Short summary

This study aims to test whether retinal vascular oxygen saturation and retinal vascular fractal dimension are reliable non-invasive markers of disease activity in a 3-month prospective study of treatment-naïve patients with neovascular age-related macular degeneration (nAMD).

---

Rationale

nAMD is the most common cause of severe visual impairment in the elderly population. The disease is characterized by choroidal neovascularization (CNV) that leads to visual loss caused by retinal fluid, haemorrhage, pigment epithelial detachment and fibrosis.

Vascular endothelial growth factor (VEGF) is an important regulator of angiogenesis and vascular permeability in nAMD. VEGF-inhibitory treatment has been established as standard of care, but long-term repeated intravitreal treatment is a huge burden for patients and the health care system. Due to the high-demanding treatment regimen in nAMD, it is vital to identify non-invasive biomarkers for disease activity in order to provide individualized treatments.

Given the well-known correlation between VEGF and hypoxia, a pivotal role has been suggested for ischaemia in nAMD. As the major part of the oxygen in the blood is bound to haemoglobin, oxygen saturation in retinal vessels could provide important information on oxygenation of the inner retina. The vascular oxygen saturation of the inner retina is a functional marker that can be measured non-invasively by a spectrophotometric retinal oximeter.

Diabetic retinopathy (DR) is a well-known retinal ischaemic disease. Jorgensen et al demonstrated that higher levels of DR were associated with increased intravascular oxygen saturations. This indicates an inverse relation between the retinal oxygen saturation in the tissue and in the vessels. The reason for this could be a diffusion barrier that permits the oxygen to enter the ischaemic retinal tissue or maldistribution of the blood flow due to capillary non-perfused areas and shunts. We recently found a lower retinal venous oxygen saturation 3 month after panretinal photocoagulation (PRP) in patients with proliferative diabetic retinopathy (PDR) that responded well to treatment as compared to those with progressing disease despite photocoagulation. We speculate that successful PRP leads to higher oxygen saturation in the retinal tissue (with a corresponding lower oxygen level to be measured in the vasculature).

In a case-control study by Geirsdottir et al, it was demonstrated that the retinal vascular oxygen saturation increases with age in patients with nAMD as opposed to healthy individuals who in time will have decreased retinal oxygen saturation. This underscores the findings from DR, that patients with retinal ischaemic diseases have higher vascular oxygen saturations. Even though these results indicate some proof-of-concept, there have never been any prospective studies regarding the correlation between retinal vascular oxygen saturations and treatment-response in patients with nAMD.

The retinal vascular fractal dimension is another non-invasive marker of the retinal vasculature. Fractal patterns are common in nature and include phenomenon like snowflakes, leaves, and costal lines. It is a global structural measurement of the retinal vasculature that summarizes the density of the retinal vasculature into a single parameter: the retinal vascular fractal dimension.

We have demonstrated a lower retinal vascular fractal dimension in patients with type 1 diabetes for those who have had PRP as compared to those who were laser-naïve. This is in accordance with other studies regarding structural changes in retinal vessels after photocoagulation, and we speculate that this was caused by a lower metabolic demand of the retina after treatment aimed to increase the oxygen saturation of the retina. Also, in a 16-year prospective study we found that each 0.01-point decrement in a retinal vascular fractal dimension was predictive of a 17 per cent higher risk of long-term PDR. These results indicate that lower retinal vascular fractal dimensions are a potential marker of ischaemic retinal disease.

This study aims to test whether retinal vascular oxygen saturation and retinal vascular fractal dimension are reliable non-invasive markers of disease activity in treatment-naïve patients with nAMD.

---

Description of the cohort

We will include 50 newly diagnosed treatment-naïve eyes in patients with nAMD (cases) as well as 50 age- and gender-matched persons without AMD (controls). For cases, measurements of retinal oxygen saturation and fractal dimension will be performed at both eyes at baseline and month 3. For controls, measurements will be performed in one eye (preferably the right eye) at baseline only.

Controls will be selected amongst patients receiving cataract surgery, and the measurements will be made at the follow-up examination the day after the surgery.

Cases will not be excluded because of age or gender.

All patients will be recruited at the Department of Ophthalmology Odense University Hospital, at their respective preliminary examinations for nAMD and cataract surgery.

Cases:
Inclusion: Newly diagnosed treatment-naïve patients with nAMD
Exclusion: Previously laser-treatment at the eye with nAMD

Controls:
Inclusion: Age 50-90 years, Age- and gender-matched with cases
Exclusion: No AMD at the examined eye, No previously laser treatment at the examined eye

Data and biological material

At baseline, data will be collected on age, gender, history of eye diseases, best-corrected visual acuity (BCVA) (as measured by the Early Treatment Diabetic Retinopathy Study scale). Measurement of both retinal oxygen saturation and retinal vascular fractal dimension is based upon pictures of the retina, taken with specialized cameras and hereafter analyzed with specialized software. Retinal oxygen saturation will be measured (centered at optic disc and macula) by Oxymap T1 with the Oxymap Analyzer software (Oxymap, Reykjavik, Iceland), and retinal vascular fractal dimension will be determined by semiautomatic software (Singapore Institute Vessel Assessment-Fractal, version 4.0, Singapore).

At month 3, retinal oxygen saturation and retinal vascular fractal dimension will be measured again in both eyes of patients with nAMD.

Collaborating researchers and departments

Department of Ophthalmology, Odense University Hospital and Faculty of Health Science, University of Southern Denmark

• Medical student and pregraduate research fellow Ditte Borup Jakobsen

Department of Ophthalmology, Odense University Hospital

• Professor and Head of Research Unit Jakob Grauslund, MD, PhD
• PhD-student and Daily supervisor Thomas Lee Torp, MD
  

Department of Ophthalmology, National Hospital Reykjavik, Iceland

• Professor Einar Stefansson, MD, PhD,

The Ophthalmic Image Analysis Unit, Moorfields Eye Hospital, London

• Professor Tunde Peto, MD, PhD