Impact of variability in carbamazepine raw materials on drug release

Abstract

Variability in raw materials presents a challenge for pharmaceutical companies. The varying physicochemical properties can critically influence drug release and bioavailability of the final dosage form. Therefore, a strategy to control this variability is required. In this study the well-established antiepileptic drug carbamazepine (CBZ) was selected as the model drug as it presents one example where variability in raw materials has been linked to bioinequivalence and clinical failures. CBZ shows poor solubility, low potency, and a narrow therapeutic index. Furthermore, CBZ exhibits at least four polymorphic forms and it transforms into the less soluble CBZ dihydrate in water. The purpose of this work was to study the impact of variability in CBZ samples of four different suppliers on the drug release and to suggest a strategy to deal with the sample variability. Thus, the CBZ samples were characterized at preformulation as well as at a formulation level. Polymorphism and morphology of CBZ samples were analyzed by differential scanning calorimetry, X-ray powder diffraction, sieve analysis, and scanning electron microscopy. CBZ samples were characterized by a unidirectional dissolution method measuring disc intrinsic dissolution rate (DIDR) of CBZ raw material and initial drug release in presence of the tablet fillers microcrystalline cellulose (MCC) and mannitol (30–90% drug load). Furthermore, CBZ samples were recrystallized in 1% polyvinylpyrrolidone ethanol solutions as an approach to reduce the sample variability. At the formulation level, a high-dose CBZ tablet was developed with the aim of a tablet formulation that is robust towards the variability in CBZ samples and that conforms to the USP requirements of CBZ tablets for immediate release. Therefore, the superdisintegrant crospovidone (CrosPVP) and the dry binder hydroxypropyl cellulose (HPC) were used, as both are reported to inhibit transformation to CBZ dihydrate. The tablet filler was MCC. All CBZ samples were of p-monoclinic form but differed in their polymorphic purity, particle size, morphology, and intrinsic dissolution rate. The DIDR profiles showed high variability among the CBZ samples. Two inflection points could characterize individual transformation behavior of anhydrous CBZ to CBZ dihydrate. Presence of MCC reduced drug release variability. Recrystallizing CBZ resulted in strongly reduced variability in dissolution and tablet strength and the transformation to CBZ dihydrate was inhibited. However, particle size and morphology could not be controlled and drug release from binary mixtures with MCC presented deviation for one of the recrystallized CBZ samples. For the tablet formulation the optimal condition was with 6% HPC and 5% CrosPVP, where tablet properties of all CBZ samples were at least 70 N tablet hardness, less than 1 min disintegration, and within the USP requirements for drug release. Nonetheless, dissolution curves of the various CBZ samples differed. Excluding the additive sodium laurylsulfate required by the USP monograph and analyzing the optimized tablet formulation in water only, the dissolution curves of the various CBZ samples could not be distinguished anymore (ANOVA, p > 0.05). The impact of variability in CBZ raw materials on the drug release could be characterized by an individual transformation behavior to the CBZ dihydrate. The applied unidirectional dissolution method can be suggested as a straightforward monitoring tool in preformulation studies conforming to the basic tenet of quality by design of FDA’s PAT initiative. To allow a certain variability in CBZ raw materials, it is suggested to incorporate the excipients CrosPVP, HPC, and MCC into the design of a CBZ tablet formulation. The strategy proposed of how to control the variability in CBZ samples includes the monitoring at preformulation level combined with the design of a robust tablet formulation.