This review was undertaken at the Army Institute of Public Health, Maryland, and the Center for Wildlife Health, Tennessee, where the two authors are employed.
Dr G. J. Hickling. The Center for Wildlife Health, 274 Ellington Plant Sciences, University of Tennessee Institute of Agriculture, Knoxville, TN 37996, USA. Email: email@example.com
Since its emergence in the north-eastern and upper mid-western United States in the 1970s, Lyme disease, caused by Borrelia burgdorferi, has captured the public’s attention as the nation’s most prevalent vector-borne zoonotic disease. In contrast, recent publications on tick-pathogen systems in the eastern United States, and findings from Department of Defense investigations of ticks found biting military personnel, indicate that residents of the south-eastern United States are primarily at risk from emerging diseases caused by tick-borne pathogens other than B. burgdorferi. The risk of contracting these diseases varies greatly among states as a consequence of regional variation in the abundance of the key vector tick species. Moreover, this risk is changing, because tick distributions are in flux. To improve health outcomes, health providers need better information and awareness regarding which tick species bite humans in each state and which zoonotic pathogens are prevalent in these ticks. Effective diagnosis, treatment, control and reporting of tick-borne disease in the south-eastern United States require that health providers think ‘beyond Lyme’ and consider the marked regional differences in the tick species that bite humans and in the pathogens that these ticks carry.
• The risk of contracting several tick-borne diseases, including Lyme disease, varies greatly among eastern US states because of regional variation in the abundance of key vector tick species.
• Moreover, these patterns of geographic risk are changing because tick distributions are in flux.
• Improving health provider knowledge of tick species distributions, and of tick-borne diseases other than Lyme disease, would lead to better diagnosis, treatment and reporting of these diseases, particularly in the south-eastern United States.
Lyme disease has been spreading in the north-eastern and upper mid-western United States since the 1970s and has become the nation’s most prevalent vector-borne zoonotic disease. Concurrently, the south-eastern United States has seen emergence and resurgence of other tick-borne diseases (TBDs), several only recently identified. Here, we review the close association of selected zoonotic pathogens with particular tick species. We then use reports from the literature and data collected at military installations throughout the eastern United States to document marked regional variation in the abundance of these tick species and show that this leads to corresponding state-by-state variation in which TBDs pose the greatest risk to human health. We conclude by discussing the implications of these regional differences for education and outreach efforts concerning TBDs in the eastern United States.
Recent literature on tick-pathogen systems in the eastern United States was reviewed. This is an active area of research, with one-third of the articles cited here published in 2009–2011. In addition, previously unpublished findings from the Department of Defense (DOD) Human Tick Test Kit Program are reported. The DOD Human Tick Test Kit Program is a tick identification and polymerase chain reaction (PCR) testing service for ticks found biting military personnel, dependents, and DOD civilian employees. The methods of tick identification and PCR used have been described previously in Stromdahl et al. (2001, 2003, 2011).
In this review, the ‘south-eastern’ United States refers to the region where Amblyomma americanum is the most abundant tick species. This region comprises Virginia, North Carolina, South Carolina, Georgia, Florida, Alabama, Mississippi, Arkansas, Tennessee and Kentucky, together with parts of bordering states with similar vector tick ecologies (southern Maryland, southern Missouri, eastern Kansas, eastern Oklahoma and eastern Texas).
Within this region, changes in tick distribution and abundance have been apparent in recent years, although these have been difficult to quantify. Tick surveillance efforts have varied in intensity and methodology, so there are problems and limitations with aggregating studies. Collection methods are biased for certain species and life stages, and distribution information is affected by the amount of effort now and in the past to sample multiple habitats and hosts, throughout the year, with methods that provide as unbiased a sample as possible. In particular, lack of detection of a tick species at a particular time of sampling is not proof of that species’ absence from the survey area. Despite these limitations, strong trends are apparent in the available data, and we highlight these to support our argument that the increasing risk of tick-borne pathogens in the United States is related, in part, to regional changes in tick abundance and distribution.
Tick Species Found Biting Humans in the Eastern United States
The five tick species most likely to bite humans in the eastern United States, and the zoonotic pathogens that they are known to transmit, are summarized in Table 1 and described in more detail below. There is pronounced regional variation in which species are most commonly found on humans; in particular, A. americanum dominates in south-east, whereas Ixodes scapularis dominates in the North (Fig. 1).
Table 1. Main vectors of tick-borne human diseases in the eastern United States, with examples of the zoonotic pathogens and diseases associated with each tick species
aCoxiella burnetii, the agent of Q fever, is also potentially vectored by one or more of the tick species listed here.
bFormerly human monocytic ehrlichiosis (HME).
cThe role of this tick in the tularaemia transmission cycle is not completely understood.
dFormerly human granulocytic ehrlichiosis (HGE) and human granulocytic anaplasmosis (HGA).
