2015 Volume 79 Issue 4 Pages 854-861
Background: Because there is limited information on fetal bradyarrhythmia associated with congenital heart defects (CHD), we investigated its prognosis and risk factors.
Methods and Results: In our previous nationwide survey of fetal bradyarrhythmia from 2002 to 2008, 38 fetuses had associated CHD. Detailed clinical data were collected from secondary questionnaires on 29 fetuses from 18 institutions, and were analyzed. The 29 fetuses included 22 with isomerism, 4 with corrected transposition of the great arteries (TGA) and 3 with critical pulmonary stenosis; 14 had complete atrioventricular block (AVB), 8 had second-degree AVB, and 16 had sick sinus syndrome; 5 died before birth, and 10 died after birth (5 in the neonatal period). Neonatal and overall survival rates for fetal bradyarrhythmia with CHD were 66% and 48%, respectively. Pacemaker implantation was needed in 17 cases (89%). Beta-sympathomimetics were administered in utero in 13 cases and were effective in 6, but were not associated with prognosis. All cases of corrected TGA or ventricular rate ≥70 beats/min survived. A ventricular rate <55 beats/min had significant effects on fetal myocardial dysfunction (P=0.02) and fetal hydrops (P=0.04), resulting in high mortality.
Conclusions: The prognosis of fetal bradyarrhythmia with CHD is still poor. The type of CHD, fetal myocardial dysfunction, and fetal hydrops were associated with a poor prognosis, depending on the ventricular rate. (Circ J 2015; 79: 854–861)
Fetal bradyarrhythmia is an uncommon but life-threatening disease, especially in cases of complete atrioventricular block (AVB), which has a poor prognosis even in the postnatal period because of fetal hydrops and severe heart failure.1–4 Fetal bradyarrhythmia may occur with and without associated congenital heart defects (CHD). In fetuses without CHD, the association of AVB with maternal anti-Ro/Sjögren’s syndrome A antibodies is well established.5,6 These antibodies damage the conduction system and the myocardium, and lead to AVB and myocarditis, respectively, resulting in a poor prognosis. We previously reported that fetal hydrops is associated with a poor prognosis in fetuses without CHD7 and we suggested that myocardial dysfunction, rather than the severity of bradyarrhythmia, was associated with fetal hydrops. In contrast, in fetuses with CHD, congenital bradyarrhythmia is thought to result from the structural defect itself.8,9 An embryological insult in the early gestational phase has been proposed to be the pathogenic mechanism of conduction system defects. Heterotaxy and corrected transposition of the great arteries (TGA) often lead to discontinuity between the atrioventricular node and the ventricular conduction tissues, resulting in bradyarrhythmia,10 and sick sinus syndrome (SSS) often occurs with heterotaxy and worsens the bradyarrhythmia.
Editorial p 761
In the fetuses with bradyarrhythmia, an association with CHD was reported as a significant risk for the prognosis.11–17 However, there is limited information on the risks related to outcomes of fetal bradyarrhythmia associated with CHD. Although Jaeggi et al reported that in all survived cases there was a ventricular rate >60 beats/min, the data were not based on multivariate analysis because of the limited number of cases, especially survivors.18 Therefore, in this study we investigated the prognosis and risk factors for fetuses with bradyarrhythmia associated with CHD based on a large cohort in a Japanese database in recent years.
Our primary nationwide questionnaire survey has been described in detail.7 In brief, data were collected using questionnaires sent to Departments of Perinatology and Pediatric Cardiology at 750 institutions in Japan over 7 years (2002–2008). The response rate was 61% (455 institutions). Among 128 fetuses previously identified from 52 institutions, 38 (30%) had CHD, of which additional data were collected in 32 cases in a secondary survey.
