- Email: [email protected]
Hypoxia-induced fetoplacental vasoconstriction in perfused human placental cotyledons
Volume 157 Number 5
tematic evaluation of clinical diagnostic criteria. Littleton, Massachusetts: PSG Publishing Co, 1977;60:217. 4. Quass L, Robrecht D, Kaltenbach FJ. The mean arterial pressure versus rollover test predictors of hypertension in pregnancy. In: Samour MB, Symonds EM, Zuspan FP, El-Tomi N, eds. Pregnancy hypertension. Cairo: Ain Shams University Press, 1982:145. 5. Phelan JP. Enhanced prediction of pregnancy-induced hypertension by combining supine pressor test with mean arterial pressure of middle trimester. AM ] OBSTET GvNECOL 1977;129:397-400. 6. Oney T, Kaulhansen H. The value of the mean arterial pressure in the second trimester (MAP-2 value) as a predictor of pregnancy-induced hypertension and preeclampsia. A preliminary report. Clin Exp Hypertens [A] 1983;B2:21l.
Average mean arterial pressure and later eclampsia
7. MoutquinJM, Rainville C, Giroux L, eta!. A prospective study of blood pressure in pregnancy: prediction of preeclampsia. AMJ 0BSTET GYNECOL 1985;151:191-6. 8. Moller B, Lindmark G. Eclampsia in Sweden, 1976-1980. Acta Obstet Gynecol Scand 1986;65:307-14. 9. Campbell DM, Templeton AA. Is eclampsia preventable? In: Bonnar J, MacGillivray I, Symonds EM, eds. Pregnancy hypertension. Baltimore: University Park Press, 1980:483-8. 10. Dieckmann WJ, ed. The toxemia in pregnancy. St. Louis: The CV Mosby Co, 1952:420. 11. Clark SL, Divon MY, Phelan JP. Preeclampsia/eclampsia: hemodynamic and neurologic correlations. Obstet Gynecol 1985;66:337-40.
Hypoxia-induced fetoplacental vasoconstriction in perfused human placental cotyledons Randy B. Howard, Ph.D., Tomokazu Hosokawa, Ph.D., and M. Helen Maguire, Ph.D. Kansas City, Kansas Effects of maternal hypoxia on fetoplacental vascular resistance in the human placenta were investigated in an in vitro model in which single anatomic subunits (cotyledons) from term placentas were perfused at constant flow through both fetal and maternal circuits by means of a physiologic salt solution containing dextran. Acute reduction of oxygen tension in the maternal perfusate induced prompt fetoplacental vasoconstriction that recovered rapidly on restoration of oxygen to the perfusate. The response, hypoxic fetoplacental vasoconstriction, could be repeatedly demonstrated in the same cotyledon. The time course of hypoxic fetoplacental vasoconstriction was inversely related to oxygen tension of maternal arterial and maternal and fetal venous perfusates. Maternal and fetal venous perfusate pH and Pco 2 did not change during the response. It is concluded that hypoxic fetoplacental vasoconstriction is triggered by decreased oxygen availability. It is suggested that hypoxic fetoplacental vasoconstriction may play a role in local regulation of human fetoplacental blood flow in vivo and may contribute to poor fetal prognosis in preeclampsia. (AM J 0BSTET GYNECOL 1987;157:1261-6.)
Key words: Hypoxia, fetoplacental vasoconstriction, placental perfusion A major function of the placenta is the transfer of oxygen and nutrients from maternal to fetal blood. Regulation of fetal blood flow through the placenta is not understood, but since fetoplacental vessels are not innervated, local and/or humoral mechanisms are likely
From the Department of Pharmacology, Toxicology, and Therapeutics, and the Ralph L. Smith Research Center, University of Kansas Medical Center. Supported by National Institutes of Health Grant HD 14888. Presented in part in abstract form at the Annual Meeting of the American Society for Pharmacology and Experimental Therapeutics, Indianapolis, Indiana, August 19-23, 1984. Received for publication September 22, 1986; revised June 12, 1987; accepted june 17, 1987. Reprint requests: M. Helen Maguire, Ph.D., Department of Pharmacology, Toxicology, and Therapeutics, University of Kansas Medical Center, Kansas City, KS 66103.
to be important. In the near-term sheep, occlusion of maternal placental blood flow reduced adjacent fetal placental flow, 1 and similarly ligation of a portion of the umbilical vasculature resulted in a decrease in adjacent maternal placental flow, 2 which support the concept of local re'gulatory interaction between fetal and maternal placental circulations in this species. 1 The ovine placenta is epitheliochorial and differs from the hemochorial human placenta in anatomic separation of maternal and fetal blood. Whether fetoplacental blood flow in human pregnancy is influenced by uteroplacental blood flow is not known. Demonstration of such an interaction in the human placenta would have potential significance in understanding how fetoplacental vascular resistance is controlled in normal pregnancy and in preeclampsia, a common disorder of pregnancy.
