Item

Abstract

Introduction: Hepatocellular carcinoma (HCC) is a type of cancer that typically arises from excessive inflammation in the liver. Since 2020, a new checkpoint inhibition therapy (ICI), namely Atezolizumab in combination with the VEGFR-blocker Bevacizumab (A/B), has been approved for HCC. The intestinal microbiome is important in modulating host immune responses and has been shown to influence ICI effectiveness in different cancer types, e.g. melanoma and lung cancer. Despite the therapeutic improvements in ICI over the past years, therapy response is often heterogenic and sometimes not durable. Hence, the microbiome"s predictive role on ICI efficacy for treatment of HCC requires further scrutiny. Objectives: We aimed to establish a combined predictive biomarker for short-term response to therapy with A/B in hepatocellular carcinoma. Materials & methods: We collected stool, saliva and serum samples from 50 patients with multifocal or advanced hepatocellular carcinoma whose recommended treatment is A/B. This was done before start of treatment as well as weekly (microbiome sample collection) and three-weekly (serum collection) at the time points of antibody application until the fourth antibody application after 9 weeks. Microbiome samples were subjected to whole-genome shotgun sequencing for microbiome profiling and serum samples are being subjected to targeted and untargeted metabolomics. For short-term therapy response, patients were divided into non-responders (progressive disease) whose therapy was discontinued and responders (response, stable disease) who continued A/B therapy at their first staging after 3 months. Results: In a first pilot experiment with baseline samples of 26 patients, we observed that patients who responded to therapy had higher gut microbial alpha-diversity, increased abundances of Bifidobacteriaceae and Oscillospiraceae, and reduced Enterobacteriaceae, compared to non-responders. Furthermore, a random forest model trained on a training set of pre-treatment samples was capable of correctly predicting therapy response in an independent test set of pre- treatment samples. Conclusion: Taken together, our results indicate that the baseline gut microbiome is associated with subsequent responses to immunotherapy in hepatocellular carcinoma patients, suggesting it could be useful as predictive biomarker to inform treatment choice.