All CHD cases were diagnosed prenatally using fetal echocardiography. Left or right atrial isomerism was confirmed by neonatal echocardiography, morphology of the auricle of the atrium during an operation such as pacemaker implantation (PMI), or postmortem examination. Fetal bradyarrhythmia was defined as a ventricular heart rate ≤100 beats/min. Fetal bradyarrhythmia was categorized into 3 subtypes: complete AVB, second-degree AVB (2:1 AVB), and SSS. Fetal SSS was defined as atrial heart rate ≤100 beats/min. These conditions were diagnosed prenatally using echocardiography or magnetocardiography, and confirmed in survivors by electrocardiography. Fetal myocardial dysfunction was subjectively evaluated by a pediatric cardiologist at each institution, based on contraction of the ventricles and the degree of cardiomegaly or atrioventricular valve regurgitation. In our secondary survey, the following additional perinatal data were collected: detailed anatomical features, fetal heart rate, fetal treatment, mode of pacemaker, operation, and outcomes during the follow-up period. Ventricular rate was defined as the minimum ventricular heart rate recorded in utero.
Statistical analysis was performed using STATA 11.1 (StataCorp LP, College Station, TX, USA) and JMP 10 (SAS Institute, Cary, NC, USA). Data are presented as the mean±SD or number of patients and were evaluated with Student’s t-test, or by Wilcoxon rank sum test when the data were not normally distributed. Categorical variables were evaluated using chi-squared test and Fisher’s exact test. Time to death was analyzed using the Kaplan-Meier method with log-rank test. We also conducted univariate and multivariate Cox proportional hazard model analyses. The best prediction model was selected by backward elimination with P=0.10 as a criterion for exclusion. Stepwise analysis was used to adjust for baseline variables. P<0.05 was considered significant in all analyses.
Of the 32 fetuses, 3 were excluded because of large atrial septal defect with maternal anti-Ro/Sjögren’s syndrome A antibodies, cardiac tumor and termination of pregnancy at 21 weeks of gestational age, respectively. Thus, a total of 29 fetuses were analyzed in the study. The baseline characteristics of the cases in this study compared with our previous study7 are shown in Table 1. There were 5 deaths before birth and 10 after birth (5 in the neonatal period and 5 after the neonatal period). All except 1 died from heart failure within 6 months of birth despite multidisciplinary treatment. The overall survival rate was 48% in a mean follow-up period of 36±42 months. The ventricular rate at diagnosis in fetuses with CHD did not differ significantly from that in fetuses without CHD, but the atrial rate was lower in fetuses with CHD because of a higher incidence of SSS. Fetal myocardial dysfunction was frequently associated in fetuses with CHD than in those without CHD (45% vs. 28%, P<0.01). The incidence of fetal hydrops, fetal death and neonatal survival did not differ between fetuses with or without CHD. Of 19 neonatal survival cases, a permanent pacemaker was needed in 17 (89%), which was more frequent than in fetuses without CHD (40%).
Data given as mean±SD or n (%). P<0.05, significant difference (†Student’s t-test; other data analyzed using chi-squared test and Fisher’s exact test). ‡Recorded in utero. AVB, atrioventricular block; CHD, congenital heart defects; FD, fetal death; GA, gestational age; SSS, sick sinus syndrome.
The 29 fetuses with CHD were morphologically categorized into 3 subtypes: 22 with isomerism (20 left atrial isomerism, 2 right atrial isomerism), 4 with corrected TGA, and 3 with critical pulmonary stenosis (PS; Table 2; Figure 1). There were no cases of abnormal karyotype. Of the 22 fetuses with isomerism, 15 (68%) had SSS, 9 had complete AVB, and 6 had second-degree AVB. Of the 15 cases of SSS, 5 were associated with complete AVB, and 3 were associated with second-degree AVB. There were major extracardiac anomalies that required an operation after birth in 2 cases of left atrial isomerism (1 case of biliary atresia and 1 of jejunal atresia), which were not associated with a poor prognosis. In 2 cases of left atrial isomerism, there was a single umbilical artery.
