Drug resistant tuberculosis: A diagnostic challengeM Dash
Department of Microbiology, Maharaja Krishna Chandra Gajapati Medical College and Hospital, Berhampur, Odisha, India
Correspondence Address: Source of Support: None, Conflict of Interest: None DOI: 10.4103/0022-3859.118038
Source of Support: None, Conflict of Interest: None
Tuberculosis (TB) is responsible for 1.4 million deaths annually. Wide-spread misuse of anti-tubercular drugs over three decades has resulted in emergence of drug resistant TB including multidrug-resistant TB and extensively drug-resistant TB globally. Accurate and rapid diagnosis of drug-resistant TB is one of the paramount importance for instituting appropriate clinical management and infection control measures. The present article provides an overview of the various diagnostic options available for drug resistant TB, by searching PubMed for recent articles. Rapid phenotypic tests still requires days to weeks to obtain final results, requiring biosafety and quality control measures. For newly developed molecular methods, infrastructure, training and quality assurance should be followed. Successful control of drug resistant TB globally will depend upon strengthening TB control programs, wider access to rapid diagnosis and provision of effective treatment. Therefore, political and fund provider commitment is essential to curb the spread of drug resistant TB.
Keywords: Diagnosis, drug resistant, rapid, tuberculosis
The impact of tuberculosis (TB) can be devastating even today, especially in developing countries suffering from high burdens of both TB and human immunodeficiency virus (HIV). In 2010, there were 8.8 million new cases of TB globally, causing 1.4 million deaths.  TB is a major public health problem in India, which accounts for one-fifth of the global TB incident cases. Each year nearly 2 million people in India develop TB, of which around 0.87 million are infectious cases.  It is estimated that annually around 2,80,000 (23/1,00,000 population) Indians die due to TB.  Drug resistance has enabled it to spread with a vengeance. The prevalence of multidrug-resistant tuberculosis (MDR-TB) and extensively-drug resistant TB tuberculosis (XDR-TB) are increasing throughout the world both among new TB cases as well as among previously treated ones.  Accurate and rapid diagnosis of drug-resistant TB is one of the paramount importance for instituting appropriate clinical management and appropriate infection control measures. , Fortunately, the past few years have seen an unprecedented level of funding and activity focused on the development of new tools for diagnosis of drug resistant TB. This should go a long way in helping us arrest the spread of the disease.
Sources and method included PubMed search for recent articles using MeSH terms "TB" and "resistance" and "diagnosis." Furthermore, World Health Organization (WHO) reports and national guidelines were used. The inclusion criteria for selected articles were based on relevance to the purpose of the review.
MDR-TB is a form of TB caused by a strain of Mycobacterium tuberculosis (MTB) resistant to the most potent first line anti-TB drugs, i.e., isoniazid (INH) and rifampicin (RIF). It has been estimated that India and China account for nearly 50% of the global burden of MDR-TB cases.  Approximately, 5% of all pulmonary TB cases in India may be MDR. MDR rates are low in new, untreated cases. The incidence in such cases ranges from 1% to 5% (mostly < 3%) in different parts of India. , However, during the last decade, there has been an increase in reported incidences of drug resistance in category II TB cases, particularly among those treated irregularly or with incorrect regimens and doses. In such cases, the incidence of MDR-TB varies from 11.8% to 47.1%. 
XDR-TB, is defined as TB caused by a strain of MTB that is resistant to RIF and INH as well as to any member of the quinolone family and at least one of the second line anti-TB injectable drugs, i.e. kanamycin, capreomycin or amikacin. XDR-TB was first described in 2006. Since then, there have been documented cases in 77 countries world-wide by the end of 2011.  The global prevalence of XDR-TB has been difficult to assess. The prevalence of XDR-TB has been reported from India, which varies between low, i.e., 2.4% and as high as 21.1% in HIV infected persons suffering from MDR-TB. , Treatment outcomes are significantly worse for patients with XDR-TB, compared with patients with drug-susceptible TB or MDR-TB. , In the first recognized outbreak of XDR-TB, 53 patients in KwaZulu-Natal, South Africa, who were co-infected with XDR-TB and HIV, survived for an average of 16 days, with mortality of 98%.  XDR-TB raises concerns of a future TB epidemic with restricted treatment options and jeopardizes the major gains made in TB control.
Totally, drug resistant TB (TDR-TB) or extremely drug resistant TB is resistant to all first line and second line anti-tubercular drugs. The detection of four Mumbai cases, which were resistant to all first line and second line drugs.  This kind of rapid progression of drug resistance from MDR, to XDR and TDR-TB underlines the need for rapid and accurate diagnosis of drug resistant TB.
