Clinical and Pathologic Implications of Extending the Spectrum of Maternal Autoantibodies Reactive with Ribonucleoproteins Associated with Cutaneous and Now Cardiac Neonatal Lupus from SSA/Ro and SSB/La to U1RNP

Abstract

While the relationship between maternal connective tissue diseases and neonatal rashes was described in the 1960s and congenital heart block in the 1970s, the “culprit” antibody reactivity to the SSA/Ro-SSB/La ribonucleoprotein complex was not identified until the 1980s. However, studies have shown that approximately 10-15% of cases of congenital heart block are not exposed to anti-SSA/Ro-SSB/La. Whether those cases represent a different disease entity or whether another antibody is associated has yet to be determined. Moreover, the cutaneous manifestations of neonatal lupus have also been identified in infants exposed only to anti-U1RNP antibodies. In this review, we describe what we believe to be the first case of congenital heart block exposed to maternal anti-U1RNP antibodies absent anti-SSA/Ro-SSB/La. The clinical and pathologic characteristics of this fetus are compared to those typically seen associated with SSA/Ro and SSB/La. Current guidelines for fetal surveillance are reviewed and the potential impact conferred by this case is evaluated.

1. Introduction

Neonatal lupus (NL) represents a pathologic readout of passively acquired autoimmunity that was initially described in the 1960s (cutaneous) and 1970s (congenital heart block, CHB) [1-4]. The association of neonatal lupus with anti-SSA/Ro was identified in the 1980s [3]. Identification of in utero heart block in the absence of structural abnormalities predicts the presence of maternal autoantibody responses against the specific ribonucleoproteins SSA/Ro and/or SSB/La in greater than 85% of cases [5]. However, that leaves approximately 15% of cases in which these maternal reactivities are not present, and whether those cases are “autoimmune” or represent a different clinical etiology remains unclear [5]. Prospective studies show the rate of developing CHB at approximately 2% in women with anti-SSA/Ro without a previously affected pregnancy, irrespective of maternal diagnosis [6-8]. The recurrence rate is increased by nearly tenfold in subsequent pregnancies [8, 9]. Interestingly, one study evaluated the recurrence rate in CHB absent exposure to maternal anti-SSA/Ro and/or SSB/La and found no recurrences in 36 subsequent pregnancies [10].

With regard to incidence two population-based studies were performed irrespective of the presence of maternal autoantibodies. A study from Finland places the number at 1:17,000 live births with the highest annual estimates at 1:6,500 [11]. A report from Stockholm County provided an incidence of anti-SSA/Ro autoantibody-related 2nd and 3rd degree block at 1:23,300 [12]. Although the signature cardiac disease typically manifests as CHB, which can be of varied degree but is most often complete [13], approximately 20% of affected offspring develop life-threatening cardiomyopathy [14, 15], which can also present without associated CHB [13]. The case fatality rate is approximately 18% and approaches 50% with the development of a dilated cardiomyopathy or endocardial fibroelastosis [13].

Cutaneous manifestations of neonatal lupus have been reported with anti-U1RNP [4], and there has been a previously reported case of transient 1st degree block in a mother with SLE and anti-U1RNP antibodies absent anti-SSA/Ro antibodies [16]. However, antibody reactivity solely to U1RNP has never been detected in association with advanced block in a structurally normal heart. Here we report the first case of a neonate with CHB exposed to maternal anti-U1RNP antibodies absent anti-SSA/Ro-SSB/La.

