The effects of asphyxia on renal function in fetal sheep at midgestation


  • A. E. O'Connell,

    Corresponding author
    1. Department of Physiology and Pharmacology, School of Medical Sciences, University of New South Wales, Sydney, NSW 2052, Australia
    • Corresponding author
      A. E. O'Connell: Department of Physiology and Pharmacology, School of Medical Sciences, University of New South Wales, Sydney, NSW 2052, Australia. Email:

    Search for more papers by this author
  • A. C. Boyce,

    1. Department of Physiology and Pharmacology, School of Medical Sciences, University of New South Wales, Sydney, NSW 2052, Australia
    Search for more papers by this author
  • E. R. Lumbers,

    1. Department of Physiology and Pharmacology, School of Medical Sciences, University of New South Wales, Sydney, NSW 2052, Australia
    Search for more papers by this author
  • K. J. Gibson

    1. Department of Physiology and Pharmacology, School of Medical Sciences, University of New South Wales, Sydney, NSW 2052, Australia
    Search for more papers by this author


To determine whether damage to the fetal kidneys plays a role in the formation of hydrops fetalis following a severe asphyxial episode, six chronically catheterised fetal sheep, at 0.6 gestation (90 days; term 150 days), were subjected to 30 min of complete umbilical cord occlusion. During the occlusion period, mean arterial pressure, heart rate and renal blood flow decreased (P < 0.001). There were falls in arterial pH and PO2 and a rise in PCO2 (P < 0.001). Urine flow rate decreased (P < 0.005), as did the excretion rates of sodium and osmoles (P < 0.05). However, by 60 min after release of occlusion, urine flow rate was similar to control values. By the end of day 1, most renal variables returned to normal. At post-mortem, 72 h after occlusion, all asphyxiated fetuses showed gross signs of hydrops. Body weight was higher (P < 0.05) due to fluid accumulation in the peritoneal (P < 0.001) and pleural cavities (P < 0.05) as well as subcutaneously (P < 0.05). Amniotic/allantoic fluid volume was increased (P < 0.05). Kidney histology was normal except for clusters of apoptotic cells in some proximal tubules. In conclusion, this severe asphyxial episode caused surprisingly little damage to the kidney and the changes in renal function were very transient. Thus renal damage was not important in the development of hydrops. Possibly, the midgestation fetal kidney has a limited capacity to increase urinary salt and water excretion in response to increased fluid delivery across the placenta.

Research in the area of fetal asphyxia has tended to concentrate on perinatal asphyxia using either near-term fetuses (Daniel et al. 1978; Ikeda et al. 1998), or spontaneously breathing newborn animals (Alward et al. 1978; Stonestreet et al. 1984; Gouyon et al. 1987). It is becoming increasingly clear however, that asphyxial events can occur throughout gestation (MacLennan, 1999). If transient, such an event may go unnoticed in younger fetuses, as they are able to survive longer periods of reduced oxygen supply than older fetuses (Bennet et al. 1999). Even so there could be ischaemic damage to developing organs. This may alter their pattern of development and lead to permanent alterations which could impact on adult health.

When the umbilical cord was completely occluded for 30 min in midgestational fetuses (90 days; term 150 days) these asphyxiated fetal sheep subsequently developed hydrops fetalis, or fetal oedema (Lumbers et al. 2001). Similarly, Keunen et al. (1997) noted that three of four midgestational fetuses, asphyxiated for 20 min, showed considerable generalised oedema, which they attributed to cardiovascular decompensation. Interestingly, the hydropic fetuses of Lumbers et al. (2001) showed marked activation of the renin angiotensin system (RAS). This is noteworthy as hydrops developed when Faber & Anderson (1994) infused angiotensin I (Ang I) into neprectomised fetal sheep. If the kidneys were intact, however, the infusion resulted in polyhydramnios (excess amniotic fluid) (Anderson & Faber, 1989). They concluded from this data that Ang II (the biologically active product of Ang I) promotes salt and water transfer from the mother to the fetus. Because the nephrectomised fetuses had no kidneys, they were unable to excrete the excess fluid, and it accumulated within the fetal body. It is possible that if the fetal kidneys were damaged during an asphyxial episode so that renal function was severely impaired, then this renal damage, in combination with the known activation of the RAS, could explain the development of hydrops fetalis after a midgestational asphyxial episode. The purpose of this study was to examine the renal and cardiovascular function in 90-day-old fetuses before, during and after a 30 min complete cord occlusion.


Surgical preparation

Surgical and experimental protocols were approved by The Animal Care and Ethics Committee, UNSW. At 0.6 gestation (87-91 days; term = 150 days), ewes were anaesthetised with an i.v. injection of 1 g thiopentone sodium (Pentothal 500 mg, Abbott Australasia Pty Ltd, Australia), intubated and ventilated, and anaesthesia was maintained by 2 % halothane (Fluothane, Zeneca Ltd, UK) in oxygen, which crosses the placenta to anaesthetise the fetus. Maternal antibiotics, 1.6 ml gentamicin (Gentam, 50 mg ml−1, Troy Laboratories Pty Ltd, NSW) and 2 ml procaine penicillin (Ilium Propen, 300 mg ml−1, Troy Laboratories), were given i.m. at induction.

