Critical flicker frequency test predicts overt hepatic encephalopathy and survival in patients with liver cirrhosis

Accepted Manuscript Title: CRITICAL FLICKER FREQUENCY TEST PREDICTS OVERT HEPATIC ENCEPHALOPATHY AND SURVIVAL IN PATIENTS WITH LIVER CIRRHOSIS Authors: Michele Barone, Endrit Shahini, Andrea Iannone, Maria Teresa Viggiani, Valeria Corvace, Mariabeatrice Principi, Alfredo Di Leo PII: DOI: Reference:

S1590-8658(18)30169-5 https://doi.org/10.1016/j.dld.2018.01.133 YDLD 3658

To appear in:

Digestive and Liver Disease

Received date: Accepted date:

27-9-2017 17-1-2018

Please cite this article as: Barone Michele, Shahini Endrit, Iannone Andrea, Viggiani Maria Teresa, Corvace Valeria, Principi Mariabeatrice, Di Leo Alfredo.CRITICAL FLICKER FREQUENCY TEST PREDICTS OVERT HEPATIC ENCEPHALOPATHY AND SURVIVAL IN PATIENTS WITH LIVER CIRRHOSIS.Digestive and Liver Disease https://doi.org/10.1016/j.dld.2018.01.133 This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

Predictors of hepatic encephalopathy CRITICAL

FLICKER

FREQUENCY

TEST

PREDICTS

OVERT

HEPATIC

ENCEPHALOPATHY AND SURVIVAL IN PATIENTS WITH LIVER CIRRHOSIS Michele Barone PhD 1, Endrit Shahini MD 1, Andrea Iannone MD 1, Maria Teresa Viggiani MD 1,

1

IP T

Valeria Corvace MD 2, Mariabeatrice Principi MD 1, Alfredo Di Leo PhD 1.

Gastroenterology Unit, Dept. of Emergency and Organ Transplantation (D.E.T.O.), University of

2

SC R

Bari, Italy. Gastroenterology Unit, San Camillo Hospital, Manfredonia (Foggia), Italy.

U

Corresponding Author Alfredo Di Leo, MD, PhD

N

Gastroenterology Unit,

A

Dept. of Emergency and Organ Transplantation (D.E.T.O.),

M

University of Bari, Italy.

Telephone: +390805592577; Fax: +390805593251.

ED

Email address: [email protected].

PT

ABSTRACT

CC E

Background: A critical flicker frequency (CFF)≤39Hz identifies cirrhotic patients with minimal hepatic encephalopathy (mHE) and predicts the risk of both overt hepatic encephalopathy (oHE) and mortality in patients with previous episodes of decompensation and/or oHE.

A

Aims: Herein, we evaluated the effectiveness of CFF in predicting the first episode of oHE and

survival in cirrhotics who had never experienced an episode of oHE. Methods: Our cohort study of 134 patients and 150 healthy subjects were examined. A CFF>39Hz was considered normal and pathological when ≤39Hz. The median follow up was 36 months. Results: At baseline, all controls had CFF>39Hz. Ninety-three patients had a CFF>39Hz and 41 1

Predictors of hepatic encephalopathy had a CFF≤39Hz. The prevalence of CFF≤39Hz significantly increased with the progression of the Child-Pugh class (p=0.003). Moreover, the risk of oHE was increased by CFF≤39 (p <0.001, by log-rank test) [HR=7.57; CI(3.27-17.50); p<0.0001, by Cox model] and ammonia [HR=1.02 CI(1.01-1.03), p=0.0009]. Both a CFF value≤39 Hz and Child-Pugh class were independent predictors of mortality by Cox model [HR=1.97; CI(1.01-3.95), p=0.049; HR=3.85 CI(1.68-8.83),

IP T

p=0.003].

Conclusions: CFF predicts the first episode of oHE in cirrhotics that had never experienced oHE,

SC R

and predicts mortality risk. These findings suggest that cirrhotic patients should be routinely screened by CFF.

U

Abbreviations: CFF, critical flicker frequency; HE, hepatic encephalopathy; mHE, minimal hepatic

A

N

encephalopathy; model for end stage liver disease, MELD; overt hepatic encephalopathy, oHE.

INTRODUCTION

ED

M

Keywords: hepatonorm analyzer; liver decompensation; neurophysiological test; hyperammonemia.

