scholarly journals 5-year Recurrence Prediction After Hepatocellular Carcinoma Resection Using Deep Learning and Cox Regression Models: A Large Prospective Study

2021 ◽  
Author(s):  
Hon-Yi Shi ◽  
King-The Lee ◽  
Chong-Chi Chiu ◽  
Jhi-Joung Wang ◽  
Ding-Ping Sun ◽  
...  
Keyword(s):  
Neural Network ◽  
Hepatocellular Carcinoma ◽  
Deep Learning ◽  
Surgical Resection ◽  
Scale Score ◽  
Cox Regression ◽  
Recurrence Prediction ◽  
Feature Importance ◽  
Importance Analysis ◽  
Hcc Recurrence

Abstract BackgroundRisk of hepatocellular carcinoma (HCC) recurrence after surgical resection is unknown. Therefore, the aim of this study was 5-year recurrence prediction after HCC resection using deep learning and Cox regression models.MethodsThis study recruited 520 HCC patients who had undergone surgical resection at three medical centers in southern Taiwan between April, 2011, and December, 2015. Two popular deep learning algorithms: a deep neural network (DNN) model and a recurrent neural network (RNN) model and a Cox proportional hazard (CPH) regression model were designed to solve both classification problems and regression problems in predicting HCC recurrence. A feature importance analysis was also performed to identify confounding factors in the prediction of HCC recurrence in patients who had undergone resection.ResultsAll performance indices for the DNN model were significantly higher than those for the RNN model and the traditional CPH model (p<0.001). The most important confounding factor in 5-year recurrence after HCC resection was surgeon volume followed by, in order of importance, hospital volume, preoperative Beck Depression Scale score, preoperative Beck Anxiety Scale score, co-residence with family, tumor stage, and tumor size. ConclusionsThe DNN model is useful for early baseline prediction of 5-year recurrence after HCC resection. Its prediction accuracy can be improved by further training with temporal data collected from treated patients. The feature importance analysis performed in this study to investigate model interpretability provided important insights into the potential use of deep learning models for predicting recurrence after HCC resection and for identifying predictors of recurrence.

scholarly journals Surgical Resection Significantly Promotes the Overall Survival of Patients with Hepatocellular Carcinoma: A Propensity Score Matching Analysis

2021 ◽  
Author(s):  
Pei-Min Hsieh ◽  
Hung-Yu Lin ◽  
Chao-Ming Hung ◽  
Gin-Ho Lo ◽  
I-Cheng Lu ◽  
...  
Keyword(s):  
Risk Factors ◽  
Hepatocellular Carcinoma ◽  
Overall Survival ◽  
Propensity Score ◽  
Propensity Score Matching ◽  
Surgical Resection ◽  
Cox Regression ◽  
Survival Rates ◽  
Hbv Infection ◽  
Cox Regression Analysis

Abstract Background: The benefits of surgical resection (SR) for various Barcelona Clinic Liver Cancer (BCLC) stages of hepatocellular carcinoma (HCC) remain unclear. We investigated the risk factors of overall survival (OS) and survival benefits of SR over nonsurgical treatments in patients with HCC of various BCLC stages.Methods: Overall, 2316 HCC patients were included, and their clinicopathological data and OS were recorded. OS was analyzed by the Kaplan-Meier method and Cox regression analysis. Propensity score matching (PSM) analysis was performed.Results: In total, 66 (2.8%), 865 (37.4%), 575 (24.8%) and 870 (35.0%) patients had BCLC stage 0, A, B, and C disease, respectively. Furthermore, 1302 (56.2%) of all patients, and 37 (56.9%), 472 (54.6%), 313 (54.4%) and 480 (59.3%) of patients with BCLC stage 0, A, B, and C disease, respectively, died. The median follow-up duration time was 20 (range 0-96) months for the total cohort and was subdivided into 52 (8-96), 32 (1-96), 19 (0-84), and 12 (0-79) months for BCLC stages 0, A, B, and C cohorts, respectively. The risk factors for OS were 1) SR and cirrhosis; 2) SR, cirrhosis, and Child-Pugh (C-P) class; 3) SR, hepatitis B virus (HBV) infection, and C-P class; and 4) SR, HBV infection, and C-P class for the BCLC stage 0, A, B, and C cohorts, respectively. Compared to non-SR treatment, SR resulted in significantly higher survival rates in all cohorts. The 5-year OS rates for SR vs non-SR were 44.0% vs 28.7%, 72.2% vs 42.6%, 42.6% vs 36.2, 44.6% vs 23.5%, and 41.4% vs 15.3% (all p-values<0.05) in the total and BCLC stage 0, A, B, and C cohorts, respectively. After PSM, SR resulted in significantly higher survival rates compared to non-SR treatment in various BCLC stages.Conclusion: SR conferred significant survival benefits to patients with HCC of various BCLC stages and should be considered a recommended treatment for select HCC patients, especially patients with BCLC stage B and C disease.

