Abstract
Background
Prior research suggests that breast cancer patients with a high burden of frailty may face an increased risk of cardiotoxicity.
Objectives
This study sought to examine the association between frailty and cardiotoxicity rates in female breast cancer patients receiving adjuvant therapy after surgery.
Methods
We analyzed data from the OneFlorida+ clinical research network, focusing on breast cancer patients treated with adjuvant chemotherapy and targeted therapy from 2012 to 2022. Cardiovascular rates during adjuvant treatments were calculated based on pre-existing frailty, measured using the cumulative deficit frailty index (electronic health record frailty index). We employed multivariable Gray’s method to examine the association between frailty with cardiotoxicity.
Results
The final cohort included 2,050 patients (mean age 50.6 years), with 415 (20.2%) experiencing nonfatal adverse cardiovascular events after adjuvant therapy. The incidence of adverse cardiovascular events was 17.8% in robust, 23.2% in prefrail, and 29.4% in frail patients. In multivariable analysis, prefrail (adjusted subdistribution HR [sHR]: 1.35; 95% CI: 1.06-1.71; P = 0.015) and frail (adjusted sHR: 1.70; 95% CI: 1.11-2.61; P = 0.015) patients had a higher likelihood of experiencing adverse cardiovascular events compared with robust patients. Among non-Hispanic White and Black patients, prefrail (adjusted sHR: 1.48; 95% CI: 1.04-2.11; P = 0.031; and adjusted sHR: 1.59; 95% CI: 1.06-2.37; P = 0.024, respectively) and frail (adjusted sHR: 1.96; 95% CI: 1.10-3.50; P = 0.022; and adjusted sHR: 2.13; 95% CI: 1.11-4.10; P = 0.023, respectively) patients were more likely to experience adverse cardiovascular events compared with robust patients. No significant differences were observed in other racial/ethnic groups.
Conclusions
These findings highlight the need for close monitoring of cardiotoxicity in frail breast cancer patients undergoing adjuvant treatments to improve cardiovascular risk management.
Key Words: adjuvant treatment, breast cancer, cardiotoxicity, electronic health records, frailty, geriatric oncology, outcomes, real-world data, risk factor, treatment, women's oncology
Central Illustration
Breast cancer is the most common cancer and the second-leading cause of cancer-related death among women in the United States.1 In 2024, an estimated 310,720 women in the United States will be diagnosed with breast cancer, and 42,250 will die from the disease.1 Over recent decades, the development and use of systemic adjuvant treatment has significantly reduced breast cancer mortality by 10% to 30%.2, 3, 4, 5 However, these treatments, particularly chemotherapy and targeted therapy, can lead to cardiotoxicity, a severe adverse effect that damages the cardiovascular system, which can cause heart failure, arrhythmias, myocardial infarction, and even death. Cardiotoxicity contributes to approximately 35% of mortality from causes other than breast cancer in older breast cancer patients.6 Previous studies have shown that 5% to 34% of breast cancer patients receiving chemotherapy or targeted therapy experienced cardiotoxicity.7, 8, 9, 10
Approximately 80% of breast cancer patients are diagnosed at 50 years of age or older, a stage at which the risk of chronic illnesses begins to increase.1 Frailty, a common age-related risk factor and a single global public health issue, has received considerable attention because it explains the variability in health among older adults.11 Frailty is characterized by a decline in physical function, increased vulnerability to stressors, and a higher risk of adverse outcomes, such as falls, disability, hospitalization, and death.12 Among breast cancer patients, frailty prevalence ranges from 5% to 71%, with frail breast cancer patients facing a substantially higher risk of all-cause mortality compared with their nonfrail counterparts.13, 14, 15 Understanding the impact of frailty on cardiotoxicity in breast cancer patients can better inform clinicians and patients about prognosis throughout the continuum of cancer care, especially among older adults.16 Previous prospective studies have shown that the cumulative deficit frailty index is associated with increased adverse cardiovascular events, independent of traditional risk factors.17,18 However, to our knowledge, there are no observational studies examining associations between frailty and adverse cardiotoxicity-related cardiovascular events in breast cancer patients, particularly in the context of adjuvant treatments.
In this study, we aimed to assess how the risk of cardiotoxicity varies with the level of frailty in breast cancer patients who received adjuvant chemotherapy and targeted therapy. We used electronic health records (EHRs) from the OneFlorida+ clinical research network,19 a large network within the Patient-Centered Clinical Research Network (PCORnet). This comprehensive data source enabled us to investigate the relationship between frailty and cardiotoxicity in breast cancer patients in a real-world setting.
Methods
Data source and study population
We obtained patient-level EHR data from the OneFlorida+ clinical research network, which includes 16.8 million patients in Florida, 2.1 million in Georgia, and 9,800 in Alabama. The OneFlorida+ EHR data follow the PCORnet common data model, providing detailed information on demographics, diagnoses, procedures, observations (including vital signs), prescriptions, laboratory tests, immunizations, and more.
From the OneFlorida+ EHR data, we identified women diagnosed with breast cancer between January 1, 2012, and April 30, 2022, who underwent surgery within 1 year of their breast cancer diagnosis. Breast cancer cases were identified using International Classification of Diseases–Ninth Revision (ICD-9) and International Classification of Diseases–Tenth Revision (ICD-10) codes (ICD-9: 174.x; ICD-10: C50.x), while subsequent surgeries were identified using Current Procedural Terminology and ICD-9/ICD-10 procedure codes (Supplemental Table 1).
In this cohort, we further identified patients who received adjuvant chemotherapy and targeted therapy after surgery. This included treatments such as anthracyclines, human epidermal growth factor receptor 2 (HER2)–targeted therapy, cyclin-dependent kinase 4/6 (CDK4/6) inhibitors, and 5-fluorouracil (5-FU) or capecitabine. These therapies were identified using Healthcare Common Procedure Coding System, RxNorm concept unique identifier, and National Drug Code codes (Supplemental Table 1).
These adjuvant treatments were specifically selected based on their relevance to cardiotoxicity as determined by clinical experts. The first date of adjuvant treatment after surgery was defined as the index date, while the last date of adjuvant treatment was defined as the date on which adjuvant therapy was administered, followed by a 90-day period without any subsequent adjuvant therapy. We excluded breast cancer patients whose first adjuvant treatment was more than 1 year after surgery and further excluded those who met any of the following criteria: 1) experienced any of the adverse cardiovascular events (see Primary Outcome) before the index date; 2) had a diagnosis of another cancer within 1 year before the index date; 3) lacked encounter records within 1 year before the index date or within 1 year after the last treatment date; or 4) had a duration of chemotherapy or targeted therapy that exceeded 2 years.
