start / graph / ana-q143-drug-exposure-and-clonal-replacement-are-completely-confounded-across-this-patient-pair

AnalysisstrongQ143-menin-under-drug-in-infants

Drug exposure and clonal replacement are completely confounded across this pair - the relapse sample is a different clone in a different lineage, and the paper's own reading is that the drug killed the cells the diagnostic sample was made of

The two samples are not the same cells before and after a drug. The drug wiped out the B-cell leukaemia; what grew back was a myeloid leukaemia from the same child. So anything that looks different between the two could be the drug, or could simply be that these are different cells.

The same thing, for a scientist

The design that makes this pair valuable is also what limits it. 148752 is diagnostic B-ALL; 152985 is relapse AML from the same child after a lineage switch on revumenib. The authors write: *"This suggests the menin inhibitor effectively eliminated the B-ALL cells by targeting the oncoprotein, yet the leukemia persisted in the AML state."* A fall in any chromatin signal between the two aliquots therefore has **at least four sufficient explanations**: pharmacological displacement; a surviving clone that expresses less oncoprotein (which is the paper's own separate headline about lineage-switching samples generally); a myeloid rather than lymphoid chromatin landscape; and differing sample composition or blast fraction, which is nowhere recorded. This design cannot separate them, and no analysis of these files can.

Go deeper

The paper's own three statements, which together make the confound explicit

"the menin inhibitor effectively eliminated the B-ALL cells by targeting the oncoprotein, yet the leukemia persisted in the AML state" - the surviving population is not the treated population 1.
"In lineage-switching samples the oncoprotein levels are reduced" (abstract) - lineage-switching samples have low oncoprotein independent of drug 1.
The MLL::ENL oncogene is still expressed in 152985, confirmed by RT-PCR with two primer sets, though at lower levels - so the fusion transcript is present while its chromatin signal is not 1. That is the most interesting fact in the pair and the one a pure drug-effect reading would not predict.

What survives the confound and what does not

Does not survive: any statement of the form 'revumenib reduced menin occupancy by X%'. That is a between-clone comparison wearing a pharmacodynamic label.

Survives: statements internal to one sample - menin's enrichment relative to an accessibility reference within 148752, or the ranking of channels within 152985 - and statements about which channel changes relative to the others, since all five channels experienced the same clonal replacement. That is why this beat's drug-arm claims are written as within-sample comparisons between channels rather than as between-sample percentages.

A fifth confound nobody can bound: blast fraction

Neither GEO nor the paper records the blast percentage of either aliquot. A diagnostic infant B-ALL marrow is typically blast-dominated; a relapse sample after therapy may not be. If 152985 contains a larger fraction of normal haematopoietic cells, every fusion-driven signal is diluted by that fraction alone, and the dilution applies to all five channels. This is unbounded and should be assumed to contribute.

Facts and figures

authors own sentence	the menin inhibitor effectively eliminated the B-ALL cells by targeting the oncoprotein, yet the leukemia persisted in the AML state
fusion transcript in relapse	still expressed, confirmed by RT-PCR with two primer sets, at lower levels
what survives	within-sample channel comparisons, and between-channel relative change
what does not	any statement of the form 'revumenib reduced menin occupancy by X%'
status	run

confounded with drug exposure

[
  "clonal replacement",
  "lineage switch B-ALL to AML",
  "intrinsically lower oncoprotein levels in switched samples",
  "unrecorded blast fraction / sample composition"
]

How this could bite us back

The pair will be cited as 'menin occupancy before and after a menin inhibitor in a patient'. It is that only in the loosest sense, and the looser sense is the one that travels.
If a clinical pharmacodynamic assay is designed on this template - biopsy at diagnosis, biopsy at relapse - it inherits this confound by construction. The assay that answers the pharmacodynamic question is an on-treatment biopsy of the SAME disease state, not a relapse sample.

Still open

Are the diagnostic and relapse populations clonally related? No genotyping of the two aliquots is presented, and for this patient the raw sequencing is public so the question is answerable from the deposit.
What was the blast fraction of each aliquot? Unrecorded, and it bounds every dilution argument.
The MLL::ENL transcript is still expressed while its chromatin signal is gone. Is the protein present and displaced, or is it not translated? No protein measurement exists for this sample.

