Oncofusion

start / graph / drug-prednisolone

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

Prednisolone - the active steroid, the drug every resistance experiment in this disease actually used, and the one given into the spinal fluid

also called prednisolone · PRDL · PRED (in vitro) · prednisolone sodium succinate · prednisolone disodium phosphate · Di-Adreson-F · delta-1-hydrocortisone · metacortandralone

Prednisolone is the form of the steroid that actually works - prednisone has to be turned into it by the liver first. Almost everything we know about why infant leukaemia cells survive steroids was measured on prednisolone in a dish, and in the Interfant protocol prednisolone is also the steroid injected into the spinal fluid.

The same thing, for a scientist

Prednisolone is the 11-hydroxy, receptor-active glucocorticoid; prednisone is its 11-keto prodrug [see `drug-prednisone`]. In Interfant-06 prednisolone is listed as a drug intervention distinct from prednisone and occupies a different route: it (or hydrocortisone) is the **intrathecal** steroid given with every IT dose from induction day 1 through maintenance week 15, while prednisone is the systemic prophase agent [E1]. **Its scientific importance is larger than its protocol footprint.** Infant ALL blasts are >580-fold resistant to prednisolone in the MTT assay against 425 older children with common/preB ALL, with L-asparaginase 12-fold resistant alongside, and t(4;11) cases a further >7.4-fold more prednisolone-resistant than other 11q23 rearrangements [E2]. Four mechanisms have been reported in MLL-rearranged infant ALL specifically, all measured with prednisolone: high MCL-1 [E3], S100A8/S100A9 suppression of glucocorticoid-evoked cytosolic Ca2+ [E4], Src(FYN/LCK)-driven phosphorylation of annexin A2 requiring S100A10/p11 [E5], and a PI3K/FCGR1B axis found by Connectivity Map screening [E6]. Each is separately reversible in vitro and none is sufficient, which is the honest summary of the field.

Go deeper
Why this is a separate entry from prednisone, and why that is not pedantry

They are different molecules. Prednisone carries an 11-keto group and has negligible affinity for the glucocorticoid receptor; prednisolone carries the 11-hydroxy group and is the ligand. Hepatic 11beta-HSD1 reduces one to the other, and 11beta-HSD2 reverses it [see drug-prednisone E3].

The consequences for reading this literature:

  • In cell culture there is no liver. Every in-vitro glucocorticoid-resistance study in MLL-rearranged infant ALL therefore used prednisolone, and several of them say 'prednisone' in the title or the framing sentence while reporting prednisolone in the methods 356. The papers are not wrong; a reader who copies the framing word into a claim is.
  • In Interfant-06 they are different drugs on different routes 1. Writing 'the protocol gives prednisone/prednisolone' erases the fact that one is systemic for seven days and the other is intrathecal for two years.
  • The prognostic factor and the resistance mechanism are therefore measured on different molecules in different compartments. drug-prednisone carries the prognostic half; this entry carries the mechanism half.

This is a rule-69-class trap in chemistry rather than in gene symbols: two strings, two real referents, a curated relationship between them, and a literature that uses them interchangeably when the systems it works in cannot.

The four reported resistance mechanisms, and what each actually shows

All four come from the same disease - MLL-rearranged infant ALL - which is unusually good provenance for a mechanism claim in this graph.

mechanismwhat was shownwhat reverses it in vitro
MCL-1 high 3expression associated with in-vitro prednisolone resistance and with in-vivo prednisone poor responseRNAi knockdown, partially
S100A8/S100A9 high 4enforced expression causes prednisolone resistance by suppressing glucocorticoid-evoked free cytosolic Ca2+Src inhibitor PP2, via S100A8/A9 downregulation
Src(FYN/LCK) -> annexin A2 phosphorylation, p11/S100A10-dependent 5shRNA against ANXA2, FYN, LCK or S100A10 each sensitised markedlyFYN/LCK-selective Src inhibitors
PI3K / FCGR1B 6in-silico Connectivity Map screen predicted LY294002; validated, and FCGR1B knockdown alone sensitisedLY294002 and other PI3K inhibitors

