4.3.1 Decision pathway for Algorithm 3 – DST for second-line drugs for people with RR-TB or MDR-TB

Tests

  • Recently targeted NGS tests have been recommended that can detect mutations associated with resistance to BDQ, FQ, LZD, CFZ, PZA, RIF, INH, EMB, AMK and STR.
    • The use of targeted NGS tests to detect resistance does not eliminate the need for conventional phenotypic DST, which will be necessary for determining resistance to anti-TB agents that have not been assessed or where performance of resistance detection was suboptimal by the targeted NGS test, and for monitoring the emergence of additional drug resistance.
    • Targeted NGS tests are suitable for use at the central laboratory or NTRL level. They may also be used at the regional level if the appropriate infrastructure, human resources and QA systems are available. Implementation of targeted NGS testing depends on the availability of a reliable specimen transport system and an efficient mechanism for reporting results.
  • A low complexity automated NAAT is recommended for the detection of resistance to FQs:
    • It can be used in PLHIV, children and people with extrapulmonary TB.
    • The first-in-class test, the Xpert MTB/XDR test, provides results in under 2 hours, requires minimal hands-on time, can be used at the peripheral level and provides results simultaneously for FQ, INH, ETO and AMK. This test requires a 10-colour GeneXpert instrument, unlike the current Xpert MTB/RIF and Xpert Ultra test, which use the 6-colour GeneXpert instruments. The new instruments can be linked to existing 6-colour GeneXpert systems through a common computer.
    • The use of low complexity automated NAATs to detect FQ resistance does not eliminate the need for conventional phenotypic DST, which will be necessary for determining resistance to other anti-TB agents and for monitoring the emergence of additional drug resistance.
  • The SL-LPA is recommended for the detection of resistance to FQs and should be used where available:
    • The diagnostic accuracy of SL-LPA is similar when it is performed directly on sputum or on cultured isolates. SL-LPA can be used on smear-positive or smear-negative specimens, although a higher indeterminate rate will occur with smear-negative specimens.
    • SL-LPA is only recommended for use with sputum specimens or MTBC isolates. The laboratory testing of other specimen types should rely on culture and phenotypic DST.
    • SL-LPA is suitable for use at the central or NTRL level. It may also be used at the regional level if the appropriate infrastructure, human resources and QA systems are available. Implementation of SL-LPA testing depends on the availability of a reliable specimen transport system and an efficient mechanism for reporting results.
    • The use of SL-LPA to detect FQ resistance does not eliminate the need for conventional phenotypic DST, which will be necessary for determining resistance to other anti-TB agents and monitoring the emergence of additional drug resistance.
    • Culture-based phenotypic DST is still required and recommended for drugs that are used in RR/MDR-TB regimens and for which there are either no rapid molecular DST methods or where their performance is suboptimal (e.g. BDQ, LZD, Pa, CS, CFZ and DLM).

