ChemMORT: an automatic ADMET optimization platform using deep learning and multi-objective particle swarm optimization.

Drug discovery and development constitute a laborious and costly undertaking. The success of a drug hinges not only good efficacy but also acceptable absorption, distribution, metabolism, elimination, and toxicity (ADMET) properties. Overall, up to 50% of drug development failures have been contributed from undesirable ADMET profiles. As a multiple parameter objective, the optimization of the ADMET properties is extremely challenging owing to the vast chemical space and limited human expert knowledge. In this study, a freely available platform called Chemical Molecular Optimization, Representation and Translation (ChemMORT) is developed for the optimization of multiple ADMET endpoints without the loss of potency (https://cadd.nscc-tj.cn/deploy/chemmort/). ChemMORT contains three modules: Simplified Molecular Input Line Entry System (SMILES) Encoder, Descriptor Decoder and Molecular Optimizer. The SMILES Encoder can generate the molecular representation with a 512-dimensional vector, and the Descriptor Decoder is able to translate the above representation to the corresponding molecular structure with high accuracy. Based on reversible molecular representation and particle swarm optimization strategy, the Molecular Optimizer can be used to effectively optimize undesirable ADMET properties without the loss of bioactivity, which essentially accomplishes the design of inverse QSAR. The constrained multi-objective optimization of the poly (ADP-ribose) polymerase-1 inhibitor is provided as the case to explore the utility of ChemMORT.

ABC-Net: a divide-and-conquer based deep learning architecture for SMILES recognition from molecular images.

Structural information for chemical compounds is often described by pictorial images in most scientific documents, which cannot be easily understood and manipulated by computers. This dilemma makes optical chemical structure recognition (OCSR) an essential tool for automatically mining knowledge from an enormous amount of literature. However, existing OCSR methods fall far short of our expectations for realistic requirements due to their poor recovery accuracy. In this paper, we developed a deep neural network model named ABC-Net (Atom and Bond Center Network) to predict graph structures directly. Based on the divide-and-conquer principle, we propose to model an atom or a bond as a single point in the center. In this way, we can leverage a fully convolutional neural network (CNN) to generate a series of heat-maps to identify these points and predict relevant properties, such as atom types, atom charges, bond types and other properties. Thus, the molecular structure can be recovered by assembling the detected atoms and bonds. Our approach integrates all the detection and property prediction tasks into a single fully CNN, which is scalable and capable of processing molecular images quite efficiently. Experimental results demonstrate that our method could achieve a significant improvement in recognition performance compared with publicly available tools. The proposed method could be considered as a promising solution to OCSR problems and a starting point for the acquisition of molecular information in the literature.

BioNet: a large-scale and heterogeneous biological network model for interaction prediction with graph convolution.

MOTIVATION: Understanding chemical-gene interactions (CGIs) is crucial for screening drugs. Wet experiments are usually costly and laborious, which limits relevant studies to a small scale. On the contrary, computational studies enable efficient in-silico exploration. For the CGI prediction problem, a common method is to perform systematic analyses on a heterogeneous network involving various biomedical entities. Recently, graph neural networks become popular in the field of relation prediction. However, the inherent heterogeneous complexity of biological interaction networks and the massive amount of data pose enormous challenges. This paper aims to develop a data-driven model that is capable of learning latent information from the interaction network and making correct predictions. RESULTS: We developed BioNet, a deep biological networkmodel with a graph encoder-decoder architecture. The graph encoder utilizes graph convolution to learn latent information embedded in complex interactions among chemicals, genes, diseases and biological pathways. The learning process is featured by two consecutive steps. Then, embedded information learnt by the encoder is then employed to make multi-type interaction predictions between chemicals and genes with a tensor decomposition decoder based on the RESCAL algorithm. BioNet includes 79 325 entities as nodes, and 34 005 501 relations as edges. To train such a massive deep graph model, BioNet introduces a parallel training algorithm utilizing multiple Graphics Processing Unit (GPUs). The evaluation experiments indicated that BioNet exhibits outstanding prediction performance with a best area under Receiver Operating Characteristic (ROC) curve of 0.952, which significantly surpasses state-of-theart methods. For further validation, top predicted CGIs of cancer and COVID-19 by BioNet were verified by external curated data and published literature.

The association between trajectory of serum cholesterol, statin dosage, and the risk of recurrent biliary stone diseases.

