Keywords
pregnancy, preeclampsia, total body water, intravascular volume, interstitial volume, renin–angiotensin–aldosterone system, natriuretic hormones, pressure natriuresis, Na + K + -ATPase, osmoregulation, vasopressin
Editors’ comment : The role of the renin-angiotensin-aldosterone system (RAAS) in preeclampsia has intrigued investigators for decades. Chesley had a chapter devoted to this topic in his first edition and he would have been delighted by some of the novel twists that have occurred in this important area. The discovery of agonistic autoantibodies (AT1-AA) is very much in line with his extreme interest on why preeclamptic women are hypersensitive to angiotensin II. He recognized that angiotensin II had a very short half-life in the circulation, and that peptide fragments (e.g., angiotensin III) also were vasoactive, but Chesley may have been surprised to learn that Ang 1-7 has potent vasodilatory activity. In the current edition we combine the subjects of the RAAS, AT1-AA, and body fluid volumes in preeclampsia as logical extensions of this intricate regulatory system .
Introduction
The renin angiotensin-aldosterone system (RAAS) is one of the most evolutionarily conserved blood pressure and volume regulating systems in vertebrates. Importantly, during human pregnancy, there is a ~12 kg weight gain, as well as a 30–50% increase in extracellular fluid, plasma, blood volume, and total body water. Also striking is a resetting of the thresholds for vasopressin secretion and thirst, plasma osmolality averaging 10 mOsm below nonpregnant levels (see Chapter 16 ). There is a marked stimulation of the RAAS and other potent mineralocorticoids that accompanies these changes, while opposing salt-retaining influences are increases in GFR and a rise in the plasma concentration of several natriuretic hormones. Therefore, pregnancy is a sensitive state relying on a multifactoral autoregulation of blood pressure control mechanisms and body fluid volume homeostasis.
Alterations in the RAAS accompany the development of hypertension. Preeclamptic women have long been known to have increased vascular sensitivity to angiotensin II in the absence of elevated angiotensin II or plasma renin activity. Recently alterations in either the vasodilatory Ang 1-7 or activating autoantibodies to the angiotensin II type I receptor (AT1-AA) have been noted to occur during preeclampsia. AT1-AA were found to be present in the serum of preeclamptic women at much higher levels than sera from nonpregnant women or pregnant women who went on to have normal pregnancies. Therefore, in recent years much research has been performed to determine the contribution of AT1-AA to the pathophysiology associated with preeclampsia. AT1-AA bind to and activate the AT1-receptor and induce signaling in vascular cells, including activating protein 1, calcineurin, and nuclear factor kappa-B activation, which can be blocked by an AT1 receptor antagonist. This signaling results in increased reactive oxygen species, sFlt-1 production and plasminogen activator inhibitor-1, and endothelin-1, all of which have been implicated in preeclampsia. More recent studies reveal an important role for AT1-AA in causing the increase in renal and vascular sensitivity to angiotensin II (Ang II). In addition to being elevated during preeclampsia, AT1-AA have also been reported to be increased in postpartum women. Hubel and colleagues demonstrated that AT1-AA correlated with insulin resistance and sFlt-1 and do not regress completely after delivery. Although these autoantibodies have been linked to poor placentation and abnormal renal function, their role in the hypertensive state of preeclampsia has yet to be fully elucidated. Furthermore, the importance of AT1-AA after preeclampsia, especially in the context of increased cardiovascular risk, remains to be determined.
Resolving the ongoing debate concerning the cause of the increased plasma volume (does it represent “underfill,” “overfill,” or “normal fill”?) could have important implications for management of complications during preeclampsia. Plasma volume generally decreases in preeclampsia (most marked in eclampsia), while interstitial water (edema) may increase further or remain unchanged. Levels of all components of the RAAS are decreased compared to normal pregnancy while the increased incidence of AT1-AA is associated with increased severity of the disease. Along with the controversial AT1-AA there is an evolving literature on the role of the dilating peptide Ang 1–7 in normal pregnancy and preeclampsia. Finally, evolving thoughts regarding treating preeclampsia with sodium loading or plasma volume expansion, a challenging view, were revisited in light of the recent discovery of genetic mutations leading to inefficient aldosterone production in preeclampsia. Contrarily, the possible presence of ouabain-like factors in animal models and women with preeclampsia is also noted. Furthermore, animal models of preeclampsia in which AT1-AA are suppressed are mentioned and are the subject of further investigation as we plunge forward to seek better, more innovative and safe therapeutics for managing edema and alterations in blood pressure during preeclampsia.
