Peter K. Kaiser, MD: Numerous studies have shown that intravitreal anti-VEGF injections for the treatment of nonproliferative diabetic retinopathy (NPDR) in the absence of diabetic macular edema (DME) is associated with reduced diabetic retinopathy (DR) progression, 2-step or greater reduction in Diabetic Retinopathy Severity Scores (DRSS), and delay in onset of vision threatening complications such as DME. However, practice pattern surveys reveal that many eyes with NPDR are followed in the real world and not treated with anti-VEGF (Figure 1).1 What is your take on why we see this divide between clinical trials and real-world practice?
Figure 1. The dichotomy of treating NPDR in real-world practice: evidence supports intravitreal anti-VEGF injections, but practice pattern surveys indicate most patients are followed with monitoring.1
David S. Boyer, MD: There could be several explanations. First, NPDR patients typically have good vision, so it may be difficult to motivate someone who is not experiencing any symptoms to start injections. Second, patients are anxious about starting injections and especially so when they are not convinced they have active disease. Third, as the PANORAMA and Protocol W studies showed, there was no significant visual improvement noted in the group treated with intravitreal anti-VEGF injections versus the untreated group.2,3 Fourth, the treatment burden adds to whatever other medical interactions these patients have due to their diabetes.
Dr. Kaiser: What is your preferred approach for managing an eye with severe NPDR? Is there an ideal candidate for anti-VEGF injections?
Jeffrey S. Heier, MD: There are simple things we can do for patients with NPDR as retina specialists, including talking to them about the importance of systemic glycemic and blood pressure control. We usually observe severe NPDR unless the patient is motivated to start treatment, and then we can discuss anti-VEGF therapy. The data on anti-VEGF therapy are overwhelmingly positive for these patients, yet few are started on therapy due to the treatment burden.
Dr. Kaiser: Various studies comparing panretinal photocoagulation (PRP) to anti-VEGF injections have shown an improved risk/benefit for the latter. If that is the case, why is PRP still used?
Dr. Heier: The easy answer is patients’ adherence, or lack thereof. We tend to see more adherence issues and higher rates of loss to follow up among patients with diabetes. Combination anti-VEGF therapy and PRP can be used to stabilize the eye in case treatment is interrupted. But we are walking a tight rope: PRP provides long-term control but damages vision in the peripheral field and affects night vision, among other potential complications.
Caroline R. Baumal, MD: Another factor driving the use of PRP laser is the limitation of long-term outcomes related to the current data. It is known that earlier treatment of symptomatic center-involved DME is associated with better visual prognosis. To date, it is not clear whether that is the case in eyes with severe NPDR without DME.
Dr. Kaiser: There is an inflammatory component to DR. Are steroids an option in DR without DME?
Dr. Heier: Steroids are an option in limited cases, but the safety implications are well known, including cataracts and IOP elevation. There appears to be a higher risk for steroid response in patients with diabetes, but the pseudophakic patient who you know is not a steroid responder, the dexamethasone implant (Ozurdex, Allergan, an AbbVie Company) may be a consideration. The presence of macular edema indicates active inflammation, so steroids become much more of an option in that setting.
Dr. Baumal: Intravitreal corticosteroids may also induce regression of retinal neovascularization in PDR, and this may have utility for patients when anti-VEGF use is relatively contraindicated, for example during pregnancy. However, there is limited clinical use for intravitreal steroid in eyes with NPDR without DME.
Dr. Kaiser: To summarize, it sounds like we are left with active monitoring as the preferred option for the majority of NPDR patients because the active treatment options engender so many questions.
Dr. Baumal: While watchful waiting is a common paradigm, there are some key learnings about the natural history of severe NPDR from the PANORAMA study to be considered. Just over 50% of sham-treated (observed) eyes developed vision-threatening complications by 100 weeks, which closely replicates findings from the ETDRS studies that were performed over 3 decades ago.2 It is telling that the rate of progression from severe NPDR to PDR with vision-threatening complications has not improved with better systemic medical therapies for diabetes.
There may be some scenarios when offering treatment for severe NPDR without DME may be considered. For example, a patient who had complications in the fellow eye related to PDR may benefit from anti-VEGF treatment to reduce DR severity. It may also be considered in a patient with clinical signs of rapidly progressing NPDR severity and poor systemic control of diabetes. Anti-VEGF injections may be preferred over panretinal laser treatment for PDR in patients who cannot tolerate or cooperate with the laser procedure, in patients who have lost peripheral vision in one eye from laser, or in patients with a high risk of developing DME.
