Cancer Curation Project

Overview

Comprehensive curation of cancers for the dismech knowledge base, prioritizing malignancies with precise genetic etiology and well-characterized pathophysiological progression. This project focuses on cancers where: 1. Single-gene or well-defined fusion gene drivers are established 2. Molecular mechanisms are understood at the pathway level 3. Genotype-phenotype correlations inform treatment decisions 4. Progressive stages of tumorigenesis are documented

This is an evergreen project: beyond the defined tiers, "Other" expansion categories capture additional molecularly-defined cancers and emerging biomarkers as the field evolves.

Selection Criteria

Cancers are prioritized based on: - Genetic precision: Clear driver mutations, fusion genes, or germline predisposition - Pathway clarity: Well-understood molecular cascades (e.g., RAS-MAPK, Wnt/β-catenin, HIF/VHL) - Therapeutic relevance: Targeted therapies directly addressing genetic defects - Educational value: Paradigmatic examples of cancer biology (e.g., two-hit hypothesis, oncogene addiction)

Existing Cancer-Related Entries in KB (4)

Cancers to Curate

Tier 1: Paradigmatic Single-Gene Driver Cancers (Highest Priority)

These represent textbook examples of cancer genetics with direct therapeutic implications.

Tier 2: Hereditary Cancer Syndromes (High Priority)

Well-characterized germline mutations with defined progression pathways.

Tier 3: Cancers with Clear Molecular Subtypes (Medium Priority)

Defined molecular classifications guiding precision medicine.

Tier 4: Pediatric Cancers with Fusion Genes (Medium Priority)

Defined by characteristic chromosomal translocations.

Tier 5: Molecularly-Defined Breast Cancer Subtypes (High Priority)

Modern oncology treats these as distinct disease entities requiring different therapeutic approaches.

Tier 6: Molecularly-Defined Lung Cancer Subtypes (High Priority)

Lung cancer is now treated as multiple distinct molecular diseases.

Tier 7: Molecularly-Defined Melanoma and Skin Cancers (High Priority)

Tier 8: Molecularly-Defined GI Cancers (High Priority)

Tier 9: Molecularly-Defined Brain Tumors (High Priority)

WHO 2021 classification requires molecular markers for diagnosis.

Tier 10: Molecularly-Defined Hematologic Malignancies (High Priority)

Tier 11: Environmental Etiology Cancers (High Priority)

Cancers with strong environmental/infectious causation - important for prevention.

Tier 12: Other Molecularly-Defined Solid Tumors (Medium Priority)

Tier 13: Pediatric/Rare Sarcomas (Medium Priority)

Molecular Subtypes as Discrete Entities

Modern precision oncology recognizes that cancers historically grouped together are actually distinct diseases requiring different treatments. This KB models them as separate entities:

Breast Cancer Example

Instead of one "Breast Cancer" entry with subtypes, we have: - HER2-Positive Breast Cancer: Defined by ERBB2 amplification; treated with trastuzumab, pertuzumab, T-DXd - Triple-Negative Breast Cancer: ER-/PR-/HER2-; treated with immunotherapy + chemotherapy - ER-Positive Breast Cancer: Hormone receptor-driven; treated with endocrine therapy + CDK4/6 inhibitors - PIK3CA-Mutant Breast Cancer: Specific mutation; treated with alpelisib

Rationale

1. Different biology: Molecular alterations drive fundamentally different tumor behavior
2. Different treatment: Each requires distinct therapeutic approaches
3. Different prognosis: Outcomes vary dramatically by molecular subtype
4. Regulatory recognition: FDA approvals are increasingly biomarker-specific
5. Clinical practice: Oncologists make treatment decisions based on molecular profile

Environmental Cancers

Cancers with strong environmental/infectious etiology demonstrate: - Gene-environment interactions (e.g., aflatoxin + HBV → TP53 R249S) - Prevention opportunities (HPV vaccination, H. pylori eradication) - Distinct molecular signatures from sporadic counterparts

Key Pathways to Document

Tumor Suppressor Pathways

1. RB1/E2F pathway - Cell cycle checkpoint control
2. TP53 pathway - DNA damage response, apoptosis
3. APC/Wnt/β-catenin - Intestinal stem cell regulation
4. VHL/HIF - Oxygen sensing, angiogenesis
5. BRCA/DNA repair - Homologous recombination

Oncogenic Signaling Cascades

1. RAS-RAF-MEK-ERK (MAPK) - Cell proliferation
2. PI3K-AKT-mTOR - Cell growth, survival
3. JAK-STAT - Cytokine signaling
4. Hedgehog - Developmental signaling
5. Notch - Cell fate determination

