi H, Gores GJ,Nyberg SL. Caspase inhibition reduces apoptotic deathof cryopreserved porcine hepatocytes. Hepatology.2001;33:1432–40.

34. Kono Y, Yang S, Letarte M, Roberts EA. Establishment ofa human hepatocyte line derived from primary culture ina collagen gel sandwich culture system. Exp Cell Res.1995;221:478–85.

35. McCloskey P, Edwards RJ, Tootle R, Selden C, RobertsE, Hodgson HJ. Resistance of three immortalizedhuman hepatocyte cell lines to acetaminophen and N-acetyl-p-benzoquinoneimine toxicity. J Hepatol.1999;31:841–51.

36. McCloskey P, Tootle R, Selden C, Larsen F, Roberts E,Hodgson HJ. Modulation of hepatocyte function inan immortalized human hepatocyte cell line followingexposure to liver-failure plasma. Artif Organs. 2002;26:340–8.

37. Gerlach JC, Schnoy N, Encke J, Smith MD, Muller C,Neuhaus P. Improved hepatocyte in vitro maintenancein a culture model with woven multicompartment capil-lary systems: electron microscopy studies. Hepatology.1995;22:546–52.

38. Flendrig LM, la Soe JW, Jorning GG, Steenbeek A,Karlsen OT, Bovee WM, Ladiges NC, te Velde AA, Cha-muleau RA. In vitro evaluation of a novel bioreactorbased on an integral oxygenator and a spirally woundnonwoven polyester matrix for hepatocyte culture assmall aggregates. J Hepatol. 1997;26:1379–92.

39. Morsiani E, Pazzi P, Puviani AC, Brogli M, Valieri L,Gorini P, Scoletta P, Marangoni E, Ragazzi R, AzzenaG, Frazzoli E, Di Luca D, Cassai E, Lombardi G, CavallariA, Faenza S, Pasetto A, Girardis M, Jovine E, Pinna AD.Early experiences with a porcine hepatocyte-based


bioartificial liver in acute hepatic failure patients. Int JArtif Organs. 2002;25:192–202.

40. Matsumura KN, Guevara GR, Huston H, Hamilton WL,Rikimaru M, Yamasaki G, Matsumura MS. Hybrid bioar-tificial liver in hepatic failure: preliminary clinicalreport. Surgery. 1987;101:99–103.

41. Chen SC, Hewitt WR, Watanabe FD, Eguchi S, Kahaku E,Middleton Y, Rozga J, Demetriou AA. Clinical experi-ence with a porcine hepatocyte-based liver support sys-tem. Int J Artif Organs. 1996;19:664–9.

42. Watanabe FD, Mullon CJ, Hewitt WR, Arkadopoulos N,Kahaku E, Eguchi S, Khalili T, Arnaout W, ShackletonCR, Rozga J, Solomon B, Demetriou AA. Clinical experi-ence with a bioartificial liver in the treatment of severeliver failure. A phase I clinical trial. Ann Surg.1997;225:484–91; discussion 491–4.

43. Samuel D, Ichai P, Feray C, Saliba F, Azoulay D,Arulnaden JL, Debat P, Gigou M, Adam R, Bismuth A,Castaing D, Bismuth H. Neurological improvement dur-ing bioartificial liver sessions in patients with acute liverfailure awaiting transplantation. Transplantation.2002;73: 257–64.

44. Demetriou AA, Brown RS, Jr., Busuttil RW, Fair J,McGuire BM, Rosenthal P, Am Esch JS, 2nd, Lerut J,Nyberg SL, Salizzoni M, Fagan EA, de Hemptinne B,Broelsch CE, Muraca M, Salmeron JM, Rabkin JM, Met-selaar HJ, Pratt D, De La Mata M, McChesney LP,Everson GT, Lavin PT, Stevens AC, Pitkin Z, SolomonBA. Prospective, randomized, multicenter, controlledtrial of a bioartificial liver in treating acute liver failure.Ann Surg. 2004;239:660–7; discussion 667–70.

45. Ellis AJ, Hughes RD, Wendon JA, Dunne J, LangleyPG, Kelly JH, Gislason GT, Sussman NL, Williams R.Pilot-controlled trial of the extracorporeal liver assistdevice in acute liver failure. Hepatology.1996;24:1446–51.

46. Millis JM, Kramer DJ, O’Grady J, Heffron T, Caldwell S,Hart ME, Maguire P. Results of phase I trial of the extra-corporeal liver assist device for patients with fulminanthepatic failure. Am J Transplantation. 2001;1(suppl1):391.

47. van de Kerkhove MP, Di Florio E, Scuderi V, Mancini A,Belli A, Bracco A, Dauri M, Tisone G, Di Nicuolo G,Amoroso P, Spadari A, Lombardi G, Hoekstra R, CaliseF, Chamuleau RA. Phase I clinical trial with the AMC-bioartificial liver. Academic Medical Center. Int J ArtifOrgans. 2002;25:950–9.

48. Sauer IM, Zeilinger K, Obermayer N, Pless G, GrunwaldA, Pascher A, Mieder T, Roth S, Goetz M, Kardassis D,Mas A, Neuhaus P, Gerlach JC. Primary human livercells as source for modular extracorporeal liver sup-port–a preliminary report. Int J Artif Organs.2002;25:1001–5.

49. Mundt A, Puhl G, Muller A, Sauer I, Muller C, Richard R,Fotopoulou C, Doll R, Gabelein G, Hohn W, Hofbauer R,Neuhaus P, Gerlach J. A method to assess biochemicalactivity of liver cells during clinical application of extra-corporeal hybrid liver support. Int J Artif Organs.2002;25:542–8.

50. Sauer IM, Kardassis D, Zeillinger K, Pascher A,Gruenwald A, Pless G, Irgang M, Kraemer M, Puhl G,Frank J, Muller AR, Steinmuller T, Denner J, Neuhaus P,Gerlach JC. Clinical extracorporeal hybrid liver sup-port–phase I study with primary porcine liver cells.Xenotransplantation. 2003;10:460–9.

51. Chamuleau RA, Poyck PP, van de Kerkhove MP. Bioar-tificial liver: its pros and cons. Ther Apher Dial.2006;10:168–74.

52. Jalan R, Williams R. Bio-artificial liver support for acuteliver failure: should we be using it to treat patients?Transplantation. 2002;73:165–6.

53. Baquerizo A, Mhoyan A, Shirwan H, Swensson J, BusuttilRW, Demetriou AA, Cramer DV. Xenoantibodyresponse of patients with severe acute liver failureexposed to porcine antigens following treatment with abioartificial liver. Transplant Proc. 1997;29:964–5.

54. Abouna GM, Boehmig HG, Serrou B, Amemiya H,Martineau G. Long-term hepatic support by intermittentmulti-species liver perfusions. Lancet. 1970;2:391–6.

55. Nyberg SL, Hibbs JR, Hardin JA, Germer JJ, Platt JL, PayaCV, Wiesner RH. Influence of human fulminant hepaticfailure sera on endogenous retroviral expression in pighepatocytes. Liver Transpl. 2000;6:76–84.

56. Patience C, Takeuchi Y, Weiss RA. Infection of humancells by an endogenous retrovirus of pigs. Nat Med.1997;3:282–6.

57. Martin U, Kiessig V, Blusch JH, Haverich A, von derHelm K, Herden T, Steinhoff G. Expression of pig endo-genous retrovirus by primary porcine endothelial cellsand infection of human cells. Lancet. 1998;352:692–4.

58. Wilson CA, Wong S, Muller J, Davidson CE, Rose TM,Burd P. Type C retrovirus released from porcine primaryperipheral blood mononuclear cells infects human cells.J Virol. 1998;72:3082–7.

59. Nyberg SL, Hibbs JR, Hardin JA, Germer JJ, Persing DH.Transfer of porcine endogenous retrovirus across hollowfiber membranes: significance to a bioartificial liver.Transplantation. 1999;67:1251–5.

60. Martin U, Winkler ME, Id M, Radeke H, Arseniev L,Takeuchi Y, Simon AR, Patience C, Haverich A, SteinhoffG. Productive infection of primary human endothelialcells by pig endogenous retrovirus (PERV). Xenotrans-plantation. 2000;7:138–42.

61. Heneine W, Tibell A, Switzer WM, Sandstrom P, RosalesGV, Mathews A, Korsgren O, Chapman LE, Folks TM,Groth CG. No evidence of infection with porcine endo-genous retrovirus in recipients of porcine islet-cell xeno-grafts. Lancet. 1998;352:695–9.

62. Paradis K, Langford G, Long Z, Heneine W, Sandstrom P,Switzer WM, Chapman LE, Lockey C, Onions D, Otto E.Search for cross-species transmission of porcine endo-genous retrovirus in patients treated with living pig tissue.The XEN 111 Study Group. Science. 1999;285:1236–41.

63. Patience C, Patton GS, Takeuchi Y, Weiss RA, McClure MO,Rydberg L, Breimer ME. No evidence of pig DNA or retro-viral infection in patients with short-term extracorporealconnection to pig kidneys. Lancet. 1998;352:699–701.

64. Arroyo V. Review article: albumin in the treatment ofliver diseases-new features of a classical treatment. Ali-ment Pharmacol Ther. 2002;16(Suppl 5):1–5.

65. Ruiz-del-Arbol L, Monescillo A, Jimenez W, Garcia-PlazaA, Arroyo V, Rodes J. Paracentesis-induced circulatorydysfunction: mechanism and effect on hepatic hemody-namics in cirrhosis. Gastroenterology. 1997;113:579–86.

66. Gines A, Fernandez-Esparrach G, Monescillo A, Vila C,Domenech E, Abecasis R, Angeli P, Ruiz-Del-Arbol L,Planas R, Sola R, Gines P, Terg R, Inglada L, Vaque P,Salerno F, Vargas V, Clemente G, Quer JC, Jimenez W,Arroyo V, Rodes J. Randomized trial comparing albu-min, dextran 70, and polygeline in cirrhotic patients with


ascites treated by paracentesis. Gastroenterology.1996;111:1002–10.

67. Sort P, Navasa M, Arroyo V, Aldeguer X, Planas R, Ruiz-del-Arbol L, Castells L, Vargas V, Soriano G, Guevara M, GinesP, Rodes J. Effect of intravenous albumin on renal impair-ment and mortality in patients with cirrhosis and sponta-neous bacterial peritonitis. N Engl J Med. 1999;341:403–9.

68. Duvoux C, Zanditenas D, Hezode C, Chauvat A, MoninJL, Roudot-Thoraval F, Mallat A, Dhumeaux D. Effects ofnoradrenalin and albumin in patients with type I hepa-torenal syndrome: a pilot study. Hepatology. 2002;36:374–80.

69. Ortega R, Gines P, Uriz J, Cardenas A, Calahorra B, DeLas Heras D, Guevara M, Bataller R, Jimenez W, ArroyoV, Rodes J. Terlipressin therapy with and without albu-min for patients with hepatorenal syndrome: results of aprospective, nonrandomized study. Hepatology.2002;36:941–8.

70. Curry S, Mandelkow H, Brick P, Franks N. Crystal struc-ture of human serum albumin complexed with fatty acidreveals an asymmetric distribution of binding sites. NatStruct Biol. 1998;5:827–35.

71. Evans TW. Review article: albumin as a drug-biologicaleffects of albumin unrelated to oncotic pressure. AlimentPharmacol Ther. 2002;16(Suppl 5):6–11.

72. Mitzner SR, Stange J, Klammt S, Peszynski P, Schmidt R,Noldge-Schomburg G. Extracorporeal detoxificationusing the molecular adsorbent recirculating system forcritically ill patients with liver failure. J Am Soc Nephrol.2001;12 (Suppl 17):S75–82.

73. Stange J, Ramlow W, Mitzner S, Schmidt R, KlinkmannH. Dialysis against a recycled albumin solution enablesthe removal of albumin-bound toxins. Artif Organs.1993;17:809–13.

74. Stange J, Mitzner S. A carrier-mediated transport oftoxins in a hybrid membrane. Safety barrier between apatients blood and a bioartificial liver. Int J Artif Organs.1996;19:677–91.

75. Hughes R, Ton HY, Langley P, Davies M, Hanid MA,Mellon P, Silk DB, Williams R. Albumin-coated Amber-lite XAD-7 resin for hemoperfusion in acute liver failure.Part II: in vivo evaluation. Artif Organs. 1979;3:23–6.

76. Stange J, Mitzner S, Ramlow W, Gliesche T, Hickstein H,Schmidt R. A new procedure for the removal of proteinbound drugs and toxins. Asaio J. 1993;39:M621–5.

77. Falkenhagen D, Strobl W, Vogt G, Schrefl A, LinsbergerI, Gerner FJ, Schoenhofen M. Fractionated plasmaseparation and adsorption system: a novel system forblood purification to remove albumin bound substances.Artif Organs. 1999;23:81–6.

78. Rifai K, Ernst T, Kretschmer U, Bahr MJ, Schneider A,Hafer C, Haller H, Manns MP, Fliser D. Prometheus–anew extracorporeal system for the treatment of liverfailure. J Hepatol. 2003;39:984–90.

79. Herget-Rosenthal S, Lison C, Treichel U, et al. Citrateanticoagulated modified fractionated plasma separationand adsorption: first clinical efficacy and safety data inliver failure (Abstract). J Am Soc Nephrol. 2003;14:729A.

80. Mitzner SR, Klammt S, Peszynski P, Hickstein H, KortenG, Stange J, Schmidt R. Improvement of multiple organfunctions in hepatorenal syndrome during albumin dia-lysis with the molecular adsorbent recirculating system.Ther Apher. 2001;5:417–22.

81. Sorkine P, Ben Abraham R, Szold O, Biderman P, KidronA, Merchav H, Brill S, Oren R. Role of the molecular

adsorbent recycling system (MARS) in the treatment ofpatients with acute exacerbation of chronic liver failure.Crit Care Med. 2001;29:1332–6.

82. Stange J, Mitzner SR, Risler T, Erley CM, Lauchart W,Goehl H, Klammt S, Peszynski P, Freytag J, Hickstein H,Lohr M, Liebe S, Schareck W, Hopt UT, Schmidt R.Molecular adsorbent recycling system (MARS): clinicalresults of a new membrane-based blood purificationsystem for bioartificial liver support. Artif Organs.1999;23:319–30.

83. Schmidt LE, Svendsen LB, Sorensen VR, Hansen BA,Larsen FS. Cerebral blood flow velocity increases duringa single treatment with the molecular adsorbents recir-culating system in patients with acute on chronic liverfailure. Liver Transpl. 2001;7:709–12.

84. Schmidt LE, Sorensen VR, Svendsen LB, Hansen BA,Larsen FS. Hemodynamic changes during a single treat-ment with the molecular adsorbents recirculating systemin patients with acute-on-chronic liver failure. LiverTranspl. 2001;7:1034–9.

85. Mitzner SR, Stange J, Klammt S, Risler T, Erley CM,Bader BD, Berger ED, Lauchart W, Peszynski P, FreytagJ, Hickstein H, Loock J, Lohr JM, Liebe S, Emmrich J,Korten G, Schmidt R. Improvement of hepatorenal syn-drome with extracorporeal albumin dialysis MARS:results of a prospective, randomized, controlled clinicaltrial. Liver Transpl. 2000;6:277–86.

86. Sen S, Mookerjee RP, Cheshire LM, Davies NA, WilliamsR, Jalan R. Albumin dialysis reduces portal pressureacutely in patients with severe alcoholic hepatitis. JHepatol. 2005;43:142–8.

87. Stange J, Mitzner SR, Klammt S, Freytag J, Peszynski P,Loock J, Hickstein H, Korten G, Schmidt R, Hentschel J,Schulz M, Lohr M, Liebe S, Schareck W, Hopt UT. Liversupport by extracorporeal blood purification: a clinicalobservation. Liver Transpl. 2000;6:603–13.

88. Jalan R, Sen S, Steiner C, Kapoor D, Alisa A, Williams R.Extracorporeal liver support with molecular adsorbentsrecirculating system in patients with severe acute alco-holic hepatitis. J Hepatol. 2003;38:24–31.

89. Heemann U, Treichel U, Loock J, Philipp T, Gerken G,Malago M, Klammt S, Loehr M, Liebe S, Mitzner S,Schmidt R, Stange J. Albumin dialysis in cirrhosis withsuperimposed acute liver injury: a prospective, con-trolled study. Hepatology. 2002;36:949–58.

90. Hassanein T, Tofteng F, Brown RSJ, McGuire BM, LynchP, Mehta R, et al. Efficacy of albumin dialysis (MARS) inpatients with cirrhosis and advanced grades of hepaticencephalopathy: a prospective, controlled, randomizedmulticenter trial. Hepatology. 2004; 40:726A–727A.

91. Stange J, Hassanein TI, Mehta R, et al. Short-term survi-val of patients with severe intractable hepatic encephalo-pathy: The role of albumin dialysis. Hepatology.2005;42:286A.

92. Novelli G, Rossi M, Pretagostini R, Poli L, Novelli L,Berloco P, Ferretti G, Iappelli M, Cortesini R. MARS(Molecular Adsorbent Recirculating System): experiencein 34 cases of acute liver failure. Liver. 2002; 22(Suppl2):43–7.

93. Novelli G, Rossi M, Pretagostini R, Novelli L, Poli L,Ferretti G, Iappelli M, Berloco P, Cortesini R. A 3-yearexperience with Molecular Adsorbent Recirculating Sys-tem (MARS): our results on 63 patients with hepaticfailure and color Doppler US evaluation of cerebral per-fusion. Liver Int. 2003;23(Suppl 3):10–5.


94. Novelli G, Rossi M, Pretagostini M, Pugliese F, RubertoF, Novelli L, Nudo F, Bussotti A, Corradini S, Martelli S,Berloco PB. One hundred sixteen cases of acute liverfailure treated with MARS. Transplant Proc. 2005;37:2557–9.

95. Lahdenpera A, Koivusalo AM, Vakkuri A, HockerstedtK, Isoniemi H. Value of albumin dialysis therapy insevere liver insufficiency. Transpl Int. 2005;17:717–23.

96. Felldin M, Friman S, Backman L, Siewert-Delle A, Hen-riksson BA, Larsson B, Olausson M. Treatment with themolecular adsorbent recirculating system in patientswith acute liver failure. Transplant Proc. 2003;35:822–3.

97. Felldin M, Friman S, Olausson M, Backman L, CastedalM, Larsson B, Henriksson BA, Siewert-Delle A. [Liverdialysis using MARS in acute hepatic failure. Promisingresults in a pilot setting]. Lakartidningen. 2003;100:3836–8, 3841.

98. Abraham RB, Szold O, Merhav H, Biderman P, KidronA, Nakache R, Oren R, Sorkine P. Rapid resolution ofbrain edema and improved cerebral perfusion pressurefollowing the molecular adsorbent recycling system inacute liver failure patients. Transplant Proc. 2001;33:2897–9.

99. Schmidt LE, Wang LP, Hansen BA, Larsen FS. Systemichemodynamic effects of treatment with the molecularadsorbents recirculating system in patients with hyper-acute liver failure: A prospective controlled trial. LiverTranspl. 2003;9:290–7.

100. Awad SS, Swaniker F, Magee J, Punch J, Bartlett RH.Results of a phase I trial evaluating a liver supportdevice utilizing albumin dialysis. Surgery. 2001;130:354–62.

101. Tsai MH, Chen YC, Wu CS, Ho YP, Fang JT, Lien JM,Yang C, Chu YY, Liu NJ, Lin CH, Chiu CT, Chen PC.Extracorporal liver support with molecular adsorbentsrecirculating system in patients with hepatitis B-asso-ciated fulminant hepatic failure. Int J Clin Pract.2005;59:1289–94.

102. Novelli G, Rossi M, Ferretti G, Nudo F, Bussotti A,Mennini G, Novelli L, Ferretti S, Antonellis F, MartelliS, Berloco PB. Molecular adsorbent recirculating sys-tem treatment for acute hepatic failure in patients withhepatitis B undergoing chemotherapy for non-Hodg-kin’s lymphoma. Transplant Proc. 2005;37:2560–2.

103. El Banayosy A, Kizner L, Schueler V, Bergmeier S,Cobaugh D, Koerfer R. First use of the MolecularAdsorbent Recirculating System technique on patientswith hypoxic liver failure after cardiogenic shock. AsaioJ. 2004;50:332–7.

104. Lamesch P, Jost U, Schreiter D, Scheibner L, Beier O,Fangmann J, Hauss J. Molecular adsorbant recirculat-ing system in patients with liver failure. TransplantProc. 2001;33:3480–2.

105. Manz T, Ochs A, Bisse E, Strey C, Grotz W. Liver sup-port – a task for nephrologists? Extracorporeal treat-ment of a patient with fulminant Wilson crisis. BloodPurif. 2003;21:232–6.

106. Sen S, Felldin M, Steiner C, Larsson B, Gillett GT,Olausson M, Williams R, Jalan R. Albumin dialysisand Molecular Adsorbents Recirculating System(MARS) for acute Wilson’s disease. Liver Transpl.2002;8:962–7.

107. Kreymann B, Seige M, Schweigart U, Kopp KF, ClassenM. Albumin dialysis: effective removal of copper in apatient with fulminant Wilson disease and successful

bridging to liver transplantation: a new possibility forthe elimination of protein-bound toxins. J Hepatol.1999;31:1080–5.

108. McIntyre CW, Fluck RJ, Freeman JG, Lambie SH. Use ofalbumin dialysis in the treatment of hepatic and renaldysfunction due to paracetamol intoxication. NephrolDial Transplant. 2002;17:316–7.

109. Siewert-Delle A, Henriksson BA, Baeckmann L. Albu-min dialysis with the MARS for a patient with acuteliver failure due to paracetamol intoxication: a casereport. Z Gastroenterol. 2001;39(Suppl):48.

110. Shi Y, He J, Chen S, Zhang L, Yang X, Wang Z, Wang M.MARS: optimistic therapy method in fulminant hepaticfailure secondary to cytotoxic mushroom poisoning–acase report. Liver. 2002;22(Suppl 2):78–80.

111. Catalina MV, Nunez O, Ponferrada A, Menchen L,Matilla A, Clemente G, Banares R. [Liver failure due tomushroom poisoning: clinical course and new treat-ment perspectives]. Gastroenterol Hepatol. 2003;26:417–20.

112. Koivusalo AM, Yildirim Y, Vakkuri A, Lindgren L,Hockerstedt K, Isoniemi H. Experience with albumindialysis in five patients with severe overdoses of para-cetamol. Acta Anaesthesiol Scand. 2003;47:1145–50.

113. Novelli G, Rossi M, Pretagostini R, Poli L, Peritore D,Berloco P, Di Nicuolo A, Iappelli M, Cortesini R. Use ofMARS in the treatment of acute liver failure: prelimin-ary monocentric experience. Transplant Proc. 2001;33:1942–4.

114. Stange J, Hassanein TI, Mehta R, Mitzner SR, BartlettRH. The molecular adsorbents recycling system as aliver support system based on albumin dialysis: a sum-mary of preclinical investigations, prospective, rando-mized, controlled clinical trial, and clinical experiencefrom 19 centers. Artif Organs. 2002;26:103–10.

115. Kellersmann R, Gassel HJ, Buhler C, Thiede A,Timmermann W. Application of Molecular AdsorbentRecirculating System in patients with severe liver fail-ure after hepatic resection or transplantation: initialsingle-centre experiences. Liver. 2002;22(Suppl 2):56–8.

116. Mullhaupt B, Kullak-Ublick GA, Ambuhl P, MaggioriniM, Stocker R, Kadry Z, Clavien PA, Renner EL. Firstclinical experience with Molecular Adsorbent Recircu-lating System (MARS) in six patients with severe acuteon chronic liver failure. Liver. 2002;22(Suppl 2):59–62.

117. Delafosse B, Garnier E, Dumortier J, Boillot O. EdouardHerriot experience with the MARS in 5 patients experi-encing hepatic failure. Z Gastroenterol. 2001;39:38.

118. Sen S, Jalan R, Morris EC, Steiner C, Mackinnon S,Williams R. Reversal of severe cholestasis caused bychronic graft-versus-host disease with the MARSliver-support device. Transplantation. 2003;75:1766–7.

119. Notohamiprodjo M, Banayosy A, Kizner L, Schuller V,Korfer R. One year experience with MARS therapy inpatients with multiorgan failure in Cardiac Center BadOyenhausen. Z Gastroenterol. 2001;39(Suppl):51.

120. Sturm E, Franssen CF, Gouw A, Staels B, Boverhof R, DeKnegt RJ, Stellaard F, Bijleveld CM, Kuipers F. Extra-corporal albumin dialysis (MARS) improves cholestasisand normalizes low apo A-I levels in a patient withbenign recurrent intrahepatic cholestasis (BRIC).Liver. 2002;22(Suppl 2):72-5.

121. Mullhaupt B, Kullak-Ublick GA, Ambuhl PM, StockerR, Renner EL. Successful use of the Molecular


Adsorbent Recirculating System (MARS) in a patientwith primary biliary cirrhosis (PBC) and treatmentrefractory pruritus. Hepatol Res. 2003;25:442–446.

122. Doria C, Mandala L, Smith J, Vitale CH, Lauro A,Gruttadauria S, Marino IR, Foglieni CS, Magnone M,Scott VL. Effect of molecular adsorbent recirculatingsystem in hepatitis C virus-related intractable pruritus.Liver Transpl. 2003;9:437–43.

123. Macia M, Aviles J, Navarro J, Morales S, Garcia J. Effi-cacy of molecular adsorbent recirculating system forthe treatment of intractable pruritus in cholestasis. AmJ Med. 2003;114:62–4.

124. Bellmann R, Graziadei IW, Feistritzer C, SchwaighoferH, Stellaard F, Sturm E, Wiedermann CJ, Joannidis M.Treatment of refractory cholestatic pruritus after livertransplantation with albumin dialysis. Liver Transpl.2004;10:107–14.

125. Mela M, Mancuso A, Burroughs AK. Review article:pruritus in cholestatic and other liver diseases. AlimentPharmacol Ther. 2003;17:857–70.

126. Sen S, Ratnaraj N, Davies NA, Mookerjee RP, CooperCE, Patsalos PN, Williams R, Jalan R. Treatment ofphenytoin toxicity by the molecular adsorbents recir-culating system (MARS). Epilepsia. 2003;44:265–7.

127. Sen S, Ytrebo LM, Rose C, Fuskevaag OM, Davies NA,Nedredal GI, Williams R, Revhaug A, Jalan R. Albumindialysis: a new therapeutic strategy for intoxicationfrom protein-bound drugs. Intensive Care Med. 2004.

128. Sen S, Davies NA, Mookerjee RP, Cheshire LM, HodgesSJ, Williams R, Jalan R. Pathophysiological effects ofalbumin dialysis in acute-on-chronic liver failure: Arandomized controlled study. Liver Transpl. 2004;10:1109–19.

129. Sauer IM, Goetz M, Steffen I, Walter G, Kehr DC,Schwartlander R, Hwang YJ, Pascher A, Gerlach JC,Neuhaus P. In vitro comparison of the molecular adsor-bent recirculation system (MARS) and single-passalbumin dialysis (SPAD). Hepatology. 2004;39:1408–14.

130. Seige M, Kreymann B, Jeschke B, Schweigart U, KoppKF, Classen M. Long-term treatment of patients withacute exacerbation of chronic liver failure by albumindialysis. Transplant Proc. 1999;31:1371–5.

131. Sauer IM, Zeilinger K, Pless G, Kardassis D, TheruvathT, Pascher A, Goetz M, Neuhaus P, Gerlach JC. Extra-corporeal liver support based on primary human livercells and albumin dialysis–treatment of a patient withprimary graft non-function. J Hepatol. 2003;39:649–53.

132. Kramer L, Bauer E, Schenk P, Steininger R, Vigl M,Mallek R. Successful treatment of refractory cerebraloedema in ecstasy/cocaine-induced fulminant hepaticfailure using a new high-efficacy liver detoxification

device (FPSA-Prometheus). Wien Klin Wochenschr.2003;115:599–603.

133. Krisper P, Haditsch B, Stauber R, Jung A, Stadlbauer V,Trauner M, Holzer H, Schneditz D. In vivo quantifica-tion of liver dialysis: comparison of albumin dialysisand fractionated plasma separation. J Hepatol.2005;43:451–7.

134. Evenepoel P, Laleman W, Wilmer A, Claes K, Maes B,Kuypers D, Bammens B, Nevens F, Vanrenterghem Y.Detoxifying capacity and kinetics of prometheus–a newextracorporeal system for the treatment of liver failure.Blood Purif. 2005;23:349–58.

135. Evenepoel P, Laleman W, Wilmer A, Claes K, KuypersD, Bammens B, Nevens F, Vanrenterghem Y. Pro-metheus versus molecular adsorbents recirculating sys-tem: comparison of efficiency in two different liverdetoxification devices. Artif Organs. 2006;30:276–84.

136. Steiner C, Zinggrebe A, Viertler A. Experiences withMARS therapy in liver disease: analysis of 385 patientsof the International MARS Registry. Hepatology. 2003;38:239A.

137. Sen S, Steiner C, Williams R, Jalan R. Artificial liversupport: Overview of Registry and controlled clinicaltrials. In: Arroyo V, Forns X, Garcia-Pagan JC, Rodes J,eds. Progress in the treatment of liver diseases. Barce-lona: Ars Medica; 2003:429–35.

138. Mallet VO, Mitchell C, Mezey E, Fabre M, Guidotti JE,Renia L, Coulombel L, Kahn A, Gilgenkrantz H. Bonemarrow transplantation in mice leads to a minor popu-lation of hepatocytes that can be selectively amplified invivo. Hepatology. 2002;35:799–804.

139. Kobayashi N, Fujiwara T, Westerman KA, Inoue Y,Sakaguchi M, Noguchi H, Miyazaki M, Cai J, TanakaN, Fox IJ, Leboulch P. Prevention of acute liver failurein rats with reversibly immortalized human hepato-cytes. Science. 2000;287:1258–62.

140. Seternes T, Sorensen K, Smedsrod B. Scavengerendothelial cells of vertebrates: a nonperipheral leuko-cyte system for high-capacity elimination of wastemacromolecules. Proc Natl Acad Sci USA. 2002;99:7594–7.

141. Auth MK, Joplin RE, Okamoto M, Ishida Y, McMasterP, Neuberger JM, Blaheta RA, Voit T, Strain AJ. Mor-phogenesis of primary human biliary epithelial cells:induction in high-density culture or by coculture withautologous human hepatocytes. Hepatology. 2001;33:519–29.

142. Sirica AE, Gainey TW. A new rat bile ductular epithelialcell culture model characterized by the appearance ofpolarized bile ducts in vitro. Hepatology. 1997;26:537–49.


5Glucose Sensors and Insulin Pumps: Prospectsfor an Artificial Pancreas

Martin Press

Introduction

Insulin is an extraordinary hormone. When weeat, as any medical student will tell you, insulinstimulates glucose uptake into insulin-sensitivetissues (muscle, fat, liver). What they oftenappreciate less is that this effect requires thehigh insulin levels which occur only transientlyafter meals. When we are not eating, there is acontinued need for glucose homeostasis, andinsulin then, at much lower concentrations, reg-ulates glucose production, predominantly fromglycogen stores in the liver. Patients are oftenbaffled that they can wake up with higher glucoselevels than those they went to bed with despitenot having eaten anything, but hepatic glucoseproduction, regulated by insulin, is the maindeterminant of fasting blood glucose levels. Insu-lin, in effect, does the work of two hormones.

It does so by having two quite different dose–response curves in different organs (Figure 5-1).The steep part of these respective curves is seenat concentrations which are some 10-fold apart.Thus, at fasting insulin concentrations, smallchanges in insulin, such as are constantly seenas the pancreatic b-cells respond to minutechanges in glucose concentrations, have a neg-ligible effect on glucose uptake. However, thissame fluctuation in insulin levels modulatessignificant changes in hepatic glucose produc-tion and is thus able to rapidly correct anytendency for glucose levels to stray outside avery narrow physiological range. Following ameal, not only is glucose uptake stimulated,but hepatic glucose production is also almostcompletely suppressed.

It is not clear why it is that the liver has notevolved to be able to measure glucose levels itselfand has to rely on the pancreatic b-cells to ‘tell’ ithow much glucose to produce. So long as thepancreatic b-cells are working well it presents noproblems. But if a diabetic patient gives himselftoo much insulin, the resultant high-circulatinginsulin levels are read by the liver as the signalthat glucose levels are too high and hepatic glucoseproduction promptly falls, resulting in hypoglyce-mia. Conversely, insulin deficiency as a result ofpancreatic b-cell failure is the signal to the liverthat glucose levels are low, with the result thathepatic glucose production increases inappropri-ately. In the extreme situation of diabetic ketoaci-dosis, when the pancreatic b-cells fail completely,the liver is pouring glucose into the circulation totry to correct a non-existent state of hypoglycemia.

Conventional Insulin Treatment

In vivo, insulin is secreted in pulses every10–15 min. Given that the half life of insulin inthe circulation is less than 5 min, this results inwide short-term variation in insulin concentra-tions which enables minute-to-minute regulationof blood glucose levels. While the replacement ofother endocrine hormones with long half lives,such as thyroxine, does not present a great clini-cal problem, the idea of replacing a hormone suchas insulin whose concentrations are constantlychanging is another matter altogether and indeedit is amazing that in patients with Type 1 diabeteswe manage as well as we do with just a few injec-tions a day.


77

Traditional insulin therapy has in the pastinvolved relatively ‘dirty’ insulin preparations,purified from animal pancreases. When insulinwas first introduced into clinical practice in the1920s this did not matter as the impressive pointwas that one was preventing what would other-wise be inevitable death from diabetic ketoaci-dosis. However, the impure insulin preparationsextracted from animal pancreases elicit animmunological response which involves insulinantibodies which tend to buffer sudden changesin insulin concentrations and may cause immu-nological reactions at the injection sites. Withthe appreciation that a more sophisticatedapproach was necessary in order to achieve ade-quate metabolic control to prevent diabeticcomplications and optimize quality of life, theshortcomings of the old insulin preparationsbecame even more apparent.

In the 1970s, so-called mono-componentinsulins were introduced which contained veryfew impurities. These were extracted from pigs,whose insulin is only 1 amino acid differentfrom man (compared to bovine insulin, which

is 3 amino acids different). As a result, theyelicited less in the way of an antibody response.Since then, there has been an almost completemove toward recombinant DNA technology tosynthesize insulin de novo in the laboratory. Theresultant insulin is 100% pure and is identical tonaturally produced human insulin. Indeed, itsoriginal marketing bore a strong resemblance tothat of Coca Cola in being ‘the real thing’.

Because soluble, ‘regular’ insulin synthesizedin this way elicits little or no significant antibodyresponse, it tends to be shorter acting thanextracted insulin. However, the recombinantDNA technology has been taken one step furtherand, by means of substitutions, deletions, orrearrangements within the molecule, manufac-turers have been able to produce insulins witheven shorter durations of action. There are cur-rently three genetically modified ultra-short act-ing insulins on the British market, lispro, aspart,and glulisine, each the result of different mole-cular modifications, but all with essentially thesame pharmacokinetics.

The physiological significance of this geneticengineering is that we now have insulins whoseduration of action is sufficiently short as tomimic the narrow insulin peaks secreted phy-siologically with meals. Soluble (regular) insulinas injected exists as a hexamer and, since it takessome 30–40 min to dissociate into monomersand enter the circulation, the injection needs tobe given 30–40 min before the patient wants toeat. This causes practical problems, especiallywith children. Even then, the biological effectof the insulin is still not fast enough to preventa rapid rise in blood glucose levels, so thatpatients have to be encouraged to eat smallmeals and then a snack 3 or 4 h later, when themeal has gone but the insulin is still working, ifthey wish to avoid hypoglycemia before the nextmeal.

In contrast, with ultra-short acting analoguesthe dissociation is so rapid that there is analmost immediate rise in circulating insulinlevels and the elevation in insulin levels lastsfor only 3 or 4 h. When the meal effect is gone,so is the insulin. This means that the insulin can(indeed, must) be taken immediately before thepatient eats, and mid-morning, mid-afternoon,and bedtime snacks are not necessary. Thepatient can eat (within reason) whatever theywish and adjust the insulin dose to the carbohy-drate content of the meal so that the old idea ofneeding to restrict carbohydrate intake has been

2

Glu

cose

tur

nove

r (m

g/kg

/min

)

INSULINfed rangefasting range

2

Glu

cose

tur

nove

r (m

g/kg

/min

)

INSULINfed rangefasting range

PERIPHERAL UPTAKEPERIPHERAL UPTAKE

HEPATIC PRODUCTIONHEPATIC PRODUCTION

Figure 5-1. Glucose kinetics (with kind permission of OxfordUniversity Press). Glucose uptake is stimulated by insulin, but only atthe elevated concentrations seen after meals. At low insulin levels,glucose uptake is not zero because of insulin-independent uptake bytissues such as the brain, in particular. Glucose production, in contrast, isregulated at the low insulin levels typically seen in the fasting state, inwhich glucose production is the main determinant of fasting glucoselevels.


replaced with a philosophy of learning the car-bohydrate content of foods and matching insu-lin doses to the carbohydrate intake.

As emphasized above, these high-insulinpeaks are present for only a few hours aftereach meal. In between meals, low levels of insu-lin are still necessary to regulate hepatic glucoseproduction and maintain normoglycemia. Withconventional insulin regimens, this has alwaysproved difficult due partly to the vagaries ofabsorption of depot insulin preparations withmarked variation from one injection to thenext and partly to the fact that insulin levelsfollowing an injection inevitably peak and thenwane. The latter is particularly a problem over-night when insulin requirements are typicallylow in the first part of the night but rise in thesecond part (the so-called ‘dawn phenomenon’),just when the effect of an intermediate actinginsulin is waning.

What is ideally required therefore, is a systemof administering insulin which provides a rela-tively constant ‘basal’ infusion of insulin aroundthe clock, with ‘boluses’ of insulin which aregiven just before the patient eats. This is whatthe insulin pumps currently available seek toprovide. They give a constant trickle of insulinas a basal rate with extra as boluses whenevertold to do so. They can be programmed to give adifferent basal rate at different times of the day(a higher rate from 5 am until 8 am, for example)while at the same time giving complete flexibilitywith respect to the number of times the patienteats and the dose of insulin needed to balancethe amount of carbohydrate the meals or snackscontain. A recent study showed that metaboliccontrol was a direct function of the number ofinsulin boluses given1.

Open Loop Insulin Pumps: History

The concept of a portable pump, permanentlyattached to the patient and under the patient’scontrol, was first developed in the late 1970s. Thefirst publication on ‘continuous subcutaneousinsulin infusion’ (CSII), by Dr John Pickup ofGuy’s Hospital, was published in the British Med-ical Journal in 19782 and the pump that he used isnow in the Science Museum in London.

At that time, small pumps were used to givein-patients morphine or other drugs for rela-tively short periods of time. They were designedto be left on the bedside table. Some of the first

attempts to give insulin from a portable pumpinvolved simply strapping these same pumps tothe patient’s belt but they were desperatelyheavy and crude. Smaller, lighter pumps wererapidly developed. The first studies in Englandwere done predominantly with the Mill Hill infu-ser, while those in the USA used a purpose builtpump made by Auto-Syringe in New Hampshire(Figure 5-2). Dean Kamen, who dropped out ofcollege to start Auto-Syringe in 1979, hasbecome better known recently for inventing theSegway personal transporter, from whichGeorge Bush famously fell off. He is in thegreat tradition of eccentric inventors.3 He soldAuto-Syringe after a few years to Baxter Inter-national for $30 million and, with the money,bought an island off the coast of Connecticut,declared independence from the United Statesand appointed the founders of Ben and Jerry’s asjoint Ministers of Ice Cream!

MiniMed, the company which has dominatedthe development of diabetes technology sincetheir first pump was launched in 1983, was alsovery much associated with one particular man.Mr. Al Mann personally funded the developmentof insulin pumps and glucose sensors by Mini-Med for 20 years and without his personal com-mitment it is very doubtful that an artificialpancreas would even be on the horizon at thispoint.

While it was the technology that grabbed theheadlines, equally important was the change inmindset whereby it was left to the patient to takeresponsibility for their diabetes, followingchanges in their blood glucose levels with self-administered finger pricks and a portable glucosemeter. This in turn arose from Dr Clara Lowy’sstudy, also published in 1978,4 on women whomonitored their diabetic control during preg-nancy and who, to the doctors’ astonishment,chose to continue to monitor their blood glucoselevels afterwards because it put them in control.

All pumps currently available commerciallyuse broadly the same principles. Inside a casethe size of a small mobile phone is a battery anda small electric motor which drives a syringepump. This delivers a near-continuous infusionof insulin through a long tube to a plastic can-nula which is inserted through the skin intosubcutaneous adipose tissue. The cannula istypically changed twice a week while the insulinreservoir has to be topped up at intervals whichdepend on the insulin requirements of thepatient.

Glucose Sensors and Insulin Pumps 79

Several companies have brought insulinpumps to market over the years, and at presentin Britain, there are at least four available. Theoriginal Auto-Syringe pump had only two but-tons, labeled ‘basal’ and ‘bolus’ (Figure 5-2). Thebasal rate could be adjusted in steps of 1 unit per24 h and it was a long time before other manu-facturers offered such fine adjustment, whichis necessary because the slope of the dose–response curve of glucose production versusinsulin is so steep (Figure 5-1). Other aspects ofpump development consisted largely of impro-ved safety features and more sophisticated pro-grammes for varying the shape and duration ofbolus delivery. Roche offers the improved conve-nience of prefilled insulin syringes. Most centersnow use ultra-short acting insulin analoguesrather than insulin itself to take advantage of thenarrowness of the post-meal insulin peak andthe speed with which it is possible to changeeffective circulating insulin levels.

Open Loop Insulin Pumps: Pros and Cons

Compared to what is now regarded as a conven-tional regimen of multiple daily injections(MDI), pumps have both advantages and disad-vantages. Thus, there is no need with the pumpto inject yourself, but on the other hand youhave to replace the giving set and cannula twicea week. There is no peaking or waning of insulin

effect as seen overnight and in the inter-prandialperiods with conventional regimens. On theother hand, if there is an interruption in insulinsupply for more than an hour or two, there israpid metabolic decompensation leading tohyperglycemia and ketosis since there is no sub-cutaneous insulin reserve.

Most studies comparing CSII and MDI havedescribed either a reduction in glycated hemo-globin without a corresponding increase in theincidence of hypoglycemia or a reduction inhypoglycemia without a concomitant rise inHbA1c.

5 However, to most patients on an insulinpump, the major advantage relates to flexibilityand issues of lifestyle and quality of life ratherthan directly to the diabetic control itself.

Although CSII offers the convenience ofbeing able to programme the pump to allow forreproducible fluctuations in insulin sensitivity,most centers start with a single daily basal rate,which is typically some 20–30% less than thepatient has previously been receiving fromtheir long-acting insulin because of betterabsorption. The pump makes it possible toallow for diurnal variation in insulin sensitivity(we are most insulin resistant in the early hoursof the morning as the result of growth hormonesecretion during deep sleep in the early part ofthe night). Disetronic (since taken over byRoche) took this concept to extremes with theirphilosophy of programming frequent changes inbasal rate around the clock to match changes in

Figure 5-2. Original Auto-Syringeinsulin pump, circa 1981. A smallsyringe filled with insulin was con-nected to a catheter inserted into sub-cutaneous fat via a fine tube. Thesyringe was driven by an electricmotor the speed of which was depen-dent on the ‘basal’ and ‘bolus’ settingsof the pump.


average diurnal insulin sensitivity. Thisapproach failed to take into account the factthat everyone is different and no one exactlyfollows the average. In practice, a single 24 hconstant basal rate usually works surprisinglywell and in order to do better, there is no alter-native but to work out an individualized pro-gramme of basal rates for each patient. Eventhen, the necessary diurnal changes in basalrates vary from one day to another in a givenpatient, which is why perfect control is notattainable with an open loop system.

Overnight basal rates are adjusted to givesatisfactory mean fasting glucose levels. So longas there is no mid-morning snack, morningbasal levels are adjusted according to pre-lunchglucose levels and so on. Some centers omit theprevious meal and meal bolus to adjust the basalrate but this may change insulin sensitivity andgives the whole procedure a spurious accuracy.Patients are not machines and do not giveexactly the same results if tested more than once.

Meal boluses are calculated on the basis ofmeal carbohydrate content and we are not theonly unit to have obtained excellent results bygiving patients an intensive educational course,modeled on the principles of DAFNE (DosageAdjustment For Normal Eating) and BERTIE(Bournemouth’s Education Resources for Train-ing in Insulin and Eating), to teach them toestimate the doses they are likely to need for aparticular meal.

Just as with the basal rates, it is necessary tomodify and adapt the formula used to calculatemeal boluses in a particular patient. Thus, forexample, the patient may start with, say, 1 unitper 10 g carbohydrate ingested and increase ordecrease this as necessary according to theirresponse. Sometimes it may be necessary tovary the ratio to allow for fluctuations in diurnalinsulin sensitivity. Thus, for example, if they eat60 g of carbohydrate for both breakfast andlunch, many patients find that while 6 units ofinsulin will be sufficient at lunchtime, they mayneed 7 or 8 units to cover breakfast. This mayalso reflect some ‘carry-over’ of insulin from thebreakfast dose. Supper usually needs more insu-lin in any case because for most patients it is thebiggest meal of the day.

The real beauty of the pump, however, is thatit gives such great flexibility with respect to thetiming and amount of meals. With a conven-tional insulin regimen, once the insulin injectionhas been given, one has to eat to match the

insulin or risk hypoglycemia. Conversely, ifone eats more than one had intended when theinjection was given, one has to either accept thatglucose levels will run high or give oneself anextra injection. With the pump, it is perfectlypossible to give a bolus to cover one’s maincourse and only when dessert comes to decidewhether to give extra insulin to cover it. Extrasmall boluses of insulin to cover between-mealsnacks can be given without any problem.

When we exercise, circulating insulin levelstypically fall by some 30% and this, together withan increase in glucagon, adrenaline, and growthhormone, modulates the mobilization of thenecessary metabolic fuels. In patients withType 1 diabetes, there is no way of reducinginsulin levels, which are determined by the pre-vious injection of insulin, except with the pump.As alluded to above, the lack of a subcutaneousinsulin reservoir is a drawback if the insulinsupply from the pump is interrupted, with oneexception. This is exactly what is wanted duringexercise.

Patients taking injections have little choicebut to take extra sugar to prevent hypoglycemiawhen they exercise. However, giving sets onmodern pumps have a quick release connectorwhich allows patients to simply disconnect thepump and take it off when they exercise. If theyare going to exercise for more than half an hour,they should keep the pump on but can turn thebasal rate down to mimic what normal b-cellswould do.

Modern ‘open-loop’ insulin pumps are in nosense automatic. Just as a car does not go round acorner unless you turn the steering wheel, so apump has to be told how much insulin to give.This in turn means that the patient has to doregular finger-prick glucose estimations. Unsur-prisingly, the more finger-pricks the patient does,the better the control. One of the fashionable‘tools’ available to patients on pumps who areprepared to do more than the minimum numberof finger-pricks is to give correction doses tobring down a high glucose level. The disadvan-tage of this is first that it is difficult, when givinginsulin without food, not to give too much, andovershoot into hypoglycemia, and second that ittends to make patients err on the cautious sidewith their meal doses. One might add that itencourages patients to do 8 or 10 finger-pricks aday, which some of us might think would bespending too much time and energy on theirdiabetes and too little on the rest of their lives!


The latest version of Medtronic’s pumpincorporates a ‘Bolus Wizard’ to help the patientto work out how much insulin they need (Rochemarket a similar bolus calculator and othermanufacturers are also developing ‘smartpumps’). It takes into account not only whatyour blood sugar level is and how much carbo-hydrate you are planning to eat but also calcu-lates and allows for the amount of insulin stillacting in your circulation from your previousmeal bolus. Many patients find this ingeniouspiece of software invaluable although they haveto realize that the Wizard’s suggested doses areno more than a recommendation and they needto have the confidence to override the sugges-tions when necessary. Furthermore, high post-prandial glucose levels as a result of not takingenough insulin with the previous meal comedown anyway, and the idea of chasing themdown faster carries the danger of making thepatient more neurotic about the risk to theirhealth of a brief post-prandial peak than isappropriate. I am personally much more con-cerned about a high fasting glucose valuebecause it will probably have been high all night.

There are two other drawbacks which areunique to the pump. The first is cost. Mostmanufacturers charge roughly £2,500 for thepump itself but on top of this is the cost ofgiving sets and cannulas, which comfortablyadd another £1,000 per year, not to mentionthe strips for the glucose meter for the extrafinger-pricks which are necessary if one is totake full advantage of what the pump has tooffer. Both the pump manufacturers and theglucose meter manufacturers, like mobilephone manufacturers, tend to give the basicequipment away quite cheaply for what it isand then make their money on the supplies,sticks, or phone calls, respectively. In Britain,recommended targets for diabetes controlwhich made the patient eligible for reimburse-ment for pumps and pump supplies were pub-lished in 2003. Patients who could not achievelevels of glycated hemoglobin below 7.5% (or6.5% if they had other cardiovascular risk fac-tors) ‘without disabling hypoglycemia’ weredeemed eligible. The levels chosen meant thata far higher percentage of patients with Type 1diabetes were eligible than the 2 or 3% currentlyon insulin pumps in the United Kingdom.Clearly there would be very substantial implica-tions for the health-care budget if all suchpatients were started on CSII.

Finally, there is the issue of visibility and bodyimage: the concept that one is wearing a badge ofillness. Even if the pump is invisible to an out-sider (and there is no reason why it should notbe), it is a constant reminder to the patient him-self of the fact that they have a medical problem.When pumps first became available on a largescale in the United States in the early 80s, the firstpeople to want them were, unsurprisingly, thosefor whom this was not a big issue and who werekeen to be part of the new technological revolu-tion. Twenty-five years later, pumps are old hat(at least the open loop pumps available commer-cially) and it is much more a question of theattitude of the professional group managing thepatient and whether the money will be reim-bursed which decides whether a given patientgoes onto an insulin pump.

Some groups start patients with newly diag-nosed Type 1 diabetes on pumps from the be-ginning, even though such patients are usuallyrelatively easy to manage for the first year or twobecause they have significant residual b-cell func-tion. The logic for this is first that this is what theyare going to need in due course and the soonerthey get used to the idea the better, and second,that there is evidence that diabetic control in thefirst years of diabetes has a disproportionate influ-ence on the risk of complications in later years.

