1. Long-term data on the effect of pregnancy on the progression of cardiac disease are few, and well conducted follow-up studies would be of great value for counseling.
2. Any deterioration in the function of bioprosthetic valves caused during pregnancy is small.
3. There is some evidence of greater long-term deterioration in function in women with a systemic right ventricle, pulmonary hypertension, peripartum cardiomyopathy, and Marfan syndrome.
The challenges of pregnancy for women with heart disease have been well documented. In earlier chapters, there has been a focus on fully assessing the underlying heart disease and developing a risk profile for each particular patient. With such a risk profile, we can provide prepregnancy counseling (Chapter 1) and use that risk profile as we plan the management and the monitoring through the pregnancy (Chapter 6). We have become adept at assessing the risk to the woman during pregnancy and early postpartum, the risk to her baby during the pregnancy, and the risk of recurrence of congenital heart disease in the offspring. We have put together management plans for a number of specific lesions, as detailed in their particular chapters, and have looked at the risks and benefits of invasive intervention if ongoing medical management is insufficient (Chapter 8). However, the information that we have accumulated for management during that year of pregnancy and early postpartum on how to achieve a healthy baby and a live mother has not been translated into the information that we need in order to speak to the woman and her partner about the long-term consequences. They need to know the prognosis for the particular cardiac lesion and whether that prognosis will be adversely affected by the pregnancy. This first year of pregnancy and early postpartum period is just the beginning of the woman’s life as a mother. She wants to be alive and as healthy as possible for as long as possible to fully care for her child. As stated by Colman et al. in the first edition of this book, better information about long-term outcomes after pregnancy would not only provide better-informed risk assessment for the woman but might also justify the early institution of medical therapy or even the consideration of invasive intervention prior to pregnancy.
There are some good data on the medium-term complications after pregnancy in women with congenital heart disease.[2–4] In a study of 318 women having 405 pregnancies, a number of cardiac events were documented over a median follow-up of 2.6 years. These events included death, arrhythmias, and congestive heart failure. A number of risk factors for late complication were identified: New York Heart Association (NYHA) class >II and/or cyanosis; subaortic ventricular dysfunction; subpulmonary ventricular dysfunction and/or pulmonary regurgitation; left heart obstruction; and adverse cardiac events before pregnancy and/or during pregnancy. For a woman with no risk factors, the 5-year risk of a late cardiac event was 7%, while it was 23% for a woman with one risk factor and 44% for a woman with more than one risk factors. Although this study did identify factors associated with a higher risk of late cardiac event, there was no matched control group and so no conclusion could be made as to the impact of the pregnancy itself on the risk of these late cardiac events.
As detailed in previous chapters, concerns have been expressed about not always having enough good data when it comes to the short-term management of a wide range of cardiac problems through pregnancy and early postpartum. Unfortunately, we have even less information about whether there will be any increased deleterious effect of the pregnancy on the woman’s underlying heart disease. In fact, the only comment about long-term issues in the recent European Society of Cardiology guidelines was about the long-term cardiovascular consequences of pregnancy-induced hypertension.
The difficulty in determining whether there are any long-term consequences of being pregnant for the woman with heart disease is that we have not been able to separate the effect of the simple passage of time on the heart disease from the direct impact of the pregnancy. We know that a woman with Fontan circulation, systemic right ventricle, or pulmonary hypertension (PH) is expected to have a decreased lifespan. The question is how much shorter that life will be for the woman who chooses motherhood and how much additional disability she may face in the long term.
In women without heart disease, we know that the cardiovascular and hemodynamic effects of pregnancy resolve gradually during the postpartum period.[6,7] When it comes to valvar heart disease such as rheumatic mitral stenosis, there are multiple reports as to medical management and interventional management during pregnancy.[8,9] However, although there have been reports of long-term follow-up after mitral valvuloplasty in pregnancy, there are no long-term data about the effects of pregnancy on the long-term worsening of medically managed mitral stenosis (although such worsening is not expected, it has not been ruled out). On the other hand, in women with more complex heart disease, there has been more debate about whether there is any long-lasting change in ventricular function after pregnancy.[10,11] In a retrospective study of 53 patients and 40 controls, Uebing et al. showed no significant change in right ventricular or left ventricular function after a mean follow-up of 2.9 years. For the patients with tetralogy of Fallot (ToF), there was an increase in right ventricular size but no change in functional class during follow-up. In a prospective study of 35 patients without a control group, Cornette et al. did demonstrate a reduction in systolic function as well as progressive diastolic dysfunction during pregnancy that persisted for 6 months after pregnancy. How much longer these changes would persist is unknown and the clinical significance of these findings remains unclear. Pregnancy for the woman with a univentricular heart (Fontan physiology) poses special problems regarding cardiac output response and arrhythmia, with an increased incidence of miscarriage and fetal growth restriction.[2,3] There are few published data on pregnancy in these patients and no published data on long-term outcome after pregnancy, although a recent review did recommend follow-up echocardiography to assess “potential postpartum cardiovascular sequelae”.
