Background: The early detection of cytomegalovirus (CMV) after liver transplantation may form the basis of a preemptive strategy for prevention of active CMV disease.
Methods: We prospectively analyzed the clinical use of weekly quantitative polymerase chain reaction-(PCR) based plasma viral load determinations and the antigenemia assay for predicting the development of active CMV disease in 97 consecutive liver transplant recipients.
Results: CMV disease occurred in 21/97 patients. Using a positive cut-off of >400 copies/ml plasma, PCR had a sensitivity of 100%, specificity 47.4%, positive predictive value 34.4% and negative predictive value 100% for prediction of CMV disease. Respective values for a positive antigenemia (>0 positive cells/slide) were 95.2, 55.3, 37.0, and 97.7%. Different cut-off points for a positive test were analyzed using receiver-operating characteristic (ROC) curves. The optimal cut-off for viral load was in the range of 2000-5000 copies/ml (sensitivity 85.7%, specificity 86.8%, PPV 64.3%, NPV 95.7% for >5000 copies/ml). The optimal cut-off for antigenemia was in the range of four to six positive cells/slide. Mean peak viral load in symptomatic patients was 73,715 copies per/ml versus 3615 copies/ml in patients with asymptomatic CMV reactivation (P<0.001). In a multivariate logistic regression analysis of risk factors for CMV disease (CMV serostatus, acute rejection, and induction immunosuppression), peak viral load and peak antigenemia emerged as the only significant independent predictors of CMV disease (for PCR, odds ratio=1.40/1000 copy/ml increase in viral load, P=0.0001; for antigenemia odds ratio=1.17/1 positive cell/slide).
Conclusions: Plasma viral load by quantitative PCR is useful for predicting CMV disease and could be used in a preemptive strategy.