Effect of Hypercarbia and Isoflurane on Brain Cell Death and Neurocognitive Dysfunction in 7-day-old Rats

• DOI: 10.1097/aln.0b013e31819c7140
• Publication Date: 2009-05-21T15:58:41Z
• Authors: Greg Stratmann; Laura D. V. May; Jeffrey W. Sall; Rehan S. Alvi; Joseph S. Bell; Brandi K. Ormerod; Vinuta Rau; Joan F. Hilton; Ran Dai; Michael T. Lee; Kavel H. Visrodia; Ban Ku; Emanuel J. Zusmer; Jeremy Guggenheim; Atoosa Firouzian
• Keywords: Neurocognitive; Hypercarbia; Memory Impairment; Minimum alveolar concentration

MILLIONS of neonates undergo anesthesia for surgical or diagnostic procedures each year.1,2Anesthesia kills neonatal brain cells several animal species, including primates,3–6and causes long-term neurocognitive dysfunction.3,5,6Consequently, anesthesia-induced apoptotic cell death is a growing concern within the community, public, and regulatory authorities alike.7However, causal link between dysfunction lacking. If such relationship existed, degree should predict severity dysfunction. We therefore studied if 0, 1, 2, 4 h isoflurane at 1 minimum alveolar concentration (MAC) postnatal day (P)7 rats previously observed that arterial carbon dioxide tension spontaneously ventilating isoflurane-treated P7 increases shortly after induction general anesthesia; therefore, we also tested hypercarbia alone affects in rat it alters function.Pavlovian fear conditioning well-established method studying learning memory.8,9In typical fear-conditioning experiment, rodent taught to associate neutral cue (e.g. , tone) with an aversive event shock). After rodents have been trained one more pairings tone shock, they learn predicts shock. Rodents will exhibit their innate response (i.e. freezing, immobility) even when presented without Memory association shock assessed by quantifying freezing behavior tone. In addition surrounding environment made up visual, auditory, olfactory, tactile cues (the context) When are given multiple not only freeze tone, context which training took place.The basolateral amygdala, located medial temporal lobe, essential acquisition conditional fear. The amygdala site convergence information about its predictive occurs.10,11The hippocampus assembles representation communicates amygdala. roles these two structures well characterized. Lesions block context, whereas lesions leave intact.12,13The differential contribution allows us perform different tests memory training: (1) measurement second distinct from providing assessment effect on amygdala-dependent, hippocampal-independent task; (2) absence anesthetics hippocampal-dependent task.In spatial reference task Morris water maze, find submerged platform escape maze. Due local visual olfactory would indicate position, use distal cues. This requires formation allocentric depends neocortex.14,15Both (learning) retention can be working executive function, uses placed location day. integration hippocampus, cortex thalamus, among others.16–19Here show three interventions cause significant regions, namely 2 as dioxide. Of these, caused deficit.All experiments were conducted approval Institutional Animal Care Use Committee University California, San Francisco. Rats (n = 17) litters containing male 162) female 2) Sprague-Dawley (from Simonsen Laboratories, Gilroy, CA, Charles River Davis, CA) cross-fostered before starting experiment. pups 7 days old, sham anesthesia. treatment, groups control (4 anesthesia, dioxide) distributed equally litters.Anesthesia was described elsewhere.20Briefly, administered least 10 anesthetized simultaneously. MAC determined tail clamping every 15 min. Physiologic variables assessed, temperature controlled 36.5°C.Control glove box exposure anesthetic agent but otherwise identical conditions animals group. Thirty kept dioxide–enriched atmosphere rule out effects fraction inspired varied 30% mimic groups. assumption long breathing did appear depressed, less than mmHg higher first tension.21Cardiac puncture blood draws performed under (0.05 mm 1% lidocaine) injected subcutaneously.Tissue preparation FluoroJade (FJ) staining accompanying article this issue Anesthesiology.20FJ stains scanned photographed 4× tile scan 20× areas appreciable death. mapped according Paxinos' atlas brain.22Data fold increase control. Density FJ-positive (FJ+) animals, representing physiologic death, provided reference. All FJ+ counted slices structure interest, results averaged. individually thalamic nuclei combined analysis paracentral, centromedial, paraventricular, mediodorsal (central lateral), intermediodorsal, rhomboid, reuniens (including ventral reuniens), lateral posterior (laterocaudal, laterorostral, mediocaudal, mediorostral). Cortical layers 2-6 hippocampal subregions analyzed cornu ammonis (CA)-1, CA-2, CA-3, pyramidal layer, dentate gyrus.Eight weeks underwent conditioning. Four rats, counterbalanced group assignment, time. chambers (length, 32 cm; width, 25 height, cm) constructed clear acrylic. grid floor used deliver composed 19 stainless steel bars, diameter spaced 16 center. These floors connected delivery system (Med Associates, St. Albans, VT). wiped pine-scented cleaner (5% Pine Scented Disinfectant; Midland, Inc., Sweetwater, TN) session. room place illuminated overhead fluorescent bulbs, ventilation fan background noise (65 db). appearance, odor, texture comprised context.After 3-min baseline exploratory period chambers, received (2000 Hz, 90 db)–shock (1 mA, s) separated Freezing, all movement except necessary respiration, defensive reliable measure learned fear.12Each animal's scored 8 s during observation period, percentage calculated using formula 100 × f/n, where f number events per n total observations rat.The next day, For test, once again chamber min (in transported separate plastic pots (height, 14 diameter, 15.5 room. test triangular shape acrylic 28 sidewalls 22 45-degree angle. equipped speaker down acetic acid (1%; Fisher Scientific, Louis, MO) appeared dark being lit single 30-Watt red bulb. A kind white db) noise. 