If preeclampsia is the “disease of theories,” one of the most far-fetched of these is that agonistic autoantibodies may participate as direct mediators of increased vascular sensitivity and may induce alterations in volume homeostasis. The role of agonistic autoantibodies in Graves disease is well established and, increasingly, circulating immunoglobulins have been associated with hypertension. Their possible role in preeclampsia is explored in this chapter. Agonistic autoantibodies directed at the angiotensin II AT1 receptor were first detected in patients with malignant hypertension. When these were observed in a hypertensive woman with a history of preeclampsia, it prompted a cohort investigation for such antibodies in archived sera from preeclamptics. These antibodies induced AT1 receptor signaling and functioned in immuno blots as well as non-agonistic commercial antibodies; moreover, they were cleared rapidly after delivery. The epitope shows signaling events in vascular smooth muscle cells and trophoblasts that could contribute to the development and signs of preeclampsia. Moreover, AT1-AA have also been observed in a reduced uterine perfusion rat model of hypertension and in pregnant rat models of hypertension induced by elevated cytokines such as TNF-α, IL-6, and IL-17. Once expressed in pregnancy, AT1-AA are capable of eliciting sFlt-1 production in human trophoblasts and in pregnant rodents. Furthermore, hypertension resulting from AT1-AA during pregnancy is caused by activation of the endothelin-1 and placental oxidative stress pathways. Newer evidence suggests an interplay between the AT1-AA, Ang II and the AT1 receptor activating ET-1 and oxidative stress pathways and culminating in hypertension. Knowledge gained from such research has validated the utility of rat and mouse disease models that are relevant for target identification and new therapies for preeclampsia.
The seemingly bizarre concept that circulating agonistic antibodies might mediate vasospasm in preeclampsia stemmed from a serendipitous observation in a single patient. When coupled to our knowledge of autoimmunity, immune tolerance, and the remarkable changes in the RAAS during pregnancy it led to a novel hypothesis. What has evolved in the past decade is an integrated framework of observations that fits into the concept of faulty angiogenesis as a precursor for phenotypic preeclampsia. The improbable agonistic autoantibody story warrants a brief review.
Autoantibodies that stimulate G protein-coupled receptors have long been accepted as causing diseases of the thyroid gland. Agonistic antithyrotropin receptor immunoglobulins (TSAb) that mimic the action of thyrotropin (TSH) mediate the endocrine manifestations of Graves disease and stimulate extraocular muscle fibroblast proliferation and differentiation in its associated ophthalmopathy. Nonetheless, the immunological mechanisms of TSAb production remain obscure. Recently, Kim-Saijo et al. produced a transgenic mouse using patient-derived TSAb that may prove useful as a model for Graves disease. Autoantibodies directed against specific epitopes in the insulin receptor exist, but these only rarely cause recurrent hypoglycemia and a severe form of insulin resistance (type B insulin resistance). In cancer chemotherapy, the generation of autoantibodies against the death receptors DR4 and DR5 of tumor necrosis factor-alpha-related apoptosis-inducing ligand (TRAIL) are interesting therapeutic targets, since agonistic antibodies against DR5 and DR2 induce apoptosis in cancer cells. Furthermore, agonistic anti-CD40 antibodies profoundly suppress the immune response to infection with lymphocytic choriomeningitis virus. Thus agonistic antibodies directed against cell-surface receptors appear to have an in vivo functional capacity to instill or suppress disease. A goal of this chapter will be to review the notion that such antibodies cause or amplify pathological cardiovascular responses in preeclampsia.
Body Fluid Volumes
Pregnancy is a physiologic process whereby repeated adjustments in intracellular and extracellular volume occur to maintain the steady state. Each new steady-state value is then held within relatively narrow limits, that is, these changes are sensed as normal and “defended” in face of variations in fluid and sodium intake. There is a 30–50% increase in extracellular fluid (ECF), plasma, and blood volume, associated with 30–50% increases in cardiac output, glomerular filtration rate (GFR) and renal blood flow.
