Purchase this article with an account.
Gloria J Hong, Jagger C Koerner, Marguerite C Weinert, Sharon F Freedman, David K Wallace, J Wayne Riggins, Keith J Gallaher, Sasapin Grace Prakalapakorn; Pattern of retinal vascular changes before treatment for retinopathy of prematurity. Invest. Ophthalmol. Vis. Sci. 2019;60(9):6537.
Download citation file:
© ARVO (1962-2015); The Authors (2016-present)
Retinopathy of prematurity (ROP) is a leading cause of childhood blindness. To measure vascular features in retinal images, semi-automated computer programs (eg ROPtool) have been created to more objectively describe the dilation and tortuosity of blood vessels. Studies using vessel analysis programs found that the rates of change in dilation and tortuosity were greater for infants who developed type 1 ROP than infants who did not. In this study, we used ROPtool to analyze retinal images of infants longitudinally screened and treated for ROP to quantitatively evaluate patterns of retinal vascular changes before treatment.
Using retinal images collected as part of a prospective study evaluating the use of a narrow-field retinal camera for ROP screening, we included images from 37 imaging sessions of left eyes from 8 infants screened and treated for type 1 ROP. We used ROPtool to analyze ≤2 blood vessels per quadrant using the following indices: tortuosity index (TI), maximum TI (Tmax), dilation index (DI), maximum DI (DImax), combination dilation/tortuosity indices: sum of adjusted indices (SAI), maximum SAI (SAImax), tortuosity-weighted plus (TWP), and maximum TWP (TWPmax). Using the nonparametric Wilcoxon test, indices were compared over time between temporal and nasal sides, superior and inferior sides, and between individual quadrants. An analyzable quadrant was one in which ≥1 blood vessel could be traced by ROPtool for ≥1 disc diameter in length.
Eight of 99(8.1%) infants were treated for ROP. Average number of imaging session/infant=4.6 (144(97.3%) had 3-4 analyzable quadrants). Differences were statistically significant 1 week before treatment between (1) temporal and nasal sides [TI(p=0.02), Tmax(p=0.007), DI(p=0.04), SAImax(p=0.004), TWPmax(p=0.004)] and (2) superotemporal and superonasal quadrants [TI(p=0.04), Tmax(p=0.02), DI(p=0.04), SAImax(p=0.03), TWPmax(p=0.02)].
In infants treated for ROP, temporal vessels were more dilated and tortuous than nasal vessels 1 week before treatment, especially superotemporally versus superonasally. This vascular pattern may help alert examiners to which infants will develop severe ROP needing treatment.
This abstract was presented at the 2019 ARVO Annual Meeting, held in Vancouver, Canada, April 28 - May 2, 2019.
Mosaic of left eye images with ROPtool overlay 7 (A), 4 (B), & 1 week (C) before ROP treatment. Temporal vessels became more dilated & tortuous than nasal vessels 1 week before treatment, especially superotemporally vs superonasally.
This PDF is available to Subscribers Only