Under aseptic conditions, graded polyvinyl catheters (0.9 mm o.d., 0.5 mm i.d.; 1.2 mm o.d., 0.8 mm i.d.; 1.5 mm o.d., 1.0 mm i.d.) were placed into a fetal femoral artery and both tarsal veins. The fetal bladder was catheterised suprapubically and a catheter (2.7 mm o.d., 1.5 mm i.d.) placed in the amniotic cavity. A doppler flow probe (1.6 mm i.d. cuff, subminiature 20 MHz piezoelectric transducer; Iowa Doppler Products, USA) was placed around the fetal left renal artery and an inflatable occluder (In Vivo Metric, Ukiah, California, USA), placed around the umbilical cord. Approximately 300 ml of warm 0.9 % saline was placed into the uterus prior to closure. Catheters were placed into a maternal femoral vein and artery (2.7 mm o.d., 1.5 mm i.d.). All catheters and probes were exteriorsed via the maternal flank. The same dose of antibiotics, as above, was administered to the fetus (via the amniotic cavity), and ewe (i.m.) at the end of surgery.

After surgery, sheep were housed in metabolic cages in a temperature controlled room with other ewes. They had free access to food (1200 g chaff and 300 g oats) and water. For 3 days, gentamicin and penicillin were given daily to fetus and ewe and catheters were flushed with heparanised (100 u ml−1) 0.9 % saline. Experiments were carried out 3-4 days after surgery.

Experimental protocol

Experiments were conducted on the fetus at 91-95 days gestation. Ewes were randomly assigned to either the sham group, or the occlusion group.

On day 1, fetal and maternal blood pressures and heart rates, and intra-amniotc pressure were measured continuously using pressure transducers (Biotrans II Dome/Bio BTR-3, Daralan Pty Ltd T/AS The Critical Assist Group, NSW, Australia) connected to a polygraph (Model 79D, Grass Instrument Co., Quincy, MA, USA). Renal blood flow (RBF) was measured continuously using a 545C-4 directional pulsed Doppler flowmeter (Bioengineering, University of Iowa, IO, USA). These data were collected using an IBM compatible PC and a National Instruments interface card (model 371).

The fetal bladder was drained for at least 45 min before the experiment. Urine volumes were measured half-hourly. Maternal and fetal blood samples were taken at 45 min (3.5 ml) and 75 min (1.5 ml). After 90 min, the umbilical occluder was rapidly inflated with a known quantity of saline (3.5-4 ml). The success of occlusion was demonstrated by the sharp increase in fetal mean arterial pressure (MAP) and decrease in fetal heart rate (FHR) (Bennet et al. 1999), and by hypoxaemia, hypercapnia and acidaemia in a fetal blood sample (0.6 ml) taken 5 min into occlusion. Another fetal blood gas was taken at 25 min (0.6 ml). At 30 min, if the MAP was less than 12 mm Hg, the occluder was deflated (n= 3). If this MAP was not reached, the occlusion period was extended until such time as MAP did attain this level, or 40 min was reached (n= 2). However, if MAP reached ≈7 mm Hg, the occluder was deflated regardless of the time (n= 1; time 29 min). A 4 h recovery period followed. Maternal and fetal blood samples were taken at 105 min (2 ml) and 225 min (3 ml) post-occlusion. In the control group, the occluder was not inflated. Otherwise, the experiment was identical.

On days 2 and 3, fetal urine was collected for 2 × 30 min experimental periods and maternal and fetal blood samples (2 ml) were taken at 45 min. On day 4 the same variables as in day 1 were measured over 90 min (3 × 30 min periods). Blood samples were taken at 45 min (2 ml) and 75 min (4.5 ml).


At the end of day 4, the ewe and her fetus were killed by i.v. pentobarbitone sodium (5 g, Lethabarb, Virbac (Australia), NSW). The whole uterus was carefully removed so no fluid was lost. Fetal weight, crown-rump length, abdominal girth and tail length were measured. The fetus was photographed. To quantify the extent of the skin oedema, calipers were used to measure the thickness of the ear and skin folds at the shoulder and rump regions. The amniotic/allantoic fluid volume was measured.

Before any organs were extracted, pleural and peritoneal fluids were collected. Specific organs and tissues, i.e. kidneys, heart, adrenals, liver, lungs, spleen, brain, gut, tail, cotyledons and the membranes and umbilical cord, were weighed and a slice of the left kidney was fixed in buffered formalin and stained with haematoxylin and eosin for histological examination. If the fetus was male, the scrotum was also weighed. Tissue samples were snap-frozen in liquid nitrogen before being stored at −80°C for later analysis.

Biochemical analysis

Fetal and maternal blood gases and pH were measured at 37°C using a blood gas analyser (ABL 700 Series, Radiometer Pacific Pty Ltd) and corrected to 39.5°C. Plasma glucose and lactate levels, as well as concentrations of potassium and sodium, were also measured using this analyser.

Urine and plasma osmolalities were determined using a Fiske One-Ten Osmometer (Fiske Associates, MA, USA). Sodium and potassium concentrations in urine were measured using a FLM3 Flame Photometer (Radiometer Pacific Pty Ltd). The level of creatinine in the plasma and urine was measured by Laverty Pathology (Mayne Health, Sydney, Australia). Protein concentrations in plasma and urine were determined by a Lowry protein assay (Lowry et al. 1951), using a UV/visible spectrometer (Unicam 5600 series, ATI, Cambridge, UK).

