Combining interferon-γ release assays and metagenomic next-generation sequencing for diagnosis of pulmonary tuberculosis: a retrospective study Authors: Yanyan Liu, Miaohong Fang, Chenxi Yuan, Yi Yang, Liang Yu, Yasheng Li, Lifen Hu, Jiabin Li Published: 2024 ABSTRACT BackgroundRapid diagnosis of pulmonary tuberculosis (PTB) is urgently needed. We aimed to improve diagnosis rates by combining tuberculosis-interferon (IFN)-γ release assays (TB-IGRA) with metagenomic next-generation sequencing (mNGS) for PTB diagnosis.MethodsA retrospective study of 29 PTB and 32 non-TB patients from our hospital was conducted between October 2022 and June 2023. Samples were processed for TB-IGRA and mNGS tests according to the manufacturer’s protocol.ResultsThe levels of IFN-γ release in PTB patients were significantly higher than those in non-TB patients (604.15 ± 112.18 pg/mL, and 1.04 ± 0.38 pg/mL, respectively; p < 0.0001). Regarding presenting symptoms or signs, cough and thoracalgia were less common in PTB patients than in non-TB patients (p = 0.001 and p = 0.024, respectively). Total protein and albumin levels in the sera of PTB patients were significantly elevated compared to non-TB patients (p = 0.039 and p = 0.004, respectively). The area under the ROC curve (AUC) for TB-IGRA in PTB diagnosis was 0.939. With an optimal IFN-γ cut-off value of 14.3 pg/mL (Youden’s index 0.831), sensitivity was 86.2% and specificity was 96.9%. ROC curve analysis for mNGS and TB-IGRA combined with mNGS showed AUCs of 0.879 and 1, respectively. The AUC of TB-IGRA combined with mNGS was higher than that of TB-IGRA and mNGS alone.ConclusionsTB-IGRA combined with mNGS may be an effective method for diagnosing tuberculosis, and can be used in the clinical diagnosis of PTB.Supplementary InformationThe online version contains supplementary material available at 10.1186/s12879-024-10206-5. Background Tuberculosis (TB) has been recognized as a global public health emergency for the past two decades and remains the leading cause of death among adults worldwide due to an infectious disease [1]. According to the Global Tuberculosis Report 2020, more than 10 million new TB and 1.4 million deaths have been reported to the World Health Organization (WHO) [2]. TB is caused by Mycobacterium tuberculosis, an airborne infectious disease, with pulmonary tuberculosis (PTB) accounting for approximately 80% of all cases [3]. Bacteriological culture are the gold standards for TB diagnosis. However, the low positivity rate and long incubation period pose challenges for clinical diagnosis [4]. The morbidity and spread of TB can be reduced by early detection and accurate diagnosis, which are key elements of TB control [5]. Despite wide use in China, the TB antibody (TB-Ab) test generally achieves the diagnostic rate of < 70% [6]. The Xpert MTB/RIF assay is a commercially available, real-time PCR-based assay, which was recommended by the WHO in 2010 as an initial diagnostic test in individuals suspected of having TB, especially in high-burden settings [7]. However, this method has limited sensitivity for detecting PTB, false positivity, and restricted utility [2]. Therefore, a rapid and sensitive technology is needed to improve positivity rate. The tuberculosis-interferon (IFN)-γ release assays (TB-IGRA) demonstrates high specificity and sensitivity making it a promising diagnostic tool for TB [8]. The assay detect inflammatory cytokine of IFN-γ released by T cells, which is an ex vivo blood test for the T cell immune response [9]. However, TB-IGRA detection results might be influenced by the immune function of the individual. Patients with compromised immune systems might have false-negative results [10]. Additionally, high levels of inflammation could lead to an increased release of IFN-γ in response to non-TB antigens, which could potentially lead to a false-positive result [11]. Severe inflammation could suppress the immune system, which might result in a false-negative TB-IGRA test if the immune system was not responding adequately to the TB antigens [12]. Metagenomic next-generation sequencing (mNGS) plays an important role in clinical diagnosis and rational drug use [13–15]. The absence of culture and pathogen detection results