Mohammad AskarPublic working copy · medical review pending

Intraocular lymphoma · 088

Vitreoretinal lymphoma: recognise the masquerade and plan the biopsy

New or atypical uveitis in an older adult warrants a broad differential. Imaging raises suspicion; coordinated tissue diagnosis and systemic staging determine the next step.

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.

Absence of macular oedema or synechiae does not diagnose lymphoma, and a normal brain MRI does not exclude ocular disease. Combine clinical findings, cytology and ancillary tests; choose treatment with haematology according to fitness, renal function and disease extent.

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.

01Define the disease and its population

Primary intraocular lymphoma (PIOL) is an umbrella term; primary vitreoretinal lymphoma (PVRL) mainly involves vitreous and retina and is usually diffuse large B-cell lymphoma. It belongs to the primary central nervous system lymphoma (PCNSL) spectrum and may first appear in the eye before brain lesions become detectable. 1 3

The original describes rarity, more women than men, mean onset at 50–60 years and bilateral disease in 80–90%, sometimes initially unilateral or asymmetric. Its less-than-1% share of intraocular tumours lacks a specified source population. The quoted 4–6% of intracranial tumours and 1–2% of extranodal non-Hodgkin lymphomas concern PCNSL rather than a measured share attributable to PVRL. These original figures should not be applied to every clinic population. 1

The original also reports CNS involvement in 60–80%, on average 29 months after ocular onset, and mean survival of 58 months. Their cohorts, treatment era and definitions remain unresolved, so they are not individual predictions. EHA–ESMO cites a broad 56–90% CNS-dissemination range within 30 months from heterogeneous, largely retrospective evidence; the differing figures cannot be treated as interchangeable estimates. 1 3

02Recognise clues without making them exclusion rules

Consider PVRL in new uveitis in an older adult, persistent or recurrent vitreous haze, or poor steroid response. A transient steroid response can also occur. Painless floaters and blurred vision may precede dramatic signs; acuity can appear better than expected from the haze, but this is not a diagnostic rule. 1 2

The anterior chamber may be quiet or contain cells, including atypical lymphoid cells in an analysed specimen. Large mutton-fat keratic precipitates in the first teaching case show that a granulomatous appearance does not exclude lymphoma. Vitreous infiltrates may form sheets described as aurora borealis, strings of pearls or snowballs; yellowish subretinal infiltrates and granular RPE changes add suspicion without proving the cause. 1 2

Cystoid macular oedema and posterior synechiae may be less conspicuous than in typical inflammatory uveitis. Their absence is a clue, not a diagnostic criterion; their presence does not exclude PVRL. The comparison table therefore qualifies the original absolute “no” entries. Infection, sarcoidosis, malignancy and other masquerade causes remain in the differential. 1 2

Recognise clues without making them exclusion rules
FeaturePVRL clues and limitsInflammatory uveitis clues and limits
Ocular injectionOften absent, but redness can occur.Can occur; silent disease can lack it.
Anterior chamberOften quiet, but cells are possible.Cells or irritation may occur.
Keratic precipitatesMay be absent or cellular/grey/diffuse; mutton-fat appearance does not exclude.Granulomatous or nongranulomatous patterns.
SynechiaeOften absent, not an exclusion rule.Possible, not obligatory.
VitreousCellular infiltrates, sometimes sheets or pearls.Cells and protein-related haze; appearances overlap.
RetinaSubretinal or retinal infiltrates; not diagnostic alone.Vasculitis, retinal necrosis, subretinal fluid or oedema according to cause.

