Abstract
Purpose::
To determine whether the locations and/or extents of localized retinal nerve fiber layer (RNFL) defects are related to the visual field abnormalities by frequency doubling technology (FDT) perimetry.
Methods::
Fifty-four eyes of 54 consecutive patients with a localized RNFL defect on RNFL photographs but a normal visual field, assessed by Humphrey C30-2 SITA-Standard program, were selected. Patient with diurnal intraocular pressure of < 21 mmHg without any medication was included. Approximation of the defect to the macula (Angle α) and width of the defect (Angle ß) were measured on RNFL photograph. Patients were divided into two groups according to the results of reliable Humphrey Matrix 30-2 threshold tests; an abnormal FDT perimetry group (Group A, n=29) and a normal FDT perimetry group (Group B, n=25). Age, systemic disease, highest IOP, central corneal thickness (CCT), Angle α, and Angle ß were compared in these two groups.
Results::
Age, systemic disease, highest IOP, and CCT were not different between the two groups. Angle α was significantly smaller in Group A than in Group B (47.0 ± 16.4º vs. 65.4 ± 14.4º, P < .001) and Angle ß was significantly larger in Group A than in Group B (29.4 ± 9.0º vs. 15.8 ± 7.9º, P < .00 1) (Mann-Whitney U test). Multivariate logistic regression analysis showed that only Angle ß was associated with FDT perimetry abnormalities (P = .002). In the area under receiver operating characteristic curve of Angle ß, highest sensitivity (86.2%) and specificity (76.0%) for FDT perimetry abnormalities was obtained at a cut-off value of 21º.
Conclusions::
Localized RNFL defects closer to the macula and wider in width were found to be correlated with FDT perimetry abnormalities. However, defect width was more strongly associated with FDT perimetry abnormalities than defect location.