Severe Thrombocytopenia Is Associated With Bleeding in Hepatocellular Carcinoma Patients Receiving Antiangiogenic Therapy

ABSTRACT Aims Thrombocytopenia and antiangiogenic agents increase bleeding risk in hepatocellular carcinoma (HCC). We aimed to identify a baseline platelet threshold for bleeding risk stratification. Methods This single‐center retrospective cohort included 489 HCC patients treated with antiangiogenic agents from January 2018 to June 2022 without prophylactic platelet‐increasing interventions. Bleeding events were defined using International Society on Thrombosis and Hemostasis criteria. Fine and Gray competing‐risk models were used to explore platelet‐count cutoffs associated with bleeding. Associations were assessed with logistic regression and Chi‐square/Fisher's exact tests, adjusting for key confounders including high‐risk esophageal and gastric varices (EGV). Severe thrombocytopenia was defined as baseline platelets < 50 × 10 9 /L. Results The bleeding rate was 11.0% (54/489). Baseline platelet counts were lower in patients with bleeding than in those without (median 116 × 10 9 /L vs. 140 × 10 9 /L; p = 0.015). Bleeding was more frequent with severe thrombocytopenia (< 50 × 10 9 /L) than with platelet counts ≥ 50 × 10 9 /L (22.9% [8/35] vs. 10.1% [46/454]; p = 0.021). Although platelet count < 50 × 10 9 /L was associated with bleeding in univariable analyses, it was not independently predictive after multivariable adjustment; high‐risk EGV remained the dominant independent risk factor. Conclusion Baseline platelets < 50 × 10 9 /L is a useful risk‐stratification marker (not an independent predictive biomarker) for bleeding during antiangiogenic therapy in HCC, and this threshold should be interpreted alongside variceal risk assessment, particularly in patients with high‐risk EGV.