Tuberculosis (TB) is a transmissible disease caused by the bacterium Mycobacterium tuberculosis (MTB). The bacterium normally targets the lungs causing pulmonary TB (PTB). It can also infect and spread to the other parts of the body like kidney, spine and brain causing extra pulmonary TB (EPTB) there. It usually spreads through the air by means of respiratory droplets through the sneeze or cough of an infected person. For diagnosis of pulmonary TB (PTB) which is more common, samples from respiratory tract like sputum, induced sputum, broncho aleveolar lavage or lung biopsy is usually done. Diagnosis of extrapulmonary TB (EPTB) depends on the site of infection, e.g. stool sample for intestinal TB to detect Mycobacterium avium as in the case of AIDS patients; other samples may include biopsies, aspirates, pus, urine, and sterile body fluids like cerebrospinal, synovial, pleural, pericardial, and peritoneal fluids (1-3).

Tuberculosis continues to be the major cause of mortality despite all sorts of developments in medicare including diagnostics and medicines. It poses continued threat to the health programs and services in resource poor set-ups. Co-infection with HIV is the new factor that is responsible for the escalation of TB (9% of global TB incidence) (4). Majority of infection in human results in an asymptomatic latent infection and about one in ten latent infection progresses to active TB if left untreated, killing more than 50 percent of those infected. The situation becomes aggravated further due to its co-infection with HIV and incomplete treatment worsening it further to relapse cases and emergence of drug resistance in the form of multi drug resistant (MDR) and extensively drug resistant (XDR)-TB. MDR-TB develops due to bacterial resistant to first line TB drugs like rifampicin (RIF) and isoniazid (INH), while XDR-TB occurs due to bacterial resistant to fluoroquinone (FLQ), amikacin (AMK), kanamycin (KAN) and Capreomycin (CPM). Tuberculosis is difficult to diagnose due to the late onset of symptoms, so that a person acts as a reservoir of TB bacilli with potential to infect others. Similarly, co-infection with HIV makes TB difficult to detect via conventional sputum smear microscopy (SSM) / X-ray. Late diagnosis coupled with incomplete treatment escalates it to MDR and XDR-TB. Although MDR-TB is curable via second line anti-TB drugs, therapy is ineffective, marred with toxic effects and require almost 2 years of treatment which is certain to have its side effect (1-3).

20^th century has seen spurt of development in TB diagnostics and medicare from conventional sputum microscopy to Nucleic acid amplification test (NAAT) based molecular and biosensing based protocols, culminating into genome sequencing. This has generated a ray of hope for the timely and accurate diagnosis and treatment, hence eradication of TB. However, it is still the giant killer and major cause of mortality and morbidity. Main cause of concern is the emergence of TB bacilli in the regions where it is thought to be completely eradicated. Irrespective of the evolution of fancy and improvised technologies for TB diagnosis SSM and culture based (Lowenstein-Jensen (LJ) media) are still the most reliable and preferred diagnostics of choice for detecting AFB causing active TB, particularly for resource limiting set up of developing countries (2-5).

Timely detection and effective treatment are essential for managing tuberculosis. However, the rate of global case detection is considerably low. The diagnosis of pediatric TB and EPTB is particularly difficult and there are, as yet, no rapid screening tests that can help finding active cases in the target population. Tuberculosis diagnosis is primarily based on the culture method and non-culture methods. Recently, it is complemented by cutting edge molecular and genome sequencing techniques that have increased the sensitivity of TB diagnosis manifold and also provide epidemiological information about it prevalence, INTRODUCTION of new strains or cause of relapse etc (2-3).

This systematic review has incorporated a decadal growth in TB diagnostics from conventional SSM to NAAT based molecular and sequencing based protocols. The review also highlights the TB diagnostics in varied set-ups focusing on challenges of customized diagnostics. In addition, review elucidates on currently available and upcoming protocols and barriers to their successful uptake. The review further focuses on the advantage and disadvantages of each technique / protocol and intends to complement earlier reviews and reports, stimulating discussion and informing potential opportunities for market intervention to improve access to effective TB diagnostics.

