Objective
We assessed the impact of pregnancy on long-term cardiac outcomes in women with prior surgery for congenital pulmonary valve anomalies.
Study Design
Data on all reproductive age women with prior pulmonary valve repair or replacement, cared for at a tertiary institution over a 10-year period, were analyzed. Kaplan-Meier curves and proportional hazards models were estimated to assess the impact of pregnancy and multiparity on a composite long-term adverse outcome defined as death, heart failure, or unanticipated cardiac surgery. Peripartum cardiac complications were also assessed.
Results
Thirty-three parous and 20 nulliparous, nonpregnant controls with primary pulmonary valve replacement or repair were identified. Among the parous women, there were 95 pregnancies (median, 3.0; 1–10) resulting in 81 live births. Peripartum cardiac complications occurred in 28 (29.8%; 95% confidence interval, 20.4–39.2) of the pregnancies. A composite adverse long-term cardiac outcome occurred in 17 of 33 parous women, over 417 person-years (4 per 100 person-years) and 1 of 20 nulliparous women over 258 person-years (0.4 per 100 person-years); women with pregnancies were more likely at any point in time to have a composite long-term adverse cardiac outcome compared with nulliparous controls. Women with 2 or more pregnancies were more likely to have a composite adverse cardiac outcome than those with less than 2 pregnancies (hazard ratio, 8.8; 95% confidence interval, 1.5–50.3).
Conclusion
Peripartum cardiac complications are common in women with prior pulmonary valve repair or replacement. Pregnancy appears to increase the risk of long-term adverse cardiac outcomes in these patients when compared with nulliparous controls.
Congenital abnormalities of the right ventricular outflow tract range from more simple lesions such as isolated pulmonary stenosis to more complex lesions including tetralogy of Fallot (ToF). With ToF, pulmonary stenosis is accompanied by the presence of a ventricular septal defect as well as an overriding aorta. Given the many advances in cardiosurgical techniques that have occurred over the last few decades, patients with the above-mentioned congenital cardiac anomalies typically survive well into adulthood, thus reaching child-bearing age. Despite prior surgical intervention, these patients typically have residual cardiac lesions, which may include stenosis or insufficiency of the pulmonary valve. Such lesions may be exacerbated by the increase in right ventricular cardiac output that occurs during pregnancy.
As more of these women opt to pursue child-bearing, physicians are being asked to counsel them regarding both pregnancy and cardiac outcomes. Initial data suggested that women with prior surgical repair of the right ventricular outflow tract did well during pregnancy with a relatively low rate of complications and normal recovery in the immediate peripartum period. More recently, however, several authors have reported adverse peripartum cardiac outcomes in these women including arrhythmias and heart failure with some suggestion that this effect may be cumulative with multiple pregnancies. We sought to evaluate peripartum outcomes and assess the impact of pregnancy on long-term cardiac outcomes in women with prior surgical intervention on the pulmonary valve for congenital abnormalities.
Materials and Methods
This was a retrospective study of all women (older than 18 years) with prior surgical intervention for congenital anomalies of the pulmonary valve cared for at a tertiary institution over a 10 year period from 2001 to 2011. Participants were ascertained from the Adult Congenital Heart Disease database at our institution. Patients were divided into 2 groups: (1) those with a history of a pregnancy attaining 20 weeks’ gestation and (2) a nulliparous, nonpregnant control group. Patients were excluded from further analysis if there was documentation of an underlying genetic syndrome including 22q11. This study was approved by the University of Utah Institutional Review Board.
Demographic, obstetrical, neonatal, and cardiac data were abstracted from the electronic medical record where applicable. Recorded demographic data included current age and body mass index (BMI) at last follow-up and for the pregnant cohort, age at each pregnancy, outcome of each pregnancy, prepregnancy BMI, and number of pregnancies.
Recorded obstetric complications included: preeclampsia, gestational diabetes, preterm premature rupture of membranes (PPROM), placental abruption, and/or perinatal death. Other recorded obstetric variables included gestational age at delivery, indication for and mode of delivery, neonatal birthweight, presence or absence of neonatal congenital heart disease, and neonatal survival.
Recorded cardiac variables for the pregnancy group included New York Heart Association (NYHA) functional class at time of conception, time of delivery, and at last follow-up, cardiac medications at the time of conception, initiation of cardiac medications during the pregnancy, presence of a right ventricle to pulmonary artery (PA) conduit including type of conduit (homograft vs heterograft valved conduit), and age of conduit at pregnancy. Peripartum (antepartum or within 6 weeks of delivery) complications were also recorded.
