Authors: Laurie B. Griffin, Rachel Sinkey, Alan Tita, Dwight J. Rouse
Categories: Article, antihypertensive medication, hypertensive disorders of pregnancy, indication creep, preeclampsia
Source: American journal of obstetrics & gynecology MFM
Authors: Laurie B. Griffin, Rachel Sinkey, Alan Tita, Dwight J. Rouse
Hypertensive disorders of pregnancy, including gestational hypertension and preeclampsia, affect approximately 13% of all pregnancies and are a major cause of maternal and neonatal morbidity and mortality worldwide. Although the treatment of preeclampsia with severe features has been well established on the basis of randomized controlled data, international society guidelines vary on the treatment of gestational hypertension and preeclampsia without severe features. The American College of Obstetricians and Gynecologists recommends against the use of antihypertensive agents for nonsevere hypertension (blood pressure of <160/110 mm Hg) in both gestational hypertension and preeclampsia without severe features given a lack of level 1 evidence in support of treatment and the theoretical risk of masking of disease progression or causing adverse fetal effects, such as growth restriction. However, with the publication of the Chronic Hypertension in Pregnancy trial, (CHAP) which demonstrated the benefit of treatment of nonsevere chronic hypertension, “indication creep” or the application of a treatment outside the population of proven benefit is being observed with the use of antihypertensive medication for the treatment of nonsevere hypertension in gestational hypertension and preeclampsia without severe features. The use of antihypertensive treatment in this population without a definitive trial and no clearly defined safety protocols is potentially dangerous and could, at worst, lead to maternal and fetal harm or, at best, provide benefit in ways that are hard to assess and, thus, interfere with efforts to generate definitive evidence to change practice guidelines, denying many pregnant patients optimal care. It is imperative that a definitive trial be performed performed prior to the widespread use of antihypertensive treatment for gestational hypertension or preeclampsia without severe features.
Hypertensive disorders of pregnancy (HDPs), including gestational hypertension and preeclampsia, increased 2-fold between 2007 and 2019 and now occur in approximately 13% of pregnancies in the United States (US).^1^ HDPs have short- and long-term implications for both maternal and child health and are directly responsible for 6% to 26% of maternal deaths worldwide.^2^ Affected mothers have a 2-fold increased risk of coronary heart disease, heart failure, and stroke.^3,4^ HDPs disproportionately affect women of color and their infants, with 14% to 16% of non-Hispanic Black and American Indian or Alaskan Native individuals being diagnosed with HDPs compared with 12% of White women.^1,5,6^ Furthermore, Centers for Disease Control and Prevention (CDC) 2007–2016 data reveal that HDPs disproportionately lead to pregnancy-related deaths among Black (8.2%), American Indian or Alaskan Native (12.8%), and Hispanic (9.7%) women compared with White women (6.2%).^7^ Infants born to mothers with HDPs have more neonatal complications than infants born to normotensive mothers, including respiratory and hematologic complications, decreased birthweight, and increased mortality.^8^
Hypertension in pregnancy can be dichotomized as chronic or pregnancy associated. The diagnosis of chronic hypertension (cHTN) is made if a patient meets the criteria for the American College of Cardiology (ACC)/American Heart Association (AHA) stage 2 hypertension (systolic blood pressure [SBP] ≥140 mm Hg or diastolic blood pressure [DBP] ≥90 mm Hg) before pregnancy or before 20 weeks of gestation.^9,10^ In contrast, when new-onset hypertension occurs at ≥20 weeks of gestation, it is categorized along a continuum of increasing severity from nonsevere HDPs (gestational hypertension or preeclampsia without severe features) to severe HDPs (preeclampsia with severe features or eclampsia).^11^ Gestational hypertension is defined as new-onset hypertension (SBP ≥140–159 mm Hg and/or DBP ≥90–109 mm Hg) without proteinuria or end-organ damage. Preeclampsia without severe features is defined as new-onset mild hypertension with proteinuria in the absence of end-organ damage; preeclampsia with severe features is defined as preeclampsia with new-onset severe hypertension (SBP ≥160 and/or DBP ≥110) and/or end-organ damage. End-organ damage includes visual changes, severe headaches, thrombocytopenia, renal insufficiency, liver dysfunction, pulmonary edema, heart failure, seizures (ie, eclampsia), maternal stroke, or placental abruption with its risks of fetal hypoxemia, permanent neurologic deficits, or perinatal death.^11^ The Table provides an overview of the classification of hypertension in pregnancy. Proper diagnosis of hypertension in pregnancy is crucial to subsequent management as treatment algorithms and timing of delivery are governed by the specific hypertension diagnosis.
