Rodent-Associated Bartonella Febrile Illness, Southwestern United States

Jonathan Iralu; Ying Bai; Larry Crook; Bruce Tempest; Gary Simpson; Taylor McKenzie; Frederick Koster


Emerging Infectious Diseases. 2006;12(7):1081-1086. 

In This Article

Materials and Methods

From July 1993 to June 2001, 114 patients 15–78 years of age were referred by their physicians for assistance in diagnosing a febrile illness with a duration <12 days at the time of admission. One hundred patients were hospitalized in New Mexico, 10 in Arizona, and 4 in Colorado. All patients were hospitalized on the basis of the attending physician's decision concerning severity of illness, the possibility of hantavirus infection in the prodrome phase, and the need for diagnostic studies, supportive care, and presumptive antimicrobial-drug therapy. At the time specimens were collected, results of conventional microbiologic assays and diagnostic serologic analysis were negative or unavailable.

Patients were divided into 4 clinical groups according to conventional diagnostic results ( Table 1 ). Seventy-six patients (group A) had an acute undifferentiated febrile illness without pulmonary, cardiac, or renal manifestations. Twelve patients (group B) had bacterial lobar pneumonia (11 patients) or acute respiratory distress syndrome (1 patient) diagnosed by typical signs and symptoms, hypoxemia, pulmonary infiltrates, and prompt clinical response to β-lactam antimicrobial drugs.[16,17] Twelve patients (group C) had hantavirus cardiopulmonary syndrome diagnosed by strip immunoblot serology[18] and reverse transcription–polymerase chain reaction (RT-PCR) of serum.[19] Fourteen patients (group D) had an acute febrile syndrome without pulmonary manifestations and with a diagnosis established by conventional blood culture, serology, or PCR; this group included 3 patients with Escherichia coli sepsis, 2 with E. coli pyelonephritis, 3 with Rocky Mountain spotted fever, 1 with acute Staphylococcus aureus aortic valve endocarditis, 1 with bubonic plague, 1 with acute Q fever, 1 with parvovirus infection, 1 with acute rheumatic fever, and 1 with acute lupus erythematosis. All patients (except those in group D) had at least 2 negative blood cultures, negative spinal fluid cultures and cytometrics when appropriate, negative hantavirus serologic results (except group C), and negative serologic results for plague, tularemia, Q fever, spotted fever, and Ehrlichia species ordered at the discretion of the attending physician. Except for hypertension (5 patients) and chronic alcoholism (12 patients), no patient had underlying disease such as diabetes, malignancy, or HIV infection. The charts were reviewed retrospectively by the investigators. The study was approved by the institutional review boards of the University of New Mexico and the Navajo Nation.

Citrated and clotted blood was collected within 24 hours of admission from 90 patients (acute-phase sample), 7–42 days after admission from 10 patients, and at admission and during convalescence from 14 patients (all in group A). Plasma was immediately frozen at –80ºC. An IFA was performed as previously described.[13] All antigens were prepared at the Bacterial Zoonoses Branch, Centers for Disease Control and Prevention (CDC), Fort Collins, Colorado.

Vero E6 monolayers were infected separately with 1 of 9 strains of Bartonella: 3 strains (B. quintana, B. henselae, and B. elizabethae) were isolated from humans and 6 strains were isolated from the meadow vole (Microtus pennsylvanicus), white-throated woodrat (N. albigula), deer mouse (Peromyscus maniculatus), cotton rat (Sigmodon hispidus), Ord kangaroo rat (Dipodomys ordi), and rock squirrel (Spermophilus variegatus). Plasma was diluted 1:32 in phosphate-buffered saline, placed in antigen-containing wells, incubated at 37ºC for 30 minutes, washed, and incubated at 37ºC for 30 minutes with rabbit antihuman immunoglobulin (Ig) conjugated with fluorescein isothiocyanate. Positive samples were then tested in serial 2-fold dilutions on monolayers infected with 1 of 9 Bartonella strains. Mouse hyperimmune sera were produced by injection of BALB/c mice with the same Bartonella strains that were used for the antigen preparations. These sera were used as IFA-positive controls (titers >1,000 in each assay). Results were tabulated without knowledge of the patient's clinical status.


Comments on Medscape are moderated and should be professional in tone and on topic. You must declare any conflicts of interest related to your comments and responses. Please see our Commenting Guide for further information. We reserve the right to remove posts at our sole discretion.
Post as: