In their article, Singh et al. describe their protocol on preemptive therapy in CMV pp65-antigen positive liver transplant recipients, 1 which was implemented 13 years ago and led to a decrease of CMV disease. In their view, preemptive therapy is superior to general prophylaxis, although prospective randomized studies that compare these two approaches do not exist.
However, the authors do not mention the third possibility, which is symptom-triggered treatment. Two randomized, controlled studies comparing preemptive therapy with symptom-triggered treatment or placebo showed that only few CMV-positive patients will develop disease even without preemptive treatment. In the study of Paya et al., CMV-PCR was used to monitor preemptive therapy; only five out of 34 PCR-positive patients (14.7%) developed symptoms in the placebo arm. 2 . We observed six cases of CMV-disease in 30 pp65-antigen positive patients (20%) who did not receive preemptive treatment. 3 Therefore, up to 85% of patients would be treated unnecessarily. Most of the patients who developed disease in the mentioned studies had the high-risk constellation (DΟ©RΟͺ). In addition, an Italian study showed that 22 out of 32 patients (69%) with less than 50/200,000 pp65antigen positive cells cleared the virus spontaneously. 4 Preemptive therapy can only be effective if it is guided by an assay with a high positive predictive value for CMV disease. We have recently shown in 135 consecutive liver transplant recipients that this could be achieved by using a cut-off for the pp65-antigen assay and/or the quantitative PCR. Application of preemptive therapy, guided by a CMV assay with a high positive predictive value in combination with a low incidence of CMV-disease, will be highly cost effective compared to general prophylaxis. 5 Using a balanced immunosuppression and modern CMV assays at our center, the incidence of CMV disease decreased from 9.4% between 1988 and 1993 to 2.7% between 1993 and 2000 without any prophylaxis and with only temporary use of preemptive therapy. In addition, CMV had no negative impact on patient and graft survival. 6 Unlike the Gordian knot, CMV infection is not a knot that defied all solutions to untie it over several decades. After liver transplantation, the problem of handling CMV infection has already been solved in different ways. Depending on several influence factors, such as the net immunosupression, logistics for CMV surveillance, and most of all the overall incidence of CMV disease, each transplant center has to analyze its own experience to figure out the optimal and most economic HCMV management, and general recommendations should not be given. In our own experience, which is paralleled to other centers, symptomatic CMV infection after liver transplantation does neither represent a frequent nor an insolvable problem.