Southern tick-associated rash illness (STARI)

INTRODUCTION — During the 1980s, clinicians in Missouri and the southeastern United States described a new illness characterized by the presence of a rash typical of erythema migrans (EM) and mild flu-like symptoms that was temporally associated with a bite by the Lone Star tick (Amblyomma americanum) (picture 1). EM-like skin lesions appeared at the site of these tick bites, but serologic tests in such patients consistently failed to show evidence of infection with Borrelia burgdorferi, a finding that fits with the fact that the vector for Lyme disease (Ixodes scapularis) (picture 2) was unknown or uncommon in these locations [1,2]. Experimental studies subsequently revealed that A. americanum is an incompetent vector for B. burgdorferi sensu stricto, the cause of Lyme disease in the United States [3]. (See "Epidemiology of Lyme disease".)

Over the next 20 years, this illness was given a variety of names, including Southern Lyme disease and Masters disease. The most widely used name, Southern tick-associated rash illness (STARI), may be misleading since this syndrome and its presumed vector are now known to occur in other regions such as the Midwest and the Mid-Atlantic states, as well as several New England states [4].

At present, STARI should be considered to be a syndrome of unproven cause that is clinically diagnosed on the basis of its characteristic skin rash, mild clinical course, and occurrence in areas where Lyme disease is either uncommon or not present, but where A. americanum ticks are known to be present. Until there is a definitive diagnostic test for this illness and until the causative organism has been definitively isolated from humans, much of our understanding about epidemiology, clinical features, and treatment will remain speculative.

CAUSATIVE ORGANISM — STARI is presumed to be transmitted to humans by the Lone Star tick, but available data do not conclusively support any single agent as the etiology of STARI.

Early studies using polymerase chain reaction (PCR) testing led to a preliminary hypothesis that STARI is caused by a spirochete tentatively named B. lonestari. Initial descriptions of this organism were based on PCR amplification of the flagellin B gene and 16S rRNA from ticks collected from patients with STARI or from ticks that were collected in regions where STARI was recognized [3,5-8]. (See 'Epidemiology' below.)

B. lonestari has also been grown in vitro from A. americanum ticks collected in Georgia [9]. Light and electron microscopic studies of this single isolate revealed that the organism is a spirochete that is morphologically similar but not identical to B. burgdorferi sensu lato (picture 3). In addition, PCR amplification of DNA from this tick isolate had an essentially identical sequence to amplification products of 16S rRNA made from Lone Star ticks collected in different geographic locations and from blood samples from white-tailed deer killed in several southeastern states [3,5,8-10].

However, there is only one published report of finding PCR products typical of B. lonestari in a sample taken from a human. In 2001, a patient with typical features of STARI who had an embedded A. americanum tick in the center of an erythema migrans (EM)-like skin lesion was identified [3]. Simultaneous PCR testing of a skin biopsy sample and the imbedded tick in two geographically separate laboratories yielded an identical amplification product. This PCR product had an identical sequence to PCR amplification products of the flagellin B gene and 16S rRNA recovered from Lone Star ticks by other investigators in different locations where STARI was recognized. Attempts to culture Borrelia from a portion of the skin biopsy were unsuccessful, and serial testing of the patient's serum failed to show evidence of infection with B. burgdorferi. Other investigators have not been able to duplicate this finding [11].

Taken together, the mixed results reported in the literature suggest there may be multiple potential triggers for this primarily clinical syndrome. (See 'Clinical features' below.)

EPIDEMIOLOGY — Human illness with typical clinical features of STARI have been described in patients from Mississippi, Missouri, Maryland, Georgia, South Carolina, North Carolina, and New Jersey [4,6,12-16]; in addition, a case was described in a patient who was bitten by a tick on Long Island [17]. (See 'Clinical features' below.)

Spirochetes resembling Borrelia spp have been visualized in Lone Star ticks collected in several geographic areas, including Alabama, Missouri, New Jersey, North Carolina, Nebraska, Oklahoma, and Texas [10,18,19]. Molecular evidence that such spirochetes are B. lonestari has been reported by investigators studying A. americanum ticks. (See 'Causative organism' above.)

