Mohammad AskarPublic working copy · medical review pending

Posterior uveitis and endophthalmitis · 087

Posterior uveitis: retinal vasculitis and endophthalmitis

Retinal vessel inflammation, ischaemia, masquerade disease and intraocular infection need different investigations. Use the phenotype to guide the next question, rather than treating it as proof of cause.

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.

Retinal vasculitis has many associations, so the work-up hunts for infection, systemic disease and masquerade tumours; FA shows leakage and true non-perfusion, while OCT and OCTA cannot replace it. Endophthalmitis is an intraocular infection, not simply panuveitis: delayed postoperative and endogenous forms need cultures, organism-appropriate antimicrobials and often surgery, and a negative culture does not prove sterility.

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.

01Definitions: what retinal vasculitis is and is not

Retinal vasculitis involves retinal arteries, veins and/or capillaries. Primary or idiopathic disease has no identified associated cause after appropriate assessment; secondary disease accompanies an infection, systemic inflammation or a defined ocular syndrome, sometimes panuveitis. The original calls isolated disease rare and secondary disease commonest; their relative frequency depends on the population and definitions. 1

Phlebitis predominantly affects veins. Sheathing or cuffing is a whitish coat around vessels; it does not by itself prove active inflammation. Masquerade disorders imitate primary uveitis, especially tumours; infarction and infection may enter the differential and can themselves provoke inflammation. Global ischaemia means extensive retinal underperfusion. Classify the vessel involved and central versus peripheral location. 1

Two shortcuts are wrong. Not every uveitis is autoimmune, and not every retinal vasculitis is a systemic vasculitis. The causes include infection, systemic inflammatory disease, primary ocular syndromes, drugs and neoplasia, so the diagnosis is a differential, not a reflex label. A masquerade such as intraocular lymphoma stays on the list, especially in older patients or when inflammation resists treatment. 1 2

02Causes and classification: a structured differential

The causes are broad and include almost any form of systemic inflammation or infection, so a structured approach matters. Among infectious causes, bacterial examples are tuberculosis (TB), syphilis (lues) and borreliosis; viral examples include herpesviruses and COVID-19; fungal infection, parasitic infection such as toxoplasmosis and rickettsia also belong here. The original adds that vaccinations can trigger uveitis. Such reports are often temporal associations, and timing alone does not prove causation. 1 2

Among non-infectious causes the original lists systemic autoimmune disease, namely sarcoidosis, Behçet disease, systemic lupus erythematosus (SLE) and granulomatosis with polyangiitis (GPA, formerly Wegener). Primary ocular syndromes such as birdshot retinochoroidopathy, drug-induced disease and neoplastic masquerade, especially intraocular lymphoma, complete the list. 1

Classification runs along two axes. By vessel type, disease may be arterial, venous (phlebitis) or capillary. By location it may be central or peripheral. These labels organise the differential, but vessel type and location predict a cause only loosely: an arterial pattern is not proof of one disease and a venous pattern is not proof of another. 1 2

The practical consequence is that the big mimickers, TB, sarcoidosis and syphilis, must be considered in any uveitis because they can imitate almost any picture. Infection must be excluded before immunosuppression escalates, and a masquerade must be excluded before chronic steroid treatment is accepted. 1 2

03Epidemiology and prognosis: original figures with their limits

The original reports an incidence of about 2-3 cases per 100000 persons per year, an association with other uveitis manifestations in 15-55% of cases, and a tendency to affect younger patients. These are original statements: the source population, case definition and denominators were not verified for this draft, so they must not be presented as current global rates. A systematic review that pooled 97 studies and 7619 patients underlines how heterogeneous the reported populations and definitions are, so no single prevalence or incidence figure is universally validated. 1 2

