How many children, and why a platform
A disease of roughly 160 American infants a year will never support a conventional development programme on its own. The mechanism, though, is shared across many small populations, which is the whole argument for building a platform rather than a single drug.
The break happens in the same 8.3 kb of KMT2A in every one of these patients. What it joins to is what differs — and it changes the disease completely. AFF1 accounts for 48.3% of infant ALL and 0.2% of paediatric AML. That is why no single partner supports a development programme on its own, and why the shared mechanism does.
| stratum | n | AFF1 | MLLT1 | MLLT3 | MLLT10 | AFDN | EPS15 | ELL | KMT2A-PTD | other |
|---|---|---|---|---|---|---|---|---|---|---|
| infant ALL | 987 | 48.3 | 24 | 16 | 6 | 0.3 | 2.2 | · | · | 15 |
| infant AML | 197 | 1.5 | 1 | 28.4 | 27.4 | 1.5 | 1 | 15.2 | · | 16 |
| paediatric ALL | 530 | 40.6 | 18.9 | 17.2 | 3.2 | 5.8 | 1.5 | · | · | 26 |
| paediatric AML | 465 | 0.2 | 6 | 36.6 | 23.4 | 9 | 1.1 | 6.7 | 1.5 | 28 |
| adult ALL | 647 | 82.7 | 10.2 | 1.2 | 0.5 | 1.5 | 0.6 | 0.1 | 0.1 | 14 |
| adult AML | 441 | 1.1 | 3.6 | 24.5 | 10 | 10.2 | 1.6 | 11.6 | 25.6 | 25 |
Cells are the percentage of that stratum carrying that partner; · means the partner was not reported in that stratum, and other counts the remaining rearrangements, which run to more than a hundred distinct partner genes across the cohort. This is a referral cohort, not a population sample, so a column describes the composition of KMT2A-rearranged disease — not how often it occurs. The KMT2A recombinome of acute leukemias in 2023 · PMID 37019990
Two properties are left off the chart on purpose. Every one of these entries carries published counter-evidence, and every one records its own open questions — both are requirements of the schema rather than findings, so a bar for either could not move, and a statistic that cannot move cannot tell you anything. 39 of 40 additionally carry a bite_back note, which is required only where an entry touches an intervention.
0.2% of profiled sarcomas, median age 38, and every number here has a denominator that is not the population
About two in every thousand sarcomas sent for gene sequencing carry a KMT2A rearrangement, and the typical patient is around forty rather than a baby. Both numbers come from tumours that doctors chose to sequence, not from a whole population, so they are floors rather than true rates.
172 adults with KMT2A-r AML: translocation subtype did not predict survival, so treat all KMT2A-r as adverse
In 172 adults with this leukemia, which partner gene was involved made no difference to how long they lived. The authors concluded that every version should simply be classed as high risk.
465 KMT2A-r B-ALL cases from 1 month to 89 years: partner predicts MRD clearance independently of age
By studying leukemias from babies to eighty-nine-year-olds together, this study could ask whether the partner gene matters once age is accounted for. It does: the partner still predicted how fast the leukemia cleared, even after adjusting for age and how mature the cells were.
756 children with 11q23/MLL-rearranged AML: 5-year EFS runs from 11% to 92% depending on the partner
Doctors from eleven countries pooled 756 children with this leukemia and asked whether the partner gene mattered. It mattered enormously: survival ranged from about one in ten to more than nine in ten depending on which partner was involved.
At the DNA level almost every KMT2A-rearranged patient has a reciprocal allele - 94% across partners, and 5% have nothing else
Nearly every child with this kind of leukaemia carries both halves of the chromosome swap in their DNA - not just the famous half. In about one in twenty, the famous half is missing altogether and only the mirror-image gene is there.
AUGMENT-101: the partner composition of the trial that approved a menin inhibitor, with KMT2A::AFF1 at n=2 evaluable
The trial that got the first menin inhibitor approved published which partner gene each patient had. Only two of the fifty-seven patients whose response counted had the partner this project's whole mechanistic story is built around, and a quarter had no partner identified at all.
CCG 1953 (n=115 infants): the first trial to genotype every infant found the partner does not predict outcome
This was the first cooperative-group trial to test every enrolled baby for which partner gene was involved. Despite doing the test on everyone, the partner made no measurable difference to how the babies did.
Childhood leukaemia worldwide, ages 0-14
Leukaemia is the commonest cancer of childhood: about 46 children in every million aged 0-14 are diagnosed each year, which works out to roughly 90,000 children worldwide. Infants under one are a small slice of that, which is why an infant-only drug programme looks unfundable while a childhood-leukaemia platform does not.
Children's Cancer Hospital Egypt (n=150 KMT2A-r): EFS 50% to 17% across four partner groups in a single-centre cohort
A single large hospital in Egypt followed 150 children with this leukemia and grouped them by partner gene. Survival ran from half to one in six depending on the group, though extra chromosome changes mattered too.
CIBMTR post-transplant AML (n=426 KMT2A-r, curated cytogenetics): no outcome difference by fusion partner
In the largest registry study of transplant outcomes in this leukemia, with the chromosome reports checked by hand, the partner gene made no difference to what happened after transplant.
