Retinal Nerve Fiber Layer Defects on Ultrawide Field Imaging and Spectral-Domain Optical Coherence Tomography: Significance for Glaucoma Detection in Adult Diabetic Patients

Abstract

Purpose : To compare retinal nerve fiber layer (RNFL) defects observed on ultrawide field imaging (UWFI) with spectral-domain optical coherence tomography (OCT) and Humphrey visual field (HVF) findings to determine their significance in adult diabetic patients.

Methods : Monoscopic color and red-free (green channel) UWFI were reviewed independently by two glaucoma specialists (BJS and AH) for the presence (none, localized, diffuse) and location (clock-hour) of RNFL defects. Disparities were adjudicated by consensus. Unweighted kappa was used to calculate agreement of RNFL defects identified on UWFI with an abnormality on the OCT-RNFL sector map. Sensitivity and specificity for detection of RNFL defects on UWFI using OCT as a reference standard was also calculated. A subset of patients with HVF within 6 months of UWFI were reviewed for the presence of a glaucomatous field deficit.

Results : 103 eyes of 54 patients had OCT-RNFL and UWFI within 6 months. Demographic data are summarized in Table 1. RNFL was gradable in 93/103 (90.3%) of UWFI. A RNFL defect was seen in 26/93 (27.9%; 16 localized, 10 diffuse) UWFI scans compared to 40/93 eyes (51.6%; 19 localized, 21 diffuse) that had a RNFL abnormality on the sector map of the OCT-RNFL. 21 of 26 eyes with a defect on UWFI (80.8%) were associated with an OCT-RNFL abnormality (kappa=0.49±0.086), in which the location of the defect observed correlated with the OCT in 15/21 (71.4%; Figure 1). Among the 10 UWFI scans where the RNFL was not gradable, 8 (80%) had a RNFL defect on OCT. Sensitivity and specificity of UWFI for detection of RNFL defects was 52.5% and 90.6%, respectively.

55 eyes from 29 patients had a HVF within 6 months of UWFI. Of these, 46/55 (83.6%) were reliable. 18 of these 46 (39.1%) were abnormal. Of the 5 eyes with a RNFL defect on UWFI and a HVF within 6 months, 3 eyes (60.0%) had an abnormal HVF. Of the 15 eyes with a OCT-RNFL abnormality and a HVF within 6 months, 10 eyes (66.7%) had an abnormal HVF.

Conclusions : RNFL defects seen on UWFI have high specificity for abnormalities on OCT with over 60% being associated with HVF abnormalities. Observation of RNFL defects or inability to grade the RNFL may warrant further evaluation for glaucoma among diabetic patients imaged with UWFI in teleretinal programs where ancillary testing for glaucoma is often unavailable.

This is an abstract that was submitted for the 2017 ARVO Annual Meeting, held in Baltimore, MD, May 7-11, 2017.

 

 

Figure 1