4Evidence base for policy formulation

Publication Details

Publicly available data on the pre-clinical and clinical development of bedaquiline were reviewed. These included toxicity, dosing and pharmacokinetic studies, drug–drug interaction (DDI) studies, an early bactericidal activity study, safety studies, a pivotal Phase IIb clinical trial and an (ongoing) single arm open-label trial.13,14

A total of 265 subjects participated in 11 Phase I trials with bedaquiline (208 subjects were enrolled in eight single-dose trials evaluating bedaquiline doses up to 800 mg; and 57 subjects were enrolled in three multiple-dose trials evaluating bedaquiline doses up to 400 mg daily with a maximum treatment duration of 15 days). The Phase I trials provided a basic understanding of bedaquiline's pharmacokinetic characteristics, DDI potential, and short-term safety/tolerability in healthy subjects and in a special population of moderately hepatic-impaired subjects. A double-blind, single-dose trial was conducted to evaluate the effect of a single supra-therapeutic (800 mg) dose of bedaquiline on the QT corrected (QTc) interval.

A Phase IIa, 7-day extended early bactericidal activity trial in 75 patients with drug-susceptible TB (evaluating doses up to 400 mg bedaquiline daily) was conducted to evaluate clinical antimycobacterial activity of bedaquiline.

The bedaquiline Phase II programme encompassed 2 Phase IIb clinical trials: C208 and C209. Trial C208 consisted of two stages, of which Stage 1 was an exploratory study and Stage 2 was a multi-centre, stratified, randomized, double-blind placebo-controlled trial, serving as a pivotal proof-of-efficacy study. Study C209 is a single-arm, open label trial (ongoing).

4.1. Evidence for the efficacy of bedaquiline in the treatment of MDR-TB

Evidence for efficacy derives from the C208 Stage 2 trial, in which subjects aged 18 to 65 years with newly diagnosed MDR-TB – enrolled from 15 sites in Brazil, India, Latvia, Peru, the Philippines, the Russian Federation, South Africa and Thailand – were randomized in a 1:1 ratio to receive bedaquiline 400 mg, or placebo, daily for the first two weeks, followed by 200 mg bedaquiline, or placebo, three times per week for the remaining 22 weeks.15 In both the bedaquiline and placebo arms, patients received a five-drug MDR-TB background medication regimen (BR) consisting of fluoroquinolones (mainly ofloxacin), aminoglycosides (mainly kanamycin), pyrazinamide, ethionamide, ethambutol, and cycloserine/terizidone in various combinations. After 24 weeks, subjects continued the BR of MDR-TB therapy until a total treatment duration of 96 weeks was achieved. The total duration of the study was 120 weeks. All subjects presented in the data sets completed Week 72 (the pre-determined study data cut-off point) and also Week 120 (end of study).

The primary efficacy endpoint for C208 Stage 2 was time to sputum culture conversion16 in commercial liquid culture (MGIT™ 960 Mycobacterial Detection System, Becton Dickinson Diagnostic systems, USA) during the 24-week investigational treatment period, evaluated after all subjects had completed the 24-week investigational treatment period, or discontinued earlier. In the primary efficacy analysis, subjects who discontinued before week 24 were considered as not having culture converted (censored at the last culture visit, i.e. missing = failure). Primary efficacy analysis was based on a modified intention to treat (mITT) population, which excluded subjects who had drug-susceptible TB, XDR-TB or unconfirmed MDR-TB (based on susceptibility tests taken prior to randomization), or had missing or negative baseline cultures, or who were positive at baseline, but had no post-baseline culture results. The mITT population was composed of 132 subjects (66 in each of the bedaquiline and placebo groups). The median time to culture conversion was 83 days (95%CI: 56, 97) in the bedaquiline group compared to 125 days (95%CI: 98, 168) in the placebo group. Primary analysis at Week 24 using the Cox proportional hazards model (adjusted for lung cavitation and pooled centre) showed a statistically significant difference in time to culture conversion between the two treatment groups in favour of bedaquiline: HR=2.44 [1.57, 3.80] (p<0.0001).