within 24–48 h is an advantage of mNGS [16, 17]. M. tuberculosis and other potential pathogens can be detected using the mNGS to enhance the diagnostic accuracy for mixed infections [18–21]. However, M. tuberculosis’s intracellular bacterium and thick capsule can lead to false-negative results in the pathogen detection by the mNGS test [22, 23]. The broad detection capability of mNGS might lead to the identification of clinically insignificant microorganisms, potentially causing unnecessary treatment [24]. According to the WHO’s report on TB, no single test is 100% accurate [2]. Therefore, combining TB-IGRA and mNGS as a supplementary diagnostic method could potentially enhance tuberculosis diagnosis by leveraging the strengths of both tests. This combined approach could be particularly beneficial in complex cases, such as patients with negative or indeterminate TB-IGRA results. In this study, we investigated the sensitivity and specificity of PTB detection using TB-IGRA, mNGS, and clinical laboratory indices. This study aimed to develop improved strategies for PTB diagnosis by combining multiple parameters. Materials and methods Patients and sample collection This study was a retrospective cohort study. We reviewed patients with pulmonary infections from the First Affiliated Hospital of Anhui Medical University between October 2022 and June 2023 using a hospital electronic medical record system (Fig. 1). All enrolled patients were classified as having PTB or non-TB according to the Clinical Diagnosis Standards for TB in China (WS 288–2017). The Chinese diagnostic criteria for TB, outlined in the “WS288-2017” standard, provide a comprehensive approach to diagnosing pulmonary tuberculosis: (1) Suspected Cases: Defined by specific criteria including epidemiological history combined with clinical symptoms or imaging findings; (2) Clinical Diagnosis Cases: Identified by imaging findings along with clinical symptoms, tuberculin skin test results, or auxiliary examination results; (3) Confirmed Cases: Diagnosed by positive sputum smear microscopy, culture, or pathological diagnosis of tuberculosis lesions. The inclusion criteria were as follows: (1) PTB: Patients with pulmonary infection and positive mycobacterial tuberculosis cultures; (2) Non-TB: Patients with pulmonary infection negative for and nontuberculous mycobacterial (NTM), with no bacteriological (sputum smear microscopy, mycobacterial sputum culture, or nucleic acid amplification assays) or radiological evidence of PTB and NTM, and no history of TB and NTM. Patients with the following characteristics were excluded from this study: (1) patients aged less than 18 years, (2) those who received anti-TB treatment for > 2 weeks before admission to our hospital, and (3) those with incomplete information. Information on age, gender, smoking history, fever, cough, thoracalgia, wheezing, white blood cells (WBCs), C-reactive protein (CRP), neutrophils, procalcitonin (PCT), lymphocytes, erythrocyte sedimentation rate (ESR), D-dimer, total protein, albumin, interferon-γ (IFN-γ), smear-negative PTB, Chest CT scans, ten immunity tests, TB-IGRA and mNGS were obtained from medical records. This study was approved by the Ethics Committee of the First Affiliated Hospital of the Anhui Medical University (PJ2023-13-35). The Ethics Committee agreed to a waiver of informed consent because this was a retrospective study. Clinical trial number was not applicable. Fig. 1Flow diagram of cases inclusion and exclusion Flow diagram of cases inclusion and exclusion TB-IGRA test All subjects were tested using TB-IGRA according to the manufacturer’s instructions (Wantai Biology Ltd., Beijing, China). Briefly, were injected 1mL heparinized whole blood from each patient was injected into three special culture tubes (T/P/N) for TB-IGRA: T (test tube) coated with a recombinant fusion protein of CFP-10 and ESAT-6 (M. tuberculosis-specific antigens), P (positive control tube) containing phytohemagglutinin (PHA), and N (negative control tube). All the tubes were then incubated for 22 ± 2 h at 37 °C, then centrifuged at 3000 × g for 15 min, and the serum was collected. First, 20 µL of the serum was diluted, and 50 µL of the sample plasma was added to the sample wells, with