The diagnostic yield of brushing versus non-brushing bronchoalveolar lavage fluid for Xpert MTB/RIF in patients with suspected tuberculosis: a prospective self-controlled study

Background Bronchoalveolar lavage fluid (BALF) Xpert Mycobacterium tuberculosis and rifampicin resistance (Xpert MTB/RIF) is increasingly used as an important diagnostic method for sputum smear-negative pulmonary tuberculosis (PTB) in many countries. However, no studies have investigated whether collecting BALF with or without brushing from the lesion site affects the diagnostic yield of Xpert MTB/RIF for PTB. This study aims to evaluate the impact of Xpert MTB/RIF testing on BALF collected with and without brushing on the diagnostic yield for PTB. Methods In a self-controlled study design, BALF samples collected before and after brushing were obtained from 124 suspected PTB patients from Chongqing University Fuling Hospital between September 2020 and September 2023. All specimens were subjected to MGIT960 liquid culture, acid-fast bacillus (AFB) smear, and Xpert MTB/RIF testing. Using clinical diagnostic results as the reference standard, the sensitivity and specificity of AFB smear, MGIT960 culture, and Xpert MTB/RIF assays for MTB detection were compared between pre- and post-brushing BALF. Subgroup analyses stratified by chest CT features and lavage lung segments were conducted. Results Of the 124 suspected cases, 110 (88.7%) were confirmed as active tuberculosis. The diagnostic yield differed substantially among methods, with Xpert MTB/RIF demonstrating the highest sensitivity by detecting 103 cases (93.6%). Notably, a 5.6% discordance rate was observed between paired pre- and post-brushing samples, contributing to the significantly higher sensitivity of post-brushing Xpert MTB/RIF testing compared to pre-brushing testing (93.6% vs. 87.3%, P = 0.014). In contrast, the sensitivity of BALF AFB smear and mycobacterial culture showed no significant improvement after brushing (P > 0.05. Furthermore, subgroup analysis revealed that the increase in sensitivity was particularly pronounced in patients with a tree-in-bud sign on chest CT (92.0% vs. 76.0%, P = 0.025). Conclusions Xpert MTB/RIF testing of post-brushing BALF enhances diagnostic sensitivity for PTB, particularly in patients with tree-in-bud CT findings. Prioritizing post-brushing specimens is recommended to improve early diagnosis rates, addressing diagnostic delays in challenging cases. This approach optimizes TB diagnostic workflows for timely clinical management.