Lone star tick (Amblyomma americanum)
Ehrlichia chaffeensis infectionb
Ehrlichia ewingii infection
‘Panola Mountain Ehrlichia’
Ehrlichiosis, human, undetermined
Spotted fever rickettsiosis
Southern tick-associated rash illness
Red meat allergy
American dog tick (Dermacentor variabilis)
Spotted fever rickettsiosis
Gulf Coast tick (Amblyomma maculatum)
Spotted fever rickettsiosis
Brown dog tick (Rhipicephalus sanguineus)
Spotted fever rickettsiosis
Blacklegged tick (Ixodes scapularis)
Anaplasma phagocytophilum infectiond
Ehrlichia muris-like infection
The lone star tick, A. americanum, is by far the most abundant human-biting tick in the south-eastern United States (Merten and Durden, 2000). All life stages – adult, nymph and larva – are active during the warmer months and all attack humans aggressively. Multiple concurrent tick bites from this species are commonplace; ∼15% of individuals submitting A. americanum to the DOD Human Tick Test Kit Program submit multiple ticks (EYS, pers. obs.).
Amblyomma americanum larvae (often referred to as ‘seed ticks’) are not known to transmit human pathogens but nevertheless can cause significant discomfort. A female A. americanum will oviposit a mass containing thousands of eggs, and after hatching, the resulting larvae remain clumped on vegetation where they quest for hosts. A passing human can be infested with hundreds of larvae from a brief swipe of the vegetation. Larval bites produce erythematous pruritic papules in response to tick salivary chemicals, and attached larvae are so small that they may not be seen or correctly identified by tick bite victims and physicians. Consequently, the papules may be treated as a rash or insect bites and the larvae not removed (Elston, 2006; Fisher et al., 2006). Similarly, nymphal A. americanum populations can be huge, with encounter rates exceeding 500 per hour reported by soldiers training in the field (Stromdahl et al., 2000).
There has been longstanding speculation as to whether A. americanum vectors B. burgdorferi, the agent of Lyme disease (Barbour, 1996). As described in our later section on disease aetiology, there is good evidence that it does not. Questing and on-host A. americanum collected from the field are very rarely infected with B. burgdorferi. Moreover, laboratory transmission experiments indicate that A. americanum larvae feeding on infected hosts: (i) rarely become infected with B. burgdorferi; (ii) are highly inefficient at maintaining B. burgdorferi through the moult; and (iii) are incapable of transmitting the spirochete to susceptible hosts (references in Tsao, 2009; Feder et al., 2012).
An important consideration for healthcare professionals is that the bite of A. americanum is associated with the development of southern tick-associated rash illness (STARI), a Lyme disease-like syndrome of unknown aetiology that is well documented in the south-east (Masters et al., 2008) and emerging (but likely under-diagnosed) along the Atlantic seaboard (Feder et al., 2011, 2012). Bites of A. americanum also have been associated with food allergy; some tick bite victims develop an IgE antibody response, presumably to compounds in tick saliva, that later triggers anaphylaxis or urticaria when red meat is consumed (Commins et al., 2011). An oligosaccharide (alpha-galactose) is present in the tissues, muscle, fat and blood of non-primate mammals, and eating their meat creates a risk for anaphylaxis in individuals who have acquired a tick bite-induced IgE response to alpha-galactose.
The American dog tick, Dermacentor variabilis, frequently bites humans (Fig. 1) and traditionally has been associated with the transmission of R. rickettsii, the agent of RMSF. Indeed, this species was found throughout a recent focus of RMSF cases in south-eastern states (Openshaw et al., 2010). However, molecular investigations of D. variabilis from areas reporting RMSF have found an almost complete absence of pathogenic R. rickettsii and instead have detected non-pathogenic SFG rickettsiae such as ‘R. amblyommii’ and R. montanensis (Moncayo et al., 2010; Stromdahl et al., 2011). Many past ‘RMSF-positive’ human diagnoses are now suspected to be serological cross-reactions with these other rickettsiae. In 2010, the US Centers for Disease Control and Prevention (CDC) changed their case reporting category from ‘RMSF’ to ‘Spotted Fever Rickettsiosis (including RMSF)’ (Openshaw et al., 2010).
There are several reasons why D. variabilis is considered to have limited ability to transmit RMSF. First, the RMSF agent, R. rickettsia, is a virulent pathogen that has lethal effects on Dermacentor spp. tick hosts; consequently, it is found at very low prevalence in tick populations. A very high proportion of immatures of a closely related species, D. andersoni, experimentally infected with R. rickettsii failed to moult to adults or died before taking the adult meal (Niebylski et al., 1999). Second, D. variabilis produces rickettsiostatic proteins that help to protect the tick from rickettsial infection (Ceraul et al., 2007, 2008, 2011). Third, the immature stages of D. variabilis do not attack humans – the adults do, but they are large and relatively conspicuous (Stromdahl et al., 2011).
Dermacentor variabilis is a widespread tick that overlaps the distribution of I. scapularis (Merten and Durden, 2000) and thus also of Lyme disease, which inevitably has led to public concern as to whether D. variabilis is spreading that disease. Compounding this concern, viable B. burgdorferi spirochetes can occasionally be found in engorged D. variabilis removed from infected hosts and in questing ticks collected from vegetation. Indeed, B. burgdorferi DNA sequences have been amplified from D. variabilis ticks removed from humans (Piesman and Happ, 1997; Stromdahl et al., 2001).
