Corneal epithelial cell viability following exposure to ophthalmic solutions containing preservatives and/or antihypertensive agents

Introduction: This in-vitro study compared the toxicity of bimatoprost 0.01% containing benzalkonium chloride (BAK) 0.02% with other commercial BAK-free or BAK-containing prostaglandin analogs. Methods: Six test solutions were evaluated: travoprost 0.004% with polyquaternium-1 0.001% (PQ), PQ, bimatoprost 0.01% with BAK 0.02%, latanoprost 0.005% with BAK 0.02%, tafluprost 0.0015% preservative free (PF), and BAK 0.02%. Phosphate-buffered saline (PBS) was the live control and 70% methanol was the dead control. Confluent human corneal epithelial cells were incubated with test solutions (diluted 1:5 or 1:10 with PBS) or control solutions for 10 or 25 min, after which cells were fluorescently labeled to distinguish live and dead cells. Data were expressed as a percentage of PBS live-cell fluorescence for automated readouts. Live and dead cells were manually counted for numeric analyses. Results: For 1:5 and 1:10 dilutions using automated readout, cells exposed to bimatoprost with BAK, latanoprost with BAK, and BAK alone demonstrated significant reductions in the live cell signal compared with PBS, travoprost with PQ, and PQ alone (all P < 0.001). They also demonstrated significantly greater toxicity than tafluprost PF for 1:5 dilutions (all P < 0.001) and 1:10 dilutions (P ≤ 0.02), except for 1:10-diluted bimatoprost with BAK (P = 0.41). For 1:5 dilutions using manual cell count, cells exposed to bimatoprost with BAK demonstrated significant reductions in the percentage of live cells compared with PBS (P = 0.02). For 1:10 dilutions using manual cell count, cells exposed to bimatoprost with BAK, latanoprost with BAK, and BAK alone demonstrated significantly greater toxicity than PBS, travoprost with PQ, PQ alone, and tafluprost PF (all P ≤ 0.03). No significant differences were observed among PBS, travoprost with PQ, and PQ alone under any test conditions (P ≥ 0.63). Conclusion: This study demonstrated that BAK-containing solutions, including bimatoprost 0.01% with BAK, were toxic to human corneal epithelial cells, whereas BAK-free solutions showed little to no evidence of toxicity.