In this episode of Evidence-Based Retina, Rishi P. Singh, MD, FASRS, sits down with Baruch D. Kuppermann, MD, PhD, who discusses the ERLYDEX study on switching diabetic macular edema patients from anti-VEGF therapy to a dexamethasone implant after a suboptimal response, and how he makes that call in his own practice.
Dr. Kuppermann is a professor and chair of the Gavin Herbert Eye Institute at the University of California, Irvine.
ERLYDEX fast facts:
- ERLYDEX is a retrospective, post-hoc analysis of 167 diabetic macular edema (DME) patients with a suboptimal anti-VEGF response across seven practices, comparing continued anti-VEGF injections against a switch to dexamethasone implant (OZURDEX, AbbVie)1,2
- At weeks 4 and 12, the OZURDEX group had greater visual gains than the group that was continued on anti-VEGF:1,2
- 45% of patients vs. under 10% gained 10 letters
- 20 to 25% of patients vs. under 5% gained 15 letters
- Intraocular pressure (IOP) elevation occurred in fewer than 20% of dexamethasone-implant eyes within the 12-week study, usually manageable with one or two drops; cataract risk could not be assessed, given the short follow-up1,2
Suboptimal responders to anti-VEGF treatment
Anti-VEGF response in DME varies widely, and Dr. Kuppermann breaks patients into three broad groups based on how well they respond:
- One-third are robust anti-VEGF responders whose retinas dry completely.
- The middle third has a mediocre response, with residual fluid but reasonably good vision.
- A final third has suboptimal responses or non-response, with persistent macular thickening after multiple injections.
He notes that many in this group also have elevated inflammatory chemokines and cytokines alongside VEGF, which could make a switch to another class of drugs, such as steroids, beneficial.3
“It's that group of suboptimal responders that I think are the ones that are most likely to derive a benefit from some other class of drugs and in this case steroids is one.”
Dr. Singh adds that step therapy requirements can prolong the treatment process by requiring sequential trials of bevacizumab and other agents. Steroids, by contrast, have been available for years, and data from the ERLYDEX study could better guide decisions about when to switch some patients to a dexamethasone implant.
ERLYDEX data
Within the seven practices and 167 patients in ERLYDEX, some continued anti-VEGF injections after that initial suboptimal response, while others switched to a dexamethasone implant (OZURDEX). Because the study was retrospective, investigators used best recorded visual acuity rather than ETDRS testing.
Outcomes were compared at weeks 4 and 12 after the switch:1,2
- In the DEX implant group, nearly 45% of patients gained 10 letters of vision and 20 to 25% of patients gained 15 letters or more
- In the group that continued on anti-VEGF, under 10% gained 10 letters of vision and under 5% of patients gained 15 letters or more
OZURDEX safety profile
Regarding side effects in the DEX implant group, Dr. Kuppermann says the 12-week study could not assess cataract risk, since cataract formation typically takes a year or more to develop. IOP elevation, however, was visible within that window and occurred in fewer than 20% of eyes.1,2 He and Dr. Singh agree that this is typically manageable, usually controlled with one or two drops, and rarely requires surgical intervention.
He notes that in this study, the rate is even lower than 20%, in part because most patients have received only one OZURDEX implant.1,2 He adds that it's rare for IOP elevation to appear after the third, fourth, or fifth implant, and that patients who tolerate the early implants tend to be free of that risk going forward. Both the drying effect and the pressure effect peak around week 8, and one way to manage elevated pressure is simply not reinjecting right away, letting the steroid effect wear off.
“I would even be so bold to argue that if we had a steroid that didn't cause either IOP elevation or cataracts, it might be the first line therapy for DME because the impact when you do eventually switch…is so impressive anatomically that you say…why didn't I do it before?”
How Dr. Kuppermann approaches this in his own practice
Asked how this translates to his own clinic, Dr. Kuppermann says his main biomarker is excess macular thickness on OCT rather than vision, since vision is something he can't directly control, while leakage is.
To track that biomarker, he gets an OCT at every visit and gives three loading doses of anti-VEGF therapy, at baseline, 4 weeks, and 8 weeks, watching for roughly a 20% drop in thickness with each injection. If that trend holds, even without full resolution, he's satisfied and continues the course. But if the first two injections reduce thickness only by about 10% and the third produces no additional change, he begins considering whether the patient needs a different drug class, such as a DEX implant.
That decision, he says, rarely takes more than 6 months, and can come sooner, within 3 months, if other markers point toward a chronic process, such as hyperreflective foci or soft exudates. Still, he starts every patient on anti-VEGF and won't switch to steroids without compelling evidence of a poor response.
Dr. Singh reinforced that the data in this study would help inform retina specialists' treatment strategies, since current anti-VEGF agents target VEGF but leave cytokines largely unaddressed, and steroids remain a valuable treatment option.
Key takeaways
- DME patients fall into three broad anti-VEGF response groups, per Dr. Kuppermann.
- Robust responders
- Mediocre responders with residual fluid but reasonably good vision
- Suboptimal responders and non-responders
- In the retrospective ERLYDEX analysis, switching suboptimal anti-VEGF responders to OZURDEX produced substantially greater visual gains at weeks 4 and 12 than continuing anti-VEGF.1,2
- IOP elevation occurred in fewer than 20% of Ozurdex eyes within 12 weeks in this cohort and was generally manageable; cataract risk could not be assessed due to the short follow-up period.
- Dr. Kuppermann starts patients on anti-VEGF therapy and tracks excess macular thickness on OCT at baseline, 4 weeks, and 8 weeks, aiming for roughly a 20% reduction per injection. He considers switching drug classes if the first two injections reduce thickness only by about 10% and the third produces no further change, a decision that typically takes 3 to 6 months.
This article was written by Sonia Kelley, OD, MS, based on the recorded video from Drs. Singh and Kuppermann.