1261
1262
Howard, Hosokawa, and Maguire
Preeclampsia is typically associated with placental ischemia," and chronic hypoxia has been suggested as a factor contributing to the narrowing and obliteration of fetal arteries observed in placentas from patients with preeclampsia.• In the present study we asked whether uteroplacental hypoxia might affect human fetoplacental vascular resistance and thus modify fetal blood flow through the placenta. Effects of human placental hypoxia could not be determined in vivo because studies of placental function in vivo are limited to therapeutic and noninvasive procedures. Moreover, we did not have access to an animal model of the villous hemochorial placenta, such as the subhuman primate. We therefore used an in vitro model, the isolated dual-perfused cotyledon from human term placenta.' The human placenta is made up of a number of discrete anatomic subunits or cotyledons that can be isolated and perfused via both fetal and maternal circuits.' Maternal circulation in the human placenta is not confined to a separate vasculature but flows into the intervillous space and directly bathes the villi of the fetal placenta. In the placental model perfused in vitro, cannulas inserted through the maternal surface of the cotyledon deliver perfusate to the intervillous space': Careful placement of the cannulas is necessary to achieve adequate perfusion of the intervillous space. Using this preparation, we have determined dose-dependent effects of autacoids and humoral agents on fetoplacental vascular resistance. 6·' In this article we report that perfusion of the maternal circuit of the in vitro placental cotyledon with hypoxic medium causes reversible fetoplacental vasoconstriction. Material and methods Placentas from normal full-term pregnancies were obtained immediately after delivery. A single cotyledon from each placenta was set up for constant flow perfusion of the fetal circuit and intervillous space with Earle's salt solution to which 40 gm/L of dextran (approximate molecular weight 40,000) was added.'- 7 The upper chamber of the perfusion apparatus was open to allow access to the maternal surface of the perfused cotyledon. Preparation and perfusates were maintained at 37° C. Perfusate flow rates were monitored with Gilson No. 1 flowmeters. Fetal perfusate was gassed with 94% nitrogen, 6% carbon dioxide. Normoxic maternal perfusate was gassed with 95% oxygen, 5% carbon dioxide, and hypoxic maternal perfusate was gassed with 95% nitrogen, 5% carbon dioxide in a separate reservoir. Effective placement of maternal cannulas was monitored by measurement of fetal venous gas tensions and pH. Maternal flow rate was increased, if necessary, to achieve fetal venous pH of 7.30 to 7.35, Pco 2 of 39 to 47 mm Hg, and Po 2 > 100 mm Hg. Gas and pH values of perfusates were measured with a Radiometer
l\ovember 1987 Am J Obstet Gynecol
BMS 3 blood gas analyzer. Maternal and fetal arterial and venous perfusates were sampled with glass syringes from ports in the perfusion lines. Maternal venous perfusate was sampled by a syringe from the perfusate outflow above the maternal surface adjacent to the maternal cannulas. Perfusion pressures were monitored via Statham P23ID transducers and recorded on a Gilson ICT-2H duograph; perfusion pressure of the fetal circuit was used as an index of fetal vascular resistance. Experimental procedure. Control conditions of normoxie maternal perfusion were established. Hypoxia was then imposed by substitution of the normoxic maternal perfusate by hypoxic maternal perfusate for 8 to 15 minutes; perfusion with normoxic maternal perfusate was then reinstituted. Transit time from reservoir to intervillous space was 1 to 2 minutes. A period of 3 to 4 minutes was allowed to elapse between recovery and the next hypoxic challenge of the same duration. At the end of each experiment, a 1% solution of Coomassie blue in saline solution was infused via the fetal arterial cannula, and the cotyledon demarcated by the dye was dissected and weighed. Perfusate flow rates are expressed as ml min_, gm _,_ Perfusion medium. Earle's salt solution had the following composition (mmol/L): NaCl 116.4; NaHC0 3 26.2; KC15.36; CaCl2 1.80; MgSO. 0.81; NaH 2 P04 1.01; and glucose 5.5. NaCl was Fisher biologic grade. Other salts were the best analytic grade available. Dextran, molecular weight 38,000 to 42,000, was obtained from Chemical Dynamics Corp., New Jersey, or ICN, Ohio. Data analysis. Results are given as mean :!:: standard error of the mean. Differences between means of perfusate gas and pH values before and during maternal hypoxia were evaluated by a one-way analysis of variance F test with 4 and 15 degrees of freedom; BMDP7D 9 statistical software was used for computations. Correlation coefficients, partial correlation coefficients,'" and multiple-regression analysis were computed with BMDP6R and BMDP2R statistical software. 9 Results Weights of the 44 cotyledons used in the study ranged from 8 to 27 gm with a mean of 14.8 :!:: 0.7 gm. Control fetal vascular resistance in the preparations was 115 ± 6 mm Hg min gm ml- 1 , and mean flow rates of maternal and fetal perfusates were 1.35 ± 0.005 and 0.41 :!:: 0.02 ml min-' gm-', respectively. Perfusates had the following control gas and pH values: maternal arterial: pH 7.393 ± 0.003, Pco 2 34.8 ± 0.3 mm Hg, Po 2 540:!:: 4 mm Hg (n = 16 to 18 cotyledons); maternal venous: pH 7.358 ± 0.007, Pco 2 40.6 ± 0.4 mm Hg, Po 2 355 :!:: 14 mm Hg (n = 11 to 14 cotyledons); fetal arterial: pH 7.342 ± 0.007, Pco 2 42.9 :!:: 0.6 mm Hg, Po2 27 ± 2 mm Hg (n = six cotyledons); and fetal venous: pH 7.357 :!:: 0.003, Pco2 42.6 :!:: 0.4
Hypoxia-induced fetoplacental vasoconstriction
Volume 157 Number 5
A
-a
t
sa ___. x 120E
EE ......,
a. a.
95%02 5%C02
95%N2 6%C02 0
40
I "'------'
:
: - - - -...._ __
13 · i ~ min
:am·ln: I
soE~
.
3Q
t
I - - - " " "I ....
'5.
1
I
~ Q.
8 a.. a..
95%N 2 5%C02
7.41
•
+e
e
+e
I
I
7.30
L__j
1 min
1263
40E
30
I
•
I
e
t
I
••
600 6
4 2
0
0 0
4
8
Time
12
16
20
(min)
Fig. 1. A, Typical effects on maternal (upper tracing) and fetal (lower tracing) perfusion pressure observed when the maternal circuit of the human placental cotyledon was perfused with hypoxic medium. Arrows indicate the start of 9 minutes of hypoxic maternal perfusion (95% N 2 , 5% C0 2 ) and restoration of normoxic perfusion (95% 0 2 , 5% C0 2 ). B, Time course of averaged pressor responses to three sequential hypoxic stimuli of 10-minute duration in one cotyledon. (6 PP = Change in perfusion pressure [mm Hg], bars indicate standard error of the mean.)
mm Hg, Po 2 198 :±: 5 mm Hg (n = 42 cotyledons). The mean arteriovenous increase in fetal perfusate Po 2 of approximately 170 mm Hg indicated that a significant fraction of oxygen entering the maternal side of the cotyledon was transferred across the placenta. Perfusion of the maternal circuit of the cotyledons with hypoxic perfusate induced prompt reversible vasoconstriction of the fetoplacental vasculature as evidenced by a rise in fetal perfusion pressure 2 to 4 minutes after perfusion with the hypoxic perfusate commenced, which was followed by a drop in pressure 2 to 4 minutes after delivery of normoxic maternal perfusate was reinstituted (Fig. 1, A). Fetoplacental vasoconstriction induced by maternal hypoxia could be repeatedly elicited in the same cotyledon (Fig. 1, B). Return to baseline perfusion pressure was complete or nearly complete with recovery from each response (Figs. 1, B, 2, and 3). This reversible response is described as hypoxic fetoplacental vasoconstriction. Development of the fetal pressor response to maternal hypoxia and recovery of baseline perfusion pressure paralleled changes in the oxygen tension of maternal and fetal perfusates. Fig. 2 shows the fall in rna-
a.. a..
