INTRODUCTION

Spotted fever group (SFG) rickettsiosis is an infectious, systemic disease caused mainly by the Gram-negative bacterium Rickettsia rickettsii (Wolbach 1919) Brumpt 1922, although other species of the same genus may cause a milder disease. These bacteria are mainly transmitted by ticks Amblyommasculptum (Berlese, 1888) and Amblyommaaureolatum (Pallas, 1772), which acquire them from an infected host during hematophagy, especially during the rickettsia' bloodstream phase, amplifying transmission to new hosts in subsequent blood repast. In addition to the human species, a wide range of domestic and wild mammals, including carnivores, rodents and herbivores, can be infected, and each of these taxa participates with different competences in transmission cycle and in the local maintenance of rickettsiae. Regardless of hematophagy, transovarian and transstadial transmission allows the spread of rickettsia to its progeny, but in the case of R. rickettsii infecting A. sculptum, the levels of vertical transmission do not ensure bacterium maintenance without the participation of amplifying vertebrate hosts.^2,3 In endemic areas for such rickettsiae, where capybaras (Hydrochoerus hydrochaeris, Linneaus, 1766) occur, the tick population is predominantly A. sculptum, while in non-endemic areas, there is a population balance of this species with Amblyomma dubitatum (Neumann, 1899). These invertebrates tend to have larger populations in degraded areas covered by grasses¹.

Domestic horses and free-ranging capybaras have been described as preponderant for the local persistence of rickettsia in urban and peri-urban environments. Horses appear to be less competent as reservoirs than capybaras, but both are capable of sustaining the local transmission cycle for tick vectors and, consequently, for humans. Regarding capybaras, an ecological phenomenon has added to their competence as a reservoir: the growing urbanization in Brazil since the end of the last century. The largest rodent in the world has become accustomed to human presence, gradually expanding its presence in urban centers with hydrology favorable to its presence, as it prefers riparian forest associated with freshwater courses and collections, where they find shelter and feed on riverine vegetation.^4,5,6

These SFG have been sparsely diagnosed in Minas Gerais municipalities since 1945. In Juiz de Fora, specifically, the first confirmed case was described in 1995. Since then, intermittent outbreaks of Brazilian spotted fever (BSF) in humans have been diagnosed by health agencies in the municipality, as well as horses, dogs and ticks infected with R. rickettsii in several regions.^7,8

As the participation of horses, dogs and their ticks in transmission cycle was already well characterized at the time, after successive episodes of SFG in Juiz de Fora since 1995, the City Hall carried out a program to fumigate horses in the municipality from 2007 to 2011, notably in the so-called “carrier” horses. These are informal workers who use horses for animal traction, performing small cargo transport services where other forms of transport are unfeasible, costly or non-existent. Usually, these workers live in poor regions and, commonly, horses “live” together with people in a complex equine stall room, intensely facilitating horse-tick-human transmission, being such an interaction a risk factor for the disease occurrence in humans, as well as the proximity to capybaras⁸.

The hydrological situation of the city of Juiz de Fora favors the presence of capybaras in the urban environment. The city is cut in its lower part by the Paraibuna river valley, which receives several streams and streams tributaries of all the hills and mountains surrounding the city. Thus, capybaras perennially occupy the Paraibuna valley and can be seen in other water bodies in the city. Occupation by capybaras is considered a risk for rickettsia transmission to humans, due to its central ecological role in the bacterium transmission cycle. Government agencies are considering the culling of capybaras in the urban perimeter as a way of controlling SFG in humans (Rafael Veríssimo Monteiro, personal communication), following examples from other situations⁹.

