The analysis of choroidal thickness measured by SD-EDI-OCT in aggressive posterior retinopathy of prematurity, threshold retinopathy of prematurity after laser photocoagulation and full-term children without treatment

This study aims to examine choroidal thickness by Spectral-Domain-Enhaced Depth Imaging -Optical Coherence Tomography (SD-EDI-OCT) in children with a history of aggressive posterior retinopathy of prematurity (AP-ROP), Threshold disease (THD) and full-term children and assess the impact of choroidal thickness on visual acuity. All infants (n = 57) aged 4-<7 years treated AP-ROP (Group-1, n=18), THD (Group-2, n=17) and healty full-term children (Group-3, n=22) examined in our outpatientclinic and had a SD-EDI-OCT as well as visual acuity (VA), gestational age (GA), birth weight (BW), axial length (AXL) , spherical equivalent (SPE), subfoveal choroidal thickness were included in the study. A retrospective review of records was performed and potential risk factors influencing postoperative VA were evaluated by using linear multivariate logistic regression. We performed horizontal SD-EDI-OCT crosshair scans through the fovea and and subfoveal choroidal thicknesses (SFCT) was measured. SFCT, VA, GA, BW were significant different between the groups. AXL was different between Group-1/-2 and Group-1/-3 but not between Group-2 and -3. In multivariate regression analysis for Group-1, -2, and -3 SFCT (Group-1 β=-0.594, p=0.036; Group-2 β=-0.677 p=0.006; Group-3 β= 0.568, p=0.002) was detected as significant independent factor for influencing LogMar VA. For Group-1 and -2 (Group-1 (p=0.002) and Group-2 (p=0.003)) SFCT was detected as significant independent factor for influencing postoperative VA unfavourably. For Group-3 GA, AXL and SFCT (p<0.004, p<0.001, p<0.0001) were detected as significant independent factors for influencing VA favourably. There was negative correlation between SFCT and AXL in all groups. Spherical equivalent was different between the groups without infleuncing VA in multivariate regression analysis. The used regression model for Group-1, -2 and -3 explain the variation of the independent risk factor SFCT with 64.7 %, 54.2 % versus 67.7 % for 3 groups. Furthermore the varition for the dependent risk factor LogMar VA for Group-1, -2, and -3 is explained with 46 %, 53 % versus 77 %. Our data demonstrate SFCT as independent risk factor for influencing postoperative VA (dependent risk factor) unfavourably. But above-stated values expressed as a percentage demonstrate that still more predictors may influence dependent factor LogMar VA. Therefore prospective randomized clinical trials including systemic risk factors especially in very low birth children like Group-1 are necessary to clarify the association of choroid circulation with ROP and VA in these group of children because of the implication that thinner choroid is associated with worse vision in these patients .