Comprehensive Genomic Profiling Alters Clinical Diagnoses in a Significant Fraction of Tumors Suspicious of Sarcoma.

Öfverholm I, Wallander K, Haglund C, Chellappa V, Wejde J, Gellerbring A, Wirta V, Renevey A, Caceres E, Tsagkozis P, Mayrhofer M, Papakonstantinou A, Linder-Stragliotto C, Bränström R, Larsson O, Lindberg J, Lin Y, Haglund de Flon F

Clin. Cancer Res. 30 (12) 2647-2658 [2024-06-14; online 2024-04-04]

Tumor classification is a key component in personalized cancer care. For soft-tissue and bone tumors, this classification is currently based primarily on morphology assessment and IHC staining. However, these standard-of-care methods can pose challenges for pathologists. We therefore assessed how whole-genome and whole-transcriptome sequencing (WGTS) impacted tumor classification and clinical management when interpreted together with histomorphology. We prospectively evaluated WGTS in routine diagnostics of 200 soft-tissue and bone tumors suspicious for malignancy, including DNA and RNA isolation from the tumor, and DNA isolation from a peripheral blood sample or any non-tumor tissue. On the basis of specific genomic alterations or absence of presumed findings, WGTS resulted in reclassification of 7% (13/197) of the histopathologic diagnoses. Four cases were downgraded from low-grade sarcomas to benign lesions, and two cases were reclassified as metastatic malignant melanomas. Fusion genes associated with specific tumor entities were found in 30 samples. For malignant soft-tissue and bone tumors, we identified treatment relevant variants in 15% of cases. Germline pathogenic variants associated with a hereditary cancer syndrome were found in 22 participants (11%). WGTS provides an important dimension of data that aids in the classification of soft-tissue and bone tumors, correcting a significant fraction of clinical diagnoses, and identifies molecular targets relevant for precision medicine. However, genetic findings need to be evaluated in their morphopathologic context, just as germline findings need to be evaluated in the context of patient phenotype and family history.

Bioinformatics Support and Infrastructure [Collaborative]

Bioinformatics Support, Infrastructure and Training [Collaborative]

Clinical Genomics Stockholm [Service]

PubMed 38573684

DOI 10.1158/1078-0432.CCR-24-0384

Crossref 10.1158/1078-0432.CCR-24-0384

pii: 742915