Sero‐epidemiology investigation of Coxiella burnetii in domestic ruminants throughout most Greek regions

Abstract Q fever is not considered as a public health problem in Greece where most regions are considered as Coxiella burnetii free possibly because of the low interest for this agent. Our objective was to conduct a large‐scale study to investigate the sero‐epidemiology of C. burnetii in domestic ruminants throughout the most of Greek regions. We tested serum samples obtained from goats, sheep and bovines from different regions of Greece. All sera were tested for C. burnetii IgG antibodies by a commercial ELISA according to the manufacturer's recommendations. We tested 1,173 goats and sheep obtained from 177 different herds and totally 194 (17%) animals from 78 (44%) herds were positive for C. burnetii. Positive animals were present in seven (88%) different regions and seropositivity varied widely among these regions. The highest percentage was observed in Peloponnese (44%), where all the tested herds presented animals with C. burnetii antibodies. Ιn all Aegean Islands except the island of Limnos we detected goats and sheep positive for C. burnetii with seroposivity varying between 2% in Kos to 37% in Rhodes. Finally, in 22 (85%) Greek prefectures we found C. burnetii IgG‐positive animals whereas in 14 (54%) prefectures more than 50% of tested herds had seropositive animals. We also tested 28 cows from five different herds in Macedonia and Aegean Islands and six (21%) of them, obtained from two (40%) herds were positive. Considering the importance of C. burnetii for public health, our data reflect the lack of awareness by veterinarians, physicians and competent authorities as we provide evidence of C. burnetii seropositivity in productive animals throughout the most of Greek territories. Due to the increased risk of inhalation of the bacterium by people who entered the affected farms we raise the question of Q fever emergence in Greece.


| INTRODUC TI ON
Q fever is a worldwide zoonosis caused by an obligate intracellular bacterium, Coxiella burnetii (Angelakis & Raoult, 2010). In human, clinical findings in Q fever infection are often confusing, and primary infection is asymptomatic in approximately 60% of cases. Cardiovascular complications are the main risk of C burnetii infection, including acute and chronic endocarditis and vascular infections (Million et al., 2016;Million, Walter, Thuny, Habib, & Raoult, 2013). Indeed, Q fever endocarditis is associated with surgery for 15% to 73% of patients, causes death in 5% to 65% of patients, and induces a large number of relapses when it is inadequately treated (Prudent, Lepidi, Angelakis, & Raoult, 2018).
C. burnetii could also cause obstetric complications including abortion or foetal malformations in pregnant women (Angelakis et al., 2012).
The main reservoirs of C. burnetii are cattle, sheep, and goats. In most cases, human contamination occurs from inhalation of aerosolized bacteria that are spread in the environment from animal birth products (Angelakis & Raoult, 2010) and findings suggest the role of wind in the transmission of C. burnetii between ruminants and humans (Pandit, Hoch, Ezanno, Beaudeau, & Vergu, 2016;Tissot-Dupont, Amadei, Nezri, & Raoult, 2004). Moreover, introduction of new animals into herds has been identified as a risk factor of C. burnetii infection and it is known that trade between cattle herds occurs frequently and sometimes over long distances (Nusinovici et al., 2014). In livestock, infections caused by C. burnetii are usually asymptomatic although the disease has been implicated in abortion, stillbirths, endometritis, mastitis and infertility (Arricau-Bouvery & Rodolakis, 2005). Recently, there has been an increased awareness of Q fever as an economically important pathogen due to a rise in the frequency of reported outbreaks and the economic impact of Q fever has on commercial livestock operations in the form of lost animal reproductive productivity and herd death (Enserink, 2010).
The importance of Q fever, in terms of public health, increased after the outbreak in the Netherlands, where more than 4,000 people became ill and 50,000 animals were slaughtered to control the epidemic (Van Der Hoek et al., 2012).
Although the classification of C. burnetii by the CDC as a potential bioterrorism agent resulted in the disease becoming reportable in many countries (Eldin et al., 2017), Q fever is not considered as a public health problem in Greece and few human cases have been recorded (Kokkini et al., 2013). However, we recently raised the question of the under-diagnosis of human C. burnetii infections in Greece (Karageorgou et al., 2020). Previously, it was found that the C. burnetii genotype MST32 is circulating in sheep and goat at eight different areas of Northern Greece (Chochlakis et al., 2018). However, most Greek regions are still considered as Q fever free possibly because of the low interest for this agent and to the best of our knowledge only two sero-epidemiological studies have been previously conducted to estimate C. burnetii in domestic ruminants in Greece (Filioussis et al., 2017;Pape et al., 2009). In association with the Greek Ministry of Rural Development and Food, we conducted a large-scale pilot study throughout the most of Greek regions to determine for the first time if Q fever is a concern in domestic ruminants in Greece.

