Maternal Anti-Ro and Anti-La Antibody–Associated Endocardial Fibroelastosis
Background— Maternal anti-Ro and anti-La antibodies are associated with congenital heart block (CHB). Although endocardial fibroelastosis (EFE) has been described in isolated cases of autoantibody-mediated CHB, the natural history and pathogenesis of this disease are poorly understood.
Methods and Results— We retrospectively reviewed the clinical history, echocardiography, and pathology of fetuses and children with EFE associated with CHB born to mothers positive for anti-Ro or anti-La antibodies at 5 centers. Thirteen patients were identified, 6 with a prenatal and 7 with a postnatal diagnosis. Six mothers were positive for anti-Ro and anti-La antibodies, and 7 were positive for anti-Ro antibodies only. Only 1 mother had autoimmune disease. Severe ventricular dysfunction was seen in all fetal and postnatal cases. Four fetal and 3 postnatal cases had EFE at initial presentation. However, 2 fetal and 4 postnatal cases developed EFE 6 to 12 weeks and 7 months to 5 years from CHB diagnosis, respectively, even despite ventricular pacing in 6 postnatal cases. Eleven (85%) either died (n=9) or underwent cardiac transplantation (n=2) secondary to the EFE. Pathologic assessment of the explanted heart, available in 10 cases, revealed moderate to severe EFE in 7 and mild EFE in 3 cases, predominantly involving the left ventricle. Immunohistochemistry in 4 cases (including 3 fetuses) demonstrated deposition of IgG in 4 and IgM in 3 and T-cell infiltrates in 3 cases, suggesting an immune response by the affected fetus or child.
Conclusions— EFE occurs in the presence of autoantibody-mediated CHB despite adequate ventricular pacing. Autoantibody-associated EFE has a very high mortality rate, whether developing in fetal or postnatal life.
Received August 23, 2001; revision received December 6, 2001; accepted December 19, 2001.
Endocardial fibroelastosis (EFE), whether congenital or acquired, is a rare and poorly understood disease of the endomyocardium, with most patients progressing to end-stage heart failure and death.1–3⇓⇓ The pathologic features of EFE, considered the gold standard for diagnosis, consist of collagen and elastin deposition, ventricular hypertrophy, and diffuse endocardial thickening.4,5⇓ The etiology of EFE remains unclear but may include viral infections,6,7⇓ primary carnitine deficiency,8 and metabolic disorders9 and can be hereditary.4,10⇓ It has been suggested that EFE may be secondary to an autoimmune process, which may be initiated antenatally in some cases. The inflammatory response within the myocardium may lead to smooth muscle hypertrophy and collagen and elastin fiber deposition. The net effect on the myocardium is left ventricular dilatation and hypertrophy, reduced ventricular compliance, and progressive endocardial fibrosis.11
The association between maternal anti-Ro and anti-La antibodies and congenital heart block (CHB) has been well documented.12 Small case series have described infants diagnosed with both CHB and EFE, but a causative link between these 2 diseases has not yet been identified.13 It has been suggested that EFE in patients with CHB is a secondary phenomenon attributable to long-standing heart block, bradycardia, and congestive heart failure. However, not all patients with autoantibody-mediated CHB develop EFE, suggesting that EFE may occur by a different mechanism.
In the present study, we identified a cohort of fetuses and infants with the diagnosis of EFE and CHB born to mothers positive for anti-Ro or anti-La antibodies who were managed at several institutions. The primary objective of the study was to describe the echocardiographic and pathological findings and clinical outcomes of these patients. The secondary objective of the study was to document the association between maternal anti-Ro and anti-La antibodies and EFE.
This was a multicenter, historical cohort study that included patients from the Hospital for Sick Children, Toronto, Canada, the University of California San Francisco, San Francisco, Calif, the Izaak Walton Killam-Grace Health Center, Halifax, Canada, the Children’s Hospital of Pittsburgh, Pa, and British Columbia’s Children’s Hospital, Vancouver, Canada. Patients with CHB were included in the study if they had a fetal or postnatal diagnosis of EFE confirmed at necropsy or by echocardiography and were born to mothers with anti-Ro or anti-La antibodies. Patients were excluded from the study if there was evidence of structural heart disease known to be associated with secondary EFE.
