Fetal brain toxicity & anesthetic agents: yet another ... NOV 1220 Segal.pdfScott Segal Professor of...
Transcript of Fetal brain toxicity & anesthetic agents: yet another ... NOV 1220 Segal.pdfScott Segal Professor of...
Scott Segal Professor of Anesthesiology
Chair, Department of Anesthesiology Tufts University School of Medicine
Fetal brain toxicity & anesthetic agents: yet another reason to avoid
general anesthesia?
Behavioral teratogenicity
� Enduring behavioral changes without gross structural defects
� Common anesthetics can cause long-term behavioral abnormalities
NMDA antagonism and brain injury
Science 1999; 283:70-74
Ikonomidou et al. 1999
� Seven-day-old rats ¡ Intra-peritoneal vehicle or dizocilpine (MK801),
NMDA antagonist ¡ Brains stained
÷ TUNEL (fragmented DNA, sign of apoptosis) ÷ Silver (cell death)
TUNEL staining
Ikonomidou, Science 1999; 283:70-74
Ikonomidou et al. 1999
Ikonomidou et al. 1999
� Drug given lasted 4-6 hr � Same effect seen with PCP, ketamine � Effects sensitive to gestational age or post-
natal age ¡ Rats synaptogenesis peaks at 7 days age ¡ Corresponds to third trimester and postnatal
period in humans
� Rats exposed to “triple cocktail”: N2O, midazolam, isoflurane
� Day 7 (peak synaptogenesis), 6 hr anesthetic � Sacrificed for histology or studied later for
learning ability
J Neuroscience 2003
Jevtovic-Todorovic 2003
Control Silver Caspase 3
Morris water maze
Jevtovic-Todorovic 2003
Radial arm maze
Jevtovic-Todorovic 2003
Other anesthetics
� NMDA antagonists ¡ Ketamine ¡ N2O
� GABAA agonists ¡ Midazolam, diazepam ¡ Pentobarbital, thiopental ¡ Propofol ¡ Isoflurane, sevoflurane ¡ Ethanol
Mechanisms
� NMDA ¡ Glutamate signaling important for synapto-
genesis, NMDA antagonists block
� GABA ¡ Increases chloride influx, cell protection in adult ¡ Causes chloride efflux, calcium influx in
developing brain ¡ Unclear how apoptosis triggered
Criticism (1)
� Large doses, prolonged exposures � Post-natal neurodevelopment in rats � No surgical stimulation accompanies the
anesthetics � Hemodynamics, respiratory status, glucose
not been uniformly observed or controlled ¡ GA causes hypoglycemia in rats ¡ Hypoglycemia impairs immature brain
development
Loepke AW, McGowan FX, Soriano SG. Anesth Analg 2008 (June)
Criticism (2)
� Neuroplasticity implies regeneration of neurons lost to apoptosis is likely
� Some effects are only observed when exposure occurs on certain developmental days and not others
� Species specificity ¡ Mice, rabbits, piglets, sheep intact after isoflurane
4-6 hr ¡ Most behavioral consequences have only been
observed in rats
Loepke AW, McGowan FX, Soriano SG. Anesth Analg 2008 (June)
Criticism (3)
� Human brain development slower, would require 2 wk exposure to be equivalent
� Human experience with neonatal anesthesia spans decades and does not demonstrate significant neurotoxicity
� Some anesthetics (e.g. xenon), are NMDA antagonists but cause no toxicity ¡ May even be protective
Loepke AW, McGowan FX, Soriano SG. Anesth Analg 2008 (June)
Responses to criticism
� Doses are sub-anesthetic in species studied ¡ Anesthetic dose ketamine 5 mg/kg humans, 80 mg/
kg mouse ¡ Isoflurane MAC 1.15% humans, 2.26% in mouse
� Brain slices, cell culture confirm results ¡ No blood flow, hemodynamic, glucose issues
� ISF studies in other species experiments flawed ¡ Must look within hrs for apoptosis ¡ Staining methods critical
� Non-human primates show effects in <8 hr (ethanol) or 5.5 hr (ketamine)
Jevtovic-Todorovic V., Olney, JW. Anesth Analg 2008 (June)
What about human record of safety?
� Ethanol clearly causes human degeneration ¡ But unrecognized for centuries ¡ Only appreciated 2° craniofacial abnormalities
� Rapid recovery from anesthesia is not reassuring ¡ Seen in rat, guinea pig, monkey models as well ¡ Little investigation of long-term cognitive effects
Jevtovic-Todorovic V., Olney, JW. Anesth Analg 2008 (June)
� Guinea pigs: longer gestation, more brain development in utero, larger animal
� Exposed to “triple cocktail”: 75% N2O, midazolam (1 mg/kg IM), isoflurane (0.55%) vs. fentanyl (“sham”) vs. control
� 4 hr anesthetic � Fetal (2 hr post anesthesia) or neonatal
(35-40 d post) brains studied for histology
Brain Pathology 2008
Rizzi 2008
Rizzi 2008
Rizzi 2008
Brain toxicity in animal models
� Not confined to apoptosis ¡ Neurogenesis, microgliosis ¡ Dendritic spine formation, synaptic function
� Not confined to post-natal exposure ¡ Guinea pigs: peak effect in 2nd trimester ¡ Rhesus monkeys: 24 hr ketamine in utero ¡ Rats 2nd trimester: behaviorally abnormal as
adults (Palanisamy et al. 2011)
� Timed pregnant rats (day 14) exposed to 1.4% isoflurane or 100% O2 for 4 h
� Offspring subjected to multiple behavioral tests at 8 weeks (young adult)
� Exposed adults abnormal in radial arm maze and elevated plus maze ¡ Impaired spatial working memory, reduced
anxiety
Anesthesiology 2011 (Mar); 114: 521–8
Brain toxicity in animal models
� Not confined to rats ¡ Mice, guinea pigs ¡ Nonhuman primates (ketamine in rhesus monkey)
Neurotoxicology and Teratology 33 (2011) 220–230
� Rhesus monkeys exposed/unexposed to ketamine 24 hr ¡ Day 5-6 postnatal ¡ Returned to mothers, weaned @ 7 months ¡ Began standardized cognitive training
Paule et al., 2011
Human data
Anesthesiology 2009 (April); 110:796 – 804
Wilder et al.
