Telephone: 312-503-3453; FAX: 312-908-6192
Article first published online: 28 JAN 2009
Copyright © 2009 American Association for the Study of Liver Diseases
Volume 15, Issue 2, pages 184–192, February 2009
How to Cite
Sotil, E. U., Gottstein, J., Ayala, E., Randolph, C. and Blei, A. T. (2009), Impact of preoperative overt hepatic encephalopathy on neurocognitive function after liver transplantation. Liver Transpl, 15: 184–192. doi: 10.1002/lt.21593
See Editorial on Page 121
Presented at the 42nd Annual Meeting of the European Association for the Study of the Liver, April 11-15, 2007, Barcelona, Spain, and published in abstract form [J Hepatol 2007;46(suppl 1):S35-S36].
- Issue published online: 28 JAN 2009
- Article first published online: 28 JAN 2009
- Manuscript Accepted: 12 JUN 2008
- Manuscript Received: 15 APR 2008
- AASLD Transplant Hepatology Fellowship grant to E.U.S
In the current Model for End-Stage Liver Disease allocation system, patients are at risk of suffering repeated episodes of hepatic encephalopathy (HE) while waiting for an orthotopic liver transplantation (OLT); the posttransplantation impact of these episodes has not been well explored. We evaluated the cognitive function and quality of life in a group of OLT recipients (n = 25) who had suffered from overt HE prior to their procedure (HE-PreLT group) and compared their performance to that of a similar group of patients (n = 14) without overt HE (No HE-PreLT group) as well as to controls. Patients were selected from a cohort of 280 patients who underwent OLT during this period; the presence of clinical confounders excluded many of the remaining subjects. Demographic and clinical characteristics were balanced among groups. At an average of 18 months after OLT, we administered 2 neuropsychological batteries [Psychometric Hepatic Encephalopathy Score (PHES) test battery and Repeatable Battery for the Assessment of Neuropsychological Status (RBANS)]; a pyschophysiological test (critical flicker frequency); and the SF-36 quality of life score. The HE-PreLT group scored below controls in 5 of 6 cognitive domains tested by RBANS, 3 of 6 PHES subtests, as well as the critical flicker frequency test. The No HE-PreLT group scored below the controls in 1 of the 6 cognitive domains tested by RBANS. The more severe neurocognitive abnormalities seen in the HE-PreLT group did not appear to affect quality of life, as lower values than normative data were only found in 1 of the 8 SF-36 scales. In conclusion, neurocognitive abnormalities were more severe in liver transplant recipients that had suffered from overt HE prior to OLT. Prospective studies of neurocognitive function pre-OLT and post-OLT are needed to fully determine the impact of such abnormalities. Liver Transpl 15:184–192, 2009. © 2009 AASLD.
Liver allograft allocation changed in the United States with implementation of the Model for End-Stage Liver Disease (MELD)-based system in February 2002 (www.unos.org). As a result, hepatic encephalopathy (HE) ceased to be used as a criterion to prioritize patients on the transplant list. In a large cohort of patients with cirrhosis, the presence of HE did not independently influence short-term survival.1 Elimination of subjective criteria to stage mental abnormalities was viewed as an important goal of the new classification system. However, this change has come at a cost for subjects with HE. Patients with liver failure and episodic encephalopathy, the most common clinical manifestation of HE, are at risk of developing repeated bouts of encephalopathy. In a series of 200 patients with cirrhosis related to hepatitis C, recurrent encephalopathy was the leading cause of hospital readmission.2
In view of this new allocation paradigm, we decided to examine whether patients awaiting orthotopic liver transplantation (OLT) who develop bouts of HE before the procedure confront a different posttransplantation course. Neurological complications remain a common problem after OLT,3 affecting as many as 13%-47% of transplant recipients.4 Cognitive dysfunction can be detected using neuropsychological and neurophysiological testing.5–17 These alterations could arise as a consequence of prior HE but multiple variables may influence neurological function in this population. The role of anoxic intraoperative complications, immunosuppressive medication, comorbidities,3 and deficiencies of other neuroactive agents, such as thiamine,18 should be considered.
