Abstract
Objective. To study anti-Ro/La-negative congenital heart block (CHB).
Methods. Forty-five fetuses with CHB were evaluated by analysis of anti-Ro/La antibodies using sensitive laboratory methods.
Results. There were 9 cases of anti-Ro/La-negative CHB; 3 died (33.3%). Only 3 (33.3%) were complete in utero and 5 (55.5%) were unstable. No specific etiology was diagnosed. Six infants (66.6%) were given pacemakers. There were 36 cases of anti-Ro/La-positive CHB. All except 2 infants (94.4%) had complete atrioventricular block in utero. Ten died (27.8%), one (2.7%) developed severe dilated cardiomyopathy, and 26 (72.2%) were given pacemakers.
Conclusion. Nine of the 45 consecutive CHB cases (20%) were anti-Ro/La-negative with no known cause. They were less stable and complete than the anti-Ro/La positive cases.
Most cases of congenital heart block (CHB) detected in utero and unrelated to structural cardiac abnormalities are associated with anti-Ro/La antibodies, although the percentages of positive cases vary1–5. Descriptions of the prevalence and outcome of anti-Ro/La-negative atrioventricular (AV) block also tend to differ2–10. According to some investigators, mortality and morbidity are similar to those in anti-Ro/La-positive AV block3,4,7,8, while others report spontaneously reversible CHB and a more favorable course6,9,10.
MATERIALS AND METHODS
Forty-five consecutive fetuses with AV block were observed from 1990 to 2007 in 5 tertiary referral centers in Northern Italy (2 rheumatological, 2 cardiological, and one obstetric clinic). The inclusion criteria were congenital AV block detected in utero or at birth by fetal echocardiography and electrocardiogram. The exclusion criteria were structural cardiac abnormalities, congenital long QT syndrome11, mothers who had taken drugs during pregnancy that could induce fetal bradycardia, mothers who had had infectious diseases during pregnancy, or who had tested positive to hepatitis B/C viruses or human immunodeficiency virus, or to IgM anticytomegalovirus, Herpes or rubella virus and toxoplasma at the beginning of pregnancy.
Maternal sera were collected when CHB was detected and at delivery and tested for autoantibodies to Ro/SSA and La/SSB ribonucleoproteins using ELISA. Sera were tested a second time at the Padua University Hospital rheumatology laboratory employing a custom-designed counterimmunoelectrophoresis (CIE) method12; the fine specificities for 52-kDa and 60-kDa anti-Ro/SSA and 48-kDa anti-La/SSB were determined using a commercial ELISA (Diamedix, Delta Biologicals, Rome, Italy) and a line-blot assay (Inno-Lia, Innogenetics, Ghent, Belgium). Maternal sera negative to anti-Ro/SSA and anti-La/SSB antibodies were tested for confirmation by immunoblotting analysis using human salivary gland cell lysates, by ELISA using recombinant Ro52 protein, and by immunoprecipitation analysis with Ro60 and La in vitro translated proteins13, all at the laboratory of the University of Florida Department of Oral Biology. Maternal sera were also tested at the Padua laboratory for a battery of autoantibodies, including antinuclear (ANA), anti-dsDNA, anti-extractable nuclear antigens (ENA), anticardiolipin, and anti-ß2-glycoprotein I antibodies.
Statistical analysis was done using SPSS software, version 14.0.
RESULTS
Forty-five fetuses with CHB were examined. Thirty-six were born to anti-Ro/La-positive mothers (80%) and 9 to anti-Ro/La-negative mothers (20%). Negative maternal sera lacked reactivity to both Ro/SSA and La/SSB according to ELISA, CIE, and line-blot assays. These results were confirmed by the University of Florida Department of Oral Biology.
Anti-Ro/La-negative CHB infants (Table 1)
Five (55.5%) of the 9 infants were female. Three developed complete AV block in utero (Patients 1, 2, 3) and one also presented congenital sensorineural deafness. Two others had a stable second-degree AV block (Patients 4 and 5). Two fetuses had a second-degree AV block, one progressing to complete block soon after birth and the other at 3 months (Patients 6 and 7). The block alternated with normal sinus rhythm in the other 2 infants (Patients 8 and 9) and reverted to a stable normal sinus rhythm in Patient 9. Five blocks were unstable, changing their degree (nos 3,6,7,8,9).
Six blocks (66.6%) were detected in utero. Three (33.3%) were diagnosed at birth (Table 1 and 2) when cesarean delivery was needed because of fetal bradycardia. Six were given pacemakers. Two presented signs of heart failure in utero and 3 died (33.3%) shortly after birth. The 6 survivors had a mean age of 5.5 ± 3.5 years at the end of followup. Echocardiography showed no signs of cardiomyopathy or myocarditis.
Anti-Ro/La-negative mothers
Eight mothers were asymptomatic and one had photosensitivity and Raynaud’s phenomenon. All were negative for ANA, anti-dsDNA, anti-ENA, and antiphospholipid antibody. Eight had further pregnancies, with no recurrences. One had a family history of AV block (11.1%).
