Meta-learning algorithm development to predict outcomes in patients with hepatitis C virus-related hepatocellular carcinoma.

BACKGROUND AND STUDY AIMS: Prediction of prognosis and treatment outcomes for patients with hepatocellular carcinoma (HCC) is complex for most patients. Machine learning predictive analysis can be used to explore the rich information in electronic health records to discover hidden patterns and relationships. We aimed to develop a noninvasive algorithm for predicting outcome treatment options for patients with HCC. PATIENTS AND METHODS: This cross-sectional study included 1298 patients with Hepatitis C virus-related HCC attending an HCC multidisciplinary clinic, Kasr Al-Aini Hospital, Cairo University, between 2009 and 2016. Using machine learning analysis, we constructed Reduced Error Pruning (REP) decision tree algorithms and applied Auto-WEKA to select the best classifier out of 39 algorithms. RESULTS: The REP-tree algorithm predicted HCC management outcomes with a recall (sensitivity) of 0.658 and a precision (specificity) of 0.653 using only routine data. 854 (65.8%) instances were correctly identified, and 444 (34.2%) instances were incorrectly classified. Out of 31 attributes, liver decompensation was selected by REP-tree as the best predictor of HCC outcome (root node). With Auto-WEKA, the random subspace classifier was chosen as the best predictive algorithm with a recall (sensitivity) of 0.750 and a precision (specificity) of 0.75. There were 974 (75%) correctly classified instances and 324 (25%) incorrectly classified instances, which was better than REP-tree. CONCLUSION: Machine learning analysis explores data to discover hidden patterns and trends and enables the development of models to predict HCC treatment outcomes utilizing simple laboratory data. The random subspace classifier predicted the outcome more accurately than REP-tree.

Predictors of recurrence and survival of hepatocellular carcinoma: A prospective study including transient elastography and cancer stem cell markers.

BACKGROUND AND STUDY AIMS: To investigate whether the measurement of liver stiffness (LSM) using fibroscan and the serum Cancer Stem Cells (CSC): Ep-CAM and cytokeratin-19, could predict the recurrence of hepatocellular carcinoma (HCC) and their impact on clinical outcome and overall survival. PATIENTS AND METHODS: This is a prospective study, including 179 HCV-related HCC patients. All patients were treated following the BCLC guidelines. All HCC patients had transient elastography, measurements of Ep-CAM and cytokeratin-19 before and six months post-treatment. We looked for predictors of recurrence and performed a survival analysis using Kaplan-Meier estimates. RESULTS: TACE was the most common procedure (77.1%), followed by microwave ablation (15.6%). Complete ablation was achieved in 97 patients; 55 of them developed HCC recurrence. After treatment, LSM increased significantly with a significant reduction in CSCs levels in complete and partial response groups. The median time to observe any recurrence was 14 months. LSM increased significantly post-treatment in patients with recurrence versus no recurrence. Higher levels of CSCs were recorded at baseline and post-treatment in patients with recurrence but without statistical significance. We used univariate analysis to predict the time of recurrence by determining baseline CK-19 and platelet levels as the key factors, while the multivariate analysis determined platelet count as a single factor. The univariate analysis for prediction of overall survival included several factors, LSM and EpCAM (baseline and post-ablation) among them, while multivariate analysis included factors such as Child score B and incomplete ablation. CONCLUSION: Dynamic changes were observed in LSM and CSCs levels in response to HCC treatment and tumour recurrence. Child score and complete ablation are factors that significantly affect survival.