Use of metagenomic next-generation sequencing for accurate diagnosis of tuberculous pleurisy: a retrospective cohort study

Introduction

Tuberculous pleurisy (TP) is the main cause of infectious pleural effusion and unilateral pleural effusion in developing countries (1). Notably, the clinical manifestations of TP vary, which include effusion and pleural thickening that can result in persistent lung damage (2). A key principle in tuberculosis (TB) control is the timely and accurate diagnosis so anti-TB treatment can be quickly implemented to prevent the spread of infection (3). Currently, clinical diagnosis of TP involves several assays, such as microbiological culture, acid-fast bacilli staining (AFS) test, Xpert MTB/RIF (Xpert Mycobacterium tuberculosis/rifampicin) assay and T-SPOT test. However, each assay has its own limitations and it is difficult to provide rapidity and accuracy at the same time (4). Conventional acid-fast bacilli (AFB) culture exhibits low positivity rates and prolonged turnaround times in pleural fluid samples (5). Moreover, Xpert MTB/RIF targets only Mycobacterium tuberculosis complex (MTBC)-specific gene sequences and cannot detect non-tuberculous mycobacteria (NTM) (6). Pleural biopsy is considered the gold standard for the diagnosis of TP. However, there are limitations with the histopathological diagnosis in differentiating between MTBC and NTM. Infections with MTBC or NTM both can yield positive AFS and changes of granulomatous inflammation. Additionally, these infections may show similar clinical symptoms and radiological features, which often result in misdiagnosis of NTM infections; nevertheless, the treatment regimens for these two types of infections are significantly different (7). Misdiagnosis of these two types of pathogens can lead to poor clinical outcomes and increase potential transmission risks. Therefore, more precise diagnostic methods are needed in clinical practice to distinguish between these two different types of pathogen infections.

Metagenomic next-generation sequencing (mNGS) is a test method that directly extracts the genomic nucleic acids of all microbes from environmental or clinical samples and analyzes the entire microbiome compositions and gene functions without bias using high-throughput sequencing technology (8). This powerful approach is able to detect genetic information of a wide range of microbes, including bacteria, viruses, fungi, and parasites, in a single sequencing process within 24–48 hours (9,10). Due to its high sensitivity and efficiency, mNGS has been considered as a valuable method in clinical practice, with broad potential in detecting various infectious pathogens (11). Studies have shown that in bronchoalveolar lavage fluid (BALF), compared to traditional methods, mNGS offers high sensitivity and rapid results, and can aid in the diagnosis of mycobacterial infections including those that are drug resistant, thereby facilitating early diagnosis and personalized treatment (12,13). However, the diagnosis of tuberculous pleuritis typically begins with pleural fluid testing, yet conventional microbiological assays demonstrate limited sensitivity due to the pauci-bacillary nature of TP. Furthermore, the granulomatous tissue formed in TP is based on the body’s immune response to microbial antigens, and the relative scarcity of bacteria at the site of the lesion reduces the sensitivity of routine pathology examinations such as AFS test. The application of mNGS in formalin-fixed paraffin-embedded (FFPE) of pleural biopsy are still unclear. Therefore, this study collected FFPE samples of pleural biopsy tissues from TP patients for mNGS analysis to evaluate the diagnostic value of mNGS technology in FFPE specimens for TP detection. We present this article in accordance with the STROBE reporting checklist (available at https://jtd.amegroups.com/article/view/10.21037/jtd-2024-2050/rc).

Methods

Ethical statement and informed consent

The study was conducted in accordance with the Declaration of Helsinki and its subsequent amendments. The study was approved by Medical Ethics Committee of The First Affiliated Hospital of Soochow University [No. (2023) IRB-350] and the informed consent from each patient or the individual patient’s family was waived because there were no new interventions for the patients and the information was anonymized.

Subjects

This was a retrospective cohort study. Patients admitted to The First Affiliated Hospital of Soochow University and suspected of TP from April 1, 2018 to February 1, 2023 were screened. The patients underwent thoracoscopic surgery or internal medicine thoracoscopy, and had histopathological support for the diagnosis. Their FFPE samples of pleural biopsy were collected for mNGS assay detection. General clinical characteristics and history information such as gender and age were gathered. Clinical and laboratory indicators including T-SPOT, AFS of pleural biopsy samples, C-reactive protein (CRP), lymphocyte ratio in pleural effusion, protein, adenosine deaminase (ADA), lactate dehydrogenase (LDH) and histopathological data were also considered. See Figure 1 for details. TP was diagnosed in accordance with the WS 288-2017 Health Industry Standard of the People’s Republic of China (14).