Ideally, patients would undergo a detailed neurological evaluation prior to transplantation and be followed longitudinally after OLT. A pretransplantation evaluation is particularly difficult in patients with episodic encephalopathy, in whom acute events preclude the performance of specific tests. Limited studies in patients with minimal encephalopathy after OLT suggest improvement, but not complete normalization, of neurological function.6, 8, 17 In the present study, we examined subjects who had undergone OLT approximately 1.5 years prior to testing. We studied a cohort with similar confounding variables and divided the patients according to the presence of prior documented admissions for episodic HE. We administered 2 batteries of neuropsychological tests, 1 psychophysiological evaluation, and an evaluation of quality of life (QOL). Our results suggest that episodes of HE pre-OLT are associated with more neurocognitive abnormalities 1.5 years after the procedure.
PATIENTS AND METHODS
The study was carried out at the outpatient Post–Liver Transplantation Clinic of a tertiary referral hospital between December 1, 2005 and August 30, 2006. The protocol was approved by the Institutional Review Board and all participants provided written informed consent. Participants (n = 39) between 18 and 70 years of age with a history of cirrhosis who had received a liver transplant between July 1, 2003 and December 31, 2005, and had had regular postoperative follow-up were included in the study. Over this period of time, a total of 280 liver transplantations were performed at our institution. Twenty healthy controls were also included.
Twenty-five participants had a history of overt HE prior to OLT that required the use of ammonia-lowering medications (HE-PreLT group) and 14 had no history of overt HE prior to OLT (No HE-PreLT group). The HE-PreLT patients had inpatient or outpatient documentation of mental status changes and/or asterixis, and had been treated with ammonia-lowering medications (lactulose, neomycin, metronidazole, or rifaximin) prior to OLT. Participants who had received treatment for HE but for whom no clear documentation of HE was found upon review of the medical records were not included in the HE-PreLT group. One participant with no clear documentation of HE that was taken off ammonia-lowering treatment and did not suffer any episodes of overt HE was included in the No HE-PreLT group.
Liver transplant recipients were excluded from the study if they met any of the following criteria: presence of neurological or major psychiatric conditions; use of psychoactive medications within a month prior to testing; regular alcohol use within 6 months prior to testing; recurrence of cirrhosis or significant allograft dysfunction; decompensated pulmonary disease, cardiac disease, or renal failure; decompensated diabetes mellitus; or significant ophthalmological disease that precluded completion of neuropsychological testing.
The healthy control group was comprised of 20 subjects similar in age to the liver transplant recipients, who did not suffer from any chronic medical condition that could affect cognitive or motor function (including hypertension and diabetes mellitus). Control subjects with a history of conditions with no effect on neurological or motor function, such as osteoporosis or gastroesophageal reflux disease, were allowed to participate in the study. Control subjects were excluded if they had taken any psychoactive medications within 1 month prior to testing or if alcohol consumption exceeded 2 alcoholic drinks per day.
Demographic characteristics (age, gender, education level, native English speaker status, and employment status) and clinical characteristics (etiology of liver disease leading to transplantation, MELD score prior to OLT, date of transplant, type of graft received, comorbid conditions, and current medication use including immunosuppressant medications) were obtained from questionnaires administered to the participants and from review of the medical records.
Two neuropsychological test batteries, the Psychometric Hepatic Encephalopathy Score (PHES) and the Repeatable Battery for the Assessment of Neuropsychological Status (RBANS), were used in this study.
PHES Test Battery
PHES scores a neuropsychological battery composed of five different tests: Number Connection Tests A (NCT-A) and B (NCT-B), Line Tracing test, Serial Dotting test, and Digit Symbol test.19 The tests detect changes in attention and psychomotor speed, areas most affected by HE. A PHES score is obtained based on sum of the results of each test. While validation of this approach by other centers is awaited, its use in research has been temporarily recommended by an expert panel on hepatic encephalopathy for assessment of minimal HE.20 As U.S. population normative data for PHES scores is not available, the PHES test battery was administered to both groups of patients and to controls.
