OCT biomarkers inform prognosis; they do not alone prove that steroids work better. Separate raised pressure from glaucoma, and monitor both.
Terminology and draft status
Terminology review is not complete. Selected actual WHO UMD queries and their limits are recorded in terminology.md; they do not approve the whole translation.
01Recognise burden without turning a survey into an individual risk
DME can occur alongside diabetic retinopathy and threaten working-age vision; it is not the universal leading cause in every region. The original gives 66 million European adults with diabetes, 72 million by 2050 and one third undiagnosed. These historical estimates lack a verified population and source here. 1
The original working-age range 20–64 years, mean DME diagnosis age 52, over 70% employed, at least 20 medical visits yearly and 50% fearing sight loss most are unverified cohort statements. Cardiac, renal and diabetic-foot disease, travel and carers can add burden; missed visits need practical support, not blame. 1
The lecture says eye checks at least every 2 years and only half examined after 2 years of diabetes. Neither is a universal schedule or current coverage estimate: screening depends on diabetes type, previous retinal findings, risk and national guidance; new symptoms require earlier review. 1
02Connect vascular leakage and cellular dysfunction with limits
Hyperglycaemia, microvascular damage, hypoxia, VEGF and inflammation interact with the blood–retinal barrier. The original “no inflammation without hypoxia and vice versa” is an oversimplified model. It does not establish a compulsory sequence or prove an inflammatory phenotype in one eye. 1
The original animal comparison is VEGF threefold, normalising in 1 day, versus inflammatory mediators 120-fold for 14 days. The experiment, assays and denominators remain unverified. These numbers cannot rank human disease mechanisms, choose a drug or define a clinical interval. 1
Müller cells support retinal architecture, light transmission and ionic/water homeostasis through aquaporin and potassium channels. The original links channel loss to cellular swelling and diffuse thickening, while endothelial tight-junction injury permits extracellular leakage and cystic spaces. OCT morphology does not uniquely identify intracellular versus extracellular origin. 1
Experimental claims of persistent endothelial injury after VEGF exposure, minimal anti-VEGF effect on Müller degeneration and steroid-driven regeneration/channel recovery remain mechanistic hypotheses here. They do not establish irreversible human treatment failure or prove that a steroid restores vision better. 1
03Treat the visual problem and reassess an incomplete response
Optimise metabolic and vascular risk management with the treating team. The German S3 guideline supports anti-VEGF for centre-involving DME with relevant visual impairment when morphology suggests functional benefit. With preserved good vision, controlled observation is a different pathway; selected noncentral focal laser differs from panretinal laser for proliferative disease. 2
Protocol V randomised 702 adults with centre-involving DME and vision 20/25 or better to aflibercept, laser or observation. At 2 years at least 5-letter loss occurred in 16%/17%/19%; observation included rescue aflibercept for prespecified visual worsening. Aflibercept versus observation risk difference was −3% (95% CI −11% to 4%). This supports selected observation with reliable review, not equivalence in eyes with worse vision. 3
Original response claims are 30–65% suboptimal or delayed, about 40% without gain in year 1, real-world +4.7 letters versus trial +11–12, and US 52% stopping after roughly 6 months. Their cohorts, definitions and exposure are unverified and the comparisons are not randomised. Review diagnosis, ischaemia, traction, structural damage, injection delivery and access before attributing failure solely to inflammation. 1
Vitrectomy may address vitreomacular traction or other surgical pathology. It is not routine treatment for every nontractional DME. The original treatment menu of observation, intravitreal anti-VEGF/steroids, focal or panretinal laser and PPV must be matched to its own indication rather than treated as interchangeable options. 1
04Use OCT associations without inferring a treatment interaction
Describe cysts, subretinal fluid (SRF), hyperreflective foci (HRF), disorganisation of retinal inner layers (DRIL) and thickness in context. HRF can have several substrates; OCT does not directly prove activated microglia or the dominant inflammatory pathway. The original five-column plus/minus matrix is retained below as an unvalidated teaching model. 1
A systematic review included 27 studies, 28 reports and 75 biomarkers. No biomarker associated with improved vision or change in vision had at least moderate certainty. HRF and DRIL were among features associated with worse acuity; confounding and inconsistent definitions remained. Prognostic association does not prove that steroid treatment outperforms anti-VEGF in that subgroup. 4
Original cyst diameter >250 µm and central thickness >400 µm are not universal switch thresholds. Multiple HRF, large cysts or persistent SRF can prompt a careful reassessment, but their presence alone does not validate the original early-switch recommendation. Discuss residual vision potential, treatment exposure and individual safety. 1
| Biomarker | Original inflammation model | Original chronicity model | Original poor-prognosis model | Interpretation and limits |
|---|---|---|---|---|
| Large cysts >250 µm | −/+ with hyperreflective material | + | −/+ | Original steroid-benefit inference unvalidated; assess structure and course. |
