Antihypertensive therapy in pregnancy

Download Antihypertensive therapy in pregnancy

Post on 14-Jul-2016




0 download

Embed Size (px)


  • Ad*DMe20E-m


    InHymweocthRethtensinprancastamgrawhfolsivstugreffstrspthAntihypertensive Therapy in PregnancyJason G. Umans, MD, PhD, FACP*, and Marshall D. Lindheimer, MD, FACP

    dressivision of Nephrology and Hypertension, Georgetown University dical Center, 6PHC, 3800 Reservoir Road, NW, Washington, DC 007, USA.


    rrent Hypertension Reports 2001, 3:392399rrent Science Inc. ISSN 1522-6417pyright 2001 by Current Science Inc.

    troductionpertension in pregnancy remains a major cause of

    orbidity and mortality in the mother and fetus in bothstern and developing nations. Much of this morbiditycurs in women with chronic hypertension, especially inose who develop superimposed pre-eclampsia. In 1976dman et al. [1] published a landmark report describingeir randomized trial of methyldopa to treat chronic hyper-

    sion in pregnant women. There have been other trialsce then, some of excellent quality, and considerable

    ogress has been made towards understanding the etiologyd pathophysiology of the hypertensive disorders compli-ting gestation, particularly pre-eclampsia [2]. Yet, as wert this millennium, there are few guidelines on how to

    anage the most challenging of pregnant patients, thevida with pre-existing essential hypertension, includingen and how to treat with antihypertensive drugs. In thelowing sections we review evidence guiding antihyperten-e therapy in pregnant women; assess the contributions ofdies that shed light on the mechanisms of hypertension in

    avidas; direct the reader to several recent and ongoingorts that use the tools of meta-analysis to provide auctured review of this inadequate literature; highlightecific treatment recommendations of the expert panelsat have recently opined on this subject [35]; and

    present recommendations for critically needed research inthe area. Not discussed here are current recommendationsfor clinical surveillance of hypertensive gravidas, treatment ofeclamptic seizures, risk factors for pre-eclampsia or strategiesfor its prevention, or guidelines for counseling these patients;these have all been recently reviewed elsewhere [2,6,7].

    Maternal and Fetal Risks of Hypertension in PregnancyRegarding risk, we first must ask the following: 1) what arethe short-term risks of hypertension during pregnancy, 2)do available treatment trials allow us to advocateantihypertensive use for their control, 3) can we adequatelyjassess the comparative safety and efficacy of available anti-hypertensives, and 4) can we define appropriate therapeuticgoals and treatment protocols?

    The major maternal risk associated with underlyinghypertension that might justify pharmacotherapy is super-imposed pre-eclampsia, whose complications account formost, but not all, of the morbidity ascribed to chronichypertension during pregnancy [8]. Other risks that couldconceivably be modified by treatment include placentalabruption, accelerated hypertension leading to hospital-ization or to target organ damage, and cerebral vascularcatastrophes [9]. It is also conceivable that treatmentmight avert the fetal risks that include death, growthretardation, neonatal morbidity, and early delivery, thelatter occurring in many cases due to concerns regardingmaternal safety. Unfortunately, we have little data tosupport most of the above assumptions, and we needunequivocal answers to these questions for the followingreasons: First, although the diagnosis of superimposedpre-eclampsia is often difficult in the setting of chronichypertension (the careful clinician should always err infavor of cautious overdiagnosis), several well-conductedstudies suggest that superimposed pre-eclampsia willcomplicate at least 15% to 20% of pregnancies in womenwith blood pressures greater than 140/90 mm Hg, withincreased r i sk a t h igher va lues [9 ,10] . Second,hypertension doubles the incidence of placental abrup-tion [9,11], and in one large prospective study the riskwa s f u r t h e r i n c r e a s e d t h r e e f o l d w h e n c h r o n i chypertension was complicated by superimposed pre-eclampsia [9]. Finally, chronic hypertension is associated,with remarkable consistency, with a threefold increase inperinatal mortality [11], along with impaired fetalgrowth and neonatal outcome; early delivery, perinatal

    Human pregnancy, normally characterized by systemic vasodilation and modest hypotension, can be complicated by underlying maternal hypertension and several unique hypertensive disorders, including pre-eclampsia. Although well-designed and adequately powered clinical trials are critically needed, there have been several recent meta-analyses of this large literature, along with consensus statements and treatment guidelines from three distinct multidisciplinary groups of clinicians and investigators. In this paper we review recent analyses and guidelines, advising on our current approach to antihypertensive therapy in pregnant women.