A rate, atrial rate; AAI, atrium paced, atrium sensed, and pacemaker inhibited in response to sensed beat; IIAVB, second-degree AVB; AVVR, atrioventricular valve regurgitation; Beta, β-sympathomimetics; CAVB, complete AVB; CAVVR, common AVVR; cPS, critical pulmonary stenosis; cTGA, corrected transposition of the great arteries; DDD, dual-chamber inhibits and triggers; DDI, dual-chamber inhibits; LAI, left atrial isomerism; LD, late death (>28 days after birth); MR, mitral regurgitation; NND, neonatal death (≤28 days after birth); PMI, pacemaker implantation; RAI, right atrial isomerism; TR, tricuspid regurgitation; V rate, ventricular rate (minimum ventricular rate=minimum ventricular heart rate recorded in utero); VVI, ventricle paced, ventricle sensed, and pacemaker inhibited in response to sensed beat. Other abbreviation as in Table 1.
Prenatal and postnatal courses of fetal bradyarrhythmia with congenital heart defects (CHD). Flow diagram shows the 3 subtypes and their outcomes. PMI, pacemaker implantation; PS, pulmonary stenosis; TCPC, total cavopulmonary connection; TGA, transposition of the great arteries.
All 5 fetal deaths were those with isomerism (4 left atrial isomerism, 1 right atrial isomerism) and associated fetal hydrops. In these cases, the ventricular rate ranged from 30 to 45 beats/min (median, 40) and all fetuses died in utero at 21–35 weeks of gestation (median, 32). Among the 8 neonatal and late deaths in cases of isomerism, 7 fetuses had fetal myocardial dysfunction, of which 3 had fetal hydrops. Of the 22 cases of isomerism, 11 fetuses had a common atrioventricular canal, of which 9 had common atrioventricular valve regurgitation ≥2nd degree. In the 13 neonatal survivors with isomerism, 12 needed PMI; 4 underwent a total cavopulmonary connection at ages of 1.4, 1.6, 4.8, and 7.0 years, respectively, and 4 underwent functional repair. Of the 4 cases of corrected TGA, there were 2 of complete AVB without SSS, and 2 of second-degree AVB with and without SSS. None of the survivors of corrected TGA had fetal myocardial dysfunction or fetal hydrops; all survived, and 3 needed PMI. All 4 had PS (3 mild, 1 severe), and 3 had a ventricular septum defect, for which 2 underwent functional repair at 0.9 and 2.0 years old, respectively. Of the 3 with critical PS, all had complete AVB without SSS, and in 2 cases PMI was needed. All had duct-dependent pulmonary circulation. Of 2 cases of fetal myocardial dysfunction and fetal hydrops, one neonate died soon after birth with no response to resuscitation, and the other survived with functional repair through open pulmonary valvotomy at 5 days old.
A β-sympathomimetic agent (continuous maternal infusion of ritodrine hydrochloride) was administered in utero in 13 cases, including 10 with left atrial isomerism, 1 with right atrial isomerism, 1 with corrected TGA, and 1 with critical PS. The ventricular heart rate increased by >10% in 6 cases (5 left atrial isomerism, 1 critical PS), and in 3 cases there was a temporary response. In the 6 responsive cases, 2 fetuses survived and 4 died after birth (1 in the neonatal period, 3 after the neonatal period). Of the 7 non-responsive cases, 2 fetuses survived and 5 died (3 in utero, 2 in the neonatal period). No fetus was treated with steroids or with an invasive intrauterine procedure such as PMI.
The type of bradyarrhythmia was not associated with prognosis. All fetuses with a ventricular rate ≥70 beats/min survived. Kaplan-Meier survival curves for the primary endpoint of all-cause mortality showed that fetal hydrops, fetal myocardial dysfunction and a ventricular rate <55 beats/min (Figure 2) were significant risk factors for death (P<0.01, log-rank test). The results of univariate and multivariate Cox model analyses (Table 3) demonstrated that development of fetal myocardial dysfunction (hazard ratio [HR], 14.06; 95% confidence interval [CI]: 1.42–139.10, P=0.02), fetal hydrops (HR, 10.71; 95% CI: 1.15–100.00, P=0.04) and ventricular rate had significant effects on death (HR, 0.94; 95% CI: 0.88–1.00, P=0.04). A ventricular rate <55 beats/min had significant effects on fetal myocardial dysfunction (HR, 22.14; 95% CI: 1.60–306.79, P=0.02) and fetal hydrops (HR, 7.04; 95% CI: 1.14–43.58, P=0.04).
