Abstract
Background:
Breast cancer survival is increasing, making late effects such as cardiovascular disease (CVD) more relevant. The purpose of this study was to evaluate incident CVD following breast cancer diagnosis among long-term survivors and to investigate possible risk factors for CVD.
Methods:
A population-based cohort of 6,641 breast cancer survivors diagnosed between 1997 and 2009 who survived at least 10-years was identified within the Utah Cancer Registry (UCR). In addition, 36,612 cancer-free women from the general population, matched by birth year and state, were identified within the Utah Population Database (UPDB). Cox Proportional Hazards models were used to calculate CVD hazard ratios (HRs) for >10 to 15, and >15 years.
Results:
Long-term breast cancer survivors had an increased risk of newly diagnosed diseases of the circulatory system (HR 1.32, 99% CI 1.00–1.75) from 10 to 15 years following cancer diagnosis compared with the general population. No increased CVD risks were observed after fifteen years. Breast cancer survivors with Charlson Comorbidity Index score ≥2 had a significantly higher risk of diseases of the circulatory system (HR 2.64, 95% CI 1.08–6.45) beyond 10 years following breast cancer diagnosis. Similarly, older age, obesity, lower education, and family history of CVD and breast cancer were risk factors for heart and circulatory system diseases among long-term breast cancer survivors.
Conclusion:
Risk of CVD compared to the general population was moderate among this cohort of long-term breast cancer survivors between 10 to 15 years since cancer diagnosis. Awareness of CVD risks is important for breast cancer survivors.
Keywords: Cancer Survivorship, Breast Cancer, Cardiovascular Diseases, Late Effects, Matched Cohort Study
Precis:
Risk of cardiovascular diseases between 10 to 15 years since cancer diagnosis was moderate among this cohort of long-term breast cancer survivors compared to the general population. Awareness of CVD risks is important for breast cancer survivors.
Introduction
The number of breast cancer survivors in the United States (U.S.) is expected to surpass 4 million by 2026.1 Nearly 3.8 million breast cancer survivors in the U.S. were alive in 2020, with an additional 276,480 new cases diagnosed that year.1 The continued improvement in survival for breast cancer is due in part to breast cancer treatment effectiveness, and early screening practices and detection.2 Hence the 5-year survival rate of breast cancer was 90%, based on 2010–2016 data.3 With the long-term survival following breast cancer diagnosis, late effects of treatment such as cardiovascular disease (CVD) are a concern.
Previous population-based studies of CVD risks between 4 and 10 years following a breast cancer diagnosis in breast cancer survivors compared to general population cohorts reported inconsistent findings.4–9 One questionnaire-based study that reported CVD outcomes >10 years after breast cancer diagnosis did not find an increased heart disease risk among breast cancer survivors compared with cancer-free women from the general population; there were few CVD events >15 years after cancer diagnosis to estimate risks.8
Among current risk factor studies for cardiovascular outcomes among breast cancer survivors, nearly all were based on follow-up time <10 years, were limited to older women over 65 years of age and Medicare recipients10–18 or those diagnosed before 2002,11,19 and none had cancer-free comparisons. The purpose of our study is to evaluate incident cardiovascular diseases following breast cancer diagnosis among a contemporary cohort of long-term breast cancer survivors compared to a general population cohort of women without cancer and to investigate possible risk factors for late onset CVDs among breast cancer survivors.
Methods
Study Population
Women diagnosed with a first primary breast cancer (primary site ICD-O-3 C50.0 to C50.9) identified within the Utah Cancer Registry (UCR) were included in the study. For eligibility, women had to be a) Utah residents, b) ≥18 years of age at the time of diagnosis, and c) diagnosed between 1997 and 2017 with breast cancer. Breast cancer survivors were matched on birth year (±2 years) and birth state with up to five cancer-free women from the general population. Eligibility criteria for the general population were that women had lived in Utah since 1996 or were ≤30 years of age at the time they first moved to Utah to assure adequate follow-up time for the women from the general population to become a case. A total of 783 breast cancer survivors were excluded for unknown stage (Figure 1). We also restricted the study population to women with >10 years of follow-up (which also meant that the diagnosis year started in 2009), and the final sample size was 6,641 breast cancer survivors and 36,612 women in the general population cohort.
Figure 1.
Exclusion criteria for breast cancer survivors and the general population.
Data Sources and Outcome Measures
The Utah Population Database (UPDB) contained records from the UCR, Utah driver’s license, statewide vital, demographic, and family history information that link to medical information. All cancer patients from the Utah Cancer Registry are linked to the Utah Population Database. UPDB uses record linking IBM® InfoSphere® QualityStage software to perform probabilistic records linking to various databases, including the UCR. Variables from the UCR included: age at cancer diagnosis (years), birth year, race, ethnicity, rural residence, treatment (surgery, chemotherapy, radiotherapy, and endocrine therapy), tumor grade, ER status, PR status, second primary cancers, laterality, and AJCC staging. Variables from the UPDB included: education, baseline body mass index, baseline Charlson Comorbidity index (from the ICD diagnosis codes), family history of breast cancer, family history of cardiovascular disease, and tobacco use (from ICD diagnosis codes).
