Donor immune cells vs. recipient tissue after allogeneic stem cell transplant
Graft-versus-host disease begins before any donor cell has arrived. The conditioning regimen — chemotherapy, total body irradiation, or both — that ablates the recipient's marrow to make room for the graft also inflicts collateral damage on epithelial barriers, particularly the gastrointestinal mucosa, skin, and liver. This is "Step One" of the classic three-step Ferrara model of acute GVHD pathophysiology.
James Ferrara's widely used framework divides acute GVHD pathophysiology into three sequential, mechanistically distinct steps:
Step 1 — Tissue damage from the underlying disease and conditioning regimen activates host APCs and triggers a pro-inflammatory "cytokine storm."
Step 2 — Donor T-cell activation: mature donor T cells within the graft recognize host alloantigens presented by activated APCs, undergo costimulation, and proliferate.
Step 3 — Effector phase: activated donor T cells (CTLs, Th1, Th17) and soluble inflammatory mediators (TNF-α, IFN-γ, cellular Fas/FasL and perforin/granzyme pathways) converge on target organs, producing the clinical syndrome.
This conceptual model, though a simplification of overlapping and continuous biology, remains the dominant teaching framework because each step maps to a distinct therapeutic opportunity — reducing conditioning intensity, blocking costimulation, or neutralizing effector cytokines.
Conditioning serves two purposes: eradicating residual malignancy and immunosuppressing the recipient enough to permit donor stem cell engraftment. Two broad strategies exist:
• Myeloablative conditioning (MAC): high-dose chemotherapy (e.g., cyclophosphamide, busulfan) with or without total body irradiation (TBI, typically 12–14 Gy). Produces profound marrow ablation and strong anti-leukemic effect, but causes extensive mucosal and endothelial injury — the single strongest driver of Step 1 tissue damage.
• Reduced-intensity conditioning (RIC) / non-myeloablative regimens: lower doses (e.g., fludarabine-based) that rely more on the graft-versus-leukemia (GVL) effect than on direct cytotoxicity for disease eradication. Better tolerated in older or comorbid patients, and associated with less severe Step 1 injury and somewhat lower acute GVHD severity — though relapse risk can be higher.
The intensity of conditioning correlates directly with the magnitude of tissue damage, DAMP release, and downstream GVHD severity — a dose-response relationship that is central to modern regimen selection.
Regimen-related toxicity to the gut is thought to be the single most important initiating event in acute GVHD: damaged intestinal epithelium allows translocation of bacterial lipopolysaccharide (LPS) into the portal and systemic circulation, amplifying the cytokine storm through Toll-like receptor 4 signaling on host APCs — a "second hit" layered on top of direct radiation/chemotherapy injury.
Dying and stressed recipient cells release damage-associated molecular patterns (DAMPs) — ATP, uric acid, high-mobility group box 1 (HMGB1), heat-shock proteins, and mitochondrial DNA — which engage pattern-recognition receptors (Toll-like receptors, NLRP3 inflammasome) on host antigen-presenting cells.
This DAMP signaling, reinforced by translocated microbial LPS (a pathogen-associated molecular pattern, or PAMP), triggers:
• Upregulation of MHC class I and II on host APCs • Increased expression of costimulatory molecules (CD80/CD86, CD40) • Secretion of TNF-α, IL-1β, and IL-6 — the classic "cytokine storm" cytokines • Recruitment and further activation of dendritic cells, macrophages, and Langerhans cells in skin and gut
The net effect is a fully "licensed," highly immunostimulatory host APC compartment — primed and waiting to present alloantigen to the donor T cells that arrive with the graft days later. Without this priming step, donor T cells encounter far less costimulatory signal and GVHD is substantially attenuated, which is why reducing conditioning intensity (or damage to the gut specifically) is a validated GVHD-mitigation strategy.
Even when a donor is "HLA-matched," the graft is never truly identical to the recipient. Mature donor T cells infused with the stem cell product survey activated host APCs and recognize disparities in major or, far more commonly, minor histocompatibility antigens as foreign — triggering the central alloimmune activation event of GVHD.
HLA matching (typically at the 8/8 or 10/10 allele level across HLA-A, -B, -C, -DRB1, and sometimes -DQB1) is the primary criterion for selecting a donor, and it dramatically reduces — but never eliminates — GVHD risk.
Even in a fully HLA-matched sibling or unrelated donor pair, GVHD still occurs in 30–50% of transplants because of minor histocompatibility antigens (miHAs): peptides derived from polymorphic, non-HLA genes that differ between donor and recipient and are presented on the recipient's own HLA molecules. Classic examples include HA-1, HA-2, and HY antigens (encoded on the Y chromosome, relevant in male-recipient / female-donor pairs).
