(M-054) Preclinical data exposures underpredict clinical physiologically active doses for bispecific TCEs in solid tumors indications; are there better metrics? - insights from a small mechanistic MBMA

Dean Bottino, Ph. D. – Senior Scientific Director, Takeda Pharmaceuticals; Khem Ghusinga, Ph. D. – Senior Scientist, Applied BioMath / Certara; Scott Gruver, Ph. D. – Senior Scientist, Applied BioMath / Certara; Jess Wu, Ph. D. – Senior Scientist, Takeda Pharmaceuticals

Objectives:
We explore the utility of mechanistic modeling for translation of preclinical data to clinical pharmacologically active doses for CD3 bispecific biologics in solid tumor indications.

Methods: We gathered publicly available data from all CD3 bispecific molecules in development for solid tumor indications up to March 2023. The four solid-tumor-indication CD3 bispecific molecules that had enough data to conduct the study were: tebentafusp [1], tarlatamab [2], pasotuxizumab [3], and ERY974 [4]. The data on these TCEs included: in vitro cytotoxicity assay data, mouse TGI data, cyno or human PK data, as well as clinical trial results. Mechanistic PKPD models that focused on target expression, binding and trimer formation, and PK were constructed and outputs/metrics for translating preclinical results to clinical results for clinically tested physiologically active doses (PADs) were identified. Several metrics of translation were explored, and preclinical vs clinical outcome correlations were determined by using linear regression on a log-log scale.

Results: Three metrics gave good correlations: in vitro EC50 versus projected tumor concentration, in vitro projected trimers per target cell (TpT) ET50 versus human projected tumor TpT, and mouse TGI projected TpT versus human projected tumor TpT. The results indicate that both EC50 and mouse TGI TpT underpredict the PAD by two to three orders of magnitude, while in vitro ET50 overpredicts the PAD by an order of magnitude. Additionally, mouse TGI max TpT showed best correlation with human predictions.

Conclusions: The correlation results indicate that all three metrics have value, the mouse TGI max TpT suggesting a better relationship than the others. More importantly, the in vitro exposure metric was two to three orders of magnitude lower than the PAD exposure for human. That is, the typical approach of using in vitro EC10-EC50 to determine a clinical starting dose defines a much lower PAD than what is observed in practice, suggesting the need for a more aggressive preclinical to clinical translation approach. Exploration of more compounds is needed to strengthen these arguments.

Citations: [1] KIMMTRAK® (tebentafusp-tebn) label: https://www.accessdata.fda.gov/drugsatfda_docs/label/2022/761228s000lbl.pdf
[2] Ahn, M.-J.; Cho, B.C.; Felip, E.; Korantzis, I.; Ohashi, K.; Majem, M.; Juan-Vidal, O.; Handzhiev, S.; Izumi, H.; Lee, J.-S.; et al. Tarlatamab for Patients with Previously Treated Small-Cell Lung Cancer. N. Engl. J. Med. 2023, 389, 2063–2075
[3] Hummel HD, et al. Pasotuxizumab, a BiTE(®) immune therapy for castration-resistant prostate cancer: Phase I, dose-escalation study findings. Immunotherapy. 2021;13:125–141. doi: 10.2217/imt-2020-0256
[4] Safran et al., Abstract CT111: Results of a phase 1 dose escalation study of ERY974, an anti-glypican 3 (GPC3)/CD3 bispecific antibody, in patients with advanced solid tumors. Cancer Res 1 July 2021; 81 (13_Supplement): CT111. https://doi.org/10.1158/1538-7445.AM2021-CT111