nueva página del texto (beta)
Inglés (pdf)
Artículo en XML
Referencias del artículo
Traducción automática
Enviar artículo por email
Citado por SciELO 
Similares en
SciELO 
Permalink
Introduction
After the shutdown period caused by the coronavirus disease (COVID-19) pandemic, economic activities are resuming while the number of new cases is still growing in some countries. Moreover, a future second wave of infections is a latent possibility as well. Diagnostic testing is crucial to suppress the spread of COVID-19 in people who have symptoms of respiratory illness, as well as in asymptomatic exposed subjects, and convalescent patients who may still be infectious. The current gold-standard test is reverse transcription-polymerase chain reaction (RT-PCR) using a nasopharyngeal swab (NPS)1. It detects parts of the severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) genome explicitly from a few hundreds of viral particles in a swab sample2,3. However, NPS must be obtained by a trained health worker; also, it is uncomfortable for the patient who may bleed, sneeze or cough, posing a high risk of virus transmission to the person who is taking the sample and is affected by the global shortage of swabs. In addition, children may be uncooperative, increasing the risk of nasal trauma1.
Saliva is an attractive alternative specimen because the patient can collect it in a non-invasive way, and the results from saliva samples complied with those from NPS in 31-100% of samples obtained from positive patients4-13. The SARS-CoV-2 particles in saliva can result from the migration of infected cells to the lower and upper respiratory tract14 or from infection in the oral cavity, especially the tongue15.
Diagnostic testing of NPS is a well-established procedure1-3. However, little is known about the optimal condition to test saliva. Collection of specimens has been obtained by passive drooling (100% match between NPS/saliva)8, spitting the first saliva in the morning (92% match)7, coughing up saliva from the throat before spit (85-92% match)4,5,9,11, with the use of buccal swabs (82% match)13, and directly from the salivary gland canal (31% match)10. Passive drooling may be more suitable for children because not all of them can easily cough up, and buccal swabs may be convenient in infants who are too young to spit.
The high analytical sensitivity of the RT-PCR test may be hampered in the clinic due to time-dependent changes in nasopharynx and mouth viral load. For example, in patients incubating the virus, those in which viral shedding is migrating from upper to lower respiratory tract, or individuals recovering from the infection14.
Knowledge about the sensitivity of diagnostic testing in saliva along the course of infection may be useful to adopt this biofluid as an alternative specimen and to address the potential of virus transmission by infected persons. In the present study, changes in the capacity of saliva to report the presence of the SARS-CoV-2 virus were evaluated as a function of storage conditions of the sample, conditions of the mouth, and evolution of infection.
Methods
Participants and sample collection
All participants were positive in an NPS RT-PCR assay. Saliva specimens from two hospitalized children and two adolescents at the Hospital Infantil de Mexico Federico Gómez were collected. Participants were asked to gently spit approximately 2 ml of saliva in sterile 50 ml centrifuge plastic tubes. From the two adolescent participants, a unique saliva sample was collected in their first day of hospitalization and aliquots were kept at 4°C, −30°C or at room temperature up to 48 h. Two children provided several samples during their hospitalization. After collection, closed tubes were kept in a hard plastic box at room temperature until RNA was extracted (no more than 4 h). One adult recovering at home provided two samples per day: one of the first salivae in the morning and the other after brushing teeth, having breakfast, and brushing teeth again for 4 consecutive days. In this case, closed tubes were placed into a plastic bag with a hermetic seal, which in turn was placed into a closed hard plastic box and kept in the fridge. On completion of the sampling, the closed box was transported to the laboratory.
Virus detection
Viral RNA was extracted with the spin column-based Kit RNAv-090 from BioPure. For RT and amplification of the viral genes, the GeneFinder COVID-19 PLUS RealAmp Kit (MFK-45, GeneFinder) was used. If Ct (cycle threshold) was < 40 for at least one viral gene, the result was positive. If only one viral gene was amplified with Ct > 40, the result was undetermined. If the internal control showed a clear amplification and any viral gene showed a signal, the result was negative.
