Systemic immune profiling in hepatocellular carcinoma: navigating etiological heterogeneity and selection bias

This correspondence provides a methodological commentary on the recent study by Nishio et al, which used single-cell RNA sequencing (CITE-seq) to identify circulating immune biomarkers for combination immunotherapies in advanced hepatocellular carcinoma (HCC). While the original study offers high-resolution insights into systemic immunity, we identify three critical areas that warrant further consideration to ensure the clinical validity of the proposed biomarkers. First, we discuss how etiological heterogeneity (viral vs non-viral HCC) may confound regimen-specific immune signatures. Second, we highlight the potential for "selection bias" and statistical inflation of effect sizes resulting from the balanced 1:1 responder to non-responder study design, as well as the inherent mathematical artifacts in compositional single-cell data analysis. Third, we question the biological applicability of modeling stable ligand-receptor interactions (CellChat) within the high-flow environment of peripheral circulation. We conclude that future validation in unselected, prospective cohorts is essential to confirm whether these circulating immune subsets can serve as robust clinical predictors in real-world settings.