Diagnosis of Acute Graft-versus-Host Disease

The first descriptions of acute graft versus host disease (aGVHD) were made by scientists involved in the early experimental transplants of allogeneic spleen cells into irradiated mice. To differentiate this new illness from the initial problems observed in the mice related to primary radiation sickness, this murine aGVHD was referred to as “secondary disease” (1). Today, aGVHD, or graft-versus-host disease occurring within 100 days of a stem cell transplant (distinguishing the acute from the chronic forms of the disease), remains a major problem for the transplant physician and his patient. The pathophysiology of aGVHD can be considered a three-step process (2): Damage from chemo-radiotherapy. Host antigen-presenting cell activation and amplification of donor T cells. Target cell apoptosis via cellular and inflammatory mediators. Overall, this process results in varying degrees of tissue damage to the skin, liver, gut, and less often to the eyes. The various clinical grading systems for aGVHD do not consider the ocular complications, which include photophobia, hemorrhagic conjunctivitis, and pseudo-membrane formation, even though patients with eye involvement may have a poorer prognosis (3). Most transplant units grade the severity of aGVHD using the system described by Glucksberg (4) or a variation of the same. The grading score takes into account the extent and type of skin lesions, the degree of jaundice and the volume of diarrhea. Over many years of clinical usage, this system has proved to be useful with high-grade (III and IV) patients having a much poorer outcome than those with low-grade disease (I and II). However, the International Bone Marrow Transplant Registry Severity Index may give an even more accurate prediction of outcome (5). aGVHD cannot always be considered as disadvantageous, particularly for patients transplanted for malignant diseases such as leukaemia. These patients benefit from the graft versus leukaemia effect of the transplant, which as yet cannot be separated out from the GVHD effect. Patients with aGVHD are less likely to relapse after stem cell transplant than patients with no GVHD. The etiology of aGVHD is complex. There are those factors relating to the potential donor and recipient. The most important of which is the presence of HLA mismatch; in general, the greater the degree of HLA mismatch, the greater the risk of aGVHD. The age of the recipient is also an important factor. Under the age of 20 years, approximately 20% of patients will experience grade II–IV aGVHD, as opposed to 80% of patients over 50 years of age who receive grafts following a fully myeloablative induction regimen. Less important donor/recipient factors are sex-mismatching and donor parity. Stem cell source and induction regimen can also affect the incidence of aGVHD. Umbilical cord cells appear to be associated with less aGVHD than other sources of stem cells and “nonmyeloablative” induction similarly reduces the risk GVHD presumably by causing less tissue damage than more traditional induction therapies. Finally, immune modulation reduces the incidence of aGVHD. This involves the use of immunosuppressive prophylaxis to both the recipient's immune system and the immunoreactive cells in the transplant. Stem cell transplantation has rarely been undertaken without the prophylactic use of drugs like cyclosporine, methotrexate and more recently tacrolimus and mycophenolate mofetil to prevent unacceptable rates of GVHD. Serotherapy with antithymocyte globulin, physical methods to remove donor T-cells and more general lymphocyte depletion using monoclonal antibodies such as Campath, have made increasingly mismatched stem cell transplants possible. But this has not been without problems, notably increased rates of primary graft failure and loss of graft versus malignancy effect associated with an increased risk of relapse. The improved strategies for aGVHD prophylaxis have made alternative donor stem cell transplants a safer option. Increasingly mismatched donor stem cells are being used and even one-haplotype, mismatched parents can be considered as donors. Procedure planning has become more complex as the donor pool has increased. Each transplant has to be individually designed particularly with respect to the disease being treated, the desirability of graft versus malignancy effect and the age of the patient. Although prevention of aGVHD has markedly improved over the last two decades, the treatment of established aGVHD has not really changed and corticosteroids remain the mainstay of therapy. Moderate dose regimens of prednisolone 2 mg/kg/day have been shown to be as effective as those using higher dosage schedules and are safer. However, two problems remain: the lack of success of second line therapy for aGVHD and the clinical problem of accurate diagnosis. In many situations, the diagnosis is not difficult and can be made clinically from the appearance of a typical rash in a patient, who has recently begun to engraft. However, as more “alternative” donors have been used and immunosuppression has been correspondingly escalated, infections, particularly viral infections, have become increasingly common. At the same time the new molecular techniques for the diagnosis of viral infections have greatly increased the number of patients diagnosed with viral reactivation or de novo infection. Clinicians are faced with patients with symptoms that could be aGVHD but might equally well be infection. Patients with isolated diarrhea or a rising bilirubin, often without a typical aGVHD rash, are a particular diagnostic problem. It can impossible at times to be sure if a patient has an infection or aGVHD or both. As increased immunosuppression would be indicated for aGVHD but not desirable for a viral infection, this clinical dilemma is an important one. There is a definite clinical need for improved diagnostic methods such as that described by Hiramaya et al. in this issue of Transplantation, which appears to offer the ability to confidently diagnose aGVHD even in the presence of an active infection.