In vitro--in vivo correlation of dissolution, a time scaling problem? Transformation of in vitro results to the in vivo situation, using theophylline as a practical example.

Abstract

Two principal approaches to demonstrating the continuous in vivo relevance of an in vitro dissolution test are outlined. The first uses the convolution technique to predict the concentration-time course in vivo; the second uses deconvolution as a mathematical tool to estimate the in vivo dissolution profile. The weighting function must be known to utilise either technique. Defined by the aim of the analysis the dose-normalized response to the oral solution is regarded as the weighting function (Impulse Response). In both cases the essential step is continuous comparison of the predicted time dependent data with actual readings of the same class. To permit the prediction of concentration-time data from in vitro dissolution data the basic equations for the transformation of the time base from in vitro to in vivo conditions are developed. The transformation is essential, since one cannot assume that the time scales for the in vitro and the in vivo experiment are definitely the same. The estimated in vivo dissolution profile using the deconvolution technique gives a hypothetical image of the true in vivo dissolution curve. Comparison with in vitro dissolution test results, using one of the equivalence testing procedures, reveals how closely and for how long the in vitro dissolution test simulates the in vivo dissolution process. For the formulation of theophylline studied, equivalence of the in vitro and the estimated in vivo dissolution profiles was not confirmed for the entire period of observation, but it was demonstrated for approximately the first 5 h. The later inequivalence is not due to possible non-linear or time-dependent kinetics of theophylline. There is a discussion of whether a change in pH, agitation of the formulation, diffusion conditions or the absorption rate constant along the gastrointestinal tract might explain the biphasic linear correlation of the in vitro and in vivo data observed.