key: cord-0718537-7je1i8qd authors: Muenchhoff, Maximilian; Mairhofer, Helga; Nitschko, Hans; Grzimek-Koschewa, Natascha; Hoffmann, Dieter; Berger, Annemarie; Rabenau, Holger; Widera, Marek; Ackermann, Nikolaus; Konrad, Regina; Zange, Sabine; Graf, Alexander; Krebs, Stefan; Blum, Helmut; Sing, Andreas; Liebl, Bernhard; Wölfel, Roman; Ciesek, Sandra; Drosten, Christian; Protzer, Ulrike; Boehm, Stephan; Keppler, Oliver T title: Multicentre comparison of quantitative PCR-based assays to detect SARS-CoV-2, Germany, March 2020 date: 2020-06-18 journal: Euro Surveill DOI: 10.2807/1560-7917.es.2020.25.24.2001057 sha: 8558091193008b80490f4ed52ae0e00aac4be05e doc_id: 718537 cord_uid: 7je1i8qd Containment strategies and clinical management of coronavirus disease (COVID-19) patients during the current pandemic depend on reliable diagnostic PCR assays for the severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2). Here, we compare 11 different RT-PCR test systems used in seven diagnostic laboratories in Germany in March 2020. While most assays performed well, we identified detection problems in a commonly used assay that may have resulted in false-negative test results during the first weeks of the pandemic. Containment strategies and clinical management of coronavirus disease (COVID-19) patients during the current pandemic depend on reliable diagnostic PCR assays for the severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2). Here, we compare 11 different RT-PCR test systems used in seven diagnostic laboratories in Germany in March 2020. While most assays performed well, we identified detection problems in a commonly used assay that may have resulted in false-negative test results during the first weeks of the pandemic. Strategies to limit the severe pandemic and to manage coronavirus disease (COVID-19) patients strongly depend on readily available, accurate and reliable RT-PCR assays to detect the genome of the causative agent acute respiratory syndrome coronavirus 2 (SARS-CoV-2) in biosamples. The first full-length SARS-CoV-2 genome sequence was made publicly available in early January 2020 [1] and, soon after, various RT-PCR assays were reported by academic laboratories, public health agencies and diagnostics companies [2] [3] [4] [5] [6] . Their overall performance and relative sensitivity are largely unclear. The aim of this study was to compare the inter-laboratory and inter-method sensitivity of different RT-PCR assays by providing a blinded, frozen dilution series of a nucleic acid extract of a highly positive biosample to seven different diagnostic laboratories in Germany in March 2020. Nucleic acids were pooled from multiple extractions of one SARS-CoV-2-positive stool sample using the QIAsymphony DSP Virus/Pathogen Kit (Qiagen, Hilden, Germany). This stool sample was from a 5-year-old child with COVID-19 [7] and was chosen because of high initial PCR signals and sufficient sample availability to generate large quantities of eluate for further distribution. Of note, no PCR inhibition was observed for detection of the spiked-in extraction RNA control (QuantiNova IC Probe Assays Red 650, Qiagen). A 1:10 dilution series was prepared and aliquots were labelled in a blinded fashion to be shipped on dry ice to participating laboratories in March 2020. Participants were instructed to perform the diagnostic assays used at their centre for SARS-CoV-2 detection in quadruplicate using 5 µL of the aliquot per reaction. All results were reported back to the initiating laboratory (Laboratory 1) before the results were unblinded. The details of all these PCR-based assays are summarised in Table 1. In parallel, samples were quantified using the One- Step RT-digital droplet (dd)PCR Advanced Kit for Probes (BioRad, Feldkirchen, Germany) on the BioRad QX200 platform. Primer and probe sequences were used for detection of the SARS-CoV-2 nucleocapsid gene (N) as published by the Centers for Disease Control and Prevention (CDC) [2] Digital droplet PCR quantification of the distributed dilution series of nucleic acid eluate of SARS-CoV-2-positive clinical material, Germany, March 2020 The undiluted sample showed between 4,325 and 5,015 SARS-CoV-2 RNA copies per reaction using 5 µL of eluate for the CDC N1, N2, N3 and Charité E protocols, but only 850 and 1,951 RNA copies for the Charité N and P primer/probe combinations ( Figure 1A) , respectively, indicating a lower sensitivity of the latter. The 1:10 dilution series displayed good linearity down to a calculated concentration of 0.4 RNA copies per reaction at the 10 −4 dilution for both the CDC N1 and N2 primer/probe combinations ( Figure 1B ). Result interpretations from the seven participating laboratories are summarised in Figure 2 displaying the number of replicates scored positive by the respective laboratory for each method and dilution. Most methods reliably detected the sample at the 10 −3 dilution, which is equivalent to ca 5 RNA copies for the CDC N1, N2, N3 and Charité E reactions based on the absolute quantification by ddPCR. Of note, the Seegene Allplex 2019-nCoV Assay gave negative results for all four replicates in the E gene at the 10 −3 dilution, while reporting positive results for N and RdRp (Laboratory 1). According to the manufacturer's instructions at the time of analysis, this would have been interpreted as an inconclusive result. Of note, the RdRp primer/probe did not show any positive result at the 10 −4 dilution. Driven by false-negative results for samples with low PCR-positivity using the original Charité RdRp reaction (see below and others [8, 9] ), we compared the primer/ probe sequences with currently available SARS-CoV-2 genomes. When compared with all genomes available on GISAID (9,184 SARS-CoV-2 genomes on 15 April 2020, Supplement), the regions used for amplification in the CDC and Charité protocol are highly conserved: Only 1.55%, 0.45% and 2.4% of genome sequences contain any kind of mismatch within the primer/ probe regions of the CDC N1, N2 and N3 protocols, respectively, and 0.25%, 0.29% and 0.67% in the primer/probe regions of the Charité E, RdRp and N pro tocols, respectively. The Charité RdRp reverse primer contains an ambiguity base at position 15,519 that does not match the reference sequence (Wuhan-Hu-1/2019), with an S (i.e. G or C) instead of T for the reverse complement (Supplementary Figure S1 ). The other ambiguity base at 15,528 showing Y (i.e. C or T) should be changed to T because the currently circulating viruses have a T at this position and no polymorphisms were detected in any of the 9,184 sequences submitted to date (accession date: 15 April 2020). Based on computation using Primer Express v3.0 (Applied Biosystems, Dreieich, Germany) annealing temperatures were predicted to be 64 °C for the RdRp forward and 51 °C for the RdRp reverse primer of the Charité protocol. This temperature difference may result in reduced PCR efficiency. To address this issue, modified RdRp primers were synthesised as shown in Supplementary Figure S1 and tested in comparison with the original primers. Testing the dilution series with these modified RdRp primers (see above and Supplementary Figure S1 ) yielded positive results for two additional dilution steps (10 -3 and 10 -4 ) compared with the original Charité RdRp primers ( Figure 1 ). To further compare the sensitivity of these modified RdRp primers with the original version of the Charité RdRp primers and the Charité E and the CDC N1 reaction, we retested 28 eluates of clinical respiratory specimens from the diagnostic unit at Laboratory 1 that had shown crossing point (Cp) values > 35 in the initial CDC N1 reaction. Using the original version of the confirmatory Charité RdRp primers, 16 of 28 samples tested negative, but 11 of these showed positive results using the modified primers ( Figure 3) . Overall, the detection by the Charité E, modified Charité RdRp, and CDC N1 reactions were robust. Notably, six and seven of these 28 respiratory samples scored negative or at the limit of detection (Cp = 40) in the Charité E and modified Charité RdRp reactions, while only one sample came up negative in retesting in the CDC N1 reaction (p = 0.04 and p = 0.02, chi-squared-test comparing Charité E and modified RdRp to CDC N1, respectively). Of note, in a routine clinical setting, the CDC N1 reaction also detected SARS-CoV-2 RNA in nucleic acid extracts from 37 of 83 sera (45%) from COVID-19 patients in intensive care units, with a positive correlation of their Cp values with those of the corresponding respiratory material (Spearman Rank correlation co-efficient r=0.4285, p (two-tailed) < 0.0001 (data not shown)). The majority of RT-PCR assays for SARS-CoV-2 examined in this study detected ca 5 RNA copies per reaction, reflecting a high sensitivity and their suitability for screening purposes world-wide. A reduced sensitivity was noted for the original Charité RdRp gene confirmatory protocol, which may have impacted the confirmation of some COVID-19 cases in the early weeks of the pandemic. The protocol needs to be amended to improve the sensitivity of the RdRp reaction. The CDC N1 primer/probe set was sensitive and robust for detection of SARS-CoV-2 in nucleic acid extracts from respiratory material, stool and serum from COVID-19 patients. A new coronavirus associated with human respiratory disease in China Information for laboratories about coronavirus (COVID-19) Detection of 2019 novel coronavirus (2019-nCoV) by real-time RT-PCR Coronavirus disease (COVID-19) technical guidance: Laboratory testing for 2019-nCoV in humans Comparison of seven commercial RT-PCR diagnostic kits for COVID-19 Rapid establishment of laboratory diagnostics for the novel coronavirus SARS-CoV-2 in Bavaria Clinical and epidemiological features of a family cluster of symptomatic and asymptomatic SARS-CoV-2 infection Analytical sensitivity and efficiency comparisons of SARS-COV-2 qRT-PCR primer-probe sets Comparative performance of SARS-CoV-2 detection assays using seven different primer-probe sets and one assay kit We acknowledge the authors, originating and submitting laboratories of the sequences from GISAID's EpiCov database on which this research is based. All submitters of data may be contacted directly via the GISAID website (www.gisaid.org). This work was supported by LMUexcellent funding of LMU München. We thank Mikrogen for participating in this investigator-initiated study. HM, DH, NA, RK, MW, SZ performed experiments. RW, DH, AB, HR, AS, BL, SC, CD, UP analysed data. HB, AG and SK assembled SARS-CoV-2 genome sequences and extracted variants for the analysis of the primer binding regions. MM and OTK wrote the manuscript.All authors discussed the results and commented on the final manuscript. None declared. MM, HM, HN, NG-K, SB and OTK designed the study. MM, This is an open-access article distributed under the terms of the Creative Commons Attribution (CC BY 4.0) Licence. You may share and adapt the material, but must give appropriate credit to the source, provide a link to the licence and indicate if changes were made.Any supplementary material referenced in the article can be found in the online version.