Results
The clinical characteristics of participants are shown in table 1.
Table 1 Clinical characteristics of participants
| Age (years) | Gender | Days after onset of respiratory symptoms* | Other morbidities | Experiment |
|---|---|---|---|---|
| 16 | Male | 5 | Recovering from renal transplant | Stability at different temperatures (Fig. 1) |
| 15 | Male | 4 | Acute lymphoblastic leukemia | Stability at different temperatures (Fig. 1) |
| 4 | Male | 4 | Hepatoblastoma | Saliva sampling along hospitalization (Table 2) |
| 4 | Female | 5 | Hydronephrosis | Saliva sampling along hospitalization (Table 2) |
| 56 | Male | 3 | None | Before and after tooth brushing and breakfast (Fig. 2) |
*Before the first sample was taken.
Storage of saliva samples at room temperature, 4°C or −30°C preserves the viral genome for 2 days
Aliquots of samples from two participants were analyzed the day of collection and after storage, for 1 and 2 days at room temperature, in the fridge, or the freezer to test the stability of the viral genome in saliva during 48 h after collection. Positive results were consistently obtained in all samples without changes in the Ct values of viral genes (Fig. 1).
Figure 1 Saliva preserves the viral genome. The saliva of two participants was aliquoted and kept at different temperatures before RT-PCR analysis. The corresponding Ct values were plotted as a function of time. Participant 1: room temperature; , 4°C; and ◻, −30°C. Participant 2: , room temperature; , 4°C; and , −30°C. The amplified viral genes were (A) Orf-1ab, (B) S, and (C) N; (D) hRP as an internal control. RT-PCR: reverse transcription-polymerase chain reaction; Ct: cycle threshold; hRP: human RNAse P.
The SARS-CoV-2 virus can be detected in saliva a few minutes after brushing teeth and feeding
One of the aspects to consider when using saliva as a specimen is if the first salivae in the morning are better than those collected at any hour of the day and if perturbations in the oral cavity affect the sensitivity of the assay. Therefore, the presence of the virus was compared between samples of the first saliva in the morning, and approximately 15 min after having breakfast and brushing teeth for 4 consecutive days by an adult participant recovering at home. Paired samples were available only for days 1, 2, and 4, in which the viral genome was detected in both conditions with similar Ct values (Fig. 2A). On day 3, the sample of the first saliva in the morning was negative and not enough for analysis after having breakfast and tooth brushing.
Figure 2 Detection of the viral genome in the saliva is not affected by brushing teeth and feeding. (A) The viral gene N or (B) the internal control RNAse P from human samples were RT-PCR amplified using the first saliva in the morning before (black bars) or after oral hygiene and having breakfast (grey bars) for four consecutive days. On day three, the first saliva in the morning was negative, and not enough sample was collected for analysis after oral hygiene and breakfast.
The SARS-CoV-2 virus was not consistently detected in the saliva of two children during their hospitalization
Saliva may be a good specimen to report the presence of the virus during the early stages of infection, and optimally, it should work during the disease progression and should become negative when the patient is not infected anymore. Therefore, samples from two children were analyzed during their hospitalization periods (5 and 12 days, respectively). In the first case, the test resulted positive only on day one (Table 2). In the second case, three out of six samples resulted positive on days 1, 2, and 6 (Table 2).
Table 2 Time course of SARS-CoV-2 detection in the saliva of two hospitalized children
| Days of hospitalization | |||||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| 1 | 2 | 3 | 4 | 5 | 6 | 7 | 8 | 9 | 10 | 11 | 12 | 13 | |
| Patient 1 | + | U | − | − | − | NT | NT | NT | NT | NT | NT | NT | NT |
| Patient 2 | + | + | NT | NT | − | + | NT | NT | NT | U | NT | NT | − |
U: undetermined result; SARS-CoV-2: severe acute respiratory syndrome coronavirus 2; NT: not tested.
