Two groups invoke LIN28B to explain the SAME epidemiological fact in OPPOSITE directions - and neither claim is the one this graph has been repeating
also called why is KMT2A::AFF1 an infant leukaemia · the LIN28B epidemiological argument · fetal factor and short latency
KMT2A::AFF1 leukaemia starts in the womb but babies usually get ill only after they are born, and almost never later in childhood. One group says a fetal gene called LIN28B is what makes the leukaemia possible; another says the very same gene is what holds it back until birth, when the gene switches off. Both are explaining the same timing puzzle with the same molecule, pointing opposite ways.
The same thing, for a scientist
The defining epidemiology of KMT2A::AFF1 is: **initiated in utero, presenting overwhelmingly in infancy, with a very short latency and few or no cooperating mutations.** A fetal-restricted permissive factor is the obvious candidate explanation, and **LIN28B has been proposed for that role twice, with opposite signs.** **Direction 1 (permissive).** Ling et al.: human fetal liver CD34+ cells fail to transform on KMT2A::AFF1 induction without LIN28B; *'LIN28B activity may help explain why KMT2A::AFF1 leukaemias are so aggressive, making it a potential target'* [E1]. **Direction 2 (protective).** Okeyo-Owuor et al., on MLL-ENL: LIN28B suppressed AML initiation; *'This may explain why MLL rearrangements often occur before birth in human infant leukemia patients, but transformation usually does not occur until after birth, when LIN28B levels decline'* [E2]. Eldeeb et al. reach the same framing - *'developmentally restricted expression of LIN28B provides a layer of protection against MYB-dependent AML'* [E3]. **And a precision this graph has been getting wrong.** Ling et al.'s own sentence explains **aggressiveness**, not infant-restriction. The abstract nowhere claims LIN28B explains *why the disease is an infant disease*. That inference is the reader's, and this beat found it stated as though it were the paper's in more than one place - including in the brief that dispatched this beat.
The two quotations side by side, because the symmetry is the finding
| Ling 2026 (human, B-lineage, KMT2A::AFF1) | Okeyo-Owuor 2019 (mouse, myeloid, MLL-ENL) | |
|---|---|---|
| epidemiological fact being explained | KMT2A::AFF1 iALL is initiated in utero, very poor prognosis, few cooperating mutations | MLL rearrangements arise before birth but transformation usually follows birth |
| role assigned to LIN28B | permissive - required for transformation to occur at all | protective - suppresses initiation; transformation waits for LIN28B to decline |
| the authors' own words | 'LIN28B activity may help explain why KMT2A::AFF1 leukaemias are so aggressive, making it a potential target in LIN28B-expressing leukemias' | 'This may explain why MLL rearrangements often occur before birth in human infant leukemia patients, but transformation usually does not occur until after birth, when LIN28B levels decline' |
| what the explanandum actually is | aggressiveness, not infant-restriction | the delay between rearrangement and transformation |
Note the third row against the fourth. The two groups are not even explaining the same thing. Ling et al. explain severity; Okeyo-Owuor et al. explain timing. Those are compatible claims about different explananda that have been read as competing claims about one - and the competition is real only if one also assumes LIN28B is doing the same job in both lineages, which this beat's hyp-fetal-state-is-permissive-for-lymphoid-and-protective-for-myeloid-kmt2a-transformation treats as an open question.
What a fetal-restricted factor can and cannot explain about infant leukaemia
It CAN explain, in principle:
- Why the disease needs no cooperating mutations: the fetal cell supplies self-renewal and proliferative competence that an adult cell would need a second hit to acquire.
- Why latency is short: the permissive window and the initiating lesion coincide in time.
- Why the same fusion in an older child behaves differently: this graph already records that fetal-specific expression is retained in infant KMT2A::AFF1 ALL and NOT in childhood KMT2A::AFF1 ALL.
