Fusion-positive clones that never become leukaemia: persistence in remission, spontaneous clearance, and silent rearrangements in healthy people
also called covert preleukaemic clone · spontaneous regression of preleukaemia · fusion without disease
Carrying the fusion is not the same as having leukaemia. There are people who carry a fusion-positive clone for years without being ill, one documented baby whose fusion-positive clone simply vanished, and evidence that broken-and-rejoined KMT2A DNA turns up in healthy adults without ever producing a working fusion protein.
The same thing, for a scientist
Three independent observations establish that a fusion-positive haematopoietic clone can exist without disease, and each has a different strength. (1) **Persistence in remission**: clonotypic genomic RUNX1::RUNX1T1 sequences remained detectable in complete clinical remission in 3 of 3 patients tested, i.e. the founding lesion outlives the leukaemia [E1]. (2) **Spontaneous clearance**: a clonal KMT2A::MLLT1-positive population in the healthy monozygotic co-twin of an affected infant was serially documented in marrow and blood and then disappeared, temporally associated with virus-induced cytopenias; the co-twin was well three years later [E2]. This is the only published case of its kind and is the single most important datum for the PI's question about whether fusions are in flux. (3) **Silent rearrangement in healthy people**: inverse PCR detected KMT2A rearrangements in peripheral lymphocytes of 49% of 100 healthy adults, predominantly t(4;11)-type, with **no detectable t(4;11) transcript in anyone** [E3] — weak evidence, but if true it relocates the rate-limiting step from breaking the DNA to producing an expressed fusion in a competent cell. The composite picture is a bottleneck between rearrangement and disease that is partly cell-intrinsic and possibly partly immune.
Why the clearance case deserves more attention than it gets1 deeper
Chuk et al. did not merely observe a discordant twin pair; they demonstrated the presence and then the absence of the clone by serial sampling of a healthy child 2. That converts 'discordance' from an absence of evidence into evidence of elimination. The timing coincided with viral illness and cytopenias, which the authors read as immune-mediated clearance. Nothing in the paper measures an immune response, so the mechanism is a hypothesis; but a hypothesis with a documented human end-point is worth more than a mouse model with none. If a fetal or neonatal immune compartment can clear KMT2A-fusion-positive cells, that is an endogenous therapeutic mechanism nobody is trying to reproduce.
What would make this generalisable
One case cannot support a rate. The design that could: systematic, consented, serial screening of the healthy monochorionic co-twins of index infant KMT2A-r cases, which is the only population enriched enough (near-unity prior) to make prospective observation feasible. Even a handful of such co-twins followed with a patient-specific genomic-junction assay would produce the first clearance-versus-progression fraction in humans.
The bottleneck is not the break
The genesis beat established several routes by which KMT2A can be cut: topoisomerase II cleavage, apoptotic nucleases, endonuclease G under replication stress (mechanism-top2-cleavage-at-kmt2a, mechanism-apoptotic-nuclease-cleavage-kmt2a, mechanism-replication-stress-endog-cleavage-kmt2a). If breaks are as common as 3 suggests, then the scarcity of infant leukaemia (about 1 in 30,000 births, epi-fusion-positive-newborns-versus-leukemia-risk) is not explained by the scarcity of breaks. It must be explained downstream: by the break happening in the wrong cell, by the join being out of frame, by the cell dying, or by the clone being cleared. Each of those is a different prevention target, and the field does not currently know which one carries the load.
- If fusion-positive clones can persist in remission, a fusion-transcript MRD assay used to trigger salvage therapy could commit children to transplant on the strength of a clone that was never going to relapse.
- If endogenous immune surveillance clears preleukaemic clones, then profound lymphodepletion during infant induction may remove the control mechanism at the exact moment residual disease is lowest.
- Conversely, if the bottleneck is cell-intrinsic rather than immune, immunotherapy adds nothing at the preleukaemic stage and the resource should go elsewhere.
- Is spontaneous clearance immune-mediated, and if so by which compartment? A single follow-up case with paired immunophenotyping and TCR/NK repertoire sequencing at the time of clearance would transform a curiosity into a therapeutic lead.
- Does a KMT2A::AFF1 clone persist in patients in long-term remission, as RUNX1::RUNX1T1 does [E1]? Patient-specific genomic-junction ddPCR on banked remission samples from Interfant survivors is cheap, answerable now, and would change how MRD positivity is interpreted.
- Are the KMT2A rearrangements reported in healthy adults [E3] real? A modern replication using translocation-capture sequencing rather than inverse PCR, on 100 healthy donors, would either kill the claim or establish that breaking is not the bottleneck.
- Does the reciprocal derivative chromosome get lost during the preleukaemic phase, and does its loss mark the transition to overt disease?
- E1moderate confidence 3 of 5: One strong study, or several smaller consistent ones.paper
Clonotypic genomic fusion sequences persist in patients who are in complete clinical remission, so a fusion-positive clone can coexist with health.
