Seroprevalence of Canine Ehrlichiosis and Microscopic Screening for Canine Babesiosis in Dogs in Harare, Zimbabwe, 2016-2017

A cross-sectional study was done to determine ehrlichiosis seroprevalence and babesiosis prevalence in dogs that were presented to selected veterinary clinics in Harare. Sera from randomly selected dogs were tested for antibodies to Ehrlichia spp. using an enzyme-linked immunosorbent assay while microscopy of peripheral blood smears was used to confirm babesiosis. Overall, 75.2% (88/117, 95% CI: 66.2–82.5) of sera samples tested were positive to Ehrlichia spp. antibodies while the prevalence of canine babesiosis was 47.9% (56/117, 95% CI: 38.6–57.3). Age, breed, and sex were found not to be associated with the two disease conditions (p > 0.05). Most of the dogs with babesiosis (82.1%, 46/56) were also positive to Ehrlichia spp. antibodies. Hypoalbuminaemia (53.8%, 63/117), anaemia (53.0%, 62/117) and thrombocytopaenia (40.2%, 47/117) were the most common laboratory findings. Thrombocytopaenia and hypoalbuminaemia was more pronounced in dogs with babesiosis only while anaemia was more marked in dogs with babesiosis and positive to Ehrlichia spp. antibodies.


Introduction
Canine ehrlichiosis, a potentially fatal disease of dogs is caused by Ehrlichia species. e disease has acute, subclinical and chronic stages [1] and clinical findings in dogs vary with the stage of the infection [2]. Clinical signs observed in the acute phase of the disease include fever, anorexia, oculo-nasal discharges, vomiting, weight loss, hepatosplenomegaly, lymphadenopathy and, rarely epistaxis and haemorrhage [1,3]. e chronic phase is marked by epistaxis, haematuria, petechiae, ecchymosis distributed over skin surface, respiratory distress, ocular abnormalities, and CNS signs [1]. Previous canine ehrlichiosis studies in Zimbabwe showed an overall seroprevalence of 42% [4]. Dogs are naturally infected with several species; E. canis, E. chaffeensis, and E. ewingii [5] with E. canis being the most common and causing the most severe clinical disease in Africa and Asia [6]. Although several Ehrlichia spp. are able to cause natural disease in dogs, only E. canis and E. ruminantium are known to occur in southern Africa [6]. However, serological evidence of antibodies against E. chaffeensis, E. ewingii, and E. ruminantium from dog sera in South Africa and Zimbabwe has been documented [7][8][9][10]. Some studies from Venezuela and Costa Rica have suggested that E. canis might be zoonotic [11,12].
Canine babesiosis is a disease of worldwide significance that causes fever, haemolytic anaemia, haemoglobinuria and death [13]. It is an important disease of domestic and wild canidae [14]. e most common clinical signs associated with babesiosis are anorexia, fever, depression/lethargy, pale mucosae, splenomegaly, and weight loss [13]. Canine babesiosis studies in Zimbabwe are limited, with two studies reporting a prevalence of 6.9% and 26% [4,15]. e disease is caused by three strains of the large Babesia canis namely, B. canis, B. rossi, and B. vogeli; the small B. gibsoni and the microbabesiae; B. microti, B. vulpes and B. conradae [16]. In Africa, the small-sized Babesia has been reported in East and North Africa [13,17,18] with the rest of Africa reporting the large-sized Babesia [13,19,20] and there is no report of the microbabesiae [21,22]. However, currently there is no literature on Babesia spp. infecting dogs in Zimbabwe.
Dogs can have concurrent infections with various Babesia, Bartonella, Ehrlichia and Rickettsia species [23] and those with a heavy tick exposure can be infected at a higher rate with multiple and potentially zoonotic tick-borne pathogens [23]. Worldwide, tick-borne diseases are an important cause of morbidity and mortality in dogs with the brown dog tick, Rhipicephalus sanguineus being implicated as a vector of E. canis, B. vogeli, and B. gibsoni [1,22]. Hence, the transmission occurs when R. sanguineus takes a blood meal from the dog [1,2]. Concurrent infections of E. canis with Babesia spp. have been reported [4,24,25] leading to more severe case outcomes [26]. e epidemiology of canine tick-borne diseases may change due to the effects of climate change and the ease of international travel [27]. Studies about the prevalence of Babesia and Ehrlichia spp. co-infections in dogs in Zimbabwe are limited [4]. e first objective of this study was to determine the seroprevalence of ehrlichiosis and the prevalence of babesiosis.
e second objective was to determine the prevalence of Ehrlichia spp. seropositivity in dogs with babesiosis and the common clinicopathological findings.

