844 Background: Homologous recombination deficiency (HRD) scores quantify genomic instability based on chromosomal alterations such as loss of heterozygosity, telomeric allelic imbalance, and large scale transitions. HRD is emerging as a potential biomarker in gastrointestinal (GI) cancers, correlating with sensitivity to DNA damaging agents, immune response modulation, and clinical outcomes. ARID1A, a chromatin remodeling gene frequently altered in GI malignancies, may play a prognostic role, although its interaction with HRD status remains unclear. Methods: Using The Cancer Genome Atlas (TCGA), we analyzed 7 individual GI cancers: colon (n=271), rectum (n=84), stomach (n=396), pancreatic (n=154), hepatocellular carcinoma (n=343), cholangiocarcinoma (n=36), and esophageal carcinoma (n=161). HRD scores were dichotomized at the median into HRD-high and HRD-low groups. ARID1A expression was similarly dichotomized and analyzed within each cancer type, adjusting for HRD status using multivariable Cox proportional hazards models. Hazard ratios (HRs) and 95% confidence intervals (CIs) were computed for overall survival. Results: In hepatocellular carcinoma, HRD high status was associated with significantly worse overall survival (HR 4.26; 95% CI 1.04–17.3; p=0.043). In contrast, pancreatic adenocarcinoma showed improved survival with HRD-low status (HR 0.61; 95% CI 0.38–0.97; p=0.039). Other tumor types, including colon, rectum, stomach, cholangiocarcinoma, and esophageal carcinoma, showed no significant differences in survival based on HRD status. When adjusting for HRD, ARID1A expression level was not significantly associated with overall survival in any GI cancer subtype. Hazard ratios ranged from 0.72 in colon adenocarcinoma to 1.58 in cholangiocarcinoma, but none reached statistical significance. In hepatocellular carcinoma, ARID1A-high expression was nearly absent, precluding meaningful survival analysis. Conclusions: HRD status exhibits tumor specific prognostic associations in GI cancers. HRD-high status appears deleterious in hepatocellular carcinoma, while HRD-low status may be protective in pancreatic cancer. ARID1A expression, however, does not independently impact survival when HRD is accounted for. These findings support further exploration of HRD as a prognostic and potentially predictive biomarker in GI malignancies and suggest limited utility for ARID1A expression alone in prognostication.
Jain et al. (Sat,) studied this question.