4.1.1 Decision pathway for Algorithm 1 – mWRD as the initial diagnostic test for TB

Tests

The mWRDs appropriate for this algorithm include the Xpert MTB/RIF, Xpert MTB/RIF Ultra, Truenat MTB, Truenat MTB Plus and TB-LAMP tests and moderate complexity automated NAATs.

  • “Xpert MTB test” designates either the Xpert MTB/RIF test or the Xpert MTB/RIF Ultra (hereafter referred to as “Xpert Ultra”) test. The individual tests are named when describing test-specific features. The Xpert Ultra test has an additional semiquantitative category called “trace”, caused by small amounts of bacterial DNA.
  • The Truenat MTB and MTB Plus assays use the same results categories as the Xpert MTB/ RIF assay, and the decision pathway for the Truenat tests is the same as that for the Xpert MTB/RIF test.
  • Moderate complexity automated NAATs include a number of tests for detection of MTBC as well as RIF and INH resistance. Currently, tests from four manufacturers have been reviewed and recommended, but only for use on respiratory samples:
    • Abbott RealTime MTB and Abbott RealTime MTB RIF/INH (Abbott);
    • FluoroType MTBDR and FluoroType MTB (Bruker/Hain Lifescience);
    • BD MAX MDR-TB (Becton Dickinson); and
    • cobas MTB and cobas MTB-RIF/INH (Roche).
  • Simultaneous detection of MTBC and RIF resistance applies to the Xpert MTB/RIF, Xpert Ultra and moderate complexity automated NAAT tests. The Truenat MTB or MTB Plus test requires subsequent testing with Truenat MTB-RIF Dx. Some moderate complexity automated NAATs (Abbott RealTime MTB and cobas MTB) have the option of simultaneous or manual reflex testing for resistance to RIF or INH (or both). For the two-step procedures, the same DNA sample that was isolated for the initial test is used.
  • The TB-LAMP test only detects MTBC and it requires a fresh specimen to be collected for subsequent testing with an alternative molecular test or phenotypic DST to detect RIF resistance.

General considerations

WHO recommends the use of an mWRD (Xpert MTB/RIF, Xpert MTB/RIF Ultra, Truenat MTB, Truenat MTB Plus, Truenat MTB-RIF Dx, TB-LAMP or moderate complexity automated NAAT) as the initial diagnostic test, rather than microscopy or culture, for all individuals with signs and symptoms of TB. This includes all newly presenting symptomatic individuals; it may also include people who are on treatment or have been previously treated, if the person is being evaluated for possible RR-TB or Hr-TB (e.g. non-converters at the end of the intensive phase of treatment despite treatment adherence) or for a new or continuing episode of TB (e.g. relapse cases or people treated previously, including those who had been lost to follow-up). TB programmes should transition to replacing microscopy as the initial diagnostic test with mWRDs that show a higher sensitivity for the diagnosis of TB as well as simultaneous detection of resistance to RIF (and, for moderate complexity automated NAATs, INH as well)

This algorithm is designed to be used with any of the mWRDs for the detection of MTBC, although the algorithm may require minor modification based on which mWRD is used and in which population. The description and WHO recommendations for each test are presented in Section 2. Only special considerations as they relate to the algorithms are provided here.

