Since January 2020 Elsevier has created a COVID-19 resource centre with

free information in English and Mandarin on the novel coronavirus COVID-

19. The COVID-19 resource centre is hosted on Elsevier Connect, the
company's public news and information website.

Elsevier hereby grants permission to make all its COVID-19-related

research that is available on the COVID-19 resource centre - including this
research content - immediately available in PubMed Central and other
publicly funded repositories, such as the WHO COVID database with rights
for unrestricted research re-use and analyses in any form or by any means
with acknowledgement of the original source. These permissions are
granted for free by Elsevier for as long as the COVID-19 resource centre
remains active.
 ll

Review
The origins and potential future of SARS-CoV-2
variants of concern in the evolving COVID-19 pandemic
Sarah P. Otto1,*, Troy Day2, Julien Arino3, Caroline Colijn4, Jonathan Dushoff5, Michael Li6, Samir Mechai7,
Gary Van Domselaar8,9, Jianhong Wu10, David J.D. Earn11, and Nicholas H. Ogden7
1Department of Zoology and Biodiversity Research Centre, University of British Columbia, Vancouver, BC V6T 1Z4, Canada
2Department of Mathematics and Statistics, Department of Biology, Queen’s University, Kingston, ON K7L 3N6, Canada
3Department of Mathematics and Data Science Nexus, University of Manitoba, Winnipeg, MB R3T 2N2, Canada
4Department of Mathematics, Simon Fraser University, Burnaby, BC V5A 1S6, Canada
5Department of Biology and M.G. DeGroote Institute for Infectious Disease Research, McMaster University, Hamilton, ON L8S 4L8, Canada
6Public Health Risk Sciences Division, National Microbiology Laboratory, Public Health Agency of Canada, Guelph, ON N1G 3W4, Canada
7Public Health Risk Sciences Division, National Microbiology Laboratory, Public Health Agency of Canada, St. Hyacinthe, QC J2S 2M2,

Canada
8National Microbiology Laboratory – Public Health Agency of Canada, Winnipeg, MB R3E 3R2, Canada
9Department of Medical Microbiology and Infectious Diseases, University of Manitoba, Winnipeg, MB R3E 0J9, Canada
10Laboratory for Industrial and Applied Mathematics, York University, Toronto, ON M3J 1P3, Canada
11Department of Mathematics and Statistics and M.G. DeGroote Institute for Infectious Disease Research, McMaster University, Hamilton,

ON L8S 4K1, Canada

*Correspondence: otto@zoology.ubc.ca
https://doi.org/10.1016/j.cub.2021.06.049

SUMMARY

One year into the global COVID-19 pandemic, the focus of attention has shifted to the emergence and spread
of SARS-CoV-2 variants of concern (VOCs). After nearly a year of the pandemic with little evolutionary change
affecting human health, several variants have now been shown to have substantial detrimental effects on
transmission and severity of the virus. Public health officials, medical practitioners, scientists, and the
broader community have since been scrambling to understand what these variants mean for diagnosis, treat-
ment, and the control of the pandemic through nonpharmaceutical interventions and vaccines. Here we
explore the evolutionary processes that are involved in the emergence of new variants, what we can expect
in terms of the future emergence of VOCs, and what we can do to minimise their impact.

Introduction site per day7, roughly 5-fold lower than influenza A/H3N2 (10.9 3
The December 2020 announcement by Public Health England1 10
6 nucleotide substitutions per site per day; https://nextstrain.
that the variant B.1.1.7 exhibited a large number of mutations org8; accessed 22 May 2021). Thus, across the 30,000 base-
and a significant increase in transmission (50–100%2) rattled pair genome of SARS-CoV-2, approximately 20 genetic changes
the world and served as a wake-up call on the importance of occur per year within a lineage.
VOCs. Until then, there was little evidence that any mutations Not all mutations that arise will persist long enough to be
to the RNA genome of SARS-CoV-2 substantially increased included in this estimate of the molecular clock of SARS-CoV-
viral fitness3. Within weeks, additional VOCs were reported 2. Assuming that synonymous mutations are largely neutral while
with similar characteristics: greater than expected numbers of non-synonymous mutations persist with probability u, Wang
mutations and signatures of enhanced transmission (for et al.9 used codon-based likelihood methods to estimate u =
example, B.1.351 in South Africa4 and P.1 in Brazil5). 0.56 among human SARS-CoV-2. Thus, roughly half of the mu-
The genomes of all viruses accumulate mutations over time. tations that alter amino acids are quickly lost. Given the propor-
However, the pace of mutation accumulation and the conse- tion of sites subject to synonymous (pS = 0.22) and non-synony-
quences for transmission and disease in the host population mous (pN = 0.78) mutations, we can thus back-calculate the rate
depend on a range of factors, including the mutation rate and at which mutations must have occurred, before selection acted,
the impacts of mutation on viral dynamics within and between in- to be 2.84 3 10
6 (that is, 1.87 3 10
6/(pS + u pN)). This correc-
dividual hosts. Together these factors determine the emergence tion is slight because selection has had little time to eliminate
and spread of viral variants and the evolution of pandemics. The anything but the most deleterious mutations accumulated over
genomes of RNA viruses are particularly prone to mutation6. But the pandemic’s short time frame. Over longer evolutionary time
SARS-CoV-2, like related coronaviruses, encodes a proof- frames, such as that observed between SARS-CoV-2 and the
reading domain (ExoN) that reduces its mutation rate relative most closely related bat sequences, Wang et al.9 find that non-
to RNA viruses that do not (such as influenza, HIV, and hepatitis synonymous mutations are 14 times less likely to persist (u =
C viruses). The pace of change of the SARS-CoV-2 genome has 0.039) than what we see today among the virus in humans, as
thus been estimated at 1.87 3 10
6 nucleotide substitutions per selection has slowly eliminated weakly deleterious mutations.

R918 Current Biology 31, R918–R929, July 26, 2021 Crown Copyright ª 2021 Published by Elsevier Inc.
 ll
Review

Box 1. Variant detection and terminology.

As viruses accumulate mutations, it can be challenging to know how to refer to the diverse virus forms that arise. A ‘variant’ is any
virus with a different sequence from other viruses. A ‘lineage’ refers to viruses that are more closely related — cousins on the phylo-
genetic tree of the virus. A ‘strain’ is reserved for a broader grouping of viruses with different properties; for example, SARS-CoV-1
and SARS-CoV-2 represent different strains of coronavirus.
The World Health Organization71 has recently established guidelines for referring to viral variants for SARS-CoV-2. A ‘variant of
interest’ (VOI) contains mutations thought likely to alter the phenotypic properties of the virus, with documented community trans-
mission or international spread. Because VOIs might alter transmission rates or disease progression, impacting human health, they
should be monitored closely.
In addition, a ‘variant of concern’ (VOC) has an established and detrimental effect on human health, with mutations that increase
viral transmission rates, cause more severe health outcomes, and/or reduce the efficacy of public health measures such as vacci-
nation. Systematic genomic surveillance is needed to allow the epidemiological analysis necessary to determine whether a VOI
(that is, a variant of potential impact) should be designated a VOC (that is, one of demonstrated impact).
In many cases, variants have been observed to increase in frequency, but this alone is not sufficient information to document an
increased transmission rate and justify designation of a VOC. Founder effects and transmission to a new segment of the population
can lead, by chance, to the rapid growth of a lineage. Shifts in travel patterns over time can also cause changes in the frequency of
viral lineages. Uneven sampling presents another challenge, especially if samples are more likely to be sequenced if they are
suspected to be VOI or VOC. Particularly confounding is an immense statistical problem associated with multiple comparisons;
with 2 million genomes sequenced and frequencies tracked in hundreds of regions across the globe, random changes in variant
frequency may often be confused with a transmission advantage. Verifying a transmission advantage requires repeatedly showing
that a variant increases in frequency, over multiple weeks and regions, as documented for B.1.1.7 in the UK1,2,29.

