Intrauterine asphyxia and the breakdown of physiologic circulatory compensation in fetal sheep

Intrauterine asphyxia and the breakdown of physiologic circulatory compensation in fetal sheep Barry S. Block, MD: Donald H. Schlafer, DVM, PhD: Richard A. Wentworth, PhD; Lois A. Kreitzer, MS: and Peter W. Nathanielsz, MD, PhDC

Chicago, Illinois, and Ithaca, New York In response to acute hypoxemia, the fetus invokes physiologic compensatory mechanisms that cause a preferential redistribution of the circulation to sustain the brain, heart, and adrenal gland and maintain blood flow to the placenta. These mechanisms are available for a limited time and eventually the fetus is no longer able to maintain preferential perfusion and decompensation occurs. To identify the relationship between hypoxemia with severe acidemia and the breakdown of circulatory compensation, we decreased uterine blood flow in 10 chronically instrumented pregnant sheep. We measured fetal blood gases and pH, arterial and central venous pressures, heart rate, combined ventricular output, and regional blood flow distribution during hypoxemia with severe acidemia and when a fixed-baseline sustained bradycardia (agonal) heart rate pattern developed. Hypoxemia with severe acidemia was characterized by markedly decreased blood flow to most organs; however, the preferential perfusion of the brain, heart, adrenal gland, and placenta was still present. An agonal heart rate pattern was characterized by complete cardiovascular collapse. This study demonstrates that circulatory compensation is present in fetal sheep affected by deficiency of oxygen delivery despite hypoxemia with severe acidemia. (AM J OBSTET GVNECOL 1990;162:1325-31.)

Key words: Pregnancy, obstetrics, hypoxia, microspheres Intrauterine fetal asphyxia is a leading cause of fetal death and newborn cardiovascular and respiratory depression. 1 Asphyxia is characterized by hypoxemia with acidemia commonly associated with deficient oxygen delivery to the fetus. Decreased oxygen delivery of maternal origin has been produced experimentally in pregnant ewes by reducing the maternal arterial oxygen content,2.3 restricting uterine blood flow:' 5 or a combination of both. 6 Combined Po. reduction and uterine blood flow restriction has the advantage that the chemoreceptor and/or reflex response to acute hypoxemia is enhanced when resting Po. is reduced. 7 In response to acute hypoxemia, the fetus invokes physiologic compensatory mechanisms that produce a redistribution of the circulation to sustain the brain, heart, and adrenal gland and maintain blood flow to the placenta." These mechanisms are available for a limited time, and eventually the fetus is no longer able

From the Section of Maternal-Fetal Medicine, Department of Obstetrics and Gynecology, the University of Chicago: and the Department of Pathologyb and the Laboratory for Pregnancy and Newborn Research, Department of Physiology,' the College of Veterinary Medicine, Cornell University. Supported in part by grant HD 21350 and The Harold Wetterberg Foundation. Received for publication August 7, 1989; revised November 20, 1989; accepted january 18, 1990. Reprint requests: Barry S. Block, MD, Department of Gynecology and Obstetrics, Lama Linda University, Lama Linda, CA 92350. 611119486

to maintain preferential perfusion and decompensation occurs. The purpose of this study was to identify the relationship between hypoxemia with severe acidemia and the breakdown of physiologic circulatory compensation.

Material and methods Ten mixed-breed Rambouillet-Columbia pregnant ewes of known gestational age were included in this study. Surgery was performed with halothane general anesthesia after a 24-hour fast. The maternal abdomen was prepared by an aseptic technique, and the uterus was exposed through a midline abdominal incision. A pelvic limb of the fetus was exteriorized through an incision made in the uterus. Polyvinyl catheters (41100inch inside diameter) were advanced into the fetal descending aorta through the femoral artery, into the inferior vena cava through the femoral vein, and, through a separate incision, into the ascending aorta through the fetal carotid artery and into the superior vena cava through the jugular vein. A polyvinyl catheter was placed in the amniotic cavity, and the incision was closed. The maternal neck was prepared by an aseptic technique and polyvinyl catheters were placed in the carotid artery, jugular vein, and trachea. A No. 6 Fogarty balloon-tip catheter (Edwards Laboratories, Santa Ana, Calif.) was placed into the descending aorta through the femoral artery, with the tip distal to the renal arteries. The ewe was given 250 mg ampicillin