  • The Xpert MTB/RIF and Xpert Ultra tests have specific recommendations for extrapulmonary TB and children being evaluated.
    • These tests are recommended for use with CSF (preferred sample for TB meningitis), lymph node aspirates and lymph node biopsies. In addition, the Xpert MTB/RIF test is recommended for pleural fluid, peritoneal fluid, pericardial fluid, synovial fluid or urine, as an initial diagnostic test for the corresponding extrapulmonary TB disease. Blood may also be used as a specimen for HIV-positive adults and children with signs and symptoms of disseminated TB. Sample-specific optimization steps may be needed to achieve optimal results, especially for paucibacillary samples.
    • Minor variations for use of these tests with children with signs and symptoms of pulmonary TB include the following:
    • Xpert MTB/RIF can be used as the initial diagnostic test for pulmonary TB with sputum, gastric aspirate, nasopharyngeal aspirate or stool samples, whereas Xpert Ultra is recommended for use with sputum and nasopharyngeal aspirate specimens; and
    • in settings with a pretest probability of 5% or more and an Xpert MTB/RIF or Ultra negative result on the initial test, repeated testing with Xpert MTB/RIF or Ultra with the same or different specimen types (for a total of two tests) may be used; otherwise, repeat testing is not recommended.
  • The Truenat MTB and MTB Plus test considerations include the following:
    • There is uncertainty about use of this test in PLHIV, because insufficient data were available on the performance of these tests in PLHIV. The indirect data on test performance in people who were smear-negative were used to extrapolate the recommendation to use in PLHIV.
    • In children, sufficient data were available to recommend the use of these tests with sputum samples only. There were no data on how these tests performed with other specimens.
    • The performance of these test for the detection of extrapulmonary TB is unknown.
  • The TB-LAMP test is recommended as a replacement test for sputum-smear microscopy; it would be suitable for use in settings that have a low prevalence of HIV and MDR-TB. Considerations for the use of this test include the following:
    • In populations with a high burden of MDR-TB, TB-LAMP should not replace the use of rapid molecular tests that detect RIF resistance (e.g. Xpert MTB/RIF), because TB-LAMP does not provide any information on RIF resistance.
    • In populations with a high prevalence of HIV, TB-LAMP should not replace the use of rapid molecular tests that have a higher sensitivity for detection of TB (e.g. Xpert Ultra).
    • Considerations for the moderate complexity automated NAATs include the following:
    • The recommendations apply to PLHIV. Performance on smear-negative samples was reviewed but was only available for TB detection, not for RIF and INH resistance. Data stratified by HIV status were not available.
    • The recommendations apply to adolescents and children based on the generalization of data from adults. An increased rate of false negative or indeterminate results may be found with paucibacillary TB disease in children.
    • Extrapolation to use in people with extrapulmonary TB and testing on non-sputum samples was not considered because data on diagnostic accuracy of technologies in the class for non-sputum samples were limited.
  • The mWRDs are not recommended as tests for monitoring treatment, because the presence of dead bacilli may generate a positive result. Instead, microscopy and culture should be used for monitoring, in accordance with national guidelines and WHO recommendations.

Algorithm 1 describes the collection of one initial specimen to be used for mWRD testing and the collection of additional specimens as needed. For operational issues, programmes may consider collecting two specimens (e.g. spot and morning sputum samples, or two spot specimens) from each person routinely, instead of only collecting a second specimen when additional testing is needed. If two specimens are collected, the first should be tested promptly using the mWRD. The second specimen may be used for the additional testing described in the algorithm (e.g. repeat mWRD testing for failed tests or follow-on resistance testing, or for smear microscopy or culture as a baseline for treatment monitoring).

  • If only one specimen can be collected (e.g. if tissue biopsy samples are difficult or impossible to obtain repeatedly), the TB diagnostic algorithm should be modified to prioritize testing with the mWRD. If additional TB testing is warranted, one option is to consider using any portion of the sample remaining after the mWRD for other tests (e.g. culture, histology, LPA and DST). Alternatively, the sample could be processed for culture and the same sediment could be used for the mWRD, culture and other tests. Clinical decisions should be made based on clinical judgement and the results of available laboratory tests.

With respect to the detection of MTBC, mWRD results are typically reported as “MTB not detected”, “MTBC detected”, “no result”, “error” or “invalid”. Within the “MTBC detected” result group, some mWRDs provide semiquantitative results (high, medium, low or very low). The Xpert Ultra test has an additional semiquantitative category called “trace”.

  • Each of the semiquantitative categories of MTBC detected, including “trace”, is considered as bacteriological confirmation of TB.
  • Xpert Ultra “trace”: for PLHIV and children who are being evaluated for pulmonary TB, and for individuals being evaluated for extrapulmonary TB, the “MTBC detected trace” result is considered as bacteriological confirmation of TB.
  • In HIV-negative, symptomatic adults with a recent history of TB treatment (i.e. completed <5 years ago), Xpert Ultra “trace” results (and occasionally other molecular mWRD “MTBC detected very low”) may be positive not because of active TB but because of the presence of non-viable bacilli. Clinical decisions must be made on all available information and clinical judgement.

With respect to the detection of RIF and INH resistance, the mWRDs report the results as RIF or INH “resistance detected”, “not detected” or “indeterminate”. For the assays for which resistance detection relies on the absence of binding of wild-type reporter probes to amplicons (e.g. Xpert MTB/RIF and Truenat MTB-RIF Dx) it may be more appropriate to state that resistance is inferred rather than detected.

