Hot Topic

Graft Versus Host Disease

Considerations in Apoptotic Pattern of Injury


Receive notification when new Hot Topics are published:

Click CC to turn on closed captioning.

Published: September 2013

Print Record of Viewing

Dr. Lam-Himlin describes the histologic features and clinical correlation of graft versus host disease and considerations for diagnosis.

Presenter: Dora M. Lam-Himlin, MD

  • Senior Associate Consultant, Division of Anatomic Pathology
  • Assistant Professor of Laboratory Medicine and Pathology at Mayo Clinic in Arizona

Questions and Feedback

Contact us: .


Download the PDF

Thank you for the kind introduction. It is a privilege to speak to you about the histologic diagnosis of graft versus host disease and other diagnostic considerations in apoptotic pattern of injury. The content of this lecture is intended for pathologists in practice and in training, and the scope will cover diagnosis of graft versus host disease as well as the differential diagnoses.

I have no conflicts of interest to disclose.

There are 4 primary objectives that viewers should achieve at the end of this presentation.

  1. The first is the ability to outline the histologic features of graft versus host disease (GVHD).
  2. Secondly, viewers should be able to list the differential diagnoses for apoptotic pattern of injury.
  3. The third objective is to recognize that GVHD is a clinical correlation diagnosis.
  4. In doing so, by the end of this presentation, viewers will be able to list the clinical and medication considerations in the diagnosis of GVHD.

This is the overview, or road map, for our discussion over the next 20 minutes.

I will start with a basic review of GVHD, including a discussion of the pathophysiology, clinical features, endoscopic findings, and histology and grading. Following this, we will discuss apoptotic pattern of injury through a series of 4 case examples.

To whet your appetite for the upcoming discussion, and to introduce the concept of apoptosis as a nonspecific pattern of injury, observe these four examples of colonic mucosa with abundant apoptoses. These examples each show colonic crypts with markedly increased apoptotic bodies, and look remarkably similar. However, each of these 4 cases has distinctly different etiologies. Keep this in mind as we move into our discussion of GVHD.

Graft versus host disease is the donor T-lymphocyte mediated destruction of host tissues that occurs following allogeneic transplant. Hematopoietic transplant is the most likely to cause GVHD, with a frequency of 40-80% in transplant patients. Donor lymphocyte transfusions, such as might be seen in whole blood transfusion, can also cause GVHD. Rare reports of GVHD secondary to solid organ transplantation have been recorded, although this is exceptionally rare.

Patients present with gastrointestinal tract symptoms, the most common of which include nausea, vomiting, and diarrhea. Abdominal pain may or may not be a significant symptom. Some patients present with gastrointestinal bleeding, which can be significant enough to require transfusion support. More commonly affected than the gastrointestinal tract is the skin, which can show a sunburn-like rash, particularly in the face and chest areas. The luminal gastrointestinal tract is the next most common site of involvement, followed by the liver.

Endoscopic findings are essentially nonspecific, and can range from showing patchy erythema and ulceration in severe cases to normal in more than 40% of cases of GVHD. The best diagnostic material for evaluation of GVHD are biopsies of the duodenum and colon. These sites are the most sensitive for diagnosis. It is worthy to note that severity of GVHD can vary from site to site and the effects of GVHD can be patchy. As a result, concordance rates of GVHD grading between upper and lower GI biopsies are poor. As such, some institutions include both upper and lower gastrointestinal biopsies in their GVHD protocol. However, it is not uncommon to receive only upper or only lower GI biopsies for the evaluation of GVHD.

Histologically, apoptosis is the hallmark of this disease, and historically pathologists have graded GVHD on a 4-tiered grading system. I will first describe these criteria, and in the next few slides we can review some photomicrograph examples.

The mildest form, grade 1 GVHD, is defined as the presence of increased apoptoses in the crypt epithelium. While the normal number of apoptoses has been reported as 1 per 10 crypts in the colon and up to 1 per 4 crypts in the duodenum, a practical strategy to keep in mind is that apoptoses in the setting of high clinical suspicion are likely to represent GVHD. Note, for example, that the Stanford criteria for GVHD require only that a single apoptotic body is present in each tissue fragment. When struggling with an equivocal case of low grade GVHD, investigation into the patient’s clinical symptoms may be helpful; for example, the presence of a skin rash is a useful indicator that the patient has some degree of GVHD. In addition, it is also helpful to keep in mind that low grade GVHD, such as grade 1 GVHD, is not always treated. Rather, clinicians may choose to watch and wait.

