Short-Term Ocular Biometric Changes as Predictors of Long-Term Angle Closure Progression.

Importance: Baseline ocular biometrics of the anterior segment forecast progression from primary angle closure suspect (PACS) to primary angle closure (PAC). As ocular biometrics change with aging, it is also important to understand the progression risk associated with these longitudinal anatomical changes. Objective: To assess 18-month ocular biometric changes as risk factors for progression from PACS to PAC between 36 and 72 months. Design, Setting, and Participants: This retrospective cohort study analyzed data from the prospective Zhongshan Angle Closure Prevention (ZAP) Trial, which was a randomized clinical trial conducted from September 2008 to November 2018. Participants were aged 50 to 70 years and had bilateral PACS, defined as inability to visualize pigmented trabecular meshwork in 2 or more quadrants on gonioscopy. Participants were evaluated at baseline, at 2 weeks, and at 6, 18, 36, 54, and 72 months for progression to PAC, defined as development of intraocular pressure greater than 24 mm Hg, peripheral anterior synechiae, or acute angle closure. Untreated eyes without progression at or before 18 months were eligible. Cox regression models assessed risk factors for progression. Data were analyzed from November 2023 to June 2024. Main Outcome and Measure: Progression from PACS to PAC between 36 and 72 months. Results: A total of 785 untreated eyes (759 without progression, 26 with progression) of 785 participants were analyzed (mean [SD] age, 58.7 [5.0] years; 651 females [82.9%], 134 males [17.1%]). In univariable Cox models, baseline trabecular-iris space area at 500 μm (TISA500) and 18-month change in lens vault (ΔLV) and TISA at 750 μm (ΔTISA750) were associated with progression. In age-adjusted multivariable Cox models, baseline TISA500 (hazard ratio [HR], 1.28 per -0.01 mm2; 95% CI, 1.09-1.50; P = .006) and ΔLV (HR, 1.22 per 0.1 mm; 95% CI, 1.07-1.41; P = .008) (concordance index, 0.73) or baseline TISA500 (HR, 1.31 per -0.01 mm2; 95% CI, 1.11-1.54; P = .003) and ΔTISA750 (HR, 1.06 per -0.01 mm2; 95% CI, 1.02-1.10; P = .009) (concordance index, 0.73) were more predictive than baseline TISA500 alone (HR, 1.27 per -0.01 mm2; 95% CI, 1.09-1.49; P = .007) (concordance index, 0.69). A multivariable model with categorical TISA500 in the lowest quartile (<0.031 mm2; HR, 2.65; 95% CI, 1.20-5.86; P = .03) and ΔLV in the highest quartile (>0.663 mm; HR, 2.70, 95% CI, 1.23-5.93; P = .02) independently conferred greater risk of progression (concordance index, 0.69). Conclusions and Relevance: Short-term (18-month) changes in LV and TISA750 were associated with long-term (36-72 month) angle closure progression in untreated PACS eyes, suggesting that monitoring changes in lens size or position and angle width could augment predictive performance of baseline ocular biometrics for long-term angle closure progression.