(692a) Molecular Simulation Studies of Quantitative Correlation Between the Stability of Drug Nanocrystals Suspension System and Certain Properties of Both the Drug API and the Stabilizer

                                                            Wusheng Zhu, Frank Romanski, and M. Silvina Tomassone

                          Department of Chemical and Biochemical Engineering, Rutgers University, Piscataway, New Jersey 08854

Molecular simulations have been applied in the suspension system of poorly water-soluble (BCS Class II) drug nanocrystals in presence of a polymer or surfactant stabilizer. Since the binding energy between the drug crystal surfaces and the stabilizer is believed to link the inhibitory effect of the stabilizer that reflects the relative stability of the drug suspension system, comparison of the simulated drug-stabilizer binding energy, which is evaluated as a statistical average over different cleavage drug crystal faces, is utilized to gauge the relative stabilization capability of the suspension system. To efficiently screen an effective stabilizer for a given drug API that will result in good suspension stability, our research work is to find out the correlation between the drug-stabilizer binding energy and certain properties of the API and the stabilizer. If such a correlation is mapped out, an effective stabilizer can be easily chosen for a given API achieving good suspension stability without doing numerous real screening tests. In previous studies, we have built up quantitative correlations for two cases: i) correlation between the drug-stabilizer binding energy and certain properties such as supersaturation ratio and the molar water solubility of the drug API in presence of the same stabilizer; ii) correlation between the drug-stabilizer binding energy and certain properties such as hydrophobicity and molecular area of the stabilizer absorption on the same drug API nanocrystals. The current ongoing work suggests that a hierarchical high dimensional representation model can be employed to build up the full correlation between the binding energy and the properties of both the drug API and the stabilizer. The correlations obtained for the same stabilizer or the same drug API act as the first order approximation in the high dimensional representation model. The second order correlation can be fitted by simulated binding energies between various stabilizers adsorption on various drug APIs.