The use of spin columns kits for viral RNA extraction from saliva is a feasible option for the SARS-CoV-2 detection assay
The high viscosity of some saliva samples prevents accurate pipetting, and its high protein content may clog to some extent the silica resin of spin columns during the extraction procedure. Despite this, using raw saliva without any transport media or additive, we could detect the presence of the viral genome. Overall, we analyzed samples from five different participants. As diagnosis was initially confirmed with a positive result from an NPS, we obtained a 100% agreement between the first saliva sample taken from each participant and the NPS test. This result is similar to those reported in other studies using spin columns8 or magnetic particles-based automated technology for viral RNA isolation4-7,9,11,13 (Table 3).
Table 3 Comparison of SARS-CoV-2 detection in NPS and saliva.
| Match between saliva/NPS results | Sample collection | Viral RNA extraction procedure | Age group | Reference |
|---|---|---|---|---|
| 85% (49/58) | Cough up the first saliva in the morning. Posterior oropharyngeal | Magnetic particles | Adults | 4 |
| 88% (37/42) | Deep cough 3-5 times before spitting | Magnetic particles | Adults | 5 |
| 85% (33/39) | Spit gently. Use of transport media 1:1 | Magnetic particles | Adults | 6 |
| 92% (35/38) | Spit the first saliva in the morning | Magnetic particles | Adults | 7 |
| 100% (25/25) | Passive drooling | Spin column kit | Adults | 8 |
| 87% (20/23) | First saliva in the morning. Posterior oropharyngeal | Magnetic particles | Adults | 9 |
| 31% (4/13) | Cotton swabs from salivary gland canal | | Adults | 10 |
| 92% (11/12) | Cough out saliva from throat | Magnetic particles | Adults | 11 |
| 73% (8/11) | | | Children | 12 |
| 82% (9/11) | Buccal swabs on the buccal mucosa (bilateral) | Magnetic particles | Children | 13 |
Discussion
Although saliva is an attractive option as a specimen for SARS-CoV-2 detection, several methodological issues must be systematically explored before it can be used in the clinic.
At present, the presence of the virus is addressed by the detection of its genome, although it does not necessarily mean the presence of active infective viral particles16. However, the stability of viral particles is desirable for diagnostic testing because it relaxes the conditions of storage and transport to the laboratory. Our results demonstrated that, for samples with Ct values within the range of positive results (26-32), the viral genome is stable for 2 days if kept at room temperature, in the refrigerator, or after a freeze-thaw cycle. These conditions cover the most common types of storage and transport of clinical samples. The RNase activity of saliva17 does not reach the viral genome, suggesting the presence of intact viral particles at the different tested temperatures.
Regarding conditions of the oral cavity, our results indicate that the first saliva in the morning is as good as that obtained ≈15 min after breakfast and brushing teeth for the detection of the virus. This result supports the notion that saliva is a robust biofluid for COVID-19 diagnosis. Regardless, these results must be confirmed with more participants. It is important to note that one out of the seven samples shown in figure 2 was negative (first saliva in the morning at day 3), indicating that even during the early days of infection, the saliva test is not 100% reliable.
In two reports of SARS-CoV-2 detection in the saliva of children12,13, the virus becomes undetectable in saliva by days 8-10 after the onset of symptoms, while it was still detectable in NPS. Han et al. reported a steeper decline of viral load in saliva compared to NPS12. Accordingly, one of the two children showed a very fast negativization. The other had a prolonged hospitalization, but positive results were consistently obtained only during the first two determinations (Table 2). These results are consistent with the idea that saliva may be a good option for COVID-19 diagnosis only during the early stages of infection. Further studies are necessary to determine the time-dependent changes of viral clearance in the mouth of children.
At present, most of the reported comparisons between NPS and saliva use magnetic particles-based automated technology to extract viral RNA (Table 3). However, not all laboratories have such instrumentation. The use of spin-column kits only needs a microfuge and, unquestionably, a biosafety level two laboratory setting. Although a high number of paired samples is required to compare the sensitivity of two extraction methods, our results indicate that the simple benchtop extraction method is sensitive enough to reveal the presence of the viral genome.
