This report presents the serological evidence of recent or probable rickettsial infection, in most cases, probable co-infection, in almost 6 % of patients from a prospective, clinical investigation (Study 1), where patients were recruited on the basis of EM and/or general signs of infection (fever, headache, muscle pain) following known or probable tick bite (Table ). Three patients had IgG antibodies at the first visit, probably due to a past infection, and another 17 patients showed seroconversion or a 4-fold rise in IgG titre at the second visit, of which seven had a recent or current infection and ten a past infection, of which five were judged as probable infection. IgG and IgM antibodies normally appear 3–10 days after disease onset and peak after 3–4 weeks. Treatment within 2–5 days of disease onset may also inhibit antibody production. Seventeen of the 20 patients had only IgM titre at the first visit, while 10/20 were seronegative for IgM at the second visit, which shows the importance of paired sera for the analysis of both IgG and IgM antibodies to establish a reliable serological diagnosis. The specificity of the serological response showing the presence of IgG antibodies to Rickettsia-specific protein was demonstrated by WB analysis in three of the seroreactive samples (S2) (Fig. ).
The study also shows that Rickettsia
spp. infection occurs either as a single infection or as a co-infection in patients with EM or serological evidence of Borrelia
spp. or Anaplasma
sp. infection. Of the 206 patients in the study, 174 were recruited on the basis of EM and 32 because of flu-like symptoms in combination with a preceding tick bite. All patients who were seroreactive for Rickettsia
spp. in paired sera presented different disease symptoms comparable to those of LB (Table ). The observed symptoms were similar and gave no guidance in relation to the causative agent. The variability of the clinical picture in LB has been highlighted in a recent report showing that asymptomatic B. burgdorferi
infections, documented by seroconversion, were found more often than symptomatic infections in individuals bitten by a B. burgdorferi
-infected tick [18
]. No corresponding study has been carried out regarding Rickettsia
, but in the present study, all patients with proven rickettsial antibodies presented varying degrees of symptoms. However, the symptoms of Rickettsia
spp. or B. burgdorferi
infection are, in both cases, quite general and do not allow us to distinguish between the agents.
EM is typically regarded as a clinical sign consistent with LB [19
]. Fifteen of the 19 Rickettsia
spp.-seroreactive patients showed EM, of whom eight were serologically positive also for Borrelia
spp. EM occurred also in six of the seven (6/206, 2.9 %) patients who were seroreactive only for Rickettsia
spp., which may indicate that either the Rickettsia
infection was causative or that Lyme infection did not produce antibody development. There are only a few previous reports in which EM or erythema resembling EM has been associated with rickettsial infection [20
]. Further serological examination including several agents in prospective clinical cases may provide more guidance as to the causative agent. Skin biopsy for PCR and/or immunohistochemistry and the detection of Rickettsia
spp. and/or Anaplasma
sp. organisms are also likely to be of value in clarifying the cause. Because several tick-borne agents give similar symptoms and, as the present study indicates, may occur as co-infections or separately, the task of providing a correct diagnosis is complex. In patients exposed to ticks presenting with unspecific symptoms, there is reason to consider the choice of antibiotics if other infectious agents are not excluded [14
]. If EM is the only found symptom and causes other than LB can be excluded, it is likely that phenoxymethylpenicillin is sufficient as the drug of choice. Thus, serological testing is important for a full clinical assessment of underlying causes.
Study 2 gives a similar picture for 16 seroreactive patients with symptoms involving the skin, joints (culture-negative arthritis), headache and cough. The reason for the cough symptom requires further study, but it is known that rickettsiosis gives pulmonary vasculitis, which could be a possible explanation.
Screening for Rickettsia
spp. from different localities in Sweden, using PCR, has demonstrated a mean infection prevalence of 1.5–17.3 % for R. helvetica
in I. ricinus
ticks, including all stages, proving that Sweden is an endemic area for this agent and that the risk of infection is consistent with the tick’s distribution [4
Previous smaller serosurveys in Sweden have shown IgG antibodies to Rickettsia
spp. in the serum of up to 4.4 % of tick-exposed subjects, compared to 0.6 % in healthy blood donors [15
]. In a prospective study of Swedish recruits who trained in the coastal areas, 8.9 % showed seroconversion compared to the proportion of 9.2 % found to be seroreactive in France in a group of forestry workers, and in a study from Laos, 2.9 % of adults admitted to hospital because of fever showed seroconversion to R. helvetica
The current more extensive study demonstrates both seroconversion and significant rise of titres for Rickettsia spp. in single-infected individuals as well as in those patients co-infected with other tick-borne agents known to present similar clinical symptoms. The complexity of the clinical picture needs to be considered when diagnosing the causative agent and selecting appropriate treatment. It also demonstrates a need for further studies.