OFFLU Avian Influenza Matching (AIM) Report, October 2023

OFFLU AVIAN INFLUENZA MATCHING (OFFLU-AIM) REPORT 23 October 2023 OFFLU avian influenza matching report, October 2023 Page 2 Contents Summary ..................................................................................................................................................... 3 Introduction ................................................................................................................................................ 5 Findings ...................................................................................................................................................... 7 Clade 2.3.4.4b......................................................................................................................................... 7 Clade 2.3.2.1a......................................................................................................................................... 9 Clade “2.3.2.1c” ...................................................................................................................................... 9 Influenza A(H7N6) in South Africa ......................................................................................................... 10 Conclusions............................................................................................................................................... 10 References ................................................................................................................................................. 10 Annex 1 (A guide to assessing antigenic characteristics of avian influenza viruses at national/sub-national level laboratories) ................................................................................................ 13 Annex 2 (Antigenic Cartography Results) ............................................................................................. 15 Summary ................................................................................................................................................ 15 Introduction ............................................................................................................................................ 16 Vaccine seed strains .............................................................................................................................. 16 Methods, analyses and results .............................................................................................................. 16 Generation of chicken antisera .............................................................................................................. 16 Hemagglutination Inhibition (HI) assay .................................................................................................. 18 Antigenic Cartography ........................................................................................................................... 18 Chicken antisera panel .......................................................................................................................... 19 Antigenic maps ...................................................................................................................................... 20 Table of antigenic distances .................................................................................................................. 21 Main findings .......................................................................................................................................... 22 2.3.2.1a .................................................................................................................................................. 22 2.3.4.4b .................................................................................................................................................. 22 Recommendations and next steps for OFFLU AIM............................................................................... 23 References............................................................................................................................................. 24 Acknowledgements .................................................................................................................................. 26 OFFLU avian influenza matching report, October 2023 Page 3 Summary Effectiveness of vaccination of poultry against high pathogenicity avian influenza (HPAI) depends on usage of appropriate vaccines at the recommended doses and age. Antigenic variation in field strains may render vaccines less effective. This is particularly true for vaccines that depend primarily on the humoral immune response, such as killed antigen (inactivated) adjuvanted vaccines. OFFLU-AIM is designed to provide information on possible antigenic changes in HPAI viruses that could reduce their effectiveness. This report focuses on goose/Guangdong/1/96-lineage A(H5Nx) viruses (gsGD). The information in this report is derived from multiple sources including antigenic studies conducted at OFFLU reference laboratories. Clade 2.3.4.4b A(H5N1) HPAI viruses pose the greatest avian influenza threat globally. They are circulating across Eurasia, parts of Africa, and the Americas. Most activity at present (October 2023) has been reported from Northern Europe with a recent increase in cases in North America. These viruses are still present in other regions, causing disease in wild birds, wild mammals and, in some places, poultry. They are expected to expand their range to other parts of Europe, North America and Africa in the autumn of 2023, largely via carriage of viruses by wild birds. At present, there is no information available to suggest that significant antigenic variant clade 2.3.4.4b viruses are circulating that would interfere with vaccination. This requires use of vaccines based on related clade 2.3.4.4 viruses (including antigens derived from 2.3.4.4b and 2.3.4.4c viruses) and those based on other clades/antigens that have been demonstrated to be efficacious against clade 2.3.4.4b viruses in experimental studies. These vaccines are likely to provide a high degree of protection, as long as sufficient doses are given, and other diseases do not result in immunosuppression at the time of vaccination. Differences in response to vaccines in types of poultry other than chickens need to be recognized. Vaccine trials conducted in France and the Netherlands demonstrated protection from disease and transmission in poultry against clade 2.3.4.4b viruses isolated in these two locations in 2021. The French study involved mule ducks vaccinated with either a self-replicating