Comparison of the Ct-values for genomic and subgenomic SARS-CoV-2 RNA reveals limited predictive value for the presence of replication competent virus.

SARS-CoV-2 is the causative agent of the acute respiratory disease COVID-19. In addition to the full length positive-sensed, single-stranded genomic RNA (gRNA), viral subgenomic RNAs (sgRNAs) that are required for expression of the 3' region of the genome are synthesized in virus-infected cells. However, whether these sgRNA-species might be used as a measure of active virus replication and to predict infectivity is still under debate. The commonly used methods to monitor and quantitate SARS-CoV-2 infections are based on RT-qPCR analysis and the detection of gRNA. The infectivity of a sample obtained from nasopharyngeal or throat swabs is associated with the viral load and inversely correlates with Ct-values, however, a cut-off value predicting the infectivity highly depends on the performance of the assay. Furthermore, gRNA derived Ct-values result from nucleic acid detection and do not necessarily correspond to active replicating virus. We established a multiplex RT-qPCR assay on the cobas 6800 omni utility channel concomitantly detecting SARS-CoV-2 gRNA, sgRNA, and human RNaseP-mRNA used as human input control. We compared the target specific Ct-values with the viral culture frequency and performed ROC curve analysis to determine the assay sensitivity and specificity. We found no advantage in the prediction of viral culture when using sgRNA detection compared to gRNA only, since Ct-values for gRNA and sgRNA were highly correlated and gRNA offered a slightly more reliable predictive value. Single Ct-values alone only provide a very limited prediction for the presence of replication competent virus. Hence, careful consideration of the medical history including symptom onset has to be considered for risk stratification.