Efficacy of biomarker-matched therapy in clinical trials for advanced gastrointestinal cancers: A pooled analysis of SCRUM-Japan studies.

Abstract

753 Background: We conduct the SCRUM-Japan MONSTAR-SCREEN, a nationwide molecular profiling project to facilitate the enrollment of patients with advanced solid tumors into matched clinical trials based on identified biomarkers. Here, we investigated clinical outcomes of patients with gastrointestinal cancers in this project. Methods: The SCRUM-Japan GI-SCREEN was launched for gastrointestinal cancers in 2015, followed by GOZILA, MONSTAR-SCREEN-1, and MONSTAR-SCREEN-2 for advanced solid tumors. The used profiling assay for genomic alterations was Oncomine Comprehensive Assay for tumor tissue in GI-SCREEN, Guardant360 for plasma in GOZILA, FoundationOne CDx for tissue and FoundationOne Liquid CDx for plasma in MONSTAR-SCREEN-1, and CARIS MI Profile for tissue in MONSTAR-SCREEN-2. Patients were treated in clinical trials or practice based on identified biomarkers. We analyzed data from patients with gastrointestinal cancers in this project. Results: Of 11,408 patients enrolled in our project as of May 22, 2023, 555 (5.0%) were enrolled in matched clinical trials based on identified biomarkers. The major cancer types that had matched clinical trials included colorectal (63.2%), biliary tract (12.4%), gastric (n=9.2%), esophageal (n=6.5%), and pancreatic cancer (n=4.7%). The objective response rate (ORR), median progression-free survival, and median overall survival (OS) for patients in matched trials were 27.8% (95% CI, 24.0% to 31.9%), 3.0 months (95% CI, 2.8 to 3.7 months), and 14.4 months (95% CI, 13.0 to 16.1). The major treatment lines in which investigational drugs were administered were third (24.6%) and fourth (16.0%) line. Evaluating by each drug target given to at least 20 patients, therapies targeting HER2 had the highest ORR of 46%, followed by PD-1/PD-L1 (33%), BRAF (30%), and MEK (28%). Antibody-drug conjugates demonstrated the highest ORR with 48.8%, followed by monoclonal antibodies at 37.7% and small molecule inhibitors at 23.0%. Overall, patients who received matched therapy in clinical trials or practice had significantly longer OS than those who did not (hazard ratio, 0.76; 95% CI 0.69 to 0.84; P <0.01). Conclusions: Our nationwide molecular profiling project has facilitated the enrollment of patients with advanced gastrointestinal cancers in clinical trials. Furthermore, it demonstrated a survival benefit by providing patients matching targeted therapy.