RNA virus vaccine or a subunit vaccine. The vaccines providing protection in layers in the Netherlands were HVT vector vaccines. These experimental studies may not reflect the situation in the field where other factors can impact on vaccine effectiveness but did demonstrate good protection under laboratory conditions. Elsewhere, updated vaccine antigens were introduced for clade 2.3.4.4b viruses in China (e.g., re-14) and Egypt based on strains from 2020 and 2018, respectively. Additional studies have been conducted in turkeys but have not been reported formally. For those intending to vaccinate turkeys, we recommend discussions with OFFLU or other experts when determining appropriate methods and candidate products to consider for vaccination. Killed antigen vaccines developed against earlier gsGD clades, such as clade 2.2, clade 2.1, 2.3.2.1, clade 1 or clade 7, and their derivatives, are unlikely to afford consistent protection against clade 2.3.4.4b viruses and should only be used as a component of vaccines in places where viruses in these other clades persist. Antigenic differences also exist between clade 2.3.4.4h and 2.3.4.4b viruses that would reduce efficacy of inactivated vaccines based on the former. Clade 2.3.2.1a viruses are present in South Asia where vaccination is used officially in Bangladesh, with some informal usage of vaccination elsewhere. Up to 2018, viruses from Bangladesh demonstrated reasonable antigenic stability but insufficient evidence is available on recent viruses. Tests should be done against recent strains from South Asia to assess whether antigenic changes have occurred that might affect efficacy of existing avian vaccines. Inactivated vaccines based on clade 2.3.4.4b would not be expected to provide appropriate protection against clade 2.3.2.1a viruses. OFFLU avian influenza matching report, October 2023 Page 4 Clade “2.3.2.1c” 1 viruses are still circulating in SE Asia, including Cambodia, Laos and Vietnam. Further studies are needed on the antigenic characteristics of recent viruses, however genetic information suggests these viruses are likely to have remained antigenically stable. Inactivated vaccines based on Clade 2.3.4.4b would not be expected to provide consistent protection against clade 2.3.2.1c viruses. Other viruses referred to as clade “2.3.2.1c” in Indonesia, and more recently as 2.3.2.1e, are still present in Indonesia and Timor Leste but few have been isolated or fully characterized recently. Again, further studies are needed to assess antigenic characteristics of these recent viruses. As with other clade 2.3.2.1 viruses, inactivated vaccines based on clade 2.3.4.4b viruses would not be expected to provide consistent protection against these viruses. Viruses belonging to multiple gsGD HA clades, including recently introduced clade 2.3.4.4b viruses, are co-circulating in some countries. It is still not known whether pre-existing strains will persist in these places once clade 2.3.4.4b becomes established. Nevertheless, it appears that clade 2.3.4.4h has largely been replaced by clade 2.3.4.4b in Asia. If two antigenically distinct strains persist, consideration should be given to introducing a bivalent vaccine, as has been done in some countries in the past, to ensure optimal protection against both circulating strains. When vaccines are used, post-vaccinal antibody responses should be monitored to ensure an appropriate response to vaccination. If using killed antigen vaccines, flocks should be monitored throughout the life of the vaccinated flock. Booster doses should be provided if titres fall below levels expected to prevent disease. Antibody titres are less reliable indicators of likely protection when using vector vaccines alone. It is also essential to monitor for infection in vaccinated flocks. Any viruses detected in vaccinated flocks should be characterised and assessed for any potential adaptations that might reflect antigenic changes. Nothing in this report should be construed as an endorsement of a specific product and the information provided is general in nature. Additional support on appropriate vaccines can be obtained from FAO, WOAH and OFFLU. 1 Clade 2.3.2.1c viruses have evolved into different sublineages since they were last formally classified in 2014. An updated, standardised clade nomenclature for viruses that previously fell within clade 2.3.2.1c is being developed but has not yet been accepted formally. Nevertheless, viruses from Indonesia that were previously referred to as 2.3.2.1c have been referred to as 2.3.2.1e in OFFLU and WHO documents for several years (see footnote 2), based on advice from molecular geneticists, given they differ from those found in mainland SE Asia. Further changes in classification are expected for viruses within clade 2.3.2.1c and OFFLU will advise once the new nomenclature has been accepted and adopted. OFFLU avian influenza matching report, October 2023 Page 5 Introduction Currently, some countries using vaccines against avian influenza have mechanisms in place to monitor antigenic changes in field strains. However, this information is not always shared internationally or made publicly available. The goal of OFFLU Avian Influenza Matching (AIM) project (OFFLU-AIM) is to provide improved information on antigenic characteristics of avian influenza viruses to support vaccination programs against avian influenza. It will also assist countries utilising vaccines to stay informed about potentially significant changes in the antigenic characteristics of circulating strains. This information is also of value for countries considering introduction of vaccination as an additional preventive/control measure. OFFLU-AIM serves as a platform where current epidemiological, genetic, and antigenic information on AI viruses circulating in different geographic regions is analysed and made freely accessible. Information comes from a range of sources including published information, data from countries monitoring for avian influenza viruses including potential antigenic variants, vaccine challenge trials, specific antigenic cartography and other intelligence that warrants consideration by countries using or considering vaccination. Multiple countries are using vaccination against gsGD HPAI viruses, given the difficulties encountered in detecting and preventing all infections and the threat of zoonotic infections. Early adopters of vaccination include China, Egypt, Indonesia, Vietnam and Bangladesh – all places where the virus was well entrenched in poultry before vaccination commenced. Several other countries used vaccines briefly in poultry and unregistered vaccine is being applied in some other countries, a practice