Oncofusion

start / graph / concept-why-pml-rara-became-curable

Explainerstrong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.Q11-comparative

Why PML::RARA became curable — and what KMT2A::AFF1 does not share with it

also called APL lesson · differentiation therapy paradigm · degrader paradigm

Acute promyelocytic leukaemia went from the deadliest leukaemia to the most curable without chemotherapy, because its fusion protein happened to have two rare features: a chemical slot that vitamin A fits into, and a structure that arsenic makes the cell destroy. KMT2A::AFF1 has neither, which is the single most important comparison in this whole project.

The same thing, for a scientist

APL is the only fusion-driven leukaemia cured by drugs that act directly on the fusion oncoprotein. Two independent properties made it tractable: (i) PML::RARA retains the RARA ligand-binding domain, a genuine nuclear-receptor pocket, so a physiological ligand (ATRA) at pharmacological dose evicts corepressors and forces differentiation; and (ii) the PML moiety makes the fusion a substrate for ligand- and arsenic-induced degradation, so ATRA plus arsenic trioxide destroys the oncoprotein and clears leukaemia-initiating cells rather than merely differentiating blasts. Differentiation and LIC eradication are separable, and it is degradation that produces cure. KMT2A::AFF1 has no ligand pocket and no known partner-encoded degradation handle, so neither arm of the APL strategy transfers.

Go deeper
Feature 1 — a retained ligand-binding pocket, which is an accident of which gene broke1 deeper

RARA is a nuclear hormone receptor. Its ligand-binding domain is a well-formed hydrophobic pocket evolved to hold a small lipophilic molecule, and the t(15;17) breakpoint retains it intact. So the fusion arrives with a drug site already built in, and the drug (all-trans retinoic acid) is a natural metabolite that needs no medicinal chemistry at all.

The oncogenic mechanism is that PML forces RARA into a homodimer that binds retinoic-acid response elements with abnormally high avidity and holds corepressors even at physiological ligand concentration. Pharmacological ATRA overcomes that, swaps corepressor for coactivator, and the promyelocyte differentiates. The centrality of the dimer is demonstrated directly: forced RARA homodimers alone are sufficient to prime mice for APL-like disease 1, which tells us the fusion's essential contribution is dimerisation of a ligandable receptor.

Why this is not generalisable

Fewer than a handful of recurrent fusion partners contribute a nuclear-receptor ligand-binding domain. Across the paediatric leukaemia atlas, PML::RARA is the only one. So this feature is not a lesson about fusions; it is a lesson about how lucky APL was, and a warning against reasoning from it.

Feature 2 — the fusion is degradable, and degradation, not differentiation, is what cures1 deeper

This is the deeper and more transferable lesson, and it is often mis-told. Retinoic acid triggers both differentiation and PML::RARA catabolism. Arsenic trioxide acts on the PML moiety and also degrades the fusion. In mouse APL, ATRA arrests leukaemia-initiating cells and clears them in vivo; ATRA plus arsenic synergise to clear LICs through cooperative PML-RARA degradation, and that combination does not enhance differentiation 2. In a PLZF-RARA model, retinoic acid causes full differentiation but neither LIC loss nor remission, which dissociates the two phenotypes cleanly. A cAMP-dependent phosphorylation site in PML-RARA is required for ATRA-induced degradation and LIC clearance.

The clinical payoff is exactly what that biology predicts: in the randomised APL0406 trial, ATRA plus arsenic trioxide gave 100% complete remission and 97% two-year event-free survival versus 86% for ATRA plus chemotherapy, with less haematological toxicity — a chemotherapy-free cure 3. The general principle, stated by the authors of the degradation work, is that therapy-triggered destruction of an oncoprotein may be a general way to eradicate cancer-initiating cells 24.

Two things were left out of the original version of this entry, and both cut against the analogy. First, the degradation handle is a chemistry, not a strategy: arsenic works because it binds cysteine pairs in the zinc fingers of the PML RBCC, and that binding is what drives oligomerisation, UBC9 recruitment, SUMOylation and destruction. Second, degradation therapy is not resistance-proof. Two of fifteen relapsed APL patients treated with arsenic were clinically resistant and both carried missense mutations in the PML B2 domain; the mutant fusion no longer SUMOylated, no longer oligomerised and no longer partitioned into the insoluble fraction on arsenic, while remaining a functional oncoprotein 7. PML-A216V is now trackable by ddPCR before overt resistance appears. Escape at the drug-binding surface with the oncogenic mechanism intact is exactly the BCR-ABL1 failure mode, and it happens at the degradation handle too.

