HER2+ Syngeneic Murine Models for Brain Metastasis and CNS Drug Testing

Investigators report engineering immunocompetent, syngeneic murine mammary carcinoma models to express human HER2 (wild-type and the exon 20 insertion HER2YVMA) and to track spontaneous brain metastasis using a GFP tag. Using both 4T1 (BALB/c background) and EO771 (C57BL/6 background) backbones, the paper describes a platform for model development and characterization alongside therapy-testing readouts intended for brain-metastasis quantification. The authors present the resulting HER2+ syngeneic murine models as enabling side-by-side comparisons of tumor behavior and CNS endpoint measurements under a shared experimental workflow. The report then spans model construction, imaging methodology, and treatment response patterns as complementary elements of that workflow.

To generate the lines, the authors describe lentiviral transduction of parental 4T1 and EO771 cells with constructs encoding human HER2WT-GFP or HER2YVMA-GFP, followed by puromycin selection and GFP-based identification of HER2-expressing cells. They report validating HER2 surface localization by immunofluorescence and confirming HER2 protein expression by Western blot, with higher HER2 phosphorylation in HER2YVMA versus HER2WT. In vivo, they observed divergent implantation behavior: 4T1-HER2 tumors had 100% take, whereas EO771-HER2 tumors showed lower take rates (15–30%); the investigators addressed this by excising established EO771 tumors and transplanting small tumor pieces to achieve consistent implantation. For spontaneous brain-metastasis quantification, brains were harvested at day 33 after tumor cell injection for ex vivo imaging. In aggregate, the authors frame the two syngeneic backbones as showing distinct growth and CNS-dissemination behavior under their protocol.

Brain metastasis detection was reported as an ex vivo fluorescence endpoint based on GFP signal from harvested whole brains imaged on an IVIS platform with the skull removed. The acquisition parameters were specified as excitation 480 nm and emission 520 nm, using a 1-second exposure time, binning factor of 4, and a 12.5 cm field of view, with fluorescence quantified in Living Image software to support cross-sample comparability. The authors also note practical constraints of GFP-based readouts, describing that fluorescence can underestimate micrometastatic burden or miss deeper parenchymal lesions because of signal attenuation, and they mention outlier handling in quantitative analysis via a Grubbs test in their dataset. Together, these details outline a standardized endpoint strategy for capturing CNS dissemination in the presented models.

When trastuzumab was evaluated as monotherapy in 4T1 models, the authors report no significant reduction in primary tumor volume, alongside a reduction in whole-brain GFP signal consistent with lower metastatic burden by their imaging metric. They quantify the brain-signal change as a 17% ± 8% decrease in 4T1-HER2WT and a 26% ± 7% decrease in 4T1-HER2YVMA relative to saline controls. In separate combination experiments, they report that trastuzumab plus tucatinib combined with immune checkpoint blockade (anti–PD-1 and anti–CTLA-4) suppressed primary tumor growth, increased tumor-infiltrating CD4+ T cells, and prolonged survival in EO771-HER2YVMA, whereas similar benefits were not observed in 4T1-HER2YVMA under the conditions tested. The authors interpret these patterns as illustrating heterogeneous therapeutic responsiveness across immunocompetent model contexts.

Antibody–drug conjugate experiments were presented as a contrast between activity in human xenografts and more limited responses in the newly engineered syngeneic models. In the xenograft setting, the authors report that trastuzumab deruxtecan drove complete regression in high-HER2 BT474 tumors, while trastuzumab emtansine showed a more limited activity pattern centered on high-HER2 disease and no efficacy in the HER2− MDA-MB-231 model where T-Dxd reduced tumor volume. In the syngeneic 4T1/EO771 HER2-overexpressing models, they describe limited or variable activity at the tested regimen (10 mg/kg IV on days 0 and 7) in small cohorts (n=3), including a partial single-mouse response in EO771-HER2WT with T-Dxd and no observed efficacy with T-DM1 in that setting. The paper’s discussion links these findings to interpretive caveats the authors articulate, including that HER2 surface levels alone did not appear to predict ADC efficacy and that pharmacokinetics, blood–brain barrier exposure, dose optimization, and larger cohorts were considerations for translational interpretation of T-Dxd and T-DM1 testing in CNS-relevant preclinical workflows.

Key Takeaways:

• The authors describe 4T1 and EO771 syngeneic murine mammary carcinoma lines engineered to express human HER2WT or HER2YVMA with GFP, with spontaneous CNS dissemination read out by ex vivo whole-brain IVIS fluorescence imaging at a defined harvest timepoint.
• Reported responses differed by model: in the 4T1 models, the study reported no reduction in primary tumor size with trastuzumab monotherapy, alongside a decrease in whole-brain GFP signal as measured by their imaging assay; meanwhile, trastuzumab+tucatinib plus anti–PD-1/anti–CTLA-4 was reported to show immune and survival-associated effects in EO771-HER2YVMA but not in 4T1-HER2YVMA.
• Across ADC experiments, T-Dxd was reported to show broader activity than T-DM1 in xenografts, whereas syngeneic 4T1/EO771 testing showed limited/variable activity at the stated dosing and small cohort size, alongside author-noted PK/BBB and dose/cohort-size caveats.