PII-117 - EXPOSURE-RESPONSE OF TRASTUZUMAB DERUXTECAN (T-DXD) IN SUBJECTS WITH HER2-LOW METASTATIC BREAST CANCER

Z. Lu¹, H. Li², R. Wada², L. Li¹, E. Kamiyama³, P. Vaddady¹, M. Abutarif¹, T. Garimella¹, A. Khatri¹; ¹Quantitative Clinical Pharmacology, Daiichi-Sankyo, Inc, ²QuanTx Consulting, ³Quantitative Clinical Pharmacology, Daiichi Sankyo Co., Ltd..

Background: Exposure-response (ER) analysis of T-DXd, a HER2-targeting antibody-drug conjugate, with DXd, a topoisomerase I inhibitor payload, in HER2-low metastatic breast cancer (mBC) population was conducted.

Methods: Exposure-efficacy (EE) analysis was performed using data from HER2-low mBC subjects in a Phase 3 trial (N =362), DESTINY-Breast04 (DB-04). Exposure-safety (ES) analyses of T-DXd were performed using pooled data from 10 Phase 1-3 trials (N = 1675), including DB-04 and HER2-positive BC subjects.
Confirmed objective response (ORR), Grade3+ (G3+) adverse events, G3+ anemia, G3+ neutropenia, and G3+ thrombocytopenia were analyzed by logistic regression; progression-free survival (PFS), overall survival (OS), and AE of any grade and G3+ interstitial lung disease (ILD) were analyzed by time-to-event Cox regression. T-DXd or DXd exposure metrics and patient-specific covariates (e.g., demographics, disease and patient characteristics were tested.

Results: EE analysis identified no statistically significant relationships between T-DXd or DXd exposure and PFS (9.9-month median PFS) at the 5.4 mg/kg dose in DB-04. Shallow but significant (p < 0.01), relationships were observed between increasing T-DXd trough concentration and OS and ORR. Model-estimated OS probability at 360 days was 88% at the median exposure and changed by ±6% (82% to 93%) over the extremes of exposure (5th to 95th percentile [P]). The estimated ORR was 52% at median exposure and changed by ±14% (42% to 65.5%) over exposure range of 5th to 95th P.
ES analysis showed statistically significant relationships between T-DXd exposure and ILD and between DXd exposure and all other safety endpoints. Model-predicted AE rates were similar (within ±5%) for T-DXd 5.4 mg/kg dose between HER2-low and HER2-positive BC. Despite higher T-DXd or DXd exposure in pharmacokinetic analysis, high bodyweight subjects had similar predicted AE rates as other subjects.

Conclusion: Model-based ER analyses support 5.4 mg/kg T-DXd dose in HER2-low BC based on strong efficacy (9.9-month median PFS and OS probability of 88% at 1 year) and shallow EE relationships, and comparable AE rates for 5.4 mg/kg dose in HER2-low and HER2-positive BC.