CALCULATION OF PARTIAL COMPONENTS OF BIOAVAILABILITY IN SLOW-RELEASE FORMULATIONS USING MODEL-INDEPENDENT METHODS

In certain slow release formulations the total dose of drug is subdivi ded into at least two fractions with different release capacities; the se may display different degrees of bioavailability. In this type of f ormulation, calculation of the amount of biavailability by conventiona l methods does not permit one to differentiate the contribution of the different dose fractions to the total bioavailability of the formulat ion. This paper develops a model-independent method that permits optim ization of the different components of bioavailability regarding amoun t and rate, combining convolution/deconvolution methods as well as non -linear regression. The method calculates bioavailability parameters b y optimization of a prescribed input function in the form of a Laplace transform using the MULTI(FILT) program. The serum level curve after the administration of a Slow release formulation is used as a response function obtained by convolution between a prescribed input function and a weighting function expressed as a polyexponential. The method wa s tested using simulated serum level data with added random error. In all cases the bioavailability parameters were calculated with good pre cision.