Gene targeting strategy for B-NDG MHC I/II DKO mice plus. The murine B2m and H2-Ab1 gene were knocked out while a fused gene composed of murine B2m and Fcgrt gene was inserted after the signal peptide sequence of murine Fcgrt gene in B-NDG MHC I/II DKO mice plus.

• B-NDG MHC I/II DKO Mice Plus provide a GvHD-reduced immunodeficient model for human PBMC engraftment studies.

Comparison of GvHD severity after human PBMC engraftment in B-NDG mice, B-NDG B2m KO mice plus, and B-NDG MHC I/II DKO mice plus. Five-week-old female B-NDG mice, B-NDG B2m KO mice plus, and B-NDG MHC I/II DKO mice plus were irradiated with 1.0 Gy and intravenously engrafted with human PBMCs from three healthy donors on day 0 (5×10⁶ cells/mouse, n=6). Survival, body weight, and clinical GvHD scores were monitored. Compared with B-NDG mice and B-NDG B2m KO mice plus, B-NDG MHC I/II DKO mice plus showed prolonged survival, improved body-weight maintenance, and reduced GvHD scores across donors. Values are expressed as mean ± SEM.

• B-NDG MHC I/II DKO Mice Plus can extend the study window for human PBMC engraftment studies.

B-NDG MHC I/II DKO Mice Plus supports an extended experimental window after human PBMC engraftment. B-NDG MHC I/II DKO mice plus were intravenously engrafted with human PBMCs on day 0 (1×10⁷ cells/mouse, n=6). Survival, body weight, and clinical GvHD scores were monitored over time. Mice survived up to 96 days after engraftment and maintained body weight for an extended period. Clinical GvHD scores increased gradually but remained moderate, with no obvious GvHD-related clinical symptoms other than body-weight loss during most of the observation period. Values are expressed as mean ± SEM..

• Sustained human immune-cell reconstitution was maintained for up to 16 weeks, dominated by human T cells.

Human PBMCs were successfully reconstituted in B-NDG MHC I/II DKO mice plus. Mice were intravenously engrafted with human PBMCs on day 0 (1×10⁷ cells/mouse, n=6). Peripheral blood was collected at the indicated time until week 16. (A) Frequency and (B) absolute numbers of reconstituted human immune-cell subsets. Human CD45+ cells increased from week 2 and remained detectable through week 16. Human T cells were the dominant population, with CD4+ T cells, CD8+ T cells, Tregs, and DCs subsets also detected. Values are expressed as mean ± SEM.

• Anti-hCD3/hHER2 BsAb analog showed strong antitumor efficacy in the huPBMC-engrafted NCI-N87 CDX model.

Efficacy evaluation of anti-hCD3/hHER2 bispecific antibody in a huPBMC-engrafted NCI-N87 CDX model using B-NDG MHC I/II DKO mice plus. NCI-N87 human gastric cancer cells were subcutaneously inoculated into female B-NDG MHC I/II DKO mice plus. Human PBMCs were intravenously engrafted after tumor inoculation. When tumor volume reached approximately 100 mm³, mice were randomized and treated intraperitoneally with control antibody, trastuzumab analog, or anti-hCD3/hHER2 BsAb analog. (A) Tumor volume. (B) Body weight. (C) Frequency of human CD45+ cells in peripheral blood at the endpoint. Anti-hCD3/hHER2 bispecific antibody markedly inhibited tumor growth, while body weight was maintained during treatment. These results demonstrate that huPBMC-engrafted B-NDG MHC I/II DKO mice plus can be used to evaluate T cell–engaging bispecific antibody efficacy in CDX models. Values are expressed as mean ± SEM.

• AMG509 showed strong antitumor efficacy in the huPBMC-engrafted 22Rv1 CDX model.

Tumor growth, body weight change, and terminal tumor weight in 22Rv1 tumor-bearing huPBMC-B-NDG MHC I/II DKO mice plus. Female B-NDG MHC I/II DKO mice plus were first engrafted with human PBMCs (1×10⁷), followed 8 days later by subcutaneous inoculation of 22Rv1 cells (5×10⁶). When tumors reached 80–120 mm³, mice were randomized and treated intraperitoneally with PBS or AMG509 analog at 0.01 or 0.1 mg/kg once weekly for three doses (n = 6). AMG509 analog at 0.1 mg/kg significantly inhibited tumor growth and reduced terminal tumor weight, while 0.01 mg/kg showed no meaningful antitumor activity. No obvious body-weight toxicity was observed at 0.1 mg/kg. Values are expressed as mean ± SEM. Significance was determined by one-way ANOVA test. *P < 0.05, **P < 0.01, ***P < 0.001.

• AMG509 analog at the efficacious dose increased human IL-2, indicating T-cell activation, without inducing broad CRS-like cytokine elevation.

Serum cytokine profiling after AMG509 analog treatment in 22Rv1 tumor-bearing huPBMC-engrafted B-NDG MHC I/II DKO mice plus. Serum cytokines were measured 4 hours after dosing with PBS or AMG509 analog. AMG509 analog at 0.1 mg/kg significantly increased human IL-2, indicating T-cell activation, while IFNγ, IL-6, IL-10 and TNFα were not broadly elevated at this efficacious dose. These results suggest no overt systemic CRS-like cytokine response under the tested conditions. Values are expressed as mean ± SEM.

