Mohammad AskarPublic working copy · medical review pending

Diabetic retinopathy · 123

Diabetic retinopathy: stages, oedema and the emergencies that change management

Read the stage, decide whether the macula or the ischaemic drive is the problem, and treat the systemic disease alongside — with the limits of each study kept visible.

Public working copy · medical review pending. Source, translation and medical checks remain open. Read the limitations in the text and references. These educational materials do not replace individual medical assessment. Published: 2026-11-18T19:45:52+01:00. Internal project references lead to this notice. They are not publicly accessible primary sources; the source limitations remain stated in the text.

Stage NPDR using any one 4-2-1 criterion without new vessels. PDR treatment extends beyond high-risk disease. Anti-VEGF is first line for most visually impaired centre-involved DMO, while selected good-vision eyes may be monitored with a rescue plan. In NVG, injections buy time while pressure and ischaemia are addressed. Aspirin evidence is not a reason to change prescribed antithrombotic treatment yourself.

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.

01Assess the retina and the systemic risks together

Diabetic retinopathy (DR) is a leading cause of preventable blindness in working age and a marker of systemic microangiopathy. Retina, kidney and nerve share the same small-vessel injury, so the eye finding is also a message about the heart, the kidney and blood pressure. Three concepts organise the lecture: follow non-proliferative DR (NPDR) according to stage; control neovascularisation in proliferative DR (PDR); and treat diabetic macular oedema (DMO), an important cause of visual loss, to preserve function as well as improve anatomy. 1 2

Treatment goals are therefore threefold: preserve vision by controlling oedema and preventing macular haemorrhage; make new vessels regress so that vitreous haemorrhage and traction are avoided; and keep the systemic base stable with the diabetes team. The original claim that any injection or laser is “doomed to fail” if HbA1c and blood pressure are neglected is too strong: ocular treatment still helps, and systemic control lowers risk alongside it. HbA1c describes roughly the last three months. The “legacy effect” means that years of poor control leave a risk that a short period of improvement before surgery does not immediately erase. That is an argument for long-term control and for not delaying urgently needed eye treatment while waiting for a perfect HbA1c. 1 2

Aspirin and anticoagulants: in the ETDRS, 3711 patients with mild-to-severe NPDR or early PDR were randomised to aspirin 650 mg daily or placebo; aspirin neither prevented high-risk PDR nor increased vitreous haemorrhage. This is randomised evidence about aspirin, not a safety guarantee for every anticoagulant, and it is not a reason to stop or start any such drug yourself. 7

Screening schedules depend on age and diabetes type. ADA 2026 recommends the first comprehensive dilated examination within 5 years of adult type 1 diagnosis and at type 2 diagnosis. For youth with type 1 diabetes, start after 3–5 years of disease once age ≥11 or puberty has begun, whichever comes earlier. Pre-existing diabetes needs assessment before pregnancy and in the first trimester, with follow-up during pregnancy and up to 1 year postpartum according to disease; gestational diabetes alone is a different situation. Symptoms require assessment regardless of the screening date. 20 21

02Ischaemia, VEGF and what treatment can and cannot change

The chain is short and worth memorising: capillary occlusion causes tissue hypoxia; the hypoxic retina secretes vascular endothelial growth factor (VEGF); VEGF increases vessel permeability and drives new vessel growth. Pericyte loss is an early histological event in diabetic microangiopathy and destabilises the blood–retinal barrier. The original description of this as an irreversible “point of no return” is too absolute: pericyte loss precedes capillary occlusion, and barrier function and oedema can improve with treatment. 1 2

Anti-VEGF suppresses the permeability and growth signal while treatment is effective; panretinal photocoagulation (PRP) reduces the peripheral retinal drive. Anti-VEGF can be an alternative to PRP in selected PDR with reliable repeated treatment and follow-up, or an adjunct. PRP is more durable but does not eliminate recurrence or the need for review. In NVG, rapid regression after injection still needs treatment of the ischaemic cause and pressure. 2 5 12 20

