Graft versus host disease, commonly called GVHD, is a condition in which transplanted immune cells from a donor attack the recipient’s own tissues. It occurs most often after an allogeneic hematopoietic stem cell transplant (sometimes called a bone marrow transplant), where the donor and recipient are different people. The transplanted immune cells, particularly T cells, recognize the recipient’s body as foreign and mount an immune assault that can damage the skin, gut, liver, lungs, eyes, and mouth. Despite decades of refinement in transplant medicine, as many as half of recipients develop clinically significant GVHD with current prevention strategies, making it one of the most common and consequential complications of this lifesaving procedure.
How the Immune Attack Begins
Before a stem cell transplant takes place, patients undergo conditioning with chemotherapy, radiation, or both. This treatment is necessary to make room for the donor’s cells and to destroy the underlying disease, but it inflicts serious collateral damage on the body’s tissues. Dying cells release molecular alarm signals, and damage to the intestinal lining allows bacteria and bacterial fragments to leak into the bloodstream. These danger signals activate the innate immune system and create an inflammatory environment that primes donor T cells to react aggressively once they arrive.
The process is often described in three overlapping phases. First, the conditioning regimen causes tissue injury and inflammation. Second, the donor’s T cells become activated when they encounter the recipient’s cells and recognize them as foreign. Third, those activated T cells and the inflammatory molecules they produce attack target organs. The release of endogenous danger signals like ATP and uric acid, along with bacterial components leaking through damaged gut barriers, directly fuels this cascade.
Acute and Chronic Forms
GVHD comes in two broad categories that differ in timing, biology, and the organs they tend to hit hardest. Acute GVHD classically appears within the first 100 days after transplant, though it can arise later in some settings. It tends to be driven by a vigorous, direct T cell attack on three main targets: skin, the gastrointestinal tract, and the liver. A skin rash is frequently the earliest sign. Gut involvement causes nausea, diarrhea, and abdominal pain, while liver involvement shows up as rising bilirubin and liver enzymes.
Chronic GVHD usually develops later and behaves more like an autoimmune disease. Its underlying biology involves a broader dysfunction of the immune system, including damage to the thymus (which normally trains immune cells to distinguish self from non-self), failure of regulatory T cells that would ordinarily keep inflammation in check, and the generation of antibodies against the recipient’s tissues. The resulting damage often includes fibrosis, the pathological thickening and scarring of tissues. Chronic GVHD can affect virtually any organ system, with the skin, eyes, and mouth among the most commonly involved sites.
Which Organs Are Affected and How
Skin involvement is a hallmark of both acute and chronic GVHD. In the acute form, it often starts as a flat red rash that can spread widely. In chronic GVHD, skin changes fall into two broad categories: nonsclerotic presentations, which resemble conditions like lichen planus with discoloration and surface changes, and sclerotic presentations, which involve thickening and hardening of the skin similar to scleroderma.
The gastrointestinal tract is one of the most dangerous sites of involvement. Acute gut GVHD causes watery or bloody diarrhea that can become life-threatening through dehydration and infection. The gut is also where much of the early inflammatory cascade begins, as damage to the intestinal barrier lets bacterial products amplify the immune response.
Liver GVHD can be tricky to pin down because its appearance overlaps with other post-transplant problems like infections and drug toxicity. In some cases, liver GVHD is histologically indistinguishable from drug-induced liver injury. Clinicians watch for a steady, silent rise in bilirubin and liver enzymes, which research has shown can precede a full-blown liver GVHD episode by about five to twelve days, depending on the marker. That early creep in lab values, even before a patient feels anything, can serve as a useful early warning.
Oral chronic GVHD deserves its own mention because it is common and significantly affects daily life. It causes inflammation, thinning, and ulceration of the tissues lining the mouth (resembling oral lichen planus), dryness from salivary gland damage (resembling Sjögren syndrome), and in some cases restricted mouth opening due to scarring of oral and surrounding tissues. These symptoms can make eating, drinking, and dental care painful and complicated.
The eyes are another frequent target of chronic GVHD. Dry eye disease in this setting can become severe, with damage to the corneal surface and significant pain. Conjunctival fibrosis, where scar tissue forms on the inner surface of the eyelids, is a marker of advanced ocular involvement.
Why Genetic Mismatches Matter Beyond HLA
Matching donors and recipients by their HLA type (the major tissue-compatibility markers) is the most fundamental step in reducing GVHD risk. But even when HLA matching is perfect, GVHD can still develop, and a large part of the explanation lies in minor histocompatibility antigens. These are small protein differences between donor and recipient that arise from genetic variation across the genome.
