Treatment of juvenile myelomonocytic leukemia (JMML) increasingly follows the biology of each child’s disease: germline or somatic origin, secondary mutations, DNA methylation, clinical behavior and response to therapy. These help determine whether a child can be observed, should go rapidly to transplantation or may benefit from targeted drugs. For a cancer that affects barely one child per million each year, this is remarkable.
A rare childhood leukemia
JMML is a childhood myeloid malignancy with an estimated incidence of about 1.2 cases per million children per year. It accounts for approximately 1% of pediatric leukemias and occurs predominantly in infancy and early childhood, with a median age at diagnosis of around 2 years. Boys are affected about twice as often as girls.
JMML does not behave like a conventional acute leukemia. Blasts are usually not the dominant population. The disease produces persistent proliferation of monocytic and granulocytic cells, often with striking enlargement of the spleen and liver. The fifth edition of the World Health Organization classification therefore places JMML among myeloproliferative neoplasms, and the International Consensus Classification places it among pediatric and/or germline mutation-associated disorders.
One pathway, many origins
More than 90% of cases carry alterations in PTPN11, NRAS, KRAS, NF1 or CBL, all of which activate RAS-MAPK signaling. This constitutive signaling produces the characteristic myelomonocytic expansion. Somatic PTPN11 mutations are the largest subgroup, at approximately 35%-38% of cases.
The origin of the mutation can matter as much as its name. Around 55%-65% of JMML is associated with acquired somatic RAS-pathway mutations, and ~ 25%-30% occurs in children with germline RASopathy-associated alterations. The most relevant inherited conditions are neurofibromatosis type 1, CBL syndrome and Noonan syndrome.
Somatic PTPN11-mutated and NF1-associated JMML generally behave aggressively. Some children with germline CBL abnormalities experience spontaneous resolution, and selected NRAS-mutated cases may follow an indolent course. In infants with Noonan syndrome and germline PTPN11 alterations, myeloproliferation can also resolve spontaneously, and it must be distinguished from aggressive sporadic JMML.
How JMML presents
Hepatosplenomegaly, enlargement of the spleen and liver, occurs in ~80%-90% of patients. Fever, pallor, lymphadenopathy, skin lesions, bruising or bleeding, recurrent infections and poor weight gain are also common. Leukemic infiltration of the lungs occasionally causes respiratory symptoms.
The blood count typically shows leukocytosis with persistent absolute monocytosis, anemia and thrombocytopenia. Immature granulocytes and nucleated red blood cells may circulate in peripheral blood, and fetal hemoglobin can be elevated for age. The bone marrow is generally hypercellular, with prominent myelomonocytic proliferation.
In a very young child, fever, leukocytosis, lymphadenopathy and splenomegaly can initially resemble severe infection. Persistent monocytosis and thrombocytopenia, together with the characteristic molecular findings, separate JMML from reactive childhood monocytosis.
Diagnosis
Diagnosis combines the hematologic phenotype with molecular testing. The hallmark is persistent peripheral blood monocytosis of at least 1 × 10⁹/L, with blasts plus promonocytes below 20% in peripheral blood and bone marrow. BCR::ABL1-positive leukemia and other genetically defined diseases that can mimic JMML must be excluded.
Next-generation sequencing confirms clonality when it finds a RAS-pathway alteration. It can also distinguish sporadic JMML from a germline predisposition syndrome and uncover additional abnormalities associated with aggressive disease. Monocyte-subset flow cytometry, which is useful in adult CMML (expansion of the classical CD14++/CD16− monocyte fraction), is not an established diagnostic criterion for JMML.
Cases without one of the five canonical drivers deserve broader investigation. Some harbor less common alterations affecting RAS signaling, or kinase fusions. Current classification no longer treats molecularly unexplained JMML-like disease as automatically equivalent to canonical JMML.
Predicting the course
Older risk assessment relied heavily on age, platelet count, fetal hemoglobin and cytogenetics. Secondary epigenetic mutations, or further RAS-pathway lesions, mark clonal evolution and have been associated with adverse outcomes. In a contemporary transplant cohort, 62% of children had karyotypic abnormalities or additional mutations in these genes or the RAS pathway.
DNA methylation profiling has emerged as one of the strongest prognostic tools. Patients fall into low-, intermediate- and high-methylation groups. Low methylation is enriched for biologically favorable disease, and high methylation is associated with inferior outcomes.
The information addresses a difficult clinical question: which children can safely be observed, and which need definitive therapy quickly? A favorable driver is less reassuring when the disease has acquired additional molecular abnormalities or an adverse methylation profile.
How Is JMML treated?
Unlike childhood AML, JMML has no established curative induction chemotherapy regimen. Conventional cytotoxic therapy can reduce leukocytosis, organomegaly and disease burden, but it has not consistently improved survival or prevented relapse after transplantation.
Selected children can start with careful observation, particularly those with forms associated with spontaneous regression (some germline CBL-associated and selected NRAS-mutated disease). Most children with aggressive JMML need allogeneic hematopoietic stem cell transplantation (HSCT), the established curative treatment.
