Discussion
This population PK analysis was established using rich and sparse PK
samples collected from 6 clinical trials including 72 healthy volunteers
and 147 HCV infection patients. A two-compartment model with a
sequential zero-first order absorption and first-order elimination could
described the PK profiles of yimitasvir well. After a stepwise forward
inclusion and backward exclusion covariate screen procedure,
statistically significant covariates in the final model were food status
on Ka and F, gender and ALT on CL/F, and disease status on Td.
Results from phase 1 SAD study in healthy volunteers suggested that the
exposure (AUC and Cmax) of yimitasvir increased less
than proportionally in the range of 30-600 mg. Similar exposures were
found between 400 mg and 600 mg groups indicating a limited absorption.
Less than dose proportionality was also observed in phase 2 study in
patients after oral administrations of 100 or 200 mg yimitasvir in
combination with 400 mg sofosbuvir. Sofosbuvir did not affect the
pharmacokinetic properties of yimitasvir, which had been proven by other
drugs targeting HCV NS5A protein like velpatasvir [9] and ledipasvir
[10]. However, the exposure of yimitasvir was dose-proportional in
phase 1b trial in patients with chronic HCV genotype 1 infection after
single and multiple oral dosing from 30 to 200 mg. The mechanism behind
the difference in the PK linearity of yimitasvir between phase 1b and
the other trials is unclear. One significant difference in clinical
trial design was that subjects in phase 1b trial administered yimitasvir
in the evening after dinner for more than 4 h, while subjects in other
trials took the drug in the morning under fasted state. The inefficiency
of gastric emptying resulted in decrease of absorption rate with a
median time to peak concentration (Tmax) of 4-12 h and a
40% AUC decrease when compared to phase 1 SAD trial. The decrease in
exposure may contribute to the linearity of yimitasvir in phase 1b
trial.
Food status was a significant covariate affecting the absorption rate
(Ka) and bioavailability (F). In this population PK analysis, food was
classified into three categories. Subjects who took the drug under
fasted and fed (high-fat meal) condition were classified as food = 0 (as
reference) and food = 2, respectively, while subjects in phase 1b trial
who administered yimitasvir in the evening after dinner more than 4 h
were classified as food = 1. Administration with a high-fat meal or in
the evening after dinner more than 4 h resulted in a 28.9% and a 38.5%
decrease in AUCss, a 10.0% and 3.52% decrease in
Ctrough,ss, and a 46.1% and 58.9% decrease in
Cmax,ss, respectively (Figure 4). The results from the
population PK analysis were consistent with the results from previous
studies. The decrease in steady state exposures might be due to the
solubility of yimitasvir. Yimitasvir exhibited a pH-dependent solubility
profile. In the fasted condition, yimitasvir solubility increased due to
a low gastric pH in the stomach. However, a high-fat meal resulted in
higher gastric pH for drug dissolution reduction [14]. Whether the
decreased exposure after a high-fat meal would be clinically significant
was not clear. It was recommended that yimitasvir was administered at
least 2 h before or after a meal.
Gender was found to be a significant covariate for yimitasvir CL/F.
Female subjects had lower apparent clearance than male subjects. The
steady state exposures (AUCss,
Ctrough,ss, Cmax,ss) in female subjects
were 15.1-54.0% higher than male subjects. Baseline ALT was another
significant covariate on Cl/F. The apparent clearance decreased with the
increase of baseline ALT.
There was no difference in PK parameters between patients and healthy
volunteers, except for parameter Td, which was a little longer in
healthy volunteers than that in patients. The difference in Td between
the two kinds of population did not result in exposure differences.
In phase 2 clinical trial, yimitasvir 100 or 200 mg was administered
once daily for consecutive 12 weeks in combination with 400 mg
sofosbuvir in HCV infection patients. The primary endpoint of SVR12
rates were achieved 100% in both 100 mg yimitasvir/400 mg sofobuvir and
200 mg yimitasvir/400 mg sofobuvir groups. The adverse reaction rates
were comparable between 100 mg (35.9%) and 200 mg (36.9%) groups. The
most common adverse reactions were neutropenia (3.9%), leukopenia
(3.1%), hypercholesterolemia (3.1%) and fatigue (3.1%). All of these
adverse reactions were grade 1 or 2 in severity. No dose-response
relationship for efficacy and safety was identified in phase 2 clinical
trial. Considering the high response rate and favorable safety profiles
of yimitasvir in phase 2 clinical study, the significant covariates of
gender and baseline ALT on clearance was not clinically relevant.
Except for gender, other baseline demographics such as age and BW had no
effect on yimitasvir PK properties. It seems that there is no need to
adjust dosage based on these factors. But due to the small sample size
of phase 2 study, this conclusion should be further validated in a
larger patient population.
This is the first time to establish a population PK model of a new HCV
NS5A inhibitor, yimitasvir, in Chinese healthy volunteers and patients
with chronic HCV genotype 1 Infection. It is very helpful for us to know
statistically significant covariates affecting the pharmacokinetic
property of yimitasvir. This work can be used as a basis for subsequent
population PK model development including more patient data from phase 3
trial.
The population pharmacokinetic model was demonstrated to be appropriate
and effective to describe the pharmacokinetics of yimitasvir in Chinese
population. Food status, disease status (healthy volunteers vs
patients), gender and baseline ALT were identified as statistically
significant covariates to affect yimitasvir pharmacokinetics. High-fat
meal decreased absorption rate and bioavailability, so it is recommended
to take yimitasvir at least 2 h before or after a meal. Considering the
favorable safety profile of yimitasvir and 100% SVR12 rate in phase 2
study, the impact of gender and ALT on yimitasvir exposure was not
considered clinically relevant. This conclusion should be further
validated in a larger patient population from phase 3 clinical trial.