OCT criteria offer favorable performance in discriminating ODE from ONHD

Analyses of optical coherence tomography images from eyes with established diagnoses of optic nerve head drusen or optic disc edema and unaffected controls showed qualitative and quantitative criteria had reasonable sensitivity and specificity for discriminating between the two pathologies.

Key Points

Columbia, MO-Qualitative and quantitative criteria can be established to differentiate optic disc edema (ODE) from optic nerve head drusen (ONHD) on optical coherence tomography (OCT), said Lenworth N. Johnson, MD, professor of ophthalmology and neurology, Mason Eye Institute, University of Missouri-Columbia.

He presented the findings from a study in which OCT parameters were analyzed and compared in three groups of 20 eyes each that were normal or had well-documented ONHD or ODE from papilledema or another optic neuropathy. The results showed that eyes with ODE and ONHD could be differentiated with reasonable sensitivity and specificity using qualitative criteria relating to the appearance of the internal optic nerve head contour and of changes in thickness of the peripapillary subretinal hyporeflective space (SHYPS). Quantitative measurements of retinal nerve fiber layer (RNFL) thickness and of SHYPS thickness offered even better discriminative value.

"Differentiating ODE from ONHD is sometimes difficult, even when using fluorescein angiography, B-scan ultrasound, and CT scan," Dr. Johnson said. "Our study suggests OCT may offer diagnostic utility in these cases, particularly using measurements of RNFL and SHYPS thickness to identify eyes with ODE. However, our sample size is relatively small, and our findings need to be validated in additional studies."

OCT images were obtained using a specific device (Stratus OCT, Carl Zeiss Meditec), and the cases included in the study were selected from those seen over a 3-year period. In eyes with ODE, the OCT images were obtained during the acute phase when edema was present. ONHD had been diagnosed based on the presence of two or more of the following features: visible optic disc ONHD, autofluorescence on fundus photographs, calcification on B-scan ultrasound or CT scan, normal intracranial pressure on lumbar puncture, and persistence of disc elevation on follow-up exams.

The OCT scans were performed by a single technician. RNFL thickness was measured in all four quadrants on images from the fast RNFL scan. SHYPS thickness was measured from the fast optic disc scan images at distances of 0.75, 1.5, and 2 mm from the disc center.

"One limitation of our study was that the quantitative measurements were made manually with a caliper and high-resolution images displayed on the computer monitor," Dr. Johnson said. "Although we established fairly good reproducibility for test-retest differences for this method, it is still prone to variability. Incorporation of an internal caliper for the SHYPS measurement from OCT manufacturers would be helpful to addressing this issue."

Qualitative analyses

For the qualitative analyses, a diagnosis was assigned based on review of the image from the fast optic disc scan using the following criteria: ODE-elevation of the ONH with a smooth internal contour and a gradual tapering in SHYPS thickness to form a recumbent "lazy" V pattern; ONHD-an elevated ONH with a "lumpy-bumpy" contour and a rapid and abrupt decline in the SHYPS; normal-absence of any of these features.

Each image was reviewed by three examiners who were unaware of the true diagnosis.

In the qualitative analysis, OCT had a sensitivity and specificity of 63% in discriminating ODE from optic disc ONHD. Most errors occurred with incorrect classification of eyes with mild ODE as having ONHD or vice versa, noted Dr. Johnson.

In the quantitative analyses, mean RNFL thickness in all four quadrants was significantly greater in eyes with ODE than in those with ONHD. Receiver operating characteristic (ROC) curves plotted from the data showed that a cutoff of >86 µm for nasal RNFL had a sensitivity of 80% and specificity of 70% for discriminating between eyes with ODE and those with ONHD, Dr. Johnson said.

SHYPS thickness at all three measured radii also was significantly greater in eyes with ODE than in those with ONHD. In the ROC analysis, a SHYPS thickness cutoff of >127 µm had 75% sensitivity and 90% specificity in differentiating eyes with ODE from those with ONHD.