Could Automated Objective Measurements Acquired at the Preoperative Stage Estimate the Corrected Distance Visual Acuity after Corneal Cross-Linking in Keratoconus?

Abstract

Objective markers describing corneal optical density (COD), thinnest corneal thickness (TCT), and anterior (ARC) and posterior (PRC) surface radii over the 3 mm thinnest region of the cornea were investigated to provide a model for estimating corrected distance visual acuity (CDVA) after corneal cross-linking (CXL) in keratoconus. CDVA, COD, TCT, ARC, and PRC were monitored (using Pentacam™) over 1 year in patients with (1) keratoconus treated with routine CXL (2) relatively stable untreated keratoconus, and (3) age/gender-matched controls. In group 1 (n = 77), the median logMAR CDVA (mode, interquartile range) improved significantly (p < 0.01) from 0.26 (0.22, 0.12-0.65) to 0.07 (0.00, 0.02-0.21). The mean (±standard deviation, 95% confidence interval) COD (in 0-100 grey scale units) in the 0-2 mm central anterior corneal region (0-2 ant), TCT (µm), ARC (mm), and PRC (mm) changed significantly (p < 0.01), from 21.2 (± 3.70, 20.4-22.0), 454 (± 40.0, 446-462), 6.49 (± 0.71, 6.33-6.65), and 4.81 (± 0.65, 4.66-4.96) to 31.5 (± 9.19, 29.5-33.6), 423 (± 49.3, 412-434), 6.78 (± 0.80, 6.60-6.98), and 4.74 (± 0.64, 4.59-4.88), respectively, but remained stable in groups 2 (n = 23) and 3 (n = 24). Significant relationships (p < 0.01) were uncovered between postop CDVA and preop values of COD, TCT, ARC, and PRC. Multilinear regression revealed significant correlations between CDVA at 1 year and preop COD, TCT, ARC, and PRC (r2 = 0.533, r20-2ant = 0.126, r2TCT = 0.321, r2ARC = 0.506, r2PRC = 0.467). Including preop CDVA further enhanced this correlation (r2 = 0.637, r2LogMAR CDVApreop = 0.566). CXL improved CDVA, increased COD and ARC, and reduced TCT and PRC. The chance of correctly estimating the CDVA at 1 year after CXL using preoperative markers of COD, TCT, ARC, and PRC is 53%, improving to 64% with the inclusion of preoperative CDVA. Objective measurements taken at the preoperative screening stage may be useful to estimate the likely postoperative CDVA when preoperative CDVA measures are unreliable or unobtainable. ClinicalTrials.gov identifier, NCT06522789.