Emerging real-world evidence of anti-VEGF for diabetic macular oedema

Reviewed by Dr Marko Lukic and Dr Thomas A Ciulla

One-year results from a retrospective cohort study that assessed the effectiveness of aflibercept (Eylea, Regeneron Pharmaceuticals) in treatment-naïve patients with diabetic macular oedema (DMO), conducted at Moorfields Eye Hospital, London, compare favourably to outcomes seen in randomised clinical trials (Figures 1 and 2).¹ Follow-up over 36 months further confirmed maintenance of vision long term with continued retreatment if needed.²

At 3 years, eyes with worse visual acuity (VA) at baseline (<69 early treatment diabetic retinopathy study [ETDRS] letters), accounting for 58% of the study cohort, achieved greater than a two-line vision gain (+11.27 letters) from baseline, while stable vision was maintained in eyes having initially better VA (69–80 letters). Some 89% of patients overall received the recommended treatment initiation protocol of five consecutive monthly injections.

Lead author of both studies, Dr Marko Lukic, explained: “Clinical trials provide what I term the ideal results with respect to treatment efficacy and safety. Real-world studies provide us with more realistic results of treatment effectiveness in everyday practice using practical protocols.”

He added: “At month 36, one quarter of eyes gained at least three lines of vision from baseline and the final mean VA was around 69 ETDRS letters, meeting driving criteria.”

Overall, the mean VA gain was 9.9 letters with a mean of 6.9 injections over 12 months. The independent DRCR.net Protocol T study showed better improvements in VA with an average of nine injections in the first year.³ However, following strict protocols suggested by clinical trials is difficult to maintain in real-life settings, remarked Dr Lukic.

He explained: “We mostly commence treatment of new patients with DMO using intravitreal aflibercept injections and always aim to provide five initial monthly injections and then treatment as needed, which for most patients means three bimonthly injections and an average of seven to eight injections in year one. It is important to secure the maximum vision improvement in the first year and, after that, to maintain stable vision with retreatment as needed.”

“We divided our study cohort into two subgroups according to baseline VA and showed that those with worse baseline VA (<69 letters, 20/50 or worse) had greater vision gains than those with better presenting vision. However, those with baseline VA ≥69 letters still achieved good maintained vision. And overall, when you compare these two subgroups, those who commenced treatment earlier (i.e., with better VA at baseline) had better results in terms of final VA letter score (approximately 75 vs. 63 letters, respectively, at month 36).”

If continuity of anti-vascular endothelial growth factor (VEGF) treatment for DMO is maintained and if there is no presence of significant macular ischaemia, then Dr Lukic believes good vision outcomes can be routinely achieved in everyday practice. Adherence to the initiation protocol of monthly loading followed by bimonthly injections in year one is extremely important, he said.

Dr Lukic added: “Diabetic macular oedema is a chronic condition and regular retreatment is needed for most patients. Retreatment beyond year one is led largely by optical coherence tomography appearance, i.e., those with persistent fluid or who develop new fluid continue to receive retreatment.”

“Some patients received more than three injections over both year two and year three. On the other hand, 13% of aflibercept-treated eyes received no further injections beyond the loading phase by the end of 3 years since initiation of treatment.”

There is always a small subgroup of patients who do not respond as expected and who may need to be switched to another treatment (e.g., 3/102 aflibercept-treated eyes in the cited 12-month study), Dr Lukic said. “Steroids may have a role in selected cases, for example, pseudophakic eyes with refractory or persistent DMO despite continued anti-VEGF therapy.”

Initial treatment intensity matters

In a large real-word study in the United States, it was found that patients with DMO consistently achieved poorer vision outcomes and received fewer anti-VEGF injections than those in randomised clinical trials. A retrospective electronic medical record (EMR) database analysis of 28,658 treatment-naïve eyes treated with anti-VEGF for DMO from 2013 to 2018 found a mean VA gain of 4.2 letters with a mean of 6.4 injections over 1 year.⁴ This large sample of DMO patient eyes included all-comers who received at least one anti-VEGF injection and had 12-month follow-up data.

Half of all patient eyes received six or fewer injections over the year. Vision outcomes correlated with treatment intensity and, as expected, mean letter gains were higher in those with poorer starting vision. Outcomes similar to clinical trials were achieved in a subgroup of patients with moderately severe baseline visual impairment (20/70 to 20/200), showing a 1-year gain of 10.3 letters with ≥10 injections (from a mean baseline VA of 47.2 letters).

