Drilling Down for New Insights

Subanalyses of key studies reveal trends that may aid treatment decisions for DME

BY VIRGINIA PICKLES, CONTRIBUTING EDITOR

Optomap image courtesy of Optos

One of the unknowns in treating diabetic macular edema is whether or not initial therapy with an anti-VEGF agent will produce the expected positive outcome in a particular patient. Absent a reliable early predictor, clinicians begin anti-VEGF therapy knowing some patients will have an inadequate response, signaling the need for a different course of treatment. To date, however, there has been no definitive guidance regarding how many anti-VEGF injections should be administered before a patient is deemed a nonresponder.

“Diabetic macular edema has a natural history that progresses from being driven primarily by permeability to being driven primarily by inflammation,” says Pravin U. Dugel, MD, clinical professor, University of Southern California, and managing partner of Retinal Consultants of Arizona. “When patients come in, we don’t know who is where. So we treat them with anti-VEGF monotherapy, which works very well in the permeability phase, but not so well if the disease has become multifactorial and is in the inflammatory phase.”

This distinction is particularly relevant, Dr. Dugel says, because other FDA-approved treatments are available, namely two intravitreal steroid implants: Ozurdex (dexamethasone, Allergan) and Iluvien (fluocinolone, Alimera Sciences). “With these alternative therapies available, it would be helpful for us to predictably know if a patient will do well with continued anti-VEGF monotherapy or if we should consider switching therapies,” he says.

A post-hoc analysis of 3-year data from the Diabetic Retinopathy Clinical Research Network’s Protocol I study¹ provides some insight for clinicians.

PREDICTING LONG-TERM RESPONSE

Dr. Dugel presented an overview of the EARLY (Early Anti-VEGF Response and Long-term Efficacy) analysis at the Late-Breaking Developments session held during Retina Subspecialty Day at the 2015 meeting of the American Academy of Ophthalmology.² The analysis, supported by Allergan, assessed best-corrected visual acuity (BCVA) response to Lucentis (ranibizumab, Genentech) in 3 cohorts: eyes that gained less than 5 letters; eyes that gained 5 to 9 letters; and eyes that gained 10 or more letters. Unadjusted differences in mean BCVA change from baseline were -0.3, +6.9, and +15.2, respectively, at 12 weeks, and +3.0, +8.2, and +13.8, respectively, at 3 years. After multivariate adjustment, significant correlation remained between BCVA gain at week 12 and at years 1 and 3 (p <0.001).

“For the first time, we are able to provide insight into predicting long-term response to anti-VEGF therapy early in the course of treatment,” Dr. Dugel says. “Essentially, this analysis provides a basis for understanding how patients will respond after 3 injections, allowing us to consider alternate modes of treatment for those who are not doing well.

“Our analysis also shows that the overall treatment benefit is quite low in eyes that do not respond well to anti-VEGF monotherapy initially but continue to receive monthly injections,” Dr. Dugel continues. “Therefore, we know that if we don’t switch to the appropriate therapy in time, not only will patients not benefit, but we may be leaving vision on the table. The point is, we really need to customize treatment for diabetic macular edema according to the individual patient, and the insight gained from this analysis will help us better understand which treatment may benefit each patient.”

What are clinicians taking away from the EARLY analysis findings? We asked Antonio Capone, Jr, MD, and Thomas A. Albini, MD, to weigh in.

BUILDING ON PRIOR STUDIES

“Findings from the EARLY analysis continue to build the case that patients with diabetic macular edema need to be treated with the most effective therapy for them individually as early as possible in the course of their disease,” says Dr. Capone, a partner at Associated Retinal Consultants, Royal Oak, Mich., and a professor at William Beaumont Hospital-Oakland University School of Medicine, Auburn Hills, Mich. “We have data from as far back as the pivotal studies for Lucentis³ demonstrating that patients who were in the sham arm and had treatment delayed for a year — in other words, those who weren’t treated as effectively as possible as early as possible in retrospect — never gained the same level of vision as patients who were treated promptly. From that body of data, we know that the sooner we treat with an appropriate therapy the better, not just for prompt return of vision but also for the ultimate level of vision obtained.