2.1. Case Report

A 40-year-old white female was diagnosed with mixed connective tissue disease (MCTD) at age fourteen when she presented with arthritis and Raynaud’s syndrome. She was found to have a positive antinuclear antibody, low positive anti-Smith antibodies, and high titer antibodies to U1RNP. In addition to MCTD, she subsequently experienced a pulmonary embolus while on estrogen-containing birth control pills but did not have antibodies to cardiolipin, Beta-2glycoprotein I, or the lupus anticoagulant. A genetic screen identified heterozygosity for one mutation, G20210A, in prothrombin, homozygosity for the MTHFR gene C677T mutation, and homozygosity for plasminogen activity inhibitor-1 4G/5G. She had one previous pregnancy in 2011, which resulted in the birth of a healthy girl at term by vaginal delivery. Prior to and during that pregnancy she consistently tested negative for anti-SSA/Ro and anti-SSB/La antibodies. She conceived again in 2015, and anti-SSA/Ro and anti-SSB/La antibodies remained undetectable. Medications continued during pregnancy were enoxaparin 40 mg sc daily and aspirin 81 mg daily. A routine obstetrical sonogram at 28 weeks revealed a fetal heart rate in the 130s. However, during the next scheduled visit at 30 2/7 weeks the fetal heart rate was noted to be in the 50s, and a sonogram confirmed bradycardia. A female child weighing three pounds and twelve ounces was delivered by emergency C-section with APGAR scores of 4 and 7 at 1 and 5 minutes, respectively. An electrocardiogram revealed complete (3rd degree) heart block. The child was initially placed on nasal CPAP and subsequently intubated, given isoproterenol, and transferred to a local level IV neonatal intensive care unit. Rapid deterioration ensued with respiratory distress syndrome and signs of pulmonary hypertension. She was given poractant alfa and started on an oscillator with inhaled nitric oxide (iNO). Epinephrine was started in lieu of isoproterenol but discontinued due to multiple premature ventricular contractions/trigeminy, and dopamine was started. Broad-spectrum antibiotics (ampicillin and ceftriaxone) were initiated in addition to hydrocortisone for hemodynamic support. An echocardiogram on the first day of life showed a moderately dilated left ventricle with moderately decreased biventricular systolic dysfunction. There was no pericardial effusion. There was a small patent ductus arteriosus with a trivial left-to-right shunt. Repeat echocardiogram on day 4 showed improved left ventricular systolic function and right-sided function was within the lower limits of normal. An epicardial pacemaker was placed on day 5 and set to VV1 90, and patent ductus arterial ligation performed on day 12. The hospital course was complicated by concern for necrotizing enterocolitis, which was managed conservatively; a non-occlusive aortic thrombus that extended inferiorly from the renal vessels presumed secondary to an umbilical venous line; a transverse sinus thrombosis, which raised the possibility of a hypercoagulable state; a right grade 1 and left grade 2 interventricular hemorrhage; cholestasis; and renal failure with anasarca requiring peritoneal dialysis. Care was withdrawn at day 28 for hypotension despite maximum vasopressor support and broad-spectrum antibiotics. An autopsy was performed.

2.2. Autopsy

A limited postmortem examination was performed. The autopsy revealed severe anasarca, pleural effusions, and pulmonary findings consistent with immaturity. The heart was structurally normal but enlarged (22.7 gm, normal range 12.4 ± 2.8 gm). Initial dissection of the heart was done along the lines of blood flow, and additional dissection was performed to obtain the AV node and adjacent conducting system for histologic analysis. Histologically, there was loss and calcification of the myocytes in the area of the bundle of His and extending into the proximal Purkinje fibers (Fig 1A,B). The AV node was not involved in the destructive process (Fig 1C). Areas of early myocyte calcification were seen in the septal myocardium likely related to a secondary process during the time of cardiac support (Fig 1D).

Figure 1.

A – Low-power view of the bundle of His (circled) demonstrating myocyte loss, scarring and calcification confined to the conducting system (100x original magnification, hematoxylin & eosin). B – Higher-power view of the conducting system showing calcified remnants of myocytes (arrows) and dense collagen (200x, H&E) C – High-power view of the AV nodal tissue showing no evidence of pathology in this location (200x H&E). D – Focal areas of myocyte calcification of the non-conducting septal myocardium were appreciated (200x, H&E).

Maternal blood was sent to the Research Registry for Neonatal Lupus at New York University School of Medicine for evaluation of antibodies specific for 60 kD SSA/Ro (native protein), 52 kD SSA/Ro (recombinant protein), and 48 kD SSB/La (recombinant protein) as previously described [17] and were found to be negative, Table 1. Anti-U1A of the U1RNP complex was evaluated using an enzyme-linked immunosorbent assay and confirmed to be positive, Table 1.

