Abstract 7537: “Cross-over Effect” in survival curves with immune-checkpoint inhibition: Identifying ideal candidates for immunotherapy-only approaches in specific tumor histologies

Abstract Immune checkpoint inhibition (ICI), a cornerstone of modern immunotherapy (IO), has revolutionized cancer treatment, offering significant survival benefits across various malignancies. Initial trials with ICI in refractory cancers demonstrated remarkable efficacy compared to standard-of-care (SOC) treatments. With the integration of ICI into combination therapies, including chemotherapy and multi-tyrosine kinase inhibitors (TKI), its role in frontline, early-stage, and locally advanced cancers has grown. However, the potential of an immunotherapy-only strategy, especially in advanced malignancies, warrants better biomarker selected approaches given the inconsistencies with existing biomarkers. This study investigates the "cross-over” effect observed in Kaplan-Meier survival curves from randomized trials comparing ICI with non-ICI agents where the survival curves, assumed to be proportionally constant distinctly cross over each other during the on-treatment monitoring period. We initiated a meta-analysis investigating randomized trials comparing ICIs with non-ICI historical SOC agents and finally analyzed data from 38 large clinical trials. Trials were selected based on their comparative nature with overall survival (OS) and progression-free survival (PFS) data as outcomes. A positive “cross-over” in Kaplan-Meier curves was identified when divergence occurred for at least three consecutive months. Cross-over in PFS and OS curves were observed in 53% (20/38) and 43% (16/37) of trials, respectively. All cross-overs were one-sided with the ICI arm crossing over the non-ICI arm with a superior tail. Notably, 100% of trials with OS curve cross-over involved ICIs against chemotherapy. A significant association (Fischer’s exact test, p=0.01) was found between OS curve cross-over and adenocarcinoma histologies, in contrast to trials with no cross-over effect, predominantly featuring squamous cell carcinomas. This pattern persisted within various organ sites such as the lung, esophagus, and cervix where both adenocarcinomas and squamous histological patterns are seen. Cross-over of Kaplan-Meier curves blatantly violates the cox-proportional hazard assumption, potentially underestimating the impact of the ICI-only approach compared to alternative therapies. These observations highlight substantial biological heterogeneity in tumor responses to ICI versus chemotherapy. The cross-over effect, particularly pronounced in gland-forming tumors (adenocarcinomas), underscores the urgent need to investigate mechanisms of ICI resistance in these tumor types and incorporate tumor histology and longitudinal tumor growth dynamics into existing predictive models for biomarkers for IO response and help identify patients who would benefit the most from immunotherapy-only approaches. Citation Format: Shubhank Goyal, Anish Thomas, Parth Anil Desai. “Cross-over Effect” in survival curves with immune-checkpoint inhibition: Identifying ideal candidates for immunotherapy-only approaches in specific tumor histologies [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2024; Part 1 (Regular Abstracts); 2024 Apr 5-10; San Diego, CA. Philadelphia (PA): AACR; Cancer Res 2024;84(6_Suppl):Abstract nr 7537.