• AMG509 analog increased the frequencies of Ki-67+ and TIM-3+ human T-cell subsets.

The peripheral blood human immune-cell profiling in the AMG509 analog efficacy study using 22Rv1 tumor-bearing B-NDG MHC I/II DKO mice plus. Peripheral blood was collected at the efficacy-study endpoint, 5 weeks after human PBMC engraftment, and analyzed by flow cytometry. Human CD45+ leukocytes were detected in all groups, and CD3+ T cells were the dominant human immune-cell population. The T-cell compartment showed a CD8+ T-cell–dominant profile, with relatively low CD4+ T-cell and Treg frequencies. AMG509 analog at 0.1 mg/kg was associated with increased Ki-67+ and TIM-3+ T-cell subsets, particularly in CD8+ T cells, indicating enhanced peripheral T-cell proliferation and checkpoint/exhaustion-associated activation. Values are expressed as mean ± SEM.

• Ki-67+, PD-1+ and TIM-3+ T-cell subsets were detectable, supporting the presence of proliferating and activated/checkpoint-associated human T cells in peripheral blood during antitumor efficacy evaluation.

The absolute number analysis of peripheral blood human immune-cell subsets in the AMG509 analog efficacy study using 22Rv1 tumor-bearing B-NDG MHC I/II DKO mice plus. Peripheral blood was collected at the efficacy-study endpoint, 5 weeks after human PBMC engraftment, and analyzed by flow cytometry. Human CD45+ leukocytes and CD3+ T cells were detected in all groups, confirming successful human PBMC reconstitution. The human T-cell compartment was mainly driven by CD8+ T cells, while CD4+ T cells and Tregs remained relatively low. Values are expressed as mean ± SEM.

• AMG509 analog increased the frequencies of Ki-67+ and TIM-3+ T-cell subsets, indicating enhanced intratumoral T-cell proliferation and activation/checkpoint-associated phenotypes.

Human tumor-infiltrating lymphocytes in 22Rv1 tumor-bearing huPBMC-B-NDG MHC I/II DKO mice plus were analyzed by flow cytometry at the study endpoint. Tumor tissues were collected at the efficacy-study endpoint and analyzed by flow cytometry. Human CD45+ immune cells were detectable in tumors, and CD3+ T cells were the predominant human immune-cell population. AMG509 analog at 0.1 mg/kg did not markedly increase total T-cell infiltration, but increased Ki-67+ and TIM-3+ cells within CD3+, CD4+ and CD8+ T-cell subsets, with increased CD69+ and PD-1+ subsets also observed. Values are expressed as mean ± SEM. Significance was determined by one-way ANOVA test. *P < 0.05, **P < 0.01, ***P < 0.001.

• No overt AMG509 analog–specific severe organ toxicity was observed in liver, spleen, or lung tissues.

Organ weight and H&E histopathology in the AMG509 analog efficacy study using 22Rv1 tumor-bearing huPBMC-engrafted B-NDG MHC I/II DKO mice plus. Liver, spleen, and lung tissues were collected for analysis. (A) Organ weights. (B) Representative H&E staining images. Liver weight was slightly increased in AMG509 analog–treated groups, while spleen and lung weights showed no obvious treatment-related changes. Inflammatory cell infiltration was observed in liver and lung tissues, as indicated by red arrows. Similar findings were also present in PBS controls, suggesting these changes were mainly related to the huPBMC-engrafted model background rather than AMG509 analog–specific organ toxicity. Values are expressed as mean ± SEM. Significance was determined by one-way ANOVA test. *P < 0.05, **P < 0.01, ***P < 0.001.

• High-dose AMG509 analog induced a transient decrease in rectal temperature, suggesting a CRS-like response under high-dose treatment.

CRS-like tolerability assessment of AMG509 analog in 22Rv1 tumor-bearing huPBMC-engrafted B-NDG MHC I/II DKO mice plus. Mice were treated intravenously with PBS or AMG509 analog at 0.1, 1, or 10 mg/kg as a single dose (n=3). Rectal temperature was monitored before and after dosing. AMG509 analog at 10 mg/kg induced a transient decrease in rectal temperature at 6 hours, suggesting a high-dose CRS-like response. Values are expressed as mean ± SEM.

• AMG509 analog induced acute cytokine responses, including increased IFNγ, IL2, IL10, and TNFα, supporting the use of B-NDG MHC I/II DKO mice plus for CRS-like cytokine response assessment.

Serum cytokine profiling after single-dose AMG509 analog treatment in 22Rv1 tumor-bearing huPBMC-B-NDG MHC I/II DKO mice plus. Mice were treated intravenously with PBS or AMG509 analog at 0.1, 1, or 10 mg/kg, and serum cytokines were measured at 2, 6, and 24 hours after dosing. AMG509 analog induced acute cytokine responses, including increased IFNγ, IL2, IL10, and TNFα, mainly at 1 and/or 10 mg/kg. IL2 peaked early and declined by 24 hours, whereas IFNγ remained elevated at 24 hours in high-dose groups. IL6 did not show a consistent dose-dependent increase. Values are expressed as mean ± SEM.