Visual loss is a significant risk. About 30% of patients are said to reach acuity below 20/200 (6/60). Non-ischaemic vasculitis is described as having a generally good prognosis, with an 8-year study reporting that most patients retained acuity of 1.0 or better, whereas ischaemic vasculitis has a poor prognosis, with about 70% of patients falling below 0.1. The cohort, its selection and its follow-up are not detailed here; the figures are kept as original values, not as a guarantee for an individual eye. 1 2

The original also claims that history alone can reach the correct diagnosis in up to 90% of cases, often without extensive laboratory testing. That is a statement about the value of a careful history, not a validated diagnostic yield. It does not mean that 90% of real patients are diagnosed by history alone, and it does not justify omitting targeted tests when the history points to infection or systemic disease. 1

04History and examination: asking the questions that change management

The history is the most important diagnostic tool. Systemically, ask about oral or genital ulceration (pointing to Behçet disease), rash, unexplained weight loss and night sweats (TB, lymphoma), joint pain or arthritis, and lymphadenopathy (sarcoidosis, TB). Ask about neurological symptoms such as tinnitus, headache, loss of consciousness and dizziness, and about changes in bowel habit. 1

Cover infection risk directly: recent viral or other infection including COVID-19, travel, sexual contacts (syphilis remains important and can present aggressively and atypically), and contact with animals such as cats and dogs. Review new medications or treatment changes. These questions are cheap, and several of them redirect the whole work-up. 1 2

On fundoscopy, look for vascular sheathing and cuffing as whitish lines following the vessels, and for haemorrhages; intraretinal blood suggests ischaemia. Ischaemic signs include cotton-wool spots, which may be extensive and suggest systemic vasculitis, and retinal infarcts. Disc swelling is non-specific and occurs in many inflammations. It must be distinguished from the hazy view of the disc produced by marked vitritis; that obscured view is not by definition a pseudopapilloedema. Slit-lamp examination looks for cells in the anterior chamber and vitreous (vitritis). 1

History and examination set the differential, but they do not replace targeted investigation. A negative review of systems lowers probability; it does not exclude infection or masquerade disease when the ocular phenotype is suspicious. That is why the lecture moves from history to angiography and, in selected cases, to tissue. 1 2

05Fluorescein angiography: leakage, non-perfusion and their limits

Fluorescein angiography (FA) is an indispensable tool for the uveitis specialist. It assesses the extent of disease, identifies ischaemic zones and quantifies inflammatory activity. Leakage appears as hyperfluorescence along the vessels that increases in intensity and extent in late phases, indicating damage to the blood-retina barrier. 1

Occlusion is seen as absent filling of vessel segments. Non-perfusion or ischaemia appears as dark, so-called black areas, the dropout zones. A dark area is not automatically non-perfusion: haemorrhage or another blocking material can hide underlying fluorescence, so blocking must be distinguished from true non-perfusion before ischaemia is diagnosed. 1

FA also shows macular oedema as a cystoid, petaloid pattern in the late phase and disc leakage as hyperfluorescence of the optic disc. The pattern may suggest a cause. Arterial leakage is associated with viral and protozoal infection and with systemic vasculitides such as SLE and GPA, and it tends to produce occlusion and severe ischaemia. Capillary involvement produces exudation and haemorrhage and is typical of syphilis and Whipple disease. Venous leakage, the commonest form, may be focal (sarcoidosis, multiple sclerosis) or diffuse (idiopathic retinal vasculitis). These associations are patterns, not proof of a specific cause. 1 2

OCT assesses oedema and OCTA maps flow-related signal, but neither demonstrates dye leakage. The original description of replacing FA for macular assessment applies only to selected questions, not all vasculitis. Use complementary imaging according to the question; normal structural OCT does not exclude active vascular leakage. 1

06Complications: macular oedema, neovascularisation and ischaemia

Macular oedema is an important cause of visual loss in posterior uveitis; macular ischaemia can add irreversible damage even when thickness improves. OCT evaluates fluid and structure, OCTA can help assess perfusion, and FA remains useful for leakage and vascular activity. These are complementary questions rather than an automatic replacement of one test by another. 1