Molecular basis of drug resistance
RIF acts by binding to the beta-subunit of the ribonucleic acid (RNA) polymerase (coded for by the rpoB gene), inhibiting RNA transcription. Subsequent deoxyribonucleic acid (DNA) sequencing studies have shown that more than 95% of RIF resistant strains have mutations in an 81-base pair region (codons 507-533) of the rpoB gene. INH inhibits enoyl-acyl carrier protein (ACP)-reductase (coded by the inhA gene), which is involved in mycolic acid biosynthesis. INH is also a "pro-drug," which is converted to its active form by the catalase-peroxidase enzyme (coded by katG gene). Resistant mutants can be due to different regions of several genes, including binding of activated INH to its inhA target, the activation of the pro-drug by katG or by increased expression of the target inhA. Point mutations in codon 315 of the katG gene have been found in 50-90% of high-level INH resistant strains while 20-35% of low-level INH resistant strains have been reported mutations in inhA regulatory region and 10-15% have mutations in the ahpC-oxyR intergenic region (often together with katG mutations in other regions). 
MTB is an extremely slow growing organism. Using the standardized drug susceptibility testing (DST) with conventional methods, 8-12 weeks are required to identify drug resistant TB on solid media (i.e. Lowenstein-Jensen [LJ] medium). In general, these methods assess inhibition of MTB growth in the presence of antibiotics to distinguish between susceptible and resistant strains. As the results usually take weeks, inappropriate choice of treatment regimen may result in death such as in case of XDR-TB (especially in HIV co-infected patients). In addition, delayed diagnosis of drug resistance results in inadequate treatment, which may generate additional drug resistance and continued transmission in community. The most common medium for the agar proportion method in resource limited countries is LJ medium; however, the Clinical and Laboratory Standards Institute, (CLSI) considers this medium to be unsuitable for susceptibility testing due to uncertainty about the potency of drugs following inspissation and also because components present in the eggs or the medium may negatively affect some drugs. Both Centers for Disease Control and Prevention and CLSI recommend that Middle brook 7H10 agar supplemented with oleic albumin dextrose catalase (OADC) be used as the standard medium for the agar proportion assay.
Rapid phenotypic methods for diagnosis of drug resistant TB
Rapid automated liquid based culture and susceptibility tests
Automated liquid culture systems such as Bactec radiometric system (Bactec 460TB; Becton Dickinson, USA), non-radiometric systems MB/BacT ALERT (BioMerieux, France), Versa Trek (Trek Diagnostic System, USA) and mycobacteria growth indicator tube (MGIT 960; Becton Dickinson, USA) are more sensitive and shorter turnaround time than solid media cultures. These are also less labor intensive and therefore, less vulnerable to manual errors. But, automated systems still requires few weeks to obtain final results.  Also these instruments are costly, require maintenance and can be extremely difficult for most public health laboratories in developing countries.
Nitrate reductase assay (Griess method)
The NRA is a liquid or solid medium technique that measures nitrate reduction by members of the MTB complex to indicate growth and to indicate resistance to INH and RIF. Since the NRA method uses nitrate reduction as a sign of growth, results are detected earlier than by examination of microcolonies on solid medium. However, the Griess reagent kills the organisms when added to the tubes, so multiple tubes must be inoculated if further testing is necessary. In addition, not all members of the MTB complex reduce nitrate, so the presence of nitrate-negative acid fast bacilli (AFB) may require further testing.
Thin layer agar cultures and TK medium
Thin layers of middle brook 7H11 solid agar medium are used to detect microcolonies by conventional microscopy. It can be adapted for the rapid detection of drug resistance directly from sputum samples, but requires average turnaround time of 11 days. 
Newly developed test such as TK medium (Salubris Inc., USA) is a colorimetric system that indicates growth of mycobacteria by changing the color of the growth medium. Metabolic activity of growing mycobacteria changes the color of the culture medium and this allows for an early positive identification before bacterial colonies appear. Unfortunately, there is insufficient published evidence on the field performance of these tests in developing countries. 
Microscopic observation drug susceptibility assay
The MODS assay is based on characteristic cord formation of MTB that can be visualized microscopically ("strings and tangles" appearance) in liquid medium with or without antimicrobial drugs (for DST).  The test sensitivity is better than traditional methods using LJ media with a turnaround time of 7 days for culture and drug susceptibility testing (C-DST) for INH and RIF. It is cheap, simple and fairly accurate.  Biosafety level-3 facilities are required if the plates are opened for further testing, such as for the confirmation of the identification of TB. Laboratories using MODS assay require a functioning biosafety cabinet, a safety centrifuge, an incubator, an inverted light microscope for observation of mycobacterial growth and supplemented liquid media (Middle brook 7H9 broth with OADC and PANTA).