03Classification: the 4-2-1 rule and the border to PDR

Grading follows the ETDRS photographic severity scale, which has 13 levels and was built from the power of haemorrhages, microaneurysms, IRMA and venous beading to predict progression. Clinical practice uses the simplified five-stage international scale: no retinopathy, mild, moderate and severe NPDR, and PDR. 3 4

Severe NPDR requires any one of the 4-2-1 criteria, without PDR signs: more than 20 intraretinal haemorrhages in each of 4 quadrants, definite venous beading in at least 2 quadrants, or prominent intraretinal microvascular abnormalities (IRMA) in at least 1 quadrant. They are alternatives, not a requirement to meet all three. Any neovascularisation of the disc (NVD) or elsewhere (NVE) makes the eye PDR. 3 4 22

The lecture's figure of “up to 50% conversion to PDR within one year” is a historical teaching figure for severe NPDR. It is useful for explaining why review intervals shorten, but it is not a universal rate for every stage, population or level of control, and this draft did not re-verify the exact percentage in the sources read. 1 3

IRMA are intraretinal irregular vascular loops; NVE extends through the internal limiting membrane (ILM) onto the retinal surface or into the vitreous. IRMA usually leak little on fluorescein angiography, but leakage and evolution towards neovascularisation can occur. Combine the vascular appearance, leakage and OCT location rather than treating any single sign as infallible. 1 3 9

04PDR: which eye is high-risk and when to laser

PDR means any neovascularisation, at the disc or elsewhere. Not every PDR is high-risk, and this difference decides urgency. The Diabetic Retinopathy Study identified four risk factors, and eyes with three or more — the “high-risk characteristics” — had a much higher two-year risk of severe visual loss. The classical high-risk features are NVD of about one-quarter to one-third disc area, any NVD with vitreous or preretinal haemorrhage, and NVE of at least half disc area with haemorrhage. The “one-quarter to one-third” range comes from that historical DRS context; it is a treatment threshold, not a description of all PDR, and a smaller NVD with haemorrhage is still high-risk. 6

ETDRS report 9 compared early scatter photocoagulation with deferral. Scatter was not recommended for mild or moderate NPDR, should be considered when retinopathy is more severe, and usually should not be delayed once high-risk PDR is present. The same report documented short-term adverse effects on visual acuity and visual field, greatest immediately after treatment and smaller with less extensive scatter. 5

PRP is offered for PDR beyond the high-risk subgroup; NICE recommends offering it at first diagnosis, with greater urgency for high-risk features or unreliable attendance. Anti-VEGF can be an alternative in selected PDR or an adjunct, but requires reliable follow-up, repeated dosing and attention to traction. The historical high-risk thresholds guide urgency, not a rule to withhold treatment from all other PDR. 5 12 20

05Diabetic macular oedema: treat the centre and the function, not a fixed number

DMO is an important cause of visual loss in diabetes. Centre involvement matters for reading vision; OCT describes retinal thickness, intraretinal and subretinal fluid and hyperreflective foci. Anatomy informs the decision but is not a substitute for visual function. The original rule “centre involvement plus vision below 0.7 (about 20/30)” describes a common situation, not a legal threshold. The decision combines centre involvement, duration, functional symptoms and the fellow eye; a patient with 1.0 vision and new metamorphopsia may be treated, but the symptom alone does not automatically require an injection. 2 8 9

Protocol V randomised 702 eyes with centre-involved DMO and vision 20/25 or better to initial aflibercept, laser or observation. At 2 years, loss of at least 5 letters occurred in 16%, 17% and 19%, respectively, without a statistically clear difference; this does not prove equivalence. Rescue aflibercept was needed in 25% of laser and 34% of observation eyes. These selected good-vision eyes had structured follow-up and rescue criteria; the findings do not justify unmonitored waiting or apply to every visually impaired eye. 8

