- Email: [email protected]
Lymph flow rate response to angiotensin II is decreased in pregnant sheep
Lymph flow rate response to angiotensin II is decreased in pregnant sheep Guillermo J. Valenzuela, MD, and Lawrence D. Longo, MD Loma Linda, California Pregnancy in humans is associated with a number of physiologic changes including interstitial fluid retention (edema) and a decrease in the systemic vascular response to infused angiotensin II. In nonpregnant sheep angiotensin II increases the lymph flow rate by what appears to be a direct effect on the lymphatic vessels. The purpose of this study was to test the hypothesis that during pregnancy the lymph flow rate response to angiotensin II infusion is decreased in relation to that of the nonpregnant state. We speculate that a decrease in lymph flow may explain the interstitial fluid retention observed during human pregnancy. In nine nonpregnant and five pregnant chronically catheterized ewes, we infused angiotensin II at rates of 0.1,10, and 1000 ng/kg/min during a 5-minute period, with intervals of at least 15 minutes between doses. At the highest angiotensin II dose, peak lymph flow rate increased 286% in pregnant ewes compared with an increase of 344% in the nonpregnant sheep (p < 0.05). No changes occurred in the intravascular volume, plasma or lymph protein concentration, or venous pressure. The arterial pressure responses to angiotensin II were decreased in pregnant sheep (p < 0.05). These results are compatible with a model for fluid retention in pregnancy in which a decreased lymph flow rate plays a significant role in interstitial fluid retention. (AM J OBSTET GVNECOL 1989;161 :1615-9.)
Key words: Angiotensin II, fluid balance, pregnancy, lymph flow
Human pregnancy is associated with a number of cardiovascular changes, including increases of approximately 40% in blood volume, about 50% in cardiac output, and the retention of several liters of fluid in the interstitium.' The fluid distribution across the capillaries follows the balance between the hydrostatic and oncotic pressures on both sides of the capillaries as described by Starling'S law." The resulting forces are modified by the area of the surface of interchange and the protein permeability of the capillary walls. 3 Fluid and protein, representing -30% of the total protein mass in the intravascular system, are transferred into the interstitium and returned to the systemic circulation by the lymphatic system. Thus one may speculate that relatively small changes in the lymph flow rate could produce significant retention of protein and fluid, thus altering the forces that affect the interstitial space oncotic and interstitial fluid pressure. We have previously postulated that angiotensin II may have a direct effect on lymphatic flow in nonpregnant ewes; a concept supported by the findings of Dabney et al. 5 of the effects of angiotensin II in the dog's hind-paw lymphatics. From the Division of Perinatal Biology and the Department of Obstetnc) and Gynecology, Loma Linda Unlverstty. Presented at the Thlrty-fifth Annual Meeting of the SOCIety for Gynecologzc Investtgation, Baltimore, Maryland, March 17-20,
1988.
Reprint requests: GUlllenno J. Valenzuela, MD, Department of Obstetncs and Gynecology, San Bernardino County Medtcal Center, 780 E. Gilbert St., San Bernardino, CA 92404. 6/6116510
During pregnancy, several animal species, including sheep6 and man/ experience a decrease in systemic arterial pressure response to angiotensin II infusion, despite the fact that circulating levels of angiotensin II in the pregnant sheep during angiotensin II infusion are not statistically different from those in nonpregnant animals. s Therefore we tested the hypothesis that lymph flow response to angiotensin II infusion is decreased in pregnancy as compared with nonpregnant controls. Such a finding would suggest that during pregnancy the lymphatic system behaves in a manner similar to that of the systemic vascular system with a decreased responsiveness to vasoactive hormones. We hypothesize that if lympahtic contractility is decreased and if at the same time interstitial compliance is increased, more protein and fluid would remain within the interstitium and appear clinically as edema. Material and methods We used nine nonpregnant and five pregnant ewes, with body weights of 30 to 80 kg. The animals were obtained from a local breeder and housed in individual pens with artificial light (14 hours of light and 12 hours of dark) and ambient temperature of 22° C. The data on the nonpregnant animals have been described in detail elsewhere" Surgery was performed with the animals under general anesthesia as described in detail elsewhere! Briefly, anesthesia was induced with intravenous Biotal (500 mg) and maintained by inspiration of 1.5% halothane 1615
1616
Valenzuela and Longo
December 1989 Am J Obstet Gynecol
200
iiJ z .... ...J UJ Ul
180 •
III
a: UJ >
160
0
!! UJ
a:
140
::J
Ul Ul UJ
a: Q.
