Equine Granulocytic Anaplasmosis, A Neglected Disease: Risk Factors Associated with Prevalence of Antibodies in Equines

Laboratório de Sanidade Animal, Universidade Estadual do Norte Fluminense Darcy Ribeiro (UENF), Campos dos Goytacazes, RJ, Brazil, 28013-602 Laboratório de Reprodução e Melhoramento Genético Animal, Universidade Estadual do Norte Fluminense Darcy Ribeiro (UENF), Campos dos Goytacazes, RJ, Brazil, 28013-602 Instituto do Noroeste Fluminense de Educação Superior (INFES), Universidade Federal Fluminense (UFF), Av. João Jasbick, s/n, Aeroporto, Santo Antônio de Pádua, RJ, Brazil, 28.470-000


INTRODUCTION
Anaplasma phagocytophilum was reclassified as a gramnegative bacterium of granulocytic cells, including neutrophils The aim of the present research was to verify the prevalence of anti-A. phagocytophilum antibodies in Equidae from Rio de Janeiro state, Brazil, and identify possible risk factors for infection, taking into account geographic distribution, the management practices of the properties, and factors related to individual disease cases to contribute to the surveillance of this agent in the country and worldwide.

MATERIALS AND METHODS
Blood of 612 equines was collected from the jugular vein in serum tubes and maintained at +4°C. All sample procedures were performed according to the routine of the Municipal Institute Jorge Veitsman in the use of animals. The samples were centrifuged at 4.000 rpm for 10 min, and the separated sera were stored at −20°C. The sample size was selected based on the total number of 121,594 Equidae reported in the state [14], an expected prevalence of 17.03% and an acceptable prevalence of up to 25%. Of these sera, anti-A. phagocytophilum antibodies of the IgG class were searched through IFAT using the IFA Equine Antibody-Fuller Laboratories© kit, following the manufacturer's instructions, with the aid of a Zeiss immunofluorescence microscope (AXIOSKOP 40). Sera that reacted at a dilution of 1:80 were considered positive samples.
In all properties where the samples were collected, an epidemiological questionnaire was administered, giving preference to the veterinarians responsible for the herds, followed by the treaters or trainers and finally, the owners. To facilitate data organization, the inquiry was divided into four parts according to the classification of its variables, such as: (1) "geographic distribution" in mesoregions, municipalities and sample collection properties; (2) "management" in relation to the practices adopted in the properties, such as the breeding system, production system, water supply, quality of management, type of food, food supplementation, mineral supplementation, regular veterinary assistance, vaccination, hygienic and sanitary conditions, receipt of animals for breeding, reproductive management and acaricide application; (3) "signs of infection" based on questions regarding factors related to the disease, such as history of anemia in the property, presence of dogs, association with breeding of small ruminants, association with cattle breeding, presence of wild animals, commercialization of Equidae, frequent trips and participation in events and/or gatherings; (4) "individualized inquiry" with information from each animal, namely, gender, age range, equine origin, equestrian activity and individual cases of anemia.
From the seroreactivity of the animals and the data collected in epidemiological records, risk factors were evaluated. To verify the relative risks (Rr) relating to the EGA and the various variables, the non-parametric chi-square test and the Fisher exact test were used to determine the odds ratio. To determine the real risk factors when considering multiple factors, the logistic regression was calculated with a 95% confidence interval using the SAS program [15].

Geographic Distribution
Out of the six mesoregions analyzed, only the Northwest was characterized not as a risk factor but rather as a protective factor for AGE (Table 1). Regarding the municipalities, seven of them (Rio de Janeiro, Friburgo, Teresópolis, Rio Bonito, Campos, Barra do Pirai and Paulo de Frontin) were considered at risk; however, the municipality of Vassouras (3.8%) presented protective characteristics for the infection, and among all the investigated municipalities, São Francisco do Itabapoana was the only municipality where anti-A. phagocytophilum antibodies were not detected in the 10 (1.6%) investigated animals ( Table 1). Out of the 33 properties, only one (PEC-PM -6.7%) showed protective characteristics against A. phagocytophilum infection ( Table 1).

Management
Out of the 13 analyzed parameters related to management, four presented statistical significance regarding the prevalence of anti-A. phagocytophilum antibodies in equine sera, namely, food supplementation (P=0.0001), hygienic conditions (P=0.0136), reproductive management (P=0.0038), and application of acaricides (P=0.0428) ( Table 2).

