Blood Culture–Negative Endocarditis, Morocco

We investigated the microorganisms causing blood culture–negative endocarditis (BCNE) in Morocco. We tested 19 patients with BCNE by serologic methods, molecular methods, or both and identified Bartonella quintana, Staphylococcus aureus, Streptococcus equi, and Streptococcus oralis in 4 patients. These results highlight the role of these zoonotic agents in BCNE in Morocco.

B lood culture-negative endocarditis (BCNE) can occur when the patient has previously received antibiotic drugs or in the presence of slow-growing or intracellular microorganisms (1,2). In Morocco, epidemiologic data on endocarditis are fragmentary and show that this disease is frequently associated with rheumatic heart disease (3). The aim of this multicenter preliminary study was to investigate the microorganisms causing BCNE in Morocco by using a multimodal strategy.
During June 1, 2009-September 1, 2010, we prospectively included all patients with BCNE seen in 3 health centers in the Morocco regions of Casablanca and Marrakech. The specimens were then referred to the Institut Pasteur du Maroc in Casablanca before being transported to Marseille, France, where diagnostic assays were performed as described by Fournier et al. (4). For each studied patient, the physician in charge completed a questionnaire. Answers were not obtained for all questions from all patients.
We extracted bacterial DNA from excised valves or EDTA blood by using the QIAmp Tissue kit (QIAGEN, Hilden, Germany) and performed PCR (5). We examined paraffin-embedded heart valves and used hematoxylin and eosin stain for histopathologic features (5). To detect microorganisms within tissues, we systematically performed the Giemsa, Gram (Brown-Brenn and Brown Hopps), periodic-acid Schiff, Grocott-Gomori, Warthin-Starry, Gimenez, and Ziehl-Nielsen stains (4). For patients for whom results of all other techniques remained negative, we performed autoimmunohistochemistry as described by Lepidi et al. (6).
Our prospective study enabled the identification of 19 patients (Table 1): 11 men, 7 women, and 1 person of unspecified gender. Mean age was 40.26 years (range 22-57 years). Among these, 6 lived in urban areas (socioeconomic conditions unknown) and 7 in periurban communities under conditions of poverty. No information on residential environment could be obtained for the remaining 6 patients.  In Morocco, cases of BCNE represent two thirds of all cases of infectious endocarditis and constitute a major problem of diagnosis and management of patients (1). In other countries, the negativity of blood cultures in BCNE may be explained mostly by the administration of antimicrobial drugs before blood culture collection or by the causative role of fastidious microorganisms, as is the case for zoonotic pathogens, such as C. burnetti, B. quintana, and B. henselae (7). However, in Morocco, cases of endocarditis caused by these zoonotic pathogens are poorly diagnosed. Previous studies have demonstrated the role of C. burnetii, but we could find no study that directly detected Bartonella in BCNE.

RESEARCH LETTERS
In the neighboring country of Algeria, B. quintana is considered to be the most common agent of infectious endocarditis, with a prevalence of 15.6%, compared with 5% in countries in Europe (8). This difference is likely explained by differences in living conditions (9). B. quintana infections occur preferentially in disadvantaged populations (homeless) infected by body lice. In our study, 7 patients for whom information on their residential environment was available lived in conditions of poverty and poor hygiene, including the patient who had B. quintana endocarditis.
Bartonella spp. can infect any heart valve and cause destructive valvular lesions often requiring surgical replacement. Despite the high antimicrobial drug susceptibility of these bacteria, the mortality rate from Bartonella endocarditis may reach 31% (10). In addition, the recommended treatment for these infections is a combination of doxycycline and gentamicin, which is different from that for endocarditis caused by other common bacterial species, and therefore a specific diagnosis is highly desirable (10). However, the small number of patients included in our study precludes the issuance of a definitive recommendation to include serologic testing, molecular testing, or both for zoonotic pathogens in patients with endocarditis in Morocco.