Formulation design of tablets using response surface method and boosting tree
Toyama University Graduate School of Medicine and Pharmacology Education Department (Pharmacology) Formulation Design Study Course Visiting Assistant Professor Yoshihiro Hayashi
Toyama University Graduate School of Medicine and Pharmacology Education Department (Pharmacology) Formulation Design Study course Visiting Professor Yoshinori Okishi
Since the physical properties of the drug substance contained in the tablet exert a large influence on the quality of the tablet, it is important to clarify the relationship between the drug substance property and the tablet physical property in the formulation design of the tablet.However, in order to grasp the properties of the drug substance, it is necessary to multilaterally evaluate the physical properties, and it is a complicated system in which synergistic effect and antagonistic effect occur between each physical property.Therefore, the use of statistical analysis is useful for understanding the relationship between the original drug properties and tablet physical properties.In this lecture, we introduce two study cases applying response surface method and boosting tree to formulation design.
1. Relationship evaluation of responsive surface of drug substance tablet and placebo tablet
Several kinds of tablets with different mixing ratios of additives were prepared according to the experimental design method to prepare response curves of tablet hardness and disintegration time.Then, as a result of quantitatively comparing each response surface, we found that there is a correlation between the response curves of drug substance-containing tablets and placebo combination tablets.By applying this result, we demonstrated the possibility of estimating the response surface of drug substance-containing tablets from a limited experimental point by utilizing the data of placebo-containing tablets.
2. Assessment of relationship between original drug properties and tablet physical properties by boosting tree
Various original drug properties (molecular weight, partition coefficient, particle size distribution, bulk density, Hausner ratio, moisture content, elastic recovery rate, etc.) were evaluated for 81 kinds of drug substance.And by applying the boosting tree which is one of the machine learning functions installed in JMP Pro, we modeled the relationship between the original drug properties and tablet hardness with high accuracy.In this study, we showed that the boosting tree is a useful method for clarifying important drug substance properties.