calibration solution added to standard wells. The wells were mixed and incubated for 60 min at 37 °C. Subsequently, the was added about 50 µL of enzyme-labeled antibody was added to both sample and standard wells, mixed, and incubated for 60 min at 37 °C. After washing the wells five times, 50 µL of chromogenic solutions A and B were added and incubated for 15 min at 37 °C. Finally, the absorbance was measured at 450 nm using a microplate reader. Standard curves were prepared for each experiment. The ELISA immunosorbent assay was used to measure IFN-γ levels. Positive samples were identified according to the criteria listed in Supplementary Table 1. Metagenomic next-generation sequencing and analysis DNA samples were extracted and purified using a QIAamp DNA Micro Kit (QIAGEN, Hilden, Germany) according to the manufacturer’s recommendations. The samples were transcribed using the QIAamp Viral RNA Mini Kit (QIAGEN, Hilden, Germany) and the SMART MMLV Reverse Transcriptase kit (Takara Biotechnology Co. Ltd., Dalian, China). The concentration and quality of the extracts was determined using Qubit 4.0 (Thermo Fisher Scientific, MA, United States). DNA and RNA libraries were constructed using the QIAseq Ultralow Input Library Kit (QIAGEN, Hilden, Germany) and TruePrep DNA Library Prep Kit (Vazyme, Jiangsu, China), respectively. Sequencing of the libraries was performed on the NextSeq 550 platform (Illumina, San Diego, CA, USA) was used to sequence libraries. High-quality sequencing data were generated by removing low-quality, low-complexity, and short reads (< 35 base pairs). Bowtie 2 was used to obtain clean reads by mapping human reads to the human reference genome (hg38). Then, Burrows-Wheeler Aligner software was used to align the clean sequences to the microbial pan-genome database. Meanwhile the same procedure and bioinformatics analysis were used for mNGS of the negative and positive controls. The number of specific reads and reads per million (RPM) were calculated. For detected bacteria and fungi, an mNGS result was defined as positive when the genome coverage of detected sequences belonging to this microorganism ranked top 10 of the same kind of microbes or when RPMsample/RPMNTC>10 if RPMNTC≠0, and the microorganism was not detected in the negative control (NTC). For viruses, mNGS was considereda positive result when at least one specific read was mapped to a species or when RPMsample/RPMNTC was > 5 if RPMNTC≠0 and the virus was not detected in the NTC. Statistical analyses All statistical analyses were performed using SPSS (version 17.0; SPSS Inc., Chicago, IL, USA). Data are presented as mean ± standard error of the mean (SEMs). Differences in PCT, age, WBCs, lymphocytes, CRP, neutrophils, ESR, D-dimer, total protein, albumin, and IFN-γ levels was assessed by the Mann-Whitney U test. The chi-square (χ2) test was used to assess the differences in smoking history, gender proportion, fever, cough, thoracalgia, wheezing, smear-negative PTB, and chest CT findings. Receiver operating characteristic (ROC) curves were used to plot sensitivity versus 1-specificity (evaluated at several different diagnostic thresholds of IFN-γ concentrations). Youden’s index was used to determine the optimal cutoff threshold. The differences of the ROC curve (AUC) between the three groups were compared using MedCalc 22.0 software. Statistical significance was set at p < 0.05. Results General characteristics of the enrolled cohort Between October 2022 and June 2023, 89 patients with suspected pulmonary infections at the First Affiliated Hospital of Anhui Medical University were enrolled in this study. All enrolled patients underwent analysis using TB-IGRA and mNGS test. A total of 28 patients were excluded due to the loss of key clinical data (n = 20), lack of raw sequence data (n = 7), and duplication (n = 1). Eventually, 61 eligible patients were enrolled, comprising 29 patients with PTB and 32 patients without TB (Fig. 1). Comparison of clinical parameters in PTB and non-TB patients The mean age of the 29 PTB patients was 50.0 ± 3.23 years, and the mean age of the 32 non-TB patients was 58.1 ± 3.16 years. Regarding presenting symptoms or signs, cough and thoracalgia were significantly less common