key: cord-0984917-uz91cd6h authors: Leung, Kenneth Siu-Sing; Ng, Timothy Ting-Leung; Wu, Alan Ka-Lun; Yau, Miranda Chong-Yee; Lao, Hiu-Yin; Choi, Ming-Pan; Tam, Kingsley King-Gee; Lee, Lam-Kwong; Wong, Barry Kin-Chung; Ho, Alex Yat-Man; Yip, Kam-Tong; Lung, Kwok-Cheung; Liu, Raymond Wai-To; Tso, Eugene Yuk-Keung; Leung, Wai-Shing; Chan, Man-Chun; Ng, Yuk-Yung; Sin, Kit-Man; Fung, Kitty Sau-Chun; Chau, Sandy Ka-Yee; To, Wing-Kin; Que, Tak-Lun; Shum, David Ho-Keung; Yip, Shea Ping; Yam, Wing-Cheong; Siu, Gilman Kit Hang title: A territory-wide study of early COVID-19 outbreak in Hong Kong community: A clinical, epidemiological and phylogenomic investigation date: 2020-03-31 journal: nan DOI: 10.1101/2020.03.30.20045740 sha: 2f8fb557a6ac5b05a6dd3c0f8b7295a136d4e63a doc_id: 984917 cord_uid: uz91cd6h Background: Initial cases of coronavirus disease 2019 (COVID-19) reported in Hong Kong were mostly imported cases from Mainland China. However, most cases reported in February 2020 were local infections with unknown source, indicating local community transmissions. This study aimed to report the clinical, epidemiological and phylogenomic characteristics of the local cases of COVID-19 in our community. Methods: We extracted the demographic, clinical and epidemiological data from 50 COVID-19 patients, who accounted for 53.8% of the cases in Hong Kong by the end of February 2020. We used both Nanopore and Illumina platforms to perform whole-genome sequencing (WGS) of the severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) from these patients. Phylogenetic relatedness among these local cases and their placement in the global phylogeny were examined. The evolutionary rate and divergence time of transmission were also determined. Findings: Of these 50 patients, only three (6.0%) had visited Wuhan while 43 (86.0%) did not have recent travel records. The average interval from symptom onset to hospital admission was 8.5 days. The most common signs and symptoms on admission were cough (74.0%) and fever (58.0%). Radiographic abnormality was found in 46 (92.0%) patients. Three (6.0%) patients required ICU admission. Phylogenetic analysis concurred with epidemiological investigation that 42 (84.0%) cases could be grouped into six transmission clusters. Forty-four (88.0%) cases harboured a common mutation Orf3a G251V. Global phylogeny of SARS-CoV-2 revealed that most (88.0%) cases in Hong Kong were clustered in two subclades with the strains from other countries. The estimated time to the most recent common ancestor (tMRCA) of COVID-2019 outbreak in Hong Kong was December 24, 2019 with an evolutionary rate of 3.04x10-3 substitutions per site per year. The reproduction number value was 1.84 as of February 28, 2020 in Hong Kong. Interpretation: We provided a territory-wide overview of COVID-19 in Hong Kong, which has borders connecting to Mainland China. Transmission in closed settings especially during family and religious gatherings is a hallmark of the recently reported cases. The reproduction number value indicated an ongoing outbreak in the community. Social distancing and vigilant epidemiological control are crucial to the containment of COVID-19 transmission Fifty COVID-19 patients were included with 54·0% being female and the mean age was 55·2 (table 1) . Three were considered as imported cases as the patients stayed in Wuhan patients presented with cough. Fever was presented in 58·0% of the patients on admission, 1 8 5 but gradually developed in 64·0% patients during hospitalisation. Other less common 1 8 6 symptoms include muscle-ache(25·0%), sore throat(24·0%), shortness of breath(24·0%), and 1 8 7 diarrhoea(14·3%)(table 2). Two(4·0%) patients were asymptomatic throughout the study 1 8 8 period. Regarding radiological examination, 27(54·0%) showed bilateral pneumonia, 1 8 9 11(22·0%) unilateral pneumonia, and 17(34·7%) multiple mottling and ground-glass opacity. None of the patients was co-infected with other respiratory viruses or fungi. Two patients 1 9 1 were culture-positive for Klebsiella aerogenes and Ralstonia pickettii in their sputum 1 9 2 specimens. Both presented with ARDS and acute respiratory injury accompanied with septic 1 9 3 shock or acute renal injury, and required ICU admission. Of the 50 cases, 42(84·0%) could be clustered based on their epidemiological linkages(figure 1 9 5 1). Six transmission clusters(Clusters 1-6) were identified. Cluster 1 involved a family of 1 9 6 four members. The father, who travelled to Guangdong, China in late January 2020, was 1 9 7 believed to infect his wife and subsequently their daughter and son-in-law in a family 1 9 8 All rights reserved. No reuse allowed without permission. author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprint (which was not peer-reviewed) is the . 