COVID salivary diagnostics: A comparative technical study

Abstract Since the beginning of the coronavirus disease 2019 (COVID‐19) pandemic, molecular diagnostics of severe acute respiratory syndrome coronavirus 2 (SARS‐CoV‐2) have taken center stage in the detection of infected individuals for isolation purposes but also in the mass surveillance as a preventive strategy to contain the virus spread. While nasopharyngeal swabs (NPS) have remained the golden standard substrate, salivary diagnostic for SARS‐CoV‐2 has been proposed as an alternative and noninvasive measure in vulnerable individuals. Nevertheless, there is a widespread assumption that salivary reverse‐transcription polymerase chain reaction (RT‐PCR) does not match the quality of testing using NPS and particular care should be taken in respect to food or beverage intake, when sampling saliva. Our study indicates that without any precaution in the selection of 190 patients, nor restriction over the time window of sampling, there is 99% match in the COVID‐19 positivity between NPS and saliva when using RT‐PCR, with a reported Delta in thermal cycles (Cts) values for the viral genes Envelope (E) and Open reading frame 1ab (Orf1ab) between 0 and 2, a 98.7% sensitivity and 100% specificity. This high accuracy is maintained in pooling configurations that can be used for mass‐testing purposes in professional and educational settings. The further advantage to using crude saliva as compared to NPS or mouthwash is that direct methods yield robust results. Overall, our study validates and promotes the use of salivary diagnostic for COVID‐19 eliminating the need of a medical practitioner for the sampling, resolving the unpleasantness of the NPS intervention and empowering the patient to do self‐testing in times of need.

Nasopharyngeal swabs (NPS) have been widely used as collection material in first-line diagnostics of COVID-19. Besides the invasiveness of the approach, limiting the repeatability of the testing, in particular, for vulnerable age segments, the NPS vary significantly based on the operator, the site of sampling, and the viral load of the swabbed tissue. 4 The shortage of NPS as collection agents profoundly affects the sensitivity of reverse-transcription polymerase chain reaction (RT-PCR) testing resulting in up to 30% of false negatives. There is numerous evidence that saliva could be used as an alternative and noninvasive biofluid suitable for SARS-Cov2 genetic diagnostics. [5][6][7] Saliva is a peripheral biofluid produced by the parotid, sublingual, and submandibular glands and rich in DNA, RNA, and proteins. The salivary glands, in particular the parotid glands due to proximity of the nasopharyngeal tract, function as a sink for respiratory pathogenic species that are released in saliva making this biofluid ideal for pathogen testing. Furthermore, recent evidence indicates that the time or method of sampling does not interfere with the results of the genetic test, and saliva can be kept at 4°C for 24 h without degradation giving great flexibility in the biofluid sampling and triaging. 8 We have also previously demonstrated that the saliva can be used as a microbial classifier for dementia progression 9 and it is also found to detect Influenza viruses A and B. 10 This and other reports strongly support the use of saliva for detecting infectious diseases. Several Universities and Institutions around the world have demonstrated that saliva COVID-19 diagnostic is the only acceptable way of routine monitoring and controlling the virus spread by performing more efficient contact tracing in the long run 11 and preparing for the flu season in the fall/winter. Based on the accumulated evidence and with the outbreak of the highly infectious Omicron variant, in the winter of 2021, the CDC has recommended the use of saliva testing over NPS, to facilitate mass-testing. 12 The primary objective of this study has been to evaluate whether saliva is a surrogate biofluid to NPS in symptomatic individuals without control about the collection time or oral care and to assess the clinical performance and analytical sensitivity of RT-PCR detection of SARS-Cov-2 in saliva compared to that of NPS.
Secondly, our study aimed to compare the simple heat-released RNA protocol (extraction-free) to the magnetic beads and spin column RNA extraction protocol followed by RT-qPCR in both NPS and saliva samples. Finally, we evaluated the sensitivity of different pooling strategies for saliva samples in detection of SARS-CoV-2 and assessed an optimal pool size for mass testing in public establishments, according to the measure adopted by the Swiss government during the pandemic surveillance. were patients with severe respiratory symptoms, inability to give consent, inability to follow procedures or insufficient knowledge of project language. Severity of the respiratory disease was not reported. The viral prevalence of SARS-Cov-2 in this period was between 2% and 10% in the tested incoming patients.