Kaplan-Meier survival curves for the primary endpoint (all-cause mortality) for ventricular rate ≥55 beats/min and <55 beats/min. The follow-up period was defined as the time from diagnosis in utero to death. The time to death was analyzed using the Kaplan-Meier method with log-rank test. P<0.05 indicates a significant difference. bpm, beats/min.
P<0.05, significant difference. †Minimum ventricular heart rate recorded in utero; ‡minimum atrial heart rate recorded in utero. CI, confidence interval; HR, hazard ratio. Other abbreviation as in Tables 1,2.
There was a total of 12 cases of PMI in the neonatal period, of which 6 neonates survived and 6 died. Myocardial dysfunction tended to be more frequently associated in the cases of death than in survivors (100% vs. 33%, P=0.06). The mean ventricular rate at birth in the 2 groups was 61 beats/min. The pacing rate was decided by the heart rate giving maximum systolic blood pressure, except in one case in which the pacing rate was blindly configured to 150 beats/min. The mean pacing rate was significantly higher in the cases of death than in survivors (133 beats/min vs. 107 beats/min, P=0.04).
This study is one of the largest to investigate the prognosis and risk factors for fetal bradyarrhythmia associated with CHD. We found that the ventricular rate, type of CHD, myocardial dysfunction and fetal hydrops were important factors in predicting the outcome of fetuses with CHD. Fetuses with bradyarrhythmia associated with CHD had a worse prognosis than those without CHD, but our data also showed that those without risk factors had a relatively better clinical course. Hence, the outcome should not simply be predicted by the presence of CHD, but by associated factors in each case. For example, fetuses with corrected TGA and AVB had a good prognosis and those with ventricular heart rate ≥70 beats/min or good cardiac function had favorable outcomes. This is important information for clinicians and families in determining the appropriate management of affected fetuses.
A ventricular heart rate at diagnosis <55 beats/min had significant effects on fetal myocardial dysfunction and fetal hydrops, resulting in high mortality. Groves et al found that a ventricular heart rate <55 beats/min is a risk factor for fetal or neonatal death in fetuses with complete AVB without CHD,19 but more recent multivariate analyses by Maeno et al and Jaeggi et al indicated that a ventricular heart rate <55 beats/min was not a risk factor.20,21 Our previous study of fetal complete AVB without CHD also showed that the fetal ventricular heart rate was not associated with prognosis.7 In contrast, our current study revealed that the ventricular heart rate was related to the outcome in fetuses with bradyarrhythmia with CHD. Jaeggi et al also found that a fetal heart rate <60 beats/min was not only associated with fetal hydrops, but predicted an adverse outcome in complete AVB with major CHD.18 Thus, measurement of the ventricular heart rate is very important for predicting the outcome and prognosis of fetuses with bradyarrhythmia associated with CHD.
We found that the type of CHD was strongly associated with prognosis. Fetuses with corrected TGA and AVB had a good prognosis, as found in previous studies.11–14 A poor prognosis of fetal bradyarrhythmia with CHD seems to be related to a high incidence of left isomerism.15–20 The 5-year survival rates in all cases of left isomerism is low, ranging between 65% and 84%, even in recent studies, because of associated serious cardiac lesions.22–25 High intrauterine mortality in cases of left atrial isomerism with complete AVB is often preceded by development of heart failure.20,21 A high incidence of poor cardiac function and hydrops may partially explain the poor prognosis. We suggest that myocardial dysfunction could be ascribed to bradycardia and intrinsic ventricular dysfunction proceeding from the CHD.
