Predictors of long-term progression in the early manifest glaucoma trial.

PURPOSE: To determine progression factors at the end of the Early Manifest Glaucoma Trial (EMGT) based on all EMGT patients and evaluate separately patients with higher and lower baseline intraocular pressure (IOP; median split).

DESIGN: Cohort of clinical trial participants.

PARTICIPANTS: Patients with early open-angle glaucoma randomized to argon laser trabeculoplasty plus betaxolol (n = 129) or no immediate treatment (n = 126), examined every 3 months for up to 11 years.

METHODS: Cox proportional hazard analyses, expressed by hazard ratios (HRs) and 95% confidence intervals (CIs).

MAIN OUTCOME MEASURE: Time to progression, defined by perimetric and photographic disc criteria.

RESULTS: Overall progression was 67% when follow-up ended (median, 8 years). Treatment approximately halved progression risk (HR, 0.53; 95% CI, 0.39-0.72); results were similar for patients with higher and lower baseline IOP (HRs, 0.41 and 0.55). Baseline progression factors (HRs, 1.51-2.12; P<0.01) were higher IOP, exfoliation, bilateral disease, and older age, as previously reported. New baseline predictors were lower ocular systolic perfusion pressure in all patients (< or =160 mmHg; HR, 1.42; 95% CI, 1.04-1.94), cardiovascular disease history (HR, 2.75; 95% CI, 1.44-5.26) in patients with higher baseline IOP, and lower systolic blood pressure (BP) (< or =125 mmHg; HR, 0.46; 95% CI, 0.21-1.02) in patients with lower baseline IOP. Postbaseline progression factors were IOP levels at follow-up, with 12% to 13% average increase per millimeter of mercury in all patients (HRs, 1.12-1.13 per mmHg higher) and similar results in patients with higher and lower baseline IOP (HRs, 1.15 and 1.13 per mmHg higher). Disc hemorrhages (HR, 1.02; 95% CI, 1.01-1.03 per percent higher frequency) also predicted progression. Thinner central corneal thickness (CCT) (HR, 1.25; 95% CI, 1.01-1.55 per 40 microm lower) was a new significant factor, a result observed in patients with higher baseline IOP (HR, 1.42; 95% CI, 1.05-1.92 per 40 microm lower) but not lower baseline IOP, with significant IOP-CCT interaction.

CONCLUSIONS: Treatment and follow-up IOP continued to have a marked influence on progression, regardless of baseline IOP. Other significant factors were age, bilaterality, exfoliation, and disc hemorrhages, as previously determined. Lower systolic perfusion pressure, lower systolic BP, and cardiovascular disease history emerged as new predictors, suggesting a vascular role in glaucoma progression. Another new factor was thinner CCT, with results possibly indicating a preferential CCT effect with higher IOP.