Quantitative and qualitative magnetic resonance imaging-based analysis to predict postoperative disease progression in hepatocellular carcinoma.
AI interpretation is pending for this paper.
Open original publication →What the AI sees
Not AI summarized yet.
Research significance
Pending deeper interpretation.
Source abstract
BACKGROUND: Hepatocellular carcinoma (HCC) is one of the most common abdominal malignancies and the third leading cause of cancer-related death globally. Predicting the postoperative prognosis of HCC remains an important area of clinical research. Many qualitative magnetic resonance imaging (MRI) indicators have been used for noninvasive and global assessment of HCC for diagnostic and prognostic purposes. However, the prognostic value of quantitative MRI parameters has not been extensively validated in multicenter cohorts. OBJECTIVE: This study aimed to verify whether MRI-based imaging features could improve prognostic prediction in patients with HCC. We developed prognostic models to predict the progression-free survival (PFS) of patients with HCC. METHODS: In this multicenter cohort study, 203 patients diagnosed with early- to intermediate-stage HCC were divided into a training dataset and an external validation dataset. In addition to clinical indicators, the candidates included qualitative and quantitative MRI data. Three noninvasive models were developed: the Clinical Model, the Radiological Model, and the Combined Model. The patients were stratified according to the risk of disease progression. Model performance was assessed through discrimination and calibration analyses, decision curve analysis was performed, and a clinical application was designed. In addition, the subgroups at different stages of the disease were analyzed, and PFS was compared between the high- and low-risk groups. RESULTS: Prognostic nutritional index, tumor-feeding vessels, mosaic architecture and the signal enhancement ratio (SER) parameter were identified as independent predictors of disease progression in patients with HCC. The Combined Model demonstrated superior performance compared with the Clinical and Radiological Models. The areas under the curve (AUCs) in the training dataset were 0.822, 0.748, and 0.795, whereas those in the validation dataset were 0.819, 0.758, and 0.677, respectively. The Combined Model also demonstrated favorable calibration and greater net clinical benefit on decision curve analysis. Its predictive ability remained consistent regardless of the Child-Pugh classification or Barcelona Clinic Liver Cancer stage. CONCLUSION: This study highlights the important role of quantitative and qualitative MRI assessments in evaluating the prognosis of patients with HCC and emphasizes the predictive value of SER parameter. However, these findings should nevertheless be interpreted with caution given the relatively limited sample size, despite the use of bootstrap internal validation and external validation.