Preclinical Proof of Concept for Decentralized Manufacturing of a MAGE-A4/CD8α-Expressing Autologous T-Cell Therapy for Solid Tumors

Introduction: Uzatresgene autoleucel (uza-cel) is a next-generation T-cell therapy targeting melanoma-associated antigen (MAGE) A4. Uza-cel uses the same engineered T-cell receptor (TCR) as afamitresgene autoleucel (afami-cel), which has shown efficacy in synovial sarcoma (NCT04044768; D'Angelo SP. Lancet. 2024;403:1460) with an additional CD8α co-receptor that data suggest improves potency and broadens the immune response against non-sarcoma MAGE-A4-expressing tumors (Anderson VE. J Immunother. 2023;46:132). Preliminary efficacy and a manageable safety profile were seen in the Phase 1 SURPASS trial (NCT04044859) of uza-cel in patients with MAGE-A4-expressing advanced solid tumors. Overall response rates per Response Evaluation Criteria in Solid Tumors (RECIST) v1.1 were seen in six of 15 (40%) ovarian, three of four (75%) head and neck, and four of seven (57%) urothelial cancers as of August 14, 2023, leading to SURPASS expansion cohorts and the Phase 2 SURPASS-3 trial focusing on these tumor types. TCR T-cell therapy can have a manufacturing time that leads to intervals of several weeks between apheresis and treatment infusion, including in SURPASS, where uza-cel has been manufactured centrally. This lag time could be detrimental for patients with advanced tumors. We investigated suitability of a decentralized, rapid, and fresh cell manufacturing platform to produce T cells expressing the MAGE-A4 TCR and CD8α co-receptor compared with the established central manufacturing process for uza-cel.

Methods: We tested a rapid, enclosed, fully automated system that uses fresh cells, with monitoring software and quality control testing, which can be deployed close to treating clinical sites for faster cell therapy delivery. T cells were manufactured at the respective labs from the same Good Manufacturing Practice-grade lentiviral vector and leukapheresis material from three healthy donors by the Galapagos decentralized manufacturing (GDM) platform or the current centralized process (CCP) that uses frozen cells. Both labs assessed transduction efficiency, vector copy number, and cytotoxicity (measured by xCELLigence real-time cell analysis and Incucyte live cell analysis) of the T cells. Cryopreserved GDM T cells and uza-cel were exchanged between sites for direct comparisons of interferon γ production, cytotoxicity, and cell phenotypes (assessed by flow cytometry with a panel of markers). Transduced CD4+/- cells were sorted by flow cytometry, followed by single-cell RNA sequencing to assess differences in gene expression profiles between manufacturing processes.

Results: GDM and CCP manufacturing times were 6 and 10 days, respectively. Transduction efficiency was 40.1-63.5% and 39.3-61.1%, cell yields were 0.407-0.758x10⁹ and 2.00-2.20x10⁹, and viral copy number per transduced cell was 4.89-7.07 and 6.03-8.57 by GDM and CCP, respectively. Both labs found that GDM T cells produced more interferon γ in response to MAGE-A4-expressing target cells at all effector:target (E:T) ratios. Both labs observed similar cytotoxicity between manufacturing methods at higher E:T ratios, whereas GDM T cells showed increased cytotoxicity at lower E:T ratios and maintained cytotoxicity over time better. GDM T cells comprised fewer apoptotic cells, more cells displaying a stem cell memory phenotype, and a broadly altered immunophenotype. GDM T cells had a less differentiated, more proliferative gene expression profile and had greater proliferative capacity. GDM T cells had a less cytotoxic gene expression profile but killed more quickly.

Conclusions: This study provides proof of concept that the GDM platform can produce MAGE-A4/CD8α-expressing T cells with features that may result in improved efficacy and durability of response in the clinic compared with the existing manufacturing procedure. Importantly, the GDM platform could significantly reduce total vein-to-vein time for eligible patients from several weeks to ~1 week. The GDM system will be used to supply MAGE-A4/CD8α-expressing T cells for a proof-of-concept clinical trial to assess their safety and efficacy in patients with advanced head and neck cancer.

Acknowledgments: Alasdair Gunn, Alicia Derrac Soria, Elizabeth Evans, Laura Hudson, Olusegun Oshota, Will Spinner, Rebecca Chamberlain, Clare Patrick, Jack Turner, Charlotte Lucy, Catherine Mercer, George Zervakis.

Herman:Adaptimmune Ltd: Current Employment, Current equity holder in publicly-traded company, Current holder of stock options in a privately-held company. Gobessi:Galapagos: Ended employment in the past 24 months. Sand:Galapagos: Current Employment. Wagner:Galapagos: Current Employment. Yuan:Galapagos: Current Employment. Davis:Adaptimmune Ltd: Current Employment. Donaldson:Adaptimmune: Current Employment, Current equity holder in publicly-traded company. Butler:Adaptimmune Ltd: Current Employment, Current equity holder in publicly-traded company. Bath:Adaptimmune: Current Employment. Harris:Adaptimmune: Current Employment. Golden:Adaptimmune: Current Employment, Current equity holder in publicly-traded company. Tipping:Adaptimmune Ltd: Current Employment, Current equity holder in publicly-traded company. Sanderson:Adaptimmune: Current Employment, Current equity holder in publicly-traded company, Current holder of stock options in a privately-held company. Mellors:Galapagos: Current Employment. Debnam:Adaptimmune Ltd: Current Employment, Current equity holder in publicly-traded company.