Objective
The importance of maternal autoantibody levels in congenital heart block and elucidation of maternal factors that may reduce disease burden require further clarification.
Study Design
Pregnancies complicated by maternal anti-Ro antibodies from 2007 through 2011 were retrospectively reviewed.
Results
In all, 33 women were followed up throughout pregnancy. Semiquantitative maternal anti-La levels were significantly higher in pregnancies complicated by fetal heart block of any degree (median difference, 227.5; P = .04), but there was no difference in maternal anti-Ro levels. In all, 94% of fetuses maintained normal conduction when the mother was treated with hydroxychloroquine or daily prednisone therapy throughout pregnancy, compared to 59% in the untreated group (odds ratio, 0.1; P = .04).
Conclusion
Pregnancies complicated by fetal heart block did not have higher levels of maternal anti-Ro antibodies. Maternal anti-La level may be a useful predictor of fetal heart block. Maternal treatment with either hydroxychloroquine or daily low-dose prednisone throughout pregnancy may provide a protective effect.
Pregnant women who are positive for anti-Ro/SSA antibodies have approximately a 2% risk for the fetus developing congenital complete heart block. This risk increases to 16% when a previous child has had heart block and approaches 55% when maternal hypothyroidism is present in combination with anti-Ro antibodies. It has been demonstrated that mothers with antibodies against the 52-kd Ro protein have longer atrioventricular (AV) conduction times and are at greater risk. A recent prospective cohort study suggests that the amount of anti-Ro antibody, rather than simply its presence, is associated with damage to the fetal conduction tissue.
Despite this improved understanding, clinical management of these mothers and fetuses remains challenging. Two prior studies have shown that PR prolongation does not necessarily precede more advanced heart block. With no surrogate markers for fetal heart block known, frequent screening via fetal echocardiography of fetuses exposed to maternal anti-Ro antibodies to trend the mechanical PR interval remains a common clinical practice. The mechanical PR interval, obtained by using pulsed-wave Doppler to measure the time from the onset of atrial systole to the onset of ventricular systole, is a validated approach for fetal rhythm assessment. While routine use of dexamethasone for prophylaxis is not recommended due to its association with significant adverse events, several anecdotal reports demonstrate regression or stabilization of first- or second-degree fetal heart block after maternal steroid therapy. Recent work by Izmirly et al suggests that women with systemic lupus erythematosus who are treated with hydroxychloroquine throughout pregnancy have a lower risk of having a child affected by cardiac neonatal lupus and that hydroxychloroquine use may protect against the recurrence of cardiac manifestations in subsequent pregnancies. Additionally, beneficial effects of maternal antiinflammatory treatment have not been firmly established by prospective studies and routine use of these medications is not currently standard of care.
Given these continued challenges, the goals of this study were to determine the rate of congenital heart block in fetuses exposed to maternal Anti-Ro/SSA antibodies at our institution, determine the correlation between fetal heart block and maternal autoantibody levels, and investigate the efficacy of maternal therapy with antiinflammatory medications. We hypothesized that pregnancies complicated by congenital heart block would be associated with higher maternal levels of anti-Ro antibody. We also hypothesized that mothers treated with a daily antiinflammatory medication throughout pregnancy would be less likely to have a fetus with congenital heart block. To test these hypotheses, we conducted a retrospective review of appropriate patients managed at our institution.