Prevalence of Anti-A. Phagocytophilum Antibodies in Equidae
The Open Microbiology Journal, 2020, Volume 14 221 season were considered less susceptible than animals receiving no supplementation ( Table 2).
It was observed that 83.5% of equines received regular veterinary assistance, and 67.8% were vaccinated for at least six types of diseases, specifically, rabies, equine influenza, tetanus, encephalomyelitis, equine herpes virus, and leptospirosis ( Table 2).
Considering the hygienic and sanitary conditions of the facilities where the Equidae were reared, it was observed that animals reared in good (44.9%) and regular (47.1%) conditions are more susceptible to A. phagocytophilum infections than animals that live under excellent (3.9%) and poor (4.1%) conditions ( Table 2).
Properties that do not use any kind of reproductive technology are more likely to present seroreactive animals. The use of acaricides has been found to be a protective factor; untreated animals are twice as likely to contract A. phagocytophilum infection ( Table 2).

Evidence of Infection
Regarding the presence of factors that may favor the infection, no significant differences were observed among the eight evaluated parameters: Anemia history on the property, presence of dogs, association with other creations, cattle breeding association, presence of wild animals, commercialization of equidae, frequent trips and participation events and / or gatherings ( Table 3).

Individualized Inquiry
Considering the Equidae individually, based on the five parameters evaluated, the origin of the Equidae and the individual case reports of anemia were the ones that presented significant differences (P=0.0098) in their variables regarding the prevalence of anti-A. phagocytophilum antibodies ( Table  4). Animals born on the property and those from another state showed significant differences (P=0.0005) in A. phagocytophilum infection compared with animals of other origins ( Table 4). Out of the 612 horses that were evaluated, 101 animals with anemia and 511 without anemia were observed. A lack of anemia is considered a protective factor for the presence of anti-A. phagocytophilum antibodies ( Table 4).
In the present research, it was found that race and species influence A. phagocytophilum infection in Equidae in the state of Rio de Janeiro, although no seroreactive Pampa and Paint Horse animals were observed ( Table 5).  Superscript equal letters in parentheses in the column do not differ statistically.

Multivariate logistic Regression
Taking into account the prevalence of anti-A. phagocytophilum antibody seroreactivity and the negative correlations of the influence of one factor on others, it was observed that property and origin ( Table 6) are the true risk factors for infection.

