Has anyone tested an XPO1 inhibitor in a KMT2A-rearranged system? Yes - four preclinical studies and zero clinical ones, counted with the queries and the controls
also called XPO1 inhibitor in KMT2A-r bounded search · selinexor KMT2A counted literature
We asked whether anyone had ever put an XPO1-blocking drug into KMT2A-rearranged leukaemia, and counted the answer instead of guessing. Four preclinical studies exist, all combinations, all in myeloid models; no clinical trial has ever selected patients by this genotype.
The same thing, for a scientist
**The dispatch brief for this beat predicted that an XPO1 inhibitor had probably never been tested in a KMT2A-rearranged system and that this would be 'a real and cheap gap'. That prediction is wrong, and the literature wins.** **Queries run 2026-09-20, PubMed E-utilities esearch, `querytranslation` read back on every one:** | query | records | |---|---:| | `(selinexor OR eltanexor OR "KPT-330" OR "KPT-8602" OR "XPO1 inhibitor" OR "XPO1 inhibitors" OR "XPO1 inhibition" OR "exportin-1 inhibitor" OR "nuclear export inhibitor" OR leptomycin) AND (KMT2A OR "MLL-rearranged" OR "MLL rearranged" OR "MLL-r" OR "mixed lineage leukemia")` | **2** | | `selinexor[tiab] AND (KMT2A[tiab] OR MLL[tiab])` | **4** | | `XPO1[tiab] AND (KMT2A[tiab] OR MLL[tiab])` | **10** | | `(selinexor[tiab] OR "XPO1 inhibitor"[tiab] OR eltanexor[tiab]) AND (menin[tiab] OR revumenib[tiab] OR ziftomenib[tiab])` | **3** | **Operand controls (rule 67), because a small intersection needs its operands sized:** `selinexor[tiab]` = **684**; `KMT2A[tiab]` = **1,652**; `XPO1[tiab] AND leukemia[tiab]` = **145**. Large operands, small but non-zero intersection - the instrument returns ones and the answer is a real small number, not a query failure. **Name list (rules 71-72):** selinexor, eltanexor, KPT-330, KPT-8602, XPO1 inhibitor / inhibitors / inhibition, exportin-1 inhibitor, nuclear export inhibitor, leptomycin. Singular, plural and nominalisation all run. Hyphenated compound codes used as written. **What the hits actually are, after title-checking every one:** 1. **PMID 41959134** (bioRxiv 2026, preprint) - **ziftomenib + selinexor** in KMT2A-r and NPM1-m AML: cell lines, primary CD34+ KMT2A-r progenitors, MV4;11 and primary PDX in NSG mice [E1]. 2. **PMID 40373471** (Transl Oncol 2025) - **triptolide + selinexor** in MLL-R AML, in vitro, ex vivo and in vivo, via Rap1/Raf/MEK-mediated MYC down-regulation [E2]. 3. **PMID 40340946** (J Transl Med 2025) - **selinexor + JQ1** in AML including an **MLL-AF9 xenograft**, CDX and PDX, via dual C-MYC inhibition [E3]. 4. **PMID 35668193** (Nat Cancer 2022) - **selinexor + ipatasertib** outperformed chemotherapy and chemotherapy-plus-selinexor in a **syngeneic MLL-AF9-driven mouse AML model**; the mechanism is drug-induced P2RY2-AKT signalling as a resistance consequence of XPO1 inhibition [E4]. **Four studies, four different partners, and every one is a combination in a MYELOID model.** Zero monotherapy studies in a KMT2A-rearranged system. Zero lymphoid models. Zero infant models. **Zero clinical trials selecting on KMT2A rearrangement** - the paediatric selinexor phase 1 (NCT02212561) was genotype-agnostic. **So the gap is not 'nobody has tried'. The gap is narrower and more actionable:** nobody has tested XPO1 inhibition in **KMT2A-rearranged B-lineage** disease, nobody has tested it as **monotherapy** in this genotype, and nobody has measured **XPO1 chromatin occupancy** in a KMT2A-rearranged cell.
Rule 80 applied: what this count cannot bound
These are title-and-abstract counts. A selinexor experiment that appears only in a supporting panel of a paper about something else is invisible to every query above, and mechanism measurements in this field usually are supporting panels. The specific measurement most likely to be hiding that way is an XPO1 or CRM1 ChIP in a KMT2A-rearranged line - it would sit in a supplementary figure of a paper titled after something else entirely.
