[Cytomegalovirus pneumopathies. What role should be given to cytomegaloviruses isolated from blood and bronchoalveolar lavage fluid in AIDS and from organ and bone marrow grafts?].

Cytomegalovirus (CMV) is often suspected as the causal agent in lung disease occurring in various immunodepressive states: AIDS, organ transplantation, bone marrow graft. The mechanisms involved in these three situations is however quite different. The role played by the cytopathogenic effect of the virus and the immune reaction of the host vary considerably depending on the underlying immunodepression. Thus, the criteria allowing to distinguish between CMV infection (presence of the virus or anti-CMV antibodies, no clinical signs) and CMV disease (generalized or organ specific disease resulting from the pathogenic effect of CMV replication) lack precision. The aim of this review of the literature is to assess the implicated immunovirology mechanism and thus the diagnostic (and thus therapeutic) criteria of CMV lung diseases. There is a graduation scale from AIDS, to organ transplantation and bone marrow allograft in the degree of immune reaction implicated in the lung disease and thus the need and timing of antiviral treatment. In AIDS, an interstitial pneumonia, associated with an isolation of CMV (whatever the sample origin, blood, bronchoalveolar lavage or the isolation technique) does not usually implicate treatment. Treatment may be indicated in rare cases (advanced stage immunodepression, high virus titre, endothelial involvement) or in cases in which the infection is also located in other organs. For organ transplantation, observation of CMV in blood or lavage samples in a patient with clinical or radiological signs would justify treatment. For lung transplantation, more so than for any other organ, treatment should be started early whenever respiratory signs are associated with evidence of CMV infection. Finally, in bone marrow allografts, the high rate of failure when pneumonitis has become patent implicates starting treatment immediately upon diagnosis of CMV infection. The strategy proposed here is based on a certain rationale but can be open to discussion. Controlled clinical trials are required to determine the most rigorous and coherent attitude. Finally, within the framework of the diseases examined here, search for lung disease caused by cytomegalovirus should not mask other organ localizations in, for example, the retina, the digestive tract.