120
...J
....
a:
UJ I-
100
a:
0
115
TIME
(MIN)
Fig. 1. Mean ± SEM arterial pressure (expressed as percent over baseline) is presented for pregnant
(_ _ ) and nonpregnant (--) animals. Angiotensin II infusion is represented by interval between vertical dotted lmes; the doses used were 0.1,10, and 1000 ng / min / kg, respectively. The nonpregnant animals responded with higher arterial pressures (p < 0.05).
balanced with oxygen. Tygon catheters (inside diameter 1.27 mm) were placed through a neck incision in the thoracic duct, and one catheter was also placed in one of the internal jugular vein tributaries. The connection between these two vessels was made externally so as to allow the lymph to circulate undisturbed while allowing us easy access to the lymphatic vessels at the time of the experiments. We also catheterized a femoral artery and vein. Splenectomy was performed through a left flank incision. An incision was made 2 em parallel to the left rib cage, centered about the midaxillary line. Once the incision was made into the peritoneal cavity, the loose fascial planes attached to the splenic capsule were released by blunt dissection. The spleen was then delivered through the incision, stretching the pedicle containing the splenic artery and vein; this pedical was then dbubly clamped and tied with 1-0 nylon sutures. Occasionally other small vascular connections at the inferior pole were also ligated. Average blood loss from splenectomy was <50 ml, and the procedure added < I 0 minutes to the surgery. Postoperatively the animals received heparin, 30,000 IV intravenously daily and 15,000 IV subcutaneously twice a day. Experimental protocol. The experiments were conducted 4 to 14 days after surgery. We measured lymph flow rate by draining the lymph into a continuously weighed vial -30 em below the level of the thoracic duct. The rate was calculated every 10 seconds from the differential of the vial's weight by the use of an online program. We returned the lymph to the animal every 10 minutes, by infusing the vial's content into the jugular vein. Samples for hematocrit and plasma and lymph protein concentration were obtained during the control period and immediately before and after each
of the angiotensin II infusions. Protein concentrations were determined by a hand-held refractometer (American Optical, Buffalo). Arterial and venous pressures were measured continuously and averaged every 30 seconds; the values were stored for later analysis. After a control period of 30 minutes, we infused angiotensin II at a dose of 0.1 ng/kg animal weight per minute for 5 minutes into the jugular vein. Then after a 15-minute observation period, we infused a dose of 10 ng / kg / min for 5 minutes followed by an observation period of 30 minutes. We then administered a third angiotensin II infusion at a rate of 1000 ng/kg/min for 5 minutes. Finally, all functions were measured for 40 minutes. Data analysis. The data are presented as average ± SEM. To eliminate interanimal variability, the arterial pressure and the lymph flow rate are presented as percent over initial control. Group comparisons were performed by unpaired t test, and the changes with time for each variable were tested with a two-way analysis of variance (separated for pregnant and nonpregnant animals). The lymph flow rate responses were also analyzed by measuring the total area under the curve for pregnant and nonpregnant animals. We plotted the responses to angiotensin II infusion and cut out and weighed the area under the curve for both pregnant and nonpregnant ewes. Also we measured the amount oflymph collected during the first 10 minutes after the angiotensin II infusion for pregnant and nonpregnant animals. Results The values for pregnant and nonpregnant animals during the control period are presented in Table
Effects of angiotensin on lymph flow
Volume 161 !';umber 6, Part I
1617
500
i1) Z I-t
400
...J UJ Ul
'"
III
a:
UJ
300
> 0
!! :J: 0 ...J LL
200'
I
CL
::£
>-
100
...J
0
115
0 TIME
(MIN)
Fig. 2. Average ± SEM lymph flow rates (expressed as milliliters per minute over baseline) responses to angIotensin II are presented in nonpregnant (_ _ ) and pregnant (--) animals. Vertical dotted hnes represent angiotensIn II infusion doses of 0.1, 10. and 1000 ng/min/kg. There was no effect at lower doses, but at higher doses the response was signincantly lower for pregnant ewes (p < 0.05). 300 r I
i1) Z I-t
...J UJ Ul
'"
III
a: UJ
>
o
!! 3:
o
...J LL I
CL
::£
>-
...J
200
I! I
I
",~_ l_~'~ I
i
NONPREGNANT
A II
PREGNANT
(1000 NG/KG/MIN)
EFFECT ON LYMPH FLOW
Fig. 3. Excess of lymph flow In response to the higher angiotensin II infusion dose (expressed as % over baseline) for pregnant and nonpregnant animals. The difference is significantly lower for pregnant animal (p < 0.05).
