Published in Retina

Real-World Durability vs. Trial Durability: Closing the Gap

This is editorially independent content supported by advertising from Regeneron
5 min read

Join Drs. Rishi Singh and Jayanth Sridhar to compare safety and efficacy data on second-generation anti-VEGF agents from clinical trials and real-world practice.

Pivotal anti-vascular endothelial growth factor (VEGF) trials seldom reflect the pace of a busy injection clinic. Trial participants are generally treatment-naive, adherent, and managed on a fixed protocol; in routine practice, patients miss visits, arrive pretreated, and have intervals extended based on clinical judgment.
This gap explains why many retina specialists view durability data for second-generation agents with skepticism. Registry data on aflibercept 8 mg now provide an early test of whether trial-level intervals translate to routine care.
In this episode of Evidence Based Retina, Rishi Singh, MD, FASRS, sits down with Jayanth Sridhar, MD, chief of ophthalmology at Olive View-UCLA Medical Center, to compare the durability of aflibercept 8 mg in the PULSAR and PHOTON trials with real-world outcomes from the American Academy of Ophthalmology IRIS Registry.

Anti-VEGF durability fast facts

  • In an IRIS Registry analysis of eyes with neovascular age-related macular degeneration (AMD), about one-third had anti-VEGF injection intervals of less than 8 weeks at the end of year 1. By the end of year 3, 42% had discontinued treatment, defined as no injection for more than 6 months.1
  • A 2026 network meta-analysis of 16 studies involving 6,758 patients found no statistically significant differences in best-corrected visual acuity (BCVA) change among aflibercept, ranibizumab, bevacizumab, brolucizumab, and faricimab for neovascular AMD.2
  • In PULSAR, 1,011 patients with neovascular AMD were randomized to receive aflibercept 8 mg every 12 or 16 weeks, or 2 mg every 8 weeks. At week 48, both 8-mg regimens were noninferior (+6.7 and +6.2 letters vs. +7.6).3
  • In PHOTON, aflibercept 8 mg given every 12 or 16 weeks after 3 monthly loading doses led to noninferior week-48 BCVA gains in diabetic macular edema (DME; +8.8 and +7.9 letters, respectively) compared with aflibercept 2 mg given every 8 weeks after 5 loading doses (+9.2 letters).4
  • Aflibercept 8 mg is formulated at 114.3 mg/mL, compared with 40 mg/mL for the 2-mg dose, increasing the injection volume from 0.05 mL to 0.07 mL.5

How aflibercept 8 mg durability holds up outside the protocol

Extended labeled intervals matter only if they withstand the conditions of routine practice.
The 2-year trial results set the benchmark against which registry data should be evaluated:
  • PULSAR (neovascular AMD): Among patients who completed week 96, 87% in the every-12-week arm had a last assigned interval of at least 12 weeks. In the every-16-week arm, 78%, 53%, and 31% qualified for intervals of at least 16, 20, and 24 weeks, respectively. Through week 96, the mean number of active injections was 9.7 and 8.2 with 8 mg, compared with 12.8 with 2 mg.5
  • PHOTON (DME): Among patients completing week 96, 88% (every-12-week arm) and 83% (every-16-week arm) maintained intervals of 12 or more and 16 or more weeks, respectively; 43% and 47% qualified for 20 or more weeks. Mean active injections were 9.5 and 7.8 with 8 mg vs. 13.8 with 2 mg.6

"We have trials where patients are treatment-naive, well treated, and they don't miss many visits compared to our normal population."—Rishi P. Singh, MD, FASRS

Registry durability in treatment-naive and switched eyes

IRIS Registry data presented at AAO 2025 included more than 1,300 treatment-naive neovascular AMD eyes, nearly 400 treatment-naive DME eyes, as well as more than 8,300 neovascular AMD eyes and more than 2,200 DME eyes switched from aflibercept 2 mg, followed for approximately 58 weeks.7
  • Treatment-naive eyes reached an injection interval of approximately 12 weeks in both neovascular AMD and DME, with median gains of 4 and 5 letters, respectively, among eyes starting at 20/50 or worse; overall, vision was maintained.7
  • Switched neovascular AMD eyes extended their intervals by approximately 2 weeks with stable vision; an expanded cohort of 11,284 switched eyes showed the same median extension, with visual acuity unchanged regardless of the pre-switch interval.7,8
  • Switched DME eyes (2,280 eyes; median follow-up, 57 weeks) extended their intervals by a median of 2 weeks, reaching a median of 11.9 weeks with vision maintained. Treatment-naive DME eyes starting at 20/50 or worse gained a median of 5 letters.7,9
Treatment-naive registry eyes landed near the 12-week interval seen in most trial patients. Switched neovascular AMD eyes gained a more modest 2 weeks or so, which is expected; PULSAR enrolled only treatment-naive patients,5 whereas switch decisions in practice tend to target eyes already on short intervals or with an incomplete response to 2 mg.
Switched DME eyes, by contrast, came closer to trial-level durability.9 Other limitations include roughly 1 year of follow-up, visual acuity recorded without consistent best correction, and electronic records that may not capture full treatment histories.7 Of note, these analyses have so far been reported as meeting presentations.