BACKGROUND: Statins may reduce the risk of recurrent gallstone disease by decreasing bile cholesterol saturation and pathogenicity. However, limited studies have investigated this issue. This study aimed to assess whether statin doses and serum cholesterol levels were associated with a decreased risk of recurrent biliary stone diseases after the first event index, with a follow-up time of 15 years. METHODS: Based on the Chang Gung Research Database (CGRD) between January 1, 2001, and December 31, 2020, we enrolled 68,384 patients with the International Classification of Diseases, Ninth and Tenth Revision codes of choledocholithiasis. After exclusions, 32,696 patients were divided into non-statin (<28 cDDD, cumulative defined daily doses) (n = 27,929) and statin (≥28 cDDD) (n = 4767) user groups for analysis. Serum cholesterol trajectories were estimated using group-based trajectory modeling (n = 8410). RESULTS: The statin users had higher Charlson Comorbidity Index (CCI) scores than the non-statin users. Time-dependent Cox regression analysis showed that statin use >365 cDDD was associated with a significantly lower risk of recurrent biliary stones (adjusted hazard ratio [aHR] = 0.28, 95% CI, 0.24-0.34; p < 00.0001), acute pancreatitis (aHR = 0.24, 95% CI, 0.17-0.32, p < 00.0001), and cholangitis (aHR = 0.28, 95% CI, 0.25-0.32, p < 00.0001). Cholecystectomy was also a protective factor for recurrent biliary stones (aHR = 0.41, 95% CI, 0.37-0.46; p < 00.0001). The higher trajectory serum cholesterol group (Group 3) had a lower risk trend for recurrent biliary stones (aHR = 0.79, p = 0.0700) and a lower risk of cholangitis (aHR = 0.79, p = 0.0071). CONCLUSION: This study supports the potential benefits of statin use and the role of cholecystectomy in reducing the risk of recurrent biliary stone diseases.

Learning to SMILES: BAN-based strategies to improve latent representation learning from molecules.

Computational methods have become indispensable tools to accelerate the drug discovery process and alleviate the excessive dependence on time-consuming and labor-intensive experiments. Traditional feature-engineering approaches heavily rely on expert knowledge to devise useful features, which could be costly and sometimes biased. The emerging deep learning (DL) methods deliver a data-driven method to automatically learn expressive representations from complex raw data. Inspired by this, researchers have attempted to apply various deep neural network models to simplified molecular input line entry specification (SMILES) strings, which contain all the composition and structure information of molecules. However, current models usually suffer from the scarcity of labeled data. This results in a low generalization ability of SMILES-based DL models, which prevents them from competing with the state-of-the-art computational methods. In this study, we utilized the BiLSTM (bidirectional long short term merory) attention network (BAN) in which we employed a novel multi-step attention mechanism to facilitate the extracting of key features from the SMILES strings. Meanwhile, SMILES enumeration was utilized as a data augmentation method in the training phase to substantially increase the number of labeled data and enlarge the probability of mining more patterns from complex SMILES. We again took advantage of SMILES enumeration in the prediction phase to rectify model prediction bias and provide a more accurate prediction. Combined with the BAN model, our strategies can greatly improve the performance of latent features learned from SMILES strings. In 11 canonical absorption, distribution, metabolism, excretion and toxicity-related tasks, our method outperformed the state-of-the-art approaches.

MG-BERT: leveraging unsupervised atomic representation learning for molecular property prediction.

MOTIVATION: Accurate and efficient prediction of molecular properties is one of the fundamental issues in drug design and discovery pipelines. Traditional feature engineering-based approaches require extensive expertise in the feature design and selection process. With the development of artificial intelligence (AI) technologies, data-driven methods exhibit unparalleled advantages over the feature engineering-based methods in various domains. Nevertheless, when applied to molecular property prediction, AI models usually suffer from the scarcity of labeled data and show poor generalization ability. RESULTS: In this study, we proposed molecular graph BERT (MG-BERT), which integrates the local message passing mechanism of graph neural networks (GNNs) into the powerful BERT model to facilitate learning from molecular graphs. Furthermore, an effective self-supervised learning strategy named masked atoms prediction was proposed to pretrain the MG-BERT model on a large amount of unlabeled data to mine context information in molecules. We found the MG-BERT model can generate context-sensitive atomic representations after pretraining and transfer the learned knowledge to the prediction of a variety of molecular properties. The experimental results show that the pretrained MG-BERT model with a little extra fine-tuning can consistently outperform the state-of-the-art methods on all 11 ADMET datasets. Moreover, the MG-BERT model leverages attention mechanisms to focus on atomic features essential to the target property, providing excellent interpretability for the trained model. The MG-BERT model does not require any hand-crafted feature as input and is more reliable due to its excellent interpretability, providing a novel framework to develop state-of-the-art models for a wide range of drug discovery tasks.

