The natural history and transmission potential of asymptomatic SARS-CoV-2 infection

Vietnam contaiment approach

Since January 2020, various control measures, including isolation of confirmed cases, contact-tracing, airport quarantine and social distancing have been implemented in Vietnam with increasing stringency as the pandemic progressed worldwide (Figure 1) [14, 15]. Accordingly, anyone known to have been in contact with a confirmed COVID-19 case, or having travelled to Vietnam from a COVID-19 affected country, were isolated for ≥14 days at a designated isolation center.

• Download figure
• Open in new tab

Figure 1:

Timelines of containment strategies applied in Vietnam and in Ho Chi Minh City since the beginning of 2020 as the epidemic/pandemic progresses alongside the implementation of SARS-CoV-2 RT-PCR testing and the duration of the clinical study.

Note to figure 1: *Initially China, followed by Korea, other European counties (Italy, France, UK etc) and the USA

From the second week of March 2020, all isolated individuals were subject to serial SARS-CoV-2 nasopharyngeal throat swab (NTS) screening by real time RT-PCR. A confirmed case was established if two independent RT-PCR assays (E gene and RdRP RT-PCR assays) were positive [16]. Confirmed cases were admitted to a designated COVID-19 hospital for follow-up until they recovered and/or had at least two consecutive days with negative SARS-CoV-2 RT- PCR NTS [17].

Setting

The Hospital for Tropical Diseases (HTD) is a tertiary referral infectious diseases hospital responsible for receiving and treating patients with COVID-19 in southern Vietnam. From January 2020 to the first week of April, HTD was responsible for RT-PCR screening of 80% of quarantined people in HCMC.

In addition to its main campus in the centre of HCMC, HTD has two designated 300-bed centres for the care of confirmed/suspected cases with COVID-19, namely Cu Chi and Can Gio Hospitals, located approximately 60 km to the West and East, respectively, of HCMC (Figure 2A). The present study was conducted at Cu Chi Hospital.

• Download figure
• Open in new tab

Figure 2:

Settings and the clinical study.(A) Map showing the location of the Hospital for Tropical Diseases (HTD) main campus and its designated COVID-19 centres in Cu Chi where the clinical study was conducted, and Can Gio, Flowchart illustrating the results of RT-PCR screening of quarantined people between March 10^th, 2020 and April 5^th, 2020, and the enrolment of patients in the clinical study

Note to Figure 2:

*Extrapolated from data extracted from the HTD database system. During this period, HTD tested a total of 11,052 cases, accounting for 80% of isolated people in HCMC.

**The remaining cases were either treated at the main campus of HTD in HCMC or at the other designated isolation center (Can Gio Hospital).

^#One was transferred to the main campus of HTD 10 hours after admission and two were transferred from the main campus of HTD after 5 and 6 days of hospitalization and were not enrolled because of enrolment competition.

Maps were obtained from https://mapchart.net/.

Patient erollment and data collection

We enrolled individuals with confirmed SARS-CoV-2 infection admitted to Cu Chi Hospital from March 10^th to April 4^th, 2020. From each participant, we prospectively collected demographic and clinical data, travel history and information concerning contact with confirmed COVID-19 cases, using standardized paper case record forms.

We collected NTS, combining them into a single tube containing 1ml of viral transport medium. NTS were taken daily from enrolment to hospital discharge (Figure 2B). Additionally, a saliva sample was obtained at enrolment. After collections, clinical samples were stored at −4°C at the study site and were then transferred to the HTD laboratory in HCMC within 4 hours for analysis.

Viral RNA extraction and SARS-CoV-2 RT-PCR analysis

We manually extracted viral RNA from 140 ul of NTS and saliva samples (if volume was sufficient for testing) using the QIAamp viral RNA kit (QIAgen GmbH, Hilden, Germany), and then recovered the cleaned up RNA in 50 ul of elution buffer provided with the kit. Since we enroled patients who had a confirmed diagnosis by two independent RT-PCRs (Egene and RdRp assays) as per the WHO recomendation [16], we used E gene assay for testing of samples collected from enrolment onward. Real time RT-PCR were carried out as previously described [16].