The immense number of currently active cases — 12.1 million population and point to natural selection for the virus to better
globally (https://www.worldometers.info/coronavirus; accessed replicate or to evade a weakened immune system, as well as
2 July 2021) — has greatly increased the opportunity for viruses the intensive antibody therapies received by these patients11–13.
with distinct characteristics to evolve. Several viral lineages have A disproportionate fraction of mutations in many of these pa-
now been reported that have potential impacts (so called ‘variant tients are clustered in the gene encoding Spike, the protein
of interest’, VOI) or demonstrated impacts (‘variant of concern’, that juts out of the virus and binds to the ACE2 receptor that al-
VOC) on disease transmission and human health (see Box 1 for lows entry into host cells: for example, 57% of 15 changes in the
terminology). At the time they emerged, the first three VOCs final sample reported by Choi et al.11 and 33% of 9 changes in
(B.1.1.7, B.1.351 and P.1) exhibited almost twice as many the final nose and throat sample reported by Kemp et al.13,
changes to their genomes as other contemporaneous SARS- compared to an expectation of only 13% based on the length
CoV-2 lineages, with 23 mutations characterizing B.1.1.710, 21 of Spike. Many of these mutations were found in the receptor
in B.1.3514, and 23 in P.15. The B.1.617.2 variant causing surging binding domain, which is essential for host entry. This concentra-
case numbers in India and recently designated a VOC (Table 1) tion in Spike may reflect, in part, relaxed selective constraints
also exhibits an excess of mutations, roughly similar to B.1.1.7 within immunocompromised individuals, but many of these mu-
(Figure S1). tations are non-synonymous (for example, all eight mutations in
Spike reported by Choi et al.11), consistent with selection favour-
Factors at play in the emergence and spread of VOCs ing changes to the Spike protein. Furthermore, several mutations
The establishment and spread of such variants depend on two have arisen in parallel in different patients, suggesting selection
main factors: what happens within the individuals in whom the for those changes15. For example, McCarthy et al.14 report four
mutations arose and what happens afterwards as the virus trans- independent cases of patients bearing deletions in the same
mits among individuals. amino-terminal domain of Spike, which is important for host
Processes occurring within individuals entry and antibody evasion. These non-random patterns of
Mutations arise as viruses replicate within an infected individual, genomic changes suggest that selective pressures, alongside
and thus new variants initially face selective forces within that in- mutation, strongly shape viral evolution within immunocompro-
dividual. For SARS-CoV-2, these within-individual evolutionary mised individuals. Immunocompromised patients should be pro-
processes have been best documented in immunocompromised tected from COVID-19 by prioritizing their contacts for vaccina-
patients11–14. These patients maintain high viral loads over pro- tion, and any patient with a prolonged infection should be treated
longed periods of time, allowing more opportunities for viral with great care as a potential source of new variants.
replication and selection and leading to elevated substitution The changes observed in the three main VOCs to date (B.1.1.7,
rates (see elevated branch lengths in, for example, refer- P.1, B.1.351) echo the evolutionary changes observed in immuno-
ences11,13, which reflect a larger than expected number of sub- compromised individuals, with a larger than expected number of
stitutions). By sequencing the virus at multiple time points, these mutations that are non-randomly distributed across the genome.
studies have documented rapid changes to the composition of This pattern led Public Health England1 to suggest that B.1.1.7
the viral population within a patient, over the course of days. might have originated in an immunocompromised individual.
These rapid changes are faster than expected by drift in a large Many mutations in these VOCs are also concentrated in Spike

Current Biology 31, R918–R929, July 26, 2021 R919

 Table 1. SARS-CoV-2 variants of concern and variants of interest.
R920 Current Biology 31, R918–R929, July 26, 2021

Pango Nextstrain First detection VOI or WHO71 Observed clinical effect

lineage clade8 location VOC designation Mutations of interest on S protein Transmissibility Virulence Antigenicity
2
B.1.1.7 20I/501Y.V1 United Kingdom VOC Alpha N501Y, E484K*, P681H, 50–100% higher 39–72% No impact on NBA

ll
D614G more lethal41 Minimal impact on NBSa
B.1.351 20H/501Y.V2 South Africa VOC Beta N501Y, K417N, E484K, 20–113% higher72 – Moderately reduced NBAa
D614G Reduced NBSa
P.1 20J/501Y.V3 Brazil/Japan VOC Gamma N501Y, K417T, E484K, 70–140% higher, 20–90% Moderately reduced NBAa
D614G evades immunity more lethal5 Reduced NBSa
21–46% more5
B.1.427 and 21C/S:452R United States VOI Epsilon L452R, 18–22% higher73 – Moderately reduced NBAa
B.1.429 (California) D614G Reduced NBSa
B.1.525 21D United States VOI Eta A67V, E484K, D614G, Q677H, F888L – – Potentially reduced NBAa
(New York)/Nigeria Potentially reduced NBSa
B.1.526 21F United States VOI Iota L5F*, T95I, D253G, S477N*, E484K, – – Moderately reduced NBAa
(New York) D614G, A701V* Reduced NBSa
B.1.617.1 21B/S:154K India VOI Kappa (T95I), G142D, E154K, L452R, E484Q, Secondary attack – Potentially reduced NBAa
D614G, P681R, Q1071H rates similar to Potentially reduced NBSa
B.1.1.774
B.1.617.2 21A/S:478K India VOC Delta T19R, G142D*, 156del, 157del, R158G, Secondary attack – Potentially reduced NBAa
L452R, T478K, D614G, P681R, D950N rates higher than Potentially reduced NBSa
B.1.1.7; household
transmission 64%
higher than B.1.1.7
(26–113% higher)74
B.1.617.3 20A India VOI T19R, G142D, L452R, E484Q, D614G, – – Potentially reduced NBAa
P681R, D950N Potentially reduced NBSa
P.2 20J Brazil VOI Zeta E484K, D614G, V1176F – – Potentially reduced NBAa
Reduced NBSa
Here we have used the variant classification of the Centre for Disease Control of the United States and modified from the CDC data table, which lists additional mutations and information about
clinical effectsa. Nextstrain clade names start with the year of origin and distinguish lineages that reach a global frequency of 20%, so the name of a lineage can change if it increases in frequency8.
NBA: Neutralization by antibodies (monoclonal antibodies in therapeutic use); NBS: Neutralization by convalescent and/or post-vaccination sera (variant relative to non-variant). VOI and VOC des-
ignations vary over time (e.g. B.1.427 and B.1.429 were previously designated as VOC by the CDCa, accessed 13 June 2021) and by country (e.g. Canada designates the entire B.1.617 clade as a
VOCb) because of different evaluations of the existing evidence.
*Mutation detected in some sequences within the lineage. Ranges give 95% confidence intervals or credible intervals, except for the transmission rate of B.1.1.7 (a consensus estimate across
models2) and for P.1 (50% Bayesian credible intervals).
a
https://www.cdc.gov/coronavirus/2019-ncov/cases-updates/variant-surveillance/variant-info.html; ‘moderate’ includes modest decreases and/or cases where alternative antibody treatments
remain available.
b
https://www.canada.ca/en/public-health/services/diseases/2019-novel-coronavirus-infection/health-professionals/testing-diagnosing-case-reporting/
sars-cov-2-variants-national-definitions-classifications-public-health-actions.html; (all accessed 2 July 2021).

Review
 ll
Review

1.0 suggest that few viruses found subsequent infections; for

example, only 1–8 virions17. As a result, there is little genetic di-
Probability of establishment

versity, allowing little opportunity for evolutionary change within

0.8
typical infections17,18. Instead, drift is likely to dominate which vi-
ruses cause new infections, with one main exception: the elimi-
0.6 Contact/activity Poisson nation of strongly deleterious mutations. Indeed, Lythgoe et al.17
heterogeneity Transmission
heterogeneity report a similar reduction in the number of nonsynonymous mu-
0.4 tations observed within individuals (u = 0.55) as estimated from
different individuals (u = 0.56)9, suggesting that the elimination of
strongly deleterious mutations primarily occurs within individual
0.2
hosts.
Elevated mutation rates can also be caused by genetic
changes in SARS-CoV-2 that increase the error rate during viral
0.00 0.05 0.10 0.15 0.20 0.25
replication19. Takada et al.19 showed that mutation P203L in
Probability of transmission given a contact, T
NSP14 alters the ExoN proofreading domain and nearly doubles
Current Biology the mutation rate. Such ‘mutator’ lineages could also generate
variants with rare combinations of new mutations, but to date,
Figure 1. The role of heterogeneity in the probability that a variant these mutators have not risen in frequency, possibly because
establishes within a population.
Illustrated here is a predominantly susceptible population, with an average of the higher numbers of deleterious mutations that result19.
number of ten contacts per case and no competition for susceptible hosts Another potential source of novel variants within individual in-
between the variant and non-variant. If the number of contacts per individual is fections is recombination. There is debate about the extent of
Poisson distributed and there is a constant chance of infection per contact, the
probability of establishment rises with the chance of infection per contact as
recombination in SARS-CoV-220,21, in part because detecting
shown by the black curve. Here, variants are not expected to persist unless the recombination is challenging when there is relatively little
transmission probability is above 10%, as only then are cases expected to give genomic variation22. All circulating SARS-CoV-2 viruses are
rise to at least one new case (Rt > 1). If cases vary in their infectiousness, then similar and closely related by descent from the virus that first in-
variants are less likely to establish because more cases fail to have any onward
transmission (blue curve, where we assume that half of the cases are three fected humans in late 2019. Furthermore, artefacts generated
times as infective as the other half). If, however, there is variability in contact during genomic sequencing and assembly can mimic recombi-
number or activity level, variants are more likely to establish because in- nation and may explain some of the apparent recombination
dividuals with more contacts are more likely to get infected and then more
likely to pass on the variant (red curve, assuming the contact distribution is
events21. That said, a recent analysis provides compelling evi-
negative binomial with a dispersion parameter of k = 3). Because the disease dence that sampled genomes from the UK arose by recombina-
spreads more easily among the subset of active people, heterogeneity in tion during a period when B.1.1.7 and non-variant lineages were
contacts also reduces the critical transmission probability above which
both prevalent23. Recombination is a concern because it gener-
establishment is possible (red curve rises above zero earlier, causing Rt > 1).
(Based on methods in reference75.) ates new combinations, potentially bringing together compo-
nents of different lineages in a way that benefits the virus,
although no fitness advantage has been detected for the re-
(35% of the 23 mutations in B.1.1.710). All three contain a specific combinants observed to date23.
deletion (Orf1ab:3675–3677del) affecting NSP6, a protein Processes occurring among individuals
involved in intracellular processes that alter the balance between For those SARS-CoV-2 mutations that do make it out of the body
viral replication and clearance. All three lineages also contain and cause new infections, the probability that the variant be-
S:N501Y, and two VOCs (B.1.351 and P.1) bear mutations in com- comes established depends on its transmissibility in the popula-
mon at three other sites in spike (S:L18F, S:K417N/T, S:E484K; tion in which it emerges, as well as the nature and extent of con-
where this notation indicates, for example, ‘S’ for the spike tacts among individuals and chance events. For many infectious
gene, ‘L’ for the original amino acid leucine, ‘18’ for the amino diseases, a small fraction of individuals tends to be responsible
acid position, and ‘F’ for the mutant amino acid phenylalanine). for a large fraction of the transmission events (the so-called ‘80/
Given the number of characteristic mutations in these three 20 rule’24, in which 80% of new infections are thought to be
VOCs and the genome size, there is a <5% probability that we caused by 20% of cases). This pattern has also been observed
would see the same site mutated in any of the variants just by for SARS-COV-225,26 and is sometimes referred to as ‘overdis-
chance. These parallel changes are thus a strong signal that se- persion’, meaning that the variation in the number of new infec-
lection played a role in the initial emergence of the VOCs, espe- tions generated by different individuals is larger than expected
cially given the weak constraints generally observed9. (typically with reference to a Poisson distribution).
We have focused above on immunocompromised individuals Overdispersion can arise through various processes, and the
because there is limited opportunity for evolution within typical details have a strong effect on the probability that a new variant
infections (that is, those within immunocompentent hosts). establishes within a population (by established, we mean that the
With exponential growth of the population of viruses within a variant is present in enough active cases that it is very unlikely to
host, most mutations at peak infectivity would have occurred be lost by chance). If overdispersion arises because individuals
in the last round or two of replication, leaving little time for selec- vary in infectivity, with some cases transmitting much less often
tion within that individual (this is the principle underlying the and others much more often than expected (for example, ‘super-
classic experiment of Luria and Delbrück16). Furthermore, esti- spreaders’), then this variability tends to decrease the probability
mates of the bottleneck size between transmission events of establishment of a variant27 (Figure 1, compare black and blue

