Analysis of the outcomes of three different cross‐linking protocols for treatment of paediatric keratoconus: A multicentre randomized controlled trial

Abstract Purpose To analyse long‐term outcomes of standard cross‐linking (SCXL), accelerated cross‐linking (ACXL) and transepithelial cross‐linking (TCXL) in the treatment of progressive paediatric keratoconus regarding stability, safety and efficacy. Methods This prospective multicentre randomized controlled trial included 97 eyes of 97 paediatric patients with stages I–III ABCD keratoconus grading system, who were randomized into three groups; SCXL group (control group, n = 32; 3 mW/cm 2 × 30 min/5.4 J/cm 2 ), ACXL ( n = 33; 18 mW/cm 2 × 5 min/5.4 J/cm 2 ) and TCXL ( n = 32; 18 mW/cm 2 × 5 min/5.4 J/cm 2 ). Subjective refraction, uncorrected and corrected visual acuity, keratometry and pachymetry measurements using corneal topography were recorded preoperatively and 1, 2 and 3 years postoperatively. Results SCXL group exhibited significant successive improvements in the mean visual, refractive and keratometric parameters throughout the entire postoperative 3 years while ACXL group exhibited significant improvements in visual and keratometric parameters in the first postoperative year that remained stable in second and third postoperative years. TCXL group exhibited significant progressive deterioration in all mean parameters compared to SCXL and ACXL ( p < 0.0001). Both SCXL and ACXL revealed final 100% success rate with good stability while TCXL revealed final 22% failure rate with keratoconus progression ( p < 0.0001). Conclusion SCXL and ACXL were comparable in halting keratoconus progression and achieved good stability and safety; however, SCXL was more efficient than ACXL as it yielded greater significant postoperative mean visual, refractive and keratometric improvements achieving smoother corneal remodelling. Both SCXL and ACXL were much superior to TCXL. SCXL is the best CXL treatment option for paediatric keratoconus while ACXL is a good and effective alternative.