The primary CVD outcome measures were identified by available International Classification of Disease (ICD) coding, specifically, ICD-9 and ICD-10 diagnosis codes through the statewide Ambulatory Surgery database; Inpatient Hospital Claims database from the Utah Department of Health, which includes information on primary, secondary, and additional diagnosis; and hospital discharge information. We also used the electronic medical record (EMR) data from the two biggest healthcare providers in Utah, the University of Utah Health and Intermountain Healthcare. The Clinical Classification Software (CCS) as a part of the International Classification of Diseases (9th Revision) was created by the Healthcare Cost and Utilization Project (HCUP) that collapses over 14,000 diagnosis codes into clinically meaningful categories for the outcome measures.20 Specifically, as a multi-level hierarchical grouping system, CCS includes 72 cardiovascular outcomes grouped as follows: a) level 1 (diseases of circulatory system), b) level 2 (hypertension, diseases of the heart, cerebrovascular diseases, diseases of the arteries, arterioles, and capillaries, and diseases of the veins and lymphatics), c) level 3 (twenty-six outcomes) such as cerebrovascular disease, and d) level 4 (forty outcome) such as diseases of the arteries, arterioles, and capillaries (Table 3).
Table 3.
Circulatory System Disease Risk at >10–15 and >15 Years After Cancer Diagnosis in Breast Cancer Survivors in Comparison With a General Population Cohort of Women #
| >10 – 15 years | >15 years | |||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Cardiovascular Diseases | Patients | General Population | Hazard Ratios (99% Confidence Interval) | Patients | General Population | Hazard Ratios (99% Confidence Interval) | ||||||
| N (%) | N (%) | N (%) | N (%) | |||||||||
| 7 Diseases of the Circulatory System | 231 | (33.4) | 361 | (26.2) | 1.32 | (1.00 to 1.75) † | 59 | (30.0) | 88 | (27.0) | 0.90 | (0.46 to 1.77) |
| 7.1 Hypertension | 375 | (14.0) | 991 | (13.3) | 1.07 | (0.89 to 1.30) | 138 | (14.0) | 300 | (12.7) | 0.90 | (0.63 to 1.41) |
| 7.1.1 Essential Hypertension | 252 | (9.1) | 645 | (8.3) | 0.81 | (0.62 to 1.05) | 38 | (3.7) | 76 | (2.9) | 0.50 | (0.19 to 1.33) |
| 7.1.2 Hypertension with comp./Secondary Hypertension | 175 | (3.0) | 521 | (2.4) | 1.02 | (0.76 to 1.37) | 24 | (0.9) | 85 | (1.0) | 0.50 | (0.17 to 1.57) |
| 7.3 Cerebrovascular Disease | 277 | (4.9) | 836 | (4.0) | 1.24 | (1.00 to 1.53) † | 114 | (4.8) | 319 | (4.0) | 0.90 | (0.54 to 1.35) |
| 7.3.1 Acute Cerebrovascular Disease | 99 | (1.6) | 306 | (1.3) | 1.00 | (0.67 to 1.48) | 18 | (0.7) | 43 | (0.5) | 0.49 | (0.11 to 2.32) |
| 7.3.2 Occlusion or Stenosis of Precerebral Arteries | 61 | (1.0) | 199 | (0.8) | 0.82 | (0.48 to 1.39) | 13 | (0.5) | 32 | (0.3) | 0.70 | (0.15 to 3.48) |
| 7.3.4 Transient Cerebral Ischemia (transient ischemic attack) | 54 | (0.9) | 226 | (1.0) | 0.79 | (0.48 to 1.30) | 13 | (0.5) | 32 | (0.4) | 0.60 | (0.13 to 2.94) |
| 7.4 Diseases of the Arteries, Arterioles, and Capillaries | 659 | (17.0) | 1561 | (12.8) | 1.42 | (1.23 to 1.64) † | 259 | (17.4) | 612 | (14.5) | 1.08 | (0.80 to 1.44)† |
| 7.4.1 Peripheral and Visceral Artherosclerosis | 150 | (2.6) | 407 | (1.8) | 1.08 | (0.78 to 1.49) | 27 | (1.1) | 72 | (0.8) | 0.80 | (0.29 to 1.80) |
| 7.4.2 Aortic, Peripheral, and Visceral Artery Aneurysms | 36 | (0.6) | 103 | (0.4) | 0.96 | (0.48 to 1.94) | –* | (0.3) | 23 | (0.2) | 1.30 | (0.31 to 5.12) |