General considerations

  • Two BDQ-containing regimens are recommended for the treatment of MDR/RR-TB (9):
    • An all-oral 6-month regimen composed of BDQ, Pa and LZD, with or without MFX (BPaL or BPaLM) in people suffering from MDR/RR-TB or MDR/RR-TB with additional resistance to FQs (pre-XDR-TB).
    • An all-oral regimen of 9 months composed of BDQ, LFX or MFX, ETO, EMB, INH (high-dose), PZA and CFZ (4–6 months of BDQ-LFX-ETO-EMB-PZA-high-dose INH-CFZ / 5 months of LFX-CFZ-PZA-EMB) for individuals not eligible for BPaL or BPaLM.
    • Individualized all-oral longer regimens, designed using the WHO priority grouping of medicines, may still be used for people with MDR/RR-TB who do not meet the eligibility criteria for the BDQ-containing regimens.
  • WHO guidelines stress the importance of DST before treatment, especially for medicines for which mWRDs are available.
    • WHO-recommended rapid molecular tests are available for the detection of mutations associated with resistance to FQs (a low complexity automated NAAT, FL-LPA, SL-LPA and targeted NGS tests) and mutations associated with BDQ resistance (targeted NGS tests). The recently recommended targeted NGS tests can also detect mutations associated with resistance to some of the other drugs included in MDR-TB regimens (e.g. LZD, CFZ, PZA, AMK and STR).
    • WHO recommends a new molecular test for PZA resistance detection belonging to the class “high complexity reverse hybridization NAAT”. Its use is limited to culture isolates. Alternatively, pncA sequencing should be performed when available. In a quality-assured laboratory, with careful attention to inoculum preparation, a susceptible phenotypic DST result using MGIT for PZA can be used to guide the inclusion of PZA in a DR-TB treatment regimen (Web Annex C).
    • Reliable phenotypic DST methods are available for RIF, INH, FQs, BDQ, CFZ, Pa, CS, LZD, AMK and DLM. Testing algorithms that rely on culture and phenotypic DST are described in the relevant WHO policy framework (53) and technical manual (Web Annex C). Member States should ensure that there is capacity for DST for drugs used for treatment and for which reliable testing is available.
    • No reliable phenotypic DST methods are available for EMB, ETO/prothionamide, or imipenem-cilastatin/meropenem; hence, results should not be used for clinical decision-making.
    • If phenotypic DST to second-line drugs is not available in-country, specimens or isolates may be shipped to an external laboratory for testing (e.g. a WHO supranational reference laboratory [SRL]). Material transfer agreements and import or export permits may be needed.
    • Currently, the availability of phenotypic DST for BDQ and LZD is limited in many settings, and resistance levels are likely to be low. There is, however, increasing evidence that BDQ resistance occurs even in unexposed people at a level of 1.4–3.4% (54). BDQ is a core drug for DR-TB treatment and is included in the revised definition of XDR-TB. Thus, building testing capacity to test this and other drugs used in treatment (e.g. LZD, Pa, CS, CFZ and DLM) is essential. If resistance is suspected during treatment and DST is not available, the strains should be referred to a TB SRL for further testing.
    • Initiation of treatment should not be delayed while waiting for the results of DST

