Alzheimer's disease risk factor lymphocyte-specific protein tyrosine kinase regulates long-term synaptic strengthening, spatial learning and memory

c-Triadem: A constrained, explainable deep learning model to identify novel biomarkers in Alzheimer's disease

Alzheimer's disease (AD) is a neurodegenerative disorder that requires early diagnosis for effective management. However, issues with currently available diagnostic biomarkers preclude early diagnosis, necessitating the development of alternative biomarkers and methods, such as blood-based diagnostics. We propose c-Triadem (constrained triple-input Alzheimer's disease model), a novel deep neural network to identify potential blood-based biomarkers for AD and predict mild cognitive impairment (MCI) and AD with high accuracy. The model utilizes genotyping data, gene expression data, and clinical information to predict the disease status of participants, i.e., cognitively normal (CN), MCI, or AD. The nodes of the neural network represent genes and their related pathways, and the edges represent known relationships among the genes and pathways. Simulated data validation further highlights the robustness of key features identified by SHapley Additive exPlanations (SHAP). We trained the model with blood genotyping data, microarray, and clinical features from the Alzheimer's Neuroimaging Disease Initiative (ADNI). We demonstrate that our model's performance is superior to previous models with an AUC of 97% and accuracy of 89%. We then identified the most influential genes and clinical features for prediction using SHapley Additive exPlanations (SHAP). Our SHAP analysis shows that CASP9, LCK, and SDC3 SNPs and PINK1, ATG5, and ubiquitin (UBB, UBC) expression have a higher impact on model performance. Our model has facilitated the identification of potential blood-based genetic markers of DNA damage response and mitophagy in affected regions of the brain. The model can be used for detection and biomarker identification in other related dementias.

c-Triadem: A constrained, explainable deep learning model to identify novel biomarkers in Alzheimer's disease

Alzheimer's disease (AD) is a neurodegenerative disorder that requires early diagnosis for effective management. However, issues with currently available diagnostic biomarkers preclude early diagnosis, necessitating the development of alternative biomarkers and methods, such as blood-based diagnostics. We propose c-Triadem (constrained triple-input Alzheimer's disease model), a novel deep neural network to identify potential blood-based biomarkers for AD and predict mild cognitive impairment (MCI) and AD with high accuracy. The model utilizes genotyping data, gene expression data, and clinical information to predict the disease status of participants, i.e., cognitively normal (CN), MCI, or AD. The nodes of the neural network represent genes and their related pathways, and the edges represent known relationships among the genes and pathways. We trained the model with blood genotyping data, microarray, and clinical features from the Alzheimer's Neuroimaging Disease Initiative (ADNI). We demonstrate that our model's performance is superior to previous models with an AUC of 97% and accuracy of 89%. We then identified the most influential genes and clinical features for prediction using SHapley Additive exPlanations (SHAP). Our SHAP analysis shows that CASP9, LCK, and SDC3 SNPs and PINK1, ATG5, and ubiquitin (UBB, UBC) expression have a higher impact on model performance. Our model has facilitated the identification of potential blood-based genetic markers of DNA damage response and mitophagy in affected regions of the brain. The model can be used for detection and biomarker identification in other related dementias.

G-Protein Signaling in Alzheimer's Disease: Spatial Expression Validation of Semi-supervised Deep Learning-Based Computational Framework

Systemic study of pathogenic pathways and interrelationships underlying genes associated with Alzheimer's disease (AD) facilitates the identification of new targets for effective treatments. Recently available large-scale multiomics datasets provide opportunities to use computational approaches for such studies. Here, we devised a novel disease gene identification (digID) computational framework that consists of a semi-supervised deep learning classifier to predict AD-associated genes and a protein-protein interaction (PPI) network-based analysis to prioritize the importance of these predicted genes in AD. digID predicted 1,529 AD-associated genes and revealed potentially new AD molecular mechanisms and therapeutic targets including GNAI1 and GNB1, two G-protein subunits that regulate cell signaling, and KNG1, an upstream modulator of CDC42 small G-protein signaling and mediator of inflammation and candidate coregulator of amyloid precursor protein (APP). Analysis of mRNA expression validated their dysregulation in AD brains but further revealed the significant spatial patterns in different brain regions as well as among different subregions of the frontal cortex and hippocampi. Super-resolution STochastic Optical Reconstruction Microscopy (STORM) further demonstrated their subcellular colocalization and molecular interactions with APP in a transgenic mouse model of both sexes with AD-like mutations. These studies support the predictions made by digID while highlighting the importance of concurrent biological validation of computationally identified gene clusters as potential new AD therapeutic targets.

