Imatinib: one molecule, three unrelated rare diseases, twenty years of cross-subsidy
also called imatinib platform precedent · one drug many orphan indications
Imatinib was built for one leukemia caused by the BCR-ABL1 fusion, then turned out to also treat stomach tumours driven by a different kinase and a rare blood disorder driven by a third. One molecule ended up serving several separate rare diseases because it blocked a family of related targets.
The same thing, for a scientist
Imatinib is the field's oldest and most complete demonstration that a single mechanism-directed molecule can serve multiple unrelated rare populations. It was developed against BCR::ABL1 in chronic myeloid leukemia, where 53 of 54 patients in the first-in-human chronic-phase cohort achieved complete haematologic response. It then produced partial responses in 53.7% of 147 patients with advanced KIT-driven gastrointestinal stromal tumour, and normalised eosinophil counts in 9 of 11 patients with hypereosinophilic syndrome, in whom the FIP1L1::PDGFRA fusion kinase was identified as the target. Three orphan populations, one asset, one shared structural feature: an ATP-binding site conformation common to ABL, KIT and PDGFRA.
The three populations2 deeper
| Disease | Driver | Result | Source |
|---|---|---|---|
| Chronic myeloid leukemia, chronic phase | BCR::ABL1 | 53/54 complete haematologic response at 300 mg or more daily | 1 |
| Advanced gastrointestinal stromal tumour | activating KIT | 53.7% partial response, 27.9% stable disease (n=147) | 2 |
| Idiopathic hypereosinophilic syndrome | FIP1L1::PDGFRA | 9/11 responses lasting more than three months | 3 |
None of these populations knew about the others. The FIP1L1::PDGFRA fusion was discovered because patients responded to imatinib, a striking inversion of the usual order: the drug found the target 3.
What the precedent actually teaches
Three lessons, one of which is uncomfortable.
- Shared structural chemistry generalises across diseases that share nothing else. Imatinib binds an inactive-conformation ATP pocket shared by ABL, KIT and PDGFRA. The analogous claim for this project is that menin's peptide-binding pocket is shared by every KMT2A fusion and by NPM1-mutant AML, which is exactly what
platform-one-asset-two-populations-menindocuments. - The second and third indications were not planned. They were discovered after approval, by clinicians giving an available drug to patients with a biologically plausible disease. That argues for making mechanism-directed agents available and characterised, not for trying to predict every indication in advance.
- The uncomfortable one: the first indication was the commercially large one. CML is far commoner than GIST or hypereosinophilic syndrome. Imatinib's rare indications were subsidised by a bigger first market, exactly as KMT2A-rearranged leukemia is now being subsidised by NPM1-mutant AML. A platform of this kind has historically needed an anchor indication large enough to pay for it, and infant KMT2A-rearranged leukemia can never be that anchor.
Where imatinib did reach children, and how long it took
Imatinib did eventually reach children with Philadelphia-chromosome-positive ALL, with long-term follow-up from Children's Oncology Group study AALL0031 showing durable benefit when imatinib was added to intensive chemotherapy 4. That paediatric result is roughly a decade after the adult CML result.
A decade is the honest estimate of the lag between an adult mechanism-directed approval and a convincing paediatric frontline dataset, absent a deliberate intervention to shorten it. Revumenib was approved in November 2024. On the imatinib timetable, a frontline infant dataset arrives in the mid-2030s. The children diagnosed between now and then are the cost of that lag, and shortening it is a legitimate object of this project (see hyp-menin-inhibitors-belong-frontline-in-infants).
| asset | imatinib |
|---|---|
| shared structural feature | inactive-conformation ATP-binding pocket common to ABL1, KIT and PDGFRA |
| anchor indication was largest | true |
| adult to paediatric lag years | 10 |
indications — the numbers behind the chart
[
{
"disease": "chronic myeloid leukemia",
"driver": "BCR::ABL1",
"key_result": "53/54 complete haematologic response",
"pmid": "11287972",
"title": "Efficacy and safety of a specific inhibitor of the BCR-ABL tyrosine kinase in chronic myeloid leukemia.",
"authors": "Druker BJ, Talpaz M, Resta DJ, et al.",
"venue": "N Engl J Med",
"year": 2001,
"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."
