Oncofusion

start / graph / drug-thiotepa

Drugmoderate confidence 3 of 5: One strong study, or several smaller consistent ones.Q53-transplant

Thiotepa - the third agent that reaches the central nervous system completely, and whose alkylating metabolites are excreted through the skin into a nappy

also called thiotepa · THIO · TT · thiophosphoramide · N,N',N''-triethylenethiophosphoramide · TESPA · TSPA · Tepadina (trade name) · TEPA / triethylenephosphoramide (the active desulfurated metabolite) · ethylenimine alkylator (class)

Thiotepa is added to transplant conditioning because it gets into the fluid around the brain and spinal cord as completely as into the blood, which matters in a leukaemia that often reaches there. It and its active breakdown product also come out through sweat, so in a baby it collects under a nappy and burns the skin - which is why children on this drug are bathed several times a day.

The same thing, for a scientist

Thiotepa is a lipophilic ethylenimine alkylator that is itself active and is also desulfurated by CYP2B6 and CYP3A4 to TEPA (triethylenephosphoramide), an alkylator in its own right with a longer half-life than the parent. In 19 children studied at 25-75 mg/m2, thiotepa and TEPA both achieved cerebrospinal-fluid to plasma AUC ratios of essentially unity (1.01 and 0.95), and thiotepa showed saturable, dose-dependent clearance falling from 28.6 to 11.9 L/m2/h across that range [E1]. Thiotepa is the third agent of both FORUM chemoconditioning backbones, BU/THIO/FLU and TREO/THIO/FLU [E2]. Its characteristic toxicity is cutaneous, because the drug and its metabolites are excreted in sweat and concentrate in occluded skin: in 15 consecutive paediatric patients on treosulfan-thiotepa-fludarabine, every one developed an acute skin reaction, 80% in the inguinal area and genitalia, with a mean affected body surface area of 71.8% - and erosions were more common in children under 2 years of age [E3].

Go deeper
Complete CNS penetration, in a disease that goes to the CNS1 deeper

The 1989 paediatric phase I study measured thiotepa and TEPA simultaneously in plasma and cerebrospinal fluid. In a child with a Rickham reservoir the CSF:plasma AUC ratios were 1.01 for thiotepa and 0.95 for TEPA, and lumbar and ventricular concentrations were 'nearly identical to simultaneous plasma concentrations' 1.

A ratio of 1.0 is unusual. Most cytotoxics used in this disease achieve a small fraction of plasma exposure in the CSF, which is why intrathecal therapy exists at all. Infant KMT2A-rearranged ALL has a high rate of CNS involvement at diagnosis, and Interfant-06 gives at least 12 and often 14 or more intrathecal injections for that reason. A conditioning agent that reaches the CNS at plasma concentration is therefore doing something the rest of the regimen cannot - and Interfant-06's Bu/Cy/Mel conditioning contains no such agent 4.

That is a hypothesis-shaped observation rather than a finding. No study has compared CNS relapse rates after thiotepa-containing versus thiotepa-free conditioning in infants, and the CSF data are from 1989 in children with solid tumours.

Saturable clearance: the dose-exposure relationship is not a line

Thiotepa clearance fell from 28.6 L/m2/h at 25 mg/m2 to 11.9 L/m2/h at 75 mg/m2, with 'evidence of the saturation of TP [TEPA] production' 1. So exposure rises faster than dose, and the conversion to the active metabolite saturates at the same time - two non-linearities pointing in opposite directions.

Conditioning doses are far above the 25-75 mg/m2 range studied there. Any mental model in which doubling the thiotepa dose doubles the exposure, or doubles the TEPA, is wrong in both directions, and there is no paediatric population pharmacokinetic model of conditioning-dose thiotepa in infants to replace it with.

The nappy problem, measured

Thiotepa is 'highly lipophilic' and it and its active metabolites 'may be excreted in part via skin in patients receiving high-dose therapy' 5. In occluded, moist, warm skin the excreted alkylator sits against the epidermis and burns it.

A prospective dermatological assessment of 15 consecutive paediatric patients on treosulfan-thiotepa-fludarabine found 3:

findingfrequency
any acute skin reaction15/15
erythematous patches, inguinal area and genitalia80%
diffuse hyperpigmentation73%
erosions, inguinal area and buttock47%
xerosis and desquamation40%
neck and axillae involvement40%
mean affected body surface area71.8%

And the line that makes this an infant entry rather than a paediatric one: 'Erosions were more prevalent in children younger than 2 years of age.'

