Oncofusion

start / graph / drug-mitoxantrone

Drugstrong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.Q51-backbone

Mitoxantrone - the Interfant-06 protocol says it is less cardiotoxic than daunorubicin, and the largest survivor cohort ever assembled says it is about seventeen times worse per milligram

also called mitoxantrone · MTZ · MX · mitozantrone · mitoxantrone hydrochloride · Novantrone · anthraquinone · dihydroxyanthracenedione

Mitoxantrone is a close cousin of the red anthracyclines, used in the experimental arm of the infant protocol and in relapse. The protocol tells doctors it is gentler on the heart than daunorubicin; the biggest long-term survivor study ever done found the opposite, by a very large margin.

The same thing, for a scientist

Mitoxantrone is an anthraquinone - a DNA intercalator and topoisomerase II poison lacking the anthracycline sugar - given in Interfant-06 as **12 mg/m2/dose on days 1, 3 and 5 of MAE** (36 mg/m2 total), the second block of the experimental myeloid-style consolidation arm [E1]. It is also the agent behind the largest single randomised improvement in relapsed childhood ALL: in ALL R3, substituting mitoxantrone for idarubicin in reinduction raised 3-year progression-free survival from 35.9% to **64.6%** (P = 0.0004) and overall survival from 45.2% to 69.0% (P = 0.004), and the randomisation was stopped early [E2]. It is in current infant induction regimens as well: the recent multicentre phase 1/2 infant study used mitoxantrone 8 mg/m2 on days 8-9 of induction in place of an anthracycline [E3]. **And the protocol's cardiac guidance is wrong.** Interfant-06 chapter 7 states 'Mitoxantrone is less cardiotoxic than daunorubicin' [E1]; the pooled CCSS/DCOG-LATER/St Jude analysis of 28,423 five-year survivors gives late-cardiomyopathy equivalence ratios to doxorubicin of **0.6 for daunorubicin and 10.5 for mitoxantrone** [E4] - a per-milligram ratio of roughly 17.5 in the opposite direction, with the authors concluding that the conventional 4:1 haematologic equivalence 'significantly underestimates' mitoxantrone's long-term cardiomyopathy association.

Go deeper
The contradiction, stated exactly, because a clinician could act on it

What the protocol says, chapter 7, under Mitoxantrone 1:

If functional cardiac changes occur during treatment, monitor ECG or echocardiography more frequently but do not reduce the dose. Mitoxantrone is less cardiotoxic than daunorubicin and cumulative anthracycline dose should not exceed safe limits.

What the evidence says 4: equivalence ratios to doxorubicin for late-onset cardiomyopathy by age 40, from 28,423 five-year survivors with 399 cardiomyopathy events over a median 20 years - daunorubicin 0.6 (95% CI 0.4-1.0), mitoxantrone 10.5 (95% CI 6.2-19.1). The authors state that the widely used haematologic-toxicity equivalence of 4:1 underestimates mitoxantrone's cardiomyopathy association.

Applied to this protocol. The experimental arm's 36 mg/m2 of mitoxantrone is roughly 378 mg/m2 doxorubicin-equivalent; the standard arm's 180 mg/m2 of daunorubicin is roughly 108 mg/m2 doxorubicin-equivalent. risk-anthracycline-cardiotoxicity-infant-leukemia computes the whole-arm totals and gets about 504 against about 108 - a four- to five-fold difference in predicted lifetime cardiomyopathy exposure between two randomised arms whose six-year disease-free survival was indistinguishable (39.3% vs 36.8%, P = .47) 5.

The resolution is chronology, not error. The protocol is the SKION version of April 2006; Feijen et al. published in 2019. The 2006 statement reflected the haematologic-equivalence assumption that was standard then. It is nonetheless a statement in a live protocol document that is wrong in direction, and a reader of that document today would be misled.

Where it sits, and where else it earns its place

Interfant-06: MAE block, 12 mg/m2/dose IV over 1 hour on days 1, 3 and 5, with cytarabine 100 mg/m2 q12h days 1-10 and etoposide 100 mg/m2 days 1-5 1. Experimental arm only; the block exists because the randomisation asked whether AML-style consolidation helps, and it did not 5.

