Characteristics of Hard Tick Relapsing Fever Caused by Borrelia miyamotoi, United States, 2013–2019

Borrelia miyamotoi, transmitted by Ixodes spp. ticks, was recognized as an agent of hard tick relapsing fever in the United States in 2013. Nine state health departments in the Northeast and Midwest have conducted public health surveillance for this emerging condition by using a shared, working surveillance case definition. During 2013–2019, a total of 300 cases were identified through surveillance; 166 (55%) were classified as confirmed and 134 (45%) as possible. Median age of case-patients was 52 years (range 1–86 years); 52% were male. Most cases (70%) occurred during June–September, with a peak in August. Fever and headache were common symptoms; 28% of case-patients reported recurring fevers, 55% had arthralgia, and 16% had a rash. Thirteen percent of patients were hospitalized, and no deaths were reported. Ongoing surveillance will improve understanding of the incidence and clinical severity of this emerging disease.

Phylogenetically, B. miyamotoi is a relapsing fever group Borrelia (18).Diseases caused by this diverse group of spirochetes are differentiated by their vector, such as louseborne relapsing fever, transmitted by body lice, and tickborne relapsing fever or soft tick relapsing fever, transmitted by soft-bodied (argasid) ticks in several areas, including the western United States (19).B. miyamotoi is an agent of hard tick relapsing fever (HTRF), although resulting illness has also been referred to as B. miyamotoi disease.In the United States, B. miyamotoi is transmitted by I. scapularis ticks in the Northeast and Midwest (20,21) and by I. pacificus ticks on the Pacific Coast (22).Those tick species also transmit the causative agents of Lyme disease (23), anaplasmosis (24), babesiosis (25), Powassan virus disease (26), and a form of ehrlichiosis (27).Data from tick testing indicate that the geographic range of B. miyamotoi is similar to that of those pathogens (28,29).
The incidence of HTRF caused by B. miyamotoi and its public health role are largely unknown.In the United States, prevalence of B. miyamotoi in Ixodes spp.ticks is relatively low, but consistent across geographic regions at ≈2% (3,29).A seroprevalence evaluation conducted in several states in the northeastern United States in 2018 suggested that 2.8% of persons might have evidence of previous infection, compared with 11% of persons who had evidence of previous Lyme disease (30).Data from previous case series suggest that HTRF caused by B. miyamotoi most often manifests as a nonspecific febrile illness.Among identified cases, fever, myalgia, arthralgia, and headache are common, but recurring fevers similar to those documented in patients who have soft tick relapsing fever are relatively uncommon (4%-11% of total) (10,17).Immunocompromised persons who have HTRF might have more severe symptoms, including meningoencephalitis (11,16,31).
Specific laboratory diagnosis of B. miyamotoi infection is achieved through PCR detection of B. miyamotoi DNA (10,17,32).Serologic reactivity to surface proteins, especially glycerophosphodiester phosphodiesterase (GlpQ), is also used, but reactivity is not specific to B. miyamotoi infection or HTRF (33,34).GlpQ is found in all relapsing fever group borreliae but not in the B. burgdorferi sensu lato species that cause Lyme disease (34).However, GlpQ cannot distinguish between B. miyamotoi infection and infections caused by other relapsing fever group Borrelia spp., including agents of soft tick relapsing fever.In addition, related GlpQ proteins are found in common bacterial pathogens, such as Haemophilus influenzae and Escherichia coli, further reducing specificity of those serologic assays (34).
After initial cases of HTRF were identified in the United States, several states that had a high incidence of Lyme disease and other Ixodes-transmitted illnesses initiated public health surveillance to clarify Borrelia miyamotoi, transmitted by Ixodes spp.ticks, was recognized as an agent of hard tick relapsing fever in the United States in 2013.Nine state health departments in the Northeast and Midwest have conducted public health surveillance for this emerging condition by using a shared, working surveillance case definition.During 2013-2019, a total of 300 cases were identified through surveillance; 166 (55%) were classified as confirmed and 134 (45%) as possible.Median age of case-patients was 52 years (range 1-86 years); 52% were male.Most cases (70%) occurred during June-September, with a peak in August.Fever and headache were common symptoms; 28% of case-patients reported recurring fevers, 55% had arthralgia, and 16% had a rash.Thirteen percent of patients were hospitalized, and no deaths were reported.Ongoing surveillance will improve understanding of the incidence and clinical severity of this emerging disease.
the epidemiology of this novel tickborne condition.We summarize available information on HTRF as ascertained through public health surveillance efforts in the United States beginning in 2013.

