(657a) Prediction of Tablet Dissolution By Process Parameters in Continuous Manufacturing

Tablet dissolution is an attribute that is critical to the product quality and should be well controlled in order to produce an acceptable pharmaceutical product. Tablet dissolution testing is a lengthy procedure that is time-consuming, expensive, and requires media preparation. In addition, it is a destructive test, and the sample will no longer be available for further characterization. Hence, there has been a growing interest in the development of non-destructive predictive tools for tablet dissolution to enable the real time release testing (RTRt) strategy. Application of such a strategy would not only lower laboratory and personnel costs but also increase product quality. A statistical model for predicting tablet dissolution profiles will enable rapid techniques in continuous assessment of tablets ensuring the desired quality attributes.

In this work, we developed a statistical model to predict tablet dissolution profile in a continuous direct compaction manufacturing line. Tablets consisted of model API, Acetaminophen (APAP), lactose monohydrate as the filler, and magnesium stearate as lubricant. Multivariate effects were analyzed using Quality-by-Design methods. The experiment was based on a fractional factorial design with 30 experiments, including 3 repeated center points, to study the influence of four process variables – drug concentration (%), compaction force (KN), blending speed (rpm), and feed frame speed (rpm) – on the final product. Dissolution profiles of 180 tablets (six tablets for each experimental run) were obtained and fit into the Weibull model (including parameters of α and β), and the values of α and β were calculated. Multi variate analysis of variance (MANOVA) was used to determine the statistical differences between the dissolution profiles. The results indicated that compaction force and API concentration were the main parameters affecting tablet dissolution in this process. Therefore, a predictive model was developed to correlate the dissolution model parameters α and β to the process parameters. The model was verified using an independently manufactured validation set. The values of α and β of the validation set were estimated by the model. The dissolution profiles of these tablets were predicted using the estimated values of α and β. The predicted dissolution profiles obtained from the model were in agreement with the experimental results using the similarity factor f₂ value greater than 50. This work establishes an approach to the real time release testing (RTRt) of pharmaceutical tablets and the possibility of assessing the quality characteristics of tablets non-destructively.