When future historians examine the numerous events that have contributed to the rapid emergence of knowledge related to infectious diseases during the latter portion of the 20th century, it is probable that research contributions related to infections caused by Bartonella species will be considered of unique medical importance. Although predated by a substantial body of important observations, two manuscripts1,2 published in 1990 in the same issue of the New England Journal of Medicine set the stage for a scientific revolution in our understanding of Bartonella infections. Relman and colleagues used polymerase chain reaction (PCR) to amplify bacterial DNA from lesions of AIDS patients with bacillary angiomatosis.1 When compared with known eubacterial sequences, the uncultured organisms were most closely related to Rochalimaea quintana, the cause of Trench Fever. Independently, Slater and colleagues described the isolation of a fastidious gram-negative organism in blood cultures from immunocompromised and non-immunocompromised febrile patients.2 The organisms described in these two manuscripts were subsequently characterized as B. quintana or a newly defined species B. henselae. During the ensuing nine years, research efforts have begun to elucidate the importance of Bartonella species as veterinary and human pathogens.

In 1993, Brenner and colleagues3 provided evidence to justify reclassification of the four species in the genus Rochalimaea. Previously designated Rochalimaea species were united with the genus Bartonella, and named Bartonella quintana, Bartonella vinsonii, Bartonella henselae, and Bartonella elizabethae. This reclassification removed the family Bartonellaceae from the order Rickettsiales, and resulted in the transfer of these organisms from the family Rickettsiaceae to the family Bartonellaceae. In 1995, Birtles and colleagues4 proposed the unification of the genus Grahamella with the genus Bartonella, resulting in the addition of five Bartonella species: B. talpae, B. peromysci, B. grahamii, B. taylorii, and B. doshiae. Historically, Grahamella species have been characterized as arthropod-transmitted hemotropic gram-negative bacteria of small mammals, fish, and birds, which are of no known pathogenic consequence for larger mammals. The role of rodents as reservoir hosts for Bartonella species is an important area of current investigation. In effect, these reclassifications eliminated the previous genera, Rochalimaea and Grahamella.

In 1992, Schwartzman5 summarized the rapidly expanding clinical spectrum of human illness caused by infections due to Bartonella organisms. In more recent years, Bartonella species have become associated with several distinct clinical syndromes in people: bacillary (pertaining to bacilli or rodlike forms) angiomatosis,6–8 bacillary peliosis (extravasation of blood) hepatis,7,8 relapsing fever with bacteremia,2,9 cat scratch disease (CSD),10–12 endocarditis,13– 15 granulomatous hepatosplenic syndrome,16 retinitis and swelling of the optic nerve,17 osteolytic lesions,18,19 and granulomatous pneumonia, which we reported in a chronically-ill renal transplant patient.20 A role for B. henselae in the pathogenesis of AIDS encephalopathy has also been proposed.21

In 1993, Bartonella infection was associated with human endocarditis for the first time. Within a year, B. quintana, B. elizabethae and B. henselae were shown to cause endocarditis in people, providing one of several examples that illustrate the rapid recognition of new clinical entities associated with human Bartonella infections.13–15 As a result of reviewing these reports, we isolated and characterized a novel Bartonella subspecies from a dog with endocarditis.22,23 Subsequently, we have obtained serologic evidence to support tick transmission of B. vinsonii subsp. berkhoffii to dogs. We have also identified 7 additional cases of endocarditis and a case of granulomatous lymphadenitis associated with B. vinsonii infection in dogs. The medical importance of members of the genus Bartonella as canine pathogens or the role of dogs as a reservoir for human infection awaits additional study.