PT

Hepatic encephalopathy (HE) is a potentially reversible disorder that occurs in cirrhotic patients with neuropsychiatric abnormalities and motor disturbances that range from mild alterations of

CC E

cognitive and motor functions to coma and death 1-2. This condition has been linked to alterations of the gut flora that increase the production of gut-derived toxins, such as ammonia and indoles,

A

leading to a condition of endotoxemia associated to systemic and cerebral inflammation subclinical expression of HE is a defined minimal hepatic encephalopathy (mHE)

5-7

3-4

. The

. The latter

condition is characterized by the presence of various quantifiable neurophysiological and neuropsychological deficits when specific diagnostic tools, such as the paper-and-pencil tests and its variants, as well as critical flicker frequency (CFF), are used 8-11.

2

Predictors of hepatic encephalopathy The visual test based on CFF measures the frequency (Hz) when impression of fused light turns to a flickering one 5,11. This neurophysiological test has an elevated specificity and reproducibility, with only little biases due to training effects and daytime variability, making it particularly useful in establishing the efficacy of a putative therapeutic treatment 7,11-17. It has been demonstrated that CFF is able to predict not only the risk of overt HE development but

cirrhotics and patients with a previous history of overt HE

11,15,18

IP T

also the risk of mortality in a cohort of patients including a discrete number of Child-Pugh class C . Similarly, CFF predicts the risk

19-20

SC R

of overt HE in cirrhotics that have undergone transjugular intrahepatic portosystemic shunt (TIPS) .

On these premises, we screened a large number of consecutive cirrhotic patients that underwent

U

CFF evaluation at baseline. Thereafter, their clinical status was monitored as outpatients every 6

N

months in order to study the occurrence of the first episode of overt HE and evaluate survival. The

A

novelty of our study relies on the fact that the evaluation of CFF value as risk factor for both the

M

development of the first episode of oHE and the liver related mortality was performed in non-

ED

alcohol related cirrhotic patients that had never experienced previous episodes of hepatic

A

CC E

PT

encephalopathy, an aspect that was not considered in previous studies.

3

Predictors of hepatic encephalopathy METHODS This was an observational study performed by the medical and nursing staff of the Gastroenterology Unit, Department of Emergency and Organ Transplantation (D.E.T.O.) of the “Aldo Moro” University of Bari, Italy. Informed consent was obtained from all participants and all procedures were in agreement with the ethical guidelines of the Declaration of Helsinki with approval received

IP T

by the Local Ethical Committee (protocol #: A100349/DS).

SC R

Patients

From September, 2010 to February, 2013, cirrhotic patients without clinical signs of HE, who were followed as outpatients, were evaluated. Only outpatients were considered in order to increase the

U

probability of selecting compensated cirrhotics without overt HE 1,2.

N

The diagnosis of liver cirrhosis was formulated on the basis of one or more of the following

A

parameters: clinical data, laboratory parameters, upper gastrointestinal endoscopy, imaging

M

techniques (ultrasonography), and liver elastography. Liver biopsy was performed only when clinical, laboratory and instrumental exams provided inconclusive data.

ED

The inclusion criteria were: liver cirrhosis without evidence of encephalopathy and no previous

PT

history of overt HE; schooling ≥ 5 years; age >18 years; BMI >18.5 kg/m2, and stable body weight in the previous 3 months. The exclusion criteria were: current or previous presence of overt HE

CC E

(West-Haven criteria) 1,2, current or past alcohol consumption (higher than 20g daily in women and 30g in men) 1,20-22, gastrointestinal bleeding and spontaneous bacterial peritonitis in the previous six weeks, significant comorbidities such as cardiac, respiratory or renal failure; previous transjugular

A

intrahepatic portosystemic shunt, electrolyte imbalance as hyponatremia (Na <125 mg/dl), neurological diseases, not liver-related metabolic encephalopathies, BMI <18.5kg/m2, more than 5% loss of body weight in the last 3 months, illiteracy, color blindness or severe visual disturbances (cataracts, diabetic retinopathy), hepatocellular carcinoma or other malignancies. As concerning the presence of significant comorbidities, 7 patients (5.2 %) suffered diabetes, 13 (9.7%) hypertension, 4

Predictors of hepatic encephalopathy 27 (20.1%) had a BMI ≥25 kg/m2, 21 (15.7%) dyslipidemia, whereas a history of a cardiovascular disease not causing heart failure was present in 11 patients (8.2%). Moreover, no patient had taken lactulose, lactitol, rifaximine or neomycin in the previous 3 months and none of them received antibiotics or psychotropic drugs in the week prior to the study. In addition, we assessed CFF in 150 subjects who were admitted as outpatients for gastrointestinal symptoms and resulted not affected

IP T

by liver diseases to further support the specificity of 39 Hz as cut-off value to recognize individuals with neurophysiological disturbances. Their demographical characteristics (age, 64.0 ± 8.2 years;

SC R

M/F, 105/45; grade of scholarity, 8.8 ± 3.4 years) were similar to those observed in our cirrhotic patients.