scholarly journals mTOR Expression in Liver Transplant Candidates with Hepatocellular Carcinoma: Impact on Histological Features and Tumour Recurrence

2019 ◽  
Vol 20 (2) ◽  
pp. 336 ◽  
Author(s):  
Marta Guerrero ◽  
Gustavo Ferrín ◽  
Manuel Rodríguez-Perálvarez ◽  
Sandra González-Rubio ◽  
Marina Sánchez-Frías ◽  
...  
Keyword(s):  
Hepatocellular Carcinoma ◽  
Liver Transplant ◽  
Cox Regression ◽  
Cell Signalling ◽  
Mtor Pathway ◽  
Tumour Recurrence ◽  
Recurrence Rates ◽  
Explanted Liver ◽  
Peritumoral Tissue ◽  
Hcc Recurrence

(1) Background: The mammalian target of rapamycin (mTOR) pathway activation is critical for hepatocellular carcinoma (HCC) progression. We aimed to evaluate the mTOR tissue expression in liver transplant (LT) patients and to analyse its influence on post-LT outcomes. (2) Methods: Prospective study including a cohort of HCC patients who underwent LT (2012–2015). MTOR pathway expression was evaluated in the explanted liver by using the “PathScan Intracellular Signalling Array Kit” (Cell Signalling). Kaplan-Meier and Cox regression analyses were performed to evaluate post-LT HCC recurrence. (3) Results: Forty-nine patients were included (average age 56.4 ± 6, 14.3% females). Phospho-mTOR (Ser2448) was over-expressed in peritumoral tissue as compared with tumoral tissue (ΔSignal 22.2%; p < 0.001). The mTOR activators were also increased in peritumoral tissue (phospho-Akt (Thr308) ΔSignal 18.2%, p = 0.004; phospho-AMPKa (Thr172) ΔSignal 56.3%, p < 0.001), as they were the downstream effectors responsible for cell growth/survival (phospho-p70S6K (Thr389) ΔSignal 33.3%, p < 0.001 and phospho-S6RP (Ser235/236) ΔSignal 54.6%, p < 0.001). MTOR expression was increased in patients with multinodular HCC (tumoral p = 0.01; peritumoral p = 0.001). Increased phospho-mTOR in tumoral tissue was associated with higher HCC recurrence rates after LT (23.8% vs. 5.9% at 24 months, p = 0.04). (4) Conclusion: mTOR pathway is over-expressed in patients with multinodular HCC and is it associated with increased post-LT tumour recurrence rates.

Deep learning to develop transcriptomic model for survival prediction in TCGA patients with hepatocellular carcinoma.

2020 ◽  
Vol 38 (15_suppl) ◽  
pp. e14057-e14057
Author(s):  
Hao Yu ◽  
Wei Dai ◽  
Chi Leung Chiang ◽  
Shisuo Du ◽  
Zhao-Chong Zeng ◽  
...  
Keyword(s):  
Hepatocellular Carcinoma ◽  
Deep Learning ◽  
Creatinine Level ◽  
Clinical Data ◽  
Cox Regression ◽  
P Value ◽  
Survival Prediction ◽  
Signal Pathways ◽  
Test Set ◽  
Hidden Layer