The use of an mWRD to detect resistance to RIF or INH (or both) does not eliminate the need for conventional culture-based phenotypic DST, which will be necessary to determine resistance to other anti-TB agents and to monitor the emergence of additional drug resistance.

Decision pathway

  1. Collect a good-quality specimen and transport it to the testing laboratory. Conduct the mWRD. For individuals being evaluated for pulmonary TB, the following specimens may be used: induced or expectorated sputum (preferred), bronchoalveolar lavage, gastric lavage or aspirates, nasopharyngeal aspirates and stool samples. (For information on which specimens may be used with which mWRD, see Section 2.2 above or individual WHO policy statements.)
  2. If the mWRD result is “MTB not detected” 1 A, re-evaluate the person and conduct additional testing in accordance with national guidelines
    • Further investigations for TB may include chest X-ray, additional clinical assessments, additional mWRD testing or culture and clinical response following treatment with broad-spectrum antimicrobial agents (FQs should not be used).
    • In children with signs and symptoms of pulmonary TB in settings with a high pretest probability (>5%) and a negative Xpert MTB result on the first initial test, repeat the Xpert MTB test for a total of two tests. The tests may use the same specimen types or different specimen types (e.g. one sputum specimen and one nasopharyngeal aspirate sample).
    • The performance of the other mWRDs in repeat testing is not known.
    • Consider the possibility of clinically defined TB (i.e. TB without bacteriological confirmation). Use clinical judgement for treatment decisions.

unfig-1

  1. If the mWRD result is “MTBC detected, RIF resistance not detected” and “INH resistance not detected” or INH results are unknown:1 B
    • Initiate the person on an appropriate regimen using first-line TB drugs in accordance with national guidelines.
    • Request additional DST in the following cases:
      • Molecular or phenotypic DST for INH is indicated particularly (see Algorithm 4 for follow-up testing):
        • if the person has been treated with INH or is a contact of a person known to have Hr-TB; or
        • if there is high prevalence of INH resistance that is not associated with RIF resistance (i.e. Hr-TB or poly-resistance, not MDR-TB) in this setting.

          If the INH resistance is “not detected” by the moderate complexity automated NAAT and the person has a high risk of Hr-TB, phenotypic DST for INH should be performed because 6–14% of INH resistance can be missed by current molecular tests.

      • Molecular or phenotypic DST for RIF resistance may be requested if the person is at risk of having RR-TB despite the initial mWRD result showing “susceptibility”. Sometimes, these anomalous results may be due to sample labelling errors and a repeat test may resolve the issue. False RIF-susceptible Xpert MTB results can occur but are uncommon (1–5% of RIF-resistant TB cases tested) and the level of such results depends on the epidemiologic settings. In contrast, phenotypic DST for RIF, especially using liquid culture, is associated with a higher proportion of false-susceptible results (41). The updated critical concentration for RIF which should be used will reduce, but not eliminate, this issue. Sequencing should be performed when available, and should cover not only the RIF-resistance-determining region (RRDR) but regions outside as well (e.g. codons 170 and 491).

    • If additional molecular or phenotypic testing is performed:

      • The molecular and phenotypic testing may be performed in different laboratories. Perform these tests in parallel – do not wait for the results of one test before initiating another test.

      • The molecular and phenotypic DST may be performed using the specimen (direct DST) or using bacteria recovered by culture (indirect DST). Direct DST is preferred for molecular testing, whereas indirect DST may be preferred for phenotypic DST, because of technical issues related to producing an appropriate inoculum and loss to contamination.

      • A rapid molecular test is preferred using WHO-recommended tests. Mutation interpretation can also be found in the WHO catalogue of mutations (4). DNA sequencing has proven useful in many cases but WHO has not yet evaluated it for clinical use.

      • Culture-based phenotypic DST for INH and RIF requires 3–8 weeks to produce a result. Phenotypic DST may be useful for evaluating people with a susceptible molecular result, particularly in populations with a high pretest probability for resistance to INH.