U

Study design

N

A total of 181 consecutive cirrhotic patients were examined from September, 2010 to February,

A

2014. Severity of liver disease was evaluated by Child–Pugh and model for end-stage liver disease

M

(MELD) score. Among them, 134 met the eligibility criteria and underwent CFF evaluation along with 150 healthy subjects. Among the 134 patients, using a CFF cut-off value of 39 Hz, 41 CFF

ED

positive (≤39 Hz) and 93 CFF negative (>39 Hz) patients whose demographic characteristics and causes of cirrhosis are described in Table 1, were identified. After the enrollment, all patients

PT

underwent routine physical and clinical examination as well as laboratory analyses (hemogram,

CC E

coagulogram, serum electrolyte levels, renal and liver function tests) and ecotomography every six months. At the same time, they also received a complete neurologic examination and determination of arterial ammonia levels. Finally, an accurate anamnesis was performed in order to establish if the

A

patient was admitted to a hospital for oHE or had developed signs of oHE during the last interval of 6 months. In the latter case, the patient/relative was accurately interviewed to inquire if clear psychomotor (flapping tremor) and/or behavioral (somnolence, disorientation, uninhibited behavior) alterations had previously occurred. To rule out the presence of any major psychiatric disorders, a psychiatric evaluation was performed. Furthermore, patients were invited to come for visitation 5

Predictors of hepatic encephalopathy with a cohabitant that could report any suspect of neurological disturbance. Since the goal of our study was to evaluate the predictive value of CFF on the development of oHE and survival, we did not repeat CFF evaluations during the follow up. The follow up lasted for a median time of 36 months (IQR: 12-47) and ended in June, 2016. On the other hand, the date of death or liver transplantation was obtained by the official records of our Health System in April,

IP T

2017. To know the causes of death, either patient’s relatives or the public register of the National Institute of Statistic were consulted. To calculate the probability of death or transplantation, patients

SC R

were observed for a median time of 49 months (IQR: 18-61). Critical flicker frequency (CFF) assessment

U

CFF was conducted by the Hepatonorm analyzer (HE Flicker Diagnostics, Düsseldorf, Germany).

N

Flicker frequencies were measured 10 times at the same daytime (at afternoon, 15:00-16:00), and

A

the mean value was calculated to avoid little errors due to daytime variability. All evaluations were

M

performed in a dark room after 5 minutes of adaptation 14.

Measurements of the CFF threshold were performed by intrafoveal stimulation with a luminous

ED

diode by decreasing the frequency of light from 60 Hz downward, the CFF threshold was determined as frequency when impression of fused light turned to a flickering one. The CFF was

PT

considered abnormal when the value was ≤39 Hz 5,7,11.

CC E

Statistical analysis

Normal distribution of continuous variables was analyzed by Kolmogorov–Smirnov and Lilliefors

A

tests. Continuous variables were expressed as mean ± standard deviation and comparisons were performed by t-test. Categorical variables were expressed as percentages and comparisons between the two CFF groups were performed by Chi-squared. Pearson’s correlation test was performed to evaluate possible correlations of CFF values with either ammonia levels or Child-Pugh classes or MELD score.

6

Predictors of hepatic encephalopathy Time-to-event outcomes (mortality and first episode of overt HE) were evaluated using the Kaplan– Meier method. The univariate phase of analysis was performed using the Log-rank test to evaluate the association between each covariate and the outcome. The Log-rank test for trend was also used for ordinal and continuous variables. All covariates showing a P-value <0.25 at univariate analysis were included in the multivariate proportional hazards model to estimate any influence on the

IP T

association between CFF group and the outcome variables. In the multivariate phase of our analysis we also assessed the collinearity between parameters that could have a possible influence on the

SC R

outcomes. Both unadjusted and adjusted hazard ratios (HR) with their 95% confidence intervals (95% CI) were estimated for each variable included in the multivariate model. Statistical significance was set at p <0.05.

U

Statistical analysis was performed using the Statistical Analysis Software (SAS Institute Inc., Cary,

A

CC E

PT

ED

M

A

N

NC).