e14057 Background: This study aimed to investigate the prognostic value of transcriptome and clinical data of Hepatocellular carcinoma (HCC) patients for overall survival (OS) by deep learning method. Methods: A total of 371 HCC patients with 20530 level three RNA-sequencing data were from The Cancer Genome Atlas (TCGA). Cox-nnet model, a deep learning model through an artificial neural network extension of the Cox regression model, was used for OS prediction. The patients were randomly split into train-set and test-set (7:3). In train-set, the significant genes associated with OS under univariate Cox regression were considered for modeling. Clinical parameters, including age, gender, pathologic stage, child pugh classification, creatinine level etc. were also considered. The Cox-nnet model was developed by cross-validation. Its discrimination was determined by the concordance index (CI) in the independent test-set and compared with multivariable Cox regression. The clustering method Uniform Manifold Approximation and Projection (UMAP) was used for revealing biological information from the hidden layer in the model. Results: In the train-set (n = 259), 1505 genes and two clinical variables (child pugh score and creatinine level) were significantly associated with OS (adjusted P-value < 0.05). To avoid overfitting, only 40 most significant genes were included in the Cox-nnet model. In the test-set (n = 112), the CI of Cox-nnet (0.76, se = 0.04) is better than the CI of multivariable Cox regression (0.71, se = 0.05). The difference between good or poor survival subgroups classified by Cox-nnet was remarkably significant ( P-value = 1e-4, median OS: 80.7 vs. 25.1 months). In the Cox-nnet model with all significant variables, the weights in the hidden layer were clustered by UMAP into 3 positive clusters and 2 negative clusters, which are enriched in GO/KEGG. The “cell cycle” and “complement and coagulation cascades” are the most important signal pathways in positive and negative clusters, respectively. Conclusions: Combining transcriptomic and clinical data, and with deep learning algorithm, we built and validated a robust model for survival prediction in HCC patients. Our study would be useful to explore the clinical implications in survival prediction and corresponding genetic mechanisms. Clinical trial information: 5U24CA143799, 5U24CA143835, 5U24CA143840, 5U24CA143843, 5U24CA143845, 5U24CA143848, 5U24CA1438.

scholarly journals Predicting Hepatocellular Carcinoma With Minimal Features From Electronic Health Records: Development of a Deep Learning Model

JMIR Cancer ◽  
10.2196/19812 ◽  
2021 ◽  
Vol 7 (4) ◽  
pp. e19812
Author(s):  
Chia-Wei Liang ◽  
Hsuan-Chia Yang ◽  
Md Mohaimenul Islam ◽  
Phung Anh Alex Nguyen ◽  
Yi-Ting Feng ◽  
...  
Keyword(s):  
Neural Network ◽  
Hepatocellular Carcinoma ◽  
Deep Learning ◽  
Electronic Health Records ◽  
Network Model ◽  
Neural Network Model ◽  
Learning Model ◽  
Health Records ◽  
Electronic Health ◽  
Deep Learning Model

Background Hepatocellular carcinoma (HCC), usually known as hepatoma, is the third leading cause of cancer mortality globally. Early detection of HCC helps in its treatment and increases survival rates. Objective The aim of this study is to develop a deep learning model, using the trend and severity of each medical event from the electronic health record to accurately predict the patients who will be diagnosed with HCC in 1 year. Methods Patients with HCC were screened out from the National Health Insurance Research Database of Taiwan between 1999 and 2013. To be included, the patients with HCC had to register as patients with cancer in the catastrophic illness file and had to be diagnosed as a patient with HCC in an inpatient admission. The control cases (non-HCC patients) were randomly sampled from the same database. We used age, gender, diagnosis code, drug code, and time information as the input variables of a convolution neural network model to predict those patients with HCC. We also inspected the highly weighted variables in the model and compared them to their odds ratio at HCC to understand how the predictive model works Results We included 47,945 individuals, 9553 of whom were patients with HCC. The area under the receiver operating curve (AUROC) of the model for predicting HCC risk 1 year in advance was 0.94 (95% CI 0.937-0.943), with a sensitivity of 0.869 and a specificity 0.865. The AUROC for predicting HCC patients 7 days, 6 months, 1 year, 2 years, and 3 years early were 0.96, 0.94, 0.94, 0.91, and 0.91, respectively. Conclusions The findings of this study show that the convolutional neural network model has immense potential to predict the risk of HCC 1 year in advance with minimal features available in the electronic health records.

scholarly journals Comparison of overall survival on surgical resection versus transarterial chemoembolization with or without radiofrequency ablation in intermediate stage hepatocellular carcinoma: A propensity score matching analysis

2020 ◽  
Author(s):  
Chih-Wen Lin ◽  
Yaw-Sen Chen ◽  
Gin-Ho Lo ◽  
Yao-Chun Hsu ◽  
Chia-Chang Hsu ◽  
...  
Keyword(s):  
Hepatocellular Carcinoma ◽  
Overall Survival ◽  
Radiofrequency Ablation ◽  
Propensity Score ◽  
Propensity Score Matching ◽  
Surgical Resection ◽  
Cox Regression ◽  
Cox Regression Analysis ◽  
Tace Group ◽  
The Impact

