Looking back on the COVID-19 pandemic in an elite sports team using whole genome sequencing

Objectives The aim of this study was to investigate the effectiveness of infection control measures to prevent transmission of SARS-CoV-2 within a professional sports team using whole genome sequencing. Design Prospective cohort study. Methods 74 players and staff members of a Dutch professional male football team were followed from August 2020 until May 2021. A set of health and safety measures were introduced and all participants underwent regular SARS-CoV-2 RNA testing. All positive samples were subsequently sequenced (Nanopore sequencing) to assess whether infections were acquired within the training center or in the community. Results Throughout the study period, 13 participants tested positive for SARS-CoV-2. The phylogenetic analysis revealed 2 clusters (of 2 and 3 cases respectively), indicating that 3/13 cases (23%) acquired infection from another player or staff member. The first cluster was diagnosed upon enrolment, thus transmission had occurred prior to the implementation of health and safety protocols. Finally, 4 cases were diagnosed prior to symptom onset, emphasizing that frequent testing leads to early detection and isolation. Conclusions These data show that a combination of regular testing and basic control measures can prevent outbreaks of COVID-19 in a professional sports team. Whole genome sequencing is an important tool to distinguish between infections introduced from the community and infections transmitted between athletes.


Introduction
Elite athletes have suffered an immense impact from the COVID-19 pandemic. Public health restrictions have led to the suspension of competitions and international events, as well as interruptions of regular training activities. In a survey among more than 12,000 athletes in 142 countries, less than 40% was able to maintain training activities during lockdown at regular levels. 1 In elite athletes insufficient training stimuli for periods longer than 4 weeks have been shown to cause respiratory, circulatory and metabolic changes, all of which affect performance. 2,3 Furthermore, among athletes who acquired COVID-19, approximately 14% had a symptom duration of >28 days. 4 European football leagues were interrupted for a short period of time and resumed in summer 2020. Since then, professional sports clubs have been faced with the challenge of developing protocols to create safe conditions during training and competitions. Feyenoord is a top-ranked football club based in Rotterdam, the Netherlands. Within the club, including athletes and staff, national regulations were followed with respect to hygiene measures and isolation of confirmed cases and high-risk contacts. As social distancing was not possible at all times, an additional set of measures was implemented to minimize the risk of transmission among its players and staff members.
The aim of this prospective cohort study was to evaluate the effectiveness of additional infection control measures in a group without the possibility of continuous social distancing. These measures were introduced during the 2020-2021 season, prior to the rollout of COVID-19 vaccination programs. To assess the effectiveness, it was essential to distinguish between case introductions from the community and transmission events within the training center. For that reason, SARS-CoV-2 genome sequence analysis was performed on all positive samples. At the time of writing the COVID-19 pandemic is transitioning into an endemic phase. It is important to take note of the lessons learned and experiences gained during this time, as they will be crucial in responding adequately to future outbreaks.