Figure 1 Research flowchart. The study design roadmap is as above. Out of 22 NTP patients, 11 cases were malignant pleural effusion, 6 pulmonary infection and 5 pleural effusion due to other causes. FFPE, formalin-fixed paraffin-embedded samples; mNGS, metagenomic next-generation sequencing; NTP, non-tuberculous pleurisy; TP, tuberculous pleurisy.

Inclusion criteria: (I) patients with pleural effusion who had uncertain preoperative diagnosis and underwent pleural biopsy to obtain pathological tissues; (II) sufficient pathological FFPE samples from patients can be sent for mNGS assay.

Exclusion criteria: insufficient samples for mNGS assay.

Experimental methods

DNA extraction and mNGS

The 1 to 5 sections of FFPE samples were scraped and transferred to a 1.5 mL centrifuge tube. Nucleic acid extraction was performed using the Universal Genomic DNA Rapid Extraction Kit (DK806-02, LifeFeng, Shanghai, China) according to the manufacturer’s protocol. DNA quantity and quality were evaluated using Qubit (Thermo Fisher Scientific, Austin, TX, USA) and NanoDrop (Thermo Fisher Scientific, Wilmington, DE, USA), respectively. DNA libraries were constructed with the Hieff NGS OnePot II DNA Library Prep Kit for MGI (Yeasen Biotechnology, Shanghai, China) following the provided instructions. Library quality control was conducted using an Agilent 2100 Bioanalyzer (Agilent Technologies, Santa Clara, CA, USA). Sequencing was performed on a DIFSEQ-200 platform (Dinfectome, Nanjing, China) in 50 bp single-end mode. No template negative controls (NTCs) were incorporated throughout the extraction, library preparation, and sequencing processes.

Raw sequencing data were first processed using Bcl2fastq2 (version 2.20) for demultiplexing. Following demultiplexing, low-quality reads, adapter-contaminated sequences, duplicate reads, and short fragments (<36 bp) were filtered out to generate high-quality clean data. Host sequences were removed by mapping the filtered reads to the human reference genome (hs37d5) using Bowtie2 software (version 2.2.6). Reads that failed to align with the human genome were subsequently subjected to pathogen identification by aligning against a microbial genome database. This database comprised genome sequences of bacteria, fungi, viruses, and parasites, which were downloaded from the National Center for Biotechnology Information (NCBI, https://www.ncbi.nlm.nih.gov/).

Statistical analysis

SPSS 25.0 and GraphPad prism 8.0 software were used for statistical analysis. The statistical description of continuous variable data was expressed as mean ± standard deviation if normal distribution was satisfied, and as median and quartile if not satisfied. The t-test or rank sum test was used for hypothesis test. Categorical variable data were expressed by the number of cases and rate, and the Chi-squared test or Fisher’s precision probability test was used for hypothesis test. The diagnostic accuracy parameters, including sensitivity [true positives/(true positives + false negatives)] and specificity [true negatives/(true negatives + false positives)], were calculated with 95% confidence interval (CI) using Wilson score method (optimal for boundary proportions) or exact binomial methods (for n<20 samples). All statistical analysis employed two-sided tests with a predetermined significance threshold of α=0.05.

Results

Patient characteristics

A total of 73 patients with clinical symptoms suspicious for TP diagnosed at The First Affiliated Hospital of Soochow University during April 1, 2018 to February 1, 2023 were included. Among these patients, 51 cases were clinically diagnosed with TP based on pathological changes showing granulomatous or caseous necrotic lesions after thoracoscopic surgery or internal medicine thoracoscopy, while the remaining 22 cases were diagnosed with NTP. Out of these, 11 cases were diagnosed as malignant pleural effusion, 6 cases as pulmonary infection, 2 cases as eosinophilic pneumonia, and 3 cases as pleural effusion due to other causes. Of 51 TP patients with a median age of 52.0 years, 33 (64.71%) cases were male and 5 (9.80%) cases possessed a history of previous TB. Of 22 NTP patients with a median age of 64.0 years, 13 (59.09%) cases were male and none had TB previously. There was significant difference in CRP, ADA and lymphocyte percent between the two groups (P=0.001, P<0.001, P=0.001). There was no significant difference in effusion protein content and LDH (P=0.94, P=0.85). See Table 1 for details.