that OFFLU does not recommend. Some places (e.g., France and Hong Kong SAR) use preventive vaccination, despite the virus not being present in poultry, because of the high risk of infection. The increased threat posed by clade 2.3.4.4b viruses since 2020 has been a catalyst for increased consideration or adoption of vaccination. In 2023, France (vaccination of mule ducks used for foie gras production) and several countries in South and Central America have introduced vaccination. Most of the vaccines used at present are adjuvanted killed antigen vaccines containing whole virus, usually produced by reverse genetics. Other vaccines are also in use or under development, including viral vector vaccines such as those using herpesvirus of turkeys (HVT) as the vector, subunit vaccines (used currently in France), self-amplifying RNA vaccines, DNA vaccines, virus-like particles, and replicon particles. Some of these vaccines, especially those containing a live virus can produce broader cross protection against heterologous strains of A(H5Nx) HPAI virus because they also stimulate cell mediated immunity (see for example Palya et al 2018). This means that antigens used in these vaccines do not necessarily require updating as frequently as those in killed antigen vaccines; they can tolerate some changes in antigenicity. Nevertheless, it is still critical to monitor flocks vaccinated with these vaccines to detect infection and to characterise any viruses found in vaccinated flocks, by undertaking a thorough assessment of their antigenic characteristics. The focus of this first OFFLU-AIM report is on HPAI viruses of the H5Nx subtype in the goose/Guangdong/1/96 lineage (gsGD). These HPAI viruses emerged in China in 1996 and have been circulating as high pathogenicity viruses since that time. They have evolved through point mutations into multiple 5th order HA clades and have reassorted to produce multiple genotypes. They have undertaken several intercontinental waves of transmission since 2005 due mainly to carriage by wild birds, presumably after spillback of infection from domestic poultry in places where wild birds and poultry mix freely. Of the initial 10 clades that emerged in the early 2000s only viruses within clade 2 and its derivatives have persisted and evolved with the viruses detected recently belonging to clades 2.3.4.4b (widespread globally except Oceania), 2.3.2.1a (South Asia), “2.3.2.1c” (lower Mekong including Cambodia, Laos and Vietnam with related strains in Indonesia and Timor Leste) and possibly 2.3.4.4h. Vaccines targeting each of these clades are being used in at least one country (2.3.4.4b – multiple countries, 2.3.2.1a – Bangladesh, “2.3.2.1c” – Vietnam; “2.3.2.1c” – Indonesia; 2.3.4.4h - China). OFFLU avian influenza matching report, October 2023 Page 6 This report should be read in conjunction with the OFFLU and WHO vaccine composition meeting reports covering avian influenza and other potentially zoonotic influenza A viruses detected from February to September 2023 2 . The focus of the VCM report is preparation of candidate vaccine antigens for use in humans in the event of an influenza pandemic but it provides important information on viruses detected recently. It also provides information on antigenic and genetic characteristics of viruses with the former performed using ferret antisera produced against human candidate vaccine reference antigens, not chicken sera and avian vaccine antigens. Sera from ferrets can produce different results to those raised in chickens. OFFLU-AIM is building capacity to test viruses using antisera raised in chickens against avian vaccine antigens, or surrogate antigens when the vaccine strain is not available. In this first report some initial studies were conducted using viruses detected in 2020, 2021 and 2022, predominantly belonging to clade 2.3.4.4b (see Annex 2). This work will be expanded in subsequent years. Two types of events are relevant when considering antigenic changes in avian influenza viruses. The first is the evolution of viruses against which vaccines are already being used within the country or region. Since gsGD HPAI viruses emerged they have undergone considerable antigenic variation, especially in countries where these viruses remain entrenched. Examples include the antigenic variation that occurred in Egyptian clade 2.2 viruses and Indonesian clade 2.1 viruses, and their derivatives (Arafa et al 2012, Swayne et al, 2015). Antigenic change has occurred rapidly in some places whereas in other places the same vaccine antigen has been used and remained appropriate over multiple years (e.g., vaccines used in Viet Nam against clade 1 viruses and their derivatives from 2005 onwards until this clade was replaced by clade 2.3.2.1c). The second is the introduction of a strain of virus that differs significantly from existing viruses and vaccine antigens. In other words, antigenic changes developed elsewhere, and the novel strain was imported, usually, but not always, via wild birds. This is exemplified by the recent introduction of A(H5N1) clade 2.3.4.4b to Indonesia where vaccines in use are designed to protect against A(H5N1) clade “2.3.2.1c” viruses that still circulate there. In some countries the new strain replaced the existing strain(s) but in others both pre-existing and new strains co-circulate or the new strain is restricted to only part of the country (e.g., clade 2.3.4.4b in Indonesia). The information in this document is generic in nature and depends on reports of information from individual countries on circulating viruses. It does not imply preference for specific vaccines or products. All countries considering vaccination should use the information in this document as a supplement to data collected locally on circulating strains and on available vaccines when determining the most appropriate vaccines and vaccination strategies to apply. We urge all countries to continue sharing information on novel strains of virus, especially those suspected of being antigenic variants (such as viruses that are circulating in well vaccinated poultry) so that OFFLU can continue to provide up-to-date information on viruses in circulation. Where possible, and especially if there are suspicions of antigenic variants, viruses and gene sequences should be shared with OFFLU reference laboratories to allow full antigenic and genetic characterization. Much of the information in this document is of greater relevance for killed antigen vaccines (still the most commonly used avian influenza vaccines globally) than vaccines that are based on a live vector virus or others that also stimulate cell mediated immunity and do not rely primarily on antibody mediated