1. During the control period the artenal pressure
was significantly higher for the nonpregnant ewes (P < 0.05). There were no statistical differences between the groups for lymph flow rate, venous pressure, hematocrit, plasma, or lymph protein concentration. Fig. 1 presents the percent change of arterial pressure over baseline in association with angiotensin II infusion for both pregnant and nonpregnant ewes; there was a statistically higher pressor response at all angiotensin II doses in the nonpregnant animals (P < 0.05). Fig. 2 presents the lymph flow response for both groups expressed as the percent change over baseline. There was no change in the lymph flow rate at
the two lower doses of angiotensin II even though arterial pressure was elevated in the nonpregnant animals. In absolute terms, after the maximum angiotensin II dose, lymph flow rate increased from a baseline of 2.46 ± 0.69 to 2.98 ± 0.7 mil min in pregnant animals, as compared with the change from baseline of 2.93 ± 0.15 to 5.07 ± 0.57 ml!min for the nonpregnant group (p < 0.05). The ratio of the total area under the curve for nonpregnant! pregnant response was 1.65 (p < 0.05). The excess lymph flow during the first 10 minutes after angiotensin II infusion at the highest dose was also greater for the nonpregnant animals (Fig. 3) (p < 0.05).
1618 Valenzuela and Longo
December 1989 Am J Obstet Gynecol
Table I. Values during 3D-minute control period Nonpregnant Pregnant (mean ± SEM) (mean ± SEM)
Arterial pressure (mm Hg) Venous pressure (mmHg) Lymph flow (mil min) Hematocrit (%) Plasma protein (gm/dl) Lymph protein (gm/dl)
P Value
lOO.3 ± 3.7
84 ± 4.2
<0.05
7.9 ± 1.1
9.1 ± 0.7
NS
2.93 ± 0.15
2.46 ± 0.69
NS
28.84 ± 1.17 6.3 ± 0.8
26.24 ± 3.66 6.8 ± 0.2
NS NS
4.4 ± 0.5
5.1 ± 0.6
NS
There were no changes with time for venous pressure, hematocrit, plasma protein, or lymph protein in association with angiotensin II infusion for the pregnant or nonpregnant ewes (p > 0.1).
Comment An understanding of fluid dynamics during pregnancy is important because appropriate fluid retention is associated with improved perinatal outcome. 9 • 10 Furthermore, many clinical conditions such as preterm labor, heart disease, kidney disease, and preeclampsia are characterized by alterations of the total amount of fluid retained or its distribution. For instance, preterm labor is characterized by an expansion of blood volume less than that of the normal pregnant women who is delivered at term. In contrast, preeclampsia is characterized by a total fluid retention within normal limits, but it is distributed chiefly to the interstitium when the disease is more pronounced. 10 The present finding of a relatively decreased lymph flow in response to angiotensin II infusion in pregnant animals as compared with nonpregnant animals is relevant to the problem of fluid dynamics during pregnancy. This is because it is compatible with pregnancyassociated hyporesponsiveness of the lymphatics to pressor hormones as a possible mechanism for interstitial fluid retention. For instance, a 5% lower fluid transport by the lymphatics may result in an interstitial fluid retention of -100 mil day; thus in a few days it could produce the retention of a significant amount of fluid and protein. One might argue that the increased lymph flow resulted from the increase in arterial pressure observed with the angiotensin II infusion. However, angiotensin II acts on the smaller arterioles and does not affect veins,11 so we would not expect an effect on the capillary hydrostatic pressure. The lack of venous pressure change can be used as indirect evidence that the hydrostatic capillary pressure was not significantly changed. If the origin of the increase in lymph flow were the intravascular fluid, we would have ex-
pected a change in the micro hematocrit in splenectomized animals. Furthermore, in experiments in which a greater transfer of fluid is obtained by increasing the local blood flow, there is a delay before the increase in lymph flow is apparent. 12 In addition, there were no changes in interstitial fluid pressure (as reflected by a lack of changes in venous pressure) or in the plasma or lymph protein concentrations (therefore the oncotic pressures remained stable). Thus with these direct and indirect measurements all parameters suggest that the increased lymph flow did not originate from the systemic circulation. Another possibility is that the excess lymph originated by mobilization of interstitial fluid; however, the lack of change in venous pressure strongly suggests that interstitial fluid pressure remained stable during the angiotensin II infusion. We speculate that the decreased response to angiotensin II seen in pregnancy results in a decrease in lymph flow. Although the absolute values for lymph flow (expressed as milliliters per minute per kilogram) are not different for pregnant and nonpregnant animals in these results or in our previous study,13 if a correction factor for the placenta, fetus, and amniotic fluid weights is used, the pregnant animals display a 20% lower lymph flow rate. The mechanism may involve a down-regulation of angiotensin II receptors similar to that thought to occur during pregnancy in rabbits. 