Identifying rapid glaucoma progressors at baseline using 10–2 visual field testing

Purpose To examine the added value of baseline 10-2 visual field (VF) testing for predicting rapid 24-2 VF progression rates in primary open-angle glaucoma (POAG). Design Prospective observational cohort study Subjects Subjects were participating in a prospective longitudinal study at our VA medical center and all were diagnosed with POAG (repeatable VF loss with corresponding structural damage in at least one eye) with 24-2 mean deviation (MD) better than -12dB. Methods After baseline 10-2 and 24-2 VF testing, a minimum of five good-quality 24-2 VF tests over at least 3 years were required. Longitudinal 24-2 VF testing was completed every 4 to 6 months and rapid progression was defined as MD deterioration equal to or worse than -0.5 dB, -0.75 dB, and -1.00 dB per year. Pair-wise, receiver operating characteristic (ROC), and linear mixed-effects regression analyses were employed. Results In 218 eyes of 117 subjects, 9.7 (±2.4) 24-2 VF tests were completed over 6.8 (±1.7) years. Rapid MD progression rates of ≤–0.5 dB, ≤-0.75 dB, and ≤–1.0 dB per year were present in 54 (25%), 21 (10%), and 13 (6%) eyes respectively. In multivariable regression analyses without 24-2 MD, baseline 10-2 MD (OR=0.41 [CI: 0.20,0.85], p=0.016) and central 24-2 VF defect (OR=2.6 [CI: 1.2,5.9], p=0.018) independently predicted rapid progression (≤-0.50 dB/year). In analyses with 24-2 MD but without 10-2 MD, only central 24-2 VF defect (OR=3.4 [CI: 1.0,11.0], p=0.049) independently predicted rapid progression. Sequential application of 10-2 and 24-2 MD Youden cut-offs resulted in false positive error rate of 14% and false negative error rate of 7% with only 4 eyes in the fastest progression groups falsely labeled as low risk. Absence of 10-2 VF defect at baseline was protective for developing rapid progression (odds ratio: 0.15 [95% CI: 0.07,0.31]) with only two eyes in the fastest progression groups lacking baseline 10-2 VF loss. Conclusions Baseline 10-2 VF metrics in isolation and in combination with 24-2 VF metrics predicted future rapid 24-2 VF progression with reasonably good precision. 10-2 VF testing may have unique value for baseline glaucoma risk assessment.