Main

An important adaptive response of the fetus when oxygenation becomes compromised is the well described protective process whereby blood flow is centralized in favor of the brain, heart, and adrenals(1–6). For the ovine fetus studied near term with acutely induced short-term hypoxemia, the blood flow increase to the brain, heart, and adrenals is such that oxygen delivery to these organ tissues is well maintained(1–3). With sustained hypoxemia of several hours duration, the blood flow increase to these so-called vital organs is seen to be maintained(5), although with onsetting fetal acidemia and a progressive decline in blood O2 content, oxygen delivery to the brain will fall. As reviewed by Jensen and Berger(6), these blood flow changes variably involve local endothelial factors, the autonomic nervous system, and both parasympathetic and sympathetic pathways, and the release of circulating vasoactive substances with specific vascular and metabolic effects. Among the latter, there is clear evidence for activation of the pituitary adrenal axis, resulting in increased ACTH and cortisol concentrations(6–9), which allegedly help to maintain arterial blood pressure during fetal hypoxemia(10). Of note, we have recently shown that ACTH also affects adrenal cortical blood flow in the ovine fetus during late gestation(11), suggesting that the increase in adrenal blood flow observed during fetal hypoxia may in part relate to functional activation in association with changes in plasma ACTH concentrations.

Although hypoxia-related blood flow changes have been well characterized for the ovine fetus studied near term, there is limited information on the younger fetus. In studies at approximately 0.6 to 0.7 of gestation by Iwamotoet al.(12) and Jensen and Berger(6), centralization of blood flow was evident during short-term compromises in fetal oxygenation with increases to the brain, heart, and adrenals, although the overall response appeared less than that reported near term. This may relate to an immaturity of those mechanisms underlying the cardiovascular response to hypoxia at younger gestational ages as reported by others(9, 13) and indeed suggested by the lower values for hypoxemia-stimulated catecholamines in the study of Iwamoto et al.(12). However, there has been no study of the circulatory response of the preterm fetus to a sustained hypoxic insult and whether the “centralization” of blood flow is maintained. Nor is there information on associated endocrine changes which might relate to organ specific changes in blood flow. We have therefore studied the effect of sustained hypoxia of several hours' duration with resulting metabolic acidosis on the regional blood flow of the preterm ovine fetus at 0.75 of gestation and sought correlation with associated changes in pituitary, thyroid, and adrenal hormones.

METHODS

Surgical procedures. Twelve premature fetal sheep of mixed breed were surgically prepared for study between 107 and 110 d of gestation(term, 147 d). The surgical procedures and postoperative care of the animals have previously been described(14). Polyvinyl catheters(V3, Bolab, Lake Havasu City, AZ) were placed in the brachiocephalic artery and the sagittal sinus for blood sampling, in the inferior vena cava for injection of microspheres, and in the amniotic cavity (V11, Bolab) for measurements of amniotic pressure. A polyvinyl catheter (V11, Bolab) was also placed in the maternal femoral vein. Postoperatively, animals were maintained in individual cages suitable for continuous monitoring and were allowed at least 3 d to recover from surgery before the experiments began. Animal care followed the guidelines of the Canadian Council on Animal Care.

Physiologic measurements. Each animal was studied during a 2-h normoxic control period, an experimental period of 8 h with sustained hypoxemia, and a 1-h period of recovery. A continuous recording was obtained for fetal arterial blood pressure, fetal heart rate, and amniotic fluid pressure as previously described(14).