Analysis of data

Fetal blood pressure was corrected for intra-amniotic pressure. Using RBF and MAP expressed as a percentage of their control values (RBFc, MAPc), renal vascular resistance relative to control (RVR) was calculated using the following formula:

display math

On day 1, data from the three 30 min control periods were averaged to obtain single values. Following this, data were analysed for each 30 min. On each of days 2, 3 and 4, 30 min periods were averaged to form one value. For most renal variables, results are presented as percentage of control values (% control). All results are expressed as means ± standard error of the mean (s.e.m.). Unless stated, n= 6 for the occlusion group and n= 5 for the sham group. Within each group, means were compared using SPSS (SPSS/PC; SPSS Inc., Chicago, IL, USA) by analysis of variance (ANOVA) for repeated measures and post-mortem data were analysed by Student's unpaired t test. A Student-Newman-Keuls test was used to determine the periods that were different when ANOVA reached significance.


Fetal cardiovascular response (Figs 1 and 2).

Figure 1.

Fetal blood pressure and heart rate during and after occlusion

A, fetal mean arterial pressure (MAP) for day 1, including the 30 min occlusion period (35 min for two animals), and day 4. a Different from control and all other occlusion periods, P < 0.05; b different from all occlusion periods, P < 0.001; c different from control, 5, 10 and 15 min, P < 0.05; d different from control, P < 0.05. There was no change in the sham group. B, fetal heart rate (FHR) for day 1, including the 30 min occlusion period (35 min for two animals), and day 4. a Different from control and all other occlusion periods except 10 min, P < 0.001; b different from control and all other occlusion periods except 5 and 15 min, P < 0.025; c different from control and all other occlusion periods except 10 and 20 min P < 0.025; d different from control and all other occlusion periods except 15 and 25 min, P < 0.025; e different from control and all other occlusion periods except 20 min, P < 0.01; f different from control and all other occlusion periods except 35 min, P < 0.05; g different from control and all other occlusion periods except 30 min, P < 0.05; h different from control in sham fetuses, P < 0.05. Sham (•, n= 5), occlusion (▴, n= 6). Values expressed as mean +s.e.m. Occlusion time periods are 5 min averages. Day 4 (D4) compared to control period only.

Figure 2.

Fetal renal blood flow (RBF) normalised to control for fetuses subjected to 30 min of umbilical cord occlusion

* Different to all periods, P < 0.001. Values expressed as mean +s.e.m. (n= 5).

In the control period, there were no differences between the groups in fetal arterial systolic, diastolic and mean pressures as well as heart rate (FHR). When the umbilical cord was occluded, there was a short immediate increase in fetal mean arterial pressure (MAP, Fig. 1A), followed by a progressive decrease to 11.9 ± 1.4 mmHg (P < 0.001, lowest MAP = 7.4 mmHg). Heart rate declined progressively to 61 ± 3 beats min−1 (lowest HR = 40 beats min−1) at the end of occlusion (P < 0.001, Fig. 1B). MAP and FHR did not change in the sham group. After deflation of the occluder, there was a small rebound hypertension in the occlusion group to 36 ± 2 mmHg at 30 min post-occlusion, (not shown), then blood pressure returned to control values. FHR returned to control values by 30 min. Only small changes in heart rate occurred during day 1 in the sham group. However, in both groups, blood pressure was higher on day 4 than during control.

During occlusion, renal blood flow (RBF, Fig. 2) decreased to 25 ± 5 % of control (P < 0.001). It was restored to control levels within 30 min of cuff deflation. Unfortunately, only one sham animal had a working flow probe. In this animal renal blood flow remained relatively stable throughout the day (data not shown). RBF was not measured on subsequent experimental days.

The relative renal vascular resistance (RVR), in the occlusion group, increased in the occlusion period to be 370 ± 91 % of control (P < 0.001, Fig. 3) and remained above control for the rest of the day (P < 0.05). RVR was not calculated for days 2, 3 and 4. In the one sham fetus with a working flow probe, RVR stayed relatively stable during day 1 (data not shown).

Figure 3.

Renal vascular resistance relative to control on day 1 in fetuses subjected to 30 min umbilical cord occlusion (n= 5)

* Different from all periods, P < 0.001; ‡ different from control period, P < 0.05. Values expressed as mean +s.e.m.

Fetal blood gases, pH and electrolyte concentrations

There were no differences between the groups during the control period (Table 1). When the umbilical cord was occluded, fetal arterial PO2 dropped to 6.2 ± 0.9 mmHg (P < 0.001) and PCO2 had increased to 158 ± 5 mmHg (P < 0.001) by 25 min. Arterial pH decreased to 6.72 ± 0.02 (P < 0.001). All three returned to control values by 2 h of recovery. Plasma bicarbonate levels decreased in the occlusion period, were at normal at 2 h, but were elevated at 4 h when compared to 2 h (P < 0.01). In the occlusion period there was a considerable rise in potassium concentration (P < 0.001) and a small, but significant rise in sodium levels (P < 0.05). Consequently, the plasma sodium/potassium ratio decreased during occlusion and remained lower than control for the rest of the day. Fetal haematocrit, plasma osmolality and plasma protein concentration remained stable over day 1. Minimal changes occurred in the sham group over the entire experimental period (Table 1).

