Track the sub-1% cooperating-lesion clones through menin inhibition, because nobody has measured what the drug selects for
also called duplex MRD of RAS and FLT3 subclones under revumenib · does menin inhibition select the RAS-mutant clone?
A menin inhibitor is now approved for this leukaemia, and drug-resistant cells making up less than one cell in three hundred at diagnosis are known to grow back after chemotherapy. Nobody has checked which of those hidden cells a menin inhibitor spares. This experiment does that by sequencing the same patient's samples very deeply, before, during and after treatment.
The same thing, for a scientist
**The gap.** Duplex sequencing has shown that infant and childhood KMT2A-rearranged ALL carries drug-resistant clones - NRAS G13D/G12S, FLT3 D835H - at **0.06% to 0.34% of diagnostic cells** that expand at relapse after conventional chemotherapy, and that the relapse clone persists as a minor subclone throughout treatment before expanding in the last month of disease [E1]. Separately, menin inhibitors are now in use in KMT2A-rearranged leukaemia, and the two published investigations of resistance in KMT2A::AFF1 implicate **KMT2C/D** and unexplained upfront resistance, with emergent RAS and TP53 comutations explicitly **not sufficient** to cause resistance in vitro [E2]. **No study has measured what menin inhibition does to the sub-1% cooperating-lesion reservoir in a patient.** That is a strange gap, because the theoretical concern is specific and testable: menin inhibition acts on the fusion's transcriptional programme, and a RAS-mutant subclone has an alternative proliferative input. If any clone is positioned to survive menin inhibition, it is that one. **Design.** Serial sampling of patients receiving a menin inhibitor for KMT2A-rearranged acute leukaemia - within an existing trial or an expanded-access cohort - with targeted duplex sequencing of KRAS, NRAS, PTPN11, NF1, FLT3, PIK3CA, PIK3R1, TP53, IKZF1 and **MEN1** at baseline, at each response assessment, and at progression. Fusion-quantified MRD in parallel at every timepoint to give a leukaemic denominator. **Primary readout.** The trajectory of each variant's **cellular prevalence among leukaemic cells** across timepoints, not its raw allele fraction. A clone that holds constant as a fraction of a shrinking leukaemia is being spared; a clone that rises as a fraction is being selected; a clone that falls with the bulk is being killed. These are three different clinical situations that a single post-treatment genotype cannot distinguish, which is why the serial design is the whole experiment. **MEN1 is on the panel deliberately.** MEN1 point mutations are the canonical on-target menin-inhibitor resistance mechanism, and including them lets the study answer whether cooperating-lesion selection and on-target resistance are alternative routes or co-occur.
What each possible result would mean
| result | interpretation | consequence |
|---|---|---|
| RAS-mutant clones rise in cellular prevalence under menin inhibition | the drug selects for signalling-independent escape | the case for an upfront MEK or SHP2 combination strengthens sharply, and would be the first patient-level evidence for it |
| RAS clones fall with the bulk | the fusion programme is required even in RAS-mutant cells | supports menin monotherapy; weakens the combination rationale that the one published in-vivo test already failed to support on survival |
| RAS clones hold constant while the bulk falls | a spared but not expanding reservoir | argues for maintenance or consolidation strategy rather than combination |
| MEN1 mutations appear and RAS clones do not move | on-target resistance dominates | the cooperating-lesion question is a distraction and the graph should say so |
| nothing detectable moves and relapse occurs anyway | resistance is epigenomic | consistent with the KMT2C/D observation and with the finding that comutations were insufficient in vitro; genotype-guided combination is the wrong frame entirely |
Every row is informative, which is unusual and is the main argument for running it.
Why this is achievable without a new trial
Menin inhibitors are in clinical use in KMT2A-rearranged acute leukaemia and serial marrow sampling is already protocol-mandated for response assessment in every menin-inhibitor study. The additional requirement is DNA from timepoints that are already collected, plus a duplex panel that the Lund group has already built and run in this exact disease 1.
