INTRODUCTION

TOP ABSTRACT INTRODUCTION MATERIAL AND METHODS RESULTS DISCUSSION REFERENCES

FETAL GROWTH RETARDATION (FGR) is of major clinical concern because there is a close relationship between the degree of growth retardation and perinatal morbidity and mortality (26). One of the causes of FGR is placental insufficiency, which is usually accompanied by a decreasing amount of amniotic fluid, resulting in oligohydramnios. This is ascribed to a reduction in fetal urine flow, and it has been shown that the hourly urine production rate in the growth-retarded human fetus is reduced (28, 29). However, the mechanism leading to a diminished urine flow in the growth-retarded fetus is not yet understood. It has been suggested that the concomitant fetal hypoxemia, caused by placental insufficiency, causes a decrease in renal blood flow (1) due to redistribution of cardiac output in favor of the brain, heart, and adrenals (3, 23, 36). The decrease in renal blood flow would lead to reduced glomerular filtration and hence decreased fetal urine production and eventually to oligohydramnios (8).

In the human fetus, renal blood flow and urine production rate cannot be measured directly. With color Doppler-ultrasound, blood flow velocity waveforms of the renal arteries are used as an index of blood flow resistance downstream. Urine production rate is estimated from consecutive ultrasound measurements of bladder size. Ultrasound Doppler studies of blood flow in the renal artery in the human growth retarded fetus show conflicting results (1, 29), and data on fetal urine production rate are not univocal (28, 31).

In the ovine fetus, renal blood flow and urine production rate can be measured directly and continuously. Furthermore, whereas urine production rate in the human fetus is found to depend on behavioral states (30), in the ovine fetus no relation was found between urine production rate and high- or low-voltage electrocortical activity (5). Therefore, fetal renal responses to hypoxemia can be much more accurately studied in the sheep fetus than in the human fetus.

Although human FGR is associated with chronic moderate hypoxia (3), most animal studies concern the fetal renal responses to acute hypoxia. In those studies, hypoxia is often severe and accompanied by acidemia (35). Moreover, the results of those studies are conflicting. The studies in which microspheres were used showed a decrease in renal blood flow (33, 37), whereas those in which electromagnetic flow probes were used showed an increase (27). Also, results concerning ovine fetal urine production rate during hypoxemia are conflicting. Depending on the severity of hypoxia and the way it was applied, a decrease or an increase or no change in fetal urine flow has been reported (11, 40).

The lack of chronic hypoxia studies prompted us to investigate the effect of sustained moderate hypoxia on renal blood flow and renal function in the near-term ovine fetus. The experiments consisted of 48 h of isocapnic hypoxia, not resulting in acidemia, but sufficient to produce redistribution of blood flow in favor of the brain at the expense of the carcass. Renal blood flow and urine production rate were continuously measured by means of ultrasonic blood flow transducers.

	MATERIAL AND METHODS

TOP ABSTRACT INTRODUCTION MATERIAL AND METHODS RESULTS DISCUSSION REFERENCES

Surgical Procedures

Two double-lumen catheters were placed in the amniotic cavity for continuous recording of intra-amniotic pressure and to return the collected fetal urine. After the uterus was closed, two pairs of Teflon-insulated stainless steel wire electrodes (N12-50F-405-0, New England Wire, Lisbon, NH) were sewn in the myometrial wall to record electromyographic activity. Maternal catheters were placed in the left femoral artery and vein for blood sampling and administration of antibiotics. All catheters and leads were tunneled to the ewe's left flank, and the abdominal wall was closed. A small catheter was implanted in the ewe's trachea for administration of nitrogen during the hypoxia studies.

After surgery, 500 ml of warm Ringer lactate was infused into the amniotic cavity to replenish the lost amniotic fluid. Before and after surgery, antibiotics [cefuroxime (Zinacef), Glaxo, Zeist, the Netherlands] were given to the ewe and the fetus (1,200 mg iv to the ewe, 100 mg iv to the fetus, and 200 mg into the amniotic cavity). Analgesics were administered to the ewe after surgery [metalmizol (Novalgin), Hoechst, München, Germany, 4 ml iv and 4 ml im]. Guidelines for care and use of animals as approved by the local animal committee were followed.

