Chronic Beryllium Disease — Comprehensive Research Report

1. Disease Information

Chronic beryllium disease (CBD, "berylliosis") is a granulomatous interstitial lung disease driven by a beryllium-specific, cell-mediated (delayed-type, Type IV) immune response to inhaled beryllium. Think of it as the immune system deciding a chunk of metal is a permanent enemy it can't digest and building little fortified walls (granulomas) around it forever — a grudge that slowly stiffens the lung. It's the textbook gene-environment disease: you need both the exposure and the right immune-presentation genotype.

Key identifiers: - MONDO: MONDO:0015274 (chronic beryllium disease) — already in the entry. - MeSH: D001607 ("Berylliosis"). - ICD-10: J63.2 (Berylliosis). - ICD-11: under occupational/inorganic-dust pneumoconioses (CA60.1 region) — verify the exact leaf against the current ICD-11 browser; codings have shifted. - OMIM: No Mendelian OMIM entry — CBD is a complex susceptibility trait, not a single-gene disease. The relevant genetics live in HLA-DPB1 allele associations, not an OMIM disease number. - Orphanet: A rare-disease entry exists for berylliosis; confirm the ORPHA number in the current Orphadata pull rather than trusting memory.

Synonyms: berylliosis, chronic berylliosis, chronic beryllium lung disease, beryllium granulomatosis. (The entry's synonym list is good.)

Data derivation: Disease-level, aggregated. The knowledge here comes from occupational cohort studies (nuclear-weapons, aerospace, machining, ceramics workers), medical-surveillance registries (BeLPT screening programs at DOE facilities), immunology/structural-biology studies, and the 2014 ATS official statement — not from a single-patient EHR source.

2. Etiology

Primary cause (environmental): Inhalation of airborne beryllium — dust, fume, or fine particulate — is the sole necessary exposure. No beryllium, no CBD. It's an occupational/environmental toxicant acting as an immunogen (CHEBI: beryllium / beryllium(2+) cation; ECTO: exposure to beryllium). The entry correctly lists the at-risk industries: beryllium ore extraction and metal machining, aerospace/defense, nuclear-weapons and reactor work, electronics and semiconductors, dental-alloy and ceramics fabrication, and metal recycling.

Crucially, and unlike the classic fibrogenic pneumoconioses (silicosis, asbestosis), risk is not simply cumulative-dose-dependent. Even brief or low-level exposures can sensitize a genetically susceptible person; particle size, solubility, and surface area strongly modulate immunogenicity. Take-home ("paraoccupational") exposure on work clothes and near-facility community exposure are documented.

Genetic risk factors: - HLA-DPB1 Glu69 is the dominant heritable risk factor. Alleles encoding glutamic acid at position 69 of the DPβ chain (HLA-DPB102:01, 17:01, 06:01, etc.) markedly raise risk of both sensitization and disease. Richeldi et al. (Science 1993, PMID:8105536) — the founding observation: "97% [of CBD cases] expressed the HLA-DPB1*0201-associated glutamic acid (unaffected population, 30%; P < 0.001) at residue 69." - Higher-affinity Glu69 alleles (notably *17:01) and gene copy number / expression level further stratify risk (see McCanlies, Silveira, and the E69 genotype-exposure work, e.g. PMC8760148). - A minority susceptibility route via HLA-DR (glutamate at DRβ position 71, and an HLA-DRPheβ47 marker in Glu69-negative individuals; PMC1198259*) exists for the ~15–20% of patients lacking DP-Glu69.

Environmental/host modifiers: Higher airborne concentration, respirable particle fraction, and soluble beryllium salts increase risk; smoking is not a clear risk factor for CBD itself but suppresses BeLPT responses (see §10). Age, sex, and family history are not established independent risk factors beyond the HLA genetics.