Phage based assay
Phage amplification-based test (FAST Plaque-Response, Biotech Laboratories Ltd. UK) has been developed for direct use on sputum specimens. Drug resistance is diagnosed when MTB is detected in samples that contain the drug (i.e., RIF). When these assays performed on MTB culture isolates, they have shown high sensitivity and variable specificity, but the evidence is lacking about the accuracy when they are directly applied to sputum specimens.  It also requires high standards of biosafety and quality control.
Luciferase reporter phages are genetically-modified phages containing the flux gene encoding firefly luciferase. This catalyzes a reaction producing light in the presence of the luciferin substrate and adenosine triphosphate; light is only produced in the presence of viable mycobacteria. Detection of light released from viable mycobacteria can be achieved by a luminometer or photographic film within 2-4 days from culture.  The luminometer readout is more sensitive and enables quantification of results while the use of Polaroid photographic film offers a lower-tech approach with lower sensitivity.
Rapid molecular methods for diagnosis of drug resistant TB
Since the publication of genome details of MTB H37Rv strain in 1998, these have been utilized in development of nucleic acid amplification (NAA) tests for diagnosis of drug resistant TB. A number of NAA tests are now available, manual and automated, commercial and in the house, with varying performance characteristics. Real-time polymerase chain reaction (RT-PCR) and line probe assays (LPAs) have been commercialized and widely used in clinical laboratories merit special mention, detailed below.
Molecular tools are based on identification of specific mutations responsible for drug resistance, which are detected by the process of NAA in conjunction with electrophoresis, sequencing or hybridization. Direct sequencing techniques such as RT-PCR that uses wild-type primer sequences to amplify genes and enable the use of specific probes to identify mutations. Recently introduced semi-quantitative nested RT-PCR, which integrates and automates sample processing and simultaneously detects MTB and RIF resistance within the single-use disposable cartridges. A study examined 1,730 patients with suspected drug-sensitive or MDR pulmonary TB across Peru, Azerbaijan, South Africa and India. There was sensitive detection of MTB and RIF resistance directly from untreated sputum in < 2 h with minimal hands-on time.  The WHO has recently supported the use of this system as an initial diagnostic test in respiratory specimens of patients with high clinical suspicion of having TB or who could be MDR.  These tests are very expensive require adequate maintenance and calibration of the equipment. Further, the specificity for RIF resistance in populations where MDR-TB is rare requires confirmation by repeat testing and thus increases cost.
LPAs are a family of novel DNA strip tests that use both PCR and reverse hybridization methods. In these assays, a specific target sequence is amplified and applied on nitrocellulose membranes. Specific DNA probes on the membrane hybridize with the amplified sequence applied on it. Color conjugates make the amplified target sequences appear as colored bands. These tests have been designed to identify MTB and simultaneously detect genetic mutations related to drug resistance both from clinical samples as well as culture isolates. These tests able to identify MTB complex and simultaneously detect genetic mutations in the rpoB gene region related to rifampin resistance. The LPA strip consists of 10 oligonucleotide probes: One is specific for the MTB complex, five are partially overlapping wild-type probes that span the region at positions 509-534 of the rpoB gene and four probes are specific for amplicons carrying the most common rpoB mutations (D516V, H526Y, H526D and S531L). According to a recent review, the sensitivities of the LPA are above 90% for clinical isolates with 99-100% specificity [Table 1]. It identifies MTB complex and simultaneously detects mutations in the rpoB gene as well as mutations in the katG gene for high-level INH resistance. These tests can also detect mutations in the inhA gene for low-level INH resistance and mutations in the gyrA, rrs and embB genes for 2 nd line anti-tubercular drugs fluoroquinolones, aminoglycosides and ethambutol respectively [Table 2].  The newly developed assay may represent a reliable tool for detection of fluoroquinolones, amikacin, capreomycin and ethambutol resistance. LPA strip can use culture isolates and smear positive sputum as specimen and provide results in 1-2 days. A recent laboratory evaluation study from South Africa estimated the accuracy of the GenoType MTBDRplus assay performed directly on AFB smear-positive sputum specimens.  It showed high sensitivity, specificity, positive and negative predictive values for detection of RIF and INH resistance [Table 1]. However, a meta-analysis on this assay found that sensitivity estimates for INH resistance were comparatively modest. , In general, LPAs are expensive and require dedicated equipments, reagents and facilities. Molecular drug resistance testing has technical limitations, i.e., LPAs can detect only well characterized drug resistance alleles, not every drug resistance allele can be discriminated via current tests and silent mutations (which do not confer drug resistance) can be detected by probes leading to misclassification of drug resistance. In addition, molecular tests cannot determine the proportion of drug resistant bacteria within a mixed population of cells (i.e., wild type and drug resistant). Cross contamination with amplicons generated from previous tests can be problematic especially when the tests have been employed in laboratories without appropriate staff training and quality control.
