In vitro evaluation of cytotoxic potential associated to antimicrobial agents commonly used in clinical ophthalmological practice

Abstract Purpose Topical antibiotics may cause cytotoxic effects on eye surface, especially when the integrity of ocular epithelia is a major concern (e.g. corneal damage). Here we investigated the potential cytotoxicity of one antiseptic (povidone‐iodine) and three antibiotics commonly used in ophthalmology (netilmicin, levofloxacin and chloramphenicol). Methods Human Corneal Epithelial cells (HCE) were exposed to all compounds (1.5, 3 or 6 mg/ml) for different lengths of time, i.e. 0.5, 1, 3, 6 and 24h. Cell viability was then assessed by MTT assay. Cell viability cut‐off was fixed at 50% (DB‐ALM: Protocol n°17). Statistics were by two‐way ANOVA. Results None of the compounds tested at 1.5 mg/ml had any significant cytotoxic effect, except for levofloxacin producing a 66% cell mortality at 24h. At 3 mg/ml, in addition to levofloxacin, chloramphenicol too produced a submaximal cytotoxic effect at 24h (76%). Netilmicin was the only to be devoid of any cytotoxic effect at 6 mg/ml while all the other compounds produced significant cytotoxic effects with varying onset times, chloramphenicol and povidone‐iodine showing the earliest onset (see figure). Conclusions Results demonstrate that levofloxacin and chloramphenicol produce cytotoxic effects on HCE even at concentrations that are lower than those used in clinical ophthalmological practice. Conversely, povidone‐iodine only showed cytotoxic effects at 6 mg/ml, that is the concentrations of eye drops available on the market. Most importantly, netilmicin was found to be endowed with an advantageous cytocompatibility profile being devoid of any cytotoxicity even at concentration exceeding those currently used in clinical ophthalmological practice (3 mg/ml). Further studies are warranted to investigate the potential cytotoxic effects of antimicrobial agents tested in a pulsed‐treatment setup in order to mimic the human dose regimen thus accounting for potential biases due to prolonged cell exposure.