Over 480 000 cases of multidrug-resistant (MDR) tuberculosis (TB) occur every year globally, 9% of them being affected by extensively drug-resistant (XDR) strains of Mycobacterium tuberculosis. The treatment of MDR/XDR-TB is unfortunately long, toxic and expensive, and the success rate largely unsatisfactory (<20% among cases with resistance patterns beyond XDR).
The aim of this review is to summarise the available evidence-based updated international recommendations to manage MDR/XDR-TB, and to update the reader on the role of newly developed drugs (delamanid, bedaquiline and pretomanid) as well as repurposed drugs (linezolid and meropenem clavulanate, among others) used to treat these conditions within new regimens.
A nonsystematic review based on historical trials results as well as on recent literature and World Health Organization (WHO) guidelines has been performed, with special focus on the approach to managing MDR/XDR-TB.
The new, innovative global public health interventions, recently approved by WHO and known as the “End TB Strategy”, support the vision of a TB-free world with zero death, disease and suffering due to TB. Adequate, universally accessed treatment is a pre-requisite to reach TB elimination. New shorter, cheap, safe and effective anti-TB regimens are necessary to boost TB elimination.
The new WHO post-2015 End TB Strategy will support the efforts that research on new drugs and regimens requires http://ow.ly/LnJER
In 2014, the World Health Organization (WHO) estimated 9 million new tuberculosis (TB) cases had occurred globally in 2013, 480 000 of them being affected by multidrug-resistant (MDR) Mycobacterium tuberculosis strains . MDR-TB is defined as resistance in vitro to at least isoniazid and rifampicin, while extensively drug-resistant (XDR)-TB is resistant to at least one fluoroquinolone and one injectable second-line anti-TB drug in addition to isoniazid and rifampicin [2–5].
In “hot spots” (e.g. in some former Soviet Union countries), between 20% and 30% of the new TB cases are infected by MDR-TB strains (the global average is 3.5%) , while a proportion up to 50% is observed among retreated cases (figs 1 and 2) [3, 6, 7]. As MDR/XDR-TB clinical outcomes are largely suboptimal and their treatment very long, toxic and expensive, these difficult-to-treat cases are considered a serious threat to TB control and elimination [4, 5, 8, 9].
A recent study demonstrated that while in Germany the MDR-TB treatment-related costs exceed €50 000 , in Europe, the average cost to treat a single XDR-TB case is over €160 000 . The largest meta-analytic study presently available revealed that MDR-TB treatment success is only 54% (with 15% death, 8% failure/relapse and 23% default). When the drug resistance profile is beyond XDR (with increasing complexity), the outcomes are unfortunately lower: treatment success ranges from 40% to 19%, failure/relapse from 15% to 54% and death from 15% to 35% [4, 5].
Every day, clinicians managing these cases face relevant challenges that include frequent occurrence of adverse events, problems in patients' adherence, lack of clinical experience, and limited availability of adequate diagnostics and second-line anti-TB drugs (in some countries). The risk of acquiring further drug resistance is therefore real.
WHO has recently launched its innovative “End TB Strategy” , supporting the TB elimination strategy and the vision of a TB-free world with zero death, disease and suffering due to TB [10–12]. The new strategy clearly supports universal access to high-quality MDR-TB diagnosis and treatment. However, since the market launch of rifampicin in the early 1960s, no new anti-TB drug has been specifically developed until recently, while significant progress has been achieved in the area of diagnostics (e.g. Xpert MTB/RIF test; Cepheid, Sunnyvale, CA, USA) [1, 13]. The need for new drugs and regimens is obvious [14–16].
The aim of this review is to summarise, based on evidence, how to design an anti-TB regimen, and the updated international recommendations to manage MDR/XDR-TB, and to update the reader on the role of newly developed drugs (delamanid, bedaquiline and pretomanid) as well as repurposed drugs (linezolid and meropenem clavulanate, among others) used to treat these conditions within new regimens. A nonsystematic review based on historical trials results as well as on recent literature and WHO guidelines has been performed, with special focus on the approach to managing MDR/XDR-TB.
Designing a regimen to treat TB
The treatment regimens, approved TB drugs and the dosage of anti-TB drugs recommended by the evidence-based WHO guidelines (presently under revision) are summarised in tables 1 and 2. “New” and “retreatment” cases are clearly separated, 30 days of previous anti-TB treatment being the cut-off . New TB cases (irrespective of HIV status) should be treated for the first 2 months (intensive phase) with isoniazid, rifampicin, pyrazinamide and ethambutol, followed by isoniazid and rifampicin for the remaining 4 months (continuation phase) . The daily dosage is recommended (although the three times weekly dosing can be used during the continuation phase under directly observed therapy) as well as the fixed-dose combinations .
