Scholarlink

Purpose <p>Hepatitis B virus (HBV) infection is a major risk factor for hepatocellular carcinoma (HCC). Although nucleoside/nucleotide analogue (NA) therapy significantly reduces this risk, a subset of patients still develop HCC. This study aimed to identify high-risk groups by evaluating the comprehensive utility of multidimensional factor combinations.</p> Methods <p>We analyzed 565 patients with chronic HBV infection who underwent liver stiffness (LS) measurement between 2009 and 2024. Independent risk factors for HCC were identified using multivariate logistic regression. In 84 patients who underwent testing before and after NA therapy, the annual rate of change in parameters was evaluated to assess dynamic risk prediction.</p> Results <p>Multivariate analysis identified male sex, LS, total bilirubin, gamma-glutamyl transpeptidase, and fibrosis-4 index as independent predictors, with LS being the most significant (cut-off value: 1.54&#xa0;m/s; area under the curve: 0.811). Diagnostic accuracy (69.4%) was increased to 79.1% by applying a platelet count (PLT) cut-off value of 18.1 × 10<sup>4</sup>/μL to the high-LS group, which was the next most-associated with LS. Rate of change in LS (%LS) was the most powerful predictor (p = 0.0015) in the longitudinal analysis. A %LS cut-off of − 13.1% yielded an area under the receiver operating characteristic curve of 0.914. Combining LS-PLT criteria with %LS resulted in a sensitivity of 90.9% and overall accuracy of 92.9% for identifying high-risk patients.</p> Conclusion <p>Integrating cross-sectional LS and PLT measurements with the longitudinal rate of %LS significantly enhances HCC risk stratification. This approach facilitates personalized surveillance and optimized management for patients with chronic hepatitis B.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Superior accuracy of a novel multidimensional index integrating static and dynamic liver stiffness for hepatocellular carcinoma risk prediction in chronic hepatitis B

  • 期刊论文

  • 发表日期：

    2026年7月9日

查看全文
  • Yusuke Sano,
  • Naoto Kawabe,
  • Hiroko Sugiyama,
  • Yutaka Sasaki,
  • Keisuke Osakabe,
  • Eizaburo Ohno,
  • Teiji Kuzuya,
  • Senju Hashimoto,
  • Yoshiki Hirooka,
  • Naohiro Ichino

摘要

Purpose

Hepatitis B virus (HBV) infection is a major risk factor for hepatocellular carcinoma (HCC). Although nucleoside/nucleotide analogue (NA) therapy significantly reduces this risk, a subset of patients still develop HCC. This study aimed to identify high-risk groups by evaluating the comprehensive utility of multidimensional factor combinations.

Methods

We analyzed 565 patients with chronic HBV infection who underwent liver stiffness (LS) measurement between 2009 and 2024. Independent risk factors for HCC were identified using multivariate logistic regression. In 84 patients who underwent testing before and after NA therapy, the annual rate of change in parameters was evaluated to assess dynamic risk prediction.

Results

Multivariate analysis identified male sex, LS, total bilirubin, gamma-glutamyl transpeptidase, and fibrosis-4 index as independent predictors, with LS being the most significant (cut-off value: 1.54 m/s; area under the curve: 0.811). Diagnostic accuracy (69.4%) was increased to 79.1% by applying a platelet count (PLT) cut-off value of 18.1 × 104/μL to the high-LS group, which was the next most-associated with LS. Rate of change in LS (%LS) was the most powerful predictor (p = 0.0015) in the longitudinal analysis. A %LS cut-off of − 13.1% yielded an area under the receiver operating characteristic curve of 0.914. Combining LS-PLT criteria with %LS resulted in a sensitivity of 90.9% and overall accuracy of 92.9% for identifying high-risk patients.

Conclusion

Integrating cross-sectional LS and PLT measurements with the longitudinal rate of %LS significantly enhances HCC risk stratification. This approach facilitates personalized surveillance and optimized management for patients with chronic hepatitis B.