Keratoconus is an ectatic corneal disorder characterized by a progressive noninflammatory thinning of the corneal stroma, which results in corneal protrusion, irregular astigmatism, and decreased vision [
Since first described by Luce [
Very recently, corneal visualization Scheimpflug technology (Corvis ST, Oculus, Wetzlar, Germany) has been developed to evaluate corneal biomechanics. This instrument displays corneal deformation in real time and records the deformation parameters for analyzing corneal biomechanics [
In the current study, we compared the corneal biomechanical parameters of keratoconus patients and normal controls using measurements obtained with the Corvis ST and estimated the sensitivity and specificity of these parameters for discriminating keratoconus corneas from normal corneas. To further evaluate the results obtained with the Corvis ST, we also applied Pentacam corneal tomography (Oculus, Wetzlar, Germany) to measure the anterior segment parameters.
This prospective comparative study included 120 eyes of 107 participants: 60 keratoconic eyes from 47 keratoconus patients (the KC group) and 60 normal eyes from 60 controls (the control group). One randomly selected eye of each participant in the control group and one or two keratoconic eyes in the KC group were examined. A diagnosis of keratoconus was made if the eye had an irregular cornea determined by distorted keratometry mires or distortion of the retinoscopic or ophthalmoscopic red reflex and at least one of the following slit-lamp signs: Vogt’s striae, Fleischer’s ring with an arc >2 mm, or corneal scarring consistent with keratoconus [
All participants underwent a complete ophthalmic examination, including a detailed assessment of uncorrected distance visual acuity, corrected distance visual acuity, slit-lamp microscopy, and fundus examination, intraocular pressure using Goldmann applanation tonometry (IOP-GAT, Haag-Streit, Koenz, Switzerland), corneal topography (Allegro Topolyzer; Wavelight AG, Germany), corneal tomography (Pentacam), and corneal biomechanics (Corvis ST). All measurements were taken between 09:00 and 17:00 by 2 trained ophthalmologists during the same visit. Three effective results were obtained from each instrument and the mean was utilized for analyses.
The Pentacam system (software version 1.18r15) measured the corneal tomography using a rotating Scheimpflug camera as described preciously [
The Pentacam output parameters were flat, steep, and mean keratometry; astigmatism; central corneal thicknesses; anterior chamber depth, volume, and angle; and corneal volumes at 3, 5, 7, and 10 mm (CV3 to CV10).
The Corvis ST (software version 1.00r30) allows noninvasive imaging of the cornea’s dynamic deformation response to a puff of air. A high-speed Scheimpflug camera records the deformation with full corneal cross-sections, which are then displayed in slow motion on a control panel (Figure
Cornea at stasis and maximum concavity in normal ((a) and (b)) and keratoconic ((c) and (d)) corneas. The convexity of the cornea at stasis and DA at maximum concavity is greater in the keratoconus than the normal corneas.
Corvis ST output. The output includes the IOP, CCT, and corneal biomechanical characteristics (applanation time, length, and velocity, time to the highest concavity and curvature radius, peak distance, and deformation amplitude).
During the deformation response, a precisely metered air pulse causes the cornea to move inward or flatten (the phenomena of corneal applanation), that is, the first applanation. The cornea continues to move inward until reaching a point of the highest concavity. Because the cornea is viscoelastic, it rebounds from this concavity to another point of applanation (the second applanation) and then to its normal convex curvature. The Corvis ST records throughout the deformation process and therefore gains information concerning the cornea’s viscoelastic properties and stiffness, as well as recording standard tonometry and pachymetry data [
In the current study, we also used Goldmann applanation tonometry to measure the IOP and the Pentacam to detect the CCT, although the Corvis ST can measure both IOP and CCT.
Statistical analyses were performed with SPSS version 17.0 software (SPSS for Windows, Chicago, IL). The Kolmogorov-Smirnov test was used to check for a normal distribution of quantitative data, which are here provided as the mean and standard deviation (SD). Differences between data were evaluated using Welch’s modified Student’s two-sample
The mean age of patients in the KC group was
Comparison of tomography and biomechanical parameters between the KC and control group, mean ± SD (range).