1325

1326 Block at al.

HEART RATE (bpm)

May 1990 Am J Obstet Gynecol

I

:l-'

:r-

NITROGEN INFUSlDN

100

CAROllO ARTERY PRESSURE (mmHo)

I

•

1• •

1 .1111 ••••

NORIIOXIA

FLUSH

NITROGEN INFUSION

I HEART RATE (IIpm)

AORTIC DCCL.U8ION

~~~.Jrv--

:[-""'"-"\~

r-~r------------------

10 CAROTID ARTERY PRESSURE (mmHo)

HYPOXE.U

f-

1 min

-l

(ASPHYXIA)

AGONAL

Fig. 1. Fetal heart rate and carotid artery blood pressure during maternal nitrogen infusion and aortic occlusion. Upper two panels, Responses during hypoxemia. Lower two panels, Responses during occlusion show asphyxia (not performed in this fetus) and agonal heart rate pattern study periods.

and 900 mg chloramphenicol intravenously twice daily, and the fetus was given 50 mg chloramphenicol intravenously and intraamniotically twice daily. Blood gas values and pH from the fetal ascending aorta were measured with an automated analyzer (ABL2; Radiometer, Copenhagen) and corrected to 39° C. Percent saturation of hemoglobin was measured with a hemoximeter (OSM2; Radiometer). Arterial oxygen content was calculated as follows: Oxygen content (mil dl) = 1.34 (ml/ gm hemoglobin) X (hemoglobin concentration) (gm/ dl) x hemoglobin oxygen saturation (%/100). Maternal and fetal arterial blood pressures and amniotic fluid pressure were determined with strain-gauge transducers (P23Db; Statham/Gould, Oxnard, Calif.) and recorded continuously on a polygraph (Dynograph R611; Beckman Instruments, Inc., Fullerton, Calif.). Fetal arterial and venous blood pressures were corrected by subtracting amniotic fluid pressure. Fetal heart rate was derived from the arterial pressure pulse by a cardiotachometer (9857B input coupler; Beckman). Fetal combined ventricular output and regional blood flow distribution were measured according

to the radionuclide-labeled microsphere technique. s Briefly, one of five 15 IJ. radiolabeled microspheres (New England Nuclear, Boston) was selected at random from cerium 141, chromium 51, niobium 95, ruthenium 103, and tin 113. The microspheres were suspended in a solution containing 10% dextran and 0.05% polysorbate 80 (Tween 80). After ultrasonic dispersion, approximately 2 X 106 microspheres were injected during a 15-second period into the superior and inferior vena cava. Reference blood samples were withdrawn simultaneously from the ascending and descending aorta. Experimental protocol. No experiments were performed in the first 5 days after surgery. Gestational age at study ranged between 124 and 132 days. After 30 minutes of baseline measurements, the first regional blood flow determination was performed. Nitrogen,6 L/min, was infused into the maternal trachea9 for 20 minutes and the second regional blood flow determination was performed. The infusion of nitrogen was continued and the balloon catheter in the aorta was inflated to the extent that has previously been shown

Volume 162 Number 5

to virtually completely occlude uterine blood flow. '" We used a sling to help the hind limbs support maternal weight and were able to extend the time of occlusion until fetal decompensation occurred. The third regional blood flow determination was performed when a stable bradycardia developed in either of two distinct circumstances. Group I (hypoxemia with severe acidemia, n = 5) until hypertension, known to accompany uterine flow occlusion, was no longer maintained and mean arterial blood pressure returned to baseline values (Fig. 1). Group II (agonal heart rate pattern, n = 5) until a fixed-baseline sustained bradycardia (agonal pattern") developed (Fig. 1). The aortic occluder was then released and recordings continued to determine if the fetus would recover; however, no further regional blood flow measurements were performed. At the end of the experiment the ewe was killed and the uterus and its contents