Protective factors: Genetically, absence of a Glu69 allele is the main "protective" state — DP molecules lacking Glu69 cannot coordinate Be²⁺ and don't present it (mechanistic basis in §6). Environmentally, protection is entirely about exposure reduction: enclosed processes, wet machining, respiratory protection, and the lowered OSHA permissible exposure limit of 0.2 µg/m³ (8-hr TWA) under the 2017 final rule (PMID:28071878). No dietary/nutritional protective factor is established.

Gene–environment interaction: CBD is arguably the cleanest human example of a defined HLA-restricted gene × environment interaction — a specific MHC-II pocket residue (Glu69) is required to convert an inhaled metal into a T-cell antigen. Neither factor alone produces disease.

3. Phenotypes

CBD is insidious and adult-onset (occupational latency of years to decades). Severity and progression are variable. Suggested HP terms (the entry already carries most of these):

Additional phenotypes worth considering for completeness: exercise-induced hypoxemia / reduced gas exchange (often the earliest physiologic abnormality, sometimes preceding resting PFT changes — cardiopulmonary exercise testing catches it first), skin granulomas / beryllium ulcers at sites of dermal beryllium implantation (a documented extrapulmonary manifestation), and in end-stage disease cor pulmonale / right heart failure (HP:0001648-ish pulmonary hypertension secondary to fibrosis). Beryllium can also produce contact dermatitis. Hepatic and other systemic granulomas are rare.

Onset/severity/progression: adult-onset; severity mild→severe and variable; course insidious then chronic-progressive in the subset that advances; frequency among affected individuals for the core respiratory phenotypes is high but the entry appropriately omits precise frequency: bands where quantitative support is thin.

Quality-of-life impact: progressive exertional limitation, oxygen dependence in advanced disease, and lifelong therapy burden. No CBD-specific validated QOL instrument is standard; generic respiratory/ILD tools (SF-36, St. George's Respiratory Questionnaire) are used in studies — flag as not-CBD-specific.

4. Genetic / Molecular Information

• Susceptibility gene: HLA-DPB1 (HGNC — the entry uses hgnc:4940; verify this HGNC number resolves to HLA-DPB1, as HLA HGNC IDs are easy to transpose). Also HLA-DPA1 (the paired α chain), and HLA-DRB1 for the minority DR-restricted route.
• Risk "variants": these are classical HLA alleles / polymorphic residues, not ClinVar-style pathogenic point mutations. The functional unit is the codon-69 residue of DPB1 (Glu69 = risk; non-Glu69 = low risk). This is a germline susceptibility polymorphism, not somatic, and not "pathogenic" in the ACMG/AMP sense — it's an immune-response allele. So most of the ClinVar/gnomAD/COSMIC machinery in the template is N/A; the right resources are the IMGT/HLA database and HLA allele-frequency references.
• Functional consequence: gain of an antigen-presentation function — Glu69 creates an acidic pocket that binds Be²⁺. Mechanistically it's a "toxic gain of presentation," not loss-of-function.
• Dose/copy effect: susceptibility scales with Glu69 allele copy number and DP surface expression (homozygotes and high-expressing haplotypes at higher risk).
• Modifier genes: candidate immunogenetic modifiers include TNF-α promoter polymorphisms (−308), TGF-β1, and other cytokine variants associated with disease severity/progression in some cohorts, though replication is uneven — cite cautiously.
• Epigenetics / chromosomal abnormalities: No established disease-defining DNA-methylation signature, and no aneuploidy/translocation involvement. Treat as not applicable / not established.

5. Environmental Information

• Environmental factor: beryllium and beryllium-containing compounds (metal, oxide, alloys such as copper-beryllium, ceramics) as respirable dust/fume. CTD is a good structured source for beryllium–gene interactions.
• Lifestyle: No causal lifestyle factor. Smoking is a notable confounder — cigarette smoke suppresses lymphocyte proliferation and can cause false-negative BeLPT, so it affects detection more than causation.
• Infectious agents: None. CBD is non-infectious. (This is a key contrast with tuberculosis, whose caseating granulomas it superficially resembles.)

