Management of Hepatic Encephalopathy in the Hospital

Abstract

Hepatic encephalopathy (HE) develops in about 50% of patients with cirrhosis and is one of the features of decompensated cirrhosis. The inpatient incidence of HE is approximately 23,000/year and management of these patients is common for internists and subspecialists. Treatment of the hospitalized patient with HE has changed in recent years. Treatment entails two phases, induction and maintenance of remission. Most cases of significant hepatic encephalopathy are precipitated by infection, gastrointestinal bleeding, medications or other culprits. All patients should be evaluated for secondary triggers of HE and treatment should be initiated with a non-absorbable disaccharide (i.e. lactulose) in most cases. Rifaximin (off-label) can be added in patients not responding to lactulose. Neomycin is a less preferable alternative to rifaximin, due to its side effect profile. Other therapies including zinc, LOLA, and branch chain amino acids can be considered for patients not responding to a disaccharide and non-absorbable antibiotic. Large portosystemic shunts may be embolized in patients with medically refractory recurrent or severe HE with otherwise well compensated cirrhosis. Molecular Adsorbent Recirculating System is now available for patients with severe hepatic encephalopathy who do not respond to medical therapy. It is critically important that patients hospitalized with significant hepatic encephalopathy continue a maintenance medication(s) at the time of dismissal to prevent further episodes. Patients with a 1^st time episode of HE can be placed on lactulose and careful instruction should be provided to patient and caregiver about titration of dose to achieve 3 bowel movements per day. Patients with recurrent HE episodes despite lactulose benefit from the addition of rifaximin which decreases the frequency of recurrent HE episodes and related hospitalizations. Lastly, patients and their families should be counselled about the risk of motor vehicle accidents which requires mandatory reporting to department of motor vehicles in some states.

Introduction

Hepatic encephalopathy (HE) is a significant neuropsychiatric syndrome that most commonly occurs in decompensated cirrhosis. Clinical features range from clinically imperceptible symptoms in minimal hepatic encephalopathy which require neuropsychometric testing to identify, to a comatose state in the worst cases.¹ The Working Party for Hepatic Encephalopathy established nomenclature for HE in 1998.² Type A hepatic encephalopathy refers to HE secondary to Acute Liver Failure, Type B refers to enteric hyperammonemia (without liver disease), and Type C is associated with chronic liver disease. The severity of HE is graded using the West Haven Criteria (Grade I-IV), but alternative terminology has been suggested and gained some traction. In the new lexicon called SONIC (Spectrum of Neuro-cognitive Impairment in Cirrhosis) covert hepatic encephalopathy (CHE) includes minimal and Grade I HE, and overt hepatic encephalopathy (OHE) encompasses Grade II-IV HE. (Table 1). Episodic HE develops over a short time frame and can fluctuate whereas persistent HE impairs day to day executive function. Most patients with episodic, overt (≥ Grade II) HE will require management in the hospital, which is the focus of this review.

Table 1.

Hepatic Encephalopathy Grades

GRADE	Impairment		SONIC Criteria
0	Normal	Normal	Normal
MHEa	Normal examination findings. Subtle changes in work or driving	Minor abnormalities of visual perception or on psychometric or number tests
1	Personality changes, attention deficits, irritability, depressed state	Tremor and incoordination
2	Changes in sleep-wake cycle, lethargy, mood and behavioral changes, cognitive dysfunction	Asterixis, ataxic gait, speech abnormalities (slow and slurred)
3	Altered level of consciousness (somnolence), confusion, disorientation, and amnesia	Muscular rigidity, nystagmus, clonus, Babinski sign, hyporeflexia
4	Stupor and coma	Oculocephalic reflex, unresponsiveness to noxious stimuli

^aMHE=minimal hepatic encephalopathy

Hepatic encephalopathy eventually occurs in 50% of cirrhotics.^3,4 Hepatic encephalopathy portends a worse survival for patients compared to similar patients without HE, even after accounting for the Model for End Stage Liver Disease (MELD) score.⁵ The development of hepatic encephalopathy merits consideration of liver transplantation. Whether treatment of HE alters survival is unknown. Treatment of HE continues to be a significant area of investigation. Currently, non-absorbed disaccharides (ex. lactulose; lactitol) and non-absorbable antibiotics (ex. neomycin; rifaximin) represent the mainstay of treatment. (Table 2)

Table 2.