Open-loop pumps have advantages over mul-tiple injections,5 particularly in patients whoselife style varies considerably from one day to thenext. When we put people on pumps, we alwaysstress that this is an experiment and that, if theydo not like it, they can go back to injectionswithout having lost anything. Very few peopledo. Clearly the patients who go onto pumpshave been selected because we think they willsuit them but in practice, once a patient has gotused to being attached to a gadget 24 h a day thereis rarely any looking back.

Unfortunately, even with the most highlymotivated patient, it is still not possible toachieve perfect control. Blood glucose is influ-enced by things of which we still have littleunderstanding so that even a patient who eatsexactly the same amount of exactly the samefood, and who takes exactly the same amountof exercise at the same time each day, will havedifferent glycemic profiles from one day to thenext. Instability and unpredictability are thehallmarks of Type 1 diabetes. This high variancemeans that attempts to achieve completely nor-mal average glucose levels are inevitably limited


by the resultant increase in the frequency ofhypoglycemic episodes.

This is quite simply the nature of Type 1 dia-betes and does not imply failure on the part ofeither the patient or the medical team. Thus, thelandmark Diabetes Control and ComplicationsTrial (DCCT), which attempted to achieve thebest possible control with either pumps or multi-ple injections, managed only an average HbA1c,reflecting the average blood glucose for the past 2months or so, of 7%, compared to an upper limitof normal for subjects without diabetes of 6.1.6

This was because the incidence of severe hypo-glycemia (defined as requiring the intervention ofothers) was 62 per 100 patient years (compared to19 in the conventionally treated control group).Thus, the markedly increased incidence of hypo-glycemia prevented normalization of HbA1c, andit is hardly surprising that intensive treatment,while it reduced the incidence of complications,failed to prevent them altogether.

Intraportal Insulin Delivery

In the 1990s, three manufacturers (Siemens,Infusaid, and Minimed) explored whether therewere advantages to infusing the insulin into theperitoneal cavity rather than into fat. The think-ing was first that this would mimic physiologicaldelivery of insulin into the portal vein so that theliver would see much higher concentrationsthan other organs, and second that an artificialpancreas would probably involve a fullyimplanted system, so that this would be thenext logical step. Both Siemens and Infusaiddropped out in the 1990s because of prohibitivedevelopment costs while MiniMed went on to doa clinical trial of a closed loop system which usedan intravenous sensor and intraperitoneal insu-lin delivery which was abandoned after a fewyears because of sensor problems.

Intraperitoneal insulin delivery has onemajor clinical advantage, which is a muchreduced risk of hypoglycemia. At the sametime, it has several major disadvantages. Itrequires a small operation to implant thepump. The insulin reservoir has to be refilledevery few months in the hospital and furthersurgery is needed every few years to replace thebatteries. There were major problems with insu-lin aggregation and blockage of the catheter tubewith the specially manufactured 400 U/ml insu-lin which was necessary, necessitating a redesign

of the pump with a side-port which could beused to flush the catheter. Intraperitoneal insu-lin delivery is arguably the best route by whichto give insulin but at present it is out of fashionbecause of these drawbacks and current devel-opment work on artificial pancreases is focusingon subcutaneous insulin delivery with ultra-short-acting insulin analogues.

Need for Closed Loop System

Although subcutaneous insulin infusion pumpshave been available since the early 1980s, therewere major issues with the development of con-tinuous glucose sensors and they did notbecome commercially available until 1999.They showed vividly just how bad glucose con-trol is in patients with Type 1 diabetes. In anunselected group of patients, Bode showed thatpatients with Type 1 diabetes spend only 14.5 hper day in the euglycemic range (70–180 mg/dl,3.9–10.0 mmol/l), with 2.3 h below and 7.2 habove these limits.7 Apart from the increasedrisk of complications from the hyperglycemicvalues, patients will feel less well as the resultof the fluctuation of their glucose levels whileepisodes of hypoglycemia may compromisetheir quality of life still further.

In order to achieve near-normal glucose con-trol, and in particular to achieve it without hypo-glycemia, one needs a closed loop system wherebyblood glucose levels are monitored on a minute-to-minute basis and the amount of insulin infusedinto the circulation is also varied on a minute-to-minute basis to keep glucose levels normal. Inthe fasting state, insulin levels need to changevery rapidly to counter changes in glucose levels.Furthermore, immediately after meals, when largeamounts of glucose enter the blood stream fromthe digestion of starch, the magnitude of the neces-sary increase in insulin levels needs to be consid-erable. Conversely, an immediate fall in insulinlevels is necessary in response to exercise to facil-itate the mobilization of metabolic fuels. All ofthese are of course exactly what healthy pancreaticb-cells do very effectively.

Transplants

Replacement of the b-cells is perfectly possibleeither by transplanting a whole pancreas or,because the patient with Type 1 diabetes needs


only the b-cell containing islets, by isolatingand transplanting the islets alone. Whole pan-creas transplantation carries a high chance oflong-term insulin independence. The operationis reserved almost exclusively for patients need-ing a kidney transplant for end-stage diabeticnephropathy who will require a kidney.

Pancreatic islet transplantation is less invasivesince the islets comprise only some 2–3 ml andthey can be transplanted without the need forsurgery. However, it is not easy to isolate suffi-ciently large numbers of viable islets from a cada-ver pancreas. To compensate for the fact that somany islets are lost during the procedure of iso-lation and transplantation, more than one trans-plant is usually necessary, making the procedurecurrently somewhat extravagant. Although dia-betes is stabilized and serious hypoglycemia isabolished, long-term insulin independence isachieved in only a minority of patients.

Most importantly, however, the need for long-term immunosuppression, with the inevitabledrug side effects and increased incidence of infec-tions and malignancies, makes either procedureless than ideal. One is effectively exchanging oneproblem for another and unless the diabetes iscausing major problems this exchange is not jus-tified. Furthermore, the shortage of donor organsmeans that transplantation is not going to be thesolution except for a small minority of patientsfor the foreseeable future until a source of b-cellsin quantity becomes available.

Closed Loop Insulin Pumps: History

The attraction of closed loop insulin pumps isthat one will in theory have the advantages of atransplant without the drawbacks. The credit forthe first closed-loop insulin pumps goes toMichael Albisser in Canada8 and Ernst-FriedrichPfeiffer in Germany in 1974. These initial experi-ments led to the development of the Biostator,9 alarge and cumbersome machine suitable only forbedside use (Figure 5-3). Heparin was infusedvery slowly through the central lumen of a dou-ble lumen catheter inserted into an antecubitalvein, while blood was withdrawn, only slightlymore quickly, through the other. One was thusable to withdraw heparinized blood withoutanticoagulating the patient. To avoid exsangui-nating the patient, the flow rate had to be veryslow, which in turn introduced a significantdelay to the glucose measurements.

A glucose analyzer on one side of themachine contained the pump to withdraw theblood and measure glucose concentrations,while on the other side was a barely less cum-bersome insulin infusion set-up. In the middlewas a keyboard and what, by today’s standards,was a very primitive computer. The apparatuscould only be used for relatively briefacute procedures such as surgery or labor andthe need to miniaturise the system was obvious.At the time, it was thought that it might bepossible to miniaturise the glucose analyzerand insulin infuser, but it was felt that itwould be the impossibility of miniaturizing acomputer that would preclude the developmentof a portable artificial pancreas!

Nevertheless, when open-loop pumps cameinto widespread use in the early 1980s, it waswidely believed that closed-loop pumps were

Figure 5-3. The Biostator dates from the mid-1970s and constitutedone of the first attempts at a closed loop insulin pump. Glucose levelswere measured on blood continuously withdrawn from an indwellingvenous catheter. A computer with basic algorithms regulated the rate ofdelivery of insulin into a second catheter. Interestingly, it was the computerwhich was felt at the time to be the component which was unlikely ever tobe miniaturized to the point at which a closed loop pump might becomeportable. (With kind permission of Prof Freckmann at the Institute ofDiabetes Technology).


only a matter of a few years away, and indeed thefirst description from Japan of a wearable closed-loop system in man in 198210 reinforced thisbelief. In that paper, the rate limiting step wasperceived to be that of biocompatibility, meaningthat the sensor had to be changed every fewdays. One felt at the time that the Japanesewould soon solve the problem. They have notdone so, however, and instead one now acceptsthat subcutaneous sensors have to be changed, justlike insulin pump cannulae, up to twice a week.

The closed-loop comprises three components:the glucose sensor, the insulin pump, and thecomputer software telling the pump how muchinsulin to give. Unsurprisingly, each of these hasits own problems. The fact that it has taken solong since 1982 vividly attests to these difficulties.

Glucose Sensors

Many attempts have been made to develop a non-invasive way to measure blood glucose. It would, ineffect, allow patients to do finger-prick glucosemeasurements without having to prick their fingers.This would allow much more frequent checks onglucose levels which in turn would make it possibleto monitor not only single levels but also trends.Unfortunately, despite attempts to use many differ-ent physical principles, including near-infraredspectroscopy, changes in tissue refractive errors(and hence in light scattering), fluorescence decaytimes, and many others, none of these methods hasproved as accurate as the invasive methods. Themethods simply have too little specificity andhence have to contend with too great a levelof background noise. The reader is referredto the review in John Pickup’s textbook ifthey wish to go into these methods in moredetail11.

The Glucowatch represented a way to mea-sure glucose outside the body by a supposedlynon-invasive method. In this technique, the gad-get worn on the wrist applied a voltage betweentwo electrodes which resulted in salt, water, andits contained glucose being sucked through theskin by a process of reverse iontophoresis.Unfortunately, it could only be used for a matterof hours at a time, tended to cause a mild burnand stopped working if the patient sweated.Given that sweating is a feature of hypoglycemia,exactly the situation in which one most needs anaccurate glucose measurement, this representeda major drawback and Animas has now with-drawn the product.

The commonest implanted glucose sensorsuse glucose oxidase, which catalyses the reactionbetween glucose and water to form gluconic acidand hydrogen peroxide:

glucoseþ O2 ! gluconic acidþ H2O2

If a voltage is applied to the electrode, theH2O2 is oxidized with the release of twoelectrons:

H2O2 ! O2 þ 2Hþ þ 2e�

In other words, a current is generated which isproportional to the ambient glucose concentration.This is the same reaction as is used on many of thestrips used for finger-prick glucose estimation.

One way to use this technology to give acontinuous glucose sensor is the microdialysistechnique which has been developed commer-cially by Menarini (www.menarini.com) as theGlucoDay. The patient wears a small pumpwhich circulates saline through a loop of micro-dialysis tube implanted under the abdominalskin. As the saline traverses the semipermeablemicrodialysis tube it picks up glucose in propor-tion to ambient tissue concentrations. The glu-cose estimation is performed within themachine itself, ex vivo, and a continuous tracingof tissue glucose levels is recorded.

The main drawbacks are the sheer bulk of theequipment necessary to house the pump and thetwo reservoirs (for the saline and the waste),the fact that one cannot shower and the factthat the tubing has to be changed in the hospitalevery couple of days. After each change, it mayneed many hours to reequilibrate before calibra-tion. The slower the flow through the tubing thebetter the equilibration with interstitial fluid andhence the accuracy, but the longer the delay whichmay be as much as 30 min and which causesproblems with the response time of the appara-tus, especially in the context of closed-loop insu-lin pumps. A newer model is in clinical trials.

The field of sensor technology has in practicebeen dominated by the products originally mar-keted by MiniMed, which became Medtronic in2001. The first commercially available sensorbecame available in 1999 and was called theContinuous Glucose Monitoring System(CGMS), although strictly speaking it performsmeasurements every 10 s which are averagedevery 5 min so that it is only near continuous.A hollow plastic tube is introduced into


subcutaneous fat using a steel needle in the mid-dle as introducer. The glucose oxidase is aroundthe tip of the tube and is connected via a wire tothe ‘monitor’ on the belt, which applies thenecessary voltage to the electrode and recordsthe resultant current (Figure 5-4).

As with all sensors, calibration is a potentialproblem. Ideally one wants to calibrate at timepoints which give a wide range of glucosevalues. At the same time, there is a delaybetween blood and tissue glucose levels ofsome 10–15 min so that one also wants theglucose level to be stable at the time the calibra-tion is done. In practice, these conditions arehard to meet but, clinically, absolute precisionmay be less important than the ability to detecttrends in changes in glucose concentrations.The accuracy may break down at low glucose

levels, when tissue glucose uptake reduces localconcentrations more than blood levels, but inother circumstances tissue glucose is a suffi-ciently accurate guide to blood levels.

The original Minimed CGMS could store dataover a 3-day period, but it then had to be down-loaded through a PC to find out what glucoselevels had been doing over the past 72 h. When itwas first marketed, the hope was that it wouldrevolutionize the management of Type 1 dia-betes by identifying where the problems laywith the insulin doses being used. In practice,consistent patterns in diurnal glucose profilescan usually be identified with old-fashioned fin-ger-pricks and what the sensors usually did wasto demonstrate, if one did not already know it,that no 2 days are the same for the patient withType 1 diabetes and that it is not their fault thatthey cannot achieve perfect control.

The thinking behind the CGMS was not to useit on its own but to develop it to the point that itcould be used to drive an insulin pump. Indeed,there was a certain symmetry to a patient wear-ing a pump on one side of their belt and a CGMSmonitor on the other. However, when Medtronicwent forward to clinical trials, they chose insteadto go for a fully implanted system using theintraperitoneal pump described above, drivenvia a subcutaneous wire by an intravenous sen-sor inserted through the subclavian vein andsuspended in the superior vena cava just abovethe right atrium to give an intravenous/intraper-itoneal (iv/ip) system.

Clinical trials with this system began in 2001and very promising results were reported, parti-cularly from Eric Renard’s group in Montpelier.However, although it proved possible to use thesystem quite successfully to control overnightand inter-prandial glucose levels, the responsetime of the sensor was not quick enough torespond to the rapid changes in insulin require-ments encountered immediately after a meal or,just as importantly, to reduce insulin infusionrates when glucose levels fell. This in turnresulted from the fact that there had been pro-blems with the fragility of the sensors in theturbulence of vena caval blood flow. The sensorhad had to be strengthened and the price paidfor greater rigidity had been a slower responsetime.

This highlighted the whole question of whatone was trying to achieve. As indicated in theintroduction, insulin effectively behaves as twoseparate hormones. When we eat, we ‘tell’ the

Figure 5-4. Minimed’s original Continuous Glucose MonitoringSystem (CGMS). The ‘monitor’ on the belt applies a voltage to thesubcutaneous sensor and collects output current from the electrode. Inthis version, data were downloaded onto a PC at the end of the 3-daymonitoring period and related to insulin doses, meals, activity, etc.retrospectively. Reproduced by kind permission of Medtronic Ltd.


pancreas that we are going to do so and theb-cells are primed in terms of insulin secretionincreasing even before food has been ingested.Furthermore, secretion of incretins, releasedfrom the gut in response to food, changes thealgorithm regulating the amount of insulinsecreted such that more insulin is secreted at agiven glucose level. While it is perfectly possiblewith a system of ‘meal announcement’ to tell aninsulin pump that a meal is imminent and thealgorithms need to change, the goal of the pro-ject was a fully automatic system which did notneed any input from the patient and was nottherefore at risk from human error.

If this was the goal, then the trial was a fail-ure. Even though intraperitoneal insulin deliv-ery has clear advantages, the fully implantedartificial pancreas has, at least for the timebeing, been abandoned in favor of the triedand tested subcutaneous sensor driving anexternal pump giving insulin subcutaneously(an sc/sc system).

Current State of Play

Pumps

It is the 30th anniversary of the first insulinpumps. During this time, much effort has goneinto their development. Although new genera-tions of pump continue to evolve, conventionalpumps have now become sophisticated and reli-able machines.

The major pump manufacturers and theirpumps are (alphabetically):

1. Animashttp://www.animascorp.com/

2. Cozmohttp://www.delteccozmo.com/

3. Minimed Paradigmhttp://www.minimed.com/

4. Roche Accu-Chek (and they continue to mar-ket the Disetronic D-tron Plus, which has theadvantage of pre-filled insulin syringes)http://www.disetronic-usa.com/

The websites all have a pronounced NorthAmerican slant, which is appropriate given thefact that more than 20% of patients with Type 1diabetes are on pumps in the United States com-pared to less than 5% in Britain (with otherEuropean countries somewhere in between).This in turn represents the historical fact that

pumps have only recently been financed in theUnited Kingdom. Pumps currently cost between£2,300 and £2,700 plus roughly £1,000/year forconsumables.

Those interested in future developmentsin insulin pumps may wish also to visit the web-sites of Omnipod, made by Insulet (http://www.myomnipod.com/), which does away withtubing but is not yet approved in Europe, theh-Patch from Valeritas (http://www.valeritas.com/h_patch.shtml), which is primarily aimedat Type 2 diabetes, and the tiny nanopump devel-oped by Debiotech, recently acquired by Animas(http://www.debiotech.com/debiotech. html).

Sensors

In 1982, when the first closed-loop artificial pan-creas paper was published, the limiting factor toits long-term success was perceived to be one ofbiocompatibility. Twenty-five years on, this hasnot been solved but the potential need to replacesensors has been accepted and the focus hasswitched elsewhere.

Recently, real-time sensors have becomeavailable so that the patient can see at aglance what their glucose level is at any onetime and, just as importantly, how fast it ischanging and in what direction. Three compa-nies currently have real-time continuous glu-cose sensors approved by the FDA. They are(alphabetically):

1. Abbott FreeStyle Navigator (Figure 5-5)http://www.continuousmonitor.com/

2. DexCom Seven (so called because it isapproved for use for 7 days)http://www.dexcom.com/html/dexcom_pro-ducts.html

3. Medtronic MiniMed Guardian Real-time(Figure 5-6)http://www.minimed.com/products/insulinpumps/index.html

A detailed table comparing the respective productscan be found athttp://www.childrenwithdiabetes.com/continu-ous.htm.

All do essentially the same thing. They differslightly in the size of the electrode and the fre-quency with which it is suggested they should bechanged. All incorporate a small transmitter anda receiver with a screen and can be used ashypo-/hyper-glycemia alarms. This latter


function is not perfect, however, because if youset the alarm to go off when glucose levels fallbelow, say, 4 mmol/l, the sensor delay may meanthat you are not given enough warning and arealready confused by the time you are being toldto take action. If on the other hand, you set thealarm threshold higher, it is constantly going offat normal glucose levels. The sound of the warn-ing beeps from glucose sensors has alreadybecome commonplace at gatherings of scientistsworking in the field! Nevertheless, the importantpoint is that these sensors give the patient anidea not only of what their blood glucose levels isbut also of whether it is rising or falling.

Although links between sensor manufacturersand pump manufacturers are in the offing, Mini-med is currently the only manufacturer to appearin both lists. In fact, they have recently developedthe Paradigm Real-Time system, in which it is thepump which collects and displays the glucosedata transmitted from the sensor, obviating theneed for a second ‘box’ (Figure 5-6). Although thepump is not driven by the sensor, this representsa step toward a fully integrated artificial pancreas.For those patients not using sensors, they havealso now developed (in collaboration with Bayer)the Contour Link, a glucose meter able to trans-mit finger-prick glucose data to the Paradigmpump for recording and trend analysis.

While every diabetes meeting currently includesa favorable presentation on the benefits of real-

time glucose monitoring, it remains to be seenwhether in due course drawbacks will also becomeapparent. The past year has seen many presenta-tions on the benefits to the patient but for everypatient who can respond appropriately to a givenchange in glucose level there may be at least oneother patient who reacts inappropriately and forwhom a real-time sensor could potentially cause asmany problems as it solves. A clinical trial of out-comes is currently in progress.

There are also, of course, major cost impli-cations. MiniMed’s current UK costs are £2,200for the system plus £375 per 10 sensors.Because it is so expensive, we currently haveapproval to use real-time sensors only on asmall number of patients with particularlyunstable diabetes and frequent hypoglycemiathough the use of one for a few days may alsobe useful in chosen cases to identify unrecog-nized problems.

Once sensors are used to drive pumps, thesafety issues will become even more importantand far more data will be needed on the accuracyand reliability of the sensors. Clearly if a sensortells a pump that the glucose level is 12 mmol/lwhen it is in fact only 6 the consequences for thepatient could be extremely dangerous. Exactlywhat safety measures will need to be taken, orwhether a pump will need more than one sensor(so that if they disagree with each other an alarmis triggered), remains to be seen.

Figure 5-5. Abbott FreeStyle Navi-gator. The small unit on the left isattached to the skin and covers theelectrode itself. The tissue glucose con-centration is transmitted to a hand-held unit which displays not only theactual glucose level but also the factthat it is falling and displays an alarmfor a risk of impending hypoglycemia.Reproduced by kind permission ofAbbott Diabetes Care.


Control Algorithms

While the manufacturers cited above can nowprovide the hardware both for the sensors andfor the pumps, the attention now turns to thesoftware connecting the two and the necessarycontrol algorithms.

Although many systems have been explored,the two which have been most thoroughly stu-died are the Model Predictive Control (MPC), anadaptive system capable of ‘learning’ to someextent, and the Proportional-Integral-Derivative(PID) system which adds together functionsreflecting the current glucose level, how far it isfrom the desired target and how fast it is

changing. Whereas the PID system is essentiallyretrospective, looking back over previous glu-cose levels to come up with a suggested insulininfusion rate, the MPC system is more forwardlooking, basing its recommendations for insulininfusion rates on the patient’s own responses togiven insulin infusion rates in the past. It is saidto have advantages in systems with long delaysand at extremes of glucose levels. A detailedreview of this highly specialized area can befound in Roman Hovorka’s excellent recentreview.12

One of the major issues with feedback loops isthe effect of a delay at any point in the system.Hovorka has estimated that an iv sensor linked toan ip pump has a delay of about 70 min betweenthe glucose sensing and the insulin effect (com-pared to a delay of only 30 min with a nativepancreas) and this rises to almost 100 min withan sc sensor and sc insulin delivery. Delays in afeedback loop tend to cause oscillation and thequestion is whether it will ever be possible toachieve good enough control with a single pro-gramme with such a long delay.

Clinical experience with closed-loop pumpsremains very limited with only about 100 patientdays at this point. Good glycemic control hasbeen shown to be possible with a closed-loop solong as the subject does not eat, because thevariation in the amount of insulin needed ismodest, but problems arise following a mealwhen, if insulin levels do not increase promptlyas soon as food is ingested, an unacceptably largehyperglycemic excursion results. More impor-tantly, if insulin levels do not fall fast enoughwhen glucose levels start to fall, the patient runsa real risk of overshooting into hypoglycemia.

It is theoretically possible to circumvent thisproblem with a hybrid, semi-automatic systeminvolving ‘meal announcement’ whereby thepatient gives the pump a signal that they areabout to eat which changes the computer algo-rithm. This mimics the physiological situationin which the pancreatic b-cells do not operatewith a single control algorithm. When you eat,the K cells in the stomach and upper small bowelsecrete GIP (glucose sensitive insulinotropicpolypeptide), while the L-cells further down thegut secrete GLP-1 (glucagon-like peptide, Type 1).These incretins ‘tell’ the b-cells that you have eatenand change the algorithm so that, at a given glu-cose level, the b-cells secrete more insulin. In fact,even before you eat, neural signals lead to an

Figure 5-6. Medtronic MiniMed Paradigm Real-time monitorwith MiniLink. The sensor has been inserted into the subcutaneoustissue in the right iliac fossa. This patient is also wearing a paradigminsulin pump infusing insulin continuously via the tubing and a cannulainto the subcutaneous tissue of the left iliac fossa. Tissue glucose levels,transmitted from the sensor via the MiniLink, are collected in the pumpwhich displays not only the actual tissue glucose levels but also the rateof change. At this stage, insulin delivery rates cannot be regulated bysensor glucose data. Reproduced by kind permission of Medtronic Ltd.


increase in insulin secretion before blood glu-cose levels begin to rise.

The idea of making an artificial pancreaswhich is more independent than the native pan-creas and needs only a single algorithm seemssomewhat ambitious. With hybrid systems, itwill be necessary to persuade the regulatoryauthorities that an artificial pancreas which isnot fully automatic and which relies on humaninput will still prove to be safe. The Yale grouphas successfully used a system whereby there isonly one algorithm but the patient gives a pre-meal priming dose of insulin 15 min before themeal to raise insulin levels more quickly thanwould otherwise occur.13

Although the control algorithms are thepiece of the puzzle that there is least experienceof, one feels quite positive because seriousmathematicians are now putting a lot ofthought and effort into the design of appropri-ate software. One important new advance isthat, in contrast to the old physiological ‘graybox’ mathematical modeling, the FDA has justgiven approval for an in silico metabolic simu-lator which makes it possible to compare dif-ferent control algorithms without the need foranimal trials.

Closed Loop Pumps and the JDRFArtificial Pancreas (AP) Project

In 1982, with open-loop pumps already availableand the publication of the first closed-looppaper, one felt that closed-loop pumps mightonly be 5 years away. Yet it took until 1999before satisfactory glucose sensors becameavailable. In 2001, when Minimed began trialsof their fully implanted closed loop system, oneagain felt that closed-loop pumps were only afew years away. These studies were able toachieve good control, at least so long as thepatients did not eat, but were abandoned dueto problems with the sensor.

Against this background, the announce-ment in late 2005 by the Juvenile DiabetesResearch Foundation (JDRF), whose missionstatement is a cure for Type 1 diabetes, ofmajor support for its Artificial Pancreas Pro-ject is a major cause for optimism. Not onlywill this ensure long-term funding and supportfor the project but, just as importantly, since itis now a charity rather than a pharmaceutical

company making the running, it will alsoprovide great help to clear the legal and regu-latory hurdles which an artificial pancreasnecessarily entails.

Conclusion

The fact that we have reached this stage owes alot to Mr. Al Mann, who founded Minimed in1980 and who personally financed the project formany years at a time when other companieswere giving up. As we reach the 30th anniversaryof insulin pumps, we have several manufac-turers making portable, sophisticated, and reli-able pumps.

Glucose sensors have proved much moredifficult and it took all of the 1980s and1990s until the first glucose sensor wasapproved. Even now, there remain problemsboth with their accuracy and, more particu-larly, with their reliability. The early attemptat a trial for a closed-loop pump in 2001 usedan intravenous sensor (and intra-peritonealinsulin delivery) but reliability problemshave meant that intravenous sensors are nolonger available. The delay inherent in aclosed-loop system which uses subcutaneoussensors and subcutaneous insulin deliverycurrently being developed for an artificialpancreas introduces additional problemsbecause it introduces such a long delay intothe feedback loop. If such a system is to work,it seems likely that it will have to be semi-automatic. That is to say, the pump is auto-matic through the night and between mealsbut relies on patient intervention to deal withmeal boluses. Such an approach is lookingreally very promising.13

By the time a safe and successful artificialpancreas becomes a reality, countless millionswill have been spent on the project and it will behard to recoup these development costs. Onehopes that companies will not even try to doso. The recent setting up by the JDRF of its AP(Artificial Pancreas) project has revitalized thewhole field. The hope is that within the next 5–10years the artificial pancreas will provide thesolution which gives all patients with Type 1diabetes the prospect of near-normal glucosecontrol and hence the best possible quality oflife and a greatly reduced risk of diabeticcomplications.


References

1. Kerr D, James J, Nicholls H. Technologies as therapeuticdevices. What do we expect from users of insulin pumptherapy? Infusystems Int. 2008;7:1–4.

2. Pickup JC, Keen H, Parsons JA, Alberti KGMM: Contin-uous subcutaneous insulin infusion: an approach toachieving normoglycaemia. BMJ. 1978;i:204–207.

3. Juice: The Creative Fuel that drives Today’s World ClassInventors. Evan I Schwartz. Harvard Business SchoolPress; 2004.

4. Sonksen PH, Judd SL, Lowy C. Home monitoring ofblood-glucose. Method for improving diabetic control.Lancet. 1978;i:729–32.

5. Pickup J, Mattock M, Kerry S: Glycaemic control withcontinuous subcutaneous insulin infusion comparedwith intensive insulin injections in patients with Type 1diabetes: meta-analysis of randomised controlled trials.BMJ 324:705, 2002.

6. The Diabetes Control and Complications Trial ResearchGroup. The Effect of Intensive Treatment of Diabeteson the Development and Progression of Long-Term

Complications in Insulin-Dependent Diabetes MellitusN Engl J Med. 1993;329(14):977–86.

7. Bode BW, Stubbs HA, Schwartz S, Block JE. Glycemiccharacteristics in continuously monitored patientswith Tye 1 and Typ2 diabetes. Diabetes Care. 2005;28:2361–2366.

8. Albisser AM, Leibel BS, Ewart TG et al. An artificialendocrine pancreas. Diabetes. 1974;23:389–96.

9. Pfeiffer E-F, Kerner W. The artificial endocrine pancreas:its impact on the pathophysiology and treatment ofdiabetes mellitus. Diabetes Care. 1981;4:11–26.

10. Shichiri M, Kawamori R, Yamasaki Y, Hakui N, Abe H.Wearable artificial endocrine pancreas with needle-typeglucose sensor. Lancet. 1982;2:1129–1131.

11. Textbook of Diabetes, ed. Pickup JC and Willams G.Chapter 34. Third edition, Blackwell, 2003.

12. Hovorka R. Continuous glucose monitoring and closed-loop systems. Diabet Med.. 2006;23:1–12.

13. Weinzimer SA, Steil GM, Swan KL, Dziura J, Kurtz N,Tamborlane WV. Fully Automated Closed-Loop InsulinDelivery Versus Semiautomated Hybrid Control inPediatric Patients With Type 1 Diabetes Using an Artifi-cial Pancreas. Diabet Care. 2008;31:934–939.


6From Basic Wound Healing to Modern Skin Engineering

L.C. Andersson, H.C. Nettelblad and G. Kratz

Introduction

The skin is our largest organ and it simulta-neously fulfills many functions required ineveryday life. The most important function ofthe skin is to act as a barrier between the bodyand the external environment. Many immunecells are also present in the skin to provideprotection against invading micro-organisms.The skin is in addition an important participantin the physiological water homeostasis and it isone of the main ways the body regulates itstemperature. Normal skin has sensation withprotective value; a physical tensile capacityallowing for body mobility and it provides theexternal contour as well as the texture and thecolor of the exterior.

Over the millennia mankind has been fasci-nated by the body’s ability to heal itself.Throughout history the main goal has alwaysbeen to cover skin defects and to close woundsin order to protect the human body. Skin graft-ing and flap surgical procedures have been usedin order to facilitate wound healing for thou-sands of years. Due to the advances in criticalcare and resuscitation, patients who, in the past,would have died in the acute phase are nowsurviving. Consequently there is a much greaterneed for high-quality skin substitutes. The clin-ical demand has driven newer technologies,building upon principles learned using cadaverand autografts, to the creation of engineeredskin substitutes using living allograft cells aswell as the combining of technologies to createcomposites – the most advanced products and atpresent the closest products to living skin. Anautologous split thickness skin graft currentlystill remains superior to all commercially

available skin substitutes with regards to itsunique qualities. There are, however, situationswhere we prefer to use non-autologous substi-tutes, either because the patient may be toofragile for the added surgical trauma of harvest-ing split-skin grafts or due to the fact that thereis not enough skin which could be harvested, forexample, post-burn.

Structure of the Skin

The skin consists of two principal layers: theepidermis and the dermis. The epidermis con-sists of a keratinized layer with no vital cellsand cell layers of keratinocytes which producekeratin. The epidermis provides protectionagainst micro-organisms and loss of fluids.The epidermis within the basal cell layer con-tains melanocytes that produce melanin, ourskin pigmentation. The dermis mainly consistsof fibroblasts and interstitial connective tissue.The fibroblasts produce collagen and elastine,which provide the tensile strength of the skincomposite. The dermis also harbors hair follicles,which are lined by epidermal cells and are loca-lized adjacent to sebaceous glands, which pro-duce sebum. There are also sweat glands, bloodvessels, and nerves in the dermis (Figure 6-1).The dermis provides the tensile strength andelasticity that allows mobility of the skin.1

Historical Overview

Our ancestors struggled to treat wounds anddefects inflicted either by nature or by acts ofman. Skin substitutes in the form of xenografts


93

were first used to provide wound coverage as farback as 1500 BC (frog skin). In the 1600s, lizardskin was used in western cultures and this pro-gressed to the use of mammalian skins as well ascadaver skin during the twentieth century,including dog and rabbit skin and the still-usedpigskin products.2–4 Facial deformities such asdefects after the amputation of the nose were thetype of problems that received the most atten-tion. A famed practitioner, Sushruta (circa 600BC), described operations for reconstruction ofthe nose without any form of anesthesia. Tissuefrom the forehead was dissected but remainedattached to the area between the eyebrows andwas then turned down in order to reconstruct anose and left intact for 3–4 weeks. The nasalbridge was then divided and the nose wasimproved in its shape. It was further describedthat the donor defect in the forehead healedrather well and left very little deformity.5 Duringthe Roman Empire, Celsus (25 BC–50 AD) raisedsmaller flaps which involved the skin and the fat(pieces of skin and fat, with blood circulationstill attached to the body in at least one area) inorder to reconstruct skin defects in facial areas.6

Over the years, many others have contributed tonew techniques and variations involving differ-ent types of flaps in order to correct defects with,but it was not until the beginning of the

nineteenth century that the concept of skingrafting (auto grafting: a skin piece from onepart of the body was cut loose with no attach-ment and no circulation, then grafted onto awell-vascularized wound on the same indivi-dual) became clinically used. Sir Astley Cooperremoved skin from an amputated thumb andused it to cover the stump defect with as a full-thickness skin graft in 1817.7 Nevertheless, skingrafting was not fully recognized and acceptedfor clinical use until the last quarter of the nine-teenth century. In 1869, Reverdin reported thatthe healing of granulating wounds was improvedby so-called ‘‘epidermic grafts’’.8 In 1874,Thiersch advocated the use of larger sheets ofepidermal grafts to cover wounds with and alsoemphasized the importance of an epidermalcomponent in the graft with a small amount ofdermis (Figure 6-2). The grafts used were thinsplit thickness skin grafts containing both theepidermis and the parts of the dermis and thesepieces of skin were tangentially excised andgrafted onto wounds and skin defects. Theactual donor sites were then left to heal sponta-neously through epithelialization from thedepth and from the sides of the donor defect,which was itself superficial enough and the heal-ing would be achieved in approximately 2weeks.9 Xenografts gave way to homografts in

Figure 6-1. The skin in cross section.


the form of allograft (cadaveric graft) and auto-graft. The modern development of intensive careand improved surgical techniques enabled skinand fat flaps to be developed further into tissueflaps involving, e.g., skin, fat, tendon, muscle,and bone tissue. The tissue could be harvested inone part of the body and transferred to anotherbody region in order to cover wounds with. Theblood supply after having been disconnected atthe original site was re-established via microsur-gery. Many other developments in modern med-icine also contributed to the overall need forlarger wound areas to be covered with skin, e.g.,burns and trauma10 Our increased understandingof immunology and wound healing has also beena contributing factor of great importance in mod-ern wound management. The clinical demand ofskin has driven the bio-engineering of skin, inthe form of skin substitutes and skin cultivation,which will be integrated, more and more intoclinical practice in the future.11

Wound Healing

The body will immediately respond with anacute inflammation if the skin is acutelydamaged. Many physiological substances are

released at this stage in order to initiate thenormal wound-healing process. Granulation tis-sue of fibroblasts and new capillaries will growinto the wound. Collagen will be produced toprovide stability and strength and an epithelia-lization from the wound edges will start. Thereare many factors that contribute to a normalwound-healing process and many others,which disturb this important process. An acutewound may not progress in the normal wound-healing process and instead become chronic.The understanding of wound healing constantlyincreases and many new treatments which facil-itate wound healing are continuously beingdeveloped.9

Epidermal Wound Healing

The epidermal cells divide and migrate ifdamage to the epidermis has occurred. Awound no deeper than after the harvest of asplit-skin graft (Figure 6-3) will heal sponta-neously (epithelialization) if no infection arises,usually within 2–3 weeks with the correct type ofwound dressing which protects from externalbacteria, viruses, and particles. Damage to theepidermis does not normally result in

Figure 6-2. The skin in cross section - indicating the depth of a Thiersch graft.

From Basic Wound Healing to Modern Skin Engineering 95

contractive scarring but some pigmentation var-iation may be seen as a long-term result. Thehealing of the epidermis will proceed both fromthe wound edges as well as from the underlyingtissues since there are keratinocytes which linethe hair follicles. It can take several years for thebasal cell membrane to remodel and for themelanocytes to migrate into the damaged area.9

Dermal Wound Healing

The healing of a wound, which extends deepdown into the dermis, at the same depth aswhen a full-thickness skin graft is harvested,will take a lot longer than the healing of anepidermal superficial wound (Figure 6-4).There will be no epithelialization from the dee-per tissues due to lack of keratinocytes in theseplanes, but healing takes place solely from thewound edges. Before the epithelialization fromthe wound edges can take place, a layer of gran-ulation tissue consisting of fibroblasts and capil-laries will form in order to reinstitute the dermis.The actual epithelialization from the woundedges will then spontaneously start on top ofthe granulation tissue. The wound should

decrease by 10–15% of the original size perweek in order to represent normal wound heal-ing.12 The actual healing time of deep dermal ordeeper tissue wounds will vary, all depending onthe size of the wound and external interferingfactors, e.g., foreign bodies, micro-organisms,and the patients’ general condition. Large der-mal wounds may need surgical intervention withdebridements and skin grafting in order to heal.The dermal fibroblasts will produce collagenand scar tissue will form. The scar tissue in thedermis will over time be remodeled, but thescarred skin will contract and become less flex-ible and smoother than normal skin is. Thedermal tensile strength may also be reduced inscarred skin. The skin mobility in scarred skinafter dermal healing will be reduced and skinscar contractions may have to be surgically‘‘released’’ or corrected in order to improve thefunction of a specific area of the body. Thescarred skin after damage to the dermis alsoappears different in pigmentation than an indi-vidual’s normal skin. The whole maturing pro-cess of a dermal scar can take several years andsurgical interventions may be necessary duringparts of this process in order to improve apatient’s function and appearance.9 These

Figure 6-3. The skin in cross section -indicating the depth of a split-thickness graft.


types of clinical problems can especially be seenin burn injuries to the hands, due to the largeflexibility and mobility of the hand. It is ofteninitially necessary to debride and excise deeperhand-burns and to cover the healthy bleedingwounds with superficial skin grafts in order toavoid secondary infections and contractive scar-ring with reduced hand function. Further cor-rective surgery may still be required later, butthe early excision of the dead tissue is veryimportant as it significantly enhances thechances of good hand function (Figure 6-5).13

Clinical Management of Wounds

Modern wound-healing treatments consist ofdebridement, negative pressure wound therapy,topical growth factors, the use of bio-engineeredtissue, reconstructive wound-closure techniques(skin grafts, local flaps, pedicle flaps and micro-surgical free flaps), and hyperbaric oxygen. It isimportant to have at least one of the physicianson the multidisciplinary wound team familiarwith modern wound-care techniques. Debride-ment is the basis of all wound-healing strategies.Debriding a wound is defined as removing

necrotic tissue, foreign material, and bacteriafrom an acute or chronic wound, all of whichinhibits the wound healing.12 An acute woundhas yet to progress through the sequentialstages, a chronic wound has become ‘‘stuck’’ inone of the wound-healing stages. The advent ofnegative pressure wound therapy has reducedthe number of non-healing wounds. The nega-tive pressure device can provide rapid formationof granulation tissues and its ability to decreaseedema helps prepare the wound for simplewound-closure techniques.14

Pre-debridement Assessment

The goal of treating any type of wound is tocreate an environment that is conductive to nor-mal and timely healing. The process begins withthe identification of a correct diagnosis of thewounds’ etiology and continues with optimizingthe patient’s medical condition, including bloodflow to the wound site. A proper wound assess-ment before debridement is essential, the originand age of the wound should be determined, thepatient’s general condition (nutrition, smoking,local circulation, etc.) must also be taken into

Figure 6-4. The skin in cross section indicating the depth of a full-thickness graft.


account and a careful medical history should beobtained to identify any possible wound-healinginhibiting medical conditions. It is also clearlyimportant to understand the extent of a wound,specifically with regards to wound depth and theinvolvement of deeper tissue, e.g., tendons andbone. An underlying osteomyelitis could feed awound with bacteria if undiscovered.15 Magneticresonance imaging (MRI) is today the gold stan-dard way to diagnose osteomyelitis. An infectedtendon and fascia can, in the same way, be afocus of infection. The wound-healing processand the take of a skin graft may be disturbed ifthe bacterial concentration on a wound is higherthan 105 per gram of tissue. A further increase ofthe bacterial load could spread from a localsuperficial infection to a systemic one whichleads to septicemia. The application of topicalantibiotics such as silver sulfadiazine, mafenideacetate, and silver nitrate all help to lower thebacterial count and reduce the risk of sepsis andcan these be used before wound debridement.16,17

Debridement

Different debridement techniques include sur-gery, topical agents, and bio-surgery. The mostimportant surgical step in treating any wound isto perform adequate debridement to remove allforeign material and unhealthy or non-viabletissue until the wound edges and base consistonly of normal, soft, well-vascularized healthytissue.12 Only an atraumatic surgical technique(sharp dissection, skin hooks, bipolar cautery,etc.) should be used in order to avoid damagingthe underlying healthy tissue.18 This underlyinghealthy tissue will become the basis for thewound-healing progression. A chronic woundhas to be converted by debridement to an acutewound so that it can then proceed through thenormal healing phases. Frequent debridementsremove the inhibitors of wound healing such asmetalloproteases, including the collagenasematrix metalloproteinase 1 and 8 and elastaseand it allows the growth factors to function more

Figure 6-5. Deeply burnt hand (left) with exposed tendons before debridment and six months later after excision and skin grafting with meshedsplit-skin grafts.


effectively.19 An immediate debridement isnecessary when a necrotizing fasciatis or anascending cellulitis occurs in conjunction witha wound. No other treatment could achieve bac-terial control in such a wound status. Biologicaldebriding agents such as maggots are an effec-tive alternative to surgical debridement inpatients who cannot go to the operating theatrefor medical reasons. The maggots secreteenzymes that dissolve necrotic tissue and thebiofilm that surrounds bacteria. This forms anutrient-rich liquid that larvae can feed on.They are placed on wounds and covered with asemipermeable dressing. Debridement by mag-gots is painless but the sensate patient can feelthe larvae moving. Importantly maggots help tosterilize wounds because they consume all bac-teria regardless of their resistance to antibioticsincluding methicillin-resistant Staphylococcusaureus (MRSA).20–23

Post-debridement Treatment

After debridement a clear, well-vascularizedwound should be kept in the optimal environ-ment for healing (moist, clean, and vascular-ized) in order to enable the use of growth factors.Moist healing has been shown to be far morerapid than healing under an eschar or in otherdry conditions. In this environment, the woundbase can support and promote successful

collagen deposition, angiogenesis, epithelializa-tion, and wound contracture; the result shouldbe the formation of healthy red granulation tis-sue with neoepithelialization at the borders.Epithelializing wounds are characterized by apink neoepithelium that usually creeps in fromthe edges.12 A well-vascularized wound should,after debridement, heal by secondary intentionor accept a skin graft.24–26 Other surgicalwound-closure techniques may also be required,e.g., flaps, all depending on the wound’s locationand extent. Exposed bone should if possible becovered with vascularized tissue, which is trans-ferred into the wound in conjunction with thedebridement. Preferably should a muscle–fat–skin flap be used in order to heal the wounddefect, as well as to help the bone tissue ‘‘fight’’a threatening osteomyelitis (Figure 6-6).

Dressings of Wounds

There is no single dressing suitable for all typesof wounds and often a number of different dres-sing types will be needed throughout the healingprocess. Dressings containing silver havereturned in advanced wound care and arefound used in conjunction with many pro-ducts.27,28 Silver ions kill a broad spectrum ofbacteria including methicillin-resistant Staphy-lococcus aureus (MRSA), Vancomycin-resistantEnterococcus, and Pseudomonas aeruginosa.12

Figure 6-6. Chronic wound with osteomyelitis and exposed bone (sternum) after heart surgery (left). Debridment and wound closure with apedicled musculo-cutaneus rectus abdominis flap (right).


Modern Skin Engineering

Tissue engineering is an interdisciplinary scien-tific field that combines the principles of lifescience and engineering toward the develop-ment of biologic substitutes that will serve torestore, maintain, or improve tissue function.29

First, specific cells must be harvested, isolated,and expanded in tissue culture. The secondcomponent of tissue engineering focuses on thescaffold for tissue structure. Collaboration withbiomedical engineers is critical for the develop-ment of novel scaffolds that will have the optimalcombination of immunological compatibility,sufficient mechanical strength, and biodegrad-ability (Figure 6-7). Third, the implanted con-struct should be incorporated into the healingtissue helping to restore structure and func-tion.10 Various tissues have been engineeredbut the most developed are skin substitutesand a number of these produces have US/FDAapproval and are currently available.2 The needfor skin grafting has been the driving forcebehind the development of these skin substi-tutes.11 The skin is the largest single organ ofthe human body and although composed of onlytwo specialized tissue layers, it remains a recon-structive challenge in many cases when compro-mised. The philosophy of replacing like with likehas contributed to the development of the treat-ment strategies used today. The simplest treat-ment is secondary closure, which is when awound is left to heal spontaneously. This doesnot always lead to a good functional and estheticresult and more complex tissue reconstructionmay be necessary. Advanced type of tissue

reconstruction in the form of distant-free tissuetransfer can lead to superior functional results,but sets large demands on the patient and thesurgical team. A large wound area due to its sizemay not be possible to treat with the patients’own tissue. A layer of tissue, which is mechani-cally similar to the original skin over a recon-struction site, has to be used in order to com-plete the reconstruction and to avoid failure.This has made bioengineered skin substitutes asolution that has received much attention lately.Skin substitutes are a heterogeneous group ofsubstances that aid in the temporary or perma-nent closure of many types of wounds. Depend-ing on wound and product characteristics, dif-ferent skin substitutes may be chosen. Artificialskin substitutes, products of tissue engineering,consist of microengineered, biocompatible,polymer matrix in combination with cellular,and/or extracellular elements such as collagen.30

An ideal skin substitute should possess the phy-sical characteristics and function of normal skinwhile healing takes place. However, no perfect orideal skin substitute yet exists.2

Characteristics and Clinical Useof Skin Substitutes

Skin substitutes are a heterogeneous class oftherapeutic devices that vary in their biologyand application. Although there is no singleperfect skin substitute, certain characteristicscan be considered when evaluating alternatives.A long shelf-life and easy storage makes theproduct readily available. The substitute shouldbe easy to prepare and apply without intensivetraining. Flexibility of thickness allows the pro-duct to be tailored to every type of wound. Theproduct should be able to withstand a hypoxicwound bed and have a degree of resistance toinfection in order to allow relatively ischemictissues to be candidates for application. Theideal skin substitute should have resistance totensile forces and provide permanent and long-term wound stability. It should reproduce bothcomponents of the skin (epidermis and dermis)and provide no antigenicity that could compro-mise the graft or host or present difficulties withfuture applications.2 Because no single productmeets all these criteria, each patient caserequires careful evaluation before choosing theappropriate treatment. Although many acuteand chronic wounds may benefit from a tailoredFigure 6-7. Preparation of a bio-engineered skin substitute.


multidisciplinary approach that utilizes one ormore of the products mentioned, each patientshould be evaluated for other possible therapiesbefore the use of skin substitutes. Adequateassessment of patient-related factors, surgicaldebridement, and infection control will firsthave to be performed. Different techniquessuch as local or regional flaps or microvascular(free tissue) transplantation should be consid-ered and the incorporation of a skin substitutecan be included in the patient’s treatment plan.If a patient is not considered a candidate forwound or defect reconstruction, creative appli-cations of skin substitute technologies may notonly significantly benefit a patient, but also bethe only option for wound closure.2

Types of Skin Substitutes

Xenografts are tissues transplanted from onespecies onto another species, used as a tempor-ary graft (e.g., frog skin and lizard skin). Porcineproducts are the most commonly used xeno-grafts today.3,4 Permacol and OASIS are amongthe currently available products.31,32

Allografts are grafts transplanted betweengenetically non-identical individuals of the samespecies. Most human skin substitute allograftscome from cadaveric sources. Allografts fall intothree categories: epithelial/epidermal, dermal,and composite (epidermal and dermal). Withinthese three categories, they may be acellular, cel-lular/living, or cellular/nonliving. AlloDerm is acommercial available acellular dermal allograftthat was initially developed for skin defects, butit has also been used on burns and soft tissuereplacement.33–35 This product retains dermalelements and the basement membrane allowedkeratinocytes to migrate into the material. A sub-sequent split-thickness graft or cultured kerati-nocyte graft is added after neovascularization ofthe neodermis (Figure 6-8) for epidermal cover.Clinical use of acellularized human cadaveric der-mis and ultra-thin skin grafts has shown goodclinical results in face, hand, and foot burns.33–35

Graftjacket, Neoform, and DermaMatrix arefurther acellular dermal allografts available.36,37

ICX-SKN, TransCyte (also known as Dermagraft-TC) and Dermagraft are cellular dermal allograftswhich use a scaffold of dermal collagen that isseeded with neonatal fibroblasts to stimulate cellswithin the host’s wound to promote healing.38–44

Composite allograft products are the mostadvanced and closest products to living skinthat are currently commercially available. Apli-graft and Orcelare are the currently available pro-ducts. Apiligraf has been used in the treatment ofepidermolysis bullosa (EB). In a study of ninepatients with 96 sites of skin loss, 90–100% heal-ing was observed by 5–7 days with clinically nor-mal appearing skin in place by days 10–14.45,46

OrCel, a composite bilayer produce with neonatalkeratinocytes impregnated onto a coated non-porous sponge composed of type I bovine col-lagen has also been used for wound coverage aftercontracture release in EB patients.47,48 The FDAhas approved its use for reconstruction or treat-ment of recessive dystrophic EB of the hands andskin graft donor sites in these patients. Studiesevaluating its use in chronic venous and diabeticlower extremity ulcers are ongoing.

Autografts are tissues grafted to a new posi-tion on the same individual. They are commonlydivided into three main categories:2

(1) Split-thickness skin grafts (STSGs) whichcontain the epidermis and a variable thick-ness of the upper layers of dermis, leavingthe remaining layers of dermis in place toheal by secondary epithelialization from thewound edges and keratinocytes within theadnexa of the deeper dermis.

