key: cord-0759065-h35qdqa9 authors: Leuzinger, K.; Roloff, T.; Gosert, R.; Soegaard, K.; Naegele, K.; Rentsch, K.; Bingisser, R.; Nickel, C.; Pargger, H.; Bassetti, S.; Bielicki, J. A.; Khanna, N.; Tschudin Sutter, S.; Widmer, A.; Hinic, V.; Battegay, M.; Egli, A.; Hirsch, H. H. title: Epidemiology of SARS-CoV-2 Emergence Amidst Community-Acquired Respiratory Viruses date: 2020-07-08 journal: nan DOI: 10.1101/2020.07.07.20148163 sha: 43f82b306e4fa0efc342e43dd4b04028a610c70a doc_id: 759065 cord_uid: h35qdqa9 Background. SARS-CoV-2 emerged in China in December 2019 as new cause of severe viral pneumonia (CoVID-19) reaching Europe by late January 2020. We validated the WHO-recommended assay and describe the epidemiology of SARS-CoV-2 and community-acquired respiratory viruses (CARVs). Methods. Naso-oropharyngeal swabs (NOPS) from 7663 individuals were prospectively tested by the Basel-S-gene and the WHO-based E-gene-assay (Roche) using Basel-N-gene-assay for confirmation. CARVs were tested in 2394 NOPS by multiplex-NAT, including 1816 together with SARS-CoV-2. Results. Basel-S-gene and Roche-E-gene-assays were concordant in 7475 cases (97.5%) including 825 (11%) positive samples. In 188 (2.5%) discordant cases, SARS-CoV-2 loads were significantly lower than in concordant positive ones and confirmed in 105 NOPS. Adults were more likely to test positive for SARS-CoV-2, while children were more likely to test CARV-positive. CARV co-infections with SARS-CoV-2 occurred in 1.8%. SARS-CoV-2 replaced other CARVs within 3 weeks reaching 48% of all detected respiratory viruses followed by rhino/enterovirus (13%), influenzavirus (12%), coronavirus (9%), respiratory syncytial (6%) and metapneumovirus (6%). Conclusions. The differential diagnosis for respiratory infections was broad during the early pandemic, affecting infection control and treatment decisions. We discuss the role of pre-existing immunity and competitive CARV replication for the epidemiology of SARS-CoV-2 infection among adults and children. Patients presenting with influenza-like illness to the outpatient department or emergency 98 department of the University Hospital Basel or the University of Basel Children's Hospital 99 were enrolled in this analysis of prospectively collected results on respiratory virus panel 100 and/or SARS-CoV-2 testing between 1. January and 29. March 2020. 101 102 Clinical samples and total nucleic acid extraction All statistical data analysis was done in R (https://www.r-project.org/), and Prism (version 8; 154 Graphpad Software, CA, USA) was used for data visualization. Statistical comparison of 155 non-parametric data was done using Mann-Whitney U test, and Bonferroni correction was 156 applied for multiple comparisons. 157 158 Ethics statement 159 The study was conducted according to good laboratory practice and in accordance with the 160 Declaration of Helsinki and national and institutional standards and was approved by the 161 ethical committee (EKNZ 2020-00769) . 162 All rights reserved. No reuse allowed without permission. (which was not certified by peer review) 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 July 8, 2020 . . https://doi.org/10.1101 /2020 Results 163 To independently evaluate the WHO-recommended assay, we designed two different single step RT-QNAT assays targeting the S-gene and the N-gene. We observed close clustering 165 of the complete SARS-CoV-2 genomes and specifically its S-and N-gene target sequences, 166 clearly separating from the corresponding HCoV genome sequences (Supplementary 167 Figure 1 ). We found no insertions or deletions in either target, and only a single SNP in the 168 probe-binding site of the S-gene RT-QNAT in one of 3323 (0.03%) sequences, at a central 169 position not predicted to affect the assay performance (Supplementary Table 2) . 170 171 To cross-validate the WHO-Roche-E-gene and the Basel-S-gene without reporting delay, we 172 analyzed all submitted NOPS directly in parallel. From 9 th until 29 th of March 2020 (calendar 173 week 11 to 13), 7663 samples were submitted from 354 (5%) pediatric and 7309 (95%) adult 174 patients ( Table 1) . Most patients had presented to primary care and outpatient clinics (74%), 175 while 26% of cases originated from secondary and tertiary care units including 3% from 176 intensive care ( Table 1) Figure 1A) . Indeed, 666 (72%) NOPS extracts had 186 SARS-CoV-2 loads of more than 1 million copies (c)/mL UTM in the S-gene RT-QNAT 187 (median 7.2 log10 c/mL, IQR 5.8 -8.4; Figure 1B) . Conversely, the Ct-values were 188 significantly higher for discordant results indicating low viral loads ( Figure 1A) . Thus, the 189 Basel-SCoV2-S-111bp had a sensitivity of 99.7% (95% CI: 95% -100%) and specificity of 190 All rights reserved. No reuse allowed without permission. (which was not certified by peer review) 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 July 8, 2020 . . https://doi.org/10.1101 /2020 99.0% (95% CI: 91% -100%). Taken together, 930 (12.1%) SARS-CoV-2 infections were 191 confirmed and further analyzed. Among SARS-CoV-2-positive patients, male gender was 192 more prevalent (49% vs. 44%; p=0.002) and the median age was higher (49 vs. 43 years; 193 p<0.001) compared to those with a negative test result (Table 1) . However, higher patient 194 age was not associated with higher SARS-CoV-2 loads (Spearman's r=0.034, p=0.30). 