Chapter 19: Drug Resistance and Drug Resistance Detection
by Anandi Martin and Françoise Portaels
Drug resistance in tuberculosis (TB) is a matter of great concern for TB control programs since there is no cure for some multidrug-resistant TB (MDR-TB) strains of M. tuberculosis. There is concern that these strains could spread around the world, stressing the need for additional control measures, such as new diagnostic methods, better drugs for treatment, and a more effective vaccine. MDR-TB, defined as resistance to at least rifampicin (RIF) and isoniazid (INH), is a compounding factor for the control of the disease, since patients harboring MDR strains of M. tuberculosis need to be entered into alternative treatment regimens involving second-line drugs that are more costly, more toxic, and less effective.
Moreover, the problem of extensively drug resistant (XDR) strains has recently been introduced. These strains, in addition to being MDR, were initially defined as having resistance to at least three of the six main classes of second-line drugs (aminoglycosides, polypeptides, fluoroquinolones, thioamides, cycloserine, and para-aminosalicylic acid) (CDC 2006). More recently, at a consultation meeting of the World Health Organization (WHO) Global Task Force on XDR-TB, held in Geneva, a revised laboratory case definition was agreed: “XDR-TB is TB showing resistance to at least rifampicin and isoniazid, which is the definition of MDR-TB, in addition to any fluoroquinolone, and to at least 1 of the 3 following injectable drugs used in anti-TB treatment: capreomycin, kanamycin and amikacin.” (http://www.who.int/tb/xdr/taskforcereport_oct06.pdf). XDR-TB now constitutes an emerging threat for the control of the disease and the further spread of drug resistance, especially in HIV-infected patients, as was recently reported (Gandhi 2006). For this reason, rapid detection of drug resistance to both first- and second-line anti-tuberculosis drugs has become a key component of TB control programs.
19.2. Drug resistance surveillance
19.2.1. Benefits and recommendations
The surveillance of drug resistance in TB is a critical component of the monitoring system of the disease. The benefits of drug resistance surveillance are numerous and include the strengthening of laboratory networks, the evaluation of TB control program performance, and the collection of important data for appropriate treatment strategies. Furthermore, global drug resistance surveillance identifies areas of high resistance, warning the health authorities to initiate the appropriate correction measures. To adequately establish drug resistance surveillance at a national level, three recommendations have been provided: the sampled specimens should be representative of the patients from the area under study and the sample size should be statistically determined to allow standard epidemiological analysis; the patient’s history should be obtained and medical records carefully reviewed to determine whether the patient has received previous treatment in order to distinguish primary from acquired resistance; and the laboratory techniques used for determining the drug susceptibility to anti-tuberculosis drugs should be selected from those that are internationally recommended (WHO/IUATLD 1998). In 1996, the WHO and the International Union Against Tuberculosis and Lung Disease (IUATLD) launched the Global Project on Drug Resistance Surveillance based on data collected and reported by an international network of laboratories acting as Supranational Reference Laboratories. The network includes twenty-six Supranational Reference Laboratories distributed in the five WHO regions and is coordinated by the Prince Leopold Institute of Tropical Medicine in Antwerp, Belgium.