So the bound is: zero in titles and abstracts; supporting figures not searched. Two papers were opened beyond their abstracts in this beat (the two whose full text this project already holds, PMID 31755865 via PMC6874418 and PMID 41959134's abstract record); the remainder were read at abstract level, and that is stated in every evidence caveat.
Rule 77 applied: the query was also run without the disease term
XPO1[tiab] AND leukemia[tiab] returns 145 records, and selinexor[tiab] alone returns 684. Scoping to KMT2A removes 99% of that literature, and the removed part is mostly multiple myeloma, DLBCL, NPM1-mutant AML and CLL. Two of the most mechanistically relevant papers for this beat - the NPM1c schedule result (PMID 36037515) and the NPM1-fusion XPO1 dependence (PMID 39443736) - are invisible to any KMT2A-scoped query and are the reason the scoping check matters here. They are cited in mechanism-npm1c-crm1-hoxa-recruitment and mechanism-nes-crm1-chromatin-addressing.
Searched, read and rejected
Recorded per the brief's bounded-rejection rule, so a future cycle does not repeat the search without knowing the decision:
- PMID 40647396 (Cancers 2025, NPM1-mutated AML review) - returned by
XPO1 AND (KMT2A OR MLL) AND mutation; a clinical review, no primary measurement relevant here. Not cited. - PMID 38306602 (Blood Adv 2024, MLLT10 structural variants in paediatric AML) - matched on XPO1 only incidentally; it is an outcome study of MLLT10 fusions, already covered by this graph's PICALM::MLLT10 layer. Not cited.
- PMID 34743404 (FEBS J 2023, molecular regulators of HOXA9) - a genuinely on-topic review that names 'NPM1c-XPO1' in its list of HOXA9 regulators. Used as a map only; every claim it would support is taken from a primary source instead.
- PMID 32609823 / 37317978 (Falini, NPM1-mutated AML reviews) - authoritative but reviews; superseded here by the primary Brunetti, Oka and Pianigiani papers.
- PMID 40806458, 40712110, 36001451, 33007990 and the rest of the E571K retrieval - PMBL, Hodgkin, ATLL and cfDNA detection papers. Real and on-topic for the hot spot, out of scope for KMT2A. One (PMID 33451349) was kept because it carries the structural and functional characterisation.
What would change the conclusion: any paper reporting XPO1 inhibition in a KMT2A-rearranged LYMPHOID model, or any trial registering KMT2A rearrangement as an eligibility criterion for a SINE. Neither was found.
| script | manual; PubMed E-utilities esearch via scratchpad helper, 2026-09-20 |
|---|---|
| status | run |
| date | 2026-09-20 |
| tool | NCBI E-utilities esearch (db=pubmed) |
| result summary | Four preclinical studies test an XPO1 inhibitor in a KMT2A-rearranged system: ziftomenib+selinexor (preprint), triptolide+selinexor, selinexor+JQ1, selinexor+ipatasertib. All are combinations, all myeloid, none lymphoid, none infant, none monotherapy, and no clinical trial selects on KMT2A rearrangement. |
| bound | titles and abstracts only; supporting figures not searched (rule 80) |
| brief premise corrected | The dispatch brief predicted the combination was untested. It is tested, four times, and the correct gap is narrower. |
inputs
[ "PubMed E-utilities esearch", "querytranslation read back on every query" ]
name list searched
[ "selinexor", "eltanexor", "KPT-330", "KPT-8602", "XPO1 inhibitor", "XPO1 inhibitors", "XPO1 inhibition", "exportin-1 inhibitor", "nuclear export inhibitor", "leptomycin" ]
disease terms searched
[ "KMT2A", "MLL-rearranged", "MLL rearranged", "MLL-r", "mixed lineage leukemia", "MLL[tiab]" ]
counts — the numbers behind the chart
{
"drug_list AND KMT2A_list": 2,
"selinexor[tiab] AND (KMT2A[tiab] OR MLL[tiab])": 4,
"XPO1[tiab] AND (KMT2A[tiab] OR MLL[tiab])": 10,
"selinexor[tiab] (operand control)": 684,
"KMT2A[tiab] (operand control)": 1652,
"XPO1[tiab] AND leukemia[tiab] (operand control)": 145,
"SINE AND menin-inhibitor names": 3
}remaining gaps
[ "XPO1 inhibition in a KMT2A-rearranged LYMPHOID model - zero studies", "XPO1 inhibitor monotherapy in a KMT2A-rearranged system - zero studies", "XPO1 or CRM1 ChIP in any KMT2A-rearranged cell - zero reports found", "any trial with KMT2A rearrangement as a SINE eligibility criterion - zero" ]
- The one study that looked for adaptive resistance found it: XPO1 inhibition upregulates P2RY2 and activates AKT in AML, which is a built-in escape route present before any treatment begins.
