Abstract
PURPOSE: To evaluate the long-term efficacy of corneal crosslinking using the standard protocol (S-CXL) and the accelerated protocol (A-CXL) over a period of 8 years. METHODS: This retrospective cohort study included 61 eyes from 61 patients with progressive keratoconus (KC). They were treated with either the S-CXL (3*30, N = 16) or the A-CXL (9*10, N = 45). Visual acuity and corneal parameters, including higher-order aberrations (HOAs), corneal tomography, and corneal densitometry, were recorded prior to the intervention and at four follow-up visits using Scheimpflug tomography. A linear mixed model with Bonferroni correction was used to analyze the longitudinal data. RESULTS: Anterior corneal curvature parameters were significantly reduced in both groups at 3 and 8 years (all p ≤ 0.019). Central and thinnest corneal thickness decreased significantly after 12 and 36 months (p < 0.001). Surface variance and vertical asymmetry indices remained unchanged postoperatively, despite the significant reduction at 8 years in the A-CXL group (p < 0.001). The greatest increase in corneal optical density was observed in both groups at 6 months in the anterior layer up to a diameter of 10 mm, which roughly normalized at 12 months. HOA coma was reduced in both groups at 36 and 96 months (all p ≤ 0.012). In contrast, only the S-CXL group showed a borderline significant improvement in visual acuity after 96 months (p = 0.049) which should be interpreted cautiously due to differences in sample size and baseline characteristics. The rate of postoperative complications was comparable in both protocols. CONCLUSION: In an 8-year long-term follow-up, both S-CXL and A-CXL successfully halted KC progression. Non-inferiority of the accelerated protocol was demonstrated.