Matthew R. Starr, MD
Vitreoretinal Surgery Fellow
Wills Eye Hospital, Philadelphia, PA
Dr. Raj Maturi of the Midwest Eye Institute, presented for the first time, the results of the DRCR network’s Protocol W: “A Randomized Trial of Intravitreous Anti-VEGF for Prevention of Vision Threatening Complications of Diabetic Retinopathy” on the first day of the 2021 ARVO Annual Meeting. He presented the findings of the recently published report in JAMA Ophthalmology detailing the outcomes of eyes with NPDR either treated with intravitreal aflibercept or sham injections.
This pivotal trial only included eyes with severe NPDR in at least one eye with a DRSS between 43 and 53, no evidence of neovascularization on FA, VA 20/25 or better, no DR treatment within 12 months, and eyes had to have less than 4 prior injections and never received PRP. Eyes were randomized to either sham or aflibercept with the primary outcome being cumulative probability of developing PDR or CI DME with VA loss ( as defined by an increase of at least 10 microns in CST compared to baseline and VA loss greater than 10 letters at 1 visit or 5-9 letters for 2 consecutive visits). The treatment algorithm was monthly injections for the first 3 months and then Q4 months for the first 2 years. After 2 years, eyes were only injected if the DR was graded as worse than mild based on clinical exam. Sham eyes were only treated if they developed high risk PDR or vitreous hemorrhage.
The study included 399 eyes with 200 in the aflibercept arm and 199 in the sham cohort. There were similar baseline hemoglobin A1c, visual acuity, CST, and DR severity with about 25% of each arm having eyes with severe NPDR at baseline. At 2 years, the incidence of DME with vision loss or PDR was 44% in the sham cohort and 16% in the treatment arm with a Hazard ratio of 0.32. Analyzing only the development of PDR, there was an incidence of 33% in the sham cohort and 14% in the aflibercept group, HR 0.34. Lastly, only evaluating the incidence of DME with vision loss at 2 years, there were 15% of eyes in the sham cohort and 4% in aflibercept cohort with a HR of 0.36.
When stratified by baseline DRSS, Dr. Maturi reports that there is “definitely a relationship here.” With the lowest DRSS levels having much lower rates of developing PDR or DME. Interestingly, despite regular aflibercept injections, 36% of eyes with baseline severe NPDR went on to develop PDR or DME (versus 68% in sham group).
Despite the improvement in DRSS and decreased rates of PDR and DME, there were no differences in VA at 2 years between the two cohorts, nor were there any differences in the number of eyes with 5, 10, or 15 letter improvements at 2 years. Additionally, the CST difference at 2 years was insignificant at 2 years.
Even though the VA benefits were negligible, this report provides an interesting natural history on the progression of NDPR to PDR in eyes that did and did not receive intravitreal anti-VEGF injections. Despite routine treatment, a significant proportion of eyes with severe NPDR will develop PDR. These eyes may require more frequent monitoring than Q4 months.
Dr. Maturi concluded that the proportion of eyes with moderate to severe NPDR was much lower in treatment arm. However, this treatment did no confer any VA benefit. The four year results will be critical in determining long term VA benefits (if any) of routine aflibercept injections.