In this way, the conditions that describe the favorable scenario for SFG transmission to humans can be summarized as: i) proximity to transitional vegetation areas, with greater presence of grasses and low shrubs; ii) proximity to capybaras and horses; iii) presence of transmitting ticks, Amblyomma sculptum. All these conditions are present in Juiz de Fora.^9,10

Considering the conditions described above, this research aimed to investigate the possibility of a geospatial overlap between the location of human cases of spotted fever group rickettsiosis and mammalian reservoirs in the urban perimeter of Juiz de Fora. To achieve these objectives, we: i) georeferenced the human cases registered in the urban perimeter of Juiz de Fora, from 2003 to 2020; ii) we verified the geographic overlap of SFG cases with the areas of occurrence of horses and capybaras in the urban perimeter of the city, to estimate which of these hosts offers the greatest risk of transmission to humans; iii) we verified, through spatial analysis tools, the potential existence of aggregations of human cases in time and geographic space that could reveal areas in the city with greater risk of occurrence of SFG.

METHODS

Human cases

SFG human cases in the urban perimeter of Juiz de Fora were georeferenced, using public information and/or registered in the Information System of Notifiable Diseases (SINAN - Sistema de Informação de Agravos de Notificação) of the Ministry of Health, from 2003 to 2020. According to Resolution 510/2016 of the Brazilian National Health Council (CNS), this type of study is exempt from an ethical license. One of us (JGCJ) was primarily responsible for epidemiological surveillance and epidemiological screening of 80% of cases, in addition to collecting data from SINAN. From these sources, in addition to the geographical coordinate relating to the affected person’s place of residence, information was collected on whether the probable infection site of the case was autochthonous or not in the municipality, gender, whether there was hospitalization, the type of diagnosis used and the final result (death/cure). For georeferencing, the WGS 84 datum, and the QGIS^® 3.16 software and the Rstudio 1.1^® platform running R^® version 3.5.3 were used for map generation and statistical analysis. Fisher’s exact test was used to compare the rates of death from SFG between the sexes, with a significance level at p<0.05.

Capybara location

To map the geographic occurrence of capybaras, observations from two different sources were included, obtained from September 2016 to December 2017:

a Information on the collection of capybaras in the urban perimeter made by the fire department teams, IBAMA (Brazilian Institute for the Environment and Renewable Natural Resources) and IEF (State Institute of Forestry, Minas Gerais), georeferenced from the collection address (5 locations).

b Information obtained from the Information System for Wild Health (SISS-GEO) application, managed by the Biodiversity and Wild Health Institutional Platform/Fiocruz, for monitoring wild animals and generating epizootic alerts. The application captures the geographic coordinates by GPS of the animal’s photographic record in real time, in addition to other observations.¹¹

Domestic horse location

The locations of sightings of domestic horses found in the city were mapped, georeferenced from Google Earth^® satellite image. Cart horses, although they circulate throughout the city, occupy fixed points of waiting for work, often on the banks of the Paraibuna river. Four of the known points have been geolocated. The places defined as tick fumigation points for domestic horses were also mapped, chosen by the city hall precisely because they are places of concentration of cart horses for grazing, walking and/or waiting for requisition for work. Tick fumigation took place from 2007 to 2011, from June to November, with intervals of approximately 21 days between each tick fumigation, at nine different points, also including the horses eventually parked on the banks of the Paraibuna river.

Epidemiological and geospatial analyzes

Two epidemiological and geospatial analyzes were performed concerning human cases. First, the density of observations in the study area was assessed, a 1^st order property of the case distribution pattern, capable of revealing local clusters of cases. For this analysis, a heatmap or Kernel density map was generated, with a parabolic density probability function, based on geographic occurrence of human cases of BSF in the urban perimeter of Juiz de Fora, from 2003 to 2020 (Heatmap plugin QGIS 3.10). A distance of 2,000 m was used as the influence radius of each point in the model, based on the fact that the average daily walking distance of a normal person is around 4,000 m. This heatmap (Kernel) intends to reveal an eventual overlap of the area of daily displacement of human cases of SFG that occurred in the analyzed period. Areas of great overlap would be areas of greater risk of exposure to rickettsia contamination. The property of the 2^nd order of distribution pattern of cases was also assessed, i.e., whether each case can interfere with the occurrence of others. To this end, we calculated the Average Nearest Neighbor (ANN) of observed cases and compared it with a Monte Carlo simulation (600 iterations), based on the random generation of cases in the geographic occurrence window delimited by the cases that have already occurred. We compare the observed ANN with the average obtained in the 600 iterations, generating the probability that the observed cases present a certain pattern of geographic aggregation (random, uniform or aggregated).^12-14

