Comparative study of Van Herick method and gonioscopy method for evaluating anterior chamber in patients with retinal vein occlusion

Objective  To compare the detection rate of narrow anterior chamber angle (NACA) and the consistency in assessing anterior chamber angle width between the slit-lamp Van Herick (VH) method and gonioscopy in patients with retinal vein occlusion (RVO).

Methods  A diagnostic test study was performed. A total of 267 consecutive patients with RVO (534 eyes) diagnosed in the Department of Ophthalmology, Peking University Third Hospital between October 2011 and April 2021 were enrolled. Peripheral anterior chamber depth was assessed using the slit-lamp VH method, and single-mirror gonioscopy was performed concurrently. Gonioscopy served as the gold standard, and NACA was defined as grade 1-2 according to the VH grading system. For gonioscopy, the Scheie classification system was adopted: NACA was defined as narrow grade Ⅲ to Ⅳ, and an eye was diagnosed with NACA if narrow angles were present in ≥3 quadrants. The differences in NACA detection rates between the two methods were compared across four subgroups: overall RVO patients, patients with isolated RVO, RVO patients complicated with primary angle closure (PAC), and RVO patients complicated with primary angle-closure glaucoma (PACG).  

Results  In the overall RVO cohort, the NACA detection rates were 27.5% for gonioscopy and 14.0% for VH method, with a statistically significant difference ( P<0.01); the two methods exhibited good consistency (Kappa=0.57, P=0.03); the sensitivity and specificity of VH method for NACA were 49.0% and 99.2%. Among 244 eyes of patients with RVO, the NACA detection rates by gonioscopy and VH method were 18.0% and 7.4%, respectively, with a statistically significant difference ( P<0.001), and the consistency was moderate (Kappa =0.46, P=0.07); the sensitivity and specificity of VH method for NACA were 36.4% and 99.0%. For 232 fellow eyes of RVO patients, the NACA detection rates by gonioscopy and VH method were 15.5% and 7.3%, respectively, with a statistically significant difference ( P<0.001), and the consistency was moderate (Kappa=0.56, P=0.07); the sensitivity and specificity of VH method for NACA were 44.4% and 99.5%. In 30 RVO-affected eyes complicated with PAC or PACG, the NACA detection rates were 100% for gonioscopy and 70.0% for VH method; the sensitivity of VH method for NACA was 70.0%. For 28 fellow eyes of RVO patients complicated with PAC or PACG, the NACA detection rates were 100% for gonioscopy and 67.9% for VH method; the sensitivity of VH method for NACA was 67.9%.

Conclusions  The VH method shows good consistency with gonioscopy in evaluating NACAs in RVO patients. While the VH method has high specificity, it exhibits limited sensitivity, potentially resulting in missed diagnosis. For RVO patients at high risk of PACG, combined gonioscopy is recommended as a priority.