Drug susceptibility testing (DST) (rapid and/or conventional) is strongly recommended by WHO in all cases and particularly for those previously treated . While awaiting DST results, in settings with a medium or low probability of MDR-TB, retreatment cases could initially be treated with an empiric regimen including isoniazid, rifampicin, pyrazinamide, ethambutol and streptomycin for 2 months, followed by isoniazid, rifampicin, pyrazinamide and ethambutol for 1 month, and isoniazid, rifampicin, and ethambutol for 5 months.
Designing a regimen to treat MDR-TB
Based on their effectiveness and safety, WHO identified five groups of anti-TB drugs (table 3) . A stepwise approach is recommended to design treatment regimens for MDR-TB cases (table 4) [9, 19, 20]. Although DST-guided individualised regimens have to be preferred, standardised regimens are used in setting where DST is not available. They are based on representative patient population data, ensuring that treatment regimens are likely to include a sufficient number of active drugs (four) to ensure treatment success.
Rational MDR-TB treatment should include a minimum of four active drugs: a later-generation fluoroquinolone (moxifloxacin, gatifloxacin or levofloxacin) plus an injectable aminoglycoside (amikacin, capreomycin or kanamycin) plus any first-line drug to which the isolate is susceptible (e.g., pyrazinamide) plus the addition of one drug from group 4 (cycloserine, p-aminosalicylic acid (PAS), terizidone, prothionamide or ethionamide) (table 4). The drugs belonging to group 5 should be included whenever four active drugs are not available from previous groups. When using injectables, the minimum duration of the intensive phase is 8 months, the continuation phase lasting for 12–18 months, for a total treatment duration of at least 20 months. As a rule, 18 months need to be added to the date of the first negative culture to define the final treatment duration. In the case of failure to achieve culture conversion, the underlying causes (incorrect drug dosage, quality of drug supply, nonadherence factors, malabsorption and comorbidity) need to be identified and, possibly, corrected.
Unfortunately, the TB armamentarium is presently based on a limited number of effective drugs and the “optimised background regimens” (OBRs) used in clinical practice have not been tested in randomised controlled trials: clinicians are often forced to add some toxic old drugs (e.g. PAS, ethionamide and cycloserine) to more effective drugs that have been developed for indications other than TB.
Linezolid was prescribed “off label” before reliable data from multicentre observational studies were made available . At a later stage, systematic (individual and nonindividual) reviews and experimental data further defined the linezolid profile [22–27].
Linezolid, a first-generation oxazolidinone, demonstrated clinical effectiveness in most difficult-to-treat drug-resistant cases, although the frequency and severity of adverse events (i.e. peripheral neuropathy, optic neuropathy, gastrointestinal disorders and myelosuppression) limit its long-term use . A recent prospective randomised trial enrolling XDR-TB patients failing previous chemotherapy demonstrated the efficacy of a reduced linezolid dosage (300–600 mg per day), confirming previous findings : 87% of all enrolled patients achieved bacteriological conversion within 6 months . As four patients acquired resistance during treatment (three of them receiving 300 mg per day), additional evidence is necessary to assess the optimal dose and adequate duration of treatment. Interesting studies have been conducted to prevent adverse events while maintaining the efficacy of linezolid using intermittent dosing and of increasing linezolid concentration in combination with clarithromycin (table 5) [28, 29]. An innovative study from the Netherlands  suggested that clarithromycin can boost the blood levels of linezolid, allowing administration of lower doses with fewer adverse events and economic savings. A recent individual-data meta-analysis  provided updated evidence on efficacy, safety and tolerability of linezolid, and indirect evidence that a proper treatment drug monitoring (TDM)) approach to drug dosage can reduce linezolid toxicity (fig. 3) .
Recently, although not yet approved by regulatory authorities, another oxazolidinone drug (sutezolid) is attracting interest, being better tolerated .
Meropenem clavulanate is active in vitro against M. tuberculosis, showing a good tolerability profile. Meropenem and clavulanate together have a potent in vitro activity against M. tuberculosis, as clavulanate inhibits the extended-spectrum β-lactamase (BlaC) produced by TB bacilli, which generally hamper the activity of β-lactam antibiotics like meropenem. In a recently published case–control study, meropenem/clavulanate added to an OBR regimen containing linezolid (at the dose of 1 g three times a day) achieved high smear and culture conversion in MDR/XDR-TB (table 5) . Further studies are presently ongoing to define the role of meropenem/clavulanate within the group 5 drugs.