  • Antihypertensive Therapy in Pregnancy Umans and Lindheimer 393

    death, and neonatal complications are even more frequentinwiba

    teecmof14apanprclimrhtreprmbe"hnoqudein ina



    perfusion pressure. However, despite the accord between

    women with either superimposed pre-eclampsia orth target organ damage as evidenced by proteinuria atseline [9].Currently, it appears that treatment of underlying hyper-

    nsion does not prevent superimposed pre-lampsia, abruptio placenta, or reliably decrease perinatalortality; however, these conclusions are based on the results several small trials whose results vary considerably [11]. The major apparent benefit of antihypertensive therapypears to be decreased occurrence of severe hypertensiond hypertension necessitating hospital admission later inegnancy [1214]. This more limited outcome remainsnically important since blood pressures as low as 170/110m Hg are clearly associated with cerebrovascular hemor-age in pregnant women, leading experienced clinicians toat such values as a medical emergency; improved bloodessure control also provides the reassurance required to per-it safer prolongation of the pregnancy, with its attendantnefit to the neonate. Importantly, since the relative risks ofard" morbid endpoints are low in mild hypertension, andne of the available trials are either comprehensive or ade-ately powered, our ability to advocate specific guidelinespends more on considerable clinical experience [35]combination with a critical evaluation of an unfortunatelydequate literature [2,6,1116], than on clear data.Enthusiasm for aggressive control of underlying mild

    pertension is further tempered by relatively unknownal and remote childhood risks of antihypertensive drugposure in utero. In this respect, while several recent smalldies have assessed some measures of fetal and neonatal

    owth, development, and function [1719], thentinued preference of many workers for methyldopa ase first-line agent for blood pressure control in pregnancyates to its apparent safety, demonstrated, albeit in smallmbers, through 7.5 years follow-up of children exposed

    utero [20].The Australasian Society for the Study of Hypertension

    Pregnancy has advocated drug therapy to maintainaternal pressures of less than 140/90 mm Hg [4], aal shared by the Canadian Hypertension Society forme groups of women at perceived excess risk [5]. Weefer the recommendations of the National High Bloodessure Education Program (NHBPEP) Working Group Hypertension in Pregnancy [3], with somewhat

    gher threshold pressures for (re)instituting treatment50160/100110 mm Hg), and somewhat less stringentgets for blood pressure control (but would treat at lowerels in select patients, such as those with underlying renal

    sease). This preference is in accord with a recent meta-ression analysis of results from 14 trials that suggested

    at tighter control of maternal mean arterial pressureight contribute to fetal growth restriction, irrespective ofe specific agents used [21]. This conclusion gains physio-gic plausibility from the presumed inability of intervil-us placental blood flow to autoregulate with reduced

    this metaregression analysis and the NHBPEP WorkingGroup guidelines, we believe that a definitive prospectivetrial specifically focused on the maternal and fetal effects ofdiffering levels of targeted (and achieved) blood pressurecontrol has yet to be performed and is critically needed.

    Maternal Hemodynamics During Normal and Hypertensive PregnancyNormal pregnancy is marked by early systemic vasodila-tion; decrements in systemic vascular resistance are so largethat mean arterial pressure falls by approximately 10 mmHg despite 40% to 50% increases in blood volume andcardiac output [2]. Blood pressure is maximallydecreased by midpregnancy, increasing gradually towardsterm. Women with underlying essential hypertension mayexhibit an even greater early gestational fall in bloodpressure, by as much as 15 to 20 mm Hg [2,22], so as toeither obscure recognition of their underlying condition orto make continued treatment unnecessary.

    The hypertension in pre-eclampsia is characterized byprimary intense systemic vasoconstriction, as determinedby invasive hemodynamic measurements [2,23]. Thevasoconstriction is associated with modest decrements ofcardiac output, but with normal left ventricular fillingpressures. Curiously, there is now evidence that somewomen destined to develop pre-eclampsia exhibit evengreater than normal increments in cardiac output prior tothe onset of hypertension, their cardiac outputs fallingwith onset of the systemic vasoconstriction and hyperten-sion that characterize overt disease [24]. Women whodevelop (nonproteinuric) gestational hypertension sharethis early exaggerated increase in cardiac output butmaintain their hyperdynamic circulation, with lowperipheral resistance, throughout pregnancy. Easterling etal. [25] hypothesized that treatment of this early excessivecardiac output with atenolol would prevent subsequentpre-eclampsia; this prediction was supported by a pilotstudy of 56 hemodynamically selected and initiallynormotensive gravidas [25]. However intriguing, thispreliminary report neither offers guidance on thetreatment of already hypertensive women, nor allows us tobalance the fetal risks of more widespread -blocker useagainst possible maternal benefit. Pre-eclampsia is furthercharacterized by markedly increased sympathetic outflow[26], which, although not a likely cause of the hyperten-sion, has been taken by many as a mechanistic justificationfor the initial selection of agents such as methyldopa.