14 Another possible mechanism could be the release and/or synthesis of compounds with vasodilator capability produced by angiotensin II infusion during pregnancy; it is well known that blocking prostaglandin production produces a disappearance of the angiotensin II resistance of pregnancy. 15 Mechanisms other than alterations in the lymph flow rate can be postulated to explain edema formation during pregnancy. For instance, estrogen increases capillary permeability, thus augmenting the fluid and protein transferred across the capillary wall. The administration of fluid also produces an increase in the capillary passage of fluid; however, the steady state is maintained by a severalfold increase in the lymph flow rate, resulting in no net fluid retention in the interstitium. 2 For fluid and protein to be retained in the interstitium, it is also necessary that there be changes in the interstitial fluid space compliance. In this manner the interstitial space may contain more fluid without increasing the pressure. Otherwise, the elevation in interstitial fluid pressure would produce an increase in the lymph flow rate. Estrogen infusion produces fluid retention, mainly in the interstitium,16 and the decrease of physiologic concentration of estrogen by castration in sheep decreases interstitial fluid space compliance. 17 The fluid and protein retention occur until a critical interstitial fluid volume is reached beyond which the interstitial pressure increases at the same rate as that
Effects of angiotensin on lymph flow
Volume 161 Number 6, Part 1
of a nonpregnant animal. At this time the lymph flow rate and the response to fluid infusion return to normal, as demonstrated by us in the pregnant sheep.13 In conclusion, pregnancy in sheep is associated with a decreased lymph flow rate response to angiotensin II infusion, a phenomenon similar to the response observed in the systemic circulation. This hyporesponsiveness may playa role in the fluid retention of pregnancy.
8.
9. 10. 11.
REFERENCES 1. Metcalfe], Stock MK, Barron DH. Maternal physiology during gestation. In: Knobil E, Neill], eds. Physiology of reproduction. New York: Raven Press, 1988:2145-76. 2. Guyton AC, ed. Textbook of medical physiology. 6th ed. Philadelphia: WB Saunders, 1981 :370-82. 3. Renkin EM. Some consequences of capillary permeability to macromolecules: Starling'S hypothesis reconsidered. Am] Physiol 1986;250:H706-1O. 4. Valenzuela G], Hewitt CW, Graham AD. A II infusion increases thoracic duct lymph flow in chronically catheterized sheep. Am] PhysioI1987;252:R853-8. 5. Dabney]M, Buehn M], Dobbins DE. Constriction oflymphatics by catecholamines, carotid occlusion, or hemorrhage. Am] Physiol 1988;255(Heart Circ Physiol 24): H514-24. 6. Rosenfeld CR, Gant NF. The chronically instrumented ewe: a model for studying vascular reactivity to A II in pregnancy.] Clin Invest 1981;67:486-92. 7. Chesley LC, Talledo E, Bohler CS, Zuspan FP. Vascular reactivity to A II and norepinephrine in pregnant and
12. 13. 14. 15.
16.
17.
1619
nonpregnant women. AM] OBSTET GYNECOL 1965;91: 837-42. Naden RP, Coultrup S, Arant BS, Rosenfeld CR. Me'tabolic clearance of angiotensin II in pregnant and nonp~egnant sheep. Am] Physiol 1985;249(Endocrinol Metab 12):E49-55. Longo LD, Hardesty ]S. Maternal blood volume: measurements, hypothesis of control, and clinical considerations. Rev Perinat Med 1984;5:35-59. Goodliri RC, Quaife MA. The significance, diagnosis and treatment of maternal hypovolemia as associated with fetall maternal illness. Semin Perinatol 1981;5: 163-74. Flemming ]T, Harris PD, Joshua IG. Endogenous prostaglandins selectively mask large arteriole constriction to angiotensin II. Am] PhysioI1987;253:HI573-80. Lewis GP, Winsey N]P. The action of pharmacologically active substances on the flow and composition of cat hind limb lymph. Br] Pharmacol 1970;40:446-60. Valenzuela G], Wood LL, Brace RA. Thoracic duct lymph flow in the pregnant sheep and responses to blood volume expansion. Am] Physiol 1986;250:RI095-8. Brown GP, Venuto RC. Angiotensin II receptor alterations during pregnancy in rabbits. Am] Physiol 1986; 251 :E58-65. Everett RB, Worley R], MacDonald PC, Cant NF. Effect of prostaglandin synthetase inhibitors on pressor response to angiotensin II in human pregnancy. ] Clin Endocrinol Metab 1978;46:1007-10. Ueda S, Fortune V, Bull BS, Valenzuela G], Longo LD. Estrogen effects on plasma volume, arterial blood pressure, interstitial space, plasma proteins, and blood viscosity in sheep. AM] OBSTET GYNECOL 1986;155:195-201. Valenzuela G], Brace RA, Longo LD. Lymphatic and vascular responses to fluid infusion in castrated and noncastrated sheep. Am] PhysioI1987;252:RllI4-8.
