Jd. Bonny et M. Kyowa, USE OF IN-VITRO RELEASE TESTS FOR THE PREDICTION OF THE IN-VIVO BEHAVIOR AND THE DEVELOPMENT OF FLUCYTOSINE CONTROLLED-RELEASE CAPSULES, Journal of pharmaceutical sciences, 84(5), 1995, pp. 619-623
The antifungal substance flucytosine (5-fluorocytosine) was chosen as
a model drug for studying the development of a controlled-release caps
ule dosage form based mainly on in vitro drug release tests. A solutio
n and various capsule formulations of flucytosine were orally administ
ered to beagle dogs, and the resulting plasma concentration-time profi
les were determined. Special attention was directed to multiple-unit h
ard-gelatin capsules that exhibited a controlled-release effect in vit
ro, but not in vivo. The respective plasma profile was treated by the
technique of numerical deconvolution to obtain the corresponding in vi
vo release profile (i.e., the drug release from the dosage form in the
gastrointestinal tract of the dog). The influence of different experi
mental conditions of the USP paddle method on the release profile was
investigated. By adding surfactant to the dissolution medium and omiss
ion of the sinker device, which is used to prevent the capsule from fl
oating, the in vitro release rate markedly increased and correlated we
ll with the in vivo release. Based on this adjusted in vitro test, the
formulation was modified to achieve a more pronounced controlled-rele
ase effect. The plasma profile resulting from the reformulated multipl
e-unit capsules was in good agreement with the predictions made by num
erical convolution of the in vitro release profile.