Investigating the various predictive values of POLE/POLD1 mutations for response to immune checkpoint inhibitors (ICIs) in different solid tumors.

Abstract

2606 Background: Both POLE and POLD1 encode the catalytic subunit of polymerase enzyme complexes involved in DNA replication and repair. The mutations of POLE and POLD1 have been shown to be oncogenic and lead to DNA repair defects and elevated tumor mutation burden (TMB). And patients with POLE/POLD1 mutations are more likely to benefit from ICIs therapy. Previous studies have shown that TMB has divergent predictive value for response to ICIs therapy in different cancer types. We hypothesized that the associations between POLE/POLD1 mutations and ICIs efficacy are also varied in different solid tumors. Therefore, we explored the prediction values of POLE/POLD1 mutations in some cancer types. Methods: The ICIs treatment cohort from Memorial Sloan Kettering Cancer Center (MSKCC) was selected to analyze the association of POLE/POLD1 mutations with ICIs efficacy. TCGA cohort was enrolled for characterizing tumor infiltrating lymphocytes (TILs) with CIBERSORT. The patients were classified into two groups: POLE/POLD1 mutations (Mut) and wildtype (WT). Overall survival (OS) after ICIs therapy was estimated with Kaplan-Meier method. Results: In MSKCC pan-cancer dataset, patients with POLE/ POLD1 mutations had significantly longer median OS after ICIs therapy (34.00 vs 19.00 months, P = 0.0143), indicating that POLE/POLD1 mutations were associated with better immunotherapy outcomes. Then we analyzed the predictive roles in each cancer type. Notably, we found the associations of POLE/POLD1 mutations with longer median OS in NSCLC (Undefined vs 12.00 months, P = 0.05) and esophagogastric cancer (27.00 vs 15.00 months, P = 0.05). However, the associations between POLE/POLD1 mutations and ICIs efficacy were not observed in bladder cancer, melanoma, glioma, head and neck cancer, renal cell carcinoma, and colorectal cancer. Furthermore, our data showed that the median TMB was significantly higher in the Mut group for NSCLC (20.2 vs 6.9 muts/Mb, P < 0.0001) and esophagogastric cancer (21.4 vs 5.6 muts/Mb, P < 0.0001). In TCGA esophagogastric cancer cohort, POLE/POLD1 mutations were correlated with decreased naive B cells (P = 0.0306) and increased activated memory CD4+ T cells (P = 0.0224). In TCGA NSCLC cohort, POLE/POLD1 mutations were correlated with elevated gamma delta T cells (P = 0.0219). These data suggested that POLE/POLD1 mutations were also involved in the infiltration of some immune cells. Conclusions: Although POLE/POLD1 mutations were associated with better ICIs efficacy for all the enrolled patients, the prolonged OS was only found in the Mut group for NSCLC and esophagogastric. These data suggested that POLE/POLD1 mutations may be a useful predictor for ICIs efficacy in these two types of cancer. Moreover, POLE/POLD1 mutations were correlated with the level of TILs in NSCLC and esophagogastric. The finding was consistent with the efficacy of immunotherapy.