RBANS is a validated battery of neuropsychological tests designed to detect mild cognitive dysfunction.21 It can be administered in 30 minutes and has good correlation with longer psychological tests such as the Wechsler Adult Intelligence Scale-Third Edition and Wechsler Memory Scale-Third Edition.22 It consists of a total scale and 5 subscales that test different neuropsychological domains. Each subscale is obtained from the score of 2 or more subtests. The subtests administered are as follows: List Learning and Story Memory for Immediate Memory subscale; Figure Copy and Line Orientation for Visuospatial/Constructional subscale; Digit Coding and Digit Span for Attention subscale; Picture Naming and Semantic Fluency for Language subscale; and List Recall, Story Recall, List Recognition, and Figure Recall for Delayed Memory subscale. RBANS raw scores are compared to normative data by decade of life obtained from a large U.S. population sample and transformed into a standard score. As U.S. normative data is available, RBANS was administered to both groups of patients but not to controls.
RBANS is a sensitive test for the detection of mild cognitive dysfunction in patients with psychiatric conditions such as schizophrenia22 and neurological conditions such as stroke,23 Alzheimer's, and Huntington's dementia.21 By assessing several neuropsychological domains, this battery is able to evaluate which areas of the brain are most affected by cognitive dysfunction and can categorize cognitive dysfunction as either cortical or subcortical.21 While its capacity to detect mild cognitive dysfunction and its ability to evaluate several neuropsychological domains makes it an ideal tool for use in the evaluation of mild hepatic encephalopathy, it has only been used for the assessment of cognitive dysfunction in end-stage liver disease in a few studies.24, 25
Critical flicker frequency (CFF), is a psychophysiological test that has recently been reported to be both sensitive and specific for the detection of overt and minimal HE in patients with cirrhosis.26 We administered the test to both groups of patients and the control group. CFF reflects the efficiency of the visual apparatus as well as the arousal capacity of the cerebral cortex.26, 27 It has been shown to correlate with a number of neuropsychological tests that assess attention, concentration, visuopraxis, and psychomotor speed, areas of cognition frequently affected by HE.26 A portable apparatus (Hepatonorm Analyzer; R&R MEdi-Business, Freiburg, Germany) was used. A luminous diode with a wavelength of 650 nm, passing through a concave-convex lens system, is used for intrafoveal stimulation. Starting with a frequency of light pulses at 60 Hz, the frequency at which the fused light is perceived as a flickering light is recorded as the CFF threshold.26 The test was performed during daylight hours in a quiet room, using a mean of 10 measurements for a final CFF value.
Quality of Life
Liver transplant recipients completed the Medical Outcomes SF-36 Short Version, a 36-item self-administered questionnaire. SF-36, a commonly used QOL measurement instrument,28 provides 2 summary scores: physical component and mental component; and 8 health status scales: Physical Functioning, Role Limitations due to Physical Problems, Bodily Pain, General Health Perceptions, Vitality, Social Functioning, Role Limitations due to Emotional Problems, and Mental Health (SF-36 Health Survey Manual and Interpretation Guide; QualityMetric, Lincoln, RI)
Diagnosis of Depression
Depression is a common disorder in liver transplantation patients before OLT29 and to a lesser degree after OLT.14 Cognitive symptoms (diminished ability to concentrate or think) are one of the Diagnostic and Statistical Manual of Mental Disorders, Fourth Edition diagnostic criteria for a major depressive episode. As the presence of depression-related cognitive symptoms could affect the results of the neuropsychological tests, a depression scale, the Beck Depression Inventory, second edition (BDI-II)30 was administered to both groups of liver transplant recipients. The BDI-II is a 21-item self-reporting questionnaire that can identify patients as having mild, moderate, or severe depression.