Table 1. Blood gas values and plasma composition for all experimental days
   Day 1 Day 2Day 3Day 4
  1. Values expressed as mean ±s.e.m. Occlusion samples were taken after 25 min of complete cord occlusion. Sham (S) n= 5; occlusion (O) n= 6 except day 4, n= 5. Comparisons were made within day 1. Days 2, 3 and 4 were compared with control only. |Different from control, P < 0.05, ‖P < 0.01; *different from all other periods on day 1, P < 0.05, **P < 0.001; ¶different from 2 h, P < 0.05, ¶¶P < 0.01; ‡different from day 2, P < 0.05; §different from day 3, P < 0.05; °different from day 4, P < 0.05.

PO2 (mmHg)S21.8 ± 1.221.4 ± 1.220.0 ± 2.021.4 ± 0.920.9 ± 1.522.4 ± 1.523.5 ± 1.7
 O22.0 ± 0.76.2 ± 0.9**21.2 ± 0.620.4 ± 1.123.7 ± 1.124.7 ± 1.0|25.1 ± 1.11
PCO2 (mmHg)S51.0 ± 1.150.3 ± 0.651.0 ± 1.451.2 ± 1.152.7 ± 0.7|§°51.3 ± 0.949.5 ± 0.4|‡§
 O51.6 ± 1.3157.5 ± 4.7**50.7 ± 1.351.4 ± 1.250.9 ± 1.249.0 ± 1.5|48.9 ± 1.4|
pHS7.36 ± 0.067.35 ± 0.017.35 ± 0.017.34 ± 0.017.35 ± 0.017.35 ± 0.007.34 ± 0.00
 O7.36 ± 0.016.72 ± 0.02**7.35 ± 0.027.38 ± 0.027.37 ± 0.017.36 ± 0.017.36 ± 0.01
HCO3(mmol l−1)S27.5 ± 0.726.7 ± 0.627.0 ± 0.726.7 ± 0.727.9 ± 0.627.3 ± 0.525.7 ± 0.4|‡§
 O27.9 ± 0.518.8 ± 0.6**26.7 ± 1.329.4 ± 1.2¶¶28.1 ± 0.726.9 ± 0.8‡26.9 ± 0.9‡
Haematocrit (%)S24 ± 123 ± 124 ± 223 ± 223 ± 223 ± 125 ± 1
 O27 ± 228 ± 227 ± 227 ± 226 ± 226 ± 227 ± 2
Plasma protein (mg ml−1)S24.5 ± 1.126.3 ± 1.523.7 ± 0.727.0 ± 1.925.6 ± 1.0
 O24.9 ± 0.724.9 ± 0.922.8 ± 0.723.5 ± 1.125.3 ± 1.4
Osmolality (mosm kg−1)S293 ± 2296 ± 1299 ± 2302 ± 2302 ± 2294 ± 2
 O294 ± 1296 ± 2297 ± 2300 ± 3305 ± 2|‡°293 ± 1
Na+ (mmol l−1)S137 ± 1138 ± 0.5138 ± 1138 ± 1138 ± 0.5137 ± 0.2137 ± 0.2
 O138 ± 0.2139 ± 1*137 ± 0.4138 ± 1¶136 ± 1|136 ± 1|136 ± 0.3|
K+ (mmol l−1)S3.02 ± 0.152.86 ± 0.15|¶2.98 ± 0.162.92 ± 0.142.96 ± 0.193.12 ± 0.212.92 ± 0.20
 O3.13 ± 0.085.65 ± 0.14**3.43 ± 0.09‖3.27 ± 0.093.45 ± 0.06|3.47 ± 0.09|3.30 ± 0.05|‡§
Na+/K+S46 ± 249 ± 2|47 ± 248 ± 247 ± 345 ± 348 ± 3
 O44 ± 125 ± 0.8*40 ± 1*42 ± 1*39 ± 1|39 ± 1|41 ± 1|

In the occlusion group, PO2 was higher and PCO2 was lower on days 3 and 4 than during control. Arterial pH was unchanged, but plasma bicarbonate was lower on days 3 and 4 than day 2. Plasma osmolality on day 3 was increased with respect to all other days (P < 0.05). Plasma sodium levels were lower on days 2, 3 and 4 than control (P < 0.05) while potassium levels remained higher. As a result of these changes, the sodium/potassium ratio was reduced on days 2-4 (P < 0.005).

Plasma glucose and lactate levels (Fig. 4)

Figure 4.

Plasma concentrations of glucose (A) and lactate (B)

** Different from all periods, P < 0.001, different from control, *P < 0.05; ‡ different from control (Con) and occlusion (Occl), P < 0.05. Sham (open bars) n= 5; occlusion (filled bars) n= 6 except at 4 h, when n= 5. Values expressed as mean +s.e.m.

