Equine Influenza A Virus

Overview and Taxonomy of Equine Influenza A Virus

Equine influenza A virus (EIV) is a highly contagious, acute respiratory pathogen of equids, representing one of the most significant viral threats to the global equine industry. As a member of the family Orthomyxoviridae, genus Influenzavirus A, EIV is characterized by a negative-sense, single-stranded, segmented RNA genome. This genomic architecture, comprising eight distinct segments, is the fundamental driver of the virus’s evolutionary capacity, enabling both antigenic drift, the gradual accumulation of point mutations, and, less frequently in equids, antigenic shift through genetic reassortment [11, 27]. The virus is classified based on the antigenic properties of its two major surface glycoproteins: hemagglutinin (HA) and neuraminidase (NA). To date, two subtypes have been definitively associated with endemic disease in horses: H7N7 and H3N8 [8, 11, 12]. The H7N7 subtype, first isolated in Prague in 1956 (A/equine/Prague/1/56), was historically significant but is now widely considered to be extinct in circulating equid populations, with no confirmed isolations for several decades [8, 26]. The last substantial evidence of H7N7 activity was serological detection in unvaccinated horses in the 1990s, and subsequent global surveillance has failed to identify its continued circulation, leading to its removal from vaccine strain recommendations by the World Organisation for Animal Health (WOAH) [8, 26]. Consequently, the contemporary global burden of equine influenza is exclusively attributable to the H3N8 subtype, which emerged in the early 1960s and has since undergone extensive evolutionary divergence [8, 11].

The taxonomy of contemporary H3N8 EIV is defined by a hierarchical phylogenetic structure rooted in two major lineages: the Eurasian lineage and the American lineage [8, 11]. The American lineage, which has become globally dominant, underwent a critical evolutionary radiation in the late 1980s, giving rise to the Kentucky, Argentina, and Florida sublineages [8, 22]. The Florida sublineage, in particular, has been the predominant circulating lineage worldwide since the early 2000s. In 2002, a significant divergence event within this sublineage produced two distinct, co-circulating clades: Florida Clade 1 (FC1) and Florida Clade 2 (FC2) [8, 16, 17]. This bifurcation has profound implications for vaccine efficacy and epidemiological tracking. FC1 viruses have historically been dominant in the Americas, but have demonstrated a remarkable capacity for transcontinental spread, causing major epizootics in Europe, Asia, and the Middle East in recent years [1, 2, 5, 10, 15]. For instance, the incursion of FC1 into Europe in late 2018, after a near-decade of absence, resulted in widespread outbreaks, displacing the previously dominant FC2 strains [10, 15]. Conversely, FC2 viruses have been more prevalent in Europe and Asia, although their detection in Europe has waned since 2018 [9, 10, 16]. The WOAH Expert Surveillance Panel (ESP) currently recommends that vaccines include both an FC1-like strain (e.g., A/equine/South Africa/4/2003 or A/equine/Ohio/2003) and an FC2-like strain (e.g., A/equine/Richmond/1/2007) to ensure broad coverage against circulating field viruses [10, 17, 18].

The molecular basis for this taxonomic classification and the virus’s continued success lies in the relentless evolution of its surface glycoproteins, particularly the HA. The HA protein is the primary target of the host’s protective humoral immune response, and as such, it is under intense selective pressure from population immunity, a phenomenon known as antigenic drift [8, 11]. This process is characterized by the accumulation of non-synonymous mutations in the HA1 domain, which encodes the globular head region containing the receptor-binding site and five major antigenic sites (A through E) [5, 18]. Mutations within these antigenic sites can alter the three-dimensional conformation of the protein, allowing the virus to evade pre-existing antibodies induced by prior infection or vaccination [8, 13]. This antigenic evolution is the primary driver of periodic vaccine breakdown, where outbreaks occur in populations with ostensibly high vaccination coverage [8, 13, 15]. A seminal example of this was the multifocal outbreak in Argentina in 2018, where 76% of affected horses had been vaccinated with a product containing a strain that was antigenically mismatched to the circulating FC1 virus, providing clear field evidence of vaccine failure [13]. Similarly, studies comparing the 2019 European FC1 strains (e.g., A/equine/Tipperary/1/2019) to the WOAH-recommended FC1 vaccine strain A/equine/South Africa/4/2003 demonstrated 2.5- to 6.3-fold reductions in neutralizing antibody titers, underscoring the continuous need for surveillance and vaccine strain updates [10]. Specific substitutions, such as T163I and N188T in antigenic site B, and N63D in antigenic site E, have been identified in post-2015 and post-2018 FC1 strains, respectively, highlighting the ongoing molecular evolution of the virus [5].

Beyond the HA, the other seven genomic segments contribute to the virus’s biology, host range, and pathogenicity. The NA glycoprotein, while also a target of the immune response, is less variable than HA but still undergoes antigenic drift [20]. The internal genes, which encode the polymerase complex (PB2, PB1, PA), the nucleoprotein (NP), the matrix proteins (M1, M2), and the non-structural proteins (NS1, NS2/NEP), are generally more conserved but are not static. For example, the PB1-F2 protein, a virulence factor encoded by an alternative reading frame in segment 2, has been shown to vary in length and function among EIV strains. A 2018 French FC1 strain, A/equine/Paris/1/2018, was found to encode a longer, 90-amino-acid variant of PB1-F2 with enhanced ability to disrupt mitochondrial membrane potential, potentially contributing to increased pathogenicity [5]. Furthermore, the NS1 protein is a key antagonist of the host interferon response, and truncations in its C-terminal domain have been associated with attenuation, a feature exploited in the development of live-attenuated influenza vaccines (LAIVs) [23, 24]. The NP is the primary target of the equine Mx1 restriction factor, an interferon-induced host protein that inhibits viral replication. Adaptive mutations in NP, such as G34S and H52N, have been identified that confer resistance to equine Mx1, demonstrating a dynamic host-pathogen arms race that shapes viral evolution [4].

The segmented nature of the influenza genome also allows for reassortment, a process where two different influenza viruses co-infect a single cell and exchange gene segments. While less common in equids than in swine or avian species, reassortment has been documented and is a critical mechanism for generating novel strains. A notable example is the 2005 Italian outbreak strain (A/equine/Bari/2005), which was identified as a multi-reassortant virus, possessing HA and M genes from the FC2 lineage while the remaining six segments originated from the Eurasian lineage [21]. This event highlights the potential for co-circulating lineages to generate novel genotypes. More significantly, reassortment has been implicated in the host-range shift of EIV to dogs. Phylodynamic analyses suggest that the H3N8 canine influenza virus (CIV), which emerged in Florida in the early 2000s, originated from a reassortant virus of the FC1 lineage [25]. This interspecies transmission event, which established a stable, enzootic canine lineage, is a stark reminder of the pandemic potential of influenza A viruses [3, 25]. The molecular adaptation involved specific mutations in the HA, such as W222L, which enhanced binding to canine-specific sialic acid receptors (Neu5Gc and sialyl Lewis X), overcoming a key host-range barrier [3]. The host range of EIV is not limited to horses and dogs; experimental and field evidence has demonstrated infection in cats [6], Bactrian camels [7], and serological evidence suggests potential spillover to other species [14]. However, the zoonotic risk to humans from equine H3N8 viruses is currently considered negligible, as human sera show very low levels of cross-reactive antibodies and the viruses replicate poorly in human airway epithelia [19, 27]. Nevertheless, the WOAH and the World Health Organization (WHO) continue to monitor EIV as part of a broader One Health surveillance framework for influenza viruses with pandemic potential.

In summary, the taxonomy of equine influenza A virus is a dynamic reflection of its evolutionary history, defined by the emergence, divergence, and global dissemination of the H3N8 subtype, particularly the Florida sublineage clades. The virus’s segmented genome and error-prone RNA polymerase fuel continuous antigenic drift, necessitating vigilant global surveillance and periodic updates of vaccine strains by bodies like the WOAH to maintain protective immunity in the world’s horse populations.

Molecular Pathogenesis of Equine Influenza A Virus

The molecular pathogenesis of equine influenza A virus (EIV) is a multifaceted process orchestrated by the coordinated actions of its eight negative-sense RNA segments, encoding at least ten distinct proteins. This intricate molecular machinery enables the virus to circumvent host defenses, replicate efficiently in the equine respiratory epithelium, and, under selective pressure, undergo antigenic drift that perpetuates its circulation even in vaccinated populations. A comprehensive understanding of these molecular events is critical for vaccine strain selection, therapeutic intervention, and pandemic preparedness, particularly given the virus's documented capacity for interspecies transmission to dogs, cats, and camels [3, 6, 7]. The World Organisation for Animal Health (WOAH) and the World Health Organization (WHO) recognize EIV as a pathogen of significant economic and zoonotic concern, underscoring the necessity of dissecting its molecular mechanism of disease.

Receptor Engagement and Host Tropism

The initial, defining molecular event in EIV infection is the binding of the haemagglutinin (HA) glycoprotein to sialic acid receptors on the surface of equine respiratory epithelial cells. Unlike human influenza viruses which preferentially bind α2,6-linked sialic acids, equine H3N8 viruses exhibit a strong tropism for α2,3-linked sialic acids, particularly those conjugated with N-glycolylneuraminic acid (Neu5Gc) [3, 32]. This receptor specificity is not merely a binary switch; it involves nuanced recognition of the glycan substructure. The presence of sialyl Lewis X (SLeX) motifs, in addition to the glycosidic linkage type, constitutes a critical molecular barrier for host range. Wen et al. (2018) elegantly demonstrated that a single amino acid substitution, W222L, in the receptor binding site of the HA of an equine-origin A(H3N8) virus dramatically increased binding avidity to Neu5Gc and SLeX motifs, which are abundant in the submucosal glands of the canine trachea [3]. This landmark study provided a molecular mechanism for the host-range shift of EIV from horses to dogs, a phenomenon that led to the establishment of a enzootic canine influenza virus lineage around 2002 [3, 25]. The HA protein's receptor binding domain consequently acts as a critical molecular gatekeeper, and mutations within this domain (e.g., in the 220-loop) can alter sialic acid linkage preference, potentially expanding tropism to other mammalian species [20]. The successful infection of cats and Bactrian camels with H3N8 EIV further highlights that the molecular barriers to cross-species transmission are not absolute and can be breached through specific adaptive mutations in the HA receptor binding pocket [6, 7, 27].

The Intracellular Molecular Battle: Replication, Restriction, and Countermeasures

Following receptor-mediated endocytosis and HA-mediated membrane fusion at low pH, the viral ribonucleoprotein (vRNP) complexes are released into the cytoplasm and trafficked to the nucleus, where transcription and replication occur. The viral RNA-dependent RNA polymerase, composed of PB1, PB2, and PA subunits, is a primary molecular arbiter of host range and pathogenesis. Codon usage bias and evolution of the polymerase genes (PB1, PB2, PA) are heavily influenced by natural selection pressures exerted by the equine host, with specific lineage-defining amino acid substitutions identified in Florida clade 2 viruses circulating in Europe and Asia since 2007 [31]. The PA subunit, in particular, appears to be under less translational pressure compared to the PB1 and PB2 subunits, suggesting differential contributions to host adaptation [31].

The host cell mounts a formidable intrinsic antiviral defense, and EIV has evolved sophisticated countermeasures. The equine Mx1 protein (eqMx1), an interferon-inducible dynamin-like GTPase, restricts influenza A virus replication by targeting the viral nucleoprotein (NP). Fatima et al. (2019) demonstrated that eqMx1 is uniquely specialized, effectively inhibiting the polymerase activity of non-equine influenza viruses but having a more restricted capacity against equine-origin strains [4]. Crucially, the NP of EIV has acquired adaptive mutations that confer resistance to eqMx1. The specific substitutions G34S and H52N in the H3N8JL89 strain NP were shown to substantially reduce Mx1 restriction, representing a molecular arms race where the virus evolves to evade this host restriction factor [4]. This dynamic is a powerful driving force in NP evolution and may serve as a robust indicator of pandemic potential.