Patients seen at our institution's post–liver transplantation clinic were invited to participate in the study. All participants provided informed consent prior to their participation. All tests were performed during a single 1-hour to 2-hour ambulatory session scheduled at the participant's convenience.
SPSS version 12.0 software was used for data analysis. Continuous variables are reported as mean ± standard deviation. Categorical variables were reported as number of participants and percentages. Chi-square and Fisher's exact test were used to evaluate categorical data. Kolmogorov-Smirnov Z was used to determine normality. For 3-group comparison analysis of variance (ANOVA), analysis of covariance (ANCOVA), and Bonferroni correction in post hoc analysis were used when the continuous variables were normally distributed and had equal variances, while Welch-ANOVA, Welch-ANCOVA, and Tamhane's T2 for post hoc analysis were used for continuous variables normally distributed with unequal variances. Student t test was used for continuous variable 2-group comparison. Comparison with normative values for RBANS and SF-36 was performed with 1 sample T test. Kruskal-Wallis and Wilcoxon rank sum were used for nonnormally distributed continuous variables. P values below 0.05 with correction for multiple testing, if applicable, were considered statistically significant.
Demographic and Clinical Characteristics
Participants' demographic characteristics are shown in Table 1. There were no differences with respect to age, education level, or percentage of native English speakers between both groups of patients and controls. There was a statistically significant difference in gender between both group of patients and controls, with 30% of subjects being male in the control group versus 76% and 64% in the HE-PreLT group and the No HE-PreLT group, respectively (chi-square, P < 0.05). Education level was similar in all subjects with most of them having completed at least 1 year of college. Only 3 patients in the HE-PreLT group were nonnative English speakers.
|HE-PreLT (n = 25)||No HE-PreLT (n = 14)||Controls (n = 20)|
|Age (years)||57 ± 8||51 ± 11||52 ± 9|
|Education by group|
|<High school||1 (4%)||0||0|
|High school||5 (20%)||3 (21%)||2 (10%)|
|>High school||19 (76%)||11 (79%)||18 (90%)|
|Education level‡||11.16 ± 2.36||11.77 ± 1.88||12.25 ± 1.80|
|Native English speaker||22 (88%)||14 (100%)||20 (100%)|
Comparison of clinical characteristics between the HE-PreLT group and the No HE-PreLT group (Table 2) revealed no statistically significant differences in etiology of liver disease leading to transplantation, type of graft received, time from OLT to testing date, use of immunosuppressant medication, or presence of pharmacologically-treated hypertension or diabetes mellitus. Both groups of patients differed in regard to MELD score (without exception points) prior to transplantation. The HE-PreLT group's mean MELD score was 24.4 ± 8.9 while the No HE-PreLT group's MELD score was 16.4 ± 9.8 (Students t test, P = 0.013). There was a tendency toward a higher number of patients who received combined liver/kidney transplantation in the HE-PreLT group (24% versus 0%) but this did not reach statistical significance (Fisher's exact test, P = 0.071).
|HE-PreLT (n = 25)||No HE-PreLT (n = 14)|
|HCV||9 (36%)||4 (29%)|
|HCC||3 (12%)||3 (21%)|
|ETOH||9 (36%)||2 (14%)|
|PSC/PBC||2 (8%)||5 (36%)|
|Others||7 (28%)||5 (36%)|
|MELD score (without exception points)||24.4 ± 8.9†||16.4 ± 9.8|
|Type of graft|
|Cadaveric OLT||15 (60%)||11 (79%)|
|OLT + kidney transplant||6 (24%)||0|
|LDLT||4 (16%)||3 (21%)|
|Months from OLT to testing (range)||17 ± 8.1 (6-31.8)||18.7 ± 11 (6-33.6)|
|Tacrolimus||12 (48%)||9 (64%)|
|Tacrolimus/MMF||10 (40%)||5 (36%)|
|Pharmacologically treated diabetes||2 (8%)||3 (21%)|
|Pharmacologically treated HTN||8 (32%)||4 (29%)|
|Pharmacologically treated diabetes and HTN||4 (16%)||1 (7%)|
|Other capacity (self)||8%/8%||7%/7%|
|BDI-II total score||6.2 ± 6.0||5.6 ± 5.9|
|Mild depression||5 (20%)||0|
|Moderate depression||2 (8%)||2 (14%)|
The BDI-II was administered to both groups of transplant recipients to ensure that cognitive function was not affected by depression. In the HE-PreLT group, 28% of patients showed mild or moderate depression versus 14% in the No HE-PreLT group (chi-square, P = not significant). No patient in either group had severe depression.
