Biomechanical and Tomographic Outcomes in Pediatric Keratoconus Treated with Conventional Epithelium-Off Corneal Collagen Cross-Linking.

Background and

Objectives

Pediatric keratoconus may progress rapidly, and treatment decisions are often made before prolonged observation is possible. This study evaluated 24-month visual, tomographic, pachymetric, and biomechanical outcomes after conventional epithelium-off corneal collagen cross-linking (CXL) using Pentacam tomography and Corvis ST dynamic Scheimpflug analysis. Materials and

Methods

This single-center observational longitudinal cohort included 28 eyes of 23 patients aged 13-18 years treated at the OCULENS Clinic, Cluj-Napoca, Romania, between 2019 and 2023. Because the study had no untreated or alternative-treatment control group, postoperative changes were interpreted as associations after CXL rather than as proof of causality. Baseline, 6-, 12-, and 24-month values were analyzed for UCVA, BCVA, Kmax, thinnest pachymetry, SP-A1, deformation amplitude (DA), first and second applanation times (A1T and A2T), highest concavity (HC) radius, and BAD-D. Repeated-measures ANOVA was used after assessment of within-eye difference normality; findings were interpreted cautiously because the analysis was eye-based, the cohort was small, and multiple outcomes were examined.

Results

UCVA improved from 0.53 ± 0.16 to 0.44 ± 0.16 logMAR and BCVA from 0.31 ± 0.11 to 0.25 ± 0.11 logMAR (both p < 0.001). Mean Kmax decreased modestly from 54.36 ± 3.11 D to 53.41 ± 2.79 D, while SP-A1 increased from 84.69 ± 4.75 to 97.39 ± 5.11 (both p < 0.001). Thinnest pachymetry showed early postoperative thinning followed by partial recovery by 24 months. DA decreased, A1T increased, A2T decreased, HC radius increased, and BAD-D decreased significantly. Kmax and SP-A1 were inversely correlated at all visits (r = -0.714 to -0.773; all p < 0.001), but these correlations were considered exploratory. Post-24-month retreatment and keratoplasty-related events were recorded descriptively and were not included in the formal 24-month model.

Conclusions

Within the prespecified 24-month analytic window, conventional epithelium-off CXL was associated with stabilization, modest visual and tomographic improvement, and a concordant biomechanical stiffening signal. The results should be interpreted as cautious observational findings rather than definitive evidence of long-term stability, because of the small sample, eye-level analysis, absence of a control group, limited follow-up, and lack of formal repeatability testing.