Low-dose vs standard-dose nintedanib at initiation in fibrosing interstitial lung disease: a target-trial emulation of treatment persistence and lung function.

Abstract

Background: Early intolerance to nintedanib may shorten time spent on treatment. Whether initiating treatment at 100 mg vs 150 mg twice daily improves treatment persistence for 12 months without compromising lung function is uncertain.

Objectives: To estimate the effects of 100 mg vs 150 mg at initiation on 12-month time on treatment and ∼1-year lung function.

Methods: An observational emulation of an individually randomized target trial in a single-center new-user cohort was conducted. Baseline confounding was addressed using stabilized inverse probability of treatment weighting (IPTW) with 1st to 99th percentile trimming. The primary estimand was the difference in the restricted mean time on treatment (RMST) over 12 months. Key secondary estimands were the IPTW-adjusted absolute risk difference (RD) in 12-month treatment discontinuation (accounting for competing death) and the mean difference in change in forced vital capacity (ΔFVC) %predicted at ∼12 months among survivors. Uncertainty was quantified by patient-level bootstrapping.

Results: Among 172 initiators (100 mg, n = 94; 150 mg, n = 78), 100-mg initiation extended the 12-month RMST by 52.9 days (95% CI, 11.8-97.1 days) and decreased the 12-month discontinuation risk (cumulative incidence function, 0.135 vs 0.281; RD, 0.146). The ΔFVC %predicted was broadly similar among survivors, but the estimates were imprecise.

Conclusions: In this target-trial emulation, starting at 100 mg twice daily improved 12-month treatment persistence. One-year FVC %predicted estimates among survivors were broadly similar but imprecise and should be interpreted with caution; a start-low, escalate-as-tolerated policy warrants prospective evaluation.