Research and analysis

SARS-CoV-2 variant surveillance and assessment: technical briefing 55

Updated 22 September 2023

Applies to England

Summary

This report has been published to share the detailed variant surveillance analyses which contribute to the variant risk assessments and designation of new SARS-CoV-2 variants. This specialist technical briefing contains early data and analysis on emerging variants and findings have a high level of uncertainty. Unless stated otherwise, this technical briefing uses a data cut-off of 18 September 2023 to allow time for analyses.

During periods when technical briefings are being published by the UK Health Security Agency (UKHSA), the routine variant prevalence and growth rate reports will be included in the technical briefing.

Variant risk assessment BA.2.86 (Variant Technical Group of 19 September 2023)

1. Based on pseudovirus and live virus data from multiple UK and international centres, BA.2.86 is distant antigenically from previous Omicron viruses and is likely to have similar antibody escape to XBB.1.5 in the UK population context (moderate confidence).
2. Based on data from 2 laboratories, BA.2.86 appears to have slightly higher human ACE2 binding affinity than those XBB variants tested (low confidence). ACE2 binding affinity may be a factor in transmissibility.
3. Based on preliminary in vitro data and a small amount of animal data, there are no signals of concern regarding change in phenotype compared to other Omicron lineages (low confidence). However, no epidemiological assessment of clinical severity in patients is possible and the correlation between these laboratory studies and clinical severity is only partially understood.
4. BA.2.86 continues to transmit in the UK as evidenced by continued sporadic cases with no travel history in almost all regions (moderate confidence). It is circulating within a mix of variants which are antigenically distinct but which have similar or slightly greater level of escape from population immunity. Given this mixture and the characteristics of BA.2.86, it is plausible that the incidence of this variant may increase, but using the current low number of sequenced cases we cannot predict whether it will outcompete other circulating variants.

Real-world vaccine effectiveness studies on BA.2.86 and circulating variants will be published when sufficient data is available. Routine weekly updates on this emergent variant will now pause and reporting will return to standard variant prevalence updates.

Data summary

BA.2.86

As of 18 September 2023, there are 48 BA.2.86 sequenced cases in England, an increase of 11 cases in the last week.

Cases were found in:

• the East of England
• London
• the North East
• the North West
• the South East
• the South West
• the West Midlands

Of 48 total cases in England, 10 were hospitalised (6 have unknown hospitalisation status), and no deaths due to COVID-19 have been reported. Most sequenced cases in England are people tested in hospital and the hospitalisation rate cannot be estimated from this data. These cases include 30 sequenced as part of an investigation into an outbreak in a care home which reported a high attack rate. Of the remaining 18 cases, all were sporadic apart from which 2 may be epidemiologically linked.

As of 18 September 2023, 6 BA.2.86 cases have been reported by Public Health Scotland. There are no cases reported by Wales or Northern Ireland.

New cases continue to be identified globally. As of 18 September 2023, there are 139 sequences from human cases identified as BA.2.86 available in GISAID from 15 countries. The earliest collection date is 24 July 2023, and the most recent collection date is 9 September 2023. Of the 139 sequences, 46 are from the UK. The remaining UK sequence data will be uploaded to GISAID in due course where it meets the relevant quality control thresholds. There are also multiple reports of BA.2.86 detected in wastewater in other countries, though limited sequence data is available.

Neutralisation studies

Data from 4 UK labs shows neutralisation titres against BA.2.86 are slightly higher than titres against XBB.1.5 in 3 of 4 UK serum group data, and lower in one serum group.

ACE2 binding affinity data

Two international independent labs (Cao and Veesler) have shared preliminary ACE2 binding affinity measurements of BA.2.86. Published data is not yet available for these studies. BA.2.86 spike shows high human ACE2 binding affinity, in line with or above previous Omicron sublineages. Weak ACE2 binding affinity has previously been associated with more poorly transmissible variants (for example XBB); however, it is unclear if stronger ACE2 binding affinity is associated with greater transmissibility beyond a certain point.

Higher ACE2 affinity may allow the virus to tolerate further mutations while still maintaining the ability to transmit. It is important to continue to monitor BA.2.86 for further potential antigenic changes going forward.

SARS-CoV-2 variant prevalence in England (UKHSA-designated variants)

As of 14 September 2023, classification for sequences with a specimen date between 28 August 2023 and 3 September 2023 inclusive is as follows:

• 37.0% V-23JUL-01 (EG.5.1)
• 30.0% V-23APR-01 (XBB.1.16)
• 24.5% V-22OCT-02 (XBB)
• 5.5% V-23JAN-01 (XBB.1.5)
• 1.8% V-22DEC-01 (CH.1.1)

The remaining sequences (1.2%) were from lineages that have not been designated as a variant by UKHSA or are of insufficient quality to assign.