Risks of adverse events for users of proton-pump inhibitors plus aspirin or clopidogrel in patients with aspirin-related ulcer bleeding.

BACKGROUND AND AIM: Clopidogrel is widely prescribed for patients with of aspirin-related upper gastrointestinal bleeding (UGIB) history. This study aimed to compare the risk of a major adverse cardiovascular event (MACE), UGIB, and mortality between aspirin and clopidogrel in patients at risk of bleeding. METHODS: We analyzed adult patients at high risk of UGIB following aspirin-related bleeding for secondary MACE prevention between 2000 and 2012. Secondary prevention was for those patients who had ever been hospitalized for cardiovascular disease and reused aspirin or changed to clopidogrel after discharge. Study endpoints were recurrence of MACE, UGIB, and death in 90 days of follow-up. The associations between study outcomes and the use of clopidogrel (vs aspirin) were analyzed. RESULTS: Among 947 eligible patients, 653 reused aspirin (in combination with a proton-pump inhibitor), and 294 were treated with clopidogrel (in combination with a proton-pump inhibitor) after discharge for UGIB. Compared with aspirin treatment, clopidogrel showed an increased risk of MACE (adjusted hazard ratio [aHR] 1.65; 95% confidence interval [CI] 0.87-3.12) and UGIB (aHR 1.25; 95% CI 0.66-2.36), but without statistical significance in 90 days' follow-up. Clopidogrel use was associated with greater than four times the risk of any cause of mortality (aHR 4.84; 95% CI 1.59-14.75), but the significance did not hold in propensity score-matched cohort analysis (P = 0.06). CONCLUSIONS: A nonsignificant difference between clopidogrel and aspirin for short-term MACE prevention as well as UGIB recurrence was found in the present study. Further research to assess 90-day mortality would assist clinical decision making.

Incidence and risk factors of colonoscopic post-polypectomy bleeding and perforation in patients with end-stage renal disease.

BACKGROUND AND AIMS: Colonoscopic polypectomy in end-stage renal disease (ESRD) patients are at risks of post-polypectomy bleeding and perforation, but evidences are limited. This study aimed to determine the incident polypectomy complications among ESRD patients. METHODS: In the nationwide ESRD cohort, a propensity score matched case-control study design was conducted to assess risk associated with post-polypectomy bleeding and perforation using the Taiwanese National Health Insurance Research Database from 1997 to 2013 for adults aged 40 years and older; 7011 ESRD and 19 118 non-ESRD patients met the study criteria. A total of 5302 patients in each group were matched for further analyses. The primary endpoint was post-polypectomy bleeding or bowel perforation in 30 days. The secondary endpoint was mortality and length of hospital stay for the bleeding complications requiring hospitalization. RESULTS: Overall incidences of post-polypectomy bleeding or perforation in patients with ESRD was higher than the non-ESRD group (5.83% vs 1.78%, P < 0.0001) in the matched cohort. High risk of adverse outcomes was associated with ESRD (adjusted odds ratio [aOR], 2.38, 95% confidence interval [CI], 1.85-3.05), female patient (aOR, 1.7, 95% CI, 1.37-2.11), history of acute myocardial infarction (aOR, 1.91, 95% CI, 1.1-3.32), liver disease (aOR, 1.79, 95% CI, 1.37-2.34), diabetes (aOR, 1.45, 95% CI, 1.16-1.82), cancer (aOR, 1.4, 95% CI, 1.09-1.81), inpatient setting (aOR, 13.19, 95% CI, 9.73-17.88), and prior use of clopidogrel (aOR, 1.61, 95% CI, 1.03-2.52) and warfarin (aOR, 2.03, 95% CI, 1.21-3.41). CONCLUSIONS: End-stage renal disease was associated with approximately twofold higher risk of colonoscopic post-polypectomy bleeding or perforation and should be cautiously performed in this special population cohort.

The role of antibiotics in upper gastrointestinal bleeding among cirrhotic patients without major complications after endoscopic hemostasis.