The one feature that might transfer: make the fusion degradable on purpose

APL got degradation for free from PML. Modern chemistry can install it: PROTACs, molecular glues and bifunctional degraders convert a binding event into a degradation event. The strategic implication for the lab is precise and actionable — a menin inhibitor is an occupancy drug, and the APL lesson says occupancy forces differentiation while degradation eradicates. A KMT2A-fusion degrader (or a menin degrader, which would deplete the adaptor rather than block it) is the closest available analogue of the APL mechanism and is a first-rank Phase-2 hypothesis.

Does KMT2A::AFF1 share anything? A line-by-line answer
Feature that made APL curablePresent in KMT2A::AFF1?
Retained ligand-binding pocket on the fusionNo. The only pocket-bearing catalytic module of KMT2A is the SET domain, and it is always lost at the breakpoint. AFF1 contributes no pocket.
Retained enzyme active siteNo. Neither half is an enzyme in the fusion.
Partner-encoded degradation handleNo, and worse than absent. KMT2A fusions are actively degradation-resistant: wild-type MLL is cleared bimodally by SCF(Skp2) at S phase and APC(Cdc20) at late M phase, degradation initiates in the N-terminal ~1400 aa that every fusion retains, and the prevalent MLL fusions tested resist it 6. This is not a missing handle; it is a deleted disposal route.
Forced dimerisation as the essential oncogenic actUnclear, and worth testing. APL, t(8;21) and inv(16) all depend on a partner-supplied oligomerisation interface; whether AFF1 supplies an analogous obligate self-association is untested.
An obligate partner with a real pocketYes — this is the one thing it has. Menin. That is why menin inhibition works at all, and it is the lab's hypothesis.
Disease biology that is a single differentiation block, curable by releasing itPartly. Menin inhibitors do induce differentiation and differentiation syndrome, exactly as ATRA does, which is encouraging; but responses are transient, consistent with occupancy without degradation.

The honest conclusion: KMT2A::AFF1 shares APL's phenotype (a releasable differentiation block) but not its chemistry (no pocket, no degradation handle). The lab inherits APL's strategic idea — destroy the driver, do not merely occupy it — rather than any of its molecules.

And that strategic idea is weaker than this entry originally claimed. Three corrections from the cycle-33 counter-pass:

  1. Degradation is escapable. PML B2-domain mutations give clinical arsenic resistance with the oncoprotein intact 7. A degrader is a drug with a binding site, and binding sites mutate. The hope that a degrader would 'escape the interface resistance mutations that defeat revumenib' has a published counterexample in the very disease this analogy is drawn from.
  2. KMT2A fusions are the hardest possible substrate for this strategy. Degradation of wild-type MLL initiates in the retained N-terminal ~1400 aa and the fusions resist it 6. APL's driver happened to be unusually easy to destroy; KMT2A's is selected to be hard.
  3. APL's cure rate is partly a selection artefact. 29% of unselected APL patients die within 30 days, mostly of haemorrhage, a third without ever receiving ATRA 8. Any argument of the form 'APL went from worst to best, therefore KMT2A-r can' should be discounted by the fraction of APL patients the cure never reaches.
Correction (cycle 33): what was withdrawn from this entry

Three statements in the original entry have been weakened or withdrawn on evidence that already existed when it was written.

  • Withdrawn: the implication, carried in transferable_lesson and in open question 2, that a degrader would be immune to the interface-resistance problem. Arsenic, the founding degradation therapy, is defeated by PML B2-domain point mutation in patients 7. data.degradation_is_not_resistance_proof records this.
  • Sharpened: 'no known partner-encoded degradation handle' understated the problem. KMT2A fusions are degradation-resistant, and the resistance maps to the retained N-terminus 6.
  • Qualified: 'the most curable leukaemia' is a statement about trial populations; unselected 30-day mortality is 29% 8.

What survives, and is if anything strengthened: differentiation and eradication are separable 2; occupancy alone is not enough; and the ligand pocket was an accident of which gene broke.

