Seroprevalence and associated risk factors for vector-borne pathogens in dogs from Egypt

Background Dogs play an important role as reservoirs of many zoonotic vector-borne pathogens worldwide, yet reports of canine vector-borne diseases (CVBDs) in Egypt are scarce. Methods Serum samples were collected from pet dogs (n = 500) of the three most common breeds (German Shepherd, Rottweiler and Pit Bull) in five Governates of Cairo (n = 230), Giza (n = 110), Al-Qalyubia (n = 60), Al-Gharbia (n = 60) and Kafr El-Sheikh (n = 40) with a hot desert climate. The presence of antibodies to Anaplasma spp. (A. phagocytophilum, A. platys), Ehrlichia spp. (E. canis, E. chaffeensis, E. ewingii), Borrelia burgdorferi (s.l.) and Dirofilaria immitis were assessed using IDEXX SNAP® 4Dx® ELISA tests. For each pathogen, risk factors (i.e. geographical area, keeping condition, sex, age, breed, tick infestation, weekly sanitation of dog enclosures and application of ectoparasiticides) were evaluated by logistic regression approach. Results In total, 18.2% (n = 91, 95% CI 15.1–21.8) of dogs scored seropositive for at least one pathogen, the most frequent being Ehrlichia spp. (n = 56; 11.2%; 95% CI 8.7–14.3) followed by Anaplasma spp. (n = 33; 6.6%, 95% CI 4.7–9.1), Borrelia burgdorferi (s.l.) (n = 9; 1.8%, 95% CI 0.9–3.4) and D. immitis (n = 7; 1.4%, 95% CI 0.9–2.9). In the tested population, 15.4% (95% CI 12.5–18.8) of dogs were exposed to a single pathogen while 2.4 (95% CI 1.4–4.2) and 0.4% (95% CI 0.1–1.4) were simultaneously exposed to two or three pathogens, respectively. Major risk factors associated with VBDs were living outdoors (Anaplasma spp., P = 0.0001; Ehrlichia spp., P = 0.0001), female sex (Ehrlichia spp., P = 0.005), German Shepherd breed (Anaplasma spp., P = 0.04; Ehrlichia spp., P = 0.03), tick infestation (Anaplasma spp., P = 0.0001; Ehrlichia spp., P = 0.0001; B. burgdorferi (s.l.), P = 0.003; D. immitis, P = 0.02), irregular sanitation (Anaplasma spp., P = 0.0001; Ehrlichia spp., P = 0.0001; B. burgdorferi (s.l.), P = 0.002; D. immitis, P = 0.01) and not using ectoparasiticides (Anaplasma spp., P = 0.0001; Ehrlichia spp., P = 0.0001; B. burgdorferi (s.l.), P = 0.007). Conclusion To our knowledge, this is the first large-scale seroepidemiological study of CVBDs in Egypt. Considering that all of the detected pathogens are potentially zoonotic, effective ectoparasite control strategies, regular examination of pet dogs and successful chemoprophylaxis are advocated.

