(P 189) PROGNOSTIC IMPACT OF NEW YORK ESOPHAGEAL SQUAMOUS CELL CARCINOMA-1 EXPRESSION AND HUMAN LEUKOCYTE ANTIGEN SUBTYPES IN METASTATIC SYNOVIAL SARCOMA

Objective: Synovial sarcoma (SS) is a rare mesenchymal tumor characterized by a specific chromosomal translocation that results in the formation of an oncogenic SS18::SSX1/2/4 fusion gene. In the metastatic setting, therapeutic options are limited to chemotherapy and anti-angiogenic agents, and prognosis remains poor. The New York Esophageal Squamous Cell Carcinoma-1 (NY-ESO-1) is a cancer-testis antigen aberrantly expressed in various cancers and commonly in SS. Letetresgene autoleucel (lete-cel), an adoptive T-cell receptor therapy, is currently being investigated in clinical trials in previously treated metastatic SS (mSS) patients (pts) with NY-ESO-1 positive tumors who express at least one of the human leukocyte antigen (HLA) HLA-A*02-01/05/06 (HLA*A) alleles, with encouraging preliminary results. The impact of NY-ESO-1 expression and HLA type on the prognosis of mSS pts is not known. This study aims to assess the impact of NY-ESO-1 expression and HLA subtype, independently and together, on overall survival (OS) in mSS.

Methods: This study is a retrospective clinico-biological cohort of patients with mSS identified within the Netsarc+ database of the French Sarcoma Group. Included pts were adults with histologically confirmed diagnosis of mSS from 2000 to 2020, and availability of a representative formalin-fixed paraffin-embedded archival tumor sample from the primary tumor. Each case was reviewed by an expert pathologist of the Netsarc+ network to ensure quality/quantity control criteria (at least 20% of tumor cells, sample size surface area >5 mm2, and depth >90 µm). NY-ESO-1 status was determined by immunohistochemistry, positive NY-ESO-1 expression was defined as ≥30% of cells that are 2+ or 3+. HLA*A subtypes were determined by RNA seq using OptiType. Survival distributions were estimated by Kaplan–Meier method and compared between biomarker expression groups using the log-rank test. The prognostic value of each covariate on OS was evaluated using univariate analyses; all significant parameters were included in the multivariate model. P-values are presented for descriptive purposes.

Results: One hundred and nine mSS pts met study criteria; NY-ESO-1 was expressed in 65/106 cases (61.3%) and HLA*A in 41/94 cases (44%); 23/91 pts had an NY-ESO-1+/HLA*A status (25%). At the time of the data analysis, 88 pts were deceased and the median follow-up was 69 months.

Of the 106 pts with NY-ESO-1 expression available, 59 were men and 49 were < 45 years old at the time of metastases. The primary tumor was ≥5 cm in 84.9% of cases, was located in limbs in 60.4% of pts, and presented as poorly differentiated in 9.4% of pts. At diagnosis, 22.6% of pts were metastatic. The metastasis-free interval was < 1 year for 38.3% of localized pts. Metastatic site involved only one organ in 86.8% of pts. The lung was the only site of metastasis for 70.7% of pts. In the metastatic setting, 73.6% of pts received at least two lines of systemic treatment.

In an unadjusted analysis, OS was observed to be significantly longer in pts whose tumors expressed NY-ESO-1 (Figure 1). Prognostic impact of NY-ESO-1 expression lost its significance after adjustment by histology, site of primary tumor, metastatic-free interval, and location of first metastases (Table 1).

OS was longer in patients with HLA*A type compared to others (Figure 1). This difference was not statistically significant.

When considering the dual positive patients with NY-ESO-1+/HLA*A+ status, the median OS was 36.5 months versus 26.4 months (Figure 2) for the rest of the cohort; this difference was not statistically significant. When considering the 2L+ dual positive population who would be eligible to receive lete-cel, the median OS difference was also not statistically significant (12.9 vs 10.2 months).

Conclusion: This is the first study to examine the prognostic impact of both NY-ESO-1 and HLA*A type in mSS patients, and the first prognostic study in this population using real-world data and archival tissue. Assessing efficacy of adoptive T-cell therapies in highly selected patients, versus standard treatment through randomized trials, is difficult. In this context, it is essential to ensure that expression of NY-ESO-1 and HLA*A type does not modify the natural history of the disease.

Unadjusted results report a longer OS in patients whose tumors express NY-ESO-1, even if the prognostic impact of this antigen disappears in the multivariate analysis. Lack of significance may be related to the small size of the cohort. There were no statistically significant differences in survival outcomes based on HLA*A type nor on the combination of HLA*A type and NY-ESO-1 expression, including in the 2L setting.