Bartonella spp. seroprevalence in tick-exposed Swedish patients with persistent symptoms

Background Bartonella spp. are emerging pathogens transmitted by arthropod vectors, possibly including ticks. We have investigated signs of bartonellosis in Swedish patients with presumed tick-bite exposure and symptom duration of at least 6 months. Methods Serological testing for Bartonella henselae and Bartonella quintana was performed in 224 patients. Symptoms, tick exposure, evidence of co-infection and previous treatments were evaluated. Seropositive patients were compared to a matched group (twofold larger and negative serology) from the same study cohort. Results Seroprevalence was 7% for B. henselae and 1% for B. quintana, with one patient testing positive to both agents. Tick bites were reported by 63% of the patients in the seropositive group and 88% in the seronegative group and presumed tick exposure was more common in the seronegative group. Animal contact was equally common in both groups, along with reported symptoms. The most common symptoms were fatigue, muscular symptoms, arthralgia and cognitive symptoms. Exposure to co-infections was evenly distributed in the seropositive and seronegative groups. Conclusions Antibodies to Bartonella were more common in this cohort of patients than in cohorts of healthy Swedish blood donors in previous studies but lower than those in blood donors from southern Europe. Positive Bartonella serology was not linked to any specific symptom, nor to (suspected) tick-bite exposure. Graphical abstract

and Europe [9]. This tick is also the vector for Borrelia burgdorferi [9,10], and co-infections in ticks have been demonstrated, a phenomenon that may result in the transmission of more than one pathogen during a tick bite [5,11,12]. Bartonella henselae and other Bartonella spp. have been demonstrated in the blood of patients exposed to ticks [13].
Human bartonellosis can present with a wide range of symptoms and diseases [14][15][16][17]. For example, B. quintana is the cause of trench fever with recurring fever, headache and bone pain [18], and B. henselae may cause cat-scratch disease with lymphadenopathy, fever and myalgia [19]. Atypical presentations of cat-scratch disease have also been reported with neurological and/ or rheumatological symptoms [19][20][21]. Both B. quintana and B. henselae may cause endocarditis. Other Bartonella spp. have also been reported as the cause of endocarditis in a few patients [4]. Moreover, asymptomatic bacteremia with Bartonella spp. has been demonstrated in humans [22] and animals [7].
Patients suffering from persistent unexplained symptoms sometimes attribute these to a previous tick bite [23]. If Bartonella spp. are transmitted by ticks, an investigation for Bartonella infection may be relevant in this patient population. Therefore, we examined a cohort of Swedish patients with suspected previous tick exposure for the prevalence of antibodies against B. henselae and B. quintana and analyzed their epidemiological, clinical and baseline demographic data compared to a group of seronegative patients attending the same clinic.

Study population
We purposely selected participants from an exploratory study of human tick-borne infections conducted at the Center for Vector-borne Infections (CVI), Uppsala University Hospital, Uppsala, Sweden between October 2015 and December 2018 [24]. A total of 224 patients were enrolled in the principal study during this period. All patients completed standardized questionnaires on their symptoms and tick exposure and underwent a standardized medical and laboratory examination at the outpatient clinic. Patients had to fulfill at least four of seven predefined inclusion criteria of which symptom duration of > 6 months was mandatory. The other criteria were age ≥ 18 years; suspected tick-borne infection based on previous tick exposure; symptoms; laboratory findings; previous treatment for tick-borne infection; and/or suspicion of co-infection with other tick-borne infections. A summary of data on these patients has previously been published [24].

Bartonella serology
Patient sera were analyzed as part of routine diagnostics at the Uppsala University Hospital, Uppsala, Sweden for IgG antibodies against B. henselae and B. quintana by indirect immunofluorescence assay (IFA) using the Anti-Bartonella henselae/quintana IIFT Mosaic kit from Euroimmun AG (Lübeck, Germany) according to the manufacturer's instructions. Titers at 1:64 were considered to be the limit value for B. henselae and B. quintana and titers at 1:128 or higher were considered to indicate seropositivity. Positive samples were titrated to end titer.

Microbiological testing
All patients were also examined for other tick-borne infections known to be present in Sweden, such as serological testing for Borrelia burgdorferi, Borrelia afzelii and Borrelia garinii (Euroimmun ® , Lübeck, Germany) and Anaplasma phagocytophilum (Focus Diagnostics ® , Cypress, CA, USA). Serological tests (IFA) for antibodies against Babesia microti and Babesia divergens were performed at the Public Health Agency of Sweden, Solna, Stockholm and the National Institute for Public Health and the Environment (RIVM), Bilthoven, the Netherlands. Serological testing (IFA) and PCR assays for Rickettsia spp. and tick-borne encephalitis virus (Immunozym FSME IgM and IgG, respectively; Progen Biotechnik GmbH, Heidelberg, Germany) and the PCR assay for Candidatus Neoehrlichia mikurensis in blood were performed at Sahlgrenska University Hospital, Gothenburg, Sweden as previously described [24].