Facts and figures
comparison axiswhat made a fusion oncoprotein directly druggable
fusionfusion-pml-rara
cure mechanismdegradation of the fusion oncoprotein, which clears leukaemia-initiating cells; differentiation alone is insufficient
transferable lessonOccupancy forces differentiation, degradation eradicates -- so a degrader is worth building. QUALIFIED in cycle 33: a degrader is still an occupancy drug with an extra step, and the founding degradation therapy (arsenic) is defeated in patients by PML B2-domain point mutation with the oncoprotein intact (PMID 21613260). And the substrate is adverse: KMT2A fusions resist the SCF(Skp2)/APC(Cdc20) turnover that clears wild-type MLL, with degradation initiating in the N-terminal ~1400 aa every fusion retains (PMID 17908926).
non transferablethe ligand pocket and the natural-product drug were accidents of which gene broke
implication for kmt2aprioritise degrader modalities (fusion-directed or menin-directed) alongside occupancy inhibitors, and test whether the KMT2A fusion has any induced-degradation handle at all
tractability features
[
  {
    "feature": "retained nuclear-receptor ligand-binding pocket",
    "present": true,
    "drug": "all-trans retinoic acid",
    "shared_with_kmt2a_aff1": false
  },
  {
    "feature": "partner-encoded degradation handle",
    "present": true,
    "drug": "arsenic trioxide (plus ATRA)",
    "shared_with_kmt2a_aff1": false
  },
  {
    "feature": "forced oligomerisation as the essential oncogenic act",
    "present": true,
    "drug": "none directly",
    "shared_with_kmt2a_aff1": null
  },
  {
    "feature": "obligate partner with a real small-molecule pocket",
    "present": false,
    "drug": "n/a",
    "shared_with_kmt2a_aff1": true
  },
  {
    "feature": "releasable single differentiation block",
    "present": true,
    "drug": "ATRA",
    "shared_with_kmt2a_aff1": true
  },
  {
    "feature": "degradation handle is resistance-proof",
    "present": false,
    "drug": "arsenic trioxide (defeated by PML B2-domain mutation)",
    "shared_with_kmt2a_aff1": false
  }
]
degradation is not resistance proof
{
  "evidence": "PML B2-domain missense mutations in PML-RARA confer clinical arsenic resistance (2/15 relapsed APL patients), abolishing arsenic-induced SUMOylation, oligomerisation and insolubility while the oncoprotein persists",
  "pmid": "21613260",
  "tracking": "PML A216V detectable by ddPCR before overt clinical resistance (PMID 30651632)",
  "implication": "a degrader is an occupancy drug with an extra step; it inherits drug-binding-surface resistance",
  "title": "Missense mutations in PML-RARA are critical for the lack of responsiveness to arsenic trioxide treatment.",
  "authors": "Goto E, Tomita A, Hayakawa F, et al.",
  "venue": "Blood",
  "year": 2011,
  "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."
}
kmt2a fusion is degradation resistant
{
  "mechanism": "wild-type MLL cleared by SCF(Skp2) at S phase and APC(Cdc20) at late M phase; degradation initiates in the N-terminal ~1400 aa retained by every fusion; prevalent MLL fusions tested are resistant",
  "pmid": "17908926",
  "implication": "the APL degradation arm is being proposed against the one substrate class selected for escape from degradation",
  "title": "Bimodal degradation of MLL by SCFSkp2 and APCCdc20 assures cell cycle execution: a critical regulatory circuit lost in leukemogenic MLL fusions.",
  "authors": "Liu H, Cheng EH, Hsieh JJ",
  "venue": "Genes Dev",
  "year": 2007,
  "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."
}
apl cure rate is conditional on surviving to randomisation
{
  "unselected_30_day_mortality_pct": 29,
  "haemorrhagic_fraction_of_early_deaths_pct": 41,
  "never_received_atra_among_early_deaths_pct": 35,
  "population": "Swedish Adult Acute Leukemia Registry 1997-2006, n=105 APL",
  "pmid": "21502956",
  "title": "Continuing high early death rate in acute promyelocytic leukemia: a population-based report from the Swedish Adult Acute Leukemia Registry.",
  "authors": "Lehmann S, Ravn A, Carlsson L, et al.",
  "venue": "Leukemia",
  "year": 2011,
  "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."
}
How this could bite us back
  • Differentiation therapy in a high blast burden causes differentiation syndrome, which is fatal if unmanaged; any KMT2A strategy that succeeds at releasing the block inherits this toxicity.
  • If cure in APL requires degradation and we can only achieve occupancy in KMT2A-r, we should expect deep remissions followed by relapse from a surviving initiating cell, which is exactly what the menin-inhibitor data show so far.
  • Chasing the APL analogy too literally risks a decade spent looking for a pocket that the KMT2A fusion structurally cannot have.
  • If 29% unselected early death is the honest denominator for APL, then 'APL was cured' is partly a statement about which patients reach a protocol, and the same selection flatters any KMT2A-r result reported from relapsed/refractory trial populations.
  • A KMT2A-fusion degrader would have to out-compete a protein that has already deleted three of its own disposal routes, so the achievable depth of depletion, not the potency of the binder, is the likely binding constraint.
Still open
  1. Is there any small molecule or molecular glue that induces degradation of a KMT2A fusion protein, and if not, is the fusion inherently resistant to degradation because it lost the PHD2/PHD3 turnover module?
  2. Would a menin degrader outperform a menin occupancy inhibitor -- and, given that arsenic is defeated by PML B2-domain mutation, on what basis would we expect a degrader to escape MEN1 interface mutations rather than simply select for mutations at the degrader's own binding site?
  3. Is the differentiation syndrome seen with menin inhibitors a sign that we are on the APL path, or a sign that we are getting the differentiation arm without the eradication arm?
  4. Can the arsenic mechanism be repurposed conceptually — is there any partner-encoded structural feature in a KMT2A fusion (an AFF1 low-complexity region, an exposed junction) that a covalent or oxidative agent could exploit the way arsenic exploits PML?
  5. What is the actual half-life of KMT2A::AFF1 versus wild-type KMT2A in one leukaemic cell? Every degradation argument in this graph, for and against, rests on a measurement nobody has published.
Every source behind this entry (8)
  1. E1strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.paper