6. Mechanism / Pathophysiology

This is the heart of the entry and it's already modeled cleanly against the granuloma_formation module. The causal chain:

Step 1 — Deposition & uptake (trigger). Inhaled beryllium particles deposit in distal airways/alveoli and are engulfed by alveolar macrophages (CL:0000583) and dendritic cells (CL:0000451). Beryllium is poorly soluble and biopersistent — the macrophage cannot degrade or clear it, so it becomes the "persistent indigestible stimulus" that converts a normal, self-limited macrophage response into a chronic one. (Conforms to granuloma_formation#Persistent Indigestible Stimulus.) GO: chronic inflammatory response (GO:0002544).

Step 2 — Neoantigen formation (mechanism). Ionic Be²⁺ is incorporated into the peptide-binding groove of HLA-DP, coordinated by acidic residues of the DPβ chain and a bound self-peptide. This is the structural punchline from Clayton et al. (Cell 2014, PMID:24995984): "the T cell ligand is created when a Be2+ cation becomes buried in an HLA-DP2/peptide complex, where it is coordinated by both MHC and peptide acidic amino acids… the TCR does not interact with the Be2+ itself, but rather with surface changes induced by the firmly bound Be2+." So beryllium is not a classical covalent hapten — it's buried at the MHC–peptide interface, remodeling the surface into a neoantigen. This is what the paper means by "bridging allergic hypersensitivity and autoimmunity." GO: antigen processing and presentation via MHC class II (GO:0002495).

Step 3 — Glu69-restricted presentation (mechanism). Presentation is restricted to DP molecules bearing Glu69 (the acidic residue that coordinates the cation). Amicosante/Fontenot (PNAS 2000, PMID:11050177): beryllium presentation to CD4⁺ T cells "underlies disease-susceptibility HLA-DP alleles." Bill et al. (J Immunol 2005, PMID:16272364) pinned it to a single β-chain residue — "beryllium presentation… is dependent on a single amino acid residue of the MHC class II beta-chain" (Glu69 on DP, the homologous Glu71 on DR). The crystal structure of HLA-DP2 (PNAS 2010, PMID:20356827) showed the unique solvent-exposed acidic pocket.

Step 4 — CD4⁺ T-cell sensitization (mechanism). TCR recognition drives clonal expansion of beryllium-specific CD4⁺ memory/effector T cells (CL:0000545 T-helper 1 cell as the closest CL anchor). This expanded population is beryllium sensitization (BeS) — the state that precedes organ disease and is measured by the BeLPT. GO: T cell proliferation (GO:0042098). Sensitization is necessary but not sufficient for CBD.

Step 5 — Compartmentalized Th1 cytokine response (amplifier). On re-encounter in the lung, these cells polarize to Th1 and secrete IFN-γ, TNF, and IL-2; they accumulate as effector-memory Th1 cells in the bronchoalveolar space (Fontenot, JCI 2002 — target-organ localization of memory CD4⁺ T cells; Fontenot & Maier review, PMID:18317020). GO: type II interferon production (GO:0032609), tumor necrosis factor production (GO:0032640).

Step 6 — Macrophage recruitment & activation (amplifier). IFN-γ + TNF classically activate lung macrophages and recruit monocytes; TNF is the non-redundant organizer of the granulomatous response. (Conforms to granuloma_formation#Th1 and TNF-Driven Macrophage Recruitment and Activation.) GO: macrophage activation (GO:0042116).

Step 7 — Epithelioid/giant-cell transformation (central effector). Macrophages become epithelioid cells (CL:0002150) and fuse into multinucleated giant cells (CL:0000647) — the asteroid- and Schaumann-body-containing giant cells shared with sarcoid. GO: syncytium formation by cell-cell fusion (GO:0000768).