in PTB patients than in non-TB patients (p = 0.001 and p = 0.024, respectively; Table 1). The levels of total protein and albumin in the sera of PTB patients were significantly elevated compared to those in non-TB patients (p = 0.039 and p = 0.004, respectively). However, CRP levels in PTB patients were lower than those in non-TB patients (p = 0.024; Table 1). No significant differences in the other parameters were identified between the PTB and non-TB groups (Table 1). All of the patients were negative for ten immunity tests, including HIV, Hepatitis A, Hepatitis B, Hepatitis C, and so on. The lung infections of non-TB group were caused by viruses, bacteria, and fungi other than tuberculosis, including 14 Human herpesvirus types 4, 6 Streptococcus pneumoniae, 5 Human cytomegaloviruses, 5 Klebsiella pneumoniae, 5 Pneumocystis yersoni, 4 Pseudomonas aeruginosa, 3 Aspergillus flavus, 3 Human herpes virus type 1, 3 Hemophilus, 2 Stenotrophomonas maltophilia, 2 Candida albicans, 2 Escherichia coli, 1 Aspergillus sydowii, Human polyoma virus type 3, 1 Cryptococcus neoformans, 1 Human herpesvirus type 6B, 1 Bacillus cereus complex group, 1 Human herpesvirus type 7, 1 Actinomyces odontolyticus and 1 Acinetobacter baumannii. Multiple pathogens in some non-TB patients were detected by mNGS. The specific bacteria responsible for the non-TB group were listed in Supplementary Table 2. Table 1Demographic and clinical characteristics of PTB and non-TB patientsClinical parametersPTB (n = 29)Non-TB (n = 32)p value*Age (years)§50.0 ± 3.2358.1 ± 3.160.078Gender0.099 Male (n, %)12 (41.4%)20 (62.5%) Female (n, %)17 (58.6%)12 (37.5%)Smoke history (n, %)2 (6.9%)4 (12.5%)0.458Presenting symptoms or signs Fever (n, %)14 (48.3%)20 (62.5%)0.264 Cough (n, %)9 (31.0%)23 (71.9%)0.001 Thoracalgia (n, %)8 (27.6%)18 (56.3%)0.024 Wheezing (n, %)5 (17.2%)11 (34.4%)0.129Laboratory parameters WBCs (*109/L)§7.24 ± 0.617.48 ± 0.640.977 Neutrophils (*109/L)§5.58 ± 0.585.60 ± 0.670.707 Lymphocyte (*109/L)§1.03 ± 0.081.37 ± 0.190.180 PCT (ng/L)§0.09 ± 0.021.01 ± 0.510.374 CRP (mg/L)§31.29 ± 7.5765.58 ± 14.950.024 ESR (mm/h)§41.34 ± 7.5931.63 ± 4.540.649 D-dimer(µg/L)§1.63 ± 0.413.11 ± 1.040.484 Total protein (g/L)§64.99 ± 1.5360.23 ± 1.320.039 Albumin (g/L)§37.25 ± 0.9532.83 ± 1.000.004 Smear-negative PTB20 (69.0%)32 (100%)0.0001Chest CT scan0.608 Bilateral22 (75.9%)26 (81.2%) Unilateral7 (24.1%)6 (18.8%)Notes PTB, pulmonary tuberculosis; non-TB, patients with pulmonary infection negative for tuberculosis; WBCs, white blood cell; PCT, procalcitonin; CRP, C-reactive protein; ESR, erythrocyte sedimentation rate; CT, Computed Tomography§Data are shown as means ± SEMs*From Chi-square (χ2) test and Mann-Whitney U test. Statistical significance was set at p < 0.05 Table 1Demographic and clinical characteristics of PTB and non-TB patientsClinical parametersPTB (n = 29)Non-TB (n = 32)p value*Age (years)§50.0 ± 3.2358.1 ± 3.160.078Gender0.099 Male (n, %)12 (41.4%)20 (62.5%) Female (n, %)17 (58.6%)12 (37.5%)Smoke history (n, %)2 (6.9%)4 (12.5%)0.458Presenting symptoms or signs Fever (n, %)14 (48.3%)20 (62.5%)0.264 Cough (n, %)9 (31.0%)23 (71.9%)0.001 Thoracalgia (n, %)8 (27.6%)18 (56.3%)0.024 Wheezing (n, %)5 (17.2%)11 (34.4%)0.129Laboratory parameters WBCs (*109/L)§7.24 ± 0.617.48 ± 0.640.977 Neutrophils (*109/L)§5.58 ± 0.585.60 ± 0.670.707 Lymphocyte (*109/L)§1.03 ± 0.081.37 ± 0.190.180 PCT (ng/L)§0.09 ± 0.021.01 ± 0.510.374 CRP (mg/L)§31.29 ± 7.5765.58 ± 14.950.024 ESR (mm/h)§41.34 ± 7.5931.63 ± 4.540.649 D-dimer(µg/L)§1.63 ± 0.413.11 ± 1.040.484 Total protein (g/L)§64.99 ± 1.5360.23 ± 1.320.039 Albumin (g/L)§37.25 ± 0.9532.83 ± 1.000.004 Smear-negative PTB20 (69.0%)32 (100%)0.0001Chest CT scan0.608 Bilateral22 (75.9%)26 (81.2%) Unilateral7 (24.1%)6 (18.8%)Notes PTB, pulmonary tuberculosis; non-TB, patients with pulmonary infection negative for tuberculosis; WBCs, white blood cell; PCT, procalcitonin; CRP, C-reactive protein; ESR, erythrocyte sedimentation rate; CT, Computed Tomography§Data are shown as means ± SEMs*From Chi-square (χ2) test and Mann-Whitney U test. Statistical significance was set at p < 0.05 Demographic and clinical characteristics of PTB and non-TB patients Notes PTB, pulmonary tuberculosis; non-TB, patients