Long-standing accepted thresholds for initiation of treatment for diabetic (DR) include center-involving diabetic macular edema (CI-DME) with visual acuity (VA) impairment and proliferative DR (PDR).^1,2 DR, however, is a phenotypically variable disease, and other thresholds for initiation of interventional treatments are less clearly defined. Two large, prospective, randomized trials have recently reported top-line data that directly address two of these outstanding thresholds.

Here, I discuss these trials and how the data may be able to be applied to clinical practice.

PANORAMA

PANORAMA is a Phase 3, double-masked, randomized study³ of the efficacy and safety of intravitreal aflibercept in patients with moderately severe to severe non-proliferative diabetic retinopathy (NPDR) or diabetic retinopathy severity scale (DRSS) levels 47 and 53.⁴ Historically, these high-risk NPDR eyes are recognized to be at risk of progression to PDR over one year, or 26% to 52% based on the landmark Early Treatment Diabetic Retinopathy Severity (ETDRS) trial of the 1980s and 1990s.⁵

PANORAMA randomized 402 patients equally to three arms: sham, aflibercept every 16 weeks following four loading doses (three monthly doses and one dose at an eight-week interval, Q16) or aflibercept every eight weeks following five monthly loading doses (Q8). There were two prespecified primary endpoints: weeks 24 and 52. The week-52 endpoint was the proportion of patients improving two or more steps on the DRSS in the Q16 and Q8 arms compared to sham.

At baseline, enrolled eyes were well sighted, with mean vision of approximately 20/25. At the week-52 mark, the proportion of patients achieving a two- or more step improvement in DRSS was approximately 65% and 80% in the Q16 and Q8 arms, respectively, compared to 15% with sham — highly statistically significant differences. There were minimal changes in VA through 52 weeks, as these eyes entered the trial with good central VA and no CI-DME. From a clinically relevant safety perspective, the proportion of patients developing prespecified, potentially vision-threatening clinical events appeared meaningfully different between the groups. Approximately 41% of sham patients compared to approximately 10% of aflibercept patients developed PDR or CI-DME.

DRCR NETWORK PROTOCOL V

DME can be clinically sub-divided into three relevant categories: CI-DME with VA impairment, CI-DME with good VA and non-CI-DME.¹ While a tremendous amount of prospective data supports pharmacologic intravitreal treatment of CI-DME with VA impairment with either VEGF inhibitors^6,7 or corticosteroids,^{8, 9} limited prospective data has been available to guide treatment of CI-DME with good VA in the anti-VEGF era.

Recently, the Diabetic Retinopathy Clinical Research (DRCR) Network reported findings from Protocol V, a multicenter, randomized clinical trial addressing the timing of treatment initiation among eyes with CI-DME and good VA.¹⁰ Study eyes (n=702) were randomized into three arms: observation (n=236), macular laser photocoagulation (n=240) or aflibercept (n=226). The primary outcome was ≥ 5-ETDRS letter VA decrease from baseline at two years.

Protocol V was conducted at 91 sites across North America. Subjects were 18 years of age or older with type 1 or 2 diabetes. Study eyes had CI-DME with BCVA of 20/25 or better (79 or more ETDRS letters). CI-DME was defined as central subfield thickening (CST) ≥305 µm in women and ≥320 µm in men by optical coherence tomography (OCT) using a Heidelberg Spectralis. Prior laser photocoagulation or intravitreal injection therapy for DME within the past 12 months was exclusionary. Investigators and participants were not masked.

Eyes randomized to aflibercept received one injection at baseline and continued to receive monthly injections while VA or CST was improving or worsening (defined as ≥5-letter VA or ≥10% CST change) from either of the last two visits. Injections were deferred if the eye did not improve or worsen over two visits and either: CST was below the definition of CI-DME and VA was 20/20 or better; or at least 24 weeks had passed since injections were initiated. If deferral of injection occurred at three consecutive visits after 24 weeks, the follow-up interval was extended to eight weeks and then to 16 weeks if deferral criteria continued to be met. Injections were resumed if VA or CST worsened. Eyes in the laser arm received macular laser at baseline with re-treatment at 13-week intervals if indicated.

