Skip to main content

Foundational Trials

Bass

Image: Shutterstock / ChaNaWiT

Publication

  • Title: Rifaximin Treatment in Hepatic Encephalopathy
  • Acronym: None
  • Year: 2010
  • Journal published in: New England Journal of Medicine
  • Citation: Bass NM, Mullen KD, Sanyal AJ, et al. Rifaximin treatment in hepatic encephalopathy. N Engl J Med. 2010;362(12):1071-1081.

Context & Rationale

  • Background
    • Overt hepatic encephalopathy (HE) is a recurrent neuropsychiatric complication of cirrhosis with major impacts on hospitalisation, function, and survival.
    • After an initial episode, recurrence risk is high, and “secondary prophylaxis” is a routine goal in chronic liver disease management.
    • Nonabsorbable disaccharides (particularly lactulose) were widely used as first-line prophylaxis, but tolerability (diarrhoea, bloating) and adherence are variable, and the evidentiary base for long-term prevention was actively debated around the time of this trial.
    • Systemic antibiotics (e.g., neomycin, metronidazole) can reduce HE episodes but are limited by systemic toxicity and antimicrobial stewardship concerns.
    • Rifaximin is a minimally absorbed, gut-selective antibiotic; before this study, robust placebo-controlled evidence for preventing recurrent overt HE and reducing HE-related admissions was limited.
  • Research Question/Hypothesis
    • In adults with cirrhosis and a recent history of recurrent overt HE who are in remission, rifaximin 550 mg twice daily (added to usual care) reduces the risk of “breakthrough” overt HE compared with placebo over 6 months.
    • Key corollary hypothesis: rifaximin reduces HE-related hospitalisations.
  • Why This Matters
    • Clinically meaningful prevention would reduce recurrent neurocognitive deterioration and the need for HE-driven admissions, with downstream benefits for patients, carers, and health systems.
    • A positive, methodologically robust, multicentre placebo-controlled trial could justify widespread adoption of rifaximin for secondary prophylaxis and inform guideline recommendations.

Design & Methods

  • Research Question: Among adults with cirrhosis and recurrent overt HE in remission, does rifaximin 550 mg twice daily, compared with placebo (with usual care permitted), prolong time to first breakthrough overt HE over 6 months and reduce HE-related hospitalisations?
  • Study Type: Randomised, double-blind, placebo-controlled, parallel-group, multicentre phase 3 trial; stratified by geographic region; outpatient setting (70 sites across USA, Canada, and Russia).
  • Population:
    • Setting: Outpatients with cirrhosis; enrolment 2005–2008 across USA (n=205), Canada (n=14), and Russia (n=80).
    • Inclusion criteria: Age ≥18 years; cirrhosis; ≥2 episodes of overt episodic HE (Conn score ≥2) within the previous 6 months; remission at enrolment (Conn score 0–1); MELD score ≤25.
    • Key exclusions (selected): expected liver transplantation within 1 month; precipitating condition within the previous 3 months (including gastrointestinal haemorrhage and portosystemic shunt); chronic renal insufficiency with serum creatinine >2.0 mg/dL; respiratory insufficiency; haemoglobin <8 g/dL; serum sodium <125 mmol/L; serum potassium <2.5 mmol/L; serum calcium >10 mg/dL; intercurrent infection or active spontaneous bacterial peritonitis.
    • Episode-count nuance: prior HE episodes precipitated by gastrointestinal haemorrhage requiring ≥2 units transfusion, use of medications, haemodialysis, or central nervous system injury were not counted toward the “≥2 episodes” inclusion criterion.
  • Intervention:
    • Rifaximin 550 mg orally twice daily for up to 6 months (168 days).
    • Study drug discontinued at the time of the first breakthrough episode of overt HE (or earlier for adverse events/withdrawal/other protocol reasons).
    • Usual care permitted, including lactulose; several potentially confounding therapies were prohibited (including other antibiotic therapy and other listed agents used for HE management).
  • Comparison:
    • Matching placebo orally twice daily for up to 6 months (168 days), with the same discontinuation rules and background care approach (including permitted lactulose use).
  • Blinding: Double-blind (participants and investigators) with placebo control; implications: reduced performance and ascertainment bias, though the primary endpoint incorporated clinician-rated HE scales.
  • Statistics: Power calculation: 200 patients (100 per group) were estimated to provide >80% power to detect a reduction in breakthrough HE from 70% (placebo) to 50% (rifaximin) using a two-sided α=0.05. Primary analysis: intention-to-treat; time-to-event analyses used Kaplan–Meier methods and Cox proportional-hazards modelling with stratification by geographic region.
  • Follow-Up Period: 6 months (to day 168) with scheduled assessments (days 7 and 14, then every 2 weeks); patients who withdrew early for reasons other than HE were contacted at 6 months to ascertain outcomes where feasible.