Current Biology 31, R918–R929, July 26, 2021 R921

 ll
Review

1.0 London lower transmission, on average31). Even for those who do have
Frequency of SGTF

Northeast symptoms, roughly half of the transmission events are estimated

0.8 Northwest to occur before individuals exhibit symptoms (‘pre-symptomatic’
East Midlands transmission), at least in areas where individuals are encouraged
0.6
West Midlands and able to self-isolate32. Most symptomatic individuals recover
0.4 East of England
from COVID-19, but even so the infection fatality rate is high
Southeast
0.2 Southwest at about 0.68% (95% CI: 0.53–0.82%) globally33. In many
Yorkshire infectious diseases, direct selection is expected to reduce viru-
lence34. However, because severe symptoms and mortality
Dec. 14

Jan. 18
Sept. 7

Nov. 9
Oct. 5

from COVID-19 typically occur weeks after peak viral load, direct
selection against virulence is expected to be extremely
Current Biology
weak35–37. By contrast, selection strongly favours an increased
transmission rate in the early stages of a pandemic, when the
Figure 2. The rise in frequency of B.1.1.7 in nine regions of England.
Data are weekly measures of the fraction of Pillar 2 tests that showed SGTF,
abundance of susceptible hosts is high35,38–40. These general
taken from Public Health England Technical Briefing 5 (http://www.gov.uk/ predictions are also borne out in SARS-CoV-2-specific models,
government/publications/investigation-of-novel-sars-cov-2-variant-variant-of- which additionally show that selection favours shorter latent pe-
concern-20201201).
riods prior to infectivity, higher infectiousness of asymptomatic
individuals, and prolonged infectious periods37.
curves). On the other hand, if the overdispersion occurs because Although the death rate caused by SARS-CoV-2 may be under
some individuals have many fewer contacts or much lower activ- relatively weak direct selection, mutations that affect transmis-
ity levels and others many more, this tends to increase the prob- sion rate or other disease attributes can have correlated effects
ability of establishment (Figure 1, red curve compared to a Pois- on mortality. Variants with even a moderately enhanced trans-
son distributed number of contacts in black). mission rate can readily spread during a pandemic, whether
The key difference between the two cases is that when over- they increase or decrease death rates. Thus, the evolution of
dispersion is due to chance events affecting the course of infec- virulence depends strongly on the nature of the mutations that
tion, then there is no relationship between the likelihood of arise. For example, if increased transmission rate is due to a
an individual acquiring a variant infection and that individual’s higher viral load, more severe illness may result. Alternatively, if
propensity to transmit to others. As a result, this simply intro- mutations cause milder symptoms and so raise activity levels
duces additional noise into the system, decreasing the probabil- of infected individuals, they may increase transmissibility but
ity of establishment (blue curve). By contrast, with overdisper- reduce severity and death rates.
sion due to contact or activity heterogeneity (red curve), Unfortunately, of these possibilities, B.1.1.7 is both more
variants are more likely to infect highly connected and active in- transmissible and more virulent41,42. Based on structural model-
dividuals who also have a higher propensity to transmit by virtue ling, the changes in the Spike protein in B.1.1.7 are predicted to
of their high connectivity. This process, whereby the most active cause conformational changes that allow the viruses to bind
individuals contribute disproportionately to transmission, also more easily to cells in the respiratory tract of an uninfected per-
explains why targeting vaccines and nonpharmaceutical inter- son43 (and so a lower infectious dose might be needed). Data
ventions to essential workers is particularly effective at reducing also suggest that infected individuals have a slightly higher viral
cases28. load and longer infectious period44. However, for B.1.1.7 at least,
After a favoured variant becomes established within a popula- a shorter serial interval does not explain the observed spread in
tion, its spread through a population rises in a more deterministic the UK2. The increased transmission has been accompanied by
fashion. Tracing the week-after-week spread of B.1.1.7 across increased virulence, with an estimated 64% higher mortality rate
many regions of the UK in great detail was made possible (95% CI: 32–104%)41,42. Currently unknown is which of the mu-
because of a deletion in the spike gene of B.1.1.7 (S:H69– tations in B.1.1.7 are responsible for the observed changes in
V70del) that caused one of three molecular probes to ‘drop transmission and virulence, which are selectively favoured, and
out’ in the PCR test for COVID-19 used in the UK (referred to which have risen by hitchhiking despite being neutral or even
as S-gene target failure’, or SGTF). By analysing data on the pro- deleterious.
portion of PCR tests in which this probe dropped out, B.1.1.7 While public health measures have reduced the reproductive
was estimated to be 50–100% more transmissible than previ- number (Rt) of non-variant SARS-CoV-2 to below 1 in many
ously circulating variants, depending on the model2,29. The countries, these measures were not initially strong enough
higher probability of transmission increases the likelihood of to control the spread of B.1.1.7, allowing it to become the pre-
B.1.1.7 establishing and spreading in new areas (Figure 1). dominant variant in many countries within months. Where
Within three months of the Public Health England announce- B.1.1.7 has established (say at 5–10% frequency), the subse-
ment, over 100 countries had reported B.1.1.730, despite quent rise in frequency has been fast. In the UK, the odds that
increased restrictions on travel from the UK. a case was caused by B.1.1.7 (that is, the frequency of B.1.1.7
The way that selection is expected to act on new SARS-CoV-2 divided by the frequency of non-variants) doubled every
variants is affected by the particular features of COVID-19, 1–2 weeks depending on the region (Figure 2). The average
including the extreme variability in disease presentation. About reproductive number is expected to increase alongside the rise
17% of cases never exhibit symptoms (‘asymptomatic’ cases31), in variant frequency, requiring even stronger public health mea-
and yet they can still infect others, although less often (42% sures to control the disease. This recent history is now being

R922 Current Biology 31, R918–R929, July 26, 2021

 ll
Review

repeated with B.1.617.2, which is estimated to be 64% more B.1.351 evade monoclonal antibody therapies, at least for
transmissible within households than B.1.1.7 (Table 1). some antibodies (Table 1). Many also exhibit weaker inhibition
Processes occurring among species by neutralizing antibodies in serum samples from vaccinated
Predicting the course of SARS-CoV-2 evolution is further individuals54, requiring 6.4-fold higher titers of antibodies to
complicated by the broad potential host range and known trans- neutralize B.1.351 and 3.5-fold to neutralize P.1 after two doses
missibility to animals45. Were transmission in domesticated and/ of Moderna (mRNA-1273). Real-world estimates also find
or wild animals to become an important reservoir for SARS- slightly lower vaccine effectiveness against known infections
CoV-2, then selection for greater transmission in these animals by variants B.1.351 (Pfizer BNT162b2)55 and B.1.617.2 (Pfizer
could lead to different disease attributes. For example, evolution BNT162b2 and AstraZeneca ChAdOx1)56. Nevertheless, vac-
of transmissibility of West Nile virus in American robins (a key cine effectiveness increases with the second dose and/or
reservoir host) resulted in increased virulence to crows46. For over time, reaching high levels of protection against infection
SARS-CoV-2, a host jump in Denmark from humans to farmed (>75%) and almost complete protection against severe dis-
mink and back to humans led to a highly mutated variant (‘Clus- ease55.
ter 5’), which was less easily neutralized by antibodies from Over the long term, the accumulation of multiple mutations
convalescent sera47. Cluster 5 initially reached a high frequency in SARS-CoV-2 will eventually reduce the efficacy of immune
(30%) in communities near affected farms, but through massive responses in a large proportion of vaccinated and naturally im-
control efforts, including the culling of 3–4 million mink, appears mune people (Box 2). How long might we expect immunity to
to have been eradicated47. As concluded by Larsen et al.47, last? In another coronavirus (229E), antibodies within blood
keeping susceptible animals in such large and dense popula- collected in the 1980s were shown to decline in neutralizing
tions poses a threat to human health by presenting an opportu- ability after a decade of evolutionary changes (‘antigenic
nity for viral adaptation. drift’57), suggesting years of immunity. But with so many active
Future evolution COVID-19 cases, further bursts of mutations, as seen in immu-
What types of variants are we likely to see over the next year? nocompromised patients and via passage through animals,
Although it is impossible to answer this question precisely, should be expected, accelerating evolutionary change. Alto-
evolutionary models can help us explore the possibilities. gether, it seems plausible that boosters may be needed within
As long as circulating strains cannot infect recovered or immu- the next year or two, with longer gaps between boosters once
nized individuals, then models predict that, as the number of sus- case numbers fall globally. Fortunately, updating mRNA vac-
ceptible individuals falls, selection for increased transmissibility cines to target new variants is relatively straightforward, and
will become weaker, with selection increasingly favouring pro- clinical studies are already underway for variant-specific vac-
longed infectious periods35,37–40. cines. For example, Moderna recently reported on phase 2 tri-
However, as the proportion of the population that is immu- als with mRNA-1273.351, a vaccine that matches changes in
nologically naı̈ve shrinks through infection and immuniza- Spike observed in the B.1.351 variant58. That said, the results
tion, selection will also increasingly favour variants that are showed very modest increases in antibodies targeting B.1.351
partially or wholly able to overcome the immunity of previously following two doses of the standard Moderna vaccine with
protected individuals (Box 2). In some countries, particularly mRNA-1273.351, suggesting that variant-specific vaccines
with limited medical and public-health resources, estimates may primarily boost previous immune responses rather than
suggest that most of the population may now have been in- convey new immune responses to the variant epitopes. Target-
fected, becoming naturally immune and increasing the relative ing escape variants with vaccines that target other proteins,
fitness of such immune-escape variants48. like the nucleocapsid, might be an effective alternative
The evolution of resistance to vaccines typically occurs more strategy.
slowly and is observed less often than the evolution of resistance Even without processes that make multiple changes more
to therapeutic agents like antimicrobials49. Both natural immu- likely, selection can favour the accumulation of mutations one
nity and vaccine-induced immunity generally result in a broad at a time that together reduce detection by the immune system
range of antibodies and T cells that recognize distinct parts and increase the risk of reinfection. The gradual build-up of
of the virus, particularly Spike and nucleoproteins50,51. Even escape mutants is particularly likely in regions where there are
though current mRNA vaccines only encode Spike proteins, a both high case numbers and high numbers of immunized individ-
range of antibody and T-cell responses are produced, recog- uals (Box 2).
nizing different parts (‘epitopes’) of the protein52, making it less In summary, the emergence of new variants should be ex-
likely that single mutational changes will substantially reduce pected, initially driven by selection for greater transmissibility
protection. Even with the multiple mutations observed in the and a longer duration of infectivity and then increasingly shaped
currently circulating VOCs, substantial natural and vaccine- by selection to evade immune responses, enabling transmission
induced antibody responses, and especially T-cell responses, to vaccinated and naturally immunized people. Although there
are observed against the variants tested (B.1.1.7, B.1.351, and currently appears to be only weak selection on virulence, viru-
P.152,53). Importantly, a mutant virus that escapes the immune lence may increase (or decrease) if coupled by mutation with
response of one individual may not escape the immune response these other features (as a pleiotropic side effect or through link-
of another. age). The further emergence of variants with multiple genetic
Although full escape from vaccine-induced immunity in changes, due to infections in immunocompromised individuals,
SARS-CoV-2 has not yet been documented, variants displaying mutators, and/or recombination, is expected and should be
partial escape are already here. P.1, B.1.617.2, and especially closely monitored.

Current Biology 31, R918–R929, July 26, 2021 R923

 ll
Review

Box 2. Escape mutations.

Each infection carries the risk of giving rise to a mutation that evades resistance in individuals who have specific immunity due to
either prior infection or vaccination. To illustrate the main factors determining this risk, we simplify SARS-CoV-2 dynamics and
consider a population consisting only of S susceptible individuals, I infected, and R resistant individuals, ignoring case heteroge-
neity (presymptomatic, asymptomatic, vaccination type and dose, etc.; see reference67 for a more detailed model focused on
within-individual adaptation). Although these variables all change over time, we focus on a region with a small current fraction
of active infections, so that S and R can be treated as constant over the short term, allowing us to focus on the dynamics of
the infectious class I and escape variants I .

Before escape variants are present, new cases arise at a rate b S I per day and are cleared at rate k (Box 2 figure, below). Each new
infection has a chance m of giving rise to an escape variant by mutation. Because we are considering the first such mutation, the
ability of the variant to infect resistant individuals is only partial, reducing the transmission rate to resistant individuals from b to p b.
At this snapshot in the epidemic, the number of standard cases and the number of escape-variant cases are expected to change at
rates:
dI
= ð1
mÞ b S I
k I (1)
dt

dI
= m b S I + b S I + p b R I
k I (2)
dt

The chance that an escape mutation appears on any particular day is thus proportional to ðm b S IÞ. While cases remain roughly
constant in number, the time until the first escape mutation appears is approximately exponentially distributed with a mean of
1=ðm b S IÞ days. The waiting time is thus shorter in locations with ineffective control measures (higher b), many circulating cases
(higher I), and many remaining susceptibles (higher S).

Once an escape mutant has appeared, equation (2) indicates that the mutant strain will increase in frequency at a faster rate than
the original strain, raising the reproductive number for the new variant from Rt = ðb SÞ=k to Rt = ðb S + p b RÞ=k. Preventing the
selective accumulation of partially resistant mutations thus requires reducing contacts between cases and resistant individuals
R I , boosting resistance where possible by vaccinating naturally infected individuals and completing recommended vaccine
doses (reducing p), and persisting with public health measures that reduce transmission in general (reducing b).

This basic model also clarifies the challenges inherent in deciding whether to spread available vaccines to more people in single
doses or to complete two-dose regimes. Vaccinating more people with a single dose can decrease the number of infections circu-
lating in a region (reducing I and I ) and reduce the reproductive number of all variants (by reducing S), but it also has the potential to
increase the probability of onward transmission to partially resistant individuals (increasing p R and the selective advantage of escape
variants). The right policy thus depends on resistance after one versus two doses and the impact on the number of severe cases.
IS

(1 – )
N
S I R
p I* R

N
I* R*
I*S
Current Biology

A simplified model of the evolution of the first step escape mutant (I*) using an SIR model framework.
The transmission rate b to susceptible individuals and the clearance rate k are assumed to be equal for both variants. Only the escape mutant is partially able to
infect resistant individuals (at a rate reduced by a factor p relative to the rate of infecting susceptible individuals). The model considers only the first step in the
accumulation of mutations that increase the ability to reinfect otherwise resistant individuals but could be extended to model subsequent steps by mutation.

R924 Current Biology 31, R918–R929, July 26, 2021

 ll
Review
Weekly new cases per 100,000 people

Effective ongoing surveillance also provides information about

200
evolutionary change during the pandemic. Understanding the
impact of these evolutionary changes can be important for guid-
150 Total ing control measures and for solidifying public support for these
Non-VOC measures. For example, as in many regions, B.1.1.7 recently
VOC increased in prevalence in Ontario Canada (Figure 3). During
100 the initial spread (from January–February 2021), however, the to-
tal new daily case count steadily decreased, leading many peo-
ple to believe that the pandemic was under control. But this
50 steady decrease in total case count masked the fact that control
measures at the time were insufficient to control B.1.1.7. As a
result, once the prevalence of B.1.1.7 reached a high enough
value, it began to dominate the total case count, causing the
020

overall daily case count to start rising once again (Figure 3).

021
021

21
4, 20

Although models predicted this evolutionary dynamic and the

10, 2

25, 2
21, 2

subsequent spike in cases based on the selective differences

Dec.

Mar.