(2) Full-thickness skin grafts (FTSGs) which con-tain the epidermis and the entire dermis.49

These types of grafts are preferred in areas

Figure 6-8. Histology from biopsy showing a neodermis withingrowth of capillaries, collagen, and fibroblasts after use of a cellfree dermal substitute.


where significant scarring or contracture ofthe grafts would provide harmful estheticor functional consequences. Because of thelimited supply of FTSG donor sites they areusually reserved for reconstructing woundsof the head, neck, hands, and genitals.

(3) Cultured autologous skin substitutes whichare frequently referred to as cultured epider-mal autografts (CEAs). This nomenclatureincludes epidermal grafts and excludes der-mal/epidermal grafts. The CEAs are graftedonto a wound and the healing of the newepidermis is initiated (Figure 6-9). Epicel andLaserskin are among the currently availableproducts.50,51 Cultured skin substitute (CASS)is a CEA with the addition of a cultured auto-logous dermal layer, making it a more anato-mically correct skin substitute. This product isstill in clinical trials but it does in theoryrepresent the most advanced autologous skinsubstitute available. The product is created byculturing autologous fibroblasts and keratino-cytes with collagen and glycosaminoglycansubstrates.52,53

A synthetic monolayer substitute Suprathel is amonolayer acellular synthetic dressing which hasproven to decrease pain when used on donorsites.54,55

Synthetic biolayer substitutes are acellularproducts engineered without allogenic cells,they function as dermal templates and promotein growth of host tissues to repair defects orcreate a neodermis. After in growth, skin graftingcan be performed on top of the new dermis. Theyalso contain a removable silicone epidermal layer

to help protect the wound from moisture loss andcontamination. Commercially available productsare Biobrane and Integra. Biobrane is a biosyn-thetic skin substitute consisting of a bilaminatemembrane of nylon mesh bonded to a thin layerof silicone, which is, coated with porcine type 1collagen-derived peptides (dermal analogue).The silicone layer functions as temporary epider-mis.56–58 Integra Bilayer Matrix Wound Dressingis a synthetic bilayer acellular skin substitutecomposed of an outer silastic sheet (epidermalanalogue) with a matrix composed of bovine col-lagen and glycosaminoglycan (dermal analo-gue).59–67 The silastic sheet provides temporarycover before skin grafting is performed.

What Does the Future Hold in ModernSkin Engineering?

The wound-healing treatments with debride-ment remain very important. Infection controlis vital for any healing. None of the available skinsubstitutes has been shown to be superior toautologous split-skin grafting (Figure 6-10).Existing substitutes do also have problemswhich include graft take, infection, immuneand allergic reactions, and the need for a secondprocedure. We can today cultivate keratino-cytes, fibroblasts as well as melanocytes invitro and the cultivated cells are transplantedonto patients in different ways and in combina-tion with skin substitutes, in order to facilitatewound healing and to provide patients with bet-ter esthetic and functional results (Figures 6-11,6-12). The skin cannot regenerate itself and it

Figure 6-9. Spraying of cultivated cells onto dermal skin.Figure 6-10. Late result of skin-grafted right hand compared to non-injured left hand.


Figure 6-11. Cultivated Fibroblasts (left) and Keratinocytes (right) in vitro.

Figure 6-12. Vitiligo, before and after grafting of cultivated melanocytes.


does heal with scarring which sometimes con-tracts severely. The future involves furtherunderstanding of the wound-healing mechan-isms as well as further discoveries of the differ-ent functions of the different skin cells. Thediscovery of plasticity of cells (cells ability totransdifferentiate) may also influence the devel-opment of new types of skin substitute productswith no antigenicity.68–73 The ‘‘key’’ to the sti-mulation of regeneration of the skin couldreduce/eliminate skin scarring and provide uswith better protection against micro-organismswhich continue to threaten our existence.

References

1. Rudolph R, Ballantyne DL. Skin grafts. In: McCarthy J(ed). Plastic Surgery: Vol 1 General Principles. Philadel-phia, PA: WB Saunders, 1990:221–25.

2. Shores JT, Gabriel A, Gupta S. Skin substitutes andalternatives: a review. Adv Skin Wound Care. September2007; Vol. 20 No. 9. PP. 493–508.

3. Piccolo N, Piccolo-Lobo M, and Piccolo-Daher M. Use offrogskin as a temporary biological dressing, Proc AmBurn Assoc. 1992;24.

4. Chiu T, Burd A. ‘‘Xenograft’’ dressing in the treatment ofburns. Clin Dermatol. 2005;23:419–23.

5. Bhishagratna K. The Sushruta Samhita: An EnglishTranslation Based on Original Texts. Cosmo Publica-tions 2006.

6. Spencer WG. Celsus De Medicina, with English transla-tion. Vol. 3. Cambridge, MA: Harward University press.1938.

7. Balch CM, Marzoni FA. Skin transplantation during thepre-Reverdin era, 1804–1869. Surg Gynecol Obstet.1977;144:766.

8. Reverdin JL. Greffes epidermiques; experience faite dansle service de M. le docteur Guyon, a l’Hopital Necker,pendant 1869. Bull Soc. Imperiale Chir. Paris Series 2,Vol.10, published in 1870.

9. Peacock EE, Cohen IK. Wound healing. In: McCarthy J(ed). Plastic Surgery: Vol 1 General Principles. Philadel-phia, PA: WB Saunders, 1990;161–178.

10. Chong AKS, Chang J. Tissue Engineering for the HandSurgeon. A Clinical Perspective. J Hand Surg. March2006;31A:3.

11. Mooney DJ, Mikos AG. Growing new organs. Sci Am.1999;280:60–65.

12. Attinger CE, Janis JE, Steinberg J, Schwartz J, Al-Attar A,Couch K. Clinic approach to Wounds: Debridement andWound Bed Preparation Including the Use of Dressingsand Wound-Healing Adjuvants. Plastic and Reconstruc-tive Surgery. Clinical Approach to Wounds. June Supple-ment 2006;117:7S.

13. Zellweger G. Die Behandlung Der Verbrennungen.1985;118–121.

14. Defranzo AJ, Argenta LC, Marks M, et al. The use ofvacuum-assisted closure therapy for the treatment oflower-extremity wound with exposed bone. PlastReconstr Surg. 2001;108:1184.

15. Grayson ML, Gibbons GW, Balogh K, et al. Probing tobone in infected pedal ulcers: A clinical sign of osteomy-litis in diabetic patients. JAMA. 1995;273:721.

16. Fox CL. Silver sulfadiazine, a new topical therapy forPseudomonas in burns. Arch Surg. 1968;96:184.

17. Kucan JO, Robson MC, Heggers JP, et al. Comparison ofsilver sulfadiazine, povidone-iodine and physiologicalsaline in the treatment of chronic pressure ulcers. J AmGeriatr Soc. 1981;29:232.

18. Edgerton MT. The Art of Surgical Technique, Baltimore:Williams & Wilkins; 1988.

19. Trentgrove NJ, Stacey MC, Macauley S, et al. Analysis of theacute and chronic wound environments: The role of pro-teases and their inhibitors. Wound Repair Regen. 1999;7:442.

20. Courtenay M, Church JC, Ryan TJ. Larva therapy inwound management. J R Soc Med. 2000;93:394.

21. Sherman RA, Hall MJ, Thomas S. Medicinal maggots: Anancient remedy for some contemporary afflictions. AnnRev Entomol. 2000;45:55.

22. Wollina U, Karte K, Herold C. Biosurgery in woundhealing: The renaissance of maggot therapy. J Eur AcadDermatol Venerol. 2000;14:285.

23. Sherman RA, Sherman J, Gilead L, et al. Maggot therapyin outpatients. Arch Phys Med Rehab. 2001;81:1226.

24. Krizek TJ, Robson MC. The evolution of quantitative bac-teriology in wound management. Am J Surg. 1975;130:579.

25. Robson MC, Heggers JP. Delayed wound closure basedon bacterial counts. J Surg Oncol. 1970;2:379.

26. Robson MC, Heggers JP. Bacterial quantification of openwounds. Mil Med. 1969;134:19.

27. Tredget EE, Shankowsky HA, Goeneveld A, et al. Amatched-pair, randomised study evaluating the efficacyand safety of Acticoat silver coated dressing for treat-ment of burn wounds. J Burn Care Rehabil. 1998;19:531.

28. Yin HQ, Langford R, Burrell RE. Comparative evaluationof the antimicrobial activity of Acticoat antimicrobialdressing. J Burn Care Rehabil. 1999;20:195.

29. Langer R, Vacanti JP. Tissue engineering. Science.1993;260:920–926.

30. Enoch S, Grey JE, Harding KG. ABC of wound healing.Recent advances and emerging treatments. BMJ. 22 April2006;332:962–965.

31. Harper C. Permacol: clinical experience with a new bio-material. Hosp Med. 2001;62:90–5.

32. Hodde JP, Ernst DM, Hiles MC. An investigation of thelong-term bioactivity of endogenous growth factor inOASIS Wound Matrix. J Wound Care. 2005;14:23–5.

33. Wainwright DJ. Use of an acellular allograft dermalmatrix (Alloderm) in the management of full-thicknessburns. Burns. 1995;21:243–8.

34. Munster AM, Smith-Meek M, Shalom A. Acellular allo-graft dermal matrix: immediate or delayed epidermalcoverage? Burns. 2001;27:150–3.

35. Sheridan R, Choucair R, Donelan M, Lydon M, Petras L,Tomkins R. Acellular allodermis in burns surgery: 1-yearresults of a pilot trial. J Burn Care Rehabil. 1998;19:528–30.

36. Brigido SA, Boc SF, Lopez RC. Effective managementof major lower extremity wounds using an acellularregenerative tissue matrix: a pilot study. Orthopedics.2004;27(Suppl):s145–9.

37. Schoepf C. Allograft safety: efficacy of the tutoplast pro-cess. Implants- Int Magazine Oral Implant. 2006;7.

38. Lukish JR, Eichelberger MR, Newman KD, et al. The use ofa bioactive skin substitute decreases length of stay forpediatric burn patients. J Pediatr Surg. 2001;36:1118–21.


39. Demling RH, Desanti L. Management of partial thicknessfacial burns (comparison of topical antibiotics and bio-engineered skin substitutes) Burns. 1999;25:256–61.

40. Mansbridge J, Liu K, Patch R, Symons K, Pinney E.Three-dimensional fibroblast culture implant for thetreatment of diabetic foot ulcers: metabolic activity andtherapeutic range. Tissue Eng. 1998;4:403–14.

41. Harding KG, Moore K, Phillips TJ. Wound chronicityand fibroblast senescence-implications for treatment.Int Wound J. 2005;2:364–8.

42. Boyd M, Flasza M, Johnson PA, Roberts JS, Kemp P.Integration and Persistence of an investigationalhuman living skin equivalent (1CX-SKN) in human sur-gical wounds. Regen Med. 2007;2:369–76.

43. Bello YM, Falabella AF, Eaglstein WH. Tissue-engi-neered skin. Current status in wound healing. Am JClin Dermatol. 2001;2:305–13.

44. Purdue GF, Hunt JL, Still JM Jr. et al. A mulitcenterclinical trial of a biosynthetic skin replacement, Derma-graft-TC, compared with cryopreserved human cadaverskin for temporary coverage of excised burn wounds. JBurn Care Rehabil. 1997;18(1 Pt 1):52–7.

45. Fivenson DP, Scherschun L, Cohen LV. Apligraft in thetreatment of severe mitten deformity associated withrecessive dystrophic epidermolysis bullosa. Plast ReconstrSurg. 2003;112:584–8.

46. Fivenson DP, Scherschun L, Choucair M, Kukuruga D,Young J, Shwayder T. Graftskin therapy in epidermolysisBullosa. J Am Acad Dermatol. 2003;48:886–92.

47. Still J, Glat P, Silverstein P, Griswold J, Mozingo D. Theuse of a collagen sponge/living cell composite material totreat donor sites in burn patients. Burns. 2003;29:837–41.

48. Sibbald RG, Zuker R, Coutts P, Coelho S, Williamson D,Queen D. Using a dermal skin substitute in the treatmentof chronic wounds secondary to recessive dystrophicepidermolysis bullosa: a case series. Ostomy WoundManage. 2005;51(11):22–46.

49. Wright KA, Nadire KB, Busto P, Tubo R, McPherson JM,Wentworth BM, Alternative delivery of keratinocytesusing a polyurethane membrane and the implicationsfor its use in the treatment of full-thickness burn injury.Burns. 1998;24:7–17.

50. Carsin H, Ainaud P, Le Bever H, et al. Cultured epithelialautografts in extensive burn coverage of severely trau-matized patients: a five year single- centre experiencewith 30 patients. Burns. 2000;26:379–87.

51. Lobmann R, Pittasch D, Muhlen I, Lehnert H. Autolo-gous human keratinocytes cultured on membranes com-posed of benzyl ester of hyaluronic acid for grafting innonhealing diabetic foot lesions: a pilot study. J DiabetesComplications. 2003;17:199–204.

52. Harriger MD, Warden GD, Greenhalgh DG, Kagan RJ,Boyce ST. Pigmentation and microanatomy of skinregenerated from composite grafts of cultured cells andbiopolymers applied to full-thickness burn wounds.Transplantation. 1995;59:702–7.

53. Boyce ST, Goretsky MJ, Greenhalgh DG, Kagan RJ, Rie-man MT, Warden GD. Comparative assessment of cul-tured skin substitutes and native skin autograft for treat-ment of full-thickness burns. Ann Surg. 1995;222:743–52.

54. Schwarze H, Kuntscher M, Uhlig C, et al. Suprathel, anew skin substitute, in the management of donorsites of split-thickness skin grafts: Results of a clin-ical study. Burns. 2007; E-pub:doi: 10.1016/j.burns.2006. 10.393.

55. Banes AJ, Compton DW, Bomhoeft J, et al. Biologic,biosynthetic, and synthetic dressings a temporarywound covers: a biochemical comparison. J Burn CareRehabil. 1986;7:96–104.

56. McHugh TP, Robson MC, Heggars JP, Phillips LG, SmithDJ Jr, McCollum MC. Therapeutic efficacy of Biobrane inpartial- and full-thickness thermal injury. Surgery.1986;100:661–4.

57. Feldman DL, Rogers A, Karpinski RH. A prospective trialcomparing Biobrane DuoDERM and Xeroform for skingraft donor sites. Surg Gynecol Obstet. 1991;173:1–5.

58. Prasad JK, Feller I, Thomson PD. A Prospective con-trolled trial of Biobrane versus scarlet red on skin graftdonor areas. J Burn Care Rehabil. 1987;8:384–6.

59. Grzesiak JJ, Pierschbacher MD, Amodeo MF, MalaneyTI, Glass JR. Enhancement of cell interactions with col-lagen/gylcosaminoglycan matrices by RGD derivatiza-tion. Biomaterials. 1997;18:1625–32.

60. Stern R, McPherson M, Longaker MT.Histologic study ofartifical skin used in the treatment of full-thicknessthermal injury. J Burn Care Rehabil. 1990;11:7–13.

61. Komorowska-Timek E, Gabriel A, Bennett DC, et al.Artificial dermis as an alternative for coverage of com-plex scalp defects following excision of malignanttumours. Plast Reconstr Surg. 2005;115:1010–7.

62. Machens HG, Berger AC, Mailaender P, Bioartificialskin. Cells Tissue Organs. 2000;167:88–94.

63. Orgill DP, Straus FH 2nd, Lee RC. The use of collagen-GAG membranes in reconstructive surgery. Ann NYAcad Sci. 1999;888:233–48.

64. Fitton AR, Drew P, Dickson WA. The use of a bilaminateartificial skin substitute (Integra) in acute resurfacingof burns: an early experience. Br J Plast Surg. 2001;54:208–12.

65. Klein MB, Engrav LH, Holmes JH, et al. Management offacial burns with a collagen/glycosaminoglycan skin sub-stitute-prospective experience with 12 consecutive patientswith large, deep facial burns. Burns. 2005;31: 257–61.

66. Groos N, Guillot M, Zilliox R, Braye FM. Use of anartificial dermis (integra) for the reconstruction ofextensive burn scars in children. About 22 grafts. Eur JPaediatr Surg. 2005;15:187–92.

67. Jeschke MG, Rose C, Angele P, Fuchtmeier B, Nerlich MN,Bolder U. Development of new reconstructive techniques:use of Integra in combination with fibrin glue and nega-tive-pressure therapy for reconstruction of acute andchronic wounds. Plast Reconstr Surg. 2004;113:525–30.

68. Brazelton TR, Rossi FM., Kesher, GI, Blau, HM. Frommarrow to brain: expression of neuronal phenotypes inadult mice. Science. 2000;290:1775–1779.

69. Mezey E, Chandross KJ, Harta G, Maki RA, McKercherSR. Turning blood into brain: cells bearing neuronalantigens generated in vivo from bone marrow. Science.2000;290:1779–1782.

71. Orlic D, Kajstura J, Chimenti S, Jakoniuk I, AndersonSM, Li B, et al. Bone marrow cells regenerate infractedmyocardium. Nature. 2001;410:701–705.

72. Sigurjonsson OE, Perreault MC, Egeland T, Glover JC.Adult human hematopoietic stem cells produce neuronsefficiently in the regenerating chicken embryo spinalcord. Proc Natl Acad Sci USA. 2005;102:5227–5232.

73. Wurmser AE, Nakashima, K, Summers RG, Toni N, D,Amour KA, Lie DC, et al. Cell fusion-independent differ-entiation of neural stem cells to the endothelial lineage.Nature. 2004;430:350–356.


7Artificial Sphincters

Austin Obichere and Ibnauf Suliman

Introduction

Sphincters in broad term refer to circularmuscles within the gastrointestinal or urogen-ital tract that regulate the release of bowelcontents or urine to the outside. Incontinence(fecal or urinary) is due to inability to controlstool or urine expulsion as a result of disease,accident, or injury to the anal and sphincterurethrae muscles, respectively. The concept ofartificial sphincter use in surgery is not newand was first conceived over 50 years ago torestore continence either by surgically fash-ioning/transposing muscle or by implantingprosthetic materials to achieve normal bowelor urinary action.

This chapter will review the origin and evolu-tion of artificial sphincters in colorectal surgery.Its role in the clinical management of end-stagefecal incontinence and as an alternative toabdominal wall colostomies will be discussed,and promising future strategies for restoringnormal anal defecation examined. A detailedaccount of the artificial urinary sphincter is out-side the scope of this chapter and will be dis-cussed elsewhere.

Origin of Artificial Sphincters

The emergence of artificial sphincters in sur-gery followed attempts to either surgically fash-ion or implant prosthesis to achieve urinary orbowel control and avoid diverting urinarytubes/catheters and abdominal wall stomas.The evolution of the artificial bowel sphinc-ter arguably takes it origin from the work of‘M. Pillore’, a French surgeon who successfully

performed a caecostomy for complete bowelobstruction caused by rectal carcinoma in1776. Unfortunately, the patient died 20 daysafter the operation and an autopsy revealedgangrene of the small bowel due to two poundsof mercury he had ingested, believing thiswould overcome his obstruction. However, thebirth of a left iliac colostomy was conceivednearly two decades later in 1793 by Duret – ina desperate attempt to relieve intestinal obs-truction in an infant 3 days old with imperfo-rate anus. He had initially performed lumbarcolostomy upon the dead body of a baby2 weeks old, but abandoned this approach ashe feared leakage of meconium into the perito-neum. We are told that the operation was asuccess and that the infant lived to a ripe oldage of 45 years. Other notable achievementsof this remarkable pioneer surgeon includeanchoring the mesocolon with suture to pre-vent stoma retraction. He recognized theoccurrence of stomal prolapse and initiateda crude form of continence through bowelcleansing of stool via the ‘artificial anus’ usingplain water containing two drops of syrup ofrhubarb.

The contribution of these pioneer surgeonsfirmly established the colostomy as a success-ful treatment strategy for managing malignantor end-stage benign colorectal disorders thathas not needed to change for over 200 years.However, despite much progress in surgicaltechnique, stoma design, and management,fecal incontinence and perceived abnormalbody image remain major pitfalls impairingpatient quality of life and has resulted in arelentless quest for better fecal effluent con-trol with surgically fashioned or prostheticbowel sphincters.


107

Surgically Fashioned Bowel Sphincters

In selected patients, construction of an artificialor neo-anal sphincter can be achieved by trans-posing preserved fascia or skeletal muscle inclose proximity to the anal canal. The resultingneo-sphincter constitutes an acceptable alterna-tive to a permanent colostomy, thus avoiding thepsychological issues relating to perceived abnor-mal body image of a stoma. Such a neo-sphinctermay be described as passive, dynamic, orphysiological.

Passive Neo-sphincter

Harvey Stone conceived the idea of using pre-served fascia as a subcutaneous purse stringsuture that was drawn up snugly around theanus for treating anal incontinence of operativeor traumatic origin. Several years later in 1929,he modified this procedure, anchoring the twoloops of fascia to the gluteus maximus muscle onopposite sides so that the anus could be tigh-tened by contraction of the muscle.1 It is ofinterest that much earlier, at the turn of the20th century, Chetwood described a techniquewhereby the gluteus maximus muscle was trans-posed for sphincter reconstruction whileChittenden in 1930 promoted another varietyof the passive neo-sphincter utilizing bilateralgluteus maximus flaps that were sutured ante-rior to the rectum thus suspending it in a ham-mock of muscle.2,3

The gluteus maximus muscle gained furtherpopularity in reconstructive surgery for analincontinence when Bistrom (1944) described amethod of detaching a part of the origin of themuscle and fashioning an opening throughwhich the rectum was pulled through to createa sphincter mechanism.4 It soon became evidentthat the direct pull of the parallel muscle bundleseven at normal resting tone provided somedegree of continence that could be reinforcedvoluntarily and that the normal physiologicallength tension relationship was re-established.Furthermore, the muscle is well innervated andvascularized, which permits transfer of a seg-ment with little risk of developing ischemia.

More recently, construction of an anal sphinc-ter mechanism using either the origin or theinsertion of the gluteus maximus muscle hasbeen reported.5,6 The former being a modificationof the original technique described by Bistrom,

involved creation of bilateral muscular slings thatwere split to encircle the rectum in order to sta-bilize each other at a proper resting tension andfiber length. In the case of the latter, muscle wasdetached from its insertion on the linear aspera ofthe femur to create a double scissor-like dia-phragm around the rectum which was thenanchored to the ischial tuberosities. However,despite reported claims of improved anal conti-nence by proponents of these two methods, theyhave failed to gain widespread acceptance andhave so far been restricted to isolated cases.

Utilization of the gracilis muscle to create acontinent neo-anal sphincter in four childrenwith end-stage anal incontinence was describedby Pickrell in 1952 and emerged as the favoredoperation because the muscle was not bulkyand did not impair certain motor functionssuch as climbing stairs. In his original article,Pickrell described transposition of a singlegracilis muscle in a clockwise fashion 3608surrounding the rectum and fixed to the con-tra-lateral ischial tuberosity with suture.7

Modifications of this technique now include abilateral muscle wrap which adopts either oneof two configurations; a complete circumferen-tial wrap using both muscles or the creation of asling behind the rectum with a single gracilismuscle reminiscent of the puborectalis, whilethe other was rotated in a clockwise fashion toencircle the rectum.

These operations were designed primarily forthe management of end-stage fecal incontinencedue to defective anal sphincter function asopposed to attempts in restoring anal conti-nence after abdomino-perineal excision of therectum. However, passive transposition of a sin-gle gracilis after surgical ablation of the rectumhas also been reported, whereby an opening wascreated in the muscle through which the colonwas pulled through and anchored to the peri-neum. The transposed gracilis was secured toadjacent deep perineal muscles, the presacralaponeurosis, and the contralateral ischial tuber-osity.8 In one series, use of adductor longusmuscle to create a neo-anal sphincter after rectalexcision has also been described.9

Indications

Passive artificial muscle sphincters were ori-ginally designed for the treatment of fecalincontinence, but have recently been added


to the armamentarium at the disposal of colo-proctologist where total anorectal reconstruc-tion may be required for the followingconditions;

• End-stage fecal incontinence

• Abdomino-perineal excision of rectum

• Congenital anorectal disorder (Imperforateanus/Spina bifida)

• Sphincter injury/trauma.

Types of Passive Neo-sphincters

A variety of skeletal muscles have so far beensuccessfully transposed to create a passive neo-anus. The common denominator in all cases isthat the muscle utilized in the creation of apassive artificial anus must carry its own neuro-vascular supply and lie in close proximity to theanal canal. Skeletal muscles harboring thesecharacteristics are as follows;

• Gracilis (unilateral or bilateral)

• Gluteus maximus (unilateral or bilateral)

• Adductor longus (unilateral)

• Satorius (unilateral)

• Rectus femoris (unilateral)

Complications

A major drawback of passive anal sphincterreconstructive techniques is that fecal conti-nence is short-lived because striated skeletalmuscle fatigue easily and cannot sustain pro-longed voluntary contraction. This led toattempts by others to discover new techniquesand strategies designed to create fatigue resis-tant-striated skeletal muscles and resulted inelectrically stimulated neo-sphincters (fatigueresistant muscles).10,11

Reports of muscle necrosis of the harvestedneo-sphincter have largely been attributed topoor muscle quality or surgical technique. Thegracilis which is by far the most commonlyused muscle, receives its neurovascular supplyin the upper third of the muscle thereby allow-ing transposition around its proximal pedicle.Ischemic injury therefore occurs as a result ofdirect trauma to the neurovascular bundle dur-ing muscle harvesting or indirectly, due toexcessive pedicle tension from re-routing,peri-operative hematoma formation, or tissueedema.

Infection remains the main harbinger offailure of passive neo-sphincters requiring revi-sional surgery where initial conservative treat-ment with antibiotics with or without woundlavage/debridement fails. A variety of organismsboth gut-derived and skin microbes have beenidentified but colononization by methicillin-resistant Staphylococcus aureus (MRSA) meansthat failure is near irreversible. Sometimes dueto overwhelming sepsis such as necrotizingfascitis and associated muscle necrosis, sal-vage is not possible and neo-sphincter re-construction using the contra-lateral muscleshould be considered and re-scheduled to a laterdate.

Poor-quality muscle either inherent or due toatrophy from fatigue will result in poor outcomeand incontinence. Most patients thereforerequire self medication with constipating agentsto attain acceptable continence followed by theassistance of self-administered enema to evacu-ate stool. Understandably, achieving the rightbalance can be difficult and tiresome for manypatients.

Dynamic/Electrically StimulatedNeo-sphincters

It was discovered in animal experiments and inhumans that repeated electrical stimulation ofskeletal muscle from external sources resulted inconversion from type 2 (easily fatigued) to type 1(fatigue-resistant) muscle fibers.12,13 This adaptiveresponse to a generated electrical pulse allowssustained voluntary contraction of the neo-analsphincter, enhancing the fecal continence achievedby a passive muscle wrap.

A significant advance in this technique waspioneered by Cor Baeten and Norman Williamswho independently developed different types ofimplantable device to induce this vital changein muscle property, thereby improving neo-sphincter function and making this sort of sur-gery a practical reality.10,11 Baeten developedan intra-muscular (albeit perineural) mode ofmuscle stimulation, whereas Williams directlystimulated the nerve trunk to the gracilis. Theformer technique has proved more robust inmost surgeon’s hands as it avoids problemswith electrode displacement which seems com-moner when the nerve trunk is stimulateddirectly (Figure 7-1).

Artificial Sphincters 109

Indications

The electrically stimulated neo-sphincter hasbeen shown to be more effective than the passivemuscle wraps in terms of continence control andquality of life. It is more likely to produce higherresting anal pressures than the non unstimu-lated neo-sphincter, and hence improve conti-nence. Despite sharing similar indications withits predecessor (see above), dynamic neo-sphincters have gained popularity particularlyamong highly motivated young adults withsphincter injury from trauma or accident, whoare determined to avoid an abdominal wallstoma.

Types of Dynamic Neo-sphincters

The two skeletal muscles most frequently uti-lized in the construction of surgically fashio-ned neo-sphincters are the gracilis and gluteusmaximus. The former is generally preferredbecause it is less bulky, easier to transpose,and does not impair motor function like walk-ing up stairs. On the other hand, dissection andseparation of the gluteus into two parts withcareful preservation of the inferior gluteal neu-rovascular supply to the distal half, is more chal-lenging and has deterred widespread use of thismuscle.

Five different configurations of gracilo-plasty (alpha, gamma, epsilon, split sling,and double wrap) have been described andin one study that compared the first threetypes, it emerged that the alpha wrap gener-ates lower neo-anal pressures than gammaand epsilon.14

Dynamic Neo-sphincter FollowingAbdomino-perineal Excision of Rectum

The last decade has seen attempts by surgeons toconstruct a dynamic neo-sphincter followingsurgical excision of the rectum for carcinoma,with some degree of success.12,14–17 All authorsadopt chronic low-frequency electrical stimula-tion for muscle conversion even though thetiming of its onset and duration vary amonginvestigators. Most experts favor constructionof the neo-sphincter with double gracilis musclewrap at the same time as perineal stoma forma-tion, usually without a defunctioning stoma pre-sumably because it allows early stimulation ofthe transposed muscle.12,16,18 But Williams andcolleagues employed a three-staged approachthat included perineal colostomy formationwith vascular delay of the distal gracilis muscle,creation of a stimulated graciloplasty, and clo-sure of the defunctioning stoma.15

The gracilis vascular delay procedure,whereby the distal perforating vessels to themuscle are ligated, allows new channels to formfrom the proximal major vessel and was thoughtto reduce the risk of distal muscle ischemia aftertransposition.19 However, the emergence ofdirect anatomical and physiological evidencefrom experimental studies showing only onearterial system within the gracilis muscle ques-tions previous practice involving an additionaloperation for vascular delay, implying that thisapproach may now be redundant.12

The main indication for rectal excision andneo-sphincter reconstruction in these circum-stances has been in the treatment of distal rectalcancer. Initial concerns regarding the risk ofjeopardizing the possible cure of the patient inreturn for avoiding a colostomy have so far beenrefuted by some studies which have shown thatit is an oncologically safe operation. One studyreviewed 20 patients who had abdomino-perinealreconstruction with a double dynamic gracilo-plasty after rectal excision for low rectal cancerfound no local recurrence, but revealed fourpatients who developed distant metastasis.12 Inanother report of 12 patients who had electri-cally stimulated graciloplasty after abdomino-perineal excision of the rectum, 10 had rectaladenocarcinoma of which 6 were Dukes stageA, 3 were B, and only 1 was stage C.14,20 Therewas no local recurrence at a median follow-upof 54 months (range, 3–79), but pulmonary

gracilismuscle(oldposition)

gracilismuscle(newposition)

pulse generator

magnet

leads

Figure 7-1. Electrically stimulated neo-sphincter.


metastases developed in the patient with DukesC rectal cancer. Furthermore, in a personal ser-ies of 81 patients over a 10-year period of which37 surviving patients were followed for a meanperiod of 79 months, the study concluded thatrestorative perineal graciloplasty did not reducethe effectiveness of abdomino-perineal resectionin the cure of cancer.17

Functional Outcome of DynamicGraciloplasty

Total anorectal reconstruction with electricallystimulated graciloplasty after abdomino-perinealexcision of the rectum may be primary (recon-struction of the neo-anus in the same operationas the abdomino-perineal resection) or second-ary (neo-sphincter reconstruction several yearsafter a Miles resection). The former method isby far the most frequently performed, perhapsdue to the anticipated increased morbidity withthe latter. In seven patients who had asecondary procedure a mean of 8.5 years aftera Miles resection, it was found that althoughcontinence was regained in five patients, theysuffered numerous complications, additionaloperations, long hospital stay, and subsequentre-admissions.21

The reported functional results followingreconstruction of a neo-anorectum with gracilo-plasty (stimulated or unstimulated) after excisionof the rectum for cancer, vary considerably inthe literature. One series evaluated 26 out of31 patients for continence at a mean follow-upof 40 months, revealing continence to liquid andsolid stool in 22 patients or 85%.18 Mercati andcolleagues found that all seven patients who hadunstimulated graciloplasties in their series weresatisfied with the level of continence achievedwith the neo-anus, although they acknowledgedthat these patients suffered a slight-to-moderatefecal incontinence causing a persistently wet anusthat required the use of pads.16 In another study,‘successful’ functional outcome was reported in 8of 15 patients (53%) who were continent to solidand liquid stool, but continence in these circum-stances included occasional fecal soiling.12

Contrary to the above findings which supporta favorable functional outcome, others havereported episodes of incontinence to solidstool, persistent soiling due to liquid stool neces-sitating wearing of sanitary pads at all times, anddifficulty with stool evacuation requiring the aid

of regular enemas.14 This disparity in reportedcontinence may be attributed to differing tech-niques employed by investigators as exemplifiedby Williams and colleagues, who favored astaged approach with a single gracilis musclewrap, while others constructed the neo-sphincterwith a double gracilis wrap at the same time asformation of a perineal colostomy.12,15,16,18 Inaddition, there is no standardized or validatedtool available to measure functional outcomeafter anorectal reconstruction with the dynamicgraciloplasty, thereby encouraging comparisonof studies in this area.

It is of interest that amalgamation of datafrom the literature would appear to suggestthat the double gracilis neo-anus gives betterfunctional results compared to the single musclewrap. However, the contrasting methodsemployed in these studies in determining thepatients’ level of continence with a neo-sphinc-ter has been overlooked, and obvious failure intaking account of the effect of observer bias.Moreover, one group recently reviewed morbid-ity and functional outcome in 15 patients with amean follow-up of 28 months who had under-gone double dynamic graciloplasty after abdo-mino-perineal resection. They found that of 12patients available for assessment, five out of sixcases without neo-sphincter stenosis, had ‘goodcontinence’, indicating that the double dynamicgraciloplasty was associated with a high rate ofstenosis which compromised functional out-come.22 The reason why this is so has beenattributed to the forces generated by asymme-trical contraction of both gracilis muscles. Giventhese conflicting reports in the literature, it isclear that there are areas of significant variationamongst experts pertaining to graciloplasty con-figuration, timing of surgery, and implantationof the electrical device and function. Unques-tionably, it would seem that the true functionaloutcome of these procedures remain uncertain.

Complications

The dynamic neo-sphincter operation is techni-cally demanding has a long-learning curve andcarries considerable morbidity. It has thereforebeen recommended this type of surgery be con-fined to specialist colorectal centers. Complica-tions may also be due to the technical challengesencountered during harvesting and creation ofthe neo-anus or device related.


As with the passive neo-sphincter, musclenecrosis from direct and indirect causes ofischemia along with reports of infection arealso encountered following construction of thedynamic neo-sphincter. Neo-sphincter stenosisis a problem that was described recently and isassociated with the double gracilis wrap thatappears to generate asymmetrical pressureforces around the neo-anus.22,23

The device-related complications are numer-ous and range from mechanical failure which isfrequently the result of complete battery dis-charge to the local effects of the device itself.Localized pain or discomfort, device or leadmigration, and erosion of skin lead disconnec-tion or fracture are some of the well-documentedproblems of the device. Repeated operations andultimately explantation of the device are some-times necessary to address some of these draw-backs contributing to further overall morbidityin these patients.

Inability to evacuate stool is commoner fol-lowing neo-anal reconstruction with dynamicgraciloplasty than with the passive neo-sphincter,presumably because the former has been shownto generate higher neo-anal pressures on mano-metry. Nonetheless, it is a problem that leavesthe patient entirely reliant on enemas to achi-eve stool expulsion despite terminating elec-trical discharges from the pulse generator tothe neo-sphincter by interrupting the establi-shed electrical circuit with a magnet. Given thehigh morbidity associated with neo-sphincterreconstruction, all patients are likely to benefitfrom a centralized service with the necessaryexpertise on site, along with a plan that incorpo-rates pre-operative counseling and a selectiveapproach.

Physiological Neo-sphincter

The passive and dynamic neo-sphincters offerlittle or no body image disturbance unlike anabdominal wall stoma. They also provideadded benefit in terms of continence controlvia the neo-anus, albeit an effective biological‘thirsch’ wire. Unfortunately, these patients relyon regular enemas or irrigation of the neo-anusfor the evacuation of ‘formed’ stool and areincontinent to liquid fecal discharge. Further-more, the loss of normal reservoir capacityprovided by the excised rectum and lack of asensory component either through disease or

surgery means that these types of neo-sphincterswill always fall short of normal physiologicalfunction.24

Recent reports from animal experimentsdescribe a new generation of neo-sphincterreconstruction capable of voluntary action fol-lowing cross-innervation of the nerve supplyingthe transplanted muscle and the pudendal nerve.Sato and colleagues created a neo-sphincter withthe biceps femoris muscle (equivalent of thehuman gracilis muscle) in 22 dogs after rectalextirpation and performed anastomosis betweenthe nerve to the muscle and the pudendalnerve.25 Following nerve regeneration, theydemonstrated successful muscle conversionfrom type 2 to type 1 fibers with apparent satis-factory defecatory function. Similarly, Congi-liosi and others rendered three groups of10 dogs incontinent and compared passive,dynamic, and a physiological neo-anal wrapwith pudendal nerve anastomosis.26 Theyobserved that the physiological group produceda functional anal sphincter superior to the othertwo methods. More recently one novel studyshowed experimentally that it was possible toreconstruct a functional external anal sphincterusing a free latissmus dorsi flap with pudendalneuromicrovascular anastomoses in nine mon-grel dogs rendered fecally incontinent.27 Whatthese studies seem to imply is that the pudendalnerve potentially has a crucial role in attempts torestore native innervation and voluntary analsphincter function.

Preliminary studies in human cadavers haveshown that it is possible to create a physiologicalneo-sphincter using the gluteus maximusmuscle – inferior half of the muscle supplied bythe inferior gluteal neurovascular bundle.28

Another group also demonstrated in humancadavers that it was possible to technicallyreconstruct a physiological neo-sphincter usingthe gracils muscle with pudendal nerve anasto-mosis.29 The first clinical case of the ‘voluntary’neo-sphincter after abdomino-perineal resec-tion of the rectum has shown encouragingearly results for a technique which may signal anew more efficient alternative to the dynamicneo-sphincter28 (Figure 7-2). Long-term data in10 out of a series of 19 patients from a singlecenter over an 8-year period indicate that volun-tary defecation was achieved in all 10 patientswithout the need for enemas or irrigation. Inaddition, better continence scores and qualityof life were demonstrated in this group when


compared to 27 historical controls who hadabdomino-perineal excision of the rectumand end colostomy formation.30 Sato and hiscolleagues proposed that the resulting restora-tion of sensitivity after pudendal nerve ana-stomosis, allows recognition of the need todefecate and might explain the promisingresults so far reported following constructionof the physiological neo-anus.

Indications

Physiological neo-sphincters share similar indi-cations to passive and dynamic types and incommon with the other two rely on skeletalmuscle that can be transposed along with theirown neuro-vascular bundle. The main contra-indication, therefore, would include patientswith impaired nerve or muscular function fromneuromuscular disorders. So far, all 19 patientswho have undergone physiological neo-sphinc-ter reconstruction have been diagnosed with lowrectal cancer treated by abdomino-perineal exci-sion of the rectum.30 It is foreseeable that thisapproach may play a role in the surgical man-agement of end-stage fecal incontinence due toother benign conditions provided that furtherdata confirm the efficacy of this promisingtechnique.

Types

A single center experience utilizing the lowerhalf of the gluteus maximus muscle with itsown inferior gluteal neuro-vascular bundleaccount for all the patients in the literaturewho have had anal continence restored by phy-siological neo-sphincter reconstruction. How-ever, use of the gracilis muscle and free latissmusdorsi flaps with pudendal neuromicrovascularanastomoses hold future promise.27,29 It isexpected that other skeletal muscles with theirown neurovascular supply may be adopted inthe future or perhaps preferred to the gluteusmuscle so long as more studies concur that thetechnique is feasible and is a superior alternativeto other types of neo-sphincters.

Complications

With the exception of device-related complica-tions, the physiological neo-sphincters presentsimilar problems encountered with the dynamicneo-anus.

Muscle ischemia and necrosis resulting frominfection, intrinsic skeletal muscle characteris-tics, or poor harvesting techniques are othercomplications that occur without exception inall neo-sphincters. However, the physiological

Figure 7-2. Physiological neo-sphincterillustrating pudendal and inferior glu-teal nerve anastomosis and gluteusmuscle wrap around the colon.


neo-sphincter whose function is dependent onpudendal nerve growth to re-connect with themuscle, presents a potentially unique problem ofdisuse muscle atrophy. This phenomenon com-monly occurs during the period required foradequate nerve regeneration. Given that the esti-mated peripheral nerve regeneration rate is1 mm/day, recovery of muscle function mayfail or be delayed, depending on the distancebetween the site of nerve anastomosis and themuscle end-plate. To overcome this potentialdrawback of muscle atrophy, it has been sug-gested that direct electrical stimulation of themuscle during the time required for nervegrowth may reverse this process.30 The evidencesupporting the prevention of disuse muscleatrophy in the physiological neo-sphincter hasbeen extrapolated from data demonstratingcontinued neo-sphincter muscle activity by pro-grammed electrical stimulation using an exter-nal device.

Experimental Organ Transplantation

The quest for anal continence after abdomino-perineal excision of the rectum has led a numberof surgeons to create a neo-anus reconstructedfrom muscle, or even an anal artificial sphinc-ter.31 However, clinical results have not beenaltogether satisfactory, partly because thesemethods were intended to deal with the strengthof the anal sphincter alone, reminiscent of abiological Thiersch.

Anal continence is not just the result of con-tractile activity in the anal sphincter, but alsodepends on good sensation, reasonable capacity,and a colon and rectum that are neither tooactive nor too inert. Furthermore, the ability tobe continent and the ability to evacuate are tosome extent mutually antagonistic. This meansthat many of the above surgical attempts torestore continence have been bedeviled by defe-catory problems because the patient has no ideawhen the rectum is full and because contractilerectal activity and deactivation of whateverdevice is being used to control continence arenot coordinated.

Anorectal transplantation should have thepotential to solve a number of the above pro-blems. Parts designed for a specialized purposewould seem on first principles to have a betterchance of restoring best function compared totransposed muscle or implanted plastics, and

the pudendal nerve, being a mixed motor andsensory nerve, should be able to restore sensa-tion, motor function, and perhaps reflexes.

What has been most exciting about noveltechniques employing pudendal nerve anasto-mosis is the prospect of restoring a sensorycomponent. Many of the clinical problems ofthe dynamic neo-sphincters relate to the lack ofknowledge of the need to defecate through lossof a normal sensory pathway. Anorectal trans-plantation with pudendal nerve anastomosistherefore holds out the prospect of potentiallynormal anorectal function. But it is beset by itsown difficulties, not least the apparent extremetechnical difficulty involved surgically, added towhich are the issues of returning physiologicalfunction and immunological considerations.

Evolution of Organ Transplantation:Essential or for Quality of Life?

Organ transplantation has traditionally beenaimed at restoring life to patients who wouldotherwise face imminent death from fatal dis-ease or failure of a vital organ. This concept wasreinforced by transplantation of the heart, lungs,liver – and to a lesser extent kidneys and smallintestine, where imminent demise from organfailure might be avoided with long-term dialysisand total parenteral nutrition, respectively.There is no doubt that organ transplantationnot only saves lives but might also provide therecipient with a new lease of life and freedom toperform normal daily activities with improvedquality of life.

The last two decades have seen patient qualityof well-being increasingly recognized as animportant outcome measure of surgical practiceand it is not in the least surprising why someinvestigators have now been prepared to con-sider transplantation of a non-essential organprimarily to improve quality of life – as exem-plified by recent clinical transplantation of thelarynx, womb, hand, and face.32–35

Permanent colostomies following anorectalamputation for cancer can be devastating forsome people who would rather contemplate sui-cide than accept life passing bowel contentsthrough an artificial opening on their abdominalwall.36–38 It remains debatable whether currentalternative surgical techniques such as the peri-neal stoma and neo-sphincters, which avoid acolostomy, might reasonably be considered a


sufficient answer because of persisting defeca-tory problems which impair quality of life. Thephysiological neo-sphincters using pudendalnerve anastomosis may be an advance becausethe pudendal nerve might not only permit reflexsphincter function but also allow the patient tosense the need to defecate.

It is envisaged that anorectal transplanta-tion is a possible strategy that would replacethe anorectal structural unit after abdomino-perineal resection, if it were technically andbiologically feasible. The resulting preserva-tion of body image and potential return ofnormal anal defecation might improve patientquality of life far beyond currently availablemethods, thereby reinforcing calls that trans-plantation of a non-essential organ can beperformed for quality of life reasons alone.39

Assessing the Feasibility of AnorectalTransplantation

Anatomy and Function of the PudendalNerve. The pudendal nerve is a mixed nervecarrying sensory, motor, and autonomic fibersfrom three roots derived from the second, third,and fourth (S2–S4) ventral rami of the sacralplexus. Normally the three roots unite to formtwo cords, which unify to create a main trunk,commencing at the upper border of the ischialspine and carrying sensory fibers to genitalia,muscular branches to the sphincter urethrae,perineum, and external anal sphincter. Thepudendal nerve trunk passes back between thepiriformis and the coccygeus muscle and curlsaround the sacrospinous ligament before run-ning forward to lie within the pudendal canal(Alcock’s) on the inferior lateral wall of theischiorectal fossa.40

The pudendal nerve has three main branches:inferior rectal, which constitutes the motor sup-ply to the external anal sphincter and controlsvoluntary sphincter action; dorsal nerve of thepenis/clitoris, which innervates the genitals; andthe common perineal nerve, giving muscularbranches to the perineal muscles and sphincterurethrae. The inferior rectal nerve branchemerges from the posterior end of Alcock’scanal, before division of the pudendal trunkinto its terminal branches (dorsal nerve ofpenis/clitoris and common perineal) whichoccurs within the canal itself 41 (Figure 7-3).

Like all anatomical structures, the pudendalnerve is not exempt from anomalies involvingany part of its entire course. A thorough under-standing of variations of the main nerve trunkand its branches is vital for accurate nerve selec-tion in any attempt to reconstruct the ‘physiolo-gical’ anorectum following rectal excision.

Surgical Approach. Most surgeons are unfami-liar with pudendal nerve anatomy because itsexposure is infrequently encountered in clinicalpractice. When access is attempted, the surgeonmay find it difficult, with risk of injury to thenerve and its branches. Moreover, interest todate in the pudendal nerve has largely focused

Figure 7-3. Normal course of pudendal nerve and branches.


on the successful application of non- and mini-mally invasive techniques for the diagnosis andtreatment of pudendal neuropathy, a recognizedcause of pelvic floor, rectal, and anal sphincterdisorders, rather than surgery.42

Surgical access to the pudendal nerve inhumans has mostly been limited to cadavericstudies and occasional attempts at pudendalnerve decompression for cases of unexplainedanal or perineal pain. Pudendal nerve trunkanomalies have been demonstrated, but sur-geons wishing to develop a reproducible surgicalapproach have little to guide them.43 Indeed,some investigators acknowledge that optimalexposure of the pudendal nerve and its brancheswithout injury is a potential obstacle to attemptsto reconstruct a functional anal sphincter com-plex after rectal extirpation.26

One study examined pudendal nerve anat-omy in cadavers seeking to identify potentialanomalies of clinical significance and todescribe a new, simple, reproducible approachfor maximal exposure of the nerve and itsbranches that might contribute to improvedaccess for functional reconstructive proce-dures.44 They found that a simple four step sur-gical approach was not only simple but alsoreproducible and gave maximal pudendalnerve exposure as follows

Step 1: The mid-point of the lateral borderof the sacrum, the ischial tuberosity, and theipsilateral greater trochanter of the femur wereidentified and marked with indelible ink

(Figure 7-4). These surface landmarks were cho-sen because they were easily identifiable anddefine the anatomical region of interest.

Step 2: A vertical incision through skin andsubcutaneous tissue was extended from the midsacrum to the ischial tuberosity, then curvedupward and laterally to the greater trochanterto create a flap that was reflected to reveal obli-que fibers of the distal half of gluteus maximusarising from its sacral origin and directed down-ward (Figure 7-5).

Step 3: A longitudinal incision was madethrough the exposed gluteus maximus onto thesacrum and extended distally to the ischial tuber-osity so that lateral retraction of this muscleexposed the glistening fibers of the sacrotuberous

Figure 7-4. Surface landmarks forpudendal nerve exposure in humancadaver.

Figure 7-5. Exposure of gluteus maximus muscle.


ligament and inferior gluteal neurovascular bun-dle arising from its free edge (Figure 7-6).

Step 4: The sacrotuberous ligament was thendivided at its distal attachment and drawn out-ward along its free border to reveal the pudendalneurovascular bundle in a connective tissue tun-nel (Alcock’s canal) (Figure 7-7).

The study concluded that the aforementionedfour-step approach, permitted optimal exposureof the pudendal nerve and its branches therebyfacilitating restoration of neural innervation toanal sphincter reconstructive procedures. Simi-larly, another group explored the anatomicbases of the functional graciloplasty with puden-dal nerve anastomoses in seven adult humancadavers.45 They showed that re-innervation ofthe gracilis muscle transposed around the anuswith pudendal end-to-side anastomoses was fea-sible and predicted eventual clinical application.

Nerve Regeneration and Attainment of RestoredFunction. Peripheral nerve regeneration andthe return of normal physiological functionafter transection or injury have been the subjectof numerous experimental and clinical studies.Promising experimental results in large animalssuggest that a mixed peripheral nerve like thepudendal nerve following controlled injury, asin transection, can regenerate to innervate thetarget organ and potentially restore automaticfunction.25,26 This return in function is almostuniversally assessed by qualitative techniques asaccurate quantitative tests are impossible tointerpret in animals, despite their crucial rolein the clinical assessment of regeneration.However, more recently, quantitative methodsutilizing qol tools and radiological imaging toassess defecatory function have been adoptedin patients who have undergone anorectal

Figure 7-6. Exposure of sacrotuberousligament.

Figure 7-7. Maximal exposure ofpudendal neurovascular bundle.


reconstruction with the physiological neo-sphincter after rectal excision for cancer.