Variant growth modelling

EG.5.1.6, XBB.1.16.15, EG.5.1.3, XBB.1.16.6, and EG.5.1.1 have growth advantages over 5% in the model (English data). However, the sample sizes for the majority of lineages are relatively small. The lineage BA.2.86 does not appear in this analysis as it did not meet the criteria of 50 sequenced cases and 1.5% lineage prevalence in the past 6 weeks.

BA.2.86 (V-23AUG-01)

UK cases

There are 48 confirmed cases of BA.2.86 in England.

Of these, 30 cases were identified as part of a care home outbreak investigation in the East of England.

There are an additional 18 cases identified through routine surveillance:

• 1 in the East of England
• 7 in London
• 2 in the North East
• 3 in North West
• 1 in the South East
• 1 in the South West
• 1 in the West Midlands

Two cases have unknown region. Of 11 cases with known information on epidemiological links, 2 cases may be epidemiologically linked. None of the 11 cases with travel information had recent travel history.

Of 48 cases, 10 cases were hospitalised, 2 were tested in an emergency department, and 6 cases have unknown hospitalisation status. There were no known deaths due to COVID-19 among these cases. This data cannot be used to estimate a hospitalisation rate since most testing and genomic surveillance is among people in hospital.

Figure 1. Number of confirmed BA.2.86 cases by specimen date as of 18 September 2023

The data used in this graph can be found in the accompanying spreadsheet.

Figure 2. Age-sex breakdown of BA.2.86 cases as of 18 September 2023

The data used in this graph can be found in the accompanying spreadsheet.

Scotland has reported 6 confirmed cases of BA.2.86, an increase of 1 since the previous technical briefing. Further information about cases is published on the Public Health Scotland webpages.

Global cases

As of 18 September 2023, there are 139 sequences from 137 human cases identified as BA.2.86 available in GISAID. This is an increase of 39 sequences and 38 cases since the last technical briefing. The sequences are from 15 countries:

• Australia (1)
• Canada (2)
• Denmark (1)
• France (78)
• Germany (1)
• Israel (4)
• Japan (2)
• Portugal (2)
• South Africa (17)
• South Korea (1)
• Spain (9)
• Sweden (15)
• Switzerland (1)
• the UK (46)
• the US (15)

One sequence from South Africa was produced from an isolate from an existing case, and one UK sample has been sequenced twice. The earliest collection date is 24 July 2023, and the most recent collection date is 9 September 2023. Figure 3 shows the proportion of sequences from human cases in GISAID that are identified as BA.2.86 by week of collection; the duplicate sequences from South Africa and the UK have been removed. An additional 5 partial sequences from wastewater samples are available from Thailand.

Figure 3. Proportion of sequences uploaded to GISAID since 1 July 2023 that are identified as BA.2.86

Sequences are grouped by week of collection and coloured by world region. Sequence data from the UK has been shown separately to those from other European countries. Where the collection date is not available, the submission date has been used. The proportion of sequences which are BA.2.86 is indicated by the bar (left-hand axis). The total number of sequences in GISAID per week is shown by the black line (right-hand axis), and the total number of BA.2.86 sequences within each week is shown by the number above each bar. Data as of 18 September 2023.

The data used in this graph can be found in the accompanying spreadsheet.

Laboratory investigations

Neutralisation studies

UK laboratories have provided data on live virus and pseudovirus neutralisation studies. This data has been combined in Figure 4, along with a description of the data sets below.

The amalgamation of the data sets was performed by Derek Smith and Sam Turner’s team at the University of Cambridge.

The neutralisation titres for each group are shown against a selection of variants, alongside publicly available data from pre- and post-vaccination sera from the Moderna XBB.1.5 vaccine.

In these data sets, neutralisation titres against BA.2.86 are slightly higher than titres against XBB.1.5 in 3 of 4 UK serum group data, and lower in one serum group. This suggests that the strength of antibody mediated protection against BA.2.86 in the UK population is likely to be similar than against the recently circulating XBB.1.5 variant.

Figure 4. Combined UK neutralisation assay data for serum groups

Serum groups tested are described in the text below. Serum group A is from the Screaton group at Oxford University, B from the Pirbright Institute, C from Vaccine Development and Evaluation Centre (VDEC) UKHSA, D from Glasgow Centre for Virus Research (CVR). Pre and Post are Moderna pre- and post-titres for the whole cohort.

Supplementary data is not available for this figure.

