Multivariate toxicity screening of liposomal formulations on a human buccal cell line.

The influence of various formulation factors on the in vitro cellular toxicity of liposomes on human buccal cells (TR146), were studied by using the concept of statistical experimental design and multivariate evaluation. The factors investigated were the type of main phospholipid (egg-PC, DMPC, DPPC), lipid concentration, the type of charge, liposome size, and amount and nature of the charged component (diacyl-PA, diacyl-PG, diacyl-PS, stearylamine (SA), diacyl-TAP) in the liposomes. Both full factorial design and D-optimal designs were created. Several significant main factors and interactions were revealed. Positively charged liposomes were shown to be toxic. The toxicity of negatively charged liposomes was relatively low. Diacyl-TAP was less toxic than SA, and DPPC was less toxic than DMPC. Low level of positively charged component was favourable and essential when using egg-PC as the main lipid. The amount of negatively charged component, the liposome size, and the total lipid concentration did not affect the toxicity within the experimental room. DPPC appeared to be a good candidate when formulating both positively and negatively charged liposomes with low cellular toxicity. The concept of statistical experimental design and multivariate evaluation was shown to be a useful approach in cell toxicity screening studies.