Article Text
PDF
Abstract
Objectives Autoimmune congenital heart block (CHB) is associated with placental transcytosis of maternal autoantibodies directed against Ro/SS-A and La/SS-B. However, only about 2% of children born to mothers with the respective antibodies are affected, indicating that further risk factors exist, which are not yet fully understood. In this study, we investigated whether a maternal type I interferon (IFN) signature represents a risk factor for the development of CHB.
Methods Blood samples, clinical data and serological parameters from 9 women with CHB pregnancies, 14 pregnant women with antibodies against Ro/SS-A but without a CHB complication and another 30 healthy pregnant women as controls were studied. SIGLEC1 expression was measured by flow cytometry and was correlated to plasma IFN-α levels measured by ELISA, and IFN-γ-induced protein 10 (IP-10) levels measured by Bio-Plex technique.
Results Mothers of affected children had a significantly higher expression of SIGLEC1 (p=0.0034) and IFN-α (p=0.014), but not of IP-10 (p=0.14, all MWU) compared to mothers of unaffected children. SIGLEC1 and IFN-α expression were reduced by hydroxychloroquine and oral glucocorticoids.
Conclusions High expression of SIGLEC1 in pregnant women with autoantibodies against Ro/SS-A indicates an enhanced risk for CHB development, and these women may benefit especially from IFN-α directed therapy, for example with hydroxychloroquine.
- congenital heart block
- neonatal lupus syndrome
- SIGLEC1
- sialoadhesin
- interferon-α
Statistics from Altmetric.com
Introduction
Autoimmune congenital heart block (CHB) is a severe heart disease belonging to the spectrum of neonatal lupus syndrome that occurs in approximately 1 in every 20 000–30 000 pregnancies and is associated with a high fetal morbidity and mortality. Placental transcytosis of maternal IgG antibodies directed against the ribonucleic complexes Ro (SS-A) and La (SS-B) is a key factor in the pathogenesis, resulting in inflammation-mediated conduction abnormalities of the fetal heart. The most common manifestation of CHB is fibrotic remodelling of the atrioventricular (AV) node with subsequent AV block, but extranodal abnormalities such as endocardial fibroelastosis, cardiomyopathy or septal defects related to carditis are also possible cardiac manifestations.1
Although the connection between CHB and maternal autoantibodies was first described nearly four decades ago,2 the cascade of inflammation and factors ultimately leading to the fibrotic cardiac remodelling are still poorly understood. Immunohistological and in vitro analyses have shown a role of cardiac IgG accumulation, apoptosis and macrophage activation.3 However, maternal IgG against Ro and/or La alone are not sufficient for the initiation of disease development: in women bearing these autoantibodies, the risk for CHB is only about 2%, with a recurrence rate in subsequent pregnancies estimated at 16%.1
Around 2/3 of mothers with affected children have no defined autoimmune disease at the time of pregnancy,1 and in these cases, the detection of an abnormal fetal heart rate in routine ultrasound check-ups often precedes the measurement of maternal autoantibodies. The remaining mothers are mostly diagnosed with either systemic lupus erythematosus (SLE) or primary Sjögren’s syndrome (pSS),1 which are both pathogenically driven by an upregulation of type I interferons (IFN).4 Until now, a potential role of type I IFN in autoimmune CHB has not been investigated, which was therefore the aim of this study. For this purpose, we measured IFN-α and its response protein SIGLEC1 expressed on CD14+ monocytes in both affected and unaffected mothers. SIGLEC1, also known as sialoadhesin and CD169, was one of the most prominently upregulated genes in the IFN signature of patients with pSS5 and SLE6 and has been shown to correlate with disease activity.7 8 Additionally, we measured IFN-γ-induced protein 10 (IP-10), which is regulated by types I and II IFN and has also been shown to correlate with disease activity in SLE.7
Materials and methods
More information on material and methods is available in the online supplementary material.