Key Results

This trial was not stopped early. Treatment and follow-up proceeded to the planned 6-month time horizon.

Outcome Rifaximin (n=140) Placebo (n=159) Effect p value / 95% CI Notes
Breakthrough overt HE episode (primary) 31/140 (22.1%) 73/159 (45.9%) HR 0.42 95% CI 0.28 to 0.64; P<0.001 Breakthrough defined by worsening Conn score/asterixis criteria; study drug discontinued at first breakthrough; number needed to treat reported as 4 for 6 months.
HE-related hospitalisation 19/140 (13.6%) 36/159 (22.6%) HR 0.50 95% CI 0.29 to 0.87; P=0.01 Key secondary endpoint; number needed to treat reported as 9 for 6 months.
Discontinued study drug (any reason) 52/140 (37.1%) 93/159 (58.5%) Not reported Not reported Most discontinuations were driven by breakthrough HE (28/140 vs 69/159).
All-cause deaths during study 9/140 (6.4%) 11/159 (6.9%) Not reported Not reported Deaths adjudicated as related to progression of underlying liver disease in most cases.
Any adverse event 112/140 (80.0%) 127/159 (79.9%) Not reported Not reported Overall adverse event frequency was similar between groups.
Clostridioides difficile infection 2/140 (1.4%) 0/159 (0.0%) Not reported Not reported Both events occurred in patients with multiple pre-existing risk factors; study drug was continued with successful treatment.
  • Rifaximin reduced the risk of breakthrough overt HE over 6 months (22.1% vs 45.9%; HR 0.42; 95% CI 0.28 to 0.64; P<0.001).
  • Rifaximin reduced HE-related hospitalisation (13.6% vs 22.6%; HR 0.50; 95% CI 0.29 to 0.87; P=0.01).
  • Subgroup analyses were presented graphically; treatment effect was directionally consistent, but some smaller strata had non-significant within-stratum testing (e.g., MELD 19–24 and those not receiving lactulose at baseline).