30-s tones 60 s. removed additional 30 order half treatment vice versa . Freezing observers blinded assignment training, both tests.A (diameter, 10.3 circular pool 180 depth, 50 filled warm (24°C) opaque water. Two sessions (7 apart). At beginning session, onto then trials (intertrial interval, s), released six pseudorandomly assigned release points (facing wall; see fig. 1). On subsequent trials. Animals locate hidden platform. allotted time, guided experimenter. either case, To minimize any bias associated location, equal numbers four locations duration (even Training until able averaged over session five sessions. Time reach (latency), path length, swimming speed, time-integrated distance EthoVision video tracking (Noldus Instruments, Wageningen, Holland) set analyze ten samples second.To assess probe trial tank 3 last (when criterion achieved). During 60-s trial, proportion time spent target quadrant versus other quadrants separately intervals, crossings recorded analyzed.After appearance changed completely relocating maze new remote examine presence gross sensorimotor deficits, visible protruding inch above surface marking edge bright tape could seen clearly system. (three session) relocated session.23The dependent trials.For task, rearranged repositioning below positions (odd 1) reoccurring average fourth design sought promote strategies rather taxon strategies.23,24One free swim, allowed explore found, remain s, Performance swim scored. chosen facing wall tank. drop incorporate short, medium, swim. apply center pool, case same locations. (after sessions). speed Instruments) second.When achieved (day 12), indicating understood rules delay inserted difficulty. 13, 1-h introduced; 14, extended h. (latency) platform, (spatial training). 12 (1-min delay), 13 (1-h (4-h delay) memory.Data reported median interquartile range values experiments. answer question dose-dependently, differences explored Kruskall-Wallis degrees freedom revealed Dunn's correction comparisons determine threshold end, post hoc compared evaluated Wilcoxon signed rank test. differs 4-h group, sum used. strength correlation bregma density, Spearmann coefficient Cell data Prism MacIntosh 4.0 (GraphPad Software Diego, CA).Fear scores expressed medians (±interquartile range) collapsed portion (to account habituation stimulus). Group (baseline postshock pretone tone-freezing, freezing) Kruskall Wallis comparisons. Fear Inc.).Water means (± SE). tasks, measures latency, highly positively correlated (r > 0.74; P < 0.05 groups); distances clarity economy. (latency, criterion, scores) two-way variance (ANOVA) Newman-Keuls tests. variable exhibited heterogeneity variance, Mann–Whitney U Data Statistica (V.6; StatSoft, Tulsa, OK) Stata version 10.0 (StatCorp, College Station, TX). Alpha levels 0.05.For assessment, minimal size required detect difference 40% 80% power significance level behavioral experiments, sample greater include margin error. SAS 9 (SAS Institute, Cary, NC) Proc POWER calculations.The type detail no hypoxia hypoglycemia occurred Rapid moderate hypercapnia (fig. 2A) acidosis 2B) developed immediately mortality 25%, most dying hour 2-h discontinuation anesthetic. tensions pH dioxide-control matched those isoflurane–treated 2).Of detectable severely affected 3). isoflurane, density 45 360 times laterodorsal-ventrolateral nuclei, respectively. thalamus region. mediodorsal, laterodorsal, posterior, reuniens, central amount substantial (between 20-fold 192-fold increase) significantly although unplanned always statistically significant. submedius, paraventral (table Despite anteroventral anteromedial included statistical analyses due small usable areas.The (P 0.01 0.001 control, test; 4). Neither isoflurance nor had similar distribution death.In (CA1-3 gyrus), 5A). (1,802 cells/μl) differ (2,769 cells/μl). One 2.5-fold controls, 5-fold increase. suggests dioxide, mild mostly dose-dependent manner, h.In 2–6 cortex, neither 5B). maximum cortical approximately 7.5-fold isoflurane.In mammillothalamic tract, there dioxide–treated However, treatments increased 5C).Additional regions implicated deficit here, fornix, entorhinal, perirhinal, parahippocampal cortices, mamillary bodies,25had little groups.Baseline onset tone-shock presentation generally low regular controls 0.01, 6A). Administration elicited robust increasing short-term 6B).On low; however, unlike previous found 6C). Tone showed high interindividual variability 6D) (not shown).Context variable, 6E). There i.e. shown).ANOVA (F(4,52)= 1.79; 0.001; 7A) latency that, acquired position rate, slightly worse (session 2; 0.05) better (sessions 3; 0.01) few Interestingly, achieve (15 trials) (F(4,81)= 2.89; 0.027; 7B). exposed 0.05, Newman–Keuls test). 