White-tailed deer may be one of the vertebrate reservoir hosts for B. lonestari. All three stages of A. americanum feed on these animals, and inoculation of white-tailed deer with B. lonestari has been shown to result in spirochetemia, as assessed with blood smears. In one report, these deer showed evidence of infection by polymerase chain reaction (PCR) within four days of inoculation, and subsequent PCR tests remained positive for up to 28 days [20]. In another study, 7 of 80 samples obtained from deer killed in 17 separate sites in Arkansas, Georgia, and North and South Carolina had evidence of B. lonestari infection using PCR amplification for the Borrelia flagellin gene. Furthermore, the flagellin gene sequence amplified from deer blood was identical to the DNA sequences previously amplified from ticks collected in Alabama, Missouri, Tennessee, and Texas and from the single human patient with STARI who had PCR products typical of B. lonestari [3,8].

If subsequent studies confirm that B. lonestari is a cause of clinical STARI and white-tailed deer are the intermediate host, or that salivary antigens of A. americanum can cause a similar syndrome, it is likely that human disease will be found to occur in other areas where both the tick and reservoir host are common.

CLINICAL FEATURES — The diagnosis of STARI is currently based upon the presence of an erythema migrans (EM)-like rash following a history of tick bite that occurs in areas where the vector of Lyme disease is uncommon or unknown [21].

The clinical spectrum of STARI was illustrated in a report of 14 patients seen over a two-year period in central North Carolina [12]. All of the patients had a recent history of EM-like skin lesions after removing attached ticks, and eight remembered removing a tick from the exact site where the rash developed. The median time from tick detection to onset of rash was 12 days with a range of 2 to 21 days. In every case in which the tick was available for subsequent study, it was identified as A. americanum. The illness was mild in all cases; only one had fever and all recovered without sequelae.

Skin lesions in patients with STARI may be single or multiple. The lesions are characteristically erythematous with irregular borders and central clearing (picture 4). Some patients have a papule in the center of their EM-like lesions; many recall removing a tick at the exact site of this papule. The skin lesions may expand centrifugally over several days. At the time of diagnosis, the diameter of such lesions has ranged from 3 to 10 cm.

Patients with STARI often complain of headache, generalized aching, fatigue, and/or nausea, but these symptoms as well as the rash disappear following treatment with tetracyclines. Thus far, no long-term complications or evidence of recurrence following treatment has been recognized.

In a prospective study, the clinical and epidemiologic features of 21 patients from Missouri with EM-like skin lesions were compared to 97 patients from New York with EM-like lesions [22]. The two groups of patients had substantial and statistically significant clinical differences. Compared to patients from New York, patients from Missouri with EM-like skin lesions were:

●More likely to recall a recent tick bite (86 versus 20 percent)

●Less likely to be symptomatic (19 versus 76 percent)

●Less likely to have multiple skin lesions (5 versus 27 percent)

●More likely to have smaller lesions (mean size 8.3 cm versus 16.4 cm)

●More likely to have lesions with central clearing (76 versus 22 percent)

●More likely to recover rapidly after antibiotic treatment (at three-month follow-up visit 0 versus 21 percent with persistent symptoms)

The authors of this report concluded that these differences were further evidence that STARI is a distinct and different illness than Lyme disease.

LABORATORY FINDINGS — Although enzyme-linked immunosorbent assay (ELISA) testing is equivocal or weakly positive for B. burgdorferi antibodies in some patients with STARI, Western blot patterns show profiles that are inconsistent with acute B. burgdorferi infection. In addition, patients with STARI do not develop antibodies to rickettsial or ehrlichial antigens. Thrombocytopenia and leukopenia do not typically occur and routine laboratory tests of renal and hepatic function are usually normal.

PATHOLOGY — The histologic findings in skin biopsies from patients with STARI are similar to the histologic findings from biopsies of erythema migrans (EM) lesions of patients with Lyme disease. Superficial and deep infiltrates with lymphocytes are present in both conditions. However, the skin lesions of patients with STARI lack plasma cells whereas plasma cells are often present in EM lesions of patients with Lyme disease. In addition, silver stains for the presence of spirochetes are characteristically negative in biopsies of skin lesions in patients with STARI [15].

