Bioequivalence study of stressed and nonstressed hard gelatin capsules using amoxicillin as a drug marker and gamma scintigraphy to confirm time and GI location of in vivo capsule rupture

Purpose: Evaluate if crosslinked hard gelatin capsules (HGCs) having differ ent in vitro dissolution profiles changed in vivo release times or altered bioavailability of a drug marker; assess if a two-tier dissolution test (wi th and without enzyme) predicted in vivo performance. Methods. Two classifications of stressed HGCs were artificially produced by exposure to formaldehyde (HCHO). HGCs were categorized as, a) pass/pass (p /p) which met in vitro dissolution criterion (75% drug dissolution at 45 mi n), b) moderately crosslinked fail/pass (f/p) which failed dissolution crit erion in the absence of enzymes and passed in the presence of enzymes, and c) severely crosslinked fail/fail (f/f) which failed in vitro standards wit h or without enzymes. A sh-way, single dose bioequivalence study (n = 10) a dministered the three HGCs under the fasted and fed condition. In vivo caps ule rupture and GI transit were monitored via gamma scintigraphy, and blood samples were collected through six hours. Results. Each crosslinked HGC was bioequivalent to the control p/p capsule when using AUC(0-infinity) and Cmax for comparison. Mean in vivo disintegra tion of the pip capsule was 7 +/- 5 min for the fasted condition and 11 +/- 7 min for the fed condition. In vivo rupture for the f/p capsule was 22 c 1+/- min and 23 +/- 11 min for the fasted and fed studies, respectively, wh ile the f/f HGC ruptured at 31 +/- 15 min and 71 +/- 19 min under the faste d and fed condition, respectively. Onset of amoxicillin absorption was depe ndent on in vivo HGC rupture and subsequent entry of thr released radioacti ve marker into the small intestine. Consequently, fasted Tmax values were s ignificantly later for the f/p HGC (1.62 +/- 0.53 hr) and f/f HGC (1.85 +/- 0.58 hr) as compared to the pip HGC (1.17 +/- 0.30 hr). Fed Tmax values we re statistically different only for the f/f capsule (2.55 +/- 0.44 hr) wher e Tmax values for thr pip and f/p HGCs under the fed condition were 1.50 +/ - 0.47 hr and 1.60 +/- 0.46 hr, respectively. Conclusions. A two-tier dissolution procedure that retested a crosslinked h ard gelatin capsule with addition of gastric or intestinal enzymes provided an adequate in vitro indicator of the formulation's in vivo performance. T he observed delays in the onset of amoxicillin absorption and Tmax for the severely crosslinked f/f HGC was attributed to delayed in vivo capsule rupt ure, however, this delay did not adversely change AUC(0-infinity) nor Cmax.