Abstract Background: Patients with Barcelona Clinic Liver Cancer (BCLC) stage B hepatocellular carcinoma (HCC) are recommended to undergo transcatheter arterial chemoembolization (TACE). However, TACE in combination with radiofrequency ablation (RFA) is not inferior to surgical resection (SR), and the benefits of surgical resection (SR) for BCLC stage B HCC remain unclear. Hence, this study aims to compare the impact of SR, TACE+RFA, and TACE on analyzing overall survival (OS) in BCLC stage B HCC. Methods: Overall, 428 HCC patients were included in BCLC stage B, and their clinical data and OS were recorded. OS was analyzed by the Kaplan-Meier method and Cox regression analysis. Results: One hundred forty (32.7%) patients received SR, 231 (53.9%) received TACE+RFA, and 57 (13.3%) received TACE. The OS was significantly higher in the SR group than that in the TACE+RFA group [hazard ratio (HR): 1.78; 95% confidence incidence (CI): 1.15-2.75, p=0.009]. The OS was significantly higher in the SR group than that in the TACE group (HR: 3.17; 95% CI: 2.31-4.36, p<0.0001). Moreover, the OS was significantly higher in the TACE+RFA group than that in the TACE group (HR: 1.82; 95% CI: 1.21-2.74, p=0.004). The cumulative OS rates at 1, 3 and 5 years in the SR, TACE+RFA, and TACE groups were 89.2%, 69.4% and 61.2%, 86.0%, 57.9% and 38.2%, and 69.5%, 37.0% and 15.2%, respectively. After propensity score matching, the SR group still had a higher OS than those of the TACE+RFA and TACE groups. The TACE+RFA group had a higher OS than that of the TACE group. Conclusion: The SR group had higher OS than the TACE+RFA and TACE groups in BCLC stage B HCC. Furthermore, the TACE+RFA group had higher OS than the TACE group.

scholarly journals Factors influencing early recurrence of hepatocellular carcinoma after curative resection

2020 ◽  
Vol 48 (8) ◽  
pp. 030006052094555
Author(s):  
Yu Zhu ◽  
Lingling Gu ◽  
Ting Chen ◽  
Guoqun Zheng ◽  
Chao Ye ◽  
...  
Keyword(s):  
Hepatocellular Carcinoma ◽  
Curative Resection ◽  
Cox Regression ◽  
Postoperative Recurrence ◽  
Early Recurrence ◽  
Tumor Diameter ◽  
Cox Regression Analysis ◽  
Factors Influencing ◽  
Linear Manner ◽  
Hcc Recurrence

Objective To identify the factors influencing early recurrence in patients with hepatocellular carcinoma (HCC) after curative resection. Methods Clinical data for 99 patients with HCC undergoing curative resection were analyzed. The clinicopathological factors influencing early recurrence were analyzed by Cox regression. Results Twenty-five of 99 patients (25.3%) suffered from early recurrence. There were significant differences between patients with and without recurrence in terms of tumor diameter, tumor capsular integrity, and preoperative alpha fetoprotein level. Cox regression analysis revealed that a tumor diameter >2.6 cm and preoperatively increased total bilirubin (TBL) level were risk factors for postoperative recurrence, while tumor capsular integrity had a protective effect on postoperative recurrence. After adjusting for preoperative TBL level and tumor capsular integrity, the risk of HCC recurrence was markedly increased in line with increasing tumor diameter in a non-linear manner. Conclusion Tumor diameter >2.6 cm and preoperatively increased TBL level are associated with a higher risk of early recurrence after curative resection in patients with HCC, while tumor capsular integrity is associated with a lower risk of early recurrence.

scholarly journals Online Critical Unit Detection and Power System Security Control: An Instance-Level Feature Importance Analysis Approach

2021 ◽  
Vol 11 (12) ◽  
pp. 5460
Author(s):  
Junyu Ren ◽  
Li Wang ◽  
Shaofan Zhang ◽  
Yanchun Cai ◽  
Jinfu Chen
Keyword(s):  
Deep Learning ◽  
Power Systems ◽  
Operating Conditions ◽  
Security Assessment ◽  
Dynamic Security ◽  
Security Control ◽  
Feature Importance ◽  
Importance Analysis ◽  
Viable Approach ◽  
Critical Unit

Rapid and accurate detection of critical units is crucial for the security control of power systems, ensuring reliable and continuous operation. Inspired by the advantages of data-driven techniques, this paper proposes an integrated deep learning framework of dynamic security assessment, critical unit detection, and security control. In the proposed framework, a black-box deep learning model is utilized to evaluate the dynamic security of power systems. Then, the predictions of the model for specific operating conditions are interpreted by instance-level feature importance analysis. Furthermore, the critical units are detected by reasonable local interpretation, and the security control scheme is extracted with a sequential adjustment strategy according to the results of interpretation. The numerical simulations on the CEPRI36 benchmark system and the IEEE 118-bus system verified that our proposed framework is fast and accurate for specific operating conditions and, thereby, is a viable approach for online security control of power systems.

