A novel multicompartment dissolution system was developed by modifying a conventional six-vessel United States Pharmacopoeia dissolution system to study the dissolution and possible precipitation of poorly soluble weak bases after oral administration. The modified system includes a "gastric" compartment, an "intestinal" compartment, an "absorption" compartment, and a reservoir to simulate the dissolution and absorption in the gastrointestinal tract. Dissolution profiles of 50-mg dipyridamole (pK(a) 6.0, 12.5) tablet (2 * 25 mg Persantine tablets), 25- and 50-mg cinnarizine (pK(a) 1.95, 7.5) powders, which are poorly soluble weak bases, were generated in the system using dissolution medium with different pHs in the "gastric" compartment. The in vitro dissolution results were compared with the in vivo oral exposure data in humans. For both dipyridamole and cinnarizine, the in vitro dissolution using the multicompartment system was able to predict the pH effect on oral exposure. The results from the multicompartment system are more closely correlated with the in vivo data, compared with that from the conventional dissolution test. The system showed that although both dipyridamole and cinnarizine completely dissolved in the gastric compartment at lower pH, approximately 36% (at 25-mg dose) and 40% (at 50-mg dose) of cinnarizine precipitated in the "intestinal" compartment whereas the precipitation of dipyridamole was <10% of the initial dose. The difference in the amount "absorbed" between these two compounds in vitro is therefore primarily attributed to the precipitation potential, although no in vivo data are available to confirm this result. The difference in the amount precipitated may be explained by the lower solubility and consequently higher degree of supersaturation of cinnarizine in the "intestinal" compartment.
(c) 2004 Wiley-Liss, Inc.