In September 2006, a 56-year-old woman was hospitalized in Falun, Sweden, after 3 weeks of illness with gradually worsening headache and fever. She had no lymphadenopathy, rash, eschar, or history of tick bite or tick exposure. Radiographs showed a small, retrocardial pulmonary infiltrate, but the patient had few, if any, respiratory symptoms. Laboratory tests showed elevated C-reactive protein (56–128 mg/L), slightly elevated to reference-level leukocyte count (9,800–12,000 cells/μL), slightly low thrombocyte count (150,000 cells/μL), and reference values of hemoglobin and of aspartate and alanine aminotransferases. Cerebrospinal fluid (CSF) showed a slight pleocytosis (28 cells × 10⁸/L, of which 18 ×10⁸/L were mononuclear cells) but was otherwise within reference limits. CSF was stored at –20°C in a regular freezer and thawed only when used 1 year later. Negative results were obtained for blood and CSF cultures and for investigation for herpesviruses, tick-borne encephalitis, and enteroviruses. Urine was negative for Legionella and pneumococcal antigens, and serum was negative for antibodies against Borrelia burgdorferi. Computed tomography images of the brain and sinuses were unremarkable.

Intravenous administration of cefuroxime had no effect on the fever. Because atypical pneumonia was suspected, treatment was changed after 3 days to doxycycline (100 mg 2×/day). After 2–3 days the patient’s fever was gone, and she slowly recovered. The treatment was continued for 10 days. At a follow-up visit 1 year later, the patient was still well but had been asthenic for several months. No antibodies against Mycoplasma pneumoniae or Coxiella burnetii were found at the follow-up visit, and no other possible causative agent was confirmed. After giving informed consent, the patient was retrospectively included in an ongoing project that involved searching for fastidious organisms.

The patient’s previously frozen CSF was divided into 2 aliquots; bacterial DNA was extracted by using a MagNa Pure Kit (Roche Diagnostics GmbH, Mannheim, Germany) according to the manufacturer’s instructions. A genus-specific real-time PCR, as described by Stenos et al. (12), was used to detect SFGR. The PCR was performed in a Lightcycler 1.0 Real-Time PCR System (Roche Diagnostics GmbH) by using an LC Taqman Master Kit (Roche Diagnostics GmbH), primers, and TaqMan probe targeting the citrate synthase gene (12). To minimize risk for contamination, 0.25 µL LC uracil-DNA glycosylase (Roche Diagnostics GmbH) was included in each reaction. The positive control contained purified DNA of R. helvetica originally isolated from a domestic I. ricinus tick (3); the negative control contained sterile water. Positive samples were further analyzed by using 3 nested PCRs that amplify the 17kDa, outer membrane protein B (ompB), and 16S rRNA gene fragments as previously described (3,13,14) (Table). Amplification was conducted in a DNA thermal cycler (Hybaid, Ashford, UK) and a MJ Mini Gradient Thermal Cycler (Bio-Rad, Hercules, CA, USA), and expected fragment sizes were confirmed by gel electrophoresis in 2% agarose. Direct cycle sequencing analysis of both strands of nested PCR products was performed at the Center for Genomics and Bioinformatics, Karolinska Institutet, Stockholm, Sweden.

Rickettsial DNA was amplified by real-time PCR from both CSF aliquots. Positive samples were further examined by using nested PCRs. The sequences obtained were 165 (17kDa) and 253 bp (ompB) and shared 100% similarity with the corresponding gene sequences of R. helvetica (GenBank accession nos. EU407139, EU407140).

Figure

Figure. Rickettsiae in infected Vero cells. Sample from cerebrospinal fluid of patient with subacute meningitis, Sweden, 2006. Gimenez stain; original magnification x1,000.

To isolate the pathogen, we injected CSF from the frozen aliquot in volumes of 10 μL in a 25-cm³ flask into confluent monolayers of Vero cells and 80 μL in the other (15). After incubation, the cell culture was maintained in Dulbecco modified Eagle medium containing 10% fetal calf serum and kept in a humid cell chamber in 5% CO₂, at 32°C, to allow rickettsiae to multiply. All cell lines and reagents were checked weekly for growth or bacterial contamination. Detection of growing rickettsiae was monitored by using Gimenez staining and an immunofluorescence assay of cells collected after centrifuging the medium and staining with rabbit antirickettsial hyperimmunserum and Alexa Fluor 488 goat antirabbit immunoglobulin (Ig) G (H+L) conjugate (Invitrogen, Carlsbad, CA, USA) as secondary antibody (Figure).

After 6 weeks, many intracellular bacteria were observed in the cells. Rickettsial DNA was verified by real-time PCR (12). The sequences obtained by nested PCR for the 17kDa and ompB genes in the isolate grown in Vero cells were identical to the sequences of the isolates obtained from the CSF (3,13,14). Amplification and partial sequencing of the 16S ribosomal RNA gene of the isolate produced fragments of 1,400 and 750 bp, respectively, which were 100% homologous to fragments of the deposited 16S ribosomal DNA sequence of R. helvetica from ticks (GenBank accession no. L36212).

SFGR antigen prepared from isolates grown in Vero cells of R. helvetica from an I. ricinus tick and from the patient was applied to each well of microscope slides. The antigen was dried, fixed in acetone, and incubated with serial dilutions of serum or CSF, as previously described (7). The positive control was serum from a patient with proven Mediterranean spotted fever and end-point IgG titers of 160 (provided by the Swedish Institute for Infectious Disease Control); the negative control was phosphate-buffered saline and serum from 3 healthy blood donors. IgG was detected by fluorescein isothiocyanate–conjugated γ-chain–specific polyclonal rabbit antihuman IgG (DakoCytomation A/S, Glostrup, Denmark). Microimmunofluorescence assay showed IgG end titers of 160 and 320 in the early-phase serum sample when the isolates of R. helvetica from tick and patient, respectively, were used as antigens. No antirickettsial IgG was detected in CSF when either isolate was used as antigen.

For patients with fever and headache but no rash or eschar, diagnosis is difficult and can probably not be based only on epidemiologic, clinical, and standard laboratory criteria. It therefore seems that in SFGR-endemic areas, SFGR should routinely be included in the differential diagnosis of cause of meningitis. Appropriate antimicrobial drug therapy is essential for prompt recovery and prevention of complications.

SFGR isolation is usually not available in ordinary hospital laboratories and is too time-consuming to be a diagnostic alternative in clinical settings. Although this patient’s CSF had been stored in a regular freezer for 1 year, the rickettsial organisms were still viable.

PCR seems to be the most practical way to diagnose a suspected central nervous system disorder such as meningitis. The amplified nucleotide sequences were long enough to exclude other related rickettsial species. For example, the differences from other related rickettsiae were 10 and 5 nt for the R. monacensis 17kDa and ompB gene fragments, respectively, and 8 nt for R. slovaca ompB. Our study suggests that R. helvetica may cause infection of the CNS. When seeking to diagnose possible agents of meningitis, the usefulness of PCR and the relevance of the broader clinical spectrum of acute febrile illness caused by R. helvetica should be considered.