This study was approved by the University of Florida Institutional Review Board.
Primary outcome
The primary study outcome was the onset of new nonfatal adverse cardiovascular events during adjuvant treatment. We included 4 groups of adverse cardiovascular events potentially associated with cardiotoxicity: heart failure, cardiac arrhythmia, cardiomyopathy, and atrial fibrillation/flutter (see Supplemental Table 2). These events were selected based on the expertise of cardiologists and oncologists, who reviewed the mechanisms of cardiotoxicity in breast cancer treatment and related adverse cardiovascular outcomes.20
We identified these events in the EHRs of our study population using International Classification of Diseases codes (Supplemental Table 2). Only events that occurred during adjuvant treatment or within 90 days after the last dose of adjuvant treatment were included in our analysis. The time frame for assessing adverse cardiovascular events after adjuvant treatment in breast cancer patients varies across studies, ranging from several months to several years, depending on the study design and objectives.21, 22, 23, 24 We chose a 90-day period to directly attribute the observed events to adjuvant treatment, minimizing the impact of potential confounders that could emerge over a longer follow-up.
Patients were followed from the index date until the first occurrence of an adverse cardiovascular event, death, or censoring. Censoring occurred at 90 days after the last date of adjuvant treatment or at a maximum follow-up of 730 days, whichever came first.
Primary exposure
The primary exposure was frailty, measured by the electronic health record frailty index (eFI),25,26 a deficit-accumulation frailty index validated for use with EHRs.27,28 Prior research has demonstrated that the eFI accurately represents frailty patterns and provides valuable prognostic information for mortality among older adults.27,28 In this analysis, the eFI included the following 43 items (termed deficits) from EHRs:
-
•
Diagnoses (24 items): anemia, rheumatoid arthritis or osteoarthritis, renal disease, dizziness or vertigo, dyspnea, falls, fragility or fracture, hearing impairment, hypotension, dementia, osteoporosis, Parkinson’s disease, peptic ulcer, pulmonary disease, skin ulcer, thyroid disease, urinary incontinence, urinary system disease, blindness and other vision defects, weight loss, depression, mild liver disease, moderate or severe liver disease, and chronic pain.
-
•
Laboratory test results (15 items): estimated glomerular filtration rate, high-density lipoprotein cholesterol, total cholesterol, triglycerides, potassium, sodium, aspartate aminotransferase, mean corpuscular volume, blood urea nitrogen, calcium, albumin, total protein, alkaline phosphatase, hemoglobin, and glucose.
-
•
Vital signs (4 items): systolic blood pressure, diastolic blood pressure, body mass index, and smoking.
Cutoffs for laboratory results and vital signs used for deficit scoring are described in Supplemental Table 3.25
We deliberately excluded diabetes and hypertension from the eFI calculation, as they are strongly correlated with adverse cardiovascular events, which could artificially inflate the association between frailty and cardiotoxicity. Instead, they were included as covariates in the statistical models (see Covariates).
Consistent with standard eFI construction procedures, we required at east 30 nonmissing items to classify patients as having a robust health status.25,29 The eFI was calculated as the unweighted sum of scores for each deficit, divided by the total number of nonmissing deficits.25 The resulting eFI score ranges from 0 to 1, with higher values indicating a greater burden of frailty. For this study, eFI was calculated using EHR data from the year before the index date and categorized into 3 groups: robust (0 ≤ eFI ≤ 0.10), prefrail (0.10 < eFI ≤ 0.21), and frail (0.21 < eFI ≤ 1), based on prior literature.15,25
Covariates
Covariates included age at cancer diagnosis, race/ethnicity, health insurance, adjuvant treatment type, year adjuvant therapy started, rurality of residency, census tract poverty, baseline hypertension and diabetes status, and baseline health care utilization (ie, outpatient and inpatient visits). These factors were chosen based on a priori knowledge of their associations with both exposure and outcome.
Age at breast cancer diagnosis was categorized into 5 groups: <50, 50 to 59, 60 to 69, 70 to 79, or ≥80 years. Race/ethnicity was classified as non-Hispanic White, non-Hispanic Black, non-Hispanic other, Hispanic, or unknown. Insurance coverage was grouped into 3 categories: Medicare, Medicaid or uninsured, or other health insurance, which included commercial insurance, managed care, federal/state/local government insurance, workers’ compensation, and other. We categorized adjuvant treatment regimens into 4 groups based on cardiotoxicity mechanisms: anthracycline, HER2-targeted therapies, CDK4/6 inhibitors, and 5-FU and capecitabine. For each category, we compared the risk of adverse cardiovascular events between users and nonusers. Rurality of residency was determined by linking the patients’ latest zip codes in the EHRs to the rural-urban commuting area (RUCA) codes.30 Patients were categorized as urban (RUCA code 1) or nonurban (RUCA codes 2-10). Census tract–level poverty, defined as the percentage of the population below the poverty line, was determined by linking patients’ zip codes to the Census Bureau’s American Community Survey. Categories included <10%, 10% to 19%, and ≥20%.
Additionally, we obtained baseline hypertension (ICD-9: 401-405; ICD-10: I10, I11-I13, I15) and diabetes (ICD-9: 250; ICD-10: E10-E14) status, as well as baseline health care utilization from the patients’ EHR data within 1 year prior to the index date. Hypertension and diabetes were included because they are known risk factors for adverse cardiovascular events in breast cancer patients.31,32
Health care utilization was measured by the number of outpatient and inpatient visits. Due to the low number of patients with multiple inpatient visits, we categorized the number of inpatient visits as counts into 3 groups (0, 1, >1).
Last, to account for changes in chemotherapy selection and regimens over the study period, we included the year when adjuvant therapy started as a covariate in our analyses.
Statistical analysis
We summarized the distributions of study characteristics in the overall population and by eFI group (ie, robust, prefrail, or frail). For continuous variables, we reported mean ± SD if normally distributed or median (Q1-Q3) if skewed. For categorical variables, we reported count and percentage. Normality of continuous data was assessed using the Kolmogorov-Smirnov test.
To compare distributions of study characteristics across eFI groups, we used analysis of variance or the Kruskal-Wallis test for continuous variables, and the chi-square or Fisher exact test for categorical variables. We calculated cumulative incidence with 95% CI and created cumulative incidence curves to visualize the risk of adverse cardiovascular events across different eFI groups, considering all-cause death as a competing risk. Differences between the cumulative incidence curves across eFI subgroups were compared using Gray’s method.
