PRECISE Study
Research type
Research Study
Full title
Can deep phenotyping of retinal images predict response to intravitreal aflibercept therapy in Patients with Neovascular Age-Related Macular Degeneration?
IRAS ID
251170
Contact name
Sobha Sivaprasad
Contact email
Duration of Study in the UK
2 years, 3 months, 31 days
Research summary
Wet (neovascular) age related macular degeneration is the most common cause of blindness in the elderly. Anti-vascular endothelial growth factor drugs such as aflibercept are injected into the eye as the standard of care for this condition. Patients require monthly injections for 3 months and then seen and injected every 8 weeks based on a fixed protocol. However, the response rate with the drug is different. Human graders have attempted detailed grading of the retinal image scans (OCT and OCTA) taken as part of their care to identify responders. To date, we have not been able to precisely determine the markers of response. In this project, we aim to consent the 3000 patients across 10 NHS sites to collect their 4 consecutive retinal scans (OCT in 3000 and OCTA in approximately 1000 patients) done at initiation of treatment during their routine care and send these images to computer scientists in UCL as well as IBM, Australia and Europe. IBM has very powerful computers that may give a better idea of treatment response using artificial intelligence. A risk model can then be created to better inform patients of their expected response after 3 loading injections at an individual level.
REC name
London - Stanmore Research Ethics Committee
REC reference
19/LO/1385
Date of REC Opinion
21 Aug 2019
REC opinion
Favourable Opinion