Currently, as a result of a multi-stakeholder initiative, two new anti-TB drugs were approved by the US Food and Drug Administration and by the European Medicines Agency: bedaquiline and delamanid [39, 40]. In addition, pretomanid, sutezolid (mentioned earlier), SQ109 and benzothiazinones will be discussed.
Bedaquiline, belonging to the diarylquinoline class of antibiotics (table 6), selectively targets the proton pump of ATP synthesis, leading to inadequate ATP synthesis, which is necessary for bacterial metabolism . The minimal inhibitory concentration (MIC) of bedaquiline against M. tuberculosis is very low, and its bactericidal activity in the murine model is superior to that of that of isoniazid and rifampicin . The results of two trials (phase II) suggested that a standard 2-month treatment regimen with bedaquiline yielded high culture conversion rates, rapid sputum culture conversion and low acquired resistance to companion drugs in newly diagnosed MDR-TB cases [41, 51].
Based on the available evidence, both WHO and the US Centers for Disease Control and Prevention recently issued recommendations that support the use of bedaquiline, at a dose of 400 mg daily for 2 weeks, then 200 mg three times a week for 22 weeks, added to OBR (as per WHO recommendations) to treat MDR-TB in adults when the following conditions are met: pharmacovigilance is in place, informed consent is ensured and QT monitoring is possible [52, 53]. As safety concerns for this drug remain because of an increased risk of death and QT prolongation, in the absence of additional evidence, the simultaneous use of bedaquiline and delamanid is not recommended [52, 53].
Delamanid and pretomanid
Delamanid and pretomanid belong to the nitroimidazole class of antibiotics, presently undergoing phase II and phase III clinical trials (table 6). They inhibit the synthesis of mycolic acids, which are components of the cell envelope of M. tuberculosis. Delamanid showed, both in vitro and in vivo, a high activity (and significant early bactericidal activity (EBA)) in adult cases affected by pulmonary TB [54, 55]. In a phase II randomised, placebo-controlled, multinational clinical trial in patients with MDR-TB, delamanid administered for 2 months at two different drug doses (100 and 200 mg twice daily) in addition to OBR achieved a significant increase in culture conversion compared with placebo (45% in the 100-mg group versus 42% in the 200-mg group and 29% in the placebo group) . In a subsequent open-label extension trial for MDR/XDR-TB, the mortality rates in patients receiving delamanid for at least 6 months were 1% versus 8.3%, in those treated for 2 months or less . Although delamanid was well tolerated, QT prolongation was more frequently reported in patients receiving delamanid against those receiving placebo . Based on the available evidence, WHO recommends the use if delamanid at the dose of 100 mg twice daily for 6 months, added to OBR in adults, when pharmacovigilance is in place and informed consent ensured .
A recent retrospective analysis of clinical studies carried out so far showed that the greatest reduction in mortality (around four-fold) occurred among patients treated with delamanid for ≥6 months versus those treated for ≤2 months . Although anecdotal evidence suggests that delamanid is effective and safe in children , two clinical trials (www.clinicaltrials.gov identifier numbers NCT01859923 and NCT01856634) are studying delamanid in the treatment of paediatric MDR-TB.
Pretomanid (previously known as PA-824) has a low MIC for M. tuberculosis, comparable to that of isoniazid. This drug was studied as part of different potential new regimens. Diacon and co-workers [40, 41, 44, 46] assessed the 14-day early bactericidal activity of a regimen composed of pretomanid, moxifloxacin and pyrazinamide, which proved to be significantly higher than that of bedaquiline alone, bedaquiline plus pyrazinamide and bedaquiline plus pretomanid (but not to pretomanid plus pyrazinamide), and comparable to that of the standard treatment regimen (isoniazid, rifampicin and pyrazinamide with streptomycin or ethambutol). Interestingly, the addition of pyrazinamide increased the activity of both bedaquiline and pretomanid.
In a new phase IIb trial, the bactericidal activity of a new 8-week regimens including moxifloxacin, pretomanid (100 or 200 mg, according to the arm), pyrazinamide and clofazimine was compared to that of the standard anti-TB regimen to treat sputum smear positive patients with both drug-susceptible and drug-resistant TB. The new regimen yielded higher bactericidal activity than the current WHO-recommended regimen after 2 months of treatment and was well tolerated (no episode of QT interval exceeding 500 ms was identified) .
