Applicability of saliva in the diagnosis of COVID-19 : a review

In 2020, the World Health Organization (WHO) classified COVID-19 as a global pandemic. Since then, there is a need for new methods to facilitate the diagnosis and control of this disease. Currently, reverse transcription followed by real-time polymerase chain reaction (rRT-PCR) of respiratory samples obtained by swabs represents the gold standard in the qualitative detection of Sars-CoV-2 infection. However, this type of collection has several disadvantages, making saliva a potential tool for the diagnosis of COVID-19. Thus, the aim of this study is to evaluate, through a systematic review of current scientific literature, the applicability of saliva for the diagnosis of COVID-19 in comparison to current methods. A search was carried out in MEDLINE, SciELO, Scopus and Web of Science databases, using descriptors, strategies and pre-established criteria by two independent evaluators, followed by a manual search in the references of articles selected for full reading. The research strategies identified 476 studies and 1 study was added through manual search. After analysis, 200 articles were excluded because they were duplicated among results found in databases. With the completion of the screening process, 12 articles were included in this review. It was concluded that it is necessary to produce new studies in order to obtain even more reliable and effective data about the use of saliva in the diagnosis of COVID-19. However, studies have shown that this material can be an excellent alternative sample for the detection of SARS-CoV-2.


Introduction
In December 2019, a series of cases of pneumonia was observed in the city of Wuhan, specifically in the Chinese province of Hubei, with clinical presentations very similar to viral pneumonia. Subsequently, analysis of the genomic sequencing of samples from the lower respiratory tract indicated the appearance of Coronavirus Disease 2019 , caused by the severe acute respiratory syndrome virus 2 (Sars-CoV-2) (Huang et al., 2020;Randad et al., 2020).
On March 11, 2020, the World Health Organization (WHO) classified COVID-19 as a global pandemic (WHO, 2020 (a)). Since then, the infection has spread rapidly around the world, mainly due to the high contagion rate of the virus (Bulut & Kato, 2020). Until July 14, 2020, the COVID-19 pandemic was responsible for affecting approximately 13 million individuals and causing more than 570 thousand deaths (WHO, 2020 (b)).
According to current research, reverse transcription followed by real-time polymerase chain reaction (rRT-PCR) of respiratory samples represents the gold standard in the qualitative detection of SARS-CoV-2 infection (Azzi et al., 2020). The rRT-PCR method applied in the Development, v. 9, n. 9, e636997991, 2020 (CC BY 4.0) | ISSN 2525-3409 | DOI: http://dx.doi.org/10.33448/rsd-v9i9.7991 4 diagnosis of COVID-19 is based on the amplification of viral RNA in several cycles until there is enough genetic material to be detected, with the purpose of effectively diagnosing individuals infected with SARS-CoV-2 (Mesa & Castillo, 2020).
However, although this monitoring standard is widely used in the world, it has a number of disadvantages, such as exposure of health professionals to a disease with high risk of nosocomial transmission during sample collection, excessive spending on personal protective equipment (PPE), discomfort for the patientsince swabs are inserted deep into the nose or mouth to collect naso and oropharyngeal samples, respectively, generation of aerosols by inducing coughing and sneezing, in addition to some situations in which collection is contraindicated, as in cases of coagulopathic patients (Tajima, Suda, & Yano, 2020;Ceron et al., 2020;Ng et al., 2020).
Saliva is a complex of multiglandular secretions mainly composed of peeled oral epithelial cells, gingival crevicular fluid, metabolites, hormones and electrolytes, in addition to large number of proteins, such as immunoglobulins. Since the current method has several disadvantages and is invasive, saliva can be an excellent alternative sample for the diagnosis of COVID-19, as it has been increasingly used for the purpose of assessing human health. Thus, with the development of appropriate methods of collecting and processing samples, saliva will provide useful clinical information about the disease, facilitating diagnosis, management and control of COVID-19 (Ceron et al., 2020;Ng et al., 2020;Woźniak, Paluszkiewicz, & Kwiatek, 2019).
Studies have shown that saliva can be a reliable tool in the diagnosis of COVID-19. For Pasomsub et al. (2020) and Azzi et al. (2020), the saliva RT-PCR test has high sensitivity and performance comparable to the current most used method, showing the importance of saliva and the need for further research to confirm its potential diagnostic value. Therefore, saliva could facilitate the diagnosis of the disease, given the simplicity of sample collection and good diagnostic performance (Martina et al., 2020).
Thus, the present study aims to evaluate, through a systematic review of the current scientific literature, the applicability of saliva for the diagnosis of COVID-19 compared to the current commonly used methods.