Demographic patient data collected included date of birth, sex, date diagnosed with EFE, date diagnosed with CHB, date of pacemaker insertion and type of pacemaker, ventricular heart rate at time of diagnosis of CHB and at time of death, follow-up, or cardiac transplantation, date of death, and cause of death (Table 1). Maternal data collected included presence of anti-Ro or anti-La antibodies, age at pregnancy, underlying diagnosis of autoimmune diseases, such as systemic lupus erythematosus or Sjogren’s syndrome, parity, history of miscarriages and other children with EFE or CHB, and the use of oral or intravenous corticosteroids during the pregnancy for maternal or fetal reasons.
Two investigators (L.E.N. and L.K.H), blinded to the pathology results, reviewed all available fetal and postnatal echocardiograms at time of diagnosis of EFE and at time of death or cardiac transplantation. Echocardiographic parameters examined included left ventricular ejection fraction by either M-mode or Simpson’s rule,14 left ventricular shortening fraction, and severity of mitral regurgitation graded as absent, mild, moderate, or severe. Measurements of the left ventricular end-diastolic dimension (LVED), right ventricular end-diastolic dimension (RVED), interventricular septal dimension, and left ventricular posterior wall dimension were recorded from M-mode tracings or 2-dimensional images in centimeters. The fetal and postnatal echocardiogram measurements of M-mode cardiac dimensions were compared with standard normal values.15,16⇓ The ratio of the left to right ventricular end-diastolic dimension was calculated. The presence or absence of fibrosis of the endomyocardium was identified as areas of echogenicity on the endocardial surface of the right ventricle, left ventricle, left atrium, mitral valve leaflets, and tension apparatus and was graded as mild, moderate, or severe depending on the thickness and the extent of myocardial involvement.
The fetal and postnatal pathology reports and slides from autopsy specimens, explanted hearts after cardiac transplantation, or endomyocardial biopsies were reviewed by 1 of 2 pathologists (G.P.T. or J.B.M.M.), who were blinded to clinical data for shape of heart (defined as globular or nonglobular), severity of endocardial fibrosis (graded as mild, moderate, or severe), extent of EFE (defined as diffuse or focal), thickness of EFE (defined as thin or thick), and depth of EFE in millimeters. The presence or absence of EFE in the right atrium, left atrium, right ventricle, left ventricle, mitral valve, and papillary muscles was noted. Where possible, the atrioventricular (AV) node was examined for abnormalities.
Available 5-μm formalin-fixed, paraffin-embedded tissue sections were mounted on positively charged microscopic slides. Tissue sections were then baked overnight at 60°C, dewaxed in xylene, and hydrated to distilled water through decreasing concentrations of alcohol. All immunohistochemical procedures were performed on the Ventana Gen II auto-immuno in situ stainer (Ventana Medical Systems), a closed system using the avidin biotin complex detection system, using the DAB (3-3′-Diaminobenzidine) Ventana Detection System (No. 250-001) as the chromogen substrate. All tissue sections were treated for endogenous peroxidase and biotin. The counterstain of preference was hematoxylin for nuclear detail. Sections were stained using alkaline phosphatase–conjugated antibodies directed against human IgG, human IgM, human IgA, CD20, and CD43. For the IgG, IgM, and IgA staining, rabbit polyclonal anti-human–specific antibody was used (DAKO, Carpinteria, Calif). Monoclonal mouse anti-human antibody was used for CD20 and CD43 (DAKO) using mouse IgG1 κ isotypes. Sections were stained with a 3-step automated detection system as outlined by the manufacturer. For IgG, IgM, and IgA staining, the primary antibody was rabbit polyclonal anti-human class–specific antibody, whereas for CD20 and CD43 staining, the primary antibody was a mouse monoclonal anti-human IgG1 κ isotype. The secondary antibody was a biotinylated goat anti-mouse IgG and anti-rabbit IgG, as required. The third step was peroxidase-labeled streptavidin, DAB chromogen, substrate, and copper enhancer. The omission of the primary antibody was used as a control as indicated by the manufacturer, and, in addition, for the murine monoclonal antibodies, an isotype-specific control was used (either anti-CD8 or anti-CD69). In addition, the hearts of two 20-week fetuses (elective terminations for social indications), one 28-week fetus (demise after placental abruption), one 1-month-old infant (sudden infant death syndrome), and one 6-month-old infant (death related to acute asphyxia) were used as control cases. All 5 cases were stained for IgG, IgA, IgM, CD20, and CD43 using the identical staining techniques described above (see Table 2).
