Cause of high variability in drug dissolution testing and its impact on setting tolerances

Considering a variable mixing/stirring and flow pattern in a drug dissolution vessel as a likely source of high variability in results, experiments were conducted using USP paddle apparatus by placing (aligned to the walls) a metal strip (1.7 mm thick x 6.4 mm wide) in a dissolution vessel. The metal strip forces the undisintegrated tablet to settle about 3 mm away from the centre, facilitates spread of disintegrated material and diminishes the cone formation at the bottom of the vessel. To assess the impact of this altered environment in the vessel, but still maintaining the vessel dimensions within required specifications, drug release characteristics were evaluated for products having different formulation/manufacturing attributes. Tests were conducted with calibrator tablets (USP prednisone and salicylic acid tablets and FDA proposed NCDA No. 2 prednisone tablets) and two commercially available products (250 mg amoxicillin capsules and 5 mg glibenclamide tablets). Except for the glibenclamide tablet product, all products gave significantly (P<0.01) higher dissolution results with vessels containing metal strip than without. The extent of increased dissolution with the metal strip varied from product to product i.e. USP prednisone tablet was the smallest (14.%) and NCDA No. 2 was the largest (88.4%). Based on the results obtained from this study, it is concluded that employing the current apparatuses, in many cases products will provide lower than anticipated results which may not be reflective of the product drug release characteristics. Test-to-test variability, within or between laboratories, can also be very high depending on the settling position of the product once dropped in the vessel and/or due to slight aberration in the walls of the vessel by altering the extent of spread of disintegrated material at the bottom of the vessel. Thus, dissolution testing will require wider tolerances to be useful for comparison of batch-to-batch or interlaboratory results. Crown copyright (C) 2001 Published by Elsevier Science B.V. All rights reserved.