Cardiovascular biomarkers for the prediction of adverse cardiovascular events and mortality in patients with cancer

Introduction Patients with cancer are prone to develop thromboembolic and atherothrombotic events. Advances in anti-cancer therapies lead to prolonged cancer-specific survival, putting patients at increasing risk of major adverse cardiovascular events (MACE). Hence, tools to personalize risk stratification to predict and eventually prevent cardiovascular complications in patients with cancer are an unmet medical need. Here, we aimed to assess the predictive utility of cardiovascular biomarkers in a representative population of oncologic patients.

Method In a single-center prospective cohort study (Vienna Cancer & Thrombosis Study - CATS), cardiovascular biomarkers (lipoprotein(a), NT-proBNP, P-selectin, E-selectin, L-selectin, ICAM-1, VCAM-1, sLOX-1) were explored for their predictive utility towards MACE (i.e., myocardial infarction, ischaemic stroke, cardiovascular death), cardiovascular-, and all-cause mortality. Patients were followed for MACE for 2 years, whereas a prolonged follow-up period (5 years) for mortality outcomes was feasible due to data availability from the national death registry. MACE and cardiovascular mortality were analysed by competing-risk regression, accounting for non-cardiovascular death as competing outcome event, adjusting for age, sex and smoking status. All-cause mortality was analysed in Cox regression analysis, adjusting for age, sex, cancer stage and type.

Results In total, 2,192 adult patients with newly diagnosed or recurrent cancer were included (median age: 62 years; 53% male). Over a median follow-up of 23 months, 50 MACEs occurred (cumulative 1- and 2-year incidence: 1.7% [95% confidence interval, CI: 1.2-2.3] & 2.4% [1.8-3.1]). MACE-risk was independently associated with baseline levels of NT-proBNP (sub-distribution hazard ratio, SHR, per double: 1.28 [95%CI: 1.06-1.54]), ICAM-1 (SHR: 1.53 [1.06-2.20], and L-selectin (SHR: 0.63 [0.44-0.90]), but not for lipoprotein(a), P-selectin, VCAM-1, or sLOX-1. An additive predictive effect of biomarkers was observed in the derivation of a point-based prediction score ([Fig. 1]). Long-term cardiovascular mortality was independently associated with levels of NT-proBNP (SHR: 1.42 [95%CI: 1.17-1.72]) and ICAM-1 (1.62 [1.05-2.48]). Accordingly, risk of all-cause mortality independently increased with higher NT-proBNP (HR: 1.16 [95%CI: 1.10-1.22]) and ICAM-1 (1.15 [1.06-1.25]). No significant association with cardiovascular mortality or all-cause mortality was observed for the remainder of biomarkers ([Fig. 2]).

Fig. 1  Additive effect of biomarkers for the prediction of MACE in patients with cancer.; Patients were assigned +1 point each for levels of NT-proBNP ≥75th percentile, ICAM-1 ≥75th percentile, and L-selectin ≤25th percentile. Cumulative incidences were obtained in competing risk analysis, accounting for non-cardiovascular mortality as the competing outcome event.

Fig. 2  Overall survival estimates (2A) and cumulative risk of cardiovascular death (2B) according to dichotomized biomarker levels. Biomarkers were dichotomized at the 75^th percentile of distribution, comparing patients below this cut-off (“low”), to those at or above the cut-off (“high”). Overall survival estimates were obtained in Kaplan-Meier analysis, whereas cumulative risk of cardiovascular death was analysed in a competing-risk framework, accounting for non-cardiovascular-cause mortality as competing outcome event.

Conclusion Cancer patients are at substantial risk of cardiovascular events, with a predictive utility of NT-proBNP, L-selectin, and ICAM-1 for short-term MACE-risk. Further, NT-proBNP and ICAM-1 independently predict long-term risk of cardiovascular- and all-cause mortality. Our findings suggest a possible benefit of incorporating cardiovascular biomarkers in personalized risk prediction models to develop individualized cardiovascular prevention strategies in oncologic patients.

Conflict of Interest

No conflicts of interest to disclose.

Publication History

Article published online:
20 February 2023