Methodology for using oral dose pharmacokinetic data to select drugs for prolonged release formulations and validation of the method using simulated data

A computer aided method for selecting drugs as potential candidates for oral prolonged release formulations has been previously published. In order to decide whether trial formulations were warranted, prolonged release dosing was simulated to find all release rates and doses producing successful regimens in every subject for whom clinical pharmacokinetic data were available. Intravenous dose data were required because the models and their parameter values could not be assessed with oral doses. The new method uses model independent equations derived from rapidly absorbed oral doses to simulate the administration of prolonged release formulations. Using these equations, repetitive oral dosing regimens were reiteratively simulated to search for all successful release rates and doses in every available subject. Results were validated by comparison to those obtained with the published method using theophylline and hypothetical drugs, which included examples of flip‐flop and vanishing exponentials. When the oral reference dose data were reliable, results were similar even when the model independent parameter values did not agree with those used to generate the oral reference dose data. Differences were observed only when the random errors in the oral reference dose data were large. If more than one model independent equation provided equivalent fits to the oral reference dose data, the final results were similar independent of which equation was employed.