Evaluation of real-world early response of DMO to aflibercept therapy to inform future clinical trial design of novel investigational agents

Abstract

New clinical trials for diabetic macular oedema (DMO) are being designed to prove superiority over aflibercept when this agent is already very effective in improving visual acuity (VA) and DMO. The aim of this study was to determine the optimal inclusion–exclusion criteria for trials to aim for superiority in visual outcomes with newer agents. As Phase 1 studies are short duration, we aimed to evaluate the early response of aflibercept in a real-world cohort initiated on monthly aflibercept for 3 consecutive injections and observed the effects at 4 months. The sub-optimal responders were pre-defined based on different cut-offs for VA and central sub-field thickness (CST). 200 patients with treatment naïve DMO treated with 3 loading doses of aflibercept were included in the study. We found that those presenting with baseline VA of 35–54 ETDRS letters (n = 43) had higher proportion of sub-optimal responders compared to other categories (p < 0.001). Patients with baseline CST of less than 400 µm (n = 96) responded less well functionally and anatomically to loading dose than eyes with baseline CST of 400 µm or more (n = 104, p = 0.02), indicating that eyes with CST ≥ 400 µm is another inclusion criteria. There was minimal correlation between change in CST and change in VA at 4 months (r = − 0.27), suggesting that both these inclusion criteria are non-exclusive. However, for maximal efficacy, patients that meet both these inclusion criteria are more likely to show benefit from an alternative intervention. New trials should aim to include patients with treatment naïve DMO with VA between 35–54 letters and CST of 400 µm or more when aflibercept is used as the comparator.

Predictive effect of TCED-HFV grading and imaging biomarkers on anti-VEGF therapy in diabetic macular edema

“…However, 14.8% of the eyes exhibited a decrease in BCVA of more than five letters. As in previous studies [ 7 , 12 , 13 ], our findings confirm a difference in the response of affected eyes to anti-VEGF treatment. We observed a significant correlation between the number of injections and the anatomical results.…”

“…Eyes present with VMT or ERM at baseline are less likely to reach the endpoint of CST < 300 μm. However, the decrease in CST does not reflect the final VA results, nor is it a clear predictor of VA [ 12 , 22 , 23 ]. An analysis of the RIDE and RISE trials revealed that patients with no significant changes in fovea thickness after treatment with anti-VEGF had similar visual acuity and DR improvements to those with immediate retinal thinning [ 24 ].…”

Predictive effect of TCED-HFV grading and imaging biomarkers on anti-VEGF therapy in diabetic macular edema

“…However, 14.8% of the eyes exhibited a decrease in BCVA of more than ve letters. As in previous studies [10], our ndings con rm a difference in the response of affected eyes to anti-VEGF treatment. We observed a signi cant correlation between the number of injections and the anatomical results.…”

“…However, the decrease in CST does not re ect the nal VA results, nor is it a clear predictor of VA [19][20][21]. An analysis of the RIDE and RISE trials revealed that patients with no signi cant changes in fovea thickness after treatment with anti-VEGF had similar visual acuity and DR improvements to those with immediate retinal thinning [22].…”

Indicators of Visual Prognosis in Diabetic Macular Oedema

“…However, in real-life studies with ranibizumab, eyes with baseline VA less than 37 letters fail to achieve good VA that met driving standards. This outcome was also noted in a study on early response to aflibercept after 3 loading injections in real-life suggesting that there are more eyes with irreversible structural changes in eyes with VA less than 37 letters [ 10 ]. Based on available evidence, early treatment with anti-VEGF agents is recommended.…”