Graft-versus-host disease in a transplanted infant - the only antileukaemic benefit and the main cause of death are the same process at different grades
also called GVHD · graft-versus-host disease · aGVHD · cGVHD · acute graft versus host disease in infants · GVHD prophylaxis Interfant-06
When donor immune cells attack the child's own tissues it is called graft-versus-host disease. A mild version is good, because the same cells also attack any leftover leukaemia; a severe version kills more children than it saves. Doctors cannot aim for the mild version, and in a baby the drugs used to control it leave an immune system that has never met anything before.
The same thing, for a scientist
GVHD is the clinical face of alloreactivity and is therefore inseparable from graft-versus-leukaemia. The CIBMTR ALL analysis defines the window precisely: grade I-II acute GVHD without chronic GVHD gives the best overall survival (mortality HR 0.83-0.76 versus no GVHD), while grade III-IV acute GVHD raises non-relapse mortality 2.69- to 3.91-fold and abrogates the relapse protection entirely [E1]. In the largest infant dataset, 252 children with acute leukaemia diagnosed under 1 year receiving unrelated cord blood had 40% +/- 3% grade II-IV acute GVHD at day 100 and 13% +/- 2% chronic GVHD at 4 years, with 1-year transplant-related mortality of 23% [E2]. Interfant-06 prophylaxes with ciclosporin alone for matched sibling donors and ciclosporin plus short-course methotrexate plus ATG-Fresenius for matched non-sibling donors, tapering ciclosporin from day +60 in the absence of symptoms [E3]. Conditioning intensity contributes: melphalan exposure above a cohort median of 6.81 mg*h/L was associated with more acute GVHD in paediatric transplant, plausibly through mucosal injury initiating alloreactivity [E4].
The grade window, and why it cannot be aimed at
From 5,215 ALL transplants 1:
| GVHD state | relapse | non-relapse mortality | net overall survival |
|---|---|---|---|
| none | reference | reference | reference |
| grade I-II acute, no chronic | lower (HR 0.49-0.69) | not raised | best (mortality HR 0.83-0.76) |
| grade III-IV acute | lower | 2.69-3.91x higher | worse - protection abrogated |
| chronic GVHD | lower | raised | beneficial only in advanced disease |
There is no intervention that produces grade I-II acute GVHD and stops. Prophylaxis is titrated to prevent GVHD, treatment is escalated once it appears, and the grade is an emergent property of donor-recipient mismatch, graft composition, conditioning intensity and the recipient's mucosal barrier. The therapeutic target is a grade, and the available controls are all-or-nothing.
What Interfant-06 actually does about it
Chapter 6.6 3:
| donor | prophylaxis |
|---|---|
| matched sibling (MSD) | ciclosporin A from day -1, 1.5 mg/kg i.v. twice daily, switching to 3 mg/kg/day orally in two doses when tolerated; levels measured and dose adjusted; tapered from day +60 by about 20% of the initial dose per week if no GVHD |
| matched donor (MD) | ciclosporin as above plus methotrexate 10 mg/m2 i.v. on days +1, +3 and +6 with leucovorin 15 mg/m2 on days +2, +4 and +7 plus ATG-Fresenius 20 mg/kg on days -3, -2 and -1 |
Two details worth noticing. Ciclosporin is monitored and busulfan is not - the protocol mandates ciclosporin level measurement and dose adjustment while making busulfan level measurement conditional on being 'possible' 3. The immunosuppressant gets therapeutic drug monitoring; the agent with the narrow therapeutic window and the veno-occlusive risk does not.
And methotrexate appears here in a third role. The graph already holds drug-methotrexate as an antileukaemic agent given intrathecally and systemically in Interfant-06. On days +1, +3 and +6 after transplant it is an immunosuppressant, at 10 mg/m2, with leucovorin rescue - a different dose, a different purpose and a different rescue strategy from anything in the frontline protocol.
The infant-specific part, and the size of what is not known
What is measured. In 252 children with acute leukaemia diagnosed before one year of age and transplanted with unrelated cord blood: grade II-IV acute GVHD 40% at day 100, chronic GVHD 13% at four years, transplant-related mortality 23% at one year 2. In 40 infants transplanted at a single centre between 1982 and 2003, 20 of 40 developed acute GVHD and 7 developed chronic GVHD 5.
What is not measured, and is widely assumed. This beat found no study comparing GVHD incidence or severity between infants and older children with the same donor type and prophylaxis. The clinical belief that younger recipients have less GVHD is not supported by anything retrieved here, and the cord blood figures above cannot test it because cord blood itself carries a low chronic GVHD risk by graft type.