The binding constraints are not technical:
- Sample access. Serial marrow DNA from menin-inhibitor-treated patients sits with trial sponsors and consortia, and infant KMT2A-rearranged ALL patients on these agents are few.
- Input mass. Duplex sequencing consumes molecules to build consensus; a deep-remission marrow may not contain enough leukaemic genomes to estimate a subclone's prevalence. This is not a flaw - a timepoint with too few leukaemic genomes is genuinely uninformative about clone composition and should be reported as such rather than imputed.
- The denominator. Without fusion-quantified MRD at each timepoint, a falling variant allele fraction is uninterpretable, because it falls whenever the leukaemia falls.
| design | Serial-sample correlative study within existing menin-inhibitor trials or expanded-access cohorts in KMT2A-rearranged acute leukaemia. Targeted duplex (error-corrected) sequencing of KRAS, NRAS, PTPN11, NF1, FLT3, PIK3CA, PIK3R1, TP53, IKZF1 and MEN1 at baseline, at each protocol response assessment and at progression, with fusion-quantified MRD at every matched timepoint. |
|---|---|
| readout | Per-variant cellular prevalence AMONG LEUKAEMIC CELLS across timepoints, classified as selected (rising fraction), spared (constant fraction) or killed (falling with bulk). Secondary: whether MEN1 resistance mutations and cooperating-lesion selection co-occur or are alternatives. |
| controls | Fusion-quantified MRD as the leukaemic denominator at every timepoint; matched germline; pre-specified minimum leukaemic-genome input per timepoint with uninformative timepoints reported as such rather than imputed; where material allows, a transcriptional or chromatin readout on the same timepoints so that an epigenomic resistance mechanism is not misread as a null. |
| cost tier | med |
| timeline | 18-36 months, dependent entirely on sample access from menin-inhibitor-treated cohorts rather than on assay development |
| status | proposed |
| prerequisite | access to serial marrow DNA from menin-inhibitor-treated KMT2A-rearranged patients, which sits with trial sponsors and consortia |
| informative in every arm | true |
- A positive result - RAS clones selected by menin inhibition - would argue for adding a MEK inhibitor upfront to infants, in a disease where the one in-vivo test of that combination reduced burden without extending survival, and where the only open paediatric MEK protocol excludes children under one year.
- Finding a 0.1% resistant clone in a patient in remission creates an immediate clinical pressure to act on a measurement whose clinical significance is unestablished, in infants, where the available actions are toxic.
- A null result would be read as reassurance about menin monotherapy when it may only mean the resistance mechanism is not in the panel; the design must therefore carry a non-genomic readout or its null is uninterpretable.
- Deep sequencing of remission samples will find variants of uncertain significance in normal haematopoiesis, and in an infant those raise germline and predisposition questions that the consent for a correlative resistance study may not cover.
- Do any menin-inhibitor trials in KMT2A-rearranged leukaemia bank serial marrow DNA with consent permitting deep resequencing? If not, the experiment is blocked for reasons that have nothing to do with science.
- How many leukaemic genomes are recoverable from a deep-remission marrow in an infant, and is that enough to estimate a 0.1% subclone?
- Should the panel include KMT2C and KMT2D at the DNA level, given that the one mechanistic lead in pretreated KMT2A::AFF1 resistance is their downregulation rather than their mutation?
- If cooperating-lesion clones are neither selected nor killed by menin inhibition, what is the right frame for combination design in this disease at all?
- E1moderate confidence 3 of 5: One strong study, or several smaller consistent ones.paper
Targeted duplex sequencing of 20 infants and 7 children with KMT2A-rearranged ALL, including 5 longitudinal and 6 paired relapse sample sets, identified drug-resistant clones carrying NRAS G13D or G12S or FLT3 D835H at 0.06% to 0.34% of diagnostic cells that expanded at relapse, and showed the relapse clone persisting as a minor subclone from diagnosis through treatment before expanding during the last month of disease; 62% of the identified cancer-associated mutations were below the resolution of standard NGS.
