172 adults with KMT2A-r AML: translocation subtype did not predict survival, so treat all KMT2A-r as adverse
also called Issa 2021 Blood Cancer Journal · MD Anderson KMT2A-r adult AML series
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.
The same thing, for a scientist
A retrospective MD Anderson analysis of 172 adults with KMT2A-rearranged AML, compared with 522 age-matched patients with diploid AML, found that in multivariable analysis translocation subtype did not predict survival, whereas age and allogeneic HSCT did. KMT2A-r AML had worse outcomes than diploid AML both at diagnosis and at relapse, most starkly after second salvage (median OS 2.4 versus 4.8 months, P<0.0001). Therapy-related KMT2A-r AML did worse than de novo (median OS 0.7 versus 1.4 years, P<0.0001), and allo-HSCT in first remission was associated with 5-year OS 52% versus 14% (P<0.0001). The stated conclusion is that AML risk stratification guidelines should include all KMT2A-r as adverse - an explicit endorsement of the class abstraction in adults.
This is the class assumption stated by a clinical group
The menin-interface bet in this graph is a mechanistic class claim. Here is the clinical version of the same claim, arrived at independently from outcome data: partner identity does not carry prognostic information in adult KMT2A-r AML, therefore stratify on the rearrangement, not the partner 1. What makes this entry load-bearing is that it is a negative reported as a conclusion, not a subgroup that nobody examined.
It disagrees with the paediatric data, not with itself
The same partners that show no prognostic signal here separate childhood AML by 24 EFS points 2 and childhood ALL across a 50-point range 3. Two readings survive: either the partner effect is real and adult disease biology or adult treatment intensity swamps it, or the paediatric effect is an artefact of confounding with age, protocol and comutation. Nothing in this beat distinguishes those two readings, and that is the central gap it found.
Where the prognostic weight went instead
Mutation burden carried some of it: patients with two or more comutations had lower OS 1. Subsequent adult series localise it more precisely to KRAS, DNMT3A and TP53 4. In adults, comutation looks like the axis that partner is in children - which is itself a testable claim, because it predicts that a paediatric partner analysis adjusted for RAS-pathway status should shrink.
| metric | overall survival by KMT2A translocation subtype in adults |
|---|---|
| value | no independent effect of translocation subtype in multivariable analysis |
| unit | hazard ratio (not significant) |
| population | adults with KMT2A-rearranged AML |
| n total | 172 |
| comparator cohort | 522 age-matched diploid AML patients |
| region | United States (single institution) |
| year | 2021 |
| prespecified | no - retrospective series |
| source | PMID 34588432 |
independent prognostic factors
[ "age", "allogeneic HSCT in CR1" ]
other findings — the numbers behind the chart
{
"allo_hsct_cr1_5y_os_pct": 52,
"no_allo_hsct_5y_os_pct": 14,
"therapy_related_median_os_years": 0.7,
"de_novo_median_os_years": 1.4,
"second_salvage_median_os_months_kmt2ar": 2.4,
"second_salvage_median_os_months_diploid": 4.8
}- The clinical conclusion drawn here - treat every KMT2A-r as adverse regardless of partner - is the same abstraction the menin-interface bet makes, so the two claims will rise or fall together and cannot be used to corroborate each other.
- Allo-HSCT in CR1 carried a 38-point OS difference in this series; any menin-inhibitor benefit in adults will be measured against a transplant effect far larger than the partner effect, and partner-stratified response will be entangled with transplant eligibility.
- How many adults carried each partner? Without per-arm n the null cannot be converted into an upper bound on the partner effect.
- Does the adult partner null survive adjustment for therapy-related versus de novo status, which is itself strongly partner-associated?
- Is the child-adult discordance a biology difference or a power difference? A pooled individual-patient analysis stratified by age would settle it and does not exist.
