Mechanistic Viral Kinetic (VK) modeling of a triple therapy for Hepatitis C: a protease inhibitor (Danoprevir) in combination with Pegasys® (PEG-IFN alpha-2a ) and Copegus® (Ribavirin)
Introduction and Objectives:
Danoprevir (DNV) is an inhibitor of the HCV NS3/4A protease, an enzyme required for viral replication, that demonstrated a potent in-vitro and in-vivo activity against HCV, and is currently in Phase 2 development for the treatment of Chronic Hepatitis C (CHC) in combination with standard of care (SOC): Pegasys® (PEG-IFN alpha-2a ) and Copegus® (Ribavirin – RBV). Ritonavir (RTV) is used at low, sub-therapeutic doses, to enhance the exposures of HIV protease inhibitors. Co-administration of DNV with low dose RTV is being investigated as a strategy to enhance DNV pharmacokinetics and potentially pharmacodynamic effect. The goal of this analysis was to describe the effect of DNV with/without RTV and with/without SOC on HCV RNA viral loads using a mechanistic VK model. The VK model is based on the prey-predator principle where the hepatocytes are the prey and the virus is the predator. This model addresses the host–virus–drug interactions with system parameters and drug effect parameters as displayed in Figure 1 and equations 1 – 4. The structure of the mechanistic VK model has been previously developed using a large viral load database composed of 2,100 CHC patients treated with PEG-IFN, alone or in combination with ribavirin .
The effect of DNV on the host-virus interactions has been incorporated in the VK model by using available PK and viral load data from treatment-naive, non-cirrhotic patients with HCV genotype 1 infection in phase 1 and 2 clinical studies (NSHC-002, NP22660 and NV21075 ). In study NSHC-002, patients were randomized to 15 days of DNV 100 or 200 mg, q12h or q8h (N=32). In study NP22660 (phase 1b), patients were randomized to 15 days of DNV 100, 200 mg q12h or 200 mg q24h in combination with RTV 100 mg and PegIFN 180 µg/week plus RBV 1000/1200 mg/day (N=24). In study NV21075 (phase 2b), patients were randomized to 12 weeks of (A) DNV 300 mg q8h, (B) 600 mg q12h, or (C) 900 mg q12h in combination with PegIFN 180 µg/week plus RBV 1000/1200 mg/day (n=129). System parameters and SOC treatment effect were fixed based on model parameters determined from the previously well-validated model . DNV treatment effect was investigated mechanistically by including an exposure-dependent inhibition and/or induction on various model parameters. The estimation of the parameters was performed using MONOLIX .
In the HCV VK model for SOC, the PEG-IFN and RBV exposure variables were the weekly dose and the daily body-weight adjusted dose., respectively. DNV pharmacokinetics is highly variable and consequently individual exposure parameters were used instead of the dose to describe the effect of DNV. The exposure PK variables evaluated were daily AUC and the trough concentration. The effect of DNV was initially implemented as an exposure-dependent inhibition of the virus production; additional mechanisms of action have also been considered: i.e. an effect on the death rate of infected cells and on the clearance of the virion. The extended HCV VK model for DNV effect described the data well and was qualified using Visual Predictive Check procedure.
A mechanistic viral kinetic model was used to describe the effect of DNV alone and in combinations with SOC and with or without co-administration of RTV and on HCV viral load. This model represents a key element of a modeling and simulation frame work used in the development of DNV and potentially other direct acting antiviral agents. Model-based simulations will be performed to further guide DNV drug development.
where target cells T, produced at a rate s, become infected at an infection rate b, and replicate and die at rate constants r and d, respectively. In turn, infected cells are produced at a rate constant b and lost with a rate d. Free non-replicating virions are produced from infected cells at a rate p and are cleared with a free-virion clearance rate c. The effect of PEG-IFN was described by inhibition of the virus production by dose dependant fraction ε. RBV effect was described by causing a fraction of the newly produced virion to become non-infectious by dose dependant fraction ρ.