BackgroundGene-modified autologous T cells expressing NY-ESO-1c259, an affinity-enhanced T-cell receptor (TCR) reactive against the NY-ESO-1-specific HLA-A*02-restricted peptide SLLMWITQC (NY-ESO-1 SPEAR T-cells; GSK 794), have demonstrated clinical activity in patients with advanced synovial sarcoma (SS). The factors contributing to gene-modified T-cell expansion and the changes within the tumor microenvironment (TME) following T-cell infusion remain unclear. These studies address the immunological mechanisms of response and resistance in patients with SS treated with NY-ESO-1 SPEAR T-cells.MethodsFour cohorts were included to evaluate antigen expression and preconditioning on efficacy. Clinical responses were assessed by RECIST v1.1. Engineered T-cell persistence was determined by qPCR. Serum cytokines were evaluated by immunoassay. Transcriptomic analyses and immunohistochemistry were performed on tumor biopsies from patients before and after T-cell infusion. Gene-modified T-cells were detected within the TME via an RNAish assay.ResultsResponses across cohorts were affected by preconditioning and intra-tumoral NY-ESO-1 expression. Of the 42 patients reported (data cut-off 4June2018), 1 patient had a complete response, 14 patients had partial responses, 24 patients had stable disease, and 3 patients had progressive disease. The magnitude of gene-modified T-cell expansion shortly after infusion was associated with response in patients with high intra-tumoral NY-ESO-1 expression. Patients receiving a fludarabine-containing conditioning regimen experienced increases in serum IL-7 and IL-15. Prior to infusion, the TME exhibited minimal leukocyte infiltration; CD163+ tumor-associated macrophages (TAMs) were the dominant population. Modest increases in intra-tumoral leukocytes (≤5%) were observed in a subset of subjects at approximately 8 weeks. Beyond 8 weeks post infusion, the TME was minimally infiltrated with a TAM-dominant leukocyte infiltrate. Tumor-associated antigens and antigen presentation did not significantly change within the tumor post-T-cell infusion. Finally, NY-ESO-1 SPEAR T cells trafficked to the TME and maintained cytotoxicity in a subset of patients.ConclusionsOur studies elucidate some factors that underpin response and resistance to NY-ESO-1 SPEAR T-cell therapy. From these data, we conclude that a lymphodepletion regimen containing high doses of fludarabine and cyclophosphamide is necessary for SPEAR T-cell persistence and efficacy. Furthermore, these data demonstrate that non-T-cell inflamed tumors, which are resistant to PD-1/PD-L1 inhibitors, can be treated with adoptive T-cell based immunotherapy.Trial registrationClinicalTrials.gov, NCT01343043, Registered 27 April 2011.

102 Background: MAGE-A4 is a cancer/testis antigen with expression in many solid tumors promoting cell growth by preventing cell cycle arrest and apoptosis. This study (NCT03132922) evaluated safety and efficacy of SPEAR T-cells directed against the MAGE-A4 peptide expressed in 9 tumor types. Methods: This Phase I dose-escalation, expansion trial evaluated patients (pt) who were HLA-A*02 positive with advanced cancers that expressed the MAGE-A4 protein. Autologous T-cells from eligible patients were isolated, transduced with a lentiviral vector containing the MAGE-A4c1032 TCR, and expanded. Prior to ADP-A2M4 infusion, pts received a lymphodepletion regimen of cyclophosphamide and fludarabine. Cohorts 1, 2, 3, and expansion were to receive transduced cell doses of up to: 0.12 × 109, 1.2 × 109, 6 × 109, and 10 x 109, respectively. Results: As of 23 October 2019, 9 pts were treated in dose escalation with no DLTs; 25 pts were treated in expansion. Median age was 56.5 yr (range: 31-78). All pts received prior chemotherapy. Most common ( > 30%) AEs ≥Grade 3 were lymphopenia, leukopenia, neutropenia, anemia, thrombocytopenia, and febrile neutropenia. Two pts had trial-related deaths (aplastic anemia and CVA) leading to modification of the lymphodepletion regimen and eligibility criteria. In Cohort 3/expansion (28 pts), Best Overall Response was PR (7), SD (11), PD (5), non-evaluable (5). All PRs, at the time of data cut-off, were in patients with synovial sarcoma. Transduced T-cells were detectable in all patients. Tumor infiltration of SPEAR T-cells was detectable in several cohort 3/expansion pts. Conclusions: The Phase I single agent ADP-A2M4 trial will complete enrollment shortly and updated data will be presented. ADP-A2M4 shows promising efficacy and a manageable safety profile at a dose range of 1.2 – 10 × 109. Clinical activity in various tumors has led to a separate on-going low dose radiation sub-study of this trial, a Phase II trial in sarcoma (SPEARHEAD-1, NCT04044768), and a Phase I trial (SURPASS, NCT04044859) with ADP-A2M4CD8, a next-generation SPEAR T-cell targeting MAGE-A4. Clinical trial information: NCT03132922.

