Tuberculosis Diagnosis: Updates and Challenges

Abstract

Tuberculosis (TB) is caused by a single infectious agent, Mycobacterium tuberculosis, and a public health concern due to increased cases of drug-resistance and high mortality rates. Rapid identification of tuberculosis is necessary for its early treatment and to prevent the emergence of drug-resistant strains. For effective management of patients, rapid, cost-effective, and point-of-care (POC) diagnostic methods are required. The commonly used screening and identification methods are clinical examination, radiography, sputum smear microscopy, culture method, serological method, and tuberculin skin test. In addition, several molecular methods such as NAAT based GeneXpert, loop-mediated isothermal amplification (LAMP), line probe assay (LPA), whole genome sequencing (WGS) and other non-invasive methods of lateral flow urine lipoarabinomannan assay (LF-LAM) and eNose assays are developed. Sputum smear microscopy, Xpert MTB/RIF, and LED-Fluorescence microscopy (LED-FM) are the preferred methods to use in peripheral laboratories. The non-invasive methods of tuberculosis diagnosis are more beneficial in patients from whom collecting sputum sample is difficult particularly in children and HIV co-infected patients. Molecular methods can simultaneously identify the pathogen, M. tuberculosis, and mutations in drug-resistance genes. Even though, many advanced methods are currently available, accurate and affordable diagnostic method for tuberculosis is still challenging. Here, we review and highlight the uses and challenges of currently available conventional and advanced diagnostic methods of tuberculosis screening and diagnosis.

1. Introduction

Until the coronavirus (COVID-19) pandemic, tuberculosis (TB) was the leading infectious disease, ranking above HIV/AIDS. Tuberculosis is an ancient bacterial infection and genetic evidence indicates that the causative infectious agent, Mycobacterium tuberculosis (Mtb), is infecting humans for more than 40,000 years and have originated from animal domestication [1]. Tuberculosis is a communicable disease and Mtb spreads through air droplets during coughing and sneezing. The global prevalence of TB in 2019 is approximately 10 million people. Global TB report of World Health Organization (WHO) has reported 5,946,816 pulmonary tuberculosis (PTB) patients, 206,030 patients with MDR/rifampicin-resistant (RR), and 12,350 patients with extensively drug-resistant (XDR) TB globally in 2019 [2]. The cases of pulmonary tuberculosis are higher than extrapulmonary cases. The majority of TB patients (about 90%) are found to be adults, with more cases being men than women [3]. Globally, among four people, approximately one demonstrates an immunological reaction to Mtb infection and they may either remain dormant or progress to an active infection. Patients who are infected with TB but not having significant signs of the active disease were previously defined to have latent TB and more recently changed to TB infection. Tuberculosis can affect the lungs, which is named pulmonary tuberculosis (PTB), or other organs, named extrapulmonary tuberculosis (EPTB) [4].

WHO recommends to utilize the TB screening tests initially to screen high-risk individuals and further identification of pulmonary tuberculosis for rapid diagnosis and early treatment initiation [5, 6]. Rapid and feasible diagnostic methods are required to screen and diagnose active TB cases, HIV positive patients, workers having current or past history of silica exposure, identify cases in high endemic regions and having limited access to healthcare facilities [7]. Systematic diagnostic methods to screen and diagnose TB include physical examination for signs and symptoms, Chest-X Rays, conventional culture method followed by antibiotic susceptibility testing, molecular tests such as Xpert MTB/RIF [7]. WHO recommended TB diagnostic methods include light emitting diode (LED) microscopy method, BACTEC Mycobacterium Growth Indicator Tube (MGIT) 960 system, Xpert MTB/RIF, lateral flow urine lipoarabinomannan assay (LF-LAM), loop-mediated isothermal amplification (LAMP), line probe assay (LPA). Sputum smear microscopy, Xpert MTB/RIF, and LED-Fluorescence microscopy (LED-FM) are the preferred methods to use in peripheral laboratories. A summary of commonly used screening and diagnostic tests is presented in Table 1 [8]. Even though, several advanced diagnostic methods are available, rapid and accurate diagnosis of TB is still challenging [9], particularly in regions with high endemic TB. In this review we provide an overview of various currently available methods to screen and diagnose tuberculosis.