Read the pattern, not the list. Three of the four converge on Src-family kinase signalling (PP2 in 4, FYN/LCK in 5, and the SFK/FLT3 inhibitor RK-20449 in an independent group's PDX work 7), and two converge on BCL-2-family dependence (MCL-1 in 3; BH3 profiling showing BCL-2 dependence, with venetoclax completing the kill, in 7). The most credible synthesis available is that steroid resistance in this disease is an apoptotic-threshold problem held up by kinase signalling, and that is what target-mcl1-with-bcl2-and-dexamethasone-infant-kmt2a-all is about.

What no one has shown: that any of these is downstream of the fusion. Every mechanism here is an association or a manipulation in MLL-rearranged cells; none is a demonstration that KMT2A::AFF1 or KMT2A::MLLT1 installs it. The proB immunophenotype confounds all of it 2.

Infant-specific pharmacology and the intrathecal route

Prednisolone's systemic kinetics are nonlinear because it binds transcortin (CBG) saturably, so a dose increase raises free drug more than proportionally - an unusual property for the graph to carry, since almost every other backbone agent scales linearly. In children, free prednisolone exposure after oral prednisone varies enough that the drug was undetectable at 12 h in some patients and still present at 24 h in others [see drug-prednisone E7], and no such study exists in infants.

The intrathecal use is the infant-specific one and it is the least studied. Interfant-06 specifies prednisolone or hydrocortisone intrathecally at every IT timepoint 1 - an either/or that the protocol does not resolve and that is left to the centre. This beat found no comparison of intrathecal prednisolone against intrathecal hydrocortisone on any endpoint, in any age group, and records that as a bounded absence in E8 rather than assuming equivalence.

