Subtypes of primary angle closure disease based on age-independent anterior segment optical coherence tomography parameters.

Abstract

Objective: To identify subgroups of angle closure disease by considering age-independent anterior segment parameters.

Methods: Anterior-segment optical coherence tomography (ASOCT) was performed in primary angle closure suspect (PACS) and primary angle closure glaucoma (PACG) patients. Clustering analysis using age-independent parameters, anterior chamber width (ACW), anterior vault (AV), posterior corneal arc length (PCAL), and iris area was performed. The optimum number of subgroups was determined using Bayesian Information Criterion and subjects were classified into subgroups by Gaussian Mixture Model methods.

Results: A total of 650 PACS and 411 PACG were analysed. The optimal number of subgroups of the combined PACS and PACG dataset was 3. Subgroup 1 (n = 186, 29.3%) has the largest anterior chamber dimension with large AV and total anterior chamber area, subgroup 2 (n = 16, 2.5%) has the widest ACW and shallowest anterior chamber depth (ACD), and subgroup 3 (n = 432; 68.1%) has large iris area with the smallest anterior chamber dimensions, characterised by a small ACW, AV, and PCAL. Subgroup 3 comprised a significantly greater proportion of PACG compared to PACS (74.2% vs 64.6%, p = 0.04) while subgroup 1 had the greatest proportion of PACS ≥ 70 years old, yet to have progressed to PACG.

Conclusion: We identified 3 subgroups of angle closure eyes, each characterised by distinct structural components based on ASOCT. A greater proportion of older PACS yet to have progressed to PACG belonging to the subgroup with the largest anterior chamber dimensions suggests that a more spacious anterior chamber may be associated with PACS that remains stable.