The cause and significance of such changes have been debated for decades by three schools of thought. One advocates that the alterations are secondary to primary arterial vasodilatation causing “underfill” or decreased effective volume. The second view, called “normal fill,” implies that the gravida senses her volume as “normal” at every new steady state, and reacts appropriately to sodium restriction or surfeit. The final concept, “overfill,” views pregnant women as overexpanded with intravascular volume. Each camp agrees, however, that primary renal sodium and water retention is presumably responsible for the volume changes associated with normal pregnancy.
Preeclampsia
The sources of peripheral edema, a sign of excessive interstitial ECF, are more complex in gravidas than in nonpregnant women. There are two forms of edema in pregnancy – “normal” edema reflects a physiological ECF volume increase, whereas “abnormal” or “pathological” edema in preeclampsia occurs when fluid has shifted from the vascular to the interstitial space (due, for example, to pathological “leakiness” of vessels (see Chapter 9 )) – but the two etiologies are currently clinically indistinguishable. In women with “pathological” edema, total ECF is increased. In the second half of gestation, both pedal and pretibial edema can be detected in the majority of pregnant women, occurring more commonly as the day proceeds, often disappearing with recumbency. In addition to this dependent edema, many women develop edema in the hands and face as pregnancy progresses; again, its frequency reflects the care with which edema is sought, an incidence of>60% being described in the classic 1941 monograph by Dexter and Weiss. Even those women who fail to manifest overt edema have increases in lower limb volume (i.e., subclinical edema).
Chesley, in the first edition of this text, suggested that the primary reason for increases in peripheral edema in normal pregnancy was the decrease in plasma oncotic pressure. Values for oncotic pressure decrease from an average level of 370 mm H 2 O (27 mm Hg) in nonpregnant individuals to 345 mm H 2 O (25 mm Hg) in early pregnancy and 300 mm H 2 O (22 mm Hg) in late gestation. Figure 15.1 combines Chesley’s summary during gestation with a study by Zinaman et al. performed at delivery. Note that the lowest values are reached in the hours that follow delivery, the period of time when women with preeclampsia are most at risk of both pulmonary and cerebral edema, if fluid administration is not monitored with appropriate care.
Kogh et al. have shown that reductions in plasma oncotic pressure on the order of that encountered in normal pregnancy can be associated with a significant increase in the rate of fluid extravasation from capillaries. The reduction in plasma oncotic pressure during pregnancy is mainly due to the ~1 g decrement in plasma albumin in normal gestation. The usual explanations of this decline are that it is a dilutional phenomenon related to the increase in plasma water, as well as in the decrease in plasma tonicity of ~10 mOsm. However, such explanations may be simplistic as little is known of the production and disposal of albumin in normal gestation, and while the levels of some circulating solutes decline along with the decrease in tonicity, others do not. Finally, women with “normal” peripheral edema and otherwise normal pregnancies also have the greatest weight gain and plasma volume expansion as well as larger neonates with lower perinatal mortality rates than those with less fluid retention.
Appreciation of changes in oncotic pressure during gestation is extremely relevant to clinical practice. Preeclamptic women with peripheral edema appear to be at risk of developing interstitial and frank pulmonary and cerebral edema. This risk is increased by fluid challenges, particularly the intravenous infusion of crystalloids at the time of delivery, and the danger is enhanced in the immediate puerperium when fluid shifts from the expanded interstitial compartment of normal gestation back into the circulation. For example, Benedetti and Carlson have suggested that the risk of pulmonary edema can be predicted by measuring colloid oncotic pressure, a suggestion supported by the observation that the lowest values are measured in the early postpartum period, the period of greatest risk of this complication. Given that values in preeclamptic women are the lowest, they would naturally be the group at greatest risk of this life-threatening complication.
Finally, Hytten, as far back as 1970, wrote that edema is so common in pregnancy that it is not a useful diagnostic criterion to use in the diagnosis of preeclampsia. Still, hypertension plus edema continued to be used to diagnose the disease (often compromising the value of research reports), only disappearing from the diagnostic criteria of all major classifications at the commencement of this millennium.