with pulmonary infection negative for tuberculosis; WBCs, white blood cell; PCT, procalcitonin; CRP, C-reactive protein; ESR, erythrocyte sedimentation rate; CT, Computed Tomography §Data are shown as means ± SEMs *From Chi-square (χ2) test and Mann-Whitney U test. Statistical significance was set at p < 0.05 Comparison of IFN-γ release levels in PTB and non-TB patients The levels of IFN-γ release in PTB and non-TB patients were 604.15 ± 112.18 pg/mL, and 1.04 ± 0.38 pg/mL, respectively. PTB patients exhibited significantly higher IFN-γ release than non-TB patients (p < 0.0001; Table 2). Table 2Comparison of IFN-γ levels in PTB and non-TB patientsVariablesPTB (n = 29)Non-TB (n = 32)p value*IFN-γ (pg/mL)§604.15 ± 112.181.04 ± 0.38< 0.0001Notes PTB, pulmonary tuberculosis; non-TB, patients with pulmonary infection negative for tuberculosis; IFN-γ, interferon-γ; n, number§Data are shown as means ± SEMs*Comparison of IFN-γ level between PTB and non-TB patients by Mann-Whitney U test. Statistical significance was set at p < 0.05 Table 2Comparison of IFN-γ levels in PTB and non-TB patientsVariablesPTB (n = 29)Non-TB (n = 32)p value*IFN-γ (pg/mL)§604.15 ± 112.181.04 ± 0.38< 0.0001Notes PTB, pulmonary tuberculosis; non-TB, patients with pulmonary infection negative for tuberculosis; IFN-γ, interferon-γ; n, number§Data are shown as means ± SEMs*Comparison of IFN-γ level between PTB and non-TB patients by Mann-Whitney U test. Statistical significance was set at p < 0.05 Comparison of IFN-γ levels in PTB and non-TB patients Notes PTB, pulmonary tuberculosis; non-TB, patients with pulmonary infection negative for tuberculosis; IFN-γ, interferon-γ; n, number §Data are shown as means ± SEMs *Comparison of IFN-γ level between PTB and non-TB patients by Mann-Whitney U test. Statistical significance was set at p < 0.05 Comparison of IFN-γ release levels in PTB patients with positive and negative sputum smears The levels of IFN-γ release in PTB patients with positive and negative sputum smears were 1243.35 ± 412.44 pg/mL, and 425.56 ± 106.70 pg/mL, respectively. PTB patients with positive sputum smears had significantly higher IFN-γ release than those with negative sputum smears (p = 0.023; Table 3). Table 3Comparison of IFN-γ release levels in PTB patients with positive and negative sputum smearsVariablesPTB with positive sputum smears (n = 9)PTB with negative sputum smears (n = 20)p value*IFN-γ (pg/mL)§1243.35 ± 412.44425.56 ± 106.700.023Notes PTB, pulmonary tuberculosis; IFN-γ, interferon-γ; n, number§Data are shown as means ± SEMs*Comparison of IFN-γ level between PTB patients with positive and negative sputum smears by Mann-Whitney U test. Statistical significance was set at p < 0.05 Table 3Comparison of IFN-γ release levels in PTB patients with positive and negative sputum smearsVariablesPTB with positive sputum smears (n = 9)PTB with negative sputum smears (n = 20)p value*IFN-γ (pg/mL)§1243.35 ± 412.44425.56 ± 106.700.023Notes PTB, pulmonary tuberculosis; IFN-γ, interferon-γ; n, number§Data are shown as means ± SEMs*Comparison of IFN-γ level between PTB patients with positive and negative sputum smears by Mann-Whitney U test. Statistical significance was set at p < 0.05 Comparison of IFN-γ release levels in PTB patients with positive and negative sputum smears Notes PTB, pulmonary tuberculosis; IFN-γ, interferon-γ; n, number §Data are shown as means ± SEMs *Comparison of IFN-γ level between PTB patients with positive and negative sputum smears by Mann-Whitney U test. Statistical significance was set at p < 0.05 ROC curve analysis of the sensitivity and specificity of TB-IGRA, mNGS, and TB-IGRA combined with mNGS ROC curve analysis for IFN-γ showed that the area under the AUC for PTB diagnosis was 0.939 (95%CI: 0.876-1), with a sensitivity of 86.2% and specificity of 96.9% when the recommended cut-off value of 14.3 pg/mL (Youden’s index 0.831) was applied. When a cutoff value of 32.12 pg/mL was used, the sensitivity was 82.8% and the specificity was 100%, although Youden’s index slightly decreased to 0.828. ROC curve analysis for mNGS showed that the AUC for PTB diagnosis was 0.879 (95%CI: 0.782–0.977). ROC curve analysis for TB-IGRA combined with mNGS showed that the AUC for PTB diagnosis