Tuberculosis Diagnosis by Metagenomic Next-generation Sequencing on Bronchoalveolar Lavage Fluid: a cross-sectional analysis

• Main Authors: Xi Liu, Yuanli Chen, Hui Ouyang, Jian Liu, Xiaoqing Luo, Yayi Huang, Yan Chen, Jinmin Ma, Jinyu Xia, Li Ding
• Format: Article
• Language: English
• Published: Elsevier 2021-03-01
• Series: International Journal of Infectious Diseases
• Subjects: Tuberculosis; Metagenomic Next-generation Sequencing; Bronchoalveolar lavage fluid; Diagnostic performance
• Online Access: http://www.sciencedirect.com/science/article/pii/S120197122032587X

Background: Metagenomic next-generation sequencing (mNGS) is an effective diagnostic method for infectious diseases, however, its clinical utility for tuberculosis (TB) diagnosis remains to be demonstrated. Methods: A total of 322 bronchoalveolar lavage fluid (BALF) samples were collected from 311 suspected and confirmed pulmonary TB patients and tested by mNGS, acid-fast bacillus (AFB) smear by microscopy, Xpert® MTB/RIF (Xpert), mycobacterium culture and bacterial/fungal culture. Diagnostic performance of mNGS was compared with conventional methods for detection of Mycobacterium tuberculosis complex (MTBC) and other pathogens in BALF. Underlying factors associated with positive detection in pulmonary TB patients were investigated. Results: mNGS, Xpert and culture presented a high proportion of complete matching for MTBC detection (244/322, 75.8%). In pulmonary TB patients pre-treatment the sensitivity of MTBC detection by mNGS, Xpert, culture and smear was 59.9% (85/142), 69.0% (98/142), 59.9% (85/142) and 24.6% (35/142), respectively, and 79.6% overall; MTBC was detected by mNGS in 33.2% (5/34) Xpert and culture negative samples. Positive MTBC detection by mNGS was affected by Vitamin D, erythrocyte sedimentation rate, TB initial treatment/retreatment, and cavity in chest imaging (χ2 = 37.42, P < 0.001), but not by prior anti-TB therapy within 3 months. mNGS was able to detect new potential pathogens in 8.7% (28/322) of samples. Conclusions: Combining mNGS with conventional detection methods could increase the detection rate for MTBC. Additionally, mNGS could identify pathogens in a non-targeted approach for better diagnosis of coinfection.