key: cord-0819643-7f52ofkn
authors: Marais, G. J. K.; Naidoo, M.; Hsiao, N.-y.; Valley-Omar, Z.; Smuts, H.; Hardie, D. R.
title: The implementation of a rapid sample preparation method for the detection of SARS-CoV-2 in a diagnostic laboratory in South Africa
date: 2020-08-07
journal: nan
DOI: 10.1101/2020.08.06.20169276
sha: b7cc42da9f72e4716c28411522a740dc1b0aad90
doc_id: 819643
cord_uid: 7f52ofkn

The SARS-CoV-2 pandemic has resulted in shortages of both critical reagents for nucleic acid purification and highly trained staff as supply chains are strained by high demand, public health measures and frequent quarantining and isolation of staff. This created the need for alternate workflows with limited reliance on specialised reagents, equipment and staff. We present here the validation and implementation of such a workflow for preparing samples for downstream SARS-CoV-2 RT-PCR using liquid handling robots. The rapid sample preparation technique evaluated, which included sample centrifugation and heating prior to RT-PCR, showed a 97.37% (95% CI: 92.55-99.28%) positive percent agreement and 97.30% (95% CI: 90.67-99.52%) negative percent agreement compared to nucleic acid purification-based testing. This method was subsequently adopted as the primary sample preparation method in the Groote Schuur Hospital Virology Diagnostic Laboratory in Cape Town, South Africa.

Severe Acute Respiratory Syndrome Coronavirus 2 (SARS-CoV-2), an emergent 37 betacoronavirus, was identified as a novel causative agent of severe pneumonia in Wuhan, 38

China in 2019 [1] . The capacity for person-to-person transmission was soon identified and the 39 ensuing pandemic has caused more than seventeen million cases at the time of submission [2] . 40 41 Currently, diagnostic testing for SARS-CoV-2 relies on molecular techniques, primarily 42 reverse-transcriptase polymerase chain reaction (RT-PCR), from respiratory specimens [3] . 43

The specialised equipment and reagents required to offer these tests at scale has placed 44 . CC-BY-NC-ND 4.0 International license It is made available under a perpetuity.

is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint

The copyright holder for this this version posted August 7, 2020. . https://doi.org/10.1101/2020.08.06.20169276 doi: medRxiv preprint 3 significant strain on worldwide supply chains of reagents. Public health measures put in place 45 in numerous countries, including travel restrictions, have further made planning for sustainable 46 service delivery difficult as laboratory stock orders may not be filled on time. These issues 47 motivate for the use of diagnostic workflows that favour locally or readily available reagents 48 to, at least partially, insulate supply chains from fluctuations in global demand and evolving 49 travel limiting public health measures. To address these issues, a number of laboratories have 50 successfully developed alternative sample preparation techniques which limit reagent needs 51 and avoid complex nucleic acid (NA) purification protocols [4] [5] [6] . There is also a significant 52 cost saving when the reagent-free direct heating method, as described by Fomsgaard and 53 Rosenstierne [4] , is used which will become critical if economic fallout from the pandemic 54 intensifies. Staff shortages in the laboratory are an inevitability as social distancing 55 requirements are implemented in concert with increasing demand for diagnostic testing. SARS-56

CoV-2 outbreaks in the laboratory environment may also introduce unpredictable shortages of 57 critical staff further limiting the capacity of laboratories to offer predictable test turnaround 58 times. The necessary influx of new staff, who may have limited training or training in a related 59 field, can further compromise the reliability of diagnostic laboratory services as the capacity 60 for oversight and quality control is hindered by rapidly evolving testing demands and workflow 61 instability due to reagent shortages and potentially unreliable testing kits due to limited 62 regulatory oversight [7] . All these factors highlight the need for automated workflows that limit 63 the number of laboratory staff-dependent steps and in particular steps requiring specialised 64 training. Automation further limits human error such as sample switches and cross-65 contamination and are generally amenable to greater degrees of workflow control due to 66 traceable instrument log files. is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint

The copyright holder for this this version posted August 7, 2020. is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint

The copyright holder for this this version posted August 7, 2020. Standard diagnostic testing sample preparation included placing NP or OP swabs in a 2ml 100

