Background The results of the first randomized controlled trial comparing cyclophosphamide (CYC) and mycophenolate (MMF) for systemic sclerosis-related interstitial lung disease (SSc-ILD) demonstrated that MMF had improved safety and tolerability profiles compared with CYC over 2 years (Tashkin ACR Abstract 2015); however, the comparative long-term effects of these agents are unknown.
Objectives To compare survival of patients randomized to MMF versus CYC in Scleroderma Lung Study (SLS) II and to determine which baseline factors predict survival.
Methods All SSc-ILD patients in SLS II (Comparing 2 years of MMF versus 1 year of CYC followed by 1 year of placebo) were eligible for inclusion. During the 2-year trial, mortality outcomes were collected by the individual study sites and sent to the SLS II morbidity and mortality committee for adjudication. Following the trial, deaths were ascertained from reports by the next of kin or the coordinator/principal investigator at the study site, as well as reports from online obituaries and the National Death Index. Kaplan Meier curves were generated and the log-rank test was used to compare survival rates. Cox proportional hazards modeling was used to determine which baseline demographic and disease features, as well as which molecular markers predict survival.
Results Of the 142 patients in SLS II, contact was established (or mortality was confirmed) with 88%. The median follow up was 3.3 years from randomization. 16 deaths occurred during the 2-year trial (5 MMF; 11 CYC), and an additional 11 deaths occurred after the trial (6 MMF; 5 CYC). One patient underwent lung transplantation (CYC). The majority of deaths were deemed to be due to respiratory failure (44%), followed by congestive heart failure (7%) and lung cancer (7%). Patients assigned to MMF had a trend for improved survival at 1- (HR 3.8; P=0.09), 2- (HR 2.1; P=0.1), and 3-years (HR 1.9; P=0.1) post-randomization. However, by 3.5 years, there was no difference in survival between treatment arms (Figure 1). In univariate analysis, male gender (HR 2.2; P=0.04), age (HR 1.1; P<0.001), disease duration (HR 0.7; P=0.02), and baseline DLCO (HR 0.9; P=0.02), were associated with increased risk of death. Of the molecular markers, higher baseline level of the chemokine CCL18 was associated with increased mortality (HR=1.003; P=0.05). In the multivariate analysis, the best 1-variable model for predicting survival was age (HR 1.1; P<0.001). The best 2-variable model for predicting survival was age (HR 1.1; P<0.001) and disease duration (HR 0.7; P=0.02), and the best 3-variable model for predicting survival was age (HR 1.1; P<0.001), disease duration <3 years from the onset of the first non-Raynaud's symptom (HR 4.6; P=0.008), and baseline FVC (HR 0.9; P=0.004).
Conclusions Treatment with MMF for SSc-ILD appears to be associated with improved survival during the first 3 years following treatment initiation; however, beyond this time frame, there is no clear MMF survival advantage. Of the factors associated with SSc-ILD survival in prior observational cohorts, increased age, shorter disease duration, and lower baseline FVC were the most important predictors of mortality in SLS II participants.
Disclosure of Interest None declared