As GVHD progresses in severity to grade 2, one can observe the presence of crypt abscesses and individual crypt damage. When crypts begin to drop out and focal mucosal necrosis is present, this is considered grade 3 GVHD. Finally, grade 4 GVHD is defined by the presences of diffuse mucosal necrosis and denudation of the mucosa. Remember that these histologic findings exist along a spectrum, and by necessity some degree of subjectivity in interpretation will be present.

While most pathologists are accustomed to the 4-tiered grading system for GVHD, a separate National Institute of Health consensus recommendation for the reporting of GVHD exists. This reporting mechanism utilizes the degree of confidence that a pathologist feels in diagnosing the presence or absence of GVHD by reporting absent, possible, probable, and unequivocal GVHD. Although this method of reporting has been adopted in some centers, many clinicians prefer to receive the numerical grading system described previously.

Finally, before we move on to histologic examples of GVHD, it is important to note that longstanding low-grade GVHD can cause crypt distortion which can be mistaken as chronicity changes of inflammatory bowel disease. Before diagnosing a bone marrow transplant patient with new onset inflammatory bowel disease, investigate their history of GVHD.

The following several slides include examples of histologic findings in GVHD. It is not possible to identify apoptotic bodies at this low power view of the colonic mucosa. However, at this scanning magnification, one can note the slight hyperchromasia of the deep crypt epithelium –the area which we know is the proliferative compartment. This is a helpful clue at low power and further investigation of these hyperchromatic areas will be the highest yield for apoptotic bodies.

At higher magnification, one can appreciate the presence of multiple apoptotic bodies in close proximity to one another. These clusters of dying epithelial cells contain numerous minute fragments of nuclear debris that are deeply basophilic. There is no significant crypt damage or destruction in this example, which is sufficient for a diagnosis of grade 1 GVHD in the proper clinical context.

Another example of apoptotic bodies with preservation of crypt architecture is seen here. This example shows abundant apoptotic bodies, which are seen as small round dark bodies surrounded by a white halo. A common question from pathologists is, “How can one differentiate apoptotic bodies from lymphocytes, which may also appear as small dark round bodies with white halos?”  This is an important question to ask, and has a relatively simple solution. One can search for a known lymphocyte in the lamina propria. Take, for example, the 2 small lymphocytes found in the lamina propria of this photo at the far left. Use the nuclei of these inflammatory cells as the benchmark size, which should be about 10 microns in diameter. Compare this nuclear size to the body in question. Apoptotic bodies will be smaller in size by comparison. Numerous examples exist within this photo and show a wide range of sizes for apoptotic bodies.

Although I frequently emphasize that apoptotic bodies are found in clusters in GVHD, single cell apoptosis is also seen. These 2 examples highlight the subtlety of single cell apoptoses. Again, note the presence of small dark round nuclear debris surrounded by a white halo.

This example of GVHD grade 2 shows a crypt abscess and early crypt damage with early crypt dropout.

Higher power magnification of the previous photo confirms the presence of abundant apoptotic bodies.

Grade 3 GVHD is characterized by loss of crypts, as seen here. Note the architectural distortion that is present, with areas of crypt dropout.

This example of severe grade 4 GVHD shows complete loss of crypts and epithelium. Only granulation-type tissue is present.

With the key features of GVHD in mind, let’s take a look at our 4 example cases. Remember, these four cases all demonstrate a pattern of injury that exhibits prominent apoptotic bodies, but all 4 have different etiologies.

Case 1. This photograph shows a colonic crypt with abundant apoptotic bodies. In the absence of any clinical information, a pathologist can only comment that an apoptotic pattern of injury is present and list the differential diagnoses. However, further information is available which reveals that this biopsy is from a 32-year-old man who is status post bone marrow transplant and presents with diarrhea. The clinicians ask pathology to rule out GVHD. While the finding of apoptoses is compatible with a diagnosis of GVHD, further examination of the biopsy reveals enlarged cells with densely eosinophilic nuclear inclusions, such as seen in the bottom photo.

The diagnosis in this case is cytomegalovirus infection (CMV). It is the most common GI pathogen in immunocompromised patients who often present with complaints of dysphagia, odynophagia, and diarrhea. The endoscopic finding is of punched-out ulcers, while histology shows a neutrophilic infiltrate, ulceration, and nuclear and cytoplasmic inclusions. The virus preferentially infects endothelium, stromal cells, and macrophages. Treatment is with antiviral therapy and the key to diagnosis is identification of characteristic viral inclusions. Immunohistochemistry may also be necessary in some cases, and is often an up-front immunostain in GVHD protocols.

