Q&A: Yasha S. Modi, MD, on retinal nonperfusion in diabetic retinopathy

Photo of Yasha S. Modi, MD, at Retina World Congress 2025

Yasha S. Modi, MD, associate professor of ophthalmology at NYU Langone Health, discusses highlights from his presentation, "Retinal Nonperfusion in Diabetic Retinopathy," which he presented during the 2025 Retina World Congress annual meeting with the Ophthalmology Times team.

Note: This conversation has been lightly edited for clarity.

Ophthalmology Times: You presented at Retina World Congress 2025. What did your presentation focus on?

Yasha S. Modi, MD: At Retinal World Congress, I presented a talk on retinal nonperfusion and its prognostic value. What we were looking at was really to go and say, relative to the DRSS, kind of like the gold standard for diabetic retinopathy staging, is there value in looking at leakage and nonperfusion? And whether you look at the posterior pole or the far periphery, sort of has that which has been described in the DRC, our protocol, Protocol AA. What we see is that by ultra-wide field FA, both nonperfusion and leakage, independent of the diabetic retinopathy scoring system, carries an increased risk of progressing to more advanced forms of diabetic retinopathy. So this becomes clinically relevant when you think you have somebody with what would otherwise be moderately severe or very severe NPDR, and in order to really, truly understand their risk, I believe that we probably need imaging in the form of ultra wide field imaging and perhaps earlier treatment, whether that's in the form of anti-VEGF medications or panretinal photocoagulation before a complication occurs.

OT: When you think about your time in ophthalmology, how have advances in imaging impacted the diagnosis and therefore treatment on a whole, or more specifically in diabetic retinopathy?

Modi: When we think about these prognostic factors, both leakage and nonperfusion, what's incredible is that the ETDRS back in 1991 identified that these were risk factors, but now with more modern imaging, with ultra widefield FA imaging, not only do we understand them to be risk factors, but potentially areas for target treatment. So that may be, if you see a ton of areas of nonperfusion, maybe you're going to selectively target your PRP to that region.

OT: Does anti-VEGF alter nonperfusion?

Modi: So when we think about anti-VEGF and its effect on nonperfusion, I think there's a really interesting literature that we learned by looking back at clinical trials. And from the earliest RISE and RIDE studies, which FDA-approved ranibizumab for the treatment of diabetic macula edema, what we have learned is that when you give anti-VEGF on a monthly basis, nonperfusion remains relatively stable, as opposed to the sham arm, where nonperfusion continues to grow.

Now, if you look at Charlie Wyckoff's study from recovery, he asked the question, is it dose dependent? In other words, if you're injecting the medication on a quarterly basis versus a monthly basis, is there some value in doing it less frequently? And the answer is that nonperfusion continued to progress when you gave aflibercept every quarter, but when you give it every month, it appeared to have some stabilizing effects on nonperfusion.

The caveat is that at no point in time in any study have we been able to definitively demonstrate a improvement to nonperfusion. So whatever level of nonperfusion we start with, that's effectively what we're going to end with in the best case scenario. But it does sort of provide some sort of evidence to say that anti-VEGF may modulate progression of nonperfusion when given very frequently, but doesn't reverse it.

OT: What might you say to a resident or a fellow, specific to diabetic retinopathy?

Modi: When we summarize these points, it's valuable to think, if you're, for instance, a resident or you're a fellow. How should this data modify the way in which you approach patients with diabetic retinopathy. And if you look into the eye of somebody who ultimately fulfills the sort of classically taught 4-2-1 rule of 4 dot plot hemorrhages, 2 quadrants of venous speeding, 1 quadrant of Irma, if you see that that should be an automatic trigger to get an ultra-wide field FA, and then you can make a much more meaningful decision as to exactly how at risk your patient is.