Reduction of quaternary ammonium-induced ocular surface toxicity by emulsions: an in vivo study in rabbits

PURPOSE To evaluate and compare the toxicological profiles of two quaternary ammonium compounds (QAC), benzalkonium chloride (BAK), and cetalkonium chloride (CKC), in standard solution or cationic emulsion formulations in rabbit eyes using newly developed in vivo and ex vivo experimental approaches. METHODS Seventy eyes of 35 adult male New Zealand albino rabbits were used in this study. They were randomly divided into five groups: 50 microl of phosphate-buffered saline (PBS), PBS containing 0.02% BAK or 0.002% CKC (BAK Sol and CKC Sol, respectively), and emulsion containing 0.02% BAK or 0.002% CKC (BAK Em and CKC Em, respectively) were applied to rabbit eyes 15 times at 5-min intervals. The ocular surface changes induced by these eye drops were investigated using slit-lamp examination, flow cytometry (FCM), impression cytology (IC) on conjunctiva, and corneal in vivo confocal microscopy (IVCM). Standard immunohistology in cryosections was also examined for cluster of differentiation (CD) 45+ infiltrating and terminal deoxynucleotidyl transferase-mediated dUTP-nick end labeling (TUNEL)+ apoptotic cells. RESULTS Clinical observations and IVCM showed that the highest toxicity was induced by BAK Sol, characterized by damaged corneal epithelium and a high level of inflammatory infiltration. BAK Em and CKC Sol presented moderate effects, and CKC Em showed the lowest toxicity with results similar to those of PBS. Conjunctival imprints analyzed by FCM showed a higher expression of RLA-DR and TNFR1 markers in BAK Sol-instilled eyes than in all other groups, especially at 4 h. Immunohistology was correlated with in vivo and ex vivo findings and confirmed this toxicity profile. A high level of infiltration of CD45+ inflammatory cells and TUNEL+ apoptotic cells was observed in limbus and conjunctiva, especially in QAC solution-receiving eyes compared to QAC emulsion-instilled eyes. CONCLUSIONS The acute administration of 15 instillations at 5 min intervals was a rapid and efficient model to assess quaternary ammonium toxicity profiles. This model showed the highest toxicity, induced by the BAK solution, and the lowest level of toxicity, induced by the CKC emulsion. These in vivo and ex vivo experimental approaches demonstrated that ocular surface toxicity was reduced by using an emulsion instead of a traditional solution and that a CKC emulsion was safe for future ocular administration.