Another frontline restriction factor is tetherin (BST-2/CD317), which physically tethers budding virions to the cell surface, preventing their release. Equine tetherin (eqTHN) exhibits remarkably high restriction activity against influenza A viruses, even more so than human tetherin. The shorter cytoplasmic tail of eqTHN contributes to this heightened activity [30]. In response, EIV employs its HA and NA glycoproteins as antagonists. Wang et al. (2018) identified that specific amino acids in HA (13T and 49L) and NA (32T and 80V) of a human pandemic H1N1 strain were sufficient to overcome eqTHN restriction when expressed in the context of a recombinant virus [30]. This demonstrates that the HA and NA of influenza A viruses, while primarily mediating receptor binding and release, also serve as critical countermeasures against host restriction factors, a key molecular battleground determining species susceptibility.

The non-structural protein NS1 is a master regulator of the host antiviral response, functioning primarily to suppress type I interferon (IFN-α/β) induction. NS1 achieves this by sequestering double-stranded RNA (dsRNA) to prevent activation of RIG-I and by directly inhibiting the 2'-5' oligoadenylate synthetase (OAS)/RNase L pathway and protein kinase R (PKR) [24]. The NS1 protein of equine H3N8 viruses often carries a naturally truncated C-terminus, which has significant implications for virulence. Na et al. (2016) demonstrated that a recombinant virus bearing the NS gene from a Korean H3N8 EIV strain, which encoded a truncated NS1, was dramatically attenuated in mice, inducing no weight loss or histopathological lesions. This attenuation was associated with significantly lower levels of pro-inflammatory cytokines (IL-6, CCL5, IFN-γ), suggesting that the full-length NS1 C-terminal domain is a critical virulence determinant that exacerbates the host inflammatory response and contributes to lung pathology [23, 24]. The truncated NS1 likely impairs the virus's ability to efficiently subdue the host's innate immune response, thereby limiting pathogenesis.

The PB1-F2 protein, a small protein encoded by an alternative reading frame of the PB1 segment, is another potent virulence factor. Kleij et al. (2024) identified that a 2018 French EIV strain (A/equine/Paris/1/2018) encoded a longer variant of PB1-F2 (90 amino acids) compared to other Florida clade 1 strains (81 amino acids) [5]. This extended PB1-F2 variant exhibited enhanced abilities to depolarize the mitochondrial membrane potential (ΔΨm) and permeabilize synthetic membranes. This mitochondrial targeting activity is a key molecular mechanism by which PB1-F2 promotes apoptosis and contributes to the exacerbated pathogenicity observed in that particular outbreak [5]. This finding underscores that virulence can be encoded not only in the major surface glycoproteins but also in subtle variations within accessory proteins.

Molecular Determinants of Secondary Bacterial Infection

The molecular pathogenesis of EIV extends beyond direct viral cytopathology to include a profound predisposition to secondary bacterial pneumonia, most commonly caused by Streptococcus equi subsp. zooepidemicus. The molecular basis for this lethal synergy is rooted in the virus's tropism for ciliated epithelial cells and goblet cells. Muranaka et al. (2012) demonstrated that experimental EIV infection in horses leads to extensive degeneration and necrosis of ciliated epithelia and a significant reduction in goblet cell numbers by post-infection days 2 and 3 [28]. This histological damage translates into a functional loss of mucociliary clearance, the primary physical defense mechanism of the respiratory tract. The resulting accumulation of mucus and compromised clearance provide a nutrient-rich environment and a permissive niche for bacterial colonization and proliferation. The presence of bacterial DNA was confirmed in lung tissue only from day 7 onwards, coinciding with the histological emergence of suppurative bronchopneumonia [28]. Furthermore, Yamanaka et al. (2006) showed that treatment with oseltamivir phosphate, an NA inhibitor, not only reduced viral shedding and pyrexia but also significantly reduced the bacterial counts of S. zooepidemicus in bronchoalveolar lavage fluid [29]. This strongly suggests that reducing viral replication, and thereby the extent of epithelial damage, directly mitigates the risk of secondary bacterial invasion, confirming that viral virulence is the molecular driver of this polymicrobial disease.

Antigenic Drift as a Molecular Mechanism of Immune Evasion

The relentlessness of EIV as a veterinary pathogen is largely attributable to its capacity for antigenic drift, the gradual accumulation of amino acid substitutions, particularly in the globular head domain of HA which contains major antigenic sites (A, B, C, D, and E). These sites are the primary targets of neutralizing antibodies. Molecular surveillance has repeatedly documented that circulating EIV strains possess numerous amino acid differences compared to WOAH-recommended vaccine strains. For instance, the 2019 outbreak strains in Europe (A/equine/Tipperary/1/2019 and A/equine/Essex/1/2019) showed a 2.5- to 6.3-fold reduction in neutralization by antisera raised against the vaccine strains A/equine/South Africa/4/2003 and A/equine/Ibaraki/1/2007 [10]. This molecular change has direct clinical consequences, leading to vaccine breakdown as witnessed in multifocal outbreaks in vaccinated horse populations in Argentina in 2018 and in France in late 2018 [13, 15].

Specific substitutions have been pinpointed as key molecular events driving antigenic change. The T163I substitution in antigenic site B of the HA was identified in a French 2018 strain (A/equine/Paris/1/2018) [5]. The N188T mutation, also in antigenic site B, has become fixed in post-2015 FC1 strains [5]. In FC2 strains circulating in Egypt, a specific cluster of substitutions was identified that grouped them within the FC2-144V subgroup, distinct from the vaccine strains used locally [9]. The Saudi Arabian 2019 strains revealed ten amino acid substitutions in the HA1 domain alone compared to clade 1 vaccine strains [1]. This constant, low-level antigenic variation accumulates to the point where vaccine-induced antibodies are no longer able to effectively neutralize field viruses, even when the vaccine is regularly updated. The molecular evolution of the HA is thus a direct and primary driver of periodic epizootics, highlighting the need for continuous global molecular surveillance to guide vaccine strain selection as per WOAH expert surveillance panel recommendations [8, 34].

Molecular Basis of Interspecies Transmission and Reassortment

The segmented nature of the influenza A virus genome renders EIV susceptible to genetic reassortment, a process where co-infection of a single cell with two different influenza viruses can generate a novel progeny virus with a new constellation of gene segments. This is a powerful molecular mechanism for rapid evolution and host-switching. Evidence for reassortment in EIV is significant. The H3N8 canine influenza virus originated from a reassortant virus of the Florida-1 clade [25]. In Pakistan, phylogenetic analysis of the NP and M internal genes from 2015-2016 equine isolates revealed high nucleotide identity (99.7-100%) with an avian H7N3 virus (A/avian/Pakistan/2004), suggesting that natural reassortment between equine and avian influenza viruses occurred in this region, likely facilitated by mixed farming systems [33]. Furthermore, a 2005 strain from Italy (A/equine/Bari/2005) was characterized as a multi-reassortant, possessing HA and M genes from the Florida 2 sublineage while clustering with Eurasian lineage viruses in its other six segments [21]. These findings demonstrate that the equine host can act as a mixing vessel, allowing the generation of novel genotypes through reassortment with avian strains. This capacity for molecular exchange poses a constant threat of introducing new genes (e.g., a novel NA subtype or a more virulent polymerase complex) into the circulating equine H3N8 backbone, potentially leading to unpredictable shifts in pathogenicity, host range, or transmissibility. The continued surveillance of the full genome, beyond just HA and NA, is therefore a molecular necessity for pandemic preparedness [5, 14, 27].

Epidemiology and Global Outbreak Dynamics

The global epidemiology of equine influenza A virus (EIV) is a complex tapestry woven from viral evolution, host population dynamics, international commerce, and the variable application of vaccination strategies. While the disease was first recognized in the mid-20th century, the modern landscape is dominated by the H3N8 subtype, with the original H7N7 strain having vanished from circulation in the late 20th century [8, 11, 12, 26]. The contemporary period, particularly from the early 2000s onward, has been defined by the emergence and global dissemination of the Florida sublineage, which has split into two distinct clades, Clade 1 (FC1) and Clade 2 (FC2), that now exhibit divergent geographic distributions and epidemiological behaviors [8, 11, 42]. The World Organisation for Animal Health (WOAH) Expert Surveillance Panel (ESP) has established a framework for monitoring these circulating strains and providing annual recommendations for vaccine strain updates, underscoring the global regulatory and economic importance of this pathogen.

Historical Lineage Dynamics and the Florida Sublineage Emergence

The evolutionary trajectory of EIV H3N8 is marked by a critical divergence from the earlier Eurasian and American lineages [16, 22]. Following the emergence of the Florida sublineage around 2002, the virus underwent a clade separation that has fundamentally shaped outbreak patterns. FC1, which evolved as a relatively cohesive lineage, became predominantly established in the Americas, while FC2, which itself further diverged into European and Asian sublineages, circulated primarily in Europe and parts of Asia [16, 22]. This dichotomy, however, has proven to be dynamic rather than static. Phylogeographic analyses of Asian isolates from 2007 to 2017 revealed that while FC1 outbreaks in Asia were caused by independent introductions from the Americas, FC2 strains in mainland Asia formed an autochthonous monophyletic group, suggesting endemic circulation within a local host population [16]. This pattern of independent introductions versus sustained local transmission has profound implications for control strategies, as endemicity requires a fundamentally different approach than managing sporadic incursions.

The emergence of the FC1 and FC2 dichotomy was accompanied by specific molecular signatures. Comprehensive analyses of the polymerase genes (PB2, PB1, PA) from 1963 to 2015 revealed group-specific consensus amino acid substitutions that drove the divergence of EIVs, with FC2 viruses exhibiting particularly notable variations that led to a phylogenetically distinct group originating from A/equine/Richmond/1/07 [31]. Furthermore, the polymerase genes demonstrate a weak codon usage bias, with natural selection being the dominant evolutionary force rather than mutation pressure, indicating that the virus is under strong adaptive constraints imposed by the equine host [31]. This selective pressure extends to host-specific restriction factors; for instance, equine Mx1 protein targets the viral nucleoprotein (NP), and adaptive mutations in NP (e.g., G34S and H52N) can confer resistance to this innate immune factor, suggesting a continuous arms race between the virus and equine host defenses that shapes viral evolution at a population level [4].

The 2018–2019 Transatlantic Shift: FC1 Invasion of Europe

One of the most dramatic epidemiological events in recent EIV history occurred between late 2018 and 2019, when FC1 strains, which had been almost exclusively circulating in the Americas, suddenly spread across Europe, causing widespread outbreaks [10, 15, 46]. This incursion was particularly significant because France, which had a well-established surveillance network (the RESPE network), had reported no clinical EI cases between 2015 and late 2018, largely attributed to high vaccination coverage (87.6% of the tested population had antibody titers above the clinical protection threshold) [15, 46]. The FC1 incursion demonstrated that even a well-vaccinated population is vulnerable to antigenic drift. Serological studies using a DIVA (Differentiating Infected from Vaccinated Animals) test, possible due to the predominant use of a recombinant canarypox-vectored vaccine in France, confirmed that the pre-2018 immunity was almost entirely vaccine-induced, with no evidence of natural infection in young horses [15]. The 2018–2019 FC1 viruses showed significant antigenic differences from the WOAH-recommended FC1 vaccine strains (e.g., A/equine/South Africa/4/2003 and A/equine/Ohio/2003), with virus neutralization tests using post-infection horse antisera demonstrating 2.5- to 6.3-fold reductions in antibody titers against the emerging strains [10]. Since this incursion, FC2 viruses have not been detected in Europe since 2018, suggesting that the FC1 lineage may have outcompeted FC2 in the European equid population [10]. This event serves as a paradigm for how antigenic drift can precipitate a sudden shift in global EIV epidemiology, overriding established vaccination programs.

