Diagnostic accuracy of metagenomic next-generation sequencing in pulmonary tuberculosis: a systematic review and meta-analysis

Abstract

Background: Metagenomic next-generation sequencing (mNGS) has emerged as a promising tool in clinical practice due to its unbiased approach to pathogen detection. Its diagnostic performance in pulmonary tuberculosis (PTB), however, remains to be fully evaluated.

Objective: This study aims to systematically review and Meta-analyze the diagnostic accuracy of mNGS in patients with PTB.

Methods: We conducted a literature search in PubMed (MEDLINE), Web of Science, Cochrane, and EMBASE databases, including studies published up to 2024. Studies comparing the diagnostic accuracy of mNGS with other methods such as Xpert-MTB/RIF and Mycobacteria tuberculosis (MTB) culture using bronchoalveolar lavage fluid (BALF), sputum, and lung biopsy tissue were included. Preclinical studies, review articles, editorials, conference abstracts, and book chapters were excluded. Statistical analysis was performed using Rev-man5, R package metabias, and Stata software.

Results: Thirteen studies met the inclusion criteria and were included in the meta-analysis. The pooled sensitivity and specificity of mNGS for PTB were 83% (95% CI: 69-91%) and 99% (95% CI: 92-100%), respectively. Subgroup analyses revealed that in BALF, mNGS demonstrated a pooled sensitivity of 73% (95% CI: 61-82%) and specificity of 98% (95% CI: 92-100%); in the sputum, the pooled sensitivity was 60% (95% CI: 38-87%) with a specificity of 99% (95% CI: 96-100%); and in the lung biopsy tissue, the pooled sensitivity was 71% (95% CI: 38-95%) and the specificity was 98% (95% CI: 93-100%). For Xpert-MTB/RIF, the pooled sensitivity and specificity were 72% (95% CI: 53-85%) and 100% (95%CI: 100-100%), respectively. Subgroup analyses demonstrated that in BALF, Xpert-MTB/RIF exhibited a pooled sensitivity of 69% (95% CI: 53-81%) and a specificity of 100% (95% CI: 77-100%). The pooled sensitivity and specificity of mycobacteria culture were 50% (95% CI: 36-64%) and 100% (95% CI: 83-100%), respectively. Subgroup analyses indicated that in BALF, the pooled sensitivity of mycobacteria culture was 44% (95% CI: 37-52%) with a specificity of 100% (95% CI: 8-100%); in the sputum, the pooled sensitivity was 42% (95% CI: 21-65%) and the specificity was 100% (95% CI: 100-100%). When combining mNGS with Xpert-MTB/RIF, the pooled sensitivity and specificity were 79% (95% CI: 40-97%) and 98% (95% CI: 95-100%), respectively.

Conclusion: mNGS demonstrates similar diagnostic accuracy to Xpert-MTB/RIF in PTB and outperforms mycobacteria culture in terms of sensitivity. Furthermore, mNGS exhibits good detection capabilities across various PTB clinical samples.

Systematic review registration: PROSPERO CRD42023427586.

Keywords: Bronchoalveolar lavage fluid (BALF); Diagnosis; Metagenomic next-generation sequencing (mNGS); Pulmonary tuberculosis (PTB); Xpert-MTB/RIF.

Fig. 1

Flow chart of literature retrieval

Fig. 2

Categorized bar charts depicting risk of bias and applicability concerns in 13 included studies utilizing QUADAS-2. QUADAS-2 Quality Assessment of Diagnostic Accuracy Studies-2

Fig. 3

Forest plot displaying the sensitivity and specificity of mNGS across all pulmonary samples for the diagnosis of PTB

Fig. 4

Forest plot illustrating the sensitivity and specificity of mNGS in BALF for the diagnosis of PTB

Fig. 5

Forest plot depicting the sensitivity and specificity of mNGS in sputum samples for the diagnosis of PTB

Fig. 6

Forest plot showing the sensitivity and specificity of mNGS in the lung biopsy tissue for the diagnosis of PTB

Fig. 7

Forest plot outlining the sensitivity and specificity of Xpert-MTB/RIF across all pulmonary samples for the diagnosis of PTB

Fig. 8

Forest plot demonstrating the sensitivity and specificity of Xpert-MTB/RIF in BALF for the diagnosis of PTB

Fig. 9

Forest plot representing the sensitivity and specificity of culture methods in all pulmonary samples for the diagnosis of PTB

Fig. 10

Forest plot highlighting the sensitivity and specificity of culture methods in BALF for the diagnosis of PTB

Fig. 11

Forest plot portraying the sensitivity and specificity of culture methods in sputum samples for the diagnosis of PTB

Fig. 12

Forest plot exhibiting the sensitivity and specificity of mNGS combined with Xpert-MTB/RIF in all pulmonary samples for the diagnosis of PTB

Fig. 13

Summary receiver operating characteristic (SROC) plot encapsulating data from studies reporting Both the Sensitivity and Specificity of mNGS in BALF