The Structural and Scientific Gaps Still Undermining DME Care: Rahul Khurana, MD

Sustained drug delivery, neuroprotection, and better coordination between retina specialists and primary care physicians represent the most urgent unmet needs in diabetic macular edema (DME) management, and the next generation of therapies must target all 3, according to Rahul Khurana, MD, partner and president/CEO of Northern California Retina Vitreous Associates and volunteer clinical associate professor in ophthalmology at the University of California, San Francisco.

Real-world undertreatment during the critical loading phase continues to undermine outcomes that would otherwise be achievable with current agents, and clinicians are too often mislabeling slow responders as nonresponders, which triggers premature switches before patients have had an adequate treatment trial, he explained. He also discussed best-practice treat-and-extend implementation, the structural barriers fragmenting DME care across practice settings, and the pipeline advances he believes could reshape the treatment paradigm.

In the first part of the interview with The American Journal of Managed Care^® (AJMC^®), Khurana discussed the clinical blind spots perpetuating vision loss, the socioeconomic forces shaping drug selection, and the biological complexity driving research toward combination and next-generation therapies.

This interview has been lightly edited for clarity.

AJMC: Durability is one of the most persistent challenges in DME management. How do you think about the tradeoff between an agent's efficacy ceiling and its maintenance burden when initiating or switching therapy?

Khurana: The efficacy-durability tradeoff in DME is best understood not as a simple ceiling-versus-burden calculation but as an optimization across peak efficacy, durability of effect, and the probability that a patient will sustain a treatment long enough to realize a drug's potential. We know from across all the different treatments that if you maintain an intensive treatment regimen, all agents are very effective. The problem, as I mentioned before, is patient adherence. When you have to give an injection every month, it's hard for patients. Many of our diabetic patients are working and find it difficult to take time off to maintain such a long treatment regimen.

That's where durability becomes an advantage—being able to achieve similar effects with 3 treatments versus 6 treatments. That's where some of the newer agents, such as high-dose aflibercept and faricimab, are attractive, because they've been shown to allow treatment intervals a little longer than with traditional agents.^1,2 Now, an important caveat is that even these newer drugs have not been definitively proven to extend intervals, because in those trials they were often compared to 2-mg aflibercept with different dosing strategies, leaving no opportunity for direct comparison. But having the option to use different agents that may minimize treatment burden goes a long way in improving patient adherence and ultimately patient outcomes

AJMC: The treat-and-extend model has become common practice for anti–vascular endothelial growth factor (anti-VEGF) injections, but it's not universally applied. What does best-practice implementation of that approach look like, and where do clinicians most often shortcut it in ways that cost patients?

Khurana: Best-practice treat-and-extend implementation has really stemmed more from our management of neovascular age-related macular degeneration, and I think it's been applied in DME in cases where I'm not sure it's necessarily always the most appropriate approach. The reason is that we know within diabetic macular edema that once you complete that initial intensive injection phase, you often don't need to keep treating as intensively. We have very good data from the DRCR Retina Network showing that in the long term, patients don't need as many injections.³

That said, I do think many clinicians use treat-and-extend in DME practice, and when they do, the 3 non-negotiable pillars are, first, a complete loading phase; second, optical coherence tomography [OCT]–driven interval adjustment using defined criteria; and third, very conservative extension increments.

One of the problems in practice is that the loading phase is often shortened. We know from the original studies that most patients received 5 to 6 loading doses, yet in real-world practice, patients are barely receiving 4 treatments in an entire year. We're missing out on the loading phase that is really key to reversing the damage from diabetic macular edema. I tell most of my patients who expect to be treated with one injection: this disease has developed over years; we can't simply reverse it with 1 treatment, and often the first year will require a lot of injections. Most patients average 7 to 8. But once we make that investment, we can reap the benefits in the years ahead.

The failure to complete the loading phase is why we're not achieving optimal visual acuity outcomes in real-world practice. It leads to more recurrences, greater potential for irreversible vision loss, and ultimately suboptimal outcomes.

AJMC: If you had to identify the single biggest gap between what current approved therapies can achieve and what patients with DME actually need, what would it be?

Khurana: There are 2 gaps. The first is treatment burden. Studies have shown unequivocally that the more treatments you give, the better the outcomes, yet in the real world, that's simply not happening. Even as we've developed agents that potentially last longer—moving from monthly to every 2 months to every 3 months and adopting treat-and-extend—we're still not seeing those results. The first major gap is the absence of more sustained drug delivery that can help with the loading phase. There's exciting investigational work in that area, both through drug delivery devices and gene therapy.

The second large gap is our inability to prevent progressive neurodegenerative retinal damage that continues despite successful control of vascular leakage. Even in patients who are being treated properly and are coming in monthly or very regularly, we have seen that over time some continue to lose vision. There is likely a neurodegenerative component to that loss, and we are still in the very early stages of understanding it.

AJMC: There is a subset of patients who are underresponders or nonresponders to anti-VEGF. How do you identify those patients early, and what does management look like for them today?

Khurana: Identifying anti-VEGF underresponders and nonresponders in DME requires distinguishing between true biological resistance and slow response—a distinction that fundamentally changes management.

One of the most consequential clinical errors is premature switching or escalation before the loading dose is complete. In AMD and some other diseases, we’ve been spoiled, and we've come to expect that 3 injections constitute a loading phase. The reality in diabetic macular edema is that many more injections are required. We sometimes label a patient a nonresponder after only 3 injections, when in reality they may be a slow responder. There is strong data from the DRCR showing that patients with persistent fluid whom we might consider treatment failures have actually done well—gaining approximately 10 letters—demonstrating that persistent fluid alone does not indicate failure.⁴

I think another challenge is the lack of consistent terminology for what constitutes a nonresponder. In practice, many of us think, if the fluid is not completely resolved, the patient is a nonresponder. But when you look at visual acuity—where the fluid was before versus where the patient is now—they have had a response and often a very good visual outcome.