The mechanism is not mysterious and neither is the countermeasure - frequent bathing and linen changes during and after infusion, with topical care. A single centre that implemented written preventive guidelines after one severe case reported only one further episode of significant cutaneous toxicity across 26 subsequent paediatric thiotepa transplants, 'even with double or triple transplantation protocols', and argued that such instructions 'should be detailed in the Summary of Product Characteristics' 5.

This is one of the few places in this beat where a serious infant-specific toxicity has an effective, cheap, non-pharmacological mitigation that is not in the label.

Prodrug or not? Both, and that is the point

Thiotepa is not a classical prodrug. The parent molecule is an active alkylator, and CYP2B6/CYP3A4 desulfuration produces TEPA, which is also an active alkylator with a longer half-life (4.3-5.6 h against 1.34-2.0 h for the terminal phase of thiotepa) 1.

This is a third pattern alongside the two the graph already holds. Cyclophosphamide is an inactive prodrug requiring CYP activation. Fludarabine monophosphate is a prodrug dephosphorylated then re-phosphorylated to its active form. Thiotepa is an active parent with an active metabolite, so total alkylating exposure is the sum of two species with different half-lives, and inhibiting or inducing CYP2B6 shifts the balance between them without abolishing activity.

The consequence for an infant is that CYP2B6 expression is developmentally regulated and low at birth, so the ratio of thiotepa to TEPA exposure in a two-month-old is not the ratio in a ten-year-old - and no study reported here measures it.

Facts and figures
targetDNA - ethylenimine ring-opening alkylation producing interstrand cross-links, by BOTH the parent thiotepa and its desulfurated metabolite TEPA.
modalitysmall molecule lipophilic ethylenimine alkylator; an ACTIVE parent with an ACTIVE metabolite (TEPA, via CYP2B6/CYP3A4 desulfuration) - neither a classical prodrug nor a single species
role in this diseasethird agent of both modern chemoconditioning backbones (BU/THIO/FLU, TREO/THIO/FLU); NOT a component of Interfant-06 or JPLSG MLL-10 infant conditioning
cns penetrationCSF:plasma AUC ratio 1.01 for thiotepa and 0.95 for TEPA - essentially complete, unusual among cytotoxics used in this disease
pharmacokinetic nonlinearityclearance falls from 28.6 to 11.9 L/m2/h between 25 and 75 mg/m2, with saturation of TEPA production; exposure rises faster than dose while conversion to the active metabolite saturates
half livesthiotepa terminal 1.34-2.0 h; TEPA 4.3-5.6 h
vocabularies searchedPubMed [tiab] on 2026-09-20: thiotepa OR thiophosphoramide OR 'N,N-triethylenethiophosphoramide' 2,527; Tepadina 1. Generic-name retrieval is adequate; the trade name is effectively unpublished.
infant specific
[
  "Excreted in sweat and concentrated in occluded skin, so the inguinal and buttock area under a nappy is the worst-affected site: erythema 80%, erosions 47%, mean affected BSA 71.8% in a paediatric series",
  "Erosions explicitly more prevalent in children under 2 years of age",
  "CYP2B6, which performs the desulfuration to TEPA, is developmentally regulated and low at birth, so the thiotepa:TEPA exposure ratio in an infant is unlikely to match that in an older child - and has not been measured",
  "The effective countermeasure is nursing care (frequent bathing, linen changes during and after infusion), which is not in the Summary of Product Characteristics"
]
key ae
[
  "cutaneous toxidermia with hyperpigmentation and erosions, concentrated in occluded skin",
  "mucositis",
  "myelosuppression",
  "hepatotoxicity",
  "CNS toxicity at high dose, consistent with complete CNS penetration",
  "secondary malignancy (alkylator class)"
]
resistance
[
  "glutathione S-transferase conjugation",
  "enhanced interstrand cross-link repair",
  "CYP2B6 variation alters the thiotepa:TEPA balance without abolishing alkylating activity, since both species are active",
  "no KMT2A-rearrangement-specific thiotepa measurement exists"
]
trials
[
  "NCT01949129"
]
How this could bite us back
  • The drug reaches the brain at plasma concentration. That is the reason to use it and also the reason to fear it: an agent with a CSF:plasma ratio of 1.0 delivers full alkylating exposure to a brain that is in its most rapid period of development, and the neurocognitive consequences in an infant have not been separated from those of the rest of the regimen.
  • A toxicity delivered by the nappy. The worst-affected sites are precisely the occluded, moist areas a baby spends the day in, and the countermeasure is nursing frequency rather than dose - so the toxicity is partly a function of how a unit is staffed.
  • Two active species, one name. Inducing or inhibiting CYP2B6 shifts the thiotepa:TEPA ratio without removing activity, so a drug interaction that would abolish a classical prodrug's effect here merely redistributes it between two molecules with different half-lives and different tissue distribution.
  • Non-linear clearance means a dose error does not scale. At saturating doses a 20% overdose is more than a 20% overexposure, and there is no paediatric conditioning-dose model to predict by how much.
  • The evidence for thiotepa's contribution is structurally unobtainable from the trial that made it standard. FORUM holds thiotepa constant in both chemoconditioning arms, so the agent became part of the international standard without ever being the variable in a randomised comparison.
Still open
  1. Does complete CNS penetration of a conditioning alkylator reduce CNS relapse in infant KMT2A-rearranged ALL, where CNS involvement at diagnosis is common? No thiotepa-containing versus thiotepa-free comparison exists in this population, and Interfant-06's conditioning has no CNS-penetrant component.
  2. What is the thiotepa:TEPA exposure ratio in an infant, given that CYP2B6 is developmentally regulated and low at birth? Total alkylating exposure is the sum of two active species with different half-lives, and nobody has measured the split below one year.
  3. Thiotepa clearance saturates between 25 and 75 mg/m2. What does the dose-exposure curve look like at conditioning doses in a patient weighing 7 kg? The only paediatric CSF and clearance data are from 1989 at a tenth of the relevant dose.
  4. Would a written skin-care protocol during and after thiotepa infusion reduce erosion rates in infants? The mitigation is cheap, non-pharmacological, absent from the label, and supported only by one uncontrolled before-and-after series.
Every source behind this entry (5)
  1. E1moderate confidence 3 of 5: One strong study, or several smaller consistent ones.paper