03Use imaging to target diagnosis and staging

OCT can show intraretinal or subretinal infiltrates, irregular or indistinct retinal pigment epithelium (RPE), hyperreflective foci in the posterior vitreous and sub-RPE material above Bruch membrane. The original says this material typically does not cross Bruch membrane; a structural pattern alone cannot establish the tumour type. 1

Vertical hyperreflective lesions may raise suspicion, but the original “highly specific” designation lacks a verified accuracy study here. Fundus autofluorescence (FAF) and fluorescein angiography (FFA) complement OCT and examination. None replaces coordinated cytological diagnosis; a compatible scan is not permission to start a uniform cancer regimen. 1 2 3

Arrange prompt brain MRI and, according to findings, orbital imaging and neurological/haematological staging. Normal initial brain imaging does not exclude ocular lymphoma or later CNS involvement. Coordinate additional systemic and cerebrospinal-fluid investigations with the specialist team rather than assuming that isolated ocular presentation needs no staging. 2 3

04Case 1: revisit a diagnosis that does not fit

A teaching case describes a woman with seronegative rheumatoid arthritis and previous breast and bowel cancer after restarting DMARD treatment. In August 2023 blurred vision was attributed to hydroxychloroquine after only 1 week of exposure. That timing is not typical of chronic hydroxychloroquine retinopathy and should prompt reassessment rather than reliance on the drug label alone. 1

A later assessment labelled the disease dry AMD. After 3 months, diabetic eye screening attributed vision down to hand movements to cataract. Referral was delayed from January to July 2024 before uveitis-clinic assessment; keratic mutton-fat precipitates, anterior-chamber cells and multiple retinal lesions did not fit a simple lens-opacity explanation. 1

After another 2 months she developed right facial numbness, headache and slurred speech; MRI showed a pontine mass suspicious for CNS lymphoma. Vitreous biopsy confirmed aggressive B-cell lymphoma. Methotrexate and rituximab were started, but death from treatment complications was reported. 1

The original calls the nearly 1-year delay decisive for prognosis. The course illustrates missed opportunities to reconsider a diagnosis, but an uncontrolled case cannot establish that a particular delay caused death or that earlier treatment would certainly have prevented it. The actionable lesson is prompt escalation when the examination and presumed cause disagree. 1

05Cases 2 and 4: normal staging does not end follow-up

A 77-year-old woman had 1 month of visual loss and right panuveitis. Early suspicion in the emergency department led to a normal brain MRI, uveitis consultation and prompt vitreous biopsy confirming B-cell lymphoma. Timely haematology treatment and a favourable course were reported; this illustrates coordinated referral but does not prove the original assertion that the intervention saved her life. 1

Another case describes a 78-year-old man with renal disease referred for chorioretinitis. Vision rapidly fell to light perception with high intraocular pressure and iris neovascularisation; empirical antiviral treatment failed. OCT showed a large sub-RPE lesion between RPE and Bruch membrane, and vitreous biopsy confirmed B-cell lymphoma. 1

No systemic or CNS involvement was found at that staging, and systemic chemotherapy was selected to treat ocular disease and address future risk. The original aim of protecting the fellow eye and preventing CNS involvement is a therapeutic intention, not a guaranteed effect. Renal disease is particularly relevant when considering methotrexate-based treatment. 1 3

06Cases 3 and 5: infection and amyloid can imitate lymphoma

A 70-year-old wheelchair user with diabetes and a pacemaker was initially treated for vitreous haemorrhage. Severe postoperative panuveitis with pearl-string vitreous opacities and snowballs raised lymphoma suspicion. A penile lesion during admission redirected the investigation toward syphilis; the original reports Treponema pallidum identification in a work-up including vitreous biopsy, but the exact test and specimen route remain unverified. 1

A 2-week course of intravenous penicillin was reported to resolve inflammation and improve vision substantially. This case illustrates the importance of infection in the differential, not a regimen for every patient or proof that only men are at risk. Ocular syphilis can occur in any sex and age group; coordinate confirmatory testing and specialist treatment. 1

A fifth case had vision of hand movements and no fundus view after cataract surgery. Sepsis, renal and hepatic damage and a pacemaker were part of the history. The vitreous specimen was submitted for lymphoma versus amyloidosis and reportedly supported ocular amyloid, later associated with cardiac amyloidosis. 1