Mar.
Jan.

observed in the UK (Figure 4), accurate surveillance data for

B.1.1.7 were not readily available to demonstrate unequivocally
Current Biology what was happening. Consequently, it was challenging to galva-
nize support for enhancing control measures proactively until the
Figure 3. Weekly new cases per 100,000 people for Ontario, Canada. total daily case count began to climb once again.
Data displayed as total count, those due to non-VOC, and those due to VOC Since the Public Health England announcement1, countries
(primarily B.1.1.7, as measured by a PCR test for the N501Y mutation).
Plots are moving seven-day averages, using data from https://covid19-
worldwide have also increased mandatory testing before and/
sciencetable.ca/ontario-dashboard/. The decreasing total case count be- or after travel, as well as quarantine periods following travel,
tween 21 January, 2021 and early March, 2021 masked an underlying increase which can slow the spread of VOCs into regions where they
in the case count due to VOCs.
have yet to establish. Enhanced measures to reduce transmis-
sion, including travel restrictions and lockdown measures that
What are the implications of ongoing evolution for prohibit various activities, have subsequently reversed the rise
control strategies during the pandemic? in case numbers in many regions (Figure 4).
The emergence of VOCs and the ongoing evolution of SARS- Detection of immune-escape VOCs will pose specific chal-
CoV-2 have potentially important implications for how best to lenges. Preliminary data suggest that vaccinated individuals
control the pandemic. We first consider implications for detect- who become infected are less likely to feel symptoms, making
ing VOCs and then turn to different potential control measures breakthrough infections less likely to be detected, although se-
that can help slow their emergence and spread. vere infections continue to occur60. To detect and break trans-
Detection and surveillance of VOCs mission chains involving immune-escape variants, backwards
Identifying VOCs as they emerge allows for the earlier implemen- contact tracing61, as well as regular testing of asymptomatic in-
tation of control measures. Whole-genome sequencing followed dividuals, may be essential tools, followed by enhanced control
by bioinformatic analysis can identify lineages that are increasing measures to prevent onward transmissions. Distinguishing
in frequency or that can infect previously immune individuals. escape mutations from chance breakthrough cases will also be
Coupled with structural modelling of potential effects on viral a challenge. Flagging transmission chains that occur between
function and the ability to evade antibodies, variants of interest vaccinated individuals, as well as whole-genome sequencing
(VOIs) can be identified (Box 1). Systematic surveillance and of breakthrough cases, will help detect new immune-escape
contact tracing are then needed to track a VOI to determine VOCs earlier. Hampering this effort, however, is the fact that
the mechanisms by which it is increasing in frequency, as well vaccination status, reason for sequencing, contact history, and
as whether it causes altered disease characteristics, such as case information (for example, severity) are typically not shared
increased transmission rates, longer periods of infectiousness, alongside genomic data, limiting the power to detect escape var-
greater virulence, or the capacity to infect previously immune in- iants and prevent their global spread.
dividuals. Assessing changes to the infectious period of a variant Finally, internationally coordinated genomic surveillance will
is, for example, critical for setting effective quarantine intervals. be important for identifying which variants to include in future
Once a variant of interest is elevated to the status of VOC booster vaccines, with an eye to prioritizing immune-escape
(Box 1), reducing its spread also requires effective genetic sur- VOCs and highly divergent SARS-CoV-2 lineages.
veillance. Importantly, the higher the fraction of cases that are Control measures to slow the evolution of new variants
genetically typed, the easier it will be to contain a VOC59, Although we cannot generally reduce the rate of mutation, we
because the variant is more likely to be detected early when may be able to reduce the risk that multiple mutations arise. In
there are fewer VOC infections. For this reason, rapid PCR- particular, we can provide a halo of protection around immuno-
based methods that detect a VOC (for example the SGTF that al- compromised individuals by prioritizing vaccination of their
lowed detection of B.1.1.7 in the UK1) are an important tool, households and caregivers, reducing their risk of exposure. If
alongside whole genome sequencing, to detect VOCs early long-lasting infections do occur, enhanced surveillance and
enough to isolate cases and prevent establishment. contact tracing should be considered to prevent onward

Current Biology 31, R918–R929, July 26, 2021 R925

 ll
Review

British Columbia Alberta Figure 4. The spike in case numbers in the

3000 3000 spring of 2021 was predicted by models of
VOC dynamics.
Daily report

Daily report
2000 2000 Case numbers in Canadian provinces (black cir-
cles; data up to 8 March, 2021) were fit using a
1000 1000 dynamic modeling approach, either ignoring VOC
(purple) or allowing the spread of VOC with a
0 0
transmission advantage of 50% (grey). In each
Oct Nov Dec Jan Feb Mar Apr May Oct Nov Dec Jan Feb Mar Apr May
panel, the VOC is introduced a week before the
date of the first publicly reported case in each
Saskatchewan Manitoba province (vertical dashed line) with initial numbers
1000 1000 set to match the observed VOC numbers in early
March. Subsequent case numbers, which were
Daily report

Daily report
750 750
not used in the model fits, are shown as hollow
500 500
circles. The spike in cases led to various emer-
250 250 gency restrictions (vertical solid lines), which
0 0
subsequently brought cases down over the next
Oct Nov Dec Jan Feb Mar Apr May Oct Nov Dec Jan Feb Mar Apr May couple of weeks. Poor model predictions in a
couple of provinces are likely due to migration
among provinces and/or a low sampling rate for
Ontario Quebec VOC (for example, genomics now indicates that
5000 5000
B.1.1.7 was in Manitoba at least 19 days earlier
4000 4000
Daily report

Daily report

than the first reported case). (Based on model fits

3000 3000 using the Public Health Agency of Canada/
2000 2000 McMaster model76.)
1000 1000
0 0
Oct Nov Dec Jan Feb Mar Apr May Oct Nov Dec Jan Feb Mar Apr May

Current Biology

transmissions. Heightened measures to reduce onward trans- Over the near future, epidemic control by nonpharmaceutical
mission may also be warranted for lineages associated with interventions will be replaced by control via vaccination,
higher-than-expected mutation rates (for example, mutants increasing selection in favour of immune escape. Reducing
that alter proofreading capacity19). case numbers during this time will continue to be important.
More broadly, we can reduce the rate of emergence of new In particular, reducing exposure of immunized individuals to
VOCs and slow the spread of existing ones by reducing overall active cases should slow the evolution of immune-escape vari-
case numbers through vaccination at a global scale and by main- ants (Box 2) by reducing their selective advantage. Vaccinating
taining or enhancing the nonpharmaceutical interventions that whole communities or workplaces at once, coupling vaccination
have contributed to controlling the pandemic (case detection campaigns with temporary stay-at-home orders to reduce com-
and isolation, contact tracing and quarantine, masking and per- munity transmission, and communicating the risks of exposure
sonal distancing, and improved ventilation)62,63. Having low case to vaccinated individuals are all potential actions that could
numbers makes it easier to test and genotype a high fraction of reduce the evolution of immune escape variants, by reducing
cases and increases the efficacy of contact tracing measures to contact rates between infectious and immunized individuals.
stop onward transmission64. Furthermore, mathematical models Having effective public health measures also reduces the likeli-
predict that measures that reduce contact rates with susceptible hood that recombinant variants arise by driving down the chance
individuals will not only slow the spread of SARS-CoV-2 overall of coinfection.
but will also reduce the relative advantage of variants that Dosing strategies
have a transmission advantage37. Thus, the measures taken to Data suggest that individuals who have received only a single
reduce contacts and limit the number of COVID-19 cases may dose of the Pfizer BNT162b2 vaccine or who have natural immu-
have the added benefit of slowing the rate at which VOCs with nity elicit a weaker antibody response to VOCs than those who
a transmission advantage overtake the wild type. This predicted have received two vaccine doses, although T-cell responses
pattern, with selection weakening as stringency measures are were similar for variants and non-variants52. Recent reports
increased, appears to be borne out in data for B.1.1.7 from En- also indicate lower real-world effectiveness of vaccines against
gland (Figure 5) and British Columbia, Canada (Figure S2). VOCs after one dose (Pfizer: 29.5% [95% CI: 22.9–35.5] for
Importantly, the existence of VOCs means that scientists and B.1.351 versus 16.9% [95% CI: 10.4–23.0] for B.1.1.755; Pfizer:
public health officials cannot group all cases together when 33.2% [95% CI: 8.3–51.4] for B.1.617.2 versus 49.2% [95% CI:
inferring the dynamics of SARS-CoV-2 and predicting future 42.6–55.0] for B.1.1.756; AstraZeneca 32.9% [95% CI: 19.3–
case loads. Each VOC will have its own reproductive number 44.3] for B.1.617.2 versus 51.4% [95% CI: 47.3–55.2] for
(or growth rate) and, consequently, different requisite measures B.1.1.756). These studies, though preliminary, also find that effi-
for effective control. Additional restrictions (for example, further cacy rises for all variants following the second dose. At this time,
closures of schools, businesses, and leisure venues65) can however, it is unknown how much of the rise in efficacy against
slow the spread and reduce the impact of VOCs, despite their VOCs after the second dose is due to the booster shot itself
higher transmission advantage29. By flattening both non-VOC and how much is due to the maturation of the immune response
and VOC COVID-19 curves, more people can be vaccinated given more time since the first dose. At face value, these data
and protected before peak case numbers are reached. indicate that completing the recommended number of doses

