1756 - Combination Therapy Optimization in Gastrointestinal Cancers using Multi-omic Molecular Profiling

Background

The majority of new cancer drugs in development target molecular pathways expressed in increasingly smaller subsets of patients. Combination trials will dictate even further enrichment and smaller percentages of eligible patients. Multi-omic molecular profiling (MOMP) from a national research consortium applied to 6 large oncology pipelines permits prospective feasibility and prioritization of clinical research projects.

Methods

5377 gastrointestinal cancers that underwent multi-omic molecular profiling (MOMP) were retrospectively analyzed to estimate the potential actionability of single agents and pairwise combinations. Patient-specific actionability was assessed using a heuristic algorithm to match therapeutic agents across 6 oncology pipelines to biomarkers implicated in preclinical and clinical studies.

Results

Single agent actionability was highest across all 5377 patient tumors for agents targeting PI3K/AKT/mTOR signaling (24%), replicative stress via Chk1 (19%), DNA damage repair via PARP/ATR/ATM (15%), RAF/MEK/ERK signaling (13%), PD-1/PD-L1 (7%), and NOTCH signaling (7%). For combinatorial actionability, we found that PI3K/AKT/mTOR pathway alterations often co-occurred with biomarkers associated with sensitivity to agents targeting PARP (9%), NOTCH (6%), and PD-1/PD-L1 (5%). Other promising immunotherapy combinations for specific GI cancer types included MEK, ERK, Chk1, ATR, hedgehog and porcupine inhibitors. We present these analyses in the context of current oncology pipelines.

Conclusions

Use of a computational based analysis of multi-omic molecular tumor profiling data could transform drugdevelopment strategy, agent and drug pipeline prioritization in cancer trial design by providing a mathematically rationalized agent match based on real world molecular data. Before embarking on molecularly enriched single agent or combination clinical trials, prospective analyses should be performed to determine feasibility of accrual, potential clinical utility, and target-specific or disease-specific designs.

Clinical trial identification

Legal entity responsible for the study

Georgetown University Hospital, Lombardi Cancer Center.

Funding

Georgetown University Hospital, Lombardi Cancer Center.

Editorial Acknowledgement

Disclosure

E. Blais: Employee: Perthera, Inc. with stock equity holdings/options; Co-inventor: Perthera-assigned patents from and can receive royalties and licensing fees. M. Pishvain: Consultant and Chief Medical Officer: Perthera; Compensation and stock: Perthera. E. Petricoin: Co-founder: Perthera, Inc., Ceres Nanosciences, Inc.; Board of directors and stock equity holder of both companies; Consultant: Perthera, Inc., Ceres Nanosciences, AVDX Investors Group, LLC and AzGen Inc; Coinventor on Perthera-assigned and University-assigned and US Government assigned patents and can receive royalties and licensing fees from both. J.L. Marshall: Scientific advisory board of Perthera, Inc. and Caris Life Sciences. All other authors have declared no conflicts of interest.