Cardiovascular Diseases Following Breast Cancer: Towards a Case-by-Case Assessment Through a Prediction Risk Score Model in 943 Patients.

OBJECTIVES

To identify patients at high risk of developing cardiovascular disease through the identification of risk factors among a large population of breast cancer women and to assess the performance of Abdel-Qadir risk prediction model score.

MATERIALS AND METHODS

The medical records and baseline characteristics of all patients/tumors diagnosed with breast cancer from 2010 to 2011 in a French comprehensive cancer center were collected. Cardiovascular events were defined as arterial and cardiac events, atrial fibrillation and venous thromboembolism occurring during the 5-year follow-up. Abdel-Qadir multivariable prediction model for major adverse cardiovascular events were used with the concordance index (c-index) score to assess calibration by comparing predicted risks to observed probabilities.

RESULTS

Among the 943 breast cancer patients included, 83 patients (8.8%) presented with at least one cardiovascular event, leading to a cumulative incidence of 0.07 at 5 years (95% confidence interval [CI], 0.055-0.088). The cumulative incidence of atrial fibrillation at 5 years was 0.01 (95% CI, 0.005-0.018). Factors associated with the occurrence of cardiovascular events were pre-existing cardiovascular diseases including high blood pressure (hazard ratio [HR]=1.78, 95% CI=1.07-2.97, P=0.028), acute coronary syndrome (HR=5.28, 95% CI: 2.16-12.88, P<0.05) and grade 3 Scarff-Blool-Richardson (HR=1.95, 95% CI: 1.21-3.15, P=0.006). With a c-index inferior to 0.7, the Abdel-Qadir score was not fully validated in our population.

CONCLUSION

These findings call for the assessment of the performance of risk prediction models such as Abdel-Qadir score coupled with other factors such as Scarff Bloom and Richardson grading in order to identify patients at high risk of experiencing cardiotoxicity.