Use of Eltrombopag Exposure-Platelet Response Relationship for Dose Optimization in Patients with Chronic HCV-Infection with and without Interferon
Jianping Zhang, Daphne D. Williams, Katy P. Moore
GlaxoSmithKline, Research Triangle Park, North Carolina, USA
Objectives: Eltrombopag is the first oral small-molecule, non-peptide thrombopoietin receptor agonist. It is under development for the treatment of thrombocytopenia in patients with hepatitis C virus (HCV). A population pharmacokinetic/pharmacodynamic (PK/PD) model was developed to characterize the effect of eltrombopag on platelet counts in HCV patients with and without IFN (peginterferon alfa-2a and peginterferon alfa-2b) and ribavirin treatment and to identify covariates that influence eltrombopag PK and PD. The final PK/PD model was used to guide the dose selection for Phase III studies in HCV patients to achieve and maintain sufficient platelet counts to initiate and complete the full course of IFN therapy.
Methods: PK/PD data were constructed from 3 studies including 63 healthy subjects, 41 subjects with HCV, and 24 subjects with mild to severe hepatic impairment (HI). Dosing of eltrombopag in these studies ranged from 5 to 75 mg QD as a single dose or repeat doses for up to 16 weeks. Nonlinear mixed effects modeling was conducted using NONMEM V. Influential covariates were identified using step-wise forward addition and backward elimination technique. A sequential analysis approach was applied to characterize the PK/PD relationship between plasma eltrombopag concentrations and platelet counts in patients with HCV. Visual predictive check and non-parametric bootstrap were implemented for final model evaluation. Platelet response (>50, >75, >90, >200 Gi/L) at daily eltrombopag doses of 12.5, 25, 50, 75 and 100 mg with and without IFN were simulated for HCV patients using the final population PK/PD model.
Results: Eltrombopag PK was described by a two-compartment linear model with sequential zero-order and bolus input in absorption compartment and first-order absorption. The typical oral clearance (CL/F) and volume of distribution (Vc/F) of eltrombopag were 0.71 L/hr and 8.72 L, respectively, for a healthy subject. Vc/F correlated with body weight, consistent with the theoretical allometric scaling for volume of distribution. Age and aspartate aminotransferase (AST) correlated negatively with CL/F. Subjects with severe HI displayed a 46% reduction in rate of absorption (Ka). Overall, CL/F was 66% and 18% lower in subjects with HCV and HI, respectively, when compared to healthy subjects. Concomitant IFN had no effect on eltrombopag PK. The relationship between plasma eltrombopag concentrations and platelet response was well captured by a catenary cell production and loss model, with an estimated maximum stimulation (Smax) of 6.4 fold increase in platelet production and 50% of maximum stimulation achieved at 17.2 µg/mL (SC50). The model also characterized the inhibitory effect of IFN on platelet counts. PK/PD simulations supported 50 mg QD as an appropriate starting regimen for patients with HCV at which >75% of patients should achieve >75 Gi/L platelet counts after 2 weeks of eltrombopag alone treatment. Following initiation of IFN therapy, platelet counts were reduced by 40% from peak levels. At steady state, ³78% of patients who initiated IFN therapy were predicted to maintain platelet counts above 50 Gi/L over the course of IFN therapy with concomitant eltrombopag. A 25mg dose increase in patients with insufficient platelet response was predicted to improve their platelet counts. Following the dose adjustment, >70% of patients who did not respond (platelet counts <75 Gi/L) initially were predicted to respond, and >25% of patients who would have stopped IFN therapy due to reduction of platelet counts (<50 Gi/L) were predicted to complete the full course of IFN therapy.
Conclusions: The relationship between eltrombopag exposure and platelet response in HCV patients with and without IFN was well characterized by the population PK/PD model. Simulations based on the model optimized dosing strategies for the Phase III studies. By minimizing thrombocytopenia these strategies should enable more HCV patients to achieve adequate platelet counts in order to initiate and maintain IFN therapy.