DIAGNOSIS — At present, there is no serologic test available to detect STARI or exposure to Amblyomma ticks. The diagnosis remains a clinical one, based on compatible symptoms and physical examination findings in a person with a history of a tick bite from a known endemic area of the Lone Star tick. (See 'Epidemiology' above and 'Clinical features' above.)

Novel diagnostics for STARI based on host serum metabolomics are in development [23], but until there is a reliable laboratory method to diagnose STARI, it is likely that some patients with erythema migrans (EM)-like skin lesions due to STARI will be misdiagnosed as having Lyme disease if they reside in areas where the vectors of both illnesses coexist. This phenomenon may be partly responsible for the increase in the reported incidence of Lyme disease in North Carolina and other states where the vector of B. burgdorferi is uncommon [24]. However, the frequency that such diagnostic confusion occurs has not yet been studied, nor has the potential clinical consequences of unnecessary testing and treatment for Lyme disease in patients with STARI.

TREATMENT — For patients with EM-like lesions and a potential tick exposure, we suggest antimicrobial therapy. Although the benefit of treatment for STARI has not been established, it is important that treatment of early Lyme disease not be delayed. However, some clinicians and patients may prefer close observation rather than treatment if they are confident about the absence of an exposure to Ixodes ticks [25].

Our preferred treatment is doxycycline 100 mg twice daily for 10 days. Patients with presumed STARI described in case series and case reports appear to have responded promptly to therapy with doxycycline [22,26]. For patients with a contraindication to using doxycycline, amoxicillin 500 mg orally three times a day or cefuroxime 500 mg orally twice daily for 14 days are reasonable alternatives since the organism causing STARI may be a spirochete of the Borrelia family, and the primary goal is often to avoid undertreatment of early Lyme disease [15]. (See "Treatment of Lyme disease", section on 'Early localized disease (single erythema migrans)'.)

Follow-up serological testing for Lyme disease is not necessary in such patients since the preceding regimens are highly effective for both STARI and early Lyme disease and because early treatment of such patients may prevent antibody development, even in patients with B. burgdorferi infection. (See "Diagnosis of Lyme disease".)

SUMMARY AND RECOMMENDATIONS

●History – During the 1980s, clinicians in Missouri and the Southeastern United States described a new illness characterized by the presence of a rash typical of erythema migrans (EM) and mild flu-like symptoms that was temporally associated with a bite by the Lone Star tick (Amblyomma americanum). This infection was described as the Southern tick-associated rash illness or "STARI." (See 'Introduction' above.)

●Etiology – While STARI is widely presumed to be transmitted to humans by the Lone Star tick, definitive evidence for an etiologic agent of this syndrome remains elusive. (See 'Introduction' above and 'Causative organism' above.)

●Epidemiology – Most cases of human illness with typical clinical features of STARI have been described in patients from Mississippi, Missouri, Maryland, Georgia, South Carolina, and North Carolina. All three stages of the A. americanum tick feed on white-tailed deer, which may serve as the reservoir/trigger for this infection. (See 'Epidemiology' above.)

●Clinical manifestations – STARI is characterized as a mild illness with onset of typical EM-like skin lesions, which can be single or multiple. Patients often complain of headache, generalized malaise, and nausea. (See 'Clinical features' above.)

●Diagnosis – The diagnosis of STARI is currently based upon the presence of an EM-like rash following a history of tick bite that occurs in areas where the vector of Lyme disease is uncommon or unknown. At present, there is no diagnostic test available to detect STARI. (See 'Diagnosis' above.)

●Management – For patients with an EM-like lesion and a potential tick exposure, we suggest antimicrobial treatment (Grade 2C). Although the benefit of treatment for STARI has not been established, it is important that treatment of early Lyme disease not be delayed, and it is often difficult to distinguish STARI from early Lyme. We prefer treatment with doxycycline (100 mg twice daily for 10 days); however, in patients with a contraindication to using doxycycline, amoxicillin (500 mg three times daily) or cefuroxime (500 mg orally twice daily) for 14 days are reasonable alternatives. (See 'Treatment' above.)