Factors influencing haze formation and corneal flattening, and the impact of haze on visual acuity after conventional collagen cross-linking: a 12-month retrospective study

Abstract

Background: Our aim was to determine associations of pachymetry, keratometry, and their changes with haze formation and corneal flattening after collagen cross-linking, and to analyse the relationship between postoperative haze and visual outcome. Methods: Retrospective analysis was performed on 47 eyes of 47 patients with keratoconus using the Pentacam HR Scheimpflug camera before and 1, 3, 6 and 12 months after cross-linking. Corneal backscattered light values in grey scale unit were recorded in the anterior, center and posterior corneal layers and in four concentric rings. Surface area- and thickness-corrected grey scale unit values were assessed with an additional calculation. Friedman test with post hoc Wilcoxon signed-rank test was used to analyse changes in visual acuity, pachymetry, keratometry and densitometry. Spearman’s rank correlation test was used to detect correlations of haze formation and corneal flattening with pachymetry, keratometry and their postoperative change. Generalized estimating equations analysis was used to investigate the influence of densitometry values on postoperative visual acuity after controlling for the effect of preoperative keratometry. Results: One year after treatment, significant flattening was observed in maximum and mean keratometry readings (p < 0.001). Significantly increased densitometry values were observed in three central rings compared to baseline (post hoc p < 0.0125). According to receiver operating characteristic curve, densitometry value of the anterior layer of 0–2 mm ring was the most characteristic parameter of densitometry changes after cross-linking (area under the curve = 0.936). Changes in haze significantly correlated with preoperative maximum keratometry (R = 0.303, p = 0.038) and with the changes in maximum keratometry (R = -0.412, p = 0.004). Changes in maximum keratometry correlated with preoperative maximum keratometry (R = -0.302, p = 0.038). Postoperative haze had a significant impact on uncorrected and best corrected distance visual acuity (β coefficient = 0.006, p = 0.041 and β coefficient = 0.003, p = 0.039, respectively). Conclusions: Our findings indicate that in more advanced keratoconus more significant corneal flattening effect parallel with haze formation can be observed after cross-linking. Despite significant reduction of keratometry, postoperative corneal haze may limit final visual acuity.