It CANNOT explain, on current evidence:
- Why the disease presents after birth rather than in utero. Congenital leukaemia is rare; most KMT2A-rearranged infants present in the first year of life, not at delivery. A purely permissive fetal factor predicts in utero disease, and the protective reading (direction 2) predicts the observed postnatal presentation better. This is the strongest argument for direction 2 and it is an epidemiological argument, not an experimental one.
- Why some fetuses with the rearrangement never develop leukaemia. A permissive factor present in every fetus cannot by itself select which one transforms.
- Why KMT2A::AFF1 specifically. LIN28B is fetal in every fetus and every lineage; the fusion is what is rare.
The honest position: a fetal-restricted permissive factor is a necessary-condition explanation, and the epidemiology needs a sufficient-condition explanation as well. Nothing in this beat supplies one.
The precision error, named so it can be fixed
Claim as it has circulated in and around this project: Ling et al. propose a fetal-restricted permissive factor as the explanation for why KMT2A::AFF1 is an infant leukaemia.
What the abstract actually says: 'Thus, LIN28B has an essential role in normal human fetal B-lymphopoiesis, and is necessary for the initiation of KMT2A::AFF1 iALL in human fetal cells in the absence of co-operating mutations. LIN28B activity may help explain why KMT2A::AFF1 leukaemias are so aggressive, making it a potential target in LIN28B-expressing leukemias.'
The difference: the paper claims necessity for initiation in fetal cells, and offers LIN28B as an explanation for aggressiveness. It does not claim LIN28B explains the age restriction of the disease. The age-restriction inference is available and reasonable, but it is an inference, and this graph should attribute it to itself rather than to the paper.
This matters because the age-restriction claim is the one that would make LIN28B a therapeutically selective target, and it is the weaker of the two. It is also, unhelpfully, the claim that direction 2 contradicts most directly. Attributing the stronger-sounding version to the source hides both problems at once. The full text is paywalled and absent from PMC (verified 2026-09-23), so the paper's own discussion of age restriction - if it has one - is unreadable here.
| what ling et al actually claim | necessity for initiation IN FETAL CELLS, and LIN28B as a candidate explanation for AGGRESSIVENESS. The abstract does NOT claim LIN28B explains why the disease is age-restricted to infancy. |
|---|---|
| the attribution error | The age-restriction explanation has been attributed to Ling et al. in and around this project, including in the brief that dispatched this beat. It is a reasonable inference and it is the graph's own, not the paper's. The distinction matters because the age-restriction claim is what would make LIN28B selectively targetable, and it is the claim direction 2 contradicts most directly. |
| what neither direction explains | Why THIS fetus and not that one. A factor present in every fetus is a necessary-condition explanation; the epidemiology also needs a sufficient-condition explanation, and this beat found none. |
| read depth | E1 abstract only and paywalled; E2, E3 abstract; E4 full text |
| shown for fusion | Direction 1: KMT2A::AFF1, human. Direction 2: KMT2A::MLLT1 (MLL-ENL), mouse. No study has tested both directions in one system. |
epidemiological facts to be explained
[ "KMT2A::AFF1 infant ALL is initiated in utero", "presentation is overwhelmingly in the first year of life, NOT at birth - congenital/fetal leukaemia is exceedingly rare", "latency between rearrangement and disease is short compared with adult KMT2A-rearranged leukaemia", "few or no cooperating driver mutations, unlike childhood ALL", "very poor prognosis" ]
direction 1 permissive
{
"claim": "LIN28B is required for KMT2A::AFF1 to transform a human fetal cell; the fetal state supplies what would otherwise need a second hit",
"explains": "the absence of cooperating mutations; possibly aggressiveness",
"does_not_explain": "why disease presents after birth rather than in utero; why most rearranged fetuses never transform",
"source": "PMID 42213620",
"system": "human fetal liver CD34+, endogenous CRISPR translocation, B-lineage"
}direction 2 protective
{
"claim": "LIN28B suppresses initiation; transformation waits until LIN28B declines postnatally",
"explains": "the in-utero-rearrangement / postnatal-presentation gap; the rarity of fetal leukaemia",
"does_not_explain": "why LIN28B is required for transformation in the human B-lineage system; and its own necessity arm failed (fetal Lin28b deletion did not accelerate AML)",
"source": "PMID 31405949, PMID 36763502",
"system": "mouse transgenic MLL-ENL, myeloid"
}- If direction 2 is right even partly, then a LIN28B inhibitor given to an infant is removing a brake, and the intervention that follows from the biology is LIN28B restoration rather than inhibition - an opposite-signed programme with an opposite modality.