All-photonic kinase inhibitors: light-controlled release-and-report inhibition

Light-responsive molecular tools targeting kinases affords one the opportunity to study the underlying cellular function of selected kinases. In efforts to externally control lymphocyte-specific protein tyrosine kinase (LCK) activity, the development of release-and-report LCK inhibitors is described, in which (i) the release of the active kinase inhibitor can be controlled externally with light; and (ii) fluorescence is employed to report both the release and binding of the active kinase inhibitor. This introduces an unprecedented all-photonic method for users to both control and monitor real-time inhibitory activity. A functional cellular assay demonstrated light-mediated LCK inhibition in natural killer cells. The use of coumarin-derived caging groups resulted in rapid cellular uptake and non-specific intracellular localisation, while a BODIPY-derived caging group predominately localised in the cellular membrane. This concept of release-and-report inhibitors has the potential to be extended to other biorelevant targets where both spatiotemporal control in a cellular setting and a reporting mechanism would be beneficial.

Early-life iron deficiency persistently disrupts affective behaviour in mice

BACKGROUND/OBJECTIVE

Iron deficiency (ID) is the most common nutrient deficiency, affecting two billion people worldwide, including about 30% of pregnant women. During gestation, the brain is particularly vulnerable to environmental insults, which can irrevocably impair critical developmental processes. Consequently, detrimental consequences of early-life ID for offspring brain structure and function have been described. Although early life ID has been associated with an increased long-term risk for several neuropsychiatric disorders, the effect on depressive disorders has remained unresolved.

MATERIALS AND METHODS

A mouse model of moderate foetal and neonatal ID was established by keeping pregnant dams on an iron-deficient diet throughout gestation until postnatal day 10. The ensuing significant decrease of iron content in the offspring brain, as well as the impact on maternal behaviour and offspring vocalization was determined in the first postnatal week. The consequences of early-life ID for depression- and anxiety-like behaviour in adulthood were revealed employing dedicated behavioural assays. miRNA sequencing of hippocampal tissue of offspring revealed specific miRNAs signatures accompanying the behavioural deficits of foetal and neonatal ID in the adult brain.

RESULTS

Mothers receiving iron-deficient food during pregnancy and lactation exhibited significantly less licking and grooming behaviour, while active pup retrieval and pup ultrasonic vocalizations were unaltered. Adult offspring with a history of foetal and neonatal ID showed an increase in depression- and anxiety-like behaviour, paralleled by a deranged miRNA expression profile in the hippocampus, specifically levels of miR200a and miR200b.

CONCLUSION

ID during the foetal and neonatal periods has life-long consequences for affective behaviour in mice and leaves a specific and persistent mark on the expression of miRNAs in the brain. Foetal and neonatal ID needs to be further considered as risk factor for the development of depression and anxiety disorders later in life.Key MessagesMarginal reduction of gestational alimentary iron intake decreases brain iron content of the juvenile offspring.Early-life ID is associated with increased depression- and anxiety-like behaviour in adulthood.Reduction of maternal alimentary iron intake during pregnancy is reflected in an alteration of miRNA signatures in the adult offspring brain.

miRNA-132/212 Deficiency Disrupts Selective Corticosterone Modulation of Dorsal vs. Ventral Hippocampal Metaplasticity

Cortisol is a potent human steroid hormone that plays key roles in the central nervous system, influencing processes such as brain neuronal synaptic plasticity and regulating the expression of emotional and behavioral responses. The relevance of cortisol stands out in the disease, as its dysregulation is associated with debilitating conditions such as Alzheimer's Disease, chronic stress, anxiety and depression. Among other brain regions, cortisol importantly influences the function of the hippocampus, a structure central for memory and emotional information processing. The mechanisms fine-tuning the different synaptic responses of the hippocampus to steroid hormone signaling remain, however, poorly understood. Using ex vivo electrophysiology and wild type (WT) and miR-132/miR-212 microRNAs knockout (miRNA-132/212-/-) mice, we examined the effects of corticosterone (the rodent's equivalent to cortisol in humans) on the synaptic properties of the dorsal and ventral hippocampus. In WT mice, corticosterone predominantly inhibited metaplasticity in the dorsal WT hippocampi, whereas it significantly dysregulated both synaptic transmission and metaplasticity at dorsal and ventral regions of miR-132/212-/- hippocampi. Western blotting further revealed significantly augmented levels of endogenous CREB and a significant CREB reduction in response to corticosterone only in miR-132/212-/- hippocampi. Sirt1 levels were also endogenously enhanced in the miR-132/212-/- hippocampi but unaltered by corticosterone, whereas the levels of phospo-MSK1 were only reduced by corticosterone in WT, not in miR-132/212-/- hippocampi. In behavioral studies using the elevated plus maze, miRNA-132/212-/- mice further showed reduced anxiety-like behavior. These observations propose miRNA-132/212 as potential region-selective regulators of the effects of steroid hormones on hippocampal functions, thus likely fine-tuning hippocampus-dependent memory and emotional processing.

