Substitution of ethambutol with linezolid during the intensive phase of treatment of pulmonary tuberculosis: a prospective, multicentre, randomised, open-label, phase 2 trial.

BACKGROUND: Linezolid improves the treatment outcomes of multidrug-resistant tuberculosis substantially. We investigated whether use of linezolid instead of ethambutol increases the proportion of sputum culture conversion at 8 weeks of treatment in patients with pulmonary tuberculosis.

METHODS: We did a phase 2, multicentre, randomised, open-label trial for patients with pulmonary tuberculosis at the three affiliated hospitals to Seoul National University and National Medical Center (Seoul-Seongnam, South Korea). Patients, aged 20-80 years, with a positive sputum for pulmonary tuberculosis, but without resistance to rifampicin, and current treatment administered for 7 days or fewer, were randomly assigned at a 1:1:1 ratio into three groups. The control group received ethambutol (2 months) with isoniazid, rifampicin, and pyrazinamide. The second group used linezolid (600 mg/day) for 2 weeks and the third group for 4 weeks instead of ethambutol for 2 months. We used a minimisation method to randomise, and stratified according to institution, cavitation on chest radiographs, and diabetes. The primary endpoint was the proportion of patients with negative culture conversion of sputum in liquid media after 8 weeks of treatment. The results of this trial were analysed primarily in the modified intention-to-treat population. The trial is registered with ClinicalTrials.gov, number NCT01994460.

FINDINGS: Between Feb 19, 2014, and Jan 13, 2017, a total of 429 patients were enrolled and 428 were randomly assigned into either the control group (142 patients), the linezolid 2 weeks group (143 patients), or the linezolid 4 weeks group (143 patients). Among them, 401 were eligible for primary efficacy analyses. In the modified intention-to-treat analyses, negative cultures in liquid media at 8 weeks of treatment were observed in 103 (76·9%) of 134 control patients, 111 (82·2%) of 135 in the linezolid 2 weeks group, and 100 (75·8%) of 132 in the linezolid 4 weeks groups. The difference from the control group was 5.4% (95% CI -4·3 to 15·0, p=0·28) for the linezolid 2 weeks group and -1·1% (-11·3 to 9·1, p=0·83) for the linezolid 4 weeks group. Numbers of patients who experienced at least one adverse event were similar across the groups (86 [62·8%] of 137 in control, 79 [57·2%] of 138 in the linezolid 2 weeks group, and 75 [62·0%] of 121 in the linezolid 4 weeks group). Resistance to linezolid was not identified in any patient.

INTERPRETATION: Higher rates of culture conversion at 8 weeks of treatment with short-term use of linezolid were not observed. However, safety analyses and the resistance profile suggested the potential role of linezolid in shortening of treatment for drug-susceptible tuberculosis.

FUNDING: Ministry of Health and Welfare, South Korea.