Hepatic encephalopathy predicts early post-transplant cognitive and functional impairment: The Livcog cohort study

INTRODUCTION

HE affects 40%–80% of patients with decompensated cirrhosis based on various estimates.1,2 HE is the liver decompensation associated with the highest health care use, while the quality of life decrement associated with HE is well described. Despite some patients with decompensated cirrhosis undergoing liver transplant (LT), cognitive function may not fully recover after LT in up to one third of cases, though longitudinal studies have been limited. It is also unclear whether other comorbidities contribute to persistent cognitive impairment (CI), such as pre-existing Alzheimer disease and related dementia neuropathology masked by liver disease prior to LT. CI is a top symptom of concern identified by liver transplant recipients (LTRs) and affects daily function, adherence to medical regimens, and quality of life.1,3,4

Large prospective studies evaluating the impact of HE, medical comorbidities, and pretransplant factors that affect early post-transplant function are limited. The LivCog cohort fills this gap by prospectively evaluating longitudinal changes in cognition, objective and subjective function, quality of life, self-management, and clinical outcomes.5,6 In this preliminary study, we sought to quantify pre-LT and early post-LT CI and investigate the associations between pretransplant HE and post-transplant delirium, nonhome discharge, and cognition as at 3 months after LT.

METHODS

Detailed methods of the LivCog cohort have been published previously.5,6 In brief, this is a post-transplant longitudinal cohort study following patients for 2 years after LT at 4 large transplant centers in Pennsylvania, Texas, and Illinois. The inclusion criteria are pragmatic enrolling adult patients 18 or older who are English-speaking or Spanish-speaking on the LT waitlist or LTRs, including simultaneous liver-kidney recipients. Assessments of cognition occur in a subset of pre-LT patients and for all patients at 1–3, 12, and 24 months after LT. A single Institutional Review Board at the University of Pennsylvania has approved all study procedures (Protocol# 850519). Written informed consent is obtained in person or via electronic means. All research was conducted in accordance with the Declarations of Helsinki and Istanbul.

Measures and outcomes

The LivCog cohort collects detailed data on age, race, sex, ethnicity, and education. Pretransplant complications are abstracted by trained staff from the electronic health record and include liver disease etiology: alcohol-associated liver disease, alcohol-associated hepatitis, viral, cholestatic, other liver disease, and peritransplant complications, including postoperative delirium, and nonhome discharge. HE diagnosis is obtained from outpatient hepatology notes, inpatient discharge summaries, and/or outpatient prescriptions for lactulose or rifaximin. Outpatient HE and the highest grade of pre-LT HE hospitalization are also abstracted. Detailed data on pretransplant medical history (eg, diabetes, hypertension, kidney disease, dyslipidemia, and smoking) and laboratory parameters at waitlisting and LT are also collected. Cognitive function is serially assessed via the Montreal Cognitive Assessment (MoCA); a well-validated test of global cognition, and the National Institutes of Health Toolbox (NIHTB), a validated toolkit attention that measures a composite total score and individual subscores on attention executive function, memory, processing speed, and language.⁷ The self-reported, 12-item everyday cognition scale assesses functional and cognitive abilities. An additional question assessed the prevalence of having a memory or other thinking concern.⁸ 12-item everyday cognition scale has been validated for use in Alzheimer disease to track progression of functional and cognitive abilities.⁹

Analysis

For this analysis, descriptive statistics were calculated for all variables: percentages for categorical data and means (M), SD, medians, IQR for continuous data. Chi-squared tests and t tests compared patient characteristics by pretransplant HE status. To evaluate the effects of pretransplant HE or hospitalized pretransplant HE on perioperative delirium, nonhome discharge, and post-transplant cognitive function at 3 months, multivariable models, logistic regression for categorical variables, and linear regression for continuous variables, were fit adjusting for age, education, post-LT liver frailty index, liver disease etiology, hypertension, dyslipidemia, diabetes, smoking, body mass index and MELD-Na components at waitlisting. Covariates were chosen a priori based on the literature review and were limited due to sample size considerations. Analyses were conducted with Stata 15 (Statacorp, College Station, TX). p values of <0.05 were considered statistically significant.