Hepatic Encephalopathy Treatment Options

Drug	FDAa Approved	Approval Year
Neomycin	Yes	1970
Lactulose	Yes	1976
Metronidazole	No	–
Vancomycin	No	–
Rifaximin	Yes	2010
MARSb	Yes	2013

^aFDA= Food and Drug Administration,

^bMARS=Molecular Adsorbent Recirculating System

Hospitalization for episodic OHE, or the development of OHE during hospitalization is common. In the U.S. Nationwide Inpatient Sample, the inpatient incidence of HE ranged from 20,918 (2005) to 22,931 (2009).⁶ Up to 80% of OHE episodes are precipitated by an event such as infection or gastrointestinal bleeding. Management of the hospitalized patient with episodic OHE is directed at correcting the underlying precipitant and providing pharmacologic treatment that reduces ammonia-genesis.

Most patients will require maintenance medications at the time of hospital dismissal as secondary prophylaxis for episodic OHE. Data suggest that many patients do not receive maintenance medication at/after dismissal. An abstract presented at AASLD annual meeting in 2012 characterized a subset of insurance claims for patients by ICD-9 code for HE (572.2) and compared this to prescriptions filled during the calendar years 2009-2011. For years 2009 (n=13,623), 2010 (n=15,529) and 2011(n=16,328), 89.2%, 87.8%, and 86.4% had inpatient claims for HE respectively, and 60.3%, 62.3%, and 63.9% did not receive ongoing treatment.^7,8 Volk and colleagues also described a high readmission rate (69%) among a cohort of patients with decompensated cirrhosis (n=402) where one of the most common reasons for preventable readmission was recurrent HE due to lack of education on, or inappropriate use of lactulose.⁹ Thus, more attention should be focused on assuring patients are prescribed and educated about maintenance medication for secondary prevention of OHE at the time of hospital dismissal.¹⁰ In this review, we summarize the evidence on the optimal medical treatments for patients who have been hospitalized for an episode of episodic OHE and suggest treatment algorithms for induction and maintenance of remission (secondary prophylaxis). Liver transplantation for hepatic encephalopathy is not covered in this review.

The search strategy employed for this review included search terms ‘hepatic encephalopathy’ AND ‘treatment’ in pubmed, with filters of human and last 10 years. Bibliographies were also manually searched as well as abstracts from recent liver conferences. Clinical trials were emphasized when discussing evidence for each modality.

Evidence for Induction Therapies to Treat Episodic, Overt Hepatic Encephalopathy

Non-Absorbable Disaccharides

Lactulose (β-galactosidofructose) and lactitol (β-galactosidosorbitol) reduce ammonia levels by acidification of the colon with resultant conversion of ammonia to ammonium, shifting the colonic flora from urease to non-urease producing bacterial species, and by its cathartic effect. Non-absorbable disaccharides have demonstrated variable efficacy in clinical trials. An often-cited meta-analysis performed in 2004 showed that non-absorbable disaccharides were superior to placebo but did not improve survival. When only high quality trials were included in this meta-analysis, non-absorbable disaccharides had no effect on HE. There have been no trials of non-absorbable disaccharides versus placebo since these results. Despite the mixed results, lactulose still remains the 1^st line therapy for treatment of acute, episodic OHE. Decades of clinical experience with lactulose speaks to its effectiveness to reverse episodic OHE except in all but the most severe of cases. The discordance between the efficacy in clinical trials and real life effectiveness is manifold and includes heterogeneity of types of HE (minimal vs. overt vs. chronic), differences in the prognostic importance of HE precipitants, and subjectivity of HE assessment tools amongst others. Clinical guidelines recommend lactulose or lactitol as 1^st line therapy.¹⁰

Neomycin, Metronidazole and other Antibiotics

Neomycin is a poorly absorbed aminoglycoside used to decrease gut bacteria-derived ammonia and is FDA approved for use in ’acute’(episodic), overt HE but not chronic HE. There are multiple older studies that explore the efficacy of this agent in HE. One of the earliest studies from Conn et al. in 1978 was a randomized, controlled trial comparing neomycin-sorbitol to lactulose in ‘acute’ (episodic) HE demonstrating no difference between the 2 therapies.¹¹ One of the most recent studies from 1992 compared neomycin 6g/day (n=20) to placebo (n=19) for episodic HE in a double blind randomized, trial and found no significant difference in treatment failures (2 per group) or time to resolution of HE in the neomycin arm (39.11 +/− 23.04 hours) vs. placebo. (49.47 +/− 21.92).¹² In general, the evidence for neomycin in episodic OHE is weak, and its use is complicated by the risk of oto- and nephro-toxicity. Although FDA approved, the decreased efficacy and side effects compared to other therapies limit its clinical utility. Other small trials have evaluated metronidazole and vancomycin and suggested some benefit but the risk of neurotoxicity and vancomycin-resistant enterococci colonization, respectively, hamper any enthusiasm about using these agents as mainstays in the armamentarium for HE.^13-16,17,18

