Comprehensive mutational profiling identifies new driver events in cutaneous leiomyosarcoma

Abstract Background Cutaneous leiomyosarcoma (cLMS) is a rare soft tissue neoplasm, showing smooth muscle differentiation, that arises from the mesenchymal cells of the dermis. To-date, genetic investigation of these tumours has involved studies with small sample sizes and limited analyses that identified recurrent somatic mutations in RB1 and TP53, copy number gain of MYCOD and IGF1R, and copy number loss of PTEN. Objectives To better understand the molecular pathogenesis of cLMS, we comprehensively explored the mutational landscape of these rare tumours to identify candidate driver events. Methods In this retrospective, multi-institutional study, we performed whole-exome sequencing and RNA sequencing on 38 cases of cLMS. Results TP53 and RB1 were identified as significantly mutated, thus, represent validated driver genes of cLMS. COSMIC mutational signatures SBS7a/b and DBS1 were recurrent, thus, ultraviolet light exposure may be an aetiological factor driving cLMS. Analysis of significantly recurrent somatic copy number alterations, which represent candidate driver events, found focal (<10Mb) deletions encompassing TP53 and KDM6B, and amplifications encompassing ZMYM2, MYOCD, MAP2K4 and NCOR1. A larger (24 Mb) recurrent deletion encompassing CYLD was also identified as significant. Significantly recurrent broad copy number alterations, involving at least half of a chromosome arm, included deletions of 6p/q, 10p/q, 11q, 12q, 13q and 16p/q, and amplification of 15q. Notably PTEN is located on 10q, RB1 on 13q and IGFR1 on 15q. Fusion gene analysis identified recurrent CRTC1/3::MAML2 fusions, as well as many novel fusions in individual samples. Conclusions Our analysis of the largest number of cLMS cases to-date highlights the importance of large cohort sizes and the exploration beyond small targeted gene panels when performing molecular analyses, as it allowed a comprehensive exploration of the mutational landscape of these tumours and identification of novel candidate driver events. It also uniquely afforded the opportunity to compare the molecular phenotype of cLMS with LMS of other tissue types, such as uterine and soft tissue LMS. Given that molecular profiling has resulted in the development of novel targeted treatment approaches for uterine and soft tissue LMS, our study now allows the same opportunities to become available for patients with cLMS.