Scolaris Content Display Scolaris Content Display

Cochrane Database of Systematic Reviews Protocol - Intervention

Aminoglycosides and metronidazole for people with cirrhosis and hepatic encephalopathy

Collapse all Expand all

Abstract

This is a protocol for a Cochrane Review (Intervention). The objectives are as follows:

To assess the benefits and harms of aminoglycosides and metronidazole versus placebo, no intervention, or other pharmacological interventions on mortality, morbidity, and adverse events for people with cirrhosis and hepatic encephalopathy.

Background

Description of the condition

Cirrhosis is a chronic disorder of the liver associated with a variety of conditions such as excessive alcohol intake, infection with hepatitis B and hepatitis C, and obesity. In susceptible people, the damage caused to the liver results in the development of fibrosis (scar tissue) which replaces normal liver tissue, intermixed with areas of regeneration (nodules) which form as the liver attempts to repair itself.

Hepatic encephalopathy is the term used to describe the spectrum of neuropsychiatric complication which can arise in people with cirrhosis (Ferenci 2002; Dhiman 2010; Vilstrup 2014) and defined, most recently, as brain dysfunction caused by liver insufficiency or portal systemic shunting of blood, or both (Vilstrup 2014).

The spectrum of changes in brain functioning associated with hepatic encephalopathy ranges from the clinically 'indiscernible' to the clinically 'obvious'. The term 'minimal' hepatic encephalopathy is used to describe people with cirrhosis who are 'clinically normal' but in whom psychometric testing shows impairment of attention, decision making, and memory functions, while electrical recordings show slowing of their brain waves. There is no standard approach to the diagnosis of minimal hepatic encephalopathy; various psychometric and neurophysiological tools may be used, depending on local expertise. The term 'overt' hepatic encephalopathy is used to describe people with cirrhosis who show clinical evidence of neuropsychiatric abnormalities including alterations in intellect, behaviour, motor function, and consciousness (Weissenborn 1998; Ferenci 2002). The clinical changes may arise over a period of hours or days in people who have previously been stable. This so‐called 'episodic' hepatic encephalopathy may be precipitated by a number of diverse events such as gastrointestinal bleeding, infection, or constipation, although in 50% of instances no obvious cause can be identified. The diagnosis of overt hepatic encephalopathy is a clinical one, and scales such as the West Haven Criteria and Glasgow Coma Scale may be used to determine severity. Between episodes of hepatic encephalopathy, people may return to normal, although some degree of impairment may remain (Bajaj 2010). Less frequently, people present with persistent neuropsychiatric abnormalities that remain stable over time; many have extensive spontaneous portal‐systemic shunting of blood which develops as a complication of their cirrhosis or else a surgically created or transjugular intrahepatic portosystemic shunt undertaken to treat this problem.

The exact reason why people with cirrhosis develop hepatic encephalopathy is unknown, but the accumulation of toxic material, such as ammonia, which is produced in the gastrointestinal tract is thought to play a key role. These toxins are normally detoxified in the liver but this process fails when liver function deteriorates and when blood is shunted around the liver following the development of portal‐systemic shunting. The unprocessed toxins then enter the systemic circulation and impinge on the brain.

The prevalence of overt hepatic encephalopathy varies in relation to the stage of the underlying liver disease (Bajaj 2011). Thus, at the time of diagnosis with cirrhosis, the prevalence is of the order of 10% to 14% (Saunders 1981; Jepsen 2010), while in people with more longstanding cirrhosis, it is nearer 20% (D'Amico 1986; de Jongh 1992; Zipprich 2012). Overall, it is likely that 40% of people with cirrhosis will develop overt hepatic encephalopathy at some point during their disease evolution (Amodio 2001). The prevalence of minimal hepatic encephalopathy varies from 22% to 75% depending on the diagnostic paradigms used (Montagnese 2004); approximately 50% in people with previous overt hepatic encephalopathy retain features compatible with a diagnosis of minimal hepatic encephalopathy (Sharma 2010; Lauridsen 2011).

