Anti-thymocyte globulin - a rabbit antibody against human T cells, given to infants at a fixed mg/kg dose against a T-cell mass that nobody measures
also called ATG · anti-thymocyte globulin · antithymocyte globulin · ATG-Fresenius · ATG-Fresenius S · Grafalon (current name of ATG-Fresenius) · Thymoglobulin (rabbit anti-thymocyte globulin, different product) · ATGAM (equine, different product) · serotherapy (class)
This is an antibody made in rabbits that destroys human T cells. It is given in the three days before a transplant from anyone other than a matched brother or sister, to stop the child rejecting the graft and to reduce the donor cells attacking the child. The dose is set by the child's weight, but what it has to destroy is the child's T cells - and those do not scale with weight.
The same thing, for a scientist
ATG is polyclonal serotherapy: rabbit immunoglobulin raised against human T-lymphoblasts, given before transplant to deplete recipient T cells (preventing rejection) and, because it persists into the post-transplant period, incoming donor T cells (reducing GVHD). Interfant-06 gives ATG-Fresenius S 20 mg/kg on days -3, -2 and -1 for matched non-sibling donors only, alongside ciclosporin and short-course methotrexate; matched sibling recipients receive ciclosporin alone [E1]. The product is described in the protocol as *'obtained from rabbits immunised with human T-lymphoblasts of the Jurkat cell-line'* [E1]. The central dosing problem is structural: ATG is consumed by its target, so exposure depends on the recipient's absolute lymphocyte count, and the pre-lymphodepletion absolute lymphocyte count is an independent predictor of graft rejection in paediatric transplant (9.4% above 1.275 x10^9/L versus 1.1% below) [E2]. A weight-based dose therefore delivers a different effective exposure to two infants of the same weight with different lymphocyte counts.
Two jobs, opposite directions, one dose
ATG is given before the graft and has a long half-life, so it acts on two different cell populations in sequence:
- Before day 0 it depletes recipient T cells. That prevents immunologic rejection of the graft.
- After day 0 residual ATG depletes donor T cells as they arrive. That reduces GVHD - and reduces graft-versus-leukaemia by the same action.
The balance between those two effects is set by how much ATG is still circulating when the graft goes in, which depends on how much was consumed by the recipient's own lymphocytes. Too much residual ATG and the graft is disarmed; too little and it is rejected or attacks. A fixed 20 mg/kg over three days does not control that balance, and Interfant-06 specifies no monitoring 1.
This graph records the corresponding concern for fludarabine under drug-fludarabine: the conditioning regimen contains two agents whose job is to suppress T cells, one dosed per square metre and one per kilogram, in a disease where the antileukaemic benefit of donor T cells is unproven.
Who gets it, and the asymmetry that creates
Interfant-06 gives ATG only to recipients of a matched non-sibling donor (MD) 1. A child with an HLA-identical sibling receives ciclosporin alone.
So two infants on the same protocol, conditioned identically with busulfan, cyclophosphamide and melphalan, receive different degrees of T-cell depletion purely on the basis of whether they happen to have a matched sibling. That is a clinically reasonable rule - unrelated grafts carry more alloreactive risk - and it means donor availability, a fact about the family, becomes a determinant of how much graft-versus-leukaemia the child can have. Any analysis comparing outcomes by donor type in this trial is confounded with serotherapy exposure.
The protocol also flags an acute hazard: 'Emergency medicines need to be ready for immediate intervention, and frequent examination of the vital signs is required' during ATG administration, reflecting cytokine-release and anaphylactoid reactions to a foreign species' immunoglobulin 1.
Why a weight-based dose is the wrong shape for this drug
ATG is a target-mediated drug: it binds and is consumed by the T cells it destroys. Its effective exposure therefore scales with the mismatch between dose and lymphocyte mass, not with body weight.
The clinical consequence is visible in the rejection data. In 410 paediatric allogeneic transplants, an absolute lymphocyte count at or above 1.275 x10^9/L before lymphodepletion carried a graft-rejection rate of 9.4% against 1.1% below that threshold - one of five independent predictors of rejection 2. A high lymphocyte count consumes serotherapy and lymphodepleting drug, leaving less to complete the job.