- A therapeutic rationale resting on age restriction is resting on the weaker of the two claims in the source paper, and on an inference the paper's abstract does not make.
- A fetal-restricted target is self-limiting: it disappears as the patient ages. A relapse after the programme has lapsed would be untouched by a drug against it, which is the worst possible timing for a disease whose relapses are its main cause of death.
- When exactly does LIN28B fall in HUMAN haematopoiesis, and does the fall coincide with the peak incidence of KMT2A::AFF1 presentation? Both directions make predictions about this curve and neither has plotted it. Human fetal-to-infant HSPC atlases exist and this is a re-analysis, not an experiment.
- Does the full text of Ling et al. make an age-restriction argument? Unreadable here because the paper is paywalled and not in PMC. Resolving it would settle whether this entry's attribution correction is needed or is itself an artefact of reading only the abstract.
- Why is congenital KMT2A::AFF1 leukaemia rare if the fetal state is permissive? This is the single cleanest observation favouring direction 2, and no entry in this graph addresses it.
- Is there an epidemiological dataset with age-at-presentation resolved finely enough within the first year to test whether incidence tracks the decline of a fetal programme? SEER-type registries may be too coarse, and this beat did not look.
- E1strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.paper
Ling et al. state that LIN28B has an essential role in normal human fetal B-lymphopoiesis and is necessary for the initiation of KMT2A::AFF1 infant ALL in human fetal cells in the absence of co-operating mutations, and that LIN28B activity may help explain why KMT2A::AFF1 leukaemias are so aggressive, making it a potential target in LIN28B-expressing leukaemias; they frame infant ALL as initiated in utero, most often by KMT2A rearrangement, carrying a very poor prognosis despite a lack of the additional oncogenic driver mutations common in childhood ALL.
The fetal specific gene LIN28B is essential for human fetal B-lymphopoiesis and initiation of KMT2A::AFF1 infant leukemia. · Ling RE, Welsh AM, Hamer L, Jackson T, Neil E, Elliott N, et al. · Blood 2026What this evidence cannot show
ABSTRACT ONLY and paywalled - Europe PMC returns isOpenAccess N, inEPMC N, inPMC N, verified live 2026-09-23. The quoted sentences are the abstract's closing claims, reproduced verbatim rather than paraphrased, precisely because this entry is about what the paper does and does not claim. The paper's discussion section, where an age-restriction argument would normally appear, could not be read - so the absence of an age-restriction claim is an absence FROM THE ABSTRACT, not from the paper.
Evidence that argues against this (1)
- The same gene is invoked with the opposite sign to explain the same class of epidemiological observation in mouse myeloid MLL-ENL. Blood Adv 2019
- E2strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.paper
Okeyo-Owuor et al. state that LIN28B, tested as a candidate accelerator of MLL-ENL leukaemogenesis, surprisingly suppressed AML initiation rather than accelerating it, and that this may explain why MLL rearrangements often occur before birth in human infant leukaemia patients but transformation usually does not occur until after birth, when LIN28B levels decline; AML initiation was most efficient when MLL-ENL was induced neonatally, and even transient suppression of MLL-ENL in neonates could prevent AML in most mice.
