Brucellosis, caused by Brucella species, is considered one of the most common bacterial zoonotic infections worldwide. The infection is endemic in many geographic areas including the Mediterranean and Middle East regions, Indian subcontinent, Mexico and parts of central and South America [1, 2]. The disease is a major cause of direct economic loss and impedes trade among countries [3]. The estimated annual economic loss due to brucellosis was about 60 million Egyptian pounds in 1995 [4]. The diagnosis of Brucella infection is usually based on clinical presentation, microbiological culture and demonstration of specific antibodies [5]. Despite being the more commonly used technique for the initial screening of the infection, serological tests have limitations because not all infected animals produce detectable levels of antibodies, and false-positive results are seen due to cross-reactivity with antigens of other bacterial species [6]. The sensitivity of bacteriological culture usually depends on the viability and number of Brucella in the sample, as well as contamination of the sample with other bacteria [7]. Conversely, culture methods are time-consuming and the handling of microorganisms is hazardous [5]. In order to overcome most of these difficulties, PCR and real-time (rt)-PCR assays have been employed for diagnosis and molecular typing of Brucella species. Various PCR assays targeting different gene loci have been successfully used for diagnosis of brucellosis [8–12]. Species identification and sub-typing of Brucella isolates are very important not only for epidemiologic surveillance, but also for investigation of outbreaks in brucellosis endemic regions [13, 14]. Multiple-Locus Variable number tandem repeat Analysis (MLVA) is a useful tool for identifying and genotyping of Brucella isolates and data could be used for epidemiological trace back investigations [15, 16].

Bovine brucellosis remains one of the most common zoonoses in Egypt, and causes great constraints to the Government in improving animal productivity. Although nearly a century has passed since the first description of Brucella in Egypt, it has not been possible to eradicate this infection. The enforcement of control measures for brucellosis in Egypt (test and slaughter, S19 vaccination) have led to a reduction of B. abortus incidence in cattle [17]. However, most of these endeavors are still not conclusive [17, 18]. Despite being endemic, little is known about the epidemiological situation of brucellosis among Egyptian dairy herds. Indeed, Brucella infection remains under-diagnosed and often underestimated. In general, the control programs of Brucella infection in animals relies mainly on vaccination with live attenuated B. abortus strain 19. This provides good levels of protection against B. abortus in cattle and B.melitensis Rev. 1is used in sheep and goats [19, 20]. When administered correctly, the two vaccines can protect livestock from brucellosis, but the vaccines still have a drawback as they retain pathogenicity and sometimes cause abortion in vaccinated animals [21] and debilitating illness in humans [22–24]. In 2006, RB51, a mutant vaccine rough strain that is devoid of the lipopolysaccharide O-side chain was developed [25]. Cattle vaccinated with this kind of vaccine remained negative in conventional brucellosis serological tests [26]. This vaccine was suggested to be more appropriate than B. abortus S19 for the control and eradication programs that relied on serological testing and removal of positive animals [27].

For the effective monitoring of bovine brucellosis, it is imperative to have reliable tests to differentiate between vaccine and field strains. Thus, many molecular approaches have been developed to detect vaccine strains [28, 29]. Although extensive reports of animal brucellosis in Egypt are currently available, the epidemiologic situation of this infection is still ambiguous and needs further investigations. The purpose of this study was to analyze the epidemiologic pattern of bovine brucellosis in a well managed dairy farm in Egypt through a two year study period.

Four Brucella isolates were recovered from the collected four specimens of the aborted fetus. These isolates were typed as one vaccine strain (from liver specimens) and three isolates of B. abortus field strain (from spleen, abomasum and genital organs) by using Bruce-ladder PCR (Table 3) while B. abortus was confirmed by AMOS PCR. The three field strain isolates showed the same MLVA-16 genotype. B. abortus was also identified in the sample of the farm veterinarian on the basis of a combination of serological tests (RBT, ELISA and FPA) and rt-PCR assay. Cows that were positive in at least two serological tests or one serological test plus the rt-PCR (n = 27) were removed from the farm in July 2013 while RB51 vaccine was further administered to all non pregnant heifers and parturient cows (30-45 days post calving). In May 2014, two hundred and sixty three samples gave negative results by using rt-PCR, RBT, ELISA and FPA; while three samples were seropositive.