BACKGROUND AND AIM: Antibiotic prophylaxis should be instituted for cirrhotic patients with upper gastrointestinal bleeding (UGIB), but the benefit on compensated patients remains undetermined. We aimed to compare the clinical outcomes between cirrhotic patients without major complications with UGIB with and without antibiotic prophylaxis. METHODS: We conducted this population-based cohort study by using Taiwanese Longitudinal Health Insurance Database 2000 (LHID2000, between 1997 to 2013), aged 18 years or older with a hospital discharge diagnosis of cirrhosis (n = 64,506), UGIB (n = 7,784), and endoscopic therapy (n = 2,292). After strict exclusions, 1205 patients were enrolled and were divided into antibiotic exposure (n = 558) and non-exposure (n = 647) groups. The outcomes were rebleeding and mortality. RESULTS: After completing the analysis adjusted by death, the rebleeding rates within 4 weeks were significantly lower in patients with antibiotic prophylaxis (3.05% versus 6.03%, P = 0.0142) and those with endoscopic therapy (0.72% vs 3.09%, P = 0.0033) but not significant after 3 months and onwards. Male patients aged > 55, high CCI score â‰§ 4, and UGIB of variceal etiologies were benefited from rebleeding. The use of antibiotics did not significantly impact 6-week mortality (adjusted hazard ratio: 1.07, 95%CI: 0.41~2.75; P = 0.8943). Old age, multiple comorbidities, and UGIB of variceal etiologies were risk factors of all-cause mortality. CONCLUSIONS: The current study suggested that cirrhotic patients without major complications who suffered from UGIB were benefited by the use of antibiotics to prevent rebleeding within 4 weeks after endoscopic treatment of UGIB especially for those with age > 55, high CCI score â‰§ 4, and UGIB of variceal etiologies.

A 14 day esomeprazole- and amoxicillin-containing high-dose dual therapy regimen achieves a high eradication rate as first-line anti-Helicobacter pylori treatment in Taiwan: a prospective randomized trial.

BACKGROUND: The first-line eradication rate of standard triple therapy for Helicobacter pylori infection has declined to <80%, and alternative therapies with >90% success rates are needed. Inconsistent eradication rates were reported for proton pump inhibitor- and amoxicillin-containing high-dose dual therapy. OBJECTIVES: We performed a prospective, randomized controlled study to assess the efficacy of esomeprazole- and amoxicillin-containing high-dose dual therapy and investigated the influencing clinical factors. PATIENTS AND METHODS: We recruited 240/278 eligible H. pylori-infected patients after exclusion. They were randomly assigned to 14 day high-dose dual therapy (esomeprazole 40 mg three times daily and amoxicillin 750 mg four times daily for 14 days; EA group) or 7 day non-bismuth quadruple therapy (esomeprazole 40 mg twice daily, clarithromycin 500 mg twice daily, amoxicillin 1 g twice daily and metronidazole 500 mg twice daily for 7 days; EACM group). Urea breath tests were followed up 8 weeks later. RESULTS: The eradication rates for the EA and EACM groups were 91.7% (95% CI = 85.3%-96.0%) and 86.7% (95% CI = 79.3%-92.2%) (P = 0.21) in ITT analysis; and 95.7% (95% CI = 90.2%-98.6%) and 92.0% (95% CI = 85.4%-96.3%) (P = 0.26) in PP analysis. The adverse event rates were 9.6% versus 23.0% in the two groups (P = 0.01). The H. pylori culture positivity rate was 91.8%. The antibiotic resistance rates were amoxicillin, 0%; clarithromycin, 14.6%; and metronidazole, 33.7%. CONCLUSIONS: A 14 day esomeprazole- and amoxicillin-containing high-dose dual therapy achieves a high eradication rate as first-line anti-H. pylori therapy, comparable to that with 7 day non-bismuth quadruple therapy but with fewer adverse events.

The efficacy of limited endoscopic sphincterotomy plus endoscopic papillary large balloon dilation for removal of large bile duct stones.

