Abstract
Purpose :
To determine whether nasal step visual field (VF) defects in glaucoma, confirmed with conventional perimetry, are truly horizontal and repeatable.
Methods :
Glaucoma patients in a prospective longitudinal study were included if: (i) all 5 VF locations (24-2, Humphrey Field Analyzer, HFA) immediately above or below the horizontal midline had a pattern standard deviation value P<0.5%, (ii) no mirror-image adjacent VF locations abnormal, and (iii) no non-edge VF locations in the non-affected hemifield with a pattern standard deviation value P<1%. A custom high-density VF test pattern containing 46 test points was created with a microperimeter (CenterVue Compass). Stimuli were located at 0, 3, 9, 15, 21, and 27 degrees of horizontal eccentricity, with 1, 5, 7, 9, 11, and 13 stimuli located at 1 degree of vertical separation, respectively at each horizontal eccentricity (Fig. 1). The scotoma angle was defined with a best-fit line with the highest vertical difference in sensitivity at each horizontal eccentricity and the fovea (Fig. 2). All eyes were tested twice on the same day to determine the reproducibility of the scotoma angle.
Results :
We included 10 (3 left and 7 right) eyes of 10 patients with a median age of 73.5 years (range: 65 to 83 yrs) and a median HFA 24-2 VF mean deviation of -9.52 dB (range: -3.69 dB to -15.94 dB). The scotoma angle varied from -8.17° to +6.28°, with the smallest (unsigned) angle being 1.33° and the median being 4.22°. The test-retest difference of the two measurements of scotoma angle had a median of 0.25° (range: 0° to 3.84°), while pointwise test-retest difference had a median (IQR) of 0 (0 to 2) dB (Fig. 2).
Conclusions :
Glaucomatous hemifield defects determined by conventional perimetry do not respect the horizontal midline. These findings are confirmed by highly reproducibility values in both pointwise sensitivity and scotoma angle. These findings have implications for mapping structural parameters to visual field sensitivity in individual patients.
This abstract was presented at the 2019 ARVO Annual Meeting, held in Vancouver, Canada, April 28 - May 2, 2019.