Data analysis

For viral-load associated analysis, in the absence of quantitative RT-PCR results, we use cycle threshold (Ct) values as surrogates. We used the t-test to compare the difference in measured Ct values obtained at enrolment between the two groups, Wilcoxon signed-rank test to compare the measured Ct values between NST and saliva, and chi-square test to compare two proportions. We compared the trend in the detection probability of SARS-CoV-2 and the viral RNA load in NTS between asymptomatic and symptomatic individuals. For the detection probability, we fitted a logistic regression model that quantifies the probability to test positive over time. We used generalized estimating equations (geepack package in R [18]) to correct for the repeated measurements per individual. We assumed that those who left the study earlier, after several days with a negative test result, remained negative until day 19. For the trend in Ct, we used a zero-inflated mixed effects model for semi-continuous data (GLMMadaptive package in R). Since the Ct data have an upper threshold of 40.5, we used the transformation Y=40.5-Ct. Hence, an undetectable viral load was given the value zero. We report the mean value on the original scale, which is a weighted combination of the value 40.5 for those that test negative and the measured value of those that test positive. Additionally, we compared the measured Ct values between the two groups; for this we used a random effects model. Further details can be found in the supplementary code. Note that inference with respect to measured Ct values should be interpreted with caution, because the subgroup of negative values is selectively excluded. Apart from R, we also performed the analysis in SPSS V23.0 (IBM Corp, NY, US) and generated the figures using GraphPad PRISM^® V5.04 (GraphPad Software Inc, CA, US) and R.

Ethics

This clinical study received approvals from the Institutional Review Board of the HTD and the Oxford Tropical Research Ethics Committee of the University of Oxford. Study participants gave their written informed consent.

RT-PCR screening of quarantined people

Between March 10^th and April 4^th, 2020, approximately 14,000 people were referred to one of nine designated quarantine centres deployed across HCMC, and were screened for SARS-CoV- 2 by RT-PCR of NTS. Forty-nine people had a positive test, accounting for 96% (49/51) of all reported cases in HCMC during the same period. The other two self-presented to local hospitals after falling ill. Of these 49, 33 (67%) were admitted to Cu Chi Hospital, and 30 (30/33, 91%) agreed to participate in the clinical study (Figure 2B).

Baseline characteristics of the study participants

Of the 30 study participants, 16 were imported cases (i.e. they acquired the infection outside of Vietnam, Supplementary Figure 2), and 14 acquired the infection locally; all 14 had an epidemiological link with two community transmission clusters occurring in HCMC during the study period. Of the locally acquired infections, 7 (50%) were asymptomatic; while 6 (38%) of the imported cases were asymptomatic. Those with locally acquired infection were more likely to be male than those with infection acquired outside of Vietnam (Table 2).

Seventeen of the participants had mild respiratory disease; i.e. no requirement for supplementary oxygen during hospitalization. None of the participants developed severe disease. The other 13 patients had no symptoms or signs of infection throughout their hospital admission. The demographic and laboratory characteristics of the two groups were similar at enrolment (Table 1 and 2). A small proportion of symptomatic patients presented with diarrhea and/or lost their sense of smell. None of the 30 participants had abnormal findings on chest radiographs.

Viral detection in NTS and saliva samples

Compared with symptomatic patients, those with asymptomatic infection were less likely to have detectable SARS-CoV-2 in NTS samples collected at enrolment (8/13 (62%) vs. 17/17 (100%) P=0.02). However, 4/5 patients whose NTS collected at enrolment were negative had an NTS positive result in one of the subsequent sampling days, but with a high Ct value (Supplementary Figure 2), suggesting these patients had low viral load in their respiratory samples.

Of the 30 study participants, 27 (90%) had a saliva sample collected at enrollment with sufficient volume for RT-PCR analysis. SARS-CoV-2 RNA was detected in 20/27 (74%) available saliva, 7/11 (64%) in the asymptomatic and 13/16 (81%) in the symptomatic group (P=0.56). There was one patient, whose NTS collected at enrolment was negative but saliva was positive. Accordingly, a combination of both NTS and saliva samples collected at enrolment slightly increased the diagnostic yield of samples collected at enrolment of the asymptomatic group

Quantification of viral RNA in NTS and saliva at enrolment

At enrolment, among those who were RT-PCR positive, the viral loads measured in NTS and saliva were similar in asymptomatic and symptomatic patients (Figure 3A). However, among asymptomatic patients who had both saliva and NTS collected, higher viral load was observed in NTS than in saliva (P=0.031) (Figure 3B). A similar trend was observed for symptomatic cases (P=0.064) (Figure 3B).