Retinal ischaemia can stimulate vascular endothelial growth factor (VEGF) and neovascularisation, as in proliferative diabetic retinopathy. Selected diffuse disease with neovascularisation may need panretinal photocoagulation (PRP), while focal disease may need sectoral treatment. The extent of non-perfusion alone is not a universal laser indication; assess activity, new vessels and visual risk. 1

Laser is a tool, not an automatic response to every non-perfused area. Many eyes with peripheral non-perfusion are monitored, with control of the underlying disease and of inflammation, and laser is reserved for eyes that develop neovascularisation or are at high risk of it. The decision depends on the extent and location of ischaemia, the presence of new vessels and whether the underlying process is controlled. Prognosis tracks ischaemia: the more ischaemic the retina, the worse the visual outlook. 1 2

07Disease pictures and teaching cases: sarcoidosis, Behçet, toxoplasmosis, lymphoma

The big mimickers, TB, sarcoidosis and syphilis, must be considered in every uveitis because they can imitate any picture. Case 1, sarcoidosis: ocular findings include yellow deposits on the vessels (wax droppings) and retinal infiltrates; systemically there may be a characteristic rash and bilateral hilar lymphadenopathy on chest imaging. 1 2

Case 2, Behçet disease: ocular findings include marked venous occlusions (phlebitis); systemically there are oral ulcers and a positive pathergy phenomenon, a skin reaction to a needle prick. Pathergy is optional in the ICBD criteria and contributes one point; it has not been removed, but a negative test does not exclude Behçet disease. The EULAR 2025 update abstract and the earlier systematic review support early systemic immune control in selected severe Behçet disease with retinal vasculitis. They do not establish a universal steroid-free 48-hour rule, and a negative interferon-gamma release assay (IGRA) does not exclude all infection. 1 4

Case 3, toxoplasmosis: focal necrotising retinitis with intense vitritis produces the classic headlight-in-the-fog appearance. Importantly, toxoplasmosis can also present primarily as vasculitis without an obvious retinitis lesion, so a quiet-looking retina does not exclude it. Intraocular lymphoma is an important masquerade, especially in older patients or when inflammation is resistant to treatment; it often presents as chronic vitritis or retinal infiltrates, and diagnosis requires vitreous biopsy. 1 2

These pictures are suggestive, not diagnostic. The phenotype narrows the differential, the history and tests weigh the alternatives, and tissue is reserved for cases where masquerade disease or an unresponsive process must be resolved. That sequence, pattern then probability then tissue when needed, prevents both shotgun testing and premature immunosuppression. 1 2

08Vitreous biopsy: what the original technique can and cannot tell us

For an adequate specimen, for example when lymphoma is suspected, simple needle aspiration is often insufficient. The original recommends a vitrectomy-assisted biopsy. A single trocar is placed and, with the vitrectome cutting and aspirating, a larger volume of undiluted vitreous (1.5-2 ml) is collected. 1

The original then describes light pressure on the globe during cutting, followed by restoring intraocular pressure with balanced salt solution (BSS) or air, and sending the specimen to a specialised laboratory experienced in cytopathological processing. This is an unvalidated original technical description, not a universal surgical protocol: the volume, the pressure and the handling are centre-specific, and light pressure on the globe must not be read as advice to compress an eye unsafely. In practice a specialist uses a sterile technique with a stable, controlled intraocular pressure. 1

Specimen handling determines whether the result is usable. Cytology, immunocytochemistry, flow cytometry and PCR each have requirements, and flow cytometry in particular needs fresh, viable cells. Coordinate with the laboratory before the procedure, label the specimen clearly and transport it promptly. A non-diagnostic sample is a real outcome, and it may require a repeat procedure rather than a change in diagnosis. 1 2

09Endophthalmitis: delayed postoperative and endogenous infection

Endophthalmitis is a severe intraocular infection, not simply inflammation of all layers of the uvea. The term describes infection of the intraocular contents, and panuveitis is a descriptive label for the inflammation it causes rather than the definition. Sterile inflammation, and inflammation with a negative culture, are not the same as a sterile eye: cultures can be negative in genuine infection. 1 3