Because miHAs are ordinary self-proteins with amino-acid polymorphisms rather than foreign pathogens, donor T cells that recognize them have not been centrally deleted in the donor's thymus — they simply never encountered the recipient's variant peptide before. This is why "HLA-matched" transplants are never truly "antigen-matched," and why sex-mismatched transplants (male recipient, female donor) carry a measurably higher GVHD risk due to HY-antigen disparity.
Because miHA disparity is unavoidable even in perfectly HLA-matched pairs, some degree of donor-anti-host alloreactivity is essentially built into allogeneic transplantation — GVHD risk can be reduced and managed, but with current technology it cannot be fully eliminated without also eliminating the beneficial graft-versus-leukemia effect that relies on the same alloreactive T-cell repertoire.
Full donor T-cell activation requires two coordinated signals delivered at the immune synapse formed with an activated host APC:
• Signal 1 — antigen specificity: the donor T-cell receptor (TCR) engages the host peptide-MHC complex. In GVHD this peptide is a minor (or, in mismatched transplants, major) histocompatibility antigen rather than a pathogen-derived epitope.
• Signal 2 — costimulation: CD28 on the T cell binds CD80/CD86 on the APC; CD40L (CD154) on the T cell engages CD40 on the APC. Because the conditioning-primed host APC is highly "licensed" (Stage 1), costimulatory ligand density is abnormally high, producing an unusually strong activating signal.
Signal 1 without signal 2 typically induces T-cell anergy rather than activation — a principle exploited therapeutically by costimulation-blockade strategies (e.g., abatacept, a CTLA-4-Ig fusion protein) now used in some GVHD prophylaxis regimens to blunt this step selectively.
Activated donor T cells undergo rapid IL-2-driven clonal proliferation, expanding alloreactive clones ten- to one-hundred-fold within the first week after infusion. Differentiation is shaped by the local cytokine milieu (itself a product of Step 1 inflammation):
• Th1 cells: driven by IL-12, produce IFN-γ and TNF-α; classically associated with acute GVHD and macrophage activation • Th17 cells: driven by IL-6 and TGF-β, produce IL-17; implicated particularly in gut and lung GVHD and steroid-refractory disease • Cytotoxic T lymphocytes (CTLs, CD8+): kill target cells directly via perforin/granzyme exocytosis and Fas–FasL engagement • Regulatory T cells (Tregs): a counter-regulatory population that, when insufficient in number or function relative to effector T cells, fails to restrain the alloresponse — Treg deficiency/dysfunction is a major research target for GVHD prevention
The balance between effector subsets and regulatory T cells — rather than any single pathway — ultimately determines whether clinically significant GVHD develops.
In the effector phase, activated donor T cells leave secondary lymphoid organs, traffic through the bloodstream, and home preferentially to skin, gastrointestinal epithelium, and liver — the three classic GVHD target organs. Direct cytotoxicity and cytokine-mediated inflammation combine to produce organ-specific, cumulative tissue injury.
Why do donor T cells preferentially attack skin, gut, and liver rather than the whole body uniformly? Organ tropism is directed by chemokine receptor and integrin expression acquired during activation:
• Gut homing: α4β7 integrin binds MAdCAM-1 on intestinal endothelium; CCR9 responds to the gut-specific chemokine CCL25 • Skin homing: cutaneous lymphocyte antigen (CLA) binds E-selectin on dermal endothelium; CCR4 and CCR10 respond to skin-derived chemokines • Liver: less tissue-specific homing machinery is defined, but hepatic sinusoidal endothelium is highly permissive to activated lymphocyte trafficking and antigen presentation via resident Kupffer cells and bile duct epithelium expressing MHC
These three organs share a common feature that likely explains their selective vulnerability: all are barrier epithelia in continuous contact with microbial antigens or metabolic stress, making them sites of high baseline APC activity and adhesion-molecule expression — exactly the substrate alloreactive T cells are primed to exploit.
Once in the target organ, effector donor T cells damage host tissue through two principal cytotoxic pathways:
• Perforin/granzyme pathway: CD8+ CTLs release perforin, which polymerizes to form pores in the target cell membrane, allowing granzyme B to enter and trigger caspase-dependent apoptosis • Fas–FasL pathway: FasL on the effector T cell engages Fas (CD95) on target epithelial cells, activating the extrinsic apoptotic cascade — particularly important in gut crypt epithelial injury
These are amplified by soluble inflammatory mediators — TNF-α and IFN-γ — secreted by Th1 cells and activated macrophages, which upregulate MHC and adhesion molecules on target tissue (further increasing visibility to donor T cells), induce direct epithelial apoptosis, and activate a feed-forward inflammatory loop sometimes termed the "cytokine cascade," amplifying tissue injury well beyond the numbers of T cells physically present.