In utero origin of t(8;21) AML1-ETO translocations in childhood acute myeloid leukemia · Wiemels JL, Xiao Z, Buffler PA, Maia AT, Ma X, Dicks BM, et al. · Blood 2002What this evidence cannot show
Only 3 patients assessed, and with RUNX1::RUNX1T1 rather than a KMT2A fusion. Persistence in remission has not been shown for KMT2A::AFF1, where the equivalent observation would be far more clinically consequential.
Evidence that argues against this (1)
- The same backtracking chemistry, applied with an explicit detection limit, mostly returns nothing in AML. In a series using one identical method on neonatal blood spots from 12 non-infant ALL and 13 childhood AML patients, at a sensitivity requiring 1-3 positive cells, clonotypic markers were recovered in 3 of 12 ALL and 0 of 13 AML cases, and the authors state this is equally compatible with a lower clone load as with a lower rate of prenatal origin. Persistence of a clonotypic fusion sequence in remission in 3 patients is therefore a positive finding at the edge of assay capability, and the complementary negative - clones that persist below one cell per sample - is unmeasurable by the same method. Burjanivova T et al. 2006
- E2moderate confidence 3 of 5: One strong study, or several smaller consistent ones.paper
A documented clonal KMT2A::MLLT1-positive population in a healthy infant cleared spontaneously, temporally associated with viral illness and cytopenias.
Discordance of MLL-rearranged (MLL-R) infant acute lymphoblastic leukemia in monozygotic twins with spontaneous clearance of preleukemic clone in unaffected twin · Chuk MK, McIntyre E, Small D, Brown P · Blood 2009What this evidence cannot show
n=1. Immune mediation is inferred from timing alone; no immunological assay is reported. Three years of follow-up cannot exclude later disease. Publication bias favours reporting the surprising discordant pair, so the true discordance rate cannot be read off this.
Evidence that argues against this (1)
- Spontaneous clearance of a clonal neonatal myeloproliferation is documented at scale in a different genotype, and there the mechanism is developmental rather than immune. In a Brazilian series of 77 neonatal leukaemias, 42 congenital cases were associated with Down syndrome, 27 of those underwent spontaneous remission and were reclassified as transient myeloproliferative disorder, and GATA1 mutations were found in 70%. Transient abnormal myelopoiesis regresses as the fetal-liver context that supports it disappears after birth, with no immunological trigger required. That is a competing explanation the single KMT2A::MLLT1 twin case cannot exclude, since the inference to immune mediation there rests on timing alone with no immunological assay reported. Moura SV et al. 2015
- E3weak confidence 2 of 5: Preliminary, small, indirect, or not yet peer reviewed.paper
KMT2A rearrangements were detected by inverse PCR in lymphocytes from 49% of healthy adults, with no detectable t(4;11) fusion transcript in any subject.
MLL leukemia-associated rearrangements in peripheral blood lymphocytes from healthy individuals · Brassesco MS, Montaldi AP, Gras DE, de Paula Queiroz RG, Martinez-Rossi NM, Tone LG, Sakamoto-Hojo ET · Genetics and Molecular Biology 2009What this evidence cannot show
Inverse PCR is highly prone to template-switching artefacts at these frequencies; n=100; low-visibility journal; no independent replication. The negative RT-PCR result is at least internally consistent (a genuine expressed t(4;11) clone at 49% prevalence would be impossible to reconcile with disease rates).
Evidence that argues against this (1)
- The far more stringent cord-blood work on a different fusion found essentially no confirmable positives after sorting, which is a general warning that PCR-based prevalence claims in this area do not replicate. Lausten-Thomsen U et al. 2011
- ← related toRUNX1::RUNX1T1 (AML1-ETO), t(8;21)(q22;q22)
Marker-versus-disease bound (Miyamoto 1996, PMID 8639850): AML1/ETO transcripts persisted in all 18 t(8;21) patients in remission for 12-150 months, in 2.1% of clonogenic colonies across four lineages, so the fusion transcript is a clonal marker rather than a disease state.
Counter-evidence (Greaves 2003, PMID 12791663): the twin-leukaemia literature is dominated by CONCORDANT monozygotic infant pairs, so the single documented clearance case does not establish the natural history of a screen-detected carrier -- neither body of case reports has a denominator
the fusion is stable once disease exists, but a preleukaemic clone can still be cleared
clearance and silent rearrangement both widen the gap between carriers and cases
if breaks are common, the bottleneck to disease lies downstream of cleavage
a true positive may be a child who needed nothing
most fusion-positive newborns never develop leukemia
- related to →How many children this affects
Counter-evidence (Moura 2015, PMID 24991719): 27 of 42 Down-syndrome congenital leukaemias regressed spontaneously and were reclassified as transient myeloproliferative disorder with GATA1 mutations in 70% -- spontaneous clearance of a clonal neonatal myeloproliferation is documented at scale with a DEVELOPMENTAL rather than immune explanation