The pathophysiology of nerve injury has notbeen fully elucidated but current knowledgewould seem to implicate diminished viabilityor death of Schwann cells, axonal degeneration,depletion of nerve or target-derived neuro-trophic factors, and the loss of primary sensory(dorsal root ganglia) and motor neurons.46

Predictably, attempts to improve nerve regen-eration would therefore have to rely on thesuccessful manipulation of the aforementionedpathophysiological processes triggered by nerveinjury.47

Nerve growth factor (NGF), neuropetide-3(NT-3), and neuropeptide-Y (NT-Y) are just afew of the neurotrophic factors that have beenshown to play a role in the maintenance andsurvival of nerve cells, in addition to mimickingthe effect of target organ-derived trophic factorson neuronal cells.48 Convincing evidence of neu-ronal regeneration using exogenous neuro-trophic factors has been demonstrated, largelyin small laboratory animals, although equallyencouraging results employing similar techni-ques have been reported in non-human pri-mates.49 The outcome of ongoing clinical trialsis awaited.

Most investigators recognize that the specifi-city of re-innervation is the most importantdeterminant of successful nerve regeneration,but there is controversy whether the regenerat-ing axons are physically guided by basal laminartubes or are drawn directly by neurotrophicfactors originating from the distal nerve stumpand targets which influence the sorting of regen-erating axons.46 However, it appears that bothfactors come into play following injury of theperipheral nervous system. A major drawbackencountered after nerve injury is the frequentretraction of fibers resulting in a shortfall inlength which must be accommodated to allowsuccessful reinnervation of target or end organs.To this end, the use of nerve autograft to fill asignificant gap remains the gold standard,47 butwhere an autograft is not possible, an allograftor autogenous biodegradable nerve conduitshave so far shown promise in experimentalstudies.50–52

Where anorectal transplantation with puden-dal nerve anastomosis following surgical abla-tion of the rectum is shown to be technically andbiologically feasible, regeneration of the recipi-ent’s pudendal nerve and reinnervation of the

graft with the attendant immunological chal-lenges will become the key to success. It is alsoanticipated that ongoing research leading to bet-ter understanding of the host rejection response,advances in immunotherapy, and the develop-ment of neurotrophic factors will ultimatelytransform application of allogenic nerve graftsfrom the laboratory into the clinical setting.

Technological Advances Which MakeAnorectal Transplantation Feasible

Tremendous technological advances over thelast 30 years have made organ transplantation afeasible alternative treatment for end-stageorgan failure. In many cases, it is considered aless expensive form of treatment for a failed ordiseased organ, which adds to the human andclinical interests among investigators and hasresulted in remarkable progress in terms ofgraft survival and recipient well-being. Thisimproved outcome owes much to better surgicaltechniques and in particular microsurgerywhich allows anastomosis of vessels and nerverepair of structures measuring less than 1 mm indiameter.

Improvements in graft survival and functionhave also resulted from better matching at theDR site of HLA locus of the major histocompat-ibility complex (MHC), coupled with advancesin harvested graft preservation techniques.

Perhaps the greatest stride forward in termsof graft survival has been in the area of immunetherapy and modulation of the host response toforeign antigen. In 1978, the clinical advent ofa novel immunosuppressive agent, cyclosporinA, transformed organ transplantation with imp-roved outcome.53 Further progress soon followedwith the introduction of Tacrolimus (Prograf,FK506), a fungal metabolite which acts in a simi-lar way to Cyclosporin A by inhibiting the earlieststeps of T-cell activation.54

Randomized clinical trials now indicatethat Tacrolimus in comparison to CyclosporinA results in a further decrease in the incidenceof rejection.55–57 The development of newerdrugs in this area offers even greater pro-spects of achieving one of the ultimate goalsin organ transplantation: to protect the trans-planted graft and recipient from rejection andadverse effects so well that transplantation forquality of life indications becomes widelyaccepted.


Anorectal Transplantation

Despite the arguments in favor of anorectaltransplantation as an alternative to abdominalwall colostomy after excision of the rectum, thereare no data in the literature that have examinedthe feasibility of this new concept in humans.However, one study explored the technical feasi-bility of orthotopic allotransplantation of theanorectum and assessed graft viability after abdo-mino-perineal excision of the rectum in an animalmodel.57

Pre-operative Preparation. Large White Land-race pigs were selected for this study becausepreliminary cadaveric dissection had shownthat females had a well-developed externalsphincter muscle complex surrounding theanal canal and vagina, in contrast to males,in which the muscle was less bulky andencircled the anal canal only. In view of thispeculiar anatomy, and ease of operation, fourfemales (22–42 kg) provided donor anorectumto four recipient males (29–39 kg) havingreceived pre-medication (xylazine 1 mg/kgand ketamine 5 mg/kg) 1 h before standardgeneral anesthesia (1–2% halothane in 4 Loxygen and 1 L nitric oxide per minute)(Figure 7-8). Intravenous access was estab-lished in each recipient and 4.3% dextrose/saline given at a rate of 100 ml/h, duringand after transplantation.

Donor Operation. Commencing in the left lat-eral position, a circumano-vaginal incision wasextended on both sides in a posterior lateraldirection (Figure 7-9). Preliminary cadavericdissections had shown that mobilizationbetween the rectum and the coccyx, with detach-ment of the gluteus maximus muscle along theincision margin, provided optimal exposure ofthe pudendal neurovascular bundle arising fromthe lateral pelvic wall.

The external anal sphincter around both anusand vagina was dissected free and detached fromthe pubic bone (Figure 7-10).

The animal was then placed in the supineposition and laparotomy with division of thesymphysis pubis performed. The inferiormesenteric artery and vein were prepared fordivision and the operation was suspendedwhile the recipient was prepared. When thiswas accomplished, intravenous heparin 200units per kg body weight was given, and thepudendal structures divided as far proximal aspossible to facilitate subsequent re-anastomosisin the recipient. The colon was then transected,followed by division of the inferior mesentericartery and vein, releasing the specimen. Benchdissection was then performed to detach thevagina from the donor specimen (Figure 7-11).The mesenteric artery was cannulated and per-fused with 200 ml heparinized saline (5000 unitsheparin in 1000 ml normal saline) and preservedat 0–48C to await transplantation.

Figure 7-8. Illustration of pig understandard general anesthesia.


Recipient Operation. The perineal phase wassimilar to the donor operation, except insteadof dealing with vagina, these pigs being male, theexternal anal sphincter encircling the anal canalwas stripped free from the cavernosum musclewhich was preserved. The abdominal phaseomitted division of the pubic symphysis. Thenext steps were heparinization of recipient(200 units/kg); transperineal introduction ofthe anorectal graft; rectal anastomosis at theproximal end (single layer of interrupted 3/0polyglactin) to stabilize the graft before micro-surgical anastomosis of inferior mesentericartery and vein (10/0 nylon in all pigs); laparot-omy wound closure; microsurgical pudendalartery anastomosis (10/0 nylon in 2 pigs); bilat-eral pudendal nerve anastomosis (10/0 nylon byepineural technique in three pigs) (Figure 7-12);and closure of the perineum (3/0 prolene)(Figure 7-13). After transplantation, all recipientanimals were given intramuscular analgesia(carpofen 4 mg/kg) every 4 h.

Recorded Variables Post Transplantation. Theduration in minutes of the operation andischemic times were recorded by an observerand various parameters of anastomosed struc-tures (vein, artery and nerve) were measured

with a millimeter grid scale using a microscopeat �10 magnification. Orthotopic allotransplan-tation of the anorectum was attempted in fourshort-term live pig models, and before they weresacrificed 24 h later, observation at laparotomyrevealed two pink grafts, one slightly dusky buthealthy graft and one outright failure, reflectingthe state of the mesenteric vessels, which werepatent in three and thrombosed in one. Histolo-gical examination of these observed findingsshowed no difference between the control biop-sies and the three cases with satisfactory mesen-teric flow. However, gross ischemia was presenthistologically in the failed case.

The study concluded that anorectal transplan-tation was technically feasible in a short-term pigmodel and called for further long-term studies toassess the return of normal defecation, preven-tion of sepsis, and overcome rejection issuesbefore it can be considered in humans. It is ofinterest that in this model, the animals did notreceive antibiotic or immuno-suppressive ther-apy, a reasonable strategy given that the earliestsigns of acute rejection may take up to 1 week todevelop in the colon.58 One possible explanationfor this delayed large bowel antigenicity com-pared to the small intestine is the relativepaucity of lymphoid tissue in the colon.59 This

Figure 7-9. Exposure of perineum inpig and circumano-vaginal incision.


phenomena may facilitate a good outcome offuture attempts at anorectal transplantation, per-haps because fewer immunological challenges areencountered.

Indications

Although clinically unproven, anorectal trans-plantation would be a reasonable alternativestrategy in people requiring rectal extirpationfor cancer, trauma, or end-stage benign disease.This novel approach is beset with numerousobstacles not least the technical, immunological,and biological challenges which to date areuntested in humans. However, it is foreseenthat over the next decade the expected expan-sion in organ transplantation for quality of lifereasons alone as opposed to saving life will driverenewed interest in this area as investigatorscontinue in their quest to restore normal defeca-tion albeit physiological, following surgical exci-sion of the rectum.

Complications

The drawbacks of anorectal transplantation canbroadly be divided into problems related tothe technique (organ harvesting/preservation,donor/recipient operations) and the immunolo-gical and biological challenges.

The technique for anorectal transplanta-tion has yet to be elucidated in humans butknowledge of the normal immune response toforeign protein is fundamental to our under-standing of the complex immune-biologicalevents triggered by organ transplantation;that must be conquered not only to preventrejection but also to improve survival andfunction of allogenic graft. This immune-biological challenge may take the form ofimmune response to foreign antigen, allograftrejection, and major histocompatibility com-plex (MHC).

Immune Response to Foreign Antigen. Allotrans-plantation induces an immune response in the

Figure 7-10. Dissection en bloc of the anorectum, pudendal neurovascular bundle, vagina, and uterus.


recipient which may be antigen specific (T andB cell mediated) or non-specific (macrophageand cytokine mediated), but frequently involvesa combination of these two mechanisms. Anti-gens are presented to recipient T cells eitherdirectly (antigen delivered by donor antigenpresenting cell) or indirectly (antigen deliveredby recipient antigen presenting cell). The acti-vated T cell via chemical messengers recruitsother T and B cells to proliferate, differentiate,and develop effector function which then causesdamage in an antigen-specific manner. Theeffector function of T cells arises throughenhanced cytotoxic activity of natural killercells, interleukin and interferon, while B cell

effector function results in proliferation and dif-ferentiation to plasma cells which secrete speci-fic antibody resulting in complement fixation orantibody-mediated cytotoxicity. On the otherhand, the non-specific immune responseinvolves the activation and recruitment ofmacrophages, cytokines, and other non-specificeffector cells by their interaction with T cellswhich causes direct damage of grafts.

Allograft Rejection. The rejection of allograftmay occur by one of three mechanisms: cell-mediated, which is orchestrated by circulatingT cells; antibody mediated following activationof plasma cells; and other mechanisms whichremain poorly understood but are believed toinvolve cytokines such as tumor necrosis factor(TNF). Based on extensive work in human renalallografts, the speed of organ rejection followingtransplantation can be classified into three:

• Hyperacute: occurring within minutes to sev-eral hours after transplantation due to pre-formed circulating antibodies.

• Acute: occurring within days to a few monthsafter transplantation and may be the result ofcell-mediated immunity

• Chronic: occurring more than 6 months aftertransplantation and is due to persistentimmune response or the long-term sideeffects of drugs.60

Because rejection can be seen to occur atvarying intervals after transplantation it is gen-erally assumed to be the result of the differentmechanisms outlined above, but it is more likelythat all three events are simultaneously acti-vated, albeit to varying degrees. It is reasonableto predict that virtually any cell, cytokine, orchemical mediator that is capable of tissuedestruction may play an important role in theensuing immunological sequelae after organtransplantation, which might explain why ithas so far proved difficult completely to controlthe immune response. Attempts to improvegraft survival and function which entail suppres-sion of the entire immune system in recipientsmight impede activation of those agents whichseek to destroy the transplanted organ, butwould leave the patient fatally exposed to infec-tious diseases and oncogenic viruses.

Major Histocompatibility Complex (MHC). Adonor graft carries on its surface transplantationantigens known as major histocompatibility

Figure 7-11. Donor anorectal graft.


complex (molecules which present antigens to Tlymphocytes, MHC) located on the short arm ofchromosome six. In humans, these antigenswhich govern rejection are known as humanleucocyte antigens (HLA). The human leococyteantigen contains three regions which encode fortype I, II, and III antigens. Types I and II arehighly polymorphic or variable and trigger spe-cific T cell response, while the less polymorphictype III antigen has no role in initiating specificT cell responses, but nonetheless plays animportant part in immunity through inductionof cytokines and the complement system. Therelevance of HLA compatibility between donorand recipient is that when patients are wellmatched particularly at the DR site of the HLAII locus of the major histocompatibility com-plex, outcome in terms of graft survival andfunction fare better than those mismatched.60

Prosthetic Bowel Sphincter

Artificial Bowel Sphincter

The artificial urinary sphincter (AUS) wasintroduced into clinical practice over threedecades ago for the management of genuinestress urinary incontinence. The AMS-721(American Medical Systems) also known asthe Brantley Scott AUS initially consisted offour main components (a reservoir, inflatableocclusive cuff, inflating, and deflating pump

mechanism). Further modification of thisdevice resulted in the development of theAMS 800 series in 1982, which had threecomponents (balloon, cuff, and pump) andhas retained its role as the ‘gold standard’artificial urinary sphincter or prototypefrom which other types of artificial sphincterdevices have evolved.

In 1987, Christiansen and Lorentzen werecredited with successfully implanting what wasessentially a urological device in a single patientfor the management of anal incontinence of neu-romuscular origin. The prosthesis consisted ofthree components – a balloon to regulate pres-sure, a pump representing the control assembly,and a cuff which mirrored the anal sphincter(Figure 7-15). They demonstrated that the devicegenerated a sufficient pressure (66–74 mmHg)which lay within the normal range of restingsphincter pressure, allowing closure of the analcanal without mechanical failure, infection of thedevice, or local erosion of tissue.61

Other investigators have also utilized the‘AMS 800’ artificial urinary sphincter (AmericanMedical Systems) for the management of analincontinence secondary to benign disease.62–64

Description of the AMS 800 ArtificialBowel Sphincter

The device is filled with fluid and is implantedentirely within various parts of the body in

Figure 7-12. Illustration of anasto-moses of pudendal neurovascular bun-dle by microsurgery.


either men or women with severe fecal incon-tinence not amenable to other forms of treat-ment. It is designed to control stool expulsionand it is claimed to simulate normal sphincterfunction by opening and closing the anal or neo-anal canal at the control of the patient. Thedevice is made from solid silicone elastomerthat is inert and minimizes the risk of rejectionby the body. The cuff is implanted around theanal or neo-anal canal, while the pump is placedin the scrotum or labium. The pressure regulat-ing balloon is implanted in the lower abdomen,under the muscle layer, and just above the pubicsymphysis. It is normally filled with sterile solu-tion usually saline that can be imaged with plainx-ray.

Device Activation/Deactivation

In the activated or closed position, the cuff isfilled with fluid and gently squeezes shut theanal or neo-anal canal to maintain fecal con-tinence. To defecate, the system is deactivatedby squeezing and releasing the soft lower partof the pump several times. This leads to thetransfer of fluid from the cuff to the balloon,leaving an empty cuff that no longer appliespressure to the bowel and permits expulsionof stool. The fluid that is transferred to theballoon creates pressure within the balloonwhich pushes the fluid back into the cuff torestore continence. It takes several minutesfor the cuff to re-fill, presumably sufficient

Figure 7-13. Closure of perinealwound in recipient.


time for fecal evacuation to take place beforethe anal canal is squeezed shut again by thefilled cuff.

Functional Outcome of the ArtificialBowel Sphincter

In a large personal series of 53 patients Devesaet al. evaluated the technique, and functionalresults following implantation of the Acticonartificial bowel sphincter for total anal conti-nence not amenable to sphinteroplasty or failed

sphincter repair.65 They reported that the artifi-cial anal sphincter restores continence to solidstool in almost all patients with preceding severeincontinence and that two-thirds of thesepatients achieved normal continence. Anotherstudy evaluated experience with the device in asingle institution over an 8-year period. Thirty-seven consecutive patients were prospectivelyevaluated, the majority of whom had eithersphincter disruption or neurologic disease(35/37), while only two patients had hereditarymalformation. They concluded that satisfactorycontinence was achieved following assessment

Figure 7-14a. Illustration of techni-cally satisfactory anastomoses ofmesenteric vessels before harvesting24 h post-transplantation.

Figure 7-14b. Illustration of healthylooking grafted skin in fully active/alertanimal after transplantation.


with physical examination (anal continence andrectal emptying) and manometry.66 One rando-mized clinical trial compared implantation ofthe artificial bowel sphincter versus a programof supportive therapy in 14 severely incontinent

adults. Using continence (Cleveland ContinenceScore) and quality of life (SF-36) outcome mea-sures, they found that despite device explanta-tion in one patient (14%), the artificial bowelsphincter was easy to use, effective in restoring

Figure 7-14c. Illustration of outrightfailure of graft before harvesting 24 hpost-transplantation.

Figure 7-15. Acticon artificial bowelsphincter (AMS) consisting of a cuff,pump, and balloon.


continence and improved quality of life superiorto supportive therapy.67 On the other hand, in asmaller series of 12 patients with implanted arti-ficial neo-sphincter devices for end-stage analincontinence, Wong et al. found less convincingdata in support of the artificial bowel sphinc-ter.63 They reported complications in a third oftheir patients which resulted in poor clinicaloutcome. Furthermore, a systematic review ofthe safety and effectiveness of the bowel sphinc-ter for fecal incontinence revealed that itsimplantation is of uncertain benefit and maypossibly harm many patients because of highmorbidity.68

Surprisingly, despite the relative success ofthese artificial bowel sphincters in the manage-ment of end-stage fecal incontinence (neurolo-gical disorder or irreparable sphincter defect),there is limited data in the literature evaluatingtheir use after abdomino-perineal excision of therectum for malignant disease, even though datafrom experimental studies had shown muchpromise for future clinical application.31,69,70

One plausible explanation why this is so maybe due to concerns relating to oncological clear-ance, given that most clinicians would elect todelay neo-anal reconstruction until tumor clear-ance has been confirmed histologically. For thisreason, it is therefore not surprising why cur-rently there are only a few reports of case orpersonal series that describe implantation ofthe device after Miles’ operation71,72 or as asecondary procedure in the small number ofpatients who opt for a perineal colostomy atthe time of excision of their rectal cancer.65,73

Indications

Artificial bowel sphincters may be considered asa feasible and effective management strategy forfecal incontinence due to benign or malignantdisease.

Benign Disease. Patients with end-stage fecalincontinence due to sphincter disruption(obstetrics, trauma, anal surgery) or where pre-vious attempts at sphincter repair have failed arethe most frequent indications for treatment withthe artificial bowel sphincter. Arguably, the arti-ficial bowel sphincter is currently the only sur-gical option for treatment of anal incontinencein patients with neurologic disease that affectsthe pelvic floor and muscles of the lower limb.

There are also those patients particularly in thepediatric population with congenital anomaliessuch as imperforate anus who would benefitfrom treatment with the artificial bowel sphinc-ter. In these circumstances, implantation of thesphincter device is usually deployed at the timeof the ‘pull-through’ operation or as a secondaryprocedure many months or years later.

Malignant Disease. Low rectal or anal cancerrequiring abdomino-perineal excision of the rec-tum results in an incontinent abdominal or occa-sionally perineal colostomy. Implantation of theartificial bowel sphincter either as a primary or asa secondary procedure following a Miles’ opera-tion is a feasible and effective treatment optionthat may restore anal continence, improve qualityof life, and avoid a stoma. However, experienceon this subject is limited to a handful of casereports or personal series from single institutionspossibly because of the general concern regardingthe risk of cancer recurrence alluded to above.

Contraindications

The main contraindication to implantationof the artificial bowel sphincter is active ongoingor chronic infection. Patients with pathologicalproblems that affect rectal compliance, result inchronic diarrhea, persistent fecal impaction, orhave had previous surgery resulting in scarringor impairment of the vascular supply to thebowel that precludes implantation of the devicein the pelvis. Those selected should have nophysical disability, must be medically fit to tol-erate a major abdominal operation, and shouldalso be determined and highly motivated tooperate the device correctly.

Types

The adaptation of the AMS 800 artificial urinarysphincter in 1987 for the management of fecalincontinence by Christensen et al. led to theevolution of the current artificial bowel sphinc-ter which also consists of the same three compo-nents (cuff, balloon, and pump) as the AMS 800prototype.61 More recently, other types of artifi-cial bowel sphincters have emerged such as theprosthetic artificial sphincter (PAS), a remotecontrolled artificial bowel sphincter, and theshape memory alloy (SMA). While the PAS has


been successfully tested clinically, the remotecontrolled and the SMA versions are still at thepre-clinical stage.

Prosthetic Anal Sphincter (PAS). The PASdevice was introduced almost a decade ago andinitial data from animal experiments revealedthat it produced fecal continence without causingischemia at the point of contact with the intes-tine.70 It shared a similarity with the AMS 800 inthat it also consisted of three parts (pump, bal-loon, and cuff) but its inventors claimed that PASsimulated normal action and function of the analcanal under the patients’ control by occluding theanorectum at an angle reminiscent of the pubor-ectalis. Another unique design is the pump whichhas two parts; a bulb which when squeezed emp-ties the cuff of occlusive gel permitting defeca-tion, and a button located over the dome whichwhen depressed re-fills the cuff and restores fecalcontinence. Using this device in 12 patients withsevere fecal incontinence, investigators revealedafter a median follow-up of 29 months that thedevice restored continence in 10 of 11 patients,with no device-related infective complicationsalthough one patient developed pseudomembra-nous colitis requiring device explantation.74

Remote Controlled Artificial Bowel Sphincter(ABS). One study describes pre-clinical use ofa remote-controlled artificial bowel sphincterprototype that effects fecal continence with com-fortable control.75 This device has similar designto the three part AMS 800, but in this case com-prises a micropump based on piezo-technologyallowing remote transmission of signalsenabling cuff inflation and deflation.

Shape Memory Alloy (SMA). This novel deviceutilizes a solid driving element, a combination ofshape memory alloy(SMA), and layered siliconelastomer sheets for the open and close phase ofthe artificial bowel sphincter.76 Although still atthe pre-clinical development stage, the authorsclaim that the device has few parts inside thebody and therefore can be implanted more easily.

Patient Selection and Assessment

All patients who fulfill the indications for anartificial bowel sphincter outlined above should

be considered. A full evaluation of the patient,including a thorough history and complete phy-sical examination, is essential prior to insertionof an artificial bowel sphincter. It is particu-larly important to establish that the fecalincontinence is sufficiently disabling to war-rant surgical intervention and that all otherless invasive alternatives have been previouslyexplored. A pre-operative septic screen toensure sterility must be ascertained prior todevice implantation and should include eradi-cation of MRSA organisms colonizing the skinand the nasal region. If there is any clinicalevidence of anal sepsis such as a fistula inano, device implantation should be postponedand the condition should be dealt with andcured.

Implantation of the ArtificialBowel Sphincter

Pre-operative preparation should include fullmechanical bowel preparation with antibioticand deep venous thrombosis prophylaxis. Theantibiotics should cover both aerobic and anae-robic organisms and it is advised that the entiredevice should be soaked in 120 mg of gentamy-cin solution prior to implantation. A lowermidline laparotomy incision gives good accessto the rectum and it is customary to enter themesorectal plane by dividing the pelvic perito-neal reflection on one side. Minimum dissectiondistally is required while in this plane to reachthe anorectal junction at which point a smallincision is made in the contralateral peritonealreflection to enable the sphincter cuff to encirclethe upper border of the anal canal just above thepelvic floor muscles. It is important to establishat this point that with the sphincter cuff deflated,evacuation of bowel contents is not impeded andis easily assessed by using the surgeon’s indexfinger. Implantation of the cuff component ofthe artificial bowel sphincter is concluded byclosing the peritoneal reflection to isolate thedevice from the peritoneal cavity.

A subcutaneous pouch is then created usuallyin the right iliac fossa for the control pump but itcan also be sited either in the scrotum or labiamajora. The connecting tubes from the controlpump are attached to the cuff and regulatingballoon, respectively, and the latter is left to liefree within the pelvic cavity.


The surgeon should ensure that the device issurrounded by generous subcutaneous tissuewithin the pouch to prevent damage fromtrauma, device erosion, or migration. It is alsothe operators’ responsibility to test and confirmsatisfactory functioning of the artificial bowelsphincter before the abdomen is closed.

Complications

A number of complications are associated withthe artificial bowel sphincter, while some arerelatively minor, others such as device failure,infection, or erosion present major problemsnecessitating explantation of the device.

Mechanical Failure. This is nearly alwaysrelated to the technique employed during deviceimplantation given that less than 3% of allmechanical failures are attributed to the deviceitself. The vast majority of the technical pro-blems are due to inadvertent blocking or kink-ing of the tubing system or fluid leakage fromaccidental damage causing inadequate balloonpressure. Like any other, mechanical device, theartificial bowel sphincter is subject to wearand tear that may include control pump fail-ure, disconnection of its prime components,or even damage from repeated trauma. Eithercategory will impair the normal transmissionof fluid to the cuff and prevent the squeezingshut of the intestine to maintain fecal conti-nence. These complications occur almostimmediately with failure of continence onactivating the device. A non-functioningdevice is an indication for careful inspectionand possible surgical revision where an irre-mediable problem is identified.

Infection. Uncontrolled infection inevitablyresults in explantation of the device and thebest form of treatment remains unquestionably,prevention. Because infection most frequentlyoccurs after implantation of the device, strictsterile techniques and antibiotic prophylaxisare mandatory, along with regular observationin the postoperative period to detect early signsor symptoms of sepsis. There are of course somepeople who carry a greater risk of infection suchas those with existing stomas, skin conditions,impaired immunity, and diabetes. Thesepatients should be carefully counseled and

warned that it may not be possible to re-implantthe device after explantation due to infection.The infection rates of these devices rangebetween 20 and 40% and one study whichreported experience with the device in the Uni-ted Kingdom showed that infection with methi-cillin-resistant Staphylococcus aureus (MRSA)was the most common cause of failure.77

Erosion. Erosion refers to the wearing away oftissue adjacent to the device. The immediatesurrounding organ is at greatest risk and includethe anal canal, scrotum, labium, urethra, urinarybladder, and the skin overlying the perineum orlower abdominal wall to name but a few. Thisoccurs because of ongoing sepsis, improper size,or positioning of the device, prior tissue damagefrom radiation and skin conditions; while pain,erythema, tenderness, or changes in skin textureare the earliest worrying signs. Sometimes thedevice is visible having broken through the over-lying skin in which case revision or explanationis required. In a large personal series of 53patients who received the artificial sphincterdevice for fecal incontinence, nearly 20%patients suffered cuff and/or pump erosion as alate complication.65

Migration. Any part of the device can migrateto a new location remote from the original siteof implantation. Usually poor surgical techni-que such as improper pump or regulating bal-loon placement, cuff selection, and tubinglength which can either damage adjacentremote tissue or lead to device malfunction.However, early detection will prevent long-term damage and allow revision as opposed todevice explantation.

Recurrent Incontinence. While fecal conti-nence achieved with the artificial bowelsphincter is variously reported in the litera-ture as between 60 and 90%, the true rate ofrecurrent incontinence with the device in situis not known. First, many of these patientsfail to report varying degrees of incontinencebecause they are highly motivated and deter-mined to avoid a colostomy. Second, the var-ious validated tools that are in use to measurefecal incontinence are subjective, lack unifor-mity, and measure differing aspects of fecalincontinence. Third, there is no universallyagreed definition of recurrent incontinencewith the device in situ given that few studies


document explantation of the device becauseof unsatisfactory function compared to itsremoval as a result of complications. It isalso relevant to highlight the fact that theoverwhelming majority of studies on this sub-ject are case series or single center experiencewith their inherent biases. All these factorscontribute to the present confusion and thededuction that can be drawn from the litera-ture is that the true incidence of recurrentincontinence is unknown.

Constipation/Fecal Impaction. These are com-mon problems that occur following deviceimplantation but are usually resolved by a com-bination of dietary modification and use of orallaxatives. Occasionally, regular enemas to evac-uate the neo-rectum may be required in thosepatients who encounter repeated constipation orfecal impaction.

Summary

The use of artificial sphincters in colorectal sur-gery is an acceptable management strategy torestore anal defecation in patients with end-stagefecal incontinence or following rectal excision forcancer, who would otherwise have to face life witha permanent colostomy. These sphincters may befashioned surgically or involve implantation of anartificial device. The passive neo-sphincters havelargely been superseded by the electrically stimu-lated variety which is the gold standard in terms ofefficacy, but unfortunately, carry considerablemorbidity. It is therefore recommended thatpatients considering anal reconstruction with theelectrically stimulated neo-sphincter should becarefully selected and surgery restricted to specia-list colorectal centers. However, while data per-taining to the physiological neo-sphincter is cur-rently limited, it has shown promise given theexciting prospects of restoring anal defecationunder normal voluntary control.

Anorectal transplantation has been shown tobe technically feasible experimentally, but it isbeset with formidable obstacles not least thereturning physiological function and immuno-logical considerations. Nonetheless, it is fore-seen that further interest in this novel methodof restoring normal anal defecation will be dri-ven by the quest to restore quality of life bytransplantation of none essential organs. Theimplantable artificial sphincters offer an

acceptable alternative with reported continencerates comparable to the neo-sphincters, butdevice-related complications are frequent andmay result in failure or explantation.

References

1. Stone H. Plastic operation for anal incontinence. ArchSurg. 1929;18:845.

2. Chetwood CH. Plastic operation for restoration of thesphicter ani with report of a case. Med Rec. 1902;61:529.

3. Chittenden A. Reconstruction of anal sphincter bymuscle strips from glutei. Ann Surg. 1930;92:152.

4. Bistrom O. Plastischer Erzatz das M. sphincter ani. ActaChir Scand. 1944;24:120.

5. Hentz VR. Construction of a rectal sphincter using theorigin of the gluteus maximus muscle. Plast ReconstrSurg. July 1982;70(1):82–5.

6. Bruining HA, Bos KE, Colthoff EG, Tolhurst DE. Crea-tion of an anal sphincter mechanism by bilateral proxi-mally based gluteal muscle transposition. PlastReconstr Surg. January 1981;67(1):70–3.

7. Pickrell KL, Broadbent TR, Masters FW, Metzger JT.Construction of a rectal sphincter and restoration ofanal continence by transplanting the gracilis muscle; areport of four cases in children. Ann Surg. June1952;135(6):853–62.

8. Simonsen OS, Stolf NA, Aun F, Raia A, Habr-Gama A.Rectal sphincter reconstruction in perineal colostomiesafter abdominoperineal resection for cancer. Br J Surg.May 1976;63(5):389–91.

9. Fedorov VD, Shelygin YA. Treatment of patients with rectalcancer. Dis Colon Rectum. February 1989;32(2):138–45.

10. Baeten CG, Konsten J, Spaans F, Visser R, Habets AM,Bourgeois IM, et al. Dynamic graciloplasty for treatmentof faecal incontinence. Lancet. November 9, 1991;338(8776):1163–5.

11. Williams NS, Patel J, George BD, Hallan RI, Watkins ES.Development of an electrically stimulated neoanalsphincter. Lancet. November 9, 1991;338(8776): 1166–9.

12. Geerdes B, Kurvers H, Konsten J, Heineman E, Baeten C.Assessment of ischaemia of the distal part of the gracilismuscle during transposition for anal dynamic gracilo-plasty. Br J Surg. 1997;84(8):1127–9.

13. Rosen H, Novi G, Zoech G, Feil W, Urbarz C, Schiessel R.Restoration of anal sphincter function by single-stagedynamic graciloplasty with a modified (split sling)tech-nique. Am J Surg. March 1, 1998;175(3):187–93.

14. Mander BJ, Abercrombie JF, George BD, Williams NS.The electrically stimulated gracilis neosphincter incor-porated as part of total anorectal reconstruction afterabdominoperineal excision of the rectum. Ann Surg.December 1996;224(6):702–9.

15. Williams NS, Hallan RI, Koeze TH, Watkins ES. Restora-tion of gastrointestinal continuity and continence afterabdominoperineal excision of the rectum using an elec-trically stimulated neoanal sphincter. Dis Colon Rectum.July 1990;33(7):561–5.

16. Mercati U, Trancanelli V, Castagnoli GP, Mariotti A,Ciaccarini R. Use of the gracilis muscles for sphinctericconstruction after abdominoperineal resection. Techni-que and preliminary results. Dis Colon Rectum. December1991;34(12):1085–9.


17. Cavina E. Outcome of restorative perineal graciloplastywith simultaneous excision of the anus and rectum forcancer. A ten-year experience with 81 patients. Dis ColonRectum. February1996;39(2):182–90.

18. Cavina E, Seccia M, Banti P, Zocco G. Anorectal recon-struction after abdominoperineal resection. Experi-ence with double-wrap graciloplasty supported bylow-frequency electrostimulation. Dis Colon Rectum.August 1998;41(8):1010–6.

19. Patel J, Shanahah D, Riches D. The arterial anatomy andsurgical relevance of the human gracilis muscle. J Anat.1991;176:270–272.

20. Mander BJ, Williams NS. The electrically stimulated gra-cilis neo-anal sphincter. Eur J Gastroenterol Hepatol.May 1997;9(5):435–41.

21. Rongen MJ, Dekker FA, Geerdes BP, Heineman E, BaetenCG. Secondary coloperineal pull-through and doubledynamic graciloplasty after Miles resection–feasible,but with a high morbidity. Dis Colon Rectum. June1999;42(6):776–80.

22. Rullier E, Zerbib F, Laurent C, Caudry M, Saric J. Mor-bidity and functional outcome after double dynamicgraciloplasty for anorectal reconstruction. Br J Surg.July 2000;87(7):909–13.

23. Rullier E, Laurent C, Zerbib F, Garrelon JL, Caudry M,Saric J. [Anorectal reconstruction by coloperineal ana-stomosis and dynamic double graciloplasty after abdo-mino-perineal resection]. Ann Chir. 1998;52(9):905–12.

24. Abercrombie JF, Rogers J, Williams NS. Total anorectalreconstruction results in complete anorectal sensoryloss. Br J Surg. January 1996;83(1):57–9.

25. Sato T, Konishi F. Functional perineal colostomy withpudendal nerve anastomosis following anorectal resec-tion: an experimental study. Surgery. June 1996;119(6):641–51.

26. Congilosi SM, Johnson DR, Medot M, Tretinyak A,McCormick SR, Wong WD, et al. Experimental modelof pudendal nerve innervation of a skeletal muscle neo-sphincter for faecal incontinence. Br J Surg. September1997;84(9):1269–73.

27. Schwabegger AH, Kronberger P, Obrist P, Brath E,Miko I. Functional sphincter ani externus reconstruc-tion for treatment of fecal stress incontinence usingfree latissimus dorsi muscle transfer with coaptationto the pudendal nerve: preliminary experimentalstudy in dogs. J Reconstr Microsurg. February 2007;23(2):79–85.

28. Sato T, Konishi F, Kanazawa K. Anal sphincterreconstruction with a pudendal nerve anastomosisfollowing abdominoperineal resection: report of acase. Dis Colon Rectum. December 1997;40(12):1497–502.

29. Pirro N, Sielezneff I, Malouf A, Ouaissi M, Di M, V, SastreB. Anal sphincter reconstruction using a transposed gra-cilis muscle with a pudendal nerve anastomosis: a preli-minary anatomic study. Dis Colon Rectum. November2005;48(11):2085–9.

30. Sato T, Konishi F, Endoh N, Uda H, Sugawara Y, NagaiH. Long-term outcomes of a neo-anus with a pudendalnerve anastomosis contemporaneously reconstructedwith an abdominoperineal excision of the rectum. Sur-gery. January 1, 2005;137(1):8–15.

31. Satava RM, King GE. An artificial anal sphincter. Phase 2:Implantable sphincter with a perineal colostomy. J SurgRes. March 1989;46(3):207–11.

32. Affleck J. A sound operation? Docs debate benefits, risksof first larynx transplant. Philadelphia Daily News.January 10, 1998.

33. Grady D. Medical first: a transplant of a uterus. NYTimes (Print). March 7, 2002;A1, A11.

34. Cooney WP, Hentz VR. Hand transplantation–primumnon nocere. J Hand Surg [Am]. January 2002;27(1):165–8.

35. Devauchelle B, Badet L, Lengele B, Morelon E, Testelin S,Michallet M, et al. First human face allograft: earlyreport. Lancet. July 15, 2006;368(9531):203–9.

36. Orbach CE, Tallent N. Modification of perceived bodyand of body concepts. Arch Gen Psychiatry. February1965;12:126–35.

37. Devlin HB, Plant JA, Griffin M. Aftermath of surgery foranorectal cancer. Br Med J. August 14, 1971;3(5771):413–8.

38. Wade BE. Colostomy patients: psychological adjustmentat 10 weeks and 1 year after surgery in districts whichemployed stoma-care nurses and districts which did not.J Adv Nurs 1990 Nov;15(11):1297-304.

39. Birchall M. Human laryngeal allograft: shift of emphasis intransplantation. Lancet. Feb 21, 1998;351(9102):539–40.

40. Sinnatamby C. Last’s Anatomy. 10th ed. Edinburgh:Churchill Livingstone; 1998.

41. McMinn R, Peggington J, Abrahams P. A Colour ATlas ofHuman Anatomy. 5th ed. Churchill Livingstone, 1993.

42. Shafik A. Endoscopic pudendal canal decompression forthe treatment of fecal incontinence due to pudendalcanal syndrome. J Laparoendosc Adv Surg Tech A.August 1997;7(4):227–34.

43. Sikorski A, Olszewski J, Miekos E. Anatomical consid-erations of selective pudendal neurectomy. Int UrolNephrol. 1987;19(2):159–63.

44. O’Bichere A, Green C, Phillips RK. New, simple approachfor maximal pudendal nerve exposure: anomalies andprospects for functional reconstruction. Dis Colon Rec-tum. July 2000;43(7):956–60.

45. Pirro N, Konate I, Sielezneff I, Di MV, Sastre B. Anatomicbases of graciloplasty using end-to-side nerve pudendalanastomosis. Surg Radiol Anat. December 2005;27(5):409–13.

46. Liuzzi FJ, Tedeschi B. Peripheral nerve regeneration.Neurosurg Clin N Am. January 1991;2(1):31–42.

47. Flores AJ, Lavernia CJ, Owens PW. Anatomy and phy-siology of peripheral nerve injury and repair. Am JOrthop. March 2000;29(3):167–73.

48. Terenghi G. Peripheral nerve regeneration and neuro-trophic factors. J Anat. January 1999;194(Pt 1):1–14.

49. Ahmed Z, Brown RA, Ljungberg C, Wiberg M, TerenghiG. Nerve growth factor enhances peripheral nerve regen-eration in non-human primates. Scand J Plast ReconstrSurg Hand Surg. December 1999;33(4): 393–401.

50. Trumble TE, Shon FG. The physiology of nerve trans-plantation. Hand Clin. February 2000;16(1):105–22.

51. Hazari A, Wiberg M, Johansson-Ruden G, Green C,Terenghi G. A resorbable nerve conduit as an alterna-tive to nerve autograft in nerve gap repair. Br J PlastSurg. December 1999;52(8):653–7.

52. Strauch B. Use of nerve conduits in peripheral nerverepair. Hand Clin. February 2000;16(1):123–30.

53. Calne RY, White DJ, Thiru S, Evans DB, McMaster P,Dunn DC, et al. Cyclosporin A in patients receiving renalallografts from cadaver donors. Lancet. December 23,1978;2(8104-5):1323–7.

54. Warty V, Diven W, Cadoff E, Todo S, Starzl T, Sanghvi A.FK506: a novel immunosuppressive agent. Characteristics


of binding and uptake by human lymphocytes. Trans-plantation. September 1988;46(3):453–5.

55. Fung JJ, Eliasziw M, Todo S, Jain A, Demetris AJ,McMichael JP, et al. The Pittsburgh randomized trial oftacrolimus compared to cyclosporine for hepatic trans-plantation. J Am Coll Surg. August 1996;183(2): 117–25.

56. Pirsch JD, Miller J, Deierhoi MH, Vincenti F, Filo RS. Acomparison of tacrolimus (FK506) and cyclosporine forimmunosuppression after cadaveric renal transplanta-tion. FK506 Kidney Transplant Study Group. Transplan-tation. April 15, 1997;63(7):977–83.

57. O’Bichere A, Shurey S, Sibbons P, Green C, Phillips RK.Experimental model of anorectal transplantation. Br JSurg. November 2000;87(11):1534–9.

58. Yamauchi T, Taira A, Yoshida H. Rejection of the colonin multi visceral organ transplantation in pigs. In Vivo.January 1993;7(1):9–12.

59. Nakhleh RE, Gruessner AC, Pirenne J, Benedetti E,Troppmann C, Uckun F, et al. Rejection of the colonversus ileum in a pig model of total bowel transplanta-tion. Transplant Proc. October 1996;28(5):2445–6.

60. Hakim N. Introduction to Organ Transplantation. Firsted. London: Imperial College Press; 1997.

61. Christiansen J, Lorentzen M. Implantation of artificialsphincter for anal incontinence. Lancet. August 1,1987;2(8553):244–5.

62. Christiansen J, Sparso B. Treatment of anal incontinenceby an implantable prosthetic anal sphincter. Ann Surg.April 1992;215(4):383–6.

63. Wong WD, Jensen LL, Bartolo DC, Rothenberger DA.Artificial anal sphincter. Dis Colon Rectum. December1996;39(12):1345–51.

64. Vaizey CJ, Kamm MA, Gold DM, Bartram CI, Halligan S,Nicholls RJ. Clinical, physiological, and radiologicalstudy of a new purpose-designed artificial bowel sphinc-ter. Lancet. July 11 1998;352(9122):105–9.

65. Devesa JM, Rey A, Hervas PL, Halawa KS, Larranaga I,Svidler L, et al. Artificial anal sphincter: complicationsand functional results of a large personal series. DisColon Rectum. September 2002;45(9):1154–63.

66. Michot F, Costaglioli B, Leroi AM, Denis P. Artificial analsphincter in severe fecal incontinence: outcome of pro-spective experience with 37 patients in one institution.Ann Surg. January 2003;237(1):52–6.

67. O’Brien PE, Dixon JB, Skinner S, Laurie C, Khera A,Fonda D. A prospective, randomized, controlled clinicaltrial of placement of the artificial bowel sphincter (Acti-con Neosphincter) for the control of fecal incontinence.Dis Colon Rectum. November 2004;47(11):1852–60.

68. Mundy L, Merlin TL, Maddern GJ, Hiller JE. Systematicreview of safety and effectiveness of an artificial bowelsphincter for faecal incontinence. Br J Surg. June2004;91(6):665–72.

69. Sofia CA, Rush BF, Jr., Koziol J, Rocko JM, Seebode JJ.Experiences with an artificial sphincter to establish analcontinence in dogs. Am Surg. June 1988;54(6):390–4.

70. Hajivassiliou CA, Carter KB, Finlay IG. Assessment of anovel implantable artificial anal sphincter. Dis ColonRectum. June 1997;40(6):711–7.

71. Lirici I, Di P, Ponzano H. Dynamic graciloplasty versusimplant of artificial sphincter for continent perinealcolostomy after Miles’ procedure: Technique and earlyresults. Minim Invasive Ther Allied Technol. December2004;13(5):347–61.

72. La TF, Masoni L, Montori J, Ruggeri E, Montori A.The surgical treatment of fecal incontinence withartificial anal sphincter implant. Preliminary clinicalreport. Hepatogastroenterology. Septmber 2004;51(59):1358–61.

73. Marchal F, Doucet C, Lechaux D, Lasser P, Lehur PA.Secondary implantation of an artificial sphincter afterabdominoperineal resection and pseudocontinent peri-neal colostomy for rectal cancer. Gastroenterol Clin Biol.April 2005;29(4):425–8.

74. Finlay IG, Richardson W, Hajivassiliou CA. Outcomeafter implantation of a novel prosthetic anal sphincterin humans. Br J Surg November 2004;91(11):1485–92.

75. Schrag HJ, Ruthmann O, Doll A, Goldschmidtboing F,Woias P, Hopt UT. Development of a novel, remote-controlled artificial bowel sphincter through microsys-tems technology. Artif Organs. November 2006;30(11):855–62.

76. Luo Y, Higa M, Amae S, Takagi T, Yambe T, Okuyama T, etal. Preclinical development of SMA artificial anal sphincters.Minim Invasive Ther Allied Technol. 2006;15(4):241–5.

77. Malouf AJ, Vaizey CJ, Kamm MA, Nicholls RJ. Reasses-sing artificial bowel sphincters. Lancet. June 24,2000;355(9222):2219–20.


8Cochlear Implant

George Fayad and Behrad Elmiyeh

A cochlear implant is a surgically implantedelectronic device which may benefit childrenand adults with severe to profound sensori-neural hearing.

Unlike a hearing aid, which simply makessounds louder (amplification), the cochlearimplant converts sounds into electrical signalsand sends those signals directly to the auditorynerve (hearing nerve), bypassing damagedstructures of the inner ear (hair cells). The audi-tory nerve in turn takes the signals to the brain.

A cochlear implant, despite of being anadvanced and complex technology, is unable torestore hearing fully. However, it can provide aperception of sound and be a means for com-munication when conventional hearing aids areinadequate. The extent of hearing improvementvaries between patients and is difficult to predictaccurately.

A cochlear implant is made up of two parts.The external part is worn like an external

hearing aid and consists of a microphone,which picks up sounds, a speech processor,which selectively filters sound and convertsacoustic signals into electrical signals. Thesesignals are sent through a thin cable to thetransmitter, which is a magnetic pad placedbehind the external ear and transmits the pro-cessed sound signals to the internal device byelectromagnetic induction.

The internal part is the electronic implantdevice, which is surgically placed under theskin behind the ear. This internal part has areceiver, which converts sound energy to elec-trical signals and sends them through an inter-nal cable to the electrode. The electrode is atransmitter wire with specialized ending,wound through the cochlea, stimulating thecochlear nerve directly, by passing the damaged

hair cells. The brain interprets the nerve activityas sound.

Past, Present, and Future of CochlearImplants

In the 1950s, Djourno and Eyries reported thefirst detailed description of direct stimulation ofthe auditory nerve in order to generate hearing.1

In the 1960s, more extensive experimentswith electrical stimulation to the cochlea werecarried out by researchers such as Doyle2 andSimmons3. During the 1970s, the clinical appli-cations of electrical stimulation of the auditorynerve were refined by House4 and Michelson.5

Speech processors were developed and studiesexplored the single-electrode implants’ safetyand elicited useful perception of sound. Thefirst objective, scientific assessment of implantperformance, was carried out by Bilger et al. TheBilger report showed that the implant improvedpatients’ lip reading scores, speech production,and quality of life.6

In the 1980s, the age criteria for use of theimplant lowered and several hundred childrenbenefited from this device. During the sameperiod, a multi-channel cochlear device wasdesigned and surgically implanted by GrahamClark, an Australian Otolaryngologist.7

In the years that followed, innovations inimplant technology resulted in production ofcommercially viable multi-electrode cochlearprosthesis with sophisticated, yet smaller pro-cessors and lower surgical risks.

Over the past three decades about 100,000adults and children worldwide have receivedthese devices.


133

The research and advancement of this sig-nificant technological breakthrough con-tinues. In the future, cochlear implants mayeven be able to restore or provide a normalhearing to deaf individuals. At the same time,cochlear implants might become obsolete ifresearch in molecular and gene therapy suc-ceed to stimulate the growth of new sensoryhair cells.

Cochlear Implant Services

Cochlear implant surgery requires a multidisci-plinary approach. There are specialized adultand children implant teams, which consist ofan Otolaryngologist, Audiological Scientist,Audiologist, Clinical Psychologist, and a Speechand Language Therapist. For adults, the teamincludes a Hearing Therapist, and for children,a Teacher of the Deaf and Specialist CommunityPediatrician. A successful outcome requires col-laboration from patients, families, schools, andthe implant team.

The implant teams are committed to pro-vide a high-quality and comprehensive ser-vices including extensive pre-operative

assessment and counseling, time consumingand intensive post-operative speech andhearing therapy, and follow up for adjustingand setting the device. Appreciation andrespect to different social, cultural, and lin-guistic backgrounds make these servicesaccessible to all candidates.

Common Indications for CochlearImplant

• Severe to profound sensory neural hearingloss in both ears with a functioning auditorynerve.

• Little to no benefit from conventional hearingaids.

• Living in or desiring to live in the ‘‘hearingworld’’.

• High motivation, strong commitment. andrealistic expectations.

• Strong social and educational support.

• Especially in the case of infants and youngchildren having a family willing to worktoward speech and language skills withtherapy.

Figure 8-1. A cochlear implant (with kind permission of Mrs. Raha Ansari).


Cochlear Implant Suitability

Cochlear implant surgery may be beneficial tochildren with congenital and or acquired deaf-ness, young, or middle-aged adults with hearingloss due to genetic causes, autoimmune disease,or unknown reasons; and older adults with pro-gressive hearing loss due to aging or noiseexposure.

The outcome is more favorable if an implantis fitted at a young age or soon after an acquiredhearing loss. This suggests that such a responsecan be at least partially attributed to plasticitywithin the auditory system.

However, the suitability of a patient forcochlear implantation can only be evaluated onan individual basis, taking into account apatient’s hearing history, cause of hearing loss,amount of residual hearing, speech recognitionability, general health status, and family com-mitment to the intensive rehabilitation and edu-cational process as well as follow-up testing andmonitoring of the cochlear implant. The resultsof these comprehensive assessments allow theclinicians to advise candidates on the benefitsthey may gain from the device.

Audiological evaluation includes varioustests such as determination of unaided air andbone conduction thresholds, speech discrimina-tion, and auto acoustic emissions.

Auditory brain steam response (ABR) is anelectrophysiological testing, particularly impor-tant when testing young children in order to ruleout the possibility of functional deafness.

Pre-operative Speech and Language evalua-tion demonstrates developmental language andcommunicative status. This helps in order todefine inappropriate expectations for speechand language skills following the interventionwith a cochlear implant.

Psychological evaluation is probably reallyperformed with pediatric patients but may beused in adults who present with concernsregarding cognitive status. Psychological eva-luation determines any other factors whichmay also be hindering the child’s auditory devel-opment. Accordingly appropriate help andcounseling can be provided to the parents.

Computed tomography (CT) of the temporalbone can identify any cochlear anomalies andassess whether it is possible to insert a cochlearimplant electrode into the cochlea. The informa-tion can also be used in order to choose the most

appropriate ear for implantation and the surgi-cal approach. Magnetic resonance image (MRI)examines the soft tissues carefully to establish ifthe auditory nerve is structurally sound.