Step 8 — Non-caseating granuloma assembly (effector). Compact non-caseating granulomas — epithelioid/giant-cell core cuffed by CD4⁺ T cells — form along bronchovascular bundles, interlobular septa, and hilar nodes. No central caseous necrosis, which is what makes CBD histologically indistinguishable from sarcoidosis. (Conforms to granuloma_formation#Organized Granuloma Assembly.)

Step 9 — Progressive fibrosis (consequence). Because the stimulus can't be cleared, the response persists and, in a subset, drives interstitial fibrosis — fibroblast (CL:0000057) recruitment and collagen deposition (GO:0030199) → restrictive defect, impaired gas exchange, respiratory failure, cor pulmonale. (Conforms to granuloma_formation#Tissue Containment versus Destruction and Fibrosis.)

Regulatory arm worth noting: Regulatory T cells modulate granuloma intensity in the HLA-DP2 mouse model (PMID:24912188) — a mechanistic knob that partly explains why only some sensitized people progress.

Molecular profiling: BAL from CBD patients shows a compartmentalized Th1 signature (high IFN-γ, TNF, IL-2; oligoclonal TCR expansion, e.g. Vβ-restricted repertoires). No routine metabolomic/lipidomic/proteomic diagnostic signature is established; single-cell/BAL immunophenotyping is a research tool, not clinical.

7. Anatomical Structures Affected

• Primary organ: lung (UBERON:0002048), specifically the pulmonary interstitium, alveoli/alveolar wall (UBERON:0002299 alveolus), distal airways, and bronchovascular bundles.
• Regional lymphatics: hilar and mediastinal lymph nodes (UBERON:0002509 mesenteric-node is wrong — use thoracic/hilar lymph node terms; UBERON:0002509 not applicable; consider UBERON:0000029 lymph node with a mediastinal/hilar qualifier).
• Body system: respiratory (with secondary cardiovascular involvement — pulmonary hypertension/cor pulmonale in advanced fibrosis).
• Secondary/extrapulmonary: skin (granulomas, ulcers, contact dermatitis; UBERON:0002097), and rarely liver, spleen, myocardium, salivary glands, and other sites — systemic granulomatosis is uncommon but reported.
• Tissue/cell level: epithelioid and multinucleated giant-cell macrophages, CD4⁺ Th1 lymphocytes, fibroblasts/myofibroblasts, alveolar macrophages, dendritic cells (CL terms as above).
• Subcellular: antigen presentation is at the plasma-membrane MHC-II complex (GO:0042613 MHC class II protein complex); phagolysosomal handling of biopersistent particle in macrophages (GO:0005764 lysosome). No mitochondrial/ER-specific compartment is disease-defining.
• Laterality: bilateral, diffuse, typically upper-and-mid-zone-predominant on imaging (like sarcoid), often with an upper-lobe fibrotic bias in advanced disease.

8. Temporal Development

• Onset: adult; occupational. Latency is long and variable — months to >20–40 years after first exposure, with cases documented decades after exposure ceased.
• Onset pattern: insidious/chronic. Acute beryllium disease (a distinct, high-dose chemical pneumonitis) is now essentially historical and mechanistically different — worth explicitly distinguishing from CBD in the entry if not already.
• Stages: subclinical sensitization (BeS) → early granulomatous CBD (often asymptomatic, abnormal biopsy/BAL) → symptomatic granulomatous disease → fibrotic end-stage.
• Progression rate: variable; BeS → CBD conversion runs roughly 6–8%/year in the Newman longitudinal cohort (PMID:15374840: 55 sensitized workers, mean follow-up 4.8 yr; ~31% developed CBD; the remaining ~69% stayed sensitized without disease), and a systematic review put progression at ~3.2–9.2%/year (PMID:22705916). Roughly half of sensitized individuals already have CBD at their first thorough evaluation.
• Course: chronic, lifelong. Once fibrosis is established it is irreversible — corticosteroids don't reverse scar.
• Remission: no spontaneous cure; treatment can stabilize/partially improve inflammation but disease "recrudesces with reduction of the corticosteroid dose," so relapse on tapering is characteristic.
• Critical window: exposure cessation before fibrosis is the key intervention window; early identification via surveillance BeLPT is the point of maximum leverage.

