Secondary Logo

Journal Logo

Trench Fever Unusual but Seen in Homeless and Alcoholics

Isaacs, Lawrence MD

ID Rounds

Dr. Isaacs is a clinical assistant professor of emergency medicine at the Temple University School of Medicine and the director of the emergency department at Virtua Hospital-Voorhees Division in Voorhees, NJ.

Trench fever is not something the average emergency physician thinks about, and many of us have not even heard of it. First reported in 1915 and common during World War I, trench fever was described in several reports early in the 20th century. (Trench Fever. Oxford (UK): Oxford University Press, 1919; Brit Med J 1916;12:225.)

In 1915, the organism was called Rickettsia quintana, but the name was changed to Bartonella in 1993. If the name Bartonella sounds familiar, that's because B. quintana is in the same genus as B. henselae, the organism responsible for cat scratch disease. B. quintana is an intracellular (found in erythrocytes and/or erythroblasts), gram-negative rod. Shortly after the war, the louse Pediculus humanus corporis was determined to be the vector of the disease.

Besides soldiers during wartime, homeless people living on the street or in shelters are at high risk of contracting trench fever, obviously, because the vector is the louse. Poor living conditions such as refugee camps and alcoholism also are risk factors for contracting the disease. After a mass grave of Napoleon's soldiers was uncovered in Lithuania, the DNA of B. quintana was found in the bodies of lice and the dental pulp of the soldiers' teeth. (J Infect Dis 2006;193[1]:112.) Because historical accounts claim that many of these soldiers died of infectious diseases during Napoleon's retreat from Russia, one can only speculate how many died of trench fever.

Because the vector is the human body louse, until recently the only reservoir was thought to be humans. Researchers report, however, recovered B. quintana DNA from the pulp of a domestic cat (Emerg Infect Dis 2005; 11[8]:1287), and in 2003, the cat flea was discovered to be another vector of transmission. (Emerg Infect Dis 2003;9:338.) The bacteria live in the louse's intestines, are excreted in its feces, and enter humans through broken skin, either from bites or scratches.

Three different clinical syndromes characterize the illness, the first by sudden onset of headache, dizziness, pain in the shins, and fever. Approximately three days later, the fever subsides, and four days later, a cycle of relapsing fevers begins. Each subsequent cycle is less severe than the preceding one. No deaths have been reported with this type.

The second is a persistent bacteremia characterized by consistent fevers for up to six weeks. (Mandell, Douglas and Bennett's Principles and Practice of Infectious Disease. 6th Ed. 2005. Elsevier.) There are case reports of bacteremia up to eight years after the initial infection. (Bull Acad Pol Sci (Med) 1949;7:233.) The third type is an afebrile infection which can present with a variety of nonspecific symptoms including headaches, myalgias, rash, hepatosplenomegaly, vertigo, lymphadenopathy, and arthralgias.

Back to Top | Article Outline

Serious Infections

B. quintana infection also manifests as endocarditis and bacillary angiomatosis. Endocarditis caused by B. quintana is generally found in homeless and alcoholic patients. Although vegetations are found by echocardiogram, the blood cultures are almost always negative, and in one study (Arch Intern Med 2003; 163:226), a small percentage (17%) were afebrile at time of presentation. This same study found a 12 percent mortality, even with antimicrobial and surgical treatment. The organism was determined to be the causative agent by DNA amplification, serology, or immunohistochemistry.

Bacillary angiomatosis was first described during the early years of HIV. It is a neovascular proliferative disorder similar to Kaposi's sarcoma. Although originally described in the skin and lymph nodes of HIV-positive patients, bacillary angiomatosis was later found in bone and many internal organs including the liver, spleen, brain, lungs, and uterine cervix. (Mandell, Douglas and Bennett's Principles and Practice of Infectious Disease. 6th Ed. 2005. Elsevier.) The lesions can be skin-colored, red, or purple, and can be deep or superficial. They bleed profusely if punctured.

The diagnosis of trench fever begins with clinical suspicion. Anyone who is infested with body lice or who has spent time in homeless shelters or on the street is obviously at risk. Fever and shin pain may be the tip-offs to the diagnosis. Because it is difficult to grow out of standard blood cultures (it usually takes 14 days), one of the standard diagnostic tests is indirect immunofluorescence, but polymerase chain reaction can be used when tissue biopsies are performed (heart valve and/or skin lesions).

Treatment is generally twofold. The vector first has to be eradicated, and then the bacterium. Either 1% malathion, 1% permethrin, or boiling the clothing will work on the lice. Remember that body lice stay on humans only to feed, but they live in clothing. Recently ivermectin has been shown to delouse without the other measures. (J Infect Dis, At press.)

To treat B. quintana, doxycycline and gentamicin together are the standard treatment regimen for trench fever and endocarditis. For bacillary angiomatosis, erythromycin or doxycycline is the recommended antibiotic. Interestingly, erythromycin is used because of its antiangiogenic, not antimicrobial, effect. Trench fever is not going to be a daily diagnosis, but for those of us who care for the homeless or refugees, it may be something to think about when faced with a complaint of fever with no obvious source.

© 2006 Lippincott Williams & Wilkins, Inc.