Bartonella henselae as a Cause of Prolonged Fever and Fever of Unknown Origin in Children

Richard F. Jacobs 0 1 2 Gordon E. Schutze 0 1 2 0 Clinical Infectious Diseases 1998;26:80-4 q 1998 by The University of Chicago. All rights reserved. 1058-4838/98/2601-0011$03.00 1 Received 12 March 1997; revised 4 August 1997. The protocol was reviewed and approved by the Institutional Review Board of the University of Arkansas for Medical Sciences. Financial support: This work was supported by the Horace C. Cabe Founda- tion and the Bates-Wheeler estate. Infectious Diseases, Arkansas Children's Hospital , 800 Marshall Street, Little Rock, Arkansas 72202-3591 2 From the Division of Pediatric Infectious Diseases, Department of Pediatrics, Arkansas Children's Hospital, University of Arkansas for Medical Sciences , Little Rock, Arkansas A prospective evaluation of 146 children with fever of unknown origin (FUO) and prolonged fever was performed from 1990 to 1996. FUO was defined as a documented daily temperature of 387C for at least 14 days without diagnostic signs or symptoms. Prolonged fever was defined as fever for at least 14 days and no diagnosis at the time of referral for evaluation. An established diagnosis was made for 84 (57.5%) of 146 patients. The most common infectious disease diagnoses were Epstein-Barr virus infection (22 [15.1%] of 146), osteomyelitis (14 [9.6%] of 146), bartonellosis (7 [4.8%] of 146), and urinary tract infection (6 [4.1%] of 146). Three of seven patients with confirmed Bartonella henselae infection presented with FUO and no ultrasonographic findings compatible with hepatosplenic involvement; two patients presented with FUO and hepatosplenic involvement. The relatively common finding of acute bartonellosis in this population suggests that FUO and prolonged fever in children are other presentations of infection with B. henselae. - (CSD) has been described in children presenting with persistent or prolonged fever [12, 13]. Recently, Golden [14] described two children with atypical CSD (hepatosplenic CSD) who were febrile for 2.5 and 3 weeks, respectively. Since patients with hepatosplenic CSD can present with persistent fever [14,15], this entity has been considered for inclusion in the differential diagnosis of prolonged fever in children. To date, those patients for whom CSD has been diagnosed and who have presented with FUO or prolonged fever have had hepatosplenic involvement. We describe seven children with the presentation of FUO or prolonged fever in whom infection with B. henselae was confirmed serologically. Two children presented with signs and symptoms consistent with typical CSD, and two children presented with hepatosplenic involvement; however, three children presented with FUO and no clinical or radiographic manifestations of typical or hepatosplenic CSD. Bartonellosis should be considered in the initial evaluation of FUO and prolonged fever in children with specific epidemiological exposures, especially to kittens and/or cats. Methods Since 1990, a prospective database of all children for whom FUO and prolonged fever have been diagnosed has been collected to assess the continued evolution of infectious disease diagnoses. The definition of FUO for this study included documented and recorded daily temperature of 387C (core temperature) for at least 14 days and no diagnostic signs or symptoms of an obvious clinical disease. A diagnosis of prolonged fever required the same temperature criteria plus no diagnosis at the time of referral for evaluation and no previous diagnosis or clinically apparent features of a congenital or acquired immunodeficiency. Patients were classified as having prolonged fever instead of FUO if the subsequent evaluation was indicative of a clinical disease presentation that suggested a specific diagnosis (i.e., regional lymphadenitis plus associated cat exposure CSD). A focused approach utilizing an institutional protocol for evaluation of children with FUO and prolonged fever was employed as described previously [7]. The initial evaluation was carried out in either an inpatient or outpatient setting and usually included determination of complete blood cell count, erythrocyte sedimentation rate (ESR), and hepatic enzyme levels; chest roentgenography; urinalysis and urine culture; blood culture; tuberculin skin test; and obtaining acute-phase serum samples for future use. The most common assay performed was EBV serology because of the finding of EBV infection as the most likely infectious disease diagnosis in a previous study from our institution [7]. All microbiological, virological, and serological studies were individualized on the basis of the initial history of present illness, contact and exposure history, season, physical examination, initial laboratory studies, and radiographic evaluation. All patients were followed up with the use of a temperature and symptom diary by the parents. Temperatures (rectal, oral, or axillary) were taken three times daily with a written numerical record and site of determination record and on any occasion when the parent thought that the child appeared febrile or symptomatic. Signs and symptoms were included in a narrative corresponding to the body temperature and time of day. During hospitalization, temperatures taken by a registered nurse were recorded (to include site of determination) at least once per 8hour shift. On the basis of findings of a previous study in our institution [7], an abdominal ultrasound examination was performed on all children with FUO or prolonged fever and abdominal signs or symptoms, an elevated ESR, or elevated hepatic enzyme levels. All patients with serological evidence of acute infection with B. henselae underwent an abdominal ultrasound examination. Infectious disease diagnoses. Urinary tract infections, cytomegalovirus or enterovirus infections, and blastomycosis were confirmed microbiologically. In cases of osteomyelitis, positive cultures of blood, bone aspirates, joint fluid, or biopsy specimens were considered diagnostic. All other cases of osteomyelitis were confirmed in patients with compatible signs and symptoms for whom a bone scan, CT, MRI, or a combination of these images was positive. EBV infection, bartonellosis, HIV infection, tularemia, and ehrlichiosis were diagnosed serologically by using commercial assays and definitions. EBV serology included an IgM antibody to viral capsid antigen (VCA) (indirect fluorescent antibody test: positive titer, 1:10; Gull Laboratories, Salt Lake City). Acute EBV infection was diagnosed when there was a positive titer of IgM antibody to VCA or a fourfold rise in titers of IgG antibody to VCA (indirect fluorescent antibody test: positive titer, 1:320; Grandbio, Temecula, CA). Evidence of HIV infection was diagnosed for a patient older than 15 months of age when HIV type 1/HIV type 2 serology (EIA, Abbott Laboratories, Abbott Park, IL) was positive. Tularemia was diagnosed with a positive slide and tube agglutination titer (titer in acute-phase serum of 1:160 or fourfold rise in (...truncated)