Although D. variabilis can acquire B. burgdorferi spirochetes from an infected host and maintain some alive through the moult, the tick cannot then transmit them to a new host. This is because feeding on the host triggers spirochetes to leave the protection of the tick gut and begin migrating through the tick’s haemocoel towards its salivary glands. During this migration, the spirochetes are killed by borreliacidal peptides in the haemolymph (Soares et al., 2006). Dermacentor variabilis that feed on B. burgdorferi-infected hosts are consequently unable to infect new, susceptible hosts (Sanders and Oliver, 1995; and references therein), making this tick unimportant as a vector of the Lyme disease agent (Tsao, 2009).
Dermacentor variabilis adults have been identified as competent vectors of F. tularensis (Reese et al., 2011), but the role of this tick in the tularaemia transmission cycle is not yet fully understood. The current national focus of tick-transmitted tularaemia, Arkansas and Missouri, is well within the range of D. variabilis, and D. variabilis adults routinely bite humans. Nevertheless, the overwhelming majority of tick bites in those states are from A. americanum. From 2004 to 2010, only 9% (38/432) of ticks removed from military personnel at Ft. Leonard Wood, MO, and submitted to the DOD Human Tick Test Kit Program were D. variabilis, the remainder were A. americanum. Similarly, in vegetation sampling conducted in 2009 at 79 sites within the Arkansas-Missouri tularaemia focus, only 19/4661 ticks collected were D. variabilis, 4632 were A. americanum (Brown et al., 2011). Francisella tularensis infection rates of D. variabilis reported from this area have been extremely low (<0.1%, similar to infection rates in A. americanum; Hopla, 1953; Calhoun and Alford, 1955). Although D. variabilis is not the predominant tick in the region, and infection prevalence is low, the potential for tularaemia transmission by bite of D. variabilis should not be discounted because recent laboratory experiments suggest that transmission can occur rapidly, on the first day of attachment. Paradoxically, in another epizootic of tularaemia associated with the populations of D. variabilis at Martha’s Vineyard, MA, F. tularensis infection rates of D. variabilis have been found to be as high as 2–5%; however, most human cases reported there are not the ulceroglandular type associated with the bite of a tick (Goethert and Telford, 2010).
Ixodes scapularis, the blacklegged tick (often referred to as the ‘deer tick’), is common in many parts of north-east and upper Midwest (Merten and Durden, 2000), where it frequently bites humans. The northern form was known for a time as I. dammini, a species separate from I. scapularis, the southern form. Studies in the 1990s demonstrated mating compatibility and genetic similarity of the two forms, which consequently were re-synonymized as I. scapularis (Oliver et al., 1993b, Wesson et al., 1993).
Ixodes scapularis is widespread in south-eastern States; indeed, a 1945 distribution map for the tick indicated that most populations known at that time occurred below latitude 37°N (Bishopp and Trembley, 1945). In recent decades, northern I. scapularis populations have undergone dramatic growth and geographic spread that has been attributed to habitat change and increasing populations of wildlife hosts, particularly white-tailed deer (e.g. Barbour and Fish, 1993). Additional factors must be important, however, as similar habitat and host population changes have occurred in the south-east without triggering equivalent increases in southern I. scapularis populations. Consequently, typical I. scapularis densities in south-eastern states are an order of magnitude lower than in the North (G.J. Hickling, unpublished data).
As noted previously, only a very small proportion of the ticks submitted to the DOD Human Tick Test Kit Program from the south-east were I. scapularis (Fig. 1). From 2004 to 2010, the four southernmost states in the region (Louisiana, Alabama, Florida and South Carolina) submitted a total of only 40 individual I. scapularis to the Program, whereas over the same period, nearly 3000 I. scapularis adults and nymphs were received from military personnel in mid-Atlantic, north-eastern and upper mid-western states.
In addition to low tick abundance, two other factors reduce the risk to human health of I. scapularis in the south-east. First, there are behavioural differences such that nymphal I. scapularis in the South rarely attack humans (Goddard and Piesman, 2006). This behavioural effect is apparent in the DOD Human Tick Test Kit Program data: all 40 of the I. scapularis received from the southernmost states (listed above) were adult females, whereas 3 of 26 I. scapularis submitted from North Carolina, 19 of 48 from south-eastern Virginia and 39 of 118 from northern Virginia were nymphs. Second, the three zoonotic agents commonly carried by this tick –B. burgdorferi, Anaplasma phagocytophilum and Babesia microti– are all associated with northern tick populations; prevalences of these pathogens vary from ‘very low’ to ‘undetectable’ among I. scapularis populations in the south-east. For example, for the DOD Human Tick Test Kit Program data cited above, 0 of 66 I. scapularis from the southernmost states and North Carolina were PCR-positive for any of these pathogens, whereas 4 of 48 ticks from south-eastern Virginia and 12 of 118 ticks from northern Virginia were PCR-positive for B. burgdorferi. Ixodes scapularis submitted to the Program from the mid-Atlantic, north-east and upper Midwest typically had robust (>20%) infection prevalences of B. burgdorferi, plus low but measureable rates of A. phagocytophilum and Ba. microti.
