Comparison of quantitative competitive polymerase chain reaction-enzyme-linked immunosorbent assay with LightCycler-based polymerase chain reaction for measuring cytomegalovirus DNA in patients after hematopoietic stem cell transplantation

Development of highly sensitive quantitative assays for cytomegalovirus (CMV) DNA detection is crucial for identification of immunodeficient patients at high risk of CMV disease. We designed 2 internally controlled competitive quantitative assays, enzyme-linked immunosorbent assay (ELISA)-based and real-time polymerase chain reaction (PCR) tests, using amplification of the same segment of the CMV genome. The aim of this study was to compare sensitivity, specificity, and laboratory performance characteristics of these assays. In both assays, a 159-bp segment of UL83 gene was amplified. External and internal controls were constructed by cloning the amplification product and heterogenous DNA segment flanked by target sequences for CMV-derived primers into bacterial plasmids, respectively. Real-time PCR was performed on LightCycler (Roche Diagnostics, Mannheim, Germany), and amplicons were detected using fluorescence resonance energy transfer probes. Alternatively, PCR products were labeled by digoxigenin, hybridized to immobilized probes, and detected by ELISA. The assays were tested on genomic DNA isolated from laboratory strains of CMV, QCMD control panel, and CMV DNA-positive peripheral blood DNA samples from hematopoietic stem cell transplant recipients, previously characterized by pp65 antigenemia and qualitative nested PCR. Real-time and ELISA-based PCR assays showed a linear course of 1-10(8) and 10-10(5) copies of CMV DNA per reaction, respectively. When compared with ELISA-based PCR, real-time PCR showed superiority in inter- and intra-assay reproducibility. Both assays were highly specific in detecting CMV DNA. No difference in amplification efficiency of internal or external standards and wild-type CMV DNA was found. The assays exhibited 83% concordance in CMV DNA detection from clinical samples, all discrepant samples having low CMV DNA copy numbers. There was a good correlation between viral DNA loads measured by the 2 assays. Statistically significant correlation was observed between the numbers of CMV DNA copies and pp65-positive leukocytes in the samples tested. Both variants of competitive PCR are adequately sensitive to be used for CMV DNA quantitation in clinical samples. LightCycler PCR, having superior performance characteristics and being less time-consuming, seems to be more suitable for routine diagnosis.