Equine Influenza Virus: Veterinary Reference

Overview and Taxonomy of Equine Influenza Virus: Veterinary Reference

Equine influenza virus (EIV) represents a significant and persistent etiological agent within the family Orthomyxoviridae, genus Influenza A virus. Its impact on the global equine industry is profound, extending from acute morbidity in individual animals to substantial economic losses through quarantine, cancellation of equestrian events, and trade restrictions [6, 17]. As a highly contagious pathogen, EIV is a primary cause of acute respiratory disease in equids worldwide, and its study is paramount for both veterinary clinicians and researchers focused on infectious disease dynamics [1, 4, 19]. The virus is characterized by a segmented, single-stranded, negative-sense RNA genome, a structural feature that is the root of its remarkable evolutionary capacity. This genomic architecture facilitates both antigenic drift, accumulation of point mutations, particularly in surface glycoproteins, and, less frequently in equids, antigenic shift through segment reassortment, leading to the emergence of novel viral variants [6, 19]. Understanding the complex taxonomy of EIV is not merely an academic exercise; it is a cornerstone of effective surveillance, vaccine strain selection, and the implementation of biosecurity protocols as recommended by the World Organisation for Animal Health (WOAH) [5, 6, 17].

Taxonomic Position and Subtype Classification

At its most fundamental taxonomic level, EIV is classified as an Influenza A virus, distinguished from Influenza B and C viruses by the antigenic properties of its internal nucleoprotein (NP) and matrix (M) proteins [19]. The species is further differentiated into subtypes based on the antigenic and genetic characteristics of its two major surface glycoproteins: hemagglutinin (HA) and neuraminidase (NA) [19]. Historically, two distinct subtypes have been responsible for equine influenza: H7N7 (the classical equine subtype, first isolated in Prague in 1956) and H3N8 (first isolated in Miami in 1963) [5, 6]. The H7N7 subtype, also known as equine-1, has not been isolated from equids since the late 1970s and is widely considered to be extinct in the equine population, though serological traces may persist [5]. Consequently, the entirety of contemporary equine influenza is attributable to the H3N8 subtype, which has undergone extensive and continuous evolution since its emergence [6, 11]. It is crucial to note that while EIV is primarily a pathogen of equids, its classification within Influenza A is of immense public health significance. The Orthomyxoviridae family, particularly influenza A viruses from avian and swine hosts, is monitored by global health authorities including the World Health Organization (WHO) and the Food and Agriculture Organization (FAO) due to their pandemic potential. The equine H3N8 virus has demonstrated the capacity for cross-species transmission, most notably establishing a canine influenza virus (CIV) H3N8 lineage in dogs in the United States around 2004, a stark reminder of the zoonotic and panzootic potential inherent in these viruses [3, 6, 19].

Molecular Phylogeny and Lineage Evolution of the H3N8 Subtype

The evolutionary trajectory of H3N8 EIV since its introduction into horses in 1963 provides a textbook example of phylogenetic divergence driven by immune pressure and geographic segregation. Initial molecular characterization revealed that the H3N8 viruses formed a single, relatively homogenous lineage for several decades. However, in the late 1980s, a significant genetic bifurcation occurred, separating the H3N8 EIV into two major geographically distinct lineages: the Eurasian lineage and the American lineage [5, 6, 20]. The Eurasian lineage, which circulated primarily in Europe and Asia, was subsequently outcompeted and has not been detected in recent years. The American lineage, conversely, underwent further diversification, giving rise to multiple sub-lineages including the Kentucky, South America, and, most critically, the Florida sub-lineage [5, 6, 11].

The Florida sub-lineage is the sole lineage of H3N8 EIV currently circulating in the global equine population. It is further divided into two antigenically and genetically distinct clades: Florida Clade 1 (FC1) and Florida Clade 2 (FC2) [5, 6, 16]. As of the most recent WOAH Expert Surveillance Panel recommendations, both FC1 and FC2 are recognized as the contemporary circulating strains that must be represented in all equine influenza vaccines. The emergence and antigenic divergence of these clades has been a primary driver of vaccine breakdown and outbreaks in vaccinated populations in the 21st century [5, 14]. FC1 viruses have been predominant in North America and have been responsible for major outbreaks in South Africa, Japan, and Australia following international horse movements. FC2 viruses, meanwhile, have been the dominant lineage in Europe and Asia, though both clades can be found on multiple continents due to the high volume of international equine travel [11, 17, 18]. For instance, FC1 was first reported in Italy in 2019, underscoring the dynamic and global nature of EIV circulation [7]. The genetic distinctions between these clades are predominantly located in the HA gene, especially in or near antigenic sites, which are the targets of neutralizing antibodies. Amino acid substitutions in these sites, such as A138S and G142R in antigenic site A, and R62K in antigenic site E, have been associated with reduced vaccine efficacy and the ability of new strains to infect previously vaccinated or infected horses [11, 15]. The situation is further complicated by the presence of distinct subpopulations within a single clade, as demonstrated during a 2015 outbreak in Brazil where three distinct FC1 genetic variants were isolated from a single outbreak, suggesting rapid, within-host evolution (quasispecies formation) that can complicate outbreak control [15].

Genomic Insights Beyond the Surface Glycoproteins

While HA and NA define the subtype and primary antigenic properties, analysis of the internal genes, including the matrix (M), nucleoprotein (NP), and non-structural (NS) genes, provides a more granular and often surprising view of EIV evolution and ecology. Some of the most illuminating taxonomic insights have come from these analyses, revealing reassortment events and complex phylogenetic relationships that are not apparent from surface glycoprotein analysis alone.

A landmark phylogenetic study of the M gene revealed the formation of a distinct "Asian-like group" that included not only strains from Asia but also contemporary strains from Europe that had previously been classified as FC2 based on their HA gene [2]. This finding demonstrates that phylogenetic grouping based on a single gene can be misleading. The presence of an "Asian-like" M gene in European strains suggests past reassortment events, where the HA segment maintained its FC2 identity but the M gene was acquired from a different ancestral lineage. Furthermore, this study found that the M gene of these recent strains showed homology with the obsolete H7N7 subtype, indicating that genetic remnants of the extinct H7N7 lineage may persist in the gene pool of contemporary H3N8 viruses [2]. This has profound implications for understanding viral evolution: more conserved genes like M may serve as a molecular fossil record, tracking deeper evolutionary relationships that the highly variable HA gene may obscure [2].

Analysis of the NS gene in Moroccan isolates provides another layer of taxonomic complexity. The NS gene of the A/equine/Nador/1/1997 strain clustered with the European lineage (related to Newmarket/2/93), while the NS genes of two 2004 isolates from Essaouira clustered with much older, pre-divergent strains from the 1970s [20]. This suggests that different virus populations with distinct internal gene constellations were co-circulating in Morocco, and that the 2004 outbreak may have been caused by a virus that had been evolutionarily static in its NS gene for decades [20]. The functional impact of mutations in these internal genes is an area of active research. For example, point mutations in the NEP (nuclear export protein) of the Nador strain (T33I, Q34R) were computationally predicted to enhance the protein's interaction with cellular CRM1, potentially improving the efficiency of viral ribonucleoprotein export from the nucleus and thus viral replication kinetics [8, 9]. Similarly, mutations in the HA2 subunit (the stalk of the HA protein) and the cleavage site, although not altering the monobasic cleavage site characteristic of low-pathogenicity influenza, have been identified in Moroccan strains and may play a permissive role in the viral infection process [10]. The NP and M genes of Pakistani H3N8 isolates from 2015-16 showed 99.7-100% nucleotide identity with avian influenza A (H7N3) viruses circulating in Pakistani poultry, providing strong molecular evidence for a reassortment event where the equine virus acquired internal genes from an avian source [12]. This highlights the risk posed by mixed farming practices where equids and poultry are housed in proximity, creating a "mixing vessel" environment for influenza A viruses [12, 13]. The segmented genome of influenza A is capable of generating novel genotypes through such reassortment, and the equine respiratory tract, while not a traditional mixing vessel like swine, can support the replication of both avian and mammalian-adapted viruses, as demonstrated by serological and molecular evidence of avian H3-subtype virus exposure in horses [8, 13]. These reassortment events underscore that the taxonomy of EIV is not static; it is a dynamic mosaic of gene segments with potentially disparate evolutionary histories.

Molecular Pathogenesis of Equine Influenza Virus: Genetic Diversity, Antigenic Drift, and M Gene Phylogeny

Equine influenza virus (EIV), a member of the Orthomyxoviridae family, represents a paradigm of viral adaptation within a constrained host niche. The molecular pathogenesis of EIV is fundamentally dictated by the dynamic interplay between its segmented RNA genome, the selective pressures exerted by host immunity, and the evolutionary trajectories encoded within its internal genes. Understanding these molecular underpinnings is not merely an academic exercise; it is the cornerstone of effective vaccine design, diagnostic accuracy, and global surveillance efforts, as underscored by the World Organisation for Animal Health (WOAH). The H3N8 subtype, which has exclusively circulated since the apparent extinction of the H7N7 subtype decades ago, provides a compelling case study in how a pathogen can persist, diversify, and periodically circumvent vaccine-induced immunity through a combination of antigenic drift and segment-specific phylogenetic divergence [6, 19].

The Segmented Genome and the Foundation of Genetic Diversity

The molecular architecture of EIV is the primary driver of its pathogenic potential. The virus possesses a genome of eight single-stranded, negative-sense RNA segments. This segmentation is the biological engine for genetic diversity through two principal mechanisms: mutation and reassortment. The RNA-dependent RNA polymerase (RdRp) lacks proofreading activity, leading to a high mutation rate, particularly in genes encoding surface glycoproteins. This intrinsic error-proneness is the fuel for antigenic drift. However, the segmented nature also allows for reassortment, the exchange of entire gene segments when two different influenza A viruses co-infect a single cell. While reassortment is well-documented in swine and avian hosts, its role in equine influenza evolution has been increasingly recognized. The internal genes (matrix, nucleoprotein, non-structural, and polymerase complex) can have distinct evolutionary histories from the surface glycoproteins, creating complex phylogenetic mosaics. For instance, phylogenetic analysis of the matrix (M) and nucleoprotein (NP) genes of H3N8 isolates from a 2015-2016 outbreak in Pakistan revealed a startling molecular signature: these internal genes shared 99.7-100% nucleotide identity with avian influenza A viruses (H7N3) circulating in Pakistani poultry [12]. This finding provides robust evidence for a natural reassortment event between an equine H3N8 virus and an avian H7N3 virus, most likely facilitated by the mixed farming systems common in the region. This demonstrates that the equine host is not a closed system; it can serve as a mixing vessel, leading to the emergence of novel genotypes with potentially altered pathogenic or host-range properties.