Internal Validity

  • Randomisation and Allocation: Participants were randomised 1:1 with stratification by geographic region; details of sequence generation and allocation concealment were not fully specified in the manuscript.
  • Dropout or exclusions (post-randomisation): Study drug discontinuation occurred in 52/140 (37.1%) in the rifaximin group vs 93/159 (58.5%) in placebo; discontinuation due to breakthrough HE was 28/140 (20.0%) vs 69/159 (43.4%); discontinuations due to non-breakthrough reasons were 24/140 (17.1%) vs 24/159 (15.1%).
  • Performance/Detection Bias: Double-blinding reduces co-intervention and ascertainment bias; however, the primary endpoint incorporated clinician-rated Conn score and asterixis grading, which may be vulnerable to inter-rater variation across 70 sites.
  • Protocol Adherence: Mean study drug exposure was 130.3 ± 56.5 days (rifaximin) vs 105.7 ± 62.7 days (placebo); reported mean compliance was 84.3% ± 21.5 vs 84.9% ± 22.9.
  • Baseline Characteristics: Groups were broadly similar in liver disease severity strata (MELD categories and Child–Pugh distribution) and HE history; notable imbalances included male sex (53.6% vs 67.3%) and folic acid use (36.4% vs 22.0%).
  • Heterogeneity: The trial was international (USA/Canada/Russia) with stratification by region; the overall effect was favourable, but smaller strata (e.g., Russia and higher MELD categories) reduce precision.
  • Timing: Treatment began in clinical remission and assessments were frequent early (days 7 and 14) and then fortnightly, supporting timely detection of breakthrough episodes.
  • Dose: Rifaximin 550 mg twice daily is a fixed, pragmatic regimen; dose adjustment strategies were not a trial feature, so dose–response inference is not possible.
  • Separation of the Variable of Interest: Breakthrough overt HE occurred in 31/140 (22.1%) vs 73/159 (45.9%); HE-related hospitalisation in 19/140 (13.6%) vs 36/159 (22.6%).
  • Adjunctive therapy use: Background lactulose use at baseline was high and similar (91.4% vs 91.2%); mean daily lactulose use was reported as 3.14 ± 2.096 vs 3.51 ± 2.592 cups/day. 1
  • Outcome Assessment: Outcomes were clinically defined (overt HE breakthrough and HE-related hospitalisation) with prespecified criteria; HE-related hospitalisation is relatively objective, while breakthrough HE scoring is more subjective.
  • Statistical Rigor: Intention-to-treat time-to-event methods matched the design; planned sample size was exceeded, and effect estimates were reported with hazard ratios and confidence intervals.

Conclusion on Internal Validity: Overall, internal validity appears moderate-to-strong: randomised, double-blind, placebo-controlled design and clinically meaningful outcomes support causal inference, though allocation concealment details were limited and the primary endpoint required clinician-rated assessments across many sites.

External Validity

  • Population Representativeness: The cohort reflects a common outpatient phenotype: cirrhosis with recurrent overt HE but not extremely advanced disease (MELD ≤25) and clinically stable enough for enrolment and follow-up visits.
  • Important exclusions: Recent precipitating conditions (e.g., gastrointestinal haemorrhage and portosystemic shunt within 3 months), severe renal dysfunction, major electrolyte disturbances, and active infection/spontaneous bacterial peritonitis limit applicability to patients with acutely precipitated HE or multi-organ failure.
  • Applicability to usual care: High baseline lactulose use (>90%) mirrors many real-world secondary prophylaxis strategies, supporting applicability to “add-on rifaximin” practice rather than rifaximin monotherapy.
  • Health system considerations: Resource implications (drug cost and access) and antimicrobial stewardship considerations may influence uptake across systems.

Conclusion on External Validity: Generalisability is moderate: results apply best to ambulatory cirrhosis patients with recurrent overt HE in remission (MELD ≤25), particularly those already treated with lactulose; applicability is more limited in acutely precipitated HE, severe organ dysfunction, or different care contexts.

Strengths & Limitations

  • Strengths:
    • Large, multicentre, international, double-blind, placebo-controlled design for a chronic liver complication with substantial practice variability.
    • Clinically relevant endpoints (overt HE recurrence and HE-related hospitalisation) that matter to patients and services.
    • Time-to-event methodology with clear effect estimates (hazard ratios and confidence intervals).
    • High adherence and pragmatic background care (predominant lactulose use), supporting real-world interpretability for add-on therapy.
  • Limitations:
    • No lactulose-free arm and no rifaximin-vs-lactulose head-to-head comparison, limiting inference about rifaximin monotherapy and relative efficacy.
    • Follow-up limited to 6 months for primary efficacy; long-term resistance patterns and prolonged safety are not resolved by this study alone.
    • Primary endpoint depended on clinician-rated HE scales across many sites, introducing potential measurement variability despite blinding.
    • Exclusions (e.g., recent bleeding/shunt, active infection, severe renal dysfunction) reduce applicability to higher-acuity settings, including ICU presentations.
    • Industry sponsorship and involvement in study conduct create perceived conflict-of-interest risk even when methods are rigorous.

