key: cord-1030494-wu5lm2zd
authors: Costa, R.; Bueno, F.; Albert, E.; Torres, I.; Carbonell-Sahuquillo, S.; Barres-Fernandez, A.; Sanchez, D.; Padron, C.; Colomina, J.; Lazaro-Carreno, M. I.; Breton-Martinez, J. R.; Martinez-Costa, C.; Navarro, D.
title: Upper respiratory tract SARS-CoV-2 RNA loads in symptomatic and asymptomatic children and adults
date: 2021-03-03
journal: nan
DOI: 10.1101/2021.03.03.21252814
sha: 86d029849f51ebfb72444bc67d9b347a2b7c730d
doc_id: 1030494
cord_uid: wu5lm2zd

Objectives: There is limited information comparing SARS-CoV-2 RNA load in the upper respiratory tract (URT) between children and adults, either presenting with COVID-19 or asymptomatic. Here we conducted a retrospective, single center study involving a large cohort of SARS-CoV-2 infected individuals to address this issue. Patients and Methods: A total of 1,184 consecutive subjects (256 children and 928 adults) testing positive for SARS-COV-2 RNA in nasopharyngeal exudates (NP) were included, of whom 424 (121 children and 303 adults) had COVID-19 not requiring hospitalization and 760 (135 children and 625 adults) were asymptomatic close contacts of COVID-19 patients. SARS-CoV-2 RNA testing was carried out using the TaqPath COVID-19 Combo Kit (Thermo Fisher Scientific, MS, USA). The AMPLIRUN TOTAL SARS-CoV-2 RNA Control (Vircell SA, Granada, Spain) was used for estimating SARS-CoV-2 RNA loads (in copies/mL). Results: Median SARS-COV-2 RNA loads were comparable between adults and children with COVID-19 (7.14 log10 copies/ml vs. 6.98 log10 copies/ml; P=0.094). Median SARS-CoV-2 RNA load in asymptomatic children and adults was similar (6.20 log10 copies/ml vs. 6.48 log10 copies/ml; P=0.97). Children with COVID-19 symptoms displayed SARS-CoV-2 RNA loads comparable to their asymptomatic counterparts (P=0.61). Meanwhile in adults, median SARS-CoV-2 RNA load was significantly higher in symptomatic than in asymptomatic subjects (P=<0.001), yet comparable (P=0.61) when the analysis excluded patients sampled within 48 h after symptoms onset. Conclusions: The data suggest that children may be drivers of SARS-CoV-2 transmission in the general population at the same level as adults.

CC-BY-NC 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint this version posted March 3, 2021. ; https://doi.org/10.1101/2021.03.03.21252814 doi: medRxiv preprint onset. 48 . CC-BY-NC 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint this version posted March 3, 2021. ; https://doi.org/10.1101/2021.03.03.21252814 doi: medRxiv preprint

An increasing body of evidence suggests that children are less susceptible to SARS-54

CoV-2 infection and tend to develop milder forms of COVID-19 than adults [1] . 55

Nevertheless, whether children, either symptomatic or asymptomatic, play a major role 56 in community transmission of SARS-CoV-2 compared to adults remains unclear [1] . 57

There is a consistent direct correlation between magnitude of SARS-CoV-2 RNA load 58 in the upper respiratory tract (URT) and probability of recovering live virus in cell 59 culture, in both adults and children [2] [3] [4] [5] ; hence, viral load in URT may be used as a 60 proxy for contagiousness. Supporting this assumption, transmission risk was recently 61

shown to be strongly associated with initial SARS-CoV-2 RNA levels of index cases 62 [6] . There is scarce information on how SARS-CoV-2 RNA load in UTR compares 63 between children and adults [7] [8] [9] [10] [11] , whether viral load in pediatric subjects differ across 64 ages [9,10], and whether dissimilarities in the dynamics of SARS-CoV-2 shedding in 65 URT exist between symptomatic and asymptomatic children [12] [13] [14] . Elucidation of 66 these questions is critically important for designing effective public health policies to 67 fight the pandemic. Here, to gain a further insight into these issues, we conducted a 68 retrospective, single center study involving a substantial cohort of SARS-CoV-2 69 infected children and adults, either asymptomatic or symptomatic, non-hospitalized 70 cases. 71

. CC-BY-NC 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) We amplified the β glucuronidase (GUSB) housekeeping gene to assess specimen 106 cellularity in selected specimens following a previously published protocol [18] . In 

Differences between medians across groups were compared in a pairwise fashion using 114 the non-parametric Mann-Whitney U-test, given that SARS-COV-2 RNA loads were 115 non-normally distributed. Spearman's rank test was used to test the association between 116 age and SARS-CoV-2 RNA load. Two-sided exact P-values were reported. A P-value 117 <0.05 was considered statistically significant. The analyses were performed using SPSS 118 version 20.0 (SPSS, Chicago, IL, USA). 119