Sutezolid (PNU-100480), like linezolid (see section on repurposed drugs), belongs to the oxazolidinone class of antibiotics. Their mechanism of action prevents the initiation of protein synthesis by binding to 23S RNA in the 50S ribosomal subunit of bacteria.
Having shown potent action against M. tuberculosis in the murine model , it was studied in phase I trials in humans, and appeared to be safe and well tolerated [59, 60]. Recently, a phase II clinical trial (NCT01225640) assessing safety and efficacy using EBA and whole-blood bactericidal activity was completed (table 6) .
SQ109, a 1,2-ethylenediamine, is an analogue of ethambutol. The drug is active against both drug-susceptible and drug-resistant TB by targeting MmpL3 in M. tuberculosis and specifically inhibiting the protein synthesis . In vitro, it has some synergic effects with bedaquiline and favourable interactions with sutezolid. SQ109 is presently undergoing phase II clinical trials (table 6) .
This new class of anti-TB drugs, in the pre-clinical development phase, are able to inhibit the synthesis of decaprenylphosphoarabinose, the precursor of the arabinans in the mycobacterial cell wall . Preliminary evidence suggests that BTZ043 is potent, with activity against 240 clinical isolates of M. tuberculosis, including drug-susceptible, MDR-TB and XDR-TB strains. Additive interactions and no antagonism were found between BTZ043 and rifampicin, isoniazid, ethambutol, pretomanid, moxifloxacin, meropenem with or without clavulanate, and SQ109, while synergic effects were found combining BTZ043 and bedaquiline (table 6) .
Improving the individualised approach to drug dosing: TDM
TDM is well known among clinical pharmacologists although it is not yet very popular among TB clinicians [63–65]. Based on the collection of blood samples, it allows the evaluation ex vivo of the blood concentration of a given drug, and consequently, the potential quantitative effect on the pharmacological target. The causes leading to the development of drug resistance in M. tuberculosis are well known, and include inadequate treatment (inadequate dose or dosing frequency) [66, 67], nonadherence to the prescribed regimen and pharmacokinetic (PK) variability [49, 68]. TDM has not yet been used to its full potential to improve therapy management when second-line anti-TB drugs are used of [65, 59].
Dose adjustments are important when treating patients who are slow to respond to treatment (because of inaccurate dosing, malabsorption or altered metabolism, drug–drug interactions  and PK variability ), have MDR/XDR-TB with adverse events or are intolerant to a given drug. TDM prevents, for example, the development of further drug resistance due to exposure to drug concentrations below the MIC. When managing XDR-TB and no more than four or five effective drugs are available, TDM is potentially life-saving as it can detect malabsorption and allow dose re-adjustment. Furthermore, TDM can reduce linezolid toxicity by reducing the dose necessary [23, 31].
While TDM is still considered an experimental procedure, a simple and cheap dried blood spot (DBS) method allows us to take advantage of the test even from remote settings . It needs only a small blood volume while allowing easy sampling, storage and transportation (fig. 3).
Some laboratories in the USA perform drug testing in-house (C.A. Peloquin, College of Pharmacy and Emerging Pathogens Institute, University of Florida, Gainesville, FL, USA; personal communication) ; in Europe, a proficiency testing programme is under development under the lead of the Nijmegen group in the Netherlands (C.A. Peloquin, personal communication). In the near future, a system of reference laboratories adequately covering clinical needs globally might receive the DBS by ordinary mail and rapidly send back the answer by e-mail. Although TDM is still expensive, a gradual decrease in cost is likely to occur in the future as a consequence s of an enlarged market. In addition, the reduction of the dose of drug used will pay back the TDM cost in a few days.
Improving the clinical management of MDR/XDR-TB: the European Respiratory Society/WHO TB Consilium
WHO recommends that management of MDR-TB cases is supported by a specialised team, which usually includes diverse professional perspectives (clinical, in adults and children; surgical; radiological; public health; psychological; nursing, etc.). The existence of WHO guidelines and European Standards of TB Care does not automatically guarantee appropriate treatment of MDR/XDR-TB, while rapid advice from a multidisciplinary team with clinical and management experience at national or supranational level would help to ensure proper treatment [9, 71, 72]. Similar consultation bodies dot not exist or are not used in the majority of the countries.