Test	Principle/Technology	Sensitivity (%)	Specificity (%)	Target setting	Comments
Chest X-Rays	Imaging	87	89	Secondary & tertiary centres	Even though used as screening tool for PTB, etiological agent cannot be identified.; Used to differentiate primary & secondary TB.
Sputum smear microscopy	Ziehl-Neelsen (Z-N) staining & microscopic detection of bacilli	32–94	50–99	Peripheral & reference labs	Cannot differentiate Mtb and other acid fast bacilli
LED-fluorescence method	Auramine staining & detection by fluorescent microscope	52–97	94–100	Peripheral & reference labs
TB Detect kit	BioFM-Filter-based sputum concentration & detection by kit method.	~55	88	Peripheral & reference labs	Biosafe & equipment-free method
Conventional culture method	Growth on Lowenstein–Jensen (LJ) media & identification by colony morphology & biochemical tests	93	>99	Secondary & tertiary centres	Longer turnaround time
BACTEC Mycobacterium Growth Indicator Tube (MGIT) 960 system	Liquid culture method with drug susceptibility testing	89; (smear +ve); 73; (smear −ve)	>99	Reference labs	Mtb identification requires additional laboratory tests
Tuberculin skin test (TST)	Host immune reaction to Mtb in body	87–98	74–96	Secondary & tertiary centres	False +ve in BCG vaccinated, NTM infected & high endemic regions; Low sensitivity in immune-compromised individuals
Serological tests	Detection of Mtb mycolic acid components & inflammatory biomarkers	87–92	72–83	Peripheral & reference labs	Results may vary depending upon host metabolic and disease states
Interferon gamma release; assay (IGRA)	Immune response against Mtb antigen in blood	QFT (75–84)	QFT (75–91)	Secondary & tertiary centres	Not recommended to predict active TB and treatment decisions
Lateral Flow urine lipoarabinomannan (LF-LAM)	Antigen (mycobacterial cell wall components) detection in urine	44	92	Reference labs	Cross reacts with other mycobacterial species & fungi
Xpert MTB/RIF	NAAT qPCR	98 (smear and culture +ve)	99 (smear and culture −ve)	Distric or sub-district; labs	Limited utility in resource-limited settings
Loop-mediated isothermal amplification (LAMP)	NAAT	76–80	97–98	Reference labs	Simple method to use in resource-limited settings & high endemic regions

Table 1.

Commonly used tests for screening and diagnosis of TB and detection of drug resistance.

Key: PTB – pulmonary tuberculosis, TB – tuberculosis, Mtb – M. tuberculosis, LED - light emitting diode, BCG - Bacillus Calmette Guerin, NMT - nontuberculous mycobacteria, QFT - QuantiFERON-TB Gold, NAAT - nucleic acid amplifcation test, qPCR – quantitative PCR.

2. Clinical diagnosis

The clinical manifestations of tuberculosis occur only in 5–10% of infected patients. In majority of TB patients, pulmonary tuberculosis is reported which affects mostly the lower respiratory system. The common clinical signs of pulmonary tuberculosis include hemoptysis, productive and prolonged cough, low-grade fever, loss of appetite, fatigue, night sweats, malaise, and weight loss [10]. Tubercle bacilli can also infect other body sites such as lymph nodes, kidneys, bone, joints, and meninges and is called extrapulmonary TB and its clinical signs and symptoms depend on the body sites being affected [11].

3. Radiography

Chest X-Ray (CXR) is a common diagnostic tool that differentiates primary and secondary TB. Primary TB is manifested by the presence of a single lesion in the middle or lower-right lobe and enlarged draining lymph nodes. CXR shows endogenous reactivation in the apical site and typical lymph nodes with multiple secondary tubercles. In addition, miliary lesions spread throughout the lungs [12]. Even though Chest X-Ray detects pulmonary TB, it cannot identify its etiological agent.

4. Advances in microscopy

Ziehl–Neelsen (ZN) stain is a traditional and common method and the sample is termed “smear positive” or “smear negative”, based on the presence or absence of Acid Fast Bacilli (AFB). The sensitivity of traditional ZN stain is lower and requires bacillary load of 5000–10,000 CFU/ml in sputum. In addition, it cannot differentiate Mtb and other acid-fast bacilli. The sensitivity of microscopic detection of Mtb is improved with fluorescence using carbolfuchsin and flourochromes such as Auramine-rhodamine which has been widely supported by WHO [8].

