OBJECTIVES: This study assessed the economic models of IO therapy as 1L treatment in advanced or metastatic NSCLC, with the objective to understand the challenges (eg, uncertainty in survival extrapolation, heterogeneity in PD-L1 testing methods) and make recommendations for future CEAs in this area. Here we report our initial step, to review and compare existing practices.

METHODS: A SLR was conducted following PRISMA guidelines. MEDLINE, Embase, EconLit (Jan 2008-Jan 2020), and relevant conferences (since 2016) were searched to identify CEAs of 1L IO NSCLC treatments published in English. Technology appraisals (TAs) from England, Scotland, Canada, Australia, Germany, and France were also identified. Screening and data extraction were conducted by 2 reviewers.

RESULTS: Of 1,456 references identified, 31 publications and 18 TAs met protocol-defined criteria and were selected for data extraction. Most (n=43) compared IO monotherapy or combination therapy to chemotherapy based on head-to-head trials. Seven studies compared multiple IO treatments across trials, however none considered concordance issues across PD-L1 testing methods. Eleven studies used external real-world data for survival modeling or extrapolation validation. Nineteen studies assumed that long-term treatment benefit stopped at 3 or 5 years after initiation. Partitioned survival (n=28) and Markov (n=16) were the most frequent frameworks. Two studies used individual patient-level simulations. Utilities were modeled by progression status in 12 studies, 7 applied the time-to-death approach, and 13 explored both. Eleven studies used the observed time-on-treatment distribution from the trial, while 10 used progression-free survival for treatment duration. Twelve analyses adjusted for treatment switching/crossover, while the rest considered treatment switching/crossover to represent clinical practice, and no adjustment was made. No studies applied cure models.

CONCLUSIONS: Substantial variation was observed in modeling choices. The heterogeneity in PD-L1 testing and external validation of survival extrapolation should be considered for future CEA of IO therapies in advanced or metastatic NSCLC.