Bacterial lymphadenitis at a major referral hospital in France from 2008 to 2012.

Lymph node enlargement is a common medical problem, and in a large number of patients, the causes of lymphadenopathy remain undiagnosed. We report a thorough microbiological analysis of 1,688 lymph node biopsy specimens collected in our bartonellosis reference center. We studied lymph node biopsy samples from patients with suspected regional infectious lymph node enlargement from January 2008 to December 2012. To evaluate a useful strategy for the diagnosis of infectious lymphadenitis, specimens were cultured and subjected to molecular assays. Histologic analysis was done when possible. A total of 642 (38%) biopsy specimens were infected with a bacterial agent, and quantitative PCR (qPCR) was significantly better than 16S rRNA gene PCR (rrs) for the detection of Bartonella henselae (P = 0.05), Mycobacterium tuberculosis (P = 0.05), and Mycobacterium avium (P = 0.007). Molecular assays were significantly better than bacterial cultures for the diagnosis of Francisella tularensis (P = 0.017) but were less effective for detecting M. tuberculosis (P = 0.004) and M. avium (P = 0.001). Histologic analysis was done for 412 lymph nodes, and 20% of these were compatible with an infectious lymphadenitis, whereas a neoplasm was found in 29% of these lymph nodes. M. tuberculosis was detected significantly more in female than in male patients (P = 0.01), and patients with cat scratch disease (CSD) were younger than patients with M. tuberculosis, Tropheryma whipplei, and F. tularensis. Negative rrs PCR does not exclude the diagnosis of infectious lymphadenitis. Histologic analysis of lymph node biopsy specimens is critical, as a diagnosis of infectious lymphadenitis does not preclude other concurrent diseases.