AUTHOR=Zhang Wei , Wu Di , Wang Xinping , Zhang Hua , Yu Ming 

TITLE=Development and evaluation of a nomogram prediction model for invasion and metastasis in primary liver cancer based on serum CD147 and IL-6

JOURNAL=Frontiers in Oncology

VOLUME=Volume 15 - 2025

YEAR=2025

URL=https://www.frontiersin.org/journals/oncology/articles/10.3389/fonc.2025.1524765

DOI=10.3389/fonc.2025.1524765

ISSN=2234-943X

ABSTRACT=ObjectiveThis study aims to develop a prediction model for invasive metastasis of primary liver cancer based on serum extracellular matrix metalloproteinase-inducing factor (CD147) and interleukin-6 (IL-6).MethodsBetween July 2022 and August 2024, 170 surgically treated primary hepatocellular carcinoma patients at our hospital were recruited. They were divided into a training group (n = 120) and a validation group (n = 50) at a 7:3 ratio. Univariate and multivariate logistic regression analyses were applied in the training group to identify factors related to invasive metastasis. A risk factor-based bar chart prediction model was then constructed and internally tested. Its goodness of fit was evaluated, and the model’s diagnostic efficacy was assessed using the ROC curve. Finally, decision curve analysis (DCA) was performed to evaluate the model’s clinical value.ResultsIn the training group, compared with the noninvasive metastasis group, patients in the invasive metastasis group had a significantly lower percentage of intact envelope and tumor size ≥5 cm, and significantly higher serum alpha-fetoprotein (AFP), alkaline phosphatase (ALP), C-reactive protein/albumin ratio (CAR), oncoglobulin (CEA), CD147, and IL-6 levels (all p < 0.05). After logistic multifactorial analysis, intact envelope, tumor > 5 cm, AFP, CAR, CD147, and IL-6 were identified as independent influencing factors for invasive metastasis of primary hepatocellular carcinoma (all p < 0.05). A column chart model was constructed. The C-index of the training and validation groups was 0.884 (95% confidence interval [CI]: 0.738–0.932) and 0.841 (95% CI: 0.741–0.939), respectively. The calibration curves showed good agreement between the predicted probability and the actual probability in both the training and validation groups, without significant deviation. The area under the curve (AUC) of the ROC analysis was 0.852 (95% CI: 0.824–0.979) and 0.839 (95% CI: 0.791–0.912), respectively. DCA indicated that the model had clinical application value within a certain range of threshold probabilities.ConclusionThe prediction model based on serum CD147, IL-6, and other risk factors for the invasion and metastasis of primary hepatocellular carcinoma demonstrates high diagnostic value.