| 7.4.3 Aortic and Peripheral Arterial Embolism or Thrombosis | 11 | (0.2) | 44 | (0.2) | 0.65 | (0.19 to 2.20) | –* | (0.2) | –* | (0.1) | 1.00 | (0.06 to 15.0) |
| 7.4.4.1 Hypotension | 185 | (3.1) | 465 | (2.1) | 1.13 | (0.84 to 1.54) | 30 | (1.2) | 58 | (0.7) | 1.40 | (0.56 to 3.33) |
| 7.5 Diseases of the Veins and Lymphatics | 412 | (14.1) | 900 | (11.0) | 1.31 | (1.09 to 1.59) | 126 | (11.8) | 256 | (10.0) | 1.10 | (0.71 to 1.68) |
| 7.2 Diseases of the Heart | 423 | (18.8) | 831 | (14.6) | 1.29 | (1.07 to 1.56) | 163 | (19.3) | 293 | (15.3) | 1.22 | (0.85 to 1.77)† |
| 7.2.1 Heart Valve Disorders | 249 | (4.9) | 604 | (3.3) | 0.97 | (0.74 to 1.28)† | 46 | (2.1) | 78 | (1.1) | 0.89 | (0.38 to 2.05)† |
| 7.2.2 Peri-, Endo-, and Myocarditis and Cardiomyopathy | 111 | (1.8) | 245 | (1.1) | 1.30 | (0.88 to 1.93)† | 30 | (1.2) | 40 | (0.4) | 1.77 | (0.70 to 4.44)† |
| 7.2.2.1 Cardiomyopathy | 78 | (1.3) | 177 | (0.7) | 1.34 | (0.84 to 2.14)† | 19 | (0.7) | 26 | (0.3) | 1.08 | (0.27 to 4.83) |
| 7.2.2.2 Other Peri-, Endo-, and Myocarditis | 47 | (0.8) | 99 | (0.4) | 1.21 | (0.64 to 2.29)† | 12 | (0.5) | 20 | (0.2) | 2.19 | (0.65 to 7.38) |
| 7.2.3 Acute Myocardial Infarction | 63 | (1.0) | 179 | (0.7) | 1.02 | (0.62 to 1.68) | –* | (0.3) | 36 | (0.4) | 0.20 | (0.02 to 2.78) |
| 7.2.4 Coronary Atherosclerosis and Other Heart Diseases | 186 | (3.2) | 575 | (2.9) | 0.94 | (0.70 to 1.24) | 40 | (1.8) | 95 | (1.3) | 0.79 | (0.35 to 1.83) |
| 7.2.6 Pulmonary Heart Disease | 184 | (3.1) | 486 | (2.2) | 1.00 | (0.73 to 1.35)† | 43 | (1.7) | 65 | (0.8) | 1.22 | (0.52 to 2.86) |
| 7.2.8 Conduction Disorders | 133 | (2.2) | 380 | (1.6) | 0.94 | (0.66 to 1.34) | 36 | (1.4) | 69 | (0.8) | 0.51 | (0.16 to 1.71) |
| 7.2.9 Cardiac Dysrhythmias | 335 | (7.7) | 821 | (5.7) | 0.96 | (0.76 to 1.22)† | 53 | (3.0) | 115 | (2.2) | 0.83 | (0.40 to 1.73) |
| 7.2.10 Cardiac Arrest and Ventricular Fibrillation | 38 | (0.6) | 104 | (0.4) | 1.28 | (0.79 to 2.01) | –* | (0.2) | 20 | (0.2) | 0.43 | (0.03 to 6.17) |
| 7.2.11 Congestive Heart Failure, Non-hypertensive | 190 | (3.3) | 489 | (2.2) | 0.96 | (0.71 to 1.31) | 23 | (0.9) | 66 | (0.8) | 0.54 | (0.16 to 1.80) |
Models adjusted for baseline Charlson Comorbidity Index (CCI), Body Mass Index (BMI), Baseline Tobacco Use, Race, Ethnicity, Birth Year, and Birth State.
Estimates changed when adjusting for imputed body mass index (BMI). Estimate without imputation reported.
Suppressed since the number was less than 11.
BMI imputation estimates for >10–15 years that became insignificant following imputation was observed for Cerebrovascular Disease (HR = 1.11, 99%CI 0.92 to 1.32). Changes in estimates were observed for: Diseases of the Circulatory System (HR = 1.35, 99%CI 1.08 to 1.68, and HR = 1.30, 99%CI 1.16 to 1.47), Diseases of the Arteries, Arterioles, and Capillaries (HR = 1.34, 99%CI 1.19 to 1.52, and HR = 1.28, 99%CI 1.18 to 1.39 using flexible model), Heart Valve Disorders (HR = 1.36, 99%CI 1.12 to 1.65), Peri-, Endo-, and Myocarditis and Cardiomyopathy (HR = 1.57, 99%CI 1.17 to 2.12), Cardiomyopathy (HR = 1.57, 99%CI 1.10 to 2.23, and HR = 1.76 99%CI 1.37 to 2.26 using flexible model), Other Peri-, Endo-, and Myocarditis (HR = 1.69, 99%CI 1.07 to 2.67, and HR = 1.79, 99%CI 1.30 to 2.47 using flexible model), Pulmonary Heart Disease (HR = 1.27, 99%CI 1.01 to 1.60, and HR = 1.32, 99%CI 1.13 to 1.54 using flexible model), and Cardiac Dysrhythmias (HR = 1.26, 99%CI 1.06 to 1.49, and HR = 1.36, 99%CI 1.21 to 1.52 using flexible model) following BMI imputation.
