This colon biopsy was obtained from a 25 year-old male patient with abdominal pain and intermittent
guaiac positive diarrhea on day 46 after marrow allograft transplantation for acute myelomonocytic
leukemia (AMML). The acute leukemia had developed following autologous marrow transplantation for
metastatic embryonal cell carcinoma diagnosed at age 20. Although engraftment had been prompt after
infusion of marrow from his HLA mismatched mother, graft-versus-host disease (GVHD) of skin and liver had
developed despite prophylaxis with humanized anti-CD3 antibody and methyl-prednisolone. The GVHD
resolved following the addition of cyclosporin and methotrexate therapy, but the gastrointestinal
symptoms continued and led to endoscopy and biopsy.
Diagnosis: Colon biopsy:
1) Crypt epithelial cell apoptosis and
inflammation consistent with graft-versus-host disease.
Even in the small remaining segment of H&E stained colon mucosa that you were able to examine,
alterations characteristic of both GVHD and viral infection were evident. Gastrointestinal GVHD leads
to apoptotic cell death of crypt epithelial cells, so their chromatin condenses and fragments, leaving
behind "nuclear dust" and cellular debris. Since biopsies from patients who have undergone hematopoietic
stem cell transplantation are being seen more and more frequently by pathologists in all practice
settings, let me digress briefly to point out that gut GVHD is frequently overlooked histologically.
Although it is easy to recognize when significant numbers of apoptotic crypt epithelial cells are present
as illustrated by this case and described above, frequently only a few scattered crypts display only one
or perhaps several apoptotic cells which are unassociated with a significant inflammatory infiltrate.
These focal and minimal yet diagnostic findings can be easily passed over, especially when only a small
number of thick, overstained, or artifactually torn sections are examined hurriedly.1
In contrast to the apoptotic cell destruction in GVHD, adenovirus infection is characterized by the
progressive enlargement of an amphophilic and increasingly basophilic nuclear inclusion associated with
peripheralization of chromatin and absence of cytoplasmic inclusions. The most characteristic
histological feature of adenovirus infection is the presence of cells with hyperchromatic nuclei
displaying indistinct nuclear membranes, so-called "smudge cells".
At various stages, the adenovirus cytopathic effects may resemble those seen with herpes simplex,
varicella-zoster, or cytomegalovirus infection. In the present case, there is in some sections a classic
Cowdry Type A ("owl's eye") nuclear inclusion which looks for all the world like a CMV-infected cell.
Fortunately, commercially available monoclonal antibodies now allow us to make specific viral diagnoses
immunocytochemically in fixed tissue. In the present case, there was reactivity only for adenovirus, and
it was subsequently isolated in culture from the colon biopsies as well as from the stool and urine.
After 10 days of intravenous ribavirin along with tapering of steroid therapy to reduce
immunosuppression, urine and stool cultures became negative. The patient's condition stabilized for
several weeks, but liver and gut GVHD recurred and he died on day 73 post transplant of multiorgan
failure associated with GVHD and bacterial sepsis. The autopsy was limited to needle sampling of the
liver, which showed GVHD without viral involvement.
Rapid and specific diagnosis of these viral infections has become essential as more effective
antiviral drugs have become available. Many patients now recover when given ganciclovir or foscarnet for
disseminated CMV or acyclovir for HSV or VZV. Unfortunately, there is currently no definitely effective
adenoviral therapy. Mixed results have been reported with intravenous ribavirin and vidarabine.
Cidofovir displays in vitro antiviral activity, and encouraging clinical results have recently been
reported although these have been associated with significant toxicity.2 We use ganciclovir
prophylactically in CMV seropositive patients and preemptively for CMV antigenemia. Since we introduced
ganciclovir in 1990, we have found that patients receiving it have been at significantly lower risk for
development of invasive adenoviral disease (submitted). This intriguing suggestion of a suppressive role
for this drug is supported by the presence of in vitro antiviral activity 3 as well as anecdotal
reports of its use for successful treatment.4
In the general population, adenovirus infections occur most frequently during childhood and cause
upper respiratory syndromes, conjunctivitis and enteritis. Clusters and seasonal outbreaks of infection
have been reported in communities such as military quarters, hospitals and shipyards.5 It is believed
that following primary exposure the virus persists as a latent infection in lymphoid tissues such as
tonsils and adenoids.6 Adenovirus infection in the 5233 allogeneic hematopoietic stem cell transplant
(HSCT) recipients and 1219 autologous recipients we have treated in the 20 years between 1978 and 1998
has not been associated with outbreaks in the general population, suggesting that disease has resulted
from reactivation of endogenous virus. In addition, in 21 patients with adenovirus infection of the
kidney, a majority were infected with serotypes 11, 34 and 35 (subgenus B2, which has a well-established
predilection for the urinary tract) and showed only kidney involvement at autopsy despite culture and
immunocytochemical analysis of all major organs. These findings are strongly suggestive of reactivation
of latent urinary tract infection.
The published rates of adenovirus infection in HSCT patients have varied from approximately 5 to 20%,
with higher rates reported for pediatric patients and for those given T cell depleted hematopoietic stem
cells. In an earlier publication from this center,7 we reported a 4.9% incidence in the 1051 patients
we transplanted between 1976 and 1982. Although there has been a strong suggestion that the incidence of
infection has increased in more recent years, changes in surveillance guidelines and detection techniques
have created uncertainty in this regard. When we limit analysis of our data to the seven-year period
(1991-1998) when detection methods were uniform, there is a small but significant increase in the
incidence of infection. However, there was no increase in invasive disease, perhaps as a result of the
possible protective effect of ganciclovir mentioned above. For the overall 20-year period of our study,
the incidences of infection and invasive disease respectively in allogeneic patients were 8.6% and 0.8%.
Infection was documented in 6.0% of autologous patients and there was disease in 0.5%. By multivariable
analysis of data for the allogeneic population, the risk of infection was significantly increased by
lower patient age, receipt of a second transplant, the presence of severe acute GVHD and lack of
Viral disease is a significant threat to HSCT patients, especially those who have developed
graft-versus-host disease following allogeneic transplantation. Prompt pathology studies can at times
yield a life-saving diagnosis.
- Ponec RJ, Hackman RC, McDonald GB.
Endoscopic and histologic diagnosis of intestinal graft-versus-host disease after marrow
transplantation. Gastrointest Endosc 49:612, 1999.
- Bordigoni P, Carret AS, Venard V, Witz F, Le
Faou A. Treatment of adenovirus infections in patients undergoing allogeneic hematopoietic stem cell
transplantation. Clin Infect Dis 32:1290, 2001.
- Wreghitt TG, Gray JJ, Ward KN, Salt A, Taylor
DL, Alp NJ, Tyms AS. Disseminated adenovirus infection after liver transplantation and its possible
treatment with ganciclovir. J Infect 19:88, 1989.
- Chen FE, Liang RH, Lo JY, Yuen KY, Chan TK,
Peiris M. Treatment of adenovirus-associated cystitis with ganciclovir. Bone Marrow Transplant 27:997,
- Hierholzer JC. Adenoviruses in the immunocompromised host. Clin Microbiol Rev 5:262,
- Strickler JG, Singleton TP, Copenhaver CM, Erice A, Snover DC. Adenovirus in the
gastrointestinal tracts of immunosuppressed patients. Am J Clin Pathol 97:555, 1992.
AF, Hackman RC, Fife KH, Corey L, Meyers JD. N Engl J Med 312:529, 1985.