Specific lesions with more long-term data
There are cardiac lesions in which attempts have been made to gather long-term data about the effects that pregnancy may have on late prognosis. Even in these situations, the data remain limited.
PH has been considered one of the few absolute contraindications to pregnancy. Women with PH are frequently counseled against pregnancy and are often encouraged to seek termination in the event that pregnancy occurs, because of previously reported maternal mortality rates as high as 50%. With a better understanding of PH and the use of advanced therapies, there has been an observed improvement in pregnancy outcome and a decrease in medical mortality rates (Chapter 15).[12–15] In an extensive review of cases reported in the literature, Bedard et al. were able to demonstrate an improvement in maternal mortality rates since 2005, even though mortality continued to be relatively high (20–30%), with most deaths occurring postpartum. Kiely et al. recommended following these patients for 6 months postpartum. Duarte et al. reported that 12 completed pregnancies at 5 US centers resulted in 2 maternal deaths, 1 during pregnancy and 1 occurring at 2 months postpartum. Jaïs et al. reported 20 completed pregnancies in 7 European centers. There were three maternal deaths and one catastrophic right heart failure requiring extracorporeal circulatory support and heart–lung transplantation. All four of these women had very severe pulmonary artery hypertension (mean pulmonary arterial pressure [PAP] 71±5 mmHg, mean pulmonary vascular resistance [PVR] 1667±209 dynes/sec cm–5) and the suggestion was made that such patients should not proceed with pregnancy, although it may be reasonable to continue with pregnancy for women who have pulmonary artery hypertension with controlled pressures (mean PAP 36±15 mmHg, mean PVR 500±352 dynes/sec cm–5).
We have data that suggest a permanent change in the right ventricle after pregnancy in women with repaired ToF.[10,16] We do not have similar data in the women with PH. Nevertheless, it is reasonable to be concerned that the long-term worsening in right ventricular function will be potentiated by pregnancy. Long-term studies in these patients with PH will be especially important because modern management of PH has reduced the risk of maternal deaths in pregnancy and in the early postpartum period and we need to know whether this decreased early risk of maternal deaths may be cancelled out by more rapid right ventricular deterioration in the late postpartum period. Because of the small number of women with PH who go through pregnancy, obtaining this information represents a real challenge.
We know that any patient with a systemic right ventricle, transposition of the great arteries or transposition of the great arteries with intra-atrial repair, will have progressive deterioration in the systolic function of their systemic right ventricle. This has been documented after pregnancy and with the simple passage of time. Early studies did demonstrate worsening right ventricular function with pregnancy and the concern was expressed that such worsening may not be reversible after delivery.[17,18] In a retrospective study with a median of 5 years follow-up, there was no echocardiographic worsening of right ventricular size, but there was more treatment for heart failure and more heart failure admissions for the patients who had been pregnant. In the study by Uebing et al., there was no difference demonstrated between those patients who became pregnant and the nonpregnant controls, but the group was very small. The most recent study by Bowater et al. included 18 women with 31 pregnancies compared with both a female and male set of controls with systemic right ventricle. The pregnancies were well tolerated and there were no maternal deaths and only one neonatal death. However, the mean NYHA class deteriorated from 1 prior to the pregnancy to 1.4 during pregnancy. At 1-year follow-up, the right ventricular function in the women who had been pregnant did demonstrate significant deterioration, with no evident deterioration in the control groups. However, at long-term follow-up (mean 51 months), the degree of deterioration in the right ventricular function was similar in all three groups. The authors concluded that pregnancy with a systemic right ventricle was usually well tolerated, but that there was a small reduction in functional class. There was also earlier deterioration in right ventricular function, but at long-term follow-up the degree of right ventricular dysfunction was the same as in the control groups. The clinical significance of this outcome is unclear; it is also unclear how this should affect what we tell the patient with the systemic right ventricle. Perhaps we should advise the patient who has had early worsening in systemic right ventricular systolic function not to proceed to another pregnancy, in the belief that a subsequent pregnancy may lead to further accelerated right ventricular dysfunction.