Data set A is pseudovirus neutralisation data from 19 samples collected over the previous year, with a mixture of confirmed infection backgrounds. This data was supplied by the Screaton group at the University of Oxford.

Data set B is pseudovirus neutralisation data from samples from the CONSENSUS study, a UKHSA study of SARS-CoV-2 antibody levels in older adults. Participants had received 4 doses of monovalent Wuhan vaccine, followed by a fifth dose of Wuhan /BA.1 bivalent and a sixth dose of Wuhan /BA.4-5 bivalent. Analysis was performed by the Pirbright Institute (Joseph Newman, Nazia Thakur, Tom Peacock, Dalan Bailey) with samples from UKHSA (Yung Wai Chan, Gayatri Amirthalingam, Kevin Brown)

Data set C is live virus neutralisation data using samples donated by UKHSA and the Medicines and Healthcare products Regulatory Agency (MHRA) staff members after routine vaccination. It should be noted that the assay used to generate this data had not been optimised and system suitability criteria were not in place. Genomic sequencing also showed the furin cleavage site (FCS) (s1/s2 boundary) has an unexpected amino acid change (681RWRAR). A further 3 virus isolates from additional samples are now available, none of which have the mutation at the FCS, and analysis is ongoing. The data was generated by members of the Pre-Clinical and Clinical Evaluation teams, both part of VDEC at UKHSA.

Data set D is pseudovirus neutralisation data from serum samples from 36 participants in the COVID-19 DeplOyed VaccinE Cohort Investigation (DOVE) study. This data was supplied by the CVR Glasgow with acknowledgement to Professor Brian Willet, the DOVE cohort study team, Professor Emma Thomson and the Medical Research Council (MRC).

Lateral flow device (LFD) testing data

Preliminary analyses from LFD testing do not indicate a reduction in analytical sensitivity compared to previous variants. Further information will be available in the next 2 weeks, including results from multiplex LFDs.

Variant prevalence

Testing policy and sequencing should be considered when interpreting variant data. Information about surveillance systems for England are reported by UKHSA in the National Influenza and COVID-19 surveillance report. Nationally COVID-19 healthcare pressures and incidence are not growing rapidly and in the most recent data are plateauing.

The prevalence of lineages among UK sequences by Phylogenetic Assignment of Named Global Outbreak Lineages (Pangolin) designation is presented in Figure 5. Lineages are shown if there are more than or equal to 5,000 sequences since 27 March 2023 or if they represent more than or equal to 1% of sequences within a single week over the last 6 weeks. Lineages that do not meet these criteria are combined with their parent lineage (for example, BA.2.4 is combined with BA.2).

The lineages have been assigned using the accurate Ultrafast Sample placement on Existing tRee (UShER) mode and version 1.22 of the Pangolin data.

Figure 5. Prevalence of Pango lineages in the UK sequence data with a specimen date from week beginning 27 March 2023 to week beginning 4 September 2023, as of 14 September 2023

The total number of valid sequence results per week is shown by the black line. The ‘Other’ category in this plot contains all lineages that do not meet the relevant criteria after combining smaller sub-lineages. ‘Unassigned’ are sequences that could not be assigned a lineage by Pangolin. BA.2.86 and lineages present in at least 2% of sequences in the most recent week are labelled to the right of the plot.

The data used in this graph can be found in the accompanying spreadsheet.

Variant modelling

Methods used for variant modelling have previously been described in the UKHSA prevalence and growth rate briefings.

Growth rates were based on sequences sampled through Pillar 1 testing (primarily positive tests conducted in hospital) in England (Table 1). The sampling range for both the logistic regression generalised linear model (GLM) and generalised additive model (GAM) are from 23 March 2023 to 7 September 2023. Our criteria for modelling a lineage or aggregation of lineages include:

1. Any lineage that has made up more than 1.5% of samples and with 50 sequenced cases within 6 weeks of the most recent specimen date is modelled separately.
2. Lineages with a different high-level parent will never be aggregated together.
3. Unassigned lineages are not included in this analysis.