All slides were reviewed by E.D.S and L.E.N. (blinded to clinical material) and compared with positive and negative controls for both the fetal and postnatal specimens. Each marker was recorded as either positive or negative. If positive, the strength of the reaction was graded as +, ++, or +++, depending on the extent of infiltration (diffuse or focal) and location (global, central, or peripheral).
A total of 13 patients (3 male and 10 female) of mothers positive for anti-Ro or anti-La antibodies were diagnosed with EFE. Their demographic data are represented in Table 1, with outcomes described in Figure 1. All patients had structurally normal hearts, except for 1 postnatal patient who had a small muscular ventricular septal defect. The mean maternal age of the entire cohort at the time of the pregnancy was 28.4±4.5 years. Seven mothers were primiparous, and only 1 had a history of a previous miscarriage. There was no history consistent with maternal infection and no family history of congenital heart disease or cardiomyopathy. A positive family history for CHB in a sibling was present in 2 (15%) cases.
Six of the 13 cases were diagnosed prenatally. Of those, 4 (67%) were diagnosed concurrently with EFE and CHB at a median gestational age of 24 weeks (range, 18 to 30 weeks). The other 2 developed EFE 6 and 12 weeks after the diagnosis of CHB. Four cases presented with hydrops fetalis. Four of the mothers received dexamethasone orally (4 mg per day) after the diagnosis of EFE (including 1 fetus without hydrops) in an attempt to reduce myocardial inflammation, improve ventricular function, and reduce the degree of hydrops (in 3 cases). In 1 of the 4 cases receiving dexamethasone therapy, epicardial ventricular pacing with pacemaker insertion in utero was also attempted in an effort to salvage the grossly hydropic fetus. Fetal demise occurred several hours after the procedure, despite successful ventricular pacing. All 5 fetal cases with hydrops fetalis (83%) died in utero.
Seven of the 13 cases were diagnosed with CHB and EFE postnatally. Six of the seven were paced at a median age of 4.7 months (range, 2 days to 6 years). Of those, 3 had EFE at the time of pacemaker insertion, whereas 3 developed EFE 6 months to 4 years after the pacemaker insertion. None had a diagnosis of EFE before CHB diagnosis.
Nine of the 13 (69%) patients died as a result of the EFE and severe ventricular dysfunction. The cause of death in all patients was congestive heart failure or sudden cardiac arrest. Two other patients underwent cardiac transplantation, for a total patient or organ death of 85% (Figure 1). One patient died 15 months after cardiac transplantation as a consequence of posttransplant lymphoproliferative disease. The median time from diagnosis of EFE to either death or cardiac transplantation was 7.5 days (0 to 18 days) in the fetal group and 11.9 months (9 days to 2.1 years) in the postnatal group (Figures 2 and 3⇓).
Of the postnatal diagnosis group, only 1 patient is presently alive without cardiac transplantation. This patient (now 3 years old) has recently been removed from the cardiac transplantation list and is being medically managed for congestive heart failure. The only fetus to survive was diagnosed with CHB and EFE at 24 weeks of gestation. She was semi-electively delivered by caesarian section for fetal distress at 35 weeks and required an epicardial ventricular pacemaker at 1 day of age. She was treated with intravenous immunoglobulin and pulse corticosteroids in the first few days of life. She is presently 9 months old and no longer requires antifailure medications. Her last echocardiogram demonstrated normal cardiac function, a mildly enlarged left ventricle, and echogenicity of the tricuspid valve, mitral valve, and interventricular septum consistent with mild EFE.