Statistical analysis
The Chi-square test and Fisher's exact test were conducted using IBM SPSS Statistics (SPSS, IBM Corp., Armonk, NY, USA).

Patient demographics
The study included 224 patients, of whom 16 (7%) were seropositive for antibodies against B. quintana or B. henselae, or both. An anti-Bartonella spp. IgGseronegative group of 32 patients was matched for age, sex and region of residence from the same cohort of patients investigated at the CVI. The seropositive group included 10 women and six men with a mean age of 57 (range 23-77) years; the seronegative group comprised 20 women and 12 men with a mean age of 56 (range 20-80) years. Most patients in both groups (88%) were from central Sweden, with the majority (69% of all patients) from the region of Uppsala and Stockholm. The remaining patients were from the southern part of Sweden; none came from the northern region of Sweden.

Co-infections
The search for other tick-borne pathogens revealed a possible infection with Candidatus Neoehrlichia mikurensis in one of the Bartonella-seropositive patients. Previous exposure to, but no ongoing infection of B. burgdorferi, A. phagocytophilum and Rickettsia spp. was demonstrated. The number of patients in each group is listed in Table 1. None of the patients had signs of previous exposure to or ongoing infection with tick-borne encephalitis virus, B. divergens or B. microti. There was no difference between the groups in the number of patients with previous exposure to tick-borne pathogens.

Tick-exposure
Epidemiological factors related to tick exposure are shown in Table 2. Most patients who were either seropositive or seronegative for Bartonella reported tick bites approximately once per year. Although the difference was not statistically significant, the frequency of tick bites tended to be higher in seronegative patients. Furthermore, when the prevalence of tick bites was investigated in all anti-Bartonella spp. IgG-negative patients in the entire cohort investigated at the CVI clinic (n = 208), 79% reported tick bites. All participants in the present study reported that they walked in woods and fields over the years. Patients in the seronegative group reported berry picking (P = 0.012) and being on a boat or in the archipelago (P = 0.005) more frequently than patients in the seropositive group. When we looked at tick exposure only in the group of patients with a titer > 1:64 (i.e. 7 patients) and compared these patients to the seronegative group, we found was no difference in terms of the factors related to tick exposure. Travel history was not assessed due to incomplete or missing information on this variable from the patients.

Symptoms
Reported symptoms in Bartonella-seropositive and -seronegative patients are listed in Table 3. The most frequently reported symptoms in the seropositive group were fatigue, muscular symptoms, arthralgia and sleeping problems. Muscular symptoms included muscle pain, spasms, muscle weakness and cramps, while sleeping problems included insomnia, prolonged sleep duration and daytime sleeping. In the seronegative group, fatigue was also a common symptom. Both the seropositive and negative groups rated fatigue as severe. In the seronegative group cognitive symptoms (e.g. confusion, concentration difficulties, difficulty processing new information, searching for names, forgetfulness, disorientation and dysphasia) were the most common symptoms. In three of the patients with previous Bell's palsy, the serological testing of the cerebrospinal fluid was positive for Borrelia, which was interpreted as previous neuroborreliosis.
There was no significant statistical difference between the two groups in terms of symptoms. Comparison of the symptoms in the group of patients with a positive anti-Bartonella spp. IgG titer > 1:64 (i.e. 7 patients) to those of the seronegative group revealed that cognitive symptoms were more common (P = 0.022) in the seronegative group. However, this finding should be considered uncertain due to the small size of the groups.

Antibiotic treatment
Previous antibiotic treatment in Bartonella-seropositive and Bartonella-seronegative patients is shown in Table 4. Three patients had been treated with ≥ 2 different antibiotics and three patients reported ≥ 2 courses of treatment with the same sort of antibiotics. Doxycycline was the most commonly used antibiotic, reported by 38% in the seropositive and 28% in the seronegative group. All other antibiotics were reported only by very few patients. Two patients in the Bartonella-seropositive group were recommended doxycycline treatment; one patient was positive for Candidatus Neoehrlichia mikurensis in the blood and had a titer against B. henselae of 1:128, and the other patient had a titer of 1:256 against B. henselae that persisted during follow-up 2 months later. These findings together with long-lasting diffuse symptoms and no previous treatment with antibiotics effective against Bartonella infection resulted in the recommended treatment.