Using ANOVA, a significant difference in 3 out of the 6 tests that compose PHES was found (NCT-A, NCT-B, and Digit Symbol, all P < 0.001) between the 3 groups (Fig. 1 and Table 3). Post hoc analysis (Tamhane's T2 or Bonferroni) revealed that Digit Symbol, NCT-A, and NCT-B scores were significantly worse in the HE-PreLT group versus controls. The HE-PreLT group also performed significantly worse in comparison to the No HE-PreLT group in NCT-A and Digit Symbol subtests. Composite PHES score could not be calculated, as normative values are only available for German and Spanish populations. Age can significantly affect neuropsychological tests scores. While no significant differences in age were found between the groups, the data was analyzed with ANCOVA, taking age and gender as covariates. The results did not change and the statistical significance between the groups on post hoc analysis was maintained. ANCOVA using MELD as a covariate revealed no effect of pre-LT MELD score on PHES test results.
|Domains||HE-PreLT (n = 25)||No HE-PreLT (n = 14)||Controls (n = 20)|
|NCT-A (seconds)||34.0 ± 8.3*,†||23.3 ± 8.4||19.6 ± 3.9|
|NCT-B (seconds)||98.4 ± 30.5*||76.0 ± 35.1||54.5 ± 17.0|
|Digit symbol (points)||41.2 ± 8.9*,‡||50.4 ± 9.8||54.6 ± 8.4|
|Serial dotting (seconds)||61.3 ± 25.2||59.3 ± 20.1||54.8 ± 18.1|
|Line tracing (seconds)||77.0 ± 22.7||78.1 ± 18.2||70.7 ± 26.2|
|Line tracing (errors)||11.0 ± 11.9||10.4 ± 15.0||5.4 ± 4.9|
The patients in the HE-PreLT group had cognitive dysfunction in 4 out of the 5 domains tested in RBANS as well as the RBANS total score when compared to normative data (Fig. 2). The patients in the No HE-PreLT group only showed cognitive abnormalities in one domain but the total score was significantly lower when compared to normative data (Fig, 2). When the HE Pre-OLT group was compared to the No HE Pre-LT group, no statistically significant differences were found in the domains evaluated by RBANS or in the total score (Table 4).
|Domains||HE-PreLT (n = 25)||No HE-PreLT (n = 14)|
|Immediate memory||86.4 ± 14.8||95.43 ± 17.4|
|Visuospatial/constructional||93.1 ± 13.2||90.6 ± 16.4|
|Language||92.7 ± 7.9||93.0 ± 10.5|
|Attention||93.9 ± 15.7||98.9 ± 10.7|
|Delayed memory||94.3 ± 11.2||95.9 ± 9.2|
|RBANS total scale||88.7 ± 10.1||92.5 ± 11.3|
|RBANS total percentile||26.5 ± 18.3||35.1 ± 21.3|
Language skills can significantly affect neuropsychological test results. Analysis of the scores of both neuropsychological test batteries excluding the 3 nonnative English speaker participants did not change the results.
Twenty percent of participants saw 2 lights when CFF was administered (32% of the HE-PreLT group, 14% of the No HE-PreLT group, and 10% of controls). Only data from those participants that saw 1 light was analyzed.