BMI imputation estimates that changed for >15 years were observed for: Diseases of the Arteries, Arterioles, and Capillaries (HR = 1.23, 99%CI 1.01 to 1.51), Diseases of the Heart (HR = 1.31, 99%CI 1.01 to 1.70, and HR = 1.32, 99%CI 1.13 to 1.53 using flexible model), Heart Valve Disorders (HR = 1.68, 99%CI 1.03 to 2.73, and HR = 1.53, 99%CI 1.12 to 2.08 using flexible model, tobacco use estimate was not generated for insufficient observations), and Peri-, Endo-, and Myocarditis and Cardiomyopathy (HR = 2.42, 99%CI 1.28 to 4.55, and HR = 2.44, 99%CI 1.59 to 3.75 using flexible model, tobacco use estimate was not generated for insufficient observations).
Tobacco users were identified by available ICD-9 codes for “tobacco use disorder” (305.1), available ICD-10 codes for “nicotine dependence” (F17.200, 21, 22, 29, 201, 203, 208, 209, 210, 211, 213, 218, 219, 220, 221, 223, 228, 229, 290, 291, 293, 298, and F17.299), “tobacco use disorder complicating pregnancy, childbirth, or the puerperium” (99406 and 99407), and available CPT codes for “tobacco abuse counseling” (Z71.6), “tobacco use” (Z72.0), and “nicotine poisoning” (T65.2), based on the American Academy of Family Physicians coding guidelines.21 Urban and rural location was classified based on Rural-Urban Continuum Codes (RUCC) from 2013,22 with the classification scheme that distinguishes metropolitan and nonmetropolitan areas based on county.
Follow-up time was calculated starting from the date of breast cancer diagnosis (index date), the point at which a survivor is considered to be at risk until the first cardiovascular disease incidence or censoring time (last date of follow-up, no outcome, or death). Prevalent cardiovascular disease diagnosis that occurred before the index date were excluded to allow for calculation of incident events (Supplementary Table S1). The index date of the breast cancer survivors was assigned to the women matched from the general population to allow for the calculation of follow-up time. A report by the National Association of Health Data Organization on interstate exchange of nonresident data for public health purposes and research, reported that Utah had a small percentage of residents who seek health care outside of the state; specifically, Utah residents discharged in border states ranges from 0.02% in Nevada, 0.08% in Arizona, and 0.13% in Colorado.23 Similarly, in 2019 the U.S. Census Bureau’s state-to-state migration flow, reported fairly low (approximately 3.0%) out-migration rate among Utahns.24 This allows us to capture nearly the entire diagnosis history of Utah women in our sample.
This study was approved by the University of Utah Institutional Review Board (IRB) and the Resource for Genetic and Epidemiologic Research (RGE), the oversight committee for the UPDB.
Statistical Analysis
Baseline descriptive demographic characteristics between the breast cancer survivors and general population cohorts for categorical variables were compared using Pearson Chi-square (X2) tests. To evaluate potential confounders, we assessed the three properties of a confounder. For the association between cancer diagnosis and cardiovascular diseases risk (Table 3), potential confounders that we adjusted on were: baseline Charlson Comorbidity Index (CCI), baseline body mass index (BMI), baseline tobacco use, race, ethnicity. These factors are risk factors for cardiovascular disease, associated with breast cancer diagnosis, and do not act as mediators. Smoking is not considered a strong risk factor for breast cancer, but some recent studies suggest an association.25 We adjusted on birth year and birth state to account for the matching. For the assessment of risk factors among breast cancer survivors, we adjusted on race and ethnicity, BMI, CCI, baseline tobacco, education, histology, rural residence, age at diagnosis, family history of CVDs, tumor grade, laterality, and diagnosis year, where appropriate. These factors include established and potential risk factors for CVD, the covariates are associated with the exposure of interest and do not act as mediators.
