Diabetic retinopathy (DR) is treated only after vascular lesions appear, yet the neural retina shows measurable dysfunction earlier: in longitudinal studies by the Berkeley group, retinal zones with delayed multifocal electroretinogram (mfERG) implicit time at baseline predicted the precise locations of future microaneurysms, dot haemorrhages, and sight-threatening macular oedema one to three years later. Adams and Bearse (2012) framed the resulting translational question: which interventions can alter mfERG delays and thereby reduce subsequent vascular retinopathy or oedema; whether early neural dysfunction is causal, permissive, or merely parallel to the vasculopathy; whether the location-specific predictive models generalize across populations and devices and whether improved glycaemic control reverses neurodysfunction; and whether mfERG-derived short-term endpoints can be standardized and accepted for trials and practice. As of 2026 the core remains open: the EUROCONDOR trial (449 patients, 96 weeks) found no overall effect of topical brimonidine or somatostatin on the mfERG primary endpoint (stabilization only in the subgroup with pre-existing neurodysfunction), no intervention has been shown to prevent vascular lesions by modifying neural function, human causality is unresolved, and the predictive models still lack external validation.

Background

Diabetic retinopathy (DR) is the leading cause of vision loss in working-age adults and has traditionally been managed as a disease of the retinal blood vessels: laser, intravitreal anti-VEGF drugs and steroids are applied only after microvascular lesions — microaneurysms, dot haemorrhages, or diabetic macular oedema (DME, fluid accumulation thickening the central retina) — become clinically visible. The modern view instead treats DR as a disease of the neurovascular unit, the functional coupling of neurons, glia (mainly Müller cells) and vessels, in which neural injury occurs early. The key functional evidence comes from the multifocal electroretinogram (mfERG), a non-invasive electrophysiological test that records electrical responses from many separate locations (typically 61–103 stimulus hexagons, usually consolidated into ~35 analysis zones covering the central 45° of retina) in a session of roughly eight minutes per eye. The delay of the P1 response peak of the first-order kernel (the waveform component obtained by cross-correlating the recorded signal with the local stimulus sequence), called the implicit time (IT) and attributed mainly to bipolar-cell activity in the inner nuclear layer, is the most diabetes-sensitive mfERG measure; values are expressed as age-corrected z-scores per zone.

A series of longitudinal studies from the University of California, Berkeley group established that local mfERG IT delays precede and predict the location of future vascular lesions in diabetes: baseline zone IT z-scores predicted the appearance of new microaneurysms and dot haemorrhages at the corresponding zones within one year (sensitivity 86%, specificity 84% in a 28-patient experiment; multivariate area under the receiver-operating-characteristic (ROC) curve 0.95), over two years (AUC 0.88), and over three years for persistent or recurrent retinopathy (sensitivity 88%, specificity 98% in 18 patients followed annually). Location-specific probabilities were modelled by logistic regression with generalized estimating equations to handle the correlation among zones within an eye; for example, the one-year probability of zone-level retinopathy onset in eyes without baseline retinopathy followed

log⁡ ⁣(p1−p)=−3.8+0.18 (IT z-score)+…\log\!\left(\dfrac{p}{1-p}\right)=-3.8+0.18\,(\text{IT z-score})+\ldots

and the corresponding model for onset of sight-threatening oedema within one year was

log⁡ ⁣(p1−p)=−5.39+0.37 (IT z)−0.87 (amplitude)+0.017 (systolic blood pressure)−1.93 (gender).\log\!\left(\dfrac{p}{1-p}\right)=-5.39+0.37\,(\text{IT z})-0.87\,(\text{amplitude})+0.017\,(\text{systolic blood pressure})-1.93\,(\text{gender}).

In their 2012 review, Adams and Bearse argued that these function-based, location-specific measures provide short-term predictive endpoints that could support prophylactic treatment before vasculopathy begins — a 'function-based opportunity for early treatment intervention' — and listed what would be needed to realize it: candidate interventions tested against mfERG-derived endpoints and subsequent vascular outcomes, clarification of whether neural abnormalities are causal, permissive, or parallel to the vascular lesions, generalizability and external validation of the predictive models (including whether improving long-term glycaemic control can reverse neurodysfunction), standardization and regulatory acceptance of mfERG-based surrogate endpoints, and integration with new high-resolution in vivo imaging — optical coherence tomography (OCT, cross-sectional retinal imaging), OCT angiography (OCT-A, non-invasive mapping of capillary perfusion), and adaptive optics (cellular-resolution retinal imaging) — to identify the cellular correlates of the functional delays.

The question acquired a first serious clinical test with EUROCONDOR, the European Consortium for the Early Treatment of Diabetic Retinopathy, which built a multicentre normative mfERG database and then ran a randomized trial of topically administered neuroprotective drugs in type 2 diabetes. Full-field ERG measures (including oscillatory potentials, high-frequency wavelets generated by inner-retinal amacrine cells that are among the earliest diabetes-sensitive electrophysiological changes) and mfERG are both in use, and the field's current therapeutic armamentarium — anti-VEGF agents targeting vascular endothelial growth factor — remains directed at vascular endpoints rather than at the preceding neural dysfunction.