RESULTS

A total of 34 cases of human SFG were identified in the urban perimeter of Juiz de Fora in the period from 2003 to 2020, i.e., an average frequency of two cases of SFG per year. The fatality rate was 47% among those infected (16 deaths), and there was a statistical difference in lethality between the sexes (Fisher’s exact test, p=0.03), with women presenting a lethality rate of 85.7% (6 deaths in 7 cases). All cases underwent hospitalization. Thirty of the cases had laboratory confirmation by indirect fluorescent antibody test (IFAT), while the rest underwent clinical diagnosis; however, two of these cases occurred almost simultaneously with cases of relatives who had laboratory confirmation by IFAT.

Human case geospatial distribution (2003-2020)

The temporal distribution of BSF cases in Juiz de Fora can be seen in Figure 1. It is noted that there are annual gaps in disease occurrence, while some years bring together several cases. Apparently, there is a 6-year cycle between epidemic peaks. Although cases can happen in almost any month of the year, there is a concentration of cases in the period from August to November, months that concentrated 56% of cases in the analyzed period.

Geographic occurrence of human cases in the considered period can be seen in Figure 2. The concentration of disease cases can be observed in the northwest to east quadrants, mainly in peripheral places to the Krambeck Forest. Another cluster of cases to the south can also be characterized.

Host geospatial distribution

Figure 2 shows the distribution of capybaras and domestic horses that were located in the urban perimeter of the city. Capybaras are stably distributed along the course of the Paraibuna river, but can occasionally be seen in other watercourses that cross the entire municipality. They are mostly seen in family groups, but solitary individuals can be found. Cart horses occupy several points on the edge of the Paraibuna, usually individually. The nine tick fumigation points, marked on the map in Figure 2, are distributed throughout the urban perimeter. It was not possible to map carters’ horses that are housed inside the families’ homes.

Epidemiological and geospatial analyzes

Figure 3 shows a heatmap of occurrence of human cases of BSF in the municipality of Juiz de Fora, resulting from the first-order geospatial analysis. It can be seen that, under the conditions of comparison established here, there seems to be an area that concentrates the probability of acquiring the infection in the municipality in the years analyzed. On the other hand, the second-order analysis indicated a geospatial pattern of distribution of cases of aggregate character, since the average distance to the nearest neighbor observed (712.1 m) was statistically smaller than the average obtained in the Monte Carlo simulation (1,422.7 m; p=0.002). This aggregate character indicates that, in each outbreak, the cases tend to happen close together. The neighborhoods in the urban perimeter of Juiz de Fora that are included in the hottest spot in Figure 3, relative to overlap intensities greater than 5 (Bairu, Bandeirantes, Granjas Bethânia, Manoel Honório, Progresso and Santa Terezinha), had an average Family Development Index (FDI) of 0.733 (ranging from 0.665 to 0.779), an index equivalent to category D3, low-income families.¹⁵

