New Protocol V findings offer guidance for managing DME

July 1, 2020

A new analysis from the DRCR Network Protocol V showed that many patients with center-involved diabetic macular edema and good vision do not require treatment for visual acuity loss.

This article was reviewed by Sandeep N. Shah, MD and Adam Glassman, MS.

A post-hoc analysis of data from the DRCR Network Protocol V provides new insight on outcomes of eyes with center-involved diabetic macular edema (CI-DME) and good best-corrected visual acuity (BCVA) if treated with anti-VEGF therapy only if vision worsened. The research also identified features associated with a need for intervention.

The findings were published in JAMA Ophthalmology on Feb. 20, 2020 [Glassman AR, et al; DRCR Retina Network. JAMA Ophthalmology. 2020;138(4):341-9] and presented at ARVO 2020 by Sandeep N. Shah, MD.

“Protocol V randomized adults with type 1 or 2 diabetes who had CI-DME and Snellen equivalent BCVA of 20/25 or better to initial management with aflibercept, focal/grid laser, or observation with aflibercept given in the laser and observation groups if VA decreased from baseline. At 2 years, there was no difference in VA loss between the three study arms, and mean VA was 20/20 in all groups. Based on these results, many clinicians and patients might choose initial observation for eyes with CI-DME and good VA, withholding anti-VEGF treatment unless VA worsens,” said Shah, Principal Investigator for Protocol V at the Retina Vitreous Center, Oklahoma City, OK.

“Overall, the data from Protocol V show that some patients will do quite well if managed with observation, and that using the DRCR Retina Network protocol-defined approach combining BCVA and OCT-measured central subfield thickness (CST) is an effective strategy for determining when to initiate anti-VEGF treatment," Shah added. "Knowing risk factors that predispose to the need for anti-VEGF therapy helps in deciding follow-up intervals and patient education.”

In Protocol V, a total of 236 eyes were randomized to initial management with observation. Per the protocol, patients were to be followed initially after 2 months. Anti-VEGF treatment was initiated if VA worsened by ≥10 letters or more from baseline at any visit or if there was a loss of 5 to 9 letters at two consecutive visits.

The schedule for ongoing follow-up of eyes that did not have worsening of VA was determined based on findings of the OCT. If the CST worsened, defined by an increase of at least 10% from the last visit or an increase to >400 microns, the follow-up interval was reduced to a minimum of 1 month.

However, the maximum follow-up interval was extended to every 4 months for eyes that did not have a worsening of CST.

During the 2 years of follow-up in the observation group, 19% of eyes experienced a BCVA loss of 10 or more letters at least once and 39% experienced a 5- to 9-letter loss at least once. At the next follow-up visit, however, only 32% of eyes had a sustained VA loss and began aflibercept, whereas 46% of these eyes never went on to need aflibercept during the 2 years of the trial.

“These data show how important the role of repeatability of VA change is in an overall treatment regimen,” Shah said.

At the end of the 2-year follow-up, 66% of eyes did not require anti-VEGF treatment with aflibercept and 31% of those eyes had spontaneous resolution of DME with a median VA of 20/20. Within the subgroup of eyes that received anti-VEGF treatment per study criteria, VA was still good at 2 years (median 20/25) and 70% did not lose ≥1 line of VA.

Need for intervention

The analysis of factors associated with receipt of aflibercept showed greater baseline CST, having advanced diabetic retinopathy, and need for treatment of DME in the fellow eye increased the likelihood for intervention. Eyes with baseline CST ≥300 µm were approximately twofold more likely to go on to anti-VEGF treatment than eyes with less macular thickening.

The impact of diabetic retinopathy severity was analyzed with eyes classified using the ETDRS grading system. It found that eyes with a DRSS level ≥47 (moderately severe nonproliferative diabetic retinopathy and above) were 2.22 times more likely than eyes with a DRSS level ≤43 (moderate NPDR and below) to need aflibercept treatment.

A need for DME treatment in the fellow eye within 4 months of randomization increased the likelihood that a patient would receive aflibercept in the study eye by 2.55-fold.

“However, even in these three higher risk groups, only about 50% of the eyes received anti-VEGF treatment. Therefore, although clinicians and patients should understand treatment expectations when they are in the higher risk groups, many of these eyes still will not need treatment,” said Adam R. Glassman, MS, lead author of the JAMA Ophthalmology paper and Director, DRCR.net Coordinating Center, Jaeb Center for Health Research, Tampa, FL.

Limitations

Discussing the limitations of the report, Shah identified the post-hoc nature of the analysis and that it did not evaluate the laser group, which required fewer injections than the aflibercept group.

In addition, some of the associations identified may have been due to chance considering that multiple characteristics were evaluated.

Finally, the treatment decisions were guided by VA measurements obtained with the E-ETDRS testing protocol refraction, which may differ from the approach used in clinical settings.

Sandeep N. Shah, MD
E: shah@rvcoklahoma.com
Adam Glassman, MS
E: aglassman@jaeb.org
Shah and Glassman have no relevant financial interests to disclose.