Until the index human case in 2002, A. maculatum was considered a veterinary pest and vector of veterinary disease. The tick can tolerate relatively xeric exposure, and feeds readily on pastured livestock, sometimes causing deforming injuries at feeding sites (‘gotch ear’) and lesions that permit screw-worm (Cochliomyia hominivorax) infestation. Amblyomma maculatum also can transmit Hepatazoon americaum, the agent of canine hepatazoonosis, and it is experimentally capable of transmitting E. ruminatum (formerly Cowdria ruminatum) – the agent of heartwater disease – and Leptospira pomona, the agent of leptospirosis in livestock. This tick was first described from the Gulf Coast states of Mississippi, Louisiana, Alabama, Georgia, Florida and South Carolina, with inland foci in Oklahoma, Kansas, Kentucky and Tennessee (Teel et al., 2010). More recent surveys have identified a wider distribution, including populations in Arkansas (Trout et al., 2010) and Virginia (Fornadel et al., 2011; Wright et al., 2011). It is uncertain whether the range of this tick is expanding or if more thorough surveys have detected previously unnoticed populations.
Amblyomma maculatum is reported to bite humans less frequently than does A. americanum (Felz et al., 1996; Merten and Durden, 2000; Jiang et al., 2011); however, female A. maculatum can be mistaken by tick bite victims for D. variabilis, especially if mouthparts are missing, so the attack rate and distribution of this tick may be underreported. Pheromone-mediated behaviour may play a role in reducing the frequency of human attack, as female A. maculatum are more strongly attracted to a host upon which males are feeding and releasing attraction–aggregation–attachment pheromones (Allan et al., 1991; Sleeba et al., 2010; Teel et al., 2010). In contrast to natural hosts such as deer, humans would typically not allow males to attach for long enough to attract such females.
The brown dog tick, R. sanguineus, is a peridomestic rather than wildland species, so humans need not venture into woods or grasslands to encounter it. The species is distributed globally in disjunct foci of infestation typically associated with pet or stray dogs (Estrada-Peña and Jongejan, 1999; Uspensky and Ioffe-Uspensky, 2002; Dantas-Torres, 2008, 2010). Unlike the other species discussed here, R. sanguineus can complete its life cycle indoors and will attack humans in their homes; indeed, data from the DOD Human Tick Test Kit Program suggest a predilection of this tick to feed on small children (Stromdahl et al., 2011).
Both the adult and immature stages of this tick will feed on humans and both are capable of transmitting rickettsial pathogens. Rickettsia rickettsii-infected R. sanguineus have been associated with RMSF cases from Arizona (Demma et al., 2005), California (Wikswo et al., 2008) and Georgia (Garrison et al., 2007). Rickettsia rickettsii prevalences in the brown dog ticks in these studies were higher than those reported for D. variabilis, the vector species traditionally implicated in RMSF transmission. Rhipicephalus sanguineus in southern California (Beeler et al., 2011) and in Arizona (Eremeeva et al., 2006) have also been found infected with R. massiliae, another SFG rickettsia that has recently emerged as an agent of human disease in France (Parola et al., 2008) and in Sicily (Vitale et al., 2006).
Given global warming trends, the public health burden of this tick may well increase in the south-east, as Parola et al. (2008) have demonstrated increased willingness of the immature life stages to bite humans in warm (>30°C) environments. Furthermore, at warmer temperatures, the generation time of this tick shortens, with up to four generations per year reported, which will increase the potential for population irruptions (Uspensky and Ioffe-Uspensky, 2002; Dantas-Torres, 2008). Once established in or around a home, infestations of R. sanguineus are problematic to exterminate because the tick harbours in cracks and crevices, and so is difficult to control with pesticide applications (Beeler et al., 2011).
Direct evidence of A. americanum increasing and D. variabilis decreasing is available from Panola Mt. Park, GA (latitude 33.6°N). Dermacentor variabilis was abundant in this park during the 1980s (Newhouse, 1983), with few A. americanum collected at that time. In contrast, when this same park was resurveyed two decades later, D. variabilis was rarely encountered, whereas A. americanum was collected in large numbers (Paddock and Telford, 2011).
Ixodes scapularis/Ixodes affinis
Cases of Lyme disease reported from Camp LeJeune, North Carolina (Armed Forces Health Surveillance Center, 2009), prompted an investigation of tick and Borrelia species in that area. Collection efforts revealed the presence of a second Borrelia-infected Ixodes species, I. affinis, which was morphologically so similar to I. scapularis that it had been misidentified as that species in the Camp LeJeune surveys (Harrison et al., 2010). Ixodes affinis is a Central and South American tick – previously described from Florida, Georgia and South Carolina – that appears to have invaded coastal North Carolina and more recently south-eastern Virginia (where again it was at first misidentified as I. scapularis; Nadolny et al., 2011). Intriguingly, testing of I. affinis from North Carolina has revealed these ticks to be heavily infected with B. burgdorferi (prevalences of 33–35%; Harrison et al., 2010; Maggi et al., 2010) and B. bissettii (27%; Maggi et al., 2010), whereas Borrelia spp. prevalence in sympatric populations of I. scapularis was extremely low or zero. Studies are underway to determine whether this pattern is a consequence of differences in host selection by these two Ixodes species.
Ixodes affinis rarely bites humans (none were collected from humans in Merten and Durden, 2000; M. Toliver, pers. comm.), so this tick is unlikely to directly increase Lyme disease risk. Ixodes affinis nevertheless could have an important indirect role in Lyme disease dynamics if it helps maintain high levels of B. burgdorferi in reservoir hosts that are later fed upon by I. scapularis.
