Borrelia burgdorferi exposure in coyotes: an indicator of B. burgdorferi levels in urban versus rural environments

PICO question 
Do wild coyotes in the US that are in an urban habitat compared to a rural habitat have a higher prevalence of Borrelia burgdorferi seroconversion? 
  
Clinical bottom line 
Category of research question 
Prevalence 
The number and type of study designs reviewed 
Two papers, both utilising a cross-sectional study design 
Strength of evidence 
Zero 
Outcomes reported 
The relevant studies provide very limited to no evidence towards answering this PICO question. In one, while the absolute percentage of Borrelia-antibody-positive canines (including dogs in addition to coyotes) is higher in metropolitan areas, the effect was not found to be statistically significant, possibly due to their small sample sizes. In the second study, prevalence of antibodies against Borrelia was compared between different rural habitats, but no urban coyotes were tested as a comparison and thus the PICO question cannot be evaluated 
Conclusion 
There is a knowledge gap concerning the prevalence of Borrelia in coyotes and how it differs between urban and rural environments. Wild coyotes could be used as a sentinel species of Lyme disease activity and to assess potential for domestic pet and human infections, which would inform clinical differential diagnoses as well as testing and vaccination recommendations. More studies are needed before this PICO question can be answered in a confident manner 
  
How to apply this evidence in practice 
The application of evidence into practice should take into account multiple factors, not limited to: individual clinical expertise, patient’s circumstances and owners’ values, country, location or clinic where you work, the individual case in front of you, the availability of therapies and resources. 
Knowledge Summaries are a resource to help reinforce or inform decision making. They do not override the responsibility or judgement of the practitioner to do what is best for the animal in their care. 
  



Outcomes reported
The relevant studies provide very limited to no evidence towards answering this PICO question. In one, while the absolute percentage of Borrelia-antibody-positive canines (including dogs in addition to coyotes) is higher in metropolitan areas, the effect was not found to be statistically significant, possibly due to their small sample sizes. In the second study, prevalence of antibodies against Borrelia was compared between different rural habitats, but no urban coyotes were tested as a comparison and thus the PICO question cannot be evaluated

Conclusion
There is a knowledge gap concerning the prevalence of Borrelia in coyotes and how it differs between urban and rural environments. Wild coyotes could be used as a sentinel species of Lyme disease activity and to assess potential for domestic pet and human infections, which would inform clinical differential diagnoses as well as testing and vaccination recommendations. More studies are needed before this PICO question can be answered in a confident manner How to apply this evidence in practice The application of evidence into practice should take into account multiple factors, not limited to: individual clinical expertise, patient's circumstances and owners' values, country, location or clinic where you work, the individual case in front of you, the availability of therapies and resources.
Knowledge Summaries are a resource to help reinforce or inform decision making. They do not override the responsibility or judgement of the practitioner to do what is best for the animal in their care.

The evidence
While cross-sectional studies provide a high level of evidence for prevalence questions assessing exposure to this pathogen, neither of the two studies evaluated here were directly investigating whether Borrelia burgdorferi levels vary between urban and rural environments in wild coyotes of the United States. Due to this, the evidence that they provide towards the PICO question of interest is extremely limited.
In the Olson et al. study from 2000, seroprevalence was determined for canines including both dogs and coyotes in San Diego County, CA, allowing an assessment of risk for human cases of Borreliosis to be conducted. No statistical association was found between disease prevalence and lifestyle (rural vs urban vs indoors) of the canine host, but prevalence of B. burgdorferi antibodies was higher in metropolitan dogs, with the lowest prevalence being found in the most natural habitat; they postulated this could be due to importation of seropositive animals by high mobility owners or 'brush-border habitat' styles in urban landscaping supporting tick populations (Olson et al., 2000). However, when looking solely at coyotes, the sample size is small (n = 83), and there was only one positive individual, which was from a rural environment (Olson et al., 2000).
Five years later, Foley et al. (2005) published a study that looked deeper at variation in B. burgdorferi seroprevalence by vegetation type and climate within Ixodes pacificus-infested counties of California. While Foley et al. (2005) did not directly classify the tested individuals as urban or rural, 'urban' was a habitat option and urban areas are included in mapped visualisations of the data; examining their maps, it appears no or few coyotes from urban habitats were tested. With no tested urban coyotes clearly identifiable, this study does not allow us to answer the target PICO question but does explore variation within rural environments at a finer scale (Foley et al., 2005). Prevalence was higher in areas with higher rainfall as well as in blue oak woodland, montane hardwood, and redwood vegetation habitats, with seasonal peaks in June -July and November -April (Foley et al., 2005).
There were no papers that directly addressed this PICO question, which identifies a gap in knowledge where further research is needed.

Intervention details:
• Canine sera were obtained from local veterinarians, trappers, animal shelters, and humane societies between March 1992 and March 1993. • Urban or Rural 'Lifestyle' designation was included in the data collected by the investigators.  Outcome studied: The objective of this study was to determine the seroprevalence of B. burgdorferi in San Diego county as assessed in canines, who can serve as a sentinel for human risk of Lyme disease cases. Seroconversion was measured by ELISA and confirmed via Western blot test and/or IFA at a reference laboratory.

Main findings: (relevant to PICO question):
• For wild coyotes, no association with lifestyle (urban vs rural) was found (P = 1.0 for coyotes). • Only one coyote (rural) tested positive for Lyme disease.
• While no significance of burgdorferi occurred in the coyote population, the domestic dog data showed some interesting outcomes: o The location considered the wildest, Camp Pendleton, a US Marine Corps training area, had the lowest seroprevalence in dogs (0.5%). o Seroprevalence of dogs in the metropolitan sector was the highest (4.6%). o The authors postulate this could be due to importation of seropositive animals by high mobility owners or 'brush-border habitat' styles in urban landscaping supporting tick populations.