“Normal Fill” or Resetting of the “Volumestat”
This hypothesis is supported by animal experiments that evaluate relationships between intravascular volume depletion and vasopressin release during pregnancy compared to the nonpregnant state (described below and reviewed in ). Studies in pregnant animals and humans suggest that sodium and water reabsorption in the proximal nephron (determined by indices such as fractional lithium or solute-free water clearances) are unaltered, and these species dilute their urine normally when water-loaded during gestation. Although osmotic thresholds are reset, experimental maneuvers aimed at abolishing “underfill” in both experimental animals and pregnant humans fail to alter the lower steady-state pressure, as well as the decreased osmotic threshold that occurs during gestation.
The observations concerning normal dilution and the excretion of water loads are important because failure to dilute the urine maximally and a blunted excretory response to water loading are major pathophysiologic features of the “underfill” status in diseases accompanied by hyponatremia such as cirrhosis, cardiac failure, or nephrotic syndrome. These disorders are considered prototypes of diseases in which absolute extracellular volumes are increased and “effective arterial volume” is low. Finally, most investigators describe similar sodium excretory responses to saline infusions in the pregnant and nonpregnant states in animals and humans.
Controversies about the various “fill” hypotheses are yet to be settled, and perhaps each is correct at particular stages during pregnancy. Hormone-induced vasodilatation creating temporary “underfill” in the early weeks, followed quickly by compensation to a “normal fill” state as pregnancy progresses, fits two of the three theories, whereas during the last trimester natriuretic factors predominate and restore Na balance, at least in some gravidas. The importance of settling this dispute, however, is not trivial, because a better understanding of how the pregnant woman “senses” her volume changes will impact management, particularly for gravidas with hypertensive complications and cardiac disorders.
Primary Arterial Vasodilatation (“Underfill”)
The concept of “underfill” in pregnancy partially resembles views of how cirrhotic and heart failure patients “sense” their increased ECF volumes. The following observations suggest that primary arterial vasodilatation causing arterial underfilling with secondary sodium and water excretion occurs in early gestation. Supporters of “underfill” note that systolic and diastolic blood pressures decrease early in the first trimester of pregnancy despite an increase in blood volume. The RAAS is activated early in pregnancy, an effect that would also occur with arterial underfilling due to peripheral arterial vasodilatation, while primary volume expansion would be expected to suppress these hormones. The increase in GFR and renal blood flow may precede the expanded extracellular fluid volume in pregnancy, also suggesting primary vasodilatation. Additional observations are the resetting of the osmostat and the volume depletion–vasopressin relationships (discussed below) in a direction also consistent with vascular underfilling due to systemic arterial vasodilatation.
Nitric oxide (NO) is a prime candidate as the mediator of vasodilatation in pregnancy. The resistance to angiotensin (AII), norepinephrine, and vasopressin that characterizes normal pregnancy can also be reversed by blockade of nitric oxide synthase (NOS).
Other evidence supporting “underfill” includes the observations that pregnant women manifest greater increases in aldosterone release than nonpregnant subjects in response to small quantities of AII and display exaggerated decreases in blood pressure when treated with ACE inhibitors.
Of particular interest is a study by Chapman et al. They serially studied 13 women prior to and immediately following conception, and again during gestational weeks 6, 8, 10, 12, 24, and 36 ( Fig. 15.2 ). Measured were blood pressure, cardiac output, and renal hemodynamics (inulin and para-amino hippurate clearances). Mean arterial blood pressure decreased by 6 weeks gestation, associated with an increase in cardiac output, decrease in systemic vascular resistance, and an increase in plasma volume ( Fig. 15.2 ). Renal plasma flow and GFR increased by 6 weeks gestation (see Fig. 16.2 , Chapter 16 ). Plasma renin activity (PRA) and aldosterone concentrations increased significantly by 6 weeks while norepinephrine levels did not change throughout pregnancy. Atrial natriuretic peptide (ANP) levels increased as well, first noted at 12 weeks. Plasma cyclic guanosine monophosphate (cGMP) levels decreased and cGMP clearance increased by 6 and 8 weeks, respectively. Chapman et al. strongly suggest that peripheral vasodilatation occurs early in pregnancy in association with renal vasodilatation and activation of the renin–angiotensin–aldosterone system. Volume expansion occurs early, followed by later increases in ANP, suggesting that ANP increases in response to changes in intravascular volume. The authors also confirmed the decreases in Na, Cl, and HCO serum concentrations, as well as the lowered creatinine, blood urea nitrogen, and hematocrit concentrations throughout pregnancy.