scholarly journals Hepatitis B virus pre-S2 deletion (nucleotide 1 to 54) in plasma predicts recurrence of hepatocellular carcinoma after curative surgical resection

PLoS ONE ◽  
2020 ◽  
Vol 15 (11) ◽  
pp. e0242748
Author(s):  
Chiao-Fang Teng ◽  
Tsai-Chung Li ◽  
Hsi-Yuan Huang ◽  
Wen-Ling Chan ◽  
Han-Chieh Wu ◽  
...  
Keyword(s):  
Hepatocellular Carcinoma ◽  
Hepatitis B Virus ◽  
Hepatitis B ◽  
Surgical Resection ◽  
Deletion Mutations ◽  
B Virus ◽  
Improve Patient ◽  
Generation Sequencing ◽  
Hcc Recurrence ◽  
Poor Patient Survival

Hepatocellular carcinoma (HCC) is a leading cause of cancer-related death worldwide. Despite curative surgical resection, high recurrence of HCC after surgery results in poor patient survival. To develop prognostic markers is therefore important for better prevention and therapy of recurrent HCC to improve patient outcomes. Deletion mutations over the pre-S1 and pre-S2 gene segments of hepatitis B virus (HBV) have been closely associated with recurrence of HCC after curative surgical resection. In this study, we applied a next-generation sequencing-based approach to further evaluate the association of pre-S deletion regions with HCC recurrence. We demonstrated that the pre-S2 deletion (nucleotide 1 to 54) was the most predominant deletion regions of pre-S gene in plasma of HBV-related HCC patients. Moreover, patients with the pre-S2 deletion (nucleotide 1 to 54) exhibited a significantly higher risk of HCC recurrence after curative surgical resection than those without. The pre-S2 deletion (nucleotide 1 to 54) in plasma represented a prognostic factor that independently predicted HCC recurrence with greater performance than other clinicopathological and viral factors. Our data suggest that detection of the pre-S2 deletion (nucleotide 1 to 54) in plasma may be a promising noninvasive strategy for identifying patients at high risk for HCC recurrence after curative surgical resection.

scholarly journals A Five-Gene Signature for Recurrence Prediction of Hepatocellular Carcinoma Patients

2020 ◽  
Vol 2020 ◽  
pp. 1-13
Author(s):  
Zeyu Wang ◽  
Ningning Zhang ◽  
Jiayu Lv ◽  
Cuihua Ma ◽  
Jie Gu ◽  
...  
Keyword(s):  
Hepatocellular Carcinoma ◽  
Regression Analysis ◽  
Cox Regression ◽  
Early Stage ◽  
Gene Signature ◽  
Functional Enrichment ◽  
High Recurrence Rate ◽  
Lasso Regression ◽  
Cox Regression Analysis ◽  
Recurrence Prediction

Background. Hepatocellular carcinoma (HCC) is one of the most aggressive malignancies with poor prognosis. There are many selectable treatments with good prognosis in Barcelona Clinic Liver Cancer- (BCLC-) 0, A, and B HCC patients, but the most crucial factor affecting survival is the high recurrence rate after treatments. Therefore, it is of great significance to predict the recurrence of BCLC-0, BCLC-A, and BCLC-B HCC patients. Aim. To develop a gene signature to enhance the prediction of recurrence among HCC patients. Materials and Methods. The RNA expression data and clinical data of HCC patients were obtained from the Gene Expression Omnibus (GEO) database. Univariate Cox regression analysis and least absolute shrinkage and selection operator (LASSO) regression analysis were conducted to screen primarily prognostic biomarkers in GSE14520. Multivariate Cox regression analysis was introduced to verify the prognostic role of these genes. Ultimately, 5 genes were demonstrated to be related with the recurrence of HCC patients and a gene signature was established. GSE76427 was adopted to further verify the accuracy of gene signature. Subsequently, a nomogram based on gene signature was performed to predict recurrence. Gene functional enrichment analysis was conducted to investigate the potential biological processes and pathways. Results. We identified a five-gene signature which performs a powerful predictive ability in HCC patients. In the training set of GSE14520, area under the curve (AUC) for the five-gene predictive signature of 1, 2, and 3 years were 0.813, 0.786, and 0.766. Then, the relative operating characteristic (ROC) curves of five-gene predictive signature were verified in the GSE14520 validation set, the whole GSE14520, and GSE76427, showed good performance. A nomogram comprising the five-gene signature was built so as to show a good accuracy for predicting recurrence-free survival of HCC patients. Conclusion. The novel five-gene signature showed potential feasibility of recurrence prediction for early-stage HCC.

Sign in / Sign up

Export Citation Format

Share Document