Table 1:

Reactivity of maternal sera against extractable nuclear antigens

Antibody Reactivity	Negative Control (EU) *	Positive Control (EU)	Patient (EU)
RNP UI-A	0	1018	128
NRo60	0 (< 70 cutoff)	1056	0
HisRo52	19 (< 70 cutoff)	1092	0
HisLa48	2 (< 70 cutoff)	1025	2

EU = Elisa Units

^*The positive cut offs represent 3 standard deviations above the mean of 35 healthy controls in the case of Ro60, Ro52, and La48. 10 healthy controls were tested for RNP and all gave an EU of 0. All samples are diluted at 1:1000.

3. Mixed Connective Tissue Disease and Pregnancy

The diagnosis of Mixed Connective Tissue Disease requires the presence of antibodies reactive with U1RNP [18]. Data regarding pregnancy outcomes in MCTD are limited. In 2005, a review of the literature revealed no cases of CHB [19], although it is acknowledged that intrauterine fetal demise could have been due to cardiac injury. However, in an older case series limited to patients with high titer anti-U1RNP antibodies irrespective of maternal disease status, there were no cases of CHB or any significant increase in fetal loss [20].

One case of advanced CHB in a fetus of a mother with MCTD has been reported, but this antedated the association with anti-SSA/Ro, which had not been tested [21]. It is worth noting that in several case series of CHB in which a small proportion of the mothers were described as anti-SSA/Ro-SSB/La negative, anti-U1RNP was not tested [2, 5].

4. Current guidelines for fetal surveillance

Current recommendations for antenatal surveillance of women with anti-SSA/Ro antibodies, regardless of health status, are based on the literature and experience of the treating physicians and are generally presented as guidelines rather than absolute requirements. Accordingly, all mothers with systemic lupus erythematosus (SLE), Sjogren’s syndrome, and rheumatoid arthritis should be tested for anti-SSA/Ro-SSB/La antibodies in the 1st trimester and, if negative despite having an autoimmune disease, be advised that fetal echocardiograms are unnecessary. In contrast, for patients who do have these antibodies despite having had a child with any manifestation of neonatal lupus, or for those clinically asymptomatic women with anti-SSA/Ro-SSB/La antibodies (most often identified only because of CHB in a previous pregnancy), surveillance with weekly fetal echocardiograms between 16 and 26 weeks of gestation (the vulnerable period for detection of CHB) is recommended. Thus, it is not surprising that the case presented herein had no previous echocardiograms. This could also explain the detection of complete block after 26 weeks since a prolonged PR interval, subtle evidence of myocardial dysfunction, or even 2nd degree block in the absence of bradycardia would have been missed by auscultation or a standard handheld continuous Doppler ultrasound device.

5. Pathophysiology and histologic findings of anti-SSA/Ro-SSB/La associated congenital heart block and comparison to anti-U1RNP case

With regard to the pathogenicity of anti-SSA/Ro and anti-SSB/La in the development of CHB, a major challenge has been explaining the accessibility of an intracellularly sequestered antigen such as SSA/Ro and SSB/La to extracellular maternal autoantibodies in the fetal circulation. One plausible explanation is that these antigens translocate to the cell surface during apoptosis, a physiologic process that takes place during fetal development. Consistent with this hypothesis, indeed both SSA/Ro and SSB/La have been identified at the cell surface of apoptotic fetal cardiocytes [22]. Binding of these cells results in opsonization, subsequent engulfment by macrophages, and secretion of proinflammatory and profibrotic cytokines, which may promote scarring by activation of resident fetal cardiac fibroblasts [23]. The ssRNA associated with SSA/Ro and SSB/La can activate macrophage Toll-like receptors (TLR). While the ssRNA associated with U1RNP is also capable of ligating endosomal TLRs [24], earlier publications have shown that antibodies to U1RNP do not bind apoptotic cardiocytes [22]. While this is perplexing, the case reported herein does raise the consideration that under some circumstances surface translocation of other ribonucleoproteins and subsequent reactivity with circulating autoantibodies could be pathogenic to the fetal conduction system.