Key words: Angiotensin II, fluid balance, pregnancy, lymph flow
Human pregnancy is associated with a number of cardiovascular changes, including increases of approximately 40% in blood volume, about 50% in cardiac output, and the retention of several liters of fluid in the interstitium.' The fluid distribution across the capillaries follows the balance between the hydrostatic and oncotic pressures on both sides of the capillaries as described by Starling'S law." The resulting forces are modified by the area of the surface of interchange and the protein permeability of the capillary walls. 3 Fluid and protein, representing -30% of the total protein mass in the intravascular system, are transferred into the interstitium and returned to the systemic circulation by the lymphatic system. Thus one may speculate that relatively small changes in the lymph flow rate could produce significant retention of protein and fluid, thus altering the forces that affect the interstitial space oncotic and interstitial fluid pressure. We have previously postulated that angiotensin II may have a direct effect on lymphatic flow in nonpregnant ewes; a concept supported by the findings of Dabney et al. 5 of the effects of angiotensin II in the dog's hind-paw lymphatics. From the Division of Perinatal Biology and the Department of Obstetnc) and Gynecology, Loma Linda Unlverstty. Presented at the Thlrty-fifth Annual Meeting of the SOCIety for Gynecologzc Investtgation, Baltimore, Maryland, March 17-20,
1988.
Reprint requests: GUlllenno J. Valenzuela, MD, Department of Obstetncs and Gynecology, San Bernardino County Medtcal Center, 780 E. Gilbert St., San Bernardino, CA 92404. 6/6116510
During pregnancy, several animal species, including sheep6 and man/ experience a decrease in systemic arterial pressure response to angiotensin II infusion, despite the fact that circulating levels of angiotensin II in the pregnant sheep during angiotensin II infusion are not statistically different from those in nonpregnant animals. s Therefore we tested the hypothesis that lymph flow response to angiotensin II infusion is decreased in pregnancy as compared with nonpregnant controls. Such a finding would suggest that during pregnancy the lymphatic system behaves in a manner similar to that of the systemic vascular system with a decreased responsiveness to vasoactive hormones. We hypothesize that if lympahtic contractility is decreased and if at the same time interstitial compliance is increased, more protein and fluid would remain within the interstitium and appear clinically as edema. Material and methods We used nine nonpregnant and five pregnant ewes, with body weights of 30 to 80 kg. The animals were obtained from a local breeder and housed in individual pens with artificial light (14 hours of light and 12 hours of dark) and ambient temperature of 22° C. The data on the nonpregnant animals have been described in detail elsewhere" Surgery was performed with the animals under general anesthesia as described in detail elsewhere! Briefly, anesthesia was induced with intravenous Biotal (500 mg) and maintained by inspiration of 1.5% halothane 1615
1616
Valenzuela and Longo
December 1989 Am J Obstet Gynecol
200
iiJ z .... ...J UJ Ul
180 •
III
a: UJ >
160
0
!! UJ
a:
140
::J
Ul Ul UJ
a: Q.
120
...J
....
a:
UJ I-
100
a:
0
115
TIME
(MIN)
Fig. 1. Mean ± SEM arterial pressure (expressed as percent over baseline) is presented for pregnant
(_ _ ) and nonpregnant (--) animals. Angiotensin II infusion is represented by interval between vertical dotted lmes; the doses used were 0.1,10, and 1000 ng / min / kg, respectively. The nonpregnant animals responded with higher arterial pressures (p < 0.05).