DISCUSSION

The occurrence of SFG in Juiz de Fora follows a pattern equivalent to an endemic area with disease outbreaks. In the historical sequence analyzed here, it can be noted that there are years with a higher occurrence (7 cases in the highest frequency), while in other years, there are no cases (up to 3 years without cases). It is noteworthy that the 3-year gap in occurrence of human cases coincided with the execution of the city hall’s tick fumigation program from 2007 to 2011 (Graph 1, panel A). Regarding lethality, the local rate is within the disease historical rates, although the lethal involvement in women has been statistically pointed out.¹⁶ The higher number of cases in men than in women is also in line with the traditional disease epidemiological characteristics, as men have greater exposure to transmitting ticks during their daily work. Another common aspect of SFG is the concentration of human cases from August to November of each year, when larvae and nymphs predominate in the local population of ticks. These immature forms are considered to be at greater risk for transmission, as they are of small size, which makes it difficult to identify the presence of a tick and does not alert the infected person to infestation, increasing the probability that the tick remains on individuals and the time required for rickettsiae transmission.¹⁷

The geographic distribution of cases seems to follow an aggregate pattern in time and geographic space. Over the years, the cases were concentrated in a preferred geographic region (Figures 2 and 3), and the cases tended to occur in regions close to each other or even within the same family/residential group (Average Nearest Neighbor). This could be due to: i) higher concentration of infected ticks in the region where such cases occur; ii) increased likelihood of adequate contact between ticks and people in such locations.

Two factors should be associated to justify the higher concentration of locally infected ticks: i) having vertebrate animals that support a large population of ticks; ii) that rickettsia is circulating in the region. Thus, these two factors would allow that, with proximity between people and the population of infected ticks, transmission cycle can be completed with human infection, that is, there is an increase in the probability of adequate contact between ticks and people. Other factors may be associated with this higher probability of contact between a tick and a human host, such as residents’ occupation, whether they work with horses or frequent pastures for leisure or work, risk factors already characterized for SFG. The presence of this set of factors is complete in the common scene in the countryside cities of Brazil: people living in situations of vulnerability, living together or very close to their horses’ stalls, often in peri-urban places.^9,10

In the situation described above, the participation of capybaras in occurrence of local human cases can be considered small, with the exception of places where rickettsia is introduced from these rodents. This situation does not seem to be the case in Juiz de Fora, where rickettsia has already been identified in A. sculptum in a neighboring municipality and in R. sanguineus in the municipal kennel.^7,18 However, recent research on ticks collected on the banks of the Paraibuna river revealed the presence only of Rickettsia amblyommii, a bacterium whose pathogenicity is unknown.¹⁹ This condition of lesser relevance of capybaras in the local epidemiological chain is strengthened by the decrease in human cases during the period in which the city carried out tick fumigation of cart horses, indicating that transmission from ticks parasitizing horses is the most common epidemiological situation (Figure 1, panel A). Additionally, the municipal legislation of Juiz de Fora started to prohibit the circulation of cart horses in the urban perimeter as of December 2019. Although it is premature to assess the effect of such legislation for the future, we hope that, if horses really are the main carriers of ticks transmitting BSF rickettsiae in Juiz de Fora, the annual average of cases drops to levels below the current 2 cases per year.

The spatial distribution of capybaras also does not show a spatial overlap in agreement with human cases (Figures 2 and 3). Capybaras are distributed throughout the city following the waterways, and reside permanently on the banks of the Paraibuna river. However, most human cases did not occur near this watercourse, weakening the possibility that ticks of the capybaras of this river are the main sources of transmission to humans, although these rodents possibly contribute to the maintenance and potential infection by R. rickettsii in local populations of A. sculptum, a tick species proven to be resident in endemic areas for the disease.^1,20 The presence of established immunity against rickettsia in these family groups of capybaras permanently residing in the riverbed of the Paraibuna cannot be ruled out, which would reduce rickettsia transmissibility to the tick population on the edge of the Paraibuna.²¹ Thus, the presence of these fixed groups of capybaras occupying the Paraibuna river could be attenuating the infection of R. rickettsii on the local populations of A. sculptum. The finding of R. amblyommii adds another species of rickettsia to the local population of capybaras and ticks, and this may be involved in the local immunity of these hosts against R. rickettsii, a phenomenon already characterized.^19,22

Culling of wild animals is often proposed as a way to control human and farm animal diseases, in which these animals are considered relevant in transmission cycle⁹. This approach may not always be the most appropriate, as ecological studies and mathematical simulations show that, in certain situations, the disturbance caused in the wild population subjected to such a procedure can increase disease prevalence in them, increasing the risk for human and animal populations for which disease control was desired.^23-25 Bearing this in mind, the culling of capybaras is not suggested as a strategy to control BSF in the urban perimeter of Juiz de Fora without a greater ecoepidemiological knowledge of the place, since such action may be inappropriate and/or counterproductive.