Control | KC |
| |
---|---|---|---|
Tomography | |||
Flat keratometry (diopters) | 43.27 ± 1.56 (38.5–48.8) | 48.47 ± 5.94 (40.1–69.3) | 0a |
Steep keratometry (diopters) | 44.39 ± 1.61 (39.2–49.5) | 52.09 ± 6.82 (40.8–72.9) | 0a |
|
|||
Mean keratometry (diopters) | 43.82 ± 1.54 (38.9–48.8) | 50.18 ± 6.19 (40.8–71) | 0a |
Astigmatism (diopters) | 1.12 ± 0.68 (0–3.7) | 3.63 ± 2.72 (0.1–9.3) | 0a |
|
|||
Central corneal thickness ( |
546.1 ± 30.09 (498–629) | 456.37 ± 57.45 (302–557) | 0a |
Corneal volume at 3.0 mm (mm3) | 3.94 ± 0.22 (3.6–4.6) | 3.45 ± 0.33 (2.7–4.1) | 0a |
|
|||
Corneal volume at 5.0 mm (mm3) | 11.56 ± 0.64 (10.5–13.3) | 10.57 ± 0.76 (9–12.4) | 0b |
Corneal volume at 7.0 mm (mm3) | 24.89 ± 1.4 (22.7–28.7) | 23.31 ± 1.54 (20.3–27) | 0b |
|
|||
Corneal volume at 10 mm (mm3) | 61.2 ± 3.67 (55.8–70.9) | 58.18 ± 4.01 (50.6–68.3) | 0b |
Anterior chamber angle (degree) | 39.14 ± 5.78 (28.4–63.6) | 37.25 ± 5.71 (23.6–52.5) | 0.074b |
|
|||
Anterior chamber depth (mm) | 3.17 ± 0.32 (2.2–4.03) | 3.43 ± 0.4 (2.16–4.39) | 0b |
Anterior chamber volume (mm3) | 185.2 ± 36.73 (92–276) | 203.18 ± 35.64 (124–263) | 0.007b |
|
|||
Biomechanics | |||
A-time1 (ms)c | 7.52 ± 0.43 (6.81–8.58) | 7.04 ± 0.36 (5.91–7.74) | 0b |
A-length1 (mm)d | 1.78 ± 0.27 (1.34–2.28) | 1.69 ± 0.33 (0.98–2.35) | 0.108b |
|
|||
|
0.15 ± 0.03 (0.08–0.24) | 0.17 ± 0.04 (0.11–0.26) | 0.026a |
A-time2 (ms)c | 22.18 ± 0.52 (21.27–23.3) | 22.5 ± 0.55 (21.46–23.69) | 0.001b |
|
|||
A-length2 (mm)d | 1.9 ± 0.49 (1.01–2.86) | 1.47 ± 0.46 (0.66–2.54) | 0b |
|
–0.39 ± 0.08 (–0.6 to 0.23) | –0.53 ± 0.15 (–0.88 to 0.24) | 0a |
|
|||
Highest concavity time (ms)f | 16.72 ± 0.49 (15.25–18.25) | 16.67 ± 0.94 (11.32–17.79) | 0.419a |
Highest concavity curvature (mm)g | 7.52 ± 1.05 (4.1–10.75) | 5.59 ± 2.32 (2.75–16.83) | 0b |
|
|||
Peak distance (mm)h | 4.50 ± 1.43 (2.21–5.99) | 4.50 ± 1.43 (2.15–6.29) | 0.585a |
Deformation amplitude (mm)i | 1.08 ± 0.11 (0.87–1.33) | 1.32 ± 0.19 (0.92–1.96) | 0b |
|
|||
IOP-GAT (mmHg)j | 14.85 ± 2.79 (9.5–21) | 12.09 ± 1.92 (7.5–15.5) | 0b |
aWilcoxon rank-sum test; bindependent two-sample
The ROC curve analysis showed that the DA had the greatest area under the ROC curve (AUC) among all biomechanical parameters for differentiating keratoconus from normal corneas. The AUC for the DA was 0.882 with an optimal cutoff point of 1.18 mm, sensitivity of 81.7%, specificity of 83.3%, and test accuracy of 82.5% (Figure
ROC curve (graphical plot of the sensitivity versus false positive rate) for DA. The cutoff was 1.18 mm, with 81.7% sensitivity and 83.3% specificity (test accuracy, 82.5%).
The mean DA values were
Histogram of DA for keratoconic and normal eyes.
As shown in Table
Correlation coefficient between DA and tomography parameters and intraocular pressure.
