Introduction

Q fever is a zoonosis caused by the strict intracellular bacterium Coxiella burnetii, which is maintained in the environment by animals, especially mammals [1,2,3]. Humans generally become infected through air-borne transmission of the bacterium from the envirnoment contaminated by animal reservoirs, as a spore-like form allows C. burnetii to resist different environmental conditions and disinfectants [4]. This zoonosis is found worldwide, but there is considerable uncertainty about the incidence in different regions of the world and the relative importance of the risk factors of human infection. Potential reservoirs of the pathogen are numerous in wild and domestic species [5,6,7,8]. Nevertheless, descriptive studies mainly targeted -cattle, sheep, and goats - which represent the main source of human infection in Europe, while cattle appear to be involved as major reservoirs in Canada and Japan [1, 9, 10]. Other domestic mammals, such as cats and dogs, may also play a role in the transmission to humans, for instance in Canada and Australia [11,12,13,14,15,16,17]. Although less frequent, birds, marine mammals, or game may be sometimes implicated in the occurrence of sporadic or clustered human cases [17,18,19,20].

Considering that C. burnetii infection in humans is largely asymptomatic, Q fever appears to have a limited impact on public health. Nevertheless, it could cause significant disability and possibly long-term consequences due to persistent infection. Approximately 5% of those infected with C. burnetii develop a chronic form, now called persistent focalized C. burnetii infection, that can develop in patients with heart valve defects or vascular disease, immunosuppressed patients, and pregnant women. Persistent focalized Q fever can occur from a few weeks to several years after the initial infection, with patients presenting with endocarditis, pneumonia, chronic hepatitis, and other manifestations frequently less responsive to antibiotic treatment. Different frequencies of pneumonia and hepatitis have been observed depending on the regions of the world; one hypothesis is an association with local circulating strains [20,21,22]. Moreover, since the first descriptions in the1990s, chronic fatigue syndrome (CFS) has been described as a sequel directly associated with C. burnetii infection (post-acute Q fever) [23,24,25]. As the clinical manifestations are nonspecific and the diagnosis complex, CFS is probably an underestimated disease that can become a major public health problem. Great difficulties are actually encountered in control, especially during and after large human outbreaks, such as in the Netherlands [26, 27], and in hyperendemic areas, such as in French Guiana [28••]. In addition, C. burnetii is classified as a potential inhalational bioterrorism agent [29]. These situations highlight the need for a collaborative effort from human physicians and veterinarians under the umbrella of the “one health” concept to minimize exposure among people [30,31,32].

In animals, infection is also usually asymptomatic [3]. The spectrum of clinical forms appears less diversified than in humans but is biased by a focus on reproductive disorders leading to substantial economic losses, known as coxiellosis. The association with late-term abortions, stillbirth, prematurity, and low birth weight is clearly demonstrated in sheep, goat, and cattle farms [33,34,35,36,37,38]. Other manifestations are suspected (infertility, metritis, mastitis) but evidence is still lacking [35]. For other animal species, knowledge on the clinical outcome of C. burnetii infection is very limited, but currently appears to include reproductive failure in waterbuck (Kobus ellipsiprymnus), roan antelope (Hippotragus niger), dama gazelle (Nanger dama), and water buffalo (Bubalus bubalis) and placentitis in the Pacific harbor seal (Phoca vitulina richardsi), steller sea lion (Eumetopias jubatus) and red deer (Cervus elaphus) [39]. Overall, knowledge is not fully understood about the pathogenesis and the sites (organs, cells) affected depending on the species. Authors agree on the circulation of C. burnetii strains, which appear widespread in cattle, sheep, and goats in most countries. On a technical level, the investigations were first restricted to serology. However, the available serological tests should be optimized and validated for each species [3, 40, 41]. Since the 1990s, more and more research has included direct detection by molecular analysis (PCR). The lack of standardized methods, or at least the lack of knowledge of the performance of sensitivity and diagnostic specificity, and the limited sample size of most of the existing studies could suggest variable bias in reported prevalence rates of C. burnetii infection. In addition to the identification of C. burnetii infection in different species of animals in different regions, although some results are still questionable, it was possible to acquire descriptive knowledge such as (i) the sites and routes of excretion of ruminants such as feces, urine, vaginal secretions, semen, milk, and placenta, along with serological responses; (ii) potential environmental sources (dust, manure, etc.); and (iii) molecular epidemiological data [37, 38, 42,43,44,45]. Studies have shown a higher risk of transmission following an episode of abortion, especially in small ruminants, generating massive bacterial shedding in the environment [46]. Nevertheless, the epidemiology of C. burnetii infection is still complex, with particularities according to the regions of the world. A wide range of animal species can play a role in the dissemination or maintenance of the bacterium in the environment. The effective airborne propagation ranges of C. burnetii may vary depending on meteorological and topographical factors. It is important to better understand the diversity of epidemiological situations and evolutions around the world.