Early work identified specific minor antigens, including one called HA-1, whose mismatch was strongly correlated with GVHD of grade II or higher in adult recipients of HLA-identical sibling transplants. Interestingly, this correlation was not observed in children. A broader genome-wide analysis later showed that in HLA-matched sibling pairs, the average mismatch in coding genetic variants was about 9%. Each additional one-percent increase in this genome-wide mismatch was associated with roughly a 20% increase in the risk of severe acute GVHD. These findings underline that GVHD risk is not simply an on-off switch determined by HLA; it is shaped by thousands of small genetic differences scattered across the genome.
The Gut Microbiome Connection
One of the more revealing areas of recent research involves the gut microbiome, the community of bacteria and other microbes living in the intestines. The pre-transplant conditioning regimen, along with the broad-spectrum antibiotics patients often receive, devastates this microbial community. Beneficial bacterial groups like Clostridia and Faecalibacterium decline, while harmful species like Enterococcus tend to dominate. This imbalance, known as dysbiosis, weakens the gut barrier, fuels systemic inflammation, and pushes the immune system toward aggressive, pro-inflammatory responses that worsen GVHD.
Certain bacteria produce short-chain fatty acids (SCFAs) like butyrate and propionate as byproducts of digesting dietary fiber. These molecules have potent anti-inflammatory effects, helping to maintain gut barrier integrity and supporting regulatory immune cells that tamp down excessive immune responses. Research has found that patients who went on to develop chronic GVHD had lower circulating concentrations of propionate and butyrate in their blood about 100 days after transplant, compared with patients who remained free of this complication. This finding has fueled interest in strategies to restore microbial balance, including probiotics, prebiotics, dietary fiber supplements, and fecal microbiota transplantation. While these approaches show promise in experimental settings, they are still being tested in clinical trials.
Preventing GVHD Before It Starts
Prevention, or prophylaxis, has traditionally relied on drugs that broadly suppress the immune system after transplant, most commonly calcineurin inhibitors like tacrolimus or cyclosporine combined with another immunosuppressant. These drugs reduce GVHD but at the cost of dampening the beneficial immune surveillance that helps prevent cancer relapse.
One of the most significant recent shifts in the field has been the growing use of high-dose cyclophosphamide given shortly after the transplant itself. Originally developed for transplants from half-matched (haploidentical) family donors, post-transplant cyclophosphamide works by selectively destroying the rapidly dividing T cells that have just been activated against the recipient’s tissues, while sparing the resting T cells that provide normal immune function. In haploidentical transplants, this approach reduces GVHD to acceptable rates without needing to physically remove T cells from the graft beforehand. What has been particularly striking is its effect on chronic GVHD: the incidence of chronic GVHD requiring treatment dropped to around 16% in one early study of HLA-matched transplants using this strategy.
A large comparative analysis showed that post-transplant cyclophosphamide improved overall survival when used with matched unrelated donors compared to standard calcineurin-inhibitor-based prevention, and it also improved outcomes when donors were mismatched. In other words, this approach appears to partially erase the disadvantage of using less-than-perfectly-matched donors, which has significant practical implications for patients who lack an ideal donor.
An Alternative Approach to Graft Engineering
Another strategy involves physically removing the cells most responsible for GVHD from the donor graft before infusing it. The immune cells that drive GVHD are primarily a type called alpha-beta (αβ) T cells. Selectively depleting these from the graft while leaving behind gamma-delta (γδ) T cells and natural killer (NK) cells preserves important anti-infection and anti-tumor immune functions. This selective depletion, often combined with removal of B cells, has been studied particularly in pediatric transplants and shows promise in reducing GVHD while maintaining immune recovery.
A newer frontier involves regulatory T cells, a specialized subset of immune cells whose job is to suppress excessive immune responses. Infusing donor regulatory T cells alongside the transplant could, in theory, prevent GVHD while leaving anti-tumor immunity intact. Preclinical work and early-phase human studies support this concept, and genetic engineering of regulatory T cells to improve their specificity and durability is an active area of development. Clinical trials have shown that these approaches are feasible in both fully matched and half-matched transplant settings, though they remain experimental.
Treating GVHD That Does Not Respond to Steroids
When acute GVHD develops, the first treatment is almost always corticosteroids. But a substantial proportion of patients either do not respond or relapse after an initial response, a situation called steroid-refractory GVHD that historically carried a poor prognosis and few options.
The approval of the JAK inhibitor ruxolitinib changed that landscape. In a randomized trial for steroid-refractory acute GVHD, about 62% of patients treated with ruxolitinib responded by day 28, compared with 39% receiving the best available alternative therapy. The responses were also more durable: the estimated rate of losing a response within six months was 10% with ruxolitinib versus 39% with control treatment. A separate randomized trial in steroid-refractory chronic GVHD showed similar benefits, with about half of patients responding to ruxolitinib at 24 weeks compared to about a quarter in the control group, and ruxolitinib more than tripling the time patients went without treatment failure. Data in children have been more limited, but a review of published cases reported overall response rates ranging from 45% to 87% in acute GVHD and 70% to 91% in chronic GVHD.