Global Outbreak Patterns and the Role of International Horse Movement

The movement of horses, particularly for racing and breeding, is the single most critical factor driving the global dissemination of EIV. The 2012 South American outbreak is a textbook case: first reported in Chile, the virus then spread to Brazil, Uruguay, and Argentina within three months, affecting both vaccinated and unvaccinated animals [20]. Phylogenetic analyses of the hemagglutinin (HA) and neuraminidase (NA) genes demonstrated that the Brazilian isolate clustered with contemporary viruses from the USA, strongly implicating the international movement of horses as the introduction mechanism [20]. Similarly, the 2018 multifocal outbreak in Argentina, which affected horses vaccinated with a product containing the WOAH-recommended FC1 strain, provided field evidence of vaccine breakdown, with 76% of affected horses having a documented vaccination history [13]. This outbreak was characterized by multiple amino acid changes in antigenic sites of the HA protein, reinforcing the notion that vaccine strain updates are not merely a theoretical exercise but a practical necessity for maintaining herd immunity [13].

The pattern of introduction and subsequent local spread is repeated across continents. In Asia, the 2011 epizootic in Mongolia, which affects approximately 2.1 million horses about every decade, has been linked to the introduction of viruses from international sources [36]. Whole genome sequencing of viruses from a 2019 outbreak in Saudi Arabia revealed that the strains were nearly identical to those circulating in the USA in 2019, belonging to FC1, with no evidence of reassortment but with significant amino acid substitutions in the HA signal peptide and HA1/HA2 domains compared to vaccine strains [1]. In Malaysia, a 2015 outbreak in racehorses was traced to recently imported horses from New Zealand, with the causal virus demonstrating ≥99% HA and NA gene homology with contemporary USA and Japanese strains [18, 49]. The Malaysian outbreak was controlled through rapid measures including cancellation of racing and movement restrictions, underscoring the efficacy of such interventions when implemented early [49]. Turkey experienced its first identified EIV outbreak in 2013 in racehorses, with whole-genome sequencing and antigenic analysis confirming FC1 [50]. The first detection of FC1 in the Maghreb region occurred in Algeria in 2021, where a decade-long gap in reported outbreaks was broken by an epizootic that spread across five provinces, with serological testing revealing 49.3% positivity in the sampled population [52]. In Egypt, an outbreak among Arabian mares at a national race event in 2018 yielded the first detection of FC2-144V subgroup in the Middle East, with the viruses showing 14 and 13 amino acid differences compared to vaccine strains used locally, highlighting a dangerous antigenic mismatch [9]. These examples collectively demonstrate that EIV respects no borders and that the global equine industry's interconnectedness creates a network of vulnerability.

Interspecies Transmission and Host Range Expansion

An increasingly important dimension of EIV epidemiology is its documented capacity for interspecies transmission, which has implications for both animal health and, theoretically, public health. The most significant host-range shift occurred with the emergence of H3N8 canine influenza virus (CIV) from EIV, first detected in dogs in 2004. Phylodynamic analyses date this host-range shift to approximately 2002, with the donor virus being a reassortant of the circulating FC1 EIV [25]. This event is unique because it represents a direct mammalian-to-mammalian host-range shift without an intermediate avian host. Once established in dogs, the virus spread efficiently, evolved into multiple clades, and underwent both intra- and inter-lineage reassortment, demonstrating that dogs can act as "mixing vessels" for influenza viruses [25]. The molecular underpinnings of this shift include a specific W222L mutation in the HA protein, which enhanced binding to canine-specific receptors containing sialyl Lewis X and N-glycolylneuraminic acid (Neu5Gc) motifs, which are abundant in the submucosal glands of the dog trachea [3].

Experimental infections have demonstrated that EIV can also infect cats, with all 14 experimentally infected cats showing clinical signs, shedding virus, and transmitting the infection to contact cohorts, raising the possibility of feline establishment [6]. Furthermore, natural infection of Bactrian camels in Mongolia has been documented, with phylogenetic analysis of a 2012 camel isolate (A/camel/Mongolia/335/2012[H3N8]) showing it is closely related to equine H3N8 viruses, likely representing natural horse-to-camel transmission [7]. In large-scale donkey farms in China, serological surveys have shown a 32.5% seropositivity rate, and RT-PCR detection of the virus confirmed active circulation, with the isolate showing 99.77% homology to equine strains from Northeast China [47]. The impact of EIV on the equine upper respiratory tract microbiome is also an emerging area of concern; studies in Kazakhstan have shown that environmental factors strongly shape the microbiome composition, but the presence of EIV may influence the abundance of specific taxa, potentially predisposing animals to secondary bacterial infections [38]. This is clinically relevant because secondary bacterial bronchopneumonia, often caused by Streptococcus equi subsp. zooepidemicus, is a major complication of EIV infection, and the use of antiviral agents like oseltamivir phosphate has been shown to reduce both viral shedding and secondary bacterial counts in bronchoalveolar lavage fluid [28, 29]. The risk of spillover to humans is considered low, with the World Health Organization (WHO) and WOAH maintaining surveillance but noting that equine influenza viruses have not established sustained human transmission. Serological surveys in Belgium found that only ≤12% of human sera had antibodies against equine H3 IAVs, and those viruses showed only intermediate replication efficiency in human airway epithelial cells [19]. Nevertheless, the potential for mixing vessel hosts, including pigs and perhaps dogs, to facilitate reassortment events that could generate zoonotic strains remains a concern that warrants continued vigilance [27].

Vaccination Coverage, Vaccine Failure, and the Enigma of Protection

The relationship between vaccination and outbreak dynamics is not straightforward. A systematic review and meta-analysis of EIV vaccine studies published up to 2020 calculated a mean vaccine efficacy (VE) of 50.03% (95% CI: 23.35–76.71%) for commercial vaccines, with a range of 0 to 100% [35]. This wide variability underscores the fact that vaccine efficacy is highly dependent on the match between vaccine and circulating strains, the vaccination schedule, and the immunological status of the host. Complete protection from virus shedding was achieved in only five of the 38 studies included in the meta-analysis, emphasizing that the primary goal of vaccination in the field is not to prevent infection but to reduce shedding duration and clinical severity, thereby lowering transmission potential [35]. This is supported by field studies in Argentina, where despite high vaccination coverage with a product containing the WOAH-recommended strain, a multifocal outbreak occurred, and the vaccine’s failure was attributed to antigenic drift [13]. Conversely, in the United States, a voluntary biosurveillance program analyzing 9,740 nasal swabs from 2008 to 2021 found that equids under 9 years of age with a recent history of travel were at highest risk for EIV qPCR positivity, suggesting that even in a population with some degree of vaccination, movement and age are dominant risk factors [41]. The program noted a 9.9% EIV positivity rate, with seasonal peaks in winter and spring [41].

The concept of herd immunity in equine populations is nuanced by the use of different vaccine technologies. In France, where 90% of the population had antibody titers above the protection threshold, the incursion of FC1 still caused outbreaks because the protection threshold is defined against clinical disease, not infection [15, 46]. Indeed, 60% of the vaccinated population had antibody titers that would not prevent virus shedding, allowing transmission even as clinical signs were suppressed [46]. This phenomenon, where vaccines protect the individual from severe disease but fail to prevent subclinical infection and onward transmission, is a critical epidemiological driver. It means that even in well-vaccinated populations, the virus can circulate cryptically, mutating and adapting, until a sufficiently drifted variant emerges that overcomes the protection against disease as well. The use of reverse genetics (RG) technology to rapidly update vaccine strains is a promising approach to address this; RG-derived inactivated vaccines have been shown to provide protective efficacy comparable to wild-type virus-derived vaccines in challenge studies, and the RG platform allows for rapid swapping of HA and NA genes to match emerging strains [37, 44]. The WOAH ESP’s annual strain selection process is therefore not merely a bureaucratic exercise but a critical form of global health governance that attempts to keep pace with the virus’s evolution.

Regional Endemicity and Neglected Surveillance Gaps

While much of the world has some level of EIV surveillance, significant gaps remain, particularly in Africa and parts of Asia. In Nigeria, serological surveys have consistently demonstrated widespread EIV circulation, with a 51.3% seroprevalence in polo horses from the 2021 Jos Polo Tournament (representing multiple states) [40] and 60.9% in horse stables in Kaduna [51]. In the Southwestern states of Nigeria, a 60% seropositivity was found, with higher rates in male horses and those aged 11-15 years, indicating that infection is enzootic rather than epizootic [45]. In Senegal, a 2019 epizootic in donkeys in the Foundiougne department revealed that lack of access to veterinary care and the wandering of donkeys were significant risk factors for infection, with odds ratios of 2.0 and 2.06, respectively [43]. This highlights a critical issue: in many resource-limited settings, even basic biosecurity measures and veterinary oversight are absent, allowing the virus to circulate unchecked. In Kazakhstan, serological analysis confirmed EIV circulation across multiple regions, and whole-genome sequencing of strains from 2012 revealed they belonged to FC2, but the overall prevalence and impact remain poorly characterized due to limited surveillance [38, 39].

Mongolia presents a particularly interesting case, where EIV epizootics occur approximately every 10 years, affecting up to 2.1 million horses [36]. Yet, between epizootics, active surveillance in 2016–2017 detected sporadic qRT-PCR positive horses (9 out of 680 swabs) across four aimags, suggesting low-level enzootic circulation, even though virus isolation in cell culture was unsuccessful [48]. None of the 131 surveyed herder households had vaccinated their horses, indicating that the virus persists in a naïve population through a combination of low-level transmission and periodic explosive outbreaks [48]. This situation, combined with the mixing of domestic and wild equid herds and the periodic prevalence of highly pathogenic avian influenza, has led to the characterization of Mongolia as a potential "hot spot" for novel EIV emergence [14, 48]. The presence of EIV in Bactrian camels, as documented in Mongolia, adds another layer of complexity to the ecological niche of the virus [7]. The WOAH and FAO have highlighted the need for enhanced surveillance in such regions, as they represent the most likely sites for the emergence of novel reassortants with pandemic potential.

The Role of Wind and Aerosol Transmission in Outbreak Dynamics

Recent epizootics have drawn attention to the importance of non-contact transmission mechanisms. Historically, EIV was thought to spread primarily through direct contact or fomites, but analysis of the 2010–2011 outbreak in Australia and other studies suggest that wind-aided aerosol transmission may play a significant role over distances of at least several kilometers [14]. This has profound implications for quarantine and movement control strategies, as it implies that physical separation of horses may not be sufficient to prevent transmission under certain climatic conditions. The virus is transmitted via inhalation of aerosolized droplets, and once established in the respiratory tract, it rapidly destroys ciliated epithelial cells and goblet cells, leading to a loss of mucociliary clearance and secondary bacterial infections, which facilitates further shedding and environmental contamination [28, 42]. The efficiency of aerosol transmission is enhanced by high viral titers in nasal secretions, which can be shed for up to 8 days in unvaccinated horses [29]. This mode of transmission, combined with the movement of subclinically infected vaccinated horses, creates a dynamic where outbreaks can ignite and spread with astonishing speed, as seen in the rapid pan-continental spread of the FC1 lineage in 2018–2019.

Diagnostic Methods for Equine Influenza A Virus

The accurate and timely diagnosis of equine influenza A virus (EIV) is a cornerstone of effective outbreak management, epidemiological surveillance, and informed vaccine strain selection. The clinical signs of equine influenza, pyrexia, a harsh dry cough, serous nasal discharge that may become mucopurulent, anorexia, and lethargy [12, 40], are nonspecific and can be confounded by other respiratory pathogens, including equine herpesviruses (EHV-1, EHV-4), Streptococcus equi subsp. equi, and equine rhinitis viruses A and B [41]. Consequently, laboratory confirmation is not merely advisable but essential. The diagnostic arsenal available to the equine clinician and researcher has expanded considerably over the past two decades, transitioning from classical virological methods to a sophisticated array of molecular, serological, and advanced genomic techniques. The World Organisation for Animal Health (WOAH) recommends a suite of diagnostic approaches, and the selection of a specific method must be guided by the clinical context, the stage of infection, the purpose of testing (e.g., individual diagnosis versus population surveillance), and the available laboratory infrastructure.

