The Determinants of Renal Hemodynamics in Pregnancy

The Determinants of Renal Hemodynamics in Pregnancy Christine Baylis, PhD • Studies in the rat have permitted a complete characterization of the glomerular hemodynamic changes that occur during a normal pregnancy. The rise in glomerular filtration rate (GFR) is the result of an evenly distributed increase in plasma flow to all nephrons due to vasodilation of both pre- and postglomerular resistance vessels. Pregnancy is not associated with any change in blood pressure in the glomerular capillaries; neither is there any detectable alteration in the glomerular water permeability or filtration surface area. Despite the chronically maintained renal vasodilation of pregnancy, gravid rats exhibit substantial renal reserve when challenged with an amino acid load, indicating that a residual renal vasodilatory capacity exists in the kidney of the normal pregnant animal. Despite the concomitant plasma volume expansion of pregnancy, the tubuloglomerular feedback system (a volume regulatory system that modulates GFR) remains fully active in the pregnant rat, suggesting that the plasma volume in pregnancy is not sensed as expanded by this system. The factor(s) that initiates the gestational rise in GFR is currently unknown, although since similar renal hemodynamic changes occur in the pseudopregnant rat, the stimulus is maternal rather than fetoplacental in origin. © 1987 by the National Kidney Foundation, Inc. INDEX WORDS: Glomerular filtration rate; pregnancy; renal reserve; tubuloglomerular feedback; pseudopregnancy.

I

N THE NORMAL pregnant woman, increases in glomerular filtration rate (GFR) , renal plasma flow rate (RPF), and plasma volume occur routinely, although the mechanism(s) that causes these pregnancy-related hemodynamic changes is currently unknown (Davison, pp 248-252). As indicated by Conrad (pp 253-259), the rat has proved an excellent animal model in which to study these gestational changes since increases in GFR, RPF, and plasma volume occur in the gravid rat, albeit over a greatly compressed time period (gestation in the rat lasts - 22 days). The study of renal hemodynamic changes during pregnancy in the rat requires preservation of preexisting differences in plasma volume in pregnant as compared with virgin control animals; this can be difficult under conditions of acute surgery and anesthesia. In general, however, the gestational rise in GFR and RPF is preserved in rats studied in the anesthetized, euvolemic (plasma restored) condition when prepared for micropuncture, and is quite similar to that which occurs in normovo-

From the Division of Nephrology, Department of Medicine, University of California, San Diego, La lolla. Supported by Grant No. HL31933 from the National1nstitutes of Health, Bethesda, MD. Address reprint requests to Christine Baylis, PhD, Division of Nephrology, Department of Medicine, University of California, San Diego, La lolla, CA 92093. © 1987 by the National Kidney Foundation, Inc. 0272-6386/87/0904-0004$3.00/0

260

lemic, awake, unstressed pregnant rats (Conrad, pp 253-259). This article deals mainly with information obtained using micropuncture techniques in the anesthetized, euvolemic Munich-Wistar rat, from studies designed to investigate the characteristics of the renal hemodynamic changes of pregnancy and to determine the underlying mechanisms involved in the initiation and maintenance of these changes. RENAL HEMODYNAMIC DETERMINANTS OF THE GESTATIONAL RISE IN GFR

The single nephron (SN)GFR, and therefore also the GFR, is controlled by four determinants: (1) increasing the glomerular plasma flow rate, QA will increase SNGFR; (2) the hydrostatic pressure gradient acting across the wall of the glomerulus, ~P (the difference between glomerular capillary blood pressure, PGe , and hydrostatic pressure of the filtrate in Bowman's space) favors formation of filtrate; (3) the oncotic pressure of the plasma arriving at the glomerulus (due to the presence of the plasma proteins), ITA tends to oppose filtration; and (4) the glomerular capillary ultrafiltration coefficient, Kf (the product of filtration surface area and glomerular capillary water permeability) also directly influences SNGFR under certain conditions. The normal euvolemic rat is in a state referred to as filtration pressure equilibrium, in which filtration has ceased before the end of the glomerular capillary.l In this case, SNGFR is highly dependent on plasma flow rate (QA) and is

American Journal of Kidney Diseases, Vol IX, No 4 (April), 1987: pp 260-264

RENAL HEMODYNAMIC DETERMINANTS IN PREGNANCY

• •

30

•

c

o

90 .5

E .....

c:

~

'E ...... c a:-25

III

80 III E

u.

III

Q.