Fig. 4.5. New Algorithm 3a: DST for MDR/RR-TB using targeted NGS

fig4-5

Decision pathway for Algorithm 3a – testing for BDQ and FQ resistance

  1. Promptly initiate the person on an MDR-TB regimen in accordance with national guidelines. The most recent WHO recommendation is to use an all-oral 6-month regimen composed of BDQ, Pa, LZD and MFX (BPaLM) (9). The use of an all-oral BDQ-containing treatment regimen of 9 months is now limited to those who are not eligible for the BPaLM or BPaL regimens (e.g. aged below 14 years, or pregnant or breastfeeding) (9).
  2. If molecular and phenotypic testing are performed in the same laboratory, collecting one specimen may be sufficient. If testing is performed in two laboratories, collect two specimens and conduct the molecular and phenotypic testing in parallel. Transport sputum specimens or isolates to the appropriate testing laboratory, if necessary.
  3. Conduct targeted NGS testing to detect mutations associated with resistance to BDQ and other medicines, and undertake culture in parallel 3 A .
  4. If the targeted NGS test result is indeterminate, the targeted NGS test can be repeated with a fresh sample in cases where the bacterial load is expected to give a definitive result (smear positive of high/medium grade on mWRD) and treatment decisions are based on clinical assessments, the epidemiologic situation and the results of phenotypic DST. Use the results of the targeted NGS test to modify treatment if appropriate, and select the drugs requiring phenotypic DST when the culture is positive 3 B . Note: although the results from sequencing produce information on multiple drugs simultaneously, Table 4.3.1 takes a single-drug approach for interpreting targeted NGS test results, for simplicity, although the information on the resistance or susceptibility of each of the medicines should be taken into consideration when designing a treatment regimen.
  • If the targeted NGS test detects one or more mutations associated with resistance to BDQ, evaluate the pretest probability of BDQ resistance before making clinical decisions because of the suboptimal performance of the targeted NGS test for detecting BDQ resistance (sensitivity: 67.9%, 95% CI: 42.6–93.2%; specificity: 97.0%, 95% CI: 94.3–99.7%):
    • If the risk of BDQ resistance is low (e.g. no prior BDQ exposure, the prevalence of resistance to BDQ is less than 5% in the population, or there is no history of contact with a person known to have BDQ-resistant TB) and the targeted NGS test does not detect mutations associated with resistance to FQs and LZD, continue the BDQ-containing regimen while awaiting the results of phenotypic DST.
    • If the risk of BDQ resistance is low and the targeted NGS test does detect other mutations associated with resistance to FQs, stop moxifloxacin and continue with BPaL. If resistance to LZD is detected, change to another regimen based on the targeted NGS test result and follow-up phenotypic DST results.
    • If the risk of BDQ resistance is high (e.g. prior BDQ exposure, the prevalence of resistance to BDQ is more than 5% in the population, or there is history of contact with a person known to have BDQ-resistant TB), change to an individualized regimen based on the targeted NGS test results and follow-up phenotypic DST results.
  • If the targeted NGS test does not detect mutations associated with resistance to BDQ, clinical decisions should consider the risk of BDQ resistance and the results of the targeted NGS test for other medicines, particularly for FQs and LZD. People should be closely monitored, and additional DST performed on any culture isolated at month 2 or later during treatment:
    • If the risk of BDQ resistance is low, continue the BDQ-containing regimen. The decision to perform phenotypic DST for BDQ for this group will be context dependent, taking into account the prevalence of resistance, the number of samples to be tested and the expected number missed by targeted NGS. Ideally, where resources are available, phenotypic DST would be done for all samples. Modify the regimen as appropriate if resistance to other medicines in the regimen is detected.
    • If the targeted NGS test detects other mutations associated with FQ resistance but not LZD, treat with BPaL (i.e. discontinue the use of FQ).
    • If the targeted NGS-based test detects other mutations associated with LZD resistance, change to a BDQ-containing individualized regimen based on the targeted NGS test results and follow-up phenotypic DST results.
    • If the risk of BDQ resistance is high, continue the BDQ-containing regimen while awaiting the results of phenotypic DST for BDQ. Modify as appropriate if resistance to other medicines in the regimen is detected (as described in Step 5). Closely monitor the person and conduct DST on those who remain culture positive at month 2 or later and those who are not responding clinically.

Table 4.3.1. Treatment modifications and follow-on DST for MDR/RR-TB based on results from targeted NGS

tab4-3-1

BDQ: bedaquiline; BPaL: bedaquiline (B), pretomanid (Pa) and linezolid (L); BPaLM: bedaquiline (B), pretomanid (Pa), linezolid (L) and moxifloxacin (M); CFZ: clofazimine; CS: cycloserine; DLM: delamanid; DST: drug susceptibility testing; EMB: ethambutol; ETO: ethionamide; FQ: fluoroquinolone; LZD: linezolid; M: moxifloxacin; NGS: next-generation sequencing; Pa: pretonamid; PZA: pyrazinamide; TB: tuberculosis.

a Where a targeted NGS result is indeterminate, phenotypic DST should be performed.

b Low risk: no prior BDQ exposure, the prevalence of resistance to BDQ is <5% in the population, or there is no history of contact with a person with TB with a known resistance to BDQ.

d In individuals not at high risk for resistance (e.g. no prior drug exposure, the prevalence of resistance to the specific drug is <5% in the population, or there is no history of contact with a person with TB with known drug resistance) phenotypic testing is advised but is not an absolute requirement. This is particularly true where access to phenotypic DST is limited, in which case, higher risk groups should be prioritized.

e Where possible, these drugs should only be reported for people requiring 9-month or individualized regimens. If there is no option to selectively report drugs, it should be clear that results relevant to the current regimen should be primarily considered and the additional results should be used only when needed.

f Phenotypic DST may not be more accurate than targeted NGS for these drugs. The targeted NGS results can be used for clinical management

