Abstract
INTRODUCTION: Increased metabolic activity is frequently observed in hepatocellular carcinoma (HCC). However, the impact of this increased metabolic activity on the efficacy of current treatments, such as the combination immunotherapy using atezolizumab and bevacizumab for HCC, remains unknown. METHODS: Gene expression data from mouse livers representing hepatic metabolic activity and HCC patient tumor tissues were used to identify a transcriptomic signature for high metabolic activity in HCC tumors. The hepatic metabolic signature (HMS) was used to categorize HCC patients treated with atezolizumab plus bevacizumab or sorafenib from the IMbrave150 clinical trial into high-or low-metabolic groups, using multiple statistical approaches to evaluate the clinical relevance of the signature. RESULTS: The study uncovered a robust association between high HMS and poor overall survival in HCC patients across multiple independent cohorts. Notably, high HMS patients in IMbrave150 did not show significant benefit of the atezolizumab-bevacizumab treatment in terms of overall survival or progression-free survival compared to sorafenib monotherapy. Conversely, low HMS patients demonstrated superior overall and progression-free survival outcomes with the combination regimen relative to sorafenib alone. Furthermore, an association between high HMS and features of hepatic stem cells and increased genomic instability was identified. CONCLUSION: This study provides compelling evidence that the HMS could be a predictive biomarker to identify potential HCC patients with therapeutic benefits from combination immunotherapy with atezolizumab and bevacizumab. Leveraging such predictive metabolic biomarkers may pave the way for tailored, precision medicine strategies that maximize therapeutic responses and improve outcomes for HCC patients.