Tick-borne Disease Aetiologies in the South-eastern United States
In contrast to the many detailed studies available on the causes and mechanisms of Lyme disease in northern states (see Tsao, 2009 for a recent review), the aetiologies of emerging tick-borne diseases in the south-east are incomplete, and in some instances, controversial. In addition, diagnostic testing for these emerging diseases, which is important for accurate diagnosis, treatment and reporting, suffers from the lack of specificity and poor compliance by heath providers and patients with respect to testing guidelines. Selected examples of problems arising from confusion about tick-borne disease aetiologies in the south-east are described below.
Overestimating the risk of Lyme disease in areas with abundant A. americanum: ‘Not every small tick is a “deer tick”’
In the south-east, humans are far more likely to be bitten by A. americanum than by I. scapularis (Fig. 1). For example, in 2005, of 982 ticks removed from soldiers at Ft. Pickett, VA and submitted to the DOD Human Tick Test Kit Program, 961 (98%) were A. americanum, 14 (1%) were D. variabilis and only 7 (<1%) were I. scapularis.
In the north-east, Lyme disease awareness campaigns have focused public attention on the I. scapularis nymph, as this is the infected life stage active in summer months when most disease transmission occurs (Falco et al., 1999). Although nymphal ticks are much smaller than adults, nymphal I. scapularis and A. americanum can be distinguished without magnification (Plate 1). Public awareness of the differences between the two ticks, or in some cases even of the existence of A. americanum, is lacking, so A. americanum nymphs are very frequently misidentified as I. scapularis by tick bite victims. In a study of a southern Maryland community with sympatric A. americanum and I. scapularis populations, and with biting tick submissions similar to those at Ft. Pickett (i.e. 95% of the ticks submitted for identification by the residents were A. americanum and only 3% were I. scapularis), 42% of tick bite victims were unable to distinguish these species and 34% believed that they had been bitten by I. scapularis (Armstrong et al., 2001).
Failure to differentiate among tick species has contributed to an exaggerated perception among health providers and the public of Lyme disease risk in south-eastern states, leading to over-diagnosis of Lyme disease in those states. In the Maryland study cited above, 20 of the 152 residents reporting a recent tick bite consulted a physician and 13 were prescribed antibiotics on the assumption that they were at risk of Lyme disease. Yet, of the 54 ticks submitted to the authors by those 20 patients, 98% were A. americanum (Armstrong et al., 2001).
Further findings indicating a trend towards extensive over-diagnosis of Lyme disease in the south-east comes from a comparison of data from Virginia in the DOD’s Standard Ambulatory Data Registry with the tick species submitted to the DOD Human Tick Test Kit Program from that state. From 2004 to 2008, 17 988 preliminary diagnoses of suspect Lyme disease were assigned to military cases from Virginia military treatment facilities, but only 14 (0.08%) cases of Lyme disease were subsequently confirmed (Sjoberg and Owens, 2009). During those same 4 years, 3737 A. americanum and only 114 I. scapularis were submitted by Virginia military personnel to the DOD Human Tick Test Kit Program.
The aetiology of STARI
Erythema migrans (EM), the bull’s-eye rash of Lyme disease, is an annular, macular, erythematous skin lesion characteristic of early infection with B. burgdorferi. EMs are commonly seen among patients from Lyme disease endemic regions of the United States (i.e. the north-east, upper Midwest and northern Pacific Coast). Since the 1980s, EM lesions and symptoms suggestive of Lyme disease also have been reported following bites of A. americanum from areas of the south-east where A. americanum dominates and where human-biting Ixodes species are at very low abundance (Barbour, 1996; Kirkland et al., 1997; Masters et al., 1998). This illness is termed Southern tick-associated rash illness (STARI) or Maters’ disease, in remembrance of Dr Edwin Masters, a Missouri family practitioner who reported the rash from hundreds of his patients (Masters et al., 2008).
Given the similarity of STARI symptoms to Lyme disease EMs, an infectious aetiology involving B. burgdorferi was at first suspected and investigations of the potential for B. burgdorferi transmission by A. americanum began. Low prevalences of borrelia, including occasional B. burgdorferi, were detected in A. americanum using polyclonal and monoclonal antibodies, species-specific PCR and culturing [studies from 1983 to 2002 summarized in Stromdahl et al. (2003)]. However, vector-competency studies repeatedly concluded that B. burgdorferi is unlikely to be transmitted by A. americanum (Piesman and Sinsky, 1988; Mather and Mather, 1990; Mukolwe et al., 1992; Ryder et al., 1992; Oliver et al., 1993a; Sanders and Oliver, 1995; Piesman and Happ, 1997) because of the potent borreliacidal agent present in A. americanum saliva (Ledin et al., 2005; Zeidner et al., 2009). No B. burgdorferi was detected during an extensive molecular characterization of the microbiome of the A. americanum midgut (Yuan, 2010), nor among 1621 A. americanum submitted in 2010 for testing by the DOD Human Tick Test Kit Program.