In the occluded fetuses, plasma glucose levels dropped during occlusion (from 1.4 ± 0.1 mmol l−1 to 0.5 ± 0.1 mmol l−1, P < 0.001) but returned to control values in the recovery period. Lactate levels increased dramatically in the occlusion period to 9.7 ± 1.0 mmol l−1 from a control value of 0.9 ± 0.1 mmol l−1 (P < 0.001). The concentration of lactate in the blood was still significantly increased at 2 h (P < 0.05) but returned to control values by 4 h. In sham fetuses, there was a slight, but statistically significant increase in lactate concentration at 2 and 4 h when compared to control and occlusion periods. While glucose levels remained at control levels on days 2-4 in the occlusion group, there was a slight elevation in lactate levels on days 2 and 4 (1.1 ± 0.1 mmol l−1 and 1.1 ±0.1 mmol l−1 respectively), but not day 3, when compared to control. There was no change in the sham group for days 2-4.

Fetal renal response

During the control period, there was considerable interfetus variation in a number of renal parameters (Table 2). Therefore, to examine the effects of umbilical cord occlusion, most renal variables have been expressed as a percentage of the control value (control = 100). However, it should be noted that urinary osmolality, for both the sham and occlusion groups, were in the normal ‘unstressed’ range during control (171 ± 11 mosm kg−1 and 165 ± 7 mosm kg−1 respectively; Wintour et al. 1985).

Table 2. Control values for sham and occlusion groups before standardisation
 Sham (n= 5)Occlusion (n= 6)
 Mean ±s.e.m.MinMaxMean ±s.e.m.MinMax
Urinary flow rate (ml min−1)0.14 ± ±
Urinary osmolality (mosmkg−1)171 ± 11139199165 ± 7145186
GFR (ml min−1)0.9 ± ±
Excretion rates
  Na+ (μmol min−1)8.7 ± ± 3.810.021.4
  K+ (μmol min−1)0.2 ± ±
  Osmoles (μosmol min−1)23.9 ± ± 3.333.753.9
Clearance rates
  Na+ (ml min−1)0.06 ± ±
  K+ (ml min−1)0.07 ± ±
  Osmoles (ml min−1)0.08 ± ±
  Free water (ml min−1)0.05 ± ±
Urinary protein concentration (mg ml−1)2.0 ± ±

There was a sharp decrease in urine flow rate during occlusion (to 26 ± 4 % of control, Fig. 5). Urine flow rate stayed low for the next 30 min then increased back to levels not significantly different from control. In contrast to the changes in the occlusion group, in the sham animals there was a slow decline in urine flow rate throughout the day, such that urine flow was reduced at 30, 150 and 240 min compared to control (P < 0.05). In both groups, urine flow was not different to control on days 2, 3 or 4 (Sham 184 ± 77, 168 ± 37, 73 ± 24 % control; Occlusion 66 ± 33, 107 ± 36, 113 ± 33 % control).

Figure 5.

Urinary flow rate as a percentage of control for midgestational fetal sheep undergoing complete umbilical occlusion (▴, n= 6) or for the sham group (•, n= 5)

* Different from control P < 0.005; ‡ different from control P < 0.05. Values expressed as mean ±s.e.m.

In both groups urinary osmolality tended to increase during day 1, although this rise only reached statistical significance in occluded fetuses (Fig. 6). Urinary osmolality was still elevated on day 2 (213 ± 13 mosm kg−1, P < 0.05) when compared to control (164 ± 7 mosm kg−1) and day 3 (176 ± 20 mosm kg−1) in the occluded group. On day 4, levels were slightly elevated (191 ± 15 mosm kg−1), although not different from control. The urinary osmolality remained at control levels on days 2-4 in sham fetuses.

Figure 6.

The urinary osmolality, for sham (•; n= 5) and occlusion (▴; n= 6) groups

* Different from control and occlusion periods, P < 0.05. Values expressed as mean +s.e.m.

In occluded fetuses, glomerular filtration rate (GFR, Fig. 7) was reduced at 2 and 4 h after the deflation of the occluder, but it had returned to control values on days 2, 3 and 4 (101 ± 24, 101 ± 23, 131 ± 28 % control). In sham animals, GFR remained constant over the four experimental days.

Figure 7.

Glomerular filtration rate (GFR) as a percentage of control for fetuses before and after occlusion

Sham (open bars) n= 5, occlusion (filled bars) n= 6 except at 4 h when n= 5. * Different from control, P < 0.05. Values expressed as mean +s.e.m.

Sodium and osmolar excretion rates decreased markedly during occlusion, to 25 ± 4 and 26 ± 4 % of control respectively (P < 0.05) and were still low at 2 h of recovery (47 ± 12 and 53 ± 12 % of control, P < 0.05). Potassium excretion also decreased during occlusion to 28 ± 4 % of control (P < 0.05) but was similar to control by 2 h (95 ± 21 % control). These excretion rates did not change in the sham group. In the occlusion group the clearances of sodium and potassium decreased during occlusion compared with control (P < 0.005, P < 0.05 respectively, Table 3) and were still depressed at 2 h (P < 0.05). While sodium clearance returned to control values, the clearance of potassium remained low (P < 0.05). Plasma osmolality was not measured in the occlusion period, so the clearance of osmoles and of free water could not be determined. However, at 2 h, occluded fetuses had decreased osmolar clearance when compared to control (P < 0.05), returning to normal by 4 h. Free water clearance was decreased at 2 h and 4 h for both the sham and the occlusion groups. When comparing days 2, 3 and 4 with control, there was no difference in either group.