During the normoxic control period, a Perspex chamber (volume, 4.6 cubic feet), designed to alter maternal inspired gas concentration while permitting ewes to eat and drink, was placed in the front of the monitoring cage and perfused with air at 40 L/min as previously described(5). The side facing the ewe was open, and when the chamber was in use, the open side was closed by a large cone-shaped piece of polythene sheeting with an opening placed over the ewe's head and loosely tightened around the neck. The specific gas mixture delivered to the chamber was made up from cylinders of medical air, N2, and CO2, using precision rotameters (Union Carbide, Kesbey, NJ). One hour before the initiation of induced hypoxemia, radionuclide-labeled microspheres were injected into the inferior vena cava for determination of regional organ blood flows followed immediately by simultaneous blood sampling from the fetal brachiocephalic artery and sagittal sinus. Sustained fetal hypoxemia was then induced by lowering the O2 concentration within the Perspex chamber to 8-11% with 3% CO2 added as monitored by periodic measurement of O2 and CO2 gas tensions. The oxygen concentration was progressively lowered within this range over the 1st h of study in an attempt to induce an onsetting fetal acidemia, as in our previous study with sustained hypoxemia in animals near term(5). At 1 h and again at 8 h of sustained hypoxemia, further microsphere injections were performed followed by fetal blood sampling. The animals were then returned to room air, and after a 1-h period of recovery, a final microsphere injection and fetal blood sampling were performed. All blood samples were analyzed for blood gases and pH, oxygen content, glucose, and lactate. Brachiocephalic arterial samples were also collected for measurement of IR-ACTH, cortisol, TSH, and thyroxine.

Blood gases and pH were measured using an ABL-3 blood gas analyzer with temperature corrected to 39.5°C (Radiometer, Copenhagen, Denmark). Oxygen and carbon dioxide concentrations within the Perspex chamber were also analyzed on an ABL-3 blood gas analyzer. Blood oxygen saturation and Hb were measured in duplicate by an OSM2 (Radiometer) hemoximeter. Oxygen content was then calculated using an O2 capacity of 1.36 mL of O2/g of Hb. Blood samples for hormone determination were transferred to plastic tubes and centrifuged at 4°C for 15 min at 2,000 rpm. The plasma was removed and stored at -70°C until analyzed. IR-ACTH, cortisol, TSH, and thyroxine were subsequently measured by RIA as previously described(7, 15, 16).

Regional blood flow was measured with 15-μm diameter microspheres(DuPont NEN, Boston, MA) labeled with one of four different radioisotopes(141Ce, 51CR, 85Sr, or 46Sc) using methods previously published(14). The reference sample was withdrawn from the brachiocephalic artery at a rate of 1.06 mL/min via a Harvard infusion-withdrawal pump for 2 min post-microsphere injection. With each blood flow determination, approximately 2 mL of arterial blood were withdrawn. On completion of each microsphere injection and subsequent blood sampling, the fetus was transfused with 4 mL of maternal blood.

On completion of the last microsphere injection and blood sampling, the ewe and fetus were killed, and an autopsy was performed on the fetus to validate catheter placement. The fetus was then weighed, followed by removal and weighing of fetal organs and tissues. These included the brain, heart, adrenal glands, muscle, and thyroid. The fetal brain was fixed in formalin and subsequently dissected into the following regions: right and left cerebral hemispheres, subcortical structures (thalamus, epithalamus, hypothalamus, corpus striatum, and midbrain), cerebellum, pons, medulla, pituitary, and choroid plexus. The fetal heart was trimmed free of great vessels, pericardium, and epicardial fat, and the right and left ventricular walls were dissected free. To separate the adrenal cortex from the medulla, a longitudinal incision was made through the cortex, and the medulla was teased out gently. It was recognized that complete separation of cortex and medulla would be difficult to achieve because of the interdigitation of cortical and medullary tissue that is observed at this stage of gestation(17). Muscle specimens were taken from the shoulders and from the nuchal area of the neck. All tissues were weighed separately and analyzed for radioactivity (Compugamma model 1282, LKB Wallac Oy, Turku, Finland). Regional blood flows were then calculated for upper body tissues as previously described by Makowski et al.(18). Regional flow/weight ratios (the percentage of total spheres in an anatomic area divided by the percentage weight contribution to that area) were calculated for adrenal tissues as an index of relative flow per g of tissue within a given anatomic area. Such analysis eliminates the need for a lower body reference sample, and, while ignoring absolute changes in blood flow, still serves to highlight any changes in the distribution of flow within an anatomic area. The reference blood samples and all pooled organ samples contained >400 microspheres. Although most subregions of the brain also contained >400 microspheres, the pituitary gland and choroid plexus often had less than this number, but with low count measurements for both of these tissues evenly distributed for the four study conditions.

