Tularemia Epidemic in Northwestern Spain: Clinical Description and Therapeutic Response

Tularemia is a zoonosis caused by Francisella tularensis, a small gram-negative coccobacillus that is highly infective. The principal vectors are ticks and wild rabbits, and the natural hosts are lagomorphs and other rodents. The 2 principal biogroups are Francisella tularensis biovar tularensis (type A) and Francisella tularensis biovar palaearctica (type B) [1]. Here, the clinical characteristics of 142 patients from the area of Tierra de Campos (Valladolid, northwestern Spain) diagnosed with tularemia between December 1997 and February 1998 are described. Special reference is made to the therapeutic response and to the use of ciprofloxacin in the treatment of the illness.

Patients and methods. In total, 142 patients diagnosed with tularemia were studied in the Department of Internal Medicine of the Rio Hortega Hospital (Valladolid, Spain). By use of a clinical history protocol, the following data were collected for all patients: symptoms; physical signs; and results of laboratory tests, including IgG serology for F. tularensis. Samples of exudates and lymph node aspirations (n = 25) were sent for culture in Bcye media (Oxoid) for Legionella, in a BacT-Alert blood culture bottle (Organon Teknika), and on cystine enriched chocolate agar. Direct immunofluorescence was performed on 20 aspirates.

The diagnosis of tularemia was considered positive when the patient had an F. tularensis antibody titer of IgG ⩾1 : 160 or when the patient had seroconversion or F. tularensis was isolated from a clinical sample. Six clinical forms of disease were defined: ulceroglandular, glandular, typhoidal, oculoglandular, pharyngeal, and pneumonic. The patients were treated with intramuscularly administered streptomycin at a dosage of either 1 g every 24 h or 500 mg every 12 h for 7–10 days. For patients who refused parenteral treatment or for whom aminoglycosides were contraindicated, doxycycline (100 mg orally every 12 h for 2 weeks) or ciprofloxacin (750 mg orally every 12 h for 14–28 days) was administered. The therapeutic response could not be assessed for 6 patients.

Therapeutic failure was defined by the presence of 1 of the following findings: persistence or recurrence of fever, increase in the size or appearance of new lymphadenopathies, and persistence of the constitutional syndrome with elevation of the levels of the proteins associated with the acute phase of infection [2]. The therapeutic option used in cases of treatment failure was ciprofloxacin at the above-described dose or streptomycin, if it had not been used previously.

We used the Kolmogorov-Smirnov test to compare normality of the variables. The χ² test and Fisher exact test were used for variables with a normal distribution, and the Kruskal-Wallis test and the Mann-Whitney U test were used for variables that did not fit this distribution.

Results. A total of 142 patients was included in the study. The mean age (±SD) of the patients was 52 ± 14 years (range, 14–82 years). There were more females than males (89 vs. 53, respectively). A total of 138 patients (97.2%) had previous contact with hares, 119 (83.8%) had prepared hare carcasses, and 19 (13.3%) had handled hare meat. For 4 patients (2.8%), animal contact could not be confirmed. The number of hunters affected was 16 (11.3%), all of whom were male. Tularemia was diagnosed on the basis of the following findings: culture of F. tularensis biovar palaearctica for 3 patients (2.1% of the patients and 12% of the cultures performed); seroconversion for 19 patients (13.4%); and a compatible clinical picture associated with an elevated F. tularensis antibody titer for 120 patients (84.5%). Direct immunofluorescence of aspirates had results positive for F. tularensis for 7 of the 20 patients for whom this test was carried out (35%); all these patients had high titers of positive antibodies to F. tularensis. The mean delay (±SD) between the onset of symptoms and diagnosis was 47.5 ± 31 days (range, 3–145 days).

The symptoms and signs of the patients are presented in table 1. The most frequent symptoms were asthenia and weight loss, but no relationship was found between the constitutional syndrome and other data. Fever was the most frequent sign (present in 90.8% of patients), with a mean duration of 13 days. The lymphadenopathies were localized in the epitrochlea, axillary, and cervical regions. The findings of laboratory tests were normal, except for an increase in the levels of fibrinogen and variant surface glycoprotein. Table 2 shows the distribution of the different clinical forms of disease. The ulceroglandular form was the most frequent (87 cases). Analysis of the differences between the clinical forms showed that the glandular form was associated with a lower frequency of fever and a longer diagnostic delay (mean ± SD, 82.8 ± 42.5 days vs. 43.4 ± 29.1 days for the ulceroglandular form; P = .016). The ulceroglandular form showed a higher frequency of therapeutic failure, compared with the typhoidal form (RR, 10.1; P = 0.028).

View larger version:

• In this window
• In a new window

• Download as PowerPoint Slide

Table 1

Clinical manifestations in patients with tularemia during an outbreak in Spain (n = 142).

View larger version:

• In this window
• In a new window

• Download as PowerPoint Slide

Table 2

Frequency of the 6 clinical forms among patients with tularemia in an outbreak in Spain (n = 142).

The treatments used are shown in table 3, together with the rates of therapeutic failure for the different antibiotics. There were no differences in the treatments used for the different clinical forms of disease, except for the ganglia form, for which ciprofloxacin was the predominant treatment (58% of affected patients). Therapeutic failure occurred in 32 patients (22.5%). Ciprofloxacin showed the lowest frequency of failure; however, this difference was only statistically significant in comparison with doxycycline treatment (RR, 15.75; P < .018). Treatment failure was associated with the ulceroglandular form of disease and the use of doxycycline as the initial treatment. A second course of antibiotic treatment was given to 46 patients (32 because of therapeutic failure, 12 because of adverse drug reactions, and 2 because of incorrect completion of the treatment). Ciprofloxacin (n = 34 [73.9%]), streptomycin (n = 11 [23.9%]), and gentamicin (n = 1 [2.2%]) were administered. The response was favorable for all patients except 6 (4 who received ciprofloxacin, 1 who received streptomycin, and 1 who received gentamicin), who received a third course of macrolides. The poor response in these patients was indicated by the persistence of the asthenia; all of them remained afebrile.