Exploring the interaction between T-cell antigen receptor-related genes and MAPT or ACHE using integrated bioinformatics analysis

Alzheimer's disease (AD) is a neurodegenerative disease that primarily occurs in elderly individuals with cognitive impairment. Although extracellular β-amyloid (Aβ) accumulation and tau protein hyperphosphorylation are considered to be leading causes of AD, the molecular mechanism of AD remains unknown. Therefore, in this study, we aimed to explore potential biomarkers of AD. Next-generation sequencing (NGS) datasets, GSE173955 and GSE203206, were collected from the Gene Expression Omnibus (GEO) database. Analysis of differentially expressed genes (DEGs), gene ontology (GO) functional enrichment, Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway enrichment, and protein-protein networks were performed to identify genes that are potentially associated with AD. Analysis of the DEG based protein-protein interaction (PPI) network using Cytoscape indicated that neuroinflammation and T-cell antigen receptor (TCR)-associated genes (LCK, ZAP70, and CD44) were the top three hub genes. Next, we validated these three hub genes in the AD database and utilized two machine learning models from different AD datasets (GSE15222) to observe their general relationship with AD. Analysis using the random forest classifier indicated that accuracy (78%) observed using the top three genes as inputs differed only slightly from that (84%) observed using all genes as inputs. Furthermore, another data set, GSE97760, which was analyzed using our novel eigenvalue decomposition method, indicated that the top three hub genes may be involved in tauopathies associated with AD, rather than Aβ pathology. In addition, protein-protein docking simulation revealed that the top hub genes could form stable binding sites with acetylcholinesterase (ACHE). This suggests a potential interaction between hub genes and ACHE, which plays an essential role in the development of anti-AD drug design. Overall, the findings of this study, which systematically analyzed several AD datasets, illustrated that LCK, ZAP70, and CD44 may be used as AD biomarkers. We also established a robust prediction model for classifying patients with AD.

Genetic Mapping of Behavioral Traits Using the Collaborative Cross Resource

The complicated interactions between genetic background, environment and lifestyle factors make it difficult to study the genetic basis of complex phenotypes, such as cognition and anxiety levels, in humans. However, environmental and other factors can be tightly controlled in mouse studies. The Collaborative Cross (CC) is a mouse genetic reference population whose common genetic and phenotypic diversity is on par with that of humans. Therefore, we leveraged the power of the CC to assess 52 behavioral measures associated with locomotor activity, anxiety level, learning and memory. This is the first application of the CC in novel object recognition tests, Morris water maze tasks, and fear conditioning tests. We found substantial continuous behavioral variations across the CC strains tested, and mapped six quantitative trait loci (QTLs) which influenced these traits, defining candidate genetic variants underlying these QTLs. Overall, our findings highlight the potential of the CC population in behavioral genetic research, while the identified genomic loci and genes driving the variation of relevant behavioral traits provide a foundation for further studies.

FIBCD1 is an endocytic GAG receptor associated with a novel neurodevelopmental disorder

Whole-exome sequencing of two patients with idiopathic complex neurodevelopmental disorder (NDD) identified biallelic variants of unknown significance within FIBCD1, encoding an endocytic acetyl group-binding transmembrane receptor with no known function in the central nervous system. We found that FIBCD1 preferentially binds and endocytoses glycosaminoglycan (GAG) chondroitin sulphate-4S (CS-4S) and regulates GAG content of the brain extracellular matrix (ECM). In silico molecular simulation studies and GAG binding analyses of patient variants determined that such variants are loss-of-function by disrupting FIBCD1-CS-4S association. Gene knockdown in flies resulted in morphological disruption of the neuromuscular junction and motor-related behavioural deficits. In humans and mice, FIBCD1 is expressed in discrete brain regions, including the hippocampus. Fibcd1 KO mice exhibited normal hippocampal neuronal morphology but impaired hippocampal-dependent learning. Further, hippocampal synaptic remodelling in acute slices from Fibcd1 KO mice was deficient but restored upon enzymatically modulating the ECM. Together, we identified FIBCD1 as an endocytic receptor for GAGs in the brain ECM and a novel gene associated with an NDD, revealing a critical role in nervous system structure, function and plasticity.