Facts and figures
targetNR3C1 glucocorticoid receptor (direct ligand). No gene node for NR3C1 exists in this graph as of cycle 28.
prodrugprednisone (see drug-prednisone); prednisolone is the 11-hydroxy active form
modalitysmall molecule, synthetic glucocorticoid, receptor-active form
protein binding notebinds corticosteroid-binding globulin (transcortin) saturably, so systemic pharmacokinetics are nonlinear - unusual among the backbone agents
interfant06 positionINTRATHECAL steroid (prednisolone OR therapeutic hydrocortisone) at every IT timepoint: induction days 1/15/29, protocol IB days 10/24, MARMA part I days 2/9, OCTADA(D) part I days 1/15, maintenance weeks 1/8/15. Not a systemic agent in this protocol.
in vitro resistance infant vs older childrenmore than 580-fold (MTT, 44 infants vs 425 children >1 y with common/preB ALL, P=0.001)
in vitro resistance t411 vs other 11q23a further >7.4-fold (P=0.033)
kmt2a specific evidenceSTRONG AND MECHANISTIC. Four resistance mechanisms reported in MLL-rearranged infant ALL specifically: MCL-1, S100A8/S100A9-Ca2+, Src(FYN/LCK)-annexin A2-p11, PI3K/FCGR1B. Plus independent PDX evidence for SFK/FLT3 plus BCL-2 dependence. This is the best-evidenced conventional drug in the beat.
interfant06 doseintrathecal prednisolone 6 mg if age <1 year, 8 mg if age >=1 year; substituted by hydrocortisone 12 mg or 16 mg respectively if prednisolone is unavailable for intrathecal use
infant dose adjustmentNONE. Intrathecal drugs are explicitly exempt from the 2/3 and 3/4 body-surface-area reductions that apply to every systemic agent in Interfant-06; they are dosed by age band instead.
hydrocortisone substitution ratio2:1 hydrocortisone to prednisolone by mass, asserted by the protocol without a citation
resistance
[
  "high MCL-1 (gene-mcl1)",
  "S100A8/S100A9 over-expression blocking glucocorticoid-evoked cytosolic Ca2+ flux",
  "FYN/LCK-mediated annexin A2 phosphorylation, p11/S100A10-dependent",
  "PI3K signalling; possibly via FCGR1B",
  "BCL-2 dependence revealed by BH3 profiling in PDX after SFK/FLT3 inhibition",
  "NOT classical multidrug resistance: P-glycoprotein and MRP were not higher in infants than in older children with common/preB ALL"
]
reversal agents reported in vitro
[
  "PP2",
  "FYN/LCK-selective Src inhibitors",
  "LY294002 and other PI3K inhibitors",
  "RK-20449 (SFK/FLT3)",
  "ABT-199/venetoclax",
  "gossypol and AT-101 (pan-BCL-2 family)"
]
key ae
[
  "chemical arachnoiditis and transient neurological events (intrathecal route)",
  "hyperglycaemia",
  "hypertension",
  "infection",
  "adrenal suppression"
]
trials
[
  "NCT00550992"
]
How this could bite us back
  • **Every reversal strategy in this entry ends with giving a baby a second targeted drug on top of a steroid she is already failing.** Src inhibitors, PI3K inhibitors and BCL-2 inhibitors each carry their own myelosuppression and infection risk in a population whose commonest cause of treatment-related death is infection.
  • The Src-inhibitor route selects for a BCL-2-dependent state in the same PDX system that demonstrated it works - the resistance to the resistance-breaker was found in the same paper. A two-drug proposal here should be assumed to need a third.
  • MCL-1 is the mechanism this graph already carries and the hardest one to drug: MCL-1 is essential in cardiomyocytes and in haematopoietic stem cells, and clinical MCL-1 inhibitors have been troubled by cardiac toxicity. A resistance mechanism being correct does not make it addressable.
  • Prednisolone is also given intrathecally to infants for two years, into a nervous system that is still myelinating. Nothing in this entry's resistance literature touches that exposure, and the graph should not let the mechanism work stand in for a safety assessment of the route that is actually novel in this age group.
  • If the in-vitro resistance is a property of the proB immunophenotype rather than of the fusion - which is the cited authors' own reading - then resensitisation strategies aimed at the fusion's biology are aimed at the wrong variable, and would be expected to fail in exactly the KMT2A-germline proB cases that share the phenotype.
Still open
  1. Are any of the four reported prednisolone-resistance mechanisms downstream of the KMT2A fusion itself? Every one is an association or a manipulation in MLL-rearranged cells; none has been shown to be installed by the fusion, and the proB immunophenotype confounds all of them.
  2. Four mechanisms, largely one laboratory, each partially reversible, none sufficient. Which of them survives replication by an independent group in primary KMT2A-rearranged infant samples? That single experiment would reorder this entry.
  3. Does intrathecal prednisolone differ from intrathecal hydrocortisone in CNS relapse or in neurotoxicity? Interfant-06 permits either and nobody has compared them.
  4. Is the >580-fold in-vitro prednisolone resistance a property of the bulk population or of a subclone? An MTT LC50 on a mixed population cannot tell, and the answer decides whether steroid intensification is futile or merely insufficient.
Every source behind this entry (9)
  1. E1established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.trial_registry

    Interfant-06 lists prednisolone as a drug intervention separate from prednisone, and prescribes prednisolone or therapeutic hydrocortisone intrathecally on induction days 1, 15 and 29, on days 10 and 24 of protocol IB, on days 2 and 9 of MARMA part I, on days 1 and 15 of OCTADA(D) part I, and in maintenance weeks 1, 8 and 15.

    NCT00550992
    What this evidence cannot show

    Registry record fetched 2026-09-20 from the ClinicalTrials.gov v2 API, not the protocol document; it gives route, schedule and days and no dose for any agent. The registry's intervention list is demonstrably incomplete (it omits dexamethasone, which the detailed description prescribes), so it is used here only for positive facts.