To examine the association between eFI and the risk of adverse cardiovascular events while accounting for competing risks, we built univariable and multivariable models using the Fine-Gray method. The multivariable models adjusted for the aforementioned covariates. We calculated unadjusted and adjusted subdistribution HRs (sHRs) with 95% CIs. The proportional hazards assumption was tested using the supremum tests, which confirmed that the assumption held for both the univariable and multivariable models.
Additionally, we conducted a sensitivity analysis, focusing on heart failure and cardiomyopathy, the most common complications associated with prevalent adjuvant therapies. Using the Fine-Gray competing risk model with the same covariates, we evaluated whether the associations remained consistent for these specific outcomes.
To explore the potential dose-response relationship between eFI and the risk of adverse cardiovascular events, we used a restricted cubic spline in the multivariable Fine-Gray competing risk model, with eFI = 0.21 as the reference point for the dose-response curve. The use of restricted cubic splines allows for capturing more complex, nonlinear patterns in the data, resulting in a more accurate representation of the underlying patterns. We assessed nonlinearity by comparing the model fit with restricted cubic splines to a model for the eFI, using a likelihood ratio test.
To further explore if the relationship between frailty and adverse cardiovascular events varied by race/ethnicity, we included an interaction term between eFI and race/ethnicity in the Fine-Gray competing risk model and tested its significance. We calculated adjusted sHRs with 95% CIs for frailty within each racial/ethnic group. Data processing and management were performed using Python 3.9.4 (Python Research Foundation), and all statistical analyses were conducted using SAS 9.4 (SAS Institute). Two-sided P values were calculated for all tests, and a P value of <0.05 was considered statistically significant.
Results
Characteristics of study population
We identified 24,495 female breast cancer patients who underwent surgery, of whom 4,630 received the selected adjuvant treatments within 1 year after surgery. After applying the exclusion criteria, a total of 2,025 patients (mean age 50.6 ± 11.4 years) were included in the final analysis (Figure 1).
Figure 1.
Flowchart of Patient Selection
In 2012 to 2022 electronic health records in the OneFlorida+ clinical research network, 4,630 female breast cancer patients received adjuvant treatments within 1 year after surgery. After applying exclusion criteria, a total of 2,025 patients were included in the final analysis. HCPCS = Healthcare Common Procedure Coding System; NDC = National Drug Code; RXCUI = RxNorm concept unique identifier.
Table 1 provides an overview of patient characteristics, both overall and stratified by eFI group. Overall, 61.8% of patients were classified as robust, while 32.0% were prefrail and 6.2% were frail. The majority of patients were non-Hispanic White (39.9%), residents of urban census tracts (83.3%), and lived in census tracts in which <20% of the population was below the poverty line (70.0%).
Table 1.
Patient Characteristics by Status of eFI
| Overall (N = 2,050, 100%) | Robust (n = 1,268, 61.5%) | Prefrail (n = 656, 32.0%) | Frail (n = 126, 6.2%) | P Value | |
|---|---|---|---|---|---|
| Age | <0.001 | ||||
| <50 y | 976 (47.6) | 682 (53.8) | 258 (39.3) | 36 (28.6) | |
| 50-59 y | 641 (31.3) | 359 (28.3) | 232 (35.4) | 50 (39.7) | |
| 60-69 y | 336 (16.4) | 194 (15.3) | 111 (16.9) | 31 (24.6) | |
| ≥70 y | 97 (4.7) | 33 (2.6) | 55 (8.4) | 9 (7.1) | |
| Race/ethnicity | <0.001 | ||||
| NHW | 818 (39.9) | 528 (41.6) | 240 (36.6) | 50 (39.7) | |
| NHB | 477 (23.3) | 249 (19.6) | 193 (29.4) | 35 (27.8) | |
| Hispanic | 601 (29.3) | 398 (31.4) | 170 (25.9) | 33 (26.2) | |
| NHO | 32 (1.6) | 20 (1.6) | 11 (1.7) | 1 (0.8) | |
| Unknown | 122 (6.0) | 73 (5.8) | 42 (6.4) | 7 (5.6) | |
| Census tract ruralitya | 0.031 | ||||
| Nonurban | 271 (13.2) | 149 (11.8) | 100 (15.2) | 22 (17.5) | |
| Urban | 1,707 (83.3) | 1,076 (84.9) | 528 (80.5) | 103 (81.8) | |
| Unknown | 72 (3.5) | 43 (3.4) | 28 (4.3) | 1 (0.8) | |
| Census tract povertyb | 0.110 | ||||
| <10% | 575 (28.0) | 370 (29.2) | 172 (26.2) | 33 (26.2) | |
| 10%-19% | 860 (42.0) | 540 (42.6) | 268 (40.9) | 52 (41.3) | |
| ≥20% | 534 (26.0) | 310 (24.5) | 184 (28.1) | 40 (31.8) | |
| Unknown | 81 (4.0) | 48 (3.8) | 32 (4.9) | 1 (0.8) | |
| Median number of outpatient visitsc | 14.0 | 11.0 | 17.0 | 28.0 | <0.001 |
| Number of inpatient visitsc | <0.001 | ||||
| 0 | 1,553 (75.8) | 1,026 (80.9) | 453 (69.1) | 74 (58.7) | |
| 1 | 347 (16.9) | 173 (13.6) | 147 (22.4) | 27 (21.4) | |
| >1 | 150 (7.3) | 69 (5.4) | 56 (8.5) | 25 (19.8) | |
| Hypertensionc | 710 (34.6) | 242 (19.1) | 373 (56.9) | 95 (75.4) | <0.001 |
| Diabetesc | 211 (10.3) | 43 (3.4) | 120 (18.3) | 48 (38.1) | <0.001 |
| Year adjuvant therapy started | 0.013 | ||||
| 2012-2013 | 421 (20.5) | 259 (20.4) | 135 (20.6) | 27 (21.4) | |
| 2014-2015 | 428 (20.9) | 284 (22.4) | 116 (17.7) | 28 (22.2) | |
| 2016-2017 | 472 (23.0) | 281 (22.2) | 165 (25.2) | 26 (20.6) | |
| 2018-2019 | 383 (18.7) | 215 (17.0) | 135 (20.6) | 33 (26.2) | |
| 2020-2022 | 346 (16.9) | 229 (18.1) | 105 (16) | 12 (9.5) | |
| Insurance coverage | <0.001 | ||||
| Medicare | 222 (10.8) | 92 (7.3) | 104 (15.9) | 26 (20.6) | |
| Medicaid or uninsured | 1,470 (71.7) | 955 (75.3) | 431 (65.7) | 84 (66.7) | |
| All other insuranced | 342 (16.7) | 213 (16.8) | 115 (17.5) | 14 (11.1) | |
| Unknown | 16 (0.8) | 8 (0.6) | 6 (0.9) | 2 (1.6) | |
| Treatment type | |||||
| Anthracycline | 1,197 (58.4) | 744 (58.7) | 380 (57.9) | 73 (57.9) | 0.946 |
| HER2-targeted therapy | 769 (37.5) | 487 (38.4) | 233 (35.5) | 49 (38.9) | 0.439 |
| CDK4/6 inhibitors | 18 (0.9) | 9 (0.7) | 9 (1.4) | 0 (0) | 0.186 |
| 5-FU and capecitabine | 96 (4.7) | 45 (3.6) | 47 (7.2) | 4 (3.2) | 0.001 |
| Adverse cardiovascular eventse | 415 (20.2) | 226 (17.8) | 152 (23.2) | 37 (29.4) | <0.001 |
Values are n (%). Robust = 0 ≤ eFI ≤ 0.10; prefrail = 0.10 < eFI ≤ 0.21; frail = 0.21 < eFI ≤ 1.
