Cystic Fibrosis Microbiome-directed Antibiotic Therapy Trial in Exacerbations Results Stratified (CFMATTERS): results of a multicentre randomised controlled trial.

Background: This study explores the effectiveness and safety of microbiome-directed antimicrobial therapy versus usual antimicrobial therapy in adult cystic fibrosis pulmonary exacerbations.

Methods: A multicentre two-arm parallel randomised control trial conducted across Europe/North-America enrolled 223 participants (January 2015 to August 2017). All participants were chronically colonised with Pseudomonas aeruginosa and were randomised 1:1 into two study arms. The "usual therapy" group received 2 weeks of intravenous ceftazidime 3 g thrice daily (for allergies: aztreonam 2 g thrice daily) and tobramycin 5-10 mg·kg^-1 once daily. The "microbiome-directed" group received the same usual therapy plus an additional antibiotic with greatest presumed activity against the second, third and fourth most abundant genera present in the sputum microbiome, selected by a consensus expert treatment panel. The primary outcome was change in percentage of predicted forced expiratory volume in 1 s (ppFEV₁) at 14 days post initiation of antibiotics. Secondary outcomes examined ppFEV₁ at 7 days, 28 days and 3 months; time to next exacerbation; symptom burden at 7 days; health-related quality of life (HRQoL) at 28 days; and number of exacerbations and i.v. antibiotic days at 12 months.

Results: 149 participants had an eligible exacerbation (usual therapy n=83, microbiome-directed therapy n=66). There was no difference between the groups for ppFEV₁ at day 14 (-1.1%, 95% CI -3.9-1.7%; p=0.46), or ppFEV₁ measured at other time points, or for time to next exacerbation (microbiome-directed versus usual therapy hazard ratio 0.91, 95% CI 0.60-1.38; p=0.66). The microbiome-directed group trended to have more i.v. days (median 42 days versus 28 days; p=0.08) and more subsequent exacerbations (median three versus two; p=0.044) the following year. There were no appreciable differences in symptom burden; however, HRQoL subscores were consistently worse in the microbiome-directed group (-4.3 points versus usual therapy, 95% CI -8.3--0.3 points; p=0.033).

Conclusion: The addition of a third antibiotic based on sputum microbiome sequencing analysis did not result in improved clinical outcomes.