Bacillary angiomatosis, also called epithelioid angiomatosis, is a vascular proliferative disease of the skin that is characterized by multiple, blood-filled, cystic tumors.6,7 When visceral parenchymal organs are involved, the condition is referred to as bacillary peliosis hepatis, splenic peliosis or systemic bacillary angiomatosis. These conditions, now recognized to be infectious manifestations of B. henselae or B. quintana, have been more frequently documented in immunocompromised individuals, particularly AIDS patients, but have also been reported in immunocompetent individuals. Prior to the AIDS epidemic, bacillary angiomatosis had been reported in association with tuberculosis or advanced cancer, presumably a reflection of co- existing Bartonella infection. In a study from the Greater San Francisco Bay Region of Northern California, B. henselae was isolated from the blood of 7 cats belonging to 4 patients with bacillary angiomatosis.24 An important observation in this study, which has been extended by findings from our laboratory, is that B. henselae bacteremia can persist in naturally- infected cats for periods of at least 21 months.25 Experimentally, we were able to transmit B. henselae via intravenous or intramuscular inoculation of blood, but not via intramuscular inoculation of urine sediment.26 These results strongly suggest that cats should be screened for B. henselae infection prior to use as a blood donor. Although experimental studies following World War I indicated that human urine could successfully transmit B. quintana to people, and anecdotal reports indicate exposure to cat urine may have caused CSD in people, our results fail to support transmission of B. henselae through urine. Considering that current studies have identified 25 to 40% of healthy cats as asymptomatic carriers of B. henselae, the potential of human exposure to infected cat urine in litter boxes could be substantial.

Persistent or relapsing fever, accompanied by malaise, anorexia, and weight loss has been documented frequently in immunocompromised individuals with Bartonella infections.2,7,20 Acute febrile illness with bacteremia, bacillary angiomatosis, bacillary splenitis and other chronic disease manifestations associated with Bartonella infections are being recognized with increasing frequency in immunocompetent people.9,26 Persistent B. henselae bacteremia, of approximately 3 months duration, was identified in an immunocompetent man, who was potentially infected by tick exposure.27 This report suggests that immunocompetent people can experience prolonged illness with persistent bacteremia. Recurrent bacillary angiomatosis due to presumed therapeutically refractory B. quintana infection has been reported in an AIDS patient.28

For nearly a century regional lymphadenopathy has been associated with animal contact, particularly cat scratches.5,29 Over the years, numerous microorganisms have been implicated as the cause of CSD. In 1983, small, argyrophilic (easily impregnated with silver), gram-negative, pleomorphic bacteria were seen within blood vessel walls and macrophages in lymph nodes of patients with CSD.30 In 1988, a bacteria,31 later designated Afipia felis,32 was cultured from lymph nodes of CSD patients. In the same year, Cockerell and colleagues proposed a possible association between epithelioid angiomatosis and CSD in a letter to Lancet.33 In 1992, Regnery and colleagues at the Centers for Disease Control,10 identified seroreactivity to B. henselae antigens in 88% of 41 patients with suspected CSD compared to 3% of controls. Similarly, a case-controlled Connecticut study of CSD patients and their cats34 identified a strong association with cats 12 months of age or younger, a history of a scratch or bite, contact with fleas, and seroreactivity to B. henselae antigen. Additional support that B. henselae is the predominant cause of CSD was provided when DNA was amplified from lymph node samples of 21 of 25 (84%) patients with suspected CSD, using a polymerase chain reaction assay.12 A similar study from Sweden identified B. henselae DNA, but failed to identify A. felis DNA, in a large number of patients with suspected CSD.35 We have cultured Bartonella species from 17 of 19 cats owned by 14 patients with CSD.25 In summary, recent studies indicate that B. henselae is the predominant, but not the sole cause of CSD. In 1995, Clarridge et al. isolated a novel Bartonella species,36 which was named B. clarridgeiae,37 from a cat belonging to a patient infected with HIV from whom B. henselae was isolated. Our studies have implicated B. clarridgeiae as a cause of inoculation papules, fever and regional lymphadenopathy (CSD) in 3 people. Based upon PCR- RFLP analysis of the 16S and the 16S-23S rRNA genes, approximately 10% of the isolates obtained in our laboratory from cats are B. clarridgeiae.

