Purpose To test the reliability and accuracy of a standardised non-invasive imaging algorithm using optical coherence tomography angiography (OCTA) in detecting and quantifying pharmacologically induced changes in the perilimbal vasculature.

Methods In this prospective observational imaging study, 370 angiograms of 15 normal eyes and 10 eyes with nasal pterygiums were obtained using a commercially available OCTA system with split-spectrum amplitude-decorrelation angiography technology. Postprocessing of the images to quantify the area occupied by the blood vessels was performed using the Fiji software. The main outcome measures were reliability (intraobserver and interobserver agreement) and accuracy of the standardised algorithm in detecting and quantifying relative change (vasoconstriction and vasodilatation) in the area occupied by the blood vessels after instillation of topical vasoconstrictors.

Results The intraclass correlation coefficients for intraobserver and interobserver agreement were 0.91 and 0.88 (good to excellent), respectively. In normal eyes, significant vasoconstriction was noted at 5 min (35%–47%) after application of eye-drops, which peaked at 10 min (43%–63%) and was sustained until 20 min (35%–51%), followed by gradual recovery. Greatest effect was noted with a combination of 5% phenylephrine and 0.15% brimonidine tartrate (BT) eye-drops as compared with either one or two drops of BT alone, both at the 10 min (p=0.0058) and 20 min (p=0.0375) time points. This dose-dependent temporal trend was replicated in eyes with primary nasal pterygium (p=0.31).

Conclusions The findings suggest that OCTA can reliably and accurately detect and quantify relative changes in the perilimbal vasculature in both normal eyes and in eyes with pterygium.