MA23.11 Analysis of Immune Phenotype Composition in Malignant Pleural Mesothelioma (MPM) Using Bulk RNA Sequencing

Exploiting the immune status of the tumour microenvironment (TME) is increasingly being adopted for many cancer types. Investigation into immune phenotype composition of the TME is at present lacking for malignant pleural mesothelioma (MPM) but critically important in light of the cancer’s overall poor prognosis and lack of targeted therapy as clinical standard of care. In this study, CD8+ve tumour infiltrating lymphocyte (TIL) level has been used as a starting point to compare differences in mutational patterns, histology and survival in MPM. Bulk RNA sequencing of tumour tissue from 35 MPM patients (in-house cohort) was performed. Sequencing read alignment and gene count estimation were performed using STAR (v.2.5.2b). To increase the sample size, raw data from Bueno et al. (n=211 subjects) was accessed and gene count estimations performed. In addition, the TCGA-MESO cohort (n=86 subjects) count data was included from the GDC (Genomic Data Commons) website. All count data were normalized cohort-wise using the ‘voom’ method implemented in limma package. Deconvolution of constituent immune phenotypes in the TME from the bulk RNA-sequencing data was performed by applying CIBERSORT (v.1.04) on normalized count data sets. For assessing the genetic context of observed immune phenotypes, somatic mutations were profiled using targeted sequencing of a custom gene panel for the in-house cohort. For the Bueno et al. and the TCGA-MESO cohorts, somatic mutations were either available from an overlap of whole-exome sequencing (WES) and targeted gene panel, or from WES only. A total of 27 samples (3 of 35 (8.6%), 21 of 211 (9.9%) and 3 of 86 (3.5%) from the in-house, Bueno et al. and TCGA-MESO cohorts respectively) were identified with immune phenotype enriched for CD8+ve TIL. Histological subtype distribution in the CD8+ve enriched samples was seen to be almost equivalently split between Epithelioid and Biphasic subtypes (51.85% and 48.15% respectively). Interestingly, BAP1 mutation was found to be present in only 7.7% of the samples. Considering in addition the genes NF2, SETD2, SETD6, SETDB1, TP53 and LATS1/2, mutations were only found to be present in 57.7% of the samples in total. As such >40% of samples with CD8+ve TIL do not have any mutations detected in known hotspot genes for MPM. Histological subtype is not significantly different between these ‘wild-type’ and hotspot gene(s) mutated samples. Median survival for the groups was found to be 1.85 and 0.73 years respectively. In the present study, approximately 3-10% of MPM samples were found to have enrichment for CD8+ve TIL. Nonetheless on closer examination of the genetic context, mutation patterns emerge that warrant further investigation. For samples that have TP53 (n=3) mutation or mutations in multiple hotspot genes (BAP1, NF2, SETD2, LATS2; n=1), survival understandably is lowest (0.27 years average). This raises a number of further questions including what sustains a tumour despite high CD8+ve TIL population? And more importantly with lack of tumour mutational burden what other TME signals draw effector immune cells? Further investigations, by comparing additional immune markers with copy number changes that might be present in hotspot genes, are therefore required.