Population pharmacokinetic modelling of factor IX activity after administration of recombinant factor IX in patients with haemophilia B

摘要

Purpose: The aim is to develop a pharmacokinetic model for factor IX activity (FIX) after BeneFIX (nonacog alfa, rFIX) administration and assess potential covariates using all available clinical data collected during development. Methods: The data set for model development combined observations from eight studies. Postdose FIX observations were adjusted by subtracting predose FIX if these were above the lower limit of quantification (BLQ) and all BLQ observations were removed. A population pharmacokinetic model was then developed with 4936 observations from 201 patients. Two additional studies (385 observations from 72 patients) became available and were used to evaluate the model. Results: A two-compartment model, parameterized for clearance (CL), volume of distribution of the central (V1) and peripheral (V2) compartments, and intercompartmental clearance (Q), with an effect of weight on all parameters was the final model. Weight was incorporated as a power function with exponent estimates close to conventional allometric scaling. Including interoccasion variability (IOV) on CL and V1 showed decreases in the objective function. Investigations of a full block omega matrix lead to the retention of a correlation between V2 and Q. Age was not a significant covariate with weight already included in the model. Observations in the studies used for evaluation were found to be higher than simulated values immediately after dosing, as well as a week after dosing. The differences may be due perhaps to differences in the patients enrolled in the evaluation studies (all were adults) as well as the sample collection time after dosing (longer after dosing in the evaluation studies). Conclusions: FIX is appropriately modelled as a two-compartment model after rFIX administration. When weight is included, no additional effect of age is observed. Longer times of observation after dosing may be helpful in refining the model.