Prospective evaluation of the clinical utility of different methods for the detection of human cytomegalovirus disease after liver transplantation.

Standardized human cytomegalovirus (HCMV) assays were prospectively evaluated to predict HCMV disease. In 135 consecutive adult liver transplantations, pp65-antigenemia, quantitative HCMV-DNA and qualitative pp67-messenger-RNA were determined weekly. No ganciclovir prophylaxis or preemptive treatment was used. One hundred and ten (81.5%) patients showed no HCMV-infection, 25 patients were positive in at least one of the HCMV-tests (18.5%). Four suffered from HCMV viral syndrome (3.0%) and another four from tissue invasive disease. In total, pp65-antigenemia was detected in 18, HCMV-DNA in 22 and pp67-mRNA in 18 patients. The sensitivity and negative predictive value (NPV) for HCMV-disease was 100% for all tests. The PPV for symptomatic HCMV-infection was 47% for pp67 mRNA. In contrast, the PPV of pp65-antigenemia (using a threshold of > 2/200 000 cells) and quantitative PCR (using a cutoff of > 5000 copies/mL) were 80% and 89%, respectively. A cost analysis revealed symptom-triggered or preemptive treatment was less expensive than general ganciclovir prophylaxis, if the incidence of CMV disease was low (<30%). Quantitative human cytomegalovirus (HCMV)-DNA and pp65-antigen assays have a comparable sensitivity and can therefore predict the onset of HCMV symptoms at an early stage. Compared with general prophylaxis, symptom-triggered or preemptive treatment based on one of these assays might reduce the costs and also the danger of ganciclovir resistance.