Fetal congenital heart disease - mode of delivery and obstetrical complications

Congenital heart diseases (CHDs) are the most common fetal structural anomalies, affecting nearly 1% of births annually in the United States [1‐4]. During the last decades, antenatal diagnosis of cardiac anomalies has increased due to enhanced prenatal imaging including fetal sonography and echocardiography [5‐7].

Most fetal CHD studies are grouped by the time of diagnosis, namely, before or after delivery [2, 7‐12]. With prenatal diagnosis, delivery can be planned for a tertiary center where the expected cardiac intervention will be performed following prompt multidisciplinary evaluation by the appropriate cardiac, surgical, and neonatal staff [8, 11]. However, the optimal mode of delivery in cases of fetal CHD is not well determined [10]. The presumed neonatal advantage of cesarean delivery (CD) has yet to be established and weighed against the potential maternal risk [8]. Comparisons of women with a prenatal or postnatal diagnosis of CHD showed that prenatal diagnosis decreased the likelihood of spontaneous labor [10] and increased the likelihood of a planned CD [2, 10, 13]. Nevertheless, a practice of elective CD was not found to be associated with decreased neonatal morbidity and mortality compared to attempted vaginal delivery; rather, it was associated with both higher rates of maternal morbidity as well as longer maternal and neonatal hospitalization [13‐16]. In contrast, several studies demonstrated that a prenatal diagnosis of CHD increased the odds of induction of labor but not of CD [12, 17]. Levi et al. [17] found no significant difference in CD rate between women with a prenatal or postnatal diagnosis of fetal CHD.

Few studies have compared obstetrical and outcome parameters of pregnancies complicated by fetal CHD with normal pregnancies. Some suggested that mothers carrying fetuses with CHD were at increased risk for CD [18‐20] and instrumental delivery [20]. Ge et al. [19] found that the most common indication for CD in this patient group was non-reassuring fetal heart rate (NRFHR).

Most studies of the mode of delivery in women diagnosed with fetal CHD compared CD to vaginal delivery, without considering operative vaginal delivery (OVD) and without detailing the indications for such interventions. The aim of the present study was to determine whether fetal CHD affects the mode of delivery during a trial of vaginal delivery.

Of the 237 women diagnosed with fetal CHD during the study period, 105 were eligible for the study (CHD group). The main reasons for exclusion were patent ductus arteriosus (n = 40), patent foramen ovale (n = 36), and hypoplastic heart syndrome (n = 20), followed by pulmonic stenosis (n = 16), tricuspid regurgitation (n = 7), aortic stenosis (n = 5), and mitral regurgitation (n = 1). All diagnoses were made prenatally. The most common major cardiac defect was coarctation of the aorta)25.71% of total CHDs). Other cardiac anomalies included transposition of the great arteries (22.85%), ventricular septal defect (21.9%), tetralogy of Fallot (15.23%), atrial septal defect (7.61%), Ebstein anomaly (4.76%), and atrioventricular septal defect (1.9%). The control group consisted of 511 women carrying a healthy singleton fetus within the same period.

The baseline demographic, obstetrical, and clinical characteristics of the cohort are summarized in Table 1. There were no between-group differences in maternal age, body mass index, parity, and gestational age at delivery, as dictated by the matching process. Rates of prior CD were similar in the two groups. A significantly higher proportion of women in the CHD group were smokers (4.76% vs 1.17%, p = 0.01). There was no significant difference between the groups in the rate of chronic diseases (20.95% vs 15.68%, p = 0.18), although a higher proportion of women in the CHD group had hypothyroidism (9.52% vs 3.52%, p < 0.01).

The obstetrical outcomes are presented in Table 2. There were no between-group differences in rates of gestational diabetes, hypertensive disorders of pregnancy, amniotic fluid disorders, and postpartum hemorrhage. The CHD group had a significantly higher rate of NRFHR (21.9% vs 13.3%, OR 1.82, 95% CI 1.07–3.1, p = 0.02) and chorioamnionitis (3.8% vs 0.78%, OR 5.02, CI 1.23–20.4, p = 0.01). Both these differences remained significant after adjustment for gestational week at delivery (aOR 1.84, 95% CI 1.08–3.135, p = 0.02; aOR 5.87, 95% CI 1.41–24.41, p = 0.01).

Regression analysis of the mode of delivery revealed a higher rate of OVD in the CHD group than the control group (18.09% vs 9.78%, OR 2.03, 95% CI 1.13–3.63, p = 0.01), although rates of CD were similar (9.52% vs 10.76%, OR 0.97, CI 0.47–1.98, p = 0.93). The association with OVD remained significant after adjustment for nulliparity and gestational age at delivery (OR 2.58, 95% CI 1.36–4.9, p < 0.01). The most common indication for OVD was NRFHR, with a higher prevalence in the CHD group (78.94% vs 64%), followed by prolonged second stage (21.05% vs 34%, respectively). Neither of these differences was statistically significant.