A total of 4 of 6 fetal cases had echocardiograms available for review, with 3 cases having both an initial echocardiogram at the time of diagnosis and an echocardiogram at follow-up. In all cases, there was severe right ventricular dysfunction. In 3 of 4 (75%) cases, there was both right and left ventricular dysfunction. Two patients had mild mitral regurgitation. The right and left ventricular end-diastolic and left ventricular posterior wall measurements were all within normal limits. The EFE was graded as moderate to severe in 3 of 4 (75%) cases. The only case of mild EFE in the fetal group was, in fact, the only fetal survivor. The EFE predominately involved the RV in 3 of 4 (75%) cases and was diffuse in all cases. The mean LVED/RVED ratio was 1.0±0.117 (Figure 2).
Of the postnatally diagnosed cases, a total of 6 of 7 cases had echocardiograms available for review, with 5 having both a baseline and follow-up echocardiogram. The left ventricular function was reduced in all, with a median ejection fraction of 27% (range, 6% to 55%). There was moderate mitral regurgitation in 4 cases. The right ventricular end-diastolic dimension was normal in 5 cases and increased in 1 case. The left ventricular end-diastolic and interventricular septal dimension were increased in 3 cases, whereas the left ventricular posterior wall dimension was increased in only 1 case. The EFE was graded as mild to moderate in 5 (83%) cases. There was no identifiable EFE by echocardiography in 1 case, although this was found on pathologic examination. The EFE was predominately in the LV in 5 of 6 (83%) cases, and tended to be most prominent along the LV free wall and interventricular septal areas. The mean LVED/RVED ratio was 1.9±0.6.17
Pathology was available in 12 patients (5 fetal and 7 postnatal), which included 1 patient with only a collection of histology slides (fetal group) and 1 patient with tissue from a myocardial biopsy (postnatal group). The hearts were classified as globular in only 1 fetal case and in all 6 postnatally diagnosed cases in which there was an explanted heart available for review. The AV node tissue, available in 5 cases (3 fetal and 2 postnatal), was consistently calcified and fibrotic.
Grading the severity of EFE was possible in 5 fetal and 5 postnatal cases. One case was excluded because only myocardial biopsy slides were available. These slides showed multifocal myocardial hypertrophy and fibrosis without evidence of inflammation. The second case was excluded because the heart specimen was not available. A photograph of the LV demonstrating diffuse EFE was provided for this case. The EFE was graded mild in 2 of 5 (40%) fetal cases and in 1 of 5 (20%) postnatal cases and moderate in 3 of 5 (60%) fetal cases and in 2 of 5 (40%) postnatal cases. It was graded as severe in none of the fetal cases and in 2 of 5 (40%) postnatal cases. The EFE was diffuse in all cases and was predominately seen in the left ventricle in 3 of 4 (80%) fetal and 4 of 5 (80%) postnatal cases. It was present along the endocardial surface of the left ventricle in all cases and to some extent in both right and left ventricles in 7 of 9 (77%) cases and on the papillary muscles and tension apparatus of the mitral valve in 6 of 7 (86%) cases. It was least often present in the right atrium (1 of 8, 12%) and left atrium (2 of 8, 25%).
Immunohistochemical analysis of the specimens was possible in only 4 cases, for a total of 7 sections from the AV node, right ventricle, and left ventricle (Table 2). All 4 cases were strongly positive for IgG, which was seen as a dense, diffuse infiltrate across myocardial and AV nodal tissue. Myocardial tissue stained positive as well for IgM and T cells (CD43-positive cells) in most cases (3 of 4), whereas B cells (CD20-positive cells) were rarely detected (1 of 4). Immunohistochemical staining of the control hearts showed that all 5 fetal hearts had deposition of IgG diffusely throughout the myocardium (Figure 3). The intensity of the IgG staining was greatest in the 20-week fetal hearts, but it was consistently less than the staining of the EFE cases. There was no IgG deposition on the hearts from the 1-month-old and 6-month-old infants. We did not detect IgM or IgA deposition in any of the control hearts nor evidence of CD20-positive cells.
In this study we have described a cohort of 13 fetuses and infants in whom EFE was diagnosed in association with maternal anti-Ro antibody–associated and anti-La antibody–associated CHB but in the absence of maternal autoimmune disease in all but 1 case. We found that EFE rapidly progressed to cardiac failure and death and was associated with a very high mortality or organ death rate in both the fetus and the infant, although the disease was more fulminant in the fetal group. Hydrops fetalis was present in most of the fetal cases and resulted in the intrauterine demise of almost all cases. In the postnatal group, the EFE developed despite appropriate ventricular pacing in half of the cases. Immunohistochemical staining of myocardial specimens demonstrated significant, diffuse antibody deposition on the endomyocardium in the areas confirmed to have EFE.
