Oncofusion

start / graph / target-mcl1-with-bcl2-and-dexamethasone-infant-kmt2a-all

Targetmoderate confidence 3 of 5: One strong study, or several smaller consistent ones.Q7-clinical-infant

MCL1 plus BCL-2 plus dexamethasone: a triple combination that matched induction chemotherapy in infant PDX models

also called S63845 infant ALL · MCL1 inhibitor KMT2A-r · BCL-2 MCL1 co-inhibition · venetoclax dexamethasone S63845

Blocking two survival proteins at once - BCL-2 with venetoclax and MCL1 with S63845 - alongside the steroid dexamethasone put nearly every mouse carrying an infant leukaemia graft into lasting remission, as well as three-drug chemotherapy did. On its own the MCL1 blocker did almost nothing in mice, and combined with venetoclax alone it killed two thirds of them from tumour lysis.

The same thing, for a scientist

This entry records a paper the graph previously held as unreadable and which this beat retrieved in full from Europe PMC. In KMT2A-rearranged leukaemia cell lines the MCL1 inhibitor S63845 gave IC50 values of 9-218 nM against 232 nM to >1.5 microM in KMT2A-wild-type lines, and the four most sensitive lines (IC50 <=50 nM: PER-485, PER-703, PER-494, KOPN-8) were **all derived from infants with KMT2A-rearranged ALL** [E1]. Sensitivity did not correlate with BCL-2 or MCL1 levels; BCL-XL mRNA correlated positively with IC50 and KMT2A-r cells had significantly lower BCL-XL protein, suggesting the selectivity reflects reduced BCL-XL dependence rather than MCL1 overexpression [E1]. S63845 plus venetoclax was strongly synergistic in vitro, reducing the venetoclax IC50 by up to 10,000-fold, and adding dexamethasone produced robust triple synergy [E1]. In vivo the picture inverts: S63845 monotherapy gave no significant survival benefit at 6.25, 25 or 50 mg/kg; S63845 plus venetoclax did not beat venetoclax alone and at >=12.5 mg/kg with venetoclax induced **lethal tumour lysis syndrome in 8 of 12 mice (67%)**; but the triple combination nearly doubled survival against dexamethasone alone (MLL-7: 68 vs 37 days; MLL-14: 61 vs 32 days), achieved maintained complete response in 7/7 and 7/8 mice, and matched a vincristine/dexamethasone/asparaginase induction regimen [E1][E2].

Go deeper
What was actually measured1 deeper
ObservationValueSystem
S63845 IC50, KMT2A-r lines9-218 nMcell lines
S63845 IC50, KMT2A-wt lines232 nM - >1.5 microMcell lines, p < 0.0001
Most sensitive lines (IC50 <=50 nM)PER-485 (9 nM), PER-703 (10 nM), PER-494, KOPN-8all infant KMT2A-r ALL derived
S63845 IC50, infant KMT2A-r PDX ex vivo10-228 nMshort-term PDX culture
Venetoclax IC50 shift with sublethal S63845up to 10,000-fold reductionPER-485, THP-1, PER-703
TLS with S63845 >=12.5 mg/kg + venetoclax 25 mg/kg8 of 12 mice diedNSG/NOD-SCID
Triple combination vs dexamethasone alone68 vs 37 days (MLL-7); 61 vs 32 days (MLL-14)PDX survival
Maintained complete response, triple7/7 MLL-7; 7/8 MLL-14vs 5/8 and 3/8 on venetoclax+dexamethasone
Triple vs VXL induction chemotherapyon parPDX

The in vitro-to-in vivo discordance is the most useful part of the paper and the authors pursue it: after a single 25 mg/kg dose of S63845, pro-apoptotic BIM protein fell (2 of 3 mice), and they propose reduced BIM as the reason for the loss of in vivo activity, with dexamethasone restoring pro-apoptotic priming 1.

Why this bears on the menin-interface programme

Two ways. First, it is an infant KMT2A-rearranged dependency that is not the fusion complex: MCL1 is a direct target gene of KMT2A fusion proteins, so the dependency is downstream of the fusion rather than at the interface, and an agent acting there is available whether or not a menin inhibitor works. Second, the mechanism of the triple's advantage is apoptotic priming, which is exactly the axis a menin inhibitor is expected to move by releasing differentiation. If both act by raising priming, the combination may be less than additive; if the menin inhibitor raises BIM where S63845 lowers it, it may be more. Neither has been tested.