Pronounced placental growth occurs at weeks 6–8 of gestation and is generally complete by week 12. The systemic and renal regulatory changes observed by Chapman et al. occurred well before placentation. The authors speculated that maternal factors related to ovarian function are responsible for the peripheral vasodilatation, suggesting that NO might be largely responsible for the changes they observed. Renal production of NO or other natriuretic substances could in part explain these findings .
Excessive Expansion or “Overfill”
Supporters of the “overfill” theory note that there are absolute increments in ECF volumes, suggesting that this expansion may be an epiphenomenon to other changes, primarily a marked increase in mineralocorticoid hormone levels (see below). They also note the high levels of circulating natriuretic factors and various inhibitors of the membrane pump, which are expected responses to hypervolemia. There are also increases in renal hemodynamics, and two groups of investigators have described an increased sodium excretory capacity in response to saline infusions. Although reports are anecdotal, women with cardiac or renal disorders appear more susceptible to volume overload complications, whereas healthy gravidas seem to tolerate blood loss better than nonpregnant women.
Plasma Volume in Normal Pregnancy and Preeclampsia
Plasma volume is an especially clinically significant component of the ECF as it is a major determinant of organ perfusion. There are striking increments in intravascular volume during normal pregnancy, mainly the result of plasma water, but also due to a small increase in red blood cell mass. Plasma volume has been measured in most studies with Evans blue dye, and in the older literature values appeared to rise until gestational week 30 and then decline. Those studies were in error, because late in gestation the indicator cannot attain complete mixing during the 10 minutes the gravida is positioned in a supine or sitting position. When serial studies are performed with the pregnant woman positioned in lateral recumbency, plasma volume increases were observed to commence in the first trimester, accelerate in the second, peak near gestational week 32, and remain elevated until term. The maximal increases are 1.1–1.3 L, and larger gains are observed when multiple fetuses are present.
Women with preeclampsia/eclampsia have significantly reduced plasma volumes ( Table 15.1 ). The degree of contraction appears to be an index of severity, and in this respect the greatest decreases (approaching 50% of normal pregnant values) have been reported in nulliparas with eclampsia, the convulsive phase of the disease associated with extreme severity.
| Author (Ref) | Normal Pregnancy | Preeclampsia/Eclampsia | |||
|---|---|---|---|---|---|
| Cases ( n ) | Mean (mL) a | Cases ( n ) | Mean (mL) a | % Change | |
| Werko et al. | 4 | 3865 | 9 | 3145 | −18 |
| Freis & Kenny | 7 | 4287 | 5 | 3045 | −29 |
| Rottger | 20 | 3383 | 18 | 2890 | −15 |
| Cope | 29 | 3470 | 14 | 2820 | −19 |
| Freidberg & Lutz | 10 | 3104 | 17 | 3257 | +5 |
| Kolpakova | 20 | 3309 | 15 | 2918 | −12 |
| Honger | 20 | 3800 | 19 | 3300 | −13 |
| Haering et al. | 18 | 3721 | 21 | 3148 | −15 |
| Brody & Spetz | 46 | 4245 | 34 | 4010 | −5 |
| MacGillivray | 18 | 4040 | 35 | 3535 | −12 |
| Blekta et al. | 55 | 3133 | 14 | 2590 | −17 |
| Gallery et al. | 199 | 3878* | 37 | 3383* | −13 |
| MacGillivray | 55 | 3763 | 29 | 3524 | −6 |
| Brown et al. | 54 | 3912* | 49 | 3260* | −17 |
| Silver et al. | 20 | 4070 | 20 | 3416 | −16 |
| Zeeman et al. , b | 44 | 4505 | 29* | 3215 c | −29 d |
a Mean values except for those marked with *, which are median values. Values are listed as shown or were calculated from the publication listed.
c All were eclamptic nulliparas.
d Fourteen eclamptics had a subsequent normotensive gestation in which their total blood volumes were 45% greater than that measured after the eclamptic seizure.