The presence of hepatic encephalopathy, whether minimal or overt, has a considerable impact on the execution of complex tasks, such as driving (Schomerus 1981; Bajaj 2009); health‐related quality of life (Groeneweg 1998; Montagnese 2009); patient safety (Román 2011); neurocognitive function post liver transplantation (Sotil 2009); and, ultimately, survival (D'Amico 2006; Stewart 2007; Wong 2015). It also places a considerable burden on healthcare resources (Rakoski 2012).

Description of the intervention

The non‐absorbable antibiotic rifaximin is recommended for the prevention of hepatic encephalopathy (EASL/AASLD guideline 2014a; EASL/AASLD guideline 2014b). Clinicians and researchers have previously evaluated other antibiotics, including aminoglycosides and metronidazole. Aminoglycosides (such as neomycin and paromomycin) are bactericidal antibiotics, which inhibit bacterial protein synthesis (Borovinskaya 2007). Metronidazole is also a bactericidal antibiotic, which inhibits bacterial DNA synthesis (Löfmark 2010). Aminoglycosides are particularly active against gram‐negative aerobes (Magnet 2005), while metronidazole is active against both gram‐positive and gram‐negative anaerobes (Löfmark 2010). The interventions may be administered enterally (orally or via an enteral tube) in the form of tablets or a solution. Rectal and intravenous administration is also possible. Aminoglycosides are normally poorly absorbed from the gastrointestinal tract. However, absorption may be increased in cirrhosis as the permeability of the gut increases, leading to drug‐induced ototoxicity or nephrotoxicity (Patidar 2014). Metronidazole is normally well absorbed from the gastrointestinal tract and can cause neurotoxicity in people with cirrhosis, who may have impaired clearance of the drug. As rifaximin is a relatively expensive antibiotic (Huang 2007), aminoglycosides and metronidazole could potentially be more cost‐effective interventions for the prevention and treatment of hepatic encephalopathy.

How the intervention might work

Although the exact pathogenesis of hepatic encephalopathy is unknown, hyperammonaemia is known to play a key role (Butterworth 2004). In people with cirrhosis, the liver is unable to clear the high levels of gut‐derived ammonia. As a result, organs such as the brain rely on glutamine synthesis as an alternative detoxification pathway. In the brain, ammonia is handled mainly by astrocytes, resulting in increased glutamine synthesis. Increased glutamine synthesis triggers a series of cellular events that results in astrocyte swelling, low‐grade oedema, and eventually brain dysfunction (Häussinger 2000; Albrecht 2006; Rai 2015). Other proposed effects of ammonia in the brain include blood‐brain barrier dysfunction, altered inhibitory and excitatory neurotransmission, and altered cerebral energy metabolism (Butterworth 2000). The main sources of ammonia include nitrogenous products in the diet, bacterial metabolism of urea and proteins in the colon, and deamination of glutamine in the small intestine. Antibiotics lower ammonia levels by eliminating urease‐producing gut bacteria and correcting for small intestinal bacterial overgrowth, which is frequently observed in people with cirrhosis (Yang 1998; Bauer 2001; Gunnarsdottir 2003; Pande 2009). The aminoglycoside neomycin is also a known glutaminase inhibitor (Hawkins 1994). In addition to their ammonia‐lowering effects, antibiotics may protect against hepatic encephalopathy by reducing the production of other bacteria‐derived neurotoxins, such as phenols and mercaptans, and by reducing gut bacteria translocation into the bloodstream (Jalan 2010). Reducing gut bacteria translocation into the bloodstream may reduce endotoxaemia and systemic inflammation, which are implicated in the development of hepatic encephalopathy (Patidar 2014).