regions [6], including in the Middle East and North Africa (MENA), because of the favourable climatic conditions for the perpetuation of arthropod vectors and development of canine vector-borne pathogens (CVBPs) [7]. Among CVBPs, tick-borne Ehrlichia spp., Anaplasma spp., Borrelia spp. and mosquito-borne Dirofilaria spp. are of great importance for dogs [8].
Dogs are the main reservoir hosts for zoonotic gramnegative intracellular bacteria Ehrlichia canis, E. ewingii and E. chaffeensis [9,10]. Ehrlichia canis, the causative agent of canine monocytic ehrlichiosis (CME), is transmitted by Rhipicephalus sanguineus (s.l.) and is prevalent in dog populations worldwide [11,12]. The clinical outcome of ehrlichiosis varies from mild symptoms to fatal illness in the chronic phase depending on the strain, individual immune response and presence of concomitant infections [4]. Although dogs may show nonspecific signs (e.g., fever, depression, weakness, lethargy, anorexia, weight loss and a mucopurulent nasal discharge), asymptomatic Ehrlichia infection may also occur [13][14][15].
Human and animal infections with Anaplasma species are increasingly recognized as important, occasionally emerging and potentially fatal tick-transmitted diseases of humans and animals [16,17]. Among six recognized species in the genus Anaplasma, A. phagocytophilum, the causative agent of granulocytic anaplasmosis, is diagnosed in a wide range of warm-blooded hosts including dogs, cats, horses, sheep, goats, cattle, camels and humans [18,19]. Anaplasma platys is a common VBP of dogs in MENA [20,21] and has occasionally been detected in humans [22][23][24]. Dogs usually are silent carriers of the infection [25], with clinical signs (e.g. fever, lethargy, anorexia and thrombocytopenia) sometimes described [26].
Among bacteria of the genus Borrelia, which affect different animal species including humans, Borrelia burgdorferi (s.l.) species complex causes Lyme disease, which is considered a major zoonosis for which many animals species (e.g. reptiles, rodents, wild ruminants) are reservoirs and Ixodes spp. tick the primary vector [27,28]. In dogs, B. burgdorferi (s.l.) most often causes nonspecific signs (e.g., fever, apathy, lethargy, renal damage and lymphadenopathy) but also severe arthritis and neurological disorders [29]. However, in the endemic areas the majority of seropositive dogs do not present any clinical signs of the infection although they often remain persistently infected for approximately 1 year [30].
Dirofilaria immintis (Spirurida, Onchocercidae) is the causative agent of canine heartworm disease, which is transmitted through the bite of several mosquito species worldwide [31]. Dirofilariosis may also affect other mammals including humans, leading to the formation of pulmonary nodules, which may be often confounded with pulmonary carcinoma [32]. Although most dogs infected by D. immitis-specially in endemic areas-are asymptomatic microfilaremic reservoirs, clinical signs depend on several factors, such as adult worm burden and localization [33]. The distribution of canine dirofilariosis in the MENA region, especially in North Africa, is not well known because of the paucity of epidemiological studies [31].
The aim of the current study was to provide novel information on the seroprevalence and distribution of causative agents of monocytic ehrlichiosis, granulocytic anaplasmosis, Lyme disease and heartworm disease in dogs from five Governorates of Egypt and assess the risk factors associated with the infections.

Study area
Egypt is a transcontinental country spanning the northeast corner of Africa and southwest corner of Asia. It is divided into 27 Governorates; the large regions of the Sahara desert, which constitute most of Egypt's territory, are sparsely inhabited. The investigation was conducted in Cairo ( (Fig. 1). These Governorates essentially have a hot desert climate, which is classified as BWh by the Köppen-Geiger system.

Sample collection
During 2018 and 2019, blood samples (ca. 2 ml) were collected from the cephalic or saphenous veins of 500 dogs of the three most common breeds raised in Egypt, i.e. German Shepherd, Rottweiler and Pit Bull, admitted to veterinary clinics of five cities in different Governorates, namely Naser City in Cairo (n = 230), 6th of October in Giza (n = 110), Benha in Al-Qalyubia (n = 60), Tanta in Al-Gharbia (n = 60) and Kafr El-Sheikh in the governate of the same name (n = 40). Since other breeds of dogs are rarely kept as pets in Egypt, they were excluded from this study. Dogs were grouped according to age into four groups: < 1 year (G1); between 1 and 3 (G2); between 3 and 5 (G3); > 5 years (G4). Animal data (i.e. age, sex, breed, tick infestation, weekly sanitation of the dog enclosures and tri-monthly application of ectoparasiticides) were recorded.

Statistical analysis
Chi-square test and Fisher's exact test were used to compare seropositivity to each pathogen, and the results were considered significant if P ≤ 0.05. In particular, P-values were calculated with Fisher's exact test only for variables below five (keeping condition, sex, weekly sanitation of the dog enclosures, presence of ticks on the dog body and tri-monthly application of ectoparasiticides) for D. immitis and B. burgdorferi (s.l.); all other P-values were calculated with Chi-square test. Univariable logistic regression analysis was used to evaluate the association of prevalence of each pathogen and variables of location (five Governorates), keeping condition (indoors or outdoors), sex (male or female), age (three groups), breed (three breeds), weekly sanitation of the dog enclosures, tick infestation and trimonthly application of ectoparasiticides. Variables with P ≤ 0.05 in the univariable analyses were conducted to the multivariable models. To determine the risk probability of CBVDs, the odds ratio (OR) and confidence interval (CI) of significant variables were calculated using the multivariant logistic regression model. Multiple linear regression analysis was used to determine the possible multiple collinearities of different variables included in this study. The Hosmer-Lemeshow test was calculated to assess the goodness of fit for each model. Statistical analyses were carried out using SPSS software (ver. 24.0, IBM, USA).
The risk of exposure to pathogens was significantly associated with keeping condition, sex, breed, tick infestation, weekly sanitation of the dog enclosures and tri-monthly application of ectoparasiticides ( Table 2). In particular, the risk of infection with Ehrlichia spp., Anaplasma spp. and B. burgdorferi (s.l.) was significantly associated with living outdoors, and CME was most prevalent in female dogs. Regarding breed of dogs, German Shepherds showed higher seroprevalence of Ehrlichia spp., Anaplasma spp. and B. burgdorferi (s.l.). Importantly, a significantly higher chance of seropositivity to all CVBDs was observed in dogs that lived in enclosures that were not sanitated, did not undergo ectoparasiticide application and were infested with ticks (Table 3). No statistical association was found between CVBDs and other variables. Multicollinearity analysis showed strong correlations between seropositivity to Anaplasma spp. and Ehrlichia spp. and tick infestation, not receiving adequate hygienic care and ectoparasiticides where the variance inflation factor (VIF) was 15.665 and 25.117, respectively. However, such correlations were observed for seropositivity to B. burgdorferi (s.l.) and D. immitis, i.e. VIF was 1 and 1.082, respectively.

