French Cliques: Feline Bartonellosis in Nancy

ABSTRACT & COMMENTARY

Synopsis: An epidemological survey found B. henselae and B. clarridgeiae present in stray cat populations in urban France.

Source: Heller R, et al. J Clin Micro 1997;35:1327-1331.

In the process of examining the prevalence of various infections circulating within the free-ranging cat population of Nancy, France, Heller and colleagues looked for different Bartonella strains in blood specimens obtained from 94 stray cats. Stray cats apparently form structured colonies within urban areas, and these animals had formed a total of 10 cliques. Both cultures on blood and liquid media, as well as molecular DNA sequencing, were performed.

Fifty (53%) of the cats had a positive blood culture for Bartonella. Younger cats (less than 1 year of age) were 50% more likely to be bacteremic compared with older cats. Ten percent of the cats were seropositive for feline immunodeficiency virus, but none were infected with feline leukemia virus. A single Bartonella strain was found circulating within each of six urban cat colonies, two strains were found within each of three cat colonies, and three strains were isolated from one or more of the cats in one colony.

Using molecular techniques, a total of three Bartonella strains were identified, including two separate strains of B. henselae. The third organism was identified as that of a newly recognized species, Bartonella clarridgeiae. Quite unexpectedly, this new strain was found in 30% of the infected cats. While 15 isolates of B. clarridgeiae were recovered using the BACTEC system, they generally either failed to grow or grew poorly on solid media, which may explain how this organism eluded earlier detection.

COMMENT BY CAROL A. KEMPER, MD

Bacillary angiomatosis (BA) was first described by Clay Cockerell in 1987 as a distinct, cutaneous proliferative vascular disorder in patients with HIV infection.1 Hepatic (termed hepatic peliosis by some authors), splenic, and bone marrow involvement were subsequently described in both HIV-infected patients as well as non-HIV-infected patients with profound immunodeficiency due to organ transplantation2 or immunosuppresive chemotherapies.3 Cutaneous BA has also been infrequently reported in patients with apparently normal immune function.4

Enhanced culture techniques, using either lysis centrifugation or BACTEC systems and subcultures to brain-heart infusion media or sheep blood agar and molecular-based techniques, showed that BA and hepatic peliosis were the result of infection due to either of two newly identified organisms—B. henselae and Bartonella quintana (formerly Rochalimaea henselae and Rochalimaea quintana).5,6

Remarkably, B. henselae was subsequently identified as the etiologic agent of human cat scratch disease, and B. quintana was identified as the etiologic agent of epidemic trench fever. B. quintana has also been found to cause bacteremia, osteomyelitis, and, rarely, endocarditis, especially in individuals who live in poor socioeconomic conditions, are homeless, and/or are chronic alcoholics. Another strain of Bartonella, B. elizabethae, has also been found to be a rare cause of bacteremia and endocarditis. The primary reservoir for human infection due to B. henselae is believed to be cats (or possibly their fleas), up to 70% of which have been found to be bacteremic with the organism in at least one survey. In contrast, the vector for B. quintana infection is believed to be the human louse. Although human cases of B. clarridgeiae have not been clearly identified, Heller and colleagues found a remarkably high rate of infection in stray cats in urban France which could represent a reservoir for human infection.

Case-controlled studies suggest that HIV-infected patients with profoundly low CD4 cell counts (< 100 cells/mm3) are at greatest risk for BA.7 Such patients may present with isolated cutaneous and subcutaneous disease, chronic lymphadenopathy, and/or end-organ involvement. Anemia and thrombocytopenia are common findings. B. henselae can also cause an aseptic meningitis or encephalitis in HIV-infected patients, as well as in non-HIV-infected patients with cat scratch disease.

Both B. henselae and B. quintana are remarkably similar in many respects to Bartonella bacilliformis, the etiologic agent of Oroyo fever and verruga peruana. These two diseases were forever immortalized by Daniel Carrion, a Peruvian medical student who, in his efforts to demonstrate that these diseases were different manifestations of the same infection, was zealous enough to inoculate himself with the verrugas material obtained from the cutaneous lesion of a patient and subsequently died from acute Oroya fever. Oroya fever is a febrile illness, characterized by hemolytic anemia, thrombocytopenia, lymphadenopathy, splenomegaly, and, at its most severe, encephalitis and coma, which was responsible for thousands of deaths during the construction of the Lima-Oroya railway in the 1870s. Those who recover from Oroya fever are at risk for developing the cutaneous verrugas, which have the identical hyperproliferative vascular histology as BA lesions in patients with AIDS.

Presumably all of these organisms have the capacity to cause bacteremia, but the specific histological manifestations of infection and the degree of dissemination and end-organ involvement are dependent on host immune factors. While chronic regional necrotizing lymphadenitis is often seen in persons with cat scratch disease, patients with more profound immune deficiencies, such as those with AIDS or transplant recipients, are at greater risk for widespread dissemination and end-organ disease with histological evidence of BA and hepatic peliosis. The lack of intact immunity in these patients has literally "unmasked" the ability of these organisms to cause proliferative endovascular lesions, similar to that seen in bartonellosis. The molecular basis for the endovascular proliferative capacity of this family of organisms has not yet been elucidated.

References

1. Cockerell CJ, et al. Lancet 1987;ii:654-656.

2. Kemper CA, et al. Am J Med 1990;89:216-222.

3. Myers SA, et al. J Pediatr 1992;121:574-578.

4. Tappero JW, et al. JAMA 1993;269:770-775.

5. Regnery RL, et al. J Clin Microbiol 1992;30:265-274.

6. Koehler JE, et al. N Engl J Med 1992;327:1625-1631.

7. Mohle-Boetani JC, et al. Clin Infect Dis 1996;22: 794-800.