The preterm fetus is capable of surviving prolonged periods of severe hypoxia without neural injury for much longer than at term. To evaluate the hypothesis that regulated suppression of brain metabolism contributes to this remarkable tolerance, we assessed changes in the redox state of cytochrome oxidase (CytOx) relative to cerebral heat production, and cytotoxic edema measured using cerebral impedance, during 25 min of complete umbilical cord occlusion or sham occlusion in fetal sheep at 0.7 gestation. Occlusion was followed by rapid, profound reduction in relative cerebral oxygenation and EEG intensity and an immediate increase in oxidized CytOx, indicating a reduction in electron flow down the mitochondrial electron transfer chain. Confirming rapid suppression of cerebral metabolism there was a loss of the temperature difference between parietal cortex and body at a time when carotid blood flow was maintained at control values. As occlusion continued, severe hypotension/hypoperfusion developed, with a further increase in CytOx levels to a plateau between 8 and 13 min and a progressive rise in cerebral impedance. In conclusion, these data strongly suggest active regulation of cerebral metabolism during the initial response to severe hypoxia, which may help to protect the immature brain from injury.
- umbilical cord occlusion
fetal and neonatal animals are well known to be able to survive much longer periods of hypoxia than adults (9, 22). More recently, it has become clear that preterm animals can withstand even longer periods of profound hypoxia and hypotension/hypoperfusion without neural injury than at term (17, 28, 35). In part, this remarkable tolerance to hypoxia may be related to greater anaerobic capacity in the immature brain (48), but an important additional factor is likely to be actively regulated suppression of cerebral activity. This mechanism, which is mediated by release of endogenous inhibitory neuromodulators such as γ-aminobutyric acid (GABA) and adenosine during hypoxia, recently has been shown to be active in near-term (0.8 to 0.9 gestation) fetal sheep (6, 23, 33, 49). For example, hypoxemia or adenosine infusion in near-term fetal sheep increased oxidation of cerebral cytochrome oxidase (CytOx) on near-infrared measurements, indicating a rapid reduction in electron flow down the mitochondrial electron transfer chain and, thus, a fall in metabolic activity (40, 41). Conversely, blockade of adenosine A1 receptor activity during umbilical cord occlusion in near-term fetal sheep accelerated the onset and increased the magnitude of cerebral anoxic cytotoxic edema, with subsequently increased neuronal loss (23).
This regulated suppression of metabolic rate during hypoxia or ischemia, before energy stores are depleted, has been termed adaptive hypometabolism (38). It may be of considerable importance not only before birth but also during birth itself and in other settings involving adaptation to low oxygen levels (34). The near-term fetus of the high-altitude, hypoxia-adapted llama, for example, unlike fetuses of lowland species, responds to moderate hypoxemia with only a minor increase in cerebral blood flow but a dramatic downregulation of cerebral metabolism as shown by a markedly reduced difference between cortical and core body temperatures (14). It is unknown whether this mechanism is active in the preterm fetus, although it is suggestive that adaptive hypometabolism is more pronounced in newborns and younger animals than in adults (38).
We therefore evaluated the relationship between changes in cerebral oxygenation measured using near-infrared spectroscopy (NIRS), electroencephalograph activity, the redox state of cytochrome oxidase as a measure of mitochondrial activity, cerebral metabolism as estimated from the temperature difference between brain and body, and hypoxic cytotoxic edema as measured by changes in cortical impedance during profound hypoxia induced by complete umbilical cord occlusion in preterm fetal sheep at 0.7 gestation (3, 13).
All procedures were approved by the Animal Ethics Committee of The University of Auckland. Fourteen singleton Romney/Suffolk fetal sheep were operated on at 98–99 days of gestation (0.7 gestation, term = 147 days). In terms of cerebral maturity, the 0.7-gestation fetal sheep is comparable to the human brain at 28–32 wk of gestation, before the onset of cortical myelination (2).