What the paper does not show
  • No infant patient received this. Everything is cell line, ex vivo PDX culture, or mouse.
  • Two PDX models. MLL-7 and MLL-14 carry the entire in vivo result.
  • S63845 is not a clinical agent. Clinical MCL1 inhibitors that have entered trials (AZD5991, ABBV-467) carry reported tumour lysis in adults and, for ABBV-467, cardiac troponin rises 3.
  • The haematopoietic-safety claim rests on recovery of white cells, red cells and platelets in 4 mice per group in immunodeficient animals, which does not model an infant's marrow.
  • The authors' own closing position is that KMT2A-rearranged leukaemia is highly heterogeneous and that stringent dose and patient selection, sequential scheduling and proactive TLS monitoring would be essential 1.
Facts and figures
modalityBH3 mimetic combination (MCL1 + BCL-2) with glucocorticoid
tool compoundS63845 (MCL1); venetoclax (BCL-2); dexamethasone
stagepreclinical only; no infant has received this combination
source pmid42246204
retrieved fromEurope PMC full text PMC13461956, 2026-09-19
targets
[
  "MCL1",
  "BCL2"
]
in vitro
{
  "kmt2a_r_ic50_nM": [
    9,
    218
  ],
  "kmt2a_wt_ic50_nM": [
    232,
    1500
  ],
  "most_sensitive_lines": [
    "PER-485",
    "PER-703",
    "PER-494",
    "KOPN-8"
  ],
  "pdx_ex_vivo_ic50_nM": [
    10,
    228
  ]
}
in vivo
{
  "models": [
    "MLL-7",
    "MLL-14"
  ],
  "monotherapy_effect": "none significant",
  "venetoclax_plus_s63845_tls_deaths": "8 of 12 mice at >=12.5 mg/kg",
  "triple_survival_days": {
    "MLL-7": [
      68,
      37
    ],
    "MLL-14": [
      61,
      32
    ]
  },
  "triple_mcr": {
    "MLL-7": "7/7",
    "MLL-14": "7/8"
  },
  "comparator": "on par with VXL induction chemotherapy"
}
How this could bite us back
  • S63845 monotherapy produced no objective response in any mouse; the whole effect requires dexamethasone, so the agent is a glucocorticoid potentiator and inherits every mechanism of glucocorticoid resistance.
  • Combining two BH3 mimetics caused lethal tumour lysis syndrome in 67% of mice at the higher dose, in a disease whose patients already present with enormous circulating tumour burden and the highest tumour-lysis risk in paediatric oncology.
  • MCL1 is required by cardiomyocytes and by haematopoietic stem cells; the safety data here are 4 mice per group in immunodeficient animals over weeks, which cannot speak to an infant's marrow reserve or to a heart that also faces anthracycline exposure.
  • Reduced BIM after MCL1 inhibition is an on-target adaptive response that blunts the drug in vivo within 16 hours, so resistance to this agent begins immediately and is not a late clonal event.
  • Two PDX models carry the entire in vivo result, and KMT2A-rearranged infant ALL is explicitly heterogeneous, so the response rate in an unselected infant population is unknown.
  • If venetoclax enters the frontline infant backbone via NCT06317662, any later MCL1 inhibitor arrives in patients already BCL-2-inhibited, which is precisely the pairing that was lethal in mice without a glucocorticoid.
Still open
  1. Does a menin inhibitor raise or lower BIM in infant KMT2A-rearranged cells? If it raises it, the menin inhibitor could substitute for dexamethasone in this triple; if it lowers it, the combination is antagonistic.
  2. Is the lower BCL-XL of KMT2A-rearranged cells a fusion-driven property, and does it survive menin inhibition? That would determine whether BH3-mimetic sensitivity is preserved or lost at relapse on a menin inhibitor.
  3. Can the tumour lysis seen with S63845 plus venetoclax be avoided by sequencing rather than by adding dexamethasone, and would sequencing preserve the survival benefit?
  4. Does the triple combination work in a KMT2A-rearranged PDX that has acquired menin-inhibitor resistance?
Every source behind this entry (4)
  1. E1strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.paper

    In preclinical models of infant KMT2A-rearranged ALL the MCL1 inhibitor S63845 was selectively active in vitro (IC50 9-218 nM in KMT2A-r vs 232 nM to >1.5 microM in KMT2A-wt lines; the four lines with IC50 <=50 nM were all infant KMT2A-r ALL derived), synergised strongly with venetoclax and dexamethasone, but as monotherapy or with venetoclax alone failed in vivo, with lethal tumour lysis syndrome in 8 of 12 mice at S63845 >=12.5 mg/kg plus venetoclax; the triple combination of S63845, venetoclax and dexamethasone nearly doubled survival against dexamethasone alone and matched VXL induction chemotherapy.