Download of the entire textbook
(Tuberculosis 2007, 687 pages, PDF, 8.3 MB)
19.2.2. Global trends in drug resistant tuberculosis Since the establishment of the WHO/IUATLD Global Project on Anti-tuberculosis Drug Resistance Surveillance, three global reports have been produced (WHO 1997, 2001, 2004). The first two reports covered data from 35 and 58 settings respectively. The main conclusions of those two reports were that drug-resistant TB was present in all settings surveyed, MDR-TB was identified in most settings, and good TB control practices were associated with lower or decreasing levels of resistance. The third and last report available, published in 2004, covers data from 77 settings and had the main goal of expanding knowledge of the prevalent global patterns of resistance and exploring trends in resistance over time. The data were collected between 1999 and 2002 and represented 20 % of the total global number of new smear-positive TB cases. This third report also contributes to address two issues not thoroughly dealt with in previous reports: the importance of conducting surveillance on re-treatment cases, and stressing the issue of the role of the laboratory in TB control (WHO 2004, Aziz 2006). The prevalence of drug resistance among new patients is a very important indicator for a TB control program. The prevalence of resistance among previously untreated patients also reflects program performance over a long period of time and indicates the level of transmission within the community. The prevalence of drug resistance among patients with a history of previous treatment, on the other hand, has received less attention, since surveillance of this population is more complex. Re-treatment patients are a heterogeneous group composed of chronic patients, those with treatment failure, those who have relapsed, and those who have returned after defaulting. Sometimes this population represents more than 40 % of smear-positive cases. The prevalence of drug resistance varies greatly among subgroups of this population. Chronic cases and treatment failures are at a greater risk of having resistant and MDR-TB. Relapses and default patients are more likely to have drug resistance than new cases, but are almost always at a lower risk for MDR-TB than failures and chronic cases. One of the recommendations of the last report is that all subgroups of re-treatment cases be notified separately and their outcomes reported; furthermore, surveillance of resistance should be conducted on a representative sample of this population. The second issue stressed in the third resistance report is that of the role of the laboratory. While laboratory services are fundamental for TB control, they are often the weakest components of the system. The importance of the laboratory in the control of TB should be recognized and they should be able to perform sputum smear microscopy, culture, and drug susceptibility testing of a high quality as standard components of TB control. Culture and drug susceptibility testing should be performed by national reference laboratories. Recognizing the pressing need to improve laboratory performance, a Subgroup on Laboratory Capacity Strengthening was established within the DOTS Expansion Working group in 2002 (Portaels 2006). The major objective of the subgroup is to assist high-TB burden and other countries in strengthening TB laboratory capacity and to provide high quality diagnostic services. In this third report, data were collected through routine or continuous surveillance of all TB cases (in 38 settings) or from specific surveys of sampled patients (in 39 settings). These were reported on a standard reporting form, either annually or on completion of the survey (WHO 2004). The results show that in new TB cases with data available from 75 settings (55,779 patients) the prevalence of resistance to at least one drug (any resistance) ranged from 0 % in some Western European countries to 57.1 % in Kazakhstan (median = 10.2 %). Median prevalence of resistance to individual drugs was: streptomycin (SM), 6.3 %; INH, 5.9 %; RIF, 1.4 %; and ethambutol (EMB), 0.8 %. Prevalence of MDR-TB ranged from 0 % in eight countries to 14.2 % in Kazakhstan and Israel (median = 1.1 %). The highest prevalences of MDR-TB were observed in Tomsk Oblast (Russian Federation) (13.7 %), Karakalpakstan (Uzbekistan) (13.2 %), Estonia (12.2 %), Liaoning Province (China) (10.4 %), Lithuania (9.4 %), Latvia (9.3 %), Henan Province (China) (7.8 %), and Ecuador (6.6 %). Trends in drug resistance were determined in 46 settings (20 with two data points and 26 with at least three). Significant increases in prevalence of any resistance were found in Botswana, New Zealand, Poland, and Tomsk Oblast (Russian Federation). Cuba, Hong