responses 2 Available at https://www.offlu.org/wp-content/uploads/2023/10/Avian-OFFLU-VCM-S23-OFFLU-V6-VCM-OFFLU.pdf. The WHO report that includes these data is available at https://cdn.who.int/media/docs/default-source/influenza/whoinfluenza-recommendations/vcm-southern-hemisphere-recommendation2024/202309_zoonotic_vaccinvirusupdate.pdf?sfvrsn=e78676a0_5 OFFLU avian influenza matching report, October 2023 Page 7 to vaccination for protection. The former are more likely to require updating because of antigenic changes given the latter can often provide coverage against more antigenically distant gsGD HPAI viruses strains. Nevertheless, the adjuvant in killed antigen vaccines can overcome some antigenic changes, which is one reason why vaccines for poultry do not usually require annual updates, as is the case for human Influenza A vaccines. Serological monitoring of vaccinated flocks for evidence of immunity is strongly recommended especially for vaccines that generate a strong humoral response. Antibody monitoring can be of limited value as an indicator of protection for vaccines that generate cell mediated immunity such as HVT vector vaccines, especially when the HA insert is only distantly related to field strains. Nevertheless, testing of vaccinated flocks should be undertaken to ensure the HVT vector has replicated in vaccinated birds and in situations where booster inactivated vaccines have been used following initial priming with the vector vaccine, antibody testing should be done to monitor the response. Findings Clade 2.3.4.4b At present, the major HPAI threat globally is from gsGD A(H5Nx) viruses, mainly A(H5N1), in clade 2.3.4.4b. In the past 3 months, most viral activity has centred on northern Europe and South America, with cases again persisting in wild birds during the summer and occasional spillover to poultry. New cases in commercial poultry farms have also been reported in North America in late September and early October after a pause during the summer. 3 Wild bird and mammalian cases were still being detected in North America throughout 2023. Antigenic assessments using reference chicken antisera (see Annex 2) have demonstrated some variation in response by strains from 2020 to 2022 but there was no apparent increase in the antigenic distances overall for viruses detected between the different years. Some published papers have reported changes at putative antigenic sites (e.g., Yang et al 2023). However, the changes are probably not enough to interfere with protection from a recent well-matched killed antigen adjuvanted vaccine. In some countries clade 2.3.4.4b viruses now co-exist with other endemic strains. It may be necessary to introduce multivalent vaccines covering all antigenically distinct strains as has been the case in some countries (e.g., China with a trivalent vaccine covering gsGD clade 2.3.4.4b (re-14 antigen), clade 2.3.4.4h (re-13) and H7N9 (re-4)). However, it is already evident in some countries that clade 2.3.4.4b has largely replaced some other strains (e.g., other 2.3.4.4 clades in Vietnam). Europe Recent HPAI events in Europe have been described in detail elsewhere (EFSA 2023). Viruses detected in Europe from 2023 have not been subjected to antigenic characterisation using reference sera in OFFLU reference laboratories but genetic analysis of strains suggests few significant changes at putative or possible antigenic sites. Nevertheless, if novel strains of virus are identified, especially in vaccinated flocks, additional characterisation as discussed in Annex 1 is strongly recommended. Vaccine challenge and transmission studies have been conducted in several European countries. Vaccine trials conducted in France and the Netherlands have demonstrated protection from disease and transmission in poultry against clade 2.3.4.4b viruses isolated in these two locations in 2021. The French 3 For reports from the USA sin poultry see https://www.aphis.usda.gov/aphis/ourfocus/animalhealth/animal-diseaseinformation/avian/avian-influenza/hpai-2022/2022-hpai-commercial-backyard-flocks. For Canada see https://inspection.canada.ca/animal-health/terrestrial-animals/diseases/reportable/avian-influenza/latest-bird-flusituation/investigations-and-orders/eng/1688503773556/1688503774196 OFFLU avian influenza matching report, October 2023 Page 8 study 4 involved mule ducks vaccinated with either a self-replicating RNA virus or a subunit vaccine. The vaccines providing protection in layers in the Netherlands were HVT vector vaccines 5 . These experimental studies may not reflect the situation in the field where other factors can impact on vaccine effectiveness but demonstrated good protection and prevention of transmission under laboratory conditions. Studies on vaccination in turkeys have also been undertaken in Italy with results still to be published. Results from trials with chickens may not be directly applicable to other species, including turkeys (EFSA AHAW Panel 2023b). Africa Africa has experienced multiple introductions of gsGD H5Nx clade 2.3.4.4b viruses from Eurasia since 2016. These introductions have led to the maintenance of these viruses in poultry populations, in West and southern Africa, and Egypt with the replacement of previously circulating viruses and in some cases reassortment with locally circulating endemic viruses (including A(H9N2). Many of these viruses have continued to circulate in poultry in geographically segregated regions, with some new introductions and virus detected in wild sea birds and poultry (Abolnik et al 2023, Meseko et al 2023). Antigenic characterization of recent (2023) viruses has not been reported. Some of the A(H5N8) viruses in Africa have lost a glycosylation site in the HA glycoprotein but this does not appear to be related to changes in HI titres when tested against reference antisera. Egypt also has endemic infection with clade 2.3.4.4b viruses. Potential new clade 2.3.4.4b experimental vaccines (based on A/chicken/Egypt/F71-S86C/2022(H5N8) and A/chicken/Egypt/F71F114C/2022(H5N1) were developed in Egypt using reverse genetics. The killed antigen, adjuvanted vaccines produced from these viruses provided protection against clade 2.3.4.4b A(H5N1) and A(H5N8) viruses (from 2021) but were less effective against a clade 2.2.1 virus as expected (Mahmoud et al 2023). Trials were conducted with existing vaccines in Egypt against clade 2.3.4.4b viruses when they first emerged (Kandeil et al 2018) and demonstrated imperfect protection by most of the commercially available vaccines and the need to update vaccine antigens. Another study has also confirmed that vaccines against clade 2.2 derivatives were no longer fully effective against clade 