mNGS data of FFPE specimens

mNGS assay found that, of the 51 cases clinically diagnosed as TP, 40 showed positive result, and all the 22 cases with clinical diagnosis of NTP showed negative results. The consistency of detection between mNGS and clinicopathology assays was 84.93% (62/73), including the positive consistency rate of 78.43% (40/51) and the negative consistency rate of 100% (22/22). See Table 2 for details. In the 51 TP cases, mNGS detected MTBC in 40 (78.43%) cases and NTM in 31 (60.78%) cases. Of these, 16 cases were MTBC infections alone and 7 cases were NTM infections alone, while 24 were mixed MTBC and NTM infections. See Table S1 for details. Three cases of NTM detected were Mycobacterium gordonae, a common causative agent. The clinical significance of the remaining detected NTM is not clear among the 40 patients who were positive for both MTBC and NTM or NTM alone.

Comparison of performance among mNGS, T-SPOT and AFS assays

Out of the 51 patients diagnosed with TP, 38 had their blood tested for T-SPOT, with a positive rate of 94.74% (36/38). While the positive rate of T-SPOT in 22 NTP patients was 27.27% (6/22) (P<0.001). AFS was performed on 39 cases of 51 TP cases, the positive rate of AFS was 12.82% (5/39). The above results indicate that mNGS had a higher sensitivity (78.43%, 40/51) than the conventional AFS test (12.82%, 5/39) (P<0.001) and a higher specificity (100.00%, 22/22) than T-SPOT (72.73%, 16/22) (P=0.009). See Table 3 for details. Additionally, we focused on five cases of AFS positive FFPE whose mNGS results showed three were positive for MTBC, one was positive for Mycolicibacterium phlei, and one was positive for both MTBC and Mycolicibacterium phlei. See Table S2 for details.

Discussion

Clinical diagnosis of TP is mainly based on clinical manifestations and biochemical examinations of pleural effusion. The cell type of tuberculous pleural effusion is mainly lymphocytes, and the level of ADA is often increased. The results of our study also showed that compared with NTP patients, the pleural effusion test results of TP patients suggested a higher proportion of lymphocytes and higher ADA. However, clinical diagnosis alone carries a risk of misdiagnosis. The definitive diagnosis of TP requires pathological or etiological confirmation, which remains challenging due to the low positivity rates of conventional detection methods, particularly microbiological assays (15).

The Xpert assay, recommended by the World Health Organization, has been validated for high detection efficiency of MTBC in various fresh specimens and tissues. However, Xpert is not recommended by WHO for direct application to FFPE specimens (16,17). A prospective study showed that mNGS exhibited comparable sensitivity in FFPE specimens to Xpert assay performed on fresh tissues (9). Additionally, mNGS can detect a wide range of pathogens, including MTBC and NTM, and can also identify other potential co-existing infections (18). Meanwhile, mNGS analysis can identify drug resistance genes in pathogens, which is essential for personalized anti-TB treatment.

In this research, we report diagnostic efficacy of mNGS in pleural biopsy FFPE samples from patients with suspected TP. We founded that mNGS has a remarkedly improved sensitivity (78.43%, 40/51) in detecting pathogens in FFPE of TP, and is superior to the traditional AFS testing (12.82%, 5/39). Although the T-SPOT test has a high sensitivity (94.74%, 36/38) and can aid in clinical diagnosis, its lower specificity (72.73%, 16/22) limits its application in accurately identifying specific pathogens. Furthermore, the T-SPOT test cannot differentiate between latent TB infection and active pulmonary TB. In our study, the specificity of mNGS on pleural biopsy FFPE samples is 100.00% (22/22), higher than the 88.24% reported on pulmonary FFPE specimens by Sun et al. (9). In addition, we also detected a considerable number of NTM infections, either alone or coinfected with MTBC. Seven patients were infected with NTM alone, including 1 case of Mycobacterium gordonae, 4 cases of Mycolicibacterium phlei, 1 case of Mycolicibacterium phocaicum and 1 case of Mycolicibacterium goodii combined with Mycolicibacterium fluoranthenivorans infection. These patients were treated with first-line anti-TB drugs, and the disease was under control in the follow-up survey after 1 year. The reason may be that some NTMs are sensitive to first-line anti TB regimens (19), or their pathogenicity is too weak to be self-limiting diseases. There are few reports of NTM causing pleurisy (or coinfection with MTBC) before (20,21). Whether these NTM are pathogenic bacteria or due to specimen contamination needs further exploration by mNGS in more sterile samples. However, because clinical pathology and AFS cannot distinguish MTBC and NTM, NTM infection in TP may be greatly underestimated in clinic.