Methods
The electronic search for articles was carried out in July 2020 in MEDLINE (via PubMed), Scientific Electronic Library Online (SciELO), Scopus and Web of Science Development, v. 9, n. 9, e636997991, 2020 (CC BY 4.0) | ISSN 2525-3409 | DOI: http://dx.doi.org/10.33448/rsd-v9i9.7991 5 databases, including studies published from 2015 to June 2020, without restriction of language and country. Descriptors used were "Saliva", "Salivary gland", "COVID-19" and "Coronavirus", obtained from the Medical Subject Headings (MeSH) directory, used in combination. The Table 1 expresses the results obtained in the survey carried out in the databases previously mentioned. In addition to restrict results, "articles published in the last 5 years" and "fully available free of charge" filters were selected. For the inclusion of studies, the following criteria were considered: 1-having experimental character; 2-treating saliva as potential non-invasive diagnostic method for COVID-19; 3-presenting detailed clinical information about the research carried out. On the other hand, the following exclusion criteria were considered: 1-duplicated articles; 2-those that did not meet pre-established parameters; 3-those not considering saliva as potential sample for the diagnosis of COVID-19.
After identifying articles and eliminating duplicates, the first phase of the selection consisted of analyzing the title and summary of publications. Thus, the aim was to select, at this stage, those that had in their titles any of the keywords previously established or terms relevant to the theme. In addition, with regard to reading the abstracts, those that fit the theme and met the pre-established inclusion criteria were chosen.  Development, v. 9, n. 9, e636997991, 2020 (CC BY 4.0) | ISSN 2525-3409 | DOI: http://dx.doi.org/10.33448/rsd-v9i9.7991 6 Finally, studies proceeded to the full reading phase, completing screening with 12 articles to compose this review. In addition, manual search was performed on the references of selected studies during the full reading stage in order to identify those that included the established prerequisites, but that were not included in the electronic search.
All searches were carried out by 2 independent evaluators. Articles that raised doubts were sent to a third reviewer.

Results
The search strategies in databases identified 476 studies and 1 study was added through manual search. After analysis, 200 articles were excluded because they were duplicated among results found in databases. After screening based on the reading of titles and abstracts, 256 studies were excluded, leaving 21 publications to be evaluated through full reading. After this stage, 12 articles were included in this review. The Figure 1 presents the whole trial and selection process of the articles. Research, Society and Development, v. 9, n. 9, e636997991, 2020 (CC BY 4.0) | ISSN 2525-3409 | DOI: http://dx.doi.org/10.33448/rsd-v9i9.7991 The main aspects of the 12 studies selected at the end of the last stage of the screening process are shown in Table 2. A number of advantages of the use of saliva in the diagnosis of COVID-19 were identified, in addition to the versatility regarding sample collection, configuring the potential use of this type of sampling for mass testing. Body fluid specimens were collected from the patients from hospital day 1 to 9, besides additional samples of the saliva were taken at 1 hour, 2 hours, and 4 hours after using a chlorhexidine mouthwash. The severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) viral load was determined by real-time reverse transcriptase polymerase chain reaction (rRT-PCR).
SARS-CoV-2 was detected from all the five specimens of both patients by rRT-PCR. The viral load was the highest in the nasopharynx, but it was also remarkably high in the saliva. SARS-CoV-2 was detected up to hospital day 6 (illness day 9 for patient 2) from the saliva of both patients. The viral load in the saliva decreased transiently for 2 hours after using the chlorhexidine mouthwash.
SARS-CoV-2 viral load was consistently high in the saliva; it was relatively higher than that in the oropharynx during the early stage of COVID-19. Chlorhexidine mouthwash was effective in reducing the SARS-CoV-2 viral load in the saliva for a short-term period.
Hung et al. (2020) To investigate the ideal time for the collection of saliva, speculating that a sample in the early morning, before oral hygiene and breakfast, would increase the diagnostic yield.
18 patients with previously confirmed SARS-CoV-2 infection by molecular testing.
Posterior oropharyngeal saliva was collected at 5 different time points within the same day from 18 patients with previously confirmed SARS-CoV-2 infection by molecular testing. Cycle threshold (Ct) values were compared.
There was an overall trend of lower Ct values from specimens collected in the early morning, with a gradual decrease of viral load towards nighttime, but reaching statistical significance only when compared with the specimens collected at bedtime. Eight out of 13 subjects had a higher viral load in the early morning than the rest of the