The presence of C. burnetii in several animal species linked to its zoonotic importance justifies the growing recent trend in scientific development related to coxiellosis [47]. The reservoirs of C. burnetii have been largely investigated in many countries, especially in Europe [48]. Nevertheless, the implication of wildlife in the transmission of C. burnetii to humans and livestock remains poorly understood [35, 48,49,50,51,52,53,54]. On the other hand, data on C. burnetii in Latin America are scarce, sparse, and extremely heterogeneous, although several studies have been published in the last decade in humans, especially in French Guiana and Brazil. On the one hand, French Guiana contains areas of hyperendemicity and holds the record for the highest annual incidence rate in the world [28••, 55]. Furthermore, there is a specific virulence associated with a pulmonary tropism, likely linked to a specific strain of the MST17 [22, 56,57,58,59]. On the other hand, countries such as Brazil have reported few cases, case series, and seroprevalence studies [60,61,62,63,64,65,66]. Other countries have performed only exploratory investigations, and most countries have reported none or a few human cases over the last 30 years [28••, 67]. The objective of this work is therefore to carry out a comprehensive review of the literature concerning animal C. burnetii infection in Latin America and the Caribbean.

Methodology and Objectives

Settings

We defined Latin America as Central America, the Caribbean, and South America.

Review of the Literature

A comprehensive literature review was carried out following the PRISMA Reporting Guidelines. The terms “Q fever” and “Coxiella” were searched together with the names of each of the 47 countries and territories of South America and the Caribbean in English, French, Spanish, and Portuguese in PubMed (National Institutes of Health’s National Library of Medicine) and ScienceDirect (Elsevier), and the terms “Coxiella” and “Q fever” in four languages in the databases SciELO (Scientific Electronic Library Online) and LILACS (Scientific Health Information from Latin America and the Caribbean countries), which are more specifically dedicated to the Latin American medical and scientific literature (Fig. 1). We excluded papers on other subjects than C. burnetii infection, C. burnetii infection in humans only, and studies on C. burnetii performed outside of Latin America and the Caribbean. A second time, after selecting the papers according to those inclusion criteria, the references of each paper were analyzed independently by two authors, and the most relevant ones were also included. Lastly, unpublished data concerning C. burnetii according to the knowledge of the authors were also added, including, thesis, web papers, and unpublished investigations, especially in Brazil and French Guiana. The bibliography was screened for every paper, and articles without complete text but with informative abstracts or summarized in a literature review were maintained in order not to avoid some not-indexed publications, especially old ones. The research period ranged from 1 January 1950 to 31 December 2022.

Fig. 1
figure 1

Flow chart of the study according to the Prisma criterion

Full size image

The publications were then grouped and discussed into four different groups: (1) livestock (cattle, sheep, goats, equine, swine), (2) pets (mainly dogs and cats), (3) wildlife (any taxon), and (4) ticks and other ectoparasites. For each of these four parts, the results are presented in summary tables. In the second part, a special focus was made on five countries where the infection is more often reported: French Guiana, Brazil, Argentina, Chile and Colombia, and where research on animal fever has been the most abundant.

Results/Discussion

After applying the criteria, 113 studies in French, Spanish or Portuguese or English were selected: (i) 84 published articles found in the databases; (ii) 16 articles not found in the databases but found through the bibliography review of the articles in the previous item, and (ii) 13 articles acquired from other sources, including four theses developed in Brazil and French Guiana, including four PhD theses (Fig. 1). Among the 21 countries or territories in continental Latin America, 16 (76%) had at least one publication related to C. burnetii detection in animals, and nine (35%) among the 26 Caribbean countries and territories. The most productive country was Brazil (N = 51), followed by French Guiana (N = 21), Colombia (N = 16), Argentina, Chile, and Uruguay (N = 8), respectively. Five countries had between two and four studies, and fourteen countries or territories had only one study. About 30% of the studies have been published between 2020 and 2022. The more abundant literature was available for livestock, followed by ticks, wildlife, and at last pets.

Animal Coxiella burnetii Infection in Latin America

Livestock

During the study period, there were 64 publications concerning C. burnetii from livestock in 20 countries and territories of Latin America and the Caribbean, mainly from Brazil (N = 18), Colombia (N = 10), and French Guiana (N = 6) (Fig. 2). Most of them concerned cattle, sheep, and goats, but also included zebu cattle, buffalo cattle, alpaca, horses, swine, and chicken (Table 1). Most of the studies were seroprevalence surveys in livestock, especially cattle. In cattle, seroprevalence rates varied a lot according to regions and studies, from 0% in the Brazilian state of Mato Grosso do Sul and in several Islands from the lesser Caribbean to 31.8% in Duque de Caxias, state of Rio de Janeiro [72, 83, 101]. Some authors also calculated the herd prevalence in addition to the apparent individual prevalence, with a variation from 46.9 to 82% [102, 105, 122]. Seroprevalence in goats varied a lot, from 0%, in the Caribbean Islands, in French Guiana or Piauí state, Brazil, to 35% in Nuevo Leon, Mexico, 50% in Itaboraí, Brazilian state of Rio de Janeiro, and 60.6% in Venezuela [76, 78, 101, 107, 109, 111, 121]. In most of the studies, the seroprevalence was relatively low, between 0 and 3% [80, 98, 100, 101]. At last, seroprevalence in sheep was highly variable also, from 0% in Argentina and the Caribbean Islands Trinidad and French Guiana, to 17.3% in Uruguay, 26.3% in St. Kitts, and 66% in Itaboraí, Brazilian state of Rio de Janeiro [76, 113,114,115,116].

Fig. 2
figure 2

Map of the publications on Coxiella burnetii infection in livestock in Latin America and the Caribbean

Full size image
Table 1 Data on Coxiella burnetii detection in livestock, swine and equine in Latin America
Full size table

Some had investigated the presence of the C. burnetii genome on vaginal swabs [86, 89, 98, 109], fetuses or placenta [86, 88••, 89, 117], serum or blood samples [66, 76, 99], and finally, milk and dairy products [71, 76, 81, 84, 85, 96, 98, 123]. Vaginal carriage of C. burnetii has been found in a few studies and implicated as a cause of abortion for different species [63, 88••, 118,119,120].