Receptor Tyrosine Kinases

1. BCR-ABL - CML paradigm
2. KIT/PDGFRA - GIST
3. RET - MTC, MEN2
4. EGFR/HER2 - Lung, breast cancer
5. ALK - Lung cancer, lymphoma

Research Sources (2025-2026)

Recent developments informing this curation: - AACR Cancer Progress Report 2025 - ecDNA in 17% of cancers - Genetic background sets trajectory of cancer evolution (2025) - Germline-somatic interactions - Blood cancer protective variant (2026) - CHIP protection - VHL signaling in ccRCC (2024) - HIF-2α targeting - Li-Fraumeni syndrome sperm donor case (2025) - 100% female penetrance

STATUS

Tier 1: Paradigmatic Single-Gene Drivers (6/6) ✓

• [x] Chronic Myeloid Leukemia (BCR-ABL1) - Created 2026-01-24, 60.4% compliance
• [x] Retinoblastoma (RB1) - Created 2026-01-24, 55.9% compliance
• [x] Gastrointestinal Stromal Tumor (KIT/PDGFRA) - Created 2026-01-24, 45.9% compliance
• [x] Medullary Thyroid Carcinoma (RET) - Created 2026-01-24
• [x] Clear Cell Renal Cell Carcinoma (VHL) - Created 2026-01-24
• [x] Ewing Sarcoma (EWS-FLI1) - Created 2026-01-24

Tier 2: Hereditary Cancer Syndromes (6/6) ✓

• [x] Li-Fraumeni Syndrome (TP53) - Created 2026-01-24
• [x] Familial Adenomatous Polyposis (APC) - Created 2026-01-24
• [x] Von Hippel-Lindau Disease (VHL) - Created 2026-01-24
• [x] Multiple Endocrine Neoplasia Type 2 (RET) - Created 2026-01-24
• [x] Hereditary Breast and Ovarian Cancer Syndrome (BRCA1/2) - Created 2026-01-24
• [x] Neurofibromatosis Type 1 (NF1) - Created 2026-01-24

Tier 5: Molecularly-Defined Breast Cancer (0/4)

• [ ] HER2-Positive Breast Cancer
• [ ] Triple-Negative Breast Cancer
• [ ] ER-Positive/HER2-Negative Breast Cancer
• [ ] PIK3CA-Mutant Breast Cancer

Tier 6: Molecularly-Defined Lung Cancer (0/8)

• [ ] EGFR-Mutant NSCLC
• [ ] ALK-Rearranged NSCLC
• [ ] ROS1-Rearranged NSCLC
• [ ] KRAS G12C-Mutant NSCLC
• [ ] MET Exon 14 Skipping NSCLC
• [ ] RET-Rearranged NSCLC
• [ ] BRAF V600E-Mutant NSCLC
• [ ] Small Cell Lung Cancer

Tier 7: Molecularly-Defined Melanoma/Skin (0/6)

• [ ] BRAF V600-Mutant Melanoma
• [ ] NRAS-Mutant Melanoma
• [ ] KIT-Mutant Melanoma
• [ ] Uveal Melanoma
• [ ] Basal Cell Carcinoma
• [ ] Merkel Cell Carcinoma

Tier 8: Molecularly-Defined GI Cancers (0/8)

• [ ] MSI-High Colorectal Cancer
• [ ] BRAF V600E-Mutant Colorectal Cancer
• [ ] HER2-Positive Colorectal Cancer
• [ ] HER2-Positive Gastric Cancer
• [ ] Claudin 18.2+ Gastric Cancer
• [ ] Hepatocellular Carcinoma
• [ ] Cholangiocarcinoma, IDH-Mutant
• [ ] Cholangiocarcinoma, FGFR-Altered

Tier 9: Molecularly-Defined Brain Tumors (0/6)

• [ ] IDH-Mutant Astrocytoma
• [ ] IDH-Mutant Oligodendroglioma
• [ ] Glioblastoma, IDH-Wildtype
• [ ] H3 K27-Altered Diffuse Midline Glioma
• [ ] Medulloblastoma, WNT-Activated
• [ ] Medulloblastoma, SHH-Activated

Tier 10: Molecularly-Defined Hematologic Malignancies (0/10)

• [ ] FLT3-Mutant AML
• [ ] NPM1-Mutant AML
• [ ] IDH-Mutant AML
• [ ] Core Binding Factor AML
• [ ] APL (PML-RARA)
• [ ] Ph+ ALL
• [ ] Ph-like ALL
• [ ] Chronic Lymphocytic Leukemia
• [ ] Mantle Cell Lymphoma
• [ ] Diffuse Large B-Cell Lymphoma

