Abstract
Purpose:
To examine the frequency of abnormal retinal nerve fiber layer thickness (RNFLT) and ganglion cell complex (GCC) measurements among healthy eyes and glaucoma suspects (GS) in a prospective longitudinal study.
Methods:
Normal and GS eyes with ≥24 month follow-up were included. Healthy eyes had no significant ocular history except uncomplicated cataract extraction and had 2 normal standard automated perimetry (SAP) exams, defined as glaucoma hemifield test within normal limits (WNL) and mean and pattern standard deviation at p>5%, prior to enrollment. GS consisted of eyes with ocular hypertension (IOP ≥24 mmHg) and normal optic discs and SAP; or glaucomatous optic neuropathy but normal SAP; or confirmed perimetric glaucoma in the fellow eye. Spectral-domain optical coherence tomography (SDOCT; Optovue Inc, Fremont, CA) was performed annually in normal and biannually in GS eyes. At each visit ≥3 scans with signal strength index ≥40 were acquired and the best scan was chosen for analysis. One eye was randomly selected for inclusion. Average, superior, inferior and sectoral RNFLT, and average, superior and inferior GCC parameters with p>5% were classified as “Within Normal Limits; WNL” and p<1% were classified as “outside normal limits; ONL”.
Results:
23 normal and 74 GS eyes were enrolled and followed up for 41.6±9.5 months. At baseline, 100% of normal and 91% of GS eyes had all RNFLT parameters classified as WNL, and 91% of normal and 88% of GS eyes had all GCC parameters classified as WNL. Among eyes with a WNL baseline RNFLT classification, no normal eyes and 22 (33%) GS eyes subsequently developed ONL classification consisting of abnormal average RNFLT (n=5), hemifield RNFLT (n=5) or sectoral RNFLT (n=21). Among eyes with a WNL baseline GCC classification, no normal eyes and 8 GS eyes (12%) subsequently developed an ONL classification consisting of average GCC (n=4), superior GCC (n=5) or inferior GCC (n=5). The rates of loss for all RNFLT and GCC parameters were similar (p>0.05) between the 2 groups. No eyes developed abnormal SAP during follow-up.
Conclusions:
Specificity in this sample of healthy eyes was very high for RNFLT (100%) and GCC (91%) parameters. It is unclear whether the high detection rate for abnormal sectoral RNFLT among GS eyes represents false positives or more sensitive early detection of disease progression.
Keywords: 610 nerve fiber layer •
531 ganglion cells •
550 imaging/image analysis: clinical