Recent developments in light emitting diode technology (LED) have increased the utility of fluorescent microscopy. LED microscopes use fluorescent stains and are more sensitive in pathogen detection. To maximize the identification, multiple sputum specimens can be collected and examined on the first visit itself, rather than asking the patient at a later date. WHO has endorsed front-loading, or so-called 1-day diagnosis under defined programmatic condition [13]. ‘TBDetect’ kit is a bio-safe and fluorescent microscopy filter (BioFM-Filter) based method and increased the diagnostic efficiency of smear microscopy and LED method due to increased performance, feasibility and safety considerations. TBDetect kit concentrates sputum by filtration using BioFM-Filter. The sensitivity of TBDetect is 20%, LED-Fluorescence microscopy (LED-FM) is 16.1% and ZN microscopy is 16% [14]. This equipment-free TBDetect kit is more potential in TB diagnosis and has more utility in routine laboratory settings. However, when compared with Xpert MTB/RIF for examining pulmonary and extrapulmonary TB in specimens of sputum, urine, gastric aspirates, and others, LED-FM has less sensitivity in Mtb detection [15].

5. Advances in culture

All acid-fast bacilli are not M. tuberculosis and for definitive identification of pathogen, sputum smear microscopy and culturing of Mtb on suitable medium are required [16]. Sputum culturing is a sensitive method which can detect viable bacilli as low as 10 to 100 in volume of a few tenths of an ml. Culture method is more sensitive than sputum smear microscopy, in which sputum sample must have at least 5000 AFB/ml to get a positive result. The tubercle bacilli can be cultured on Lowenstein-Jensen (L-J) medium, egg-based medium and Ogawa’s medium [17]. The L-J medium has glycerol that improves Mtb growth, but not Mycobacterium bovis, whereas sodium pyruvate enhances M. bovis and few strains of drug-resistant Mtb culture in the medium [18].

WHO recommends the use of dual medium, solid medium (e.g. Lowenstein–Jensen or Middlebrook 7H11) and liquid medium (e.g. for use with the BACTEC Mycobacterium Growth Indicator Tube (MGIT) 960 system) to increase the sensitivity, specificity and avoid contamination and reduce turnaround time. The use of liquid culture medium in identification and drug-susceptibility testing (DST) was endorsed by WHO in 2007. BACTEC™MGIT™960 system contains liquid Middlebrook 7H9 medium that detects the increasing fluorescence signals, every 60 minutes, automatically as oxygen is consumed by growing bacilli. Oxygen quenches the fluorescent compound present at the bottom of MGIT. Growing Mtb uses oxygen in the MGIT and subsequently, the fluorescent compound is detected [19]. Currently used two models of BACTEC system are BACTEC460 and BACTEC MGIT960. BACTEC MGIT960 is a user-friendly system with having non-radiometric and continuous signal monitoring system and is more advanced than BACTEC460 [20]. The MGIT technology yields result in less than 8 days. Addition of Streptomycin, INH, rifampicin and ethambutanol at critical concentrations allows Automated MGIT technology to detect drug susceptibility. Mycobacterial culture remained the gold standard method for detection and drug susceptibility testing [4]. However, in MGIT method species identification of Mtb requires additional laboratory tests, and hence its utility is limited. Micro MGIT system is more advanced as it does not need any special instrument other than UV lamp for fluorescence detection and hence utility of Micro MGIT is more in resource-limited settings [21].

6. Serological tests

Several commercially available antibody-based TB diagnostic tests are on the market, but clinical validation and current test performance are poor. Serum biomarkers are considered potential in diagnosing TB. Devising specific and accurate biomarkers which are consistent in different HIV status, ethnicities, and sites of TB infection is difficult; however, C-reactive protein, interferon-γ, interferon-γ inducible protein-10, fibrinogen, α2-macroglobulin, matrix metalloproteinase-9, transthyretin, complement factor H, and tumor necrosis factor-α have shown as potential biomarkers with 92% sensitivity and 72% specificity in detecting TB [22]. Trehalose esters of mycolic acids of Mtb cell wall lipids have been used in serological tests to diagnose PTB and the assay showed 87% sensitivity and 83% specificity [23]. If commercial methods are developed for these biomarkers, the serum assay would be rapid and effective in determining whether the patient needs further diagnostic testing. Most of these biomarkers are inflammatory markers and vary widely among patients depending upon their metabolic and disease states. Even though WHO is recommending against using commercial serological assays, antibodies, and combinations of antigens in the test panel improved performance of TB screening tests [24].