Rifaximin

The role of rifaximin in the treatment of episodic OHE is contentious. Rifaximin is not FDA approved for treatment of episodic OHE, only for secondary prevention of overt HE. Older trials that evaluated rifaximin for episodic HE used different comparators and generally enrolled small numbers of patients with ‘acute’, chronic, or unclear acuity of HE. Bucci et al. (1993) compared rifaximin (1200mg/day) to fixed dose lactitol (30grams/day) in 30 and 28 patients with moderate to severe HE, respectively.¹⁹ It was not clear whether these patients had overt or chronic HE. Endpoints were HE grade, neuropsychometric tests, and ammonia at days 3, 6, 9, 12, and 15. Overall, there did not appear to be any clinically important differences between groups at day 15 across all endpoints considered, but rifaximin did seem to result in faster improvement when compared to lactitol. A randomized controlled trial of rifaximin versus placebo for acute HE (grades I-III) showed improvement in a composite outcome (including mental status, neuropsychometric testing, EEG findings, and ammonia) at dose of 1200 and 2400mg/day.²⁰ A randomized, double-blinded, double dummy, controlled trial compared rifaximin 1200mg (n=50) vs. lactitol 60g/day(n=53) and demonstrated an approximately 80% efficacy in both arms after 5-10 days of treatment. An open label, randomized trial in Koreans compared rifaximin 1200mg (n=32) to lactulose 90cc/day (n=22) in overt HE grade I-III and found no difference in improvement in HE grades or composite endpoints.²¹ Rifaximin was better tolerated in most studies, as anticipated. Table 3 summarizes data in trials comparing lactulose/lactitol to rifaximin which generally demonstrate similar outcomes in both treatment groups. Given the small number of trials with methodologic flaws, the question of using rifaximin as monotherapy for episodic OHE remains unanswered.

Table 3.

Disaccharides versus Rifaximin for Treatment of Hepatic

Reference	Method	HE Type	Intervention	Outcome Measure	Results
Bucci 1993	N=58 DBa, RCTb	Unknown	Rifaximin 1200mg Lactulose 30g	Mental status, asterixis, cancellation test, Reitan test, EEGc, ammonia	Improvement in MSd, ammonia, cancellation Test, EEG more pronounced and faster onset in rifaximin.
Festi 1993	N=21 RCT, OL	Grade I HEe	Rifaximin 1200mg Lactulose 40 g	HE sxf, ammonia, serum sodium	HE sx improvement comparable; faster onset for asterixis, ammonia, Na+f in rifaximin group
Massa1993	N=40 DB, RCT	Grade 2-3 HE	Rifaximin 400mg TID, Lactulose 20grams thrice/day	PSEg Index	Mental status improvement favored rifaximin as did ammonia level (faster drop), Reitan test, and earlier change in EEG
Mas 2003	N=103 DB, RCT	Grade 1-3 Acute HE	Rifaximin 1200mg, Lactitol 60mg (titrated)	PSE Index	Comparable overall global improvements in both groups (≈ 80%) with rifaximin patients demonstrating better HE grade, ammonia level, and EEG (statistically significant), and higher % with complete HE resolution (53% vs. 37%)
Paik 2005	N=54 OLh, RCT	Grade 1-3	Rifaximin 1200mg, Lactulose 90g	HE index, mental status, asterixis, Reitan test	Improvementsfor both, no statistically significant difference for any outcome

Encephalopathy: Summary of Trials

^aDB=double blind,

^bRCT=randomized controlled trial,

^cEEG=electroencephalogram,

^dMS=mental status,

^eHE=hepatic encephalopathy,

^fNa+=sodium,

^gPSE=portosystemic encephalopathy,

^hOL=open label

Rifaximin use for episodic OHE, in addition to lactulose, has become increasingly common despite a previous lack of evidence. Fortunately, a recent RCT (n=120) was conducted by Sharma et al. comparing rifaximin with lactulose to lactulose and placebo in patients with OHE.²² Eighty percent of patients had severe HE, grade III or IV and 70% were Child Class C with the remainder Child Class B. Patients in the lactulose and rifaximin group had a higher proportion of complete reversal of HE (76% vs. 50.8%, p<0.004), shorter hospital stays, and a striking improvement in 10 day mortality (49.1% vs 23.8%, p<0.05). The very high mortality in the lactulose plus placebo arm raises some concerns about the validity of this study, which should be repeated in a larger number of patients at multiple sites. In the meantime, this is likely the best data that will be available for some time, with which to make evidence based treatment decisions.

