Hepatocellular carcinoma macroscopic gross appearance on imaging: predictor of outcome after transarterial chemoembolization in a real-life multicenter French cohort.

ABSTRACT

BACKGROUND: Conventional transarterial chemoembolization (cTACE) with lipiodol is widely performed in patients with hepatocellular carcinoma (HCC) unsuitable for curative treatment. Additional tumor parameters such as HCC macroscopic appearance based on imaging might be helpful for transarterial chemoembolization prognostication and management. PATIENTS AND METHODS: A total of 405 patients with HCC who underwent cTACE between 2008 and 2016 from a real-life multicenter French cohort were retrospectively reviewed. Tumors were classified into two macroscopic types according to HCC gross appearance on imaging nodular versus non-nodular. The study population was stratified into two groups derivation and validation cohorts. Independent prognostic factors of survival based on multivariate cox regression models were determined and then assessed in the validation set. Thereafter, time to progression (TTP) and radiological response rate were investigated for each prognostic factors of survival. RESULTS: Median overall survival (OS) was 35 months for Barcelona Clinic Liver Cancer (BCLC) stage A, 22 months for BCLC stage B and 12 months for BCLC stage C patients (P < 0.0001). The corresponding TTP for these patients was 12 (7-17) months, 5 (3-6) months and 1.2 (1.2-3) months (P < 0.0001). Multivariate analysis revealed that tumors size and number, non-nodular type, alpha-fetoprotein, aspartate aminotransferase serum levels and impairment of performance status-1 were independent predictors of survival among the study groups. Non-nodular type was the most powerful factor that influences OS, TTP and radiological response rate for the recommended transarterial chemoembolization candidates. TTP was consistent with OS within each stage. CONCLUSION: HCC macroscopic appearance on imaging is a determinant predictor of outcome after cTACE in a real-life multicenter cohort.