Detection of occludable angle with anterior segment optical coherence tomography and Pentacam as non-contact screening methods

Abstract

Abstract Purpose To evaluate diagnostic capacity for occludable anterior chamber angle detection with anterior segment optical coherence tomography (AS-OCT) and Pentacam. Methods Observational cross-sectional study with AS-OCT and Pentacam. AS-OCT measures: angle opening distance from Schwalbe line (SL) perpendicular (AOD-SL-Perp) and vertical to iris (AOD-SL-Vert), and iridotrabecular angle (ITA). Pentacam measures: anterior chamber depth (ACD), anterior chamber volume (ACV), and anterior chamber angle (ACA). We analysed Spearman’s correlation with gonioscopic classification. Area under receiver operating characteristic curves (AUCs) for occludable angle detection were compared. Agreement between iridocorneal values of methods was evaluated. Results Seventy-four left eyes of 74 patients. Correlation between temporal AS-OCT and gonioscopy: 0.83 (p < 0.0001) AOD-SL-Perp temporal, 0.82 (p < 0.0001) AOD-SL-Vert temporal, and 0.69 (p < 0.0001) ITA temporal. Correlation between AS-OCT nasal and gonioscopy: 0.74 (p < 0.0001) AOD-SL-Perp nasal, 0.74 (p < 0.0001) AOD-SL-Vert nasal, and 0.70 (p < 0.0001) ITA nasal. Correlation of Pentacam with temporal gonioscopy: 0.57 (p < 0.0001) ACD, 0.56 (p < 0.0001) ACV, and 0.63 (p < 0.0001) ACA. Correlation of Pentacam with nasal gonioscopy: 0.47 (IC 0.27–0.73, p < 0.0001) ACD, 0.49 (p < 0.0001) ACV, and 0.56 (CI 0.38–0.7, p < 0.0001) ACA. AS-OCT AUCs: AOD-SL-Perp temporal 0.89 (CI 0.80–0.95), AOD-SL-Vert 0.87 (CI 0.77–0.94), ITA temporal 0.88 (CI 0.78–0.94), AOD-SL-Perp nasal 0.83 (CI 0.72–0.91), AOD-SL-Vert nasal 0.87 (CI 0.77–0.94), and ITA nasal 0.91 (IC 0.81–0.96). Pentacam AUCs: ACD 0.76 (CI 0.64–0.85), ACV 0.75 (CI 0.63–0.84), and ACA 0.84 (CI 0.74–0.92). No statistical differences between different AUCs. Intraclass correlation coefficient (ICC) of ACA (Pentacam) with ITA temporal (AS-OCT) 0.59 and with nasal ITA nasal (AS-OCT) 0.65. Conclusion Both systems show high capacity for non-contact occludable angle detection. But agreement between methods is moderate or low.