Prednisone - the prophase drug, the prognostic factor, and a prodrug that has never itself been tested on a KMT2A-rearranged blast
also called prednisone · PRED · PDN · Deltasone · prednisone phase · prednisone prophase · 1,4-pregnadiene-17alpha,21-diol-3,11,20-trione
Prednisone is the steroid every infant with leukaemia gets first, alone, for seven days, and how fast the leukaemia melts in that week is one of the strongest predictors of whether the child will be cured. It is also a prodrug: the liver has to turn it into prednisolone before it does anything at all.
The same thing, for a scientist
In Interfant-06 prednisone is given orally or intravenously three times daily on days 1-7 as a single-agent prophase, with intrathecal methotrexate and intrathecal **prednisolone** on day 1 [E1]. The cytoreduction achieved in that window defines the prednisone good/poor responder split, and poor prednisone response is one limb of the conjunction (with KMT2A rearrangement and age under 6 months) that defines the Interfant-06 high-risk group [E1][E2]. Pieters et al. report KMT2A rearrangement as the most important prognostic factor for EFS, followed by age, WBC count and prednisone response [E2]. **The pharmacology matters and is routinely elided.** Prednisone is an 11-keto prodrug with negligible affinity for NR3C1; hepatic 11beta-hydroxysteroid dehydrogenase type 1 reduces it to prednisolone, which is the receptor ligand. In healthy men 76% of an administered dose is recycled between the two steroids and oral prednisone bioavailability is about 62% [E3]. Consequently **every in-vitro measurement of steroid resistance in this disease used prednisolone** - the >580-fold MTT resistance of infant ALL blasts [E4], the MCL-1 association [E5] - while the in-vivo prognostic factor is prednisone. The graph should not treat those as one quantity, and this entry exists so that it does not.
Where it sits in the backbone
From the registered Interfant-06 protocol 1: prednisone orally or IV three times daily on days 1-7, as the only systemic cytotoxic agent in that window, with MTX and prednisolone intrathecally on day 1. On day 8 the steroid is switched to dexamethasone for the five-week remission induction; prednisone is not given again anywhere in the protocol.
So prednisone's whole systemic career in this protocol is seven days, and its purpose after cytoreduction is measurement. The prophase is a single-agent in-vivo drug-sensitivity assay whose readout is used to allocate risk.
The registry lists prednisone and prednisolone as two separate DRUG interventions on this study 1. That is not a duplicate record - they occupy different routes: prednisone is the systemic prophase agent, prednisolone (or hydrocortisone) is the intrathecal steroid given at every IT timepoint from day 1 of induction through week 15 of maintenance.
What this beat could not establish: the mg/m2 of the prophase. The registry's detailed description gives route, frequency and days and no dose 1, and the Interfant-06 primary publication is not open access. The dose is deliberately left blank here rather than supplied from a sibling protocol or from recall.
The prognostic weight, stated exactly
Interfant-06's own abstract names the prognostic hierarchy: KMT2A rearrangement first, then age, then WBC count, then prednisone response 2. The risk model then uses prednisone response as a criterion, not merely a covariate: high risk is the conjunction of KMT2A rearrangement and age under 6 months and (WBC >= 300 x10^9/L and/or poor prednisone response) 1. An infant can be pushed into the high-risk arm - and therefore toward transplant in first remission - by the prednisone prophase alone.
A caution this graph should carry. Prednisone response is a composite readout of absorption, hepatic conversion, receptor signalling and apoptotic competence. Attributing a poor prednisone response to intrinsic glucocorticoid resistance of the blast is the usual reading and it is an inference: nothing in the prophase separates a blast that cannot die from an infant who did not activate enough prodrug. See drug-prednisolone for what has actually been measured on the blast.
Infant pharmacology: the conversion step is the part nobody measured in infants
Hepatic 11beta-HSD1 activity, oral absorption and plasma protein binding (prednisolone binds transcortin saturably, which is why its kinetics are nonlinear) are all developmentally regulated, and the first year of life is when they change fastest 6. The only paediatric population-PK study of prednisolone after oral prednisone this beat found enrolled kidney-transplant recipients aged 3.4 to 17.2 years and reported large interindividual variability, with bodyweight and ciclosporin cotreatment as covariates 7. No infant is in it.
That is the infant-specific finding for this drug, and it is an absence: the prophase is a dose-response measurement made on a population whose capacity to activate the drug has never been characterised. A protocol that uses a fixed prednisone dose to classify risk is assuming conversion is uniform across 0-12 months.
