Outcomes of systemic sclerosis with overlapping myositis: a propensity score-matched cohort study
Lekshmi Minikumari Rahulan, Nithin Gowda AV, Abhishek Gollarahalli Patel, Sakir Ahmed, Zafar Neyaz, Vimal Kumar Paliwal, Vikas AgarwalAbstract
Objectives
Systemic sclerosis (SSc) carries the highest disease-specific mortality among rheumatic diseases, with cardiac involvement and interstitial lung disease (ILD) as leading causes of death. We aimed to determine whether overlapping myositis independently contributes to cardiac involvement and mortality in SSc.
Methods
We did a 1:1 propensity score matched retrospective cohort study of 880 SSc patients (2013 ACR/EULAR criteria) at a single centre. SSc myositis required proximal weakness and elevated muscle enzymes without alternative causes. Cardiac involvement used Scleroderma Clinical Trials Consortium criteria (pericardial effusion, ventricular dysfunction, arrhythmia and conduction disease). We used standard and Firth penalized logistic regression for cardiac predictors, Cox regression for survival and Fisher’s exact test for cause-specific mortality. All P-values were Benjamini–Hochberg adjusted.
Results
Of 880 patients, 92 (10.4%) had myositis. After matching, 92 pairs (n = 184) were analysed. Myositis independently predicted cardiac involvement (standard aOR = 19.99, 95%CI 2.19–182.26, P = 0.008; Firth OR = 12.84, 95%CI 2.31–98.47, P = 0.006). Higher mRSS also predicted cardiac involvement (aOR = 2.34 per 10 points, P = 0.006). Anti-Scl70 was less common in myositis (25.0% vs 50.0%, P = 0.009). Overall mortality did not differ (13.0% vs 9.8%, P = 0.620). In non-myositis deaths, ILD caused 33.3%; in myositis deaths, cardiac failure caused 25% and no ILD-related deaths (Fisher’s P = 0.229). Cardiac involvement was significantly associated with mortality [aHR = 8.09 (1.36–48.26), P = 0.022], independent of myositis.
Conclusion
Overlapping myositis in SSc is associated with an increased risk of cardiac involvement and a shift in the cause-of-death profile away from ILD, without a detectable difference in overall survival over the observed median 21-month follow-up. Systematic cardiac surveillance is needed in these patients.