Composite versus individual biomarkers for predicting clinical benefit to PD-1 blockade in NSCLC 85 3 Methods Patients, endpoints and samples In this study, 162 patients with pathologically confirmed stage IV NSCLC were eligible for efficacy analysis. All enrolled patients started second or later line monotherapy nivolumab, administered intravenously at a dose of 3mg/kg every two weeks for at least one dose, between October 2014 and August 2017 at the Netherlands Cancer Institute/Antoni van Leeuwenhoek hospital (NKI-AVL), the Netherlands. Patients with tumors harboring known sensitizing EGFR mutations or ALK translocations were excluded from treatment. A randomization process was employed to allocate patients into a training and validation cohort. This randomization was stratified by treatment outcome at 6 months and at 12 months. Since we could only generate gene expression data in 68/162 (42%) of patients’ tumors, additional stratification was done by whether mRNA expression analysis was performed or not. Stratification for missing values of other biomarkers was not performed, as the number of excluded samples per biomarker remained relatively low, ranging from 1 to 32 (see Fig. S1 and later in this section). Response to treatment was evaluated according to the Response Evaluation Criteria in Solid Tumors (RECIST) version 1.1. Patients with progressive disease (PD) who were not evaluable for response assessment were designated by the treating physician as having PD. The primary clinical outcome was Disease Control (DC), defined as achieving a complete response (CR), partial response (PR) or maintaining stable disease (SD) at the 6-month mark following the initiation of treatment. As a secondary outcome measure, DC 12m was employed, representing the persistence of CR, PR or SD for a duration of 12 months or more. This secondary endpoint aimed to serve as an indicator of the long-term efficacy to PD-1 blockade therapy. Pretreatment formalin-fixed paraffin embedded (FFPE) tumor tissue samples were collected from all patients. Written informed consent for the research usage of material, not essential for diagnostic purposes, was obtained from each patient by an institutionally implemented opt-out procedure. The study was conducted in accordance with the Declaration of Helsinki. The data was accessed for research purposes after the approval by the Institutional Review Board (IRB) of the Netherlands Cancer Institute on January 11, 2018 (CFMPB586). After K.H., M.M., R.D.S., M.M.H., E.F.S. and K.M. retrieved archived tumor samples and response data from medical records, all patients were pseudonymized. PD-1T TIL and PD-L1 tumor proportion score (TPS) data for 94 samples as well as tertiary lymphoid structures (TLS) and CD20+ B cell data for 91 samples were used from
RkJQdWJsaXNoZXIy MTk4NDMw