From the inception of the DOD Human Tick Test Kit Program, A. americanum have been tested for borreliae. At first, the target was B. burgdorferi, later, ‘B. lonestari’. Over time, lack of evidence of disease transmission reduced ‘B. lonestari’ to the role of a bystander and testing for it ended in 2009. In 2010, the DOD Human Tick Test Kit Program made a final intensive effort to screen A. americanum for borreliae, testing all 1,621 ticks submitted using a PCR that employed generic Borrelia flagellin gene primers (Barbour et al., 1996). This effort yielded 24 generic positive samples that were then tested further using specific PCRs for ‘B. lonestari’ (Bacon et al., 2004) and B. burgdorferi (Straubinger, 2000). A small number (9/1,621) were positive in the ‘B. lonestari’ PCR, but none were positive for B. burgdorferi.
Of the 15 tick samples positive in the generic Borrelia PCR, but negative in both the ‘B. lonestari’- and B. burgdorferi-specific PCRs, 10 were sent to Ibis Biosciences for further analysis using a multilocus PCR electrospray ionization mass spectrometry (PCR/ESI-MS) Borrelia identification and genotyping assay (Crowder et al., 2010). PCR/ESI-MS analysis determined that two samples were Borrelia-negative and four were positive for ‘B. lonestari’ (two of these also contained DNA from an additional relapsing-fever Borrelia). The remaining three samples, and one of the ‘B. lonestari’-positive samples, were positive for the B. burgdorferi flagellin primer but negative for seven other Borrelia primers. Attempts to clone and sequence the flagellin amplicon from these samples were unsuccessful. The PCR/ESI-MS assay targets the same region of the flagellin gene used in the initial screening (Barbour et al., 1996), so amplicon contamination from the positive control could have been responsible for these flagellin primer detections.
Consequently, at this time, the aetiology of STARI remains unknown. Possibilities include an unknown infectious agent or an inflammatory process. A recent study describes an unusual allergic reaction in patients who have experienced bites of A. americanum (Commins et al., 2011). The tick’s saliva apparently prompts the bite victim to produce IgE antibodies that can later be triggered by a similar oligosaccharide present following the consumption of red meat. A related allergic response, perhaps to another component of A. americanum saliva, is one possible explanation for the symptoms seen in STARI patients.
Areas where rash-associated tick species overlap
In recent decades, the range of A. americanum has expanded northwards along the eastern seaboard and population abundances are climbing in several north-eastern states (Paddock and Yabsley, 2007; and references therein). Northern cases of STARI can therefore be expected to increase in Lyme disease endemic areas. In aggregate, there are differences in the clinical presentation of groups of patients with STARI versus Lyme disease (Wormser et al., 2005b), but it can be impossible to distinguish the EM-like skin lesion of STARI from that of Lyme disease for a particular patient (Feder et al., 2011). This creates the potential for diagnostic confusion in states where both I. scapularis and A. americanum are becoming abundant. Useful diagnostic features of STARI may include a lack of plasma cells at biopsy and a lack of response of the EM-like lesion to oral antibiotic therapy (Feder et al., 2011). Nevertheless, the most reliable method of distinguishing STARI from B. burgdorferi infection is by documenting that the patient was bitten by an A. americanum rather than an Ixodes tick.
Three species of Ehrlichia– transmitted by A. americanum, E. chaffeensis, E. ewingii and PME – are known to cause human disease (Table 1). These pathogens were previously considered to be of only veterinary importance (or in the case of PME were unknown). Recognition of their zoonotic potential is very recent, coming in 1986 for E. chaffeensis (Maeda et al., 1987), 1996 for E. ewingii (Buller et al., 1999) and 2005 for PME (Reeves et al., 2008). Knowledge and awareness of these potentially life-threatening diseases are consequently low among some health providers and the public (Morgan, 2010). Failure to correctly diagnose and treat ehrlichiosis (which with prompt antibiotic therapy normally resolves) can result in deaths of healthy young individuals (e.g. Martin et al., 1999; Rooney et al., 2001).
In 1993, the largest outbreak of ehrlichiosis to that date occurred in a recently developed retirement community in a forested region of Tennessee’s Cumberland Plateau. A subsequent CDC investigation (Standaert et al., 1995) identified tick bites, golfing, exposure to wildlife and failure to apply insect repellent as risk factors for infection. Despite subsequent efforts by the community’s managers to mitigate the problem, residents continue to self-report tick-associated health problems and a recent survey obtained prevalence estimates of 1%E. chaffeensis, 6%E. ewingii and 2% PME in 253 questing adult A. americanum from the community (Harmon, 2010). Comparison of these data with other surveys for R. rickettsii (Moncayo et al., 2010) and B. burgdorferi (Rosen, 2009) indicates that Ehrlichia species are the most prevalent zoonotic pathogens in Tennessee ticks.
Consistent with Harmon (2010), Paddock and Yabsley (2007) cite several studies indicating that E. ewingii occurs in reservoir and vector populations in the south-east at frequencies similar to or greater than infection with E. chaffeensis. Confirmed cases of human disease caused by E. ewingii are uncommon, however, and tend to be associated with previously immune-suppressed patients. Paddock et al. 2005 (cited in Paddock and Yabsley, 2007) have suggested that E. ewingii causes milder illness than E. chaffeensis, particularly in persons without pre-existing immune suppression, and so fewer E. ewingii-infected patients may seek medical attention.