C. burnetii infection has been reported elsewhere in other domestic mammals, though less frequently than in livestock [124]. In Latin America, studies among domestic animals, including cattle (including zebu cattle and buffalo cattle), goats, and sheep, were scarce. Thus, we identified one study with 25% seropositivity for C. burnetii in horses in Colombia in 1977 and 3.4% in French Guiana in 2007 [108, 125]. Pigs have been found to be positive in two studies: 1.9% in French Guiana in 2007, 11.3% in Trinidad in 1996, and 0 to 21.2% in Uruguay in 1984 and 1985 [115, 116], although their role in the epidemiology of C. burnetii remains unclear.

Finally, results vary from region to region, and it is difficult to make broad generalizations about trends on the continent. One may hypothesize a correlation between the seroprevalence of C. burnetii infection and animal density. The justification for the higher prevalence of cattle would be in places where there is intensive animal farming [126, 127]. In any case, estimated seroprevalence data are very approximate, which makes it necessary to work on protocols and monitoring devices for C. burnetii infections.

Pets

A total of 18 studies from seven countries were performed on detecting C. burnetii in pets-dogs and cats, mainly in French Guiana and Brazil (Table 2). These studies included five unpublished works, one thesis in French Guiana and Brazil, respectively, and three unpublished works originating from the investigation of the epidemiology and veterinary department of the French Army following outbreaks in military personnel (Fig. 3). Dogs were found positive in the indirect immunofluorescence assay (IFA) for C. burnetii in Argentina, Brazil, and French Guiana, with prevalence from 1.8 to 21.7% according to the region, the study, and the cutoff of the titer of IgG [107, 128, 129, 131, 134]. In Brazil and French Guiana, some dogs have been found to be positive, with higher prevalence rates around human Q fever cases [76, 77, 107, 130]. Although some of these dogs were inquired about during confirmed human Q fever outbreak investigations, they do not seem to be at the origin of these human clusters [76, 107, 130]. Cats have been studied in only three regions :Argentina, Brazil, and French Guiana [77, 107,108,109, 128,129,130,131, 134] with six cats positive, five in Argentina and one in Rio de Janeiro, Brazil [77, 128]. No evidence of C. burnetii infection was found in dogs using serology or PCR tests on vaginal swabs in Colombia, Martinique, Nicaragua, and Uruguay [97, 134, 135].

Table 2 Data on Coxiella burnetii in pets in Latin America
Full size table
Fig. 3
figure 3

Map of the publications on Coxiella burnetii infection in pets in Latin America and the Caribbean

Full size image

Aside from livestock, pet animals, especially dogs and cats, in close contact with humans are important potential reservoirs of C. burnetii during urban Q fever outbreaks [124]. Some cases of human coxiellosis have been reported from infected dogs and cats as a source of infection [14, 138]. At the same time, several pets were found to be positive in the investigation of human clusters without these animals being incriminated in the transmission to humans [76, 107, 130].

Wildlife

A total of 24 studies, to which we must add the unpublished results of two investigations, were published on wildlife from only five countries of Latin America: Chile, Colombia, Brazil, French Guiana, and Uruguay (Fig. 4). Some studies have been performed through serodiagnosis in various mammal species, and most of them were negative, as shown in Table 3. Nevertheless, some species have shown seropositivity, meaning contact with the bacterium, such as free-living cervids (Blastocerus dichotomous and Mazama gouazoubira) in Mato Grasso do Sul, São Paulo, Goiás, and Paraná, Brazil, rodents and marsupials in French Guiana (Proechimys sp., Philander opossum, Didelphis marsupialis [107, 151]) and Pampas deer (Ozotocero sbezoarticus) in the Department of Maldonado of Uruguay. Thus, recent studies conducted in Brazil showed serological evidence of C. burnetii infection in Wiedomys pyrrhorhinos rodents and Didelphis albiventris marsupial from Ceará and Pernambuco states, respectively [131]. These animals were collected from areas where dogs were also found seroreactive to C. burnetii, suggesting that wild and peridomestic cycles of C. burnetii can be connected by rodents and other wild mammals, a scenario identified in other regions where peridomestic rather than wild cycles have a high impact in coxiellosis cases [131, 157]. The use of molecular biology, mainly PCR targeting the repetitive element IS1111, helped identify new positive mammals (Table 4). The results reported in Ct (cycle treshold) were mostly superior to 35, which corresponds to traces in bacterial loads. The animal groups with molecular evidence of C. burnetii in Latin America were Chiroptera (7 species), Rodentia (6 species), Suina (2 species), and one species for Pilosa, Cingulata, and Perissodactyla, respectively (Table 4).