2.3.4.4b viruses (Salaheldin et al 2022). A new vaccine antigen was introduced by a commercial company to combat clade 2.3.4.4b viruses (Ibraham et al 2021) based on a 2018 virus. Antigenic characterisation of A(H5N8) viruses was undertaken (Kandeil et al 2022) and some antigenic changes were detected in flocks that had been vaccinated with poorly matched vaccines, demonstrating the importance of only using appropriately matched vaccines. Information on antigenic characteristics of recent (2023) clade 2.3.4.4b viruses from Egypt is not available. North America In the past three months viral activity remained low in North America with only a small number of wild bird cases reported, predominantly from the northwest of the US and occasional positive cases in live bird markets in the northeast of the country (New York and New Jersey). Since late September, multiple cases in commercial poultry have been reported from Alberta, Canada and South Dakota and Utah in the USA. There are no reports regarding the antigenic characteristics of these recent viruses, but earlier viruses appeared to have few if any mutations in known antigenic sites. 4 https://www.anses.fr/en/results-vaccination-ducks-%20avian-influenza 5 https://www.wur.nl/en/research-results/research-institutes/bioveterinary-research/show-bvr/two-vaccines-effectiveagainst-bird-flu.htm OFFLU avian influenza matching report, October 2023 Page 9 Somewhat surprisingly, clade 2.3.4.4b viruses in the US remained relatively similar to clade 2.3.4.4.c antigenically, based on results using reference antisera (see Annex 2) despite several differences at putative antigenic sites. Similar results were obtained using ferret antisera (Kandeil et al 2023). Central and South America A(H5N1) gsGD clade 2.3.4.4b viruses were first detected in Central and South America in 2022 and have spread to the southern tip of South America, with cases in wild birds and sea lions reported in southern Chile and Argentina. They continue to circulate in wild birds and poultry although some countries have declared freedom from infection in poultry. Viruses reported so far are similar to those detected in North America with multiple incursions (JiminezBluhm et al 2023, Ruiz-Saenz et al 2023) Recent isolates are not available from countries where vaccines have been deployed and it will be important to continue to monitor viruses in areas where vaccination is being used and to fully characterize them antigenically. Asia Clade 2.3.4.4b viruses are widespread across much of Asia. Limited information is available on the antigenic characteristics of recent viruses, however those for which sequence data are available have few changes at likely antigenic sites. China updated vaccine antigens in early 2022 following the re-introduction of clade 2.3.4.4b viruses (Shi et al 2023). Clade 2.3.4.4b viruses appear to have largely replaced other clades in China. Studies performed on wild bird isolates from 2021 in China demonstrated that the Re-14 vaccine antigen was an appropriate match for these viruses (Tian et al 2023). Antigenic assessments on clade 2.3.4.4b viruses from Cambodia from 2021 demonstrated differences that cannot be explained readily by changes at putative antigenic sites (see Annex 2). Further studies will be undertaken on these viruses. Clade 2.3.2.1a Viruses in this clade have been circulating in South Asia since 2011. They have persisted despite several incursions of clade 2.3.4.4 viruses. A study was conducted on effectiveness of vaccines against clade 2.3.2.1a viruses from Bangladesh that suggested some antigenic variation in viruses from crows but less so in poultry (Kwon et al 2021). No recent results of antigenic characterization are available although some variation at putative antigenic sites has been reported in the OFFLU VCM report. Further studies are needed to assess vaccine efficacy against recent clade 2.3.2.1a viruses detected in South Asia. Killed antigen vaccines containing antigens derived from clade 2.3.4.4 viruses would likely provide sub-optimal protection against clade 2.3.2.1a virus based on the antigenic distance between the two viruses (see results in Annex 2). Clade “2.3.2.1c” Viruses referred to in-country as clade 2.3.2.1c and in OFFLU and WHO publications as clade 2.3.2.1e2, are circulating in Indonesia and Timor Leste but few recent viruses have been detected or fully characterised. Indonesia has a system in place for updating vaccines when antigenic variants emerge (Hartaningsih et al 2015). Vaccines targeting these viruses would be unlikely to provide appropriate protection against clade 2.3.4.4.b viruses now present in Indonesia if these become more widespread. Further studies are needed on the antigenic characteristics of recent clade “2.3.2.1c” viruses from Cambodia, Viet Nam and Laos, however genetic information suggests these viruses are likely to have remained antigenically stable. Inactivated vaccines based on Clade 2.3.4.4b would not be expected to provide consistent protection against clade 2.3.2.1c viruses. OFFLU avian influenza matching report, October 2023 Page 16 Antigenic cartography enables the integration of phenotypic data from multiple laboratories, providing a standardised assessment and comparative analysis of globally circulating gsGD lineage H5 viruses against commonly used vaccine antigens. Large antigenic distances from antisera raised against surrogate vaccine antigens to currently circulating strains might indicate limited protection from vaccination in the field. OFFLU will continue to raise chicken antisera against more vaccine antigens and test additional clades of contemporary circulating strains. In the next report this will explicitly ensure contemporaneously circulating viruses in 2023/24 are included. Introduction This annex provides information on one of the key goals of OFFU-AIM which was to establish standardized systems for antigenic cartography for gsGD HPAI A(H5Nx) viruses relevant for poultry. This work summarises progress made in this area including interpretation of the findings. Vaccine seed strains Strains of virus were selected based on vaccines that are being or have been used in the past. The following table provides information on vaccine seed strains and also surrogate strains that were used when the vaccine strain was not available. Methods, analyses and results Generation of chicken antisera A standardised panel of chicken sera and homologous antigens were generated at IZSVe and APHA, immunising birds with adjuvant as this likely most closely reflects the chicken immune response when vaccinated. Sera were generated using antigens from viruses which are genetically similar to vaccine seed strains, which we