With the excellent survival of women with repaired ToF, increasing numbers of these women are becoming pregnant (Chapter 14). Although pregnancy is generally well tolerated in this group, a number of factors have been described as affecting maternal and offspring outcome.[21–23] Long-term data are limited. Kamiya et al. described persisting increase in right ventricular volumes at 6 months postpartum and made the further observation that right ventricle size tends to become more dilated after the second and third pregnancies, but the lack of a control group makes it difficult to be certain whether this was directly related to the pregnancy itself. In a study with matched controls, Uebing and colleagues did describe a persistent increase in right ventricular size, but no decrease in right ventricular systolic function among women with repaired tetralogy who became pregnant, compared with tetralogy controls without pregnancy. There was no increase in cardiovascular complications reported in this group during the mean follow-up of 2.9 years. Uebing et al. used echocardiography for the assessment of right ventricular size and systolic function. Assenza et al. reported the long-term effects of pregnancy on right ventricular remodeling in women with repaired ToF ; using cardiovascular magnetic resonance imaging, they determined that right ventricular end-diastolic volume increased in those women who had been pregnant, as compared with women with repaired tetralogy and no previous pregnancy. This was especially true for those women who already had a severely dilated right ventricle at baseline. Nevertheless, there was no change in right ventricular systolic function and there was no adverse cardiovascular outcome in this group after a mean follow-up of 41 months. In a long-term follow-up study (median follow-up of 7.5 years) of 33 pregnant women who had undergone pulmonary valve repair or replacement for tetralogy or pulmonary stenosis, Metz et al. observed a significant increase in the number of patients suffering congestive heart failure or requiring cardiac surgery, when compared with nonpregnant controls. Furthermore, women who had had two or more pregnancies were even more likely to have a long-term adverse cardiac event.
As outlined in Chapter 16 on cardiomyopathy, previous work has concentrated on maternal and neonatal outcomes in patients with dilated cardiomyopathy (DCM) and in particular peripartum cardiomyopathy (PPCM). With improved diagnosis and treatment of PPCM, maternal mortality has fallen. Thus, counseling patients on the risk posed by pregnancy should be based on the specific type of cardiomyopathy and the clinical state of the patient, such as their NYHA functional class, size of the left ventricle, and ejection fraction. Currently, there is still relatively little literature on the effect of pregnancy on the long-term progression of the various types of cardiomyopathy other than DCM and PPCM. Grewal et al. suggest that pregnancy worsens the outcomes in DCM patients in the short to medium term compared with similar nonpregnant patients. They prospectively enrolled women with known idiopathic (84%) or doxorubicin-induced (16%) DCM between 1994 and 2008 (excluding women with PPCM). Nonpregnant women were matched with 32 pregnant women for age, diagnosis, degree of left ventricular dysfunction, and year of initial clinic visit. Event rates were low in patients with normal or mild left ventricular dysfunction (left ventricular ejection fraction [LVEF] ≥45%). However, comparing the 18 women with moderate or severe left ventricular dysfunction (LVEF <45%) over a 16-month period from the start of pregnancy, event-free survival was worse in the pregnant women (28±11% vs 83±10%), suggesting that pregnancy had a deleterious effect. Of note, the nonpregnant group of patients were on significantly more medication during the follow-up period; thus, it may be that cessation or reduction of anti-failure medication during pregnancy in women with pre-existing left ventricular dysfunction contributes to adverse outcome. Katsuragi et al. described outcomes in 29 patients who were diagnosed with DCM before conception or in the first 7 months of pregnancy and followed up until a year after delivery. Fifteen patients were known to have DCM prior to pregnancy and all remained in NHYA class I throughout the follow-up period. Out of ten patients who were diagnosed during pregnancy and did not terminate the pregnancy, the NYHA class deteriorated in all but one patient, with three dying from severe heart failure (one in the third trimester and two within 6 months of delivery). There were eight cases of termination (four in each patient group), with a fractional shortening of 15.2±3.1%; none of these patients showed a deterioration in cardiac function or NYHA class at 20 months. Thus, the short- to medium-term outcome appears to be worse for patients who present with DCM during pregnancy and then continue with it. These patients could be considered to have pregnancy-associated cardiomyopathy (PACM).
Chapter 16 described our current knowledge of long-term outcomes in PACM and PPCM in detail: the main predictor of maternal prognosis is the degree of left ventricular dysfunction at the time of diagnosis, whereas the predictor of recurrence of PPCM in future pregnancies is the degree of recovery of left ventricular systolic function. As with all heart failure studies, maximal medical anti-failure therapy has continued to change during the era of referenced literature; thus, even if we continue to follow up the current identified cohorts of patients with PPCM or DCM, our newly identified patients may not follow the same clinical course. With further advances in the understanding of the pathophysiology of PPCM, additional studies are required to assess the impact of new therapies such as bromocriptine.
Finally, the increasing recognition of genetic etiologies for cardiomyopathies has resulted in a population of women with known genetic mutations, who have yet to develop any clinical disease; some clinicians would now advise the screening of first-degree relatives of patients with PPCM to identify covert DCM, as well as other forms of familial DCM. As yet there are no outcome data on the potential effects of pregnancy in this subclinical population, particularly as to whether pregnancy may accelerate the time to clinical presentation.