Viral RNA Detection: The Reigning Gold Standard

Real-time reverse transcription polymerase chain reaction (RT-qPCR) has been established as the primary diagnostic tool for the detection of EIV, owing to its exceptional sensitivity, specificity, and rapid turnaround time [1, 2, 41]. This technique targets highly conserved regions of the influenza A genome, most commonly the matrix (M) gene, allowing for the detection of all influenza A virus subtypes, including both H3N8 and the now-extinct H7N7 [53, 58]. The assay is exquisitely sensitive, capable of detecting viral RNA from nasopharyngeal swabs even during the early or late stages of infection when viral loads may be low [32]. A large-scale voluntary surveillance program in the United States, which processed over 9,700 nasal swabs from 2008 to 2021, demonstrated a 9.9% EIV qPCR-positivity rate, highlighting its utility in monitoring endemic circulation and identifying risk factors such as age under nine years and recent travel history [41]. Similarly, outbreaks in Chile [2], Saudi Arabia [1], and Malaysia [49] were rapidly confirmed and characterized using RT-qPCR.

Conventional RT-PCR and multiplex RT-PCR formats have also been developed, offering a cost-effective alternative for laboratories that may lack real-time instrumentation. One notable multiplex approach, designed using dual priming oligonucleotide technology, successfully amplifies all eight genomic segments of the H3N8 virus in a single reaction, facilitating comprehensive genotypic analysis without the need for multiple individual assays [53]. Point-of-care (POC) molecular diagnostics, such as the cobas Influenza A/B & RSV test for the cobas Liat system, represent a significant advancement for field deployment. This system detects influenza A virus RNA within 20 minutes and has demonstrated sensitivity equivalent to that of laboratory-based RT-qPCR for equine strains, although a rate of invalid results (7.7% in one study) indicates a need for further optimization [58].

A crucial adjunct to nucleic acid detection is the pre-analytical concentration of viral particles. Novel sialoglyco particulates, engineered to display N-glycolylneuraminyl-α-(2→3)-N-acetyllactosamine (Neu5Gcα2,3LacNAc) residues, the preferred receptor for equine influenza virus [3], have been shown to dramatically increase the detection sensitivity of RT-qPCR. When applied to nasal swab samples from infected horses, these particulates captured the virus with high affinity, resulting in a significant increase in the number of detectable viral gene copies, particularly in samples with low viral loads [32].

Virus Isolation: The Timeless but Resource-Intensive Method

While molecular detection has largely supplanted virus isolation for routine diagnosis, the isolation of infectious virus remains indispensable for detailed antigenic and genetic characterization, vaccine strain development, and basic research. The traditional method involves inoculation of clinical specimens into the allantoic cavity of 9- to 11-day-old embryonated chicken eggs (ECEs). This system is well-established, and many field strains, including those from recent outbreaks in Argentina [13] and Chile [2], have been successfully isolated using this technique. However, isolation can be challenging, particularly when the initial viral load is low, or the sample quality is poor [55].

To enhance the success rate of isolation, researchers have refined the use of chorioallantoic membranes (CAMs). A study comparing the detection of EIV in amniotic/allantoic fluid (AF) versus CAM homogenates by RT-qPCR found that CAM homogenates were far more sensitive, allowing for the identification of viral replication even when the AF remained negative by both haemagglutination assay and RT-qPCR. This approach facilitates the targeted passaging of CAM homogenates, ultimately yielding high-titer virus stocks that would otherwise be missed [55]. Furthermore, the development of reverse genetics (RG) systems has revolutionized our ability to generate vaccine viruses and study viral biology. RG viruses, which combine the haemagglutinin (HA) and neuraminidase (NA) of contemporary field strains with the six internal genes of a high-growth donor virus like A/Puerto Rico/8/34 (PR8), exhibit superior growth properties in eggs, enabling rapid vaccine strain updates in response to antigenic drift [37, 44].

Serological Diagnostics: Unveiling Historical and Immunological Status

Serological assays are critical for determining prior exposure, evaluating vaccine-induced immunity, and conducting large-scale epidemiological studies. The haemagglutination inhibition (HI) test and single radial haemolysis (SRH) assay remain the workhorses for measuring antibody responses, particularly those directed against the HA protein, which is the primary target of protective immunity [40, 45, 54].

The HI test is a relatively simple, low-cost assay that quantifies the ability of serum antibodies to prevent the agglutination of red blood cells by the virus. It is the most widely used method for assessing post-vaccination antibody titers and for antigenic characterization of field strains against a panel of ferret antisera [20, 40]. However, the HI test has limitations, including inter-laboratory variability and the requirement for fresh red blood cells, which can be a logistical hurdle. The SRH assay is considered by many to be more reproducible, as it is less sensitive to the presence of non-specific inhibitors. It measures the zone of haemolysis produced by the diffusion of complement-fixing antibodies from a serum sample in an agarose gel containing virus-sensitized red blood cells [15, 54]. Large-scale serosurveillance studies, such as those conducted in France involving over 3,000 horses, have successfully used SRH to demonstrate that high vaccination coverage (87.6% of horses with titers above the clinical protection threshold) was correlated with an absence of clinical EIV for several years prior to 2018 [15, 46].

Enzyme-linked immunosorbent assays (ELISAs) offer advantages of high throughput and objective, automated readout. Commercial ELISAs targeting the influenza A nucleoprotein (NP) are invaluable for screening purposes. These NP-ELISAs have been employed in seroprevalence studies across Africa, revealing infection rates of 55.9% in Saudi Arabia [56], 60.9% in Nigeria [51], and 49.3% in Algeria [52]. Importantly, NP-ELISAs can serve a differentiating infected from vaccinated animals (DIVA) function when used in conjunction with vaccines that do not contain the NP antigen, such as the canarypox-vectored ProteqFlu® vaccine. In the French cohort, the combination of SRH and NP-ELISA confirmed that the strong antibody responses observed in young horses were due to vaccination alone, not natural infection, thus validating the absence of widespread EIV circulation [15, 46].

A more recent and sophisticated serological tool is the pseudotyped virus neutralization test (PVNT). This assay uses replication-incompetent lentiviral or retroviral particles pseudotyped with the EIV HA (and optionally NA) to measure functional neutralizing antibodies. The PVNT does not require the handling of live EIV, making it a safer and more versatile option. It has shown excellent agreement with both SRH and HI tests for detecting antibodies induced by vaccination and experimental infection, and it appears to be more sensitive, particularly when antibody levels are low [54, 57]. This enhanced sensitivity could prove invaluable for the early detection of waning immunity or for distinguishing low-level cross-reactive responses.

Advanced Genomic Characterization: Sequencing and Phylogenetics

Beyond mere detection, the modern diagnostic laboratory plays a pivotal role in tracking viral evolution and guiding vaccine policy. This is achieved through genomic sequencing, which has evolved from Sanger sequencing of single genes to next-generation sequencing (NGS) of the complete viral genome. Full-genome sequencing is now considered essential for identifying reassortment events, detecting the emergence of novel mutations in antigenic sites, and performing phylodynamic analyses that reveal global migration patterns [5, 16, 22].

The HA1 domain of the HA gene is the primary target for routine antigenic surveillance, as mutations in this region are responsible for antigenic drift. Recent studies have identified substitutions in antigenic sites B and E of post-2018 Florida clade 1 strains, such as T163I and N188T in site B and N63D in site E, which correlated with reduced neutralization by antisera raised against older vaccine strains [5, 10]. The NA gene is also sequenced to monitor for mutations that could affect antiviral susceptibility or antigenicity [18]. Long-read nanopore sequencing platforms, such as the Oxford Nanopore MinION, have emerged as a powerful, portable, and cost-effective tool for generating complete genomes in near real-time. This technology was successfully used to characterize the 2018 French outbreak strain, A/equine/Paris/1/2018, revealing an unusually long PB1-F2 protein (90 amino acids) with enhanced pro-apoptotic activity, a virulence marker that would have been missed by targeted HA-only sequencing [5]. Similarly, nanopore sequencing has been applied to archival formalin-fixed tissues, allowing for the retrospective genomic characterization of a 2005 Italian outbreak strain as a multi-reassortant between Florida clade 2 and Eurasian lineages [21]. These advanced genomic tools are not merely academic; they provide the critical, data-driven evidence required by the WOAH Expert Surveillance Panel (ESP) to recommend updates to vaccine strain composition, ensuring that vaccines remain relevant against a continuously evolving pathogen [10, 16].

Genomic Characterization and Antigenic Evolution

The genomic architecture of the equine influenza A virus (EIV) is a paradigm of adaptive evolution, driven by the interplay of a segmented RNA genome, error-prone RNA-dependent RNA polymerase, and intense selective pressure from host immunity and vaccine-induced antibodies. The H3N8 subtype, which has completely supplanted the H7N7 subtype (last isolated in the late 1970s and considered extinct in the field) [8, 26], serves as the primary model for understanding the molecular mechanisms underpinning equine influenza persistence and periodic vaccine failure. The virus possesses a negative-sense, single-stranded RNA genome comprising eight segments, encoding at least 11 proteins, including the surface glycoproteins hemagglutinin (HA) and neuraminidase (NA), the polymerase complex (PB2, PB1, PA, and the accessory PB1-F2), the nucleoprotein (NP), the matrix proteins (M1 and M2), and the non-structural proteins (NS1 and NEP/NS2) [5, 11, 24]. The evolutionary dynamics of these segments are not uniform; the HA gene, in particular, undergoes relentless antigenic drift, while internal genes may experience reassortment, host-specific adaptation, and codon usage bias shaped predominantly by natural selection [31, 60].

The Hemagglutinin Gene: Epicenter of Antigenic Drift and Host Range

The HA glycoprotein is the primary target of neutralizing antibodies and the principal determinant of viral attachment to sialic acid receptors. In equine H3N8 viruses, the HA is cleaved into HA1 and HA2 subunits, with the globular HA1 head domain containing the receptor-binding site (RBS) and five major antigenic sites (A, B, C, D, and E) that are the foci of immune-driven selection [5, 8]. The World Organisation for Animal Health (WOAH) Expert Surveillance Panel (ESP) has long recognized that the continuous accumulation of amino acid substitutions in these antigenic sites erodes vaccine efficacy, necessitating periodic updates of vaccine strains [10, 13, 15]. The Florida sublineage, which emerged around 2002, has since diverged into two major clades: Florida clade 1 (FC1) and Florida clade 2 (FC2) [8, 16]. This divergence is not merely phylogenetic; it reflects distinct antigenic trajectories and geographical circulation patterns. FC1 viruses have predominantly circulated in the Americas and have periodically erupted into Europe and Asia, while FC2 viruses have been more prevalent in Europe and Asia, with evidence of an autochthonous Asian sublineage that emerged around 2006 [16, 22].

Detailed genomic characterization of FC1 strains from the 2019 Saudi Arabian outbreak revealed a considerable number of amino acid substitutions in the HA protein relative to the WOAH-recommended FC1 vaccine strain A/equine/South Africa/4/2003. Specifically, 10 substitutions were identified in the HA1 domain and 4 in the HA2 domain, including changes in the signal peptide region [1]. Similarly, analysis of French FC1 strains circulating in 2018–2019 identified critical substitutions such as T163I in A/equine/Paris/1/2018 and N188T in post-2015 strains, both located within antigenic site B, a region known to be immunodominant in horses [5]. Furthermore, antigenic site E exhibited modifications in post-2018 strains, with the N63D substitution [5]. These changes are not silent; they have measurable antigenic consequences. Virus neutralization tests using equine antisera raised against the vaccine strains A/equine/South Africa/4/2003 and A/equine/Ibaraki/1/2007 demonstrated that antibody titers against the 2019 European FC1 isolates (A/equine/Tipperary/1/2019 and A/equine/Essex/1/2019) were 2.5- to 6.3-fold lower than titers against the homologous vaccine strains [10]. This reduction in cross-neutralization is a classic hallmark of antigenic drift and provides a molecular explanation for the vaccine breakdown observed in multifocal outbreaks in Argentina in 2018, where 76% of affected horses had been vaccinated with a product containing the WOAH-recommended strain [13].