Time tested diagnosis for active tuberculosis (TB) involves detection of Mycobacterium tuberculosis complex bacilli in specimens from the respiratory tract for PTB or specimens from other bodily sites for EPTB. Microscopic analysis of sputum smear detects AFB by Ziehl-Neelsen(Z-N) staining (6). Sputum test is good indicator of bacterial load hence the infectivity. It is the most common, rapid and economical diagnostic method for active PTB diagnosis but it has lower sensitivity, when bacterial load is less than 10,000 bacilli / ml of sputum (Table 1). However, sensitivity can be improved by fluorescent staining (auramine-rhodamine) coupled with LED microscopes but it fails to distinguish between tuberculous and nontuberculous bacteria. Other limitations include prevalent epidemiological factors in the region of interest and skill of laboratory technician (7-8) (Table 1). Laserson et al. (2005) (9) in their study of Viatnamese immigrants have observed that incubation of sputum with detergent C18-Carboxypropylbetaine (CB-18) for 24 hours improves the sensitivity of AFB staining by 20-30% when compared to direct auramine and Z-N staining (9). Sputum analysis cannot detect EPTB. There might be a possibility that bacilli resides in cavities of lungs and may not be there in sputum. Decontamination process of sputum for mycobacterial culture may lead to the loss of bacilli from the sputum. In addition, HIV patients and children might found it difficult to produce sputum. These constraints severely limit the utility of this very simple and economical diagnostics method¹⁰.

Table 1. Imaging techniques

TBDx developed by signature mapping medical sciences, USA is an automated high-thoughput system with the capacity of 200 slides. It utilizes robotic loading of stained slides. Results are rapid in few minutes in the form of high resolution digital image (8, 11-12) (Table 1).

Priority based treatment depending on the severity of symptoms have the potential to discriminate amongst those at risk from the people bearing symptoms of PTB but do not actually have TB. Chest X-ray (CXRs) serves as a potent algorithm in mathematical analysis of TB epidemiology. CXR integrated with user friendly, robust interface has evolved into an automated algorithm to quantify abnormalities from digital CXR (DCXR) images. DCXR can be utilized for both active and passive TB diagnosis. It can be interfaced with other such platforms like as Xpert® MTB/RIF for rapid detection of active TB particularly in HIV co-infection setup. It is an economical tool with high sensitivity and specificity and could be a tool of need in low resource set-up (Table 1). Delft Imaging Systems (Netherlands) have come up with a software to score CXR image at the rate of one per minute. Fourth version of Computer aided detection for tuberculosis (CAD4TB) (Delft Imaging Systems) can also be integrated with DCXR platform (13-14). EasyPortable DCXR (Delft Imaging Systems) fits the requirement of limited resource set up. It includes an X-ray detector, CAD4TB software and Picture Archive and Communication System (PACS) with a capacity to hold 50,000 DCXR images and can operate from a laptop (2) (Table 1).

Advenio (Chandigarh, India) has launched riView-TB, a CAD4TB enabled automated image analysis software similar to DCXR. This platform provides easy to interpret images without prior training in clinical radiography. It is slated to be improvised for simultaneous diagnosis of pneumonia and silicosis to complement PTB diagnosis (2) (Table 1).

CellScope, a mobile cell microscope (CITRIS, University of California, Berkeley, San Francisco) is another innovation in the series of next generation microscopy. It is a fluorescence microscope that gives output in the form of high resolution enlarged digitized fluorescent images of blood samples to detect rod shaped bacterium (8, 15) (Table 1).

A good quality sample in sufficient amount is a prime requirement for accurate diagnosis. This becomes a limitation with children and HIV patients who cannot cough up enough sputum. Lung flute from Medical Acoustics (USA) allows the patient to collect their own sample when they are not able to expectorate enough volume. It consists of a plastic device held via lips to exhale air into it. This generates vibrations in the lungs that soften and liquefy the sputum enabling patient to cough out more sputum (16).