The model fit for any lineage with a positive growth rate (with 95% confidence intervals (CIs) that do not cross zero) is shown in Figure 6. The lineages that were estimated to have a positive growth rate advantage over 5% with reasonable uncertainty (CIs not below zero in the GAM) are:

• EG.5.1.6 (26.52%, GAM)
• XBB.1.16.15 (21.48%, GAM)
• EG.5.1.3 (18.37%, GAM)
• XBB.1.16.6 (10.53%, GAM)
• EG.5.1.1(6.73%, GAM)

Table 1. Growth rate (GR) of English sequence lineages as of 7 September 2023†

Lineage*	Lineage Group Composition**	Pillar 1 Sample Size***	Weekly growth rate advantage (GAM)	Estimated prevalence¥ (GAM)	Weekly growth rate advantage (GLM)
EG.5.1.6 (XBB.1.9.2.5.1.6)	EG.5.1.6 (84.75%); HV.1 (15.25%)	59	26.52% (95% CI: 22 to 31.04)	2.97% (95% CI: 2.08 to 4.22)	24.51% (95% CI: -13.6 to 62.61)
XBB.1.16.15	XBB.1.16.15 (100%)	140	21.48% (95% CI: 17.23 to 25.73)	6.37% (95% CI: 4.82 to 8.38)	6.51% (95% CI: -18.56 to 31.58)
EG.5.1.3 (XBB.1.9.2.5.1.3)	EG.5.1.3 (100%)	184	18.37% (95% CI: 13.86 to 22.88)	7.69% (95% CI: 5.94 to 9.9)	4.56% (95% CI: -18.09 to 27.2)
XBB.1.16.6	XBB.1.16.6 (100%)	178	10.53% (95% CI: 5.64 to 15.42)	6.65% (95% CI: 4.6 to 9.51)	11.52% (95% CI: -10.06 to 33.11)
EG.5.1.1 (XBB.1.9.2.5.1.1)	EG.5.1.1 (97.08%); HK.3 (2.51%); HK.2 (0.21%); HK.1 (0.21%)	479	6.73% (95% CI: 5.32 to 8.14)	15.98% (95% CI: 12.02 to 20.93)	6.5% (95% CI: -8.52 to 21.51)
XBB.1.16.1	XBB.1.16.1 (62.79%); FU.1 (25.58%); FU.2 (11.63%)	86	2.27% (95% CI: -9.12 to 13.67)	2.08% (95% CI: 1.22 to 3.53)	11.44% (95% CI: -23.56 to 46.45)
EG.5.1 (XBB.1.9.2.5.1)	EG.5.1 (94.85%); EG.5.1.5 (3%); EG.5.1.2 (2.15%)	233	0.37% (95% CI: -5.55 to 6.29)	6.27% (95% CI: 4.32 to 9.01)	1.76% (95% CI: -0.91 to 4.43)
XBB.1.16.11	XBB.1.16.11 (100%)	141	-0.02% (95% CI: -7.65 to 7.62)	4.56% (95% CI: 2.26 to 8.98)	-11.25% (95% CI: -36.18 to 13.67)
XBB.1.16	XBB.1.16 (66.67%); XBB.1.16.21 (6.8%); XBB.1.16.9 (6.12%); XBB.1.16.2 (3.91%); XBB.1.16.3 (3.57%)…	588	-3.14% (95% CI: -8.14 to 1.86)	11.8% (95% CI: 8.82 to 15.61)	-15.77% (95% CI: -30.21 to -1.32)
EG.5.1.4 (XBB.1.9.2.5.1.4)	EG.5.1.4 (100%)	128	-9.62% (95% CI: -16.5 to -2.74)	4.12% (95% CI: 2.63 to 6.4)	7.89% (95% CI: -17.56 to 33.35)
XBB.1.5	XBB.1.5 (45.99%); XBB.1.5.59 (10.22%); XBB.1.5.71 (6.57%); XBB.1.5.11 (5.11%); XBB.1.5.72 (4.38%)…	137	-10.58% (95% CI: -17.49 to -3.68)	2.51% (95% CI: 1.79 to 3.51)	-4.55% (95% CI: -32.9 to 23.8)
GE.1 (XBB.2.3.10.1)	GE.1 (100%)	192	-11.75% (95% CI: -20.07 to -3.43)	3.86% (95% CI: 2.56 to 5.79)	-11.91% (95% CI: -34.18 to 10.37)
EG.6.1 (XBB.1.9.2.6.1)	EG.6.1 (100%)	64	-12.38% (95% CI: -25.62 to 0.85)	1.08% (95% CI: 0.57 to 2.05)	18.35% (95% CI: -20.71 to 57.4)
FL.1.5.1 (XBB.1.9.1.1.5.1)	FL.1.5.1 (98.15%); HN.1 (1.85%)	108	-15.5% (95% CI: -24.99 to -6)	2.52% (95% CI: 1.51 to 4.18)	5.61% (95% CI: -22.43 to 33.66)
XBB.2.3.11	XBB.2.3.11 (93.28%); GS.1 (6.72%)	119	-27.82% (95% CI: -40.23 to -15.42)	1.87% (95% CI: 0.95 to 3.67)	-25.53% (95% CI: -53.5 to 2.45)

*Listed parent lineages include all sub-lineages, other than those explicitly modelled.