Zinc

Zinc deficiency is common in cirrhosis. In a recent clinical trial, zinc deficiency prevalence was 96% among patients with a median MELD=12.²³. Ammonia is converted to urea by ornithyl transcarbamylase in the liver and is combined with glutamate by glutamine synthetase in the skeletal muscle to form glutamine. Both ammonia-reduction pathways are impaired by zinc deficiency. Treatment with zinc has been shown to enhance the formation of urea from ammonia and amino acids.²⁴ Zinc therapy has been the focus of 4 randomized, controlled trials, producing heterogeneous results. Three of these trials are now 20 years old or more.^25-27 The most recent study from Takuma et al. in 2010 randomized patients with cirrhosis and HE Grade I-II refractory to standard treatment, to receive zinc treatment (n=39) in addition to lactulose and branch chain amino acids(BCAA), versus no zinc (n=40) with BCAA and lactulose.²³ Patients were followed for 6 months to determine the effect on quality of life and HE. HE improved in 21(54%) vs. 10 (26%) in the zinc vs. no zinc arms, respectively, with 16 (41%) zinc-treated patients improving to HE grade zero. Other endpoints including ammonia level, psychometric testing, and QOL were met in the zinc treatment arm. While this study had some flaws including lack of blinding and the zinc formulation (contained l-carnosine), it is the only recent study with reasonable evidence to suggest the benefit of zinc. There is not enough data to define the optimal dose of zinc. Zinc is relatively well tolerated with a rare side effects of dyspepsia, and copper deficiency (with chronic, high dose use), and can decrease the effectiveness of ciprofloxacin if taken at the same time (take zinc 2 hours before or 6 hours after ciprofloxacin).

L-Ornithine L-Aspartate

L-ornithine-L-aspartate (LOLA) is a compound salt that stimulates ornithine transcarbamolyase and carbamoyl phosphate synthetase, and is a substrate for the formation of urea. LOLA also works by stimulating glutamine synthesis in the skeletal muscle and consequently lowering ammonia. LOLA has been most thoroughly evaluated in the setting of chronic HE. Randomized, placebo controlled, double-blind studies were performed in Germany using an intravenous and an oral form of LOLA in patients with chronic HE, both of which demonstrated improvements in the number connection tests, ammonia values, and HE parameters (mental state gradation and Portosystemic Encephalopathy Index).^28,29 Episodic, recurrent HE was specifically excluded from these studies. Little data exists for management of episodic OHE or prophylaxis. One study from Pakistan evaluated LOLA as adjunctive treatment versus placebo in patients who were allowed to receive standard medical treatment (SMT).³⁰ Patients with Grade II HE or above had improvement in HE grade on SMT + LOLA (79%) vs. SMT + Placebo (55%) which was significant (p=0.019). In other countries, LOLA is used as adjuvant therapy, but this medicine is not available in the U.S.

Branch Chain Amino Acids

The plasma amino acid profile in patients with cirrhosis is altered with a decrease in branch chain amino acids (BCAA) and increase in aromatic amino acids. The BCAAs are a source of glutamate which helps to metabolize ammonia in skeletal muscle. The benefits of BCAA in liver disease have been investigated for several decades. BCAA supplementation may improve albumin synthesis, decrease insulin resistance, decrease hepatocellular carcinoma, and improve immune function.³¹ Two randomized controlled trials demonstrated that BCAA improved important composite endpoints of death/hospitalization metrics in one study and hepatic failure, variceal bleeding, HCC, and mortality in a 2^nd study.^32,33 BCAA has been studied in HE as well, to a lesser degree. An early study indicated that BCAA was effective for latent (minimal) HE.³⁴. More recently, BCAA was administered to 58 patients with one previous episode of OHE, and compared to 58 patients receiving maltodextrin.³⁵ There was no significant difference in the frequency of recurrent HE. However, there was significant loss to follow-up in this trial, so results must be interpreted cautiously. Currently, the European Society for Clinical Nutrition and Metabolism (ESPEN) recommend use of 1.2g/kg/day of protein for compensated cirrhosis, and 1.5g/kg/day in decompensated cirrhosis. This recommendation was based on the results of a randomized controlled trial of a normal protein diet (1.2g/kg/day) vs. a restricted diet demonstrating no effect on the outcome of episodic HE, but increased muscle breakdown in the low protein diet group.³⁶ Additionally, ESPEN provide a grade A recommendation for the use of standard protein supplementation in patients with HE grade ≤ 2 and BCAA preparations for HE grade III IV.³⁷

Percutaneous Embolization of Large Portosystemic Shunts

Two large retrospective series have been published demonstrating efficacy and safety of embolization of large portosystemic shunts in medically refractory HE. In the European multi-center cohort study (n=37), 59% of patients were free of HE within 100 days and 48% were HE-free over an average of 2 years after embolization.³⁸ There was one hepatic capsular hemorrhage, otherwise there were no other major peri-procedural complications. Long term safety appeared good with no increase in variceal bleeding events. In the largest US series (n=15), 90% of cirrhotics improved at 2 months post procedure.³⁹ One patient developed an infected hepatic cyst 2 weeks post-procedure, otherwise there were no significant complications attributed to the procedure. The median MELD in both studies was 13. Logistic regression performed in the European study suggested that patients with MELD >11 were at risk of HE recurrence after shunt embolization.

