Generalizability of immune checkpoint inhibitor trials to real-world patients with advanced non-small cell lung cancer

Abstract

IntroductionBased on data from randomized controlled trials (RCTs), immune checkpoint inhibitors (ICI) are standard-of-care in advanced non-small cell lung cancer (aNSCLC). However, trial eligibility criteria are restrictive, and participants and outcomes may not represent the wider population. We aim to assess the generalizability of key phase III RCTs to real-world patients. MethodsAmong aNSCLC patients enrolled in the Embedding Research (and Evidence) in Cancer Healthcare (EnRICH) program between 26/6/17–18/2/21, we assessed the proportion of patients who fulfilled key trial eligibility criteria: performance status (PS) 0–1, normal laboratory results, no EGFR/ALK mutations, no exclusionary comorbidities (previous cancer, conditions necessitating steroid use, autoimmune diseases, HIV, hepatitis B/C, tuberculosis, interstitial lung disease, organ transplantation). We defined patients who met all assessed criteria as trial-typical and describe ICI uptake and overall survival (OS). ResultsOf 454 patients (median age 71 years, 42.1% female), 30% were trial-typical. Less than half received ICI (47.6%), with trial-typical patients more likely to receive ICI (69.1% vs 38.4%, adjusted odds ratio 3.77, 95% CI 2.40–5.91). Median OS was 10.2 and 5.4 months in patients receiving first- and second-line ICI, respectively. Rationalizing trial criteria to include patients with PS ≤ 2 and exclude those with targetable mutations, steroid use, autoimmune diseases, interstitial lung disease, tuberculosis or organ transplantation increased the proportion of trial-typical patients to 57.3%. ConclusionsLandmark phase III RCTs in aNSCLC have limited generalizability. OS of real-world patients receiving ICI is shorter than reported in trials. Novel ICI trials should consider broader eligibility criteria to improve their generalizability.