were removed, dissected, and weighed, and correct catheter placement was confirmed. The tissues were then carbonized and counted with an automated "{-counter and multiple-channel pulse height analyzer (Gamma Trac 2250; Searle Analytic, Elk Grove Village, 111.). The statistical method used to compare the mean values in the experimental periods for group I and group II fetuses was a two-factor analysis of variance for repeated measures and the Newman-Keuls test.'2 The level of significance chosen was p < 0.05. As expected, the mean values for the baseline and hypoxemia study periods were not significantly different between group I and group II fetuses. The data presented in Tables I and II for the baseline and hypoxemia study periods represent pooled values. The rate-pressure product, known to correlate with myocardial work in the fetus, ,:I was calculated as: peak systolic blood pressure times heart rate. Results A typical recording of fetal heart rate and arterial blood pressure response in a fetus with hypoxemia during maternal aortic occlusion is shown in Fig. 1. The response in this fetus demonstrates that decreased arterial P0 2 (fmm 22.7 to 12.6 mm Hg produced by maternal infusion of nitrogen) results in heart rate and blood pressure changes proportional to the duration of hypoxemia. For comparison with the agonal heart rate pattern regional blood flow determination that was performed, the point at which a hypoxemia with severe acidemia regional blood flow determination would have been performed (i.e., asphyxia) is also indicated. The cardiovascular, blood gas, and acid-base state measurements for all fetuses are summarized in Table I. Infusion of nitrogen directly into the maternal trachea (hypoxemia) decreased the P0 2 and oxygen con-

Asphyxia and fetal decompensation

1327

tent approximately 50% without significantly affecting the fetal pH or Pco 2 • Occlusion of the uterine artery (average time of occlusion: group I [hypoxemia with severe acidemia] = 6.1 ± 2.8 minutes: group II [agonal pattern] = 11.0 ± 4.5 minutes) was associated with a fetal P0 2 and oxygen content that was less than 25% of the baseline values. The fetal heart rate decreased in proportion to the degree of hypoxemia in both groups. The mean fetal arterial blood pressure was stable during both hvpoxemia alone and hvpoxemia with severe acidemia but decreased to 50% of the baseline value when an agonal heart rate pattern developed. The mean hemoglobin content of fetal blood increased approximately 25C)t, and the central venous pressure increased sixfold during study periods of both the hypoxemia with severe acidemia and agonal heart rate pattern. The fetal measurements of combined ventricular output and regional blood flow distribution are summarized in Table II. The fetal combined ventricular output and placental blood flow were stable during hypoxemia but decreased 60% during hypoxemia with severe acidemia and more than 90% when an agonal heart rate pattern developed. The blood flow to the brain, heart, and adrenal gland increased during both hypoxemia and hypoxemia with severe acidemia but was markedly decreased when an agonal heart rate pattern developed. The proportion of combined ventricular output perfusing the brain, heart, and adrenal gland was increased during hypoxemia with severe acidemia and then returned to baseline values when an agonal heart rate pattern developed. Blood flows to the carcass, kidney, and spleen were stable during hypoxemia but decreased markedly during both hypoxemia with severe acidemia and when an agonal heart rate pattern developed. The proportion of combined ventricular output perfusing the placenta was stable during both hypoxemia and hypoxemia with severe acidemia but decreased to 25% of the baseline values when an agonal heart rate pattern developed. The