The present study tracked the longitudinal changes in the topographic findings in patients with mild KC. Oshika et al.
20 reported changes in the corneal refractive parameters over several years in patients with KC who had no history of penetrating keratoplasty. In their study,
20 the subjects were not adjusted for disease stage. Because KC is a progressive disease with various stages according to its natural course, mild KC and advanced KC may have different progression rates. Therefore, we set the inclusion criteria as mild KC of grade I or II by the Amsler-Krumeich classification,
11,12 which is based on disease evolution, and we tracked the progression of mild KC.
Corneal curvature is the clinical variable most commonly used to monitor the change in KC disease severity.
3,5,13 Several index-based classification methods based on corneal topography systems
3,13 as well as the Amsler-Krumeich classification system use the central K value for grading KC severity. In the Collaborative Longitudinal Evaluation of Keratoconus (CLEK) study,
13 whether the first definitive apical clearance lens and the flatter keratometric reading increased ≥3.00 D was used to evaluate definitive disease progression. In our study, only nine patients showed ≥3.00 D increase in the central K during the follow-up period. Therefore, we modified the criteria to ≥1.50 D increase in the central K, and we expanded the number of patients in the progression group for a more detailed statistical evaluation.
Kaplan-Meier analysis revealed that progression occurred in approximately 50% of patients with mild KC, with an increase in the central K of ≥1.50 D during 12 years (
Fig. 3). In the study by Shirayama-Suzuki et al.,
8 who investigated the progression of suspected KC compared with true KC, progression to true KC occurred in >20% of the eyes over 6 years. Li et al.
9 reported that 50% of the clinically normal fellow eyes progressed to clinical KC within 16 years. Holland et al.
21 stated that bilateral disease would be found in most patients if they were observed for a sufficient period. Thus, longer follow-ups may result in finding more progression to severe KC, although the natural course of KC appeared to be slow in most of the patients in our study.
Patients who showed increases ≥1.50 D in central K were considered the progression group, and the characteristics of this group were analyzed. Mean age and baseline BCVA were significantly lower in the progression group than in the nonprogression group (
P = 0.001 and
P = 0.045, respectively;
Table 2) In the univariate analysis, age younger than 30 years was a significant predictor for KC progression. This was consistent with the results of the CLEK study,
13 in which younger age and poorer high-contrast manifest refraction visual acuity at baseline predicted the rate of corneal curvature change. However, other demographic parameters, such as sex, history of atopy, and refractive sphere/cylinder, were not significantly different from the parameters in the nonprogression group. Many studies have demonstrated a link between KC and atopic disease during the past half century.
22 –24 The largest controlled study also found that the proportion of patients with a history of atopic disease was significantly higher in those with KC.
25,26 However, a history of atopic disease may not be associated with KC progression, though it has a positive impact on disease development. In accordance with our study, McMahon et al.
13 showed that a history of atopic disease is not associated with the rate of corneal curvature change in patients with KC. Additionally, in a study by Davis et al.,
14 a history of atopy was not involved in the changes in visual acuity in patients with KC.
Among the elevation topographic indices, anterior difference ≥0.04 mm, irregularity index 3 mm ≥6.5 D, and irregularity index 5 mm ≥6.0 D at the baseline examination and annual change rate of anterior BFS ≥0.1 D/y were significant independent predictors of KC progression. Nilforoushan et al.
27 showed that patients with suspected KC had multiple distinguishing characteristics on the anterior and posterior corneal surfaces according to the elevation topography and that they had a higher anterior maximum elevation and larger differences between the highest and lowest points on the Orbscan IIz (Bausch & Lomb Surgical) posterior elevation.
Irregularity indices at 3 and 5 mm reflect optical surface irregularities, which are proportional to the SD of the axis-independent surface curvature.
16 Irregularity indices are calculated automatically by the Orbscan IIz (Bausch & Lomb Surgical) software, based on a statistical combination of the standard deviations of the mean and toric curvatures.
28 Several studies have demonstrated that these irregularity indices are significantly higher in corneas with suspected KC than in normal corneas.
