Systematic enumeration of menin-inhibitor resistance mechanisms: seven entries written, thirteen candidates rejected with reasons

[
  "NCBI eutils esearch with [Title/Abstract] field tags across 14 query families (menin+resistance; each named compound; lineage switch; efflux/ABCB1; pharmacokinetics/CYP3A4; MEN1 copy number; fusion partner; microenvironment; stemness/differentiation; ENL/YEATS; clonal evolution/cfDNA; CRISPR/resistance screens; KMT2C/KMT2D/UTX; venetoclax combinations)",
  "NCBI efetch for every PMID used, with title, authors, journal, year and DOI read back",
  "raw ClinicalTrials.gov v2 API: NCT04065399 plus a 66-record sweep of all menin-inhibitor studies with whyStopped read on every non-active record",
  "data/graph/nodes/*/*.json for the existing-id diff and for the four pre-existing resistance entries",
  "docs/absence-vs-library-report.json from scripts/check-absence-vs-library.py --min-overlap 2"
]

[
  "resistance-jund-liberation-menin",
  "resistance-tp53-apoptotic-priming-menin",
  "resistance-mll34-utx-tumour-suppressor-failure",
  "resistance-lineage-switch-under-menin-inhibition",
  "resistance-enl-marked-epigenetic-memory",
  "resistance-swi-snf-bet-dependent-state",
  "resistance-cross-resistance-between-menin-inhibitors",
  "hyp-gmp-destination-paradox-menin"
]

[
  {
    "candidate": "Drug efflux (ABCB1 / P-glycoprotein / ABCG2) as menin-inhibitor resistance",
    "reason": "No literature exists. `(menin OR revumenib OR ziftomenib)[tiab] AND (efflux OR ABCB1 OR P-glycoprotein OR ABCG2)[tiab]` returns exactly 2 records, both unrelated: menin and cholesterol in colorectal cancer (PMID 37627154) and epigenetic targets in taxane-resistant prostate cancer (PMID 38346967). The named BET-inhibitor resistance precedent in MLL-AF9 that this graph already cites was explicitly shown NOT to be efflux-mediated, which is the closest the field comes to addressing the question and points away from it."
  },
  {
    "candidate": "Microenvironment- or niche-mediated protection from menin inhibition",
    "reason": "Named as a resistance category in a 2026 review (PMID 42733002) but no primary study supports it in this disease. `(menin inhibitor OR menin inhibition)[tiab] AND (bone marrow stroma OR mesenchymal OR cytokine OR co-culture)[tiab]` returns 3 records, none of which tests stromal protection against a menin inhibitor. Writing an entry from a review's category heading would be exactly the error this project has been burned by."
  },
  {
    "candidate": "Hypoxic niche protecting KMT2A-rearranged blasts from menin inhibition",
    "reason": "The evidence points the OPPOSITE way. The only study testing hypoxia in KMT2A-rearranged AML under a menin inhibitor reports that hypoxia induction SENSITISED cells to venetoclax plus MI-503, and that the combination works partly by repressing HIF-1A and its target HDAC9 (PMID 38053201, [E4]). A hypoxia-protection entry would have been written against the data."
  },
  {
    "candidate": "MEN1 copy-number change, MEN1 amplification or target re-expression as resistance",
    "reason": "No literature. `MEN1[tiab] AND (amplification OR copy number)[tiab] AND leukemia[tiab]` returns 2 records: a 2004 review of MLL transformation (PMID 15663355) and a 2002 FISH study of MEN1 allelic deletion in AML (PMID 12210500), neither involving a drug. There is also a mechanistic argument against it: menin is REQUIRED for the leukaemogenic activity of KMT2A fusions, and the clinical resistance mutations are missense changes that specifically preserve the menin-MLL1 interaction (PMID 36922589) — the tumour cannot afford to lose menin, which is why every observed escape keeps it."
