Seroprevalence Survey of Equine Lyme Borreliosis in France and in Sub-Saharan Africa

E quine borreliosis is a vector-borne disease transmitted by ticks belonging to the Ixodes genus, specifically Ixodes ricinus in metropolitan France. Its causative agent, Borrelia burgdorferi sensus lato (s.l.), is a flagellated spirochete. It is responsible for a human multisystemic disease with a great diversity of clinical signs and several variations in the course of the disease (Strle and Stanek 2009). In short, the disease can be divided into three stages. During the first stage, there is a skin rash, known as erythema migrans, the most typical clinical feature of the disease. The second stage is characterized by central and/or peripheral nervous system involvement known as Lyme neuroborreliosis; cardiac disorders are also described in this stage. Finally, in the third stage, chronic arthritis develops, especially in the knees. In horses, most of the infections are subclinical. The most frequently associated signs are arthritis and uveitis (Butler et al. 2005).

The objective of the study is to assess the seroprevalence of B. burgdorferi infection in French and African horses and to detect risk factors. In the following, B. burgdorferi will be used for B. burgdorferi (s.l.), which mainly includes in Europe Borrelia burgdorferi (s.s.), Borrelia afzelii, and B. garinii. The field work was carried out from 2005 to 2006. Blood samples (n = 570) were collected in 13 sites in metropolitan France, French Guiana, Ivory Coast, Chad, Gabon, Djibouti, and Congo (Democratic Republic). On each site, 30 horses were collected if the number was more than 30 horses. Otherwise, all horses were selected. All the horses were healthy at the time of sampling. The horses were born and bred in rural areas of northwest and center of France and then transferred, at the age of 3, in their definitive facilities in different suburban environments. The blood samples were tested using an ELISA Dot-blot method (SNAP^® 4 Dx; IDEXX Laboratories, Westbrook, ME). The results were compared by using chi-square test (α = 5%).

The results are shown in Table 1. There is no positive horse in Africa or in French Guiana. The seroprevalences in the eastern, center-western, and southern regions of metropolitan France are, respectively, 48%, 31%, and 12%. The prevalence among males, females, and geldings are, respectively, 17%, 34%, and 33%. Thirty-eight percent of the horses and 32% of the ponies tested positive. The classification in different age-classes (0–5 years old, 6–10 years old, 11–15 years old, 16–20 years old, and more than 21 years old) showed seroprevalences of 28%, 30%, 35%, 36%, and 38%, respectively. The statistical comparison shows no difference due to age, gender, or type (horse/pony). Some locations in the northeast are more predisposed to have seropositive horses (p < 0.01): Mourmelon and Bitche. It is probably due to the ability of the horses to be outside in paddocks. In contrast, horses in the south region are less seropositive, probably due to a smaller presence of the vector in this hot and dry area.

In Europe, B. burgdorferi circulates in endemic areas between I. ricinus ticks and a large number of vertebrate hosts upon which ticks feed (Wilske 2005). The vector of equine borreliosis in metropolitan France is also I. ricinus. This hard tick lives preferentially in humid areas such as the borders of woods or fields and where big game is plentiful. In the south of France, the climate is quite hot and dry, but there are some humid places near rivers. The reservoirs of B. burgdorferi are small rodents and probably cervidae and some birds. In Poland, a national survey made on 395 horses showed a 26% seroprevalence (ELISA) (Stefancíková et al. 2008). In Turkey where the geographic distribution of Ixodes spp. is limited, the equine seroprevalence (n = 300) is only 6% (Bhide et al. 2008). Our study confirms that seroprevalence is higher in the northeast of France where climate is more humid than in the south. Our seropositive horses were all imported from the north of France; therefore they were probably infected before their arrival to the southeast. However, even though Ixodes ticks are rare in the southeast, they can be found in certain areas. In Africa and in French Guiana, where the vector is absent, the seroprevalence is 0%. In addition, our study confirmed the absence of circulation of B. burgdorferi in Africa and French Guiana. In Ivory Coast, although serologies were positive in man, no confirmed diagnosis (culture, molecular biology) was established (Kouassi-M'Bengue et al. 2003).

The commercial kit SNAP 4 Dx uses antigen-specific conjugate. Thus, it can be used with blood samples of other species than the canine recommended by the manufacturer. As a result, we used this test to assess the prevalence of antibodies against B. burgdorferi in horses. The findings indicated that the test kits had a fair sensitivity (63%) and a very high specificity (100%) for horses infected with B. burgdorferi (Johnson et al. 2008). The relevance of this simple and rapid test (20 min) resided in its association with the screening of Anaplasma phagocytophilum infection. During our survey, seroprevalence was of 19% in the east, 16% in center-west, and 0% in southeast, French Guiana, and Africa (Maurizi et al. in press). These two diseases are transmitted by the same vector. The Ixodes coinfection has already been demonstrated in France (Halos et al. 2006).

In conclusion, this study shows that around 33% of the horses in metropolitan France have antibodies against B. burgdorferi. That is why equine borreliosis is part of the differential diagnosis of fever with unknown origin that includes other tick-borne diseases such as babesiosis. On the epidemiological level, the horse is not suspected of being a reservoir. However, it is a good infection sentinel as ticks often stay attached to him more than 1 day. This regular monitoring of horses would detect epidemiological changes. Moreover, the direct transmission by urine of infected animals is suspected since shedding of viable B. burgdorferi has been evidenced in equine urine. This aspect, still partialy known, is interesting from a zoonotic point of view (Manion et al. 1998).