Medical evaluation is also necessary to assesspatient’s general health and suitability for thecochlear implant surgery.

Ethical Issues

Cochlear implants for congenitally deaf childrenare often considered to be most effective whenimplanted at a young age; hence they areimplanted before the recipients can decide forthemselves.

Much of the strongest objection to cochlearimplants has come from the deaf community,which consists largely of pre-lingually deaf peo-ple whose first language is sign language. Indi-viduals who are deaf and the deaf community donot consider deafness as a disability and indeedcelebrate their deaf culture like all languages do.However, there has been a greater acceptance ofcochlear implant technology by the deaf com-munities over the last decade.

Cochlear Implant Surgery

The device is surgically implanted under generalanesthesia, and the operation usually takes 1–3 h.

First a small area of hair behind the ear isshaved and an incision is made on the skinaround the back the ear. Some of the mastoidbone positioned behind the ear is drilled, withthe aid of a surgical microscope, to get into themiddle ear through which the cochlea, located inthe inner ear, is accessed.

A few millimeters of the bone behind thewound is removed to seat the cochlear device.This is reduced the bump felt under the skin.

Having placed the device, a cochleostomy, asmall opening into the cochlea is created; andthe electrode is delicately threaded into it.

Prior to closure of the wound, an audiologicalphysicist tests the implant’s function.

At the end of the operation, a dressing isapplied. Often the patient is kept in the hospitalone night for observation.

An x-ray from the site of the operation will betaken within 1 week of the surgery to check theposition of the cochlear implant’s electrode.

Cochlear Implant 135

Initial fitting takes place approximately4–6 weeks after the surgery when the woundhas fully healed and the device is switched onfor the very first time. Adjustments will berequired in the subsequent appointments.

Complications of Cochlear Implant

Cochlear implant surgery is a safe procedure andserious complications are rare. This is due to newerand safer cochlear devices as well as improvedsurgical techniques. The pre-operative immuniza-tion against pnemococcal meningitis reduces therisk of this rare complication even further.

In addition to complications associated withthe general anesthesia, there are risks related tothe surgery itself, include bleeding, infection, tin-nitus, poor hearing outcome, temporary dizzi-ness secondary to the vestibular system damage,numbness in the area of the scar, taste distur-bance due to chorda tympani injury, and damageto the facial nerve that can cause muscle weak-ness, or, in worst cases, disfiguring paralysis.

There are also risks associated with the implant,such as mechanical or electrical failure and rejec-

tion. These problems may require removal orreplacement of the implant.

References

1. Djourno A, Eyries C, Vallancien B.De l’excitation electri-que du nerf cochleaire chez l’homme, par induction adistance, a l’aide d’un micro-bobinage inclus a demeure.CR Soc Biol (Paris). 1957;151:423–425.

2. Doyle JH, Doyle JB Jr,Turnbull. Electrical stimulation ofeight cranial nerve. Arch Otolaryngol. October 1964;80:388–91.

3. Simmons FB. Electrical stimulation of the auditory nervein man. Arch Otolaryngol. July 1966;84(1):2–54.

4. House WF, Urban J. Long term results of electrodeimplantation and electronic stimulation of the cochlea inman. Ann Otol Rhinol Laryngol. July–August 1973;82(4):504–17.

5. Michelson RP, Merzenich MM, Schindler RA, SchindlerDN. Present status and future development of the cochlearprosthesis. Ann Otol Rhinol Laryngol. July–August1975;84(4 Pt 1):494–8.

6. Bilger RC, Black FO. Auditory prostheses in perspective.Ann Otol Rhinol Laryngol Suppl. 1977;86(3Pt. 2 Suppl.38):3–10.

7. Clark GM, Patrick JF, Bailey Q. A cochlear implant roundwindow electrode array. J Laryngol Otol. Febraury1979;93(2):107–9.


9Stem Cells and Organ Replacement

Natasa Levicar, Ioannis Dimarakis, Catherine Flores, Evangelia I Prodromidi, Myrtle Y Gordonand Nagy A Habib

Introduction

Significant research activities in tissueengineering or regenerative medicine (theterm recently used) field started in the 1970sand there is currently a great excitement overthe possibility of replacing damaged bodyparts through regenerative medicine. Potentialstrategies to replace repair and restore thefunction of the damaged tissues or organsinclude stem cell transplantation, transplanta-tion of tissues engineered in the laboratory, andthe induction of regeneration by the body’sown cells. It is believed that novel cellular ther-apeutics can perform better than any medicaldevice, recombinant protein, or chemical com-pound. Possible candidate cells to be usedinclude autologous primary cells, cell lines,and various stem cells including bone marrow(BM) stem cells, cord blood stem cells, fetalcells, and embryonic stem (ES) cells. Stemcells are defined by their capacity for self-renewal and multilineage differentiation, mak-ing them uniquely situated as a powerful tool totreat a wide variety of diseases. In recent years,advances in stem cell biology, includingembryonic and adult stem cells, have madethe prospect of tissue regeneration a potentialclinical reality and several studies have shownthe great promise that stem cells hold for ther-apy.1,2 In this review, we will discuss the use ofembryonic and adult stem cells in treatment forliver and kidney failure, diabetes, myocardialinfarction, and neuronal disorders.

Embryonic Stem Cells

ES cells are isolated from the inner cell mass of5-day-old blastocysts.3 They represent a poten-tial source of cells with remarkable properties:practically unlimited self-renewal and ability todifferentiate into many specialized cell types.4

They are pluripotent and able to give rise tocells belonging to all three germ layers: ecto-derm, endoderm, and mesoderm.5 Neuronsand skin have been formed from ES cells, indi-cating ectodermal differentiation.5,6 ES cells areable to differentiate to cardiac cells, bone, carti-lage, and endothelial cells indicating mesoder-mal differentiation.7–9 Moreover, they are able togive rise to pancreatic cells, indicating endoder-mal differentiation.10 However, many existingapproved ES cell lines have been cultured onmouse feeder cell layers, which can supplymany needed growth factors. This has exposedthe human ES cells to potential risk of viraltransfer between species and has limited theirclinical potential.11 Another major limiting fac-tor for their usefulness in clinical therapy lies intheir tumorigenicity when introduced in vivo.The injected ES cells formed teratomas wheninjected subcutaneously in NOD/SCID mice.5

The ability of these cells to potentially differ-entiate into any cell type has generated greathopes for regenerative medicine. However,since their initial isolation, research on thesecells has been hampered or banned in somecountries because of ethical concerns aboutdestroying human embryos, and therefore


137

human life, to obtain them.12 The ethical debatenow focuses on defining the stage at which life isconsidered to begin and from what point theembryo should be protected. These obstacleshave generated the search for the alternativesources of cells such as adult stem cells.

Adult Stem Cells

Like ES cells, adult stem cells possess self-renewalcapacity and multilineage differentiation poten-tial.13 Adult stem cells are present in approxi-mately 1–2% of the total cell population within aspecific tissue and are vital in the maintenance oflocal tissue homeostasis by continuously contri-buting to tissue regeneration, replacing cells lostduring apoptosis or direct injury. Adult stem cellsare usually quiescent and are held in an undiffer-entiated state within their niche until they receivea stimulus to differentiate.14 More recent devel-opments have proved that adult stem cells residein nearly every tissue, including the BM, brain,digestive system, skin, retina, muscles, pancreas,and liver.15 However, there is a controversy aboutwhether cells isolated from a particular tissueoriginated in that tissue or are circulating BMstem cells.16,17

The best characterized and most widelyunderstood adult stem cells are hematopoieticstem cells (HSC), which sustain the formation ofthe blood and immune systems throughoutlife and were first identified in 1961.18 The BMcompartment is largely made up of HSC andcommitted progenitor cells, non-circulatingstromal cells (called mesenchymal stem cells(MSC)) that have the ability to develop intomesenchymal lineages.19,20 It was previouslythought that adult stem cells were lineagerestricted, but recent studies demonstrated thatBM-derived progenitors in addition to hemato-poiesis also participate in regeneration ofischemic myocardium,21 damaged skeletalmuscle22 and neurogenesis.23 MSC can be iso-lated as a growing adherent cell population andcan differentiate into osteoblasts, adipocytes,and chondrocytes.20

Renal Failure

The kidney is a complex organ with filtration,reabsorptive, and secretory functions as well asendocrine/metabolic activity. The importance of

the kidney as a life-sustaining organ is high-lighted in situations in which kidney functionis compromised or even lost. Without any func-tioning kidney, death can occur within a fewdays. Impaired renal function includes abnorm-alities of both glomerular filtration and renaltubular reabsorption and secretion. Tubularfunction tends to be mainly disrupted by meta-bolic insults to the tubular cells (for example,ischemia or toxins) and could result in tubuloin-terstitial fibrosis and eventually acute renalfailure (ARF). ARF is quite common and affectsup to 7% of all hospitalized patients, with amortality rate that ranges from 20 to 70%depending on the studied population.24 Glomer-ular function can be disrupted by diseases thatalter glomerular structural arrangements (forexample, structural damage to glomerular base-ment membrane, endothelium, epithelium, ormesangium) and ultimately lead to chronicrenal failure (CRF). Chronic injury of the kidneyis responsible for the majority of cases of end-stage renal disease. The means by which kidneyfunction can be replaced in humans includedialysis and renal allogeneic transplantation.Dialysis is life saving but often poorly tolerated.Transplantation of human kidneys is limited bythe availability of donor organs. In addition,mortality rates still remain high despite the exis-tence of these types of treatment. Therefore,there is clearly need for new replacement thera-pies for kidney patients.

Stem Cell Therapy for Kidney Diseases

The potential of human ES cells to generatemesodermal tissue is encouraging for renalcell differentiation. Indeed, ES cells have thepotential to differentiate into renal progenitorcells. Yamamoto et al.25 recently showed thatmurine ES cells had the potential to give rise tomesonephric ducts and ureteric buds in terato-mas. In contrast, Steenhard and colleagues26

reported 50% integration of undifferentiatedES cells (E12–E13) into the tubules of embryonickidneys without evidence for teratomas.Conversely, they found only rare evidencefor integration of injected ES cells into the glo-merular epithelial tufts that form in organ cul-ture. Kobayashi and co-workers created Wnt-4transformed murine ES cells and showed in vivoand in vitro that these assembled into tubule-like formations and expressed aquaporin-2.27


In addition, cells differentiated from mouse EScells via embryoid bodies have been shown toexpress markers of renal lineages in vitro28 andin vivo.29 In the presence of nephrogenic factors,such as retinoic acid, activin-A, and bone mor-phogenetic protein-7 (BMP-7), ES cells can formrenal epithelial cells that are capable of integrat-ing into a developing kidney with very highefficiency.30

The use of therapeutic cloning for replace-ment of transplantable kidney tissue in vivowas recently demonstrated by Lanza and collea-gues,31 who employed a nuclear transfer techni-que to generate cloned bovine fetuses from anadult animal. Renal cells were then isolated fromtheir E56 cloned fetus, passaged and expandedin vitro and seeded onto a collagen-coated poly-carbonate membrane that was then subcuta-neously implanted into the nuclear donor. Therenal cells self-aggregated into ‘‘renal units’’within the membrane, were subsequently vascu-larized by the genetically identical host, andultimately appeared to produce a concentratedurine-like fluid. No evidence of acute rejectionof these bioartificial renal units was observed.

Growing research and interest in the use ofextra-renal cells for replacing renal structuresand functions has been lately described. Cellsexogenous to the kidney possess the ability todifferentiate into multiple cell lineages andcan become incorporated into a number oforgans as part of the normal processes ofcell turnover or organ repair. The adult BMseems to be the source of extra-renal stemcells. However, it remains unknown whetherBM-derived cells invade the kidney and dif-ferentiate into renal cell types, remain in thekidney and are induced to differentiate fol-lowing tissue injury, or whether they firsthome to the BM and are then recruited tothe damaged kidney. Most studies investigat-ing BM plasticity have used human patientsor experimental animals that have received akidney or BM transplant. However, differentinjury models used, different methods ofdetecting engrafted cells, and different popu-lations of BM transplanted are some of thereasons for controversial observations in thisfield. Moreover, the exact mechanism of BMcontribution to renal tissues is not clear withthe scientific community being dividedbetween transdifferentiation of BM cells tokidney cells or fusion of BM cells with term-inally differentiated renal cells.32

Endothelium

In the kidney, endothelial cells are present inlarge vessels and in the abundant network ofperitubular capillaries. In addition, a lymphaticnetwork, thought to increase with renal damage,is lined by endothelial cells. Highly differen-tiated endothelial cells are also found in theglomerulus. During vascular rejection, donorendothelial cells become the predominantfocus of the immune attack since the blood ves-sels of a transplanted organ are the interfacebetween donor and recipient. Endothelialdamage is also an important feature of chronicallograft nephropathy, the most common causeof long-term graft loss. High numbers of circu-lating endothelial cells are found in renal trans-plant patients, indicating ongoing damage of theendothelium.33 Repair of damaged endotheliumin renal allografts by circulating recipient cellswas reported over 30 years ago.34 Sinclair35 alsoshowed that extensive acute damage of the renalvasculature in an allograft may be repaired byhost cells, while less severely damaged graftscould be restored by proliferating neighboringdonor endothelial cells. This was confirmedmany years later by Lagaaij and co-workers,36

who showed that endothelial cells of the recipi-ent could replace those of the donor in humankidney grafts severely damaged by vascularrejection. The number of recipient cellsdecreased with lesser degrees of tissue damage.Studies of renal transplantation in rats indicatedthat endothelial chimerism could be inducedalso by ischemia or toxicity caused by immuno-suppressive treatment.37 Recently, it was shownin human renal transplants that potential lym-phatic progenitor cells derive from the circula-tion and incorporate into the growing lymphaticvessel.38 Rookmaaker and colleagues39 firstreported that endothelial cells, contributing torenal vessels in a female patient with thromboticmicroangiopathy, were derived from the BM of amale donor. Dekel et al.40 recently showed thattransplantation of adult CD34þ-enriched HSCinto ischemic and growing human kidneys hada role in vasculogenesis.

The origin of the glomerular endothelium intransplanted human kidneys is less clear,although in rodents there is firm evidence thatBM contributes to glomerular endothelium.Cornacchia and co-workers41 showed that glo-merular endothelial progenitor cells werederived from the BM in mice. Furthermore,

Stem Cells and Organ Replacement 139

BM-derived cells were shown to participate inglomerular endothelial repair and contribute tomicrovascular repair after induction of reversi-ble anti-Thy 1.1 glomerulonephritis in a rat allo-geneic BM transplant model.42 Conversely, othergroups did not find any evidence for BM-derivedendothelial cells in rats with or without specificglomerular injury, when they stained engraftedcells with endothelial markers.43,44 Later, it wasshown that intrarenal administration of culture-modified BM mononuclear cells reducedendothelial injury in Thy-1 nephritis, presum-ably due to incorporation of BM-derivedendothelial cells into the glomerular lining andproduction of angiogenic factors.45

In the context of severe irreversible glomeru-lar damage induced by anti-Thy-1.1 antibodyfollowed by unilateral nephrectomy in the rat,BM cell infusion improved renal functionand glomerular hemodynamics and reducedmortality in chimeric animals with this type ofdisease.46,47 In addition, BM-derived cells differ-entiated into glomerular endothelial cells duringglomerular healing after the induction of mouseHabu-snake venom nephritis.48

Interstitium

The renal interstitium consists of extracellularmatrix (ECM), fibroblasts, vascular pericytes,and inflammatory cells. Normally, fibroblastshave a role in wound healing and repair throughthe ECM molecules and other proteins they pro-duce. Activated fibroblasts or myofibroblasts,also express alpha-smooth muscle actin (�-SMA)and intermediate filaments, desmin, and vimen-tin. In the kidney, interstitial fibrosis is a com-mon feature of both tubular and glomerularinjury and high numbers of interstitial myofi-broblasts are usually strong predictors of pro-gressive renal fibrosis and kidney failure.

Fibroblasts may also originate from the BMand migrate via the peripheral blood to populatethe kidney. Recipient circulating smooth muscleprecursor cells have been found in vascular andinterstitial compartments of renal allograftsundergoing chronic rejection.49 Such cellsappear transiently attached to the tubular base-ment membrane in mice after uranyl acetate-induced ARF, suggesting that they might beinvolved in promoting cellular recovery inassociation with monocytes or macrophages.50

Direkze et al.51 found evidence for BM-derived

(myo)fibroblasts in kidneys of sex-mismatchedBM-transplanted mice following paracetamoladministration at a toxic dosage. In a mousemodel of ARF induced by ischemia reperfusion,it was shown that about 6.5% of �-SMA-positiveinterstitial cells came from donor BM, suggestinga role for BM-derived cells in renal interstitialfibrosis.52 Conversely, BM-derived interstitialcells did not make a significant contribution tocollagen I synthesis in mice transplanted with BMfrom transgenic donors, in which the transgenewas under the control of a promoter for the �2chain of collagen I and injured by unilateral ure-teric obstruction.53

Tubular Epithelium

Each part of the tubular system has importantfunctions, which are dependent upon normalcellular function of tubular epithelial cells(TEC). Proximal tubules in the cortex are mainlyinvolved in the selective reabsorption of varioususeful components of the glomerular filtrate tothe bloodstream. The adult tubular epitheliumhas the potential to regenerate following acutedamage induced by ischemic or toxic insults.Cells from outside the kidney, for instance cellsoriginating from the BM, can potentially hometo the injured epithelium and be triggered todifferentiate.

In humans, observations of BM participationin tubular repair are mainly limited to renaltransplantation studies, as very rarely can renalbiopsies of a BM-transplanted patient beobtained. Extra-renal cells possible of BM originhave been shown to participate in tubular regen-eration after ARF in human renal transplants.Two groups have detected between 0.6 and 6.8%and approximately 1%, respectively, of Y chro-mosome-containing tubular cells in kidneys ofmale patients, who received a renal transplantfrom a female patient.54,55 Nishida and collea-gues56 have also reported the contribution of BMcells to renal regeneration in a 7-year-old femalepatient who received a male BM transplant, sug-gesting a clinical application for BM cells.

Sex-mismatched transplantation of wholeBM, predominantly in mice, has generated a lotof controversy over BM plasticity for the kidney.In female recipients of wild-type whole BM,Y chromosome-positive tubular cells expressingepithelial markers have been reported withoutinduction of renal injury.55 Further evidence for


BM contribution to tubular cells has been pro-vided by Fang and co-workers,57 who showedthat about 10% of regenerating tubular cellscome from the BM and are capable of DNAsynthesis after folic acid-induced injury andtreatment with granulocyte-colony stimulatingfactor (G-CSF). On the contrary, Szczypka andcolleagues58 have demonstrated that BM-derived cells rarely contribute to repair of renaltubules in uninjured or folic acid-treated mousekidneys. In addition, inducing renal ischemiareperfusion injury (IRI) in chimeric mice didnot result in tubular regeneration by BM cells,as examined by three separate detectionapproaches.59,60

Morigi and colleagues61 have found that theMSC population has the primary capacity toparticipate in renal tubule repair after toxicinjury, whereas the HSC mediate a lesser protec-tive effect. Similar findings have been reportedby Herrera and his group62 following transplan-tation of GFP-expressing MSC into mice thatwere subjected to glycerol-induced tubularinjury. However, others have found that purifiedMSC did not incorporate into tubules ofischemic mice, although renal function wasimproved, probably due to immunomodulatorymechanisms rather than renewal of tubularcells.59,60,63,64 The role of HSC in renal repairhas also been debated, with some groups report-ing no evidence for HSC repopulating the kidneyusing GFP as a marker, but without inducing anyspecific renal damage.65,66 Conversely, followingHSC transplantation from b-gal transgenicROSA26 mice into wild-type recipients, two stu-dies described the presence of X-gal positivecells in cortex and medulla that stained positivefor epithelial markers after renal IRI.67,68 In oneof these studies, it was also shown that sortedHSC have a functional role in the repair of thekidney as they could partially reverse blood ureain injured chimeric mice.67 By changing thedonor mouse strain from ROSA26 to eGFPtransgenic mice, Lin et al.52 later suggested thattubular regeneration after IRI is predominantlydue to indigenous proliferating renal cells withHSC making only a small contribution.

Glomeruli

Glomerular mesangial cells provide structuralsupport in the glomerular capillary tufts. Themesangial cells are modified smooth muscle

cells and regulate blood flow through theglomerular capillaries by their contractile activ-ity. Mesangial cells may be injured by immuno-logical insults and play a key role in thedevelopment of scarring in many progressiveglomerular diseases, including glomerulone-phritis and diabetic nephropathy. Mesangialcell injury manifested as mesangiolysis, protei-nuria, proliferation, and hypercellularity, if sus-tained, may lead to impaired renal function.Therefore, the repair process following glomer-ular injury needs a well-balanced recruitment ofglomerular endothelial and mesangial cells.

In recent years, the role of BM as a reservoirfor mesangial cells during kidney repair hasbeen well investigated. Mesangial progenitorcells have been shown to derive from the BM ofGFP transgenic donor rats in vivo, and they stainfor desmin and respond to angiotensin II stimu-lation in vitro.43,44 In addition, BM-derived cellshave been shown to incorporate into glomerulias specialized glomerular mesangial cells afterfolic acid injury in mice.58 However, in thesestudies the exact stem cell population in theBM which provides these progenitor cells hasnot been elucidated. Later studies sought todefine the exact origin of the engrafting cells inthe BM and showed that cloned progeny of asingle HSC can transdifferentiate into glomeru-lar mesangial cells, which respond to angioten-sin II.69 In contrast, other groups using a slightlydifferent method of HSC purification failed todetect donor-derived glomerular mesangial cellsin mice transplanted with single HSC.66

In the disease setting, contribution of BM todevelopment and progression of glomerulo-sclerosis has been very elegantly shown in twostudies. Cornacchia and colleagues41 havedemonstrated that mouse mesangial cell pro-genitors originate from the BM and transmitboth the sclerotic lesions and the glomerularhypertrophy of donor kidneys to normal glo-meruli. In diabetic nephropathy, where mesan-gial cell proliferation and mesangial sclerosis arepathologic hallmarks, Zheng et al.70 have shownthat BM-derived mesangial cell progenitorsfrom type II diabetic mice may also transfer thedisease genotype and phenotype into normalrecipients, which develop albuminuria andsevere glomerular lesions.

Recently, several groups have providedevidence for a functional role of BM-derivedmesangial cell progenitors in improvementof renal disease. Guo et al.71 showed that


transplantation of wild-type BM prevents orattenuates progression of mesangial sclerosis inthe Wt1þ/– mouse model of renal disease, mostprobably by replacement of mesangial cellswith BM cells.71 BM transplantation has beenshown to attenuate mesangial cell activation inThy 1.1 nephritic rats45 and reduce mortalityand pathological findings in fatal progressiveglomerulosclerosis.47 However, whether thesecells originated from resident BM cells, and/orinjected BM cells was not determined in thesestudies.

The origin of glomerular epithelium andespecially podocytes has been less well studiedcompared to other parts of the nephron. Podo-cytes form part of the glomerular filtration bar-rier and support the capillary tuft and regulateglomerular filtration. Podocytes also have pha-gocyte-like functions, as they remove any largemolecules trapped in the outer layers of thefilter. They are predominantly terminally differ-entiated cells with little or no capacity ofrenewal.

Poulsom and colleagues55 were the first toobserve that cells located in the periphery of aglomerulus resembled podocytes and stainedwith vimentin and could be BM-derived. How-ever, other specific phenotypic markers were notused to identify them unequivocally. Since then,a few groups supported a small contribution ofBM cells to podocyte regeneration in mice bystaining them with antibodies to a specific podo-cyte marker, WT-1 protein.71 An alternative andpossibly efficient way to use cell therapy in renaldisease may be to condition or otherwise manip-ulate stem cells before their delivery in vivo.

A recent study showed that culturing canineBM-derived cells on plastic, type IV collagensubstrate, which resembles matrix ordinarilyseen by podocytes in vivo, promotes their partialdifferentiation in vitro into a podocyte-like phe-notype.72 Whether these cultured cells couldever become functional podocytes in vivo isunknown, but this in vitro conditioning strategywith stem cells clearly underscores the impor-tance of cell–matrix interrelationships in achiev-ing appropriate phenotypes.

Recent studies have also reported that wholeBM transplantation has a functional effect in amouse model for Alport’s syndrome. Alport’ssyndrome is a progressive disease that ulti-mately leads to renal failure. There is no specifictherapy and it is fatal, if renal replacement ther-apy is not provided once kidneys fail to function.

In the murine model, mice that lack the �3 chainof collagen IV (Col4�3–/–) develop progressiveglomerular damage leading to renal failure. Theproposed mechanism is that podocytes fail tosynthesize normal GBM, so the collagen IV net-work is unstable and easily degraded. Tworecent studies have shown for the first timethat BM cells replace defective podocytes andproduce Col4�3, thereby improving functionaland structural parameters of renal disease inAlport mice.73,74 Although, the lineage of BM-derived cells that were recruited to the damagedglomeruli was not clearly established in thesestudies, preliminary data suggest that BM-derived MSC fail to reverse disease in thismodel,73,75 suggesting that the HSC compart-ment might be exerting the beneficial effectsobserved. Although more experimental dataare needed before BM transplantation can beapplied to patients, these observations providenew hope for treating Alport’s disease by stemcell therapy.

These data demonstrate that BM proves ahighly promising, ethically accepted, and readilyaccessible source of adult stem/progenitor cellsfor the remodeling of injured kidney.

De Novo Kidney Creation

Growing new organs in situ by implanting devel-oping animal organ anlagen/primordia repre-sents a novel solution to the problem of thelimited supply of human donor organs thatoffers advantages relative to transplanting EScells or xenotransplantation of developedorgans. Renal primordia (metanephroi) trans-planted into animal hosts undergoorganogenesis in situ, become vascularized byblood vessels of host origin, and exhibit excre-tory function.76 Metanephric mesenchymal tis-sue harvested from donors at an early stage indevelopment (E13-E15) was transplanted intoneonatal mouse kidney and was shown to createfunctioning nephrons.77,78 Transplantation ofdeveloping metanephroi into adult rats hasresulted in functional chimeric kidneys vascu-larized by the host and producing urine afteranastomosis to the host’s ureter.79 Metanephroican be also preserved prior to transplantationand differentiate in situ before being introducedinto animal hosts to develop into functionalkidneys.80,81 In addition, human metanephroican be induced in vivo to grow and develop


into mature nephrons following transplantationinto mice.82 Moreover, xenotransplanted earlykidney embryonic precursor cells, of bothhuman and pig origin, can form functional min-iature kidneys in mice, although it was notshown whether the urine they produced hadbeen concentrated by the structures formed.83

These studies demonstrate the possibility ofintegrating new filtering nephrons into kidneys.Critical to providing organ function replace-ment through cell therapy is the need for theisolation and growth in vitro of specific cells,such as kidney progenitors, that could poten-tially create new kidney segments or entirenephrons appropriate for transplantation. Avery promising study recently showed thathuman MSC injected into in vitro cultureddeveloping rodent embryos and subjected tofurther organ culture could be reprogrammedto contribute to kidney cell lineages, such aspodocytes and tubular epithelial cells.84

Artificial Kidney (See Chapter 3)

Hemodialysis replaces some of the filtrationfunctions of the kidney but does not recapitulatethe endocrine/metabolic activities of renal cells.To address this deficiency, a bioartificial tubule,which uses cultured epithelial progenitor cellson appropriate membranes and biomatrices thatconfer immunoprotection and long-term func-tional performance, has been developed. Humeset al.85 have created a synthetic hemofiltrationcartridge and a renal tubule assist device (RAD)containing human or porcine cells in an extra-corporeal circuit. Cells are grown as confluentmonolayers along the inner surface of hollowfibers within a standard hemofiltration car-tridge.86,87 The RAD has been used for the treat-ment of acutely uremic dogs and increasesexcretion of ammonia and enhances glutathionemetabolism and 1,25(OH)2D3 production dur-ing 24 h of treatment.

A Phase I/II trial in 10 humans with ARF andmulti-organ failure has been safely conductedand RAD treatment resulted in declines in pro-inflammatory cytokines (G-CSF, IL-10, andIL-6/IL-10 ratios).88 A controlled, randomizedPhase III trial is currently underway. The RADwas recently shown to modify sepsis to improvesurvival in ARF.89 However, existing bioartificialkidneys are large and they use intensive labor,making the miniaturization of the bioartificial

kidney a prerequisite before applying this noveltechnology. The differentiated growth of humantubular epithelial cells on thin-film and nano-structured materials strongly suggests that thisminiaturization will be soon feasible.90,91

Diabetes

Diabetes mellitus affects over 6% of the popula-tion worldwide and the World Health Organiza-tion expects that the number of diabetic patientswill increase to 300 million by 2025 (57). Type Idiabetes is a chronic disease and results fromautoimmune-mediated destruction of insulin-secreting b-cells in the islets of Langerhans ofthe pancreas. Once activated the continueddestruction of b-cells leads to a progressive lossof insulin, then to clinical diabetes, and finally inalmost all affected to a state of absolute insulindeficiency.92 Type II diabetes is due to systemicinsulin resistance and reduced insulin secretionby pancreatic b-cells. Although the etiology oftype II diabetes remains obscure, obesity and asedentary lifestyle are the most common epide-miologic factors associated with development ofthe disease. The absence of insulin is life-threa-tening, thus requiring diabetic patients to takedaily hormone injections. However, insulininjections do not adequately mimic beta cellfunction, which results in the development ofdiabetic complications such as neuropathy,nephropathy, retinopathy, and diverse cardio-vascular disorders. Long-term normalization ofglucose metabolism is a prerequisite for preven-tion of secondary complications and to date hasonly been achieved with transplantation of thewhole organ or with a reasonable number ofislets. However, the chief limitation to trans-plantation, either whole gland or islets, is thepaucity of donors. Islet transplantation is mainlylimited because of the difficulty in obtainingsufficiently large numbers of purified isletsfrom cadaveric donors and treatment can beoffered to only an estimated 0.5% of needy reci-pients.93 Additionally, differentiated b-cells can-not be expanded efficiently in vitro and senescerapidly.94 Several approaches are now beinginvestigated to generate insulin-producing cellseither by genetic engineering of b-cells or byutilizing various b-cell precursor cells, andstem/progenitor cells. Insulin-secreting cellscould be transplanted into patients to helpmaintain blood glucose homeostasis, reduce


the burden of diabetes-related complications,and overcome the limitation of donor organs.

Stem Cell Therapy for Diabetes

ES cells have been proposed as a potential sourcefor cell replacement therapy in the treatment ofdiabetes. Several studies have demonstrated thatby manipulating culture conditions and usinggrowth and transcription factors of the b-celllineage (in particular Pdx-1 and Pax4), ES cellscan differentiate in vitro into insulin-producingcells.10,95,96 Soria et al.97 transplanted an insulin-secreting cell clone from undifferentiated ESinto the spleen of streptozotocin-induced dia-betic animals. Although transplanted animalscorrected hyperglycemias within 1 week, anintraperitoneal glucose tolerance test showed aslower recovery in transplanted versus controlmice. Fujikawa et al.98 cultured and differen-tiated embryonic cells into insulin and c-peptidepositive cells. When the cultured cells weretransplanted into diabetic mice, they reversedthe hyperglycemic state for approximately3 weeks, but the rescue failed due to immatureteratoma formation. An encapsulated solidtumor was found at each transplanted site in 6of 10 transplanted mice.

The risk of teratoma formation would need tobe eliminated before ES cell-based therapies forthe treatment of diabetes are considered. Pro-mising data regarding the potential of BM stemcells to reconstitute the b-cell endocrine portionof the pancreas has been produced by Ianus andco-workers.99 Irradiated female wild-type micewere injected with BM stem cells expressingenhanced green fluorescent protein (EGFP)under the control of the murine insulin promo-ter. Up to 3% of total cells per islet were found toexpress EGFP at 4–6 weeks post-transplantation.Further RT-PCR analysis of sorted cells showedthe expression of b-cell markers including insu-lin I, insulin II, GLUT2, IPF-1, HNF1�, HNF1b,HNF3b, and Pax-6. Finally, demonstration ofglucose-dependent and incretin-enhanced insu-lin secretion was reported as proof of function-ality. However, another group using a similarstrategy failed to show any GFP positive cellswithin the pancreatic parenchyma of trans-planted animals.100 As this was not informativeper se of pancreatic engraftment, another seriesof experiments was conducted using BMDS cellsexpressing GFP under the control b-actin

promoter. Despite the large degree of engraft-ment none of the GFP positive cells co-expressedinsulin or the b-cell transcription factors Pdx-1or Nkx6.1, while >99.9% expressed the pan-hematopoietic marker CD45 as well as myeloidantigens. The authors concluded that althoughBM stem cells demonstrated efficient pancreaticengraftment, a hematopoietic cell fate wasalmost exclusively retained. Hess et al.101 usedmurine model of streptozotocin-induced pan-creatic damage to induce hyperglycemia andtransplanted the mice with BM-derived stemcells expressing GFP. They observed that pan-creatic injury was a prerequisite for BM stemcells taking on an insulin-producing phenotypewithin the pancreatic parenchyma, an observa-tion also made by other researchers.102 Theyhave detected only up to 2.5% donor-derivedinsulin positive cells. Moreover there was noexpression of Pdx-1 in these cells. Also, a largeproportion of donor cells documented in ductalor islet regions were of endothelial lineage, thusassociating the regenerative process with var-ious endothelial interactions. Similarly, Lechneret al.103 failed to detect donor-derived cells inanimals with induced pancreatic damage andtreated with BM stem cells. Kang et al.104 showedthat HSC transplantation prevents diabetes inNOD mice but does not contribute to significantislet cell regeneration once disease was estab-lished. Successful BM engraftment before dia-betes onset prevented disease in all mice for1 year after transplantation. However, despiteobtaining full hematopoietic engraftment inover 50 transplanted mice, only one mousebecame insulin independent. However, BMstem cells transplanted into the renal capsuleand the distal tip of the spleen of streptozoto-cin-diabetic mice reversed the existing hyper-glycemia and the animals’ ability to respond toin vivo glucose challenges. Furthermore,Banerjee et al.105 showed that multiple injectionsof BM stem cells improved glycemic control instreptozotocin-treated mice. Ende et al.106 trans-planted human umbilical cord blood mononuc-lear cells into non-obese diabetic mice withautoimmune type I diabetes were able to reduceblood glucose levels and improve survival com-pared to untreated animals.

Tang et al.107 pre-differentiated BM stem cellsinto insulin-producing cells, transplanted themin streptozotocin-induced diabetic mice, andachieved reversal of hyperglycemia andimproved metabolic profiles in response to


intraperitoneal glucose tolerance testing. Ohet al.107,108 transplanted BM stem cells transdif-ferentiated into insulin-producing cell aggre-gates and showed lowered circulating bloodglucose levels and maintained comparativelynormal glucose levels in mice for up to 90 dayspost-transplantation. These studies indicatedthat pre-transplantation in vitro cell manipula-tion may provide a useful tool for deliveringlarge numbers of predefined cells, thus avoidingcomplications such as suboptimal delivery ratesand cell differentiation down non-pancreaticpathways.

Clinical Studies

The first reported data on cellular therapy fortype I and II diabetes were presented by Fernan-dez Vina et al.109,110. They have treated 23patients with type I diabetes with CD34þCD38–

cells isolated from BM and followed them up for90 days. After 90 days from cell transplantation,the blood sugar decreased by 9.7% and c-peptidesignificantly increased by 55%. It was alsoobserved that the requirement for exogenousinsulin, taken daily by the patients, decreasedby 17% suggesting that autologous BM stemcells could improve pancreatic function. Similarresults were obtained for type II diabetespatients where autologous CD34þCD38– cellswere transplanted via spleen artery into 16patients. Ninety days post-transplantation theblood sugar significantly decreased by 27%,while c-peptide and insulin increased by 26and 19%, respectively. Even more impressive isthe fact that 90 days post-transplantation, 84%of treated patients did not need any more antidiabetic drugs or insulin.

Liver Diseases

Liver diseases impose a heavy burden on societyand affect approximately 17% of the popula-tion.111 The main causes of cirrhosis, the endresult of long-term liver damage, are hepatitisB and C and alcohol abuse. At the cirrhotic stage,liver disease is considered irreversible and theonly solution is orthotopic liver transplantation(OLT). However, the increasing incidence ofliver disease, the widening donor–recipientgap, and the poor outcome in patients not sup-ported by liver transplantation mean that there

is obviously a demand for new strategies to sup-plement OLT.

New strategies such as cell therapy are underinvestigation. Potential sources include theexpansion of existing hepatocytes, ES cells, pro-genitor/stem cells in the liver, and BM stem cells.Hepatocyte transplantation has been accom-plished in small number of patients with meta-bolic deficiencies and acquired liver disease.112–114

The major limiting factor is the inability to pro-duce large quantity of hepatocytes and to keepthem ready for use on demand, short duration ofthe functioning replacement cells, and the needfor immunosuppression.115,116 These limitationsgenerated the interest in stem cells as a stableand expandable source, expected to provide largenumbers of cells for hepatic disorders.

Stem Cell Therapy for Liver Diseases

Several studies have shown in vitro hepatic dif-ferentiation of ES cells, showing expression ofalpha-feto protein (AFP), albumin, alpha-1-antitrypsin, and transthyretin.117,118 Co-culture with Thy1-positive cells in vitro inducedthe maturation of AFP-producing cells isolatedfrom ES cell cultures into hepatocytes.119 More-over, AFP-positive cells isolated from culturedES cells differentiated into hepatocytes whentransplanted into livers of apolipoprotein-E-(ApoE) or haptoglobin-deficient mice.120

Yamada et al.121 and Chinzei et al.122 demon-strated that ES cells growing in embryoid bodiesexpress mRNAs for albumin, AFP, and othermature hepatocyte markers. They also incorpo-rate into hepatic plates, produce albumin, andmorphologically resemble adjacent hepatocyteswhen transplanted into mice.

More work has been done on adult extrahe-patic stem cells such as HSC. Petersen et al.123

transplanted male BM stem cells into injuredlivers of female mice. They demonstrated thatregenerated hepatic cells were of BM origin byusing Y chromosome, dipeptidyl peptidase IVenzyme, and L21-6 antigen as markers to iden-tify donor-derived cells. Grompe et al.124 usedFAH(–/–) mice, an inducible animal model oftyrosinemia type I, a lethal hereditary liver dis-ease and showed repopulation of injured liver bydonor-derived BM cells. Twenty-two weeks aftertransplantation one-third of the liver comprisedBM-derived cells, suggesting that BM stem cellscontribute to hepatocyte generation, when


regenerative potential of hepatocytes isimpaired. Sakaida et al.125 showed that BMstem cells can reduce liver fibrosis, probably byexpressing matrix metalloproteases, whichenable degradation of hepatic scars. Malletet al.126 induced hepatic apoptosis in mice byJO2 antibody, the murine anti-Fas agonist andtransplanted unfractionated BM cells expressingBcl-2 under the control of a liver-specific pro-moter. BM-derived hepatocytes expressing Bcl-2were only seen in the liver of the mice, whichreceived JO2 antibody injections. Moreover, inmice with induced liver cirrhosis, 25% of therecipient liver was repopulated in 4 weeks byBM-derived hepatocytes.127 Similarly, murineHSC converted into viable hepatocytes withincreasing liver injury and restored liver func-tion 2–7 days after transplantation, suggestingthat HSC contribute to liver regeneration bydifferentiating into functional hepatocytes.128.However, some studies have shown that BMstem cells can repopulate liver even in theabsence of liver injury. Theise et al.129 identifiedup to 2.2% donor-derived hepatocytes whenthey transplanted BM or CD34þlin– cells intoirradiated mice without acute liver injury. HSCtransplanted into irradiated mice engrafted inseveral organs, including liver, gastrointestinaltract, bronchus and skin of recipient animals,and generated albumin-expressing hepatocyte-like cells.130

In contrast, several other studies failed toshow the contribution of BM stem cells to liverregeneration. HSC reconstituted blood leuko-cytes of irradiated mice, but did not contributeto non-hematopoietic tissues, including liver,brain, kidney, gut, and muscle.66 Several groupsfailed to show a significant contribution of BMstem cells to liver regeneration using variousliver injury models.131 Although the contribu-tion of HSC to hepatocyte lineages in vivo stillremains divisive, the differences between thestudies may in part reflect the types of cellsused, different injury models used and themethod used to detect engrafted stem cells.

Several studies have also shown the presenceof cells of BM origin in the human liver. Alisonet al.132 examined livers from female patientsthat received BM transplantation and founddonor-derived hepatocytes, suggesting thatextrahepatic stem cells can colonize the liver.Theise et al.133 identified hepatocytes and cho-langiocytes of BM origin in archival autopsy andbiopsy liver specimens. Using double-staining

analysis, they found a large number of engraftedhepatocytes (4–43%) and cholangiocytes(4–38%), derived from extrahepatic circulatingstem cells, suggesting BM stem cells can replen-ish large numbers of hepatic parenchymal cells.However, in a similar study, Korbling et al.134

found only 4–7% BM-derived hepatocytes. Ng etal.135 identified only small proportion of donor-derived hepatocytes (1.6%) in liver allografts,most donor-derived cells were macrophages/Kupffer cells. Two other studies did not detectany BM-derived hepatocytes at all.136,137 Thedifferences in the published studies could bedue to use of different techniques to identifyrecipient derived hepatocytes in transplantedpatients. Also, various markers can be used forhepatocyte identification and the accuracy of themethods used for identification is variable.

Clinical Studies

Recent success in a few studies using stem celltreatment for liver diseases has raised hopes forthe patients that cell therapy may be feasibletreatment for their disease. Only three clinicalstudies using BM stem cells for liver disease havebeen published so far. In the first study, threepatients with large central hepatobiliary malig-nancies were treated with autologous CD133þ

cells.138 CD133þ cells highly enriched from auto-logous BM were selectively implanted to the left-lateral portal branches subsequent to selectiveportal vein embolization of right liver segments.Mean daily hepatic growth determined by CTscan volumetry rates was 2.5-fold higher com-pared to the patients that had been subjected toportal vein embolization without CD133þ appli-cation. These data suggested that stem cell ther-apy enhances and accelerates liver regenerationand may bear the potential for augmentation ofliver regeneration before extensive hepatect-omy. The second study using adult HSC in treat-ment of liver diseases was performed by ourgroup.139 We treated five patients with liverinsufficiency and mobilized their stem cells byG-CSF. Between 1� 106 and 2� 108 CD34þ cellswere injected into the either portal vein or hepa-tic artery. Three of the five patients showedimprovement in serum bilirubin and four offive in serum albumin. Clinically, the procedurewas well tolerated with no observed procedure-related complications. The data suggested thatstem cells contributed to the regeneration of


liver damage and are encouraging for the futuredevelopment of stem cell therapy for liver dis-eases. In the third study, nine liver cirrhosispatients were treated with autologous BM.140

Mononuclear cells (CD34þ, CD45þ, c-kitþ)were infused via the peripheral vein the totalnumber of the infused cells was 5.20þ/–0.63 �109. They reported of significant improvementin serum albumin levels, total protein, andimproved Child–Pugh score.

Myocardial Infarction/Heart Disease

Coronary artery disease with its clinical sequelaecontinues to be one of the leading causes ofcongestive heart failure. Following myocardialinfarction, cardiomyocyte death and segmentalscarring eventually contribute to the develop-ment of replacement fibrosis. The impact ofmyocardial fibrosis on the remaining healthymyocardium may range from subclinical to det-rimental. The remodeling process initiated willeventually lead to impairment of left ventricularfunction. In an attempt to reverse the naturalprogress of the disease, several alternative novelapproaches are being considered. These includecellular transplantation in which de novo intro-duced cells are expected to regenerate/repairareas of diseased myocardium.

Stem Cell Therapy for IschemicHeart Diseases

The main types of stem cells currently beinginvestigated include embryonic, umbilical cordblood, and adult stem cells. Undebatable prooffor the ability of ES cells to differentiate intocardiomyocytes cannot be other than the spon-taneous contraction observed in embryoidbodies.8 Besides spontaneous ‘‘beating’’, cellsfrom these areas exhibit typical immunopheno-typical, molecular, and electrophysiologicalproperties of cardiomyocytes.8,141 Followingintramyocardial transplantation into rodentmodels of myocardial infarction, engraftmentwith subsequent cardiac-lineage differentiationhas been reported by many groups; andsustained improvement in cardiac function wasalso noted.142,143 Recent data also suggest thatES cells are capable of homing to myocardialinjury following peripheral delivery by being

chemoattracted to locally released cytokines.144

Although a high degree of in vitroelectromechanical coupling has been documen-ted for ES cells and cardiomyocytes,145 there arestill concerns regarding underlying arrhythmo-genic potential.146

Although the totipotent/pluripotent ES cellsmay seem to be the ideal candidate, their clinicalapplication is not foreseeable in the near future.Serious ethical considerations remain the maindrawback; other issues complicating clinicaladaptation include insufficient availability,associated disadvantages of allograft transplan-tation, and unpredictable electrical behavior aswell as the risk of tumor formation. The lack ofthese side effects probably with the exception ofarrhythmogenicity has allowed the majority ofavailable adult stem cell types to enter the clin-ical trial setting.

Extensive preclinical work in small and largeanimal infarction models preceded the intro-duction of adult stem cell therapy in the experi-mental clinical setting. This ongoing body ofresearch established feasibility and providedthe necessary support via demonstration ofmyocardial regeneration/repair and functionalimprovement.147,148 The use of crude BM versuspre-selected stem cell populations remains con-tentious, the main argument being synergy ver-sus specification, respectively. By definitioncrude preparations are less concentrated inpotent stem cells and encompass a variety ofdifferent cell types. This in turn may exposesubjects to unjustifiable procedure-associatedrisks with potentially unwanted sideeffects.149,150 On the other hand, pre-selectedpopulations may require to undergo multiplepopulation doublings in vitro prior to deliveryin order to achieve sufficient cell counts.151 Pro-longed passaging in culture is not without risksince replicative senescence,152 changes in mul-tipotentiality,153 and spontaneous transforma-tion154 have all been associated with long-termin vitro culture. Finally, data concerning thearrythmogenic potential of transplanted MSChas emerged from in vitro experimental work.155

The sole assay available at present for identi-fication of HSC remains the reconstitution of thehemopoietic system of a myeloablated host. Sur-face markers such as CD34 or the more primitiveCD133 help in distinguishing subpopulationsenriched in HSC in humans; these include theCD34þCD38– cell population as well as the sidepopulation. The latter cell type may reside in


various organs besides the BM and there is con-flicting data as to the origin of these cells withsome evidence supporting the fact of them actu-ally being BM-derived HSC rather than tissuespecific.156,157 As murine progenitor cells donot express the same surface markers, lineagedepletion (Lin–) is the main criterion to identifyenriched populations.158 In Orlic’s landmarkpaper,21 transplanted Lin–c-kitþ were demon-strated to form new myocytes, endothelial cells,and smooth muscle cells, leading thus to de novomyocardial regeneration. Similar data have alsobeen produced with human peripheral bloodCD34þ cells.159 Nonetheless, controversyremains as to whether HSC are capable of trans-differentiation into cardiomyocytes160,161 oreven confer any functional improvementwhatsoever.162

Endothelial progenitor cells (EPC) or angio-blasts and HSC are thought to share a commonprecursor, the hemangioblast.163 Studies haveshown the connection between angioblastsresiding within the BM and neovasculariza-tion.164,165 EPC are recruited to sites of ischemicinjury and promote angiogenesis. Fazel and col-leagues166 suggested that this is achieved viaregulation of the myocardial levels of variousangiogenic cytokines such as the angiopoietins1 and 2 and vascular endothelial growth factor(VEGF). Absolute numbers and functional activ-ity of EPC appear to be inversely correlated withrisk factors for coronary artery disease;167,168 anissue to consider for the design of clinical studiesif one bears in mind target populationcharacteristics.

MSC have two main characteristics: the abil-ity to adhere to culture dishes as well as todifferentiate into a variety of tissues under theappropriate conditioning. In addition, theyappear uniformly negative for typical ‘‘hemato-poietic’’ surface markers. Apart from innateplasticity, MSC may exhibit immunomodulatoryeffects169 that may allow for allogeneic in vivotransplantation with minimal risk of immunerejection.170 Possible co-delivery of MSC withBM-derived mononuclear cells in order to max-imize clinical effect has also been proposed.148

A very appealing adult BM-derived popula-tion of high plasticity was reported by the Min-nesota group.171 These cells which are known asmultipotent adult progenitor cells (MAPCs)were isolated from MSC cultures and maintainthe ability to differentiate in vitro in cells of thethree germ layers. Only recently another group

has reported the identification of a similar sub-population from adult human BM.151 Intramyo-cardial transplantation of these cells in a rodentmyocardial infarction model lead to in vivo dif-ferentiation into multiple lineages including car-diac, endothelial, and smooth muscle cellphenotypes.

Clinical Studies

From the clinical standpoint, a plethora of stu-dies have been published incorporating BM –and blood-derived stem cells. Design heteroge-neity (cell type, number of delivered cells, etc.)along with the fact that majority of studies con-sist of small cohorts with short follow-up war-rant cautious interpretation of the results. Theroute of administration may also vary depend-ing mostly on the underlying disease. Techni-ques involving transepicardial, intravascular(both intracoronary and intravenous), andtransendocardial delivery have been developedand are in clinical use. Although the majority oflarge randomized studies demonstrate improve-ment in left ventricular ejection fraction, onemust notice that two of these show minimal tono clinical benefit.

Mechanisms of Action

The exact mechanisms that govern the func-tional improvement observed in animal andhuman studies remain a highly debatableissue. The inability for routine pathologicalexamination of specimens in the clinical trialsetting necessitates data collection from experi-mental models to investigate this. The initial –and to a certain extent simplistic– consensuswas to attribute all benefit to the innate plasti-city of transplanted stem cells. Cells committedto cardiomyogenic lineages have been consid-ered to be produced from delivered cells viatransdifferentiation,21,172 fusion with residentend-differentiated cells,173 or even a combina-tion of both mechanisms. 174 Although bothtransdifferentiation and fusion provide anacceptable theoretical platform at the cellularlevel for the documented functional recovery,it is by now widely accepted that their collectiveclinical effect is probably negligible due to theirdiminutive frequency of occurrence.