The secondary endpoint for C208 Stage 2 was the proportion of patients with culture conversion. The proportion of subjects with culture conversion at Week 24 (i.e. 24-week responders [missing = failure]) was 78.8% in the bedaquiline arm and 57.6% in the placebo arm (p = 0.008, based on a logistic regression model with only treatment as covariate). Similar analyses were conducted at Week 72 and Week 120. The percentage of responders (missing = failure) at Week 72 (i.e. the time point attained by all Stage 2 subjects at the interim analysis who were ongoing in the trial) was 71.2% in the bedaquiline group and 56.1% in the placebo group (p= 0.069). Utilizing all available efficacy data up to end of study (Week 120), the percentage was 62.1% in the bedaquiline group and 43.9% in the placebo group (p= 0.035).

Efficacy was further evaluated using WHO-recommended treatment outcome definitions applied to Week 120 final data. Cure was defined as: “at least five consecutive negative cultures from samples collected at least 30 days apart in the final 12 months of treatment; if only one positive culture is reported during that time, a patient may still be considered cured, provided that this positive culture is followed by a minimum of three consecutive negative cultures taken at least 30 days apart”. In the bedaquiline arm, 38/66 (57.6%) subjects were categorized as cured, compared to 21/66 (31.8%) in the placebo arm (p=0.003).

Table 1. Summary of evidence for the efficacy of bedaquiline in the treatment of MDR-TB.

Table 1

Summary of evidence for the efficacy of bedaquiline in the treatment of MDR-TB.

4.2. Evidence for the safety of bedaquiline in the treatment of MDR-TB

The safety database covered non-clinical aspects (pharmacology and toxicology) during pre-clinical development, and human experience in Study C208 (pivotal randomized control trial, double-blind placebo-controlled) and Study C209 (single arm, open label). The intention to treat (ITT) population in each of these studies was used for the description of safety. A total of 160 subjects contributed to ITT analysis, 79 in the bedaquiline arm and 81 in the placebo arm.

Similar numbers of patients in the bedaquiline group and placebo group reported adverse events (AEs) (Table 2). The most frequently reported AEs in the bedaquiline group (from both controlled and uncontrolled trials) were nausea, arthralgia, headache and vomiting. Additional AEs identified were, in order of frequency: dizziness, increased transaminases, myalgia, diarrhoea and QT prolongation on electrocardiogram (ECG). AEs of at least grade 3 were similar in both groups: 28/102 (27.5) in the bedaquiline group and 24/105 (22.9) in the placebo group. Main safety concerns included QT prolongation and cardiac events, hepatic events, and deaths.

Table 2. Summary of adverse events of interest.

Table 2

Summary of adverse events of interest.

Cardiovascular safety (Trial C208: pooled experience Stage 1 and Stage 2)

Mean QTcF 17 increases were observed in both the pooled bedaquiline (‘Any bedaquiline’) and pooled placebo (‘Any placebo’) groups, but they were more pronounced in the Any bedaquiline group: more patients had QTcF values above 450 ms (26.6% versus 8.6%) and more patients had QTcF increases >60 ms from reference values (9.1 % versus 2.5%). There were no reports of Torsade de Pointes events, and no reported fatalities from sudden death. Bedaquiline, in multiple dosing, can prolong the QT interval and the risk is highest during the treatment phase, but could extend beyond the treatment period. The use of bedaquiline with QT-prolonging medications increases the risk of prolonged QT interval, i.e. QTcF prolongation from multiple QTcF prolonging drugs could be additive (e.g. clofazimine).

Table 3. QT prolongation during treatment as reflected by worst QTcF.

Table 3

QT prolongation during treatment as reflected by worst QTcF.