Antigenic Drift: The HA Gene as a Moving Target

The hemagglutinin (HA) glycoprotein is the primary target of the host's protective antibody response. Consequently, it undergoes relentless positive selection for amino acid substitutions that alter antigenic sites, a process known as antigenic drift. The H3 HA gene of EIV has evolved into distinct lineages since its introduction into horses in 1963. The initial divergence gave rise to the Eurasian and American lineages, with the American lineage subsequently diversifying into the Florida sublineage, which now dominates globally. Within the Florida sublineage, two co-circulating clades, Clade 1 (FC1) and Clade 2 (FC2), are recognized by WOAH as the epidemiologically relevant strains that must be included in updated vaccines [5, 6].

The clinical consequences of antigenic drift are starkly illustrated by the 2019 epidemic in Great Britain, where a significant proportion of confirmed cases (18%) occurred in horses that were up-to-date on their vaccinations [21]. This vaccination breakdown is a direct result of a mismatch between the vaccine strains and the circulating field virus. Detailed molecular analysis of FC1 strains isolated during a 2015 outbreak in Brazil identified nine amino acid substitutions in the HA protein compared to the WOAH-recommended FC1 vaccine strain (A/eq/Ohio/01/03) [15]. Critically, two of these substitutions were located in antigenic site A (A138S and G142R) and one in site E (R62K). These are canonical antibody-binding regions, and alterations at these residues can significantly reduce the affinity of pre-existing vaccine-induced antibodies for the new viral variant. The study also documented the presence of three distinct genetic variants and multiple quasispecies within the same outbreak, suggesting that EIV can evolve rapidly even during a single epizootic event, further complicating containment [15].

The selective pressure on HA is not uniform across all residues. In a comprehensive study of Argentinean isolates spanning nearly three decades (1985–2012), the accumulation of amino acid substitutions mapped directly to the evolutionary timeline. Early South American clade 1 (SA1) viruses from the 1990s carried substitutions at antigenic site B (Q189N, Q190E, E193K) [11]. By the time FC1 viruses were detected in Argentina in 2012, five additional substitutions had accrued relative to the 2003 FC1 reference strain, with further micro-evolution evident in the form of two distinct subpopulations (K14A vs. K14T/M70V) among the 2012 isolates [11]. This demonstrates a continuous, stepwise process of antigenic change. The failure to update vaccines promptly can have profound consequences, as seen in Japan, where a vaccine selection committee determined in 2013 that including a FC2 strain was necessary. Comparative testing identified A/equine/Yokohama/aq13/2010 as the most suitable candidate, while A/equine/Carlow/2011 was rejected due to unstable antigenic characteristics [16], highlighting that not all field strains are equal in their ability to serve as effective vaccine antigens.

Intragenic Drivers of Pathogenesis: Beyond the Surface Glycoproteins

Pathogenesis is not solely a function of host immune evasion; it is also determined by the intrinsic fitness and replicative capacity of the virus, which are governed by internal genes. The nucleoprotein (NP) and matrix (M) genes, while less variable than HA, undergo mutations that can influence viral replication, assembly, and host adaptation. The non-structural protein 1 (NS1) is a key virulence factor that antagonizes the host's innate interferon response, thereby facilitating viral replication. Analysis of Moroccan EIV strains revealed lineage-specific substitutions in the NS1 protein. The 1997 isolate (Nador/1/97) had 12 amino acid changes compared to the ancestral A/equine/Miami/63 strain, while the two 2004 isolates (Essaouira) shared a single common mutation at residue S228P [20]. The critical finding was that these mutations were located outside the RNA-binding domain and effector domain of NS1, leading the authors to postulate that the protein's capacity to inhibit cellular defenses was likely preserved [20]. This suggests that while the NS gene evolves, its core virulence function is conserved under strong negative selection.

Further insights into molecular adaptation come from in silico modeling of the nuclear export protein (NEP or NS2), which is critical for the translocation of viral ribonucleoprotein (vRNP) complexes from the nucleus to the cytoplasm. Comparing Moroccan strains to a reference, the A/equine/Nador/1/1997 strain harbored two specific mutations (T33I and Q34R) in the CRM1-binding domain of NEP [8, 9]. Computational modeling predicted that these mutations enhance the binding affinity of NEP to the equine CRM1 transport protein, thereby facilitating more efficient nucleocytoplasmic trafficking of vRNPs. This represents a direct molecular adaptation to the equine cellular environment, potentially increasing viral replication kinetics and virulence [8].

M Gene Phylogeny: A Marker of Deep Evolutionary History

While the HA gene is the primary target for clade classification, analysis of more conserved internal genes, particularly the M gene, has revealed a more nuanced and historically informative phylogenetic structure. The M gene encodes the matrix protein 1 (M1), which underlies the viral envelope, and the ion channel M2 protein. A seminal analysis of the M gene from Polish EIV isolates and all available global sequences uncovered the formation of a distinct phylogenetic cluster termed the "Asian-like group" [2]. This group was not defined by geography; it included not only strains from Asia but also contemporary European isolates. The defining feature of this group was the presence of 12 specific nucleotide substitutions that were crucial for its evolutionary differentiation [2].

Perhaps the most provocative finding from M gene phylogeny is the evidence of a genetic link between the extinct H7N7 subtype and contemporary H3N8 viruses. The study observed detectable homology in the M gene sequences of the Asian-like H3N8 group and the historical H7N7 strains [2]. This molecular "ghost" of a past lineage persisting in the genome of the currently circulating virus suggests that reassortment events between H7N7 and H3N8 viruses may have occurred in the past, allowing conserved genetic elements from the old phylogenetic group to be retained. This underscores a crucial principle: the high variability of the HA gene can obscure deep evolutionary relationships that are preserved in more conserved genes like M [2, 12]. Phylogenetic analysis of the M gene, therefore, provides a more robust and stable framework for tracking the long-term evolutionary history and potential reassortment events in the EIV lineage [2]. The existence of a genetically distinct Asian-like M gene group co-circulating alongside viruses with a different M gene background adds a layer of complexity to our understanding of EIV population structure and evolution that would be invisible through HA analysis alone.

Epidemiology of Equine Influenza Virus: Global Distribution, Clade Dynamics, and Asian-like Group Emergence

Equine influenza virus (EIV) represents one of the most significant respiratory pathogens affecting equid populations worldwide, and its epidemiological landscape has undergone profound transformations since the emergence of the H3N8 subtype in 1963 [6, 19]. The virus, an influenza A virus belonging to the family Orthomyxoviridae, has demonstrated remarkable evolutionary plasticity, driven by its segmented genome and the continuous accumulation of point mutations that facilitate antigenic drift. The global distribution of EIV is not uniform; rather, it reflects a complex interplay of international horse movement, vaccination practices, biosecurity measures, and the intrinsic evolutionary dynamics of the virus itself. Understanding the epidemiological patterns of EIV is essential not only for designing effective surveillance and control strategies but also for anticipating the emergence of novel strains that may threaten both equine health and, through cross-species transmission, other mammalian hosts.

Global Distribution and Epidemiological Patterns

The global distribution of EIV is characterized by its endemicity in most regions with substantial equid populations, punctuated by periodic epizootics that often follow the introduction of antigenically distinct strains into susceptible populations. The virus has been documented across all continents where horses are maintained, though the intensity of circulation and the frequency of outbreaks vary considerably based on vaccination coverage, population density, and the intensity of international movement [6, 19]. The World Organisation for Animal Health (WOAH) maintains surveillance mechanisms to monitor the global spread of EIV, recognizing the disease as a significant threat to international trade and equestrian activities. The H3N8 subtype, which has entirely supplanted the earlier H7N7 subtype that has not been isolated for decades, is the sole circulating subtype in contemporary equine populations [5, 6].

International movement of horses, particularly through air transport, has been repeatedly identified as a critical driver of the global dissemination of EIV [17]. The 2007 outbreak in Australia, which affected over 75,000 horses and cost an estimated one billion Australian dollars, was traced to the importation of an infected vaccinated stallion from Japan, with the virus ultimately originating from the United States [17]. Similarly, the introduction of clade 1 viruses into South Africa in 2003 and Japan in 2007 underscores the role of long-distance travel in facilitating intercontinental viral spread [17]. The risk is compounded by the phenomenon of infected vaccinated horses, which may exhibit minimal clinical signs yet shed sufficient virus to initiate outbreaks in naïve or partially immune populations [17]. In the United States, a voluntary upper respiratory biosurveillance program operating from 2008 to 2021 analyzed 9,740 nasal swab submissions and reported an EIV qPCR-positivity rate of 9.9%, with young equids (<9 years of age), recent travel history, and seasonal occurrence in winter and spring emerging as significant risk factors [22]. This surveillance infrastructure, while voluntary, provides critical insights into the endemic circulation of EIV in North America and underscores the need for continuous monitoring.

Regional epidemiological studies have revealed substantial heterogeneity in EIV circulation patterns. In Argentina, phylogenetic analyses of H3N8 viruses detected between 1985 and 2012 demonstrated four distinct introductions of virus, presumably from North America, giving rise to Group VIII, South American clades 1 and 2, and Florida clade 1 [11]. The demographic reconstruction indicated a remarkable increase in viral diversity in 2006, likely attributable to the co-circulation of different lineages, followed by an abrupt decline after 2009, which may have been connected to the incorporation of Florida clade 2 strains into vaccines [11]. In Brazil, an outbreak in 2015 at a veterinary school hospital in São Paulo involved twelve isolates belonging to Florida clade 1, with hemagglutinin amino acid sequences revealing nine substitutions compared to the recommended vaccine strain, including two in antigenic site A and one in antigenic site E [15]. The identification of three distinct genetic variants and eleven variants within four quasispecies provided direct evidence of within-outbreak viral evolution, likely facilitated by suboptimal vaccine-induced immunity [15]. In Croatia, a major outbreak in 2015 followed a large horse fair in Bjelovar, spreading to more than 20 stud farms and caused by a Florida sublineage clade 2 virus imported through movement of asymptomatic carrier animals [27]. Post-outbreak seroprevalence in continental Croatia was only 12.3%, with vaccination coverage below 10%, highlighting the vulnerability of populations with inadequate immunization [27].

