Transportability of patient outcomes from a US clinical trial to real-world populations - a case study using Lung-MAP S1400I (NCT02785952)

Abstract

2.AbstractBackgroundThe external validity of results from clinical trials to routine clinical practice is often questioned. This is sometimes because certain real world patient groups are excluded or underrepresented in clinical trials, or because standards of care in trials are different from those in real-world populations globally. This lack of external validity of trial results manifests as an efficacy-effectiveness gap. In this study, we aim to address the question of whether it is possible to extend results from a clinical trial to real-world populations across different countries. To do this, we use the Lung-MAP nonmatch sub-study S1400I trial (NCT02785952) as a case study.SettingSquamous cell lung carcinoma is a subtype of non-small cell lung cancer (NSCLC) accounting for 25-30% of cases. Compared to other NSCLC subtypes such as adenocarcinoma, the presence of actionable genetic variants is less common and there are fewer targeted therapies available for advanced/metastatic NSCLC (aNSCLC) of squamous subtype. Patients with squamous aNSCLC who progress on front-line chemotherapy commonly receive immunotherapy using immune checkpoint inhibitors such as nivolumab. The Lung-MAP nonmatch sub-study S1400I (NCT02785952) compared overall survival (OS) in patients with recurrent/stage IV squamous NSCLC randomized to receive either nivolumab monotherapy or nivolumab + ipilimumab combination therapy and found no significant difference in mortality rates between these groups. The trial included patients from the United States only.ObjectivesThe goal of this study is to evaluate the transportability of results fromNCT02785952in United States patients to real-world populations in the United States, Germany, France, England and Japan. Using individual-level data for OS fromNCT02785952, we will adjust for baseline characteristics from published studies of real-world populations in these countries and benchmark the predicted OS against Kaplan-Meier estimates reported by these studies for patients with squamous cell aNSCLC treated with nivolumab. Sensitivity analyses for unmeasured prognostic variables will be performed.