Tier 11: Environmental Etiology Cancers (0/12)

• [ ] Malignant Mesothelioma (asbestos)
• [ ] Cervical Cancer (HPV)
• [ ] HPV-Positive Head and Neck SCC
• [ ] HPV-Negative Head and Neck SCC
• [ ] Nasopharyngeal Carcinoma (EBV)
• [ ] Gastric Cancer, H. pylori-Associated
• [ ] EBV-Associated Gastric Cancer
• [ ] Burkitt Lymphoma (EBV)
• [ ] Kaposi Sarcoma (HHV-8)
• [ ] Adult T-Cell Leukemia/Lymphoma (HTLV-1)
• [ ] Aflatoxin-Related HCC
• [ ] Arsenic-Related Cancers

Tier 12: Other Molecularly-Defined Solid Tumors (0/10)

• [ ] FGFR-Altered Urothelial Carcinoma
• [ ] BRCA-Mutant Prostate Cancer
• [ ] MSI-High Endometrial Cancer
• [ ] NTRK Fusion-Positive Cancers
• [ ] Thyroid Cancer, BRAF-Mutant
• [ ] Thyroid Cancer, RET-Fusion
• [ ] Pheochromocytoma/Paraganglioma
• [ ] Polycythemia Vera
• [ ] Primary Myelofibrosis
• [ ] Essential Thrombocythemia

Tier 13: Pediatric/Rare Sarcomas (0/8)

• [ ] Alveolar Rhabdomyosarcoma
• [ ] Embryonal Rhabdomyosarcoma
• [ ] Synovial Sarcoma
• [ ] Neuroblastoma
• [ ] Wilms Tumor
• [ ] Desmoplastic Small Round Cell Tumor
• [ ] Clear Cell Sarcoma
• [ ] Dermatofibrosarcoma Protuberans

Progress Summary

• Total planned: 78+ cancers (original 28 + 50 expansion + open-ended)
• Completed: 12/78 (15.4%)
• In progress: 0

Expansion Rationale (50 new entries)

1. Molecularly-defined subtypes as discrete entities (Tiers 5-10, 12): Modern oncology recognizes that molecular markers define distinct disease entities requiring different treatments. HER2+ breast cancer is a fundamentally different disease from triple-negative breast cancer, not just a "subtype."

2. Environmental etiology cancers (Tier 11): Cancers with strong infectious/environmental causation demonstrate gene-environment interactions and are critical for prevention strategies.

3. WHO classification alignment: Brain tumors (Tier 9) now require molecular markers for diagnosis per WHO 2021 CNS classification.

4. Actionable alterations: Priority given to cancers with FDA-approved targeted therapies or immunotherapies linked to molecular features.

Expansion Categories

• [ ] Other single-gene driver cancers; e.g. SMARCB1-deficient tumors, SDH-deficient GIST
• [ ] Other hereditary cancer syndromes; e.g. Cowden, Peutz-Jeghers, MEN1, NF2, tuberous sclerosis
• [ ] Other molecular breast cancer subtypes; e.g. CDH1-mutant lobular, GATA3-mutant
• [ ] Other molecular lung cancer subtypes; e.g. HER2-mutant, NRG1 fusion, SMARCA4-deficient
• [ ] Other molecularly-defined sarcomas; e.g. CIC-rearranged, BCOR-rearranged
• [ ] Other virus-associated cancers; e.g. HBV-associated HCC, HIV-associated lymphomas
• [ ] Emerging biomarker-defined cancers; e.g. TMB-high, HRD-positive agnostic

NOTES

2026-01-24 (Tier 2 Evidence Integration Session)

Tier 2 Evidence Integration Complete:

Deep research completed and PMID evidence integrated for all 6 Tier 2 hereditary cancer syndromes:

Compliance improvement (weighted scores): - Li-Fraumeni Syndrome: 37.0% → 47.8% - FAP: 41.9% → 48.8% - VHL: 36.7% → 40.8% - MEN2: 43.2% → 50.0% - HBOC: 37.5% → 45.0% - NF1: 43.6% → 51.3%

All 6 files pass validation (schema + terms + references).

2026-01-24 (Evidence Integration Session)

Tier 1 Evidence Integration Complete:

Deep research completed and PMID evidence integrated for all 6 Tier 1 cancers:

All 6 files pass validation (schema + terms + references).