The receptor-binding specificity of the HA is another critical axis of evolution. Equine influenza viruses preferentially bind to α2,3-linked sialic acids, which are abundant in the equine respiratory tract. However, the substructure of the receptor glycan, including the type of sialic acid (N-acetylneuraminic acid [Neu5Ac] vs. N-glycolylneuraminic acid [Neu5Gc]) and the presence of fucosylated motifs like sialyl Lewis X (SLeX), can impose barriers to interspecies transmission [3]. The landmark host-range shift of equine H3N8 to dogs, which led to the establishment of a canine influenza virus (CIV) lineage, was facilitated by a single amino acid substitution, W222L, in the HA. This mutation increased binding avidity to canine-specific receptors containing Neu5Gc and SLeX motifs, which are abundant in the submucosal glands of the dog trachea [3]. Phylodynamic analyses have dated this host-range shift to approximately 2002 and revealed that the emergent CIV was a reassortant virus derived from the circulating FC1 lineage [25]. This event underscores the zoonotic and epizootic potential of equine H3N8 viruses, which have also been shown to infect cats, Bactrian camels, and to replicate in human airway epithelial cells, albeit with limited seroprevalence in humans [6, 7, 19, 61]. The WOAH and the World Health Organization (WHO) continue to monitor these spillover events, as the equine H3N8 lineage possesses a theoretical, albeit low, pandemic risk [19, 27].

The Neuraminidase and Internal Genes: Reassortment, Virulence, and Host Restriction

While the HA gene is the primary target of vaccine-induced immunity, the NA glycoprotein also undergoes antigenic drift and contributes to viral fitness. The NA enzymatic activity cleaves sialic acid receptors, facilitating viral release and spread. Comparative analyses of FC1 and FC2 strains have identified lineage-specific amino acid substitutions in NA, some of which are located in antigenic regions [1, 13]. However, the internal genes, PB2, PB1, PA, NP, M, and NS, are equally important for understanding viral pathogenesis, host adaptation, and the potential for reassortment.

The polymerase genes (PB2, PB1, PA) are key determinants of host range and replication efficiency. Comprehensive codon usage bias analyses of polymerase genes from EIV isolates spanning 1963 to 2015 revealed that natural selection, rather than mutational pressure, is the dominant evolutionary force shaping these genes [31]. Lineage-specific consensus amino acid substitutions were identified that led to the divergence of EIVs into pre-divergent, Eurasian, and Florida sublineages. Notably, the FC2 viruses circulating in Europe and Asia since 2007 have undergone major variations in the polymerase genes, leading to the emergence of a phylogenetically distinct group originating from A/equine/Richmond/1/2007 [31]. The PB1-F2 protein, an accessory protein encoded by an alternative reading frame in the PB1 segment, is a virulence factor that can induce apoptosis by disrupting the mitochondrial membrane potential (ΔΨm). Strikingly, the French FC1 strain A/equine/Paris/1/2018 was found to encode a longer, 90-amino-acid variant of PB1-F2, compared to the 81-amino-acid variant typical of other FC1 strains. This longer variant exhibited enhanced abilities to abolish mitochondrial membrane potential and permeabilize synthetic membranes, suggesting that it may contribute to the increased pathogenicity observed during the 2018–2019 outbreak in France [5]. This finding highlights the importance of whole-genome sequencing, as virulence determinants can reside in segments beyond HA and NA.

The nucleoprotein (NP) is a target of host restriction factors, particularly the equine Mx1 protein (eqMx1). Equine Mx1 is an interferon-induced dynamin-like GTPase that inhibits influenza A virus replication by targeting the viral NP. Adaptive mutations in NP, such as G34S and H52N, have been shown to confer resistance to eqMx1, allowing the virus to evade this innate immune barrier [4]. The evolution of NP under this selective pressure is a dynamic process that can influence the pandemic potential of emerging strains. The matrix protein M2, encoded by segment 7, is the target of the antiviral drug amantadine, although resistance mutations can emerge rapidly. The non-structural protein NS1 is a multifunctional virulence factor that antagonizes the host interferon response. Naturally occurring truncations of the NS1 protein, such as those identified in a Korean H3N8 isolate (A/equine/Kyonggi/SA1/2011), have been associated with attenuation of virulence in mouse models, suggesting that the C-terminal domain of NS1 contains critical virulence determinants [23, 24].

Reassortment and Phylogeographic Dynamics

The segmented nature of the influenza genome allows for reassortment, a process that can generate novel genotypes with altered antigenicity, virulence, or host range. While the FC1 and FC2 clades have largely evolved as separate lineages without evidence of inter-clade reassortment in the HA and NA genes, reassortment involving internal genes has been documented [16]. A notable example is the 2005 Italian strain A/equine/Bari/2005, which was identified as an inter-lineage multi-reassortant. Phylogenetic analysis of all eight segments revealed that the HA and M genes clustered with the FC2 lineage, while the remaining six genes clustered with the older Eurasian lineage [21]. This finding demonstrates that reassortment between co-circulating lineages can occur and may contribute to the emergence of viruses with novel properties. Furthermore, evidence from Pakistan suggests that equine H3N8 viruses can reassort with avian influenza A viruses. The NP and M genes of Pakistani equine isolates from 2015–2016 showed 99.7–100% nucleotide identity with avian H7N3 viruses circulating in Pakistani poultry, indicating a reassortment event likely facilitated by mixed farming practices [33]. Such interspecies reassortment events are of significant concern to the Food and Agriculture Organization (FAO) and the WHO, as they could generate viruses with pandemic potential.

Phylogeographic analyses have revealed the global dissemination patterns of EIV. The South American continent has experienced multiple independent introductions of EIV from North America, with the most recent common ancestors for each outbreak estimated to have circulated in North America at least one year prior to their detection in South America [22]. The 2012 South American outbreak, which affected vaccinated and unvaccinated horses across Chile, Brazil, Uruguay, and Argentina, was caused by an FC1 virus closely related to contemporary US strains. Despite reports of vaccine breakdown, antigenic analysis using ferret antisera failed to identify significant antigenic drift, suggesting that other factors, such as inadequate vaccine coverage, poor biosecurity, and high-density horse populations, contributed to the rapid spread [20]. In Asia, a distinct phylogeographic pattern has emerged. While FC1 outbreaks in Asia are typically caused by independent introductions from the Americas, the FC2 strains circulating in mainland Asia have formed an autochthonous monophyletic group with a common ancestor dated to 2006, providing evidence of sustained endemic circulation within a local horse population [16]. This Asian FC2 lineage has developed a characteristic amino acid signature pattern in the HA1 protein, including changes in or near antigenic sites, which may have implications for vaccine efficacy in the region [9, 16]. The detection of FC2-144V subgroup viruses in Arabian mares in Egypt in 2018, and the subsequent introduction of FC1 into the Maghreb region of Algeria in 2021, illustrate the dynamic and expanding geographic range of these lineages [9, 52].

Antigenic Cartography and Implications for Vaccine Strain Selection

The practical consequence of genomic evolution is antigenic drift, which necessitates periodic updates of vaccine strains. The WOAH ESP reviews global surveillance data annually and makes recommendations for vaccine strain composition. The current recommendation includes an FC1 strain (e.g., A/equine/South Africa/4/2003-like or A/equine/Ohio/2003-like) and an FC2 strain (e.g., A/equine/Richmond/1/2007-like) [8, 17]. However, the emergence of new variants, such as the post-2018 FC1 strains that have spread across Europe and displaced FC2, has raised concerns about the continued efficacy of existing vaccines [10, 15]. Antigenic cartography using hemagglutination inhibition (HI) and virus neutralization (VN) assays with post-infection ferret or horse antisera is the gold standard for quantifying antigenic distances between vaccine and circulating strains. Studies have consistently shown that the 2019 FC1 strains are antigenically distinct from the 2003–2007 vaccine strains, with 2.5- to 6.3-fold reductions in cross-reactivity [10]. This antigenic drift is driven by the accumulation of substitutions in antigenic sites A, B, and E [5].

To address this challenge, reverse genetics (RG) technologies have been developed to rapidly generate vaccine seed viruses that match circulating strains. By combining the HA and NA genes from a contemporary field strain with the six internal genes from a high-growth donor virus (e.g., A/Puerto Rico/8/34), RG-derived vaccines can be produced quickly and at high titers in embryonated chicken eggs [37, 44]. Challenge studies in horses have demonstrated that RG-derived inactivated vaccines provide protective efficacy comparable to that of wild-type virus vaccines, significantly reducing viral shedding, pyrexia, and clinical signs [44]. Furthermore, live-attenuated influenza vaccine (LAIV) platforms, based on a temperature-sensitive master donor virus, have been developed and shown to induce robust humoral and cellular immunity, including cross-protection against both FC1 and FC2 strains when used in a bivalent formulation [17, 59]. These platforms offer the flexibility to update vaccine strains rapidly in response to antigenic drift, a critical capability given the accelerating pace of EIV evolution. The ongoing surveillance and genomic characterization of EIV, as championed by the WOAH and national reference laboratories, remain the cornerstone of effective equine influenza control.

Vaccine Efficacy and Immunoprophylaxis Strategies

The control of equine influenza A virus (EIV) through vaccination represents one of the most complex and dynamic challenges in veterinary immunoprophylaxis. Despite over five decades of vaccine availability since the late 1960s, the virus continues to cause periodic outbreaks in both vaccinated and unvaccinated equine populations globally, underscoring the limitations of current immunization strategies and the relentless evolutionary pressure exerted by the pathogen [8, 35]. A comprehensive systematic review and meta-analysis encompassing 43 controlled trials of EIV vaccination and challenge studies in previously naïve equids, spanning publications up to December 2020, calculated a mean vaccine efficacy (VE) of only 50.03% (95% CI: 23.35–76.71%) for commercial vaccines, with an extraordinary range from 0% to 100% across individual studies [35]. This striking heterogeneity in protective efficacy is not merely a statistical artifact but a reflection of profound biological variability stemming from antigenic mismatch between vaccine strains and circulating field viruses, differences in vaccine platform technologies, adjuvant systems, vaccination schedules, and the inherent immunological status of the target population.

The Antigenic Drift Imperative: Vaccine Strain Mismatch and Breakthrough Infections

The principal biological driver of suboptimal vaccine efficacy is the continuous antigenic drift of the EIV hemagglutinin (HA) glycoprotein, particularly within the globally dominant H3N8 subtype. The World Organisation for Animal Health (WOAH) Expert Surveillance Panel has long recognized that the Florida sublineage, which emerged around 2002, has diverged into two distinct clades, clade 1 (FC1) and clade 2 (FC2), that now circulate with distinct geographical and temporal patterns [8, 10, 16]. The implications for vaccine strain selection are profound. For example, while FC1 strains were historically predominant in the Americas, they suddenly spread across Europe from late 2018 to 2019, causing numerous outbreaks in populations where FC2 had been the dominant circulating lineage [10]. Virus neutralization tests using equine antisera raised against the WOAH-recommended FC1 vaccine strain A/equine/South Africa/4/2003 and the Japanese vaccine strain A/equine/Ibaraki/1/2007 demonstrated that antibody titers against the 2019 European FC1 isolates (A/equine/Tipperary/1/2019 and A/equine/Essex/1/2019) were 2.5- to 6.3-fold lower than titers against the homologous vaccine strains [10]. This quantitative reduction in neutralizing capacity is a direct consequence of amino acid substitutions accumulating in antigenic sites of the HA globular head, particularly site B, where substitutions such as T163I in A/equine/Paris/1/2018 and N188T in post-2015 strains have been documented [5].

