Abstract 1070 Poster Session IV, Tuesday, 5/4 (poster 248)

Background: the role of steroid use for bronchopulmonary dysplasia (BPD) treatment on the pathogenesis of ROP remains controversial. Since the number of surviving very premature infants is increasing, the incidence of BPD did not drop as expected after the use of surfactant. This leads to an increased postnatal use of DEX in more and more immature infants. Meanwhile, a trend towards a rise in the incidence of ROP has been observed.

Objectives: the goal of this study is to assess the impact of post-natal DEX on the incidence and severity of ROP in relation to the other recognized risk factors.

Study Design: prospective study of cohort, blinded for ophthalmologic examinations.

Methods: all infants < 33 wks gestational age (GA) and/or < 1500g birthweight (BW) entering our tertiary care unit from 10/15/1997 to 11/20/1998 have been involved in the study. They were divided into 2 groups: Gr I, exposed to systemic DEX; Gr II, non exposed group. All infants underwent our routine ophthalmologic follow-up except for one extra examination in Gr I, just before DEX treatment when indicated for BPD. Fifteen parameters recognized as "risk factors" in the literature have been analyzed as confounding variables. Development of retinal vascularization, ROP incidence and severity have been closely noticed. After comparison between the 2 groups, the impact of DEX was studied using stepwise logistic regression analysis.

Results: 276 infants were born within the study period; 35 died, 19 were excluded for malformations, lack of parental consent or ophthalmologic examination before DEX treatment; 17 did not complete ophthalmologic follow-up yet. 40 infants exposed to DEX and 165 non exposed have been analyzed. GA was 28±2.4 wks (mean±SD) in Gr I and 30.6 ± 1.8 in Gr II (p<0.001), BW was 1067±401 g in Gr I and 1425±320 in Gr II (p<0.001); 86% showed hemodynamic disturbances requiring treatment within the first 2 days of life in Gr I vs 44% in Gr II (p<0.001); 68% had BPD in Gr I vs 14% in Gr II (p<0.01); 65% showed ROP in Gr I vs 10.9% in Gr II (p<0.001). Ophthalmologic examination showed no delayed vascularization, 44 infants had ROP: 25 unilateral, 19 bilateral. Maximum ROP stages (Stg) were: Stg1 n=9, Stg2 n=27, Stg2+ n=1, Stg 3 n=6, Stg5 n=1 (ie 3.9% of the infants had severe ROP).

Multivariate analysis showed an increased risk for ROP of 8.1 {2.4-483} for hemodynamic disturbances (adjusted Relative Risk {95% Confidence Interval}), of 4.1 {1.5-64.9} for DEX exposure, of 3.7 {1.4-50.5} for BW < 1000g and of 3.5 {1.3-40.7} for BPD. No other factors were found significant.

Conclusion: this study demonstrated an impact of DEX on increased risk for ROP, after adjustment for the other significant risk factors. No delay in vascularization was observed and no increase in severity. More vascular tortuosity was noticed for infants exposed to DEX which could be consistent with an angiogenetic effect of steroids on retinal vascular development.