Maintenance

Physiological Measurements

Blood gas analysis was performed at hourly intervals and more frequently (every 15-30 min) during the induction of hypoxia. Each day, recordings were made for 6 h between 0830 and 1600. Control, hypoxemia, and recovery measurements were made on consecutive days. Fetal renal blood flow, carotid blood flow, arterial blood pressure, fetal heart rate, fetal urine production rate, and intra-amniotic pressure were measured continuously. All signals were recorded on an eight-channel recorder (Gould, model 8188, Cleveland, OH) and also stored on disk in an online computer system (Poly, Inspector Research, Amsterdam, The Netherlands) at a sampling rate of 10 Hz.

Glomerular filtration rate (GFR) was determined by using polyfructosan (Inutest, Bournoville, Almere, The Netherlands). At the start of the daily 6-h measurement period, a priming dose of polyfructosan (25 mg/ml) in isotonic saline was infused for 20 min at a rate of 26.67 ml/h, followed by a maintenance dose of polyfructosan (25 mg/ml) at a rate of 6.67 ml/h for the remaining time of the daily measurement. An equilibration period of 3 h was allowed for the compounds to reach steady-state plasma concentration. Then, three 1-h urine collections were made with midpoint blood samples. Urine and plasma osmolality were determined by freezing-point depression (Osmomat 030 Gonotec, Salm & Kipp, Breukelen, The Netherlands), and serum and urine samples were stored at 20°C until analysis. The anthrone method (16) was used to measure polyfructosan in serum and urine. GFR and filtration fraction (FF) were calculated with standard formulae.

At the end of the daily 6-h measurement, fetal plasma was taken and stored at 80°C until catecholamines were measured, using reversed phase chromatography and electrochemical detection (38).

At the end of experiments the animals were killed by overdoses of pentobarbital sodium. The ultrasonic flow probes were checked for their zero flow value, and experimental flow data were corrected if necessary. Fetuses were towel-dried, weighed, placement of catheters was verified, and kidneys were removed and weighed.

Data and Statistical Analysis

	RESULTS

TOP ABSTRACT INTRODUCTION MATERIAL AND METHODS RESULTS DISCUSSION REFERENCES

During the control period, fetal arterial blood gas values were within the normal ranges usually obtained in our laboratory (42) and in agreement with other studies (37). Fetal body weight at autopsy was 2.7 ± 0.1 kg, and total kidney weight was 24.7 ± 2.1 g, with no significant difference between the instrumented right kidney and the left kidney.

Fetal Blood Gas and Circulatory Responses to Hypoxemia

View larger version (34K): [in this window] [in a new window]   Fig. 1.   Mean arterial pressure (MAP), fetal heart rate, arterial oxygen saturation (SaO2), urine production rate, and renal blood flow (right kidney) of fetal sheep (n = 15) during 4 consecutive days. Data were collected for 6 h each day, always between 0830 and 1530. bpm, Beats/min.

Fetal carotid blood flow, which was measured in four animals, was significantly increased during the entire hypoxia period, from 65.4 ± 2.4 to 85.6 ± 8.5 ml/min (P < 0.01), and returned to prehypoxic levels during the recovery period (61.1 ± 3.5 ml/min). The increase in fetal carotid blood flow was significantly correlated with the decrease in fetal Pa_O2 (r is 0.71, P < 0.01, Fig. 2).

View larger version (11K): [in this window] [in a new window]   Fig. 2.   Relationship between fetal carotid blood flow and arterial O2 pressure (PaO2) expressed as changes during hypoxemia compared with prehypoxemic levels (r is 0.71, P < 0.01).

Fetal Renal Responses to Hypoxemia

1

1

View larger version (20K): [in this window] [in a new window]   Fig. 3.   Representative example of time-related responses of renal blood flow during 48 h of hypoxemia in 1 fetus. Note that renal artery blood flow increased almost immediately after initiation of hypoxemia and remained elevated all the time, although large fluctuations were observed. Also, renal artery blood flow recovered almost immediately after termination of hypoxemia.

