Primary Central Nervous System Lymphoma

Complex MONDO:0002571 Pathograph 28 Show in embeddings browser Non-Hodgkin Lymphoma Central Nervous System Neoplasm

Primary central nervous system lymphoma is an aggressive diffuse large B-cell lymphoma whose defining feature is where it is not: the brain, spinal cord, leptomeninges and eyes are its exclusive sites at diagnosis, and demonstrable disease outside the central nervous system excludes the diagnosis. That restriction is the entity, and it is why staging deliberately looks for systemic disease it hopes not to find. The tumour arises from a B cell that has acquired a small, stereotyped set of somatic lesions — MYD88 L265P and CD79B together in the majority of cases — which switch on Toll-like-receptor and B-cell-receptor signalling without any ligand and drive canonical NF-kB constitutively. Clonal evolution studies of paired primary and relapse samples place those mutations, with TBL1XR1 and BCL6 rearrangement, in a common progenitor cell held in a memory B-cell state, and place the immune-escape lesions much later. The late group is the characteristic one: loss of the HLA locus at 6p21, B2M mutation, and gain or rearrangement of the PD-L1/PD-L2 locus at 9p24.1 all remove the tumour from T-cell view inside a compartment that was already immunologically sheltered. WHO's 2022 classification groups PCNSL with primary testicular and vitreoretinal large B-cell lymphoma on exactly this basis, as lymphomas of immune-privileged sites. Clinically it is a subacute mass lesion with a deceptive imaging signature — homogeneous enhancement and restricted diffusion rather than the ring enhancement of glioblastoma or abscess — and a notorious sensitivity to corticosteroids, which can make the tumour vanish radiographically and render a subsequent biopsy uninterpretable. Treatment is high-dose methotrexate-based induction followed by consolidation, and the consolidation choice is the one that matters most for what survivors are left with: autologous transplant preserves cognition where whole-brain radiotherapy erodes it.

Ask OpenScientist

Ask a research question about Primary Central Nervous System Lymphoma. OpenScientist will conduct autonomous deep research using the Disorder Mechanisms Knowledge Base and PubMed literature (typically 10-30 minutes).

Submitting...

Do not include personal health information in your question. Questions and results are cached in your browser's local storage.

1
Mappings
11
Pathophys.
11
Phenotypes
1
Gaps
28
Pathograph
7
Genes
4
Medical Actions
2
Differentials
1
Trials
1
Models
1
Deep Research
🏷

Classifications

ICD-O Morphology
Lymphoma
Harrison's Part
ONCOLOGY HEMATOLOGY
🔗

Mappings

MONDO
MONDO:0003655 cerebral lymphoma Not Yet Curated
skos:narrowMatch MONDO rdfs:subClassOf
MONDO asserts cerebral lymphoma as a direct subclass of primary central nervous system lymphoma; it is this entity restricted to the cerebral hemispheres. Recorded here so the narrower concept resolves to this entry.
?

Discussions and Knowledge Gaps

1
Does the transforming B cell acquire its lesions inside the CNS, or transform systemically and then home there?
KNOWLEDGE GAP cns_tropism_unresolved
Attached to
pathophysiology#Acquisition of MYD88 and CD79B Driver Mutations pathophysiology#Perivascular Infiltration of CNS Parenchyma
The disease is defined by a location whose origin is not established. Two accounts are current and are not mutually exclusive: that the clone is selected and retained in the CNS by chronic B-cell-receptor engagement with CNS self-proteins, or that transformation happens outside the CNS and the clone homes in afterwards via chemokine gradients. The clonal-evolution data complicate rather than settle it, because a common progenitor in a memory B-cell state implies a systemic reservoir while the immune-escape lesions that arise late look like adaptation to a compartment the tumour is already in. Until this is resolved, the entry's chain deliberately does not commit to an anatomical location for the initiating lesion.
Show evidence (1 reference)
PMID:37322012 SUPPORT Other
"Pathophysiology is incompletely understood, although a central role seems to comprise immunoglobulins binding to self-proteins expressed in the central nervous system (CNS) and alterations of genes involved in B cell receptor, Toll-like receptor and NF-κB signalling."
States both that the mechanism is unsettled and what the leading antigen-selection account proposes.
⚙

Pathophysiology

11
Acquisition of MYD88 and CD79B Driver Mutations
The initiating lesion, and an unusually stereotyped one. MYD88 L265P and CD79B ITAM mutations occur together in most cases, and clonal-evolution analysis of paired primary and relapse specimens places them, alongside TBL1XR1 mutation and BCL6 rearrangement, in a common progenitor cell that is held in a memory B-cell state before germinal-centre re-entry. This is the origin node: the transforming event is somatic, and the cell it happens in is bound as a memory B cell rather than as a generic B cell, because that is the state the cited clonal-evolution analysis places the common progenitor in. A single cell type is bound deliberately: binding both the generic and the specific term would make the cell-of-origin derivation report two origins, which would be an artefact of granularity rather than the genuine lump/split signal that report is for.
memory-state common progenitor B cell CL:0000787 Cell Ontology (CL) Relation: this pathophysiological event involves this cell type This pathophysiological event involves memory-state common progenitor B cell, annotated with memory B cell (CL:0000787). CL:0000787 is a cell type from the Cell Ontology.
Genetic context variant_origin: SOMATIC functional_impact_category: GAIN_OF_FUNCTION
Downstream
Ligand-Independent BCR and Toll-Like Receptor Signaling Both lesions act on the same receptor-proximal step: MYD88 L265P on the TLR/IL-1R adapter, CD79B on the B-cell receptor ITAM.
Show evidence (4 references)
PMID:30723112 SUPPORT Human Clinical
"Combined WES and targeted sequencing identified MYD88 mutation in 67% (42 of 63) of patients, CDKN2A biallelic loss in 44% (16 of 36), and CD79b mutation in 61% (22 of 36)."
Gives the frequencies of the three defining lesions in a sequenced cohort.
PMID:30723112 SUPPORT Human Clinical
"Phylogenetic analysis of paired primary and relapsed specimens identified MYD88 mutation and CDKN2A loss as early clonal events."
Establishes these as initiating rather than late lesions, which is what makes this the origin node.
PMID:36971477 SUPPORT Human Clinical
"All LBCL-IP sample pairs were clonally related, and both tumors developed from a common progenitor cell (CPC) with MYD88 and TBL1XR1 mutations and/or BCL6 translocations in 30/33 cases, indicating that these are early genetic events."
Identifies the common progenitor cell and the lesions it carries.
+ 1 more reference
Ligand-Independent BCR and Toll-Like Receptor Signaling
MYD88 L265P allows the myddosome to assemble without receptor ligation, and mutation of the CD79B ITAM tyrosine removes the negative-feedback step that normally terminates B-cell receptor signalling. Both convert a transient, antigen-gated signal into a continuous one. This is a qualitative change in regulation rather than a quantitative increase, which is why the pathway modifiers here are GAIN_OF_FUNCTION rather than INCREASED.
B cell receptor signaling pathway GO:0050853 Gene Ontology (GO) Relation: this pathophysiological event involves this biological process This pathophysiological event involves B cell receptor signaling pathway (GO:0050853), qualified as gain of function. GO:0050853 is a biological process from the Gene Ontology. ⇑ GAIN OF FUNCTION toll-like receptor signaling pathway GO:0002224 Gene Ontology (GO) Relation: this pathophysiological event involves this biological process This pathophysiological event involves toll-like receptor signaling pathway (GO:0002224), qualified as gain of function. GO:0002224 is a biological process from the Gene Ontology. ⇑ GAIN OF FUNCTION
Downstream
Constitutive Canonical NF-kB Activation
Show evidence (2 references)
PMID:37322012 SUPPORT Other
"Pathophysiology is incompletely understood, although a central role seems to comprise immunoglobulins binding to self-proteins expressed in the central nervous system (CNS) and alterations of genes involved in B cell receptor, Toll-like receptor and NF-κB signalling."
Names the three pathways this node and the next sit on, and is candid that the overall mechanism is not fully worked out.
PMID:28619981 SUPPORT Human Clinical
"Bruton tyrosine kinase (BTK) links the B-cell antigen receptor (BCR) and Toll-like receptors with NF-κB."
States the receptor-to-NF-kB connection this node and the next one model, and identifies the node in the chain that the BTK inhibitors act on.
Constitutive Canonical NF-kB Activation
The convergence point. Continuous signalling through both receptors, with CARD11 mutation and 3q12.3 gain driving NFKBIZ in a subset, holds canonical NF-kB transcriptionally active and sustains the survival and proliferation programme the tumour depends on. It is also the node the BTK inhibitors act on, which is why they work in a disease with no other targeted option.
canonical NF-kappaB signal transduction GO:0007249 Gene Ontology (GO) Relation: this pathophysiological event involves this biological process This pathophysiological event involves canonical NF-kappaB signal transduction (GO:0007249), qualified as gain of function. GO:0007249 is a biological process from the Gene Ontology. ⇑ GAIN OF FUNCTION
Downstream
Unchecked Clonal B-Cell Proliferation
Show evidence (1 reference)
PMID:30723112 SUPPORT Human Clinical
"PCNSL is characterized by frequent mutations within the B-cell receptor and NF-κB pathways."
States the pathway convergence that defines the disease's genomics.
Aberrant Somatic Hypermutation and CDKN2A Loss
A parallel arm rather than a downstream consequence: off-target activation-induced cytidine deaminase activity mutates PIM1, BTG1, BTG2 and other loci, while CDKN2A is lost by 9p21.3 deletion, mutation or promoter methylation. The result is loss of the p16INK4a/p14ARF brake on the cell cycle on top of the signalling lesions, and progressive genomic instability.
Genetic context variant_origin: SOMATIC functional_impact_category: LOSS_OF_FUNCTION
Downstream
Unchecked Clonal B-Cell Proliferation
Show evidence (2 references)
PMID:35646048 SUPPORT Human Clinical
"The most common mutations were identified in IGLL5 (68%), PIM1 (63%), MYD88 (55%), CD79B (42%), BTG2 (39%), PCLO (39%), KMT2D (34%), and BTG1 (29%) genes."
Lists the aberrant-somatic-hypermutation target genes alongside the signalling drivers in a whole-genome/exome cohort.
PMID:30723112 SUPPORT Human Clinical
"Copy-number analysis demonstrated frequent regions of copy loss (ie, CDKN2A), with few areas of amplification."
Records CDKN2A loss as the dominant copy-number event, against a copy-number landscape with little amplification.
Acquisition of Immune-Escape Lesions
The genetically distinctive arm of this disease, and the one that separates it from nodal DLBCL. Loss of the HLA locus at 6p21 by deletion or copy-neutral loss of heterozygosity, B2M mutation, and gain or rearrangement of the PD-L1/PD-L2 locus at 9p24.1 remove the tumour from both MHC-restricted recognition and T-cell effector function. Clonal-evolution analysis places these as late events, arising independently in primary and relapse samples rather than being inherited from the common progenitor.
Genetic context variant_origin: SOMATIC functional_impact_category: LOSS_OF_FUNCTION
MHC class II protein complex GO:0042613 Gene Ontology (GO) Relation: this pathophysiological event involves this cellular component This pathophysiological event involves decreased MHC class II protein complex (GO:0042613). GO:0042613 is a cellular component from the Gene Ontology.
Downstream
Escape from T-Cell Immunosurveillance
Show evidence (1 reference)
PMID:36971477 SUPPORT Human Clinical
"Genetic alterations in genes involved in immune escape (HLA, CD274/PDCD1LG2) were predominantly unique in primary and relapse samples and thus considered late genetic events."
Places the immune-escape lesions late in the clonal hierarchy and shows they arise independently in each tumour.
Escape from T-Cell Immunosurveillance
The functional consequence, and it compounds rather than creates the problem: the CNS is already an immune-sheltered compartment, so a tumour that also loses MHC display and overexpresses checkpoint ligands is doubly hidden. Reactive T cells, macrophages and microglia are present in the lesion but do not clear it. WHO's grouping of this disease with primary testicular and vitreoretinal large B-cell lymphoma rests on this shared biology.
CD8-positive, alpha-beta T cell CL:0000625 Cell Ontology (CL) Relation: this pathophysiological event involves this cell type This pathophysiological event involves CD8-positive, alpha-beta T cell (CL:0000625). CL:0000625 is a cell type from the Cell Ontology. microglial cell CL:0000129 Cell Ontology (CL) Relation: this pathophysiological event involves this cell type This pathophysiological event involves microglial cell (CL:0000129). CL:0000129 is a cell type from the Cell Ontology.
Downstream
Perivascular Infiltration of CNS Parenchyma
Show evidence (2 references)
PMID:36971477 SUPPORT Human Clinical
"Large B-cell lymphoma of immune-privileged sites (LBCL-IP) arise in immune sanctuaries including the testis and central nervous system (CNS)."
States the immune-sanctuary framing that groups this disease with its testicular and vitreoretinal counterparts.
PMID:37322012 SUPPORT Other
"Other factors such as T cells, macrophages or microglia, endothelial cells, chemokines, and interleukins, probably also have important roles."
Supports the presence and relevance of the reactive cellular compartment, while marking its role as probable rather than established.
Unchecked Clonal B-Cell Proliferation
Sustained NF-kB survival signalling with the cell-cycle brake removed. The clone that results is a diffuse large B-cell lymphoma by morphology and immunophenotype, non-germinal-centre in the great majority of cases, and it accounts for over 95% of lymphomas arising primarily in the CNS.
neoplastic B cell CL:0000236 Cell Ontology (CL) Relation: this pathophysiological event involves this cell type This pathophysiological event involves neoplastic B cell, annotated with B cell (CL:0000236). CL:0000236 is a cell type from the Cell Ontology.
Downstream
Perivascular Infiltration of CNS Parenchyma
Show evidence (1 reference)
PMID:37530337 SUPPORT Human Clinical
"Primary central nervous system diffuse large B-cell lymphoma (PCNS-DLBCL) is an uncommon extranodal lymphoma that accounts for more than 95% of all the CNS lymphomas."
Establishes DLBCL as the histology of essentially the whole entity.
Perivascular Infiltration of CNS Parenchyma
The tumour grows in angiocentric cuffs around CNS microvessels rather than as a discrete encapsulated mass, which is why resection has no role and why the lesion is diffusely infiltrative well beyond its enhancing margin. Deep periventricular, basal ganglia, thalamic and corpus callosum involvement is characteristic and prognostically adverse.
brain UBERON:0000955 Uberon multi-species anatomy ontology (UBERON) Relation: this pathophysiological event occurs in this anatomical location This pathophysiological event occurs in brain (UBERON:0000955). UBERON:0000955 is an anatomical location from the Uberon multi-species anatomy ontology. spinal cord UBERON:0002240 Uberon multi-species anatomy ontology (UBERON) Relation: this pathophysiological event occurs in this anatomical location This pathophysiological event occurs in spinal cord (UBERON:0002240). UBERON:0002240 is an anatomical location from the Uberon multi-species anatomy ontology.
Animal models
Canine primary CNS B-cell lymphoma Dog partially recapitulates Reproduces a primary B-cell lymphoma confined to the CNS in an outbred species, without engineered lesions.
Downstream
Mass Effect and Blood-Brain Barrier Disruption
Neoplasm of the nervous system
Increased CSF protein concentration
Show evidence (1 reference)
PMID:35096360 SUPPORT Human Clinical
"Most PCNSL occurred in the brain, followed by the spinal cord."
Gives the anatomical distribution bound on this node.
Mass Effect and Blood-Brain Barrier Disruption
Expanding perivascular tumour, vasogenic oedema and a leaking blood-brain barrier together produce the clinical syndrome. Which deficit appears is a matter of where the lesion sits rather than of anything specific to the tumour, which is why presentation is heterogeneous and frequently mistaken for stroke, demyelination or a psychiatric disorder before imaging.
Downstream
Seizure
Headache
Papilledema
Atypical behavior
Memory impairment
Ataxia
Hemiparesis
Aphasia
Show evidence (1 reference)
PMID:37322012 SUPPORT Other
"Clinical presentation varies depending on the involved regions of the CNS."
States that the clinical picture is determined by lesion location rather than by a tumour-specific syndrome.
Vitreoretinal Dissemination
Spread to the eye, itself a second immune-privileged compartment behind the blood-ocular barrier. It is common enough that ophthalmic examination is part of staging, and a substantial minority of affected eyes are asymptomatic, so it is found by looking rather than by complaint.
retina UBERON:0000966 Uberon multi-species anatomy ontology (UBERON) Relation: this pathophysiological event occurs in this anatomical location This pathophysiological event occurs in retina (UBERON:0000966). UBERON:0000966 is an anatomical location from the Uberon multi-species anatomy ontology.
Downstream
Blurred vision
Show evidence (1 reference)
PMID:37322012 SUPPORT Other
"Primary central nervous system lymphoma (PCNSL) is a diffuse large B cell lymphoma in which the brain, spinal cord, leptomeninges and/or eyes are exclusive sites of disease."
Establishes the eye as one of the disease's defining compartments.
Loss of EBV-Specific T-Cell Surveillance
A separate route into the same tumour, and the reason this entry is not purely a story about MYD88. In HIV infection, transplant immunosuppression and congenital immunodeficiency, failure of EBV-specific T-cell control permits outgrowth of an EBV-transformed B-cell clone without the same dependence on the somatic driver mutations above. The resulting disease is EBV-positive and is regarded as biologically distinct from the sporadic immunocompetent form, while converging on the same infiltrative endpoint.
Downstream
Unchecked Clonal B-Cell Proliferation Reaches the same proliferating clone by viral transformation under absent immune control rather than by the MYD88/CD79B route.
Show evidence (3 references)
PMID:36960939 SUPPORT Human Clinical
"Among 309 CNSL patients aged ≥60, 11.7% had EBV + tumors of which 72.2% were solid organ transplant (SOT)-related post-transplant lymphoproliferative disorders (PTLD)."
Quantifies the EBV-positive fraction and ties most of it to transplant-associated immunosuppression, which is the population this branch describes.
PMID:36960939 SUPPORT Human Clinical
"Younger age, SOT or autoimmune disease, and immunosuppressive treatment correlated highly with EBV-positivity."
Establishes the association between impaired immune control and EBV positivity that this node asserts.
PMID:37322012 SUPPORT Other
"Patients receiving prolonged immunosuppressive agents (such as those undergoing solid organ transplantation or those with autoimmune disorders) are the major at-risk population for PCNSL"
Names the at-risk population this branch is scoped to.
⬡

Pathograph

Use the checkboxes to hide or show graph categories. Hover nodes for evidence and cross-linked metadata.
Pathograph: causal mechanism network for Primary Central Nervous System Lymphoma Interactive directed graph showing how pathophysiology mechanisms, phenotypes, genetic factors and variants, experimental models, environmental triggers, and treatments relate through causal and linked edges.
●

Phenotypes

11
Eye 2
Papilledema HP:0001085 Human Phenotype Ontology (HP) Relation: this clinical feature is this phenotype This clinical feature is Papilledema (HP:0001085). HP:0001085 is a phenotype from the Human Phenotype Ontology.
Show evidence (1 reference)
PMID:37322012 SUPPORT Other
"Clinical presentation varies depending on the involved regions of the CNS."
Supports the raised-pressure presentation this sign belongs to; the source does not report its frequency.
Blurred vision HP:0000622 Human Phenotype Ontology (HP) Relation: this clinical feature is this phenotype This clinical feature is Blurred vision (HP:0000622). HP:0000622 is a phenotype from the Human Phenotype Ontology.
Show evidence (1 reference)
PMID:37322012 SUPPORT Other
"Primary central nervous system lymphoma (PCNSL) is a diffuse large B cell lymphoma in which the brain, spinal cord, leptomeninges and/or eyes are exclusive sites of disease."
Establishes ocular involvement as within the disease definition, which is what this visual phenotype reflects.
Metabolism 1
Increased CSF protein concentration HP:0002922 Human Phenotype Ontology (HP) Relation: this clinical feature is this phenotype This clinical feature is Increased CSF protein concentration (HP:0002922). HP:0002922 is a phenotype from the Human Phenotype Ontology.
Show evidence (1 reference)
PMID:38937027 SUPPORT INDIRECT Other
"In rare cases, cerebrospinal fluid (CSF) or vitreous humour might aid in providing a cytological diagnosis."
Supports CSF as a diagnostic compartment in this disease; the cited source addresses its cytological and biomarker yield rather than protein concentration specifically, so this is an indirect support.
Nervous System 7
Hemiparesis HP:0001269 Human Phenotype Ontology (HP) Relation: this clinical feature is this phenotype This clinical feature is Hemiparesis (HP:0001269). HP:0001269 is a phenotype from the Human Phenotype Ontology.
Show evidence (1 reference)
PMID:37322012 SUPPORT Other
"the most common clinical presentation is focal neurological deficit followed by altered mental status"
Establishes focal neurological deficit as the leading presentation, of which motor weakness is the commonest form.
Aphasia HP:0002381 Human Phenotype Ontology (HP) Relation: this clinical feature is this phenotype This clinical feature is Aphasia (HP:0002381). HP:0002381 is a phenotype from the Human Phenotype Ontology.
Show evidence (1 reference)
PMID:37322012 SUPPORT Other
"the most common clinical presentation is focal neurological deficit followed by altered mental status"
The same statement of the leading presentation; aphasia is its dominant-hemisphere form.
Atypical behavior HP:0000708 Human Phenotype Ontology (HP) Relation: this clinical feature is this phenotype This clinical feature is Behavioral and personality change, annotated with Atypical behavior (HP:0000708). HP:0000708 is a phenotype from the Human Phenotype Ontology.
Show evidence (1 reference)
PMID:37322012 SUPPORT Other
"Clinical presentation varies depending on the involved regions of the CNS."
Supports a location-determined clinical picture, of which behavioural change is the frontal manifestation; the source does not quantify it.
Memory impairment HP:0002354 Human Phenotype Ontology (HP) Relation: this clinical feature is this phenotype This clinical feature is Memory impairment (HP:0002354). HP:0002354 is a phenotype from the Human Phenotype Ontology.
Show evidence (1 reference)
PMID:30785830 SUPPORT Human Clinical
"Cognitive impairment was observed after WBRT, whereas cognitive functions were preserved or improved after ASCT."
Documents the treatment-related arm of this phenotype and the consolidation choice that determines it.
Seizure HP:0001250 Human Phenotype Ontology (HP) Relation: this clinical feature is this phenotype This clinical feature is Seizure (HP:0001250). HP:0001250 is a phenotype from the Human Phenotype Ontology.
Show evidence (1 reference)
PMID:37322012 SUPPORT Other
"Clinical presentation varies depending on the involved regions of the CNS."
The cited review ties the manifestations to lesion location; it does not give a seizure frequency, so none is recorded here.
Headache HP:0002315 Human Phenotype Ontology (HP) Relation: this clinical feature is this phenotype This clinical feature is Headache (HP:0002315). HP:0002315 is a phenotype from the Human Phenotype Ontology.
Show evidence (1 reference)
PMID:37322012 SUPPORT Other
"Clinical presentation varies depending on the involved regions of the CNS."
Supports a location-determined presentation; the source does not quantify headache specifically.
Ataxia HP:0001251 Human Phenotype Ontology (HP) Relation: this clinical feature is this phenotype This clinical feature is Ataxia (HP:0001251). HP:0001251 is a phenotype from the Human Phenotype Ontology.
Show evidence (1 reference)
PMID:37322012 SUPPORT Other
"Clinical presentation varies depending on the involved regions of the CNS."
Supports the location-dependent deficit of which this is the posterior-fossa form.
Neoplasm 1
Neoplasm of the nervous system HP:0004375 Human Phenotype Ontology (HP) Relation: this clinical feature is this phenotype This clinical feature is Neoplasm of the nervous system (HP:0004375). HP:0004375 is a phenotype from the Human Phenotype Ontology.
Show evidence (1 reference)
PMID:35096360 SUPPORT Human Clinical
"Most PCNSL occurred in the brain, followed by the spinal cord."
Records the anatomical distribution of the tumour.
🧬