Discussion
Currently, reverse transcription followed by real-time polymerase chain reaction (rRT-PCR) of respiratory samples collected by swabs represents the gold standard in the qualitative detection of Sars-CoV-2 infection (Azzi et al., 2020). However, this type of collection has several disadvantages, making saliva a potential material for diagnosis due to its non-invasive nature. In this way, diagnosis, management and control of COVID-19 can be facilitated ; Ceron et al., 2020).
Saliva is an easily accessible fluid that can be self-collected through a non-invasive procedure and can be beneficial for mass-scale seroprevalence studies. In addition, virus entry is mainly via the upper respiratory tract and antibodies contained in saliva can provide a first barrier to the entry of Sars-CoV-2. As a result, saliva has been studied as a potential diagnostic tool and is expected to replace other materials, such as serum or naso and oropharyngeal smears for the diagnosis of COVID-19 (Faustini et al., 2020;Sri Santosh et al., 2020).
For these reasons, it is of fundamental importance to consider factors that may influence the analysis of viral load when using this specimen for the diagnosis of COVID-19. Assessing the best method and time for collection, in addition to the choice and purity of the antigen, are essential elements for a more accurate identification of the presence of Sars-CoV-2. In this sense, using saliva from the posterior oropharynx is considered a promising method due to the sensitivity and the viral load rate found in this type of sample, which is justified by the predilection of the virus by the respiratory tract ; To et al., 2020 (b);Zheng et al., 2020;Yoon et al., 2020;Hung et al., 2020;Chen et al., 2020).
Other authors also consider the use of saliva obtained by spitting into a container for analyzing the presence of the virus (Tajima et al., 2020;Pasomsub et al., 2019;Zheng et al., 2020;Yoon et al., 2020;Wyllie et al., 2020). Considering the methods above, it is worth Development, v. 9, n. 9, e636997991, 2020 (CC BY 4.0) | ISSN 2525-3409 | DOI: http://dx.doi.org/10.33448/rsd-v9i9.7991 16 mentioning that saliva is an extremely valuable material for the diagnosis of the disease, especially due to the easy collection by patients themselves, making it a potential method for obtaining more real data regarding the pandemic, avoiding underreporting by promoting mass testing of the population.
However, other methods were also considered for the acquisition of diagnostic material from the same fluid. In studies carried out by Azzi et al. (2020) and Faustini et al. (2020), samples collected from the 25 and 63 included patients, respectively, were obtained using techniques called drooling technique and passive dribble. This method enables collecting only oral fluids, excluding secretions from the lower respiratory tract or from the oropharynx mucosa (Golatowski et al., 2013). In addition, Randad et al., (2020) evaluated samples obtained from saliva collected from the space between the gum and the tooth, which is enriched with gingival crevicular fluid (GCF). For this reason, the composition of this material is very similar to serum due to the presence of elements of the immune system in the GCF (Taylor & Preshaw, 2016;Brandtzaeg, 2013). Furthermore, with regard to the comparison between the main collection methodsspitting and sputum -the results obtained by Yoon et al. (2020) suggest greater sensitivity in tests that analyzed sputum samples. However, this result cannot be affirmed with total certainty due to the limited number of patients included in the study, and further studies are necessary in order to concretely establish this information.
Currently, some studies have shown the existence of better times for the collection of saliva, but little is known about the most advantageous time for this collection. In studies carried out by Tajima et al. (2020) and Hung et al. (2020), morning saliva samples showed higher viral load in the early morning compared to other times of the day, corroborating the hypothesis that morning saliva has greater sensitivity for the diagnosis of COVID-19. Although several factors affect the viral load rate, the advantage of collecting saliva in this period is justified because, during sleep, the ciliary movement of the trachea and the descent of nasopharynx fluids cause this sample to be contaminated with sputum, thus increasing the amount of viruses. In addition, the authors agree that collection should be carried out before breakfast and tooth brushing in order to obtain the greatest efficiency. Although the study sample is very small, its results are very promising and must be taken into account, but further studies should be carried out to investigate the best time for saliva collection.
In addition, some authors highlight the high sensitivity and the comparable or even superior performance of saliva in relation to the current method that uses swabs (Pasomsub et al., 2019). According to studies by Wyllie et al. (2020), the excellent performance of saliva occurs mainly when there is early hospitalization and becomes more consistent during hospitalization and prolonged recovery. In the same study, it was observed that 2 asymptomatic health professionals showed negative swabs results; however, the detection of SARS-CoV-2 was found when they performed tests with saliva, which suggests, according to them, that saliva may be an appropriate and even more sensitive alternative in the process of identifying asymptomatic and pre-symptomatic patients.
Similar results were obtained in studies by Azzie et al. (2020), performed with sample collected from individuals previously tested positive for SARS-CoV-2 using rRT-PCR with saliva samples. Of the group composed of 25 patients, 2 of them had negative results for swabs in tests performed on the same day, although saliva confirmed the infection. This suggests that patients discharged from hospital, after negative results for respiratory swabs, can transmit the virus through saliva as a result of inconsistent results.
There are several advantages of using saliva for the diagnosis of COVID-19. As samples are easily supplied by patients, the contact between contaminated individuals and health professionals is reduced, which consequently reduces the risks of nasocomial transmission. For the same reason, the use of PPE by professionals during collection is not necessary, thus reducing excessive spending on public health. In addition, unlike saliva and due to the use of deep swabs, the current method not only generates discomfort, but can also induce coughing and / or sneezing, responsible for the production of aerosols (Tajima et al., 2020;; Ceron et al., 2020;Ng et al., 2020).

Conclusions
When considering that most studies do not have high number of patients, it is not possible to state, in fact, that saliva can be used as the main diagnostic material. However, studies have shown that saliva can be excellent alternative sample for the detection of SARS-CoV-2, facilitating the diagnosis and control of COVID-19. From this perspective, further studies should be developed in order to obtain even more reliable and effective data on tests performed and sample collection methods.