Early diagnosis of keratoconus: what difference is it making?

Why early diagnosis?

In corneal refractive surgery, early diagnosis of keratoconus (subclinical asymptomatic keratoconus) is of great importance in patients seeking surgery because it can prevent progression of the pathology after surgery and make it symptomatic (corneal ectasia), thus creating the need for diagnostic tests that provide high sensitivity (ability to detect the disease in affected subjects) with the objective that no asymptomatic subclinical keratoconus ends up not being diagnosed and thus not undergo corneal refractive surgery. Recently, Uçakhan et al1 investigated several Pentacam parameters in subclinical keratoconus, keratoconus and normal eyes. They found that the Scheimpflug system could differentiate between ectatic and normal eyes. In this study, the optimum cut-off point for posterior elevation was found to be 26.5 µm (97.7% sensitivity and 81.0% specificity). In a study by Mihaltz et al2, receiver operating characteristic (ROC) curve analysis indicated that posterior elevation was the most important criterion in the diagnosis of keratoconus. A threshold value of 15.5 µm had sensitivity of 95.1% and specificity of 94.3% for differentiating normal eyes from keratoconus. These authors found lower pachymetry readings in subclinical, early and moderate keratoconus; however, they did not find significant differences in these parameters between subclinical keratoconus and normal eyes.

In the general population, the prevalence of keratoconus is 1/2000.3 4 If the prevalence of a particular disease in a population is low, the positive predictive value tends to be low given that having a greater number of healthy people increases the number of false positives.

As the prevalence of keratoconus in the general population is low, to reach a conclusive diagnosis in early stages we would have to reconfirm the result with a second independent test since a single positive result would not be conclusive.

However, with the goal of designing screening protocols that improve the ratio of cost-effectiveness we are led to search for diagnostic criteria that maximise the prevalence of the disease in certain population groups in order to increase the positive predictive value of these diagnostic tests.

The prevalence in patients with astigmatism greater than 2 Dp increased to 14.1% (6.3% of eyes had keratoconus and 7.8% had subclinical keratoconus).

What diagnostic tools do we currently rely on?

The computerised videokeratography for keratoconus diagnosis was first introduced in the 1980s. These systems used analysis of Placido disk images to compute anterior corneal curvature. A zone of increased corneal power surrounded by zones of decreasing corneal power, inferior-superior asymmetry in corneal power and skewing of the steepest radial axes above and below the horizontal meridian was established as characteristic of keratoconus on videokeratography maps. Several indices have been developed to discriminate keratoconus and subclinical keratoconus from normal eyes, as well as progression, thus improving the sensitivity of these tests. Some of the most popular include the keratoconus prediction index of the Klyce-Maeda group and the Rabinowithz KISA index per cent.5 6

The development and increase of refractive surgery techniques caused an increase in ectatic disease, and so in the 1990s more sophisticated systems were developed; known as eye imaging instruments, these tools are used to detect suspicious corneas and to make an early diagnosis.

Orbscan I (Bausch & Lomb, Rochester, New York, USA) is an eye imaging instrument which uses slit scanning technology to provide wide-field pachymetry and anterior and posterior elevation maps as well as keratometry. Later, Orbscan II combined slit scanning with Placido-based topography analysis making it more sensitive to early diagnosis. Measurements to be taken into consideration with this instrument in order to detect keratoconus in normal subjects are: maximum rear elevation compared with the best fit sphere (BFS), irregularity in the central 3 and 5 mm zones as well as pachymetry. To assess progression in patients with keratoconus, we consider increase in apex elevation, displacement of the corneal apex, decrease in thinnest-point pachymetry and an increase in the mean simulated keratometry minimum.