The epidemiological significance of donkeys as hosts for EIV was brought into sharp focus by an outbreak in rescue donkeys in Colorado in 2020, where contemporary Florida clade 1 virus caused severe disease, with survival in donkeys less than one year of age being only 16.6% compared to 85.7% in older animals [25]. This finding emphasizes the importance of prenatal vaccination protocols across all equid species, not solely horses. In Senegal, a study of risk factors for EIV infection in donkeys identified lack of access to veterinary care and wandering behavior as significant predictors of infection, with odds ratios of 2.0 and 2.06, respectively, while attendance at rural markets and young age were not associated with increased risk [23]. These data illustrate that epidemiological risk factors may differ substantially between managed horse populations and free-roaming equid populations, necessitating tailored control strategies.

Clade Dynamics and Molecular Evolution

The molecular epidemiology of EIV H3N8 is fundamentally structured around the concept of clades and sublineages, which reflect the progressive accumulation of amino acid substitutions in the hemagglutinin (HA) glycoprotein, the primary target of protective immunity. Following the divergence of the H3N8 subtype into Eurasian and American lineages, the American lineage subsequently diversified into Kentucky, South America, and Florida sublineages [5, 6]. Since the early 2000s, the Florida sublineage has become globally dominant, bifurcating into two distinct clades: Florida clade 1 (FC1) and Florida clade 2 (FC2) [5, 6, 24]. The WOAH Expert Surveillance Panel recommends that all commercial vaccines include representatives of both FC1 and FC2 strains to ensure broad protection against circulating viruses [5, 17]. However, surveillance data indicate that many commercially available vaccines have not been updated to include FC2 strains, a gap that has contributed to vaccine breakdown in the field [5, 14].

The 2019 equine influenza epidemic in Great Britain provided a stark illustration of the consequences of mismatched vaccine strains and inadequate population immunity. A retrospective descriptive study of 412 confirmed cases across 234 infected premises revealed that 72% of confirmed cases were unvaccinated, and only 18% were vaccinated, with the median age of affected horses being five years [21]. The epidemic occurred in two phases, with the first between January and April and the second extending through August. Critically, only 23% of affected premises quarantined new arrivals, 37% had isolation facilities, and 57% of resident horses were vaccinated [21]. These data underscore that even in a country with robust veterinary infrastructure and mandatory vaccination for competition horses, gaps in coverage and biosecurity create conditions conducive to widespread outbreaks. In Ireland, a detailed analysis of racing yards affected by EIV outbreaks within a four-week period demonstrated vaccine breakdown across all products in 33.8% of horses with up-to-date vaccination records, with 66.7% of these horses not having received a booster within the previous six months [26]. The study concluded that annual booster vaccination should not be relied upon as the sole preventive measure and that increasing booster frequency, particularly in young horses, may be beneficial [26].

The molecular mechanisms underlying clade diversification are rooted in the continuous antigenic drift of the HA gene. In Argentinean strains, amino acid substitutions at antigenic sites B and D were identified in viruses belonging to South American clades 1 and 2, while Florida clade 1 viruses detected in 2012 contained five substitutions compared to the FC1 reference strain, including two at antigenic site A and one at site E [11]. The observation of two distinct subpopulations among the Argentinean FC1 isolates, one with a substitution at K-14A and the other with K-14T and M70V, suggests ongoing diversification even within a single introduction event [11]. In the Brazilian outbreak, the HA amino acid sequences showed nine substitutions compared to the vaccine strain, with four changes altering the hydrophobicity of the hemagglutinin molecule, a property that can influence receptor binding and viral fitness [15]. The detection of multiple variants within quasispecies populations provides evidence of rapid evolution under selective pressure from vaccine-induced immunity [15].

The Asian-like Group: A Paradigm Shift in Phylogenetic Classification

The traditional paradigm of EIV phylogenetics, which relies predominantly on HA gene sequences for clade assignment, has been challenged by the discovery of the Asian-like group through analysis of the matrix (M) gene. A seminal study by Kwaśnik et al. analyzed the M gene sequences of Polish EIV isolates alongside all available M sequences in GenBank and identified a distinct phylogenetic cluster that the authors termed the "Asian-like group" [2]. Critically, this group was not confined to isolates from Asia; rather, it included strains isolated in Europe that had previously been classified as belonging to Florida clade 2 based on their HA gene sequences [2]. Twelve nucleotide substitutions in the M gene were determined to be crucial for defining this group, and the analysis revealed homology between the M gene of Asian-like group strains and H7N7 strains, the extinct subtype that once circulated in equids [2].

The implications of this finding are profound. It suggests that the M gene, which is more conserved than the HA gene, may retain evolutionary signatures of past lineages that are not apparent in the rapidly evolving HA gene. The presence of "traces" of H7N7 in the M gene of contemporary H3N8 strains may indicate a link between the old phylogenetic group and recent strains, possibly reflecting past reassortment events [2]. This finding underscores the importance of analyzing multiple genomic segments for comprehensive phylogenetic reconstruction, as reliance solely on HA may obscure important evolutionary relationships. The Asian-like group, therefore, does not appear to be assigned to a specific geographical region but rather represents a lineage defined by M gene ancestry that is circulating in both Asia and Europe [2].

The emergence of the Asian-like group concept is further supported by analyses of internal genes from other regions. In Pakistan, EIV isolates from an outbreak in 2015-16 were typed as H3N8 and grouped with Florida clade 1 based on HA and NA gene sequences, but analysis of the NP and M internal genes revealed high similarity (99.7-100%) with avian influenza A viruses (A/avian/Pakistan/H7N3/2004) isolated in Pakistan [12]. Phylogenetic analysis showed clustering of the Pakistani equine isolates with avian viruses, suggesting that reassortment between equine H3N8 viruses and avian influenza A H7N3 viruses had occurred, most likely facilitated by mixed farming systems where horses and poultry are kept in close proximity [12]. Two major substitutions, F63L and K243R, were recorded in the matrix gene of the Pakistani isolates compared to the closely related avian virus, while the NP gene did not acquire any amino acid substitution, indicating that the reassortment event involved exchange of internal gene segments while maintaining the equine-origin surface glycoproteins [12]. This finding not only corroborates the importance of internal gene analysis for understanding EIV evolution but also raises concerns about the potential for the emergence of novel reassortants with altered host range or pathogenicity.

Further evidence of the complexity of EIV evolution comes from analyses of the non-structural (NS) gene in Moroccan isolates. Strains A/equine/Nador/1/1997 and A/equine/Essaouira/2/2004 and A/equine/Essaouira/3/2004, isolated during outbreaks in 1997 and 2004, showed differential clustering: the 1997 strain grouped with European lineage strains such as A/equine/Newmarket/2/93, while the 2004 strains showed high homology with strains that had circulated before 1990, belonging to the pre-divergent phase [20]. Amino acid sequence comparison revealed that the Nador/1997 strain had 12 substitutions in the NS1 protein compared to the reference strain A/equine/Miami/1963, none of which were located in the key residues of the RNA-binding domain or the effector domain, suggesting that the virulence of the Moroccan strains may be maintained despite these mutations [20]. Analysis of the nuclear export protein (NEP) in these same strains identified two mutations (T33/I and Q34/R) in the CRM1-binding domain of the Nador/1997 strain, which enhanced the interaction between NEP and CRM1 proteins, facilitating nucleocytoplasmic trafficking and potentially increasing viral replication efficiency in equine cells [8, 9].

The Asian-like group phenomenon, coupled with evidence of reassortment involving internal genes from avian influenza viruses, suggests that the evolutionary trajectory of EIV is more complex than previously appreciated. The detection of H7N7 ancestry in the M gene of contemporary strains [2], the identification of avian-origin internal genes in equine isolates from Pakistan [12], and the persistence of pre-divergent lineage signatures in strains from Morocco [20] collectively indicate that EIV evolution involves not only antigenic drift of surface glycoproteins but also periodic reassortment and retention of conserved gene segments from ancestral lineages. These findings have direct implications for vaccine design and strain selection, as they suggest that the genetic diversity of circulating EIV may be greater than that captured by HA-based surveillance alone. The integration of whole-genome sequencing into routine surveillance programs, as advocated by the Centers for Disease Control and Prevention (CDC) and the Food and Agriculture Organization (FAO) for influenza viruses generally, would provide a more complete picture of EIV evolution and emergence, enabling the timely detection of novel variants with pandemic potential. The continuous monitoring of EIV at the genomic level, combined with serological surveillance using tools such as the recently developed NP-cELISA [4], will be essential for maintaining effective control of this economically significant pathogen.

Clinical Presentation and Pathophysiology of Equine Influenza Infection

Equine influenza (EI), caused by the equine influenza virus (EIV), is a highly contagious, acute respiratory disease of equids that, while rarely fatal in otherwise healthy adults, imposes a substantial burden on the global equine industry through morbidity, disruption of competition and breeding activities, and the implementation of costly biosecurity and quarantine measures [6, 24, 30]. The clinical syndrome is the direct result of a complex interplay between viral virulence factors, host immune status, and environmental stressors, and a thorough understanding of both the outward clinical signs and the underlying pathophysiological mechanisms is essential for accurate diagnosis, effective management, and informed vaccination strategies [6, 19].

Incubation Period and Onset of Clinical Signs

Following natural exposure via inhalation of aerosolized respiratory secretions from an infected equid, the incubation period is typically short, ranging from 1 to 3 days, with an average of 24 to 48 hours [19, 24]. The brevity of this period underscores the high viral replicative efficiency within the equine respiratory epithelium. The initial phase of infection, often subclinical in a partially immune host, involves viral attachment to and entry into ciliated columnar epithelial cells lining the upper and lower respiratory tracts. The hemagglutinin (HA) glycoprotein on the viral envelope binds to sialic acid receptors on host cells, with equine viruses exhibiting a preferential affinity for α2,3-linked sialic acids, which are abundant on equine respiratory epithelial cells [19].

Clinical Manifestations: A Spectrum of Severity

The hallmark of acute equine influenza in a fully susceptible, immunologically naïve animal is the sudden onset of a high, biphasic fever, often reaching 39–41°C (102–106°F) within 24–48 hours of infection [6, 24]. This pyrexia is accompanied by profound depression, anorexia, and a harsh, dry, non-productive cough that can persist for several weeks [24, 30]. The cough, a distinctive and often pathognomonic clinical feature, results from extensive damage to the ciliated respiratory epithelium, leading to airway inflammation and irritation. In addition to pyrexia and cough, affected horses develop a profuse, initially serous, and later mucopurulent nasal discharge, reflecting the progression of epithelial necrosis and secondary bacterial colonization [6]. Other common clinical signs include serous ocular discharge, conjunctivitis, submandibular lymphadenopathy (enlargement and tenderness of the lymph nodes under the jaw), and tachypnea [19, 24]. On thoracic auscultation, increased bronchovesicular sounds may be detected, and in more severe cases, crackles and wheezes indicative of lower airway involvement and secondary bacterial pneumonia can be heard [6].

