Detecting "occult" viral infection after marrow transplantation.

In this issue of the Journal, BARBERÁ et al. [1] report on the detection of previously unsuspected Cytomegalovirus (CMV) pulmonary infection in the lungs of marrow transplant recipients who died with evidence of diffuse alveolar damage [1]. Detection of such "occult" infections is of great importance and would have major implications for the care of marrow recipients. Viral pneumonia has been the leading infectious cause of death following marrow transplantation, and CMV assumed the role of "the Captain of Death" among the viruses. Sixty percent of diffuse pneumonias have been due to CMV [2]. For the period between 30 and 150 days after transplantation, this rate is increased to 80% [3]. Other herpes group viruses, such as herpes simplex virus (HSV) and varicella zoster virus (VZV), have been less commonly seen because of the prophylactic administration of acyclovir. To place this paper in perspective, it is important to understand the pathogenesis of CMV pneumonia after marrow transplantation. The incidence of CMV pneumonia has varied according to various epidemiological features of the patient. The major factor is the patient's previous experience with CMV infection as evidenced by the pretransplant serology. Among seropositive patients (who presumably carry latent virus), an incidence of CMV infection (defined as excretion of virus) of approximately 75% and an incidence of CMV pneumonia as high as 25% have been reported [2]. The incidence both of infection and pneumonia among seropositive patients seem unaffected by marrow donor serology or by transfusion practices. It is presumed that most, if not all, CMV infection occurring in seropositive patients is due to reactivation of latent virus during periods of reduced CMVspecific CD8+ cytotoxic T-lymphocyte (CTL) activity [2, 4, 5]. In seronegative patients with seronegative marrow donors, the risk of infection is entirely attributable to blood product exposure, and a 30–40% incidence of primary infection has been observed. The incidence of CMV pneumonia among seronegative patients has been 10% or less. These risks can be eliminated by use of screened seronegative or filtered blood products [6]. Other factors also affect the risk of CMV pneumonia. The most prominent of these is the occurrence of acute graft-versus-host disease (GVHD) among allograft recipients, which increases the risk of CMV pneumonia by twofold or more [2]. Whether therapy for GVHD further increases this risk is less clear. In multivariate analyses, CMV pneumonia is more common among older patients and possibly among patients receiving more intensive conditioning regimens or receiving methotrexate prophylaxis for GVHD. As a corollary to the increased risk associated with acute GVHD, patients not subject to GVHD have a lower incidence of pneumonia. Thus, CMV pneumonia is less common among recipients of autologous transplants (about 4%) and exceedingly uncommon among recipients of twin marrow [7, 8]. CMV pneumonia was previously fatal in over 85% of affected marrow recipients. Multiple experimental treatment modalities were unsuccessful in altering outcome. Several centres have reported favourable responses to combination therapy with ganciclovir and high-titre antiCMV immunoglobulin [9–11]. Various treatment regimens have been used successfully, usually initially involving administration of ganciclovir 2.5 mg·kg-1 t.i.d. for at least 2 weeks, together with CMV immunoglobulin 400–500 mg·kg-1 3–5 times weekly for 2–3 weeks. In some series, continued therapy with lower dose ganciclovir and CMV immunoglobulin for a period of several weeks after successful therapy has been suggested in order to avoid early relapse. Thus, CMV pneumonia has been of primary concern as a cause of morbidity and mortality among marrow recipients. However, up to 12% of patients develop diffuse pneumonia in which no infection can be detected. The name "idiopathic pneumonia syndrome" has recently been applied to describe the clinical presentation of these apparently noninfectious processes [12]. Concern that undetected viral infection may account for a proportion of these idiopathic pneumonias has persisted in the marrow transplant community. As reported in this issue of the Journal, BARBERÁ et al. [1] from Barcelona examined the necropsy lung tissue of 19 marrow transplant recipients, who displayed histology of diffuse alveolar damage (DAD), for evidence of viral infection. These cases had no characteristic histological features suggesting viral pneumonia, such as Cowdry type A intracytoplasmic inclusion bodies, other than the presence of DAD. In situ hybridization with biotinylated nucleic acid probes and immunohistochemistry with antibodies to CMV, HSV, Epstein Barr virus (EBV) and adenovirus were performed and detected CMV deoxyribonucleic acid (DNA) and/or protein antigen in six cases. Notably, and consistent with the epidemiology of the infection, only necropsy tissue from patients with a history of allogeneic marrow transplant and acute GVHD were positive for CMV. Despite the fact that 13 of these patients had undergone bronchoalveolar lavage (BAL), CMV infection had been suspected in only a single case (on the basis of serological changes). Necropsy viral studies from cases with normal lung histology were negative. EDITORIAL