Background Voriconazole is an antifungal agent effective on most keratitis causative fungi with an excellent transcorneal penetration. Voriconazole eyedrops (VED) are unavailable in Europe and are usually compounded in hospital pharmacies. New eyedrop containers emerged on the hospital market (eg, high density polyethylene bottles available in trays (CAT)) for which few stability data are available, or Novelia bottles which innovative insert maintains sterility after opening (no stability data available).

Purpose To collect data on VED stability in 3 different containers in order to switch if necessary: amber glass, HPDE bottles and Novelia bottles stored frozen (−20°C) and refrigerated once thawed.

Material and methods 3 batches of 1% VED (10 mL) were aseptically compounded under a laminar flow hood from injectable Vfend (Pfizer) and sterile water for injection (Baxter), and stored at −20°C in amber glass (n=32; Gravis), HDPE (n=32; CAT) or Novelia (n=31; Nemera) bottles. The stability study was done according to the GERPAC-SFPC stability study guidelines. At each time point, the visual aspect was checked and voriconazole concentration (using a stability indicating HPLC-UV-diode array detector method), pH and osmolality were measured. Non-visible particle counts (by light obscuration particle count test), sterility and absence of racemisation (impurity D–(2S,3R)-voriconazole–detected by chiral HPLC) were assessed at the beginning and end of the study. Parameters were measured: when stored for 3 months at −20°C; then thawed, after 15 days at +2 to +8°C, with comparison of two thawing methods (+2 to +8°C for 6 hours or 25°C for 2 hours). Statistical analysis were performed using non-parametric tests (α < 5%) to compare containers.

Results During storage, the concentration was between 95.2±1.4% and 103.6±1.3% of the initial concentration (Co) (NS); 15 days after thawing, the concentration was between 97.1±1.6% and 98.6±0.8% of Co (NS). pH remained stable (NS). Osmolality was slightly higher in glass than in plastic containers (p=0.003). Sterility was preserved. Count of ≥10 µm particles remained <80/mL. Degradation product areas increased by a maximum of 1.45 and remained unquantifiable. No impact of the thawing method was evidenced. Impurity D was not detected.

Conclusion VED remained stable for up to 3 months at −20°C and for 15 days after thawing, with no notable difference between the three containers, allowing us to choose the most suitable.

No conflict of interest