The prevention window: the break happens in a fetus, and we know when it CLOSES but not when it opens
also called in utero window problem · when could a preventive agent be given · gestational timing of the KMT2A break
Any plan to stop the KMT2A break from happening has to happen before the baby is born, because the fusion is already there at birth. What nobody knows is how early in pregnancy the break occurs, which means nobody can say when a preventive drug would have to be given.
The same thing, for a scientist
This entry exists because every discussion of preventing the KMT2A break skips the scheduling question. The window's **closing** is established: clonotypic KMT2A::AFF1 genomic junctions are recoverable from the patient's own archived neonatal blood spot, so the translocation is present at birth in children later diagnosed between five months and two years [E1]. The window's **opening** is not established by any direct measurement. The nearest bounds are indirect. Concordant infant twin leukaemias carry the same unique, non-constitutive KMT2A rearrangement, proving a single in-utero founding cell seeded the co-twin through the shared monochorionic circulation [E2], and intraclonal chromosomal diversification in such twins is itself datable to the prenatal period [E3] - so the event precedes the shared circulation and leaves time for divergence before birth. The tissue bound is developmental: definitive haematopoiesis in the human fetal liver runs through gestation with a documented shift in haematopoietic composition and in HSC/MPP differentiation potential as gestation proceeds [E4], so the cell of origin exists only within a gestational interval and the break must fall inside it. **The honest summary is an interval that is open at one end: sometime after fetal liver haematopoiesis is established, and before birth.** A preventive agent would have to be given across most of a pregnancy, to a woman with no identifiable risk factor, for a disease affecting roughly one infant in 38,000.
Why an unknown opening is worse than a late one1 deeper
If the break were known to occur in, say, a two-week interval, a prophylactic would need to cover two weeks. Because the opening is unknown, coverage must span from the establishment of fetal liver haematopoiesis to term - the majority of gestation, and the majority of organogenesis. Every toxicity argument in hyp-block-kmt2a-break-machinery-upstream is multiplied by that duration. A drug that is acceptable for two weeks in the third trimester is a different proposition from the same drug across the second trimester.
There is a second, subtler cost. Because nobody knows when the break happens, nobody can design an observational study that would detect it happening - the sampling schedule is undefined. This is why the field has backtracking studies, which observe the consequence at birth, and no prospective studies, which would observe the event.
What would date it
Three approaches, none of which has been applied to KMT2A.
- Molecular clock on the clone. The number of somatic mutations accumulated in the fusion-positive clone by birth, against a known fetal mutation rate, would date the founding cell. Infant KMT2A-rearranged ALL has one of the lowest somatic mutation burdens of any sequenced cancer, with a mean of 1.3 non-silent mutations in the predominant clone, which makes this hard but also means the number is small enough to count exactly.
- Twin divergence. Concordant twins who diverge chromosomally before birth 3 bound the event before the divergence; systematic subclone dating in twin pairs would narrow it.
- Serial fetal sampling in a model. In a humanised fetal-liver system, staged exposure with a rearrangement readout would establish which gestational stage is permissive.
exp-human-cd34-kmt2a-breakage-assayis the readout; the staged fetal material is the hard part.
The one place the window is reachable
Therapy-related KMT2A-rearranged leukaemia. There the break occurs during scheduled, supervised exposure to a topoisomerase II poison, at a known time, in a patient already under care, with a measurable counterfactual. Everything that makes the infant window unreachable is reversed. This graph's mechanism-top2-cleavage-at-kmt2a already asks whether a TOP2B-sparing poison could be designed, and mechanism-tdp2-repair-of-trapped-top2-complexes records why the obvious repair-inhibition moves would make things worse.
The framing that follows: prevention research in this disease should be done where the window is open, and the mechanistic knowledge exported to the infant case only if and when a target that is safe in a fetus ever appears.
| window closes | at birth - clonotypic junction present in the neonatal blood spot (E1) |
|---|---|
| window opens | UNKNOWN; no direct measurement dates the founding translocation within gestation |
| latency after birth | diagnosis between 5 months and 2 years in the founding backtracking series; twin presentations as early as 7 weeks |
| consequence for intervention | a prophylactic would have to cover most of gestation, in an unselected pregnant population, for a disease incidence of roughly 1 in 38,000 infants |
| window that IS reachable | therapy-related KMT2A-rearranged leukaemia, where the exposure is scheduled and supervised |
indirect bounds
[ "before the shared monochorionic circulation in concordant twins (E2)", "early enough for chromosomal subclone divergence to occur prenatally (E3)", "within the gestational interval in which the fetal liver haematopoietic compartment exists (E4)" ]
ways to date it that have never been tried
[ "molecular clock on the fusion-positive clone using somatic mutation count at birth", "systematic subclone dating across concordant twin pairs", "staged gestational exposure in a humanised fetal liver system with a rearrangement readout" ]
- An unknown window opening means a prophylactic must cover most of gestation, including organogenesis, which converts any fetal toxicity from a manageable risk into a disqualifying one.
- Because the event cannot be dated, no prospective study can be designed to observe it, so the evidence base for any preventive claim will stay retrospective and therefore weak.
- Publicising a prenatal 'window' invites maternal behaviour change during pregnancy with no validated target, which is the exact harm this graph's exposure-avoidance entry judged not feasible.
- A clone seeded in utero can spontaneously clear, so even a successful measurement of fusion incidence at birth would overestimate the preventable disease burden.