The classic explanation of the decreased intravascular volume in hypertensive disorders is that the fluid is “chased” out by the increase in vasoconstrictor tone that results in increased blood pressure. The increase in sympathetic tone that may occur during preeclampsia favors this notion. Schobel et al. concluded that preeclampsia is a state of sympathetic overactivity which reverts to normal after delivery. Their data suggest that the increases in peripheral vascular resistance and blood pressure that characterize this disorder are mediated, at least in part, by a substantial increase in sympathetic vasoconstrictor activity. Thus, sympathetic nervous system overactivity in preeclampsia may contribute to abrogated vasodilatation expected in normal pregnancy.
Contrary to the above, Gallery et al. demonstrated that low plasma volume precedes the rise in blood pressure and other clinical manifestations of preeclampsia by several weeks ( Fig. 15.3 ). Their serial study also demonstrated that the increment in plasma volume in preeclamptic women and those whose blood pressure remained normal was similar in the second trimester of pregnancy but plasma volume contraction occurred thereafter, still preceding the development of overt disease by weeks. These observations appear to counter explanations that the decreased volumes reflect the effect of vasoconstriction on efflux of fluid from vessel walls or via the kidney (so-called “pressure naturesis” associated with hypertension). Finally, once overt preeclampsia appears, further volume contraction occurs, and this appears proportional to the severity of the disorder ( Fig. 15.3 ).
Intervention studies are rare in preeclampia; nevertheless one study compared low vs. high salt in 2077 pregnant women in 1958. The interventions consisted of advice to either increase or reduce the salt intake with meals. Surprisingly, the authors observed a lower incidence of toxemia, edema, perinatal death, and bleeding during pregnancy in those told to consume more salt. Extra salt in the diet seems to be essential for the health of a pregnant woman, her fetus, and the placenta. In contrast to guidelines for nonpregnant women, salt restriction is not advised in pregnant women, even in those with a history of hypertension. In 2000 the Cochrane database investigated the effects of advice on salt consumption during pregnancy on prevention and treatment of preeclampsia and its complications and confirmed that advising reduced salt intake cannot be recommended. Experimental in vivo data that salt reduction is not protective during pregnancy were obtained by Giradina et al. A low-salt diet significantly increased arterial pressure and vascular reactivity in pregnant and hypertensive-pregnant rats (RUPP). The authors speculate that the observed phenotype is caused by an increase in calcium entry from the extracellular space with a low-salt diet.
Novel Salt Concept
Recently, novel findings in metabolism suggested that salt is implicated not only in blood pressure control and volume homeostasis, but also in immune regulation.
The traditional concept is that sodium excretion by the kidney is the critical pathway regulating fluid status, determining the level of intra-arterial pressure and blood presssure control. Machnik et al. proposed an alternative mechanism that implicates macrophages as mediators of sodium storage in the subdermis. Sodium chloride can be stored without accumulation of water at hypertonic concentrations in interstitial proteoglycans. They investigated underlying mechanisms of salt-induced hypertension in rats and identified immune cells as principal sensors of salt load. Tissue macrophages express tonicity enhancer binding protein in response to the detected local hypertonicity and via activation of vascular endothelial growth factor-C (VEGF-C). As a result, they increase the density of the lymph-capillary network in the skin and enhance production of nitric oxide in the skin vessels, thus managing extracellular volume and blood pressure homeostasis. This finding does not overrule the renal regulatory function. It indicates rather that there also exists a local complementary extrarenal mechanism for electrolyte, volume, and blood pressure balance.
Human data on the role of VEGF-C-mediated salt homeostasis corroborate the findings in rats. The researchers explored the role of the VEGF-C–macrophage–lymphangiogenesis pathway as an extrarenal homeostatic mechanism in proteinuric chronic kidney disease patients and in healthy controls under a high-salt diet. VEGF-C levels along with blood pressure were elevated both in patients and in healthy individuals who ingested a high-salt diet.
Another proof that salt activates the immune system has been recently found in a study looking into the triggering mechanisms of autoimmune diseases. The authors showed that increased dietary salt intake is an environmental risk factor and generates pathogenic Th17 cells relevant for the development of autoimmune diseases. These data are based on the concept that the salt concentration in the interstitium and lymphoid tissue is considerable higher than in the plasma, approaching levels as high as 250 mM. However, the role of this concept in pregnancy and preeclampsia has yet to be determined.
