Keratoconus is the most prevalent ectatic corneal disorder, principally manifested as progressive thinning and steeping of the cornea (1, 25). Due to the high acceptance rate of corneal refractive surgeries, early detection of this disorder has been a challenging topic and represents a substantial area of interest. Studies revealed that keratoconus corneas have a lower deformation resistance and higher deformation amplitude after the air-puff force than normal ones. Thus, the theory of biomechanics alteration as the pathogenesis of keratoconus was developed, and in-vivo biomechanical parameters have been evaluated as the primary diagnostic criteria in keratoconus suspect corneas (21, 26).
Consistent with previous studies (20, 27), DA ratio, ARTh, SPA1, CBI, and TBI showed significant differences between normal and subclinical keratoconus groups. Apart from A2-length, which in theory we expect higher values in the suspect corneas and conversely presented lower values compared to normal eyes, arrays of the parameters were consistent with the fact that subclinical keratoconus corneas have decreased viscoelastic and stiffness and increased distensibility (28). Nevertheless, the reported results of A2-length were in line with other studies (29). Following IOP and CCT adjustment, A2-length, SPA1, and DA ratio lost their significant differences. However, a recent study among the Chinese population (20) stated that after adjusting for IOP and CCT, DA ratio, and ARTh lost their significant divergence between the normal and subclinical groups, but SPA1 and CBI remained significantly different between the groups. The mean estimated value of the peak distance (PD) at the highest concavity presented significantly higher in the normal group after the adjustment, which is contradictory to other studies that exhibited lower peak distance rate compared to subclinical keratoconus eyes (28).
Regarding the association between CCT and the biomechanical parameters, the outcomes of studies vary. A retrospective study on 184 normal eyes and 28 subclinical keratoconus eyes stated significant (all p-values < 0.001) correlations of CCT with DA, A1-length, A2-velocity, and radius at the highest concavity, and no significant correlations (p-value > 0.05) with A2-length, A1-velocity and peak distance (30). A cross-sectional clinical study among South Asian population confirmed significant correlation of CCT with A1-velocity (r = -0.299), DA ratio (r = -0.554), ARTh (r = 0.453), SPA1 (r = 0.649), CBI (r = -0.366), and TBI (r = -0.239), among the normal eyes (12). Furthermore, a recent study on Chinese population revealed significant (all p-value < 0.001) correlation of CCT with following Corvis ST parameters among both normal and subclinical keratoconus eyes; Integrated radius (rNL= -0.41, rSKN = -0.43), DA ratio (rNL = -0.56, rSKN = -0.49), SPA1 (rNL = 0.63, rSKN = 0.72), CBI (rNL = -0.51, rSKN = -0.4). They also stated no significant correlation between CCT and ARTh among both groups (20).
On the subject of the discriminatory value of the Corvis ST parameters, studies reported varied AUROC curve values and optimum threshold rates. Compared to the results of our study, Heidari et al. (27), in a study with quite similar inclusion criteria (subclinical keratoconus defined without the necessity of clinical keratoconus diagnosis in the other eye), reported higher AUROC curve values for SPA1 (0.779) and DA ratio (0.742), and lower AUROC curve values for ARTh (0.718), CBI (0.758), and TBI (0.828) in differentiating subclinical keratoconus from normal eyes. However, due to the use of BAD-D value in the subclinical keratoconus inclusion criteria of our study, the TBI accuracy should be discussed with caution. A novel study on 47 keratoconus suspect eyes, defined as frank keratoconus fellow eyes with normal or borderline topographically/tomographically features, described TBI (AUROC: 0.946), SPA1 (0.833), and CBI (0.822) as valuable discriminators of topographically/tomographically borderline fellow eyes from normal ones (31). Also, Ambrosio et al. (5) stated high discriminatory value of TBI (AUROC: 0.985, cut-off: 0.29, sensitivity: 0.904, specificity: 0.960), and CBI (AUROC: 0.822, cut-off: 0.07, sensitivity: 0.681, specificity: 0.823) in distinguishing very asymmetric ectasia with normal topography (VAE-NT) eyes from normal ones. Another study on 79 normal and 27 VAE-NT eyes among the Chinese population confirmed the discriminatory values of TBI (AUROC: 0.928, cut-off: 0.38, Youden index: 0.753), and CBI (AUROC: 0.860, cut-off: 0.27, Youden index: 0.642) for distinguishing early keratoconus from normal eyes (32). Other studies that evaluated the TBI discriminatory value in distinguishing VAE-NT from normal eyes reported AUROC values from 0.751 with a cut-off value of 0.259 (33) to 0.925 with a cut-off value of 0.16 (34).
Kataria et al. (12), in a cross-sectional clinical study among South Asian participants, reported a good discriminatory value of SPA1 (AUROC: 0.762, cut-off: 93.74, sensitivity: 66, specificity: 83) and CBI (AUROC: 0.775, cut-off: 0.01, sensitivity: 68, specificity: 77) in distinguishing VAE-NT from normal eyes. Compared to our results, Ren et al. conducted a study among Chinese population and reported lower AUROC curve values for CBI (AUROC: 0.703, cut-off: > 0.05, Youden index: 0.38) and ARTh (AUROC: 0.618, cut-off: ≤ 434.02, Youden index: 0.3), and higher AUROC curve values for DA ratio (AUROC: 0.684, cut-off: > 4.47, Youden index: 0.33), and SPA1 (AUROC: 0.753, cut-off: ≤ 107.3, Youden index: 0.4) in distinguishing subclinical keratoconus from normal eyes (20).
Ethnic diversity and discrepancy in the utilized inclusion criteria among the studies could be the reason for the resulting variations in the reported AUROC curve values and optimum cut-off values. Considering that most of the previous studies evaluated the non-Caucasian population and included the presence of clinical keratoconus in the fellow eye as one of the subclinical keratoconus inclusion criteria (35), the current study aimed to evaluate the diagnostic value of biomechanical parameters in subclinical keratoconus (normal appearing corneas with suspicious topographic/tomographic features) cases. The small sample size and the lack of follow-up for subclinical keratoconus cases are the major limitations of our study.