calculated rate-pressure product at rest was 9800 ± 1200 beats/min/mm Hg. In fetuses in group I (hypoxemia with severe acidemia) the rate-pressure pmduct was stable during both hypoxemia and aortic occlusion because the systolic blood pressure increase offset the heart rate decrease. After the balloon occluder was deflated and oxygen delivery to the uterus was restored, the rate-pressure pmduct increased (Fig. 2). In fetuses in group II (agonal pattern) the rate-pressure product was also stable during hypoxemia but decreased 50% during aortic occlusion. In addition there was a progressive decline in calculated rate-pressure product even after oxygen delivery to the uterus was restored, and none of the fetuses in group II recovered.

1328

Block et al.

May 1990 Am J Obstet Gynecol

Table I. Fetal heart rate, carotid blood pressure, and arterial blood gases and acid-base state during baseline, hypoxemia, hypoxemia with sever acidemia, and agonal study periods

=

Groups I and II (n Baseline

Heart rate (beats/min) Mean carotid presure (mm Hg) Central venous pressure (mm Hg) PO z (mm Hg) Oz content (mlldl) Pcoz (mm Hg) Hemoglobin (gm/dl) pH Base excess (mEq/L) Buffer base (mEq/L)

161 51 2.9 23.1 9.2 51.8 11.7 7.32 -0.6 47.1

± ± ± ± ± ± ± ± ± ±

I

13 7 1.4 3.8 3.1 2.4 1.3 0.04 2.0 1.6

10)

Hypoxemia

144 50 6.8 13.7 4.5 49.5 12.0 7.33 -2.0 44.7

± 18* ±

7

± 2.4 ± 2.3* ± 1.3* ± 2.6 ± 1.3 ± 0.06 ± 3.0 ± 3.2

Group I (n = 5): Hypoxemia with severe acidemia

105 43 18.0 4.0 2.9 90.5 14.0 7.00 -10.6 36.8

± 9*

± ± ± ± ±

± ± ±

±

7

1.9* 2.3* 0.4* 11* 1.4* 0.10* 4.3* 4.3*

Group II (n = 5):Agonal heart rate pattern

81 25 18.8 2.3 2.5 III

13.0 6.86 -20.2 27.0

± 14* ± 9* ± 4.3* ±0.6* ± 0.6* ± 27* ± 1.7* ± 0.13* ± 4.6* ± 4.9*

Values are mean ± 1 SD.

*p< 0.01 (hypoxemia, hypoxemia with severe acidemia, or agonal vs baseline).

Table II. Fetal combined ventricular output and its distribution and individual organ blood flows during baseline, hypoxemia, hypoxemia with acidemia, and agonal study periods

Hypoxemia

Group I (n = 5): Hypoxemia with severe acidemia

586 ± 104

644 ± 159

254 ± 73*

62 ± 37*

158 ± 46

198 ± 78

67 ± 35*

7 ± 7*

Groups I and II (n Baseline

Combined ventricular output (mil min/kg fetus) Umbilical blood flow (ml/min/kg fetus) Percent distribution of combined ventricular output Adrenal gland Brain Carcass Heart Kidney Placenta Spleen Organ blood flow (mil min/ 100 gm) Adrenal gland Brain Carcass Heart Kidney Spleen

I

=

0.08 2.8 45.5 2.0 1.8 33.4 1.4

± 0.05 ± 1.0 ± 5.6 ± 0.9 ± 0.6 ± 5.5 ± 0.6

0.16 4.5 42.1 4.8 2.1 36.7 0.8

314 158 39 252 134 353

± 166 ± 43

762 234 35 560 168 240

± ± ± ±

9 76 48 161

10)

± ± ± ± ± ± ±

0.10 1.6 6.8 2.7 0.7 5.9 0.6

± 314t ± 66t ± 11

± 209t

± 46 ± 145

0.28 6.8 34.0 11.5 2.3 28.6 0.4 545 185 12 597 75 35

± 0.13*

± 2.6t ± 12.7 ± 4.7* ± 1.8 ± 8.9 ± 0.3t ± ± ± ± ± ±

177t 74 5* 244t 75t 27t

Group II (n = 5): Agonal heart rate pattern

0.25 2.9 68.2 8.3 0.4 7.8 0.5 79 7 5 28 4 30

± 0.25t

± 2.1 ± 15.4t

± 5.3* ± 0.3t ± 5.8* ± 0.7t

± ± ± ± ± ±

52t 6* 3* 20t 5* 40t

Values are mean ± 1 SD.

*p < 0.01, (hypoxemia, hypoxemia with severe acidemia, or agonal vs baseline). tp < 0.05 (hypoxemia, hypoxemia with severe acidemia, or agonal vs baseline).

Comment

We determined fetal heart rate, combined ventricular output, regional blood flow distribution, and mean arterial blood pressure at rest, during acute hypoxemia, during hypoxemia with severe acidemia, and when an agonal heart rate pattern developed. Hypoxemia produced by maternal tracheal infusion of nitrogen decreased P0 2 and oxygen content by about 50% without causing a significant change in pH or Pco 2 • The fetal cardiovascular effects of hypoxemia included de-

