Characterization of the tumor immune microenvironment (TIME) and somatic landscape in gastrointestinal (GI) malignancies with <i>MTAP</i> deletions (del).

Abstract

835 Background: PRMT5 is a synthetic lethality target in patients (pts) with MTAP del and early phase trials are underway with PRMT5 inhibitors. Additionally, MTAP del are associated with a less immunogenic TIME and reduced efficacy of immunotherapy, but research has primarily not been focused on GI malignancies. Thus, we investigated the TIME and somatic landscape in GI malignancies with MTAP del. Methods: From the Tempus Database, we retrospectively analyzed de-identified next-generation sequencing data from pts across GI malignancies, including pancreatic (n=11,217), gastroesophageal (GEJ, n=5,803), cholangiocarcinoma (CCA, n=3,244), and colorectal (CRC, n=17,537) cancers. Tumors were sequenced with the Tempus xT DNA (648-gene panel) and xR RNA assays. MTAP del were defined as two-copy losses. Somatic alterations (alt), immune cell infiltration predicted from gene expression patterns, PD-L1 from IHC, TMB, and MSI were evaluated. Chi-squared/Fisher’s Exact tests or Kruskal-Wallis tests were used to assess statistical significance (p&lt;0.05, q&lt;0.05 for false discovery rate correction for multiple testing). Results: MTAP del were identified in 14.8%, 11.3%, 0.9%, and 7.3% of pancreatic, CCA, CRC, and GEJ cancers, respectively. Of these, 98/93%, 99/91%, 92/89%, and 98/89% had co-occurring CDKN2A/B loss. In pancreatic, CCA, and CRC pts, MTAP del was associated with a reduced proportion of B cells and CD4 T cells, and there were higher percentages of macrophages vs pts with MTAP WT status (p&lt;0.001 for all). Reductions in proportion of CD8 T cells were also associated with MTAP del in pancreatic and CCA pts (p&lt;0.001 for both). Lower TMB (p=0.036) and MSI-H status (p=0.009) were found in CRC pts with MTAP del. GEJ with MTAP del also exhibited significantly lower MSI-H status than MTAP WT GEJ (p=0.014). SMAD4 alterations , a marker of reduced immune infiltrates, were more prevalent in pts with MTAP del across GI malignancies (q&lt;0.005). In the CCA cohort, there was a higher percentage of BRAF alt and FGFR2 fusions in pts with MTAP loss (q&lt;0.001, q=0.028), while KRAS alt were higher in pancreatic cases with MTAP loss (q&lt;0.001). Conclusions: This is the largest analysis of the TIME and somatic landscape of MTAP loss across GI malignancies. In pts with MTAP del and pancreatic cancer, CCA, and CRC, we observed a less immunogenic TIME pattern, indicating the evaluation of immunotherapy implications in these GI malignancies with MTAP del is warranted. Our findings are hypothesis-generating, providing further rationale to study synthetic lethality and novel combinatorial therapeutic strategies in GI malignancies with MTAP del.