Background/Purpose: Mortality and worsening of organ function would be desirable endpoints for clinical trials in systemic sclerosis (SSc). However, these events are relatively rare, making clinical trial design with these endpoints challenging. To enrich for patients with these events, predictive factors have to be identified. The aim of this study was therefore to identify predictive factors in a population of patients with diffuse SSc from the large European Scleroderma Trials and Research (EUSTAR) group database.

Methods: Inclusion criteria were diagnosis of diffuse SSc and a follow-up after 9–15 (12±3) months. This timeframe was chosen to reflect typical clinical trial design. Disease worsening/organ progression was fulfilled if any of the following events occurred: new renal crisis, decrease in forced vital capacity (FVC) ≥10%, new left ventricular ejection fraction (LVEF) <45% or decrease in LVEF by >10% for patients with baseline LVEF <50%, new pulmonary (arterial) hypertension on echocardiography, or death. These parameters had been defined by expert nominal group technique. Methodologically, two main limitations had to be addressed: (1) the problem of missing data and (2) the low number of events, which prohibits the simultaneous exploration of the set of predictors in a regression model. We addressed these issues by (a) imputing multiple predictors on the basis of different algorithms (multiple imputation), and by (b) using least absolute shrinkage and selection operator (LASSO) regression, thus 42 clinical parameters were entered as predictors into the analysis.

Results: Of the 1451 patients who met the inclusion criteria, 706 had complete data available on all parameters for disease worsening; there were no clinically meaningful differences between patients with and without complete data. Of 706 patients originally evaluated, 228 (32.3%) had disease progression, most of which was either a decrease in FVC (103 patients, 14.6%) or death (92 patients, 13.0%) within the observation period (12±3 months). Of the 42 clinical parameters introduced into the model as outcome predictors, eight remained in the final regression model which was chosen by the Bayesian information criterion (Table 1). The probability, in our model, e.g. for a 60-year-old patient with diffuse SSc, lung fibrosis and active digital ulcers as well as muscle weakness to develop organ progression within the next 12 months increased to >60% compared with 32.3% in the total population. Bootstrap with 10000 repetitions successfully validated the model. The Receiver Operating Characteristic was 0.711 for the final model and 0.705 for the validation. The maximum absolute error in predicted probability was 0.026.

Conclusion: The use of the predictive factors presented here could enable cohort enrichment with patients at risk for overall disease worsening in clinical trials.