** The top 5 contributing lineages to the modelled group in the most recent 6 weeks (27 July 2023 to 7 September 2023). More than 5 sublineages are indicated by ‘…’

*** Sample size is for Pillar 1 samples in England in the most recent 6 weeks (27 July 2023 to 7 September 2023).

¥ Estimated prevalence for the 7 September 2023.

† Sampling range for both logistic regression and generalised additive models (GAM) is from 23 March 2023 to 7 September 2023.

CI = confidence intervals

Figure 6. Modelled prevalence of lineage groups with a growth rate advantage over other circulating lineages

The black line shows the central estimate and blue shaded regions the 95% confidence intervals. Points show the national level proportions, with the size being indicative of the sample size for that particular lineage. The grey portion of the ribbon denotes that this period of time is likely to be backfilled with more sequenced cases, making proportions unreliable.

Supplementary data is not available for this figure.

Published information on variants

On 1 April 2022 UKHSA amended its variant classification system. Further details are available in technical briefing 39.

SARS-CoV-2 routine variant data update covers surveillance data and sequencing coverage data on all other variants up to 25 March 2022.

The variant collection page gives content on variants, including previous technical briefings. Technical briefings are published periodically.

The UKHSA variant definition repository contains the previous genomic definitions for UKHSA declared variants.

Sources and acknowledgments

Data sources

Data used in this investigation is derived from the CLIMB and UKHSA genomic programme data set, the UKHSA Second Generation Surveillance System, the Secondary Uses Service data set, Emergency Care Data Set, the UKHSA Case and Incident Management System and GISAID.

Authors of this report

Amelia Kelly

Bassam Hallis

Deborah Williamson

Derek Smith

Eileen Gallagher

Emma Thomson

Gavin Dabrera

Gavin Screaton

James Gilbert

Joseph Newman

Kate Agami

Katy Davidson

Laura Lopez Pascua

Martyn Fyles

Meaghan Kall

Meera Chand

Natalie Groves

Nurin Iwani Binti Abdul Aziz

Sam Turner

Sue Charlton

Susan Hopkins

Susie Dunachie

Tom Peacock

Tom Ward

Variant Technical Group members

Chair

Meera Chand (UKHSA)

Genomics and bioinformatics

Andrew Rambaut (University of Edinburgh)

Eileen Gallagher (UKHSA)

Leena Inamdar (UKHSA)

Matt Holden (Public Health Scotland)

Natalie Groves (UKHSA)

Nicholas Loman (UKHSA and University of Birmingham)

Richard Myers (UKHSA)

Virology and immunology

Bassam Hallis (UKHSA)

Deborah Williamson (UKHSA)

Derek Smith (University of Cambridge)

Emma Thomson (University of Glasgow and London School of Hygiene and Tropical Medicine)

Gavin Screaton (University of Oxford)

Katie Binley (Northern Ireland Public Health Agency)

Lance Turtle (University of Liverpool)

Maria Zambon (UKHSA)

Ravi Gupta (University of Cambridge)

Susanna Dunachie (University of Oxford)

Thomas Peacock (Pirbright Institute and Imperial College London)

Wendy Barclay (Imperial College London)

Epidemiology and modelling

Chris Williams (Public Health Wales)

Daniela de Angelis (University of Cambridge)

Erik Volz (UKHSA and Imperial College London)

Fergus Cumming (UKHSA)

Gavin Debrera (UKHSA)

Jamie Lopez-Bernal (UKHSA)

Jim McMenamin (Public Health Scotland)

John Edmunds (London School of Hygiene and Tropical Medicine)

Julia Gog (Scientific Pandemic Influenza Group on Modelling and University of Cambridge)

Kimberly Marsh (Public Health Scotland)

Meaghan Kall (UKHSA)

Neil Ferguson (Imperial College London)

Nick Watkins (UKHSA)

Susan Hopkins (UKHSA)

Thomas Finnie (UKHSA)

Tom Ward (UKHSA)

International epidemiology

Chris Lewis (Foreign, Commonwealth and Development Office)

Acknowledgements

The authors are grateful to those teams and groups providing data for these analyses including: the National Health Service, CLIMB, the Wellcome Sanger Institute, Health Protection Data Science teams, the Genotype to Phenotype Consortium, Medical Research Council Biostatistics Unit, the Francis Crick Institute, Cambridge and Imperial College, London.

We are also grateful to the SIREN and CONSENSUS studies as well as Kings College London for providing materials for laboratory analysis.