The severity and duration of clinical signs are highly variable and dependent on several critical factors, including the immune status of the horse (vaccination history), age, the presence of concurrent infections, and the specific strain of EIV [6, 26]. In unvaccinated or inadequately vaccinated horses, the disease is typically more severe, with high fever, profound lethargy, and a persistent cough that may last for 2–3 weeks [21, 24]. The impact on performance is considerable; even after the resolution of acute clinical signs, a period of exercise intolerance and suboptimal performance can persist for several weeks due to residual airway inflammation and damage [24, 28]. In contrast, horses with partial immunity from prior infection or vaccination can exhibit a milder syndrome, characterized by a low-grade fever, transient cough, or even a completely subclinical infection [17, 26]. However, critically, these partially immune horses can still shed significant quantities of virus, serving as a source of infection for naïve cohorts, a phenomenon that has been repeatedly implicated in international disease spread via air transport [17]. Data from the 2019 epidemic in Great Britain highlighted that a substantial proportion of confirmed cases were in vaccinated horses, with 18% of cases having an up-to-date vaccination record, and 33.8% of horses in affected yards experiencing vaccine breakdown, underscoring the reality of incomplete clinical protection conferred by vaccination [21, 26].

Species-Specific Considerations: The Donkey

While the clinical presentation in horses is well-characterized, a growing body of evidence from species like the donkey reveals important nuances. The clinical signs of EIV infection in donkeys are broadly similar to those in horses, including pyrexia, nasal discharge, and coughing [23, 25]. However, a significant and alarming difference lies in the prognosis for young animals. A retrospective case series of an outbreak in rescue donkeys in the United States, caused by a Florida clade 1 H3N8 virus, demonstrated a stark disparity in survival: survival was only 16.6% (1/6) in donkeys less than one year of age, compared to 85.7% (6/7) in animals older than one year [25]. This exceedingly high mortality in foals, which occurred in the context of seronegative dams, underscores the critical importance of prenatal vaccination protocols for all equid species, including donkeys, and suggests that immunological naivety in young donkeys renders them particularly vulnerable to severe, fatal influenza pneumonia [25]. The risk factors for infection in donkey populations, such as lack of access to veterinary care and the free-roaming nature of animals, further complicate control efforts in endemic regions [23].

Pathophysiology: The Molecular and Cellular Basis of Disease

The pathophysiology of equine influenza is rooted in the virus's ability to hijack the host cell's machinery for replication while simultaneously evading and subverting the host immune response. The segmented, negative-sense RNA genome of EIV, belonging to the Orthomyxoviridae family, encodes for a suite of proteins that mediate these effects [19].

1. Viral Entry and Replication: The process begins with the attachment of the HA glycoprotein to α2,3-linked sialic acid receptors on the surface of respiratory epithelial cells. The virus is then internalized via receptor-mediated endocytosis. The low pH within the endosome triggers a conformational change in the HA protein, facilitating fusion of the viral and endosomal membranes and release of the viral ribonucleoprotein (vRNP) complexes into the cytoplasm. These vRNPs are then transported to the nucleus, where viral RNA replication and transcription occur. The viral neuraminidase (NA) glycoprotein plays a crucial role later in the replication cycle by cleaving sialic acid residues on the host cell surface, enabling the release of newly assembled virions from the infected cell and preventing viral aggregation [19]. The reassortment potential of the segmented genome is a major driver of viral evolution, leading to antigenic drift and the emergence of new strains, such as the Florida clade 1 and 2 sublineages, which have necessitated regular updates to vaccine formulations as recommended by the World Organisation for Animal Health (WOAH) [5-7, 16].

2. Host Immune Evasion and Virulence Factors: Beyond the structural proteins HA and NA, the virus encodes non-structural proteins that are critical for pathogenesis. The NS1 protein is a potent antagonist of the host's innate immune system. It functions primarily by inhibiting the production of type I interferons (IFN-α/β), which are key antiviral cytokines. By binding to and sequestering double-stranded RNA (dsRNA), a byproduct of viral replication, NS1 prevents the activation of cellular sensors such as RIG-I and MDA5, thereby suppressing the downstream interferon signaling cascade [20]. Mutations in the NS1 gene, while often not affecting the critical RNA-binding and effector domains, can subtly influence the potency of this immune suppression [20]. The viral nuclear export protein (NEP/NS2) is another non-structural protein essential for viral replication. It orchestrates the nuclear export of newly synthesized vRNPs from the nucleus to the cytoplasm, where viral assembly occurs. Specific mutations in the NEP protein, such as the T/33/I and Q/34/R substitutions identified in a Moroccan H3N8 isolate, have been shown to enhance the binding affinity between NEP and the cellular export factor CRM1, thereby facilitating more efficient nucleocytoplasmic trafficking and potentially increasing viral replication fitness [8, 9].

3. Tissue Damage and Clinical Sequelae: The culmination of these viral processes is widespread destruction of the ciliated respiratory epithelium. The loss of this critical mucociliary escalator function is the primary pathophysiological event. It leads to an impaired ability to clear mucus, debris, and inhaled pathogens, resulting in the characteristic cough and accumulation of nasal exudate. The disruption of the epithelial barrier also predisposes the animal to secondary bacterial infections, most commonly caused by opportunistic bacteria like Streptococcus equi subsp. zooepidemicus or Pasteurella spp., which can progress to bronchopneumonia [6, 19]. The systemic signs of fever and depression are driven by the release of pro-inflammatory cytokines, including interleukins (IL-1, IL-6) and tumor necrosis factor-alpha (TNF-α), which act on the hypothalamus to induce pyrexia and trigger acute phase responses [29]. Indeed, experimental studies have demonstrated significant changes in acute phase parameters like complement activation and erythrocyte sedimentation rate following vaccination or infection, confirming the systemic nature of the inflammatory response [29]. The cleavage site of the HA0 precursor protein into HA1 and HA2 by host proteases is a key determinant of viral tropism and pathogenicity. EIVs possess a monobasic cleavage site (e.g., PEKQI-R-GL), which is cleaved only by trypsin-like proteases present specifically in the respiratory tract, thus restricting infection to these tissues and explaining the lack of systemic spread in horses, unlike highly pathogenic avian influenza viruses which have a multi-basic cleavage site [10].

The Role of Vaccination in Modulating Pathophysiology

Clinical protection, as opposed to sterile immunity, is the primary goal of equine influenza vaccination. Vaccination aims to generate robust humoral and cell-mediated immune responses that can limit viral replication and shedding, thereby reducing the severity of clinical disease and the duration of infectiousness [17]. Antibodies against the HA protein, particularly those measured by the single radial hemolysis (SRH) assay, are considered the primary correlate of protection. An SRH antibody level of 85 mm² or greater is widely considered the threshold for clinical protection, reducing the likelihood of pyrexia and severe coughing [31]. However, the 2019 European epidemic, which notably involved vaccinated horses, demonstrated that even vaccinated populations with antibody levels above this threshold could experience breakthrough infections and clinical disease [21, 26]. This phenomenon is attributed to the relentless antigenic drift of the HA glycoprotein, whereby accumulating amino acid substitutions in antigenic sites A, B, D, and E allow the virus to evade vaccine-induced antibodies [11, 15]. A study of a Brazilian outbreak in 2015 identified four substitutions in the HA of circulating Florida clade 1 viruses that altered the hydrophobicity of the protein and were located in key antigenic sites, enabling infection of vaccinated horses [15]. Consequently, WOAH expert panels continuously monitor circulating strains and recommend updates to vaccine compositions, with the current requirement being that vaccines must include representative strains from both Florida clade 1 and Florida clade 2 of the H3N8 subtype to be effective [5, 7, 14, 16]. The failure to do so leaves the equid population vulnerable to outbreaks, even in the presence of high vaccination coverage, as the epidemiological data from the 2019 British and Irish outbreaks starkly illustrate [21, 26].

Diagnostic Approaches for Equine Influenza Virus: Molecular Assays, Rapid Antigen Detection, and Microfluidic Immunofluorescence

The accurate and timely diagnosis of equine influenza virus (EIV) is critical for implementing effective quarantine measures, guiding therapeutic interventions, and monitoring the evolutionary dynamics of circulating strains, particularly within the context of international horse movement and competition. As the World Organisation for Animal Health (WOAH) emphasizes, the socio-economic impact of EIV outbreaks, ranging from disrupted racing calendars to costly trade restrictions, demands a diagnostic arsenal that balances sensitivity, specificity, speed, and field-deployability. Contemporary diagnostic strategies for EIV have expanded significantly beyond traditional virus isolation and serological profiling, now encompassing a sophisticated hierarchy of molecular assays, rapid antigen detection (RAD) platforms, and emerging microfluidic technologies. Each modality carries distinct advantages and limitations that must be carefully weighed against the clinical context, the stage of infection, and the objectives of surveillance.

Molecular Assays: The Gold Standard for Sensitivity and Genomic Characterization

Real-time reverse transcription polymerase chain reaction (rRT-PCR) and quantitative RT-PCR (RT-qPCR) remain the cornerstone of EIV diagnosis in both reference laboratories and large-scale surveillance programs. The unparalleled sensitivity of these molecular techniques, capable of detecting as few as 50 femtograms or approximately 3 × 10³ copies of genomic RNA per reaction, allows for the identification of viral RNA well before the onset of clinical signs and often for several days after the resolution of symptoms [33]. This window is epidemiologically invaluable, as subclinically shedding horses are a primary vector for international virus spread, a phenomenon well-documented in outbreaks linked to air transport of vaccinated but infected animals [17].

The selection of genetic targets for molecular assays is a subject of ongoing refinement. While the hemagglutinin (HA) gene is the primary target for subtype differentiation and phylogenetic lineage assignment, particularly for distinguishing between Florida clade 1 (FC1) and Florida clade 2 (FC2) strains, other genomic segments provide critical insights. The matrix (M) gene, for instance, has been leveraged to identify the formation of an "Asian-like" group among strains that previously clustered in the Florida 2 clade based on HA analysis, revealing complex evolutionary pathways that may escape detection by HA-based typing alone [2]. Furthermore, the M gene's relative conservation makes it an excellent target for pan-influenza A detection [12]. The nucleoprotein (NP) gene also serves as a robust target, with avian H7N3-like internal genes identified in EIV isolates from Pakistan, underscoring the potential for reassortment events between equine and avian influenza A viruses in regions practicing mixed farming [12].