Food but not water was withdrawn 18 h before surgery. Ewes were given 5 ml of Streptocin (250,000 IU/ml procaine penicillin and 250 mg/ml dihydrostreptomycin; Stockguard Labs, Hamilton, New Zealand) intramuscularly for prophylaxis 30 min before the start of surgery. Anesthesia was induced by intravenous injection of Alfaxan (3 mg/kg alphaxalone; Jurox, Rutherford, NSW, Australia), and general anesthesia was maintained using 2–3% halothane in O2. Under anesthesia, a 20-gauge catheter was placed in a maternal front leg vein, and the ewes were given a saline infusion to maintain fluid balance. Depth of anesthesia, maternal heart rate, and respiration were monitored by trained anesthesiology staff.
All surgical procedures were performed using sterile techniques (4). Following a maternal midline abdominal incision and exteriorization of the uterus and either the top or bottom half of the fetus, catheters were placed in the left fetal femoral artery and vein, right axillary artery, and the amniotic sac. An ultrasonic blood flow probe (size 3S; Transonic Systems, Ithaca, NY) was placed around the left carotid artery to measure carotid blood flow as an index of global cerebral blood flow (19, 23, 52). Two small, flexible fiber optic probes for NIRS recordings were placed biparietally on the skull 3.0 cm apart, 1.5 cm anterior to bregma. They were secured using rapid-setting dental cement (Rocket Red; Dental Ventures of America, Anaheim, CA). Through burr holes, two pairs of electroencephalograph (EEG) electrodes (AS633-5SSF; Cooner Wire, Chatsworth, CA) were placed on the dura over the parasagittal parietal cortex (5 and 10 mm anterior to bregma and 5 mm lateral) and secured with cyanoacrylate glue. To measure cortical impedance, we placed a third pair of electrodes (AS633-3SSF; Cooner Wire) over the dura 5 mm lateral to the EEG electrodes (21). A reference electrode was sewn over the occiput. A temperature probe (IT-18 thermometer, resolution 0.02°C; Physitemp, Clifton, NJ) was placed over the parasagittal dura 20 mm anterior to bregma. The burr hole was sealed, and the skin over the fetal skull was secured with cyanoacrylate glue. A second temperature probe was placed in the esophagus at the level of the right atrium as a measure of core body temperature.
A pair of electrodes was sewn over the fetal chest to measure the fetal electrocardiogram (ECG). An inflatable silicone occluder was placed loosely around the umbilical cord near its abdominal insertion (In Vivo Metric, Healdsburg, CA). All fetal leads were exteriorized through the maternal flank. A maternal long saphenous vein was catheterized to provide access for postoperative care and euthanasia. Antibiotics (80 mg of gentamicin; Rousell, Auckland, New Zealand) were administered into the amniotic sac before closure of the uterus.
A separate group of three fetal sheep were instrumented to measure local cortical blood flow (5). In this group a four-part composite probe (diameter ∼400 μm) containing emitting and receiving laser Doppler channels, a Po2 electrode, and a thermocouple was placed in the right parietal cortex ∼5 mm lateral to the midline and 5 mm posterior to the coronal suture, to a depth of 5 mm below the dura, in the gray matter of the cortex (Oxford Optronix, Oxford, UK) (23). An ultrasonic flow probe (3S; Transonic Systems) was placed on the left carotid artery, near the angle of the jaw. Vascular catheters, an ECG lead, and an umbilical cord occluder were placed as described above.
Postoperatively, all sheep were housed in separate metabolic cages with access to water and food ad libitum in a temperature-controlled room (16 ± 1°C, humidity 50 ± 10%) with a 12:12-h light-dark cycle. Five days of postoperative recovery was allowed before experiments commenced. During this time antibiotics were intravenously administered to the ewe daily for 4 days [600 mg of benzylpenicillin sodium (Novartis, Auckland, New Zealand) and 80 mg of gentamicin]. Fetal catheters were maintained patent by continuous infusion of heparinized saline (20 U/ml at 0.2 ml/h); the maternal catheter was maintained by daily flushing with heparinized saline.