New horizons in drug discovery of lymphocyte-specific protein tyrosine kinase (Lck) inhibitors: a decade review (2011-2021) focussing on structure-activity relationship (SAR) and docking insights

Lymphocyte-specific protein tyrosine kinase (Lck), a non-receptor Src family kinase, has a vital role in various cellular processes such as cell cycle control, cell adhesion, motility, proliferation, and differentiation. Lck is reported as a key factor regulating the functions of T-cell including the initiation of TCR signalling, T-cell development, in addition to T-cell homeostasis. Alteration in expression and activity of Lck results in numerous disorders such as cancer, asthma, diabetes, rheumatoid arthritis, atherosclerosis, and neuronal diseases. Accordingly, Lck has emerged as a novel target against different diseases. Herein, we amass the research efforts in literature and pharmaceutical patents during the last decade to develop new Lck inhibitors. Additionally, structure-activity relationship studies (SAR) and docking models of these new inhibitors within the active site of Lck were demonstrated offering deep insights into their different binding modes in a step towards the identification of more potent, selective, and safe Lck inhibitors.

Western diet-induced fear memory impairment is attenuated by 6-shogaol in C57BL/6N mice

Dementia is a progressive cognitive diminution impeding with normal daily activities that is constantly on the increase. Currently, the estimated prevalence is 50 million affected people worldwide, a figure expected to triple within the next 30 years. While the pathophysiology of the different types of dementia is complex, likely involving the interplay between multiple genetic and environmental factors, strong evidence points towards an important link between diet and cognitive health. Here we examined the consequences of high-fat, high-sugar Western diet (HFSD)-induced obesity on cognitive performance in the fear conditioning task in mice and explored a possible beneficial effect of 6-shogaol (6S), an active constituent of ginger, in this model. Chronic exposure to HFSD significantly enhanced body weight gain in C57BL/6N mice and this effect was prevented by treatment with 6S. HFSD + vehicle-treated mice presented with a selective deficit in cued fear memory, which was not observed in HFSD + 6S-treated animals. The findings of this study provide first evidence for a beneficial effect of 6S on HFSD-induced obesity and emotional memory deficit in mice.

Beyond TCR Signaling: Emerging Functions of Lck in Cancer and Immunotherapy

In recent years, the lymphocyte-specific protein tyrosine kinase (Lck) has emerged as one of the key molecules regulating T-cell functions. Studies using Lck knock-out mice or Lck-deficient T-cell lines have shown that Lck regulates the initiation of TCR signaling, T-cell development, and T-cell homeostasis. Because of the crucial role of Lck in T-cell responses, strategies have been employed to redirect Lck activity to improve the efficacy of chimeric antigen receptors (CARs) and to potentiate T-cell responses in cancer immunotherapy. In addition to the well-studied role of Lck in T cells, evidence has been accumulated suggesting that Lck is also expressed in the brain and in tumor cells, where it actively takes part in signaling processes regulating cellular functions like proliferation, survival and memory. Therefore, Lck has emerged as a novel druggable target molecule for the treatment of cancer and neuronal diseases. In this review, we will focus on these new functions of Lck.

The exploration of novel Alzheimer's therapeutic agents from the pool of FDA approved medicines using drug repositioning, enzyme inhibition and kinetic mechanism approaches

Novel drug development is onerous, time consuming and overpriced process with particularly low success and relatively high enfeebling rates. To overcome this burden, drug repositioning approach is being used to predict the possible therapeutic effects of FDA approved drugs in different diseases. Herein, we designed a computational and enzyme inhibitory mechanistic approach to fetch the promising drugs from the pool of FDA approved drugs against AD. The binding interaction patterns and conformations of screened drugs within active region of AChE were confirmed through molecular docking profiles. The possible associations of selected drugs with AD genes were predicted by pharmacogenomics analysis and confirmed through data mining. The stability behaviour of docked complexes (Drugs-AChE) were checked by MD simulations. The possible therapeutic potential of repositioned drugs against AChE were checked by in vitro analysis. Taken together, Cinitapride displayed a comparable results with standard and can be used as possible therapeutic agent in the treatment of AD.

