
ET Adherence Raises Cardiometabolic Risk, Improves Survival in HR+ Breast Cancer
Five-year AI adherence was tied to more hypertension, dyslipidemia, and diabetes in HR+ breast cancer, yet adherence also lowered all-cause mortality.
Long-term adherence to adjuvant endocrine therapy was associated with a higher 10-year risk of hypertension, dyslipidemia, and diabetes among postmenopausal women with early-stage hormone receptor–positive breast cancer, but it was also associated with substantially lower breast cancer–specific and all-cause mortality, according to findings from a prospective cohort study published in JNCI Cancer Spectrum.1
The increase in cardiometabolic risk was driven largely by aromatase inhibitors (AIs), which were associated with higher incidence of all 3 risk factors, whereas tamoxifen adherence was associated only with an increased risk of hypertension.
Compared with women who were never adherent, 5 years of adherence to any endocrine therapy was associated with an absolute 10-year cumulative incidence difference of 8.81% (95% CI, 4.25%-13.40%) for hypertension, 5.16% (95% CI, 0.31%-10.00%) for dyslipidemia, and 2.58% (95% CI, 0.02%-5.14%) for diabetes. The estimated 10-year cumulative incidence of hypertension was 28.6% (95% CI, 26.4%-30.9%) with 5-year adherence vs 19.8% (95% CI, 15.8%-23.8%) with nonadherence. Corresponding rates were 30.9% (95% CI, 28.6%-33.1%) vs 25.7% (95% CI, 21.4%-30.0%) for dyslipidemia and 12.1% (95% CI, 10.9%-13.4%) vs 9.54% (95% CI, 7.28%-11.80%) for diabetes.
At the same time, 5-year adherence was associated with a 10.7% (95% CI, 7.74%-13.70%) lower absolute 10-year cumulative incidence of all-cause death, a 5.37% (95% CI, 3.36%-7.38%) lower incidence of breast cancer death, and a 2.89% (95% CI, 1.24%-4.53%) lower incidence of cardiovascular disease (CVD) death. Among nonadherent patients, the 10-year cumulative incidence of all-cause death was 28.3% (95% CI, 25.6%-31.0%) vs 17.6% (95% CI, 16.3%-18.9%) among those adherent for 5 years.
“AI adherence was associated with increased cardiometabolic risk, whereas [tamoxifen] was associated only with hypertension risk,” the investigators, led by Jorge R. Ledesma, PhD, MPH, a postdoctoral delivery science research fellow at Kaiser Permanente Northern California, wrote. “Despite these risks, [adjuvant endocrine therapy] was associated with improved survival.”
When the analysis was stratified by therapy type, 5-year AI adherence was associated with a higher absolute 10-year risk of hypertension (9.37%; 95% CI, 4.54%-14.20%), dyslipidemia (6.84%; 95% CI, 1.76%-11.90%), and diabetes (3.29%; 95% CI, 0.58%-6.00%) vs nonadherence. Five-year tamoxifen adherence was associated with a similar increase in hypertension risk (9.38%; 95% CI, 2.71%-16.10%), but not with dyslipidemia (–3.02%; 95% CI, –8.84% to 2.79%) or diabetes (0.36%; 95% CI, –3.32% to 4.04%).
In a direct comparison of adherent patients, AI use was associated with a 9.86% (95% CI, 5.09%-14.60%) higher absolute 10-year incidence of dyslipidemia than tamoxifen, with 10-year rates of 32.5% (95% CI, 29.8%-35.3%) and 22.7% (95% CI, 18.8%-26.6%), respectively. The difference in diabetes incidence between AI and tamoxifen users, 2.93% (95% CI, –0.36% to 6.21%), was not statistically significant, and hypertension incidence was identical in either group at 29.2%. There were no statistically significant differences in any mortality outcome between 5-year AI and tamoxifen adherence.
Obesity appeared to modify some of these associations. Among patients with a body mass index of 30 or higher, 5-year tamoxifen adherence was associated with an increased 10-year risk of diabetes (cumulative incidence difference, 11.1%; 95% CI, 3.32%-18.90%), whereas among patients without obesity, tamoxifen adherence was associated with a decreased risk (–4.42%; 95% CI, –6.71% to –2.12%). The investigators hypothesized that tamoxifen’s known impairment of insulin sensitivity may be insufficient to cause diabetes in healthy individuals but could tip those with greater baseline insulin resistance into the disease.
The study included 8365 postmenopausal women (defined as 50 years and older at diagnosis) with stage I to III hormone receptor–positive breast cancer from the Pathways Heart Study, a prospective cohort within Kaiser Permanente Northern California, who initiated endocrine therapy between 2005 and 2013. Adherence was assessed annually using pharmacy dispensing data and defined as a proportion of days covered of at least 80%. Investigators estimated 10-year cumulative incidence under 6 adherence duration strategies (0-5 years) using sequentially doubly robust estimation, adjusting for demographic, clinical, lifestyle, and tumor characteristics, and treated all-cause death as a competing risk.
The median age at diagnosis was 64 years, 90.9% of patients had stage I or II disease, and 32.5% received chemotherapy. Overall, 42.7% of patients were continuously adherent for at least 5 years, 12.3% for 3 to 4 years, and 12.9% for 1 to 2 years; 12.8% were nonadherent, and 19.4% had mixed adherence patterns. Most patients (67%) received only an AI, 10% received only tamoxifen, and 10% switched between the 2. Findings were largely consistent in sensitivity analyses using alternative adherence thresholds of 60% and 70% and a stricter postmenopausal age cutoff of 55 years.
The investigators noted that cumulative incidence curves showed the largest differences in risk emerging after 5 years of therapy, which may explain why earlier studies with shorter follow-up did not detect some of these associations. They suggested that the curves highlight windows for targeted cardiometabolic monitoring and early intervention, including adherence to American Cancer Society recommendations for healthy eating and physical activity, without compromising the survival benefit of therapy.
Notably, adherence was associated with higher incidence of cardiometabolic risk factors but lower CVD mortality. The authors proposed that adherent patients who developed these risk factors may have had greater engagement with the health system, which could have attenuated differences in subsequent CVD death, and called for mediation analyses to clarify the mechanisms.
Limitations included possible residual confounding from factors not captured in the electronic health record, such as diet, physical activity, cancer recurrence, and health care access, as well as a lack of data on reasons for discontinuation. Adherence was based on dispensing rather than actual medication-taking behavior, and outcomes were defined by clinical algorithms rather than physiological biomarkers. Additionally, patients were treated between 2005 and 2013, before extended-duration endocrine therapy and adjuvant CDK4/6 inhibitors became more common for higher-risk disease, and results from an integrated health care system may not generalize to other populations.
Reference
Ledesma JR, Nugent J, Cheng R, et al. Endocrine therapy adherence and 10-year cardiometabolic risk in hormone receptor-positive breast cancer. JNCI Cancer Spectr. 2026;10(6):pkag095. doi:10.1093/jncics/pkag095
Related to this article