    Evidence that argues against this (1)
    • The registry's either/or - 'prednisolone or therapeutic hydrocortisone' - means the IT steroid a given infant received is not recoverable from the protocol, only from centre records. Any claim about intrathecal prednisolone exposure in Interfant-06 is therefore a claim about a subset of unknown size. Pieters R et al. 2019
  2. E2strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.paper

    In MTT assays on primary samples, 44 infants with ALL were more than 580-fold resistant to prednisolone (P = 0.001) and 12-fold resistant to L-asparaginase (P = 0.001) compared with 425 children over 1 year with common/preB ALL, while being 1.9-fold MORE sensitive to cytarabine and 1.7-fold more sensitive to 2-chlorodeoxyadenosine; t(4;11)-positive cases were a further >7.4-fold more prednisolone-resistant than other 11q23 abnormalities (P = 0.033).

    In vitro drug-resistance profile in infant acute lymphoblastic leukemia in relation to age, MLL rearrangements and immunophenotype. · Ramakers-van Woerden NL, Beverloo HB, Veerman AJ, Camitta BM, Loonen AH, van Wering ER, et al. · Leukemia 2004
    PMID 14712291DOI
    What this evidence cannot show

    MTT on primary samples, 2004. The >580-fold and >7.4-fold figures are censored: many infant samples did not reach LC50 at the highest concentration tested, so the ratio is a lower bound, not a point estimate. The authors' own conclusion is that the proB immunophenotype carries much of the effect and that 'the role of MLL cannot be excluded' - i.e. this is not cleanly a fusion effect. P-glycoprotein, MRP and LRP expression were not higher in infants, so classical multidrug resistance does not explain it.

    Evidence that argues against this (1)
    • Age within the proB or MLL-rearranged groups did not affect resistance in this same study (infants under 1 year vs children over 1 year with the same immunophenotype were indistinguishable), so the resistance travels with phenotype and rearrangement rather than with infancy. An entry that reads this as 'infants are steroid-resistant' states the weaker and less useful version of the finding. Ramakers-van Woerden NL et al. 2004
  3. E3moderate confidence 3 of 5: One strong study, or several smaller consistent ones.paper

    High-level MCL-1 expression is associated with prednisolone resistance in vitro in MLL-rearranged infant ALL and with in-vivo prednisone poor response, and MCL-1 knockdown by RNA interference led only to partial prednisolone sensitisation.

    Association of high-level MCL-1 expression with in vitro and in vivo prednisone resistance in MLL-rearranged infant acute lymphoblastic leukemia. · Stam RW, Den Boer ML, Schneider P, de Boer J, Hagelstein J, Valsecchi MG, et al. · Blood 2010
    PMID 19965632DOI
    What this evidence cannot show

    Association plus a partial-rescue knockdown, not a sufficiency experiment. The authors state explicitly that the data 'strongly imply that high-level MCL-1 expression is not the sole mechanism'. MCL-1 is also high in more favourable childhood ALL subtypes in the same analysis, so it is not KMT2A-specific.

    Evidence that argues against this (1)
    • An independent group's BH3 profiling of MLL-ALL patient-derived xenografts identified BCL-2 dependence, not MCL-1 dependence, as the mechanism of treatment resistance, and venetoclax (a BCL-2-selective inhibitor that does not inhibit MCL-1) completed the kill when combined with an SFK/FLT3 inhibitor and glucocorticoid. Two different BCL-2-family dependencies are reported in the same disease by different assays. de Groot AP et al. 2021
  4. E4moderate confidence 3 of 5: One strong study, or several smaller consistent ones.paper

    Enforced expression of S100A8/S100A9 in prednisolone-sensitive MLL-rearranged ALL cells rapidly produced prednisolone resistance by suppressing prednisolone-induced free-cytosolic Ca2+, and the Src kinase inhibitor PP2 sensitised otherwise resistant cells by downregulating S100A8 and S100A9.