Study Location, Design and Collection of Blood Samples.
is study was conducted in urban Harare, Zimbabwe where a cross-sectional study was employed to collect blood samples from dogs between October 2016 and March 2017. e blood samples were collected from dogs presented for routine elective surgery or ill-health at randomly selected private veterinary practices. A systematic random sampling technique was used to select dogs presented at the selected private veterinary practices; the first dog being selected using simple random sampling and every tenth dog therea er. e selected dogs were restrained manually and whole blood was collected from the cephalic vein into 5 ml plain and ethylene diamine tetra-acetic acid (EDTA) tubes. Serum obtained through centrifugation at 2500 rpm for 10 minutes using a Sigma 3E-1 centrifuge (Sigma Harz, Germany) was stored at −20°C prior to use for Ehrlichia spp. serological testing. e EDTA blood was used for complete blood counts and microscopy of peripheral blood from an ear vein was used for detection of Babesia piroplasms from Giemsa-stained thin blood smears. Data recorded during blood samples collection included age, sex, and breed of the dog.

Testing for Ehrlichia spp. Antibodies and Babesia
Piroplasms.
e ImmunoComb® Canine Ehrlichia Antibody Test Kit (Biogal-Galed Laboratories, Israel) was used to detect IgG antibodies against Ehrlichia spp. from the collected dog sera. e test was carried out according to the manufacturer's instructions (http://www.biogal.co.il) and as previously described [28,29]. e results were read with a calibrated colour Comb Scale (graded S0-S6), which was provided with the test kit. A scale of S3, which is equivalent to a positive immune response at a titre of 1 : 80 by an indirect fluorescent antibody (IFA) test, was considered as the "cut-off " level of IgG antibodies (http://www.biogal.co.il). Hence, in this study serum samples with a Comb Scale score of ≥S3 (i.e. ≥1 : 80 titre) were considered to be positive for Ehrlichia spp. antibodies.
Microscopy was used for the detection of Babesia piroplasms. Giemsa-stained, thin peripheral blood smears were prepared from peripheral blood from an ear vein. Two peripheral blood smears were made and from each, a minimum of 100 fields was microscopically examined for the presence of Babesia piroplasms. Careful examination of the Giemsastained thin peripheral blood smears was done by well-trained and experienced technologists.

Haematology and Clinical
Chemistry. EDTA blood of dogs testing positive for Babesia and Ehrlichia spp. antibodies was subjected to a complete blood count using an automated haematology analyzer (BC-2008 Vet-Shenzhen Mindray Bio-medical Electronics, China). e blood samples were analyzed to measure haematocrit (HCT), total number of erythrocytes (RBC), haemoglobin concentration (HB), mean erythrocyte volume (MCV), mean corpuscular haemoglobin content (MCH), mean haemoglobin concentration in erythrocytes (MCHC), erythrocyte distribution width (RDW), and total number of platelets (PLT). e leukogram measurement included total number of leukocytes (WBC), absolute number of neutrophils (NEU), lymphocytes (LYM), monocytes (MONO), and eosinophils (EOS). An automated chemistry analyzer (Humastar 180®-Human GmbH, Wiesbaden, Germany) was used to measure albumin and globulin. Reference values for all the parameters were adopted from Diagnopath Veterinary laboratory (Pvt) Ltd, Harare.

Data Analysis.
e recording and editing of the data was performed using Microso Excel®. A statistical package called Epicalc (2000) version 2, was used to analyze all the raw data. e total number of positive/seropositive animals was calculated from the total number of samples tested over the study period and expressed as a percentage. Positive/ seropositive animals were examined in relation to sex, breed, and age. e Chi-square test was used to measure differences in proportions between generated categories of babesiosis and ehrlichiosis status and values of 푝 < 0.05 were considered as significant. Seropositivity/positivity was also analyzed according to HCT, RBCs, MCV, hemoglobin, MCHC, platelets, WBCs, total protein, and albumin. Association between the above blood parameters was evaluated by calculating the Chisquare test, relative risk and the 95% confidence interval using Epicalc version 2.

Ethical Considerations
Ethical approval for use of dogs and for all protocols in this study was obtained from the Ethical and the Higher Degree committees of the Faculty of Veterinary Science reference number VEHDC 2016/05. e purpose of this study was well explained to the veterinary practitioners stationed at all the veterinary clinics in Harare, who all expressed consent to participate in the study. Verbal and written consent was obtained from owners whose dogs were selected for the study. Standard operating procedures were followed for collection of blood samples.