Anti-Ro/La-positive CHB infants (Table 3)
Twenty-five (69.4%) cases were female. None of the 34 presenting with complete AV block in utero reverted to a lesser degree (Table 2 and 3), but the 2 infants with incomplete block did. One of the fetuses with a second-degree AV block reverted to a normal sinus rhythm after therapy with high-dose dexamethasone and was born with a first-degree block that remained stable throughout the followup. The other fetus, with a second-degree AV block, reverted to an alternating pattern between second-degree block and normal sinus rhythm during treatment with dexamethasone 4 mg daily and weekly plasmapheresis; at age 27 months she developed a complete AV block.
AV block was always diagnosed in utero on the basis of fetal bradycardia (mean gestational age 23.03 weeks). There were 2 cases of sudden death in utero and 4 were aborted, 3 with severe heart failure. Three died immediately after birth due to heart failure and a fourth, who was given a pacemaker shortly after birth, died suddenly at age 21 months. Twenty-six neonates (72.2%) were given a pacemaker (Table 2 and 3); 4 of them died. Twenty-six infants were alive at the end of the followup, with mean age 6.6 ± 4.5 years. Echocardiography showed myocarditis in utero in 2 cases (5.5%). One male received a heart transplant at age 17 months because of severe dilated cardiomyopathy.
Anti-Ro/La-positive mothers
Eleven mothers were asymptomatic; 12 had Sjögren’s syndrome, 11 undifferentiated connective tissue disease, and 2 systemic lupus erythematosus. Thirteen (36.1%) were anti-La/SSB-positive. Only 4 had been diagnosed with connective tissue disease before the index pregnancy. Twelve had further pregnancies and there were no recurrences. There was a family history in one case (2.7%)14.
DISCUSSION
Cardiologists have been skeptical in the past about the existence of immune-mediated CHB, and some rheumatologists may still be doubtful about the existence of CHB not associated with anti-Ro/La antibodies. The exact percentages of anti-Ro/La antibody positivity in mothers of fetuses with CHB detected in utero and not associated with structural cardiac abnormalities are not known. Immunological studies using the most sensitive laboratory methods give positive results close to 100%, but these mothers were selected mainly from rheumatology centers2. Cardiological studies report lower percentages of anti-Ro/La-positive cases but do not always provide details about the laboratory methods3,6–10. Schmidt, et al3 and Maeno, et al7 reported 35% of anti-Ro/La negativity but did not indicate the laboratory tests. Using Ouchterlony double-diffusion and quantitative radioligand assays, Villain, et al5 reported that 55/111 cases (49.5%) of complete AV block detected before age 15 years were anti-Ro/La-negative. They also reported that 6/56 blocks (10.7%) detected in utero were anti-Ro/La-negative, and another 25 had to be excluded because their antibody status was not known5. Lopes, et al15 reported that 41/57 (71.9%) mothers of fetuses with isolated AV block “were seropositive for antinuclear antibodies, most often anti-Ro antibodies,” but did not indicate the test method. In agreement with our findings that report showed that spontaneous regression of AV block was possible in cases of anti-Ro/SSA-negative block and that mortality was similar in anti-Ro/SSA-positive and negative blocks, so these blocks too call for close followup.
Referral and ascertainment bias might have skewed our results. However, we used state of the art laboratory methods and took the extra step of investigating further the initially negative reactivity for anti-Ro/SSA and La/SSB. In fact there were concerns that the negative reactivity was the result of low sensitivity of some assays or of failure to detect conformational epitopes (e.g., Ro60), which are readily detectable with immunoprecipitation of radiolabeled recombinant Ro60. By including these extra steps, we are confident about the negative reactivity reported here.
It has been suggested that cases of complete in utero AV block are often anti-Ro/La-positive, while incomplete blocks are generally negative5,6,8–10. Our study is the first to specifically compare anti-Ro/La-negative CHB with positive cases (Table 2) using sensitive laboratory methods. Twenty percent of our consecutive CHB cases were negative, but only 8.1% were negative when the analysis was limited to AV blocks that became complete in utero.
As no definite etiology (such as tumors, myocarditis, drugs) was found in the cases of anti-Ro/La-negative CHB, these must be considered “idiopathic.” It is possible that there was some viral involvement, but it goes beyond the aims of our study to assess the role of viruses in the pathogenesis of CHB, whether anti-Ro/La-positive or negative. That one anti-Ro/La-negative female infant presented congenital sensorineural deafness is an interesting finding that warrants further study, since it has been correlated with congenital long QT syndrome11.
The early diagnosis of Ro/La-positive CHB cannot be explained by the fact that the mothers were attending rheumatology centers, as 32 of the 36 mothers were unknown to us before the blocks were detected.
The mortality rate was similar in the anti-Ro/La-negative and positive groups (33.3% and 27.8%, respectively), in agreement with Lopes, et al15, but the causes were different. No anti-Ro/La-negative fetuses developed cardiomyopathy. The cardiological courses of the cases of negative blocks were partially different for later gestational age and because they presented less often as complete, but often had a changing or progressive course.
While these findings have all the limitations of an observational study, there can be no doubt about the antibody negativity of the 9 cases we found. Although no single variable clearly distinguished Ro/La-positive from negative cases of CHB, the latter were often incomplete, were detected later during pregnancy, and were less stable.
Footnotes
-
Dr. Chan was supported in part by US National Institutes of Health grant AI47859.
- Received July 1, 2009.
- Accepted for publication March 2, 2009.