DISCUSSION
For the diagnosis of A. phagocytophilum infections, the IFAT-paired serology method was chosen because it is considered the "gold standard" by the World Health Organization in the search for anti-A. phagocytophilum antibodies [16], in addition to being the routine technique for serological monitoring of infection in reference laboratories when the objective is bacterium tracking in communities [17]. Moreover, the diagnosis of the bacterium in the peripheral blood is limited, even in acute infections, due to the short duration of the bacteremia, which makes the isolation in smears or molecular biology difficult. In the chronic phase, all these procedures are even more difficult due to the possible elimination of the bacterium from the organism over time [18,4,19].
The observed prevalence of 20.3%, distributed throughout the state of Rio de Janeiro, without observation of clinical signs of the disease, suggests that EGA is being underdiagnosed or confused with other diseases, such as piroplasmosis, equine infectious anemia, rhabdomyolysis, and leptospirosis. The prevalence observed in this study (22.3%) was observed by other authors [5] in Denmark, who also did not observe clinical signs of the disease. In addition, it is known that in a seroreactive population, the infection has been occurring for more than three weeks, and antibodies persist for a long period, therefore blood samples can be collected more than three months after infection initiation [20,12,5,16].
Both breeding and production systems were not determinant in A. phagocytophilum infection in horses in this research, although another study [21] observed in horses in Sweden that grazing time is correlated with positive seroreactivity to anti-A. phagocytophilum antibodies and that the extensive and semi-intensive systems presented higher seroreactivity than confined groups.
The quality of management used in the rearing and the origin of the water the Equidae received did not influence the prevalence of A. phagocytophilum infection. The result relating to water was expected because this agent is not water-borne; however, unlike in this study, another study [22] observed that horses reared in poorly managed conditions are more susceptible to disease.
It is known that some nutritional deficiencies and mineral imbalances in horses may lead to anemia, immunodepression, and concomitant diseases [23], but the types of food or mineral supplementation provided to animals were not determining factors in the infection by A. phagocytophilum in the equines of this research. The animals not supplemented during the dry season presented the highest rates of seropositivity among the groups surveyed, which may be associated with a malnutrition condition that may favor immunodepression and acquisition of opportunistic infectious agents, according to another study [24].
The presence of permanent veterinary care is a favorable factor for the prophylaxis and control of several diseases in farm animals [23]; however, the presence of veterinarians on the property and the number of vaccines applied were not determining factors in the control of A. phagocytophilum infection in animals.
Seroreaction to anti-A. phagocytophilum antibodies was detected in 23.7% of Equidae on properties that frequently received breeding animals, with no significant difference related to this practice (P=0.3624), and a significant difference (P=0.0038) in seroreaction prevalence was observed in Equidae (33.2%) that did not receive any type of reproductive management, a result also observed in another study [5].
In Brazil, especially in the Southeast region, the records found of tick-borne diseases are for borreliosis and spotted fever, which have a mortality of up to 30% in humans [25]. These same authors mention that these diseases have worldwide relevance, especially in countries of the northern hemisphere. In this research, it was observed that most horses that were analyzed received tick treatment (86.9%), and a significantly higher prevalence (P=0.0428) of seroreactivity was observed within the untreated group than within those treated with ectoparasiticides. In addition, tick treatment was considered a protective factor, and it was observed that untreated animals are twice as likely as treated animals to contract EGA. Another study [24] observed that horses from various regions of Central America that are diagnosed as positive for anti-A. phagocytophilum antibodies are parasitized by the same tick species, Amblyomma sculptum (synonymy A. cajennense) and A. nitens, which are also reported to parasitize the Equidae from the Southeast region of Brazil. The maintenance of the A. phagocytophilum biological cycle is known to depend on tick density and the tick species involved [26]. The 16.5% prevalence of seroreactivity in treated animals in relation to the total analyzed may suggest incorrect antiparasitic control contributing to tick resistance and vector permanence in the herd [27,28]. A collective history of anemia is not a useful parameter for suspected EGA in the herds, as the difference in the prevalence of seroreaction in anemic compared with nonanemic herds was not statistically significant (P=0.2097). Similar results were also observed in another study [4] when evaluating horses that were experimentally infected with A. phagocytophilum for four months. All animals developed acute clinical signs such as fever, inappetence, ataxia, and hind limb edema; however, anemia was not a problem, generally speaking. However, this variable was characterized as a risk factor when individually assessed, contrary to the observation for the history of anemia in the properties evaluated in the present research.
Dogs carrying A. phagocytophilum have already been detected in Rio de Janeiro [29]; in addition, the importance of dogs as a source of the infection for the vectors is known. However, in this research, there was no characterization of the presence of this animal species on equine farms as significant for A. phagocytophilum infection (P=0.3814).
As for conviviality with small ruminants, the findings of this research did not recognize this factor as significant for the prevalence of seroreaction against A. phagocytophilum (P=0.2500); however, other authors, when studying the prevalence of this pathogen in properties of Italy, observed that horses raised in association with small ruminants present significantly higher rates of infection [22]. Specifically to the presence of cattle, although 27.4% of the equines examined in this study live with these animals, this variable was not significantly (P=0.1792) associated with the positivity index, suggesting that the cattle were not relevant as hosts or reservoirs of A. phagocytophilum in the investigated properties. Another study observed the presence of Rhipicephalus microplus and non-infection by Ixodes ticks in cattle when integrated or alternating grazing between these and horses [24], reporting that integrated breeding of these two species does not pose a risk to Equidae in terms of the prevalence of infection by A. phagocytophilum.
In the epidemiological survey in this research, the presence of wild animals was observed in 95.6% of the investigated properties, with 19.1% of Equidae being seroreactive ( Table 3). According to other authors [30], even in regions where the prevalence of EGA is low, tick monitoring, prevention, and control programs should be established, mainly in areas near forests where there are several species of the genus Ixodes, which infest the wild fauna. According to other studies, ticks of genera Ixodes and Amblyomma are also commonly found in wild rodents, lagomorphs, cervids, marsupials, carnivores, reptiles, and birds [31]. However, the presence of wild animals was not considered a significant factor (P=0.4539) for the presence of A. phagocytophilum infection in the Equidae of Rio de Janeiro state.
Although commercialization, frequent travel, and participation in events and/or gatherings are stress factors for animals and can contribute to immunodepression and consequently infectious diseases [21,32], the respective prevalences of 13.1%, 16.7%, and 17.7% of seroreactivity associated with these variables in this research indicated that these factors are not significant in A. phagocytophilum infection (P=0.1406, P=0.9752, P=0.1680, respectively); however, these activities were considered a risk factor for disease in horses [21].
Out of the 612 Equidae investigated from the state of Rio, there was no significant difference in seroreactivity for A. phagocytophilum between sexes (P=0.0984), corroborating the findings of authors of a study [33], who also did not observe differences in infection according to the sex of the animal in 733 horses studied in the region of Minnesota and Wisconsin (USA).
By stratifying the examined equine herd according to age group, it was found that the highest numbers of seroreactive animals corresponded to foals and adult animals, evidencing that any individual is subject to reinfection with detectable serological titers, with no statistically significant difference in the prevalence of seroreactivity of the animals (P=0.8586). However, a study found a significant correlation (P≤0.03) between age over 11 years in horses and the presence of anti-A. phagocytophilum antibodies [5], and another one concluded that the prevalence increases with age [32]. Thus, age constitutes an important risk factor due to the higher prevalence of tick exposure throughout an animal's life.
Equidae from the property and from other states presented a significantly higher prevalence of seroreaction to A. phagocytophilum, when compared to animals from other origins (P=0.0005). This is one of the most important variables in this study because the logistic regression analysis shows the possibility of foci of infection located in properties or microregions, as observed for other diseases, such as equine infectious anemia [34].
Seroreactive animals were found for all equine aptitude categories evaluated in this research. However, no aptitude category was significant for seroreaction against A. phagocytophilum, which further raises concern about this zoonosis with possible EGA dispersal because the variety of equestrian activities intensify equine transit, which may favor the spread of the disease [25], including its spread to humans.
In this research, the prevalence of seroreactive animals in properties with a history of anemia was not significant (P=0.2097); however, anemia, as individually assessed in equines, was significantly associated with infection (P=0.0098), and the absence of anemia in animals was a protective factor against infection. In addition, another study stated that A. phagocytophilum is responsible for important hematological complications, such as alterations in erythropoiesis, suppression of erythroid progenitor cell proliferation, erythroid cell apoptosis and reduced iron availability, which may contribute to the development of anemia and cytopenias [35].
Regarding the equine breeds analyzed from this research in the state of Rio de Janeiro, significant differences in seroreactivity for anti-A. phagocytophilum antibodies were observed among breeds (P=0.0401). In contrast, other authors did not observe significant differences when evaluating 23 equine breeds in the United States [33]. According to other studies [33,21], horses used for racing are an important risk factor. Another risk factor to consider is that genetically pure breeds, such as English thoroughbred, Holstainer, and Mangalarga, are more sensitive to tick infections [36]. Although studies in asinines and mules are scarce, a study found 6% seroreactivity (6/100) in asymptomatic asinines tested in Italy [37]. This result, together with the 22.2% (2/9) seroreactivity in the mules surveyed in the state of Rio de Janeiro, corroborates that this animal species also participates in the epidemiological chain of EGA.
Comparing the distribution of the prevalence of seroreaction against A. phagocytophilum in the properties and its prevalence according to the origin of the equines and to the other variables where significance was observed in the seroreaction observed in this research, multivariate logistic regression analysis indicated that the distribution of infection on properties (P<0.0001) and the origin (P=0.0095) of Equidae were the real risk factors for A. phagocytophilum infection. These results corroborate those in researches conducted in Europe [21,32].

