Abstract
Interpreting genotypic tests can help clinicians to take right and timely therapeutic decisions http://ow.ly/jGHH30hcB6U
Introduction
Treating multidrug-resistant (MDR-) and extensively drug-resistant (XDR-) tuberculosis (TB) is a difficult task for any clinician: there are few therapeutic options, the treatment is very long (up to 2 years), complicated by frequent adverse events, and expensive [1, 2].-Recently, the debate in the scientific community has been focused on the role and contribution of the new (bedaquiline and delamanid) [3–7] and re-purposed anti-TB drugs (in particular, linezolid, clofazimine and carbapenems) [8–11]. Furthermore, the World Health Organization (WHO), in its 2016 revised MDR-TB guidelines, has recommended a 9–12 month “shorter regimen” for patients not previously exposed and with Mycobacterium tuberculosis (MTB) strains susceptible (or likely to be) to the drugs composing the regimen (with the exception of isoniazid) [12]. The regimen is composed of an initial 4–6-month phase of kanamycin, moxifloxacin, prothionamide, clofazimine, pyrazinamide, high dose isoniazid, and ethambutol followed by 5 months of moxifloxacin, clofazimine, pyrazinamide and ethambutol; it benefits from the use of isoniazid at high dose.
The WHO, in order to ensure that eligible patients have access to the shorter regimen recommends that (WHO-endorsed) rapid genetic tests are performed [12]. The clinician, ideally, has the results of the rapid tests in a few days and can, therefore, decide to initiate the shorter regimen, or, if the patient is not eligible, a longer regimen either standardised (according to each country's national recommendations) or individualised based on a number of drugs likely to be effective. After 4–8 weeks the drug susceptibility test (DST) for second-line drugs is made available and the clinician has the possibility to modify the regimen, and, if necessary, to shift from the shorter to the longer regimen.
Evidence on sensitivity and specificity of rapid drug susceptibility tests is available in the literature, but their interpretation in order to make clinical decisions is often difficult. Evidence on confidence to diagnose resistance based on both phenotypic and genotypic tests was scanty before the publication of the article by Miotto et al. [13] in this issue of the European Respiratory Journal. This study is pivotal in making the point on what is known and on directing future research. However, it is not yet of immediate application for clinicians managing MDR/XDR-TB cases.
We aim to provide simple criteria to help clinicians interpreting the available genetic tests before taking therapeutic decisions on their patients. They are based on the new mutation grading system presented by Miotto et al. [13] in this journal and on a review of the recent literature.
What is the information provided by WHO-recommended tests?
Xpert MTB/RIF and the next-generation assay Xpert MTB/RIF Ultra (Cepheid, Sunnyvale, CA, USA) are fully automated nucleic acid amplification assays that detect MTB and mutations affecting the rifampicin resistance determining region (RRDR, codons 426–452) of the rpoB gene directly from clinical specimens. Real-time polymerase chain reaction (PCR) and melting temperature-based analysis are used by the two assays, respectively, to target the RRDR wild-type sequence (no mutant probes targeted).
Line probe assays (LPAs) are based on the PCR amplification of specific fragments of the MTB genome, followed by reverse hybridisation of the PCR products to oligonucleotide probes immobilised on nitrocellulose strips. Resistance is detected by lack of binding to wild type probes and also by binding to probes targeting specific mutations. Commercial LPAs include: GenoType MTBDRplus V2 (Hain Lifescience, Nehren, Germany) and Nipro NTM + MDRTB detection kit 2 (Nipro Corporation, Tokyo, Japan) for MTB rifampicin and isoniazid resistance determination (rpoB RRDR, katG region Ser315, inhA promoter); GenoType MTBDRsl V1 and V2 (Hain Lifescience) for identification of MTB mutations associated with fluoroquinolone (gyrA quinolone resistance determining region, QRDR, plus gyrB QRDR in version two) and second-line injectable drug (rrs region 1400, plus eis promoter in version two excluding the ethambutol resistance-conferring target embB, included in the previous version) resistance [14].
In 2016, the WHO endorsed the MTBDRsl V2, that includes the additional targets eis and gyrB, as initial test to detect resistance to fluoroquinolones and second-line injectables instead of phenotypic DST. This test's results would suffice to inform the critical decision to initiate (or not) the new shorter regimen versus the standard one [13]. The clinical interpretation of the test is summarised in table 1.
Why is this study important?
Miotto et al. [13] developed an expert, consensus-driven, standardised approach for the interpretation of mutations in MTB associated with drug resistance, based on a systematic review of genotypic and phenotypic drug-resistance literature data and on a statistical procedure to grade mutations. Liquid- and solid-media conventional DST and their combination, and Wayne enzymatic assay for pyrazinamide, were considered the reference standards for the analysis [13].