Relapse. ALL R3 is the strongest randomised result for any conventional agent in this graph's paediatric-ALL evidence: 216 children randomised, idarubicin versus mitoxantrone in reinduction, stopped early in December 2007 for a difference in progression-free and overall survival, with the benefit arising mainly from fewer disease events (HR 0.56, 0.34-0.92) rather than fewer toxic deaths 2. Patients were aged 1-18, so no infant is in it.

Current infant induction. The 2026 phase 1/2 infant study used mitoxantrone 8 mg/m2 on days 8 and 9 of induction and days 1 and 2 of reinduction, with dexamethasone and pegaspargase 3 - i.e. mitoxantrone has moved from the experimental consolidation arm into front-line infant induction in at least one contemporary design.

Facts and figures
targetgene-top2a
target noteSchema requires a single target id. Mitoxantrone is an anthraquinone DNA intercalator and topoisomerase II poison and does not distinguish TOP2A from TOP2B. It is NOT an anthracycline - it lacks the sugar - and grouping it with anthracyclines is a convention of dose-equivalence tables, not a chemical identity.
modalitysmall molecule, anthraquinone (dihydroxyanthracenedione); DNA intercalator and topoisomerase II poison
interfant06 positionMAE block, experimental arm only: 12 mg/m2/dose IV over 1 hour on days 1, 3 and 5 (36 mg/m2 total)
interfant06 dose12 mg/m2/dose
infant dose adjustmentx 2/3 under 6 months, x 3/4 from 6 through 12 months
late cardiomyopathy equivalence ratio to doxorubicin10.5
late cardiomyopathy equivalence ci6.2 to 19.1
doxorubicin equivalent of the mae block mg per m2378
protocol statement contradictedInterfant-06 SKION v01 (April 2006) chapter 7 states 'Mitoxantrone is less cardiotoxic than daunorubicin' and directs clinicians not to reduce the dose for functional cardiac changes. PMID 30703192 (2019) gives equivalence ratios of 10.5 for mitoxantrone and 0.6 for daunorubicin, i.e. roughly 17.5-fold the opposite way, with the whole 95% CI excluding the protocol's direction. The protocol predates the analysis by thirteen years.
kmt2a specific evidenceNONE AGENT-SPECIFIC. Mitoxantrone is in the protocol and in a current infant induction regimen; no measurement of mitoxantrone sensitivity in KMT2A-rearranged or infant blasts was found. The ALL R3 randomised superiority is in children aged 1-18 with no infants enrolled.
key ae
[
  "myelosuppression",
  "mucositis",
  "blue-green discoloration of urine and sclerae",
  "late dilated cardiomyopathy at an equivalence ratio of 10.5 to doxorubicin",
  "secondary acute myeloid leukaemia (topoisomerase II poison class effect)"
]
resistance
[
  "no infant- or KMT2A-specific mechanism identified in this beat",
  "P-glycoprotein efflux",
  "reduced topoisomerase II expression"
]
trials
[
  "NCT00550992",
  "NCT02553460",
  "ISRCTN45724312"
]
How this could bite us back
  • **A live protocol document tells the treating clinician that this drug is less cardiotoxic than daunorubicin and not to reduce the dose for cardiac changes, and the best available evidence puts it about seventeen times worse per milligram.** That is not a theoretical error; it is guidance at the bedside of an infant.
  • The experimental arm of Interfant-06 delivered roughly four to five times the doxorubicin-equivalent cardiotoxicity of the standard arm for no disease-free survival benefit. Half of the medium- and high-risk infants randomised in that trial carry that exposure for life.
  • Mitoxantrone is a topoisomerase II poison, so the same second-leukaemia mechanism that makes etoposide and the anthracyclines aetiological agents in this disease applies to it - and it is being given to infants whose leukaemia was caused by a KMT2A rearrangement in the first place.
  • The strongest randomised result for mitoxantrone in paediatric ALL comes from a trial that stopped early for benefit in children aged 1-18. Stopping early inflates effect sizes, and the enrolled population excludes exactly the age group this graph is about.
  • Moving mitoxantrone from the experimental consolidation arm into front-line infant induction, as a current study does, front-loads the agent with the worst cardiomyopathy equivalence into the youngest patients at the point where dose reduction by the 2/3 rule is already being applied for reasons nobody has justified.
Still open
  1. Has the Interfant-06 cardiac guidance been revised in Interfant-21 or in later versions of the Interfant-06 document? This beat read the SKION April 2006 version only, and the sentence it contains is one a clinician acts on.
  2. Does the ALL R3 mitoxantrone benefit hold in infants? The trial excluded them by design, and infants are the group whose relapse outcomes are worst and whose remaining heart-years are most.
  3. If the MRD-by-consolidation interaction in Interfant-06 is real, is the mitoxantrone-containing arm worth roughly 378 mg/m2 of doxorubicin-equivalent cardiotoxicity in the high-MRD subgroup? That is a quantified trade-off nobody has published.
  4. Would dexrazoxane protect against mitoxantrone specifically? The randomised paediatric cardioprotection evidence is for doxorubicin, and the agent with the worst equivalence ratio has never been paired with it in a trial.
Every source behind this entry (5)
  1. E1established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.paper