For chronic GVHD that has failed multiple prior treatments, a drug called belumosudil targets a different signaling pathway (ROCK2) involved in fibrosis and immune dysregulation. In patients with severe chronic GVHD who had already tried a median of three prior treatment lines, belumosudil produced best overall response rates in the range of 74% to 77%. Responses occurred across all affected organs, including difficult-to-treat sites like the lungs, liver, and fibrotic skin. Belumosudil has also shown improvements in ocular GVHD, with over half of patients in one study showing measurable improvement in eye surface disease and roughly two-thirds reporting symptom improvement after six months of treatment.
The Graft-Versus-Leukemia Paradox
Here is the uncomfortable truth that makes GVHD such a difficult problem to solve: the same donor immune cells that cause GVHD also provide one of the most powerful anti-cancer effects in transplant medicine. When donor T cells attack the recipient’s tissues, they often attack remaining cancer cells too. This graft-versus-leukemia (GVL) effect is a major reason allogeneic transplants cure cancers that chemotherapy alone cannot.
The paradox is that most strategies to prevent or treat GVHD also reduce the GVL effect. Broadly depleting or suppressing donor T cells prevents GVHD but often leads to higher relapse rates. Finding treatments that reduce GVHD without crippling GVL has been a central challenge in the field for decades, and routine methods to achieve this reliably are still lacking. The regulatory T cell approaches mentioned earlier are attractive partly because they appear to suppress GVHD without blocking the cytotoxic T cell activity needed to kill cancer cells, though this balance remains under investigation.
How Doctors Grade and Monitor the Disease
Standardized grading systems are essential for comparing results across transplant centers and clinical trials. An international task force has recommended the MAGIC criteria for acute GVHD and the NIH 2014 criteria for chronic GVHD as the most comprehensive frameworks currently available. Acute GVHD is staged by the severity of involvement in each target organ (skin, gut, liver) and then assigned an overall grade from I (mild) to IV (life-threatening). Chronic GVHD is scored by the number and severity of organs involved, classified as mild, moderate, or severe.
Beyond clinical grading, blood-based biomarkers are gaining traction as a way to predict outcomes. Two markers linked to gut damage, ST2 and REG3α, have been validated as predictors of death from causes other than cancer relapse in patients with acute GVHD. A combined algorithm using both markers correctly classified about 75% of patients by risk at the time of GVHD diagnosis. This kind of risk stratification could eventually guide decisions about how aggressively to treat, allowing doctors to intensify therapy early for high-risk patients and potentially spare low-risk patients unnecessary toxicity.
GVHD in Children Versus Adults
The underlying biology of GVHD appears to be the same in children and adults, but there are meaningful clinical differences. Children generally have lower rates of chronic GVHD than adults, and the earlier finding that certain minor antigen mismatches correlated with GVHD in adults but not in children hints at age-related differences in immune reactivity. Children also face unique long-term challenges: GVHD and its treatments can affect growth, bone development, hormonal function, and the ability to attend school and participate in normal childhood activities.
A study of young long-term survivors of allogeneic transplant found that about 31% still had active chronic GVHD. Prior acute GVHD increased the risk of persistent chronic disease, while pretransplant T cell depletion reduced it. Chronic GVHD severity was associated with a greater burden of symptoms, and transplant survivors as a whole reported lower quality of life compared to population norms. These findings highlight that GVHD is not just an acute medical emergency but a chronic condition that can shape daily life for years.
How the Field First Recognized GVHD
The discovery of GVHD grew out of an unexpected place: research into the effects of nuclear radiation after World War II. Scientists studying total-body irradiation in mice found that animals given lethal doses of radiation could be rescued by infusing bone marrow from another animal, but many of these rescued mice later sickened and died from what was initially called “secondary disease.” Over time, researchers realized that this delayed illness was caused by the transplanted immune cells attacking the recipient, and the condition was renamed graft-versus-host disease. Understanding GVHD was, in a real sense, a prerequisite for making bone marrow transplantation viable in humans, and much of the early translational work still informs how preclinical models are designed today.
Mouse models remain the most common preclinical tool for studying GVHD, supplemented by large-animal models. But translating findings from these models to humans has been inconsistent. Differences in immune system composition, microbiome, and conditioning regimens between species mean that not every promising therapy in mice works in people. This translational gap is one reason progress in GVHD treatment, while real, has been slower than in some other areas of immunology.