Background: Hepatocellular carcinoma (HCC) is associated with poor prognosis and high mortality. Glypican-3 (GPC3) is a heparan sulfate proteoglycan overexpressed in 70-80% of HCC and is considered a potential therapeutic target for HCC. We modified T cells with a fusion protein of anti-GPC3 single-chain fragment variable (scFv) linked to CD3ε, which incorporates into the native T cell receptor/CD3 complex forming chimeric anti-GPC3 scFv-CD3ε engineered T cells (CT0180). Preclinically, CT0180 showed competitive antitumor activity but lower cytokine release compared to 28ζ or BBζ chimeric antigen receptor T cells. Methods: This is an open-label, dose-escalation/exploration phase I study to investigate CT0180 in patients with GPC3-positive advanced HCC (NCT04756648). The study objectives are to evaluate the safety, preliminary efficacy, and cellular pharmacokinetics of CT0180. Eligible patients underwent sequential apheresis, lymphodepletion, and cell infusion. Antiangiogenic drug monotherapy was allowed after apheresis as bridging therapy. Lymphodepletion regimen comprised fludarabine 25 mg/m² and cyclophosphamide 300 mg/m² daily for 3 days. Five dose levels (DLs, range 10×10⁶-600×10⁶ cells) with up to 3 cycles were explored using i3 + 3 design, and intra-patient dose-escalation was allowed. Data are reported as of 01-Feb-2023. Results: From Feb-2021 to Jul-2022, 7 patients with hepatitis B virus-related HCC were treated with CT0180 (one patient each at 10×10⁶ and 30×10⁶ DLs, 3 at 100×10⁶ DL, and 2 at 300×10⁶ DL). The median age was 46 (28–74) years. Patients had prior 2 lines or more systemic therapy, with at least one antiangiogenic tyrosine kinase inhibitor and/or anti-PD-1/PD-L1 immunotherapy. Most common grade 3-4 adverse events were hematologic toxicities, ie lymphopenia and neutropenia, which were considered to be related to lymphodepletion. No dose-limiting toxicities, immune effector cell-associated neurotoxicity syndrome, or treatment-related deaths occurred. Grade 1 cytokine release syndrome was observed in 6 patients; tocilizumab was given in one patient and no glucocorticoids were used. All 7 patients were evaluable for efficacy, in which 2 achieved partial response (PR) (30×10⁶ and 300×10⁶ DL) and 3 achieved stable disease (SD) (10×10⁶, 100×10⁶ and 300×10⁶ DL). One patient sustained PR for 6.7 months, and one patient sustained SD for 6.1 months with follow-up ongoing. Median follow-up time was 15.9 months, and median overall survival was 11.6 months with 3 patients alive at last follow-up. CT0180 transgene copy number ranged 47–4487 copies/μg genomic DNA and peaked on either D3 or D7. Interleukin-6 increased and peaked on D1 after CT0180 infusion in most patients. Conclusions: CT0180 demonstrated manageable safety profile and promising antitumor potential. Further exploration of CT0180 in HCC is needed. Clinical trial information: NCT04756648.