. CC-BY-NC 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint this version posted March 3, 2021. ; https://doi.org/10.1101/2021.03.03.21252814 doi: medRxiv preprint

We first compared initial SARS-CoV-2 RNA load in NP from symptomatic pediatric 122 and adult patients. Specimen collection was carried out at a median of 2 days (range, 0-123 10 days) and a median of 3 days (range, 0-10 days), respectively, after symptoms onset. 124

The data are shown in Fig. 1A . The range of estimated SARS-COV-2 RNA loads 125 appeared comparable between children and adults; nevertheless, a trend towards a lower 126 median viral RNA load was observed in children compared to adults (6.98 log 10 127 copies/ml and 7.14 log 10 copies/ml and), although the difference did not reach statistical 128 significance (P=0.094). 129

We next compared initial SARS-CoV-2 RNA load in children and adults by time of NP 130 sampling after symptoms onset. Since SARS-CoV-2 RNA load peaks within the first 48 131 h after COVID-19 clinical presentation [18] , we split each patient group into two 132 subgroups (<3 days/≥3 days). As anticipated, SARS-CoV-2 RNA load was significantly 133 higher in NP specimens collected within 48 h after onset of symptoms than in those 134 obtained later on, in both children (median, 7.46 log 10 copies/ml vs. 5.17 log 10 135 copies/ml; P=<0.001) and adults (7.81 log 10 copies/ml vs. 6.45 log 10 copies/ml; 136 P=0.002) (Fig. 2) . Interestingly, SARS-CoV-2 RNA load measured within 48 h after 137 symptoms onset was comparable (P=0.263) between children and adults, whereas those 138 determined at later times (>48 h) were significantly lower in children (P=0.002). 139

Finally, we compared initial SARS-CoV-2 RNA loads across age groups conventionally 140 defined for children (infants, toddlers, preschoolers, school-aged, and adolescents) and 141 arbitrarily set for adults (18 to 65 years/>65 years). Pairwise comparison analyses are 142 shown in Fig. 1B . Overall, there were no between-group differences in either children or 143 adults (P=>0.14 for all pairwise comparisons). Moreover, as shown in Fig. 3A and 3B , 144 . CC-BY-NC 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) 

A wide range of SARS-CoV-2 RNA loads were detected in asymptomatic children and 148 adults (Fig. 4A) , likely reflecting the broad spectrum of NP collection times after 149 exposure to the presumed index case which, it should be noted, was not dissimilar 150 between children and adults. SARS-CoV-2 RNA loads in asymptomatic children 151 (median, 6.20 log 10 copies/ml) and adults (median, 6.48 log 10 copies/ml) were 152 comparable in magnitude (P=0.97). Likewise, no differences in SARS-CoV-2 RNA 153 loads were observed across pediatric ages or between adults aged ≤ 65 years or older 154 (Fig. 4B ) and no correlation was found between age and SARS-CoV-2 load (Rho, 155 0.066; P=0.44 for children and Rho, 0.020; P=0.62 for adults ( Fig. 3C and 3D) . 156

Children with COVID-19 symptoms displayed slightly higher SARS-CoV-2 RNA load 159 than their asymptomatic counterparts (Fig. 5A) , although statistical significance was not 160 reached (P=0.61). In adults, median estimated SARS-CoV-2 RNA load was 161 significantly higher in symptomatic than asymptomatic subjects (P=<0.001) (Fig. 5B) , 162 nevertheless, it was comparable (P=0.61) when patients sampled within 48 h after 163 symptoms onset were excluded from the analysis (Supplementary Fig. 1) . 164

We previously reported that SARS-CoV-2 could not be cultured from NP specimens 167 returning C T >25 (<5.9 log 10 copies/ml) by the TaqPath COVID-19 RT-PCR [16]. We 168 . CC-BY-NC 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The data are shown in Supplementary Fig. 2 . Overall, the percentage of NP specimens 170 returning SARS-CoV-2 N RT-PCR C T s below the abovementioned threshold was 171 similar for symptomatic children and adults (P=0.28) and was also comparable between 172 asymptomatic children and adults (P=0.87). Among children, that percentage appeared 173 higher for those aged under 3 years. For most age groups the percentage was higher in 174 symptomatic than in asymptomatic subjects, although these differences did not reach 175 statistical significance. 176

participants 178

To assess the quality of NP specimens collected from children and adults regarding 179 cellularity, we randomly selected 30 samples from each population group (n=60) that 180 were matched in SARS-CoV-2 RNA load (median 5.70 log 10 copies/ml; range 3.5-11.6 181 log10 copies/mL in specimens from children; median, 6.60 log 10 copies/ml; range, 2.2-182 10.9 log 10 copies/ml in specimens from adults; P=0.99). These specimens were assayed 183 with an in-house designed RT-PCR amplifying the housekeeping GUSB gene. The C T 184 of NP samples obtained from children and adults did not differ significantly (median 185 C T , 28.1; range, 24.8-32.7; and median C T , 29.0; range, 25.2-31.7, respectively, P=0.3), 186

suggesting that SARS-CoV-2 RNA loads measured in the two population groups were 187 not biased by differences in cellularity across NP specimens. RNA loads in asymptomatic children than in those with mild to moderate In this regard, it must be stressed that in our study asymptomatic individuals were tested 237 relatively soon after exposure, whereas in Kociolek's the authors admit a potential 238 population bias towards lower SARS-CoV-2 loads due to an excessive number of 239 remote infections detected via screening programs (i.e. hospital pre-admission). 240