Anti-VEGF is first line for most centre-involved DMO with visual impairment, not automatically every good-vision eye. Options include aflibercept, ranibizumab and the dual VEGF-A/Ang-2 inhibitor faricimab. YOSEMITE/RHINE included 1891 participants: protocol-defined schedules, fixed 8-week treatment or personalised faricimab intervals of 4–16 weeks supported non-inferiority at 1 year against 8-week aflibercept within a 4-letter ETDRS margin. This industry-funded trial does not establish superiority or validate every real-world treat-and-extend schedule. 11 20

Consider intravitreal steroids in selected persistent DMO or when anti-VEGF is unsuitable; pseudophakia alone is not an automatic indication. Dexamethasone implant (Ozurdex) and triamcinolone differ in formulation and local authorisation. Assess pressure rise, cataract, contraindications and the exact product monitoring requirements; no blanket licence or safety equivalence is implied. 10

Focal or grid laser keeps specific indications: non-centre oedema, localised leaking microaneurysms, and eyes where injections are not feasible. It is no longer first line for centre-involved oedema. OCT angiography adds information about the foveal avascular zone (FAZ). An enlarged FAZ suggests macular ischaemia and predicts a poorer response, but it is not a fixed ceiling on vision, and OCT-A is prone to projection, segmentation and signal artefacts. Treat the patient, not the FAZ measurement alone. 2 9

06Neovascular glaucoma and vitreous haemorrhage: two emergencies

Neovascular glaucoma (NVG) is an emergency. New vessels on the iris (NVI) and in the angle (NVA) grow into a fibrovascular membrane that contracts and closes the angle. Both components matter: the pressure and the ischaemia. Initial management lowers intraocular pressure with topical treatment and systemic acetazolamide where appropriate, gives anti-VEGF to regress the rubeosis, and adds topical steroids and cycloplegia for inflammation. Anti-VEGF buys time; it does not replace PRP, which treats the ischaemic drive. A network meta-analysis of 17 randomised trials (1311 eyes) compared anti-VEGF regimens in NVG, all with PRP, and supports anti-VEGF as part of combined treatment rather than as monotherapy. 14 15 16

If surgery is needed, glaucoma drainage implants (Ahmed, Baerveldt) are often preferred in ischaemic, rubeotic eyes, but this is not a universal rule; trabeculectomy still has a role in selected eyes. The original argument that implants are better because they need no iridectomy names one practical point, not the main reason: the decisive issues are conjunctival condition, previous surgery, target pressure and the chance of haemorrhage or fibrin. 14 15

When vitreous haemorrhage (VH) obscures the fundus, assess urgently and use B-scan to look for detachment or associated pathology. A negative scan cannot reliably exclude every small retinal tear: repeat assessment and retinal review may be necessary. A tear or detachment needs prompt specialist planning, while the urgency and choice of intervention depend on cause, macular threat and progression. Ghost cell glaucoma follows chronic vitreous haemorrhage: degenerated, rigid red cells block the trabecular meshwork. It is a reason to consider timely pars plana vitrectomy (PPV) in selected eyes with persistent haemorrhage, but it is not an automatic indication for surgery in every VH. 17

07Surgery and special situations

Fibrovascular membranes cause tractional retinal detachment (TRD). Macula-threatening or progressive detachment may require PPV with segmentation and delamination; stable extramacular traction is assessed individually. Gas or silicone oil tamponade depends on breaks, configuration and operative findings, and is not mandatory for every TRD. Preoperative anti-VEGF reduces vascularity but can make membranes contract abruptly — the “crunch” phenomenon — and worsen a TRD. The original instruction to operate “mandatorily within three to seven days” is a traditional precaution rather than a validated rule; the risk depends on how fibrotic the membranes are, and timing must be coordinated with the actual surgical list, sometimes earlier. 1 2