Why it is important to do this review

From 1957 to 1966, the aminoglycoside neomycin was the recommended treatment for people with hepatic encephalopathy (Patidar 2014). After 1966, clinicians primarily used non‐absorbable disaccharides based on randomised clinical trials and clinical experience (Elkington 1969; Germain 1973; Heredia 1987; Heredia 1988; Grandi 1991; Horsmans 1997). A recent systematic review found that non‐absorbable disaccharides are associated with beneficial effects on mortality and morbidity (Gluud 2016a; Gluud 2016b). At present, we have limited data on the effectiveness of aminoglycosides and other antibiotics such as metronidazole and vancomycin, possibly because clinicians introduced and subsequently abandoned the intervention before it could be assessed using current norms in evidence‐based medicine (Patidar 2014). Moreover, concerns about organ toxicity and antibiotic resistance, particularly vancomycin resistance, have yet to be properly addressed (Leise 2014). However, antibiotics are still commonly used, in several countries, to treat hepatic encephalopathy, and the European Association for the Study of the Liver (EASL)/American Association for the Study of Liver Diseases (AASLD) guidelines report favourably on both neomycin and metronidazole (EASL/AASLD guideline 2014a; EASL/AASLD guideline 2014b).

Three trials found that neomycin may be as effective as lactulose in improving signs of hepatic encephalopathy (Conn 1977; Atterbury 1978; Orlandi 1980), and one study found that metronidazole may be as effective as neomycin (Morgan 1982). Two subsequent trials found that the aminoglycoside paromomycin may be as effective as the more widely used antibiotic rifaximin (Testa 1985; Parini 1992). In light of current practice guidelines and the available evidence, a thorough assessment of the use and safety of antibiotics in hepatic encephalopathy is warranted. One systematic review published in 2004 found no difference between antibiotics and non‐absorbable disaccharides (Als‐Nielsen 2004). However, the review did not include randomised clinical trials comparing antibiotics versus placebo or no interventions. In addition, methodological issues and lack of statistical power weakened the conclusions. This review will evaluate the beneficial and harmful effects of aminoglycosides and metronidazole using the evidence available.

Objectives

To assess the benefits and harms of aminoglycosides and metronidazole versus placebo, no intervention, or other pharmacological interventions on mortality, morbidity, and adverse events for people with cirrhosis and hepatic encephalopathy.

Methods

Criteria for considering studies for this review

Types of studies

We will include randomised clinical trials irrespective of blinding, publication status, or language in our analyses of benefits and harms. Due to the risk of bias, we will only include quasi‐randomised studies and observational studies in analyses of harms. We will not specifically search for observational studies for inclusion in this review (which is a known limitation).

Types of participants

We will include people with cirrhosis, regardless of sex, age, and aetiology of the underlying liver disease. We will include people with overt or minimal hepatic encephalopathy and people at risk of developing hepatic encephalopathy (primary or secondary prevention). If we identify trials including subsets of relevant participants, we will only analyse outcomes for participants who fulfil our inclusion criteria.

Types of interventions

The intervention comparisons will be:

  • Aminoglycosides versus placebo, no intervention, or other pharmacological interventions (erythromycin, rifaximin, or non‐absorbable disaccharides);

  • Metronidazole versus placebo, no intervention, or other pharmacological interventions (erythromycin, rifaximin, or non‐absorbable disaccharides).

We will include randomised clinical trials irrespective of the dose, treatment duration, and mode of administration, and allow co‐interventions if they are administered equally in each allocation arm. We will include separate pair‐wise comparisons of the agents under study, should we identify multi‐arm trials (e.g., trials comparing an aminoglycoside versus metronidazole versus placebo).A separate review will evaluate the antibiotic rifaximin versus placebo, no intervention or non‐absorbable disaccharides for people with cirrhosis and hepatic encephalopathy (Kimer 2015). Therefore, we will not include these intervention comparisons in the present review.

Types of outcome measures

We will evaluate outcomes at the maximum duration of follow‐up.

Primary outcomes

  • All‐cause mortality.

  • Hepatic encephalopathy defined as development of hepatic encephalopathy or lack of improved manifestations of hepatic encephalopathy.

  • Serious adverse events defined as any untoward medical occurrence that leads to death, is life threatening or requires hospitalisation or prolongation of hospitalisation (ICH GCP 1997). We will analyse adverse events as a composite outcome (Gluud 2017).