Human infections with these three zoonotic ehrlichieae are officially reportable diseases (as ‘E. chaffeensis infection’, ‘E. ewingii infection’ or ‘Ehrlichiosis, undetermined’ in the case of PME; http://www.cdc.gov/osels/ph_surveillance/nndss/phs/infdis2011.htm). Infections with these three ehrlichieae can induce antibodies that may cross-react and confuse serology, and E. ewingii and A. phagocytophilum morulae are indistinguishable visually. PCR is the only specific diagnostic test available for E. ewingii infection (Thomas et al., 2009); unless a broad-range PCR is used, these infections will not be detected.
Novel rickettsiae of unknown pathogenicity
Until the early 1980s, RMSF was the most commonly recognized tick-borne disease in the United States As discussed previously, many of these past ‘RMSF’ diagnoses may have resulted from serological cross-reaction with rickettsiae other than R. rickettsii, particularly ‘R. amblyommii’ and R. montanensis found in A. americanum. Screening of A. maculatum submitted to the DOD Human Tick Test Kit Program has identified several additional unusual rickettsiae of undetermined pathogenicity, including ‘R. andeanae’ and R. felis (Jiang et al., 2011). Furthermore, rickettsial outer-member protein B gene (ompB) sequences obtained from three I. scapularis nymphs in North Carolina (Smith et al., 2010) were similar to sequences reported for R. massiliae, a suspected human pathogen in Italy (Vitale et al., 2006). Their recent detection in North Carolina ixodid tick populations leaves open the possibility that undiagnosed febrile illnesses associated with tick bites could be due to newly emerging rickettsial zoonoses. More evidence, such as sequencing of additional gene targets, is needed to determine whether these ‘novel’ rickettsiae are indeed pathogenic.
Human babesiosis in the south-east?
Babesia microti is the infectious agent most frequently transmitted by blood transfusion in the United States. Transmission by transfusion is more likely than by tick bite – transfusion-transmitted cases have been reported from south-eastern states (Herwaldt et al., 2011; Leiby, 2011). There is no evidence to date, however, to suggest that the south-east is endemic for either Ba. microti or Ba. duncani, another species associated with rare cases of human disease in western states (Conrad et al., 2006). Despite reports from a variety of commercial and private laboratories of positive serologies in humans from south-eastern states for Ba. microti and Ba. duncani, these organisms have never been detected in ticks collected in the south-east, nor has a vector cycle been described (Prince et al., 2010). The DOD Human Tick Test Kit Program tests all I. scapularis received using a PCR targeting Ba. microti (Tonnetti et al., 2009); the only PCR-positive ticks identified to date have been from north-eastern and upper mid-western states. Therefore, considerable caution must be exercised in interpreting Babesia-positive laboratory results in the south-east in the absence of travel/transfusion history. Public health investigators nevertheless continue to be concerned and open to the possibility of locally acquired babesiosis, and as a result, an autochthonous babesiosis human case was recently identified in Tennessee in an individual infected with the ‘MO-1’ strain of Babesia, a strain more closely related to the European Babesia divergens strain than to Ba. microti or Ba. duncani (Moncayo, 2010).
Between 1969 and 2002, six tick-borne diseases – babesiosis, Lyme disease, E. chaffeensis infection (formerly HME), E. ewingii infection, A. phagocytophilum infection (formerly HGE) and R. parkeri infection – were identified and characterized. These discoveries doubled the number of tick-borne diseases known in North America (Paddock and Yabsley, 2007). STARI, with unknown aetiology, was also identified during this period.
Concurrently, the populations of ticks that vector these diseases were in flux. Ixodes scapularis was expanding in distribution and abundance out of endemic foci in the upper Midwest and north-east (Steere et al., 2004), while A. americanum was expanding northwards out of the south-east. Irrupting deer populations, and associated habitat changes, helped drive this flux (Barbour and Fish, 1993; Paddock and Yabsley, 2007). For A. americanum, expanding wild turkey populations may also have played a role (Paddock and Yabsley, 2007).
Given changing risk of tick-borne disease in the south-east, and the close association of specific pathogens with specific tick species, it is problematic that many healthcare providers and members of the public have little awareness that there are multiple tick species, and that the species identity of an attacking tick is an important differential for correct diagnosis of a subsequent infection. There is also lack of awareness of several of the tick-borne diseases described above. For example, ehrlichieae are the most common zoonotic pathogens found in Tennessee ticks (Harmon, 2010); yet, informal classroom surveys indicate that few wildlife-major undergraduates have heard of ehrlichiosis or can name more than one tick-borne disease (GJH, pers. obs.). Perhaps, more concerning, in a survey of 36 East Tennessee healthcare providers, 62% of those responding characterized their knowledge of ehrlichiosis as ‘none or weak’, 20% reported having no knowledge at all of ehrlichiosis, two-thirds did not consider the possibility of tick-borne disease test cross-reactivity to the agents of unknown pathogenicity and one-third did not consider identifying the species of attacking tick to be useful in disease diagnosis (Morgan, 2010).