BACKGROUND: The removal of large bile duct stones (> 15 mm) by conventional endoscopic sphincterotomy (EST) and endoscopic papillary balloon dilation (EPBD) can be challenging, requiring mechanical lithotripsy (ML) in addition to EST or EPBD. The primary complication of ML is basket and stone impaction, which can lead to complications such as pancreatitis and cholangitis. The present study aims to investigate the efficacy of limited EST plus endoscopic papillary large balloon dilation (EST-EPLBD) for large bile duct stone extraction with an extent of cutting < 1/2 the length of the papillary mound. METHODS: We enrolled 185 patients with ≥15 mm bile duct stones who received EST, EPLBD and limited EST-EPLBD treatment from January 1, 2010 to February 28, 2018, at Kaohsiung Chang Gung Memorial Hospital (Kaohsiung, Taiwan). All patients were categorized into three groups: EST group (n = 31), EPLBD group (n = 96), and limited EST-EPLBD group (n = 58). The primary outcome variables were the success rate of complete stone removal and complications. RESULTS: The limited EST-EPLBD group exhibited a higher success rate of the first-session treatment compared with the EST and EPLBD groups (98.3% vs. 83.9% vs. 86.5%; P = 0.032) but required a longer procedure time (32 (12-61) min vs. 23.5 (17-68) min vs. 25.0 (14-60) min; P = 0.001). The need for ML during the procedure was 4 (12.9%) in the EST group, 10 (10.4%) in the EPLBD group and 2 (3.4%) in the limited EST-EPLBD group. Post-procedure bleeding in the EST group was more common than that in the limited EST-EPLBD group (9.7% vs. 0%; P = 0.038). Furthermore, dilated bile duct was the only risk factor for bile duct stone recurrence in the limited EST-EPLBD group. CONCLUSIONS: Limited EST-EPLBD exhibits a higher success rate but requires marginally longer procedure time for the first-session treatment. Furthermore, dilated bile duct is the only risk factor for bile duct stone recurrence in patients undergoing limited EST-EPLBD.

Clinical characteristics and prognosis of HCC occurrence after antiviral therapy for HCV patients between sustained and non-sustained responders.

BACKGROUND: Hepatitis C virus (HCV)-infected patients who achieved sustained virologic response (SVR) may still develop hepatocellular carcinoma (HCC). The characteristic of HCC and the prognosis between SVR and non-SVR patients were not well known. METHODS: Among 1884 HCV-infected patients who were treated with pegylated IFN plus ribavirin therapies, 122 patients developed HCC during follow-up were enrolled in this study. Laboratory data were collected before and at least 1 year after IFN-based therapy, as well as the latest follow-up. RESULTS: Both SVR and non-SVR patients had similar risk factors to develop HCC, but with a little difference. Liver cirrhosis plays a key role in HCC occurrence in both groups. Among the patients who developed HCC, non-SVR patients had significantly higher total bilirubin, higher FIB-4, lower pre-treatment platelet count, higher pre-treatment AFP levels and higher proportion of cirrhosis than SVR patients before occurrence of HCC. After curative treatment, SVR patients had lower recurrence and longer overall survival than non-SVR patients by Kaplan-Meier analysis. Multivariate analysis revealed that APRI ≥0.7 was the independent risk factor for HCC recurrence; and AFP ≥20 ng/ml post IFN therapy, as well as HCC recurrence were the independent risk factors of mortality. CONCLUSION: Liver cirrhosis plays a key role in HCC occurrence after antiviral therapies. SVR patients may have lower HCC recurrence and longer survival rates than non-SVR patients. Only APRI was associated with HCC recurrence; and post-IFN AFP and HCC recurrence were predictive of subsequent mortality independently.

The Role of Non-Selective ß-Blockers in Compensated Cirrhotic Patients without Major Complications.

Background and Objectives: Non-selective ß-blockers (NSBB) could prevent decompensation and hepatocellular carcinoma (HCC) in cirrhotic patients with clinically significant portal hypertension (CSPH), but remained uncertain for compensated cirrhotic patients without major complications. We aimed to compare the clinical outcomes between propranolol users and non-users of a CC group without major complications. Material and Methods: We conducted this population-based cohort study by using the Taiwanese Longitudinal Health Insurance Database 2000. Propranolol users (classified as cumulative defined daily dose (cDDD)) and non-PPL users were matched with a 1:1 propensity score in both cohorts. Results: This study comprised 6896 propranolol users and 6896 non-propranolol users. There was no significant impact on the development of spontaneous bacterial peritonitis between the two groups (aHR: 1.24, 95% confidence interval (CI): 0.88~1.75; p = 0.2111). Male gender, aged condition, and non-liver related diseases (peripheral vascular disease, cerebrovascular disease, dementia, pulmonary disease, and renal disease) were the independent risk factors of mortality. PPL users had significantly lower incidence of HCC development than non-users (aHR: 0.81, p = 0.0580; aHR: 0.80, p = 0.1588; and aHR: 0.49, p < 0.0001 in the groups of 1-28, 29-90, and >90 cDDD, respectively). Conclusion: The current study suggested that high cumulative doses of propranolol could decrease the risk of hepatocellular carcinoma among compensated cirrhotic patients without major complications. Further large-scale prospective studies are still required to confirm the findings in this study. Results: It remained uncertain whether non-selective ß-blockers (NSBB) could prevent decompensation and hepatocellular carcinoma (HCC) in compensatory cirrhotic patients without major complications. This study aimed to compare the clinical outcomes between propranolol users and non-users of the CC group without major complications.