Fetal mean arterial blood pressure (MAP), corrected for maternal movement by subtraction of amniotic fluid pressure (Novatrans II, MX860; Medex, Hilliard, OH) (31), ECG, and EEG were recorded continuously. The blood pressure signal was collected at 64 Hz and low-pass filtered at 30 Hz. The analog fetal EEG signal was filtered with a first-order high-pass filter at 1.6 Hz and a low-pass filter with the cut-off frequency set with the −3 dB point at 30 Hz, and it was digitized at 256 Hz (using analog-to-digital cards; National Instruments, Austin, TX). The intensity and frequency were derived from the intensity spectrum signal between 1 and 20 Hz. The total EEG intensity (power) was normalized by log transformation [dB; 20× log (intensity)], and data from left and right EEG electrodes were averaged to give mean total EEG activity. The 90% spectral edge of the EEG, i.e., the frequency below which lay 90% of the EEG intensity, was calculated from the spectra. The impedance signal was also extracted as previously described (53). The impedance of a tissue rises concomitantly as cells depolarize and fluid shifts from the extracellular to the intracellular space, and thus impedance is a measure of cytotoxic edema. Data were collected by computer and stored to disk for off-line analysis.
Experiments were conducted at 103–104 days of gestation. Fetal MAP, carotid blood flow, fetal heart rate (FHR) derived from the fetal ECG, and NIRS signals were recorded continuously from 60 min before occlusion to 60 min afterward. Fetuses were randomly assigned to the sham control (i.e., instrumented fetuses not subjected to cord occlusion, n = 7) or the occlusion group (n = 7). Fetal asphyxia was induced by rapid inflation of the umbilical cord occluder for 25 min with sterile saline, using a volume previously shown to totally compress the umbilical cord (4). Successful occlusion was confirmed by the rapid onset of bradycardia, a rise in MAP, and changes in pH and blood gas measurements.
Samples of arterial blood were collected 15 min before occlusion and 20 min after the start of occlusion for pH and blood gas determination (Ciba-Corning Diagnostics 845 blood gas analyzer and co-oximeter; East Walpole, MA) and for glucose and lactate measurements (YSI model 2300; Yellow Springs, OH). At the end of the experiment the ewes and fetuses were killed by an intravenous overdose of pentobarbitone sodium (9 g) to the ewe (Pentobarb 300; Chemstock International, Christchurch, New Zealand).
Concentration changes in fetal cerebral deoxyhemoglobin ([Hb]), oxyhemoglobin ([HbO2]), and CytOx were measured using a NIRO 500 spectrophotometer (Hamamatsu Photonics, Hamamatsu City, Japan), and data were recorded by computer for off-line analysis (3, 4). The principles of NIRS have been described previously (45). Briefly, near-infrared light at four different wavelengths between 775 and 908 nm was carried to the fetal head through a fiber optic bundle. Emerging light was collected by the second optode and transmitted to the spectrophotometer. Changes in cerebral [Hb], [HbO2], and [CytOx] were calculated from the modified Lambert-Beer law by using a previously established algorithm that describes optical absorption in a highly scattering medium (45, 56). The NIRS measures obtained are relative changes from zero, not absolute changes.
Two key parameters were calculated: total hemoglobin (THb, the sum of HbO2 and Hb) and delta hemoglobin (DHb, the difference between HbO2 and Hb). THb is related to cerebral blood volume (CBV) by the cerebral hematocrit: CBV = [THb]/(H × R), where H is the arterial hematocrit and R is the cerebral-to-large vessel hematocrit ratio, assumed to be 0.69 (56). THb is an index of CBV, given a stable blood hemoglobin and hematocrit (51). DHb is a volume-weighted average of total intravascular oxygenation in the brain (10).
Brain blood flow and heat production.
Changes in the heat production of the brain were calculated as an index of global cerebral metabolism by using the Fick principle (24). The difference between the temperature of arterial blood supplying the brain and the brain tissue itself gives the temperature increase resulting from brain metabolism (14, 24). Multiplying this difference by carotid blood flow provides an index of relative heat production by the brain. As used presently, the validity of the method depends on left carotid artery flow accurately characterizing the flow through the brain tissue under study (19, 20, 23, 27, 52). It also requires that the temperature probe placed on the dura accurately reflects brain tissue temperature and that the core body temperature measured near the right atrium is a valid index of preductal arterial blood temperature.