It is classified as exogenous or endogenous. Exogenous acute postoperative endophthalmitis appears within days of an intervention; chronic postoperative endophthalmitis appears weeks to months later. Endogenous endophthalmitis arises by haematogenous spread from an infection focus in the body. A sterile intraocular inflammation without a demonstrable organism is described separately. 1

Chronic postoperative endophthalmitis begins insidiously weeks to months after surgery, usually cataract surgery, with a recurrent, low-grade (indolent), often granulomatous uveitis. A typical sign is a white plaque in the capsular bag behind the intraocular lens (IOL). It is often misdiagnosed and treated as sterile postoperative uveitis. Low-virulence organisms enclosed in the capsular bag drive it; the commonest are Cutibacterium acnes (formerly Propionibacterium acnes), Staphylococcus epidermidis and fungi such as Candida. 1 3

A high index of suspicion is decisive when postoperative uveitis does not respond to steroids or keeps recurring. A YAG capsulotomy can release the enclosed organisms into the vitreous and markedly worsen the inflammation, so the diagnosis should be considered before laser. Diagnosis uses anterior chamber or vitreous sampling, with culture and PCR; capsular material may also be sampled. 1 3

Selected chronic postoperative infections need vitrectomy, organism-appropriate antimicrobials and capsular management, sometimes IOL/capsular removal when persistent. Vancomycin capsular wash is not a universal solution and does not treat fungal infection. IDSA2016 ocular recommendations82–91 support joint ophthalmic/infectious-disease care for Candida, with systemic and selected intravitreal antifungals and vitrectomy; much of that evidence is low quality. The original good ocular/excellent systemic prognosis describes uncomplicated exogenous cases, not a guarantee. 1 3

Endogenous endophthalmitis is a severe intraocular infection arising from haematogenous spread of bacteria or fungi from a systemic focus. Risk factors in the original include immunosuppression (diabetes, malignancy, HIV, chemotherapy, systemic steroids), major surgery such as abdominal surgery, endoscopy and dental procedures, and intravenous drug use. The patient is often systemically very unwell and hospitalised with fever and sepsis, but not always, and endogenous endophthalmitis should not be dismissed because the patient looks stable. Gram-positive organisms such as Streptococcus spp. and Gram-negative organisms such as Neisseria, Haemophilus influenzae and E. coli are described; fungi also occur. These organisms and exposures are context, not a frequency table. 1 3

Clinically there is severe acute visual loss, often to light perception, with hypopyon, severe vitritis and often retinal or subretinal abscesses. Blood cultures and a vitreous or anterior chamber sample should be obtained promptly when feasible, but sampling must not delay empirical treatment. Treatment is immediate intravitreal broad-spectrum antibiotics for suspected bacterial infection, with selected high-dose systemic antibiotics and treatment of the systemic source; fungal disease needs appropriate antifungals and selected vitrectomy. The prognosis is variable and often poor: retinal detachment, chronic hypotony and phthisis bulbi are common ocular complications, and mortality is driven by the underlying sepsis rather than being inevitable in every patient. 1 3

A retrospective six-eye C. acnes series reported favourable final vision but frequent need for additional treatment; its literature review also compared different capsular strategies. Selection, small samples and nonrandom treatment prevent a causal ranking or a guarantee that IOL removal cures every infection. 5

The historical EVS randomized acute postoperative bacterial cases and found vitrectomy benefit in the light-perception-only subgroup; its systemic-antibiotic comparison did not improve final visual outcome. Those results do not establish management of delayed capsular infection, endogenous sepsis or fungal disease, and cannot justify withholding systemic treatment of a systemic source. 6