Before transplantation, therapy aims to control the disease and bridge the child to HSCT, not to cure it. Very young children are especially vulnerable, because intensive cytotoxic treatment carries substantial toxicity.
Azacitidine before transplant
Azacitidine provided the first strong prospective evidence for such an approach. In the multicenter phase II AZA-JMML-001 trial, Niemeyer and colleagues treated 18 newly diagnosed children with azacitidine before planned HSCT, at 75 mg/m² daily for seven days of each 28-day cycle. After three cycles, 11 of 18 patients achieved a clinical partial response, and 17 ultimately proceeded to transplantation.
The trial also produced an unexpected observation. Responses were particularly frequent in low- and intermediate-methylation disease, and high baseline methylation did not predict greater sensitivity to a hypomethylating drug.
Targeting RAS directly
The near-universal dependence of JMML on RAS-MAPK signaling made a clear case for MEK inhibition, and it has now been tested prospectively. In a Children’s Oncology Group phase II study, Stieglitz and colleagues gave the MEK1/2 inhibitor trametinib to children with relapsed or refractory RAS-mutated JMML. Five of 10 treated patients responded, an objective response rate of 50%, and four children went on to HSCT.
Resistance remains a problem. Three children developed progressive disease, and clonal evolution was observed during therapy in some cases. Blocking MEK can suppress RAS-driven proliferation without eliminating genetically evolving JMML stem and progenitor cells. This supports combining pathway-directed treatment with other therapies, and using targeted agents as a bridge to transplantation in aggressive disease.
Transplant outcomes
Modern results are considerably better than the roughly 50% survival often quoted from older JMML series. Meyran and colleagues analyzed 119 children who underwent a first allogeneic HSCT in France between 2002 and 2021. Five-year overall survival was 73.6%, event-free survival was 66.4% and treatment-related mortality was 9%. The 5-year cumulative incidence of relapse was 24.6%, making recurrent JMML the principal cause of treatment failure.
Four factors independently predicted poorer overall survival:
- age above 2 years at diagnosis
- a monocyte count above 7.2 × 10⁹/L
- additional genetic alterations
- an interval of at least six months between diagnosis and transplantation
Children with three or more of these factors had a 5-year overall survival of 34.2%, compared with 100% in the small subgroup with none.
The timing signal does not prove that every child should be transplanted within six months, because timing depends on disease characteristics and clinical circumstances. It does argue against unnecessary delay once aggressive JMML has been identified and transplantation is indicated.
Relapse after HSCT can still be salvageable, because JMML is sensitive to immune attack. Options include reducing immunosuppression and, in selected patients, a second transplant. The persistent relapse rate is also why research now focuses on molecular disease control before and after HSCT.
What comes next
Azacitidine has prospective evidence as pretransplant therapy, and trametinib has shown activity against relapsed or refractory RAS-mutated disease. Current studies are combining the two approaches earlier in treatment, including MEK inhibition with azacitidine, and testing strategies that adjust treatment intensity to molecular risk.
You can also read:
FAQ
Can JMML ever disappear without treatment?
Yes, but only in selected molecular subtypes. Some children with germline CBL-associated disease and a subset with NRAS mutations can have spontaneous regression. Noonan syndrome-associated myeloproliferation can also resolve during infancy. This unusual behavior is one reason molecular and germline testing matter before committing a child to intensive therapy.
Can JMML be mistaken for an infection?
Yes. JMML often develops in very young children with fever, lymphadenopathy, leukocytosis, monocytosis and hepatosplenomegaly, all of which can initially suggest infection. Persistent monocytosis, thrombocytopenia, immature myeloid cells and characteristic molecular abnormalities should prompt investigation for JMML when the clinical picture does not resolve as expected.
Is JMML inherited?
Usually not, but germline predisposition is unusually important in JMML. Children with neurofibromatosis type 1, CBL syndrome and certain RASopathies can develop JMML or related myeloproliferation. Finding a mutation in blood therefore does not automatically mean it arose only in the leukemia; determining whether it is germline can affect diagnosis, prognosis and family counseling.
Why isn't JMML treated with standard AML chemotherapy?
JMML is biologically different from AML, and intensive AML-type chemotherapy has not consistently produced durable disease control or prevented relapse after transplantation. Cytotoxic drugs may reduce disease burden, but for aggressive JMML, allogeneic stem cell transplantation remains the established curative treatment. Pretransplant therapy is increasingly focused on controlling JMML effectively without unnecessary toxicity.
Can targeted therapy replace stem cell transplantation in JMML?
Not currently. The MEK inhibitor trametinib produced responses in 5 of 10 children with relapsed or refractory RAS-mutated JMML in a phase II study, demonstrating that the pathway driving JMML can be therapeutically targeted. The study was small, resistance occurred, and several responding children subsequently underwent transplantation. Current research is testing whether targeted therapy can be moved earlier and combined with agents such as azacitidine.