:f
I
I 0
4
8
Time
12
16
20
(min)
Fig. 2. Time courses of a typical fetal pressor response and associated maternal arterial perfusate gas and pH values induced by 10-minute hypoxic challenge in a single cotyledon. Arrows indicate start of hypoxic maternal perfusion (95% N 2 , 5% C0 2 ) and normoxic maternal perfusion (95% 0 2 , 5% C0 2 ). (6 PP = Change in fetal perfusion pressure [mm Hg].)
ternal arterial Po 2 from 540 to 25 mm Hg that accompanies the onset of hypoxic fetoplacental vasoconstriction and the subsequent return to control level Po 2 associated with recovery from the response. The onset of hypoxic fetoplacental vasoconstriction and recovery from hypoxic fetoplacental vasoconstriction were also reflected in the fall and rise, respectively, of Po 2 in both maternal and fetal venous perfusates as shown for a typical experiment in Fig. 3. Oxygen tensions in both maternal and fetal venous perfusates varied significantly during hypoxic fetoplacental vasoconstriction (fetal venous F = 490.48, p < 0.0001; maternal venous F = 70.85, p < 0.0001). As indicated in Fig. 3, the pressor response commenced as the fetal venous Po 2 fell below approximately 100 mm Hg and continued to rise as fetal venous Po 2 fell to below 40 to 50 mm Hg 2 to 4 minutes after reinstitution of normoxie perfusion. No significant change from prehypoxic maternal venous or fetal venous pH was observed during hypoxic fetoplacental vasoconstriction (compare Fig. 3; fetal venous, F = 0.12, p = 0.97; maternal venous, F = 0.47, p = 0.76). No significant change from prehypoxic values occurred in either fetal venous
1264 Howard, Hosokawa, and Maguire
~7.3r
0C)
0.
00. Q. Q.
<3
7.3
45[ 40
.. ·--..
November 1987 Am J Obstet Gynecol
.. -i-.:t---f
95%N 2 5%C0 2
95%02 5%C02
a-i-= ~---
.
_.._T_t .._-----I t
1
1
.• • • . . •
!
•of
!
• r: -.383
•
Ol
......
0
..
•
-;
r: .299
::r::
E
I 0
::!
4
8
Time
12
16
20
22
(min)
Fig. 3. Time courses of fetal (broken lines) and maternal (solid lines) venous perfusate gas and pH values induced by three hypoxic stimuli of 11-minute duration and a typical pressor response in the same cotyledon. Arrows indicate start of maternal hypoxic (95% N 2 , 5% CO,) and maternal normoxic (95% 02o 5% C02 ) perfusion. Gas and pH values are mean ::!: standard error of the mean; where not shown standard error of the mean is within the data point. (6 PP = Change in perfusion pressure [mm Hg].)
Pco, or maternal venous Pco 2 during the response (fetal venous F 0.25, p = 0.90; maternal venous F = 0.28, p = 0.89) (Fig. 3). Furthermore, in two preparations subjected to repeated hypoxia, neither Pco, nor pH of fetal venous perfusate measured before hypoxic vasoconstriction differed significantly from pH and Pco, measured during peak hypoxic vasoconstriction within each preparation (Table 1). However, fetal venous Po, during hypoxic constriction fell to less than 25% of the prehypoxia control level (Table 1). Similarly, comparison of gas tensions and pH values 'of maternal venous perfusates before and during repeated hypoxic fetoplacental vasoconstriction in three different cotyledons showed no difference in pH and Pco 2 before and during the responses but showed a twofold to threefold fall in Po, (data not shown). In 55% of the cotyledons studied, the maximum extent of vasoconstriction to hypoxic stimuli of 9- to 11minute duration was observed with the first hypoxic stimulus, but in 45% of the cotyledons the magnitude of the response increased with repeated hypoxia until the maximum response observed in the cotyledon was achieved with the third or fourth stimulus. The reason for this difference between cotyledons is not presently understood. The maximum hypoxic vasoconstriction for each cotyledon, which was normalized as pressure
!N
0
-
~~~~~··~-
.•
';. 150
u.
A
!:::. PP /min
.a
1.2
1.s
(mm Hg/minl
Fig. 4. Relationships between the normalized fetal pressor response to hypoxic challenge (6 PP min· 1, mm Hg min- 1) and each of three different parameters of perfusion: cotyledon weight (wt), maternal flow per gm (QM gm· 1, ml min- 1 gm - 1), and control (i.e., before hypoxic challenge) fetal venous Po2 (F,Po 2) in 41 different cotyledons. (r = simple correlation coefficient.)
increase per minute of hypoxic stimulus, ranged from 0.3 to 1.6 mm Hg min·\ with a mean of 0.76 ± 0.05 mm Hg min- 1 (n = 44 cotyledons). It was thought that the variation between cotyledons in the normalized response to maternal hypoxia might be related to parameters of the preparation such as cotyledon weight, maternal and fetal flow rates (ml min- 1 gm- 1), control fetoplacental vascular resistance (mm Hg min gm ml- 1), or control perfusate gas tensions and pH. Regression analysis showed that the normalized hypoxic pressor response indeed correlated with cotyledon weight, maternal flow rate, and control (normoxic) fetal venous Po,, as shown in Fig. 4 and by the simple correlation coefficients given in the figure. However, the response did not correlate with fetal venous pH (r = 0.04 7), fetal venous Pco, (r = 0.09), fetal perfusate flow rate (r = 0.141 ), or fetoplacental vascular resistance (r = 0.036). Further statistical analysis was performed to elucidate how the normalized hypoxic pressor response was related to cotyledon weight, maternal flow rate, and normoxic fetal venous Po,. Specifically, partial correlation coefficients were computed to assess the extent to which each of these three parameters were (uniquely)
Hypoxia-induced fetoplacental vasoconstriction
Volume 157 Number 5
1265
Table I. Comparison of fetal venous perfusate pH, Pco 2 , and Po 2 during normoxic maternal perfusion before hypoxic challenge with pH, Pco 2 , and Po, at the peak pressor response to hypoxic maternal perfusion in two dual-perfused placental cotyledons Cotyledon No.
2
Duration of hypoxia (min)
Fetal venous pH n*
10
7
11
4
Control
7.372
± 0.001
7.320 ± 0.005
I
Fetal venous Pco2
Hypoxia
Control
7.373 ± 0.001 7.316 ± 0.01
41.3 ± 0.3 43.1 ± 0.7
I
Hypoxia
41.3 ± 0.2 44.2 ± 1.0
Fetal venous Po, Control
196 ± 2 155 ± 2
I
Hypoxia
47 ± 3 31 ± 2
Values are mean ± SEM. *Number of sequential hypoxic stimuli; replicate three to five measurements of pH, Pco 2 , and Po, were made in control and hypoxia perfusates sampled before and during each pressor response.
