Abstract A17: Predictions of comparative clinical outcomes for checkpoint inhibitor combo therapies and mechanistic targets in first-line NSCLC

Abstract

Abstract The success of checkpoint inhibitors for immunotherapy has ushered in a new era of cancer treatment. While response rates of approximately 20% are observed with checkpoint monotherapy in NSCLC, it remains one of the most deadly forms of cancer. Combinatory therapy approaches have been rapidly implemented, with the selection of agents largely based on theoretical considerations (Kargl, 2017 #19227). The high number of new clinical trials of anti-PD1 with a combination therapy raises the concern that adequate patient enrollment may become a barrier to trial completion and approval. Hence, drug candidate selection and identification of patients most likely to benefit from treatment are emerging needs of the field. Further, irregularity in patient enrollment criteria and trial designs prevents the direct comparison of regimens, impeding clinical decision making. To address rationale immunotherapy decisions for first- line NSCLC patients, we developed a mechanistic, systems biology model to compare the efficacy of 4 candidate therapies in combinations with an anti-PD1 therapy. Our model captured mechanisms underlying the pro- and anti-inflammatory influences of the innate immune system on Tregs and T cell effector function. “Simulated patients” were then created based on patient level data of 30-60 baseline measurements from preclinical and clinical publications in squamous and nonsquamous NSCLC patients. The model permeated all vectors and combinations of data. Each unique set of permutations that fit within clinical monotherapy data from platinum doublet chemotherapy SOC, anti-PD1, anti-CTLA4, and EGFR inhibitor trials in NSCLC represented a simulated patient. Over 200 patients were simulated to create a virtual trial population. As a performance validation, our model blindly predicted the NSCLC results from published combination trials of anti-PD1 + CTLA4 and anti-PD1 + EGFR. The model then was used to predict results for an anti-PD1 + Lag3 and anti-PD1 + CCL2, for which no trial results are available. We conducted a direct comparison between the co-therapy regimens, controlling for trial design and using the same simulated patient population for each trial arm. The large and diverse simulated patient population allowed us to evaluate trends in the characteristics of the tumor microenvironment of differential responders/nonresponders on one regimen vs. another, such as the immune cell composition, biomarker levels, and cytokine levels. Citation Format: Kadie Vanderman, Andrew Stine, Steven Chang. Predictions of comparative clinical outcomes for checkpoint inhibitor combo therapies and mechanistic targets in first-line NSCLC [abstract]. In: Proceedings of the AACR Special Conference on Tumor Immunology and Immunotherapy; 2018 Nov 27-30; Miami Beach, FL. Philadelphia (PA): AACR; Cancer Immunol Res 2020;8(4 Suppl):Abstract nr A17.