Fetal cortical blood flow.
It was not possible to combine placement of NIRS probes and laser Doppler probes in individual animals. We therefore examined changes in local cortical blood flow with laser Doppler (5, 24) in a separate group and compared the time course with that of carotid blood flow. Since laser Doppler measures relative changes in blood flow, changes are expressed as percent changes from baseline.
Off-line analysis of the physiological data was performed using customized Labview programs. Data were analyzed using SPSS for Windows (SPSS, Chicago, IL). For the analysis of the occlusion and recovery data, the baseline period was taken as the mean of the 60 min before occlusion. For between-group comparisons, two-way analysis of variance (ANOVA) for repeated measures was performed using 1-min averaged data. When statistical significance was found between groups or between group and time, analysis of covariance (ANCOVA) was used to compare selected time points by using the baseline control period before occlusion as a covariate. Statistical significance was accepted when P < 0.05. Data are means ± SE.
There was no difference in baseline blood gases, acid base status, glucose and lactate concentrations, FHR, MAP, EEG, and NIRS signals between the sham control group and occlusion group before occlusion. Occlusion was associated with marked hypoxemia and severe mixed respiratory and metabolic acidosis as shown in Table 1.
Pressure, heart rate, and carotid blood flow.
Following the start of occlusion, MAP rapidly increased to a peak of 55 ± 1 mmHg at 4 min of occlusion compared with 36 ± 1 mmHg in the sham control group (P < 0.001, Fig. 1, top). Thereafter, MAP progressively fell, reaching a nadir of 13 ± 1 mmHg compared with 38 ± 1 mmHg in the sham control group (P < 0.001). Following release of the occluder, MAP rapidly increased and was transiently elevated between 7 and 14 min after the end of occlusion (P < 0.005) and again between 47 and 56 min (P < 0.05).
FHR rapidly fell at the onset of occlusion and was significantly lower than that of the sham control group for the entire occlusion period (P < 0.001, Fig. 1, middle), reaching a nadir of 64 ± 5 beats/min at the end of occlusion, compared with 177 ± 4 beats/min in the sham control group. Following release of the occluder, there was a brief rebound tachycardia between 4 and 21 min after the end of occlusion (P < 0.001).
There was no significant change in carotid blood flow during the first 10 min of occlusion, although there was a trend toward lower values compared with the sham control group. Thereafter, carotid blood flow progressively fell to reach a nadir of 9.1 ± 1.5 ml/min compared with 29.7 ± 2.6 ml/min in the sham control group (P < 0.001, Fig. 1, bottom). Following release of the occluder, carotid blood flow rapidly increased, with transient hyperperfusion between 6 and 9 min after the end of occlusion (P < 0.05), followed by resolution to sham control group values.
EEG and impedance.
EEG intensity (power) was rapidly suppressed at the onset of occlusion to 4.4 ± 0.6 dB compared with 14.2 ± 0.1 dB in the sham control group (P < 0.001, Fig. 2, bottom) and remained suppressed throughout the occlusion and in the immediate reperfusion period. EEG spectral edge frequency also was suppressed during occlusion and in the immediate reperfusion period (P < 0.001, Fig. 2, middle). Average frequency during occlusion was 4.8 ± 0.3 Hz compared with 38.2 ± 1.1 Hz in the sham control group (P < 0.001).
Cortical impedance initially rose slowly at the onset of occlusion and was significantly elevated after 4 min compared with that of the sham control group (P < 0.05, Fig. 2, top), reaching a 10% increase above the sham control group after 12 min. Impedance then rose rapidly, reaching a peak of 48 ± 6% above the sham control group 2–3 min after the end of occlusion (P < 0.0001). The increase in impedance then gradually resolved but remained slightly elevated compared with the sham control group until 30 min after release of occlusion (P < 0.05).
