Work

Comparison of corneal pachymetry evolution after accelerated corneal crosslinking in keratoconus eyes using anterior segment optical coherence tomography and Scheimpflug imaging

Year: 2025
Type: article
Study design: Observational Study
Abstract: PURPOSE: , 10 min) in keratoconus (KC) eyes as measured by Pentacam HR Scheimpflug imaging (Oculus, Germany) and Casia 2 (Tomey, Japan) anterior segment optical coherence tomography (AS-OCT). METHODS: 218 eyes covering the full topographic KC (TKC) severity spectrum that underwent A-CXL between 2017 and 2021 were examined at five time points (before, 6 weeks, 6 months, 1 year after A-CXL and at the latest follow-up) using both Pentacam HR and Casia 2. Central (CCT) and thinnest (TCT) corneal thickness were assessed. Corneal densitometry and TKC classification were assessed by Pentacam HR only. Intraoperative CCT evolution was monitored using ultrasound (Pachymeter SP-3000, Tomey, Japan) to ensure that CCT did not fall below 400 µm. RESULTS: CCT and TCT measured by Pentacam HR were significantly lower than preoperatively during the first postoperative year, while Casia 2 CCT and TCT measurements were not significantly lower than preoperatively at any time. The highest agreement of Pentacam HR and Casia 2 measurements occurred six weeks after A-CXL. Corneal densitometry peaked six weeks after A-CXL and decreased thereafter without reaching pre-A-CXL values. Ultrasound revealed significant CCT thinning during A-CXL. CONCLUSION: Corneal Scheimpflug and AS-OCT based thickness measurements are not interchangeable. Corneal thinning after A-CXL detected by Pentacam HR was not observed in Casia 2 measurements. Since higher densitometry values did not correlate with a greater inter-device measurement difference, Scheimpflug-based corneal thickness measurements should be interpreted cautiously during the first year after A-CXL.
Language: English
Field: Medicine
Open Access status: hybrid
Estimated APC paid: $3,790
OpenAlex ID: W4417019191
FWCI: 0.543
Cites: 43
Cited by: 1
Related to: