Clinico-genomic characteristics of clinical trial participation and its impact on clinical outcome in metastatic NSCLC: A nationwide database analysis in Japan.

Abstract

8626 Background: The clinico-genomic factors influencing clinical trial participation and their impact on clinical outcomes remain unclear. We investigated which clinico-genomic characteristics predict clinical trial participation in patients with metastatic NSCLC and whether trial participation improves clinical outcomes compared to nonparticipation, using a nationwide database. Methods: We retrospectively analyzed 2,966 patients with metastatic NSCLC who underwent comprehensive genomic profiling (CGP) testing from March 2019 to December 2023 in the Center for Cancer Genomics and Advanced Therapeutics, a nationwide database. Multivariable logistic regression identified clinico-genomic factors associated with trial participation. Multivariable Cox model compared OS between trial participants and non-trial participants, adjusting for age, sex, smoking status, performance status, liver/brain metastases, FDA-approved/potentially druggable genes, PD-L1 tumor proportion score (TPS). Overall response rate (ORR) by the line of therapy was evaluated. Results: Of 2,966 patients, 167 (6%) participated in clinical trials. In the multivariable analysis, EGFR mutation (mut), non-squamous (sq) histology, and male sex were associated with a higher likelihood of trial participation, whereas STK11 mut was associated with a lower likelihood of trial participation. After stepwise selection, biomarker factors ( EGFR mut, KRAS G12C mut, RET fusion, MET exon 14 skipping mut, PD-L1 TPS ≥ 50%) explained 72% of the increased trial participation odds, while clinical factors (age < 65, male sex, non-sq histology) accounted for 28%. Among patients with lung adenocarcinoma (LUAD), KRAS G12C and male sex predicted higher trial participation; among those with lung sq cell carcinoma (LUSC), KRAS G12C predicted higher trial participation. Participation did not confer an OS benefit in the overall NSCLC cohort (HR, 0.94; 95% CI, 0.74–1.19), in LUAD (HR, 1.03; 95% CI, 0.78–1.37), or in EGFR -mut patients (HR, 0.99; 95% CI, 0.52–1.88), but was significantly associated with improved OS in LUSC (HR, 0.29; 95% CI, 0.10–0.78). ORR was not significantly different between trial participants and non-trial participants in 1L (49% vs. 57%, P=0.25), 2L (30% vs. 34%, P=0.75), and 4L (25% vs. 19%, P=0.56) therapy lines, but was significantly higher among trial participants in 3L (46% vs. 23%, P=0.002) and ≥5L (48% vs. 18%, P=0.0001). Conclusions: Biomarker factors contributed to 72% of the likelihood of trial participation in patients with metastatic NSCLC, underscoring the importance of CGP. Clinical trial participants exhibited survival outcomes comparable to those of nonparticipants. Their higher ORR in later lines of therapy suggests that clinical trial participation may be a potent therapeutic option, particularly after standard treatment.