CJ

z en

20

40

h

CI

J:30 E E

•

~20 :J en en ~

a. 10

o

~

____

L -_ _ _ _ _ _ _ _~_ _~

Virgin

12d pregnant

Fig 1. Single nephron glomerular filtration rate (SNGFR) and some of its determinants in the virgin and midterm 12-day-pregnant euvolemic MunichWistar rat. Upper panel: SNGFR (0 ); glomerular plasma flow rate (.). Lower panel: mean glomerular hydrostatic pressure gradient (.
also affected by changes in the hydrostatic pressure gradient (..:lP) and systemic oncotic pressure, ITA; however, provided that Kr remains above a certain "threshold " value, further increases in this variable will not further increase the value of SNGFR . l All these dynamic determinants of SNGFR can be either directly measured or calculated from measurements made in micropuncture studies. Micropuncture experiments in the euvolemic midterm pregnant (day 12) Munich-Wistar rat have indicated that the substantial increase in superficial cortical SNGFR, shown in the upper panel of Fig 1, results exclusively from the parallel increases in glomerular plasma flow rate. 2 As

261

shown in Fig 1 (lower panel), neither the glomerular hydrostatic pressure gradient, ..:lp, nor the systemic oncotic pressure, ITA' was changed at day 12 of pregnancy, and the gravid rats remained at filtration pressure equilibrium, as indicated by the equality between .:lP and the postglomerular arteriolar oncotic pressure, ITE (Fig 1, lower panel). Although not shown, arterial blood pressure (AP) was unaltered at this stage of pregnancy which, together with the constancy of glomerular capillary blood pressure, indicated that equivalent or proportional dilation occurred in both the afferent and efferent arteriolar resistance vessels. Because 12-day-pregnant rats remained at filtration pressure equilibrium, an exact value of Kf could not be calculated 1; however, the estimated minimum value was similar in virgin and 12-day-pregnant animals 2 and small changes in the "true" value of Kf would have no effect on SNGFR in the presence of filtration pressure equilibrium . 1 In these same studies, whole kidney GFR was also measured and the increases noted in the 12-day-pregnant rat were in proportion to the increases in SNGFR, implying an evenly distributed increase in filtration to all glomeruli. In addition , the increase in RPF was proportional to the increased QA' which suggests an evenly distributed renal vasodilation throughout the kidney. 2 Several other micropuncture studies in the euvolemic pregnant rat at earlier and late stages of gestation have indicated that the gestational rise in GFR is evident quite early, at between six and nine days of pregnancy, and declines toward the nonpregnant value close to term 2.3; this agrees with much of the other published literature in the rat and pregnant women (Conrad, and Davison, pp 248-252). One significant finding from these micropuncture studies is that no sustained rise occurs in glomerular capillary blood pressure during gestational " hyperfiltraton ," ie, during the pregnancy-related rise in GFR .2.3 This contrasts with observations reported in one study, in the hydropenic (probably somewhat volume-contracted) and saline volume-expanded 15- to 16-day-pregnant Munich-Wi star rat in which significant increases in glomerular blood pressure and thus ..:lP were reported in pregnancy, while all rats were also at filtration pressure disequilibrium with Kr being lower in the pregnant than in the virgin state. 4 The alterations in ..:lP and Kr offset each other and the increased SNGFR must therefore have resulted

262

from the marked increases in QA also reported in this study.4 Although no euvolemic animals have been studied at this stage of gestation, the constancy of PGC and the normal values of Kf in 6-, 9-, 12-, and 19-day-pregnant euvolemic animals 2 .3 suggest that the low K f and high ~P reported in these studies by Dal Canton et al 4 was the result of some aspect of the experimental preparation rather than pregnancy per se. CHARACTERISTICS OF THE ALTERED RENAL HEMODYNAMICS OF PREGNANCY