refer to throughout the document as surrogate vaccine seed strains. For each antiserum, a batch of the homologous virus was generated in SPF embryonated chicken eggs and the allantoic fluid inactivated by beta-propiolactone (BPL). Batches had a minimum antigen content equal to 120.000 HAU (e.g., 100 ml with a titer of 1:32). Batches underwent innocuity testing by three blind passages in SPF embyronated chicken eggs. Specific pathogen free (SPF) chickens (n=10/antigen) were inoculated via the intramuscular route with 1ml BPL inactivated avian influenza antigen (non-infectious antigen) plus Montanide as an adjuvant (inoculation 1). (Virus/adjuvant [30/70%]). At 14 days (+/- 3 days) post prime inoculation, birds were boosted (inoculation 2) with the same antigen preparation and procedure. Around 20 days after first inoculation, blood was collected from the birds via wing bleeds and used to assess antibody status against the inoculated antigen. If antibody titres were below useful levels (usually <1/16 HAIUs/25 ul) birds were boosted again (inoculation 3). Bleeding and boosting was repeated another 1 time if necessary (inoculation 4). At a maximum of day 36, all birds were euthanised and blood collected via heart bleeds under terminal anaesthesia. The serum was separated from red blood cells and heat treated at 56°C for 30 min prior to being lyophilized for storage and distribution. For each batch of serum and antigen generated, half the volumes were shipped between IZSVe and APHA. OFFLU avian influenza matching report, October 2023 Page 17 Table 1: Seed strains of vaccines which are currently understood to be in use or have been used in the past. Where available clade information and references have been included. Asterisks (*) denote strains for which surrogate antigens have been included in AIM. Seed strain Clade Reference *Surrogate included in AIM A/duck/Potsdam/1402/86 LPAIV Fereidouni et al., 2009 A/chicken/Vietnam/C58/2004* 1 Boltz et al., 2009 A/Vietnam/1194/2004/1 A/turkey/turkey/2005-like* 2.2 Palya et al., 2018 A/Turkey/turkey/2005 A/chicken/Egypt/Q1995D/2010 2.2.1.1 Kayali et al., 2016 A/chicken/Egypt/D10552B/2015 2.2.1.2 Gomaa et al., 2019 A/chicken/Egypt/M2583D/2010 2.2.1.2 A/Chicken/Egypt/RG-173CAL/2017 2.2.1.2 Ibrahim et al., 2022 A/duck/China/E319-2/2003 2.3.2 Cavalcanti et al., 2017 A/duck/Guangdong/S1322/2010* 2.3.2.1 Shi et al., 2022 A/chicken/Nepal/T360/20 14 A/duck/Sukoharjo/BBVW-14289/2012* “2.3.2.1c” Indriani et al., 2014 A/mynah/Indonesia(Austr iaQ)/13064792-010/2013 A/Hubei/1/2010 2.3.2.1a A chicken/Tanggamus/03171107665/2017 2.3.2.1c A/Duck/VietNam/QB7412 “2.3.2.1c” rgCA2/2.3.2.1d 2.3.2.1d Kang et al., 2022 A/Chicken/Liaoning/SD007/2017 2.3.2.1f Shi et al., 2022 A/duck/Anhui/1/2006* 2.3.4 Shi et al., 2022 A/Anhui/1/2005 A/Waterfowl/Korea/S57/2016 2.3.4.4 Kuruppuarachchi et al., 2022 A/duck/France/161108 h/2016* 2.3.4.4b Grasland et al., 2023 A/mallard/Georgia/DT093 82/2017 A/green-winged teal/Egypt/877/2016* 2.3.4.4b Goma et al., 2019 A/Mute Swan/Croatia/102/2016 A/whooper swan/Shanxi/4-1/2020* 2.3.4.4b Shi et al., 2022 A/duck/Cambodia/f4k241 D3/2021 A/chicken/ME-2018* 2.3.4.4b Ibrahim et al., 2022 A/mallard/Georgia/DT093 82/2017 A/gyrfalcon/Washington/401886/2014-like* 2.3.4.4c Bertran et al., 2017 A/gyrfalcon/Washington/ 41088/6/2014 A/chicken/Guangzhou/3/2013 2.3.4.4g Shi et al., 2022 rgES3/2.3.4.4h 2.3.4.4h Kang et al., 2022 A/duck/Guanzou/S4184/2017 2.3.4.4h Shi et al., 2022 A/duck/Fujian/S1424/2020 2.3.4.4h Shi et al., 2022 OFFLU avian influenza matching report, October 2023 Page 18 Hemagglutination Inhibition (HI) assay Antigenic properties of avian influenza A H5 viruses were characterised using established harmonised approaches to HI assays at APHA and IZSVe using chicken mono-specific antisera. These HI assays were conducted as described in the WOAH Manual of Diagnostic Tests and Vaccines for Terrestrial Animals (WOAH, 2023) with a modification in the steps involving the incubation of chicken red blood cells with antigens (incubation time of 1 hour at a temperature of 4°C). The antigen containing 4 hemagglutinating units was confirmed by back titration. In the HI assay, sera were diluted two-fold from a starting dilution of 1:2 and tested in duplicate. Positive antisera and negative control sera obtained from specific pathogen free chickens were also tested for each assay. Antigenic Cartography Antigenic cartography methods as described by Smith et al., (2004), were previously used to characterise equine, swine and human influenza A viruses and also avian H5 HPAI and LPAI viruses (Smith et al., 2008; Lewis et al., 2011; Lorusso et al., 2010). Cartography was used to quantify and visualise the HI data and to characterise the relationships between surrogate vaccine seed strains and circulating strains. Map testing determined that the data were most robustly visualised in 3 dimensional antigenic maps, as described by Lewis et al., (2021). Maps were viewed and analysed in R Studio using the Racmacs package version 1.1.35). Inconsistent HI titers were excluded from analysis. For information regarding map generation and interpretation please contact an appropriate reference laboratory. Genetic analysis was carried out using manually curated datasets using Mafft v7.490 (Katoh and Standley, 2013) to align nucleotide and amino acid sequences using default settings. Alignments were trimmed from the H5 start codon to the start of the cleavage site. Phylogenetic analysis was undertaken generating maximum likelihood phylogenetic trees in iqtree2 V2.2.2.6 (Nguyen et al., 2015) to infer H5 clade designations. Strains in antigenic maps were coloured according to viral clades. Treetime v0.9.2 was used to infer ancestral sequences Sagulenko et al., (2017). The NADC IAV bioinformatics toolkit (Zhang et al., 2017) was used to analyse and compare sequences for differences in putative antigenic sites. OFFLU avian influenza matching report, October 2023 Page 19 Chicken antisera panel Table 2: Viruses used to generate post infection chicken antiserum antisera. Clade information is included. Asterisks (*) denote antigens which have been used as a surrogate representative of a vaccine antigen. Virus Isolate Name Clade A/Vietnam/1194/2004/1* 1 A/turkey/turkey/2005* 2.2 A/chicken/Nepal/T360/2014* 2.3.2.1a A/mynah/Indonesia(AustriaQ)/13064792-010/2013* “2.3.2.1c” A/Anhui/1/2005* (RG) 2.3.4 A/chicken/Czech Republic/1175-1-20VIR465-1/2020 2.3.4.4b A/chicken/Bulgaria/722-1-22VIR778-1/2021 2.3.4.4b A/mallard/Georgia/DT09382/2017* 2.3.4.4b A/mute Swan/Croatia/102/2016* 2.3.4.4b A/duck/Cambodia/f4k241D3/2021* 2.3.4.4b A/great skua/Scotland/B07779/2021 2.3.4.4b A/gyrfalcon/Washington/41088-6/2014* 2.3.4.4c