Fig. 4
figure 4

Map of the publications on Coxiella burnetii infection in wildlife in Latin America and the Caribbean

Full size image
Table 3 Studies on Coxiella burnetii in wildlife in Latin America
Full size table
Table 4 Summary of wildlife species with molecular evidence for Coxiella burnetii in Latin America
Full size table

Nonflying Mammals

In French Guiana, the first studies about C. burnetii infection in wildlife date back to the late 1990s, following the discovery of the first human cases. This enthusiasm for the search for an animal reservoir of C. burnetii in this small French territory of about 300,000 inhabitants (200,000 inhabitants at the time of the first investigations) and an area of 83,846 km² is due to the real public health burden represented by human Q fever and the occurrence of several clusters in the French military personnel, linked to a very proactive Nation Reference Center in Marseille, in mainland France. Furthermore, these investigations have been enhanced, as all the first studies on animal infection with C. burnetii seem to rule out cattle as a source of human infection, so the origin of the epidemics had to be found in wildlife [107, 108, 158]. Although a positive three-toed sloth (Bradypus tridactylus) was found positive [109, 154], other animals, such as capybara (Hydrochoerus hydrochaeris), were tested and found negative [109, 152], weakening the former hypothesis claiming that the sloth was the single reservoir [151, 159]. Indeed, other species have been found positive between 2005 and 2014 following numerous investigations [153], and several species were found to be positive through the analysis by IS1111 PCR of the muscular juice of animals collected from hunters. Thus, one nine-banded armadillo (Dasypus novemcinctus), three white-lipped peccaries (Tayassu pecari), three collared peccaries (Pecari tajacu), one south American tapir (Tapirus terrestris), and two other capybaras (Hydrochoerus hydrochaeris) have already been found to be positive, most of them with Ct superior to 35, but never published. One spiny rat (Proechimys cuvieri) was also found positive on liver samples among 29 other rodents [153]. A study is currently underway in French Guiana that has collected over 1000 samples from animals, especially tissues and stools, to better characterize the wildlife reservoir of C. burnetii [160].

In Brazil, the first study found on the search for C. burnetii in wildlife is a thesis, which is an investigation in domestic animals, wild animals, and arthropods in Itaboraí in the state of Rio de Janeiro around cases of suspected C. burnetii infection [77]. Afterward, in a study developed in eight municipalities in the state of Rio de Janeiro, in the Atlantic Forest, between 2007 and 2012, the prevalence of C. burnetii DNA in 131 different species of wild rodents was 4.6% [139]. The positive rodents belonged to the species Akodon cursor (3/32: 9.4%), Mus musculus (1/19: 5.3%), Oxymycterus dasytrichus (1/12: 8.3%), and Oligoryzomys nigripes (1/16: 6.3%) captured in forest edges and near human dwellings in the municipalities of Piraí and Valença [139]. Other studies have been conducted on other species of nonflying mammals, but all with negative results. Thus studies using PCR have also been performed in several taxa: free-living cervids, free-living wild boars, and different species of Xenarthra in Brazil [141, 145, 146], Darwin’s fox (Lycalopex fulvipes) in Chile [149], different species of wild canids, rodents, and marsupials in Brazil, French Guiana, and Uruguay [107, 109, 115, 116, 131, 142], and in reptiles and amphibians in French Guiana without the detection of C. burnetii DNA [107, 109, 143].

Chiroptera

Concerning bats, although the role of these animals in the enzootic cycles of C. burnetii has received little attention worldwide, it was possible to find studies developed in Brazil and Chile [147•]. The first study to characterize this proteobacteria in Latin America involved 119 bats from 21 species captured in preserved areas in Rio de Janeiro, Bahia, and Santa Catarina, Brazilian from 2014 to 2015. Coxiella burnetii was PCR-detected in specimens of two species from the genus Artibeus: the great fruit-eating bat (Artibeus lituratus) (3 C. burnetii-positive/14 bats) and the fringed fruit-eating bat (Artibeus fimbriatus) (1/7) from two different regions: Jacarepaguá (Rio de Janeiro state) (3/44; 7%) and Serra do Tabuleiro State Park (Santa Catarina state) (1/28; 4%) [140]. In the second study, the first report of C. burnetii in Chile, PCR positivity of 9.0% (5/55) was observed in Mexican free-tailed bats or Brazilian free-tailed bats (Tadarida brasiliensis) sampled in three regions of the country [147•]. Lastly, a third study was published in 2022 using IS1111-qPCR to detect C. burnetii in blood samples from 126 bats captured in 2014, 2015, and 2018 in the Macaregua cave, Colombia [161]. Molecular evidence of C. burnetii was found in 6.3% of the samples: 3/49 (6.1%) Seba’s short-tailed bats (Carollia perspicillata), a widespread frugivorous bat; 2/35 (5.7%) ghost-faced bats (Mormoops megalophylla); and 3/42 (7.1%) Trinidadian funnel-eared bats (Natalus tumidirostris), an insectivorous bat.

There is a strong popular rumor that bats are reservoirs and transmitters of C. burnetii to humans in French Guiana. However, no study supported this hypothesis, and it is difficult to know where this rumor originated. In one of the very first studies conducted in the late 1990s, it was shown that seeing bats near one’s home was independently associated with the occurrence of Q fever in patients with fever compared to patients supported for dengue fever [107]. Nevertheless, the same team that found the various nonflying mammals positive found six bats belonging to four species positive to C. burnetii, spread over the whole territory but with high Ct (> 35): three Seba’s short-tailed bat (Carollia perspicillata), one Parnell’s mustached bat (Pteronotus parnellii), a lesser spear-nosed bat (Phyllostomus elongatus), and a lesser bulldog bat (Noctilio albiventris) [153].