Data analysis. Results obtained from the 12 animals for the four study time points are presented as grouped means and SEM. Regional oxygen delivery to tissues was calculated as the product of arterial oxygen content and regional blood flow. The data were examined by analysis of variance for repeated measurements, followed by Scheffe's test if a significant F ratio was obtaine (p < 0.05). Significance for data that were not normally distributed was determined by a nonparametric test (Wilcoxon test). Data are also presented relating hormonal values to blood gas changes and regional blood flow for given endocrine tissues, with correlation coefficients determined by regression analysis. Data on cerebral metabolism for glucose, lactate, and oxygen have previously been published(19).

RESULTS

Fetal cardiovascular parameters and arterial blood gas, pH, and oxygen content measurements from the control, hypoxic, and recovery periods are shown in Table 1. Fetal arterial Po2 and oxygen content were significantly decreased at 1 h of hypoxic study with a further decrease when measured at 8 h. There was a small but significant fall in fetal arterial Pco2 likely because of maternal hyperventilation, despite the added 3% inspiratory carbon dioxide. Fetal arterial pH, although little changed after 1 h of induced hypoxemia, was variably decreased to 7.22 ± 0.03(p < 0.01) by 8 h (range, 7.07-7.38), reflecting a metabolic acidosis as indicated by the corresponding change in base excess. All of these values showed a return toward control levels when measured at 1 h of recovery. Mean arterial blood pressure at the time of metabolic measurements showed little change throughout the study. Fetal heart rate, although somewhat decreased at 1 h and increased at 8 h of induced hypoxemia, was likewise not significantly changed from control values. However, fetal heart rate was significantly increased (p < 0.01) when measured at 1 h of recovery.

Table 1 Fetal cardiovascular and arterial blood gas, pH, and oxygen content measurements
Full size table

Hormonal measurements. Fetal plasma IR-ACTH, although variably increased in five of the animals when measured at 1 h of hypoxic study, was increased in all animals at 8 h of study, with a mean increase of greater than 12-fold (p < 0.01) (Table 2). Fetal plasma cortisol, although low throughout the study, was likewise increased in all animals when measured at 8 h of hypoxic study, with a mean increase of greater than 4-fold (p < 0.05) (Table 2). Mean plasma IR-ACTH and cortisol values had fallen at 1 h of recovery, although continued to be significantly elevated compared with the control period. For all measurements (12 fetuses and 4 measurements), ACTH values were highly correlated with the degree of fetal acidemia as reflected by base excess values (r = -0.87, p < 0.001), whereas slightly less so with the degree of fetal hypoxia as reflected by arterial oxygen content values (r = -0.67, p < 0.01). Cortisol values showed a modest correlation to both base excess values (r = -0.52,p < 0.05) and arterial oxygen content values (r =-0.59, p < 0.01) and to ACTH values (r = 0.46,p < 0.05). Fetal plasma TSH and thyroxine showed no change in response to the 8 h of sustained hypoxemia and remained little changed throughout the study (Table 2).

Table 2 Fetal hormonal measurements
Full size table

Regional blood flow and oxygen delivery measurements. Mean blood flow measurements from the organs and tissues examined from the control, hypoxic, and recovery periods are shown in Table 3. Data on oxygen delivery are given in Table 4.

Table 3 Fetal organ and tissue blood flow measurements
Full size table
Table 4 Fetal organ and tissue oxygen delivery measurements
Full size table

Brain. Blood flow to the brain showed a marked increase in response to induced hypoxemia from the control value of 102 ± 7 mL·min-1·100 g-1 to 172 ± 20 mL·min-1·100 g-1 when measured after 1 h(p < 0.01), and with an additional small increase to 193 ± 12 mL·min-1·100 g-1 when measured after 8 h(p < 0.01). Oxygen delivery to the brain was initially maintained as the increase in brain blood flow was sufficient to compensate for the fall in arterial O2 content. However, by 8 h, cerebral oxygen delivery was significantly decreased (p < 0.01), as the small further increase in blood flow was insufficient to compensate for the continued drop in arterial oxygen content. Of note, for all measurements at this time, cerebral oxygen delivery was well correlated with the degree of fetal acidemia as reflected by base excess values (r = 0.81, p < 0.01). These parameters showed a return toward control levels when measured at 1 h of recovery, although oxygen delivery to the brain continued to be depressed.