    The renin-angiotensin system is activated in normalhuman pregnancy [27]. Even in pre-eclampsia, whereangiotensin II levels are lower than in normal gestation,there may be simultaneous upregulation of AT1 receptors,accompanied by mechanistically fascinating evidence forthe production of AT1 receptor agonistic autoantibodies[28]. Thus, angiotensin converting enzyme (ACE)

  • 394 Antihypertensive Therapy: Patient Selection and Special Problems

    inhibitors or AT receptor blockers might have seemedatwofetwidy(liret[2faiprFocoofnosm"ssuuncirdrat wiACindrusthatprfin



    Methyldopa decreases the subsequent incidence of severe1tractive antihypertensive agents for use in pregnantmen. Unfortunately, ACE inhibition not only increasedal wastage in animal studies, but it has been associatedth a specific fetopathy in humans (including renalsgenesis and calvarial hypoplasia), oligohydramnioskely a result of fetal oliguria), intrauterine growthardation, and often-fatal neonatal anuric renal failure

    9,30]. AT1 blockers also lead to fetal and neonatal renallure in the rat, leading us to similarly reject their use inegnancy [31]. In spite of these concerns, noted by the USod and Drug Administration and each of the recentnsensus panels [35], there have been several reports ACE inhibition during pregnancy. A case report [33]ted eventually reversible neonatal renal failure, while aall series, whose authors suggested low-dose captopril as

    alvage therapy" for refractory hypertension, notedrprisingly good neonatal outcome [34]. We remainconvinced, and do not believe that there are clinicalcumstances that would currently warrant the use of theseugs late in human pregnancy. We note that many womenrisk for hypertension during pregnancy, particularly thoseth underlying diabetes mellitus, may benefit from use ofE inhibitors prior to conception. Since all cases of ACE

    hibitor-associated fetopathy or renal failure occurred withug use in the latter two trimesters, it seems reasonable toe these drugs when appropriate, counseling women thatey should either change to alternate agents whentempting to conceive, or discontinue them early inegnancy, not resuming these agents until they areished nursing their infants.

    nsiderations in the Selection of Specific tihypertensive Agents in Pregnancy

    noted above, early pregnancy is profoundly vasodilatord hypotensive in most women with stage 1 or 2 hyper-sion (140179/90109 mm Hg); indeed most women

    th such mild to moderate hypertension will be able toarkedly decrease or eliminate their need for antihyper-

    sives early in gestation. Conversely, many cliniciansnsider failure of this early hemodynamic adaptation toegnancy to be an ominous prognostic factor. We willxt focus our discussion on agents used for nonemergentood pressure control during pregnancy, turningerward to drugs used for urgent control of higher or

    ore refractory elevations of blood pressure, usuallyarer to delivery.

    al agents for initial blood pressure controlethyldopa remains the preferred agent for initial bloodessure control [3], since no modern antihyperten-e has proven superior, or shares its history of clinical

    fety, bolstered by prospective long-term follow-up. It hasen assessed in both prospective placebo-controlled andmparative trials in pregnant women [24,1114].

    hypertension, while being well tolerated by the motherand without apparent adverse effects on uteroplacental orfetal hemodynamics [34] or on fetal well being. A recentmeta-analysis, which found no effect of antihypertensivetherapy on perinatal mortality (23 trials, RR, 0.71 [0.461.09]), suggested that other drugs might be superior tomethyldopa in preventing this outcome [14]. The markedheterogeneity in these mostly small trials tempers our will-ingness to abandon consensus recommendations favoringmethyldopa based on an ongoing meta-analysis in theabsence of an adequately powered randomized trial. (Inaddition, these poorer perinatal outcome results may beconfounded by the fact that many of the methyldopastudies were performed earlier when neonatal salvage wasless developed.) Most important to us are data, albeit froma small number of patients, that birthweight, neonatalcomplications, and development during the first year weresimilar in children exposed to methyldopa or placebo, andboth intelligence and neurocognitive development wereunimpaired a...