The real-world consequences of this antigenic drift are starkly illustrated by multifocal outbreaks in regularly vaccinated horse populations. In Argentina during 2018, an outbreak affecting horses that had been vaccinated with a product containing the WOAH-recommended FC1 strain revealed that 76% of affected animals had a documented history of vaccination, providing compelling field evidence of vaccine breakdown [13]. Phylogenetic analysis of the HA and neuraminidase (NA) genes from the Argentinian outbreak strains revealed multiple amino acid changes at antigenic sites when compared to the vaccine strain, confirming that the virus had evolved sufficiently to evade vaccine-induced immunity [13]. Similarly, the 2012 South American epidemic, which spread rapidly across Chile, Brazil, Uruguay, and Argentina, affected horses vaccinated with both outdated antigens and the WOAH-recommended FC1 strain A/equine/South Africa/4/03, although subsequent antigenic analysis using ferret antisera failed to demonstrate obvious antigenic differences, suggesting that other factors such as poor movement control and suboptimal vaccination coverage contributed to the outbreak [20]. These epidemiological observations underscore a critical principle: vaccine efficacy in the field is not solely a function of antigenic match but is modulated by host factors, vaccine formulation, and management practices.

Vaccine Platform Technologies: Comparative Immunological Mechanisms

The contemporary EIV vaccine landscape encompasses a diverse array of platform technologies, each with distinct immunological mechanisms, advantages, and limitations. The most widely used products include whole inactivated virus (WIV) vaccines, modified live-attenuated influenza vaccines (LAIVs), recombinant canarypox-vectored vaccines, and emerging technologies such as virus-like particles (VLPs) and reverse genetics (RG)-derived vaccines.

Whole Inactivated Virus Vaccines have been the mainstay of equine influenza prophylaxis for decades. These vaccines typically contain adjuvants such as montanide oil ISA 206, saponin, or Mycobacterium phlei extract to enhance immunogenicity. A comparative study evaluating these adjuvants demonstrated that the montanide oil ISA 206-adjuvanted WIV vaccine induced hemagglutination inhibition (HI) antibody titers that persisted for up to 10 months post-inoculation, with superior stability at 4°C for over one year and at room temperature for six months [65]. However, the inherent limitation of WIV vaccines is their reliance on the induction of humoral immunity, primarily neutralizing antibodies directed against the HA globular head, which is precisely the region undergoing the most rapid antigenic drift. The meta-analysis of 21 studies involving commercial vaccines, which are predominantly WIV formulations, revealed a mean VE of only 50.03%, highlighting the inadequacy of this approach for achieving sterilizing immunity [35].

Live-Attenuated Influenza Vaccines represent a paradigm shift in equine influenza immunoprophylaxis, leveraging the induction of both humoral and cellular immune responses, including mucosal IgA and cytotoxic T lymphocyte (CTL) responses that target more conserved internal viral proteins. The first reverse genetics-generated H3N8 EIV LAIV, based on a temperature-sensitive (ts) master donor virus, demonstrated remarkable efficacy in both mouse models and the natural horse host. A single intranasal administration of this LAIV conferred protection against wild-type H3N8 challenge, with the critical advantage that the vaccine backbone can be rapidly updated to incorporate the HA and NA genes of emerging strains [59]. Building on this concept, a bivalent LAIV incorporating HA and NA from both FC1 (A/equine/Ohio/2003) and FC2 (A/equine/Richmond/1/2007) strains was developed and tested in horses. Prime-boost vaccination with this bivalent LAIV was safe and conferred protection against challenge with both clade 1 and clade 2 wild-type viruses, as evidenced by significant reductions in clinical signs, fever, and virus excretion [17]. This approach directly addresses the WOAH recommendation that vaccines should include representative strains from both Florida sublineage clades to maximize protection against currently circulating viruses [17].

Recombinant Canarypox-Vectored Vaccines, such as ProteqFlu™, offer a unique advantage in their DIVA (Differentiating Infected from Vaccinated Animals) capability, which is invaluable for surveillance and outbreak management. The vaccine encodes the HA genes of both an American lineage (FC1) and a Eurasian lineage strain, providing broader antigenic coverage. During Australia’s first and only equine influenza outbreak in 2007, which was caused by a Florida clade 1 strain (A/equine/Sydney/07) that was antigenically distinct from the vaccine strains, horses vaccinated with ProteqFlu™ demonstrated clinical protection despite the mismatch [63]. Serological analysis confirmed that vaccination induced cross-reactive HI antibodies to the outbreak strain, and although interferon-gamma (IFN-γ) responses were detected following in vitro re-stimulation of peripheral blood mononuclear cells, no significant difference in cellular responses was observed between vaccinated and unvaccinated groups, suggesting that humoral cross-reactivity was the primary mediator of protection [63]. The widespread use of this vaccine in France has been credited with maintaining high population-level immunity, with a large-scale seroepidemiological study of 3,004 horses revealing that 87.6% possessed antibody titers above the clinical protection threshold, and 83% of seropositive young horses lacked antibodies to the EIV nucleoprotein, confirming that their immunity was vaccine-induced rather than from natural infection [15]. This robust vaccine coverage likely explains the absence of detected EIV circulation in France between 2015 and late 2018, until the incursion of a novel FC1 strain challenged this immunity [15].

Reverse Genetics-Derived Vaccines represent a technological breakthrough for rapid vaccine strain updating. The standard approach involves replacing the HA and NA genes of a high-growth master donor virus (typically A/Puerto Rico/8/34, PR8) with those of contemporary EIV strains, creating a 6:2 reassortant that grows to high titers in embryonated chicken eggs while presenting the antigenic targets of the circulating virus [37, 44]. A formalin-inactivated vaccine derived from an RG virus containing the HA and NA of A/equine/Tipperary/1/2019 (a 2019 European FC1 strain) and the six internal genes of PR8 was evaluated in a challenge study using 12 naïve yearlings. Both the RG-derived and wild-type virus-derived vaccines significantly reduced viral shedding and clinical signs compared to controls, with no significant differences in protective efficacy between the two vaccine types [44]. Importantly, the RG virus demonstrated superior growth properties in embryonated chicken eggs, facilitating large-scale vaccine production [37]. This technology enables vaccine manufacturers to respond rapidly to antigenic drift, potentially reducing the lag time between the emergence of a new antigenic variant and the availability of an updated vaccine from years to months.

Virus-Like Particle Vaccines represent an emerging platform that combines the safety profile of inactivated vaccines with the immunogenicity of whole virions. VLPs comprising the HA, NA, and M1 proteins of H3N8 EIV, produced using the insect cell-baculovirus expression system, demonstrated morphology and structure indistinguishable from native virus particles [64]. Equine hyperimmune serum raised against these VLPs provided both prophylactic and therapeutic efficacy in a mouse passive immunization and challenge model, suggesting that VLP-based vaccines could offer a safe and effective alternative to egg-grown vaccines, particularly for horses with egg protein allergies [64].

Immunological Correlates of Protection and Vaccine-Induced Immunity

The establishment of robust immunological correlates of protection is essential for vaccine evaluation and strain selection. The single radial hemolysis (SRH) assay and the hemagglutination inhibition (HI) test remain the gold standards for measuring vaccine-induced antibody responses, with an SRH antibody level of approximately 85 mm² or an HI titer of ≥1:32 generally considered indicative of clinical protection [15, 46]. However, these thresholds are not absolute, and protection against infection (sterilizing immunity) requires substantially higher antibody levels than protection against clinical disease. The meta-analysis of vaccine efficacy revealed that complete protection from virus shedding, as measured by the absence of detectable virus in nasopharyngeal swabs, was achieved in only 5 of the 38 studies analyzed, underscoring the difficulty of preventing infection even with effective vaccines [35].

The humoral immune response to EIV vaccination is characterized by the production of HA-specific neutralizing antibodies that block viral attachment to sialic acid receptors on host epithelial cells. However, the continuous evolution of the HA glycoprotein, particularly within the five major antigenic sites (A through E) on the HA1 domain, allows emerging strains to escape neutralization by antibodies raised against older vaccine strains. For example, the Saudi Arabian strains isolated during the 2019 outbreak exhibited 10 amino acid substitutions in HA1 and 4 in HA2 compared to the FC1 vaccine strains, with several of these substitutions located within or near antigenic sites [1]. Similarly, the FC2 strains detected in Egypt in 2017-2018, which represented the first detection of the FC2-144V subgroup in the Middle East, showed 14 amino acid differences from the Egyptian vaccine strain A/equine/Egypt/6066NANRU-VSVRI/08 (H3N8) and 9 differences from the WOAH-recommended FC2 strain A/equine/Richmond/1/2007 [9].

Cellular immune responses, particularly CD8+ cytotoxic T lymphocytes targeting conserved internal proteins such as the nucleoprotein (NP) and matrix (M1) protein, offer the potential for cross-protective immunity against heterologous strains. The concurrent administration of a recombinant canarypox EIV vaccine and an inactivated equine herpesvirus type 1 (EHV-1) vaccine was shown to significantly enhance EIV-specific IFN-γ production in horses, without compromising humoral responses, suggesting that strategic combination vaccination could potentiate cellular immunity [62]. However, the role of T-cell responses in protection against EIV challenge remains incompletely defined, and the meta-analysis did not identify any studies that systematically evaluated cellular immune correlates of protection [35].

Vaccination Strategies and Schedules: Optimizing Immunoprophylaxis

The optimal vaccination schedule for equine influenza remains a subject of debate, with recommendations varying by country, horse use, and risk assessment. The WOAH-recommended primary vaccination course for naïve horses consists of two doses administered 4-6 weeks apart, followed by a third dose at 5-6 months, with booster vaccinations every 6-12 months thereafter [8, 12]. However, the immunological basis for these intervals is increasingly being questioned. A study evaluating the immunogenicity of an RG-derived inactivated vaccine in Thoroughbred yearlings revealed that the first vaccination induced poor antibody responses, particularly in the RG vaccine group, but that the second and third vaccinations elicited robust HI titers that were maintained for up to 28 weeks after the first dose [37]. This pattern of delayed seroconversion is consistent with the need for multiple antigen exposures to achieve protective antibody levels, particularly in immunologically naïve animals.

The issue of concurrent vaccination with other biologics is of practical importance, as horses frequently receive multiple vaccines simultaneously. A study comparing concurrent versus consecutive administration of an inactivated EIV vaccine and a modified-live EHV-1 vaccine in Thoroughbred racehorses found that the percentage of horses showing a twofold or greater increase in HI titers against a heterologous FC2 strain was significantly higher in the consecutive administration group (75%) compared to the concurrent group (37%) [66]. This suggests that concurrent administration may interfere with the immune response to EIV, possibly due to antigenic competition or interference from the live viral vector, and that separating vaccinations by 1-2 weeks may be preferable, especially for naïve horses or those with incomplete vaccination histories [66].

The concept of "accelerated" vaccination schedules has been explored in the context of outbreak response. During the 2007 Australian outbreak, horses were vaccinated with ProteqFlu™ using an accelerated regimen, and subsequent evaluation of IFN-γ responses in peripheral blood mononuclear cells demonstrated that cellular immunity was detectable following in vitro re-stimulation, although no significant difference was observed between accelerated and conventional schedules [63]. This finding suggests that while accelerated schedules may be necessary in emergency situations, they may not confer superior immunological protection compared to standard regimens.