The absence of fetal surveillance notwithstanding, it is also possible that the pathogenesis of anti-U1RNP-mediated injury in this case is not identical to that occurring with anti-SSA/Ro antibodies, accounting for a later vulnerable period. Of relevance, the histopathology of this anti-U1RNP-exposed fetus with CHB did not reveal the characteristic fibrosis of the AV node. However this histologic picture is not unique, as a similar description has been reported for two post-natal deaths associated with complete block and maternal anti-SSA/Ro-SSB/La antibodies [25]. One revealed fibrosis of the bundle of His and the other foci of microscopic calcification in the atrial septum in proximity to but not involving the AV node [25]. In addition, there was no evidence of a mononuclear cell infiltrate (particularly macrophages or giant cells) that has also been consistently described in anti-SSA/Ro associated CHB. This again may suggest a distinct pathologic entity or simply reflect death that occurred postpartum, a month after the initial detection of block.

6. Implications of case

The implications of this case report pertain to women who are of childbearing age and have anti-U1RNP antibodies and not anti-SSA/Ro-SSB/La. This at-risk population would include asymptomatic women with anti-U1RNP antibodies in addition to those with connective tissue diseases in which anti-U1RNP antibodies are present such as MCTD, SLE, inflammatory myositis and scleroderma. Population data reveal anti- U1RNP antibodies in 0.20% of women [26]. Women with MCTD by definition have anti-U1RNP antibodies [18]. To address the frequency of anti-U1RNP antibodies in isolation, particularly in SLE, we obtained data from the Manhattan Lupus Surveillance Program, a population-based surveillance program initiated to determine the incidence and prevalence of SLE. Anti-U1RNP antibodies were present in isolation in 21% of cases of SLE, a higher prevalence than anti-SSA/Ro alone, Table 2. Data on the presence of anti-U1RNP in inflammatory myositis supports 5.5% [27] and in scleroderma approximately 8% [28] but whether these are isolated reactivities or observed in combination with anti-SSA/Ro and/or anti-SSB/La antibodies was not reported. In aggregate, these data imply that many more women may be at risk of having a child with CHB than previously appreciated. Consequently, fetal surveillance with weekly echocardiograms may eventually be recommended for many more women, creating a substantial economic burden to health care. For asymptomatic mothers in whom CHB is identified in an otherwise structurally normal heart, anti-U1RNP antibodies should be sought if antibodies to the SSA/Ro-SSB/La complex are negative. Prospective studies analogous to those done on anti-SSA/Ro antibody positive pregnancies [6-8] are needed to provide the incidence of anti-U1RNP-associated CHB. That this newly identified antigenic target is also a ribonucleoprotein is likely more than coincidental and further reinforces that antibody insult involves drivers of the fetal innate immune system.

Table 2:

Frequency of anti-SSA/Ro, anti-SSB/La and anti-U1RNP positivity in 745 American College of Rheumatology classification criteria-confirmed cases of systemic lupus erythematosus in which all three antibodies were tested

Anti-SSA/Ro	Anti-SSB/La	Anti-U1 RNP	Percent
+	+	+	8.1
+	+	−	9.7
+	−	+	9.8
+	−	−	13.0
−	+	+	0.7
−	+	−	1.5
−	−	+	20.5
−	−	−	36.8

Highlights.

• Identification of in utero heart block in the absence of structural abnormalities predicts the presence of maternal autoantibody responses against the specific ribonucleoproteins SSA/Ro and/or SSB/La in greater than 85% of cases.

• Cutaneous manifestations of neonatal lupus have been reported with anti-U1RNP.

• This review contains the first case report of congenital heart block associated with anti-U1RNP antibodies absent the traditionally associated neonatal lupus antibodies to anti-SSA/Ro and/or anti-SSB/La.

• Clinicians should consider testing for maternal anti-U1RNP in congenital heart block cases in which anti-SSA/Ro and/or anti-SSB/La are negative and there is no structural etiology.