Cardiovascular Risk and Prediction Model of Vascular Endothelial Growth Factor Tyrosine Kinase Inhibitors in Patients with Liver Cancer

Major adverse cardiovascular events (MACE) in patients with liver cancer receiving vascular endothelial growth factor tyrosine kinase inhibitors (VEGF-TKI) remain uncertain. This retrospective cohort study aimed to evaluate the association between VEGF-TKIs and MACE and develop a risk prediction score. Using the Taiwan Cancer Registry linked with the Taiwan National Insurance Claim Database, 11,960 VEGF-TKI users were matched to 11,960 non-users by age and sex. Cause-specific hazard ratios (HRs) were estimated using multivariable Cox models with competing risk of non-cardiovascular death; propensity score (PS)-adjusted or PS-matched models were additionally conducted. During follow-up, the incidence of MACE was 49.5 versus 28.3 per 1,000 person-years in VEGF-TKI users and non-users. VEGF-TKI use was associated with increased MACE risk in multivariable analysis (adjusted HR 1.31; 95% CI 1.14–1.50); however, PS-adjusted and PS-matched models show non-significant associations. Because proportional hazards assumptions were violated, a 60-day landmark analysis was conducted, in which the increased risk was again significant across analytic approaches, including multivariable adjusted (HR 1.92; 95% CI 1.63–2.27), PS-adjusted (HR 1.35; 95% CI 1.00-1.82) and PS-matched models (HR 1.30; 95% CI 1.03–1.63). A point-based score was developed using a multivariable sub-distribution hazard model and assessed by a time-dependent c-index. The score demonstrated good discrimination with c-indices of 79.9% (3 months) and 74.2% (6 months). VEGF-TKIs were associated with increased risk of MACE among liver cancer patients. Landmark analyses indicated a persistently elevated risk among patients surviving beyond early treatment. A risk score based on prior cardiovascular history may help identify high-risk patients.