To analyse the long-term efficacy of corneal cross-linking (CXL) in children with keratoconus and to describe criteria for the detection of pseudoprogression.Methods:
Evaluation of retrospectively collected corneal topography (Placido system) and tomography (Scheimpflug system) results after CXL in children (age 18 or younger). Twenty-five patients (33 eyes) were included. Follow-up was assessed after 1, 2, 3 and 4 years. Progression was defined as an increase in maximal keratometry (Kmax) of at least one dioptre (D) in 1 year.Results:
For the entire group, mean Kmax prior to CXL was 55.3 ± 7.3D and decreased significantly (p = 0.00001) after 1 year to 53.4 ± 7.4D. In 23 patients, the progression could be halted. Five cases of presumed progression were identified. One case showed marked steepening in Kmax 4 years after CXL, but the topographic parameters were unchanged. The tomography was repeated and showed that Kmax was stable. Two cases with limbal vernal keratoconjunctivitis (VKC) worsened both in corneal tomography and topography. After resolution of the limbal inflammation, the Kmax values returned to the values before the inflammation. We found two cases of true progression both of which had advanced keratoconus prior to CXL with a preoperative Kmax of 64.4, respectively, 75.1D.Conclusion:
Our results confirm that CXL is effective in stabilizing keratoconus in children. True progression after CXL could only be verified in two of 33 eyes in a follow-up period of 37.5 months (SD ± 10 months). Two different measuring methods can help to detect diagnostic discrepancies and prevent false conclusions. Moreover, limbal vernal changes can cause transient pseudoprogression, reversible upon sufficient treatment.