9. Inheritance and Population

• Epidemiology: CBD is uncommon and occupational. Cross-sectional prevalences among exposed worker cohorts: beryllium sensitization ~0.8–12% and CBD ~0.1–8%, depending on job/exposure intensity (machinists and ceramics workers highest; lower-exposure nuclear R&D sites ~2–3% sensitization). Population-level prevalence in the general (non-exposed) public is effectively negligible.
• Inheritance pattern: Not Mendelian. It's a multifactorial/HLA-restricted susceptibility requiring environmental exposure. The heritable component is the HLA-DPB1 Glu69 (and minor DR) association — best modeled with relationship_type: SUSCEPTIBILITY, exactly as the entry does. Penetrance, expressivity, anticipation, mosaicism, founder effects, consanguinity, and carrier frequency in the classical genetics sense are N/A; the analogous concept is Glu69 allele frequency (~30–40% of the general population carries a Glu69 allele, vs ~80–97% of CBD patients — i.e., the allele is common but disease requires exposure + likely higher-affinity alleles/copy number).
• Demographics: determined by occupation, not ethnicity — exposed workforces skew historically male, but that reflects the industries, not a biological sex effect. Geographic distribution tracks beryllium industry (US DOE nuclear-weapons complex, aerospace hubs, and beryllium-processing regions). No endemic geography in the infectious sense.

10. Diagnostics

The diagnostic dyad: documented beryllium exposure + demonstrated beryllium-specific immune sensitization + granulomatous pathology. Per the 2014 ATS official statement (PMID:25398119):

• Beryllium Lymphocyte Proliferation Test (BeLPT) — the pivotal test. Patient blood (or BAL) lymphocytes are cultured with beryllium salts; proliferation indicates sensitization. Blood BeLPT single-test sensitivity ~61.5%, specificity ~90.8%; split-sample (duplicate) testing raises sensitivity to ~76% at some cost to specificity. BAL BeLPT is more sensitive for organ disease but can be falsely negative in smokers or the immunosuppressed. This is the test that distinguishes CBD from sarcoidosis (which is BeLPT-negative).
• Bronchoscopy with transbronchial (or surgical) biopsy — to demonstrate non-caseating granulomas / mononuclear interstitial infiltrate; BAL typically shows a lymphocytosis with elevated CD4:CD8 ratio and a positive BAL BeLPT.
• Imaging: chest CT/HRCT — upper/mid-zone nodular and reticular opacities, ground-glass, septal thickening, hilar/mediastinal adenopathy; can be normal early. RadLex terms apply.
• Pulmonary function testing: restrictive (or obstructive/mixed) pattern with reduced DLCO; cardiopulmonary exercise testing detects gas-exchange abnormality earliest.
• Genetic testing: HLA-DPB1 Glu69 typing is used in research and risk stratification but is not a stand-alone diagnostic — the allele is too common in the general population. It supports susceptibility, not diagnosis.
• Differential diagnosis: sarcoidosis (the big one — clinically/histologically identical; the exposure history + BeLPT is what separates them), tuberculosis and other infectious granulomas (caseating), hypersensitivity pneumonitis, other pneumoconioses, granulomatosis with polyangiitis.
• Screening: workplace medical-surveillance BeLPT programs (DOE, aerospace) for asymptomatic exposed workers — the standard secondary-prevention tool. LOINC codes exist for BeLPT-type lymphocyte proliferation results.

Omics/liquid-biopsy diagnostics: not clinically applicable; research only.