195 Moreover, SARS-CoV-2 was detected in 14 (4%) of 354 children compared to 916 (12%) of 196 the non-pediatric patients (p<0.05). 197 To investigate the epidemiology of CARVs and SARS-CoV-2 during the first phase of the 199 pandemic, we identified all NOPS (n=2394) from patients with influenza-like illness, which 200 had been tested by CARV-multiplex-NAT between January 1 st (calendar week 1) and March 201 29 th 2020 (calendar week 13). In 942 (39%) cases, at least one pathogen had been detected 202 including 95 with two (3.9%), and 9 with 3 (0.1%) pathogens. The weekly prevalence rates 203 for SARS-CoV-2 and CARVs revealed a fluctuating CARV activity until calendar week 7 204 followed by a steep increase in CARVs, which declined after week 10, when SARS-CoV-2 205 detection rates rose sharply (Figure 2A) . This was also reflected in the cumulative rates 206 ( Figure 2B ; histogram) reaching 48% for SARS-CoV-2 by calendar week 13, followed by 207 rhinovirus (13%) and influenzavirus (12%) ( Figure 2B ; pie chart). Restricting the analysis to 208 1816 NOPS, from which both, CARV-multiplex-NAT and SARS-CoV-2 RT-QNAT had been 209 requested, the cumulative SARS-CoV-2 detection was 17% after rhinovirus (22%) and 210 influenzavirus (20%) (Figure 2C ; pie chart). The weekly detection rates revealed that 211 SARS-CoV-2 largely replaced all other CARVs except rhinovirus ( Figure 2C) . 212 213 Unlike for SARS-CoV-2, the CARV detection rate was significantly higher in children than in 214 adults (P<0.001; Table 2 ). This significant effect also prevailed when SARS-CoV-2 and 215 CARV positive cases where analyzed together and when excluding rhinovirus-infected 216 cases from the analysis ( Table 2 ; see also below). Analyzing the age distribution of CARV-217 positive cases (Figure 3) , we found higher detection rates of adenovirus, parainfluenzavirus, 218 All rights reserved. No reuse allowed without permission. (which was not certified by peer review) 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 July 8, 2020. Table 3) . Moreover, adenovirus-positive 221 patients were significantly younger than patients testing positive for other CARVs or SARS-222 CoV-2 (P<0.001; Table 3 ). Among adults (1554/1816; 85.5%), no significant age differences 223 were observed for patients testing positive for any CARV, but patients being positive for 224 SARS-CoV-2 tended to be older than patients testing positive for any other CARV (P<0.01; 225 SARS-CoV-2 was detected together with at least one other CARV, which consisted of a 232 single pathogen in 15 cases, namely rhinovirus (n=5), human coronaviruses (n=5), 233 parainfluenzavirus (n=3) and influenzavirus (n=2), and more than one CARV detection in 2 234 cases ( Table 4) . Overall, CARV detection was associated with a high negative predictive 235 value of 98.1% for SARS-CoV-2 infection. The negative predictive value for SARS-CoV-2 236 infection was higher in CARV-positive children (99.0%; ≤16 years) than adults (97.1%; >16 237 years). Conversely, a negative multiplex-NAT result after the first detected SARS-CoV-2 238 case was associated with a rapidly increasing likelihood to be positively tested for SARS-239 CoV-2 from 1% in calendar week 9 to 48% in calendar week 13 (Figure 2) . 240 All rights reserved. No reuse allowed without permission. (which was not certified by peer review) 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 July 8, 2020. . https://doi.org/10. 1101 /2020 CARVs are at their yearly seasonal peak including influenzavirus, respiratory syncytial virus, 243 and human coronaviruses. Our analysis from the start to the peak of the pandemic wave of 244 SARS-CoV-2 in Northwestern Switzerland has three major findings. 