The original “pathognomonic lesions” wording does not supply amyloid typing or primary pathology records. Neither the opacities nor the cardiac history alone establishes the amyloid type. Tissue interpretation and systemic assessment determine what the finding means; imaging resemblance should widen the differential, not replace it. 1

07Plan the biopsy with the laboratory

Diagnostic vitrectomy with conventional cytology is central when PVRL is suspected. Coordinate immunocytochemistry or immunohistochemistry with CD markers, flow cytometry, PCR/clonality, MYD88 mutation analysis and IL-10/IL-6 testing according to specimen and laboratory capability. An IL-10:IL-6 ratio above 1 can support suspicion but does not establish or exclude every case by itself. 2 3

A 28-expert Delphi consensus using a 75% agreement threshold recommends considering withdrawal of systemic corticosteroids at least 2 weeks before biopsy to improve diagnostic yield. This requires clinical feasibility and agreement with the treating team; it is not advice to stop essential steroids abruptly or compromise a threatened eye. Consensus guidance is not a prospective diagnostic-accuracy trial. 2

The original specifies delivery within 10 minutes or preservation in CytoLyt. Treat this as a laboratory-dependent handling proposal, not a universal recipe: fixation suitable for cytology may destroy cells needed for flow cytometry or prevent another assay. Agree the fresh and fixed aliquots, volume, transport and rapid processing before sampling. 1 2 3

A negative or scant first specimen does not reliably exclude lymphoma. Reassess clinical suspicion, previous steroid exposure and specimen quality, and discuss repeat or alternative sampling with the specialist team. Communicate the suspected diagnosis before the procedure so the laboratory can prioritise the limited material. 2 3

08Choose ocular and systemic treatment together

Refer promptly to haematology/oncology for staging and joint treatment planning. High-dose methotrexate, cytarabine (ARA-C) and rituximab appear in systemic regimens, but their use depends on CNS involvement, fitness, renal function and the selected protocol. They are not an interchangeable standard prescription for every ocular presentation. 1 3

For fit PVRL patients, EHA–ESMO recommends high-dose methotrexate-based induction with low-level, largely retrospective disease-specific evidence (IV, C); intravitreal methotrexate can be added for rapid local regression. Selected intravitreal rituximab or ocular radiotherapy and consolidation strategies require specialist discussion. Ocular radiotherapy and whole-brain radiotherapy are different decisions; neither is automatically indicated for everyone. 3

Patients unable to tolerate systemic chemotherapy may receive selected local or alternative treatment. Local control alone does not prevent CNS relapse, while the preventive benefit of upfront systemic treatment in isolated ocular disease remains uncertain. Follow both ocular response and systemic/CNS status; explain benefit, toxicity and uncertainty rather than promising protection of the fellow eye or future brain disease. 3

What are the main learning points?

Absence of macular oedema or synechiae does not diagnose lymphoma, and a normal brain MRI does not exclude ocular disease. Combine clinical findings, cytology and ancillary tests; choose treatment with haematology according to fitness, renal function and disease extent.

When is assessment urgent?

Rapid visual loss, severe eye pain or pressure rise, or new facial numbness, headache or speech disturbance requires urgent assessment. Suspected lymphoma needs prompt specialist coordination; do not change corticosteroids without the treating team.

References and access status

  1. 1
    Immutable original export; primary intraocular lymphoma teaching cases

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

  2. 2
    Consensus Recommendations for the Diagnosis of Vitreoretinal Lymphoma

    Complete NCBI abstract read10November2026;28-expert Delphi/75%agreement, not a prospective accuracy trial. Full report/supplement unread.

  3. 3
    EHA–ESMO2024: Primary central nervous system lymphomas, selected PVRL sections

    Selected PVRL diagnosis/staging paragraphs and PVRL treatment section/recommendations read10November2026. Mainly retrospective PVRL-specific evidence; entire guideline, supplements and visual figure appraisal not completed.