Methods
This prospective study followed a cohort of 74 unvaccinated players and staff members from August 2020 until May 2021. From the onset of the 2020-2021 season, the club started a collaboration with the Erasmus University Medical Center department of Viroscience, to implement a set of infection control regulations and perform SARS-CoV-2 RNA and serological testing and analysis. This study was approved by the Erasmus MC Medical Ethical Committee (MED-2021-0447).
At the start of this study, players and staff were extensively informed about which preventive measures each individual could take to prevent infection as much as possible. Emphasis was also placed on the fact that regular testing for possible contamination does not prevent contamination. From August 2020 until May 2021 all players and staff members in physical contact with players underwent weekly or biweekly testing for the presence of SARS-CoV-2 RNA in nasopharyngeal swabs collected by a clinical virologist. The implemented health and safety protocols were in agreement with previously published recommendations (Table 1). 5,6 All participants had 24/7 access to (telephone) consultations with the club physician and the clinical virologist. They were instructed to immediately report any symptoms or contacts with symptomatic individuals, subsequently underwent SARS-CoV-2 RNA testing, and awaited their results at home. Symptomatic family members were tested at their municipal health facility or in selected cases by the clinical virologist. Individuals with confirmed SARS-CoV-2 infection were instructed to remain at home for at least 7 days from symptom onset and resumed their activities only upon complete resolution of symptoms.
The training center, including dressing rooms, a fitness center, treatment rooms, wellness and lounge areas, and a restaurant, comprised a total surface area of 2247 m 2 . The frequency of cleaning and disinfection of indoor surfaces was increased to 2 times per day from the onset of the season. In all indoor spaces, players and staff members followed instructions to maintain a 1.5 m distance from others and to conduct regular hand hygiene. The use of face masks was obligatory in any indoor situation where maintaining distance was not possible such as the treatment of players by a physical therapist. For road transfers, a second bus was used to decrease the number of individuals per vehicle and special attention was given to the layout of the different busses. Players who had been through an infection were divided between the two busses as much as possible.
Real-time polymerase chain reaction (RT-PCR) and transcriptionmediated amplification (TMA) detection of SARS-CoV-2 on nasopharyngeal swabs were performed using the SARS-CoV-2 test on a Cobas® 6800 system (Roche Diagnostics) or Aptima SARS-CoV-2/FLU assay or Aptima Sars-CoV-2 assay (Panther R ). Subsequently, all positive samples were sequenced using a SARS-CoV-2 specific multiplex PCR for Nanopore sequencing. 7 Libraries were generated using native barcoding kits from Nanopore EXP-NBD104 and EXP-NBD114 and SQK-LSK109, and sequenced on a R9.4 flow cell. The resulting raw sequence data were demultiplexed using Porechop. 8 Primers were trimmed and a reference-based alignment was performed. The consensus genome was extracted and positions with a coverage < 30 were replaced with an N. 9 All sequences were aligned using MUSCLE. 10 Sequences with >10% Ns were excluded. The alignment was manually checked for discrepancies after which IQ-TREE 11 was used to perform a maximum likelihood phylogenetic analysis under the GTR + F + I + G4 model as the best predicted model using the ultrafast bootstrap option with 1000 replicates. The phylogenetic trees were visualized in Figtree. 12 A serological survey was conducted at the start of the study to assess the SARS-CoV-2-specific immunity status of the study population. Sera were tested for the presence of IgG antibodies against SARS-CoV-2 spike protein using either a quantitative chemiluminescence immunoassay (LIAISON-XL® SARS-CoV-2 TrimericS IgG assay, DiaSorin) or a qualitative ELISA (Beijing Wantai Biological Pharmacy Enterprise Co., Ltd.). Results were interpreted as positive according to instructions of the manufacturers (> 33.8 BAU/mL and optical density ratio > 1, respectively).

Results
The population characteristics are shown in Table 2. Between August 2020 and May 2021 a cohort of 74 players and staff members was followed-up for an average of 254 days. In a serosurvey performed at the beginning of the study, 11 participants (14.9%) had serological evidence of previous infection. Throughout the study period, 13 participants (17.6%) were tested positive for SARS-CoV-2 by PCR. Of these, 4 cases were diagnosed prior to symptom onset and 4 cases were positive at the time of enrolment. All cases were mild and the reported symptoms included fatigue, fever, myalgia, loss of smell, nasal congestion, and headache, with a duration ranging from 2 days to 5 weeks. No cases developed long COVID-19. All players who tested positive for SARS-CoV-2 RNA underwent an electrocardiogram (ECG), cardiac ultrasound, and MRI to evaluate myocardial pathology. No myocardial abnormalities were observed on imaging or ECG. Fig. 1 shows a timeline of positive test results with a phylogenetic analysis to assess whether cases were linked. The date of diagnosis and PANGO lineage for each case are shown in Supplementary Table 1. Two clusters of cases were identified within the cohort. The first cluster (N = 2, cases 2 and 3), involved two players who were diagnosed upon enrollment. This implies that transmission must have occurred before the implementation of health and safety measures, which limits the cluster's ability to inform on the effectiveness of these measures. The second cluster (N = 3, cases 6,7, and 8) involved three players. All three individuals tested positive on the same day, making it difficult to determine the source and secondary cases. One player reported a match as a potential moment of transmission. Since a negative test was required to participate in the match, it is possible that the infection was acquired from a player on the opposing team. Sequence analysis could not be performed for 5 positively tested participants due to a low viral load in the diagnostic sample.
Seven participants received a COVID-19 vaccine before the end of the study, with the majority receiving their vaccinations in late April or May, just prior to the study's conclusion. There were no instances of vaccine breakthrough infections observed among these participants.

