Predicting the Occurrence of Cachexia in Patients with BCLC Stage B/C Hepatocellular Carcinoma Receiving Systemic Therapy: A Nomogram Based on Common Clinical Parameters.

BACKGROUND AND

OBJECTIVE

During the disease course of patients with BCLC B/C hepatocellular carcinoma (HCC) receiving systemic therapy, approximately half of the patients will develop cachexia. Therefore, early identification of which patients are likely to develop cachexia is of crucial significance.This study aims to construct and validate a nomogram for predicting the risk of cachexia in this population based on common clinical parameters.

PATIENTS AND

METHODS

This retrospective single - center study involved 906 patients managed at the Fifth Medical Center of Chinese PLA General Hospital from January 2020 to December 2023. Baseline clinical imaging data, biochemical indicators, and relevant clinical data of patients before systemic treatment were collected. All patients were followed up to record treatment regimens and document weight changes for cachexia diagnosis. The data were stratified into a training cohort and a validation cohort. In this study, LASSO regression alongside univariate and multivariate Logistic regression analyses were utilized to ascertain independent risk factors linked to cachexia occurrence, and further to construct and validate a diagnostic nomogram.

RESULTS

This nomogram incorporates predictors such as patient age, maximum size of intrahepatic lesions, extrahepatic metastasis, neutrophil-to-lymphocyte ratio (NLR), and total bile acids, demonstrating good predictive performance. In the training and validation cohorts, its Harrell's concordance index (C-index) reached 0.865 (95% CI: 0.836-0.895) and 0.820 (95% CI: 0.768-0.871), respectively. Calibration curves demonstrated strong consistency between the nomogram's predicted outcomes and the actual measured values, and decision curve analysis (DCA) further substantiated its clinical applicability.

CONCLUSION

This nomogram shows good predictive performance and can effectively identify high-risk individuals, but it is limited by its single-center retrospective design and requires further verification and optimization through multicenter prospective studies.