Background Anti-centromere antibodies (ACA) typically associate with limited cutaneous disease subset (lcSSc) in scleroderma, and are protective for renal crisis (RC) and interstitial lung disease (ILD). Only 5–7% of ACA positive patients have the diffuse subset (dcSSc). Both antibody specificity and disease subset may influence disease phenotypic expression and organ manifestations.
Objectives To describe demographic and clinical manifestations of ACA positive patients with dcSSc (ACA+diffuse), comparing with two other subsets: ACA positive with lcSSc (ACA+limited) and ACA negative with dcSSc (non-ACA diffuse).
Methods Retrospective study of consecutive patients ACA+diffuse from a single large referral center between 2001–2015. Comparative analysis was made with 160 consecutive patients ACA+limited and 260 consecutive patients non-ACA diffuse. Data were obtained by medical records review.
Results 1313 patients ACA+ were identified, 36 (2.7%) had dcSSc. The peak modified Rodnan skin score was lower in ACA+diffuse compared with non-ACA diffuse (24.0±9.9 vs 27.4±10.4, p=0.031), but occurred later in disease course (88.8±77.2 vs 30.7±33.1 months, p<0.001).
Survival in ACA+ was similar for both subsets with 5, 10 and 15 years survival rates of 96%, 84% and 73% in ACA+limited and 94%, 80% and 72% in ACA+diffuse. Non-ACA diffuse had lower survival of 85%, 72% and 55% at 5, 10 and 15 years respectively, although the difference from ACA+diffuse was not significant.
ACA+diffuse had higher incidence of ILD than ACA+limited (p=0.018), but significantly lower than non-ACA diffuse (p=0.003). During follow-up at 5 years, 15% of ACA+diffuse developed ILD (3% - ACA+limited; 36% - non-ACA diffuse). At 15 years, cumulative incidence of ILD was 26% in ACA+diffuse (5% - ACA+limited; 49% - non-ACA diffuse).
More patients developed pulmonary hypertension (PH) in ACA+diffuse group (27.8%), than in the other groups (12.5% - ACA+limited; 11.9% - non-ACA diffuse), though this was mainly due to longer follow-up of ACA+diffuse. Cumulative incidence of PH in ACA+diffuse was not different from the other groups. At 5 years, 8% of ACA+diffuse had PH (6% - ACA+limited, 6% - non-ACA diffuse), while at 15 years cumulative incidence was 24%, 20% and 18% respectively.
Similarly, cardiac involvement was more frequent in ACA+diffuse (8.3%) vs 1.9% in ACA+limited and 6.5% in non-ACA diffuse, but differences were not significant adjusting for time of follow-up. In contrast, incidence of RC was higher in non-ACA diffuse (13.9%), affecting only 2 patients in ACA+diffuse (5.6%) and none in ACA+limited.
Conclusions This study confirms that ACA diffuse subset is infrequent, but has a distinct clinical phenotype. Despite the dcSSc, these patients have a more insidious onset of skin and major organ involvement, which may represent a therapeutic window for early intervention. We confirm that ACA+ has a protective role and is associated with lower incidence of ILD and RC and better survival than expected in dcSSc. The role of ACA in patients with dcSSc as a possible phenotype modifier needs to be explored further in larger prospective studies.
Disclosure of Interest None declared