- Can the founding translocation be dated within gestation by counting somatic mutations in the fusion-positive clone at birth against a fetal mutation rate? With a mean of 1.3 non-silent mutations the count is small enough to be exact and nobody has attempted it.
- Does the permissive compartment for KMT2A::AFF1 exist throughout fetal liver haematopoiesis or only during part of it? The atlas shows the compartment changes across gestation; nobody has asked which stage is permissive.
- Is there any archived serial fetal sampling - amniocentesis, chorionic villus, fetal blood - in a child later diagnosed with KMT2A-rearranged infant leukaemia? A single such case would date one event directly, and such samples exist in biobanks.
- Would a prospective cord-blood screen for KMT2A fusion in a birth cohort be technically feasible at the required sensitivity? The graph's `epi-fusion-positive-newborns-versus-leukemia-risk` records that this has never been done for KMT2A and that the ETV6::RUNX1 precedent may not transfer.
- E1established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.paper
Patient-specific clonotypic KMT2A::AFF1 genomic breakpoint junctions were recovered from archived neonatal blood spots of children later diagnosed with infant leukaemia, establishing that the translocation is present at birth.
Backtracking leukemia to birth: identification of clonotypic gene fusion sequences in neonatal blood spots · Gale KB, Ford AM, Repp R, et al. · Proceedings of the National Academy of Sciences of the United States of America 1997What this evidence cannot show
A small backtracking series using a clone-specific nested PCR that can only detect the junction it was designed around. It fixes the closing of the window and says nothing about its opening.
Evidence that argues against this (1)
- Presence at birth does not universally mean a short postnatal interval: the same backtracking approach demonstrated protracted postnatal natural histories in childhood leukaemia, with clonotypic sequences recovered from birth spots of children diagnosed many years later. For KMT2A the latency is months, which is the contrast that matters here, but the general inference from 'present at birth' to 'no postnatal window' does not hold across subtypes. Maia AT et al. 2004
- E2established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.paper
Concordant infant twin leukaemias carry the same unique, non-constitutive KMT2A rearrangement, which can only arise if a single founding cell in one fetus seeded the co-twin in utero.
In utero rearrangements in the trithorax-related oncogene in infant leukaemias · Ford AM, Ridge SA, Cabrera ME, et al. · Nature 1993What this evidence cannot show
Three twin pairs. Establishes that the event precedes the shared monochorionic circulation, which is an upper bound on lateness, not a date.
Evidence that argues against this (1)
- Concordance in infant KMT2A-rearranged twins is not universal, which complicates any inference from twin data about timing or sufficiency: a fusion-positive preleukaemic clone was documented in an unaffected co-twin and then spontaneously cleared. A clone can be seeded in utero and then disappear, so the presence of the event before birth does not fix either its date or its consequence. Chuk MK et al. 2009
- E3strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.paper
Concordant infant twin leukaemia can show a shared stem line with extensive divergent chromosomal subclones, dating the onset of intraclonal diversification to the prenatal period.
Prenatal chromosomal diversification of leukemia in monozygotic twins · Kempski H, Mensa-Bonsu KA, Kearney L, et al. · Genes, chromosomes & cancer 2003What this evidence cannot show
Cytogenetic subclone divergence dates diversification, not the founding translocation, and twin pairs are a special and rare setting. It narrows the interval from one side only.
Evidence that argues against this (1)
- Prenatal diversification argues the founding event is early enough for subclones to accumulate, but infant KMT2A-rearranged ALL has one of the lowest somatic mutation frequencies of any sequenced cancer, with a mean of 1.3 non-silent mutations in the predominant clone. Very little sequence-level divergence therefore accumulates before diagnosis, which limits how far any clock argument built on mutation accumulation can date the event. Andersson AK et al. 2015
- E4strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.paper
Human fetal liver definitive haematopoiesis, profiled by single-cell transcriptomics across roughly 140,000 liver cells, shows a shift in haematopoietic composition during gestation away from being predominantly erythroid, accompanied by a parallel and functionally validated change in the differentiation potential of HSC/MPPs.
Decoding human fetal liver haematopoiesis · Popescu DM, Botting RA, Stephenson E, et al. · Nature 2019What this evidence cannot show
A developmental atlas, cited here only to establish that the fetal liver haematopoietic compartment changes substantively across gestation. It says nothing about KMT2A, about breakage, or about which gestational stage is permissive for the translocation.
Evidence that argues against this (1)
- Using a developmental atlas to bound a break window assumes the target cell is the fetal liver HSC/MPP compartment, and that assumption is itself a position rather than a measurement. The graph's own cell-of-origin entries place the initiating compartment at the fetal liver LMPP, and a shift in HSC/MPP differentiation potential across gestation could equally mean the permissive window is narrow OR that the compartment persists throughout - the atlas cannot distinguish these. Popescu DM et al. 2019
the scheduling constraint that bounds every version of this hypothesis
- depends on →In-utero origin, proven: clonotypic KMT2A::AFF1 junctions found in the patient's own newborn blood spot
backtracking fixes the closing of the window at birth
latency after birth is months; the interval before birth is undated
- depends on →Twin concordance: near-total for infant KMT2A-rearranged leukaemia, about 5% for childhood ALL overall
twin data bound the event before the shared monochorionic circulation
the compartment's gestational existence is the only tissue-level bound on when the break can occur