Antitumor Activity Associated with Prolonged Persistence of Adoptively Transferred NY-ESO-1 c259T Cells in Synovial Sarcoma

Abstract

We evaluated the safety and activity of autologous T cells expressing NY-ESO-1^c259, an affinity-enhanced T-cell receptor (TCR) recognizing an HLA-A2-restricted NY-ESO-1/LAGE1a-derived peptide, in patients with metastatic synovial sarcoma (NY-ESO-1^c259T cells). Confirmed antitumor responses occurred in 50% of patients (6/12) and were characterized by tumor shrinkage over several months. Circulating NY-ESO-1^c259T cells were present postinfusion in all patients and persisted for at least 6 months in all responders. Most of the infused NY-ESO-1^c259T cells exhibited an effector memory phenotype following ex vivo expansion, but the persisting pools comprised largely central memory and stem-cell memory subsets, which remained polyfunctional and showed no evidence of T-cell exhaustion despite persistent tumor burdens. Next-generation sequencing of endogenous TCRs in CD8⁺ NY-ESO-1^c259T cells revealed clonal diversity without contraction over time. These data suggest that regenerative pools of NY-ESO-1^c259T cells produced a continuing supply of effector cells to mediate sustained, clinically meaningful antitumor effects.Significance: Metastatic synovial sarcoma is incurable with standard therapy. We employed engineered T cells targeting NY-ESO-1, and the data suggest that robust, self-regenerating pools of CD8⁺ NY-ESO-1^c259T cells produce a continuing supply of effector cells over several months that mediate clinically meaningful antitumor effects despite prolonged exposure to antigen. Cancer Discov; 8(8); 944-57. ©2018 AACR.See related commentary by Keung and Tawbi, p. 914This article is highlighted in the In This Issue feature, p. 899.

Trial registration: ClinicalTrials.gov NCT01343043.

Figure 1.

Temporal pattern of clinical response and overall clinical outcomes. A, OS for patients who received NY-ESO-1^c259T-cell treatment (Kaplan—Meier plot). B, Swimlane plot shows clinical outcomes of treated patients. Patients 206 and 261 received less than the protocol-specific dose (<1 billion cells). Living subjects were censored for survival on March 30, 2017. Patient 208 discontinued study on June 16, 2016, due to development of chemotherapy-associated acute myelogenous leukemia. C, Spider plot illustrates protracted kinetics of maximal NY-ESO-1^c259T-cell mediated antitumor responses.

Figure 2.

Transient radiographic progression and tumor trafficking of NY-ESO-1^c259T cells. A, At the designated times following infusion of NY-ESO-1^c259T cells, miliary synovial sarcoma involving the lungs of patient 201 demonstrates radiographic progression on day 2, PR at week 4, and then CR at week 12, which persisted for 9 months. B, Radiographic progression on day 2 shown in A was associated with fever and elevated circulating C-reactive protein. C, Patient 206 presented with tumor enlargement, increased pleural fluid, fever, and constitutional symptoms on day 60 following NY-ESO-1^c259T-cell infusion. D, The pleural mass highlighted in yellow in C was resected and pleural fluid drained. NY-ESO-1^c259T cells identified by dextramer binding via flow cytometry were present within the tumor mass (left) and within the pleural fluid (right).

Figure 3.

NY-ESO-1^c259T cells show increased expansion in responding patients and persist with functional capacity to kill NY-ESO-1 targets for several months. A, Peak vector copies/μg of DNA demonstrating T-cell expansion in nonresponders (n = 6) versus responders (n = 6). B, T-cell persistence in nonresponders (n = 6) versus responders (n = 6). C, Enumeration of CD4⁺ (top) and CD8⁺ (bottom) NY-ESO-1^c259T cells based upon binding to NY-ESO-1 pentamer within the manufactured cell product (MP) and in peripheral blood from patient 202 at designated time points. D, Peripheral blood CD3⁺ T cells flow sorted from PBMCs obtained from patient 202 at 28 months postinfusion (blue line) or from a healthy control transduced with NY-ESO-1^c259T cells (red line) demonstrate equivalent killing of A375 HLA-A2⁺NY-ESO-1⁺ target cells. Data plotted as mean of duplicate replicates ± SD.

Figure 4.

The NY-ESO-1^c259T-cell pool evolves from one composed predominantly of effector cells in the manufactured product to a long-term persisting pool composed predominantly of TCM and TSCM cells. A, Flow cytometry of pentamer binding CD4⁺ (top) and CD8⁺ (bottom) NY-ESO-1^c259T cells expressing CCR7 and CD45RA from patient 202 within the manufactured cell product (MP) and at the designated time points following infusion. B and C, Summarized data for 4 individual patients (B) and composite data from the entire cohort (C).

Figure 5.

Longitudinal analysis of TCRBV sequences within sorted cell subsets. All cell samples are from patient 201. A, Clonality of each collected population at each time point. B, Average number of distinct long-lived clones (dark blue) and randomly selected control clones (light blue) present within each cell subset in the apheresis product. Over 1,000 bootstrapped samples from each group are selected then normalized to the total number of cells in each bootstrapped sample (19). C–G, The same as B, except clones represent NY-ESO-1^c259T cells in the manufactured product and at time points indicated in each graph. H, Cell counts of 11 long-lived clones elected based upon the presence of TCRBV sequences at each time point. Individual clones are connected across time, according to the legend at the top right. The cell subset containing an individual clone at a particular time point is denoted by the colored bars, according to the legend at the bottom right.