Historically, atypical manifestations of CSD have included tonsillitis, encephalitis, cerebral arthritis, transverse myelitis, granulomatous hepatitis and/or splenitis, osteolysis, pneumonia, pleural effusion, and thrombocytopenic purpura. Previously, Bartonella infection would not have been considered a likely differential diagnosis by the physician in patients lacking a history of lymphadenopathy or animal contact. As evidenced by reports in the past four years, the spectrum of human disease associated with the genus Bartonella is likely to expand, requiring periodic reassessment as new information becomes available.

Whether members of the genus Bartonella are pathogenic for cats or contribute to previously described instances of argyrophilic bacteria in lymph nodes of cats with persistent lymphadenopathy38 or to peliosis hepatis39 remains to be determined. Although B. henselae bacteremia can be documented in 25 to 41% of healthy cats,24,25 we have observed self-limiting febrile illness of 48 to 72 hours duration, mild to moderate transient anemia, and transient neurologic dysfunction in cats experimentally infected with B. henselae by blood transfusion.26 Self-limiting fever also occurred in 4 B. henselae bacteremic cats from the same household following minor surgical procedures. Due to the high percentage of chronically bacteremic healthy cats in the United States, establishing a cause and effect relationship between disease manifestations and bacteremia in cats will require epidemiologic analysis.

Bartonella henselae and B. bacilliformis can be found within erythrocytes, whereas B. quintana maintains an epicellular location on the erythrocyte.40 Cell lysis, using a lysis centrifugation technique, greatly facilitates bacterial isolation from blood. Although organisms within the genus Bartonella are fastidious and slow-growing, they can be cultured successfully with agar plates containing 5% defibrinated rabbit or sheep blood, that are maintained at 35oC in a high humidity chamber with a 5% CO2 concentration. In our experience, bacterial colonies may not be visible until 10 to 56 days after inoculation of the agar plate. Because members of the order Rickettsiales are cultivable only in their host cells or living tissues, cultivation of Bartonella species on bacteriologic media, in conjunction with DNA divergence studies, provided additional justification for removal of members of the genus Bartonella from the order Rickettsiales. Bartonella are small, curved, gram-negative rods. In tissues, the organisms stain positively with silver stains such as the Warthin-Starry stain.

The diagnosis of Bartonella infection should be confirmed by culturing the organism from blood or tissues such as lymph node or heart valve or by amplifying DNA from tissues using PCR. Seroconversion, detected by IFA or ELISA techniques, can be documented in people with acute disease and in cats following experimental infection. The kinetics of the serologic response to B. henselae antigens in chronically-infected experimental cats is highly variable in degree and duration.26 A seroepidemiologic survey, incorporating 577 samples from throughout North America identified an overall prevalence of 28%, with prevalence rates ranging from a low of 4–7% in the Midwest and great plains region to 60% in the southeast.41 High seroprevalence rates correlate with warm, humid climates conducive for the environmental maintenance of cat fleas,41,42 which have been shown to be capable of transmitting B. henselae from cat to cat.43 A seroepidemiologic study of 592 cats in Baltimore, MD identified an overall seroprevalence rate of 14.7%, with a much higher prevalence in feral cats (44.4%) as compared to pet cats donated to the City Municipal Animal Shelter (12.2%).44 Although providing useful epidemiologic data, serologic test results may be of limited clinical utility for several reasons. We have been unable to detect B. henselae-specific antibodies in some bacteremic cats and have been unable to culture Bartonella from some cats in which antibodies are detectable. Negative blood cultures obtained from cats seroreactive to B. henselae antigen may reflect low level bacteremia or the timing of the blood culture, since experimentally-infected cats experience a relapsing bacteremia. Numerous naturally-infected cats are persistently bacteremic, generally in conjunction with low antibody titers.25 In our experience, high antibody titers generally correlate with positive blood cultures. The extent of serologic cross reactivity to Bartonella species requires additional clarification.45 In unpublished studies, we have demonstrated co-infection with B. henselae and B. clarridgeiae in cats.