Genetic Associations

7
MYD88
Gene: MYD88 hgnc:7562 HUGO Gene Nomenclature Committee (hgnc) Relation: this disease-associated gene is this gene This disease-associated gene is MYD88 (hgnc:7562). hgnc:7562 is a gene from the HUGO Gene Nomenclature Committee. relationship_type: CAUSATIVE variant_origin: SOMATIC
Show evidence (1 reference)
PMID:30723112 SUPPORT Human Clinical
"Combined WES and targeted sequencing identified MYD88 mutation in 67% (42 of 63) of patients, CDKN2A biallelic loss in 44% (16 of 36), and CD79b mutation in 61% (22 of 36)."
Gives the mutation frequency in a sequenced PCNSL cohort.
CD79B
Gene: CD79B hgnc:1699 HUGO Gene Nomenclature Committee (hgnc) Relation: this disease-associated gene is this gene This disease-associated gene is CD79B (hgnc:1699). hgnc:1699 is a gene from the HUGO Gene Nomenclature Committee. relationship_type: CAUSATIVE variant_origin: SOMATIC
Show evidence (1 reference)
PMID:35646048 SUPPORT Human Clinical
"The most common mutations were identified in IGLL5 (68%), PIM1 (63%), MYD88 (55%), CD79B (42%), BTG2 (39%), PCLO (39%), KMT2D (34%), and BTG1 (29%) genes."
Gives the CD79B mutation frequency alongside the other recurrent lesions.
CDKN2A
Gene: CDKN2A hgnc:1787 HUGO Gene Nomenclature Committee (hgnc) Relation: this disease-associated gene is this gene This disease-associated gene is CDKN2A (hgnc:1787). hgnc:1787 is a gene from the HUGO Gene Nomenclature Committee. relationship_type: CAUSATIVE variant_origin: SOMATIC
Show evidence (1 reference)
PMID:30723112 SUPPORT Human Clinical
"Phylogenetic analysis of paired primary and relapsed specimens identified MYD88 mutation and CDKN2A loss as early clonal events."
Establishes CDKN2A loss as an early rather than late clonal event.
PIM1
Gene: PIM1 hgnc:8986 HUGO Gene Nomenclature Committee (hgnc) Relation: this disease-associated gene is this gene This disease-associated gene is PIM1 (hgnc:8986). hgnc:8986 is a gene from the HUGO Gene Nomenclature Committee. relationship_type: CAUSATIVE variant_origin: SOMATIC
Show evidence (1 reference)
PMID:35646048 SUPPORT Human Clinical
"The most common mutations were identified in IGLL5 (68%), PIM1 (63%), MYD88 (55%), CD79B (42%), BTG2 (39%), PCLO (39%), KMT2D (34%), and BTG1 (29%) genes."
Gives the PIM1 mutation frequency in a whole-genome/exome cohort.
B2M
Gene: B2M hgnc:914 HUGO Gene Nomenclature Committee (hgnc) Relation: this disease-associated gene is this gene This disease-associated gene is B2M (hgnc:914). hgnc:914 is a gene from the HUGO Gene Nomenclature Committee. relationship_type: CAUSATIVE variant_origin: SOMATIC
Show evidence (1 reference)
PMID:36971477 SUPPORT Human Clinical
"Genetic alterations in genes involved in immune escape (HLA, CD274/PDCD1LG2) were predominantly unique in primary and relapse samples and thus considered late genetic events."
Places the immune-escape lesion class, which B2M belongs to, late in the clonal hierarchy.
CD274
Gene: CD274 hgnc:17635 HUGO Gene Nomenclature Committee (hgnc) Relation: this disease-associated gene is this gene This disease-associated gene is CD274 (hgnc:17635). hgnc:17635 is a gene from the HUGO Gene Nomenclature Committee. relationship_type: CAUSATIVE variant_origin: SOMATIC
Show evidence (1 reference)
PMID:31471540 SUPPORT Human Clinical
"Copy number alterations (CNAs) of 9p24.1 occur frequently in Hodgkin lymphoma, primary mediastinal large B-cell lymphoma (PMBCL), primary central nervous system lymphoma, and primary testicular lymphoma, resulting in overexpression of PD-L1 and sensitivity to PD-1 blockade-based immunotherapy."
Names PCNSL among the diseases where this locus is recurrently altered and states both the consequence and its therapeutic implication.
CARD11
Gene: CARD11 hgnc:16393 HUGO Gene Nomenclature Committee (hgnc) Relation: this disease-associated gene is this gene This disease-associated gene is CARD11 (hgnc:16393). hgnc:16393 is a gene from the HUGO Gene Nomenclature Committee. relationship_type: CAUSATIVE variant_origin: SOMATIC
Show evidence (1 reference)
PMID:28619981 SUPPORT Human Clinical
"The only PCNSL with complete ibrutinib resistance harbored a mutation within the coiled-coil domain of CARD11, a known ibrutinib resistance mechanism."
Demonstrates the CARD11 lesion in PCNSL and its functional consequence of bypassing upstream BTK inhibition.
💊

Medical Actions

4
High-Dose Methotrexate-Based Induction Chemotherapy
Action: high-dose methotrexate-based chemotherapyNCI Thesaurus (NCIT) Relation: this treatment is this clinical intervention This treatment is high-dose methotrexate-based chemotherapy, annotated with Chemotherapy (NCIT:C15632). NCIT:C15632 is a clinical intervention from the NCI Thesaurus. Ontology label: Chemotherapy NCIT:C15632
Agent: methotrexate CHEBI:44185 Chemical Entities of Biological Interest (CHEBI) Relation: this treatment uses this therapeutic agent This treatment uses methotrexate (CHEBI:44185). CHEBI:44185 is a therapeutic agent from Chemical Entities of Biological Interest. cytarabine CHEBI:28680 Chemical Entities of Biological Interest (CHEBI) Relation: this treatment uses this therapeutic agent This treatment uses cytarabine (CHEBI:28680). CHEBI:28680 is a therapeutic agent from Chemical Entities of Biological Interest. thiotepa CHEBI:9570 Chemical Entities of Biological Interest (CHEBI) Relation: this treatment uses this therapeutic agent This treatment uses thiotepa (CHEBI:9570). CHEBI:9570 is a therapeutic agent from Chemical Entities of Biological Interest. rituximab NCIT:C1702 NCI Thesaurus (NCIT) Relation: this treatment uses this therapeutic agent This treatment uses rituximab (NCIT:C1702). NCIT:C1702 is a therapeutic agent from the NCI Thesaurus.
Platform: Small molecule
The backbone. Methotrexate at high dose crosses the blood-brain barrier, which is the property the whole regimen is built around. The MATRix combination adds cytarabine, rituximab and thiotepa and roughly doubles the complete remission rate against methotrexate and cytarabine alone.
Mechanism Target:
Unchecked Clonal B-Cell Proliferation — Cytotoxic against the proliferating clone; the rituximab component adds CD20-directed killing.
Show evidence (2 references)
PMID:27132696 SUPPORT Human Clinical
"At median follow-up of 30 months (IQR 22-38), patients treated with rituximab and thiotepa had a complete remission rate of 49% (95% CI 38-60), compared with 23% (14-31) of those treated with methotrexate-cytarabine alone"
The randomised comparison establishing the four-drug regimen over the two-drug backbone.
PMID:37322012 SUPPORT Other
"Standard of care includes methotrexate-based polychemotherapy followed by age-tailored thiotepa-based conditioned autologous stem cell transplantation and, in patients unsuitable for such treatment, consolidation with whole-brain radiotherapy or single-drug maintenance."
States the overall treatment sequence this induction opens.
Autologous Haematopoietic Stem Cell Transplantation Consolidation
Action: autologous haematopoietic stem cell transplantationNCI Thesaurus (NCIT) Relation: this treatment is this clinical intervention This treatment is autologous haematopoietic stem cell transplantation, annotated with Autologous Hematopoietic Stem Cell Transplantation (NCIT:C16039). NCIT:C16039 is a clinical intervention from the NCI Thesaurus. Ontology label: Autologous Hematopoietic Stem Cell Transplantation NCIT:C16039
Platform: Cell therapy
Thiotepa-based conditioning followed by autologous transplant. In the PRECIS randomisation it gave better two-year progression-free survival than whole-brain radiotherapy and, more importantly for survivors, preserved or improved cognition where radiotherapy impaired it.
Mechanism Target:
Unchecked Clonal B-Cell Proliferation — Permits myeloablative consolidation dosing against residual clone, with stem-cell rescue.
Show evidence (1 reference)
PMID:30785830 SUPPORT Human Clinical
"The 2-year progression-free survival rates were 63% (95% CI, 49% to 81%) and 87% (95% CI, 77% to 98%) in the WBRT and ASCT arms, respectively."
The randomised progression-free survival comparison against radiotherapy consolidation.
Whole-Brain Radiotherapy Consolidation
Action: whole-brain radiotherapyNCI Thesaurus (NCIT) Relation: this treatment is this clinical intervention This treatment is whole-brain radiotherapy, annotated with Radiation Therapy (NCIT:C15313). NCIT:C15313 is a clinical intervention from the NCI Thesaurus. Ontology label: Radiation Therapy NCIT:C15313
Platform: Radiotherapy
Effective, and increasingly reserved for patients who cannot have a transplant. The reason is not efficacy but delayed neurotoxicity: cognitive decline, gait disturbance and incontinence that can arrive years later and that transplant consolidation avoids.
Mechanism Target:
Perivascular Infiltration of CNS Parenchyma — Irradiates the whole compartment the tumour infiltrates, rather than a resectable margin.
Show evidence (1 reference)
PMID:30785830 SUPPORT Human Clinical
"Cognitive impairment was observed after WBRT, whereas cognitive functions were preserved or improved after ASCT."
The cognitive outcome that has moved this from first choice to fallback in transplant-eligible patients.
Bruton Tyrosine Kinase Inhibition
Action: Bruton tyrosine kinase inhibitor therapyNCI Thesaurus (NCIT) Relation: this treatment is this clinical intervention This treatment is Bruton tyrosine kinase inhibitor therapy, annotated with Pharmacotherapy (NCIT:C15986). NCIT:C15986 is a clinical intervention from the NCI Thesaurus. Ontology label: Pharmacotherapy NCIT:C15986
Agent: tirabrutinib NCIT:C102876 NCI Thesaurus (NCIT) Relation: this treatment uses this therapeutic agent This treatment uses tirabrutinib (NCIT:C102876). NCIT:C102876 is a therapeutic agent from the NCI Thesaurus. ibrutinib NCIT:C81934 NCI Thesaurus (NCIT) Relation: this treatment uses this therapeutic agent This treatment uses ibrutinib (NCIT:C81934). NCIT:C81934 is a therapeutic agent from the NCI Thesaurus.
Platform: Small molecule
Tirabrutinib and ibrutinib target BTK in the B-cell receptor arm the disease depends on, and they work in relapsed and refractory disease where little else does. The limitation is durability rather than response: responses are frequent but median duration is measured in months.
Mechanism Target:
Ligand-Independent BCR and Toll-Like Receptor Signaling — Blocks BTK immediately downstream of the mutant receptor complex, which is why the drug class matches this disease's genetics.
Show evidence (2 references)
PMID:38690230 SUPPORT Human Clinical
"The overall response rate was 63.6% (95% CI: 47.8-77.6) with complete response (CR), unconfirmed CR, and partial response in 9, 7, and 12 patients, respectively."
Gives the response rate in relapsed and refractory disease.
PMID:38690230 SUPPORT Human Clinical
"The median duration of response (DOR) was 9.2 months, with a DOR rate of 19.8%; the median progression-free survival (PFS) and median overall survival (OS) were 2.9 months and not reached, respectively"
Records the short duration of response that is this class's main limitation.
🌍

Environmental Factors

1
Prolonged iatrogenic immunosuppression
exposure to prolonged immunosuppressive drug therapy Relation: this environmental factor is this exposure This environmental factor is exposure to prolonged immunosuppressive drug therapy.
Left unbound after searching. ECTO's drug-exposure terms cover named agents and classes rather than the state of sustained therapeutic immunosuppression, which is what this exposure is; ECTO:0000509 (exposure to drug) is the only close hit and is too general to say anything. Recorded as free text rather than bound to a term that would not carry the meaning.
Solid organ transplantation and long-term immunosuppressive treatment for autoimmune disease. This is not a cause of the sporadic disease and is not modelled as one; it is the exposure that permits the EBV-driven branch, and the link is drawn onto that node rather than onto the tumour as a whole.
Show evidence (1 reference)
PMID:37322012 SUPPORT Other
"Patients receiving prolonged immunosuppressive agents (such as those undergoing solid organ transplantation or those with autoimmune disorders) are the major at-risk population for PCNSL"
Establishes this exposure as the principal identified risk factor for the disease.
Mechanism Target:
PREDISPOSES Loss of EBV-Specific T-Cell Surveillance — Sustained pharmacological suppression of T-cell function is the route by which EBV-specific control is lost in the transplant and autoimmune populations.
Show evidence (1 reference)
PMID:36960939 SUPPORT Human Clinical
"Younger age, SOT or autoimmune disease, and immunosuppressive treatment correlated highly with EBV-positivity."
Ties immunosuppressive treatment specifically to the EBV-positive form this edge targets.
🔬

Biochemical Markers

1
Cerebrospinal fluid interleukin-10
Show evidence (2 references)
PMID:27924864 SUPPORT Human Clinical
"Using a CSF IL-10 cutoff value of 8.2 pg/ml, the diagnostic sensitivity and specificity were 95.5% and 96.1%, respectively"
Gives the diagnostic performance of the marker at a stated cutoff.
PMID:27924864 SUPPORT Human Clinical
"An increased CSF IL-10 level at diagnosis and post-treatment was associated with poor Progression free survival (PFS) for patients with PCNSL"
Records the prognostic dimension, which is why this is modelled as a biomarker rather than only as a diagnostic step.
🔬

Diagnosis

2
Stereotactic brain biopsy
The diagnostic gold standard, with one caveat that matters more here than in most tumours: corticosteroids are rapidly cytolytic to lymphoma cells and can make the lesion disappear radiographically and become uninterpretable histologically. They should be withheld before biopsy whenever it is safe to do so.
Show evidence (1 reference)
PMID:38937027 SUPPORT Other
"Suspicion raised on brain MRI must be confirmed by a histopathological diagnosis of a tumour specimen collected by stereotactic biopsy."
States the diagnostic pathway and the primacy of tissue diagnosis.
CSF MYD88 L265P and interleukin-10 assay
The minimally invasive adjunct, and the reason the driver mutation is clinically useful beyond its biology. MYD88 L265P by PCR together with CSF IL-10 gives a usable diagnostic signal where biopsy is hazardous or non-diagnostic.
Show evidence (1 reference)
PMID:38937027 SUPPORT Other
"Several independent studies have shown that MYD88Leu265Pro and IL-10 can be easily assessed in peripheral blood, plasma, aqueous and vitreous humour, and CSF of patients with PCNSL with substantial sensitivity and specificity, especially when evaluated in combination."
States the combined-biomarker approach and the compartments it can be run in.
📈

Progression

1
Outcome
Five-year overall survival across the SEER period was roughly a third, having improved substantially since the 1970s but not for patients over 60. A quarter to a half of those who respond will relapse.
Show evidence (2 references)
PMID:35096360 SUPPORT Human Clinical
"The 5-year overall survival rates in SEER 9 registries and SEER 18 registries were 30.5% and 37.4%, respectively."
Gives population-level survival from the two registry series.
PMID:37322012 SUPPORT Other
"Despite available treatments, 15-25% of patients do not respond to chemotherapy and 25-50% relapse after initial response."
Gives the refractory and relapse fractions that sit behind the survival figures.
📊

Prevalence

1
United States, SEER registries 1975-2017
Annual Incidence 0.5 per 100,000 per year 1–9 per 1,000,000 per year
Rose from 0.1 to 0.5 per 100,000 per year across the study period. The rate recorded here is the end-of-period figure; the trend itself is the more informative number and is quoted in the evidence.
Show evidence (1 reference)
PMID:35096360 SUPPORT Human Clinical
"Incidence rate increased from 0.1/100,000 to 0.5/100,000 with an AAPC of 5.3% from 1975 to 2017."
Gives both the incidence rate and its trend over four decades.
🔀

Differential Diagnoses

2

Conditions with similar clinical presentations that must be differentiated from Primary Central Nervous System Lymphoma:

Glioblastoma and other high-grade glioma
Overlapping Features The main radiological mimic, and the distinction changes management completely — glioma is resected, PCNSL is not. Imaging separates them imperfectly: PCNSL enhances homogeneously with restricted diffusion and lower relative cerebral blood volume, glioblastoma ring-enhances with higher perfusion. Biopsy settles it.
Show evidence (1 reference)
PMID:38937027 SUPPORT Other
"Suspicion raised on brain MRI must be confirmed by a histopathological diagnosis of a tumour specimen collected by stereotactic biopsy."
Supports tissue diagnosis as the step that resolves the imaging differential.
Secondary CNS involvement by systemic DLBCL
Overlapping Features Not a mimic but a definitional boundary, and the reason whole-body staging is mandatory rather than optional. Systemic lymphoma with CNS deposits is a different disease with different biology and management; finding extra-CNS disease reclassifies the case out of this entry.
Show evidence (1 reference)
PMID:37322012 SUPPORT Other
"PCNSL should be distinguished from lymphoma categories other than DLBCL primary arising in the CNS and from the secondary CNS lymphomas, which are diagnosed in patients with systemic DLBCL who have CNS involvement."
States the boundary directly.
🔬

Clinical Trials

1
NCT01011920 PHASE_II COMPLETED
IELSG32, the randomised trial that established the MATRix induction regimen and compared whole-brain radiotherapy with autologous transplant as consolidation.
Show evidence (1 reference)
PMID:27132696 SUPPORT Human Clinical
"This study is registered with ClinicalTrials.gov, number NCT01011920."
Ties the registration identifier to the published trial.
🐁

Animal Models

1
Canine primary CNS B-cell lymphoma
A spontaneous rather than engineered model, and included for that reason. Primary CNS lymphoma occurs naturally in dogs at a few percent of intracranial primary neoplasms, which makes it one of the few settings where the disease's CNS restriction arises without being imposed by the experiment. Single case reports rather than a characterised colony, so it supports the existence of the natural phenotype and little more.
Species
Dog
Genotype
Naturally occurring, no engineered genotype
Publication
PMID:23115372
Pathograph links
Perivascular Infiltration of CNS Parenchyma partially recapitulates fidelity: low observes: tissue Reproduces a primary B-cell lymphoma confined to the CNS in an outbred species, without engineered lesions.
Limitations: Case-report evidence only, with no molecular characterisation, so whether the canine disease shares the MYD88/CD79B genetics that define the human entity is unknown. It supports the natural occurrence of the anatomical phenotype and cannot be used to argue anything about mechanism.
Show evidence (1 reference)
PMID:23115372 SUPPORT Model Organism
"Canine primary central nervous system lymphomas constitute about 4% of all intracranial primary neoplasms, but comprehensive histopathologic classifications have rarely been carried out."
Establishes the natural occurrence and frequency in dogs, and states the characterisation gap that limits the model.
{ }

Source YAML

click to show 
name: Primary Central Nervous System Lymphoma
creation_date: "2026-09-09T12:00:00Z"
category: Complex
synonyms:
- PCNSL
- primary CNS lymphoma
- primary diffuse large B-cell lymphoma of the CNS
- PCNS-DLBCL
- primary brain lymphoma
description: >-
  Primary central nervous system lymphoma is an aggressive diffuse large B-cell
  lymphoma whose defining feature is where it is not: the brain, spinal cord,
  leptomeninges and eyes are its exclusive sites at diagnosis, and demonstrable
  disease outside the central nervous system excludes the diagnosis. That
  restriction is the entity, and it is why staging deliberately looks for
  systemic disease it hopes not to find.

  The tumour arises from a B cell that has acquired a small, stereotyped set of
  somatic lesions — MYD88 L265P and CD79B together in the majority of cases —
  which switch on Toll-like-receptor and B-cell-receptor signalling without any
  ligand and drive canonical NF-kB constitutively. Clonal evolution studies of
  paired primary and relapse samples place those mutations, with TBL1XR1 and BCL6
  rearrangement, in a common progenitor cell held in a memory B-cell state, and
  place the immune-escape lesions much later. The late group is the
  characteristic one: loss of the HLA locus at 6p21, B2M mutation, and gain or
  rearrangement of the PD-L1/PD-L2 locus at 9p24.1 all remove the tumour from
  T-cell view inside a compartment that was already immunologically sheltered.
  WHO's 2022 classification groups PCNSL with primary testicular and
  vitreoretinal large B-cell lymphoma on exactly this basis, as lymphomas of
  immune-privileged sites.

