Comparative Immunology, Microbiology and Infectious Diseases 37 (2014) 1–9

Contents lists available at ScienceDirect

Comparative Immunology, Microbiology and Infectious Diseases journal homepage: www.elsevier.com/locate/cimid

Vaccination using phase I vaccine is effective to control Coxiella burnetii shedding in infected dairy cattle herds Anne-Frieda Taurel a,b,c,∗ , Raphaël Guatteo a,b , Anne Lehebel a,b , Alain Joly c , Franc¸ois Beaudeau a,b a

INRA, UMR1300 Biologie, Epidémiologie et Analyse de Risque en santé animale, BP 40706, F-44307 Nantes, France LUNAM Université, Oniris, Ecole nationale vétérinaire, agroalimentaire et de l’alimentation Nantes-Atlantique, UMR BioEpAR, F-44307 Nantes, France c Union Bretonne des Groupements de Défense Sanitaire, F-56000 Vannes, France b

a r t i c l e

i n f o

Article history: Received 5 September 2012 Received in revised form 30 September 2013 Accepted 3 October 2013 Keywords: Coxiella burnetii Shedding Dairy herds Vaccination Oxytetracycline Milk

a b s t r a c t The effectiveness of the vaccination of dairy cows combined or not with antibiotics (i.e. oxytetracycline) to control Coxiella burnetii (Cb) shedding at herd level was investigated in 77 Q fever clinically affected herds. In addition to nulliparous heifers’ vaccination, one out of the four following medical strategies was randomly assigned to dairy cows in each herd: vaccination (using a phase I vaccine) alone, vaccination combined with oxytetracycline, oxytetracycline alone or nothing. Their effectiveness to reduce Cb load in quarterly samples of bulk tank milk (BTM) and of pooled milk of primiparous (MP) was assessed through logistic hierarchical models. A significant reduction in Cb load was observed in herds where the vaccination of ≥80% of dairy cows was implemented; whereas the use of antibiotics was uneffective. Our findings support the interest of a whole vaccination strategy and provide evidence for decreasing the use of antibiotics in dairy cattle herds. © 2013 Elsevier Ltd. All rights reserved.

1. Introduction Coxiella burnetii is the agent responsible for Q fever, a worldwide distributed zoonosis [1], that can infect a wide variety of organisms such as mammals, birds, arthropods [2]. Humans and animals get infected mainly by inhaling contaminated aerosols produced by infected shedder ruminants [2]. In Humans, Q fever is mainly asymptomatic, but can induce in case of acute infection mainly flue like illness and in case of chronic infection endocarditis and abortion in pregnant women. Between 2007 and 2011 in the Netherlands, a Q fever outbreak involved more than 3500 human cases, 24 leading to death [3,4], confirming the

∗ Corresponding author at: UMR Oniris-INRA1300 BioEpAr, Atlanpole – La Chantrerie, BP 40706, F-44307 Nantes Cedex 03, France. Tel.: +33 240 68 78 21; fax: +33 240 68 77 68. E-mail address: [email protected] (A.-F. Taurel). 0147-9571/$ – see front matter © 2013 Elsevier Ltd. All rights reserved. http://dx.doi.org/10.1016/j.cimid.2013.10.002

putative impact of Q fever in terms of public health. Ruminants when infected are also mainly asymptomatic, but abortions, especially in small ruminants [5], metritis [6] and cases of infertility are reported [7]. Thus, any measures which aims at preventing or reducing C. burnetii shedding in infected ruminants will decrease the risk of transmission between ruminants and the risk of exposure and infection for the nearby population and will have an interest in terms of both animal and public health. When infected, ruminants shed the bacteria mainly through milk [8]. They can also shed through urine, feces, vaginal mucus and birth products and thus fulfill the environment with contaminated aerosols [5,8–11]. To control C. burnetii infection in ruminants, 2 types of measures are available: non-medical and medical measures. Non-medical measures rely on (i) hygienic measures focused mainly around calving and/or abortion and (ii) management of slurry. These latter aimed at limiting the spread of bacteria to reduce the associated exposure. But the effectiveness of such control actions is poorly

2

A.-F. Taurel et al. / Comparative Immunology, Microbiology and Infectious Diseases 37 (2014) 1–9