Every source behind this entry (3)

E1strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.paper
The relapse AML sample is a clonally and lineally different population from the diagnostic B-ALL sample, by the authors' own account, and the fusion transcript is still expressed in it.
MLL oncoprotein levels influence leukemia lineage identities · Janssens DH, Duran M, Otto DJ, Wu W, Xu Y, Kirkey D, Mullighan CG, Yi JS, Meshinchi S, Sarthy JF, Ahmad K, Henikoff S · Nature Communications 15:9341 2024
PMID 39472576 DOI
What this evidence cannot show
Clinical narrative plus qPCR/RT-PCR on three biological replicates; no clonal sequencing of the two aliquots is presented, so 'different clone' is inferred from phenotype and from the drug's selective effect rather than measured by genotype.
Evidence that argues against this (1)
- Lineage switch in KMT2A-rearranged leukaemia can occur in a clonally RELATED population rather than an unrelated one - the same rearrangement persists across the switch - so 'different clone' may overstate it; the switched population can be a descendant of the diagnostic one rather than a separate leukaemia. Tirtakusuma R et al. 2022
E2established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.paper
Lineage switching in KMT2A-rearranged B-ALL is an established escape route under lineage-directed pressure and is not specific to menin inhibition.
Acquisition of a CD19-negative myeloid phenotype allows immune escape of MLL-rearranged B-ALL from CD19 CAR-T-cell therapy · Gardner R, Wu D, Cherian S, Fang M, Hanafi LA, Finney O, Smithers H, Jensen MC, Riddell SR, Maloney DG, Turtle CJ · Blood 127:2406-2410 2016
PMID 26907630 DOI
What this evidence cannot show
CD19 CAR-T context, not menin inhibition; establishes the phenomenon and its mechanism class, not its frequency under this drug.
Evidence that argues against this (1)
- Tirtakusuma et al. attribute lineage switching in MLL/AF4 leukaemia to specific epigenetic regulator gene lesions, which would make the switched clone genetically distinguishable at diagnosis rather than a purely drug-driven selection - and this deposit has no genomics on either aliquot to check that. Tirtakusuma R et al. 2022
E3strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.dataset
Neither the blast fraction nor the cellular composition of either aliquot is recorded in GEO or in the paper.
GSE252378_family.soft.gz - the per-sample SOFT records for GSE252378, read raw with line numbers · Janssens DH et al. · NCBI GEO FTP 2024
What this evidence cannot show
An absence over the metadata and the paper's Methods; the Supplementary Information and Source Data were not opened, so a composition table could exist there.
Evidence that argues against this (1)
- The paper's Supplementary Information and Source Data were not opened in this beat, and a per-patient clinical table with blast percentages is exactly the kind of content that lives there - so this absence is bounded by the full text and the metadata only, and should be re-checked before it is quoted as complete. Janssens DH et al. 2024

Connected to

depends on →148752 and 152985 are one patient before and during revumenib, not two treated patients - so the drug-free arm this corpus said did not exist is sample 148752
the confound is a property of the pair
related to →GMP-like state — the destination of therapy-driven lineage switch and, separately, the state most sensitive to menin inhibition
the relapse clone activates the GMP-like program the authors describe
related to →Mixed phenotype acute leukaemia (MPAL) and acute leukaemia of ambiguous lineage
lineage switch is the extreme of the lineage-plasticity spectrum this disease occupies
← depends onIn the revumenib-exposed relapse sample nothing is enriched above a non-specific reference - not menin, not the fusion N-terminus, not ENL - and the sample's own top-ranked promoters are mitochondrial pseudogenes
the clonal-replacement confound
← depends onMenin does not outlast the fusion under revumenib in primary infant material - 0.586 against 0.485 at oncoprotein peaks, a ratio of 1.21 against a pre-registered threshold of 1.5 - so the fusion-independent menin plateau seen after a degron does not reproduce here
the confound bounds this result
← depends onENL redistributes rather than persists: it is the only channel that holds more of its signal at the myeloid-program peaks than at the B-lineage ones, by 2.06x in my instrument and 1.26x in the authors' own - and in absolute terms it falls further than menin does
the lineage switch is what defines the two peak classes
← depends onIn a patient, the menin-fusion association is demonstrable before the drug and unmeasurable after it - and the asymmetry is what this pair establishes, not a rate of menin loss
the confound
← depends onSpike-in-normalised menin CUT&RUN with paired N- and C-terminal KMT2A antibodies in an infant KMT2A::AFF1 sample, on and off drug, in the same disease state
the confound this design removes by splitting one aliquot