The effect of controlled release tablet performance and hydrogel strength on in vitro/in vivo correlation.

The impact of controlled release (CR) formulations having different gel strength values (gamma) on in vivo tablet performance and the in vitro/in vivo correlation of the formulations was investigated. The CR tablets containing either hydroxypropyl methylcellulose (HPMC), hydroxypropyl cellulose (HPC), or carbomer were formulated with theophylline and Fast Flo lactose to produce tablets with a polymer content of 8 and 30% w/w. gamma was measured using a previously reported method. Male beagle dogs were utilized. Results showed that dissolution profiles were similar for all three polymers at the same % w/w level of polymer, irrespective of media (DI H2O, 0.1 N HCl, and pH 6.8 phosphate buffer). Mean gamma values were significantly different (p < or = 0.05) and were in order of HPMC K100MP > HPC HXF > carbomer 971P (same 30% w/w) with absolute gamma values at 30% w/w in DI H2O of 6600, 4600, and 1600 ergs/cm3, respectively. Drug profiles in plasma for the 30% HPMC K100MP tablets were consistent with in vitro dissolution profiles and gamma values. Plasma profiles for the 30% HPC HXF tablets were similar in vivo as the HPMC tablets. Plasma profiles for the 30% carbomer 971P formulation showed much higher drug concentrations (compared to HPMC and HPC) in vivo in all dogs. This findings is not consistent with the slow drug release found in the dissolution profiles but consistent with its low in vitro gamma values. Assessment of the predictability of a level A in vitro/in vivo correlation was quantified by absolute mean percent prediction error (PE). Formulations having gamma approximately 6000 ergs/cm3 have acceptable PE < 20%, and low standard deviation (sigma). Results showed that gamma values of CR hydrogel tablets in vitro will affect the in vivo performance (i.e., absorption kinetics of the drug) of the tablets and were also found to better assess (compared to in vitro dissolution profiles alone) the predictability of in vitro/in vivo correlations (level A and multiple level C).