    Interfant-06 gives mitoxantrone 12 mg/m2/dose IV over 1 hour on days 1, 3 and 5 of the MAE block (36 mg/m2 total) in the experimental arm, and its chapter 7 toxicity guidance states that if functional cardiac changes occur during treatment the clinician should monitor more frequently but not reduce the dose, because 'Mitoxantrone is less cardiotoxic than daunorubicin'.

    What this evidence cannot show

    A trial protocol document, not a peer-reviewed publication, and specifically the SKION (Dutch) national version of April 2006; the trial was amended during its course (asparaginase provisions in MARMA and OCTADA(D) among others), so later versions may differ. Doses given are PLANNED doses; delivered cumulative dose per patient is lower wherever a course was omitted, delayed or curtailed by a stopping rule, and the trial report does not publish delivered dose per patient. Retrieved and text-extracted by this beat on 2026-09-20 (90 pages, 194,979 characters); page numbers are the PDF's own. The quoted sentence is verbatim from the Mitoxantrone paragraph of chapter 7 of the SKION April 2006 version, read from the extracted text of the PDF on 2026-09-20. Later protocol versions and later Interfant protocols may have revised it; this beat checked only this version.

    Evidence that argues against this (1)
    • The claim in that sentence is contradicted in direction by the largest analysis of late cardiomyopathy in childhood cancer survivors: equivalence ratios to doxorubicin of 0.6 for daunorubicin and 10.5 for mitoxantrone, with the authors stating that the conventional 4:1 haematologic equivalence significantly underestimates mitoxantrone's long-term cardiomyopathy association. The protocol predates that analysis by thirteen years. Feijen EAM et al. 2019
  2. E2strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.paper

    In ALL R3, an open-label randomised trial in 22 UK and Irish and nine Australian and New Zealand centres, 216 children aged 1-18 with first relapse of ALL were randomised to idarubicin or mitoxantrone in reinduction; estimated 3-year progression-free survival was 35.9% (95% CI 25.9-45.9) with idarubicin versus 64.6% (54.2-73.2) with mitoxantrone (P = 0.0004) and 3-year overall survival 45.2% versus 69.0% (P = 0.004); randomisation was stopped in December 2007, and the difference arose mainly from fewer disease events (HR 0.56, 0.34-0.92, P = 0.007) rather than fewer adverse treatment effects (HR 0.52, 0.24-1.11, P = 0.11).

    PMID 21131038DOI
    What this evidence cannot show

    **Patients were aged 1-18 years, so this trial contains no infants** and its result cannot be read directly into infant ALL. Randomisation was stopped early for benefit, which inflates effect-size estimates. The comparator is idarubicin, not daunorubicin and not no anthracycline, so the trial establishes a relative ranking within the class and not an absolute case for mitoxantrone. Relapsed disease, with subsequent transplant allocated by MRD, so the reinduction agent is not the only difference between the arms' downstream therapy.

    Evidence that argues against this (1)
    • Long-term follow-up of the same trial in patients with late bone-marrow relapses was published subsequently, and any durable assessment of the mitoxantrone benefit has to be set against the cardiomyopathy equivalence ratio of 10.5, which acts over decades and which the trial's 3-year endpoints cannot see. Parker C et al. 2019
  3. E3strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.paper

    A multicentre single-arm phase 1/2 study at 18 hospitals in the USA and Canada enrolled 50 infants aged 1 year or younger with newly diagnosed ALL (35 with KMT2A rearrangement) and used a chemotherapy backbone containing mitoxantrone 8 mg/m2 IV once daily on days 8 and 9 of induction and days 1 and 2 of reinduction, with dexamethasone 5 mg/m2 twice daily and pegaspargase 2500 units/m2, plus bortezomib and vorinostat.

