The menin-inhibitor plus PARP-inhibitor experiment is genuinely unrun, but its rationale should be DNA repair, not HOXA9 - and the evidence comes from a disease where menin is a tumour suppressor
also called experiment H · menin PARP · MEN1 homologous recombination · talazoparib menin
This project recommended an untried drug combination on the reasoning that menin inhibitors should mimic switching off HOXA9, and a properly controlled search confirms nobody has run it. The check also turned up a better reason to try it from a different cancer, where losing menin directly cripples DNA repair - but there menin is protective, so the logic runs the opposite way round.
The same thing, for a scientist
Ledger experiment **H** - *a menin inhibitor should phenocopy HOXA9 inhibition and sensitise to PARP inhibitors, since HR proficiency in this disease is HOXA9-driven* - was checked with a bounded query. **The absence survives**: 5 hits, **0 on topic**, against controls of 1,275 (PARP inhibitor AND leukemia) and 5 (HOXA9 AND PARP). **But one of the five reframes the rationale.** PMID 35171113 shows that in pancreatic neuroendocrine tumours, **MEN1 loss of function itself impairs double-strand break repair** via a homologous-recombination defect with reduced BRCA2 expression, and MEN1-mutant cells are more sensitive to the strongly trapping PARP1 inhibitor talazoparib. **That is a direct menin-to-HR link that does not go through HOXA9 at all**, and it is a cleaner rationale for the combination than the one this graph has been using. **The direction is the problem**: in PanNET menin is a tumour suppressor being LOST, whereas in KMT2A-rearranged leukaemia menin is an oncogenic cofactor being pharmacologically displaced. Whether displacing menin from chromatin reproduces the repair defect caused by deleting it is exactly the untested step.
The bounded query, and the two false positives worth knowing about
| query | hits |
|---|---|
| TARGET menin inhibitor AND (PARP OR olaparib OR talazoparib OR niraparib OR veliparib) | 5 |
| control: PARP inhibitor AND leukemia | 1,275 |
| control: HOXA9 AND PARP | 5 |
All five target hits fail on reading, and two fail in instructive ways:
- PMID 40016386 (MI-503 in osteosarcoma) matches because the paper reports increased cleaved PARP.
cl-PARPis an apoptosis readout, not a PARP inhibitor. Anymenin AND PARPquery will collect apoptosis westerns. - PMID 23343084 matches because 'Menin' is the surname of an author - Menin E, on a grape-seed-extract paper about Her-2 in liver and breast lines. A person's name collided with the protein's name.
The remaining three are a breast-cancer review, a menin-ASO TNBC paper that proposes combining with PARP inhibitors without testing it, and the PanNET paper below.
Why the PanNET result is the better rationale, and why its direction is a problem
The graph's current rationale is indirect: menin inhibition lowers HOXA9, HOXA9 drives HR proficiency, therefore HR falls, therefore PARP inhibitors work. Three steps, each measured in different systems.
The PanNET result is one step: MEN1 loss reduces BRCA2 and impairs double-strand break repair directly, and talazoparib sensitivity follows. If that mechanism is a property of menin rather than of the tissue, it predicts the same combination in leukaemia without needing HOXA9 at all - and it predicts it more strongly, because it removes two inferential steps.
The direction is the live objection and it is not a small one. In PanNET, MEN1 is a bona fide tumour suppressor and the cells have lost the protein entirely. In KMT2A-rearranged leukaemia, menin is an oncogenic cofactor and a menin inhibitor displaces it from the KMT2A interface while leaving the protein present and free to do everything else it does. A pocket blocker is not a knockout, and this graph has made that exact distinction before in the H3K27me3 argument. So the PanNET finding raises the prior on the combination and does not establish it.
What it changes practically: the experiment should measure an HR endpoint - RAD51 foci, BRCA2 level, a DSB repair assay - and not only viability, because if the mechanism is repair the readout should be repair. The current framing would have run a synergy screen and learned nothing about why.
bounded query
{
"target_hits": 5,
"on_topic": 0,
"controls": {
"PARP inhibitor AND leukemia": 1275,
"HOXA9 AND PARP": 5
},
"run": "PubMed esearch 2026-09-20"
}false positive classes
[ "cleaved PARP as an apoptosis readout", "'Menin' as an author surname (PMID 23343084)" ]
- This entry raises the prior on a drug combination in children and does not establish it. The supporting result is a GENETIC KNOCKOUT of MEN1 in a tumour where menin is a suppressor; a menin inhibitor blocks one interface and leaves the protein in place. Nobody should read a PanNET talazoparib result as a paediatric leukaemia rationale without the repair endpoint being measured first.
- And PARP inhibitors are not benign in this population. Any proposal to combine them with a menin inhibitor in infants must carry the haematological toxicity of both, in patients already marrow-suppressed, which the ledger's one-line version of this experiment does not.
- Does a pharmacological menin inhibitor reduce BRCA2 or RAD51 foci in a KMT2A-rearranged line? That is the one-step version of the PanNET result and it decides whether the rationale transfers.