Deton Corp. (USA) has come up with one liter bag to collect microdroplets presumed to contain MTB bacilli. These microdroplets are used for diagnosis (3). DNA Genotek Inc. (Ottawa, Canada) has developed OMNIgene® SPUTUM to liquefy and decontaminate the sputum. This facilitates its transport without cold chain for further diagnosis via microscopy, culture, Xpert® MTB/RIF and other molecular diagnostic protocols. prepIT® MAX technology from DNA Genotek Inc. (Ottawa, Canada) enables release of nucleic acid from MTBC cells via chemical lysis (3).

PrimeStore Molecular Transport Medium® (PS-MTM®) from Longhorn Vaccines & Diagnostics (USA) provides a NAAT enabled solution to inactivate, lyse and stabilize MTB DNA without cold chain. The genetic material produced can be utilized for range of other TB tests. PS-MTM® has further diversified into range of products like PrimeSwab™ and PS-MTM® for sample preparation, PrimeSwab™ and PS-MTM® for DNA extraction (PrimeXtract™) and PrimeMix® for real time PCR assay. Molzym GmbH & Co. KG (now Molzym) has launched the MTB-DNA Blood kit (25 to 50 reactions) that removes excess human genomic DNA and concentrates MTB DNA (3).

Tuberculosis may remain latent in an infected patient before reactivation. This is one amongst the most challenging aspect in TB diagnostics. The advent of ELISPOT assay has provided promise to test for latency. This test involves the quantification of interferon production from whole blood or peripheral blood mononuclear cells (PBMC) stimulated with either purified protein derivative (PPD) or specific antigens (17-19). Currently available QuantiFERON®-TB Gold Test (Cellestis Ltd, WA, USA) for latent and active TB utilizes ESAT-6 or CFP-10 antigens (absent from BCG vaccine) (Table 4). Although the assay got the FDA approval it is yet to be commercialized. MPB skin patch test is another test with very high specificity and sensitivity (20-21). This test is free from reader bias, do not give false positive results due to booster phenomena, free from the effect of prior BCG vaccination and require only one patient visit to give sample and get result (22) (Table 4).

Table 4. Culture based phenotypic assays

Genotypic methods target the nucleic acid of the mycobacterium. Ribosomal rRNA, which has universal occurrence, abundance, and conserved specific sequences can be a very useful genetic marker. Accuprobe (Gen-Probe Inc) detection method targets rRNA. This method can detect M tuberculosis complex, M avium, M intracellulare, M avium complex, M gordonae, and M kansasii. Accuprobe is simple and rapid with accuracy around 90%. However, it covers a limited range of species (Table 4). Peptide nucleic acids (PNA) based assay provides another approach to differentiate M. tuberculosis complex from NTM cultures²⁹. PCR based restriction enzyme analysis of the hsp65 also provides an economical alternative for mycobacterium detection from culture. However, technique is complicated and labor intensive (22,30). Similarly, sequencing 16sRNA can also provide rapid detection of mycobacterium but it is labor intensive (22, 31).

Detection of antibodies like lipoarabinomannans (LAM, cell wall component of Mycobacteria), A60, 38Kd and 16 Kd against MTB is a simple and economical method for EPTB detection. However, inconsistency in results as covered by meta-analyses and systematic reviews hampers its commercialization (3,32-33).

Determine™ TB LAM Ag rapid assay (Alere Inc.) is an immune-chromatographic strip that detects LAM antigen in urine. Surprisingly, a ‘negative result’ does not rule out the possibility of TB. Test is also vulnerable to ‘false positives’ due to contamination with dust or faeces. For test to be accurate, the CD4 counts/mm3 should not be less than 100 and best results are obtained with CD4 counts around 200 counts/mm (3, 32-35). It detects NTMs in addition to MTBC along with the site of mycobacterial infection (Table 5).

Table 5. Immunological biomarkers

Analysis of metabolic pathways for TB detection is another breakthrough in technology that is rapid, safe and useful for TB diagnosis in children and HIV patients. Maiga et al (2012) (6) have developed Rabbit urease breath test to analyze the sensitivity and specificity of urease based detection of M. tuberculosis (6). This test involves administration of urea labeled with isotope as substrate. Urea metabolized to ₁₃C-CO₂ can be detected in exhaled breaths using IR spectrometers. Signal shows good relation with bacterial load that served well for diagnostic purposes and treatment monitoring (6). The test is under clinical trial to evaluate its efficacy.