    Forced RARA homodimers are sufficient to prime mice for APL-like leukemia, showing that the essential oncogenic contribution of PML in PML::RARA is to dimerise a ligandable nuclear receptor.

    Forced retinoic acid receptor alpha homodimers prime mice for APL-like leukemia · Sternsdorf T, Phan VT, Maunakea ML, et al. · Cancer Cell 2006
    PMID 16473276DOI
    What this evidence cannot show

    Engineered dimerisation constructs with long latency that prime for rather than fully reproduce human APL, so PML-specific contributions such as nuclear-body disruption and degradation sensitivity are not excluded by this experiment.

    Evidence that argues against this (1)
    • If PML's only essential contribution were to dimerise RARA, then PML would be scaffolding and the therapy would be aimed elsewhere - but the curative drug binds PML itself. Arsenic binds directly to cysteine residues in the zinc fingers of the RBCC domain of PML and PML-RARA, and that binding INDUCES PML oligomerisation, increasing recruitment of the SUMO-conjugating enzyme UBC9 and driving SUMOylation and degradation. The PML moiety is therefore both the dimerisation module and the drug-binding module, and the second of those is the one that made APL curable. 'The essential oncogenic contribution of PML is to dimerise a ligandable nuclear receptor' names the wrong half for the purposes of this entry's argument. Zhang XW et al. 2010
  2. E2strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.paper

    Retinoic acid and arsenic synergise to clear APL leukaemia-initiating cells through cooperative PML-RARA degradation, while that combination does not enhance differentiation, and a PLZF-RARA model differentiates fully without LIC loss or remission.

    Eradication of acute promyelocytic leukemia-initiating cells through PML-RARA degradation · Nasr R, Guillemin MC, Ferhi O, et al. · Nature Medicine 2008
    PMID 19029980DOI
    What this evidence cannot show

    Mouse APL models and ex vivo LIC assays; LIC frequency is inferred from transplantation, and the degradation-versus-differentiation dissociation is cleanest in the non-PML (PLZF-RARA) comparator rather than in human disease.