View larger version (11K): [in this window] [in a new window]   Fig. 4.   Correlation between changes in fetal renal blood flow due to decreased PaO2 during hypoxemia. Fetal renal blood flow was inversely correlated with fetal PaO2 (r is 0.69, P < 0.0001).

Fetal urine flow. Fetal urine flow during the control measurements was within normal ranges for fetuses of this age (24). During hypoxemia (Table 2) urine flow did not change significantly from control values (0.63 ± 0.10 and 0.56 ± 0.06 ml/min, respectively). Continuous measurement of fetal urine flow showed that also during the onset of hypoxemia no changes in urine flow occurred (Fig. 1). During the course of the first day of hypoxemia fetal urine flow and urine osmolality tended to increase slightly, whereas a tendency to decrease was observed during the second day of hypoxemia.

The GFR and the FF showed a nonsignificant decrease during the course of the hypoxemia period, whereas during the recovery period GFR and FF started to increase toward control levels, but again, data did not reach significance.

Renal oxygen delivery. During the first 24 h of hypoxemia, renal oxygen delivery was maintained (7.4 ± 1.3 and 6.9 ± 1.7 ml · min¹ · 100 g¹ during control and hypoxemia day 1, respectively), but during the second 24 h of hypoxemia, oxygen delivery (5.8 ± 1.1 ml · min¹ · 100 g¹) was significantly decreased compared with the control period (P < 0.01).

Renal function. The osmotic clearance (C_osm) and free water clearance (C_H2O) did not change significantly during the 48 h of hypoxemia nor during the recovery period.

Catecholamines. Fetal plasma epinephrine and norepinephrine levels were measured in six animals. Compared with control values, plasma epinephrine increased in the 48 h of hypoxemia from 263 ± 168 to 377 ± 217 pg/ml and plasma norepinephrine increased from 520 ± 313 to 971 ± 547 pg/ml, but the difference did not reach significance.

	DISCUSSION

TOP ABSTRACT INTRODUCTION MATERIAL AND METHODS RESULTS DISCUSSION REFERENCES

Our findings indicate that prolonged fetal hypoxemia, of a degree not resulting in acidemia and/or a change in fetal Pa_CO2, causes an immediate and sustained increase in renal blood flow, whereas fetal urine production rate remains almost unchanged. These observations were made in the near-term chronically instrumented fetal sheep, which provides a convenient and generally accepted model for examining the circulatory and renal responses to hypoxemia (4). However, when investigating the renal responses, the slow recovery of fetal renal function from the stress of surgery has to be taken in consideration. We and others (17) experienced that recovery of renal function takes ~5 days, and therefore we did not start measurements until 6-7 days after surgery.

In the present study we induced moderate fetal hypoxia for 48 h, with virtually no changes in blood acid base balance, but sufficient to cause redistribution of cardiac output. The level of hypoxemia achieved in the present study was associated with a considerable increase in common carotid artery flow, indicating a substantial increase in blood flow to the fetal brain (39). Redistribution of blood flow in favor of the fetal brain occurs at the expense of peripheral organs (23). Indeed, in a previous study in fetal lambs we found at a similar degree of hypoxemia, a decrease in abdominal aorta flow of 16% with a concomitant increase in fetal common carotid blood flow of 31%, using the same ultrasonic blood flow technique (unpublished results).