R926 Current Biology 31, R918–R929, July 26, 2021

 ll
Review
Selective advantage of B.1.1.7

London von der Leyen (President of the European Commission), ‘‘none of

1.2 Northeast us will be safe until everyone is safe.’’
Northwest
1.0
East Midlands
0.8 West Midlands Conclusions
0.6 East of England As COVID-19 transitions from a pandemic to an endemic dis-
Southeast ease, VOCs present new global challenges to health by virtue
0.4 Southwest of increased transmissibility and virulence and evasion of natural
0.2 Yorkshire
and vaccine-induced immunity. In this article we have explored
65 70 75 80 85 90 the selective forces that shape how VOCs emerge and become
Stringency index established. We also identify possible steps that we can take to
Current Biology
limit their emergence and, when they do arise, their impact. Mov-
ing forward, we must also consider how SARS-CoV-2 transmits
Figure 5. Selective advantage of B.1.1.7 declines with increasing
social restrictions in the UK. to and amongst other animal species, placing both them and us
The same data as in Figure 2 but now the selective advantage of B.1.1.7, as at further risk. It will therefore be important to adopt a multidisci-
measured by the weekly change in log(frequency SGTF/frequency of non- plinary One Health approach70 for future pandemic management
SGTF), is plotted against the stringency index of restrictions during that week
that accounts for the interrelated nature of human, animal, and
in the UK (taken from http://www.bsg.ox.ac.uk/research/research-projects/
covid-19-government-response-tracker). All data points corresponding to a ecosystem health.
frequency of less than 10% were excluded to ensure SGTF data predominantly
reflect the presence of B.1.1.7.
SUPPLEMENTAL INFORMATION

Supplemental information can be found online at https://doi.org/10.1016/j.

for two-dose vaccines reduces the risk of transmission of partial- cub.2021.06.049.
escape variants.
Over the short-term, however, the reduction in case ACKNOWLEDGEMENTS
numbers made possible by administering one dose to more
people is likely to both save lives and reduce the chance The authors would like to thank Art Poon for providing the deviations from the
molecular clock used in Figure S1B, Hongru Wang and Rasmus Nielsen for
that more resistance mutants arise in the first place66 (Box providing estimates of pS to estimate the mutation rate, and two referees for
2). As long as one dose provides sufficient immunity, reducing helpful suggestions. Funding was provided by the Natural Sciences and Engi-
the burden of disease by vaccinating as many people as neering Research Council of Canada to S.P.O. (RGPIN-2016-03711) and T.D.
(RGPIN-2013-217636), by Canada 150 Research Chair support for C.C., by
possible with one dose and delaying the second dose better the Public Health Agency of Canada to M.L., S.M., G.V.D., and N.H.O., by
prevents the evolution of immune escape variants than more the CIHR 2019 Novel Coronavirus (COVID-19) rapid research program to
slowly vaccinating the population with two doses at a short in- J.W., and by the M.G. DeGroote Institute for Infectious Disease Research at
McMaster University to J.D. and D.J.D.E.
ter-dose interval66,67.
Another important factor to consider when determining the
dosing regime comes from recent studies showing that delaying DECLARATION OF INTERESTS

the second dose enhances immunity, with elevated antibody

The authors declare no competing interests.
levels following a longer dosing interval (12 weeks) than a shorter
interval (<6 weeks) for both the Pfizer68 and the AstraZeneca69
vaccines. Delayed second doses might thus increase the selec- REFERENCES
tive advantage of partial escape mutations over the short term
but reduce it over the longer term given the more robust immune 1. Public Health England (2020). Investigation of novel SARS-CoV-2 variant:
responses observed with a delayed booster68,69. Variant of Concern 202012/01. Version 1, release date 21/12/2020, https://
www.gov.uk/government/publications/investigation-of-novel-sars-cov-2-
Reopening variant-variant-of-concern-20201201.
As we move to reopen societies, rebuild economies, and resume
2. Volz, E., Mishra, S., Chand, M., Barrett, J.C., Johnson, R., Geidelberg, L.,
social activities, contact rates will increase. This is expected to Hinsley, W.R., Laydon, D.J., Dabrera, G., O’Toole, A., et al. (2021). Assess-
increase the selective advantage of VOCs that are more trans- ing transmissibility of SARS-CoV-2 lineage B.1.1.7 in England. Nature 593,
missible and/or more able to infect immunized individuals 266–269.
(Figure 5). Increased activity will also make contact tracing 3. van Dorp, L., Richard, D., Tan, C.C.S., Shaw, L.P., Acman, M., and Bal-
more difficult, as will the potential rise in asymptomatic cases loux, F. (2020). No evidence for increased transmissibility from recurrent
among vaccinated individuals60. Coupling reopening with mutations in SARS-CoV-2. Nat. Commun. 11, 5986–5988.

expanded use of rapid testing and environmental sampling 4. Tegally, H., Wilkinson, E., Giovanetti, M., Iranzadeh, A., Fonseca, V.,
would allow cases, and particularly asymptomatic breakthrough Giandhari, J., Doolabh, D., Pillay, S., San, E.J., Msomi, N., et al. (2021).
Detection of a SARS-CoV-2 variant of concern in South Africa. Nature
cases, to be identified and controlled. Border measures, with 592, 438–443.
testing before, during, and after travel, along with quarantine
measures, will continue to be important while cases remain 5. Faria, N.R., Mellan, T.A., Whittaker, C., Claro, I.M., Candido, D.D.S., Mis-
hra, S., Crispim, M.A.E., Sales, F.C.S., Hawryluk, I., McCrone, J.T., et al.
high globally to slow the spread of immune-escape variants (2021). Genomics and epidemiology of the P.1 SARS-CoV-2 lineage in
and other VOCs that emerge. Ultimately, though, variants will Manaus, Brazil. Science 372, 815–821.
continue to emerge until cases are globally controlled through, 6. Holmes, E.C. (2009). The Evolution and Emergence of RNA Viruses (Ox-
for example, equitable access to vaccines. In the words of Ursula ford: Oxford University Press).

Current Biology 31, R918–R929, July 26, 2021 R927

 ll
Review
7. Vasilarou, M., Alachiotis, N., Garefalaki, J., Beloukas, A., and Pavlidis, P. 25. Endo, A., Abbott, S., Kucharski, A.J., and Funk, S. (2020). Estimating the
(2021). Population genomics insights into the first wave of COVID-19. overdispersion in COVID-19 transmission using outbreak sizes outside
Life 11, 129. China. Wellcome Open Res. 5, 67.

8. Hadfield, J., Megill, C., Bell, S.M., Huddleston, J., Potter, B., Callender, C., 26. Althouse, B.M., Wenger, E.A., Miller, J.C., Scarpino, S.V., Allard, A.,
Sagulenko, P., Bedford, T., and Neher, R.A. (2018). Nextstrain: real-time Hebert-Dufresne, L., and Hu, H. (2020). Superspreading events in the
tracking of pathogen evolution. Bioinformatics 34, 4121–4123. transmission dynamics of SARS-CoV-2: Opportunities for interventions
and control. PLoS Biol. 18, e3000897.
9. Wang, H., Pipes, L., and Nielsen, R. (2020). Synonymous mutations and
the molecular evolution of SARS-CoV-2 origins. Virus Evol. 7, veaa098. 27. Lloyd-Smith, J.O., Schreiber, S.J., Kopp, P.E., and Getz, W.M. (2005).
Superspreading and the effect of individual variation on disease emer-
10. Rambaut, A., Loman, N., Pybus, O., Barclay, W., Barrett, J., Carabelli, A., gence. Nature 438, 355–359.
Connor, T., Peacock, T., Robertson, D.L., and Volz, E., on behalf of the
COVID-19 Genomics Consortium UK (2020). Preliminary genomic charac- 28. Mulberry, N., Tupper, P., Kirwin, E., McCabe, C., and Colijn, C. (2021).
terisation of an emergent SARS-CoV-2 lineage in the UK defined by a Vaccine rollout strategies: The case for vaccinating essential workers
novel set of spike mutations. virological.org, https://virological.org/t/ early. medRxiv. https://doi.org/10.1101/2021.02.23.21252309.
preliminary-genomic-characterisation-of-an-emergent-sars-cov-2-lineage-
in-the-uk-defined-by-a-novel-set-of-spike-mutations/563. 29. Davies, N.G., Abbott, S., Barnard, R.C., Jarvis, C.I., Kucharski, A.J., Mun-
day, J.D., Pearson, C.A.B., Russell, T.W., Tully, D.C., Washburne, A.D.,
11. Choi, B., Choudhary, M.C., Regan, J., Sparks, J.A., Padera, R.F., Qiu, X., et al. (2021). Estimated transmissibility and impact of SARS-CoV-2 lineage
Solomon, I.H., Kuo, H.-H., Boucau, J., Bowman, K., et al. (2020). Persis- B.1.1.7 in England. Science 372, eabg3055.
tence and evolution of SARS-CoV-2 in an immunocompromised host.
New Engl. J. Med. 383, 2291–2293. 30. O’Toole, Á., and Hill, V. (2021). Global report investigating novel coronavi-
rus haplotypes: B.1.1.7. https://cov-lineages.org/global_report_B.1.1.7.
12. Avanzato, V.A., Matson, M.J., Seifert, S.N., Pryce, R., Williamson, B.N., html. (Accessed 26 March 2021.)
Anzick, S.L., Barbian, K., Judson, S.D., Fischer, E.R., Martens, C., et al.
(2020). Case study: prolonged infectious SARS-CoV-2 shedding from an 31. Byambasuren, O., Cardona, M., Bell, K., Clark, J., McLaws, M.-L., and
asymptomatic immunocompromised individual with cancer. Cell 183, Glasziou, P. (2020). Estimating the extent of asymptomatic COVID-19
1901–1912. and its potential for community transmission: systematic review and
meta-analysis. JAMMI 5, 223–234.
13. Kemp, S.A., Collier, D.A., Datir, R.P., Ferreira, I.A.T.M., Gayed, S., Jahun,
A., Hosmillo, M., Rees-Spear, C., Mlcochova, P., Lumb, I.U., et al. (2021). 32. Tindale, L.C., Stockdale, J.E., Coombe, M., Garlock, E.S., Lau, W.Y.V., Sar-
SARS-CoV-2 evolution during treatment of chronic infection. Nature 592, aswat, M., Zhang, L., Chen, D., Wallinga, J., and Colijn, C. (2020). Evidence
277–282. for transmission of COVID-19 prior to symptom onset. eLife 9, e57149.