Mineralocorticoids and the Renin-Angiotensin-Aldosterone Axis
Levels of several antinatriuretic hormones, primarily the mineralocorticoids aldosterone and desoxycorticosterone, increase markedly during gestation. Considerable research has also focused on the RAAS, because it is intimately involved in the renal control of salt and water balance. Prorenin levels increase quite early, probably the reason for the increment in circulating total renin levels. AII is also increased, an expected consequence of the increased renin and angiotensin production, and one reason for the high levels of aldosterone (an increased sensitivity of the adrenals to AII, noted previously, is another reason). Figure 15.4 summarizes a detailed study by Wilson et al. in which plasma renin substrate activity (PRA), plasma and urinary aldosterone, and urinary sodium and potassium excretion were measured throughout gestation. Sequential increases in substrate and activity, starting early in gestation, are shown more precisely in the study by Chapman and colleagues, whose subjects were tested before conception and during the initial weeks of pregnancy ( Fig. 15.2 ). Of interest in the study by Wilson et al. is that salt excretion (reflecting intake) was similar throughout gestation compared to postpartum measurements, suggesting that inadequate sodium intake does not account for stimulation of the RAA axis.
Of further interest is that seemingly high levels of all components of the RAA axis, and other potent mineralocorticoids, are not constant but respond appropriately when volume status is manipulated. Thus circulating levels of renin, angiotensin, and aldosterone decrease after saline infusion or during a high-salt diet, or increase further following dietary sodium restriction or administration of diuretics. Also, in one study, inhibition of aldosterone biosynthesis resulted in a diuresis and subtle signs of volume depletion, the salt loss already apparent when aldosterone excretion, though decreasing, was substantially greater than nonpregnant levels. Thus, in pregnancy the renin–angiotensin system does not function autonomously as some have postulated, but around a new set point. In essence, the high levels of aldosterone, often exceeding those measured in nonpregnant patients with primary aldosteronism, are appropriate in pregnancy and respond to homeostatic demands. There is a corollary view, however, that this reset is a two-component system, nonsuppressible and unresponsive to physiologic stimuli – a view to our knowledge that remains untested.
In addition to a marked rise in filtered sodium (possibly straining reabsorptive mechanisms), there are changes, too, in hormone levels and/or in autacoid function that theoretically enhance renal sodium excretion during pregnancy. These include increased circulating levels of oxytocin, vasodilating prostaglandins, melanocyte stimulating hormone, progesterone, and natriuretic peptides. Progesterone levels increase markedly in pregnancy and because the affinity of this hormone for the mineralocorticoid receptor exceeds that of aldosterone, some have queried how the latter functions as a mineralocorticoid in the face of such high circulating progesterone levels. The answer suggested is that progesterone actually becomes an antinatriuretic influence via metabolism to deoxycorticosterone (DOC) at extra-adrenal sites. In fact, because renal steroid 21-hydroxylase activity (the enzyme that enhances conversion of progesterone to DOC) is particularly high in pregnancy, the considerable portion of maternal DOC produced in the vicinity of the renal receptors enhances sodium reabsorption.
Studies by Shojaati et al. suggest that women who develop preeclampsia have gene mutations leading to inefficient aldosterone production (primarily a decrease in aldosterone synthase, CYP11B2). These authors suggest this results in inefficient volume expansion and thus poor placental perfusion in early pregnancy, and preeclampsia later in gestation. This work reminds us of the 1958 paper by Robinson in the Lancet, where salt loading throughout gestation decreased the incidence of preeclampsia, and saline infusions temporarily improved blood pressure. Indeed, in a recent case report from the same group who suggested the aldosterone production may be inefficient in preeclampsia, salt loading throughout gestation was purported to have prevented preeclampsia.
