TO STUDY CORNEAL BIOMECHANICS USING CORVIS ST IN NORMALAND KERATOCONUS CORNEAS

Abstract

Aim: To compare various parameters derived during Corvis ST (CoST) measurement in normal, forme fruste keratoconus (FFKC) and keratoconus (KC) subjects. Methods: 102 eyes of 79 participants of which 43 eyes from 43 age-matched controls, 19 eyes of 15 FFKC patients and 40 eyes from 27 KC patients were included in the study. Standard technique for measurement by CoST was followed and advanced edge detection algorithm was applied to derive multiple parameters. Receiver operating curves (ROC) were set up to separately identify the predictive accuracy of the various CoST parameters to detect FFKC and KC compared to normal eyes and presented as area under the ROC (AUROC) along with its standard error and 95% CI. Results: The mean age of the participants was 25.5 + 8.3 years (median= 25 years, IQR= 21 – 29 years, range= 11-54 years) and 50 (63%) were men. In univariate ordinal and logistic regression models, increment in HCDA and A2T were found to significantly increase the likelihood of KC while increase in HCR, A1T and A2V were associated with lower likelihood of a diagnosis of KC. There were no statistically significant differences between normal eyes and those with FFKC in terms of the CoST parameters. An HCR value of < 6.02mm had the highest AUROC and showed a very high sensitivity and specificity for differentiating KC from normal eyes. Conclusion: Five CoST parameters, viz. deflection amplitude, highest concavity radius, first and second applanation time and applanation velocity at second moment showed high sensitivity and specificity in differentiating normal from KC eyes. Highest concavity radius was found to be most sensitive and specific for differentiating KC from normal corneas.