Clinical application of chest CT image screening combined with nanopore targeted sequencing in improving the diagnosis rate of pulmonary tuberculosis patients

Authors:
Open Access
Publisher website
Google Scholar
Abstract
Background: How to increase the diagnosis rate of pulmonary tuberculosis (PTB) has always been a major challenge in medicine. Compared with conventional microbiological tests (CMTs), nanopore targeted sequencing (NTS) is undoubtedly a faster and more effective diagnostic method. This study explores the feasibility of combining chest CT imaging screening with NTS to improve PTB diagnosis. Method: This retrospective observational study was conducted from March 2023 to December 2024, enrolling 380 patients from a single center who underwent bronchoalveolar lavage fluid (BALF) analysis and NTS testing. We compared the diagnostic accuracy of radiologists with varying experience levels to determine the most suitable preliminary screening protocol for PTB. Additionally, we compared NTS results with CMTs (including mycobacterial culture, acid-fast bacilli smear and Xpert MTB/RIF) to evaluate differences in diagnostic performance between NTS and CMTs. Results: We conducted a horizontal comparison of PTB diagnostic accuracy among three radiologists with different levels of experience. The most experienced radiologist had the lowest missed diagnosis rate. When combining cases suspected by all three radiologists (202 cases), the missed diagnosis rate dropped to as low as 9.09% (27.27% [95% CI: 16.77–40.70%] vs. 30.91% [95% CI: 19.73–44.51%] vs. 16.36% [95% CI: 8.38–29.82%] vs. 9.09% [95% CI: 3.80–19.84%], p = 0.019). Furthermore, we analyzed the diagnostic performance of NTS and CMTs in PTB patients and found that NTS had significantly higher sensitivity than Xpert MTB/RIF, mycobacterial culture, and acid-fast bacilli smear (89.09% [95% CI: 78.17–95.02%] vs. 78.18% [95% CI: 65.44–87.12%] vs. 23.64% [95% CI: 13.99–36.99%] vs. 16.36% [95% CI: 8.38–28.82%], p < 0.001). ROC analysis demonstrated that incorporating NTS into the testing of suspected PTB specimens significantly improved diagnostic capability (AUC = 0.990, 95% CI: 0.968–1.000). Conclusion: Radiologists’ experience in PTB imaging diagnosis can reduce missed diagnoses. Additionally, combining NTS with CMTs helps improve early diagnosis in suspected PTB patients. The combined use of NTS and Xpert MTB/RIF yielded the highest diagnostic accuracy (AUC 0.990).

This publication has 37 references indexed in Scilit: