Clinical and autoantibody profiles of systemic sclerosis patients: A cross-sectional study from North India

Plain Language Summary

We conducted this cross-sectional study to determine the clinical phenotypes and frequency of associated autoantibodies in systemic sclerosis patients at a tertiary-care centre. Hundred and six such patients consented to this study and had a mean age at symptom onset of 40 ± 13 years with a median symptom duration of 6 years. Around 71.7% of our patients had interstitial lung disease. Around 58.5% of our cohort had diffuse cutaneous involvement which was significantly associated with anti-Scl70 antibodies, digital ulcers and the presence of interstitial lung disease. The presence of anti-Scl70 antibodies was a risk factor for the presence of interstitial lung disease (ILD) and digital ulcers, while anti-centromere antibodies seemed to be negatively correlated with ILD, but were a risk factor for calcinosis and pulmonary arterial hypertension. The combination of diffuse cutaneous disease and Scl70 antibodies was the strongest predictor of ILD and digital ulceration. Patients with antibodies against RNP68, sm/RNP and Ku had significantly more musculoskeletal involvement. Renal involvement was noted in only two patients. We concluded that North Indian patients have some phenotypic differences as compared to their Caucasian counterparts with a larger proportion of patients presenting with ILD and Scl70 antibodies. The role of minor antibodies deserves special mention and future studies.

Introduction

Systemic sclerosis (SSc), or scleroderma, is an idiopathic systemic autoimmune disease characterized by the triad of vasculopathy, immune dysfunction and fibrosis.¹ The prevalence of SSc has been reported to be as high as 276 to 658 per million^2,3 in epidemiological studies from the United States of America (USA), while studies from Asia have reported lower prevalence rates of around 56 per million.⁴ A study by Mintz et al. retrospectively analyzed data from ANA-positive patients in an Indian setting between 1996 and 2006 and estimated the prevalence to be around 120 per million population.⁵

The pathogenetic processes in SSc have been extensively studied and some autoantibodies seem to have strong associations with organ involvement. The two major phenotypic subtypes are based on the extent of skin involvement; data from various studies have supported the claim that the diffuse and limited cutaneous subtypes might be clinically distinct entities with differences in severity of organ involvement as well as their autoantibody associations. The most common disease-specific autoantibodies reported in patients are anti-Topoisomerase I (anti-Scl70) and anti-centromere antibodies (Anti-Centromere),⁶ but other multiple autoantibodies have shown a greater association with SSc than with other autoimmune diseases.

There is a dearth of data from large cohorts in India that may help to uncover the true prevalence of SSc in the population and thus help to understand some of the geographic and ethnic differences in autoantibody profiles and phenotypes of patients noted from around the globe.⁷ Being a rare disease with considerable heterogeneity in presentation, SSc leaves a rather large gap in the characterization of its autoantibody associations and correlation with clinical phenotypes. We designed this single-centre, cross-sectional study from Medanta—the Medicity Hospital, Gurgaon, India to study the clinical and serological characteristics of SSc patients and uncover any differences between patients of North India and those from other countries.

Materials and Methods

Study Design: This cross-sectional study was conducted at the Department of Rheumatology and Clinical Immunology at Medanta—the Medicity Hospital, Gurgaon, from August 2017 to June 2019.

Patient Details: We collected the data of all consecutive patients attending our outpatient or inpatient departments from August 2017 to June 2019, who were classified as SSc as per the American college of rheumatology/European league against rheumatism 2013 criteria⁸ for clinical trials. We excluded pregnant females, patients less than 18 years of age and those who did not give consent to be included in this study.