The specific design of primers and probes is paramount. A study developing a novel rRT-PCR assay targeting the H3 gene in Kazakhstan demonstrated that meticulous optimization of annealing temperatures, primer concentrations, and probe chemistry could yield an assay with absolute specificity for EIV, showing no cross-reactivity with other equine respiratory pathogens such as equine herpesviruses or Streptococcus equi [33]. This is a non-trivial achievement, as the differential diagnosis of equine respiratory disease frequently requires a syndromic panel. Indeed, large-scale voluntary surveillance programs in the United States, which processed nearly 10,000 nasal swabs between 2008 and 2021, have relied on multiplex qPCR panels to simultaneously test for EIV, equine herpesvirus-1 (EHV-1), equine herpesvirus-4 (EHV-4), Streptococcus equi subspecies equi, and equine rhinitis viruses A and B [22]. The results from such programs, which identified a 9.9% EIV qPCR-positivity rate, with younger, travelling horses in winter and spring at highest risk, demonstrate the indispensable role of molecular surveillance in defining regional risk factors and informing biosecurity protocols [22].

Beyond conventional rRT-PCR, the field is advancing toward isothermal amplification methods and next-generation sequencing (NGS). Loop-mediated isothermal amplification (LAMP) offers a simplified, rapid alternative to PCR, requiring only a water bath and yielding results in under an hour. While the sensitivity of LAMP for influenza A is slightly lower than rRT-PCR, approximately 0.1 ng/sample compared to 0.01 ng/sample for the M gene, its robustness and minimal equipment requirements make it an attractive screening tool in resource-limited settings [36]. More transformative is the application of metagenomic NGS for unbiased detection and whole-genome sequencing directly from clinical specimens. This approach not only confirms the presence of the virus but also provides complete genome sequences that are essential for tracking antigenic drift, identifying reassortment events, and ensuring the vaccine strains recommended by WOAH are epidemiologically relevant [18, 37]. Targeted NGS methods using universal amplification primers have been successfully applied to equine specimens, yielding high-quality coverage for HA and NA subtyping, though the choice of reference genome and sequencing depth (≥10 million reads for samples with Ct values >30) critically influences recovery and accuracy [18, 35].

Rapid Antigen Detection Kits: Balancing Speed Against Sensitivity

While molecular assays excel in sensitivity, their reliance on sophisticated thermal cyclers, skilled personnel, and often several hours to complete limits their utility at the point of care, particularly during outbreak scenarios involving large numbers of horses at equestrian events or sales. Rapid antigen detection (RAD) kits, based primarily on immunochromatography (lateral flow) or silver amplification immunochromatography, provide results in 10–15 minutes and are widely used by equine practitioners for field-side decision-making. However, the comparative performance of these kits against the RT-qPCR gold standard reveals a significant trade-off in sensitivity.

A comprehensive evaluation of seven commercial human influenza RAD kits, conducted under experimental conditions, reported sensitivities ranging from a modest 54% to 63% against RT-qPCR [32]. The kit with the highest performance (Quick Chaser Auto Flu A, B) achieved a sensitivity of 63%, while others demonstrated significantly lower detection rates. This variability is even more pronounced when testing samples from naturally infected horses, where viral loads may be lower or more variable than in controlled experimental infections. The fundamental limitation is biological: RAD kits detect viral nucleoprotein, requiring a threshold viral load, often corresponding to a Ct value of <27–30, which may not be present during the incubation period, late in the disease course, or in vaccinated horses that are shedding reduced quantities of virus [1, 32]. During the 2019 epidemic in Great Britain, which involved over 400 confirmed cases and a significant proportion of vaccinated horses, reliance on RAD kits alone would likely have missed a substantial number of infected animals, particularly those with waning immunity or those infected with antigenically drifted strains [21, 26].

Despite these limitations, RAD kits retain a critical role as ancillary diagnostic tools. Their high specificity (typically >95%) means a positive result is highly actionable, allowing for immediate isolation of the index case. The World Organisation for Animal Health (WOAH) and national equine federations recognize this utility, particularly in situations where molecular confirmation will be delayed by sample transport [32]. The practical reality is that RAD kits are often the first line of defense in the field, and their use has been shown to reduce the lag time between clinical suspicion and intervention, thereby limiting the spread within and between premises [34]. The key takeaway for veterinarians is that a negative RAD test does not rule out EIV infection, and a confirmatory RT-qPCR should be performed on any horse with clinical signs consistent with influenza, especially in high-risk populations or during known outbreaks.

Microfluidic Immunofluorescence: Bridging the Diagnostic Gap

The persistent tension between the sensitivity of molecular methods and the speed of RAD kits has driven innovation toward rapid, high-performance platforms. Microfluidic immunofluorescence represents a promising intermediate technology that seeks to combine the operational simplicity of an antigen test with analytical performance approaching that of PCR. A landmark evaluation of a microfluidic immunofluorescence assay kit, originally designed for the detection of influenza A/B and SARS-CoV-2 in human nasopharyngeal specimens, successfully demonstrated its application to EIV [1]. The assay, which operates with a 12-minute turnaround time, utilizes a microfluidic cartridge to capture viral nucleoprotein from a swab sample, followed by detection via fluorescently labeled antibodies. The quantitation of fluorescence signal provides a semi-quantitative readout that is less subjective than the visual interpretation of lateral flow bands.

In head-to-head comparisons with RT-qPCR, the microfluidic immunofluorescence assay achieved a sensitivity of 60.7% when evaluating nasopharyngeal swabs from horses experimentally infected with EIV. This performance was superior to a standard immunochromatography cartridge, which exhibited only 53.6% sensitivity, and comparable to a silver amplification immunochromatography kit, which also reached 60.7% [1]. More importantly, the microfluidic platform demonstrated exceptionally high specificity; it did not cross-react with equine coronavirus, equine herpesviruses, or a panel of seven bacterial pathogens, ensuring that positive results are highly reliable [1]. The ability to detect 11 distinct EIV strains, encompassing both H3N8 and potentially other subtypes, suggests broad reactivity against circulating lineages.

The mechanism underlying the enhanced sensitivity of microfluidic immunofluorescence lies in the physics of the assay. The microfluidic channel creates a high surface-area-to-volume ratio, facilitating efficient capture of viral antigens from the sample flow. The fluorescence detection system, integrated into a compact reader, amplifies the signal from the captured antigen-antibody complexes, reducing the limit of detection compared to colorimetric lateral flow methods. For field veterinarians, this translates into an instrument that can be deployed to a stable or racetrack and operated with minimal training, yielding results in a time frame that permits immediate clinical and biosecurity decisions. The assay is also designed to accept nasopharyngeal swabs, the specimen type recommended by WOAH for optimal viral recovery, making it directly compatible with existing sampling protocols [1].

The practical implications of this technology are profound. During the 2019 EIV epidemic in Great Britain and the concurrent outbreaks in Ireland, rapid identification of index cases was hampered by the reliance on less sensitive point-of-care tests and the logistical delays associated with sending samples to central laboratories [21, 26]. A microfluidic deployment could have accelerated the recognition of vaccine breakdown, which was documented in 33.8% of horses with up-to-date vaccination records in the Irish outbreak [26]. Furthermore, the semi-quantitative nature of the fluorescence readout could potentially allow veterinarians to estimate relative viral load, providing insights into the contagiousness of an individual animal and guiding decisions about isolation length. This is a critical advantage over binary positive/negative RAD kits, as horses with higher viral burdens are more likely to be sources of environmental contamination and onward transmission.

However, the technology is not without limitations. The sensitivity of 60.7%, while superior to some RAD kits, still implies that nearly 40% of infected animals could be missed, particularly if sampled early or late in infection when viral loads are declining. The cost of the microfluidic cartridges and the reader itself remains higher than that of disposable lateral flow cassettes, which may be a barrier to adoption in some practice settings or in low- and middle-income countries where EIV is endemic [1, 34]. Nevertheless, as manufacturing scales and the technology matures, the cost is likely to decrease. The microfluidic immunofluorescence assay represents a meaningful step toward closing the diagnostic gap, offering a rapid, specific, and moderately sensitive tool that can enhance the effectiveness of field-based outbreak management. Its integration into a tiered diagnostic approach, where a positive result prompts immediate action and a negative result in a suspicious case triggers an RT-qPCR, offers a pragmatic and powerful strategy for the control of equine influenza.

Vaccination Strategies and Immune Response to Equine Influenza Virus

The comprehensive management of equine influenza (EI) hinges upon a robust and continuously adaptive vaccination paradigm, given the substantial antigenic plasticity inherent to the H3N8 equine influenza virus (EIV). Vaccination remains the cornerstone of prophylaxis, intended not merely to mitigate clinical disease but to curtail viral shedding and thereby impede transmission dynamics. However, the efficacy of this strategy is perpetually challenged by the virus’s capacity for antigenic drift, the heterogeneity of vaccine platforms, and variable compliance with recommended schedules. This section delves into the immunological underpinnings of vaccine-induced protection, evaluates the strategic deployment of vaccination campaigns in the face of evolving viral lineages, and examines the critical interplay between host immunity, vaccine composition, and epidemiological outcomes.

The Immunological Correlates of Protection and Humoral Response Dynamics

Protection against EIV is predominantly mediated by the humoral immune response, specifically antibodies directed against the hemagglutinin (HA) glycoprotein, which neutralize viral infectivity and prevent cellular entry. The gold-standard serological correlates of protection are the hemagglutination inhibition (HI) assay and the single radial hemolysis (SRH) test. An SRH antibody titre of 85 mm² is widely recognized as the threshold for clinical protection, above which horses are significantly less likely to exhibit pyrexia, coughing, or nasal discharge upon challenge. A titre exceeding 150 mm² is often associated with sterilizing immunity, defined by the complete abrogation of viral replication and shedding [31]. These thresholds are not merely academic; they serve as practical benchmarks for evaluating vaccine immunogenicity and for making evidence-based decisions regarding booster intervals.

The kinetic profile of the antibody response following vaccination is critically dependent on the vaccine platform and adjuvant system. Horses receiving a primary course with an ISCOM matrix-adjuvanted vaccine, as documented in the Hong Kong racing population, demonstrate a rapid and robust SRH response, with average titres reaching approximately 113 ± 34 mm² in resident horses. Notably, a significant incremental increase in titre was observed following each biannual booster, peaking at an average of 128 ± 42 mm² [31]. This consistency underscores the value of a structured, regularly boosted vaccination schedule in maintaining population-level immunity. Conversely, studies from the United Kingdom have identified troubling divergence from datasheet-compliant protocols; only 7.7% of veterinary practices adhered to recommended intervals between the second and third primary vaccination [38]. Such non-compliance risks leaving horses in a window of suboptimal immunity, particularly vulnerable to infection before the full maturation of the memory B-cell pool.

