| JOURNAL HOME | CME HOME | THIS MONTH | PAST ISSUES | ETOC | COLLECTIONS |
| AUTHORS | REVIEWERS | EDITORIAL BOARD | FEEDBACK | RSS | HELP |
| ||||||||||||||
| ||||||||||||||
|
|
|
|
||||||||||||
|
||||||||||||||
From the *Clinical Science at South Bristol, Child Health, University of Bristol, St. Michael's Hospital, Bristol, UK; University of Wales Swansea, Singleton Park, Swansea, UK; and Department of Biochemistry, Children's Hospital, Bristol, UK.
Address correspondence and reprint requests to J. Dingley, Department of Anaesthetics, Morriston Hospital, Swansea SA6 6NL, UK. Address e-mail to j.dingley{at}swansea.ac.uk.
Abstract
BACKGROUND: Hypothermia (HT) reduces neuronal injury after perinatal asphyxia. The anesthetic gas xenon (XE) may enhance this effect. We investigated the sedative and respiratory effects of variable XE concentrations at 37°C normothermia (NT) or 32°C HT after a hypoxic-ischemic (HI) insult to determine the concentration at which XE was a respiratory depressant in spontaneously breathing 7-day-old rat pups.
METHODS: (I) In three control groups, the effects of fasting at NT and HT were investigated. (II) Six groups were subjected to a HI insult (left carotid ligation then 90 min breathing 8% oxygen); three then breathed Air, 50%Xe or 70%Xe for 5 h at NT (NTAir, NT50%Xe, NT70%Xe), while three breathed identical mixtures during HT (HTAir, HT50%Xe, or HT70%Xe), in addition to a control group. Blood gases, glucose, and lactate were measured. Sedation (spontaneous movement/respiratory rate) was recorded.
RESULTS: Blood chemistry data were successfully obtained from 70 pups. (I) Pups maintained normal blood gas, glucose, and lactate values after 9 h fasting at NT or HT. (II) After HI insult, in comparison with control and NTAir groups, 70%Xe at both NT and HT produced higher PCO2 and lower pH values while the HTAir and HT50%Xe groups only had lower pH values. The HT70%Xe combination produced the highest PCO2 and lowest pH values (56.8 mm Hg, 7.35, respectively) and the greatest sedative effect.
CONCLUSION: After HI insult, 70%Xe at both NT and HT induced sedation, respiratory depression, CO2 retention, and a decrease in pH relative to air and control groups. The effects were largely avoided with 50%Xe.
Hypoxic-ischemic (HI) brain injury, such as perinatal asphyxia, can result in lifelong motor and cognitive impairment.1–3 These injuries are characterized by an encephalopathy created by a primary insult, followed by a self-sustaining destructive cascade over hours or days.1 This cascade suggests that an effective post-insult therapy for babies might limit the eventual damage. However, no clinical intervention, except hypothermia (HT), has been shown to alter neurological outcome in babies.4–6 There is still room for improvement as only 1 in 6 children so treated will benefit, which leads us to consider and seek combination therapies. The noble gas xenon (XE), which has received a marketing license as an anesthetic drug in Europe, is also showing great promise as a neuroprotectant in experimental studies.7–9 It is attractive for combination therapy with HT as it is almost chemically inert and free of adverse clinical side effects or toxicity, particularly fetotoxicity.10–12 In addition, XE exhibits rapid onset/offset characteristics and has long been used in neonates for radiological studies.13 It antagonizes the N-methyl-d-aspartate receptor and possibly others, reducing receptor over-stimulation triggered cell death.14–16 There may be additional mechanisms as XE appears more protective than drugs that antagonize these receptors alone. XE has been shown to be neuroprotective both in vitro and in vivo in neonatal rats when administered both before,17 during,8 and after a HI insult.7 The published human trials of HT involved neonates at high risk of brain injury in which 90% needed mechanical ventilation.4,5,18 It may be desirable in the future to use XE-HT therapy in neonates at less extreme risk. Consequently, a significant proportion of future candidates may be breathing spontaneously.