Renal Reserve in Pregnancy: Response to Amino Acid Infusion As discussed, the renal vasculature in gravid rats is chronically vasodilated as compared with that of the virgin animal. Amino acid infusion produces a pure renal vasodilation, resulting in proportional increases in SNGFR and QA without substantial alterations in the other determinants of glomerular filtrationS and may therefore be used to assess the remaining vasodilatory capacity of the kidney or "the renal reserve; the ability of the kidney to further raise its GFR and RPF in response to an additional vasodilator stimulus." Preliminary studies have shown that in the already vasodialated, midterm pregnant rat, exhibiting the maximal gestational rise in GFR and SNGFR, administration of an amino acid infusion leads to further marked increases in SNGFR and QA' due to proportional reductions in afferent and efferent arteriolar resistances. 6 The percentage of increase in GFR was similar in the relatively vasoconstricted virgin and vasodilated pregnant rat with similar amino acid loads, although the more vasodilated kidney might be predicted to have already exhausted some of its renal reserve. Thus, these studies indicate that despite the chronic renal vasodilation of pregnancy, additional, vasodilatory stimuli can be superimposed on the kidneys of gravid animals . 6 In one preliminary report in the normal pregnant woman, a meal of meat protein, given in the third trimester, evoked a substantial acute rise in the endogenous creatinine clearance that was superimposed on the chronic gestational rise in GFR.7 These findings are in agreement with observations in the rat indicating a residual renal reserve in normal pregnancy, although they await confirmation using inulin as a marker of GFR. (In this respect, preliminary observations of unpublished data by Beilin, W M .Barron, and M.

CHRISTINE BAYLIS

D. Lindheimer, noted at this conference, that inulin and p-aminohippurate clearances rise marginally, if at all, is of interest and we await completion of the study.) The converse observation, demonstrating that the gestational rise in GFR can be superimposed on states of chronic, underlying renal vasodilation is discussed by Davison in his article on pregnancy in renal transplant recipients (pp 374-380) .

Tubuloglomerular Feedback in Pregnancy The tubuloglomerular feedback system provides a mechanism whereby the rate of early distal nephron tubular fluid movement controls the rate of filtration (SNGFR) at the same nephron . This system may be viewed primarily as a volume regulatory mechanism since an increase in SNGFR will produce an increased distal fluid delivery which in turn will lead to a feedback-evoked renal vasoconstriction . This will cause reduction back to the original value of the SNGFR, thus protecting the organism from excess fluid losses. In some states of chronic volume expansion, the vasoconstrictor activity of the tubuloglomerular feedback system becomes suppressed so that very little feedbackevoked change in SNGFR is observed over a wide range of distal fluid rates. 8 This presumably allows the organism to maintain an elevated GFR while excreting the excess volume load. The activity of the tubuloglomerular feedback system can be investigated experimentally by placing a block in the midproximal tubule and microperfusing the late proximal tubule, thus varying the rate of fluid delivery to the early distal nephron while collecting fluid for SNGFR measurement from an early proximal convolution. Studies have been performed in the midterm pregnant rat, which exhibits significant volume expansion, to investigate whether tubuloglomerular feedback suppression leads to a "permissive" increase in GFR in the midterm pregnant rat. 9 As shown in Fig 2, the virgin rat reduces SNGFR by -30% when flow rate is varied over the physiologic range, between 10 and 20 nUmin. At day 12 of pregnancy, despite the significant ( - 30 %) gestational plasma volume expansion, suppression of the tubuloglomerular feedback system is not observed. Indeed, as shown in Fig 2, the vasoconstrictor activity of the system seems to operate at normal or even slightly supernormal levels. 9 Therefore, the gestational rise in GFR cannot be

263

RENAL HEMODYNAMIC DETERMINANTS IN PREGNANCY

36 32 28 SNGFR nl/min

24 20 16

L 0

10

20

30

40

VLp • nl/min Fig 2. Tubuloglomerular feedback curves for euvolemic 12-day-pregnant and virgin Munich-Wistar rats. After introduction of a wax block into the next to last surface convolution, the late proximal tubule was perfused at varying flow rates (VLP over a range from 0 to 40 nLlmin, and the early proximal single nephron glomerular filtration rate (SNGFR) was measured at each perfusion rate. (Reprinted with permission. 9 )

due to a permissive effect due to volume expansion-induced feedback suppression; rather, the tubuloglomerular feedback system is acting as though extracellular fluid volumes were normal or even slightly contracted, a subject discussed further in the articles by both Barron (pp 296-302) and by Lindheimer et al (pp 270-275).

nancy in the rat in which maternal pituitary prolactin maintains the pregnancy. 2 Studies in the Munich-Wistar rat have established that in pseudopregnancy induced by a sterile mating the GFR and outer cortical SNGFR were increased by - 30 % as compared with the virgin value. 2 Furthermore, the altered determinants of SNGFR, measured by micropuncture, were virtually identical in pregnant and pseudopregnant rats, strongly suggesting that the initiation of the gestational rise in GFR is due to maternal rather than fetoplacental factors. 2 Studies by Matthews and Taylor, however, have implicated a major role for some placental factor in maintenance of the increased GFR later in the rat pregnancy. 2 The nature of the maternal factor responsible for the increased GFR of pregnancy is currently unknown. Progesterone, when administered chronically to female rats, produces a chemical form of pseudopregnancy but does not evoke any alterations in GFR; therefore, it is unlikely that progesterone is directly involved in mediating the gestational rise in GFR.2 The current data on prolactin 10