CDK4/6 = cyclin-dependent kinase 4/6; eFI = electronic health record frailty index; HER2 = human epidermal growth factor receptor 2; NHB = non-Hispanic Black; NHO = non-Hispanic other; NHW = non-Hispanic White; 5-FU = 5-fluorouracil.
Determined by linking the patient’s latest zip code to the rural-urban commuting area codes.
Determined by linking the patient’s latest zip code to the Census Bureau’s American Community Survey and categorizing percent population below poverty into 3 groups: <10%, 10%-19%, or ≥20%.
Measured within <1 year prior to the first date of adjuvant therapy.
Included commercial insurance, managed care, federal/state/local government insurance, workers’ compensation, and other types.
Occurred during adjuvant treatment or within 90 days after the last dose of adjuvant treatment.
Within 1 year prior to the index date, patients had a median of 14.0 outpatient visits, with most patients (75.8%) having no inpatient visits. Hypertension was present in a little over one-third of the patients (34.6%), while about one-tenth (10.3%) had diabetes. The majority (71.7%) were on Medicaid or uninsured. The most prevalent adjuvant treatments were anthracycline (58.4%) and HER2-targeted therapy (37.5%). The distributions of patient characteristics differed significantly across eFI groups, with the exception of census tract poverty and the types of adjuvant treatment, including anthracycline, HER2-targeted therapy, and CDK4/6 inhibitors.
For our primary outcome, the overall percentage of nonfatal adverse cardiovascular events among breast cancer patients receiving adjuvant therapy was 20.2% (n = 415). This percentage increased significantly with worsening frailty, rising from 17.8% in the robust group to 23.2% in the prefrail group and 29.4% in the frail group. Among 415 adverse cardiovascular events, there were 295 cases of cardiac arrhythmia, 54 cases of heart failure, 54 cases of cardiomyopathy, and 12 cases of atrial fibrillation and flutter.
Cumulative incidence rates of adverse cardiovascular events by frailty
The Central Illustration shows the cumulative incidence curves of nonfatal adverse cardiovascular events stratified by frailty. The overall median follow-up time was 217 days (Q1-Q3: 154-291 days). As shown in the figure, there was a significant difference in the cumulative incidence rates of adverse cardiovascular events (Gray’s method, P < 0.001), with higher rates observed in more frail patients.
Central Illustration.
Cumulative incidence of Adverse CV Events by Frailty
Solid lines represent cumulative incidences of adverse cardiovascular (CV) events for robust (green line), prefrail (orange line), and frail (red line) patients. The 2-year cumulative incidence rates of adverse CV events were 23.7% (95% CI: 20.6%-27.0%) for robust patients, 29.6% (95% CI: 24.9%-34.4%) for prefrail patients, and 42.6% (95% CI: 26.0%-58.3%) for frail patients. eFI = electronic health record frailty index; sHR = subdistribtution HR.
The 1-year cumulative incidence rates of adverse cardiovascular events were 21.4% (95% CI: 18.6%-24.2%) for robust patients, 26.7% (95% CI: 22.6%-31.0%) for prefrail patients, and 35.8% (95% CI: 24.9%-47.0%) for frail patients. The 2-year cumulative incidence rates were 23.7% (95% CI: 20.6%-27.0%) for robust patients, 29.6% (95% CI: 24.9%-34.4%) for prefrail patients, and 42.6% (95% CI: 26.0%-58.3%) for frail patients. The complete cumulative incidence curves by frailty, accounting for all-cause death as a competing risk, are shown in Supplemental Figure 1.
Results from multivariable analysis using Gray’s method
Results from the univariable and multivariable analysis using the Fine-Gray method are summarized in Table 2. Adjusting for covariates, prefrail (adjusted sHR: 1.35; 95% CI: 1.06-1.71; P = 0.015) and frail (adjusted sHR: 1.70; 95% CI: 1.11-2.61; P = 0.015) breast cancer patients were significantly more likely to experience nonfatal adverse cardiovascular events during adjuvant treatment compared with robust patients.
Table 2.