There is some evidence that the neuroprotective effect of XE is dose-related, and so it may prove desirable to use as high a XE concentration as possible.19,20
Although it might be possible to give up to 75% XE, 50% would be a more realistic compromise since encephalopathic newborns can develop pulmonary hypertension and have sometimes aspirated meconium requiring an increased Fio2.
Inhaled anesthetics may cause respiratory depression,21 hypercapnia, and cerebral vasodilation, which might be unhelpful after HI injury. The purpose of this study was to determine the effects of HT and XE, alone or in combination, on spontaneous respiration and activity after a HI insult. We investigated this in a modification of a well-known neonatal animal model of HI.22
METHODS
Animal Preparation
All procedures were conducted under Home Office License in accordance with United Kingdom (UK) guidelines (License number 30/2554). Wistar rat pups, of either gender, from size-culled litters (to 12 pups each) were used.
Temperature Measurements
Within each group, an additional pup was designated to carry a rectal temperature probe (the "sentinel" rat). Rectal temperature closely follows brain temperature within 0.1°C in this model.22 The temperature in the chambers was adjusted to maintain the desired rectal temperature in the sentinel rat of each chamber.
Activity Measurements
A subset of pups was labeled for visual identification and 5 min epochs throughout the 5 h intervention period were filmed. The films were blindly assessed and number of forelimb movements and respiratory rate counted in 4 to 6 pups in each of three groups (normothermia [NT]Air, NT50%Xe and NT70%Xe).
Conduct of Experiment
The pups were randomized and paired by weight and sex in two sets of investigations: (I) three non-HI control groups to evaluate the metabolic effects of separation from the dam at NT and HT, followed by (II) three NT and three induced HT interventional groups receiving various XE/temperature regimes for 5 h post-HI insult and an additional group of non-ligated controls.
At the end of these interventions, the rats were killed and immediate blood gas, glucose and lactate analyses of the mixed arteriovenous blood samples were performed as described below.
Method of Producing the Insult in the HI Groups
At postnatal day (P) 7 (day of birth designated day 1), experimental pups were exposed to a HI insult in a variation of the Rice-Vannucci model.23 Immediately after this insult, each group received different interventions.
Anesthesia/Operative Procedure
The pups were anesthetized by inhaling halothane/nitrous oxide. The left common carotid artery was divided between double ligatures of silk suture (6-0) and the duration under anesthesia was 8.7 (3.89) (Mean ± sd) min.
The pups were left to recover for up to 1.5 h before being exposed to hypoxia (premixed 8.00% oxygen 92% N2) for 90 min in a temperature-controlled chamber at a target Trectal of 36.7°C.
Chamber Design
After the 90 min period of hypoxia, the pups were exposed to their allocated gas mixture/temperature regimes for 5 h in a circular chamber of 2.2 L volume. This chamber was surrounded by a water jacket, through which a coil of the tubing carrying fresh gas to the chamber also passed. The desired rectal temperature of the sentinel rat in each chamber was maintained by varying the temperature of this water jacket.7 The chamber was divided radially into 13 individual compartments by wire mesh screens so that pups could receive an interventional gas/temperature regime together as a group (12 + 1 "sentinel").
Gas Administration
The chambers were supplied with gas within a closed recirculating system propelled by a roller pump at 600 mL/min to conserve Xe. The gas passed through a soda-lime container to remove any CO2 and then a humidifier en-route to the chamber. This closed loop design meant that a high flow could be maintained through the chamber while still maintaining a low fresh Xe consumption. The composition of the gas entering each chamber was monitored at 1-min intervals. The oxygen fraction was measured using a standard anesthesia gas analyzer (Capnomac, Datex-Ohmeda, Finland), which was also used to demonstrate the absence of any CO2 (gain/drift <0.2 kPa for CO2 measurement over the range 0–8 kPa). The XE concentration was measured using an ultrasonic analyzer (Minison, Thomas Swan Scientific Equipment Ltd. Cambridge, UK) used previously.7,24,25 Sample gases were dried before the analyzers by passage through a 10 mL container of silica gel beads (Geejay Chemicals Ltd. Bedfordshire, UK).