unfig-2

  1. If the mWRD test result is “MTBC detected, RIF resistance not detected and INH resistance detected” (currently only applicable to the moderate complexity automated NAATs):
    • Initiate the person on an appropriate Hr-TB regimen in accordance with national guidelines. The WHO recommendation for people with Hr-TB is treatment with RIF, ethambutol (EMB), PZA and levofloxacin (LFX) for a duration of 6 months (9).
    • Follow Algorithm 4 for Hr-TB:
      • Additional DST for RIF may be required in settings where RIF-resistant mutations outside the RRDR are common. The decision on the choice between phenotypic testing or sequencing will depend on the type of mutation expected. In places where the rpoB I491F mutation is common, sequencing is preferred because phenotypic DST, even with the lower CC, will still miss many resistant infections; in other cases (e.g. V170F) phenotypic testing is appropriate (27).
  2. If the mWRD result is “MTBC detected, RIF resistance detected”, 1C an MDR-TB risk assessment is needed, irrespective of the INH result. People at high risk for MDR-TB include people who have been previously treated (e.g. those who had been lost to follow-up, relapsed or experienced a failed treatment regimen); non-converters (e.g. people who were smear-positive at that end of the intensive phase of treatment for drug susceptible TB); contacts of people with MDR-TB; and any other groups at risk for MDR-TB identified in the country. In high MDR-TB burden countries, every person with TB is considered to be at high risk of having MDR-TB.
  • If the person is at high risk of having MDR-TB and the test result shows RIF resistance, initiate them on a regimen for MDR/RR-TB in accordance with national guidelines. Follow Algorithm 3 for additional testing.
  • If the person is not at high risk of having MDR-TB, repeat using an mWRD with a second sample. To aid interpretation, the initial instrument output for the result can be reviewed when available. Probe binding delay and samples having a low bacillary load, which can be inferred from the semiquantitative readings (e.g. low and very low), have been associated with increased false resistance in some settings (42–44).
    • If the second test also indicates RIF resistance, initiate an MDR/RR-TB regimen in accordance with national guidelines and WHO recommendations, and follow Algorithm 3 for additional testing.
    • If the mWRD result for the second sample is “MTBC detected, RIF resistance not detected”, initiate treatment with a first-line regimen in accordance with national guidelines. In most situations, false positive RIF-resistant results due to technical performance of the assay are rare; however, such results may occur because of laboratory or clerical errors. It is assumed that the repeat test will be performed with more caution, that the result of the second test will be correct, and that the result of the first test may have been due to a laboratory or clerical error. Mixed infections in high-burden settings could also explain such discordance; therefore, people should be closely followed up and repeat tested if the response to first-line treatment is poor. If an INH resistance or susceptibility result is available, interpret the result and follow-up as described in Algorithm 4.
    • In the event that the mWRD result for the second sample is “MTBC detected, RIF resistance is inconclusive”, the person will require further investigation. A possible mixed infection may explain such a scenario. History of prior treatment and TB contact history should be reassessed. The decision to manage a person as having Hr-TB or MDR/ RR-TB will need to be based on further investigation that includes phenotypic DST to RIF and INH and, where available, DNA sequencing. A third mWRD should be performed to decide on the initial therapy; the person should be closely followed up while awaiting the final definitive results and the appropriate algorithm should be followed.
    • In the event that a moderate complexity automated NAAT was performed and INH results are also available, this could be useful to provide certainty. INH resistance is associated with RIF resistance, and the finding of INH resistance should prompt further investigation to exclude RIF resistance.
  • For all people with RR-TB or MDR-TB, conduct additional investigations to assess resistance to the drugs being used in the treatment regimen. Rapid detection of FQ resistance is essential in determining the regimen to be selected. The recent addition of a low complexity automated NAAT for the detection of FQ resistance provides a rapid and accurate peripheral level solution that can be performed directly on specimens. Phenotypic (culture and DST) and molecular (e.g. SL-LPA and DNA sequencing) methods are available to evaluate drug resistance beyond RIF and INH. Rapid molecular methods are preferred. However, for resistance detection to some of the new and repurposed drugs, phenotypic DST is the only currently available option; thus, two separate specimens may be required.
    • MDR/RR-TB regimens rely on the use of either FQs (submit a sample for molecular testing for FQ resistance – see Algorithm 3a) or BDQ (submit a sample for molecular testing for BDQ resistance – see Algorithm 3b).
    • Ideally, a specimen from each person should be submitted for DST for each of the drugs used in the regimen for which there is a reliable testing method. However, do not delay treatment initiation while waiting for DST results (e.g. phenotypic DST can take weeks or even months to provide results).
    • Any positive culture recovered during treatment monitoring that is suggestive of treatment failure should undergo DST for the drugs used in the treatment regimen.