balanced with oxygen. Tygon catheters (inside diameter 1.27 mm) were placed through a neck incision in the thoracic duct, and one catheter was also placed in one of the internal jugular vein tributaries. The connection between these two vessels was made externally so as to allow the lymph to circulate undisturbed while allowing us easy access to the lymphatic vessels at the time of the experiments. We also catheterized a femoral artery and vein. Splenectomy was performed through a left flank incision. An incision was made 2 em parallel to the left rib cage, centered about the midaxillary line. Once the incision was made into the peritoneal cavity, the loose fascial planes attached to the splenic capsule were released by blunt dissection. The spleen was then delivered through the incision, stretching the pedicle containing the splenic artery and vein; this pedical was then dbubly clamped and tied with 1-0 nylon sutures. Occasionally other small vascular connections at the inferior pole were also ligated. Average blood loss from splenectomy was <50 ml, and the procedure added < I 0 minutes to the surgery. Postoperatively the animals received heparin, 30,000 IV intravenously daily and 15,000 IV subcutaneously twice a day. Experimental protocol. The experiments were conducted 4 to 14 days after surgery. We measured lymph flow rate by draining the lymph into a continuously weighed vial -30 em below the level of the thoracic duct. The rate was calculated every 10 seconds from the differential of the vial's weight by the use of an online program. We returned the lymph to the animal every 10 minutes, by infusing the vial's content into the jugular vein. Samples for hematocrit and plasma and lymph protein concentration were obtained during the control period and immediately before and after each
of the angiotensin II infusions. Protein concentrations were determined by a hand-held refractometer (American Optical, Buffalo). Arterial and venous pressures were measured continuously and averaged every 30 seconds; the values were stored for later analysis. After a control period of 30 minutes, we infused angiotensin II at a dose of 0.1 ng/kg animal weight per minute for 5 minutes into the jugular vein. Then after a 15-minute observation period, we infused a dose of 10 ng / kg / min for 5 minutes followed by an observation period of 30 minutes. We then administered a third angiotensin II infusion at a rate of 1000 ng/kg/min for 5 minutes. Finally, all functions were measured for 40 minutes. Data analysis. The data are presented as average ± SEM. To eliminate interanimal variability, the arterial pressure and the lymph flow rate are presented as percent over initial control. Group comparisons were performed by unpaired t test, and the changes with time for each variable were tested with a two-way analysis of variance (separated for pregnant and nonpregnant animals). The lymph flow rate responses were also analyzed by measuring the total area under the curve for pregnant and nonpregnant animals. We plotted the responses to angiotensin II infusion and cut out and weighed the area under the curve for both pregnant and nonpregnant ewes. Also we measured the amount oflymph collected during the first 10 minutes after the angiotensin II infusion for pregnant and nonpregnant animals. Results The values for pregnant and nonpregnant animals during the control period are presented in Table
Effects of angiotensin on lymph flow
Volume 161 !';umber 6, Part I
1617
500
i1) Z I-t
400
...J UJ Ul
'"
III
a:
UJ
300
> 0
!! :J: 0 ...J LL
200'
I
CL
::£
>-
100
...J
0
115
0 TIME
(MIN)
Fig. 2. Average ± SEM lymph flow rates (expressed as milliliters per minute over baseline) responses to angIotensin II are presented in nonpregnant (_ _ ) and pregnant (--) animals. Vertical dotted hnes represent angiotensIn II infusion doses of 0.1, 10. and 1000 ng/min/kg. There was no effect at lower doses, but at higher doses the response was signincantly lower for pregnant ewes (p < 0.05). 300 r I
i1) Z I-t
...J UJ Ul
'"
III
a: UJ
>
o
!! 3:
o
...J LL I
CL
::£
>-
...J
200
I! I
I
",~_ l_~'~ I
i
NONPREGNANT
A II
PREGNANT
(1000 NG/KG/MIN)
EFFECT ON LYMPH FLOW
Fig. 3. Excess of lymph flow In response to the higher angiotensin II infusion dose (expressed as % over baseline) for pregnant and nonpregnant animals. The difference is significantly lower for pregnant animal (p < 0.05).
1. During the control period the artenal pressure
was significantly higher for the nonpregnant ewes (P < 0.05). There were no statistical differences between the groups for lymph flow rate, venous pressure, hematocrit, plasma, or lymph protein concentration. Fig. 1 presents the percent change of arterial pressure over baseline in association with angiotensin II infusion for both pregnant and nonpregnant ewes; there was a statistically higher pressor response at all angiotensin II doses in the nonpregnant animals (P < 0.05). Fig. 2 presents the lymph flow response for both groups expressed as the percent change over baseline. There was no change in the lymph flow rate at
the two lower doses of angiotensin II even though arterial pressure was elevated in the nonpregnant animals. In absolute terms, after the maximum angiotensin II dose, lymph flow rate increased from a baseline of 2.46 ± 0.69 to 2.98 ± 0.7 mil min in pregnant animals, as compared with the change from baseline of 2.93 ± 0.15 to 5.07 ± 0.57 ml!min for the nonpregnant group (p < 0.05). The ratio of the total area under the curve for nonpregnant! pregnant response was 1.65 (p < 0.05). The excess lymph flow during the first 10 minutes after angiotensin II infusion at the highest dose was also greater for the nonpregnant animals (Fig. 3) (p < 0.05).