At the end of these 5 h interventions, immediate mixed arteriovenous blood gas measurements were performed as described below. Lactate and glucose were also measured in selected groups.
Blood Gas Sampling and i-STAT Analysis
A validated portable cartridge-based analyzer was used (i-STAT Corporation, Abbott Diagnostics Division) in which a drop of blood is applied to a test cartridge. The i-STAT accuracy is typically ±5% for CO2 and some variables are measured directly while others are derived by calculation (Appendix).
The mixed blood sample was obtained by decapitation and use of a capillary sampling tube. A pure arterial blood sample (for example by left ventricular puncture) was not practical in these tiny pups for two reasons. First, their small size meant that the volume of blood obtainable by any means rarely exceeded 0.2 mL (total blood volume at this age being less than 1 mL). The arterial portion would again be smaller. Second, to obtain an arterial blood sample would have required sternotomy, cardiac puncture and therefore anesthesia under UK ethical guidelines. XE is very rapidly eliminated from the body via the lungs so the delay between pup removal from the chamber and securing an arterial blood gas sample by sternotomy/left ventricular puncture would have allowed the pup to both wake up and self-correct any XE-induced respiratory depression. The arteriovenous sampling method allowed the blood sample to be obtained immediately on removal of a rat from the chamber (into the air) before any change in gas values had time to develop.
Since it was only possible to recover enough blood from each rat pup for one gas analysis, there was no possibility of a repeat attempt if cartridge analysis failed.
The blood gas results were not temperature corrected. The rationale for this is that the "normal" acid-base ranges change as body temperature changes and it is reasonable to conclude that the pH and PCO2 values measured at 37°C will best reflect the normality or otherwise of the in vivo acid-base status regardless of the actual body temperature of the subject.26
Statistical Analysis
Data were checked for normal distribution before being analyzed using repeat measures ANOVA with Bonferroni's correction for multiple comparisons. The software package SPSS version 15 was used. A P value of <0.05 was considered significant.
RESULTS
i-STAT blood biochemistry data were successfully obtained from 70 pups.
Non-HI Control Experiments: Effect of Separation
The mean lactate values in all three (separated HT, nonseparated and separated NT) control groups were similar and within normal limits at 1.71 (0.27), 1.25 (0.06), 1.40 (0.17) mmol/L, respectively. The blood glucose levels were also normal and similar among these groups at 5.57 (0.75), 6.63 (0.04), and 5.04 (0.38) mmol/L, respectively (Mean ± sem).
There was no significant difference in measured blood gas variables among any of these three control groups (Fig. 1).
|
Post-Insult Gas Exchange Experiments
Temperature
Rectal temperature during the 90-min HI insult, and subsequent 5 h exposure to the gas regimes at NT or HT, are shown in Table 1. All insults were performed at 36.5°C to 36.8°C rectal temperature and there were no significant differences in temperatures during HI among the groups. In the HT groups, rectal target temperature was reached within 30 min and steadily maintained close to 32°C for the remainder of the 5-h treatment period.
|
Mixed Arteriovenous Blood Gas Analyses
The data for further analysis have been restricted to that which have actually been measured rather than derived, although for completeness, we also present the bicarbonate (derived from two measured values) and the PO2 values (although very sensitive to arteriovenous mixing effects) in Table 2.
|
At NT those pups breathing 70% XE had mean PCO2 values significantly higher than both the NTAir and non-HI control groups (P < 0.001 and P = 0.005, respectively). However, these differences were not seen with those breathing 50% XE at NT. The same pattern was seen when HT was applied; mean PCO2 when breathing 70% XE being significantly higher than NTAir and non-HI control groups (P < 0.001 and P = 0.007, respectively). However, these differences were not seen with those breathing 50% XE during HT (Fig. 2A).