7

Predictors of hepatic encephalopathy RESULTS In our 150 controls, age, sex distribution and scholarity were similar to those observed in the whole cirrhotic population (62.6 ± 9.6 vs. 62.3 ± 9.9 years, p=0.81; M/F 115/35 vs. 97/37, p=0.48; 8.8 ± 3.4 vs. 8.2 ± 3.3 years, p=0.16) and the CFF value was always >39 Hz. Moreover, their CFF mean value was similar to that observed in CFF negative cirrhotic patients (supplementary figure I).

IP T

Demographic and clinical characteristics of CFF positive and CFF negative patients are reported in Table 1. These two groups at baseline were similar for age, sex distribution, scholarity and etiology.

SC R

However, baseline ammonemia and MELD score were significantly higher in CFF positive as compared to CFF negative patients (p=0.002 and p=0.004, respectively) and for this reason, the possible correlation of CFF with either ammonemia or MELD score was evaluated (supplementary

U

figure II). In both cases, the correlation was scarce (r=0.33 and 0.31, respectively). Interestingly, the

N

prevalence of CFF positive patients in the three Child-Pugh classes showed a progressive

A

significant increase with the worsening of the Child-Pugh class (p=0.003) (Figure 1).

M

A total of 19 out of 41 CFF positive patients and 11 out of 93 CFF negative patients developed oHE

ED

during the median follow up period of 36 months (IQR: 12-47). In most cases (80%), the first episode of oHE was the main cause of hospital admission, and the diagnosis of oHE was included

PT

in our medical records during the subsequent visit, at the follow up. In the remaining 6 patients, the diagnosis of oHE was made by ourselves during the follow up visit as reported in materials and

CC E

methods. According to Kaplan-Meier and log-rank test, the likelihood of developing a first episode of overt HE was higher in patients with CFF ≤ 39 Hz as compared to patients with CFF >39 Hz

A

[HR=18.45; CI (7.36-46.3), p <0.001] (Figure 2). At the multivariate analysis, patients with a CFF value ≤ 39 Hz had 7.57 higher risk of oHE as compared to those with a CFF value >39 Hz (95% CI: 3.27 to 17.50), after adjusting for age, Child-Pugh, MELD and serum ammonia (Table 2). Moreover, a modest but highly significant increase of oHE risk [HR 1.02, CI (1.01-1.03); p= 0.0009] was observed at the increase of serum ammonia levels.

8

Predictors of hepatic encephalopathy Finally, the possibility that CFF value could influence mortality risk was considered. As shown in figure 3, patients with a CFF ≤39 Hz had a significantly increased risk of mortality as compared to patients with CFF >39 Hz [HR=2.44, CI (1.03-5.80), p=0.043]. At the multivariate analysis, patients with a CFF value ≤ 39 Hz have a 1.97 higher risk of death compared to those with a CFF value >39 Hz (95%CI: 1.01 to 3.95), after adjusting for age, sex, Child, MELD and serum ammonia

IP T

(Table 3). Moreover, Child-Pugh class was an independent predictor of mortality risk [HR= 3.85 (1.68-8.83), p=0.003].

SC R

Among the patients who died during follow up, the cause of death were: de novo hepatocellular carcinoma (2 CFF ≤39 Hz and 7 CFF >39 Hz), variceal bleeding (2 CFF ≤39 Hz and 4 CFF >39 Hz), liver decompensation (6 CFF ≤39 Hz and 10 CFF >39 Hz), hepatorenal syndrome (2 CFF ≤39

U

Hz and 2 CFF >39 Hz). Furthermore, 2 with CFF ≤39 Hz and 3 with CFF >39 Hz underwent liver

A

CC E

PT

ED

M

A

N

transplantation.

9

Predictors of hepatic encephalopathy DISCUSSION Minimal

hepatic

encephalopathy

(mHE)

is

characterized

by

neuropsychological

and

neurophysiological alterations occurring in cirrhotic patients, in the absence of features of overt encephalopathy. It significantly interferes with the normal functioning of daily routine activity, and

IP T

also impairs health-related quality of life 21-22. This pathological condition also predisposes patients to motor vehicle accidents since it reduces attention skills and delays reaction time 23-25. In addition,

SC R

the presence of mHE increases the risk of developing overt HE 2,11,14.

Specific diagnostic tools, such as the paper-and-pencil tests and its variants, as well as critical flicker frequency (CFF), have been used to recognize mHE. However, the former methods have

N

cultural background, which all limit their reliability 2,5.