    In a phase I study of 19 children, thiotepa and its major active metabolite TEPA both penetrated cerebrospinal fluid essentially completely (CSF:plasma AUC ratios 1.01 and 0.95 in a patient with a Rickham reservoir, with lumbar and ventricular concentrations nearly identical to simultaneous plasma concentrations); thiotepa showed dose-dependent clearance falling from 28.6 L/m2/h at 25 mg/m2 to 11.9 L/m2/h at 75 mg/m2, with evidence of saturation of TEPA production; TEPA's half-life (4.3-5.6 h) exceeded thiotepa's (terminal 1.34-2.0 h).

    Phase I and pharmacokinetic evaluation of thiotepa in the cerebrospinal fluid and plasma of pediatric patients: evidence for dose-dependent plasma clearance of thiotepa. · Heideman RL, Cole DE, Balis F, Sato J, Reaman GH, Packer RJ, Singher LJ, Ettinger LJ, et al. · Cancer Research 1989
    PMID 2491958
    What this evidence cannot show

    1989, 19 children with refractory solid malignancies, at doses of 25-75 mg/m2 - an order of magnitude below the conditioning doses used before transplant, in an era of assay technology and in a population unlike an infant with leukaemia. The headline CSF:plasma ratio of 1.01 comes from ONE patient with an indwelling reservoir; the broader statement that lumbar and ventricular concentrations were nearly identical to plasma rests on a small number of paired samples. No PMID-linked DOI exists for this record. Nothing here was measured in an infant: the age distribution of the 19 children is not given in the abstract.

    Evidence that argues against this (1)
    • Complete CSF penetration is a pharmacokinetic property and not a demonstrated clinical benefit in this disease. No comparison of CNS relapse after thiotepa-containing versus thiotepa-free conditioning exists in infant KMT2A-rearranged ALL, and Interfant-06 reports that the CNS relapse rate on its protocol was already low without cranial irradiation, so the headroom for a CNS-penetrant conditioning agent to help may be small. Pieters R et al. 2006
  2. E2strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.paper

    Thiotepa is the common third agent of both chemoconditioning backbones compared within FORUM - busulfan/thiotepa/fludarabine in 180 children and treosulfan/thiotepa/fludarabine in 128 - which gave indistinguishable 3-year overall survival (0.71 vs 0.72), event-free survival (0.60 vs 0.55) and GVHD outcomes.

    Comparable outcomes after busulfan- or treosulfan-based conditioning for allo-HSCT in children with ALL: results of FORUM. · Kalwak K, Moser LM, Poetschger U, Bader P, Kleinschmidt K, Meisel R, Dalle JH, Yesilipek A, et al. · Blood Advances 2025
    PMID 39602342DOI
    What this evidence cannot show

    Because thiotepa is present in BOTH arms of this comparison, the study says nothing whatever about thiotepa's own contribution - it is held constant by design. No randomised trial has tested adding or removing thiotepa from a paediatric conditioning regimen. All patients were over 4 years old at HSCT.