|
At NT, the pH in those pups breathing 70% XE was significantly lower than those in the separated NT controls and the NTAir groups, respectively (P < 0.001 in both comparisons). Under HT, the groups breathing Air, 50% XE and 70% XE all had pH values significantly lower than the NTAir group (P = 0.001, P = 0.001, P < 0.001, respectively). The same pattern was seen when the pH values of these three HT groups were compared with those of the non-HI controls; again in all three groups the pH was significantly lower (P = 0.01, P = 0.005, P = 0.001, respectively) (Fig. 2B).
Sedative Effects
In the NT groups, both respiratory rate and spontaneous movements were significantly reduced in the pups breathing 70%Xe when compared with those breathing air or 50%Xe. There was a similar but nonsignificant trend seen in the 50% group when compared with those breathing air alone. Spontaneous movement was more depressed than respiratory rate suggesting 70% Xe has a sedative effect in spontaneously breathing rat pups (Table 3).
|
DISCUSSION
In the control groups, neither prolonged separation from the dam or HT caused any metabolic problems, which is consistent with previous findings.27
With respect to the blood gas analyses, it must be remembered that the blood samples were, and had to be, arteriovenous mixtures due to the very small total blood volume of the neonatal rats (<1 mL at P7).
Before interpreting these values, we must consider which variables are similar in both the systemic arterial and venous circulations (i.e., useful for interpretation) and those which are dissimilar (i.e., of limited use as arteriovenous mixed values will vary according to mixing ratio of each blood type). We know that systemic arterial and venous pH values are similar unless there is severe circulatory failure.28,29 Typically, the central venous pH is quoted as being lower than the arterial by approximately 0.03.30 The venous PCO2 is also similar to the arterial value.30 The pH and PCO2 are therefore useful for interpretation. We included the calculated bicarbonate (HCO3) value in Table 2 as it is derived from these two measured values alone. As with the PCO2, venous and arterial HCO3 values are very similar.31
All groups had low PO2 values due to the mixed arteriovenous nature of the samples, the lowest value being in the juvenile control group. It is difficult to speculate on the reason for this but perhaps this was a physiological reflection of the fact that this was the least stressed group having not received an earlier HI brain insult.
Five hours of breathing 50% XE at NT after a unilateral HI insult to the brain appeared to cause minimal respiratory depression, relative to both a similarly insulted group breathing air alone and uninjured control animals breathing air or XE. In contrast, if pups breathe 70% XE for 5 h post-insult, there is respiratory depression with CO2 retention and a consequent decrease in pH as well as sedation with reduced spontaneous movements. Under conditions of HT, there was again CO2 retention in the HT70%Xe group, the pH decreasing, not only in this group but also in the HT50%Xe and HTAir groups. Mean lactate levels were within normal limits for all treatment groups.
It is worth noting the clinical observation during these experiments that, whenever XE was in use, the pups always appeared pinker and less peripherally cyanosed than the NT-air group. We found a clear sedative effect of 70% Xe at NT with a 75% reduction in spontaneous movements. Although most inhaled volatile anesthetic vapors have peripheral vasodilator effects, XE appears to have little effect on the peripheral resistance in adults.25,32,33 An alternative explanation to enhanced peripheral perfusion could be that there was less stress-induced peripheral vasoconstriction as these pups were the most sedated groups.
XE-induced respiratory depression has been reported in humans breathing as little as 33% XE in radiology suites.34 The minimum alveolar anesthetic concentration (MAC) for XE in adult humans was estimated in 1969 as 71% of an atmosphere35 and more recently as 63%.36
The MAC of XE in adult rats has been estimated as 95%.37,38 The neonatal value is not known; however, we do know the neonatal/adult MAC ratio for other anesthetics. For example, the MAC of volatile anesthetics in human neonates ranges from 20% to 60% higher than the equivalent adult value,39,40 and in 9-day-old Wistar rats, the MAC is almost twice the adult value for halothane, isoflurane, sevoflurane.41 If XE follows these trends, then the MAC in neonatal rats and humans might be as high as 190% and 85%–114%, respectively. Furthermore, for common volatile anesthetics, MAC decreases with HT.42,43
The respiratory depression seen in the most affected (HT70%Xe) group was not particularly severe, with a pH at the low end of the normal range. Since there is limited evidence that higher doses of XE confer greater neuroprotection,19,20 then it is possible that clinically we might tolerate a degree of respiratory depression at least for short periods in return for some additional benefit, if this is subsequently proved to be the case.