View larger version:

• In this window
• In a new window

• Download as PowerPoint Slide

Table 3

Initial treatment administered and treatment failure rates for the different antibiotics given to patients with tularemia in an outbreak in Spain (n = 142).

Discussion. Tularemia is a zoonosis caused by F. tularensis. No previous outbreak had been reported in Spain. The most notable epidemiologic finding was contact with hares. The microorganism was isolated from dead hares during the epidemic [3]. The illness was contracted through the handling and processing of the hares, but the infection mechanism of the hares is not known. The most likely hypothesis is the importation of infected hares from countries where tularemia is endemic. Other possibilities, such as natural migration or transmission by birds, are unlikely because the epidemic remained localized in a specific geographical region.

The principal port of entry was the skin. The pathogen usually enters through small lesions in the skin, although it is possible that it can penetrate through grossly intact skin [4]. Glandular involvement with no visible cutaneous lesions was found in 13 patients, which may support the latter theory. It seems impossible to rule out unnoticed small cuts or trivial skin lesions. Other possible routes of infection include the inhalation of aerosol droplets from the skin of dead animals, leading to the pneumonic or typhoidal forms of disease [3]. Additional routes of infection, such as tick bites, the eating of undercooked, contaminated meat, and the drinking of contaminated water, were not observed in the present study. Also, the mechanism of infection of F. tularensis biovar palaearctica, which was isolated during the outbreak, is predominantly contact with infected hares and not arthropod bites [2]. This pathogenetic mechanism explains the clear predominance of the ulceroglandular form of tularemia.

One outstanding finding of this study is the delay between the onset of symptoms and diagnosis and initiation of therapy, which would explain the differences between the clinical findings of the present study and those described in the literature [5]: specifically, the absence of rigors and the presence of a constitutional syndrome. The delay was due to the previous absence of this illness in Spain.

Diagnosis was made on the basis of a compatible clinicoepidemiologic picture and an F. tularensis antibody serum titer of ⩾1 : 160. F. tularensis was isolated in 3 patients, from a lymph node in 1 and from cutaneous ulcers in the other 2; this proportion was similar to that in other studies [5]. The US Centers for Disease Control and Prevention requires seroconversion for diagnosis [1]. Only 13.4% of the patients we studied fulfilled this requirement; most of the group had high titers at diagnosis. This was because of the delay from the onset of the clinical symptoms—6 weeks—the point at which the antibodies against F. tularensis peaked [6]. However, given that antibodies to F. tularensis are found in <1% of the general population in nonendemic areas [5], and taking into account the risks of culture (i.e., the risk of infection for the laboratory technician), the serological tests are the diagnostic technique of choice in these areas.

One of the most notable features of this study was the use of ciprofloxacin in the treatment of the illness. The recommended agents for treatment of tularemia include aminoglycosides (streptomycin and gentamicin) and tetracyclines [7]. The strains of F. tularensis isolated in this study showed a high sensitivity to third-generation cephalosporins, aminoglycosides, macrolides, and quinolones [8]. The fluoroquinolones are an excellent alternative. Syrjäla et al. [9] demonstrated that 10 strains of F. tularensis have in vitro susceptibility to ciprofloxacin, norfloxacin, ofloxacin, and pefloxacin. In the English-language literature, 10 reports have been published of patients treated with orally administered quinolones; all these patients were cured [10].

Ciprofloxacin was used for 66 patients in this study, and better results were obtained with this drug than with the usual therapeutic regimens. It was used as second-line treatment after the failure of other treatments in 34 patients, and it achieved a cure in 30. Among patients who did not respond to ciprofloxacin therapy, the persisting symptom was asthenia, but fever did not recur in any of them. There were no significant differences between the groups of patients in which the different treatments were used, except for the diagnostic delay, which was greatest in the ciprofloxacin group; the other epidemiologic, clinical, and analytic variables were similar.

The prognosis for our patients was good; no patient died. The strain of F. tularensis biovar palaearctica that was responsible for this epidemic is less virulent than biovar tularensis and causes a milder clinical syndrome [1]. However, a high proportion of patients failed to respond to therapy. Therapeutic failure has been related to the presence of fluctuating lymphadenopathies, short courses of treatment, the presence of severe involvement, associated chronic illness, and a delay in the initiation of antibiotic therapy. In the literature, there are no clear, established criteria for deciding on a second course of treatment after treatment failure. Further courses of antibiotics are indicated [2, 11] when the patient has persistent fever, when new adenopathies appear, when existing adenopathies increase in size, and when constitutional syndrome with elevated markers of inflammation persists.

The therapeutic failure rate found in this study was not related to the diagnostic delay, as has been reported elsewhere [12]. There was, however, an association with the use of tetracyclines as initial treatment, a finding that has already been reported [12], and with the ulceroglandular clinical form of disease. There is a hypoxic acid medium within a fluctuating lymph node that impedes the action of the aminoglycosides. In conclusion, on the basis of our results, we believe that ciprofloxacin should be considered as first-line treatment for tularemia because of its efficacy, its relatively few side effects, and its ease of administration. We realize, however, that our was not a blind, randomized study and that therefore our results should be treated with caution.

• Received September 11, 2000.
• Accepted December 28, 2000.

• © 2001 by the Infectious Diseases Society of America