2026-01-24 (Initial Session)

Tier 1 Complete - All 6 paradigmatic cancers curated:

1. Chronic Myeloid Leukemia (BCR-ABL1) - 60.4% compliance - Deep research with falcon, key refs: PMID:11287972, PMID:11423618 - Covers BCR-ABL1 fusion, TKI therapy, disease phases

2. Retinoblastoma (RB1) - 55.9% compliance - Key ref: PMID:5279523 (Knudson two-hit hypothesis) - Paradigm for tumor suppressor gene concept

3. Gastrointestinal Stromal Tumor (KIT/PDGFRA) - 45.9% compliance - KIT/PDGFRA mutations, imatinib, avapritinib

4. Medullary Thyroid Carcinoma (RET) - RET proto-oncogene, MEN2 syndromes, selpercatinib/pralsetinib

5. Clear Cell Renal Cell Carcinoma (VHL) - VHL/HIF pathway, belzutifan (HIF-2α inhibitor)

6. Ewing Sarcoma (EWS-FLI1) - Fusion transcription factor oncogene

Project created with comprehensive research on cancers with precise genetic etiology.

Key selection rationale: - Tier 1 represents "textbook" cancers where single genetic events drive disease (BCR-ABL in CML, KIT in GIST, etc.) - ideal for demonstrating oncogene addiction - Tier 2 covers hereditary syndromes with germline mutations - demonstrates two-hit hypothesis and cancer predisposition - Tier 3 includes common cancers with well-defined molecular subtypes guiding therapy - Tier 4 focuses on pediatric cancers defined by fusion genes - often "clean" genetic drivers - Tier 5 includes additional genetically-defined cancers for completeness

Pathway themes to emphasize: 1. Tumor suppressor loss (RB1, TP53, APC, VHL) 2. Oncogenic kinase activation (BCR-ABL, KIT, RET) 3. Transcription factor fusions (EWS-FLI1, PAX3-FOXO1) 4. DNA repair deficiency (BRCA1/2, MMR genes) 5. Epigenetic dysregulation (IDH1/2 mutations)

Recommended curation order: Start with Tier 1 as these are the clearest examples of genotype-phenotype relationships and have FDA-approved targeted therapies demonstrating mechanism-based treatment.

2026-01-24 (Schema Enhancements)

Added NCIT term support:

1. BiomarkerDescriptor with NCIT validation for biochemical entries
2. GeneProductDescriptor with NCIT validation (rooted at NCIT:C26548 Gene Product) for pathophysiology
3. gene_products slot on Pathophysiology class for fusion proteins, oncoproteins, etc.

Terms added to cancer files: - CML: NCIT:C16325 (BCR/ABL1 Fusion Protein) - GIST: NCIT:C17328 (KIT receptor) - MTC: NCIT:C18539 (RET receptor)

Future work - Post-composition for complex aberrations:

Explore OWL-style post-composition for fusion proteins and complex molecular aberrations that aren't in NCIT: - EWS-FLI1 fusion protein (Ewing Sarcoma) - not currently in NCIT as protein entity - VHL/pVHL protein (ccRCC) - gene exists, protein entity not found - RB1/pRB protein (Retinoblastoma) - gene exists, protein entity not found - Complex fusions with variant breakpoints (e.g., BCR-ABL p190 vs p210)

Consider using the qualifiers slot or creating a dedicated composition mechanism to express:

# Example post-composition for missing fusion proteins
gene_products:
- preferred_term: EWS-FLI1 fusion protein
  description: Fusion of EWSR1 and FLI1 gene products
  qualifiers:
    - predicate: has_part
      value: {term: {id: NCIT:C17476, label: RNA-Binding Protein EWS}}
    - predicate: has_part
      value: {term: {id: NCIT:C17770, label: Friend Leukemia Virus Integration 1}}

This would allow precise annotation of fusion proteins even when pre-coordinated terms don't exist.

Added Histopathology Section (2026-01-24):

New dedicated histopathology slot on Disease class for microscopic tissue findings:

1. HistopathologyFindingTerm dynamic enum - rooted at: - NCIT:C35867 (Morphologic Finding) - patterns, dysplasia, necrosis - NCIT:C18000 (Histologic Grade) - Fuhrman, Nottingham, GIST grades - HP:0025461 (Abnormal cell morphology) - rosettes, inclusion bodies

2. HistopathologyFindingDescriptor class for term binding

3. HistopathologyFinding class with: name, finding_term, description, frequency, diagnostic, evidence, notes, context

Example histopathology entries added: - Retinoblastoma: Flexner-Wintersteiner rosettes (HP:0031927), Homer Wright rosettes (HP:0031926), Fleurettes (NCIT:C35950) - GIST: Spindle Cell Pattern (NCIT:C53643), Low Mitotic Activity (NCIT:C35961)

Rationale: Histopathology findings are microscopic observations distinct from clinical phenotypes. They often define tumor subtypes, inform prognosis, and require tissue sampling. Keeping them separate from phenotypes provides cleaner data modeling.