Taewook Sung1, Jihyun Jeon1, Sung Young Lee2, SuYeon Kim2, Seongyeon Kang2, Sung Ho Kim2, Heung Tae Kim2, Soyoung Lee1*, Jung-woo Chae1,3*, Hwi-yeol Yun1,3*
1College of Pharmacy, Chungnam National University, Daejeon, Republic of Korea 2ImmuneOncia Therapeutics, Inc., Seoul, Republic of Korea 3Department of Bio-AI convergence, Chungnam National University, Daejeon, Republic of Korea *Those of authors contributed equally as correspondence.
Objectives: The emergence of anti-programmed cell death ligand 1 (PD-L1) inhibitors marks a sophisticated approach in cancer therapeutics, strategically addressing immune evasion mechanisms within tumors. However, immune-related adverse events have been reported in the clinical field. IMC-001 is a fully human PD-L1 recombinant monoclonal antibody that strongly binds to PD-L1 to inhibit its binding to programmed cell death protein 1 (PD-1) or B7-1 (CD80). IMC-001 showed robust dose-dependent efficacy and no evidence of toxicity ranging from 1 mg/kg to 10 mg/kg in animal models without toxicity in cynomolgus monkeys. Also, in Phase 1 clinical trial, which is dose escalation study ranging 2.5 mg/kg to 20 mg/kg, IMC-001 showed no dose-limiting toxicity (DLT). This study aims to optimize IMC-001 dosage regimen through population pharmacokinetic (PK) modeling to minimize adverse effects in human cancer treatment.
Methods: To optimize the dosage regimen for IMC-001, PK modeling was conducted using the data from Phase 1 and Phase 2 Clinical trial. Phase 1 involved dose escalation (2.5, 5, 10, 15, 20 mg/kg) with five patients per group, while Phase 2 which focused exclusively on the 20 mg/kg dosage level with a cohort of 22 patients, both phases comprising multiple-dose, 1-hour IV infusions. Target-Mediated Drug Disposition (TMDD) modeling, incorporating FcRn recycling, was utilized, with quasi-steady-state (QSS) approximation simplifying equations. Covariate searching was performed to identify relevant factors. PK simulations compared PK parameters of AUCtau, Cmax, and Ctrough across Q2W, Q4W, Q6W, and Q8W dosage regimens. The simulations were conducted multiple dosage, with administration repeated six times to achieve steady state. These simulations were performed using NONMEM 7.5.
Results: The final model incorporates the FcRn recycling compartment within the TMDD modeling framework. The covariate search results indicated that no statistically significant covariates were identified, consequently precluding their inclusion in the final model. To incorporate physiological factors, four parameters were fixed. The dissociation constant between PD-L1 and IMC-001 was set at 0.00155 (μmol/L), while the dissociation constant between FcRn and IMC-001 was established at 71.3 (μmol/L). The total concentration of FcRn was determined to be 0.291 (μmol/L), and the degradation for the IMC-001 and PD-L1 complex was fixed at 0.225 (1/h). The estimated parameters were also feasible to clinical trial data. Using this final model, comparisons between Q2W and other intervals were conducted based on AUCtau, Cmax, and Ctrough. We assessed other dose intervals to achieve comparable systemic exposure to the 20 mg/kg Q2W regimen, which were the effective dose in the Phase 2 trial. The simulated mean values were as follows: for AUCtau, the values for the Q2W, Q4W, Q6W, and Q8W IV 20 mg/kg regimens were 81598.5, 82426.2, 85476.7, and 86093.8 (ng*h/mL), respectively. For Cmax, the values were 582.3, 491.2, 470.2, and 465.9 (ng/mL). Ctrough values were 130.8, 46.9, 25.3, and 15.9 (ng/mL), respectively. Among various simulations, compared to Q2W IV 20 mg/kg regimen, other regimens showed similar AUCtau, while having lower peak level of IMC-001. Although simulated trough level was also lower in other intervals, the efficacy of IMC-001 is anticipated to be achieved by systemic exposure. Therefore, IMC-001 20 mg/kg Q2W regimen could be replaced by Q4W, Q6W and Q8W regimens with lower risk of having pharmacokinetic related safety issues due to higher peak level.
Conclusions: The PK model derived from this study demonstrates accuracy in forecasting the pharmacokinetic profile of IMC-001. Transitioning from a Q2W regimen to Q4W, Q6W, or Q8W regimens is predicted to offer a means of mitigating pharmacokinetic-related safety issues while preserving efficacy. Subsequent research aimed at linking pharmacodynamic models to the PK model could bolster our simulation findings, thereby providing insights for formulating an optimal regimen targeted at minimizing adverse effects.
References:
[1] Weize H et al. (2022). Development of pediatric physiologically-based pharmacokinetic model to support recommended dosing of atezolizumab in children with solid tumors.
[2] Yao-Yun F et al. (2019). Human FcRn Tissue Expression Profile and Half-Life in PBMCs.
[3] Kim JH et al. (2007). Kinetics of FcRn-mediated recycling of IgG and albumin in human: Pathophysiology and therapeutic implications using a simplified mechanism-based model.
[4] Chetty M et al. (2014). Prediction of the Pharmacokinetics, Pharmacodynamics, and Efficacy of a Monoclonal Antibody, Using a Physiologically Based Pharmacokinetic FcRn Model.
[5] Ilean C et al. (2022). Effects of small molecule‑induced dimerization on the programmed death ligand 1 protein life cycle.
Reference: PAGE 32 (2024) Abstr 10927 [www.page-meeting.org/?abstract=10927]
Poster: Drug/Disease Modelling - Oncology