14. McCarthy, K.R., Rennick, L.J., Nambulli, S., Robinson-McCarthy, L.R., 33. Meyerowitz-Katz, G., and Merone, L. (2020). A systematic review and
Bain, W.G., Haidar, G., and Duprex, W.P. (2021). Recurrent deletions in meta-analysis of published research data on COVID-19 infection fatality
the SARS-CoV-2 spike glycoprotein drive antibody escape. Science rates. Int. J. Infect. Dis. 101, 138–148.
371, 1139–1142.
34. Bull, J.J. (1994). Virulence. Evolution 48, 1423–1437.
15. McCormick, K.D., Jacobs, J.L., and Mellors, J.W. (2021). The emerging
plasticity of SARS-CoV-2. Science 371, 1306–1308. 35. Frank, S.A. (1996). Models of parasite virulence. Q. Rev. Biol. 71, 37–78.

16. Luria, S., and Delbruck, M. (1943). Mutations of bacteria from virus sensi- 36. Day, T. (2003). Virulence evolution and the timing of disease life-history
tivity to virus resistance. Genetics 28, 491–511. events. Trends Ecol. Evol. 18, 113–118.

17. Lythgoe, K.A., Hall, M., Ferretti, L., de Cesare, M., MacIntyre-Cockett, G., 37. Day, T., Gandon, S., Lion, S., and Otto, S.P. (2020). On the evolutionary
Trebes, A., Andersson, M., Otecko, N., Wise, E.L., Moore, N., et al. (2021). epidemiology of SARS-CoV-2. Curr. Biol. 30, R849–R857.
SARS-CoV-2 within-host diversity and transmission. Science 372,
eabg0821. 38. Lenski, R.E., and May, R.M. (1994). The evolution of virulence in parasites
and pathogens: reconciliation between two competing hypotheses.
18. Braun, K., Moreno, G.K., Wagner, C., Accola, M.A., Rehrauer, W.M., J. Theor. Biol. 169, 253–265.
Baker, D.A., Koelle, K., O’Connor, D.H., Bedford, T., Friedrich, T.C., and
Moncla, L.H. (2021). Limited within-host diversity and tight transmission 39. Day, T., and Gandon, S. (2007). Applying population-genetic models in
bottlenecks limit SARS-CoV-2 evolution in acutely infected individuals. theoretical evolutionary epidemiology. Ecol. Lett. 10, 876–888.
bioRxiv, https://doi.org/10.1101/2021.04.30.440988.
40. Bull, J.J., and Ebert, D. (2008). Invasion thresholds and the evolution of
19. Takada, K., Ueda, M.T., Watanabe, T., and Nakagawa, S. (2020). Genomic nonequilibrium virulence. Evol. Appl. 1, 172–182.
diversity of SARS-CoV-2 can be accelerated by a mutation in the nsp14
gene. bioRxiv, https://doi.org/10.1101/2020.12.23.424231. 41. Davies, N.G., Jarvis, C.I., CMMID COVID-19 Working Group, Edmunds,
W.J., Jewell, N.P., Diaz-Ordaz, K., and Keogh, R.H. (2021). Increased mor-
20. Richard, D., Owen, C.J., van Dorp, L., and Balloux, F. (2020). No detect- tality in community-tested cases of SARS-CoV-2 lineage B.1.1.7. Nature
able signal for ongoing genetic recombination in SARS-CoV-2. bioRxiv, 593, 270–274.
https://doi.org/10.1101/2020.12.15.422866.
42. Challen, R., Brooks-Pollock, E., Read, J.M., Dyson, L., Tsaneva-Atanasova,
21. Ignatieva, A., Hein, J., and Jenkins, P.A. (2021). Ongoing recombination K., and Danon, L. (2021). Risk of mortality in patients infected with SARS-
in SARS-CoV-2 revealed through genealogical reconstruction. bioRxiv, CoV-2 variant of concern 202012/1: matched cohort study. BMJ 372, n579.
https://doi.org/10.1101/2021.01.21.427579.
43. Teruel, N., Mailhot, O., and Najmanovich, R. (2020). Modelling conforma-
22. VanInsberghe, D., Neish, A.S., Lowen, A.C., and Koelle, K. (2021). Recom- tional state dynamics and its role on infection for SARS-CoV-2 spike pro-
binant SARS-CoV-2 genomes are currently circulating at low levels. bio- tein variants. bioRxiv, https://doi.org/10.1021/2020.12.16.423118.
Rxiv, https://doi.org/10.1101/2020.08.05.238386.
44. Calistri, P., Amato, L., Puglia, I., Cito, F., Di Giuseppe, A., Danzetta, M.L.,
23. Jackson, B., Rambaut, A., Pybus, O.G., Robertson, D.L., Connor, T., and Morelli, D., Di Domenico, M., Caporale, M., Scialabba, S., et al. (2021).
Loman, N.J., on behalf of the COVID-19 Genomics UK Consortium (2021). Infection sustained by lineage B.1.1.7 of SARS-CoV-2 is characterised
Recombinant SARS-CoV-2 genomes involving lineage B.1.1.7 in the UK. by longer persistence and higher viral RNA loads in nasopharyngeal
virological.org, https://virological.org/t/recombinant-sars-cov-2-genomes- swabs. Int. J. Infect. Dis. 105, 753–755.
involving-lineage-b-1-1-7-in-the-uk/658.
45. Damas, J., Hughes, G.M., Keough, K.C., Painter, C.A., Persky, N.S.,
24. Woolhouse, M.E., Dye, C., Etard, J.-F., Smith, T., Charlwood, J.D., Gar- Corbo, M., Hiller, M., Koepfli, K.-P., Pfenning, A.R., Zhao, H., et al.
nett, G.P., Hagan, P., Hii, J.L., Ndhlovu, P.D., Quinnell, R.J., et al. (1997). (2020). Broad host range of SARS-CoV-2 predicted by comparative and
Heterogeneities in the transmission of infectious agents: implications for structural analysis of ACE2 in vertebrates. Proc. Natl. Acad. Sci. USA
the design of control programs. Proc. Natl. Acad. Sci. USA 94, 338–342. 117, 22311–22322.

R928 Current Biology 31, R918–R929, July 26, 2021

 ll
Review
46. Brault, A.C., Huang, C.Y., Langevin, S.A., Kinney, R.M., Bowen, R.A., Ra- bert-Dufresne, L., Mones, E., Lehmann, S., and Ahn, Y.-Y.
61. Kojaku, S., He
mey, W.N., Panella, N.A., Holmes, E.C., Powers, A.M., and Miller, B.R. The effectiveness of backward contact tracing in networks. Nat. Physics
(2007). A single positively selected West Nile viral mutation confers 17, 652–658.
increased virogenesis in American crows. Nat. Genet. 39, 1162–1166.
62. Ludwig, A., Berthiaume, P., Orpana, H., Nadeau, C., Diasparra, M.,
47. Larsen, H.D., Fonager, J., Lomholt, F.K., Dalby, T., Benedetti, G., Kristen- Barnes, J., Hennessy, D., Otten, A., and Ogden, N. (2020). Assessing
sen, B., Urth, T.R., Rasmussen, M., Lassaunière, R., Rasmussen, T.B., the impact of varying levels of case detection and contact tracing on
et al. (2021). Preliminary report of an outbreak of SARS-CoV-2 in mink COVID-19 transmission in Canada during lifting of restrictive closures
and mink farmers associated with community spread, Denmark, June to using a dynamic compartmental model. Can. Commun. Dis. Rep. 46,
November 2020. Euro. Surveill. 26, 2100009. 409–421.