  Clinically it is a subacute mass lesion with a deceptive imaging signature —
  homogeneous enhancement and restricted diffusion rather than the ring
  enhancement of glioblastoma or abscess — and a notorious sensitivity to
  corticosteroids, which can make the tumour vanish radiographically and render a
  subsequent biopsy uninterpretable. Treatment is high-dose methotrexate-based
  induction followed by consolidation, and the consolidation choice is the one
  that matters most for what survivors are left with: autologous transplant
  preserves cognition where whole-brain radiotherapy erodes it.
disease_term:
  preferred_term: primary central nervous system lymphoma
  term:
    id: MONDO:0002571
    label: primary central nervous system lymphoma
mappings:
  mondo_mappings:
  - term:
      id: MONDO:0003655
      label: cerebral lymphoma
    mapping_predicate: skos:narrowMatch
    mapping_source: MONDO rdfs:subClassOf
    mapping_justification: >-
      MONDO asserts cerebral lymphoma as a direct subclass of primary central
      nervous system lymphoma; it is this entity restricted to the cerebral
      hemispheres. Recorded here so the narrower concept resolves to this entry.
parents:
- Non-Hodgkin Lymphoma
- Central Nervous System Neoplasm
classifications:
  icdo_morphology:
    classification_value: Lymphoma
  harrisons_chapter:
  - classification_value: ONCOLOGY_HEMATOLOGY
pathophysiology:
- name: Acquisition of MYD88 and CD79B Driver Mutations
  description: >-
    The initiating lesion, and an unusually stereotyped one. MYD88 L265P and
    CD79B ITAM mutations occur together in most cases, and clonal-evolution
    analysis of paired primary and relapse specimens places them, alongside
    TBL1XR1 mutation and BCL6 rearrangement, in a common progenitor cell that is
    held in a memory B-cell state before germinal-centre re-entry. This is the
    origin node: the transforming event is somatic, and the cell it happens in
    is bound as a memory B cell rather than as a generic B cell, because that is
    the state the cited clonal-evolution analysis places the common progenitor
    in. A single cell type is bound deliberately: binding both the generic and
    the specific term would make the cell-of-origin derivation report two
    origins, which would be an artefact of granularity rather than the genuine
    lump/split signal that report is for.
  biological_scale: MOLECULAR
  genetic_context:
    variant_origin: SOMATIC
    functional_impact_category: GAIN_OF_FUNCTION
  cell_types:
  - preferred_term: memory-state common progenitor B cell
    term:
      id: CL:0000787
      label: memory B cell
  downstream:
  - target: Ligand-Independent BCR and Toll-Like Receptor Signaling
    causal_link_type: DIRECT
    description: >-
      Both lesions act on the same receptor-proximal step: MYD88 L265P on the
      TLR/IL-1R adapter, CD79B on the B-cell receptor ITAM.
  evidence:
  - reference: PMID:30723112
    reference_title: "MYD88 L265P mutation and CDKN2A loss are early mutational events in primary central nervous system diffuse large B-cell lymphomas."
    supports: SUPPORT
    evidence_source: HUMAN_CLINICAL
    snippet: >-
      Combined WES and targeted sequencing identified MYD88 mutation in 67% (42
      of 63) of patients, CDKN2A biallelic loss in 44% (16 of 36), and CD79b
      mutation in 61% (22 of 36).
    explanation: >-
      Gives the frequencies of the three defining lesions in a sequenced cohort.
  - reference: PMID:30723112
    reference_title: "MYD88 L265P mutation and CDKN2A loss are early mutational events in primary central nervous system diffuse large B-cell lymphomas."
    supports: SUPPORT
    evidence_source: HUMAN_CLINICAL
    snippet: >-
      Phylogenetic analysis of paired primary and relapsed specimens identified
      MYD88 mutation and CDKN2A loss as early clonal events.
    explanation: >-
      Establishes these as initiating rather than late lesions, which is what
      makes this the origin node.
  - reference: PMID:36971477
    reference_title: "Large B-cell Lymphomas of Immune-Privileged Sites Relapse via Parallel Clonal Evolution from a Common Progenitor B Cell."
    supports: SUPPORT
    evidence_source: HUMAN_CLINICAL
    snippet: >-
      All LBCL-IP sample pairs were clonally related, and both tumors developed
      from a common progenitor cell (CPC) with MYD88 and TBL1XR1 mutations
      and/or BCL6 translocations in 30/33 cases, indicating that these are early
      genetic events.
    explanation: >-
      Identifies the common progenitor cell and the lesions it carries.
  - reference: PMID:36971477
    reference_title: "Large B-cell Lymphomas of Immune-Privileged Sites Relapse via Parallel Clonal Evolution from a Common Progenitor B Cell."
    supports: SUPPORT
    evidence_source: HUMAN_CLINICAL
    snippet: >-
      the CPC contains genetic alterations that support prolonged
      survival/proliferation and retention in a memory B-cell state, followed by
      germinal center reentry, aSHM and immune escape
    explanation: >-
      Supports binding the memory B-cell state on this node as the cell in which
      the initiating lesions are held.
- name: Ligand-Independent BCR and Toll-Like Receptor Signaling
  description: >-
    MYD88 L265P allows the myddosome to assemble without receptor ligation, and
    mutation of the CD79B ITAM tyrosine removes the negative-feedback step that
    normally terminates B-cell receptor signalling. Both convert a transient,
    antigen-gated signal into a continuous one. This is a qualitative change in
    regulation rather than a quantitative increase, which is why the pathway
    modifiers here are GAIN_OF_FUNCTION rather than INCREASED.
  biological_scale: MOLECULAR
  biological_processes:
  - preferred_term: B cell receptor signaling pathway
    modifier: GAIN_OF_FUNCTION
    term:
      id: GO:0050853
      label: B cell receptor signaling pathway
  - preferred_term: toll-like receptor signaling pathway
    modifier: GAIN_OF_FUNCTION
    term:
      id: GO:0002224
      label: toll-like receptor signaling pathway
  downstream:
  - target: Constitutive Canonical NF-kB Activation
    causal_link_type: DIRECT
  evidence:
  - reference: PMID:37322012
    reference_title: "Primary central nervous system lymphoma."
    supports: SUPPORT
    evidence_source: OTHER
    snippet: >-
      Pathophysiology is incompletely understood, although a central role seems
      to comprise immunoglobulins binding to self-proteins expressed in the
      central nervous system (CNS) and alterations of genes involved in B cell
      receptor, Toll-like receptor and NF-κB signalling.
    explanation: >-
      Names the three pathways this node and the next sit on, and is candid that
      the overall mechanism is not fully worked out.
  - reference: PMID:28619981
    reference_title: "Ibrutinib Unmasks Critical Role of Bruton Tyrosine Kinase in Primary CNS Lymphoma."
    supports: SUPPORT
    evidence_source: HUMAN_CLINICAL
    snippet: >-
      Bruton tyrosine kinase (BTK) links the B-cell antigen receptor (BCR) and
      Toll-like receptors with NF-κB.
    explanation: >-
      States the receptor-to-NF-kB connection this node and the next one model,
      and identifies the node in the chain that the BTK inhibitors act on.
- name: Constitutive Canonical NF-kB Activation
  description: >-
    The convergence point. Continuous signalling through both receptors, with
    CARD11 mutation and 3q12.3 gain driving NFKBIZ in a subset, holds canonical
    NF-kB transcriptionally active and sustains the survival and proliferation
    programme the tumour depends on. It is also the node the BTK inhibitors act
    on, which is why they work in a disease with no other targeted option.
  biological_scale: MOLECULAR
  biological_processes:
  - preferred_term: canonical NF-kappaB signal transduction
    modifier: GAIN_OF_FUNCTION
    term:
      id: GO:0007249
      label: canonical NF-kappaB signal transduction
  downstream:
  - target: Unchecked Clonal B-Cell Proliferation
    causal_link_type: DIRECT
  evidence:
  - reference: PMID:30723112
    reference_title: "MYD88 L265P mutation and CDKN2A loss are early mutational events in primary central nervous system diffuse large B-cell lymphomas."
    supports: SUPPORT
    evidence_source: HUMAN_CLINICAL
    snippet: >-
      PCNSL is characterized by frequent mutations within the B-cell receptor and
      NF-κB pathways.
    explanation: >-
      States the pathway convergence that defines the disease's genomics.
- name: Aberrant Somatic Hypermutation and CDKN2A Loss
  description: >-
    A parallel arm rather than a downstream consequence: off-target
    activation-induced cytidine deaminase activity mutates PIM1, BTG1, BTG2 and
    other loci, while CDKN2A is lost by 9p21.3 deletion, mutation or promoter
    methylation. The result is loss of the p16INK4a/p14ARF brake on the cell
    cycle on top of the signalling lesions, and progressive genomic instability.
  biological_scale: MOLECULAR
  genetic_context:
    variant_origin: SOMATIC
    functional_impact_category: LOSS_OF_FUNCTION
  downstream:
  - target: Unchecked Clonal B-Cell Proliferation
    causal_link_type: DIRECT
  evidence:
  - reference: PMID:35646048
    reference_title: "Whole-Genome/Exome Sequencing Uncovers Mutations and Copy Number Variations in Primary Diffuse Large B-Cell Lymphoma of the Central Nervous System."
    supports: SUPPORT
    evidence_source: HUMAN_CLINICAL
    snippet: >-
      The most common mutations were identified in IGLL5 (68%), PIM1 (63%), MYD88
      (55%), CD79B (42%), BTG2 (39%), PCLO (39%), KMT2D (34%), and BTG1 (29%)
      genes.
    explanation: >-
      Lists the aberrant-somatic-hypermutation target genes alongside the
      signalling drivers in a whole-genome/exome cohort.
  - reference: PMID:30723112
    reference_title: "MYD88 L265P mutation and CDKN2A loss are early mutational events in primary central nervous system diffuse large B-cell lymphomas."
    supports: SUPPORT
    evidence_source: HUMAN_CLINICAL
    snippet: >-
      Copy-number analysis demonstrated frequent regions of copy loss (ie,
      CDKN2A), with few areas of amplification.
    explanation: >-
      Records CDKN2A loss as the dominant copy-number event, against a
      copy-number landscape with little amplification.
- name: Acquisition of Immune-Escape Lesions
  description: >-
    The genetically distinctive arm of this disease, and the one that separates
    it from nodal DLBCL. Loss of the HLA locus at 6p21 by deletion or
    copy-neutral loss of heterozygosity, B2M mutation, and gain or rearrangement
    of the PD-L1/PD-L2 locus at 9p24.1 remove the tumour from both MHC-restricted
    recognition and T-cell effector function. Clonal-evolution analysis places
    these as late events, arising independently in primary and relapse samples
    rather than being inherited from the common progenitor.
  biological_scale: MOLECULAR
  genetic_context:
    variant_origin: SOMATIC
    functional_impact_category: LOSS_OF_FUNCTION
  cellular_components:
  - preferred_term: MHC class II protein complex
    modifier: DECREASED
    term:
      id: GO:0042613
      label: MHC class II protein complex
  downstream:
  - target: Escape from T-Cell Immunosurveillance
    causal_link_type: DIRECT
  evidence:
  - reference: PMID:36971477
    reference_title: "Large B-cell Lymphomas of Immune-Privileged Sites Relapse via Parallel Clonal Evolution from a Common Progenitor B Cell."
    supports: SUPPORT
    evidence_source: HUMAN_CLINICAL
    snippet: >-
      Genetic alterations in genes involved in immune escape (HLA,
      CD274/PDCD1LG2) were predominantly unique in primary and relapse samples
      and thus considered late genetic events.
    explanation: >-
      Places the immune-escape lesions late in the clonal hierarchy and shows
      they arise independently in each tumour.
- name: Escape from T-Cell Immunosurveillance
  description: >-
    The functional consequence, and it compounds rather than creates the problem:
    the CNS is already an immune-sheltered compartment, so a tumour that also
    loses MHC display and overexpresses checkpoint ligands is doubly hidden.
    Reactive T cells, macrophages and microglia are present in the lesion but do
    not clear it. WHO's grouping of this disease with primary testicular and
    vitreoretinal large B-cell lymphoma rests on this shared biology.
  biological_scale: CELLULAR
  cell_types:
  - preferred_term: CD8-positive, alpha-beta T cell
    term:
      id: CL:0000625
      label: CD8-positive, alpha-beta T cell
  - preferred_term: microglial cell
    term:
      id: CL:0000129
      label: microglial cell
  downstream:
  - target: Perivascular Infiltration of CNS Parenchyma
    causal_link_type: DIRECT
  evidence:
  - reference: PMID:36971477
    reference_title: "Large B-cell Lymphomas of Immune-Privileged Sites Relapse via Parallel Clonal Evolution from a Common Progenitor B Cell."
    supports: SUPPORT
    evidence_source: HUMAN_CLINICAL
    snippet: >-
      Large B-cell lymphoma of immune-privileged sites (LBCL-IP) arise in immune
      sanctuaries including the testis and central nervous system (CNS).
    explanation: >-
      States the immune-sanctuary framing that groups this disease with its
      testicular and vitreoretinal counterparts.
  - reference: PMID:37322012
    reference_title: "Primary central nervous system lymphoma."
    supports: SUPPORT
    evidence_source: OTHER
    snippet: >-
      Other factors such as T cells, macrophages or microglia, endothelial cells,
      chemokines, and interleukins, probably also have important roles.
    explanation: >-
      Supports the presence and relevance of the reactive cellular compartment,
      while marking its role as probable rather than established.
- name: Unchecked Clonal B-Cell Proliferation
  description: >-
    Sustained NF-kB survival signalling with the cell-cycle brake removed. The
    clone that results is a diffuse large B-cell lymphoma by morphology and
    immunophenotype, non-germinal-centre in the great majority of cases, and it
    accounts for over 95% of lymphomas arising primarily in the CNS.
  biological_scale: CELLULAR
  cell_types:
  - preferred_term: neoplastic B cell
    term:
      id: CL:0000236
      label: B cell
  downstream:
  - target: Perivascular Infiltration of CNS Parenchyma
    causal_link_type: DIRECT
  evidence:
  - reference: PMID:37530337
    reference_title: "Primary central nervous system lymphoma: Comprehension of cell-of-origin subtypes."
    supports: SUPPORT
    evidence_source: HUMAN_CLINICAL
    snippet: >-
      Primary central nervous system diffuse large B-cell lymphoma (PCNS-DLBCL)
      is an uncommon extranodal lymphoma that accounts for more than 95% of all
      the CNS lymphomas.
    explanation: >-
      Establishes DLBCL as the histology of essentially the whole entity.
- name: Perivascular Infiltration of CNS Parenchyma
  description: >-
    The tumour grows in angiocentric cuffs around CNS microvessels rather than as
    a discrete encapsulated mass, which is why resection has no role and why the
    lesion is diffusely infiltrative well beyond its enhancing margin. Deep
    periventricular, basal ganglia, thalamic and corpus callosum involvement is
    characteristic and prognostically adverse.
  biological_scale: TISSUE
  locations:
  - preferred_term: brain
    term:
      id: UBERON:0000955
      label: brain
  - preferred_term: spinal cord
    term:
      id: UBERON:0002240
      label: spinal cord
  downstream:
  - target: Mass Effect and Blood-Brain Barrier Disruption
    causal_link_type: DIRECT
  - target: Neoplasm of the nervous system
    causal_link_type: DIRECT
  - target: Increased CSF protein concentration
    causal_link_type: INDIRECT_KNOWN_INTERMEDIATES
  evidence:
  - reference: PMID:35096360
    reference_title: "Primary central nervous system lymphoma in the United States, 1975-2017."
    supports: SUPPORT
    evidence_source: HUMAN_CLINICAL
    snippet: >-
      Most PCNSL occurred in the brain, followed by the spinal cord.
    explanation: >-
      Gives the anatomical distribution bound on this node.
- name: Mass Effect and Blood-Brain Barrier Disruption
  description: >-
    Expanding perivascular tumour, vasogenic oedema and a leaking blood-brain
    barrier together produce the clinical syndrome. Which deficit appears is a
    matter of where the lesion sits rather than of anything specific to the
    tumour, which is why presentation is heterogeneous and frequently mistaken
    for stroke, demyelination or a psychiatric disorder before imaging.
  biological_scale: ORGANISM
  downstream:
  - target: Seizure
    causal_link_type: DIRECT
  - target: Headache
    causal_link_type: DIRECT
  - target: Papilledema
    causal_link_type: INDIRECT_KNOWN_INTERMEDIATES
  - target: Atypical behavior
    causal_link_type: DIRECT
  - target: Memory impairment
    causal_link_type: DIRECT
  - target: Ataxia
    causal_link_type: DIRECT
  - target: Hemiparesis
    causal_link_type: DIRECT
  - target: Aphasia
    causal_link_type: DIRECT
  evidence:
  - reference: PMID:37322012
    reference_title: "Primary central nervous system lymphoma."
    supports: SUPPORT
    evidence_source: OTHER
    snippet: >-
      Clinical presentation varies depending on the involved regions of the CNS.
    explanation: >-
      States that the clinical picture is determined by lesion location rather
      than by a tumour-specific syndrome.
- name: Vitreoretinal Dissemination
  description: >-
    Spread to the eye, itself a second immune-privileged compartment behind the
    blood-ocular barrier. It is common enough that ophthalmic examination is part
    of staging, and a substantial minority of affected eyes are asymptomatic, so
    it is found by looking rather than by complaint.
  biological_scale: TISSUE
  locations:
  - preferred_term: retina
    term:
      id: UBERON:0000966
      label: retina
  downstream:
  - target: Blurred vision
    causal_link_type: DIRECT
  evidence:
  - reference: PMID:37322012
    reference_title: "Primary central nervous system lymphoma."
    supports: SUPPORT
    evidence_source: OTHER
    snippet: >-
      Primary central nervous system lymphoma (PCNSL) is a diffuse large B cell
      lymphoma in which the brain, spinal cord, leptomeninges and/or eyes are
      exclusive sites of disease.
    explanation: >-
      Establishes the eye as one of the disease's defining compartments.
- name: Loss of EBV-Specific T-Cell Surveillance
  description: >-
    A separate route into the same tumour, and the reason this entry is not
    purely a story about MYD88. In HIV infection, transplant immunosuppression
    and congenital immunodeficiency, failure of EBV-specific T-cell control
    permits outgrowth of an EBV-transformed B-cell clone without the same
    dependence on the somatic driver mutations above. The resulting disease is
    EBV-positive and is regarded as biologically distinct from the sporadic
    immunocompetent form, while converging on the same infiltrative endpoint.
  biological_scale: ORGANISM
  downstream:
  - target: Unchecked Clonal B-Cell Proliferation
    causal_link_type: INDIRECT_KNOWN_INTERMEDIATES
    description: >-
      Reaches the same proliferating clone by viral transformation under absent
      immune control rather than by the MYD88/CD79B route.
  evidence:
  - reference: PMID:36960939
    reference_title: "EBV-positive PCNSL in older patients: incidence, characteristics, tumor pathology, and outcomes across a large multicenter cohort."
    supports: SUPPORT
    evidence_source: HUMAN_CLINICAL
    snippet: >-
      Among 309 CNSL patients aged ≥60, 11.7% had EBV + tumors of which 72.2%
      were solid organ transplant (SOT)-related post-transplant
      lymphoproliferative disorders (PTLD).
    explanation: >-
      Quantifies the EBV-positive fraction and ties most of it to
      transplant-associated immunosuppression, which is the population this
      branch describes.
  - reference: PMID:36960939
    reference_title: "EBV-positive PCNSL in older patients: incidence, characteristics, tumor pathology, and outcomes across a large multicenter cohort."
    supports: SUPPORT
    evidence_source: HUMAN_CLINICAL
    snippet: >-
      Younger age, SOT or autoimmune disease, and immunosuppressive treatment
      correlated highly with EBV-positivity.
    explanation: >-
      Establishes the association between impaired immune control and EBV
      positivity that this node asserts.
  - reference: PMID:37322012
    reference_title: "Primary central nervous system lymphoma."
    supports: SUPPORT
    evidence_source: OTHER
    snippet: >-
      Patients receiving prolonged immunosuppressive agents (such as those
      undergoing solid organ transplantation or those with autoimmune disorders)
      are the major at-risk population for PCNSL
    explanation: >-
      Names the at-risk population this branch is scoped to.
phenotypes:
- category: Clinical
  name: Neoplasm of the nervous system
  description: >-
    The lesion itself: one or more enhancing, diffusion-restricted intra-axial
    masses, most often supratentorial and frequently deep or periventricular.
  phenotype_term:
    preferred_term: Neoplasm of the nervous system
    term:
      id: HP:0004375
      label: Neoplasm of the nervous system
  evidence:
  - reference: PMID:35096360
    reference_title: "Primary central nervous system lymphoma in the United States, 1975-2017."
    supports: SUPPORT
    evidence_source: HUMAN_CLINICAL
    snippet: >-
      Most PCNSL occurred in the brain, followed by the spinal cord.
    explanation: >-
      Records the anatomical distribution of the tumour.
- category: Clinical
  name: Hemiparesis
  description: >-
    Focal motor deficit, the commonest way this disease presents. Which deficit
    appears follows the lesion, so hemiparesis and aphasia are two faces of the
    same thing rather than separate syndromes.
  phenotype_term:
    preferred_term: Hemiparesis
    term:
      id: HP:0001269
      label: Hemiparesis
  evidence:
  - reference: PMID:37322012
    reference_title: "Primary central nervous system lymphoma."
    supports: SUPPORT
    evidence_source: OTHER
    snippet: >-
      the most common clinical presentation is focal neurological deficit
      followed by altered mental status
    explanation: >-
      Establishes focal neurological deficit as the leading presentation, of
      which motor weakness is the commonest form.
- category: Clinical
  name: Aphasia
  description: >-
    Language deficit with dominant-hemisphere involvement, the other common form
    of the focal presentation.
  phenotype_term:
    preferred_term: Aphasia
    term:
      id: HP:0002381
      label: Aphasia
  evidence:
  - reference: PMID:37322012
    reference_title: "Primary central nervous system lymphoma."
    supports: SUPPORT
    evidence_source: OTHER
    snippet: >-
      the most common clinical presentation is focal neurological deficit
      followed by altered mental status
    explanation: >-
      The same statement of the leading presentation; aphasia is its
      dominant-hemisphere form.
- category: Clinical
  name: Atypical behavior
  description: >-
    Personality change, apathy and behavioural disturbance, often the earliest
    complaint and a common reason the diagnosis is initially missed.
  phenotype_term:
    preferred_term: Behavioral and personality change
    term:
      id: HP:0000708
      label: Atypical behavior
  evidence:
  - reference: PMID:37322012
    reference_title: "Primary central nervous system lymphoma."
    supports: SUPPORT
    evidence_source: OTHER
    snippet: >-
      Clinical presentation varies depending on the involved regions of the CNS.
    explanation: >-
      Supports a location-determined clinical picture, of which behavioural
      change is the frontal manifestation; the source does not quantify it.
- category: Clinical
  name: Memory impairment
  description: >-
    Cognitive decline at presentation, and separately a late consequence of
    whole-brain radiotherapy in survivors.
  phenotype_term:
    preferred_term: Memory impairment
    term:
      id: HP:0002354
      label: Memory impairment
  evidence:
  - reference: PMID:30785830
    reference_title: "Radiotherapy or Autologous Stem-Cell Transplantation for Primary CNS Lymphoma in Patients 60 Years of Age and Younger: Results of the Intergroup ANOCEF-GOELAMS Randomized Phase II PRECIS Study."
    supports: SUPPORT
    evidence_source: HUMAN_CLINICAL
    snippet: >-
      Cognitive impairment was observed after WBRT, whereas cognitive functions
      were preserved or improved after ASCT.
    explanation: >-
      Documents the treatment-related arm of this phenotype and the consolidation
      choice that determines it.
- category: Clinical
  name: Seizure
  description: >-
    Less frequent than in glioma, which is consistent with the deep and
    periventricular location typical of this tumour rather than a cortical one.
  phenotype_term:
    preferred_term: Seizure
    term:
      id: HP:0001250
      label: Seizure
  evidence:
  - reference: PMID:37322012
    reference_title: "Primary central nervous system lymphoma."
    supports: SUPPORT
    evidence_source: OTHER
    snippet: >-
      Clinical presentation varies depending on the involved regions of the CNS.
    explanation: >-
      The cited review ties the manifestations to lesion location; it does not
      give a seizure frequency, so none is recorded here.
- category: Clinical
  name: Headache
  description: >-
    Raised intracranial pressure from mass effect and vasogenic oedema.
  phenotype_term:
    preferred_term: Headache
    term:
      id: HP:0002315
      label: Headache
  evidence:
  - reference: PMID:37322012
    reference_title: "Primary central nervous system lymphoma."
    supports: SUPPORT
    evidence_source: OTHER
    snippet: >-
      Clinical presentation varies depending on the involved regions of the CNS.
    explanation: >-
      Supports a location-determined presentation; the source does not quantify
      headache specifically.
- category: Clinical
  name: Papilledema
  description: >-
    The fundoscopic sign of raised intracranial pressure, and the reason ocular
    examination is part of the initial assessment for a second reason beyond
    vitreoretinal disease.
  phenotype_term:
    preferred_term: Papilledema
    term:
      id: HP:0001085
      label: Papilledema
  evidence:
  - reference: PMID:37322012
    reference_title: "Primary central nervous system lymphoma."
    supports: SUPPORT
    evidence_source: OTHER
    snippet: >-
      Clinical presentation varies depending on the involved regions of the CNS.
    explanation: >-
      Supports the raised-pressure presentation this sign belongs to; the source
      does not report its frequency.
- category: Clinical
  name: Ataxia
  description: >-
    Gait and limb ataxia with cerebellar or brainstem involvement, which is the
    less common posterior-fossa presentation.
  phenotype_term:
    preferred_term: Ataxia
    term:
      id: HP:0001251
      label: Ataxia
  evidence:
  - reference: PMID:37322012
    reference_title: "Primary central nervous system lymphoma."
    supports: SUPPORT
    evidence_source: OTHER
    snippet: >-
      Clinical presentation varies depending on the involved regions of the CNS.
    explanation: >-
      Supports the location-dependent deficit of which this is the
      posterior-fossa form.
- category: Clinical
  name: Blurred vision
  description: >-
    Blurring and floaters from vitreoretinal involvement. A substantial share of
    eyes with disease are asymptomatic, so a normal visual history does not
    exclude ocular involvement.
  phenotype_term:
    preferred_term: Blurred vision
    term:
      id: HP:0000622
      label: Blurred vision
  evidence:
  - reference: PMID:37322012
    reference_title: "Primary central nervous system lymphoma."
    supports: SUPPORT
    evidence_source: OTHER
    snippet: >-
      Primary central nervous system lymphoma (PCNSL) is a diffuse large B cell
      lymphoma in which the brain, spinal cord, leptomeninges and/or eyes are
      exclusive sites of disease.
    explanation: >-
      Establishes ocular involvement as within the disease definition, which is
      what this visual phenotype reflects.
- category: Laboratory
  name: Increased CSF protein concentration
  description: >-
    Raised CSF protein, usually with pleocytosis and normal glucose. It is part
    of the IELSG prognostic score rather than merely a diagnostic finding.
  phenotype_term:
    preferred_term: Increased CSF protein concentration
    term:
      id: HP:0002922
      label: Increased CSF protein concentration
  evidence:
  - reference: PMID:38937027
    reference_title: "Molecular diagnosis of primary CNS lymphoma in 2024 using MYD88(Leu265Pro) and IL-10."
    supports: SUPPORT
    evidence_source: OTHER
    snippet: >-
      In rare cases, cerebrospinal fluid (CSF) or vitreous humour might aid in
      providing a cytological diagnosis.
    explanation: >-
      Supports CSF as a diagnostic compartment in this disease; the cited source
      addresses its cytological and biomarker yield rather than protein
      concentration specifically, so this is an indirect support.
    directness: INDIRECT
genetic:
- name: MYD88
  notes: >-
    Somatic, not germline. The L265P hotspot allows the MYD88-IRAK4-IRAK1
    myddosome to assemble without receptor ligation. It is also the basis of the
    CSF liquid-biopsy assay, which is unusual — the same lesion serves as driver,
    diagnostic marker and drug target.
  gene_term:
    preferred_term: MYD88
    term:
      id: hgnc:7562
      label: MYD88
  relationship_type: CAUSATIVE
  variant_origin: SOMATIC
  evidence:
  - reference: PMID:30723112
    reference_title: "MYD88 L265P mutation and CDKN2A loss are early mutational events in primary central nervous system diffuse large B-cell lymphomas."
    supports: SUPPORT
    evidence_source: HUMAN_CLINICAL
    snippet: >-
      Combined WES and targeted sequencing identified MYD88 mutation in 67% (42
      of 63) of patients, CDKN2A biallelic loss in 44% (16 of 36), and CD79b
      mutation in 61% (22 of 36).
    explanation: >-
      Gives the mutation frequency in a sequenced PCNSL cohort.
- name: CD79B
  notes: >-
    Somatic ITAM-domain mutation, most often co-occurring with MYD88 L265P. The
    pairing is what places most of these tumours in the MCD genomic subtype.
  gene_term:
    preferred_term: CD79B
    term:
      id: hgnc:1699
      label: CD79B
  relationship_type: CAUSATIVE
  variant_origin: SOMATIC
  evidence:
  - reference: PMID:35646048
    reference_title: "Whole-Genome/Exome Sequencing Uncovers Mutations and Copy Number Variations in Primary Diffuse Large B-Cell Lymphoma of the Central Nervous System."
    supports: SUPPORT
    evidence_source: HUMAN_CLINICAL
    snippet: >-
      The most common mutations were identified in IGLL5 (68%), PIM1 (63%), MYD88
      (55%), CD79B (42%), BTG2 (39%), PCLO (39%), KMT2D (34%), and BTG1 (29%)
      genes.
    explanation: >-
      Gives the CD79B mutation frequency alongside the other recurrent lesions.
- name: CDKN2A
  notes: >-
    Lost by 9p21.3 deletion, mutation or promoter hypermethylation. An early
    clonal event rather than a late one, which is why it sits on the initiating
    arm of the chain rather than at progression.
  gene_term:
    preferred_term: CDKN2A
    term:
      id: hgnc:1787
      label: CDKN2A
  relationship_type: CAUSATIVE
  variant_origin: SOMATIC
  evidence:
  - reference: PMID:30723112
    reference_title: "MYD88 L265P mutation and CDKN2A loss are early mutational events in primary central nervous system diffuse large B-cell lymphomas."
    supports: SUPPORT
    evidence_source: HUMAN_CLINICAL
    snippet: >-
      Phylogenetic analysis of paired primary and relapsed specimens identified
      MYD88 mutation and CDKN2A loss as early clonal events.
    explanation: >-
      Establishes CDKN2A loss as an early rather than late clonal event.
- name: PIM1
  notes: >-
    Mutated by aberrant somatic hypermutation rather than by a hotspot, and among
    the most frequently altered genes in sequenced cohorts.
  gene_term:
    preferred_term: PIM1
    term:
      id: hgnc:8986
      label: PIM1
  relationship_type: CAUSATIVE
  variant_origin: SOMATIC
  evidence:
  - reference: PMID:35646048
    reference_title: "Whole-Genome/Exome Sequencing Uncovers Mutations and Copy Number Variations in Primary Diffuse Large B-Cell Lymphoma of the Central Nervous System."
    supports: SUPPORT
    evidence_source: HUMAN_CLINICAL
    snippet: >-
      The most common mutations were identified in IGLL5 (68%), PIM1 (63%), MYD88
      (55%), CD79B (42%), BTG2 (39%), PCLO (39%), KMT2D (34%), and BTG1 (29%)
      genes.
    explanation: >-
      Gives the PIM1 mutation frequency in a whole-genome/exome cohort.
- name: B2M
  notes: >-
    Beta-2-microglobulin, the invariant light chain of MHC class I. Mutation or
    loss of heterozygosity removes class I from the cell surface outright, which
    is a more complete escape than reduced HLA expression.
  gene_term:
    preferred_term: B2M
    term:
      id: hgnc:914
      label: B2M
  relationship_type: CAUSATIVE
  variant_origin: SOMATIC
  evidence:
  - reference: PMID:36971477
    reference_title: "Large B-cell Lymphomas of Immune-Privileged Sites Relapse via Parallel Clonal Evolution from a Common Progenitor B Cell."
    supports: SUPPORT
    evidence_source: HUMAN_CLINICAL
    snippet: >-
      Genetic alterations in genes involved in immune escape (HLA,
      CD274/PDCD1LG2) were predominantly unique in primary and relapse samples
      and thus considered late genetic events.
    explanation: >-
      Places the immune-escape lesion class, which B2M belongs to, late in the
      clonal hierarchy.
- name: CD274
  notes: >-
    PD-L1, at 9p24.1 with PDCD1LG2 and JAK2. Copy gain or rearrangement of the
    locus overexpresses the checkpoint ligands. The same lesion is the reason
    PD-1 blockade is a rational thing to try in this disease, so it is both an
    escape mechanism and a therapeutic handle.
  gene_term:
    preferred_term: CD274
    term:
      id: hgnc:17635
      label: CD274
  relationship_type: CAUSATIVE
  variant_origin: SOMATIC
  evidence:
  - reference: PMID:31471540
    reference_title: "Amplification of 9p24.1 in diffuse large B-cell lymphoma identifies a unique subset of cases that resemble primary mediastinal large B-cell lymphoma."
    supports: SUPPORT
    evidence_source: HUMAN_CLINICAL
    snippet: >-
      Copy number alterations (CNAs) of 9p24.1 occur frequently in Hodgkin
      lymphoma, primary mediastinal large B-cell lymphoma (PMBCL), primary
      central nervous system lymphoma, and primary testicular lymphoma, resulting
      in overexpression of PD-L1 and sensitivity to PD-1 blockade-based
      immunotherapy.
    explanation: >-
      Names PCNSL among the diseases where this locus is recurrently altered and
      states both the consequence and its therapeutic implication.
- name: CARD11
  notes: >-
    Gain-of-function mutation activates NF-kB below the receptor, which makes it
    a route to the same output that does not depend on the receptor being
    engaged. That position is why it also confers resistance to BTK inhibition:
    the drug acts upstream of the lesion.