unfig-9

  1. If the mWRD gives a result of “MTBC detected, RIF indeterminate” 1 D , the person will require further investigation. The interpretation and follow-up testing for Xpert Ultra differs from that for other mWRDs. With any of the mWRDs, the initial result of “MTBC detected” should be considered as bacteriological confirmation of TB. The person should be initiated on an appropriate regimen using first-line TB drugs in accordance with national guidelines, unless the person is at high risk of having MDR-TB (in which case, they should be initiated on an MDR-TB regimen). In most settings, for the purpose of making treatment decisions, a history of prior TB treatment is not sufficient to indicate that the person is at high risk of having MDR-TB.
  • For most mWRDs, an “MTBC detected, RIF resistance indeterminate” result is generally caused by a paucibacillary TB load in the sample; in such cases, retesting a fresh specimen is recommended.
    • If the result of the second mWRD is “MTBC detected, RIF resistance not detected”, follow Step 3. If the result is “MTBC detected, RIF resistance detected”, follow Step 5.
    • In some cases, testing a second sample, which might also contain very few bacteria, may generate a result of “MTBC detected, RIF indeterminate” or “MTB not detected”. In these situations, additional investigations such as culture and phenotypic DST or molecular testing of the isolate or sequencing may be needed to confirm or exclude resistance to RIF, because the indeterminate result provides no information on resistance. “MTBC detected (non-trace), RIF indeterminate” results obtained with the Xpert Ultra test (especially those with high or medium semiquantitative results) may be due to the presence of large deletions or multiple mutations in the RRDR or mutations that pose a challenge with mutation analysis software (45).
    • The Ultra melt curves from “MTBC detected (non-trace), RIF indeterminate” samples should be reviewed (preferably by an advanced Xpert user or supervisor), including a review of the amplification of the probes and melt curve profile (45). 
      • Melt curves that suggest the presence of a large deletion or multiple mutations in the RRDR should be interpreted as “RIF resistance detected”. In such cases, follow Steps 6.1 and 6.2.
      • If the melt curve is not consistent with the presence of a large deletion or multiple mutations in the RRDR, the result is interpreted as “indeterminate”. In such cases, follow Step 7 for additional testing.
      • If the semiquantitative result is high or medium, FL-LPA or DNA sequencing may be useful. 
    • Culture and phenotypic DST, FL-LPA or DNA sequencing may be performed for follow-up testing, to confirm or exclude RIF resistance.

unfig-4

  1. If the Xpert Ultra test result is “MTBC detected trace” 1 E , additional considerations are needed. However, WHO suggests not repeating Xpert Ultra testing in adults who have an initial Xpert Ultra trace result to confirm the result.
  • Review the clinical characteristics to determine the person’s age, HIV-infection status and history of TB treatment, and determine whether the samples are pulmonary or extrapulmonary.
    • PLHIV include individuals who are HIV-positive or whose HIV status is unknown but who present with strong clinical evidence of HIV infection, reside in settings where there is a high prevalence of HIV or are members of a group at risk for HIV. For all those with unknown HIV status, perform HIV testing in accordance with national guidelines.
    • Children are defined as those aged under 15 years.
    • Individuals with a history of recent TB treatment include those who successfully completed a course of therapy within the past 5 years. The likelihood of a false positive mWRD test result is highest immediately after completing treatment, and slowly declines with time (46, 47). Those who initiated but did not complete therapy and those who failed therapy should be considered as being at high risk of having MDR-TB; such people require careful clinical evaluation.
    • Xpert Ultra is recommended for use with CSF, lymph nodes and tissue specimens. Data are limited for the test’s performance with other extrapulmonary samples.
    • Health workers must endeavour to obtain a reliable history of TB treatment, recognizing that some people may not communicate their treatment history because of stigma or, in the case of migrants, concern over legal status.