1 0 gathering. Clusters 2 and 3 were family clusters of local infection with unknown source. Both 1 9 9 clusters involved three household members without recent travel history. Cluster 4 was a 2 0 0 super-spreading event(SSE) associated with a barbecue and hotpot party of 19 family 2 0 1 members in late January. Ten had developed symptoms two days after the party. A colleague 2 0 2 of one infected member, who was not present at the party, was also diagnosed to have 2 0 3 COVID-19. Cluster 5 initiated from a resident of a public estate, who was diagnosed on 2 0 4 January 30. Eleven days later, three members of a household living in the same building as, 2 0 5 but 10-storey below from, the index case were also infected. Two household members 2 0 6 attended a family gathering of 29 people at a Chinese seafood restaurant during their 2 0 7 incubation period. Three were diagnosed consecutively around two weeks after the gathering. Additionally, a Filipino domestic helper of one infected member, who was absent in the 2 0 9 family gathering, was also infected. The earliest reported case of Cluster 6 was a woman who were also tested positive for SARS-CoV-2. Consensus genomes of all 50 cases were constructed based on Nanopore sequencing and genome with 550× and 132× coverage for Nanopore and Illumina platforms respectively. The 2 1 8 consensus genome size was ~29.9 kbp with GC content ~38%. The genomes were highly conserved with the first SARS-CoV-2 genome with an average sequence identity of (table 3) . G251V in Orf3a is the most frequent amino acid substitution 2 2 2 All rights reserved. No reuse allowed without permission. author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprint (which was not peer-reviewed) is the . with 44/50(88·0%) of the samples harbouring this mutation, followed by Orf1ab 1 2 December 11, 2019 to January 5, 2020) with an evolutionary rate 3.04×10 -3 substitutions per 2 4 8 site per year(95% BCI: 2·04×10 -3 to 4·09×10 -3 substitutions per site per year). Based on our 2 4 9 demographic data, time interval from symptom onset to hospital admission was ~8·5 days. The estimated reproduction number was 1.84(95% BCI: 1.37 to 2.35). This study provides a territory-wide overview of early COVID-19 outbreak in Hong Kong, an clinical, epidemiological, phylogenomic, and phylodynamic data. Fever was reported to be the most common symptom since 81·8%-98·0% of the patients had by Guan et al found that fever was only presented in 43·8% of the patients on admission(2). In this study, fever was identified in only 58·0% of the patients on admission, but developed 2 6 0 in 64.0% after hospitalisation. Afebrile patients might be missed if the surveillance case 2 6 1 definition relied solely on fever. Therefore, laboratory surveillance has been extended to in-2 6 2 patients and out-patients with respiratory symptoms. ICU admission was relatively shock. Of note, both patients had bacterial co-infection, which was absent in other cases. In 2 6 6 addition to virulence factors of the pathogens, the host immune status, old age, and presence 2 6 7 of chronic illness might be associated with enhanced disease severity. Immune supportive 2 6 8 treatment and prompt antibiotics administration might reduce complications and mortality. In Hong Kong, initial cases recorded in January 2020 were mostly imported cases. Since building, but 10-storey above, through a faulty sewage pipe setup. Based on phylogenetic 2 9 8 analysis, viral genomes in Cluster 5 shared a similar genetic distance from the reference 2 9 9 genome and were assigned to the same branch of the tree. This supports a potential 3 0 0 transmission linkage among these cases. high degree of genome similarity suggests that these cases might be originated from a single February after being found associated with a series of confirmed cases connected with the 3 1 8 worship hall. He was completely asymptomatic throughout the study period. Phylogenetic analysis showed that this case was closest to the root of the cluster(top portion with strains ongoing. Here we demonstrated the possibility of "hidden spreader" as a source of COVID-3 2 3 All rights reserved. No reuse allowed without permission. author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprint (which was not peer-reviewed) is the . 