sHRs Estimating Association Between Frailty and Adverse Cardiovascular Events
| Unadjusted sHR (95% CI) | P Value | Adjusted sHR (95% CI) | P Value | |
|---|---|---|---|---|
| Frailty (eFI) | ||||
| Prefrail vs Robust | 1.34 (1.09-1.65) | 0.006 | 1.35 (1.06-1.71) | 0.015 |
| Frail vs Robust | 1.81 (1.28-2.55) | <0.001 | 1.70 (1.11-2.61) | 0.015 |
| Age | ||||
| 50-59 y vs <50 y | 1.24 (1.00-1.54) | 0.050 | 1.28 (1.02-1.61) | 0.035 |
| 60-69 y vs <50 y | 0.90 (0.67-1.21) | 0.468 | 0.96 (0.68-1.35) | 0.810 |
| ≥70 y vs <50 y | 1.28 (0.83-1.98) | 0.263 | 1.56 (0.91-2.69) | 0.107 |
| Race/ethnicity | ||||
| NHB vs NHW | 1.42 (1.12-1.80) | 0.003 | 1.42 (1.10-1.82) | 0.006 |
| Hispanic vs NHW | 0.96 (0.75-1.22) | 0.712 | 0.98 (0.76-1.25) | 0.848 |
| NHO/unknown vs NHW | 0.78 (0.51-1.20) | 0.262 | 0.74 (0.48-1.15) | 0.180 |
| Census tract ruralityb | ||||
| Nonurban vs urban | 0.96 (0.73-1.26) | 0.779 | 1.02 (0.77-1.36) | 0.883 |
| Unknown vs urban | 0.62 (0.31-1.24) | 0.177 | 0.84 (0.17-4.20) | 0.832 |
| Census tract povertya | ||||
| 10%-19% vs <10% | 0.94 (0.74-1.18) | 0.574 | 0.90 (0.70-1.14) | 0.370 |
| ≥20% vs <10% | 0.95 (0.74-1.24) | 0.722 | 0.88 (0.67-1.15) | 0.347 |
| Unknown vs <10% | 0.65 (0.35-1.20) | 0.170 | 0.88 (0.20-3.81) | 0.869 |
| Number of outpatient visitsd | 1.01 (1.00-1.02) | 0.013 | 1.01 (0.99-1.02) | 0.385 |
| Number of inpatient visitsd | ||||
| 1 vs 0 | 0.95 (0.73-1.25) | 0.727 | 0.90 (0.69-1.19) | 0.471 |
| >1 vs 0 | 1.32 (0.94-1.85) | 0.114 | 1.04 (0.72-1.51) | 0.817 |
| Hypertensiond | 1.14 (0.93-1.39) | 0.207 | 0.93 (0.72-1.19) | 0.564 |
| Diabetesd | 1.09 (0.80-1.49) | 0.580 | 0.84 (0.60-1.19) | 0.338 |
| Year adjuvant therapy started | ||||
| 2014-2015 vs 2012-2013 | 1.58 (1.14-2.18) | 0.006 | 1.67 (1.20-2.31) | 0.002 |
| 2016-2017 vs 2012-2013 | 1.71 (1.25-2.35) | <0.001 | 1.84 (1.33-2.54) | <0.001 |
| 2018-2019 vs 2012-2013 | 1.74 (1.25-2.42) | <0.001 | 1.92 (1.37-2.69) | <0.001 |
| 2020-2022 vs 2012-2013 | 1.41 (0.99-2.00) | 0.054 | 1.68 (1.16-2.43) | 0.006 |
| Insurance coverage | ||||
| Medicare vs all other insurancec | 1.11 (0.75-1.65) | 0.598 | 0.99 (0.63-1.57) | 0.978 |
| Medicaid/uninsured vs all other insurance | 1.23 (0.94-1.62) | 0.131 | 1.29 (0.95-1.74) | 0.099 |
| Unknown vs all other insurance | 3.68 (1.95-6.96) | <0.001 | 3.79 (1.98-7.26) | <0.001 |
| Anthracycline | 0.92 (0.76-1.13) | 0.439 | 1.11 (0.36-3.39) | 0.856 |
| HER2-targeted therapy | 1.15 (0.94-1.41) | 0.168 | 1.23 (0.40-3.78) | 0.716 |
| CDK4/6 inhibitors | 0.89 (0.31-2.56) | 0.835 | 0.98 (0.22-4.32) | 0.982 |
| 5-FU and capecitabine | 0.54 (0.30-0.99) | 0.048 | 0.64 (0.24-1.68) | 0.363 |
Robust = 0 ≤ eFI ≤ 0.10; prefrail = 0.10 < eFI ≤ 0.21; frail = 0.21 < eFI ≤ 1.
Abbreviations as in Table 1.
Determined by linking patient’s latest zip code to the Census Bureau’s American Community Survey and categorizing percent population below poverty into 3 groups: <10%, 10%-19%, or ≥20%.
Determined by linking patient’s latest zip code to the rural-urban commuting area codes.
Other insurance included commercial insurance, managed care, federal/state/local government insurance, workers’ compensation, and other types.
Measured within 1 year prior to the first date of adjuvant therapy.
Patients diagnosed between 50 and 59 years of age had a higher likelihood of adverse cardiovascular events than those diagnosed before 50 years of age (adjusted sHR: 1.28; 95% CI: 1.02-1.61; P = 0.035). Additionally, non-Hispanic Black patients were at a higher risk of adverse cardiovascular events compared with non-Hispanic White patients (adjusted sHR: 1.42; 95% CI: 1.10-1.82; P = 0.006).
Results from the restricted cubic splines indicated that the risk of adverse cardiovascular events increased in a monotonic dose-response pattern with eFI (Figure 2), with the likelihood ratio test showing no evidence of nonlinearity (P nonlinearity for cancer survivors = 0.197). In a sensitivity analysis that focused on heart failure and cardiomyopathy as the adverse cardiovascular events, we observed similar associations between frailty and adverse cardiovascular events in the multivariable analysis using the Fine-Gray method (Supplemental Table 4). Prefrail and frail patients were more likely to experience heart failure and cardiomyopathy compared with robust patients (adjusted sHR: 1.56; 95% CI: 1.01-2.41; P = 0.043).
Figure 2.
Restricted Cubic Splines Depicting Dose-Response Relationships Between eFI and Adverse Cardiovascular Events
The electronic health record frailty index (eFI) represents the cumulative deficit eFI. The competing-risk model adjusts for age at diagnosis, race/ethnicity, insurance, census tract–level rurality, census tract–level poverty, baseline hypertension and diabetes status, and baseline health care utilization, year adjuvant therapy started, and adjuvant treatment regimen. The solid red line shows fitted values, with dashed black lines representing the 95% CIs. The green dotted line is the reference line, with eFI = 0.21 as the reference in the curve.
In the interaction model, we observed a significant interaction term between eFI and race/ethnicity (P = 0.043). As a result, we reported the adjusted sHR for eFI within each racial/ethnic group in Table 3. Frailty was associated with higher rates of adverse cardiovascular events in non-Hispanic White and non-Hispanic Black breast cancer patients. Among non-Hispanic White breast cancer patients, prefrail individuals had an adjusted sHR of 1.48 (95% CI: 1.04-2.11; P = 0.031), and frail individuals had an adjusted sHR of 1.96 (95% CI: 1.10-3.50; P = 0.022), indicating a higher likelihood of experiencing adverse cardiovascular events during adjuvant treatment. Similarly, among non-Hispanic Black patients, prefrail individuals had an adjusted sHR of 1.59 (95% CI: 1.06-2.37; P = 0.024) and frail individuals had an adjusted sHR of 2.13 (95% CI: 1.11-4.10; P = 0.023).