• Download figure
• Open in new tab

Figure 3:

Comparison of Ct values of SARS-CoV-2 real time RT-PCR obtained from nasopharyngeal swabs (NTS) and saliva. (A) Data include results of RT-PCR analysis of all available NTS and saliva collected from 30 participants at enrolment, (B) Data only include results of RT-PCR analysis of paired NTS and saliva of 6 asymptomatic and 12 symptomatic patients who had both sample types collected at enrolment.

Quantification of viral RNA and duration of viral detection in NTS

During follow-up, Ct-values differed between the two groups (P=0.027 for difference over first 19 days, Figure 4A), with asymptomatic patients having lower viral load than the symptomatic patients. If restricted to RT-PCR positive samples, viral RNA abundance was similar to slightly lower in the asymptomatic participants (P=0.086) (Supplementary Figure 3).

Analysis of the probability of RT-PCR positivity showed asymptomatic participants had a lower probability of having a positive RT-PCR result (i.e. a faster viral clearance) than symptomatic participants (P<0.001 for difference over first 19 days, Figure 4B). This difference was most pronounced during the first week of follow-up. After this period, the probability of detection quickly fell to almost zero in both groups. The majority of the positive patients were weakly positive (Ct value >32) in this period.

• Download figure
• Open in new tab

Figure 4:

Trends in Ct values and viral detection probability in nasopharyngeal throat swabs over the course of hospitalization; (A) Changes of Ct values, relatively reflect the level of viral load, (B) Dynamics of viral detection probability from enrolment onward

Note to Figure 4: Each dot represents one observed value (A) or the mean value (i.e. frequency) per day (B), lines indicate mean and shades indicate 95% confidence interval.

Presumed transmission from asymptomatic carriers

Fourteen participants were identified to have an epidemiological link with two community-transmission clusters occurring in HCMC during the study period. Cluster #1 had three patients participating in the present study. Of these three, 2 had contact with a confirmed case on March 2^nd, who was not enrolled in this study because this patient was admitted to a different hospital. Subsequently, one developed fever, runny nose and sore throat on March 12^th, 2020, suggesting an incubation period of 10 days, and tested positive for SARS-CoV-2 on March 13^rd, 2020. The other had no fever or any signs/symptoms suggestive of infection and was positive for SARS-CoV-2 on March 14, 2020. Two days later, a colleague of these two cases developed mild respiratory symptoms, including runny nose and loss of sense of smell, and tested positive for SARS-CoV-2 on March 17^th, 2020.

Cluster #2 included 11 study participants, including 7 with asymptomatic infection (Figure 5). We identified a transmission chain involving an asymptomatic participant (patient #19) who was positive for SARS-CoV-2 on March 25 (Ct value of NTS: 24, and saliva: 28). Subsequently, a contact of this case (patient #22) was positive for SARS-CoV-2 on March 23^rd (Ct value of NTS: 23, and saliva: 34), although this contact did not develop symptoms. Furthermore, on March 27^th, a contact of both patient #19 and #22 (patient #27) presented with cough and sore throat, with positive NTS for SARS-COV-2. Additionally, patient #26, contact of patient #22, who was also a contact of patient #19, was confirmed with SARS-CoV-2 on March 20^th, also without any symptoms. An additional transmission chain from cluster #2 was recorded between patient #24 and #29, both of whom were asymptomatic (Figure 5).

• Download figure
• Open in new tab

Figure 5:

Illustration of cases with an epidemiological link with community transmission cluster #2.

Red circles indicate symptomatic patients, and blue circles indicate asymptomatic cases. Patients sitting on the open circle are people of the cluster who first had contacts with each other on the same day. Patient numbers correspond to the numbers presented in Supplementary Figure 2.