Genetic Profiling in Soft Tissue Sarcoma

University dissertation from Lund University, Faculty of Medicine Doctoral Dissertation

Abstract: Soft tissue sarcomas (STS) are a heterogeneous group of highly malignant mesenchymal tumors that account for ~1% of all malignancies. Frequent heterogeneity and pleomorphism along with suboptimal diagnostic reproducibility and insufficient prognostic markers make clinical management of these tumors difficult. This thesis has applied microarray-based gene expression and copy-number profiling to STS. The studies provide clues to the genetic pathways involved in STS development and identify profiles linked to diagnosis and prognosis. The results from Study I that concerns intratumor versus intertumor heterogeneity of gene expression profiles in malignant fibrous histiocytoma (MFH) and leiomyosarcoma (LMS), suggest that intratumor heterogeneity may be particularly relevant in small tumor series and thus serve as a reminder to run larger sample sets for increased reliability. Study II established expression patterns related to the SS18-SSX fusion variants and metastatic potential in synovial sarcoma (SS). The differential expression of various developmental genes, transcription factors, histones, and metallothioneins suggests that the gene fusion variants have distinct downstream effects. In Study III, 177 STS of mixed histopathological subtypes were profiled using cDNA microarrays. Distinct gene expression patterns were identified in subtypes with specific translocations or mutations. Herein, frequent upregulation of developmental genes, from e.g. the Wingless and Hedgehog signaling pathways, was demonstrated. The more pleomorphic STS showed overexpression of genes involved in proliferation, adhesion, motility and protein degradation. Moreover, a prognostic signature partly characterized by hypoxia-related genes was identified within the pleomorphic STS. Study IV applied array-based comparative genomic hybridization in MFH and LMS, and demonstrated extensive genetic complexity with multiple recurrent gains and losses, novel amplifications and homozygous deletions. Losses in chromosomal regions 6q14 and 7q36 provided prognostic information independent of previously established risk factors. In summary, these studies demonstrate the potential of genetic profiling in STS and herein, define intratumor heterogeneity, demonstrate that gene fusion variants in SS yield different downstream effects, identify diagnostic and prognostic subsets within STS, and in the pleomorphic tumors, discern prognostically important alterations within the plethora of genetic aberrations that characterize many STS.