Rocky Mountain Spotted Fever, Colombia

We investigated 2 fatal cases of Rocky Mountain spotted fever that occurred in 2003 and 2004 near the same locality in Colombia where the disease was first reported in the 1930s. A retrospective serosurvey of febrile patients showed that >21% of the serum samples had antibodies against spotted fever group rickettsiae.

B etween July 1934 and August 1936, sixty-fi ve cases of Rocky Mountain spotted fever (RMSF), 62 of them fatal, were reported from Tobia, Colombia (1). No reports of this disease (known locally as Fiebre de Tobia) have been produced from Colombia since, and currently RMSF is generally not included in the differential diagnoses of febrile syndromes.
We recently confi rmed RMSF as the cause of death for 2 patients by PCR (2,3), sequencing, immunohistochemical tests (4), and culture (5) ( Table 1). The fi rst patient was a 32-year-old pregnant woman (26 weeks), who had abdominal pain, headache, and fever in December 2003; pharyngitis was diagnosed, and she received amoxicillin with no improvement. A cutaneous macular rash, hepatomegaly, hyperbilirubinemia, leukocytosis, and thrombocytopenia (50,000/μL) subsequently developed. She then experienced respiratory failure and died. One of her relatives, as well as 2 dogs, had died a few days earlier with similar symptoms. Another sick dog rapidly recovered after receiving doxycycline.
The second patient was a 31-year-old previously healthy man who went to the local hospital in May 2004 with fever and severe headache; dengue was diagnosed clinically. Three days later, he became stuporous and was admitted to the hospital. Within a short period, seizures developed and he became comatose. He died a few hours later.
The 2 patients lived near the towns of Villeta and Tobia, Cundinamarca, Colombia. The histopathologic fi ndings of both patients were similar and consisted of vascular congestion; interstitial edema; frequent nonoccluding thrombi (mainly in the lungs); and multiple foci of perivascular lymphocytic and monocytic infi ltration in all viscera, including the brain. The lungs showed marked interstitial infl ammatory infi ltrates. Immunohistochemical analysis showed rickettsiae in the microvascular endothelium of all studied organs, including brain, liver, spleen, and lungs of both patients (Figure).
Several weeks after these events, we collected and identifi ed adult male and female ticks from the farms and surroundings where the patients had lived (Table 1). We found ticks of the species Amblyomma cajennense, a known vector of spotted fever group rickettsioses in Latin America (6)(7)(8)(9), and Rhipicephalus sanguineus, recently documented as a vector for Rickettsia rickettsii (10).
To begin to clarify the magnitude of spotted fever group rickettsioses as a public health problem in Colombia, we tested the following samples for spotted fever group rickettsiae by immunofl uorescence assay (IFA) (11): 1) 64 serum samples from a national Colombian surveillance system (2001)(2002)(2003)(2004)) that studies malaria, dengue, and yellow fever (Instituto Nacional de Salud, Colombia); and 2) 96 serum samples from a regional (the state where the reported patients lived) surveillance system (2000-2001) for dengue (Secretaria de Salud de Cundinamarca, Colombia). Serum samples showing distinctly fl uorescent rickettsiae at a >1:64 dilution were considered positive. We found immunoglobulin G (IgG) and IgM antibodies against spotted fever group rickettsiae (R. rickettsii was used as antigen) but not against typhus group rickettsiae (R. typhi was used as antigen) ( Table 2). These data suggest that spotted fever group rickettsioses may be a frequent cause of febrile illnesses, not only in the state where the reported patients lived but also in various other regions of Colombia. Since there is strong cross-reactivity among rickettsial species when IFA is used as an antibody-detection technique, other spotted fever group rickettsiae, including those recently described in Latin America (R. parkeri and R. felis) could explain the assay results (12,13). Furthermore, most of these patients received a clinical diagnosis of dengue, an endemic disease in Colombia that appears to have become an umbrella diagnosis under which other diseases are assigned. A similar situation was recently described in Mexico (14).
RMSF in Colombia is seldom considered in the differential diagnosis for febrile disease; possible causes include the lack of an adequate diagnostic infrastructure and the invisibility of tick-and fl eaborne infectious diseases in most medical curricula. The problem is further compounded by the presence of numerous agents (many transmitted by arthropod vectors) that produce nonspecifi c febrile syndromes during the early stages of the disease. Most of those agents are viruses that, unlike rickettsiae, have no specifi c treatment; thus, physicians might not feel compelled to use antimicrobial agents. Given the lack of appropriate and inexpensive diagnostic tests that are useful in the acute stage and that can be implemented in small rural hospitals, the best diagnostic tool available to healthcare personnel is clinical suspicion based on knowledge of the clinical manifestations (15), ecology, and epidemiology of rickettsio-ses. Physicians in areas where RMSF is endemic should consider prescribing a course of empirical treatment with doxycycline in patients who have high fever, severe headache, and myalgia, even in the absence of rash or history of tick bite, as both are frequently absent in RMSF. Such a treatment will not harm a patient with dengue or other viral infections and is likely to save the life of a patient infected with R. rickettsii.