Under the current epidemic conditions, as a golden standard to diagnose COVID-19 suspected cases, real-time fluorescent RT-PCR targeting SARS-CoV-2 is routinely implemented in most large- and medium-sized laboratories in China (14–16). Several types of specimens can be used to detect SARS-CoV-2 nucleic acids, such as respiratory nasal swabs, oropharyngeal swabs, sputum, and others (17–19). However, many factors might affect SARS-CoV-2 nucleic acid detection's analytical performance, such as types of virus transport media, methods of specimen pretreatment and template preparation, specimen storage temperature, and duration. Conventionally, it is better to use SARS-CoV-2 positive specimens to evaluate the performance of real-time fluorescent RT-PCR kits (8, 9). However, considering the biosafety and lack of positive specimens in most laboratories outside the Hubei province in China, it is critical to find alternative methods to evaluate SARS-CoV-2 nucleic acid detection's analytical performance.
As the virus infects human respiratory epithelial cells, SARS-CoV-2 particles could be co-extracted with human cells containing the IRC genes that could be detected by most commercially available real-time fluorescent RT-PCR kits (10, 11). Therefore, the abundance of IRC genes could be theoretically used to represent the extraction or preparation efficiency of SARS-CoV-2 nucleic acids. Compared with conventional strategies using specimens from COVID-19 confirmed cases, the current evaluation methods using specimens from healthy individuals are much more straightforward, safer, and more efficient to perform evaluations, namely in routine clinical laboratories that cannot implement demanding biosafety measures. Based on the above considerations, the specimens of oropharyngeal swabs collected from healthy individuals were used to evaluate the analytical performance of SARS-CoV-2 nucleic acid detection, in which the indexes were the ΔCq values between IRC genes of a specimen and QC. Serial experiments were conducted to evaluate various factors that might affect analytical performance, such as methods of specimen pretreatment and template preparation, types of virus transport medium, storage temperature, and duration. Being extremely valuable in theoretical and practical significance, the aforementioned ΔCq values' analytical performances were confirmed by pseudovirus and COVID-19 specimens.
Compared with both MB and OS methods of template preparations, the CF methods showed the best analytical performance. Using a more extensive collection of initial volumes of specimens, the CF methods exhibited theoretical advantages and practical results compared to OS methods (Fig. 1–4). Although using the same initial volumes of specimens, the CF methods also showed better performance than the MB methods (Fig. 1–4), which might be associated with the relatively lower recovery rate of MB methods (19). Notably, in our clinical practice, the CF methods also exhibited reliable results for one another commercial available real-time fluorescent RT-PCR kit (Daan Gene, Guangdong, China; data not shown).
In the experiments using templates containing pseudovirus fragments, the Cq values of IRC or N genes were close, which further confirmed that the MB methods had low extraction efficiencies because larger initial volumes of specimens were used in MB methods. However, compared with MB and OS methods, it was interesting to observe that the CF methods exhibited lower amplification efficiency for N genes (Fig. 4A and 4C), which was inconsistent with the data for the IRC genes (Fig. 1–3 and 4B). The reasons for these results might be that the pseudovirus particle could not be deposited well in CF methods, which would result in a low amplification. However, the above problems do not exist in clinical specimens. The reason might be that the SARS-CoV-2 mainly existed in the epithelial cells or adhered to the cell surface. Therefore, the SARS-CoV-2 particles could be co-deposited with the epithelial cells when CF methods were used (10, 11). The abovementioned speculation was confirmed by subsequent experiments using clinical COVID-19 specimens, in which both IRC and N genes exhibited better amplification than the CF method, and the amplification of ORF1ab genes was exhibited only in CF methods (Fig. 4D). The previous results suggested that methodology evaluations based on human IRC genes could be implemented as acceptable strategies to explore the analytical performance of SARS-CoV-2 nucleic acid detection.
A standard limitation of the PCR-based method is failed amplification due to the presence of PCR-inhibitory substances in the specimens, such as the heme compounds found in blood, aqueous and vitreous humor, heparin, urine, or polyamines, among other compounds (20). Based on the current comparative analysis on various types of virus transport media, it was essential to select the kinds of virus transport media when the CF or OS method of template preparation is planned to be used in real-time fluorescent RT-PCR targeting specific pathogens. For example, due to the presence of an amplification inhibitor, i.e., guanidine salt (21, 22), the analytical performance of the NX medium was always lower than that of both NS and YK media (Fig. 2A). Considering the biosafety of RT-PCR detection targeting SARS-CoV-2, it is better to inactivate pathogens using water or a metal bath before any procedures because there are no significant differences between the various specimens pretreatment (Fig. 1). For the other factors that might affect SARS-CoV-2 nucleic acid detection's analytical performance, it is better to implement the analysis in time (Fig. 3) and release SARS-CoV-2 particles from oropharyngeal swabs by vortex mixing (Fig. 2B).