There were statistically significant differences between the groups with the use of ANOVA (P = 0.001): HE-PreLT group, 40.1 ± 2.9; No HE-PreLT group, 40.8 ± 2.2; and controls, 43.6 ± 3.0. Post hoc analysis revealed differences between both groups of patients and controls. When the data was adjusted for age, there were still statistically significant differences between the groups (ANCOVA, P < 0.05). Post hoc analysis (Bonferroni) adjusted for age revealed that only the HE-PreLT group had a lower CFF value than controls (Fig. 3).
Using the SF-36 questionnaire, we did not find significant differences in QOL between the HE-PreLT group and the No HE-PreLT group in any of the 8 health status scales or the composite mental and physical scores (Table 5).
|HE-PreLT (n = 25)||No HE-PreLT (n = 14)|
|PF||72.7 ± 24.8||76.1 ± 19.7|
|RP||53.1 ± 39.9||64.3 ± 37.7|
|BP||69.0 ± 25.9||68.4 ± 18.0|
|GH||71.3 ± 17.4||66.2 ± 17.4|
|VT||67.9 ± 18.4||64.8 ± 19.9|
|SF||80.2 ± 24.7||79.1 ± 23.8|
|RE||79.2 ± 35.2||87.2 ± 27.9|
|MH||76.0 ± 18.6||78.3 ± 21.4|
|SF-36 PCS||44.0 ± 11.1||48.2 ± 13.2|
|SF-36 MCS||52.9 ± 9.1||53.2 ± 10.0|
When each group's values were compared with normative values (Fig. 4A,B), the HE-PreLT group scored lower than the normative values (adjusted by decade of life) in the Role Limitations due to Physical Problems scale (1-sample Student t test, P < 0.05). This scale evaluates problems with work or other daily activities as a result of physical health (SF-36 Health Survey Manual and Interpretation Guide).
In the present study, we tested a cohort of subjects who had undergone OLT an average of approximately 1.5 years prior to testing. The cohort represents approximately 15% of subjects undergoing OLT at our institution during that period; many other patients met exclusion criteria for entry into the study. Our results reveal the presence of more pronounced cognitive dysfunction after OLT among patients with overt HE prior to OLT.
Alterations in neurocognitive function have been previously reported in patients at varying intervals after OLT.5–17 In only 3 of 19 published studies,10, 12, 13 the presence of overt HE prior to OLT was reported, a factor whose influence on postoperative results was not analyzed. Several variables could influence neurocognitive function after OLT, including intraoperative complications, the development of cerebral pontine myelinolysis, the use of immunosuppressive medication, and severe recurrence of the original liver disease.3 We were interested in the influence of overt HE on postoperative neurocognitive function for 2 reasons. First, the allocation of organs using the MELD score has increased the number of patients with severe liver failure coming to OLT, with overt HE being the cause of multiple admissions.2 Second, long-lasting effects of overt HE on brain function would have an important implication for organ allocation as well as raise conceptual questions regarding the pathogenesis of HE.
The 2 groups of transplantation patients had similar baseline and clinical characteristics. Both were on similar immunosuppressive regimens, were tested at equivalent times after OLT, and suffered from similar comorbidities, such as hepatitis C, hypertension, and diabetes. The difference in MELD score between groups was evaluated via covariate analysis and did not appear to affect the overall evaluation of neurocognitive function.
Two sets of neuropsychological batteries were administered. The PHES battery was proposed as a useful tool to evaluate patients with cirrhosis,31 as it focuses on alterations of attention and psychomotor function, the main psychological domains affected by HE.32 The other, RBANS, has been widely used to assess neuropsychological status in multiple medical conditions,21 though it is seldom used in patients with liver disease.24, 25 The PHES battery has the advantage of using tests widely employed in patients prior to OLT, though in the absence of U.S. normative values, its results were also compared to a control group of age-matched individuals. RBANS has well-characterized normative values, based on a group of more than 600 normal subjects of varying ages. It tests 5 different cognitive domains, with the test interpretation based on specific domain alterations and a total score.