  },
  {
    "candidate": "Rebound or re-expression of HOXA9/MEIS1 on continued therapy",
    "reason": "No literature at all: `(menin inhibitor OR menin inhibition)[tiab] AND (HOXA9 OR MEIS1)[tiab] AND (rebound OR re-expression OR reexpression OR restoration)[tiab]` returns zero, and the field's actual finding is the reverse — resistant cells show SUSTAINED on-target suppression of the canonical targets (PMID 41027490). That sustained-suppression observation is recorded inside resistance-mll34-utx-tumour-suppressor-failure, where it belongs."
  },
  {
    "candidate": "KMT2A partner identity determining menin-inhibitor response",
    "reason": "Tested once and came out negative: three AMLs with the high-risk partners MLLT10 and AFDN, plus two infant ALL samples, were all sensitive to menin inhibition in patient-derived xenografts ([E5], PMID 41027490). No trial has reported response stratified by partner or by breakpoint, and this graph already records that absence in mechanism-different-breaks-different-biology. Partner-specific chromatin binding patterns do exist (PMID 39472576) and correlate with ALL-versus-AML prevalence, but nobody has connected them to drug response."
  },
  {
    "candidate": "Differentiation escape / re-acquisition of a stem programme as menin-inhibitor resistance",
    "reason": "Not documented, and one direct result points the other way: GMP-like primary AML shows the GREATEST ex vivo sensitivity to a menin-KMT2A inhibitor ([E6], PMID 39704147), so a more differentiated myeloid state is associated with more drug effect, not less. The stem-compartment literature that does exist is about SENSITISATION — leukaemia stem cells of MLL-rearranged AML have enhanced guanine nucleotide biosynthesis whose blockade drives differentiation and sensitises to menin inhibition (PMID 40102405) — which is the inverse framing. The tension between the two directions is real enough to be worth a hypothesis, and is written as hyp-gmp-destination-paradox-menin rather than as a speculative resistance entry."
  },
  {
    "candidate": "SETD2 / H3K36me3-supported transcription elongation as a menin-independent buffer",
    "reason": "The single relevant study (PMID 42170936) shows that a SETD2 inhibitor SYNERGISES with revumenib and that SETD2 inhibition sensitises leukaemia cells by decreasing H3K36me3. That is a combination-rationale paper, not a demonstration that H3K36me3 causes resistance. Recorded here because the mechanistic link to LEDGF/PSIP1 PWWP reading of H3K36me is attractive and someone will re-derive it; the honest status is untested."
  },
  {
    "candidate": "Guanine nucleotide biosynthesis / IMPDH2 as a metabolic support of resistance",
    "reason": "Same category as SETD2. PMID 40102405 shows IMPDH2 inhibition reduces LEDGF and menin protein, impairs MLL-fusion complex formation and chromatin binding, and sensitises to menin inhibitors. Sensitisation by blockade is not evidence that the intact pathway causes resistance; no study reports high guanine nucleotide flux in resistant cells."
  },
  {
    "candidate": "Immunoproteasome function as a resistance mechanism",
    "reason": "PMID 38049829 shows immunoproteasome function maintains oncogenic gene expression in KMT2A-complex-driven leukaemia. That is a vulnerability, not an escape route; nothing links it to menin-inhibitor failure."
  },
  {
    "candidate": "Loss of the fusion / loss of the drug target at relapse",
    "reason": "Contradicted by this graph's own sourced entry. The KMT2A fusion is truncal and stable, retained through relapse, through lineage switch and through immunophenotypic escape from CD19-directed therapy (mechanism-fusion-clonal-stability-diagnosis-to-relapse). Target loss therefore is not available as a route in this disease, which is one of the few structural advantages the menin strategy has."
  },
  {
    "candidate": "Residual fusion occupancy at saturating drug",
    "reason": "Already covered in this graph as hyp-apparent-authority-residual-fusion-occupancy plus exp-residual-occupancy-cutrun-saturating-menin. A duplicate resistance entry was not written. However this beat found a human observation that bears directly on it and appears in neither node — see cross_beat_findings below."