Molecular Adsorbent Recirculating System (MARS)

Molecular Adsorbent Recirculating System (MARS) was introduced in 1999, based on the concept of albumin dialysis. This system was designed to remove protein/albumin bound toxins such as bilirubin, bile acids, nitrous oxide and endogenous benzodiazepines (amongst others) and also removes non-protein bound ammonia that accumulates in liver failure. Over 45,000 MARS treatments had been administered to approximately 15,000 patients. While the effect of MARS on survival for patients with liver failure remains in question, 3 studies have shown improvements in hepatic encephalopathy. The most recent, and largest trial (RELIEF Trial) enrolled 189 patients with acute on chronic liver failure (ACLF) and evaluated MARS plus standard medical therapy (SMT) vs. SMT alone, on the primary endpoints of 28 day and 90 day LTx free survival. Survival endpoints were not met, but safety was demonstrated. There was a higher proportion of patients with MELD >20 (78.9% vs. 69.7%, p=0.16)) and SBP (14.4% vs. 6.7%, p=0.94) at baseline in the MARS treatment group. The proportion of patients with HE Grade III-IV HE improvement to HE Grade 0-I was higher in MARS treated patients (15/24; 62.5%) compared to standard medical therapy (13/34; 38.2%) which trended toward statistical significance (p=0.07).⁴⁰ In another study designed specifically to evaluate the effect of MARS on HE, 70 patients with Grade III (56%) and grade IV HE (44%) were randomized to receive MARS + SMT vs. SMT alone. The primary endpoint was met whereby a higher proportion of patients had a 2 grade improvement in HE in the MARS arm (mean 34%) vs SMT arm (19%) with a p value=0.044 and more rapid improvement (p=0.045).⁴¹ MARS was also well tolerated in this trial. In the smallest study, MARS had a statistically significant impact on improvement of HE in 9 patients with alcoholic hepatitis and HE (without standard medical therapy) and decreased circulating aromatic phenolic amino acids.⁴² The FDA has approved use of the MARS system for HE related to decompensation of chronic liver disease. Exclusion criteria in the 2 largest trials included active hemorrhage, active infections, severe cardiopulmonary disease, renal replacement therapy, and hemodynamic instability amongst others, which will decrease the applicability of MARS. Also, MARS may reduce the bioavailability of certain antibiotics.⁴³ Cost is also a major concern with this modality. Nevertheless, MARS appears to be a viable option for patients with severe HE unresponsive to standard medical therapy.

Evidence for Secondary Prophylactic/Maintenance Strategies for Hepatic Encephalopathy

Lactulose

The results of 2 recent trials provided solid evidence for the practice of secondary prevention with lactulose alone, or lactulose and rifaximin. Sharma and colleagues performed an open label study randomizing 140 patients to placebo or daily lactulose after recovery from an episode of overt hepatic encephalopathy.⁴⁴ After drop out and protocol violations, 19.6% (12/61) of lactulose-treated patients experienced recurrent OHE vs. 46.8% (30/64) in the placebo (p=0.001) over a median follow-up time of 14 months. A similarly designed trial by Agrawal et al. with patients recovered from OHE received lactulose (n=80), probiotics (n=77), or no therapy (n=78).⁴⁵ Patients in the probiotic arm received 3 capsules per day containing 112.5 billion viable lyophilized bacteria per capsule containing 4 strains of Lactobacillus, 3 strains of Bifidobacterium, and one strain of Streptococcus salivarius subspecies thermophilus. The intention to treat results demonstrated a significantly lower rate of OHE for lactulose (37.5%) and probiotic (45.4%) as compared to no treatment (64.1%). Ideally, these results should be reproduced in studies in other countries/centers. The role of probiotics as stand-alone maintenance treatment is also interesting and deserving of further investigation.