At baseline, mean age was 59 years, CST was 311 µm and mean VA was 85.2 ETDRS letters (approximately 20/20 Snellen equivalent). Two-year completion rate excluding deaths was notably high at 92% (625/681), and the median number of visits through two years was 18, 11 and 12 in the aflibercept, laser and observation groups, respectively. Median number of aflibercept injections through two years in the aflibercept group was eight. In the laser arm, 32% (n=77) received additional macular laser during following up.

PROTOCOL V FINDINGS

Before the trial began, it was hypothesized that one possible finding from this study was that patients in the observation and laser arms would be more likely to experience VA loss than patients randomized to aflibercept. Therefore, the trial structure allowed eyes randomized to observation or laser to receive aflibercept if prespecified VA worsening criteria were met. Specifically, aflibercept was initiated if VA decreased from baseline by ≥ 10 ETDRS letters at one visit or by 5 to 9 letters at two consecutive visits. Through two years, this endpoint was met in 34% (80/326) and 25% (60/240) of eyes in the observation and laser arms, respectively. Eyes randomized to laser were less likely than eyes randomized to observation to initiate aflibercept therapy (P=0.01). Among eyes that lost VA and initiated aflibercept treatment, the median number of injections before the two-year endpoint was seven and nine in the laser and observation arms, respectively, notably similar to the median of eight injections given to the aflibercept population as a whole.

Despite 25% to 34% of eyes in the laser and observation arms experiencing a clinically meaningful decrease in VA during the trial, at the two-year time point, the percentage of eyes with at least a 5-letter VA decrease was 19% (39/208), 17% (36/212) and 16% (33/205) in the observation, laser and aflibercept groups, respectively; there were no statistically significant differences between the groups. The percentage of eyes with VA of 20/20 or better at two years was 66%, 71% and 77%, in the observation, laser and aflibercept groups, respectively, with only the difference between observation and aflibercept being statistically significant (P=0.03). This was the only secondary visual outcome among five that resulted in a statistically significant difference between any of the arms. A prespecified exploratory area under the curve analysis found small absolute differences that were statistically significant between the aflibercept (+1.5 letter), laser (0 letters, P<0.001 vs. aflibercept) and observation (-0.4 letters, P<0.001 vs. aflibercept) groups.

WHAT THIS MEANS FOR CLINICAL PRACTICE

Well-controlled Phase 3 trials such as VISTA/VIVID and RIDE/RISE have demonstrated overwhelming superiority of intravitreal anti-VEGF therapy over macular laser or observation for the management of CI-DME with VA loss.^6,7,11 Such data provides strong support to treating CI-DME with VA loss on a population basis.

Similarly, the landmark Diabetic Retinopathy Study found that treatment of high-risk PDR was far superior to observation.² Subsequently, PDR is a well-accepted tipping point at which therapy for DR is initiated, guided more recently by the possible utility of pharmacotherapy.^12,13

In comparison, the decision to initiate therapy for both high-risk NPDR without DME and CI-DME with good VA requires a more detailed consideration of the risk/benefit ratio on an individualized basis. In both situations, prospective data can now be used to support observation, laser (PRP for high-risk NPDR; macular laser for CI-DME with good VA) or intravitreal anti-VEGF therapy.

Many factors play into incorporation of these data into our daily clinical practices, some common to both patient populations. First, on average, patients enrolled in PANORAMA and Protocol V had reasonable control of their diabetes and blood pressure. Furthermore, these patients were selected from clinical practices because of their willingness to attend frequent, lengthy clinical visits and because of their ability to adequately provide informed consent. Therefore, the studied populations may not be completely representative of patients with diabetes in routine clinical care, where non-compliance¹⁴ is one of many well-recognized challenges to optimizing long-term outcomes.

Critical to keep in mind through all of these analyses is that, ultimately, vision and visual function are what matter the most to our patients. While we as physicians look carefully at imaging biomarkers gleaned from OCT, fundus photographs and angiograms for guidance on how to treat patients, the relationship between these and VA is limited,¹⁵ and patients fundamentally do not care what their imaging looks like; they care about their vision. Patients want to see well today, tomorrow and for the rest of their lives.