It is commonly believed that the fetal hypoxic response of a reduced peripheral blood flow includes the kidneys. In the fetus, however, normal blood flow to the kidneys is already much lower than in the newborn and in the adult. A further reduction in case of hypoxemia might result in ischemic-hypoxic damage. Therefore, we believe that the widely accepted concept suggesting a decrease in renal blood flow as a fetal response to hypoxia is questionable. We found a considerable increase in renal artery blood flow during fetal hypoxemia, which remained elevated for the entire 48 h of hypoxia. Furthermore, a strong inverse correlation was found between fetal Pa_O2 and renal blood flow (Fig. 4). A similar correlation has been reported by Iwamoto and Rudolph (21) in fetal lambs that were spontaneously hypoxemic. They also found an increase in renal blood flow when arterial O₂ content decreased during conditions in which no significant differences were seen in pHa and Pa_CO2 compared with normoxemic fetuses. In those fetal sheep renal oxygen delivery was maintained, a phenomenon that was also apparent in the present study, albeit only for the first 24 h of hypoxemia. During these initial 24 h of hypoxemia, we found no changes in renal function, whereas in the second 24-h period of hypoxemia, when renal oxygen delivery became significantly reduced, both GFR and FF decreased.

Our results seem at variance with the concept that fetal hypoxemia is associated with diminished renal blood flow. However, the appreciation of the fetal renal response to hypoxemia is complicated by the various techniques and experimental conditions employed (35). Fetal hypoxemia can be induced by occlusion of the umbilical cord, by reducing blood flow to the uterine arteries, by embolization of the placenta, or by lowering the inhaled oxygen concentration of the ewe. Factors beside fetal hypoxemia, such as acidemia and hypercapnia, can have specific influences on fetal renal function. The use of different techniques yields different degrees of fetal hypoxemia with variable combinations of acidemia and hypercapnia.

In the present study there was neither progressive acidemia nor hypercapnia, which made it possible to study the effects of almost pure hypoxemia on fetal renal blood flow without all the concomitant changes that are usually associated with hypoxemia studies in the sheep fetus.

Another difference between former studies on the renal response of the fetus to hypoxia and the present study is the continuous blood flow measurement we used instead of the use of microspheres. Although a well-established method, an essential limitation of the microsphere method is the intermittent character of the technique and the restricted number of observations. Continuous measurement offers the online ability to recognize rapid changes in blood flow immediately during the entire observation period. Hitherto, only two studies on ovine fetal renal blood flow during acute hypoxemia have been reported where continuous measurement techniques were used (13, 27). It is striking that in those two acute studies, an increase in fetal renal blood flow was also observed. Chronic hypoxemia studies using continuous measurement techniques have not been reported in the ovine fetus.

Why most of the studies using microspheres found the opposite effect remains unknown, but may be at least partly explained by essential differences between the methods: the continuous ultrasonic and electromagnetic methods measure the blood flow in the artery supplying the kidney, whereas the microsphere technique uses only the microspheres trapped in the microvessels of the kidney to assess blood flow. During hypoxemia, redistribution of intrarenal blood flow takes place in favor of the inner medulla, because this area is highly vulnerable to hypoxemia (14). Blood flow to the renal medulla is derived mainly from the efferent arterioles of the juxtamedullary glomeruli. The increase in medullary blood flow will be underestimated by the usually applied microsphere technique (32). This is partly explained by the separation of red blood cells from the plasma somewhere between the renal artery and the medullary circulation (18, 32), a process called "plasma skimming."

Theoretically, the observed differences between the microsphere and the Transonic method could also be the result of a deleterious effect of the probe on the innervation of the kidney. Direct stimulation of the renal nerves produces a decrease in renal blood flow and an increase in renal vascular resistance (34). After renal denervation, there is an early vasodilatory response to hypoxemia (~10 min in duration) followed by progressive vasoconstriction (33), and sympathectomy resulted in no change in blood flow to the fetal kidneys during hypoxemia (23). Therefore, it is unlikely that the observed increase in fetal renal blood flow is caused by damage to the renal nerve.

Two fetal lambs did not recover from hypoxemia and showed progressively decreasing Pa_O2 levels, although the ewes fully recovered. In these animals, fetal renal blood flow did not decrease during recovery to the control level but increased further. Within 24 h after replacing the nitrogen administration to the ewe with air, parturition started and blood flow recording of the fetal renal arteries became impossible.

