Cadazolid for the treatment of Clostridium difficile infection: results of two double-blind, placebo-controlled, non-inferiority, randomised phase 3 trials

Background Cadazolid is a novel quinoxolidinone antibiotic developed for treating Clostridium difficile infection. We aimed to investigate the safety and efficacy of cadazolid compared with vancomycin in patients with C difficile infection. Methods IMPACT 1 and IMPACT 2 were identically designed, multicentre, double-blind, placebo-controlled, noninferiority, randomised phase 3 trials. IMPACT 1 was done in Australia, Brazil, Canada, France, Germany, Italy, the Netherlands, Peru, Poland, Romania, Spain, and the USA, and IMPACT 2 was done in Argentina, Belgium, Brazil, Canada, Chile, Croatia, Czech Republic, Greece, Hungary, Israel, Romania, Slovakia, South Korea, the UK, and the USA. Patients (aged 18 years or older) with mild-to-moderate or severe C difficile infection (diarrhoea with positive glutamate dehydrogenase and toxin A or B enzyme immunoassays) were randomly assigned (1: 1) with a randomisation list stratified by centre and C difficile infection episode type (block size of four), and allocation was masked to investigators and participants. Patients received either oral cadazolid 250 mg twice daily with vancomycin-matching placebo capsule four times daily or oral vancomycin 125 mg four times a day with cadazolid-matching placebo suspension twice daily for 10 days, with 30 days of follow-up. The primary efficacy outcome was non-inferiority (margin -10%) of cadazolid versus vancomycin for clinical cure in the modified intention-to-treat and per-protocol populations. Clinical cure was defined as resolution of diarrhoea with no additional treatment for C difficile infection. These trials are registered with ClinicalTrials.gov, numbers NCT01987895 (IMPACT 1) and NCT01983683 (IMPACT 2). Findings Between March 28, 2014, and March 24, 2017, for IMPACT 1, and Dec 13, 2013, and May 2, 2017, for IMPACT 2, 1263 participants were randomly assigned to receive cadazolid (306 in IMPACT 1 and 298 in IMPACT 2) or vancomycin (326 in IMPACT 1 and 311 in IMPACT 2). In the modified intention-to-treat population in IMPACT 1, 253 (84%) of 302 had clinical cure in the cadazolid group versus 271 (85%) of 318 in the vancomycin group. In IMPACT 2, 235 (81%) of 290 versus 258 (86%) of 301 had clinical cure. In the per-protocol population, 247 (88%) of 282 versus 264 (92%) of 288 had clinical cure in IMPACT 1 and 214 (87%) of 247 versus 237 (92%) of 259 in IMPACT 2. Non-inferiority for clinical cure to vancomycin was shown in IMPACT 1 but not in IMPACT 2 (IMPACT 1 treatment difference: -1.4 [95% CI -7.2 to 4.3] for modified intention to treat and -4.1 [-9.2 to 1.0] for per protocol; IMPACT 2: -4.7 [-10.7 to 1.3] for modified intention to treat and -4.9 [-10.4 to 0.6] for per protocol). The safety and tolerability profiles of the two antibiotics were similar. Interpretation Cadazolid was safe and well tolerated but did not achieve its primary endpoint of non-inferiority to vancomycin for clinical cure in one of two phase 3 C difficile infection trials. Therefore, further commercial development of cadazolid for C difficile infection is unlikely. Copyright (c) 2019 Elsevier Ltd. All rights reserved.