ABSTRACT Objectives To assess the performance of newly developed polymerase chain reaction (PCR) primers to detect severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) RNA, using gel electrophoresis and sequencing. Our results were compared against those obtained with the primers developed by Charité Berlin and ones commercially available in the ApplexTM SARS-CoV-2 assay. Design Evaluation study Setting This evaluation study was conducted at the Erasmus MC an academic hospital in the southwest of the Netherlands. Samples were obtained from a Medical Diagnostic Center also stationed in the South-West of the Netherlands that offers routine microbiology diagnostics (e.g., serology, molecular testing, bacterial cultures) for approximately 1,500 primary health care facilities. The primer sequences were designed by BioCoS, a biotechnology company providing bioinformatics services for biomarker discovery and primer design. Participants 150 symptomatic patients suspicious for a SARS-CoV-2 infection who presented themselves at a general practitioner or at a geriatric specialist were included. Main outcome measures Presence or absence of SARS-CoV-2 RNA in oro-nasopharyngeal swabs as detected by RT-(q)PCR, gel electrophoresis and sequencing of the PCR amplicons after which the positive predicted value (PPV), negative predicted value (NPV), positive percentage agreement (PPA) and negative percentage agreement (NPA) of each primerset was determined. Results Gel electrophoresis of RT-(q)PCR amplicons and sequencing methods demonstrated that the newly discovered and designed triplet STAMINA primersets by BioCoS in the ORF1ab (PPV,100%; NPV, 80%), E- (PPV 100%; NPV 73.85%) and N-gene (PPV 100%; NPV 60%) harbored an increased PPA compared to the triplet Charité Berlin primersets designed in the RdRp- (PPV 100%; NPV 67.61%), E- (PPV 100%; NPV 71.64%) and N-gene (PPV 96.97%; NPV 39.17%), by using the AllplexTM SARS-CoV-2 assay as a criterion standard. Moreover, calculating the PPA by using our own constructed composite reference as a standard confirmed that the STAMINA primersets outperformed the Charité Berlin primersets, which came with a trade-off in NPA. Sequencing of the RT-(q)PCR amplicons revealed the presence of aspecific products e.g., Homo sapiens, bacteria and viruses other than SARS-CoV-2, but excluded the presence of related coronaviruses in the amplicons generated with the STAMINA primersets. Conclusion This evaluation study reveals that reliable detection of SARS-CoV-2 RNA using RT-(q)PCR critically depends on primer design and PCR test parameters. Moreover, our work shows that the newly developed primers, despite outperforming the ones designed by Charité Berlin in PPA, are still suboptimal to detect SARS-CoV-2 RNA.

Confidential sequences will be published in a later version