48. Watson, O.J., Abdelmagid, N., Ahmed, A., Elhameed, A.E.A.A., Whittaker, 63. Tang, B., Scarabel, F., Bragazzi, N.L., McCarthy, Z., Glazer, M., Xiao, Y.,
C., Brazeau, N., Hamlet, A., Walker, P., Hay, J., Ghani, A., et al. (2020). Heffernan, J.M., Asgary, A., Ogden, N.H., and Wu, J. (2020). De-escalation
Characterising COVID-19 epidemic dynamics and mortality underascer- by reversing the escalation with a stronger synergistic package of contact
tainment in Khartoum, Sudan (London: Imperial College London), tracing, quarantine, isolation and personal protection: feasibility of pre-
https://doi.org/10.25561/84283. venting a COVID-19 rebound in Ontario, Canada, as a case study. Biology
9, 100.
49. Kennedy, D.A., and Read, A.F. (2017). Why does drug resistance readily
evolve but vaccine resistance does not? Proc. R. Soc. B 284, 20162562. 64. Tupper, P., Otto, S., and Colijn, C. (2020). Fundamental limitations of con-
tact tracing for COVID-19. medRxiv, https://doi.org/10.1101/2020.12.15.
50. Le Bert, N., Tan, A.T., Kunasegaran, K., Tham, C.Y., Hafezi, M., Chia, A., 20248299.
Chng, M.H.Y., Lin, M., Tan, N., Linster, M., et al. (2020). SARS-CoV-2-spe-
65. Ng, V., Fazil, A., Waddell, L., Turgeon, P., Otten, A., and Ogden, N.H.
cific T cell immunity in cases of COVID-19 and SARS, and uninfected con-
(2021). Modelling the impact of shutdowns on resurging SARS-CoV-2
trols. Nature 584, 457–462.
transmission in Canada. R. Soc. Open Sci. 8, 210233.
51. Grifoni, A., Weiskopf, D., Ramirez, S.I., Mateus, J., Dan, J.M., Moder- 66. Cobey, S., Larremore, D.B., Grad, Y.H., and Lipsitch, M. (2021). Concerns
bacher, C.R., Rawlings, S.A., Sutherland, A., Premkumar, L., Jadi, R.S., about SARS-CoV-2 evolution should not hold back efforts to expand
et al. (2020). Targets of T cell responses to SARS-CoV-2 coronavirus in hu- vaccination. Nat. Rev. Immunol. 21, 330–335.
mans with COVID-19 disease and unexposed individuals. Cell 181, 1489–
1501. 67. Saad-Roy, C.M., Morris, S.E., Metcalf, C.J.E., Mina, M.J., Baker, R.E., Far-
rar, J., Holmes, E.C., Pybus, O.G., Graham, A.L., Levin, S.A., et al. (2021).
52. Skelly, D.T., Harding, A.C., Gilbert-Jaramillo, J., Knight, M.L., Longet, S., Epidemiological and evolutionary considerations of SARS-CoV-2 vaccine
Brown, A., Adele, S., Adland, E., Brown, H., et al. (2021). Vaccine-induced dosing regimes. Science 372, 363–370.
immunity provides more robust heterotypic immunity than natural infection
to emerging SARS-CoV-2 variants of concern. Research Square, https:// 68. Parry, H., Bruton, R., Stephens, C., Brown, K., Amirthalingam, G., Hallis,
doi.org/10.21203/rs.3.rs-226857/v1. B., Otter, A., Zuo, J., and Moss, P. (2021). Extended interval BNT162b2
vaccination enhances peak antibody generation in older people. medRxiv,
53. Tarke, A., Sidney, J., Methot, N., Zhang, Y., Dan, J.M., Goodwin, B., https://doi.org/10.1101/2021.05.15.21257017.
Rubiro, P., Sutherland, A., da Silva Antunes, R., Frazier, A., et al. (2021).
Negligible impact of SARS-CoV-2 variants on CD4+ and CD8+ T cell reac- 69. Voysey, M., Clemens, S.A.C., Madhi, S.A., Weckx, L.Y., Folegatti, P.M.,
tivity in COVID-19 exposed donors and vaccinees. bioRxiv, https://doi. Aley, P.K., Angus, B., Baillie, V.L., Barnabas, S.L., Bhorat, Q.E., et al.
org/10.1101/2021.02.27.433180. (2021). Single-dose administration and the influence of the timing of the
booster dose on immunogenicity and efficacy of ChAdOx1 nCoV-19
54. Wu, K., Werner, A.P., Koch, M., Choi, A., Narayanan, E., Stewart-Jones, (AZD1222) vaccine: a pooled analysis of four randomised trials. Lancet
G.B., Colpitts, T., Bennett, H., Boyoglu-Barnum, S., Shi, W., et al. 397, 881–891.
(2021). Serum neutralizing activity elicited by mRNA-1273 vaccine. New
Engl. J. Med. 384, 1468–1470. 70. Ogden, N.H., Wilson, J.R., Richardson, D.M., Hui, C., Davies, S.J., Kum-
schick, S., Le Roux, J.J., Measey, J., Saul, W.C., and Pulliam, J.R.C.
55. Abu-Raddad, L.J., Chemaitelly, H., and Butt, A.A., on behalf of the Na- (2019). Emerging infectious diseases and biological invasions: a call for
tional Study Group for COVID-19 Vaccination (2021). Effectiveness of a One Health collaboration in science and management. R. Soc. Open
the BNT162b2 COVID-19 vaccine against the B.1.1.7 and B.1.351 vari- Sci. 6, 181577.
ants. New Engl. J. Med. 385, 187–189.
71. World Health Organization (2021). COVID-19 Weekly Epidemiological Up-
56. Lopez Bernal, J., Andrews, N., Gower, C., Gallagher, E., Simmons, R., date (25 February 2021), https://apps.who.int/iris/handle/10665/339859.
Thelwall, S., Stowe, J., Tessier, E., Groves, N., Dabrera, G., et al.
72. Pearson, C., Russell, T.W., Davies, N.G., Kucharski, A.J., CMMID COVID-19
(2021). Effectiveness of COVID-19 vaccines against the B.1.617.2
working group, Edmunds, W.J., and Eggo, R.M. (2021). Estimates of severity
variant. medRxiv, https://doi.org/10.1101/2021.05.22.21257658.
and transmissibility of novel SARS-CoV-2 variant 501Y.V2 in South Africa
(Centre for Mathematical Modelling of Infectious Diseases Repository),
57. Eguia, R.T., Crawford, K.H.D., Stevens-Ayers, T., Kelnhofer-Millevolte, L.,
https://cmmid.github.io/topics/covid19/sa-novel-variant.html.
Greninger, A.L., Englund, J.A., Boeckh, M.J., and Bloom, J.D. (2021). A hu-
man coronavirus evolves antigenically to escape antibody immunity. PLoS 73. Deng, X., Garcia-Knight, M.A., Khalid, M.M., Servellita, V., Wang, C., Mor-
Pathog. 17, e1009453. ris, M.K., Sotomayor-González, A., Glasner, D.R., Reyes, K.R., Gliwa, A.S.,
et al. (2021). Transmission, infectivity, and neutralization of a spike L452R
58. Wu, K., Choi, A., Koch, M., Ma, L., Hill, A., Nunna, N., Huang, W., Oes- SARS-CoV-2 variant. Cell 184, 3426–3437.
treicher, J., Colpitts, T., Bennett, H., et al. (2021). Preliminary analysis of
safety and immunogenicity of a SARS-CoV-2 variant vaccine booster. 74. Public Health England (2021). SARS-CoV-2 variants of concern and vari-
medRxiv, https://doi.org/10.1101/2021.05.05.21256716. ants under investigation in England: Technical briefing 15 (London: Public
Health England), https://assets.publishing.service.gov.uk/government/
59. Otto, S.P., Phillips, B., and Colijn, C. (2021). Are there COVID variants uploads/system/uploads/attachment_data/file/993198/Variants_of_
amongst us. BC COVID-19 Modelling Group blog, https://bccovid- Concern_VOC_Technical_Briefing.pdf.
19group.ca/post/covid-variants-amongst-us/.
75. Alexander, H.K., and Day, T. (2010). Risk factors for the evolutionary emer-
60. Teran, R.A., Walblay, K.A., Shane, E.L., Xydis, S., Gretsch, S., Gagner, A., gence of pathogens. J. R. Soc. Interface 7, 1455–1474.
Samala, U., Choi, H., Zelinski, C., and Black, S.R. (2021). Postvaccination
SARS-CoV-2 infections among skilled nursing facility residents and staff 76. Bolker, B., Li, M., Dushoff, J., Earn, D., So, M., Levine, Z., Gharouni, A.,
members – Chicago, Illinois, December 2020-March 2021. Morb. Mortal. Papst, I., and Walker, S. (2021). McMasterPandemic (Initial release for
Wkly. Rep. 70, 632–638. Zenodo), https://doi.org/10.5281/zenodo.4946092.

Current Biology 31, R918–R929, July 26, 2021 R929