Safety and preliminary efficacy of a novel armored mesothelin-targeted CAR-T therapy in advanced solid tumors.

Abstract

2543 Background: UCMYM802 is an autologous mRNA-electroporated CAR-T cell therapy targeting mesothelin (MSLN), incorporating a lymphocyte-antigen-presenting cell costimulatory (LACO-Stim) molecule. This first-in-human study aimed to assess its safety, tolerability, preliminary efficacy, and pharmacokinetic/pharmacodynamic (PK/PD) profiles. Methods: This phase I study enrolled patients with metastatic or recurrent MSLN-positive solid tumors who failed standard therapies. Three dose levels were evaluated: 1×10 8 (n = 1), 5×10 8 (n = 1), and 1×10 9 CAR-T cells (n = 7). No lymphodepletion was required prior to infusion. Patients received up to 4 weekly infusions. Primary endpoints were safety and determination of the maximum tolerated dose (MTD). Secondary endpoints included objective response rate (ORR) by RECIST v1.1, disease control rate (DCR), and PK/PD parameters. Results: As of Nov. 2025, 9 patients were enrolled and treated (median age 67, range 52-67). Tumor types included pancreatic cancer (n = 3), peritoneal malignant mesothelioma (n = 1), adenocarcinoma of unknown primary, consistent with a gynecologic or peritoneal origin (n = 1), extrahepatic cholangiocarcinoma (n = 1), cholangiocarcinoma (n = 1), ovarian cancer (n = 1), and lung cancer (n = 1). All patients received at least one infusion. The 1×10 9 dose level was defined as the MTD, with one dose-limiting toxicity (DLT) observed (grade 4 cytokine release syndrome [CRS] with hypotension). Treatment-related adverse events (TRAEs) occurred in all patients, most commonly CRS (88.9%), fever (88.9%), decreased lymphocyte count (77.8%), anemia (77.8%), prolonged prothrombin time (66.7%), and hypoxia (55.6%). Grade ≥3 CRS occurred in 25% of patients. Among seven efficacy-evaluable patients, two patients achieved partial response (PR; one each in 5×10 8 and 1×10 9 cohorts), and two had stable disease (SD; one each in 1×10 8 and 1×10 9 cohorts). The ORR was 28.6% (2/7) and DCR was 57.1% (4/7). Notable responses included a PR in gynecologic or peritoneal origin adenocarcinoma and a PR at 6 months (converted to SD at 7.5 months) in extrahepatic cholangiocarcinoma. PK analysis showed peak CAR-T expansion (by copy number) at 1-hour post-infusion, with the highest and most sustained exposure after the first infusion. PD analysis indicated IFN-γ elevation correlating with CAR-T peak expansion. CAR-T exposure and activation marker expression showed a dose-dependent trend. Conclusions: UCMYM802 demonstrated a manageable safety profile consistent with expected CAR-T-related toxicities, primarily CRS. Preliminary anti-tumor activity was observed in heavily pretreated patients with MSLN-positive advanced solid tumors, with encouraging signals in gynecologic or peritoneal origin adenocarcinoma and biliary tract cancers. The recommended dose for expansion is 1×10^9 CAR-T cells. These results support continued investigation of UCMYM802. Clinical trial information: NCT06256055 .