7. Interferon gamma release assays

Past or current Mtb infection can be detected by measuring T-cell mediated interferon-gamma that are secreted following subsequent stimulation with specific Mtb antigens. QuantiFERON-TB Gold In-Tube assay (QFT-GIT, Cellestis Ltd., Australia) and the T-SPOT.TB (Oxford Immunotec, UK) are the two commercially available interferon-gamma release assay (IGRA) kits. These assay kits detect Mtb infection in blood samples by detecting plasma levels of immune cells secreted gamma interferon. The interferon gamma-specific Mtb antigens include early secretory antigenic target-6 (ESAT-6), culture filtrate protein 10 (CFP-10), and tuberculosis 7.7 antigens. The blood sample is considered TB positive if the gamma interferon levels are above a specified threshold [13, 25]. However, previous studies conducted in children and adults across the globe have reported that these two IGRA assays cannot differentiate latent Mtb infection from active tuberculosis disease and are less efficient than tuberculin skin test (TST). IGRA-based QIAreach QFT is a new and simplified version of QFT-PLUS. QIAreach QFT provides qualitative analysis using a fluorescence lateral flow reader, transportable, user-friendly and does not require well-trained personnel [26]. However, few reviews and meta-analyses have reported neither IGRA nor TST is highly accurate in predicting active tuberculosis. IGRA is not recommended for treatment decisions for suspected cases of TB instead UK National Institute for Health and Clinical Excellence guidelines and WHO suggest using IGRA as a supportive screening test in diagnostic laboratories particularly in low-and middle-income countries having high tuberculosis and HIV-infected patients [27, 28]. A negative result of IGRA assay may not definitively exclude active TB infection in a high-risk group. Compared to TSTs, IGRAs are more specific in infections caused by non-tuberculosis mycobacteria but they are wrongly marketed as a confirmatory TB test despite their limited clinical utility. IGRAs are more expensive and need tedious sample processing for accuracy in results. Even though IGRA evaluation studies and its recommendations required many years, effort, funding, and resources across the globe, it did not have a major contribution to successful tuberculosis control [13]. A recent meta-analysis has shown that heparin-binding hemagglutinin (HBHA) is a latency-associated Mtb antigen and HBHA-induced IGRA can be a promising diagnostic test to differentiate latent and active TB [29]. IGRA testing is also a cost-effective screening method [30].

8. Tuberculin skin test

Tuberculin skin test (TST) was introduced around 100 years ago but is still in use for an initial TB screening to detect Mtb exposure in many countries. In TST purified protein derivative (PPD) tuberculin is injected intradermally and after 48–72 h the induration is measured (≥5 mm is considered positive) at the injected site to detect the individual’s immune response. The accurate TST result depends on well-experienced personnel for intradermal injection of PPD and its interpretation. The test is a simple and suitable method to detect Mtb exposure in geographical areas with rare TB cases, the test may be false positive in individuals from high endemic areas, vaccinated to Bacillus Calmette Guerin (BCG) and infected with non-tuberculous mycobacteria. However, TST has low sensitivity in immune-compromised patients [31].

9. GeneXpert

The most significant development toward tuberculosis diagnosis was NAAT-based GeneXpert. It is a real-time PCR-based multifunctional, automated, point-of-care (POC), user-friendly diagnostic system. In GeneXpert M. tuberculosis complex and rifampicin (RIF) resistance (targets rpoB gene for RIF resistance and associated M. tuberculosis–specific flanking regions) can be simultaneously detected in 2 h of time. The sensitivity of GeneXpert is higher than the sputum smear microscopy and culture method. The assay has utility in detecting extra pulmonary tuberculosis with sensitivities of 53–95%. In addition to high sensitivity, the assay provides high specificity without any cross-reaction with nontuberculous mycobacteria. Among children having pulmonary tuberculosis, GeneXpert rapidly detects all the smear positive and 61% of smear-negative samples after two induced sputum samples. GeneXpert is successfully used for routine screening of patients before antiretroviral treatment [13].

Even though GeneXpert detects RIF resistance in initial multicentre evaluation with high sensitivity and specificity, few studies have reported that rpoB gene sequencing and other methods have detected false-positive RIF resistance in areas of low RIF resistance prevalence [32, 33]. Although GeneXpert is a long-awaited development in TB diagnosis, it may not be feasible in settings with a lack of infrastructure for working on real-time PCR and computers. The requirement of annual maintenance of equipment is another hindrance. To overcome the challenges, corrective measures are introduced, including the revision of diagnostic device software and redesigning cartridge oligonucleotide probes and the newer software and combination of the oligonucleotide probes, called G4 version cartridge is released [34].

