Abstract Chimeric antigen receptor (CAR)-T cell therapy has improved outcomes for patients with multiple myeloma (MM), but its broader use is restricted by manufacturing complexities and treatment-related toxicities. AZD0120 is a dual-targeting B-cell maturation antigen (BCMA)/CD19 CAR-T cell therapy manufactured via the rapid FasTCAR process. We developed a dual-targeting “loop” CAR that incorporates a novel humanized anti-BCMA single-chain variable fragment (scFv), clone SG, and an FMC63-derived anti-CD19 scFv. This AZD0120 CAR preserved functional binding to both antigens and conferred robust in vitro and in vivo cytotoxicity while maintaining single-antigen reactivity. Conventional manufacture of CAR-T cells with the AZD0120 CAR (AZD0120C) yielded cells with minimal tonic signaling, limited responsiveness to soluble BCMA, and preservation of naïve/stem cell memory-enriched phenotypes, yet robust cytokine production upon BCMA + target engagement. AZD0120C demonstrated cytotoxicity comparable to benchmark BCMA CAR-Ts across MM lines in vitro and showed strong in vivo expansion and tumor control in xenograft models. FasTCAR manufacturing – designed to shorten vein-to-vein timelines and enrich less-differentiated phenotypes – further enhanced in vivo performance: AZD0120 consistently achieved superior tumor control and greater CAR-T expansion vs AZD0120C across disseminated MM.1S, NALM-6, and JeKo-1 models, with superior efficacy observed at lower cell doses. Collectively, these data support clinical evaluation of AZD0120 as a differentiated BCMA/CD19 CAR-T cell therapy with the potential to improve disease control and patient access in MM. Key Points AZD0120 is a dual-targeting CAR-T that displays a favorable anti-myeloma functional profile and co-targets a source of potential relapse The FasTCAR process yields T N/SCM -rich CAR-T populations, promotes in vivo expansion and achieves potent tumor control in xenograft models

Abstract Introduction: Small-cell lung cancer (SCLC) is aggressive with poor prognosis, and current immunotherapies have limited efficacy. Delta-like ligand 3 (DLL3), expressed in ∼80% of SCLC with restricted normal-tissue expression, is a promising CAR T target. We developed nanobody-based, autologous DLL3 CAR T cells armed with a dominant-negative TGFβ receptor II (DNR) to resist TGFβ-driven immunosuppression and enhance antitumor activity. Methods: Candidate DLL3 binders were derived from a camelid-immunized phage display library and engineered into second-generation CARs. Constructs were tested in vitro for specific lysis with target cell lines. Lead binders were screened for cross-reactivity against Notch ligands. The DLL3 epitope was mapped, and membrane-proximal binders targeting EGF3-6 were prioritized. Off-target interactions were assessed using a membrane proteome array. Multiple CARs—with varying binders/biparatopic combinations, co-stimulation domains, and hinges—were compared in short-term and repeated antigen-stimulation assays across cells with varying DLL3 levels in vitro. Lead CAR T cells with or without DNR were compared in vitro and in vivo. Top constructs advanced to testing in multiple CDX and PDX models to evaluate antitumor efficacy and safety. Manufacturing was compared between a 3-day FasTCAR and an 8-day conventional process. Results: Three VHH binders demonstrated high DLL3 specificity, with no cross-reactivity to Notch family ligands and no off-target binding, and similar affinity for mouse DLL3. The lead biparatopic CAR construct, B2, demonstrated superior cytotoxicity and durability in short-term and repeated antigen-stimulation assays. When armored with DNR, B2-DNR CAR T cells showed augmented persistence versus the unarmored counterparts. In vivo safety assessments showed no body-weight loss in CAR T-treated mice across multiple dose levels, suggesting a wide therapeutic index. T cells were present in the mouse pituitary pars intermedia, where DLL3-positive cells are enriched, but their DLL3 expression was largely intracellular rather than on the cell surface. B2 did not show on-target, off-tumor toxicity. B2-DNR CAR T cells rapidly cleared established tumor xenografts with high and low DLL3 expression in immunodeficient mice, correlating with robust intratumoral CD3+ T-cell infiltration. In a TGFβ-rich PDX model, B2-DNR CAR T cells outperformed the unarmored counterpart, supporting the functional benefit of DNR. The 3-day FasTCAR manufacturing process demonstrated greater efficacy and durability than the 8-day process. Conclusions: The camelid nanobody-derived, DLL3-targeted CAR T-cell therapy manufactured by a 3-day FasTCAR process demonstrates the potential to elicit deep and durable antitumor responses with a favorable safety profile. Preclinical data for the lead candidate, B2-DNR, support further clinical development for the treatment of SCLC. Citation Format: Qi Dong, Wenjie Yin, Xiangling Dai, Tao Wang, Guangyao Zhu, Manli Yin, Yu Zou, Yu Yang, Di Wu, Mark Cobbold, Lianjun Shen. Nanobody-based, DLL3-directed FAST-CAR T-cell therapy for small cell lung cancer (SCLC) [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2026; Part 1 (Regular Abstracts); 2026 Apr 17-22; San Diego, CA. Philadelphia (PA): AACR; Cancer Res 2026;86(7 Suppl):Abstract nr 1526.