Discussion
Many patients suffer from persistent medically unexplained symptoms, and if these patients have been exposed to ticks, an investigation into possible tick-borne infection may be advisable. Whether Bartonella infection should be included in this investigation has yet to be determined. In the present study of Swedish patients with presumed tick exposure and persistent unexplained symptoms, we demonstrated a slightly higher seroprevalence against B. henselae and B. quintana than reported in previous studies of healthy blood donors [25,26]. However, a positive Bartonella serology could not be linked to any specific symptom or to a higher prevalence of any epidemiological risk factor related to tick exposure.
Ticks are considered to be one of the most important vectors in Europe, but infected patients may not always have noted a tick bite [27]. In several studies, PCR assays have demonstrated the presence of Bartonella spp. in ticks (including I. ricinus), both in Europe and the USA [1,6,12]. However, in Europe, the prevalence of Bartonella spp. in ticks varies greatly from only 0.6% in Denmark [28] to 11.8% in Germany and 38.2% in France [7]. In Sweden, the presence of Bartonella spp. in ticks has been investigated in two studies: one on 167 ticks [29] collected in central Sweden and the second on 1663 ticks collected in Sweden and the Åland Islands, Finland [30]. Both studies failed to demonstrate Bartonella spp. in the examined ticks.
The seroprevalence of antibodies against Bartonella spp. differs between countries. In Europe, the seroprevalence against B. henselae in healthy populations is higher in the central and southern countries, ranging from 19% in Germany to 23% in Austria to 57% in Croatia [31][32][33]. In Sweden, the seroprevalence against B. henselae is 3% in orienteers [25], 14% in drug addicts [34] and 29% in a homeless population [35], and ranges from 1 to 1.6% in healthy blood donors [25,26]. A high seroprevalence against Bartonella spp. in populations of homeless and addicts is also known from other studies [22,36,37]. The present study demonstrated a seroprevalence of 7% against B. henselae, which is higher than that reported previously in Sweden (except for homeless persons), even in persons at high risk of exposure to tick bites (such as orienteers). The prevalence of antibodies against B. quintana was 1% in the current study, which is somewhat higher than previously demonstrated in healthy blood donors (0.2-0.3%) [25,26] but lower than in values reported for orienteers (3%) and a homeless population (4.2%) [25,35]. One patient in the present study had antibodies against both B. henselae and B. quintana, which may represent cross-reactivity; the cross-reactivity between these species has been shown in other studies to be high [38,39]. Unfortunately, information on travel history was missing for many patients so no conclusions could be made on the role of travel on the seroprevalence against Bartonella spp.
The patients included in the present study were all being investigated for possible tick-borne infections. Hence, all had most likely been exposed to ticks to some extent. However, our investigation of the different epidemiological factors associated with tick exposure did not show a higher prevalence of any factor in patients with positive Bartonella serology. Rather, there was a tendency to a higher prevalence of some factors in the group with negative serology; these included, in particular, a higher frequency of gardening, berry picking in forested areas and being on a boat in the archipelago. Also, previous tick bites tended to be more common in the seronegative group, as did possession of a cat. This latter finding was against expectation, as it was expected that the seropositive group would have more contact with cats given that cat scratches are the primary mode of transmission of B. henselae to humans [19]. However, one previous serological study of cats in Sweden showed a seroprevalence of 1% against B. henselae [40], and when blood from cats was investigated for B. henselae with culture or PCR assays the prevalence was 2% [41]. Co-infections are known to occur both in ticks and humans, including Bartonella spp. [11,42]. Most patients in our study showed signs of previous exposure to at least one other tick-borne pathogen, with B. burgdorferi being the most common. No difference between the two groups was noted for co-infections. As a result of the study design, a high prevalence of exposure to various tick-borne pathogens is not surprising.
Cat-scratch disease is caused by B. henselae and usually presents with fever and enlarged lymph nodes. However, atypical symptoms occur in up to 20% of patients, mostly in adults [19]. Neurological symptoms, such as meningoencephalitis, myelitis and neuroretinitis, have been described in up to 10% of cases [19], and musculoskeletal manifestations, including myalgia and arthralgia/arthritis, also appear in up to 10% of patients [43]. An investigation of patients examined by a rheumatologist demonstrated a high Bartonella seroprevalence and a high rate of positive Bartonella PCR in the blood [21]. The symptoms most frequently reported by patients in our study were fatigue, arthralgia, muscular symptoms, sleeping problems and cognitive symptoms. These symptoms were reported at similar frequencies from both seropositive and seronegative patients, with the exception of cognitive symptoms, which were more common in the seronegative group. In our seropositive patients, fatigue was the single most common symptom (frequency 94%).

Conclusions
We have demonstrated a somewhat higher seroprevalence of antibodies against B. henselae and B. quintana in presumed tick-exposed patients than reported previously in blood donors in Sweden. However, Bartonellaseropositive patients did not have a higher prevalence of epidemiological risk factors associated with tick exposure, nor did they have more tick bites, when compared to Bartonella-seronegative patients from the same study cohort. Symptoms were also the same in both seropositive and seronegative patients. Accordingly, these findings do not add support to ticks being vectors for Bartonella spp. transmission. Why the prevalence of Bartonella antibodies is higher in this population remains to be determined.