The efficiency of murine MLL-ENL-driven leukemia initiation changes with age and peaks during neonatal development. · Okeyo-Owuor T, Li Y, Patel RM, Yang W, Casey EB, Cluster AS, Porter SN, Bryder D, Magee JA · Blood Adv 2019What this evidence cannot show
Abstract read; full text not opened. MOUSE, TRANSGENIC MLL-ENL, MYELOID - so the epidemiological claim it makes about HUMAN infant leukaemia is an extrapolation from a mouse model, exactly as the competing claim is an extrapolation from a human in vitro model. Neither is an epidemiological study. The abstract also reports that transplantation stress mitigated the developmental barrier, which weakens the argument that the fetal window is intrinsically restrictive.
Evidence that argues against this (1)
- The epidemiological claim here is an extrapolation from a mouse transgenic myeloid model to human infant ALL, which crosses species, lineage, partner and lesion type at once - exactly the importation this project repeatedly documents as unreliable. It is not an epidemiological study and should not be weighed as one. Blood 2026
- E3strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.paper
Eldeeb et al. propose that developmentally restricted expression of LIN28B provides a layer of protection against MYB-dependent AML, and describe the influence of prenatal signals on leukaemogenesis - such as those mediated by the fetal-specific protein LIN28B - as controversial.
A fetal tumor suppressor axis abrogates MLL-fusion-driven acute myeloid leukemia. · Eldeeb M, Yuan O, Guzzi N, Thi Ngoc PC, Konturek-Ciesla A, Kristiansen TA, Muthukumar S, Magee J, Bellodi C, Yuan J, Bryder D · Cell Rep 2023What this evidence cannot show
Abstract read; full text not opened. Mouse, transgenic MLL-ENL, myeloid, supraphysiological LIN28B. The word 'controversial' in the authors' own abstract is quoted here because it is the field's own assessment of exactly the question this entry is about, and it predates the Ling result by three years.
Evidence that argues against this (1)
- This is the same collaboration as E2 and largely the same model, so the two are not independent corroborations of the protective reading. Counting them as two lines of evidence overstates the weight on that side of the contradiction. Blood Adv 2019
- E4strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.paper
Li et al. state that a majority of infant leukaemias are driven by MLL rearrangements arising before birth, that infant leukaemias harbour few additional cooperating mutations, and that the threshold of genetic change required for infant leukaemia initiation and the time required to cross it are both low compared with MLL-rearranged adult leukaemias - observations they take to suggest MLL rearrangements transform fetal haematopoietic progenitors more efficiently than adult progenitors; they also note that fetal leukaemias are exceedingly rare despite prenatal mutation acquisition.
LIN28B promotes differentiation of fully transformed AML cells but is dispensable for fetal leukemia suppression. · Li Y, Mendoza-Castrejon J, Patel RM, Casey EB, Denby E, Bryder D, Magee JA · Leukemia 2024What this evidence cannot show
FULL TEXT READ (PMC10912017). These are the paper's framing statements in its introduction and discussion rather than its own measurements, so they carry the field's consensus rather than new data. The sentence that does the most work here - 'fetal leukaemias are exceedingly rare despite prenatal mutation acquisition' - is the epidemiological observation that the protective reading explains and the permissive reading does not.
Evidence that argues against this (1)
- These are framing statements from an introduction and discussion, carrying field consensus rather than new measurement, and the paper's own data contradict the strongest version of that consensus - its fetal Lin28b deletion did not accelerate AML. Framing sentences in a paper are the least-reviewed part of it. Leukemia 2024
the epidemiology being explained: in-utero initiation, postnatal presentation, short latency, no cooperating mutations
invoked with opposite signs by two groups for the same class of observation
the epidemiological contradiction the lineage hypothesis was written to reconcile
fetal expression retained in infant but not childhood KMT2A::AFF1 ALL is the graph's existing support for the permissive reading
what the experiment shows versus what it is quoted as showing
- related to →KMT2A::AFF1 (MLL-AF4), t(4;11)(q21;q23)
the index fusion whose age restriction is the explanandum