Abstract
Purpose :
Corneal crosslinking (CXL) is a procedure that halts the progression of keratoconus. The first CXL required 30 mins for ultraviolet-A (UVA) irradiation but an accelerated method that reduces the irradiation time has been developed. In this study, we retrospectively compare 5 year outcomes of conventional and accelerated CXLs.
Methods :
Twenty-two eyes of 21 progressing keratoconus patients (12 males and 9 females, 22.0±5.7year-old) were investigated. Epithelial off CXL was performed in all cases. Following a 0.1% riboflavin instillation, UVA was irradiated at 3.0 mW/mm2 for 30 minutes in the conventional CXL (C-CXL) group (12 eyes, 11 patients), and at 18.0 mW/mm2<font size="1"> </font>for 5 minutes in the accelerated CXL (A-CXL) group (10 eyes, 10 patients). The best-corrected visual acuity (BCVA), steepest keratometric value (Ks), thinnest corneal thickness (TCT) and corneal endothelial cell density (ECD) were analyzed before and up to 5 years after CXL. Delta Ks were calculated by subtracting the postoperative Ks from the preoperative Ks. The demarcation line depth (DLD) was measured by CASIA® (Tomey) at 1 month, and corneal haze was quantified using the densitometry from the Pentacam® (Oculus) up to 1 year after CXL.
Results :
The BCVA, TCT, and ECD were not significantly different between both groups and also unchanged in both groups postoperatively. The preoperative Ks were 53.01±3.66D in the C-CXL group and 48.76±5.72D in the A-CXL group (p=0.015), and 49.86±1.36D and 46.94±2.71D at 5 years (p=0.075). When we analyzed the delta Ks, C-CXL showed significantly more flattening compared to A-CXL at 1 year (-1.36±1.56D in C-CXL and 0.67±1.18D in A-CXL; p= 0.009) and the difference increased thereafter (-3.21±3.79D and 0.61±1.70D; p=0.026 at 5 years). 7 eyes (58.3%) from C-CXL group showed flattening of their Ks by more than 1.0D at 5 years, despite no eye in the A-CXL exhibiting this. The DLD was 300.6±13.9 µm in C-CXL and 306.9±14.0 µm in A-CXL (p=0.223). The corneal densitometry was significantly higher after C-CXL than A-CXL from 1 month (20.7±3.0 and 15.4±1.4; p=0.002) to 1 year postoperatively (16.4±4.3 and 10.4±0.8, respectively; p=0.001).
Conclusions :
Both C-CXL and A-CXL were effective at halting the progression of keratoconus. A-CXL caused less haze and risks for long term continuous flattening.
This abstract was presented at the 2019 ARVO Annual Meeting, held in Vancouver, Canada, April 28 - May 2, 2019.