Discussion
This study demonstrates that a combination of regular testing and basic control measures can prevent outbreaks of COVID-19 in a professional sports team. Through sequencing of the SARS-CoV-2 genome, we were able to distinguish between infections that were introduced from the community and infections that were transmitted within the study population. Our data show that introductions from the community are likely to occur, but weekly testing led to early detection (31% was diagnosed prior to symptom onset) and thereby prevention of secondary cases.
Data from Germany and Qatar support that the safe practice of professional football is possible in spite of ongoing viral circulation. The German Bundesliga reported 14 infections among over 1700 players and staff members during the months of May -July 2020 (5 cases per 100,000 per week), which was a period of low incidence in the general population. 13 A cohort study of 1337 players and staff in Qatar detected 85 cases during a 9-week period (457 cases per 100,000 per week), transmission during training or matches was deemed improbable. 14 In our study we found a similar incidence: 17.6% of the players and staff members acquired COVID-19 during the study period (484 cases per 100,000 per week).
Two main strengths of this study were the use of whole genome sequencing and the longer follow-up time which included periods of heavy viral circulation in the general population. Because players and staff members continuously move between the training center and private environments, determining the source of infection with genomics was essential to evaluate the effectiveness of the preventive measures. Indeed, the majority of infections that were detected were community-acquired, only 3 study participants acquired infection from another player or staff member (23% of all positives).
At the time of this study, mass vaccination programs had not been rolled out yet and the dominant circulating SARS-CoV-2 variants were wild-type (harboring the D614G mutation) and alpha (B.1.1.7). Since then, several game-changing events have altered the COVID-19 pandemic landscape. The advantage of studying transmission in a highly susceptible (unvaccinated) population is that transmission events were not blocked by vaccine immunity. Therefore, it is more likely that the absence of outbreaks can indeed be attributed to effective prevention. The emergence of variants that escape vaccine-induced immunity underline the relevance of these data. 15 As our protocol included regular (weekly) testing, it is possible that some cases remained undiagnosed, furthermore, cases diagnosed elsewhere and outside of the scope of this study were not included in the analysis. Although the detection rate may be higher with more frequent testing, our data show that little benefit could be expected in terms of further reductions in transmission. Finally, in several respiratory samples with low SARS-CoV-2 viral load, it was not possible to perform sequence analysis. With one exception (case 12 in Fig. 1), the distance in time between these cases was sufficiently large to discard the possibility of epidemiological relatedness.
Outbreaks of respiratory viruses pose an important challenge to athletes around the world. In many sports, social distancing and the use of masks are not feasible. Here, using whole genome sequencing, we demonstrate that weekly PCR testing combined with a basic set of behavioral rules leads to early detection and isolation of COVID-19 cases, allowing professional football players to safely play throughout the season. Compliance of players and staff is essential and should be emphasized. To ensure the safety of teammates, it is important to take responsibility for minimizing exposures in private settings and to remain consistently vigilant for potential symptoms. Like any strong opponent, this coronavirus can only be controlled by team play.

Conclusions
Through the use of whole genome sequencing, we demonstrate that a combination of regular testing and basic control measures can prevent outbreaks of COVID-19 in a professional sports team. SARS-CoV-2 introductions from the community are likely to occur, but weekly testing led to early detection and prevention of secondary cases.