Kong SAR, and Thailand reported significant decreases over time. Tomsk Oblast (Russian Federation) and Poland reported significantly increased prevalences of MDR-TB. Decreasing trends in MDR-TB were observed in Hong Kong SAR, Thailand, and the USA. Among previously treated cases with data available from 66 settings (8,405 patients) the median prevalence of resistance to at least one drug (any resistance) was 18.4 %, with the highest prevalence being 82.1 % in Kazakhstan. Median prevalence of resistance to individual drugs was: INH, 14.4 %; SM, 11.4 %; RIF, 8.7 %; and EMB, 3.5 %. The median prevalence of MDR-TB was 7.0 %. The highest prevalence of MDR-TB was reported in Oman (58.3 %) and Kazakhstan (56.4 %). Countries of the former Soviet Union had a median prevalence of resistance to the four drugs of 30 %, compared with 1.3% in all other settings. However, these data should be interpreted with caution given the small number of subjects tested in some settings. Trends in drug resistance in this group were determined in 43 settings (19 with two data points and 24 with at least three data points). A significant increase in the prevalence of any resistance was observed in Botswana. Cuba, Switzerland, and the USA showed significant decreases. The prevalence of MDR-TB significantly increased in Estonia, Lithuania, and Tomsk Oblast (Russian Federation). Decreasing trends were significant in Slovakia and the USA. The annual incidence of MDR-TB cases was estimated in 69 settings. In most Western and Central European countries, the estimated incidence was fewer than 10 cases each. Estonia, Latvia, Lithuania and two Oblasts in the Russian Federation were estimated to have between 99 and 248 MDR-TB cases. For Henan and Huber Provinces of China, more than 1,000 cases each were estimated, and for Kazakhstan and South Africa, more than 3,000. The report also evaluated RIF resistance as a predictor of MDR-TB, in order to explore the significance of rapid testing for RIF resistance to identify cases likely to have MDR-TB. The positive predictive value, a function of the sensitivity and specificity of RIF resistance testing and the prevalence of MDR-TB and non-MDR-TB RIF resistance, was highest among previously treated cases in settings with high MDR-TB prevalence and low non-MDR-TB RIF resistance. The report also confirmed that, globally, more isolates were resistant to INH than to any other drug (range 0-42 %). INH and SM resistance were more prevalent than RIF or EMB resistance. Resistance to INH, SM, RIF and EMB was the most prevalent pattern among previously treated cases and the proportions of isolates resistant to three or four drugs were significantly greater than among new cases, suggesting an amplification of resistance. It appears that monoresistance to either INH or SM is the main gateway to the acquisition of additional resistance. Tables 19-1 and 19-2 below show a summary of the prevalence of drug resistance and MDR-TB in new TB cases and previously treated patients, respectively, according to the five WHO regions in the world. Table 19-1: Median prevalence of drug resistance, polyresistance and MDR-TB among new TB cases by region (%) Region Any resistance Polyresistance MDR-TB Africa 7.1 1.3 1.4 Americas 9.7 2.1 1.1 Eastern Mediterranean 9.9 2.5 0.4 Europe 8.4 1.1 0.9 South-East Asia 19.8 4.0 1.3 Western Pacific 11.4 2.5 0.9 Overall median 10.2 1.9 1.1 Adapted from Reference WHO, 2006 Table 19-2: Median prevalence of drug resistance, polyresistance and MDR-TB among previously-treated TB cases by region (%) Region Resistance Polyresistance MDR-TB Africa 16.7 1.8 5.9 Americas 24.6 3.7 7.0 Eastern Mediterranean 63.3 5.8 48.3 Europe 15.9 2.6 4.7 South-East Asia 39.9 7.3 20.4 Western Pacific 32.8 6.1 15.5 Overall median 18.4 3.2 7.0 Adapted from Reference WHO, 2006 19.3. Methods for detection of drug resistance Early detection of drug resistance constitutes one of the priorities of TB control programs. It allows initiation of the appropriate treatment in patients and also surveillance of drug resistance. Detection of drug resistance has been performed in the past by so-called ‘conventional methods’ based on detection of growth of M. tuberculosis in the presence of the antibiotics. However, due to the laboriousness of some of these methods, and most of all, the long period of time necessary to obtain results, in recent years new technologies and approaches have been proposed. These include both phenotypic and genotypic methods. In many cases, the genotypic methods in particular have been directed towards detection of RIF resistance, since it is considered a good surrogate marker for MDR-TB, especially in settings with a high prevalence of MDR-TB. Genotypic methods have the advantage of a shorter turnaround