Limitations:
• When considering only coyotes, the sample size is small (total n = 83, seropositive n = 1), especially given that Borreliosis is rare. • The low number of positives limited the power of subset analysis. • Prevalence may be overinflated due to cross-reactivity of antibodies. • The authors did not utilise a vaccine free population, though they state only three dogs in the study had received the vaccine.

Intervention details:
• 215 coyotes were trapped by USDA Wildlife Services as part of a predator control program and convenience samples were taken from these animals for analysis. • 170 were tested for Borrelia burgdorferi, as blood from 45 coyotes from one county was not available when burgdorferi testing was performed. • Coyotes were split into several subsets including county, habitat, and month for analyses with varying sample sizes from 1 to 57. o 'Urban' was a habitat option and urban areas are included in mapped visualisations of the data; examining the maps, there are no urban coyotes easily identifiable. • Blood was either collected ante-mortem into ethylenediamine tetraacetic acid (EDTA) or heart clot blood was collected postmortem. • Sera was tested via indirect fluorescent antibody (IFA) for burgdorferi, with positive-IFA sera verified with Western blot testing. Outcome studied: The aim of this study was to examine spatial and temporal relationships at a smaller-scale to understand ecological drivers of disease among Anaplasma phagocytophilum and B. burgdorferiexposed coyotes. Prevalence of disease between different vegetation type and climate in Ixodes pacificus infested counties of California was investigated. Seroconversion for B. burgdorferi was tested via IFA and confirmed via Western blot test.

Main findings: (relevant to PICO question):
• There was a 18.9% seroprevalence (n = 148) with no association for sex in the tested coyotes. • B. burgdorferi occurred at higher levels in areas with higher rainfall and in blue oak woodland, montane hardwood, and redwood vegetation regions, and decreased in coastal sagebrush and cropland, providing a finer-scaled analysis within a rural category. • B. burgdorferi seroprevalence peaked seasonally (June -July and November -April).

Limitations:
• While rural coyotes were sampled in this study, urban coyotes do not seem to be included in the study population. • When split into subsets, sample sizes are smaller (n = 1-57).

Appraisal, application and reflection
Lyme disease in the United States is caused by the bacterium Borrelia burgdorferi sensu stricto (s.s.) and is vectored by Ixodes ticks, namely Ixodes scapularis in the eastern United States and Ixodes pacificus in the western United States that feed on many animal species including humans (Steere et al., 2016). Throughout the United States, Lyme borreliosis is an important emerging disease causing approximately 30,000 new human cases annually (CDC, 2019). While the majority of cases are in the northeastern/mid-Atlantic United States have occurred in all 50 states (CDC, 2019) and the highly diseased regions are expanding (Steere et al., 2016). This is also driven by climate change, allowing ticks to expand their geographic ranges and increase in abundance in recent decades presenting new and increased health risks to humans and animals (Sonenshine, 2018).
While Lyme disease is more debilitating in humans, dogs and other domestic animals can also contract Borrelia infections (Self et al., 2018). In fact, canines can act as sentinels of Lyme disease; testing for exposure to B. burgdorferi in dogs can be done during annual wellness visit examinations by veterinarians (Self et al., 2018). Veterinarians can then act as stewards of public health by reporting seroprevalence and cases of Lyme Borreliosis to the local public health department, helping to effectively track the spread of Ixodes ticks and B. burgdorferi as climate change advances. This serves not only additional animal patients that may present as cases in new areas, but also as a warning flag for human health; public health policy makers can make informed interventions if made aware of new or increasing cases of Lyme disease.
While useful, in private practice domestic animals as a sentinel may be limited because diagnostics is often limited by the permission and finances of pet owners. Coyotes (or other wild canids) are another sentinel species that can be sampled via convenience through animal control agencies, rehab facilities, and roadkill to add to the field's understanding or Borreliosis risk. While a convenience sample (use of easily available samples or samples on hand) is non-probabilistic and thus does have limitations on the amount it can inform the epidemiologic understanding of a disease, it is a good starting method that limits the challenges of wild species monitoring such as sampling effort and cost of the monitoring efforts. Despite these inherent challenges that exist in monitoring wildlife such as varying population densities and smaller sample sizes, using coyotes as a sentinel is an established practice for other pathogens such as Anaplasma phagocytophilum or Yersinia pestis, which are both zoonotic vector-borne diseases (Foley et al., 2005;and Brown et al., 2011). Testing coyotes can allow us to characterise disease levels in many different environments because coyotes will utilise and have successfully adapted to not only wilderness and rural but also urban environments (Crooks, 2002). Because coyotes utilise multiple environments and have large home ranges (Gehrt et al., 2009), they could also serve as conduits to take Borrelia-infected ticks between habitats or into new areas.
By characterising the exposure and role of Borrelia in coyotes, there is potential that daily clinical decisions of veterinarians can be informed. If veterinarians are aware of the prevalence of Borrelia in surrounding wildlife and the risk posed to their patients, it can hold an appropriate place on their differential list for clinically affected patients and can be utilised to inform testing and vaccination decisions. At the time of this writing, there is a paucity of information concerning how prevalence varies between urban and rural environments and if this can inform clinical decisions of veterinarians; more research is needed in order to determine if wild coyotes have a higher prevalence of B. burgdorferi seroconversion in an urban habitat compared to a rural habitat, and if wild coyotes are indeed a good sentinel species for B. burgdorferi in urban environments.