The dose, and the multiplier nobody has justified
The protocol document settles what the registry did not 9: prednisone 60 mg/m2/day, oral or IV, three times daily, days 1-7, and then dexamethasone 6 mg/m2/day days 8-28.
And it states the infant dose rule that governs every systemic drug in this beat:
Calculate the surface area at the start of each treatment block and then adjust. Children <6 months of age: 2/3 of the calculated dose based on surface area. Children 6 through 12 months of age: 3/4 of the calculated dose. Children >12 months of age: full dose. Dose reductions are for all drugs including glucocorticoids, but not for intrathecal drugs.
Three things follow and all three matter for this graph.
- The prognostic prophase is dosed differently across the age range it stratifies on. An infant of 5 months receives 40 mg/m2-equivalent and an infant of 7 months receives 45; the risk model then uses prednisone response, and the same model uses age 6 months as a cut-point. The dose step and the risk cut-point sit at the same age. Nothing in the trial reports separates a biological age effect from the dose step that coincides with it.
- The multipliers have no stated basis. The protocol gives 2/3 and 3/4 with no derivation, no pharmacokinetic justification and no reference, and inherits them from Interfant-99. Where the equivalent assumption has been tested - vincristine - BSA-normalised clearance in infants turned out not to differ from older children's, and the customary reduction produced measurably lower exposure [E9 counter].
- Intrathecal therapy is exempt, which is why
drug-prednisolonebehaves differently from every other agent here: its dose is banded by age (6 mg under 1 year, 8 mg from 1 year) and is not reduced at all [seedrug-prednisoloneE10].
| target | NR3C1 glucocorticoid receptor - INDIRECT: prednisone itself is an 11-keto prodrug with negligible receptor affinity and acts only after hepatic reduction to prednisolone. No gene node for NR3C1 exists in this graph as of cycle 28. |
|---|---|
| target note | The schema requires data.target. A node id was deliberately not used here because the honest target of prednisone is 'the liver, then the receptor'. See drug-prednisolone for the entry whose target is the receptor directly. |
| modality | small molecule, synthetic 11-keto glucocorticoid PRODRUG |
| active metabolite | prednisolone (see drug-prednisolone) |
| activation enzyme | 11beta-hydroxysteroid dehydrogenase type 1 (hepatic, reductase direction); 11beta-HSD2 catalyses the reverse |
| interconverted fraction healthy adults PCT | 76 |
| oral bioavailability PCT | 62 |
| interfant06 position | prednisone prophase, days 1-7, orally or IV three times daily, single systemic agent; replaced by dexamethasone from day 8 |
| interfant06 dose | 60 mg/m2/day PO or IV, days 1-7, given three times daily (protocol chapter 5.1) |
| role in risk model | poor prednisone response is one limb of the Interfant-06 high-risk conjunction (KMT2A-r AND age <6 months AND (WBC >=300e9/L and/or poor prednisone response)) |
| kmt2a specific evidence | PROGNOSTIC ONLY. Prednisone response is prognostic in KMT2A-rearranged infant ALL and is an input to risk allocation; there is no in-vitro measurement of prednisone (as distinct from prednisolone) on KMT2A-rearranged blasts, because the prodrug is inert in culture. |
| infant dose adjustment | body-surface-area dose x 2/3 under 6 months, x 3/4 from 6 through 12 months, full dose above 12 months; recalculated at the start of each block; applies to glucocorticoids and to all systemic drugs, NOT to intrathecal drugs |
| toxicity dose rule | for hypertension not controlled by antihypertensives and sodium restriction, the protocol permits reducing the glucocorticoid dose by 30-50% (chapter 7) |
key ae
[ "hyperglycaemia", "hypertension", "infection during the single-agent prophase week", "tumour lysis syndrome on cytoreduction", "adrenal suppression" ]
resistance
[ "high MCL-1 expression (associated with in-vivo prednisone poor response in MLL-rearranged infant ALL)", "S100A8/S100A9 over-expression suppressing glucocorticoid-evoked cytosolic Ca2+", "Src-mediated annexin A2 phosphorylation", "PI3K-dependent signalling (LY294002-reversible)", "every one of the above was measured with prednisolone, not prednisone" ]
trials
[ "NCT00550992" ]
- **This drug is given to a baby in the first week after a leukaemia diagnosis, with no other cytotoxic cover, and part of the point of the week is to measure her.** A prophase that fails cytoreduction is simultaneously a prognostic result and a week of near-untreated leukaemia. Any proposal to shorten, intensify or replace the prophase trades information against time, and the information is currently used to decide whether the child is sent to transplant.
