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

Data for efficacy analysis was provided by the manufacturer derived from the three trial/studies mentioned above, with data collected at various time periods and from a variety of patient populations. It should be emphasized that the study populations in Trial 208 and Study 116 are in fact derived from the cohort of Trial 204; accordingly, all three studies involved the same participants (or a subset thereof). According to published reports (10) (11) and results made available to WHO, the key endpoints used to assess the efficacy of delamanid were classified by the manufacturer as follows:

• Short-term efficacy: endpoints were measured in the pivotal two months phase II, randomized placebo-controlled clinical trial 204

  –

  Sputum culture conversion at two months (two-month SCC) – the primary endpoint, defined as a patient's achieving a series of at least five consecutive weekly cultures negative for growth of M. tuberculosis (without subsequent positive cultures). Assessment of two-month SCC using the MGIT liquid culture system served as the primary efficacy analysis;

  –

  Time to SCC – the time from treatment initiation to the first of the series of cultures that defined SCC; and

  –

  Time to detection of positive culture results using the MGIT system.

• Longer term efficacy: endpoints were measured through combination of data (using solid culture) from Trial 204, Trial 208 and Study 116, grouping patients according to the total duration of delamanid received in the various trials, irrespective of the dose received (100mg BD or 200mg BD).

  –

  Sustained SCC – Proportion of patients with no positive culture results throughout the remaining course of treatment after SCC has been achieved (consistently negative sputum cultures during the continuation phase of MDR-TB treatment ultimately defined a favorable outcome);

  –

  Favorable outcome – as assessed by the managing clinician at 24 months based on WHO treatment outcome definitions (cure, treatment completion, failure, death); and

  –

  Mortality – assessed as all-cause mortality per WHO guidelines.⁴

The primary efficacy endpoint, two-month sputum culture conversion (SCC), arose from the pivotal Trial 204. Overall, 434 patients completed the trial and the percentage of patients who completed or discontinued the trial was evenly distributed across groups. Efficacy analysis was based on a modified intent to treat (MITT) population (N=402), including randomized patients who had a positive sputum culture for TB at baseline using the MGIT culture system and who were resistant to isoniazid and rifampicin.

A higher proportion of patients treated with delamanid + OBR achieved SCC by Day 57 using the MGIT culture system than patients treated with placebo + OBR: 64/141 (45.4%) [p=0.008] in the delamanid 100mg group, 57/136 (41.9%) [p=0.04] in the delamanid 200mg BD + OBR group, and 37/125 (29.6%) in the placebo + OBR group – see Figure 2. Similar results were obtained when using solid culture (data not shown).

Figure 2

Proportion of patients with sputum-culture conversion by Day 57. Source: Gler et al, 2012 (11).

A secondary efficacy endpoint was time to SCC. Although the median time to SCC (sustained SCC achieved by Day 57) could not be calculated because fewer than 50% of patients in each group met the criteria, the Kaplan-Meier curves for time to SCC using the MGIT system showed clear separation between each delamanid BD + OBR group and the placebo + OBR group from Day 36 to Day 57 (Weeks five to eight). By the end of the two-month treatment period, the difference in SCC between the delamanid groups (both doses) and the placebo group was significant (p=0.001). The hazard ratio for increased time to conversion to a negative sputum culture as assessed with MGIT was 0.58 (95% CI: 0.39 to 0.89) in the 100mg BD group and 0.63 (95% CI: 0.42 to 0.96) in the 200mg BD group. When using solid medium, the hazard ratio for increased time to conversion to a negative sputum culture was 0.54 (95% CI, 0.36 to 0.81) in the 100mg BD group and 0.44 (95% CI, 0.29 to 0.64) in the 200mg BD group.