U

several drawbacks such as the influence of age, education, training effects, occupation, and socio-

A

On the other hand, CFF, initially used for the evaluation of other neurological diseases (the CFF

M

was devised originally as an ophthalmological test used to measure visual acuity and to screen for optic nerve lesions) 26, has been proved to be a highly reproducible method

ED

suffers a little training effect, daytime variability and low sensitivity

7,14

6,11,12

, although it also

. However, when a cut-off

value of 39 Hz is used for the diagnosis of mHE, its high specificity (94.3-100%)

7

makes this

6,11,12,27

. On the basis of the above mentioned data and considering the values of

CC E

treatments

PT

neurophysiological test particularly useful to establish the efficacy of putative therapeutic

correlation (0.44< r < 0.62) between CFF and psychometric tests reported by others value of r < 0.70 corresponds to an association <50% between the two variables

29

11,14,28

, since a

, we used only

A

CFF to diagnose mHE.

We excluded from our study patients with history of alcohol abuse since a previous study reported conflicting results on CFF mean values in alcohol-related cirrhosis

14

, and considering that an

“alcohol-related brain damage,” has been documented in alcohol abusers 30-32.

10

Predictors of hepatic encephalopathy Using CFF to screen our patients, we found 41 cirrhotics with CFF value ≤39 Hz (CFF positive) and 93 with CFF value >39 Hz (CFF negative). A group of 150 controls without evidence for acute or chronic liver diseases, no psychiatric or neurologic disorders, and no history of alcohol abuse matched for age and sex with our cohort, were also evaluated for critical flicker frequency, showing that all these subjects had CFF values >39 Hz. Both CFF positive and CFF negative cirrhotic

IP T

patients were followed up to estimate the predictive value of CFF in the development of the first episode of overt HE and for mortality occurrence.

SC R

The CFF values measured at the enrollment did not correlate either with MELD or ammonemia. The absence of correlation with the latter parameter was not surprising since other mechanisms different from hyperammonemia have been suggested to have a pathogenetic role in hepatic 33,34

. On the other hand, these results suggest that CFF evaluation represents a

U

encephalopathy

N

predictive parameter that is independent from MELD and ammonia.

A

Interestingly, a significant increase of the prevalence of CFF positivity was observed at the

ED

severity of the disease and CFF.

M

progression from lower to higher Child-Pugh classes, a result supporting a relationship between the

During follow up, as clearly shown by the Kaplan-Meier curve and the log-rank test, the likelihood

PT

of developing a first episode of overt HE was significantly higher in patients with CFF ≤ 39 Hz as compared to CFF >39 Hz. These results were also confirmed when a shorter follow up period of 18

CC E

months was considered (Supplemental Figure III) in order to compare our data with those reported in the literature

11

. Multivariate analysis after adjusting for age, Child-Pugh, MELD and serum

A

ammonia, confirmed previous results demonstrating a HR of 7.57 in CFF ≤ 39 Hz vs. CFF >39 Hz patients. In addition, serum ammonia showed a modest but highly significant increased oHE risk at multivariate analysis. As far as the possibility of CFF to predict survival, conflicting results are reported in the literature 11,18

. Some authors report that the only independent predictor of mortality is Child-Pugh class

11

,

11

Predictors of hepatic encephalopathy others report that older age, CFF ≤39 Hz and MELD are also independently associated with survival, specifying that CFF is superior to PHES in predicting survival 18. Using the Kaplan-Meier curve and the log-rank test, we found an increased risk of mortality in CFF ≤ 39 Hz vs. CFF >39 Hz patients. Moreover, multivariate analysis confirmed previous data reported 18

demonstrating that a low CFF value and Child-Pugh class were both

independent predictors of mortality with a HR of 1.97 and 3.85, respectively.

IP T

by Ampuero et al.

The present study suffers from some limitations: the findings cannot be extended to patients with

SC R

mHE assessed by PHES; the low sensitivity of CFF could represent a problem but this is compensated by a reasonable diagnostic accuracy for minimal HE; possible modifications of quality of life were not evaluated.

U

In conclusion, the present findings demonstrate that CFF is able to predict the development of the

N

first episode of overt HE and represents an independent predictor of survival in prevalently well

A

compensated cirrhotics. This CFF property is probably related to its ability to identify a subclinical

M

sign of liver decompensation. Moreover, since the predictive value of CFF is independent from

ED

MELD and Child-Pugh scores it increases the number of prognostic tools available for clinicians. In consideration of these results, we suggest that the screening of patients with CFF could be useful to

CC E

suitable therapy.

PT

identify a population that deserves a more careful clinical monitoring and/or the initiation of a

Author Agreement: On behalf of the coauthors, the corresponding author certifies that all listed

A

authors participated meaningfully in the study and approved the final version of the manuscript, including the authorship list.