Case Definition
The Centers for Disease Control and Prevention (CDC) and state health departments in areas that had a high incidence of Lyme disease jointly created an informal working surveillance case definition to identify and classify potential cases of HTRF caused by B. miyamotoi in a standardized manner.Clinical manifestations considered compatible with HTRF were broadly defined as acute onset of fever or chills with >1 of the following additional signs or symptoms: headache, sweats/chills, myalgia, arthralgia, malaise/fatigue, rash, abdominal cramps, nausea, vomiting, diarrhea, dizziness, confusion/altered mental status, photophobia, leukopenia, thrombocytopenia, or increased aminotransferase levels.
For purposes of surveillance data as summarized, we defined a confirmed case of HTRF caused by B. miyamotoi as compatible clinical manifestations with >1 of the following: isolation of B. miyamotoi from a clinical specimen; detection of B. miyamotoi DNA in a clinical specimen by using nucleic acid amplification techniques (NAAT) such as PCR; or evidence of seroconversion between acute phase and convalescent phase serum samples, including but not limited to a >4-fold change in serum antibody titer to B. miyamotoi between paired specimens.We defined a possible case as compatible clinical manifestations with >1 of the following: direct observation of spirochetes consistent with B. miyamotoi on a peripheral blood smear or detectable IgM or IgG to B. miyamotoi from a serum specimen.
We characterized cases as possible rather than probable to reflect the uncertainty of the spectrum of clinical manifestations of HTRF and the specificity of a single positive serologic titer.We excluded positive laboratory test results for which no clinical information was obtained or for which there was no associated clinical illness.Data on specific test manufacturers or antigenic targets used in serologic assays were not available to ascertain level of specificity for B. miyamotoi versus other relapsing fever Borrelia spp.

Public Health Investigation
Commercial or clinical laboratories reported positive laboratory results for B. miyamotoi in accordance with local regulations in states in which B. miyamotoi infection/HTRF was a reportable condition.Public health personnel conducted case investigations according to local practices to ascertain demographic, clinical, and exposure information to the extent possible through patient or provider interviews or medical chart reviews.

Analytic Methods
We classified symptoms as present or absent.We categorized age as <18 years, 18-64 years, and >65 years.We compared categorical and binary variables by using χ 2 or Fisher exact tests and continuous variables by using the Wilcoxon rank-sum test.We performed all statistical analyses by using SAS software (SAS Institute).This study was deemed to be a nonresearch activity by CDC under provision of public health surveillance.

Results
A total of 300 HTRF cases caused by B. miyamotoi were identified during 2013-2019 by the 9 state health departments in the Northeast and upper Midwest United States that conducted public health surveillance for this condition (Connecticut, Maine, Massachusetts, Minnesota, New Hampshire, New Jersey, Rhode Island, Vermont, and Wisconsin).The number of states in which HTRF was reportable increased from 1 in 2013 to 9 by 2019 (Figure 1).The number of cases identified annually concomitantly increased; more cases were identified during 2017-2019 (median 82, range 78-83 cases/y) than during 2013-2015 (median 9, range 8-10 cases/y) (Figure 2).
The median duration of time from symptom onset to seeking medical attention was 5 (range 0-311) days for the 69 persons for whom this information was available (Table 2).Persons who had possible illness had a longer duration from symptom onset to medical attention (median 9, interquartile range [IQR] 3-29 days) than persons who had confirmed cases (median 3, IQR 2-7 days; p = 0.03).Overall, the most common symptoms were fever (89%), fatigue (75%), headache (72%), and chills (68%).Among 64 patients  who had fever for whom a temperature was available, the median recorded temperature was 102.5°F (range 99.7°F -105.7°F).A total of 28% reported recurring fevers of some kind, 55% had arthralgia, and 16% had a rash.A description of the rash was available for 18 patients; the rash was noted to be generalized for 3 patients (1 confirmed and 2 possible cases) and focal for 15 patients (9 confirmed and 6 possible cases).An erythema migrans-like rash was reported for 1 possible HTRF case.Thrombocytopenia (51/105, 49%), increased levels of aspartate and alanine aminotransferases (40/96, 42%), and leukopenia (39/105, 37%) were common laboratory abnormalities.
All laboratory tests performed were either by using PCR or serologic analysis for B. miyamotoi infection.PCR was performed for 167 (56%) patients, and serologic analysis was performed for 137 (46%) patients (Table 3).Serologic analysis for paired serum samples was specifically performed for 19 (14%) persons.Among persons who had confirmed illness, 162/164 (99%) had a positive PCR result.Five (3%) of those persons also had a single positive IgM or IgG serologic test result, of whom 2 had a positive IgM result and none had a positive IgG or IgM/IgG combined test result.Among persons who had possible illness, 5/134 (4%) also had PCR performed; all results were negative.No microscopy or culture results were reported for any case.

