To investigate important factors in the early detection of keratoconus, this study evaluated various characteristics of corneal biomechanics and HOAs between TNC and FFKC.
Since keratoconus is a bilateral and asymmetric ocular disease, the less affected eye also has keratoconus. However, little is known about the initial corneal topography alterations in asymptomatic KC patients. Patients typically present at later stages, when they have symptoms. Klyce proposed that a fellow eye that has no clinical findings of any sort except for certain topographical changes should be diagnosed with forme fruste keratoconus[8], that “certain topographical changes” are open to question, and that Rabinowitz and colleagues found that the I-S value, which calculates the vertical gradient in corneal power across a 6 mm region, a 1.4–1.9 D gradient is consistent with a keratoconus suspect, and that a value greater than 1.9 D would be consistent with clinical keratoconus[21][22]. Current studies have slightly different definitions of the FFKC. For example, some scholars believe that topography with FFKC should have no asymmetric bowtie and no focal or inferior steepening pattern[23], while other studies have diagnosed FFKC when the patient meets the criteria for tomographic (central anterior and posterior elevations < 8 and 13 µm, respectively, with the Best Fit-Sphere (BFS) as the reference sphere) and topographic (paracentral inferior-superior dioptric asymmetry ≤ 1.4) exams[24] and the difference between FFKC and SKC (subclinical keratoconus). Miao[25] and colleagues proposed the following observations: SKC was defined as the fellow eye of clinical keratoconus with normal slit-lamp biomicroscopy but slight manifestation of topographic abnormalities, such as bow-tie patterns with skewed radial axes or inferior-superior asymmetry. FFKC was defined as the fellow eye of clinical keratoconus with normal slit-lamp biomicroscopy and no manifestation of topographic abnormalities. Based on all the above diagnoses, we proposed the following diagnostic criteria for this study: we retained the topography with no asymmetric bowtie and no focal or inferior steepening pattern and added the following diagnosis: the level of the TKC in the Pentacam was normal, shown as “-.”
In the process of exploring and diagnosing different degrees of keratoconus, it has been proposed to distinguish KC and FFKC from normal corneas by corneal topography and biomechanical parameters. Tian[26] pointed to a study in which he and his colleagues found that the majority of tomography parameters in keratoconus patients differed significantly from those in normal corneas, and other studies confirmed this conclusion. Miao discovered that the SSIv2 can distinguish KCs from normal corneas; even after being matched by CCT and bIOP, the SSIv2 continued to show greater diagnostic effectiveness for all KC groups[25]; others discovered that, when the CCT was matched, the IHD and partial DCR parameters evaluated by the Corvis ST differentiated FFKC from TNC[23]. In addition, HOAs are significantly increased in KC patients, which is the reason for the poor quality of vision in keratoconus patients with normal visual acuity[27].
The continuous progression of keratoconus is closely related to the gradual thinning of the cornea. Thin corneas are more susceptible to corneal ectasia, so it is important to consider the thickness of the cornea before performing refractive surgery to prevent corneal ectasia after surgery, which can lead to keratoconus. At present, DCR parameters are widely used in the diagnosis of ophthalmic diseases, while age, IOP and CCT affect the accuracy of DCR parameters in the diagnosis of diseases, so correction-related factors are particularly important[28][29]. Therefore, this study analyzed and compared corneal HOAs and biomechanical parameters between the TNC group and the FFKC group after matching CCTs and bIOPs by using the Corvis ST and the Pentacam. The diagnosis of FFKC in thin corneas using a combination of corneal HOAs and corneal biomechanics is novel to this study in comparison to previous research.
Our study revealed that when the CCT and bIOP matched, there were statistically significant differences in the total corneal RMS HOA, A1T, A2V, and HCT between the TNC and FFKC. The ROC curve of HCT indicates that the HCT measured by the Corvis ST can distinguish FFKC from TNC. We hypothesized that HCT may have good potential for distinguishing between FFKC and TNC when CCT and bIOP are matched.
Presently, the Scheimpflug image is mainly used to analyze the aberration of the anterior and posterior surfaces of the cornea. Some studies have reported poor reproducibility of this technique, particularly for irregular corneas[30]. However, the technique can identify very subtle corneal ectasia based on the indices generated by corneal and ocular wavefronts, which may not be detected by Placido-based neural network programs, so we speculate that corneal wavefront aberrations can generate accurate information to distinguish FFKC from TNC[31].
Twelve Pentacam-derived parameters were examined in our study; only one substantially elevated parameter in the FFKC group was the total corneal RMS HOA. The quantification of corneal HOA can be used as an objective indicator to assess the optical function of the cornea, which reflects visual function, particularly in eyes with mild to moderate corneal opacity [32]. The total corneal HOA increased with age, but the corneal spherical aberration did not significantly change with age. However, whether corneal coma increases with age remains controversial[33]. In our study, the Mann‒Whitney U test revealed that there were significant differences in age between the two groups (P < 0.05), while age in the FFKC group was lower than that in the TNC group, which sufficiently indicated that total corneal HOAs had the ability to differentiate FFKC from TNC. However, there was significant overlap between the two groups, which complicates the process of reliably and independently diagnosing FFKC.
Numerous investigations have shown that the corneal structure changes with age. It has been noted that when aging occurs, the cornea shows increased collagen fiber cross-linking and stiffness. In healthy Brazilians, Valbon et al. reported that only HCT was correlated with age[34], while in healthy Chinese subjects, Wang et al. reported that age was not independently associated with HCT and that IOP was positively associated with A1T and A2V and negatively associated with HCT[28].
In some studies, CCT and IOP were highly repeatable and reproducible in different ethnic groups, including Chinese, Hungarian, New Zealanders, and Norwegians[35][36][37][38],[28].
When the bIOP and CCT were matched, our findings demonstrated that the AUCs of the HCT, which were obtained by estimating the time at which the profile was deformed the most, were greater in the FFKC group (AUC > 0.7145), and an independent sample t test revealed a significant difference in the HCT between the FFKC and TNC groups (P < 0.05); in other words, the time at which the cornea reached the highest concavity was reduced, which is sufficient to explain the change in the deformation ability of the FRC. Jiménez et al. indicated that caffeine intake also reduced corneal deformability, as evidenced by the shorter time required to reach the HCT[39].
According to Tian's findings, compared to that of the normal group, the HCT of the clinical keratoconus group decreased, indicating a progressive loss of corneal deformation ability as keratoconus progressed[40].
In the case of different CCTs, Tian[40] found that the A1T was lower in the FFKC group than in the healthy group, while in another study [23], he and his colleagues found no significant difference between the FFKC and TNC when the CCT was controlled. Interestingly, Zhang et al.[41] reported that when the CCT was also controlled, the A1T was considerably lower in the FFKC group than in the TNC group. In our study, the A1T increased in the FFKC patients, possibly because of age, and the age of the FFKC patients was younger than that of the TNC patients. Wang et al. reported that in healthy Chinese individuals, A1T decreased significantly with increasing age, which may have caused the difference in A1T in our study [28].
Some researchers have proposed that the A2 velocity could be an additional measure of corneal elasticity[42]. In our investigation, the cornea's resilience and the time it took to attain its highest concavity varied significantly among the groups.
More specifically, A1T was greater in FFKC, whereas A2V and HCT were lower. These data indicate that corneal elasticity worsened.
To further distinguish FFKC from TCN, we performed a connected diagnosis using significantly different parameters. The set of rectified variables was used to generate a discriminant function. A1T, A2V, HCT, and total corneal HOAs were the four parameters that determined the optimal discriminant function. The discriminant function's ROC curve was 0.784, which was greater than the independent diagnosis of any parameter we calculated.
The limitation of the present study is that a validation set of eyes should be established using a larger number of eyes with FFKC to more precisely examine the sensitivity and specificity values. On the other hand, FFKC situations are difficult to find, and creating a validation set with a large number of eyes could take some time. Additionally, the outcomes were logical, and the significance levels were low.