Photo: Sarah Weldon / Shutterstock
The standard diagnostic test for severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) infection is the reverse transcriptase-polymerase chain reaction (RT-PCR) assay. Now, a new study published on the preprint server medRxiv* in July 2020 describes the use of saliva as a sample in RT-PCR for diagnosis in place of nasopharyngeal or oropharyngeal swabs.
The Need for a New Sampling Method
Since viral shedding is not continuous, multiple upper airway swabs may be needed to get an accurate result.
Secondly, the collection process for upper airway swabs often triggers sneezes and coughs, which could cause increased spread and put the individual off being tested again. Depending on the proficiency of the operator, the testing may also cause discomfort. Moreover, the technique is heavily dependent on the availability of swabs and testing media, and significant bottlenecks are already evident.
The current study thus deals with the possibility of using saliva as a non-invasive alternative sampling source. Many studies have already reported the increased sensitivity and consistency of results with saliva samples compared to upper airway swabs, but their small size was a limit to generalizability. Moreover, several of these studies did not report the specificity of the test or the limits of detection (LOD) of the virus.
The researchers aimed to study the sensitivity of the saliva, to establish the LOD, and to compare the results with upper airway swabs. The study included 110 adults between April and June 2020, of whom 61 were female, and 80% hospitalized at presentation.
This sample yielded 145 pairs of upper airway swabs and saliva samples, with 110 being collected at the time of diagnosis, 14 follow-up samples on day 2, and 6 on follow-up day 7. Only 15 participants showed up on day 28, by the time the study ended.
Of the total number of samples, there were 19 upper airway and 19 positive saliva samples. More of the positive samples came from day 2 and day 7 swabs, but the difference from the day 28 samples was not significant. The viral loads for saliva and upper airway swabs ranged from 36 to 5.4 x 107 copies/ml. However, the viral loads for saliva and upper airway swabs agreed in each pair.
Of the upper airway swabs, 13/110 were positive on day 0, but 4/14 and 4/6 on day 2 and day 7, respectively. With saliva, there were 11/110 positives on day 0, 5/14 on day 2, and 4/6 on day 7. One salivary sample was positive on day 28, of 15, but none of the upper airway swabs.
How Did the Paired Samples Agree?
There were 23 positives among the 145 paired samples, with 17 pairs being positive out of these. Three each were positive for either saliva only or upper airway swab only.
When the results are examined in more detail, out of 110 salivary samples collected on day 0, 12 were positive and 98 negative. Out of the 12 positives, 12 had a positive upper airway swab as well. Out of the 98 negatives, the upper airway swab was positive in 2. Thus, the upper airway swabs were positive in 14/110 cases and negative in 96 cases. There were 2 discordant results.
In the follow-up samples on day 2, 7, and 28, there were 8/35 positives among the salivary samples and 27 negatives. Out of the upper airway swabs, 6/35 were positive, and 29 negatives. There were 5 positive swabs showing agreement with the saliva samples and 1 positive swab in disagreement. Among the negatives upper airway swabs, 3 and 26 were in agreement and disagreement with the saliva sample results, respectively. Thus, a total of four follow-up swabs and salivary samples were in discordance.
Viral loads were low, at below 10 copies/ml in all the cases where the saliva was positive and the swab negative. The researchers attribute this to the greater dilution in the transport medium. The sourcing of the samples from individuals who had already tested positive at presentation shows that these saliva positives are true positives. Thus saliva is a more reliable source for sampling over time. This finding agrees with earlier studies of higher sensitivity for saliva and more long-term positivity.
Saliva-negative samples in discordant samples could be due to the different processing methods, with saliva undergoing freeze-thaw cycles, which could reduce the quality of the RNA.
The researchers conclude, “Overall, saliva had a good agreement to NT swabs during hospitalization (D0, D2, and D7) and after recovery (D28).” The advantages of saliva collection are apparent, including ease of sampling, higher acceptability, and independence of operator skill or technique. The downside is the more mixed nature of saliva because of the many possible contaminants from food, toothpaste, or coughed up mucus, to name just a few.
Further studies will show how the salivary viral load varies with different storage and transport conditions when preserving fluids are not used. The most significant use of this sampling method may be in low-resource settings where the number of cases is still going up, as well as for the diagnosis of young children and in repeated testing of patients where this is necessary. Urgent research is required to establish the validity of this method and allow it to be offered in certified laboratories for clinical use.
medRxiv publishes preliminary scientific reports that are not peer-reviewed and, therefore, should not be regarded as conclusive, guide clinical practice/health-related behavior, or treated as established information.