Treatment of Neovascular AMD: Artificial Intelligence in Real-world Setting

Personalized Treatment Aided by Automated Analysis of Fluid in Active Neovascular Age-related Macular Degeneration (nAMD) in a Prospective, Multicenter, Randomized Study.

The purpose of this study is to implement quantitative assessment tools for the treatment of active neovascular AMD patients in a real-world setting in order to provide advantages for both patients (treatment burden) and healthcare system (scheduling visits/treatments).

Neovascular age-related macular degeneration (nAMD) is a significant burden to health care systems in industrialized countries. Due to its chronic nature, continuous follow-up and treatment is needed to prevent significant loss of visual function in patients with nAMD. Vascular endothelial growth factor (VEGF) plays a major role in the pathomechanisms of nAMD and large multicenter trials have shown that intravitreal application of substances which intercept the VEGF pathway can interrupt the progression of nAMD and improve the visual outcome. As every single injection bears the risk of sight-threatening complications and increases the financial burden to health care providers, several studies have tested different treatment regimens, to decrease the number of applicated injections without compromising the gains in visual acuity. Thereby, strict protocols have been compared to flexible "as needed" regimens (pro re nata, PRN) and regimens with proactive increments of injection intervals (treat and extend, T&E). Studies have indicated that the outcome of anti-VEGF treatment is better in standardized clinical trials than in so-called "real world settings". This is explained by tight exclusion criteria of sponsored trials, the shorter follow-up time and the small number of patients that are treated per center, resulting in a better standard of care. PRN as well as T&E management showed disadvantages such as significant less vision gain in PRN and possible over treatment in T&E. Recently, additional treatment criteria were described to improve the patients care. Advances in diagnostic precision by SD-OCT using automated algorithms to accurately measure fluid volumes in all compartments are solid tools to determine disease activity. They allow to precisely quantifying the impact of therapeutic parameters on disease activity. Multicenter study analyses have shown that the amount of intraretinal fluid has a significant effect on vision outcome. Subretinal fluid or Pigmentepithelial detachment have been described to be less important. These findings were the basis for designing an efficient point-of-care management. Automated quantification of the fluid amount using artificial intelligence (AI) may serve as a reliable and objective method to determine the personalized point-of-care. To prove the efficacy of point-of-care management, prospective studies in real-world settings are required. More data is required to assess the outcome of real-world settings and find ways to improve treatment results, when larger amounts of patients are treated and less resources are available for decision making. The purpose of this study is to implement quantitative assessment tools for the treatment of neovascular AMD patients in a real-world setting in order to provide advantages for both patients (treatment burden) and healthcare system (scheduling visits/treatments).

Cohort 1: Quantitative assessment: Treatment decision with support of artificial intelligence software. (n=145) Cohort 2: Qualitative assessment: Treatment decision without support of artificial intelligence software. (n=145)

Patients will undergo monthly follow-up visits including fluid quantification and will be retreated in case of active disease, which is defined as: reduction of 5 EDTRS letters or more related to any (suspected) neovascular activity compared to previous visit new sub-retinal hemorrhage increase >50% in IRF volume in the central 1 mm compared to month 2 increase > 50 % in SRF volume in the central 1 mm compared to month 2 in case of NO intra- and/or subretinal fluid (=less than 10nl) in visits 2 or 3, retreat if fluid in central 1mm ≥ 10nl Presence/change of sub- and intraretinal fluid will be assessed objectively by AI software and the results will be provided during the visit to the investigator. The final decision for/against retreatment is always made by the discretion of the clinical investigator.

Patients will undergo monthly follow-up visits. Treatment will be performed in case of active disease, which is defined as: reduction of 5 EDTRS letters or more related to any (suspected) neovascular activity new sub-retinal hemorrhage any fluid In this cohort the amount of retinal fluid will not be assessed by AI software at the time of retreatment.

All patients will be treated at baseline. A loading dose of 2 additionally monthly treatments will be performed at months 1 and 2. Patients showing no intra- and/or subretinal fluid in the central 1mm subfield at month 1, no treatment will be given till any disease activity is documented. Presence/change of sub- and intraretinal fluid will be assessed objectively by AI software and the results will be provided during the visit to the investigator. The final decision for/against retreatment is always made by the discretion of the clinical investigator. Should the Investigators decision differ from the study protocol, the reason will be indicated in the CRF.

All patients will be treated at baseline. A loading dose of 2 additionally monthly treatments will be performed at months 1 and 2. Patients showing no intra- and/or subretinal fluid in the central 1mm subfield at month 1, no treatment will be given till any disease activity is documented. In this cohort the amount of retinal fluid will not be assessed by AI software at the time of retreatment.

Longitudinal changes within each group in ETDRS-BCVA and quantitative anatomic measurements in the macula assessed with noninvasive imaging

Changes in total amount of fluid in nanoliters within the central millimeter assessed with automated quantitative fluid measurement on noninvasive OCT imaging.

Inclusion Criteria Adults ≥ 50 years Active neovascular AMD (classic, occult choroidal neovascularization (CNV), RAP lesion or PCV lesion) assessed by OCT, OCTA, FA Patients who have a BCVA score better or equal 0.1 (20/200) in the study eye using ETDRS No significant fibrosis or geographic atrophy (GA) involving the fovea Willingness and ability to comply with study visits and study procedures Signed informed consent form Exclusion Criteria Hypersensitivity to Fluoresceine, Ranibizumab, Aflibercept, Brolucizumab or to any of the excipients (Polysorbate 20, Sodium dihydrogen phosphate, monohydrate, Disodium hydrogen phosphate, heptahydrate, Sodium chloride, Sucrose) Any surgical treatment of the eye within 3 months prior to baseline in the study eye History of pseudophakic cystoid macular edema (Irvine Gass Syndrome) History of glaucoma filtration surgery, corneal transplant surgery or extracapsular extraction of cataract with phacoemulsification within six months preceding Visit 0, or a history of post-operative complications within the last 12 months preceding Visit 0 in the study eye (uveitis, cyclitis etc.) History of uncontrolled glaucoma in the study eye (defined as intraocular pressure (IOP) ≥ 25 mmHg despite treatment with IOP lowering medication), or C/D Ratio >0,9 Aphakia in the study eye Presence of a retinal pigment epithelial tear involving the macula in the study eye Any concurrent intraocular condition in the study eye (e.g. advanced cataract or diabetic retinopathy) that, in the opinion of the investigator, will most likely require medical or surgical intervention during the twelve-month study period to prevent or treat visual loss that might result from that condition Active intraocular inflammation (grade trace or above) in the study eye Active or suspected ocular or periocular infection in the study eye Vitreous hemorrhage or history of rhegmatogenous retinal detachment or macular hole in the study eye Current iris neovascularization, vitreous hemorrhage, or tractional retinal detachment Evidence of current infectious blepharitis, keratitis, scleritis, or conjunctivitis in either eye Any concurrent intraocular condition in the study eye that, in the opinion of the investigator, could cause an unwanted effect on treatment efficacy, compliance or require intraocular surgery (except for cataract surgery) during the study period Presence of corneal decompensation, haze or scaring with an impact on BCVA