| SRF | + | − | −/+ | Can accompany DME; not a specific inflammatory assay or steroid selector. |
| HRF | + | + | + | Multiple substrates; association with poorer vision does not prove early steroid switching benefits. |
| DRIL | − | + | + | Inner-layer disorganisation has prognostic associations, not a validated switch rule. |
| Central thickness >400 µm | − | − | + | Original cutoff unvalidated; absence of thinning needs review without assuming no visual potential. |
05Choose an implant by indication and safety, not by one OCT sign
The EURETINA 2017 abstract supports anti-VEGF first-line and a role for steroids in chronically persistent DME; it does not establish the original automatic switch after 3–6 injections. Dexamethasone and fluocinolone acetonide (FAc) implants differ. Pseudophakia, planned cataract surgery, cardiovascular risk or difficult monthly attendance require individual assessment rather than automatic steroid selection. 5
The accessed EMA Ozurdex indication concerns adults with DME-related visual impairment who are pseudophakic or insufficiently responsive/unsuitable for non-steroid therapy. Check infection, advanced uncontrolled glaucoma and capsule-related contraindications. “No systemic side effects” is not a defensible absolute; cataract, pressure and procedural infection risks remain, and fewer injections do not eliminate monitoring. 6
The original case of a woman stable without treatment for 8 years after FAc lacks verified records and follow-up definitions. It illustrates an individual report, not a duration guarantee. No conclusion about universal implant effectiveness, absence of harm or a stopping rule follows. 1
06Distinguish ocular hypertension from damage and act by risk
Raised IOP is a risk factor; glaucoma means characteristic optic nerve damage with functional consequences, not an exclusively age-related label. OHTS randomised 1636 adults aged 40–80 with pressure 24–32 mmHg in one eye and 21–32 in the other: 5-year POAG probability was 4.4% treated versus 9.5% observed (HR 0.40,95% CI 0.27–0.59). The original approximately 91% remaining glaucoma-free is an observation-group complement, not evidence that steroid pressure rises are harmless. 7
Original FAc pressure elevation in about one third and usual control by one drop lack a universal denominator. PALADIN 24-month interim data came from 95 previously steroid-challenged patients/115 eyes; 4 pressure-lowering operations and 2 trabeculoplasties occurred after FAc. Selection and absence of randomisation limit extrapolation; do not conflate this with the separate 36-month cohort or promise that surgery is rarely needed in every eye. 8
Assess previous pressure responses, glaucoma/suspicion, family history, myopia, age and type 1 diabetes. The original observe-to-25 mmHg, start only above 25 or after damage, and refer above 25 despite 2 medicines is too rigid. Baseline, rise from baseline, optic nerve, field, trajectory and symptoms can justify earlier action; do not wait for nerve damage. 1
European consensus separates baseline susceptibility and evolving pressure risk: baseline <22 mmHg without prior events differs from ≥22 or prior events; >25 or a rise ≥10 mmHg signals high risk. It is an expert monitoring framework, not proof of one safe cutoff for every eye. Record the next pressure/nerve/field review and glaucoma referral plan individually. 9
What are the main learning points?
OCT biomarkers inform prognosis; they do not alone prove that steroids work better. Separate raised pressure from glaucoma, and monitor both.
When is assessment urgent?
Sudden visual deterioration, marked pain, redness, new floaters or a field defect after an injection require urgent ophthalmic assessment.
References and access status
- 1Immutable original export; DME teaching export, all original 85 items identical across 102–106; source attribution untraceable
Complete export read; original attribution and any recording/transcript or appointment not independently verified.
- 2German S3 diabetic retinal complications 2025
Cached selected DMEsection 2.2.6 printed 16–19 reread 14 November; not whole guideline.
- 3Protocol V randomised trial
Complete cached NCBI abstract reread 14 November 2026; full report/protocol/supplement unread.
- 4OCT biomarkers systematic review 2025
Complete cached NCBI abstract reread 14 November 2026; full report/protocol/supplement unread.
- 5EURETINA 2017 DME guideline abstract
Complete cached NCBI abstract reread 14 November 2026; full report/protocol/supplement unread.
- 6EMA Ozurdex EU product information
Cached printed 2 section 4.1–4.2 reread 14 November;4.3–4.4 safety scope from 07 November reused. No full/current FAc label appraisal.
- 7OHTS 2002 randomised ocular hypertension trial
Complete cached NCBI abstract reread 14 November 2026; full report/protocol/supplement unread.
- 8PALADIN 24 month observational interim
Complete cached NCBI abstract reread 14 November 2026; full report/protocol/supplement unread.
- 9European expert IOP monitoring consensus 2022
Complete cached NCBI abstract reread 14 November 2026; full report/protocol/supplement unread.
For patients
Diabetes can cause swelling at the centre of the retina. Treatment depends on vision, the examination and the risks of each medicine; regular follow-up matters even with a long-acting implant.
Examination and selected tests guide individual treatment and follow-up; this educational text cannot replace your own assessment.
Sudden visual deterioration, marked pain, redness, new floaters or a field defect after an injection require urgent ophthalmic assessment.
Sudden visual deterioration, marked pain, redness, new floaters or a field defect after an injection require urgent ophthalmic assessment.