},
{
"disease": "gastrointestinal stromal tumour",
"driver": "activating KIT",
"key_result": "53.7% partial response of 147",
"pmid": "12181401",
"title": "Efficacy and safety of imatinib mesylate in advanced gastrointestinal stromal tumors.",
"authors": "Demetri GD, von Mehren M, Blanke CD, et al.",
"venue": "N Engl J Med",
"year": 2002,
"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."
},
{
"disease": "hypereosinophilic syndrome",
"driver": "FIP1L1::PDGFRA",
"key_result": "9/11 responses >3 months",
"pmid": "12660384",
"title": "A tyrosine kinase created by fusion of the PDGFRA and FIP1L1 genes as a therapeutic target of imatinib in idiopathic hypereosinophilic syndrome.",
"authors": "Cools J, DeAngelo DJ, Gotlib J, et al.",
"venue": "N Engl J Med",
"year": 2003,
"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."
},
{
"disease": "paediatric Ph+ ALL",
"driver": "BCR::ABL1",
"key_result": "durable benefit added to chemotherapy",
"pmid": "24441288",
"title": "Long-term follow-up of imatinib in pediatric Philadelphia chromosome-positive acute lymphoblastic leukemia: Children's Oncology Group study AALL0031.",
"authors": "Schultz KR, Carroll A, Heerema NA, et al.",
"venue": "Leukemia",
"year": 2014,
"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."
}
]- The imatinib precedent supports the platform thesis and simultaneously predicts the failure mode this project is fighting: the anchor indication is adult, the paediatric indication arrives a decade later, and nothing in the structure corrects that on its own.
- Every one of imatinib's three indications developed resistance through target-site mutation (ABL T315I, KIT secondary mutations, PDGFRA T674I). A platform built on a single shared interface should expect a single shared class of resistance mutation, which is what MEN1 M327I and relatives already look like.
- Is there any historical example of a mechanism-directed platform whose anchor indication was paediatric, or is adult-anchored cross-subsidy the only pattern that has ever worked?
- What deliberately shortened the adult-to-paediatric lag in any oncology asset, and can that mechanism be applied to menin inhibition?
- Could a KMT2A menin inhibitor find an unanticipated third indication the way imatinib found hypereosinophilic syndrome, and what disease should be screened for menin dependency to look?
- E1established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.paper
In the first-in-human study of imatinib in chronic myeloid leukemia, complete haematologic responses occurred in 53 of 54 patients treated at doses of 300 mg or more daily in chronic phase after interferon failure.
Efficacy and safety of a specific inhibitor of the BCR-ABL tyrosine kinase in chronic myeloid leukemia · Druker BJ, Talpaz M, Resta DJ, et al. · New England Journal of Medicine 2001What this evidence cannot show
Single-arm dose-escalation in a selected interferon-refractory population; haematologic response is a surrogate for the cytogenetic and molecular endpoints that later defined CML care.
Evidence that argues against this (1)
- 53 of 54 complete haematologic responses is the most-cited number in targeted oncology and the least transferable. It was obtained in CHRONIC-phase CML, a single-driver disease in a chronic phase with an intact haematopoietic reserve. The same drug in blast crisis produced short-lived responses, and menin inhibitors are being tested in relapsed acute leukaemia, which is the blast-crisis analogue, not the chronic-phase one. New England Journal of Medicine 2001
- E2established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.paper
In an open-label randomised trial of 147 patients with advanced gastrointestinal stromal tumour, imatinib produced partial response in 53.7% and stable disease in 27.9%, with no complete responses and early resistance in 13.6%.
Efficacy and safety of imatinib mesylate in advanced gastrointestinal stromal tumors · Demetri GD, von Mehren M, Blanke CD, et al. · New England Journal of Medicine 2002What this evidence cannot show
Randomised between two doses rather than against a comparator; median follow-up only 24 weeks after response onset, so durability was unknown at publication. Early resistance in 13.6% foreshadowed the KIT secondary-mutation problem.
