EARLY DETECTION TO AVOID ECTASIA AFTER SURGERY

Identifying forme fruste keratoconus is essential to avoid ectasia after laser refractive surgery. But which is more effective – Scheimpflug corneal tomography, with its ability to image the posterior cornea and measure corneal thickness, or Placido topography, with its proven diagnostic algorithms and greater sensitivity to minute anterior corneal curvature abnormalities? Strong cases can be made for either, though combining topographic and pachymetric data may be more useful, according to presenters at Refractive Surgery Day at the annual meeting of the American Academy of Ophthalmology. Arguing for the superiority of Scheimpflug tomography, Michael W Belin MD, of the University of Arizona, Tucson, US, acknowledged that computer-based keratoscopes using Placido technology introduced by Stephen D Klyce PhD in 1984 were a monumental advance in corneal imaging. But he noted Placido imaging provides data only on the anterior corneal surface. While it can detect very subtle changes in corneal steepening, keratoconus also involves corneal thinning which may not be evident on the anterior surface. “It’s time we look below the surface to see the true extent of the pathology,†Dr Belin said. Scheimpflug tomography does this by providing not only a complete white-to-white elevation map of the anterior corneal surface, but also of the posterior corneal surface, as well as a pachymetric map of the entire cornea and a 3-D image of the anterior chamber. This allows diagnosis of subclinical keratoconus, which Dr Belin defines as a patient with normal or near normal corrected visual acuity, normal slit lamp findings and normal anterior contour, but abnormal posterior surface and abnormal pachymetric progression. Dr Belin suggested that the term “subclinical keratoconus†may be more useful than “forme fruste†because it clearly distinguishes patients with actual keratoconus from keratoconus suspects. He defines suspects as those with normal corrected visual acuity and slit lamp combined with asymmetric anterior curvature that is orthogonal in the principal meridians, normal or borderline posterior surface, and borderline pachymetric progression. The utility of posterior corneal surface mapping in detecting subclinical keratoconus was demonstrated by a 2010 study that used Scheimpflug tomography to examine 25 fellow eyes found to be topographically normal using Placido imaging in patients with so-called unilateral keratoconus. When analysed using the Belin/Ambrosio deviation index, 24 of the 25 eyes, or 96 per cent, showed tomographic abnormalities (Ambrosio R Jr et al. 2010 AAO/ISRS). “These were not suspects; these patients had keratoconus.†Diagnosing early keratoconus requires looking for what changes first, Dr Belin argued. His research suggests this is often the posterior surface, elevation at the thinnest point, pachymetric progression graphs and relational thickness, with anterior curvature changes often a late finding (Belin MW et al. Am J Ophthalmol 2011; 152(2): 157-162). “Often, normal anterior curvature combines with abnormal posterior curvature,†Dr Belin said. And Scheimpflug tomography is clearly better able to detect it than Placido, he concluded.

Placido offers proven sensitivity

Arguing that Placido topography is more accurate for spotting forme fruste keratoconus, Dr Klyce, of the Mount Sinai School of Medicine, New York City, US, disputed that the first signs of keratoconus appear at the posterior corneal surface. “This is patently false. Typically, corneas that develop ectasia don’t have abnormal posterior surfaces preoperatively. Check it in your own records.†Rather, the most reliable sign of early keratoconus is localised steepening of the anterior corneal surface, Dr Klyce said. And because Placido topography is based on analysing displacement of reflected mires, which amplify any surface irregularities, it is 20 times more sensitive to curvature changes than elevation maps produced by corneal tomography. When properly used with modern, evidence-based algorithms, it is much more reliable in detecting and correctly interpreting minute surface changes, he said. Dr Klyce addressed several potential criticisms of Placido imaging. The central “hole†of about 0.4mm in the Placido image does not necessarily result in an unmeasured area, in part because the patient’s eye moves during the procedure. If there is any doubt, a couple of additional photos can eliminate any blind spots. Dr Klyce discounted the potential for false positives due to asymmetric bow tie patterns produced by misaligning the Placido discs off the visual axis. “Two decades ago we developed algorithms to detect this.

This issue currently is not valid in topography.†If anything, the low precision of tomography is more likely to produce a misleading anterior curvature map. He showed several samples to demonstrate this. The possibility that highly aberrated or post-transplant corneas might break-up the reflected mires enough to make a map impossible also is not a problem, Dr Klyce said.

“If we are looking for very subtle changes in corneal topography, an almost smooth surface with a little inferior steepening, this is not an issue.†Dr Klyce also pointed out that Placido tomography is a mature diagnostic technology proven over decades, and even has standards documented by the American National Standards Institute and the International Standards Organization. He allowed that Placido imaging does not supply pachymetry readings, and that slit data is combined to provide it on some current systems. And pachymetry is an essential part of the screening procedure for refractive surgical candidates. But because of its superior sensitivity and established record “there is no better alternative for the accurate detection of forme fruste keratoconus than Placido disk topography.â€