Diagnosis of cytomegalovirus infection in cyclosporin-treated renal allograft recipients

The relative merits of antibody response and virus shedding as markers of cytomegalovirus (CMV) infection among cyclosporin-treated renal allograft recipients were analysed. CMV-specific antibody was assayed by IgG-specific radioimmunosorbent test (RIST) and by complement fixation test (CFT). CMV shedding was assayed by virus isolation and by the rapid test immediate early nuclear antigen detection (IENAD). RIST and CFT detected seroconversion in similar numbers of patients, but the former test was the more sensitive when CMV antibody was sought in pretransplant sera to differentiate primary from recurrent infection. IENAD detected or excluded CMV shedding for more urine specimens than virus isolation (462/515 [90%] vs. 366/515 [71%]), but the reverse applied to saliva specimens (33/57 [58%] vs. 54/57 [95%]). The high specificity of IENAD allowed positive results by IENAD to be accepted when virus isolation failed to provide a result. IENAD was, however, less sensitive than virus isolation even when specimens yielding CMV by IENAD, but no result by virus isolation, were included in the analysis (27/44 [61%] vs. 38/44 [86%]). Assays of both antibody response and virus shedding were required to maximise the diagnosis of recurrent CMV infections, but the former assay detected all primary CMV infections. The diagnostic implications of these results are discussed.