5-year recurrence prediction after hepatocellular carcinoma resection: deep learning vs. Cox regression models.

Deep learning algorithms have yet to be used for predicting clinical prognosis after cancer surgery. Therefore, this study compared performance indices and permutation importance of potential confounders in three models for predicting 5-year recurrence after hepatocellular carcinoma (HCC) resection: a deep-learning deep neural network (DNN) model, a recurrent neural network (RNN) model, and a Cox proportional hazard (CPH) regression model. Data for 725 patients who had received HCC resection at three medical centers in southern Taiwan between April, 2011, and December, 2015, were randomly divided into three datasets: a training dataset containing data for 507 subjects was used for model development, a testing dataset containing data for 109 subjects was used for internal validation, and a validating dataset containing data for 109 subjects was used for external validation. Feature importance analysis was also performed to identify potential predictors of recurrence after HCC resection. Univariate Cox proportional hazards regression analyses were performed to identify potential significant predictors of 5-year recurrence after HCC resection, which were included in the forecasting models (P < 0.05). All performance indices for the DNN model were significantly higher than those for the RNN model and the conventional CPH model (P < 0.001). The most important potential predictor of 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. The feature importance analysis performed to investigate interpretability in this study elucidated the potential use of deep learning models for predicting recurrence after HCC resection and for identifying predictors of recurrence. Further experiments using the proposed DNN model would clarify its potential uses for developing, promoting, and improving health policies for treating HCC patients after surgery.

Cloud-Based Service Information System for Evaluating Quality of Life after Breast Cancer Surgery.

OBJECTIVE: Although recent studies have improved understanding of quality of life (QOL) outcomes of breast conserving surgery, few have used longitudinal data for more than two time points, and few have examined predictors of QOL over two years. Additionally, the longitudinal data analyses in such studies rarely apply the appropriate statistical methodology to control for censoring and inter-correlations arising from repeated measures obtained from the same patient pool. This study evaluated an internet-based system for measuring longitudinal changes in QOL and developed a cloud-based system for managing patients after breast conserving surgery. METHODS: This prospective study analyzed 657 breast cancer patients treated at three tertiary academic hospitals. Related hospital personnel such as surgeons and other healthcare professionals were also interviewed to determine the requirements for an effective cloud-based system for surveying QOL in breast cancer patients. All patients completed the SF-36, Quality of Life Questionnaire (QLQ-C30) and its supplementary breast cancer measure (QLQ-BR23) at baseline, 6 months, 1 year, and 2 years postoperatively. The 95% confidence intervals for differences in responsiveness estimates were derived by bootstrap estimation. Scores derived by these instruments were interpreted by generalized estimating equation before and after surgery. RESULTS: All breast cancer surgery patients had significantly improved QLQ-C30 and QLQ-BR23 subscale scores throughout the 2-year follow-up period (p<0.05). During the study period, QOL generally had a negative association with advanced age, high Charlson comorbidity index score, tumor stage III or IV, previous chemotherapy, and long post-operative LOS. Conversely, QOL was positively associated with previous radiotherapy and hormone therapy. Additionally, patients with high scores for preoperative QOL tended to have high scores for QLQ-C30, QLQ-BR23 and SF-36 subscales. Based on the results of usability testing, the five constructs were rated on a Likert scale from 1-7 as follows: system usefulness (5.6±1.8), ease of use (5.6±1.5), information quality (5.4±1.4), interface quality (5.5±1.4), and overall satisfaction (5.5±1.6). CONCLUSIONS: The current trend in clinical medicine is applying therapies and interventions that improve QOL. Therefore, a potentially vast amount of internet-based QOL data is available for use in defining patient populations that may benefit from therapeutic intervention. Additionally, before undergoing breast conserving surgery, patients should be advised that their postoperative QOL depends not only on the success of the surgery, but also on their preoperative functional status.

Survival comparison between surgical resection and percutaneous radiofrequency ablation for patients in Barcelona Clinic Liver Cancer early stage hepatocellular carcinoma.

