297 Modeling the efficacy of NY-ESO-1 TCR T cells (letetresgene autoleucel; GSK3377794) in patients with synovial sarcoma: correlations of response with transduced cell kinetics and biomarkers

Background NY-ESO-1–specific T cells (letetresgene autoleucel [lete-cel]; GSK3377794) are autologous CD4+ and CD8+ T cells transduced to express a high-affinity T-cell receptor that recognizes NY-ESO-1 antigen in complex with HLA-A*02. NY-ESO-1 is a cancer testis antigen that is expressed in many cancers, including synovial sarcoma (SS). Study 208466 (NCT01343043) is a Phase I clinical trial that assessed the safety and efficacy of lete-cel in patients with advanced SS (presented in complementary abstract). This abstract presents correlations of transduced cell kinetics and biomarkers with response. Methods Patients with unresectable, metastatic, or recurrent SS were enrolled to 4 cohorts based on NY-ESO-1 expression levels and received different lymphodepleting regimens (LDR) prior to lete-cel infusion (N=45) (table 1). Response was assessed per RECIST v1.1. Transduced cell kinetics (persistence) were measured by quantitative PCR of transgene vector copies in DNA extracted from peripheral blood mononuclear cells. Serum cytokines were measured by Meso Scale Discovery (MSD) immunoassay. Gene expression within tumor biopsies was evaluated by Nanostring. Post hoc analyses were evaluated in a hypothesis-driven manner using logistic and linear regression. Potential determinants of peak persistence and clinical response were tested using generalized linear models. Results Higher peak persistence (Pmax) was associated (p=0.012) with response across cohorts. Higher weight-normalized cell dose (p=0.00421) and LDR (p=0.000910) were associated with Pmax according to the generalized linear model: Pmax ~ cell dose + LDR. These relationships allowed for accurate retrospective prediction of probability of response. Low LDR resulted in higher endogenous lymphocyte counts on the day of dosing, which trended with lack of response within and across cohorts. While the impact of fludarabine on IL-15 levels has been previously reported, data presented here show a novel, positive correlation between IL-15 levels pre-infusion and response (p=0.0332). Post lete-cel infusion, the concentrations of IFNγ, IL-6, and IL-2RA within the first week were increased in responders vs non-responders. The peak expression of IL-2RA within the first week showed a linear correlation to Pmax. Analysis of tumor biopsies showed good correlation between NY-ESO-1 mRNA and protein expression. Conclusions Exposure–response analysis of study 208466 reveals that efficacy appears to be driven by weight-normalized cell dose and LDR via Pmax. Biomarker correlation analysis indicates that LDR impacts the level of IL-15 pre-infusion, which correlates with response directly. IFNγ, IL-6, and IL-2RA levels appear to be promising pharmacodynamic markers. Optimizing dose and LDR may offer opportunities to maximize antitumor efficacy. Acknowledgements This study (208466) was funded by GlaxoSmithKline (GSK). Trial Registration Clinicaltrials. gov NCT01343043 Ethics Approval This study was approved by the appropriate institutional review boards and independent ethics committees.