Repeatability of interpolated corneal tomography maps in healthy and keratoconus corneas before and after crosslinking.

Abstract

Purpose: To evaluate the repeatability of interpolated Pentacam HR elevation, densitometry and anterior chamber depth (ACD) map data before and after customized corneal crosslinking (CXL) in keratoconus (KC) and healthy eyes with low-grade myopia (MY), including effects of KC severity.

Methods: Participants with progressive KC or stable MY treated with CXL were included from prospective studies. Eyes lacking complete 6- and 12-month follow-up or acceptable quality Pentacam HR examinations were excluded. Repeatability was assessed descriptively and using linear mixed-effects models (LMMs) across groups and visits. KC severity analyses used baseline maximum keratometry (K_max), both continuously and categorically (KC high/low).

Results: Ninety-five right eyes (36 KC, 59 MY) were included. Variability was generally low and greater in KC only in central elevation (anterior, p < 0.001; posterior, p = 0.001; LMM). Nevertheless, model-based limits for absolute differences between measurements, termed repeatability limits (RLs), did not exceed 1.66 μm (maximum visit-specific descriptive RL: 2.55 μm). Descriptive comparisons indicated greater KC variability overall, mixed 6-month changes and smaller group differences at 12 months, but LMMs showed no significant post-CXL repeatability changes. Selected parameters showed 4%-8% greater variability per 1-D increase in K_max, and several elevation parameters and central ACD were more variable in KC high than KC low.

Conclusion: Interpolated Pentacam HR maps were highly repeatable, with no significant post-CXL change. Variability appeared more strongly related to KC severity than diagnosis or visit. Subsequent derivation may contribute additional variability to conventional Pentacam parameters, particularly in more advanced KC.