Purpose: Siltuximab is a monoclonal antibody that binds to interleukin (IL)-6 with high affinity and specificity; C-reactive protein (CRP) is an acute-phase protein induced by IL-6. CRP suppression is an indirect measurement of IL-6 activity. Here, modeling and simulation of the pharmacokinetic (PK)/pharmacodynamic (PD) relationship between siltuximab and CRP were used to support dose selection for multicentric Castleman's disease (CD).
Methods: PK/PD modeling was applied to explore the relationship between siltuximab PK and CRP suppression following intravenous siltuximab infusion in 47 patients with B cell non-Hodgkin's lymphoma (n = 17), multiple myeloma (n = 13), or CD (n = 17). Siltuximab was administered as 2.8, 5.5, or 11 mg/kg q2wks, 11 mg/kg q3wks, or 5.5 mg/kg weekly. Simulations of studied or hypothetical siltuximab dosage regimens (15 mg/kg q4wks) were also performed to evaluate maintenance of CRP suppression below the cutoff value of 1 mg/L.
Results: A two-compartment PK model and an inhibitory indirect response PD model adequately described the serum siltuximab and CRP concentration-time profiles simultaneously. PD parameter estimates were physiologically plausible. For all disease types, simulations showed that 11 mg/kg q3wks or 15 mg/kg q4wks would reduce serum CRP to below 1 mg/L after the second dose and throughout the treatment period.
Conclusions: PK/PD modeling was used to select doses for further development of siltuximab in multicentric CD. The dosing recommendation was also supported by the observed efficacy dose-response relationship. CRP suppression in the subsequent randomized multicentric CD study was in agreement with the modeling predictions.