Pharmacokinetic-pharmacodynamic analysis of myelosuppression with raludotatug deruxtecan (R-DXd) in patients with advanced solid tumors

Objectives: Raludotatug deruxtecan (R-DXd) is a novel cadherin 6-directed antibody–drug conjugate (ADC) under clinical investigation in patients with solid tumors (DS6000-A-U101, NCT04707248) and in patients with platinum-resistant ovarian cancer (REJOICE-Ovarian01, NCT06161025). Consistent with the safety profiles of other DXd ADCs with a drug-to-antibody ratio of 8, [1,2] transient hematologic toxicities, including anemia, neutropenia, and thrombocytopenia, were among the most frequently reported treatment-emergent adverse events (AEs) for R-DXd; frequency and severity of events occurred in a dose-dependent manner. Here, we report preliminary results of the utilization of population pharmacokinetic-pharmacodynamic (PK-PD) models to characterize the dynamic changes in platelet and neutrophil counts in response to different dosing regimens of R-DXd and the use of these models to estimate the probability of Grade 3–4 hematologic AEs by dose.Methods: Time-course data of pharmacokinetics (ADC and released payload [DXd] plasma concentration), platelet, and absolute neutrophil counts were integrated from patients treated with R-DXd monotherapy at 1.6–9.6 mg/kg Q3W IV in the ongoing first-in-human Phase 1 study (DS6000-A-U101). The analysis dataset comprised 129 patients enrolled in the dose-escalation and dose-expansion phases of DS6000-A-U101 as of March 20, 2024; the majority of patients had serous ovarian cancer (N&#3f106, 82.2%) and received R-DXd at the target doses used in the ongoing Phase 2/3 study (4.8, 5.6, and 6.4 mg/kg Q3W). Two semi-mechanistic PK-PD models adapted from the literature [3,4] were built to establish a quantitative relationship between plasma payload concentration released from R-DXd and neutrophil/platelet count changes, utilizing nonlinear mixed-effect modeling. Lastly, model-based simulations were performed to estimate the incidence of Grade 3–4 hematologic events within the first 18 weeks of R-DXd treatment (6 cycles) across dose levels. Results: The PK-PD models could adequately characterize changes in platelet and neutrophil counts as a function of drug concentration in patients who received repeated 21-day cycles of R-DXd within a wide range of doses. Of note, the models accurately captured (i) the platelet/neutrophil initial nadir occurring approximately 1–2 weeks from the first dose; (ii) the slow downward drift in platelet counts over multiple cycles observed in some patients; (iii) the spontaneous recovery of cell counts as a result of dose reduction or interruption. The model predictions demonstrated a positive exposure–response relationship for Grade 3–4 hematologic AEs versus payload plasma concentration. Conclusions: The models have proven helpful in understanding the dynamics and dose-dependency of hematologic toxicity with R-DXd monotherapy. Their application can be further expanded to combination therapy to inform R-DXd dosing strategies with agents exhibiting overlapping toxicity.Citations: [1] Guo Z, et al. J Clin Pharm Ther. 2022;47(11):1837–1844[2] Jänne PA, et al. Cancer Discov. 2022;12(1):74–89 [3] Bender BC, et al. Cancer Chemother Pharmacol. 2012;70(4):591–601[4] Quartino AL, et al. Invest New Drugs. 2012;30(1):833–845.