Maternal and perinatal morbidity and mortality are increased among patients with severe forms of HDPs, and severe HDPs are a leading cause of maternal and perinatal morbidity and death in the US and globally.^12^ The course of action that poses the least medical risk to a pregnant individual with HDPs is delivery. Nevertheless, the decision to proceed with delivery is based on balancing maternal and fetal interests, as preterm delivery poses increased risks of morbidity and mortality for neonates. Two RCTS investigated, 2 randomized controlled trials (RCTs) by Odendaal et al^13^ and Sibai et al^14^ evaluated whether individuals with preeclampsia with severe features by blood pressure in the absence of end-organ damage (including eclampsia and hemolysis, elevated liver enzymes, and low platelet count) could be safely expectantly managed until 32 to 34 weeks of gestation. In both studies, the use of antihypertensive medications to lower blood pressure to <160/110 mm Hg lowered the risk of perinatal complications without increasing the rates of adverse maternal outcomes, such as eclampsia or death.^11,13,14^ This level 1 evidence led the American College of Obstetricians and Gynecologists (ACOG) to recommend that patients diagnosed with severe HDPs can be managed with antihypertensive therapy to lower blood pressure to <160/110 mm Hg and close inpatient monitoring with delivery at (1) 34 weeks of gestation (or earlier if they develop a contraindication to expectant management) or (2) at the time of diagnosis if severe HDPs occurs at ≥34 weeks of gestation.^11^
In an estimated 10% of US pregnancies, pregnant individuals are diagnosed with nonsevere HDPs.^5,15^ The average gestational age of onset of nonsevere HDPs ranges from 30 to 35 weeks.^11,16,17^ Of note, 2 RCTs investigated the optimal timing of delivery for nonsevere HDPs by comparing immediate delivery vs expectant management in nonsevere HDPs diagnosed at term (HYPITAT) and before term (HYPITAT II).^16,18^ These trials demonstrated that gestational hypertension and preeclampsia without severe features can be safely expectantly managed until 37 weeks of gestation or until the development of severe features (including blood pressure ≥160/110 mm Hg, evidence of end-organ damage, or eclampsia) in the absence of other indications for delivery. Both studies found that this approach decreased neonatal adverse outcomes without significantly jeopardizing maternal well-being.^16,18^ This informed ACOG practice guidelines for the management of nonsevere HDPs, which advise delivery at 37 weeks of gestation.^11^ Neither study addressed the question of oral antihypertensive treatment during expectant management. A small subset of participants (approximately 20%) in the HYPITAT II trial were on antihypertensive therapy at the time of randomization, but participant numbers precluded an adequately powered subgroup analysis. To date, an international consensus has not been reached regarding the use of antihypertensives for the treatment of nonsevere hypertension in gestational hypertension and preeclampsia without severe features.^19,20^ Specifically, ACOG recommends against the use of oral antihypertensive therapy for nonsevere hypertension, whereas the International Society for the Study of Hypertension in Pregnancy, National Institute for Health and Care Excellence, and Society of Obstetricians and Gynaecologists of Canada use varying hypertension thresholds for treatment ranging from 140 to 160/90 to 110 mm Hg.^20^ Importantly, these societies also use varying definitions to classify hypertension in pregnancy, complicating international consensus on best practices.^20,21^
There is a growing, but insufficient, body of evidence in support of the use of antihypertensive therapy for nonsevere HDPs. The Control of Hypertension in Pregnancy Study (CHIPS) randomized 1030 women with cHTN or gestational hypertension at 14 to 33 weeks of gestation to less tight blood pressure control (DBP <100 mm Hg) vs tight blood pressure control (DBP <85 mm Hg).^22^ The study achieved a difference in DBP of 5 mm Hg between groups and demonstrated that less tight blood pressure control increased the frequency of progression to severe hypertension (41% vs 28%, respectively; adjusted odds ratio, 1.80; 95% confidence interval, 1.34–2.38) without a difference in the primary composite outcome, including pregnancy loss, high-level neonatal care, or serious maternal complications.^22^ Furthermore, there were lower rates of thrombocytopenia and elevated transaminases in the tight blood pressure control group than in the less tight blood pressure control group.
Importantly, in the CHIPS trial, as it relates to nonsevere HDPs, only 251 patients (25%) had gestational hypertension, and no participant had a diagnosis of preeclampsia without severe features (ie, elevated, nonsevere blood pressures in the presence of proteinuria) at enrollment, precluding conclusions about potential treatment benefits for nonsevere HDPs.^22^ A recent meta-analysis of 72 small RCTs of antihypertensive therapy for nonsevere hypertension in pregnancy, including gestational hypertension, preeclampsia without severe features, and cHTN, suggested a reduced risk of progression to severe hypertension by 30% to 70% and potential improvements in fetal outcomes compared with placebo or no therapy.^23^ Specifically, the development of proteinuria or preeclampsia (odds ratio [OR], 0.73; 95% credible interval, 0.54–0.99) and fetal/neonatal death (OR, 0.54; 95% credible interval, 0.30–0.98) decreased with labetalol use compared with placebo or no treatment. Note that the intervals used were credible intervals and, thus, offer a probabilistic interpretation rather than a fixed range of plausible values seen with confidence intervals. Therefore, although promising, the inclusion of patients with either cHTN or nonsevere HDPs did not allow definitive conclusions regarding the treatment of nonsevere HDPs.