    Evidence that argues against this (2)
    • The degradation arm is escapable by point mutation, exactly like an active-site drug. Two of fifteen patients with relapsed/refractory APL treated with arsenic trioxide were clinically resistant, and both carried missense mutations in the PML B2 domain of PML-RARA. The mutant fusion failed to SUMOylate, failed to oligomerise and failed to shift from the soluble to the insoluble fraction on arsenic exposure, while remaining an intact oncoprotein. Degradation therapy therefore does not escape interface resistance; it has its own drug-binding surface and that surface mutates. Goto E et al. 2011
    • PML A216V is now detectable by droplet digital PCR before overt clinical arsenic resistance, i.e. it behaves as a selectable clone in the same way BCR-ABL1 kinase-domain mutants do. The 'degradation cannot be escaped' framing is not supported by the APL clinic itself. Alfonso V et al. 2019
  3. E3established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.paper

    In the randomised APL0406 trial, ATRA plus arsenic trioxide achieved complete remission in 100% of evaluable patients and 97% two-year event-free survival versus 86% with ATRA plus chemotherapy, with less haematological toxicity.

    Retinoic acid and arsenic trioxide for acute promyelocytic leukemia · Lo-Coco F, Avvisati G, Vignetti M, et al. · New England Journal of Medicine 2013
    PMID 23841729DOINCT00482833
    What this evidence cannot show

    Restricted to low-to-intermediate-risk APL (white-cell count 10 x 10^9/L or less), modest sample size per arm, median follow-up 34 months, and more hepatic toxicity in the arsenic arm; high-risk APL was not addressed.

    Evidence that argues against this (1)
    • APL0406's 100% complete remission is a trial-population number, not a disease number. In the population-based Swedish Adult Acute Leukemia Registry, 30 of 105 unselected APL patients (29%) died within 30 days of diagnosis (median 4 days), 41% of those early deaths were haemorrhagic, and 35% of early-death patients never received ATRA at all. Patients who die in the first week are never randomised, so the curability of APL is partly an artefact of who survives to enrol. Lehmann S et al. 2011
  4. E4moderate confidence 3 of 5: One strong study, or several smaller consistent ones.review

    The APL experience is best understood as oncoprotein-degradation therapy, and the authors of the paradigm explicitly propose degradation of the driver as a general route to eradicating cancer-initiating cells.

    Differentiation therapy revisited · de The H · Nature Reviews Cancer 2018
    PMID 29192213DOI
    What this evidence cannot show

    A review by the principal architect of the degradation model, so it argues a position; the claim that degradation generalises across cancers is an extrapolation and has not been demonstrated for any chromatin fusion.

    Evidence that argues against this (2)
    • The degradation handle is a cysteine-rich zinc-finger chemistry, not a generic strategy. Arsenic works because it binds cysteine residues in the zinc fingers of the PML RBCC and thereby drives oligomerisation, UBC9 recruitment, SUMOylation and degradation. Nothing retained in a KMT2A fusion has that architecture, so 'install degradation with modern chemistry' is a design problem from scratch rather than a repurposing of the APL mechanism. Zhang XW et al. 2010
    • The KMT2A fusion is not merely lacking a degradation handle; it is actively degradation-resistant. Wild-type MLL is cleared in two defined cell-cycle windows by SCF(Skp2) at S phase and APC(Cdc20) at late M phase, degradation initiates within the N-terminal approximately 1400 amino acids that every KMT2A fusion retains, and the prevalent MLL fusions tested were resistant to that degradation. A therapy whose whole logic is 'destroy the driver' is being proposed against a protein whose defining lesion is escape from destruction. Liu H et al. 2007
  5. E5strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.paper

    Menin inhibition in KMT2A-rearranged leukaemia produces differentiation and differentiation syndrome, phenocopying the APL differentiation arm, but responses are partial and frequently not durable.

    The menin inhibitor revumenib in KMT2A-rearranged or NPM1-mutant leukaemia · Issa GC, Aldoss I, DiPersio J, et al. · Nature 2023
    PMID 36922593DOI
    What this evidence cannot show

    Single-arm phase 1 in a heavily pre-treated population; the comparison to APL is interpretive, and many responders were consolidated with transplant so durability of drug alone is unknown. Abstract only for this cycle's counter-pass; the differentiation-syndrome comparison to ATRA remains interpretive.