Exploration of the therapeutic aspects of Lck: A kinase target in inflammatory mediated pathological conditions

Lck, a non-receptor src family kinase, plays a vital role in various cellular processes such as cell cycle control, cell adhesion, motility, proliferation and differentiation. As a 56 KDa protein, Lck phosphorylates tyrosine residues of various proteins such as ZAP-70, ITK and protein kinase C. The structure of Lck is comprised of three domains, one SH3 in tandem with a SH2 domain at the amino terminal and the kinase domain at the carboxy terminal. Physiologically, Lck is involved in the development, function and differentiation of T-cells. Additionally, Lck regulates neurite outgrowth and maintains long-term synaptic plasticity in neurons. Given a major role of Lck in cytokine production and T cell signaling, alteration in expression and activity of Lck may result in various diseased conditions like cancer, asthma, diabetes, rheumatoid arthritis, psoriasis, inflammatory bowel diseases such as Crohn's disease and ulcerative colitis, atherosclerosis etc. This article provides evidence and information establishing Lck as one of the therapeutic targets in various inflammation mediated pathophysiological conditions.

Enhanced synaptic plasticity and spatial memory in female but not male FLRT2-haplodeficient mice

The Fibronectin Leucine-Rich Transmembrane protein 2 (FLRT2) has been implicated in several hormone -and sex-dependent physiological and pathological processes (including chondrogenesis, menarche and breast cancer); is known to regulate developmental synapses formation, and is expressed in the hippocampus, a brain structure central for learning and memory. However, the role of FLRT2 in the adult hippocampus and its relevance in sex-dependent brain functions remains unknown. We here used adult single-allele FLRT2 knockout (FLRT2+/-) mice and behavioral, electrophysiological, and molecular/biological assays to examine the effects of FLRT2 haplodeficiency on synaptic plasticity and hippocampus-dependent learning and memory. Female and male FLRT2+/- mice presented morphological features (including body masses, brain shapes/weights, and brain macroscopic cytoarchitectonic organization), indistinguishable from their wild type counterparts. However, in vivo examinations unveiled enhanced hippocampus-dependent spatial memory recall in female FLRT2+/- animals, concomitant with augmented hippocampal synaptic plasticity and decreased levels of the glutamate transporter EAAT2 and beta estrogen receptors. In contrast, male FLRT2+/- animals exhibited deficient memory recall and decreased alpha estrogen receptor levels. These observations propose that FLRT2 can regulate memory functions in the adulthood in a sex-specific manner and might thus contribute to further research on the mechanisms linking sexual dimorphism and cognition.

Sustained consumption of cocoa-based dark chocolate enhances seizure-like events in the mouse hippocampus

While the consumption of caffeine and cocoa has been associated with a variety of health benefits to humans, some authors have proposed that excessive caffeine intake may increase the frequency of epileptic seizures in humans and reduce the efficiency of antiepileptic drugs. Little is known, however, about the proconvulsant potential of the sustained, excessive intake of cocoa on hippocampal neural circuits. Using the mouse as an experimental model, we examined the effects of the chronic consumption of food enriched in cocoa-based dark chocolate on motor and mood-related behaviours as well as on the excitability properties of hippocampal neurons. Cocoa food enrichment did not affect body weights or mood-related behaviours but rather promoted general locomotion and improved motor coordination. However, ex vivo electrophysiological analysis revealed a significant enhancement in seizure-like population spike bursting at the neurogenic dentate gyrus, which was paralleled by a significant reduction in the levels of GABA-α1 receptors thus suggesting that an excessive dietary intake of cocoa-enriched food might alter some of the synaptic elements involved in epileptogenesis. These data invite further multidisciplinary research aiming to elucidate the potential deleterious effects of chocolate abuse on behaviour and brain hyperexcitability.