Perinatal outcomes are summarized in Table 3. There was no significant difference between the groups in newborn sex distribution (males: 56.19% vs 54.01%, p = 0.68). The CHD group was characterized by a significantly lower median birth weight percentile (45th vs 53rd percentile, p = 0.04) and, accordingly, a significantly higher frequency of small-for-gestational-age neonates (11.42% vs 5.87%, p = 0.04, OR 2.06, 95% CI 1.02–4.18). The association did not remain significant after adjustment for gestational age at delivery (aOR 1.97, 95% CI 0.96–4.04, p = 0.06). The CHD group also had higher rates of 1-minute and 5-minute APGAR scores ≤7 (15.42% vs. 2.94%, p < 0.01, OR 3.86, CI 1.72–8.66; 6.96% vs. 0.19%, p = 0.01, OR 7.47, 95% CI 1.23–45.27), but only the difference in 1-minute APGAR score < 7 remained significant after adjustment for mode of delivery and gestational week of delivery (aOR 4.12, 95% CI 1.74–9.78, p < 0.01; aOR 5.66, 95% CI 0.89–35.63, p = 0.06). The CHD group had a significantly higher rate of meconium-stained amniotic fluid on regression analysis (21.9% vs 11.74%, p < 0.01, OR 2.1, 95% CI 1.23–3.6) and after adjustment for gestational week of delivery (aOR 2.3, 95% CI 1.33–4, p < 0.01).

Cyanotic heart defects (tetralogy of Fallot, transposition of the great arteries, Ebstein anomaly) were identified in 45 cases in the CHD group (Table 4). Comparison of this subgroup with the control group yielded no differences in baseline demographic, clinical, and obstetrical characteristics. On regression analysis, the cyanotic-CHD subgroup had a nearly significantly higher rate of OVD than the control group (20% vs 9.78%, OR 2.14, CI 0.97–4.74, p = 0.058) and a similar rate of CD (4.44% vs 10.76%, OR 0.43, CI 0.1–1.85, p = 0.26). After adjustment for nulliparity and gestational week of delivery, the association with OVD was significant (OR 2.6, 95% CI 1.06–6.35, p = 0.03).

Our results demonstrate that fetal CHD is associated with higher rate of OVD, indicated mainly for NRFHR, and has a negative impact on perinatal and obstetrical outcomes such as birth weight, APGAR score, and rates of meconium-stained amniotic fluid, chorioamnionitis, and NRFHR.

Previous studies have dealt with the importance of prenatal diagnosis of CHD [2, 7‐12]. Whether the mode of delivery in these cases differs significantly from the general population remains unclear as do the indications for non-vaginal delivery. There is little mention in the literature of the rate of OVD and of intrapartum CD following a trial of labor. Hence our findings merit attention. Additionally, in contrast to the present study, the few studies that have addressed these issues [2, 7‐12] compared the findings by time of diagnosis of CHD.

Our results are in line with the 2006 Swedish population-based study of 6343 fetuses with CHD [20]. The authors found that relative to the general population of pregnant women, women carrying a fetus with CHD had significantly higher rates of instrumental delivery (OR 1.21, 95% CI 1.10–1.34) and CD, either elective or non-elective (OR 1.91, 95% CI 1.79–2.03). However, the indications for instrumental delivery and CD were not evaluated.

Walsh et al., [18] using a similar study design to ours, compared women carrying a fetus with a major CHD to women with a non-anomalous singleton pregnancy and found that the study group had a higher rate of emergency CD, mostly attributable to NRFHR)10% vs 4.8%). The statistical significance was not maintained when the analysis was adjusted for the presence of major extracardiac anomalies. These findings were supported by another recent study demonstrating an increased risk for primary and non-primary CD in pregnancies complicated by CHD relative to non-anomalous pregnancies; again, the most common indication for primary CD was NRFHR (49% vs 24%) [19]. However, unlike the present study, these earlier investigations did not mention OVD. Furthermore, and more importantly, we included only fetuses with isolated CHD without major extracardiac malformations.

Indeed, we believe that our restricting the study group to cases of isolated CHD and excluding cases of elective CD may account for why the CD rate was not significantly higher in the CHD group than the control group (9.52 and 10.76%, respectively).

Our analysis of the obstetrical and perinatal outcomes of women with a diagnosis of fetal CHD is supported by earlier studies [18, 20, 25]. Walsh et al. [18] also showed that NRFHR is more common in cases of fetal CHD than in controls, and Cedergren et al. [20] reported a higher rate of meconium aspiration, fetal distress, and small-for gestational age among fetuses with CHD.

There are several potential mechanisms that might explain our results. First, fetuses with a cardiac anomaly are at increased risk of heart rate abnormalities and are therefore more prone to distress and, consequently, a higher likelihood of emergency delivery. Ueda al [24]. speculated that cardiac abnormalities are associated with abnormalities in fetal heart rate patterns and found that fetuses with CHD have a higher rate of emergency CD due to NRFHR. Second, fetuses with CHD may have lower tolerance and less reserve for labor, either owing to the cardiac anomaly itself or to overall smallness; therefore, their risk of OVD or CD due to NRFHR is higher [18]. Finally, as in the case of elective CD, physicians may tend to recommend emergency delivery due to their perception of fetal vulnerability. This is supported by several studies showing that a prenatal diagnosis of CHD could affect the mode of delivery and is associated with a higher rate of elective CD [2, 10].

Our study demonstrated that OVD, especially for an indication of NRFHR, is more common in pregnancies complicated by fetal CHD than in the general population, after exclusion of elective cesarean sections. In addition, fetal CHD was associated with higher rates of small for gestational age, meconium-stained amniotic fluid, 1-minute and 5-minute APGAR scores ≤7, NRFHR, and chorioamnionitis.

Physicians should be alerted to these findings and take them into consideration when deciding on mode of delivery. They may also be of help when counseling mothers with a prenatal diagnosis of CHD to regarding the possibility of a safe trial of labor. However, further large prospective studies are warranted.