OFFLU avian influenza matching report, October 2023 Page 20 Antigenic maps Figure 2: A 3 dimensional antigenic map showing the evolutionary relationships of H5 HPAI Gs/Gd lineage viruses. Each antiserum is represented by a cube: black cubes represent an antiserum raised using a virus which represents a surrogate vaccine seed strain, pink cubes represent antisera raised to quantify diversity. Antigens are represented as balls and are coloured by clade according to the key. Each square represents one antigenic unit. One antigenic unit is representative of a 2 fold different in HA assay titer. The corresponding table of antigenic distances can be found in table 4. OFFLU avian influenza matching report, October 2023 Page 21 Table of antigenic distances Table 3: A table of antigenic distances generated from the antigenic maps. One antigenic unit is equal to a twofold decrease in HA titer. Distances were coloured using a heat map. Vaccine antigens are labelled next to their surrogate chicken antisera. Antigens are ordered by clade, year and subtype. OFFLU avian influenza matching report, October 2023 Page 22 Main findings The focus was on clade 2.3.4.4b viruses from 2020 (7 viruses), 2021 (23 viruses) and 2022 (15 viruses). Two clade 2.3.2.1a viruses were also assessed. Note that one antigenic unit represents a twofold difference in HI assay titre. There were no antisera tested that were <3 antigenic units from genetic clade representative 2.3.4.4b and genetic clade representative 2.3.2.1a viruses. This indicates that it may be necessary to use multivalent H5 vaccines in places where multiple antigenically different viruses are co-circulating. 2.3.2.1a Contemporary viruses from Asia were >6 units from both genetic clade representative 2.3.4.4b and genetic clade representative 2.3.4.4c chicken antisera. Both antigens were <3 antigenic units from chicken antisera raised against genetic clade representative 2.3.2.1 viruses. Antigens were 4-6 antigenic units from clade 2.3.4 antigens. Both antigens were 5 antigenic units from the clade 2.2 antisera. Antigens were between 3-5 antigenic units from the clade 1 antisera. This indicates that inactivated vaccines using antigens from clades 2.3.4.4b, 2.3.4.4c, 2.3.2, 2.2, and 1 viruses may not provide protection against clade 2.3.2.1a viruses. 2.3.4.4b Clade 1, 2.2 and 2.3.4 antisera were at least 4 antigenic units from genetic clade representative 2.3.4.4b viruses. Clade 2.3.2.1 antisera was at least 3 antigenic units from genetic clade representative 2.3.4.4b viruses. This indicates that inactivated vaccines using these antigens would probably not provide sufficient protection against the contemporary genetic clade representative 2.3.4.4b viruses, as has been demonstrated in vaccine challenge studies. Some antigenic diversity is seen among the contemporary genetic clade representative 2.4.4.4b viruses. All were <6 antigenic units from genetic clade representative 2.3.4.4b chicken antisera with most viruses showing <3 antigenic units from genetic clade representative 2.3.4.4b chicken antisera. Antigens tested were >4 antigenic units from clade 1, 2.2 and 2.3.4 antisera. Most antigens tested were <4 antigenic units from the clade 2.3.4.4c chicken antiserum. All antigens tested have >3 antigenic units from clade 2.3.2.1 chicken antisera. Although there is some antigenic variation in clade 2.3.4.4b viruses, there did not appear to be significant changes when comparing isolates from 2020 to those from 2021 and 2022, suggesting the viruses have been relatively stable antigenically during this period (see Table 4). Table 4 provides the GMT values for antigenic difference against three reference 2.3.4.4b strain and one reference 2.3.4.4c strain. OFFLU avian influenza matching report, October 2023 Page 23 Table 4 Antigenic distance – Clade 2.3.4.4b viruses by year of isolation Year Count Antiserum A* (2.3.4.4b) Antiserum B* (2.3.4.4b) Antiserum C* (2.3.4.4b) Antiserum D* (2.3.4.4c) 2020 GMT 7 3.14 2.43 3.14 2.71 2020 Min 2 1 2 2 2020 Max 4 3 4 4 2021 GMT 23 3.13 2.04 2.74 3 2021 Min 1 0 1 1 2021 Max 6 4 4 6 2022 GMT 15 2.73 1.87 2.56 2.33 2022 Min 1 1 2 1 2022 Max 4 3 4 4 *Antiserum A A/Mute Swan/Croatia/102/2016 which is a surrogate for A/green-winged teal/Egypt/877/2016 (one of the earliest clade 2.3.4.4b viruses from Europe/Middle East) Antiserum B A/mallard/Georgia/DT09382/2017 which is a surrogate for A/chicken/ME-2018(an antigen used in some Egyptian-origin vaccines) Antiserum C A/duck/Cambodia/f4k241D3/2021 which is a surrogate for A/whooper swan/Shanxi/4-1/2020 (Chinese Re-14 antigen) Antiserum D A/gyrfalcon/Washington/41088/6/2014 (early clade 2.3.4.4c virus from North America) A/mallard/Georgia/DT09382/2017 appeared to be slightly closer to test isolates than other clade 2.3.4.4b reference strains. A/gyrfalcon/Washington/41088/6/2014 (clade 2.3.4.4c) appears similar antigenically to clade 2.3.4.4b viruses in terms of antigenic distance. The reason for the larger antigenic distance for the virus from Cambodia is being examined given it does not appear to be due to multiple changes at putative antigenic sites. Recommendations and next steps for OFFLU AIM 1. Viruses from a broader geographical representation should be tested including other clades which are circulating in poultry such as clade “2.3.2.1c”. 2. Sera should be raised against viral strains from other subtypes which represent vaccine antigen seed strains including H9 viruses. Lyophilised reagents (antisera and inactivated viruses) generated by APHA and IZSVe can be shared, upon request, with selected partners, targeting other WOAH and FAO international reference laboratories to enable the generation of comparative data assessing potential alteration in virus antigenicity using standardised reagents and methodologies under a defined quality framework. It should be noted however that careful between laboratory standardization needs to be undertaken before data from additional laboratories can be reliably represented in the analyses. Results will allow for continuous monitoring of OFFLU avian influenza matching report, October 2023 Page 24 antigenic changes in currently circulating viruses and inform the continuous expansion and updating of sera panels with global representation and relevance. Further to establishing harmonised methods across defined laboratories, antigenic data generated through this activity will complement data generated or presented by FAO or WOAH members and OFFLU partners including regional surveillance data, epidemiological data including evidence of vaccine breakdown and experimental studies. References Bertran, K., Balzli, C., Lee, D.