29,30
Among the corneal thickness topographic variables, thinnest pachymetry <350 μm at the baseline examination was a significant predictor for KC progression, which is in agreement with the definition of KC as noninflammatory thinning and protrusion (
Table 6). However, the difference between central pachymetry and thinnest pachymetry or decentration of the thinnest pachymetry (horizontal and vertical displacement) was not associated with KC progression. This may be related to the fact that horizontal and vertical displacement of the thinnest pachymetry on the pachymetry map is commonly associated with poor patient fixation or operator centration during acquisition of the Orbscan IIz (Bausch & Lomb Surgical) image.
27
Among keratometric variables, the yearly change rate of SKmax ≥0.15 D/y, SKmin ≥0.2 D/y, and central K ≥0.1 D/y predicted KC progression. However, baseline keratometric values were not associated with KC progression.
Additionally, significantly deeper ACD was found at the baseline examination in the progression group than in the nonprogression group. The yearly change rate of ACD was also larger in the progression group, and it was significant in the multivariate analysis. In agreement with our study results, Emre et al.
31 reported that significantly increased anterior chamber parameters, including ACD, were found according to KC severity. Kovacs et al.
32 also reported that the ACD increased significantly in a KC group compared with a control group. The increased ACD in patients with KC may be explained by the fact that the most specific changes in KC curvature are steepening and protrusion of the cornea; this deformation occurs in both the anterior and the posterior corneal surfaces of KC eyes.
In this study, we also evaluated longitudinal changes in astigmatism using vector analysis. We plotted the astigmatism of patients at baseline and at the last examination using a double-angle plot, as was done in the study of de Toledo et al.,
33 who evaluated the long-term progression of astigmatism after penetrating keratoplasty for treating patients with KC.
18,33 To the best of our knowledge, ours is the first study to evaluate the characteristics of astigmatic changes by vector analysis in patients with progressive KC.
A higher percentage of patients had astigmatism in the WTR area at the baseline examination (
Fig. 4) that changed to an oblique astigmatism area at the last examination (
Fig. 5). Considering the finding that the progression group showed more changes toward oblique astigmatism (
Fig. 6) with a larger mean absolute value of the DV than the nonprogression group (
P = 0.04), progression of KC seems to accompany the axial changes in the steep meridian from WTR astigmatism to oblique astigmatism over time. In addition, these axial changes toward oblique astigmatism may to be associated with the increase of SRAX because the proportions of patients with SRAX greater than 21° was significantly increased at the last examination compared with the baseline examination (
P < 0.001).
In agreement with our findings, de Toledo et al.
33 also stated that increasing oblique astigmatism after penetrating keratoplasty suggests recurrent KC. Significant variability in the orientation and magnitude of evolving astigmatism was observed in the study of Piñero et al.,
34 who evaluated corneal astigmatic changes by vector analysis occurring in KC corneas during a 3-year follow-up period, although the magnitude of the refractive and corneal astigmatism increased. Further studies with longer follow-up are needed to confirm the direction of vectorial changes of astigmatism in patients with KC.
In summary, mild KC tended to progress in approximately 25% of patients, and this progression lasted 3.5 years on average. Kaplan-Meier analysis revealed that the median time to increase the central K by ≥1.50 D was 12 years. Among the topographic indices, anterior difference ≥0.04 mm, irregularity index 3 mm ≥6.5 D, irregularity index 5 mm ≥6.0 D, thinnest pachymetry <350 μm at the baseline examination, and yearly change rate of anterior BFS ≥0.1 D/y, central K ≥0.1 D/y, SKmax ≥0.15 D/y, SKmin ≥0.2 D/y, and ACD ≥0.0 mm/y were significant predictors for KC progression. Additionally, the axial changes in the steep meridian from WTR astigmatism toward oblique astigmatism were associated with KC progression. These indices may be used as parameters to assess KC progression and diagnose KC. Further studies are needed to determine the cutoff criteria of these parameters for distinguishing subclinical KC and the different stages of clinical KC.