Regional blood flow for all structures within the brain showed a similar percent increase from normoxic control values of approximately 70% at 1 h, and approximately 90% at 8 h of hypoxic study, except that for the pituitary gland, where the increase in blood flow was much smaller, and that for the choroid plexus, where blood flow actually fell. These blood flow values all showed a return toward control levels when measured at 1 h of recovery, except that for the choroid plexus which continued to be significantly decreased. For all measurements (12 fetuses and 4 measurements), pituitary blood flow showed no correlation with IR-ACTH or TSH values, nor indeed, to either base excess or to arterial oxygen content values.

Heart. Blood flow to the heart was significantly increased in response to induced hypoxemia, from the control value of 211 ± 15 mL·min-1·100 g-1 to 422 ± 54 mL·min-1·100 g-1 at 1 h (p < 0.01) and with a further increase to 720 ± 72 mL·min-1·100 g-1 when measured at 8 h (p< 0.01). The increase in myocardial blood flow was such that oxygen delivery was well maintained throughout the period of hypoxic study. Blood flow to the heart, although returning toward control levels, continued to be elevated during the early period of recovery, resulting in a significant increase in oxygen delivery to the myocardium at this time (p < 0.05).

Regional blood flow within the heart was higher to the right ventricle when compared with the left ventricle, for both control and hypoxic study measurements. However, the percent change from normoxic control values was similar for both ventricles throughout the study.

Upper body muscle. Blood flow to the skeletal muscles of the upper body showed no significant change during the 8 h of hypoxic study. As such, oxygen delivery to skeletal muscle fell progressively during hypoxemia, and continued to be significantly decreased at 1 h of recovery.

Thyroid. Blood flow to the thyroid gland also showed a marked increase in response to induced hypoxemia, from the control value of 312± 38 mL·min-1·100 g-1, to 533 ± 90 mL·min-1·100 g-1 when measured after 1 h(p < 0.01), an increase comparable to that of the brain. However, thyroid blood flow fell thereafter and was significantly less than normoxic control values when measured at 1 h of recovery. Although oxygen delivery to the thyroid gland was thus maintained at 1 h of hypoxic study, a significant decrease was noted when measured at both 8 h of induced hypoxemia and at 1 h of recovery.

Adrenal. Blood flow to the adrenal cortex relative to that of the adrenal medulla was 0.06 ± 0.01 during the period of normoxemia, indicating that at this gestational age blood flow to the adrenal cortex, per 100 g of tissue, is considerably less than that to the adrenal medulla. However, with induced hypoxemia of 8-h duration there was a 3-fold increase in adrenal cortical flow relative to that of the medulla, indicating a differential effect on these vascular beds. This increase in cortical flow relative to that of the medulla continued to be evident when measured at 1 h of recovery. For all measurements (12 fetuses and 4 measurements), adrenal cortical blood flow relative to medullary blood flow did show a modest correlation to ACTH values (r = 0.59, p < 0.01) as well as to base excess values (r = -0.59, p < 0.01). There was no correlation to cortisol values.