Table 3. The clearance (C) of Na+, K+, osmoles and free water for all days as a percentage of control for sham (S) and occlusion (O) groups
  Day 1Day 2Day 3Day 4
% Control ControlOcclusion2h4h   
  1. Values expressed as mean ±s.e.m. Comparisons done within day 1 while days 2, 3 and 4 were compared only with control. §Different from control, P < 0.05, §§P < 0.005; ‡different from 4h P < 0.05; *different from all periods on day 1 P < 0.05. Sham n= 5; occlusion n= 6, except for CNa and CK on days 2 and 3, n= 5 and day 4, n= 4.

CNaS10087 ± 2158 ± 1750 ± 15§226 ± 105170 ± 3298 ± 34
 O10026 ± 4‡§§49 ± 12§77 ± 1890 ± 47143 ± 73173 ± 60
C K S100114 ± 2881 ± 2284 ± 25137 ± 71130 ± 2780 ± 32
 O10015 ± 2*86 ± 20§75 ± 12§144 ± 61166 ± 77185 ± 59
C Osm S10071 ± 2263 ± 19203 ± 100156 ± 2786 ± 30
 O10054 ± 12§75 ± 1579 ± 37115 ± 57134 ± 43
C H2O S10049 ± 13§39 ± 12§221 ± 106148 ± 4254 ± 22
 O10050 ± 14§51 ± 14§50 ± 3692 ± 2790 ± 24

Because of the fall in GFR, the filtered load of sodium was reduced in the occluded fetuses at 2 and 4 h after occlusion (control 254 ± 58 μmol min−1; 2 h 130 ± 30 μmol min−1; 4 h 113 ± 30 μmol min−1; P < 0.05). The reabsorption rate for sodium also fell (control 238 ± 57 μmol min−1; 2 h 122 ± 29 μmol min−1; 4 h 101 ± 27 μmol min−1; P < 0.05) and hence the fractional reabsorption of sodium was unchanged (control 93 ± 1 %; 2 h 93 ± 2 %; 4 h 89 ± 2 %).

In both groups, the concentration of protein in the urine did not differ from control values at any experimental time (Table 2).

Post-mortem data

The post-mortem data includes an additional fetus in the occluded group in which renal function was not studied (99 days gestation, 4 days post-occlusion), and three unoperated twins in the sham group. Fetal age at post-mortem was the same in both groups (sham 95 ± 1 days; occlusion 96 ± 1 days).

All fetuses in the occlusion group showed visible signs of hydrops (Fig. 8). This was verified by the higher fetal body weight (P≤ 0.05) and abdominal girth in the occlusion group (P < 0.01; Table 4). Nose-rump length (NRL), tail length and most organ and tissue weights were the same in both groups, however, scrotal weight as well as tail weight were greater in occluded fetuses (P < 0.05, Table 4).

Figure 8.

Photograph of an occluded fetus (1270 g, bottom) and its unoperated twin (930 g, top) aged 99 days

Table 4. Organ and tissue weights at post mortem
 Control n Occlusion n
  1. All values expressed as mean ±s.e.m.*P < 0.05; **P < 0.01; ***P < 0.001

Body Weight (g)808 ± 4481020 ± 82*7
Organ Weights (g)
  Adrenal (total)0.18 ± 0.0280.22 ± 0.027
  Brain17.2 ± 0.8817.1 ± 0.96
  Cotyledons388 ± 428435 ± 507
  GIT22.6 ± 2.8821.1 ± 2.87
  Heart (total)6.8 ± 0.386.5 ± 0.57
  Kidney (total)9.0 ± 0.788.6 ± 0.46
  Liver54.9 ± 3.3857.2 ± 3.77
  Lung30.3 ± 2.2826.7 ± 1.57
  Scrotum5.6 ± 1.3511.9 ± 0.1*2
  Spleen1.6 ± 0.181.4 ± 1.27
  Tail3.0 ± 0.384.2 ± 0.5*7
  Umbilical cord and167 ± 258228 ± 327
Lengths (cm)
  Nose-rump37.2 ± 0.9838.9 ± 0.87
  Girth22.2 ± 0.6825.2 ± 1.0*7
  Tail8.3 ± 0.488.3 ± 0.47
Fluid Volume (ml)
  Amniotic/allantoic320 ± 694676 ± 124*5
  Peritoneal5.9 ± 1.0820.1 ± 2.5***6
  Pleural2.5 ± 0.389.6 ± 3.5*7
Skin Thickness (mm)
  Shoulder2 ± 084 ± 1*6
  Rump3 ± 187 ± 1**6
  Ear1 ± 171 ± 05

The fluid volumes measured at post-mortem (amniotic/allantoic, peritoneal and pleural) were higher in the occlusion group (Table 4). Also, the thickness of the skin at the shoulder and the rump regions was higher (P < 0.01).

