• Written By:
    Cornea/External Disease

    This prospective longitudinal study demonstrates the long-term ability of corneal crosslinking (CXL) to slow keratoconus progression in pediatric patients.

    Study design

    Investigators enrolled 62 eyes of 47 children, aged 18 years and younger, who demonstrated progression (i.e., an increase in Kmax by at least 1 D in the preceding year). Clinicians used an epithelium-off Siena (modified Dresden) protocol for all patients.

    Evaluation included visual parameters, Scheimpflug corneal tomography and OCT demarcation line measurements.

    Outcomes

    At 10 years, UDVA improved from 0.45 to 0.23 logMAR (P=0.0001) and CDVA improved from 0.14 to 0.1 logMAR (P=0.019). Both parameters were significantly improved over baseline at nearly every visit during the 10-year follow up. Stability was recorded in nearly 80% of patients.

    After year 8, average Kmax improvement was no longer statistically significant compared with baseline values. Approximately 14% of eyes had a Kmax increase of 1.2 D at the 3-year follow up, returning to baseline values despite earlier improvements. Between year 7 and 10, 13 eyes (9 patients) showed a Kmax progression of more than 1 D.

    Four eyes (2 children) underwent a second CXL treatment and 2 eyes (2 patients) underwent deep anterior lamellar keratoplasty (DALK). Moderate haze was present in 6 eyes (4 patients), without affecting visual acuity.

    Limitations

    The study was limited by a relatively small sample size.

    Clinical significance

    This study presents long-term outcomes (10 years) of CXL in a pediatric population. CXL effectively slowed progression in this vulnerable age group.

    Nevertheless, there was a 24% progression rate in these eyes. Precautions against eye rubbing need to be enforced as eyes undergoing retreatment were in children with persistent eye rubbing behavior. Physicians should also review the possible need for corneal transplant surgery in select patients to improve their visual outcomes.