Thus, even though the potential role of these animals as a source of infection for humans and other animals, including cattle, is still unknown, these reports suggest the existence of a complex C. burnetii transmission cycle involving a large number of wild mammals [162]. Therefore, further studies must be conducted to better understand their role in C. burnetii cycles. The results of this research on species other than domestic ruminants are difficult to interpret because the choices and times of the samples are not homogeneous, and the PCR tests performed in this way on a wide variety of biological matrices (various organs and fluids) allow first-line screening but not a controlled quantification for analytical epidemiology. Methodological developments are required.

Ticks and Ectoparasites

There has been an increasing interest in tick-borne pathogens, including C. burnetii, these last few years (Fig. 5). Eleven countries have published at least one study about C. burnetii in ticks in Latin America and the Caribbean, with a huge number of studies coming from Brazil, especially in the last 3 years (N = 19), followed by French Guiana (N = 7) and Colombia (N = 5). Although the importance of ticks in the epidemiology of C. burnetii infection remains debatable [9], there is no doubt that ticks can be infected by C. burnetii and that they can, therefore, act as vectors [6, 163,164,165]. In natural conditions, several C. burnetii strains have been successfully isolated from wild ticks, including from a few South American tick species (Table 5) [6, 173]. The Fiocruz group published a study in which Amblyomma sculptum and Rhipicephalus sanguineus were PCR-positive for C. burnetii [173]. For instance, in the case study by Pacheco et al. (2013) in Argentina, C. burnetii infection in two tick species belonging to the Amblyomma genus, A. tigrinum and A. parvum, was confirmed using hemolymph tests, isolation in Vero cells, and multilocus DNA sequencing. The strain At12 was typed as ST 73 in this publication. In Cuba, Noda and colleagues found six Amblyomma mixtum from 2 tick pools positive for C. burnetii using IS1111 qPCR [184]. In Brazil and French Guiana, specimens of another Amblyomma tick species, A. geayi, collected on three-toed sloth infected by C. burnetii were also infected [109, 167]. Therefore, a sylvatic cycle based on C. burnetii tick-borne transmission seems to be sustainable. However, C. burnetii is probably far more frequently transmitted through the airborne route than through ticks [6]. Although most of the studies did not find any sign of infection by C. burnetii in South American tick species, recent observations, based on advances in molecular and cell biology, showed that ticks, including those found in South America, commonly harbor Coxiella-like endosymbionts (Coxiella-LE), closely related but genetically distinct to C. burnetii. These Coxiella-LE are almost exclusively confined to ticks and, according to current knowledge, pose a much lower infection risk to vertebrates than C. burnetii [6]. Extensive molecular surveys have consistently revealed that Coxiella-LE predominates in most tick species investigated thus far, with at least two-thirds of tick species being naturally infected [156, 168, 187, 191]. Most importantly, Coxiella-LE has been commonly misidentified as C. burnetii [6, 192••, 193, 194]. Several C. burnetii detection methods are in use, but many are not efficient enough to clearly distinguish between C. burnetii and Coxiella-LE [192••, 193,194,195]. Based on the identification of IS1111 in both C. burnetii and the Coxiella-LE, Mares-Guia et al. (2018) developed a nested PCR assay, considering that IS1111 amino-acid sequences in endosymbionts revealed to be genetically divergent, including degraded copies likely to be nonfunctional, providing a specific C. burnetii detection [173, 194].

Fig. 5
figure 5

Map of the publications on Coxiella burnetii infection in ticks in Latin America and the Caribbean

Full size image
Table 5 Studies on Coxiella burnetii in ticks and ectoparasites in Latin America
Full size table

Highlights on the Epidemiological Situation of Specifically Selected Countries

French Guiana

French Guiana is a French overseas territory located in the northeast corner of the South American continent. The territory of 83,534 km2 is more than 90% covered by the Amazon rainforest, and the population of approximately 300,000 inhabitants (www.insee.fr) is concentrated on the coast, particularly in Cayenne and its surroundings. The human Q fever situation in French Guiana is unique, with the highest incidence in the world, with a peak in the mid-2000s at more than 150 cases per 100,000 inhabitants per year and a stabilization since 2009 of between 25 and 35 cases per 100,000 inhabitants per year, which is much higher than in the rest of Latin America where the disease is barely reported, making French Guiana a hyper-endemic zone [28••, 55, 67]. For comparison, Q fever incidence in mainland France was 1.9/100,000 inhabitants in 2008–2011, and the incidence reached 69/100,000 inhabitants per year in the Netherlands during the largest outbreak ever in 2009, in the population living within 5 km of an infected dairy goat farm, and 6/100,000 inhabitants per year beyond that distance [21, 196]. The distribution of cases is heterogeneous and is mainly concentrated in the capital, Cayenne, and its surroundings, although a recent seroprevalence study suggests that the infection may be more widely distributed across the country than previously reported [197]. A unique strain was identified a few years ago, found exclusively in French Guiana, MST17, both in humans and animals, when the strain was sequenced [59, 109]. One hypothesis is that the incidence of Q fever in humans is “over-reported” in French Guiana compared to other countries because the more virulent MST17 strain causes a less asymptomatic disease (with atypical symptoms) and that Guianese is therefore “over-diagnosed” compared to other countries. While lung disease usually represents 30 to 50% of cases, it affects more than 90% of patients in this French territory [21, 57]. Finally, one of the most questioning specificities to date of coxiellosis in French Guiana is its animal reservoir, which seems to contrast sharply with the rest of the world. Concerning domestic ruminants, usually considered the main source of transmission of the pathogen to humans, two serological surveys were carried out in the 1990s and 2000, respectively, suggesting a low seroprevalence in ruminants [107, 108, 152]. Ad hoc investigations on farms have been systematically negative, including investigations around cases of Q fever in farmers. The occurrence of several outbreaks of Q fever on the outskirts of Cayenne had raised the possibility of the bacterium being transported by sea winds from certain nearby herds, but this hypothesis has to be corroborated. That is why research has shifted for more than 20 years away from livestock to focus on the search for the bacterium in wildlife surrounding human cases [109, 151, 154]. Nowadays, several individuals from various species of nonflying mammals and bats have been found positive (Table 4), although mostly with Ct > 35, which means that the C. burnetii load is detectable but below the limit of quantification. We do not exactly know the link between these species and transmission to humans, apart from the positive sloth incriminated in the epidemic among the military families of Tiger Hill [109, 151]. Moreover, with the exception of the military clusters reported earlier, the numerous cases of Q fever reported in French Guiana each year are generally isolated cases, without any clustering of cases, which does not facilitate the understanding of the origin of the transmission. One possibility would be the persistence of the bacterium in the environment as a result of different wildlife species emitting it, and that humans could become infected by inhalation of this dust as a result of favorable circumstances like leveling work near the house, gardening, particularly the use of a brush cutter, sweeping, etc. [57, 107, 154].