Neonates with severe birth asphyxia in a specialist baby unit would already be mechanically ventilated, in which case, with suitable equipment, XE could be administered without concern over respiratory depression.
A more interesting question, hypothetical at this stage, concerns those babies more modestly affected with no mechanical ventilation requirement who might normally be managed/monitored in an incubator and allowed to breathe spontaneously. If, for example, a relatively short XE exposure was found to be protective, as has been suggested, would we deliver this via a spontaneous breathing method (by adding XE to the gas in a special incubator for example) or would we deliberately intubate, ventilate and subsequently extubate a neonate purely for the purpose of administering a XE based therapy? These issues cannot be fully resolved at present but these are questions we are seeking to answer, therefore, we suggest it might be reasonable to deliver up to 50% Xe in spontaneously breathing neonates, proceeding with care above this value. The rapid reversibility (emergence) characteristics of XE would offer a distinct safety advantage in such a scenario.
For XE neuroprotection studies in spontaneously breathing neonatal rats, we suggest that a concentration of up to 50% Xe can be safely administered at NT or HT for at least several hours without CO2 retention. Since the MAC in neonatal rats may be anything up to twice the value for human neonates, for human studies it would be wise to start with <50% Xe, increasing in small steps from this point. In addition, this XE concentration may need reducing if HT is added to the regime.
ACKNOWLEDGMENTS
We are grateful to SPARKS (UK) and The Laerdal Foundation for Acute Medicine (Norway) for supporting the study.
The authors gratefully acknowledge the support of Mr. Wolfgang Schmehl of Linde Gas Therapeutics for donation of the xenon used in this study. Dr. Dingley is a Board member of a University of Wales College of Medicine spin-out company involved in the development of delivery systems for medical gases including xenon.
Measured Values
pH is measured by direct potentiometry. In the calculation of results for pH, concentration is related to potential through the Nernst equation.
Partial pressure of carbon dioxide (PCO2) is measured by direct potentiometry.
Partial pressure of oxygen (PO2) is measured amperometrically. The sensor produces a current proportional to the oxygen concentration.
The three measures discussed above were each measured at 37.0°C.
Glucose is measured amperometrically. Oxidation of glucose produces hydrogen peroxide H2O2, which is then oxidized at the electrode to produce a current proportional to the sample glucose concentration.
Lactate is measured amperometrically. The enzyme lactate oxidase selectively converts lactate to pyruvate and H2O2. The H2O2 is oxidized at a platinum electrode to produce a current proportional to the sample lactate concentration.
Calculated Values
All other values are calculated. For example, the bicarbonate (HCO3) value is calculated using the Henderson-Hasselbach equation and the base excess using a similar equation. Oxygen saturation (SaO2) is calculated from measured PO2, PCO2 and pH, and assumes that normal adult hemoglobin is the dominant hemoglobin in the sample (using the oxygen-hemoglobin dissociation curve). Even in the best circumstances SaO2 values calculated from a measured PO2 and an assumed oxyhemoglobin dissociation curve may differ significantly from the direct measurement.44 In this neonatal study, an assumed adult dissociation curve will clearly give inappropriate SO2 values. Furthermore, our samples were mixed arteriovenous giving very variable PO2 (and therefore SO2 values) depending on the relative proportions of arterial and venous blood in the sample.
Footnotes
Accepted for publication November 2, 2008.
REFERENCES
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
|
