Pharmacokinetics of Melphalan After Intravitreal Injection in a Rabbit Model.

Purpose: Although widely used for vitreous seed control in retinoblastoma patients, currently there are no data on melphalan pharmacokinetics after intravitreal injections. Therefore, in this study, we characterized the ocular and systemic disposition of melphalan after intravitreal injection in the rabbit eye. Methods: New Zealand rabbits received a single intravitreal injection of 15 mu g of melphalan. Vitreous, aqueous, retina, and blood samples were collected at different times up to 12 h after the injection. Melphalan was quantitated in the biological samples using a validated high-performance liquid-chromatography technique and pharmacokinetic parameters were calculated by means of compartmental models. Results: Model-predicted melphalan maximum vitreous, aqueous, and retina concentrations were 7.8 mu g/mL, 0.024 mu g/mL, and 9.8 mu g/g tissue, respectively, attained immediately and at 0.8 and 0.25 h after intravitreal injection. Melphalan vitreous concentrations were higher than 0.3 mu g/mL for 5 h after dosing. The elimination half-life from the vitreous, aqueous humor, and retina was 1.0, 0.2, and 1.2 h, respectively. Aqueous exposure [area under the curve (AUC)] was only 0.7% of that of the vitreous AUC. Melphalan concentrations in the retina were still detectable 12 h after dosing, while plasma exposure was under the limit of quantitation. Conclusion: Intravitreal administration of 15 mu g melphalan leads to pharmacological vitreous levels with low aqueous exposure. Melphalan concentrations in the retina were measurable up to 12 h after dosing, but we report nondetectable systemic exposure in the rabbit. The results correlate with the clinical features of retinoblastoma patients that show control of vitreous seeds without systemic toxicity using intravitreal melphalan.