The rickettsiae to some extent resemble small bacteria but are not stained with Gram stain and cannot be cultured on artificial media. These organisms are spread only by insect vectors that have fed on blood from a patient with the disease, not from personal contact with a patient. Blood cultures can sometimes isolate and identify rickettsiae, especially in Rocky Mountain spotted fever. The blood specimens should be frozen and sent to the laboratory packed in dry ice. Since artificial culture media are not available, chick embryo or live animal inoculation must be done. Very few laboratories will perform rickettsial culture. Therefore, serologic tests or other procedures by far overshadow culture as diagnostic aids. The most commonly used procedure is the Weil-Felix reaction. This test takes advantage of the fact that certain rickettsial diseases produce antibodies that also react (or cross-react) with antigen contained in certain strains of Proteus bacteria. These Proteus groups are called OX-19 and OX-K. Titers of 1:80 are suspicious, and titers of 1:160 are definitely significant. Antibodies appear 7-10 days after onset of illness. Rickettsial diseases that may be diagnosed by means of the Weil-Felix reaction are the following:

Rickettsial Diseases

Unfortunately, there are serious limitations to the Weil-Felix test. First, there are a fairly large number of borderline false positive results as well as occasional outright false positives. Since the Weil-Felix reaction depends on Proteus antigen, urinary tract infection by Proteus should be ruled out if the Weil-Felix test result is positive. Second, about two thirds of patients with Rocky Mountain spotted fever or Brill’s disease (recrudescent typhus) have false negative reactions. In these two diseases, other serologic tests (e.g., the microimmunofluorescence procedure) are preferred to the Weil-Felix test.

Rocky Mountain spotted fever

Rocky Mountain spotted fever occurs in spring and summer, predominantly in the eastern two thirds of the United States (the area east of the Rocky Mountains). The disease is transmitted through the bite of an infected tick, with an incubation period of 3-12 days. Fever, severe frontal headache, severe myalgias, and a macular or maculopapular skin rash that develops on the third to fifth day and eventually develops a petechial component are the most characteristic clinical findings, with the rash typically involving the palms and soles. Nausea, vomiting, and diarrhea are reported in about 60% of patients and abdominal pain is reported in about 35%. The WBC count is usually normal, but thrombocytopenia is said to be common. In severe cases disseminated intravascular coagulation may develop. Serologic tests include the Weil-Felix test, CF test, microimmunofluorescence (MIF) test, latex agglutination (LA) test, and tissue biopsy with immunofluorescent staining. The Weil-Felix test has very poor sensitivity and is not specific, and the antibody does not develop until some time between 7-21 days after infective contact (compared with incubation period of 3-12 days). Complement fixation is more specific but detects 50% or fewer cases and is very laborious to perform, the antibody does not appear until 14-21 days after infection, and antibody elevations persist for years. Antibodies detected by the MIF and LA tests begin to appear 7-10 days after onset of illness. These tests are reported to detect considerably more than one half of Rocky Mountain spotted fever cases but are available only in reference laboratories or public health laboratories. The MIF test is currently considered the gold standard. LA is reported to detect about 80%-85% of patients detected by MIF. Skin biopsy of a lesion with immunofluorescent antirickettsial antibody studies on the tissue can also be done but the degree of sensitivity is controversial (50%-70%) and the test is not widely available.

Cat scratch disease

This condition has two possible etiologies. One is a small bacillus that stains faintly gram-negative, and the other is a member of the Rickettsial family. About 80% of patients affected are children (about 90% are less than 18 years old and about 70% are less than 10 years old). About 85% (range, 83%-90%) have a history of contact with cats, most of which (about 80%) are kittens or less than 1 year old. Cat inoculation via scratch (57%-83% of patients), bite, or lick is followed in 4-14 days (range, 3-50 days) by development of a primary inoculation site pustule or papule in 54%-93% of patients. The primary lesion persists 1-3 weeks but can persist up to 1-3 months. Regional lymphadenopathy develops in 1-3 weeks after appearance of the primary site lesion; the nodes are unilateral in about 90% (range, 85%-95%) of patients and are tender in about 80%. Cat scratch disease is possibly the most common cause of chronic lymphadenopathy in children. The node or nodes are located in the head and neck area in about 25% of cases; the axilla (occasionally the epitrochlear area) in about 45%, and the groin in about 20%. They generally regress in size in 2-4 months but can persist longer. Adenopathy is accompanied by fever and malaise in about 35% of patients (range, 30%-59%), but the fever is usually (90%) less than 102°F (38.9°C). The lymph nodes suppurate in 10%-30% of cases and may drain in 10%-20% of cases. Parinaud’s oculoglandular syndrome (unilateral conjunctivitis with regional lymphadenitis) is present in 4%-6% of patients. The lymph nodes often raise the clinical question of mycobacterial infection and sometimes of malignant lymphoma. Biopsy specimens display a characteristic lesion consisting of lymphocytes and also histiocytes with scanty cytoplasm surrounding a central area containing segmented neutrophils that usually becomes necrotic. The lesions may be single or can be stellate (branching) due to coalescence of the original lesions. The same histologic pattern can be found in tularemia and in lymphogranuloma venereum. It is extremely difficult to find the organisms by Gram stain, and more success is obtained with the Warthin-Starry silver stain. Interestingly, the more commonly used silver stain methods such as methenamine silver are not useful. Even with Warthin-Starry, the organisms may be difficult to find or may be absent. The best areas are the center of suppurative granulomas. A skin test has been described using antigen derived from lymph node pus, but the antigen is not readily available.

In 1988, scientists at the Armed Forces Institute of Pathology (AFIP), using special media, cultured a gram-negative rod from several patients with cat scratch disease which was subsequently given the name Afipia felis. However, only a few years later it was found that considerably more patients with cat scratch disease had antibodies to a rickettsial organism named Rochalimaea henselae than evidence of Afipia felis infection. Later, R. henselae was isolated by culture from several patients with cat scratch disease and now appears to be the most frequent cause. Both organisms can be isolated using chocolate agar in 5% CO2 atmosphere; but growth may take up to 6 weeks. Both organisms also grow in Haemophilus test medium.

The Rochalimaea genus includes two species, Rochalimaea quintana (etiology of trench fever) and R. henselae. R. henselae is now considered the etiologic agent of several conditions. One is bacillary angiomatosis, which is focal proliferation of small blood vessels, endothelial cells, and connective tissue cells (somewhat similar in appearance to Kaposi’s sarcoma), and which is found predominantly in the skin but sometimes in the liver or spleen, most often in patients with HIV-1 infection. R. quintana can also cause these lesions. Another condition is peliosis hepatis (multicystic angiomatoid structures with endothelium not seen by ordinary microscopy; usually in the liver). It also causes occasional cases of septicemia, most often in HIV-infected patients, and is one cause of cat scratch fever. The organism is a small gram-negative curved bacillus that requires special culture conditions for isolation. Diagnosis can be made by Warthin-Starry silver stains from tissue biopsies and culture on special media noted previously. There have been two published reports of a fluorescent antibody test for R. henselae and for A. felis. The sensitivity of these tests is not yet established. Nucleic acid probe methods have been used experimentally and will probably also become commercially available in the future, although most likely in many patients the disease will be treated empirically.