Quantification of fluorescein leakage induced by DL-2-aminoadipic acid in an experimental model of chronic neovascular leakage

Abstract

Purpose : We have previously described chronic vascular leakage and retinal pathology induced by intravitreal (IVT) administration of DL-2-aminoadipic acid (DLAAA) in nonhuman primates and its angiographic correlates to wet age related macular degeneration and diabetic retinopathy. To employ the DLAAA model and compare candidate therapeutics versus existing standards of care we have evaluated novel graded and computer-based methods to quantify leakage and treatment response.

Methods : Eyes of adult African green monkeys (Chlorocebus sabaeus) received IVT injection of DLAAA (5 mg/200μL). Animals underwent weekly or biweekly fundus photography and fluorescein angiography (FA) to monitor DLAAA-induced pathology and leakage. Fourteen regions of interest (ROIs) were used to delineate inner (1), mid (2) and outer (3) portions of the superior (S), nasal (N), inferior (I) and temporal (T) retina as well as the fovea and optic nerve head. Fluorescein signal intensity within ROIs was evaluated in each eye using ImageJ (v1.51). Graded scoring was also employed to assess the size and spread of fluorescein leakage within the retina.

Results : DLAAA administration induced retinal degeneration and neovascularization with stable, persistent vascular leakage evident. Graded scoring of FA images confirmed stable leakage, largely confined to the central retina from week 7 onwards. Variability in background fluorescein signal intensity was evident in the same eye over time despite leakage severity remaining stable in the absence of treatment. Background normalization was required to compare leakage changes over time. Computer-based analysis yielded greater sensitivity when evaluating differences in severity of leakage between central and peripheral retinal regions over time, but this approach necessitates consistent anatomical positioning during image acquisition and when overlaying ROI maps to conduct intensity measurements.

Conclusions : Quantitation of changes in leakage severity is a key element in employing the DLAAA model to compare candidate therapies against existing standards of care. The relatively stable pathology and leakage induced by DLAAA beyond week 8 supports the use of computer-based analysis strategies but care must be taken in ensuring fluorescein measures are normalized to background signal intensity, and that anatomical landmarks are consistently referenced.

This abstract was presented at the 2019 ARVO Annual Meeting, held in Vancouver, Canada, April 28 - May 2, 2019.