Diabetic papillopathy is disc swelling in a person with diabetes, usually with mild visual loss, and it is a diagnosis of exclusion. The differential includes raised intracranial pressure, compressive and infiltrative optic neuropathy, ischaemic optic neuropathy, optic neuritis, disc drusen, vitreous traction and sinus thrombosis. Not every swollen disc needs MRI and MRV, but red flags do: severe or progressive loss, other neurological signs, unilateral swelling with a pale disc, or an atypical course. Do not attribute disc swelling to diabetes before that assessment. A retrospective single-centre series included 37 eyes of 22 patients: 15/22 patients (68.2%) were bilateral, while 16/37 eyes (43.2%) had centre-involved DMO. Patient and eye denominators differ; selection and small sample size limit generalisation. 19

Pregnancy can accelerate DR: assess early, monitor according to findings and treat sight-threatening disease; PRP is established. Intravitreal anti-VEGF can enter the systemic circulation. Human pregnancy data are limited, largely observational and confounded; they establish neither fetal safety nor a definite teratogenic effect. Decide with the obstetric team using the exact product information and maternal and fetal risks. Cataract surgery can worsen DMO. Optimising the macula before surgery is reasonable and often helpful, but preoperative injection is not mandatory for every eye with treated DMO; the decision depends on activity, duration and the planned surgery. A silicone intraocular lens matters mainly if later vitrectomy with silicone oil is likely, because it complicates the view and encourages oil adherence; it is not a general contraindication. 12 13 18

08Quick review: six rows and their limits

Use the six-row table to connect finding, meaning and next step. Any one 4-2-1 criterion without new vessels can define severe NPDR; PDR treatment is not restricted to high-risk disease; and anti-VEGF in NVG is a temporary component of treatment. Read the red-flag column as a prompt, not a complete rule. The closing advice of the original lecture still holds after correction: see the patient as a whole, ask about HbA1c and blood pressure, do not repeat the aspirin myth, and be able to explain why the surgical choice in NVG is individual. 1 3 4 6 7 14

Quick review: six rows and their limits
Finding / stageLeading sign and testTreatment principleExam red flag and limit
Severe NPDR (4-2-1, no new vessels)4-2-1 on fundus examinationClose review; consider PRP in selected severe NPDR, not only after PDR developsThe ~50% one-year figure is historical orientation; any new vessel makes it PDR
PDRNVD/NVE with leakage on FA; OCT for the maculaPRP for PDR; greater urgency at high risk; anti-VEGF an alternative or adjunct in selected eyesImpending vitreous haemorrhage; not every PDR is high-risk
Centre-involved DMOVisual loss, metamorphopsia; OCT confirms centre involvementAnti-VEGF for most visually impaired eyes; selected good-vision observation, steroid or focal laserAn enlarged FAZ limits prognosis but is not a fixed ceiling
Neovascular glaucomaHigh IOP, rubeosis iridis, angle new vesselsLower IOP, anti-VEGF, PRP; glaucoma surgery individualisedAnti-VEGF is temporary; the ischaemic drive still needs PRP
Vitreous haemorrhage / TRDObscured fundus: B-scan; a negative scan does not exclude every small tearRetinal assessment; PPV according to cause, progression and macular threat“Crunch” after anti-VEGF: plan the date individually
Diabetic papillopathy (diagnosis of exclusion)Disc oedema; assess for other causesObserve only after assessment; treat concurrent DMODisc swelling is not automatically diabetic; red flags need imaging
What are the main learning points?

Stage NPDR using any one 4-2-1 criterion without new vessels. PDR treatment extends beyond high-risk disease. Anti-VEGF is first line for most visually impaired centre-involved DMO, while selected good-vision eyes may be monitored with a rescue plan. In NVG, injections buy time while pressure and ischaemia are addressed. Aspirin evidence is not a reason to change prescribed antithrombotic treatment yourself.

When is assessment urgent?

Sudden visual loss, a painful red eye, pressure rise or rubeosis, dense vitreous haemorrhage, suspected detachment, marked inflammation or infection, and neurological warning signs require urgent assessment. Do not attribute an acute problem to diabetes without excluding other causes.