time, no need for growth of the organism, the possibility of direct application in clinical samples, lower biohazard risks, and the feasibility of automation; however, not all molecular mechanisms of drug resistance are known. Phenotypic methods, on the other hand, are in general simpler to perform and might be closer to implementation on a routine basis in clinical mycobacteriology laboratories. The following section describes the phenotypic and genotypic methods as well as the new methodologies recently proposed for drug resistance detection in TB. 19.3.1. Conventional phenotypic methods In general, phenotypic methods assess inhibition of M. tuberculosis growth in the presence of antibiotics to distinguish between susceptible and resistant strains. This is possible since M. tuberculosis isolates from patients never treated before are very uniform in their level of susceptibility, as shown by the narrow ranges of minimal inhibitory concentrations (MIC) of the main anti-tuberculosis drugs (Heifets 1996). The classical definition for a drug resistant M. tuberculosis strain is that it displays a degree of susceptibility significantly lower than that of a wild strain that has never been in contact with the drug (Canetti 1963, Canetti 1969). Phenotypic methods based on cultivation of M. tuberculosis in the presence of antibiotics have been most commonly performed on egg-based or agar-based solid media, and can also be performed as a direct or indirect method. For the direct method, antibiotic-containing and control media are inoculated with a decontaminated and concentrated clinical specimen, while for the indirect method the antibiotic-containing and control media are inoculated with a bacterial suspension of the isolated strain. There are three conventional phenotypic methods for drug susceptibility testing based on solid media: the proportion method, the resistance ratio method and the absolute concentration method (Canetti 1963, Canetti 1969, Kent 1985). More recent methods are based on liquid media including the BACTEC radiometric and the Mycobacterial Growth Indicator Tube methods. The proportion method The proportion method is the most commonly used method worldwide amongst the three methods mentioned above. It allows the precise determination of the proportion of resistant mutants to a certain drug. Briefly, several 100-fold serial bacilli dilutions are inoculated into drug-containing and drug-free (control) media. One of those dilutions should produce a number of colonies that is easy to be counted. The number of colonies obtained in the drug-containing and control media are enumerated and the proportion of resistant mutants is then calculated. When performed in Löwenstein-Jensen medium tubes, the test is first read after 28 days of incubation at 37°C. If the proportion of resistant bacteria is higher than 1 % for isoniazid, rifampicin and para-aminosalycilic acid, or 10 % for the other drugs, the strain is considered resistant and the results are final; otherwise, the test is read again at 42 days of incubation to assess if the strain is susceptible to a certain drug (Heifets 2000). If the test is performed on agar, a Middlebrook 7H10/11 is used and the medium is incubated in a 10 % CO2 atmosphere. Results are interpreted after 21 days of incubation or even earlier if they show the strain to be resistant (Kent 1985). The critical concentrations of the main drugs used in the proportion method are shown in Table 19-3. Table 19-3: Critical concentration of main antibiotics in the proportion method (µg/mL) Antibiotic Löwenstein-Jensen 7H10 agar 7H11 agar Isoniazid 0.2 0.2, 1.0 0.2, 1.0 Rifampicin 40.0 1.0 1.0 Ethambutol 2.0 5.0 7.5 Streptomycin 4.0 2.0 2.0, 10.0 Pyrazinamide 100 – – PAS 0.5 2.0 8.0 Kanamycin 20.0 5.0 6.0 Ethionamide 20.0 5.0 10.0 Ofloxacin 2.0 2.0 2.0 Capreomycin 20.0 10.0 10.0 Cycloserine 40.0 – – Adapted from: Kent 1985; WHO/CDS/TB/2001.288; and NCCLS 2000 The resistance ratio method This method is based on the resistance ratio, which corresponds to the MIC of a test strain divided by the MIC of the drug-susceptible reference strain H37Rv tested at the same time. Thus, it compares the resistance of an unknown strain with that of a standard laboratory strain. For the performance of the test, parallel sets of tubes containing two-fold dilutions of the tested drug are then inoculated with a standardized inoculum of both test and reference strain. Reading of the test is performed after 4 weeks of incubation at 37°C. Tubes containing 20 or more colonies are considered as positive for growth and the MIC is defined as the lowest concentration of drug in the presence of which the number of colonies is lower than 20. An isolate with a resistance ratio value of 2 or less is considered