Purpose: High-frequency applied cetalkonium chloride (CAC) and benzalkonium chloride (BAC) 0.02% did not hamper corneal healing in a living rabbit model of induced corneal erosion. In contrast, the ex vivo eye irritation test (EVEIT) shows inhibition of healing for these substances. In a systematic ex vivo reproduction of the in vivo experiments, we discuss the background of these differences. Methods: Excised rabbit corneas (n = 5 per group) were cultured in artificial anterior chambers (EVEIT). Four erosions were induced for each cornea before starting regular 21 installations/day over 3 days of (1) CAC containing eye drops (Cationorm®), (2) 0.02% BAC. Corneal fluorescein staining, quantification of glucose-/lactate consumption, and histology were performed. Results: BAC 0.02% treated corneas showed increased epithelial lesions from 10.13 ± 0.65 mm2 to 10 ± 0.8 mm2 on day 0, to 86.82 ± 5.18 mm2 (P < 0.0001) by day 3. After a trend toward smaller lesions for CAC on day 1, erosion sizes increased significantly by day 3 from 9.82 ± 0.30 mm2 to 29.51 ± 16.87 mm2 (P < 0.05). For 1 cornea, corneal erosions nearly disappeared on day 3 (0.89 mm2). Corneal lactate increased significantly for BAC and CAC, whereas glucose concentrations were unchanged. Histology revealed disintegration of the corneal structures for both compounds. Conclusions: The data underline the EVEIT as a predictive toxicity test to show side effects in a time-compressed manner. The consistency of these predictions was previously demonstrated by the EVEIT for BAC, phosphate buffer, and others. The EVEIT is suited for a chronic application prediction of tolerability and toxic side effects of eye drops in particular, and other chemicals in general.
Keywords: EVEIT; epithelial healing; ocular toxicity prediction; ophthalmic preservatives tolerability.