(P 231) IDENTIFICATION OF NEOANTIGEN FROM PECOMA-LIKE NEOPLASM PROVIDES RATIONALE FOR TREATMENT WITH CELLULAR THERAPY

Objective: Fusion-driven sarcomas are rare soft tissue neoplasms that occur primarily in young patients. In general, they are genotypically simple, have a low tumor mutational burden (TMB) relative to other sarcoma subtypes, and are almost completely insensitive to immune checkpoint blockade (ICB). An important exception is alveolar soft part sarcoma (ASPS), which are characterized by the presence of ASPSCR1::TFE3 fusions. While ASPS do have a low TMB, they are exceptionally responsive to ICB for reasons that are not currently understood.

TFE3 is a master transcription factor involved in cell proliferation and angiogenesis and is implicated in a variety of cancer types. In addition to ASPS, TFE3 fusions are found in a subset of clear cell renal cell carcinoma (ccRCC) and in rare PEComa-like neoplasms. Given the responsiveness of ASPS to ICB, we hypothesized that the ASPSCR1::TFE3 fusion functions as a particularly immunogenic neoantigen. To test this, we performed ex vivo culture of tumor-infiltrating lymphocytes (TIL) from a patient with a PEComa-like neoplasm with a type 2 ASPSCR1::TFE3 fusion. After TIL growth, we tested whether they could recognize the fusion junction region and other potentially immunogenic epitopes.

Methods: TIL from a treatment naïve 49-year-old patient with a PEComa-like neoplasm were cultured from a resected liver metastasis using previously described methods. Individual TIL cultures were screened against peptides and tandem minigenes (TMGs) representing nonsynonymous mutations identified from whole exome sequencing and relevant tissue differentiation antigens (TDAs) gp100 and MELAN-A. In addition to whole exome sequencing, we used transcriptomic data to identify epitopes from gene fusion junction regions. An autologous EBV transformed B-cell line (EBV-B) was established for use as antigen presenting cells (APCs). APCs were electroporated with TMGs or pulsed with peptides and then co-
cultured overnight with TIL fragment cultures. Reactive T-cells were identified by interferon gamma production (ELISPOT or electrochemiluminescence). In the cases of T-cell recognition, tumor specificity was tested by comparing reactivity between mutant and wild type peptide titrations.

Results: The PEComa-like neoplasm was found to harbor only 24 expressed nonsynonymous point and frameshift mutations, where a typical epithelial carcinoma might harbor 100-200). Moreover, we identified 5 gene fusions including the expected Type 2 ASPSCR1::TFE3 fusion (Table 1). Between nonsynonymous mutations, fusions and TDAs we created a screening library of peptides and TMGs encompassing 42 epitopes. Eleven (92%) of 12 individual TIL cultures were successfully grown from a sample of liver metastasis. Strikingly, 10 (91%) of 11 TIL cultures were found to contain reactivity to a peptide representing INPP4A p.S433L (Figure 1 A and B). As a point of comparison, in epithelial carcinomas reactivity to a single neoantigen is often restricted to only 1 or 2 TIL cultures (4-8% of the total). To test the specificity of this recognition and due to the low background, TIL fragment culture 1 and 3 were selected for comparison of mutant and wild-type peptide titrations and demonstrated specificity for the
mutant peptide (Figure 1C and D).

Conclusion: Despite a low tumor mutational burden, TIL isolated from a PEComa-like neoplasm contained a strong reactivity to a private neoantigen, demonstrating that PEComa-like neoplasms can be immunogenic and serving as evidence that supports treatment with immunotherapy. The lack of reactivity to the recurrent ASPSCR1-TFE3 fusion or established TDAs limits the generalizability of this finding to other patients.