- The high-risk allocation that prednisone response feeds into leads to allogeneic transplant in first remission with busulfan/cyclophosphamide/melphalan conditioning. A misclassification here is not academic: it is a transplant a child did or did not receive.
- Every mechanistic finding about 'glucocorticoid resistance' in this disease - MCL-1, S100A8/A9, annexin A2, PI3K - was measured on prednisolone in culture, while the clinical criterion is prednisone in an infant. Proposals to resensitise (Src inhibitors, BCL-2 family inhibitors, PI3K inhibitors) are aimed at a target defined in one system and validated against an endpoint from another.
- Steroid intensification is the obvious answer to steroid resistance and it is the one this disease can least afford: infants in induction die of infection. The most recent prospective infant ALL trial reported four treatment-related deaths, all infectious, with grade 3-4 infection in 44% of patients during induction.
- A prophase week is also a week of selection. If glucocorticoid resistance in KMT2A-rearranged ALL is partly a property of a subclone rather than of the bulk, the prophase may enrich for exactly the cells that will relapse, and the response measurement would then be reporting on the population that is about to be deleted.
- What is the distribution of prednisolone exposure after a fixed prednisone prophase dose across the 0-12 month age range? Until that is measured, 'prednisone poor response' conflates a resistant blast with an infant who did not activate the drug, and the two have opposite implications for therapy.
- Does prednisone response add prognostic information beyond end-of-induction MRD in the KMT2A-rearranged group? Interfant-06 uses it as a risk criterion and MLL-10 does not, and no head-to-head exists.
- Would the Interfant high-risk group change if response were measured on day-8 prednisolone exposure rather than on day-8 blast count? This is answerable retrospectively from stored samples and has never been attempted.
- Is 11beta-HSD1 expressed in KMT2A-rearranged blasts themselves, and could intracellular conversion contribute to or blunt local prednisolone exposure? Not addressed anywhere in the literature this beat searched.
- E1established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.trial_registry
In the registered Interfant-06 protocol, prednisone is given orally or intravenously three times daily on days 1-7 as a single-agent prophase with intrathecal methotrexate and intrathecal prednisolone on day 1; the steroid then switches to dexamethasone for remission induction on days 8-28; the registry lists prednisone and prednisolone as two separate drug interventions, and it states no dose for the prophase.
Interfant-06: International Collaborative Treatment Protocol for Infants Under One Year With Acute Lymphoblastic or Biphenotypic Leukemia (NCT00550992) · ClinicalTrials.gov 2007What this evidence cannot show
A registry record, not the protocol document. The detailed description gives route, schedule and days but no mg/m2 for any agent. The intervention list on this record also omits dexamethasone although the detailed description prescribes it on days 8-28, so the intervention list is demonstrably incomplete and must not be read as an exhaustive drug inventory; it is used here only for the positive fact that prednisone and prednisolone are listed separately. Fetched 2026-09-20 from the ClinicalTrials.gov v2 API.
Evidence that argues against this (1)
- Interfant-06's risk criteria are stated in the opposite direction in the primary publication's abstract, which defines high risk using age OLDER than 6 months, while the registry and the protocol's own MRD paper both give age UNDER 6 months. The prednisone-response criterion is common to all three, but the group it selects differs. This discrepancy is already recorded in `concept-interfant-06-treatment-backbone` and is not resolved here. Pieters R et al. 2019
- E2strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.paper
In Interfant-06 (651 infants under 1 year, international phase III), KMT2A rearrangement was the most important prognostic factor for event-free survival, followed by age, white blood cell count and prednisone response.
Outcome of Infants Younger Than 1 Year With Acute Lymphoblastic Leukemia Treated With the Interfant-06 Protocol: Results From an International Phase III Randomized Study. · Pieters R, De Lorenzo P, Ancliffe P, Aversa LA, Brethon B, Biondi A, et al. · J Clin Oncol 2019What this evidence cannot show
Read from the abstract only; the full text is not open access and returned no PMCID. The ordering is the authors' own multivariable summary and this beat did not see the model, the covariate coding, or whether prednisone response was entered as a binary or a continuous blast count. Prednisone response is also partly collinear with the other three factors, since KMT2A-rearranged infants respond worse.