The European Respiratory Society (ERS), which is in the frontline of TB elimination in Europe [73–78], launched the electronic ERS/WHO TB Consilium in Vienna, Austria, in September 2012, under a Memorandum of Understanding signed with the WHO Regional Office for Europe. The overall aim of this initiative is to provide scientifically sound, evidence-based advice to national consilia and individual clinicians on how to manage drug-resistant TB and other difficult-to-treat TB cases, including co-infection with HIV and paediatric cases [14, 15, 79], in order to prevent development of further drug resistance, and ensure monitoring and evaluation of clinical practices in the different countries (prevention, diagnosis and treatment). The platform ensures a cost-free, multilingual (English, Portuguese, Russian and Spanish), Internet-based consultation system able to provide suggestions on clinical management of complicated TB cases in less than 3 days.
In collaboration with the WHO Regional Office for Europe, a new function aimed at supporting clinicians treating trans-border migrants affected by TB has been recently launched. Furthermore, new functionality is under development (in collaboration with public organisations), aimed at allowing patients to request specific advice directly for their TB infection or disease to a physician, a nurse or a psychologist, and/or to ask for motivational support from other patients who survived, in order to support the TB elimination roadmap [73–75].
The pipeline to design new regimens
Fluoroquinolones are considered pivotal in reducing treatment duration. A recent randomised trial, however, showed a higher relapse rate of 15% and 11% in the 4-month thrice-weekly regimens of gatifloxacin or moxifloxacin with isoniazid, rifampicin and pyrazinamide (2GHRZ3/2GHR3 or 2MHRZ3/2MHR3), respectively, compared with 6% in the standard 6-month regimen (2EHRZ3/4HR3) during the 24 months after the end of treatment .
Three noninferiority trials are presently assessing the use of fluoroquinolones to shorten the treatment duration to 4 months. The OFLOTUB Trial (NCT00216385) is comparing the standard 6-month regimen with a regimen consisting of a 2-month intensive phase with gatifloxacin substituting ethambutol, followed by a 2-month maintenance phase of gatifloxacin, isoniazid and rifampicin (2GHRZ/2GHR). Follow-up data have been obtained but the results have not yet been published [81–83]. REMox TB (NCT00864383) is comparing standard 6-month therapy with two study regimens (2 months of moxifloxacin, isoniazid, rifampicin and pyrazinamide followed by 2 months of moxifloxacin, isoniazid and rifampicin (2MHRZ/2MHR) or 2 months of ethambutol, moxifloxacin, rifampicin and pyrazinamide followed by 2 months of moxifloxacin and rifampicin (2EMRZ/2MR)). RIFAQUIN (ISRCTN44153044) is comparing the standard 6-month regimen with two study regimens (2 months of daily ethambutol, moxifloxacin, rifampicin and pyrazinamide followed by 2 months of twice-weekly moxifloxacin and rifapentine (2EMRZ/2PM2) or 4 months of once-weekly moxifloxacin and rifapentine (2EMRZ/4PM1) in a maintenance phase).
The very recent study by Dawson et al.  opens new perspectives on what regimens can be designed with the newly available anti-TB drugs . This new phase IIb trial compared the bactericidal activity of 8-week regimens including moxifloxacin and pretomanid (100 or 200 mg, according to the arm) plus pyrazinamide against the standard anti-TB regimen to treat sputum smear positive patients with both drug-susceptible and drug-resistant TB.
The bactericidal activity of the of the 8-week regimens was higher than that of the current WHO-recommended regimen in both drug-susceptible and drug-resistant TB after 2 months of treatment. The experimental treatment was well tolerated and no episode of QT interval exceeding 500 ms was identified.
The study showed that rifampicin-sparing regimen might work (allowing easier and safer treatment of HIV-positive cases with protease inhibitors) while achieving rapid sputum culture conversion and reduced transmission M. tuberculosis with the community. Given the potential to shorten treatment duration, hopes exist to improve patient adherence.
In spite of the progress achieved so far, much needs still to be done to improve our approach to clinical trials of new anti-TB drugs and regimens. The surrogate markers currently adopted to measure the efficacy of a given drug are old and need complicated statistical approaches to interpolate the scientific evidence . Furthermore, the time necessary to assess the pharmacological profile of a new drug is still very long and the possibility of enrolling vulnerable persons into the trials is a barrier we need to overcome soon . The new WHO post-2015 End TB Strategy will support the efforts that research on new drugs and regimens requires.
Conflict of interest: None declared.
- Received April 1, 2015.
- Accepted April 3, 2015.
- The content of this work is ©the authors or their employers. Design and branding are ©ERS 2015
This article is open access and distributed under the terms of the Creative Commons Attribution Non-Commercial Licence 4.0