1618 Valenzuela and Longo
December 1989 Am J Obstet Gynecol
Table I. Values during 3D-minute control period Nonpregnant Pregnant (mean ± SEM) (mean ± SEM)
Arterial pressure (mm Hg) Venous pressure (mmHg) Lymph flow (mil min) Hematocrit (%) Plasma protein (gm/dl) Lymph protein (gm/dl)
P Value
lOO.3 ± 3.7
84 ± 4.2
<0.05
7.9 ± 1.1
9.1 ± 0.7
NS
2.93 ± 0.15
2.46 ± 0.69
NS
28.84 ± 1.17 6.3 ± 0.8
26.24 ± 3.66 6.8 ± 0.2
NS NS
4.4 ± 0.5
5.1 ± 0.6
NS
There were no changes with time for venous pressure, hematocrit, plasma protein, or lymph protein in association with angiotensin II infusion for the pregnant or nonpregnant ewes (p > 0.1).
Comment An understanding of fluid dynamics during pregnancy is important because appropriate fluid retention is associated with improved perinatal outcome. 9 • 10 Furthermore, many clinical conditions such as preterm labor, heart disease, kidney disease, and preeclampsia are characterized by alterations of the total amount of fluid retained or its distribution. For instance, preterm labor is characterized by an expansion of blood volume less than that of the normal pregnant women who is delivered at term. In contrast, preeclampsia is characterized by a total fluid retention within normal limits, but it is distributed chiefly to the interstitium when the disease is more pronounced. 10 The present finding of a relatively decreased lymph flow in response to angiotensin II infusion in pregnant animals as compared with nonpregnant animals is relevant to the problem of fluid dynamics during pregnancy. This is because it is compatible with pregnancyassociated hyporesponsiveness of the lymphatics to pressor hormones as a possible mechanism for interstitial fluid retention. For instance, a 5% lower fluid transport by the lymphatics may result in an interstitial fluid retention of -100 mil day; thus in a few days it could produce the retention of a significant amount of fluid and protein. One might argue that the increased lymph flow resulted from the increase in arterial pressure observed with the angiotensin II infusion. However, angiotensin II acts on the smaller arterioles and does not affect veins,11 so we would not expect an effect on the capillary hydrostatic pressure. The lack of venous pressure change can be used as indirect evidence that the hydrostatic capillary pressure was not significantly changed. If the origin of the increase in lymph flow were the intravascular fluid, we would have ex-
pected a change in the micro hematocrit in splenectomized animals. Furthermore, in experiments in which a greater transfer of fluid is obtained by increasing the local blood flow, there is a delay before the increase in lymph flow is apparent. 12 In addition, there were no changes in interstitial fluid pressure (as reflected by a lack of changes in venous pressure) or in the plasma or lymph protein concentrations (therefore the oncotic pressures remained stable). Thus with these direct and indirect measurements all parameters suggest that the increased lymph flow did not originate from the systemic circulation. Another possibility is that the excess lymph originated by mobilization of interstitial fluid; however, the lack of change in venous pressure strongly suggests that interstitial fluid pressure remained stable during the angiotensin II infusion. We speculate that the decreased response to angiotensin II seen in pregnancy results in a decrease in lymph flow. Although the absolute values for lymph flow (expressed as milliliters per minute per kilogram) are not different for pregnant and nonpregnant animals in these results or in our previous study,13 if a correction factor for the placenta, fetus, and amniotic fluid weights is used, the pregnant animals display a 20% lower lymph flow rate. The mechanism may involve a down-regulation of angiotensin II receptors similar to that thought to occur during pregnancy in rabbits. 14 Another possible mechanism could be the release and/or synthesis of compounds with vasodilator capability produced by angiotensin II infusion during pregnancy; it is well known that blocking prostaglandin production produces a disappearance of the angiotensin II resistance of pregnancy. 15 Mechanisms other than alterations in the lymph flow rate can be postulated to explain edema formation during pregnancy. For instance, estrogen increases capillary permeability, thus augmenting the fluid and protein transferred across the capillary wall. The administration of fluid also produces an increase in the capillary passage of fluid; however, the steady state is maintained by a severalfold increase in the lymph flow rate, resulting in no net fluid retention in the interstitium. 2 For fluid and protein to be retained in the interstitium, it is also necessary that there be changes in the interstitial fluid space compliance. In this manner the interstitial space may contain more fluid without increasing the pressure. Otherwise, the elevation in interstitial fluid pressure would produce an increase in the lymph flow rate. Estrogen infusion produces fluid retention, mainly in the interstitium,16 and the decrease of physiologic concentration of estrogen by castration in sheep decreases interstitial fluid space compliance. 17 The fluid and protein retention occur until a critical interstitial fluid volume is reached beyond which the interstitial pressure increases at the same rate as that
Effects of angiotensin on lymph flow
Volume 161 Number 6, Part 1
of a nonpregnant animal. At this time the lymph flow rate and the response to fluid infusion return to normal, as demonstrated by us in the pregnant sheep.13 In conclusion, pregnancy in sheep is associated with a decreased lymph flow rate response to angiotensin II infusion, a phenomenon similar to the response observed in the systemic circulation. This hyporesponsiveness may playa role in the fluid retention of pregnancy.