Evidence that argues against this (1)
- Prognostic weight is protocol-specific. In the JPLSG MLL-10 trial, which stratified on KMT2A status, age and CNS leukaemia rather than on prednisone response, the significant multivariable factors for poor prognosis were female sex and MRD >= 0.01% at the end of early consolidation - prednisone response does not appear. A factor that is decisive in one risk model can be absent from another that reaches a comparable or better 3-year EFS (66.2% for KMT2A-r). Tomizawa D et al. 2020
- E3established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.paper
Prednisone and prednisolone interconvert extensively in humans: after an oral prednisone dose of 0.8 mg/kg and an intravenous prednisolone dose of 0.66 mg/kg in 12 healthy men, bioavailability was about 62% and 76% of the dose was recycled between the two steroids, with the formation clearance of prednisolone from prednisone much greater than the reverse.
Bioavailability and reversible metabolism of prednisone and prednisolone in man. · Garg V, Jusko WJ · Biopharm Drug Dispos 1994What this evidence cannot show
Twelve healthy adult men, 1994, single dose. Nothing here is measured in infants, in children, or under the hepatic and nutritional conditions of leukaemia induction. It establishes the direction and magnitude of the prodrug relationship in adults and nothing about its variance in the first year of life.
Evidence that argues against this (1)
- Prednisolone's own pharmacokinetics are nonlinear and highly variable in children even when the conversion step is bypassed conceptually: in 39 paediatric kidney-transplant recipients aged 3.4-17.2 years, free prednisolone was undetectable at 12 h in some patients and still measurable at 24 h in others, with ciclosporin cotreatment reducing clearance by 67%. Extensive interconversion therefore does not imply predictable exposure. de Truchis C et al. 2023
- E4strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.paper
Infant ALL blasts are more than 580-fold resistant to prednisolone in the MTT assay compared with 425 children over 1 year with common/preB ALL (P = 0.001), and t(4;11)-positive cases are 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 2004What this evidence cannot show
This is a PREDNISOLONE measurement, not a prednisone measurement, and it is cited in this entry only to mark that distinction. 44 infants; a >580-fold figure is a censored estimate arising because many infant samples never reached 50% kill at the highest concentration tested, so the true ratio is unbounded above rather than equal to 580. The authors' own conclusion attributes much of the profile to the proB immunophenotype rather than to MLL rearrangement.
Evidence that argues against this (1)
- The in-vitro resistance profile that anchors every steroid claim in this beat was measured with PREDNISOLONE, not prednisone, because prednisone is a prodrug and is essentially inert in cell culture where there is no hepatic 11beta-HSD1 to convert it. So the famous '>580-fold resistant' number is a prednisolone number, and the protocol's prophase drug is prednisone. The link between them is a pharmacokinetic inference, not a measurement in these patients. Ramakers-van Woerden NL et al. 2004
- E5moderate confidence 3 of 5: One strong study, or several smaller consistent ones.paper
MCL-1 expression was significantly higher in MLL-rearranged infant patients who showed a poor response to prednisone in vivo than in prednisone good responders, and high MCL-1 was associated with prednisolone resistance in vitro; knocking MCL-1 down produced only 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 2010What this evidence cannot show
The in-vivo arm of this result is the only place in the cited literature where the word prednisone (rather than prednisolone) attaches to a measurement on patient blasts, and it is an association with the clinical prophase response, not a drug-exposure measurement. The authors themselves state the result 'strongly implies that high-level MCL-1 expression is not the sole mechanism'. Cohort sizes for the in-vivo comparison are not given in the abstract.
Evidence that argues against this (1)
- At least three further mechanisms have been reported for glucocorticoid resistance in the same disease by the same group - S100A8/S100A9 suppression of prednisolone-evoked cytosolic Ca2+, Src-driven annexin A2 phosphorylation, and a PI3K-dependent axis found by chemical genomic screening - so MCL-1 is one of several partially sufficient explanations rather than the mechanism. Spijkers-Hagelstein JA et al. 2012
- E6moderate confidence 3 of 5: One strong study, or several smaller consistent ones.paper
The pharmacokinetics of drugs used to treat infant ALL are shaped by rapid first-year developmental changes in renal maturation, body composition, blood composition and the ontogeny of drug-metabolising enzymes and transporters, and current dosing guidelines are largely extrapolated from paediatric or adult standards by body-surface-area or age adjustment.
Effect of developmental changes on pharmacokinetics of drugs used in the treatment of infant acute lymphoblastic leukaemia-A comprehensive review. · de Kluis T, Brigitha LJ, Hanff LM, Huitema ADR, van der Sluis IM, Diekstra MHM · Br J Clin Pharmacol 2025What this evidence cannot show
A narrative review, cited as a map of which developmental processes are relevant and for the statement that infant dosing is extrapolated rather than derived. It is not a measurement and it is not specific to prednisone. Reviews are weak evidence for any particular number and are used here only for the framing claim.