Secondary outcomes

  • Health‐related quality of life: the overall score based on the quality‐of‐life questionnaires used in individual randomised clinical trials.

  • Non‐serious adverse events: all events that do not fulfil the criteria for a serious adverse event.

Exploratory outcomes

  • Number Connection Test results.

  • Blood ammonia concentrations.

Search methods for identification of studies

Electronic searches

We will search The Cochrane Hepato‐Biliary Group (CHBG) Controlled Trials Register (Gluud 2017), Cochrane Central Register of Controlled Trials (CENTRAL) in the Cochrane Library, MEDLINE Ovid, Embase Ovid, Science Citation Index Expanded (Web of Science), and LILACS. Preliminary search strategies with the expected time spans of the searches are given in Appendix 1. We also plan to search Chinese, Japanese, and Russian databases if the CHBG is able to provide us with access to these databases.

Searching other resources

We will scan the reference lists of relevant articles, conference proceedings (from the annual meetings of the United European Gastroenterology (UEG), the American Gastroenterological Association (AGA), and the AASLD. We will also write to the principal authors of trials and the pharmaceutical companies involved in the production of the relevant antibiotics. We will search online trial registries such as ClinicalTrials.gov (clinicaltrials.gov/), the European Medicines Agency (EMA) (www.ema.europa.eu/ema/), the World Health Organization (WHO) International Clinical Trial Registry Platform (www.who.int/ictrp), and the Food and Drug Administration (FDA) for ongoing or unpublished trials (www.fda.gov). In addition, we will search Google Scholar using the terms 'hepatic encephalopathy' AND ('aminoglycosides' OR 'metronidazole' OR 'antibiotics').

Data collection and analysis

All three authors (RJ, LG, and MM) will participate in the selection of randomised clinical trials and independently extract data and assess bias as described below. We will resolve disagreements through discussion.

Selection of studies

All three authors (RJ, LG, and MM) will participate in the searches and review the search results based on screening of titles and abstracts. We will retrieve and list the full‐text articles of potentially relevant references. All authors will independently read the potentially eligible papers and participate in the final selection of randomised clinical trials. For randomised clinical trials described in more than one publication, we will select the paper with the longest duration of follow‐up as our primary reference. We will list excluded randomised clinical trials with the reasons for exclusion in the 'Characteristics of excluded studies' table. If we identify randomised clinical trials published in languages other than English, we will use professional translation services or obtain help from medical professionals who are fluent in the language of the publication.

Data extraction and management

All three authors (RJ, LG, and MM) will collect data independently using a customised data extraction form piloted on ten randomised trials evaluating interventions for hepatic encephalopathy. We will r write to the authors of included trials for information that is not included in the published reports. We will gather the following data on the included trials and list details in the 'Characteristics of included studies' table:

  • Design (cross‐over or parallel), setting (number of clinical sites and outpatient clinic or hospital department), country of origin;

  • Participants: mean age, proportion of men, alcoholic liver disease, viral hepatitis, minimal hepatic encephalopathy, overt hepatic encephalopathy, previous overt hepatic encephalopathy;

  • Interventions: type, dose, duration of therapy, mode of administration.

We will also gather data on all outcomes, including the criteria used in the assessment of hepatic encephalopathy, and bias control. We will obtain help from professional services as well as medical professionals to translate papers published in languages other than English. Medical professionals will double check the translations and double‐check our data extraction from these trials.

Assessment of risk of bias in included studies

We will assess bias control using the domains described in the CHBG module (Gluud 2017) and classify the risk of bias for separate domains as high, unclear, or low (Higgins 2011). We will assess all bias domains for each randomised clinical trial and base our primary conclusions on the results of our primary outcomes with a low risk of bias in the overall assessment.

Allocation sequence generation

  • Low risk of bias: sequence generation achieved using computer random number generation or a random number table. Drawing lots, tossing a coin, shuffling cards, or throwing dice will be adequate if performed by an independent person but not otherwise.

