Tularemia is a zoonotic disease caused by the Gramnegative coccobacillus, Francisella tularensis. This organism can be transmitted from contact with infected animals (e.g. rabbits, squirrels, deer) or from insect bites (e.g. deerflies, ticks).1 Although the common clinical illnesses include the ulceroglandular, glandular, oculoglandular, oropharyngeal, typhoidal and pneumonic forms, meningitis is extremely rare. We report a case of tularemic meningitis and a review of the current literature. This case should remind clinicians that meningitis caused by F. tularensis can occur and be difficult to diagnose.
Case report. KH, a previously healthy 4-year-old Caucasian female child, presented to her primary care physician with a 2-day history of malaise, fever to 102°F, cough and painful postauricular swelling. After this initial evaluation she received a dose of intramuscular ceftriaxone and treatment was begun with oral amoxicillin. She subsequently presented to Arkansas Children's Hospital for further evaluation the next day. Her past medical history was unremarkable and she specifically denied exposure to sick contacts, trauma, animals or tick bites.
On physical examination at this time, KH was febrile (40.5°C) but was alert and cooperative. She had a 1- by 1-cm erythematous, nontender scalp papule with a purulent base and a 2- by 3-cm warm, painful right postauricular lymph node with bilateral otitis media. Laboratory evaluation included a white blood cell count (WBC) of 23 500/mm3 (21% neutrophils and 71% band forms), a platelet count of 259 000/mm3 and a negative tularemia antibody titer. Blood cultures were obtained and the patient was given intravenous nafcillin and cefotaxime empirically for suspected cellulitis, adenitis and acute otitis media with effusion. An ultrasound examination of the neck the next morning revealed only inflammatory changes in the anterior cervical chain, with overlying cellulitis but no focal abscesses. Although KH developed increased activity and appetite, she remained febrile and continued to complain of postauricular pain.
Because of her persistent fever and pain, on the sixth day of hospitalization a repeat ultrasound examination and computed tomography scan of the neck was performed which again revealed only the existing lymphadenopathy without a focal abscess. On the seventh day a culture of the scalp wound was obtained, the tularemia antibody titers were again negative, and antibiotic therapy was changed to clindamycin therapy and empirically given gentamicin for presumed ulceroglandular tularemia. The Gram-stained smear of this ulcerative lesion revealed no organisms.
Overnight KH became increasingly irritable and lethargic and developed projectile emesis. A lumbar puncture was performed. The cerebrospinal fluid (CSF) analysis revealed a WBC of 1570/mm3 with a differential of 19% neutrophils, 10% band forms and 66% lymphocytes, 5% monocytes, and a red blood cell count of 320 cells/mm3, a protein of 120 mg/dl and a glucose of 34 mg/dl. The serum sodium value dropped from 134 mg/dl to 120 mg/dl. A repeat chest radiograph also revealed the development of right middle lobe pneumonia. Because of the tentative diagnosis of bacterial meningitis clindamycin and gentamicin were discontinued, and she was treated with cefotaxime and vancomycin for presumed bacterial meningitis. A second lumbar puncture was performed 2 days later because of persistent fevers, irritability and mental status changes. Her CSF WBC was 1135/mm3 with 35% neutrophils, 18% band forms, 41% lymphocytes, 6% monocytes and 765 red blood cells/mm3. The CSF protein value at that time had risen to 369 mg/dl and the CSF glucose was 37 mg/dl. On Hospital Day 10 the initial wound culture grew a Gram-negative coccobacillus preliminarily identified as nontypable Haemophilus influenzae. On Hospital Day 13 the initial CSF culture was growing the same Gram-negative organisms and these were found to be not H. influenzae but F. tularensis. Antimicrobial therapy was changed to gentamicin (6 mg/kg/day) and doxycycline (4 mg/kg/day). The identification of the organisms was confirmed by the State Health Department and the Centers for Disease Control and Prevention.
KH improved markedly during the next 48 h with stabilization of serum electrolytes and improved mental status. After 6 days of gentamicin and doxycycline therapy a repeat lumbar puncture revealed a CSF WBC of 265/mm3 with 2% neutrophils, 89% lymphocytes and 9% macrophages, a red blood cell count of 265/mm3 with a CSF protein concentration of 179 mg/dl and glucose of 40 mg/dl. Before her discharge home on Hospital Day 21 the tularemia titers were positive at 1:320. Other than the initial CSF all subsequent CSF cultures remained negative. She was treated with a 10-day course of intravenous gentamicin and a 3-week course of oral doxycycline. Physical examination at follow-up 6 weeks later was normal with the exception of a small crusted lesion on the vertex of her scalp consistent with a healing tick bite.