A very interesting notion is that transplantedcells may promote restoration of resident car-diac stem cell niches. This falls under the widersuggestion that stem cells exert their beneficialeffect via paracrine pathways. In a rat model ofmyocardial infarction, transplantation of MSCinduced greater expression of proangiogenicand homing cytokines while reducing theexpression of the proapoptotic protein Bax com-pared with medium-treated control hearts.175

Stem cell recruitment in a murine model ofmyocardial infarction was also shown to estab-lish a proangiogenic milieu in the infarct borderzone leading neoangiogenesis and the formationof an extensive myofibroblast-rich repair tis-sue.166 By providing improvement in localblood supply,176 in addition to further recruit-ment of resident and peripheral stem cells to thesite of injury via paracrine factor secretion,recovery of the local microenvironment is sup-ported. An interaction also exists with the ECMleading to the induction of reverse remodeling.

Stem Cells and Cardiac-RelatedBioengineering

Steady advancements in the field of pediatriccardiac surgery have increased the demand forexogenous myocardial ‘‘building blocks’’, mainlyin the form of right ventricular outflow tractconduits as well as transannular patches. Syn-thetic biomaterials, heterologous glutaralde-hyde-preserved bovine pericardium/jugularvein, and glutaraldehyde-preserved homograftscurrently represent the mainstay of optionsavailable to surgeons. Several limitations renderthese implants far from ideal; inability to followthe recipients’ growth may limit considerablyconduit life span and make repeated surgerynecessary. From a physiological point of view,failure of electromechanical integration deprivesthe pulmonary arterial circulation of pulsatileblood flow. Similarly, both mechanical andavailable biological valves (xenografts or allo-grafts) are linked with a need for life-long antic-oagulation associated morbidity and limiteddurability respectively.

While production of a total bioengineeredheart remains wishful thinking, many researchgroups are addressing the construction of viablemyocardial patches. The aim remains a viableautologous construct that may contract, achieve

electromechanical coupling, and integratewithin the native vascular network. Althoughencouraging data utilizing fetal or neonatalcardiomyocytes177–179 are being gradually pro-duced, the nature of these cells probably pre-cludes their transition to the clinical level.Cardiomyocytes are difficult to obtain and havelimited capacity for proliferation, with such con-structs not reflecting the actual cellular organi-zation of the myocardial wall. On the other handstem cells are easily obtained and expanded aswell as having by definition the capacity to dif-ferentiate into a variety of tissue types includingendothelial, smooth muscle, and nerve cells.After seeding BrdU-labeled MSC into porousacellular bovine pericardium, Wei et al.180 pro-ceeded to repair a surgically created rightventricular myocardial defect in a rat model.Both epicardial and endocardial surfaces wereneo-mesothelialized and neo-endothelialized,respectively, avoiding thus the formation ofadhesions and thrombi. Evidence of tissueregeneration (neo-muscle fibers, neo-capillaries, and smooth muscle cells) within theMSC patch was documented with many of thesecells staining positive for BrdU. Unfortunately,the MSC patch appeared akinetic via echocar-diography, possibly due to an inadequate num-ber of cardiomyocytes present within the patch.

Heart valve tissue engineering incorporates(autologous) cell seeding onto biodegradablescaffolds, anticipating gradual replacement ofthe scaffold by cells and matrix produced by theimplanted cells. Proof of principal was first pub-lished more than a decade ago by seeding a poly-glycolic acid scaffold with a mixed population ofautologous endothelial cells and fibroblasts; thiswas subsequently transplanted as an orthotopicsingle pulmonary valve leaflet in an ovinemodel.181 Hoerstrup et al. showed that MSCwere a reliable cell source for heart valve tissueengineering.182 The created trileaflet valvedemonstrated morphological features andmechanical properties similar to those of nativevalves, while MSC differentiation into cells ofmyofibroblast phenotype was reported. Basedon previous in vitro work,183 Sutherland and col-leagues demonstrated prolonged (>4 months) invivo function of a stem cell-engineered valve con-structed from MSC and a biodegradable heartvalve scaffold.184 Even more interestingly, thesequential seeding of biodegradable leaflet scaf-folds with human fetal mesenchymal progenitorsand umbilical cord blood-derived endothelial


progenitor cells lead to constructs comparablewith native heart valve leaflets.185 Growing outof this important contribution, prenatal chorio-nic villus sampling may allow surgeons to per-form autologous valve replacements immediatelyafter birth.

Stem Cells and Pulmonary Regeneration

Lung parenchyma is organized around a com-plex 3-D array of proximal and distal airwayslined with numerous types of distinct epithelia.Architectural complexity alongside multiple insitu cell lineages is the main limitations to recon-struct pulmonary units via tissue engineering.Resident stem cell populations have beendescribed to exist within the proximal air-ways,186,187 distal airways,188,189 and thealveoli.190,191 The actual stemness of these popu-lations remains at issue, as the term ‘‘reparativecells’’ has been recently proposed to indicate cellpopulations involved in lung remodeling follow-ing injury.192 BM-derived stem cells have beendescribed to populate the pulmonary parench-yma,193,194 while recruitment of non-residentstem cells appears to be associated withinjury.195–197 Interestingly, cells recruited toinjury do not always participate in tissue repair;circulating fibrocytes have the opposite effect, asthese cells are known to contribute to the patho-genesis of pulmonary fibrosis.198 Distal lungepithelial cell progenitors have been generatedfrom ES cells199 but small cell yields along withethical considerations of ES cell research. Prob-ably make this approach very remote at presenttoclinical transition.Mondrinosandco-workers200

recently demonstrated in vitro generation ofbranching sacculated 3-D pulmonary tissue con-structs with mixed populations of murine fetalpulmonary isolates, cultured in 3-D hydrogels inthe presence of tissue-specific growth factors.Seeding of ovine somatic lung progenitor cellsonto polyglycolic acid or Pluronic F-127 scaf-folds leads to production of identifiable pulmon-ary structures;201 although polymers performedlikewise in vitro, Pluronic F-127 scaffolds gener-ated less tissue inflammation in vivo resulting intissue organization comparable to normal pul-monary parenchyma.

From the clinical front, stem cell therapy isabout to make its debut in pulmonary disease.202

Peripheral mobilized autologous mononuclear

cells engineered with human nitric oxidesynthase will be directly infused within the pul-monary arterial tree of patients with idiopathicpulmonary arterial hypertension. Extensive pre-clinical work has shown restoration of the integ-rity of the pulmonary microcirculation as well asreversal of established pulmonary hypertensionin relevant animal models. Of all developments,this trial from Toronto is probably the mostanticipated in the field of pulmonary stem celltherapy.

Central Nervous System Diseases

Replacement cell therapy in the central nervoussystem (CNS) is arguably the most complextopic in regenerative medicine. Lessons learnedfrom countless studies have begun to outlinewhat is required of the transplanted cells inorder to be successful. Within the CNS, cellsmust migrate to the site of injury, and differenti-ate into the appropriate cell type according todisease. Differentiated cells must integrate withexisting neural cells, create functional synapseformations with existing circuitry, and restoresaltatory conductivity when appropriate. Neuro-transmitter release and uptake must have a reg-ulatory mechanism. In most degenerative dis-eases of the CNS, a concerted regeneration ofmultiple types of neural cells must occur forfunctional recovery. In vitro investigation hasdemonstrated that ES cells, fetal derived stem,and progenitor cells, as well as cells fromnon-neural sources are capable of exhibitingphenotypic, genetic, and electrophysiologicalcharacteristics of neural cells. Animal modelsof CNS diseases are proof positive that stemcells are capable of restoring appropriate motorfunction. However, the molecular mechanismsof repair remain unknown making long-termeffects of transplantation uncertain. Here, wediscuss the feasibility of stem cell therapy inCNS regeneration in the most likely candidatesfor clinical use, Parkinson’s disease (PD), andstroke.

Parkinson’s Disease

Parkinson’s disease (PD) is a neurodegenerativedisease characterized by a progressive lossof mesencephalic dopaminergic neurons.Symptoms include tremor, rigidity, and


hypokinesia.203 The pathology of PD becomesmore complicated as there are thought to besymptoms which are due to subsequent degen-eration of other neuronal systems. Current treat-ments include administration of L-DOPA, deepbrain stimulation, dopamine (DA) agonists,and inhibitors of DA breakdown.204,205 Currenttreatments temporarily alleviate physicalsymptoms of Parkinson’s, but do not impedeprogression of the disease. In addition, thereare drawbacks to available therapy whichinclude dyskinesia and the loss of the efficacywith disease progression. Ideally, clinical cell-based therapy should provide long-lastingmajor improvements of mobility, suppressionof dyskinesias, and improvement of symptomsthat are resistant to other treatments such asbalance problems.

Cell transplantation therapy would be idealin diseases like PD where the pathology isrestricted to only one part of the brain, theA9 region of the striatum, and the damage isto a specific subset of dopaminergic neurons.Transplantation experiments of fetal neuraltissue in animals and humans have enabledresearchers to identify the requirements thatneed to be satisfied by stem cells in order toachieve clinical recovery in PD. First, the cellsmust have the capacity for regulated dopa-mine synthesis and release while showingphenotypic, morphological, and electrophysi-cal properties of dopaminergic neurons of thesubstantia pars compacta. Second, in animalmodels of PD, these cells must be able toameliorate motor symptoms while survivingover long-term periods. A third and impor-tant requirement is that grafted dopaminergicneurons must demonstrate functionality, inte-grating into the hosts’ terminal network ofnerve cells.204,205

Stem Cell Therapy in Parkinson’s Disease

Neurons exhibiting a dopaminergic phenotypereportedly have been generated from mouse,monkey, and human ES cells as well as fromfetal rodent and human neural stem cells (NSC)and non-neural tissue. Strategies for efficientinduction of dopamine neurons from ES cellshave been developed using forced gene expres-sion, growth factor cocktails, retinoic acid, andco-culture experiments. For example, nuclear

receptor related-1 (Nurr-1) is a transcriptionfactor which plays a role in the differentiationof dopaminergic neurons; driving Nurr-1expression in ES cells successfully induces thedopaminergic phenotype in vitro.206 In ratmodels of PD, transplants of ES cells over-expressing Nurr-1 successfully integrated intothe site of lesion and expressed tyrosine hydro-xylase (TH), the enzyme involved in dopaminesynthesis.206 Another example of inducingdopaminergic differentiation of ES cells is bystromal cell-derived inducing activity(SDIA).207 Co-culture of ES cells with PA6 stro-mal cell line yields dopamine secreting THþneurons, which upon transplantation into6-OHDA PD animal models, graft into the stria-tum and continue to express TH.207,208

Although hopeful results have been demon-strated using mouse ES cells, issues surround-ing the genetic stability of ES cells need to beaddressed before ES cells can be used in humantrials. Long-term culture of mouse ES cells canlead to chromosomal aberrations which havealso been observed in mid-term culturedhuman ES cell lines, implying potential detri-mental long-term effects associated with ES celltransplantation.209,210

Stem cells derived from tissues other thanthe brain are also currently being explored asa possible source for autologous transplant.Increasing numbers of investigations areexploring the possibility of generating dopa-minergic neurons from post-natal stem cellsderived from other tissues such as BM, per-ipheral blood, and umbilical cord.204,211,212

The transdifferentiation of HSC and MSCinto neural cells is very controversial, thedifferentiation of cells from one-germ lineageto another still raises doubts despite growingevidence of alternative sources of tissue com-mitted neural cells such as in BM.212 In vitro,multipotent BM-derived stem cells are cap-able of exhibiting neuron-like morphologyand phenotype and expressing several neuro-nal-specific markers including TH.204,211,213

Transplantation experiments using GFP-labeled BM-derived stem cells demonstratemigration, differentiation, and long-term sur-vival of these cells in rodent models ofPD.204,211,212 Umbilical cord stem cells havealso been successfully induced to differentiateinto THþ neurons, using neuron-conditionedmedia supplemented with FGF-8 and sonichedgehog.214


Clinical Studies

During development, fetal ventral mesencephalictissue contains a mixture of two subtypes ofdopaminergic neurons, making it an ideal candi-date for transplantation therapy. Difficulties inidentifying surface markers which are differen-tially expressed between neuronal subtypes, cre-ates a problem in trying to purify a population ofdopaminergic neurons specific to the substantianigra pars compacta. Using transgenic mice andretrograde axonal tracing, Thompson et al. 215 hasdemonstrated that dopaminergic neurons whichinnervate the substantia nigra pars compacta(SNpc) almost exclusively express the transcrip-tion factor Girk2. The SNpc is the A9 region of thebrain which is affected during Parkinson’s dis-ease. These results imply axonal guidancemechanisms take place in the CNS, suggestingthat cell transplantation therapies for PD mayonly be efficient if transplanted fetal cells arecapable of differentiating into the correct DAphenotype specific to the SNpc. Researchers inthe clinical setting do not yet have an accuratemeasure of the neuronal cell type which engraftsin transplantation. As a marker of graft viability,clinical studies measure fluorodopa (18F-dopa)uptake on positron emission (PET). However,this measures total F-dopa uptake and cannotdistinguish between the uptake from endogenousneurons and the grafted neurons.216 Results fromclinical trials remain unclear because of contrast-ing results from different groups.

In open-label clinical trials, human fetal ven-tral mesencephalic tissue transplants haveshown long-term survival and restoration ofdopaminergic networks in the striatum.216–220

However, it is not entirely clear if transplantedcells are able to synthesize and secrete dopaminein a regulated manner. Grafts contained a mixedpopulation of neural cells, 5–10% of which weredopaminergic neurons205 thus, it is unclearwhether the success of engraftment was duespecifically to transplanted dopaminergic neu-rons, or possibly due to a role of accompanyingglial cells. Increasing evidence points to animportant role of astrocytes during neuronalfate specification of neural stem and precursorcells.221 Initial transplantations of fetal braintissue were promising, and in the best case, apatient showed marked symptomatic relief forup to 10 years post-transplant, without the needfor additional anti-Parkinsonian medicationsuch as L-dopa.218 In a 59-year-old patient who

received two transplants from seven donors,post-mortem histopathological investigation ofthe grafted area showed TH immunoreactivecells and dopaminergic fibers proximal to thegrafted area and extended to host tissue indicat-ing that dopaminergic processes crossed thegraft–host interface and innervated the Parkin-sonian striatum.216 Despite graft survival andintegration, some patients have developedpost-transplant dyskinesia even in the absenceof anti-Parkinsonian medication.218 It has beenspeculated that this may be due to uneven inner-vation of grafted cells, inflammation, differentpopulations of DA neurons, or inappropriatesynaptic graft–host connections.218,219,222 Thisalludes to a lack of proper regulatory mechan-isms of DA synthesis and release.

Although open-label clinical trials havedemonstrated the viability and innervations offetal transplants, results from more recent dou-ble-blind trials showed contrasting evidence,stating that no significant alleviation of motorsymptoms of PD were found after fetal mesen-cephalic tissue transplantation. Two recent dou-ble-blind placebo-controlled clinical trials bothreported that the grafts failed to improve pri-mary endpoints in PD patients.218–220 Betweenthe two trials, 73 patients in total were dividedinto two groups, one of which received fetalnigral transplantation, and the other a shamsurgery, in both studies, only the neurosurgeonwas aware of the type of surgery conducted oneach patient. There were no significant differ-ences in the overall treatment effect between thetransplantation group and the sham surgerygroup in either study, as both failed to meet theprimary end point. The primary outcome wasthe difference in the UPDRS (Unified Parkin-son’s Disease Rating Scale) motor score betweenthe baseline reading prior to surgery and finalphysician visits post-operatively.

A big problem with transplantation of fetalmesencephalic tissue is that 15% of patientsdevelop dyskinesia.219,220,223 It was thought thatovergrowth of grafted cells may lead to excessDA release subsequently causing dyskinesias.However, no correlation has been foundbetween graft-induced dyskinesia and highlevels of F-dopa post-operatively.223 Dyskinesiamay have also resulted from abnormal regula-tory mechanisms of DA synthesis and releasefrom transplanted cells.223 There are conflictingreports which compare putaminal F-dopauptake in grafted patients with and without


dyskinesias. Ma et al.224 reported that in dyski-netic patients, there is an imbalance betweendopaminergic innervation in the ventral anddorsal putamen. In contrast, Olanow et al.220

reported no such regional differences of F-dopalevels between patients with and without graft-induced dyskinesia. Differences in outcomebetween the open-label trials and the double-blind trials have not been explained as thereare many differences among the studies such aspatient selection criteria, immunosuppressionregimes, and donor tissue preparation.Although it is promising that fetal nigral trans-plants survive, integrate with host tissue, andexpress TH, fetal cell transplantation cannot berecommended at this time as it has not shownlong-term symptomatic relief nor has it beenshown to stop progression of the disease.

Although results of cell transplantation stu-dies seem hopeful and cell integration can beseen, there is no proof of dopamine secretionby the exogenous cells. In addition, there has notbeen a clear demonstration that DA neuronsgenerated in vitro can efficiently reinnervatethe striatum, release dopamine in a regulatedmanner, and provide functional recovery fromPD once transplanted into animal models. Theoptimal number of grafted cells requiredremains unknown, and efforts to standardizefetal tissue transplantations have failed to pinpoint the number of donor required for humantransplant.218 Although functional recovery istemporarily observed after transplantation, per-haps further investigations should focus on themechanisms of symptomatic relief and exploit-ing endogenous means of dopaminergic neuro-nal repair. Investigators have observed anincrease in neurogenesis after delivery of lesionto the dopaminergic system, while other studiesfind that lesion to the DA system stimulates aglial response rather than an increase in neuro-nal generation.205,225

Stroke

In stroke, a cerebral artery is blocked, causing afocal ischemia and loss of neurons and glia.Subsequently, multifactorial motor, sensory,and cognitive capacities are impaired, andthere is neither a cure nor an effective means ofpromoting recovery. Thus, therapy which haspotential of neuron regeneration is attractive.There are two main arms of investigation in the

search for stem cell therapy in stroke: (i) directlytransplanting stem and progenitor cells and(ii) stimulating the proliferation of endogenousneural stem cells which would ideally migrate tothe penumbra region and replace degeneratedneurons and glia.

Stem Cell Therapy in Stroke

ES cells have been successfully transplanted inmodels of PD, before their therapeutic potentialin cerebral ischemia was explored. The challengehere is that unlike PD, where damage is specificto dopaminergic neurons in a defined region ofthe brain SNpc, stroke affects glia and varyingtypes of neurons, leaving the brain susceptible todamage followed by the possibility of inflamma-tion and glial scar. Thus, specifically definingwhich neural cell type to transplant remainschallenging.

Neural precursor cells (NPC) can be derivedfrom ES cells, expanded as neurospheres and,upon attachment to a substrate, can be differen-tiated into mixed cultures of glia and matureneurons. ES cell-derived NPC transplanted intoa rodent stroke model survive and differentiateinto glia and electrophysiologically mature neu-rons of different neurotransmitter sub-types.226,227 Tracking NPC using a GFP tagshowed that after transplantation, these cellssurvived and matured into glutaminergic,gabaergic, dopaminergic, serotonergic, and cho-linergic neurons.226,227 In stroke models, ES cellsgenetically modified to over-express the anti-apoptotic gene bcl-2 have increased survivalrelative to normal ES cells. Also, there is anincrease in the ES cell-derived neuronal fractionof the transplanted cells. After focal ischemia,animals that received Bcl-2 over-expressing EScells demonstrated significant functional recov-ery relative to animals that received non-geneti-cally modified ES cells and control animals thatreceived culture medium.228 There are two con-troversial findings with regard to neural differ-entiation from cells originating in non-neuraltissue. First, several groups have independentlyidentified stem and progenitor cells whichexpress neural lineage markers circulating inthe peripheral blood and residing in theBM.211,212,229,230 Second, cells derived fromnon-neural tissues such as cord blood, umbilicalcord, peripheral blood, and BM have beeninduced to have neural phenotypes in


vitro.213,231–234 Are there stem cells in non-neural tissue that are truly plastic? Or are inves-tigators differentiating pre-determined neuralcommitted progenitor cells?

In rat stroke models, intravenous administra-tion of CD34þ hUCBCs within 1 day of ischemicinduction, lead to significant improvement offunctional recovery. CD34þ hUCBCs selectivelymigrated to the injured tissue, survived, anddifferentiated in the brain but there was no sig-nificant reduction in the lesion volume. Themajority of cells were found at the ischemicboundary,235 perhaps due to the lack of bloodsupply in the lesion itself. In lieu of replacingdamaged cells and integrating into host circui-try, the administration of hUCBCs and theirmigration to injury may instead stimulatetrophic factors, endogenous neuroprotection,and brain recovery mechanisms.235 The samegroup has shown that after stroke, intravenousadministration of BM stromal cells enhancesangiogenesis at the ischemic boundary of thehost brain, which is important in promotingrepair and restoration of blood flow to theischemic area. The administration of BM stro-mal cells promotes endogenous secretion ofVEGF and the upregulation of VEGF receptor 2(VEGFR2), which mediate angiogenesis.236 Ani-mal transplantation studies have shown thatMSC transplantation into ischemic brain modelsleads to functional recovery, providing relief insensory-motor symptoms caused bystroke.225,237 However, close immunohistologi-cal examination of post-mortem tissue showsevidence that functional recovery may havebeen mediated by molecular signals secreted bythe transplanted MSC.225 Although MSC providea therapeutic effect, genetically modified MSCprogrammed to hypersecrete trophic factorssuch as brain-derived neurotrophic factor(BDNF) and glial cell line-derived neurotrophicfactor (GDNF) have a greater therapeutic effectrelative to non-genetically altered MSC.237,238

Cerebral ischemia induces the release of severalneurotrophic factors in the brain includingGDNF and BDNF which are thought to haveneuroprotective properties.238 Treatment withintravenous BDNF after stroke improves func-tional motor recovery in animal models.239

BDNF rescues motoneurons, hippocampal neu-rons, and dopaminergic neurons from braininjury, and its upregulation after focal cerebralischemia suggests that BDNF may have a neuro-protective role in stroke.240 Topical treatment of

focal ischemia with GDNF significantly reducesinfarct volume.241 Investigators have thusexplored the potential of transfecting MSC withvectors containing BDNF and GDNF.237,242

BDNF gene-modified human MSC (BDNF-MSC) intravenously administered to rats withcerebral occlusion reduced lesion volume andincreased functional improvement compared toanimals that received MSC alone.242 In addition,intravenous administration of MSC transfectedwith GDNF (GDNF-MSC) also showed a greaterreduction in the lesion volume, increase inGDNF in the penumbra region, and improve-ment in behavioral defects.237,242 Althoughthese results show promise in gene-modifiedMSC transplantation therapy for stroke, thelong-term effects of receiving gene-modifiedcells remain uncertain.

Clinical Study

There has been one safety and feasibility trial forcell transplantation in stroke. This open-labeltrail consisted of 12 patients with at least abasal ganglionic infarct. Patients received LBS-Neurons (from Layton Bio-Science) which areproduced from the NT2/D1 human precursorcell line and induced to differentiate into neu-rons. Two million cells per injection were admi-nistered stereotactically, and patients receivedbetween 1 and 3 injections (2–6 � 106 cells).Six out of 12 patients showed an improved func-tional outcome, but three patients deteriorated,exhibiting decreased motor performance rela-tive to baseline performance prior to surgery.Although this trial may have proved that trans-plantation is safe, its efficacy is poor and therewere no consistent signs of improvement inpatients.243

Conclusions

Despite a great amount of interest and develop-ment in stem cell biology and therapy in the pastfew years, many questions are yet to be answeredbefore stem cell therapy can be applied to itsfullest potential in the clinic. Which cells shouldbe used in order to maximize potential benefitsand what is the best method of their delivery?The ideal route for administering stem cells hasstill yet to be determined, but it is important totake certain factors into consideration. The


strength of homing signals may vary in differentclinical scenarios. In more acutely ischemic sce-narios, the stem cells may be administered eitherperipherally or locally through the circulatorysystem. When the homing signals may be lessintense, injection of the cells directly into thedamage tissue may produce a more favorableoutcome. It also remains to be determined howcell survival can be optimized and how cells canbe tracked once delivered to ensure that theyreach the right location. An important issueconcerning the therapeutic use of stem cells isthe quantity of cells necessary to achieve anoptimal effect. Ideally, cells should expandextensively in vitro have minimal immunogeni-city and be able to reconstitute tissue whentransplanted into damaged tissue. It must alsobe defined which patients are suitable for thistherapy and which stem cell types are the mosteffective given the underlying pathology.Currently, it is not possible to fully anticipatelong-term side effects, since most of the trials arevery recent. In addition, there is growing evi-dence that transplanted cells not just simplyreplace missing tissue but also trigger localmechanisms to initiate a repair response andsecret factors useful for tissue protection orneovascularization.

Ultimately, large clinical trials will have to beconducted and at the same time, we need tocontinue the basic research to elucidate theunderlying mechanism of stem cell therapy.

References

1. Assmus B, Schachinger V, Teupe C, Britten M, LehmannR, Dobert N, et al. Transplantation of Progenitor Cellsand Regeneration Enhancement in Acute MyocardialInfarction (TOPCARE-AMI). Circulation. December 10,2002;106(24):3009–17.

2. Wollert KC, Meyer GP, Lotz J, Ringes-Lichtenberg S,Lippolt P, Breidenbach C, et al. Intracoronary autolo-gous bone-marrow cell transfer after myocardial infarc-tion: the BOOST randomised controlled clinical trial.Lancet. July 10–16, 2004;364(9429):141–8.

3. Thomson JA, Itskovitz-Eldor J, Shapiro SS, Waknitz MA,Swiergiel JJ, Marshall VS, et al. Embryonic stem cell linesderived from human blastocysts. Science. November 6,1998;282(5391):1145–7.

4. Brivanlou AH, Gage FH, Jaenisch R, Jessell T, Melton D,Rossant J. Stem cells. Setting standards for humanembryonic stem cells. Science. May 9, 2003;300(5621):913–6.

5. Reubinoff BE, Pera MF, Fong CY, Trounson A, Bongso A.Embryonic stem cell lines from human blastocysts:

somatic differentiation in vitro. Nat Biotechnol. April2000;18(4):399–404.

6. Schuldiner M, Yanuka O, Itskovitz-Eldor J, Melton DA,Benvenisty N. Effects of eight growth factors on thedifferentiation of cells derived from human embryonicstem cells. Proc Natl Acad Sci USA. October 10,2000;97(21):11307–12.

7. Kaufman DS, Hanson ET, Lewis RL, Auerbach R,Thomson JA. Hematopoietic colony-forming cellsderived from human embryonic stem cells. Proc NatlAcad Sci USA. September 11, 2001;98(19):10716–21.

8. Kehat I, Kenyagin-Karsenti D, Snir M, Segev H, Amit M,Gepstein A, et al. Human embryonic stem cells candifferentiate into myocytes with structural and func-tional properties of cardiomyocytes. J Clin Invest.August 2001;108(3):407–14.

9. Levenberg S, Golub JS, Amit M, Itskovitz-Eldor J, LangerR. Endothelial cells derived from human embryonicstem cells. Proc Natl Acad Sci USA. April 2,2002;99(7):4391–6.

10. Assady S, Maor G, Amit M, Itskovitz-Eldor J, SkoreckiKL, Tzukerman M. Insulin production by humanembryonic stem cells. Diabetes. August 2001;50(8):1691–7.

11. Gearhart J. New human embryonic stem-cell lines–moreis better. N Engl J Med. March 25, 2004;350(13):1275–6.

12. Robertson JA. Human embryonic stem cell research:ethical and legal issues. Nat Rev Genet. January2001;2(1):74–8.

13. Weissman IL. Stem cells: units of development, units ofregeneration, and units in evolution. Cell. January 7,2000;100(1):157–68.

14. Fuchs E, Tumbar T, Guasch G. Socializing with theneighbors: stem cells and their niche. Cell. March 19,2004;116(6):769–78.

15. Slack JM. Stem cells in epithelial tissues. Science. Febru-ary 25, 2000;287(5457):1431–3.

16. Kucia M, Ratajczak J, Reca R, Janowska-Wieczorek A,Ratajczak MZ. Tissue-specific muscle, neural and liverstem/progenitor cells reside in the bone marrow,respond to an SDF-1 gradient and are mobilized intoperipheral blood during stress and tissue injury. BloodCells Mol Dis. January–February 2004;32(1):52–7.

17. Kucia M, Ratajczak J, Ratajczak MZ. Bone marrow as asource of circulating CXCR4þ tissue-committed stemcells. Biol Cell. February 2005;97(2):133–46.

18. Till JE, McCulloch E. A direct measurement of the radia-tion sensitivity of normal mouse bone marrow cells.Radiat Res. February 1961;14:213–22.

19. Bianco P, Riminucci M, Gronthos S, Robey PG. Bonemarrow stromal stem cells: nature, biology, and poten-tial applications. Stem Cells. 2001;19(3):180–92.

20. Pittenger MF, Mackay AM, Beck SC, Jaiswal RK, DouglasR, Mosca JD, et al. Multilineage potential of adulthuman mesenchymal stem cells. Science. April 2,1999;284(5411):143–7.

21. Orlic D, Kajstura J, Chimenti S, Jakoniuk I, AndersonSM, Li B, et al. Bone marrow cells regenerate infarctedmyocardium. Nature. April 5, 2001;410(6829):701–5.

22. Gussoni E, Soneoka Y, Strickland CD, Buzney EA, KhanMK, Flint AF, et al. Dystrophin expression in the mdxmouse restored by stem cell transplantation. Nature.September 23, 1999;401(6751):390–4.

23. Mezey E, Chandross KJ, Harta G, Maki RA, McKercherSR. Turning blood into brain: cells bearing neuronal


antigens generated in vivo from bone marrow. Science.December 1, 2000;290(5497):1779–82.

24. Nash K, Hafeez A, Hou S. Hospital-acquired renal insuf-ficiency. Am J Kidney Dis. May 2002;39(5):930–6.

25. Yamamoto M, Cui L, Johkura K, Asanuma K, Okouchi Y,Ogiwara N, et al. Branching ducts similar to mesonephricducts or ureteric buds in teratomas originating frommouse embryonic stem cells. Am J Physiol Renal Physiol.January 2006;290(1):F52–60.

26. Steenhard BM, Isom KS, Cazcarro P, Dunmore JH,Godwin AR, St John PL, et al. Integration of embryonicstem cells in metanephric kidney organ culture. J Am SocNephrol. June 2005;16(6):1623–31.

27. Kobayashi T, Tanaka H, Kuwana H, Inoshita S, TeraokaH, Sasaki S, et al. Wnt4-transformed mouse embryonicstem cells differentiate into renal tubular cells. BiochemBiophys Res Commun. October 21 2005;336(2):585–95.

28. Kramer J, Steinhoff J, Klinger M, Fricke L, Rohwedel J.Cells differentiated from mouse embryonic stem cells viaembryoid bodies express renal marker molecules. Differ-entiation. March 2006;74(2–3):91–104.

29. Vigneau C, Zheng F, Polgar K, Wilson PD, Striker G.Stem cells and kidney injury. Curr Opin Nephrol Hyper-tens. May 2006;15(3):238–44.

30. Kim D, Dressler GR. Nephrogenic factors promote differ-entiation of mouse embryonic stem cells into renal epithe-lia. J Am Soc Nephrol. December 2005;16(12):3527–34.

31. Lanza RP, Chung HY, Yoo JJ, Wettstein PJ, Blackwell C,Borson N, et al. Generation of histocompatible tissuesusing nuclear transplantation. Nat Biotechnol. July2002;20(7):689–96.

32. Poulsom R, Prodromidi EI, Pusey CD, Cook HT. Celltherapy for renal regeneration – time for some joined-up thinking? Nephrol Dial Transplant. December2006;21(12):3349–53.

33. Reinders ME, Rabelink TJ, Briscoe DM. Angiogenesisand endothelial cell repair in renal disease and allograftrejection. J Am Soc Nephrol. April 2006;17(4):932–42.

34. Williams GM, Alvarez CA. Host repopulation of theendothelium in allografts of kidneys and aorta. SurgForum. 1969;20:293–4.

35. Sinclair RA. Origin of endothelium in human renal allo-grafts. Br Med J. October 7, 1972;4(831):15–6.

36. Lagaaij EL, Cramer-Knijnenburg GF, van Kemenade FJ,van Es LA, Bruijn JA, van Krieken JH. Endothelial cellchimerism after renal transplantation and vascularrejection. Lancet. January 6, 2001;357(9249):33–7.

37. Xu W, Baelde HJ, Lagaaij EL, De Heer E, Paul LC, BruijnJA. Endothelial cell chimerism after renal transplanta-tion in a rat model. Transplantation. November 15,2002;74(9):1316–20.

38. Kerjaschki D, Huttary N, Raab I, Regele H, Bojarski-Nagy K, Bartel G, et al. Lymphatic endothelial progenitorcells contribute to de novo lymphangiogenesis in humanrenal transplants. Nat Med. February 2006;12(2):230–4.

39. Rookmaaker MB, Tolboom H, Goldschmeding R,Zwaginga JJ, Rabelink TJ, Verhaar MC. Bone-marrow-derived cells contribute to endothelial repair after throm-botic microangiopathy. Blood. February 1, 2002;99(3):1095.

40. Dekel B, Shezen E, Even-Tov-Friedman S, Katchman H,Margalit R, Nagler A, et al. Transplantation of humanhematopoietic stem cells into ischemic and growing kid-neys suggests a role in vasculogenesis but not tubulogen-esis. Stem Cells. May 2006;24(5):1185–93.

41. Cornacchia F, Fornoni A, Plati AR, Thomas A, Wang Y,Inverardi L, et al. Glomerulosclerosis is transmitted bybone marrow-derived mesangial cell progenitors. J ClinInvest. December 2001;108(11):1649–56.

42. Rookmaaker MB, Smits AM, Tolboom H, Van ’t Wout K,Martens AC, Goldschmeding R, et al. Bone-marrow-derived cells contribute to glomerular endothelial repairin experimental glomerulonephritis. Am J Pathol.August 2003;163(2):553–62.

43. Imasawa T, Utsunomiya Y, Kawamura T, Zhong Y,Nagasawa R, Okabe M, et al. The potential of bone mar-row-derived cells to differentiate to glomerular mesangialcells. J Am Soc Nephrol. July 2001;12(7):1401–9.

44. Ito T, Suzuki A, Imai E, Okabe M, Hori M. Bone marrowis a reservoir of repopulating mesangial cells duringglomerular remodeling. J Am Soc Nephrol. December2001;12(12):2625–35.

45. Uchimura H, Marumo T, Takase O, Kawachi H, ShimizuF, Hayashi M, et al. Intrarenal injection of bone marrow-derived angiogenic cells reduces endothelial injury andmesangial cell activation in experimental glomerulone-phritis. J Am Soc Nephrol. April 2005;16(4):997–1004.

46. Ikarashi K, Li B, Suwa M, Kawamura K, Morioka T, YaoJ, et al. Bone marrow cells contribute to regeneration ofdamaged glomerular endothelial cells. Kidney Int. May2005;67(5):1925–33.

47. Li B, Morioka T, Uchiyama M, Oite T. Bone marrow cellinfusion ameliorates progressive glomerulosclerosis inan experimental rat model. Kidney Int. February2006;69(2):323–30.

48. Hayakawa M, Ishizaki M, Hayakawa J, Migita M,Murakami M, Shimada T, et al. Role of bone marrowcells in the healing process of mouse experimental glo-merulonephritis. Pediatr Res. August 2005;58(2):323–8.

49. Grimm PC, Nickerson P, Jeffery J, Savani RC, Gough J,McKenna RM, et al. Neointimal and tubulointerstitialinfiltration by recipient mesenchymal cells in chronicrenal-allograft rejection. N Engl J Med. July 12,2001;345(2):93–7.

50. Sun DF, Fujigaki Y, Fujimoto T, Yonemura K, Hishida A.Possible involvement of myofibroblasts in cellular recov-ery of uranyl acetate-induced acute renal failure in rats.Am J Pathol. October 2000;157(4):1321–35.

51. Direkze NC, Forbes SJ, Brittan M, Hunt T, Jeffery R,Preston SL, et al. Multiple organ engraftment bybone-marrow-derived myofibroblasts and fibroblasts inbone-marrow-transplanted mice. Stem Cells. 2003;21(5):514–20.

52. Lin F, Moran A, Igarashi P. Intrarenal cells, not bonemarrow-derived cells, are the major source for regenera-tion in postischemic kidney. J Clin Invest. July2005;115(7):1756–64.

53. Roufosse C, Bou-Gharios G, Prodromidi E, Alexakis C,Jeffery R, Khan S, et al. Bone marrow-derived cells do notcontribute significantly to collagen I synthesis in a mur-ine model of renal fibrosis. J Am Soc Nephrol. March2006;17(3):775–82.

54. Gupta S, Verfaillie C, Chmielewski D, Kim Y, RosenbergME. A role for extrarenal cells in the regeneration followingacute renal failure. Kidney Int. October 2002;62(4):1285–90.

55. Poulsom R, Forbes SJ, Hodivala-Dilke K, Ryan E, WylesS, Navaratnarasah S, et al. Bone marrow contributes torenal parenchymal turnover and regeneration. J Pathol.September 2001;195(2):229–35.


56. Nishida M, Kawakatsu H, Shiraishi I, Fujimoto S, GotohT, Urata Y, et al. Renal tubular regeneration by bonemarrow-derived cells in a girl after bone marrow trans-plantation. Am J Kidney Dis. November 2003;42(5):E10–2.

57. Fang TC, Alison MR, Cook HT, Jeffery R, Wright NA,Poulsom R. Proliferation of bone marrow-derived cellscontributes to regeneration after folic acid-inducedacute tubular injury. J Am Soc Nephrol. June2005;16(6):1723–32.

58. Szczypka MS, Westover AJ, Clouthier SG, Ferrara JL,Humes HD. Rare incorporation of bone marrow-derivedcells into kidney after folic acid-induced injury. StemCells. 2005;23(1):44–54.

59. Duffield JS, Bonventre JV. Kidney tubular epithelium isrestored without replacement with bone marrow-derived cells during repair after ischemic injury. KidneyInt. November 2005;68(5):1956–61.

60. Duffield JS, Park KM, Hsiao LL, Kelley VR, Scadden DT,Ichimura T, et al. Restoration of tubular epithelial cellsduring repair of the postischemic kidney occurs inde-pendently of bone marrow-derived stem cells. J ClinInvest. July 2005;115(7):1743–55.

61. Morigi M, Imberti B, Zoja C, Corna D, Tomasoni S,Abbate M, et al. Mesenchymal stem cells are renotropic,helping to repair the kidney and improve function inacute renal failure. J Am Soc Nephrol. July2004;15(7):1794–804.

62. Herrera MB, Bussolati B, Bruno S, Fonsato V, RomanazziGM, Camussi G. Mesenchymal stem cells contribute tothe renal repair of acute tubular epithelial injury. Int JMol Med. December 2004;14(6):1035–41.

63. Lange C, Togel F, Ittrich H, Clayton F, Nolte-Ernsting C,Zander AR, et al. Administered mesenchymal stem cellsenhance recovery from ischemia/reperfusion-inducedacute renal failure in rats. Kidney Int. October2005;68(4):1613–7.

64. Togel F, Hu Z, Weiss K, Isaac J, Lange C, Westenfelder C.Administered mesenchymal stem cells protect againstischemic acute renal failure through differentiation-independent mechanisms. Am J Physiol Renal Physiol.July 2005;289(1):F31–42.

65. Krause DS, Theise ND, Collector MI, Henegariu O,Hwang S, Gardner R, et al. Multi-organ, multi-lineageengraftment by a single bone marrow-derived stem cell.Cell. May 4, 2001;105(3):369–77.

66. Wagers AJ, Sherwood RI, Christensen JL, Weissman IL.Little evidence for developmental plasticity of adult hema-topoietic stem cells. Science. Sep 27, 2002;297(5590):2256–9.

67. Kale S, Karihaloo A, Clark PR, Kashgarian M, Krause DS,Cantley LG. Bone marrow stem cells contribute to repairof the ischemically injured renal tubule. J Clin Invest.July 2003;112(1):42–9.

68. Lin F, Cordes K, Li L, Hood L, Couser WG, Shankland SJ,et al. Hematopoietic stem cells contribute to the regen-eration of renal tubules after renal ischemia-reperfusioninjury in mice. J Am Soc Nephrol. May 2003;14(5):1188–99.

69. Masuya M, Drake CJ, Fleming PA, Reilly CM, Zeng H,Hill WD, et al. Hematopoietic origin of glomerularmesangial cells. Blood. March 15, 2003;101(6):2215–8.

70. Zheng F, Cornacchia F, Schulman I, Banerjee A, ChengQL, Potier M, et al. Development of albuminuria andglomerular lesions in normoglycemic B6 recipients of

db/db mice bone marrow: the role of mesangial cellprogenitors. Diabetes. September 2004;53(9):2420–7.

71. Guo JK, Ardito TA, Kashgarian M, Krause DS. Preven-tion of mesangial sclerosis by bone marrow transplanta-tion. Kidney Int. September 2006;70(5):910–3.

72. Perry J, Tam S, Zheng K, Sado Y, Dobson H, Jefferson B,et al. Type IV Collagen Induces Podocytic Features inBone Marrow Stromal Stem Cells In Vitro. J Am SocNephrol. January 2006;17(1):66–76.

73. Prodromidi EI, Poulsom R, Jeffery R, Roufosse CA,Pollard PJ, Pusey CD, et al. Bone Marrow Derived-Cells Contribute to Podocyte Regeneration and Ame-lioration of Renal Disease in a Mouse Model ofAlport Syndrome. Stem Cells. November 2006;24(11):2448–55.

74. Sugimoto H, Mundel TM, Sund M, Xie L, Cosgrove D,Kalluri R. Bone-marrow-derived stem cells repair base-ment membrane collagen defects and reverse genetickidney disease. Proc Natl Acad Sci USA. 2006 May 9,2006;103(19):7321–6.

75. Ninichuk V, Gross O, Segerer S, Hoffmann R, RadomskaE, Buchstaller A, et al. Multipotent mesenchymal stemcells reduce interstitial fibrosis but do not delay progres-sion of chronic kidney disease in collagen4A3-deficientmice. Kidney Int. July 2006;70(1):121–9.

76. Hammerman MR. Renal organogenesis from trans-planted metanephric primordia. J Am Soc Nephrol. May2004;15(5):1126–32.

77. Woolf AS, Palmer SJ, Snow ML, Fine LG. Creation of afunctioning chimeric mammalian kidney. Kidney Int.November 1990;38(5):991–7.

78. Woolf AS, Hornbruch A, Fine LG. Integration of newembryonic nephrons into the kidney. Am J Kidney Dis.June 1991;17(6):611–4.

79. Rogers SA, Lowell JA, Hammerman NA, HammermanMR. Transplantation of developing metanephroi intoadult rats. Kidney Int. July 1998;54(1):27–37.

80. Hammerman MR. Growing new kidneys in situ. Clin ExpNephrol. September 2004;8(3):169–77.

81. Rogers SA, Hammerman MR. Transplantation of meta-nephroi after preservation in vitro. Am J Physiol RegulIntegr Comp Physiol. August 2001;281(2):R661–5.

82. Dekel B, Amariglio N, Kaminski N, Schwartz A, Goshen E,Arditti FD, et al. Engraftment and differentiation ofhuman metanephroi into functional mature nephronsafter transplantation into mice is accompanied by a pro-file of gene expression similar to normal human kidneydevelopment. J Am Soc Nephrol. April 2002;13(4):977–90.

83. Dekel B, Burakova T, Arditti FD, Reich-Zeliger S,Milstein O, Aviel-Ronen S, et al. Human and porcineearly kidney precursors as a new source for transplanta-tion. Nat Med. January 2003;9(1):53–60.

84. Yokoo T, Ohashi T, Shen JS, Sakurai K, Miyazaki Y,Utsunomiya Y, et al. Human mesenchymal stem cells inrodent whole-embryo culture are reprogrammed to con-tribute to kidney tissues. Proc Natl Acad Sci USA. March1 2005;102(9):3296–300.

85. Humes HD, Fissell WH, Weitzel WF, Buffington DA,Westover AJ, MacKay SM, et al. Metabolic replacementof kidney function in uremic animals with a bioartificialkidney containing human cells. Am J Kidney Dis. May2002;39(5):1078–87.

86. Humes HD, MacKay SM, Funke AJ, Buffington DA. Tis-sue engineering of a bioartificial renal tubule assist


device: in vitro transport and metabolic characteristics.Kidney Int. June 1999;55(6):2502–14.

87. Nikolovski J, Gulari E, Humes HD. Design engineeringof a bioartificial renal tubule cell therapy device. CellTransplant. July–August 1999;8(4):351–64.

88. Humes HD, Weitzel WF, Bartlett RH, Swaniker FC,Paganini EP, Luderer JR, et al. Initial clinical results ofthe bioartificial kidney containing human cells in ICUpatients with acute renal failure. Kidney Int. October2004;66(4):1578–88.

89. Tiranathanagul K, Brodie J, Humes HD. Bioartificialkidney in the treatment of acute renal failure associatedwith sepsis. Nephrology (Carlton). August 2006;11(4):285–91.

90. Fissell WH. Developments towards an artificial kidney.Expert Rev Med Devices. March 2006;3(2):155–65.

91. Fissell WH, Manley S, Westover A, Humes HD,Fleischman AJ, Roy S. Differentiated growth of humanrenal tubule cells on thin-film and nanostructuredmaterials. Asaio J. May–June 2006;52(3):221–7.

92. Devendra D, Liu E, Eisenbarth GS. Type 1 diabetes:recent developments. Bmj. March 27, 2004;328(7442):750–4.

93. Lechner A, Habener JF. Stem/progenitor cells derivedfrom adult tissues: potential for the treatment of dia-betes mellitus. Am J Physiol Endocrinol Metab. Feb2003;284(2):E259–66.

94. Halvorsen TL, Beattie GM, Lopez AD, Hayek A, LevineF. Accelerated telomere shortening and senescence inhuman pancreatic islet cells stimulated to divide invitro. J Endocrinol. July 2000;166(1):103–9.

95. Hori Y, Rulifson IC, Tsai BC, Heit JJ, Cahoy JD, Kim SK.Growth inhibitors promote differentiation of insulin-producing tissue from embryonic stem cells. Proc NatlAcad Sci USA. December 10, 2002;99(25):16105–10.

96. Miyazaki S, Yamato E, Miyazaki J. Regulatedexpression of pdx-1 promotes in vitro differentiationof insulin-producing cells from embryonic stem cells.Diabetes. April 2004;53(4):1030–7.

97. Soria B, Roche E, Berna G, Leon-Quinto T, Reig JA,Martin F. Insulin-secreting cells derived from embryo-nic stem cells normalize glycemia in streptozotocin-induced diabetic mice. Diabetes. February 2000;49(2):157–62.

98. Fujikawa T, Oh SH, Pi L, Hatch HM, Shupe T, PetersenBE. Teratoma formation leads to failure of treatmentfor type I diabetes using embryonic stem cell-derivedinsulin-producing cells. Am J Pathol. June 2005;166(6):1781–91.

99. Ianus A, Holz GG, Theise ND, Hussain MA. In vivoderivation of glucose-competent pancreatic endocrinecells from bone marrow without evidence of cell fusion.J Clin Invest. March 2003;111(6):843–50.

100. Taneera J, Rosengren A, Renstrom E, Nygren JM, SerupP, Rorsman P, et al. Failure of transplanted bone mar-row cells to adopt a pancreatic beta-cell fate. Diabetes.February 2006;55(2):290–6.

101. Hess D, Li L, Martin M, Sakano S, Hill D, Strutt B,et al. Bone marrow-derived stem cells initiate pan-creatic regeneration. Nat Biotechnol. July 2003;21(7):763–70.

102. Mathews V, Hanson PT, Ford E, Fujita J, Polonsky KS,Graubert TA. Recruitment of bone marrow-derivedendothelial cells to sites of pancreatic beta-cell injury.Diabetes. January 2004;53(1):91–8.

103. Lechner A, Yang YG, Blacken RA, Wang L, Nolan AL,Habener JF. No evidence for significant transdifferen-tiation of bone marrow into pancreatic beta-cells invivo. Diabetes. March 2004;53(3):616–23.

104. Kang EM, Zickler PP, Burns S, Langemeijer SM,Brenner S, Phang OA, et al. Hematopoietic stem celltransplantation prevents diabetes in NOD mice butdoes not contribute to significant islet cell regenerationonce disease is established. Exp Hematol. June2005;33(6): 699–705.

105. Banerjee M, Kumar A, Bhonde RR. Reversal of experi-mental diabetes by multiple bone marrow transplanta-tion. Biochem Biophys Res Commun. March 4,2005;328(1):318–25.

106. Ende N, Chen R, Reddi AS. Effect of human umbilicalcord blood cells on glycemia and insulitis in type 1diabetic mice. Biochem Biophys Res Commun. Decem-ber 17, 2004;325(3):665–9.

107. Tang DQ, Cao LZ, Burkhardt BR, Xia CQ, LitherlandSA, Atkinson MA, et al. In vivo and in vitro character-ization of insulin-producing cells obtained from mur-ine bone marrow. Diabetes. July 2004;53(7):1721–32.

108. Oh SH, Muzzonigro TM, Bae SH, LaPlante JM, HatchHM, Petersen BE. Adult bone marrow-derived cellstrans-differentiating into insulin-producing cells forthe treatment of type I diabetes. Lab Invest. May2004;84(5):607–17.

109. Fernandez Vina R, Saslavsky J, Andrin O, Vrsalovick F,Ferreyra de Silva J, Ferreyra O, et al. First word reporteddatas from Argentina of implant and cellular therapywith autologous adult stem cells in type 2 diabeticpatients (Teceldiar study 1). 4th ISSCR Annual Meet-ing, Toronto; 2006.

110. Fernandez Vina R, Andrin O, Saslavsky J, Ferreyra deSilva J, Vrsalovick F, Camozzi L, et al. Increase of ‘c’peptide level in type 1 diabetics patients after directpancreas implant by endovascular way of autologousadult mononuclears CD34þCD38 (–) cells (Teceldiab 2study). 4th ISSCR Annual Meeting, Toronto; 2006.

111. Health Mo. Annual Report of the Chief Medical Officer.London: Ministry of Health, UK; 2001.

112. Fox IJ, Chowdhury JR, Kaufman SS, Goertzen TC,Chowdhury NR, Warkentin PI, et al. Treatment of theCrigler-Najjar syndrome type I with hepatocyte trans-plantation. N Engl J Med. May 14, 1998;338(20):1422–6.