Abstract Background: AZD6750 is a CD8α-guided IL-2 immunocytokine engineered to enhance the therapeutic index of IL-2 by reducing aldesleukin-associated toxicities, while preserving the efficacy that has been compromised with recent clinical IL-2 mutein strategies. This should result in a safer, more effective IL-2 with the potential to benefit larger numbers of patients and allow for the utilization of IL-2 in synergistic combinations. Combinations of anti-PD-1 with IL-2 therapies are currently being explored in the clinic with both preliminary clinical and nonclinical evidence that this combination can generate improved responses. As such, a potentially effective therapy is the combination of AZD6750 plus rilvegostomig, a monovalent, Fc-reduced, bispecific IgG1 antibody against PD-1 and TIGIT receptors, to enhance IO activity. Methods: In vitro, antigen-specific tumor cell cytolysis was evaluated after treatment with AZD6750, rilvegostomig, or a combination of both therapies. Ex vivo, non-small cell lung cancer (NSCLC) patient-derived tumor samples were treated with AZD6750, rilvegostomig, or a combination of both therapies, and the activity evaluated by IFN-γ secretion. In vivo, NSG mice were implanted subcutaneously with antigen-expressing tumors, engrafted with expanded human PBMCs containing enriched antigen-specific CD8+ T cells, and treated with AZD6750, rilvegostomig, a combination of both therapies, or an isotype control and tumor growth was monitored. Results: In vitro, AZD6750 in combination with rilvegostomig improved antigen-specific cytolysis of tumor cells compared with AZD6750 monotherapy (EC50 p<0.01). In addition, ex vivo treatment of primary NSCLC tumors with rilvegostomig plus AZD6750 drove an increase in IFN-γ secretion compared with either AZD6750 (2-fold) or rilvegostomig (6-fold) monotherapies, highlighting the potential of this combination to augment functional responses in tumor infiltrating lymphocytes. In vivo, AZD6750 plus rilvegostomig resulted in a statistically significant reduction in tumor growth rate when compared to an isotype IgG (all studies), rilvegostomig (all studies), or AZD6750 (in 2 out of 3 studies) in an antigen-specific humanized mouse tumor model. Conclusions: These preclinical data demonstrate that AZD6750 combined with rilvegostomig can enhance anti-tumor immune responses, leading to greater tumor cell killing. The Phase 1 study investigating AZD6750 in select advanced or metastatic solid tumors is currently ongoing (NCT07115043). Citation Format: Matthew J. Elder, Aidan H. Riley, Sin Lih Tan, Jerome Mastio, Bruno Frederico, Fabien Garcon, Hena Khalique, Georgina Bowyer, Paul Chariou, Nicholas M. Durham, Maria Broggi, Emily Hsiue, Simon Rodney, Jonathan B. Fitzgerald, Nadia Luheshi, Mark Cobbold, Saso Cemerski, Simon J. Dovedi. AZD6750, a CD8α-guided IL-2 immunocytokine effectively combines with rilvegostomig, a PD-1/TIGIT bispecific antibody, to enhance endogenous immunity [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2026; Part 1 (Regular Abstracts); 2026 Apr 17-22; San Diego, CA. Philadelphia (PA): AACR; Cancer Res 2026;86(7 Suppl):Abstract nr 4340.

Deficiencies in T cells are a major limitation for cancer immunotherapy. Polymorphonuclear myeloid-derived suppressor cells (PMN-MDSC) are key mediators of immune suppression, contributing to poor clinical outcomes and resistance to checkpoint inhibitors (CPI). Excessive peroxidation of polyunsaturated phospholipids, especially phosphatidylethanolamines (PE), is a critical stage of the ferroptotic cell death process. PMN-MDSC are undergoing ferroptosis in tumors. Here, we show that initiation of ferroptosis causes PMN-MDSCs to release oxidized PE (oxPE) via extracellular vesicles, shielding themselves from immediate death while triggering fatal injury to neighboring T cells. The accumulation of oxPE and oxidized bis-monoacylglycetophosphates (oxBMP) in the endo-lysosomes of T cells increases lysosomal membrane permeability (LMP), resulting in T cell death. In vivo inhibition of ferroptosis in PMN-MDSCs reduces LMP, increases intratumoral T cell numbers, and promotes antitumor effects across multiple mouse models. Targeting ferroptosis reverses CPI resistance in mice. Analysis of tumor samples from cancer patients revealed association of T cell depletion with ferroptotic PMN-MDSC and T-cell depletion in regions enriched for PMN-MDSCs, with T cell LMP closely associated with PMN-MDSC proximity. This translated to worse clinical outcome in patients with high ferroptosis signature. These findings indicate that PMN-MDSC ferroptosis drives T cell loss, highlighting a potential therapeutic target for cancers with high PMN-MDSC infiltration.