| Sample collection
From January 2015 to December 2019 we tested serum samples, following communication with local veterinarians, obtained from goats, sheep and bovines from different regions of Greece. The participation to our study was voluntary and we encouraged veterinarians to sample animals with a clinical diagnosis of any of the following adverse pregnancy outcomes: abortion, premature delivery, stillbirth and weak offspring. In order to facilitate the participation and increase the number of tested farms, we did not force veterinarians to collect other than the serum samples information. Serum samples were collected from each animal suspected for Q fever and within 6 hr of collection were sent to our laboratory stored at −20°C for further analysis.

| Ethics statement
Animal sampling were conducted according to the guidelines of the Greek National Veterinary Agency.

| Serology assays
The serum antibody IgG titres against C. burnetii were assayed using a Q fever indirect ELISA kit (IDEXX Q-Fever (C. burnetii) Antibody Test, IDEXX Laboratories, Inc), with a mixture of C. burnetii phases I and II antigen according to the manufacturer's instructions. The manufacturer had previously validated the ELISA kit estimating sensitivity of 100% and specificity of 95%, respectively. Briefly, the serum samples were incubated at room temperature for 30 min and then diluted in a 1:20 sample diluent. Negative and positive controls were always included while examining serum samples. ELISA results were obtained by comparing the optical density (OD) of the sample well with the OD from the positive control. For each sample we calculated the ELISA index S/P % according to the manufacturer's instructions using a photometer at a wavelength of 450 nm as follows: S/P % = 100 × (OD value of the sample tested − OD value of the negative control)/ (OD of the positive control − OD of the negative control). Sera samples with S/P % <30% were considered negative, whereas samples with S/P % ≥40 were considered positive. A sample with 30% ≤SP % <40% was considered suspect and reanalysed again.

| Statistical analysis
Student t or χ 2 tests were performed using Epi Info version 6.0 software (Centers for Disease Control and Prevention, Atlanta, GA, USA). Means were compared using analysis of variance or the Kruskal-Wallis test, on the basis of results of the Bartlett test for inequality of population variances. Proportions were compared using the Mantel-Haenszel χ 2 or Fisher exact tests when the expected value of a cell was <0.05. A p value < .05 was considered significant.

| RE SULTS
We tested 1,173 goats and sheep obtained from 177 different herds located in eight out of the nine geographic regions of Greece ( Figure 1a). Overall, 194 (17%) animals from 78 (44%) herds were seropositive for C. burnetii (Table 1) We also tested 28 cows from five different herds in Macedonia and Aegean Islands and six (21%) of them, obtained from two (40%) herds were positive (Table 2). We did not find significant difference in C. burnetii seropositivity between goats-sheep and cows (p = .4).
Similarly, we did not find difference between herds with goat and sheep versus bovine herds positives for C. burnetii (p = .9).