Fig. 4.6. Algorithm 3b: DST for people with MDR/RR-TB (limited or no targeted NGS capacity)

tab4-6a

Decision pathway for Algorithm 3b – testing for FQ resistance

  1. Promptly initiate the person on an MDR-TB regimen in accordance with national guidelines. The most recent WHO recommendation is to use an all-oral 6-month regimen composed of BDQ, Pa, LZD and MFX (BPaLM) (9). The use of an all-oral BDQ-containing treatment regimen of 9–12 months is now limited to those not eligible for the BPaLM or BPaL regimens (e.g. aged below 14 years, or pregnant or breastfeeding) (9).
  2. If molecular and phenotypic testing are performed in the same laboratory, collecting one specimen may be sufficient. If testing is performed in two laboratories, collect two specimens and conduct the molecular and phenotypic testing in parallel. Transport sputum specimens or isolates to the appropriate testing laboratory, if necessary.
  3. Conduct the low complexity automated NAAT or SL-LPA to detect mutations associated with FQ resistance. Note that targeted NGS tests can also detect mutations associated with resistance to FQ. If a targeted NGS test is available, follow Algorithm 3a for interpretation of results and follow-up actions.
  4. If the low complexity automated NAAT or SL-LPA detects one or more mutations associated with resistance to FQs and:
  • the individual is on a BPaLM regimen, discontinue MFX and continue with BPaL treatment while awaiting the results of the phenotypic DST;
  • the individual is on the 9-month all-oral regimen, change to an individualized longer regimen, designed using the WHO priority grouping of medicines (9):
    • the first-in-class low complexity automated NAAT (Xpert MTB-XDR) provides results for INH, FQs, ETO and AMK, and can be used to inform individualized regimen selection;
    • collect a specimen and submit for phenotypic DST to the WHO Group A, B and C drugs (e.g. for BDQ, Pa and LZD), if phenotypic DST is not already being done as described in Step 6; and
    • perform DST for MFX at the clinical breakpoint to determine the potential use of highdose (800 mg) MFX for treatment (9) (Web Annex C).
  1. If the low complexity automated NAAT or SL-LPA is negative for mutations associated with resistance to FQs:
  • Continue people on the all-oral BDQ-containing MDR-TB regimen (6- or 9-month), while awaiting the results of the phenotypic DST (Step 6).
  • In settings with high underlying prevalence of resistance to FQs or for people considered at high risk of resistance, refer a specimen for culture and phenotypic DST for FQs, because the sensitivity of the low complexity automated NAAT and SL-LPA to detect mutations associated with FQ resistance is about 93% and 86%, respectively. The phenotypic DST should include testing for resistance to the FQs used in the country. The phenotypic DST should also include testing at the clinical breakpoint to inform individualized drug selection. Modify the regimen as necessary, based on the phenotypic DST results.
  1. Perform culture and phenotypic DST for each of the drugs included in the treatment regimen for which there are accurate and reproducible methods. For the preferred regimens, reliable phenotypic DST methods when performed in a quality-assured laboratory are available for BDQ, LZD, Pa, CS, FQs, CFZ, PZA and INH (Web Annex C). A WHO-recommended molecular test for PZA resistance detection is available (high complexity reverse hybridization NAAT) but is currently limited to use on culture isolates
  • If the isolate is susceptible to all drugs, continue the person on the preferred MDR-TB regimen.
  • If resistance to a drug is detected, refer to Table 4.3.1 to guide treatment modification. Given that results for phenotypic DST are slow, reassess the person’s response when these results become available. The decision to change from a shorter to the longer MDR-TB regimen should consider the phenotypic DST result and clinical response. Monthly monitoring is important, and the person should be closely followed up.
  1. For all people with TB, ensure that treatment monitoring includes the collection of samples for culturing as described in the WHO consolidated guidelines (9). Any positive culture suggestive of treatment failure should undergo phenotypic DST. Modify the regimen as necessary, based on the phenotypic DST results.
  • WHO recommends that all people with TB being treated with an MDR-TB regimen be monitored for treatment response using sputum culture and sputum-smear microscopy. It is desirable for sputum culture to be repeated at monthly intervals.
  • Although the risk of treatment failure increases with each additional month without bacteriological conversion, no discrete cut-off point has been defined that could serve as a reliable marker of a failing regimen. The choice of cut-off point will depend on the clinician’s desire to minimize the risk of failure and, in particular, to limit the risk of prolonging a failing regimen.

Book navigation