Typical variability in drug dissolution testing: study with USP and FDA calibrator tablets and a marketed drug (glibenclamide) product.

To evaluate variability in drug dissolution testing 28 laboratories analyzed USP calibrators, US FDA prednisone tablets and a marketed glibenclamide tablet product. The experiments were conducted using paddle and basket methods at 50 (calibrators) and 75 (glibenclamide) rpm. The media employed were deaerated by equilibrating at 37 degrees C for 24 h and by the USP recommended method. The 95% CI values for percent drug release for the USP calibrator tablets were similar to the reported tolerances for the USP Acceptance Ranges; however, individual results from 15 of 28 laboratories suggest that the apparatus would not comply with the USP Apparatus Suitability Criteria. For FDA prednisone calibrator tablets, percent drug release using equilibrated medium was different (P=0.003) than by the USP recommended method. For the glibenclamide tablet results, a CV of 14-37% was observed, depending upon the sampling time and the type of apparatus employed. The results indicate that failure to meet the USP Dissolution Apparatus Suitability Test may not truly mean that the apparatus is 'out of compliance'. Due to the high variability in dissolution testing, in many cases the impact of formulation or manufacturing changes on drug release characteristics may not be observed, in particular with multi-point profiles.