Hepatic events (Trial C208: pooled experience Stage 1 and Stage 2)

There was a higher incidence of events related to hepatic disorders in the Any bedaquiline group (9 subjects, 8.8%) compared to the Any placebo group (2 subjects, 1.9 %). Increases in transaminases accounted for the majority of these reported events. An analysis to identify cases of severe liver toxicity revealed 1 case of a patient who experienced concurrent >3-fold elevation of aspartate aminotransferase (AST) and >2-fold elevation in total bilirubin, but was confounded by reported alcoholic hepatitis and concurrent intake of hepatotoxic background medications.

Table 4. Investigator-reported hepatic events.

Table 4

Investigator-reported hepatic events.


Four deaths were reported from the C208 Stage 1 trial: 2 out of 23 subjects (8.7%) in the bedaquiline arm and 2 out of 24 subjects (8.3 %) in the placebo arm. In the C208 Stage 2 trial, twelve deaths were reported in total (irrespective of when deaths occurred). Of these, 10/79 (12.7%) came from the bedaquiline group and 2/81 (2.5%) from the placebo group (p=0.017) (ITT analysis). In the bedaquiline group, 8 of the 10 deaths occurred in culture converters. TB was the cause of death in the two placebo-arm deaths and in 5 of the 10 bedaquiline-arm deaths (all occurred off bedaquiline treatment). Counting deaths strictly at the 120 weeks cut-off point reveal nine deaths in the bedaquiline and one death in the placebo group. There was no discernible pattern between death and culture conversion, relapse, microbiological response, susceptibility to drugs used in the BR, HIV status, or severity of disease. Despite detailed descriptive line listings of all deaths, the reason(s) for the imbalance were not clear.

Table 5. Trial C208 Stage 2: Causes of death.

Table 5

Trial C208 Stage 2: Causes of death.

4.3. Cost effectiveness

Modelling of the incremental cost-effectiveness of adding bedaquiline to WHO-recommended MDR-TB regimens was conducted by an independent consultant contracted by WHO for review by the EG. The model assumed that bedaquiline would be added to treatment for all patients starting MDR-TB treatment. Data from WHO were available on current MDR-TB treatment costs (excluding programme costs) and effectiveness in several high TB burden settings. Several scenarios were explored to appraise the cost-effectiveness of bedaquiline in these settings. Under the model assumptions, the bedaquiline-containing regimens were assessed as relatively cost-effective in most settings, but results were ambiguous in low-income settings, and highly dependent on the assumptions made about the generalizability of trial results to routine settings. The EG noted that further analysis would be needed to test the robustness of the assumptions in various settings and to separately assess affordability. As the recommendation of the EG is to use bedaquiline for only selected sub-groups of the full MDR-TB patient population (as opposed to all patients with MDR-TB that were considered in the cost-effectiveness analysis), the cost-effectiveness model needs to be further refined such that results are available for these sub-groups specifically.



Janssen Pharmaceutical Companies. TMC207 (bedaquiline) treatment of patients with MDR-TB (NDA 204–384). Briefing document to the Anti-Infective Drugs Advisory Committee Meeting; 28 November 2012; 2012. All documents available at: http://www​.fda.gov/AdvisoryCommittees​/CommitteesMeetingMaterials​/Drugs/Anti-InfectiveDrugsAdvisoryCommittee/ucm293600​.htm.


References for all documents available on bedaquiline can be found at the website indicated in page 1 of this document.


This dose regimen was selected based on non-clinical safety and microbiology data as well as safety and pharmacokinetic results from several Phase I clinical trials with bedaquiline, and early bactericidal activity results from the earlier Phase IIa trial C202.


Defined as: “two consecutive negative cultures from sputa collected at least 25 days apart (as well as all intermediate cultures), and this culture negativity was not followed by a confirmed positive MGIT culture (or a single positive sputum result after the subject completed the trial), and the subject did not discontinue up to the time point being analyzed”.


QTcF: QT interval corrected for heart rate according to the Fridericia method.