As sputum smear microscopy and culture methods are having several limitations, WHO recommended Xpert (MTB/RIF or MTB/RIF Ultra) or Truenat (MTB or MTB Plus) in TB suspected individuals. These biomolecular diagnostic tests are also recommended to identify extra pulmonary TB and pediatric TB cases. This cartridge system is nucleic acid amplification tests (NAATs) based method and within 2 h of turnaround time detects the presence of Mtb DNA as well as mutations in rpoB which is a rifampicin drug resistance-associated gene [35, 36, 37]. Xpert assays are successful in diagnosing PTB in adults with 89% sensitivity and 99% specificity. The Xpert MTB/RIF Ultra assay is less specific than Xpert MTB/RIF assay as it fails to differentiate between dormant and active TB DNA samples. In addition, these assays are less sensitive in detecting Mtb in children and patients co-infected with HIV and extrapulmonary TB [38, 39, 40]. To overcome the limitations of high cost, and need of an uninterrupted power supply and to make these methods affordable in rural areas, several smaller and battery-operated technologies are in process. Currently, GeneXpert-Omni (Omni;Cepheid) is a promising, cost-effective, portable tool for widespread use in peripheral healthcare settings. It ensures point-of-care and is portable hence reducing the cost and time involved in transporting the specimens to central laboratories.

In addition to Omni, Cepheid is developing Xpert MTB/XDR assay to cover the detection of resistance to INH, FQL, ethionamide (ETH), and SLID. Similar to other Xpert assays, it is also a NAAT-based assay detecting 16 clinically significant mutations associated with resistance in 90 minutes. When compared to phenotypic drug sensitivity testing (pDST), it has 94% sensitivity and 100% specificity in detecting drug resistance [41]. Several large-scale multicentre clinical trials are currently ongoing in establishing it as a follow-on test to existing methods of Xpert MTB/RIF and MTB/RIF Ultra. The diagnostic performance of Xpert MTB/XDR needs to be improved for the early identification of drug resistance and shorter drug regimens.

10. Loop-mediated isothermal amplification (LAMP)

LAMP is a rapid, easy, inexpensive, and highly specific NAAT-based method used to diagnose infectious diseases. It utilizes different sets of primers (minimum 4) which can identify the target sequence by recognizing distinct regions (minimum 6) of target gene giving high amplification efficiency with a sensitivity of 76–80% and 97–98%. It is a single-step amplification-based strand displacement reaction of approximately 15–60 min at a constant temperature of 65°C and subsequent amplicon detection by visual inspection of incorporated fluorescence [8]. It is a simple method and does not use any expensive reagents or equipment for result interpretation and can be used as a rapid diagnostic tool in resource-limited settings and high endemic regions [42, 43]. For diagnosing EPTB, LAMP assay has higher sensitivity [44].

11. Line probe assay

Line probe assay (LPA) is a rapid PCR-based method that amplifies DNA from Mtb and immobilization of oligonucleotides on a strip. In the presence of gene mutations, immobilized oligonucleotides emit a colorimetric signal indicating isoniazid or rifampicin drugs resistance as well as drug-sensitive strains in the sample [45]. LPAs are efficient to detect drug-resistant strains of Mtb in smear-positive samples [46] and has optimal diagnostic accuracy in smear-negative TB cases [47]. WHO has endorsed LPA as the initial detection method of multidrug-resistant TB (MDR) for isoniazid and rifampicin resistance in both pulmonary and extrapulmonary TB patients [48]. LPA-based commercial products for TB drug resistance include INNO-LiPA Rif TB Kit (Innogenetics, Zwijndrecht, Belgium), GenoType MTBDRplus (Hain Lifesciences-Bruker, Nehren, Germany), and Nipro NTM + MDRTB II (Osaka, Japan). Fluoroquinolone (FLQs) resistance can be detected by a more sensitive and new-generation LPA method (GenoType MTBDRsl version 2.0; Hain Lifesciences-Bruker) [8]. SL-LPA MTBDRs1 ((Hain Lifescience, Germany) is a second-generation line probe assay. MTBDRs1 (version 1.0) detects mutations in gyrA, rrs, embB genes and version 2.0 detects additional mutations in gyrB and eis promoter region. Even though LPA is a cost-minimizing method [49], its accuracy varies, and WHO has limited its utility in XDR-TB surveillance [48].

