Investigating Time-Varying Predictor Effects on Cardiovascular Outcomes in Breast Cancer Survivors

PURPOSE: Relevant factors can have shifting prognostic impacts on cardiovascular disease (CVD) occurrences among in patients with breast cancer (BC) over time. CVD incidence and its driving factors vary among different CVDs. We examined the time to the first occurrence of heart attack, atrial fibrillation, embolic stroke, angina pectoris, embolism, peripheral vascular disease, and heart failure (HF). We particularly focused on the influence of molecular subtype, adjusting for age, tumor stage, and radiation therapy. METHODS: The 36,605 women diagnosed with BC from the Norwegian Cancer Registry were included. Cox regression analyses were performed for the first time for six CVDs, with death treated as a competing risk. The association between the time to first CVD diagnosis and the patient's molecular subtype was calculated. Because the Cox proportional hazard assumption was not met, a random survival forest (RSF) analysis was conducted. RESULTS: The association between the time to the first CVD and the patient's molecular subtype differed for each CVD and was non-linear for HF. The time-varying cardiovascular risk in human epidermal growth factor receptor 2 (HER2)-positive versus HER2-negative breast cancer patients reflects differences in treatment, biology, and patient profiles. HER2-positive patients face early cardiotoxicity due to targeted therapies and are closely monitored, while HER2-negative patients, often older with higher baseline CVD risk, may experience delayed detection due to less routine cardiac surveillance. In ranking the factors with respect to their predictive importance for time to first HF, molecular subtype emerged as the second most important factor, followed by age. CONCLUSION: Time to HF depends on the molecular subtype in a time-dependent manner. RSF analyses can identify complex relationships between predictors and survival without the Cox proportional hazard assumption, providing important insights into how patient and treatment factors are associated with time to CVD.