References and access status

  1. 1
    Immutable original export; diabetic retinopathy teaching export

    Complete export read; original attribution and any recording/transcript or appointment not independently verified.

  2. 2
    German S3 guideline on retinal complications of diabetes 2025 (AWMF 045-024)

    Historical selected access reused from earlier review ledgers and editorially compared; no fresh complete retrieval or full-source appraisal claimed. Exact prior scope is recorded in evidence.md.

  3. 3
  4. 4
  5. 5
    Early Treatment Diabetic Retinopathy Study Research Group. Early photocoagulation for diabetic retinopathy. ETDRS report number 9. Ophthalmology 1991;98(5 Suppl):766-785. PMID 2062512

    Complete primary NCBI abstract independently read 15 November 2026; full report, protocol, tables and supplements not appraised.

  6. 6
  7. 7
    Early Treatment Diabetic Retinopathy Study Research Group. Effects of aspirin treatment on diabetic retinopathy. ETDRS report number 8. Ophthalmology 1991;98(5 Suppl):757-765. PMID 2062511

    Complete primary NCBI abstract independently read 15 November 2026; full report, protocol, tables and supplements not appraised.

  8. 8
  9. 9
    OCT biomarkers systematic review 2025 (diabetic macular oedema). PMID 40803536

    Historical selected access reused from earlier review ledgers and editorially compared; no fresh complete retrieval or full-source appraisal claimed. Exact prior scope is recorded in evidence.md.

  10. 10
    EMA Ozurdex EU product information (dexamethasone intravitreal implant)

    Historical selected access reused from earlier review ledgers and editorially compared; no fresh complete retrieval or full-source appraisal claimed. Exact prior scope is recorded in evidence.md.

  11. 11
    YOSEMITE/RHINE faricimab phase 3 one-year report. PMID 35085503

    Complete primary NCBI abstract independently read 15 November 2026; full report, protocol, tables and supplements not appraised.

  12. 12
    NICE NG242: diabetic retinopathy recommendations (2024)

    Selected indexed NICE NG242 recommendations 1.5.1–1.5.5 independently read 15 November 2026; direct page access returned 403. Earlier DMO access is historical reuse; not a complete guideline appraisal.

  13. 13
    ESCRS recommendations for cataract surgery (executive web text)

    Historical selected access reused from earlier review ledgers and editorially compared; no fresh complete retrieval or full-source appraisal claimed. Exact prior scope is recorded in evidence.md.

  14. 14
    European Glaucoma Society. Terminology and Guidelines for Glaucoma, 6th edition (2025)

    Historical selected access reused from earlier review ledgers and editorially compared; no fresh complete retrieval or full-source appraisal claimed. Exact prior scope is recorded in evidence.md.

  15. 15
    RCOphth. Retinal vein occlusion guideline, executive summary (January 2022), neovascular/established NVG recommendations

    Historical selected access reused from earlier review ledgers and editorially compared; no fresh complete retrieval or full-source appraisal claimed. Exact prior scope is recorded in evidence.md.

  16. 16
  17. 17
    Shaikh N, Srishti R, Khanum A, et al. Vitreous hemorrhage — causes, diagnosis, and management. Indian J Ophthalmol 2023;71:28-38. PMID 36588205

    Complete primary NCBI abstract independently read 15 November 2026; full report, protocol, tables and supplements not appraised.

  18. 18
  19. 19
  20. 20
    ADA Standards of Care 2026, section 12: retinopathy

    Selected official indexed recommendation/definition read 15 November 2026; not the entire document or evidence reviews.

  21. 21
    ADA Standards of Care 2026, section 14: youth screening

    Selected official indexed recommendation/definition read 15 November 2026; not the entire document or evidence reviews.

  22. 22
    AAO Diabetic Retinopathy Preferred Practice Pattern 2025: international severe NPDR definition

    Selected official indexed recommendation/definition read 15 November 2026; not the entire document or evidence reviews.

Original lecture link