Expectorate sputum was obtained from participants in the TB group. pecific clinical specimens, such as pus, cerebrospinal fluid, or pleural fluid, were obtained from participants in the EPTB group according to the hospital standard operating procedure for AFB direct smear and Mtb culture. On the same day, each included participant was also required to submit a random urine specimen with a clean-catch collection technique25.

Sputum, cerebrospinal fluid, pus, or pleural fluid specimens were assessed for AFB with the Ziehl Neelsen (ZN) technique (ST Reagensia Company, Jakarta, Indonesia) according to the standard protocol26. Remaining sputum specimens were decontaminated using a standard N-acetyl-L-cysteine-NaOH (NaLC-NaOH) method and then concentrated via centrifugation at 3000 g for 15 minutes26 before culturing on Lowenstein-Jensen (LJ) medium (bioMérieux SA, Marcy l’Etoile, France). Subsequently, Mtb culture growth was observed weekly for 3 to 8 weeks and was considered positive when mycobacterial growth was >1 colony-forming unit (CFU).

The urine specimen was collected in a sterile container and directly centrifuged at 1006 g for 15 minutes to separate the low molecular weight analytes and then kept at -80°C for further examination of the Mtb antigen cocktail (ESAT6, CFP10, and MPT64). The cocktail was detected using the semiquantitative immunochromatography test (ICT) method (Jei Daniel Biotech Corp., Taiwan), in which 100 µL of sample buffer was placed into the specimen collection box with 100 µL of urine sediment and mixed well by dropper for 30 – 60 seconds then left to stand for 30 minutes. Four ~60 µL drops were applied to the “S” region of the testing card, and the result was read after 30 minutes using an automatic reader (JD Reader) to determine the Mtb antigen cocktail concentration of the clinical specimen. Quality control and instrument assurance were performed according to the manufacturer’s recommendations.

The urinary Mtb antigen cocktail concentration was assessed against the clinical TB, EPTB, and LTBI diagnoses. Population characteristics are presented descriptively: (1) as numbers and percentages for categorical data such as age, clinical symptoms, and gender; (2) as median and range (minimum-maximum) for numerical data. The significance of differences in median urinary Mtb antigen cocktail concentration among groups was evaluated using the Kruskal-Wallis test. The significance of differences in urinary Mtb antigen cocktail concentration between the following pairs of groups was independently assessed using the Mann-Whitney U test: (a) TB vs. EPTB; (b) TB vs. LTBI; (c) EPTB vs. LTBI. We considered tests with p < 0.05 to be significant. All data analysis was performed using SPSS software v.23 (IBM, Armonk, NY, USA).

Figure 1

Table 1

Variable	Tuberculosis Group
	TB n = 52	EPTB n = 22	LTBI n = 25
Gender (n, %)
Male	34 (65.4)	15 (68.2)	11 (44.0)
Female	18 (34.6)	7 (31.8)	14 (56.0)
Age (year)
Median	36	27	25
Range (min-max)	17 – 73	18 – 53	15 – 58
BMI (kg/m2)
Median	19.0	18.0	22.1
Range (min - max)	14.4 – 26.2	14.2 – 26.7	17.7 – 33.6
Symptoms (n, %)
Cough	46 (88.5)	7 (38.9)	NA
Fever	27 (71.1)	12 (37.5)	NA
Breathless	16 (42.1)	10 (31.3)	NA
Chest pain	16 (42.1)	6 (18.8)	NA

Abbrev: NA: not applicable; TB: active pulmonary tuberculosis; EPTB: extra-pulmonary tuberculosis; LTBI: latent tuberculosis infection

Figure 2

A total of 474 participants were enrolled in this study, and their positivity in direct smear (AFB) and Mtb cultures was 23% and 30% for TB and EPTB, respectively. These findings underscore the low positivity of AFB and culture tests in standard TB and EPTB diagnostics. However, the positivity of the standard IGRA diagnostic test for LTBI was 59% (Figure 1). Among participants in the EPTB group, four had spondylitis, nine had meningitis, eight had pleuritis, and one had abdominal.