documented in the literature. Regarding medical measures, an experimental study conducted in goats [12] reported that the use of a phase I vaccine in non-infected then challenged animals was associated with a very strong reduction in the incidence of abortions and vaginal shedding of Cb and with an absence of milk shedding (compared to nonvaccinated or vaccinated goats with a phase II vaccine). Field studies have been also implemented in naturally infected flocks. Vaccination with the phase I vaccine was also associated with prevention and/or reduction of shedding in ruminants [13,14], especially when applied on primiparous and/or susceptible goats [15,16] or susceptible non-pregnant dairy cows [17]. Besides vaccination, antibiotherapy, mainly based on the use of tetracyclines, is also frequently implemented in routine practice either at drying off to prevent late abortion or at calving to prevent the shedding peak. Nevertheless, little evidence supports these regimens, given the contradictory results reported in the literature in a restricted number of studies with different designs, which make them hardly comparable [14,18–21]. However, the use of tetracyclines is associated with a prevention of vaginal shedding at calving in dairy cows, when injected once at drying off (2 months to 3 weeks before calving) [14,19]. Considering the risk of antibioresistance, already reported for tetracyclines in human medicine [22], providing evidence for a rationale use of antibiotics is crucial. Considering medical strategies, on the use of vaccination and antibiotics has recently been reported to prevent or reduce Cb shedding at calving time in the vaginal mucus of dairy cows [14]. Nevertheless, the effectiveness of such medical strategies to prevent and/or reduce Cb shedding at herd level is still unknown. To assess the effectiveness of medical measures at herd level, a first option could be to sample all the animals within herds to assess the dynamics over time of the within-herd prevalence of shedders. However, in cattle, the heterogeneity of Cb shedding pattern (intermittent, sporadic, persistent) and the non-concomitancy of shedding routes reported [8] led us to consider this option as not sustainable. Another option to assess the effectiveness of medical measures at herd level could be then to use some samples representative of the whole herd. Milk has been reported as the more frequent and persistent shedding route in ruminants [5,8]. Additionally real time PCR applied to bulk tank milk has been reported as reliable to assess the withinherd prevalence of milk shedders [23]. Thus, pooled milk samples such as BTM or pooled samples of subpopulation of cows could be an alternative to investigate a putative reduction of Cb load shed at herd level. Therefore, the aim of this study was to assess and compare the effectiveness at herd level of four different medical strategies, in dairy cows, combining or not vaccination and antibiotherapy to reduce the bacterial load measured in pooled milk samples from Q fever clinically affected dairy herds. 2. Material and methods This work has been conducted in compliance with the recommendation of the STROBE Initiative (Strengthening

the Reporting of Observational Studies in Epidemiology) on cohort study [24]. 2.1. Herds and animals From February 2008 to June 2010, 136 dairy herds infected with C. burnetii have been recruited in the West of France among those experiencing abortions and included in the abortion control program implemented by the local animal health services. To be included in the protocol, herds had to fulfill the following criteria: (i) at least 50% of seropositive animals (using ELISA) within a sample of at least 6 cows and an abortion attributable to C. burnetii (based on a positive result using real time PCR on placenta) in the previous month or (ii) a bulk tank milk tested positive using real time PCR. Given the absence of DIVA ELISA (Differentiating Infected and Vaccinated Animals), no vaccination against Q fever for at least 5 years should have been performed. Moreover, the herds should not have implemented systematic therapeutics based on tetracyclines to treat reproductive disorders in the last 6 months before inclusion. A consent form was signed by all the farmers. All the females of 12 months and more, for which a calving was expected during the 13 months follow up, have been included in the study. 2.2. Experimental design 2.2.1. Nature and allocation of treatments At inclusion time, one strategy among 4 has been randomly assigned to each herd. For ethical reasons and given that vaccination using phase I vaccine was already demonstrated as effective to prevent shedding on susceptible animals [12,17] such as nulliparous animals [25], all the nulliparous heifers (12 months and more) were vaccinated regardless the strategy attributed to the herd. Therefore, only cows were concerned by the randomly attributed herd strategy. The four different strategies applied on dairy cows were: (i) vaccination with a phase I vaccine, (ii) vaccination and antibiotherapy based on the use of oxytetracycline, (iii) antibiotherapy alone or (iv) none of these treatments. Vaccination consisted in 2 injections (4 mL each) of primovaccination 3 weeks apart and an annual booster (Coxevac, Ceva Santé Animale, Libourne, France) administered by the veterinary practitioners. In herds allocated to a strategy containing antibiotics (Tenaline LA, (Ceva Santé Animale, Libourne, France) 20 mg/kg oxyetracycline at each injection), 1 individual regimen among five was randomly attributed to each cow: antibiotic at (i) drying off, at (ii) calving, at (iii) drying off and calving, at (iv) twice at drying off 15 days apart, or (v) twice at drying off 15 days apart and at calving. In order to improve the compliance especially for antibiotic administration, the farmers received a reminder text message on their cell phone on the expected day of intervention (based on the estimated dates of drying off and calving provided by the milk recording scheme). 2.2.2. Samples and laboratory analysis At inclusion time, blood samples were collected by the practitioner from all the animals older than 12 months to estimate the within-herd seroprevalence.

A.-F. Taurel et al. / Comparative Immunology, Microbiology and Infectious Diseases 37 (2014) 1–9

Samples were immediately sent to the laboratory (Institut Départemental d’Analyses et de Conseil, Nantes, France) to be tested using the Q fever LSI ELISA kit (LSI, Lissieu, France), reported to be the most sensitive ELISA available thanks to the ovine strains from which it has been made [27], according to the manufacturer’s instructions. The results were expressed in an optical density Sample/Positive control (S/P) ratio. A serum sample was considered as positive when the S/P ratio in serum was >40, and seronegative otherwise according to the manufacturer’s instruction. We assumed that the putative existence of few persistent milk shedder cows could lead to persistent PCR positive bulk tank milk (BTM), although a reduction in the prevalence of shedders could occur. Then, in addition to BTM samples, we also performed pooled milk samples from primiparous cows (MP) only. As animals are considered only in their first lactation, we assumed that the chance to observe a reduction in the bacterial load in MP associated with the implementation of a medical strategy was higher. To measure a reduction of the bacterial load in those samples, BTM and MP samples were performed during 18 months at a quarterly (59% of herds) or monthly frequency (41% of herds). All samples were sent to IDHESA laboratory for detection and quantification, by real time PCR analysis (TAQVET® C. burnetii kit, LSI), of the load of Cb presents. Samples were considered as positive when