  • Unclear risk of bias: not described.

  • High risk of bias: sequence generation method was not random. We will only include such studies for assessment of adverse events.

Allocation concealment

  • Low risk of bias: allocation by a central and independent randomisation unit; administration of coded, identical drug containers/vials or sequentially numbered, opaque, sealed envelopes.

  • Unclear risk of bias: not described.

  • High risk of bias: allocation sequence was likely to be known to the investigators who assigned the participants. We will only include such studies for assessment of adverse events.

Blinding of participants and personnel

  • Low risk of bias: blinding of participants and personnel performed adequately using a placebo. We will define lack of blinding as not likely to affect the evaluation of mortality (Hrobjartsson 2001; Savovic 2012a; Savovic 2012b).

  • Unclear risk of bias: insufficient information to assess blinding.

  • High risk of bias: no blinding or incomplete blinding.

Blinding of outcome assessors

  • Low risk of bias: blinding of outcome assessors performed adequately using a placebo. We will define lack of blinding as not likely to affect the evaluation of mortality (Hrobjartsson 2001; Savovic 2012a; Savovic 2012b).

  • Unclear risk of bias: insufficient information to assess blinding.

  • High risk of bias: no blinding or incomplete blinding.

Incomplete outcome data

  • Low risk of bias: missing data unlikely to make treatment effects depart from plausible values. The investigators used explicit methods, such as intention‐to‐treat analyses with multiple imputations or carry‐forward analyses, to handle missing data.

  • Unclear risk of bias: insufficient information.

  • High risk of bias: results were likely to be biased due to missing data.

Selective outcome reporting

  • Low risk of bias: trial reported clinically relevant outcomes (mortality, hepatic encephalopathy, and serious adverse events). If we have access to the original trial protocol, the outcomes selected should be those called for in that protocol. If we obtain information from a trial registry (such as www.clinicaltrials.gov), we will consider that information reliable only if the investigators registered the trial before inclusion of the first participant.

  • Unclear risk of bias: not all predefined outcomes were reported fully, or it was unclear whether data on these outcomes were recorded or not.

  • High risk of bias: one or more predefined outcomes were not reported.

For‐profit bias

  • Low risk of bias: trial appears to be free of industry sponsorship or other type of for‐profit support.

  • Unclear risk of bias: insufficient information about support or sponsorship.

  • High risk of bias: trial received funding or other support from a pharmaceutical company.

Other bias

  • Low risk of bias: trial appears free of other biases including: medicinal dosing problems or follow‐up (as defined below).

  • Unclear risk of bias: trial may or may not be free of other factors that could put it at risk of bias.

  • High risk of bias: there were other factors in the trial that could put it at risk of bias such as the administration of inappropriate treatments being given to the controls (e.g. an inappropriate dose) or follow‐up (e.g. the trial included different follow‐up schedules for participants in the allocation groups).

Overall bias assessment

  • Low risk of bias: all domains were at low risk of bias using the definitions described above.

  • High risk of bias: one or more of the bias domains were of unclear or high risk of bias.

Measures of treatment effect

We will analyse dichotomous data using risk ratios (RR) and continuous outcomes using mean differences (MD), both with 95% confidence intervals (CI).

Unit of analysis issues

We will include data from the first treatment period of cross‐over trials (Higgins 2011).

Dealing with missing data

We will extract data on all randomised participants to allow intention‐to‐treat analyses. We will conduct a worst‐case scenario analysis including all participants with missing outcomes as failures (Higgins 2008), and extreme worst‐case and best‐case scenario analyses in which participants in the experimental arm are counted as failures and participants in the control arm are counted as successes and vice versa (Gluud 2017).

Assessment of heterogeneity

We will visually inspect forest plots and express heterogeneity as I² values using the following thresholds: 0% to 40% (unimportant), 40% to 60% (moderate), 60% to 80% (substantial), and greater than 80% (considerable), and include information in the 'Summary of findings' tables (Additional tables).

