Frequency-Doubling Threshold Perimetry in Glaucoma, Glaucoma Suspect, and Normal Eyes

Abstract

Purpose This study was undertaken in order to compare frequency-doubling threshold perimetry (FDT) results with standard achromatic automated perimetry (SAP) results in normal eyes, ocular-hypertensive (OHT) eyes, glaucoma-suspect eyes, and eyes with primary open-angle glaucoma (POAG) for detection of visual field loss and severity by location. Methods: Ninety patient eyes (29 OHT, 35 suspect, and 26 POAG) and 42 normal eyes were evaluated by FDT and SAP. Standard visual fields were done using the normal protocols. FDT thresholds were measured with 16 ten-degree targets, which covered most of the Humphrey-program 24-2 field area. Each target consisted of a 0.25-cycle per degree sinusoidal grating flickering at 25 Hz in counterphase. Abnormality for both tests was determined by comparison with the manufacturer’s internal normative database. A variety of definitions of “abnormal on FDT” were evaluated. The results below used the criteria that two adjacent points outside the 95% probability limits as abnormal. Thresholds for the two tests were also compared by visual-field location to determine whether defective areas overlapped. Results: Of the 61 suspect or POAG eyes, 75% (46) were abnormal on FDT and 48% (29) on SAP; 44% (27) were abnormal on both tests, with results overlapping in at least one quadrant. FDT showed more abnormal quadrants (2.74 ± 1.26) than SAP (1.70 ± 0.91). There was a significant correlation between mean deviation for SAP and that for FDT in patient eyes (p< 0.001). Conclusions: Frequency-doubling perimetry had a sensitivity of 75% when specificity was set to 90%; however, a larger number of OHT eyes were identified as abnormal than would be expected to convert to glaucoma (48%). Longitudinal follow-up of these eyes will help determine the predictive value of FDT for the development of glaucoma.