Rifaximin

In 2010, Bass et al. published results from a randomized controlled trial of rifaximin (n=140) vs. placebo (n=159) for secondary prevention of episodic OHE in adult patients with ≥ 2 previous episodes of un-precipitated HE and were in remission at the time of enrollment.⁴⁶ The MELD score was required to be ≤ 25 for participants. Over 90% of patients in both arms were taking lactulose. Patients experienced a nearly sixfold reduction in breakthrough-HE in the rifaximin group (31/140) compared with placebo (73/159) with a hazard ratio of 0.42 (95% Cl 0.28-0.64, p <0.001). The authors also reported a 50% reduction in hospitalizations for the rifaximin group (19/140) as compared with placebo (36/159). Adverse events were very similar between the rifaximin and placebo groups. Based on data from this phase III study, rifaximin was approved by the FDA for secondary prevention of OHE. It is important to reiterate that this trial did not address rifaximin use for episodic HE. A trial to evaluate the safety and efficacy of rifaximin in cirrhotics with MELD ≥ 25 is underway (NCT01846663). Another ongoing trial is examining rifaximin plus lactulose versus rifaximin alone for secondary prevention (NCT01842581).

Patients with advanced MELD scores or CTP B/C cirrhosis will require secondary prophylaxis after recovery from spontaneous or precipitated episodic HE. Child-Pugh Class A cirrhotics who are otherwise well compensated but develop isolated OHE in the setting of infection or GI bleeding, represent a more challenging decision. Some of these patients may not need or benefit from long-term maintenance treatment with lactulose but data to drive patient selection for this strategy is not readily available. Decisions about secondary prophylaxis for these patients are made on a case-by-case basis with no guidelines for best care. Ideally, these types of patients would receive neuropsychometric testing for covert HE within a few weeks after leaving the hospital. However, neuropsychometric testing for covert HE are not practical for most patients and a simple clinical tool to detect covert HE is badly needed. Patients with one previous, precipitated (ie. GI bleeding) OHE event, otherwise well compensated(Child Pugh A), and have lingering CHE as identified by specialized testing, might warrant maintenance ongoing treatment with lactulose. However, this potential strategy requires further investigation before incorporating into standard clinical practice.

Approach to Induction and Maintenance Treatment of Hepatic Encephalopathy

Approach to Induction and Maintenance Treatment for the First Episode of Episodic Overt Hepatic Encephalopathy, West Haven Grade I-II (Figure 1)

Figure 1.

Management of First Episode of Hepatic Encephalopathy, Grade 1-2

^aHE=hepatic encephalopathy, ^brifaximin is not FDA approved for overt hepatic encephalopathy

The first step in management of episodic OHE is evaluation for the typical precipitants of OHE including gastrointestinal bleeding, infections, new medications such as opioids or benzodiazepines, constipation, diarrhea, dehydration, alkalosis or hypokalemia, and hypoxemia. Up to 80% of patients may have a precipitant. When a precipitant is found, management of the precipitant along with concomitant lactulose therapy is recommended. For those not responding to initial treatment, it is important to re-evaluate the diagnosis of OHE, review the possibility of other precipitants, and ensure that the patient is having 3-4 stools/day while on lactulose. For example, a head CT may be necessary in a patient who has a new focal neurologic defect on re-examination. If the diagnosis of OHE is correct, no other precipitants are found, and bowel movements are adequate, then neomycin or rifaximin could be added to lactulose. Neomycin is FDA approved for this indication, but there are obvious concerns over nephrotoxicity, ototoxicity, and to some degree, efficacy. Rifaximin use in this setting is off-label . However, with the recent RCT demonstrating improved HE outcomes and a mortality benefit in patients on lactulose with rifaximin, we would recommend using rifaximin preferentially. Patients who recover from a first episode of OHE generally require lactulose maintenance therapy, especially if Child Pugh Class B/C.

Approach to Induction and Maintenance Treatment for Recurrent (≥2 episodes) Episodic Overt Hepatic Encephalopathy (Figure 2)

Figure 2.

Management of Recurrent Episode of Hepatic Encephalopathy, Grade 1-2

^aHE=hepatic encephalopathy, ^brifaximin is not FDA approved for overt hepatic encephalopathy, ^cPS=portosystemic, ^dL-ornithine-L-aspartate, ^eBCAA=branch chain amino acids, ^fMELD=Model for End Stage Liver Disease Score,