related to the pressor response while statistically controlling the influences of the other two. Partial correlation coefficients of the pressor response with fetal venous Po,, cotyledon weight, and maternal How rate (ml min- 1 gm- 1) were 0.449, 0.325, and -0.087, respectively, thus suggesting that control fetal venous Po 2 was the parameter most highly related to the normalized pressor response independent of the other two parameters. However, the partial correlation coefficient of 0.325 associated with cotyledon weight suggests that it too may be uniquely related to the magnitude of the normalized hypoxic pressor response. These conclusions were confirmed by multiple-regression analysis for the normalized pressor response (R = 0.6035). Of "the three individual regression coefficients, those for fetal venous l'o2 (p < 0.01) and cotyledon weight (p < 0.05) were statistically significant. Comment
In this study the effect of acute uteroplacental hypoxia on human fetoplacental vascular resistance has been investigated with the isolated dual-perfused placental cotyledon as an in vitro experimental model. Several parameters of this in vitro model resembled those reported for the in vivo placenta. Specifically, maternal and fetal How rates of 1.35 ± 0.005 and 0.41 :!:: 0.02 ml min- 1 gm- 1 were similar to How rates calculated for maternal and umbilical blood in vivo, that is, 1.0 and 0.6 ml min- 1 gm- 1 , respectively (calculated using reported uteroplacental" and umbilical'' How rates and assuming a placental weight of 500 gm), and gas tensions and pH of the fetal arterial perfusate approximated those found in vivo. 13 With this model we demonstrated that maternal hypoxia induced reversible fetoplacental vasoconstriction. The response was inversely related to maternal arterial perfusate oxygen tension and was associated with a reversible drop in Po, of maternal and fetal venous perfusates. However, neither pH nor Pco, of the venous perfusates changed during the hypoxic response, which suggests that the mechanism of the response does not involve change in
glucose metabolism. We conclude that decreased oxygen availability is the primary triggering factor in hypoxic fetoplacental vasoconstriction since oxygen tension was the only variable changed. A fivefold variation in the size of the normalized hypoxic vasoconstriction between different cotyledons was observed and was shown by statistical analysis to be correlated with two prestimulus perfusion parameters, control fetal venous Po, and cotyledon· weight. These findings suggest that cotyledons that transferred oxygen most effectively across the placenta and larger cotyledons were more sensitive to maternal hypoxia. Reasons for the increased sensitivity are unknown. However, it may be speculated that general function was better preserved in those cotyledons that transferred oxygen to the fetalcirculation more effectively, and that as a result sensitivity to lack of oxygen was greater. On the other hand, the basis of the positive correlation between cotyledon size and sensitivity to maternal hypoxia is less readily explained. It is unlikely to be the result of better maternal perfusion in the larger cotyledons, since analysis of interrelationships of perfusion parameters under control conditions has shown that maternal How rate (ml min- 1 gm- 1) correlated negatively with cotyledon Weight. 11 Again, it may be that function is better preserved at term in the larger cotyledons. Although we have demonstrated hypoxic fetoplacental vasoconstriction in an artificially perfused placental preparation in vitro, the response could be important in vivo, for example, in acute local blood How regulation in which it may result in a redistribution of fetoplacental blood How from hypoxic to normoxic or better perfused regions of the placenta. In this regard, the response may be analogous to the phenomenon of hypoxic pulmonary vasoconstriction, which is important in the adjustment of ventilation-perfusion ratios in the lung. We have shown that verapamil and lipoxygenase inhibitors antagonize hypoxic fetoplacental vasoconstriction, 15 which suggests that the response is mediated, at least in part, by release of lipoxygenase products.
1266
Howard, Hosokawa, and Maguire
Fetoplacental vasoconstriction in response to maternal hypoxia has not been previously described, although it is reminiscent of the reduction of fetal placental blood flow observed 24 hours after maternal placental occlusion in the near-term sheep. 1 However, we know of no evidence that the ovine response to occlusion of maternal placental flow occurs acutely and reversibly. An acute pressor response to abolition of maternal perfusion can be demonstrated in the in vitro placental preparation and is reversible on restoration of maternal perfusion. 16 Hypoxic fetoplacental vasoconstriction in the in vitro cotyledon resulted from acute change in maternal oxygen tension. It seems possible that in vivo, chronic placental hypoxia such as has been postulated to occur in preeclampsia3 • 1 may cause a chronic vasoconstrictor response that may contribute to fetal mortality and growth retardation associated with this disease. Furthermore, release of hypoxia-induced vasoconstrictive mediators may be important in the etiology of preeclampsia. We thank Dr. C. R. King and the staff of the delivery room of Bell Memorial Hospital for their help in obtaining the placentas, and Dr. J. Levy for performing data analysis.
November 1987 Am J Obstet Gynecol
5. 6. 7.
8.
9.
10. 11.
12.
13. 14.
REFERENCES l. Stock MK, Anderson DF, Phernetton TM, McLaughlin MK, Rankin JHG. Vascular response of the fetal placenta to local occlusion of the maternal placental vasculature. J Devel Physiol 1980;2:339-46. 2. RankinJHG, Goodman A, Phernetton T. Local regulation of uterine blood flow by the umbilical circulation. Proc Soc Exp Bioi Med 1975;150:690-4. 3. Chesley LC. Hypertensive disorders in pregnancy. New York: Appleton-Century-Crofts, 1978:455-66. 4. Las Heras J, Baskerville JC, Harding PGR, Haust MD.
15.
16.
Morphometric studies of fetal placental stem arteries in hypertensive disorders of pregnancy. Placenta 1985;6: 217-28. Schneider H, Panigel M, Dancis J. Transfer across the perfused human placenta of antipyrine, sodium and leucine. AM j 0BSTET GYNECOL 1972; 113:822-8. Hosokawa T, Howard RB, Maguire MH. Conversion of angiotensin I to angiotensin II in the human fetoplacental vascular bed. Br J Pharmacal 1985;84:237-41. Howard RB, Hosokawa T, Maguire MH. Pressor and depressor actions of prostanoids in the intact human fetoplacental vascular bed. Prostaglandins Leukotrienes Med 1986;21 :323-30. Maguire MH, Howard RB; Hosokawa T, Poisner AM. Effects of some autacoids and humoral agents on human fetoplacental vascular tesistance candidates for local regulation of fetoplacental blood flow. Trophoblast Res 1987 (In press). Dixon WJ. BMDP statistical software. Berkeley: University of California Press, 1983. Draper NR, Smith H. Applied regression analysis. New York: John Wiley & Sons, 1981:265-6. Page EW. The placenta and fetus; trophoblastic diseases. In: Benson RC, ed. Current obstetric and gynecologic diagnosis and treatment. Los Altos, California: Lange Medical Publications, 1978:533-59. Gill RW, Trudinger BJ, Garrett WJ, Kossoff G, Warren PS. Fetal umbilical venous flow measured in utero by pulsed doppler and B-mode ultrasound. I. Normal pregnancies. AM j 0BSTET GYNECOL 1981; 139:720-5. Longo L. Disorders of placental transfer. In: Assali NS, Brinkman CR, eds. Pathophysiology of gestation. New York: Academic Press, 1972:1-65. Howard RB, Levy J, Hosokawa T, Maguire MH. Interrelationships of perfusion parameters in the dualperfused human placental cotyledon. Trophoblast Res 1987;2:585-96. Howard RB, Hosokawa T, Maguire MH. Hypoxiainduced fetoplacental vasoconstriction in cotyledons of human term placentas: antagonism by selected drugs [Abstract]. Pharmacologist 1984;26: 144. Kitagawa H, Siegel P, Maguire MH. Ischemia-induced vasoconstriction and adenosine release in perfused human placental cotyledons [Abstract]. Pharmacologist 1987; 29:197.
tematic evaluation of clinical diagnostic criteria. Littleton, Massachusetts: PSG Publishing Co, 1977;60:217. 4. Quass L, Robrecht D, Kaltenbach FJ. The mean arterial pressure versus rollover test predictors of hypertension in pregnancy. In: Samour MB, Symonds EM, Zuspan FP, El-Tomi N, eds. Pregnancy hypertension. Cairo: Ain Shams University Press, 1982:145. 5. Phelan JP. Enhanced prediction of pregnancy-induced hypertension by combining supine pressor test with mean arterial pressure of middle trimester. AM ] OBSTET GvNECOL 1977;129:397-400. 6. Oney T, Kaulhansen H. The value of the mean arterial pressure in the second trimester (MAP-2 value) as a predictor of pregnancy-induced hypertension and preeclampsia. A preliminary report. Clin Exp Hypertens [A] 1983;B2:21l.