113. Strom SC, Fisher RA, Thompson MT, Sanyal AJ, ColePE, Ham JM, et al. Hepatocyte transplantation as abridge to orthotopic liver transplantation in terminalliver failure. Transplantation. February 271997;63(4):559–69.

114. Bilir BM, Guinette D, Karrer F, Kumpe DA, Krysl J,Stephens J, et al. Hepatocyte transplantation in acuteliver failure. Liver Transpl. January 2000;6(1):32–40.

115. Ohashi K, Park F, Kay MA. Hepatocyte transplantation:clinical and experimental application. J Mol Med.November 2001;79(11):617–30.

116. Najimi M, Sokal E. Liver cell transplantation. MinervaPediatr. October 2005;57(5):243–57.

117. Hamazaki T, Iiboshi Y, Oka M, Papst PJ, Meacham AM,Zon LI, et al. Hepatic maturation in differentiatingembryonic stem cells in vitro. FEBS Lett. May 18,2001;497(1):15–9.

118. Jones EA, Tosh D, Wilson DI, Lindsay S, Forrester LM.Hepatic differentiation of murine embryonic stem cells.Exp Cell Res. January 1, 2002;272(1):15–22.


119. Ishii T, Yasuchika K, Fujii H, Hoppo T, Baba S, Naito M,et al. In vitro differentiation and maturation of mouseembryonic stem cells into hepatocytes. Exp Cell Res.September 10 2005;309(1):68–77.

120. Yin Y, Lim YK, Salto-Tellez M, Ng SC, Lin CS, Lim SK.AFP(þ), ESC-derived cells engraft and differentiateinto hepatocytes in vivo. Stem Cells. 2002;20(4):338–46.

121. Yamada T, Yoshikawa M, Kanda S, Kato Y, Nakajima Y,Ishizaka S, et al. In vitro differentiation of embryonic stemcells into hepatocyte-like cells identified by cellular uptakeof indocyanine green. Stem Cells. 2002;20(2):146–54.

122. Chinzei R, Tanaka Y, Shimizu-Saito K, Hara Y, Kaki-numa S, Watanabe M, et al. Embryoid-body cellsderived from a mouse embryonic stem cell line showdifferentiation into functional hepatocytes. Hepatology.July 2002;36(1):22–9.

123. Petersen BE, Bowen WC, Patrene KD, Mars WM,Sullivan AK, Murase N, et al. Bone marrow as a poten-tial source of hepatic oval cells. Science. May 14,1999;284(5417):1168–70.

124. Grompe M, Lindstedt S, al-Dhalimy M, Kennaway NG,Papaconstantinou J, Torres-Ramos CA, et al. Pharma-cological correction of neonatal lethal hepatic dysfunc-tion in a murine model of hereditary tyrosinaemiatype I. Nat Genet. August 1995;10(4):453–60.

125. Sakaida I, Terai S, Yamamoto N, Aoyama K, Ishikawa T,Nishina H, et al. Transplantation of bone marrow cellsreduces CCl4-induced liver fibrosis in mice. Hepatol-ogy. December 2004;40(6):1304–11.

126. Mallet VO, Mitchell C, Mezey E, Fabre M, Guidotti JE,Renia L, et al. Bone marrow transplantation in miceleads to a minor population of hepatocytes that can beselectively amplified in vivo. Hepatology. April2002;35(4):799–804.

127. Terai S, Sakaida I, Yamamoto N, Omori K, Watanabe T,Ohata S, et al. An in vivo model for monitoring trans-differentiation of bone marrow cells into functional hepa-tocytes. J Biochem (Tokyo). October 2003;134(4):551–8.

128. Jang YY, Collector MI, Baylin SB, Diehl AM, Sharkis SJ.Hematopoietic stem cells convert into liver cells withindays without fusion. Nat Cell Biol. June2004;6(6):532–9.

129. Theise ND, Badve S, Saxena R, Henegariu O, Sell S,Crawford JM, et al. Derivation of hepatocytes frombone marrow cells in mice after radiation-induced mye-loablation. Hepatology. January 2000;31(1):235–40.

130. Wang X, Ge S, McNamara G, Hao QL, Crooks GM, NoltaJA. Albumin-expressing hepatocyte-like cells developin the livers of immune-deficient mice that receivedtransplants of highly purified human hematopoieticstem cells. Blood. May 15, 2003;101(10):4201–8.

131. Kanazawa Y, Verma IM. Little evidence of bone mar-row-derived hepatocytes in the replacement of injuredliver. Proc Natl Acad Sci USA. September 30,2003;100(Suppl 1):11850–3.

132. Alison MR, Poulsom R, Jeffery R, Dhillon AP, QuagliaA, Jacob J, et al. Hepatocytes from non-hepatic adultstem cells. Nature. July 20, 2000;406(6793):257.

133. Theise ND, Nimmakayalu M, Gardner R, Illei PB,Morgan G, Teperman L, et al. Liver from bone marrowin humans. Hepatology. July 2000;32(1):11–6.

134. Korbling M, Katz RL, Khanna A, Ruifrok AC, RondonG, Albitar M, et al. Hepatocytes and epithelial cells ofdonor origin in recipients of peripheral-blood stemcells. N Engl J Med. Mar 7, 2002;346(10):738–46.

135. Ng IO, Chan KL, Shek WH, Lee JM, Fong DY, Lo CM,et al. High frequency of chimerism in transplantedlivers. Hepatology. October 2003;38(4):989–98.

136. Fogt F, Beyser KH, Poremba C, Zimmerman RL,Khettry U, Ruschoff J. Recipient-derived hepatocytesin liver transplants: a rare event in sex-mismatchedtransplants. Hepatology. July 2002;36(1):173–6.

137. Wu T, Cieply K, Nalesnik MA, Randhawa PS, SonzogniA, Bellamy C, et al. Minimal evidence of transdifferen-tiation from recipient bone marrow to parenchymalcells in regenerating and long-surviving human allo-grafts. Am J Transplant. September 2003;3(9):1173–81.

138. am Esch JS, 2nd, Knoefel WT, Klein M, Ghodsizad A,Fuerst G, Poll LW, et al. Portal application of autolo-gous CD133þ bone marrow cells to the liver: a novelconcept to support hepatic regeneration. Stem Cells.April 2005;23(4):463–70.

139. Gordon MY, Levicar N, Pai M, Bachellier P, DimarakisI, Al-Allaf F, et al. Characterisation and Clinical Appli-cation of Human Cd34þ Stem/Progenitor Cell Popula-tions Mobilised into the Blood by G-Csf. Stem Cells. July2006;24(7):1822–30.

140. Terai S, Ishikawa T, Omori K, Aoyama K, Marumoto Y,Urata Y, et al. Improved liver function in liver cirrhosispatients after autologous bone marrow cell infusiontherapy. Stem Cells. October 2006;24(10):2292–8.

141. Wei H, Juhasz O, Li J, Tarasova YS, Boheler KR.Embryonic stem cells and cardiomyocyte differentia-tion: phenotypic and molecular analyses. J Cell MolMed. October–December 2005;9(4):804–17.

142. Min JY, Yang Y, Sullivan MF, Ke Q, Converso KL, ChenY, et al. Long-term improvement of cardiac function inrats after infarction by transplantation of embryonicstem cells. J Thorac Cardiovasc Surg. February2003;125(2):361–9.

143. Hodgson DM, Behfar A, Zingman LV, Kane GC, Perez-Terzic C, Alekseev AE, et al. Stable benefit of embryonicstem cell therapy in myocardial infarction. Am J PhysiolHeart Circ Physiol. September 2004;287(2):H471–9.

144. Min JY, Huang X, Xiang M, Meissner A, Chen Y, Ke Q,et al. Homing of intravenously infused embryonic stemcell-derived cells to injured hearts after myocardialinfarction. J Thorac Cardiovasc Surg. April2006;131(4):889–97.

145. Kehat I, Khimovich L, Caspi O, Gepstein A, Shofti R,Arbel G, et al. Electromechanical integration of cardio-myocytes derived from human embryonic stem cells.Nat Biotechnol. November 2004;22(10):1282–9.

146. Zhang YM, Hartzell C, Narlow M, Dudley SC, Jr. Stemcell-derived cardiomyocytes demonstrate arrhythmicpotential. Circulation. September 3,2002;106(10):1294–9.

147. Ye L, Haider H, Sim EK. Adult stem cells for cardiacrepair: a choice between skeletal myoblasts and bonemarrow stem cells. Exp Biol Med (Maywood). January2006;231(1):8–19.

148. Minguell JJ, Erices A. Mesenchymal stem cells and thetreatment of cardiac disease. Exp Biol Med (Maywood).January 2006;231(1):39–49.

149. Yoon YS, Park JS, Tkebuchava T, Luedeman C, LosordoDW. Unexpected severe calcification after transplanta-tion of bone marrow cells in acute myocardial infarc-tion. Circulation. June 29 2004;109(25):3154–7.

150. Kang HJ, Kim HS, Zhang SY, Park KW, Cho HJ, KooBK, et al. Effects of intracoronary infusion of peripheral


blood stem-cells mobilised with granulocyte-colony sti-mulating factor on left ventricular systolic function andrestenosis after coronary stenting in myocardial infarc-tion: the MAGIC cell randomised clinical trial. Lancet.March 6, 2004;363(9411):751–6.

151. Yoon YS, Wecker A, Heyd L, Park JS, Tkebuchava T,Kusano K, et al. Clonally expanded novel multipotentstem cells from human bone marrow regenerate myo-cardium after myocardial infarction. J Clin Invest. Feb-ruary 2005;115(2):326–38.

152. Stenderup K, Justesen J, Clausen C, Kassem M. Aging isassociated with decreased maximal life span and accel-erated senescence of bone marrow stromal cells. Bone.December 2003;33(6):919–26.

153. Vacanti V, Kong E, Suzuki G, Sato K, Canty JM, Lee T.Phenotypic changes of adult porcine mesenchymalstem cells induced by prolonged passaging in culture.J Cell Physiol. November 2005;205(2):194–201.

154. Rubio D, Garcia-Castro J, Martin MC, de la Fuente R,Cigudosa JC, Lloyd AC, et al. Spontaneous human adultstem cell transformation. Cancer Res. April 15,2005;65(8):3035–9.

155. Chang MG, Tung L, Sekar RB, Chang CY, Cysyk J, DongP, et al. Proarrhythmic potential of mesenchymal stemcell transplantation revealed in an in vitro coculturemodel. Circulation. April 18 2006;113(15):1832–41.

156. Majka SM, Jackson KA, Kienstra KA, Majesky MW,Goodell MA, Hirschi KK. Distinct progenitor popula-tions in skeletal muscle are bone marrow derived andexhibit different cell fates during vascular regeneration.J Clin Invest. February 2003;111(1):71–9.

157. Summer R, Kotton DN, Sun X, Fitzsimmons K, Fine A.Translational physiology: origin and phenotype of lungside population cells. Am J Physiol Lung Cell Mol Phy-siol. September 2004;287(3):L477–83.

158. Osawa M, Hanada K, Hamada H, Nakauchi H. Long-term lymphohematopoietic reconstitution by a singleCD34-low/negative hematopoietic stem cell. Science.July 12, 1996;273(5272):242–5.

159. Yeh ET, Zhang S, Wu HD, Korbling M, Willerson JT,Estrov Z. Transdifferentiation of human peripheralblood CD34þ-enriched cell population into cardio-myocytes, endothelial cells, and smooth muscle cellsin vivo. Circulation. October 28, 2003;108(17):2070–3.

160. Murry CE, Soonpaa MH, Reinecke H, Nakajima H,Nakajima HO, Rubart M, et al. Haematopoietic stemcells do not transdifferentiate into cardiac myocytes inmyocardial infarcts. Nature. April 8, 2004;428(6983):664–8.

161. Balsam LB, Wagers AJ, Christensen JL, Kofidis T,Weissman IL, Robbins RC. Haematopoietic stem cellsadopt mature haematopoietic fates in ischaemic myo-cardium. Nature. April 8, 2004;428(6983):668–73.

162. Deten A, Volz HC, Clamors S, Leiblein S, Briest W,Marx G, et al. Hematopoietic stem cells do not repairthe infarcted mouse heart. Cardiovasc Res. January 1,2005;65(1):52–63.

163. Choi K, Kennedy M, Kazarov A, Papadimitriou JC,Keller G. A common precursor for hematopoietic andendothelial cells. Development. February 1998;125(4):725–32.

164. Asahara T, Masuda H, Takahashi T, Kalka C, Pastore C,Silver M, et al. Bone marrow origin of endothelial pro-genitor cells responsible for postnatal vasculogenesis in

physiological and pathological neovascularization. CircRes. August 6, 1999;85(3):221–8.

165. Kalka C, Masuda H, Takahashi T, Kalka-Moll WM,Silver M, Kearney M, et al. Transplantation of ex vivoexpanded endothelial progenitor cells for therapeuticneovascularization. Proc Natl Acad Sci USA. March 28,2000;97(7):3422–7.

166. Fazel S, Cimini M, Chen L, Li S, Angoulvant D, Fedak P,et al. Cardioprotective c-kitþ cells are from the bonemarrow and regulate the myocardial balance of angio-genic cytokines. J Clin Invest. July 2006;116(7):1865–77.

167. Vasa M, Fichtlscherer S, Aicher A, Adler K, Urbich C,Martin H, et al. Number and migratory activity ofcirculating endothelial progenitor cells inversely corre-late with risk factors for coronary artery disease. CircRes. July 6, 2001;89(1):E1–7.

168. Hill JM, Zalos G, Halcox JP, Schenke WH, WaclawiwMA, Quyyumi AA, et al. Circulating endothelial pro-genitor cells, vascular function, and cardiovascularrisk. N Engl J Med. February 13, 2003;348(7):593–600.

169. Jiang XX, Zhang Y, Liu B, Zhang SX, Wu Y, Yu XD, et al.Human mesenchymal stem cells inhibit differentiationand function of monocyte-derived dendritic cells.Blood. May 15, 2005;105(10):4120–6.

170. Amado LC, Saliaris AP, Schuleri KH, St John M, Xie JS,Cattaneo S, et al. Cardiac repair with intramyocardialinjection of allogeneic mesenchymal stem cells aftermyocardial infarction. Proc Natl Acad Sci USA. August9, 2005;102(32):11474–9.

171. Jiang Y, Jahagirdar BN, Reinhardt RL, Schwartz RE,Keene CD, Ortiz-Gonzalez XR, et al. Pluripotency ofmesenchymal stem cells derived from adult marrow.Nature. July 4, 2002;418(6893):41–9.

172. Kajstura J, Rota M, Whang B, Cascapera S, Hosoda T,Bearzi C, et al. Bone marrow cells differentiate in car-diac cell lineages after infarction independently of cellfusion. Circ Res. January 7, 2005;96(1):127–37.

173. Nygren JM, Jovinge S, Breitbach M, Sawen P, Roll W,Hescheler J, et al. Bone marrow-derived hematopoieticcells generate cardiomyocytes at a low frequencythrough cell fusion, but not transdifferentiation. NatMed. May 2004;10(5):494–501.

174. Zhang S, Wang D, Estrov Z, Raj S, Willerson JT, Yeh ET.Both cell fusion and transdifferentiation account for thetransformation of human peripheral blood CD34-posi-tive cells into cardiomyocytes in vivo. Circulation.December 21, 2004;110(25):3803–7.

175. Tang YL, Zhao Q, Qin X, Shen L, Cheng L, Ge J, et al.Paracrine action enhances the effects of autologousmesenchymal stem cell transplantation on vascularregeneration in rat model of myocardial infarction.Ann Thorac Surg. July 2005;80(1):229–36; discussion36–7.

176. Kamihata H, Matsubara H, Nishiue T, Fujiyama S,Tsutsumi Y, Ozono R, et al. Implantation of bonemarrow mononuclear cells into ischemic myocardiumenhances collateral perfusion and regional functionvia side supply of angioblasts, angiogenic ligands,and cytokines. Circulation. August 28, 2001;104(9):1046–52.

177. Zhao YS, Wang CY, Li DX, Zhang XZ, Qiao Y, Guo XM,et al. Construction of a unidirectionally beating 3-dimensional cardiac muscle construct. J Heart LungTransplant. August 2005;24(8):1091–7.


178. Furuta A, Miyoshi S, Itabashi Y, Shimizu T, Kira S,Hayakawa K, et al. Pulsatile cardiac tissue grafts usinga novel three-dimensional cell sheet manipulation tech-nique functionally integrates with the host heart, invivo. Circ Res. March 17, 2006;98(5):705–12.

179. Baar K, Birla R, Boluyt MO, Borschel GH, Arruda EM,Dennis RG. Self-organization of rat cardiac cells intocontractile 3-D cardiac tissue. Faseb J. February 2005;19(2):275–7.

180. Wei HJ, Chen SC, Chang Y, Hwang SM, Lin WW, LaiPH, et al. Porous acellular bovine pericardia seededwith mesenchymal stem cells as a patch to repair amyocardial defect in a syngeneic rat model. Biomater-ials. November 2006;27(31):5409–19.

181. Shinoka T, Breuer CK, Tanel RE, Zund G, Miura T, MaPX, et al. Tissue engineering heart valves: valve leafletreplacement study in a lamb model. Ann Thorac Surg.December 1995;60(6 Suppl):S513–6.

182. Hoerstrup SP, Kadner A, Melnitchouk S, Trojan A, EidK, Tracy J, et al. Tissue engineering of functional trileaf-let heart valves from human marrow stromal cells. Cir-culation. September 24, 2002;106(12 Suppl 1):I143–50.

183. Perry TE, Kaushal S, Sutherland FW, Guleserian KJ,Bischoff J, Sacks M, et al. Thoracic Surgery DirectorsAssociation Award. Bone marrow as a cell source fortissue engineering heart valves. Ann Thorac Surg.March 2003;75(3):761–7; discussion 7.

184. Sutherland FW, Perry TE, Yu Y, Sherwood MC, RabkinE, Masuda Y, et al. From stem cells to viable autologoussemilunar heart valve. Circulation. May 31,2005;111(21):2783–91.

185. Schmidt D, Mol A, Breymann C, Achermann J,Odermatt B, Gossi M, et al. Living autologous heartvalves engineered from human prenatally harvested pro-genitors. Circulation. July 4, 2006;114(1 Suppl):I125–31.

186. Hong KU, Reynolds SD, Watkins S, Fuchs E, Stripp BR.In vivo differentiation potential of tracheal basal cells:evidence for multipotent and unipotent subpopula-tions. Am J Physiol Lung Cell Mol Physiol. April 2004;286(4):L643–9.

187. Hong KU, Reynolds SD, Watkins S, Fuchs E, Stripp BR.Basal cells are a multipotent progenitor capable ofrenewing the bronchial epithelium. Am J Pathol. Feb-ruary 2004;164(2):577–88.

188. Reynolds SD, Hong KU, Giangreco A, Mango GW,Guron C, Morimoto Y, et al. Conditional clara cellablation reveals a self-renewing progenitor function ofpulmonary neuroendocrine cells. Am J Physiol LungCell Mol Physiol. June 2000;278(6):L1256–63.

189. Boers JE, Ambergen AW, Thunnissen FB. Number andproliferation of clara cells in normal human airwayepithelium. Am J Respir Crit Care Med. May 1999;159(5 Pt 1):1585–91.

190. Adamson IY, Bowden DH. The type 2 cell as progenitorof alveolar epithelial regeneration. A cytodynamicstudy in mice after exposure to oxygen. Lab Invest.January 1974;30(1):35–42.

191. Evans MJ, Cabral LJ, Stephens RJ, Freeman G. Transforma-tion of alveolar type 2 cells to type 1 cells following expo-sure to NO2. Exp Mol Pathol. February 1975;22(1):142–50.

192. Weiss DJ, Berberich MA, Borok Z, Gail DB, Kolls JK,Penland C, et al. Adult stem cells, lung biology, and lungdisease. NHLBI/Cystic Fibrosis Foundation Workshop.Proc Am Thorac Soc. May 2006;3(3):193–207.

193. Pereira RF, Halford KW, O’Hara MD, Leeper DB,Sokolov BP, Pollard MD, et al. Cultured adherent cellsfrom marrow can serve as long-lasting precursor cellsfor bone, cartilage, and lung in irradiated mice. ProcNatl Acad Sci USA. May 23, 1995;92(11):4857–61.

194. Pereira RF, O’Hara MD, Laptev AV, Halford KW,Pollard MD, Class R, et al. Marrow stromal cells as asource of progenitor cells for nonhematopoietic tissuesin transgenic mice with a phenotype of osteogenesisimperfecta. Proc Natl Acad Sci USA. February 3,1998;95(3):1142–7.

195. Abe S, Boyer C, Liu X, Wen FQ, Kobayashi T, Fang Q,et al. Cells derived from the circulation contribute tothe repair of lung injury. Am J Respir Crit Care Med.December 1, 2004;170(11):1158–63.

196. Rojas M, Xu J, Woods CR, Mora AL, Spears W, RomanJ, et al. Bone marrow-derived mesenchymal stem cellsin repair of the injured lung. Am J Respir Cell Mol Biol.August 2005;33(2):145–52.

197. Macpherson H, Keir P, Webb S, Samuel K, Boyle S,Bickmore W, et al. Bone marrow-derived SP cells cancontribute to the respiratory tract of mice in vivo. J CellSci. June 1, 2005;118(Pt 11):2441–50.

198. Phillips RJ, Burdick MD, Hong K, Lutz MA, Murray LA,Xue YY, et al. Circulating fibrocytes traffic to the lungsin response to CXCL12 and mediate fibrosis. J ClinInvest. August 2004;114(3):438–46.

199. Rippon HJ, Polak JM, Qin M, Bishop AE. Derivation ofdistal lung epithelial progenitors from murine embryo-nic stem cells using a novel three-step differentiationprotocol. Stem Cells. May 2006;24(5):1389–98.

200. Mondrinos MJ, Koutzaki S, Jiwanmall E, Li M,Dechadarevian JP, Lelkes PI, et al. Engineering three-dimensional pulmonary tissue constructs. Tissue Eng.April 2006;12(4):717–28.

201. Cortiella J, Nichols JE, Kojima K, Bonassar LJ, DargonP, Roy AK, et al. Tissue-engineered lung: an in vivo andin vitro comparison of polyglycolic acid and pluronicF-127 hydrogel/somatic lung progenitor cell constructsto support tissue growth. Tissue Eng. May 2006;12(5):1213–25.

202. http://controlled-trials.com/ISRCTN14519481/ISRCTN14519481 (accessed 8 October 2006).

203. Arenas E. Stem cells in the treatment of Parkinson’sdisease. Brain Res Bull. April 2002;57(6):795–808.

204. Hellmann MA, Djaldetti R, Israel Z, Melamed E. Effectof deep brain subthalamic stimulation on camptocor-mia and postural abnormalities in idiopathic Parkin-son’s disease. Mov Disord. September 13, 2006.

205. Lindvall O, Kokaia Z, Martinez-Serrano A. Stem celltherapy for human neurodegenerative disorders-how to make it work. Nature Medicine. July 2004;10(Suppl:):S42–50.

206. Kim J-H, Auerbach JM, Rodriguez-Gomez JA, Velasco I,Gavin D, Lumelsky N, et al. Dopamine neurons derivedfrom embryonic stem cells function in an animal modelof Parkinson’s disease. Nature. 2002;418(6893):50–6.

207. Kawasaki H, Mizuseki K, Nishikawa S, Kaneko S, KuwanaY, Nakanishi S, et al. Induction of midbrain dopaminergicneurons from ES cells by stromal cell-derived inducingactivity. Neuron. October 2000;28:31–40.

208. Morizane A, Takahashi J, Takagi Y, Sasai Y, HashimotoN. Optimal conditions for in vivo induction of dopami-nergic neurons from embryonic stem cells through


stromal cell-inducing activity. J Neurosci Res. Septem-ber 15, 2002;69(6):934–9.

209. Draper JS, Smith K, Gokhale P, Moore HD, Maltby E,Johnson J, et al. Recurrent gain of chromosomes 17qand 12 in cultured human embryonic stem cells. NatBiotechnology. 2004;22(1):53–4.

210. Longo L, Bygrave A, Grosveld F, Pandolfi PP. The chro-mosome make-up of mouse embryonic stem cells ispredictive of somatic and germ cell chimaerism. Trans-genic Res. September 1997;6(5):321–8.

211. Blondheim N, Levy Y, Ben-Zur T, Burshtein A, CherlowT, Kan I, et al. Human mesenchymal stem cells expressneural genes, suggesting a neural predisposition. StemCells Dev. April 2006;15(2):141–64.

212. Kucia M, Zhang YP, Reca R, Wysoczynski M,Machalinski B, Majka M, et al. Cells enriched in mar-kers of neural tissue-committed stem cells reside in thebone marrow and are mobilized into the peripheralblood following stroke. Leukemia. 2005;20(1):18–28.

213. Jiang Y, Henderson D, Blackstad M, Chen A, Miller RF,Verfaillie CM. Neuroectodermal differentiation frommouse multipotent adult progenitor cells. Proc NatlAcad Sci USA. September 302003;100 Suppl 1:11854–60.

214. Fu YS, Cheng YC, Lin MY, Cheng H, Chu PM, Chou SC,et al. Conversion of human umbilical cord mesenchy-mal stem cells in Wharton’s jelly to dopaminergicneurons in vitro: potential therapeutic applicationfor Parkinsonism. Stem Cells. January 2006;24(1):115–24.

215. Thompson L, Barraud P, Andersson E, Kirik D,Bjorklund A. Identification of dopaminergic neuronsof nigral and ventral tegmental area subtypes in graftsof fetal ventral mesencephalon based on cell morphol-ogy, protein expression, and efferent projections. J Neu-rosci. July 6, 2005;25(27):6467–77.

216. Kordower JH, Freeman TB, Snow BJ, Vingerhoets FJ,Mufson EJ, Sanberg PR, et al. Neuropathological evi-dence of graft survival and striatal reinnervation afterthe transplantation of fetal mesencephalic tissue in apatient with Parkinson’s disease. N Engl J Med. April 27,1995;332(17):1118–24.

217. Clarkson ED, Freed CR. Development of fetal neuraltransplantation as a treatment for Parkinson’s disease.Life Sci. October 29, 1999;65(23):2427–37.

218. Correia AS, Anisimov SV, Li JY, Brundin P. Stem cell-based therapy for Parkinson’s disease. Ann Med.2005;37(7):487–98.

219. Freed CR, Greene PE, Breeze RE, Tsai WY, DuMouchelW, Kao R, et al. Transplantation of embryonic dopa-mine neurons for severe Parkinson’s disease. N Engl JMed. March 8, 2001;344(10):710–9.

220. Olanow CW, Goetz C, Kordower JH, Stoessl AJ, Sossi V,Brin M, et al. A double-blind controlled trail of bilateralfetal nigral transplantation in Parkinson’s disease. AnnNeurol. August 2003;54(3):403–14.

221. Song Z-M, Undie AS, Koh PO, Fang YY, Zhang L,Dracheva S, et al. D1 dopamine receptor regulation ofmicrotubule-associated protein-2 phosphorylation indeveloping cerebral cortical neurons. J Neurosci. July15, 2002;22(14):6092–105.

222. Hagell P, Piccini P, Bjorklund A, Brundin P, RehncronaS, Widner H, et al. Dyskinesias following neural trans-plantation in Parkinson’s disease. Nat Neurosci. July2002;5(7):627–8.

223. Piccini P, Pavese N, Hagell P, Reimer J, Bjorklund A,Oertel WH, et al. Factors affecting the clinical outcomeafter neural transplantation in Parkinson’s disease.Brain. December 2005;128(Pt 12):2977–86.

224. Ma N, Stamm C, Kaminski A, Li W, Kleine HD, Muller-Hilke B, et al. Human cord blood cells induce angiogen-esis following myocardial infarction in NOD/scid-mice.Cardiovasc Res. April 1, 2005;66(1):45–54.

225. Zhao LR, Duan WM, Reyes M, Verfaillie CM, Low WC.Immunohistochemical identification of multipotentadult progenitor cells from human bone marrow aftertransplantation into the rat brain. Brain Res Brain ResProtoc. March 2003;11(1):38–45.

226. Buhnemann C, Scholz A, Bernreuther C, Malik CY,Braun H, Schachner M, et al. Neuronal differentiationof transplanted embryonic stem cell-derived precursorsin stroke lesions of adult rats. Brain. December2006;129(pt12):3238–48.

227. Takagi Y, Nishimura M, Morizane A, Takahashi J,Nozaki K, Hayashi J, et al. Survival and differentiationof neural progenitor cells derived from embryonic stemcells and transplanted into ischemic brain. J Neurosurg.August 2005;103(2):304–10.

228. Wei L, Cui L, Snider BJ, Rivkin M, Yu SS, Lee CS, et al.Transplantation of embryonic stem cells overexpres-sing Bcl-2 promotes functional recovery after transientcerebral ischemia. Neurobiol Dis. June–July2005;19(1–2):183–93.

229. Goolsby J, Marty MC, Heletz D, Chiappelli J, Tashko G,Yarnell D, et al. Hematopoietic progenitors expressneural genes. Proc Natl Acad Sci USA. December 9,2003;100(25):14926–31.

230. Habich A, Jurga M, Markiewicz I, Lukomska B, Bany-Laszewicz U, Domanska-Janik K. Early appearance ofstem/progenitor cells with neural-like characteristics inhuman cord blood mononuclear fraction cultured invitro. Exp Hematol. July 2006;34(7):914–25.

231. Munoz-Elias G, Woodbury D, Black IB. Marrow stro-mal cells, mitosis, and neuronal differentiation: stemcell and precursor functions. Stem Cells. 2003;21(4):437–48.

232. Ortiz-Gonzalez XR, Keene CD, Verfaillie CM, Low WC.Neural induction of adult bone marrow and umbilicalcord stem cells. Curr Neurovasc res. July2004;1(3):207–13.

233. Padovan CS, Jahn K, Birnbaum T, Reich P, Sostak P,Strupp M, et al. Expression of neuronal markers indifferentiated marrow stromal cells and CD133þstem-like cells. Cell Transplant. 2003;12(8):839–48.

234. Woodbury D, Schwarz EJ, Prockop DJ, Black IB. Adult ratand human bone marrow stromal cells differentiate intoneurons. J Neurosci Res. August 15, 2000;61(4): 364–70.

235. Chen J, Sanberg PR, Li Y, Wang L, Lu M, Willing AE,et al. Intravenous administration of human umbilicalcord blood reduces behavioral deficits after stroke inrats. Stroke. November 2001;32(11):2682–8.

236. Chen SL, Fang WW, Ye F, Liu YH, Qian J, Shan SJ, et al.Effect on left ventricular function of intracoronarytransplantation of autologous bone marrow mesenchy-mal stem cell in patients with acute myocardial infarc-tion. Am J Cardiol. July 1, 2004;94(1):92–5.

237. Horita Y, Honmou O, Harada K, Houkin K, Hamada H,Kocsis JD. Intravenous administration of glial cell line-derived neurotrophic factor gene-modified humanmesenchymal stem cells protects against injury in a


cerebral ischemia model in the adult rat. J Neurosci Res.September 22 2006;84(7):1495–504.

238. Kurozumi K, Nakamura K, Tamiya T, Kawano Y, IshiiK, Kobune M, et al. Mesenchymal stem cells that pro-duce neurotrophic factors reduce ischemic damage inthe rat middle cerebral artery occlusion model. MolTher. January 2005;11(1):96–104.

239. Schabitz WR, Berger C, Kollmar R, Seitz M, Tanay E,Kiessling M, et al. Effect of brain-derived neurotrophicfactor treatment and forced arm use on functionalmotor recovery after small cortical ischemia. Stroke.April 2004;35(4):992–7.

240. Schabitz WR, Schwab S, Spranger M, Hacke W. Intra-ventricular brain-derived neurotrophic factor reducesinfarct size after focal cerebral ischemia in rats. J CerebBlood Flow Metab. May 1997;17(5):500–6.

241. Zhang WR, Hayashi T, Iwai M, Nagano I, Sato K,Manabe Y, et al. Time dependent amelioration againstischemic brain damage by glial cell line-derived neuro-trophic factor after transient middle cerebral arteryocclusion in rat. Brain Res. June 8, 2001;903(1–2):253–6.

242. Nomura T, Honmou O, Harada K, Houkin K, HamadaH, Kocsis JD. I.V. infusion of brain-derived neuro-trophic factor gene-modified human mesenchymalstem cells protects against injury in a cerebral ische-mia model in adult rat. Neuroscience. 2005;136(1):161–9.

243. Kondziolka D, Wechsler L, Goldstein S, Meltzer C,Thulborn KR, Gebel J, et al. Transplantation of culturedhuman neuronal cells for patients with stroke. Neurol-ogy. August 22, 2000;55(4):565–9.


10Composite Tissue Transplantation: A Stage BetweenSurgical Reconstruction and Cloning

Earl R. Owen and Nadey S. Hakim

There is no doubt the entire brief and brillianthistory of transplantation is about a temporary,far from ideal, and only partially successful solu-tion to the desperate attempt of our professionto keep people alive and functioning. With thepresent advances in nanosurgery, miniscopes,cloning, stem cell culturing and placement,laser solder bonding, ultrasonics, chemoge-nomics, and microrobotics involving physicaland chemical manipulations of even single cellcomponents, seasoned transplantation research-ers in the most advanced laboratories are lookingfor better therapy than transplants to replaceworn out, diseased, or failing tissues. These newmethods, now being developed, will take overfrom surgeons having to perform allografts,which still require immunosuppression of thebody’s incredibly evolved, individually persona-lized immunological recognition system.

The first Composite Tissue Allograft (CTA)was magnificently unsuccessful! It is, however,suitably emblazoned into the ‘‘literature’’, parti-cularly as a color feature and transplantationicon as there are approximately 29 beautifulpaintings since the 1600s, several of which arein the Vatican, of the event that made Saints outof Doctor’s Damian and Cosmos. In the 4thcentury AD, they are reported to have trans-planted the leg of a dead black man to replacethat of a white amputee, despite the obvious lackof immunosuppressives at the time. For such anoutrageous attempt, even by a surgeon, to ‘‘playGod’’, these worthy experimental surgeons were,of course, tried, condemned, and crucified bytheir conservative colleagues at that time.

The second lower limb CTA was also a mag-nificent failure. Alexis Carrel from Lyon,

France,1 received the Nobel Prize for Medicinein 1912. Carrel used his earlier researched vesselanastomotic technique,2 together with Dr Guthriefrom USA to actually ‘‘swap’’ (or double allograft)the left hind legs of a black and white pair ofgreyhound dogs in an initially surgically success-ful experiment.3 In the limited time they survived,these canine CTAs were sketched and the illustra-tion of the standing dogs appeared in theNew York newspapers at the time.

The first modern attempt at a genuinely spe-cific human limb or hand allograft4 was an inevi-table failure. In Ecuador in 1964, ignoring anyneed for early immunosuppression and basedon a mistake of family relationship between thedead donor and live hand amputee, this allograftfailed after the surgery despite the recipientbeing later airlifted to a Florida hospital forImuran and steroid therapy too late to be useful.

The frontier of CTA was considered to havebeen crossed, in a blaze of publicity, on Septem-ber 23, 1998, in Lyon, France, when our team ofinternational surgeons and specialists per-formed a terminal forearm and hand allograft.5

The donor was a French road traffic accidentvictim and the recipient a New Zealand-bornman who had lost his dominant right hand inan accident while he was in jail some 14 yearspreviously. It had been an objective of at leasttwo of the surgeons to aim their lifetime researchtoward CTAs. For the past 31 years, they hadbeen accumulating microsurgical and immuno-logical drug expertise and techniques with theirunit’s laboratory research programs as well astheir department’s clinical experiences duringthe development of both transplantation andmicrosurgery.


165

It is pertinent to consider how many scientificinnovations had to merge in order that thedaunting Frontier of Composite Tissue Allo-grafts (CTAs) be crossed.

Medical innovation tends to occur almostsimultaneously in different parts of the world.Thus, it was in the 1950s that alert surgicalminds became enthralled with the idea of oper-ating on smaller and smaller parts of the body,which led to microsurgery. Simultaneouslyintelligent surgeons with increasing researchopportunities started to attempt to reconstructor replace parts of the body beginning withorgans that were in terminal failure. The surgicalworld from then on saw the fields of microsur-gery and transplantation merge. It was on theeast coast of America that Julius Jacobson andAnasto Suarez defined the word and performedthe earliest practice of microvascular surgery byconsistently anastomosing arteries of 1 mm dia-meter.6 At the same time, Dr Sun Lee of Koreawas experimenting on the west coast with simi-lar microvascular techniques not only to joinsmall vessels but also to perform organ trans-plants in rodents using his amazing surgicalskills in research, although his medical degreesnever allowed him to practice in the UnitedStates.7 Sun Lee advanced the knowledge andscope of transplantation and immunology bytechnically describing and carrying out trans-plantations of the kidney, liver, spleen, stomach,heart, heart and lung, uterus, ovary, intestines,and even the pancreas in his years of surgicalrodent research.8 He proved this could be donecheaper and faster and with more certaintyacross larger series than those experimentswhich were also being attempted all around theworld in larger animals such as rabbits, dogs,and pigs.

The early history of successful kidney trans-plantation occurred during this 1950s intellectualawakening of medicos and Murray’s outstand-ingly successful kidney transplantations9 usingidentical twins as donors from 1954 onwards initself spurred the already exciting development ofimmunology.

A definition for CTA is simply the simulta-neous allotransplantation of multiple heteroge-neous livable nonorgan bodily tissues. It must,by definition, include the supplying of a func-tioning blood supply, and preferably a nervesupply, but skin does not necessarily need tobe included in the CTA to have the compositetissues satisfy the definition.

The First Successful HandTransplant (CTA)

The recipient of the first successful human handCTA had sought a possible hand transplant inseveral countries and joined a group of similarone- and two-handed amputees in Australiahoping for transplants who were otherwise fitand healthy throughout 1997. Up until Septem-ber 1998, they had been tutored by one of us inSydney, Australia, on all elements of the possibleapproaching procedure.

Their extremities were color photographed,measured, X-rayed, angiograms were taken andtheir forearm muscles, tendons, and nerves care-fully examined, tested, and well mapped. Theywere commenced on forearm muscle strength-ening exercises and their motives, states ofmind, determination, flexibility, personalitystrength, sanguinity, and absence of family psy-chiatric history evaluated by psychiatric andpsychological specialists.

Their complete medical history and routinepre-transplantation blood and serum biologywere assessed. Apart from the lecture dis-cussions these potential recipients were givenhomework of all the relevant review articlesand books on transplantation to enlightenthem thoroughly on the ‘‘pros and cons’’ ofwhat they were about to undertake.10 Their legalposition as regards to possible complications,and their legal agreements to actually give theirfull permission and co-operation to follow theproposed protocols were drawn up and theygave their signed agreements. This includedthat they set up a system for the early monthsof post-operation care, when they would need tobe able to contact their treating doctors in anyemergency.

The teams so far performing hand and faceCTAs have had healthy recipients, who are fit anddisease free, highly motivated, and co-operative.Being quite healthy, they differ from the usualdebilitated and infirm allograft organ recipient.They have agreed that they will fully co-operatewith their physicians and therapists to give them-selves and their team the best chance for theallograft to succeed. They realize how importantit is to stay closely in touch with their treatingteam members with regular follow-ups.

By June 1998, the five most appropriateamputees, four men and one woman, wereadjudged ready to not only be knowledgeable


and eager candidates for this possibly physicallyand personality disturbing ordeal, but also beready-to-board an aircraft to fly to Europe at thereceipt of a phone call signifying that a suitablycompatible donor was available.

The international team, four already in France,two in Australia, and one each from Italy andEngland met by arrangement in Lyon andawaited an appropriate donor. While waitingthey rehearsed the operation procedures withthe various already organ allograft-experiencedphysicians, nurses, therapists, operating roomstaff, and administrators. Prior to this alignmentof experienced expert talent, the permissionsrequired to actually plan and carry out thisprocedure had been sought, explained, and even-tually approved by all the authorities involved.This actually involved people from the highestauthorities in the French Parliament throughNational, Civil, State, local Government, Univer-sity, and Hospital managements and even includ-ing the obtaining of the approval of the FrenchPatient’s Protection Association.

We were privileged to have had the co-operation of the France-wide organization ofthe French Ministry of Health’s EtablissementFrances des Greffes, wherein the well-trainedand smoothly functioning seven sectional teamsof doctors and nurses are alerted to all France’spossible organ/tissue donors. This network(which also connects to the Eurotransplant sys-tem) keeps the thorough computerized detailsof the potential donors online and matchesthem with appropriate recipient waiting lists.It not only swiftly and efficiently notifiesthe awaiting transplant teams who harvest thedonor materials, but also delicately andcompassionately approaches and comfortsand supports the donor’s relatives at the pointof donation.

Previously in Australia, discussions for ahand transplantation from a dead man to a liv-ing man had been arranged, debated, andapproved by the most senior Australian repre-sentatives of all the world’s major religions.

Over the past few years, in individual coun-tries as well as international gatherings of trans-plantation experts, the feelings invoked by theconcept of hand and other CTAs had swungaround to show less opposition and finally defi-nite statements that the ‘‘time was now ripe’’.11

Most of the debates were over the obvious factthat a hand or face transplant was not a life-saving one but rather a ‘‘quality of life’’ enabling

and effecting procedure. The other contentioushurdle was the argument that the powerful drugsto prevent rejection had very serious side effectssuch as diabetes induction, or the increased sus-ceptibility to infection and cancer, and evenperhaps other life-threatening conditions, andwould possibly need to be taken for the rest ofthe recipient’s life.

It is often thought that there must be a bigdifference between CTAs and the now morefamiliar and regular ‘‘solid’’ organ allografts.But, are the solid organ allografts actually basi-cally only the one tissue? With very few excep-tions all the cells of all tissues have both a bloodsupply and have or are within close range of anerve supply. A ‘‘solid’’ organ such as a liver haskilometers of its three main tubular structures,arteries, veins, and ducts, as well as nerves andconnective tissues and a remarkable variety ofdiffering sheets of specialized liver cells carryingout unique manufacturing, reprocessing, andexcretory functions. The exploration of themicroanatomy, physiology, and biochemistryof the amazing variety of liver cells is far fromcomplete, as is the knowledge of its recuperativepowers, yet the success of liver allografts with allits many tissues was the basis for our drugregime for our CTAs.

Before we decided on the actual date to carryout the preparations for this, our first single-hand allograft, we were encouraged that aCTA’s rejection of the hand with its basicallyuntried human-skin-grafted component couldbe controlled by our selected cocktail of immu-nosuppressive drugs by several earlier groups’successes. Giumberteau and Baudet’s team hadsuccessfully allografted a digital muscle flexionsystem for the hand in 1989,12 and Hoffman andKirschner had successfully begun allograftingknee joints13 and the femoral diaphysis14 follow-ing the lead of Chiron’s adventurous operativetechnique.15

Early attempts to transplant skin across evenminor histocompatibility barriers, first, withinthe same animal species by Medawar, had pro-ven how difficult it could be.16 In the following50 years of immunology research when skintransplants and allografts had been attempted,it was proven that skin tissue rejection was themost difficult rejection to control. We knew thatthis could be our major hurdle. The team leaderhad to deal with the other members of the inter-national team who were keen to start our CTAseries (we had permission in mid-1998 to go

Composite Tissue Transplantation: A Stage Between Surgical Reconstruction and Cloning 167

ahead with the first five cases in France), with abilateral amputee, to refute the possible objec-tions that to do just a one-handed allograftmight be considered an unnecessary procedureand that a single-handed man could still carryout a productive and worthwhile occupation.The team leader argued that the area of skin ontwo hands might prove to be too much skin areafor the chosen drugs regime to protect, and sothere was a much greater chance for success ifonly one hand with its already considerable sur-face area of skin was attempted. As we had in ouroverall team a leading French dermatologistwith transplantation experience and the use ofan antirejection drug ointment available to bol-ster skin rejection suppression if necessary, wefelt confident to progress to obtain an appropri-ate donor. This took longer to arrange than wehad anticipated despite the fact that in France,all citizens who die are eligible in French law tobe donors of organs or tissues unless they havespecifically written in to the Department ofHealth to refuse such permission.

Whereas an organ allograft is not seen by therecipient, a hand or face allograft is seen as soonas the patient awakens from the operation’sanesthetic. So not only the donor’s blood typeand, preferentially, tissue types should match,but so should the skin color, texture, and hairdistribution as well as the actual size and shapeof the donor parts.

Surgical Technique

Human ingenuity encompassed the trephiningof skulls to release cerebral hematomas some5000 years ago in Egypt, as testified by thespecimens viewable in Medical Museumstoday. The late 1950s, the 1960s, and early1970s were exciting times for experimental sur-geons worldwide and the early enthusiasts fromall over the world corresponded and formedmeetings to start the International Microsurgi-cal Society and the Transplantation Society,both in 1967, which saw the coming togetherof extraordinary medical and scientific minds.So in the past 50 years skillful surgeons havereplanted, reconstructed, and repositionedcompletely severed human bones, tendons,nerves, muscles, intestines, skin, tongue,limbs, joints, hands, digits, scalps, ears, lar-ynxes, and penises, requiring delicate anasto-mozing techniques.

CTA recipients are subjected to all routinepre-transplant investigations with the addi-tion of the suitable special tests to specificallydefine the available functional anatomy andphysiology and thus the perceived capacity toactivate the allograft post-operatively. Thedetailed surgery of those first successful sin-gle and double forearm and hand allograftshas been well described by the members ofthe French/Australian and the Austrian,American, and Chinese teams.17

The actual surgical procedure in the Septem-ber 1998 operation was to first carefully dissectand mark the donor’s end tissues, whose bloodvessels had previously been perfused with coldUniversity of Wisconsin’s solution at harvesting.The lower forearm donor bones were osteoto-mized at the chosen lengths to match the lengthsto the recipient’s desired both arms of equalnatural length. After application of a tourniquet,the donor’s and recipient’s bones were fixedtogether with metal plates and 4.5 mm screwsto stabilize all further surgery. The radial andulnar arteries and then the cephalic and thelargest veins were available and were carefullyanastomosed with 8/0 microsutures. The pre-viously applied tourniquet was released and thegraft reperfused with recipient’s blood. Thistook slightly longer than we expected, as thedonor tissues had been cold preserved for 12 h,so we warm wet wrapped the area once it was innormal color and left it to warm up while wetook a 20-min break. After a celebratory hotdrink, we rescrubbed and gowned and returnedto carefully suture the tendons and muscles inlayers and microsurgically the ulnar and mediannerves. This was done 20 and 21 cm, respec-tively, proximal to the wrist skin crease, to takeadvantage of the shorter healthy remaining reci-pient’s and the freely excessive amount ofdonor’s nerves. To encourage strong boneunion, we harvested iliac crest autologous can-cellous bone chips and spread them around theosteosynthesis sites. The recipient’s skin wassutured to the generously shaped donor skin,even including a wedge of split skin graft fromthe recipient’s right thigh, as we had learnedfrom 28 years of arm replantations around thedistal forearm to allow space for possible initialswelling or compartment syndrome. We left twoPenrose drains in site, the wrist slightlyextended, the elbow flexed 458, and the distallimb gently bandaged, splint supported, andelevated.


Obviously more donor tissue than would berequired was harvested in each case as the actualcondition of the available viable and actionabletissue in the recipient’s limb stumps was neverprecisely known until seen at operative dissec-tion. In over 28 cases of the so-called ‘‘handallografts’’ now surviving it would appear thatlonger lengths of donor nerve tissue than wereoriginally planned in these cases was routinelyused and we believe this does give a better even-tual functional result, as recipient’s nerve tissuehas usually suffered functionally regressive par-tial degradement.

When the patient awoke and immediatelysaw his hand he exclaimed, ‘‘I have my handback!’’, but it was not until the local anestheticto his right axilla had worn off and he tried outgentle flexion of the fingers that he was con-vinced. Physiotherapy for two sessions a daystarted only 10 h after the surgery and rehabili-tation by controlled motion passive- and active-supervised exercises were continued for3 months. Psychological support daily was lessintensive after 3 weeks had passed. Skin biopsieswere taken once a week together with a programof close observation in case it might be neededmore frequently, should commencement ofrejection be suspected. No surgical complica-tions occurred in this case, who was correctlysupervised in Lyon for the first 3 post-operativemonths. He then went off, by himself, quitecontrarily to the pre-operatively psychologicalassessments by experts both in Australia and inFrance and went his own way. He would notkeep regular follow-up sessions, could only becontacted by mobile phone, and that too occa-sionally. He learned about the onset of mildrejection by observing his mild pink skin rashand treated himself by increasing his daily doseof prednisone for 1 or 2 days until the rashsubsided. At the occasional times when one ofus could examine him, we were surprised howrapidly the patient’s nails and his ulnar andmedian nerves grew. These were checked firstby Tinel’s sign, and then with delicate nerveconduction studies as the feeling grew down tohis fingertips in the first 12 months. At 2 yearspost-operation he was found touring the UnitedStates where he deliberately then ceased hisimmunosuppressives, and then demanded thatthe team amputate the slowly rejecting part.When this was carried out 5 months later onthe still very healthy patient some 28 monthspost-operation, and the transplanted tissues

were then examined, only the skin was actuallyproved to be rejecting. All this team’s later dou-ble hand allographs were far more co-operativeand have done very well.18 They use their handsand fingers well at daily tasks and at work andreport that their lives have been transformedback to being normal.

One of the most intriguing understandingsthat has come out of the hand transplantationprogram is the insight science that has gainedinto the way the brain can adapt to changingcircumstances. That there really is brain areaplasticity was shown in progressive fMRIs frompre-operative to post-operative observationsthat peripheral input can modify cortical handrepresentational organization in both sensoryand motor regions. Relatively blank areas ofcortex in the previously amputated patient’sbrain regained activity as the allografted handsregained activity when seen as early as 6 monthspost-operatively and continued to be more acti-vated as the movements and feelings matured.19

Immunology

The choice of immunosuppression in our handallograft program was not an easy one. The teamdiscussed many possibilities and rejected pre-operative complete irradiation of the recipientsbone marrow, but settled for commencing onantilymphocyte antibody just prior to theattachment of the CTA and then utilizing whatwe called a ‘‘cocktail’’ of well-proven drugs thatsustained liver allografts. Several followingCTAs in the United States, Austria, and Chinaalso adopted this ‘‘starting with a clean plate’’(having cleared out the recipient’s lymphocytes)approach which had been well tried in manyorgan allograft programs worldwide.