To assess the long-term efficacy of delamanid for MDR-TB beyond the first two months of treatment, the manufacturer combined data from Trial 204, Trial 208 and Study 116 to evaluate the proportion of patients who achieved sustained SCC and final treatment outcomes at 24 months (including mortality). In total, follow-up could be assessed in 421 (87.5%) out of the 481 patients initially randomized in Trial 204. To compare treatment outcomes, patients were grouped as follows:

• Patients treated with delamanid 100mg BD or 200mg BD (+ OBR) were grouped together. The rationale for this was that short-term treatment results (two-month SCC) were similar for those participants who received 100mg BD and 200mg BD;
• Patients who participated in Trial 208 and received six months of delamanid (any dose) were grouped together with those who had received placebo, delamanid 100mg BID, or delamanid 200mg BD during Trial 204 and designated as the “≥6month” administration group. Individuals in this group thus received six or eight months of delamanid. The manufacturer's rationale for this was that treatment outcomes and sustained SCC results were similar among those who had received six vs. eight months of delamanid.
• Patients who only contributed to trial 204 and received delamanid for two months or who received placebo were pooled to make a delamanid “≤2 month” administration group, because of comparable demographics and baseline characteristics and similar final treatment outcome. Hence individuals in this group received two months of delamanid or no delamanid at all.

Analyses performed by the manufacturer to evaluate the long-term efficacy of delamanid compared patients who received delamanid for ≥6 months with patients receiving delamanid for ≤2 months. The key endpoints for assessing longer term efficacy included:

• Sustained SCC: By the end of treatment, 90.9% (130/143) of patients in the delamanid ≥6months treatment group achieved sustained SCC compared to 70.9% (112/158) of patients in the delamanid ≤2months treatment group.
• Favourable outcome: The proportion of patients who showed a favourable treatment outcome (i.e. confirmed microbiological cure by solid culture or treatment completion) at the end of the 24-month treatment period was significantly higher in the ≥6 month group (74.5%, 95% CI: 67.7–80.5) than in the ≤2 month group (55.0%, 95% CI: 48.3–61.6) (p < 0.0001).
• Mortality: The reported mortality rate was lower in the delamanid ≥6 months treatment group (two deaths, 1%) compared to the delamanid ≤2 months treatment group (19 deaths, 8.3%) (p < 0.001).

Table 1

Favourable treatment outcome and mortality at 24 months for MDR-TB patients treated with delamanid.

In order to clarify the participation of patients in the above analysis, WHO requested the manufacturer to provide treatment outcomes at 24 months (using WHO-recommended treatment outcome definitions) stratified by delamanid dose and duration of exposure for patients who had been followed up within Study 116. The results are provided in Table 2.

Table 2

Outcome at 24 months for patients consenting to participate in Study 116 who were treated with delamanid 100mg BD or 200mg BD + OBR for two, six or eight months or with OBR alone, using WHO treatment outcome categories (N=421).

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

Twelve phase I trials in healthy subjects have been conducted. Pooled data provided by the manufacturer showed that treatment-emergent adverse effects (TEAEs) with the highest incidence in subjects who received delamanid were headache (20.9%, 88/422), nausea (11.6%, 49/422), and dizziness (8.8%, 37/422).

In total, 887 individuals have been exposed to delamanid in clinical trials. Overall, 22.1% (196/887) of delamanid-treated patients had a cumulative exposure of longer than six months (180 days). Most frequent were headache, abdominal pain and insomnia. The incidence and distribution of TEAEs occurring in 10% of patients in the two delamanid groups in Trial 204 are shown in Table 3 The only clinically relevant TEAE with a difference in incidence between the delamanid+OBR treatment groups compared to the placebo+OBR group was QT prolongation.⁵ Other TEAEs varied in occurrence but were present in similar proportions in the delamanid+OBR and the placebo+OBR groups. Most frequent were nausea, vomiting, and dizziness.

Table 3

Incidence of adverse events (occurring in 10% of patients in either delamanid group and with greater frequency than in the placebo group).