Discussion
Data presented indicate that dog populations (i.e. 18.2%) in Egypt are exposed to CVBP, therefore posing threats to their own health and to people. This is largely due to the wide distribution of Rh. sanguineus (s.l.), the most common tick species infesting dogs in Egypt and a vector of canine ehrlichiosis and anaplasmosis [40][41][42]. Though little information is available about the prevalence of CVBDs in the MENA region, dogs often act as reservoirs of VBPs with prevalence of 18.8% in Qatar [55], 24.5% in Saudi Arabia [24], 38.1% in Iraq [21], 46.9% in Iran [20], 69.7% in Algeria [48], 73% in Israel [56] and 83.8% in Morocco [46]. In Egypt CVBDs were mostly observed in urban Governorates (i.e. Giza, Cairo and Al-Qalyubia) where keeping pet animals is more common, also indicating that dogs and people in these regions are at higher risk of acquiring VBDs of canine origin.
The most prevalent pathogen diagnosed in this study was Ehrlichia spp. An E. canis seroprevalence of 41% was reported in dogs from Cairo and Alexandria [34] where DNA of E. canis was also detected in Rh. sanguineus (s.l.) ticks collected on dogs [45]. In addition, the presence of E. canis is supported by the wide distribution in Egypt of the brown dog tick R. sanguineus (s.l.) [40], which is the recognized vector for this species [57]. Although the employed test could detect exposure to E. canis, E. chaffeensis and E. ewingii, in a previous study only E. canis, but not E. ewingii or E. chaffeensis, was molecularly diagnosed in blood of 39 seropositive dogs from Cairo, Giza and Al-Qalyubia [58]. Nonetheless, since E. ewingii and E. chaffeensis have been diagnosed in dogs, ticks and human patients from different countries in the African continent, e.g. Cameroon, Mali, Uganda and South Africa [59][60][61],   further investigations should be carried out to characterize the species of Ehrlichia infecting dogs in Egypt.
To the best of our knowledge this is the first report of sero-reaction to A. phagocytophilum/A. platys in dogs from Egypt (6.6%). It seems that A. phagocytophilum infection of dogs is not prevalent in some regions of MENA, also considering that previous studies from Saudi Arabia, Qatar, Iraq and five regions of Iran failed to detect the infection [20,21,55,62]. However, both A. phagocytophilum and A. platys were detected in Rh. sanguineus (s.l.) blood-feeding on dogs in Cairo and Giza [44,45] and A. phagocytophilum in blood of five human patients in the Nile Delta [63]. The detection of nine dogs (1.8%) seropositive to B. burgdorferi (s.l.) suggests that the pathogen circulates in Egypt as indicated by previous studies where prevalence ranged from 23% [36] to 71.4% [37]. In addition, some tick species, e.g. Rh. sanguineus (s.l.), Rh. annulatus, Hyalomma excavatum, Hy. dromedarii, Amblyomma lepidum and Ornithodoros savignyi, scored molecularly positive for DNA of B. burgdorferi (s.l.) [36,64]. The circulation of B. burgdorferi (s.l.) in Egypt has also been demonstrated [36,63,65,66]. In addition, tick-borne relapsing fever (TBRF) caused by Borrelia persica, Borrelia microti, Borrelia latyschewii and Borrelia baltazardi is also endemic in MENA [67] including Egypt where relapsing fever Borrelia spp. were detected in Ornithodoros savignyi ticks and sera of camel, sheep, goat, cattle and buffalo [68,69]. In particular, B. persica, which is transmitted by Ornithodoros tholozani, can infect dogs and has been reported from Egypt, Israel, Iran, Pakistan and former USSR Asian republics including Uzbekistan [70,71]. Considering that yet available commercial point-of-care diagnostic kits such as SNAP ® 4Dx ® employed in this study do not detect TBRF in dogs, complementary tests for dogs living in the endemic areas are recommended.