    Identifier note: the deep-research report for this entry gave hgnc:16412 for
    CARD11, which is NLRC4. The correct value used here was taken from the HGNC
    API.
  gene_term:
    preferred_term: CARD11
    term:
      id: hgnc:16393
      label: CARD11
  relationship_type: CAUSATIVE
  variant_origin: SOMATIC
  evidence:
  - reference: PMID:28619981
    reference_title: "Ibrutinib Unmasks Critical Role of Bruton Tyrosine Kinase in Primary CNS Lymphoma."
    supports: SUPPORT
    evidence_source: HUMAN_CLINICAL
    snippet: >-
      The only PCNSL with complete ibrutinib resistance harbored a mutation
      within the coiled-coil domain of CARD11, a known ibrutinib resistance
      mechanism.
    explanation: >-
      Demonstrates the CARD11 lesion in PCNSL and its functional consequence of
      bypassing upstream BTK inhibition.
prevalence:
- population: United States, SEER registries 1975-2017
  measure_type: ANNUAL_INCIDENCE
  prevalence_class: BAND_1_9_PER_1000000
  rate_per_100000: 0.5
  rate_denominator: POPULATION_PER_YEAR
  notes: >-
    Rose from 0.1 to 0.5 per 100,000 per year across the study period. The rate
    recorded here is the end-of-period figure; the trend itself is the more
    informative number and is quoted in the evidence.
  evidence:
  - reference: PMID:35096360
    reference_title: "Primary central nervous system lymphoma in the United States, 1975-2017."
    supports: SUPPORT
    evidence_source: HUMAN_CLINICAL
    snippet: >-
      Incidence rate increased from 0.1/100,000 to 0.5/100,000 with an AAPC of
      5.3% from 1975 to 2017.
    explanation: >-
      Gives both the incidence rate and its trend over four decades.
progression:
- phase: Outcome
  notes: >-
    Five-year overall survival across the SEER period was roughly a third, having
    improved substantially since the 1970s but not for patients over 60. A
    quarter to a half of those who respond will relapse.
  evidence:
  - reference: PMID:35096360
    reference_title: "Primary central nervous system lymphoma in the United States, 1975-2017."
    supports: SUPPORT
    evidence_source: HUMAN_CLINICAL
    snippet: >-
      The 5-year overall survival rates in SEER 9 registries and SEER 18
      registries were 30.5% and 37.4%, respectively.
    explanation: >-
      Gives population-level survival from the two registry series.
  - reference: PMID:37322012
    reference_title: "Primary central nervous system lymphoma."
    supports: SUPPORT
    evidence_source: OTHER
    snippet: >-
      Despite available treatments, 15-25% of patients do not respond to
      chemotherapy and 25-50% relapse after initial response.
    explanation: >-
      Gives the refractory and relapse fractions that sit behind the survival
      figures.
diagnosis:
- name: Stereotactic brain biopsy
  description: >-
    The diagnostic gold standard, with one caveat that matters more here than in
    most tumours: corticosteroids are rapidly cytolytic to lymphoma cells and can
    make the lesion disappear radiographically and become uninterpretable
    histologically. They should be withheld before biopsy whenever it is safe to
    do so.
  evidence:
  - reference: PMID:38937027
    reference_title: "Molecular diagnosis of primary CNS lymphoma in 2024 using MYD88(Leu265Pro) and IL-10."
    supports: SUPPORT
    evidence_source: OTHER
    snippet: >-
      Suspicion raised on brain MRI must be confirmed by a histopathological
      diagnosis of a tumour specimen collected by stereotactic biopsy.
    explanation: >-
      States the diagnostic pathway and the primacy of tissue diagnosis.
- name: CSF MYD88 L265P and interleukin-10 assay
  description: >-
    The minimally invasive adjunct, and the reason the driver mutation is
    clinically useful beyond its biology. MYD88 L265P by PCR together with CSF
    IL-10 gives a usable diagnostic signal where biopsy is hazardous or
    non-diagnostic.
  evidence:
  - reference: PMID:38937027
    reference_title: "Molecular diagnosis of primary CNS lymphoma in 2024 using MYD88(Leu265Pro) and IL-10."
    supports: SUPPORT
    evidence_source: OTHER
    snippet: >-
      Several independent studies have shown that MYD88Leu265Pro and IL-10 can be
      easily assessed in peripheral blood, plasma, aqueous and vitreous humour,
      and CSF of patients with PCNSL with substantial sensitivity and
      specificity, especially when evaluated in combination.
    explanation: >-
      States the combined-biomarker approach and the compartments it can be run
      in.
biochemical:
- name: Cerebrospinal fluid interleukin-10
  notes: >-
    A diagnostic biomarker rather than a disease mechanism, and an unusually good
    one for a compartment that is hard to sample. Raised CSF IL-10 separates PCNSL
    from other CNS disease with high sensitivity and specificity, and the
    IL-10/IL-6 ratio improves the separation from CNS infection specifically. It
    also carries prognostic information, which a purely diagnostic marker would
    not.
  evidence:
  - reference: PMID:27924864
    reference_title: "Cerebrospinal Fluid IL-10 and IL-10/IL-6 as Accurate Diagnostic Biomarkers for Primary Central Nervous System Large B-cell Lymphoma."
    supports: SUPPORT
    evidence_source: HUMAN_CLINICAL
    snippet: >-
      Using a CSF IL-10 cutoff value of 8.2 pg/ml, the diagnostic sensitivity and
      specificity were 95.5% and 96.1%, respectively
    explanation: >-
      Gives the diagnostic performance of the marker at a stated cutoff.
  - reference: PMID:27924864
    reference_title: "Cerebrospinal Fluid IL-10 and IL-10/IL-6 as Accurate Diagnostic Biomarkers for Primary Central Nervous System Large B-cell Lymphoma."
    supports: SUPPORT
    evidence_source: HUMAN_CLINICAL
    snippet: >-
      An increased CSF IL-10 level at diagnosis and post-treatment was associated
      with poor Progression free survival (PFS) for patients with PCNSL
    explanation: >-
      Records the prognostic dimension, which is why this is modelled as a
      biomarker rather than only as a diagnostic step.
environmental:
- name: Prolonged iatrogenic immunosuppression
  description: >-
    Solid organ transplantation and long-term immunosuppressive treatment for
    autoimmune disease. This is not a cause of the sporadic disease and is not
    modelled as one; it is the exposure that permits the EBV-driven branch, and
    the link is drawn onto that node rather than onto the tumour as a whole.
  exposure_term:
    preferred_term: exposure to prolonged immunosuppressive drug therapy
  notes: >-
    Left unbound after searching. ECTO's drug-exposure terms cover named agents
    and classes rather than the state of sustained therapeutic immunosuppression,
    which is what this exposure is; ECTO:0000509 (exposure to drug) is the only
    close hit and is too general to say anything. Recorded as free text rather
    than bound to a term that would not carry the meaning.
  influences_mechanisms:
  - target: Loss of EBV-Specific T-Cell Surveillance
    environmental_effect: PREDISPOSES
    causal_link_type: DIRECT
    description: >-
      Sustained pharmacological suppression of T-cell function is the route by
      which EBV-specific control is lost in the transplant and autoimmune
      populations.
    evidence:
    - reference: PMID:36960939
      reference_title: "EBV-positive PCNSL in older patients: incidence, characteristics, tumor pathology, and outcomes across a large multicenter cohort."
      supports: SUPPORT
      evidence_source: HUMAN_CLINICAL
      snippet: >-
        Younger age, SOT or autoimmune disease, and immunosuppressive treatment
        correlated highly with EBV-positivity.
      explanation: >-
        Ties immunosuppressive treatment specifically to the EBV-positive form
        this edge targets.
  evidence:
  - reference: PMID:37322012
    reference_title: "Primary central nervous system lymphoma."
    supports: SUPPORT
    evidence_source: OTHER
    snippet: >-
      Patients receiving prolonged immunosuppressive agents (such as those
      undergoing solid organ transplantation or those with autoimmune disorders)
      are the major at-risk population for PCNSL
    explanation: >-
      Establishes this exposure as the principal identified risk factor for the
      disease.
animal_models:
- name: Canine primary CNS B-cell lymphoma
  species: Dog
  genotype: Naturally occurring, no engineered genotype
  publication: PMID:23115372
  description: >-
    A spontaneous rather than engineered model, and included for that reason.
    Primary CNS lymphoma occurs naturally in dogs at a few percent of intracranial
    primary neoplasms, which makes it one of the few settings where the disease's
    CNS restriction arises without being imposed by the experiment. Single case
    reports rather than a characterised colony, so it supports the existence of
    the natural phenotype and little more.
  modeled_mechanisms:
  - target: Perivascular Infiltration of CNS Parenchyma
    relationship: PARTIALLY_RECAPITULATES
    fidelity: LOW
    model_scale: TISSUE
    description: >-
      Reproduces a primary B-cell lymphoma confined to the CNS in an outbred
      species, without engineered lesions.
    limitations: >-
      Case-report evidence only, with no molecular characterisation, so whether
      the canine disease shares the MYD88/CD79B genetics that define the human
      entity is unknown. It supports the natural occurrence of the anatomical
      phenotype and cannot be used to argue anything about mechanism.
    evidence:
    - reference: PMID:23115372
      reference_title: "Primary central nervous system B-cell lymphoma in a young dog."
      supports: SUPPORT
      evidence_source: MODEL_ORGANISM
      snippet: >-
        Canine primary central nervous system lymphomas constitute about 4% of
        all intracranial primary neoplasms, but comprehensive histopathologic
        classifications have rarely been carried out.
      explanation: >-
        Establishes the natural occurrence and frequency in dogs, and states the
        characterisation gap that limits the model.
treatments:
- name: High-Dose Methotrexate-Based Induction Chemotherapy
  description: >-
    The backbone. Methotrexate at high dose crosses the blood-brain barrier,
    which is the property the whole regimen is built around. The MATRix
    combination adds cytarabine, rituximab and thiotepa and roughly doubles the
    complete remission rate against methotrexate and cytarabine alone.
  therapeutic_modality: SMALL_MOLECULE
  treatment_term:
    preferred_term: high-dose methotrexate-based chemotherapy
    term:
      id: NCIT:C15632
      label: Chemotherapy
    therapeutic_agent:
    - preferred_term: methotrexate
      term:
        id: CHEBI:44185
        label: methotrexate
    - preferred_term: cytarabine
      term:
        id: CHEBI:28680
        label: cytarabine
    - preferred_term: thiotepa
      term:
        id: CHEBI:9570
        label: Thiotepa
    - preferred_term: rituximab
      term:
        id: NCIT:C1702
        label: Rituximab
  target_mechanisms:
  - target: Unchecked Clonal B-Cell Proliferation
    description: >-
      Cytotoxic against the proliferating clone; the rituximab component adds
      CD20-directed killing.
  evidence:
  - reference: PMID:27132696
    reference_title: "Chemoimmunotherapy with methotrexate, cytarabine, thiotepa, and rituximab (MATRix regimen) in patients with primary CNS lymphoma: results of the first randomisation of the International Extranodal Lymphoma Study Group-32 (IELSG32) phase 2 trial."
    supports: SUPPORT
    evidence_source: HUMAN_CLINICAL
    snippet: >-
      At median follow-up of 30 months (IQR 22-38), patients treated with
      rituximab and thiotepa had a complete remission rate of 49% (95% CI 38-60),
      compared with 23% (14-31) of those treated with methotrexate-cytarabine
      alone
    explanation: >-
      The randomised comparison establishing the four-drug regimen over the
      two-drug backbone.
  - reference: PMID:37322012
    reference_title: "Primary central nervous system lymphoma."
    supports: SUPPORT
    evidence_source: OTHER
    snippet: >-
      Standard of care includes methotrexate-based polychemotherapy followed by
      age-tailored thiotepa-based conditioned autologous stem cell
      transplantation and, in patients unsuitable for such treatment,
      consolidation with whole-brain radiotherapy or single-drug maintenance.
    explanation: >-
      States the overall treatment sequence this induction opens.
- name: Autologous Haematopoietic Stem Cell Transplantation Consolidation
  description: >-
    Thiotepa-based conditioning followed by autologous transplant. In the PRECIS
    randomisation it gave better two-year progression-free survival than
    whole-brain radiotherapy and, more importantly for survivors, preserved or
    improved cognition where radiotherapy impaired it.
  therapeutic_modality: CELL_THERAPY
  treatment_term:
    preferred_term: autologous haematopoietic stem cell transplantation
    term:
      id: NCIT:C16039
      label: Autologous Hematopoietic Stem Cell Transplantation
  target_mechanisms:
  - target: Unchecked Clonal B-Cell Proliferation
    description: >-
      Permits myeloablative consolidation dosing against residual clone, with
      stem-cell rescue.
  evidence:
  - reference: PMID:30785830
    reference_title: "Radiotherapy or Autologous Stem-Cell Transplantation for Primary CNS Lymphoma in Patients 60 Years of Age and Younger: Results of the Intergroup ANOCEF-GOELAMS Randomized Phase II PRECIS Study."
    supports: SUPPORT
    evidence_source: HUMAN_CLINICAL
    snippet: >-
      The 2-year progression-free survival rates were 63% (95% CI, 49% to 81%)
      and 87% (95% CI, 77% to 98%) in the WBRT and ASCT arms, respectively.
    explanation: >-
      The randomised progression-free survival comparison against radiotherapy
      consolidation.
- name: Whole-Brain Radiotherapy Consolidation
  description: >-
    Effective, and increasingly reserved for patients who cannot have a
    transplant. The reason is not efficacy but delayed neurotoxicity: cognitive
    decline, gait disturbance and incontinence that can arrive years later and
    that transplant consolidation avoids.
  therapeutic_modality: RADIOTHERAPY
  treatment_term:
    preferred_term: whole-brain radiotherapy
    term:
      id: NCIT:C15313
      label: Radiation Therapy
  target_mechanisms:
  - target: Perivascular Infiltration of CNS Parenchyma
    description: >-
      Irradiates the whole compartment the tumour infiltrates, rather than a
      resectable margin.
  evidence:
  - reference: PMID:30785830
    reference_title: "Radiotherapy or Autologous Stem-Cell Transplantation for Primary CNS Lymphoma in Patients 60 Years of Age and Younger: Results of the Intergroup ANOCEF-GOELAMS Randomized Phase II PRECIS Study."
    supports: SUPPORT
    evidence_source: HUMAN_CLINICAL
    snippet: >-
      Cognitive impairment was observed after WBRT, whereas cognitive functions
      were preserved or improved after ASCT.
    explanation: >-
      The cognitive outcome that has moved this from first choice to fallback in
      transplant-eligible patients.
- name: Bruton Tyrosine Kinase Inhibition
  description: >-
    Tirabrutinib and ibrutinib target BTK in the B-cell receptor arm the disease
    depends on, and they work in relapsed and refractory disease where little
    else does. The limitation is durability rather than response: responses are
    frequent but median duration is measured in months.
  therapeutic_modality: SMALL_MOLECULE
  treatment_term:
    preferred_term: Bruton tyrosine kinase inhibitor therapy
    term:
      id: NCIT:C15986
      label: Pharmacotherapy
    therapeutic_agent:
    - preferred_term: tirabrutinib
      term:
        id: NCIT:C102876
        label: Tirabrutinib
    - preferred_term: ibrutinib
      term:
        id: NCIT:C81934
        label: Ibrutinib
  target_mechanisms:
  - target: Ligand-Independent BCR and Toll-Like Receptor Signaling
    description: >-
      Blocks BTK immediately downstream of the mutant receptor complex, which is
      why the drug class matches this disease's genetics.
  evidence:
  - reference: PMID:38690230
    reference_title: "Three-year follow-up analysis of phase 1/2 study on tirabrutinib in patients with relapsed or refractory primary central nervous system lymphoma."
    supports: SUPPORT
    evidence_source: HUMAN_CLINICAL
    snippet: >-
      The overall response rate was 63.6% (95% CI: 47.8-77.6) with complete
      response (CR), unconfirmed CR, and partial response in 9, 7, and 12
      patients, respectively.
    explanation: >-
      Gives the response rate in relapsed and refractory disease.
  - reference: PMID:38690230
    reference_title: "Three-year follow-up analysis of phase 1/2 study on tirabrutinib in patients with relapsed or refractory primary central nervous system lymphoma."
    supports: SUPPORT
    evidence_source: HUMAN_CLINICAL
    snippet: >-
      The median duration of response (DOR) was 9.2 months, with a DOR rate of
      19.8%; the median progression-free survival (PFS) and median overall
      survival (OS) were 2.9 months and not reached, respectively
    explanation: >-
      Records the short duration of response that is this class's main
      limitation.
clinical_trials:
- name: NCT01011920
  phase: PHASE_II
  status: COMPLETED
  description: >-
    IELSG32, the randomised trial that established the MATRix induction regimen
    and compared whole-brain radiotherapy with autologous transplant as
    consolidation.
  evidence:
  - reference: PMID:27132696
    reference_title: "Chemoimmunotherapy with methotrexate, cytarabine, thiotepa, and rituximab (MATRix regimen) in patients with primary CNS lymphoma: results of the first randomisation of the International Extranodal Lymphoma Study Group-32 (IELSG32) phase 2 trial."
    supports: SUPPORT
    evidence_source: HUMAN_CLINICAL
    snippet: >-
      This study is registered with ClinicalTrials.gov, number NCT01011920.
    explanation: >-
      Ties the registration identifier to the published trial.
differential_diagnoses:
- name: Glioblastoma and other high-grade glioma
  description: >-
    The main radiological mimic, and the distinction changes management
    completely — glioma is resected, PCNSL is not. Imaging separates them
    imperfectly: PCNSL enhances homogeneously with restricted diffusion and lower
    relative cerebral blood volume, glioblastoma ring-enhances with higher
    perfusion. Biopsy settles it.
  evidence:
  - reference: PMID:38937027
    reference_title: "Molecular diagnosis of primary CNS lymphoma in 2024 using MYD88(Leu265Pro) and IL-10."
    supports: SUPPORT
    evidence_source: OTHER
    snippet: >-
      Suspicion raised on brain MRI must be confirmed by a histopathological
      diagnosis of a tumour specimen collected by stereotactic biopsy.
    explanation: >-
      Supports tissue diagnosis as the step that resolves the imaging
      differential.
- name: Secondary CNS involvement by systemic DLBCL
  description: >-
    Not a mimic but a definitional boundary, and the reason whole-body staging is
    mandatory rather than optional. Systemic lymphoma with CNS deposits is a
    different disease with different biology and management; finding extra-CNS
    disease reclassifies the case out of this entry.
  evidence:
  - reference: PMID:37322012
    reference_title: "Primary central nervous system lymphoma."
    supports: SUPPORT
    evidence_source: OTHER
    snippet: >-
      PCNSL should be distinguished from lymphoma categories other than DLBCL
      primary arising in the CNS and from the secondary CNS lymphomas, which are
      diagnosed in patients with systemic DLBCL who have CNS involvement.
    explanation: >-
      States the boundary directly.
discussions:
- discussion_id: cns_tropism_unresolved
  kind: KNOWLEDGE_GAP
  prompt: >-
    Does the transforming B cell acquire its lesions inside the CNS, or transform
    systemically and then home there?
  attaches_to:
  - pathophysiology#Acquisition of MYD88 and CD79B Driver Mutations
  - pathophysiology#Perivascular Infiltration of CNS Parenchyma
  rationale: >-
    The disease is defined by a location whose origin is not established. Two
    accounts are current and are not mutually exclusive: that the clone is
    selected and retained in the CNS by chronic B-cell-receptor engagement with
    CNS self-proteins, or that transformation happens outside the CNS and the
    clone homes in afterwards via chemokine gradients. The clonal-evolution data
    complicate rather than settle it, because a common progenitor in a memory
    B-cell state implies a systemic reservoir while the immune-escape lesions
    that arise late look like adaptation to a compartment the tumour is already
    in. Until this is resolved, the entry's chain deliberately does not commit to
    an anatomical location for the initiating lesion.
  evidence:
  - reference: PMID:37322012
    reference_title: "Primary central nervous system lymphoma."
    supports: SUPPORT
    evidence_source: OTHER
    snippet: >-
      Pathophysiology is incompletely understood, although a central role seems
      to comprise immunoglobulins binding to self-proteins expressed in the
      central nervous system (CNS) and alterations of genes involved in B cell
      receptor, Toll-like receptor and NF-κB signalling.
    explanation: >-
      States both that the mechanism is unsettled and what the leading
      antigen-selection account proposes.
notes: >-
  Entry level and MONDO binding. This is curated at the WHO entity level as
  primary CNS lymphoma rather than at the anatomical level of the stub that
  prompted it. MONDO asserts cerebral lymphoma (MONDO:0003655) as a direct
  subclass of MONDO:0002571, so the narrower term is recorded as a
  skos:narrowMatch rather than given its own entry.

