Drug-resistant tuberculosis treatment

Table 3.2. Relative risk for treatment failure or relapse, and death (versus 
treatment success), 2018 IPD meta-analysis for longer MDR-TB regimens and 
delamanid Trial 213 (intent-to-treat population)
a
Medicine
Treatment failure or 
relapse versus treatment 
success
Death versus treatment 
success
Number 
treated
Adjusted odds 
ratio (95% CL)
Number 
treated
Adjusted odds 
ratio (95% CL)
A
Levofloxacin or 
moxifloxacin 
3143
0.3 (0.1–0.5)
3551
0.2 (0.1–0.3)
Bedaquiline
1391
0.3 (0.2–0.4)
1480
0.2 (0.2–0.3)
Linezolid
1216
0.3 (0.2–0.5)
1286
0.3 (0.2–0.3)
B
Clofazimine 
991
0.3 (0.2–0.5)
1096
0.4 (0.3–0.6)
Cycloserine or 
terizidone 
5483
0.6 (0.4–0.9)
6160
0.6 (0.5–0.8)

---

Recommendations 
30
Medicine
Treatment failure or 
relapse versus treatment 
success
Death versus treatment 
success
Number 
treated
Adjusted odds 
ratio (95% CL)
Number 
treated
Adjusted odds 
ratio (95% CL)
C
Ethambutol 
1163
0.4 (0.1–1.0)
1245
0.5 (0.1–1.7)
Delamanid
289
1.1 (0.4–2.8)
b
290
1.2 (0.5–3.0)
b
Pyrazinamide
1248
2.7 (0.7–10.9)
1272
1.2 (0.1–15.7)
Imipenem–
cilastatin or 
meropenem
206
0.4 (0.2–0.7)
204
0.2 (0.1–0.5)
Amikacin 
635
0.3 (0.1–0.8)
727
0.7 (0.4–1.2)
Streptomycin
226
0.5 (0.1–2.1)
238
0.1 (0.0–0.4)
Ethionamide or 
prothionamide
2582
1.6 (0.5–5.5)
2750
2.0 (0.8–5.3)
P-aminosalicylic 
acid
1564
3.1 (1.1–8.9)
1609
1.0 (0.6–1.6)
Other 
medicines
Kanamycin
2946
1.9 (1.0–3.4)
3269
1.1 (0.5–2.1)
Capreomycin
777
2.0 (1.1–3.5)
826
1.4 (0.7–2.8)
Amoxicillin–
clavulanic acid
492
1.7 (1.0–3.0)
534
2.2 (1.3–3.6)
CL: confidence limit; GDG: Guideline Development Group; IPD: individual patient data; MDR-TB: multidrug-resistant
tuberculosis.
a
See also text, 
Table 3.3
and Annex 3, Annex 4 and Annex 5 for more detail on how the estimates were derived and the additional 
factors considered by the GDG when reclassifying medicines for use in longer MDR-TB regimens as shown in 
Table 3.1
.
b
The values are the unadjusted risk ratios, as defined by the study investigators of Trial 213 by month
24.
Regarding PICO question 4 (MDR/RR-TB, 2018) (number of agents likely to be effective), the 
analysis showed that in longer MDR-TB treatment regimens, the risk of treatment failure, relapse 
and death was comparable when the treatment started with four, five or six medicines likely to be 
effective. The analysis also showed that patients who took three agents in the continuation phase – 
the situation expected when starting with four agents and stopping the injectable agent at the end 
of the intensive phase – fared no worse than those who took four agents in the continuation
phase.
Given that drug–drug interactions, pill burden and likelihood of adverse events all increase with 
the number of agents in a regimen, it would be desirable to give patients the minimum number of 
medicines necessary to obtain comparable levels of relapse-free cure. When deciding on the minimum 
number of agents to recommend, the GDG considered analyses that included injectable agents in the 
regimens, while fully cognizant that future longer regimens are expected to be increasingly injectable 
free. Moreover, it was important to provide for situations in which more than one medicine is stopped 
at some point during treatment, either because of its indication for use – bedaquiline and delamanid 
on-label use is 6 months – or because of tolerability (particularly linezolid; 
Table 3.3
) (72), meaning 
that for most of its duration, the regimen would contain two key agents fewer than at the start. While 
bedaquiline use beyond 6 months is referred to as off-label use, new evidence on the safety profile 
of bedaquiline use beyond 6 months was available to the GDG in 2019. This evidence supports the 
safe use of bedaquiline beyond 6 months in patients who receive appropriate schedules of baseline 

---

WHO consolidated 
guidelines 
on
tuberculosis: 
drug-resistant tuberculosis treatment
31
and follow-up monitoring. The use of bedaquiline beyond 6 months continues to be off-label use; 
thus, best practices in off-label use still apply. 
The 2018 IPD included experience from over 300 patients who were treated with linezolid for at 
least 1 month, mostly at a dose of 600 mg/day, with information on duration of use. About 30% only 
received linezolid for 1–6 months, but over 30% received it for more than 18 months, and these 
patients had the lowest frequency of treatment failure, loss to follow-up and death. A plot of linezolid 
duration and treatment failure suggests that the optimal duration of use would be around 20 months, 
corresponding to the usual total duration of a longer MDR-TB regimen. However, such an analysis does 
not account for survivorship bias, meaning that those who complete the full length of treatment are 
more likely to have a successful outcome, given that deaths and losses to follow-up occur earlier. No 
clear pattern could be discerned for type of adverse event and duration of use, although a few cases 
were reported with optic neuropathy, known to be associated with long-term use of linezolid (73), 
whereas haematological toxicity was reported regardless of duration of
use.