Another surveyor, Pentacam (Oculus, Wetzlar, Germany), is based on the Scheimpflug principle, which provides a 3D mapping of the cornea from 25 000 actual lifting points, including measurement of anterior and posterior surfaces of the cornea, corneal thickness and angle anterior chamber. A prominent feature of Pentacam is the improved map for ectasia detection analysis by Belin/Ambrosio; in order to calculate an elevation, a reference sphere (BFS) is used with a radius depending on the K. Next, the sphere (BFS) is optimised, excluding the central area of 4 mm, thus highlighting pathologies and facilitating the detection of subclinical keratoconus. This surveyor gives us an idea of the aberrations in eyes with keratoconus; there will be a significant increase in vertical coma, primary coma and coma aberrations.7 Another index that facilitates early diagnosis of keratoconus versus normal eyes is the pachymetric progression index.8

These surveyors complement each other in daily clinical practice enabling an increase in the detection of subclinical keratoconus. Differences were detected in the values of both instruments with slightly smaller values in Pentacam versus Orbscam I.9 However, no differences were detected in the minimum pachymetric value but instead in the subsequent lifting of the best sphere fit in spite of similar curvature radii in Pentacam versus Orbscam II. This difference can be useful in order to avoid refractive surgery in patients with subclinical keratoconus.10

Thanks to all these indices, surveyors are currently the most sensitive test available to us and the most highly specific. But the search for new methods that allow us to make a diagnosis as early as possible goes even further. With the methods currently developed, we have greatly increased the capacity to detect corneas that are suspicious of keratoconus, although with it brings an increase in false positives in our clinical daily practice.

Wavefront aberrometers (Wavefront—Abbott, WaveLight Analyzer—Alcon, Maxwell—Ziemer or Zywave—Technolas PV) allow for evaluation of the Zernike coefficients of the anterior surface and posterior corneal and higher order aberrations, being increased in patients with keratoconus or subclinical keratoconus.11–13

Until recently, evaluation of corneal biomechanics was feasible only in vitro. Today we can measure using the Ocular Response Analyzer (Reichert Inc., Depew, New York, USA) to detect lower levels of corneal hysteresis and corneal resistance factor, both in keratoconus patients and healthy relatives of patients with keratoconus.14 15 Alternative parameters and their combination provide greater accuracy.16 17

The spectral-domain optical coherence tomography (OCT) (Spectralis, Heidelberg Engineering) provides details about the distribution of corneal epithelial and stromal thicknesses. The epithelium and stroma in keratoconic eyes were thinner inferotemporally and thicker supranasally compared with healthy eyes. The distribution pattern was more distinct in epithelium than in stroma.18 The OCT for ultrahigh resolution (Artemis, Arc Scan) detects a smaller thickness of the epithelium and Bowman layer, showing an epithelial pattern like a donut, with a more thickened periphery and thinned in the centre of eyes with keratoconus enabling for an earlier diagnosis.19–21 The polarisation-sensitive OCT (PS-OCT 1300, Thorlabs) can detect an increase in the average face phase retardation of the posterior surface of the cornea in keratoconus patients due to changes in the lamellar structure of collagen fibres, which may be useful in early diagnosis.22

Confocal microscopy provides a study of the microstructure of the cornea. In patients with keratoconus or subclinical keratoconus, the anterior and posterior stromal keratocyte densities are lower and stromal nerve diameter is higher, although these changes are not specific to the disease. This should be considered as an additional technique.23 24

The potential of Fourier Transform Infrared (FTIR) spectroscopy is combined with advanced chemometric processing of spectral data for the analysis of raw tears. This technique can be performed in patients with various eye diseases including keratoconus. Patients with keratoconus and patients with non-specific inflammatory eye diseases show variations in spectral ranges mainly attributed to lipids and carbohydrates. This analysis can be used as a complement to early diagnosis.25

The Optical Quality Analysis System, a double pass system (Visiometrics SL, Terrassa, Spain), has been used for the detection of keratoconus and subclinical keratoconus. The Strehl ratio and modulation transfer function measurements are seen as altered in these patients.26

What are the challenges ahead?

The challenges we face in the future are treatments that increase the effectiveness in halting disease progression in asymptomatic subclinical stages and preserving visual quality and highly specific diagnostic equipment to avoid applying unnecessary treatments in false positives.