The authors identified 286 mutations classified as determining high, moderate, minimal, indeterminate, or “no-association” corrected confidence for predicting resistance to rifampicin (rpoB), isoniazid (katG, inhA-mabA, furA, mshA), ethionamide/prothionamide (inhA-mabA, ethA, mshA), ofloxacin/moxifloxacin/levofloxacin (gyrA, gyrB), pyrazinamide (pncA), streptomycin (rpsL, rrs, gidB, tap, whiB7), amikacin (rrs), capreomycin (rrs, tlyA) and kanamycin (rrs, eis, whiB7). The detection of these mutations can be used as rule-in criteria for predicting phenotypic drug resistance at current critical concentrations and, therefore, for guiding treatment decisions, developing molecular diagnostic DST assays, and interpreting the existing ones. This is the first study providing a standardised and statistical evidence-based approach to characterising drug resistance-associated MTB genetic mutations.
How to make clinical decisions based on the results of rapid tests and DST
Considering these graded mutations will enable the clinician to rule-in resistance with indisputable confidence (100% specificity), covering most of the resistant cases for at least the core drugs. Importantly, the use of this system will allow to overcome the well-known limitations of the conventional phenotypic testing for some drugs hampering an accurate and prompt choice of eligible regimens. TB laboratory staff and clinicians would simply investigate the presence of such mutations (by molecular approaches, e.g. LPAs) to include/exclude drugs for treatment (figure 1), considering the diagnostic performance reported in table 2.
Implications to use drugs at high dose in case of drug resistance
Isoniazid (INH)
If inhA mutations only are detected, even normal doses (e.g. 5 mg per day per kg body weight) of INH could be used; high doses (10 mg·kg−1 or more) are likely to be effective [15, 16]. If katG mutations only are detected, use of high doses is an option. Most katG mutations (other than 315 codon) confer moderate resistance (minimum inhibitory concentration (MIC) 1–5 µg·mL−1) that might be treated with higher doses of the drug; even the most common S315T variant leads to a variable range of resistance [17]. In the absence of additional mutations affecting the inhA gene (and ethA gene, so far uniquely detectable by sequencing approaches), ethionamide can be considered an option for the intensive phase of the shorter regimen. If inhA + katG mutations are concurrently detected, INH drug use should be avoided, since these patterns are linked to high resistance levels.
It would be crucial to determine whether the concurrent presence of inhA + katG mutations is associated with specific MTB genotypes as this might imply a limited use of shorter regimens in settings where these lineages are more represented.
Rifampicin–rifabutin (RMP-RFB)
If low level resistant rpoB mutations (MIC of 1–2 μg·mL−1) are identified, the clinician could consider either the use of high doses RMP (12–20 mg·kg−1) or a switch to RFB (for which susceptibility is common). Considering genotypes, D435Y, D435V, S441L, H445L plus other nucleotide variants, insertions and deletions affecting codons 435 and 445 are associated with low level RMP resistance and RFB susceptibility. Such mutations are included in the grading system developed by Miotto et al. [13] that, therefore, might be informative also for the use of RFB, although more data are needed. Nevertheless, critical concentrations used to determine RFB resistance are under evaluation, potentially leading to a redefinition of phenotypically resistant cases [18–21].
Moxifloxacin (MFX)
If low level resistant gyrA mutations (MIC ≤1 μg·mL−1) are identified, use of higher doses of MFX might be beneficial. Common gyrA mutations, such as A90 V, S91P, D94A and D94Y, are linked to low MIC levels and could enable the use of the drug at higher doses (800 mg·day−1) [22–25].
Implications for the use of the “shorter regimen”
The genetic tests can guide the clinician to identify the eligible patients for the shorter regimen or to decide for a traditional individualised regimen. This choice is made possible by our increasing knowledge of the genetic mechanisms that are the basis for MTB drug resistance and the study by Miotto et al. [13] represents a cornerstone.
High dosages of core anti-TB drugs, such as moxifloxacin, demonstrated their therapeutic efficiency [26]. On the other hand, the recent results from the STREAM stage 1 controlled trial has shown that adverse effects are possible; therefore, safety of higher doses needs to be assessed. [27]. While we promote the use of genetic tests to guide critical treatment decisions, we need to know more on the phenotypic meaning of genetic mutations. MIC testing is required to link unequivocally these genetic mutations with phenotypic resistance levels examining different genetic backgrounds and large strain collections.
Given the wide information on drug resistance potentially provided by genetic investigation, any commercial molecular test based on sequencing of large genomic regions (or even better whole genome sequencing) would be decisive for guiding treatment of MDR- and XDR-TB cases.
Acknowledgements
The authors wish to thank Lia D'Ambrosio and Rosella Centis for the editorial support in finalising the manuscript.
Footnotes
Published online Dec 28, 2017; republished Jan 03, 2018 with amendments to the layout of table 1.
Conflict of interest: None declared.
- Received November 6, 2017.
- Accepted December 11, 2017.
- Copyright ©ERS 2017