Imaging markers for diabetic macular oedema: Forecasting visual outcomes beyond retinal thickness

By
Enrico Borrelli
October 18, 2026

Overview

In a post hoc analysis of 60 KINGFISHER trial eyes, baseline central subfield thickness and an OCTA composite of superficial perfusion density and vessel length density were independently associated with week-52 visual acuity; baseline hyperreflective foci volume was associated with a smaller visual-acuity gain. The findings support investigation of structural and vascular imaging markers beyond retinal thickness, while the study’s sample size, optional OCTA acquisition, device restriction, and model limitations constrain interpretation.

Background

Anti-VEGF therapy has altered the natural history of diabetic macular oedema (DMO), but individual visual outcomes remain difficult to predict. Central subfield thickness explains only a limited proportion of visual-acuity variation, and anatomical and functional outcomes may not correspond closely. Structural OCT studies have evaluated markers including disorganisation of the retinal inner layers, external limiting membrane and ellipsoid zone integrity, and intraretinal hyperreflective foci. OCT angiography (OCTA) also enables quantitative assessment of retinal capillary networks and diabetic macular ischaemia across separate capillary plexuses.

Data Highlights

Measure or finding	Reported result
Study population	60 eyes from the KINGFISHER trial with available baseline spectral-domain OCT and OCTA measurements
Mean visual acuity	62.9 letters at baseline; 75.4 letters at week 52
Baseline visual acuity predictor	Greater baseline central subfield thickness was the only independent predictor reported
Week-52 visual acuity associations	Baseline central subfield thickness and a composite of superficial perfusion density and vessel length density were independently associated
Visual-acuity gain association	Higher baseline hyperreflective foci volume was independently associated with a smaller gain
Model variance explained	19%–32%; models were not adjusted for multiple comparisons
Systematic review	27 studies evaluated 75 candidate biomarkers; five were associated with reduced visual acuity at two or more time points, with moderate certainty of evidence

Key Findings

In the KINGFISHER post hoc analysis, mean visual acuity increased from 62.9 letters at baseline to 75.4 letters at week 52.
Greater baseline central subfield thickness was the only independent predictor of baseline visual acuity reported in the analysis.
Baseline central subfield thickness and a composite index of superficial perfusion density and vessel length density were independently associated with week-52 visual acuity.
Higher baseline hyperreflective foci volume was independently associated with a smaller visual-acuity gain.
A systematic review and meta-analysis of 27 studies and 75 candidate biomarkers found five associated with reduced visual acuity at two or more time points, with moderate certainty of evidence; long-term investigations identified DRIL and outer retinal alterations as features closely associated with vision after years of anti-VEGF therapy.
The source reports important limitations: OCTA was not required by the parent protocol, analysis was restricted to Heidelberg Spectralis scans, models explained 19%–32% of variance, and multiple comparisons were not adjusted for.

Clinical Implications

The reported associations identify OCT and OCTA measures for further prognostic evaluation, but the source does not establish transferable clinical thresholds or individual-level prediction. OCTA measurements depend on device, scan protocol, segmentation and slab definitions, artefact removal, and binarisation methods, so measurements and thresholds may not be interchangeable across centres.

Conclusion

The analysis adds OCTA vascular measures and hyperreflective foci volume to the imaging features associated with visual outcomes in DMO. Larger, prespecified imaging studies and consistent measurement methods are needed to validate these exploratory associations.