Subsequently, the Chronic Hypertension and Pregnancy (CHAP) trial demonstrated that treatment of nonsevere cHTN reduced severe preeclampsia and preterm birth.^24^ The CHAP trial was an RCT of 2408 pregnant individuals with mild cHTN at more than 60 US sites. Participants were randomized to usual care (no antihypertensive therapy unless blood pressure ≥160/105 mm Hg to goal <160/105 mm Hg) vs antihypertensive therapy to lower blood pressure to <140/90 mm Hg. The incidence of the primary composite outcome—a composite of preeclampsia with severe features, medically indicated preterm birth at <35 weeks of gestation, placental abruption, or neonatal death—was lower in those randomized to antihypertensive therapy (<140/90 mm Hg: 30.2% vs 37.0%; adjusted risk ratio (aRR), 0.82; 95% confidence interval, 0.74–0.92). The number needed to treat to prevent 1 primary outcome event was 14.7 (95% confidence interval, 9.4–33.7). In addition, a 20% reduction in preeclampsia with severe features (aRR, 0.80; 95% confidence interval, 0.70–0.92) was observed.^24^ Although the results of the CHAP trial led ACOG and the Society for Maternal-Fetal Medicine (SMFM) to recommend targeted blood pressure control to <140/90 mm Hg for pregnant women with cHTN,^17,25^ ACOG and SMFM did not alter the recommendations for the management of nonsevere HDPs, which affects far more women in pregnancy than cHTN (10% vs 2%, respectively).^10,11^
There is no adequately powered randomized trial to support oral antihypertensive treatment of nonsevere HDPs. Therefore, data to date are insufficient to change society guidelines for the management of nonsevere HDPs. Instead of treating nonsevere HDPs, ACOG recommends expectant management consisting of blood pressure monitoring, serial laboratory evaluation, symptom screening, antenatal fetal surveillance, and ultrasonographic growth assessment.^11^ Delivery is recommended at 34 weeks of gestation or sooner if maternal or neonatal status deteriorates (ie, progression to severe preeclampsia) or at the time of development of severe features if after 34 weeks of gestation. Support cited for these recommendations includes (1) a lack of definitive evidence for the benefit of antihypertensive medications in nonsevere HDPs, (2) concerns that treatment of hypertension in nonsevere HDPs may mask worsening maternal disease, and (3) treatment of nonsevere hypertension in nonsevere HDPs that may lower arterial perfusion to the fetus, causing fetal growth restriction or small-for-gestational-age neonates because of maternal hypotension.^11,19^ To date, all evidences suggest that treatment of nonsevere HDPs would decrease the incidence of severe HTN, an effect that, in theory, would decrease the adverse events associated with severe HTN and may decrease the progression of end-organ damage in the form of thrombocytopenia and elevated transaminases. However, by decreasing the incidence of severe HTN, there is a theoretical risk that treatment will mask signs of worsening disease, which would otherwise prompt earlier delivery. This may lead to the unintended consequence of an increased incidence of obstetrical emergencies, such as the development of eclampsia because of the prolongation of pregnancy in individuals who may otherwise have been delivered for severe hypertension. Therefore, although the treatment of nonsevere HDPs will likely lead to a decreased incidence of severe hypertension, whether the prevention of severe hypertension is strictly of benefit to pregnant individuals or has unintended harmful effects remains unknown. Given the high prevalence and morbidity of nonsevere HDPs, the same rigor and urgency to obtain definitive evidence for the management of nonsevere HDPs should be pursued, as was done for cHTN in pregnancy. The 2021 AHA Scientific Statement on Hypertension in Pregnancy calls for “prospective, adequately powered studies” to evaluate the effect of oral antihypertensive therapy for nonsevere HDPs.^26^
“Indication creep” is a well-documented phenomenon that occurs when a treatment with proven benefit in a specific population is inappropriately applied to an untested population.^27–29^ Indication creep is ubiquitous in all fields of medicine and has been documented to cause harm and lead to unnecessary use of healthcare resources and dollars.^28^ For example, the use of therapeutic hypothermia for cardiac arrest for unwitnessed cardiac arrests has not been shown to be effective or cost-effective,^28^ and the application of risperidone, a medication approved for the treatment of schizophrenia, for the treatment of behavioral disorders in patients with dementia was found to increase mortality risk.^27,30^
In obstetrics, indication creep is not uncommon and has been documented for interventions, such as antenatal late preterm steroid use,^29^ and is of increasing concern with new screening guidelines in areas, such as cervical length screening.^31^ Of note, 1 extreme example of indication creep for medication use in obstetrics with serious consequences was the use of thalidomide, a medication developed for insomnia, that became a popular therapy for morning sickness and resulted in significant birth defects in fetuses before its removal from the market.^32^ Although antihypertensive medications, such as labetalol and nifedipine, have been demonstrated to be safe in pregnancy and not associated with birth defects, no definitive data exists to ensure that oral antihypertensive treatment of nonsevere HDPs does not mask disease progression and thereby increase maternal risk or affect fetal growth or well-being.^11,19^