"Registry studies capture the broadest kind of group of patients: all the patients who are receiving the drug, not just the patients who met inclusion criteria for a trial and/or were more likely to come in for their visits and be more responsive in a clinical trial."—Jayanth Sridhar, MD

Safety results

The larger injection volume and greater drug exposure raised concerns about IOP elevation, retinal pigment epithelium (RPE) tears in eyes with large pigment epithelial detachments (PEDs), and intraocular inflammation.
In PULSAR, ocular adverse events in the study eye occurred in 39%, 38%, and 39% of the 8-mg every-12-week, 8-mg every-16-week, and 2-mg groups, respectively, through week 48.3 In PHOTON, ocular treatment-emergent adverse events in the study eye occurred at similar rates across groups through week 96 (37.1% to 45.4%).6

"One of the best things we've seen is safety, even though it was not designed to be a safety study, eight milligrams was very safe; it is well-tolerated in the real world."—Jayanth Sridhar, MD

In closing

Registry data narrow the trial-to-practice gap more for treatment-naive patients than for patients who have switched.
For a treatment-naive eye, clinicians starting aflibercept 8 mg can reasonably target a 12-week interval. For an eye switched after years of 2-mg therapy, extending the interval by 2 weeks while maintaining stable vision is more realistic.
Peer-reviewed publication of these registry cohorts, with longer follow-up and best-corrected visual acuity, will show whether this early pattern holds.
For more information on second-generation anti-VEGF agents, check out: Decreasing Treatment Burdens With Second-Generation Anti-VEGF Agents.
EBR spotify banner
  1. MacCumber MW, Yu JS, Sagkriotis A, et al. Antivascular endothelial growth factor agents for wet age-related macular degeneration: an IRIS registry analysis. Can J Ophthalmol. 2023;58(3):252-261. doi:10.1016/j.jcjo.2021.10.008
  2. Chen KY, Chan HC, Chan CM. Comparative effectiveness and safety landscape of anti-VEGF therapies for neovascular age-related macular degeneration: insights from a systematic review and network meta-analysis. Biomed Pharmacother. 2026;196:118881. doi:10.1016/j.biopha.2025.118881
  3. Lanzetta P, Korobelnik JF, Heier JS, et al. Intravitreal aflibercept 8 mg in neovascular age-related macular degeneration (PULSAR): 48-week results from a randomised, double-masked, non-inferiority, phase 3 trial. Lancet. 2024;403(10432):1141-1152. doi:10.1016/S0140-6736(24)00063-1
  4. Brown DM, Boyer DS, Do DV, et al. 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
  5. Korobelnik JF, Lanzetta P, Leal S, et al. Intravitreal aflibercept 8 mg in neovascular age-related macular degeneration: 96-week results from the randomized phase 3 PULSAR trial. Ophthalmology. 2026;133(1):39-50. doi:10.1016/j.ophtha.2025.08.022
  6. Do DV, Wykoff CC, Sivaprasad S, et al. Intravitreal aflibercept 8 mg for diabetic macular edema: ninety-six-week results from the randomized phase 2/3 PHOTON trial. Ophthalmology. 2026;133(5):577-588. doi:10.1016/j.ophtha.2025.10.028
  7. Do DV, Moll K, Murdock D, McCann E, Sherman S; ELARA Investigators. Real-world use of aflibercept 8 mg in eyes with nAMD or DME: analysis of the American Academy of Ophthalmology IRIS Registry (Intelligent Research in Sight) and ELARA study update. Presented at: American Academy of Ophthalmology Retina Subspecialty Day; October 17-18, 2025; Orlando, FL.
  8. Leng T. Real-world use of aflibercept 8 mg in eyes with neovascular age-related macular degeneration. Presented at: Retina 2026; January 17-23, 2026; Waikoloa, HI.
  9. Javaheri M, Moll K, Murdock DJ, McCann E, Sherman S. Real-world use of aflibercept 8 mg in eyes with diabetic macular edema (DME): analysis of the IRIS Registry. Invest Ophthalmol Vis Sci. 2026;67(7):2676.
Rishi P. Singh, MD, FASRS
About Rishi P. Singh, MD, FASRS

Rishi P. Singh, MD, FASRS, is the Chair of the Department of Ophthalmology at Mass General Brigham, overseeing ophthalmology across Massachusetts Eye and Ear, Massachusetts General Hospital, Brigham and Women’s Hospital, and affiliated sites. He is also a Professor of Ophthalmology at Harvard Medical School.

Previously, Dr. Singh served as Vice President and Chief Medical Officer at Cleveland Clinic Martin Health in Stuart, Florida, and as a staff surgeon at the Cleveland Clinic, where he was also Professor of Ophthalmology at the Cleveland Clinic Lerner College of Medicine in Cleveland, Ohio. He received both his undergraduate degree in medical science and his medical degree from Boston University, completing his internship at Tufts University. Dr. Singh went on to complete his ophthalmology residency at the Massachusetts Eye and Ear Infirmary/Harvard Medical School and a medical and surgical vitreoretinal fellowship at the Cole Eye Institute at the Cleveland Clinic.

Dr. Singh specializes in the management of complex retinal diseases, including diabetic retinopathy, retinal vein occlusions, retinal detachment, and age-related macular degeneration. He has authored over 300 peer-reviewed publications, books, and book chapters and serves as Principal Investigator for numerous national and international clinical trials aimed at improving outcomes for patients with retinal diseases.

He is the founder and past president of the Retina World Congress, chairs some of the largest continuing medical education meetings in retina, and serves on editorial boards and review panels for major ophthalmology journals. His leadership has extended into digital innovation, having helped lead enterprise-wide implementation of clinical technologies including Epic modules, digital informed consent, and patient-facing kiosks.

Dr. Singh has received multiple accolades for his contributions to ophthalmic research and innovation, including the Alpha Omega Alpha Research Award, the American Society of Retina Specialists Young Investigator Award, and the J. Donald Gass Beacon of Sight Award. He also leads The Center for Ophthalmic Bioinformatics, a research initiative focused on leveraging big data and artificial intelligence to advance understanding and treatment of retinal disease.

Rishi P. Singh, MD, FASRS
💙 Evidence Based Retina
Astellas
Regeneron
Bausch + Lomb Bi-Blade
EyePoint, Inc.
Apellis Pharmaceuticals
Norlase