GI GVHD is disproportionately lethal because gut crypt stem cells are a preferred cytotoxic target: their loss removes the entire regenerative capacity of the epithelium, producing mucosal denudation, loss of barrier integrity, further microbial/LPS translocation, and a self-amplifying inflammatory loop that is difficult to interrupt once established — this is why severe GI GVHD is the strongest single predictor of non-relapse mortality after transplant.
Each target organ manifests a recognizable pattern of injury:
• Skin: begins as a maculopapular exanthem, often first on palms, soles, ears, and shoulders; histology shows basal keratinocyte vacuolization and individual keratinocyte apoptosis ("satellite cell necrosis" — lymphocytes surrounding dying keratinocytes). Severe cases progress to bullae and desquamation resembling toxic epidermal necrolysis.
• Gastrointestinal tract: crypt epithelial cell apoptosis progressing to crypt abscess and complete crypt destruction/mucosal denudation; clinically produces high-volume secretory diarrhea, abdominal cramping, nausea/vomiting (upper GI GVHD), GI bleeding, and in severe cases ileus.
• Liver: lymphocytic infiltration and endothelialitis targeting small bile duct epithelium, producing a cholestatic (not hepatocellular) pattern of injury — disproportionate rise in bilirubin and alkaline phosphatase relative to transaminases, reflecting bile duct rather than hepatocyte destruction.
Once organ injury is clinically apparent, transplant physicians must quantify its severity in a standardized, reproducible way. The Glucksberg criteria — refined by the MAGIC (Mount Sinai Acute GVHD International Consortium) consensus — stage each of the three target organs from 0 to 4, then combine the organ stages into a single overall grade (I–IV) that carries strong prognostic weight.
Acute GVHD severity varies enormously — from a faint, self-limited rash to fulminant multi-organ failure — and clinical trials, treatment decisions (e.g., when to escalate beyond first-line steroids), and prognostic counseling all depend on a reproducible severity scale. The Glucksberg system (1974), still in use in modified form today, stages skin, liver, and gut independently using objective, quantifiable criteria (percent body-surface-area rash, serum bilirubin, and daily stool/diarrhea volume) rather than subjective impression, then combines them algorithmically into an overall grade.
The MAGIC consortium criteria refined the original thresholds — particularly for gut staging in adults versus pediatric patients — and are the version most widely used in contemporary clinical trials and practice.
Each organ is staged 0 (no involvement) through 4 (most severe) using organ-specific criteria (see table). The overall grade is then determined by a defined combination rule, not a simple average:
• Grade 0: no stage in any organ • Grade I: skin stage 1–2 only, no gut or liver involvement — mild disease, often outpatient-manageable • Grade II: skin stage 3, and/or gut stage 1, and/or liver stage 1 — mild-to-moderate multi-organ involvement • Grade III: gut stage 2–3 and/or liver stage 2–3 — significant organ dysfunction, typically requires inpatient systemic treatment • Grade IV: skin stage 4 and/or gut stage 4 and/or liver stage 4 — life-threatening, often with organ failure
The rule reflects an important clinical truth: gut and liver involvement drive overall severity and prognosis far more than skin involvement alone, because they reflect deeper visceral organ dysfunction rather than a cutaneous inflammatory reaction.
Overall grade is one of the strongest independent predictors of outcome after allogeneic transplant: patients with grade I GVHD have a favorable prognosis similar to patients without GVHD, while grade III–IV disease carries 1-year non-relapse mortality in the range of 40–70% even with aggressive treatment — which is why early, accurate staging directly guides how aggressively immunosuppression is escalated.
GVHD was historically split into "acute" (before day 100 post-transplant) and "chronic" (after day 100) purely by time of onset. The NIH 2014 consensus criteria replaced this arbitrary cutoff with a feature-based classification: acute GVHD is defined by its characteristic maculopapular rash, secretory diarrhea, and cholestatic hepatitis (as staged above), while chronic GVHD is a distinct, more fibrotic, autoimmune-like syndrome (affecting skin, mouth, eyes, lungs, joints, and other organs with sclerodermatous or lichenoid features) that can occur at any time, including well before day 100, and shares more biology with autoimmune disease than with the acute cytotoxic effector attack described in this simulator.
A related, overlapping category — "late acute" or "persistent/recurrent/late-onset acute GVHD" — captures cases with classic acute-type organ involvement occurring after day 100, underscoring that the biology (not the calendar) is now considered definitive.
| Product | Indication | Trial Design | Key Result |
|---|---|---|---|
Because donor T-cell alloreactivity is essentially inevitable to some degree, GVHD management operates on two fronts: prophylaxis, given to every allogeneic transplant recipient from the outset to blunt Step 2/3 activation before it starts, and treatment, escalated once clinically significant GVHD is established. Every intervention must be weighed against the loss of graft-versus-leukemia (GVL) activity that the same donor T cells provide.