Infants are the extreme case. Absolute lymphocyte counts in the first year of life are physiologically two to three times adult values, and the lymphocyte fraction of the white count is at its lifetime maximum. A weight-based ATG dose in an infant is therefore being asked to deplete an unusually large lymphocyte pool per kilogram - and this beat found no study of ATG exposure, dosing or lymphocyte-count-adjusted dosing restricted to infants.
| target | Recipient and incoming donor T lymphocytes - polyclonal rabbit immunoglobulin against multiple human T-cell surface antigens. No single-molecule target node exists in this graph; this is a polyclonal biological, not a receptor-directed agent. |
|---|---|
| modality | polyclonal xenogeneic (rabbit) immunoglobulin; serotherapy |
| role in this disease | GVHD prophylaxis and rejection prevention in Interfant-06 high-risk infants receiving a matched NON-SIBLING donor graft; not given for matched sibling donors |
| interfant06 dose | ATG-Fresenius S 20 mg/kg i.v. on days -3, -2 and -1 |
| product identity trap | ATG-Fresenius (now Grafalon) is raised against human T-lymphoblasts of the Jurkat cell line; Thymoglobulin is raised against thymocytes; ATGAM is equine. The three are different products at different mg/kg doses and a dose quoted for one is wrong for another. Interfant-06 specifies ATG-Fresenius. |
| dosing problem | ATG is target-mediated and consumed by the T cells it destroys, so effective exposure depends on recipient lymphocyte mass rather than body weight; infants have physiologically high absolute lymphocyte counts, and no infant-restricted ATG exposure or dosing study was found by this beat |
| supporting observation | pre-lymphodepletion absolute lymphocyte count >= 1.275 x10^9/L is an independent predictor of graft rejection in paediatric HCT (9.4% vs 1.1%) |
| acute hazard | cytokine release and anaphylactoid reaction to xenogeneic immunoglobulin; the protocol requires emergency medicines ready and frequent vital-sign examination |
| confounding note | Because ATG is given only to non-sibling recipients, any Interfant-06 analysis comparing outcomes by donor type is confounded with serotherapy exposure. |
key ae
[ "infusion reactions and cytokine release", "serum sickness", "profound and prolonged T-cell depletion with delayed immune reconstitution", "EBV reactivation and post-transplant lymphoproliferative disease", "increased viral reactivation generally (CMV, adenovirus)", "reduced graft-versus-leukaemia by the same mechanism that reduces GVHD" ]
resistance
[ "not applicable as a tumour-cell resistance mechanism; the relevant failure mode is inadequate exposure relative to recipient lymphocyte mass, producing graft rejection", "anti-rabbit antibodies on repeat exposure" ]
- The drug is given to reduce GVHD and reduces graft-versus-leukaemia by the identical mechanism. In a disease where the antileukaemic benefit of donor T cells has never been demonstrated, the trade is being made without either side of it quantified.
- A weight-based dose for a target-mediated drug is the wrong shape. ATG is consumed by the cells it kills, so the variable that matters is lymphocyte mass, and infants carry the highest lymphocyte counts of any age group per kilogram.
- Residual ATG disarms the incoming graft. Because it is given on days -3 to -1 and has a long half-life, the same dose that prevents rejection before day 0 depletes donor T cells after it, and nothing in the protocol controls which effect dominates.
- Profound T-cell depletion in an infant invites EBV-driven lymphoproliferative disease in a patient whose EBV status is usually naive, so the reactivation model that governs adult risk does not apply and primary infection under full immunosuppression is the real scenario.
- Three different rabbit or equine products share the abbreviation ATG at doses differing several-fold. A protocol that names the abbreviation without the product, or a reader who transfers a dose between them, produces an error of the right shape and the wrong magnitude.
- Serotherapy exposure tracks family structure. Whether an infant has a matched sibling determines whether they receive 60 mg/kg of a T-cell-depleting antibody, which makes an accident of birth a determinant of immunological outcome.
- Should ATG be dosed to absolute lymphocyte count rather than to weight in infants, whose lymphocyte counts are at their lifetime maximum? Exposure-guided ATG dosing exists in adult and paediatric practice and this beat found no infant-restricted study of it.
- How much of the Interfant-06 donor-type outcome difference is serotherapy rather than HLA matching? ATG is given only to non-sibling recipients, so the two are perfectly confounded within that protocol.
- In a disease where graft-versus-leukaemia is unproven, does ATG's suppression of donor T cells cost anything at all - or is the GVHD reduction obtained for free? This is the cleanest test available of whether GVL operates in KMT2A-rearranged infant leukaemia.
- Does residual ATG at the time of graft infusion differ systematically in infants, and would measuring it change the balance between rejection and graft disarmament?
- E1established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.book
Interfant-06 gives ATG-Fresenius S 20 mg/kg intravenously on days -3, -2 and -1 as part of GVHD prophylaxis for matched non-sibling donors only, alongside ciclosporin and methotrexate; matched sibling recipients receive ciclosporin alone; the product is rabbit immunoglobulin raised against human T-lymphoblasts of the Jurkat cell line; and emergency medicines must be ready with frequent vital-sign examination during administration.