Recently, two studies have challenged the dogma that there is no link between livestock and Q fever cases in Guiana. On one hand, Saout et al. showed that cattle had a high seroprevalence of Q fever, especially when based not on individual prevalence but on herd prevalence [122]. On the other hand, seroprevalence study for C. burnetii of almost 3000 people sampled throughout the country showed that there was a link between high prevalence and proximity to livestock farms [197]. To explain these contradictory results, the old studies focused on a small number of animals and/or herds. In addition, studies conducted in the 1990s used the complement fixation test, currently considered to be weakly sensitive [198]. Moreover, several cases of Q fever were reported among workers at the Cayenne abattoir in 2007 [108]. New studies are therefore needed in French Guiana on livestock to more strictly support these new hypotheses.

In conclusion, the epidemiological cycle of Q fever in French Guiana is still unresolved. Although several species of wild mammals have been identified as carriers of C. burnetii, these results remain anecdotal and explain neither the frequency of the disease nor the geographical and ethnic distribution of cases in the territory. Moreover, the dogma that C. burnetii was not transmitted by cattle in French Guiana seems questionable. Additional investigations of the animal reservoir, livestock, and wildlife must be carried out to allow a better understanding of the unique phenomenon observed in this small piece of Amazonia.

Brazil

In Brazil, where the first reports of C. burnetii infection occurred in the 1950s through serological tests in human and cattle samples [199], Q fever is currently subject to compulsory notification in the context of the differential diagnosis of rickettsioses. However, official data are lacking, which can impact the understanding of the epidemiology of this zoonosis in the country. Furthermore, although some human cases of Q fever have been mandatorily reported, the high incidence of febrile illnesses such as malaria, leptospirosis, dengue, Zika, and chikungunya, can lead to misdiagnosis of Q fever [66].

Among Latin American countries, Brazil has the largest number of publications, with evidence of C. burnetii circulation in livestock, pets, wild animals, and ticks. Bovines are the species with the highest average, followed by sheep, goats, dogs, wild cervids, wild rodents, bats, and ticks. This scenario is probably linked to the economic importance of the sector since Brazil is the second-largest producer of beef cattle and has a large dairy herd. Three papers are related to the detection of C. burnetii in animals with reproductive disorders [79, 82, 88••]. The study conducted in the state of Alagoas, in the northeast of Brazil, described a goat herd with high frequency in serology (55.1%) and molecular detection by conventional PCR in two placental samples from seropositive animals (8.7%) [79]. Another survey on cattle aborted fetuses reported 9.2% (7/76) of positivity at qPCR, with 3 positive fetuses from the state of Rio Grande do Sul and 4 positive fetuses from the state of São Paulo, reinforcing the widespread distribution of the pathogen in the country [88••]. In both studies, however, the lack of histopathological assay does not confirm that C. burnetii was responsible for the abortions [200]. The vast majority of studies in Brazil are based on serological and molecular detection in convenience samples. Only recently, statistical design studies are being carried out to understand better C. burnetii epidemiology [64, 79,80,81, 140, 141]. In the state of São Paulo, a study with 1515 cattle samples from 54 cities collected in 9 different slaughterhouses reported a seroprevalence of 23.8% using an in-house IFA and the detection of C. burnetii by qPCR in 12.2% of the seropositive animals, highlighting the risk of infection for abattoir workers [64]. Molecular data suggest the existence of diverse MST and MLVA genotypes of C. burnetii in the state of São Paulo (ST74) [86]. In the publication, an ST73 was also detected in samples from Argentina. These genotypes were only reported in South America, reinforcing the need for epidemiological studies aiming at the isolation and characterization of human and animal strains to better understand the real impact on animal and public health [86].

Based on a recent analysis of typing methods and their correlation with the genomic groups [201], cattle and goat strains from Brazil belong to a genomic group (GG)-III, ovine strains from Brazil are correlated with GG-I, the same as ST17 detected in French Guiana, and the Argentinian strain At12 is placed in GG-IVb-associated, a genomic group with genetic relation with the Australian strains AuQ01, AuQ02, and Namibia (ST-30), which are isolates placed on GG-X, and strains placed in the GG-IV, like the isolate Leningrad-2 (ST-7). Strains from GG-III are usually detected in cattle and are rarely involved in human infection [201, 202]. On the other hand, C. burnetii belonging to GG-I can be found across the globe and shows elevated virulence, and strains from GG-IVb and GG-X were mainly isolated from human cases, illustrating the virulence potential of Brazilian ovine strains and the Argentinian isolate At12 described by Pacheco et al. [201, 203].