“We also know from studies of other conditions in which macular edema impairs vision, such as branch and central retinal vein occlusion,⁴ that patients tend to tip their hand fairly early with regard to how they will respond to anti-VEGF therapy, and then they generally tend to ‘swim in their lane,’ so to speak,” Dr. Capone says. “In other words, if a patient is a good responder early, he tends to be a good responder throughout, i.e., at 12, 24, and 36 months; this corollary is also true for poor responders. If we can identify nonresponders or incomplete or inadequate responders earlier, as the EARLY analysis suggests, we can consider an alternative or adjunctive therapy earlier and improve outcomes. From a logic perspective, an anti-inflammatory or steroidal is the next most appropriate choice.”

CLINICAL IMPLICATIONS

Dr. Albini describes his current approach to treating diabetic macular edema as conservative. “I start the vast majority of patients with anti-VEGF therapy, giving injections every 4 weeks for the first 6 months,” he says. “If I don’t see an improvement after month 6, I consider switching to a sustained-release steroid. I am most familiar with the Ozurdex implant, which has been shown to be efficacious with manageable side effects, particularly for patients who are pseudophakic.”⁵

Although he was eager to see the visual acuity data from the EARLY analysis, Dr. Albini says he is also interested in the results of a similar analysis of Protocol I OCT outcomes, which is currently under way. “I look forward to seeing all the data in written form to better analyze and interpret it for my patients,” he says, “but from the visual acuity data discussed at the Academy, it appears the decision to switch to a steroid can be made even earlier than I had thought, i.e., at the 3-month mark.”

Dr. Capone notes, “If a patient is a prompt or intermediate responder after a first injection, I continue to treat with an anti-VEGF agent. However, I will switch a true nonresponder to a steroidal agent after the first injection if the patient is an appropriate candidate, that is, has no glaucoma or lens-status exclusions.”

THE RETINOPATHY FACTOR

The FDA’s approval last year of both Lucentis and Eylea (aflibercept, Regeneron) to treat diabetic retinopathy in patients with diabetic macular edema adds another component to any decision to switch from an anti-VEGF agent to a steroid, Dr. Albini says.

“Clinical data⁶ show a reversal in the severity of retinopathy in patients who receive serial anti-VEGF injections, and that anti-VEGF therapy is potentially as effective as panretinal photocoagulation (PRP) without the ill effects of PRP, although also lacking its permanence,” Dr. Albini says. “One of my concerns is that by switching patients to a steroid, I am not giving them the opportunity to have the reversal in their retinopathy.”

Another late-breaking development presented at the Academy meeting has helped allay Dr. Albini’s concerns. Charles C. Wykoff, MD, PhD, presented results from a new analysis of patients from the FAME study.⁷ He reported that at month 36, statistically significantly fewer eyes treated with the Iluvien (17%) experienced progression of proliferative diabetic retinopathy (PDR) versus sham-treated eyes (31%).⁸ This finding was based on fundus photographs that were graded by a masked, certified reading center, and by the incidence of the use of PRP and vitrectomy to treat PDR.

“I was impressed that there’s at least a potential from steroids to also lead to reversal in severity of retinopathy,” Dr. Albini says. “So between Dr. Dugel’s presentation and Dr. Wykoff’s presentation, I’ve become even more confident about using steroids in these diabetic patients.”

INFORMING TREATMENT DECISIONS

Dr. Albini and Dr. Capone look forward to new data from studies and subanalyses, such as those discussed here, to help inform their treatment decisions. Dr. Albini notes, “What we don’t know, that we need to know, is the relative efficacy of a steroid versus an anti-VEGF agent. There’s certainly data from Protocol I that visual acuity outcomes in pseudophakic eyes were roughly equivalent for both therapeutic options. Furthermore, the BEVORDEX study⁹ showed that PRN bevacizumab (Avastin, Genentech) produced results similar to PRN dexamethasone with significantly fewer injections in the dexamethasone eyes. Visual acuity outcomes were virtually the same in both groups and, surprisingly, OCT outcomes were better in the steroid-treated eyes than in the anti-VEGF eyes. It is comforting for me to see that the efficacy of steroids and anti-VEGF agents seems to be similar at least in these studies.”

Dr. Capone notes, “Protocol T provided important evidence to guide how we choose initial anti-VEGF therapy.¹⁰ We know from clinical experience that a proportion of patients are anti-VEGF nonresponders who may be effectively managed with a steroid as monotherapy. I am looking forward to evidence that confirms this experience.” NRP

REFERENCES