Study participants
Patients attending the outpatient unit of the Department of Rheumatology and Clinical Immunology at Charité Universitätsmedizin Berlin were included in the study. Clinical, demographic and serological parameters are summarised in table 1. Nine CHB pregnancies (defined according to Brucato et al 9) had been diagnosed previous to the referral to the outpatient department. Additionally, 14 women with anti-Ro antibodies with unaffected pregnancies (disease controls (DCs)) and 30 healthy pregnant women without antibodies against Ro or La and without a systemic autoimmune disease were included (healthy controls (HCs)). All included women had given birth at the time of analysis, and a telephone follow-up of DC pregnancies was performed that included signs of neonatal lupus syndrome and cardiac complications of the child. The median gestational age at the time of the first measurement was 25 weeks in the CHB group, 21 weeks in the DC group and 19 weeks in the HC group. Longitudinal samples and follow-up information were available from 21 of the 53 mothers (mean±SD follow-up 5.4±5.4 months). pSS was defined according to the revised American-European criteria,10 and SLE was defined according to the updated ACR criteria of 1997.11 The local ethics committee of the Charité Universitätsmedizin Berlin approved the study, and informed consent was obtained from each study participant.
Demographic data
Results
Higher expression of SIGLEC1 on CD14+ monocytes and IFN-α levels in mothers of offspring affected by CHB
In the CHB group, 9/9 mothers (100%) had elevated levels of SIGLEC1 expression, 7/9 (78%) had elevated levels of IFN-α and 7/9 (78%) had elevated levels of IP-10 compared with HC pregnancies. In the DC group, 12/14 (86%) had elevated levels of SIGLEC1 expression, 6/14 (43%) had elevated levels of IFN-α and 8/13 (62%) had elevated levels of IP-10 (plasma was not available for one patient). SIGLEC1 expression and plasma levels IFN-α, but not of IP-10, were significantly higher in the CHB group than in the DC group (figure 1). Receiver operating curve analysis between CHB and DC showed that a SIGLEC1 median fluorescence intensity (MFI) of >904 could distinguish between the groups with a sensitivity of 100% and a specificity of 64%, and a concentration of IFN-α >0.70 pg/mL with a sensitivity of 67% and a specificity of 86%.
SIGLEC1 expression on peripheral CD14+ monocytes and plasma levels of IFN-α and IP-10. (A) SIGLEC1 expression on CD14+ monocytes and IFN-α levels in plasma were significantly higher in the CHB group than in the DC group (p=0.0034 and p=0.014, both MWU). IP-10 did not significantly differ between CHB and DC (p=0.14, MWU). The dashed line represents the calculated threshold for each parameter (SIGLEC1: 358 MFI, IFN-α: 0.05 pg/mL and IP-10: 442.8 pg/mL). (B) SIGLEC1 expression correlated with IFN-α and IP-10 levels (p<0.0001, rs=0.68 and rs=0.71, SRT). (C) The median AB score, calculated by the addition of individual values for anti-Ro60 and anti-Ro52, was higher in the CHB group, but this difference was not significant (p=0.21, MWU). There was a significant difference in SIGLEC1 expression between mothers with an AB score of ≤3 and ≥4 (p=0.0099, MWU). Samples from the CHB group are indicated in red dots; samples from the DC group are indicated in black triangles. All bars represent the median. AB score, antibody score; CHB, congenital heart block; DC, disease control; HC, healthy control; IFN, interferon; IP-10, interferon-γ-induced protein 10; MFI, median fluorescence intensity.
There were five females in the DC group with a high expression of SIGLEC1. Of these, two were diagnosed with SLE and treated with hydroxychloroquine, and three were diagnosed with pSS and not under immunosuppressive or immunomodulatory medication. All five patients were stable in terms of their disease activity. A detailed subgroup analysis can be found in the online supplementary material.
Reduction of SIGLEC1 expression on CD14+ monocytes and IFN-α levels by hydroxychloroquine and oral glucocorticoids
Longitudinal information regarding treatment effects of hydroxychloroquine and glucocorticoids during pregnancy was available from five mothers with offspring affected by CHB. Both medications reduced SIGLEC1 expression and plasma levels of IFN-α (figure 2). Longitudinal analysis of CHB pregnancies showed that women with elevated levels of SIGLEC1 during pregnancy retained elevated levels after childbirth (see online supplementary table S1).
Treatment effects of hydroxychloroquine and glucocorticoids. In the group of mothers with offspring affected by CHB, the expression of SIGLEC1 on CD14+ monocytes (shown in black) and plasma levels of IFN-α (shown in grey) were reduced by both hydroxychloroquine and oral glucocorticoids, with a mean reduction of 1010 MFI SIGLEC1/1.08 pg/mL IFN-α for hydroxychloroquine and 1245 MFI SIGLEC1/1.35 pg/mL IFN-α for glucocorticoids. The mean administered daily dose was 300 mg for hydroxychloroquine and 2.5 mg dexamethasone for glucocorticoids. If treatment was withdrawn, this was done after delivery. CHB, congenital heart block; Dexa, dexamethasone; HCQ, hydroxychloroquine; IFN, interferon; MFI, median fluorescence intensity.