Although the transmission of C. burnetii through the consumption of contaminated milk from dairy animals remains controversial, studies conducted in the last decade in the state of Minas Gerais have shown samples of artisanal Minas cheese contaminated by this proteobacterium [79, 85, 87]. The risk has been assessed; in some cases, the consumption of dairy products from infected animals may induce seroconversion but no clinical manifestation [204].

In conclusion, C. burnetii infection in Brazil has been reported in different domestic and wild animal species. Although the pathogen seems widespread in the country, studies are concentrated in only six states. Hence, the prevalence of C. burnetii in animals is spatially biased. More studies are needed, especially in states without data, to better understand Q fever epidemiology in Brazil and its implications for animal and human health. Increasing Q fever knowledge would benefit preventive and control measures, reducing the risk of human infection and economic losses in livestock.

Argentina

Coxiella burnetii is a pathogen understudied in Argentina, with few or no publications in the second half of the twentieth century and early twenty-first. Serological studies have covered farm animals, pets, and humans, although the diversity of techniques and cutting-edge titles used make their comparison difficult. Only two publications use diagnosis through molecular biology, and they are also the only ones that studied ticks. Q fever is currently subject to compulsory notification.

The description of the first human cases of Q fever by serology in Argentina was carried out in 1957 in Córdoba [205] and in 1959 by a truck driver traveling through the provinces of Chaco, Formosa, Corrientes, and Missions [206]. The first human case in Buenos Aires City (BAC) was reported in 1962 (Ruggiero et al. 1962, unpublished data). Numerous serological studies were carried out in the 1950s and 1960s on different animal species, finding different seropositivities (Table 1) [68,69,70,71].

Notably, after these pioneering studies, there were no publications on the subject until 1997, when the suspicion of Q fever was reported in goats of Entre Ríos (although animals imported from Uruguay) from compatible signs by serology and was also detected in people related to the farm [207, 208]. In 1999, C. burnetii was detected in two imported cattle that were in quarantine and were then euthanized [208, 209].

Recently, clinically healthy dogs from poor neighborhoods of BAC were seropositive by indirect immunofluorescence [129], but not by IFI in cats [124, 125, 203, 205] or by molecular biology techniques in Rhipicephalus sanguineus sensu lato ticks [210]. Also, in BAC, 1/99 human sera were positive by indirect immunofluorescence [211].

In samples collected in 2004, Trezeguet et al. carried out a serological test using ELISA (phases I and II) as a screening test and FC as a confirmatory test in 840 goats of 56 farms covering practically the whole country; although only 9/840 (1,1%) were positive through FC, all belonged to the province of Buenos Aires (9/186; 4.8%) [212]. In 2005, in a sampling of 30 establishments in the province of Buenos Aires, 41 goats were detected with positive serology [208]. In a new study in 2007 covering goats from almost the entire country, Trezeguet et al. detected 11.6% (33/285) positive in goats from Buenos Aires, Catamarca, Mendoza, Río Negro, Santa Fe, and Santiago del Estero [70].

Pacheco et al. reported the detection of C. burnetii by PCR and its isolation by cell culture in Amblyomma tigrinum and Amblyomma parvum ticks from Córdoba [167].

Finally, in 2022, there was an outbreak of Q fever in slaughterhouse workers from Entre Ríos and Santa Fé [213, 214].

In conclusion, C. burnetii is a pathogen little studied in Argentina, with few or no publications in the last decade. Serological studies have covered farm animals, companion animals, and humans, although the diversity of techniques and threshold titles used makes their comparison difficult. On the other hand, only two publications use diagnosis through molecular biology, and they are also the only ones that studied ticks. It is necessary to study this neglected pathogen in depth in Argentina.

Chile

Information on C. burnetii in vertebrates and ticks in Chile is scarce. Moreover, its potential presence and distribution started to be investigated only recently. The first study analyzing a Chilean animal was published in late 2013. Archived blood samples of 30 endangered Darwin’s foxes (Lycalopex fulvipes) captured between 2009 and 2012 were analyzed by qPCR. All samples turned out to be negative [149]. Serum samples of 47 individuals of this same specie captured between 2013 and 2018 were later analyzed by means of a commercial ELISA kit, again with negative results [148]. Although this kit is not validated for this species, the combination of both negative PCR and serological results strongly suggests that this species has no regular contact with this bacterium. The cold environments of southern Chile where this fox survives, are unsuitable for most of the ixodid ticks, which partially explains the absence of this bacterium. A more spatially broad survey was performed using free-ranging dogs and the three species of fox present in Chile (Darwin’s fox, the Andean fox, Lycalopex culpaeus, and the South American grey fox, Lycalopex griseus) as sentinels. Again, none of the 358 dogs and 32 foxes surveyed from 5 different bioregions was seropositive according to the IFA test [133]. On the other hand, the analysis of blood samples from 55 bats belonging to five different species, collected opportunistically in three different Chilean regions, resulted in five positive samples by qPCR. All five samples belonged to any of the nine Brazilian free-tailed bats (T. brasiliensis) from the Metropolitan area included in the study. Although the sample size was small, the observed occurrence of this particular species in this region was remarkable [147•].

