Diagnostic Performance of Nanopore-Targeted Sequencing for Sputum Smear-Negative Pulmonary Tuberculosis: A Retrospective Study

Abstract Purpose To evaluate the diagnostic performance of nanopore-targeted sequencing (NTS) for sputum smear-negative pulmonary tuberculosis (PTB). Methods In this retrospective study, BALF from patients with suspected PTB and negative sputum smears was tested with NTS, Xpert MTB/RIF, and MGIT 960 culture at Shanghai Pulmonary Hospital (August 2022–March 2024). Sensitivity and specificity were evaluated against both microbiological and composite reference standards. Receiver operating characteristic (ROC) curves were generated and Youden’s index was employed to determine the optimal NTS cutoff. Results Among the 249 participants enrolled, 100 were microbiologically confirmed PTB, 33 were probable PTB, and 116 were TB-negative. Under microbiological reference standard, the sensitivity of NTS was 0.940 (95% CI 0.869–0.975), higher than Xpert MTB/RIF 0.840 (95% CI 0.750–0.903); P = 0.041 and MGIT 960 culture 0.640 (95% CI 0.537–0.732); both P 0.001. The specificity of NTS was 0.776 (95% CI 0.687–0.846), compared to 0.974 (95% CI 0.921–0.993) for Xpert MTB/RIF and 1.000 (95% CI 0.960–1.000) for MGIT 960 culture. Excluding patients with prior PTB history, NTS specificity increased to 0.856 (95% CI 0.766–0.916). According to Youden’s index, when four M. tuberculosis-specific sequence reads was defined as the optimal NTS cutoff for positive NTS result, the specificity of NTS rose to 0.907 (95% CI 0.827–0.954). Conclusion NTS offers rapid, highly sensitive detection of PTB in patients with low bacillary burden. Applying an optimal read-count cutoff and accounting for prior PTB history can enhance clinical utility.