Abstract Background: Ewing sarcoma (EWS) is a rare, aggressive bone and soft tissue malignancy predominantly affecting children and young adults. Despite advances in multimodal therapy, treatment options remain limited for refractory or relapsed EWS (1). Chimeric antigen receptor (CAR) T-cell therapy has shown marked efficacy in hematologic malignancies and is being developed for solid tumors. AZD0754 is a STEAP2-targeted CAR T-cell therapy incorporating a dominant-negative TGFβRII (dnTGFβRII) armoring strategy and is in Phase 1 clinical development for prostate cancer. Recent published proteomic analyses identified STEAP2 expression in EWS patient samples and cell line-derived xenografts, supporting STEAP2 as a potential candidate immunotherapeutic target (2). Methods: STEAP2 surface expression was evaluated across different human EWS cell lines by flow cytometry. AZD0754-mediated cytotoxicity was assessed in vitro using the xCELLigence real-time cell analysis platform. Based on in vitro results, A673, RD-ES, and SK-NEP-1 xenograft models were selected for in vivo efficacy studies. Mice received varying doses of AZD0754, and tumor growth and overall survival were monitored. Blood was collected at different time points for serum cytokine analysis to assess pharmacodynamic activity. Results: EWS cell lines expressed detectable cell surface expression of STEAP2 at receptor densities much lower than workhorse prostate cancer cell lines. Despite this, AZD0754 was capable of inducing antigen-dependent cytotoxicity in vitro and suppressing tumor growth in EWS xenograft models. AZD0754 exhibited dose-dependent antitumor activity that correlated with serum IFNγ levels, indicating on-target immune engagement. Conclusions: AZD0754 demonstrates robust preclinical anti-tumor activity against EWS in vitro and in vivo, with dose-dependent efficacy and corresponding cytokine levels. These findings support STEAP2 as a potential therapeutic target in EWS. STEAP2 targeting has the potential to address a critical unmet need in EWS by expanding treatment options to biologic and immune-based therapies. References1. National Center for Biotechnology Information (2021). Ewing Sarcoma. NCBI Bookshelf. 2. Mooney B, Negri GL, Shyp T, Delaidelli A, et al. Surface and global proteome analyses identify ENPP1 and other surface proteins as actionable immunotherapeutic targets in Ewing sarcoma. Clin Cancer Res.2024;30(5):1022-1037. doi:10.1158/1078-0432.CCR-23-2187. Citation Format: Peter Zanvit, Brianna Janocha, Shannon Breen, Christine Fazenbaker, Ryan Golden, Jonathan Fitzgerald, Mark Cobbold, Gordon Moody, Emily Bosco. Efficacy of AZD0754, a dominant-negative TGFβRII-armored STEAP2 CAR T-cell therapy, in Ewing sarcoma xenograft models [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2026; Part 1 (Regular Abstracts); 2026 Apr 17-22; San Diego, CA. Philadelphia (PA): AACR; Cancer Res 2026;86(7 Suppl):Abstract nr 7808.

Background. Antibody drug conjugates (ADCs) represent a transformative class of cancer therapeutics, yet the mechanisms underlying their synergy with immunotherapy remain poorly understood. We investigated the mechanistic basis of ADC combinations with T cell engagers (TCEs) and checkpoint inhibitors (CPI). Methods ADC/TCE and ADC/CPI combinations were evaluated in vitro using co culture cytotoxicity assays and synergy analyses, and in vivo in humanized mouse tumor models. Mechanistic studies employed TNF and cytokine blocking antibodies, receptor knockout cell lines, autophagy inhibitors, and siRNA silencing. Results ADC/TCE combinations produced synergistic antitumor activity independent of target antigen and payload, persisting despite ADC induced T cell loss. ADC treatment induced autophagy, upregulating TNF receptors (TNFR1/2) and mannose 6 phosphate receptor (M6PR) on tumor cell surfaces. In ADC/TCE combinations, TCE derived TNF; acting on ADC upregulated TNFRs was the primary mediator of enhanced cytotoxicity, confirmed in vivo by TNF blockade. In contrast, combinations with CPI expanded antigen specific T cells operated through a TNF independent, M6PR dependent pathway involving enhanced granzyme B uptake. Conclusions ADC induced autophagy is a unifying, target- and payload-agnostic mechanism sensitizing tumor cells to T cell mediated killing. TCEs exploit a TNF/TNFR axis, whereas antigen specific T cells leverage granzyme B/M6PR uptake. This mechanistic framework enables rational selection and design of ADC immunotherapy combination strategies.

