KMT2A::MLLT3 across 612 genotyped patients: myeloid at every age, but 248 of them have ALL, and 157 of those are infants
also called t(9;11) epidemiology · MLL-AF9 age distribution
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.
The same thing, for a scientist
From 3,401 genotyped KMT2A-rearranged patients, KMT2A::MLLT3 accounts for **612 cases**, distributed as: infant ALL 157, infant AML 56, infant other 5; paediatric ALL 91, paediatric AML 170, paediatric other 3; adult ALL 8, adult AML 108, adult other 3; and 11 unclassified by age [E1]. Expressed as shares of each clinical stratum: **infant ALL 16.0% (157/987), infant AML 28.4% (56/197), paediatric ALL 17.2% (91/530), paediatric AML 36.6% (170/465), adult ALL 1.2% (8/647), adult AML 24.5% (108/441)**; overall 11.8% of all ALL (258/2,182) and 30.4% of all AML (339/1,116) [E1]. Two readings follow, and the graph has only carried the first. **MLLT3 is the dominant AML partner at every age band**, peaking in paediatric AML. But **248 of its 597 age-and-lineage-classified cases are ALL**, 157 of them infants - so if you count patients rather than shares, the single largest KMT2A::MLLT3 group in this archive is **infant ALL**, not paediatric AML. It is also present in T-ALL (n=7 of 123 KMT2A-r T-ALL) and in MPAL (8 cases across age bands) [E1]. Outcome is covered by this graph's `epi-partner-mllt3-t9-11-favourability-claim-and-its-non-replication`; the short version verified this session is that the favourable prognosis of t(9;11) was **not confirmed** in 756 children with centrally reviewed karyotypes [E2], and that I-BFM places MLLT3 in the **non-high-risk** rather than the favourable group of a 1,130-patient cohort in which non-high-risk means 5-year EFS 54.0% and OS 70.5% [E3].
The table, as printed
Recombinome Table 1, row 2 (MLLT3), n = 612 1:
| age band | ALL | AML | other |
|---|---|---|---|
| infant (<=12 months) | 157 | 56 | 5 |
| paediatric (>12 months - 18 years) | 91 | 170 | 3 |
| adult (>18 years) | 8 | 108 | 3 |
| not age classified | 2 | 5 | 4 |
| total | 258 | 339 | 15 |
Shares of each stratum 1:
| stratum | MLLT3 share | rank among partners |
|---|---|---|
| infant ALL (n=987) | 16.0% | 3rd, behind AFF1 48.3% and MLLT1 24.0% |
| infant AML (n=197) | 28.4% | 1st, ahead of MLLT10 27.4% |
| paediatric ALL (n=530) | 17.2% | 3rd, behind AFF1 40.6% and MLLT1 18.9% |
| paediatric AML (n=465) | 36.6% | 1st, ahead of MLLT10 23.4% |
| adult ALL (n=647) | 1.2% | 4th |
| adult AML (n=441) | 24.5% | 2nd, behind KMT2A-PTD 25.6% |
The reading the graph has not carried: MLLT3 is a large infant-ALL partner
Every summary of this partner in the literature, and in this graph until now, leads with myeloid. That is correct as a statement about which disease MLLT3 dominates. It is misleading as a statement about which disease MLLT3 patients have.
- MLLT3's largest single cell in the table is infant ALL, 157 patients, larger than paediatric AML's 170 only by a whisker in the other direction - and 157 infants with ALL is a bigger group than the entire adult AML cell (108).
- 258 of 612 MLLT3 patients (42%) have ALL.
This matters for this project specifically, because the Iqbal Lab's index disease is infant B-ALL and this graph repeatedly treats MLLT3 as the myeloid comparator to MLLT1's lymphoid. On patient counts, MLLT3 is the third commonest infant ALL partner, and the infant-ALL MLLT3 group in this archive (157) is comparable to the infant-ALL MLLT10 group (59) and EPS15 group (22) combined, several times over.
