orneal collagen cross-linking (CXL) theoretically strengthens the cornea and subsequently halts or slows keratoconus progression through a photochemical reaction that alters the collagen matrix of the corneal stroma, creating riboflavin ultraviolet-A (UVA)-induced intrafibrillar and interfibrillar covalent bonds. 1 CXL represents a landmark in the management of corneal ectasia because it directly targets the underlying pathology (stromal instability stemming from collagen abnormalities) rather than only addressing the refractive consequences of the disorder. 2 Encouraging CXL outcomes have been extensively reported in the treatment of several ectatic disorders, including keratoconus, pellucid marginal degeneration, and ectasia induced after refractive surgery. 3-8 Mild to moderate progressive keratoconus is still the most common indication for CXL treatment, 6 but the literature is scarce in confirming the safety and effectiveness of this method for patients with advanced keratoconus, particularly those with stage 3 and 4 of the Amsler-Krumeich classification. 9 Therefore, the current long-term follow-up study aims to analyze the visual and topographic outcomes of standard CXL in a group of eyes with advanced keratoconus after 48 months of follow-up. PATIENTS AND METHODS This study is a retrospective review of 40 consecutive CXL procedures performed between 2008 and 2010 and identified 40 patients who met the study inclusion criteria of having advanced keratoconus and 4 years of follow-up. The procedures were performed at the Instituto Oftalmológico Paulista, São C ABSTRACT PURPOSE: To analyze the safety and efficacy of standard corneal collagen cross-linking (CXL) in advanced cases of progressive keratoconus after 4 years of follow-up. METHODS: A retrospective case series of patients with advanced progressive keratoconus (stages 3 and 4 of Amsler-Krumeich classification) underwent standard CXL treatment. The parameters examined were changes in uncorrected visual acuity (UDVA), corrected visual acuity (CDVA), keratometry values (mean, flat, steep, and apical), pachymetry, and endothelial cell count at the baseline and at 12, 24, and 48 months postoperatively. RESULTS: Forty eyes of 40 patients were enrolled in the study. The mean patient age was 22.5 years (range: 15 to 37 years). Both mean UDVA and CDVA remained stable during the time points; no statistically significant change was noted. Although a slight reduction was observed in all keratometric readings, a statistically significant reduction was only reached in the apical keratometry (P = .037) at 4 years after CXL. A significant reduction in the corneal thickness was also found (ultrasonic: 388 ± 49 to 379 ± 48 µm; slit-scanning: 362 ± 48 to 353 ± 51 µm); however, this change was likely not clinically meaningful. Endothelial cell count was not significantly different at the end of the study. Treatment failure or progression was noted in two patients (5%) over the follow-up period. CONCLUSIONS: Standard CXL treatment was safe and able to stabilize both visua...