Endophthalmitis: delayed postoperative and endogenous infection
FeatureRetinal vasculitisChronic postoperative endophthalmitisEndogenous endophthalmitis
DefinitionInflammation of the retinal vessels.Low-grade, delayed infection after eye surgery.Haematogenous spread of a systemic infection (bacteria or fungi).
Patient statusOften outpatient; systemic symptoms possible.Generally in good general condition.Often systemically very unwell and hospitalised with sepsis, but not always.
OnsetVariable, often chronic.Weeks to months postoperatively.Acute, though indolent fungal courses occur.
Clinical signsVascular sheathing, haemorrhages, ischaemia, vitritis.Recurrent uveitis, white plaque in the capsular bag.Hypopyon, severe vitritis, retinal abscesses.
DiagnosticsFA, systemic work-up, laboratory tests.High suspicion; anterior chamber or vitreous tap.Blood cultures, vitreous tap.
Common causesSystemic disease (sarcoidosis, Behçet), infections (TB, syphilis).Cutibacterium acnes (formerly P. acnes), S. epidermidis, fungi.Streptococcus spp., E. coli, other sepsis organisms; also fungi.
Treatment approachTreat the cause; selected anti-inflammatory treatment after infection assessment, and laser when indicated.Selected vitrectomy/capsular management, sometimes removal, plus organism-appropriate antimicrobial therapy.Immediate systemic and intravitreal antimicrobials; source control.
Prognosis (eye)Variable, depends on ischaemia.Often good with correct treatment.Often poor, but variable; not inevitable blindness.
Prognosis (systemic)Depends on the underlying disease.Usually good if no systemic infection; original “excellent” is not a guarantee.Poor, high mortality, though not inevitable for every patient; depends on sepsis and source control.
What are the main learning points of this lecture on posterior uveitis, vasculitis and endophthalmitis?

Retinal vasculitis has many associations, so the work-up hunts for infection, systemic disease and masquerade tumours; FA shows leakage and true non-perfusion, while OCT and OCTA cannot replace it. Endophthalmitis is an intraocular infection, not simply panuveitis: delayed postoperative and endogenous forms need cultures, organism-appropriate antimicrobials and often surgery, and a negative culture does not prove sterility.

Which findings require urgent assessment?

Sudden visual loss, severe pain, marked inflammation, a rise in intraocular pressure, a new visual field defect, or a painful deteriorating eye after surgery. Suspected infectious or neurological causes require same-day assessment.

References and access status

  1. 1
    Original export: Briefing-Dokument: Posteriore Uveitis – Vaskulitis & Endophthalmitis

    Complete original export read; recording, transcript and attribution not independently verified.

  2. 2
    ReViSe systematic review of retinal vasculitis (PubMed 40320076)

    Complete NCBI abstract read10November2026 only. Heterogeneous97studies/7619patients, not a universal incidence estimate. Full report unread.

  3. 3
    IDSA 2016 clinical practice guideline for the management of candidiasis, ocular section XII (recommendations 82-91)

    Selected ocular recommendations82–91 in sectionXII of IDSA2016 and explanatory text read10November2026; mostly low evidence, multidisciplinary Candida care. Not whole guideline.

  4. 4
    EULAR 2025 update on the management of Behçet syndrome (PubMed 41876291), with the earlier systematic literature review (PMC13425003) as read for review 033

    The2025update abstract and selected PMC13425003 review passages read7November2026 within the previously recorded scope; no fresh full guideline appraisal. No universal steroid-free48hour rule; negativeIGRA does not exclude all infection.

  5. 5
    Postoperative Endophthalmitis Caused by Cutibacterium (Formerly Propionibacterium) Acnes: Case Series and Review2021

    Complete actual NCBI abstract read13November2026; retrospective six-eye series and literature review, not randomized. Full report/figures unread.

  6. 6
    Endophthalmitis Vitrectomy Study1995: randomized acute postoperative bacterial trial

    Complete actual NCBI abstract read13November2026; historical randomized acute postoperative population, not delayed/endogenous/fungal disease. Full report/protocol unread.

Original lecture link