Background/Purpose: While prior observational studies have identified predictors of mortality in systemic sclerosis-interstitial lung disease (SSc-ILD), no studies have evaluated predictors of long-term mortality in a clinical trial, in which all patients receive standard care. The objective of this study was to identify predictors of mortality in patients who participated in the Scleroderma Lung Study (SLS) I¹ and II².

Methods: SLS I randomized 158 SSc-ILD patients to 1 year of oral cyclophosphamide (CYC) versus placebo. SLS II randomized 142 patients to 1 year of oral CYC followed by 1 year of placebo versus 2 years of mycophenolate (MMF). The FVC%-predicted and DLCO%-predicted were measured every 3 months for 2 years in both trials. 12 years after SLS I commenced, each study center contacted enrolled patients/designated surrogates to assess morbidity and mortality outcomes. Counting process cox proportional hazard modeling identified variables associated with mortality. A joint model of longitudinal FVC and survival data was used to internally validate the model. We externally validated the model using long-term mortality data from SLS II (up to 5 years of follow-up).

Results: Among SLS I patients, 43% died during the follow-up period (median follow-up: 8 years), and only 24% remained alive without organ failure. Where known, the cause of death was attributable most often to SSc. The most common type of organ failure was respiratory failure (N=31 of 33 organ failures) defined as the need for supplemental oxygen therapy (N=29) and/or lung transplantation (N=3). There was no significant difference in the time to death between patients randomized to CYC versus placebo (Figure 1). The Cox model identified the following mortality predictor variables: baseline skin score (HR 1.03; P=0.004), age (HR 1.06; P<0.0001), and the course of the FVC from baseline to 24 months (HR 0.98; P=0.022). The course of the FVC was a better predictor than the baseline FVC. The joint model identified the same variables associated with mortality. Using the SLS II data, the Cox model identified the same mortality predictor variables: baseline skin score (HR 2.08; P=0.021), age at randomization (HR 1.08; P=0.011), and the course of the FVC from baseline to 24 months (HR 0.79; P=0.020).

Conclusion: Treatment with 1-year of oral CYC for SSc-ILD did not significantly decrease long-term mortality compared with placebo. In addition to identifying traditional mortality risk factors in SSc (i.e. increased skin score and advanced age), this study found that a decline in FVC over 2 years was a better predictor of mortality than the baseline FVC. These findings suggest that early changes in surrogate measures of SSc-ILD progression may have important effects on long-term outcomes.

¹Tashkin et al. NEJM 2006.

²Tashkin et al. Lancet Resp Med 2016.

Figure 1. Time to death in patients randomized to CYC (solid line) and placebo (dotted line).