11. Outcome / Prognosis

• Natural history: variable — some patients remain stable for years with no treatment; a subset progresses to fibrotic, disabling disease and respiratory failure.
• Mortality: CBD can be fatal in advanced fibrotic disease (respiratory failure, cor pulmonale); it is a compensable occupational disease with documented excess mortality in exposed cohorts, but it is not uniformly lethal. No clean 5-/10-year survival figure applies across the disease spectrum — flag as "variable, stage-dependent."
• Morbidity/disability: progressive exertional limitation, oxygen dependence, and lifelong immunosuppressive therapy burden drive substantial disability in progressors.
• Complications: pulmonary fibrosis, pulmonary hypertension, cor pulmonale, respiratory failure, corticosteroid/immunosuppression side effects, and (rarely) systemic granulomatous involvement.
• Recovery: inflammation is partially reversible with therapy; fibrosis is not. Exposure cessation improves the trajectory but doesn't erase established sensitization.
• Prognostic factors: degree of fibrosis at diagnosis, DLCO/exercise gas exchange, extent of granulomatous burden, and continued vs ceased exposure. TNF-α and severity-associated cytokine polymorphisms are candidate molecular prognostics (unvalidated for clinical use).

12. Treatment (MAXO/NCIT terms noted)

The entry's treatment block is solid. Detail:

1. Beryllium exposure cessation — the essential first step for anyone sensitized or diseased; removes the antigen driving the T-cell response. (Not a drug — best captured as removal-from-exposure; NCIT:C49236 Therapeutic Procedure is a reasonable anchor; there isn't a crisp MAXO "exposure avoidance" term.)
2. Systemic corticosteroids (prednisone) — mainstay pharmacotherapy for symptomatic/progressive disease. Typically 3–6 months then reassess PFTs/gas exchange and taper to lowest effective dose. Suppresses granulomatous inflammation; cannot reverse fibrosis; disease recrudesces on taper, so therapy is often lifelong. (CHEBI:8382 prednisone; NCIT:C15986 Pharmacotherapy; MAXO could anchor to corticosteroid/anti-inflammatory therapy.)
3. Steroid-sparing immunosuppressants — methotrexate (e.g., 7.5 mg weekly with folic acid) and azathioprine, adapted from sarcoidosis management, to reduce steroid burden (Current Treatment of CBD review, PMC2774897). (CHEBI:44185 methotrexate.)
4. TNF-α inhibitors (infliximab) — used in refractory granulomatous disease; targets the non-redundant TNF amplifier. Maier et al. (PMID:22974830) showed infliximab "modulates an antigen-specific immune response in chronic beryllium disease." Benefit is less established than in sarcoidosis and infection risk (reactivation TB, etc.) is a real concern. The entry correctly links this to the Th1 and TNF-Driven Macrophage Recruitment and Activation node with INHIBITS. (NCIT:C20401 Monoclonal Antibody / better: infliximab-specific term if available.)
5. Supportive care — supplemental oxygen (MAXO:0000950 supportive care), pulmonary rehabilitation, vaccination, comorbidity management.
6. Lung transplantation — for end-stage fibrotic CBD refractory to medical therapy (MAXO:0010039 organ transplantation).

Pharmacogenomics: none clinically actionable specific to CBD. Experimental/tolerizing approaches — antigen-specific tolerance strategies (e.g., recombinant HLA-DP2 tolerizing beryllium-specific pathogenic T cells, PMID region ~16951350 / Falta group) are preclinical and mechanistically interesting but not clinical. No approved gene/cell/RNA therapy.

13. Prevention

• Primary prevention: exposure control is everything — engineering controls (enclosure, local exhaust, wet processing), the OSHA 0.2 µg/m³ PEL (2017 final rule, PMID:28071878), respiratory protection, dermal protection, and hygiene to prevent take-home exposure. This is the only truly effective lever.
• Secondary prevention: medical-surveillance BeLPT screening of exposed workers to catch sensitization early and remove sensitized individuals from further exposure before organ disease develops.
• Tertiary prevention: in diagnosed CBD, exposure removal + monitoring to slow progression and prevent fibrotic complications.
• Genetic screening: HLA-DPB1 Glu69 pre-employment screening is ethically contentious and not standard — it risks genetic discrimination, the allele is common, and it has poor positive predictive value. Worth flagging as a live policy debate rather than a recommendation.
• Immunization / public-health / vector control: N/A (non-infectious).

