Cytomegalovirus infection complicating immunosuppressive treatment.

Two case reports in this issue of the Journal describe remarkably similar clinical presentations. The patients (a 37 year old woman1 with systemic lupus erythematosus and a 49 year old man2 with glomerulonephritis and pulmonary haemorrhage) presented with lung involvement which improved after immunosuppressive therapy was instituted. The patients were maintaoned on lower doses of immunosuppressives and remained stable. Once their condition was under control, episodes of severe haemoptysis occurred four weeks and six weeks after initial presentation. Investigation of both patients lead to lung biopsy which revealed inclusion bodies within the lung suggestive of cytomegalovirus (CMV) infection. One patient survived while the other died. The authors discuss whether the immunosuppression given might have reactivated CMV or whether the virus was transmitted by blood transfusion. They also discuss the pathogenesis of CMV infection of the lung. Cytomegalovirus is a common infectious agent which is well adapted to hosts with normal immune responses. However, in immunosuppressed patients, latent virus can reactivate to lead to dissemination of infection and ultimately life-threatening disease. Such reactivations are the usual means by which CMV affects bone marrow transplant3'4 and acquired immune deficiency syndrome (AIDS) patients5 in whom pre-existing immunity confers little protection. In renal transplant patients, most reactivations of virus are asymptomatic while primary infections (transmitted with the donor kidney) are frequently life-threatening.6 Re-infection of patients with pre-existing antibody also occurs7 and disease can develop in a proportion. The virus can also be transmitted by blood transfusion and the serological status of the donor (seropositive) and of the recipient (seronegative) represent markers of susceptibility to CMV spread by blood transfusion.8 The pathogenesis of the various clinical conditions caused by CMV is not completely understood. Involvement of the retina and gastrointestinal tract for example, appears to be mediated by active virus replication. In contrast, CMV can be detected readily in urine and saliva, without the accompaniment of symptoms.3'9-11 In the mouse model of CMV, the development of CMV pneumonitis requires a host immune response which can be suppressed by continued administration of cyclophosphamide12 while the presence of graft versus host disease13 can turn an asymptomatic CMV infection into one with fatal outcome. What lessons can be learned from these case reports which could ultimately benefit patients? Since CMV is the single commonest cause of death in bone marrow transplant patients14 and is a major cause of mortality and morbidity in renal transplant patients,15 16 physicians caring for such individuals should already be following standard protocols to detect this virus. Those caring for other groups of immunosuppressed patients may be stimulated by the two case reports to follow such guidelines, the principles of which are straightforward. First, when a patient presents with a clinical condition which may require therapeutic immunosuppression, his antibody status to CMV should be determined. The absence of antibody indicates that he is seronegative and so not infected with this virus and every attempt should be made to give him only blood products which have been shown to lack antibodies to CMV. The presence of antibody within the patient does not indicate immunity in the protective sense but is a marker for latent CMV infection which may reactivate under the influence of immunosuppressive drugs.15 Second, once immunosuppression has been administered, serology has no further diagnostic role to play since it is absurd to assume that such patients will be able to mount appropriate immune