Radiomic Features at Contrast-enhanced CT Predict Recurrence in Early Stage Hepatocellular Carcinoma: A Multi-Institutional Study

Abstract

H epatocellular carcinoma (HCC) is the sixth leading type of cancer and the second most fatal tumor worldwide (1). For patients with early stage HCC as defined by the Milan criteria (solitary nodule 5 cm or as many as three nodules 3 cm, without macrovascular invasion and extrahepatic spread), both liver resection and liver transplant are the mainstay curative options (1,2). Although liver transplant offers definite advantages of extirpating both the tumor and the diseased liver, demand for organs far exceeds supply. Therefore, liver resection is accepted as the first-line treatment option for patients with early stage HCC and preserved liver function, whereas liver transplant is the recommended treatment for patients with decompensated cirrhosis (3). Unfortunately, HCC recurrence, including true recurrence by means of tumor dissemination and development of de novo tumors in the cirrhotic liver, occurs in 50%-60% of these patients at 5 years (4,5).Currently, HCC staging systems (eg, Barcelona Clinic Liver Cancer, Hong Kong Liver Cancer, Cancer of the Liver Italian Program, and TNM systems) occupy the central role in prognosis and therefore treatment allocation (1). Accurate risk prediction allows optimal surveillance, prevention, and management strategies for tumor recurrence; however, these systems are inadequate for predicting recurrence, and none of them provide quantifiable risk measures. Recently, a few statistical models, such as the Korean model ( 6) and pre-and postoperative Early Recurrence After Surgery for Liver Tumor (ERASL) models ( 7), have been established specifically to predict HCC

Radiomic Feature-Based Predictive Model for Microvascular Invasion in Patients With Hepatocellular Carcinoma

“…In current clinical practice, for the reason that there are no known single and highly reliable predictive factors for the incidence of MVI, combining multiple MVI status-related factors with radiomics signatures to build multivariate radiomic models is a feasible option. Previous published studies have utilized multiple imaging modal images (e.g., CT, MR, ultrasound) with radiomics algorithms to predict the incidence of MVI preoperatively and then achieve the ultimate goal of assisting in choosing an optimal surgical treatment or predicting the early recurrence of HCC to provide a reasonable treatment scheme for those patients ( 27 – 29 ). With the rapid development and improvement of radiomics technology, we are able to identify high risky HCC patients with MVI with higher accuracy and better efficacy before surgery which have great significance in terms such as select appropriate surgical therapy strategy, give assistance to confirm rational resection scope to achieve radical hepatoectomy for HCC patients rather than barely refer to tumor location, size or two-dimensional spatial relationship between intrahepatic vasculatures as conventional method did consider of the high postoperative recurrence ratio, for patients whose tumor size lower than three millimeters also susceptible to MVI positive anatomical hepatoectomy are recommended as first choice than radiofrequency ablation, for patients susceptible to MVI positive when select to perform non-anatomical hepatoectomy, the distance from resection margin tumor should beyond one millimeter to decrease the risk of postoperative recurrence ( 20 ).…”

Utility of Radiomics for Predicting Patient Survival in Hepatocellular Carcinoma With Portal Vein Tumor Thrombosis Treated With Stereotactic Body Radiotherapy

“…The area under the curve (AUC) for the clinical characteristics was 0.761 (CI: 0.641-0.881), and the AUC was 0.859 (CI: 0.770-0.948) when the radiomic features were combined with the clinical characteristics ( Figure 5). We also compared our findings with those of recent studies (Table 3) (24)(25)(26)(27).…”

Longitudinal radiomics algorithm of posttreatment computed tomography images for early detecting recurrence of hepatocellular carcinoma after resection or ablation

“…However, there were few studies based on radiomics to explore a feasible approach for surveillance of HCC patients after resection or ablation. Previous studies tried to use CT radiomics to predict the recurrence of HCC with preoperative images, for example, they combined the radiomics algorithm with clinical variables or used a single texture analysis parameter to predict early recurrence with an AUC or C-index less than 0.85 [ [18] , [19] , [20] , [21] , [22] ]. However, these studies shared the common flaws that the localization of lesions and specific time point of tumor recurrence remained unpredicted, thus making the precise intervention of recurrence impossible.…”