  WHO 2022 places this disease inside "large B-cell lymphomas of immune-privileged
  sites" alongside primary testicular and primary vitreoretinal large B-cell
  lymphoma, while the International Consensus Classification keeps it as a
  standalone entity. The entry follows the ICC here, because the immune-privileged
  grouping is a shared-mechanism family of the kind this knowledge base models
  with a Grouping rather than by merging entries. A grouping record for the three
  would be a reasonable follow-up.

  Ontology corrections to the deep-research report. The report (claude_code) was
  strong on content — 32 of 32 citations resolved with no confabulation — and
  poor on identifiers, and the two are worth separating. It offered at least
  seven wrong bindings, of which its own term-validation section caught three:

  - NCIT:C9218, offered as "Primary Central Nervous System Lymphoma", is Stage IV
    Oropharyngeal Carcinoma AJCC v6.
  - UBERON:0004087, offered as "vitreous body", is vena cava.
  - NCIT:C36044, offered as "Diffuse Large B-Cell Lymphoma", is Grade 3b
    Malignant Neoplasm, and is obsolete.

  Four more were caught only by checking each identifier against its cache or
  ontology by hand, because the validator label-checked 31 of 75 terms and skips
  gene CURIEs entirely:

  - hgnc:16412, offered as CARD11, is NLRC4. CARD11 is hgnc:16393.
  - HP:0000618, offered as "Blurred vision", is Blindness. The correct term,
    used here, is HP:0000622.
  - HP:0002153, offered as "Papilledema", is Hyperkalemia. The correct term,
    used here, is HP:0001085.
  - HP:0034332, offered as "Focal neurologic deficit", is Cognitive regression.

  One further trap is worth recording because the tooling actively pointed at it.
  The report gave CHEBI:9560 for thiotepa, and the validator's obsolescence check
  reported it as replaced by CHEBI:102166 — which is thiopental, an anaesthetic
  barbiturate, not the alkylating agent. Following the automated replacement would
  have bound the wrong drug. Thiotepa is CHEBI:9570, taken from a direct search.

  No datasets block. PCNSL has published genomic cohorts, but the accessions were
  not verified in this session and an unverified accession is worse than none.
  This is an ordinary gap rather than a reasoned exclusion, and is a good
  follow-up.

  References fetched but not cited. The deep-research run resolved 42 references
  into the cache and this entry cites 15 of them. The rest were read and set
  aside, and the reasons fall into three groups, recorded here so that a fetched
  cache is not mistaken for research nobody looked at.

  Several are about systemic DLBCL, Hodgkin lymphoma or primary mediastinal
  lymphoma rather than this disease, and were retrieved because the report
  discusses PCNSL against those comparators. PMID:31471540 is the exception that
  is cited, because it names PCNSL explicitly among the diseases carrying the
  9p24.1 lesion.

  Several cover treatment options this entry does not yet model - lenalidomide,
  checkpoint inhibitors, CAR-T, and the blood-brain-barrier penetration
  strategies. They are real and they belong here eventually; the entry currently
  carries induction, consolidation, radiotherapy and BTK inhibition, and adding
  a relapsed/refractory arm properly is a follow-up rather than something to
  bolt on.

  A few are imaging and radiomics studies supporting the differential against
  glioblastoma. The differential entry makes the weaker claim that tissue
  diagnosis is what settles it, which the cited source supports directly, so the
  imaging literature was not needed to carry it.

  Prognostic scoring (IELSG and MSKCC) is the one omission that is a genuine
  content gap rather than a scoping decision. The progression section records a
  single outcome phase with population survival and no risk stratification, and
  PMID:29541540 is cached and unused. Worth adding.
📚

References & Deep Research

Deep Research

1

Deep research results are used as seeds for research; they do not undergo the same validation as the main records and may contain errors. How we use deep research.

Evaluations and curation notes (3)

Record notes

Entry level and MONDO binding. This is curated at the WHO entity level as primary CNS lymphoma rather than at the anatomical level of the stub that prompted it. MONDO asserts cerebral lymphoma (MONDO:0003655) as a direct subclass of MONDO:0002571, so the narrower term is recorded as a skos:narrowMatch rather than given its own entry. WHO 2022 places this disease inside "large B-cell lymphomas of immune-privileged sites" alongside primary testicular and primary vitreoretinal large B-cell lymphoma, while the International Consensus Classification keeps it as a standalone entity. The entry follows the ICC here, because the immune-privileged grouping is a shared-mechanism family of the kind this knowledge base models with a Grouping rather than by merging entries. A grouping record for the three would be a reasonable follow-up. Ontology corrections to the deep-research report. The report (claude_code) was strong on content — 32 of 32 citations resolved with no confabulation — and poor on identifiers, and the two are worth separating. It offered at least seven wrong bindings, of which its own term-validation section caught three: - NCIT:C9218, offered as "Primary Central Nervous System Lymphoma", is Stage IV Oropharyngeal Carcinoma AJCC v6. - UBERON:0004087, offered as "vitreous body", is vena cava. - NCIT:C36044, offered as "Diffuse Large B-Cell Lymphoma", is Grade 3b Malignant Neoplasm, and is obsolete. Four more were caught only by checking each identifier against its cache or ontology by hand, because the validator label-checked 31 of 75 terms and skips gene CURIEs entirely: - hgnc:16412, offered as CARD11, is NLRC4. CARD11 is hgnc:16393. - HP:0000618, offered as "Blurred vision", is Blindness. The correct term, used here, is HP:0000622. - HP:0002153, offered as "Papilledema", is Hyperkalemia. The correct term, used here, is HP:0001085. - HP:0034332, offered as "Focal neurologic deficit", is Cognitive regression. One further trap is worth recording because the tooling actively pointed at it. The report gave CHEBI:9560 for thiotepa, and the validator's obsolescence check reported it as replaced by CHEBI:102166 — which is thiopental, an anaesthetic barbiturate, not the alkylating agent. Following the automated replacement would have bound the wrong drug. Thiotepa is CHEBI:9570, taken from a direct search. No datasets block. PCNSL has published genomic cohorts, but the accessions were not verified in this session and an unverified accession is worse than none. This is an ordinary gap rather than a reasoned exclusion, and is a good follow-up. References fetched but not cited. The deep-research run resolved 42 references into the cache and this entry cites 15 of them. The rest were read and set aside, and the reasons fall into three groups, recorded here so that a fetched cache is not mistaken for research nobody looked at. Several are about systemic DLBCL, Hodgkin lymphoma or primary mediastinal lymphoma rather than this disease, and were retrieved because the report discusses PCNSL against those comparators. PMID:31471540 is the exception that is cited, because it names PCNSL explicitly among the diseases carrying the 9p24.1 lesion. Several cover treatment options this entry does not yet model - lenalidomide, checkpoint inhibitors, CAR-T, and the blood-brain-barrier penetration strategies. They are real and they belong here eventually; the entry currently carries induction, consolidation, radiotherapy and BTK inhibition, and adding a relapsed/refractory arm properly is a follow-up rather than something to bolt on. A few are imaging and radiomics studies supporting the differential against glioblastoma. The differential entry makes the weaker claim that tissue diagnosis is what settles it, which the cited source supports directly, so the imaging literature was not needed to carry it. Prognostic scoring (IELSG and MSKCC) is the one omission that is a genuine content gap rather than a scoping decision. The progression section records a single outcome phase with population survival and no risk stratification, and PMID:29541540 is cached and unused. Worth adding.

Address review findings: Primary CNS Lymphoma · 2026-09-09T13:43:57Z · View source

Addressed all four blocking findings and three of five suggestions from the review of PR #11513, in a single push. Each finding was verified against the primary source before being acted on; all four were correct. Finding 1, the EBV branch asserted a mechanism with no supporting evidence. The 'Loss of EBV-Specific T-Cell Surveillance' node claimed that failure of EBV-specific T-cell control permits outgrowth of an EBV-transformed clone, and cited a PMID:37322012 sentence about distinguishing PCNSL from secondary CNS lymphoma - a nosology statement that says nothing about EBV, immunosuppression or T cells. The explanation was candid that it supported only category separation, but the node still made an unsupported causal claim. Confirmed the reviewer's observation independently: grep over the 42 references this PR adds returns no mention of EBV or Epstein-Barr in any of them. Fetched PMID:36960939, a 309-patient multicentre cohort of EBV-positive CNS lymphoma in older patients, and cited two sentences from it giving the EBV-positive fraction, the transplant-associated majority within it, and the correlation with immunosuppressive treatment. Added the PMID:37322012 immunosuppression sentence the reviewer identified as a third item. Finding 2, the commonest presentation was missing. Focal neurological deficit is the leading presentation and was absent from the phenotype list. The reason it was dropped is recorded in the entry notes: the deep-research report offered HP:0034332 for it, which is Cognitive regression. Catching that was right and dropping the phenotype instead of finding the correct term was not. Added Hemiparesis (HP:0001269) and Aphasia (HP:0002381), both labels verified against ols:hp, as the two common forms of the focal presentation, with downstream edges from the mass-effect node and the PMID:37322012 sentence naming focal deficit as the most common presentation. No frequency values were added: the frequency table the reviewer pointed at lives in the deep-research artifact rather than in a citable cached source, and a frequency without a quotable source is exactly what the earlier ZAP70 review flagged. Finding 3, the immune-escape genes were named in prose and absent from the structured section. The entry calls this arm the genetically distinctive feature of the disease and names HLA at 6p21, B2M and CD274/PDCD1LG2 at 9p24.1 in three places, while genetic: carried only MYD88, CD79B, CDKN2A and PIM1. Added records for B2M (hgnc:914), CD274 (hgnc:17635) and CARD11 (hgnc:16393), all CAUSATIVE/SOMATIC like the existing four, and cited PMID:31471540 for the 9p24.1 lesion and PMID:28619981 for the CARD11 ibrutinib-resistance mutation. HLA is deliberately not given a gene record: it is a locus spanning several genes rather than one gene, and the entry's existing prose and the PMID:36971477 snippet already carry the claim. Finding 4, 32 of 42 fetched caches were uncited. Now 15 of 42 are cited, and the notes carry a paragraph recording what was set aside and why, in three groups: references about systemic DLBCL, Hodgkin or primary mediastinal lymphoma retrieved as comparators rather than as evidence about this disease; treatment options the entry does not yet model (lenalidomide, checkpoint inhibitors, CAR-T, blood-brain-barrier strategies), which belong here but need a relapsed/refractory arm built properly rather than bolted on; and imaging/radiomics studies not needed because the differential entry makes the weaker, directly-cited claim that tissue diagnosis settles it. The notes also name the one genuine content gap rather than scoping decision: IELSG and MSKCC prognostic scoring, with PMID:29541540 cached and unused. Suggestions taken. Added a biochemical block for CSF interleukin-10 with the diagnostic performance figures and the prognostic association from PMID:27924864. Added an environmental entry for prolonged iatrogenic immunosuppression with an influences_mechanisms link at environmental_effect: PREDISPOSES onto the EBV node, which connects what had been the least-supported branch of the pathograph to a cited exposure. Its exposure_term is left unbound with the search recorded, on the same reasoning as the ZAP70 BCG entry: ECTO covers named agents and drug classes rather than the state of sustained therapeutic immunosuppression, and ECTO:0000509 is too general to carry the meaning. Added the canine primary CNS B-cell lymphoma animal model (PMID:23115372) with a PARTIALLY_RECAPITULATES link at LOW fidelity, whose limitations state plainly that the case-report evidence carries no molecular characterisation and so cannot support any claim about shared mechanism. Also strengthened the BCR/TLR node, which had rested on a review sentence conceding the pathophysiology is incompletely understood, with the PMID:28619981 opening statement that BTK links the B-cell antigen receptor and Toll-like receptors with NF-kB. One self-inflicted error caught by validation during this round: the CSF IL-10 reference_title was written from a truncated cache header rather than copied, producing 'Primary Central Nervous System Lymphoma' where the real title reads 'Primary Central Nervous System Large B-cell Lymphoma'. This is the check-reference-titles failure mode described in CLAUDE.md. Both occurrences were corrected by copying the title: field from the cache frontmatter, and check-reference-titles passes. Suggestions not taken, with reasons given in the PR reply: the immune-privileged-sites Grouping (agreed, but it is a new record for three diseases rather than an edit to this entry, so it belongs in its own PR) and prognostic scoring (a real gap, recorded in notes as such rather than added, to keep this push to the review's scope). Validation: just validate and just validate-disorders pass with 57/57 snippets verified, up from 44. check-reference-titles, check-entity-refs, check-causal-targets, check-duplicate-keys, check-qualifier-terms and check-environmental-evidence pass; check-cancer-origin still reports SOMATIC_LESION / OK / CL:0000787.