DISCUSSION

The use of the Perspex chamber to lower maternal oxygen concentrations allowed the animals continued access to food and water and results in little change in maternal ACTH and cortisol as previously reported for animals similarly studied near term(7), thus supporting this technique as a useful means of studying sustained fetal hypoxemia. Fetal plasma IR-ACTH concentrations here studied at 0.75 of gestation, increased approximately 12-fold in response to sustained hypoxemia. This is still considerably less than that we have reported for near term animals studied at 0.9 of gestation (≈50-fold increase)(7), but greater than that for immature fetuses studied at 0.6 of gestation (≈4-fold increase)(9). These findings are consistent with a maturational change in the pituitary responsiveness to corticotropin-releasing hormone, as previously reported by Challis and Brooks(20), although it is acknowledged that the sustained hypoxic insults for these three gestational aged groups are not entirely the same. A maturational change in the adrenal responsiveness to ACTH is likewise suggested, as plasma cortisol values in the present study increased approximately 4-fold in response to sustained hypoxemia, which is less than the ≈8-fold increase for near term animals(7), but greater than the negligible response in immature fetuses(9). Plasma ACTH and cortisol values were inversely related to both arterial oxygen content and base excess values, being higher the greater the degree of hypoxemia and/or acidemia, which is in keeping with the previously reported “dose response” activation of the pituitary-adrenal axis to a graded hypoxic insult(8). Conversely, neither mean plasma TSH nor thyroxine showed any significant change in response to the 8 h of induced hypoxemia. Although the hypothalamic mechanisms to stimulate TSH release are certainly operative in the ovine fetus at this gestational age, as evidenced by the response to in utero cooling(21), activation of this axis in response to hypoxemia does not appear to occur.

Blood flow to the brain was increased in a sustained manner through the 8-h period of hypoxic study. This increase in blood flow was such that, at 1 h of induced hypoxemia, cerebral oxygen delivery was maintained, consistent with the well known inverse relationship between cerebral blood flow and arterial oxygen content described for the near term ovine fetus(22). By 8 h of induced hypoxemia, cerebral oxygen delivery had decreased by approximately one-third, which may relate in part to the small fall in fetal arterial Pco2 noted at this time, and/or to the expected rightward shift in the fetal oxyhemoglobin dissociation curve with the systemic acidosis noted, which Rosenberg et al.(23) have shown would act to decrease cerebral blood flow. This is supported by the noted relationship between cerebral oxygen delivery and base excess values whereby the fall in O2 delivery was related to the degree of metabolic acidemia. Cerebral oxygen delivery is also decreased in the near term ovine fetus with sustained hypoxemia(5), with the decrease related to both the degree of acidemia and terminally to a fall in perfusion pressure.

Regional blood flow within the brain was greater to brainstem structures, both during the control and hypoxic study periods, which has also been reported by Gleason et al.(24) for animals at 0.65 gestation, and is similar to that seen in animals near term(14). However, the percent change in blood flow in response to the 8 h of sustained hypoxemia was similar for all regions of the brain (excepting the pituitary and choroid plexus), with no hierarchy of increase evident as seen for the older gestational aged fetus where that to the pons, medulla > subcortex > cerebellum, cerebral cortex(14). This developmental change in the brain's blood flow response to hypoxia may relate to regional changes in the brain's metabolic response which become evident with the appearance of well defined electrocortical states at approximately 120 d of gestation(25).

The precision with which regional blood flows within the brain can be determined is dependent on the number of microspheres trapped in the tissues and reference sample(26, 27). Although most subregions of the brain contained >400 microspheres, estimates of pituitary and choroid plexus flows were often based on smaller numbers of microspheres. Baseline flows were, however, close to those previously reported for the ovine fetus(28). Although it is recognized that entrapment of a smaller number of microspheres entails an increase in methodologic variability, this should not contribute to any biasing of results, with the changes noted in blood flow to these tissues unlikely to be due solely to differences in microsphere trapping. In the present study, arterial blood flow to the pituitary was also increased in a sustained manner through the 8-h period of hypoxic study; however, this increase was much less than that for the rest of the brain, suggesting that hypoxic regulatory mechanisms are different as previously reported in studies in adult dogs(29). There was no apparent relationship of pituitary blood flow values to IR-ACTH or TSH values during the study to suggest a metabolic basis for the blood flow change, and such a relationship seems highly unlikely given the mean changes noted for each of these parameters. Although flow to the choroid plexus was little changed when measured at 1 h of hypoxic study, in keeping with a lack of any hypoxic regulatory mechanisms, this flow was markedly decreased when measured at 8 h and continued to be decreased during the early period of recovery, the reason for which is not readily apparent.