    Evidence that argues against this (1)
    • The phenocopy is shallower than the comparison implies, and the registrational numbers are the ones to use. In the phase 2 registration analysis of AUGMENT-101, among 57 patients with centrally confirmed KMT2A rearrangement, the CR+CRh rate was 22.8% (95% CI 12.7-35.8); the differentiation that is offered as the APL parallel came with grade >=3 differentiation syndrome in 16.0%, QTc prolongation in 13.8% and febrile neutropenia in 37.2%. APL's differentiation arm converts nearly all patients; menin inhibition converts about a quarter and the toxicity that accompanies it is the same on-mechanism toxicity without the same benefit. Issa GC et al. 2024
  6. E6strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.paper

    Wild-type MLL is degraded bimodally by SCF(Skp2) at S phase and APC(Cdc20) at late M phase, degradation initiates within the N-terminal ~1400 amino acids retained by every KMT2A fusion, and the prevalent MLL fusions tested are resistant to that degradation.

    PMID 17908926DOI
    What this evidence cannot show

    Abstract read this cycle (the graph's subtraction beat holds a full-text reading). HeLa and overexpression systems, 2007, and 'tested prevalent MLL fusions' is a small panel; no head-to-head half-life measurement of fusion versus wild-type MLL in one leukaemic cell has ever been published. The direction of the result is nevertheless the opposite of what a degradation-based strategy needs.

    Evidence that argues against this (1)
    • This is presented as the decisive disanalogy - APL is cured by destroying the oncoprotein, KMT2A fusions resist degradation - and the inference assumes the degraded object has to be the fusion. It does not. At the WDR5 WIN site, an obligate partner interface of the KMT2A complex, occupancy-based inhibitors gave modest killing and no in vivo efficacy, while a structure-designed VHL-recruiting degrader of the PARTNER (MS67, with two ternary-complex crystal structures) outperformed them at suppressing WDR5-regulated transcription, at reducing the chromatin-bound fraction of MLL complex components and c-MYC, and at suppressing MLL-rearranged AML patient cells in vitro and in vivo, while being well tolerated. Degrading an obligate partner is a demonstrated route in exactly this disease, so 'the APL strategy is aimed at the worst possible substrate' holds only if the substrate must be the fusion. Yu X et al. 2021
  7. E7strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.paper

    Arsenic resistance in APL arises from PML B2-domain missense mutations in PML-RARA that abolish arsenic-induced SUMOylation, oligomerisation and insolubility while leaving the oncoprotein intact.

    Missense mutations in PML-RARA are critical for the lack of responsiveness to arsenic trioxide treatment. · Goto E, Tomita A, Hayakawa F, Atsumi A, Kiyoi H, Naoe T · Blood 2011
    PMID 21613260DOI
    What this evidence cannot show

    Abstract only. Two resistant patients out of fifteen treated, with mechanism reconstructed by overexpression in HeLa; small n, but the direction is unambiguous and has since been made clinically trackable by ddPCR (PMID 30651632).

    Evidence that argues against this (1)
    • The B2 mutations are real resistance but they are not resistance to a blockade, which changes what they predict for other fusions. Arsenic acts by binding RBCC zinc-finger cysteines and INDUCING PML oligomerisation, which drives UBC9 recruitment, SUMOylation and degradation. The resistance mutations therefore abolish a drug-induced GAIN of function, not the drug's occupancy of an obligate site. A fusion whose therapy works by blocking an interface faces a different mutational landscape from one whose therapy works by inducing an interface, and this row should not be read across to the menin-interface case without that distinction. Zhang XW et al. 2010
  8. E8established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.paper

    In an unselected population-based APL cohort, 29% of patients died within 30 days of diagnosis, 41% of those deaths were haemorrhagic, and a third of early-death patients never received ATRA.

    PMID 21502956DOI
    What this evidence cannot show

    Abstract only. Swedish adults 1997-2006, n=105 APL; predates routine ATRA-on-suspicion policy and universal front-line ATO, so modern early-death rates are lower -- but they remain far above the zero implied by trial cure rates.

    Evidence that argues against this (1)
    • The 29% figure is an aggregate and the burden is not uniform, which matters because the entry uses it as a blanket correction to the cure narrative. A later update from the same Swedish Acute Leukemia Registry covering 1997-2013 is titled 'Early death rates remain high in high-risk APL', locating the persisting problem in the high-risk stratum specifically rather than across all APL. The refinement is a strengthening of the point for high-risk patients and a weakening of it as a statement about APL as a whole. NOTE ON READING: PubMed indexes no abstract for this report, so this is cited from the title and citation record only; the stratum-specific numbers have not been read and are not quoted here. Lehmann S et al. 2017
Connected to