Physicochemical surrogates for in vitro toxicity assessment of liposomal amphotericin B

Pharmaceutical toxicity evaluations often use in vitro systems involving primary cells, cell lines or red blood cells (RBCs). Cell-based analyses ('bioassays') can be cumbersome and typically rely on hard-to-standardize biological materials. Amphotericin B (AmB) toxicity evaluations are primarily based on potassium release from RBCs and share these limitations. This study evaluates the potential substitution of two physicochemical AmB toxicity approaches for the bioassay: Ultraviolet-visible spectroscopy (UV-vis) and in vitro drug release kinetics. UV-vis spectral analyses indicated that liposomal AmB's (L-AmB) main peak position (λ_max) and peak ratio (OD₃₄₆/OD₃₂₂) are potential toxicity surrogates. Similarly, two first-order release parameters derived from USP-4 in vitro drug release analyses also provided linear relationships with toxicity. These were the initial, overall drug release rate and the ratio of loose to tight AmB pools. Positive slopes and high correlation coefficients (R² > 0.9) characterized all interrelations between physicochemical parameters and toxicity. These tests converted the manufacturing variables' nonlinear (i.e., curvilinear) relationships with in vitro toxicity to linear responses. Three different toxicity attenuation approaches (2 manufacturing, 1 formulation), covering formulation composition and process aspects, support this approach's universality. These data suggest that one or more spectral and kinetic physicochemical tests can be surrogates for L-AmB in vitro toxicity testing.