Financial support: none. Disclosures: all authors have nothing to disclose.

12

Predictors of hepatic encephalopathy Conflict of interest: All authors declare no financial, professional, or personal conflict. Conflict of interest statement Manuscript title: CRITICAL FLICKER FREQUENCY TEST PREDICTS OVERT HEPATIC ENCEPHALOPATHY AND SURVIVAL IN PATIENTS WITH LIVER CIRRHOSIS All the mentioned coauthors declare no conflict of interest:

IP T

1. Michele Barone 2. Endrit Shahini

SC R

3. Andrea Iannone 4. Maria Teresa Viggiani

U

5. Valeria Corvace

N

6. Mariabeatrice Principi

ACKNOWLEDGEMENTS

M

A

7. Alfredo Di Leo

A

CC E

PT

ED

We thank Erjon Shahini (software engineer) for statistical analysis support.

13

Predictors of hepatic encephalopathy REFERENCES 1. Ferenci P, Lockwood A, Mullen K, et al. Hepatic encephalopathy-definition, nomenclature, diagnosis, and quantification: final report of the working party at the 11th World Congresses of Gastroenterology, Vienna, 1998. Hepatology 2002;35:716-21.

IP T

2. Vilstrup H, Amodio P, Bajaj J, et al. Hepatic Encephalopathy in Chronic Liver Disease: 2014 practice guideline by the American Association for the Study of Liver Disease and the

European association for the Study of The Liver. AASLD practice guideline. Hepatology

SC R

2014;60:715-35.

3. Butterworth RF. Hepatic encephalopathy: a central neuroinflammatory disorder? Hepatology

U

2011;53:1372-6.

N

4. Skowrońska M, Albrecht J. Alterations of blood brain barrier function in hyperammonemia: an

A

overview. Neurotox Res 2012;21:236-44.

M

5. Torlot FJ, McPhail MJ, Taylor-Robinson SD. Meta-analysis: The diagnostic accuracy of critical

2013;37:527-36.

ED

flicker frequency in minimal hepatic encephalopathy. Alimentary Pharmacol Ther

6. Ampuero J, Macarena S, Montoliù C. et al. Minimal hepatic encephalopathy and critical flicker

PT

frequency are associated with survival of patients with cirrhosis. Gastroenterology 2015;

CC E

149:1483–1489.

7. Kircheis G, Hilger N, Häussinger D. Value of critical flicker frequency and psychometric hepatic

encephalopathy

score

in

diagnosis

of

low-grade

hepatic

encephalopathy.

A

Gastroenterology 2014;146:961-9.

8. Saxena N, Bhatia M, Joshi YK, et al. Electrophysiological and neuropsychological tests for the diagnosis of subclinical hepatic encephalopathy and prediction of overt encephalopathy. Liver 2002;22:190-7. 9. Bajaj JS, Hafeezullah M, Franco J, et al. Inhibitory control test for the diagnosis of minimal 14

Predictors of hepatic encephalopathy hepatic encephalopathy. Gastroenterology 2008;135:1591-1600. 10. Randolph C, Hilsabeck R, Kato A, et al. Neuropsychological assessment of hepatic encephalopathy: ISHEN practice guidelines. Liver Int 2009;29:629-35. 11. Romero-Gómez M, Cordoba J, Jover R, et al. Value of the critical flicker frequency in patients with minimal hepatic encephalopathy. Hepatology 2007;45:879–85.

IP T

12. Lunia MK, Sharma BC, Sharma P, et al. Probiotics prevent hepatic encephalopathy in patients with cirrhosis: a randomized controlled trial. Clin Gastroenterol Hepatol 2014;12:1003-8.

SC R

13. Dhiman RK, Rana B, Agrawal S, et al. Probiotic VSL#3 reduces liver disease severity and hospitalization in patients with cirrhosis: a randomized, controlled trial. Gastroenterology 2014;147:1327-37.

U

14. Kircheis G, Wettstein M, Timmermann L, et al. Critical flicker frequency for quantification of

N

low grade HE. Hepatology 2002;35:357-66.

A

15. Sharma P, Sharma BC, Puri V, et al. Critical flicker frequency: diagnostic tool for minimal

M

hepatic encephalopathy. J Hepatol 2007;47:67-73.

ED

16. Singh J, Sharma BC, Puri V, et al. Sleep disturbances in patients of liver cirrhosis with minimal hepatic encephalopathy before and after lactulose therapy. Metab Brain Dis. 2017

PT

Apr;32:595-605.

