Rhodococcus equi is a common cause of pneumonia in foals and has extensive clinical and economic consequences. Clinical disease typically affects foals between 3 weeks and 6 months of age on endemic farms (Reuss & Cohen 2015). The main clinical manifestation is chronic suppurative bronchopneumonia with lung abscesses, although many cases are accompanied by various extra-pulmonary disorders, including ulcerative enterocolitis, joint effusion and ocular disorders (Giguère et al. 2011a). Morbidity and mortality in foals are high, whereas healthy adult horses are rarely diagnosed with clinical disease (Cohen 2014; Giguère et al. 2011a, 2011b; Morresey & Waldridge 2010; Muscatello 2012a; Vazquez-Boland et al. 2013). In recent years, the awareness of the zoonotic potential of R. equi has increased, mainly in immunocompromised human patients, thus emphasising the importance of better surveillance of this pathogen (Ocampo-Sosa et al. 2007; Vazquez-Boland et al. 2013; Weinstock & Brown 2002).

Rhodococcus equi is a Gram-positive facultative intracellular bacterium that is saprophytic in soil and multiplies in herbivore faeces. The main reservoir is in the soil, but it can also replicate in equine intestines (Giguère et al. 2011b; Muscatello 2012a). Transmission is faecal–oral, yet spread via aerosol plays an important role in clinical infection. Only certain strains of R. equi are pathogenic, and several virulence genes and plasmids have been identified. In the environment, most R. equi isolates lack virulence traits, but under selection in the host environment acquisition of virulent plasmids may occur, and more virulent isolates are found. The epidemiology of R. equi exhibits a delicate balance between the environment, subclinical carriers and susceptible hosts. The main reservoir is in soil, which makes it endemic in many equine facilities, but it also appears sporadically in non-endemic locations. The risk of infection increases with animal density and dusty hot dry climates (Muscatello 2012a; Vazquez-Boland et al. 2013).

Diagnosis and prevention are challenging, and because early detection of clinical cases improves survival, much effort has been put into developing effective surveillance programmes. The gold standard is isolation of bacteria from trans-tracheal lavage of a suspected case and the detection of the virulence-associated plasmid A (VapA) gene with the aid of polymerase chain reaction (PCR) (Muscatello 2012b; Reuss & Cohen 2015). But, because R. equi is mainly an environmental organism and possibly a commensal in healthy horses, isolation of bacteria does not always correlate with clinical disease and isolation of the bacterium must be accompanied with septic lung cytology or ultrasonographic imaging of lung abscessation (Cohen 2014; Giguère et al. 2011b; Reuss & Cohen 2015). Attempts at screening and early detection have been done with routine auscultations and ultrasonography, which, although widely used worldwide, are labour intensive and nonspecific (Cohen 2014; Giguère et al. 2011b; Muscatello 2012b). Serological screening methods, mainly enzyme-linked immunosorbent assay (ELISA), using whole bacterium, crude antigens and lately VapA specific antigens, were developed for early detection of exposure to R. equi (Giguère et al. 2011b; Reuss & Cohen 2015). In general, most serological tests for antibodies against R. equi have low specificity and cannot discriminate between subclinical infection and maternally derived antibodies (Giguère & Prescott 1997; Giguère et al. 2011b; Higuchi et al. 1997). Several studies in recent years attempted to evaluate and improve the sensitivity and specificity of ELISA screening, and to determine the cut-off values for early detection of clinical cases (Giguère et al. 2003; Martens et al. 2002; Sanz et al. 2014; Witkowski et al. 2012). Despite these limitations, serological surveys may be of value in identifying endemic farms and assessing the potential for clinical infection in susceptible animals (Attili et al. 2006; Sanada, Noda & Nagahata 1992; Tel, Arserim & Keskin 2011). A serological survey of subclinically infected horses in Japan demonstrated that R. equi seroprevalence in farms may differ between clinically endemic, sporadic and ‘clean’ farms (Sanada et al. 1992).

In Israel, R. equi was first reported in two 2-month-old foals in 1998 (Steinman, Sutton & Elad 1999). Since then, the bacterium had been isolated from trans-tracheal lavage fluids in a few foals suffering from bronchopneumonia (one in 2011, none in 2012, one in 2013 and none in 2014 and 2015 http://www.vetserv.moag.gov.il/Vet/all_Publications/dochot-shnatiim/default.htm, accessed April 2017). Although climatic conditions and horse management in Israel are suitable for this bacterium, little is known about its prevalence. This study was conducted to assess the prevalence of R. equi in adult horses in Israel by using crude protein antigen–based ELISA, in order to indirectly indicate the level of exposure of foals, as access to foals for sampling was very limited.

Blood collections were performed with owners’ consent, and the study was approved by the Internal Ethics Review Committee of the Koret School of Veterinary Medicine, The Hebrew University of Jerusalem.