Cardiac data collected for the entire cohort included NYHA functional class at baseline, presence or absence of arrhythmias requiring treatment, congenital cardiac diagnoses, number of cardiac surgeries, and interval between right ventricular (RV)-PA conduit surgeries. Echocardiographic data were reviewed on all patients and the following were documented: the degree of pulmonic and tricuspid insufficiency, the degree of pulmonary valve stenosis, right ventricular size and function, and left ventricular function. For the pregnancy cohort, measurements were obtained before, during, and after pregnancy and at last follow-up. For the nulliparous group, baseline measures were obtained at the time of the first adult visit. For comparisons between the groups, baseline for the parous women was set at the time of the echocardiogram most closely preceding their first pregnancy.
A composite adverse outcome, defined as death, stroke, left ventricular dysfunction (ejection fraction less than 50%), symptomatic right ventricular dysfunction requiring treatment outside the peripartum period, or unanticipated cardiac surgery, was noted to be present or absent for all patients. Unanticipated cardiac surgery was defined as surgical or percutaneous intervention for RV-PA conduit dysfunction within 10 years of placement or surgical intervention on a cardiac valve other than the pulmonary valve. Age at follow-up was defined as the age at adverse outcome or, in patients without an adverse outcome, the age at the last clinic visit.
Pulmonary valve dysfunction was deemed present when the patient was noted to have moderate or severe pulmonary insufficiency on echocardiography. Right ventricular enlargement was categorized into 3 groups based on the following echocardiographic data: (1) none, (2) mild, and (3) moderate to severe.
All women in the cohort who achieved pregnancy were followed up throughout pregnancy and the postpartum period by the Perinatology and Adult Congenital Cardiology services.
Demographic and clinical characteristics of women were summarized as frequency and percent. Characteristics between parous and nulliparous women, and parous women with/without an adverse cardiac outcome were compared using a Student t test for continuous variables or a χ 2 or Fisher exact test for comparison of dichotomous variables.
Time (from age 19 years for nulliparous women and first pregnancy for parous women) to composite adverse outcome was estimated with a Kaplan-Meier analysis stratified by parity. Adjusted hazard ratios for composite adverse outcome were estimated with Cox proportional hazards models with time-varying age at pregnancy, complication at pregnancy, and pregnancy number. A linear mixed-effects model was estimated to test whether cardiac parameters (pulmonary valve dysfunction and right ventricular enlargement) worsened with successive pregnancies. Statistical significance was defined as a 2-sided value of P < .05.
Results
Thirty-three parous and 20 nulliparous controls with a history of primary pulmonary valve replacement or repair were identified. Among the parous women, there were 95 pregnancies (median 3.0; 1–10), resulting in 81 live births. Twenty-nine women (84.4%) were multiparous (more than 1 pregnancy beyond 20 weeks’ gestation).
The parous and nulliparous groups were similar in type of cardiac lesion, with 73% of the parous group and 80% of the nulliparous group having ToF and the remainder having primary pulmonic stenosis ( P = .36). There were no significant differences in their age, BMI, or baseline cardiac characteristics. Baseline demographics for both groups are reported in Table 1 .
| Characteristic | Parous (n = 33) | Nulliparous (n = 20) | P value |
|---|---|---|---|
| Age at event or most recent visit, mean ± SD | 32.2 ± 7.6 | 31.8 ± 12.5 | .91 |
| Current BMI (kg/m 2 ) ± SD | 27.7 ± 7.5 | 26.0 ± 7.2 | .42 |
| Number of cardiac surgeries, mean ± SD | 1.9 ± 1.2 | 2.6 ± 1.3 | .08 |
| Type of cardiac lesion, n (%) | |||
| Tetralogy of Fallot | 24 (72.7) | 16 (80.0) | .36 |
| Primary pulmonic stenosis | 9 (27.3) | 4 (20.0) | |
| Baseline NYHA classification II or greater | 0 (0) | 2 (10.0) | .14 |
| Baseline arrhythmia | 4 (12.1) | 2 (10.0) | 1.00 |
| Moderate to severe pulmonary insufficiency, n (%) | 9 (27.3) | 8 (40.0) | .34 |
| Mild to moderate pulmonary stenosis, n (%) | 9 (27.3) | 6 (30.0) | .83 |
Among the parous women, all except one were NYHA class I at the onset of each pregnancy. Worsening of NYHA functional class occurred with 16 pregnancies (16.8%). The mean age at first pregnancy was 24.1 ± 5.6 years. The parous cohort was relatively thin with a mean BMI prior to the first pregnancy of 24.0 ± 4.9 m/kg 2 . On average, each parous study subject had 2.9 ± 1.9 pregnancies and 2.4 ± 1.2 deliveries beyond 20 weeks’ gestation.
