In this prospective study, we evaluated the performance of Xpert in detecting MTB in patients with undetermined pleural effusion by using a suspension of pleural biopsy tissue specimens under direct thoracic surgery against mycobacterial cultures using MGIT 960 and CRS. To our knowledge, this is the first prospective study of its kind. As a prospective study, it followed the natural flow of patients, namely, testing pleural effusions and other clinical diagnostic methods first. Those left with unexplained pleural effusion were invited to join this study and undergo thoracoscopy if they met the enrollment criteria.
It was previously reported that the pooled sensitivity of Xpert in diagnosing TPE was between 30% and 51% [10–12]. Other studies showed that pleural biopsies taken by medical thoracoscopy were the best specimens for diagnosing TPE [9, 13, 20–22]. Previous studies using Xpert to diagnose tuberculous pleuritis with pleural biopsy revealed its pooled sensitivity as 30.5% (95% CI: 3.5–77.8) and specificity as 97.4% (95% CI: 92.1–99.3) against culture as the reference standard, though one individual study reported its sensitivity to be as high as 85.5% [12, 23]. Other studies evaluated the performance of Xpert against two CRS standards, and the sensitivity was 16.0% (95% CI: 5–36%) and 13.3% (95% CI: 4–31%); thus, the authors suggested that Xpert was not the ideal diagnostic tool for pleural tuberculosis [24]. The same team improved their methods and increased the sensitivity of Xpert in diagnosing tuberculous pleuritis to 45.0% (95% CI: 32.1–58.4) [14]. They explained that the low sensitivity in their previous study published in 2013 was probably due to the small amount of pleural tissue taken using a closed pleural biopsy.
In this study, we have achieved a sensitivity of 70.3% (95% CI: 60.9–78.6) against a CRS standard, which was significantly better than MGIT 960 whose sensitivity was 59.26% (95% CI: 49.4–68.6). The sensitivity of Xpert was 69.8% (95% CI: 57.0–80.8) against the culture. Xpert thus demonstrates huge potential as a diagnostic tool for suspected difficult to diagnose cases with tuberculous pleuritis using thoracoscopic pleural biopsy.
Tuberculous pleurisy is a serious disease caused by MTB infection. The gold standards for the diagnosis of tuberculous pleuritis include microbiological tests, including acid-fast bacilli (AFB) smear testing and culture to detect MTB in pleural effusions [25]. However, it is very difficult to find the pathogen, due to the limited number of bacteria in pleural effusions. As many previous studies were using gold standards such as AFB and culture, their sensitivity could be an overestimation. Moreover, late and missed diagnosis and treatment implies a great impact on the patient’s prognosis.
Admittedly, internal thoracoscopy is an invasive examination, but it is often with a small trauma, easy operation, and can be completed under local anesthesia, with a clear vision, safety and efficacy [26]. The visual diagnostic efficiency using thoracoscopy for tuberculous pleuritis was reported to reach as high as 93.4% through visible medical thoracoscopic biopsy [27]. Another study found that out of 833 patients with unexplained pleural effusions, who were diagnosed by thoracoscopic medical examination, yielded a diagnostic efficiency of 92.6% [28]. A meta-analysis including 17 studies and 755 patients with unexplained exudative pleural effusion revealed a pooled sensitivity of 91.0% [29]. Therefore, medical thoracoscopy can be used effectively in the diagnosis of unexplained pleural effusion.
This study showed that the sensitivity of the pleural biopsy tissue suspension for Xpert examination was 70.3% (95% CI: 60.9–78.6) against a CRS standard, which was significantly better than that of MGIT 960 whose sensitivity was 59.26% (95% CI: 49.4–68.6), and both showed specificity as 100%. The Xpert method requires only 2 h to obtain results, whereas the MGIT 960 culture takes at least 2 to 3 weeks. The Xpert test thus has a significant advantage over the MGIT 960 culture method in the time required for diagnosis. The latest tuberculosis diagnosis in China included pleural effusion with nucleic acid detection of MTB in the diagnostic criteria of tuberculous pleurisy [30]. Based on the present results, pleural biopsy tissue suspensions can also be used effectively in the future for the diagnosis of tuberculous pleuritis using Xpert and culture for confirmed positive results. As we did not find many patients with resistance, it was difficult to draw a conclusion regarding the performance of Xpert in the diagnosis of resistance.
In summary, the use of internal thoracoscopic biopsy tissue suspensions for Xpert detection demonstrated the feasibility and sensitivity to be applied in clinical settings as a diagnostic tool for tuberculous pleurisy with the following advantages: (1) taking pleural biopsy specimens under thoracic surgery for internal medicine improved the diagnostic accuracy. (2) Grinding the specimen allowed it to be applied to different settings. The biopsy specimens after grinding and liquefying could be used for MGIT 960 culture and Xpert methods. Others could be used to perform acid-fast staining smears, Lӧwenstein-Jensen culture, real-time PCR, and loop-mediated isothermal amplification according to their own conditions, along with other examination factors. This could not be achieved in the present study owing to the morphological limitations of the original specimen. (3) The positive result of the Xpert test with biopsy tissue was significantly higher than the positive rates obtained with pleural effusion, which was more meaningful for clinical diagnosis. However, the research results may be affected by the small sample size and limitations of the specimen volume obtained in a pleural biopsy. In the future, additional patient data need to be accumulated gradually, and techniques such as the Hybrid knife system should be introduced to improve the biopsy specimens with respect to the volume of the material obtained and to reduce the deviations in order to obtain further accurate research results.