Masaharu Gotoda

Investigation for the amorphous state of ER-34122, a dual 5-lipoxygenase/cyclooxygenase inhibitor with poor aqueous solubility, in HPMC solid dispersion prepared by the solvent evaporation method.

ER-34122, a poorly water-soluble dual 5-lipoxygenase/cyclooxygenase inhibitor, exists as a crystalline form. According to an Oak Ridge thermal ellipsoid plot drawing, carbonyl oxygen O (5) makes an intermolecular hydrogen bond with the hydrogen bonded to N (3) in the crystal structure. The FTIR and the solid-state 13C NMR spectra suggest that the network is spread out in the amorphous state and the hydrogen bonding gets weaker than that in the crystalline phase, because the carbonyl signals significantly shift in both spectra. When amorphous ER-34122 was heated, crystallization occurred at around 140°C. Similar crystallization happened in the solid dispersion; however, the degree of crystallization was much lower than that observed in the pure amorphous material. Also, the DSC thermogram of the solid dispersion did not show any exothermic peaks implying crystallization. The heat of fusion (ΔHf) determined in the pure amorphous material was nearly equal to that for the crystalline form, whereas the ΔHf value obtained in the solid dispersion was less than a third of them. These data prove that crystallization of the amorphous form is dramatically restrained in the solid dispersion system. The carbonyl wavenumber shifts in the FTIR spectra indicate that the average hydrogen bond in the solid dispersion is lower than that in the pure amorphous material. Therefore, HPMC will suppress formation of the intermolecular network observed in ER-34122 crystal and preserve the amorphous state, which is thermodynamically less stable, in the solid dispersed system.

Trans-3,4-dideoxyglucone-3-ene (trans-3,4-DGE), a most reactive glucose degradation product in freshly heat sterilized glucose solutions.

In our study, one or more glucose degradation products (GDPs) in freshly heat sterilized dextrose 5% in water (D5W) were found to react with a drug candidate having a β-keto amide group (Compound A) to form several drug related compounds with the same molecular weight. However the previously identified GDPs did not react with Compound A to produce the observed adducts, indicating that unidentified GDP(s) reacted with Compound A to form these adducts. Our investigation by reaction-directed fractionation of the reactive D5W with HPLC led to the identification of the reactive GDP, trans-3,4-dideoxyglucosone-3-ene (trans-3,4-DGE), responsible for producing these reaction products. The trans-3,4-DGE was identified from its derivatives of dinitrophenylhydrazine (DNPH) and acetoacetanilide and confirmed by (1) admixing Compound A with authentic trans-3,4-DGE to produce the identical impurities as admixing with freshly heat sterilized D5W, and (2) NMR analysis of the reactive fraction of glucose solutions.