Three-organ staging — skin, liver, gastrointestinal tract — combined by consensus rules into an overall clinical grade (I–IV) that guides immunosuppressive treatment intensity after allogeneic transplant
The skin is the most commonly and earliest affected organ in acute graft-versus-host disease (GVHD), typically appearing within the first 2–6 weeks after allogeneic hematopoietic stem cell transplant. Donor T-cells recognize recipient epidermal keratinocytes as foreign, driving a cytotoxic infiltrate that produces a characteristic pruritic, erythematous, maculopapular rash. Unlike liver and gut staging, which rely on laboratory or volumetric thresholds, skin staging is purely visual and quantitative: what percentage of the total body surface area (BSA) is involved?
Acute cutaneous GVHD classically begins on the palms and soles, the nape of the neck, the ears, and the upper trunk, before spreading centrifugally to involve the rest of the body. The lesions are erythematous macules and papules that can coalesce into confluent plaques. Pruritus and mild tenderness are common; a burning sensation on palms/soles is an early, easily-missed clue.
Histopathology (skin biopsy) shows a spectrum from grade 1 (vacuolar interface change of the basal keratinocyte layer) to grade 4 (complete epidermal separation from the dermis — the histologic correlate of toxic epidermal necrolysis-like disease). Clinical staging, however, is based on the bedside estimate of BSA involvement using the "rule of nines" or a body-surface diagram, not the biopsy grade — the two are related but not interchangeable.
Stage 1 — Maculopapular rash over <25% of BSA, without bullae. Often localized to palms, soles, ears, and shoulders. Frequently the sole clinical finding in otherwise mild acute GVHD.
Stage 2 — Maculopapular rash over 25–50% of BSA. The rash becomes confluent over larger regions of the trunk and extremities; pruritus intensifies.
Stage 3 — Generalized erythroderma, involving >50% of BSA. The skin takes on a diffusely red, thickened appearance; this stage carries meaningfully higher risk of progression to bullous disease.
Stage 4 — Generalized erythroderma with bullae formation and desquamation, resembling a Stevens-Johnson syndrome/toxic epidermal necrolysis phenotype. This is a dermatologic emergency: skin barrier failure predisposes to infection, fluid loss, and thermoregulatory failure, and mandates urgent escalation regardless of the other two organs.
Skin stage alone can drive a patient to overall Grade IV: bullous, desquamating GVHD (stage 4 skin) is treated as life-threatening even if liver and gut are completely spared, because of the barrier-failure risk it carries.
Hepatic acute GVHD targets the small interlobular bile duct epithelium, producing a cholestatic — rather than hepatocellular — pattern of injury. Alkaline phosphatase and bilirubin rise disproportionately to transaminases. Because bilirubin is a single, reproducible, frequently-drawn laboratory value, it was adopted as the sole staging criterion for hepatic GVHD, staged 1 through 4 by increasing threshold.
Donor cytotoxic T-lymphocytes home to the portal tracts and directly attack the biliary epithelium lining small interlobular bile ducts. Histology shows lymphocytic infiltration, bile duct epithelial apoptosis, and duct destruction — a pattern reminiscent of, and sometimes histologically indistinguishable from, other causes of duct injury such as drug toxicity, viral hepatitis, or sinusoidal obstruction syndrome (SOS/VOD). This overlap is the central diagnostic challenge in hepatic GVHD: rising bilirubin has a broad differential in the peri-transplant period.
Because duct injury impairs bile excretion, conjugated (direct) bilirubin rises first and predominates; alkaline phosphatase and gamma-glutamyl transferase (GGT) climb in parallel, while transaminases (AST/ALT) remain relatively modest unless there is concurrent hepatocellular insult. Total serum bilirubin was chosen as the staging variable specifically because it is ubiquitously measured, standardized across laboratories, and tracks disease trajectory reliably.
Stage 1 — Bilirubin 2–3 mg/dL (34–51 µmol/L). Mild, often asymptomatic cholestasis picked up on routine labs.
Stage 2 — Bilirubin 3.1–6 mg/dL (51–103 µmol/L). Mild clinical jaundice may become visible; pruritus from bile salt accumulation can begin.
