A semiparametric deconvolution model to establish in vivo-in vitro correlation applied to OROS oxybutynin

In vitro-in vivo correlation (IVIVC) models may be used to predict in vivo drug concentration-time profiles given in vitro release characteristics of a drug. This prediction is accomplished by incorporating in vitro release c haracteristics as an input function (A(vitro)) to a pharmacokinetics model. This simple approach often results in biased predictions of observed in vi vo drug concentrations, and it can result in rejecting IVIVC. To solve this problem we propose a population IVIVC model that incorporates the in vitro information and allows one to quantify possibly changed in vivo release ch aracteristic. The model assumes linear kinetics and describes the in vivo r elease as a sum of A(vitro) and a nonparametric function (A(d), a spline) r epresenting the difference in release due to in vivo conditions. The functi on A(vitro) and its variability enter the model as a prior distribution. Th e function A(d) is estimated together with its intersubject variability. Th e number of parameters associated with A(d) defines the model: no parameter s indicates perfect IVIVC, a large number of parameters indicates poor IVIV C. The number of parameters is determined using statistical model selection criteria. We demonstrate the approach to solve the IVIVC problem of an ora l extended release oxybutynin form (OROS), administered in three pharmacoki netic studies. These studies present a particular challenging case; that is , the relative bioavailability for the OROS administration is > 100% compar ed with that of the immediate-release form. The result of our modeling show s that the apparent lack of IVIVC can be overcome: in vivo concentration ca n be predicted (within or across data sets) based on in vitro release rate together with a simple form of systematic deviation from the in vitro relea se.