EyeWorld India September 2021 Issue

FEATURE EWAP SEPTEMBER 2021 15 superior to other procedures with a 94% 5-year survival rate compared to 65% with Descemet stripping automated endothelial keratoplasty (DSAEK) and 47% with penetrating keratoplasty (PK). Donald Tan, FRCSE, FRCSG, FRCOphth, FAMS, Singapore, approaches complex DMEK cases by first reconstituting the anatomy of the chamber by removing the abnormal iris and replacing it with an artificial iris. Then, Dr. Tan sutures with { gore-tex CV8 sutures, instead of 2, and this method provides better stability of the chamber and less chance for the DMEK graft to slip through the sides of the implant. With this approach, “DMEK can be more easily performed as a second stage,” says Dr. Tan. This use of the artificial iris is a new approach to anterior segment reconstruction which has been shown to be both safe and effective. Another promising advancement discussed in this session was human corneal endothelial cell injection (hCEC- injection) therapy proposed by Shigeru Kinoshita, MD, PhD, Japan. “Conceptually, one donor cornea can provide cells for 300 patients and is a safe and speedy procedure,” says Dr. Kinoshita. He specifically targets very mature differentiated cells for this type of cell injection therapy by utilizing rho-kinase (ROCK) and ROCK inhibitors. With one published paper on the safety and efficacy of h E - injection using partially mature cells, safety benefits show no immunological rejection, no uveitis, and no infection. Dr. Kinoshita also stated that the average endothelial cell density 5 years after cell- injection therapy is 1,200 cells per square millimeter. “This result is a bit superior and almost equal to DSAEK and DMEK procedures,” says Dr. Kinoshita. Jodhbir S. Mehta, BSc, MBBS, PhD, FRCOphth, FRCSEd, FAMS, Singapore, ended the session discussing new developments in artificial intelligence (AI) in the world of cornea. Dr. Mehta explained the various uses of AI in a multitude of corneal diseases, stating that “AI allows us to see what we cannot see.” One striking example of AI use in ophthalmology is the paper published by Poplin et al. (2018) in which cardiovascular risk factors such as age, gender, smoking status, systolic blood pressure, and major adverse cardiac events could be predicted from fundus photographs to an extremely high degree of accuracy by deep-learning models based on anatomical features such as optic disc or blood vessels. In corneal dystrophy, AI can assist in assessing a patient’s imaging results by analyzing slit-lamp photographs and monitoring protein aggregation in patients’ eyes. Comparing An 'ndoArt artificial D/'- lamellum, months postop. Source: Gerd Auffarth, MD, PhD, FEBO manual (human) assessment with AI showed high AI accuracy and precision. AI can offer much promise in any field, and Dr. Mehta explains that “the promise is seeing something we cannot interpret and using the technology as a cost-effective way to screen many images.” Use of AI in retinal disease is currently leading the way in ophthalmology advancements, but there are also plenty of applications for AI in anterior segment disease that we may see in the future. EWAP Editors’ note: Dr. Auffarth has interests with EyeYon Medical. Dr. Tan has interests with Santen, Network Medical Products, Eye Lens, and Zeiss. Dr. Shigeru Kinoshita has interests with several companies including CorneaGen. Dr. Fogla and Dr. Mehta declared no relevant financial interests. Human corneal endothelial cell injection (hCEC-injection) therapy. Source: Shigeru Kinoshita, MD, PhD