Comparison of the morphological and biomechanical characteristics of keratoconus, forme fruste keratoconus, and normal corneas

Purpose: To explore the feasibility of corneal morphological and biomechanical parameters for keratoconus and forme fruste keratoconus diagnosis. Methods: This case-control study included a total of 517 eyes from 408 keratoconus patients (KC group), 83 eyes from 83 forme fruste keratoconus patients (FFKC group), and 158 eyes from 158 patients with normal corneas (NL group). All subjects underwent routine ophthalmologic examinations. Pentacam and Corneal Visualization Scheimpflug Technology (Corvis ST) were used to obtain corneal morphological and biomechanical parameters. Differences between groups were compared using receiver operating characteristic (ROC) curve analysis. Results: ROC analysis showed that all corneal morphological parameters and most biomechanical parameters distinguished KC from NL, with an area under the curve (AUC) greater than 0.80, of which Belin-Ambrosio enhanced ectasia total deviation index (BAD-D) and tomographic and biomechanical index (TBI) were most efficient. The AUC for distinguishing KC from NL of the BAD-D was 0.989 and the TBI was 0.993, which were not statistically significant (DeLong et al., P= .232). The BAD-D cut-off point of 1.595 provided 95.9% sensitivity for distinguishing KC from NL with 100% specificity. The TBI cut-off point of 0.515 provided 96.7% sensitivity for distinguishing KC from NL with 100% specificity. The ability of other parameters to distinguish KC from NL was lower than that of BAD and TBI. Except for central astigmatism from the anterior corneal surface (AstigF), the AUC that distinguished FFKC from NL was 0.862. The AstigF cut-off point of 4.65 provided 73.5% sensitivity for distinguishing FFKC from NL with 99.3% specificity. Other parameters distinguished FFKC from NL with low efficiency. Among them, the AUC for distinguishing FFKC from NL of the TBI was 0.722, whose cut-off point of 0.273 provided 55.4% sensitivity for distinguishing KC from NL with 79.7% specificity. Conclusion: BAD-D and TBI have the highest efficiency, sensitivity, and specificity for distinguishing KC from NL. Except for AstigF, other corneal morphological and biomechanical parameters have a relatively low ability to distinguish FFKC from NL.