Average mean arterial pressure and later eclampsia
7. MoutquinJM, Rainville C, Giroux L, eta!. A prospective study of blood pressure in pregnancy: prediction of preeclampsia. AMJ 0BSTET GYNECOL 1985;151:191-6. 8. Moller B, Lindmark G. Eclampsia in Sweden, 1976-1980. Acta Obstet Gynecol Scand 1986;65:307-14. 9. Campbell DM, Templeton AA. Is eclampsia preventable? In: Bonnar J, MacGillivray I, Symonds EM, eds. Pregnancy hypertension. Baltimore: University Park Press, 1980:483-8. 10. Dieckmann WJ, ed. The toxemia in pregnancy. St. Louis: The CV Mosby Co, 1952:420. 11. Clark SL, Divon MY, Phelan JP. Preeclampsia/eclampsia: hemodynamic and neurologic correlations. Obstet Gynecol 1985;66:337-40.
Hypoxia-induced fetoplacental vasoconstriction in perfused human placental cotyledons Randy B. Howard, Ph.D., Tomokazu Hosokawa, Ph.D., and M. Helen Maguire, Ph.D. Kansas City, Kansas Effects of maternal hypoxia on fetoplacental vascular resistance in the human placenta were investigated in an in vitro model in which single anatomic subunits (cotyledons) from term placentas were perfused at constant flow through both fetal and maternal circuits by means of a physiologic salt solution containing dextran. Acute reduction of oxygen tension in the maternal perfusate induced prompt fetoplacental vasoconstriction that recovered rapidly on restoration of oxygen to the perfusate. The response, hypoxic fetoplacental vasoconstriction, could be repeatedly demonstrated in the same cotyledon. The time course of hypoxic fetoplacental vasoconstriction was inversely related to oxygen tension of maternal arterial and maternal and fetal venous perfusates. Maternal and fetal venous perfusate pH and Pco 2 did not change during the response. It is concluded that hypoxic fetoplacental vasoconstriction is triggered by decreased oxygen availability. It is suggested that hypoxic fetoplacental vasoconstriction may play a role in local regulation of human fetoplacental blood flow in vivo and may contribute to poor fetal prognosis in preeclampsia. (AM J 0BSTET GYNECOL 1987;157:1261-6.)
Key words: Hypoxia, fetoplacental vasoconstriction, placental perfusion A major function of the placenta is the transfer of oxygen and nutrients from maternal to fetal blood. Regulation of fetal blood flow through the placenta is not understood, but since fetoplacental vessels are not innervated, local and/or humoral mechanisms are likely
From the Department of Pharmacology, Toxicology, and Therapeutics, and the Ralph L. Smith Research Center, University of Kansas Medical Center. Supported by National Institutes of Health Grant HD 14888. Presented in part in abstract form at the Annual Meeting of the American Society for Pharmacology and Experimental Therapeutics, Indianapolis, Indiana, August 19-23, 1984. Received for publication September 22, 1986; revised June 12, 1987; accepted june 17, 1987. Reprint requests: M. Helen Maguire, Ph.D., Department of Pharmacology, Toxicology, and Therapeutics, University of Kansas Medical Center, Kansas City, KS 66103.
to be important. In the near-term sheep, occlusion of maternal placental blood flow reduced adjacent fetal placental flow, 1 and similarly ligation of a portion of the umbilical vasculature resulted in a decrease in adjacent maternal placental flow, 2 which support the concept of local re'gulatory interaction between fetal and maternal placental circulations in this species. 1 The ovine placenta is epitheliochorial and differs from the hemochorial human placenta in anatomic separation of maternal and fetal blood. Whether fetoplacental blood flow in human pregnancy is influenced by uteroplacental blood flow is not known. Demonstration of such an interaction in the human placenta would have potential significance in understanding how fetoplacental vascular resistance is controlled in normal pregnancy and in preeclampsia, a common disorder of pregnancy.
1261
1262
Howard, Hosokawa, and Maguire
Preeclampsia is typically associated with placental ischemia," and chronic hypoxia has been suggested as a factor contributing to the narrowing and obliteration of fetal arteries observed in placentas from patients with preeclampsia.• In the present study we asked whether uteroplacental hypoxia might affect human fetoplacental vascular resistance and thus modify fetal blood flow through the placenta. Effects of human placental hypoxia could not be determined in vivo because studies of placental function in vivo are limited to therapeutic and noninvasive procedures. Moreover, we did not have access to an animal model of the villous hemochorial placenta, such as the subhuman primate. We therefore used an in vitro model, the isolated dual-perfused cotyledon from human term placenta.' The human placenta is made up of a number of discrete anatomic subunits or cotyledons that can be isolated and perfused via both fetal and maternal circuits.' Maternal circulation in the human placenta is not confined to a separate vasculature but flows into the intervillous space and directly bathes the villi of the fetal placenta. In the placental model perfused in vitro, cannulas inserted through the maternal surface of the cotyledon deliver perfusate to the intervillous space': Careful placement of the cannulas is necessary to achieve adequate perfusion of the intervillous space. Using this preparation, we have determined dose-dependent effects of autacoids and humoral agents on fetoplacental vascular resistance. 6·' In this article we report that perfusion of the maternal circuit of the in vitro placental cotyledon with hypoxic medium causes reversible fetoplacental vasoconstriction. Material and methods Placentas from normal full-term pregnancies were obtained immediately after delivery. A single cotyledon from each placenta was set up for constant flow perfusion of the fetal circuit and intervillous space with Earle's salt solution to which 40 gm/L of dextran (approximate molecular weight 40,000) was added.'