Is saliva suitable as a biological fluid in relative bioavailability studies? Analysis of its performance in a 4 x 2 replicate crossover design.

The aim of the present study was to evaluate the suitability of saliva as a biological fluid in relative bioavailability (RBA) studies, with the focus on the statistical design and data variability. A randomized, open-label, four periods and two sequences (4 × 2) crossover RBA study in saliva of two phenytoin (PHT) 100 mg immediate-release capsules was performed. PHT is a narrow therapeutic index drug that has been widely used for epilepsy treatment for many years. Published information regarding its bioavailability is available, but plasma assessed. This study was designed and performed using saliva as the biological fluid and the simplest conditions that produce coherent results with previously published plasma studies. Pharmacokinetic parameters (C (max), T (max), AUC(0-t ), AUC(0-inf), C (max)/AUC(0-t ), K (e), and t (1/2)) for each volunteer at each period were calculated. Four different BE calculations were performed: individual bioequivalence, by the method of moments, and three average bioequivalence with data averaged over the two administrations and with data of periods 1-2 and 3-4. ANOVA calculation showed no significant subject-by-formulation interaction, period and sequence effects. The intra-subject variabilities were at least 20-fold lower than the inter-subject ones for C (max), AUC(0-t ) and AUC(0-inf). In all four BE calculations, the 90% CIs for the T/R ratios of studied pharmacokinetics parameters fell within the 80-125% range proposed by most regulatory agencies.