Discussion
A total of 300 cases of HTRF caused by B. miyamotoi in the United States were identified by using public health surveillance.Case-patients were most commonly male and older adults.Case investigations showed that nonspecific symptoms, including fever and headache, were common, and rash was relatively uncommon.Those clinical features are similar to those of previous large case series of HTRF (10,17), although the overall proportion of cases with recurring fevers in this report was higher, and recurring fevers were more common among possible cases than among confirmed cases.The percentage of cases associated with hospitalization in this report was lower than reported among a large case series in the northeastern United States (17), and there were no reported deaths.However, the data in this summary might include cases reflected in previous case series if those previously reported cases were captured through public health surveillance.We observed a summertime seasonal pattern for confirmed cases, similar to findings for other infections transmitted by Ixodes spp.ticks in the United States (23,35,36).The frequency of HTRF peaked later in the summer than that for Lyme disease (3,23,37).This shifted seasonality supports a role for larval blacklegged ticks in B. miyamotoi transmission to humans in the United States because those ticks are more likely to be questing for blood meals during mid-to-late summer than during other tick life stages (38), and there is documented transovarial transmission of B. miyamotoi (39).Several features differed between confirmed and possible cases.Confirmed cases occurred more commonly among older persons and among male persons than did possible cases.Confirmed cases were more frequently characterized by fever, thrombocytopenia, and increased levels of aminotransferases.Possible cases were more frequently characterized by confusion, mood disorder, abdominal pain, shortness of breath, and recurring fevers.Although possible case-patients tended to have a longer duration of illness before seeking medical care, this difference might simply reflect the bias of the case definition itself, in which direct detection by  using PCR was laboratory evidence for the confirmed case definition and a single positive serologic result was laboratory evidence for the possible case definition.Serologic analysis (particularly for IgG) is unlikely to show increased levels during acute illness (34).In addition, increased reactivity against GlpQ alone might not be a specific measure of past HTRF infection (33,34).In this study, all possible case-patients had detectable IgG, suggesting that they might have had illness onset >20 days before testing, and so results might not represent acute B. miyamotoi infection or possibly not B. miyamotoi infection at all.However, longer duration of illness before seeking medical care for the possible case-patients might enable increased opportunity for observation of recurrent fevers.We observed less striking seasonality of illness onset for possible case-patients than for confirmed case-patients.Those findings collectively decrease our confidence that the possible cases summarized here reflect acute HTRF illness.The older age distribution among confirmed HTRF cases is similar to that of anaplasmosis and babesiosis cases and differs from Lyme disease cases, even though all are transmitted in the United States by the same species of Ixodes ticks (24).Lyme disease most commonly affects children 5-14 years of age, as well as older adults (40).In contrast, anaplasmosis rarely affects children, and HTRF in children was uncommon in public health surveillance.The older age distribution of anaplasmosis is believed to reflect host susceptibility differences and immune-related factors linked to aging, rather than age-related differences in tick exposure; those potential age-based susceptibility differences might account for the older age distribution associated with persons who have HTRF.However, diverging clinical or diagnostic approaches might be used for children versus adults, such as lower levels of testing or lower clinical awareness that bias the cases identified through public health surveillance toward adults.Because the clinical manifestations of HTRF and anaplasmosis might be similar (24), increased clinical education should highlight the potential for anaplasmosis and HTRF to resemble one another.
Fifteen persons in this study reported receiving a blood transfusion or organ transplant in the 30 days before symptom onset.Although no cases of B. miyamotoi infection after blood transfusion have been documented, other tickborne pathogens, including Babesia microti, A. phagocytophilum, and Ehrlichia chaffeensis, have caused infections after blood transfusion (41,42).Spirochetemia might be higher or more prolonged for B. miyamotoi infection than for B. burgdorferi infection, suggesting that the risk for transmission from blood transfusion is greater for B. miyamotoi (17,46).Nevertheless, all 7 patients who had a confirmed infection and an available exposure history available reported a recent tick bite, suggesting that receipt of blood products or organs might simply reflect risk factors for more severe illness caused by compromised immune status, rather than a potential route of B. miyamotoi transmission.