Histological examination of the left kidney revealed no sign of inflammation, leukocyte infiltration or necrosis 72 h after occlusion. The glomeruli and blood vessels were normal in all fetuses. Small clusters of apoptotic cells were seen at focal sites along the proximal tubule in most fetuses of the occlusion group (Fig. 9). Sham fetuses, which included two of the unoperated twins, showed no apoptotic clusters. In one sham fetus the doppler flow probe obstructed the renal artery and there was coagulative necrosis of the renal cortex.

Figure 9.

Histology slides of the left kidney of a sham fetus (A) and occluded fetus (B)

Note the apoptotic cells in B, arrows, which are shown at higher magnification in C. G, glomerulus; P, proximal tubule; D, distal tubule; arrow, apoptotic cell.


The hypothesis that prompted these experiments was that the hydrops (fetal oedema) seen by others following umbilical cord occlusion in midgestation, was due to the combination of activation of the renin angiotensin system (RAS) and damage to the fetal kidneys. Renal damage was anticipated because during an asphyxial episode, the combined ventricular output of the fetus is shunted away from organs like the kidney, to the brain, heart and adrenals (Kojima et al. 1985; Ball et al. 1994; Luciano et al. 1998; Ikeda et al. 2000). This decrease in renal blood flow (RBF), in conjunction with the low PO2 associated with asphyxia, might cause ischaemic damage to the kidney, especially the renal tubules. Activation of the RAS increases the transfer of salt and water across the placenta to the fetus (Faber & Anderson, 1997; Gibson & Lumbers, 1999). Normally this excess fluid would be excreted by the fetal kidneys and hence polyhydramnios would develop (Anderson & Faber, 1989), but if the fetus was oliguric because of acute renal failure, then the excess fluid would accumulate in the fetal body. It was also anticipated that when kidney perfusion returned there might be polyuria i.e. the tubules would be unable to reabsorb as much fluid due to necrosis.

In this study, consistent with the findings of Lumbers et al. (2001), the midgestational fetal sheep developed gross hydrops when subjected to a 30 min cord occlusion (Fig. 8). Thus it is likely that the RAS was activated. However, it is probable that impaired renal function as a result of asphyxia was not involved in the pathogenesis of this form of hydrops. Apart from the small clusters of apoptotic cells, which indicate that the kidney was only slightly affected by the asphyxial episode, there were no other changes in the histological appearance of the kidney 72 h after umbilical cord occlusion (Fig. 9). The absence of an inflammatory response, i.e. a lack of leukocyte infiltration, suggests there was no ischaemic damage to the kidney tubules. This finding is in stark contrast to that of Ikeda et al. (2000) who subjected near term fetuses to partial cord occlusion for approximately 60 min or until arterial pH fell below 6.9. Tubular necrosis occurred in all the asphyxiated cases. Since the severity of the insult was comparable between the studies (pH reached 6.72 ± 0.02 in our study), the lesser degree of renal damage was probably because our fetuses were younger. In general younger fetuses cope better with asphyxia than near term fetuses. For instance, in older fetuses, complete cord occlusion is terminal at 10-12 min (Mallard et al. 1994), whereas most midgestational fetuses can withstand 30 min.

The minimal damage sustained by the kidney may explain why most renal parameters returned to control values by the end of day 1. Although urine output decreased during the occlusion period and remained low for the next 30 min, values were then similar to sham fetuses (Fig. 5). At 84-91 days gestation, fetuses produce and release arginine vasopressin (AVP), or antidiuretic hormone (ADH), in response to hypoxia (Iwamoto et al. 1989). The lack of antidiuretic response seen in our study suggests that although AVP is produced at this age, the fetal kidney may be unresponsive to it. This is probably not due to a lack of receptor development and/or maturation, as the AVP V2 receptor is present and active at 90 days gestation (Strandhoy et al. 1992). However it is likely that the levels of aquaporin-2 (AQP2, the AVP-regulated channel) are low. AQP2 mRNA is detectable at only low levels at 75 days (Butkus et al. 1999). The increasing sensitivity to AVP from 100-140 days is largely due to increasing AQP2 gene expression over this period. Therefore, if at 91-95 days gestation, AQP2 is not abundant, one would expect the effects of AVP in the present study to be minimal.

The urine flow rates on days 2, 3 and 4 showed no significant change relative to control i.e. the polyuria that we had hypothesised to occur after renal perfusion returned did not happen. This indicates that the renal tubules retained their reabsorptive capacity and therefore these fetuses did not lose excess fluid in their urine. The lack of an increase in protein in the urine also suggests the glomeruli were intact, in addition to the tubules.

In the occlusion period, the relative renal vascular resistance (RVR) was nearly four times higher than in the control period (Fig. 3). Glomerular filtration rate (GFR) is likely to have been substantially decreased due to this intense vasoconstriction and the drop in urine flow rate supports this. While renal blood flow (RBF, Fig. 2) values in the 4 h after occlusion were not different from control, they never actually reached 100 %. This was due to a sustained increase in RVR. Larger increases in afferent than efferent arteriolar resistance are likely to have occurred because there was also a sustained decrease in GFR at 2 h and 4 h. Over the next 3 days, GFR returned to control values, probably accompanied by a restoration of normal RVR.

