A novel index for diagnosing keratoconus

Miami-The Bowman’s ectasia index (BEI) may be a novel index for diagnosing keratoconus with a sensitivity and specificity of 100% for both, according to study results reported by Mohamed F. Abou Shousha, MD, PhD.

Dr. Abou Shousha and colleagues conducted the study to evaluate the use of a Bowman’s layer vertical topographic thickness map and the BEI, which is derived from the thickness maps for diagnosing keratoconus.

Forty eyes-20 normal and 20 with keratoconus-were studied.

The investigators used a custom-made ultra-high-resolution spectral-domain optical coherence tomography with an axial resolution of about 3 µm. Dr. Abou Shousha-who is a clinical instructor, Bascom Palmer Eye Institute, Miami-explained that a novel imaging technique was used to create the limbus-to-limbus Bowman’s layer vertical topographic thickness map.

The patients were instructed first to look straight ahead to facilitate imaging of the central cornea, then to look down and then up at calibrated fixation targets to image the superior and inferior portions of the cornea.

“Using custom-made software, we created a composite corneal image from one limbus to the other,” Dr. Abou Shousha said. “The epithelial layer and Bowman’s layer were delineated automatically and the thickness maps of the two structures were created.”

He demonstrated that in keratoconic corneas, there is localized inferior thinning of Bowman’s layer.

“Using the thinnest inferior point in Bowman’s layer, we found a highly significant difference between the keratoconic and control eyes,” he said. The thickness of Bowman’s layer in the control eyes was 13.2 µm compared with 7.3 µm in the keratoconic eyes (p < 0.001).

Regarding the sensitivity and specificity of using the Bowman’s layer inferior thinnest point for diagnosing keratoconus, Dr. Abou Shousha said the area under the curve (AUC) was 0.948, indicating a very high predictive accuracy for diagnosing keratoconus. Using a cutoff value of 11.5 µm, the sensitivity and specificity were both 93%.

Based on this, Dr. Abou Shousha and his colleagues devised the BEI, which can describe the relative thinning rather than the absolute thinning of the inferior Bowman’s layer.

“We compared the inferior thinnest point of the Bowman’s layer with the patient’s own normal Bowman’s layer thickness to ascertain the relative thinning,” Dr. Abou Shousha said.

The formula that was devised was Bowman’s minimal inferior thickness divided by the average thickness of the superior Bowman’s layer x 100.

“Using the BEI, we found a highly significant difference between keratoconic eyes and controls,” he reported. “When we correlated this with the astigmatic K readings, we saw that the BEI correlated very highly (p < 0.001), indicating that it is descriptive of the keratoconic severity.”

Furthermore, the AUC was 1, with a sensitivity and specificity of 100%, and 100% using a cutoff value of 70 in their case series for diagnosing keratoconus.

“Bowman’s layer vertical topographic thickness maps of patients with keratoconus disclose characteristic localized relative inferior thinning,” Dr. Abou Shousha said. “The BEI showed excellent accuracy, sensitivity, and specificity in diagnosing keratoconus, and was accurately correlated with the disease severity.”

For more articles in this issue of Ophthalmology Times eReport, click here.

To receive weekly clinical news and updates in ophthalmology, subscribe to the Ophthalmology Times eReport.