Platinum-based therapy of patients with gastrointestinal cancers that harbor homologous recombination DNA damage response (HR-DDR) mutations: A single institutional retrospective analysis.

Abstract

836 Background: HR-DDR deficiencies have been described across cancer types and associated with response to specific therapies, including platinum (Pt)-based drugs and poly ADP-ribose polymerase (PARP) inhibitors. We sought to estimate the frequency of HR-DDR gene mutations in gastrointestinal (GI) cancers and determine if these patients (pts) benefit from Pt-based therapy. Methods: A retrospective chart review was performed on 421 pts with GI cancers treated at our institution. Tumors were categorized as colorectal (CRC), gastroesophageal (GE), pancreatic (PDA), hepatobiliary (HB), or other. All had next-generation sequencing with mutations in one or more of the following HR-DDR genes: BRCA1, BRCA2, PALB2, CHEK2, ATM, ARID1A, MRE11A, NBN, BARD1, and BL M. Overall survival (OS) of pts was analyzed using Log-rank test. Results: Of those analyzed, 127 pts (84 women and 43 men) met the inclusion criteria. Forty-eight percent had primary CRC tumors, and the remainder had primary GE (17%), PDA (14%), HB (10%), and other GI (11%) tumors. Fifty-four percent had stage IV disease. Sixty-five percent received Pt-based therapy at some point during their disease course. HR-DDR analysis revealed the following mutation rates: 34% ARID1A; 16% ATM; 19% BRCA2; and 9% BRCA1. OS analysis showed no significant survival advantage based on stage or mutation type for any of the disease-specific cohorts. There was no significant OS benefit in pts who received Pt-based therapy vs. those who did not, although pts with ARID1A-mutated tumors did show a trend towards improved OS (P = 0.09). Conclusions: Despite our limited cohort, our data suggest that pts with GI tumors harboring certain HR-DDR gene mutations might benefit from Pt-based therapy, supporting the development of mutation-specific Pt-based clinical trials.