Control group | KC group | |||
---|---|---|---|---|
|
|
|
| |
Flat keratometry | −0.03 | 0.818 | 0.450 | 0 |
Steep keratometry | −0.119 | 0.365 | 0.633 | 0 |
|
||||
Mean keratometry | −0.074 | 0.573 | 0.545 | 0 |
Astigmatism | −0.232 | 0.075 | 0.608 | 0 |
|
||||
Central corneal thickness | −0.263 | 0.042 | −0.52 | 0 |
|
||||
Corneal volume at 3.0 mm | −0.263 | 0.042 | −0.431 | 0.001 |
Corneal volume at 5.0 mm | −0.262 | 0.043 | −0.264 | 0.041 |
|
||||
Corneal volume at 7.0 mm | −0.259 | 0.046 | −0.065 | 0.624 |
Corneal volume at 10 mm | −0.258 | 0.046 | 0.059 | 0.654 |
|
||||
Anterior chamber angle | 0.166 | 0.206 | −0.046 | 0.727 |
Anterior chamber depth | −0.003 | 0.983 | 0.211 | 0.105 |
|
||||
Anterior chamber volume | 0.009 | 0.943 | 0.030 | 0.823 |
IOP-GAT | −0.763 | 0 | −0.395 | 0.002 |
IOP-GAT: intraocular pressure measured by Goldmann applanation tonometry.
Scatterplots DA versus IOP-GAT (a) and CCT (b) in normal eyes and scatterplots of DA versus IOP-GAT (c) and CCT (d) in keratoconic eyes.
Keratoconus is an ectatic corneal disorder which can cause visual impairment by aggravating myopic and astigmatic conditions [
The Corvis ST monitors the deformation process of the cornea in a cross-sectional view using an ultrahigh speed Scheimpflug camera, which makes it possible to visualize dynamic changes. Because the instrument is not yet widely used, the related clinical data are very limited. We conducted this study to compare the corneal tomography and biomechanical characteristics provided by the Pentacam and Corvis ST between patients with keratoconus and age-matched controls. We found that Corvis ST offers an alternative and viable method for measuring corneal biomechanical properties. The DA had the greatest AUC among all the biomechanical parameters, but with a significant overlap between the KC and control groups.
In this study, most of the tomography characteristics of keratoconus were significantly different from those of normal corneas. The keratometry values, astigmatism, anterior chamber depth, and anterior chamber volume were significantly higher in the KC group than in the control group, whereas the corneal thickness, anterior chamber angle, and corneal volume were lower in the KC group. As in some other studies, the corneal thickness and corneal volume were significantly lower in keratoconus patients than in normal cornea [
The findings of our present study were consistent with those of Ambrosio et al. [
Although most of the biomechanical parameters were statistically different between the two groups, to differentiate keratoconus from normal corneas the DA was the most sensitive. Thus, we consider that the DA measured via Corvis ST is the most viable as a diagnostic parameter and deserves clinical attention. However, while the present study found that the DA is the best parameter for characterizing the biomechanical status of the cornea, the large overlap in ranges (1.0 to 1.4 mm) between the groups compromised its accuracy in discriminating keratoconus from normal corneas. Thus, the new biomechanical metrics should not be relied on as a stand-alone method for keratoconus diagnosis.
Our correlation analysis showed that the DA negatively correlated with IOP-GAT, CCT, and CV at 3 and 5 mm, in both groups. Hon and Lam [
Our study had some limitations. First, this was a cross-sectional study with measurements made only once. Second, the sample size was relatively small and hence statistical values may need to be interpreted with caution. Finally, the current Corvis ST software (the first generation with version 1.00r30) requires sophistication, and its outputs can only be regarded as raw data for characterizing corneas. To describe the biomechanical properties of the cornea, it is more useful to calculate the elastic modulus from the raw data. With advances in the understanding of biomechanics, the prospects for using the corneal deformation measurement will improve.
In summary, the corneal biomechanical metrics measured by the Corvis ST showed statistically significant differences between the keratoconus patients and normal controls. The DA was the most reliable indicator and may provide an additional reference for discriminating keratoconus from normal corneas. Additional research on this new technology is warranted to elucidate its full usefulness in clinical practice.
This paper has not been previously published by the authors. All authors concur with the submission.
The authors declare that there is no conflict of interests regarding the publication of this paper.
This research was supported by Grants from the National Natural Science Foundation of China (no. 81271052 and no. 31271059) and the National Basic Research Program of China (973 Program, no. 2013CB967001).