Abstract
Purpose :
To compare and study sensitivity, specificity, and agreement of various keratoconus progression criteria in crosslinking treated patients.
Methods :
We performed a retrospective study of 187 eyes that received crosslinking between 2016 and 2021 and were followed for at least 6 months. We analyzed for significant differences in visual acuity as well as topographic data taken with Pentacam AXL (Oculus, Wetzlar, Germany) using Wilcoxon signed-rank tests. We performed receiver-operating characteristic curve (ROC) analyses and obtained Youden indexes to determine cut-off values of topographic data and aberrometry analysis. We analyzed agreement between progression criteria with Cohen's kappa.
Results :
Patient’s mean age was 27.47±9.36 years and had a mean follow-up of 25.13±13.93 months. Uncorrected visual acuity before crosslinking treatment was 0.80±0.42 logAR (approx. 20/100 Snellen), and after surgery of 0.82±0.44 logMAR, without significant change (p = 0.643). Corrected visual acuity showed significant change (p = 0.018) after treatment and medical follow-up. Kmin, and Kmean showed significant differences before surgery and at last follow-up (p <0.05). The individual parameter with the highest Youden index and largest area under the curve was Kmean (AUC=0.965, Youden= 0.93, p <0.001). Agreement analyses show strong agreement between Kmean variations and combined criteria (using Kmin, Kmax, Kmean and thinnest pachymetry) to assess for keratoconus progression (k = 0.919, p <0.001).
Conclusions :
Current keratoconus progression criteria show acceptable sensitivity and specificity for assessing patients with crosslinking. Novel criteria of keratoconus progression that assess using multiple corneal modifications could bring these criteria to higher and unifying standards.
This abstract was presented at the 2024 ARVO Annual Meeting, held in Seattle, WA, May 5-9, 2024.