Create: Primary Central Nervous System Lymphoma · 2026-09-09T12:41:05Z · View source

Created kb/disorders/Primary_Central_Nervous_System_Lymphoma.yaml for MONDO:0002571. Target selection, and a MONDO defect found on the way. The stub queue offered MONDO:0016101 'neurolymphomatosis' as a candidate. It is not curatable and should not be attempted as written: MONDO's label is the human entity, but its definition is 'A transmissible viral disease of birds caused by avian herpesvirus 2', its related synonyms are Marek disease and fowl paralysis, and its asserted parent is viral infectious disease. The xrefs carry both MESH:D000077162 (Neurolymphomatosis, human) and MESH:D008380 (Marek Disease, avian), plus Orphanet:206586 (Neurolymphomatosis, human). Two distinct diseases in different species are merged under one term. Curating it from human literature would have produced an entry describing a different disease from the one the identifier names. Reported to the user; no stub change made in this session. Entity and binding. Curated at the WHO entity level as primary CNS lymphoma rather than at the anatomical level of the stub that prompted it (stubs/Cerebral_Lymphoma.yaml, MONDO:0003655). OAK confirms MONDO:0003655 is a direct rdfs:subClassOf MONDO:0002571, so the narrower term is recorded in mappings.mondo_mappings as skos:narrowMatch, which is one of the two predicates that count as curated coverage and therefore retires the stub properly rather than orphaning it. Deep research: claude_code, as requested by the user. Reference validation came back strong - 32 of 32 citations resolved, confabulation_rate 0.0, 27 of 32 assessed on topic, no unresolved identifiers. Term validation came back weak, and the split between the two is the main lesson of this entry. The report offered at least seven wrong ontology bindings. Its own term-validation section caught three: NCIT:C9218 offered as 'Primary Central Nervous System Lymphoma' is Stage IV Oropharyngeal Carcinoma AJCC v6; UBERON:0004087 offered as 'vitreous body' is vena cava; NCIT:C36044 offered as 'Diffuse Large B-Cell Lymphoma' is Grade 3b Malignant Neoplasm and is also obsolete. Four more were found only by checking each identifier by hand against cache/*/terms.csv or OLS, because the validator label-checked 31 of 75 terms and skips gene CURIEs entirely: hgnc:16412 offered as CARD11 is NLRC4 (CARD11 is hgnc:16393, confirmed against the HGNC REST API); HP:0000618 offered as 'Blurred vision' is Blindness (correct term HP:0000622); HP:0002153 offered as 'Papilledema' is Hyperkalemia (correct term HP:0001085); HP:0034332 offered as 'Focal neurologic deficit' is Cognitive regression. None of the seven was bound. One further case is worth recording because the tooling pointed at the wrong answer rather than merely failing to point at the right one. The report gave CHEBI:9560 for thiotepa. The run's obsolescence check reported that identifier as replaced by CHEBI:102166, which resolves to thiopental - an anaesthetic barbiturate, not the alkylating agent. Following the automated replacement would have bound a different drug. Thiotepa is CHEBI:9570, taken from a direct CHEBI search. All ten gene CURIEs the report supplied were checked individually against HGNC; nine were right. Named Entity Confusion preflight (just preflight-dr) returns SKIP for this entry because MONDO records no causal gene for a somatic cancer, so the gene-identity discriminator cannot run. Fell back to the manual check as the tool instructs: the report's gene census is MYD88=32, CD79B=15, CDKN2A=10, which is the PCNSL signature with no rival disease gene present. Content: 11-node causal chain from the somatic MYD88/CD79B lesion through ligand-independent BCR and TLR signalling, constitutive NF-kB, and a separately-modelled late immune-escape arm (HLA loss, B2M, 9p24.1 PD-L1/PD-L2) to perivascular CNS infiltration, with a parallel EBV branch for the immunodeficiency-associated form. 9 phenotypes, 5 genes, 4 treatments, 2 differential diagnoses, 1 clinical trial (NCT01011920, IELSG32), and a KNOWLEDGE_GAP discussion on the unresolved question of whether transformation happens inside or outside the CNS. Cell of origin derives cleanly as SOMATIC_LESION / OK / CL:0000787. An earlier draft bound both CL:0000236 (B cell) and CL:0000787 (memory B cell) on the origin node, which made just check-cancer-origin report MULTI_ORIGIN_CELL. That would have been an artefact of granularity rather than the genuine lump/split signal the report exists to surface, so a single term is bound and the reason is recorded in the node description. No GeneReviews baseline: searched and confirmed absent, which is expected for a somatic cancer. No datasets block; recorded in notes as an ordinary gap rather than a reasoned exclusion. Validation: just validate and just validate-disorders pass with 44/44 snippets verified. check-entity-refs, check-causal-targets, check-duplicate-keys and check-qualifier-terms pass on the file.

Claude Code ▸
Primary Central Nervous System Lymphoma (PCNSL): Comprehensive Research Report
claude-haiku-4-5-20251001, claude-sonnet-5 67 citations 2026-09-09T12:25:54.857525

Primary Central Nervous System Lymphoma (PCNSL): Comprehensive Research Report

1. Disease Information

Overview. Primary central nervous system lymphoma (PCNSL) is a rare, aggressive extranodal non-Hodgkin lymphoma — in the overwhelming majority of cases a diffuse large B-cell lymphoma (DLBCL) — that arises within and remains confined to the brain parenchyma, spinal cord, leptomeninges, and/or eyes (vitreoretinal compartment) at diagnosis, without evidence of systemic (nodal or extra-CNS) disease (Nature Reviews Disease Primers, 2023, PMID:37322012). In the WHO Classification of Haematolymphoid Tumours 5th edition (WHO-HAEM5, 2022) and subsequent 2024 updates, PCNSL is grouped under the new umbrella entity "large B-cell lymphomas of immune-privileged sites", alongside primary vitreoretinal large B-cell lymphoma (PVRL) and primary testicular large B-cell lymphoma (PTL), reflecting their shared biology of arising in anatomically sequestered, immune-privileged compartments. The International Consensus Classification (ICC) instead retains PCNSL as a distinct, standalone entity.

Key identifiers: - MONDO: MONDO:0002571 - ICD-O-3: 9680/3 (Diffuse large B-cell lymphoma, NOS) - ICD-10-CM: C83.3 (Diffuse large B-cell lymphoma) with site extension for CNS, or historically C71/C72 region codes when classified by anatomic site - ICD-11: 2A81.0Y or 2B33.1 region (extranodal DLBCL of CNS) - MeSH: D016543 (Lymphoma, Non-Hodgkin) combined with central nervous system neoplasm indexing; more specific term "Central Nervous System Neoplasms/lymphoma" - OMIM: No dedicated Mendelian OMIM entry (PCNSL is predominantly a sporadic somatic malignancy, not a single-gene disorder) - Orphanet: ORPHA:56163 (Primary central nervous system lymphoma is not separately Orphanet-coded as a rare disease per se, though PCNSL-related entries exist under lymphoma classifications) - NCIT: NCIT:C9218 (Primary Central Nervous System Lymphoma)

Synonyms/alternative names: Primary CNS lymphoma; PCNSL; microgliomatosis (historical term); reticulum cell sarcoma of brain (obsolete); primary brain lymphoma; primary diffuse large B-cell lymphoma of the CNS (PCNS-DLBCL); primary intraocular lymphoma / primary vitreoretinal lymphoma (PVRL, when eye-restricted).

Data provenance. Most quantitative findings summarized below are derived from aggregated disease-level resources: population-based cancer registries (SEER, CBTRUS), multicenter cohort studies, international consensus/clinical trial data (IELSG, Alliance, HOVON), and molecular genomic profiling studies (whole-exome/whole-genome sequencing cohorts). Individual-patient EHR-level data appear in a minority of biomarker validation studies (e.g., CSF IL-10/MYD88 diagnostic cohorts).

2. Etiology

Disease Causal Factors

PCNSL is fundamentally a somatic, clonal B-cell malignancy driven by acquired (not germline) genetic lesions that constitutively activate B-cell receptor (BCR) and Toll-like receptor (TLR)/MYD88-NF-κB signaling within a B lymphocyte that subsequently homes to, or transforms within, the CNS immune-privileged compartment. There is no single "cause" analogous to a Mendelian disorder; rather, disease arises from an interplay of somatic mutation acquisition, chronic antigenic/autoantigenic BCR stimulation, and — critically — a permissive or impaired local/systemic immune surveillance state.

Genetic Risk Factors

  • Somatic driver mutations (not germline predisposition): Recurrent MYD88 L265P (~50–67% of cases; adapter protein mutation causing constitutive NF-κB activation via IRAK1/4), CD79B mutations (particularly Y196; ~55–63%, most commonly co-occurring with MYD88 L265P), and PIM1 mutations (~55–59%) define the disease's dominant "MCD" genomic subtype (see Section 4 and Section 6).
  • Congenital/inherited immunodeficiency syndromes confer markedly elevated risk of secondary lymphoproliferation including PCNSL: Wiskott-Aldrich syndrome (WAS gene, X-linked), ataxia-telangiectasia (ATM gene), X-linked lymphoproliferative disease (SH2D1A/XIAP), and severe combined immunodeficiency (SCID)/common variable immunodeficiency — patients with these congenital immunodeficiencies carry an estimated ~4% lifetime risk of PCNSL.
  • No common-variant GWAS susceptibility loci for PCNSL specifically have been robustly replicated to date (distinct from the situation for some other B-cell lymphoma subtypes).

Environmental/Non-Genetic Risk Factors

  • Age: Strongest risk correlate in immunocompetent hosts — incidence rises steeply with age, peaking in the 7th–8th decade (median age at diagnosis ≈ 65–67 years).
  • HIV/AIDS infection: The single strongest acquired risk factor. Risk is inversely proportional to CD4+ T-cell count; HIV-associated PCNSL classically occurs at CD4+ counts averaging ~30 cells/µL. Pre-cART era incidence in HIV-infected persons was reported as high as 5,000-fold that of the general population, with rates of ~5 cases per 1,000 person-years (1991–1994), falling to ~0.32 per 1,000 person-years after 1999 with combination antiretroviral therapy (cART) introduction.
  • Solid organ transplantation and iatrogenic immunosuppression: Post-transplant lymphoproliferative disease (PTLD) involving the CNS occurs at a markedly younger median age (~23 years in some series) and is almost universally EBV-driven.
  • Chronic autoimmune disease and immunosuppressive drug therapy (e.g., systemic lupus erythematosus, rheumatoid arthritis on long-term immunosuppressants) increase risk.
  • No confirmed risk factors in immunocompetent individuals without evident immunosuppression. Proposed but unconfirmed/unsupported associations include prior tonsillectomy and oral contraceptive use; heavy mobile-phone use has been proposed but is not evidence-supported (Nature Reviews Disease Primers, PMID:37322012).
  • Sex: Slight male predominance in immunocompetent PCNSL (male:female ≈ 1.2:1 to 1.4:1), a pattern that is markedly amplified in HIV-associated disease (reflecting epidemiology of HIV infection itself) but reversed toward female predominance in isolated primary vitreoretinal lymphoma.

Protective Factors

  • Combination antiretroviral therapy (cART) in HIV-infected individuals is the single best-documented protective/risk-modifying intervention, having driven a >10-fold decline in HIV-associated PCNSL incidence since the mid-1990s.
  • Restoration/normalization of immune competence (e.g., withdrawal or reduction of immunosuppressive regimens in transplant recipients when clinically feasible) reduces risk.
  • No specific genetic protective variants or dietary/lifestyle protective factors have been established for PCNSL.

Gene-Environment Interactions

The central gene-environment interaction in PCNSL pathogenesis is the interplay between host immunosurveillance status (HIV-induced CD4+ depletion, iatrogenic immunosuppression, or congenital immunodeficiency) and EBV infection: virtually all HIV-associated and post-transplant PCNSL cases are EBV-genome positive, with loss of EBV-specific CD4+ T-cell effector function permitting outgrowth of EBV-transformed B-cell clones within the CNS immune-privileged niche. In immunocompetent PCNSL, EBV association is rare (a distinct immunobiological entity has been described for EBV+ PCNSL arising after immunosuppression; Blood 2021, 137(11):1468).

3. Phenotypes

Symptom onset is typically subacute, evolving over days to a few weeks, and diagnosis is frequently delayed (mean interval from symptom onset to diagnosis reported around 35–80 days across series) because of nonspecific or psychiatric presentations.

Phenotype Type Frequency Suggested HPO term
Focal neurological deficit (hemiparesis, aphasia, ataxia, sensory loss, cranial neuropathy) Sign 56–70% of intracranial cases HP:0034332 (Focal neurologic deficit) / HP:0001324 (Muscle weakness) / HP:0002317 (Unsteady gait)
Neuropsychiatric/behavioral change, personality change, cognitive decline Symptom/behavioral 32–43%; cognitive/behavioral abnormalities present at diagnosis in 50–70% overall HP:0000708 (Behavioral abnormality) / HP:0002354 (Memory impairment) / HP:0000750 (Delayed speech and language development n/a — use HP:0031466 personality change if available)
Signs of increased intracranial pressure (headache, nausea, vomiting, papilledema) Symptom/sign Common, variable HP:0002315 (Headache), HP:0002017 (Nausea), HP:0002153 (Papilledema)
Seizures Symptom ~10–15% (lower than gliomas, given deep/periventricular location) HP:0001250 (Seizure)
Visual disturbance (blurry vision, floaters) from vitreoretinal involvement Symptom ~20% at presentation overall; ~15–25% of PCNSL have vitreoretinal involvement, of which ~33% are asymptomatic HP:0000618 (Blurred vision), HP:0011805 (Vitreous floaters/vitritis-related)
Hearing loss Symptom Uncommon, cranial nerve/leptomeningeal disease HP:0000365 (Hearing impairment)
Diplopia (double vision) Symptom Occurs with cranial nerve/brainstem involvement HP:0000651 (Diplopia)
Ataxia (with cerebellar lesions) Sign Location-dependent HP:0001251 (Ataxia)
Myelopathy (weakness, sensory level, bowel/bladder dysfunction) Sign <1% of PCNSL (isolated spinal cord disease) HP:0002355 (Difficulty walking), HP:0002019 (Constipation-adjacent bladder/bowel terms as applicable)
Elevated CSF protein / CSF pleocytosis Laboratory abnormality Common on lumbar puncture HP:0002922 (Increased CSF protein), part of CSF workup

Phenotype characteristics: - Age of onset: Predominantly adult-onset/late-onset; median 65–67 years in immunocompetent disease; markedly younger (median 30s–40s) in HIV-associated disease. - Severity: Highly variable; deep periventricular/basal ganglia/corpus callosum/thalamic involvement carries worse prognosis (an IELSG adverse risk factor). - Progression: Typically progressive without treatment; can be rapidly progressive given aggressive lymphoma biology and high proliferative index. - Anatomic distribution: Intraparenchymal disease predominates (~92%), with solitary lesions in ~65% and multiple lesions in ~35%; supratentorial:infratentorial ratio ≈ 3:1. Leptomeningeal dissemination detected by CSF cytology/flow cytometry in ~15% (up to 0–50% range across studies), without demonstrated independent effect on overall survival in prospective analysis.

Quality of life impact: Cognitive and behavioral impairment at diagnosis substantially affects daily functioning; most patients show clinical improvement with successful treatment, often returning toward pre-diagnosis neurological/cognitive baseline. However, treatment-related neurotoxicity (see Section 11) — particularly following whole-brain radiotherapy (WBRT), especially combined with chemotherapy and in patients >60 years — can cause durable, sometimes progressive decline in cognition, gait, and continence, substantially impairing quality of life independent of disease status.

4. Genetic/Molecular Information

Causal/Driver Genetic Alterations (Somatic)

PCNSL is not caused by a single causal gene in the Mendelian sense; disease arises from a recurrent constellation of somatic mutations converging on BCR/TLR–NF-κB signaling and immune evasion pathways.

Gene (HGNC) Alteration Approx. frequency Functional consequence
MYD88 (hgnc:7562) p.L265P hotspot missense 50–67% Constitutive TLR/IL-1R adapter activation → IRAK1/4 recruitment → NF-κB activation (gain-of-function)
CD79B (hgnc:1699) Y196 (ITAM tyrosine) missense, most commonly co-occurring with MYD88 L265P 48–63% Disrupts BCR internalization/inhibitory signaling → chronic active BCR signaling → NF-κB
PIM1 (hgnc:8986) Missense/frameshift, often within aberrant somatic hypermutation (ASHM) targets 55–59% Serine/threonine kinase; oncogenic cooperator, prosurvival
CDKN2A (hgnc:1787) Focal deletion (9p21.3), mutation, or promoter hypermethylation ~56% (deletion); 28% (mutation) Loss of p16INK4a/p14ARF tumor suppression → deregulated cell cycle, chromosomal instability
CARD11 (hgnc:16412) Gain-of-function mutations (~15% of cases) ~15% Constitutive NF-κB activation independent of upstream BCR signal
BTG1/BTG2 (hgnc:1130/hgnc:1131) Mutation via aberrant somatic hypermutation Variable Loss of antiproliferative function
PAX5, RHOH, IGHV4-34 Recurrent targets of aberrant somatic hypermutation Variable Contributes to genomic instability and possibly self-antigen-reactive BCR repertoire (IGHV4-34 in particular)
B2M (hgnc:914, beta-2-microglobulin) Mutation/LOH (~10%) ~10% Loss of MHC class I surface expression → immune evasion
HLA region genes (6p21.32; HLA-A/-B/-C class I, HLA-DR/-DP/-DQ, TAP1 class II) Copy-neutral LOH, focal homozygous/biallelic deletion Most common recurrent chromosomal abnormality in PCNSL Loss/reduced HLA class I and II surface expression → CD8+/CD4+ T-cell immune evasion in the immune-privileged CNS niche
CD274 (PD-L1)/PDCD1LG2 (PD-L2) (9p24.1) Copy number gain (~67%), structural rearrangement/translocation (~13%) Common Overexpression of PD-L1/PD-L2 → adaptive immune checkpoint-mediated evasion
TP53 (hgnc:11998) Mutation Lower frequency than systemic DLBCL Loss of tumor suppression
BCL6 (hgnc:1001) Translocation Relatively common Germinal-center transcriptional dysregulation
MYC/BCL2 rearrangements Structural Lower frequency than systemic DLBCL Contrasts PCNSL biology from nodal "double-hit" DLBCL
ETV6 (hgnc:3495) Inactivation Documented, pathogenic significance unclear —

LymphGen/Genomic Subtype Classification

The co-occurrence of MYD88 L265P and CD79B mutations places the great majority of PCNSL within the "MCD" (MYD88/CD79B) LymphGen genetic subtype originally defined in systemic DLBCL genomic classification. One study found 59% of PCNSL samples classified as MCD, 17% as "MCD-composite," and 24% as "Other" (LymphGen not classifiable). The MCD subtype in systemic DLBCL is associated with inferior outcomes to standard immunochemotherapy, a pattern that appears to extend to PCNSL.

Molecular Clusters (proposed, Nature Reviews Disease Primers, PMID:37322012)

Four proposed molecular clusters: CS1 (hypermethylated, proliferative), CS2 (hypermethylated, immune-cold), CS3 (meningeal involvement, worst prognosis), CS4 (immune-hot, favorable outcome) — an evolving framework analogous to systemic DLBCL genomic subtyping efforts (LymphGen, cluster-based classifications).

Variant Classification/Pathogenicity

Recurrent hotspot mutations (MYD88 L265P, CD79B Y196) are consistently classified pathogenic/oncogenic drivers in ClinVar-adjacent somatic cancer variant databases (COSMIC) and functional studies. These are somatic, not germline, variants — allele frequency in population germline databases (gnomAD) is essentially zero/not applicable, as they are acquired oncogenic mutations, not inherited polymorphisms.

Epigenetic Information

  • DNA hypermethylation at CDKN2A promoter contributes to p16/p14ARF silencing.
  • Broader hypermethylator phenotypes define at least two of the four proposed molecular clusters (CS1, CS2 above), consistent with epigenetic dysregulation as a major axis of PCNSL biology alongside genetic mutation.

Chromosomal Abnormalities

  • 6p21.32 loss/HLA locus alterations — the single most frequent recurrent chromosomal abnormality in PCNSL, an immune-evasion-associated lesion characteristic of immune-privileged-site lymphomas (also seen in PTL, PVRL).
  • 9p21.3 deletion (CDKN2A locus) — ~56% of tumors.
  • 9p24.1 gain/rearrangement (CD274/PDCD1LG2, PD-L1/PD-L2 locus) — copy number gains in ~67% of cases, translocations in ~13%.
  • 3q12.3 gain (~45% of patients) — drives NFKBIZ overexpression, reinforcing NF-κB pathway activation.
  • Partial uniparental disomy and copy-neutral LOH are recurrently reported across the genome, particularly at 6p (Leukemia 2010, "Chromosomal imbalances and partial uniparental disomies in primary central nervous system lymphoma").

Suggested ontology terms: GENO terms for variant zygosity/type; GO:0007249 (I-kappaB kinase/NF-kappaB signaling); GO:0050853 (B cell receptor signaling pathway); GO:0002224 (toll-like receptor signaling pathway); CHEBI terms for relevant small molecules (see Section 12).

