Pan-cancer analysis of PBRM1 mutation and their association with immune-related biomarkers and prognosis.

Abstract

e14536 Background: Polybromo 1 ( PBRM1), which is required for the stability of the SWI/SNF chromatin remodeling complex, is often deleted or mutated in clear cell renal cell carcinoma (ccRCC). However, the prognostic value of PBRM1 in ccRCC with immunotherapy was controversial and less explored in other cancer types. Methods: The sequencing data and clinical information of patients were obtained from The Cancer Genome Atlas (TCGA) cohort (n = 10967) and the immunotherapy MSKCC cohort (n = 387). In TCGA cohort, the expression levels of PD1 were defined as high or low using a 75% cut-off. The association of PBRM1 mutation and related immune biomarkers such as PD1, TMB and immune infiltrates CD4+ and CD8+ T cells was analyzed. Results: Loss-of-function (LOF) of PBRM1 mutation ( PBRM1LOF), including frameshift, nonsense and splice, accounted for 61.4% of all somatic mutations, with a frequency of 2.2% across all cancer types. It is most common in renal clear cell carcinoma (KIRC; 23.2%, 119/512), followed by stomach adenocarcinoma (STAD; 4.1%, 18/440), uterine corpus endometrial carcinoma (UCEC; 3.8%, 20/529), and colon adenocarcinoma (COAD; 2.5%, 11/439) in TCGA cohort. In KIRC, no differences of PD1 expression (P = 0.307) and TMB (P = 0.389) were observed between patients with PBRM1LOF and non- PBRM1LOF mutations, while in STAD, UCEC and COAD, patients with PBRM1LOF mutations had significantly higher levels of PD1 (P = 0.002 in COAD, P < 0.001 in STAD and UCEC) and TMB (all P < 0.001). As for immune infiltrates, expression levels of CD4+ (P = 0.34) and CD8+ (P = 0.18) were found no difference between patients with PBRM1LOF and non- PBRM1LOF in KIRC. However, significantly higher expression levels of CD4+ (P = 0.0014) and CD8+ (P = 0.0016) in UCEC, and CD8+ in COAD (P = 0.04) were found in patients with PBRM1LOF compared with non- PBRM1LOF mutations. Further, in MSKCC cohort, a cohort of patients who had undergone immunotherapy, PBRM1LOF was shown no significant effect on patient outcome in KIRC (P = 0.42), while in other cancer types (COAD+STAD), PBRM1LOF trended towards a promising biomarker for better survival even though the difference was not statistically significant in this small cohort (P = 0.51). Conclusions: PBRM1LOF may play different prognostic roles on immunotherapy in different cancer types. In KIRC, PBRM1LOF was not associated with immune-related biomarkers as well as prognosis of immunotherapy. In COAD, STAD and UCEC, PBRM1LOF was positively associated with immune-related biomarkers and might be a promising biomarker for immunotherapy. This result needs further validation in a lager cohort.