Case 2. At low magnification, this colon biopsy shows hyperchromasia at the proliferative compartment. Higher magnification reveals the presence of abundant apoptotic bodies. Again, in the absence of any clinical information, this pattern of apoptotic injury is nonspecific, and only a differential diagnosis can be given. However, the clinical information provided for this case tells us that the biopsies were taken from a 53-year-old woman who is status post bone marrow transplant. She presented with nausea, vomiting, and diarrhea, and the clinical team has asked pathology to rule out GVHD. Further investigation reveals that this patient is taking Cellcept or mycophenolate mofetil.

CellCept and Myfortic, also known as mycophenolic acid, are immunosuppressive drugs that were originally developed for solid organ transplantation. They are now widely used for heart and lung transplantation as well as bone marrow transplantation. Medication reaction is dose dependent and patients present with gastrointestinal distress. The histology of mycophenolate-induced injury is essentially apoptotic in nature, but can mimic ischemia, GVHD, and inflammatory bowel disease. Pictured here is a severe example of mycophenolate-induced injury, which shows complete loss of epithelium and crypts. Pathologists should keep a low threshold for investigating a patient’s medication list if there is any history of transplantation or signs of apoptotic injury. Treatment is simple with either dose modification or discontinuation of the drug. The result is usually dramatic and near-immediate resolution of symptoms. It is important to note that there are not reliable histologic features to differentiate GVHD from mycophenolate-induced injury, and further investigation by the clinician is necessary in cases in which both GVHD and mycophenolate are a consideration.

Case 3. This example shows basophilia of the colonic proliferative compartment at low power, while high power examination shows abundant apoptotic bodies. Similar to our other cases, clinical information is imperative for interpretation of this case. This biopsy is from a 61-year-old man who is status post bone marrow transplantation and he presents with diarrhea, nausea, and vomiting. The clinicians request that pathology exclude GVHD. Further investigation reveals that this patient has a skin rash, has negative CMV serology, and is not taking mycophenolate.

The best diagnosis in this example is: “Colonic mucosa with increased apoptotic bodies, consistent with mild GVHD (grade 1 of 4), in the proper clinical setting.”

This brings us to the last example, case 4. These colon biopsies show apoptotic activity and come from a 46-year-old woman with severe diarrhea and abdominal discomfort. Given the nonspecific nature of the symptoms and histology, further investigation into the patient’s history is performed and reveals that the patient is receiving intravenous Taxol (paclitaxel) for vaginal adenocarcinoma.

Chemotherapeutic effect can manifest as apoptotic injury in the gastrointestinal tract. Chemotherapy results in reactive oxygen species and some medications, such as Taxol, inhibit microtubule breakdown which results in arrested mitoses. The example here shows abundant arrested mitoses, some exhibiting ring forms. The presentation, endoscopy, and histology are all nonspecific, and the key to diagnosis is in the clinical history and medication list. It is important to note that cytoreductive therapy for bone marrow transplantation can also exhibit these same changes. As a result, chemotherapeutic effect is a consideration if biopsies of the GI tract are taken within the first 21 days of bone marrow transplantation.

As you can see through these examples, the presence of apoptotic bodies in gastrointestinal biopsies is a nonspecific pattern of injury with multiple possible etiologies. Listed here for your reference are a number of situations associated with apoptotic injury. These include GVHD, acute cellular rejection of transplanted organs, chemotherapy and radiation therapy effect, zinc deficiency or any nutritional deficiency such as might be seen in fasting, inflammatory bowel disease, AIDS enteropathy, thymoma, medications, viral infections, cell-mediated immune response, and autoimmune enteropathy. The key to diagnosis in all of these cases is correlation with clinical information.

As we conclude our session today, I would like to reemphasize some of the key take-home points. Recall, apoptosis in the GI tract is not specific for GVHD. In all cases, correlation with clinical information is a must, but particularly when dealing with bone marrow transplant patients. In addition, we discussed that drug and viral effects are frequent in bone marrow transplant patients and should be kept in the forefront of your mind during signout. Finally, modification of mycophenolate dosage should have a rapid clinical response and this feature can help differentiate medication reaction from GVHD in some cases.

For those interested in learning more about this entity, selected references are included. Thank you for taking the time to join me today in our discussion of graft versus host disease and apoptotic pattern of injury. It has been my pleasure and privilege.