How to avoid a gap year

Commenting on the persistent reported gap between clinical trials and real-world outcomes, lead study investigator Dr Thomas A, Ciulla explained: “Our findings showing only limited vision improvement are attributed largely to two key factors: first, real-world patients differ from those in clinical trials and second, they are undertreated with a lower frequency of anti-VEGF injections.”

Patients with DMO can have chronic advanced oedema with ischaemia and atrophy, while some present with very good VA, Dr Ciulla said. “But even if real-world DMO patients differ from clinical trial participants, we should do everything we can for them: encourage them to comply with recommended treatment regimens and make sure they receive an appropriate level of treatment intensity.”

“Our data suggest that, to maximise real-world outcomes, the average DMO patient requires intense anti-VEGF therapy during the first year of treatment,” added Dr Ciulla. “Moreover, vision gains generally approach a plateau at a treatment frequency of nine or more injections in year one.” Slightly better overall outcomes were previously reported in another US EMR database study of DMO eyes that underwent ≥3 monthly anti-VEGF injections within 4 months of the first injection, where the mean 12-month visual improvement was 5.5 letters.⁵

Importance of complete treatment initiation schedules

In France, a prospective cohort study (APOLLON) of aflibercept in private and public clinical practice showed improvements in functional and anatomical outcomes over 12 months in both treatment-naïve and previously treated patients with DMO (Figure 3).⁶

The mean (SD) change in best-corrected visual acuity was +7.8 (12.3) letters in treatment-naïve patients and +5.0 (11.3) letters in previously treated patients at month 12, with patients receiving a mean of 7.6 injections, slightly lower than that in the aflibercept DMO registration trials. The proportions gaining 15 letters or more from baseline were 29.9% and 18.6%, respectively.

Of note, only 42.2% (n=62) of the study cohort received the complete five initial monthly doses as indicated in the label regimen. Table 1 shows visual outcomes stratified by initial treatment exposure in the treatment-naïve cohort, with better visual outcomes achieved in patients receiving the recommended treatment initiation schedule.

“Great steps forward have been made in the management of retinal disease in the past decade,” observed Dr Ciulla. “However, many of our current modalities require fixed, frequent treatments and what we ultimately need is better treatments that last longer.”

---

Marko Lukic, FEBO
E: m.lukic@nhs.net
Dr Lukic is a locum consultant ophthalmologist, National Institute for Health Research (NIHR) Biomedical Research Centre, Moorfields Eye Hospital NHS Foundation Trust and UCL Institute of Ophthalmology, London, UK. The research reported in this article was supported by the NIHR Biomedical Research Centre. He has no financial interests in the subject of this article.

Thomas A Ciulla, MD, MBA
Dr Ciulla is volunteer clinical professor of Ophthalmology at Indiana University School of Medicine, United States, a board member of Midwest Eye Institute, Indianapolis, and chief medical officer/chief development officer at Clearside Biomedical. Dr Ciulla has no financial interests in the subject of this article.

REFERENCES

1. Lukic M, Williams G, Shalchi Z, et al. Intravitreal aflibercept for diabetic macular oedema: Moorfields' real-world 12-month visual acuity and anatomical outcomes. Eur J Ophthalmol. 2020;30:557-562.
2. Lukic M, Gwyn Williams G, Shalchi Z, et al. Intravitreal aflibercept for diabetic macular oedema in real-world: 36-month visual acuity and anatomical outcomes. Eur J Ophthalmol. 19 May 2020. DOI:10.1177/1120672120925034.
3. The Diabetic Retinopathy Clinical Research Network. Aflibercept, bevacizumab, or ranibizumab for diabetic macular edema. N Engl J Med. 2015;372:1193-203.
4. Ciulla TA, Pollack JS, Williams DF. Visual acuity outcomes and anti-VEGF therapy intensity in diabetic macular edema: a real-world analysis of 28 658 patient eyes. Br J Ophthalmol. 7 Apr 2020. DOI: 10.1136/bjophthalmol-2020-315933.
5. Ciulla TA, Bracha P, Pollack J, Williams DF. Real-world outcomes of anti-vascular endothelial growth factor therapy in diabetic macular edema in the United States. Ophthalmol Retina. 2018;2:1179-1187.
6. Korobelnik JF, Daien V, Faure C, et al. Real-world outcomes following 12 months of intravitreal aflibercept monotherapy in patients with diabetic macular edema in France: results from the APOLLON study. Graefes Arch Clin Exp Ophthalmol. 2020;258:521-528.