II

MECHANISMS THAT CAUSE GESTATIONAL RISE IN GFR

The initiating cause of the pregnancy-related rise in GFR remains unknown; however, some factors have been excluded. One question relating to whether the initiating factor is maternal or fetal in origin has been answered by studies in the Munich-Wistar rat since a state of "pseudopregnancy" is easily created in the rat through mating with a sterile male. Pseudopregnancy is characterized by cessation ofthe estrus cycle following a sterile mating, an event thought to be triggered and maintained by episodic surges in pituitary prolactin production. The levels of various hormones, including progesterone, rise in pseudopregnancy, and the rate of body weight gain increases in pseudopregnant rats to resemble that which occurs in pregnancy. The pseudopregnancy is maintained for - ten to 12 days; this coincides with the first half of true preg-

-21 Yinil.

'CIOr

12.,."

'CIOr

12.,."

III

§

II

~

i ..

a:

.: 21

i

..

z: u

• -a Yinil.

'CIOr

stage of gestat i on. days.

211_

Fig 3. Summary of the change in glomerular filtration rate (GFR) (upper panel) and plasma volume (lower panel) expressed as a percentage of change as compared with the virgin value throughout the period of gestation in the Munich-Wistar rat.

264

CHRISTINE BAYLIS

are conflicting and are discussed in some detail by Conrad (pp 253-259). Possible involvement of the vasoactive hormones angiotensin II and the prostaglandins have also been largely ruled out by animal studies (Conrad) . RELATIONSHIP BETWEEN PLASMA VOLUME EXPANSION AND GESTATIONAL RISE IN GFR

Plasma volume expansion leads to increases in GFR secondary to renal vasodilation and increased renal plasma flow rate. Normal pregnancy is associated with plasma volume expansion in the rat and woman, and the plasma expansion probably contributes to the gestational rise in GFR. Figure 3 gives the time course of the percentage changes in plasma volume and GFR throughout the gestation period in the rat. The maximum increment in GFR is evidently established by day 9 of pregnancy at a

time when the plasma volume expansion is relatively slight. In contrast, late in pregnancy, when a massive increase in plasma volume is evident, the GFR is declining slightly and returning toward the virgin value. 2 Thus, a direct relationship does not exist between the gestational rise in GFR and the maternal plasma volume expansion . There is, however, some association between volume expansion and increased GFR; the elevated GFR that occurs in pseudopregnancy induced by a sterile mating is also attended by a mild plasma volume expansion. In general, plasma volume expansion is associated with increase in GFR, but the contribution of the volume expansion per se and the influence of other factors to modulate renal vascular resistance separately at various stages in pregnancy remains unclear.

REFERENCES I. Baylis C: Glomerular filtration dynamics, in Lote C (ed) : Advances in Renal Physiology. London , Croom Helm, 1986, pp 35-85 2. Baylis C: Control of volume and renal hemodynamics during pregnancy in the rat. Semin Nephrol 4:208-220, 1984 3. Baylis C: Renal effects of cyclooxygenase inhibition in the pregnant rat. Am J Physiol (in press) 4 . Dal Canton A, Conte G, Esposito C. et al : Effect ofpregnancy on glomerular dynamics: Micropuncture study in the rat. Kidney Int 22 :608-612, 1982 5. Bay lis C: Test of filtration renal reserve in 12 day pregnant (P) rats with acute glycine (G) infusion. Kidney Int 29 :380, 1986

6. Meyer TW, Ichikawa I, Zatz R, et la: The renal hemodynamic response to amino acid infusion in the rat. Trans Assoc Am Phys XCVI:76-83, 1983 7. Brendolan A, Bragantini L, Chiaramonte S, et al: Renal functional reserve in pregnancy. Kidney Int 28:232, 1985 8. Blantz RC, Pelayo JC : A functional role for the tubuloglomerular feedback mechanism . Kidney Int 25 :739-746, 1984 9. Baylis C, Blantz RC: Thbuloglomerular feedback activity in virgin and 12 day pregnant rats. Am J Physiol 249:FI69173, 1985