- A counted literature of four combination studies with four different partners is as consistent with 'selinexor needs a partner to do anything here' as with 'selinexor is a good combination agent'.
- Counting titles and abstracts cannot see a negative result that was never written up, and combination-only publication is exactly the pattern a suppressed monotherapy failure would produce.
- What is the KMT2A-rearranged subgroup outcome in the 18 SELHEM patients? The data exist and the analysis is free.
- Why are all four preclinical studies combinations? Either monotherapy was tried and failed and went unpublished, or nobody thought it worth a paper - and those two possibilities have very different implications.
- Three of the four hits converge on MYC. Is MYC the real mechanism of XPO1 inhibition in this genotype, or is it the readout everyone reaches for first?
- Would a full-text search of the four hits, plus the two chromatin-bound CRM1 papers, surface a KMT2A-rearranged XPO1 ChIP in a supplementary figure? Rule 80 says the abstract-level zero cannot answer that.
- E1weak confidence 2 of 5: Preliminary, small, indirect, or not yet peer reviewed.preprint
A 2026 preprint reports ziftomenib plus selinexor as synergistic in multiple KMT2A-rearranged and NPM1-mutant AML cell lines, in primary CD34+ KMT2A-rearranged progenitors, and in MV4;11, OCI-AML3 and primary patient-derived xenografts in NSG mice.
Combined Menin and XPO1 inhibition drive synergistic antileukemic activity in KMT2Ar and NPM1-m AML. · Uddin MH, Dhiman S, Han Y, et al. · bioRxiv 2026What this evidence cannot show
PREPRINT, not peer reviewed, abstract-level. Myeloid models only. Full entry at `mechanism-xpo1-stabilises-menin-chromatin-binding`.
Evidence that argues against this (1)
- Counting titles and abstracts cannot see a negative result that was never written up. All four hits are combinations with four different partners, which is equally consistent with 'selinexor needs a partner here' as with 'selinexor is a good combination agent' - and the only randomised test of selinexor in acute leukaemia was negative and numerically unfavourable, median OS 3.2 versus 5.6 months, HR 1.18. Sweet K et al. 2021
- E2weak confidence 2 of 5: Preliminary, small, indirect, or not yet peer reviewed.paper
Low-dose triptolide combined with selinexor was synergistic against MLL-rearranged AML in vitro, ex vivo and in vivo via Rap1/Raf/MEK/ERK-mediated MYC down-regulation.
The synergy of TPL and selinexor in MLL-R acute myeloid leukemia via Rap1/Raf/MEK pathway-mediated MYC downregulation. · Deng M, Huang P, Wang L, et al. · Translational Oncology 2025What this evidence cannot show
Preclinical, single group, AML not infant B-ALL, and the partner agent triptolide has no clinical development path in leukaemia. Already cited in `drug-selinexor`; recorded here for completeness of the count, at abstract level.
Evidence that argues against this (1)
- NAMED METHODOLOGICAL DEFEATER for the partner agent. Triptolide is an XPB/TFIIH inhibitor that blocks transcriptional INITIATION globally: Chen, Gao and Shilatifard used it precisely because one hour of treatment clears polymerase from most paused genes genome-wide. MYC is among the shortest-half-life, most initiation-dependent transcripts in a leukaemia cell, so 'MYC down-regulation' under low-dose triptolide is the expected generic consequence of shutting off initiation and does not identify a Rap1/Raf/MEK/ERK route or any interaction with XPO1. A synergy between a global initiation inhibitor and a common-essential export receptor is close to the null expectation for a viability assay. Chen F et al. 2015
- E3weak confidence 2 of 5: Preliminary, small, indirect, or not yet peer reviewed.paper
Selinexor combined with the BET inhibitor JQ1 showed strong synergy across AML models including MLL-AF9, CDX and PDX xenografts, with greater than 80% inhibition in cell lines and primary samples, C-MYC suppression at protein and mRNA level, apoptosis, cell-cycle arrest, reduced leukaemic burden in bone marrow, spleen and liver, and extended survival.
Innovative evaluation of selinexor and JQ1 synergy in leukemia therapy via C-MYC inhibition. · Wang PH, Hu CH, Fan JQ, et al. · Journal of Translational Medicine 2025What this evidence cannot show
Abstract-level. The abstract lists 'MLL-AF9' among the xenograft models without stating how many lines or animals carried it, so the KMT2A-rearranged content of this study is not quantifiable from the abstract. JQ1 is a tool BET inhibitor, not a clinical agent. Single group. Both this and E2 reach the same mechanism (MYC) by different partners, which is either convergence or a common bias toward the easiest available readout.
