Risk score for early mortality to stratify for intensive SSc therapy in the EUSTAR network: the RESIST score

Objectives: Systemic sclerosis (SSc) is associated with increased mortality. Intensive therapies have emerged for severe or refractory cases, but carry significant risks. The Risk score for Early mortality to Stratify for Intensive SSc Therapy (RESIST) score was developed to predict early mortality (<5 years) in patients eligible for such treatments.

Methods: Using the European League Against Rheumatism Scleroderma Trials and Research (EUSTAR) cohort, patients unsuitable for intensive therapy were excluded. Survival was analysed via Kaplan-Meier estimates. A multivariable Cox model with adaptive LASSO (Least Absolute Shrinkage and Selection Operator) selection was built, informed by expert opinion, univariable analyses, and literature. Model performance was evaluated using the concordance index, area under the curve (AUC) at 3 and 5 years, and calibration. Missing data were imputed, and pseudo external validation was performed on 6251 excluded patients.

Results: Of 22,059 EUSTAR patients, 4526 met the inclusion criteria; 138 died within 5 years. Deceased patients were more often male (28% vs 16%), older (53 vs 49 years), and had diffuse cutaneous SSc (dcSSc) (61% vs 35%) compared to survivors. The RESIST score included: male sex; dcSSc; age >55 years; elevated C-reactive protein; digital ulcers; modified Rodnan skin score >14; left ventricular ejection fraction <60%; and diffusing capacity of the lung for carbon monoxide <60%. This allowed patients with SSc to be discriminated into 3 groups with the following 5-year survival rates: low risk (99%, 95% CI: 98%-100%), intermediate risk (96%, 95% CI: 95%-97%), and high risk (82%, 95% CI: 78%-87%). The model showed good discrimination in both the development and validation cohorts (AUC: 0.79 [0.77-0.79] and 0.78 [0.77-0.79], respectively).

Conclusions: The RESIST score reliably predicts early mortality in patients with SSc eligible for intensive therapies and may guide treatment decisions by identifying those at high risk of death.