    PMID 41692013DOI
    What this evidence cannot show

    Single-arm, n = 50, and the primary endpoint is tolerability rather than efficacy, so it establishes that this backbone can be delivered and not that it is better. Median age 171 days. No cardiac late-effects data at a median follow-up of 4.9 years. The mitoxantrone dose and schedule are part of an investigational combination and cannot be separated from bortezomib and vorinostat.

    Evidence that argues against this (1)
    • Toxicity in this study was substantial and infectious: grade 3-4 infection in 44% of induction courses, febrile neutropenia in 42%, three episodes of bacterial sepsis plus one induction death as serious adverse events at the expansion dose level, and four treatment-related deaths all from infection. The authors state that adverse events were consistent with previous infant ALL studies, which is the point - this is the toxicity ceiling any intensification in infants runs into. Gruber TA et al. 2026
  4. E4strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.paper

    In a pooled cohort of 28,423 five-year survivors of childhood cancer (Childhood Cancer Survivor Study, DCOG-LATER, St Jude Lifetime) with 399 cardiomyopathy cases over a median 20 years of follow-up, the equivalence ratio to doxorubicin for severe, life-threatening or fatal cardiomyopathy by age 40 was 10.5 (95% CI 6.2-19.1) for mitoxantrone against 0.6 (95% CI 0.4-1.0) for daunorubicin, and the authors conclude that the current haematologic-based 4:1 doxorubicin equivalency of mitoxantrone significantly underestimates its long-term cardiomyopathy association.

    Derivation of Anthracycline and Anthraquinone Equivalence Ratios to Doxorubicin for Late-Onset Cardiotoxicity. · Feijen EAM, Leisenring WM, Stratton KL, Ness KK, van der Pal HJH, van Dalen EC, et al. · JAMA Oncol 2019
    PMID 30703192DOI
    What this evidence cannot show

    Only 265 of the 28,423 survivors received mitoxantrone, which is why the confidence interval spans 6.2 to 19.1. Retrospective cohorts treated 1962-2005, largely without dexrazoxane and with historical supportive care. The relationship for mitoxantrone is better described by a linear exponential than a linear model, so a single ratio is a simplification. The ratios are not age-stratified and most survivors were not infants.

    Evidence that argues against this (1)
    • A wide confidence interval on a small exposed subgroup is a reason to be cautious about the magnitude, not about the direction: the interval's lower bound of 6.2 is still ten times daunorubicin's point estimate of 0.6, so the protocol's claim that mitoxantrone is LESS cardiotoxic than daunorubicin is excluded across the whole interval. Feijen EAM et al. 2019
  5. E5established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.paper

    The Interfant-06 randomisation between myeloid-style consolidation (ADE then MAE, containing the mitoxantrone block) and lymphoid-style protocol IB was negative: 6-year disease-free survival 39.3% versus 36.8%, log-rank P = .47, with whole-cohort 6-year EFS 46.1% and OS 58.2%.

    PMID 31283407DOI
    What this evidence cannot show

    Read from the abstract only; the full text is not open access and returned no PMCID. A negative randomisation with 169 and 161 patients per arm has limited power to exclude a modest benefit, and the comparison is of whole multi-drug blocks, so it says nothing about mitoxantrone individually. Post-hoc MRD-stratified analysis in a later paper reported an interaction favouring myeloid-style consolidation in high-MRD patients.

    Evidence that argues against this (1)
    • The later MRD analysis of the same trial found that high end-of-induction MRD patients did better with myeloid-style consolidation and MRD-negative patients better with lymphoid-style, which converts a flat negative into a stratification hypothesis. If that interaction is real, the arm carrying roughly 378 mg/m2 of doxorubicin-equivalent mitoxantrone is the right arm for a subgroup, and the cardiotoxicity trade-off becomes a real clinical decision rather than a reason to delete the block. Tomizawa D et al. 2020
Connected to