PERFORMANCE EVALUATION OF THE ARTUS CYTOMEGALOVIRUS (CMV) PCR ASSAY FOR ANALYTICAL AND CLINICAL SENSITIVITY

Background: Cytomegalovirus is associated with significant morbidity and mortality in immunocompromised patients. Clinical virology laboratories increasingly use molecular diagnostics to detect and quantify CMV in samples from patients with risk factors for CMV disease.

Objective: To evaluate the performance characteristics of the artus^® CMV PCR assay (QIAGEN, Mississauga, Canada) and to compare it to the pp65 antigenemia assay, currently being offered in our laboratory.

Methods: OptiQuant CMV DNA Quantification Panel samples (AcroMetrix, Benicia, CA) were spiked into normal human plasma to evaluate analytical sensitivity. Total nucleic acid was extracted with an easyMAG^TM instrument (BioMerieux, St. Laurent, Canada). Extracted nucleic acid was applied to linearity, limit of detection and precision studies of the artus^® CMV PCR assay. Specificity experiments with other human herpesviruses were also performed. To compare the clinical sensitivity of the artus^® CMV PCR to the pp65 antigenemia assay, we tested 68 plasma specimens collected from 10 unique patients over a three month period.

Results: The artus^® assay was linear over a range of 500 to 50,000 copies / ml of CMV (log₁₀ = 2.70 to log₁₀ = 4.70) and the limit of detection was 50 copies / ml (log₁₀ = 1.70). No cross reactivity was observed with other human herpesviruses or normal human plasma. In our evaluation with clinical specimens, 6/10 (60%) patients had at least one positive test for CMV, with 5/6 patients demonstrating at least one positive CMV result in both tests. Overall, 17.6% (12/68) of specimens were positive for pp65 and 32.4% (22/68) were positive by the artus^® CMV PCR. Each patient that was positive by the artus^® PCR had multiple positive specimens, but 9/22 (40.9%) of the positive specimens were below the linear range of the assay.

Conclusions: These results suggest that CMV DNA can be detected in the plasma of patients at risk for CMV disease, but that a large proportion of these patients have a level of virus that is not accurately quantified by the artus^® CMV PCR. The clinical significance of these low level positives remains undefined, but may be important in monitoring the progression of CMV infection in immunocompromised patients.