Synovial sarcoma misdiagnosis and potential utility of NGS testing patient identification for autologous T cell therapy: Retrospective analysis of Tempus database.

Abstract

11543 Background: Metastatic synovial sarcoma (SS) is commonly treated with anthracycline and other chemotherapy with poor survival outcomes. Afamitresgene autoleucel (Tecelra) is a genetically modified autologous T cell therapy that was FDA approved in August 2024, based on ORR of 43.2% and median DoR of 6 months. It is indicated for previously treated HLA-A*02:01-03 and 06 allele positive patients (pts) with unresectable or metastatic SS whose tumors express MAGE-A4. SS is characterized by SS18-SSX gene fusion in > 95% of cases and this gene fusion is not found in other neoplasms. Commercial NGS testing may contribute to identification of eligible pts. Methods: Pts with diagnosis of SS and/or SS18 gene fusion were identified by retrospective analysis of the Tempus database. Demographic and disease characteristics were described for each pt and assessed for correlation with diagnosis and presence of the SS18 gene fusion. RNA was sequenced with the Tempus RS and RS.v2 assays. Gene expression values in this analysis were normalized by transcripts-per-million (TPM) and transformed to log2(TPM+1) units. The threshold of 1.62 (log2(TPM+1)) was set as the cut off for determining true positive of MAGEA4 expression. HLA type and RNA expression of MAGE-A4 was used to identify pts potentially eligible for associated T cell therapies. Results: 110 pts were analyzed, including 86 pts (78%) with SS diagnosis with SS18 fusion, 12 pts (11%) with SS diagnosis but no detected SS18 fusion (SS/SS18-), and 12 pts (11%) with SS18 fusion but a non-SS diagnosis. Age at diagnosis, sex, race, state of diagnosis, and institution type (academic vs non-academic) were not associated with misalignment between the reported diagnosis and presence or absence of the SS18 fusion. For SS/SS18- pts the location of the tumor did not differ from expected SS primary locations, and the Tempus tumor of origin predicted SS diagnosis in 2/12 pts, non-SS diagnosis in 7/12 pts, and not reported 3/12 pts. The most common predicted non-SS diagnosis was malignant peripheral nerve sheath tumor (MPNST, 5 pts). Non-SS/SS18+ pts did occur in unusual locations (Pancreas, small intestine, bone, peripheral nervous system), and 10/12 were predicted by the tempus tumor of origin algorithm to be SS. The recorded diagnosis in these 12 pts was more commonly another sarcoma subtype (10/12), rather than a non-sarcoma diagnosis (2/12). MAGE-A4 expression > 1.62 TPM was detected in 71% of all pts, 74% SS/SS18+ pt, 58% of SS/SS18- and 58% of non-SS/SS18 pts. When considering the required HLA type and MAGEA4 expression then 35% of all pts would be potentially eligible for afami-cel. Conclusions: Commercial NGS testing may increase identification of pts with SS, which is important for novel T cell therapy eligibility. A low threshold should be considered for diagnosing SS in unusual locations when the SS18 fusion is detected.