COG AAML0531/1031 (n=2,080): MLLT10 fusions are adverse, and the 5-prime partner barely changes it
Across two large childhood leukemia trials, any fusion involving the gene MLLT10 went with very poor outcomes. Which gene MLLT10 was joined to made almost no difference, which is a rare case of the partner clearly not mattering.
Fusion partner predicts the molecular subgroup almost perfectly, and survival not at all
In 84 babies with this leukaemia, which gene the KMT2A gene is joined to almost completely determines which molecular type of disease they have. It does not determine how long they live.
How common is ZNF384-rearranged leukaemia? Published answers span 1.6% to about 20% and the spread is almost entirely denominator, not biology
Depending on which group of patients you count against, ZNF384 fusions look like 2% of childhood leukaemia or 20% of a particular adult subgroup. Both numbers are correct and they are answers to different questions. The only one that behaves like a population figure comes from a Czech study that counted every child diagnosed over a decade.
How common PICALM::MLLT10 is, and in which disease - 65% T-ALL, 27% AML, and invisible to karyotype half the time
This fusion is rare overall but not rare within T-cell leukaemia, where it turns up in roughly one in eleven unselected cases. Half of the cases cannot be seen down a microscope and need a molecular test, so series that rely on chromosome pictures miss them.
How many children this affects
Leukemia in babies under one year old is rare: about 41 cases per million infants each year in the United States, which is roughly 160 babies a year. Most of those babies with lymphoblastic leukemia carry a broken KMT2A gene.
How many infants worldwide, not just in the United States
About 132 million babies are born each year worldwide, and if leukemia hits infants everywhere at the US rate, roughly 5,400 babies a year develop acute leukemia. No cancer registry anywhere actually counts that number, so it is an estimate built from two solid figures and one shaky assumption.
I-BFM childhood KMT2A-r AML (n=1,130): the 3-prime partner defines two risk groups that differ by 24 EFS points
In the largest study of children with KMT2A-rearranged myeloid leukemia, which partner gene KMT2A fused to split patients into two groups whose survival differed by about a quarter. Five specific partners marked a high-risk group, and that grouping is now written into how children are treated.
In childhood t(11;19), every lymphoblastic case was MLLT1 and every ELL case was myeloid - the one rule that lets an old cohort be de-contaminated
Two different genes on chromosome 19 can join to KMT2A and standard chromosome analysis often cannot tell them apart. In children, an RT-PCR study found that all seventeen lymphoblastic cases involved the gene this beat is about, and the only two cases of the other gene were myeloid.
Infant BCP-ALL (n=175): KMT2A::AFF1 2-year EFS 36% versus 37% for every other partner (P=0.9)
In 175 babies with B-cell leukemia studied with modern sequencing, those whose KMT2A was fused to AFF1 did exactly as well, or badly, as those with any other partner. The two survival numbers were 36% and 37%.
Infants whose leukaemia expresses the reciprocal fusion survive far better - 62.4% versus 11.7% event-free at four years
Babies whose leukaemia makes the mirror-image fusion protein do dramatically better than those whose leukaemia does not - more than five times as many are still event-free after four years. That is the opposite of what you would expect if the mirror-image protein were a second cancer driver.
Interfant-99 and Interfant-06: what they report about partners, and what they do not
The two worldwide trials for babies with this leukemia both name the KMT2A rearrangement as the strongest predictor of outcome, but neither publishes survival broken down by which partner gene was involved. That missing table is exactly the number this project needed.
JPLSG MLL-10 (n=90 infants): a 38-point EFS split built from age, CNS status and MRD, with the partner never entering the model
A Japanese trial sorted babies into risk groups using their age, whether the leukemia had reached the brain, and how fast it cleared. Those three things split survival from 94% to 57%, and which partner gene was involved was never used.
KMT2A-rearranged acute leukemia across all ages: infants are 4% of it
Around 29,000 Americans of all ages are diagnosed with acute leukemia each year, and roughly one in ten of those leukemias has a broken KMT2A gene, so about 2,900 people a year. Only about 120 of them are babies, which means infants are about four in every hundred KMT2A patients.
KMT2A::ELL in adult AML: the one partner that separates, in two independent adult cohorts
Two recent studies of adults found that one particular partner, ELL, went with noticeably better survival than the others. It is the only partner-specific survival signal that has shown up twice in adults.
KMT2A::MLLT1 is high-risk in childhood AML and the best-outcome cell in childhood T-ALL: the partner effect changes sign with lineage
The same partner gene, MLLT1, marks the worst group in childhood myeloid leukemia and the best group in childhood T-cell leukemia. A partner effect that reverses direction depending on the cell type cannot be a simple property of the partner.
KMT2A::MLLT1 outcome by age band: favourable in adult ALL, high-risk in childhood AML, and completely unstratified in the infants who are most of the disease
In adults and older children with lymphoblastic leukaemia this fusion carries a comparatively good outlook, and in children with myeloid leukaemia a bad one. In babies under one - who are more than half of all patients with this fusion - the two largest trials ever run did not report an outcome figure for it at all.