CONCLUSION
After analyzing the epidemiological factors for the presence of anti-A. phagocytophilum antibodies in Equidae from the state of Rio de Janeiro, it can be inferred that EGA is present in the equine herds of the state in an enzootic manner. Although the agent is widely distributed in all mesoregions of the state of Rio de Janeiro, the focus on properties is important in the permanence of the disease in the herds. Thus, attention should be paid to the transit of Equidae, mainly from other states, to prevent the introduction of the etiological agent A. phagocytophilum to the property and prevent it from becoming a focus of the disease. This research indicates that mules are also important as hosts or reservoirs, therefore investigations should be developed to better clarify the importance of these animals in the epidemiological chain of EGA.

AUTHOR CONTRIBUTION STATEMENT
Francisco C. R. de Oliveira: Prepared the tables and wrote the main manuscript text.

ETHICS
APPROVAL AND CONSENT TO

PARTICIPATE
The authors confirmed that the study does not have an ethical committee number, because it was a routine exams. They collect blood samples for routine examinations.

HUMAN AND ANIMAL RIGHTS
Not applicable.

CONSENT FOR PUBLICATION
Not applicable.

FUNDING
The study was funded by the CNPq (Brazil's National Council for Scientific and Technological Development), CAPES (Coordination for the Improvement of Higher Education Personnel) no 001 with, FAPERJ (Research Support Foundation of the State of Rio de Janeiro).