Molecular biomarkers of sintilimab plus lenvatinib in hepatitis-B-virus-associated hepatocellular carcinoma

The combination of immune checkpoint inhibitors and antiangiogenic drugs has shown promising efficacy in advanced hepatocellular carcinoma (HCC). However, tumor regression and progression-free survival (PFS) vary considerably among patients receiving this therapy.

To identify predictive biomarkers in HCC patients treated with sintilimab (programmed cell death protein-1 inhibitor) plus lenvatinib (tyrosine kinase inhibitor).

In this single-center study in China, patients with unresectable HCC received sintilimab every 21 days and daily oral lenvatinib. Treatment response was assessed by modified response evaluation criteria in solid tumors. Tumor biopsies underwent RNA sequencing, immune microenvironment profiling, and whole-exome sequencing. Differentially expressed genes (DEGs) and immune cell subsets between response groups were identified, followed by survival analyses. All potential predictors of PFS, together with clinical variables, were included in Cox regression to identify independent prognostic factors.

Between August 2019 and November 2021, 33 patients with hepatitis-B-virus-related HCC were enrolled; by January 2024, 13 had undergone potentially curative surgery or ablation. RNA sequencing identified 94 DEGs between responders (n = 22) and non-responders (n = 11) using Fisher's exact test or Wilcoxon rank-sum test (all P < 0.05). High long intergenic non-protein coding RNA 01554 (LINC01554) and whirlin expression were associated with longer PFS in Kaplan-Meier analysis (P < 0.05). DEG-immune cell analysis showed positive correlations with pro-B and plasma cells in responders, and negative correlations with CD4+ central memory T (Tcm), T helper 1, and natural killer T cells in non-responders; none significantly predicted PFS, although CD4+ Tcm cells approached significance (P < 0.10). Whole-exome sequencing revealed Fanconi anemia complementation group D2 mutations enriched in non-responders (P < 0.05), while cut-like homeobox 1 mutations predicted poorer PFS (P = 0.011). Cox regression identified solitary tumor [P = 0.02, hazard ratio (HR) = 0.31], high LINC01554 (P = 0.01, HR = 0.16), and elevated CD4+ Tcm cells (P = 0.05, HR = 0.29) as independent predictors of prolonged PFS.

Sintilimab plus lenvatinib showed heterogeneous efficacy in HCC. High LINC01554 expression, elevated CD4+ Tcm cells, and solitary tumors may serve as predictive biomarkers for prolonged disease control.