12. ddPCR (digital droplet PCR)

Droplet digital PCR (ddPCR) is a newly emerging technology and is being utilized in recent developments. In ddPCR, sample is diluted and partitioned into several hundred and millions of reaction chambers (Figure 1) [50]. Each separate chamber contains single or more copies of target sequence while rest do not contain target sequence and provides higher sensitivity than qPCR, detects single copy of DNA, and provides absolute quantification of gene expression [50]. This technique is used as a reference method in absolute quantification and detection of mutant DNA of drug-resistant subpopulations of Mtb [51]. The ddPCR has a higher sensitivity than quantitative or qPCR and Mtb infection can be detected in sputum and blood specimens of pulmonary and extrapulmonary TB patients [52, 53]. Though ddPCR detects a single copy of DNA per sample, is potent in absolute quantification, and works for both sputum and blood samples, it is prohibitively expensive and requires an uninterrupted power supply [25].

Figure 1.

Schematic Representation of principle of droplet digital PCR: Sample containing target sequence is partitioned into several droplets (in the magnitude of thousands) and then amplified separately. Each of these droplets are read by a dedicated droplet reader and fluorescence intensities are measured (indicative of positive/negative reactions) KEY: DNA – deoxyribonucleic acid.

13. CRISPR

Combined use of Clustered Regularly InterSPaced Repeats or CRISPR and CRISPR ASsociated nuclease 9, Cas9 is an approach involving a programmable enzyme which cuts DNA at specific sites (Figures 2 and 3) [25]. CRISPR is potential in detecting pediatric TB [54]. This test is a highly sensitive method with single-copy DNA detection, requires less sample requirement and a short turnaround time than Xpert method for detecting both pulmonary and extrapulmonary TB [55, 56]. CRISPR associated enzymes are used in SHERLOCK (specific-high sensitivity enzymatic reporter unlocking) platform which detects a single copy of RNA or DNA. This isothermal-based SHERLOCK technology can be used in places where electricity or portable readers are not available [57]. The combined LAMP and CRISPR-Cas12b detection method are more efficient in smear-negative paucibacillary TB patients [58].

Figure 2.

Schematic Representation of the CRISPR/Cas9 system: GuideRNA (gRNA) binds with Cas9 enzyme to form a complex. Cas9 endonuclease nicks the DNA a few bases upstream to a Protospacer Adjacent Motif (PAM) and mediates the cleavage of target DNA regions which are complementary to the gRNA. Key: RNA – ribonucleic acid, CRISPR/Cas9 - CRISPR (clustered regularly interspaced short palindromic repeat)-associated protein 9, crRNA - CRISPR RNA, tracrRNA - trans-activating crRNA.

Figure 3.

Schematic Representation of the CRISPR/Cas12a system: CRISPR RNA binds with Cas12a enzyme and forms a complex. Fluorescent probes (C) are in the periphery binding to quenchers (Q) with an oligonucleotide. The enzyme cleaves the target DNA and subsequently probes which activates the fluorescent signals. Key: CRISPR/Cas12a - CRISPR (clustered regularly interspaced short palindromic repeat)-associated protein 12a, DNA – Deoxyribonucleic acid.

14. MicroRNA detection

MicroRNAs have an active role in several biological processes and are used as biomarkers in diagnosis, treatment, and prognosis of a wide range of diseases including tuberculosis. Circulating mRNAs are consistent among individuals, stable, and least influenced by endogenous RNAse activity. MicroRNA expression studies in plasma samples of pulmonary TB patients and normal individuals have identified smiR-769-5p, miR-320a, and miR-22-3p as potent plasma-based biomarkers in TB diagnosis and miR-320a levels were significantly higher in drug-resistant TB [59]. In addition, plasma miRNA levels of hsa-miR-29a-3p, hsa-miR-155-5p, and hsa-miR-361-5p were found to be significantly upregulated in active tuberculosis compared to normal individuals. This plasma-based detection method is a convenient way of diagnosing TB in a population where it is difficult to obtain sputum, especially in pediatrics and extrapulmonary TB cases [60]. These circulating plasma miRNAs can further enable the differential diagnosis of latent and active TB. Diagnostic performance of miRNA can be increased by integrating serum miRNAs with diagnostic models developed by miRNA characteristics and electronic health records (HERs) [61, 62] and bioinformatics analysis [63]. Even though miRNA-based detection of TB is convenient in diagnosis in children, the method is challenging to adopt in resource-limited settings. In addition, the efficacy of the test depends on the correct sequence of miRNA. A considerable number of miRNAs are identified in children TB cases and only 7% of them are considered significant for the test [64].

