Deciphering AMD by Deep Phenotyping and Machine Learning- Prospective Study - PINNACLE

We will recruit patients with 1) intermediate AMD in one eye and advanced AMD in the non-study eye or 2) patients with bilateral intermediate AMD (where both eyes will be included). As some participants are symptomless at the stage of intermediate AMD, we will recruit from hospital databases including imaging databases and ophthalmology and optometry practices and engagement with Patient Societies e.g. the Macular Society and patient public involvement meetings. There will be four clinical sites performing detailed assessments on 50 patients each and, to increase sample size, an additional eight referral sites in the United Kingdom who will each enrol and follow 25 study patients by Spectral Domain Optical Coherence Tomography (SD-OCT) every 4 months. The acquired images from these referral sites will be sent to the Vienna Reading Centre for morphological identification of focal events. If a focal event is detected, participants will then be referred for a detailed, targeted assessment at either the University of Southampton or Moorfields Eye Hospital as detailed below. After consent, patients will undergo visual function tests (ETDRS visual acuity, microperimetry) and multimodal imaging including fundus photographs, OCT scans, OCT angiography, autofluorescence and adaptive optics imaging. The visual function tests will be repeated annually and the multimodal imaging will be done at 4 monthly intervals for a minimum of 1 year. Blood will be taken within the first year for DNA analysis. 200 patients (main cohort) will undergo dense retinal phenotyping at a minimum of 4 visits. Medical and smoking history, genotype and body mass index will also be included in the analysis as has been done previously. As well as structural tests, functional tests will be performed at baseline and end of the study using both microperimetry (a type of visual field test to create a "retinal sensitivity map" of the quantity of light perceived in specific parts of the retina) to identify focal changes and low luminance visual acuity to assess global changes. To increase sample size but make the study feasible an additional 200 patients at UK referral sites will undergo 4 monthly OCT and be referred to Southampton / Moorfields for dense phenotyping only if a focal event is detected by OCT.