Baseline Interleukin-6 and-8 predict response and survival in patients with advanced hepatocellular carcinoma treated with sorafenib monotherapy: an exploratory post hoc analysis of the SORAMIC trial

Purpose To explore the potential correlation between baseline interleukin (IL) values and overall survival or objective response in patients with hepatocellular carcinoma (HCC) receiving sorafenib. Methods A subset of patients with HCC undergoing sorafenib monotherapy within a prospective multicenter phase II trial (SORAMIC, sorafenib treatment alone vs. combined with Y90 radioembolization) underwent baseline IL-6 and IL-8 assessment before treatment initiation. In this exploratory post hoc analysis, the best cut-off points for baseline IL-6 and IL-8 values predicting overall survival (OS) were evaluated, as well as correlation with the objective response. Results Forty-seven patients (43 male) with a median OS of 13.8 months were analyzed. Cut-off values of 8.58 and 57.9 pg/mL most effectively predicted overall survival for IL-6 and IL-8, respectively. Patients with high IL-6 (HR, 4.1 [1.9-8.9], p < 0.001) and IL-8 (HR, 2.4 [1.2-4.7], p = 0.009) had significantly shorter overall survival than patients with low IL values. Multivariate analysis confirmed IL-6 (HR, 2.99 [1.22-7.3], p = 0.017) and IL-8 (HR, 2.19 [1.02-4.7], p = 0.044) as independent predictors of OS. Baseline IL-6 and IL-8 with respective cut-off values predicted objective response rates according to mRECIST in a subset of 42 patients with follow-up imaging available (IL-6, 46.6% vs. 19.2%, p = 0.007; IL-8, 50.0% vs. 17.4%, p = 0.011). Conclusion IL-6 and IL-8 baseline values predicted outcomes of sorafenib-treated patients in this well-characterized prospective cohort of the SORAMIC trial. We suggest that the respective cut-off values might serve for validation in larger cohorts, potentially offering guidance for improved patient selection.

Automated summary

Baseline IL-6 and IL-8 values were found to predict outcomes in sorafenib-treated HCC patients, suggesting potential for guiding patient selection and improving treatment outcomes.