AIM: To compare the survival outcome between surgical resection (SR) and radiofrequency ablation (RFA) for Barcelona Clinic Liver Cancer (BCLC) early stage hepatocellular carcinoma (HCC). METHODS: The retrospective study enrolled eighty-two patients with newly diagnosed BCLC early HCC (single nodule, size ≦3 cm, and Child-Pugh class A) treated either surgically (n = 46) or with RFA (n = 36) from year 2004 to 2009. The patients' survival outcomes were compared. RESULTS: There were no significant differences in overall survival (OS) rates between SR and RFA (p = 0.204). The 3- and 5-year disease-free survival (DFS) rates were 65.8 % and 53.7 % respectively, in the SR group, which were significantly higher than those in the RFA group (34.8 % and 14.9 % respectively) (p = 0.009 and p = 0.001). In subgroup analysis, the DFS was similar between RFA and SR in patients with presentation of lower platelet count (≦100,000/mL) and smaller tumor size (tumor size ≦1 cm). Multivariate analysis showed SR as a procedure type was a significant predictive factor for DFS [HR = 2.26 (CI 1.462-5.227), p = 0.002]. CONCLUSION: SR yielded similar OS but better DFS when compared to RFA for patients with BCLC early HCC (single nodule, ≦3 cm and Child-Pugh class A). In subgroup patients with lower platelet count (≦100,000/mL) and smaller tumor size (tumor size ≦1 cm), DFS was similar between both treatments.

The activation of MEK/ERK signaling pathway by bone morphogenetic protein 4 to increase hepatocellular carcinoma cell proliferation and migration.

Hepatocellular carcinoma (HCC) is one of the most common visceral malignancies worldwide, with a very high incidence and poor prognosis. Bone morphogenesis protein 4 (BMP4), which belongs to the TGF-ß superfamily of proteins, is a multifunctional cytokine, which exerts its biologic effects through SMAD- and non-SMAD-dependent pathways, and is also known to be involved in human carcinogenesis. However, the effects of the BMP4 signaling in liver carcinogenesis are not yet clearly defined. Here, we first show that BMP4 and its receptor, BMPR1A, are overexpressed in a majority of primary HCCs and that it promotes the growth and migration of HCC cell lines in vitro. We also establish that BMP4 can induce HCC cyclin-dependent kinase (CDK)1 and cyclin B1 upregulation to accelerate cell-cycle progression. Our study indicates that the induction of HCC cell proliferation is independent of the SMAD signaling pathway, as Smad4 knockdown of HCC cell lines still leads to the upregulation of CDK1 and cyclin B1 expression after BMP4 treatment. Using mitogen-activated protein/extracellular signal-regulated kinase (MEK) selective inhibitors, the induction of CDK1, cyclin B1 mRNA and protein were shown to be dependent on the activation of MEK/extracellular signal-regulated kinase (ERK) signaling. In vivo xenograft studies confirmed that the BMPR1A-knockdown cells were significantly less tumorigenic than the control groups. Our findings show that the upregulation of BMP4 and BMPR1A in HCC promotes the proliferation and metastasis of HCC cells and that CDK1 and cyclin B1 are important SMAD-independent molecular targets in BMP4 signaling pathways, during the HCC tumorigenesis. It is proposed that BMP4 signaling pathways may have potential as new therapeutic targets in HCC treatment.

Evaluation of redox statuses in patients with hepatitis B virus-associated hepatocellular carcinoma.

BACKGROUND: Excess reactive oxygen species related to neoplasia of liver has been established. Essentially, the human body has developed different antioxidant systems for defence against these attacks. To evaluate the redox status in hepatocellular carcinoma (HCC) induced by hepatitis B virus (HBV), the most important aetiological factor in Taiwan, changes in O2(.) generation, lipid peroxidation as well as antioxidant status in the blood of HCC patients with HBV carriers for more than 20 years were measured. METHODS: Superoxide anion radical (O2(.-)) generation and the levels of malondialdehyde (MDA) served as an index of lipid peroxidation along with the analyses of activities of superoxide dismutase (SOD), glutathione peroxidase (GPx) and glutathione reductase (GRx); also, glutathione status, including reduced glutathione (GSH) and oxidized glutathione (GSSG), and the levels of vitamins A, C and E were determined. RESULTS: In 54 patients, the levels of O2(.-), MDA and GSSG, and the activities of SOD and GRx of blood were significantly higher than those of 57 controls. Conversely, the levels of GSH and total GSH, and GSH/GSSG ratio, and vitamins A and C were significantly decreased. Additionally, there were no significant changes in the activity of GPx and the levels of vitamin E. CONCLUSIONS: Our data suggest that the redox statuses in patients with HBV-associated HCC were elevated or decreased in certain parameters. However, the increased activities of antioxidant enzymes may be a compensatory up-regulation and the decrease antioxidant statuses were responses to the enhanced oxidative stress in those patients.