A second, sharper consequence: essentially all of the mechanistic literature on this fusion is myeloid - the murine MLL-AF9 GMP assay, THP-1, MOLM-13, NOMO-1, Mono-Mac-6, every one monocytic. There is no widely used t(9;11) B-ALL model, although two B-ALL cell lines exist (see ana-t-9-11-cell-line-panel-accessions-fetched-and-read-back). So 42% of this partner's patients have a disease for which its experimental system does not exist.
Outcome, in one paragraph, deferring to the entry that owns it
The favourable-prognosis claim for t(9;11) originates in 298 children at a single institution, where 15 t(9;11) patients had EFS 66% against 25.9% for the other 27 KMT2A-rearranged cases 4. In 756 children with centrally reviewed karyotypes from 11 collaborative groups, overall 5-year EFS for 11q23/MLL-rearranged AML was 44% (+/- 5%) with subgroup EFS spanning 11% to 92%; the independently prognostic subgroups were t(1;11) (HR 0.1), t(6;11) (HR 2.2), t(10;11)(p12) (HR 1.5) and t(10;11)(p11.2) (HR 2.5), and the authors state plainly that they could not confirm the favourable prognosis of the t(9;11) subgroup 2. In adults, a 205-patient KMT2A-rearranged AML analysis found overall survival similar across translocations 5. In non-infant childhood ALL, t(9;11) T-ALL was the worst reported partner-by-lineage cell 6.
The current position is therefore: not favourable, not adverse, and specifically not one of the four subgroups that carry independent prognostic weight. I-BFM's operational expression of that is the non-high-risk assignment, where non-high-risk means 5-year EFS 54.0% and OS 70.5% - better than high risk (30.3% and 49.2%) and a long way from good 3.
| metric | fusion-partner distribution by age band and lineage |
|---|---|
| value | 612 |
| unit | genotyped KMT2A::MLLT3 patients |
| population | KMT2A-rearranged acute leukaemia |
| region | international referral archive |
| year | 2023 |
| source | PMID 37019990 Table 1, row 2, read from PMC10169636 on 2026-09-20 |
| cohort total | 3401 |
| t all cases | 7 |
| mpal cases | 8 |
| largest single cell | infant ALL, n=157 |
| PCT of mllt3 patients with all | 42.2 |
| outcome position | not favourable and not adverse; the favourable claim was not confirmed in 756 children and I-BFM assigns MLLT3 to the non-high-risk group (5-year EFS 54.0%, OS 70.5%, CIR 35.2%) |
| outcome entry | epi-partner-mllt3-t9-11-favourability-claim-and-its-non-replication |
by age and lineage — the numbers behind the chart
{
"infant": {
"ALL": 157,
"AML": 56,
"other": 5
},
"paediatric": {
"ALL": 91,
"AML": 170,
"other": 3
},
"adult": {
"ALL": 8,
"AML": 108,
"other": 3
},
"unclassified": {
"ALL": 2,
"AML": 5,
"other": 4
}
}totals — the numbers behind the chart
{
"ALL": 258,
"AML": 339,
"other": 15
}share of stratum PCT — the numbers behind the chart
{
"infant_ALL": 16,
"infant_AML": 28.4,
"paediatric_ALL": 17.2,
"paediatric_AML": 36.6,
"adult_ALL": 1.2,
"adult_AML": 24.5,
"all_ALL": 11.8,
"all_AML": 30.4
}- Menin-inhibitor eligibility is partner-agnostic but trial populations are not lineage-agnostic: AUGMENT-101 was 83-86% AML, while 42% of genotyped KMT2A::MLLT3 patients have ALL, so the treated t(9;11) population is not the epidemiological one.
- Extrapolating a t(9;11) result from a monocytic model to an infant-ALL t(9;11) patient crosses both lineage and age, and this partner's patient population sits mostly on the far side of that gap from its model systems.
- Do infant-ALL t(9;11) patients differ from paediatric-AML t(9;11) patients in breakpoint position within MLLT3? The archive holds both variables and the cross-tabulation has not been published.
