Measuring corneal astigmatism using OCT in keratoconus

Purpose: To measure net corneal astigmatism using optical coherence tomography (OCT) (Avanti) in individuals with keratoconus and compare the repeatability and accuracy with those obtained using Scheimpflug imaging (Pentacam HR). Setting: Casey Eye Institute, Portland, Oregon. Design: Prospective cross-sectional observational study. Methods: Net corneal astigmatism was calculated in keratoconic and normal eyes using OCT and Scheimpflug imaging with 4 settings - pupil or vertex centration settings with a 3 or 4 mm circular analytical zone. Corneal elevation maps were obtained from OCT images and fitted with the Zernike polynomials to obtain net corneal astigmatism. Manifest refraction astigmatism was used to evaluate the accuracy of net corneal astigmatism measurements. The coefficient of repeatability from 2 repeated measures was calculated. Results: 46 eyes with manifest or subclinical keratoconus and 52 normal control eyes were analyzed. For OCT measurements in keratoconus, better accuracy was achieved with pupil centration and 3 mm analytical zone; however, better repeatability was achieved with vertex centration and 4 mm analytical zone (coefficient of repeatability = 0.53 diopters, the Fligner-Killeen test with Bonferroni adjustment P <.0017). Agreement with manifest refraction was significantly better with OCT compared with that using Pentacam HR (generalized mixed-effect model with Bonferroni adjustment P <.00625). No statistically significant difference was found between instruments or settings in control eyes. Conclusions: OCT was able to measure net corneal astigmatism with better accuracy and precision in keratoconic eyes than the Pentacam HR. Measurements may be more accurate using pupil centration and a smaller analytical zone in patients with keratoconus.