[Problems of cytomegalovirus diagnostics by real-time polymerase chain reaction.]

The main goal was to explain the discrepancies between two PCR methods used for detection of cytomegalovirus (CMV) in peripheral blood samples of patients of Department of Internal Medicine-Hematooncology, University Hospital, Brno.

In past we used exon 4 of major immediate-early (MIE) gene as the target for quantitative detection of the CMV in clinical samples, but sometimes this method failed to detect the viral load in samples that were positively tested using less sensitive qualitative method targeting another region (exon 2-4) of the same gene. From January 2004 to January 2005 we totally tested samples from 363 patients. 64 patients were at least once CMV positive using quantitative method, but 20 patients were repeatedly false negative.To find the cause of this discrepancy we performed partial sequence analysis of this region (nt positions 2719-2919, GenBank M21295) and glycoprotein B (gB) genotyping. We sequenced samples from 35 patients-15 giving true positive (both in qualitative and quantitative method) and 20 giving false negative (negative in quantitative but positive in qualitative method) results in several consecutive blood samples.

The 15 true positive samples were 100% homological, whereas all 20 false negative samples showed high degree of variation from the laboratory strain AD169. These changes are not random and indicate that the two groups of patients were infected by different CMV genotypes. Moreover, sequence alignment showed similarity to laboratory strains Toledo and Towne. No preferential concordance was observed between clinical context, MIE exon 4 sequence and gB groups.

Because of high sequence variability exon 4 of MIE gene can not be used for routine diagnostics. Genetic varibility among the pathogenic strains may seriously affect its proper diagnostics.