  },
  {
    "candidate": "Pharmacokinetic under-exposure / CYP3A4 interaction as a resistance mechanism",
    "reason": "The pharmacology is real, fully documented in this graph's drug-revumenib entry (CYP3A4 substrate; label halves the dose with a strong CYP3A4 inhibitor; 270 mg BID versus 160 mg BID at 40 kg or more), and the pivotal AUGMENT-101 phase 2 dosed 163 mg q12h WITH a strong CYP3A4 inhibitor ([E7], PMID 39121437) — that is, the registration efficacy data were generated under mandatory enzyme inhibition. Ziftomenib carries its own labelled CYP3A4 fraction-metabolised question (PMID 42682186, PBPK modelling to inform labelling). Despite all that, NO case of menin-inhibitor failure attributed to under-exposure has been reported, and `(menin inhibitor OR revumenib)[tiab] AND (pharmacokinetic OR exposure OR CYP3A4)[tiab]` returns 10 records none of which describes one. Rejected as an entry and recorded as an open question: a child whose azole is stopped, or who is started on a CYP3A4 inducer, moves to a different exposure with no published guidance on what that does to response."
  },
  {
    "candidate": "UBTF tandem duplication AML response-and-resistance",
    "reason": "Real and interesting — UBTF-TD AML is menin-inhibitor-sensitive (PMID 37890156) and an NEJM correspondence reports response AND resistance in that genotype (PMID 38924737). Rejected for this beat because UBTF-TD is a distinct genotype outside the KMT2A-rearranged scope of this graph, and because the NEJM item is correspondence with no abstract in PubMed, so its mechanism could not be read without the full text. Recorded so the next agent knows it was seen and deliberately left out rather than missed."
  }
]

[
  {
    "entries": [
      "hyp-apparent-authority-residual-fusion-occupancy",
      "exp-residual-occupancy-cutrun-saturating-menin"
    ],
    "issue": "The hypothesis predicts that a reproducible subset of sites retains FUSION occupancy at saturating menin inhibitor, and lists complete loss of fusion occupancy as its first falsifier. The only human sample profiled under revumenib (PMID 39472576) shows the oncoprotein and menin reduced to UNDETECTABLE levels, with ENL persisting on oncoprotein-target loci instead. The residual-occupied-set idea survives; the identity of the occupant does not.",
    "confound": "that sample is lineage-switched, and lineage switching is independently associated with reduced oncoprotein levels in the same study, so drug saturation and switched state cannot be separated at n=1",
    "requested_action": "The owning Queen should add ENL to the predicted residual occupant set and add an ENL arm to the experiment's readout, or state why a lineage-switched sample does not bear on the prediction.",
    "read": "abstract via NCBI efetch; full text not obtained"
  },
  {
    "entries": [
      "exp-jund-displacement-coip-under-clinical-menin-inhibitor",
      "interface-menin-jund-pocket",
      "mechanism-jund-release-inverts-to-growth-promotion"
    ],
    "issue": "PMID 42371796 (Blood 2026) is filed in this graph only as an endocrine toxicity concern and as an illustration of chromatin redistribution. Its own headline finding is that menin-MLL inhibitors disrupt menin-JUND, JUND chromatin occupancy rises, JUND targets are upregulated, and this CONTRIBUTES TO RESISTANCE in MLL-rearranged leukaemia, with JUND knockout or JNK inhibition synergising and a xenograft arm. That is a leukaemia resistance result.",
    "requested_action": "No edit made. resistance-jund-liberation-menin now records the resistance reading and is edged to all three endocrine nodes. The endocrine Queen may wish to cross-reference rather than duplicate.",
    "read": "abstract via NCBI efetch; full text not obtained"
  },
  {
    "entries": [
      "resistance-adaptive-nongenetic-menin"
    ],
    "issue": "That entry's data.named_routes lists 'KMT2C/KMT2D inactivation or downregulation' among routes described as 'invisible to DNA mutation panels'. KMT2C and KMT2D loss-of-function is genetic and appears on comprehensive sequencing panels; filing it under an invisible-to-panels heading could cause a clinician to stop looking for something that is in fact testable. Separately, PMID 41734382 (Blood 2026, MITR cells with SMARCA4/BRD4/CREBBP co-dependencies and no menin mutations) is strong additional evidence for that entry and is not cited there.",
    "requested_action": "Owning Queen to qualify the detectability statement and consider adding PMID 41734382.",
    "read": "the entry in full; the paper at abstract level via NCBI efetch"
  }
]

[
  {
    "pmid": "41351513",
    "error": "DOI and author list filled in from memory in a draft; real values are 10.1182/bloodadvances.2025018178 and Chin KK, Ball BJ, Abaza Y, et al.",
    "status": "corrected before finalisation",
    "title": "Outcomes of relapsed or refractory acute myeloid leukemia after menin inhibition failure.",
    "authors": "Chin KK, Ball BJ, Abaza Y, et al.",
    "venue": "Blood Adv",
    "year": 2026,
    "title_provenance": "Title/authors/venue backfilled from PubMed esummary on 2026-09-20 because this citation stored a bare PMID with no title, which made it unscoreable by verify-citations.py. BACKFILLING IS NOT VERIFICATION: it confirms the PMID resolves to a real paper, not that this is the right paper for the claim. That still requires reading."
