Juvenile myelomonocytic leukemia

JMML is a rare cancer of the blood-forming cells in the bone marrow, and it is almost entirely a disease of babies and very young children. The marrow makes too many monocytes — a type of infection-fighting white blood cell — along with other immature cells, and they build up in the blood, the marrow, the liver and the spleen, crowding out normal blood production.

Where JMML sits in the classification system has genuinely changed, and clinicians still use different words for it. The 2016 WHO edition grouped it with the “MDS/MPN overlap” neoplasms; the 2022 WHO 5th edition moved it out and calls it a RAS-pathway-driven myeloproliferative neoplasm of early childhood; a separate 2022 consensus classification files it with paediatric and germline-associated myeloid disorders. Many clinicians still say “MDS/MPN overlap” in conversation. None of this changes how the disease is treated.

JMML is not childhood CML, and it is not a children’s version of adult chronic myelomonocytic leukemia. The old name “juvenile CML” is obsolete — by definition JMML lacks the BCR::ABL1 fusion that defines CML.

JMML is driven by over-activation of something called the RAS pathway — a relay of protein switches inside a cell that tells it when to grow and divide. A mutation jams one of those switches in the “on” position, so the cell keeps multiplying without being told to.

About 90% of children have a mutation in one of five genes in that pathway: PTPN11, NRAS, KRAS, NF1 or CBL. Which gene it is matters more here than in most leukemias, because it shapes both how fast the disease moves and whether a transplant is needed at all.

Most of these mutations are somatic — acquired inside a single blood cell after conception, present only in the leukemia, not inherited and not passed on. Some are germline, meaning present in every cell of the body and sometimes inherited. Germline PTPN11 underlies Noonan syndrome, germline NF1 underlies neurofibromatosis type 1, and germline CBL underlies CBL syndrome.

No environmental, dietary, infectious or parental cause is identified in any source we read. Nothing a parent did caused this.

What a parent usually notices first is a swollen or hard tummy, from an enlarged spleen and liver, along with a pale, tired, off-colour child, fevers or repeated infections, a rash, and easy bruising. St. Jude also lists a dry cough and breathing symptoms, because the abnormal cells can settle in the lungs.

In a European series of 110 children at diagnosis, an enlarged liver or spleen was present in 97%, swollen glands in 76%, pallor in 64%, fever in 54% and a skin rash in 36%.

The mechanism behind all of that is the same crowding-out. A marrow full of monocytes and immature cells cannot make enough red cells, which causes the pallor and tiredness, or enough platelets, which causes the bruising and bleeding — and the excess cells lodge in the spleen and liver and make them large.

How fast it moves depends on the genetics. Disease driven by PTPN11 or NF1 is often rapidly progressive. Disease driven by a germline CBL change is often slowly self-limiting. That difference is why the sequencing result matters so much and why families often wait on it before the plan becomes clear.

For most children, an allogeneic transplant — replacing the child’s blood-forming cells with a donor’s — is the main route to long-term control, rather than something held back until other treatments have failed. That is unusual, and it is the central fact about treating JMML.

Before the transplant comes conditioning, the chemotherapy given to clear the marrow. The standard European regimen combines busulfan, cyclophosphamide and melphalan. A trial comparing it against a lower-intensity regimen closed early because too many children relapsed on the lower-intensity arm.

Azacitidine, a drug that changes how genes are switched on rather than killing cells outright, is approved for children with newly diagnosed JMML and is used to control disease while a transplant is arranged. In a phase 2 trial of 18 newly diagnosed children, 11 had a partial remission after three cycles and 17 of the 18 went on to transplant.

Conventional chemotherapy has not been shown to improve survival here. In one trial, treatment produced responses in most evaluable children, but reaching a complete remission before transplant did not significantly improve survival or relapse risk. Chemotherapy is used to control the disease, not to cure it.

There is a genuine non-transplant pathway, and it is genotype-specific. Most children with germline CBL mutations do not need an urgent transplant and can be watched, with many seeing the disease settle on its own. Noonan-syndrome-associated disease with a germline PTPN11 change is self-limiting in the vast majority of cases. Some children with somatic NRAS disease regress over the long run without treatment.

The patient is a baby or a toddler, so most of what there is to describe belongs to the parents. The workup means blood counts, a marrow sample taken from the hip bone under general anaesthetic, and genetic testing — and families often wait on the sequencing before anyone can say whether the child is going urgently to transplant or being watched.

For the higher-risk genetic groups the clock is short. Published guidance recommends transplant within about three months for children with somatic PTPN11 disease who have adverse features.

Transplant means weeks in hospital: conditioning chemotherapy, the infusion, then a wait for the new cells to take hold while the child has almost no immune system at all. Isolation, transfusions, feeding support and infection risk dominate that period.

Families are sometimes told something counter-intuitive — that a degree of graft-versus-host disease is not purely a complication to be stamped out, because the same donor immune activity is part of what controls the leukemia. Immunosuppression is deliberately tapered early for that reason.

Relapse is the main way treatment fails, so surveillance continues long after discharge, including repeat marrow tests and checks on what fraction of the blood cells are the donor’s. St. Jude notes long-term effects for transplanted children including growth delays, fertility concerns and organ problems, needing follow-up into adulthood.

Central, and unusually so. For most children with JMML a transplant is the main route to long-term control rather than a last resort, which means a donor is needed for most children who are treated. This is one of the few pages in this library where that is true.

The donor is not always unrelated. A matched brother or sister is used when there is one. Unrelated adult donors and cord blood units cover the children who have no matched relative — and infants and toddlers are exactly the recipients for whom a cord blood unit works best, because a small child needs fewer cells.

An unrelated donor is not a lesser option here. In the European trial of 100 transplanted children, five-year event-free survival was 55% with an HLA-identical family donor and 49% with an unrelated donor, and the NCI reports no difference in outcome between related and unrelated donors.

Speed is part of the donor’s role in this disease specifically. PTPN11- and NF1-driven JMML often moves fast, and guidance for high-risk children puts transplant within about three months. A donor who is already registered, contactable and willing matters more here than in slower-moving diseases.