What is specific and is not about incidence at all. An infant treated for GVHD is immunosuppressed during the window in which the immune system normally builds its entire memory repertoire. Interfant-06's own supportive-care chapter anticipates this: immunoglobulin substitution to a target above 500 mg/dL until normal values are achieved, aciclovir to at least day +100, co-trimoxazole from day +14 until four weeks after the end of immunosuppression, weekly CMV PCR, and the observation that 'especially during severe GVHD a long lasting severe immunodeficiency can be expected' 3. A four-year-old recovering from GVHD is rebuilding a repertoire; a seven-month-old is building one for the first time under immunosuppression.
| kind | toxicity |
|---|---|
| severity | grade I-II acute GVHD is net beneficial in paediatric ALL; grade III-IV raises non-relapse mortality 2.69-3.91-fold and abrogates the relapse protection |
| mitigation | Interfant-06: ciclosporin alone (matched sibling) or ciclosporin + short methotrexate + ATG-Fresenius (matched donor), with ciclosporin levels monitored and tapered from day +60; plus immunoglobulin substitution, aciclovir to day +100, co-trimoxazole and weekly CMV PCR |
| not measured | No study retrieved by this beat compares GVHD incidence or severity between infants and older children holding donor type, graft source, conditioning and prophylaxis constant. The clinical belief that younger recipients have less GVHD is not supported by anything found here, and the two infant series differ by graft source in a way that makes them incomparable to each other. |
| conditioning contribution | melphalan exposure above 6.81 mg*h/L associated with more acute GVHD; busulfan overexposure above 98.0 mg*h/L associated with increased toxicity and GVHD in FORUM |
| infant specific consequence | immunosuppression during the window in which immunological memory is first built, on top of alkylator lymphoablation and any prior CD19-directed B-cell aplasia |
affects
[ "drug-busulfan", "drug-melphalan", "drug-fludarabine", "drug-thiotepa", "drug-antithymocyte-globulin" ]
interfant06 prophylaxis
{
"MSD": "ciclosporin A 1.5 mg/kg i.v. b.d. from day -1, then 3 mg/kg/day orally in two doses; levels measured and adjusted; taper from day +60 by ~20% of initial dose per week",
"MD": "ciclosporin as MSD, plus methotrexate 10 mg/m2 i.v. days +1/+3/+6 with leucovorin 15 mg/m2 days +2/+4/+7, plus ATG-Fresenius 20 mg/kg days -3/-2/-1"
}measured incidence in infants
{
"ebmt_cord_blood_n252": "grade II-IV acute GVHD 40% +/- 3% at day 100; chronic GVHD 13% +/- 2% at 4 years; TRM 23% +/- 3% at 1 year",
"single_centre_n40_marrow_1982_2003": "acute GVHD 20/40, chronic GVHD 7/40"
}- The benefit and the harm are the same process. Suppressing GVHD suppresses graft-versus-leukaemia, and there is no way to separate them pharmacologically - every intensification of prophylaxis is a deliberate reduction of the mechanism the transplant exists to deliver.
- The target is a grade and the controls are binary. Nothing available to a clinician produces grade I-II acute GVHD and holds it there, so the beneficial state is arrived at by luck rather than by design.
- An infant treated for GVHD is immunosuppressed while building its first immune repertoire, not rebuilding a lost one. Revaccination schedules, infection risk and the duration of vulnerability are all different, and the protocol addresses none of them by age.
- ATG is given to non-sibling recipients for GVHD prophylaxis and is a polyclonal T-cell depleting agent, which means the patients at higher alloreactive risk also receive the deepest ablation of the cells that would mediate any antileukaemic effect.
- Conditioning intensity feeds GVHD. Melphalan and busulfan overexposure both associate with more GVHD, so a dosing error in an unmonitored conditioning agent produces an immunological complication weeks later that will be attributed to donor mismatch.
- Chronic GVHD in a child under one year runs concurrently with growth, neurodevelopment and organ maturation, so a complication that is chronic and survivable in an adult is developmental in an infant, and none of the late-effects literature separates GVHD's contribution from conditioning's.
- Do infants have less GVHD than older children at equal donor mismatch, graft source and prophylaxis? This is widely believed, was not found measured by this beat, and is directly answerable from any large paediatric registry.
- If graft-versus-leukaemia is impaired in KMT2A-rearranged infant leukaemia, should the GVHD target in this subtype be zero rather than grade I-II? The entire prophylaxis strategy is calibrated to preserve a benefit that has not been demonstrated in this disease.
- What is the long-term immunological cost of GVHD immunosuppression in a patient with no pre-existing immune memory? Interfant-06 anticipates 'long lasting severe immunodeficiency' in prose and specifies no age-adjusted management.