14. Other Species / Natural Disease

• Natural disease: No meaningful naturally-occurring CBD in companion animals or wildlife — beryllium exposure is essentially an anthropogenic occupational phenomenon. OMIA has no CBD entry.
• Comparative biology: the disease is defined by human HLA-DP presentation, which doesn't have a direct wild-animal counterpart. Beryllium toxicity can be induced experimentally in animals, but the HLA-restricted immune disease is human-specific.
• Zoonosis / cross-species transmission: N/A.

15. Model Organisms

• Flagship model — HLA-DP2 transgenic mouse (Mucosal Immunology 2015, PMID:26129650): intratracheal beryllium oxide induces lung mononuclear infiltrates and a CD4-dependent, beryllium-specific adaptive immune response in lung and spleen, recapitulating the major features of human CBD; beryllium-responsive CD4⁺ T cells were largely TCR Vβ6⁺, and the group defined HLA-DP2-binding mimotopes recognized by beryllium-specific T cells even without beryllium present. This is the model that ties the human Glu69 genetics to an in-vivo granulomatous phenotype — evidence_source MODEL_ORGANISM.
• Regulatory-T-cell modulation in the same HLA-DP2 model (PMID:24912188) — Tregs tune granuloma intensity, a mechanistic model for variable progression.
• In vitro / cellular models: patient BAL and blood lymphocyte cultures (the BeLPT itself is a functional cellular assay), beryllium-specific CD4⁺ T-cell clones, and recombinant soluble HLA-DP2 for structural/binding work (PNAS 2010, Cell 2014) — evidence_source IN_VITRO.
• Model limitations: mouse models require the human HLA-DP2 transgene to work at all (mouse MHC-II doesn't present beryllium the human way), and murine granulomas don't fully reproduce human fibrotic end-stage disease — a legitimate HUMAN_MODEL_MISMATCH candidate for a discussion block if you want to flag translational caveats. Resources: MGI for the transgenic lines.

Key citations (verify snippets before committing, per the DR/anti-hallucination SOP)

Sources consulted: Richeldi Science 1993, Amicosante PNAS 2000, Bill J Immunol 2005, Clayton Cell 2014, ATS 2014 statement, Newman AJRCCM 2005, systematic review PMID 22705916, HLA-DP2 mouse model, Treg mouse model, infliximab in CBD, OSHA final rule, Current Treatment of CBD, E69 genotype-exposure.

Bottom line for the curation entry: the existing Chronic_Beryllium_Disease.yaml is already in genuinely good shape — the 9-node pathophysiology chain conforms correctly to granuloma_formation, the HLA-DPB1 susceptibility modeling is right, and the treatment/target_mechanisms links are sound. The main gaps I'd consider adding, if you want to push completeness: (1) a BeLPT diagnostic representation and the CD4:CD8 BAL lymphocytosis, (2) explicit skin granuloma / beryllium ulcer and pulmonary hypertension/cor pulmonale phenotypes, (3) a prevalence block (BeS 0.8–12%, CBD 0.1–8% in exposed cohorts; MEASURE_TYPE POINT_PREVALENCE, condition on exposed population in notes), (4) an HLA-DP2 transgenic mouse model/animal block with evidence_source: MODEL_ORGANISM, and (5) maybe a HUMAN_MODEL_MISMATCH discussion noting the transgene requirement and incomplete fibrosis in mice. Two small things to double-check mechanically: the HGNC id for HLA-DPB1 (hgnc:4940) and the hilar/mediastinal lymph node UBERON term if you add anatomy — those are the kind of easy-to-transpose identifiers the term validator will catch.

Want me to fold any of those additions into the YAML and run the validation stack, or keep this as a research reference for now?