The results of the present study demonstrate the high sensitivity of mNGS in detecting mycobacteria in FFPE samples of pleural biopsy, highlighting the broad applicability and reliability of this method and providing important reference value for the diagnosis of TP. If this diagnostic assay can be further validated in a broader clinical setting, it has the potential to become an important tool for improving the accuracy and efficiency of diagnosing TP and optimizing management strategies for patients suspected of this disease.

Nevertheless, it should be pointed out that there are some limitations in this study. Firstly, the sample size of this study is small. Validation in larger-scale studies to ensure the generalizability and reliability of the results is required. Secondly, limited by the resources, we were unable to use the Xpert MTB/RIF test, may affect the accuracy of our control group setup and result interpretation. Additionally, the retrospective design of this study may be subject to selection bias and limitations of retrospective data. The non-freshness and preservation methods of FFPE samples, as well as the shortcomings of FFPE samples that are easy to cause DNA fragmentation, which may affect the quality of the samples, thereby affecting the sensitivity of mNGS detection performance and analysis of drug resistance genes. Fresh tissue samples may present better sensitivity and provide more complete drug resistance gene results, which still needs follow-up research to explore. However, in clinic, patients often refuse to submit mNGS of fresh tissue in advance before pathological diagnosis because of the high price of mNGS.

Conclusions

In conclusion, mNGS showed high sensitivity in pleural biopsy FFPE specimens and also provided specific information of pathogenic microorganisms, assisting in the rapid diagnosis and precise treatment of TP.

Acknowledgments

None.

Footnote

Reporting Checklist: The authors have completed the STROBE reporting checklist. Available at https://jtd.amegroups.com/article/view/10.21037/jtd-2024-2050/rc

Data Sharing Statement: Available at https://jtd.amegroups.com/article/view/10.21037/jtd-2024-2050/dss

Peer Review File: Available at https://jtd.amegroups.com/article/view/10.21037/jtd-2024-2050/prf

Funding: This study was supported by grants from Suzhou Science and Technology Plan Project (SLJ202003 to C.J.); Suzhou Medical and Health Technology Innovation Projects (SKY2022045 to C.J.); Jiangsu Provincial Medical Key Discipline (ZDXK202201).

Conflicts of Interest: All authors have completed the ICMJE uniform disclosure form (available at https://jtd.amegroups.com/article/view/10.21037/jtd-2024-2050/coif). M.S. is an employee of Nanjing Dinfectome Technology Inc. (a for-profit company) and served as a consultant for this study. The research team has an academic collaboration with Nanjing Dinfectome Technology Inc. The other authors have no conflicts of interest to declare.

Ethical Statement: The authors are accountable for all aspects of the work in ensuring that questions related to the accuracy or integrity of any part of the work are appropriately investigated and resolved. The study was conducted in accordance with the Declaration of Helsinki and its subsequent amendments. The study was approved by Medical Ethics Committee of The First Affiliated Hospital of Soochow University [No. (2023) IRB-350] and the informed consent from each patient or the individual patient’s family was waived because there were no new interventions for the patients and the information was anonymized.

Open Access Statement: This is an Open Access article distributed in accordance with the Creative Commons Attribution-NonCommercial-NoDerivs 4.0 International License (CC BY-NC-ND 4.0), which permits the non-commercial replication and distribution of the article with the strict proviso that no changes or edits are made and the original work is properly cited (including links to both the formal publication through the relevant DOI and the license). See: https://creativecommons.org/licenses/by-nc-nd/4.0/.

