The recent report by Clancy et al on HLA class II and tumor necrosis factor α associations in neonatal lupus (1) confirms that the HLA portion of the extended haplotype DQB1*02;DRB1*03 provides the genetic predisposition for generation of anti-Ro/La antibodies that cross the placenta and bind apoptotic neonatal keratinocytes and fetal cardiocytes, thus initiating an inflammatory response. Those authors thus suggest that the risk for induction of neonatal lupus may be genetically transmitted. Congenital heart block (CHB), the main feature of neonatal lupus, has been found in 1 or more offspring of the same mother (in siblings or twins) (2, 3), but it has not been described to occur in family members of different generations, even if dated cardiology records confirm its existence (4). We recently observed a family in which both the mother and the daughter had anti-SSA/Ro antibodies and transmitted CHB to their respective infants, as described below.
A 26-year-old woman was referred to us in August 2001, during the eighth week of her first pregnancy. Sjögren's syndrome had been diagnosed 1 year earlier on the basis of dry mouth, dry eyes, anti-SSA/Ro antibodies (52 and 60 kd), and a first-degree lymphocytic infiltration in a minor salivary gland. The family history revealed CHB in her only sister. A normal heart rate and anatomy were recorded during the first fetal echocardiography performed at week 17 of gestation. At week 18 of gestation the patient exhibited swelling and erythema of her toes, particularly when standing. She did not report having pain or other symptoms, and results of arterial and venous Doppler ultrasonography were normal. Fetal ultrasonography at week 19 revealed a normal heart rate.
Sonography performed at week 22 revealed bradycardia in the fetus, which was confirmed by the finding of an advanced second-degree atrioventricular (AV) block on fetal echocardiography. At that time the original treatment plan of methylprednisolone (4 mg/day) was replaced by dexamethasone (4 mg/day) and weekly plasmapheresis: in the presence of an incomplete AV block, we used this therapy to avoid progression to a complete one. This regimen was maintained until delivery. The swelling and erythema of the patient's toes disappeared shortly after plasmapheresis was begun. Two weeks later, at week 24, fetal echocardiography showed episodes of sinus heart rate to 115–130 beats/minute along with a predominant second-degree AV block. Serial echocardiography, subsequently performed every 2 weeks, confirmed this finding. Cardiac anatomy and function in the fetus were normal, and no pericardial or pleural effusions or ascites were detected.
At week 39, a 2,440-gm girl was delivered by planned, uncomplicated cesarean section. Apgar scores were 9 after 1 minute and 10 after 5 minutes. The infant appeared healthy at birth, and no signs of cutaneous lesions were noted. Results of all standard blood tests, including IgG, IgA, and IgM levels, were normal. A 24-hour electrocardiogram (EKG) recording showed a predominant 2:1 Mobitz II second-degree AV block, with a mean heart rate of 75 beats/minute along with episodes of sinus heart rate >100 beats/minute. At 14 months of age the child developed an advanced second-degree block, which at 27 months of age progressed to a complete third-degree AV block with a mean heart rate of 72 beats/minute (ranging from 48 to 103 beats/minute), not requiring pacemaker implantation. The infant had anti-SSA/Ro antibodies (52 and 60 kd) at birth but became seronegative by 8 months of age. The levels of SSA/Ro antibodies in the mother were unaltered during pregnancy as measured by enzyme-linked immunosorbent assay (ELISA) in 6 consecutive serum samples.
Nine family members of the patient were available for study (Figure 1). The subjects, all white, were personally interviewed and clinically examined. Signs or symptoms of autoimmune diseases were sought, EKG was performed, and blood samples for HLA typing and autoantibody testing were obtained. In particular, serum IgG anti-SSA/Ro and anti-SSB/La antibodies were evaluated using an in-house counterimmunoelectrophoresis kit and by ELISA using plates coated with recombinant human 52- and 60-kd SSA/Ro and SSB/La proteins, respectively (Diamedix Delta Biologicals, Pomezia Croma, Italy).
Family pedigree and HLA haplotypes are outlined in Figure 1, in which the 2 anti-SSA/Ro–positive women (the index patient and her mother) and their daughters with CHB are indicated. Relevant findings were present in the index patient's mother. She was a 51-year-old woman with evident clinical and serologic features of a connective tissue disease, including fatigue, dry mouth, dry eyes, photosensitivity, diffuse arthralgias, oral ulcers, leukopenia with lymphocytopenia, and marked blanching of the fingers upon exposure to cold. Her antibody pattern was characterized by anti-SSA/Ro + antibodies (52 and 60 kd), antinucleolar antibodies at high titer as determined by immunofluorescence on HEp 2000 cells, and type 2 antimitochondrial antibodies as determined by immunoblotting. The first of her 2 pregnancies was normal and she gave birth to a healthy female infant (the index patient). Fetal bradycardia was detected during week 20 of her second pregnancy (May 1981). This went untreated, and at week 40 a bradycardic female infant with neonatal asphyxia was delivered. The first EKG performed on the infant, at 8 months of age, revealed a complete third-degree AV block. The worsening of the bradycardia made implantation of a pacemaker necessary when the subject reached the age of 17 years. She currently has incomplete Raynaud's phenomenon, characterized by blanching of fingers upon cold exposure. She is nonetheless negative for all examined antibodies including antinuclear, anti–extractable nuclear antigen, anti–native DNA, anticentromere, lupus anticoagulant, anticardiolipin, and antimitochondrial antibodies.
Several findings in this family study were intriguing. First, the finding of 2 women in the same family but from different generations (mother and daughter) who each had anti–52-kd SSA/Ro antibodies and delivered a female infant with CHB leads us to hypothesize that the risk for induction of anti-SSA/Ro–related CHB may be genetically transmitted.
Second, the HLA haplotype A*01;B*08;DRB1*03;DRB3*;DQB1*02 (haplotype c in Figure 1) was exhibited by all of the women in this family group and included several alleles previously reported to occur at high frequency in mothers of children with CHB (5, 6). Moreover some of these alleles (B*08, DRB1*03, and DQB1*02) are generally associated with anti–52-kd SSA/Ro antibodies (6), a reasonable marker for risk of CHB transmission (7). However, only the index patient and her mother had anti–52-kd SSA/Ro antibodies and delivered infants with CHB (Figure 1). It is interesting that the index patient's mother and 2 maternal aunts carried the same maternal and paternal haplotypes (a/c) with class II alleles in homozygous combination (Figure 1). But only the index patient's mother had anti–52-kd SSA/Ro antibodies and gave birth to a child with CHB, while her 2 sisters had no antibodies and delivered, respectively, 1 and 2 healthy infants.
It has been hypothesized that a rare allelic variant primarily associated with risk for CHB and/or with an enhanced potential for synthesizing pathogenic anti-Ro/La antibodies could occur in genetically predisposed women (5). This undefined variant was perhaps present in the mother of our index patient, who exhibited the haplotype that is reported to occur frequently in mothers of children with CHB. Furthermore, the presence of pathogenic anti–52-kd SSA/Ro antibodies also in 1 of her daughters would indicate that this variant may be genetically transmitted as a nondominant character. It is also interesting to observe that the index patient's daughter and sister, who were both affected with CHB, had the same maternal haplotype (c), while the paternal haplotypes (f and g) differed only with respect to 1 allele. The significance of this immunogenetic similarity is difficult to interpret, since some previous reports have described discordance for CHB in monozygotic twins (2, 3). No significant conclusions can be drawn from our data concerning a single family. Extending these studies on anti-SSA/Ro–related CHB to earlier generations of these families could verify our findings. However, because this antibody-transmitted disorder has only recently been defined (8, 9) and the methods used to study fetal heart rate abnormalities in the past were unreliable, this would be a difficult feat.