That and the current immunosuppressiveprotocol includes loading doses of Thymoglobu-lin1 (polyclonal antibodies) 1.25 mg/kg/d forthe first 10 days, and then three main drugsafter loading doses were continued with

• Tacrolimus (FK506)0.2 mg/kg/day keepingblood levels between 15 and 20 mgm/kg/dayin the first month;

• Mycophenolate mofetil 2/g/day;

• Prednisone 250 mg on day 1, 1 mg/kg for 10days, then tapered down to 20 mg daily,reduced to 15 mg/day at 6 months; and

• wide-spectrum antibiotics for the first 10 days.


Should early serum sickness occur, monoclonalantibodies can be administered in short courses.In some patients transient hyperglycemiaoccurred which was treated initially with insulinI.V.I. for a short time.

We somewhat suspect that a mechanism thatprotects these allografts from chimerism andgraft versus host disease and perhaps evenchronic rejection is involved. To that end cur-rent papers accepted for publications of our ownand others’ research works with co-allograftingvascularized lengths of donor femurs with ratlower limb allografts is encouraging as we havefound that larger amounts of transplanted donorbone marrow lessen the need in the rat model forimmunosuppressants. Bone marrow transplan-tation in order to create a chimerism has been afeature of our more recent allografts and those ofother authors.

Although it is well known that major histo-compatibility (MHA) antigens should ideally bematched in any allograft donor/recipient pairour first four hand allografts cases, one single,and three double hand cases (totaling sevenallografted hands) did not have even the luxuryof one of the major MHA matches. Both serolo-gic and molecular genetic testing of matchingMHAs is available and so the CTA teams arehopeful of the sort of help with allograft outcomethat has been shown to occur when there ismatching for HLA-A, for -B and for -DR inkidney organ allografts will occur in futureCTAs. We recognize that 1% of the generalpopulation, without having had a blood transfu-sion, can have preformed antibodies that couldreact to alloantigens,20 and of course check bothABO compatibility and check for any history ofprior immunological stimulation, but neitherhave nor had expected any hyperacute rejectionepisodes on our regime.

We recognize that a chronic rejection processmay develop as it can in nearly all solid organallografts, with the usual vasculopathy describedearly on as a transplant arteriosclerosis.21

We have not yet had biopsy evidence of anysuch chronic rejection so far but as we necessa-rily allograft active blood-generating donorbone marrow in these cases we are aware it hasbeen shown to continue producing blood cells.22

The best descriptions so far in the literatureon the normal, the allografted, and the allograft-rejecting skin’s three cellular layers have beenclearly outlined in detail and illustrated by ourDermatologist Jean Kanitakis in Lyon.23 There

are no really detailed reports available on otherCTAs such as the larynx and the knee joint;however, biopsies when they are reported showthe same open-vascular infiltrate of mostlyT-cell lymphocytes that we have all been familiarwith since the earliest skin allografts in micemore than 50 years ago.

In some hand-allografted patients early skinrejection occurs, seen as a distinct maculo-papular rash of pink lesions. They have a mild,nonitchy inflammatory response, and biopsiesshow a peri-vascular and dermal mononuclearheavily staining recipient cell infiltration. Thisrash is easily reversed by increasing the oralsteroids and applying topical immunosuppres-sive ointment. Maintenance doses were givento keep tacrolimus blood levels between 5 and10 mg/ml, with Solupred1 at 5 mgm/d andmycopherolate mofetil at 2 g/d.

The phenomenon of the pink rash acts as awarning system that the immunosuppressionregime is not optimum. It was first noted bythe first single-hand allograft recipient, whobegan to use the appearance of pink spots as ameans to judge his own treatment. He was ableto conserve his drug supply by voluntarily les-sening his doses in the second year post-opera-tion, until the rash appeared. He would thenadjust the doses upward and again stabilizebefore gradually dropping the dosages again.This, of course, was at the stage where he wasstaying away from his medical team for longperiods, as he was by no means an ideal patient.It may be recalled that after 16 months with hisallograft, one of us had to appear on worldwidetelevision to appeal to him, wherever he was, tocome back to any one of us in any of our coun-tries for his delayed follow up and to obtainmore drugs. Owing to his nonconformity, thispatient allowed skin rejection to progress so thatone of us had no choice but to amputate at28 months. The histopathological examinationof the allografted terminal forearm and handwas remarkable as the only microscopic signsof rejection were those of the skin.24 The patienthad no other body signs or symptoms and wasotherwise fit after this amputation.

The Face Allograft

The 2006 face allograft in Amiens, France,involved full thickness facial soft tissue finereconstruction of a woman’s face over dry bone


in a large oval defect measuring 16 � 15 cm2

from just beneath her orbicularis oculus musclesand above the nose to below the chin and jaw.This flap of tissue between skin and mouthmucosa consisted of tendons, muscles, fat,motor and sensory nerves, and vessels. Thedeceased donor tissue including the nose ofsimilar skin color and shape to that of the reci-pient provided a fully functioning face and alsoreturn to a normal life for the delightedrecipient.

Some years ago a Hollywood film called FaceOff brought to the general public a ‘‘hot’’ con-troversy that had been previously challengingTransplantation Units worldwide. In the filmthe faces of a ‘‘bad guy’’ and a ‘‘good guy’’ whowas chasing him were first cross-allografted(‘‘swapped’’), and then ‘‘swapped’’ back again atthe film’s end. This would have given the publicthe idea that such operations were quite feasibleand technically easy, with no mention of immu-nosuppression, but such intricate surgery is farfrom simple.

The deformed or damaged face has been thesubject of many clever techniques of repair, andplastic surgery history abounds with complex-staged procedures to reconstruct parts of faces.The success of the first 28 hand allograftsshowed that large areas of skin could be allo-grafted possibly allowing a CTA for those whosefacial damage could not be functionally cor-rected by even the most experienced Recon-structive Plastic Surgery teams.

Our international team considered in 1999that, following our and the world’s experiencewith CTAs, we would consider a face allograft if asuitable case occurred and came to our atten-tion, and such a case fell within the group oftissues we already had obtained permission toreconstruct by CTAs. The history of reconstruc-tions of small parts of the face goes back thou-sands of years and includes the use of tubepedicle skin transfer almost 500 years ago bythe Italian Gasparo Dagliacozzi (1557–1599).We remember the heroic efforts of plastic sur-geons to restore skin cover and shape to theseverely burned and hideously disfigured facesof those pilots in World War II, who survivedtheir cabin fires, as the work of Sir ArchibaldMcIndoe at East Grinstead Hospital, West Sussexin England, was detailed in films and reports.25

We considered the 37 published cases ofreplantation of the scalp that we found in theliterature of our Microsurgical colleagues since

the first report of successful replacement of anavulsed scalp.26 This had been followed byincreasing numbers of reports of scalp replants,which included as well as scalp such differingtissues as with a complete ear cartilage in 1978,27

and other ‘‘skin plus’’ tissues in 1983 such as thepenis, nose, ear, and scalp.28

Applications to actually perform clinical facetransplants had been submitted to two otherFrench (2004) and US (2006) authorities andhad received approval before the case of the38-year-old woman with severe facial tissueablation was referred to the Facio-Maxillary Sur-gery Department in Amiens, France, in May,2005.

The controversies about a possible ‘‘FacialTransplant’’ had been extensively discussed bytransplantation units for years and had led toethical committees of some countries and long-established surgical colleges investigating andissuing informed reports on the subject in theearly 2000s.29,30

In this case, a large family dog had found hismistress in a deep sleep and savaged her face.The almost oval bite, measuring 15 � 16 cm2

from just below the lower eyelids to below thejaw encompassing all tissue, including the nose,down to bone and teeth, was totally unsuitablefor replantation when discovered. The horrificgap, unequally present on both sides of the mid-line, would, untreated, inevitably fibrose andretract and was treated with wet dressings.After confirming the French authorities’ agree-ment to go ahead, the search for a suitable donorwas notified to the Etablissement Frances desGreffes and the Lyon and Amiens surgeonsbegan detailing and practicing the proceduresnecessary to provide a sustainable and accepta-ble CTA. The search for the right donor wasagonizingly long for both the patient and thereadied operative teams, as the retraction ofthe oval-shaped defect’s edges constricted tosuch an extent that the apparent gap lessenedconsiderably as the upper and lower teethbecame almost tightly pressed together, whenfeeding by mouth and talking became a greatdifficulty.

It took almost 7 months for a suitably com-patible donor to be discovered and the operationcommenced on November 27, 2005, in ProfessorBernard Devauchelle’s Maxillo Facial SurgeryDepartment by the team jointly led by him andProfessor Jean-Michel Dubernard. The intricatesurgery required motor facial and sensory


trigeminal nerve branch anastomoses and mus-cle and tendon and nasal reconstruction, well-practiced facial vascular anastomoses, and gentleplastic surgical skin alignment together with thebold immunosuppressive regime. The outstand-ing result is now apparent in the almost comple-tely functioning and esthetically delightful result2 years later.

The immunosuppressive drug combination,now well tried in hand CTAs was followed but inthe 10 days after the operation two separateinjections of donor bone marrow hemopoieticcells were administered, even though this boldattempt to invoke a micro-chimerism hadonly been tried experimentally. This triumph ofsurgical and immunosuppressive virtuositydemonstrates how co-operation among experi-enced specialists can achieve what may haveseemed impossible only a short time before.

Laryngeal Transplantation

The human larynx has evolved as an extraordi-narily complex of cartilages, mostly coveredwith mucosa, joints, muscles, and ligamentsand an efficient large nerve and vascular supply.Everyone can remember at least one episode oflaryngeal discomfort, such as when coughing, ora feeling of choking, or of having a sore throatand not speaking with one’s normal voice. Justunderstanding how the anatomy and physiologyof the larynx work is extremely difficult even forthose who have studied medicine. The internalmuscles shorten, lengthen, or relax the vocalcords, so we can speak and sing, whereas theupper muscles can raise and lower the Epiglottisand direct food posteriorly to the esophagus toallow the beginning of swallowing, or else allowair from both the mouth and the nasal passagesinto the trachea and on into the lungs, usuallywithout mixing up these two essential functions.

Such a complex dual purpose essentialmechanical instrument mostly works automati-cally, so when it misfires from just one sectionaldefect the physician is usually quickly consulted.If a disease such as a laryngeal cancer has pro-gressed so that laryngeal excision is required, itis comforting to know that extensive research onlaryngeal allografting has occurred since1966,31,32,33 and a partial CTA was surgicallyperformed in 1969,34 and a successful completelaryngeal CTA in 1970.35 As the first was mainlyan attempt to remove cancerous mucosa and

graft in some vascularized mucosal tissue, itwas at first seen to have survived on the immu-nosuppressive regime. Unfortunately canceroustissue, presumably from the cartilaginous bed,did overwhelm the mucosa and the patient died8 months later. It is presumed that the higherdosed, less-sophisticated immunosuppressionthen used might have encouraged the conditionsto allow freer tumor cell multiplication in thedrugs causing an absence of any body celldefense mechanism. The disappointment ofthis case dampened down enthusiasm for suchallografts. There is no man-made substitute yetfor the quite bewilderingly intricate complexthat is the human larynx, capable of such extra-ordinary control of the vocal cords, swallowing,and diameter of the shared passages for air andfood, but research continued.

The first successful laryngeal CTA occurred30 years later when in 1998 a very dedicatedgroup of Cleveland Clinic throat specialists per-formed a CTA with a large fully matched com-plex of the larynx and its pharynx as well as thethyroid and parathyroid glands, and upper fivetracheal rings. This also required anastomosis ofthe upper laryngeal nerves and one recurrentlaryngeal nerve with its complex’s completeextensive blood supply. A 40-month follow-upbiopsy of the trachea was normal. That such asurgical triumph worked and to date is stillfunctioning is most encouraging.

The immunosuppressive regime was similarfor that used in the other CTAs, except thatcyclosporine rather than tacrolimus was initiallyused in the basic three drug-sustaining cocktail.Tacrolimus was introduced after a rejection epi-sode began at 15 months post-operation andsince then the drugs have been steadily reduced.

This recipient, a no skin, but otherwise veryvaried composite tissue transplantation, suppo-sedly used his ‘‘new’’ voice 3 days post-opera-tively, and is now, over 8 years later successfullytalking, eating, and breathing within normallimits, within a one octave but quite serviceablevocal range. This operation is, like those handsand face transplantations, excellent examples ofCTAs providing near normal quality of life out-comes, which would not have been otherwiseconventionally surgically possible.

The success of this extraordinary CTA hasbeen repeated 16 times since. These highly tech-nical procedures require in the one institution afunctioning, multidisciplinary, and co-operativeteam led by strong, confident, and excellent


specialist surgeons, immunologists, physicians,and paramedics, all prepared to provide pro-longed obsessive care at every step of the lifelongjourney of the patient. The way forward in thisfield was best presented by a review in 2006 bythe US group.36

Uterine CTAs

In April 2007, the First International Sympo-sium on this subject was held in Gothenburg,Sweden. Since the first technically successfulattempt at Uterine CTA was performed37 inSaudi Arabia in April 2000, much furtherresearch has occurred, yet a fully successfulhuman uterine transplantation has not beenreported.

That first case proved that a uterus from aliving donor, together with its fallopian tubes,could be technically implanted in the recipientafter a hysterectomy, and actually later respondto combined estrogen–progesterone hormonetreatment causing thickening of the endometrialuterine tissue bed and withdrawal bleeding. Theattempt was abandoned requiring a further hys-terectomy after 99 days due to massive necrosisof the uterus due, no doubt, to vascular occlu-sion. The original surgical approach from aliving donor precluded the use of its similarlength uterine vascular supply, so saphenousvein grafts were employed to the uterine arteriesand veins on both sides and the blood shunted toand from the internal iliac arteries of the recipi-ent. The absence of adequate structural supportfixing the donor uterus and tubes in place mayalso have contributed to the vascular occlusion.The immunosuppressive protocol, similar toour CTAs, was quite successful as there was noevidence of rejection in the harvested necroseduterus.

In the time since then a great deal of subse-quent research has rationalized and optimi-zed the donor organ harvesting, from either aliving altruistic woman (a technically difficultvascular approach), or a fresh cadaver where alarge dissection is required to obtain sufficientlength of large blood vessels might prove moreappropriate.

The uterine transplantation program must beall inclusive as regards medical ethics, national,and international laws be able to be fully evalu-ated as regards both donor and recipient selec-tion and follow the proper assessment of the

efficiency of the differing drug regimes onthese tissues. Then, there are the problems ofthe effect of immunosuppressive drugs on thedevelopment of the offspring-embryo, fetus, andbaby, even including the effect of drug dose bybreast feeding. The problem of detailing earlyrejection in the pregnant uterus and some of thedrug-related uncertainties has been studied inwomen who became pregnant while on immu-nosuppressives for other organ transplants.

As it is such a tragedy when a woman’s abilityand desire to conceive children is prevented byvarious uterine abnormalities, the ingeniousresearch protocols in the literature, with thosevoiced at recent symposiums make it seeminevitable that fully successful uterine/fallopiantubes CTA will soon be performed. There arehundreds of research articles carefully debatingthe controversies involved in this field, of whichseveral seem the most pertinent and discuss thepressing issues such as operative techniques anddangers to both the women and their offspringdue to immunosuppressives, as well as the alter-natives to transplantation.38

Pregnancy outcomes after immunosuppres-sive therapy during pregnancy39 suggested morepre-term deliveries and congenital deformitiesthan the norm, but other studies in experimen-ted animals,40,41 do not reinforce this conclusioneven though some work suggests teratogeniceffects on animal offspring due to prednisolone.

Alternatives to CTAs for female sterility ofuterine origin now include many cases of suc-cessful surrogacy. Researchers have grown uterineand other specialized tissue in rabbits, and usingembryonic human stem cells and tissue engi-neering can now grow most human-specializedorgans and other tissues,42 including uterineendometria.43

The moral, ethical, religious, medical, legal,individual country, as well as the administrative,technical, child, and gender issues in thisabsorbing field all come together as the uterus,so far in human evolution, has been mandatoryto civilization. We will hear far more about thishighly debatable and essential topic in thefuture.

Peripheral Nerve Transplantation

Peripheral nerve defects are and have beenrestored as whole, cable, or fascicular replanta-tions, usually utilizing one or even both sural


nerves since 1970s.44 In 1976, the first vascular-ized nerve graft was performed by Taylor andHam in Australia.45 This early attempt was notimmediately repeated until greater numbers oftrained microsurgeons came on the scene toperfect fascicular and small vessel anastomosesunder the operating microscope. It was foundthat increasing lengths of gently harvested, non-traumatized sural nerve alone could provide thepathway for the regrowth of a patient’s own longperipheral nerves, such as, the median, ulnar,and sciatic where no immunosuppression wasrequired. This was more successful in childrenwhere one of us (E.R.O.) used lengths of suralnerve to bridge a 12-cm gap using nerve anasto-moses between the sciatic nerve and the maintrunks of the lateral popliteal and posterior tibialnerves, with such function that the child’s onlydefect several years later was the feeling to hisbig and second toe, and he grew up to walk andrun normally.

Nonetheless, extensive peripheral nervedefects where using the patient’s own nerves isimpossible or impractical can require a donor-vascularized nerve.46 A 10-year study of theresults of allotransplants for both upper andlower limb traumas have been studied, primarilyin the McKinnen team from St Louis and by ourHand Transplantation Team (pp. 29, 24, 25, 31).Patients can indeed regain almost full sensitivityand function with donor nerves also anastmo-tized with their supplying arteries up to lengthsof 19 cm on the routine three main immunosup-pressive drug regime. Positive Tinel’s tests areused to plot the down growth of viable nervetissue and on an average the drugs can be tailedright off in 18 months. That way iatrogenic com-plications can be avoided.

We have described in our hand CTAs, as hasDr S. McKinnen’s work shown for many years,that with some immunosuppressive drugs, par-ticularly Tacrolimus, the advance of nerve tissueand its activity down the grafted nerves occursfaster with the drugs than it does with or withoutan accompanying blood supply.47,48,49,50

The presence of a sustaining blood supply fora nerve graft to survive and duly function isessential. The longer the nerve length requiredthe less likely it is to pick up lengthy and ade-quate blood supply from the surrounding tissuealone, hence the need for concomitant vascularsupply. There are, of course, other factors inperipheral nerve transplantation. Some of thegrafts are interpositional rather than terminal,

and so the graft’s rates of growth may be ade-quate, but when they reach the distally severednerve end that tissue will have had no recentnerve stimulation and may have fibrosed andbecome more difficult to re-innovate.

The preservation of nerves prior to trans-plantation and the effects of the drugs on thenerve tissue and Schwann cells are becomingincreasingly important.51 While most CTAsand organ transplants require relatively largeblood vessels, needed to supply to large amountsof tissues, nerve grafts and transplants use muchsmaller blood vessels, yet the tissues making upa peripheral nerve are small and tightly packedwith their essential capillary blood supply whichis so easily obstructed by even slight pressurefrom surrounding tissue pressures.

It must therefore be emphasized that in anyCTA requiring an essential nerve componenttransplantation, the normal surgical principlesof any delicate surgery, particularly those con-cerning the absence of pressure and the use ofgravity drainage (if at all possible), must be evenmore rigorously observed than is usual.

Tongue Transplantation

Dr Rolf Evers, speaking for the Vienna GeneralHospital team led by Dr Christian Kermer andDr Franz Watzinger that carried out the firstsuccessful total tongue transplant was happywith the initial result. He was quoted worldwideon television stating that ‘‘other organs such asliver and kidneys are complicated, but the ton-gue is just muscle’’; however, we believe it to be aCTA. For it to function as a tongue it must havesignificant vascular and nerve tissues trans-planted with it. The transplant, done in July2003 following removal of a large carcinoma ofthe tongue and jaw of a 42-year-old man, showedit to be technically possible. We believe that thisbrave attempt to assist a stricken patient was notfully successful and have not as yet seen anypathological studies. We respect the delibera-tions of the Royal College of Surgeons of Eng-land who in 2003 issued a Working Party Reportsuggesting much more research into manyissues should go into tongue and face andother potential CTAs, and to date no more ton-gue transplants have been reported.52 It wouldseem that in general terms it would be perhapscounter-productive to contemplate a CTA in apatient with an advancing tumor, as the


necessary immunosuppression drugs wouldonly accelerate the growth of secondary tumors.

Penis Transplantation

The technicality of transplanting a donor penishas been well proved by the many penis replan-tations carried out since the 1970s. One of us(E.R.O.) replanted the completely amputatedpenis of a nine-year-old boy at the Royal Alexan-dra Hospital for Children in 1970, which was notpublished and only briefly mentioned in thatHospital’s Annual General Hospital Report inthe Research section, at the request of the child’sparents. He had been masturbating with a rela-tively narrow necked milk bottle in the bathroomand slipped and fell, leaving only a 1-cm stump.Brought to the hospital within an hour, and withthe experience of having successfully alreadyreplanted a 2-year-old boy’s completely ampu-tated index finger months before, the operationproceeded as a succession of anastomoses. Thesewere the urethra first and the deep fascia about itof the corpora spongiosa, both deep penilearteries, the two corporal cavernosa’s peniletunica fascia, the deep dorsal vein, and the finedorsal nerves on either side, and the very smalldorsal arteries beside, then the large dorsal veinof the penis. Some deep fascia was next and thenthe skin was joined with interrupted buriedsutures of fine microsurgery nylon with endscut very short, which was left inside. The boywas sedated and gently cool packed for over aweek and developed into a very well-behaved boyand man, capable of having quite normal erec-tions. One of us, Jean Michel Dubernard com-bined with Dr Gelet at The Edouard HerriotHospital complex in Lyon, France, performedanother functioning and normally sensitizedreplantation, also joining the dorsal nerve of thepenis. These replants were not publicized, and weredone, as were some other successful replants inChina and Taiwan, long before a highly publicizedpenile replant was completed in the United States.

The recent first penis transplant, a Chinesecase,53 had specialized microsurgery, someimmunological support, but had to be reampu-tated after 3 days. It seems, because it is techni-cally possible and because the CTAs so far haveshown that allograft skin can be defendedagainst rejection, that it is only a matter of timebefore other cases of penile allografts will becompleted and announced.

Conclusion

The second successful single-hand CTA is now8 years on, and our first three double hands allo-grafts at seven, three, and almost 1 year post-operation are functioning very well indeed. Over28 hand allografts and many other CTAs havenow been performed in many countries.

The initial hand CTAs achieved the goals ofmotor and sensory function equal to orapproaching those of properly replanted trau-matically injured hands, and certainly superiorto that of myoelectrical prostheses. Patient satis-faction and personal delight in having workingesthetically normal limbs again with good socialintegration and work potential justifies thecostly labor-intensive procedure. The drugdoses, although potentially having debilitatingside effects, can be greatly reduced over time andcan be well controlled by the astute patientunder supervision. When rejection episodesoccur, the skin of the donor is the ‘‘early warningsystem’’ that let the patient know to report backfor simple known reversal measures to resumetheir balanced state. The face CTA in France ispresently 26 months post-allograft, with anexpressive face, going about her daily life pleasedto be a normal woman. The scars are not visiblewith light make up, and the patient’s drugs arealready greatly reduced. The first laryngeal CTAis into his 8th year being very satisfied with hisextraordinary allograft which functions vir-tually normally. A partial Hand CTA consistingof half the palmar tissues and all four fingersstill in its first post-operative year has beenseen to be doing well in China, The Chinesepenis allograft is reported to be well satisfiedwith ‘‘his’’ restored manhood. Unfortunately webelieve that all the attempted knee joint CTAshave not survived.

That donated human organs and other bodystructures, such as tendons, muscles, nerves,and intestine can also join skin in being able tobe guarded from rejection is heartening anduseful as there is no other way for certain tissueloss replacements to be replaced at this time.There is thus a present need for some well-prepared patients to have CTAs to enable themto have a life worth living. All the argumentsagainst the considerable risks entailed in CTAshave been talked out and written about, andthose frontiers have been crossed without lossof life, or worsening of the situations. Experiencewith years of careful following up of CTAs shows


just how useful they are and how grateful are theindividual recipients when the protocols arescrupulously carried out.

We do believe that in the most deserving ofselect cases in a well balanced, otherwise healthybody, the potential benefits of hand and certainother CTAs outweigh the risks of early infectionand late possible malignancy, or early onset ofdiabetes. We have found out that technically andtheoretically the donor-to-recipient operationshould be, but not always is, easier than thereplacement of severed parts to the same body.This is due to the form of injury and the responseof peripheral nerves to severance and trauma,which cause a ‘‘die back’’ phenomenon withshrinkage and also tissue swelling that hampersnormal nerve conduction on the recipient side.

We believe that face CTAs are the correctform of treatment for a small select group ofpatients that cannot be otherwise adequatelyreconstructed by even brilliant conventionalplastic surgical departments.

The more we know about cell metabolism, themore we realize how much we do not know. Yethow each cell functions is the key to its healthand survival, and therefore our survival. CTAsprovide huge collections of living cells to replacedead, diseased, cancerous, or amputated parts ofour bodies. Surgery, from the point of view of thesingle cell, has always been a gross and clumsyprocedure. The more we can actually see whenwe operate, the more precise can be our applica-tion of our surgical principles. The MonocularOperating Microscope, originally used by Nylenand Holingren in Sweden54 in 1921, allowed sur-gery nearly 90 years ago on the ossicles and thelabyrinth of the ear. Later developments led tobinocular-operating microscopes with foot,mouth, and other remote controls with brilliantlighting allowing for smaller and smaller precisemicrosurgery techniques. The advances since inmicroscopy allow scanning of single living cellsand even manipulation within those cells.Science now discovers particles that are smallerthan the essential parts of a single atom and isalso researching the genetic composition as wellas the actual structure of the plethora of humancell components. Biochemists are engineeringways of influencing (turning on or off) the activ-ities of molecules within a cell component andthat could be the ‘‘surgery’’ of the future.

Perhaps surgeons will eventually be neededonly to repair gross injuries, such as happen inaccidents, wars, and catastrophic natural

disasters. As the population of this globeincreases, those with the available medical facil-ities and necessary food and water live longerlives than the previous generations.

As future genetic engineered babies grow upwithout congenital diseases or proclivities, andautomatically diagnosed ailing cells are replacedwith technologically manufactured individualstem cell derivatives by minimally invasive newmethods by remote control, the readers of thisbook may wonder how long such a cellularlyengineered human being’s life span might turnout to be.

References

1. Carrel A. La technique operatoire des anastomoses vas-culaires et al. transplantation des visceres. Lyon Med.1902;98:859.

2. Carrel A. Results of the Transplantation of blood vessels,organs and limbs. JAMA. 1908;51:1662.

3. Carrel A, Guthrie CC. Uniterminal and biterminal venoustransplantations. Surg Gynae and Obst. 1906;2:266.

4. Gilbert R. Transplant is successful with a cadaver fore-arm. Med Trib News. 1964;5:20.

5. Dubernard JM, Owen E, Herzberg G, Lanzetta M, MartinX, Kapila H, Davahra M, Hakim N. Human hand allograft:report on first six months. Lancet. 1999;353: 1315–1320.

6. Jacobson JH 2nd, Suarez E. Microsurgery in anasto-moses of small vessels. Surg Forum; Clin Congr, Vol. II.American College of Surgeons, Chicago, 1960:243.

7. Fisher B, Lee SH. Microvascular surgical techniques inresearch with special reference to renal transplantationon the rat. Surg. 1965;58:904.

8. Lee SH. Experiences with organ transplantation in therat (series begun in 1961) IX. Implantation vs Transplan-tation of the Ovary. J Res Inst Med Sci, Korea. 1974;6:112.

9. Murray JE, Merrill JP, Harrison JH. Kidney Transplanta-tion between seven pairs of identical twins. Ann Surg.1958;148:343.

10. Hakim NS. Introduction to Organ Transplantation.Imperial College Press, London; 1997. ISMB 1_ 86094-025-0.

11. Seigler M. Ethical issues in innovative surgery: should weattempt a cadaveric hand transplant in a human. Trans-plant Proc. 1998;30 (6):2779–1782.

12. Guimberteau JC, Baudet J, Panconi B, Boileau L. Humanallotransplant of a digital flexion system vascularised onthe ulnar pedicle: a preliminary report and one yearfollow-up of two cases. Plast Reconstr Surg. 1992;89:1135–1147.

13. Hofmann GO, Kirschner MH, Wagner FD, Land W,Bukren V. Allogeneic vascularised grafting of a humanknee joint with post-operative immunosuppression.Arch Orthop Trauma Surg. 1997;116:125–128.

14. Hoffman GO, Kirschner MH, Wagner FD, Brauns L,Gonschorek O, Buhran V. Allogeneic vascularised trans-plantation of human femoral diaphyses and total kneejoints – first clinical experiences. Transp Proc.1998;30:2754–2761.


15. Chiron P, Columbier JA, Tricoire JL, Puget J, Utheza G,Glock Y, Puel P. A large vascularised allograft of theFemoral Diaphysis in man. Int Orthop. 1990;14:26–272.

16. Medawar PB. The Behaviour of skin autografts and skinhomografts in rabbits. J. Anat. 1944;78:176.

17. Francois CG, Breidenbach WC, Maldonado C, KaloulidisTP, Hodges A, Dubernard JM, Owen ER, Pei G, Ren X,Barker JH. Hand Transplantation: comparisons andobservations on the first four clinical cases. Microsur-gery. 2000;20:360–371.

18. Petruzzo P, Revillard JP, Kanitakis J, Lanzetta M, HakimNS, Lefrancol SN, Owen ER, Dubernard JM. FirstHuman Double Hand transplantation: efficacy of a con-ventional immunosuppressive protocol. Clin Trans-plant. 2003;17:1–6.

19. Giraux P, Sirigu A, Schneider F, Dubernard JM. Functionalcortical reorganisation after transplantation of both handsas revealed by fMRI. Nat Neurosci. 2001;4:691–692.

20. Scorpik JC, Saloman DR, Howard RJ, Pfaff WW. Evalua-tion of Antibody Synthesis in Broadly SensitisedPatients. Transplantation. 1988; 45:95–100.

21. Billingham ME. Cardiac Transplant Arteriosclerosis.Transplant Proc. 1987;19:19–25.

22. Granger DK, Briedenback WC, Pidwell DJ, Jones JW,Baxter-Lowe LA, Kauffman CL. Lack of Donor Hypore-sponiveness and Donor Chimerism After Clinical Trans-plantation of the Hand. Transplantation. 2002; 74:1624–1630.

23. Kanitakis J. Histopathology of Human CompositeTissue Allografts. Composite Tissue Allografts HakimNS, Owen ER, Dubernard JM (eds.). Imperial CollegePress; 2006.

24. Kanitakis J, Jullien D, Petruzzo P, et al. Clinicopathologicfeatures of graft rejection in the first human hand allo-graft. Transplantation. 2003;76:688–693.

25. Banks AH. McIndoe’s Lecture of 1941 on diagnosis andtreatment of injuries of the middle third of the face.British Dent J. 1996;181 (9):346–347.

26. Case Report. Plast Reconstr Surg. March 1969;43(3):231–4.

27. Nahai F. et al. Replantation of an entire scalp and ear bymicrovascular anastomoses of only one artery and onevein. Br J Plast Surg. October 1978;31(4):339–342.

28. Strauch B, et al. Replantation of amputated parts of thepenis, nose, ear and scalp. Clin Plat Surg. Jan 1983;10(1):115–124.

29. Morris PJ, Bradley JA, Boyal L, Earley M, Hagan P,Milling M, Rumsey N. Facial transplantation: a workingparty report from the Royal College of Surgeons of Eng-land. Transplantation. 2004;77:330–338. ComiteConsul-tatif National d’Ethique, C. C. N. E. (France) www.ccneethique.fr.

30. Wiggins O, Barker J, Cunningham M, Francois C, GrossiF, Kon M, Maldonado C, Martinez S, Perez-Abadia G,Vossen M, Banis J. On the ethics of facial transplantresearch. AJOB. 2004;4:1–12.

31. Yagi M, Ogua JH, Kawasaki M, Yagi M. Autogenoustransplantation of the canine larynx. Ann Otol RhinolLaryngol. 1966;75(3):849–864.

32. Takenouchi S, Ogura JH, Kawasaki M, Yagi M. Autoge-nous transplantation of the canine larynx. Laryngoscope.1967;77(9):1644–1667.

33. Silver CE, Rosen RG, Dardik I, Eisen H, Schwibner BH,Som ML. Transplantation of the canine larynx. Ann Surg.1970;172(1):142–150.

34. Kluyskens P, Ringoir S. Follow-up of a human larynxtransplantation. Laryngoscope. 1970;80(8):1244–1250.

35. Strome M, Stein J, Esclamado R, Hicks D, Lorenz RR,Brawn W, et al. Laryngeal transplantation and 40-monthfollow-up. N Engl J Med. 2001;344(22):1676–1679.

36. Birchall MA, Lorenz RR, Berke GS, Genden EM,Haughey BH, Gilmionow M, Srome M. LaryngealTransplantation in 2005: A review. Am J Transplant.2005;6(1): 20–26.

37. Fageeh W, Raffa H, Jabbad H, Marzouki A. Transplanta-tion of the human uterus. Int J Gynecol Obstet. 2002;76:245–251.

38. Brannstrom M, Wranning CA, Racho El-Akouri R.Transplantation of the uterus. Mol Cell Endocrinol.2003;202(2):177–184.

39. Bar Oz B, Hackman R, Einarson T, et al. Pregnancy out-come after cyclosporin therapy during pregnancy: ameta-analysis. Transplantation. 2001;71(8):1051–1055.

40. Schmid BP. Monitoring of organ formation in ratembryos after in vitro exposure to azathioprine, mercap-topurine, methotrexate or cyclosporin A. Toxicology.1984;31(1):9–21.

41. Scott JR, Pitkin RM, Yannone ME. Transplantation of theprimate uterus. Surg Gynecol Obstet. 1971;133:414–418.

42. Callaghan C, Ali A, Pettigrew G. Transplant surgery. BrMed J. 2004;329:23–25.

43. Stevenson AFG. Tissue engineering: in vitro embryonalnidation in a murine endometrial construct. Indian J ExpBiol. 2003;41(6):563–569.

44. Millesi H, et al. The interfascular nerve grafting of themedian and ulnar nerves. J Bone Joint Surgery.1972;54:727.

45. Taylor GI, Ham J. The Free vascularised nerve graft. PlastReconstr Surg. 1976;57:413.

46. Bain JR. Peripheral nerve and neuromuscular allotrans-plantation: current status. Microsurgery. 2000;20:384–388.

47. Mackinnon SE, Doolabh VB, Novak CB, Trulock EP.Clinical outcome following nerve allograft transplanta-tion. Plast Reconstr Surg. 2001;107:1419–1429.

48. Russell RC, O’Brien BM, Morrison WA, Pamamull G,MacLeod A. The late functional results of upper limbrevascularization and reimplantation. J Hand Surg [Am].1984;9:623–633.

49. Owen ER, Dubernard JM, Lanzetta M, Kapila H, MartinX, Dawahra M, Hakim NS. Peripheral nerve regenerationin human hand transplantation. Transplant Proc.2001;33:1720–1721.

50. Dubernard JM, Owen ER, Lanzetta M, Hakim N. What ishappening with hand transplants? Lancet. 2001;357:1711–1712.

51. Midha R, Mackinnon SW, Becker LE. The fate ofSchwann cells in peripheral nerve allografts. J Neuro-pathol Exp Neurol. 1994;53:316–322.

52. Morris PJ, Bradley JA, Boyal L, Earley M, Hagan P,Milling M, Rumsey N. Facial transplantation: a workingparty report from the Royal College of Surgeons of Eng-land. Transplantation. 2004;77:330–338. Comite Consul-tatif National d’Ethique, C. C. N. E. (France) www.ccneethique.fr.

53. Hu, W., Lu, J., Zhang, L., et.al. ‘‘ A preliminary report ofpenile transplantation’’. European Urology. 50:851–853.2006.

54. Holmgren G. Some experiences in the surgery of oto-sclerosis. Acta Otolaryng. 1923;5:460–466.


Index

Note: The letter ‘t’ following the locators refer to tables and letter ‘f ’ following the locators refer to figure

AAbel, J., 40, 49ABS, see Artificial bowel sphincter

(ABS)ACLF, see Acute-on-chronic liver

failure (ACLF)Acute liver failure (ALF), 30,

57–62, 67–71indications of, 69pathophysiological basis

of, 69Acute lung injury

definition, 5–6epidemiology, 6management of, 6pathogenesis of, 6

Acute-on-chronic liver failure(ACLF), 57–59, 67, 69,70–71

Acute renal failure (ARF), 9–13,16, 138, 140, 143

epidemiology, 9–10optimization after

implantation, 12–13pathophysiology, 11–12risk factors, 10–11

Albisser, A.M., 84ALF, see Acute liver failure (ALF)Alison, M.R., 146Allograft rejection, 122AMC-bioartificial liver, 61fAndersson, L.C., 93Andrus, F.C., 40Anorectal transplantation, 120

assessing feasibility of, 115–118description, 119indications, 121–123pre-operative preparation, 119

recorded variables posttransplantation, 120

technological advances,118–119

ARF, see Acute renal failure (ARF)Arteriovenous fistula (AVF), 49,

50fArtificial bowel sphincter (ABS)

AMS 800, 123–124complications, 129–130contraindications, 127device activation/deactivation,

124–125functional outcome of, 125–127implantation of, 128–129indications, 127types, 127–128

Artificial circulatory supportarterial line filters, 30–31CABG, 22–23cannulation for

cardiopulmonary bypass,24–26

conduits, 23CPB safety features, 34–35cross-clamp fibrillation, 31direct gas interface

oxygenators, 28heat exchangers, 30history, 21–22long-term circulatory support,

36–37membrane oxygenators, 29–30monitoring during CPB

activated clotting time, 31–32arterial and venous

saturations, 34arterial pressure, 32–33

ECG, 34gas flows in flowmeter, 32‘Tycos’ pressure, 34urine output, 33–34venous pressure, 33venous reservoir level, 31

non-physiological aspects ofcardiopulmonary bypass,complications of

bleeding, 35ECG complications, 36kidney dysfunction, 36low cardiac output, 36neurological events, 35pulmonary complications, 36

open heart surgery, 22protecting myocardium, 31pump to substitute for heart

centrifugal pump, 28roller pump, 27–28

replicating lungs, 28of short-term CPB circuit, 26–27valve operations, 24

Artificial extracorporeal systems,63–64

albumin, importance of, 63–64devices, design of, 64–67

MARS, 64prometheus, 65–67SPAD, 64–65

MARS clinical trial, 67–69prometheus, 70safety profile of, 70SPAD clinical trial, 69–70

Artificial kidneyaccess to blood stream, 48–49anticoagulation blood

component, 48

179

Artificial kidney (cont.)basic components of, 42tblood component, 48–49capillary fiber membrane, 43fcriteria for, 51tdialyzate, 49–50

composition, 49–50Kolff Brigham dialyzer, 45,

46fwater, 49–51

future prospect of, 51–52introduction, 39membrane support structure of,

43–48semipermeable membrane,

39–42used on humans, 44f

Artificial pancreas (AP) Project,90

Artificial urinary sphincter (AUS),123

Auditory brain steam response(ABR), 135

AVF, see Arteriovenous fistula(AVF)

BBanerjee, M., 144Bellmann, R., 69Bellomo, R., 13Bilger, R.C., 133Bournemouth’s Education

Resources for Training inInsulin and Eating(BERTIE), 81

CCABG, see Coronary artery bypass

grafting (CABG)Cardiopulmonary bypass (CPB)

circuit, 22, 25–27, 30, 31complications of non-

physiological aspects,35–36

monitoring during, 31–34safety features, 34–35

Cardiovcar failurecardiogenic shock, 2–3classification of shock, 1distributive shock, 3–4hypovolemic shock, 1–2obstructive shock, 4

Central nervous system diseases,150

Chamuleau, R.A., 60, 61Chinzei, R., 145Christensen, J.L., 127Chronic renal failure (CRF), 138Cochlear implant

common indications, 134complications of, 136ethical issues of, 135past/present/future of, 133–134services of, 134suitability of, 135surgery, 135–136

Composite tissue allograft (CTA),see Composite tissuetransplantation

Composite tissue transplantationface allograft of, 170–172first successful hand transplant

of CTA, 166–169immunology of, 169–170laryngeal transplantation,

172–173penis transplantation, 175peripheral nerve

transplantation, 173–174tongue transplantation, 174–175uterine, 173

Computed tomography (CT), 135Constipation/fecal impaction, 130Continuous glucose monitoring

system (CGMS), 85–86, 86fContinuous subcutaneous insulin

infusion (CSII), 79Continuous venovenous

hemodiafiltration(CVVHDF), 15

Conventional insulin treatment,77–79

closed loop insulin pumps,84–85

glucose sensors, 85–87and JDRF artificial pancreas

(AP) project, 90current state of play, 87

control algorithms, 89–90pumps, 87sensors, 87–89

need for closed loop system, 83open loop insulin pumps

history of, 79–80intraportal insulin delivery,

83pros and cons of, 80–83transplants, 83–84

Coronary artery bypass grafting(CABG), 22–23

Cosgrove, D., 10CPB, see Cardiopulmonary bypass

(CPB)Cultured autologous skin

substitutes, 102Cultured epidermal autografts

(CEAs), 102Cultured skin substitute (CASS),

102Cuprophan, 41–42

DDawn phenomenon, 79Dekel, B., 139De Novo Kidney Creation, 142–143Devesa, J.M., 125Devile, M., 1Diabetes Control and

Complications Trial(DCCT), 83

Dimarakis, I., 137Direkze, N.C., 140Doria, C., 69Dosage adjustment for normal

eating (DAFNE), 81Dynamic neo-sphincter

abdomino-perineal excision ofrectum, 110–111

complications of, 111–112indications of, 110types of, 110

EElmiyeh, B., 137Embryonic stem cells, 137–138Ende, N., 144Endogenous Albumin-Bound

Toxins, 58tEndothelium, 139–140Epidermic grafts, 94Epidermolysis bullosa (EB), 101Experimental organ

transplantationanorectal transplantation

assessing feasibility of,115–118

description, 119indications, 121–123pre-operative preparation,

119recorded variables post

transplantation, 120

180 Index

technological advances of,118–119

artificial bowel sphincter (ABS)AMS 800, 123–124complications, 129–130contraindications, 127device activation/

deactivation, 124–125functional outcome of,

125–127implantation of, 128–129indications, 127types, 127–128

evolution for quality of life,114–115

Extracorporeal liver assist device(ELAD), 59, 60f, 61–62

Extracorporeal membraneoxygenation (ECMO), 8,21, 36

FFace allograft, 170–172Fayad, G., 137Felldin, M., 68Flores, C., 137Frankel, A., 39Fujikawa, T., 144Full-thickness skin grafts (FTSGs),

101–102

GGerlach, J.C., 59Glomerular filtration rate (GFR),

39Glomeruli, 141–142Gomez, C.M.H., 1Gordon, M.Y., 137Grompe, M., 145Guo, J.K., 141

HHabib, N.A., 137Hand allografts, 169Haemodialysis access system,

52fHakim, N.S., 165Hepatoblastoma-based ELAD,

61–62Hess, D., 144Hoerstrup, S.P., 149Human leucocyte antigens (HLA),

123Humes, H.D., 143

IInterstitium, 140Ischemia, 112–114, 139, 141,

153–154JJuvenile Diabetes Research

Foundation (JDRF), 90

KKang, E.M., 144Kidney, functions of, 40tKirwan, C., 39Kolff Brigham dialyzer, 45, 46fKorbling, M., 146Kratz, G., 93

LLaryngeal transplantation,

172–173Lechner, A., 144Levicar, N., 137Lin, F., 141, 148Liver substitution

artificial extracorporealsystems in

albumin, importance of,63–64

devices, design of, 64–67MARS, 64MARS clinical trial, 67–69prometheus, 65–67, 66f, 70safety profile, 70SPAD, 64–65SPAD clinical trial, 69–70

bioartificial systems, 59–63clinical trials of, 60–63design of devices, 59–60safety concerns, 63

current liver support systems,assessment of, 58–59

functions required of liversupport system in, 58

Liver support systemassessment of, 58–59functions required of, 58

Lung protective ventilationcorticosteroids, 7extracorporeal support, 8fluid strategy, 7high-frequency oscillatory

ventilation, 8inhaled pulmonary

vasodilators, 7–8NIV, 9

positive end expiratorypressure, 7

prone ventilation, 8ventilation, disadvantages of,

8–9weaning, 9

MMa, N., 153Macia, M., 69Magnetic resonance image (MRI),

98, 135Major histocompatibility complex

(MHC), 118, 122–123Major histocompatibility antigens

(MHA), 170Malignant disease, 127Mallet, V.O., 146MARS circuit, 65fMesenchymal stem cells (MSC),

138Methicillin-resistant

Staphylococcus aureus(MRSA), 99, 109

MHC, see MajorHistocompatibilityComplex (MHC)

Model Predictive Control (MPC),89

Modern skin engineeringconcept of, 100future of, 102–103historical overview of, 93–94introduction to, 93skin substitutes in, 100

characteristics and clinicaluse, 100

types, 101–102structure of skin in, 93wound healing in, 95

dermal, 96epidermal, 95–96

wounds, clinical management ofdebridement, 98–99dressings, 99post-debridement treatment,

99pre-debridement assessment,

97–98Molecular adsorbents

recirculating system(MARS), 64

Mono-component insulins, 78Mulholland, J., 21

Index 181

Mullhaupt, B., 69Multiorgan failure after artificial

organ implantationacute lung injury

definition, 5–6epidemiology, 6management, 6pathogenesis of, 6

acute renal failureepidemiology, 9–10optimization after

implantation, 12–13pathophysiology, 11–12risk factors, 10–11

cardiovcar failurecardiogenic shock, 2–3classification of shock, 1distributive shock, 3–4hypovolemic shock, 1–2obstructive shock, 4

current best practice guidelines,17

lung protective ventilation, 6–7corticosteroids, 7extracorporeal support, 8fluid strategy, 7high-frequency oscillatory

ventilation, 8inhaled pulmonary

vasodilators, 7–8NIV, 9positive end expiratory

pressure, 7prone ventilation, 8ventilation, disadvantages of,

8–9weaning, 9

optimization afterimplantation, 12–13

respiratory failurecauses of, 5definition of different, 4–5management of, 5

RRT as, 13–16Multiple daily injections (MDI), 80

NNephron, 39Nerve growth factor (NGF), 118Nettelblad, H.C., 93Neuropetide-3 (NT-3), 118Ng, I.O., 146NIV, see Non-invasive ventilation

(NIV)

Non-invasive ventilation (NIV), 9Nyberg, S.L., 63

OObichere, Austin, 107Oh, S.H., 145Olanow, C.W., 153Organ replacement, 137

adult cell, 138artificial kidney, 143and cardiac-related

bioengineering, 149–150central nervous system diseases,

150de novo kidney creation,

142–143diabetes, 143–144embryonic, 137–138endothelium, 139–140glomeruli, 141–142interstitium, 140for kidney diseases, 138–139Parkinson’s disease, 150–151and pulmonary regeneration,

150stem cell therapy for diabetes,

144–145clinical studies, 145liver diseases, 145

stem cell therapy for ischemicheart diseases, 147–149

clinical studies, 148mechanisms of action, 148–149

stem cell therapy for liverdiseases, 145–147

clinical studies, 146–147myocardial infarction/heart

disease, 147stem cell therapy in stroke,

153–154stem cell therapy Parkinson’s

diseaseclinical studies, 152–153stroke, 153

tubular epithelium, 140–141Owen, E.R., 165

PParkinson’s Disease, 150–151Passive neo-sphincter, 108

complications of, 109indications, 108–109types of, 109

Patel, P., 1

Penis transplantation, 175Peripheral nerve transplantation,

173–174Petersen, B.E., 145Physiological neo-sphincter,

112–113complications of, 113–114indications of, 113types of, 113

Porcine Hepatocyte-Based BAL,60–61

Press, M., 77Prometheus, 65–67, 70Proportional-integral-derivative

(PID), 89Prosthetic artificial sphincter

(PAS), 127–128Prosthetic bowel sphincter

complications of, 129contraindications, 127functional outcome, 123–127implantation, 128indications, 127types, 127

RRecombinant erythropoietin

(r-epo), 39Renal replacement therapy (RRT)

adverse effects of, 15anticoagulation, 15choice of anticoagulation, 16CVVHDF, 15extracorporeal inflammatory

mediator removal, 16indications for, 14optimal dose of hemofiltration,

16optimal treatment modalities

for, 16physiological functions of

kidney and clinicalmeasurement, 13

principle of hemofiltration, 15renal function, assessment of,

13–14vascular access, 15

Renal tubule cell assist device(RAD), 52, 143

Ross Basket, 24

SSakaida, I., 145Sen, S., 57

182 Index

Shape memory alloy (SMA),127–128, 140

Sin, D.D., 5Single-pass albumin dialysis

(SPAD), 64–65, 66fSkin

structure of, 93substitute in, 100

characteristics and clinicaluse, 100

future of, 102–103types, 101–102

Soria, B., 144Sphincters, artificial

dynamic neo-sphincter, 109complications of, 111–112dynamic graciloplasty,

functional outcome of, 111following abdomino-perineal

excision of rectum,110–111

indications of, 110types of, 110

origin of, 107passive neo-sphincter, 108

complications, 109indications, 108–109types of, 109

physiological neo-sphincter,112–113

complications, 113–114indications, 113types of, 113

prosthetic bowel sphincter, 123complications of, 129contraindications, 127functional outcome, 123–127implantation, 128indications, 127types, 127

surgically fashioned bowel, 108Split-thickness skin grafts

(STSGs), 101Stem cell therapy, 138–139

for diabetes, 144–145for ischemic heart diseases,

147–148for liver diseases, 145–146in stroke, 153–154

Suliman, I., 107

TTang, D.Q., 144Theise, N.D., 146Thompson, L., 152Tongue transplantation, 174–175

Transforming growth factor(TGF), 58

Transplantationlaryngeal, 172–173pennis, 175peripheral nerve, 173–174tongue, 174–175

Tubular epithelium, 140–141‘Tycos’ pressure, 34

UUltrafiltration, 15, 39, 42, 45, 50

VVina, R., 145

WWei, H.J., 149Williams, Roger, 57Wong, W.D., 127

YYamada, T., 145Yamamoto, M., 138

ZZheng, F., 141

Index 183

Artificial Organs

Documents

Transcript of Artificial Organs