Efficacy and Safety of Therapies for Systemic Sclerosis-Associated Interstitial Lung Disease: A Bayesian Network...

Systemic sclerosis-associated interstitial lung disease (SSc-ILD) is a leading cause of morbidity and mortality in scleroderma. Multiple immunomodulatory and antifibrotic therapies have been evaluated, but their comparative effectiveness remains uncertain.

We conducted a systematic review and Bayesian network meta-analysis of randomized controlled trials evaluating rituximab (RTX), tocilizumab (TCZ), nintedanib (NIN), cyclophosphamide (CYC), mycophenolate (MMF), and pirfenidone for SSc-ILD. The primary network was restricted to trials enrolling patients with HRCT-confirmed SSc-ILD and reporting SSc-ILD-specific outcome data; the broader dataset was analyzed as a sensitivity analysis. Primary outcomes were change in forced vital capacity (FVC) predicted and modified Rodnan skin score (mRSS). Secondary outcomes included the Health Assessment Questionnaire-Disability Index (HAQ-DI), all-cause mortality, serious adverse events (SAEs), and discontinuation due to adverse events; FVC response was analyzed as an exploratory outcome only. Certainty of evidence was assessed using GRADE.

Eight trials with HRCT-confirmed SSc-ILD (1069 patients) formed the primary network, with a broader 12-trial dataset (1451 patients) analyzed as a sensitivity analysis. In the primary network, rituximab improved FVC by 7.76% (95% CrI 2.64-12.67; low certainty) and tocilizumab by 6.47% (95% CrI 2.75-10.29; low certainty), although both estimates rest on single trials with wide credible intervals. Nintedanib showed a modest FVC difference of 1.25 percentage points (95% CrI -1.29-3.78; high certainty). Rituximab probably improved mRSS versus placebo (MD -7.50 units [95% CrI -12.77 to -2.31]), whereas tocilizumab no longer showed a clear skin benefit (MD -1.74 [95% CrI -4.68 to 1.18]). No treatment reduced mortality. In the broader network, nintedanib increased treatment discontinuation (OR 1.91 [95% CrI 1.12 to 3.26]).

In this restricted network of HRCT-confirmed SSc-ILD trials, rituximab and tocilizumab were associated with lung function preservation, and rituximab with clinically meaning skin improvement; however, these estimates derive from sparse, largely indirect evidence and should be regarded as hypothesis-generating. Nintedanib showed modest antifibrotic effects with high certainty. Mycophenolate could be assessed only indirectly and remains an area of uncertainty rather than demonstrated inferiority. Treatment selection should consider both pulmonary and extrapulmonary disease manifestations, and head-to-head trials are needed to establish optimal treatment.