Background/Purpose: In patients with systemic sclerosis (SSc), peripheral vasculopathy can promote critical ischemia and gangrene that are very severe complications with potential life-threatening consequences. Recently the DUO registry suggested a 18% prevalence of gangrene in SSc patients with digital ulcers (DUs), with smoking and a high number of DUs being predictive factors. However, little is known about gangrene in unselected SSc patients. The aim of this study was to investigate the prevalence, incidence and risk factors for gangrene in the EUSTAR cohort.

Methods: We included patients from the EUSTAR database fulfilling the ACR 1980 or the ACR/EULAR 2013 classification criteria for SSc, with at least one visit recording data on gangrene. We extracted from this database data regarding the reporting of DUs, DU history and gangrene. Centers were asked for supplementary data on traditional cardiovascular (CV) risk factors. We analyzed by uni- and multivariable logistic regression the cross-sectional relationship between gangrene and its potential risk factors such as history of DUs, cutaneous subset, disease duration, autoantibodies, classical cardio-vascular risk factors. Longitudinal data were analyzed by Cox proportional hazards regression.

Results: 1757 patients matched the inclusion criteria (age at inclusion 55.9±14.5 years, disease duration since first non-Raynaud’s symptom 7.9±10.3 years, male sex 16.7%, 24.6% diffuse cutaneous subset (DcSSc)). At inclusion, 8.9% of patients had either current or previous digital gangrene, 15.7% had current DUs and 42.7% have had DUs ever. Among the potential risk factors, older age, a history of DUs and the DcSSc subset were statistically significant risk factors in the cross-sectional multivariable model (Table).

For the longitudinal part, during the entire follow-up (median [Q1,Q3] 13.1 [9.6, 19.3] months), 16/771 patients had incident gangrene (0.9%), accounting for an incidence of 1.94/100 patient-years. All 16 patients who developed incident gangrene had previously had DUs and gangrene.  Further risk factors for incident gangrene were the DcSSc subset and longer disease duration.

Table.  Factors associated with gangrene in patients with SSc from the EUSTAR cohort: univariable and multivariable logistic regression on cross-sectional data from the first visit mentioning data on gangrene.

Conclusion: In unselected SSc patients, gangrene still occurs in 9% of SSc patients. A history of DUs and, to a lesser extent, the DcSSc subset are strongly and independently associated with gangrene, while traditional CV risk factors were not identified as risk factors. Our results confirm that gangrene is still a concern in SSc. They emphasize on the importance of microvascular SSc-associated disease in the pathogenesis of gangrene and suggest that the DcSSc subset should be prioritized for risk-stratification of the patients.