5. Environmental Information

  • Infectious agent — Epstein-Barr virus (EBV, human gammaherpesvirus 4; NCBITaxon:10376): The dominant infectious/environmental driver in immunosuppression-associated PCNSL. Virtually all HIV-associated PCNSL tumors are EBV-genome positive; EBV-driven, immunosuppression-associated PCNSL is now recognized as a distinct immunobiological entity distinguishable from sporadic immunocompetent-host PCNSL (Blood 2021, 137(11):1468). EBV latent membrane proteins and EBNA antigens drive B-cell proliferation and transformation in the setting of failed EBV-specific CD4+ T-cell immunosurveillance.
  • HIV (human immunodeficiency virus; NCBITaxon:11676): Not itself infecting the malignant B-cell clone, but the principal permissive environmental/infectious cofactor via progressive CD4+ T-cell depletion, enabling EBV-driven lymphomagenesis in the CNS.
  • Iatrogenic immunosuppressive drug exposure (calcineurin inhibitors, antimetabolites, corticosteroids in solid-organ or hematopoietic transplant recipients; disease-modifying antirheumatic drugs and biologics in autoimmune disease) — environmental/pharmacologic exposures that recapitulate the immunosuppressed-host EBV-driven pathway (post-transplant lymphoproliferative disease, PTLD, with CNS tropism).
  • Lifestyle/occupational factors: No robustly established lifestyle, dietary, occupational, or toxin exposure has been causally linked to PCNSL in immunocompetent hosts. Proposed associations (tonsillectomy history, oral contraceptive use, mobile phone radiofrequency exposure) remain unconfirmed and are not supported by consistent epidemiological evidence.

6. Mechanism / Pathophysiology

Ordered Causal Chain (Primary/Sporadic Immunocompetent-Host Pathway)

  1. A mature/late germinal-center-exit or post-germinal-center B lymphocyte acquires somatic driver mutations — most commonly MYD88 L265P and/or CD79B ITAM-domain mutations, frequently co-occurring — leading to constitutive, ligand-independent activation of TLR/IL-1R (via MYD88→IRAK1/4→NF-κB) and chronic active B-cell receptor (BCR) signaling.
  2. Constitutive MYD88/CD79B/CARD11 pathway activation, reinforced by recurrent 3q12.3 gain driving NFKBIZ overexpression, leads to sustained canonical NF-κB transcriptional activation, promoting B-cell survival and proliferation (demonstrated directly; PMID:37322012 and related molecular studies).
  3. Aberrant somatic hypermutation (targeting PIM1, PAX5, RHOH, BTG1/2, MYC, and other loci, often via off-target activity of activation-induced cytidine deaminase, AICDA) results in additional cooperating oncogenic hits and progressive genomic instability, compounded by loss of CDKN2A (via 9p21.3 deletion, mutation, or promoter hypermethylation in ~56–83% of cases combined) — leading to unchecked cell-cycle progression (p16INK4a/p14ARF loss) and further chromosomal instability.
  4. The transformed B-cell clone (often expressing self-reactive BCR repertoires, e.g., immunoglobulins recognizing CNS self-antigens such as GRINL1A, ADAP2, BAIAP2, N-hyperglycosylated SAMD14, and neurabin-I) is proposed to be selected for, and/or retained within, the CNS through chronic antigen-driven BCR engagement with CNS-resident self-proteins — an inferred mechanism supporting the "antigen-selection" hypothesis of CNS tropism, though the precise sequence of extra-CNS transformation versus intra-CNS transformation is not fully resolved (some transformation may occur systemically prior to CNS homing, per convergent evidence from PVRL biology).
  5. Concurrently, the malignant clone acquires immune-evasion lesions — most prominently 6p21.32 (HLA locus) copy-neutral LOH or focal deletion (loss of MHC class I/II surface expression), B2M mutation/LOH (~10%, complete HLA class I loss), and 9p24.1 gain/rearrangement driving PD-L1/PD-L2 overexpression — leading to evasion of CD8+/CD4+ T-cell-mediated immunosurveillance, which is especially consequential within the already immune-privileged CNS compartment (blood-brain barrier, absence of conventional lymphatics, reduced baseline antigen-presenting cell trafficking).
  6. Within the CNS parenchyma, malignant B cells characteristically infiltrate in a perivascular, angiocentric pattern, recruited in part via endothelial CXCL12/CXCL9 chemokine gradients engaging CXCR4 on neoplastic B cells, and interacting with a distinctive tumor microenvironment comprising reactive CD4+/CD8+ T cells (CD8+ T cells preferentially infiltrating tumor nests and expressing the exhaustion marker TIM-3), M1-like (CD68+CD163low) and M2-like (CD68+CD163high, PD-L1/TIM-3-expressing, centrally/perinecrotically enriched) tumor-associated macrophages/microglia, and reactive astrocytes — resulting in a locally immunosuppressive, angiogenic (aquaporin-4, galectin-3-expressing endothelium) niche that further shields the tumor from immune clearance and promotes local growth (STAT3-driven autocrine PD-L1/IDO expression in tumor cells and macrophages is one documented amplifying loop).
  7. Progressive perivascular and parenchymal tumor expansion causes local mass effect, vasogenic edema, blood-brain barrier disruption, and infiltrative destruction of adjacent white/gray matter — leading to the clinical phenotypes of focal neurological deficit, cognitive/behavioral change, and (with periventricular/leptomeningeal spread) increased intracranial pressure and CSF pleocytosis/protein elevation.
  8. In a minority of cases, dissemination occurs to the leptomeninges (CSF-borne spread, ~15% CSF-cytology/flow-positive) and/or the vitreoretinal compartment of the eye (~15–25%, another immune-privileged site sharing the blood-ocular barrier), producing multifocal CNS symptoms or visual symptoms (often asymptomatic in up to a third of cases with concomitant ocular disease), and reflecting the broader "immune-privileged site lymphoma" biology shared with PCNSL, PVRL, and primary testicular lymphoma.

Branch: Immunosuppression/EBV-Driven Pathway (HIV, transplant, congenital immunodeficiency)

In parallel to (or substituting for) steps 1–3 above, loss of EBV-specific CD4+ T-cell immunosurveillance (from HIV-mediated CD4+ depletion, pharmacologic immunosuppression, or congenital immunodeficiency) permits outgrowth of an EBV-latently-infected B-cell clone driven by viral oncoproteins (EBNA2, LMP1) rather than requiring the same degree of MYD88/CD79B-driven transformation — this EBV+ immunosuppression-associated PCNSL is now regarded as a biologically and immunologically distinct entity from sporadic immunocompetent-host PCNSL (Blood 2021, 137(11):1468), converging downstream on the same perivascular growth pattern and CNS immune-privilege exploitation described in steps 6–8 above.

Molecular Pathways

  • NF-κB signaling (canonical, via MYD88/CD79B/CARD11/NFKBIZ) — the central oncogenic driver pathway. Suggested GO term: GO:0007249 (I-kappaB kinase/NF-kappaB signaling).
  • BCR signaling pathway — GO:0050853, transduced through BTK (a druggable node targeted by ibrutinib, tirabrutinib, acalabrutinib).
  • TLR/IL-1R-MYD88-IRAK signaling — GO:0002224 (toll-like receptor signaling pathway); GO:0035666 (TRIF-independent TLR4/MYD88 pathway relevant terms as applicable).
  • JAK-STAT pathway — activated downstream of 9p24.1 amplification (JAK2 co-amplified with PD-L1/PD-L2 locus); STAT3 drives autocrine PD-L1/IDO expression.
  • PI3K-AKT-mTOR — implicated as a cooperating survival pathway in DLBCL biology generally, though PCNSL-specific data are less extensive than for MYD88/BCR pathways.

Cellular Processes

Chronic proliferative signaling with impaired apoptosis (via NF-κB-driven pro-survival gene transcription, e.g., BCL2 family members), cell-cycle dysregulation (CDKN2A loss), and active immune evasion (MHC downregulation, checkpoint ligand overexpression) dominate the cellular phenotype. Suggested GO terms: GO:0043065 (positive regulation of apoptotic process — for the "de-regulation" direction, i.e., decreased apoptosis), GO:0000082 (G1/S transition of mitotic cell cycle, dysregulated).

Protein Dysfunction

  • MYD88 L265P: gain-of-function conformational change permitting spontaneous myddosome (MYD88-IRAK4-IRAK1) assembly independent of receptor ligation.
  • CD79B Y196 mutations: disrupt the ITAM tyrosine that normally recruits LYN for BCR downregulation, converting the signal into chronic-active rather than transient BCR signaling.

Tissue Damage Mechanisms

Direct infiltrative destruction, vasogenic edema, blood-brain barrier disruption, and — importantly — treatment-related tissue injury (radiation-induced demyelination, axonal loss, gliosis, and microvascular injury following WBRT) contribute substantially to the disease's overall tissue pathology burden, sometimes exceeding the tumor's own direct damage in long-term survivors (see Section 11).

Molecular Profiling

  • Genomics/WES-WGS: Comprehensive genomic and transcriptomic landscape described in Nature Communications 2022 ("The genomic and transcriptional landscape of primary central nervous system lymphoma") and multiple subsequent LymphGen-subtyping studies (2024 medRxiv preprints on PCNSL genomic subtypes and drug sensitivity).
  • Single-cell/spatial: Emerging single-cell and multiplex immunohistochemistry studies characterize the M1/M2 macrophage-microglia dichotomy and T-cell exhaustion phenotype within the PCNSL tumor microenvironment (Biomarker Research, "Unraveling the immune microenvironment in primary CNS lymphoma," 2026).
  • CSF liquid biopsy/cell-free DNA: Digital droplet PCR (ddPCR) for MYD88 L265P in CSF and even plasma cell-free DNA is an active area of molecular diagnostic development.

Advanced Technologies

Functional genomic screening and in vitro drug-sensitivity profiling of PCNSL biopsies against kinase inhibitors has been reported (2024 medRxiv "Primary Central Nervous System Lymphoma Tumor Biopsies Show Heterogeneity in Gene Expression Profiles, Genetic Subtypes, and in vitro Drug Sensitivity to Kinase Inhibitors").

Suggested Cell Ontology (CL) terms: CL:0000236 (B cell) → transformed/neoplastic B cell; CL:0000980 (plasmablast-adjacent — not typically applicable, PCNSL retains B-cell not plasma-cell phenotype); CL:0000813 (memory T cell)/CL:0000625 (CD8-positive, alpha-beta T cell) for tumor-infiltrating exhausted CD8+ T cells; CL:0000129 (microglial cell); CL:0000235 (macrophage) with M1/M2 polarization qualifiers; CL:0002453 (oligodendrocyte, for demyelination context); CL:0000127 (astrocyte).

7. Anatomical Structures Affected

Organ level: - Primary: Brain (supratentorial > infratentorial, ratio ~3:1) — frontal, temporal, parietal, occipital lobes; deep structures (basal ganglia, thalamus, corpus callosum, periventricular white matter) are characteristic and prognostically adverse sites of involvement. Cerebellum and brainstem less commonly. - Secondary within the CNS axis: Leptomeninges (~0–50%, most series ~15% by CSF cytology/flow), spinal cord (<1%, presenting as subacute myelopathy), and the eye — vitreoretinal compartment (~15–25%, bilateral in the majority of PVRL cases). - Body systems: Nervous system (primary); ophthalmic/visual system (secondary immune-privileged extension); by definition, PCNSL spares other organ systems at diagnosis — systemic (extra-CNS) involvement is exclusionary for the "primary" designation.

Suggested UBERON terms: UBERON:0000955 (brain); UBERON:0002316 (white matter of brain — for periventricular/deep involvement); UBERON:0002037 (cerebellum); UBERON:0002298 (brainstem); UBERON:0002240 (spinal cord); UBERON:0002037/UBERON:0016540 leptomeninges (UBERON:0002360 meninges; more specifically UBERON:0002215 pia mater / arachnoid mater); UBERON:0004087 (vitreous body); UBERON:0000966 (retina).

Tissue and cell level: - Malignant lymphoid infiltrate with characteristic perivascular/angiocentric cuffing pattern around CNS microvasculature, with centroblastic-to-immunoblastic cytomorphology. - Reactive infiltrate: T lymphocytes (CD4+ peritumoral, CD8+ intratumoral), tumor-associated macrophages/microglia (M1/M2 subsets), reactive astrocytes, reactive non-neoplastic B cells. - Suggested CL terms: CL:0000542 (lymphocyte, neoplastic B-cell subtype); CL:0000980-adjacent centroblast/immunoblast morphologic descriptors; CL:0000129 (microglial cell); CL:0000127 (protoplasmic astrocyte).

Subcellular level: Nuclear translocation of NF-κB subunits (RelA/p65) as the central molecular hallmark of pathway activation; MHC class I/II trafficking defects at the endoplasmic reticulum/cell surface (relevant GO Cellular Component: GO:0005634 nucleus for NF-κB translocation; GO:0042612 MHC class I protein complex; GO:0042613 MHC class II protein complex).

Localization: Predominantly bilateral/midline-crossing (deep structures, corpus callosum "butterfly" pattern classically associated with high-grade gliomas but also seen in PCNSL) or multifocal; solitary lesions in ~65% of cases, multiple in ~35%. Vitreoretinal involvement, when present, is usually bilateral.

8. Temporal Development

Onset: - Adult/late-onset disease overwhelmingly; median age at diagnosis ≈ 65–67 years in immunocompetent PCNSL, substantially younger (30s) in HIV-associated disease, and younger still (median ~23 years) in post-transplant PCNSL/PTLD. - Onset pattern: subacute — symptoms evolve typically over days to a few weeks rather than the more chronic, insidious pattern of low-grade gliomas, but less abruptly than a stroke-like acute presentation.

Progression: - Disease course: Aggressive and progressive in the absence of treatment; untreated PCNSL carries a very short survival (historically weeks to a few months). - Staging: PCNSL does not use conventional Ann Arbor nodal staging given its extranodal, CNS-confined nature; instead, extent-of-disease workup (contrast MRI brain ± spine, CSF analysis, slit-lamp ophthalmologic exam, whole-body 18F-FDG PET/CT to exclude occult systemic lymphoma, HIV serology) defines the disease at diagnosis. - Progression rate: Variable but generally rapid without treatment; with treatment, disease course is punctuated by induction response, consolidation, and — in a substantial minority — relapse. - Relapse pattern: 15–25% of patients are primary refractory to HD-MTX-based induction; 25–50% of initial responders relapse. Median overall survival after relapse is markedly short: ~2 months for primary refractory disease and ~3.7 months for those relapsing within the first year of initial response, underscoring the poor prognosis of the relapsed/refractory setting.

Patterns: - Remission: Achievable with modern chemoimmunotherapy induction (complete response rates 17–69% depending on regimen and patient fitness) and further consolidation (autologous stem cell transplant or non-myeloablative chemotherapy), can be durable — 5-year PFS as high as 65–75% in fit patients receiving optimal thiotepa-based ASCT consolidation in clinical trials. - Critical periods/windows: The interval to diagnosis (biopsy) is a critical time-sensitive step given rapid disease evolution and the confounding effect of corticosteroids on both symptoms and diagnostic yield (steroids should be withheld pending biopsy whenever clinically feasible, as they can induce dramatic but transient radiographic and histopathologic response, obscuring diagnosis).

9. Inheritance and Population

Epidemiology

  • Incidence: Rising over recent decades — from ~0.30 per 100,000 to 0.44–0.47 per 100,000 person-years in more recent estimates; a roughly 5-fold increase in the US from 1975–2017 has been reported, driven predominantly by rising incidence among immunocompetent patients >60 years old (in ages 70–79, incidence reaches ~4.32 per 100,000). This trend contrasts with the sharp decline in HIV-associated PCNSL following cART introduction.
  • Prevalence: PCNSL constitutes ~2–4% of all primary CNS tumors, 4–6% of extranodal non-Hodgkin lymphomas, and ~1% of all lymphomas overall — a rare disease by any standard definition.
  • HIV cohort incidence: Declined from ~5 cases/1,000 person-years (1991–1994) to ~0.32/1,000 person-years post-1999 (cART era).

Inheritance Pattern

PCNSL is not a Mendelian/heritable disorder — it is a sporadic acquired somatic malignancy. There is no established autosomal dominant, autosomal recessive, X-linked, or mitochondrial inheritance pattern for sporadic PCNSL itself. The only heritable component relevant to PCNSL risk is indirect: inherited primary immunodeficiency syndromes (Wiskott-Aldrich syndrome — X-linked recessive; ataxia-telangiectasia — autosomal recessive; X-linked lymphoproliferative disease — X-linked recessive) that predispose to secondary/PCNSL-type lymphomas as a downstream consequence of immune dysfunction, not through a direct oncogenic germline variant in the lymphoma itself. - Penetrance/expressivity: Not applicable in the classic Mendelian sense; risk in immunodeficiency syndromes is better framed as a cumulative lifetime probability (~4% in the congenital immunodeficiency group cited above) rather than penetrance of a single variant. - Genetic anticipation, germline mosaicism, founder effects, consanguinity: Not established/applicable for sporadic PCNSL; may be relevant to the underlying primary immunodeficiency syndromes in isolated pedigrees but not to PCNSL as a disease entity per se.

Population Demographics

  • Affected populations: No strong ethnic/racial predisposition reported for sporadic immunocompetent PCNSL, though HIV-associated PCNSL epidemiology tracks with regional HIV prevalence and cART access disparities globally.
  • Geographic distribution: Global; incidence patterns most robustly characterized in US (SEER), European, and East Asian cohorts; HIV-associated PCNSL remains disproportionately prevalent in regions with limited cART access.
  • Sex ratio: Male:female ≈ 1.2:1 in immunocompetent disease overall; strongly male-predominant in HIV-associated disease (tracking HIV epidemiology); PVRL (isolated vitreoretinal presentation) is slightly more common in women.
  • Age distribution: Bimodal-leaning pattern — a smaller, younger peak associated with immunosuppression (HIV, transplant, median 20s–40s) and a larger, dominant peak in immunocompetent elderly patients (median 65–67 years, rising incidence with advancing age through the 70s).

10. Diagnostics

Clinical/Laboratory Tests

  • CSF analysis: Typically shows elevated leukocyte count, normal glucose, elevated protein. Flow cytometry for clonal B-cell population improves diagnostic sensitivity over cytology alone but remains imperfect (positive in a minority even with disease present).
  • CSF biomarkers (molecular diagnostics): Combined MYD88 L265P mutation detection (by allele-specific PCR or digital droplet PCR, ddPCR) plus IL-10 quantification (and IL-10/IL-6 ratio) in CSF is now established as a highly accurate, minimally invasive diagnostic adjunct, particularly valuable when biopsy is high-risk or tissue is non-diagnostic. Using a CSF IL-10 cutoff of 8.2 pg/mL, sensitivity/specificity were 95.5%/96.1%; for an IL-10/IL-6 ratio cutoff of 0.72, sensitivity/specificity were 95.5%/100.0% (Scientific Reports 2016, PMC5141427). In multivariable analysis, MYD88 positivity (OR 71.90) and elevated IL-10 (OR 2.82, P<.0001) were each independently associated with CNS lymphoma. A CSF biomarker panel combining MYD88, IL-10, and CXCL13 has been proposed for further sensitivity gains ("Molecular diagnosis of primary CNS lymphoma in 2024 using MYD88Leu265Pro and IL-10," Lancet Haematology 2024).
  • Vitreous/aqueous humor analysis (for suspected vitreoretinal involvement): cytology, IL-10/IL-6 ratio, and MYD88 L265P testing.

Imaging

  • Gadolinium-enhanced brain MRI is the standard imaging modality — the disease-defining, essential first-line study for suspicion, extent-of-disease mapping, treatment response monitoring, and recurrence detection. Lesions classically show homogeneous, avid contrast enhancement (unlike the heterogeneous ring-enhancement of glioblastoma or abscess), restricted diffusion on DWI (reflecting high cellularity/nuclear-to-cytoplasmic ratio), and relatively lower relative cerebral blood volume (rCBV) on perfusion imaging than high-grade glioma.
  • Advanced/emerging imaging: Diffusion-weighted imaging with ADC quantification, dynamic susceptibility contrast (DSC)-MRI, amino acid/FDG brain PET, dynamic contrast-enhanced (DCE)-MRI, and radiomics approaches are under active investigation to improve non-invasive diagnostic accuracy and differentiate PCNSL from glioblastoma, demyelinating disease, and infection/abscess.
  • Whole-body 18F-FDG PET/CT: Performed as part of staging to exclude occult systemic (extra-CNS) lymphoma, which would reclassify the case as secondary rather than primary CNS lymphoma.

Tissue Diagnosis (Gold Standard)

  • Stereotactic brain biopsy is the diagnostic gold standard for parenchymal disease. Corticosteroids should be withheld before biopsy whenever clinically safe, since they exert a rapid cytolytic effect on lymphoma cells, can produce dramatic (but often transient) radiographic "vanishing tumor" responses in up to half of patients, and confound both radiographic and histopathologic diagnosis.
  • Vitrectomy is the diagnostic gold standard for primary vitreoretinal lymphoma when ocular involvement predominates, sparing brain biopsy in appropriately selected patients; chorioretinal biopsy offers higher diagnostic yield than vitrectomy but is reserved for select cases owing to greater risk of visual sequelae.
  • CSF cytology/flow cytometry can obviate brain biopsy when clearly positive, though sensitivity is limited.

Histopathology/Immunohistochemistry

  • Large lymphocytes in solid sheets with characteristic perivascular, angiocentric infiltration; centroblastic-to-immunoblastic cytomorphology.
  • Immunophenotype: B-cell markers CD20+, CD79a+, PAX5+; late-germinal-center/post-germinal-center markers BCL6+ (60–100%) and MUM1/IRF4+ (90–100%); CD10+ in only 10–20% of cases; non-GCB (ABC-like) phenotype predominates overall (~90% of cases) by Hans algorithm/IHC classification; IgM and IgD are almost always expressed (reflecting failure of immunoglobulin class-switch recombination), while class-switched isotypes are rare.
  • High Ki-67/MIB-1 proliferation index typical of aggressive DLBCL.

Suggested NCIT/SNOMED-adjacent term: NCIT:C36044 (Diffuse Large B-Cell Lymphoma); the CNS-specific entity is best represented via NCIT:C9218.

