Résumé
Background & aims: Evidence for primary prophylaxis of spontaneous bacterial peritonitis (SBP) is weak and the selection of quinolone-resistant bacteria is currently of concern. Here we present results from a randomized, double-blind, placebo (PBO)-controlled trial to assess whether rifaximin (RFX) has a beneficial effect on 12-month survival in patients with severe cirrhosis and ascites.Methods: In this trial conducted at 17 French centers, patients with severe cirrhosis and grade 2 or 3 ascites and ascites protein level <15 g/L were randomized 1:1 to receive RFX 550 mg or PBO BID for 12 months, as primary prophylaxis for SBP. The primary endpoint was 12-month survival. Secondary endpoints were 3- and 6-month survival, incidence of complications of cirrhosis, and safety of RFX.Results: Between 2018 and 2022, 1957 cirrhotic patients with ascites were screened, 159 were randomized, and 152 (80/72 PBO/RFX) were analyzed in the modified intention-to-treat population. RFX did not improve 12-month (PBO/RFX : 68.1% [95% CI 56.2-78.7] / 56.6% [95% CI 43.5-67.8], p=0.74), 6-month (71.1% [95% CI 59.5-80.0] / 76.4% [95% CI 64.3-84.8]) or 3-month survivals (75.4% [95% CI 64.1-83.5]/82.6% [95% CI 71.4-89.7]), or the incidence of liver complications (SBP, encephalopathy, gastrointestinal bleeding or hepatorenal syndrome). In the per-protocol population (127 patients adherent to the study drug), a lower 12-month cumulative incidence of liver-related events in the RFX group was observed. RFX was well tolerated throughout the study.Conclusions: RFX had no beneficial effect in terms of 12-month survival or incidence of complications of liver cirrhosis in patients with severe cirrhosis and low ascitic fluid protein levels. Improving patient adherence could reduce liver complications of cirrhosis.Impact and implications: Selective gut decontamination using norfloxacin is the standard of care for secondary prophylaxis of spontaneous bacterial peritonitis (SBP). Evidence for primary prophylaxis of SBP is weaker, and fluoroquinolones have been associated with an increased risk of antimicrobial resistance (AMR). Rifaximin, a well-tolerated broad-spectrum antibiotic with a lower risk of AMR emergence, may be an alternative to norfloxacin. Our trial did not demonstrate an improvement in survival or liver complications (SBP, gastrointestinal bleeding, hepatic encephalopathy or hepatorenal syndrome) at 12 months with rifaximin as primary prophylaxis for SBP versus placebo. However, in the subgroup of patients who adhered to rifaximin, liver complications decreased. Our study highlights the importance of study drug adherence in clinical trials to adequately assess outcomes.