Amplicon size and non-encapsidated DNA fragments define plasma cytomegalovirus DNA loads by automated nucleic acid testing platforms: A marker of viral cytopathology?

ABSTRACT

Management of cytomegalovirus (CMV) in transplant patients relies on measuring plasma CMV-loads using quantitative nucleic acid testing (QNAT). We prospectively compared the automated Roche-cobas®6800-CMV and Roche-CAP/CTM-CMV with laboratory-developed Basel-CMV-UL54-95bp, and Basel-CMV-UL111a-77bp. Roche-cobas®6800-CMV and Roche-CAP/CTM-CMV were qualitatively concordant in 142/150 cases (95%). In-depth comparison revealed higher CMV-loads of the laboratory-developed assay and correlated with smaller amplicon size. After calibration to the 1.WHO-approved CMV international standard, differences were reduced but remained significant. DNase-I pretreatment significantly reduced CMV-loads for both automated Roche-CAP/CTM-CMV and Roche-cobas®6800-CMV assays, whereby 90% and 95% of samples became undetectable. DNase-I pretreatment also reduced CMV-loads quantified by Basel-CMV-UL54-95bp and Basel-CMV-UL111a-77bp, but remaining detectable in 20% and 35%, respectively. Differences were largest for 110 samples with low-level CMV-DNAemia being detectable but not-quantifiable by Roche-cobas®6800-CMV, whereby the smaller amplicon sizes yielded higher viral loads for concordant positives. We conclude that non-encapsidated fragmented CMV-DNA is the major form of plasma CMV-loads also measured by fully-automated platforms. Amplicons of <150 bp and calibrators are needed for reliable and commutable QNAT-results. We hypothesize that non-encapsidated fragmented CMV-DNA results from lysis of CMV-replicating cells and represent a direct marker of viral cell damage, which contribute to delayed viral load responses despite effective antivirals.