Although there is a paucity of data regarding the etiological association between anti-Ro or anti-La antibodies and EFE, isolated case reports have described the coexistence of EFE, CHB, and maternal autoantibodies.18,19⇓ The presence of EFE was assumed to be attributable to prolonged bradycardia and cardiac failure. However, the results from our study suggest an alternative hypothesis. EFE presented as congestive heart failure after a period of clinical stability despite adequate ventricular pacing in half of the postnatal cases, and EFE was diagnosed in most of fetal cases concurrently with a diagnosis of CHB. These observations suggest that CHB may not be the cause of EFE but rather that congestive heart failure developed as a consequence of the endomyocardial damage, separate from the CHB. Our clinical and histological findings support the concept that maternal autoantibody-associated CHB and maternal autoantibody-associated EFE are two different but associated manifestations of neonatal lupus erythematosus.20 Furthermore, we have recently encountered 3 cases of EFE cardiomyopathy in the absence of CHB among offspring of mothers with autoantibodies (Nield et al, unpublished data, 2001). These cases provide an even stronger argument for our hypothesis that EFE with myocardial dysfunction is attributable to diffuse cardiac damage induced by maternal autoantibodies. CHB is likely the result of more localized damage to the AV nodal tissue.
As is true for antenatally diagnosed structural heart disease, the fetus presenting with EFE represents a more severe spectrum of the disease. There was more involvement of both ventricles among the prenatally diagnosed cases and more myocardial disarray observed histologically. Patients who develop EFE postnatally may have a more latent form of the disease that may require a second hit, such as exposure to viral antigens or other autoantibodies. For instance, cytomegalovirus infection has recently been shown to enhance surface expression of the Ro antigen in a keratocyte cell culture model.21 Similarly, as neither CHB nor EFE occur in all fetuses and infants born to autoantibody-positive mothers, perhaps genetic predisposition, such as HLA haplotype or presence of specific antigen receptors with or without a viral trigger are also necessary to create the optimal setup for an inflammatory reaction within the myocardium.
Immunohistochemical analysis of the myocardium in these patients supports the hypothesis that EFE is secondary to the transplacental passage and direct deposition of maternal IgG on the developing endomyocardium.22 However, the 20-week hearts also had evidence of less significant IgG deposition, the importance of which is uncertain. Staining for IgG is nonspecific, and, to date, there is no technique available to stain specifically for anti-Ro and anti-La antibodies. There may be other maternal antibodies that deposit on the fetal myocardium but do not induce disease. In addition, the anti-Ro and anti-La antibody status of the control mothers is unknown.
Although IgG deposition was observed in the control cases, the finding of IgM deposition and multiple foci of T cells exclusively in hearts with EFE suggests that there is a fetal component to the immunologically mediated injury that has not previously been shown. It has been well demonstrated that the fetal immune system is capable of responding to injury by production of IgM and with a T-cell response.23,24⇓ The response of the fetal immune system to the foreign maternal IgG present on the fetal heart may play a critical role in the development of the myocardial damage. Although it is possible that the T cells may be of maternal origin, it is very unlikely, because maternal T cells are very infrequently found in the fetus and can only be detected using the polymerase chain reaction (which can detect cells at a frequency of <10−6)25,26⇓ Therefore, we suggest that the T cells were very likely of fetal origin.
