Abstract
Purpose:
To determine why glaucomatous eyes with an abnormal optic nerve cup-to-disc ratio (CDR) and corresponding visual field (VF) defect can show a normal peripapillary retinal nerve fiber layer (RNFL) thickness on frequency domain optical coherence tomography (fdOCT).
Methods:
A glaucoma specialist identified 12 eyes from 12 consecutive patients (66 to 85 years of age) with abnormal CDR and a corresponding 24-2 VF defect, but normal peripapillary RNFL thickness (within the 95% confidence limit) on the commercial fdOCT disc report (Cirrus, Carl Zeiss Meditec). Macular cube scans were also obtained. Using a manually corrected segmentation algorithm, the RNFL and retinal ganglion cell plus inner plexiform layer (RGC+) probability maps, as well as macular B-scans, were generated and compared to the disc and VF data. To test the hypothesis that damage may have been missed due to a relatively thicker baseline RNFL, a ratio of quadrant RNFL thicknesses {superior (S) + inferior (I)}/nasal (N), was calculated for each eye and compared to controls.
Results:
The eyes fell into one or more of the following categories: segmentation errors (2 eyes), early local damage missed/underestimated on the fdOCT report (3 eyes), coexisting macular pathology (4 eyes; epiretinal membrane (2), branch retinal vein occlusion (1) and macular edema (1)), subtle local thinning (SLT: 4 eyes), and a thicker baseline RNFL (2 eyes). In the latter case, the RNFL of the nasal quadrant was thicker than controls, as confirmed by a significantly smaller (S+I)/N ratio. All 4 eyes with SLT had subtle peripapillary RNFL thinning as compared to the opposite eye, in the disc region near 12 or 6 o’clock in the quadrant of question.
Conclusions:
OCT can miss marked glaucomatous damage due to segmentation errors, subtle local damage, underlying macular pathology, and thicker RNFL thickness when healthy. A comparison of RNFL thickness relative to the nasal quadrant might help identify the latter group of eyes, while a closer examination of OCT scans can help identify the former. The fact that subtle RNFL thinning occurred in the same disc locations for the 4 SLT eyes suggests an explanation. These disc locations receive significant input from regions that extend far beyond the region covered by the 24-2 VF, which might make the local defects seen on the 24-2 difficult to observe on OCT.