Stage 3 — Bilirubin 6.1–15 mg/dL (103–257 µmol/L). Overt jaundice, dark urine, pale stool; risk of coagulopathy from impaired vitamin K-dependent factor synthesis increases.
Stage 4 — Bilirubin >15 mg/dL (>257 µmol/L). Severe hyperbilirubinemia; must be distinguished urgently from sinusoidal obstruction syndrome (elevated bilirubin plus weight gain, painful hepatomegaly, and ascites) since management diverges substantially.
Bilirubin-based staging can be confounded by sepsis, total parenteral nutrition cholestasis, drug-induced liver injury (azoles, sulfa drugs), and sinusoidal obstruction syndrome — all common in the same post-transplant window. A liver biopsy or transjugular biopsy with hepatic venous pressure gradient is sometimes needed when the picture is ambiguous.
The gastrointestinal tract — particularly the crypts of the small and large intestine — is targeted by alloreactive donor T-cells that trigger a self-amplifying cytokine storm (the "cytokine storm" model of GVHD pathophysiology: tissue damage → damage-associated molecular patterns → dendritic cell activation → further T-cell priming). Crypt cell apoptosis and destruction impair fluid absorption, producing profuse secretory diarrhea whose daily volume is used as the staging surrogate for depth of mucosal injury.
Endoscopic and histologic hallmark of gut GVHD is apoptosis of the intestinal crypt epithelium — single dying cells within crypts, progressing in severe disease to crypt abscesses, crypt dropout, and complete denudation of the mucosal surface. As crypts (the intestine's stem-cell-containing regenerative units) are destroyed, the epithelium loses its capacity to absorb fluid and electrolytes, and the result is a large-volume, watery, secretory diarrhea — sometimes exceeding several liters per day in severe cases.
Upper GI involvement (esophagus, stomach, duodenum) can occur without the classic diarrhea, presenting instead as anorexia, nausea, vomiting, and dyspepsia; consensus criteria therefore also allow persistent upper-GI symptoms with a positive biopsy to satisfy stage 1 gut involvement even when stool volume is not yet elevated.
Stage 1 — Diarrhea 500–999 mL/day (adults; weight-based in pediatrics), or persistent nausea/vomiting/anorexia with a positive GI biopsy.
Stage 2 — Diarrhea 1000–1500 mL/day. Fluid and electrolyte replacement needs increase; oral intake alone is usually insufficient.
Stage 3 — Diarrhea >1500 mL/day. Large-volume losses risk hypovolemia, acute kidney injury, and severe electrolyte derangement; intravenous repletion is mandatory.
Stage 4 — Severe abdominal pain with or without ileus, or grossly bloody stool, regardless of measured volume. This stage reflects transmural/deep mucosal injury and is a marker of the most severe gut disease, often accompanied by protein-losing enteropathy.
Diarrhea volume is a surrogate, not a direct histologic measure — profound ileus can paradoxically produce low measured stool output despite catastrophic mucosal injury, which is exactly why stage 4 is defined by ileus/bleeding rather than volume alone.
The single most misunderstood aspect of acute GVHD grading is that the overall clinical grade (I–IV) is not simply "the worst organ stage." Consensus grading systems (originally Glucksberg 1974, refined by the Keystone/IBMTR criteria and later the MAGIC algorithm) combine the three organ stages using explicit rules that weigh liver and gut involvement more heavily than skin alone, because visceral organ dysfunction carries disproportionately higher risk of non-relapse mortality than skin disease of equivalent "stage number."
A patient with skin stage 4 (bullous, desquamating rash) but no liver or gut involvement is graded very differently from a patient with liver stage 4 (bilirubin >15 mg/dL) and no skin or gut disease — yet both have a "stage 4" organ. Historically both extremes map to Grade IV because either represents life-threatening single-organ failure. But intermediate combinations diverge sharply: skin stage 3 alone (generalized erythroderma, no viscera) is typically only Grade II, while liver stage 2 alone (bilirubin 3–6 mg/dL) already reaches Grade III, because early visceral organ injury has a steeper mortality curve than skin-only disease of a nominally similar stage number.