Glypican-3 (GPC3) is highly expressed in hepatocellular carcinoma (HCC), making it an attractive target for chimeric antigen receptor (CAR) T cell therapy; however, this approach has previously shown limited clinical efficacy, potentially owing to high levels of transforming growth factor-β (TGFβ) in the tumour microenvironment1-4. We therefore engineered CAR T cells with a dominant-negative TGFβ receptor II, which showed enhanced antitumour activity in preclinical studies5. Here we report findings from a first-in-human trial evaluating the safety and efficacy of C-CAR031 in patients with advanced, treatment-refractory HCC ( NCT05155189 ). Thirty-six patients received CAR T infusions at four dose levels (from 0.75 × 106 to 4.0 × 106 cells per kg). Cytokine release syndrome was reported in 34 patients, of which two cases were grade 3. Nine patients had non-haematological adverse events of grade 3 or higher. Tumour regression was observed in 32 patients, with a median best tumour reduction from baseline of 41.6% (range: 3.4-94.4%) in target lesions. The objective response rate was 44.4%, and the median duration of response was 4.4 months (95% confidence interval: 2.9-7.4). Median progression-free survival and overall survival were 4.2 months (95% confidence interval: 2.9-4.8) and 14.2 months (95% confidence interval: 10.1 to not evaluable), respectively. High-throughput analyses of tumour samples and functional validation suggested that GPC3 antigen loss and increased TGFβ levels may contribute to C-CAR031 resistance. Collectively, these results indicate that C-CAR031 has a manageable safety profile and encouraging antitumour activity in heavily pretreated patients with advanced HCC.

Abstract Background: Rilvegostomig (AZD2936) is a monovalent, bispecific humanized IgG1 monoclonal antibody that targets human TIGIT and PD-1, engineered with a triple-mutation in its fragment crystallizable (Fc)-domain to prevent Fc-effector functionality. Rilvegostomig has shown promising anti-tumor activity in NSCLC patients in a Phase I/II clinical trial (NCT04995523) and is currently being investigated in multiple Phase III trials. Here, we report the enhanced anti-tumor activity of rilvegostomig compared to a clinically approved anti-PD-1 (αPD1) mAb in ex vivo and in vivo models of head and neck squamous cell carcinoma (HNSCC). Methods: Primary HNSCC explants, derived from freshly resected patient tumors, were cultured ex vivo in a 3D platform for therapeutic agent testing. Ex vivo immune activity was assessed via IFN-γ quantification. Baseline tumor features—characterized by IHC, flow cytometry, scRNA-seq, and multiplex imaging—were used to stratify patient response and identify candidate biomarkers. A CD34+ humanized mouse model of HNSCC was developed to enable long-term engraftment of functional multi-lineage hematopoietic cells and used to investigate efficacy and effects of IO therapeutics on different components of the immune system. Results: We investigated a cohort of 42 newly diagnosed HNSCC patient resected tumor samples. These samples represent the incidence of HNSCC by anatomical sites and share similar cohort attributes as those enrolled in KEYNOTE-689. Patient-derived tumor explants maintain tumor architecture and intact tumor microenvironment (TME), including tumor cells, immune cells, stroma cells, and extracellular matrix. αPD1 monotherapy elicited immune activity in 9.5% of HNSCC samples tested in our ex vivo platform, consistent with clinical observation of major pathological response in KEYNOTE-689. In contrast, rilvegostomig induced a significantly higher rate of ex vivo immune activity (28.6%). Rilvegostomig’s activity was observed in samples derived from multiple anatomical sites, showing the highest response rate in oral cavity samples. Moreover, our analyses identified key baseline correlates that distinguish rilvegostomig activity from αPD1. Lastly, using a CD34-engrafted humanized mouse model of oral cavity HNSCC, we compared the therapeutic efficacy of αPD-1 monotherapy with rilvegostomig. In contrast to αPD-1, which produced modest effects on tumor growth, rilvegostomig induced significant tumor growth inhibition. Conclusions: Rilvegostomig elicits greater immune cell activation and broader activity in samples derived from different anatomical sites of HNSCC compared to αPD1 monotherapy in a 3D ex vivo platform and a humanized mouse model of HNSCC. These results underscore rilvegostomig's potential as a promising treatment for newly diagnosed HNSCC and support its investigation in clinical trials. Citation Format: Jun Ren, Ngan Mitchell, Marie Boutet, Vladimir Roudko, Jeremy Ratiu, Jorge Blando, Sophia Varriano, Elizabeth Galery, Rebecca Halpin, Trevor Connor, Kanam Malhotra, Bartholomew Starich, Jessica Wagner, Yi Luan, Anna Huntley, Joseph Boland, Maurizio Scaltriti, Mark Cobbold, Simon Hollingsworth, Scott Hammond, Doug Palmer, Paul Chariou, Kristen Pollizzi, Daniela Dinulescu, . Rilvegostomig elicits greater immune activation and tumor inhibition than clinically approved anti-PD-1 monotherapy in models of HNSCC [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2026; Part 1 (Regular Abstracts); 2026 Apr 17-22; San Diego, CA. Philadelphia (PA): AACR; Cancer Res 2026;86(7 Suppl):Abstract nr 5546.