The management of recurrent OHE requires a careful evaluation for precipitating factors. In addition, one must scrutinize the baseline maintenance medication regimen for these individuals. Inappropriate lactulose dose titration to achieve 3 stools per day, and non-compliance are common. Patient education or re-education is important part of the management of this scenario. If symptoms such as bloating or excessive diarrhea occur, even on small doses of lactulose, rifaximin or neomycin should be substituted for lactulose. Breakthrough OHE on an appropriate lactulose regimen calls for the addition of rifaximin (off-label) or neomycin, though rifaximin is preferred. In more challenging patients who develop recurrent episodic OHE on appropriately used lactulose and rifaximin, the provider should consider looking for a large intra-abdominal portosystemic shunt with a contrast enhanced CT. In the trial of secondary prevention with lactulose by Sharma et al, the frequency of large spontaneous shunts was 23% (32/140) but may be present in up to 70% of patients with persistent HE.⁴⁷ The risk of CT contrast nephropathy needs to be considered in these patients who often have falsely normal creatinine with tenuous renal function, “Magnetic Resonance Angiography is an alternative for patients with GFR >30ml/min/BSA. For well compensated patients with Child Pugh Class A cirrhosis and/or low MELD scores arbitrarily defined as <12-15, percutaneous embolization should be considered which can result in remission from recurrent OHE in 59-90% of patients. Recurrent OHE after shunt embolization is more likely with MELD scores >11. If a shunt is not found or high MELD precludes embolization, then additional treatments such as zinc, L-ornithine L-aspartate (where available) and branch chain amino acid feeding should be considered. It is reasonable to check zinc levels in patients with HE to help guide dosing, using a dose of zinc sulfate 220mg (contains 50mg elemental zinc) once daily in those with normal zinc levels. Once resolution of episodic OHE is achieved, all patients with recurrent OHE need to be maintained on lactulose or rifaximin or both depending on their initial therapy. For instance, a patient with breakthrough OHE on lactulose will need a maintenance regimen of lactulose and rifaximin.

Management of Severe Hepatic Encephalopathy (West Haven Criteria Grade III-IV) (Figure 3)

Figure 3.

Management of Severe Episode of Hepatic Encephalopathy, Grade 3-4

^aHE=Hepatic Encephalopathy, ^bNG=nasogastric tube, ^cCT=computed tomography, ^dEEG=electroencephalogram, ^erifaximin is not FDA approved for overt hepatic encephalopathy, ^fMELD=Model for End-Stage Liver Disease, ^gLOLA= L-ornithine-L-aspartate, ^hBCAA=branch chain amino acids, MARS= Molecular Adsorbent Recirculating System,

Grade III-IV hepatic encephalopathy is a serious condition that requires ICU monitoring. Patients who cannot protect their airway due to decreased consciousness require endotracheal intubation and mechanical ventilation. After a thorough evaluation for precipitants, patients should be given lactulose via NG tube. Lactulose at 15-30cc can be given every 1-2 hour via NG until 3 stools are achieved. If an NG or OG access is not available, then 300cc of lactulose can be given in 1 L of water as an enema [300cc of lactulose (10g/15ml) in 700cc of sterile water]. This can be repeated as necessary, though care should be taken to avoid excessively loose or voluminous stool. We recommend down-titration or temporary interruption of lactulose in that circumstance. Rifaximin should be administered with lactulose in patients with grade III or IV HE. Those who are correctly diagnosed with OHE and precipitating factors have been sought and addressed, and who do not respond to lactulose and/or non-absorbable antibiotics, should be re-evaluated to make sure the diagnosis is accurate. If the diagnosis is in question, a head CT scan and an EEG may be helpful to rule out the possibility of a CNS bleeding event or non-convulsive status epilepticus, respectively. If the diagnosis of OHE is accurate, then the patients should be evaluated for a large portosystemic shunt. A large shunt should be treated in the presence of refractory HE and relatively low MELD scores (<12-15). The majority of initial medication non-responders will not have a large shunt, and can be treated with branch chain amino acids, zinc, and where available, L-ornithine L-aspartate. Lastly, for patients not responsive to the aforementioned approach, MARS should be considered. Patients recovering from severe HE should be kept on a maintenance program of lactulose and rifaximin.