Limited data are available on transplacental treatment of bradyarrhythmia associated with CHD. In a total of 10 cases of CHD such as left atrial isomerism, transplacental treatment was attempted using several types of β-sympathomimetics, but all but one died in utero or in the early neonatal period.1,18,20,26 The fetal ventricular heart rate is increased by β-sympathomimetics in some cases, but a benefit of fetal treatment on survival has not been shown. Our retrospective multicenter study also did not demonstrate a significant benefit of the use of β-sympathomimetics on survival. This result can be partly explained by an obvious bias in this study because fetal treatment tended to be performed in severe cases of fetuses with a low heart rate and poor cardiac function (Table 2). However, it is very important to remember that long-term i.v. β-sympathomimetic treatment in mothers is reported to be associated with side-effects and may harm fetal cardiac function. Further accumulation of data is needed to clarify the beneficial and adverse effects of fetal treatment for this condition.
PMI is a crucial postnatal treatment in cases of fetal bradyarrhythmia. We did not find an effect of different modes of PMI, but a better outcome occurred with a slower heart rate setting. Previous reports in neonates with AVB without CHD showed improvement of cardiac function by reducing the pacing rate.27,28 Thus, precise assessment of the appropriate setting for PMI may be required to optimize cardiac function, especially in cases of left isomerism with abnormal ventricular structure.
We found a relatively better prognosis than in studies performed outside Japan.1,18,26,29 Although the follow-up period of 36±42 months was relatively short, the neonatal survival rate in our study was 48%, compared with only 14–18% in the previous studies. The type of arrhythmia comprised only complete AVB in the other studies, but we found the type of arrhythmia was not a risk factor for death in this study. Interestingly, another study from Japan (Maeno et al20) found similar outcomes to those in the current study. This result is partially explained by the low rates of termination of pregnancy in Japan. Furthermore, the other studies were performed 10–20 years ago and the improved outcomes may reflect the marked progress in neonatal treatment such as PMI and surgery.
The main limitation to this study relate to retrospective data selection bias and the relatively small sample size. First, the nature of a multicenter retrospective observational study using a questionnaire is such that the clinical data obtained varied among cases. Clinical management during and after the fetal period also varied among centers. However, this was a strong dataset of current clinical courses because the response rate for the primary questionnaire exceeded 60% and data were gathered from tertiary pediatric cardiac centers and several regional perinatal institutions over a wide area of Japan. Second, the diagnosis of fetal myocardial dysfunction was made comprehensively and subjectively in different institutions, so we could not show the different variables of myocardial dysfunction based on the criterion of contraction, degree of cardiomegaly, or atrioventricular valve regurgitation. Because myocardial dysfunction included multiple findings, it might be more appropriate to use each finding rather than myocardial dysfunction as 1 variable for analysis. However, we considered it very important that this subjective finding made by the physicians who performed fetal echocardiography was related statistically to the outcome. Because one of the purposes of this study was to plan a future prospective study of managing fetal bradyarrhythmia with CHD, we decided to leave this subjective finding in order to emphasize in the future study the importance of objectively measuring cardiac function. Finally, termination of pregnancy is limited after 22 weeks of gestational age in Japan. Thus, some early and severe cases in which termination of pregnancy was performed may not have reached a regional perinatal center and would not have been identified by the questionnaire, resulting in the better outcomes in Japan. Nevertheless, we think that our results reflect more accurate outcomes of natural clinical courses compared with previous studies containing the selection bias of termination of pregnancy.
Neonatal and overall survival rates for fetal bradyarrhythmia with CHD were 66% and 48%, respectively. In fetal bradyarrhythmia with CHD, the type of associated CHD, fetal myocardial dysfunction, fetal hydrops, and ventricular rate were associated with a poor prognosis. Larger prospective studies are required to explore long-term morbidity and to establish appropriate treatment strategies for fetal bradyarrhythmia with CHD.
This work was supported by a grant from the Ministry of Health, Labour and Welfare of Japan (Health and Labour Science Research Grants for Clinical Research for New Medicine).
Competing Interests: None. Ethics Approval: Approved by the Institutional Review Board of the National Cerebral and Cardiovascular Center (M21-13).