Table 3.
sHR for Frailty in Different Race/Ethnicity Groups
| Race/Ethnicity | Frailty (eFI) | Number of Adverse Cardiovascular Events (%) | Adjusted sHR (95% CI) | P Value |
|---|---|---|---|---|
| NHW (n = 818) | Robust | 137 (26.0) | — | |
| Prefrail | 88 (36.7) | 1.48 (1.04-2.11) | 0.031 | |
| Frail | 25 (50.0) | 1.96 (1.10-3.50) | 0.022 | |
| NHB (n = 477) | Robust | 68 (27.3) | — | |
| Prefrail | 72 (37.3) | 1.59 (1.06-2.37) | 0.024 | |
| Frail | 16 (45.7) | 2.13 (1.11-4.10) | 0.023 | |
| Hispanic (n = 601) | Robust | 109 (27.4) | — | |
| Prefrail | 49 (28.8) | 0.92 (0.58-1.44) | 0.714 | |
| Frail | 14 (42.4) | 1.28 (0.57-2.88) | 0.554 | |
| NHO/unknown (n = 154) | Robust | 23 (24.7) | — | |
| Prefrail | 15 (28.3) | 1.24 (0.54-2.85) | 0.624 | |
| Frail | 2 (25.0) | 0.61 (0.07-5.13) | 0.658 |
The eFI by race/ethnicity interaction term was significant in the interaction model (P = 0.043). Robust = 0 ≤ eFI ≤ 0.10; prefrail = 0.10 < eFI ≤ 0.21; frail = 0.21 < eFI ≤ 1.
Abbreviations as in Table 1.
Discussion
In this study, we analyzed EHR data from a large clinical research network within PCORnet to examine the association between frailty and cardiotoxicity in breast cancer patients undergoing adjuvant chemotherapy and targeted therapy. Our analyses indicate that 20.2% of breast cancer patients experienced nonfatal adverse cardiovascular events potentially linked to cardiotoxicity after adjuvant therapy. Higher frailty levels were associated with an increased risk of these adverse events in the multivariable analysis. We also observed that both frail and prefrail patients had a greater likelihood of experiencing adverse cardiovascular events compared with robust patients, specifically within the non-Hispanic White and non-Hispanic Black subgroups. However, these associations were not statistically significant in other racial/ethnic groups.
In our study, the observed rate of nonfatal adverse cardiovascular events among breast cancer patients receiving adjuvant therapy was 20%, a slightly lower rate than from previous studies. One retrospective EHR study reported that the incidence of cardiotoxicity reached 34% among patients treated with trastuzumab combined with anthracycline.9 In a clinical trial involving metastatic breast cancer patients, the incidence of cardiotoxicity was 27% in those receiving a combination of trastuzumab, anthracycline, and cyclophosphamide, whereas it was substantially lower for patients receiving trastuzumab alone (3%-7%).33 Our data similarly showed that the concomitant use of chemotherapy and targeted therapy was associated with a higher rate of cardiotoxicity compared with using either therapy type, aligning with previous reports.
Regarding frailty, we found that 61.8% of breast cancer patients undergoing adjuvant treatments were robust, 32.0% were prefrail, and 6.2% were frail. These results are similar to those reported in previous studies that assessed “real-world” data.34, 35, 36 A retrospective study of women 65 years of age or older with stage I to III hormone receptor–positive/HER2-negative breast cancer utilized Surveillance, Epidemiology, and End Results–Medicare claims data and claims-based algorithms to measure frailty.34 At the time of diagnosis, 56% were classified as robust, whereas 37% and 7% were identified as prefrail and frail, respectively. Another clinical trial of breast cancer patients older than 65 years of age utilized a 35-item Searle index to measure frailty, finding that 76.6% of patients were robust, 18.3% were prefrail, and 5.1% were frail.35 A separate retrospective cohort study of Medicare beneficiaries with stage I to III breast cancer measured frailty using the deficit-accumulation frailty index and found that 49.5% were robust, 29.4% were prefrail, and 21.1% were frail.36
From a biological perspective, a higher burden of frailty may indicate advanced biological aging, which has been linked to an increased risk of fatal cardiovascular events in patients with a history of cancer. For instance, Zhang et al37 estimated biological aging based on 9 blood biomarkers related to inflammation, metabolic condition, and hematopoiesis in 2002 cancer patients, finding a significant association between biological aging and cardiovascular disease–specific death.
Additionally, we observed that higher frailty levels were associated with increased rates of nonfatal adverse cardiovascular events in non-Hispanic White and non-Hispanic Black patients but not in other racial/ethnic groups, such as Hispanic patients. One possible explanation is that the distribution of individual conditions within the eFI differed significantly between our non-Hispanic and Hispanic breast cancer patients, potentially contributing to the varying effects of frailty on cardiotoxicity by race/ethnicity. Future studies could examine how the pattern of individual conditions in eFI differs across patient subgroups and how these differences may impact the rates of adverse cardiovascular events.
Furthermore, prior research suggests that factors other than frailty—such as socioeconomic conditions (eg, low literacy, lack of health insurance, poor social integration, low income, or linguistic barriers),38 a lower likelihood of receiving optimal treatments,39 and comorbidities combined with environmental factors39—may play a more significant role than frailty in contributing to adverse cardiovascular events in minority racial/ethnic groups.
Study strengths and limitations
Our study has several strengths in both design and analysis. First, the use of EHRs from a large clinical research network allowed us to identify a large cohort of breast cancer patients who underwent adjuvant therapy, allowing for a detailed assessment of their medical history in real-word settings. Second, we used a comprehensive 43-item cumulative deficit EHR frailty index, providing a robust assessment of the frailty burden in breast cancer patients. Third, our analysis employed multiple analytical approaches, including multivariable analysis using the Fine-Gray method, interaction analysis, and dose-response analysis. These approaches enhance the accuracy of our results, help identify effect modifiers, and offer insights into risk gradients across frailty levels.
Our study also has several limitations worthy of note. First, we were unable to control for certain potentially important factors associated with cardiotoxicity, such as cancer stage, hormone receptor status, radiation therapy, laterality of tumors, and functional status (eg, Eastern Cooperative Oncology Group or Knowledge and Practice Standards). Most of these factors are not available in EHRs. Second, we used a 90-day cutoff after the end of adjuvant treatments to identify adverse cardiovascular events, which may have missed later-onset events, such as heart failure and cardiomyopathy, that can develop after anthracycline exposure. Third, our findings are based on a real-world cohort of breast cancer patients who underwent adjuvant therapy after surgery, which may limit the generalizability of our results to patients in other data sources or those receiving different types of treatments.