This concept – perturbations in steroid pathways – has been explored further. Cortisol has the same affinity for the mineralocorticoid receptor as aldosterone. Cortisol availability is controlled by 11β-hydroxysteroid dehydrogenase type 2 (11β-HSD2), which inactivates cortisol into cortisone, which is unable to bind to the mineralocorticoid receptor. The 11β-HSD2 enzyme activity limits intracellular cortisol concentrations and, within the uteroplacental compartment, the transfer of cortisol into the fetal circulation. Mechanisms by which 11β-HSD2 activity is controlled include epigenetic regulation via methylation of genomic DNA, transcription, post-transcriptional modifications of 11β-HSD2 transcript half-life, and direct inhibition of enzymatic activity. Evidence exists that 11β-HSD2 expression and activity are reduced in preeclampsia and that enzyme activity correlates with factors associated with increased vasoconstriction, such as an increased Ang II receptor subtype 1 expression, and notably decreased fetal growth. Proinflammatory cytokines known to be present and/or elevated in preeclampsia regulate 11β-HSD2 activity. Shallow trophoblast invasion with the resulting hypoxemia seems to reduce 11β-HSD2 activity. A positive feedback loop exists as activated glucocorticoid receptors enhance 11β-HSD2 mRNA transcription and mRNA stability. These findings also implicate disturbed mineralocorticoid receptor signaling in preeclampsia.
Antinaturetic Peptides in Pregnancy
In normal pregnancy there is an increase in plasma atrial natriuretic peptide levels (ANP), though some investigators have failed to observe these increases. Of interest is a study by Lowe et al., who prospectively controlled both sodium intake and posture and failed to find any increase during gestation. However, ANP’s metabolic clearance rate increases, so that even were circulating levels not to rise, the hormone’s production rate has increased. ANP levels are even higher in women with preeclampsia, including those with the disease superimposed on chronic hypertension. This situation seems paradoxical, as plasma volume decreases in preeclampsia. The answer may be that ANP production responds to stimuli such as contractility (e.g., brain natriuretic protein (BNP) is produced in the heart and increased in cardiac failure), and the ANP rise in preeclampsia is more related to cardiovascular events than to volume.
Of interest is a study by Irons and colleagues noting that low doses of infused ANP produced a minimal natriuresis in normotensive third-trimester women but had no impact on sodium excretion when women were restudied 4 months postpartum. Such data suggest a heightened natriuretic system (more consistent with over- or normal rather than underfill ).
On the other hand there are studies in pregnant rats suggesting loss of natriuretic responsiveness to administered ANP during normal pregnancy. The acute natriuretic response to volume expansion is dependent on endogenous ANP release and this is also blunted in the pregnant rat. Since both ANP and NO signal through cGMP, the observations in rodents suggested that the tubular response to cGMP might be blunted, and indeed Ni and colleagues noted increased cGMP breakdown in the inner medullary collecting duct of the pregnant rat kidney (a major site of natriuretic action of ANP and NO ). This effect is ascribed to selective, local increase in abundance/activity of phosphodiesterase 5 (PDE5). These same collaborative laboratories have further reported that the natriuretic response to administered ANP in pregnant rats could be restored by local intrarenal PDE5 inhibition. This natriuretic refractoriness in pregnant rats, therefore, is cGMP-specific, since dopamine-induced natriuresis (which signals via increased cAMP) remains unblunted. As interesting as these recent findings appear, their relevance to human pregnancy needs to be established, and some differences already noted between the blunted natriuretic ANP observations in rodents versus the enhancement of sodium excretion in pregnant humans remain to be resolved. Nonetheless, a recent pilot study suggests salutary effects of sildenafil on intrauterine growth restriction.
Preeclampsia
It appears paradoxical, given the decrease of intravascular volume associated with preeclampsia, that levels of all standard elements of the RAA axis decrease in this disorder. Both aldosterone levels and AII concentrations are below those of normotensive gravidas. However, there may be relative increments in angiotensin II receptors in platelets, and in other tissues, a speculation used to explain the increased pressor sensitivity to infused AII, originally described by Gant and colleagues in 1973. Of late it seems that the presence of the AT1-AA might explain the increased sensitivity observed by Gant.
Serial studies demonstrate that activity of the RAA axis is stimulated throughout pregnancy in pregnant women with chronic hypertension, whereas in those who develop superimposed preeclampsia the RAAS functions normally early in gestation and then decreases, to levels below those in those women who do not develop preeclampsia.
The changes in the RAA axis described both in normal and in preeclamptic pregnancies are apparently not synchronized and thus the plasma aldosterone:renin ratio is increased during normal pregnancy though the slope across different salt intakes is similar ( Fig. 15.5 ). The highest ratio is observed in preeclamptic women, where the disorder seems to suppress renin and other system components more than aldosterone ( Fig. 15.5 ).