-H., Suarez, D.L., Kapczynski, D.R., Swayne, D.E., 2017. Protection of White Leghorn chickens by U.S. emergency H5 vaccination against clade 2.3.4.4 H5N2 high pathogenicity avian influenza virus. Vaccine 35, 6336–6344. https://doi.org/10.1016/j.vaccine.2017.05.051 Boltz, D.A., Douangngeun, B., Sinthasak, S., Phommachanh, P., Midouangchanh, P., Walker, D., Keating, R., Khalenkov, A.M., Kumar, M., Webster, R.G., 2009. Field assessment of an H5N1 inactivated vaccine in chickens and ducks in Lao PDR. Arch Virol 154, 939–944. https://doi.org/10.1007/s00705-009-0385-x Marcia Oliveira Cavalcanti, Eric Vaughn, Ilaria Capua, Giovanni Cattoli, Calogero Terregino, Timm Harder, Christian Grund, Carlos Vega, Francisco Robles, Julio Franco, Ayub Darji, Abdel-Satar Arafa & Egbert Mundt 2016. A genetically engineered H5 protein expressed in insect cells confers protection against different clades of H5N1 highly pathogenic avian influenza viruses in chickens, Avian Pathology, DOI: 10.1080/03079457.2016.1250866 Fereidouni, S.R., Starick, E., Beer, M., Wilking, H., Kalthoff, D., Grund, C., Häuslaigner, R., Breithaupt, A., Lange, E., Harder, T.C., 2009. Highly pathogenic avian influenza virus infection of mallards with homoand heterosubtypic immunity induced by low pathogenic avian influenza viruses. PLoS One 4, e6706. https://doi.org/10.1371/journal.pone.0006706 Grasland, B. et al, 2023.Expérimentation de vaccination des canards mulards en élevage contre un virus influenza aviaire hautement pathogène A(H5N1) clade 2.3.4.4b ANSES online access : https://www.anses.fr/fr/system/files/Rapport_evaluation_excretion_Vaccin.pdf (accessed 11.2.23) Gomaa, M.R., Khalil, A.A., Kandeil, A., Sabir, J.S.M., Kayed, A., Moatasim, Y., El Saied, M.F., El-Safty, M.M., Kayali, G., Ali, M.A., 2019. Development of an effective contemporary trivalent avian influenza vaccine against circulating H5N1, H5N8, and H9N2 in Egypt. Poult Sci 98, 6289–6295. https://doi.org/10.3382/ps/pez385 Hadfield, J., Megill, C., Bell, S.M., Huddleston, J., Potter, B., Callender, C., Sagulenko, P., Bedford, T., Neher, R.A., 2018. Nextstrain: real-time tracking of pathogen evolution. Bioinformatics 34, 4121–4123. https://doi.org/10.1093/bioinformatics/bty407 Ibrahim, M., Zakaria, S., Bazid, A.-H.I., Kilany, W.H., Zain El-Abideen, M.A., Ali, A., 2021. A single dose of inactivated oil-emulsion bivalent H5N8/H5N1 vaccine protects chickens against the lethal challenge of both highly pathogenic avian influenza viruses. Comparative Immunology, Microbiology and Infectious Diseases 74, 101601. https://doi.org/10.1016/j.cimid.2020.101601 Indriani, R., Dharmayanti, N., 2014. Prototipe virus A/Duck/ Sukoharjo/Bbvw-1428-9/2012 subtipe H5N1 clade 2.3.2 sebagai vaksin pada itik lokal. JITV 19. https://doi.org/10.14334/jitv.v19i2.1044 Kang, Y.-M., Cho, H.-K., An, S.-J., Kim, H.-J., Lee, Y.-J., Kang, H.-M., 2022. Updating the National Antigen Bank in Korea: Protective Efficacy of Synthetic Vaccine Candidates against H5Nx Highly Pathogenic Avian Influenza Viruses Belonging to Clades 2.3.2.1 and 2.3.4.4. Vaccines (Basel) 10, 1860. https://doi.org/10.3390/vaccines10111860 Katoh, K., Standley, D.M., 2013. MAFFT multiple sequence alignment software version 7: improvements in performance and usability. Mol Biol Evol 30, 772–780. https://doi.org/10.1093/molbev/mst010 OFFLU avian influenza matching report, October 2023 Page 25 Kayali, G., Kandeil, A., El-Shesheny, R., Kayed, A.S., Maatouq, A.M., Cai, Z., McKenzie, P., Webby, R., Refaey, S.E., Kandeel, A., Ali, M.A., n.d. Avian Influenza A(H5N1) Virus in Egypt - Volume 22, Number 3— March 2016 - Emerging Infectious Diseases journal - CDC. https://doi.org/10.3201/eid2203.150593 Kuruppuarachchi, K.A.P.P., Jang, Y., Seo, S.H., 2022. Evaluation of Efficacy of Oil Adjuvanted H5N6 Inactivated Vaccine against Highly Pathogenic H5N6 and H5N1 Influenza Viruses Infected Chickens. Front Biosci (Landmark Ed) 27, 268. https://doi.org/10.31083/j.fbl2709268 Lewis, N.S., Banyard, A.C., Essen, S., Whittard, E., Coggon, A., Hansen, R., Reid, S., Brown, I.H., 2021. Antigenic evolution of contemporary clade 2.3.4.4 HPAI H5 influenza A viruses and impact on vaccine use for mitigation and control. Vaccine 39, 3794–3798. https://doi.org/10.1016/j.vaccine.2021.05.060 Lewis, N.S., Daly, J.M., Russell, C.A., Horton, D.L., Skepner, E., Bryant, N.A., Burke, D.F., Rash, A.S., Wood, J.L.N., Chambers, T.M., Fouchier, R.A.M., Mumford, J.A., Elton, D.M., Smith, D.J., 2011. Antigenic and Genetic Evolution of Equine Influenza A (H3N8) Virus from 1968 to 2007▿. J Virol 85, 12742–12749. https://doi.org/10.1128/JVI.05319-11 Lorusso, A., Vincent, A.L., Harland, M.L., Alt, D., Bayles, D.O., Swenson, S.L., Gramer, M.R., Russell, C.A., Smith, D.J., Lager, K.M., Lewis, N.S., 2011. Genetic and antigenic characterization of H1 influenza viruses from United States swine from 2008. J Gen Virol 92, 919–930. https://doi.org/10.1099/vir.0.027557-0 Nguyen, L.-T., Schmidt, H.A., von Haeseler, A., Minh, B.Q., 2015. IQ-TREE: a fast and effective stochastic algorithm for estimating maximum-likelihood phylogenies. Mol Biol Evol 32, 268–274. https://doi.org/10.1093/molbev/msu300 Palya, V., Tatár-Kis, T., Walkóné Kovács, E., Kiss, I., Homonnay, Z., Gardin, Y., Kertész, K., Dán, Á., 2018. Efficacy of a Recombinant Turkey Herpesvirus AI (H5) Vaccine in Preventing Transmission of Heterologous Highly Pathogenic H5N8 Clade 2.3.4.4b Challenge Virus in Commercial Broilers and Layer Pullets. J Immunol Res 2018, 3143189. https://doi.org/10.1155/2018/3143189 Sagulenko, P., Puller, V., Neher, R.A., 2018. TreeTime: Maximum-likelihood phylodynamic analysis. Virus Evol 4, vex042. https://doi.org/10.1093/ve/vex042 Shi, J., Zeng, X., Cui, P., Yan, C., Chen, H., 2023. Alarming situation of emerging H5 and H7 avian influenza and effective control strategies. Emerging Microbes & Infections 12, 2155072. https://doi.org/10.1080/22221751.2022.2155072 Smith, D.J., Lapedes, A.S., de Jong, J.C., Bestebroer, T.M., Rimmelzwaan, G.F., Osterhaus, A.D.M.E., Fouchier, R.A.M., 2004. Mapping the antigenic and genetic evolution of influenza virus. Science 305, 371– 376. https://doi.org/10.1126/science.1097211 Smith, D.J., de Jong, J.C., Lapedes, A.S., Jones, T.C., Russell, C.A., Bestebroer, T.M., Rimmelzwaan, G.F., Osterhaus, A.D.M.E., Fouchier, R.A.M., 2008. Antigenic Cartography of Human and Swine Influenza A (H3N2) Viruses, in: Novel and Re-Emerging Respiratory Viral Diseases. John Wiley & Sons, Ltd, pp. 32– 44. https://doi.org/10.1002/9780470770672.ch4 World Organisation for Animal Health, 2023. Terrestrial Manual Online Access, - World Organisation for Animal Health. URL https://www.woah.org/en/what-we-do/standards/codes-and-manuals/terrestrialmanual-online-access/ (accessed 6.21.23).