15. Handheld electronic nose model

The handheld electronic nose model is a point-of-care and portable model and it can be used for tuberculosis screening in remote rural areas and health care settings and to rule out TB test in vulnerable populations [65]. This device detects infection via the presence of volatile organic compounds (VOC). The Aeonose (eNose BV, Zutphen, Netherlands) is an example of such model and this device has sensors and pre-concentrator. The patient needs to breathe through Aeonose via a disposable mouthpiece for 5 minutes. Breath data will be generated on the laptop and analyzed through the website of eNose. The preliminary results showed poor sensitivity and specificity in suspected tuberculosis patients. The portable and short turnaround time for results (< 5 minutes), makes the model more suitable in rural areas [66, 67].

16. Raman spectroscopy

Raman Spectroscopy (SR) is another diagnostic tool used to diagnose cancer and bacterial infections. It is a portal device and can be used in rural areas [68]. The phenomenon of Raman scattering is used to identify unique molecular markers of bacteria upon excitation with a particular wavelength of light. A combination of Surface-Enhanced Raman Spectroscopy (SERS) with a bead-beating module of a lab-on-a-chip (LOC) device can successfully differentiate Mycobacterium tuberculosis complex (MTC) and nontuberculous mycobacteria (NTM) [69]. SR with less cost, and short turnaround time is potential enough to diagnose active tuberculosis and latent tuberculosis [70]. The combination of SR with an optical microscope is useful in non-destructive identification of a single bacterial cell. SR is a rapid, user-friendly, non-invasive method for identification of pathogens. In addition, integration of machine learning can make SR a more effective TB screening method [71]. A recent study on developing PCR-based SERS has reported it can rapidly distinguish TB positive rifampicin-resistant and TB positive rifampicin-susceptible patients [72].

17. Artificial intelligence

The application of artificial intelligence (AI) along with historical methods of tuberculosis diagnosis such as Chest-X rays and smear microscopy is rapidly increasing and minimizes human errors in the interpretation of results [73, 74]. AI technology is in its initial stages which need to be validated with a large number of sample sizes. Depending upon the population used such as HIV co-infected patients or pediatrics, the results of AI vary widely [75]. Further, AI is under consideration by WHO as a technique to diagnose tuberculosis. In addition, the result images can be transferred via mobile phone technology from rural settings to server site and data can be analyzed with AI [76]. Artificial intelligence can be used to diagnose pulmonary and extrapulmonary TB [77] and to predict drug-resistant and drug-susceptible Mtb strains [78].

18. Electrochemical biosensor platform

Arginine film-based biosensor platform is a new PCR-free method used in the detection of tubercle bacilli. IS6110 gene is used as a biomarker. The probe corresponding to IS6110 gene will be immobilized on the biosensor platform and hybridized with a sputum sample or isolated DNA sample. The target gene will be identified based on electrochemical analysis using the principle of pulse voltammetry and methylene blue reduction signal measurement. The biosensor is a portable device with high sensitivity and selectivity for TB diagnosis [79]. An electrochemical device (EC) can be integrated to loop mediated isothermal amplification (LAMP) PCR for rapid detection of tubercle bacilli. LAMP-EC functions with a screen-printed graphene electrode (SPGE), redox probe, and a portable potentiostat for diagnosis of tuberculosis. The biosensor is a portable device. DNA isolation and hybridization with the sputum sample can be performed at room temperature, therefore, it can be used in rural areas [80].

19. Whole genome sequencing (WGS)

Whole genome sequencing (WGS) data has 96% concordance with culture-based drug sensitivity testing. It provides comprehensive detail of Mtb whole genome and genotypic sensitivity data to most drugs used in the treatment of MDR-TB. The correlation between genotypic results of WGS and phenotypic sensitivities is yet fully explored. WGS ideally can detect all the mutations in the genome and their functional characteristics [81]. This technique can be applied for genotypic and phenotypic characterization of organisms and profiling of drug susceptibility [82, 83] including detecting mutations in new drugs such as bedaquiline and delamanid [84]. The performance of WGS technology is further enhanced by incorporating a novel method ‘SplitStrains’ which helps to analyze WGS data of patients having mixed infections [85]. Due to high cost, the need of robust technologies, and technical expertise, initially, there was limited WGS utility in low-income countries [86]. In some countries, it is used as an important tool for case diagnosis and formulation of public health policies to trace TB contact cases in outbreak [87]. Even though NAAT and LPA methods are rapid, feasible, and accessible, detection of mutations in regions other than rifampicin resistance-associated gene rpoB is challenging. It is reported that 95% of resistance is due to mutations in this region, and WGS with the advanced genomics data on TB resistance can develop as a revolutionary to tailor TB treatment of each patient [4].