The proportion of males among participants in the TB, EPTB, and LTBI groups was 65.4%, 68.2%, and 44%, respectively, with a median age of 36, 27, and 31 years, respectively. Median BMI across the three groups was 18-22.1 and categorized as normal. Clinical symptoms, such as cough and fever, were most prevalent in the TB and EPTB groups (Table 1).

The urinary Mtb antigen cocktail was successfully detected in the TB and EPTB groups. A higher median urinary Mtb cocktail antigen concentration was found in the TB group compared to the other groups (p < 0.001; Figure 2). Independent pairwise group comparisons found higher median urinary Mtb antigen cocktail concentrations in the EPTB (p < 0.001) and LTBI groups (p = 0.015) than in the TB group. In the LTBI group, 18 of the 25 participants had a urinary Mtb antigen cocktail concentration of 0 ng/mL, while the remaining 7 had urinary Mtb antigen cocktail concentrations between 5.2 and 26.2 ng/ml.

TB infection remains a challenging global threat, particularly in laboratory diagnostics. This study explored the urinary Mtb antigen cocktail in three TB groups: TB, EPTB, and LTBI. Higher urinary Mtb antigen cocktail concentrations were found in the TB group than in the other two groups. This finding indicates that in active pulmonary TB, ESAT-6, CFP-10, and MPT-64 antigens are secreted by Mtb replication and metabolism. However, a lower urinary antigen concentration was found in the EPTB group. A previous study observed fewer Mtb cells with EPTB, called paucibacillary, compared to active-pulmonary TB4, 27, 28, reflecting the lower urinary Mtb antigen cocktail concentration in our EPTB group. Moreover, Song et al. showed higher TB antigen concentrations in cerebrospinal fluid29. Therefore, the positivity performance of urinary TB antigens provides improved diagnostic performance with active TB.

Urine is an ideal clinical specimen because it is excreted in large quantities, and the collection process does not require invasive methods. The additional value of urinary Mtb antigen cocktails to improve pulmonary TB diagnosis, particularly in peripheral health settings, has been shown30.

This study also evaluated the performance of urinary TB antigens in those with LTBI, finding negative antigen results. In those with LTBI, mycobacterial proliferation is inhibited by immune cell interactions and granuloma formation. Chronic stimulation of proinflammatory cytokines such as IFN-γ, TNF-α, IL-6, IL-12, IL-17, and IL-23 stimulates macrophages to differentiate into epithelioid and giant cells. Moreover, the chemokines CCL2, CCL3, CCL5, CXCL8, and CXCL10 induce T-cell aggregation to form granulomas. Cytokines limit primary TB infection to increase its stability. In more than 90% of latent infections, foci caseoses contain mycobacteria lined by granuloma walls. A good host cellular immune response will prevent the replication and spread of mycobacteria31, 32, 33, 34. In vitro studies have found that the ESAT-6, CFP-10, and MPT-64 antigens are not secreted when there is no replication or interaction between Mtb and host receptors22, 31, 32.

However, seven participants with LTBI had urinary Mtb antigen cocktail concentrations between 5.2 and 26.3 ng/ml, returning a median TB antigen level of 0. One possible explanation for urinary TB antigens being detected in LTBI urine is that IGRA positivity cannot differentiate between active and latent TB. Therefore, the possibility of these participants having active early-phase TB cannot be discounted31. Second, granuloma is a dynamic battleground between host and pathogen35. Macrophages egress from the lesion and spread the infection to other lung regions during latency. However, this finding is at odds with the very definition of a latent infection following the migration of infective macrophages through the airspaces. Instead, it may reflect the flaring-up of quiescent lesions during apparent “latency.” An emerging consensus is that active and latent TB has a spectrum of lesional activity.

Several challenges in diagnostic accuracy with urinary lipoarabinomannan (LAM) tests have been previously reported36 that may impact urinary TB antigens: (1) The non-sputum-based test represents an imperfect reference standard, particularly for EPTB and paucibacillary disease; (2) the probability of cross-reactivity with other pathogens and colonizing organisms; and (3) user subjectivity in test interpretation when visual reading is applicable. The standard TB diagnosis by AFB and culture was used as the reference in the active TB group. The number of growth colonies in TB and EPTB cultures that contribute to urinary antigens was not observed. Therefore, the concentration of urinary Mtb still corresponds to the positive culture.