245 First, the early pandemic phase until calendar week 10 was dominated by winter CARVs, 246 emphasizing the importance of their rapid and accurate identification due to several reasons: 247 i) significant morbidity and mortality in vulnerable patients (very young, elderly, 248 immunocompromised) [5] ; ii) specific antiviral therapy in case of influenzavirus-A/B detection; 249 iii) appropriate infection control and cohorting strategies upon hospital admission; and iv) 250 prevention of unnecessary empiric antibiotic therapy in CARV-positive patients, or treatment 251 adaptation in case of atypical bacterial agents like M. pneumoniae [10] [11] [12] [13] . In this early phase, 252 CARV detection was associated with a high negative predictive value of 98.1% for SARS-253 CoV-2 infection. 254 Second, SARS-CoV-2 almost completely replaced the seasonally circulating CARVs within 255 only 3 weeks' time. During calendar week 12 and 13, SARS-CoV-2 was practically the only 256 respiratory virus leading to a cumulative 48% runner-up position when counting all detected 257 CARVs from January 1 st , 2020. This dynamic evolution was also seen when explicitly 258 analyzing NOPS from patients with respiratory illness for whom both CARV-and SARS-CoV-259 2 testing had been requested. The weekly prevalence revealed a significant increase in SARS-260 CoV-2 detection rates while the initially increasing CARVs were curtailed. These data suggest 261 the intriguing possibility of competing risks for host infection. 262 Third, diagnosis of SARS-CoV-2 infection was highly reliable being based on three 263 independent molecular tests. Thereby, an independent validation of the WHO-endorsed E-264 gene was provided. Whereas respiratory panel testing is well validated and widely used in 265 tertiary care centers [5, 14, 15] , the response to the SARS-CoV-2 pandemic hinges on the 266 performance of a new diagnostic test for a new viral agent, and its communication within a 267 short turn-around time. To accomplish this task, we prospectively tested all NOPS directly in 268 All rights reserved. No reuse allowed without permission. (which was not certified by peer review) 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 July 8, 2020. . https://doi.org/10.1101/2020.07.07.20148163 doi: medRxiv preprint parallel with the commercial Roche-E-gene and our Basel-S-gene RT-QNAT. This outstanding 269 opportunity for independent test validation on more than 7600 patients demonstrated high 270 concordance of 97.5% between both assays including 825 (11%) SARS-CoV-2 infections, 271 which could be communicated without further delay to the treating physicians. Importantly, 272 comparison of the Ct-values revealed that the discordance mostly resulted from SARS-CoV-273 2 loads at the limit of detection. Thus, discordant results became increasingly likely at very 274 low, hence limiting viral loads in the NOPS, most likely reflecting a stochastic distribution of 275 genomes in the analyte. 276 277 A limitation of our study is the dependence on the pre-analytic steps of NOPS sampling, 278 especially in the light of the natural course of SARS-CoV-2 infection. We addressed this 279 challenge through repeated instructions and video clips demonstrating the correct use of 280 personal protection equipment, validated swab sets, and defined sampling procedures in 281 dedicated hospital areas. However, late presentation at stages of more advanced disease 282 manifesting in the lower respiratory tract requires testing of respiratory samples from the lower 283 respiratory tract, while this study was restricted to the first diagnostic testing in NOPS. Early 284 testing is clinically and epidemiologically advisable in view of high viral loads detectable 285 already early in the course in exposed pre-symptomatic and in oligo-symptomatic persons [16, 286 17] . 287 288 Since our diagnostic laboratory is serving both regional tertiary care centers for adults and 289 children, we examined the age distribution of SARS-CoV-2 and CARV infection. Indeed, 290 patients testing positive for adenovirus and respiratory syncytial virus were significantly 291 younger and more likely to be children below the age of 5 years, among whom SARS-CoV-2 292 infection remained rare, in line with other studies [18, 19] . Although SARS-CoV-2 positive 293 adults were older than patients testing positive for CARVs, the median age of >40 years in 294 CARV-positive patients suggests that similar adult populations were at risk for established 295 CARVs or for the novel SARS-CoV-2. 296 All rights reserved. No reuse allowed without permission. (which was not certified by peer review) 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 July 8, 2020. . https://doi.org/10. 