Introduction Atezolizumab plus bevacizumab is a standard first-line therapy for unresectable HCC; however, a proportion of patients develop early progressive disease (PD). Reliable pretreatment predictive biomarkers remain an unmet need. This study investigated the association between baseline serum microRNA-122 (miR-122) levels and early PD. Methods We retrospectively analyzed 64 consecutive patients with unresectable HCC and Child-Pugh class A treated with atezolizumab plus bevacizumab between October 2020 and January 2026 with available baseline serum samples. Early PD was defined as radiological progression within 6 weeks according to modified RECIST. Clinical, laboratory, and tumor characteristics were compared between patients with and without early PD. ROC analysis determined optimal cutoff values, and logistic regression identified factors associated with early PD. Results Early PD at the radiological assessment within 6 weeks of treatment initiation was observed in 16 patients (25.0%). Baseline serum AST, AFP, protein induced by vitamin K absence or antagonist II (PIVKA-II), miR-122 levels and the presence of macrovascular invasion were significantly different between early PD and non-early PD. Multivariate logistic regression analysis identified higher levels of miR-122 levels (≥0.101 fold changes) and AFP (≥332.35 μg/L) as independent predictors of early PD. Even when limited to cases without prior treatment with molecularly targeted therapy, the same results were obtained. Conclusion Baseline serum miR-122 levels were independently associated with early PD in patients with unresectable HCC treated with atezolizumab plus bevacizumab. Serum miR-122 may serve as a predictive biomarker to identify patients at high risk of early treatment failure.
Yamamoto et al. (2026) studied this question.