The differential diagnoses of uterine leiomyomas and leiomyosarcomas using DNA and RNA sequencing

Abstract

Although uterine leiomyomas and leiomyosarcomas are considered biologically unrelated tumors, they share morphologic and histologic characteristics that complicate their differential diagnosis. The long-term therapeutic option for leiomyoma is laparoscopic myomectomy with morcellation, particularly for patients who wish to preserve their fertility. However, because of the potential dissemination of undiagnosed or hidden leiomyosarcoma from morcellation, there is a need to develop a preoperative assessment of malignancy risk. Through an integrated comparative genomic and transcriptomic analysis, we aim to identify differential genetic targets in leiomyomas vs leiomyosarcomas using next-generation sequencing as the first step toward preoperative differential diagnosis. Targeted sequencing of DNA and RNA coding regions for solid tumor-associated genes was performed on formalin-fixed paraffin-embedded samples from 13 leiomyomas and 13 leiomyosarcoma cases. DNA sequencing was used to identify copy number variations, single-nucleotide variants, and small insertions/deletions. RNA sequencing was used to identify gene fusions, splice variants, and/or differential gene expression profiles. In leiomyosarcomas, tumor mutation burden was higher in terms of copy number variations, single nucleotide variants, small insertions/deletions, and gene fusions compared with leiomyomas. For copy number variations, 20 genes were affected by deletions in leiomyosarcomas, compared with 6 observed losses in leiomyomas. Gains (duplications) were identified in 19 genes in leiomyosarcomas, but only 3 genes in leiomyomas. The most common mutations (single-nucleotide variants and insertions/deletions) for leiomyosarcomas were identified in 105 genes of all analyzed leiomyosarcomas; 82 genes were affected in leiomyomas. Of note, 1 tumor previously diagnosed as leiomyosarcoma was established as inflammatory myofibroblastic tumor along this study with a novel ALK-TNS1 fusion. Finally, a differential transcriptomic profile was observed for 11 of 55 genes analyzed in leiomyosarcomas; 8.5% of initially diagnosed leiomyosarcomas showed high-confidence, novel gene fusions that were associated with these tumors. Through integrated comparative genomic and transcriptomic analyses, we identified novel differential genetic targets that potentially differentiate leiomyosarcomas and leiomyomas. This provides a new insight into the differential diagnosis of these myometrial tumors.