1101 /2020 Our study also provides intriguing observations regarding the epidemiology of SARS-CoV-2 298 in its capacity to replace circulating CARVs among adults. Notably, co-infection rates of 299 CARVs with SARS-CoV-2 were rather low as reported here and by others [8], suggesting a 300 competitive infection situation. It is presently unclear whether virus properties such as higher 301 infectiousness, facilitated transmission, or increased host susceptibility are the decisive 302 factors conferring significant advantages to the novel SARS-CoV-2 in this first wave of the 303 pandemic. Regarding the infectiousness of SARS-CoV-2, our data provide independent 304 evidence for very high viral loads in the order of 1 -100 million copies per milliliter transport 305 medium. Even if these high numbers only carry 1000-fold lower infectious units, the 306 infectious activity remains high in the patients' respiratory secretions. Notably, similarly high 307 viral loads have also been described for CARVs including influenza or respiratory syncytial 308 virus [5, [20] [21] [22] . Regarding transmission, SARS-CoV-2 is thought to behave less like 309 influenzaviruses spreading significantly by aerosols [23, 24] , but rather like respiratory 310 syncytial virus spreading by droplets, contaminated surfaces and hands [25, 26] . However, 311 aerosolization of SARS-CoV-2 may also play a role, especially when associated with high-312 velocity air streams during sneezing, singing and medical procedures [27] [28] [29] . 313 Finally, increased susceptibility of the human host to infection by this novel, presumably 314 zoonotic coronavirus remains, but is a difficult to estimate factor at this time. Already the first 315 reports from China in January 2020 indicated that SARS-CoV-2 is well adapted to the 316 human host [30] . Unlike SARS-CoV-1, SARS-CoV-2 is easily transmitted from human to 317 human already before the start of symptoms, hence facilitating the pandemic spread [17, 31, 318 32 ]. However, SARS-CoV-2 seems to be susceptible to type-1 interferons [33] , and induces 319 high amounts of interleukin-6 through excessive macrophage activation upon progression to 320 viral pneumonia, which has become a clinically relevant target of CoVID-19 treatment [34] . 321 322 What could be the underlying mechanisms for an increased susceptibility to SARS-CoV-2 323 infection competitively replacing established CARVs in a mostly adult population? Hand 324 All rights reserved. No reuse allowed without permission. (which was not certified by peer review) 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 July 8, 2020. . https://doi.org/10.1101/2020.07.07.20148163 doi: medRxiv preprint pandemic wave. We hypothesize that the decisive factors may be the differential net 327 response of the host to virus-induced unspecific innate immunity on the one hand and to 328 virus-specific adaptive immune memory on the other hand. CARV infections are known to 329 cause an innate immune response including type-1 interferons, which reduces the risk of co-330 infection by other viruses including SARS-CoV-2 [35, 36] . Since adults have been repeatedly 331 exposed to CARVs in the past, their CARV-specific immune memory may not be high 332 enough to prevent symptomatic CARV re-infection, but is readily boosted upon re-exposure, 333 hence limiting CARV replication and the associated inflammation elicited by innate immunity. 334 We propose that thereby the semi-immune mostly adult host population becomes available 335 for SARS-CoV-2 infection. Since SARS-CoV-2 is novel having little, if any, specific immune 336 memory, its replication is prolonged, evoking pronounced inflammation, delaying infection by 337 other circulating CARVs, extending transmission periods and shifting the epidemiologic 338 curve in favor of this novel agent. This differential net response of virus-induced unspecific 339 innate immunity and virus-specific adaptive immune memory may also contribute to the 340 puzzling lower infection rates seen in small children, who typically replicate CARVs in high 341 frequency and high levels, hence interfering with SARS-CoV-2. However, CARVs may differ 342 in their propensity to interfere and may be low for rhinovirus. Indeed, 46 (60%) of 77 co-343 infections involved rhinovirus. Although other (co-)factors cannot be excluded, our 344 hypothesis will be testable by analyzing, whether or not vaccines to CARVs and/or to SARS-345 CoV-2 change this competitive epidemiologic risk. Possibly, CARV interference will be 346 reduced during the summer months putting younger age populations at risk for the pandemic 347 SARS-CoV-2. 348 349 In conclusion, circulating CARVs were dominant during the first phase of the CoVID-19 350 pandemic, but rapidly replaced within 2 weeks by SARS-CoV-2. A comprehensive testing 351 strategy covering SARS-CoV-2 and CARVs is central to infection control and clinical 352 All rights reserved. No reuse allowed without permission. (which was not certified by peer review) 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 July 8, 2020. (which was not certified by peer review) 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 July 8, 2020. Switzerland for expert help and assistance. 