    PMID 42246204DOI
    What this evidence cannot show

    All preclinical. Two PDX models (MLL-7, MLL-14) carry the in vivo result; MLL-7 is venetoclax-responsive and MLL-14 venetoclax-resistant, and the triple's added benefit over venetoclax+dexamethasone was significant only in MLL-14 (p = 0.0227). S63845 is a tool compound, not a clinical agent. Immunodeficient NSG/NOD-SCID mice cannot model immune contributions or an infant's marrow reserve. BIM pharmacodynamics rest on 2 of 3 mice showing the trend. The paper carried no abstract in PubMed or Europe PMC; the full text was retrieved from Europe PMC (PMC13461956) and read in full by this beat on 2026-09-19.

    Evidence that argues against this (1)
    • In the same paper S63845 monotherapy produced no objective responses in any mouse at any dose tested, and adding it to venetoclax did not significantly extend survival over venetoclax alone in either PDX; the entire benefit requires dexamethasone to be present, which means the agent's value is as a glucocorticoid potentiator rather than as an MCL1-directed therapy. Karsa M et al. 2026
  2. E2moderate confidence 3 of 5: One strong study, or several smaller consistent ones.paper

    MCL1 is a direct target gene of KMT2A fusion proteins and is expressed at higher levels in infants than in older children with ALL, with elevated MCL1 correlating with prednisolone resistance, which is the stated rationale for targeting it in infant KMT2A-rearranged ALL.

    PMID 42246204DOI
    What this evidence cannot show

    This is the introduction's framing of prior work (its references 3-5), not a result of this paper; the underlying primary reports were not independently retrieved by this beat and should be fetched before being cited as primary evidence anywhere in this graph.

    Evidence that argues against this (1)
    • This paper's own data contradict the simple version of that rationale: S63845 sensitivity did not correlate with MCL1 protein or mRNA levels, and the selectivity tracked lower BCL-XL rather than higher MCL1. Karsa M et al. 2026
  3. E3weak confidence 2 of 5: Preliminary, small, indirect, or not yet peer reviewed.paper

    Clinical MCL1 inhibitors have entered early-phase haematological trials with tumour lysis syndrome reported in adult patients, and this beat's source paper cites cardiac troponin increases for one of them.

    PMID 42246204DOI
    What this evidence cannot show

    Reported second-hand from the discussion of the source paper (its references 12 and 13, AZD5991 and ABBV-467); the primary trial reports were not fetched by this beat, so this claim is carried at weak confidence and should not be cited elsewhere without retrieving them.

    Evidence that argues against this (1)
    • Class toxicity observed in heavily pre-treated adults does not necessarily transfer to infants, whose cardiac and renal physiology differ; nor does its absence transfer. Nobody has dosed an infant with an MCL1 inhibitor. Stutterheim J et al. 2025
  4. E4strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.review

    A phase II randomised COG trial (NCT06317662) will test replacing cytarabine with the BCL-2 inhibitor venetoclax in infant ALL, and a menin inhibitor trial of revumenib (NCT05761171) is planned.

    KMT2A-rearranged acute lymphoblastic leukemia in infants: current progress and challenges. · Stutterheim J, Van der Sluis IM, Vrenken KS, Pieters R, Stam RW · Haematologica 2025
    PMID 40207717DOI
    What this evidence cannot show

    Reported in a 2025 review; registry status should be checked directly against ClinicalTrials.gov before any claim about enrolment, arms or current status, and this beat did not do so. Read in the open-access full text at PMC12399962.

    Evidence that argues against this (1)
    • If venetoclax enters the infant backbone, the S63845 result becomes a question about adding a third BH3 mimetic to a regimen already containing one, in a setting where that pair caused lethal tumour lysis in mice without a glucocorticoid present. Karsa M et al. 2026
Connected to