Next generation sequencing reveals targetable mutations in multiple sarcoma histologies.

Abstract

11548 Background: Sarcomas are a rare and heterogenous group of cancers that arise from bone or soft tissue, and two thirds have poorly defined mutational profiles. Given their rarity, few comprehensive studies have fully characterized mutations and gene expression across sarcoma histologies correlated with clinical outcomes. Further studies are needed to determine the presence of targetable mutations that may improve patient outcomes. In this study, we explored the genomic landscape and clinical actionability of sarcoma mutations from patients enrolled in our CAUSAL (Cohort to Augment the Understanding of Sarcoma survivorship Across the Lifespan) study. Methods: Between 04/01/2022 and 01/01/2023, 481 participants, treated from 2012 – present with multiple sarcoma histologies were enrolled on CAUSAL. Next Generation Sequencing (NGS) was performed on primary or metastatic tumors from 76 patients to determine DNA mutations within a 648 gene panel. Whole transcriptome RNA sequencing (seq) provided expression profiles and RNA fusion products. Further analysis was performed using principal components analysis of RNA seq data to explore correlates amongst sarcoma subtypes. Tumor mutations were queried in ClinVar for relevance to known variants and were assigned to tiers I-IV based on the ESMO scale for clinical actionability of molecular targets (ESCAT). Tier I mutations have drug-mutation matched evidence of actionability while tier IV have only pre-clinical evidence. Results: NGS has been completed on 76 tumor samples. Sequenced tumors represented 19 histologies, with the most common ones as follows: undifferentiated pleomorphic sarcoma (15.8%), liposarcoma (9.2%), gastrointestinal stromal tumor (9.2%), osteosarcoma (7.9%), and leiomyosarcoma (7.9%). Of 76 patients, 66 (87%) had at least one mutation detected with an mean frequency of 2.74. TP53 (20/66), RB1 (14/66), and ATRX (9/66) were most commonly mutated genes. Mean (std) tumor mutation burden (TMB) was 3.4 m/MB (3.5m/MB) and one tumor had TMB of 25.3 m/MB. Of the 76 samples, 68 had RNA expression and fusion data available, of which 42 (62%) had anomalous expression changes and 11 had RNA fusions. The most overexpressed genes were NY-ESO-1 (13), LAGE-1 (9), and RET (7); the most under-expressed genes were SMARCB1 (9) and MGMT (6). 30.2% (23 of 76) of patients had potentially actionable DNA mutations, 9 had ESCAT tier I DNA mutations, 3 had tier II, 10 had tier III, and 1 had tier IV. 29.4% (20/68) of patients had potential targets based on RNA expression. 51.3% (39/76) patients had either a potential DNA or RNA target, and 6.6% (5/76) had multiple RNA or DNA targets. Conclusions: NGS revealed potentially actionable targets in over half of sarcoma patients based on ESCAT criteria. With ongoing accrual and sequencing of additional tumor specimens future analysis of CAUSAL will focus on assessing the correlation between targetable mutations and clinical outcomes.