Blood cultures are the critical element in the diagnosis of bacterial endocarditis. Endocardial vegetations exude bacteria into the bloodstream, causing continuous bacteremia and positive results on blood cultures in most patients. However, in all endocarditis series, some patients are found to have the disease despite negative cultures. The percentage of negative blood cultures varies among studies from 2% to 30%. The mean percentage appears to be about 10%, but with existing methods, probably fewer than 5% of cases of endocarditis will be culture-negative. High rates of culture-negative endocarditis from early studies probably reflect suboptimal technique and less rigorous criteria for diagnosis.
The most common cause of culture-negative endocarditis is prior antimicrobial therapy. In one large study, antimicrobial therapy reduced the incidence of positive blood cultures from 97% to 91%. Duration of antimicrobial therapy correlates with the likelihood of negative cultures. When antimicrobial therapy has been continued for several days, blood cultures usually remain negative for weeks or longer.
Several laboratory methods may be helpful when endocarditis is suspected in a patient already on antimicrobial therapy. Antimicrobial removal devices that bind antimicrobials in the serum to a resin are available. Blood must also be cultured by a routine system, as the antimicrobial removal device bottle may actually reduce the yield of some organisms.
Some studies have suggested that in patients already treated with antimicrobials, more frequent cultures should be taken. One recommendation is to obtain blood cultures every 8 to 12 hours for 3 days. Using 10 to 30 mL of blood (rather than the usual 5 mL) has also been recommended to increase culture positivity. If endocarditis is suspected in a patient who has received antimicrobial therapy, the microbiology laboratory should be asked to incubate blood cultures for at least 2 weeks.
A lysis centrifugation method has also been used to increase sensitivity in low-grade bacteremia. RBCs are lysed, and the organisms released are centrifuged. The method increases the incidence of contamination and may be less sensitive for anaerobes and streptococci.
Another important reason why some cases of endocarditis are culture-negative is that some organisms are fastidious and require special culture techniques. A group of slow-growing, gram-negative bacilli, including Haemophilus aphrophilus, Actinobacillus actinomycetemcomitans, Cardiobacterium hominis, Eikenella corrodens, and Kingella species (called as a group the HACEK bacteria), are particularly difficult to grow by standard method. These organisms require prolonged incubation and subculturing to chocolate agar.
Vitamin B and other nutritionally deficient streptococci cause about 5% of episodes of endocarditis. These organisms may fail to grow unless the medium is supplemented with pyridoxal hydrochloride or cysteine. These organisms will grow on blood agar streaked with staphylococci as satellite colonies. Nutritionally deficient streptococci show turbid growth in conventional media, so that subculturing can then be initiated (such culture methods are now routine in laboratories when endocarditis is suspected). Other unusual bacteria implicated in culture-negative endocarditis include Brucella species. The diagnosis is usually suspected by history. Culture requires special media and a carbon dioxide atmosphere. Legionella, an unusual cause of culture-negative endocarditis, has occurred almost exclusively in patients with prosthetic heart valves.
Fungal endocarditis frequently presents with negative blood cultures. Mucor species, Aspergillus, and Histoplasma can rarely cause endocarditis. Fungal endocarditis is often right-sided and may occur in drug users or patients with prosthetic valves. Vegetations are usually large, and major embolic complications are frequent. Hypertonic media and the use of arterial cultures have been recommended to improve laboratory diagnosis, but these recommendations are unproved.
Coxiella burnetii can cause Q-fever endocarditis in patients who have inhaled infected material from domesticated animals. Cultures of blood are negative. Serologic diagnosis using complement-fixing antibody is specific but not sensitive in making this diagnosis. Blood cultures must be obtained before antibiotic therapy is begun if the organism is to be isolated.
Shapiro et al. have reported a particularly well-documented case of Chlamydia psittaci endocarditis. Cultures of blood and pharyngeal specimens were positive, and Chlamydia was demonstrated in tissue by immunofluorescent stain. Unusual organisms such as the murine typhus agent and Brucella have also been diagnosed by serologic methods. The diagnosis of culture-negative enterococcal endocarditis was made in seven patients by using an immunoblotting technique directed against specific enterococcal antigen extracts.
Bartonella quintana has received increasing attention as a potentially important cause of culture-negative endocarditis. Several investigators have used polymerase chain reaction (PCR) to identify the organism in excised valves. In a multicenter international study, 22 cases of culture-negative endocarditis were reported to be caused by Bartonella species. Diagnostic studies included determination of antibody titers to Bartonella species by microimmunofluorescence, blood and vegetation culture, and amplification of Bartonella DNA from valvular tissue by PCR. Of the 22 patients studied, 13 had preexisting valvular heart disease, 11 were alcoholic, and only four owned cats.
In samples of resected heart valves from patients with culture-negative endocarditis made available at surgery, molecular techniques will be increasingly useful in determining an etiologic agent. In one study, broad-range PCR amplification of the 16S rRNA gene followed by single-stranded sequencing allowed detection of rare organisms that could not be cultured.
In patients with culture-negative endocarditis of native valves, a combination of ampicillin and aminoglycoside has been recommended as empiric therapy. This combination is effective against streptococci, including enterococci, and bacteria of the HACEK group. In patients with prosthetic heart valves and culture-negative endocarditis, both coagulase-positive and coagulase-negative staphylococci may cause the disease. Vancomycin plus aminoglycoside becomes the regimen most often recommended. Rifampin may be added as well. In patients who do not respond to antimicrobial therapy, a reassessment of etiologic agents, including fungi, Chlamydia, Coxiella, and other organisms indicated by history and epidemiology, should be pursued. Optimal antibiotic therapy for Bartonella endocarditis has not been well established. (S.L.B.)