Problem Statement

The objective is to convert the observation that localized mfERG implicit-time delays precede diabetic vascular lesions into a function-based early-intervention strategy for diabetic retinopathy. Its components are jointly necessary for that objective and must all be addressed:

  1. Causal status. Determine, in humans, whether early localized neuroretinal dysfunction (mfERG delays) is a cause of, a permissive condition for, or a parallel phenomenon to later vascular lesions — i.e., resolve the causal-versus-permissive-versus-parallel question that current longitudinal and OCT evidence (which shows only temporal precedence) cannot settle.

  2. Therapeutic modification. Identify interventions (pharmacological or otherwise) that measurably alter mfERG implicit-time delays, and test in controlled studies whether doing so reduces the subsequent incidence or progression of vascular diabetic retinopathy or macular oedema. This includes the glycaemic question: does improving long-term glycaemic control reverse or alter neuroretinal dysfunction as measured by mfERG?

  3. Validation of the predictive models. Establish the generalizability and robustness of the location-specific logistic predictive models across larger and more diverse cohorts, populations, device implementations, and clinical settings, through external (independent-cohort) validation with standardized variable coding and harmonized mfERG acquisition and analysis.

  4. Endpoint qualification. Assemble the standardization and validation evidence required for mfERG-derived, location-specific short-term endpoints to be accepted as surrogate endpoints in clinical trials and, ultimately, clinical practice — including integration with high-resolution in vivo imaging (OCT, OCT angiography, adaptive optics) to identify the cellular-level correlates of the functional delays and refine the predictive algorithms.

No single component resolves the objective on its own. The resolution is not conditioned on an affirmative finding: the causal-status component is resolved by a reasoned adjudication among causal, permissive, and parallel; the therapeutic-modification and external-validation components are resolved by the corresponding controlled-study and independent-cohort validation analyses; and the endpoint component is resolved by mfERG-derived, location-specific endpoints meeting an explicit, citable acceptance standard (for example, formal regulatory qualification as a surrogate endpoint).