susceptible, while a resistance ratio of 8 or more defines the isolate as resistant (Kent 1985, Heifets 2000). The absolute concentration method This method uses a standard inoculum of the test strain grown in a two-fold dilution drug-containing media and drug-free control. The resistance of a strain is expressed in terms of the lowest concentration of a certain drug that inhibits all or almost all the growth of the strain. The critical concentrations included in the medium are similar to the ones used in the proportion method (see Table 19-3) but the drug concentration considered as ‘critical’ should be determined in each laboratory (Heifets 2000). For the interpretation of the test, the reading is performed after 4 weeks of incubation at 37°C, or at 5-6 weeks if there is not enough growth. A strain is considered to be susceptible if the number of colonies on the drug-containing medium is less than 20 with a 3+ or 4+ (confluent) growth on the drug-free control. The BACTEC radiometric method The radiometric method is based on the commercial system BACTEC TB-460 (Becton Dickinson, Sparks, MD), which uses an enriched Middelbrook 7H9 liquid medium containing 14C-labeled palmitic acid as the sole carbon source (12B vial). Growth of the mycobacteria and consumption of the labeled fatty acid will produce 14CO2 that is detected inside the 12B vial by the BACTEC apparatus and expressed as a growth index. In the presence of a certain drug, susceptibility can be measured by inhibition of the daily increases in the growth index. For the performance of the test, a test vial containing the drug under study and a drug-free control are inoculated with a standard inoculum and incubated at 37°C. The vials are then read in the BACTEC 460-TB apparatus on a daily basis. Since two control vials are inoculated with a 100-fold serial dilution of the inoculum, results can be interpreted as in the proportion method with the 1 % proportion of growth. The BACTEC radiometric method has been approved by the Food and Drug Administration (FDA) of the United States (US) and is also considered to be the ‘gold standard’ for drug susceptibility testing to first-line anti-tuberculosis drugs (Roberts 1983, Heifets 1999). More recently, critical concentrations for second-line drugs have also been proposed and tested successfully for most drugs in a multicenter evaluation (Pfyffer 1999). The major advantage of the BACTEC radiometric method is the capacity to detect drug resistance faster than with the solid media-based methods; the major disadvantage is the cost of the system and the need for disposal of the radioactive waste from used vials. The Mycobacterial Growth Indicator Tube The Mycobacteria Growth Indicator Tube (MGIT) (Becton Dickinson, Sparks, MD) is part of the ‘new generation’ of TB diagnostic tools both in its manual version as well as in its more recently introduced automated format (Pfyffer 1997, Idigoras 2000). It is based on fluorescence detection of mycobacterial growth in a tube containing a modified Middlebrook 7H9 medium together with a fluorescence quenching-based oxygen sensor embedded at the bottom of the tube. Consumption of oxygen in the medium produces fluorescence when illuminated by a UV lamp. In the manual system, for the performance of the test a drug-containing tube and a control tube are inoculated with the standardized mycobacterial suspension and incubated at 37°C (day 0). Starting on the third day (day 2), the tubes are controlled daily with an UV lamp. The presence of an orange fluorescence in the drug-containing tube at the same time as in the control tube or within two days of positivity in the control is interpreted as resistance to the drug; otherwise, the strain is considered to be susceptible. The test is valid if the growth control gives a positive signal until the 14th day of incubation (day 12) (Palomino 1999). The MGIT system in its manual version has also been successfully used as a direct method using decontaminated clinical specimens (Goloubeva 2001). Figure 19-1: MGIT tubes showing a positive and a negative reaction The MGIT has also been recently introduced as an automated system. The BACTEC MGIT960 (Becton Dickinson, Sparks, MD) is based on the same principle of oxygen consumption and a fluorescence signal, but the tubes are incubated and controlled inside the MGIT960 apparatus. For the performance of the test, drug-containing and drug-free control vials are inoculated with a standardized inoculum of the M. tuberculosis isolate and entered into the machine in a special rack-carrier with a printed barcode; this is read by the machine when entering the tubes to identify the test and apply the adequate algorithm for susceptibility or resistance interpretation. All readings are performed