KMT2A::MLLT3 / t(9;11): the best-known partner prognosis claim, and the three cohorts that did not reproduce it
For twenty years the textbook said that one partner, AF9, carried a better outlook in myeloid leukemia. When much larger studies looked, the advantage was not there.
KMT2A::MLLT3 across 612 genotyped patients: myeloid at every age, but 248 of them have ALL, and 157 of those are infants
AF9 is the classic myeloid partner, and that is true - but in the largest genotyped series, two in five of its patients have lymphoblastic leukaemia rather than myeloid, and most of those are babies. The 'myeloid partner' label is a statement about which disease AF9 dominates, not about which disease its patients have.
Latency: months from the fusion forming in a fetus to leukaemia in an infant, against years or a decade for every other backtracked fusion
The fusion forms before birth and the leukaemia usually appears within the first year of life, sometimes within weeks. For the other childhood leukaemias that also start before birth, the gap is years and can be more than a decade. That difference is the whole argument that the KMT2A fusion does not need much help.
MLLT3 is the commonest partner in therapy-induced KMT2A leukaemia: 49 of 142, and 47 of those are t-AML
Among leukaemias caused by earlier cancer treatment, AF9 is the partner that turns up most often - a third of all such cases in the largest genotyped series. It is also the only partner whose break point inside KMT2A moves in adults specifically because of that cause.
NUP98 rearrangements are 7.2% of paediatric AML and 1.6% of adult, with 35% five-year overall survival — and the partner changes everything
About one in fourteen children with acute myeloid leukaemia has a NUP98 fusion, and only about a third of them are alive five years later, against roughly two thirds of other children. It is four to five times commoner in children than in adults.
NUP98::NSD1 without FLT3-ITD loses its prognostic significance — the confound that undermines the fusion's own reputation
NUP98-NSD1 leukaemia has a terrible reputation, but three quarters or more of these patients also carry a separate FLT3 mutation. When you analyse the ones without it, the fusion stops predicting a bad outcome — so the reputation may belong to the other mutation.
Outcome in PICALM::MLLT10: children with ALL do reasonably well, children with AML do badly, and transplant does not rescue them
Children whose CALM-AF10 leukaemia is lymphoid survive about three-quarters of the time; children whose CALM-AF10 leukaemia is myeloid survive about a quarter of the time, and a bone-marrow transplant does not change that. Adults do worse than children whatever the lineage.
Ponte-di-Legno non-infant KMT2A-r ALL (n=629): 5-year EFS spans 41.7% to 91.2% across partner-by-lineage cells
In 629 children over one year old with KMT2A-rearranged lymphoblastic leukemia, outcome differed hugely depending on both the partner gene and whether the leukemia was B-cell or T-cell. The best and worst groups were both T-cell leukemias with different partners, which is the cleanest sign that the partner itself matters.
t(11;19) is the largest named partner group in the revumenib registration trial - and the trial never resolves whether those patients have MLLT1 or ELL
In the trial that got the first menin inhibitor approved, the single biggest identified group of patients was the one with a break between chromosomes 11 and 19 - the fusion this beat is about. But the trial's records never say which of two different genes on chromosome 19 was involved, and in children with myeloid leukaemia the other one is actually more common.
The MEN1 age-penetrance curve: 7% by age 10, 52% by 20, 87% by 30
In families carrying a broken menin gene, disease appears slowly: about 7 in 100 carriers have it by their tenth birthday, half by twenty, and nearly all by forty. That slow build-up is the best available yardstick for how long menin has to be missing before something goes wrong.
The national scale: about 24 infants a year in England and Wales, about 12 in Australia
England and Wales record about 595,000 births a year and Australia about 292,000, so at the US infant leukemia rate they should see roughly 24 and 12 infant acute leukemia cases a year respectively. Australia therefore diagnoses about five infants a year with KMT2A-rearranged leukemia, which is the entire national population a trial could draw on.
Twin concordance: near-total for infant KMT2A-rearranged leukaemia, about 5% for childhood ALL overall
When one identical twin baby gets KMT2A-rearranged leukaemia, the other almost always gets it too, usually within weeks. For older children with other kinds of leukaemia the co-twin is affected only about one time in twenty, and that gap is the single strongest clue that the KMT2A fusion does almost all of the work by itself.
What fraction of fusion-positive newborns ever get leukaemia? For ETV6::RUNX1, very few. For KMT2A, nobody knows, and the biology says the answer may be most of them
For one common childhood-leukaemia fusion, far more babies are born carrying it than ever get ill, so the fusion alone is not enough. For the KMT2A fusion that this project is about, nobody has ever measured how many newborns carry it, and everything else we know suggests that a baby who carries it is in serious trouble.
Who the 460 KMT2A::MLLT1 patients actually are: 88% lymphoblastic, 53% infants, and only 47 cases of AML in the whole series
Of the 460 patients with this fusion in the largest genomic series, nearly nine in ten had lymphoblastic leukaemia and just over half were babies under one year old. Myeloid disease is the minority here - only 47 cases - which matters because the myeloid literature is where this fusion's bad reputation comes from.