In our study, the results of both batteries can be viewed as complementary. Patients with HE prior to OLT had worse performance of the PHES battery in 2 of 6 tests when compared to subjects without HE; the abnormal tests (NCT-A and Digit Symbol) mainly explore the attention domain, while the tests of motor skill (Line Tracing and Serial Dotting) showed similar values. The RBANS testing did not show differences between the 2 groups, though scores in the attention domain and overall scores were lower in the HE-PreLT group. These differences became clearer when we compared each battery to normal values. In the case of PHES, 3 of 6 tests (the 2 previous ones and NCT-B) were significantly better in controls than in the HE-PreLT group. In the case of RBANS, 4 of 5 domains in the HE-PreLT group had significant lower performance than normative controls; 1 of 5 domains (language) was different in the No HE-PreLT group as compared to normative controls.
The results of CFF, a psychophysiological test that explores the intactness of the visual pathway and the overall function of the cerebral cortex, also support the notion of a higher severity of cognitive dysfunction among the HE-PreLT group. This group, but not the No HE-PreLT group, exhibited significantly lower CFF values than the control group. While the abnormal cutoff CCF value for patients with cirrhosis has been proposed as 39 Hz,26 the mean CFF value in the HE-PreLT group was 40 Hz. A note of caution is due in the interpretation of these results. A total of 20% of the cohort saw 2 lights (rather than 1) when exposed to the stimulus, preventing performance of the test. The reasons for this “double vision” are unclear but may reflect problems with accommodation to the light stimulus.
Using the SF-36, a frequently-used QOL questionnaire, we were unable to demonstrate an effect for the cognitive alterations found in the OLT recipients on their QOL. Only the Role Limitations due to Physical Problems scale, a scale that measures the perception of problems with work or other daily activities as a result of physical health, was altered in the HE-PreLT group as compared to normative controls, while no differences were seen between the HE-PreLT group and the No HE-PreLT group. Our results are in disagreement with a recently published systematic review of assessment of QOL following liver transplantation,33 which concluded that while there is significant improvement in QOL following the procedure, deficits persist compared to the general population. Our study population was selected by strict inclusion and exclusion criteria to ensure reliable neuropsychological testing, and represented only 15% of the patients undergoing transplantation at our institution. A selection bias toward a healthier group of transplant recipients was likely introduced in the sample and may explain our QOL findings. Nonetheless, it appears that the degree of cognitive dysfunction that was seen in our patients did not significantly impact the measurements of QOL.
Our suggestion that the presence of overt HE-PreLT may affect neurocognitive function after OLT is limited by 4 factors. First, the cohort studied is relatively small, allowing for the presence of unpredictable confounders; still, we excluded a large number of OLT patients from the study due to of the presence of such confounders. Second, it is possible that the No HE-PreLT group could have suffered from minimal HE at the time of the procedure, a condition that can affect up to 50% of patients with advanced cirrhosis.34 However, the presence of minimal HE would have potentially reduced the differences between groups and made the demonstration of intergroup differences more difficult. Third, the overt HE-PreLT group had a higher MELD score prior to transplantation, indicating a higher severity of liver failure prior to OLT. While we attempted to correct for this with the use of ANCOVA with MELD as a covariate, it is possible that liver failure per se could independently explain some of the differences in neurocognitive function found in the study. Finally, we studied patients at a single point in time after OLT. Some longitudinal studies have noted a progressive improvement in neurocognitive function during the first year after OLT.14, 16 However, in a large evaluation, no differences were seen on neuropsychological testing at 1 versus 3 years post-OLT.14
The question raised by this study is an important one. If indeed episodes of overt HE before OLT affect neurocognitive function beyond the performance of the OLT, the wisdom of omitting HE from the algorithm of organ allocation needs to be reconsidered. From a pathophysiological perspective, the axiom of HE as a gliopathy without a component of neuronal cell death35 also needs to be revisited. We believe a large multicenter study is in order, in which patients are well characterized before OLT and are followed at serial intervals after the procedure. Newer tools to assess brain function, including positron emission tomography scanning and functional MR imaging, should be incorporated into such an evaluation.
The authors thank Jeanne Gottstein for her assistance in this project.