Rhodococcus equi is an important pathogen that can cause severe disease in foals. Furthermore, its zoonotic potential makes it even more important. Early recognition is crucial for successful intervention and treatment. Extended antibiotic therapy is indicated in clinical cases, and increased recognition of the development of macrolide-resistant R. equi makes it even harder to treat (Cohen 2014). Better understanding of R. equi ecology and epidemiology will increase clinician awareness and is therefore warranted. Definitive diagnosis of the disease is based on the culture of the organism from tracheal wash fluid and considered as the ‘gold standard’. However, it has some disadvantages, including difficulty in sampling and low sensitivity because of possible prior antibiotic administration and the presence of other pathogenic bacteria (Giguère & Prescott 1997). Therefore, serological tests have also been used extensively for diagnosis of the disease and/or to detect specific antibodies to the agent (Giguère et al. 2003; Takai, Kawazu & Tsubaki 1985). Positive serology may indicate exposure to R. equi, but not necessarily current infection and in most cases cannot discriminate between infections with virulent or avirulent strains or between clinical or subclinical infection (Cohen 2014; Giguère & Prescott 1997; Giguère et al. 2011b). In this study, we aimed to determine the seroprevalence of anti-R. equi antibodies in subclinically infected adult horses in Israel to indirectly indicate the level of exposure of foals, which we had very limited access to because of the low numbers and lack of cooperation from the owners.

The prevalence of R. equi exposure in healthy adult horses as seen in this study appears to be low. The observed seroprevalence of 7.6% in 2011 and 5.1% in 2014 did not differ statistically (P = 0.29). This is somewhat surprising because we speculated that the local climate, which is warm and dry most of the year, and the management, which is based on many farms on sandy paddocks, would result in higher seroprevalence. On the other hand, the limited number of clinical cases does support the results. Other serological surveys of healthy horses demonstrated slightly higher prevalences (11.7% and 14.8% of horses in different parts of Turkey, 11% in Japan and 13.45% of foals in Italy), although only the results from north-west Turkey (14.8%) (Attili et al. 2006) were statistically different from our findings (P = 0.023) (Attili et al. 2006; Cuteri et al. 2003; Sanada et al. 1992; Tel et al. 2011).

Although the population at risk for R. equi clinical infection is mainly young foals, this survey consisted almost entirely of adult horses, with very few horses under 1 year of age. The equine breeding industry in Israel is small, large-scale breeding facilities are scarce and owners are reluctant to allow sampling of foals that are not regularly handled. Most of the previous studies included some (Attili et al. 2006; Tel et al. 2011) or only (Cuteri et al. 2003) foals, although seroprevalence did not appear to be influenced by age (Attili et al. 2006; Tel et al. 2011). Serological tests used for the detection of R. equi-specific antibodies were designed to evaluate recent exposure. Because the level of antibodies in individual horses was not found to reflect the epidemiological status of the farm (Witkowski et al. 2012), ELISA results in this study and similar surveys were treated as a dichotomous parameter, to indicate exposure (Attili et al. 2006; Sanada et al. 1992; Tel et al. 2011). Little is known about the kinetics of antibody concentrations over time, but the fact that seroprevalence did not correlate with age (Attili et al. 2006; Tel et al. 2011) implies that adult horses may serve as sentinels for exposure in foals. Similar seroprevalences at both time points support the assumption that this bacterium is endemic in Israel, with probable endemic stability. This is also supported by the small number of horses that seroconverted after 3 years. The fact that some farms that were negative in 2011 had positive horses in 2014 may be ascribed to the fact that not all horses were sampled on both occasions (because of owner consent) or of new horses that were introduced to these farms during that time period.

The factors that influence R. equi incidence and acquisition of virulence are not fully understood. Exposure of foals to the bacterium early in life in combination with their immunological status seems to be crucial (Cohen 2014; Reuss & Cohen 2015). The amount of bacteria in soil and faeces does not appear to be associated with clinically infected foals, whereas factors that influence the amount of airborne particles (animal density and housing in stalls) do seem to correlate with clinical disease (Cohen 2014; Muscatello 2012a; Reuss & Cohen 2015). Our findings did not show an association between stable management and seroprevalence, and the only positive farm included at both time points keeps the horses on pasture. In addition, the presence of VapA protein is necessary, but not sufficient to cause disease in foals (Reuss & Cohen 2015). The detection of antibodies against a virulent strain may reflect the potential for acquisition of such a strain by susceptible foals, but does not necessarily correlate with its virulence de facto and the chance to develop clinical disease.

The use of a serological survey of adult horses to represent the risk of foals to acquire R. equi has several limitations. As mentioned before, seropositive results do not indicate actual infection, but only exposure to the bacterium. Also, because crude antigen was used in the ELISA, although prepared from a virulent strain, positive results may not necessarily indicate exposure to a virulent strain. Adult horses are usually not susceptible to clinical infection, and exposure in this population does not directly correlate with disease. The use of foal serum prior to nursing as a negative control also has some limitations because it lacks antibodies, while sera of negative adult horses have a nonspecific antibodies that may influence the reading. Nevertheless, because no documentation of R. equi exposure in Israel exists up to date, the results of this survey may provide some knowledge about the epidemiology of this bacterium.

Identification of R. equi-positive farms is important to raise the awareness of attending clinicians to early diagnosis and treatment of clinically affected animals. Because no effective prevention or control methods exist, the focus on farms where clinical cases occurred should be on early diagnosis using routine screening of foals by haematology, imaging or both. Although these methods also lack in specificity and may lead to unnecessary treatment in some cases, there is no preferable alternative to date (Cohen 2014).

The prevalence of R. equi exposure observed in healthy adult horses did not differ statistically between 2011 and 2014 and was found to be 7.6% and 5.1%, respectively. Although low, the detected seroprevalences indicate that the causative organism continues to be present on Israeli farms and poses a potential risk to susceptible foals. Clinicians should be aware of the presence of this pathogen on farms for early detection and treatment of clinically infected foals and immunocompromised people.