RESULTS

Among 232 LTRs enrolled from October 2022 to October 2024, the mean age was 54.4 (SD=12.0); 39.2% of LTRs were 60 and older; 33.6% were female; 42.7% received a high school education or less. Prior to LT, 106 (45.7%) had documented HE, and 71 (30.6%) had at least 1 HE hospitalization. One third of LTRs met the criteria for frailty; median MELD-Na at LT was 20 (IQR 14–27). The most common etiologies of liver disease were alcohol-associated liver disease and metabolic dysfunction–associated liver disease. Table S1, http://links.lww.com/HC9/B953, shows patient characteristics stratified by any pretransplant HE.

A total of 30.6% of patients reported a concern with memory or thinking at 1–3 months after LT; this was not associated with HE, however. Prior to LT, greater than half of LTRs (56.0%) had age-adjusted and education-adjusted MoCA scores consistent with CI (<26); this was observed in 48.9% of individuals by 3 months post-LT. As shown in Table 1, pre-LT HE (OR 3.03, 95% CI: 1.17–7.82) and pre-LT hospitalized HE (OR 2.61, 95% 1.14–6.02) were associated with pre-LT CI by MoCA (< 26). Pre-LT hospitalized HE was associated with perioperative delirium (OR 2.29, 95% CI: 1.08–4.87) and nonhome discharge after LT (OR 3.64, 95% CI: 1.48–8.88). Both pre-LT HE and hospitalized pre-LT HE were associated with numerically lower MoCA scores prior to LT and early after LT as well as 7–8 points lower T scores on the NIHTB cognitive battery.

TABLE 1.

Associations between pretransplant HE and pre-LT cognition, post-LT cognition, and functional recovery

		Any HE pretransplant				Hospitalized for HE pretransplant
	N	OR	[95% CI]		p	OR	[95% CI]		p
Categorical outcomes
MoCA CI (<26) pre-LT	127	3.03	1.17	7.82	0.022	2.61	1.14	6.02	0.024
MoCA CI (<26) 3 mo post-LT	189	1.77	0.89	3.50	0.104	1.11	0.53	2.32	0.779
Perioperative delirium	220	1.10	0.54	2.27	0.792	2.29	1.08	4.87	0.030
Nonhome discharge post-LT	220	4.21	1.68	10.57	0.002	3.64	1.48	8.88	0.004
Memory or other thinking concern present post-LT	113	1.93	0.77	4.80	0.157	1.50	0.56	4.04	0.419
Continuous outcomes	N	β coefficient	[95% CI]		p	β coefficient	[95% Conf. Interval]		p
MoCA total score pre-LTa	127	−2.02	−3.67	−0.37	0.017	−1.85	−3.61	−0.090	0.040
MoCA total score 3 Months post-LTa	189	−1.31	−2.28	−0.34	0.008	−1.11	−1.12	−0.045	0.041
NIH toolbox T score 3 mo post-LT a	53	−7.24	−13.12	−1.37	0.017	−8.03	−15.17	−0.87	0.029
ECog-12 total score post-LTb	155	0.18	0.012	0.34	0.036	0.011	−0.17	0.19	0.904

Note: Pre-LT models adjusted for age, education, liver frailty index, waitlist MELD. Post-LT candidate variables included age, education, liver frailty index, transplant MELD components, liver disease etiology including alcohol, body mass index, perioperative delirium, diabetes, hypertension, dyslipidemia, and tobacco use. Perioperative delirium was removed from several models due to collinearity with hepatic encephalopathy hospitalizations.

^a Negative β coefficient indicates worse cognitive performance.

^b Positive β coefficient indicates higher self-reported impairment.

Abbreviations: CI, cognitive impairment; ECog-12, 12-item everyday cognition scale; LT, liver transplant.

DISCUSSION

Self-rated and objective CIs are common in the early post-transplant period, and pretransplant HE increases the risk of post-transplant delirium, nonhome discharge, and objective CI pre-LT and in the early post-LT period. Patients with HE will require a higher level of care partner and transplant center support. Early LivCog data shows the important contributions of pretransplant HE to early post-transplant global function, which could affect post-transplant self-care and outcomes. Future investigations from the LivCog cohort will examine in detail the contribution of pretransplant complications, duration and severity of HE on post-LT function and longer-term cognitive recovery.