Antiviral Immunoprophylaxis: The Role of Oseltamivir

While vaccination remains the cornerstone of equine influenza control, antiviral therapy offers a complementary approach for immunoprophylaxis and treatment, particularly in outbreak settings or for high-value animals. The neuraminidase inhibitor oseltamivir phosphate (OP) has been evaluated in an experimental infection model using H3N8 EIV. In the treatment group, where OP administration (2 mg/kg twice daily for five days) was initiated immediately after the onset of pyrexia, the duration of virus excretion (mean 2.3 ± 0.6 days) and pyrexia (2.0 ± 0.0 days) were dramatically shorter than in untreated controls (6.0 ± 0.0 and 8.0 ± 1.0 days, respectively) [29]. Even in the prophylaxis group, where OP was administered once daily starting one day before viral inoculation, the periods of virus excretion (5.0 ± 0.0 days) and pyrexia (4.7 ± 1.5 days) were significantly reduced compared to controls [29]. Perhaps most importantly, both the treatment and prophylaxis groups exhibited significantly reduced bacterial counts of Streptococcus equi subsp. zooepidemicus in bronchoalveolar lavage fluid collected seven days after inoculation, indicating that antiviral therapy can mitigate the risk of secondary bacterial pneumonia, a major cause of morbidity and mortality in equine influenza cases [28, 29]. These findings highlight the potential of antiviral agents as an adjunct to vaccination, particularly for immunologically naïve animals or during the early stages of an outbreak when vaccine-induced immunity has not yet developed.

The Challenge of Vaccine Strain Selection and Global Surveillance

The WOAH Expert Surveillance Panel meets biannually to review global EIV surveillance data and recommend vaccine strain updates. The current recommendations include an FC1 strain (such as A/equine/South Africa/4/2003 or A/equine/Ohio/2003) and an FC2 strain (such as A/equine/Richmond/1/2007) [8, 10, 67]. However, the dynamic nature of EIV evolution means that these recommendations require constant reassessment. The sudden emergence and spread of FC1 strains in Europe during 2018-2019, which displaced the previously dominant FC2 strains, caught many vaccine manufacturers off guard and led to a period of increased outbreak activity in vaccinated populations [10, 15]. The full-length genome sequencing of the 2019 Saudi Arabian outbreak strains revealed that while they belonged to FC1 and showed no evidence of reassortment, they possessed a considerable number of amino acid substitutions in the HA signal peptide, HA1, and HA2 domains compared to the vaccine strains, raising concerns about potential antigenic drift [1].

The situation in Asia further illustrates the complexity of vaccine strain selection. Phylogenetic analysis of EIV strains circulating in Asia from 2007 to 2017 revealed that while FC1 outbreaks in Asia were caused by independent introductions from the Americas, the FC2 strains in mainland Asia formed an autochthonous monophyletic group with a common ancestor dating to 2006, indicating sustained endemic circulation [16]. This Asian FC2 lineage has accumulated characteristic amino acid signatures in all viral proteins, including changes located at the top of the HA1 protein within or near antigenic sites, which may have implications for vaccine efficacy in Asian horse populations [16]. The detection of FC2-144V-related EIV in Arabian mares in Egypt in 2018, representing the first report of this subgroup in the Middle East, underscores the need for continuous surveillance and vaccine strain updating in regions where EIV is endemic but surveillance infrastructure is limited [9].

Host Factors and Population-Level Immunity

The efficacy of vaccination is modulated by a complex interplay of host factors, including age, genetics, prior exposure

References

[1] Alaql FA, Alhafufi A, Kasem S, Alhammad YMO, Albaqshi HA, Alyousaf AA, et al.. Full-Length Genome of the Equine Influenza A Virus Subtype H3N8 from 2019 Outbreak in Saudi Arabia. Animals. 2022. DOI: https://doi.org/10.3390/ani12192720

[2] Mena J, Brito B, Moreira R, Tadich T, González I, Cruces J, et al.. Reemergence of H3N8 Equine Influenza A virus in Chile, 2018. Transboundary and Emerging Diseases. 2018. DOI: https://doi.org/10.1111/tbed.12984

[3] Wen F, Blackmon S, Olivier A, Li L, Guan M, Sun H, et al.. Mutation W222L at the Receptor Binding Site of Hemagglutinin Could Facilitate Viral Adaption from Equine Influenza A(H3N8) Virus to Dogs. Journal of Virology. 2018. DOI: https://doi.org/10.1128/JVI.01115-18

[4] Fatima U, Zhang Z, Zhang H, Wang X, Xu L, Chu X, et al.. Equine Mx1 Restricts Influenza A Virus Replication by Targeting at Distinct Site of its Nucleoprotein. Viruses. 2019. DOI: https://doi.org/10.3390/v11121114

[5] Kleij L, Bruder E, Raoux-Barbot D, Lejal N, Nevers Q, Deloizy C, et al.. Genomic characterization of equine influenza A subtype H3N8 viruses by long read sequencing and functional analyses of the PB1-F2 virulence factor of A/equine/Paris/1/2018. Veterinary Research. 2024. DOI: https://doi.org/10.1186/s13567-024-01289-8

[6] Su S, Wang L, Fu X, He S, Hong M, Zhou P, et al.. Equine Influenza A(H3N8) Virus Infection in Cats. Emerging Infectious Diseases. 2014. DOI: https://doi.org/10.3201/eid2012.140867

[7] Yondon M, Zayat B, Nelson M, Heil GL, Anderson B, Lin X, et al.. Equine Influenza A(H3N8) Virus Isolated from Bactrian Camel, Mongolia. Emerging Infectious Diseases. 2014. DOI: https://doi.org/10.3201/eid2012.140435

[8] Oladunni FS, Oseni S, Martínez-Sobrido L, Chambers T. Equine Influenza Virus and Vaccines. Viruses. 2021. DOI: https://doi.org/10.3390/v13081657

[9] Ahmed B, Bayoumi M, Farrag M, Elgamal M, Daly J, Amer H. Emergence of equine influenza virus H3Nx Florida clade 2 in Arabian racehorses in Egypt. Virology Journal. 2022. DOI: https://doi.org/10.1186/s12985-022-01917-9

[10] Nemoto M, Ohta M, Yamanaka T, Kambayashi Y, Bannai H, Tsujimura K, et al.. Antigenic differences between equine influenza virus vaccine strains and Florida sublineage clade 1 strains isolated in Europe in 2019.. The Veterinary Journal. 2021. DOI: https://doi.org/10.1016/j.tvjl.2021.105674

[11] Singh RK, Dhama K, Karthik K, Khandia R, Munjal A, Khurana S, et al.. A Comprehensive Review on Equine Influenza Virus: Etiology, Epidemiology, Pathobiology, Advances in Developing Diagnostics, Vaccines, and Control Strategies. Frontiers in Microbiology. 2018. DOI: https://doi.org/10.3389/fmicb.2018.01941

[12] Dionísio L, Medeiros F, Pequito M, Faustino-Rocha AI. Equine influenza: a comprehensive review from etiology to treatment. Animal Health Research Reviews. 2021. DOI: https://doi.org/10.1017/S1466252321000050

[13] Olguin-Perglione C, Vissani MA, Alamos F, Tordoya MS, Barrandeguy M. Multifocal outbreak of equine influenza in vaccinated horses in Argentina in 2018: Epidemiological aspects and molecular characterisation of the involved virus strains.. Equine Veterinary Journal. 2020. DOI: https://doi.org/10.1111/evj.13176

[14] Sack A, Cullinane A, Daramragchaa U, Chuluunbaatar M, Gonchigoo B, Gray G. Equine Influenza Virus, A Neglected, Reemergent Disease Threat. Emerging Infectious Diseases. 2019. DOI: https://doi.org/10.3201/eid2506.161846

[15] Fougerolle S, Fortier C, Legrand L, Jourdan M, Marcillaud-Pitel C, Pronost S, et al.. Success and Limitation of Equine Influenza Vaccination: The First Incursion in a Decade of a Florida Clade 1 Equine Influenza Virus that Shakes Protection Despite High Vaccine Coverage. Vaccines. 2019. DOI: https://doi.org/10.3390/vaccines7040174

[16] Miño S, Mojsiejczuk L, Guo W, Zhang H, Qi T, Du C, et al.. Equine Influenza Virus in Asia: Phylogeographic Pattern and Molecular Features Reveal Circulation of an Autochthonous Lineage. Journal of Virology. 2019. DOI: https://doi.org/10.1128/JVI.00116-19

[17] Blanco-Lobo P, Rodríguez L, Reedy S, Oladunni FS, Nogales A, Murcia P, et al.. A Bivalent Live-Attenuated Vaccine for the Prevention of Equine Influenza Virus. Viruses. 2019. DOI: https://doi.org/10.3390/v11100933

[18] Toh X, Soh M, Ng M, Yap SC, Harith N, Fernandez C, et al.. Isolation and characterization of equine influenza virus (H3N8) from an equine influenza outbreak in Malaysia in 2015. Transboundary and Emerging Diseases. 2019. DOI: https://doi.org/10.1111/tbed.13218

[19] Vandoorn E, Stadejek W, Leroux-Roels I, Leroux-Roels G, Parys A, Reeth Kv. Human Immunity and Susceptibility to Influenza A(H3) Viruses of Avian, Equine, and Swine Origin. Emerging Infectious Diseases. 2023. DOI: https://doi.org/10.3201/eid2901.220943

[20] Beuttemmüller EA, Woodward A, Rash A, Ferraz LEdS, Alfieri AF, Alfieri A, et al.. Characterisation of the epidemic strain of H3N8 equine influenza virus responsible for outbreaks in South America in 2012. Virology Journal. 2016. DOI: https://doi.org/10.1186/s12985-016-0503-9

[21] Pellegrini F, Buonavoglia A, Omar A, Diakoudi G, Lucente M, Odigie AE, et al.. A Cold Case of Equine Influenza Disentangled with Nanopore Sequencing. Animals. 2023. DOI: https://doi.org/10.3390/ani13071153

[22] Perglione CO, Golemba M, Torres C, Barrandeguy M. Molecular Epidemiology and Spatio-Temporal Dynamics of the H3N8 Equine Influenza Virus in South America. Pathogens. 2016. DOI: https://doi.org/10.3390/pathogens5040061

[23] Na W, Lyoo K, Yoon S, Yeom M, Kang B, Moon H, et al.. Attenuation of the virulence of a recombinant influenza virus expressing the naturally truncated NS gene from an H3N8 equine influenza virus in mice. Veterinary Research. 2016. DOI: https://doi.org/10.1186/s13567-016-0400-7

[24] Barba M, Daly J. The Influenza NS1 Protein: What Do We Know in Equine Influenza Virus Pathogenesis?. Pathogens. 2016. DOI: https://doi.org/10.3390/pathogens5030057

[25] He W, Li G, Wang R, Shi W, Li K, Wang S, et al.. Host-range shift of H3N8 canine influenza virus: a phylodynamic analysis of its origin and adaptation from equine to canine host. Veterinary Research. 2019. DOI: https://doi.org/10.1186/s13567-019-0707-2

[26] Karamendin K, Kydyrmanov A, Sayatov M, Strochkov V, Sandybayev N, Sultankulova K. Retrospective Analysis of the Equine Influenza Virus A/Equine/Kirgizia/26/1974 (H7N7) Isolated in Central Asia. Pathogens. 2016. DOI: https://doi.org/10.3390/pathogens5030055

[27] Abdelwhab EM, Mettenleiter T. Zoonotic Animal Influenza Virus and Potential Mixing Vessel Hosts. Viruses. 2023. DOI: https://doi.org/10.3390/v15040980

[28] Muranaka M, Yamanaka T, Katayama Y, Niwa H, Oku K, Matsumura T, et al.. Time-related Pathological Changes in Horses Experimentally Inoculated with Equine Influenza A Virus. Journal of Equine Science. 2012. DOI: https://doi.org/10.1294/jes.23.17