Myocardial blood flow increased in a stepwise manner from the control value to that measured at 1 and 8 h of hypoxic study and was well predicted by the corresponding fall in arterial oxygen content as previously reported for the ovine fetus near term(2, 3, 30). As such, oxygen delivery to the heart is well maintained at this gestational age with sustained hypoxic insults, in contrast to that for the brain. Regional blood flow within the heart was higher to the right ventricle when compared with the left ventricle as also reported for the ovine fetus near term(31), and in keeping with the higher cardiac output and thus metabolic demands of the right ventricle during fetal life(31).

Blood flow to the skeletal muscles of the upper body showed no significant change through the 8-h period of hypoxic study which is similar to our finding for the ovine fetus near term in response to sustained hypoxemia(5), but contrasts with published data on overall carcass flow, which shows a fall in perfusion in response to short-term hypoxemia for both the preterm(12) and near term(1, 2) fetus. Blood flow to the thyroid here measured under resting conditions was higher than that to the heart on a per weight basis and considerably higher than that to the brain. This would suggest a high degree of functional activity for the thyroid, in keeping with the metabolic maturation of this organ and the rapid rise in circulating thyroxine levels for the ovine fetus at this gestational age(32). Of interest then, is the blood flow response to hypoxic “stress” with an initial increase comparable to that of the brain, but a fall off thereafter such that blood flow to the thyroid was significantly less at 1 h of recovery than during the control period. However, if alterations in thyroid functional activity underlie this blood flow response, as local hypoxic regulatory mechanisms seem unlikely, this occurred in the absence of any measurable change in circulating levels of TSH or thyroxine.

Although absolute changes in blood flow for lower body tissues could not be determined in the present study, as a lower body reference catheter was not placed, regional flow/weight ratios were calculated for adrenal tissues to provide for a measure of adrenal cortical flow relative to that of the medulla. In response to the 8 h of induced hypoxemia, adrenal cortical blood flow, relative to that of the medulla, was increased 3-fold, indicating a differential effect of sustained hypoxia on these vascular beds. This finding differs from that of Jensen et al.(4) with acute hypoxic study of 30-min duration for animals near term whereby adrenal blood flow was increased similarly for both the cortical and medullary areas. However, the differential vascular effect of hypoxia on the fetal adrenal may become apparent only with more prolonged insults, as suggested by the present results, whereby little change was evident at 1 h of hypoxic study. These findings may relate to a delayed increase in the functional activity of the fetal adrenal cortex in response to the activation of the hypothalamic-pituitary-adrenal axis and is in keeping with the delayed increase in adrenal cortical flow that we have recently reported for the ovine fetus in response to a 24-h infusion of ACTH(11). Of note, values for adrenal cortical flow relative to medullary flow here measured were well correlated with those for ACTH, further supporting a coupling of blood flow to the metabolic stimulation of the adrenal cortex. These values were also well correlated with those for base excess, and thus a differential vascular effect due to blood gas changes is not precluded, nor indeed is an effect of other hormones released during hypoxia.

In response to a sustained hypoxic insult, the preterm ovine fetus at 0.75 of gestation thus demonstrates a centralization of blood flow which is qualitatively similar to that of animals studied near term. Whereas blood flow to skeletal muscle remains unchanged, that to the brain and heart are increased throughout the 8 h of induced hypoxemia. However, although myocardial oxygen delivery is well maintained throughout the 8 h of hypoxic study, that to the brain falls terminally. As such, hypoxic regulatory mechanisms with sustained hypoxemia may be better protective of the heart, with the brain more vulnerable to hypoxic injury. Although absolute blood flow to the adrenal gland was not measured, the observed increase in adrenal cortical blood flow, relative to that of the medulla and the related change in ACTH, supports a role for the functional activation of the gland in the blood flow increase with hypoxia.