Physicians in the south-east could benefit from readily accessible, up-to-date information on the availability and limitations of diagnostic tests for human exposure to the tick-borne pathogens we review here. Technology for such human testing is evolving rapidly, so we avoid making specific testing recommendations here, as these could soon be obsolete. Federal, state and local public health agencies offer information on tick-borne diseases, and state or local agencies sometimes provide tick identification and testing services to residents. Useful resources on tick-borne disease diagnostics are available online, including a manual for physicians prepared by the Massachusetts Department of Public Health (http://www.maclearinghouse.com/PDFs/BID/TM3901.pdf; accessed 28 November 2011) and a review article on tick-borne rickettsial disease by Chapman et al. (2006). Information on test availability, selection and interpretation is also provided by some of the large commercial testing laboratories (e.g. http://www.questdiagnostics.com/hcp/intguide/jsp/showintguidepage.jsp?fn=CF_Tick-borneDis.htm; accessed 28 November 2011).
Recent treatment recommendations have begun to emphasize the importance of considering the tick species and its infection status as part of the diagnostic process (e.g. Hojgaard et al., 2008; Feder et al., 2011). Even so, there are clinical presentations after A. americanum tick bites that have yet to be associated with specific aetiological agents, and there are A. americanum-borne organisms whose pathogenicities are yet unknown. Consequently, healthcare providers and researchers need to remain open to the possibility of other as-yet-unidentified disease agents and cross-reactive agents, and there is a need for more thorough characterization of the pathogens involved in unusual or unexpected disease cases. IBIS technology (Crowder et al., 2010) and other novel techniques show great promise on this front.
Our aim with this article has been to highlight the importance of considering vector and pathogen distribution and abundance when developing public health measures to reduce the risk of tick-borne disease. A consequence of the intensive media coverage afforded to Lyme disease in recent decades has been that for many Americans the combination of ‘tick’ plus ‘rash’ means only one thing: Lyme disease. As one example, in their study in coastal Maryland, Armstrong et al. (2001) noted that EM was mentioned in 65% of self-reports of Lyme disease; however, many individuals indicated that the rash was present while the tick was still attached, which is strongly the suggestive of an allergic reaction to the bite itself rather than true Lyme disease.
Members of the public should be encouraged to save any tick that they discover attached to themselves or their family in the freezer or in a small vial of alcohol. If no signs of disease develop the tick can safely be discarded after several weeks. Otherwise, if symptoms do develop, the tick species can be identified to assist in correct diagnosis of the disease and, if a disease situation warrants greater effort, the tick can be directly tested for the presence of tick-borne pathogens.
One avenue for improvement would be for agencies and other organizations providing information on tick-borne disease to put greater effort into drawing their audience’s attention to which ticks are most likely to bite humans in their local community, together with simple information (e.g. Plate 1) that would assist the public in determining which tick species and life stage have attacked them. The ‘tick identification’ web page of the University of Rhode Island’s Tick Encounter Resource Center (http://www.tickencounter.org/tick_identification; accessed 7 October 2011) is one example of an online resource that adjusts the information presented in response to the reader indicating in which region of the United States they are located. More such efforts to ensure appropriate regionalization of information are needed. Outreach material needs to be based on sound biological information; for example, this article has highlighted the value of the state-by-state information provided by the DOD Human Tick Test Kit Program. We encourage participation in the Program by an increased number of military medical care providers.
Given the potential for diagnostic confusion between the symptoms of Lyme disease and STARI, improving the knowledge of health providers regarding tick-borne disease vectors and pathogens is perhaps most acute in states such as Virginia and Maryland, where sizeable populations of I. scapularis and A. americanum now co-occur. Further south, healthcare providers have greater awareness of the risk to their patients from RMSF than from ehrlichiosis, despite the agent for the latter being far more prevalent in local ticks. This may reflect RMSF having been a long-standing problem, whereas ehrlichiosis is a recently emerging risk that older health providers would not have encountered during their formal training.
The eco-epidemiological factors that have led to the recent doubling in the number of recognized tick-borne diseases are not static, and the distribution of ticks and their pathogens remains in flux in the eastern United States. In coming years, we anticipate increased incidence of most of the diseases discussed here, and the probable emergence of additional tick-borne diseases as yet unknown. Proactive public health training and outreach are needed to address this changing risk of tick-borne disease across the south-eastern United States.
We thank Drs Mark Eshoo and Chris Crowder of Ibis Biosciences, Carlsbad, California, and Dr Marcee Toliver, for allowing us to cite their unpublished data. Dr Jean Tsao, University of Michigan, Dr Jennifer McQuiston, CDC-Atlanta, Dr Barbara J. Johnson and Ms Anna Perea, CDC-Ft. Collins and Dr Anne Kjemtrup, California Department of Public Health, provided insightful comments on drafts of the manuscript. Dr Joe Piesman and Ms. Gabrielle Dietrich, CDC-Ft. Collins, kindly supplied us with I. scapularis nymphs for high-resolution imaging. The views expressed in this article are those of the authors and do not reflect the official policy or position of the Department of the Army, Department of Defense, The National Science Foundation, or the US Government. GJH was supported in part by National Science Foundation EEID Grant No. 0914397 and by NIMBioS, the National Institute for Mathematical and Biological Synthesis.