Acknowledgments

The authors would like to thank the Municipality of Juiz de Fora for their authorization and assistance in data search and compilation.

REFERENCES

1 Luz HR, Costa FB, Benatti HR, et al. Epidemiology of capybara-associated Brazilian spotted fever. PLoS Negl Trop Dis. 2019; 13:e0007734 https://doi.org/10.1371/journal.pntd.0007734

2 Costa FB, Gerardi M, Binder LC, et al. Rickettsia rickettsii (Rickettsiales: Rickettsiaceae) Infecting Amblyommasculptum (Acari: Ixodidae) Ticks and Capybaras in a Brazilian Spotted Fever-Endemic Area of Brazil. J Med Entomol. 2020; 57(1), 308–11 https://doi.org/10.1093/jme/tjz141

3 Soares JF, Soares HS, Barbieri AM, et al. Experimental infection of the tick Amblyomma cajennense, Cayenne tick, with Rickettsia rickettsii, the agent of Rocky Mountain spotted fever. Med Vet Entomol. 2012; 26:139–51. https://doi.org/10.1111/j.1365-2915.2011.00982.x

4 Ueno TEH, Costa FB, Moraes-Filho J, et al. Experimental infection of horses with Rickettsia rickettsii.Parasit Vectors. 2016; 9:499 https://doi.org/10.1186/s13071-016-1784-y

5 Souza CE, Moraes-Filho J, Ogrzewalska M, et al. Experimental infection of capybaras Hydrochoerus hydrochaeris by Rickettsia rickettsii and evaluation of the transmission of the infection to ticks Amblyommacajennense. Vet Parasitol. 2009; 161:116-21 https://doi.org/10.1016/j.vetpar.2008.12.010

6 Nasser JT, Lana RC, Silva CMS, et al. Urbanization of Brazilian spotted fever in a municipality of the southeastern region: epidemiology and spatial distribution. Rev Bras Epidemiol. 2016; 18:299-312 https://doi.org/10.1590/1980-5497201500020002

7 Pacheco RC, Moraes-Filho J, Guedes E, Silveira I, et al. Rickettsial infections of dogs, horses and ticks in Juiz de Fora, southeastern Brazil, and isolation of Rickettsia rickettsii from Rhipicephalus sanguineus ticks. Med Vet Entomol. 2011; 25:148–55 https://doi.org/10.1111/j.1365-2915.2010.00915.x

8 Galvão MAM, Ribeiro JGL, Machado J, et al. Informe Técnico de febre maculosa. Belo Horizonte: Coordenação Estadual de Controle de Zoonoses; 2001.

9 Nunes FBP, Silva SC, Cieto AD, et al. The Dynamics of Ticks and Capybaras in a Residential Park Area in Southeastern Brazil: Implications for the Risk of Rickettsia rickettsii Infection. Vector-Borne Zoonot Oct. 2019; 711-716 http://doi.org/10.1089/vbz.2019.2479

10 Souza CE, Pinter A, Donalisio MR. Risk factors associated with the transmission of Brazilian spotted fever in the Piracicaba river basin, State of São Paulo, Brazil. Rev Soc Bras Med Trop. 2015; 48:11-7 https://doi.org/10.1590/0037-8682-0281-2014