[DHb] fell immediately after the onset of occlusion and remained significantly lower than the sham control group values for the duration of the occlusion (nadir at 10 min after the onset of occlusion, P < 0.001, Fig. 3, top). Following release of the occluder, [DHb] increased and had returned to sham control group values by 7 min after the end of occlusion. In contrast, there was no significant change in [THb] during the first 3 min after the onset of occlusion. [THb] then increased and became significantly elevated between 4 and 9 min of occlusion compared with the sham control group (P < 0.01, Fig. 3, middle). Thereafter, [THb] fell and was significantly reduced from 12 min of occlusion to the end of occlusion compared with the sham control group (P < 0.001). Following release of the occluder, [THb] gradually rose and had returned to sham control group values by 10 min after the end of occlusion.
[CytOx] was significantly elevated above the sham control group from the first minute after the onset of occlusion and remained elevated for the entire occlusion period (P < 0.001, Fig. 3, bottom). The initial rise in the first 2 min was maintained until ∼8 min (P < 0.01 vs. sham control group), followed by a further rise, reaching a stable maximum at 13 min after the start of occlusion (P < 0.01 vs. first 8 min). [CytOx] began to fall from 2 to 3 min after release of occlusion, fell in close parallel with the fall in impedance, and had reached sham control group values by 11 min after release.
Fetal temperature and cerebral heat production.
The occlusion group showed an initial, significant rise in core (i.e., body), but not brain, temperature above the sham control group during the first 6 min after the onset of occlusion (P < 0.05, Fig. 4, bottom) such that the brain-to-core temperature difference was reduced by 0.16 ± 0.01°C (P < 0.001, Fig. 4, top). With continued occlusion, both core and brain temperatures gradually fell, with a greater relative fall in core temperature such that the brain-to-core gradient widened again, reaching sham control group values in the final 5 min of occlusion. In contrast, fetal cerebral heat production, calculated as the product of carotid blood flow and the brain-to-core temperature difference, fell rapidly and remained suppressed throughout the period of occlusion (Fig. 4, middle). Following release of occlusion, fetal temperatures and the brain-to-core difference rapidly returned to sham control group values.
Local cortical blood flow.
The pattern of changes in cortical blood flow as measured by laser Doppler during the period of umbilical cord occlusion was not significantly different from that of carotid blood flow (repeated-measures ANOVA, Fig. 5). Both measures showed no significant change from baseline in the first 10 min. Cortical blood flow fell significantly from 11 min compared with baseline, and carotid blood flow, from 12 min. After the end of occlusion, there was a more rapid rise in carotid blood flow than in cortical blood flow, from 4 to 10 min, with no significant difference thereafter.
This study supports the hypothesis that preterm fetal sheep adapt to severe hypoxemia by actively suppressing brain metabolic activity. The metabolic suppression that occurs before energy failure and loss of cellular homeostasis has been termed adaptive or regulated hypometabolism (38). This type of response has been observed in late-gestation fetal sheep and llamas (6, 14, 23) and is well characterized in newborns and adults of many species (38). We now report for the first time in the preterm fetus that umbilical cord occlusion was associated with a rapid, profound fall in indexes of cerebral oxygenation, suppression of EEG activity, increased oxidation of cytochrome oxidase, and loss of the temperature difference between parietal cortex and body at a time when carotid and cortical blood flow was maintained at sham control group values, indicative of suppression of heat production by the brain.
The profound fall in intracerebral DHb (the difference between the content of oxygenated and deoxygenated hemoglobin) from the onset of occlusion as measured by NIRS is consistent with previous reports (3, 4). DHb is a well-validated measure of changes in intracerebral oxygenation (10). If arterial oxygen saturation is stable, DHb correlates closely with perfusion changes due to severe hypotension or changes in intracranial pressure (46, 50). However, as shown in the present study, during severe, prolonged hypoxia, DHb fell well before cerebral perfusion was impaired, as shown by carotid or cortical blood flow or cerebral blood volume.