While HA-specific antibodies are paramount, antibodies against the neuraminidase (NA) protein and the nucleoprotein (NP) also contribute to immunity. The NP, being more conserved across lineages, has been harnessed for novel serological tools such as the NP-based competitive ELISA (NP-cELISA). This assay demonstrated 100% sensitivity and specificity in initial validation and an 87.4% concordance rate with the HI test when applied to field sera, outperforming a commercially available kit [4]. Such assays are invaluable for large-scale serosurveillance, allowing for the differentiation between vaccinated and naturally infected animals in populations where DIVA (Differentiating Infected from Vaccinated Animals) capability is absent but where monitoring antibody prevalence remains crucial for vaccine effectiveness assessments.

Antigenic Drift, Clade Divergence, and the Imperative for Vaccine Strain Updates

The principal threat to vaccination efficacy is the relentless antigenic drift of the HA gene. The H3N8 subtype has evolved through a series of distinct phylogenetic lineages. Following the emergence of the American lineage, the Florida sublineage diverged into two co-circulating clades: Florida Clade 1 (FC1) and Florida Clade 2 (FC2). Since approximately 2010, these two clades have become globally dominant, with FC1 predominating in the Americas and parts of Asia, and FC2 circulating widely in Europe and Asia [5-7]. The antigenic distance between these clades and older vaccine strains, such as those derived from pre-divergent or Eurasian lineage viruses, is substantial. This mismatch was starkly illustrated during the 2019 epidemic in Great Britain, where a significant proportion of confirmed equine influenza cases occurred in vaccinated horses. Retrospective analysis of that outbreak revealed that 18% of confirmed cases were fully vaccinated, and among cases in vaccinated horses, 66.7% had not received a booster within the preceding six months, while 37% were overdue for an annual booster [21, 26]. These data provide compelling epidemiological evidence that waning immunity, compounded by antigenic variation, renders annual booster protocols inadequate for sustained protection, a conclusion echoed by the World Organisation for Animal Health (WOAH) which recommends biannual vaccination in high-risk populations [17, 26].

In response to this evolutionary pressure, the WOAH Expert Surveillance Panel on Equine Influenza has consistently recommended that all commercial vaccines include both an FC1 and an FC2 representative strain. Japan’s vaccine strain selection system exemplifies the rigorous evaluation required for such updates. In 2013, following a formal committee review, Japanese authorities concluded that while the existing vaccine strains did not require immediate replacement, the inclusion of an FC2 strain was essential. Comparative testing of three candidate FC2 viruses, A/equine/Carlow/2011, A/equine/Richmond/1/2007, and A/equine/Yokohama/aq13/2010, identified the Yokohama/10 strain as the most suitable, demonstrating superior growth properties in eggs and higher immunogenicity in mice [16]. Similarly, molecular characterization of EIV from outbreaks in Argentina between 1985 and 2012 revealed that the dramatic decline in viral genetic diversity observed after 2009 was temporally correlated with the incorporation of FC2 strains into locally used vaccines, suggesting that updated vaccines were exerting population-level selective pressure and suppressing transmission [11]. The failure to update vaccines at the same pace as viral mutation was a direct contributing factor to the suspension of racing in the UK in February 2019, reinforcing that vaccination must be a dynamic, surveillance-informed intervention [14].

Vaccine Platforms, Immunogenicity, and the Challenge of Mixed Vaccination Regimens

The landscape of commercial EI vaccines is heterogeneous, encompassing inactivated whole-virus preparations, ISCOM-matrix adjuvanted formulations, canarypox-vectored recombinant vaccines, and emerging platforms such as baculovirus-expressed subunit antigens and DNA vaccines. Each platform elicits a distinct immune activation pattern. The ISCOM matrix, for instance, is a potent inducer of both humoral and cell-mediated immunity, promoting antigen presentation via the MHC class I pathway and thus engaging cytotoxic T-lymphocyte responses, which are critical for clearing infected cells and reducing viral load [31].

A significant practical challenge arises from the fact that a horse may receive vaccines from different platforms over its lifetime due to changes in ownership, veterinary practice, or product availability. The compatibility of mixed vaccination regimens has rarely been studied, but preliminary data from Hong Kong, where all imported horses receive a standardized primary course of an ISCOM-adjuvanted vaccine irrespective of prior vaccination history, demonstrate that this approach can achieve robust herd immunity. The average SRH titre in resident horses was 113 mm², with less than 15% falling below the clinical protection threshold of 85 mm² [31]. This suggests that, at least with the ISCOM platform, heterologous boosting is effective. However, caution is warranted, as suboptimal immunological priming with certain inactivated vaccines may not set the stage for an optimal anamnestic response upon subsequent heterologous boosting. The development of a recombinant FC2 HA protein expressed in a baculovirus system represents a promising avenue for creating a standardized, highly specific subunit vaccine that could be rapidly updated to match circulating strains, circumventing some of the variability inherent to egg-based production and ensuring consistency in antigenic content [5].

Anatomical and Physiological Considerations: The Mucosal Immune Barrier

Vaccination strategies must also account for the fact that EIV primarily infects the epithelium of the upper and lower respiratory tract. Systemic vaccination elicits predominantly circulating IgG antibodies, which transudate into the respiratory lumen. While these are effective at neutralizing virus, they are less abundant at the mucosal surface than locally produced secretory IgA. Although parenterally administered inactivated vaccines induce minimal IgA responses, the level of systemic IgG achieved with effective adjuvants and frequent boosters, as seen in the Hong Kong cohort, appears sufficient to provide a robust barrier. The rapid decline in circulating antibodies below the protective threshold within six months of the last booster explains why biannual (every six months) revaccination is physiologically necessary for high-risk populations such as racehorses, sport horses, and breeding stock [26, 31]. The physiological stress of intense exercise, air transport, and co-mingling at equestrian events further strains the immune system, creating a permissive environment for vaccine breakthrough if antibody titres are waning [17].

Vaccine Breakdown, Herd Immunity, and the Role of Unvaccinated Populations

A sobering reality is that even the most current vaccine cannot guarantee sterilizing immunity in every individual. Vaccine breakdown is a multifactorial phenomenon involving the interplay of host factors (age, genetic background, concurrent disease, stress), viral factors (antigenic match, viral load), and vaccine factors (potency, storage, administration). In the 2015 Brazilian outbreak at a veterinary teaching hospital, nine amino acid substitutions were identified in the HA of the circulating FC1 virus compared to the vaccine strain, with two substitutions located in antigenic site A and one in site E, regions critical for antibody binding. These mutations altered the hydrophobicity of the HA surface and were associated with infection in vaccinated horses [15].

Critically, vaccine breakdown is amplified in populations with suboptimal coverage. The 2019 UK epidemic demonstrated that 72% of confirmed cases were in unvaccinated horses, and on infected premises, only 57% of the resident horse population was vaccinated [21]. These unvaccinated animals act as viral amplifiers and reservoirs, sustaining transmission chains even when vaccinated individuals are partially protected. The risk is particularly acute in donkeys and other equids, which are frequently omitted from vaccination programs despite being highly susceptible. During an EIV outbreak in a rescue facility in the United States involving FC1, mortality in donkeys under one year of age reached 83.3%, a tragedy directly attributable to the lack of prenatal dam vaccination and complete absence of herd immunity [25]. In Senegal, a case-control study identified the lack of veterinary care and free-roaming management as significant risk factors for EIV infection in donkeys, highlighting the socio-economic barriers to vaccination in certain regions [23]. A serosurvey in continental Croatia revealed a post-outbreak seroprevalence of only 12.3%, with vaccination coverage persistently below 10% [27]. These data underscore the urgent need for educational campaigns and subsidized vaccination programs targeted at non-competitive horse and donkey populations.

Molecular Markers of Vaccine Efficacy and Viral Fitness

Beyond traditional serology, molecular characterization of vaccine breakdown events has yielded insight into viral fitness adaptations. The matrix (M) gene and non-structural (NS) genes harbor mutations that can influence viral replication efficiency and host immune evasion. Phylogenetic analysis of the M gene has revealed the existence of an “Asian-like” group that is not geographically restricted, with Asian-like gene signatures present in strains recently isolated in Europe, including those classified as FC2 based on HA. This suggests that reassortment events between the HA and internal gene segments may be occurring, potentially modulating the virulence and transmissibility of antigenically matched viruses [2]. Furthermore, analysis of the NS1 protein from Moroccan equine strains identified multiple amino acid substitutions, though none were located in the critical RNA-binding domain or effector domain, indicating that while the virus can accumulate mutations without compromising its ability to inhibit host interferon responses, the potential for future adaptive mutations in these regions remains a concern [20]. These molecular surveillance data are essential for predicting whether a vaccine strain that is antigenically matched might still be rendered less effective by changes in viral replication kinetics or host antagonism conferred by internal gene segments.

Practical Implications for Vaccination Protocols

The cumulative evidence strongly indicates that reliance on an annual booster is a demonstrably insufficient strategy for high-performance or frequently transported horses. The Irish racing outbreak of 2018 clearly demonstrated that unvaccinated horses and those overdue for their six-month booster were the primary vectors of spread [26]. Consequently, regulatory bodies such as the Fédération Équestre Internationale (FEI) and the British Horseracing Authority (BHA) mandate a primary course followed by boosters at six-monthly intervals. However, the translation of these rules into practice is inconsistent; surveys of UK veterinarians found that while 86.2% treated competition horses, only 57% uniformly advised six-monthly boosters for this group, and annual vaccination was commonly advised for non-competition horses [38]. This inconsistency creates a stratified immunity profile within the national herd, where a core of highly vaccinated athletic horses lives alongside a periphery of sporadically or never-vaccinated leisure and companion animals. Data from a voluntary US surveillance program indicate that EIV risk is concentrated in young horses (less than nine years old) with a recent history of travel, emphasizing that the mobile population, even when partly vaccinated, is a conduit for transmission [22].

In conclusion, an effective equine influenza vaccination strategy must be regionally and temporally dynamic, informed by continuous molecular and antigenic surveillance, and executed with rigorous adherence to biannual booster intervals, particularly in high-risk populations. The integration of novel serological tools such as NP-cELISA for monitoring vaccine-induced antibody responses, combined with the deployment of updated multivalent vaccines containing both FC1 and FC2 antigens, represents the most robust defense against a pathogen that has repeatedly proven its capacity to exploit even minor gaps in immunological coverage.