How does hypoxemia increase renal blood flow in the fetus? Because the fetal arterial blood pressure did not change during the hypoxemia period, we conclude that the increase in fetal renal blood flow is not due to increased renal perfusion pressure, but more likely is due to changes in the resistance within the kidney. This may be achieved by local and systemic responses to hypoxemia. Acute hypoxemia results in activation of the fetal neuroendocrine system, including the hypothalamic-pituitary-adrenal axis, resulting in increased fetal plasma concentrations of catecholamines (33, 35), plasma arginine vasopressin (1, 33, 35), atrial natriuretic factor (7), adrenocorticotropic hormone (6), plasma renin (33), and cortisol (6). During prolonged hypoxemia, however, many of the endocrine changes are not sustained (19). Plasma cortisol and norepinephrine levels remain elevated, but only during severe hypoxemia accompanied by hypercapnia and transient or more pronounced acidemia (6, 19, 20). Cortisol is known to increase renal blood flow, but only in high concentrations (12). Mild to moderate hypoxemia (~20% saturation decrease) without acidemia does not seem to be an adequate stimulus to increase catecholamines (24), an observation that was confirmed in the present study, whereas isocapnic hypoxia does not change fetal plasma renin activity (41).

Prostaglandins have been implicated in the regulation of renal blood flow, especially to the medulla of the kidney, and the medulla is known to contain high levels of these hormones (32). During sustained hypoxemia, prostaglandin E₂ remains elevated for the entire hypoxemic period (20). Furthermore, blood flow to the medulla of the rat kidney after administration of the prostaglandin synthesis inhibitor indomethacin was reduced (14). Another vasodilator with direct action on the renal medulla is adenosine (32), which is increased during hypoxemia (25) and particularly improves vasa recta flow (14). Nitric oxide (NO), produced from arginine by NO synthase, which is heavily concentrated in the renal medulla, also increases medullary blood flow (14). Because hypoxemia is a potent stimulator for the release of NO (2), this vasodilator could influence the vascular tone especially in the renal microcirculation (16).

The effect of hypoxemia on fetal urine production rate has been studied in a number of sheep experiments. The results of these studies are inconsistent, probably due to differences in the methods used to induce hypoxia and the duration and the severity of the hypoxia. Acute fetal hypoxemia caused fetal urinary output to decrease (35), whereas longer periods of hypoxemia start off with an antidiureses followed by normal levels of urine production after 6 h (40). Reducing uterine blood flow to induce fetal hypoxemia resulted in an increase in urine production rate (11, 40).

During acute hypoxemia, the fall in urine flow is accompanied by a decrease in fetal GFR, FF, and C_H2O, whereas urine osmolality and C_osm increase (33, 35). In the present study, however, we have shown that chronic mild to moderate hypoxemia without changes in pHa and Pa_CO2 does not induce significant changes in fetal urine production rate, GFR, FF, C_osm, or C_H2O. Also, no transient changes in urine output were found at the onset of hypoxemia, the period that can be considered as an acute hypoxemic event. Although we did not study GFR, FF, and C_osm in the period immediately after the onset of hypoxemia, after 3 h of hypoxemia no changes were found in GFR, FF, and C_osm compared with control values. Our findings are in agreement with the results of recent studies by Cock et al. (9, 10), who also found no changes in fetal urine production or GFR during prolonged hypoxemia in the absence of acidemia in the ovine fetus. All these findings suggest that the decreased urine production rate that has been observed in the human growth-retarded fetus (28, 31) is not the result of hypoxemia per se.

We speculate that the sustained increase in renal blood flow during 48 h of moderate hypoxemia protects the vulnerable areas of the fetal kidney to ischemic-hypoxic damage, because normal fetal renal blood flow is already significantly lower than in the adult.

Perspectives

Future studies should also concentrate on the possible deleterious effects of the hypoxemia-induced increase in renal blood flow in the chronic hypoxic fetus, which considerably exceeds the relatively low normal blood flow to the kidney in the fetus (compared with renal blood flow in the postnatal situation). It could be possible that a chronic increase in renal blood flow during fetal life definitely influences the setpoint or programming of receptors involved in the regulation of circulation. This might be a mechanism explaining the observed association between undernutrition in utero and hypertension and cardiovascular disease in adult life.