Materials and Methods
This study was approved by the Duke University Hospital Institutional Review Board. Subjects were identified by reviewing the outpatient schedule and records from our fetal cardiology clinic and affiliated community pediatric cardiology clinics from 2007 through 2011. Records from our rheumatology and maternal-fetal medicine clinics were also reviewed for supplemental information. All pregnant women who were positive for anti-Ro/SSA antibodies and serially screened via fetal echocardiography throughout the pregnancy were included in the study. One subject with only 1 clinical assessment performed at our institution was excluded. Maternal information including age, parity, primary rheumatologic diagnosis, significant medical and family history, medications taken during the pregnancy, and their dosages were collected by reviewing the electronic medical record from our institution. No patients were directly contacted for data collection. Two subjects were concurrently enrolled in the 2010 Preventive IVIG Therapy for Congenital Heart Block (PITCH) trial, assessing the efficacy of maternal therapy with low-dose intravenous immunoglobulin (IVIG) as a preventive therapy for fetal heart block. Semiquantitative levels of anti-Ro, anti-La, anti-Smith, and anti-RNP antibodies were collected from immunology laboratory records from our institution, as well as via direct communication with referring providers and laboratories. Antibody levels analyzed at our institution were obtained using the AtheNA Multi-Lyte ANA-II Plus test system (Zeus Scientific Inc, Raritan, NJ) on 1 of 2 instruments, with a value of >120 arbitrary units considered positive. Our immunology laboratory performs biannual carryover and precision testing on these instruments, maintaining that each pipetting probe must have an average carryover of <3% and a coefficient of variation ≤5% to meet acceptance criteria. Results of maternal antibody levels drawn by referring physicians were obtained by direct communication with these providers and were included in the analysis only if the same methodology was used for sample processing. Each fetal echocardiogram performed during the pregnancy was independently reviewed. Analysis of the fetal echocardiograms included measurement of the mechanical PR interval, qualitative assessment of systolic function, assessment for the presence of pericardial or pleural effusions, endomyocardial fibroelastosis or other evidence suggestive of hydrops fetalis, and a careful assessment for structural heart disease. The mechanical PR interval was measured using simultaneous left ventricular inflow-outflow pulsed-wave Doppler aligned parallel to direction of blood flow. Gestational age and interval between fetal echocardiograms expressed in weeks was recorded. Infant status immediately following birth, as well as at follow-up evaluations, was also noted.
Descriptive statistics, including calculations of the mean, median, SD, and interquartile range for anti-Ro, anti-La, anti-Smith, and anti-RNP antibody levels were calculated. The distribution of the maternal autoantibody levels were assessed by standard numeric and graphic assessments. Patients with missing antibody level data or those with results available in different units of measure that could not be converted to units used for the majority of the sample were excluded from this portion of the analysis. Antibody levels in patients with fetal heart block of any degree vs those with normal conduction were compared. Differences between these groups were assessed by the Wilcoxon rank sum test. Data tables were also created demonstrating the proportion of mothers treated with daily hydroxychloroquine or prednisone who had fetuses affected by congenital heart block of any degree. Odds ratios (ORs), 95% confidence intervals (CIs), and P values for these proportional data were calculated by the Fisher exact test. Two-sided P values were used for all analyses. A level of statistical significance was set at .05 prior to data analysis. All data analysis was performed using R statistical software, version 2.13.1 (R Foundation for Statistical Computing, Vienna, Austria).
Results
In all, 33 subjects were followed up at our institution during the study period. Maternal information is summarized in Table 1 . Average subject age was 29.2 years. In all, 61% (n = 20) had a diagnosis of systemic lupus erythematosus. Other rheumatologic diagnoses included unspecified rheumatologic disease (n = 8, 24%), Sjögren syndrome (n = 3, 9%), spondyloarthritis with associated Crohn’s disease and autoimmune neutropenia (n = 1, 3%), and rheumatoid arthritis (n = 1, 3%). An average of 9 screening fetal echocardiograms was performed throughout each pregnancy. Twenty-three subjects had fetal echocardiograms performed every 2 weeks from 17-34 weeks’ gestation. Nine subjects were screened on a weekly basis for a portion of the pregnancy due to elevated concern for fetal heart block in the current pregnancy (n = 7), history of a pregnancy with fetal conduction abnormality (n = 3), history of fetal demise of unknown etiology (n = 1), or concomitant maternal hypothyroidism (n = 1). One subject whose fetus maintained normal conduction was screened every 4 weeks from 19-27 weeks’ gestation. All of the fetuses had structurally normal hearts. In all, 42% of subjects (n = 14) were treated with 200-400 mg/d of hydroxychloroquine throughout the pregnancy. Additionally, 2 patients were treated only with low-dose daily prednisone (1-20 mg once daily) and 6 subjects treated with both medications concurrently.