Poor tumor antigenicity is an important cause of non-response to immune checkpoint blockade (ICB)-therapy in many cancer patients and mandates new treatment strategies. Here we explore the use of proteolytically activated antibody-peptide epitope conjugates (APECs) to redirect the immunological effector activities of tumor-infiltrating antiviral bystander CTLs against head and neck squamous cell carcinoma (HNSCC) by loading cancer cell surface MHC I proteins with viral peptides. We find that inflationary T cell memory responses against common viral pathogens such as CMV and EBV unfold superior anti-tumor activity compared to conventional memory responses under APEC therapy in an ICB-resistant preclinical model of HNSCC. Mechanistically, APEC activation required cancer cell-intrinsic protease activity, even for proteases expressed by cells of the tumor stroma. Data mining and functional screening identified the protease activity of plasminogen activator urokinase (PLAU) as widely shared between human HNSCC cancers and as a highly efficient proteolytic activator of APECs. Furthermore, peripheral blood analysis in HNSCC patients reliably predicts the specificity, magnitude, and quality of intratumoral bystander CTL, thus allowing for the screening of HNSCC patients who may benefit most from APEC therapy. ### Competing Interest Statement MC and DGM are inventors on a patent (US9402916B2) related to the design and use of APECs for cancer therapy. SIP has received consultancy payments from Abbvie, AstraZeneca/MedImmune, CUE Biopharma, Fusion Pharmaceuticals, G1 Therapeutics, Incendia, Inovio, MSD/Merck, Newlink Genetics, Oncolys Biopharma, Parthenon Therapeutics, Recurrent Respiratory Papillomatosis Foundation, Replimune, Scopus Biopharma, Sensei Biotherapeutics, and Umoja; and research grants from Abbvie, AstraZeneca/MedImmune, CUE Biopharma, Eisai, Merck, Recurrent Respiratory Papillomatosis Foundation, Sensei Biotherapeutics, and Tesaro, outside of the submitted work. National Cancer Institute, CA240239, CA278212, CA232103 National Institute of Allergy and Infectious Diseases, https://ror.org/043z4tv69, AI123349

Introduction: BCMA-targeted CAR T-cell therapies have shown remarkable clinical benefit for patients (pts) with multiple myeloma (MM), although relapse remains a challenge. Recurrent disease tends to maintain BCMA expression, possibly indicating the existence of cancer stem cells resistant to BCMA-targeted therapies. Clonotypic CD19+ B cells from pts with MM have been shown to be resistant to anti-MM therapies. These B cells may serve as a progenitor cell reservoir, although their contribution to disease relapse is unknown. We previously reported that GC012C, a BCMA/CD19 dual-targeting CAR T-cell therapy manufactured conventionally, can eliminate more myeloma colony-forming cells in vitro than CAR T cells targeting BCMA or CD19 alone. AZD0120 (GC012F) is a clinical-stage BCMA/CD19 dual-targeting CAR T-cell therapy (using the same binders as GC012C) manufactured using the proprietary FasTCAR platform to generate younger, fitter cells, enabling robust in vivo expansion to achieve deeper and more durable clinical responses in pts with MM. Here we present the preclinical characterization and efficacy of AZD0120. Methods: AZD0120 is a second-generation CAR T-cell therapy using a 4-1BB costimulatory domain and 2 antigen binders configured in a loop orientation and manufactured on the FasTCAR platform, in which T cells are transduced with lentivirus and cryopreserved shortly after activation to retain a young CAR T-cell phenotype and maximize in vivo expansion. Lead binders to BCMA and CD19 were selected based on their cytotoxic potential against cancer cell lines with low antigen expression and are the same for GC012C and AZD0120. Binding kinetics of the CAR were measured using surface plasmon resonance, and total cellular avidity for tumor cells was evaluated by z-Movi. Binder specificity was assessed through membrane protein array and tissue cross-reactivity assays. Phenotypic and proliferative characteristics, as well as cytokine release profiles of AZD0120 and GC012C, were characterized by flow cytometry. Cytotoxicity against BCMA+ and/or CD19+ target cells was quantified by flow cytometry, luciferase-based reporter assays, or real-time cell analysis. In vivo studies were performed in immunodeficient mice bearing xenografts expressing BCMA and/or CD19 to evaluate pharmacokinetic and pharmacodynamic responses. Results: The extracellular domain of the AZD0120 CAR demonstrated high specificity for both BCMA and CD19, with no detectable off-target binding. GC012C exhibited higher avidity for Nalm6 (CD19+BCMAlow), MM.1S (CD19−BCMA+), and JeKo-1 (CD19+BCMA+) cancer cell lines versus monospecific CAR T cells. GC012C exhibited comparable or superior capacity for cytolytic activity against all 3 tumor cell lines versus monospecific CAR T cells. Lastly, GC012C demonstrated minimal IFN-γ release or CD107α mobilization in culture alone or in response to soluble BMCA versus benchmark CAR T cells. AZD0120 (manufactured with the FasTCAR platform), as compared with GC012C, was enriched for naive/stem cell memory (Tn/scm) and central memory T cells (Tcm) and demonstrated enhanced proliferation upon encountering cells expressing BCMA or CD19. On day 2 post thaw, AZD0120 showed higher Tn/scm levels than GC012C in both donor 1 (71.4% vs 12.5%) and donor 2 (35% vs 22.4%), with Tcm levels of 15.7% vs 13.6% and 49.8% vs 7.7%, respectively. AZD0120 showed potent cytotoxicity against CD19+ and/or BCMA+ cell lines in vitro, coupled with robust IFN-γ release. In mouse models, GC012C demonstrated robust antitumor activity in a disseminated MM.1S MM model and subcutaneous RPMI-8226 model, with peak CAR T-cell expansion (Cmax) observed ~day 14 post dose. However, AZD0120 had a more durable and profound antitumor response than GC012C against Nalm6 and MM.1S xenografts in vivoat comparable doses, with a 10- to 100-fold higher Cmax levels in the blood. Furthermore, AZD0120-treated mice had increased frequencies of CXCR4⁺ CAR T cells within the blood, spleen, and bone marrow, a subset that may be critical for effective MM targeting. Conclusion: The FasTCAR-based AZD0120, designed to target BCMA and/or CD19 and maximize in vivo expansion, demonstrated potent, dose-dependent, and antigen-specific cytotoxicity both in vitro and in vivo. The superior antitumor function of AZD0120 versus GC012C suggests benefit of the FasTCAR manufacturing platform and supports continued development of AZD0120 as a promising therapeutic option for pts with MM.