Discussion
An activated type I IFN system in the pathogenesis of maternal autoimmune diseases that are potentially associated with CHB, such as SLE and pSS, and the pleiotropic effects of IFNs on the immune system are well documented. Therefore, it is surprising that the type I IFN signature in the context of CHB development was not studied yet. In our cohort, all mothers bearing a child with CHB were highly positive for SIGLEC1, independently of maternal disease burden. The expression of SIGLEC1 and plasma levels of IFN-α were significantly higher than in the control group.
Although the data regarding placental passage of IFN-α are inconsistent,12 a very recent study showed that an IFN signature could also be found in the cord blood of newborns exposed to maternal anti-Ro/anti-La in utero.13 This indicates that type I IFN may have a direct role in the pathogenesis of CHB. Type I IFN are known to upregulate Ro52 and to induce apoptosis,14 15 which can potentially lead to the release of autoantigens normally located within the cytoplasm or nucleus, making them accessible to maternal autoantibodies.16 Moreover, the placental transcytosis of maternal IgG, which is mediated by FcRn receptors, might be upregulated, since type I IFNs are known to upregulate MHC class I molecules,4 of which FcRn are structural homologues.17 This might facilitate autoantibody trafficking and could explain the findings of a previous study, where higher levels of autoantibodies were found in cord blood compared with maternal blood.18 IFN-α may also be indirectly involved in the pathogenesis of CHB by regulating activated macrophages in affected cardiac tissue, resulting in tissue destruction including fibrosis.3
We describe here, for the first time, a link between CHB and IFN-α. Our data, together with data indicating that anti-Ro/La autoantibodies interrelate with type I IFN,19 and the role of these cytokines in SLE and extraglandular pSS, are consistent with the conclusion that IFN-α is likely to be a key cytokine in the CHB immunopathology. Additionally, with SIGLEC1 we describe a biomarker that is feasible for use in routine clinical settings, for example when considering immunosuppressive medication in pregnant females with known anti-Ro antibodies. SIGLEC1 also appeared to respond to interventions with hydroxychloroquine and oral glucocorticoids, mandating further studies to evaluate its usefulness for monitoring treatment response.
Our study has some limitations that will be addressed in subsequent analyses. The size of the study cohort with n=9 is small owing to the rarity of the disease but is nevertheless representative of about 180 000–270 000 pregnancies. Also, due to the observational character of this study, the maternal diagnoses and treatment differed between the CHB and the DC group, making a statistical analysis difficult. However, the females in the DC group had, overall, a more severe clinical presentation, so that an error derived from this is likely to underestimate rather than exaggerate a difference between the groups. Finally, there were five women in the DC group who had a high expression of SIGLEC1, but bore healthy children. It is therefore likely that a combination of still unknown maternal and fetal factors contributes to a risk profile, which determines the likelihood of CHB development. Further studies are needed to delineate the pathogenic role of enhanced type I IFN for CHB development. Notably, there might be a time window during gestation where this factor becomes a driver of immunopathology, as exemplified by rubella virus infection that also affects the fetal heart within a defined susceptibility phase20 and where IFN-α induction is also likely to be involved.
Acknowledgments
The authors would like to thank Carolin Scholz from the Department of Obstetrics and Gynaecology, Charité Universitätsmedizin Berlin for providing blood samples for HC pregnancies.
References
Footnotes
Contributors ARL, FS, KR, TR and TD: conception, design, analysis and interpretation of data.
ARL: drafting the article.
TD, TR and GRB: revising the manuscript critically for important intellectual content.
All authors read and approved the final manuscript.
Funding CRC initiative of the German Research Foundation (DFG) (SFB 650; TD and GRB). ARL receives a scholarship from the Leibniz Graduate School for Rheumatology (DRFZ).
Competing interests None declared.
Patient consent An own consent form approved by the ethics committee of the Charite Universitätsmedizin Berlin was used and signed by each study participant.
Ethics approval Ethics committee of the Charite Universitätsmedizin Berlin.
Provenance and peer review Not commissioned; externally peer reviewed.