The detection of seven dogs (1.4%) from Giza, Cairo and Al-Qalyubia seropositive to D. immitis corroborates an older report of microfilariae of D. immitis in blood smears from 8/19 imported German Shepherd dogs in Assiut Governorate, Upper Egypt [35]. In particular, antibodies to D. immitis were already detected from cats of Giza with prevalence of 3.4% [72]. While human cases with D. repens have been frequently reported in Egypt [73,74], D. immitis was described once in a patient [75]. Data for canine dirofilariosis by D. immitis in North Africa are limited to few reports from Tunisia [49,76], Algeria [77,78] and Morocco [46]. It has been suspected that D. immitis is absent from some Middle Eastern countries such as Israel where D. repens is present [79]. In contrast, weighted prevalence of D. immitis infection in dog populations of Turkey and Iran was estimated to be 11.32% and 11.45%, respectively [80]. However, in a recent study in five geographical regions of Iran testing 354 dogs, no microfilaremic dog was found [20].
According to our findings, dogs that were living outdoors had a higher risk of being seropositive to Ehrlichia spp., Anaplasma spp. and B. burgdorferi (s.l.) probably as an effect of the higher chances of being exposed to bites of the brown dog tick Rh. sanguineus (s.l.) and other tick species that are competent vectors of these pathogens [5]. Furthermore, as expected, dogs that did not receive adequate hygienic care and antiparasitic treatments were more likely to be affected by CVBPs. Regular application of ectoparasiticides with potent anti-feeding and fast killing effects, repellents, insect growth regulators and juvenile hormone analogues combined with environmental treatment to reduce the number of adult and juvenile ticks are key control measures in managing CVBDs [81]. Ectoparasiticides, available in several formulations, such as pour-on, spot-on, shampoos, sprays and collars, have long-lasting effects [82] and are highly recommended to dog owners to prevent CVBDs.
In this study, seropositivity to Ehrlichia spp. was more frequent in female dogs. In contrast, some studies have found higher seropositivity to CVBDs in males due to behavioural characteristics that cause greater exposure to vectors than females [83,84]. No sex-related correlation was recorded for other tested pathogens in this study.
German Shepherds showed higher seroprevalences of Ehrlichia spp. and Anaplasma spp. Although all breeds are prone to CVBDs, CME has been reported most frequently in the German Shepherds [85,86]. German Shepherd dogs and Siberian Huskies are predisposed to developing more severe clinical signs of ehrlichiosis; therefore, these breeds have a worse prognosis [14]. An experimental study showed that the cell-mediated immune response to a challenge with E. canis was reduced in German Shepherds compared to Beagle dogs [87]. Examination of other dog breeds in Egypt though are very rare, and "baladi" stray dogs will shed light on the true occurrence of CVBDs in the country.
As a limitation of the employed commercial kit, PCRpositive/antibody-negative and antibody-positive/PCRnegative dogs have been reported [88]. The first case could represent early infections before the development of antibody responses; the second case may represent cases of past infections that may have been treated or spontaneously resolved. Furthermore, this commercial kit detects both IgM and IgG antibodies against Erhlichia spp., Anaplasma spp. and B. burgdorferi (s.l.). Although chronic long-term bacteremia is characteristic for some rickettsial agents, long-term persistence of IgG could occur, not reflecting the real "infection" status in some animals. Hence, seropositivity values must be interpreted as current infection with or previous exposure to the pathogens under assessment. Finally, the prevalence of D. immitis infection in Egypt should be carefully considered as an alarm bell for the introduction of a parasite not common in Egypt but which is expanding its range of distribution in southern regions of the Mediterranean Basin such as southern Italy [89].

Conclusion
The present study is the first comprehensive study for CVBD pathogens that has been conducted in Egypt to our knowledge and confirms the presence of Ehrlichia spp., Anaplasma spp., B. burgedorferi (s.l.) and D. immitis in dogs of different regions. The connection between these VBPs and their arthropod vectors in Egypt remains largely unknown and warrants further investigation. Considering that all of the detected pathogens are known zoonotic pathogens, effective ectoparasite control strategies, regular examination of pet dogs and successful chemoprophylaxis are advocated.