I sometimes use fluorescein angiography findings to aid in the decision-making process. Signs of extensive vascular wall staining or leakage and/or retinal ischemia may prompt me to offer earlier treatment due to the high risk for progression.
Dr. Heier: Is a finding of nonperfusion a reason to start treatment?
Dr. Baumal: Currently, I do not treat isolated nonperfusion because there is a lack of clinical studies to show a benefit. A new way of categorizing DR based on data from widefield imaging and OCT angiography may be necessary in order to answer that question.
Considering the Patient Perspective
Dr. Kaiser: In patients for whom you are considering anti-VEGF injections, when do you start? How do you gauge the potential for adherence to the treatment regimen?
Dr. Boyer: The level of systemic glycemic control provides some information. An A1C hemoglobin of 9 or greater has me concerned about the patient’s compliance, and I would hesitate to start anti-VEGF injections alone. If the A1C is controlled, and if we determine that anti-VEGF is likely to have a benefit, then we explain that our objective is to prevent severe visual complications rather than to improve vision. We try to limit the number of injections needed to start, and we spread it out to eventually every 4 months.
The issue of adherence and compliance touches on a reason why there is so much optimism for treatments for NPDR that don’t require patients to come to the office. Something that could be prescribed for at-home use would facilitate earlier treatment, and the earlier we can treat, the higher the likelihood maintaining vision without complications.
Dr. Baumal: Monitoring of adherence to treatment may benefit from advances in ophthalmic imaging. Couple that with recent advances in telemedicine, fundus photography, and at-home OCT, and we may have an entirely new way of actively treating DR instead of monitoring disease.
Factors to Consider With Anti-VEGF Injections
Dr. Kaiser: There is accumulating evidence that some patients being treated for DME secondary to DR have a less-than-ideal responses to anti-VEGF injections.4,5 How do you determine that a NPDR patient is having a suboptimal response in terms of DR improvement to anti-VEGF injections?
Dr. Heier: Determining a poor response to anti-VEGF injections is a little more challenging in eyes with NPDR compared to PDR. In eyes with PDR, the treatment effect in responders is evident as early as day 1 to week 1. We don’t have that opportunity with NPDR, because we do not have verified endpoints to use for monitoring the response to treatment. For example, we know that PRP for nonperfusion does not affect progression rates,6 but we don’t have data on whether anti-VEGF injections may show a benefit except in prevention of disease progression. We need more validated endpoints and biomarkers to really change how NPDR is managed.
Dr. Kaiser: Is there any imaging we can use in the clinic to help?
Dr. Baumal: I use red-free and color widefield fundus photographs every 6 months to aid in monitoring of DR severity. I don’t use the extended DRSS system clinically, but if I do get two levels of improvement, from very severe to moderate, I consider that a success. I have access to widefield imaging through my institution, and this modality is becoming more prevalent in real-world practice.
Dr. Kaiser: Is it possible that at least some of the suboptimal response to anti-VEGF injections that we observe is because pathways other than VEGF also drive the disease process?
Dr. Boyer: The answer, in a word, is yes. In the Protocol I study, 40% of patients gained up to 12 letters of vision, and another 30% only gained around 6 letters.7 Additionally, some patients experienced no significant visual improvement. Taken together, the evidence suggests other mechanisms are involved, and that DR is a multifactorial disease. In DR and other retinal disorders, including neovascular age-related macular degeneration (AMD), the family of RGD-binding (arginine, glycine and aspartate) integrins amplify intracellular growth factor signaling, including VEGF, PDGF, βFGF, and TGF-βR, and offer another mechanism of action to treat patients.8
The DR Innovation Pipeline
Dr. Kaiser: The DR pipeline is very active, with a wide gamut of programs looking at novel targets and delivery systems (Figure 2). Conceptually, a lot of the development activity is focusing on this idea of affecting multiple signaling pathways. One proposed approach is with tyrosine kinase inhibitors (TKIs). What have we learned about TKIs and their potential for diabetic eye disease?
Dr. Heier: A previous program studied an oral TKI that was efficacious in hard-to-treat AMD patients, but the systemic toxicity ultimately resulted in halting the program.10 Certainly there is demonstration of benefit, but safety is the key concern at this point.
Dr. Baumal: The concept is interesting, but how the molecule is packaged and delivered will be crucial for demonstrating efficacy and ultimately success. Some of the work being done with polymers builds on the positive experiences we have had to date with intravitreal implants. I am aware of some phase 2 studies in AMD, but the work in DR/DME is in early stages. But if some of the positive early signals in AMD can be replicated in DR/DME, this mechanism may be promising.
