Precision Strike Strategy for Liver Diseases Trilogy with Xiao-Chai-Hu Decoction: A Meta-Analysis with Machine
Zexin Wang1, Yubing Li1, Xiaobao Wang2
1State Key Laboratory of Southwestern Chinese Medicine Resources, School of Pharmacy, Chengdu University of Traditional Chinese Medicine, Chengdu, China.
Background And Purpose:
The progression from hepatitis to liver fibrosis (LF) and ultimately to hepatic carcinoma (HCC) represents the advanced stages of various liver diseases. Currently, no universal treatment effectively addresses all three conditions. The Traditional Chinese Medicine formula Xiao-Chai-Hu decoction (XCHD) has shown promise in treating hepatitis, inhibiting LF, and serving as an adjunct therapy for HCC. This study evaluates the efficacy and optimal treatment durations of XCHD in managing these liver diseases using meta-analysis and machine learning techniques.
Methods:
Registered in the PROSPERO database (CRD42024534445), this meta-analysis systematically searched seven databases, including 54 studies with a total of 5,710 patients. Statistical analysis was performed using Stata 17.0. Five machine learning models-Random Forest (RF), XGBoost, Lasso, Multilayer Perceptron (MLP), and a stacking model combining these algorithms-were employed to analyze the data and predict the time-effect relationships. The optimal durations of XCHD treatment for the liver disease trilogy were subsequently projected.
Results:
XCHD significantly improved the primary outcome indicators for hepatitis, liver fibrosis, and HCC. Additionally, XCHD demonstrated a beneficial effect on liver dysfunction caused by these diseases. Machine learning predicted the optimal treatment durations of XCHD as 12 weeks for hepatitis, 20.31 weeks for liver fibrosis, and 12 weeks for HCC.
Conclusion:
XCHD is effective in treating the liver disease trilogy, with optimal treatment durations of 12 weeks for hepatitis and HCC, and 20.31 weeks for liver fibrosis. These findings support the potential of XCHD in developing precise clinical strategies for managing liver diseases. This study innovatively integrates meta-analysis with machine learning to determine the optimal treatment durations, providing a novel approach for evidence-based precision medicine in Traditional Chinese Medicine.