Genetic/Molecular Testing

  • Targeted mutation panels or NGS for MYD88 L265P, CD79B, PIM1, CDKN2A confirm the disease's characteristic genomic signature and can support diagnosis in ambiguous cases; not required for diagnosis when histopathology/IHC is conclusive but increasingly used both diagnostically (via CSF liquid biopsy, above) and for genomic subtype (MCD) characterization relevant to targeted therapy selection.
  • Whole-exome/whole-genome sequencing has defined the broader genomic landscape (Nature Communications 2022) but is primarily a research/discovery tool rather than routine clinical diagnostic practice at this time.

Clinical Criteria/Differential Diagnosis

Key differentials requiring exclusion include glioblastoma and other high-grade gliomas, demyelinating pseudotumor (tumefactive multiple sclerosis), toxoplasmosis and other CNS infections (especially in HIV-positive patients, where PCNSL and toxoplasmosis are the two leading causes of a ring/homogeneously-enhancing mass and can be difficult to distinguish without biopsy or EBV-PCR/thallium-SPECT adjuncts), sarcoidosis, and metastatic disease.

Screening

No population-based screening program exists for PCNSL in asymptomatic individuals, including in HIV-positive populations, where clinical vigilance for neurological symptoms at low CD4+ counts substitutes for formal screening.

11. Outcome/Prognosis

Prognostic Scoring Systems

  • International Extranodal Lymphoma Study Group (IELSG) score: Five adverse factors — age >60 years, ECOG performance status >1, elevated serum LDH, elevated CSF protein concentration, and involvement of deep brain structures. Two-year survival stratifies sharply by risk group: 80% (0–1 factors), 48% (2–3 factors), and 15% (4–5 factors).
  • Memorial Sloan Kettering Cancer Center (MSKCC) score: Based on age and Karnofsky performance status (KPS). Patients >50 years with KPS <70 have the worst prognosis (median survival ~1.1 years); patients >50 with KPS ≥70 have median survival ~3.2 years.
  • Additional emerging markers include the LDH-to-lymphocyte ratio, proposed to refine stratification within the low/intermediate MSKCC risk groups.

Survival and Mortality

  • With modern chemoimmunotherapy, PCNSL is potentially curable, particularly in patients ≤70 years who tolerate intensive induction plus consolidation.
  • Elderly patients (≥65 years) have markedly worse outcomes: population-based 1-, 5-, and 10-year survival estimated at 33–48%, 13–24%, and 10–13% respectively; fewer than half of elderly patients are alive at 1 year in some series.
  • Fit patients receiving optimal induction + thiotepa-based ASCT consolidation have achieved 5-year PFS ~65% and OS ~79% in leading trials (e.g., IELSG32-derived data), with the PRECIS trial reporting 8-year relapse-free survival of 94% after ASCT versus 48% after WBRT consolidation.

Morbidity/Function and Complications

  • Complications include treatment-related cytopenias/infection (from HD-MTX, HDC-ASCT conditioning), nephrotoxicity from methotrexate, and — most consequentially for long-term quality of life — radiation-induced neurotoxicity: delayed-onset neurocognitive decline, gait disturbance, urinary incontinence, and personality change, particularly with WBRT doses >40 Gy and in patients >60 years. Combined chemoradiotherapy carries a cumulative 5-year neurotoxicity incidence of 25–35%, with associated mortality of ~30% among those affected; median survival after neurotoxicity onset is <1–2 years. Neuroimaging in radiation neurotoxicity shows brain volume loss, cortical/subcortical atrophy, and diffuse white matter injury; autopsy findings include myelin and axonal loss, gliosis, spongiosis, and small/large vessel injury.
  • By contrast, ASCT-based consolidation preserves or improves cognitive function/quality of life relative to the general population in long-term survivors, in direct contrast to the progressive decline seen after WBRT (PRECIS trial data).

Prognostic Biomarkers

IELSG and MSKCC clinical scores remain the mainstay; molecular prognostic markers under investigation include LymphGen/MCD genomic subtype (associated with inferior chemoimmunotherapy response), 6p21.3 CN-LOH/HLA homozygous deletion, and BTG1/ETV6/TP53 mutation status (each associated with significantly shorter PFS/OS in genomic cohort studies).

12. Treatment

Pharmacotherapy — Induction (Fit Patients)

High-dose methotrexate (HD-MTX, 3.5–8 g/m²) is the essential backbone of induction, given its capacity to cross the blood-brain barrier at high dose. NCIT: methotrexate maps to NCIT:C733; treatment_term class NCIT:C15632 (Chemotherapy) or NCIT:C15986 (Pharmacotherapy) as appropriate.

Regimens (increasing intensity): - HD-MTX/cytarabine (AraC) ± rituximab (R-MTX/AraC) — CHEBI: methotrexate CHEBI:44185; cytarabine CHEBI:28680; rituximab is a monoclonal antibody (NCIT:C1454) — therapeutic_agent classification NCIT for biologics. - MATRix regimen (HD-MTX + HD-AraC + rituximab + thiotepa; CHEBI: thiotepa CHEBI:9560) — the IELSG32-validated standard, associated with significantly improved response and survival with modest additional hematologic toxicity; regimen_term candidate NCIT term for named combination protocol where available. - R-MPV (rituximab, HD-MTX, procarbazine [CHEBI:8428], vincristine [CHEBI:75261]). - MT-R (HD-MTX + temozolomide [CHEBI:41332] + rituximab).

Consolidation

  • High-dose chemotherapy with autologous stem-cell transplant (HDC-ASCT), particularly thiotepa-based conditioning, is superior to BEAM conditioning (3-year PFS 75% vs 58%). Randomized trials (IELSG32, PRECIS) show comparable 2-year PFS between ASCT (75–76%) and WBRT consolidation, but markedly better long-term relapse-free survival, cognitive preservation, and quality of life with ASCT (PRECIS 8-year RFS: 94% ASCT vs 48% WBRT).
  • Reduced-dose WBRT (e.g., RTOG1114 protocol) plus chemotherapy improves PFS versus chemotherapy alone (2-year PFS 78% vs 54%) but carries the neurotoxicity burden discussed in Section 11, and is now generally reserved for patients ineligible for ASCT.
  • Non-myeloablative consolidation chemotherapy (e.g., AraC/etoposide [CHEBI:45602], as in the Alliance trial) is an alternative for patients unsuitable for ASCT, though generally inferior PFS compared with ASCT in randomized comparison (2-year PFS 51% vs 73%).
  • NCIT treatment_term for stem-cell transplantation: NCIT:C15431 (Hematopoietic Cell Transplantation) → therapeutic_modality: CELL_THERAPY.

Elderly/Unfit Patients

Attenuated-dose HD-MTX (often ≤3.5 g/m²) with careful monitoring achieves CR rates 17–69%; consolidation typically avoids WBRT given high neurotoxicity risk in this population, favoring maintenance strategies (temozolomide, procarbazine, rituximab, lenalidomide, or ibrutinib maintenance) under active trial investigation (e.g., FIORELLA trial NCT03495960; ALLIANCE A51901 NCT04609046; NCT02313389).

Targeted/Novel Therapeutics (Relapsed/Refractory Disease)

  • BTK inhibitors: Ibrutinib (CHEBI:70925; NCIT:C64176) achieves response rates up to 50–77% in relapsed/refractory PCNSL (10/13 patients responding, including 5 CRs, in a Phase I/II MSKCC trial), though median duration of response is short (<6 months); risk of invasive fungal infection can be mitigated with isavuconazole prophylaxis. Tirabrutinib, a second-generation BTK inhibitor, achieved ORR 67% regardless of MYD88/CD79B/CARD11 mutation status.
  • Immunomodulatory agents: Lenalidomide (CHEBI:63791; NCIT:C29258) and pomalidomide (ORR 48%, median DOR 4.7 months in a Phase I study), used alone or in combination (e.g., R2I: rituximab-lenalidomide-ibrutinib, prospective trial NCT03703167).
  • Immune checkpoint inhibitors: Anti-PD-1 agents nivolumab and pembrolizumab (NCIT:C2185 immunotherapy class) — anecdotal efficacy reported; formal Phase 2 trials (NCT02857426, NCT02779101) completed with results pending/reported in subsequent literature.
  • CAR-T cell therapy: CD19-directed CAR-T (NCT04134117, tisagenlecleucel trial in PCNSL, among others) shows feasibility with some durable responses in small early-phase studies; combination with BTK inhibitors and PD-1 blockade under investigation for synergy in CNS lymphoma.
  • IRAK4 inhibitor: Emavusertib (CA-4948) — Phase I/II trial NCT03328078 in relapsed/refractory PCNSL, targeting the MYD88-driven IRAK signaling node directly.
  • Blood-brain barrier penetration strategies: TNF-alpha/NGR conjugates to permeabilize tumor vasculature; nanoparticle drug delivery; focused ultrasound — early-stage investigational approaches to improve CNS drug penetration.

Surgical/Interventional

Surgery in PCNSL is limited to diagnostic biopsy — cytoreductive resection is not standard of care (unlike glioma), as PCNSL is a chemosensitive, radiosensitive, diffusely infiltrative systemic-type malignancy rather than a mass amenable to surgical cure; resection does not improve outcomes and risks neurological morbidity, though rare exceptions (solitary accessible lesion causing critical mass effect) are individualized.

Supportive/Rehabilitative Care

Corticosteroids (dexamethasone) for symptomatic vasogenic edema (used cautiously pre-biopsy, as above); anticonvulsants for seizures; comprehensive neurocognitive rehabilitation and physical/occupational therapy for treatment-related or disease-related functional deficits (NCIT:C15302 Physical Therapy; NCIT:C121351 Occupational Therapy); best supportive care alone is appropriate for select frail/unfit patients for whom intensive therapy is not feasible.

Treatment Algorithms

Modern management follows an age/fitness-stratified algorithm: (1) fit patients receive HD-MTX-based induction (MATRix or equivalent) followed by consolidation (ASCT preferred over WBRT where feasible); (2) unfit/elderly patients receive attenuated induction with maintenance-based or reduced-intensity consolidation; (3) relapsed/refractory disease is preferentially managed on clinical trial given median OS of only 2–3.7 months with standard salvage approaches, with BTK inhibitors, immunomodulatory agents, checkpoint inhibitors, and CAR-T among emerging options.

13. Prevention

  • Primary prevention: No established primary prevention strategy exists for PCNSL in immunocompetent individuals, as no modifiable environmental or lifestyle risk factor has been robustly confirmed.
  • Immunosuppression management as risk mitigation: For immune-deficient patients, treatment/optimization of the underlying immunosuppression state constitutes the principal preventive lever — most concretely demonstrated by the dramatic decline in HIV-associated PCNSL incidence following widespread combination antiretroviral therapy (cART) adoption (from ~5 to ~0.32 cases per 1,000 person-years).
  • Secondary prevention/screening: No formal population-based or risk-group screening program exists for asymptomatic PCNSL detection, including among HIV-positive or transplant populations; management instead relies on clinical vigilance for neurological symptoms in at-risk groups and prompt neuroimaging/biopsy workup once symptoms arise.
  • Tertiary prevention: Minimizing treatment-related neurotoxicity (preferring ASCT-based consolidation over WBRT where feasible in eligible patients, using reduced-dose WBRT protocols when radiotherapy is required, and structured neurocognitive monitoring) functions as tertiary prevention against long-term disability in survivors.
  • Counseling: Genetic counseling is relevant primarily in the context of the underlying congenital immunodeficiency syndromes (Wiskott-Aldrich, ataxia-telangiectasia, XLP) that predispose to PCNSL, rather than for PCNSL as an inherited entity itself.

14. Other Species / Natural Disease

  • Dogs (Canis lupus familiaris, NCBITaxon:9615): CNS lymphoma occurs in dogs, most often as part of multicentric lymphoma with secondary CNS/leptomeningeal/choroid plexus infiltration (reported in up to ~12–30% of canine lymphoma cases with CNS involvement), though isolated primary CNS B-cell lymphoma in a young dog has also been reported (PMC3327599). Mean age of onset ~7.4 years (range 3–11); no strong breed predisposition overall, though Rottweilers show a slightly elevated risk. Forebrain is the most frequent site.
  • Cats (Felis catus, NCBITaxon:9685): CNS lymphoma is less common than in dogs, accounting for <3% of primary CNS tumors in cats; forebrain meninges (B-cell phenotype) is the most frequent site, while spinal cord lymphoma is more associated with younger cats.
  • Comparative pathology: Veterinary CNS lymphoma studies (MDPI Animals 2023, "Neuropathology of Central and Peripheral Nervous System Lymphoma in Dogs and Cats," 92-case series) demonstrate broadly analogous perivascular/leptomeningeal infiltrative patterns to human PCNSL, though most veterinary cases represent secondary/multicentric rather than strictly primary disease, limiting direct translational equivalence to human isolated PCNSL biology.
  • Orthologous genes: MYD88, CD79B, CDKN2A, and TP53 orthologs are well conserved across mammals (canine MYD88 ortholog, NCBI Gene; feline orthologs similarly annotated), supporting biological plausibility of shared BCR/TLR-NF-κB pathway dysregulation, though specific somatic mutation profiling of veterinary CNS lymphoma is not yet as well characterized as in human disease.
  • Zoonotic potential: None — PCNSL is a non-communicable malignancy; EBV itself is not classically zoonotic in this context (human-restricted gammaherpesvirus).

15. Model Organisms

Patient-Derived Xenograft (PDOX) Models

The most biologically faithful available models are orthotopic patient-derived xenografts: PCNSL patient specimens grafted into the caudate nucleus of immunodeficient nude mice achieve an ~83% engraftment success rate, whereas subcutaneous implantation fails to generate tumors — direct experimental evidence for the essential role of the brain microenvironment in PCNSL pathophysiology (Neuro-Oncology/ScienceDirect, "Primary CNS lymphoma patient-derived orthotopic xenograft model"). PDOX models recapitulate diffuse B-cell infiltration of brain parenchyma and preserve each patient's unique BCR/NF-κB pathway mutational signature, supporting their use for precision-oncology drug-sensitivity testing.

Cell-Line-Derived Xenograft Models

Earlier xenograft models (e.g., PMID for "A new xenograft model of primary central nervous system lymphoma," 1999) established feasibility of intracerebral engraftment using lymphoma cell lines in athymic/nude mice, with tumor growth monitored via bioluminescence imaging in modern iterations.

Rat Models

A nude-rat PCNSL model has been reported to reproduce both the histology and characteristic anatomic location of human CNS lymphoma, offering a larger-animal platform for some interventional/imaging studies.

Model Characteristics, Applications, and Limitations

  • Phenotype recapitulation: Orthotopic (intracerebral) engraftment is essential — subcutaneous models do not reproduce disease, underscoring that the CNS/brain microenvironment (immune privilege, blood-brain barrier, local chemokine milieu) is not merely a permissive site but an active contributor to PCNSL biology and must be modeled directly.
  • Applications: PDOX models support (a) validation of the perivascular/angiocentric growth pattern, (b) preclinical testing of blood-brain-barrier-penetrant agents (HD-MTX, novel BTK inhibitors, nanoparticle-based delivery), and (c) patient-specific drug-sensitivity screening reflecting individual BCR/NF-κB mutational profiles.
  • Limitations: Immunodeficient host models (nude/NOD-SCID mice) cannot recapitulate the tumor-immune microenvironment interactions (T-cell exhaustion, macrophage/microglia polarization, checkpoint biology) that are increasingly recognized as central to PCNSL pathophysiology and to immunotherapy response — a significant translational gap for checkpoint-inhibitor and CAR-T preclinical development, which more plausibly requires humanized or syngeneic immunocompetent models (not yet well established for PCNSL specifically in the literature surveyed here).
  • Resources: No PCNSL-specific standardized repository was identified analogous to MGI/IMPC germline knockout collections (consistent with PCNSL's somatic, non-heritable biology); available models are largely investigator-generated PDX/PDOX lines maintained at individual research centers rather than centrally cataloged.

Summary of Key Ontology Term Suggestions

Category Suggested terms
Disease identity MONDO:0002571; NCIT:C9218; ICD-O-3 9680/3
Causal genes MYD88 (hgnc:7562), CD79B (hgnc:1699), PIM1 (hgnc:8986), CDKN2A (hgnc:1787), CARD11 (hgnc:16412), B2M (hgnc:914), TP53 (hgnc:11998), BCL6 (hgnc:1001)
Biological processes (GO) GO:0007249 (NF-κB signaling), GO:0050853 (BCR signaling), GO:0002224 (TLR signaling)
Cell types (CL) CL:0000236 (B cell, neoplastic), CL:0000129 (microglial cell), CL:0000235 (macrophage), CL:0000625 (CD8+ T cell), CL:0000127 (astrocyte)
Anatomy (UBERON) UBERON:0000955 (brain), UBERON:0002240 (spinal cord), UBERON:0002360 (meninges), UBERON:0004087 (vitreous body), UBERON:0000966 (retina)
Phenotypes (HP) HP:0034332 (focal neurologic deficit), HP:0000708 (behavioral abnormality), HP:0002315 (headache), HP:0001250 (seizure), HP:0000618 (blurred vision), HP:0002922 (increased CSF protein)
Chemicals (CHEBI) CHEBI:44185 (methotrexate), CHEBI:28680 (cytarabine), CHEBI:9560 (thiotepa), CHEBI:70925 (ibrutinib), CHEBI:63791 (lenalidomide), CHEBI:41332 (temozolomide)
Treatments (NCIT) NCIT:C15632 (Chemotherapy), NCIT:C15431 (Hematopoietic Cell Transplantation), NCIT:C15313 (Radiation Therapy), NCIT:C2185 (Immunotherapy)

Notes on Evidence Gaps

  • Precise PMIDs were not retrievable for every source cited in this search-tool-mediated survey; several citations are identified by PMCID, journal/year, or DOI where PMID was not directly surfaced in search snippets — a curator populating a formal knowledge-base entry should re-verify each citation against PubMed directly before use (per this repository's evidence-sourcing discipline, every CURIE and citation must be read from a source at write time, not reconstructed from this report).
  • Germline/heritable genetic risk data for PCNSL specifically (as opposed to the associated congenital immunodeficiency syndromes) is limited; no GWAS-identified common-variant susceptibility loci were found.
  • Standardized, centrally cataloged genetic mouse models (knockout/knock-in/conditional) specific to PCNSL were not identified in this search; the field relies primarily on xenograft/PDOX approaches given the disease's somatic rather than germline genetic basis.

Sources

Reference Validation

Checked with linkml-reference-validator 0.2.1.

Outcome Count
References checked 32
Resolved 32
Unresolved (possible confabulation) 0
Unverifiable 0
References weighed for topical relevance 32
On topic 27
Off topic 0

All extracted references resolved successfully.

Term Validation

Checked with linkml-term-validator 0.4.5, through the ols: adapter.

Outcome Count
Terms checked 75
Resolved 70
Unresolved (possible confabulation) 0
Obsolete 4
Unverifiable 1
Terms whose name was checked 31
Terms named correctly 18
Terms named as a different term 6
Terms whose name is worth a second look 7

Terms the report names something else

These identifiers resolve, so nothing about them looks wrong, and the ontology calls them something unrelated to what the report calls them. That usually means the identifier is not the one the sentence needs:

  • NCIT:C9218 (3 mentions) - the report calls it "Primary Central Nervous System Lymphoma"; NCIT calls it Stage IV Oropharyngeal Carcinoma AJCC v6
  • CL:0000980 (2 mentions) - the report calls it "plasmablast-adjacent — not typically applicable, PCNSL retains B-cell not plasma-cell phenotype"; CL calls it plasmablast
  • UBERON:0002316 (1 mention) - the report calls it "white matter of brain — for periventricular/deep involvement"; UBERON calls it white matter
  • UBERON:0004087 (2 mentions) - the report calls it "vitreous body"; UBERON calls it vena cava
  • CL:0000542 (1 mention) - the report calls it "lymphocyte, neoplastic B-cell subtype"; CL calls it lymphocyte
  • NCIT:C36044 (1 mention) - the report calls it "Diffuse Large B-Cell Lymphoma"; NCIT calls it Grade 3b Malignant Neoplasm

Obsolete terms

These terms are real but deprecated. Citing one is not a fabrication; it does mean the report is naming something the ontology has retired:

  • HP:0002355 (obsolete Difficulty walking) (1 mention) - replaced by HP:0001288
  • NCIT:C36044 (Grade 3b Malignant Neoplasm) (1 mention)
  • CHEBI:9560 (CHEBI_9560) (2 mentions) - replaced by CHEBI:102166
  • CHEBI:45602 (CHEBI_45602) (1 mention) - replaced by CHEBI:45605

Terms whose name is worth a second look

The report's name for these is recognisably related to the term's own name without being one of them. A loose paraphrase reads the same way as a citation of the wrong sibling term - and so does a related synonym, which the ontology records precisely because it names something adjacent rather than the same thing - so these are listed rather than judged:

  • HP:0002922 (2 mentions) - the report calls it "Increased CSF protein"; HP calls it Increased CSF protein concentration, and lists "Increased CSF protein" among its other names
  • GO:0007249 (3 mentions) - the report calls it "I-kappaB kinase/NF-kappaB signaling"; GO calls it canonical NF-kappaB signal transduction, and lists "I-kappaB kinase/NF-kappaB signaling" among its other names
  • GO:0035666 (1 mention) - the report calls it "TRIF-independent TLR4/MYD88 pathway relevant terms as applicable"; GO calls it TRIF-dependent toll-like receptor signaling pathway, and lists "TRIF-dependent TLR signaling pathway" among its other names
  • GO:0043065 (1 mention) - the report calls it "positive regulation of apoptotic process — for the "de-regulation" direction"; GO calls it positive regulation of apoptotic process
  • GO:0000082 (1 mention) - the report calls it "G1/S transition of mitotic cell cycle, dysregulated"; GO calls it G1/S transition of mitotic cell cycle
  • CL:0002453 (1 mention) - the report calls it "oligodendrocyte, for demyelination context"; CL calls it oligodendrocyte precursor cell
  • CL:0000127 (3 mentions) - the report calls it "astrocyte", "protoplasmic astrocyte"; CL calls it astrocyte

Terms named inconsistently

The report gives these identifiers more than one name of its own:

  • CL:0000127 - called "astrocyte", "protoplasmic astrocyte"

Prefixes with no resolver

Terms carrying these prefixes were not checked either way, because no configured ontology covers them. An unrecognised prefix may name an ontology this run could not reach as easily as one that does not exist, so nothing here is evidence of fabrication: ORPHA.