17. Moratalla A, Ampuero J, Bellot P et al. Lactulose reduces bacterial DNA translocation, which

CC E

worsens neurocognitive shape in cirrhotic patients with minimal hepatic encephalopathy. Liver Int. 2017 Feb;37:212-23.

A

18. Ampuero J, Simon M, Montoliu C, et al. Minimal hepatic encephalopathy and critical flicker frequency are associated with survival of patients with cirrhosis. Gastroenterology 2015;149: 1483-9. 19. Kircheis G, Bode JG, Hilger N, et al. Diagnostic and prognostic values of critical flicker frequency determination as new diagnostic tool for objective HE evaluation in patients undergoing TIPS implantation. Eur J Gastroenterol Hepatol 2009;21:1383-94. 15

Predictors of hepatic encephalopathy 20. Berlioux P, Robic MA, Poirson H, et al. Pre-transjugular intrahepatic portosystemic shunts (TIPS) prediction of post-TIPS overt hepatic encephalopathy: the Critical Flicker Frequency is more accurate than psychometric tests. Hepatology 2014;59:622-9. 21. Arguedas MR, DeLawrence TG, McGuire BM. Influence of hepatic encephalopathy on health-related quality of life in patients with cirrhosis. Dig Dis Sci 2003;48:1622–6.

IP T

22. Groeneweg M, Quero JC, De Bruijn I, et al. Subclinical hepatic encephalopathy impairs daily functioning. Hepatology 1998;28:45-9.

SC R

23. Wein C, Koch H, Popp B, et al. Minimal hepatic encephalopathy impairs fitness to drive. Hepatology 2004;39:739-45.

24. Bajaj JS, Hafeezullah M, Hoffmann RG, et al. Navigation skill impairment: another

U

dimension of the driving difficulties in minimal hepatic encephalopathy. Hepatology

N

2008;47:596-604.

M

Gastroenterology 2009;137:1706-15.

A

25. Kircheis G, Knoche A, Hilger N, et al. Hepatic encephalopathy and fitness to drive.

ED

26. Baatz H, Raak P, De Ortueta D, et al. Practical significance of critical fusion frequency (CFF). Chronological resolution of the visual system in differential diagnosis. Ophthalmologe

PT

2010;107:715-9.

27. Sharma K, Pant S, Misra S, et al. Effect of rifaximin, probiotics, and l-ornithine l-aspartate on hepatic

CC E

minimal

encephalopathy:

a

randomized

controlled

trial.

Saudi

J

Gastroenterol 2014;20:225-32.

A

28. Ozel Coşkun BD, Ozen M. Critical flicker frequency test for diagnosing minimal hepatic encephalopathy in patients with cirrhosis. Turk J Gastroenterol 2017;28:191-196.

29. Leandro G. Bias in metanalytical research. In: Leandro G, ed. Meta-analysis in Medical Research. The handbook for the understanding and practice of meta-analysis. Malden, MA: Blackwell Publishing, Inc.; 2005:15-20. 30. Barone M. Neurophysiologic and psychometric tests in the diagnosis of low-grade 16

Predictors of hepatic encephalopathy encephalopathy. Gastroenterology 2015;148:663. 31. Sutherland GT, Sheahan PJ, Matthews J, et al. The effects of chronic alcoholism on cell proliferation in the human brain. Exp Neurol 2013;247:9-18. 32. Greg T Sutherland, Donna Sheedy, and Jillian J Kril. Using autopsy brain tissue to study alcohol-related brain damage in the genomic age. Alcohol Clin Exp Res. 2014;38:1–8.

IP T

33. Parekh PJ, Balart LA. Ammonia and Its Role in the Pathogenesis of Hepatic Encephalopathy. Clin Liver Dis 2015;19:529-537.

SC R

34. Lockwood AH. Blood ammonia levels and hepatic encephalopathy. Metab Brain Dis

A

CC E

PT

ED

M

A

N

U

2004;19:345-349.

17

Predictors of hepatic encephalopathy FIGURE LEGENDS Figure 1. Prevalence of CFF positive patients in the different Child-Pugh classes.

PT

ED

M

A

N

U

SC R

IP T

p=0.003 by Chi-squared.

CC E

Figure 2. Probability of developing the first episode of overt HE in patients with normal and altered CFF.

A

Data were obtained by the comparison of 41 CFF positive and 93 CFF negative patients. Statistical analysis by Log-rank test demonstrated a significant difference between the two groups (p <0.001).