Assessment of reporting biases

For meta‐analyses with at least 10 randomised clinical trials, we will assess reporting biases through regression analyses (Egger 1997; Harbord 2006) and funnel plots.

Data synthesis

We will perform the analyses in Review Manager 5 (RevMan 2014), STATA (Stata 2007), and Trial Sequential Analysis (Thorlund 2011; TSA 2011).

Meta‐analysis

We will stratify randomised clinical trials based on the type of control intervention (e.g. placebo or no intervention, non‐absorbable disaccharides, and antibiotics). We plan to compare the fixed‐effect and random‐effects estimates of the intervention effect. If the estimates are similar, then we will assume that any small‐study effects have little effect on the intervention effect estimate. If the random‐effects estimate is more beneficial, we will re‐evaluate whether it is reasonable to conclude that the intervention was more effective in the smaller studies. If the larger studies tend to be those conducted with greater methodological rigour, or conducted in circumstances more typical of the use of the intervention in practice, then we will report the results of meta‐analyses restricted to the larger, more rigorous studies. Based on the expected clinical heterogeneity, we expect that a number of analyses will display statistical between‐trial heterogeneity (I² greater than 0%). For random‐effects models, precision will decrease with increasing heterogeneity and CIs will widen correspondingly. Therefore, we expect that the random‐effects model will give the most conservative (and a more correct) estimate of the intervention effect. Accordingly, we plan to report the results of our analyses based on random‐effects meta‐analyses.

Trial Sequential Analysis

We will perform Trial Sequential Analysis (Higgins 2008; Wetterslev 2008; Wetterslev 2017) and define:

  • The required information size as the number of participants needed to detect or reject an intervention effect and

  • Firm evidence as being established if the Z‐curve crosses the trial sequential monitoring boundary before reaching the diversity adjusted required information size.

We will construct futility boundaries as thresholds that reflect the uncertainty of obtaining a chance negative finding in relation to the strength of the available evidence. We will perform the analyses with alpha set to 5% and power to 80%. We will repeat the analyses with random‐effects and fixed‐effect models. Based on previous evidence (Als‐Nielsen 2004; EASL/AASLD guideline 2014a; EASL/AASLD guideline 2014b), we will set the relative risk reduction to 10% (mortality), 20% (hepatic encephalopathy), and serious adverse events (20%); use the control group event proportion derived in the meta‐analysis; and model‐based diversity. We will conduct the analyses with inclusion of randomised clinical trials with a low risk of bias (if available) and repeat the analyses with alpha set to 2.5% and power to 90%.

Subgroup analysis and investigation of heterogeneity

We will perform subgroup analyses to investigate heterogeneity in randomised clinical trials stratified by the risk of bias (low risk of bias or high risk of bias); the type of hepatic encephalopathy (overt [acute or chronic], minimal or prevention); the type of aminoglycoside; and mode of administration (intravenous or enteral).

Sensitivity analysis

We will perform sensitivity analyses to evaluate the influence of missing outcome data as described above.

'Summary of findings' tables

We will use the GRADE system to evaluate the quality of the evidence of our primary and secondary outcomes. We will consider the within‐study risk of bias (methodological quality) based on individual domains as well as the overall assessment, indirectness of evidence (population, intervention, control, outcomes), unexplained heterogeneity or inconsistency of results (including problems with subgroup analyses), imprecision of effect estimate (wide CIs and as evaluated with our Trial Sequential Analyses), and risk of publication bias GRADEpro GDT. We will define the levels of certainty in the evidence as 'high', 'moderate,' 'low,' or 'very low.'

  • High certainty: this research provides a very good indication of the likely effect; the likelihood that the effect will be substantially different is low.

  • Moderate certainty: this research provides a good indication of the likely effect; the likelihood that the effect will be substantially different is moderate.

  • Low certainty: this research provides some indication of the likely effect; however, the likelihood that it will be substantially different is high.

  • Very low certainty: this research does not provide a reliable indication of the likely effect; the likelihood that the effect will be substantially different is very high.