Amsterdam 2013 Programme Satellite Meetings Registration Exhibition Virtual Exhibition Hotels Visa Letter Invitation
Search Abstracts by author or title
(results will display both Free Papers & Poster)

Corneal cross-linking and infectious keratitis: personal experience and a systematic review with a meta analysis

Poster Details

First Author: A.Abbouda SPAIN

Co Author(s):    J. Alio   D. Diaz Valle   J. Benitez del Castillo   J. Gegundez Fernandez     

Abstract Details


Collagen cross linking (CXL) of the cornea has been developed recently as a new treatment for multidrug resistant infectious keratitis. The high cost of drugs, high frequency of resistance to antibiotics, the risk of corneal melting and corneal scars makes the choice of new approaches very desirable. The photoactivation of riboflavin damages the RNA and DNA of microorganisms by oxidation processes, causing lesions in chromosomal strands. Ribo?avin induces a change in properties of the collagen and has a stiffening effect on the corneal stroma, which stabilizes it and increases its resistance to enzymatic bacteria degradation avoiding the progression of corneal melting.The aim of this study is to summarize the previously published data and evaluate the effectiveness of this treatment.


A Medline (National Library of Medicine, Bethesda, Maryland, USA) search from January 2000 to January 2013 was performed to identify all articles describing treatment of CXL in infectious keratitis. Keyword searches used the terms “Cross linking” + “Keratitis” limit to “2000–2013”, and [“Ulcerative” or “Microbial”] + “Keratitis” + “cross linking” limit to “2000–2013”. We excluded in vitro or in animal studies . Only journal articles published in English were included. Complete articles were reviewed to identify those that reported original clinical data or complications of CXL treatment.


Ultraviolet Light A and Riboflavin treatment was the same in each study. A 365-nm wavelength UVA light source was used to irradiate the ulcer, carefully avoiding the limbal area in the area of the spot diameter. The spot size of the light was adjusted to 8 mm diameter to ensure that the irradiance at the corneal surface was approximately 3 mW/cm2, as previously determined in the calibration of the unit. Drops of the B2 solution were administered every 5 minutes for the length of the procedure, The UVA light exposure and drop instillation were continued for 30 minutes. Only in the Price Vision Group centre, 36 patients were randomized to different UVA durations using a web-based random number generator. Primary outcome was healing of corneal ulcer, defined as epithelization with no progressive infiltration, block of corneal melting and secondary end point recovery of visual acuity. We evaluated the total, partial and uncomplete resolution of corneal ulcer after the procedure. Adverse outcomes: Progression of melting, corneal transplants, corneal decompensation and complications related to the procedure.


The total number of eyes treated with CXL was 104. The average age was 52.8ḟ19. At the first visit the average best corrected visual acuity (BCVA) was 0.03ḟ0.05 and at the end of follow up 0.16ḟ0.17. The infectious keratitis was associated to bacteria in 58 eyes (57%), Gram + in 44 (43%), among these 4 Mycobacterium (3.6%), and Gram – in 14 eyes (13%), Fungus in 13 eyes (12 %,), Acanthamoeba in 7 eyes (7%). In 26 eyes (25%) the microbiological culture was negative or not performed. The size of the infiltrate ranged between 0.1 and 8.5 mm with an average of 2.7mm.The mean time of re-epithelization after CXL was 20.7ḟ28.1 days (min 3-max145). It is possible to detect a long period of epithelization in older patients (Pearson -0.1, p<0.4) and in patients infected by Acanthamoeba and Aspergillus (60-140 days), Mycobacterium (120 days). We observed faster epithelization in Gram+ followed by Gram-, Acanthamoeba and fungus. The antibiotic and the antifunginal treatment were continued after CXL for 20±26.9 days. Acanthamoeba treatment was followed for 50 days after CXL. 16 eyes underwent deep or lamellar keratoplasty. Corneal transplantation was more frequent in fungus followed by acanthamoeba, Gram+ and Gram-.


Until now, there are no clear recommendations for CXL treatment even though all the papers showed good results. All patients, from the Khan study, affected by Acanthamoeba performed CXL twice (at the beginning and at 3 months) and obtained complete ulcer closure. Our findings support the use of CXL for refractory keratitis infection and we suggest using it not only in bacteria keratitis but also in Acanthamoeba and fungal keratitis without waiting for diffuse corneal melting. Longer UVA exposure duration or different irradiance may allow deeper penetration and better treatment of resistant organisms. It would be useful to modify these parameters according to the depth of infiltration, ulcer size and infectious pathogens. We estimated that the overall probability of blocking corneal melting and promoting faster epithelization in patients affected by infectious keratitis treated with CXL was 85% (95%; CI 0.77, 0.91). The estimate of response was quite similar between the studies. In conclusion, our observation supports that corneal CXL is a new way of handling infectious keratitis. The expected complications of this procedure are endothelial cell loss related to deep infiltration and reactivation of herpes simplex. CXL should be considered before undertaking emergency keratoplasty.

Financial Disclosure:


Back to previous