NIRS also can provide information on the redox state of the copper A core of cytochrome oxidase, the terminal electron acceptor of the mitochondrial electron transport chain (12, 54). Using the algorithm developed by Wray et al. (55), changes in its redox state correlate closely with flow of reducing equivalents down the phosphorylation pathway and depletion of high-energy metabolites during hypoxia or intense oxidative activity in other tissues (8, 37, 47). In the present study there was a rapid initial rise in CytOx (i.e., an increase in oxidized CytOx) over the first 2 min of occlusion. In principle, such an increase in CytOx levels could reflect either a reduction in transfer of reducing equivalents from the tricarboxylic acid cycle or an increase in consumption of high-energy metabolites as occurs during maximal oxidative metabolism in muscle (8). The former mechanism seems more likely in the context of the profound, early deoxygenation and suppression of EEG activity in the present study and suggests a rapid decrease in metabolic activity. In turn, reduced electron transfer could reflect passive failure of mitochondrial activity due to profound hypoxia or active neuroinhibition (23, 40).
Surprisingly, there is conflicting evidence for the direction of change in CytOx redox status during hypoxia, with reports of a reduction in CytOx (i.e., a fall in oxidized CytOx) during both severe hypoxia/anoxia in the piglet (47), adult pig (16), and adult rat (37) and ischemia in the near-term fetal sheep (36) but a marked increase in oxidation during moderate hypoxia in near-term fetal sheep (41), consistent with the present findings. It is improbable that these conflicting findings reflect confounding changes in hemoglobin in the brain (55), since in the present study the pattern of CytOx changes was unrelated to that of the large changes in hemoglobin signals. We might speculate that the contradictory reports reflect relatively greater active inhibition of cerebral metabolism during hypoxia in the fetus and, in particular, premature animals than postnatally. Alternatively, it may in part reflect differences in types of insult, such as the more rapid and profound depletion of substrate delivery during ischemia (36).
Supporting evidence for active inhibition of cerebral metabolism in the initial phase of hypoxia in the present study, rather than passive oxidative failure, is seen in the extremely rapid attenuation of the temperature difference between the brain and the core body temperature after the start of occlusion, consistent with previous data in near-term fetuses (6, 14, 23, 30, 33). The response occurred before the onset of the rise in cerebral impedance, i.e., before the onset of cell swelling due to depletion of cellular high-energy phosphates. It also occurred at a time when plasma levels of adenosine, a well-established suppressor of metabolic rate, increased about twofold during 5-min cord occlusions in late-gestation fetal sheep (30). Furthermore, an infusion of adenosine in late-gestation fetal sheep, to levels comparable with those during hypoxia, increased CytOx levels (40). Together with the presence of adenosine A1 receptors inside the blood-brain barrier in near-term fetal sheep (6), these findings point to adenosine as a likely mediator of prenatal adaptive hypometabolism. Other inhibitory neuromediators that may contribute to hypometabolism include noradrenergic α2-receptor activity (39, 43), GABA (49), and the endogenous GABAA receptor agonist allopregnanolone (42). An alternative mechanism (41) may be inhibition of complex I and IV of the oxidative phosphorylation pathway by increased nitric oxide during hypoxia (11, 25).
The question arises as to whether alternative mechanisms apart from adaptive hypometabolism might also explain the calculated decrease in cerebral heat production during occlusion. One possibility is that the carotid rete, which in the sheep helps cool the brain after birth (1), might remove heat from arterial blood supplying the brain. This seems unlikely, however, because before birth there is no effective sink into which heat might be deposited; the amniotic fluid is only 0.1–0.2°C cooler than fetal core temperature, and thus the temperature gradient for heat flux from the rete is small (18). Alternatively, a larger fraction of carotid flow perfusing non-brain tissues such as the face and scalp might be diverted to the brain during severe hypoxia than was evident from carotid blood flow measurements.
However, we found in a further study that the pattern of changes during occlusion was not significantly different between cortical laser Doppler and carotid blood flow, with no increase in either measure in the first 10 min of occlusion compared with baseline. Laser Doppler correlates well with microsphere measurements (5), and these results are consistent with previous microsphere studies in near-term and term fetal sheep (26, 32). Thus these data further support a rapid reduction in cerebral metabolism at the onset of hypoxia, although we cannot exclude regional heterogeneity. In contrast, there was a significantly slower recovery of cortical laser Doppler flow than carotid blood flow after release of occlusion, implying that heat production also recovered more slowly (23). This conclusion is consistent with the delay over several minutes before cerebral impedance and CytOx began to fall after release of cord occlusion. Thus we speculate that the delayed recovery of cortical flow may be secondary to continuing cytotoxic edema and consequent increased intracranial pressure.