Essentially all allogeneic transplant recipients receive prophylactic immunosuppression starting around the time of stem cell infusion, aimed squarely at blunting donor T-cell activation (Step 2) before clinical disease can develop:
• Calcineurin inhibitors (CNIs) — tacrolimus or cyclosporine: block calcineurin-dependent dephosphorylation of NFAT, preventing IL-2 gene transcription and T-cell proliferation. Almost universally combined with a second agent.
• Methotrexate (short course, days +1, +3, +6, +11) or mycophenolate mofetil (MMF): antiproliferative agents that further limit the burst of donor T-cell clonal expansion during the highest-risk early post-transplant window.
• Post-transplant cyclophosphamide (PTCy): given on days +3 and +4 after infusion, selectively eliminates rapidly proliferating alloreactive donor T-cell clones (which are dividing fastest at that time) while sparing resting/regulatory T cells. Originally developed for haploidentical (half-matched) transplants, PTCy-based prophylaxis has now become standard even in matched transplants because it markedly reduces both acute and chronic GVHD without eliminating GVL benefit as completely as broader immunosuppression.
• T-cell depletion (ex vivo, e.g., CD34 selection, or in vivo, e.g., antithymocyte globulin/alemtuzumab): physically removes or depletes donor T cells from the graft before infusion — highly effective at preventing GVHD but at the cost of higher relapse and infection risk, since the same T cells provide both GVHD risk and GVL benefit.
Once acute GVHD reaches a clinically significant grade (typically grade II or higher, or grade I skin disease that progresses), treatment escalates beyond prophylaxis:
• First-line: systemic corticosteroids, typically methylprednisolone 1–2 mg/kg/day, remain the standard of care nearly 50 years after their introduction — they broadly suppress T-cell activation, cytokine production, and downstream inflammatory signaling, inducing lymphocyte apoptosis and dampening antigen-presenting cell function.
• Response is assessed around day 3–7 and again at day 28; roughly half of patients have an inadequate response and are termed "steroid-refractory" or "steroid-dependent" (relapsing on taper) — a high-risk state historically associated with poor outcomes.
• Second-line / steroid-refractory therapy: ruxolitinib, a JAK1/2 inhibitor that blocks signaling downstream of multiple GVHD-relevant cytokine receptors (including IFN-γ and IL-6 pathways), is now the preferred, FDA-approved second-line agent based on the REACH2 (acute) and REACH3 (chronic) randomized trials, which showed superior overall response rates versus best-available-therapy. Other second-line options include extracorporeal photopheresis, mesenchymal stromal cell infusion, and additional biologic/targeted agents (e.g., anti-TNF, anti-IL-6, anti-integrin therapies) selected based on organ involvement and institutional experience.
Immunosuppression in this simulator dampens donor T-cell activity by binding and inactivating effector T cells — visually slowing their migration and halting further attack on organ tissue — mirroring how calcineurin inhibitors, steroids, and JAK inhibitors act mechanistically: not by killing the graft, but by silencing the alloreactive signaling that drives its attack on host tissue.
The same donor T cells that attack healthy recipient skin, gut, and liver also recognize and kill residual host leukemia or lymphoma cells — the graft-versus-leukemia (GVL) effect, first inferred from the observation that transplant recipients who developed GVHD had lower relapse rates than those who did not, and that T-cell-depleted grafts (which prevent GVHD most effectively) also carry the highest relapse rates.
This creates the central therapeutic tension of allogeneic transplantation: immunosuppression potent enough to fully abolish GVHD risk would also abolish much of the anti-leukemic benefit that makes allogeneic (rather than autologous) transplant preferable for many malignancies in the first place. Clinical strategies attempt to separate the two effects:
• Dose and timing optimization (e.g., PTCy, which spares regulatory and later-arriving alloreactive clones) aims to preserve more GVL activity than broad, sustained immunosuppression • Donor lymphocyte infusion (DLI) — deliberately infusing additional donor T cells after transplant to boost GVL effect in patients with relapsing or minimal residual disease, accepting some added GVHD risk in exchange • Ongoing research into separating the T-cell receptor repertoires responsible for GVHD versus GVL (since some target antigens are shared between host tissue and leukemic cells, while others — leukemia-restricted antigens — are not) aims to eventually enable T-cell products with anti-leukemic activity but minimal GVHD potential
No current prophylaxis or treatment regimen fully separates these two effects — GVHD management therefore remains a continuous balancing act between under-treatment (relapse, and GVHD-driven organ damage) and over-treatment (infection, toxicity, and loss of the curative GVL benefit).