� Olmsted county, Minnesota � 593 children undergoing GA less than age 4
¡ Halothane 88%, N2O 91% ¡ No other drugs >10%
� 4764 children with no GA � Incidence of learning disabilities K-12
¡ Reading, writing, or mathematics ¡ >1.75 SD below IQ-predicted standard score
Wilder et al.
Limitations of Wilder et al.
� Surgical procedures, diseases uncontrolled ¡ Authors: may be impossible ¡ Authors: maybe treating (surgically) is better for
development than not treating � Definition of learning disability nonstandard
¡ Authors: standard in MN � Hypoxia, hyperoxia may confound results
¡ Authors: hypoxia not linked to LD (retinopathy treatment), hyperoxia likely modest (70% N2O)
Human epidemiology
� Hernia repair < 3 yrs (N=383) vs. controls (N=5050) (DiMaggio 2009) ¡ 2.3 fold greater prevalence behavioral problem or
developmental delay
� Danish cohort hernia repair in infancy (N=2689) vs. controls (Hansen 2011) ¡ No difference in test scores grade 9 ¡ More exposed children never took tests
Twin, sibling studies
� Children exposed to 1, 2, or 3 anesthetics vs. controls (DiMaggio 2011) ¡ RR developmental/behavioral disorders
increased from 1.1 to 2.9 to 4.0 ¡ No difference between matched sibling pairs
� Monozygotic twins, only one exposed age <3 yr (N=1143 pairs) (Bartels 2009) ¡ Lower test scores age 12 in exposed group ¡ No difference between matched exposed vs.
unexposed twin pairs
Obstetric anesthesia
Anesthesiology 2009 (Aug); 111:302–10
Sprung et al, 2009
� Olmsted county, Minnesota � 5320 children born 1976 –1982 and
remaining in the community at age 5 � Cesarean delivery 497 (9.3%)
¡ General anesthesia N=193 ¡ Regional anesthesia N=304
Sprung et al., 2009
� General anesthetics used (NMDA and GABA) ¡ Total 193 patients ¡ Intravenous agents
÷ Sodium thiopental 189 ÷ Ketamine 2
¡ Inhalational agents ÷ Enflurane 15 ÷ Halothane 105 ÷ Isoflurane 6 ÷ Methoxyflurane 42 ÷ Nitrous oxide 191
CS anesthetic effect on LD
HR 0.88 (0.59–1.31)
HR 0.64 (0.44–0.92) P=0.017 vs. VD
20.8%
19.4%
15.4%
CS anesthetic effect on LD
� Controlling for anesthesia in child from birth to age 4: no change in results
� Controlling for emergency CS: no change � Controlling for other covariates of LD differing
among delivery modes: no change ¡ Induction ¡ Maternal age ¡ Maternal education ¡ Birthweight, sex ¡ Labor complications
Sprung et al: conclusions and limitations
� CS under GA does not alter LD risk vs. VD � CS under regional appears safer than VD,
perhaps because stress response is blocked � Many limitations
¡ Retrospective birth cohort design ¡ Lower CS rate than today ¡ Outdated GA agents ¡ No assessment of labor analgesia
Analgesia for vaginal delivery
� Olmstead county database 1976-82 � 1495/4684 received neuraxial analgesia
Anesth Analg 2011;112:1424–31
Flick et al. 2011
b = adj for gest age, birthweight, Apgar, surgery < age 5, maternal education c = adj for “b” + details L&D (induction, duration, forceps, inhalation anes, etc.)
Neurotoxicity: the future
� Preclinical ¡ Mechanisms (Ca++ homeostasis) ¡ Critical periods ¡ Protective strategies (!-2 agonists)
� Clinical ¡ Prospective observation (PANDA), RCTs? (GAS trial) ¡ Very difficult to design (Thomas et al., A&A, 2011)
÷ Age, type of surgery ÷ Anesthetics (RA vs. GA, sedation) ÷ Outcomes (preschool, school, adolescence, adult)
Neonatal anesthetic neurotoxicity
� Animal studies convincing and worrisome � Mechanisms appear relatively clear
¡ Is it true in human brain? ¡ Can adverse effects be blocked?
� Human data on apoptosis is nonexistent � Human epidemiology rare but concerning � Prudent in 2011 to reserve GA for
emergencies