Biosecurity and Control Measures for Equine Influenza Outbreaks: Movement Restrictions and Surveillance

The effective management of equine influenza (EI) outbreaks hinges on a multi-layered strategy that integrates stringent biosecurity protocols, scientifically informed movement restrictions, and robust surveillance systems. The highly contagious nature of the equine influenza virus (EIV), coupled with its capacity for rapid antigenic evolution and international dissemination via the movement of subclinically infected horses, necessitates a proactive and evidence-based approach to outbreak control. This section provides an exhaustive analysis of the core components of biosecurity and control, drawing on epidemiological data, diagnostic advancements, and regulatory frameworks to delineate best practices for containing and mitigating EI outbreaks.

The Epidemiological Imperative for Movement Restrictions

The primary driver of EIV dissemination, both locally and globally, is the movement of infected equids. The 2019 epidemic in Great Britain serves as a stark illustration of this principle. Epidemiological analysis of 234 infected premises (IPs) revealed that 42% of IPs reported the arrival of new horses within two weeks of a confirmed case, and only 23% of these premises implemented quarantine for new arrivals [21]. This failure to isolate incoming animals directly facilitated the rapid propagation of the virus across the country. The epidemic curve exhibited two distinct phases, with the first (January–April) driven largely by the movement of horses from high-risk events, underscoring the critical role of equestrian activities in viral spread [21]. Similarly, the 2015 outbreak in Croatia was traced directly to a major horse fair, where asymptomatic carriers introduced the virus into a largely naïve population, subsequently spreading to over 20 stud farms [27]. These events confirm that any congregation of equids, whether at competitions, sales, fairs, or breeding operations, represents a high-risk nexus for viral transmission.

The biological basis for these movement-related risks lies in the virus’s replication kinetics and the phenomenon of subclinical shedding. Vaccinated horses, while often protected from clinical disease, can still shed infectious virus, a fact repeatedly demonstrated in outbreak investigations. In the 2019 Irish outbreak, 33.8% of horses with up-to-date vaccination records experienced vaccine breakdown, and critically, 66.7% of these had not received a booster within the preceding six months [26]. This highlights that waning immunity, rather than complete vaccine failure, is a major contributor to the movement of undetected shedders. The World Organisation for Animal Health (WOAH) has long recognized this risk, recommending that horses be vaccinated between 21 and 90 days before international shipment to ensure peak antibody titers [17]. However, the absence of international standardization in import requirements, such as the United States Department of Agriculture’s (USDA) lack of specific pre-import vaccination mandates for equine influenza, creates dangerous loopholes [17]. The importation of a clade 2 virus into the USA via a vaccinated mare from Germany, which had not been vaccinated for over a year, exemplifies how gaps in international regulations can lead to transcontinental viral incursions [17].

Implementing Effective Quarantine and Zoning Protocols

A cornerstone of outbreak biosecurity is the rigorous application of quarantine and zoning. Upon suspicion or confirmation of EI, the immediate establishment of a containment zone around the infected premise is paramount. This zone should be defined based on epidemiological risk, typically a radius of 5–10 km, within which all horse movements are strictly prohibited except under exceptional circumstances (e.g., veterinary emergencies). Within the IP, a strict isolation facility must be established. The 2019 GB data indicated that only 37% of IPs had such facilities, a deficiency that likely contributed to within-herd spread [21]. Isolation should be physically separate (ideally >20 meters) from other horses, with dedicated equipment, feed, and personnel to prevent fomite transmission. The duration of isolation should be a minimum of 14 days post-resolution of clinical signs, as viral shedding can persist for 7–10 days in unvaccinated animals and potentially longer in partially immune individuals.

The concept of a “High Health, High Performance” (HHP) population, as proposed by WOAH, offers a framework for managing movement risks for elite sport horses. This system relies on stringent, continuous veterinary supervision, mandatory vaccination, and regular serological monitoring to certify a subpopulation of horses as low-risk, thereby facilitating international travel with reduced quarantine periods [17]. However, this model is predicated on the assumption of robust surveillance and vaccine efficacy. The 2019 outbreaks demonstrated that even horses under FEI (Fédération Équestre Internationale) rules, which mandate six-monthly vaccination, can be involved in transmission, particularly when booster schedules are not strictly adhered to [26, 38]. Therefore, the HHP concept must be applied with caution, recognizing that it is not a substitute for active surveillance and rapid diagnostic testing.

Surveillance: The Backbone of Early Detection and Response

Surveillance for EIV operates on multiple levels, from passive clinical reporting to active molecular and serological monitoring. The 2019 GB epidemic was detected through a combination of passive surveillance (veterinarians reporting suspicious cases) and confirmatory molecular testing [21]. However, reliance on passive surveillance alone is fraught with risk, as it depends on the clinical acumen of practitioners and the willingness of owners to report. The study in Portugal revealed that only 6% of veterinarians used laboratory tests to confirm suspected EI cases, and a significant proportion of suspected cases went unreported, leading to a substantial underestimation of the true disease burden [34]. This underreporting is a systemic weakness that undermines national and international control efforts.

Molecular Surveillance and Point-of-Care Diagnostics

The gold standard for EIV detection remains real-time reverse transcription polymerase chain reaction (RT-qPCR), which offers high sensitivity and specificity. The development of a novel RT-PCR assay for the H3 subtype in Kazakhstan, capable of detecting as few as 50 femtograms of genomic RNA, exemplifies the ongoing refinement of molecular tools for rapid, in-country diagnosis [33]. However, the logistical challenges of shipping samples to centralized laboratories and the time delay in obtaining results can impede rapid response. This has driven the development and evaluation of point-of-care (POC) diagnostic tests.

Recent evaluations of rapid antigen detection (RAD) kits, originally designed for human influenza, have shown variable but promising results for EIV. A comparative study of seven commercial kits found sensitivities ranging from 54% to 63% compared to RT-qPCR, with the Quick Chaser Auto Flu A, B kit demonstrating the highest performance [32]. A microfluidic immunofluorescence assay kit, with a 12-minute turnaround time, showed a sensitivity of 60.7% and high specificity, failing to cross-react with other equine respiratory pathogens [1]. While these POC tests are less sensitive than PCR, their utility lies in their ability to provide immediate, actionable results in the field. A positive result from a RAD kit in a horse with compatible clinical signs can trigger immediate isolation and movement restrictions, even before PCR confirmation is available. Conversely, a negative result, particularly in a high-risk scenario, should not be used to rule out infection, and a confirmatory PCR should be performed. The deployment of such tests at equestrian events, sales, and quarantine facilities could dramatically reduce the lag time between infection and intervention.

Serological Surveillance and Vaccine Monitoring

Serological surveillance, primarily through the hemagglutination inhibition (HI) assay and competitive ELISA (cELISA), provides critical data on population immunity and past exposure. The development of a novel NP-cELISA, which demonstrated 100% sensitivity and specificity in initial validation and an 87.4% concordance rate with the HI assay, represents a significant advancement [4]. This assay is simpler and more cost-effective than the HI test, making it suitable for large-scale serosurveys. Such surveys are essential for identifying gaps in vaccination coverage. For instance, a serosurvey in Croatia following the 2015 outbreak revealed that vaccination coverage remained below 10%, even after a major epidemic, highlighting a persistent risk for future outbreaks [27]. In China, a large-scale NP-cELISA survey from 2021–2023 found an average annual seroprevalence of 37.96%, indicating that a substantial proportion of the equine population remains susceptible [4].

Serological data also inform vaccine strain selection. The WOAH Expert Surveillance Panel regularly reviews circulating strains and recommends updates to vaccine composition. The failure to update vaccines in a timely manner has been implicated in several outbreaks. The 2015 Brazilian outbreak, caused by a Florida Clade 1 (FC1) virus, was linked to the use of outdated vaccine strains that did not match the circulating virus, leading to nine amino acid substitutions in the hemagglutinin (HA) protein, including changes in antigenic sites A and E [15]. Similarly, the predominance of Florida Clade 2 (FC2) viruses in Europe has necessitated the inclusion of FC2 strains in vaccines, a recommendation that many manufacturers have been slow to adopt [5, 16]. Continuous genetic surveillance of the HA gene, as performed in Argentina, Italy, and Poland, is vital for tracking the emergence of new clades and antigenic variants, such as the Asian-like group identified through M gene analysis [2, 7, 11].

Biosecurity Measures at the Premise Level

Beyond movement restrictions and surveillance, individual premise biosecurity is the final line of defense. The 2019 GB study found that only 57% of resident horses on IPs were vaccinated, and many premises lacked basic biosecurity protocols [21]. A comprehensive biosecurity plan should include:

1. Vaccination as a Biosecurity Tool: Mandatory, risk-based vaccination is the single most effective biosecurity measure. The evidence strongly supports a six-monthly booster schedule for horses at high risk of exposure (e.g., competition, breeding, and boarding facilities). The Irish outbreak demonstrated that horses vaccinated annually were significantly more likely to become infected than those boosted within the previous six months [26]. While some owners express concern about over-vaccination and adverse effects, studies have shown that biannual vaccination does not negatively impact performance [28], and the risk of adverse events is low and typically transient (e.g., local swelling, mild pyrexia) [38]. The decision to vaccinate is influenced by owner perceptions of risk, cost, and necessity, with many owners of non-competition horses choosing not to vaccinate, believing their animals are not at risk [30]. Veterinary practitioners must actively counter this misconception by emphasizing that EIV is highly contagious and can be introduced by fomites, wildlife, or asymptomatic carriers.

2. Hygiene and Fomite Control: EIV can survive on contaminated equipment, clothing, and surfaces for up to 48 hours. Strict hygiene protocols are essential. This includes the use of dedicated grooming kits, tack, and feed buckets for each horse or group of horses. Personnel should practice hand hygiene and change footwear or use footbaths containing virucidal disinfectants (e.g., quaternary ammonium compounds or accelerated hydrogen peroxide) when moving between different areas of the facility. Vehicles, especially horse trailers, should be thoroughly cleaned and disinfected after each use.

3. Management of New Arrivals: As highlighted repeatedly, the introduction of new horses is the highest-risk activity. A mandatory quarantine period of 14–21 days in a physically separate facility is non-negotiable. During quarantine, horses should be monitored daily for pyrexia and respiratory signs. Ideally, they should be tested for EIV via PCR on arrival and again before release from quarantine. This is particularly critical for horses returning from events or those originating from regions with known EIV activity.

