Article Text
Statistics from Altmetric.com
When caring for women with structural heart disease (SHD), clinicians face an apparent paradox: most women with SHD are able to complete a pregnancy without complication; however, compared with the general population, maternal risk is increased and SHD remains a major source of maternal morbidity and mortality and poor fetal outcomes.1–4 Identifying those women with SHD who are at increased risk for meaningful complications during pregnancy is critical.
During pregnancy, intravascular volume, stroke volume, heart rate and cardiac output increase, while systemic vascular resistance decreases. During pregnancy, women with congenital heart disease (CHD), cardiomyopathy or valve disease can develop heart failure due to these physiologic changes.
Preconception risk stratification allows women with heart disease to decide if they wish to conceive. Additionally, identifying patients at high risk for cardiac complications provides time to plan the optimal strategy for cardiac surveillance and management during pregnancy. Various models have been developed in order to determine which cardiac conditions place women at highest risk for cardiac complications during pregnancy (table 1).
The European Society of Cardiology's Registry on Pregnancy and Cardiac disease (ROPAC) collected information on 1321 pregnancies of women with structural and ischaemic heart disease in 28 countries. Investigators followed women though pregnancy and reported on maternal and fetal complications.5
Ruys et al6 present the heart failure data from ROPAC. Importantly, the authors report heart failure as a stand-alone end point. This has advantages compared with other studies that report composite end points (death, heart failure, arrhythmia, myocardial infarction and cerebrovascular accidents).1–3 Although composite end points are useful in allowing smaller study size, limitations exist unless each component is of equal importance to the patient and occurs in similar frequency.7 In some previous series, arrhythmias (typically supraventricular) drive the composite end point and pre-existing arrhythmia has been a strong predictor of outcome. However, an arrhythmia may not be as important an outcome to the patient as heart failure, stroke or death.
In the current ROPAC report,6 13% of women with known cardiomyopathy, structural or ischaemic heart disease developed heart failure during or immediately after pregnancy. Heart failure was more likely to develop in women with left-sided valve disease and left ventricular dysfunction, baseline cardiomyopathy, pulmonary hypertension or poorer baseline NYHA functional class. Heart failure was most likely to develop at two points in time: near the beginning of the third trimester and immediately following delivery. These data make physiologic sense as these periods are characterised by the most pronounced haemodynamic changes, including increases in vascular volume, heart rate and cardiac output. Delivery is accompanied by an abrupt autotransfusion from the placenta and increase in afterload, which explains the peak in heart failure episodes at delivery. These data support frequent and careful evaluations for at-risk women during the late second and third trimesters and longer observation after delivery of those at high risk for heart failure. In high-risk patients, it also may be reasonable to treat with diuretics in the first 24–48 h after delivery to prevent heart failure in the postpartum period.
Women who developed heart failure have an increased risk of death, maternal cardiovascular complications and fetal complications compared with women who remained free of heart failure, which confirms heart failure as an important end point. Of women with heart failure, 4.8% died compared with 0.5% of women without heart failure. Fetal death was also more common in women with heart failure than those without (4.6% vs 1.2%). Additionally, infants of women with heart failure were more likely to be born prematurely and have low birth weight.
ROPAC enrolled a diverse group of women into the study, including women with ischaemic, valvular, congenital and myopathic disease. This diversity is a strength of the study; however, diversity also introduces a challenge as to the best way to group the women for analysis. Anatomic grouping is perhaps the most intuitive but also perhaps the most problematic strategy, with patient anatomy grouped into ‘right-sided lesions,’ ‘left-sided lesions’ and ‘shunt lesions.’ While this method is conceptually simple, it may not be adequately discriminating. Left-sided lesions may be more physiologically different from one another than they are similar. For example, left-sided obstructive lesions (eg, aortic or mitral stenosis) are associated with high rates of maternal complications, whereas left-sided valve regurgitation often is well tolerated without complications.1 ,3 Neither result is surprising when considering the different underlying anatomic and physiologic abnormalities and the expected response to the physiologic changes in pregnancy.
Women also are grouped according to a modified schema based on the WHO's recommendations for contraception choice in women with cardiovascular disease. Previous authors have modified the WHO contraception criteria to create four classifications of risk for pregnancy based on the type of cardiac pathology.8 The WHO 1 risk category denotes women for whom pregnancy poses a similar risk as the general population. WHO 4 indicates that pregnancy is contraindicated due to extremely high maternal risk. Women who are categorised as WHO 2 and 3 have intermediate risk from pregnancy. The ROPAC investigators validate this categorisation strategy finding that, on multivariable analysis, women with WHO category ≥3 have a higher risk of heart failure. This important information means that WHO category can be used as a risk stratification tool in a way similar to the use of the Cardiac Disease in Pregnancy (CARPREG) score. The CARPREG score incorporates risk factors, including LV obstruction, systemic ventricular dysfunction, poor functional status, cyanosis and prior cardiac events, to identify those women at higher risk of adverse cardiovascular events during pregnancy.3 Based on this study from ROPAC, the WHO score could be used to specifically address heart failure risk using similar general risk factors.
There are still answered clinical questions for risk stratification of pregnant women with heart disease. Detailed clinical information often is incomplete in large registry studies. For example, heart failure that responds to a single dose of diuretics is very different from heart failure requiring hospitalisation or inotropic support. Individualised risk assessment is not yet possible that considers the patient's baseline characteristics, such as LVEF, pulmonary pressures and quantitative severity of valve disease. It can therefore be difficult to counsel women about what ‘heart failure’ would mean, or guide those who may have mild ventricular dysfunction, valve dysfunction or pulmonary hypertension. These data reinforce that we at least can be confident that women with normal or near-normal baseline NYHA class have a low incidence of heart failure during pregnancy even when heart disease is present.
Even though women from developing countries accounted for only 14% of the study group, the global nature of the ROPAC registry is noteworthy as other large studies have looked exclusively at women in North America or northern Europe. Women from developing countries had higher maternal mortality (3.9 vs 0.6%) and fetal death rates (6.5 vs 0.9%) compared with the overall study group. However, women in developing countries were more likely to be symptomatic preconception. Many factors may contribute to these differences in outcomes, including discrepancies in baseline characteristics or access to care.
Women seeking preconception counselling often ask whether pregnancy is likely to have long-term, irreversible health consequences. The overall mortality rate is less than 1% in the largest studies of pregnancy in women with heart disease.1–3 ,5 However, robust data on long-term morbidity—those conditions that do not resolve quickly after pregnancy—are scarce. Patients may be more likely to avoid pregnancy if they know they are at higher risk of having a permanent decrease in functional status or have an event that leaves them permanently disabled. While heart failure exacerbation is an undeniably important outcome to measure in pregnancy, it is usually able to be treated medically and may not have lingering effects. Once the physiologic changes in pregnancy have resolved, most women return to baseline functional status. However, some data suggest that late cardiac events are predicted by prepregnancy risk factors as well as events during pregnancy in women with CHD.9 Thus, pregnancy is often considered a ‘stress test for life’ and may be an important prognostic indicator for those women who successfully complete a pregnancy, but counselling women prior to conception remains imperfect.
Although our understanding of pregnancy risks in women with heart disease has significantly advanced in even the last 15 years, many important questions can only be answered through multinational collaborative studies such as ROPAC. With expanded data collection and longer-term follow-up, such studies can improve our ability to individualise our counselling and care.
Footnotes
-
Competing interests None.
-
Provenance and peer review Commissioned; internally peer reviewed.