Evaluation of "external" predictability of an in vitro-in vivo correlation for an extended-release formulation containing metoprolol tartrate

Abstract

The purpose of this study was to examine the external predictability of an in vitro-in vivo correlation (IVIVC) for a metoprolol hydrophilic matrix extended-release formulation, with an acceptable internal predictability, in the presence of a range of formulation/manufacturing changes. In addition, this report evaluated the predictability of the IVIVC for another formulation of metoprolol tartrate differing in its release mechanism. Study 1 examined the scale up of a matrix extended-release tablet from a 3-kg small batch (I) to a 50-kg large batch (II). The second study examined the influence of scale and processing changes [3-kg small batch with fluid bed granulation and drying (III); 80-kg large batch with high shear granulation and microwave drying (IV), and a formulation with an alternate release mechanism formulated as a multiparticulate capsule (V)]. In vitro dissolution of all formulations (I-V) was conducted with a USP apparatus I at pH 6.8 and 150 rpm. Subjects received the metoprolol formulations, and serial blood samples were collected over 48 h and analyzed by a validated HPLC assay using fluorescence detection. A previously developed IVIVC was used to predict plasma profiles. Prediction errors (PE) were <10% for C(max) and area under the curve (AUC) of concentration versus time for I, II, and IV. The C(max) for III was slightly underestimated (11.7%); however, the PE of the AUC was <10%. Formulation V displayed a PE for C(max) > 20% and an AUC within 5% of observed values. The low PEs for C(max) and AUC observed for I-IV strongly suggest that the metoprolol IVIVC is externally valid, predictive of alternate processing methods (IV), scale-up (II, III), and allows the in vitro dissolution data to be used as a surrogate for validation studies. However, the lack of predictability for V supports the contention that IVIVCs are formulation specific.