| Sample collection
NPS were performed on all subjects following the routine procedure 13 and sent twice to the laboratory for analysis and stored at −80°C for later analysis.
For saliva collection, subjects were asked to visualize a lemon to stimulate salivation and donate saliva. A volume corresponding to 1-2 ml of whole unstimulated saliva was collected in 50 ml sterilized conical tubes and placed in cool boxes. The collected saliva once arrived in the lab was stored at −80°C for later analysis.   Percentages of analytical sensitivity, specificity, positive predictive value, and negative predictive value with their 95% confidence intervals (CI) were calculated using the "Jeffreys" method. In all three genes, E, RdRp/S, and N, over the whole patient population a mismatch between saliva was visible at Cts higher than 30, indicating either a differential appearance of virus across the two biofluids or a discrepancy due to experimental procedure. Considering the more robust Gene E, the sensitivity of salivary diagnostic as compared to NPS was 83% and vice versa, and the specificity was 94.8% intrinsic to the biofluid. At Ct values of lower or equal 34, there were three and five samples positive for GeneE but not the others in saliva and NPS, respectively. However, when considering samples with Cts lower or equal to 30 (quantitative load identified as noncontagious) the results were the same for all the three genes and the two biofluids were more comparable with matching between saliva and NPS of 100% and a specificity of 100%.  virus with RNA extraction (Table 3 and Figure 4B), which is not the case for saliva (Table 3 and Figure 4A) and suggests that magnetic  (Table 4).

| Comparison of SARS-CoV
Overall, the better performance of the extraction-free preanalytical method for RT-PCR applies only to crude saliva and not mouthwash, prompting us to use in native saliva only for salivary detection of SARS-CoV-2 in clinical settings for mass-screening.

| Pooling saliva samples for mass population screening
To confirm the reliability of the bulk pooling method, two RT-PCR kits   Figure 6A) and Liferiver (Table 5, Figure 6B) kits, respectively, in an asymptotic way as expected by the RT-cycling method. The results confirmed the reliability of using pooling systems 5 (Delta of 1.9 and 3.6 Cts compared with individual by using Mikrogen and Liferiver kits, respectively) to 25 (Delta of 3.9 and 6.2 Cts compared to individual by using Mikrogen and Liferiver kits, respectively) to perform mass screening for example in schools (Table 5, Figure 6). Once a positive pool is identified, all samples have to be tested individually.
This is only possible when the pooling is performed in the laboratory and not externally.

| DISCUSSION
The COVID-19 pandemic has raised an urgent need of searching for fast and reliable scientific procedures, rapid result delivery and proper sample collection. In this study, we considered the important advantages of using saliva specimens in SARS-CoV-2 detection. The saliva has our attention due to its several advanced characteristics: (1) noninvasive specimen, (2) easier and safer collection, (3) a good reservoir of viruses, (4) possibility for self-collection at home, and (5) its convenience for repetitive collection in both adults and children.
Saliva has previously proved to be an ideal organic fluid for isolation In addition, as shown saliva COVID-19 molecular diagnosis is sensitive to the extraction method, with spin columns as less suitable than the bead-based or heating-based RNA extraction process.
In this study we also found a heating method compared to the extraction method, the heat treatment assay allows (1)  The direct method has also the advantage of allowing preanalytical pooling with high reliability using groups of 5-25 when doing mass surveillance in the asymptomatic population, taking a prevalence of the virus below 1%. Our experience indicates that pools of 8 saliva specimens provide an optimal balance between costs and sensitivity for mass-testing of SARS-Cov-2 in asymptomatic children. Another study previously showed similar results with saliva pools of 5 and 10 suitable with a disease prevalence of 9%. 20 Overall, our study shows that saliva is a reliable biofluid for COVID-19 detection. Saliva coupled with the direct heating method can also be used as an efficient tool for mass testing, since hundreds of samples can be diagnosed in a few hours, obtaining reliable results.
This mass testing should be applied for repetitive testing in hospitals, schools, companies, and public organizations with great convenience for the subjects and at a limited cost when using pooled samples.

| CONCLUSIONS
In our study, we presented the accuracy of salivary COVID diagnosis compared to the conventional NPS, the so-called golden standard for COVID testing by performing RT-PCR. Our study highlights that saliva coupled with the heat-released RNA method represents a very reliable tool in routine COVID-19 diagnostics. The results of the study indicate that direct saliva molecular diagnosis provides a 99% diagnostic accuracy in detecting SARS-Cov-2 infections and is a suitable replacement for NPS RT-PCR. A pooling method on saliva samples can also effectively apply in mass testing strategy as monitoring SARS-CoV-2 will remain a public health need in the coming years.