Cox Proportional Hazards models were used to calculate hazard ratios (HRs) for long-term cardiovascular outcomes from 10 to 15 years, and >15 years after breast cancer diagnosis compared with the women from the general population. We assessed the 10–15-year follow-up period, which may interest patients and clinicians to understand the possible risk of CVDs in a 5-year period after reaching 10 years post cancer diagnosis. However, we could not stratify further into the 15–20-year period due to the small number of patients for this follow-up time period. Additionally, we assessed risk factors for heart and circulatory system diseases among breast cancer survivors; therefore, we did not use the general population as the comparison group. Ninety-nine percent confidence intervals were used to account for multiple testing due to the large number of outcomes (n=72).26 Cox Proportional Hazards models were adjusted for the two matching factors, birth state and birth year, and additionally for baseline body mass index (BMI), baseline Charlson Comorbidity Index (CCI), race, and ethnicity. Proportional hazard assumption was tested by creating interactions of the predictors and time (survival time) and included in the model. Models in violation of the proportional hazard assumption were tested with flexible parametric modeling with restricted splines and reported where substantial differences existed. In addition, a modified Cox model was estimated using Fine and Gray’s model of competing risks analysis to account for death as a competing outcome rather than censoring those patients that died before experiencing an outcome.27 When inferences differed between tradition and modified Cox model, the HRs estimates were reported from both models. We also reported risk estimates for CVD outcomes over 10 years for comparison to the 10–15 and over 15 years of follow-up estimates (Supplementary Table S2). Additionally, we performed sensitivity analysis for the CVD risk estimates without including the stage IV breast cancer patients due to the difference in 5-year survival by stage (Supplementary Table S3).
Baseline CCI was calculated without cancer to avoid an overestimation of the score calculation. Baseline BMI values were calculated from height and weight information provided in the driver’s license records at least one year prior to the breast cancer diagnosis or index date. Given 30.2% of missing BMI values, BMI was imputed based on cancer diagnosis, baseline CCI, race and age at breast cancer diagnosis using the linear regression model. We estimated the HRs with and without imputation of BMI to assess the impact of imputation. When inferences differed between the two models, the HRs without imputed BMI were reported.
The a priori α level used in all statistical analysis was P < 0.05, except as noted, where P < 0.01 was considered statistically significant given the large number of cardiovascular outcomes. Analyses were performed in Statistical Software Package or STATA 15.0 (STATA Inc., College Station, TX) and Statistical Analysis System software or SAS 9.4 (SAS Institute Inc., Cary, NC).
Results
Race, ethnicity, education, family history of breast cancer, and family history of cardiovascular disease, and baseline body mass index (BMI) differed between long-term breast cancer survivors and the general population cohorts (p < 0.001; Table 1). Baseline tobacco-use distribution was similar between the two groups. The long-term breast cancer survivors had a higher CCI score than women from the general population (p = 0.002).
Table 1.
Demographic Characteristics Among 10-year Breast Cancer Survivors and the General Population
| Characteristics: N (%) | Breast Cancer (N = 6,641) |
General Population (N = 36,612) |
P-value† | ||
|---|---|---|---|---|---|
| Birth Year (years) | |||||
| <1930 | 760 | (11.4) | 3,856 | (10.5) | 0.001 |
| 1930–1939 | 1,166 | (17.6) | 6,107 | (16.7) | |
| 1940–1949 | 1,894 | (28.5) | 10,134 | (27.7) | |
| 1949–1959 | 1,856 | (28.0) | 10,733 | (29.3) | |
| >1960 | 965 | (14.5) | 5,782 | (15.8) | |
| Race | |||||
| White | 6,339 | (95.5) | 34,690 | (94.8) | <.0001 |
| Black | 11 | (0.2) | 86 | (0.2) | |
| Asian | 77 | (1.2) | 304 | (0.8) | |
| Pacific Islander | 14 | (0.2) | 70 | (0.2) | |
| American Indian/Alaskan Native | –* | (0.1) | 167 | (0.5) | |
| Multiracial | 190 | (2.8) | 701 | (1.9) | |
| Unknown | –* | (0.0) | 594 | (1.6) | |
| Hispanic | |||||
| Yes | 596 | (9.0) | 2,700 | (7.4) | <.0001 |
| No | 6,045 | (91.0) | 33,912 | (92.6) | |
| Education | |||||
| High School or Less | 940 | (14.2) | 5,148 | (14.1) | <.0001 |
| High School | 2,192 | (33.0) | 12,423 | (33.9) | |
| Some College | 1,945 | (29.2) | 11,178 | (30.5) | |
| College | 921 | (13.9) | 5,025 | (13.7) | |
| Beyond College | 643 | (9.7) | 2,838 | (7.8) | |
| Baseline Body Mass Index (kg/m2) ** | |||||
| <18.5 | 91 | (2.0) | 762 | (2.1) | <.0001 |
| 18.5–24.9 | 2,246 | (48.7) | 14,392 | (40.3) | |
| 25.0–29.0 | 1,446 | (31.3) | 11,603 | (32.4) | |
| >30.0 | 833 | (18.1) | 8,991 | (25.2) | |
| Baseline Charlson Comorbidity Index | |||||
| 0 | 4,869 | (73.3) | 27,593 | (75.4) | 0.002 |
| 1 | 1,186 | (17.9) | 6,056 | (16.5) | |
| ≥2 | 586 | (8.8) | 2,963 | (8.1) | |
| Family History of Breast Cancer ¥ | |||||
| Yes | 2,929 | (44.1) | 15,378 | (42.0) | 0.001 |
| No | 3,712 | (55.9) | 21,234 | (58.0) | |
| Family History of Cardiovascular Disease ¥ | |||||
| Yes | 4,237 | (63.8) | 25,539 | (69.8) | |
| No | 2,404 | (36.2) | 11,073 | (30.2) | <.0001 |
| Baseline Tobacco Use | |||||
| Yes | 242 | (3.6) | 1,363 | (3.7) | 0.80 |
| No | 6,399 | (96.4) | 35,249 | (96.3) | |
Two-sided Pearson Chi-square test.
