Transforming dry AMD treatment with AI
Age-related macular degeneration (AMD) is the leading cause of blindness over the age of 55 in developed countries. Since the development of anti-VEGF injections (avastin, eylea, vabysmo, etc), neovascular (wet) AMD can be successfully treated and vision loss prevented. However, non-neovascular (dry) AMD now contributes significantly to the vision loss in AMD.
Dry AMD used to be untreatable, but it is now no longer the case. In 2023, the United States Food and Drug Administration (FDA) approved Syfovre and Izervay for the treatment of geographic atrophy (GA), i.e. advanced dry AMD. The Australian Therapeutic Goods Administration followed in 2025 by granting approval for Syfovre. Furthermore, in intermediate AMD (i.e. dry AMD with more than small drusen), photobiomodulation in the form of Valeda Light Delivery System was approved for treatment by the FDA based on the Lightsite III phase III study.
Although all these new developments are exciting, I find my dry AMD patients struggling to find information on:
- how fast is my disease progressing and how likely am I going to lose vision without treatment?
- do I need treatment right now, and how is this going to help my disease?
- many of these treatments are novel, and some are not yet approved for use in New Zealand. Who will have time and expertise to transverse this complex field with me?
Current approach to dry AMD assessment is inadequate
In the public system, for a patient with dry AMD, they will be told that there is nothing one could do apart from AREDS supplements, and be discharged from the clinic with the advice that nothing is going to change quickly. Hardly reassuring!
A few of my colleagues will go a little further to score the patient on the simplified scoring scale from the AREDS study, which will give a rough estimate of the progression rate of intermediate AMD to GA. However, this does not give any information on where GA will occur (centre or non-centre) and if there is a risk of vision loss when GA occurs.
Tracking GA progression is even more difficult. Clinical trials rely on either manual segmentation or proprietary AI algorithm to segment GA from imaging. None of these are routinely available in clinical practice.
So, there is a paucity of information for our dry AMD patients on their disease progression to guide treatment decision with existing approaches.
An ideal dry AMD assessment?
The ideal solution will be an analysis of existing imaging in dry AMD, such as OCT scans, and gives a concrete assessment on the likelihood of progression to vision loss with GA, so those patients at risk of developing GA or of GA progression can be treated with PBM and Syfovre respectively. The treating ophthalmologist will then have the time and expertise to discuss available treatments, including those yet to be approved in New Zealand or those available via clinical trials, to slow the disease and preserve vision.
My approach to dry AMD
By combining AI, time and expertise, Dr Leo Sheck is closing the gap for the unmet need of dry AMD treatment. His approach has three components:
- RetInsight GA monitor - a clinically proven AI algorithm for accurately assessing areas of photoreceptor degeneration (i.e. areas at risk of GA development) and areas of RPE loss (i.e. areas that GA is already present) based on Heidelberg Spectralis OCT imaging.
- 30 minute standard appointments to discuss the implication of the analysis, and to go over in details what treatment is available.
- Expertise on GA and intermediate AMD treatment - Dr Sheck has been using Syfovre since early 2024 and he has also been in close contact with Lumithera, the company that developed the Valeda Light Delivery System.
Advantages of accurate dry AMD assessment
Personalised GA treatment for accurate patient selection to maximise vision without over-treatment
Patients with GA who are at risk of progression should consider treatment with Syfovre or Izervay. These drugs are clinically proven to slow progression, but they carry a risk of neovascular (wet) transformation affecting 5-7% of treated patients. There is also risk of inflammation including retinal vasculitis, along with the other injection risks (bleeding, infection, etc). Patients also do not particularly enjoy injections. So, case selection is of paramount importance in GA therapy.
RetInsight is a breakthrough in solving this issue. By comparing the areas of photoreceptor degeneration and RPE loss, the algorithm has been proven to identify those patients with GA that are at risk of progression and will benefit from Syfovre treatment. The data was published in Ophthalmology recently.
In my practice, I use RetInsight exactly for this purpose. In this patient, I am treating the right eye with Syfovre and you can see that the GA area (blue) is approaching but just sparing the fovea, but there is a much larger photoreceptor degeneration (green) area which is at risk of developing GA.
In the other eye, I did not initiate treatment because the GA area (blue) is already involving the fovea, and the photoreceptor degeneration (green) area is not much larger than the GA area, so the GA is unlikely to progress significantly.
Monitoring GA for treatment adherence
One issue with GA therapy is that, unlike wet AMD where successful treatment will rapidly lead to resolution of fluid and oedema, there is no visible signs on examination or OCT that will indicate therapeutic effect. One solution is to calculate the area involved (in mm^2) and to plot this metric over time, which is unfeasible until the availablity of RetInsight.
RetInsight provides a measure of the area involved automatically, so these can be plotted and a trend derived. By seeing a trend of stabilisation, this will help motivate our patients to continue on their injection therapy.
Identifying patients for PBM
PBM has been marketed as a treatment for all patients with intermediate AMD. However, intermediate AMD has a wide range of severity, and treatment should be reserved for those who are likely to progress to advanced AMD.
The size of photoreceptor degeneration (green area) on RetInsight can be used exactly for this purpose. Patients with no photoreceptor degeneration despite drusen can be monitored without treatment, as they are unlikely to progress to advanced AMD any time soon. However, those with photoreceptor degeneration, especially if these areas are enlarging or fovea involving, should consider treatment with PBM to prevent further worsening.
Frequently asked questions
Is RetInsight reliable? How come no one else is using this algorithm?
The data on RetInsight was published in Ophthalmology, most recently in 2025. The journal Ophthalmology is the leading journal in the field. The analysis was performed on the Derby and Oaks studies, the two phase III studies that led to FDA approval of Syfovre. The data was also presented in the American Academy of Ophthalmology Congress in 2023, where Dr Sheck was in attendance. Hence, he has early access to the algorithm by staying in close contact with the company.
Can my own ophthalmologist do the same for me?
Replicating the exact analysis is impossible without access to RetInsight. Furthermore, to adequately address the complexity of this rapidly evolving field requires time and expertise, investments that Dr Sheck has made to ensure its success.
I do not have medical insurance. What is the financial investment?
You can find my pricing here. A typical appointment will require a consultation fee, and fee for both OCT and Optos imaging. However, Dr Sheck does not charage extra for the advanced RetInsight analysis.
For many patients with dry AMD, the RetInsight analysis may show a reassuring result and that no further treatment is necessary. For those with more concerning findings, Dr Sheck will guide you step-by-step for the available treatments, including clinical trials.
It all sounds great. What should I do next?
Dry AMD is no longer an untreatable disease, and patients deserve the most accurate analysis on their macula status, so they can make an informed decision on their treatment. So, if you have a patient with dry AMD, please refer the patient to me because I have the technology, time and expertise to make a difference to their vision.
About Dr Leo Sheck
Book your appointment now to see Dr Sheck
Dr Sheck is a RANZCO-qualified, internationally trained ophthalmologist and runs a popular and respected practice in Auckland. He combined his initial training in New Zealand with a two-year advanced fellowship in Moorfield Eye Hospital, London. He also holds a Doctorate in Ocular Genetics from the University of Auckland and a Master of Business Administration from the University of Cambridge. He specialises in medical retina diseases (injection therapy), cataract surgery, ocular genetics, uveitis and electrodiagnostics.
