Mortality in systemic sclerosis: An international meta-analysis of individual patient data

Purpose Studies on mortality associated with systemic sclerosis have been limited by small sample sizes. We aimed to obtain large-scale evidence on survival outcomes and predictors for this disease. Methods We performed a meta-analysis of individual patient data from cohorts recruited from seven medical centers in the United States, Europe, and Japan, using standardized definitions for disease subtype and organ system involvement. The primary outcome was all-cause mortality. Standardized mortality ratios and predictors of mortality were estimated. The main analysis was based only on patients enrolled at each center within 6 months of diagnosis (incident cases). Results Among 1645 incident cases, 578 deaths occurred over 11,521 person-years of follow-up. Standardized mortality ratios varied by cohort (1.5 to 7.2). In multivariate analyses that adjusted for age and sex, renal (hazard ratio [HR] = 1.9; 95% confidence interval [CI]: 1.4 to 2.5), cardiac (HR = 2.8; 95% CI: 2.1 to 3.8), and pulmonary (HR = 1.6; 95% CI: 1.3 to 2.2) involvement, and anti-topoisomerase I antibodies (HR = 1.3; 95% CI: 1.0 to 1.6), increased mortality risk. Renal, cardiac, and pulmonary involvement tended to occur together (P<0.001). For patients without adverse predictors for 3 years after enrollment, the subsequent risk of death was not significantly different from that for the general population in three cohorts, but was significantly increased in three cohorts that comprised mostly referred patients. Analyses that included all cases in each center (n = 3311; total follow-up: 19,990 person-years) yielded largely similar results. Conclusion Systemic sclerosis confers a high mortality risk, but there is considerable heterogeneity across settings. Internal organ involvement and anti-topoisomerase I antibodies are important determinants of mortality.