19 community outbreak. It also highlights the importance of rapid quarantine of the close 3 2 4 contacts of confirmed cases regardless of the presence of signs and symptoms in order to halt 3 2 5 the COVID-19 community spread. In the evolutionary clock study, death rate δ (which refers to the duration for the case to 1 6 phylodynamic analysis which could be different from those based on epidemiological models. Finally, gap regions were observed in some consensus genomes. This is mainly because accurate information for our subsequent analyses. In conclusion, phylogenomic data were consistent with the epidemiological findings that All rights reserved. No reuse allowed without permission. author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprint (which was not peer-reviewed) is the . All rights reserved. No reuse allowed without permission. author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprint (which was not peer-reviewed) is the . author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprint (which was not peer-reviewed) is the . https://doi. org/10.1101/2020.03.30.20045740 doi: medRxiv preprint 2 2 Cluster 6 (n=13) G295V All rights reserved. No reuse allowed without permission. author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprint (which was not peer-reviewed) is the All rights reserved. No reuse allowed without permission. author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprint (which was not peer-reviewed) is the * The location of amino acid substitutions was based on SARS-CoV-2 reference genome (NC_045512.2). † The singleton cases did not have recent travel history. ‡ The patients had record of travel to Wuhan City. All rights reserved. No reuse allowed without permission. author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprint (which was not peer-reviewed) is the A total of 273 worldwide genomes as of February 28, 2020 were randomly selected and downloaded from GISAID Global Cases COVID-19 database. The phylogenetic tree was built with the samples collected in this study based on fast likelihood-based aLRT SH-like method and rooted on the earliest published genome of SARS-CoV-2 (accession no.: NC_045512.2). The Hong Kong strains showed limited genetic variability and tended to aggregate in two subclades, which were highlighted in yellow box and pink box respectively. The tree was constructed using all 50 COVID-19 cases included in this study and rooted on the earliest published genome of SARS-CoV-2 (accession no.: NC_045512.2). Bootstrap value was set at 1000×. Samples were colour-coded by epidemiological link as follows: 1) Magenta represents Cluster 1: Imported cluster; 2) Cyan represents Cluster 2: unknown source; 3) Green represents Cluster 3: local; 4) Orange represents Cluster 4: Hotpot partyrelated Superspreading event (SSE); 5) Blue represents Cluster 5: Public housing estaterelated SSE; and 6) Red represents Cluster 6: Buddha worship hall-related SSE. Case 84 and Case 102 were asymptomatic at the time of sample collection, and were marked with asterisks (*) in the diagram. Each case is indicated with a case number (see figure 1 ) followed by the date (dd/mm/yyyy) of symptom onset. All rights reserved. No reuse allowed without permission. author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprint (which was not peer-reviewed) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprint (which was not peer-reviewed) is the . https://doi.org/10. 1101 /2020 Singleton cases CLUSTER 1: Imported cluster CLUSTER 2: local cluster CLUSTER 3: local cluster CLUSTER 4: Hotpot party-related SSE CLUSTER 5: Public housing estate-related SSE CLUSTER 6: Buddha Worship hall-related SSE Apr;10(4):e1003537. author/funder, who has granted medRxiv a license to display the preprint in perpetuity.The copyright holder for this preprint (which was not peer-reviewed) is the . https://doi.org/10.1101/2020.03.30.20045740 doi: medRxiv preprint All rights reserved. No reuse allowed without permission.author/funder, who has granted medRxiv a license to display the preprint in perpetuity.The copyright holder for this preprint (which was not peer-reviewed) is the . https://doi.org/10. 1101 /2020