| D ISCUSS I ON
We provide evidence of C. burnetii presence in productive animals throughout the most of Greek territories. Only two small-scale studies have been previously performed in Greece to estimate the presence of C. burnetii antibodies in sheep and goats (Filioussis et al., 2017;Pape et al., 2009) and none in cattle. Although validated commercial ELISA kits are routinely used for the detection of C. burnetii in productive animals with high sensitivity and specificity, a limitation of our study is that we did not confirm our results by a specific immunofluorescence assay (IFA). Moreover, we did not perform molecular assays in positive samples to detect the C. burnetii genotypes that circulate in productive animals in Greece (Chochlakis et al., 2018). We believe that the determination of C. burnetii genotypes would be important as the clinical manifestations of Q fever in human depend, at least in part, on the C. burnetii genotype (Angelakis et al., 2012). Indeed, although acute clinical presentation is strain-specific, all genotypes have been associated with endocarditis in human (Angelakis et al., 2012). Another limitation is that we could not discriminate the goat and sheep samples as their origin was from mixed goat and sheep herds and no species-specific information was provided to identify the samples. Indeed, most of herds in Greece, continue to maintain traditional pastoral practices with mixed goat and sheep livestock (Figure 1b). Moreover, information about animal level and herd level demographic characteristics, livestock rearing system were not collected due to the reluctance of veterinary volunteers as was significantly increased their work.
We found that seropositivity of C. burnetti in the different prefectures of Greece is very variable from one area to another. Similar regional differences in the prevalence of the infection were observed in Bulgaria, Germany (Georgiev et al., 2013) and Iran F I G U R E 1 (a) Greek prefectures tested for Coxiella burnetii seropositivity. (b) Goat and sheep pastoral herds in Greece. Blue stars: prefectures without C. burnetii seropositive animals, red stars: prefectures with C. burnetii sero-positive animals (Nokhodian, Feizi, Ataei, Hoseini, & Mostafavi, 2017). Similarly, a study in Sweden (Ohlson et al., 2014) and a systematic review in Africa revealed that seroprevalence of C. burnetii infection is different, depending on the geographic location (Vanderburg et al., 2014). We believe that this heterogeneity could be due to differences in sample sizes among prefectures as well as due to agricultural and climatic differences among the tested areas.
Moreover, as previously (Georgiev et al., 2013) we found that seropositivity measured at the individual animal level was lower than herd's seropositivity. In each herd only a relatively low number of animals seroconverted following contact with C. burnetii. This result is somewhat surprising, given the known high rate of infectivity of C. burnetii in ruminant populations.
In Greece, as elsewhere in the world (Angelakis & Raoult, 2010;Eldin et al., 2017), we found evidence of widespread exposure to C. burnetii in domestic ruminant populations. Recently, we raised the question of the under-diagnosis of C. burnetii infections in Greece possibly due to the lack of awareness by physicians. In that study, we found evidence of C. burnetii infection in humans throughout most Greek territories as also a new endemic C. burnetii genotype (Karageorgou et al., 2020). Domestic ruminants are considered the primary reservoir for C. burnetii and most Q fever outbreaks  (Nusinovici, Frossling, Widgren, Beaudeau, & Lindberg, 2015). A higher animal density could also increase the risk of propagation by increasing the potential number of neighbouring sources of contamination (Nusinovici et al., 2014). Moreover, a high quantity of precipitation it seems that can decrease the quantity of dust in the air and thus the likelihood of transmission (Hogerwerf et al., 2012). In this context and prompted by the outbreak of Q both infected and C. burnetii free herds may be a relevant control measure to limit zoonotic risk (Courcoul et al., 2011).
In conclusion, we provide evidence of C. burnetii in productive ruminants throughout many Greek territories. We raise the question of Q fever emergence in Greece, due to the risk of inhalation of the bacterium by people who entered the affected farms. In all reported human outbreaks, animal contact with infected animals has been a consistent feature and the most likely source of C. burnetii infection (Eldin et al., 2017). However, to this date there is insufficient information to enable early prediction of large Q fever outbreaks in Greece. Considering the importance of C. burnetii for public health, our data reflect the lack of awareness by veterinarians, physicians and competent authorities alike. We believe that much remains unclear for C. burnetii and there is a need to improve the means for early detection of risk of outbreaks in Greece, the effectiveness of veterinary control measures, and determine the best follow-up strategy in territories with repeated outbreaks over several years.

CO N FLI C T O F I NTE R E S T
The authors declare no conflict of interest. Validation; Writing-original draft.

PE E R R E V I E W
The peer review history for this article is available at https://publo ns.com/publo n/10.1002/vms3.337.