Evidence that argues against this (1)
- PUBLISHED CONFLICTING RESULT about what a MYC readout can attribute. MYC is a direct target of AFF4/SEC, and Luo et al. showed that SEC recruitment to the MYC gene regulates its expression in acute myeloid AND lymphoid leukaemia cells, proposing AFF4/SEC as a therapeutic target on that basis. MYC is therefore the convergent downstream node for elongation-complex, BET and export perturbations alike, and two agents that each independently lower MYC will read as synergistic on a MYC endpoint whatever their mechanisms. The entry's own observation that E2 and E3 reach MYC by different partners is better explained as a common bias toward the easiest readout than as convergence. Luo Z et al. 2012
- E4moderate confidence 3 of 5: One strong study, or several smaller consistent ones.paper
XPO1 inhibition by selinexor activates PI3K-gamma-dependent AKT signalling in AML by upregulating the purinergic receptor P2RY2; inhibiting that axis potentiates selinexor in AML cell lines, patient-derived primary cultures and mouse models, and in a syngeneic MLL-AF9-driven mouse AML model selinexor plus ipatasertib outperformed both standard-of-care chemotherapy and chemotherapy plus selinexor.
P2RY2-AKT activation is a therapeutically actionable consequence of XPO1 inhibition in acute myeloid leukemia. · Lin KH, Rutter JC, Xie A, et al. · Nature Cancer 2022What this evidence cannot show
Abstract-level. Murine syngeneic MLL-AF9 myeloid leukaemia - the exact system this project's rule 73 warns against generalising from. Its importance here is inverted: it is the only one of the four that reports a RESISTANCE mechanism induced by XPO1 inhibition, which is a bite-back finding rather than a supporting one.
Evidence that argues against this (1)
- PUBLISHED CONFLICTING RESULT on whether the AKT escape is rate-limiting in patients. In a phase 1 of selinexor with daunorubicin and cytarabine in 21 previously untreated poor-risk AML patients, 8 of 19 evaluable patients (53%) achieved CR/CRi with no AKT-pathway agent of any kind and the MTD was not reached. If P2RY2-AKT activation were an obligate escape from XPO1 inhibition, that response rate is difficult to obtain. The finding stands as a mechanism; its framing as the reason XPO1 inhibition underperforms clinically is not supported by the one setting in which selinexor has been combined with intensive chemotherapy. Sweet K et al. 2020
- E5strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.paper
The only paediatric selinexor leukaemia trial, SELHEM (NCT02212561), enrolled 18 children with relapsed or refractory acute leukaemia without selecting on genotype, and reports no KMT2A-rearranged subgroup analysis.
Phase I Study of Selinexor, a Selective Inhibitor of Nuclear Export, in Combination With Fludarabine and Cytarabine, in Pediatric Relapsed or Refractory Acute Leukemia. · Alexander TB, Lacayo NJ, Choi JK, et al. · Journal of Clinical Oncology 2016What this evidence cannot show
Recorded in full in `drug-selinexor`. Cited here for the negative half only - that the one paediatric trial was genotype-agnostic - which is an absence-of-analysis claim and not an absence-of-effect claim. The genotypes of those 18 children exist and a retrospective look has never been published.
Evidence that argues against this (1)
- AUTHORS' OWN STATED LIMITATION, and it is a stronger objection than the absence-of-analysis point this item makes. SELHEM's stated primary endpoints are the maximum tolerated dose and recommended phase 2 dose of selinexor GIVEN WITH fludarabine and cytarabine. Any retrospective genotype-response analysis of those 18 children would therefore be a comparison of genotypes under a fludarabine/cytarabine backbone, in which a KMT2A-rearranged subgroup difference could not be attributed to selinexor. The retrospective look this entry says has never been published would, if published, be close to uninterpretable - which is a better reason for its absence than neglect. Alexander TB et al. 2016
four preclinical KMT2A-rearranged studies, all combinations, all myeloid
the genotype the counted literature is about
bounds what has and has not been tried in this genotype
both preclinical synergy reports in the KMT2A-rearranged arm terminate at MYC, which is the convergent node for elongation, BET and export perturbations alike (PMID 22547686) and, for triptolide, the expected consequence of a global TFIIH/initiation block (PMID 25561494) - convergence or a common bias toward the easiest readout