  },
  {
    "pmid": "39158067",
    "error": "author list and DOI filled in from memory in a draft; real values are Zhou X, Zhang L, Aryal S, et al. and 10.1182/blood.2023023644",
    "status": "corrected before finalisation",
    "title": "Epigenetic regulation of noncanonical menin targets modulates menin inhibitor response in acute myeloid leukemia.",
    "authors": "Zhou X, Zhang L, Aryal S, et al.",
    "venue": "Blood",
    "year": 2024,
    "title_provenance": "Title/authors/venue backfilled from PubMed esummary on 2026-09-20 because this citation stored a bare PMID with no title, which made it unscoreable by verify-citations.py. BACKFILLING IS NOT VERIFICATION: it confirms the PMID resolves to a real paper, not that this is the right paper for the claim. That still requires reading."
  },
  {
    "pmid": "40536420",
    "error": "DOI guessed in a draft; real value is 10.1158/2159-8290.CD-24-1517",
    "status": "corrected before finalisation",
    "title": "Catalytic Inhibition of KAT6/KAT7 Enhances the Efficacy and Overcomes Primary and Acquired Resistance to Menin Inhibitors in MLL Leukemia.",
    "authors": "Gordon SJV, Perner F, MacPherson L, et al.",
    "venue": "Cancer Discov",
    "year": 2025,
    "title_provenance": "Title/authors/venue backfilled from PubMed esummary on 2026-09-20 because this citation stored a bare PMID with no title, which made it unscoreable by verify-citations.py. BACKFILLING IS NOT VERIFICATION: it confirms the PMID resolves to a real paper, not that this is the right paper for the claim. That still requires reading."
  },
  {
    "pmid": "38437498",
    "error": "DOI guessed in a draft; real value is 10.1182/blood.2023022832",
    "status": "corrected before finalisation",
    "title": "BRG1/BRM inhibitor targets AML stem cells and exerts superior preclinical efficacy combined with BET or menin inhibitor.",
    "authors": "Fiskus W, Piel J, Collins M, et al.",
    "venue": "Blood",
    "year": 2024,
    "title_provenance": "Title/authors/venue backfilled from PubMed esummary on 2026-09-20 because this citation stored a bare PMID with no title, which made it unscoreable by verify-citations.py. BACKFILLING IS NOT VERIFICATION: it confirms the PMID resolves to a real paper, not that this is the right paper for the claim. That still requires reading."
  },
  {
    "pmid": "29898879",
    "error": "DOI guessed in a draft; real value is 10.1182/bloodadvances.2018018093",
    "status": "corrected before finalisation",
    "title": "Spontaneous reversion of a lineage switch following an initial blinatumomab-induced ALL-to-AML switch in MLL-rearranged infant ALL.",
    "authors": "Wölfl M, Rasche M, Eyrich M, et al.",
    "venue": "Blood Adv",
    "year": 2018,
    "title_provenance": "Title/authors/venue backfilled from PubMed esummary on 2026-09-20 because this citation stored a bare PMID with no title, which made it unscoreable by verify-citations.py. BACKFILLING IS NOT VERIFICATION: it confirms the PMID resolves to a real paper, not that this is the right paper for the claim. That still requires reading."
  }
]