Juvenile Myelomonocytic Leukemia: Manual Deep Research Summary

Disease-level modeling decision

JMML should be modeled as one disease-level mechanism graph, not split into multiple pages for every ontology subclass or driver-defined variant. The strongest support for this comes from the 2019 review that states that about 90% of patients have lesions in one of five canonical Ras-pathway genes and that these define genetically and clinically distinct subtypes rather than wholly separate diseases (PMID:30670449). Following the #1198 cancer-curation guidance, the disease file therefore uses:

a single MONDO-first disease anchor for canonical JMML
flat has_subtypes facets for the molecular-driver axis (PTPN11-mutated, NRAS-mutated, KRAS-mutated, NF1-associated, CBL-associated)
no separate dismech page for those subtype facets

Noonan syndrome-associated myeloproliferative disorder (NS-MPD) was not folded into has_subtypes for the JMML entry. The real-world series in PMID:39123476 explicitly separated NS-MPD from JMML treatment analysis because none of the NS-MPD patients required chemotherapy, and spontaneous clinical remission was observed in that group. That supports treating NS-MPD as a related but distinct disease context or differential, not as an ordinary JMML subtype facet.

Ontology grounding choices

Primary disease term: MONDO:0011908 juvenile myelomonocytic leukemia
Disease-adjacent NCIT grounding: NCIT:C9233 Juvenile Myelomonocytic Leukemia
Key phenotype terms: monocytosis, splenomegaly, hepatomegaly, leukocytosis, anemia, thrombocytopenia
Key histopathology terms: bone marrow hypercellularity, myelodysplasia
Key mechanism terms: Ras protein signal transduction, MAPK cascade, cytokine-mediated signaling pathway, DNA methylation
Key treatment terms: hematopoietic stem cell transplantation, pharmacotherapy with azacitidine and trametinib as therapeutic agents

Because the current schema exposes MONDO-specific disease mappings but not dedicated NCIT disease mappings, NCIT grounding was carried in the disease-adjacent cancer representation rather than inventing a new schema pattern locally.

Core disease biology

Canonical defining lesion space

JMML is a pediatric myelodysplastic/myeloproliferative neoplasm overlap syndrome with sustained peripheral blood monocytosis and dominant Ras-pathway biology.

PMID:32460983 describes JMML as a pediatric MDS/MPN overlap syndrome with sustained peripheral blood monocytosis and poor outcomes.
PMID:30670449 states that JMML pathobiology is characterized by constitutive activation of Ras signal transduction and that about 90% of patients harbor lesions in PTPN11, NRAS, KRAS, NF1, or CBL.
PMID:26457647 found canonical lesions in NF1, NRAS, KRAS, PTPN11, or CBL in 85% of cases and showed that additional somatic alterations at diagnosis are strongly associated with outcome.

Atomic mechanism chain supported by the literature

Ras-pathway driver lesion
Canonical driver lesions involve PTPN11, NRAS, KRAS, NF1, or CBL (PMID:30670449, PMID:26457647).
RAS/MAPK pathway hyperactivation
Ras signaling is constitutively active in JMML (PMID:30670449).
Patient-derived iPSC models confirm constitutive Ras/MAPK signaling in PTPN11- and CBL-mutant JMML cells (PMID:29884903).
GM-CSF hypersensitive signaling
GM-CSF hypersensitivity is a disease hallmark in marrow progenitors (PMID:22195407).
iPSC-derived JMML cells show constitutive GM-CSF activation with enhanced STAT5/ERK phosphorylation (PMID:23620576).
Myelomonocytic progenitor expansion
This manifests as leukocytosis, absolute monocytosis, cytopenias, and marrow hypercellularity with myelomonocytic proliferation (PMID:22195407).
Epigenetically aggressive subset
High DNA methylation identifies an aggressive biologic JMML variant with worse survival and higher post-transplant relapse risk (PMID:21406719).

Phenotype and pathology summary

The most reusable phenotype/pathology extraction came from the pathology review PMID:22195407:

marked splenomegaly and hepatomegaly
leukocytosis
absolute monocytosis
anemia
thrombocytopenia
marrow hypercellularity due to myelomonocytic proliferation
mild dysplasia

PMID:29884903 additionally highlights elevated fetal hemoglobin and GM-CSF hypersensitivity as classic diagnostic/risk features.

Treatment evidence summary

Allogeneic hematopoietic stem cell transplantation

Transplant remains the only curative treatment for most canonical JMML cases.

PMID:25435114: allogeneic hematopoietic stem cell transplant is the only curative option, but relapse and toxicity remain substantial.
PMID:30670449: most children require allogeneic hematopoietic stem cell transplantation for long-term leukemia-free survival.

Azacitidine

Azacitidine has meaningful pre-transplant cytoreductive and clinical activity.

PMID:34297046: phase 2 AZA-JMML-001 showed 61% clinical partial remission after three cycles, with 82% leukemia-free survival after subsequent HSCT at median follow-up.
Response was strongest in intermediate- or low-methylation groups in that trial.

Trametinib

MEK inhibition has become credible salvage/bridging therapy for relapsed or refractory disease.

PMID:38867349: phase II Children's Oncology Group study reported an objective response rate of 50%.
Four refractory patients were bridged to HSCT after trametinib, and three additional patients remained on off-protocol trametinib without HSCT at report time.

Subtype-axis reasoning carried into the YAML

The curated subtype axis was limited to the canonical driver-defined JMML groups:

PTPN11-mutated
NRAS-mutated
KRAS-mutated
NF1-associated
CBL-associated

These were chosen because PMID:30670449 explicitly describes them as the five genetically and clinically distinct JMML subtypes. A separate methylation-risk axis was not added to has_subtypes; methylation state was instead modeled as an orthogonal pathobiologic/risk mechanism because it alters prognosis without necessarily defining a separate clinical disease page or distinct primary causal program.