References

  1. Adams AJ, Bearse MA Jr (2012) Retinal neuropathy precedes vasculopathy in diabetes: a function-based opportunity for early treatment intervention? Clinical and Experimental Optometry 95(3):256-265. DOI 10.1111/j.1444-0938.2012.00733.x. https://doi.org/10.1111/j.1444-0938.2012.00733.x
  2. Han Y, Bearse MA Jr, Schneck ME, Barez S, Jacobsen CH, Adams AJ (2004) Multifocal electroretinogram delays predict sites of subsequent diabetic retinopathy. Investigative Ophthalmology & Visual Science 45(3):948-954. DOI 10.1167/iovs.03-1101. https://doi.org/10.1167/iovs.03-1101
  3. Han Y, Schneck ME, Bearse MA Jr, Barez S, Jacobsen CH, Jewell NP, Adams AJ (2004) Formulation and evaluation of a predictive model to identify the sites of future diabetic retinopathy. Investigative Ophthalmology & Visual Science 45(11):4106-4112. DOI 10.1167/iovs.04-0405. https://doi.org/10.1167/iovs.04-0405
  4. Bearse MA Jr, Adams AJ, Han Y, Schneck ME, Ng J, Bronson-Castain K, Barez S (2006) A multifocal electroretinogram model predicting the development of diabetic retinopathy. Progress in Retinal and Eye Research 25(5):425-448. DOI 10.1016/j.preteyeres.2006.07.001. https://doi.org/10.1016/j.preteyeres.2006.07.001
  5. Ng JS, Bearse MA Jr, Schneck ME, Barez S, Adams AJ (2008) Local diabetic retinopathy prediction by multifocal ERG delays over 3 years. Investigative Ophthalmology & Visual Science 49(4):1622-1628. DOI 10.1167/iovs.07-1157. https://doi.org/10.1167/iovs.07-1157
  6. Harrison WW, Bearse MA Jr, Schneck ME, Wolff BE, Jewell NP, Barez S, Mick AB, Dolan BJ, Adams AJ (2011) Prediction, by retinal location, of the onset of diabetic edema in patients with nonproliferative diabetic retinopathy. Investigative Ophthalmology & Visual Science 52(9):6825-6831. DOI 10.1167/iovs.11-7533. https://doi.org/10.1167/iovs.11-7533
  7. Simó R, Hernández C; European Consortium for the Early Treatment of Diabetic Retinopathy (EUROCONDOR) (2012) Neurodegeneration is an early event in diabetic retinopathy: therapeutic implications. British Journal of Ophthalmology 96(10):1285-1290. DOI 10.1136/bjophthalmol-2012-302005. https://doi.org/10.1136/bjophthalmol-2012-302005
  8. de Moraes G, Layton CJ (2016) Therapeutic targeting of diabetic retinal neuropathy as a strategy in preventing diabetic retinopathy. Clinical and Experimental Ophthalmology 44(9):838-852. DOI 10.1111/ceo.12795. https://doi.org/10.1111/ceo.12795
  9. Simão S, Costa MÂ, Sun JK, Cunha-Vaz J, Simó R; EUROCONDOR (2017) Development of a normative database for multifocal electroretinography in the context of a multicenter clinical trial. Ophthalmic Research 57(2):107-117. DOI 10.1159/000450958. https://doi.org/10.1159/000450958
  10. Santos AR, Ribeiro L, Bandello F, Lattanzio R, Egan C, Frydkjaer-Olsen U, García-Arumí J, Gibson J, Grauslund J, Harding SP, Lang GE, Massin P, Midena E, Scanlon P, Aldington SJ, Simão S, Schwartz C, Ponsati B, Porta M, Costa MÂ, Hernández C, Cunha-Vaz J, Simó R; EUROCONDOR (2017) Functional and structural findings of neurodegeneration in early stages of diabetic retinopathy: cross-sectional analyses of baseline data of the EUROCONDOR project. Diabetes 66(9):2503-2510. DOI 10.2337/db16-1453. https://doi.org/10.2337/db16-1453
  11. Sohn EH, van Dijk HW, Jiao C, Kok PH, Jeong W, Demirkaya N, Garmager A, Wit F, Kucukevcilioglu M, van Velthoven ME, DeVries JH, Mullins RF, Kuehn MH, Schlingemann RO, Sonka M, Verbraak FD, Abràmoff MD (2016) Retinal neurodegeneration may precede microvascular changes characteristic of diabetic retinopathy in diabetes mellitus. Proceedings of the National Academy of Sciences USA 113(19):E2655-E2664. DOI 10.1073/pnas.1522014113. https://doi.org/10.1073/pnas.1522014113
  12. Lynch SK, Abràmoff MD (2017) Diabetic retinopathy is a neurodegenerative disorder. Vision Research 139:101-107. DOI 10.1016/j.visres.2017.03.003. https://doi.org/10.1016/j.visres.2017.03.003
  13. Jenkins KS, Steel JC, Layton CJ (2018) Systematic assessment of clinical methods to diagnose and monitor diabetic retinal neuropathy. Journal of Ophthalmology 2018:8479850. DOI 10.1155/2018/8479850. https://doi.org/10.1155/2018/8479850
  14. Grauslund J, Frydkjaer-Olsen U, Peto T, Fernández-Carneado J, Ponsati B, Hernández C, Cunha-Vaz J, Simó R; EUROCONDOR (2019) Topical treatment with brimonidine and somatostatin causes retinal vascular dilation in patients with early diabetic retinopathy. Investigative Ophthalmology & Visual Science 60(6):2257-2262. DOI 10.1167/iovs.18-26487. https://doi.org/10.1167/iovs.18-26487
  15. Simó R, Hernández C, Porta M, Bandello F, Grauslund J, Harding SP, Aldington SJ, Egan C, Frydkjaer-Olsen U, García-Arumí J, Gibson J, Lang GE, Lattanzio R, Massin P, Midena E, Ponsati B, Ribeiro L, Scanlon P, Lobo C, Costa MÂ, Cunha-Vaz J; EUROCONDOR (2019) Effects of topically administered neuroprotective drugs in early stages of diabetic retinopathy: results of the EUROCONDOR clinical trial. Diabetes 68(2):457-463. DOI 10.2337/db18-0682. https://doi.org/10.2337/db18-0682
  16. Ba-Ali S, Larsen M, Andersen HU, Lund-Andersen H (2022) Full-field and multifocal electroretinogram in non-diabetic controls and diabetics with and without retinopathy. Acta Ophthalmologica 100(8):e1719-e1728. DOI 10.1111/aos.15184. https://doi.org/10.1111/aos.15184
  17. McAnany JJ, Persidina OS, Park JC (2022) Clinical electroretinography in diabetic retinopathy: a review. Survey of Ophthalmology 67(3):712-722. DOI 10.1016/j.survophthal.2021.08.011. https://doi.org/10.1016/j.survophthal.2021.08.011
  18. Parravano M, Fragiotta S, Costanzo E, Picconi F, Giorno P, De Geronimo D, Giannini D, Varano M, Parisi V, Ziccardi L (2024) Metabolic, microvascular, and structural predictors of long-term functional changes evaluated by multifocal electroretinogram in type 1 diabetes. Biomedicines 12(11):2614. DOI 10.3390/biomedicines12112614. https://doi.org/10.3390/biomedicines12112614
  19. Arias-Alvarez M, Sopeña-Pinilla M, Fernandez-Espinosa G, Orduna-Hospital E, Rodriguez-Mena D, Pinilla I (2025) Early detection of retinal dysfunction in type 1 diabetes without retinopathy using multifocal electroretinography. Scientific Reports 15:6379. DOI 10.1038/s41598-025-89734-2. https://doi.org/10.1038/s41598-025-89734-2
  20. Wang CN, Xia YH, Huang H, Xu YX, Zhao DM, Dong JL, Ye HJ (2026) Neurodegeneration as an early driver and therapeutic target in diabetic retinopathy. International Journal of Ophthalmology 19(8):1618-1628. DOI 10.18240/ijo.2026.08.21. https://doi.org/10.18240/ijo.2026.08.21