creased heart rate and both increased blood flow and proportion of combined ventricular output perfusing the brain, heart, and adrenal gland. Fetal mean arterial blood pressure, combined ventricular output, and placental blood flow did not change significantly from baseline values during hypoxemia. Hypoxemia with severe acidemia produced by combined maternal hypoxemia and aortic occlusion produced more pronounced fetal cardiovascular effects than did hypoxemia alone. Hypoxemia with severe ac-

Asphyxia and fetal decompensation

Volume 162 Number 5

200

1329

NITROGEN INFUSION

I(..)

~

o~

OW a:z a..:::i

150

w

wm

a:< ~1Il

mil..

fnO

100

a:(ft.

-.-

GROUP I

W I< a:

-0--

GROUP II

a..~

50

O~---'----------r---------~--------~----BASELINE HYPOXEMIA OCCLUSION RECOVERY

Fig. 2. Calculated rate-pressure product from group I and group II fetuses at rest, during hypoxemia and aortic occlusion, and after release of aortic occlusion. The symbols represent mean ± I SD for five fetuses in each group. Asterisk, p < 0.50; double asterisks, p < 0.01 (experimental versus baseline study periods).

idemia markedly decreased fetal heart rate, combined ventricular output and blood flow to the carcass, kidney, spleen, and placenta, and further increased the proportion of combined ventricular output perfusing the brain, heart, and adrenal gland. An agonal heart rate pattern produced by prolonged aortic occlusion was associated with complete cardiovascular collapse. The fetal heart rate, mean arterial blood pressure, combined ventricular output, and blood flow to all organs decreased dramatically. There are numerous studies of fetal asphyxia; however, common use of the term asphyxia to denote any untoward clinical condition is known to have caused confusion. 14 Asphyxia is associated with hypoxemia and acidosis"; however, the degree of acidosis may still vary. To avoid confusion, we have used the term hypoxemia with severe acidemia. We produced hypoxemia with severe acidemia by modification of a technique described previously. Harris et al. 6 produced hypoxemia by flowing a low oxygen gas mixture to the ewe and then reduced uterine blood flow by aortic occlusion for a short period (113 minute). Selection of the end point for the study periods of hypoxemia with severe acidemia and agonal heart rate pattern was based on the fetal heart rate and blood pressure response to uterine blood flow reduction. In this manner the degree of stress was fixed and the time of occlusion was permitted to vary. Our method is similar to that described Cohn et al.,2 in which the hypoxemic level of Po, was preselected and the duration of low oxygen-gas mixture infusion to the ewe was permitted to vary. The fetal cardiovascular responses to acute hypoxemia in our study are in agreement with the results published previously.2.4.16 In all four studies, fetal heart rate decreased, blood flow to the brain, heart, and ad-

renal gland increased, and combined ventricular output and placental blood How did not change significantly. The fetal cardiovascular responses to hypoxemia with severe acidemia in our study are in agreement with the results published previously. 1. , In all three studies, fetal heart rate decreased, the proportion of combined ventricular output perfusing the brain, heart, and adrenal gland increased, and both the combined ventricular output and placental blood flow decreased. Although the techniques used to produce hypoxemia with severe acidemia in the three studies were different, the results and degree of asphyxia among the studies were similar. Yaffe et aV produced partial (25% of baseline values) uterine blood flow occlusion for 15 minutes, and their fetuses had a P0 2 of 11 mm Hg, pH 7.14, and Pco 2 of64 mm Hg.Jensen et aI.' produced total uterine blood flow occlusion for 4 minutes, and their fetuses had a P0 2 of about 3 mm Hg, pH 7.03 and Pco 2 of 86 mm Hg. We produced