361 All rights reserved. No reuse allowed without permission. (which was not certified by peer review) 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 July 8, 2020. C. Weekly SARS-CoV-2 and CARV prevalence in NOPS tested in parallel (n=1816), 385 cumulated SARS-CoV-2 and CARV cases by calendar week 13 (pie chart). 386 All rights reserved. No reuse allowed without permission. (which was not certified by peer review) 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 July 8, 2020. . https://doi.org/10.1101/2020.07.07.20148163 doi: medRxiv preprint with the RT-QNAT assays. Patient age of CARV or SARS-CoV-2 positive patients is displayed 389 (median, 25 th and 75 th percentiles; n=1816), and compared using Mann-Whitney U test (Table 390 3; Supplementary Table 4) . 391 All rights reserved. No reuse allowed without permission. (which was not certified by peer review) 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 July 8, 2020. All rights reserved. No reuse allowed without permission. (which was not certified by peer review) 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 July 8, 2020. All rights reserved. No reuse allowed without permission. (which was not certified by peer review) 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 July 8, 2020. . SARS-CoV-2, severe acute respiratory syndrome coronavirus All rights reserved. No reuse allowed without permission. (which was not certified by peer review) 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 July 8, 2020. . Table 4 . Patients with more than one positive SARS-CoV-2 or CARV detection (n=1816). (which was not certified by peer review) 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 July 8, 2020 . . https://doi.org/10.1101 /2020 3 RT-QNAT target region ±150bp, ±100bp and ±5bp 4 SNP outside the RT-QNAT target region 5 SNP in the probe-binding site 6 SNP in the RT-QNAT target region, but not in a primer/probe-binding site All rights reserved. No reuse allowed without permission. (which was not certified by peer review) 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 July 8, 2020. . https://doi.org/10.1101/2020.07.07.20148163 doi: medRxiv preprint (which was not certified by peer review) 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 July 8, 2020. . https://doi.org/10.1101/2020.07.07.20148163 doi: medRxiv preprint Figure S1 All rights reserved. No reuse allowed without permission. (which was not certified by peer review) 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 July 8, 2020. . https://doi.org/10.1101/2020.07.07.20148163 doi: medRxiv preprint Figure S1 . Phylogenetic analysis using complete SARS-CoV-2 genome sequences, the Divergences were estimated by the Jukes-Cantor method and neighbor-joining trees were constructed with the CLC Genomic Workbench software. A. Phylogenetic analysis of complete SARS-CoV-2 genome sequences of 29'900 nucleotides (nt) in length available in the NCBI-GenBank and GISAID database (accessed on 20 th of April; n=3323). All rights reserved. No reuse allowed without permission. (which was not certified by peer review) 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 July 8, 2020. . https://doi.org/10.1101/2020.07.07.20148163 doi: medRxiv preprint Figure S1 . Phylogenetic analysis using complete SARS-CoV-2 genome sequences, the Divergences were estimated by the Jukes-Cantor method and neighbor-joining trees were constructed with the CLC Genomic Workbench software. A. Phylogenetic analysis of complete SARS-CoV-2 genome sequences of 29'900 nucleotides (nt) in length available in the NCBI-GenBank and GISAID database (accessed on 20 th of April; n=3323). All rights reserved. No reuse allowed without permission. (which was not certified by peer review) 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 July 8, 2020. . https://doi.org/10.1101/2020.07.07.20148163 doi: medRxiv preprint All rights reserved. No reuse allowed without permission. (which was not certified by peer review) 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 July 8, 2020. . https://doi.org/10.1101/2020.07.07.20148163 doi: medRxiv preprint A Novel Coronavirus from Patients with Pneumonia in China, 24. Tellier R. Review of aerosol transmission of influenza A virus 13%) Median: 6 25 th : 3 75 th : 9 IQR: 6 135 (9%) SARS-CoV-2 5 (2%) Median: 2 25 th : 1 75 th : 3 IQR: 2 143 (9%) HRSV, human respiratory syncytial virus SARS-CoV-2, severe acute respiratory syndrome coronavirus