References

1. World Health Organization. Global tuberculosis report 2024. 2024. Available online: https://www.who.int/teams/global-programme-on-tuberculosis-and-lung-health/tb-reports/global-tuberculosis-report-2024
2. Pérez P, Bello A. Management of chronic pleural tuberculosis and non-tuberculous empyema in the 21st century. Shanghai Chest 2024;8:3.
3. Trajman A, Campbell JR, Kunor T, et al. Tuberculosis. Lancet 2025;405:850-66. [Crossref] [PubMed]
4. Xu F, Wang Q, Zhang N, et al. Simultaneous diagnosis of tuberculous pleurisy and malignant pleural effusion using metagenomic next-generation sequencing (mNGS). J Transl Med 2023;21:680. [Crossref] [PubMed]
5. Hong YJ, Kim HW, Kim YS, et al. Microbiological confirmation of tuberculous pleurisy with medical thoracoscopy: targeted pleural washing and pleural biopsy. J Thorac Dis 2024;16:4904-13. [Crossref] [PubMed]
6. Penn-Nicholson A, Georghiou SB, Ciobanu N, et al. Detection of isoniazid, fluoroquinolone, ethionamide, amikacin, kanamycin, and capreomycin resistance by the Xpert MTB/XDR assay: a cross-sectional multicentre diagnostic accuracy study. Lancet Infect Dis 2022;22:242-9. [Crossref] [PubMed]
7. Dartois V, Dick T. Therapeutic developments for tuberculosis and nontuberculous mycobacterial lung disease. Nat Rev Drug Discov 2024;23:381-403. [Crossref] [PubMed]
8. Zhu Y, Gan M, Ge M, et al. Diagnostic Performance and Clinical Impact of Metagenomic Next-Generation Sequencing for Pediatric Infectious Diseases. J Clin Microbiol 2023;61:e0011523. [Crossref] [PubMed]
9. Sun WW, Dong ZW, Zhou YM, et al. Improving the identification and diagnostic efficiency of Metagenomic Next-Generation Sequencing for mycobacterial granuloma on postoperative formalin-fixed paraffin-embedded specimens. Microbes Infect 2023;25:105185. [Crossref] [PubMed]
10. Zhang J, Hou XY, Wang JY, et al. Case report: Cutaneous anthrax diagnosed using mNGS of a formalin-fixed paraffin-embedded tissue sample. Front Cell Infect Microbiol 2024;14:1329235. [Crossref] [PubMed]
11. Tan JK, Servellita V, Stryke D, et al. Laboratory validation of a clinical metagenomic next-generation sequencing assay for respiratory virus detection and discovery. Nat Commun 2024;15:9016. [Crossref] [PubMed]
12. Yin Y, Zhu P, Guo Y, et al. Enhancing lower respiratory tract infection diagnosis: implementation and clinical assessment of multiplex PCR-based and hybrid capture-based targeted next-generation sequencing. EBioMedicine 2024;107:105307. [Crossref] [PubMed]
13. Wang JZ, Yuan D, Yang XH, et al. Etiology of lower respiratory tract in pneumonia based on metagenomic next-generation sequencing: a retrospective study. Front Cell Infect Microbiol 2023;13:1291980. [Crossref] [PubMed]
14. The National Health and Family Planning Commission of the People's Republic of China. Criteria for the diagnosis of pulmonary tuberculosis (WS 288-2017). Electronic Journal of Emerging Infectious Diseases 2018;3:59-61.
15. Pai M, Dewan PK, Swaminathan S. Transforming tuberculosis diagnosis. Nat Microbiol 2023;8:756-9. [Crossref] [PubMed]
16. Rabie H, Taljaard J. Improving tuberculosis diagnosis in adults with HIV through Urine-Xpert Ultra. Lancet Glob Health 2024;12:e1917-8. [Crossref] [PubMed]
17. Mvelase NR, Mlisana KP. Xpert MTB/XDR for rapid detection of drug-resistant tuberculosis beyond rifampicin. Lancet Infect Dis 2022;22:156-7. [Crossref] [PubMed]
18. Shi CL, Han P, Tang PJ, et al. Clinical metagenomic sequencing for diagnosis of pulmonary tuberculosis. J Infect 2020;81:567-74. [Crossref] [PubMed]
19. Conyers LE, Saunders BM. Treatment for non-tuberculous mycobacteria: challenges and prospects. Front Microbiol 2024;15:1394220. [Crossref] [PubMed]
20. Yagi K, Ito A, Fujiwara K, et al. Clinical Features and Prognosis of Nontuberculous Mycobacterial Pleuritis: A Multicenter Retrospective Study. Ann Am Thorac Soc 2021;18:1490-7. [Crossref] [PubMed]
21. Shinohara T, Morizumi S, Sumitomo K. Varying clinical presentations of nontuberculous mycobacterial disease : Similar to but different from tuberculosis. J Med Invest 2021;68:220-7. [Crossref] [PubMed]