Abraham AK, et al. Culture-negative infective endocarditis. Aust N Z J Med 1984;14:223.
Among 265 endocarditis cases, 7% were culture-negative. All had received prior antimicrobial therapy.
Agarwal AK. Culture-negative infective endocarditis. Postgrad Med 1982;72:123.
Summary of basic principles of diagnosis and treatment.
Breathnach AS, et al. Culture-negative endocarditis: contribution of Bartonella infections. Heart 1997;77:474.
Describes two cases of Bartonella endocarditis: one in a homeless man, the other in a patient exposed to fleas. PCR of the excised valves was used to identify the organism.
Burnie JP, et al. Role of immunoblotting in the diagnosis of culture-negative and enterococcal endocarditis. J Clin Pathol 1987;40:1149.
Describes antibody response to antigenic extracts of four enterococcal species. Immunoblotting was used to diagnose seven cases of culture-negative enterococcal endocarditis.
Cannady PB Jr, Sanford JP. Negative blood cultures in infective endocarditis: a review. South Med J 1976;69:1420.
Good summary of the incidence of culture-negative endocarditis in different series.
Ellner JJ, et al. Infective endocarditis caused by slow-growing fastidious gram-negative bacteria. Medicine (Baltimore) 1979;58:145.
Gives case reports and microbiologic characteristics of fastidious gram-negative organisms, including Cardiobacterium, Actinobacillus, and Haemophilus species.
Goldenberger D, et al. Molecular diagnosis of bacterial endocarditis by broad-range PCR amplification and direct sequencing. J Clin Microbiol 1997;35:2733.
A promising method for diagnosis of culture-negative endocarditis. Allows identification of unusual, nongrowing organisms such as Tropheryma whippelii.
Jalava J, et al. Use of the polymerase chain reaction and DNA sequencing for detection of Bartonella quintana in the aortic valve of a patient with culture-negative endocarditis. Clin Infect Dis 1995;21:891.
The patient described had negative blood cultures and negative bacterial cultures of the resected valve. The authors used PCR and bacterial primers combined with DNA sequencing from the aortic valve vegetation. PCR was used to amplify bacterial 16S rDNA from a template of the vegetation.
Kiehn TE, et al. Comparative recovery of bacteria and yeasts from lysis-centrifugation and a conventional blood culture system. J Clin Microbiol 1983;18:300.
The lysis centrifugation system was better than broth for detecting fungi and gram-negative bacilli but was less likely to detect streptococci.
McCabe RE, et al. Prosthetic valve endocarditis caused by Legionella pneumophila. Ann Intern Med 1984;100:525.
A case report of endocarditis caused by Legionella pneumophila. This organism is a potential cause of culture-negative endocarditis in patients with prosthetic valves.
Marrie TJ, et al. Culture-negative endocarditis probably due to Chlamydia pneumoniae. J Infect Dis 1990;161:127.
Chlamydia endocarditis diagnosed by serology.
Musso D, Raoult D. Coxiella burnetii blood cultures from acute and chronic Q-fever patients. J Clin Microbiol 1995;33:3129.
Diagnosis was made by positive blood cultures and serology. Blood cultures must be obtained before the initiation of antibiotic therapy, or results will be negative.
Pesanti EL, Smith IM. Infective endocarditis with negative blood cultures. An analysis of 52 cases. Am J Med 1979;66:43.
Compares clinical features of 52 patients having culture-negative endocarditis with those who are culture-positive. Culture-negative patients tend to respond less dramatically to antimicrobial therapy.
Raoult D, et al. Diagnosis of 22 new cases of Bartonella endocarditis. Ann Intern Med 1996;125:646.
Multicenter, international study suggests that Bartonella species are an important cause of culture-negative endocarditis. Amplification of Bartonella DNA by PCR was performed by using tissue from heart valve. Of the 22 patients, 13 had predisposing valvular heart disease, and 11 were homeless. There was a high level of cross-reacting antibody to Chlamydia, making cross-adsorption necessary for proper diagnosis.
Shapiro DS, et al. Brief report: Chlamydia psittaci endocarditis diagnosed by blood culture. N Engl J Med 1992;326:1192.
A very well-documented case of C. psittaci endocarditis. The patient had been exposed to her sister’s sick parakeet.
Tunkel AR, Kaye D. Endocarditis with negative blood cultures. N Engl J Med 1992;326:1215.
An editorial accompanying a case report of Chlamydia endocarditis provides an update on etiologic agents and the diagnostic approach to culture-negative endocarditis.
Van Scoy RE. Culture-negative endocarditis. Mayo Clin Proc 1982;57:149.
Describes an approach to culture-negative endocarditis. Survival is 92% if patient responds to therapy within 1 week.
Walterspiel JN, Kaplan SL. Incidence and clinical characteristics of “culture-negative” infective endocarditis in a pediatric population. Pediatr Infect Dis 1986;5:328.
Ten-year pediatric experience with culture-negative endocarditis.
Washington JA II. The role of the microbiology laboratory in the diagnosis and antimicrobial treatment of infective endocarditis. Mayo Clin Proc 1982;57:22.
Describes the approach to the diagnostic microbiology of endocarditis, including technique of blood culture, media to use, duration of incubation, and problems with fastidious organisms.
Wright AJ, et al. The antimicrobial removal device. A microbiological and clinical evaluation. Am J Clin Pathol 1982;78:173.
An antimicrobial removal device was of little value in 87 bacteremic patients. Bacteria from 8% of patients grew bacteria in the antimicrobial removal device bottles alone.