Evidence that argues against this (1)
- Developmental immaturity does not always mean lower clearance or a need for dose reduction, and assuming it does has caused underdosing: vincristine clearance normalised to body surface area was NOT significantly different between neonates/infants and older children, so the customary infant dose reduction produced significantly lower drug exposure. Barnett S et al. 2022
- E7moderate confidence 3 of 5: One strong study, or several smaller consistent ones.paper
The only paediatric population-pharmacokinetic analysis of free prednisolone after oral prednisone identified by this beat enrolled 39 kidney-transplant recipients aged 3.4 to 17.2 years; it reported large interindividual variability with bodyweight and ciclosporin cotreatment as covariates, and contains no infants.
Prednisolone pharmacokinetics after oral prednisone administration in paediatric patients with kidney transplant. · de Truchis C, Bouazza N, Foissac F, Charbit M, Dehoux L, Lui G, et al. · Br J Clin Pharmacol 2023What this evidence cannot show
A bounded absence, not a null result. Bound: PubMed 2026-09-20, `prednisone[ti] AND prednisolone[ti] AND (pharmacokinetics[ti] OR conversion[ti] OR metabolism[ti])` returns 41 records, all title-checked; also `prednisone prednisolone interconversion 11beta-hydroxysteroid dehydrogenase pharmacokinetics` (1 record, off topic - a placental transfer study in HELLP syndrome) and `prednisolone prednisone potency clinical equivalence liver conversion children` (0 records; that zero is uninterpretable on its own because the parser expanded 'clinical' and 'care' into unrelated MeSH sets, so it is not relied on). Vocabularies searched: prednisone, prednisolone, 11beta-HSD, interconversion, bioavailability. Not searched: non-English literature, and paediatric dosing guidelines not indexed in PubMed.
Evidence that argues against this (1)
- Absence of an infant PK study is not evidence that conversion is impaired in infants. Nobody has shown a deficit; the point is only that the assumption of uniform conversion across 0-12 months is untested, and an untested assumption can be right. de Kluis T et al. 2025
- E9established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.paper
The Interfant-06 protocol document specifies prednisone 60 mg/m2/day PO or IV on days 1-7 and dexamethasone 6 mg/m2/day PO or IV on days 8-28, and directs that the body-surface-area dose be multiplied by 2/3 for children under 6 months of age and by 3/4 for children 6 through 12 months, full dose only above 12 months, with the surface area recalculated at the start of each treatment block; the reduction applies to all drugs INCLUDING the glucocorticoids but NOT to intrathecal drugs, which are dosed by age band instead.
Interfant-06 treatment protocol for infants under 1 year of age with acute lymphoblastic leukaemia - International collaborative treatment protocol, SKION version 01 (April 2006): chapter 5 chemotherapy schedule and chapter 7 modification for toxicity · Interfant-06 study group / SKION (Stichting Kinderoncologie Nederland, Dutch Childhood Oncology Group) · Interfant-06 study protocol document (90 pp.) 2006What 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 2/3 and 3/4 multipliers are stated on every chemotherapy-schedule page of the protocol and are inherited unchanged from Interfant-99, where the same rule appears in the trial scheme; the protocol gives no derivation, no pharmacokinetic justification and no reference for the two numbers.
Evidence that argues against this (1)
- The 2/3 multiplier under 6 months is an unjustified convention and there is direct evidence that this class of reduction can underdose. For vincristine, clearance normalised to body surface area was NOT significantly different between neonates/infants and older children; the lower doses given to infants produced significantly lower AUC (P = 0.047), and doses below 0.05 mg/kg gave significantly lower exposure than doses at or above it (P <= 0.0001), leading the authors to recommend that doses under 0.05 mg/kg not be used in this population. Barnett S et al. 2022
THE PROGNOSTIC FACTOR IS PREDNISONE while the measured molecule is prednisolone - two entries on purpose.
prodrug -> active form; hepatic 11beta-HSD1 reduces prednisone to prednisolone, 76% of dose recycled in healthy adults
- part of →Interfant-06 — 651 infants, and the definitive answer that myeloid-style consolidation does not help
single-agent prophase, days 1-7, oral or IV three times daily
prophase phase of the backbone
prophase agent and prognostic instrument
replaces prednisone from day 8 of Interfant-06 induction
steroid-response pathways are downregulated in the most immature blasts of the youngest patients
- ← confers resistance toSET::NUP214 adult T-ALL is gamma-delta-lineage-restricted and resistant to steroids and chemotherapy early - the same cell state CALM-AF10 occupies
corticosteroid resistance 91% versus 44% in GRAALL adults, P = .003