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 6.5 and 6.6 of the 2006 SKION protocol document, read on 2026-09-20. ATG-Fresenius (now marketed as Grafalon) is a DIFFERENT product from Thymoglobulin, which is also rabbit-derived but raised against thymocytes rather than a T-lymphoblast cell line and is dosed at roughly a fifth of the mg/kg - the two are not interchangeable and a dose quoted for one is wrong for the other. The protocol states intent, not practice, and no report of the Interfant-06 transplanted subgroup describes serotherapy actually delivered or its complications.
Evidence that argues against this (1)
- Serotherapy is not universally judged beneficial in paediatric leukaemia transplant. A retrospective comparison in childhood leukaemia found that a serotherapy-free regimen improved non-relapse mortality and immune recovery among recipients of alpha/beta T-cell-depleted haploidentical grafts, which is a different graft platform but establishes that omitting serotherapy is a live and defended option rather than an obvious deficiency. Shelikhova L et al. 2021
- E2strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.paper
In 410 paediatric and young-adult allogeneic transplants, an absolute lymphocyte count at or above 1.275 x10^9/L obtained before the start of lymphodepletion was independently associated with immunologic graft rejection (9.4% versus 1.1% below that threshold, p<.001), alongside female-donor-to-male-recipient pairing, HLA mismatch, CMV seropositivity and body-surface-area rather than model-based fludarabine dosing.
Fludarabine Pharmacokinetic Model-Based Dosing Minimizes the Risk of Graft Rejection in Pediatric Patients Undergoing Allogeneic Hematopoietic Cell Transplantation. · Dvorak CC, Waters E, Temple WC, Salinas Cisneros G, Chu J, Winestone LE, Higham CS, Kharbanda S, et al. · Transplantation and Cellular Therapy 2026What this evidence cannot show
This study is about FLUDARABINE dosing, not about ATG; the lymphocyte-count finding is cited here because it establishes that recipient lymphocyte burden determines lymphodepletion adequacy, which is the mechanism by which a weight-based serotherapy dose would misfire. It does not measure ATG exposure. Retrospective, single institution, 2008-2024, with 24 rejection events supporting a five-variable multivariable model. The 1.275 x10^9/L cut-point was derived in the same dataset. Median age 8 years; infants are within the range (0.1-26.5 years) but are not analysed separately.
Evidence that argues against this (1)
- The inference that infants need lymphocyte-adjusted ATG is this beat's reasoning from two separate facts - that ATG is target-mediated and that infant lymphocyte counts are physiologically high - and is not a finding of any study retrieved here. No study of ATG exposure or lymphocyte-count-adjusted ATG dosing restricted to infants was found, so this is an unmeasured hypothesis and is labelled as one. Naeije L et al. 2020
- E3moderate confidence 3 of 5: One strong study, or several smaller consistent ones.paper
Anti-thymocyte globulin prophylaxis reduces the onset of chronic graft-versus-host disease, with a mechanism traced in a randomised-trial cohort to a decrease in naive T-helper cells.
Anti-Thymocyte Globulin Prophylaxis Induces a Decrease in Naive Th Cells to Inhibit the Onset of Chronic Graft-versus-Host Disease: Results from the Canadian Bone Marrow Transplant Group (CBMTG) 0801 Study. · Naeije L, Kariminia A, Abdossamadi S, Azadpour S, Subrt P, Kuzeljevic B, Irvine M, Chan G, et al. · Biology of Blood and Marrow Transplantation 2020What this evidence cannot show
A correlative immunology study within the CBMTG 0801 randomised trial, which was conducted in adults; the abstract was not retrieved by this beat beyond its title, so the effect size, the ATG product and the dose are not asserted here and only the direction of the finding and its stated mechanism are claimed. Nothing in it is paediatric and nothing is infant. It is cited as the reason ATG is used, not as evidence that the Interfant-06 dose is right.
Evidence that argues against this (1)
- Reducing chronic GVHD by depleting naive T-helper cells is the same action that would reduce graft-versus-leukaemia, and in a disease where the antileukaemic benefit of donor T cells has never been demonstrated the trade is being made blind. A serotherapy-free platform in childhood leukaemia reported improved non-relapse mortality and immune recovery, which is the opposite conclusion drawn on different endpoints in a different graft setting. Shelikhova L et al. 2021
ATG-Fresenius 20 mg/kg days -3 to -1, matched non-sibling donors only
- part of →Interfant-06 — 651 infants, and the definitive answer that myeloid-style consolidation does not help
days -3 to -1 for matched non-sibling donors; not given to matched sibling recipients
depletes donor T cells as they arrive, reducing GVHD and graft-versus-leukaemia by the same action
two T-cell-suppressing agents in one regimen, one dosed per kilogram and one per square metre, neither monitored in Interfant-06