near-total uterine blood flow occlusion for 6.1 ± 2.8 minutes, and our fetuses had a Po, of 2.3 mm Hg, pH 7.00, and Pco 2 of 90.5 mm Hg. In contrast to the studies by Yaffe et aLl and Jensen et al.,' the results of our study suggest that the breakdown of physiologic circulatory compensation was not associated with hypoxemia and severe acidemia. In our study the presence of physiologic circulatory compensation during hypoxemia with severe acidemia was supported by the increased proportion of combined ventricular output perfusing the brain, heart, and adrenal gland. Yaffe et al. 1 speculated that vascular resistance of the brain, heart, and adrenal gland increased during hypoxemia with severe acidemia based on the assumption that central venous pressure does not change when

1330 Block et al.

acidemia develops. This assumption is based on the findings of Edelstone et al.,17 who found that inferior vena caval blood pressure did not change during umbilical blood flow reduction despite a statistically significant, although clinically mild, decrease in pH from 7.39 to 7.27. In the present study, measured central venous pressure did increase when severe acidemia (pH - 7.0) developed. Unfortunately, calculated vascular resistance is dependent on heart rate because of the nature of pulsatile flow ls ; therefore we were unable to provide comparisons of calculated vascular resistance based on the mean blood flow measurements made between conditions of markedly different heart rates. Jensen et al.' suggested that decentralization of the circulation was seen in asphyctic fetuses destined to die (nonsurvivors) because the total brain blood flow did not change significantly. We also noted no significant increase in total brain blood flow during hypoxemia with severe acidemia; however, in both our study and the study by Jensen et al. the proportion of combined ventricular output to the heart, adrenal gland, and brain increased during hypoxemia with severe acidemia. We propose an alternative explanation that the physiologic compensatory mechanisms were still active during hypoxemia with severe acidemia despite a decrease in combined ventricular output. In addition we were unable to confirm the finding of Jensen et al. that failure of adrenal blood flow to increase predicted subsequent fetal death. In our study the adrenal blood flow as a proportion of combined ventricular output was maintained even when cardiovascular collapse occurred. Noteworthy in the study by Jensen et al. was that three of the four non survivors were acute preparations (2 hours after surgery) and the other was spontaneously hypoxemic. Because surgery is known to adversely affect the fetus,19 comparisons with findings obtained in the postoperative period are difficult to interpret and should probably be avoided. The fetal cardiovascular status during an agonal heart rate pattern has not been reported previously, although the presence of a fixed-baseline sustained bradycardia agonal pattern preceding fetal death has been demonstrated in pregnancy in both humans ll ,2o,21 and sheep.22 During the agonal heart rate pattern study period, the fetal blood gases, pH, and heart rate measurements were more extreme than values obtained during hypoxemia with severe acidemia in human pregnancy,23 In addition, complete cardiovascular collapse was present and progressive deterioration and eventually fetal death occurred even after oxygen delivery to the uterus was restored, Clinically this may be a significant finding because it suggests that intrauterine resuscitation does not reverse an intrauterine condition characterized by an agonal heart rate pattern and indicates that operative delivery and llltensive neonatal care may be required.

May 1990 Am J Obstet Gynecol

