Once-daily oral omadacycline versus twice-daily oral linezolid for acute bacterial skin and skin structure infections (OASIS-2): a phase 3, double-blind, multicentre, randomised, controlled, non-inferiority trial

Background Pathogen resistance and safety concerns limit oral antibiotic options for the treatment of acute bacterial skin and skin structure infections (ABSSSI). We aimed to compare the efficacy and safety of once-daily oral omadacycline, an aminomethylcycline antibiotic, versus twice-daily oral linezolid for treatment of ABSSSI. Methods In this phase 3, double-blind, randomised, non-inferiority study, eligible adults with ABSSSI at 33 sites in the USA were randomly assigned (1:1) to receive omadacycline (450 mg orally every 24 h over the first 48 h then 300 mg orally every 24 h) or linezolid (600 mg orally every 12 h) for 7–14 days. Randomisation was done via an interactive response system using a computer-generated schedule, and stratified by type of infection (wound infection, cellulitis or erysipelas, or major abscess) and receipt (yes or no) of allowed previous antibacterial treatment. Investigators, funders, and patients were masked to treatment assignments. Primary endpoints were early clinical response, 48–72 h after first dose, in the modified intention-to-treat (mITT) population (randomised patients without solely Gram-negative ABSSSI pathogens at baseline), and investigator-assessed clinical response at post-treatment evaluation, 7–14 days after the last dose, in the mITT population and clinically evaluable population (ie, mITT patients who had a qualifying infection as per study-entry criteria, received study drug, did not receive a confounding antibiotic, and had an assessment of outcome during the protocol-defined window). The safety population included randomised patients who received any amount of study drug. We set a non-inferiority margin of 10%. This study is registered with ClinicalTrials.gov, NCT02877927, and is complete. Findings Between Aug 11, 2016, and June 6, 2017, 861 participants were assessed for eligibility. 735 participants were randomly assigned, of whom 368 received omadacycline and 367 received linezolid. Omadacycline (315 [88%] of 360) was non-inferior to linezolid (297 [83%] of 360) for early clinical response (percentage-point difference 5·0, 95% CI −0·2 to 10·3) in the mITT population. For investigator-assessed clinical response at post-treatment evaluation, omadacycline was non-inferior to linezolid in the mITT (303 [84%] of 360 vs 291 [81%] of 360; percentage-point difference 3·3, 95% CI −2·2 to 9·0) and clinically evaluable (278 [98%] of 284 vs 279 [96%] of 292; 2·3, −0·5 to 5·8) populations. Mild to moderate nausea and vomiting were the most frequent treatment-emergent adverse events in omadacycline (111 [30%] of 368 and 62 [17%] of 368, respectively) and linezolid (28 [8%] of 367 and 11 [3%] of 367, respectively) groups. Interpretation Once-daily oral omadacycline was non-inferior to twice-daily oral linezolid in adults with ABSSSI, and was safe and well tolerated. Oral-only omadacycline represents a new treatment option for ABSSSI, with potential for reduction in hospital admissions and cost savings. Funding Paratek Pharmaceuticals.