The effects of formulation and moisture on the stability of a freeze-dried monoclonal antibody-vinca conjugate: a test of the WLF glass transition theory

Abstract

Deacetylvinblastine (DAVLB) hydrazide, a cytotoxic vinca alkaloid, has been linked to the monoclonal antibody, KS1/4, via aldehyde residues of the oxidized carbohydrate groups on the antibody. The resulting KS1/4-DAVLB hydrazide conjugate is unstable in solution with both the acyl hydrazone linkage and the vinca moiety being subject to significant degradation, even at 5 degrees C. This necessitated the development of a freeze-dried formulation of the antibody-drug conjugate. Formulation factors considered were pH, ionic strength, buffer, excipient types, and excipient ratios. A formulation with equal weight ratios of mannitol, glycine, and conjugate in a low ionic strength phosphate buffer at near neutral pH was selected. Stability was studied at various moisture levels (1.4%, 3.0%, and 4.7%) and temperatures (5 degrees C, 25 degrees C, and 40 degrees C). Degradation was measured by size exclusion HPLC (aggregate formation) and by reverse phase HPLC (hydrolysis of hydrazone linkage and vinca decomposition). Differential scanning calorimetry (DSC) indicated that all samples were above their glass transition temperatures, Tg, when stored at 40 degrees C. When stored at 25 degrees C, only the highest moisture sample was initially above its Tg. However, due to crystallization of the excipients during storage and the resulting decrease in Tg, samples stored at 25 degrees C were also above their Tg during much of the storage period. The degradation rate, R, increased sharply with increasing temperature and with increasing moisture level. Degradation kinetics obeyed the Williams-Landel-Ferry relationship, R/Rg = exp[k(T-Tg)], where Rg is the degradation rate at Tg. For all three moisture levels and all three degradation pathways, k = 0.143.