8.
9. 10. 11.
REFERENCES 1. Metcalfe], Stock MK, Barron DH. Maternal physiology during gestation. In: Knobil E, Neill], eds. Physiology of reproduction. New York: Raven Press, 1988:2145-76. 2. Guyton AC, ed. Textbook of medical physiology. 6th ed. Philadelphia: WB Saunders, 1981 :370-82. 3. Renkin EM. Some consequences of capillary permeability to macromolecules: Starling'S hypothesis reconsidered. Am] Physiol 1986;250:H706-1O. 4. Valenzuela G], Hewitt CW, Graham AD. A II infusion increases thoracic duct lymph flow in chronically catheterized sheep. Am] PhysioI1987;252:R853-8. 5. Dabney]M, Buehn M], Dobbins DE. Constriction oflymphatics by catecholamines, carotid occlusion, or hemorrhage. Am] Physiol 1988;255(Heart Circ Physiol 24): H514-24. 6. Rosenfeld CR, Gant NF. The chronically instrumented ewe: a model for studying vascular reactivity to A II in pregnancy.] Clin Invest 1981;67:486-92. 7. Chesley LC, Talledo E, Bohler CS, Zuspan FP. Vascular reactivity to A II and norepinephrine in pregnant and
12. 13. 14. 15.
16.
17.
1619
nonpregnant women. AM] OBSTET GYNECOL 1965;91: 837-42. Naden RP, Coultrup S, Arant BS, Rosenfeld CR. Me'tabolic clearance of angiotensin II in pregnant and nonp~egnant sheep. Am] Physiol 1985;249(Endocrinol Metab 12):E49-55. Longo LD, Hardesty ]S. Maternal blood volume: measurements, hypothesis of control, and clinical considerations. Rev Perinat Med 1984;5:35-59. Goodliri RC, Quaife MA. The significance, diagnosis and treatment of maternal hypovolemia as associated with fetall maternal illness. Semin Perinatol 1981;5: 163-74. Flemming ]T, Harris PD, Joshua IG. Endogenous prostaglandins selectively mask large arteriole constriction to angiotensin II. Am] PhysioI1987;253:HI573-80. Lewis GP, Winsey N]P. The action of pharmacologically active substances on the flow and composition of cat hind limb lymph. Br] Pharmacol 1970;40:446-60. Valenzuela G], Wood LL, Brace RA. Thoracic duct lymph flow in the pregnant sheep and responses to blood volume expansion. Am] Physiol 1986;250:RI095-8. Brown GP, Venuto RC. Angiotensin II receptor alterations during pregnancy in rabbits. Am] Physiol 1986; 251 :E58-65. Everett RB, Worley R], MacDonald PC, Cant NF. Effect of prostaglandin synthetase inhibitors on pressor response to angiotensin II in human pregnancy. ] Clin Endocrinol Metab 1978;46:1007-10. Ueda S, Fortune V, Bull BS, Valenzuela G], Longo LD. Estrogen effects on plasma volume, arterial blood pressure, interstitial space, plasma proteins, and blood viscosity in sheep. AM] OBSTET GYNECOL 1986;155:195-201. Valenzuela G], Brace RA, Longo LD. Lymphatic and vascular responses to fluid infusion in castrated and noncastrated sheep. Am] PhysioI1987;252:RllI4-8.