In first, second, and third-degree relatives.
Suppressed since the number was less than 11.
Total missing BMI, n = 2,889 (6.8%): 2,025 (4.7%) of breast cancer survivors and 864 (2.0%) of general population had missing BMI values.
Approximately 27.7% and 28.4% of the breast cancer patients were diagnosed between 45–54 and 55–65 years of age, respectively (Table 2). The majority of long-term breast cancer survivors were diagnosed with stage I breast cancer (50.2%), and had estrogen (ER) and progesterone (PR) receptor overexpression (75.6% and 67.5%, respectively). The majority of long-term breast cancer survivors had surgery (99.5%), followed by radiotherapy (55.7%), and chemotherapy (43.9%).
Table 2.
Clinical Characteristics Among 10-year Breast Cancer Survivors Diagnosed 1997–2009
| Characteristics: N (%) | N = 6,641 | |
|---|---|---|
| Age at Cancer Diagnosis (years) | ||
| 19–45 | 1,057 | (15.9) |
| 45–54 | 1,837 | (27.7) |
| 55–64 | 1,889 | (28.4) |
| 65–74 | 1,303 | (19.6) |
| 75–95 | 555 | (8.4) |
| Surgery | ||
| Yes | 6,608 | (99.5) |
| No | 33 | (0.5) |
| Chemotherapy | ||
| Yes | 2,916 | (43.9) |
| No | 3,597 | (54.2) |
| Unknown | 128 | (1.9) |
| Radiotherapy | ||
| Yes | 3,699 | (55.7) |
| No | 2,861 | (43.1) |
| Unknown | 81 | (1.2) |
| Endocrine Therapy | ||
| Yes | 2,022 | (30.4) |
| No | 4,427 | (66.7) |
| Unknown | 192 | (2.9) |
| Tumor Grade | ||
| Grade I | 1,396 | (21.0) |
| Grade II | 2,788 | (42.0) |
| Grade III/IV | 2,083 | (31.4) |
| Unknown\Not stated | 374 | (5.6) |
| Estrogen-receptor (ER) Status # | ||
| Positive | 5,017 | (75.6) |
| Negative | 1,262 | (19.0) |
| Unknown | 362 | (5.4) |
| Progesterone-receptor (PR) Status # | ||
| Positive | 4,481 | (67.5) |
| Negative | 1,732 | (26.1) |
| Unknown | 428 | (6.4) |
| Laterality | ||
| Right | 3,257 | (49.0) |
| Left | 3,384 | (51.0) |
| Second Primary* | ||
| Yes | 972 | (14.6) |
| No | 5669 | (85.4) |
| AJCC Staging ** | ||
| I | 3,333 | (50.2) |
| II | 2,701 | (40.7) |
| III | 510 | (7.7) |
| IV | 97 | (1.4) |
| Rural Residence † | ||
| Yes | 671 | (10.1) |
| No | 5,970 | (89.9) |
The American Joint Committee on Cancer (AJCC).
Breast cancer patients with breast cancer as first of two or more primary cancers.
Residence was classified based on Rural-Urban Continuum Codes (RUCC) 2013.
Human Epidermal Growth Factor Receptor 2 (HER2) for cancer subtype characterization was not available until 2010.
From 10 to 15 years following breast cancer diagnosis, approximately 33.4% of breast cancer survivors had a new diagnosis of diseases of the circulatory system compared with 26.2% of the general population cohort (Table 3). Specifically, from 10 to 15 years of follow-up, breast cancer survivors had a 32% (HR = 1.32, 99% CI 1.00 to 1.75, p = 0.009, and HR = 1.31, 99% CI 0.99 to 1.73, p = 0.011 using competing risks model) higher risk of diseases of the circulatory system, 42% (HR = 1.42, 99% CI = 1.23 to 1.64, and HR = 1.39, 99% CI 1.20 to 1.61 using competing risks model) higher risk of diseases of the arteries, arterioles, and capillaries, 24% (HR = 1.24, 99% CI = 1.00 to 1.53, p = 0.009, and HR = 1.19, 99%CI 0.96 to 1.48, p = 0.035 using competing risks model) higher risk of cerebrovascular disease, and 31% (HR = 1.31, 99% CI = 1.09 to 1.59, and HR = 1.29, 99% CI 1.06 and 1.56 using competing risks analysis) higher risk of diseases of the veins and lymphatics compared with the women from the general population, adjusting for baseline CCI, BMI, tobacco use, race, ethnicity, birth year, and birth state. When we compared the risk estimates adjusting for the original BMI variable vs. imputed BMI, the estimates for heart valve disorders, cardiomyopathy, and pulmonary heart disease were significant only when adjusting for imputed BMI as specified in the footnotes (Table 3). Thus, in the tables, the estimates without BMI imputations were reported. Overall, no associations were observed for any cardiovascular diseases after more than 15 years of follow-up (Table 3).