- E1strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.paper
In 172 adults with KMT2A-rearranged AML, multivariable analysis showed that translocation subtypes causing the rearrangement did not predict survival, unlike age and allogeneic HSCT, and the authors concluded that all KMT2A-r should be classed as adverse regardless of partner.
Predictors of outcomes in adults with acute myeloid leukemia and KMT2A rearrangements. · Issa GC, Zarka J, et al. · Blood Cancer J 2021What this evidence cannot show
Single-institution retrospective series; 172 patients divided across many partners leaves individual translocation arms small and the per-arm n are not given in the abstract, so this is an unpowered null rather than a demonstration of equivalence. Referral-centre case mix, with half the cohort heavily pretreated. Comparison group is age-matched diploid AML rather than other KMT2A-r cohorts.
Evidence that argues against this (2)
- A 181-patient adolescent and adult KMT2A-r AML cohort found one partner did separate: KMT2A::ELL had superior 3-year OS versus all other subtypes (59.8% versus 39.3%, P=0.023), so the adult partner null is not unanimous. Shen H et al. 2026
- A 125-patient KMT2A-altered AML study proposed a three-tier risk model in which MLLT3/ELL-rearranged cases without PTD had 3-year OS 78.1% against 34.9% for other KMT2A-r (P=0.044), again a partner effect in adults. Chen L et al. 2025
- E2strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.paper
In childhood AML the same question is answered the other way: a partner-based high-risk group of 402 children had 5-year EFS 30.3% versus 54.0% for 728 non-high-risk children, with partner group independent of MRD 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, et al. · J Clin Oncol 2023What this evidence cannot show
Different age group, different cohort, different era mix; the paediatric study is far larger, so its positive finding and the adult null may simply reflect power rather than biology.
Evidence that argues against this (1)
- The HARMONY adult database, with 205 patients, also found OS similar across translocations (t(9;11) P=0.756), so two independent adult cohorts agree with each other and against the paediatric result. Hernández-Sánchez A et al. 2024
- E3strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.paper
In non-infant childhood KMT2A-r ALL, reported 5-year EFS spans 41.7% to 91.2% across partner-by-lineage cells in a 629-patient cohort.
Outcomes of Childhood Noninfant Acute Lymphoblastic Leukemia With 11q23/KMT2A Rearrangements in a Modern Therapy Era: A Retrospective International Study. · Attarbaschi A, Möricke A, et al. · J Clin Oncol 2023What this evidence cannot show
Different lineage and age group from the adult AML series; the extreme cells are small (n=9 and n=34) and the comparisons were not pre-specified.
Evidence that argues against this (1)
- In infants, two separate cohorts found no partner effect on EFS at all, so the paediatric positive is confined to children over 1 year and does not extend across the whole age range. Ilyasova K et al. 2025
- E4strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.paper
In 205 adults with KMT2A-r AML from the HARMONY database, overall survival was similar across translocations including t(9;11)/KMT2A::MLLT3 (P=0.756), while age over 60, secondary AML, DNMT3A mutation and KRAS mutation were the independent prognostic factors in intensively treated patients.
Rearrangements involving 11q23.3/KMT2A in adult AML: mutational landscape and prognostic implications - a HARMONY study. · Hernández-Sánchez A, González T, et al. · Leukemia 2024What this evidence cannot show
Retrospective, assembled from a multi-study database so treatment is heterogeneous; mutational data available for 185 of 205 patients. The per-translocation n are not given in the abstract, so the P=0.756 for t(9;11) is again a null of unstated power.
Evidence that argues against this (1)
- A 159-patient KMT2A-r AML series reported that the KMT2A::MLLT4/AFDN subgroup trended to inferior OS and higher relapse incidence, indicating a residual partner signal in adults once a specific partner is examined rather than tested globally. Wu Y et al. 2025
translocation subtype not predictive in 172 adults; all KMT2A-r classed adverse
adult partner null against a 24-point paediatric partner effect
prognostic weight located in comutation burden rather than partner
a partner effect in adults where the larger series reports none
the largest curated partner null, in a transplanted population