was 1. (Fig. 2) The AUC of TB-IGRA combined with mNGS was higher than that of TB-IGRA and mNGS alone (p = 0.046, and p = 0.003, respectively; Table 4). Fig. 2ROC curves of TB-IGRA, mNGS and TB-IGRA combined with mNGS for PTB diagnosis. ROC, receiver operating characteristics; TB-IGRA, tuberculosis-interferon-γ release assays; mNGS, metagenomic next-generation sequencing; PTB, pulmonary tuberculosis ROC curves of TB-IGRA, mNGS and TB-IGRA combined with mNGS for PTB diagnosis. ROC, receiver operating characteristics; TB-IGRA, tuberculosis-interferon-γ release assays; mNGS, metagenomic next-generation sequencing; PTB, pulmonary tuberculosis Table 4AUC differences of different testing methods for PTB patientsGroupsAUC difference (95% CI)Zp value*TB-IGRA vs. mNGS0.059 (-0.052, 0.170)1.0480.295TB-IGRA vs. TB-IGRA + mNGS0.081 (-0.001, 0.124)2.0120.046mNGS vs. TB-IGRA + mNGS0.121 (0.041, 0.200)2.9850.003Notes PTB, pulmonary tuberculosis; TB-IGRA, interferon-γ release assays; mNGS, metagenomic next-generation sequencing; AUC, the ROC curve*Comparison of ROC curves, Statistical significance was set at p < 0.05 Table 4AUC differences of different testing methods for PTB patientsGroupsAUC difference (95% CI)Zp value*TB-IGRA vs. mNGS0.059 (-0.052, 0.170)1.0480.295TB-IGRA vs. TB-IGRA + mNGS0.081 (-0.001, 0.124)2.0120.046mNGS vs. TB-IGRA + mNGS0.121 (0.041, 0.200)2.9850.003Notes PTB, pulmonary tuberculosis; TB-IGRA, interferon-γ release assays; mNGS, metagenomic next-generation sequencing; AUC, the ROC curve*Comparison of ROC curves, Statistical significance was set at p < 0.05 AUC differences of different testing methods for PTB patients Notes PTB, pulmonary tuberculosis; TB-IGRA, interferon-γ release assays; mNGS, metagenomic next-generation sequencing; AUC, the ROC curve *Comparison of ROC curves, Statistical significance was set at p < 0.05 Discussion The diagnosis of TB depends on the observation of clinical symptoms, smears and cultures of clinical samples, and radiographic examination [4]. However, the symptoms of TB have become increasingly insidious and the signs more atypical. Clinical diagnosis remains a major challenge due to the limitations of existing diagnostic methods and the high negative rate of cultures and smears. In this study, the TB-IGRA was combined with mNGS to determine whether this combination could reduce the rate of misdiagnoses and missed diagnoses. We evaluated the clinical characteristics and demographics of 29 patients with PTB and 32 patients without TB. We found that the levels of total protein and albumin in the sera of PTB patients were higher compared with those in non-TB patients, which might be associated with an increase in the production of various proteins to fight the infection, including acute-phase proteins. The above results were consistent with the results of previous studies [25, 26]. We also found that the levels of sera IFN-γ of PTB patients were significantly higher compared with those in non-TB patients, indicating that IFN-γ release could be influence by the immune function of the body [27, 28]. In this study, the incidence of cough and thoracalgia was significantly lower in PTB patients than in non-TB patients, indicating that pulmonary symptoms were relatively mild in these patients. It was due to the reasons that PTB patients might initially present with non-specific symptoms like fatigue, weight loss, or night sweats, rather than the more immediate and localized symptoms like cough and thoracic pain, and non-TB patients might had conditions that were more likely to present with cough and thoracalgia, such as pneumonia, bronchitis, or other respiratory infections, which are more acute and have more pronounced symptoms. It is worth noting that TB patients experienced fewer coughs, and thoracalgia might have led to delayed diagnosis because the onset of TB might have gone unnoticed. Therefore, early diagnosis is crucial for the effective treatment and prevention of disease progression. The levels of total protein and albumin in the sera of PTB patients were significantly higher than those in non-TB patients (p < 0.05). This might be due to the small sample size. However, it did not affect the overall analysis results as the inclusion criteria