RESUMEN
<p><b>OBJECTIVE</b>The influence of anti-tuberculosis (TB) treatment history on tuberculous lymphadenitis (TBLN) diagnosis is unclear. Therefore, this study aims to evaluate the diagnostic methods, including histology, microbiology, and molecular tests, used for TBLN.</p><p><b>METHODS</b>In this study, suspected patients with TBLN and having different anti-TB treatment background were enrolled. All the samples were tested simultaneously by histology, Ziehl-Neelsen (ZN) staining, mycobacterial culture (culture), Xpert MTB/RIF (xpert), real-time PCR, and high-resolution melting curve PCR (HRM). Thereafter, the performance of these methods on samples with different anti-TB treatment background was assessed.</p><p><b>RESULTS</b>In our study, 89 patients were prospectively included 82 patients with TBLN and 7 with other diseases. The overall sensitivities of Xpert, real-time PCR, histology, ZN staining, and culture were 86.6%, 69.5%, 58.5%, 43.9%, and 22.0%, respectively. The anti-TB treatment history revealed dramatic influences on the sensitivity of culture (P < 0.0001). In fact, the treatment that lasted over 3 months also influenced the sensitivity of Xpert (P < 0.05). However, the treatment history did not affect the performance of remaining tests (P > 0.05). For rifampicin drug susceptibility test (DST), the anti-TB treatment showed only significant influence on the success rate of culture DST (P = 0.001), but not on those of Xpert and HRM tests (P > 0.05).</p><p><b>CONCLUSION</b>Other tests as well as culture should be considered for patients with TBLN having retreatment history or over 1-month treatment to avoid false negative results.</p>