Comparative effects of viral-transport-medium heat inactivation upon downstream SARS-CoV-2 detection in patient samples

Abstract

Introduction. The COVID-19 pandemic, which began in 2020 is testing economic resilience and surge capacity of healthcare providers worldwide. At the time of writing, positive detection of the SARS-CoV-2 virus remains the only method for diagnosing COVID-19 infection. Rapid upscaling of national SARS-CoV-2 genome testing presented challenges: (1) Unpredictable supply chains of reagents and kits for virus inactivation, RNA extraction and PCR-detection of viral genomes. (2) Rapid time to result of <24 h is required in order to facilitate timely infection control measures.Hypothesis. Extraction-free sample processing would impact commercially available SARS-CoV-2 genome detection methods.Aim. We evaluated whether alternative commercially available kits provided sensitivity and accuracy of SARS-CoV-2 genome detection comparable to those used by regional National Healthcare Services (NHS).Methodology. We tested several detection methods and tested whether detection was altered by heat inactivation, an approach for rapid one-step viral inactivation and RNA extraction without chemicals or kits.Results. Using purified RNA, we found the CerTest VIASURE kit to be comparable to the Altona RealStar system currently in use, and further showed that both diagnostic kits performed similarly in the BioRad CFX96 and Roche LightCycler 480 II machines. Additionally, both kits were comparable to a third alternative using a combination of Quantabio qScript one-step Quantitative Reverse Transcription Polymerase Chain Reaction (qRT-PCR) mix and Centre for Disease Control and Prevention (CDC)-accredited N1 and N2 primer/probes when looking specifically at borderline samples. Importantly, when using the kits in an extraction-free protocol, following heat inactivation, we saw differing results, with the combined Quantabio-CDC assay showing superior accuracy and sensitivity. In particular, detection using the CDC N2 probe following the extraction-free protocol was highly correlated to results generated with the same probe following RNA extraction and reported clinically (n=127; R²=0.9259).Conclusion. Our results demonstrate that sample treatment can greatly affect the downstream performance of SARS-CoV-2 diagnostic kits, with varying impact depending on the kit. We also showed that one-step heat-inactivation methods could reduce time from swab receipt to outcome of test result. Combined, these findings present alternatives to the protocols in use and can serve to alleviate any arising supply-chain issues at different points in the workflow, whilst accelerating testing, and reducing cost and environmental impact.

Keywords: COVID-19; RT-qPCR testing; SARS-CoV-2; viral transport media.

Fig. 1.

Comparison of VIASURE and RealStar kits using RNA (a) C_T values generated from VIASURE ORF1ab and N gene probes and RealStar S and E probes. Individual points represent a single detected SARS-CoV-2 sample. Samples with C_T values above the dotted line are interpreted as negative. (b) △C_T for each sample was calculated relative to RealStar S C_T value to demonstrate differential probe sensitivity. Statistical differences for each gene were calculated using a one-way ANOVA with Dunnett’s multiple comparison post-hoc test comparing to clinical S (*P ≤0.05, ****P ≤0.0001). (c) Comparison of median and interquartile range for C_T values obtained using the VIASURE assay on two RT-qPCR machines, Bio-Rad CFX and Roche LC480 with no significant differences found between comparisons. (n=26).

Fig. 2.

Diagnostic kit comparison on borderline SARS-CoV-2 positive RNA samples. (a) C_T values of equivocal samples from three diagnostic workflows; RealStar, VIASURE and Quantabio XLT. Samples with C_T values above the dotted line, or ND (not detected) are interpreted as negative for SARS-CoV-2 RNA. Samplesranked by C_T value. (b) Total number of assay-probe sets that detected SARS-CoV-2 RNA per borderline sample. Results ranked by number of positive samples detected (n=18 purified patient samples).

Fig. 3.

C_T value comparison of heat inactivated material and extracted RNA from the same swab samples. Samples were heat treated at 75 °C for 10 min. RT-qPCR run with two primer-probe sets (n1 and n2) using Quantabio UltraPlex assay (n=7 patient samples).

Fig. 4.

Correlation between C_T values obtained by N1 and N2 UltraPlex assay-probe sets and Abbott assay generated C_T (n=127). Linear regression and R² correlation values for each probe are indicated.