Long-term effects of maternal immune activation on depression-like behavior in the mouse

Depression is a debilitating mental disease affecting a large population worldwide, the pathophysiological mechanisms of which remain incompletely understood. Prenatal infection and associated activation of the maternal immune system (MIA) are prominently related to an increased risk for the development of several psychiatric disorders including schizophrenia and autism in the offsprings. However, the role of MIA in the etiology of depression and its neurobiological basis are insufficiently investigated. Here we induced MIA in mice by challenge with polyinosinic:polycytidylic phosphate salt-a synthetic analog of double-stranded RNA, which enhances maternal levels of the cytokine interleukin-6 (IL-6)-and demonstrate a depression-like behavioral phenotype in adult offsprings. Adult offsprings additionally show deficits in cognition and hippocampal long-term potentiation (LTP) accompanied by disturbed proliferation of newborn cells in the dentate gyrus and compromised neuronal maturation and survival. The behavioral, neurogenic and functional deficiencies observed are associated with reduced hippocampal expression of vascular endothelial growth factor (VEGF)A-VEGFR2. IL-6-STAT3-dependent aberrant VEGFA-VEGFR2 signaling is proposed as neurobiological mechanism mediating the effects of MIA on the developing fetal brain and ensuing consequences in adulthood.

Sequestosome1/p62: a regulator of redox-sensitive voltage-activated potassium channels, arterial remodeling, inflammation, and neurite outgrowth

Sequestosome1/p62 (SQSTM1) is an oxidative stress-inducible protein regulated by the redox-sensitive transcription factor Nrf2. It is not an antioxidant but known as a multifunctional regulator of cell signaling with an ability to modulate targeted or selective degradation of proteins through autophagy. SQSTM1 implements these functions through physical interactions with different types of proteins including atypical PKCs, nonreceptor-type tyrosine kinase p56(Lck) (Lck), polyubiquitin, and autophagosomal factor LC3. One of the notable physiological functions of SQSTM1 is the regulation of redox-sensitive voltage-gated potassium (Kv) channels which are composed of α and β subunits: (Kvα)4 (Kvβ)4. Previous studies have established that SQSTM1 scaffolds PKCζ, enhancing phosphorylation of Kvβ which induces inhibition of pulmonary arterial Kv1.5 channels under acute hypoxia. Recent studies reveal that Lck indirectly interacts with Kv1.3 α subunits and plays a key role in acute hypoxia-induced Kv1.3 channel inhibition in T lymphocytes. Kv1.3 channels provide a signaling platform to modulate the migration and proliferation of arterial smooth muscle cells and activation of T lymphocytes, and hence have been recognized as a therapeutic target for treatment of restenosis and autoimmune diseases. In this review, we focus on the functional interactions of SQSTM1 with Kv channels through two key partners aPKCs and Lck. Furthermore, we provide molecular insights into the functions of SQSTM1 in suppression of proliferation of arterial smooth muscle cells and neointimal hyperplasia following carotid artery ligation, in T lymphocyte differentiation and activation, and in NGF-induced neurite outgrowth in PC12 cells.

Can the tyrosine kinase inhibitors trigger metabolic encephalopathy in cirrhotic patients?

BACKGROUND

Sorafenib is the standard treatment of advanced hepatocarcinoma (HCC) in cirrhotic patients with preserved liver function. It shares many adverse effects with other tyrosine-kinase (TK) inhibitors and antiangiogenic drugs. TK inhibitors could have a direct toxicity on CNS, both by interfering with TK-related pathways and by inhibiting angiogenesis.

AIMS

The aim of this study was to investigate whether sorafenib administration can be associated to metabolic encephalopathy in patients with cirrhosis.

METHODS

We retrospectively reviewed medical records of all cirrhotic patients treated with sorafenib for HCC afferent at our Department from January 2009 to December 2011.

RESULTS

Among 62 patients, we identified 10 patients with clinically significant cognitive impairment. Seven of these were clearly diagnosed with overt hepatic encephalopathy (HE), one with brain metastases and two with drug-related toxic-metabolic encephalopathy. These last two cases were characterized by severe cognitive impairment, mood alteration and memory deficit. Clinical exam, blood tests and brain CT excluded organic causes of encephalopathy and precipitating factors of HE. Sorafenib discontinuation was associated with complete reversal of the syndrome, which recurred on drug re-administration in one case.

CONCLUSIONS

Our study suggests that sorafenib may be a precipitating factor of metabolic encephalopathy in cirrhotic patients with advanced HCC. This neurological syndrome appears to be not responsive to the conventional treatment for HE, but it is fully reversible by drug discontinuation. It can be speculated that the potential direct neuronal action of sorafenib may represent a trigger for the onset of metabolic encephalopathy in a subset of cirrhotic patients.