unfig-5

  • For certain populations – PLHIV and children who are being evaluated for pulmonary TB; for individuals being evaluated for extrapulmonary TB using CSF, lymph nodes and tissue specimens; and for adults being evaluated for pulmonary TB, who are not at risk for HIV and who do not have a history of prior TB treatment within the past 5 years:
    • Consider the MTBC detected trace result obtained with the first specimen as bacteriological confirmation of TB (i.e. a true positive result) and use for clinical decisions.
    • Initiate the person on an appropriate regimen using first-line TB drugs, in accordance with national guidelines, unless the person is at high risk of having MDR-TB (in which case, initiate the person on an MDR-TB regimen).
    • Undertake additional investigations (e.g. culture and DST) to confirm or exclude resistance to RIF.
  • For adults being evaluated for pulmonary TB, who are not at risk of HIV and have a history of TB treatment in the past 5 years:For adults with a history of recent TB treatment or unknown treatment history, consider the possibility of the Xpert Ultra trace result being a false positive result because of the presence of non-viable bacilli.
    • Clinically re-evaluate the person and conduct additional testing (including liquid culture) in accordance with national guidelines. Consider the possibility of TB caused by reactivation, relapse or reinfection.
    • In initiating treatment, consider the clinical presentation and context of the person. Make clinical decisions based on all available information and clinical judgement.
    • Further investigations for TB may include chest X-ray, additional clinical assessments and clinical response following treatment with broad-spectrum antimicrobial agents (FQs should not be used).
      1. Repeat Xpert Ultra testing is of uncertain benefit. A recent WHO GDG recommended against repeat Xpert Ultra testing for individuals with an initial Xpert Ultra trace result for the detection of MTBC.
      2. Culture and phenotypic DST may be of benefit to detect TB and drug resistance. The trace result provides no information on RIF resistance.
  1. If the mWRD does not give a result, 1 F or gives a result of “error” or “invalid”, repeat the mWRD at the same testing site with a second specimen. Return to algorithm 1

Interpretation of discordant results

This algorithm relies on testing of a sample with an mWRD to detect MTBC and assess susceptibility to RIF. Discordance in resistance to INH is described in Algorithm 4. On occasion, follow-up testing is recommended to ensure that clinical decisions are well informed. However, discordant results may occur, usually when comparing culture-based results with molecular results. Each discordant result will need to be investigated on a case-by-case basis. General considerations are outlined below.