20. Flow cytometry assays of the M. tuberculosis: specific T-cell responses

The functional profile of T-cell mediated Mtb-specific responses to active disease and latent Mtb infection can be detected using polychromatic flow cytometry. Tuberculosis-specific flow cytometry panel comprises markers such as CD3, CD4, and CD8 which determine T-cell lineage, and interleukin 2, interferon-gamma, and tumor necrosis factor-alpha (TNF-a) antibodies as cytokine functional profile. TNF-α specific CD4 T cells are reported as a predictive marker to differentiate between active disease and the latent Mtb infection. The CD4 T-cell marker has higher sensitivity of 100% and specificity of 96% [88]. Even though flow cytometry is a highly sensitive method, antiretinal antibodies, and associated CD4⁺ T cells were not found significant in latent TB-associated uveitis or sarcoid uveitis patients [89].

21. Urine-based diagnostic tests

Urine samples ensure a non-invasive method of detection assays and are easy to collect from both adults, particularly in HIV-coinfected TB patients and children. Commercial methods are available to detect tuberculosis in urine samples. Unlike NAAT detecting Mtb DNA, lateral flow urine lipoarabinomannan (LF-LAM) test detects Mtb infection by identifying lipopolysaccharide of mycobacterial cell walls in urine samples. Even though commonly not used in many countries, LF-LAM is recommended to use in HIV-coinfected patients. LAM test helps early detection of TB and lowers TB deaths in people living with HIV (PLHIV) [90, 91]. It can be often used in low-resource settings and it is beneficial in patients in whom obtaining sputum samples is difficult. It has 42% sensitivity in HIV-coinfected having TB symptoms. However, the specificity is less as it cross-reacts with other mycobacterial species and fungi. Thus currently, it is recommended to use as an initial screening test in rural healthcare centers of high endemic areas with TB infection [92, 93].

Few studies reported lower sensitivity of LAM in patients with non-HIV infections and moderate to higher specificity in patients coinfected with HIV and having advanced immunodeficiency [94]. Many tuberculosis diagnostic tests are less sensitive in HIV-infected patients having advanced immunodeficiency but the sensitivity of LAM enzyme-linked immunosorbent assay (ELISA) is high even at lower CD4 lymphocyte cell counts. In most TB patients with CD4 cell count <50 cells/μl detectable amount of LAM antigen in urine is reported and it can be tested using TB-LAM Ag urine dip-stick assays. This advanced assay is a point-of-care lateral flow, low-cost ($3.50 per test), and highly specific in patient with advanced HIV-associated immunodeficiency [13].

22. Conclusion

Diagnosis of tuberculosis needs a rapid with possible reporting on the same day of sample collection and making a quick therapy decision. Current existing methods of the tuberculin skin test, smear microscopy, immunological test, and conventional PCR method still face several challenges for optimal diagnosis. Tuberculin skin test shows false positive results in certain populations such as patients with prior BCG vaccine, children with Mtb infection, HIV co-infected patients. Sputum AFB stain is a quick and easy method, it is not a confirmatory for Mtb as nontuberculous mycobacteria also take up the stain and even symptomatic patients may remain undiagnosed or undergo delayed diagnosis. Volatile Organic Compounds (VOC), LF-LAM and eNose assays in Mtb patients have some future potential. These non-invasive and cost-effective tests are useful in children or critically ill patients.

Conventional diagnostic methods are used as routine diagnostic methods, however, WHO recommends next-generation of NAATs as they provide fast and reliable results as point-of-care diagnostics in peripheral healthcare settings. Advanced methods of GeneXpert MTB/RIF (GX) are highly sensitive and specific one-step PCR-based methods with a short turnaround time of 2 h. NAAT-based GeneXpert is routinely used in clinical settings in many countries. In summary, in the current technological development, WHO recommends POC-NAATs (2nd generations) in addition to GeneXpert MTB/RIF and peripheral laboratories and the WGS method in at least reference laboratories in near future for TB diagnosis.

Acknowledgments

KSS thank Ms. Apoorva Jnana, PhD scholar, Department of Biotechnology, Manipal School of Life Sciences, Manipal Academy of Higher Education, Manipal, India for help.

Conflict of interest

The authors declare no conflict of interest.

Authors contributions

Both the authors contributed for the idea of the manuscript, literature review and data analysis. First draft was written by PK and critically revised by KSS. Both the authors read and approved the final version of the manuscript.