Hepatoblastoma: Manual Curation Notes

Modeling decisions used for the YAML

The dismech entry was kept at the Hepatoblastoma disease-page level rather than splitting into separate files for pure fetal, embryonal, SCU, or predisposition-associated cases.
Subtypes were modeled as flat facets rather than page-worthy child diseases: histology (epithelial, mixed epithelial and mesenchymal, fetal, embryonal, SCU-component-positive), molecular risk (C2-signature), and predisposition context (Beckwith-Wiedemann spectrum, familial adenomatous polyposis).
MONDO remains the disease anchor (MONDO:0018666 hepatoblastoma). NCIT is used where the current schema supports oncology-grounded descriptors, chiefly disease-level histopathology and biomarkers.
INI1-negative/SMARCB1-deficient small-cell liver tumors were treated as a boundary condition for hepatoblastoma curation rather than as a hepatoblastoma subtype with its own dismech page.

Core disease mechanism

The shared disease program is dominated by CTNNB1/WNT activation. Clinical and liquid-biopsy studies consistently identify CTNNB1 as the dominant recurrent driver in sporadic hepatoblastoma (PMID:32881242; PMID:33125945). This sits upstream of a broader developmental cancer program rather than defining several separate dismech pages.

Hepatoblastoma also shows frequent beta-catenin/YAP1 coactivation. Human tumor tissue demonstrates concurrent nuclear localization of beta-catenin and YAP1 in most cases, and model systems show that combined activation is tumorigenic whereas either pathway alone is insufficient for the same phenotype (PMID:24837480; PMID:30794807). This justified splitting WNT activation and YAP1 cooperation into separate but connected pathophysiology nodes.

The tumor preserves a fetal hepatic progenitor state rather than completing normal hepatocytic differentiation. Pretreatment molecular profiling supports clinically relevant subgroups defined by hepatic progenitor markers and metabolic programs (PMID:27775819). That developmental-arrest concept was modeled separately from the upstream CTNNB1/YAP1 driver nodes.

Aggressive cases show an oncofetal stem-like program and NFE2L2-linked stress adaptation. High-risk tumors are enriched for LIN28B, HMGA2, SALL4, AFP, and NFE2L2 activity, while the C2/NFE2L2-mutant context correlates with non-fetal histology, vessel invasion, metastasis, and inferior survival (PMID:27775819; PMID:33125945). Those were split into separate mechanistic nodes to keep the graph atomic.

Experimental therapy papers also support a downstream growth-support module in beta-catenin/YAP1-driven tumors. In vitro knockdown of either arm reduces growth, and rapamycin reduces tumor burden in a Yap1/beta-catenin mouse model (PMID:24837480; PMID:30863496). This was used to support a distinct proliferation node rather than folding all growth effects into the upstream signaling nodes.

Subtype and histopathology notes

Histology remains a major facet axis. Reviews and clinicopathologic series support broad separation into epithelial versus mixed epithelial-mesenchymal tumors, with fetal, embryonal, and SCU morphologies inside the epithelial axis (PMID:33981680; PMID:19072985).

The key boundary update from the more recent literature is that SCU does not by itself define a separate aggressive hepatoblastoma disease program once rhabdoid tumors are excluded. Modern COG and CHIC analyses show that INI1-negative/SMARCB1-deficient small-cell tumors should be treated as malignant rhabdoid tumor, while SCU component-positive hepatoblastoma without rhabdoid reclassification does not independently worsen outcome (PMID:34874751; PMID:36672416; PMID:31835848). That directly affected the final YAML:

SCU is modeled as a flat histology facet inside Hepatoblastoma.
INI1-negative liver tumors are described as a classification boundary, not as a Hepatoblastoma child page.
The disease-level histopathology grounding uses NCIT for Hepatoblastoma and Embryonal Neoplasm.

Predisposition contexts

Two predisposition contexts are strong enough to warrant explicit subtype facets without separate disease pages.

Beckwith-Wiedemann spectrum (BWSp): hepatoblastoma arises in the setting of 11p15 imprinting abnormalities and epigenotype mosaicism, with frequent convergence on CTNNB1 mutation in the tumor (PMID:37174013).
Familial adenomatous polyposis (FAP): APC mutation confers childhood hepatoblastoma risk and motivates infant surveillance in affected families (PMID:29719120; PMID:1329510).

These were modeled as predisposition_context subtypes instead of separate dismech entries because the shared tumor mechanism still converges on the same disease-level beta-catenin/oncofetal program.

Phenotype and biomarker notes

The most stable disease-level presentation remains painless abdominal mass with tumor-associated hepatomegaly/abdominal enlargement (PMID:33981680; PMID:16458842).

Alpha-fetoprotein (AFP) is the core biomarker. Most tumors show elevated serum AFP, which supports diagnosis and monitoring, while AFP below 100 ng/mL marks a clinically important context that overlaps with SMARCB1-reclassified rhabdoid tumors and failed conservative surgery (PMID:33981680; PMID:16458842; PMID:18166449; PMID:17661341).

CTNNB1 ctDNA is an emerging molecular biomarker that tracks tumor burden, residual disease, and response (PMID:32881242; PMID:38201440).

Treatment notes

Current management is still organized around cisplatin-based preoperative chemotherapy, complete surgical resection when feasible, and liver transplantation for unresectable liver-confined disease (PMID:28126357; PMID:25349947; PMID:16458842; PMID:12352881; PMID:34441025; PMID:17661341).

Because the schema’s oncology-specific treatment support is split between MAXO actions and NCIT regimen descriptors, the YAML keeps generic treatment actions in MAXO and uses CHEBI agents for the specific cytotoxics. An NCIT regimen term would only be preferable here if a stable hepatoblastoma-specific regimen term were already available in the local ontology slice.