The RIVET RCT: Rifamycin SV MMX improves muscle mass, physical function, and ammonia in cirrhosis and minimal encephalopathy

Abstract

Background: Minimal hepatic encephalopathy (MHE) negatively affects the prognosis of cirrhosis, but treatment is not standard. Rifamycin SV MMX (RiVM) is a nonabsorbable rifampin derivative with colonic action.

Methods: In a phase 2 placebo-controlled, double-blind randomized clinical trial patients with MHE were randomized to RiVM or placebo for 30 days with a 7-day follow-up. The primary endpoint was a change in stool cirrhosis dysbiosis ratio. Gut-brain (cognition, stool/salivary microbiome, ammonia, brain magnetic resonance spectroscopy), inflammation (stool calprotectin/serum cytokines), patient-reported outcomes (sickness impact profile: total/physical/psychosocial, high = worse), and sarcopenia (handgrip, bioelectric impedance) were secondary. Between/within groups and delta (post-pre) comparisons were performed.

Results: Thirty patients (15/group) were randomized and completed the study without safety concerns. While cirrhosis dysbiosis ratio was statistically similar on repeated measures ANOVA (95% CI: -0.70 to 3.5), ammonia significantly reduced (95% CI: 4.4-29.6) in RiVM with changes in stool microbial α/β-diversity. MHE status was unchanged but only serial dotting (which tests motor strength) improved in RiVM-assigned patients. Delta physical sickness impact profile (95% CI: 0.33 = 8.5), lean mass (95% CI: -3.3 to -0.9), and handgrip strength (95% CI: -8.1 to -1.0) improved in RiVM versus placebo. Stool short-chain fatty acids (propionate, acetate, and butyrate) increased post-RiVM. Serum, urine, and stool bile acid profile changed to nontoxic bile acids (higher hyocholate/ursodeoxycholate and lower deoxycholate/lithocholate) post-RiVM. Serum IL-1β and stool calprotectin decreased while brain magnetic resonance spectroscopy showed higher glutathione concentrations in RiVM.

Conclusions: RiVM is well tolerated in patients with MHE with changes in stool microbial composition and function, ammonia, inflammation, brain oxidative stress, and sarcopenia-related parameters without improvement in cognition. RiVM modulates the gut-brain axis and gut-muscle axis in cirrhosis.

Graphical abstract

FIGURE 1

(A) Study design. (B) CONSORT diagram. Abbreviations: HRQOL, health-related quality of life; OHE, overt hepatic encephalopathy; PHES, psychometric hepatic encephalopathy score.

FIGURE 2

Microbiome alpha and beta diversity changes. (A–C) Individual plots of Shannon diversity and delta within the rifamycin group comparing baseline to end (day 30). (D) PERMANOVA comparison of beta diversity changes between rifamycin end versus baseline. (E) PERMANOVA comparison of beta diversity changes between rifamycin end versus placebo end. Abbreviation: PERMANOVA, permutational analysis of variance.

FIGURE 3

Changes in CDR and individual microbiota using linear discriminant function effect size. (A) Linear discriminant function effect size comparison between rifamycin end versus baseline, linear discriminant function effect size comparison between rifamycin and placebo end (B), individual plots of CDR and delta within the rifamycin group comparing baseline to end (day 30) (C–E). Abbreviation: CDR, cirrhosis dysbiosis ratio.

FIGURE 4

Changes in ammonia, CPK, butyrate, and sarcopenia. (A) Repeated measures ANOVA change between venous ammonia from baseline to mid (day 15) and end (day 30), which was a significantly greater reduction in the rifamycin group. (B) Repeated measures ANOVA in CPK levels that did not change significantly throughout. (C) Repeated measures ANOVA change of stool butyrate, which was lower in the rifamycin group at baseline and came to the placebo group level after the end of the trial. (D) Delta dry lean mass from baseline to end (day 30) was higher in the rifamycin group. (E) Delta handgrip change from baseline to end (day 30) was higher in the rifamycin group. Abbreviation: CPK, creatine phosphokinase.