The frequency of recurring febrile episodes in HTRF caused by B. miyamotoi is not well understood.The percentage of patients with confirmed illness who had recurring fever (22%) was higher than those reported in a case series from Russia (11%) (10) and in a case series from the United States (4%) (17).All relapsing fever group borreliae display antigenic variation, a shift in expressed proteins that creates recurring febrile episodes (43,44).B. miyamotoi infection Low clinical awareness, limited availability of PCR testing, and limited specificity of available serologic assays make HTRF case identification challenging.Currently, diagnosis of tickborne infections requires clinicians to order tests specific to each suspected pathogen; the expanded use of multiplex direct detection assays, or tickborne panels, in commercial laboratories might improve detection of B. miyamotoi and other tickborne infections.Metagenomics approaches are also an increasing opportunity to improve direct detection of tickborne infections, including co-infections (46,47).Those approaches are particularly appealing for improved detection of B. miyamotoi infection because the spirochetes appear to be present in sufficient quantities in blood for detection by using molecular methods (17,46,48).Highlighting the usefulness of PCR-based diagnostic methods for B. miyamotoi infection, a study found that among patients with PCR-confirmed B. miyamotoi infection, the sensitivity of GlpQ IgG was <55% when assayed <20 days after illness onset, which increased to 74%-86% when assayed 21-150 days after illness onset (33).GlpQ is a common serologic target for differentiating infection with relapsing fever group borreliae from those that cause Lyme disease.However, there is limited information on its specificity for relapsing fever group borreliae, constraining its clinical usefulness.
Without sufficiently specific serologic assays, the frequency of exposure in the population and characteristics of more mild illness is difficult to ascertain.In addition, GlpQ-based assays cannot distinguish between infection with hard tick and soft tick relapsing fever borreliae, which can co-occur in some areas (i.e., along the Pacific Coast).In those circumstances, a comprehensive exposure history is necessary to direct public health intervention.However, many patients who have suspected tickborne infections, including HTRF, receive doxycycline empirically, which would effectively treat B. miyamotoi even if clinicians had not suspected this specific infection.People with mild symptoms who live in a Lyme disease-endemic area might be more likely to receive empiric therapy.
The first limitation of the surveillance data we describe is that cases identified through passive surveillance probably represent more severe disease because all persons necessarily sought medical care for an illness and obtained laboratory testing.As previously mentioned, persons who have mild symptoms or asymptomatic infections would not be detected by current public health surveillance approaches.Thus, the severity of HTRF caused by B. miyamotoi is difficult to reliably measure through this mechanism.Accordingly, the frequency of hospitalization in those data are probably an overestimate caused by inherent ascertainment bias.Second, the nature of public health surveillance activities precludes knowledge of the targets and performance of assays used by commercial laboratories.Third, B. miyamotoi might cause human infection in states where the condition is not subject to public health reporting.A recent case of HTRF caused by B. miyamotoi was identified in California; that finding, in combination with acarologic and seroprevalence assessments, suggests potential for additional cases of human illness along the Pacific Coast (28,29,49,50).Fourth, detailed data on clinical features, such as rash, clinical course and resolution, or immunocompromised status or other medical concurrent conditions, were not collected consistently as part of surveillance-based case investigations, limiting our ability to thoroughly describe the clinical course or examine the effect of immunocompromising conditions or other concurrent conditions on clinical severity or presentation of HTRF caused by B. miyamotoi infection.Fifth, data on positive laboratory findings for other tickborne diseases were not regularly compiled as part of public health surveillance; thus, these HTRF cases could reflect patients co-infected with other tickborne diseases.
Public health surveillance in the United States supports that HTRF manifests as a nonspecific febrile illness during the summer months.B. miyamotoi is among the group of pathogens transmitted to humans by Ixodes spp.ticks, and the clinical manifestations might be similar to that of other tickborne diseases in the same geographic areas.The frequency of asymptomatic or mild illness caused by HTRF that resolves without treatment is not known, nor is the potential for longer-term complications of untreated infection.

