In some rickettsioses, the inoculation eschar is the site of an intense rickettsial multiplication and thus is the preferred biopsy specimen for studying the pathologic features of Rickettsia
-induced infection and for detecting the bacteria by immunohistologic tests as well as for carrying out isolation procedures or genomic detection. As demonstrated before for many rickettsioses, the cutaneous damages at the inoculation site were histologically dominated by vasculitis and necrotic features (11
). Rickettsial invasion of endothelial cells probably represents the first step of infection (7
). Subsequent damage to the endothelium is followed by endothelial-cell activation and perivascular infiltration of lymphocytes, macrophages, and polymorphonuclear leukocytes, resulting in inflammatory vasculitis of dermal vessels, the histopathologic hallmark of rickettsial disease, and possibly thrombosis (23
). However, in contrast with the other rickettsial diseases that are characterized by perivascular infiltration of T cells and macrophages, with some B lymphocytes and few neutrophils (10,24–26
), the cutaneous damage of ATBF show vasculitis with polymorphonuclear leukocyte–rich inflammation. The predominance of neutrophils in inflammatory infiltrates may explain the importance of the local inflammation clinically observed at the site of inoculation, accompanied by the regional lymphadenitis.
By using immunohistochemical techniques, we demonstrated R. africae in cutaneous biopsy specimens from patients with ATBF. In accord with its obligate intracellular location, no extracellular organisms were observed in cutaneous biopsy specimens. Few bacterial antigens were found in vascular and perivascular locations within the cytoplasm of endothelial and inflammatory cells. In spite of the small amount of antigens detected, the inflammatory and necrotic cutaneous damage was histologically extensive. Our data indicate that R. africae replicates poorly in human tissues, likely because of its mild pathogenicity and the strong innate immune response. The local destruction of bacteria by the inflammatory reaction may explain the benign outcome of the disease.
In our study, immunohistochemical techniques had a sensitivity of 75%, a specificity of 100%, and a positive predictive value of 100%. Serologic tests, the most widely used diagnostic method for rickettsioses, had a sensitivity of 56% in early samples (12
). Although our data suggested that immunohistochemical analysis might be more sensitive than serology in early samples, statistical analysis showed no significant difference between the 2 techniques. Regarding diagnostic methods applicable to skin biopsy specimens, immunochemical techniques were also more sensitive than culture (41%), which is restricted to specialized laboratories equipped with biohazard facilities; such techniques were also more sensitive than regular PCR (47%) (27
). Moreover, the techniques exhibited a sensitivity similar to that of nested PCR (73.5%), previously found to be the most efficient diagnostic technique for spotted fever rickettsioses (27
As Mediterranean spotted fever caused by R. conorii
is endemic in the same regions of Africa as tick-bite fever, differentiation of the 2 syndromes by characterization of their etiologic agents may be useful for diagnostic and epidemiologic studies (4,28,29
). The polymorphonuclear leukocyte–rich vasculitis, which dominates the histologic features of the inoculation eschar during ATBF, could suggest the diagnosis of this rickettsiosis but is not specific. The usual method for the diagnosis of rickettsioses is serologic testing. However, serologic cross-reactions are common among the rickettsiae in the spotted fever group, particularly between R. africae
and R. conorii
). Monoclonal antibodies had been developed to R. africae
for use in assays to distinguish between R. conorii
and R. africae
in culture and skin biopsy samples (20
). In this study, we used a monoclonal antibody produced in our laboratory to distinguish ATBF from Mediterranean spotted fever at the histologic level by immunohistologic methods. We presented the pathologic description of the first series of inoculation eschars from skin biopsy specimens of patients with ATBF. We showed that cutaneous damage is dominated by vasculitis, thrombosis, cutaneous necrosis, and a polymorphonuclear leukocyte–rich inflammatory reaction. Immunohistochemical detection of rickettsial antigens may be useful in diagnosing ATBF. Pathologists should now consider ATBF, a recently rediscovered rickettsiosis, during histologic analysis of inoculation eschars, especially in patients with a recent stay in sub-Saharan Africa.