Measuring the depth of crosslinking demarcation line in vivo: Comparison of methods and devices. Issue 2 (February 2017)

• Record Type: Journal Article
• Authors: Thorsrud, Andreas; Sandvik, Gunhild F.; Hagem, Anne M.; Drolsum, Liv
• Is Part Of: Journal of cataract and refractive surgery. Volume 43:Issue 2(2017)
• Publication Date: 2017-02
• Subjects: 617.7
• DOI: 10.1016/j.jcrs.2017.01.003
• Languages: English
• ISSNs: 0886-3350

Abstract

Abstract : Purpose : To assess the interrelationship of different methods of measuring the demarcation line depth after corneal collagen crosslinking (CXL). Setting : University eye clinic, Oslo, Norway. Design : Prospective case series. Methods : Eyes having CXL for progressive keratoconus were evaluated 1 month after CXL by in vitro confocal microscopy (IVCM), optical coherence tomography (OCT), and Scheimpflug imaging. When applying IVCM, the depth of the CXL demarcation line was measured with 2 methods; that is, IVCM keratocyte disappearance and IVCM intensity increase. With OCT, the evaluations were made by measuring the depth of the corneal stromal demarcation line. Scheimpflug imaging was used with 2 depth‐measuring methods; that is manual Scheimpflug and objective Scheimpflug intensity change. The demarcation line depths in the central cornea were compared by the intraclass correlation coefficient (ICC) and pairwise comparison of the measured treated depth. If acceptable correlations (ICC > 0.7) were found, Bland‐Altman analysis was performed. Results : Twenty eyes of 20 patients were evaluated. Acceptable correlations were found between depth measurements using OCT–IVCM keratocyte disappearance (ICC = 0.80), OCT–IVCM intensity increase (ICC = 0.75), and IVCM intensity increase–IVCM keratocyte disappearance (ICC = 0.91). The Bland‐Altman plots of these 3 pairs showed sufficient levels of agreement. Using pairwise comparison of these pairs, the measured depths were in the same level by the OCT–IVCM intensity increase only ( P = .529). Conclusions : Scheimpflug images were inaccurate for measuring the CXL demarcation line depth. The 2 confocal microscopy methods and OCT images showed good correlation. Of these 3 pairs, only measurements with OCT and IVCM intensity increase depths were in the same level.