Featured RETINA Paper: Anti-VEGF Halts Progressive Retinal Nonperfusion in Diabetic Retinopathy. Conversation with Mrinali Gupta, MD.

Tomasz Stryjewski, MD
Vitreoretinal Fellow
Mass Eye and Ear

In this week’s RETINA Roundup, we spoke with Dr. Mrinali Gupta regarding her recent photoessay in RETINA entitled “Reversal of Retinal Vascular Leakage and Arrest of Progressive Retinal Nonperfusion With Monthly Anti–Vascular Endothelial Growth Factor Therapy for Proliferative Diabetic Retinopathy.”


Mrinali Gupta
Dr. Gupta

In this photoessay, Dr. Gupta and colleagues from Weill Cornell Medical College in New York demonstrate that despite adequate panretinal photocoagulation, a 59-year-old diabetic man continued to demonstrate persistent neovascularization, vascular leakage, and progressive retinal nonperfusion.

The patient underwent a course of monthly intravitreal injections of ranibizumab and by 12 months, had an impressive resolution of neovascularization and vascular leakage.

We recently caught up with Dr. Gupta to discuss her elegant paper and learn additional clinical pearls.

Stryjewski: In your photoessay, you show beautiful regression of disease after one year of monthly anti-VEGF therapy. How did you evaluate whether to inject or not each month and a related question, how would you advise a colleague who is thinking of starting an injection series for retinal nonperfusion? Is there an imaging biomarker that you follow (OCTA, FA, macular thickness, etc.) to guide your injection decision making?

Gupta: The patient in the photoessay underwent monthly intravitreal anti-VEGF injections for PDR.  For patients with PDR, based on the presence/absence of DME, the extent of neovascularization, and compliance considerations, I do combination anti-VEGF+PRP laser, PRP, or anti-VEGF alone on a case-by-case basis.  This patient had a demonstrated history of compliance for years, and he strongly preferred anti-VEGF therapy over PRP for his active PDR. For PDR, I use a similar (but not identical) algorithm as described in the DRCR.net protocol S study — with monthly intravitreal injections for 6 months, then a re-evaluation (with FA unless the NV is clinically evident without FA) and a repeat series of injections for persistent NV. Unlike the study, I treat until the NV is resolved.

Gupta MP, Kiss S, Chan RVP. Reversal of Retinal Vascular Leakage and Arrest of Progressive Retinal Nonperfusion With Monthly Anti–Vascular Endothelial Growth Factor Therapy for Proliferative Diabetic Retinopathy. RETINA: September 2018 – Volume 38 – Issue 9 – p e74–e75

In the reported case, the treatment indication was PDR, not retinal non-perfusion. However, we noticed that anti-VEGF therapy resulted not only in PDR regression, but also in an improvement in vascular leakage and a stemming of what previously had been unrelenting progressive retinal nonperfusion (in the areas previously showing marked retinal vascular leakage). The findings were compelling in his case, especially when we looked at the contralateral eye, which over the same period had no intravitreal anti-VEGF injections and had persistent leakage along the major retinal vasculature and interval progression of retinal non-perfusion.

Fortunately, ranibizumab was FDA approved at this time for diabetic retinopathy (not just PDR), and indeed there were non-PDR eyes included in the ranibizumab group in the Protocol S study. On the other hand, there are no randomized, controlled trials demonstrating efficacy of anti-VEGF injections for retinal nonperfusion in diabetic retinopathy nor any trials to guide the treatment protocol. The patient and I had a long conversation about all of this. Based on his history and concern that further progression of retinal nonperfusion could be catastrophic for his vision, the patient elected for anti-VEGF therapy specifically to try to target retinal vascular leakage noted on FA. In the absence of data to drive our treatment protocol, we have decided to use angiographic retinal vascular leakage as a biomarker with an FA roughly every 6 months. We treat monthly initially and after the vascular leakage resolves, try to extend the interval between injections, assessing with FA after 6 months to determine if the duration can needs to be shortened or further extended.

As we learn more about OCT-A, it may also be useful in cases like this, but we have not used it to guide management here as we want to catch the disease activity (leakage) before the vascular dropout occurs.

As a caveat, we of course need evidenced-based data evaluating the use of anti-VEGF injections for retinal nonperfusion (and whether they impact visual acuity outcomes long-term), the imaging metrics for monitoring these patients, and the optimal treatment protocol. The same applies for anti-VEGF injections for diabetic retinopathy in general. Numerous trials have shown reduction in diabetic retinopathy scores with anti-VEGF therapy, but there are important questions such as at what point do we initiate therapy (if using for retinopathy milder than PDR), how do we measure response, how often do we treat and what are the parameters that guide initiation/cessation of treatment?

Do you think the effect will be sustained beyond one year, or will re-treatment be necessary?

In the DRCR.net Protocol S study, only 35% of anti-VEGF treated eyes had resolution of PDR at 2 years, while only 48% had a two-step improvement in diabetic retinopathy scores, which speaks somewhat to the activity of the disease and the durability of anti-VEGF therapy. Now, neither of these metrics reflect retinal vascular leakage as in this patient. It will be interesting in the longer-term studies to see how many of the patients that achieved resolution of PDR will have recurrent PDR after cessation of anti-VEGF therapy. My sense is that the anti-VEGF injections control but do not cure the underlying pathology. However, we also know that in many patients, the diabetic retinopathy often has a variable period of disease activity and then subsequently quiets down. We hope that we can titrate this patient’s therapy to the minimal frequency of injections needed during this period of activity until he can get to that point.

Did the patient demonstrate any improved visual function with the injections?

The patient had excellent visual acuity (20/20-20/25 OU) at baseline, which he has maintained thus far. He has constricted visual fields due to ischemia (and PRP in the past), which also remains stable.

Thank you very much for the great paper and your insights!