Purpose: To evaluate different anterior segment-optical coherence tomography (AS-OCT) scanning protocols for detecting eyes with angle closure.
Methods: This cross-sectional study enrolled 2104 subjects from a community clinic in Singapore. All participants underwent AS-OCT imaging in the dark (Visante, Carl-Zeiss, Meditec, Dublin, CA), and gonioscopy with a Goldmann 2-mirror and a Sussman lens by an experienced examiner. The anterior chamber angle in a particular quadrant was considered closed on gonioscopy if the posterior trabecular meshwork could not be seen; and on AS-OCT by the presence of contact between the iris and angle wall anterior to the scleral spur. Protocols using combinations of AS-OCT scans of different quadrants of the eye were compared.
Results: A total of 1853/2104 right eyes could be graded in all 4 quadrants. In all, 380 (20.5%) eyes had closed angles in at least 2 quadrants on gonioscopy. The AS-OCT scanning protocol that only imaged the inferior quadrant showed the best area under the receiving operating characteristic curve (AUC 0.76, sensitivity 84%, specificity 69%) for detecting angle closure. Although technically easier to obtain, the protocol that only included the nasal-temporal quadrants had the lowest AUC (AUC 0.67, sensitivity 47%, specificity 88%), whereas scanning the superior-inferior quadrants showed high sensitivity for detecting angle closure (92%), but low specificity (54%).
Conclusions: Using the results obtained in a research setting, the diagnostic performance of AS-OCT varied according to the adopted scanning protocol. The inferior quadrant-only protocol showed the best diagnostic performance but the low specificity would limit its use for population screening.