Haematological Findings.
ere were no significant differences 푝 > 0.05 noted on the erythrogram and leukogram of apparently healthy and ill dogs. e changes of the erythrogram and leukogram are presented in Table 3 Veterinary Medicine International 4 indicative of exposure and not necessarily acute infection. Since no definitive diagnosis was determined, the presence of other diseases could have masked the observed erythrogram and leukogram findings. Hence, the findings of this study should be viewed in the light of its limitations. e most frequently used diagnostic technique for Ehrlichia spp. infection confirmation in the dog is serology [2]. Kahn et al. [31] showed that the indirect fluorescent antibody (IFA) and ELISA tests were equally sensitive for the early detection of IgG antibodies against E. canis. e ImmunoComb® Canine Ehrlichia Antibody Test kit has a very high sensitivity (100%) and a high specificity (94.1%) (http:// www.biogal.co.il) and hence, reduces the possibility of false positive and false negative reactions. Based on the above observations and its use in earlier studies [2,29], the ImmunoComb® Canine Ehrlichia Antibody Test kit was therefore used to determine IgG antibodies to Ehrlichia spp. in dogs in the study area. However, the use of serology was a limitation of the study; it is indicative of exposure and not necessarily acute infection, hence the stage of infection could not be determined. In addition, ehrlichiosis seroprevalence could probably have been over-estimated as dogs with Ehrlichia infection will self-cure and remain seropositive for variable periods therea er [6]. Cross-reactivity between antibodies against Ehrlichia spp. has been reported [7][8][9] and in areas where they co-exist their serological differentiation may therefore not be possible and this is a limitation to our study. e use of molecular techniques (PCR and sequencing) using parasite-specific primers would have provided a better diagnostic tool in terms of both sensitivity and specificity and along with the identification of the infecting species.
Blood smear examination is a useful diagnostic tool for clinical babesiosis in dogs and microscopy evaluation continues to be the easiest and most accessible diagnostic test for most laboratories [32]. Although microscopy is highly specific and can be used to diagnose the large forms of Babesia, the platelets count was lowest for those positive for Babesia spp. only (116.7 ± 21.5), followed by dogs positive to both Babesia and Ehrlichia spp. (134 ± 18.7) and lastly for those positive to Ehrlichia spp. antibodies only (139 ± 22.3) but the differences were not significant 푝 > 0.05 .
Eighty percent of dogs with babesiosis only, 69.6% positive to both Babesia and Ehrlichia spp. and 45.2% with ehrlichiosis only had hypoalbuminaemia as measured by a decreased value of albumin. e difference in percentages was significant 푝 < 0.05 between the dogs positive to both Babesia and Ehrlichia spp. and those positive for Ehrlichia spp. antibodies only. e hypoalbuminaemia was more pronounced in those positive to Babesia only (21.9 ± 1.5) compared to dogs positive to both Babesia and Ehrlichia spp. (23.8 ± 0.7) and Ehrlichia spp. antibodies only (24.2 ± 0.8).

Discussion
e lack of significant differences noted on the erythrogram and leukogram of apparently health and ill dogs was an unexpected finding. is was probably because the ill dogs had a low grade of babesiosis and/or ehrlichiosis infection. Due to lack of clinical details and follow-up of ill dogs, the intensity of babesiosis and/or ehrlichiosis infection could not be determined. In addition, the Babesia spp. parasite load was not determined and hence, an inability to assess the intensity of infection. Babesia vogeli causes a mild to moderate disease associated with mild laboratory changes [13]; it is possible that the ill dogs were infected by B. vogeli but the current study did not determine the Babesia spp. in babesiosis positive dogs. In naturally infected dogs, mild haematological abnormalities occur in the subclinical phase of ehrlichiosis [6,30]; it is therefore likely that apparently health dogs had a subclinical infection whilst the ill ones had a low grade ehrlichiosis infection. It is also important to note that ehrlichiosis serology is hypoalbuminaemia was constantly observed in dogs either positive to one or both pathogens. rombocytopaenia and hypoalbuminaemia was more pronounced in dogs positive to Babesia spp. only while anaemia was more marked in those positive to both Babesia and Ehrlichia spp. e severity of anaemia and thrombocytopaenia in Babesia-Ehrlichia concurrent infections was intermediate to that of individual infections [4]. In Babesia-Ehrlichia mixed infection the disease was reported to be more severe [42]. Manzillo et al. [43] reported that anaemia and thrombocytopaenia are more common in dogs with babesiosis than those with ehrlichiosis. Single infections with B. gibsoni or B. canis gave more severe haematological results than mixed infections [14]. e observed differences in severity of the laboratory abnormalities are likely dependent on Babesia/Ehrlichia spp. in individual and concurrent infections. Hence, it is difficult to ascribe haematological abnormalities to any of the Babesia/Ehrlichia spp. Tsachev et al. [37] indicated that variations in haematological profiles in Ehrlichia infected dogs may be related to differences in the virulence of Ehrlichia spp. strains, antigen heterogeneity of this bacterial agent and the clinical form of the disease. In this study, no particular stage of ehrlichiosis was selected for evaluation. e laboratory abnormalities observed, particularly anaemia and thrombocytopaenia can affect the outcome of routine surgeries [30]. Hence, before such procedures are done, testing for Ehrlichia/Babesia in dogs from the study area should be considered. According to Kelly [6], although apparently healthy, dogs subclinically infected with E. canis usually have laboratory abnormalities, especially thrombocytopaenia, anaemia, hypoalbuminaemia and leukocytosis singly or in combination. In our study, the dogs positive to Ehrlichia spp. antibodies were apparently healthy and had the above mentioned laboratory abnormalities further supporting that many natural E. canis infections are subclinical [26].
In conclusion, this study showed a high prevalence of Ehrlichia/Babesia presence in the study dogs. e haematology and biochemical profiles are similar to the results observed in other reports with the most significant abnormalities being anaemia, thrombocytopaenia and hypoalbuminaemia.
rombocytopaenia and hypoalbuminaemia was more pronounced in dogs with Babesia only while anaemia was more marked in those positive to both Babesia and Ehrlichia spp. However, there were no laboratory abnormalities which could be utilized to differentiate between individual and concurrent presence. Further studies are required to determine the Babesia and Ehrlichia spp. present in dogs in the country.
Data Availability e data used to support the findings of this study are available from the corresponding author upon request.

