Daichi Yamaguchi: No financial relationships to disclose
Objectives: S-531011, a humanized monoclonal antibody against human C-C motif chemokine receptor 8 (CCR8), is anticipated to deplete CCR8-positive tumor-infiltrating regulatory T cells, leading to a reinvigoration of antitumor immunity. CCR8 receptor occupancy (RO) by S-531011 in tumor tissues would provide the useful information to determine the optimal dose; however, RO in human tumor tissues cannot be directly measured. In this study, a quantitative systems pharmacology model was developed to explore the relationship between pharmacokinetics (PK) and RO (PK/RO model) of S-531011 for humans using our on-going clinical trial data, non-clinical data, and physiological information. Subsequently, the RO values in tumor tissues in humans were simulated using the PK/RO model to investigate the effective doses.
Methods: We developed PK/RO model of S-531011. The model consists of three compartments (central, peripheral, and tumor compartments), with S-531011, CCR8, and S-531011-bound CCR8 included in each compartment. A two-site sequential model was assumed where S-531011 has two binding sites to CCR8. S-531011 combines with CCR8 reversibly to form complex 1 (S-531011:CCR8 = 1:1), and complex 1 combines further with CCR8 reversibly to form complex 2 (S-531011:CCR8 = 1:2). Model parameters (e.g. binding affinity [Kd] and degradation rate constant of CCR8) in humans were set using the results of non-clinical studies, the estimation from the clinical data (NCT05101070), and published information. In addition, the mouse PK/RO model was developed using the data from non-clinical studies, and the predicted RO was compared with observed RO in tumors to confirm the validity of the modeling and simulation for human tumor RO. The sensitivity analysis was conducted to identify influential parameters on tumor RO and refine the model. The time courses of RO values in the tumor compartment were simulated using human PK/RO model with variabilities for sensitive parameters. As a conservative condition, lower-than-expected distribution of S-531011 from the central to tumor compartment was also tested with lower the perfusion coefficient (Pdist). These analyses were performed using MATLAB (SimBiology).
Results: The PK/RO models were developed for each of human and mouse. The simulated RO values 24 hours after the administration in tumor compartment from mouse PK/RO model were >80% in 0.05 mg/kg and almost 100% in 0.5 mg/kg, and these values were higher than the observed tumor RO values in mice. The Kd of S-531011 against human CCR8-expressing cells was identified as the sensitive parameter. The Kd value was optimized based on fitness to tumor RO values in mice, and the optimized Kd value was applied to human PK/RO model. It was predicted that the administration of S-531011 80 mg every 3 weeks (Q3W) could maintain >90% of RO in the tumor compartment. When we set Pdist as 1/10 of final model, the administration of S-531011 800 mg Q3W could maintain >90% of RO.
Conclusions: The human PK/RO model of S-531011 was developed using our on-going clinical trial data, non-clinical data, and published literatures. Based on the RO simulations using this model, it was suggested that >90% of RO in tumor tissues could be maintained in the dose range of 80 to 800 mg of S-531011 Q3W.