Patient No. | Age, y | Gravida status | Rheumatologic diagnosis | Adverse obstetrical outcomes in prior pregnancies | Autoantibody assessment timing (relative to gestation) | No. of fetal echocardiograms | Fetal rhythm diagnosis | Hydroxychloroquine therapy | Prednisone therapy |
---|---|---|---|---|---|---|---|---|---|
1 | 18 | 1 | SLE | None | +3 d | 7 | CHB a | N | N |
2 | 36 | 3 | Sjögren | CHB s/p pacer × 2 | 10 wk | 18 | CHB b | N | N |
3 | 24 | 2 | SLE | CHB, hydrops, intrauterine demise | 9 wk | 25 | CHB b | N | N |
4 | 28 | 2 | SLE | Prematurity | 16 wk | 12 | First-degree AVB | 200 mg BID | N |
5 | 31 | 2 | Unspecified | None | +1 mo | 15 | First-degree AVB | N | N |
6 | 36 | 2 | Sjögren, autoimmune hypothyroidism | SA × 1 | +7 mo | 23 | First-degree AVB; resolved second-degree AVB | N | N |
7 | 24 | 2 | SLE | Chondrodysplasia punctata | 9 wk | 7 | Normal | 400 mg qd | 15 mg qd |
8 | 25 | 1 | SLE | None | 20 wk | 6 | CHB a | N | N |
9 | 31 | 2 | SLE | SA × 1 | 4 wk | 7 | Normal | N | N |
10 | 32 | 1 | SLE | None | 13 wk | 7 | Normal | 200 mg qd | N |
11 | 27 | 1 | SLE | None | 8 wk | 8 | Normal | 200 mg BID | 5 mg qd |
12 | 26 | 2 | SLE | SA × 1 | 6 wk | 6 | Normal | N | N |
13 | 24 | 1 | SLE | None | −1 mo | 10 | Normal | 200 mg qd | N |
14 | 29 | 1 | SLE | None | 12 wk | 6 | Normal | 200 mg BID | N |
15 | 29 | 1 | Unspecified | None | 15 wk | 7 | Normal | N | N |
16 | 21 | 1 | SLE | None | 29 wk | 3 | Normal | 200 mg BID | 7.5 mg qd |
17 | 22 | 3 | SLE | None | 20 wk | 8 | Normal | N | N |
18 | 38 | 3 | ? Antiphospholipid syndrome | SA × 1 | 6 wk | 9 | Normal | N | N |
19 | 26 | 2 | Unspecified | None | 7 wk | 20 | First-degree AVB | N | N |
20 | 40 | 3 | Sjögren | None | 9 wk | 9 | Normal | N | N |
21 | 32 | 1 | Unspecified | None | 6 wk | 8 | Normal | N | N |
22 | 37 | 1 | SLE | None | −3 y | 9 | Normal | 400 mg qd | 1 mg qd |
23 | 29 | 1 | Spondyloarthritis, Crohn’s, autoimmune neutropenia | None | 9 wk | 10 | Normal | N | 5 mg qd |
24 | 39 | 2 | SLE | Stillbirth of twins | 14 wk | 6 | Normal | N | 20 mg qd |
25 | 22 | 1 | SLE | None | 9 wk | 7 | Normal | 200 mg BID | 5-10 mg qd |
26 | 27 | 1 | Unspecified | None | 17 wk | 11 | Normal | 400 mg qd | N |
27 | 32 | 5 | SLE | None | 14 wk | 2 | Normal | 400 mg qd | 15 mg qd |
28 | 40 | 3 | Unspecified | SA × 1; intrauterine demise × 1 | −1 y | 11 | Normal | N | N |
29 | 20 | 2 | SLE | None | 12 wk | 4 | Normal | 200 mg BID | N |
30 | 27 | 3 | Unspecified | None | NA | 3 | Normal | N | N |
31 | 30 | 2 | SLE | None | 25 wk | 9 | Normal | Y | N |
32 | 27 | 3 | RA | Intrauterine demise × 2 | NA | 6 | Normal | 200 mg qd | N |
33 | 37 | 2 | SLE | None | NA | 6 | Normal | N | N |
a Referred to our institution in CHB;
Fetal outcomes
Six fetuses (18%) developed congenital AV block of any degree during the pregnancy. Two fetuses rapidly progressed to complete fetal heart block at 19 and 22 weeks’ gestation, respectively. In the first case, the fetus had a normal mechanical PR interval at 18 weeks’ gestation and complete heart block by repeat assessment at 19 weeks. In the second case, the fetus developed first-degree AV block at 22 weeks at which time oral dexamethasone was prescribed. The mother developed significant nausea and vomiting and did not report this intolerance to her physicians; unfortunately, complete heart block was detected 2 days later. An additional 2 subjects were referred to our institution for evaluation with a diagnosis of complete fetal heart block that developed at 20 weeks’ gestation in both fetuses. There were no cases of intrauterine demise, endomyocardial fibroelastosis, or hydrops fetalis. All of the subjects carrying fetuses with complete heart block were treated with dexamethasone at a dose of 4 mg orally once daily. All of these fetuses remained in complete heart block at birth and 3 required insertion of a pacemaker in the neonatal period. All of these infants were still alive at the time of this report.