Conclusions
Our findings underscore the importance of conducting a geriatric assessment, with a particular focus on frailty, for breast cancer patients prior to adjuvant systemic chemotherapy or targeted therapy. Such assessments can help guide personalized treatment plans, allowing for a more tailored approach to care. Moreover, our findings emphasize the need for close monitoring of cardiotoxicity in frail breast cancer patients undergoing adjuvant treatments to improve the management of cardiovascular risks. Future research should explore the complex mechanisms connecting frailty to cardiotoxicity and develop strategies to reduce cardiovascular risks for breast cancer patients receiving adjuvant treatments.
Perspectives.
COMPETENCY IN MEDICAL KNOWLEDGE: Our findings highlight the need for close monitoring of cardiotoxicity in frail breast cancer patients undergoing adjuvant treatments for improve the management of cardiovascular risks.
TRANSLATIONAL OUTLOOK: Future research should investigate the complex mechanisms connecting frailty to cardiotoxicity and develop interventions to reduce cardiovascular risks in breast cancer patients receiving adjuvant treatments.
Funding Support and Author Disclosures
This work was supported by funding from the Florida Breast Cancer Foundation. Drs Guo and Bian were funded in part by National Cancer Institute grants 1R01CA284646, 5R01CA246418-02, 3R01CA246418-02S1, 1R21CA245858-01, 3R21CA245858-01A1S1, and 1R21CA253394-01A1; National Institute on Aging grants 1R01AG080624-01 and 5R21AG068717-02; and Centers for Disease Control and Prevention grant U18DP006512. Dr Guo was also funded in part by National Institute of Mental Health grant 5R21MH129682-02. Dr Zhang is an employee of Janssen Research and Development, and this work was initiated while he was on faculty at the University of Florida. All other authors have reported that they have no relationships relevant to the contents of this paper to disclose.
Footnotes
The authors attest they are in compliance with human studies committees and animal welfare regulations of the authors’ institutions and Food and Drug Administration guidelines, including patient consent where appropriate. For more information, visit the Author Center.
Appendix
For supplemental tables and a figure, please see the online version of this paper.
Contributor Information
Dongyu Zhang, Email: dzhan107@its.jnj.com.
Yi Guo, Email: yiguo@ufl.edu.
Appendix
References
- 1.Siegel R.L., Giaquinto A.N., Jemal A. Cancer statistics, 2024. CA Cancer J Clin. 2024;74:12–49. doi: 10.3322/caac.21820. [DOI] [PubMed] [Google Scholar]
- 2.Gillespie H.S., McGann C.J., Wilson B.D. Noninvasive diagnosis of chemotherapy related cardiotoxicity. Curr Cardiol Rev. 2011;7:234–244. doi: 10.2174/157340311799960672. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Pondé N.F., Zardavas D., Piccart M. Progress in adjuvant systemic therapy for breast cancer. Nat Rev Clin Oncol. 2019;16:27–44. doi: 10.1038/s41571-018-0089-9. [DOI] [PubMed] [Google Scholar]
- 4.Kerr A.J., Dodwell D., McGale P., et al. Adjuvant and neoadjuvant breast cancer treatments: A systematic review of their effects on mortality. Cancer Treat Rev. 2022;105 doi: 10.1016/j.ctrv.2022.102375. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Berry D.A., Cronin K.A., Plevritis S.K., et al. Effect of screening and adjuvant therapy on mortality from breast cancer. N Engl J Med. 2005;353:1784–1792. doi: 10.1056/NEJMoa050518. [DOI] [PubMed] [Google Scholar]
- 6.Bradshaw P.T., Stevens J., Khankari N., Teitelbaum S.L., Neugut A.I., Gammon M.D. Cardiovascular disease mortality among breast cancer survivors. Epidemiology. 2016;27:6–13. doi: 10.1097/EDE.0000000000000394. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Doyle J.J., Neugut A.I., Jacobson J.S., Grann V.R., Hershman D.L. Chemotherapy and cardiotoxicity in older breast cancer patients: a population-based study. J Clin Oncol. 2005;23:8597–8605. doi: 10.1200/JCO.2005.02.5841. [DOI] [PubMed] [Google Scholar]
- 8.Tarantini L., Gori S., Faggiano P., et al. Adjuvant trastuzumab cardiotoxicity in patients over 60 years of age with early breast cancer: a multicenter cohort analysis. Ann Oncol. 2012;23:3058–3063. doi: 10.1093/annonc/mds127. [DOI] [PubMed] [Google Scholar]
- 9.Guglin M., Hartlage G., Reynolds C., Chen R., Patel V. Trastuzumab-induced cardiomyopathy: not as benign as it looks? A retrospective study. J Card Fail. 2009;15:651–657. doi: 10.1016/j.cardfail.2009.04.011. [DOI] [PubMed] [Google Scholar]
- 10.Dhir A.A., Sawant S.P. Cardiac morbidity & mortality in patients with breast cancer: a review. Indian J Med Res. 2021;154:199–209. doi: 10.4103/ijmr.IJMR_879_20. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.Kojima G., Liljas A.E.M., Iliffe S. Frailty syndrome: implications and challenges for health care policy. Risk Manag Healthc Policy. 2019;12:23–30. doi: 10.2147/RMHP.S168750. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.Chen X., Mao G., Leng S.X. Frailty syndrome: an overview. Clin Interv Aging. 2014;9:433–441. doi: 10.2147/CIA.S45300. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Tsai H.-H., Yu J.-C., Hsu H.-M., et al. The impact of frailty on breast cancer outcomes: evidence from analysis of the Nationwide Inpatient Sample, 2005-2018. Am J Cancer Res. 2022;12:5589–5598. [PMC free article] [PubMed] [Google Scholar]
- 14.Wang S., Yang T., Qiang W., et al. The prevalence of frailty among breast cancer patients: a systematic review and meta-analysis. Support Care Cancer. 2022;30:2993–3006. doi: 10.1007/s00520-021-06641-8. [DOI] [PubMed] [Google Scholar]