Because of disparate results among studies and an overall lack of microbiologic data in clinical therapeutic trails, numerous issues related to treatment of Bartonella infection remain controversial. In contrast to the apparent lack of response to antimicrobial treatment in human CSD patients,46,47 bacillary angiomatosis, parenchymal bacillary peliosis, and acute Bartonella bacteremia appear to respond to antimicrobial treatment, even in immunocompromised individuals.48 Doxycycline, erythromycin and rifampin are recommended antibiotics,49 but clinical improvement has been reported following the use of penicillin, gentamicin, ceftriaxone, ciprofloxacin, and azithromycin. Treatment for 2 weeks in immunocompetent individuals and 6 weeks in immunocompromised people is generally recommended. Relapses, associated with bacteremia, have been reported in immunocompromised people despite treatment for 6 weeks. Antimicrobial efficacy has not been established for any antibiotic for eliminating B. henselae bacteremia in cats. Results from our laboratory and others indicate incomplete treatment responses in cats treated for 2 or 4 weeks with doxycycline or enrofloxacin.50,51 In contrast to other reports,50,51 cats experimentally-infected with B. henselae by blood transfusion do not develop protective immunity following homologous or heterologous challenge with B. clarridgeiae. These differences in experimental results are most likely due to alterations in virulence, induced during in vitro culture of the organism.

Based upon recent advances in our knowledge of the zoonotic potential of members of the genus Bartonella, the designations cat scratch disease and cat scratch fever may be most appropriate when considering human disease manifestations from a historical perspective. Because cat scratch disease generally denotes a self-limiting illness characterized by fever and lymphadenopathy and because the recognized spectrum of human disease manifestations associated with Bartonella infections (which may not include fever or lymphadenopathy) has expanded considerably in recent years, it is becoming obvious that the designation CSD lacks clinical, microbiologic and zoonotic utility. Although cats are a major reservoir for B. henselae and potentially B. clarridgeiae, some patients deny the possibility of a cat scratch or bite wound, or indicate no contact with cats. Transmission from environmental sources or other animal hosts is probable and the more inclusive term bartonellosis may facilitate enhanced future understanding of diseases caused by members of the genus Bartonellaceae.

Although recent research findings have substantially improved our understanding of the clinical, microbiologic and zoonotic aspects of diseases caused by Bartonella species, the exact mode of transmission, the relative role of various insect vectors such as fleas and ticks, the identification of potential reservoir hosts, and the spectrum of animal and human illnesses caused by these organisms remains largely undetermined. For example, although it is well established that the human body louse transmits B. quintana, the reservoir and mode of transmission that results in bacillary angiomatosis in the United States has not been established. The pathogenic potential of these organisms appears to be of considerable importance in dogs as well as immunocompromised and immunocompetent people.

I would like to acknowledge contributions by Dorsey Kordick, Brandee Pappalardo and Barbara Hegarty for the generation of data from my laboratory related to Bartonella infection in cats and dogs, and the Bayer Corporation, SmithKline Beecham Animal Health, Pfizer Animal Health, Intervet Inc., Heska Corporation, the Triangle Cat Fanciers, and the American Veterinary Medical Foundation for funding our work. I would also like to gratefully acknowledge the collaboration of Dr. Ken Wilson, The Veterans Affairs Medical Center and Division of Infectious Diseases, Duke University Medical Center, Drs. Don Brenner and Rus Regnery at the Centers for Disease Control and Dr. Ted Hadfield at the Armed Forces Institute of Pathology.

Key Words: Infectious diseases, Bartonella References, as cited, are available upon request to those providing a self-addressed, stamped envelope.

Recent Reviews

1. Regnery RL, Childs JE, Koehler JE. Infections associated with Bartonella species in persons infected with the human immunodeficiency virus. Clin Infect Dis 1995;21:94–98.

2. Maurin M, Raoult D. Bartonella (Rochalimaea) quintana infections. Clin Microbiol Rev 1996;9:273–292.

3. Anderson B, Neuman M. The genus Bartonella as emerging human pathogens. Clin Microbiol Rev. April, 1997.

Edward B. Breitschwerdt, DVM, Diplomate, ACVIM
Professor Veterinary Medicine
North Carolina State Veterinary School

Home