- 7 The upper chamber of the perfusion apparatus was open to allow access to the maternal surface of the perfused cotyledon. Preparation and perfusates were maintained at 37° C. Perfusate flow rates were monitored with Gilson No. 1 flowmeters. Fetal perfusate was gassed with 94% nitrogen, 6% carbon dioxide. Normoxic maternal perfusate was gassed with 95% oxygen, 5% carbon dioxide, and hypoxic maternal perfusate was gassed with 95% nitrogen, 5% carbon dioxide in a separate reservoir. Effective placement of maternal cannulas was monitored by measurement of fetal venous gas tensions and pH. Maternal flow rate was increased, if necessary, to achieve fetal venous pH of 7.30 to 7.35, Pco 2 of 39 to 47 mm Hg, and Po 2 > 100 mm Hg. Gas and pH values of perfusates were measured with a Radiometer
l\ovember 1987 Am J Obstet Gynecol
BMS 3 blood gas analyzer. Maternal and fetal arterial and venous perfusates were sampled with glass syringes from ports in the perfusion lines. Maternal venous perfusate was sampled by a syringe from the perfusate outflow above the maternal surface adjacent to the maternal cannulas. Perfusion pressures were monitored via Statham P23ID transducers and recorded on a Gilson ICT-2H duograph; perfusion pressure of the fetal circuit was used as an index of fetal vascular resistance. Experimental procedure. Control conditions of normoxie maternal perfusion were established. Hypoxia was then imposed by substitution of the normoxic maternal perfusate by hypoxic maternal perfusate for 8 to 15 minutes; perfusion with normoxic maternal perfusate was then reinstituted. Transit time from reservoir to intervillous space was 1 to 2 minutes. A period of 3 to 4 minutes was allowed to elapse between recovery and the next hypoxic challenge of the same duration. At the end of each experiment, a 1% solution of Coomassie blue in saline solution was infused via the fetal arterial cannula, and the cotyledon demarcated by the dye was dissected and weighed. Perfusate flow rates are expressed as ml min_, gm _,_ Perfusion medium. Earle's salt solution had the following composition (mmol/L): NaCl 116.4; NaHC0 3 26.2; KC15.36; CaCl2 1.80; MgSO. 0.81; NaH 2 P04 1.01; and glucose 5.5. NaCl was Fisher biologic grade. Other salts were the best analytic grade available. Dextran, molecular weight 38,000 to 42,000, was obtained from Chemical Dynamics Corp., New Jersey, or ICN, Ohio. Data analysis. Results are given as mean :!:: standard error of the mean. Differences between means of perfusate gas and pH values before and during maternal hypoxia were evaluated by a one-way analysis of variance F test with 4 and 15 degrees of freedom; BMDP7D 9 statistical software was used for computations. Correlation coefficients, partial correlation coefficients,'" and multiple-regression analysis were computed with BMDP6R and BMDP2R statistical software. 9 Results Weights of the 44 cotyledons used in the study ranged from 8 to 27 gm with a mean of 14.8 :!:: 0.7 gm. Control fetal vascular resistance in the preparations was 115 ± 6 mm Hg min gm ml- 1 , and mean flow rates of maternal and fetal perfusates were 1.35 ± 0.005 and 0.41 :!:: 0.02 ml min-' gm-', respectively. Perfusates had the following control gas and pH values: maternal arterial: pH 7.393 ± 0.003, Pco 2 34.8 ± 0.3 mm Hg, Po 2 540:!:: 4 mm Hg (n = 16 to 18 cotyledons); maternal venous: pH 7.358 ± 0.007, Pco 2 40.6 ± 0.4 mm Hg, Po 2 355 :!:: 14 mm Hg (n = 11 to 14 cotyledons); fetal arterial: pH 7.342 ± 0.007, Pco 2 42.9 :!:: 0.6 mm Hg, Po2 27 ± 2 mm Hg (n = six cotyledons); and fetal venous: pH 7.357 :!:: 0.003, Pco2 42.6 :!:: 0.4
Hypoxia-induced fetoplacental vasoconstriction
Volume 157 Number 5
A
-a
t
sa ___. x 120E
EE ......,
a. a.
95%02 5%C02
95%N2 6%C02 0
40
I "'------'
:
: - - - -...._ __
13 · i ~ min
:am·ln: I
soE~
.
3Q
t
I - - - " " "I ....
'5.
1
I
~ Q.
8 a.. a..
95%N 2 5%C02
7.41
•
+e
e
+e
I
I
7.30
L__j
1 min
1263
40E
30
I
•
I
e
t
I
••
600 6
4 2
0
0 0
4
8
Time
12
16
20
(min)
Fig. 1. A, Typical effects on maternal (upper tracing) and fetal (lower tracing) perfusion pressure observed when the maternal circuit of the human placental cotyledon was perfused with hypoxic medium. Arrows indicate the start of 9 minutes of hypoxic maternal perfusion (95% N 2 , 5% C0 2 ) and restoration of normoxic perfusion (95% 0 2 , 5% C0 2 ). B, Time course of averaged pressor responses to three sequential hypoxic stimuli of 10-minute duration in one cotyledon. (6 PP = Change in perfusion pressure [mm Hg], bars indicate standard error of the mean.)
mm Hg, Po 2 198 :±: 5 mm Hg (n = 42 cotyledons). The mean arteriovenous increase in fetal perfusate Po 2 of approximately 170 mm Hg indicated that a significant fraction of oxygen entering the maternal side of the cotyledon was transferred across the placenta. Perfusion of the maternal circuit of the cotyledons with hypoxic perfusate induced prompt reversible vasoconstriction of the fetoplacental vasculature as evidenced by a rise in fetal perfusion pressure 2 to 4 minutes after perfusion with the hypoxic perfusate commenced, which was followed by a drop in pressure 2 to 4 minutes after delivery of normoxic maternal perfusate was reinstituted (Fig. 1, A). Fetoplacental vasoconstriction induced by maternal hypoxia could be repeatedly elicited in the same cotyledon (Fig. 1, B). Return to baseline perfusion pressure was complete or nearly complete with recovery from each response (Figs. 1, B, 2, and 3). This reversible response is described as hypoxic fetoplacental vasoconstriction. Development of the fetal pressor response to maternal hypoxia and recovery of baseline perfusion pressure paralleled changes in the oxygen tension of maternal and fetal perfusates. Fig. 2 shows the fall in rna-
a.. a..
:f
I
I 0
4
8
Time
12
16
20
(min)
Fig. 2. Time courses of a typical fetal pressor response and associated maternal arterial perfusate gas and pH values induced by 10-minute hypoxic challenge in a single cotyledon. Arrows indicate start of hypoxic maternal perfusion (95% N 2 , 5% C0 2 ) and normoxic maternal perfusion (95% 0 2 , 5% C0 2 ). (6 PP = Change in fetal perfusion pressure [mm Hg].)
ternal arterial Po 2 from 540 to 25 mm Hg that accompanies the onset of hypoxic fetoplacental vasoconstriction and the subsequent return to control level Po 2 associated with recovery from the response. The onset of hypoxic fetoplacental vasoconstriction and recovery from hypoxic fetoplacental vasoconstriction were also reflected in the fall and rise, respectively, of Po 2 in both maternal and fetal venous perfusates as shown for a typical experiment in Fig. 3. Oxygen tensions in both maternal and fetal venous perfusates varied significantly during hypoxic fetoplacental vasoconstriction (fetal venous F = 490.48, p < 0.0001; maternal venous F = 70.85, p < 0.0001). As indicated in Fig. 3, the pressor response commenced as the fetal venous Po 2 fell below approximately 100 mm Hg and continued to rise as fetal venous Po 2 fell to below 40 to 50 mm Hg 2 to 4 minutes after reinstitution of normoxie perfusion. No significant change from prehypoxic maternal venous or fetal venous pH was observed during hypoxic fetoplacental vasoconstriction (compare Fig. 3; fetal venous, F = 0.12, p = 0.97; maternal venous, F = 0.47, p = 0.76). No significant change from prehypoxic values occurred in either fetal venous
1264 Howard, Hosokawa, and Maguire
~7.3r
0C)
0.
00. Q. Q.
<3
7.3
45[ 40
.. ·--..
November 1987 Am J Obstet Gynecol
.. -i-.:t---f
95%N 2 5%C0 2
95%02 5%C02
a-i-= ~---
.
_.._T_t .._-----I t
1
1
.• • • . . •
!
•of
!
• r: -.383
•
Ol
......
0
..
•
-;
r: .299
::r::
E
I 0
::!
4
8
Time
12
16
20
22
(min)
Fig. 3. Time courses of fetal (broken lines) and maternal (solid lines) venous perfusate gas and pH values induced by three hypoxic stimuli of 11-minute duration and a typical pressor response in the same cotyledon. Arrows indicate start of maternal hypoxic (95% N 2 , 5% CO,) and maternal normoxic (95% 02o 5% C02 ) perfusion. Gas and pH values are mean ::!: standard error of the mean; where not shown standard error of the mean is within the data point. (6 PP = Change in perfusion pressure [mm Hg].)
Pco, or maternal venous Pco 2 during the response (fetal venous F 0.25, p = 0.90; maternal venous F = 0.28, p = 0.89) (Fig. 3). Furthermore, in two preparations subjected to repeated hypoxia, neither Pco, nor pH of fetal venous perfusate measured before hypoxic vasoconstriction differed significantly from pH and Pco, measured during peak hypoxic vasoconstriction within each preparation (Table 1). However, fetal venous Po, during hypoxic constriction fell to less than 25% of the prehypoxia control level (Table 1). Similarly, comparison of gas tensions and pH values 'of maternal venous perfusates before and during repeated hypoxic fetoplacental vasoconstriction in three different cotyledons showed no difference in pH and Pco 2 before and during the responses but showed a twofold to threefold fall in Po, (data not shown). In 55% of the cotyledons studied, the maximum extent of vasoconstriction to hypoxic stimuli of 9- to 11minute duration was observed with the first hypoxic stimulus, but in 45% of the cotyledons the magnitude of the response increased with repeated hypoxia until the maximum response observed in the cotyledon was achieved with the third or fourth stimulus. The reason for this difference between cotyledons is not presently understood. The maximum hypoxic vasoconstriction for each cotyledon, which was normalized as pressure
!N
0
-
~~~~~··~-
.•
';. 150
u.
A
!:::. PP /min
.a
1.2
1.s
(mm Hg/minl
Fig. 4. Relationships between the normalized fetal pressor response to hypoxic challenge (6 PP min· 1, mm Hg min- 1) and each of three different parameters of perfusion: cotyledon weight (wt), maternal flow per gm (QM gm· 1, ml min- 1 gm - 1), and control (i.e., before hypoxic challenge) fetal venous Po2 (F,Po 2) in 41 different cotyledons. (r = simple correlation coefficient.)
increase per minute of hypoxic stimulus, ranged from 0.3 to 1.6 mm Hg min·\ with a mean of 0.76 ± 0.05 mm Hg min- 1 (n = 44 cotyledons). It was thought that the variation between cotyledons in the normalized response to maternal hypoxia might be related to parameters of the preparation such as cotyledon weight, maternal and fetal flow rates (ml min- 1 gm- 1), control fetoplacental vascular resistance (mm Hg min gm ml- 1), or control perfusate gas tensions and pH. Regression analysis showed that the normalized hypoxic pressor response indeed correlated with cotyledon weight, maternal flow rate, and control (normoxic) fetal venous Po,, as shown in Fig. 4 and by the simple correlation coefficients given in the figure. However, the response did not correlate with fetal venous pH (r = 0.04 7), fetal venous Pco, (r = 0.09), fetal perfusate flow rate (r = 0.141 ), or fetoplacental vascular resistance (r = 0.036). Further statistical analysis was performed to elucidate how the normalized hypoxic pressor response was related to cotyledon weight, maternal flow rate, and normoxic fetal venous Po,. Specifically, partial correlation coefficients were computed to assess the extent to which each of these three parameters were (uniquely)
Hypoxia-induced fetoplacental vasoconstriction
Volume 157 Number 5
1265
Table I. Comparison of fetal venous perfusate pH, Pco 2 , and Po 2 during normoxic maternal perfusion before hypoxic challenge with pH, Pco 2 , and Po, at the peak pressor response to hypoxic maternal perfusion in two dual-perfused placental cotyledons Cotyledon No.
2
Duration of hypoxia (min)
Fetal venous pH n*
10
7
11
4
Control
7.372
± 0.001
7.320 ± 0.005
I
Fetal venous Pco2
Hypoxia
Control
7.373 ± 0.001 7.316 ± 0.01
41.3 ± 0.3 43.1 ± 0.7
I
Hypoxia
41.3 ± 0.2 44.2 ± 1.0
Fetal venous Po, Control
196 ± 2 155 ± 2
I
Hypoxia
47 ± 3 31 ± 2
Values are mean ± SEM. *Number of sequential hypoxic stimuli; replicate three to five measurements of pH, Pco 2 , and Po, were made in control and hypoxia perfusates sampled before and during each pressor response.