Cardiac data were analyzed by pregnancy. Approximately 40% of pregnancies occurred in 15 women with an RV to PA conduit in place. Average conduit age at the time of pregnancy was 8.1 ± 1.0 years. Thirty-five of the pregnancies (37.2%) occurred in women who had undergone at least 1 additional cardiosurgical intervention beyond the initial repair of their congenital cardiac defect. Complete echocardiographic data were available for 70 pregnancies. Baseline cardiac characteristics and echocardiographic data are presented in Table 2 .
| Characteristics | Number of affected pregnancies |
|---|---|
| Cardiac characteristics (n = 95) | |
| Cardiac medication use before pregnancy, n (%) | 6 (6.3) |
| Right ventricular to pulmonary artery conduit in place, n (%) | 37 (38.9) |
| Type of conduit, n (%) | |
| Cadaveric (pulmonary/aortic homograft or allograft) | 17 (17.9) |
| Synthetic (porcine or bovine) | 20 (21.1) |
| Arrhythmia before pregnancy, n (%) | 10 (10.5) |
| Echocardiographic data (n = 70) a | |
| Pulmonary insufficiency before pregnancy, n (%) | |
| None | 28 (40.0) |
| Mild | 17 (24.3) |
| Moderate | 20 (28.6) |
| Severe | 5 (7.1) |
| Pulmonary stenosis before pregnancy, n (%) | |
| None | 47 (67.1) |
| Mild | 21 (30.0) |
| Moderate | 2 (2.9) |
| Severe | 0 (0) |
| Right ventricular enlargement before pregnancy, n (%) | 32 (45.7) |
| Right ventricular dysfunction before pregnancy, n (%) | 11 (15.7) |
| Tricuspid insufficiency before pregnancy, n (%) | |
| None | 51 (72.9) |
| Mild | 18 (25.7) |
| Moderate | 1 (1.4) |
| Severe | 0 (0) |
| Left ventricular dysfunction before pregnancy, n (%) | 2 (2.9) |
a There were 70 pregnancies with complete echocardiographic data available.
Cardiac complications occurred in 28 pregnancies (29.8%; 95% confidence interval [CI], 20.4–39.2) including new-onset congestive heart failure requiring medical therapy or bed rest in 22 (23.4%), pulmonary emboli in 5 (5.3%), bacterial endocarditis in 1 (1.1%), cerebrovascular accident in 1 (1.1%), and new-onset arrhythmias in 9 (9.6%). Progressive pulmonary valvular obstruction or insufficiency occurred in 18.8% and 34.8% of pregnancies, respectively.
Complete delivery data were available for 60 of the live-born neonates. The mean gestational age of live births was 37.1 ± 3.7 (range, 24–41) weeks. Premature delivery occurred in 23.3% (95% CI, 12.3–34.4). Three of these 14 preterm births were medically indicated for worsening cardiac status, 4 were for spontaneous preterm birth or PPROM, and the remaining 7 were medically indicated for preeclampsia. Sixteen live-born neonates (26.7%) were delivered via cesarean delivery, and 4 (6.7%) women had an operative vaginal delivery. All others had spontaneous vaginal deliveries. The average birthweight of live-born neonates was 2769.3 ± 630.1 g. Seven neonates (11.7%) were small for gestational age. Congenital heart disease was diagnosed in 5 of the neonates (8.3%) and perinatal death in 4 (6.7%). The neonatal heart disease was classified as follows: 3 with ToF, 1 with hypoplastic left heart syndrome, and 1 with pulmonary stenosis and an atrial septal defect.
Women were followed up for a median of 7.5 years (range, 1–32 years) after initial delivery. A composite long-term adverse cardiac outcome of unanticipated cardiac surgery, heart failure, or death was assessed. The long-term adverse outcome occurred in 17 of 33 (51.5%; 95% CI, 33.5–69.5) parous women over 417 person-years (4 per 100 person-years). The composite long-term adverse outcome occurred in only 1 of the nulliparous women over 258 person-years (0.4 per 100 person-years).
Some of the women had more than 1 of the components of our composite long-term adverse outcome. Details of the long-term cardiac complications are included in Table 3 . A comparison of the baseline characteristics between parous women who developed a long-term adverse cardiac outcome and those who did not is included in Table 4 . The groups differed only by the number of deliveries greater than 20 weeks’ gestation, with the adverse cardiac outcome group having a mean of 2.9 deliveries and women without an adverse outcome having a mean of 1.8 deliveries ( P = .007).