- Interfant-06 monitors ciclosporin levels and not busulfan levels. Given that busulfan overexposure independently increases GVHD, would mandatory busulfan targeting reduce GVHD more than ciclosporin adjustment does?
- E1strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.paper
In 5,215 ALL transplants reported to CIBMTR, grade I-II acute GVHD without chronic GVHD gave the best overall survival in adults and children in CR1/CR2 (mortality HR 0.83-0.76 versus no GVHD), while grade III-IV acute GVHD raised non-relapse mortality 2.69- to 3.91-fold and abrogated the protection from relapse, resulting in inferior overall survival.
The impact of the graft-versus-leukemia effect on survival in acute lymphoblastic leukemia. · Yeshurun M, Weisdorf D, Rowe JM, Tallman MS, Zhang MJ, Wang HL, Saber W, de Lima M, et al. · Blood Advances 2019What this evidence cannot show
Registry, observational, with GVHD as a time-dependent exposure - patients who relapse or die early have less opportunity to develop GVHD, which biases the relapse-protection estimate in the direction the paper reports. Conditioning regimens, donor types and prophylaxis span decades. Neither fusion status nor the infant fraction of the 1,619 paediatric patients is reported, so the grade window described here is a paediatric ALL average and has not been shown to hold in infants or in KMT2A-rearranged disease.
Evidence that argues against this (1)
- The grade window is derived from a population in which the graft-versus-leukaemia effect is demonstrated. In KMT2A-rearranged infant leukaemia it is not: the only mechanistic study in this subtype reports TRAIL resistance and low FasL sensitivity and states that a GVL effect 'has not been clearly evidenced' in these leukaemias. If the relapse-protection half of the window does not operate, the same GVHD grade would deliver only its non-relapse mortality, and the optimal target would be no GVHD at all. Inukai T et al. 2006
- E2strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.paper
In 252 children with ALL or AML diagnosed before 1 year of age receiving single-unit unrelated cord blood transplantation after myeloablative conditioning, day-100 grade II-IV acute GVHD was 40% +/- 3%, 4-year chronic GVHD 13% +/- 2%, 1-year transplant-related mortality 23% +/- 3% and 4-year leukaemia-free survival 50% +/- 3%.
Unrelated Cord Blood Transplantation for Acute Leukemia Diagnosed in the First Year of Life: Outcomes and Risk Factor Analysis. · Ruggeri A, Volt F, Locatelli F, Michel G, Diaz de Heredia C, Abecasis M, Zecca M, Vora A, et al. · Biology of Blood and Marrow Transplantation 2017What this evidence cannot show
Cord blood only, single unit, with one HLA disparity in 57% - a graft type whose chronic GVHD risk is intrinsically low, so the 13% figure is a property of the graft source as much as of the recipients' age and cannot be generalised to the unmanipulated marrow that Interfant-06 prefers. Retrospective registry data spanning 1996-2012; transplant before 2004 was independently associated with worse outcome. Median age at transplant was 1.1 years, so a substantial fraction were beyond infancy by the day of transplant. Fusion status is not reported.
Evidence that argues against this (1)
- These figures cannot be read as infant GVHD incidence because there is no age comparator in the study and the graft source drives the chronic GVHD rate. An older single-centre infant series using marrow and mismatched related or unrelated donors reported acute GVHD in 20 of 40 and chronic GVHD in 7 of 40, which are different numbers from a different graft and donor mix in the same age group. Sanders JE et al. 2005
- E3established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.book
Interfant-06 GVHD prophylaxis is ciclosporin A alone for matched sibling donors (1.5 mg/kg i.v. twice daily from day -1, switching to 3 mg/kg/day orally, with blood levels measured and dose adjusted, tapered from day +60 by about 20% of the initial dose weekly if no GVHD), and ciclosporin plus methotrexate 10 mg/m2 on days +1, +3 and +6 with leucovorin 15 mg/m2 on days +2, +4 and +7 plus ATG-Fresenius 20 mg/kg on days -3 to -1 for matched non-sibling donors; the protocol also mandates immunoglobulin substitution to a level above 500 mg/dL, aciclovir to at least day +100, co-trimoxazole from day +14 until four weeks after the end of immunosuppression and weekly CMV PCR, noting that during severe GVHD 'a long lasting severe immunodeficiency can be expected'.
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 2006What this evidence cannot show
Chapters 6.6 and 6.7 of the 2006 SKION protocol document, read on 2026-09-20. The supportive-care recommendations are explicitly prefaced by 'The guidelines for supportive care differ between centres. Here, general recommendations are given', so compliance is unknown and probably variable. The protocol states intent, not delivered care. No GVHD incidence or severity from the Interfant-06 transplanted subgroup has been published, so the effectiveness of this exact prophylaxis in this exact population is not reported anywhere.