BACKGROUND:Radiation therapy has long been a cornerstone of cancer treatment. More recently, immune checkpoint blockade has also been applied across a variety of cancers, often leading to remarkable response rates. However, photon-based radiotherapy-which accounts for the vast majority-is also known to frequently induce profound lymphopenia, which might limit the efficacy of immune system-based combinations. Proton beam therapy is known to produce a less drastic lymphopenia, which raises the possibility of greater synergy with immunotherapy. In this study, we aimed to explore the exact nature of the differential impact of the two radiation modalities upon the immune system. METHODS:We used multiparametric flow cytometry and deep sequencing of rearranged TCRb loci to investigate a cohort of 20 patients with gastrointestinal tumors who received either therapy and developed lymphopenia. RESULTS:Proton-treated patients remained relatively stable throughout treatment by most metrics considered, whereas those who received photons saw a profound depletion in naïve T cells, an increase in effector/memory populations, and a loss of TCR diversity. The repertoires of photon-treated patients underwent an oligoclonal expansion after their lymphocyte count nadirs, particularly of CD8+ Temra cells, driving this reduction in diversity. Across the entire cohort, this reduction in post-nadir diversity is inversely correlated with the overall survival time of those patients who died. CONCLUSION:This raises the possibility that increased adoption of proton-based or other lymphocyte-sparing radiotherapy regimes may lead to better survival in cancer patients.

KRAS is one of the most commonly mutated genes in human cancer, with the G12D mutation having the highest frequency, occurring in 33% and 28% of colorectal cancer (CRC) and pancreatic cancer (PDAC), respectively. KRAS G12D-containing peptides are presented by the common HLA allele HLA-A*11:01, making this cancer neoantigen a highly attractive target for the development of adoptive TCR-T cell therapy. However, the levels of cancer neoantigen presentation are typically low and combined with the immunosuppressive tumor microenvironment this compromises the activation and reactivity of TCR-engineered T cells, posing a major hurdle for these cells to eradicate tumor lesions. To establish a TCR-T cell therapy clinically intended for the treatment of solid cancers, including CRC and PDAC, we developed AZD0240, a CRISPR-engineered autologous TCR-T cell product targeting KRAS-G12D/HLA-A*11:01. To improve the reactivity of AZD0240 to the KRAS-G12D neoantigen, we developed a novel multi-armoring strategy, involving the combined disruption of a key negative regulator of T cell function and the introduction of a positive enhancement of TCR signaling. Employing this multi-armoring strategy, we observed enhanced cytokine production in responds to low levels of exogenously loaded KRAS-G12D peptide, as well as HLA-A*11:01+ tumor cells that express the KRAS-G12D mutation. Moreover, assessment of the in vitro and in vivo proliferative and cytotoxic capacity revealed that the AZD0240 armoring strategy can profoundly increase both the proliferative capacity and the anti-tumor efficacy of TCR edited T cells. Notably, the armored AZD0240 T cells were unable to respond to the KRAS G12 wild-type epitope, and no signs of increased TCR cross- or allo-reactivity, or malignant transformation were detected, derisking the clinical use of this armoring strategy. Taken together, AZD0240 is a KRAS-G12D/HLA-A*11:01 targeted TCR-T cell therapy that employs a novel multi-armoring strategy and is currently in preparation for clinical testing, with encouraging signs of pre-clinical efficacy and safety. Lianne Kok, Sander Eshuis, Xiangjun Kong, Jennifer Ma, Michelle Mojadidi, Paula Kroon, Vanessa Tubb, Ton Schumacher, Arianne Perez, Carsten Linnemann, Mark Cobbold, Jeroen van Heijst, Gavin Bendle. Pre-clinical characterization of AZD0240, a CRISPR-engineered autologous TCR-T cell product targeting KRAS-G12D/HLA-A*11:01 that is multi-armored to increase epitope sensitivity [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2025; Part 1 (Regular Abstracts); 2025 Apr 25-30; Chicago, IL. Philadelphia (PA): AACR; Cancer Res 2025;85(8_Suppl_1):Abstract nr 3478.