BACTEC TB 460 radiometric system from Becton Dickinson, Sparks, MD, USA facilitates rapid detection of M. tuberculosis thus expediting evidence based therapy (6). Liquid crystal display (LCD) based TB Breathalyser from Rapid Biosensor Systems Ltd takes less than 4 minutes to detect active TB bacilli based on the MTB specific antigen in sputum and gives output as TB positive or negative (2).

PCR is commonly used for nucleic acid amplification tests (NAAT). NAAT can detect single copy of nucleic acid in a clinical sample. However, presence of PCR inhibitors in samples can significantly compromise the sensitivity. Also, processing of samples may lead to the loss of nucleic acids (38-42). This technique is further complemented by fluorescent probe based detection or melts curve analysis, i.e. real time PCR (RT-PCR). PCR based protocol can differentiate between wild type and mutated sequence. Other methodologies include ligase chain reaction (LCR), strain displacement amplification (SDA), loop-mediated isothermal amplification (LAMP) and transcription mediated amplification (TMA) (43-45). These molecular techniques simultaneously detects resistance genes as well, e.g. DNA probe and DNA sequencing of MTB gene such as catalase (katG) or RNA polymerase (rpoB).

Mutations in katG and rpoB lead to the resistance to INH and RIF respectively. NAAT based methods are more suited for academic / research purposes than for commercial clinical purpose.

Whole genome sequencing (WGS) is recently applied to analyze molecular evolution of M. tuberculosis Beijing strain and appearance of an Iberico-American strain in Tibet. WGS provides greater resolution when compared to mycobacterial interspersed repetitive unit-variable-number tandem repeats (MIRU-VNTR, the existing protocol for molecular genotyping of circulating strains) (75-77).

QIAxel automated genotyping system from Qiagen provides high-throughput genotyping of MTBC hence providing epidemiological informations like prevalence or introduction of a particular genotype. This protocol targets variable sized regions of MTB genomic DNA using

MIRU-VNTR approach. This platform is further improvised to detect RIF and INH resistant alleles (80-81).

Unlike other bacterial pathogens that require drug resistance through plasmid acquisition, transposon or phage mediated elements, M. tuberculosis acquire drug resistance through chromosomal mutations particularly, single-nucleotide polymorphisms (SNPs) (80-81). This makes identification of drug resistant alleles difficult particularly for those lying outside of drug resistant hotspots, e.g. RIF resistance-determining region for RIF or the alleles that lie outside of rpoB and associated with resistance (77). Newer approaches to detect drug resistance using Next-generation-sequencing (NGS) involves PZA resistance via deletions, STR, INH and screening allelic variation associated with resistance to novel drugs (80,82-85). NGS is very useful in determining whether the case of relapse is because of infection with new strain or incomplete treatment (86). WGS is already applied to the patients with MDR-TB (87-88). McKinsey has even listed NGS as one of the 12 technologies that will transform our lives.

There is a co-evolution of sort between TB bacilli invading to newer regions and newer techniques / protocols as a counter offensive. Many such existing protocols are improvised and upgraded in response to the challenges and threat posed by rapidly emerging and evolving drug resistant TB bacilli. Following protocols / platforms are under development to spearhead the fight of humanity against TB:

A combine of nano- and biosensing technology is being evaluated for its use as a portable, rapid on-the-spot biosensor for MTB detection. For instance, detection of trans-renal DNA opens up new avenues for TB detection at molecular level. This technique is not yet commercialized due to the problems in the development of TB detection / readout assays.

High-resolution melting (HRM) curve analysis coupled with closed-tube RT-PCR provides a good screening method with a positive predictive value (PPV) of 100% and negative predictive value (NPV) of 99.9.%, for high throughput screening of large number of specimens in any TB laboratory. Improvized amplification techniques in conjunction with ‘molecular beacon’ approach (LATE-PCR) offer future advancements, particularly in drug resistance analysis (90-91).

IR spectrophotometer based detection of isotopic urea in exhaled breath is another future technique that holds promise. It utilizes MTB urease as a bacterial virulence factor (6).

