Comparison of Risk Factors, Their Interaction Patterns, and Scoring Systems for Liver Cancer Between Patients With and Those Without Diabetes: Retrospective Cohort Study Using Electronic Health Records and Tree-Structured Algorithms

TL;DR

This study compares liver cancer risk factors and scoring systems between people with and without diabetes, revealing distinct patterns that could improve cancer risk prediction and public health strategies.

Contribution

The study introduces diabetes-specific interaction patterns and scoring systems for liver cancer risk using tree-structured algorithms.

Findings

Alanine aminotransferase (ALT), age, sex, and triglycerides are common predictors across all groups.

Chronic viral hepatitis is a strong risk factor in diabetes but not in non-diabetes patients.

Interaction patterns differ by age, sex, and statin use, influencing liver cancer risk prediction.

Abstract

Patients with diabetes are at higher risk of developing liver cancer. Nevertheless, risk factors and their interaction patterns have rarely been compared between patients with and those without diabetes, nor have their interactions been incorporated into scoring system development.

This study aims to compare risk factors, their interaction patterns, and resulting scoring systems for liver cancer risk according to diabetes and liver disease status using tree-structured algorithms.

A retrospective cohort study was conducted using electronic health records in Hong Kong. Patients who had used public health care services between 1997 and 2021 without cancer history were identified and followed up until December 31, 2021. Scoring systems were developed based on aggregate results from individual survival trees in random survival forest, and interaction patterns among factors were separately examined using conditional inference survival tree.

Of the 190,971 patients included, 1275 developed liver cancer during follow-up (median 6.25 y). Across 4 scoring systems, alanine aminotransferase (ALT) levels, age, sex, and triglycerides were commonly chosen as predictors irrespective of diabetes and liver disease status. In the overall systems, liver cirrhosis was additionally selected as a predictor, with chronic viral hepatitis uniquely chosen in diabetes. In the absence of liver disease, fasting glucose and smoking were uniquely selected for diabetes and nondiabetes, respectively. Chronic viral hepatitis appeared as the strongest risk factor in diabetes but not in nondiabetes. Among people with diabetes, in the absence of chronic viral hepatitis, sex became the most important factor, followed by age, statin use, and ALT levels. Among people without diabetes, age became the most dominant risk factor. For older patients (>55 y), uncontrolled lipids and male sex became key risk factors in statin and nonstatin users, respectively, when the ALT level was higher (>43.4 U/L), while smoking became a key risk factor when the ALT level was lower (≤43.4 U/L). For younger patients (≤55 y), sex remained the most significant factor.

Patients with and those without diabetes exhibit distinctive interaction patterns among key factors on liver cancer risk. The resulting scoring systems reflect interaction patterns among predictors in individual survival trees. This study may help identify targets for public health interventions and provide clinical cancer risk prediction according to diabetes status.

Figures