Conflicts of Interest
e authors declare that they have no financial or personal relationships(s) that may have inappropriately influenced them in writing this article. small piroplasms are hard to observe by light microscopy which has poor to moderate sensitivity and expertise is needed [22,33]. In addition, the limit of parasites detection in a thin blood smear is reported to be parasitemias of 0.5% [34]. For these reasons, our study might have underestimated the prevalence of babesiosis in the studied dogs and more sensitive molecular PCR-based methods should be used in future studies.
Our study found a higher ehrlichiosis seroprevalence (75.2%) than that previously (52%) recorded [4]. Similarly, the current study reported a higher canine Babesia prevalence (47.9%) than an earlier report of 26% [4]. A recent PCR-based study on Babesia in apparently healthy rural dogs in the country also found a low prevalence (6.9%) of Babesia spp. antigen [15]. e prevalence was also much higher than that reported by Rautenbach et al. [35]. Differences in study areas and in testing regimes (e.g. for Ehrlichia spp.) could probably account for the variations. Babesia spp. identification was a limitation of the present study and there is no literature on the species occurring in the country. However, B. gibsoni and B. rossi have been confirmed in neighbouring Zambia [14] while B. rossi and B. vogeli are known to occur in South Africa [13]. Further studies are required to determine the tick vectors and, the Babesia and Ehrlichia spp. present in dogs in the country and also to assess their roles in the clinical and pathological picture of the diseases. In support of earlier observations in the country [4], co-presence was shown with over 80% of Babesia spp. positive dogs also being seropositive to Ehrlichia spp. antibodies. e ehrlichiosis/babesiosis concurrent prevalence found in this study (39.3%) was more than twice that was previously reported (17%) in the country [4] and this phenomenon has been reported elsewhere [25,[36][37][38]. Rhipicephalus sanguineus, the brown dog tick transmits both parasites [37] and this is likely to account for the observed findings.
Anaemia and thrombocytopaenia have been demonstrated as consistent laboratory findings in canine babesiosis and ehrlichiosis [13,25,36,39,40]. In an earlier study in the country, the two were the most common laboratory abnormalities observed in dogs with Babesia and Ehrlichia infection [4]. Despite having no signs of the disease, Babesia and Ehrlichia positive dogs in the current study had anaemia and thrombocytopaenia. However, thrombocytopaenia was observed in less than a third of dogs with either babesiosis or ehrlichiosis but in more than half of those with both. In Zambia, thrombocytopaenia was also observed in less than one-fi h of Babesia infected dogs [14]. e low number of dogs with thrombocytopaenia might be probably due to the fact that serology was performed for ehrlichiosis which indicates past exposure rather than active infection.
In this study, leukocyte abnormalities were nonspecific and most dogs had eosinopaenia and monocytopaenia while neutrophilia, lymphocytosis, and leukocytosis were observed in some of the dogs. Niwetpathomwat et al. [36,40] also observed nonspecific leukocyte abnormalities in Babesia and Ehrlichia infected dogs. e observed abnormalities of eosinopaenia, monocytopaenia and neutrophilia are in agreement with other studies for ehrlichiosis [37,39]. One of the predominant biochemical abnormalities found in dogs infected with E. canis is hypoalbuminaemia [38,41]. In this study,