Evidence that argues against this (1)
- The protocol monitors the wrong drug, judged against modern evidence. It mandates ciclosporin level measurement and dose adjustment while making busulfan level measurement conditional ('should be measured if possible'). The FORUM pharmacokinetic analysis subsequently showed that busulfan exposure outside 73.3-98.0 mg*h/L nearly doubled the event-free survival hazard and that overexposure specifically increased GVHD - so the unmonitored drug is itself a determinant of the complication the monitored drug is meant to prevent. Ben Hassine K et al. 2026
- E4moderate confidence 3 of 5: One strong study, or several smaller consistent ones.paper
In 85 paediatric alpha/beta T-cell-depleted haploidentical transplants, melphalan exposure above a cohort median of 6.81 mg*h/L was associated with increased acute GVHD, and engraftment syndrome was strongly associated with subsequent grade II-IV (41.4% versus 17.3%) and grade III-IV (34.5% versus 8.5%) acute GVHD, consistent with conditioning-induced mucosal injury initiating alloreactivity.
High Melphalan Exposure Increases the Risk of Graft-Versus-Host Disease in Pediatric Patients Undergoing Alpha-Beta T-Cell Depleted Haploidentical Transplantation. · Dvorak CC, Cho S, Salinas Cisneros G, Higham CS, Chu J, Winestone LE, Temple WC, Kharbanda S, et al. · Transplantation and Cellular Therapy 2025What this evidence cannot show
Eighty-five patients, three centres, retrospective median split of a model-predicted rather than measured exposure. The setting is alpha/beta T-cell-depleted haploidentical transplant, where GVHD biology is not that of an unmanipulated matched graft with ciclosporin and methotrexate, so transfer to the Interfant-06 setting is an inference. Mucositis itself was not measured as a mediator, so the mechanistic interpretation is the authors' hypothesis. Median age is not infant.
Evidence that argues against this (1)
- The same study found the strongest predictor of chronic GVHD to be prior acute GVHD grade (66.7% after grade II-IV versus 0% after grade 0-I) rather than melphalan exposure itself, and found no effect of melphalan exposure on immunologic rejection. The causal chain from conditioning exposure to chronic GVHD runs entirely through acute GVHD, and the dataset cannot separate a conditioning effect from an alloreactivity effect. Dvorak CC et al. 2025
- E5moderate confidence 3 of 5: One strong study, or several smaller consistent ones.paper
In 40 infants diagnosed before 12 months and transplanted between 1982 and 2003 with GVHD prophylaxis of methotrexate or ciclosporin (n=7) or both (n=33), 39 engrafted, 20 developed acute GVHD and 7 developed chronic GVHD.
Allogeneic hematopoietic cell transplantation for infants with acute lymphoblastic leukemia. · Sanders JE, Im HJ, Hoffmeister PA, Gooley TA, Woolfrey AE, Carpenter PA, Andrews RG, Bryant EM, et al. · Blood 2005What this evidence cannot show
Single centre, 40 patients over 21 years, with donors that were mismatched related in 16 and unrelated in 16 - a donor mix that would not be acceptable under Interfant-06's hierarchy and that raises GVHD risk substantially. The acute GVHD figure of 20 of 40 is not graded in the abstract, so it pools grades that have opposite net effects on survival. Thirty-nine of 40 were TBI-conditioned, a modality now prohibited in this age group, and conditioning intensity is itself a GVHD determinant.
Evidence that argues against this (1)
- A modern registry cohort of infants transplanted with unrelated cord blood reports 40% grade II-IV acute GVHD at day 100 and only 13% chronic GVHD at four years, which is a markedly lower chronic GVHD rate than 7 of 40. The difference is most plausibly the graft source rather than the era or the recipients, which is precisely why neither series can be used as an infant GVHD baseline. Ruggeri A et al. 2017
raising class I and class II raises the GVHD substrate as well as the GVL substrate
the same antigen-presentation increase is the substrate for graft-versus-host disease
KIR-ligand mismatch is chosen for graft-versus-leukaemia and is the axis a class I-raising drug would blunt
the same alloreactivity, graded: grade I-II acute GVHD is where the net survival benefit sits
measured infant incidences: 40% grade II-IV acute at day 100, 13% chronic at 4 years, 23% TRM at 1 year
ATG-Fresenius 20 mg/kg days -3 to -1, matched non-sibling donors only
methotrexate 10 mg/m2 on days +1/+3/+6 with leucovorin rescue - a third, immunosuppressive role for a drug the graph already holds as an antileukaemic agent
melphalan exposure above 6.81 mg*h/L associated with increased acute GVHD, plausibly via mucosal injury
busulfan overexposure above 98.0 mg*h/L associated with increased toxicity and GVHD in FORUM