Regarding livestock, an investigation following a Q fever outbreak that took place in 2017 among dairy farm workers resulted in the presence of the bacteria in two out of 105 raw milk samples [91]. Lastly, an unpublished report mentioned that 13 alpacas imported from Chile to China were found to be seropositive for C. burnetii during quarantine [90].

No study has yet systematically analyzed potential arthropod vectors for the presence of C. burnetii in Chile. Only a single study reported the presence of a C. burnetii-like endosymbiont in an Ornithodoros amblus female, collected from the soil near a Humboldt penguin (Spheniscus humbolti) nesting area in Isla Grande de Atacama, Chile [176].

In summary, Chile can be considered a country with a low endemicity of C. burnetii. Nevertheless, cases in humans seem to have been underestimated [215], and the bacterium is present in domestic and wild animals in the country, as confirmed in bat and cow milk samples. In consequence, veterinarians and public health authorities must be aware of potential cases of Q fever. More extensive studies on domestic animals, wildlife, and ticks are necessary in Chile to know the actual distribution and impact of Q fever in Chile.

Colombia

Until the early 2010s, publications on human Q fever in Colombia were almost nonexistent [216]. In 2006, a seroprevalence of 23.6% of antibodies against C. burnetii was reported for the first time in rural field workers in the departments of Córdoba and Sucre [217]. In 2012, two cases of Q fever were identified in Colombia; one associated with endocarditis in Medellín, Colombia, and another case in a patient with pneumonia in Cali [218, 219]. Some seroprevalence studies in cattle from several regions of Colombia were conducted between 1961 and 1981 [92, 93, 95, 125], and then in the 2010s [71, 96, 97, 99], with regard to animal infections caused by C. burnetii. The carriage of anti-C. burnetii antibodies in cattle was generally high, around 20–25% and around 5% for sheep. Colombia performed one of the rare studies by PCR in goats and sheep in Latin America in milk and vaginal swabs, sowing 6% of positivity in sheep and 0.6% in goats [98]. It is important to note that the strain found had a 100% identity with the strain CbuK Q 154 and a 99% identity with the strain Guiana Cb175. A Colombian team also published one of the rare publications in 2022 of C. burnetii in bats from Macaregua caves in the Department of Santander with 5 to 7% positivity in the bats according to the species [150•]. Finally, four studies from Colombia have been published in search of C. burnetii in ticks such as Amblyomma variegatum, Rhipicephalus microplus, and Rhipicephalus sanguineus, leading to the evidence of Coxiella-like endosymbiont [97, 156, 181,182,183]. Thus, C. burnetii is circulating in Colombia, in livestock as well as in wildlife, especially bats, and deserves the implementation of studies in various animal taxa, both in animals and humans.

Conclusions and Perspectives

It is interesting to note that the literature reviews that follow one another from one decade to the next repeat tirelessly that Q fever is ubiquitous and that it is found everywhere in the world except in New Zealand [2, 68, 69, 207]. It is difficult to know where the claim that C. burnetii has been described in every country in the world has its roots. Recently, it was shown that numerous African countries have no trace of publication on C. burnetii in humans [208]. In 2016, we published a literature review on human Q fever in South America that showed that seven countries had never reported any cases of human Q fever or seroprevalence studies according to the available literature (Belize, Costa Rica, Guatemala, Guyana, Honduras, Paraguay, Suriname) [63]. The present literature review highlights the fact that there are no available data on C. burnetii infection in animals in many countries of Latin America. Thus, among the 21 countries or territories in continental Latin America, five (24%) have no publication related to C. burnetii detection in animals (Guyana, Suriname, Honduras, Bolivia, Paraguay), and 17 (65%) among the 26 Caribbean countries and territories, among which populated islands with livestock farms such as Jamaica, Haiti, and the Dominican Republic. There is thus an urgent need for studies on C. burnetii infection in animals in Latin America and the Caribbean, as well as in humans, to better understand the dynamic of this infection in the neotropical area.

While seroprevalence studies on C. burnetii in livestock exist in various countries, studies on wildlife and domestic animals are rare. Finally, studies on ticks seem to rule out this taxon as a contributor to C. burnetii transmission, although many Coxiella-LE have been found. Pets may be particularly valuable sentinel indicators for “one health” epidemiological studies because they share human environmental exposures. A significant effort has been made in recent years, particularly in Brazil and French Guiana, but most other Latin American countries, probably due to the low reporting of human cases, do not focus their research in this direction. In order to explain the contagion cycles in hyperendemic areas, it seems important to multiply studies on domestic and wild animals in terms of descriptive epidemiology and molecular epidemiology, considering the surveillance of particular strains. It would be interesting to see if the wild reservoirs are perhaps at the origin of this hyperendemicity, on the one hand, while the domestic ones would rather be at the origin of localized epidemics. There is also a need to study synanthropic species that are neither wildlife nor domesticated, such as mice and the two Rattus species. Nevertheless, it is important to address the question of cross-border spread through wildlife and imported domestic animals. Finally, this study demonstrates the need to include C. burnetii infection in the public health surveillance systems, considering its wide spectrum of clinical manifestations, and in animal health programs, especially for animals with reproductive disorders,” and the need for better surveillance systems. The environmental aspects of the disease (reservoir and source of contamination through feces, dust, etc.) may be considered to introduce a One Health approach to this infection.