Dr. Boyer: TKIs have tremendous potential, but the previous history of failures with other TKIs is something that newer offerings will need to contend with. However, we also have data from other parts of medicine, oncology in particular, that TKI inhibition is a highly rational strategy for interrupting signaling pathways.
Dr. Kaiser: Not all TKIs are the same. Where in the cascade you target and which kinases are being inhibited by a given drug are going to make a very big difference in terms of efficacy. Delivery also matters. A previous attempt with a topically delivered TKI was unsuccessful, but having one fail doesn’t mean the whole pathway is gone. It just means that particular form of TKI, with that specific delivery method, did not work. We’re starting to see positive data from second-generation TKIs that have been formulated into better polymers.
In terms of topical approaches, previous experiences with topical agents were disappointing. The first-generation topical agents were simply TKI’s placed into solution with no thoughts whether the TKI would make it to the posterior segment. We are now seeing topical approaches that have been specifically designed to target the posterior segment. One program is an eyedrop that diffuses to the retina via the transscleral route, delivering an integrin inhibitor that targets vitronectin and other ligands (OTT166; Figure 3). Integrins have attracted significant interest because of their role in regulating cell-cell and cell-extracellular matrix interactions.
What could this mean for DR pathology, and what is the potential in regulating all those pathways?
Dr. Boyer: OTT166 affects multiple integrins, including avβ3, avβ6, and avβ8, which are known drivers of inflammation, angiogenesis, fibrosis, and leakage, which combine to contribute to DR pathophysiology.7 Targeting multiple pathways has the potential to stop the source of this growth factor signaling that amplifies multiple signaling pathways: VEGF, PDGF, βFGF, and TGF-βR.7 As well, there is some suggestion that OTT166 upregulates the Tie-2 pathway, which we know is beneficial in DR.
About OTT166
Dr. Kaiser: The idea of using a drop to treat NPDR is very reasonable given that we are watchful waiting currently. What is the evidence that OTT166 successfully reaches the retina to have an effect?
Dr. Boyer: The proof-of-concept evidence comes from a rabbit model in which therapeutic levels of OTT166 after topical administration were found to be at the highest concentration levels in the sclera and the retina/choroid compared to vitreous, aqueous, and plasma, thus supporting trans-scleral distribution.9 You can also find supporting evidence from early in-human trials in eyes with DME and AMD.11 The phase 1B studies were designed to look at safety, which was demonstrated: there were no drug-related serious adverse events or signs of corneal toxicity, other than one possibly drug-related mild case of a red eye. There was also a strong signal of biological activity, in terms of anatomic improvement and improved vision.
Dr. Heier: I was an investigator in the early-phase study. It’s difficult to extrapolate outcomes to eyes with NPDR, but there are some lessons we can learn. For instance, visual improvement, albeit minimal, was encouraging. To me, the anatomic improvement, demonstrated as resolution of fluid on OCT, was highly significant and strongly suggested that the topical application of OTT166 actually reached the retina, and then it had a positive effect in many patients once it got there (Figure 4).
The second takeaway is that there was a subgroup of patients who demonstrated a delayed response to treatment. At day 56, among responders, we observed greater reduction in mean retinal thickness compared to day 28, when treatment was stopped, which suggests durability of effect. Among previous nonresponders, there was a reduction in mean retinal thickness, which suggests a delayed response in some eyes.
Dr. Boyer: A similar phenomenon has been demonstrated with previous intravitreal integrin development programs. What impressed me with OTT166 was that we were able to see a response at day 28. When we think about treating DME with anti-VEGF, we don’t always see a response that soon. With OTT166, we’re already seeing improvement by day 28 that in some patients continues through at least day 56. As well, the safety associated with the higher dose fully justified further development.
Dr. Heier: The potential to have that kind of impact right away in early-stage DR seems incredibly promising. This drop would be something I could easily see myself considering for a wide range of NPDR patients, and possibly even some early DME eyes, where we can add anti-VEGF injections later if needed. There is potential for OTT166 to have an impact on slowing the progression to vision-threatening complications, because OTT166 has activity beyond the VEGF pathway.
Dr. Baumal: Of note, the phase 1B study enrolled a diverse population: this was an all-comers analysis, and some of these eyes had received numerous previous anti-VEGF injections. Yet, the results showed resolution of retinal thickness in this difficult-to-treat population by days 28 and 56. This is a rather impressive display of biologic plausibility in a challenging cohort.
Dr. Kaiser: The results of this study have shown evidence of treatment effect and a rapid response among responders. Again, we wouldn’t expect to see improvement in DRSS scores by 1 month after treatment with anti-VEGF injections, but that was seen in some eyes in the phase 1B study with the OTT166 eye drop (Figure 5).