Methodology: The patients were subjected to various predefined tests after obtaining informed and written consent. Demographic details, clinical features, laboratory parameters and available imaging reports of patients were analysed. Disease characteristics including disease duration and disease subtype, that is, limited cutaneous (lc) or diffuse cutaneous (dc) SSc, were recorded. Patients without skin involvement were included in the limited cutaneous group. Skin involvement was assessed by a modified Rodnan skin thickness score (mRSS) by a single assessor. Pulmonary involvement was assessed by high-resolution computed tomography (HRCT) of the thorax and, if present, further classified as non-specific interstitial pneumonia (NSIP) or usual interstitial pneumonia (UIP) according to a trained radiologist. Joint disease was defined by history or current arthralgia/arthritis. Renal involvement was assessed by serum creatinine, urine routine and microscopy and presence of scleroderma renal crisis. Gastrointestinal involvement was assessed by the history of dyspepsia, abdominal bloating, dysphagia, or frequent diarrhoea. Oesophageal Manometry was performed in selected cases with upper gastrointestinal complaints. Vasculopathy was defined by the presence of digital ulcers, gangrene, Raynaud’s phenomenon with at least a biphasic colour change in digits in response to cold exposure, or pulmonary hypertension (PH) in the absence of primary lung disease. Nailfold capillaroscopy was performed using a Video capillaroscope in selected patients. Myositis was defined by the presence of weakness and elevated muscle enzymes in the absence of other clinical or serological features suggestive of idiopathic inflammatory myopathy. Cardiovascular involvement was assessed by 2D echocardiography findings and a cut-off value of Pulmonary Arterial Systolic Pressure (PASP) above 35 mm Hg was kept as the definition of pulmonary hypertension. We defined anaemia in our population with a value of haemoglobin <12 mg/dL.

After the collection of blood samples, sera were stored at −80℃ in aliquots until being tested. Anti-nuclear antibodies (ANA) were reported by the indirect immunofluorescence (IIF) technique on Hep-2 cells. Extractable Nuclear Antigen (ENA) Profile included testing for 25 autoantibodies with a line immunoassay (LIA) method using the BlueDiverQuantrix- Anti-nuclear antibodies25 Screen IgG kit from D-tek, Belgium and a value of 6 U/mL or less for each antibody was considered as negative as per the prespecified cut-off value of this kit. Standard blood tests like complete haemogram, ESR, liver and renal function tests and creatine phosphokinase were carried out on all patients.

Statistical Analysis: Continuous quantitative data were presented in terms of means and standard deviation. Qualitative/categorical data were presented as absolute numbers and proportions. Cross tables were generated and chi-the square test was used to test associations between grouped variables. Statistical significance was set at P < 0.05. SPSS software Version 24.0 was used for statistical analysis.

Compliance with Ethical Standards: Patients were included in the study after taking informed consent. The study has been approved by Institutional Review Board at Medanta—– the Medicity, Gurgaon [MICR- 771/2017].

Results

During the defined study period, a total of 119 consecutive patients meeting the American college of rheumatology/European league against rheumatism 2013 criteria for SSc were screened and their demographic and disease parameters were noted. Of these, 106 patients consented to be included in this study; their blood samples were collected for serology and they were all included in the final analysis.

Demographic Characteristics: The mean age at evaluation of the 106 patients was 47.57 ± 12.19 years with a median symptom duration of 6 years (IQR 3-10). The mean age at onset of symptoms was 40 ± 13 years. There were 92 (86.8%) females, 46 (50%) of whom were post-menopausal and 14 (13.2%) males in our cohort with a female-to-male ratio of 6.6:1. The median modified Rodnan skin thickness score was 9. Forty-two (39.6%) patients had at least one comorbidity.

Table 1 summarizes the clinical features of our patients at evaluation.

The autoantibody profile in our cohort has been described in Table 2. We had six patients with positive anti-nuclear antibodies on immunofluorescence with no antibody positivity on the extractable nuclear antigen profile.

Organ involvement and autoantibody correlation:

Table 3 highlights the correlations between disease characteristics and autoantibody profiles of patients in our cohort.

Discussion

Data from SSc studies have reported variations in the prevalence of disease phenotypes and serologic associations from different geographic and ethnic cohorts.¹⁴ The strength of our study lies in the fact that all of our patients have been tested for the presence of 25 autoantibodies at our specialized immunology laboratory. From our search of the current literature on autoantibody profiles in scleroderma patients, no data is available from a single centre where 25 autoantibodies have been tested in a large cohort. This helped us report minor proportions of patients with relatively rare autoantibodies, although the sample size restricts us from labelling some of these rare autoantibodies as statistically significant. Table 4 compares the cohort characteristics in our study with some of the existing large scleroderma cohorts from multiple geoethnic backgrounds.