From 10 to 15 years of follow-up, 18.8% of breast cancer survivors had a diagnosis of diseases of the heart compared with 14.6% of women from the general population (Table 3). From >10 to 15 years of follow-up, breast cancer survivors had a 29% (HR = 1.29, 99% CI = 1.07 to 1.56, and HR = 1.27, 99%CI 1.05 to 1.55 using competing risks analysis) higher risk of diseases of the heart compared with the women from the general population, adjusting for baseline CCI, BMI, tobacco use, race, ethnicity, birth year, and birth state. However, we did not observe an increase in risk in any of the specific heart diseases such as cardiomyopathy, myocardial infarction or congestive heart failure in this cohort of long-term breast cancer survivors. Consistent with the primary analysis, most risk estimates for >10 to 15 years of follow-up in breast cancer survivors compared with the general population were similar for >10 years of follow-up (Supplementary Table S2). When restricted to women diagnosed with stage I-III breast cancer, increased CVD risks were observed in the >10 to 15 years of follow-up period among breast cancer survivors compared with the general population (Supplementary Table S3). Although the overall circulatory system disease risk was no longer statistically significant for stage I-III breast cancer patients, the HRs for the other CVDs did not change our inferences. Consistent with the primary analysis, no associations were observed for any CVDs for >15 years of follow-up in the sensitivity analysis (Supplementary Table S3).
Among breast cancer survivors, baseline and demographic risk factors were assessed for diseases of the heart and the circulatory system diagnosed after 10 years of follow-up (Table 4). Older age, lower education level (some high school or less), family history of cardiovascular diseases, and family history of breast cancer were observed as risk factors for both diseases of the heart and of the circulatory system among long-term breast cancer survivors. Baseline CCI was a risk factor for diseases of the circulatory system, while baseline BMI (>30 kg/m2) and white race were risk factors for diseases of the heart. Clinical characteristics and breast cancer treatments were not associated with either diseases of the heart or the circulatory system among long-term breast cancer survivors (Table 5). Breast cancer survivors with second or multiple primary cancers had a 59% higher risk (p value = 0.05) of the circulatory system disease compared with women that had breast cancer as one primary in their lifetime (Table 5).
Table 4.
Demographic and Baseline Risk Factors of Diseases of the Heart and Circulatory System 10 Years After Breast Cancer Diagnosis
| Risk Factors | Diseases of the Heart | Diseases of the Circulatory System | ||
|---|---|---|---|---|
| HR (95% CI) | HR (95% CI) | |||
| Age at Cancer Diagnosis (years) a | ||||
| 19–45 | REF | REF | ||
| 45–54 | 1.32 | (0.91 to 1.92) | 1.18 | (0.79 to 1.75) |
| 55–64 | 2.06 | (1.43 to 2.97) | 1.78 | (1.18 to 2.70) |
| 65–74 | 2.58 | (1.72 to 3.86) | 2.01 | (1.14 to 3.53) |
| 75–95 | 4.20 | (2.55 to 6.91) | 4.01 | (1.72 to 9.35) |
| P-value for trend | <.0001 | <.0001 | ||
| Baseline Body Mass Index (kg/m2) b | ||||
| <18.5 | 0.51 | (0.19 to 1.38) | 0.77 | (0.36 to 1.67) |
| 18.5–24.9 | REF | REF | ||
| 25.0–29.0 | 1.20 | (0.95 to 1.52) | 1.02 | (0.71 to 1.47) |
| >30.0 | 1.62 | (1.22 to 2.14) | 1.07 | (0.61 to 1.86) |
| P-value for trend | <.0001 | <.0001 | ||
| Baseline Charlson Comorbidity Index b | ||||
| 0 | REF | REF | ||
| 1 | 1.10 | (0.86 to 1.42) | 1.03 | (0.66 to 1.60) |
| ≥2 | 1.34 | (0.80 to 2.25) | 2.64 | (1.08 to 6.45) |
| P-value for trend | <.0001 | <.0001 | ||
| Education c | ||||
| High School or Less | 1.41 | (1.29 to 1.54) | 1.28 | (1.15 to 1.42) |
| High School | REF | REF | ||
| Some College | 0.89 | (0.83 to 0.96) | 1.00 | (0.93 to 1.08) |
| College | 0.80 | (0.73 to 0.88) | 0.99 | (0.99 to 1.09) |
| Beyond College | 0.81 | (0.72 to 0.90) | 0.85 | (0.75 to 0.95) |
| Baseline Tobacco Use d | ||||
| No | REF | REF | ||
| Yes | 0.93 | (0.74 to 1.16) | 1.06 | (0.82 to 1.39) |
| Race e | ||||
| White | REF | REF | ||
| Other | 0.74 | (0.65 to 0.84) | 0.76 | (0.67 to 0.86) |
| Hispanic e | ||||
| No | REF | REF | ||
| Yes | 1.12 | (1.01 to 1.24) | 0.97 | (0.86 to 1.09) |
| Family History of Cardiovascular Disease f | ||||
| No | REF | REF | ||
| Yes | 1.24 | (1.16 to 1.32) | 1.19 | (1.11 to 1.27) |
| Family History of Breast Cancer f | ||||
| No | REF | REF | ||
| Yes | 1.19 | (1.13 to 1.27) | 1.11 | (1.04 to 1.18) |
| Rural Residence f | ||||
| Urban | REF | REF | ||
| Rural | 0.80 | (0.63 to 1.01) | 0.83 | (0.61 to 1.12) |
Hazard Ratio (HR), Confidence Interval (CI).