Diagnostic testing algorithm for drug resistant TB
Laboratory policies and testing of patients suspected of having drug-resistant TB depend on the local epidemiology, local treatment policies, the existing laboratory capacity, specimen referral and transport mechanisms and human and financial resources. Mycobacterial culture (solid or liquid) and identification of MTB provide a definitive diagnosis of TB. Culture also provides necessary isolates for conventional DST. Thus LPAs, RT-PCRs, MODS and NRA can be used in conjunction with culture (solid and liquid) and DST [Figure 1]. 
Revised national tuberculosis control program and diagnosis of drug resistant TB
The RNTCP plans to strengthen laboratory capacity for MTB C-DST and LPA across India. To date, 35 RNTCP accredited labs including 14 LPA and 4 liquid culture labs in public and private sectors are serving patients while another 30 labs are under the process of up-gradation and accreditation under RNTCP, most of them include LPA and liquid culture for first and second line drugs.  In a policy statement released in June 2008, the WHO endorsed the use of LPA for rapid screening of patients at risk of MDR-TB and recommended the use of LPAs only on culture isolates and smear-positive sputum specimens. It is not recommended as a complete replacement for conventional C-DST.  As of January 2012, diagnosis of XDR-TB can only be confirmed at three laboratories in India, which are quality assured for second line anti-TB DST of flouroquinolones and injectable drugs. These are the National Reference Laboratories (NRL) of tuberculosis research centre/national institute for research in tuberculosis (TRC/NIRT) Chennai, National Tuberculosis Institute (NTI) Bangalore and LRS Institute, New Delhi. Routine fluoroquinolone and injectable DST (i.e., XDR-TB diagnosis) on all MDR-TB patients at the beginning of treatment has been recommended by the RNTCP National Laboratory Committee in 2011, but the capacity to conduct that testing is not yet present in most C-DST laboratories used by RNTCP. Capacity building for second line DST is being undertaken through these NRLs. 
The principles of TB control are important for the prevention of drug resistant TB; these include prompt case detection, provision of curative treatment and prevention of transmission. The WHO's "Stop TB"  directly observed treatment short course (DOTS) strategy includes supervision and support of treatment, although there is little evidence that directly observed treatment alone improves cure rates.  Ineffective drug treatment is a strong risk factor for acquired drug resistance and proper administration of anti-tubercular drugs is critical to reduce this risk. An enhanced DOTS program, DOTS-plus has been developed for managing MDR-TB in resource limited settings.  This program recommends additional facilities for C-DST for detection of drug resistant TB and provision of appropriate second line anti-tubercular drugs.
As most of the resistance arises from either inadequate or inappropriate treatment of active disease, prevention of active disease indirectly prevents drug resistance. Contact tracing including family members and health-care workers, those who are at risk of acquiring drug resistant TB  is one of the fundamental measures to detect active disease, must be aligned to all health services. Prevention of transmission in healthcare settings is difficult in places where resources are limited with no isolation facilities; one approach is to manage cases with similar resistance profiles in segregated groups. However, simple, low-cost interventions, such as opening windows and doors for adequate ventilation, can reduce transmission of TB. 
Successful control of drug resistant TB globally will depend on strengthening TB control programs, wider access to rapid C-DST along with emerging molecular diagnostic technologies and provision of effective treatment. Rapid and accurate diagnosis of drug resistant TB will require massive scaling-up of C-DST capacity and simultaneous use of molecular assays. Furthermore, all molecular tests require DNA extraction, gene amplification and detection of mutations and are, therefore, relatively expensive, demand resources and skills. These are usually unavailable in developing countries where rates of drug resistant TB are high. The challenge, therefore, is to not only develop new tools, but to also make sure that benefits of promising new tools actually reach the populations that need it most, but can least afford them. Therefore, political and fund provider commitment is essential to curb the spread of drug resistant TB.
[Table 1], [Table 2]