18

M

A

N

U

SC R

IP T

Predictors of hepatic encephalopathy

ED

Figure 3. Survival in patients with normal and altered CFF. Data were obtained by the comparison of 41 CFF positive and 93 CFF negative patients.

PT

Statistical analysis, by log-rank test, demonstrated a significant difference between the

A

CC E

two groups (p =0.043).

19

A

CC E

PT

ED

M

A

N

U

SC R

IP T

Predictors of hepatic encephalopathy

20

Predictors of hepatic encephalopathy Table 1. Demographic characteristics and causes of cirrhosis in CFF positive and CFF negative patients. CFF negative (>39 Hz)

p value

41 62.2 ± 9.7 29/12

93 62.4 ± 10.0 68/25

0.69 0.76

17 (41.5)

48 (51.6 )

HBV

9 (21.9)

11 (11.8)

HBV- HCV

3 (7.3)

2 (2.1)

HBV- HDV

1 (2.43)

4 (4.3)

1.0

NASH

8 (19.51)

15 (16.1)

0.54

AIH

1 (2.43)

4 (4.3)

1.0

PSC

1 (2.43)

3 (3.22)

1.0

PBC

0 1 (2.43)

0.13 0.17

3 (3.22)

0.55

3 (3.22)

1.0

20 (48.8)

71 (76.34)

0.001 ‡

B

18 (43.9)

21 (22.6)

0.012 ‡

3 (7.3)

1 (1.1)

0.08

11.5 ± 5.5

9.0 ± 2.5

0.004*

41.8 ± 18.6

35.4 ± 25.8

0.002*

20 (48.8)

46 (49.5)

1.0

20 (21.5)

0.24

M

A

A

Child-Pugh class

0.75

SC R

N

Cryptogenetic

U

N. of subjects Age, mean ± SD (years) Sex distribution (M/F) Etiology HCV

IP T

CFF positive (≤ 39 Hz)

C

ED

MELD score Ammonemia (µg/dl) Splenomegaly

PT

Mild ascites

5 (12.2) 5 (50)

13 (52)

1.0

Esophageal varices F2

4 (40)

11 (44)

1.0

1 (10)

1 (4)

1.0

CC E

Esophageal varices F1 Esophageal varices F3

A

Data reported were expressed as absolute number and relative percentage (in parenthesis) or mean ± standard deviation. HCV: hepatitis C virus; HBV: hepatitis B virus, HDV: hepatitis D virus, NASH: non-alcoholic steatohepatitis, AIH: autoimmune hepatitis, PSC: primary sclerosing cholangitis, PBC: primary biliary cholangitis. ‡

by Chi-squared test; *by t-test.

21

Predictors of hepatic encephalopathy Table 2. Univariate and multivariate analysis of the variables associated to overt HE.

Univariate analysis

Multivariate analysis

Variable HR (95% CI)

P value

Age

0.99 (0.96-1.03)

1.00 (0.96-1.04)

0.90

Child-Pugh score

3.67 (2.00-6.72)

1.37 (0.49-3.79)

0.55

MELD score

1.17 (1.10-1.23)

1.04 (0.93-1.15)

0.52

Serum Ammonia

1.02 (1.01-1.03)

1.02 (1.01-1.03)

0.0009

SC R

IP T

HR (95% CI)

<0.0001

CFF 1.00*

1.00*

≤ 39 Hz

7.59 (3.59-16.06)

7.57 (3.27-17.50)

N

U

> 39 Hz

A

CC E

PT

ED

M

A

*Reference group.

22

Predictors of hepatic encephalopathy Table 3. Univariate and multivariate analysis of the variables associated with mortality.

Univariate analysis

Multivariate analysis

Variable HR (95% CI)

P value

1.01 (0.97-1.04)

1.03 (1.00-1.07)

0.08

IP T

Age (years)

HR (95% CI)

0.14

Sex 1.00*

0.61 (0.27-1.40)

0.54 (0.23-1.27)

Child-Pugh score

2.56 (1.42-4.61)

3.85 (1.68-8.83)

0.003

MELD score

1.03 (0.97-1.10)

0.91 (0.81-1.05)

0.06

Serum Ammonia

1.01 (1.0-1.02)

1.01 (1.00-1.02)

0.30

U

Female

SC R

1.00*

Male

N

CFF

1.00*

≤ 39 Hz

0.91 (0.85-0.98)

A

1.00*

1.97 (1.01-3.95)

M

> 39 Hz

0.049

A

CC E

PT

ED

*Reference group.

23