- Is the 42% ALL share of this partner reproduced in a population-based registry, or is it a feature of referral for breakpoint characterisation?
- Why is there no t(9;11) B-ALL experimental system in common use when 248 of 612 genotyped patients have ALL?
- E1established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.paper
Among 3,401 genotyped KMT2A-rearranged acute leukaemia patients, KMT2A::MLLT3 accounts for 612 cases distributed as infant ALL 157 / AML 56 / other 5, paediatric ALL 91 / AML 170 / other 3, adult ALL 8 / AML 108 / other 3 and 11 unclassified; MLLT3 is 16.0% of infant ALL, 28.4% of infant AML, 17.2% of paediatric ALL, 36.6% of paediatric AML, 1.2% of adult ALL and 24.5% of adult AML, 11.8% of all ALL and 30.4% of all AML, with 7 KMT2A-r T-ALL cases and 8 MPAL cases.
The KMT2A recombinome of acute leukemias in 2023. · Meyer C, Larghero P, Almeida Lopes B, et al. · Leukemia 2023What this evidence cannot show
Read in full text from the PMC HTML page (PMC10169636) on 2026-09-20. This is a diagnostic referral archive accumulated 2003-2022 at one centre, not a population-based registry: shares reflect which cases were sent for breakpoint characterisation, and partners that standard RT-PCR panels miss are plausibly over-represented. Age bands are infant <=12 months, paediatric >12 months to 18 years, adult >18 years. 43 patients in the whole cohort had no age classification and 57 no gender.
Evidence that argues against this (1)
- A referral archive's lineage split can differ from a population's. The earlier recombinome instalment from the same centre, covering 2,345 patients, reported an age-dependent breakpoint shift within KMT2A itself, which shows the cohort composition changes as ascertainment changes and therefore that these shares are not stable population parameters. Meyer C et al. 2018
- E2strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.paper
In 756 children with 11q23/MLL-rearranged AML from 11 collaborative groups with central karyotype review, 5-year event-free survival was 44% (+/- 5%) with subgroup EFS from 11% to 92%; t(1;11)(q21;q23) (HR 0.1), t(6;11)(q27;q23) (HR 2.2), t(10;11)(p12;q23) (HR 1.5) and t(10;11)(p11.2;q23) (HR 2.5) were independent prognostic predictors, and the favourable prognosis of the t(9;11)(p22;q23) subgroup could not be confirmed.
Novel prognostic subgroups in childhood 11q23/MLL-rearranged acute myeloid leukemia: results of an international retrospective study. · Balgobind BV, Raimondi SC, Harbott J, et al. · Blood 2009What this evidence cannot show
Retrospective pooling of 11 trial groups treated on different protocols over different eras; subgroups are not prespecified and multiplicity is not controlled. 'Could not confirm' is a failure to reject, not a demonstration of equivalence, and with t(9;11) as the largest subgroup the study had most power there - which makes the negative more informative than most, but still a negative. Karyotype-defined t(9;11), not molecularly confirmed KMT2A::MLLT3.
Evidence that argues against this (1)
- The founding claim came from a single institution with a uniform protocol, where 15 t(9;11) patients had EFS 66% against 25.9% for other MLL-rearranged cases; a pooled multi-protocol retrospective can dilute a real treatment-specific effect as easily as it can correct an artefact. Rubnitz JE et al. 2002
- E3strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.paper
In 1,130 children with KMT2A-rearranged AML diagnosed 2005-2016, fusion-partner-based assignment to a high-risk group (n=402) versus a non-high-risk group (n=728) gave 5-year EFS 30.3% versus 54.0%, cumulative incidence of relapse 59.7% versus 35.2% and OS 49.2% versus 70.5%, with end-of-induction-2 flow-MRD and high-risk group independently associated with inferior outcome in multivariable analysis.