    Evidence that argues against this (1)
    • Both thiotepa-containing regimens sit inside the FORUM arm that lost to irradiation: 2-year overall survival 0.75 for chemoconditioning versus 0.91 for TBI plus etoposide, with relapse 0.33 versus 0.12. Whatever thiotepa contributes, the regimens containing it did not match TBI in children over 4. Peters C et al. 2021
  3. E3moderate confidence 3 of 5: One strong study, or several smaller consistent ones.paper

    In a prospective dermatological study of 15 consecutive paediatric patients conditioned with treosulfan-thiotepa-fludarabine and ATG, all 15 developed an acute skin reaction; erythematous patches of the inguinal area and genitalia occurred in 80%, diffuse hyperpigmentation in 73%, erosions of the inguinal area and buttock in 47%, and xerosis and desquamation in 40%, with a mean affected body surface area of 71.8%; erosions were more prevalent in children younger than 2 years of age.

    PMID 31840362DOI
    What this evidence cannot show

    Fifteen patients at a single centre in one calendar year, all with NON-MALIGNANT disease, and the regimen contains treosulfan as well as thiotepa - treosulfan also causes skin toxicity, so the study cannot attribute the reactions to thiotepa alone. There was no comparator arm without thiotepa. The statement that erosions were more prevalent under 2 years of age is reported without a denominator or a p-value in the abstract, so it is a direction and not a measured effect size. The authors emphasise that all eruptions resolved without long-term sequela, which should be read alongside the alarming 71.8% affected body surface area.

    Evidence that argues against this (1)
    • The severity is largely preventable by nursing care rather than by dose reduction, which reframes it from a drug property to a care-delivery problem. After a single centre implemented written preventive guidelines - frequent washing and linen changes during and after infusion - only one of 26 subsequent paediatric thiotepa transplant patients experienced comparable cutaneous toxicity, and the rest had only mild reactions even on double or triple transplantation protocols. Van Schandevyl G et al. 2019
  4. E4established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.book

    The Interfant-06 conditioning regimen contains no thiotepa; it is oral busulfan on days -7 to -4, cyclophosphamide on days -3 and -2 and melphalan on day -1, with ATG for matched non-sibling donors.

    INTERFANT-06: International collaborative treatment protocol for infants under one year with acute lymphoblastic or biphenotypic leukemia (protocol document, SKION version 01, April 2006) · Pieters R, de Lorenzo P, Valsecchi MG, Peters C, and the Interfant-06 study group · Interfant-06 study group / SKION protocol document 2006
    What this evidence cannot show

    Read from chapter 6.5 of the retrieved 2006 SKION protocol document on 2026-09-20. It states intent, not delivered practice, and later protocol amendments were not retrieved. The absence of thiotepa from a 2006 protocol is unremarkable - thiotepa-containing paediatric conditioning became standard later - so this is a fact about protocol chronology rather than a considered rejection of the agent.

    Evidence that argues against this (1)
    • Thiotepa's absence from Interfant-06 should not be read as a judgement against it, because it was not available as a standard component when that protocol was written. The field's subsequent convergence on thiotepa-containing backbones in both arms of FORUM suggests the omission reflects the date rather than a decision. Kalwak K et al. 2025
  5. E5moderate confidence 3 of 5: One strong study, or several smaller consistent ones.paper

    Thiotepa is highly lipophilic and it and its active metabolites may be excreted in part via the skin at high dose, producing a characteristic toxidermia; after preventive care guidelines were implemented at one centre, only one of 26 subsequent paediatric thiotepa transplant patients showed comparable cutaneous toxicity, and the authors argue such instructions should be detailed in the Summary of Product Characteristics.

    PMID 30185131DOI
    What this evidence cannot show

    A single case report with an uncontrolled retrospective follow-up of 26 patients at the same centre. The before-and-after comparison has no concurrent control and confounds the guideline with every other change in that unit over four years, including awareness effects on reporting. The index case is a 4.5-year-old with medulloblastoma on autologous transplant, not an infant with leukaemia on allogeneic transplant. 'Only mild cutaneous toxicity was observed' is a clinical impression, not a graded measurement.

    Evidence that argues against this (1)
    • The prospective dermatological study that graded skin systematically found 100% incidence of acute skin reaction with a mean 71.8% affected body surface area on a thiotepa-containing regimen, which is difficult to reconcile with 'only mild cutaneous toxicity' after a guideline. The difference may be real, or may be that systematic dermatological assessment on scheduled days detects what routine clinical care records as unremarkable. Retrospective absence of a recorded toxicity is weak evidence that it did not occur. Even-Or E et al. 2020
Connected to