We collected data on patients <18 years of age from the TB treatment programs supported by Médecins Sans Frontières in South Africa, Tajikistan, and Uzbekistan and the National TB Programme in Belarus. During November 2014–January 2017, a total of 27 children and adolescents began regimens containing bedaquiline (Table). Median age was 16 (range 10–17) years, and 15 (56%) were girls. Median weight was 50 (range 35–76) kg. No patients were HIV positive. One male patient had intrathoracic lymph node TB, and 26 patients had pulmonary TB. Diagnoses for 17 (63%) patients were confirmed by mycobacterial culture. Baseline sputum smears from 19 (70%) patients were positive for acid-fast bacilli. One boy had concomitant spinal TB osteomyelitis. Most patients (18/27, 67%) had XDR TB; 6 (22%) had MDR TB with fluoroquinolone resistance; and 3 (11%) had MDR TB with resistance to a second-line injectable drug. For the 10 patients without positive mycobacterial cultures, drug susceptibility was presumed from contact history. Thus, for all patients, the decision to use bedaquiline was based on confirmed or presumed extensive drug resistance that resulted in the inability to construct an effective treatment regimen.

The high proportion of patients with resistance to second-line drugs led to frequent use of repurposed drugs, such as linezolid (26/27, 96%) and clofazimine (26/27, 96%). Intravenous imipenem was used in some patients (4/27, 15%). Despite concerns about potential additive cardiac toxicity when combining bedaquiline and moxifloxacin, combined treatment was judged necessary for 6 (22%) children. Five of these 6 children also received clofazimine in their regimen. All cases except 1 received the recommended adult dosing regimen for bedaquiline (400 mg/d for 2 weeks, then 200 mg 3×/wk for 6 months). One 10-year-old girl (weighing 35 kg) received 300 mg/d (recommended on the basis of expert opinion) during her loading phase.

The mean duration of bedaquiline treatment of the 20 children and adolescents who completed therapy was 172 days. Regular (weekly for most) electrocardiogram monitoring was performed for all cases. Fridericia’s formula was used to correct the measured QT intervals (QTcF) for the heart rate, and cardiotoxicity was defined according to the National Cancer Institute Common Terminology Criteria for Adverse Events version 4.03 (https://evs.nci.nih.gov/ftp1/CTCAE/CTCAE_4.03_2010-06-14_QuickReference_8.5x11.pdf). Five patients had grade 3 or 4 prolongation: 2 received bedaquiline, clofazimine, and moxifloxacin and 3 received bedaquiline and clofazimine. Four patients experienced increases in QTcF >60 ms above baseline: 2 during the first month of treatment, 1 after 3 months of treatment, and 1 after 6 months of treatment. All 4 patients improved <1 month following electrolyte replacement, and no drug cessation was required. Recurrent prolongation of QTcF >500 ms was identified in 1 adolescent during the first 6 months of treatment and necessitated cessation of moxifloxacin and clofazimine after electrolyte replacement showed insufficient benefit. Following stoppage of these drugs, QTcF returned to normal. No patient experienced symptoms attributable to prolongation of QTcF during treatment with bedaquiline.

As of February 24, 2017, of the 23 patients who remained on treatment and had data available, all were culture negative; 14 of these 23 had been positive at baseline. No clinical signs suggestive of treatment failure were noted among patients of this cohort.

Our experience suggests that bedaquiline can be used safely in children >12 years of age with appropriate monitoring and could be considered in younger children in select circumstances when benefits are likely to outweigh risks. Although treatment outcomes are preliminary, we report good responses to treatment with bedaquiline among a group of children and adolescents with advanced resistance to second-line drugs. Although prolongation of QTcF was noted in some (5/27) patients when concomitant cardiotoxic drugs were used, no patient required bedaquiline cessation.

Continued reluctance to use contact history for diagnosing advanced drug resistance and limited availability of drug susceptibility testing in children remain barriers for the consideration of new drugs and use of appropriate MDR TB regimens. In addition, restricted availability of delamanid in TB programs and the perceived age restriction on the use of bedaquiline has resulted in children failing to benefit from drugs that are being used safely and successfully in adults. Although the lack of pharmacokinetic data on bedaquiline in children and adolescents must be addressed, other second-line TB drugs have been recommended and prescribed despite insufficient data on pharmacokinetics. Expanding access to bedaquiline and delamanid for children could lead to the reduction in the need for second-line injectable drugs, which are strongly associated with irreversible toxicity (5). This experience supports similar recommendations given by the US Centers for Disease Control and Prevention (9) and an international group of pediatric TB experts (10).