This study demonstrates that the breakdown of fetal physiologic circulatory compensation is not associated with hypoxemia with severe acidemia alone. The presence of circulatory compensation during hypoxemia with severe acidemia was supported by the presence of preferential perfusion of the brain, heart, and adrenal gland as a proportion of combined ventricular output. We conclude that circulatory compensation is present in fetal sheep affected by deficiency of oxygen delivery despite hypoxemia with severe acidemia, We thank Tami Myers and Vivian Surman for their excellent technical assistance. REFERENCES 1. Gruenwald P, Stillbirth and early neonatal death. In: Butler NR, Alberman E, eds. Perinatal problems: second report of the 1958 British Perinatal Mortality Survey. Edinburgh: Churchill Livingstone, 1969:163-83. 2, Cohn HE, Sacks EJ. Heymann MA, Rudolph AM. Cardiovascular responses to hypoxemia and acidemia in fetal lambs. AMJ OBSTET GYNECOL 1974;120:817-24, 3, Peeters LL, Jones MD Jr, Sheldon RE, Meschia G, Battaglia FC, Makowski EL. Relationship between oxygenation and distribution of fetal cardiac output. Gynecol Invest 1976;7:49, 4. Yaffe H, Parer JT, Block BS, Llanos AJ. Cardiovascular responses to graded reductions of uterine blood flow in the sheep fetus, J Dev Physiol 1987;9:325-36. 5, Jensen A, Hohmann M, Kunzel W, Dynamic changes in organ blood flow and oxygen consumption during acute asphyxia in fetal sheep, J Dev Physiol 1987;9:543-59, 6, Harris JL, Krueger BS, Parer JT. Mechanisms of late decelerations of the fetal heart rate during hypoxia, AM J OBSTET GYNECOL 1982;144:491-6. 7, Rudolph AM. The fetal circulation and its response to stress, J Dev Physiol 1984;6: 11-9. 8, Heymann MA, Payne BD, Hoffman JE, Rudolph AM. Blood flow measurements with radionuclide-labeled particles, Prog Cardiovasc Dis 1977 ;20:55-79. 9. Gleed RD, Poore ER. FigueroaJP, Nathanielsz PW, Modification of maternal and fetal oxygenation with the use of tracheal gas infusion. AM J OBSTET GYNECOL 1986; 155:429-35. 10. Parer JT, Krueger TR, HarrisJL. Fetal oxygen consumption and mechanisms of heart rate response during artificially produced late decelerations of fetal heart rate in sheep. AMJ OBSTET GYNECOL 1980;136:478-82, 11. Cibils LA. Clinical significance of fetal heart rate patterns during labor. AMJ OBSTET GYNECOL 1977;129:833-44. 12. Winer BJ. Statistical principles and experimental design. New York: McGraw-Hill, 1971. 13. Fisher DJ, Heymann MA, Rudolph AM, Myocardial oxygen and carbohydrate consumption in fetal lambs in utero and in adult sheep, Am J Physiol 1980;238:H399-405. 14, Denson SE. Fetal asphyxia: its impact on the neonate: an approach to understanding and anticipating complications, In: Rathi M, ed, Current perinatology. New York: Springer-Verlag, 1988:25-37, 15. GranszJP, Heimler R, Asphyxia and gestational age. Obstet Gynecol 1983;62:175-9. 16. Reuss ML, Parer JT, Harris JL, Krueger TR. Hemodynamic effects of alpha-adrenergic blockade during hypoxia in fetal sheep. AM J OBSTET GYNECOL 1982; 142:410-5. 17, Edelstone DI, Rudolph AM, Heymann MA. Effects of hypoxemia and decreasing umbilical flow on liver and ductus blood flows in fetal lambs. Am J Physiol 1980; 238:H656-63. 18, Berman W, Goodlin RC, Heymann MA, Rudolph AM,

Volume 162 Number 5

Effects of pharmacologic agents on umbilical blood flow in fetal lambs in utero. Bioi Neonate 1978;33:225-35. 19. Rudolph AM, Heymann MA. Fetal and neonatal circulation and respiration. Ann Rev Physiol 1972;36: 187-207. 20. Tushuizen PBT, StootJEG, UbachsJMH. Fetal heart rate monitoring of the dying fetus. AM J OBSTET GVNECOL 1974; 120:922-31. 21. Gaziano EP, Freeman DW. Analysis of heart rate patterns preceding fetal death. Obstet Gynecol 1977;50:578-82.

Asphyxia and fetal decompensation