We had no patients with coexisting anti-Scl70 and anti-centromere antibodies, similar to most reported populations.¹² The mean age at symptom onset was 40 ± 13 years. Data from a South Indian cohort¹⁰ had explored the possibility of a more severe disease phenotype than Caucasians, especially with a younger age at onset, while Morgan and colleagues reported higher frequencies of younger age at onset and more severe cutaneous and pulmonary disease in African Americans from their GRASP cohort¹⁵ when compared with their EUSTAR counterparts. Diffuse cutaneous involvement (dcSSc) was significantly associated with the presence of interstitial lung disease (ILD), current or past digital ulcers, lower mean BMI values and Scl70 positivity. On the other hand, limited cutaneous disease (lcSSc) was significantly associated with calcinosis and anti-centromere antibodies. Nineteen (43%) of the limited cutaneous disease patients had Scl70 antibodies and interstitial lung disease was present in 78.9% of them. Among the diffuse cutaneous involvement patients, 89% of the group with anti-Scl70 antibodies had ILD, as compared to only 62% of the ones without anti-Scl70. This sero-discordance has been described in a study from the Spanish RESCLE cohort in 2018,¹⁶ where the prevalence and severity of ILD were highest in their diffuse cutaneous involvement/anti-Scl70 group as well, while isolated pulmonary hypertension was more in the diffuse cutaneous involvement/anti-centromere group. Calcinosis was seen significantly more in the limited cutaneous disease/anti-centromere group. Interestingly, all seven of our patients who had Pm/Scl antibodies had ILD, but this number was too small for the association to be statistically significant (p = 0.085). The GRASP cohort had shown an association of Pm/Scl antibodies with limited cutaneous disease only.

The prevalence of ILD was more than twice the values reported from the EUSTAR cohort¹³ but similar to data from Indian¹⁰ and African American¹⁵ studies. This may be partly explained by a referral bias based on dyspnea or pulmonary findings by a respiratory medicine specialist, while there is a diagnostic delay in patients who may ignore early skin changes. A majority (61.3%) of our patients had anti-Scl70 antibodies which were also significantly associated with the presence of ILD (commonly the non-specific interstitial pneumonia type), lower mean BMI and digital ulcers. Anti-centromere positivity, on the other hand, correlated significantly with calcinosis, the absence of ILD and pulmonary hypertension. We reported a higher prevalence of anti-Scl70 positivity than in Caucasian (36.8%) and African American (30%) studies. In the 66 patients with PFT data available, 55% had a restrictive defect and 36.4% and 6.1% of these patients had moderate and severe restrictions of FVC, respectively. The EUSTAR cohort reported 32% of patients with FVC restriction and the South Indian study by Janardana et al. had 66% of their patients with moderate to severe FVC restriction. Given the greater proportion of diffuse cutaneous disease in our cohort, this prevalence of anti-Scl70 being higher than usual may be partly explained.

Around 26.4% of our patients had pulmonary hypertension and this is slightly less than that reported from Western India (47%) and the GRASP cohorts (30%). We have based our diagnoses on 2D Echocardiography findings, although right heart catheterization (RHC) may be a better modality in this regard. One patient with heart block and one with atrial fibrillation had Ro52 antibodies. Cardiac manifestations were not significantly associated with any of the autoantibodies. We could document renal involvement in the form of proteinuria with hypertension in only two patients which was similar to the EUSTAR data, but much lower than the West Indian¹⁰ or African American¹⁵ cohorts.

Strengths and limitations

The strength of our study is the routine testing for a set of 25 autoantibodies in all our patients, based on which we found 17 patients with RNP68 or Sm/RNP positivity and significant association with current or past synovitis. We also reported three patients with Ku antibodies, all of whom had musculoskeletal features and limited cutaneous disease. These reports of serologic profiles are lacking in most of the existing Indian and international cohorts.

Our study has certain limitations. A single centre may not be enough to capture the true prevalence of disease characteristics in the population. Referral bias has allowed us to collect more patients with diffuse cutaneous disease and anti-Scl70 positivity. To determine clinically meaningful patterns and associations in SSc, we shall need larger samples and data from multiple centres. We were also unable to provide data on RNA-Polymerase antibodies in our study, as the current line immunoassay kit does not provide for it.

Conclusion

Our cohort of North Indian patients had a similar age of onset as their European counterparts, while there was a substantially higher proportion of patients with interstitial lung disease. Two-thirds of our patients were positive for anti-Scl70 antibodies and this was strongly associated with the presence of interstitial lung disease as well as digital ulcers. All patients with anti-Pm/Scl antibodies had interstitial lung disease and antibodies against RNP68 or Sm/RNP were associated with arthritis. Anti-Ku antibodies were rarely reported and these patients had musculoskeletal involvement. The traditional understanding of dividing SSc into limited or diffuse cutaneous phenotypes is now evolving and it may be beneficial to add serologic variables to our diagnostic workup if we are to predict organ involvement and prognosis of individual patients more accurately.