Establishing the connection between maternal anti-Ro and anti-La antibodies in patients with EFE and CHB has important clinical implications. In a recent review of 102 cases of isolated CHB diagnosed at our institution over 3 decades, we have demonstrated that EFE is a significant risk factor for death in this population. Although EFE was identified in only 5% of cases with CHB, EFE cardiomyopathy accounted for one third of the overall deaths and 83% of the deaths among cases of autoantibody-mediated CHB (Jaeggi E et al, in press). The fact that patients in the latter series with CHB who are appropriately paced can develop EFE emphasizes that these patients should be followed closely by a pediatric cardiologist with serial echocardiograms throughout childhood. Such an approach would ideally identify this subgroup of patients earlier, allowing them to receive more focused medical treatment or be listed in a more timely fashion for cardiac transplantation. It is of note, however, that in our experience, echocardiography tended to overestimate the severity of EFE among fetal cases and underestimate the EFE in postnatal cases. The presence of any ventricular dysfunction may be a more useful sign of myocardial damage, although myocardial damage is likely present before there is echocardiographic evidence of dysfunction.
Given that circulating maternal autoantibodies may cause EFE, intervention with immunosuppressive agents such as corticosteroids, intravenous immune globulin, and plasmapheresis may be indicated. Anti-inflammatory agents may serve to reduce both the level of circulating autoantibodies and, if fluorinated corticosteroids are used, directly decrease the inflammation and damage to the developing endomyocardium. Previously published reports have suggested that such therapy has improved the outcome in some fetuses who present with CHB and hydrops.27,28⇓ In the postnatal population, anti-inflammatory agents have long been used in an attempt to improve the outcome of patients with acute myocarditis with some success.29,30⇓ We suggest that in cases of EFE with circulating autoantibodies, a trial of immunosuppressive therapy is warranted given the high mortality. Larger prospective studies are required to address the true efficacy of such agents in the management of autoantibody-induced EFE.
In conclusion, maternal autoantibody-induced EFE can occur in the presence of CHB despite adequate ventricular pacing and is associated with high rate of mortality or organ death among affected fetuses and infants. A fetal or infant immune response to the deposition of maternal autoantibodies on the myocardium likely contributes to the development of the disease. Additional studies are necessary to determine the true incidence of EFE of fetuses and infants born to anti-Ro antibody–positive or anti-La antibody–positive mothers and to establish optimal therapeutic options in these patients.
Financial support for this project was provided in part by the March of Dimes Research Grant Foundation (grant No. 6-FY00-252). We thank Michael Ho, who performed the immunostaining analysis of the specimens for his contribution to this manuscript.
- ↵Kelly JAD. Congenital endocardial fibroelastosis II: a clinical and pathologic investigation of those cases without associated cardiac malformations, including a report of two familial instances. Pediatrics. 1956; 18: 539–555.
- ↵Moss AJ, Adams FH. Heart Disease in Infants, Children, and Adolescents. Baltimore, Md: Williams & Wilkins; 1995: 1355–1358.
- ↵Sellers FJ, Keith JD, Manning JA. The diagnosis of primary endocardial fibroelastosis. Circulation. 1964; 29: 49–59.
- ↵Westwood M, Harris R, Burn JL, et al. Heredity in primary endocardial fibroelastosis. Br Heart J. 1975; 37: 1077–1084.
- ↵Snider AR, Serwer GA, Ritter SB. Echocardiography in Pediatric Heart Disease. St Louis, Mo: Mosby Year Book; 1997: 143–149.
- ↵Allan LD, Joseph MC, Boyd EGCA, et al. M-mode echocardiography in the developing human fetus. Br Heart J. 1982; 47: 573–583.
- ↵Sahn DJ, Lange LW, Allen HD, et al. Quantitative real-time cross sectional echocardiography in the developing human, fetus and newborn. Circulation. 1980; 62: 588–597.
- ↵Hull D, Binns BAO, Joyce D. Congenital heart block and widespread fibrosis due to maternal lupus erythematosus. Arch Dis Child. 1966; 41: 688–690.
- ↵McCue C, Mantakas ME, Tingelstad JB, et al. Congenital heart block in newborns of mothers with connective tissue disease. Circulation. 1977; 56: 82–90.
- ↵Silverman ED, Laxer RM. Neonatal lupus erythematosus. Pediatr Rheum. 1997; 23: 599–618.
- ↵Zhu J. Cytomegalovirus infection induces expression of 60KD/Ro antigen in human keratinocytes. Lupus. 1995; 4: 396–406.
- ↵Lo YMD, Lo ESF, Watson N, et al. Two-way cell traffic between mother and fetus: biologic and clinical implications. Blood. 1996; 88: 4390–4395.