The initial rise in CytOx at the start of occlusion was followed by a further secondary increase between 8 and 13 min to a stable plateau. The precise mechanism is unknown. However, the temporal relationship with several key events, including the fall in mean arterial blood pressure below baseline values and the parallel fall in carotid and cortical blood flow, a reduction in cerebral blood volume, and a rapid increase in cortical impedance indicating anoxic cerebral edema, is suggestive of further compromise of cerebral energetics, with profound tissue hypoxia and reduced substrate delivery leading to further reduction in mitochondrial electron flow, and ultimately, failure of residual production of high-energy phosphates. Thus this late plateau likely represents the maximal reduction in mitochondrial electron flow.
The initial loss of the temperature differential between brain and body resolved progressively from 6 min after the start of occlusion. This change was likely related to reduced heat removal from the brain associated with the fall in carotid and cortical blood flow following failure of cardiovascular compensatory mechanisms (17). Despite these dramatic hemodynamic changes, total heat production remained stable and suppressed, further suggesting that cerebral metabolism was maximally depressed throughout the second half of occlusion.
The marked delay between initial inhibition of cerebral activity and mitochondrial function in contrast with the later final plateau of CytOx levels and the onset of cytotoxic edema is consistent with our previous reports showing that the rise in cerebral impedance is much slower during ischemia in preterm fetal sheep than at term (15, 44). This suggests that actively regulated metabolism is effective in protecting the very immature brain for much longer than at term. The present study indicates this protection becomes of limited effectiveness after ∼8–13 min in the ovine fetus at 0.7 gestation. This timing is consistent with the duration of hypoxia required to cause injury in the preterm fetus. Whereas in the near-term fetal sheep, 10 min of umbilical cord occlusion leads to hippocampal damage (23, 29, 35), the preterm fetus at 0.6–0.65 gestation can tolerate 10–20 min of umbilical cord occlusion or ischemia without injury (17, 28, 35). In the present paradigm in the 0.7-gestation fetal sheep, we also found no injury after 10 min of occlusion and mild selective neuronal loss after 15 min of occlusion (unpublished data).
In contrast with the marked delay between the rapid increase in oxidation of CytOx and the subsequent rise in cortical impedance during occlusion, after the release of occlusion CytOx values, calculated heat production and impedance resolved to sham control group values in parallel, reciprocally with the increase in cerebral oxygenation. These data suggest that mitochondrial function was intact at the end of occlusion, with a rapid recovery of electron flux as oxygen delivery was restored. Furthermore, this finding is consistent with the observations that adenosine levels very rapidly return to normal after reoxygenation (7), suggesting that metabolism is not actively suppressed in this immediate recovery period.
In summary, the present study provides evidence of a biphasic regulation of cerebral metabolism during severe hypoxia in the very immature ovine brain. In the initial phase of adaptation, cerebral oxygenation, EEG activity, and the brain-to-core temperature differences are rapidly and profoundly diminished, whereas cytochrome oxidation is increased, indicating a reduction in electron flow down the mitochondrial chain. In the late phase, there is a further increase in oxidation of cytochrome oxidase to a plateau after 13 min, associated with a rapid increase in anoxic cytotoxic edema, suggesting that residual high-energy metabolites were depleted at this time. This late deterioration is associated with systemic compromise as shown by the development of profound hypotension and hypoperfusion. Thus these data suggest that initial adaptive brain hypometabolism is an important contributor to the remarkable tolerance to hypoxia of the preterm fetus but becomes of limited importance as severe hypoxia is continued.
We acknowledge the support of the Health Research Council of New Zealand, the Auckland Medical Research Foundation, Lottery Grants Board of New Zealand, National Institutes of Health Grant R01 HL-65494, and the March of Dimes Birth Defects Foundation.
The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked “advertisement” in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
- Copyright © 2007 the American Physiological Society