Dexamethasone and critical effect of timing on retinopathy.

PURPOSE Administration of corticosteroids soon after birth has been reported to have deleterious, protective, and no effect on retinopathy of prematurity. Conflicting results may be due to timing of corticosteroid administration. The goal of this study was to determine effects of pretreatment and late dexamethasone on retinopathy in a mouse model. METHODS The C57BL6 mouse model of oxygen-induced retinopathy (by placing animals in 75% oxygen from postnatal days 7 through 12) was used to create retinal neovascularization. Dexamethasone at 0.5 mg/kg per day was administered from day 1 through day 5 in the pretreatment group. The late-treatment group received 5 days of dexamethasone at the same dose beginning on day 12. Mice were killed at days 17 through 20, and retinal vasculature was assessed by a retinal scoring system of wholemount preparation after high-molecular-weight fluorescein-labeled dextran perfusion. In addition, retinal neovascularization was assessed by quantification of extraretinal neovascular nuclei in retinal sections. Statistical significance was defined as P: < 0.05 and was determined by the Kruskal-Wallis test, Mann-Whitney test, and Student's t-test. RESULTS Oxygen-exposed animals that received treatment with dexamethasone before oxygen exposure had an improvement in retinopathy, with a median score of 6 (5,7; 25th,75th quartiles) compared with 10 (8,11) in the untreated oxygen-exposed (P: < 0.05). The group treated late (after oxygen exposure) with dexamethasone had a median score of 10 (9,11). Pretreatment reduced extraretinal vascularization, when assessed by quantification of neovascular nuclei, to a mean +/- SEM of 19 +/- 9, significantly less than in the untreated oxygen-exposed group (55 +/- 12; P: < 0.05). No difference was observed in the late-treatment group when compared with the untreated oxygen-exposed group. Significant growth retardation, indicated by body weight, was observed in the pretreatment (P: < 0.01) and late-treatment (P: < 0. 05) groups when compared with the control group. CONCLUSIONS Timing of dexamethasone administration was critical to the inhibition of development of retinopathy in the mouse model. Degree of growth retardation, measured by body weight, also appeared to be time dependent. These data may explain the different results of clinical observations with respect to corticosteroid treatment, timing, and development of retinopathy.