BACKGROUND:Immune checkpoint inhibitors have revolutionized the treatment of solid tumors, enhancing clinical outcomes by releasing T cells from inhibitory effects of receptors like programmed cell death protein 1 (PD-1). Despite these advancements, achieving durable antitumor responses remains challenging, often due to additional immunosuppressive mechanisms within the tumor microenvironment (TME). Tumor-associated macrophages (TAMs) contribute significantly to the immunosuppressive TME and play a pivotal role in shaping T cell-mediated antitumor responses. Leukocyte immunoglobulin-like receptor subfamily B member 2 (LILRB2), expressed on myeloid cells, including TAMs, is an inhibitory receptor, which contributes to macrophage-mediated immunosuppression. In this study, we present AZD2796, a high-affinity anti-LILRB2 antibody designed to repolarize TAMs from an immunosuppressive to a proinflammatory phenotype. METHODS:Anti-LILRB2 antibodies were identified using single-B-cell encapsulation Immune Replica technology. The ability of AZD2796 to enhance proinflammatory responses from macrophages treated with CD40 ligand or lipopolysaccharide was assessed using a macrophage stimulation assay. A tumor cell/macrophage/T cell co-culture assay was developed to evaluate the effect of AZD2796, as a single agent and in combination with an anti-PD-1 antibody, on the cytolytic activity of antigen-specific T cells. In vivo assessments were then carried out to determine the ability of AZD2796 to alter tumor growth rate in mice humanized with CD34 hematopoietic stem cells. RESULTS:In preclinical assessments, AZD2796 skewed macrophage differentiation away from an immunosuppressive phenotype and enhanced the proinflammatory function of macrophages. AZD2796 significantly increased the anti-tumor response of T cells following PD-1 checkpoint blockade, while AZD2796 monotherapy reduced tumor growth in humanized mouse models. CONCLUSIONS:These findings support the potential of AZD2796 as an anti-cancer therapy, with the ability to synergize with T-cell-based therapeutics.

Background AZD5863 is a bispecific T cell engager (TCE) with high affinity to CLDN18.2 and low affinity to cluster of differentiation 3 (CD3), designed to decrease its peripheral cytokine release potential, improve the therapeutic index, and maintain potent anti-tumor activity.Methods AZD5863 was evaluated using CLDN18.2-expressing human cell lines alone or in co-cultures with human or cynomolgus monkey peripheral blood mononuclear cells to determine affinities, specificity, potency, and bystander killing activity. In vivo, AZD5863-mediated tumor growth inhibition and pharmacodynamics were evaluated in humanized mice or human CD3 transgenic mice implanted with CLDN18.2-expressing cancer cell lines.Results AZD5863 was shown to bind specifically to human and cynomolgus monkey CLDN18.2 and to CD3, with CLDN18.2 binding also conserved against the murine protein. AZD5863 mediated T cell-dependent anti-tumor activity against CLDN18.2-expressing lines, with potency significantly correlating with CLDN18.2 receptor density. Cytokine secretion induced by AZD5863, in vitro and in vivo, was lower compared with a CLDN18.2 TCE with higher affinity for CD3. AZD5863 mediated T cell-dependent bystander killing of CLDN18.2-negative cells in the presence of CLDN18.2-expressing cells, in a mechanism partly dependent on interferon gamma (IFNγ), tumor necrosis factor alpha (TNFα), and Fas ligand. In vivo, AZD5863 treatment resulted in potent tumor control in pancreatic, gastric, and esophageal models and enhanced engraftment of immune populations in a humanized model.Conclusions AZD5863 mediates potent anti-tumor activity in vitro and in vivo, while inducing limited levels of cytokines. This work improves our understanding of the mechanism of action of affinity balanced TCEs and informs the design of a phase 1 trial testing AZD5863 in gastric, pancreatic, and esophageal adenocarcinoma (NCT06005493).