This observation highlights the cross-reactivity due to Mtb antigen strains instead of other pathogenic and colonizing organisms. Choundhry et al. showed that a higher concentration of urinary Mtb antigens detected using the CFP polyclonal antigen Mtb H37Ra strain were present among TB patients37. Meanwhile, a cocktail of ESAT-6, CFP-10, MPT-64, and immunodominant Mtb antigens were also detected. Alexandra et al. previously reported that using polyclonal antibodies in cocktail form increases the sensitivity and specificity of TB examination38. The urinary TB antigens were measured with an automatic reader, eliminating possible user subjectivity in test interpretation, which can be considered a TB diagnostic test even with limited information for urinary Mtb antigens.

The urinary Mtb antigen cocktail levels were elevated in patients with active TB and undetected in those with LTBI. Therefore, this study used the antigen cocktail as a potential marker for differentiating active TB and LTBI.

AFB: acid-fast bacilli, AUC: Area Under Curve, BCG: Bacillus Calmette-Guerin, CFP: culture filtrate protein, CFP-10: culture filtrate protein 10-kDa, CFU: colony-forming unit, EPTB: extra-pulmonary tuberculosis, ESAT-6: Early secretory antigenic target protein 6-kDa, ICT: immunochromatography test, IFN: Interferon, IGRA: Interferon γ Release Assay, IL: Interleukin, LCS: Liquor cerebrospinal, LTBI: latent tuberculosis infections, MPT-64: Mtb protein 64, Mtb: M. tuberculosis, NAAT: nucleic acid amplification test, NaLC: N-acetyl-L-cysteine, PTB: pulmonary tuberculosis, RD: Regions of Differences, TB: Tuberculosis, TNF: Tumor Necrosis Factor, WHO: World Health Organization, ZN: Ziehl Neelsen

The authors are grateful to Bachti Alisyahbana for providing the clinical samples and also to the patients who participated in the study.

Dewi Kartika Turbawaty, Ida Parwati, and Adhi Kristianto Sugianli had full access to all of the data in the study and took responsibility for the data’s integrity and accuracy data analysis. Ahmad Rizal Ganiem contributed substantially to the study design, data analysis, and interpretation. Dewi Kartika Turbawaty, Ida Parwati, Adhi Kristianto Sugianli was the primary author and Dewi Kartika Turbawaty was a major contributor in writing the manuscript. All authors read and approved the final manuscript.

This study was funded by Universitas Padjadjaran through grant-in-aid for HIU for Dewi Kartika Turbawaty and Academic Leadership Grant for Professor Ida Parwati.

The corresponding author will provide the data used during the current analysis on reasonable request.

This study was approved by the Ethical Committee of the Faculty of Medicine, Universitas Padjadjaran, and Dr. Hasan Sadikin General Hospital (number 493/UN6.C1.3.2/KEPK/PN/2016 and L.B.02.01/C02/10856/VIII/2016). In addition, written informed consent was obtained from all participants.

Not applicable.

Dewi Kartika Turbawaty is an academic author, a clinical pathologist, and the Head of Clinical Pathology, staff of Microbiology and Biomolecular Department, Faculty of Medicine, Universitas Padjadjaran, Dr. Hasan Sadikin General Hospital. Ida Parwati is an academic author, a professor of the clinical pathologist, and the Head of Microbiology and Biomolecular Department, Faculty of Medicine, Universitas Padjadjaran, Dr. Hasan Sadikin General Hospital. Adhi Kristianto Sugianli is an academic author, a clinical pathologist, and staff of the Microbiology and Biomolecular Department, Faculty of Medicine, Universitas Padjadjaran, Dr. Hasan Sadikin General Hospital. Ahmad Rizal Ganiem is a Neurologist and the Head of the Neurology Department, Faculty of Medicine Universitas Padjadjaran, Dr. Hasan Sadikin General Hospital. The remaining authors declare no relevant financial disclosures or conflicts of interest.