    Elevated S100A8/S100A9 expression causes glucocorticoid resistance in MLL-rearranged infant acute lymphoblastic leukemia. · Spijkers-Hagelstein JA, Schneider P, Hulleman E, de Boer J, Williams O, Pieters R, et al. · Leukemia 2012
    PMID 22282267DOI
    What this evidence cannot show

    Cell-line and primary-sample work with enforced expression, i.e. a sufficiency experiment in an artificial expression context. PP2 is a broad Src-family tool compound and its sensitising action is attributed to S100A8/A9 downregulation by correlation, not by a rescue. No in-vivo arm.

    Evidence that argues against this (1)
    • The same group subsequently reported a different Src-dependent route to prednisolone resistance in the same disease - FYN/LCK phosphorylation of annexin A2 with p11/S100A10 as the adapter - in which the S100 protein involved is S100A10 rather than S100A8/A9. Both cannot be the mechanism, and the shared reliance on Src-family tool compounds means the two experiments are not independent. Spijkers-Hagelstein JA et al. 2013
  5. E5moderate confidence 3 of 5: One strong study, or several smaller consistent ones.paper

    In MLL-rearranged infant ALL, high ANXA2 expression discriminates prednisolone-resistant from prednisolone-sensitive patient samples, the operative event is Src-kinase phosphorylation of annexin A2 requiring the adapter p11 (S100A10), and shRNA knockdown of ANXA2, FYN, LCK or S100A10 - or exposure to FYN/LCK-selective Src inhibitors - each markedly sensitised cells to prednisolone.

    Src kinase-induced phosphorylation of annexin A2 mediates glucocorticoid resistance in MLL-rearranged infant acute lymphoblastic leukemia. · Spijkers-Hagelstein JA, Mimoso Pinhanços S, Schneider P, Pieters R, Stam RW · Leukemia 2013
    PMID 23334362DOI
    What this evidence cannot show

    In vitro only; four independent knockdowns converging is genuinely strong internal evidence, but there is no xenograft and no clinical correlation with prednisone response in vivo. The gene signature that generated the hypothesis was derived and tested in overlapping sample sets from a single group.

    Evidence that argues against this (1)
    • An independent group reached the SFK conclusion by a different route and then showed it is not sufficient: in MLL-ALL patient-derived xenografts the SFK/FLT3 inhibitor RK-20449 eliminated cells in vivo, but cells resistant to it were BCL-2-dependent, and only the triple combination of glucocorticoid, RK-20449 and venetoclax achieved complete elimination. Src inhibition alone therefore selects for a second resistance state. de Groot AP et al. 2021
  6. E6moderate confidence 3 of 5: One strong study, or several smaller consistent ones.paper

    A Connectivity Map screen against a prednisolone-resistance gene signature in MLL-rearranged infant ALL nominated the PI3K inhibitor LY294002, which with other PI3K inhibitors markedly sensitised resistant MLL-rearranged ALL cells to prednisolone in vitro; the sensitising action may be mediated by inhibition of FCGR1B, and FCGR1B knockdown alone sensitised cells.

    Chemical genomic screening identifies LY294002 as a modulator of glucocorticoid resistance in MLL-rearranged infant ALL. · Spijkers-Hagelstein JA, Pinhanços SS, Schneider P, Pieters R, Stam RW · Leukemia 2014
    PMID 23958920DOI
    What this evidence cannot show

    LY294002 is a tool compound and not a drug; it is not selective and has documented off-target activity on DNA-PK, mTOR and CK2, so 'PI3K inhibition' is naming rather than measurement here. The FCGR1B route is stated by the authors as a possibility ('may be mediated'). The signature and the validation come from the same group and overlapping samples as E5.

    Evidence that argues against this (1)
    • The same group's own earlier and later papers attribute prednisolone resistance in this disease to MCL-1, to S100A8/A9-Ca2+ and to annexin A2 phosphorylation. Four partially reversible mechanisms in one small disease from largely one laboratory is a pattern that should lower confidence in each individually until an independent group reproduces one. Stam RW et al. 2010
  7. E7moderate confidence 3 of 5: One strong study, or several smaller consistent ones.paper

    In patient-derived xenografts of infant MLL-rearranged ALL that recapitulated glucocorticoid resistance in vivo, the SFK/FLT3 inhibitor RK-20449 eliminated human MLL-ALL cells and overcame glucocorticoid resistance; BH3 profiling identified BCL-2 dependence in residual cells, and the triple combination of glucocorticoid, RK-20449 and the BCL-2 inhibitor ABT-199 (venetoclax) produced complete elimination in vivo.