This simulator uses an illustrative simplified combination rule for teaching purposes: any organ reaching its most severe stage (skin 4, liver ≥4, or GI ≥3) triggers overall Grade IV; liver ≥2 or GI ≥2 triggers Grade III; skin ≥3, or any liver/GI stage ≥1, triggers Grade II; isolated skin stage 1–2 alone gives Grade I. The exact numeric cut-points vary slightly between the Glucksberg, IBMTR/Keystone, and MAGIC consensus systems used in clinical trials — always defer to the specific protocol's published table in real practice.
Grade I — Skin stage 1–2 only; no liver or gut involvement. Represents mild, cutaneous-limited disease with an excellent prognosis.
Grade II — Skin stage 3, OR liver stage 1, OR gut stage 1 (in any combination with lower stages of the other organs). The presence of any visceral involvement, even mild, escalates the grade.
Grade III — Liver stage 2–3, OR gut stage 2–3 (regardless of skin stage). Meaningful visceral dysfunction defines this tier.
Grade IV — Skin stage 4, OR liver stage 4, OR gut stage 4 (life-threatening single- or multi-organ involvement), often with associated extreme decline in performance status.
Because the rule is disjunctive ("OR" across organs, escalating to the worst-triggering criterion), a patient can jump straight to Grade III or IV from a single severely affected organ even while the other two organs remain completely uninvolved — this is intentional, since visceral organ failure of any single system can be independently life-threatening.
Overall acute GVHD grade is one of the strongest single predictors of transplant-related (non-relapse) mortality and overall survival after allogeneic hematopoietic stem cell transplant. Because grade correlates so tightly with outcome, it directly determines initial treatment intensity: low-grade disease is managed conservatively to avoid over-immunosuppressing a patient who does not need it, while high-grade disease demands prompt, aggressive systemic therapy — often before all diagnostic confirmation is complete, because delay materially worsens outcomes.
Isolated, mild skin-limited disease (Grade I) is typically managed without systemic immunosuppression escalation beyond the patient's existing GVHD prophylaxis regimen. Topical corticosteroids (e.g., triamcinolone, clobetasol for thicker plaques), topical calcineurin inhibitors, and antihistamines for pruritus are first-line. The rationale is to avoid unnecessary systemic steroid exposure — with its attendant infection risk, hyperglycemia, myopathy, and osteoporosis — in a population whose baseline prognosis is already favorable.
Once any visceral organ is involved, or skin disease is extensive (Grade II), systemic corticosteroids — typically methylprednisolone or prednisone 1–2 mg/kg/day — become the standard first-line therapy, usually layered on top of the patient's baseline calcineurin-inhibitor-based prophylaxis. Roughly half of patients achieve a durable complete response to first-line steroids; response is typically assessed at day 3, day 7, and day 14, since early non-response is itself a strong predictor of the need for second-line therapy.
High-grade disease (Grade III–IV) carries substantially higher rates of primary steroid-refractoriness, defined as progression within 3–5 days of high-dose steroids, or failure to improve by day 7, or incomplete response by day 14–28. Steroid-refractory acute GVHD triggers second-line therapy — ruxolitinib (a JAK1/2 inhibitor) is now the preferred, FDA-approved second-line agent following the REACH2 trial, which showed superior overall response versus investigator's choice. Other historical second-line options include extracorporeal photopheresis, mycophenolate mofetil, anti-thymocyte globulin, and other biologics targeting IL-2, TNF-α, or integrin pathways.
Because Grade III–IV disease carries such steep mortality, many centers now escalate empirically at the first sign of visceral organ involvement rather than waiting for full staging confirmation, particularly when gut or liver stage is already high at presentation.
The REACH2 trial (Zeiser et al., NEJM 2020) established ruxolitinib as standard second-line therapy for steroid-refractory acute GVHD, roughly doubling overall response rate at day 28 compared with best available therapy (62% vs. 39%) — a direct downstream consequence of accurate, early grading driving faster escalation decisions.