The rates of excretion and clearance for sodium and potassium were decreased in the occlusion period, which is consistent with the depression of GFR and urine flow rate at this time. After 2 h of recovery, although urine flow rate was not different from control, excretions and clearances of sodium and osmoles were still reduced. In contrast, at 2 h, the excretion of potassium was normal though its clearance was reduced. This was possibly due to the dramatic increase in plasma concentration of this ion, which is discussed below.

It was interesting that hydrops occurred in the absence of significant renal damage and with the fetal kidneys producing urine at normal rates. This implies that hydrops occurred because midgestational fetuses may have a limited capacity to increase their urine flow rate in response to the increased influx of fluid caused by activation of the RAS. Unfortunately, there are few data on renal function at this age. Although Moritz et al. (2000) demonstrated a diuresis in response to angiotensin II infusion in fetuses at 75-85 days, this was possibly a pressure natriuresis since the dose of angiotensin II also caused a rise in blood pressure. This means that we do not know if fetuses of this age can adequately excrete a fluid load in the absence of a rise in blood pressure.

Although insufficient to prevent the formation of hydrops, it is likely that there was a small increase in urine flow in occluded fetuses during days 2-4 since the volume of amniotic/allantoic fluid doubled. For the mean rise in amniotic/allantoic fluid observed by 72 h (≈350 ml, Table 4), an increase in urinary output of only 0.082 ml min−1 would be required. This would be too small to detect, especially considering the variability between fetuses in urine flow rate, and the fact that urine was only collected for a short period each day. However, there are other factors that could have contributed to the increase in amniotic/allantoic fluid volume. These include a decrease in fetal swallowing, an increase in lung liquid production and changes in intramembranous and transmembranous flows. We did not investigate these, however, there have been studies looking at the interactions of hypoxia or asphyxia and these variables. Brace et al. (1994) examined hypoxia and fetal swallowing and found that after the hypoxic stimulus was removed, swallowing returned to normal. Hooper et al. (1988) studied lung liquid secretion and found that it followed a similar pattern: after a decrease during asphyxia, secretion returned to control values in the recovery period. Thus while alterations in swallowing and lung liquid secretion may have played a role in the increase in amniotic/allantoic volume, it seems as though it would have been small.

Another possible explanation for the development of hydrops after an asphyxial insult, other than a limited ability of the fetal kidney to excrete a salt and water load at this stage of gestation, is heart failure. In utero cardiac decompensation caused by supraventricular tachycardia, without detectable abnormalities in haematocrit, serum protein or osmolality can cause gross oedema and ascites (Stevens et al. 1982). Heart failure associated with elevated central venous pressure, caused hydrops in fetuses in which rapid, severe anaemia was induced (Blair et al. 1994). However, Lumbers et al. (2001) postulated that it was unlikely that 30 min of asphyxia in midgestation led to continued cardiac failure, as the central venous pressure was not elevated after asphyxia and the myocardium was histologically normal after 72 h of recovery.

While the renal responses to asphyxia differed from those expected, other parameters did not. The blood gas changes and cardiovascular responses to umbilical occlusion observed by us have been well described in other studies (Mallard et al. 1994; Keunen et al. 1997; Bennet et al. 1999).

Blood pressure had increased in all fetuses by day 4. This happened in both groups, so is likely to be due to developmental factors. It is well known that in the latter stages of gestation, blood pressure increases with an increasing gestational age, due to such developmental factors as increasing heart size, wall thickness, and peripheral vascular resistance (Harding & Bocking, 2001). Furthermore, gestation-dependent increases in blood pressure have been demonstrated in fetuses as young as 74-84 days (Moritz et al. 2000).

As in other asphyxial studies, the occluded fetuses developed a combined respiratory and metabolic acidosis. This acidosis explains the hyperkalaemia that was observed during the occlusion period. In acidosis, hydrogen ions move into cells, in exchange for potassium (Liebman & Edelmn, 1959). After deflation of the occluder, PCO2, bicarbonate levels and pH returned to normal allowing potassium ions to return into cells. By 4 h of recovery, potassium levels were normal. The rapid lowering of plasma potassium levels was very important, as only small increases may disrupt cardiac rhythm. In one occluded fetus, not included in this study, potassium levels rose to 6.3 mmol l−1 at 25 min occlusion. After deflation of the occluder, this fetus developed an abnormal cardiac rhythm and died 10 min later, despite administration of adrenaline.

Hydrops fetalis, not resulting from immunological causes, occurs in about 1 in 3000 pregnancies (Heinonen et al. 2000; Sohan et al. 2001). Although there are about 80 different causes of hydrops (Iskaros et al. 1997) a direct cause cannot be found in approximately 20 % of cases (Santolaya et al. 1992). We have demonstrated in this study that fetal asphyxia could account for some of these cases. Asphyxial events at midgestation could happen due to cord compression or reduced uterine blood flow, both of which can be transient. If blood flow is restored in time, the return to relative normality is possible. Such an event may go unnoticed, since we have preliminary data to indicate that hydrops can resolve by 130 days gestation in fetuses subjected to 30 min umbilical cord occlusion in midgestation.


This work was supported by a grant-in-aid from the National Heart Foundation (Australia). We are grateful to Dr Laura Bennet for her advice, and to Ms Pamela Bode for her assistance. We would also like to thank Professor Nicholas Hawkins for his assistance with histology.