ARVO 2025: Analysis of AREDS oral micronutrient supplementation on geographic atrophy growth

At ARVO 2025, in Salt Lake City, Utah, Roger Goldberg, MD, MBA, talked about his presentation on the effect of AREDS vitamins in geographic atrophy.

Video Transcript:

Editor's note: The below transcript has been lightly edited for clarity.

Roger Goldberg, MD, MBA:

Hi, I'm Roger Goldberg from Bay Area Retina Associates in Walnut Creek, California. And it's great to be here at ARVO 2025 in Salt Lake City, Utah. Today, I had the pleasure of presenting data on the effect of AREDS vitamins in geographic atrophy.

We know that the original AREDS studies, AREDS1, which kind of ran over the 1990s and was published in early around 2001 and then AREDS2, which made some slight tweaks to the formulation, ran the 2000s, and was published, I think around 2012/2013. Those showed a reduction in the progression to advanced AMD. But if you look closely at the data, that reduction in progression of advanced AMD from stage 3 or 4, kind of intermediate AMD, was largely driven by reducing the conversion to wet macular degeneration or neovascular AMD. And not really on the development or progression of geographic atrophy.
Recently, just earlier this year, there was a paper published a posthoc analysis of AREDS1 and AREDS2 that looked at, again, kind of a sub group analysis of patients who either developed GA over the course of AREDS1 and AREDS2 or who had some incipient GA and how did it progress over time. And in that posthoc analysis, which is a couple 100 patients, a couple 100 eyes, there is some suggestion that although it didn't slow the overall growth rate of geographic atrophy, and didn't have an effect on visual acuity, that perhaps it slowed the progression towards the fovea. And ultimately, when we think about treating our patients with geographic atrophy, that's the direction we really want to slow down, because we know that foveal anatomy is the most important real estate in the retina, and where the patients kind of fine central vision is located. So that was a really impactful and interesting posthoc analysis of the AREDS1 and AREDS2 studies–could just vitamins alone slow growth of GA towards the fovea.

And we said, well, let's go and look at another data set of GA patients. And can we replicate that same result? And of course, a medicine pegcetacoplan, or SYFOVRE, ran about a 1200 patient phase 3 clinical program called the OAKS and DERVY study. And about a third of those patients didn't get injections, they had sham treatment. And of course, all those patients had fellow eyes, some of whom had geographic atrophy at baseline. So here we have a relatively large population of patients who didn't receive any treatment in the OAKS and DERBY study, and we could follow them over two years. And this is with really advanced imaging, unlike AREDS, which was color fundus photography, kind of assessed on an annual basis. This was OCT and fundus autofluorescence, so really detailed reading center confirmed analyzes. And in OAKS and DERBY, about half of patients were on either AREDS1 or AREDS2. And we looked to see, does AREDS have any impact on the progression of geographic atrophy or growth towards the phobia? And the short answer is no. There is no difference between either sham or fellow eyes, so untreated population, whether they were on AREDS1 or AREDS2 or not taking any vitamin supplementation.

So in this other analysis now, also a posthoc analysis, we did not replicate the result that was seen in the AREDS1 and AREDS2 posthoc analysis. We also then looked at, well, what about in the pegcetacoplan, treated eyes? Could AREDS have any impact on those eyes. And it turns out that, again, when we look over a two year period, no difference, whether the patient was on vitamins or not in terms of pegcetacoplan,'s ability to slow geographic atrophy. Those two populations, overlap perfectly. Finally, we looked at exudative AMD, we know that the AREDS vitamins–I think, that is well established, slow or reduce the risk of developing neovascular or exudative AMD. We also know that these GA medicines, including pegcetacoplan, have an increased risk of exudative AMD development, could AREDS therapy be helpful in driving that down further, and unfortunately, we did not see any difference. So whether the patients were taking AREDS1 or AREDS2, there was really no impact on the development of Exudative AMD.

So obviously these are both the Keenan paper from AREDS1 and AREDS2, and this analysis is posthoc in nature. Lots of limitations that come with that, but we were not able to replicate that finding from AREDS1 and AREDS2. I didn't know what this–I thought would be interesting, actually, if we were able to replicate the results in the Keenan paper. Unfortunately, we weren't. I think really, what was the most interesting is really digging in on the AREDS1 and AREDS2 post hoc analysis and how frankly, confusing it is. Because in AREDS2, there was no population that didn't receive treatment. Everybody got some variation of AREDS1 plus or minus lutein and zeaxanthin or plus or minus beta carotene. So there was no real placebo or control group in AREDS2 so you sit there and scratch your head to say, well, how did they show an effect if, in that AREDS2 population, there was no control group. So I think kind of the further and deeper I got into the AREDS data, frankly, the more confusing I found it.