Introduction Non-Hodgkin lymphomas (NHL) include diverse neoplasms of the lymphoid compartment, with the majority originating from the B-cell lineage. Bispecific CD20xCD3 T cell engagers (TCEs), both monovalent and bivalent, have shown promising rates of ORRs and CRs in the clinic leading to several FDA approvals for the treatment of R/R DLBCL and FL. However, these treatments are still associated with significant toxicities, including CRS and ICANS, limiting their therapeutic window and potential for clinical combinations. Methods We engineered a first-in-class CD8-guided TCE, AZD5492, for B-NHL malignancies. AZD5492 is an asymmetric, trispecific monoclonal IgG1 antibody which harbors two Fab binding domains to CD20, one VHH binding domain to TCR, one VHH binding domain to CD8 co-receptor. AZD5492 binding, cytolytic activity, and efficacy were extensively interrogated in vitro and in vivo, in murine models and in non-human primates (NHP). Results The VHH domains of AZD5492 allow preferential engagement of CD8+ T cells through CD8/TCR binding, leading to the formation of an artificial immunological synapse with CD20+ target cells, T-cell activation and target B cell killing. Compared to conventional CD20xCD3 TCEs, which equally engage and activate CD4+ and CD8+ T cells, AZD5492 drives potent B cell killing through preferential engagement of CD8+ T cells, with reduced CD4+ T cell activation and associated cytokine production. In NSG humanized mice engrafted subcutaneously or intravenously with B cell tumors, AZD5492 conferred potent and dose-dependent anti-tumor efficacy. When compared to conventional bivalent CD20xCD3 TCEs, comparable anti-tumor efficacy was achieved with significantly less systemic cytokine production. Importantly, repeated dosing of AZD5492 was also evaluated in cynomolgus monkey for up to 1 month and induced a marked and prolonged decrease in CD19+ and CD20+ B cells in the blood and in tissues (lymph nodes, spleen, and bone marrow). Transient cytokine increases were observed post-first dose of AZD5492 but was not associated with notable toxicity. Overall, AZD5492 was well tolerated in cynomolgus monkeys and a NOAEL was established at 3 mg/kg. Conclusions AZD5492 represents a first-in-class T cell engager for the treatment of B cell lymphomas. Compared to conventional CD3-based T cell engagers, AZD5492 anti-tumor activity in murine models was associated with low cytokine release and this favorable safety profile was recapitulated in NHP. AZD5492 has therefore the potential to significantly improve therapeutic index. AZD5492 IND has been approved by the FDA and a FIH study in patients with R/R NHL and CLL is ongoing.

Prostate cancer is the second most common malignancy in males with a 5-year survival rate for metastatic castration-resistant prostate cancer of only 37%. Six-transmembrane epithelial antigen of the prostate 2 (STEAP2) is a metalloreductase with plasma membrane expression that is restricted to the prostate but is highly overexpressed across all stages of prostate adenocarcinoma. This expression profile makes STEAP2 an ideal target for a T cell engager (TCE). Despite promising efficacy, the toxicity profile of current TCEs remains challenging with patients often experiencing cytokine release syndrome and on-target, off-tumor adverse effects. This dose-limiting toxicity profile restricts their therapeutic window and with it the potential for combination therapy. We sought to improve this therapeutic index in 3 ways: 1) targeting the tumor-restricted STEAP2 antigen, 2) affinity-optimizing the CD3 binding domain and 3) integrating a CD8 binding domain to preferentially engage CD8+ T cells over CD4+ T cells, as the latter have been shown to contribute to the risk of cytokine release syndrome. Here we provide preclinical in vitro and in vivo evidence demonstrating that this design can potentially increase the therapeutic index when compared to a conventional TCE. We demonstrate that AZD6621 mediates potent, STEAP2-dependent cytotoxicity in vitro with cytotoxicity EC50 values below 10 pM. AZD6621 also induces preferential activation of CD8+ T cells over CD4+ T cells, with an approximately 5-fold increase in maximal activation of CD8+ T cells, as well as significantly reduced cytokine release compared to a conventional TCE. AZD6621 mediates complete anti-tumor regressions as a monotherapy in the STEAP2-expressing 22Rv1 subcutaneous xenograft model in humanized mice. Anti-tumor activity of AZD6621 in vivo was also observed in a C4-2 intratibial xenograft model. Furthermore, when matched for anti-tumor activity in an intravenous 22Rv1 xenograft model, systemic cytokine levels were significantly reduced in AZD6621-treated mice compared to a conventional TCE. These data provide reason to believe that AZD6621 can improve the therapeutic index of conventional TCEs via CD8-guided, affinity-optimized CD3 engagement of STEAP2 on prostate cancer cells and support a first-in-human clinical trial planned to start in 2025. Suzanne Sitnikova, Natalie Burrows, Nikolaos Ioannou, Even Walseng, Chunning Yang, Nadia Luheshi, Yariv Mazor, Mark Cobbold, Saso Cemerski, Simon Dovedi. AZD6621: Improving the therapeutic index for prostate cancer T cell engagers with a next-generation CD8-guided format [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2025; Part 1 (Regular Abstracts); 2025 Apr 25-30; Chicago, IL. Philadelphia (PA): AACR; Cancer Res 2025;85(8_Suppl_1):Abstract nr 3514.