Selective intratumoral distribution and post-T-DXd activity of K-679, an ultra-high-DAR EGFR-targeted antibody drug-loaded unimicelle conjugate (ADUC)

Background: K-679 is a compact, 20-nm EGFR-targeted antibody drug-loaded unimicelle (unimer-micelle) conjugate (ADUC) utilizing single-chain polymer nanoparticles that carry DM1 at an ultra‑high-drug-to-antibody ratio (DAR, ~50) while maintaining colloidal stability at low antibody mass. This design aims to enhance intratumoral penetration and tumor-selective payload exposure even when antigen density is low and/or heterogeneous. While antibody-drug conjugates (ADCs) have improved cancer therapy, there remains an unmet need for patients progressing after ADC treatment, particularly following trastuzumab deruxtecan (T-DXd). We investigated whether the ADUC design of K-679 enables tumor-selective pharmacokinetics (PK), extensive intratumoral distribution, spatial pharmacodynamics, and activity following prior T-DXd treatment. Methods: K-679 versus a benchmark ADC (cetuximab‑DM1 ADC, cleavable disulfide linker, DAR ~4.5) was evaluated in head-to-head comparisons using EGFR-expressing CDX (HT-29, SK‑OV-3, SK-CO-1 [KRAS G13V]) evaluated at DM1-equivalent dosing. Total DM1 (all species, measured after deconjugation) and free DM1 levels in tumor and plasma were quantified by LC‑MS/MS. At 48 h, tumor sections were immunostained for human IgG (to map conjugate distribution), CD31 (microvasculature), pHH3 (mitotic arrest), and cleaved PARP (apoptosis). In HT-29 colorectal CDX with HER2-low/EGFR-low expression, animals received T-DXd, then on Day 18 were switched to K-679 or repeat T-DXd for 3 weeks. A parallel arm received K-679 on Day 0 and Day 18. Colorectal PDX harboring KRAS G13D (EGFR IHC score 1+, patchy) were also evaluated. Results: Across CDX models at DM1-equivalent dosing, K‑679 delivered 12.7-13.7-fold higher intratumoral total DM1 and 2.3‑3.5‑fold higher intratumoral free DM1 while reducing plasma free DM1 by 4.2-8.4-fold versus the benchmark ADC, indicating tumor-selective payload release with reduced systemic deconjugation. At 48 h, K-679's IgG signal exceeded benchmark ADC despite at least an 11-fold lower antibody mass and extended into regions distant from CD31-positive vessels; pHH3 and cleaved PARP increased across the tumor, indicating mitotic arrest/apoptosis. After prior T-DXd in HT-29 xenografts, K‑679 induced robust tumor regressions, whereas repeat T-DXd did not. Consistently, two K-679 doses induced complete regressions (6/6) by Day 22, sustained to Day 39. In EGFR-low and heterogeneous colorectal PDX, a single K-679 dose inhibited tumor growth over 3 weeks. Conclusions: The ultra-high-DAR ADUC design of K-679 confers tumor-selective PK, extensive intratumoral distribution, concordant spatial pharmacodynamics, and activity after prior T-DXd. These data support clinical development for EGFR-expressing solid tumors with low/heterogeneous antigen density and post-T-DXd settings.