Among the bacterial zoonoses affecting dairy cattle in developing countries, bovine brucellosis is considered the most prevalent infection when the applied control programs have failed. The infection poses not only a risk to animals, but also represents a zoonosis with debilitating illness and severe complications in humans. In Egyptian control programs for brucellosis, RB51 B. abortus vaccine is believed to be cleared from the blood stream within three days but can sporadically induce late term abortions if administered to pregnant cows [36]. The agent can then persist in infected tissues of the cow and fetus/calf. The presence of Brucella DNA in blood samples of these rare cases was supposed to be due to intermittent bacteremia whereas the organism is located intracellularly and may result in a persistent infection for long time [37]. In the present study, an unexpected high number of rt-PCR positive blood samples were detected five months after routine vaccination. The ninety five seronegative cows with positive rt-PCR results may have had circulating rough RB51 vaccine strain DNA in their blood as some researchers were able to detect RB51 vaccine strain DNA for twenty five days in the milk of vaccinated buffaloes although cultivation was possible only for 4 days [38]. Longer circulation of vaccine DNA in the blood stream is therefore a realistic possibility for the origin of DNA found. The twenty five serologically positive cows that showed B. abortus DNA in their blood stream may have been infected by the field strain. However, the seropositive animals with negative rt-PCR results (n =10) might represent animals at different stages of infection whenever the DNA is already cleared from the blood stream. The same suggestion was previously reported by Gwida et al. [39]

In general, the use of rt-PCR for the diagnosis of brucellosis from blood samples of ruminants has been discussed controversially. It has been supposed that the sensitivity of PCR is dependent on the status of disease i.e. the chronic infected animals are regularly misdiagnosed as no B. abortus DNA is present in respective samples [12]. The presence of B. abortus DNA was, however, demonstrated in blood samples from various infected animal species and humans but with varying sensitivities [9, 40, 41]. Recently, B. abortus and B. melitensis DNA were detected in bovine milk which was collected from apparently healthy animals by using species-specific IS711 rt-PCR [42]. It has been shown that rt-PCR could be an additional tool in an outbreak situation when serological methods are still negative or it might help to identify animals which are not detected by conventional serological methods [40].

The coincidence of vaccination and the apparent infection by a field strain in this herd which was serologically negative two months prior to the administration of the vaccine highlights the fact that RB51 vaccination cannot protect a herd (or single animals) from being infected with B. abortus in an endemic area. Vaccination may only reduce economical loss for the farmer. The sense of security conveyed by vaccination may even result in loss of awareness and impaired biosecurity. It was previously speculated that the vaccine strain can transmit from vaccinated animals to pregnant animals and cause abortion in very few cases [36]. Our findings support such suspicions where vaccine and field strain are isolated from the same fetus, suggesting that transmission of RB51 vaccine strain from non pregnant animals to unvaccinated pregnant animals has taken place. Our screening system made sure that most of the female animals used for reproduction on the farm were investigated with all techniques during the investigation period of two years. Vaccination and serological investigation with subsequent removal of reactors may result in the absence of abortions from July 2013 to May 2015. Although, in the routine RBT and BAPAT screening at May 2014, three cows tested positive which were later on reactive in ELISA and FPA but negative in rt-PCR. Simply, these animals could be undetected due to the stage of infection or could be re-contaminated via new introduction of a field strain. As a special problem for eradication of brucellosis, are those female calves that were borne to Brucella positive mothers. These calves may harbor the agent during their childhood and first pregnancy and seroconvert after they have calved [43–45]. In the setting of Egypt, such animals can run undetectable for approximately 20 months. We strongly advice to remove all seropositive animals from the herd although the fact that few false positive reactors may be culled.

The farm veterinarian contracted the infection with B. abortus during his daily work on the farm, when he was not aware that the disease was introduced only recently. This finding highlights the public health hazard of Brucella infection and could be an alarm for the potential risks for personnel elsewhere. We were unable to detect the source of infection, or the way through which the agent was introduced into the farm. It is an accepted fact that even very well managed farms in endemic areas can easily become contaminated [40]. It is obvious that prevention of damage in endemic areas should involve the practices of good farm management for biosecurity and biosafety. Therefore, we highlighted some findings made during this outbreak. In general, vaccination using a vaccine which does not interfere with serology is advised. However, current vaccines can cause abortions even if all recommendations are followed. The animal owner has to be made aware of this drawback of vaccines, although it has to be stressed that the benefit of vaccination outweighs its risks. Routine screening with a sensitive and specific test (validation of test has to be done in the region where it will be used) in short intervals has to be applied. It is worthy to note that no further abortions were recorded at the farm, since applying the strategy of test and slaughter. By the end of April 2014, 120 cows were culled from the farm, being either seropositive or having concomitant ailments including mastitis, laminitis and metritis. Special concern should be directed to females of a Brucella infected mother or those females which have unknown background, as such animals can be a carrier of Brucella and may seroconvert after abortion or calving only.

We believe that RBT and BAPAT is a very useful combination of tests in Egypt, considering effectiveness and cost. If introduced, serological screening should be confirmed by ELISA and all reactors should be removed from the herd. All animals on the farm should be tested. We also suggest that PCR can be a good supplementary test for outbreak situations, as PCR can differentiate vaccine and field strains with high analytical sensitivity. We also recommend using RBT and ELISA in parallel in an outbreak situation to identify as many infected animals as possible during the initial screenings, and to repeat this test procedure for at least three successive negative tests with one month interval.