As ACOG and the SMFM updated their recommendations for the use of oral antihypertensive medication for pregnant patients with cHTN,^17,25^ we have observed that an increasing number of providers prescribe oral antihypertensive medication to patients who develop gestational hypertension or preeclampsia without severe features against ACOG guidelines and outside of the safety of a clinical trial, a phenomenon we have coined as “CHAP creep.” CHAP creep has been observed in both the liberalization of hypertension therapy for individuals who do not meet the strict diagnostic criteria of cHTN in pregnancy (eg, treatment of individuals who develop elevated BP at 21–22 weeks of gestation) or the use of antihypertensive therapy for those with clear gestational hypertension or preeclampsia without severe features (eg, treatment of individuals at 35 weeks of gestation with new-onset hypertension). The practice of CHAP creep is one we must be careful of as the implementation of novel treatments without evidence of efficacy has the potential to lead to undue harm for individuals and their fetuses and to negatively affect future pregnant individuals by making it difficult to generate definitive data in support of or against the use of the interventions. We respect the desire of clinicians to pursue interventions that minimize the morbidity associated with nonsevere HDPs. However, we must also recognize the cognitive bias, specifically confirmation bias, that exists in doing so. The SMFM calls for physicians to recognize and combat cognitive bias to optimize unbiased clinical decision-making.^33^ They specifically caution against the tendency to “overweigh evidence that supports existing beliefs and dismiss evidence that does not” and to “allot more weight to their anecdotal experience than to clinical trial evidence...to justify deviating from established guidelines or protocols.”^33^
There is a time and place for observational studies to inform expert opinion and clinical guidelines in the setting of rare diseases, and combining data from different study designs is crucial in changing clinical guidelines.^34^ However, in conditions as ubiquitous as gestational hypertension and preeclampsia without severe features, definitive level 1 evidence should be obtained before changes in practice guidelines to prevent the unintended negative consequence of indication creep and off-label use. Level 1 evidence requires support from data from at least 1 well-designed, large RCT with clear results for or against an intervention.^35^ When conditions in other specialties are compared, the evidence required to support practice is significant. For example, atrial fibrillation affects 1% to 2% of the population. The ACC/AHA first 3 management recommendations are based on level 1 evidence.^36,37^
Simply, no definitive data exists to ensure that oral antihypertensive treatment of nonsevere HDPs is beneficial and safe.^11,19^ Although preliminary data for the use of antihypertensive treatment in gestational hypertension and preeclampsia without severe features are promising, a randomized trial of antihypertensive therapy in nonsevere HDPs is urgently needed to assess the efficacy and safety under the protection of a clinical trial. A trial to answer this question is highly feasible. In the HYPITAT II trial, a Dutch practice-changing trial that evaluated the efficacy of expectant management of pregnant individuals with nonsevere HDPs to 37 weeks of gestation,^16^ women with nonsevere HDPs were enrolled beginning at 34 weeks of gestation and randomized to expectant management until 37 weeks of gestation vs immediate delivery at the time of diagnosis.^16^ Of participants randomized to expectant management, 36% delivered at <37 weeks of gestation for a maternal or fetal indication, including worsening hypertension (28%) or fetal compromise (7%). Based on data from the HYPITAT II trial and rates for adverse pregnancy outcomes in HDPs in a US population,^1,38^ conservative estimates for a primary composite outcome similar to that used in the CHAP trial (progression of disease, medically indicated preterm birth, fetal or neonatal death, or placental abruption)^24^ would be 25%. Therefore, a sample size of approximately 4000 individuals would provide ample power for a study to detect relative rate reductions of 20% from a baseline rate of 25%. We implore clinicians to use their desire to minimize the morbidity associated with nonsevere HDPs and efforts toward demanding scientific organizations support a definitive trial on antihypertensive medication for nonsevere HDPs. The results of a definitive trial, whether negative or positive, will provide level 1 evidence regarding the use of antihypertensive medication in nonsevere HDPs, improve outcomes or prevent undue harm for at least 5% to 10% of the 4 million annual US pregnancies, and definitively inform practice patterns worldwide.