Like the majority of commercially-available SARS-CoV-2 RT-PCRs, the RT-PCR 241 assays used in the current study do not co-amplify a housekeeping gene, thus precluding 242 . CC-BY-NC 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint this version posted March 3, 2021. ; https://doi.org/10.1101/2021.03.03.21252814 doi: medRxiv preprint assessment of sample cellularity. Given the widely varying quality of NP specimens 243

[26] which impacts significantly on estimated SARS-VoV-2 RNA loads [26], we 244 compared a randomly selected set of NP specimens from children and adults for their 245 cellular content using a housekeeping-gene RT-PCR set in parallel. We found 246 overlapping CTs in samples from both subject groups, making it unlikely that 247 differences in cellularity had a major impact on our results. However, only a small 248 number of NP specimens were screened for their cellular content. 249

The current study has several limitations. First, clinical outcome of asymptomatic 250 individuals, which may be determined by peak viral load, could not be ascertained in a 251 large number of participants. Second, only initial SARS-CoV-2 loads were taken into 252 consideration in the analyses, so that we could not have captured the true virus 253 replication rate on an individual basis. Third, no attempt was made to subcategorize 254 individuals according to their baseline medical condition. 255

In summary, we conclude that SARS-CoV-2 RNA loads in non-hospitalized or 256 asymptomatic COVID-19 children of all ages were comparable to those estimated in 257 adults. Our findings indicate that children may spread SARS-CoV-2 in the general 258 population at the same level as adults. 259 

This work received no public or private funds. 265 . CC-BY-NC 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint this version posted March 3, 2021. ; https://doi.org/10.1101/2021.03.03.21252814 doi: medRxiv preprint . CC-BY-NC 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. . CC-BY-NC 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review)

The copyright holder for this preprint this version posted March 3, 2021. ; https://doi.org/10.1101/2021.03.03.21252814 doi: medRxiv preprint . CC-BY-NC 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review)

The copyright holder for this preprint this version posted March 3, 2021. ; https://doi.org/10.1101/2021.03.03.21252814 doi: medRxiv preprint

Susceptibility 274 to SARS-CoV-2 Infection Among Children and Adolescents Compared With Adults: A 275 Systematic Review and Meta-analysis

Virological assessment of hospitalized patients with COVID-2019

Culture-based virus 280 isolation to evaluate potential infectivity of clinical specimens tested for COVID-19

Predicting infectious 283 severe acute respiratory syndrome coronavirus 2 from diagnostic samples

CoV-2 in nasopharynx of symptomatic neonates, children, and adolescents. Emerg 287

Characteristics and Viral RNA Detection in Children With Coronavirus Disease 2019 in 313 the Republic of Korea

CoV-2 viral RNA load dynamics in the nasopharynx of infected children

Estrategia de detección precoz

mentos/COVID19_Estrategia_vigilancia_y_control_e_indicadores.pdf

Field evaluation of a rapid antigen test (Panbio COVID-19 Ag Rapid Test Device) for

COVID-19 diagnosis in primary healthcare centres

Evaluation of a rapid antigen 326 test (Panbio COVID-19 Ag rapid test device) for SARS-CoV-2 detection in 327 asymptomatic close contacts of COVID-19 patients

Amplification of 330 human beta-glucuronidase gene for appraising the accuracy of negative SARS

RT-PCR results in upper respiratory tract specimens

viral load dynamics, duration of viral shedding, and 335 infectiousness: a systematic review and meta-analysis

SARS-CoV-2 Viral 337 Load in Upper Respiratory Specimens of Infected Patients

Asymptomatic transmission during the COVID-19 pandemic 340 and implications for public health strategies

SARS-CoV-2 Transmission From People Without COVID-19 Symptoms

Viral Shedding Among Asymptomatic and Symptomatic Patients With SARS-CoV-2

Infection in a Community Treatment Center in the Republic of Korea

Upper respiratory viral load in 349 asymptomatic individuals and mildly symptomatic patients with SARS-CoV-2 350 infection

SARS-CoV-2 Infections Among Children in the Biospecimens from Respiratory Virus-353

Exposed Kids (BRAVE Kids) Study

The authors declare no conflicts of interest. 267 . CC-BY-NC 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) CC-BY-NC 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review)

The copyright holder for this preprint this version posted March 3, 2021. ; https://doi.org/10.1101/2021.03.03.21252814 doi: medRxiv preprint