Measurable Residual Disease and Fusion Partner Independently Predict Survival and Relapse Risk in Childhood KMT2A-Rearranged Acute Myeloid Leukemia: A Study by the International Berlin-Frankfurt-Münster Study Group. · van Weelderen RE, Klein K, Harrison CJ, et al. · J Clin Oncol 2023What this evidence cannot show
The abstract does not name which partners fall in which risk group; this graph's `epi-partner-mllt3-t9-11-favourability-claim-and-its-non-replication` records MLLT3 as non-high-risk from that source. Flow-MRD at both timepoints was available for only 456 of 1,130 patients, so the MRD analysis is a subgroup. Retrospective, multi-protocol, and allo-SCT allocation was not randomised.
Evidence that argues against this (1)
- A partner-based risk dichotomy that puts MLLT3 in the better group still leaves that group with 35.2% cumulative incidence of relapse, so 'non-high-risk' is a relative statement inside a disease where a third of the better half relapse. van Weelderen RE et al. 2023
- E4moderate confidence 3 of 5: One strong study, or several smaller consistent ones.paper
The t(9;11) favourable-prognosis claim originates in a single-institution series of 298 children with primary AML in which t(9;11) was reported to have a favourable impact.
Favorable impact of the t(9;11) in childhood acute myeloid leukemia. · Rubnitz JE, Raimondi SC, Tong X, et al. · J Clin Oncol 2002What this evidence cannot show
Single institution, single era, small subgroup; this entry cites it as the origin of a claim rather than as support for it, and the detailed numbers are recorded in `epi-partner-mllt3-t9-11-favourability-claim-and-its-non-replication` which owns this question.
Evidence that argues against this (1)
- The much larger international retrospective with central karyotype review could not confirm it. Balgobind BV et al. 2009
- E5moderate confidence 3 of 5: One strong study, or several smaller consistent ones.paper
In 205 adults with KMT2A-rearranged AML from the HARMONY database, overall survival was similar across the different translocations.
Rearrangements involving 11q23.3/KMT2A in adult AML: mutational landscape and prognostic implications - a HARMONY study. · Hernández-Sánchez A, González T, Sobas M, et al. · Leukemia 2024What this evidence cannot show
Registry data, retrospective, adult only, and a null across translocations in 205 patients spread over many partners has little power to exclude a moderate partner effect.
Evidence that argues against this (1)
- An underpowered null across many small subgroups is the expected result whether or not a partner effect exists, so this cannot be read as positive evidence that partner does not matter in adults. Hernández-Sánchez A et al. 2024
- E6moderate confidence 3 of 5: One strong study, or several smaller consistent ones.paper
In non-infant childhood KMT2A-rearranged ALL from an international retrospective study, outcomes differed by partner and lineage cell, with t(9;11) T-ALL the worst reported.
Outcomes of Childhood Noninfant Acute Lymphoblastic Leukemia With 11q23/KMT2A Rearrangements in a Modern Therapy Era: A Retrospective International Study. · Attarbaschi A, Möricke A, Harrison CJ, et al. · J Clin Oncol 2023What this evidence cannot show
Very small cells: this graph's existing entry records n=9 for the t(9;11) T-ALL cell with 5-year EFS 41.7% (+/- 17.3), a confidence interval wide enough to include most of the range. Cited here for the direction of the signal, not its magnitude.
Evidence that argues against this (1)
- With n=9 and a standard error of 17.3 points, this cell is compatible with outcomes both far worse and no different from the rest of the cohort; treating it as evidence that t(9;11) is adverse in T-lineage would be over-reading it. Attarbaschi A et al. 2023
age and lineage distribution over 612 genotyped cases
commonest partner in AML at every age band
157 infant ALL cases, the partner's largest single cell
verified the 756-patient non-confirmation from the primary record
- analogous to →Who the 460 KMT2A::MLLT1 patients actually are: 88% lymphoblastic, 53% infants, and only 47 cases of AML in the whole series
the same table, the paralogous partner
- ← related toMLLT3 is the commonest partner in therapy-induced KMT2A leukaemia: 49 of 142, and 47 of those are t-AML
the therapy-induced slice of the same archive
nine myeloid lines against a partner population that is 42% lymphoid
the lineage-matched model for the partner's largest patient cell