inside the machine and the results are printed as susceptible or resistant (Ardito 2001). Many studies have now been published on the application of the MGIT system for the rapid detection of resistance to first- and second-line antituberculosis drugs (Johansen 2004, Rusch-Gerdes 2006). In all these studies, the MGIT system has shown very good results with a high correlation with the conventional methods on solid media and the BACTEC TB-460 system. The BACTEC MGIT960 system has recently been approved by the US FDA for the detection of drug resistance to first-line drugs. Other automated systems, such as those already described in Chapter 14, have been used for the rapid detection of drug resistance in M. tuberculosis, but they have not been used on a routine basis in the clinical mycobacteriology laboratory (Ängeby 2003, Ruiz 2000). Recent developments of phenotypic formats for rapid drug resistance detection will be presented in section 19.3.3 below. 19.3.2. Genotypic methods Genotypic methods for drug resistance in TB look for the genetic determinants of resistance rather than the resistance phenotype, and involve two basic steps: nucleic acid amplification such as polymerase chain reaction (PCR), to amplify the sections of the M. tuberculosis genome known to be altered in resistant strains; and a second step of assessing the amplified products for specific mutations correlating with drug resistance (García de Viedma 2003, Palomino 2005). Desoxyribonucleic acid (DNA) sequencing Sequencing DNA of PCR-amplified products has become the most widely used genotypic method for detecting drug resistance in M. tuberculosis; it is accurate and reliable and it has become the reference standard for mutation detection. It was performed several years ago by manual procedures, but in our days, it is performed with automatic sequencers (Victor 2001). DNA sequencing has been widely used for characterizing mutations in the rpoB gene in RIF-resistant strains and to detect mutations responsible for resistance to other anti-tuberculosis drugs (Telenti 1993, García de Viedma 2003, Jalava 2004). Drug resistance detection in M. tuberculosis has also been described by pyrosequencing technology (Arnold 2005, Jureen 2006). This technology is a short-read (30-50 bp) sequencing technique, which is based on the quantitative detection of pyrophosphate released following nucleotide incorporation into a growing DNA chain (Ronaghi 1999). However, not all molecular mechanisms of drug resistance for M. tuberculosis are known and it would be rather difficult and expensive to implement it routinely for the detection of drug resistance mutations for several drugs (Hazbón 2004). Solid-phase hybridization techniques There are currently two commercially available solid-phase hybridization techniques for the rapid detection of drug resistance in TB: the Line Probe Assay (INNO-LiPA Rif TB Assay, Innogenetics, Ghent, Belgium) for the detection of resistance to RIF and the GenoType MTBDR assay (Hain Lifesciences, Nehren, Germany) for the simultaneous detection of resistance to INH and RIF. The LiPA assay was introduced several years ago and is based on reverse hybridization of amplified DNA from cultured strains or clinical samples to ten probes covering the core region of the rpoB gene of M. tuberculosis, immobilized on a nitrocellulose strip (De Beenhouwer 1995). From the pattern of hybridization obtained, the presence or absence of mutated or wild regions is visualized by a colorimetric reaction and the strain can be considered as resistant or susceptible to RIF (Rossau 1997). Many studies have been conducted on the application of the LiPA assay for detection of RIF resistance; most of them have been performed on M. tuberculosis isolates and just a few have applied the test directly in sputum samples (Jureen 2004, Traore 2006). It has been proposed as a good initial indicator of multidrug resistance with a sensitivity of 98.5 % for detecting RIF resistance (Traore 2000). In a recent systematic review and meta-analysis of studies that applied the LiPA test, 12 of 14 studies performed in isolates had sensitivity greater than 95 % and specificity of 100 %. Four studies that applied LiPA directly to clinical specimens had 100 % specificity, and the sensitivity ranged from 80 % to 100 % (Morgan 2005). In a very recent and large study, not included in the meta-analysis mentioned above, the utility of the LiPA test for detecting RIF resistance was assessed in 420 sputum samples originating from different countries (Traore 2006). There was a 99.6 % concordance between the RIF resistance obtained by culture and by the LiPA test, confirming that with an adequate DNA extraction method, the LiPA test allows rapid detection of resistance to RIF