The association between steatosis and diabetes with hepatocellular carcinoma in non-genotype 3 chronic hepatitis C patients.

BACKGROUND & AIMS: Diabetes mellitus (DM) has been found to be strongly associated with an increased risk of hepatocellular carcinoma (HCC) among chronic hepatitis C (CHC) patients. Several studies have also found an association between metabolic steatosis and the risk of HCC in CHC patients, whether this latter association has been accounted for by the known relationship between DM and HCC is still unknown. METHODS: A cohort consisting of 976 non-genotype 3 patients histologically proven to have CHC and treated with interferon and ribavirin was studied. Cumulative incidence and HCC risk were analysed using the Kaplan-Meier method and Cox proportional hazard analysis. RESULTS: Hepatocellular carcinoma developed in 140 subjects over a median follow-up period of 97.3 months, while 699 patients achieved sustained virological response (SVR). According to multivariate analyses, age ≥ 60 years, advanced fibrosis and genotype 1 were identified as independent factors significantly associated with HCC development in SVR patients. Furthermore, using the absence of steatosis and absence of DM as references, the presence of steatosis without DM (HR = 2.09, 95% CI = 1.12-3.9, P = .021), the presence of DM without steatosis (HR = 2.78, 95% CI = 1.3-5.92, P = .008) and the combined presence of steatosis and DM (HR = 3.25, 95% CI = 1.44-7.33, P = .004) were identified as independent factors significantly associated with HCC development in the SVR patients. In contrast, steatosis alone, DM alone and the combined presence of steatosis and DM were not associated with HCC development in non-SVR patients. CONCLUSIONS: Steatosis and DM may be associated with HCC development in non-genotype 3 CHC patients with SVR.

Risk factors for development of hepatocellular carcinoma in patients with chronic hepatitis C without sustained response to combination therapy.

BACKGROUND/PURPOSE: Although antiviral therapy reduces development of hepatocellular carcinoma (HCC) in patients with chronic hepatitis C (CHC), HCC often develops in patients with non-sustained virologic response (non-SVR). We aimed to evaluate risk factors for HCC in HCV patients with non-SVR. METHODS: From March 2002 to December 2013, 800 patients with CHC who had received combined pegylated interferon (peg-IFN)/ribavirin (RBV) therapy without achieving SVR were enrolled. Main outcome measure was HCC development. Variables were cirrhosis, platelet count, α-fetoprotein (AFP) levels, aspartate aminotransferase (AST) to platelet ratio index (APRI), and IL28B polymorphism (CT + TT). RESULTS: One-hundred of 800 non-SVR patients developed HCC within a median 53.5-months follow-up. Cumulative incidence of HCC for all patients was 1.4%, 5.6% and 12.3%, respectively, at 1st, 3rd and 5th years of follow-up. In univariate analysis, patients who developed HCC tended to have LC (p< 0.001), lower platelet counts (<150 × 109/l, p < 0.001), higher AFP levels (≥20 ng/ml, p < 0.001), higher Fib-4 levels (p < 0.001), higher APRI levels (p < 0.001), IL 28B polymorphism (CT + TT) (p < 0.001) and higher incidence of diabetes mellitus (DM) (p = 0.019). Multivariate analysis in overall patients revealed that cirrhosis (HR: 2.94, 95% CI: 1.81-4.77, p < 0.001), IL28B rs12979860 (CT + TT) polymorphisms (HR: 3.22, 95% CI: 2.17-4.78, p < 0.001), and high APRI levels (≥2.57) (HR: 2.32, 95% CI: 1.47-3.67, p < 0.001) were independent risk factors for HCC. CONCLUSION: Liver cirrhosis, high APRI levels, and IL28B rs12979860 at baseline are independent risk factors for HCC development in patients without SVR after peg-IFN combination therapy.

Elevated body mass index is a risk factor associated with possible liver cirrhosis across different etiologies of chronic liver disease.