11 Chame M, Barbosa HJC, Gadelha LMR, et al. SISS-Geo: Leveraging Citizen Science to Monitor Wildlife Health Risks in Brazil. J Healthc Inform Res. 2019; 1:1-27 https://doi.org/10.1101/286740

12 Auchincloss AH, Gebreab SY, Mair C, et al. A review of spatial methods in epidemiology, 2000-2010. Annu Rev Public Health. 2012; 33:107-22 https://doi.org/10.1146/annurev-publhealth-031811-124655

13 Fritz CE, Schuurman N, Robertson C, et al. A scoping review of spatial cluster analysis techniques for point-event data. Geospat Health. 2013; 7:183-198 https://doi.org/10.4081/gh.2013.79

14 Althoff T, Sosic R, Hicks JL, et al. Large-scale physical activity data reveal worldwide activity inequality. Nature. 2017; 547:336–9 https://doi.org/10.1038/nature23018

15 Horta TC, Monteiro TC. Mapa social: análise da situação do desenvolvimento familiar em Juiz de Fora [Agenda Família 6mil]. Juiz de Fora: Editora Funalfa, 2012.

16 Araújo RP, Navarro MBMA, Cardoso TAO. Febre maculosa no Brasil: estudo da mortalidade para a vigilância epidemiológica. Cad Saúde Colet. 2016; 24:339-46 https://doi.org/10.1590/1414-462X201600030094

17 Del-Fiol FS, Junqueira FM, Rocha MCP, et al. A febre maculosa no Brasil. Rev Panam Salud Publica. 2010; 27(6):461–6. http://dx.doi.org/10.1590/S1020-49892010000600008

18 Guedes E, Leite RC, Prata MCA, et al. Detection of Rickettsia rickettsii in the tick Amblyomma cajennense in a new Brazilian spotted fever-endemic area in the state of Minas Gerais. Mem Inst Oswaldo Cruz. 2005; 100:841-5 https://doi.org/10.1590/S0074-02762005000800004

19 Nunes EC, Vizzoni VF, Navarro DL, et al. Rickettsia amblyommii infecting Amblyomma sculptum in endemic spotted fever area from southeastern Brazil. Mem Inst Oswaldo Cruz. 2015; 110:1058-61 https://doi.org/10.1590/0074-02760150266

20 Polo G, Acosta CM, Labruna MB, et al. Transmission dynamics and control of Rickettsia rickettsii in populations of Hydrochoerus hydrochaeris and Amblyommasculptum. PLoS Negl Trop Dis. 2017; 11: e0005613 https://doi.org/10.1371/journal.pntd.0005613

21 Ramírez-Hernández A, Uchoa F, Azevedo-Serpa MC, et al. Clinical and serological evaluation of capybaras (Hydrochoerus hydrochaeris) successively exposed to an Amblyommasculptum-derived strain of Rickettsia rickettsii. Sci Rep. 2020; 10:924 https://doi.org/10.1038/s41598-020-57607-5

22 Chen LF, Sexton DJ. What’s New in Rocky Mountain Spotted Fever? Infect Dis Clin N Am. 2008; 22:415–432 https://doi.org/10.1016/j.idc.2008.03.008

23 Carter SP, Delahay RJ, Smith GC, et al. Culling-induced social perturbation in Eurasian badgers Meles meles and the management of TB in cattle: an analysis of a critical problem in applied ecology. Proc R Soc Lond B. 2007; 274: 2769–2777 https://dx.doi.org/10.1098/rspb.2007.0998

24 Vial F, Donnelly CA. Localized reactive badger culling increases risk of bovine tuberculosis in nearby cattle herds. Biol Lett. 2012; 8: 50–53 https://dx.doi.org/10.1098%2Frsbl.2011.0554

25 Prentice JC, Marion G, White PCL, et al. Demographic Processes Drive Increases in Wildlife Disease following Population Reduction. PLoS ONE. 2014; 9(5): e86563 https://doi.org/10.1371/journal.pone.0086563