    Targeting critical kinases and anti-apoptotic molecules overcomes steroid resistance in MLL-rearranged leukaemia. · de Groot AP, Saito Y, Kawakami E, Hashimoto M, Aoki Y, Ono R, et al. · EBioMedicine 2021
    PMID 33581643DOI
    What this evidence cannot show

    PDX in NSG mice; 'complete elimination' is a xenograft endpoint over a defined observation window, not cure. RK-20449 is not a clinical agent. The glucocorticoid used in the in-vivo arms is not resolvable from the abstract, which is the exact prednisone/prednisolone/dexamethasone ambiguity this pair of entries exists to flag. Immune-deficient mice cannot model the infection toxicity that limits steroid intensification in infants.

    Evidence that argues against this (1)
    • Venetoclax's activity in this disease is not in doubt but its partner is contested: preclinical work in infant KMT2A-rearranged ALL pairs venetoclax with azacitidine rather than with a kinase inhibitor, on a different mechanistic rationale. Several venetoclax combinations are individually effective in xenografts, which weakens any one of them as evidence for the specific resistance mechanism proposed. Cheung LC et al. 2023
  8. E8weak confidence 2 of 5: Preliminary, small, indirect, or not yet peer reviewed.trial_registry

    This beat found no published comparison of intrathecal prednisolone against intrathecal hydrocortisone on any efficacy or toxicity endpoint, in any age group, although Interfant-06 permits either at every intrathecal timepoint.

    NCT00550992
    What this evidence cannot show

    A bounded absence, and a weakly bounded one. Bound: PubMed 2026-09-20; vocabularies searched were prednisolone, hydrocortisone, intrathecal, triple intrathecal therapy, and the Interfant protocol names, in combination with leukaemia and with infant. The searches were not exhaustive and the negative is recorded at weak confidence for that reason. What would change it: the Interfant-06 protocol document itself, which this beat could not obtain, may state a preference or a rationale that no indexed paper carries.

    Evidence that argues against this (1)
    • The absence of a comparison is not evidence the two are equivalent, and it is also not evidence anyone thinks they differ. Intrathecal corticosteroid choice is a long-standing centre-level convention in paediatric oncology and conventions of that kind are rarely trialled; the honest reading is that the question has not been asked, not that it has been answered either way. de Kluis T et al. 2025
  9. E10established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.paper

    Interfant-06 doses intrathecal drugs by age band rather than by surface area, and specifies intrathecal prednisolone 6 mg for age under 1 year and 8 mg from 1 year, with the substitution rule that if prednisolone is not available for intrathecal use it is replaced by 12 mg or 16 mg hydrocortisone respectively - a fixed 2:1 hydrocortisone-to-prednisolone ratio; intrathecal methotrexate is 6 mg / 8 mg and intrathecal cytarabine 15 mg / 20 mg across the same age bands.

    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 doses quoted are from the protocol IB section (day 10 and day 24 intrathecal therapy); the protocol states intrathecal doses 'according to age as indicated in the schemes' at every other phase and this beat verified the numbers at one phase only. The 2:1 hydrocortisone substitution ratio is asserted by the protocol without a citation.

    Evidence that argues against this (1)
    • A fixed 2:1 substitution ratio is a potency assumption, not a measurement, and it is being applied in the cerebrospinal fluid of an infant where neither agent's CSF pharmacokinetics has been characterised. This beat found no published comparison of intrathecal prednisolone with intrathecal hydrocortisone on any endpoint, so the substitution is untested in the compartment where it is used. de Kluis T et al. 2025
Connected to