In both PANORAMA and Protocol V, enrolled patients by and large were asymptomatic or mildly symptomatic, with excellent average VA in both populations. In practice, the bar to initiate intravitreal therapy is much higher in asymptomatic patients compared to patients with VA impairment who well recognize their blurry vision and desire to see better. While intravitreal injections are exceptionally well tolerated, there are well-recognized risks.¹⁶ We need to always appreciate the patient perspective.

Nevertheless, multiple studies have suggested that health-related quality of life (HR-QOL) may progressively decline as DR severity worsens, even among NPDR stages without DME.^17-20 Specifically, there may be a significant negative QOL inflection point before PDR development. For example, in the Los Angeles Latino Eye Study (LALES), a cohort study of 1064 patients with diabetes who underwent ocular examinations and detailed HR-QOL assessments, a DRSS level as early as moderate NPDR (level 43) was found to represent a threshold beyond which progressive decline in HR-QOL measures was observed.¹⁸ In the LALES, the correlation of decreasing HR-QOL with increasing severity of DR was significant for all QOL sub-scales considered, including driving difficulties and vision-related dependency.

At a more practical level, evidence exists that the impact of DR translates into systemic risk. For example, an epidemiologic study in Singapore involving over 10,000 people found that the presence of mild to moderate NPDR was independently associated with an increased likelihood of falling compared with persons with diabetes but without DR.¹⁸

CONCLUSION

In Protocol V, 66% to 75% of CI-DME eyes with good vision did not experience a clinically meaningful drop in VA through two years and received no aflibercept injections. In comparison, a smaller, but clinically relevant 25% to 34% of patients not treated with initial anti-VEGF therapy experienced a drop in VA. These patients subsequently received nearly the same number of aflibercept injections through the same endpoint as those initially randomized to aflibercept therapy.

This is a decision that ultimately patients need to make in the context of their life circumstances. Fortunately, we now have prospective data to inform our discussions with patients.

PANORAMA showed strong and clear anatomic benefit with aflibercept treatment with significant DR severity improvements and significant reductions in the development PDR and CI-DME compared to sham treatment; but there has yet to be evidence of functional visual improvements among treated patients. While there is data from exudative diseases including neovascular AMD and CI-DME with VA loss that the earlier one intervenes the better outcomes can be achieved at a population level,^{6, 11, 21} we have limited data at this time to show that better long-term functional outcomes or reduction in long-term treatment burden can be achieved by treating early, before eyes develop early PDR or mild CI-DME.

Both PANORAMA and Protocol V provide high-quality data to inform our discussions with patients in the process of delivering individualized care. Longer-term data will be forthcoming as PANORAMA is ultimately a two-year study, and another trial studying a similar population, the DRCR Network Protocol W, is ongoing. OM