The Role of International Regulatory Frameworks

The global nature of the equine industry necessitates international cooperation. WOAH provides the overarching framework for disease notification and trade standards. Member nations are obligated to report EI outbreaks, and the WOAH Terrestrial Animal Health Code provides detailed guidelines on import conditions, including vaccination requirements and quarantine periods [17]. However, compliance and enforcement are variable. The lack of harmonization between the import regulations of different countries, such as the USA and those of the European Union, creates vulnerabilities. Furthermore, the trend towards reducing quarantine periods for HHP horses, while economically beneficial, places an even greater burden on the accuracy of pre-export testing and the efficacy of vaccination. The Centers for Disease Control and Prevention (CDC) and the Food and Agriculture Organization (FAO) also play roles in monitoring the zoonotic potential of influenza viruses, including EIV, and in providing technical guidance for outbreak control in resource-limited settings. The 2007 Australian outbreak, which cost an estimated one billion Australian dollars and affected over 75,000 horses, serves as a stark reminder of the catastrophic economic consequences of a single incursion, underscoring the absolute necessity of maintaining robust, scientifically defensible biosecurity and surveillance systems at all levels [17].

References

[1] Kawanishi N, Kinoshita Y, Kambayashi Y, Bannai H, Tsujimura K, Yamanaka T, et al.. Performance of a Microfluidic Immunofluorescence Assay Kit for Equine Influenza Virus Antigen Detection.. Journal of Equine Veterinary Science. 2023. DOI: https://doi.org/10.1016/j.jevs.2023.104956

[2] Kwaśnik M, Gora IM, Żmudziński J, Rola J, Polak M, Rożek W. Genetic Analysis of the M Gene of Equine Influenza Virus Strains Isolated in Poland, in the Context of the Asian-like Group Formation. Journal of Veterinary Research. 2018. DOI: https://doi.org/10.2478/jvetres-2018-0057

[3] Dubovi EJ, Njaa BL. Canine influenza.. The Veterinary clinics of North America. Small animal practice. 2008. DOI: https://doi.org/10.1016/j.cvsm.2008.03.004

[4] Yang Y, Guo K, Xu L, Guo W, Dong M, Liu W, et al.. Development and application of a NP-cELISA for the detection of nucleoprotein antibodies of equine influenza virus. Microbiology spectrum. 2025. DOI: https://doi.org/10.1128/spectrum.00939-25

[5] Atwa A, Gomaa L, Elmenofy W, Amer H, Ahmed B. Expression of recombinant Florida clade 2 hemagglutinin in baculovirus expression system: A step for subunit vaccine development against H3N8 equine influenza virus. Open Veterinary Journal. 2024. DOI: https://doi.org/10.5455/OVJ.2024.v14.i1.32

[6] Branda F, Yon D, Albanese M, Binetti E, Giovanetti M, Ciccozzi A, et al.. Equine Influenza: Epidemiology, Pathogenesis, and Strategies for Prevention and Control. Viruses. 2025. DOI: https://doi.org/10.3390/v17030302

[7] Ricci I, Tofani S, Lelli D, Vincifori G, Rosone F, Carvelli A, et al.. First Reported Circulation of Equine Influenza H3N8 Florida Clade 1 Virus in Horses in Italy. Animals. 2024. DOI: https://doi.org/10.3390/ani14040598

[8] Boukharta M, Kasmi Y, Zakham F, Amri H, Ennaji MM. Prediction of genetic mutations of equine influenza virus related to adaptive determination nuclear export ribonucleoprotein complex. Journal of Basic and Applied Zoology. 2019. DOI: https://doi.org/10.1186/S41936-019-0099-X

[9] Boukharta M, Kasmi Y, Zakham F, Amri HE, Ennaji MM. Prediction of genetic mutations of equine influenza virus related to adaptive determination nuclear export ribonucleoprotein complex. Journal of Basic and Applied Zoology. 2019. DOI: https://doi.org/10.1186/s41936-019-0099-x

[10] Boukharta M, Zakham F, Touil N, Elharrak M, Ennaji M. Cleavage site and Ectodomain of HA2 sub-unit sequence of three equine influenza virus isolated in Morocco. BMC Research Notes. 2014. DOI: https://doi.org/10.1186/1756-0500-7-448

[11] Olguin-Perglione C, Golemba M, Barrandeguy M. Molecular evolution of H3N8 equine influenza virus in Argentina. Journal of Equine Veterinary Science. 2016. DOI: https://doi.org/10.1016/J.JEVS.2016.02.158

[12] 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

[13] Yu J, Yao Q, Liu J, Zhou Y, Huo M, Ge Y. Concern regarding H3-subtype avian influenza virus. Frontiers in Microbiology. 2023. DOI: https://doi.org/10.3389/fmicb.2023.1327470

[14] Durham A. Choosing an equine influenza vaccine. A & A Practice. 2019. DOI: https://doi.org/10.1136/inp.l695

[15] Favaro P, Fernandes W, Reischak D, Brandão P, Silva SO, Richtzenhain L. Evolution of equine influenza viruses (H3N8) during a Brazilian outbreak, 2015. Brazilian Journal of Microbiology. 2017. DOI: https://doi.org/10.1016/j.bjm.2017.07.003

[16] Gamoh K, Nakamura S. Update of inactivated equine influenza vaccine strain in Japan. Journal of Veterinary Medical Science. 2017. DOI: https://doi.org/10.1292/jvms.16-0558

[17] Cullinane A. Equine influenza and air transport. Equine Veterinary Education. 2014. DOI: https://doi.org/10.1111/eve.12215

[18] Mitchell PK, Cronk B, Voorhees IEH, Rothenheber D, Anderson RR, Chan TH, et al.. Method comparison of targeted influenza A virus typing and whole-genome sequencing from respiratory specimens of companion animals. Journal of Veterinary Diagnostic Investigation. 2020. DOI: https://doi.org/10.1177/1040638720933875

[19] Daly J, Kembou-Ringert JE. Equine, Canine, and Swine Influenza (Orthomyxoviridae). . 2019. DOI: https://doi.org/10.1016/b978-0-12-809633-8.20978-3

[20] Boukharta M, Azlmat S, Elharrak M, Ennaji MM. Multiple alignment comparison of the non-structural genes of three strains of equine influenza viruses (H3N8) isolated in Morocco. BMC Research Notes. 2015. DOI: https://doi.org/10.1186/s13104-015-1441-0

[21] Whitlock F, Grewar J, Newton R. An epidemiological overview of the equine influenza epidemic in Great Britain during 2019. Equine Veterinary Journal. 2022. DOI: https://doi.org/10.1111/evj.13874

[22] 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

[23] 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

[24] Rendle D, Ivens P, Bowen M, Lawrence H, Newton R, Marr C, et al.. Equine influenza: A current reference for vets in practice in the UK. UK-Vet Equine. 2019. DOI: https://doi.org/10.12968/ukve.2019.3.s3.1

[25] Ahearne M, Pentzke-Lemus LL, Romano AM, Larsen E, Watson AM, O'Fallon EA, et al.. Disease progression, pathologic, and virologic findings of an equine influenza outbreak in rescue donkeys. Journal of Veterinary Internal Medicine. 2022. DOI: https://doi.org/10.1111/jvim.16563

[26] Gildea S, Lyons P, Lyons R, Gahan J, Garvey M, Cullinane A. Annual booster vaccination and the risk of equine influenza to Thoroughbred racehorses.. Equine Veterinary Journal. 2019. DOI: https://doi.org/10.1111/evj.13210

[27] Barbić L, Savić V, Kovacevic KD, Kapetan J, Stevanović V, Kovač S, et al.. Outbreak of equine influenza in Croatia in 2015 and post outbreak epidemiological situation. Veterinarski arhiv (Tisak). 2018. DOI: https://doi.org/10.24099/vet.arhiv.0033

[28] Shipman E. The Effects of Biannual Equine Influenza Vaccine on Performance in Adult Horses. Veterinary Evidence. 2019. DOI: https://doi.org/10.18849/VE.V4I3.196

[29] Tsachev I. Contribution LABORATORY AND FIELD STUDIES ON ACUTE PHASE RESPONSE OF HORSES AFTER VACCINATION AGAINST EQUINE INFLUENZA VIRUS AND EQUINE HERPES VIRUS 4 / 1. . 2015. DOI: https://doi.org/10.15547/tjs.2015.01.012

[30] Bambra W, Daly J, Kendall N, Gardner D, Brennan M, Kydd J. Equine influenza vaccination as reported by horse owners and factors influencing their decision to vaccinate or not.. Preventive Veterinary Medicine. 2020. DOI: https://doi.org/10.1016/j.prevetmed.2020.105011

[31] Garrett D, Montesso F, Prowse-Davis L, Britt S, Fougerolle S, Pronost S, et al.. Refinement of the Equine Influenza model: the benefits of individual nebulisation for experimental infection. Journal of Equine Veterinary Science. 2016. DOI: https://doi.org/10.1016/J.JEVS.2016.02.155

[32] Kawanishi N, Kinoshita Y, Reedy S, Garvey M, Kambayashi Y, Bannai H, et al.. A comparative evaluation of seven commercial human influenza virus antigen detection kits for the diagnosis of equine influenza.. Equine Veterinary Journal. 2025. DOI: https://doi.org/10.1111/evj.14500

[33] Sandybayev N, Strochkov V, Beloussov V, Orkara S, Kydyrmanov A, Khan Y, et al.. Evaluation of a novel real-time polymerase chain reaction assay for identifying H3 equine influenza virus in Kazakhstan. Veterinary World. 2023. DOI: https://doi.org/10.14202/vetworld.2023.1682-1689

[34] Dionísio L, Medeiros F, Pequito M, Faustino-Rocha AI. Vaccination and Laboratory Diagnosis of Equine Influenza in Portugal in Comparison with Other European Countries. Medical Research Archives. 2022. DOI: https://doi.org/10.18103/mra.v10i12.3360

[35] Brito B, Frost M, Webster J, To J, Kirkland PD. Quantifying the impact of sequencing depth and reference genome choice on metatranscriptomic detection of four bovine RNA viruses.. Research in Veterinary Science. 2026. DOI: https://doi.org/10.1016/j.rvsc.2026.106125

[36] Sapachova M. Comparative Analysis of Methods of Molecular Detection of Avian Influenza Virus. Online Journal of Public Health Informatics. 2017. DOI: https://doi.org/10.5210/OJPHI.V9I1.7739

[37] Kubacki J, Fraefel C, Bachofen C. Implementation of next-generation sequencing for virus identification in veterinary diagnostic laboratories. Journal of Veterinary Diagnostic Investigation. 2020. DOI: https://doi.org/10.1177/1040638720982630

[38] Wilson A, Pinchbeck G, Dean R, McGowan C. Equine influenza vaccination in the UK: Current practices may leave horses with suboptimal immunity. Equine Veterinary Journal. 2020. DOI: https://doi.org/10.1111/evj.13377