That said, there are patients who truly do not respond to anti-VEGF therapy. On OCT, certain structural biomarkers—such as disorganization of retinal inner layers or loss of the ellipsoid zone—indicate that even with a good anatomical response, visual acuity will not recover. This reflects a very different disease process, likely driven more by ischemia or non-VEGF mechanisms.

AJMC: What investigational approaches in the DME pipeline are you most watching, and what unmet need do you most want the next wave of therapies to address?

Khurana: In the DME pipeline, there are 3 areas I'm most excited about, because they address 3 fundamental limitations of current therapies: first, treatment burden; second, VEGF-independent disease mechanisms; and third, the absence of neuroprotection. Those are the 3 areas of unmet needs and having product address them will be really helpful.

When we look at treatment burden, technologies or medications or devices with a sustained drug delivery can address that issue. Gene therapy is an exciting area with the possibility of a one-time therapy that avoids undertreatment of anti-VEGF therapy.

On VEGF-independent mechanisms: there are at least 8 other pathways involved in diabetic macular edema beyond VEGF.⁵ We have experience with inflammation, and we know steroids work but carry significant side effects, including cataracts and elevated intraocular pressure. Now, companies are targeting specific mechanisms. There is considerable excitement around IL-6 in combination with anti-VEGF.⁶ We've seen how this approach can be effective in uveitic macular edema, and there is promising work exploring it in DME as well.

The third area is neuroprotection. As I discussed, even in patients whose vascular leakage is well-controlled, some continue to experience vision loss over time. For this group, neuroprotection could be a meaningful advance.

In the pipeline, we are seeing agents that address treatment burden, VEGF-independent disease mechanisms, or the absence of neuroprotection that could really meet an unmet need and hopefully move our treatment paradigm forward.

AJMC: From a practice management standpoint, what are the structural or operational constraints that most undermine optimal DME care, even when the clinical intent is there?

Khurana: This is a really important point, because in practice we all encounter structural and organizational constraints that undermine optimal DME care. I think they cluster in 5 interconnected domains: first, insurance-driven treatment distortions; second, patient attrition from treatment burden; third, screening and referral failures; fourth, workforce capacity mismatches; and fifth, fragmented coordination between retina specialists and primary care. Each of these plays a role in not giving ideal care.

One of our biggest issues is patient-physician choices constrained by step-therapy requirements. Instead of giving optimal care that the physician knows is appropriate for that patient, they are overridden by insurance companies and prior authorization processes that don’t understand the nuances of individual patient needs. Unfortunately, that has a huge effect on which medicines we can give our patients.

On patient attrition, approximately 30% of patients with DME are lost to follow up⁷ due to issues of being unable to maintain the treatment burden or afford it. There are also issues with screening and referral failures when patients are referred too late. We know early detection leads to better outcomes. We also know that workforce capacity is an issue because many of our diabetic patients are working and they can't come in. That is a real burden to receiving the optimal care.

Finally, there is a lot of fragmented coordination between retina specialists and primary care. We know systemic control of the diabetes is essential, not only for optimal eye care but also for overall health. Better coordination with primary care around controlling the patient’s blood pressure and cholesterol and optimizing their correct diabetic regimen has the potential to improve outcomes. That coordination, if improved, could make a meaningful difference.

References

1. Brown DM, Boyer DS, Vo DV, et al; and the PHOTON Investigators. Intravitreal aflibercept 8 mg in diabetic macular oedema (PHOTON): 48-week results from a randomised, double-masked, non-inferiority, phase 2/3 trial. Lancet. 2024;403(10432):1153-1163. doi:10.1016/S0140-6736(23)02577-1

2. Wong TY, Haskova Z, Asik K, et al; and the YOSEMITE and RHINE Investigators. Faricimab treat-and-extend for diabetic macular edema: two-year results from the randomized phase 3 YOSEMITE and RHINE trials. Ophthalmology. 2024;131(6):708-723. doi:10.1016/j.ophtha.2023.12.026

3. Diabetic Retinopathy Clinical Research Network; Elman MJ, Aiello LP, Beck RW, et al. Randomized trial evaluating ranibizumab plus prompt or deferred laser or triamcinolone plus prompt laser for diabetic macular edema. Ophthalmology. 2010;117(6):1064-1077. doi:10.1016/j.ophtha.2010.02.031

4. Bressler SB, Ayala AR, Bressler NM, et al; and the Diabetic Retinopathy Clinical Research Network. Persistent macular thickening after ranibizumab treatment for diabetic macular edema with vision impairment. JAMA Ophthalmol. 2016;134(3):278-285. doi:10.1001/jamaophthalmol.2015.5346

5. Zhang J, Zhang J, Zhang C, et al. Diabetic macular edema: current understanding, molecular mechanisms and therapeutic implications. Cells. 2022;11(21):3362. doi:10.3390/cells11213362

6. Stevenson S. ARVO 2026: Vamikibart trials advance IL-6 inhibition in retinal disease. Ophthalmology Times®. May 6, 2026. Accessed June 8, 2026. https://www.ophthalmologytimes.com/view/arvo-2026-vamikibart-trials-advance-il-6-inhibition-in-retinal-disease

7. Kim JS, Lee S, Kim JY, Seo EJ, Chae JB, Kim DY. Visual/anatomical outcome of diabetic macular edema patients lost to follow-up for more than 1 year. Sci Rep. 2021;11:18353. doi:10.1038/s41598-021-97644-2