2.27; 0.0001; 7C). Surprisingly, neonatally swam slower 5 0.001), despite exhibiting performance earlier Given groups, speeds likely reflect reduced anxiety impaired ability. prolonged h) produces cognitive impairment standard task. does impair appears improve (F(4,8)= 1.40; 0.05; shown), suggesting ability motivation solve adversely Three dwell searching (F(4,12)= 0.03; 7D) (F(4,75)= 0.54; 0.05). affect learned.Latency performed, quite 1–12; 8A). Although outperform trials, ANOVA demonstrated latencies (F(4,44)= 1.14; An latency) average, (F(2,80)= 2.16; 5.70 ± 0.24; 5.11 0.08; 6.0 0.25; 5.73 0.28; 5.87 0.26 sessions), supports finding. disappeared 1.26; slowly further lower extensive training. difficulty asking remember novel 13) 14), emerged relocate 2.00; 8B). Post delay, tended 0.10); trial. Swim 8C) 5.07; 0.00002), outperforming them 0.001).There main findings study. First, Beyond threshold, dose-dependent, dose defined Substantial especially nuclei. extends death3–6,26–29to volatile clinically relevant MAC.Second, Furthermore, rostrocaudal part isoflurane-induced may hypercarbia. whole grow directionally caudal rostral inside outside.30It follows certain postmitotic age. example development, older neurons deeper younger ones superficially located. vulnerability immature toxicity function age cells, example, naturally eliminated failure make appropriate connections cells,31,32one expect clustering contain vulnerable stage.Therefore, gradient isofluane insults type. study provide direct evidence identity killed isoflurane. It possible pattern hypercarbia.Third, causing dioxide), deficit, evident tasks tasks. chose battery basis relatively knowledge anatomic involved tasks12,16–19,25,33,34and requirements hands aversiveness promoting tasks.Surprisingly, brain, improved memory, offer explanation counterintuitive As mamillothalamic tract deficit. While confirming rodents,3,5,6the inconsistencies outcome suggest might sufficient loss some recovery brain. issue, nature just opposite delayed progressive course third rat's life span.20Of contributions anesthesia.12,16–19,25,33Although, argued above, fully explain minor and/or contribute here. theoretically reached produce Another interpretation responsible findings. Besides possibilities, alternative mechanisms mediate dysfunction; attractive possibilities regard potential neurogenesis synaptic morphology function. represent forms plasticity, cellular plasticity. support hypotheses contextual has inhibition progenitor proliferation mice35and 6 specifically ablating adult-born nestin-expressing precursors,36suggesting (like demanding task) mice alcohol, mechanistically resembles agents,37decreases gyrus adulthood,38,39which consistent our decrease rats.20Anesthetic agents shown dendritic morphology,40–42and emerging extend short-term43,44and morphology.44These plausible developing far exceeds pharmacologic action anesthetic.Limitations plentiful. FJ staining, validated against methods detection extremely sensitive,45,46is specific necrotic death.45,46The model plus surgery. Tail tissue damage, inflammation, scar formation, confounder. weakness mechanistic standpoint translational because surgery children rarely. addition, endpoint, MAC, study, enhances translatability, narrow confines limitations imposed species humans. Hence, efforts duplicate scenario, important limitation undefined applicability human anesthesia.Concern translate based following line reasoning. Neurons receive survival signals synaptically connected.31,32During assumed correspond synaptogenesis,3,47–50anesthesia interrupts crosstalk neurons, withdrawal neuronal death.3,5,6,26,27,29,45,47–55Synaptogenesis occurs years birth.56Anesthesia-related synaptogenesis rodents3,5,6; mechanism operational humans birth.The current work challenges necessarily Other assumptions reasoning questioned, conjectures peak dead mature neurons.It held developmental toxicity, around rodents,48,51corresponds synaptogenesis.27,29,48,51–54The frequently cited review weight gain development across species.56Although speaks timing growth spurt, relate se occur 11–16, seems P32.57,58The probably past roughly P1053to P1447; connection questionable.A variety various monkeys1,3,5–7,26,27,29,45,47–54but neurons? short gestational (21.5 days) 16–17 age, majority younger.59Even smaller Na+–currents,61a length,60an γ-aminobutyric reuptake system,61immature chloride extrusion, consequently depolarized acid, hyperpolarized like (for Ben-Ari et al. 62). Therefore, possibility considered pathoanatomic correlate stage neuron. must morphologic criteria markers none meaningful resolution functional correlate.60Knowledge subject facilitate focused search death.As result discrepancies discussed reflecting lack convincing decline rodents, premature speculate toxicity.Even kill unclear follow. outcomes tightly linked thought intended downplay problem. finding warrants investigative problem meantime, studies vitro vivo models inform any, molecular it.In conclusion, play role unlikely rats. relevance clinical require study.The authors thank Dr. Edmond I. Eger II, M.D., Professor, Department Anesthesia, Francisco, John W. Severinghaus, Professor Emeritus, advice administration Laszlo Vutskits, Ph.D., Associate Anesthesiology, Pharmacology Intensive Care, Hospital Geneva, Switzerland, valuable comments manuscript.