Monitoring the SARS-CoV-2 outbreak in the UK

After a year of relatively uneventful evolution, the emergence and global spread of severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) variants signals an urgent need for better genetic monitoring (1). The United Kingdom (UK) has become a leader in this area. A £ 20million investment in March 2020 established the COVID-19 Genomics UK (COG-UK) consortium (2), which has produced over 200,000 SARS-CoV-2 genomes, more than double the number produced by any other country. Such a volume of data offers an unprecedented opportunity to trace which human activities are causing epidemic growth during a rapidly evolving pandemic, but also presents many bioinformatics challenges. On page 708 of this issue, du Plessis et al. (3) describe a new hybrid phylogenetic approach that integrates genetic data with epidemiological and travel data to uncover the roots of the UK’s severe spring epidemic. Notably, they find that the British epidemic results from more than 1,000 lines of transmission sown by travelers from Europe.

The study shows how last winter’s control efforts consistently lagged behind the virus, allowing SARS-CoV-2 to penetrate national borders. Their analysis of 26,000 UK footage from January to June 2020, the largest study of its kind, reveals that the UK outbreak was primarily introduced to the country by travelers from European neighbors: first Italy, then Italy ‘Spain and France. The peak of viral flow in the UK came in March as the virus spread across Western Europe, but monitoring delays have led to restrictions still focused on travelers arriving from Asia. By capturing a large number of small transmission lines that would go undetected at lower levels of virological surveillance, as well as> 1,600 singleton viruses with no offspring observed, the authors discovered an unprecedented volume of cross-border viral traffic. The genetic patterns mirrored patterns of human movement, as the number of viruses entering the UK rose and then fell after international travel fell in March.

Multiple introductions of SARS-CoV-2 by travelers from Italy, Spain and France, but not Asia, sowed the epidemic in the UK between January and June 2020.

PHOTO: TOBY MELVILLE / REUTERS

The UK is not the only country whose first focus on Asia as the epicenter of the pandemic allowed viruses to enter European sources. Genetic data has also traced the origins of the epidemics in Brazil (4), Boston (5) and New York (6) back in Europe. Travel restrictions can be very effective when rigorously enforced, but these studies collectively highlight the ease with which SARS-CoV-2 infection can occur even during small border control failures, including the repatriation of Americans from Asia at the start of the pandemic (7).

There is no silver bullet to triangulate scalability, speed, and statistical rigor, as genomic data exceeds the capacity of existing platforms. From Plessis et al. confronted with the methodological challenges encountered in the previous evolutionary analyzes of SARS-CoV-2 (8), enlarged in this case by a significantly larger data set. These challenges include a low phylogenetic signal among genetically similar viruses, exceeding the capacity of standard phylogenetic software, as well as biases that appear when other countries sequence different numbers of viruses compared to the national case count. The authors are pursuing a new approach that uses genetic data to infer the time and number of virus introductions, but uses epidemiological metadata to infer the country of origin. Better integration of genomic and epidemiological data will continue to improve outbreak responses, but can be cumbersome without open-access data repositories – for example, for fluctuations in global air transport volumes. Epidemiologists are increasingly turning to digital and cellular data from crowds to trace human movements and patterns of social contact (9).

Contact tracing has been effective in controlling early outbreaks of COVID-19, such as the first outbreak in Europe in Munich, Germany (ten), and providing key information on community transmission and the role of overcasting (11). But contact tracing is laborious and is often abandoned as epidemics develop. Genetic data can add a new dimension to these efforts by effectively determining whether two cases belong to the same line of transmission despite sampling gaps among individuals in the chain. From Plessis et al. did not explore the heterogeneities of transmission at the city level (5), but their observations reveal the growth and size-dependent extinction of hundreds of co-circulating lineages as the national epidemic was brought under control by non-pharmaceutical interventions (NPIs).

The Plessis study et al. makes use of a fraction of the UK footage generated to date. The risk of the emergence of new variants increases as populations of SARS-CoV-2 increase globally, spreading into immunocompromised, chronically infected or even non-human hosts where they experience different selection pressures. As SARS-CoV-2 becomes more evolutionary dynamic, the UK’s well-sampled data provide a resource for the global community. Denmark, Australia and other countries also have intensive SARS-CoV-2 sequencing operations. But the UK is currently the only country with over a million cases of COVID-19 that sequences over 1% of the SARS-CoV-2 genomes (UK sequences ∼5%).

The thorniest evolutionary issues require broad population-level analyzes based on continuous nationally representative sampling, with random selection of viruses to be sequenced (12). A centrally coordinated sampling strategy is a very beneficial feature of the UK virological surveillance program, although it is less quantifiable than speed or volume (2). The United States has produced the second highest number of SARS-CoV-2 genomes, but the proportion of sequenced cases varies widely between cities and states due to differences in resources. Large-scale studies become methodologically difficult when datasets are amassed from smaller studies originally designed to address other research questions, introducing bias. At times it has been difficult to assess intriguing hypotheses, for example whether SARS-CoV-2 containing the D614G spike protein mutation has spread globally due to the benefits of fitness or the random chance (13).

Variations that occur in one country quickly become a threat to neighbors. Countries must reciprocate each other’s virological surveillance efforts in a rapidly changing global viral landscape. The UK ARTIC network actively shares resources and protocols for SARS-CoV-2 sequencing. NextStrain provides a user-friendly visual platform to follow the development of SARS-CoV-2 in near real time. Numerous open access bioinformatics tools have been developed to analyze SARS-CoV-2 sequences (14). But one lesson from the UK is the importance of sustained government investment in scalable national infrastructure. Intrepid university researchers can create popular tools but struggle to expand as the amount of genomic data explodes. Global coordination would also be useful, including the universal adoption of a single nomenclature for SARS-CoV-2 lines.

The COVID-19 pandemic has galvanized long-awaited investments in promising research areas at the frontier of technology and big data. Over the past two decades, faster, cheaper, and more portable sequencing technologies and flexible bioinformatics platforms have laid the groundwork for real-time genomic epidemiology. Progress tends to be driven by public health crises, including influenza, Ebola and Zika outbreaks (15). The COVID-19 jump has begun.

Thanks: The content does not necessarily reflect the views or policies of the Department of Health and Human Services, nor does it imply endorsement by the US government.