Model predicts 10-year risk of heart failure after early-stage breast cancer

A risk prediction model was able to prospectively identify which women treated for early-stage breast cancer were at risk for heart failure or cardiomyopathy over a 10-year period, a study found.

To inform cardiac risk management in this population, researchers conducted a longitudinal cohort study of patients treated by Kaiser Permanente Southern California. All were ages 18 to 79 years with newly diagnosed invasive local (70.6%) or regional (29.4%) breast cancer in 2008 to 2020, with a median follow-up of 5.2 years. Participants (n=26,044, median age 61 years) were randomly split into derivation (60%) and validation (40%) cohorts.

The primary outcome was development of heart failure or cardiomyopathy, and participants were categorized into low-, moderate-, and high-risk groups based on the tertiles of risk. Predictors included age at cancer diagnosis; race and ethnicity; area-level socioeconomic status; local and systemic breast cancer treatments; cancer stage; obesity; and history of hypertension, diabetes, hyperlipidemia, smoking, and other cardiovascular conditions. Findings were published by JAMA Oncology on Oct. 23.

Overall, the model had good calibration and high accuracy in predicting risk in all three subgroups. Risk in the validation cohort matched estimates from the derivation cohort for identifying women at low risk (1.7%; 95% CI, 1.1% to 2.4%) versus high risk (19.4%; 95% CI. 17.3% to 21.5%) of the primary outcome at 10 years. Discrimination was also good, with a time-dependent area under the curve of 0.79 at 10 years in the validation cohort.

Women ages 65 to 79 years had a 4.9 times greater risk for heart failure or cardiomyopathy compared to women younger than age 40 years. In addition, anthracycline therapy was most strongly associated with risk of cardiac problems (adjusted hazard ratio [HR], 1.95; 95% CI, 1.62 to 2.34), followed by ERBB2-targeted therapies (adjusted HR, 1.66; 95% CI, 1.35 to 2.03). After adjustment for all other factors, there was a statistically significant association between preexisting hypertension, diabetes, smoking history, and obesity and 10-year heart failure/cardiomyopathy risk.

The model “may improve the ability to risk stratify women and better identify who may benefit from cardiac monitoring during and after cancer treatment,” the authors concluded. Limitations include that the study did not include biomarkers or cardiac imaging results, they cautioned.

An accompanying editorial noted that proactively managing the cardiovascular risk factors identified by the study is an essential component of preventive care.

“What all breast cancer survivors need is access to primary care that focuses on prevention or management of established cardiac risk factors, as well as regular clinical assessment of their functioning,” the editorialists wrote. They also called for additional validation of the model in a variety of health care and geographic settings to identify patients most vulnerable to heart failure.