SYNOPSIS
At present, B. miyamotoi is the only recognized cause of HTRF; if additional pathogens are identified, public health surveillance approaches will necessarily adapt, including through expansion of molecular testing to detect those pathogens.Infections identified through public health surveillance can enable expanded understanding of the clinical spectrum of emerging infectious diseases than what is possible through limited case series or reports, but surveillance depends on clinical suspicion, laboratory diagnostic test access, and state-based public health regulations that enable mandatory reporting of positive laboratory results to the public health system.Ongoing, coordinated public health surveillance for HTRF caused by B. miyamotoi will better define its clinical spectrum, severity, incidence, and geographic distribution, and inform associated clinical and public outreach efforts to improve recognition.However, improved access to direct detection of B. miyamotoi through unbiased and widely available PCR-based assays, as well as clinically validated serologic markers, are needed to clarify the frequency and severity of the illness.

Figure 1 .
Figure 1.US states that conducted surveillance for hard tick relapsing fever caused by Borrelia miyamotoi during 2013-2019 and year in which surveillance began.

Figure 2 .
Figure 2. Number of annual cases of hard tick relapsing fever (vertical bars) and number of states reporting cases of hard tick relapsing fever caused by Borrelia miyamotoi (line), United States, 2013-2019.The left y-axis corresponds to the vertical bars, and the right y-axis corresponds to the line; scales for the y-axes differ substantially to underscore patterns but do not permit direct comparisons.States reporting cases in that year are shown.CT, Connecticut; MA, Massachusetts; ME, Maine; MN, Minnesota; NH, New Hampshire; NJ, New Jersey; RI, Rhode Island; VT, Vermont; WI, Wisconsin.

Figure 3 .
Figure 3. Patient age distribution for confirmed (A) and possible (B) cases of hard tick relapsing fever caused by Borrelia miyamotoi identified by using public health surveillance, United States, 2013-2019.

Figure 4 .
Figure 4. Reported cases of hard tick relapsing fever caused by Borrelia miyamotoi by month of symptom onset, United States, 2013-2019.Solid line indicates total number of reported cases each month, and bars indicate number of confirmed (white) and possible (gray) cases each month.

Table 1 .
Demographic characteristics for confirmed and possible cases of hard tick relapsing fever caused by Borrelia miyamotoi identified by public health surveillance, United States, 2013-2019* *Values are no.positive/no.tested (%) unless indicated otherwise.†Missing information for 25 persons.‡By Wilcoxon rank-sum test.§Missing information for 4 persons.¶By Fisher exact test.

Table 2 .
Clinical and laboratory findings for persons who have confirmed and possible hard tick relapsing fever caused by Borrelia miyamotoi among reported cases with available laboratory findings identified by public health surveillance, United States, 2013-2019* *Values are no.(%)unless indicated otherwise.IQR, interquartile range; NA, not available.†Definedas duration from symptom onset to first seeking medical care.‡Information regarding a patient's leukocyte count was available for 68 confirmed and 37 possible cases.§Information regarding a patient's platelet count was available for 69 confirmed and 36 possible cases.¶Information regarding a patient's alanine and aspartate aminotransferase levels was available for 60 confirmed and 36 possible cases.

Table 3 .
Diagnostic results for confirmed and possible cases of hard tick relapsing fever caused by Borrelia miyamotoi reported by public health surveillance, United States, 2013-2019*