In Trial 204 and Trial 208, 74 patients experienced severe adverse events (SAEs), including death. The only clinically relevant SAE with a difference in incidence among treatment groups was QT interval prolongation, which was significantly higher in the delamanid 100mg BD + OBR group (4.3%, 7/161) and the delamanid 200mg BD + OBR group (5.6%, 9/160) than the placebo + OBR group (1.9%, 3/160). In Trial 204, the placebo-corrected, change-from-baseline QTcF⁶ increased with duration of dosing and reached 11.3 to 13.1 msec in delamanid 100mg BID + OBR and 14.1 to 15.6 msec in delamanid 200mg BID + OBR over all time points on Day 56. In Trial 208, four of 131 patients treated with delamanid 100mg BID + OBR and four of 73 patients treated with delamanid 200mg BID + OBR) had a change from baseline QTcF exceeding 60 msec. Only patients on delamanid exhibited QTcF prolongation exceeding 60 msec from baseline.

Delamanid is metabolized by cytochrome P450 enzymes like CYP3A4, and formation of its main metabolite is regulated by plasma albumin. It is neither an inducer nor an inhibitor of key drug metabolizing enzymes so is unlikely to have a significant impact on concentrations of companion drugs. Studies in healthy subjects showed no clinically-significant interactions when delamanid was co-administered with tenofovir, efavirenz or lopinavir/ritonavir.

4.3. Cost-effectiveness

The EG assessed the results of a cost-effectiveness analysis (CEA) conducted to model the incremental cost-effectiveness of adding delamanid to existing WHO-recommended MDR-TB regimens. This CEA was undertaken for different settings to allow for variation among countries across income level, the model of care used for MDR-TB treatment, and background patterns of drug resistance. It focused on the direct benefits to patients, but did not attempt to assess the indirect (and acquired) transmission benefits, nor did it assess the broader economic benefits to patients or society.

Since several analyses were conducted by the manufacturer to assess efficacy (see above), a sensitivity analysis was performed on the cost-effectiveness of delamanid when different trial data and assumptions about the translation of trial results to current practice were applied. Results showed that delamanid would be cost-effective in most environments studied. However, this interpretation needs to take into account the quality of clinical evidence as assessed by the EG (i.e. if the quality of clinical evidence is viewed as low, then likewise the evidence supporting cost-effectiveness should be regarded as low), as well as all limitations related to the assumptions made above. Of note, in settings where cure/treatment success rate is currently high, delamanid may not be cost-effective, as it may result in limited additional benefit. However, the incremental cost of delamanid introduction will not only depend on price, but also on the cost savings for retreatment as a result of an expected reduction in treatment failures. Using a conservative approach, delamanid was thus found to be cost-effective in most settings, but the quality of this evidence was considered very low, and further work would be needed to evaluate cost-effectiveness and to test the robustness of the assumptions in various settings.

Footnotes

1

References for documents available on delamanid can be found at the website indicated in page 3 of this Guidance document.

2

China, Egypt, Estonia, Japan, Latvia, Peru, the Philippines, Republic of Korea, and the United States of America.

3

Trial 208 and Study 116 are extensions of Trial 204. They are referred to by the terminology used in the published references and as provided by the manufacturer; it should be noted that these are not different/separate studies, but involved non-randomized and selected patients from the original Trial 204.

4

Generally high among MDR-TB patients with 15% reported in a large meta-analysis of MDR-TB treatment (12).

5

The heart's electrical cycle can be measured on an electrocardiogram (ECG); the QT interval is a measure of the time between the start of the Q wave and the end of the T wave representing the electrical depolarization and repolarization of the ventricles. A lengthened QT interval is a marker for potential ventricular tachyarrhythmias (such as ‘torsades de pointes’) and a risk factor for sudden death. Concomitant use of drugs that prolong the QT interval may cause additive QT prolongation, and should be avoided if possible. Of note, some of the other second-line anti-TB drugs are known to prolong the QT interval.

6

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