REFERENCES

1. Bakri SJ, Wolfe JD, Regillo CD, et al. Evidence-Based Guidelines for Management of Diabetic Macular Edema. Journal of VitreoRetinal Diseases. 2019. EPub Ahead of Print Online April 24.
2. Preliminary report on effects of photocoagulation therapy. The Diabetic Retinopathy Study Research Group. American journal of ophthalmology. 1976;81:383-396.
3. PANORAMA, Study of the Efficacy and Safety of Intravitreal Aflibercept for the Improvement of Moderately Severe to Severe NPDR. ClinicalTrials.gov :NCT02718326.
4. Grading diabetic retinopathy from stereoscopic color fundus photographs--an extension of the modified Airlie House classification. ETDRS report number 10. Early Treatment Diabetic Retinopathy Study Research Group. Ophthalmology. 1991;98(5 Suppl):786-806.
5. Fundus photographic risk factors for progression of diabetic retinopathy. ETDRS report number 12. Early Treatment Diabetic Retinopathy Study Research Group. Ophthalmology. 1991;98(5 Suppl):823-833.
6. Brown DM, Nguyen QD, Marcus DM, et al. Long-term outcomes of ranibizumab therapy for diabetic macular edema: the 36-month results from two phase III trials: RISE and RIDE. Ophthalmology. 2013;120:2013-2022.
7. Brown DM, Schmidt-Erfurth U, Do DV, et al. Intravitreal Aflibercept for Diabetic Macular Edema: 100-Week Results From the VISTA and VIVID Studies. Ophthalmology. 2015;122:2044-2052.
8. Campochiaro PA, Brown DM, Pearson A, et al. Sustained delivery fluocinolone acetonide vitreous inserts provide benefit for at least 3 years in patients with diabetic macular edema. Ophthalmology. 2012;119:2125-2132.
9. Boyer DS, Yoon YH, Belfort R, Jr., et al. Three-year, randomized, sham-controlled trial of dexamethasone intravitreal implant in patients with diabetic macular edema. Ophthalmology. 2014;121:1904-1914.
10. Baker CW, Glassman AR, Beaulieu WT, et al. Effect of Initial Management With Aflibercept vs Laser Photocoagulation vs Observation on Vision Loss Among Patients With Diabetic Macular Edema Involving the Center of the Macula and Good Visual Acuity: A Randomized Clinical Trial. JAMA : the journal of the American Medical Association. 2019. Published online April 29.
11. Wykoff CC, Marcus DM, Midena E, et al. Intravitreal Aflibercept Injection in Eyes With Substantial Vision Loss After Laser Photocoagulation for Diabetic Macular Edema: Subanalysis of the VISTA and VIVID Randomized Clinical Trials. JAMA Ophthalmol. 2017;135:107-114.
12. Gross JG, Glassman AR, Liu D, et al. Five-Year Outcomes of Panretinal Photocoagulation vs Intravitreous Ranibizumab for Proliferative Diabetic Retinopathy: A Randomized Clinical Trial. JAMA Ophthalmol. 2018;136:1138-1148.
13. Sivaprasad S, Prevost AT, Vasconcelos JC, et al. Clinical efficacy of intravitreal aflibercept versus panretinal photocoagulation for best corrected visual acuity in patients with proliferative diabetic retinopathy at 52 weeks (CLARITY): a multicentre, single-blinded, randomised, controlled, phase 2b, non-inferiority trial. Lancet. 2017;389(10085):2193-2203.
14. Obeid A, Gao X, Ali FS, et al. Loss to Follow-Up in Patients with Proliferative Diabetic Retinopathy after Panretinal Photocoagulation or Intravitreal Anti-VEGF Injections. Ophthalmology. 2018;125:1386-1392.
15. Ou WC, Brown DM, Payne JF, et al. Relationship Between Visual Acuity and Retinal Thickness During Anti-Vascular Endothelial Growth Factor Therapy for Retinal Diseases. American journal of ophthalmology. 2017;180:8-17.
16. Li AL, Wykoff CC, Wang R, et al. ENDOPHTHALMITIS AFTER INTRAVITREAL INJECTION: Role of Prophylactic Topical Ophthalmic Antibiotics. Retina. 2016;36:1349-1356.
17. Mazhar K, Varma R, Choudhury F, et al. Severity of diabetic retinopathy and health-related quality of life: the Los Angeles Latino Eye Study. Ophthalmology. 2011;118:649-655.
18. Gupta P, Liang Gan AT, Kidd Man RE, et al. Impact of Incidence and Progression of Diabetic Retinopathy on Vision-Specific Functioning. Ophthalmology. 2018;125:1401-1409.
19. Pereira DM, Shah A, D’Souza M, et al. Quality of Life in People with Diabetic Retinopathy: Indian Study. J Clin Diagn Res. 2017;11:NC01-NC06.
20. Willis JR, Doan QV, Gleeson M, et al. Vision-Related Functional Burden of Diabetic Retinopathy Across Severity Levels in the United States. JAMA Ophthalmol. 2017;135:926-932.
21. Writing Committee for the UKA-RMDEMRUG. The neovascular age-related macular degeneration database: multicenter study of 92 976 ranibizumab injections: report 1: visual acuity. Ophthalmology. 2014;121:1092-1101.

About the Author

Charles C. Wykoff, MD, PhD, is director of research for Retina Consultants of Houston and deputy chair of ophthalmology at Blanton Eye Institute, Houston Methodist Hospital in Texas.

Disclosures: Dr. Wykoff revealed relationships with Adverum; Aerpio; Alimera Sciences; Allegro; Allergan; Alnylam; Apellis; Bayer; Clearside Biomedical; DORC; EyePoint; Genentech/Roche; Kodiak; Neurotech; Notal Vision; Novartis; ONL Therapeutics; Opthea; PolyPhotonix; Recens Medical; Regeneron; Regenxbio; Samsung; Santen.