A biophotonic detection platform is being developed that utilizes reporter enzyme fluorescence to detect β-lactamase produced by MTB. This innovative technology is now being adapted for point of care (POC) use.

Becton Dickinson is in the process of designing an automated platform that will do the staining of AFB in sputum in addition to imaging. This platform will provide the sensitivity similar to the fluorescent microscopy at the rate of processing around 40 slides per day with slide scanning algorithm for scoring (2).

Xiamen Zeesan Biotech Co. Ltd (China) has developed MeltPro® Drug-Resistant TB Testing Kits as RT-PCR enabled detection of resistant alleles to RIF and INH. It is undergoing further improvisation to detect alleles resistant to ethambutol (EMB), streptomycin (STR) and fluoroquinolones (FLQs) (52-53, 92).

Genedrive® MTB assay is upgraded to include genotyping MDR-TB through RIF resistance in addition to MTBC diagnosis. This assay involves manual preparation of samples followed by RT-PCR115 (93-94).

The Northwestern Global Health Foundation (USA) in partnership with Quidel Inc. (USA) is coming up with an integrated MTBC assay. The assay involves lysing MTBC cells via heat block followed by purifying DNA using immiscible oil interface (68).

Q-POC™ platform, under development from QuantuMDx (UK) offers diagnosis for PTB.

The assay involves coughing the sputum in a cup that contains reagents for its liquification and decontamination. This is followed by concentration of MTBC cells via paramagnetic bead concentration step and quick nucleic acid amplification (4 minutes) via isothermal asymmetric PCR (2).

Tangen Biosciences Inc. (USA) is going to launch a fully integrated TB detection platform. This platform will have provision for sputum collection in a cup to capture MTB cells followed by lysis and isothermal amplification via LAMP. It comes up with the storage capacity of 10,000 results (2).

Accurate and timely diagnosis of TB is the only requirement for its control and management. Currently, TB diagnosis and treatment is plagued by factors, like inability to diagnose correctly latent and active TB; lack of precise diagnostic protocols for EPTB, pediatric TB and TB associated with HIV; prolonged diagnostic protocol aggravating the otherwise curable infection and escalating it to MDR/XDR-TB; absence of readily available testable biomarkers; Inaccessibility to the POC diagnostic facilities in the peripheral laboratories. These are further challenged by the rapidly emerging resistant strains of TB bacilli in the regions where it is thought to be eradicated or where it was non -existent historically. Need of the hour is the tools and techniques that are accessible, accurate, rapid, economical and most importantly rapid as time lag in diagnosis aggravates the infection. Although TB diagnostics have seen tremendous upsurge in the newer technologies and protocols from imaging, molecular NAAT based, immunological, culture based phenotypic methods, array based to genome sequencing protocols. Unfortunately, they offer promise in some aspect but lack or have severe limitation in other aspects, which limits their applicability on ‘one -shoe-fit-all’ basis. Another, problem with newer protocols is the reluctance of user groups to embrace it either due to lengthy approval mechanism, lack of WHO approval in Toto or simply the cost and expertise involved. There is urgent need to evolve the mechanism to accept the newer protocols on the emergency basis. As an extension to the problem, many small companies with promising diagnostic kits (in terms of good performance on heterogeneous set-ups) do not have sufficient funds to market their products or scale up their production. Also, most of the newer technologies are accepted in the developed countries but there is a need to evolve the technologies compatible to the requirement of resource limited developing countries. There is a dire need of all inclusive diagnostic protocols with applicability in wide spectrum of patients and set-ups with high sensitivity and specificity. A dynamic understanding of existing and forthcoming technologies is key in facilitating access to appropriate TB diagnostic tools through market-based interventions. As such, this review is intended to be a living document, updated as the TB diagnostics market evolves, to highlight potential.

We thankfully acknowledged the Research grant from DST-NRDMS and UGC Startup, Government of India. Funder has no direct or indirect role in plan, design or writing of the review. Grant was solely for research in TB epidemiology. PS: Planning of Review article, literature search, collection and writing. VKS: Literature search, collection of study materials and review. RSK: Literature search and collection of study materials.