Interpretation & Why It Matters

  • Add-on secondary prophylaxis
    In a high-risk recurrent HE population (≥2 recent episodes), rifaximin 550 mg twice daily reduced breakthrough overt HE despite near-universal background lactulose use, supporting an additive benefit in routine prophylaxis strategies.
  • Service-level outcomes
    The reduction in HE-related hospitalisation complements the clinical endpoint and is a key mechanistic bridge from symptom prevention to health-system impact.
  • Benefit–harm framing
    Overall adverse event rates were similar; rare C. difficile events occurred only with rifaximin, reinforcing the need for antimicrobial risk assessment and monitoring in prolonged prophylaxis.

Controversies & Subsequent Evidence

  • Trial design and background therapy: Absence of a lactulose-free arm limited inference about rifaximin monotherapy and the degree to which background lactulose contributed to outcomes; the authors framed lactulose withdrawal as unethical in this population. 2
  • Mechanistic framing and antimicrobial concerns: Contemporary editorial commentary emphasised the gut microbiota axis in HE while highlighting practical concerns (cost, long-term antimicrobial exposure, and generalisability beyond the enrolled phenotype). 3
  • Comparator evidence base: Around the time of publication, randomised evidence for lactulose in secondary prophylaxis existed but was limited in size and design features, contributing to ongoing debate about the optimal “baseline” standard of care in prevention trials. 4
  • Guideline incorporation: Post-trial practice guidelines incorporated rifaximin as add-on therapy for secondary prophylaxis (typically after recurrent episodes), aligning policy with the add-on design and outcomes of the Bass trial. 5
  • More recent guideline synthesis: Subsequent European guidance continued to recommend rifaximin for prevention of recurrent overt HE in appropriate patients, with attention to precipitating factor management and patient selection. 6
  • Extension to other prophylaxis contexts: Later randomised evidence in specific subpopulations (e.g., post-TIPS prophylaxis) expanded the prophylaxis question beyond “secondary prevention after an HE episode”, reinforcing that effect size and certainty may be context-dependent. 7
  • Synthesis of broader evidence: Systematic reviews and meta-analyses generally support rifaximin (often in combination with lactulose) for reducing recurrent HE and improving clinical outcomes, while highlighting heterogeneity in baseline therapy, trial design, and outcome definitions across the literature. 8

Summary

  • In outpatients with cirrhosis and recurrent overt HE in remission (MELD ≤25), rifaximin 550 mg twice daily reduced breakthrough overt HE over 6 months versus placebo (22.1% vs 45.9%; HR 0.42; 95% CI 0.28 to 0.64; P<0.001).
  • Rifaximin reduced HE-related hospitalisations (13.6% vs 22.6%; HR 0.50; 95% CI 0.29 to 0.87; P=0.01).
  • Most participants in both arms received lactulose at baseline (>90%), positioning the trial as evidence for add-on rifaximin rather than monotherapy.
  • Overall adverse event rates were similar; C. difficile infection was uncommon but occurred only with rifaximin (2 vs 0), and mortality was similar (9 vs 11).
  • This trial materially shifted secondary prophylaxis practice and strongly influenced subsequent guideline recommendations for rifaximin as add-on therapy in recurrent overt HE.

Further Reading

Other Trials

Systematic Review & Meta Analysis

Observational Studies

Guidelines

Notes

  • Where volume/issue/page details were not consistently available from indexed sources during preparation, DOI-linked journal landing pages provide the authoritative bibliographic record.

Overall Takeaway

The Bass trial provided the first large, double-blind, placebo-controlled evidence that rifaximin meaningfully reduces recurrent overt HE and HE-related hospitalisations over 6 months in ambulatory cirrhosis patients with recent recurrent HE—predominantly as add-on therapy to lactulose. Its clinically anchored endpoints, sizeable effect estimates, and broadly acceptable short-term safety profile established rifaximin as a core component of secondary prophylaxis and shaped subsequent guideline recommendations.

Overall Summary

  • Rifaximin 550 mg twice daily reduced breakthrough overt HE and HE-related hospitalisation over 6 months in recurrent HE patients in remission, largely on background lactulose.
  • The trial’s add-on design and outcomes underpinned later guideline endorsement of rifaximin for secondary prophylaxis after recurrent episodes.

Bibliography