related to the pressor response while statistically controlling the influences of the other two. Partial correlation coefficients of the pressor response with fetal venous Po,, cotyledon weight, and maternal How rate (ml min- 1 gm- 1) were 0.449, 0.325, and -0.087, respectively, thus suggesting that control fetal venous Po 2 was the parameter most highly related to the normalized pressor response independent of the other two parameters. However, the partial correlation coefficient of 0.325 associated with cotyledon weight suggests that it too may be uniquely related to the magnitude of the normalized hypoxic pressor response. These conclusions were confirmed by multiple-regression analysis for the normalized pressor response (R = 0.6035). Of "the three individual regression coefficients, those for fetal venous l'o2 (p < 0.01) and cotyledon weight (p < 0.05) were statistically significant. Comment
In this study the effect of acute uteroplacental hypoxia on human fetoplacental vascular resistance has been investigated with the isolated dual-perfused placental cotyledon as an in vitro experimental model. Several parameters of this in vitro model resembled those reported for the in vivo placenta. Specifically, maternal and fetal How rates of 1.35 ± 0.005 and 0.41 :!:: 0.02 ml min- 1 gm- 1 were similar to How rates calculated for maternal and umbilical blood in vivo, that is, 1.0 and 0.6 ml min- 1 gm- 1 , respectively (calculated using reported uteroplacental" and umbilical'' How rates and assuming a placental weight of 500 gm), and gas tensions and pH of the fetal arterial perfusate approximated those found in vivo. 13 With this model we demonstrated that maternal hypoxia induced reversible fetoplacental vasoconstriction. The response was inversely related to maternal arterial perfusate oxygen tension and was associated with a reversible drop in Po, of maternal and fetal venous perfusates. However, neither pH nor Pco, of the venous perfusates changed during the hypoxic response, which suggests that the mechanism of the response does not involve change in
glucose metabolism. We conclude that decreased oxygen availability is the primary triggering factor in hypoxic fetoplacental vasoconstriction since oxygen tension was the only variable changed. A fivefold variation in the size of the normalized hypoxic vasoconstriction between different cotyledons was observed and was shown by statistical analysis to be correlated with two prestimulus perfusion parameters, control fetal venous Po, and cotyledon· weight. These findings suggest that cotyledons that transferred oxygen most effectively across the placenta and larger cotyledons were more sensitive to maternal hypoxia. Reasons for the increased sensitivity are unknown. However, it may be speculated that general function was better preserved in those cotyledons that transferred oxygen to the fetalcirculation more effectively, and that as a result sensitivity to lack of oxygen was greater. On the other hand, the basis of the positive correlation between cotyledon size and sensitivity to maternal hypoxia is less readily explained. It is unlikely to be the result of better maternal perfusion in the larger cotyledons, since analysis of interrelationships of perfusion parameters under control conditions has shown that maternal How rate (ml min- 1 gm- 1) correlated negatively with cotyledon Weight. 11 Again, it may be that function is better preserved at term in the larger cotyledons. Although we have demonstrated hypoxic fetoplacental vasoconstriction in an artificially perfused placental preparation in vitro, the response could be important in vivo, for example, in acute local blood How regulation in which it may result in a redistribution of fetoplacental blood How from hypoxic to normoxic or better perfused regions of the placenta. In this regard, the response may be analogous to the phenomenon of hypoxic pulmonary vasoconstriction, which is important in the adjustment of ventilation-perfusion ratios in the lung. We have shown that verapamil and lipoxygenase inhibitors antagonize hypoxic fetoplacental vasoconstriction, 15 which suggests that the response is mediated, at least in part, by release of lipoxygenase products.
1266
Howard, Hosokawa, and Maguire
Fetoplacental vasoconstriction in response to maternal hypoxia has not been previously described, although it is reminiscent of the reduction of fetal placental blood flow observed 24 hours after maternal placental occlusion in the near-term sheep. 1 However, we know of no evidence that the ovine response to occlusion of maternal placental flow occurs acutely and reversibly. An acute pressor response to abolition of maternal perfusion can be demonstrated in the in vitro placental preparation and is reversible on restoration of maternal perfusion. 16 Hypoxic fetoplacental vasoconstriction in the in vitro cotyledon resulted from acute change in maternal oxygen tension. It seems possible that in vivo, chronic placental hypoxia such as has been postulated to occur in preeclampsia3 • 1 may cause a chronic vasoconstrictor response that may contribute to fetal mortality and growth retardation associated with this disease. Furthermore, release of hypoxia-induced vasoconstrictive mediators may be important in the etiology of preeclampsia. We thank Dr. C. R. King and the staff of the delivery room of Bell Memorial Hospital for their help in obtaining the placentas, and Dr. J. Levy for performing data analysis.
November 1987 Am J Obstet Gynecol
5. 6. 7.
8.
9.
10. 11.
12.
13. 14.
REFERENCES l. Stock MK, Anderson DF, Phernetton TM, McLaughlin MK, Rankin JHG. Vascular response of the fetal placenta to local occlusion of the maternal placental vasculature. J Devel Physiol 1980;2:339-46. 2. RankinJHG, Goodman A, Phernetton T. Local regulation of uterine blood flow by the umbilical circulation. Proc Soc Exp Bioi Med 1975;150:690-4. 3. Chesley LC. Hypertensive disorders in pregnancy. New York: Appleton-Century-Crofts, 1978:455-66. 4. Las Heras J, Baskerville JC, Harding PGR, Haust MD.
15.
16.
Morphometric studies of fetal placental stem arteries in hypertensive disorders of pregnancy. Placenta 1985;6: 217-28. Schneider H, Panigel M, Dancis J. Transfer across the perfused human placenta of antipyrine, sodium and leucine. AM j 0BSTET GYNECOL 1972; 113:822-8. Hosokawa T, Howard RB, Maguire MH. Conversion of angiotensin I to angiotensin II in the human fetoplacental vascular bed. Br J Pharmacal 1985;84:237-41. Howard RB, Hosokawa T, Maguire MH. Pressor and depressor actions of prostanoids in the intact human fetoplacental vascular bed. Prostaglandins Leukotrienes Med 1986;21 :323-30. Maguire MH, Howard RB; Hosokawa T, Poisner AM. Effects of some autacoids and humoral agents on human fetoplacental vascular tesistance candidates for local regulation of fetoplacental blood flow. Trophoblast Res 1987 (In press). Dixon WJ. BMDP statistical software. Berkeley: University of California Press, 1983. Draper NR, Smith H. Applied regression analysis. New York: John Wiley & Sons, 1981:265-6. Page EW. The placenta and fetus; trophoblastic diseases. In: Benson RC, ed. Current obstetric and gynecologic diagnosis and treatment. Los Altos, California: Lange Medical Publications, 1978:533-59. Gill RW, Trudinger BJ, Garrett WJ, Kossoff G, Warren PS. Fetal umbilical venous flow measured in utero by pulsed doppler and B-mode ultrasound. I. Normal pregnancies. AM j 0BSTET GYNECOL 1981; 139:720-5. Longo L. Disorders of placental transfer. In: Assali NS, Brinkman CR, eds. Pathophysiology of gestation. New York: Academic Press, 1972:1-65. Howard RB, Levy J, Hosokawa T, Maguire MH. Interrelationships of perfusion parameters in the dualperfused human placental cotyledon. Trophoblast Res 1987;2:585-96. Howard RB, Hosokawa T, Maguire MH. Hypoxiainduced fetoplacental vasoconstriction in cotyledons of human term placentas: antagonism by selected drugs [Abstract]. Pharmacologist 1984;26: 144. Kitagawa H, Siegel P, Maguire MH. Ischemia-induced vasoconstriction and adenosine release in perfused human placental cotyledons [Abstract]. Pharmacologist 1987; 29:197.