22. Raye JR, Killam AP, Battaglia Fe, Makowski EL, Meschia G. Uterine blood flow and 0, consumption following death in sheep. AM J OBSTE'!' GV;\;ECOL 1971; III :917-24. 23. Low JA, Pancham SR, Piercy WN, Worthington D, Karchmar J. Intrauterine fetal asphyxia: clinical characteristics, diagnosis, and significance in relation to pattern of development. AM.J OBSTET GV:-.JECOL 1977;129:857-72.

The development of hydrops fetalis in the ovine fetus after lymphatic ligation or lymphatic excision Robert L. Andres, MD, and Robert A. Brace, PhD La J olia, California Sixteen ovine fetuses underwent either ligation or excision of the left thoracic, left cervical, and left brachiocephalic lymphatic ducts. Our purpose was to test the hypothesis that interruption of lymphatic flow would lead to hydropic changes in the ovine fetus. Of the 11 animals in the group that underwent ligation, hydrops developed in 1. All five of the fetuses that underwent excision of these major lymphatic ducts were hydropic at the time of autopsy (3 to 7 days), with 62 to 502 ml of free fluid collected from the thoracic and abdominal cavities. The mean edema fluid total protein concentration in the hydropic fetuses was 2.6 gm/dl. This value was 71 % to 94% of that found in the plasma, suggesting that the fetus is capable of producing new plasma proteins at a high rate. The observation that lymphatic excision led to hydropic changes in the ovine fetus, whereas ligation did not consistently produce hydrops, suggests that fetal lymph vessels may be capable of very rapid regrowth over short distances. Thus lymphatic excision, but not ligation, produces an animal model for the study of hydrops fetalis. (AM J OeSTET GVNECOL 1990;162:1331-4.)

Key words: Hydrops fetalis, lymphatics, ovine fetus The cause of hydrops fetalis (fetal edema) often includes abnormalities in either the cardiovascular, pulmonary, hematologic, or hepatic systems. In addition, there are clinical conditions that affect the lymphatic system (e.g., cystic hygroma and pulmonary lymphangiectasia) that are associated with hydrops fetalis. I. 2 This suggests that abnormal lymphatic function in the fetus may be a cause of hydrops fetalis. Although the fetal lymphatic system has not been studied extensively, there are several observations that support the hypothesis that alterations in fetal lymph flow may lead to the development of hydrops. For example, in chronically From the Division of Perinatal Medicine, Department of Reproductive Medicine, University of California, San Diego. Supported by National Institutes of Health Grant HD 21269. Received for publication July 21,1989; revised December 28,1989; accepted January 18, 1990. Reprint requests: Robert A. Brace, PhD, Department of Reproductive Medicine (T-002), University of California, San Diego, La Jolla, CA 92093-0802. 611119493

catheterized ovine fetuses with a nonfunctional thoracic lymph duct catheter, two of four were grossly hydropic at the time of autopsy." In other studies, increases in central venous pressure (e.g., with fetal supraventricular tachycardia) were also associated with the development of hydrops!·6 Development of edema in the latter studies is consistent with the observation that an increase in central venous pressure decreases left thoracic duct lymph flow in the ovine fetus. 7 . 8 Thus a reduction in lymphatic flow may provide an explanation for the hydropic changes seen in cases of increased outflow pressure. Hydrops may develop rapidly because, with a flow rate in the left thoracic duct of 0.25 mllmin/k g 9 in the ovine fetus, or three to six times adult values,lo one would anticipate that an interruption of lymph flow would significantly affect fluid balance. The purpose of our investigation was to test the hypothesis that ligation of major lymphatic vessels would lead to the development of hydrops fetalis in the ovine fetus. As it became apparent that ligation did not con1331