Evidence that argues against this (1)
- The imatinib platform story understates how genotype-dependent each indication turned out to be, which is the part that matters when arguing a menin inhibitor will travel across fusions. In 127 patients with advanced GIST on the phase 2 study, the partial response rate was 83.5% with KIT exon 11 mutations, 47.8% with exon 9 mutations (p=.0006) and 0.0% with no detectable KIT or PDGFRA mutation (p<.0001), and only a SUBSET of PDGFRA mutant isoforms were imatinib-sensitive in vitro. One molecule reaching three diseases is therefore not the same as one molecule reaching three diseases uniformly; within each indication the benefit collapsed onto specific alleles. Heinrich MC et al. 2023
- E3established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.paper
Nine of eleven patients with idiopathic hypereosinophilic syndrome treated with imatinib had responses lasting more than three months, and the responsible target was identified as the FIP1L1::PDGFRA fusion kinase created by an interstitial 4q12 deletion, with relapse in one patient accompanied by a resistance mutation, PDGFRA T674I.
A tyrosine kinase created by fusion of the PDGFRA and FIP1L1 genes as a therapeutic target of imatinib in idiopathic hypereosinophilic syndrome · Cools J, DeAngelo DJ, Gotlib J, et al. · New England Journal of Medicine 2003What this evidence cannot show
Eleven patients. The fusion was detected in 9 of 16 patients with the syndrome and in only 5 of the 9 durable responders, so some responses were not explained by FIP1L1::PDGFRA and the mechanism in those patients remains unclear.
Evidence that argues against this (1)
- The hypereosinophilic-syndrome extension is a weaker precedent than its inclusion implies, because it worked by finding a NEW kinase fusion (FIP1L1-PDGFRA) that imatinib happened to inhibit. That is one drug matching a second target, not one target serving two populations, which is the analogy this entry is drawing for menin. New England Journal of Medicine 2003
- E4strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.paper
Long-term follow-up of Children's Oncology Group study AALL0031 showed durable benefit from adding imatinib to intensive chemotherapy in children with Philadelphia chromosome-positive acute lymphoblastic leukemia.
Long-term follow-up of imatinib in pediatric Philadelphia chromosome-positive acute lymphoblastic leukemia: Children's Oncology Group study AALL0031 · Schultz KR, Carroll A, Heerema NA, et al. · Leukemia 2014What this evidence cannot show
Non-randomised cohort comparison of sequential imatinib-intensity cohorts against historical controls, so the effect size carries the usual historical-control caveats; the population is Ph+ ALL in children generally, not infants.
Evidence that argues against this (1)
- AALL0031 is the strongest precedent in the entry and it argues for a design this project's own field is not using: benefit came from adding imatinib to INTENSIVE CHEMOTHERAPY in NEWLY DIAGNOSED disease, not from monotherapy at relapse. Cited as a precedent for multi-indication reach, it is better read as a precedent for line of therapy and combination. Leukemia 2014
Durability correction (EURO-SKI, Saussele 2018, PMID 29735299, NCT01596114): among 755 chronic-phase patients pre-selected as the best molecular responders, molecular relapse-free survival after stopping was 61% at 6 months and 50% at 24 months. Treatment-free remission is a minority outcome within an already-selected minority.
Era correction (Arico 2010, PMID 20876426): without any TKI, 610 children treated 1995-2005 reached 7-year EFS 32.0% and OS 44.9%, up from 25.0% and 36.0% in 1986-1996, so the pre-TKI 20-30% baseline is a moving number.
What is actually modelled (PDB 1IEP): a 293-residue single chain at 2.1 A mapping by SIFTS to P00520 ABL1_MOUSE from residue 229 - no BCR moiety, no SH3-SH2 cap, no myristoyl pocket, and not human.
What the patients are running on (Wang 2005, PMID 15650049): KIT mutations in 26/54 t(8;21) cases as a second hit; at remission AML1-ETO persists and the KIT mutation is gone, so the acquired kinase lesion, not the RUNX1 axis, is what therapy clears.
The crystallographic fit is not the explanation (Agafonov 2014, PMID 25218445): Abl and Src bind imatinib in near-identical conformations with vastly different affinities; selectivity comes from induced fit after binding. Also PDB 1IEP is mouse Abl1 (P00520), 293 residues, no BCR.
the anchor indication of the original platform asset
the economic reading of the same clinical story
large anchor indication cross-subsidising rare ones, with a decade paediatric lag
the original one-asset-many-indications precedent