[29] Yamanaka T, Tsujimura K, Kondo T, Hobo S, Matsumura T. Efficacy of oseltamivir phosphate to horses inoculated with equine influenza A virus.. Journal of Veterinary Medical Science. 2006. DOI: https://doi.org/10.1292/JVMS.68.923

[30] Wang M, Zhang Z, Wang X. Strain-Specific Antagonism of the Human H1N1 Influenza A Virus against Equine Tetherin. Viruses. 2018. DOI: https://doi.org/10.3390/v10050264

[31] Bera BC, Virmani N, Kumar N, Anand T, Pavulraj S, Rash A, et al.. Genetic and codon usage bias analyses of polymerase genes of equine influenza virus and its relation to evolution. BMC Genomics. 2017. DOI: https://doi.org/10.1186/s12864-017-4063-1

[32] Ogata M, Yamanaka T, Koizumi A, Sakamoto M, Aita R, Endo H, et al.. Application of Novel Sialoglyco Particulates Enhances the Detection Sensitivity of the Equine Influenza Virus by Real-Time Reverse Transcriptase Polymerase Chain Reaction.. ACS Applied Bio Materials. 2019. DOI: https://doi.org/10.1021/ACSABM.8B00813

[33] Ali A. Molecular Epidemiology of the Two Internal Genes of Equine Influenza H3N8 Virus Isolated in Pakistan 2015-16. Pakistan Veterinary Journal. 2018. DOI: https://doi.org/10.29261/PAKVETJ/2018.019

[34] . Equine influenza: evolution of a highly infectious virus. The Veterinary Record. 2018. DOI: https://doi.org/10.1136/vr.k2727

[35] Elliott S, Olufemi OT, Daly J. Systematic Review of Equine Influenza A Virus Vaccine Studies and Meta-Analysis of Vaccine Efficacy. Viruses. 2023. DOI: https://doi.org/10.3390/v15122337

[36] Yondon M, Heil GL, Burks JP, Zayat B, Waltzek T, Jamiyan B, et al.. Isolation and Characterization of H3N8 Equine Influenza A Virus Associated with the 2011 Epizootic in Mongolia. Influenza and Other Respiratory Viruses. 2013. DOI: https://doi.org/10.1111/irv.12069

[37] Ohta M, Bannai H, Kambayashi Y, Tamura N, Tsujimura K, Yamayoshi S, et al.. Growth properties and immunogenicity of a virus generated by reverse genetics for an inactivated equine influenza vaccine.. Equine Veterinary Journal. 2021. DOI: https://doi.org/10.1111/evj.13431

[38] Anarkulov E, Batanova Z, Akhmetsadykov N, Khussainov D, Strochkov V, Kalykova A. MOLECULAR-GENETIC AND MICROBIOLOGICAL ASPECTS OF RELATION BETWEEN EQUINE INFLUENZA VIRUS AND UPPER RESPIRATORY TRACT MICROBIOME OF HORSES IN KAZAKHSTAN. Ġylym ža̋ne bìlìm. 2025. DOI: https://doi.org/10.52578/2305-9397-2025-2-1-71-83

[39] Burashev Y, Strochkov V, Sultankulova K, Orynbayev M, Sansyzbay A, Sandybayev N, et al.. Complete Genome Sequencing of Two Equine Influenza A(H3N8) Virus Strains Isolated in Kazakhstan. Genome Announcements. 2018. DOI: https://doi.org/10.1128/genomeA.00574-18

[40] Omoniwa D, Edeh E, Adola JA, Oyetunde† J, Alaba BA, Oyekan O, et al.. Serological Investigation of Equine Influenza Virus in Polo Horses at the 2021 Jos Polo Tournament, Plateau State. Nigerian Veterinary Journal. 2024. DOI: https://doi.org/10.4314/nvj.v44i4.5

[41] Chappell D, Barnett D, James K, Craig B, Bain F, Gaughan E, et al.. Voluntary Surveillance Program for Equine Influenza Virus in the United States during 2008–2021. Pathogens. 2023. DOI: https://doi.org/10.3390/pathogens12020192

[42] González-Obando J, Forero J, Zuluaga-Cabrera A, Ruíz-Sáenz J. Equine Influenza Virus: An Old Known Enemy in the Americas. Vaccines. 2022. DOI: https://doi.org/10.3390/vaccines10101718

[43] Badji A, Faye A, Thior Y, Sarr S, Mbengue B, Sene A. Risk factors for infection with equine influenza virus in donkeys (Equus asinus) in Senegal. International Journal of Biological and Chemical Sciences. 2022. DOI: https://doi.org/10.4314/ijbcs.v15i5.7

[44] Ohta M, Kambayashi Y, Mita H, Kuroda T, Bannai H, Tsujimura K, et al.. Protective efficacy of a reverse genetics-derived inactivated vaccine against equine influenza virus in horses.. Vaccine. 2022. DOI: https://doi.org/10.1016/j.vaccine.2022.09.047

[45] Alaba B, Likita I, Azeez O, Mayaki A, Abiola J, Olaogun S, et al.. Seroprevalence of equine influenza in three southwestern states of Nigeria. Sokoto Journal of Veterinary Sciences. 2024. DOI: https://doi.org/10.4314/sokjvs.v22i2.8

[46] Paillot R, Pitel P, Pronost S, Legrand L, Fougerolle S, Jourdan M, et al.. Florida clade 1 equine influenza virus in France. The Veterinary Record. 2019. DOI: https://doi.org/10.1136/vr.l1203

[47] Yongfeng Y, Xiaobo S, Nan X, Jingwen Z, Wen-qiang L. Detection of the epidemic of the H3N8 subtype of the equine influenza virus in large-scale donkey farms. International Journal of Veterinary Science and Medicine. 2020. DOI: https://doi.org/10.1080/23144599.2020.1739844

[48] Sack A, Daramragchaa U, Chuluunbaatar M, Gonchigoo B, Bazartseren B, Tsogbadrakh N, et al.. Low Prevalence of Enzootic Equine Influenza Virus among Horses in Mongolia. Pathogens. 2017. DOI: https://doi.org/10.3390/pathogens6040061

[49] Gahan J, Garvey M, Samad RAA, Cullinane A. Whole Genome Sequencing of the First H3N8 Equine Influenza Virus Identified in Malaysia. Pathogens. 2019. DOI: https://doi.org/10.3390/pathogens8020062

[50] Gahan J, Garvey M, Gildea S, Gür E, Kagankaya A, Cullinane A. Whole‐genome sequencing and antigenic analysis of the first equine influenza virus identified in Turkey. Influenza and Other Respiratory Viruses. 2018. DOI: https://doi.org/10.1111/irv.12485

[51] Meseko CA, Ehizibolo D, Nwokike EC, Wungak Y. Serological evidence of equine influenza virus in horse stables in Kaduna, Nigeria. Journal of Equine Science. 2016. DOI: https://doi.org/10.1294/jes.27.99

[52] Kareche H, Valle-Casuso J, Madeline A, Froger D, Lecouturier F, Gonzalez G, et al.. Equine influenza outbreak in Eastern of Algeria in 2021: The first introduction of Florida Clade 1 to Maghreb area.. Comparative Immunology, Microbiology & Infectious Diseases. 2023. DOI: https://doi.org/10.1016/j.cimid.2023.102109

[53] Lee E, Kim E, Shin Y, Song J. Design and testing of multiplex RT-PCR primers for the rapid detection of influenza A virus genomic segments: Application to equine influenza virus.. Journal of Virological Methods. 2016. DOI: https://doi.org/10.1016/j.jviromet.2015.11.012

[54] Kinsley R, Pronost S, Bock MD, Temperton NJ, Daly J, Paillot R, et al.. Evaluation of a Pseudotyped Virus Neutralisation Test for the Measurement of Equine Influenza Virus-Neutralising Antibody Responses Induced by Vaccination and Infection. Vaccines. 2020. DOI: https://doi.org/10.3390/vaccines8030466

[55] Gora IM, Kwaśnik M, Żmudziński J, Rożek W. Chorioallantoic membranes of embryonated chicken eggs as an alternative system for isolation of equine influenza virus. Virology Journal. 2017. DOI: https://doi.org/10.1186/s12985-017-0788-3

[56] Hemida M, Perera RA, Chu D, Alnaeem A, Peiris M. Evidence of equine influenza A (H3N8) activity in horses from Eastern and Central Saudi Arabia: 2013–2015. Equine Veterinary Journal. 2018. DOI: https://doi.org/10.1111/evj.13001

[57] Scott S, Kinsley R, Temperton NJ, Daly J. The Optimisation of Pseudotyped Viruses for the Characterisation of Immune Responses to Equine Influenza Virus. Pathogens. 2016. DOI: https://doi.org/10.3390/pathogens5040068

[58] Nemoto M, Okita N, Kitahata M, Bannai H, Tsujimura K, Kinoshita Y, et al.. Evaluation of cobas Influenza A/B & RSV Test for Diagnosis of Equine Influenza.. Journal of Equine Veterinary Science. 2020. DOI: https://doi.org/10.1016/J.JEVS.2020.103249

[59] Rodríguez L, Reedy S, Nogales A, Murcia P, Chambers T, Martínez-Sobrido L. Development of a novel equine influenza virus live-attenuated vaccine. Virology. 2018. DOI: https://doi.org/10.1016/j.virol.2018.01.005

[60] Trovão N, Khan SM, Lemey P, Nelson M, Cherry JL. Comparative evolution of influenza A virus H1 and H3 head and stalk domains across host species. mBio. 2023. DOI: https://doi.org/10.1128/mbio.02649-23

[61] Sreenivasan CC, Jandhyala SS, Luo S, Hause B, Thomas M, Knudsen D, et al.. Phylogenetic Analysis and Characterization of a Sporadic Isolate of Equine Influenza A H3N8 from an Unvaccinated Horse in 2015. Viruses. 2018. DOI: https://doi.org/10.3390/v10010031

[62] Lee D, Lee E, Seo J, Ko E. Evaluation of concurrent vaccinations with recombinant canarypox equine influenza virus and inactivated equine herpesvirus vaccines. Journal of Animal Science and Technology. 2022. DOI: https://doi.org/10.5187/jast.2022.e30

[63] El-Hage C, Hartley C, Savage C, Watson J, Gilkerson J, Paillot R. Assessment of Humoral and Long-Term Cell-Mediated Immune Responses to Recombinant Canarypox-Vectored Equine Influenza Virus Vaccination in Horses Using Conventional and Accelerated Regimens Respectively. Vaccines. 2022. DOI: https://doi.org/10.3390/vaccines10060855

[64] Zhang P, Sun Y, Tan C, Ling M, Li X, Wang W, et al.. Preparation and evaluation of virus-like particle vaccine against H3N8 subtype equine influenza.. Microbial Pathogenesis. 2021. DOI: https://doi.org/10.1016/j.micpath.2021.104885

[65] Warda FF, Shosha E, Abdel-Ra’ouf AM, Kalad M. Immunogenicity of inactivated Equine Influenza (H3N8) virus vaccine with different adjuvents in equine. Benha veterinary medical journal. 2021. DOI: https://doi.org/10.21608/BVMJ.2021.66190.1350

[66] Ohta M, Bannai H, Nemoto M, Kambayashi Y, Tamura N, Tsujimura K. Antibody Responses Against Equine Influenza Virus Induced by Concurrent and by Consecutive Use of an Inactivated Equine Influenza Virus Vaccine and a Modified Live Equine Herpesvirus Type 1 Vaccine in Thoroughbred Racehorses.. Journal of Equine Veterinary Science. 2020. DOI: https://doi.org/10.1016/j.jevs.2020.103221

[67] Nemoto M, Yamanaka T, Bannai H, Tsujimura K, Kokado H. Complete Genomic Sequences of H3N8 Equine Influenza Virus Strains Used as Vaccine Strains in Japan. Genome Announcements. 2018. DOI: https://doi.org/10.1128/genomeA.00172-18