BACKGROUND/PURPOSE: Metabolic risk factors are associated with liver fibrosis. Whether or not metabolic risk factors affect the severity of liver fibrosis in patients with chronic hepatitis B (CHB), chronic hepatitis C (CHC), or risk factors associated with non-alcoholic fatty liver disease (NAFLD) remains unclear. We aimed to investigate this by transient elastography. METHODS: In this cross-sectional study, we enrolled 1513 patients who presented with chronic liver disease (CLD) at a tertiary hospital. Liver stiffness measurement (LSM) >13 kPa was used as a cutoff suggesting possible liver cirrhosis (LC). RESULTS: Possible LC was noted in 7.8% of the CHB patients, 19.9% of the CHC patients, and 11.9% of the patients with risk factors associated with NAFLD. After adjustments for biochemical and virological factors were made, BMI (per 1 kg/m2 increase) (OR: 1.17, 95% CI: 1.06-1.29, P = 0.002) was found to be an independent factor associated with possible LC in CHB patients; BMI (per 1 kg/m2 increase) (OR: 1.15, 95% CI: 1.07-1.24, P < 0.001) and diabetes mellitus (DM) (OR: 2.32, 95% CI: 1.25-4.30, P = 0.008) were found to be independent factors associated with possible LC in CHC patients; and BMI (per 1 kg/m2 increase) (OR: 1.19, 95% CI: 1.07-1.32, P = 0.002) and DM (OR: 10.35, 95% CI: 2.95-36.32, P < 0.001) were found to be independent factors associated with possible LC in patients with risk factors associated with NAFLD. CONCLUSION: Elevated BMI was an independent risk factor associated with possible LC across the three different etiologies of CLD. As such, weight loss may be beneficial in these patients.

Association between PNPLA3 (rs738409 C>G) variant and hepatocellular carcinoma in Asian chronic hepatitis C patients: A longitudinal study.

BACKGROUND/PURPOSE: Prdevious meta-analyses assess whether or not patatin-like phospholipase domain containing 3 (PNPLA3) (rs738409 C > G) was associated with increased risk of hepatocellular carcinoma (HCC) in Caucasians patients with hepatitis C virus (HCV)-related cirrhosis, these meta-analyses did not provide firm conclusions. Only one cross-sectional study involving Asian patients has previously been conducted to explore this issue. We aim to investigate this in a longitudinal cohort of Asian chronic hepatitis C (CHC) patients. METHODS: We consecutively enrolled 1011 CHC patients who underwent liver biopsy before initiating interferon-based therapy. These patients were followed-up and screened for HCC up to a median of 6.9 years. The influence of rs738409 (GG) genotype on the occurrence of HCC was assessed using the Kaplan-Meier method, then according to the multivariate Cox model. RESULTS: During follow-up, 143 (14.1%) patients developed HCC. rs738409 (GG) genotype was not associated with time-to-HCC development on multivariate Cox regression (P = 0.634). When considering the occurrence of these events over time, rs738409 (GG) genotype did not influence the risk of HCC development (log-rank = 0.12). Among 261 patients with liver cirrhosis, rs738409 (GG) genotype was not associated with time-to-HCC development on multivariate Cox regression (P = 0.737). When considering the occurrence of these events over time, rs738409 (GG) genotype did not influence the risk of HCC development (log-rank = 0.72). CONCLUSION: In this longitudinal study with liver biopsy to stage liver fibrosis, we affirm there is no influence of the rs738409 (GG) genotype on the occurrence of HCC in Asian CHC patients, including cirrhotic patients.

The correlation of controlled attenuation parameter results with ultrasound-identified steatosis in real-world clinical practice.

BACKGROUND/PURPOSE: Controlled attenuation parameter (CAP) is a method for measuring steatosis based on FibroScan. Despite observer dependency, ultrasound (US) robustly diagnoses moderate and severe steatosis. Here, we aimed to evaluate the correlation of CAP with US-identified steatosis in real-world clinical practice. METHODS: CAP and US were performed for 1554 chronic liver disease (CLD) patients. CAP was performed by two technicians, and US was performed by 30 hepatologists. The performance of the CAP as compared with the US results was assessed using the area under the receiver operating characteristic curve (AUROC). RESULTS: 532 (34.2%) of the patients had hepatitis C virus (HCV) infection, 723 (46.5%) of the patients had hepatitis B virus (HBV) infection, and the rest were patients with metabolic risk factors. CAP values were significantly correlated with the steatosis grades identified by US for all the patients (ρ = 0.497, P < 0.001), for the HBV-infected patients (ρ = 0.495, P < 0.001), for the HCV-infected patients (ρ = 0.343, P < 0.001), and for the patients with metabolic risk factors (ρ = 0.515, P < 0.001). Using CAP, the AUROC values were 0.759, 0.795, 0.715, and 0.716 for ≥moderate steatosis identified by US in, respectively, all the patients, the HBV-infected patients, the HCV-infected patients, and the patients with metabolic risk factors. The AUROC values were 0.791, 0.868, 0.807 and 0.701 for severe steatosis identified by US in, respectively, all the patients, the HBV-infected patients, the HCV-infected patients, and the patients with metabolic risk factors. CONCLUSION: CAP values were well correlated with the steatosis grades assessed by US in real-world clinical practice.

