Agreement between structural and functional tests for detecting central glaucomatous defects

Abstract

Purpose : To assess the agreement between qualitative assessment of the macula and circumpapillary RNFL (cpRNFL) using the Spectralis optical coherence tomography (OCT) Hood Glaucoma Report (HGR) and the HFA 24-2C and 10-2 test patterns for detecting central glaucomatous defects.

Methods : 127 eyes of 83 healthy, glaucoma suspect or glaucoma patients who completed OCT, 24-2C and 10-2 VF testing within six months were included.OCT cpRNFL and macula scans were qualitatively assessed based on HGR reports. Inferior and superior cpRNFL defects were defined as the presence of regions outside of normal limitsobserved on the available NSTIN plot.Inferior and superior macula defects were defined subjectively based on the color-coded ganglion cell thickness maps. Central visual field defects (CVFD)were qualitatively assessed from the 10-2 test pattern and 22 central points of the 24-2C test pattern. Superior and inferior CVFDs were defined as the presence of arcuate, partial arcuate or altitudinal defects observed on pattern deviation plots. Cohen’s K was used to assess agreement.

Results : Participant mean age was 67.6years and mean (95% CI) 24-2 MD by eye was -4.06 dB (-5.12 dB, -3.00 dB).Of 127 eyes tested, 70 (55.1%) had structural defects based on OCT reports (43 of 70 based on macula assessment and 69 of 70 based on cpRNFL assessment). 36 (28.3%) eyes had 24-2C CVFD and 52 (40.9%) had 10-2 CVFD based on PD criterion. Of these eyes, 19 had inferior structural defects (18 had macular and 19 had cpRNFL defects) with corresponding superior functional defects. A total of 36 eyes had inferior structural defects without corresponding superior 24-2C CVFDs , while 1 eye had superior CVFD without corresponding structural defects resulting in fair agreement (K=0.358; 95% CI=0.210, 0.536). 21 eyes had superior structural defects (16 had macular and 21 had cpRNFL defects )with corresponding 24-2C inferior CVFD. A total of 31 eyes had superior structural defects without corresponding functional defects while 1 eye had inferior CVFDs without corresponding superior defects resulting in moderate agreement(K=0.428; 0.287, 0.572). Similar results were observed for 10-2 testing.

Conclusions : Agreement for detecting structural and functional central defects was fair to moderate in our sample. The more structural than functional defects and more RNFL than macular defects observed needs to be confirmed in longitudinal studies.

This is a 2020 Imaging in the Eye Conference abstract.