Discussion. Meningitis caused by F. tularensis is not commonly encountered. In a series of publications from our institution involving 105 pediatric patients with tularemia since 1978, this is the first known case.1-3 Of the cases of purulent meningitis caused by F. tularensis in the literature, only 5 other cases have been described in patients ≤16 years of age4-8(Table 1) with an additional 7 cases occurring in older patients.9-15
There have been an additional 16 children described with "tularemic meningism" in a series of 28 patients.16, 17 These patients appear to have a different disease process because they were part of a short epidemic outbreak of tularemia and had only transient involvement of the central nervous system. Patient symptoms consisted mostly of a stiff neck, positive Kernig and Brudzinski signs and a severe headache that would improve after a lumbar puncture. In most cases the lumbar puncture demonstrated increased intracranial pressure without marked inflammatory signs. This transient involvement usually occurred in the early days of the disease, regressed completely and did not change the outcome of the patients.
The CSF findings in tularemic meningitis are very nonspecific. An elevated white blood cell count with a preponderance of mononuclear leukocytes, an elevated protein value and hypoglycorrhachia are usually demonstrated. This presentation can cause clinicians confusion about potential etiologic agents for their patient's disease, and many of the patients have been assumed to have meningitis caused by Mycobacterium tuberculosis, Listeria monocytogenes or a virus.4, 8, 10, 15 Meningitis caused by any of these organisms may indeed elicit a very early response with polymorphonuclear cells in the CSF which quickly change to mononuclear leukocytes. It has been speculated that the mononuclear predominance found in tularemic meningitis is a result of the intracellular nature of this organism and specific host resistance would depend mainly on cell-mediated immunity to combat the infection.8In vitro studies have demonstrated that there is uptake of F. tularensis by mononuclear cells and not by polymorphonuclear leukocytes in the absence of demonstrable antibody.18 Because specific antibodies to F. tularensis do not usually appear until ∼2 weeks after infection,19 it is possible that the shift to mononuclear cells is therefore a natural defense mechanism.
Tularemic meningitis appears to arise from hematogenous spread. This is supported by the fact that the majority of patients have multisystem involvement at the time they are diagnosed with meningitis.4-15 Oropharyngeal or ulceroglandular disease with pneumonia has been described in all of the children described to date including our own (Table 1). Our patient's presentation with ulceroglandular disease made us suspect tularemia as an initial diagnosis but we were misled by the erroneous preliminary culture result from the lesion of H. influenzae.
It is of interest to note the number of individuals who have been described with oropharyngeal involvement. This form accounts for ≤4% of all reported cases of tularemia1, 2 but has been present in 67% of the cases of meningitis in patients ≤16 years of age. This would certainly suggest consumption of poorly cooked meats as a source of infection, but the dietary histories for these individuals were not reported.4, 5, 7, 8 Most of the patients described to date have had animal contact as their major risk factor for the development of their disease. It was only recently that the first adult was believed to have contracted tularemic meningitis as a result of a tick bite.14 Although the child presented in our case had no documented history of tick bite or other animal contact, a tick was probably the source for her infection. In Arkansas ticks remain the most common source of exposure, and most ulceroglandular forms of tularemia in children involve the cervical lymph nodes because of tick bites in the scalp.1, 2
Streptomycin, gentamicin, chloramphenicol and third generation cephalosporins have been used for the treatment of tularemia. In vitro studies initially demonstrated promise for the use of third generation cephalosporins.20 Clinical experience, however, has demonstrated a failure of these agents to successfully treat F. tularensis outside of the central nervous system,21 and our current case clearly demonstrated a failure in a case involving the central nervous system. Two children and three adults have been reported to survive from tularemic meningitis.7, 8, 13-15 Two patients received a 10- to 14 day course of streptomycin and chloramphenicol,8, 13 whereas the remaining three received other regimens. A 13-month-old responded to a 10-day course of gentamicin but relapsed 5 days later. During his second hospitalization it was determined that the initial illness was indeed tularemia, and the child was then treated successfully with a 10-day course of chloramphenicol.7 A 60-year-old man received chloramphenicol and streptomycin for 3 days followed by an additional 7 days of streptomycin. Six days later he relapsed and received successful treatment with a 1-week course of streptomycin followed by tetracycline for 2 weeks.14 The last patient, a 64-year-old, received streptomycin for 9 days as part of antituberculous therapy before the diagnosis of tularemic meningitis. He was then treated with chloramphenicol for 10 days followed by an additional 7-day course of rifampin and gentamicin.15
We elected to treat our patient with gentamicin instead of streptomycin. Our previous experience with the successful treatment of tularemia with gentamicin outside of the central nervous system allowed us to feel comfortable with this approach.3 The risk of relapse with the use of streptomycin or gentamicin alone indicated the use of a second agent. Because of our previous experience with central nervous system disease and ehrlichiosis, doxycycline was chosen.22 The patient did well with this approach and successfully completed therapy without incident.