  1. mWRD result “MTBC detected other than trace”, culture negative (see Point 5 for trace):
    • The mWRD result and clinical judgement should be used to guide the treatment decision, pending additional testing.
    • The mWRD result should be considered as bacteriological confirmation of TB, if the sample was collected from a person who was not recently receiving treatment with anti-TB drugs. Cultures from individuals with pulmonary TB may be negative for several reasons, including that the person is being treated for TB (effective treatment rapidly renders MTBC non-viable), transport or processing problems have inactivated the tubercle bacilli, cultures have been lost to contamination, the testing volume was inadequate, or a laboratory or clerical error occurred. 
    • Follow-up actions may include re-evaluating the person for TB, reassessing the possibility of prior or current treatment with anti-TB drugs (including FQ use), evaluating response to therapy, and evaluating the possibility of laboratory or clerical error.
  2. mWRD result “MTB not detected”, culture positive:
    • Treatment decision should be based on the culture result. If the person started treatment based on clinical judgement, continue treatment. Record the person as having bacteriologically confirmed TB.
    • The culture-positive result should be considered as bacteriological confirmation of TB because culture is the current gold standard for the laboratory confirmation of TB. Using a sputum specimen, WRDs have a pooled sensitivity of 83–90% for detecting pulmonary TB compared with culture (48). Their sensitivity is lower in PLHIV, children and other specimen types such as CSF.
    • False positive cultures can result from a variety of causes, such as cross-contamination in the laboratory (e.g. from inappropriate specimen processing) or sample labelling problems. In well-functioning laboratories, such errors are rare.
    • Follow-up actions may include re-evaluating the person for TB, conducting additional testing using mWRDs, culturing additional samples, and evaluating the possibility of laboratory or clerical error. If the person was initiated on anti-TB therapy based on clinical judgement, evaluate the response to therapy.
  3. mWRD result “MTBC detected, RIF resistance detected”; RIF susceptible by phenotypic DST:
    • Use the mWRD result to guide treatment decisions pending additional testing.
    • Borderline resistant mutations are known to generate this discordant result, particularly in the BACTEC MGIT system (i.e. a false susceptible phenotypic result). People infected with strains carrying these mutations often fail treatment with RIF-based first-line regimens (41).
    • In some settings with a low prevalence of MDR-TB, silent mutations have been observed that generate a false-resistant mWRD result, but these are rare.
    • A review of the probe melting temperatures when available (49) or banding pattern on the FL-LPA can aid in determining of inferring the specific mutation (e.g. borderline resistant or silent).
    • False RIF-resistant results have been observed with the Xpert MTB/RIF G4 cartridge when the MTBC detected result was “very low” and associated with probe binding delay (42). Follow-up action may include mWRD testing of the culture.
    • Follow-up actions may include DNA sequencing, confirmatory testing on other mWRD testing platform, phenotypic DST using solid media and evaluation of the possibility of laboratory or clerical error.
  4. mWRD result “MTBC detected, RIF resistance not detected”; RIF resistant by phenotypic DST:
    • The treatment regimen should be modified based on the results of the phenotypic DST.
    • False RIF-susceptible mWRD results are rare but have been observed in 1–5% of RIFresistant TB cases tested with the Xpert MTB/RIF test in various epidemiologic settings. Mutations in the region of the rpoB gene sampled by the Xpert MTB tests have been shown to account for 95–99% of RIF resistance. The remainder of RIF resistance arises from mutations outside the sampled region, which produce an Xpert MTB result of “RIF resistance not detected”. In settings with a prevalent clone that harbours a mutation outside the RRDR, for example Eswatini (50), this may be more common; however, this has not been identified as a major concern in other settings (51). Surveillance to monitor emergence of such clones over time should be considered.
    • Follow-up actions may include DNA sequencing, repeating the phenotypic DST and evaluating the possibility of laboratory or clerical error.
  5. Xpert Ultra “MTBC detected trace”, culture negative:
    The interpretation of this result must consider the person’s characteristics, the specimen type and whether the person had been previously treated for TB:
    • Cultures may be negative for several reasons, including the person being treated for TB or treated with FQs, transport or processing problems that inactivated the tubercle bacilli, culture contamination or inadequate testing volume, or laboratory or clerical error.
    • The small numbers of bacilli in a sample that generates an “MTBC detected trace” result may be due to active TB disease, laboratory cross-contamination, recent exposure to (or infection with) tubercle bacilli (incipient TB), and current or past treatment for TB.
    • The FIND multicentre study revealed that many of the samples that generated results of “MTBC detected trace” and culture negative were from individuals who had completed therapy within the past 4–5 years, presumably because of the presence of small numbers of non-viable or non-replicating bacilli. Thus, “MTBC detected trace” results must be interpreted within the context of prior treatment.
  • For PLHIV and children who are being evaluated for pulmonary TB, or when extrapulmonary specimens (CSF, lymph nodes and tissue specimens) are tested, the benefits of the increased sensitivity for the detection of MTBC (i.e. true positives) outweighs the potential harm of decreased specificity (i.e. false positives).
    • The “MTBC detected trace” result is considered as bacteriological confirmation of TB (i.e. true positive results) and such people should have been initiated on therapy based on the Xpert Ultra result. Consider the possibility that the culture result was a false negative result.
    • Follow-up actions may include assessing the response to therapy (culture results are often not available for weeks after specimen collection), reassessing the possibility of prior or current treatment with anti-TB drugs (including FQ use), and evaluating the possibility of laboratory or clerical error.
  • For adults being evaluated for pulmonary TB who are not at risk of HIV, the balance of benefit and potential harm varies, based on whether the person had been treated previously for TB because of decreased specificity (i.e. false positives).
    • For individuals in whom a history of current or prior TB treatment can be reliablyexcluded:
      • Although the “MTBC detected trace” results should be considered as bacteriological confirmation of TB (i.e. true positive results), any clinical decision (e.g. to treat for TB) should be made based on all available laboratory, clinical and radiological information, and clinical judgement.
      • Consider the possibility that the culture result was a false negative result, if the samples were collected from a person who was not receiving treatment with anti-TB drugs, because of the paucibacillary nature of the sample. Follow-up actions for people placed on anti-TB therapy may include re-evaluating the person for TB, assessing the response to therapy, reassessing the possibility of prior or current treatment with anti-TB drugs (including FQ use), repeating Xpert Ultra testing, evaluating the possibility of laboratory or clerical error, and repeating culture (preferably using liquid culture).
    • For adults with a history of recent TB treatment:
      • Consider the possibility that the Xpert Ultra “MTBC detected trace” result was a false positive result because of the presence of non-viable bacilli. A culture-negative result is consistent with this possibility.
      • If these people had been initiated on anti-TB therapy based on clinical judgement, follow-up actions may include assessing the response to therapy, conducting additional testing in accordance with national guidelines, repeating culture (preferably using liquid culture), and evaluating the possibility of laboratory or clerical error

Book navigation