for PTB were positive with positive mycobacterial tuberculosis cultures, while non-TB was defined as a pulmonary infection caused by pathogens other than tuberculosis. There were no significant differences in age, sex, WBCs count, lymphocyte count, PCT level, neutrophil count, or chest CT findings between PTB and non-TB patients. Additionally, IFN-γ release in PTB patients with positive sputum smears were higher compared to PTB patients with negative smears. However, Ji et al. found there was no difference in the IFN-γ levels of smear-negative and smear-positive PTB patients (89.6% vs. 90.8%, p > 0.05) [29]. The different results might be due to the different sources of specimens and number of cases. Our results showed that the AUC of TB-IGRA, mNGS, and TB-IGRA combined with mNGS for the diagnosis was 0.939, 0.879, and 1, respectively. This indicates that the sensitivity of TB-IGRA combined with mNGS was significantly higher than that of TB-IGRA and mNGS alone (p < 0.05). Clinical studies have shown that T lymphocyte reduction can lead to a lower positivity rate of TB-IGRA in immunocompromised patients, such as those with advanced HIV infection, cancer, or those on immunosuppressive therapy [29, 30]. Addtionally, if the individual had been recently infected with TB, the immune system might not have had time to develop a detectable response to the TB antigens [4]. However, Our study did not address this issue. Additional data are required for further analysis. We found that the sensitivity was 86.2% and the specificity was 96.9%, when the optimal cut-off value for IFN-γ (14.3 pg/mL, Youden’s index 0.831) was applied. This suggests that the cut-off value for TB-IGRA is consistent with the diagnostic criteria, aligning with of previous studies [25, 28], indicating that TB-IGRA is a reliable diagnostic tool for PTB infection. Additionally, mNGS has been gradually applied to the diagnosis of TB infection due to its rapid and highly sensitive characteristics [18]. However, the specificity of mNGS is not high compared to the final clinical diagnosis [21]. The principle of TB-IGRA is that IFN-γ is secreted by T lymphocytes in PTB patients under specific antigen stimulation, making its sensitivity higher than that of mNGS, however, its specificity could be influenced by various factors, including the individual immune status, technical aspects of the test, and other concurrent infections [10]. On the other hand, mNGS could provide a broad spectrum of pathogen detection directly from clinical samples and can be complementary to TB-IGRA, especially in complex or unclear cases [19]. The main advantage of mNGS was the ability to eliminate interference from non-TB patients. Our results revealed that the combination of TB-IGRA and mNGS was superior to TB-IGRA and mNGS alone for PTB diagnosis. Therefore, it is necessary to perform the TB-IGRA combined with mNGS for PTB infections. The combination of TB-IGRA and mNGS may be helpful in evaluating and diagnosing PTB in elderly and immunocompromised patients in future studies, as well as in the patients at stages of infection that are currently difficult to diagnose. We hope that the combination of TB-IGRA and mNGS can be implemented in clinical settings. A limitation of our study was the small sample size owing to the design of this retrospective study, which obtained examination results from medical records. Meanwhile, TB-IGRA and mNGS examinations were not routinely performed at the First Affiliated Hospital of Anhui Medical University, which resulted in a substantial exclusion of subjects. Therefore, there might be selection bias in the statistics results. Further prospective studies with larger sample sizes are required to evaluate the combination of TB-IGRA and mNGS for rapid diagnosis of TB. Addtionally, the cost of TB-IGRA combined with mNGS was high, which might limit their use, especially in resource-limited settings. Conclusion TB-IGRA combined with mNGS may be an effective method for detecting PTB infection and holds significance in the clinical diagnosis of PTB, especially in mixed lung infections. Electronic supplementary material Below is the link to the electronic supplementary material. Supplementary Material 1 Supplementary Material 1