Benjamin L. Rodgers, M.D.
Robin P. Duffield, M.D.
Toby Taylor, M.D.
Richard F. Jacobs, M.D.
Gordon E. Schutze, M.D.
University of Arkansas for Medical Sciences; Arkansas Children's Hospital; Little Rock, AR
1. Jacobs RF, Narain JP. Tularemia in children. Pediatr Infect Dis J 1983;2:487-91.
2. Jacobs RF, Condrey YM, Yamauchi T. Tularemia in adults and children: a changing presentation. Pediatrics 1985;76:818-22.
3. Cross JT, Schutze GE, Jacobs RF. Treatment of tularemia with gentamicin in pediatric patients. Pediatr Infect Dis J 1995;14:151-2.
4. Pund ER, Hatcher MB. Tularemic meningitis
: report of case with post mortem observations. Ann Intern Med 1937;10:1390-8.
5. David JK Jr., Owens JN Jr. Tularemic meningitis
: report of a case and summary of previously reported cases. Am J Dis Child 1944;67:44-51.
6. Fields WS. Tularemic meningitis
: report of a case. Arch Neurol Psychiatry 1949;61:422-9.
7. Lovell VM, Cho CT, Lindsey NJ, Nelson PL. Francisella tularensis meningitis
: a rare clinical entity. J Infect Dis 1986;154:916-8.
8. Harper JL, Florman AL. Tularemic meningitis
in a child with mononuclear pleocytosis. Pediatr Infect Dis J 1986;5:595-7.
9. Bryant AR, Hirsch EF. Tularemic leptomeningitis: report of a case. Arch Pathol 1931;12:917-23.
10. Hartman FW. Tularemic encephalitis: pathology of acute tularemic with brain involvement and coexisting tuberculosis. Am J Pathol 1932;8:57-61.
11. Haizlip JO, O'Neil AE. A case of meningitis
due to bacterium tularense. JAMA 1931;97:704-5.
12. Stuart BM, Pullen RL. Tularemic meningitis
: review of the literature and report of a case with postmortem observations. Arch Intern Med 1945;76:163-6.
13. Alfes JC, Ayers LW. Acute bacterial meningitis
caused by Francisella tularensis.
Pediatr Infect Dis J 1990;9:300-1.
14. Hutton JP, Everett ED. Response of tularemic meningitis
to antimicrobial therapy. South Med J 1985;78:189-90.
15. Hill B, Sandstrom G, Schroder S, Franzen C, Tarnvik A. A case of tularemia meningitis
in Sweden. Scand J Infect Dis 1990;22:95-9.
16. Glass GBJ. An epidemic of tularemia transmitted by insects in settlements of deportation, Asino and Jaja, Siberia, Soviet Russia: report of 121 cases. Am J Med 1948;216:411-24.
17. Glass GBJ. Tularemic meningism and serous meningitis
: a report of twenty-eight cases observed during an epidemic of tularemia transmitted by insects in settlements of deportation Asino and Jaja, Siberia, USSR. Med Clin North Am 1948;32:769-78.
18. Proctor RA, White JD, Ayala E, Canonico PG. Phagocytosis of Francisella tularensis
by Rhesus monkey peripheral leukocytes. Infect Immun 1975;11:146-51.
19. Koskela P, Salminen A. Humoral immunity against Francisella tularensis
after natural infection. J Clin Microbiol 1985;22:973-9.
20. Baker CN, Hollis DG, Thornsberry C. Antimicrobial susceptibility testing of Francisella tularensis
with modified a Mueller-Hinton broth. J Clin Microbiol 1985;22:212-15.
21. Cross JT, Jacobs RF. Tularemia: treatment failures with outpatient use of ceftriaxone. Clin Infect Dis 1993;17:976-80.
22. Schutze GE, Jacobs RF. Human monocytic ehrlichiosis in children. Pediatrics 1997;100:e10.