Dynamic α-fetoprotein, platelets and AST-to-platelet ratio index predict hepatocellular carcinoma in chronic hepatitis C patients with sustained virological response after antiviral therapy.

BACKGROUND: Hepatitis C virus (HCV)-infected patients who achieve viral eradication may still develop hepatocellular carcinoma (HCC). Little is known about the impact of dynamic change of serum markers on HCC development. METHODS: We enrolled 1351 HCV-infected patients who achieved sustained virological response (SVR). Laboratory data were collected at least 1 year after IFN-based therapy and to the latest follow-up. Data on α-fetoprotein (AFP) were obtained >6 months prior to HCC development to exclude HCC-related AFP elevation. RESULTS: HCC developed in 49 patients. Risk factors for HCC in SVR patients were old age, liver cirrhosis, higher pre- and post-treatment AFP and high post-treatment AST-to-platelet ratio index (APRI). Patients with pre-AFP ≥15 ng/mL → post-AFP ≥15 ng/mL (at 1 year, 23.1%; 5 years, 42.3%) and pre-AFP <15 ng/mL → post-AFP ≥15 ng/mL (at 1 year, 25%; 5 years, 50%) had the highest risk of HCC development, followed by pre-AFP ≥15 ng/mL → post-AFP <15 ng/mL (at 1 year, 5.2%; 5 years, 7.6%) and pre-AFP <15 ng/mL → post-AFP ng/mL <15 ng/mL (at 1 year, 0.5%; 5 years, 0.9%) (P < 0.001). The pattern was similar for platelets and APRI (P < 0.001). SVR patients with pre-APRI ≥0.7 → post-APRI ≥0.7 had the highest risk of HCC development, followed by comparable risks among the other three groups. CONCLUSIONS: SVR patients with a persistently high AFP level (≥15 ng/mL) and a high APRI (≥0.7) before and after treatment had the highest incidence of HCC development. Patients with a reduction of AFP and APRI to the normal range after treatment had a markedly decreased risk of HCC. The risk was lowest for patients who kept persistently normal AFP and APRI before and after treatment.

Limited precut sphincterotomy combined with endoscopic papillary balloon dilation for common bile duct stone removal in patients with difficult biliary cannulation.

BACKGROUND: Difficult biliary cannulation in endoscopic retrograde cholangiopancreatography (ERCP) can result in failure of common bile duct (CBD) stone removal and pancreatitis. The present study aimed to report the efficacy and safety of limited precut sphincterotomy (PS) combined with endoscopic papillary balloon dilation (EPBD) for CBD stone removal in patients with difficult biliary cannulation, and the complications associated with this combined procedure. METHODS: A total of 3305 patients underwent ERCP in our hospital between October 2009 and September 2014 and 258 were diagnosed with difficult biliary cannulation. Of these 258 patients, 58 underwent limited PS combined with EPBD for CBD stone removal, and these 58 patients were included in this retrospective study. RESULTS: The overall success rate was 94.8 % (55/58), and the success rate for single-session removal was 87.9 % (51/58). The mean procedure time was 41 ± 11.48 min (range, 20-72 min). Mechanical lithotripsy was needed in 10.3 % (6/58) of patients. Procedure-related complications included bleeding in 3.4 % (2/58), pancreatitis in 8.6 % (5/58) and biliary tract infection (BTI) in 1.7 % (1/58) of patients. CONCLUSIONS: The therapeutic outcome of limited PS combined with EPBD for CBD stone removal in patients with difficult biliary cannulation was good with an acceptable complication rate. It could be an alternative to PS and "early" limited PS should be used for prompt identification of the bile duct. Limited PS combined with EPBD is safe and effective for CBD stone removal in patients with difficult biliary cannulation.