Three fetuses (9%) developed first-degree heart block in utero, with mechanical PR intervals ranging from 150-160 milliseconds. Each of these subjects was treated prophylactically with dexamethasone when prolongation of the PR interval was noted. In the case of the fetus with a mechanical PR interval of 150 milliseconds, dexamethasone was started due to an abrupt lengthening of the mechanical PR interval concurrent with an active lupus flare. Two of the 3 fetuses had regression of the mechanical PR interval to the normal range after initiation of therapy. No progression of the congenital heart block occurred throughout the pregnancy in any of the fetuses. Two of the infants had normal conduction immediately after birth and were clinically stable. One infant had continued first-degree AV block with a PR interval of 164 milliseconds by 12-lead electrocardiogram at last follow-up, but otherwise was doing well.
Interestingly, 1 fetus born to a subject with Sjögren syndrome and hypothyroidism was initially followed up for first-degree heart block, but progressed to Mobitz type 1 second-degree heart block (Wenckebach) at 19 weeks’ gestation. This subject was admitted to the hospital and treated with intravenous steroids and IVIG, with fetal rhythm subsequently reverting to first-degree heart block. The subject continued oral dexamethasone throughout the remainder of the pregnancy. The fetus remained stable in first-degree AV block throughout the remainder of the pregnancy, with persistent first-degree AV block and evidence of diastolic dysfunction by echocardiogram noted after delivery and through last assessment.
Maternal autoantibodies
Maternal autoantibody level data are summarized in Table 2 . Eighty percent (n = 24) of mothers had semiquantitative assessment of autoantibodies during the course of the pregnancy. In all, 96% of the laboratory samples included in this portion of the analysis were processed at our institution. Figure 1 demonstrates that subjects with pregnancies complicated by fetal heart block of any degree did not have significantly higher levels of anti-Ro antibodies (median difference, 38; P = .94). Maternal anti-La antibody level data are summarized in Figure 2 . Ten subjects had elevated anti-La antibody levels. All of the women with elevated anti-La antibody levels also had elevated anti-Ro antibody levels. Anti-La levels were significantly higher in pregnancies affected by fetal heart block of any degree compared to those with normal fetal conduction (median difference, 227.5; P = .04). An OR regarding elevated maternal anti-La antibody levels and development of congenital heart block was notably elevated at 6.2, however these results did not reach statistical significance (95% CI, 0.8–81.6; P = .07). There were no significant relationships found between maternal levels of anti-Smith or anti-RNP antibodies and the development of fetal heart block. There was no significant difference in anti-Ro or anti-La levels when mothers receiving antiinflammatory medications and the untreated group were compared. Two subjects, one carrying a fetus with complete heart block and the other with normal fetal conduction, also had significantly elevated antibody levels. These subjects were not included in the analysis due to the results being reported in different units of measurement that could not be accurately converted. Two subjects did not have quantitative antibody level data available, so they also were not included in this portion of the analysis.