directly from sputum samples. Figure 19-2: LiPA strips showing different mutations The GenoType MTBDR, on the other hand, detects resistance to INH and RIF in culture samples based on the detection of the most common mutations in the katG and rpoB genes respectively (Makinen 2006). It also utilizes PCR and reverse hybridization to probes immobilized on a DNA strip. In a recent study that evaluated the GenoType MTBDR assay in 143 M. tuberculosis isolates, 99 % of the MDR strains were found to have mutations in the rpoB gene and 88.4 % of strains with mutations in the codon 315 of the katG gene were also correctly identified (Hillemann 2005). The correlation with DNA sequencing was 100 %, and good sensitivity and specificity was obtained when compared to the conventional tests. As with other genotypic tests, there is interest in the application of these techniques directly to sputum samples. There are only two studies that address this issue. In the study by Hillemann et al., the GenoType MTBDR was tested directly in 42 smear-positive sputum samples obtaining a concordance of 100 % when compared to conventional drug susceptibility testing (Hillemann 2006). In another more recent study, the GenoType MTBDR was evaluated in 143 smear-positive sputum samples and it was able to correctly identify INH resistance in 48 (84.2 %) of the 57 specimens containing strains with resistance to high level of INH (0.4 µg/mL), and RIF resistance in 25 (96.2 %) of the 26 specimens containing RIF-resistant strains (Somoskovi 2006). There is currently interest in expanding these studies to TB-endemic countries to assess the usefulness of this type of assay for the rapid detection of multidrug resistance in TB (http://www.finddiagnostics.org/news/press/hain_oct06.shtml). Both solid-phase hybridization methods have proven relatively simple to perform; however, basic expertise in molecular biology and PCR techniques is required. As with other genotypic methods, the sensitivity of the test depends on the amount of DNA present in the sample, and the presence of inhibitors could also cause false-negative results (Palomino 2006). Another solid-phase reverse hybridization test for rapid detection of RIF resistance is rifoligotyping. This is an in house low-cost assay for the detection of RIF resistance-associated mutations in the rpoB gene of M. tuberculosis. The test was developed at the National Institute of Public Health and the Environment (http://www.rivm.nl/en/) in the Netherlands and initially evaluated at the Cetrángolo Hospital in Argentina (Morcillo 2002). It also involves a combination of DNA amplification and reverse-line blot hybridization. DNA of the rpoB gene of M. tuberculosis is amplified by PCR with specific primers and the PCR products are hybridized to oligonucleotides on a DNA membrane, encoding the wild type rpoB sequence, and the most frequent mutations in RIF-resistant strains. Amplified products from RIF-resistant strains will fail to hybridize to one or more of the wild type oligonucleotides, and in most cases, will hybridize to one of the mutant oligonucleotides bound to the membrane. RIF-resistant strains can be detected within a few hours with an enhanced luminescent reaction. In this evaluation, a total of 135 M. tuberculosis isolates were tested with the rifoligotyping assay and the results compared with the proportion method and the MGIT960 system. The rifoligotyping assay correctly identified 90 of the 97 RIF-resistant isolates (sensitivity 92.8 %) while all the RIF-susceptible isolates were also correctly identified. A minor modification of this assay has also been tested in a multicenter study to detect resistance to RIF, INH, SM and EMB in clinical isolates of M. tuberculosis (Mokrousov 2004). Oligonucleotides specific for wild type and mutant alleles of selected codons in the genes rpoB, inhA, ahpC, rpsL, rrs, embB, were immobilized on a nylon membrane. For validation of the test, the membranes were sent to seven laboratories in different geographical locations. The reproducibility for rpoB mutation detection was performed on a blinded set of reference DNA samples and overall concordant results were obtained. However, when further mutation analysis was performed on local strains, only 132 (85.2 %) of 155 RIF-resistant and 28 (51.0 %) of 55 EMB-resistant isolates were correctly identified. Resistance to INH was successfully identified in 16.9 % and 13.2 % of strains harboring mutations in the inhA and ahpC promoter region respectively. Likewise, mutations in rrs and rpsL conferring resistance to SM were identified in 15.1 % and 10.7 % of SM-resistant strains respectively. Nevertheless, the accuracy of this method for RIF resistance detection has recently been confirmed in another study that used a