- 15.Zhang D., Mobley E.M., Manini T.M., et al. Frailty and risk of mortality in older cancer survivors and adults without a cancer history: evidence from the National Health and Nutrition Examination Survey, 1999-2014. Cancer. 2022;128:2978–2987. doi: 10.1002/cncr.34258. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.George M., Smith A., Sabesan S., Ranmuthugala G. Physical comorbidities and their relationship with cancer treatment and its outcomes in older adult populations: systematic review. JMIR Cancer. 2021;7 doi: 10.2196/26425. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17.Farooqi M.A.M., Gerstein H., Yusuf S., Leong D.P. Accumulation of deficits as a key risk factor for cardiovascular morbidity and mortality: a pooled analysis of 154 000 individuals. J Am Heart Assoc. 2020;9 doi: 10.1161/JAHA.119.014686. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18.Hou Y., Xu C., Lu Q., et al. Associations of frailty with cardiovascular disease and life expectancy: a prospective cohort study. Arch Gerontol Geriatr. 2022;99 doi: 10.1016/j.archger.2021.104598. [DOI] [PubMed] [Google Scholar]
- 19.Hogan W.R., Shenkman E.A., Robinson T., et al. The OneFlorida Data Trust: a centralized, translational research data infrastructure of statewide scope. J Am Med Inform Assoc. 2021;29:686–693. doi: 10.1093/jamia/ocab221. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20.Valiyaveettil D., Joseph D., Malik M. Cardiotoxicity in breast cancer treatment: causes and mitigation. Cancer Treat Res Commun. 2023;37 doi: 10.1016/j.ctarc.2023.100760. [DOI] [PubMed] [Google Scholar]
- 21.Rushton M., Lima I., Tuna M., et al. Impact of stopping trastuzumab in early breast cancer: a population-based study in Ontario, Canada. J Natl Cancer Inst. 2020;112:1222–1230. doi: 10.1093/jnci/djaa054. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22.Bowles E.J.A., Wellman R., Feigelson H.S., et al. Risk of heart failure in breast cancer patients after anthracycline and trastuzumab treatment: a retrospective cohort study. J Natl Cancer Inst. 2012;104:1293–1305. doi: 10.1093/jnci/djs317. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23.Mandelblatt J.S., Huang K., Makgoeng S.B., et al. Preliminary development and evaluation of an algorithm to identify breast cancer chemotherapy toxicities using electronic medical records and administrative data. J Oncol Pract. 2015;11:e1–e8. doi: 10.1200/JOP.2013.001288. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24.Chen J., Long J.B., Hurria A., Owusu C., Steingart R.M., Gross C.P. Incidence of heart failure or cardiomyopathy after adjuvant trastuzumab therapy for breast cancer. J Am Coll Cardiol. 2012;60:2504–2512. doi: 10.1016/j.jacc.2012.07.068. [DOI] [PubMed] [Google Scholar]
- 25.Pajewski N.M., Lenoir K., Wells B.J., Williamson J.D., Callahan K.E. Frailty screening using the electronic health record within a Medicare Accountable Care Organization. J Gerontol A Biol Sci Med Sci. 2019;74:1771–1777. doi: 10.1093/gerona/glz017. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26.Clegg A., Bates C., Young J., et al. Development and validation of an electronic frailty index using routine primary care electronic health record data. Age Ageing. 2016;45:353–360. doi: 10.1093/ageing/afw039. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 27.Olivieri-Mui B.L., Shi S.M., McCarthy E.P., Habtemariam D., Kim D.H. Beyond the health deficit count: examining deficit patterns in a deficit-accumulation frailty index. J Am Geriatr Soc. 2021;69:792–797. doi: 10.1111/jgs.16955. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 28.Rockwood K., Andrew M., Mitnitski A. A comparison of two approaches to measuring frailty in elderly people. J Gerontol A Biol Sci Med Sci. 2007;62:738–743. doi: 10.1093/gerona/62.7.738. [DOI] [PubMed] [Google Scholar]
- 29.Searle S.D., Mitnitski A., Gahbauer E.A., Gill T.M., Rockwood K. A standard procedure for creating a frailty index. BMC Geriatr. 2008;8:24. doi: 10.1186/1471-2318-8-24. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 30.Cromartie J., Economic Research Service Rural-Urban Comuting Area Codes. Documentation. https://www.ers.usda.gov/data-products/rural-urban-commuting-areacodes/documentation/
- 31.Kosalka P., Johnson C., Turek M., et al. Effect of obesity, dyslipidemia, and diabetes on trastuzumab-related cardiotoxicity in breast cancer. Curr Oncol. 2019;26:e314–e321. doi: 10.3747/co.26.4823. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 32.Menna P., Minotti G., Salvatorelli E. Cardiotoxicity of targeted cancer drugs: concerns, “the cart before the horse,” and lessons from trastuzumab. Curr Cardiol Rep. 2019;21:33. doi: 10.1007/s11886-019-1121-0. [DOI] [PubMed] [Google Scholar]
- 33.Seidman A., Hudis C., Pierri M.K., et al. Cardiac dysfunction in the trastuzumab clinical trials experience. J Clin Oncol. 2002;20:1215–1221. doi: 10.1200/JCO.2002.20.5.1215. [DOI] [PubMed] [Google Scholar]
- 34.Calip G.S., Hubbard C., Nabulsi N.A., et al. Real-world patterns of treatment and recurrence by frailty status among older women with HR-positive, HER2-negative early breast cancer. J Clin Oncol. 2020;38 535-535. [Google Scholar]
- 35.Mandelblatt J.S., Cai L., Luta G., et al. Frailty and long-term mortality of older breast cancer patients: CALGB 369901 (Alliance) Breast Cancer Res Treat. 2017;164:107–117. doi: 10.1007/s10549-017-4222-8. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 36.Yan C.H., Coleman C., Nabulsi N.A., et al. Associations between frailty and cancer-specific mortality among older women with breast cancer. Breast Cancer Res Treat. 2021;189:769–779. doi: 10.1007/s10549-021-06323-3. [DOI] [PubMed] [Google Scholar]
- 37.Zhang D., Leeuwenburgh C., Zhou D., et al. Analysis of biological aging and risks of all-cause and cardiovascular disease-specific death in cancer survivors. JAMA Netw Open. 2022;5 doi: 10.1001/jamanetworkopen.2022.18183. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 38.Lindley K.J., Aggarwal N.R., Briller J.E., et al. Socioeconomic determinants of health and cardiovascular outcomes in women. J Am Coll Cardiol. 2021;78:1919–1929. doi: 10.1016/j.jacc.2021.09.011. [DOI] [PubMed] [Google Scholar]
- 39.Graham G. Disparities in cardiovascular disease risk in the United States. Curr Cardiol Rev. 2015;11:238–245. doi: 10.2174/1573403X11666141122220003. [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.