WED-248 Circulating miRNAs as predictive biomarkers of response to immunotherapy in hepatocellular carcinoma (HCC)

Background and Aims: Immunotherapy has changed the treatment of advanced HCC, but early identification of patients who do not respond remains a challenge due to the lack of predictive biomarkers. Circulating microRNAs (miRNAs) have emerged as promising non-invasive biomarkers that reflect tumor biology, disease progression, and treatment response. This study aimed to characterize the serum miRNome in patients with advanced HCC receiving immunotherapy-based systemic treatment [atezolizumab plus bevacizumab (atezo-be) or tremelimumab plus durvalumab (durva-treme)] to identify a molecular signature associated with treatment response. Method: serum samples were collected from 10 patients (discovery cohort) with advanced HCC treated with atezo-bev (n=8) or durva-treme (n=2) at 3 months (T3) after treatment start. Patients were stratified into responders (R, n=6) and non-responders (NR, n=4) based on radiological response, defined as partial or complete response according to RECIST 1.1 criteria. High-throughput miRNome profiling, differential expression, and hierarchical clustering analyses were performed to identify discriminatory signatures. For validation, serum samples were collected at multiple time points [baseline, before the second therapy cycle (T1), and T3] from an independent cohort of 20 HCC patients treated with the same regimens and analyzed by quantitative polymerase chain reaction. Results: Unsupervised hierarchical clustering of serum miRNA expression profiles, and differential expression analysis revealed clear segregation between R and NR, identifying a subset of miRNAs differentially expressed between the two groups at T3. Notably, miR-942-5p was the most statistically significant variable, exhibiting marked upregulation in responders. Other key miRNAs contributing to the discriminatory signature included the oncogenic miR-21-5p, miR-483-5p, and the hepato-specific miR-122-5p. Selected miRNAs were subsequently validated in the independent validation cohort, confirming their differential expression patterns. Conclusion: Our preliminary analysis suggests a serum miRNA signature at 3 months of therapy that may differentiate responders from non-responders. The observed differential expression pattern highlights miRNAs as candidate markers for monitoring treatment response. These findings provide a basis for further investigating the role of liquid biopsy in guiding personalized therapeutic strategies.