Driving Instructions for Patients Leaving the Hospital

Minimal Hepatic Encephalopathy (MHE) is a risk factor for motor-vehicle accidents (MVA). While the term covert HE (which encompasses MHE) has replaced the term MHE, we will use the term MHE in this section as this was the terminology used in the cited literature. It is likely that a substantial percentage of patients hospitalized for an episode of HE and who recover, will have residual minimal encephalopathy after dismissal.^48,49 Therefore, counseling about risks of driving at dismissal is an important. Some literature has amassed to quantify the magnitude of risk for MVA in patients with MHE. Bajaj et al. evaluated 167 cirrhotic patients prospectively for 1 year and identified 18 MVA’s through department of transportation records.⁵⁰ Of those 18 MVA’s, 16 (89%) occurred in patients identified to have MHE by the Inhibitory Control Test (ICT), as compared to 8/18 (44%) by standard psychometric tests (SPT). Conversely, it is important to point out that about 55% of patients who did not have an MVA were identified as having MHE. Another study assessed real world driving ability in controls (n=48), cirrhosis with no encephalopathy (n=10), minimal hepatic encephalopathy (n=27) and 14 patients with overt, grade I HE. Fitness to drive, as deemed by a driving instructor, was 87% (control), 75% (no HE), 48% (MHE), and 39% (Grade I HE).⁵¹ An earlier on-road driving study demonstrated a significant reduction in car handling, adaptation, and cautiousness among patients with MHE (n=14) compared to controls or cirrhotic patients without MHE.⁵² The instructor had to intervene on 5/14 patients with MHE to avert an accident. In summary, some patients with MHE have impaired driving skills, which can translate into real world risk of MVA. Yet, we do not have widely available and practical tools to predict who is at risk.

What are the medico-legal ramifications for the provider? No states have specific legislation that pertains to patients with hepatic encephalopathy. Only 6 states have mandatory reporting laws requiring the physician to report drivers with general medical impairment: California, Delaware, Nevada, New Jersey, Oregon, and Pennsylvania. Of the remaining 44 states, 25 provide legal immunity to physicians for reporting patients who have medical impairments.⁵³

We recommend that treating physicians consider testing for MHE if they are in tertiary centers with access to these tests. Patients found to have MHE should return to their licensing agency for a road test. It has been suggested, based on positive driving simulator evidence, that patients who have MHE and have failed a driving test could be treated with rifaximin and repeat the driving test.⁵⁴ However, this strategy cannot be endorsed for all patients at this time without more evidence. In a cost effectiveness analysis, using the ICT to diagnose MHE and treatment with lactulose to prevent MVA would be cost saving. Treatment with rifaximin would save costs in this context only if the monthly cost was <$353.⁵⁵

Conclusion

In summary, hepatic encephalopathy eventually occurs in 50% of cirrhotic patients and heralds a poor prognosis. Patients with episodic HE are primarily cared for in the hospital. Treatment of the hospitalized patient with episodic OHE can be compartmentalized into induction treatment and maintenance of remission. Lactulose remains the cornerstone of treatment for both induction and maintenance of remission. There is now evidence to support the use of rifaximin as adjunctive therapy for severe OHE but is not yet FDA approved for this indication. Alternatively, neomycin can be used as adjunctive therapy, but its side effect profile makes it a less attractive choice. Percutaneous embolization of large portosystemic shunts and MARS therapy are emerging modalities with evidence to support their use for medically refractory OHE. The vast majority of patients of require maintenance medications when dismissed from the hospital and patient/caregiver education about the role of those medications and appropriate dose titration for lactulose is crucial. Rifaximin should be added to lactulose for patients with recurrent overt hepatic encephalopathy. All patients and their families should be counseled about the risks of motor vehicle accidents. Mandatory state reporting laws should be followed. It is recommended that these patients have a fitness to drive evaluation in the outpatient setting.

Article Highlights.

• Episodic, Overt Hepatic Encephalopathy is responsible for an increasing number of hospital admissions and hospital re-admissions.

• New terminology has been suggested and is gaining traction in which West Haven Grade 0-1 HE is Covert Hepatic Encephalopathy (CHE) and Grades 2-4 HE is Overt Hepatic Encephalopathy (OHE).

• Most patients with OHE require hospital based care during the induction treatment phase followed by a maintenance treatment strategy in the outpatient setting.

• Lactulose is the mainstay of induction and maintenance treatment.

• Rifaximin, when added to lactulose has been shown to prevent episodes of OHE and hospitalization, when compared to lactulose alone.

• Rifaximin and neomycin are acceptable adjunctive therapies for treatment of patients with OHE that is not responsive to lactulose, or severe OHE, though rifaximin may be preferable due to a better side effect profile.

• Evidence for use of zinc, L-ornithine L-aspartate, and branch chain amino acids are less compelling, while there is increasing data to suggest a benefit of portosystemic shunt embolization in carefully selected patients.

Abbreviations

AASLD

American Association of the Study for Liver Disease

BCAA

branch chain amino acids

CHE

covert hepatic encephalopathy

CT

computed tomography

EEG

electroencephalogram

ESPEN

European Society for Clinical Nutrition and Metabolism

FDA

Food and Drug Administration

HE

hepatic encephalopathy

LOLA

L-ornithine L-aspartate

LTx

liver transplantation

MARS

Molecular Adsorbent Recirculating System

MHE

minimal hepatic encephalopathy

MELD

Model for End-Stage Liver Disease

OHE

overt hepatic encephalopathy

SMT

standard medical treatment