Sarstedt sample tube containing 1.5ml autoclaved 0.9% saline. If both a NP and OP swab or 101 multiple swabs of the same type was received, they were combined in a single tube. The swabs 102 were cut to fit in the tube. The tube was then vortexed for 10 seconds. The saline was used as 103 the sample input for downstream assays after which the tube was stored at 4°C. Stored tubes 104 from diagnostic samples were available for inclusion in the study. 105 106 Selected sample tubes were centrifuged at 16 000 g for 5 minutes and 50µl of the supernatant 107 was then pipetted into the wells of a 96-well PCR plate. The PCR wells were capped and the 108 plate incubated on a thermocycler at 98°C for 5 minutes followed by 4°C for 2 minutes. The is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint

The copyright holder for this this version posted August 7, 2020. is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint

The copyright holder for this this version posted August 7, 2020. is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint

The copyright holder for this this version posted August 7, 2020. . https://doi.org/10.1101/2020.08.06.20169276 doi: medRxiv preprint 8 assay. The dilution at which SARS-CoV-2 RNA could be detected with 95% confidence was 146 determined for each method by Probit analysis. The absolute analytical sensitivity of the RSP 147 method was calculated based on the relative analytical sensitivity compared to NA purification-148 based detection. Table 3 . One 163 sample that tested positive and 6 samples that tested negative were excluded. Raw data is 164 shown in the S1 Appendix. is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint

The copyright holder for this this version posted August 7, 2020. . https://doi.org/10.1101/2020.08.06.20169276 doi: medRxiv preprint Table 3. Contingency table used is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint

The copyright holder for this this version posted August 7, 2020. . https://doi.org/10.1101/2020.08.06.20169276 doi: medRxiv preprint dependent RNA-polymerase (RdRp) gene and Nucleocapsid (N) gene targets had Ct values 180 that were significantly different with a P value of <0.0001 (Fig 1) is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint

The copyright holder for this this version posted August 7, 2020. . https://doi.org/10.1101/2020.08.06.20169276 doi: medRxiv preprint

A novel coronavirus from patients with pneumonia in China

Geneva: WHO; Data as received by WHO from national authorities by 291 10:00 CEST

The laboratory diagnosis of COVID-295 19 infection: current issues and challenges

An alternative workflow for molecular detection of 298

SARS-CoV-2-escape from the NA extraction kit-shortage

and 24 April 2020 that were prepared for testing by NucliSENS® easyMag® NA purification. 196 The mean Ct difference values for each gene target (E, RdRp, N) were added to their respective 197 target Ct value. If this generated a value above 40, that gene target was deemed unlikely to be 198 detected after RSP. If all three gene targets were deemed unlikely to be detected, the sample 199 was classified as an expected negative. 198 is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprintThe copyright holder for this this version posted August 7, 2020. is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprintThe copyright holder for this this version posted August 7, 2020. The relative analytical sensitivity of the Allplex™ 2019-nCoV assay after RSP was found to 237 be 807 RNA copies per reaction. This was calculated from the 8.07-fold decrease in analytical 238 sensitivity of the RSP method compared to NucliSENS® easyMag® NA purification-based 239 testing, which has an analytical sensitivity of 100 RNA copies per reaction as per the Allplex™ 240 2019-nCoV assay package insert. The relative decrease was determined by serially diluting and 241 testing a sample with multiple replicates as shown in Table 4 . This relative loss in analytical 242 sensitivity can largely be explained by the decrease sample input volume for RSP. 243 is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprintThe copyright holder for this this version posted August 7, 2020. is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprintThe copyright holder for this this version posted August 7, 2020. . https://doi.org/10.1101/2020.08.06.20169276 doi: medRxiv preprint

The performance characteristics were deemed acceptable for clinical diagnostic use in the 253 Groote Schuur Hospital Virology Diagnostic Laboratory and allowed the laboratory to increase 254 the number of samples tested daily by a factor of 5-10 due to the decreased supply chain 255 It is made available under a perpetuity.is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprintThe copyright holder for this this version posted August 7, 2020. is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprintThe copyright holder for this this version posted August 7, 2020. . https://doi.org/10.1101/2020.08.06.20169276 doi: medRxiv preprint