e14513 Background: MUC1 is a glycoprotein that is expressed in healthy tissues on the luminal surface of simple and glandular epithelium. In tumors that arise from these cells, an alternate form with aberrant glycosylation is frequently over expressed and distinguishes tumor associated TnMUC1 from normal MUC1. We have generated a novel chimeric antigen receptor (CAR) targeting the TnMUC antigen comprised of a mouse anti-human scFv derived from the monoclonal antibody 5E5 which recognizes the epitope comprising Tn glycan of MUC1, a CD8a transmembrane region and dual CD2 and CD3z intracellular signaling domains. CD2 signaling in T-cells has been demonstrated to result in delayed exhaustion. The novel incorporation of this co-stimulatory domain may lead to enhanced persistence of the CART cells which is believed to be critical for efficacy in solid-tumors. Methods: This is a multi-center first in human Phase I study to evaluate the safety and preliminary efficacy of CART-TnMUC1-Cells for the treatment of solid-tumors. Solid-tumors included in the dose-escalation phase include metastatic treatment-resistant ovarian cancer (OC), pancreatic adenocarcinoma (PC), triple-negative breast cancer (TNBC) or non-small lung cancer (NSCLC). All patients must have TnMUC1 expression as determined by immunohistochemistry. Results: As of January 2021, a total of six patients were treated. Three in Cohort 1 (no lymphodepletion; dose = 1-2 x 107 TDN CART cells; tumors = 1 OC, 1 TNBC and 1 PC) and 3 in Cohort 2 (flu/cy lymphodepletion; dose = 1-2 x 107 TDN CART cells; tumors = 1 NSCLC and 2 OC). None of the patients treated experienced DLT’s. The trial is currently enrolling to Cohort 3 (flu/cy lymphodepletion, 5-6 x 107 TDN). No CRS, neurotoxicity, serious adverse reactions and no on-target/off-tumor toxicity was observed at these dose levels. The most common AE’s were low-grade GI symptoms (e.g., nausea, abdominal pain) in 5/6 patients (83.3%), generalized disorders (e.g., chills, fatigue) in 5/6 (83.3%) and hematologic disorders (e.g., anemia, neutropenia) in 3/6 (50%) of patients. CAR expansion was demonstrated in all patients and was improved in Cohort 2 following LD chemotherapy. Preliminary efficacy assessed by RECIST v1.1 at Day +28 demonstrate SD in all patients in Cohort 2. Conclusions: This is the first report of a novel CART-TnMUC1 construct containing a CD2 co-stimulatory domain that has been used in clinical trials for the treatment of refractory solid-tumor malignancies. While the study is still early in dose-escalation having completed only 2 of 6 planned dose levels there is no evidence of safety concerns or on-target/off-tumor toxicity. Additional safety, efficacy and biomarker data is currently being reviewed and will be presented. Clinical trial information: NCT04025216.