IRIS Registry gives glimpse into DME in real world

An analysis of data collected in the Intelligent Research in Sight (IRIS) Registry suggests that in the real-world setting, newly diagnosed diabetic macular edema (DME) is being vastly undertreated.

Not only are the majority of these patients not receiving active intervention, but those who are started on anti-vascular endothelial growth factor (VEGF) therapy seem to be getting a suboptimal number of injections, said Jeffrey R. Willis, MD, PhD.

When discussing the paper, however, Dr. Willis and members of the session’s panel noted that the findings need to be interpreted carefully. Taking into account the limitations of the registry, their comments brought forth that it is uncertain whether the data truly reflect the real-world situation. Even if they do, the study does not explain the reasons for undertreatment.

“Although DME is a leading cause of visual impairment among adults in the United States, there is limited national level data on how it is being managed,” Dr. Willis said. “In order to address this knowledge gap, our study used data from the IRIS registry to characterize real-world treatment patterns surrounding incident DME.”

Dr. Willis is now assistant medical director, Genentech, South San Francisco. He recently completed a medical retina fellowship at the UC Davis Medical Center, Sacramento, CA.

Drilling into the database

The study was a retrospective cohort study that analyzed treatment patterns during the first year after DME diagnosis for adults who were entered into the IRIS registry between July 2013 and March 2016 with newly diagnosed, treatment-naïve DME. It included 13,410 patients who met all inclusion/exclusion criteria.

The patients had a mean age of about 65 years, were nearly equally divided between men and women, and were predominantly white (63%). About one-half of the cohort had bilateral DME at presentation, and within that subgroup, mean best-corrected visual acuity (BCVA) was around 20/60. Among patients with unilateral DME, mean visual acuity was about 20/40.

The majority of patients (58%) were diagnosed with DME by a retina specialist and about 30% were diagnosed by a non-retina ophthalmologist. An optometrist made the diagnosis in the remaining cases.

During the first 28 days after diagnosis, 75% of patients were observed without receiving any treatment. When treatment was given, it was most often with anti-VEGF therapy (received by 16% of the entire cohort) followed by laser (received by 8.5% of the entire cohort). Bevacizumab (Avastin, Genentech) was the most commonly administered anti-VEGF agent followed by aflibercept (Eylea, Regeneron).

Among patients who received early intervention with anti-VEGF therapy, the mean number of injections received in the first year after diagnosis was 4.2. Almost one-half of the patients received 3 or fewer injections, Dr. Willis reported.

Looking at treatments administered beginning 28 days after initial diagnosis, the study found that 81% of patients who were initially observed continued to receive no treatment, while about 10% were started on anti-VEGF therapy and 7% were treated with laser.

A logistic and multinomial regression analysis was performed to identify factors predicting treatment within the first 28 days.

The results showed that patients were more likely to be started on anti-VEGF therapy than be observed if they were initially diagnosed by a retina specialist rather than by a non-retina provider and if they had bilateral versus unilateral DME.

The odds of receiving laser treatment rather than being observed were significantly higher among patients seen in a non-urban versus urban setting and significantly lower in patients initially diagnosed with worse vision (<20/40 versus ≥20/40).

Applying caution

Dr. Willis spoke to the strengths and limitations of the IRIS registry.

“In this study, we were able to show that the IRIS registry can be utilized to identify a very large cohort of newly diagnosed, treatment-naïve individuals with DME,” he said. “I believe the IRIS registry is an up and coming data source with promising utility for ophthalmic research that will lead to improvements in patient care.”

However, he acknowledged that its dataset is incomplete. Limitations specific to this study include the fact that DME diagnosis is based on ICD coding and OCT data were not available. “We don’t know if these patients truly had DME and if the DME was clinically significant,” Dr. Willis said.

Additionally, information was incomplete on patients’ medical history (e.g., HbA1c levels, duration of diabetes) and visual function, and it cannot be ruled out that some patients categorized as being managed with observation may have been referred for subsequent care to ophthalmologists not participating in the registry.

The point was also made that the reasons underlying the “suboptimal” number of injections are not known.

“In the world I live in, failure to get injections is not because of the providers, but it is a problem with access to care,” said Marco Zarbin, MD, PhD, Alfonse Cinotti, MD/Lions Eye Research Professor and Chair, Institute of Ophthalmology and Visual Science, Rutgers-New Jersey Medical School, Newark, NJ.

“It is a failure of how the health care system is working.”