- Can a HOXA9 rescue arm separate the BRCA2 route from the HOXA9 route? Without it a synergy result explains nothing.
- Has the combination appeared in a conference abstract or preprint? The bounded query covered PubMed only.
- E1moderate confidence 3 of 5: One strong study, or several smaller consistent ones.analysis
No study combining a pharmacological menin inhibitor with a PARP inhibitor in KMT2A-rearranged leukaemia was retrievable: a PubMed query pairing eight menin inhibitor names with five PARP inhibitor terms returned 5 records of which 0 were on topic on reading, against controls of 1,275 for PARP inhibitor AND leukemia and 5 for HOXA9 AND PARP.
Bounded absence query with controls for the menin-inhibitor plus PARP-inhibitor combination, PubMed esearch 2026-09-20 · Oncofusion knowledge graph (internal analysis) 2026What this evidence cannot show
Run 2026-09-20. **Moderate because this is a negative retrieval claim.** The drug-name list may be incomplete despite covering eight agents (rule 72), combination work often appears first in conference abstracts and preprints that PubMed does not index well, and the query was not run against ClinicalTrials.gov or bioRxiv. **Two of the five hits are query artefacts** - cleaved PARP as an apoptosis marker, and an author surnamed Menin - which shows the query shape is loose in a way that inflates rather than deflates the count.
Evidence that argues against this (1)
- A menin-ASO paper in triple-negative breast cancer explicitly proposes 'combination with chemo or PARP inhibitors' as a rationale for menin-positive tumours. So the combination is not an unconsidered idea - it is in print as a proposal in another disease, untested there too, which weakens any claim that this graph originated it. Nguyen DT et al. 2021
- E2strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.journal-article
In pancreatic neuroendocrine tumour cells, MEN1 knockout reduced double-strand break repair capability through a homologous recombination defect with reduced BRCA2 expression, and MEN1-mutant cells showed increased sensitivity to the trapping PARP1 inhibitor talazoparib in vitro.
Loss of MEN1 function impairs DNA repair capability of pancreatic neuroendocrine tumors. · Lakiza O, Lutze J, Vogle A, et al. · Endocr Relat Cancer 2022What this evidence cannot show
**The system is pancreatic neuroendocrine tumour, where MEN1 is a tumour suppressor and is genetically DELETED** - not KMT2A-rearranged leukaemia, where menin is an oncogenic cofactor pharmacologically displaced from one interface while the protein remains (rule 73). **Knockout is not pocket blockade**, which is the central caveat and the reason this is a rationale rather than a prediction. In vitro sensitivity only; combined with irradiation rather than as PARP-inhibitor monotherapy in the headline proposal. Read from the abstract and title; the full text was not obtained.
Evidence that argues against this (2)
- The direction of the menin-HOXA9 relationship in leukaemia argues the opposite way for part of this. If HR proficiency in KMT2A-rearranged disease is HOXA9-driven and menin inhibition lowers HOXA9, then HR falls by that route too - which is convergent rather than contradictory, but means a positive synergy result would NOT distinguish the BRCA2 mechanism from the HOXA9 one. The experiment needs a repair endpoint and a HOXA9 rescue arm to tell them apart, and as currently written in the ledger it has neither. Nguyen DT et al. 2021
- A SUPPORT OFFERED AND WITHDRAWN THE SAME DAY. An occupancy analysis appeared to show homologous recombination genes enriched among menin-bound promoters at 1.76x, and was briefly recorded as a third independent line for this combination. Its own housekeeping control killed it: ribosome biogenesis enriches at 1.78x and translation, OXPHOS and DNA repair all sit in the same band, so the pattern is menin's general preference for active promoters and not a repair signal. **This experiment's rationale remains the single PanNET functional result.** Oncofusion knowledge graph (internal analysis) 2026
THIRD INDEPENDENT LINE TO THE SAME PLACE: HR genes are 1.76x enriched among menin-bound promoters, found separately from the PanNET result that entry rests on.
MENIN OUTSIDE THE INTERFACE: MEN1 loss impairs double-strand break repair via reduced BRCA2 in pancreatic neuroendocrine tumour, a function the KMT2A-interface framing does not cover.
- related to →HOXA9 — required for survival of established MLL-rearranged leukemia, and dispensable for starting it
THE RATIONALE THE LEDGER USED, and the one a synergy result could not distinguish from the BRCA2 route without a rescue arm.
THE TARGET OF THE PROPOSED SECOND AGENT, and the source of two query false positives - cleaved PARP is an apoptosis readout.
SAME QUEUE, SAME SESSION: a second advertised experiment checked against the literature. This one's absence SURVIVED its bounded query, which is what a working queue looks like.
A BOUNDED ABSENCE THAT HELD: 5 hits, 0 on topic, controls at 1,275 and 5.
a third, mechanistic rationale for the menin-PARP combination that does not rest on promoter occupancy
SECOND, INDEPENDENT POINTER TO DNA REPAIR: menin-bound genes are enriched for DNA repair (226/388), arrived at separately from the MEN1-loss homologous-recombination result.