Duplex Sequencing Uncovers Recurrent Low-frequency Cancer-associated Mutations in Infant and Childhood KMT2A-rearranged Acute Leukemia. · Pilheden M, Ahlgren L, Hyrenius-Wittsten A, et al. · HemaSphere 2022What this evidence cannot show
27 patients, targeted panel, and the treatment involved was conventional chemotherapy - no patient in that series received a menin inhibitor, which is precisely the gap this experiment addresses. A variant at 0.1% cellular prevalence cannot be assigned to a leukaemic rather than a normal haematopoietic cell by sequencing alone. Abstract read; full text not retrieved for this entry.
Evidence that argues against this (1)
- The same study found infants carrying fewer cancer-associated mutations than the older children, so the sub-1% reservoir may be thinner in the infant population this experiment most wants to study - which could make the primary readout unmeasurable in the patients who matter most. Pilheden M et al. 2022
- E2moderate confidence 3 of 5: One strong study, or several smaller consistent ones.paper
In patient-derived xenografts from two patients with multiply relapsed KMT2A::AFF1 ALL who had never received a menin inhibitor, later samples were markedly less menin-inhibitor sensitive than earlier samples from the same patients despite sustained on-target transcriptional effect; emergent RAS-pathway and TP53 comutations were individually not sufficient to induce menin-inhibitor resistance in vitro, with KMT2C/D downregulation implicated in one patient, and the authors conclude that menin inhibitors should be used upfront or in early lines before substantial genomic or epigenomic evolution.
Upfront menin-inhibitor resistance in multiply pretreated leukemias. · Mahdavi L, Alikarami F, Goodrow H, et al. · Experimental Hematology 2026What this evidence cannot show
Two patients for the serial arm, five PDX models for the sensitivity arm. The targeted genomic analysis is standard-depth, so a sub-1% clone would not have been visible in it - which is the direct motivation for proposing a duplex-depth version. 'Not sufficient in vitro' is a negative in an engineered system. Abstract read; full text not retrieved for this entry.
Evidence that argues against this (1)
- If resistance is epigenomic, as the KMT2C/D observation suggests, a duplex sequencing panel will detect nothing and the experiment will return an uninformative null - so the design should be paired with a transcriptional or chromatin readout on the same timepoints wherever material allows. Mahdavi L et al. 2026
- E3moderate confidence 3 of 5: One strong study, or several smaller consistent ones.paper
Combining the menin inhibitor revumenib with the MEK1/2 inhibitor selumetinib in KMT2A-rearranged RAS-pathway-mutant AML was synergistic in vitro and reduced leukaemia burden in PDX models in vivo, but did not improve overall survival compared with menin inhibition alone, and RAS-pathway mutations were present in 43% of AML cases in a 1,750-patient Children's Oncology Group cohort where they predicted lower complete remission, worse survival and early relapse in KMT2A-rearranged AML.
Combining menin and MEK inhibition to target poor prognosis KMT2A-rearranged RAS pathway-mutant acute myeloid leukemia. · Scheidegger N, Schneider C, Alexe G, et al. · Blood Advances 2026What this evidence cannot show
AML rather than infant ALL; PDX models rather than patients; clone-level allele frequencies under treatment are not reported, which is the measurement this experiment proposes. Abstract read; full text not retrieved for this entry.
Evidence that argues against this (1)
- That the combination reduced burden without extending survival is compatible with the RAS-mutant clone being neither selected nor the cause of failure - i.e. the null result of the experiment proposed here may already be implicit in that in-vivo data, and the experiment's value is that it would make it explicit and in patients. Scheidegger N et al. 2026
what menin inhibition selects for in the sub-1% cooperating-lesion reservoir - never measured
the only design that distinguishes a response from a selection event
MEN1 is on the panel deliberately, to separate on-target resistance from cooperating-lesion selection
the same assay should include the MEN1 pocket sites