Models adjusted for:
BMI, CCI, Race/Ethnicity, Baseline Tobacco Use, Education, and Histology.
Race/Ethnicity, Baseline Tobacco Use, Education, Rural Residence, Age at Diagnosis, and Family History of CVDs.
Race/Ethnicity, and Rural Residence.
Race/Ethnicity, Education, and Rural Residence.
Rural Residence.
Race/Ethnicity.
Table 5.
Clinical Risk Factors for Disease of the Heart and Circulatory System 10 Years After Breast Cancer Diagnosis
| Risk Factors | Diseases of the Heart | Diseases of the Circulatory System | ||
|---|---|---|---|---|
| HR (95% CI) | HR (95% CI) | |||
| Diagnosis Year (years) a | ||||
| 1997–2001 | REF | REF | ||
| 2002–2005 | 0.95 | (0.75 to 1.21) | 0.95 | (0.69 to 1.33) |
| 2006–2009 | 0.87 | (0.63 to 1.20) | 0.72 | (0.46 to 1.13) |
| Surgery b | ||||
| No | REF | REF | ||
| Yes | 1.62 | (0.22 to 11.8) | 0.06 | (0.01 to 0.60) |
| Chemotherapy b | ||||
| No | REF | REF | ||
| Yes | 1.24 | (0.99 to 1.57) | 1.17 | (0.86 to 1.60) |
| Radiotherapy b | ||||
| No | REF | REF | ||
| Yes | 1.18 | (0.95 to 1.47) | 1.20 | (0.89 to 1.62) |
| Endocrine Therapy b | ||||
| No | REF | REF | ||
| Yes | 1.25 | (0.99 to 1.57) | 1.40 | (1.00 to 1.95) |
| Endocrine Receptor Status c | ||||
| ER+ and/or PR+ | REF | REF | ||
| ER−/PR− | 0.98 | (0.74 to 1.30) | 1.12 | (0.79 to 1.59) |
| Second Primary d,** | ||||
| No | REF | REF | ||
| Yes | 1.32 | (0.99 to 1.77) | 1.59 | (1.00 to 2.52) |
| Tumor Grade e | ||||
| Grade I | REF | REF | ||
| Grade II | 1.04 | (0.80 to 1.37) | 1.26 | (0.83 to 1.92) |
| Grade III | 0.92 | (0.69 to 1.24) | 0.93 | (0.59 to 1.47) |
| AJCC Stage e,* | ||||
| I | REF | REF | ||
| II | 1.00 | (0.81 to 1.24) | 1.20 | (0.88 to 1.62) |
| III | 0.95 | (0.58 to 1.57) | 0.85 | (0.41 to 1.78) |
| IV | 1.53 | (0.78 to 3.02) | 1.64 | (0.79 to 3.45) |
Hazard Ratio (HR), Confidence Interval (CI).
The American Joint Committee on Cancer (AJCC).
Breast cancer patients with breast cancer as first of two or more primary cancers.
Models adjusted for:
BMI, CCI, Race/Ethnicity, Baseline Tobacco Use, Education, and Age at Diagnosis.
BMI, CCI, Race/Ethnicity, Education, Rural Residence, Tumor Grade, Laterality, Diagnosis Year, and Age at Diagnosis.
BMI, CCI, Race/Ethnicity, Baseline Tobacco Use, and Age at Diagnosis.
BMI, CCI, Race/Ethnicity, Baseline Tobacco Use, Education, Rural Residence, Tumor Grade, Age at Diagnosis, Family History of Breast Cancer, Endocrine Therapy, Chemotherapy, and Radiotherapy.
BMI, CCI, Race/Ethnicity, Baseline Tobacco Use, Diagnosis Year, and Age at Diagnosis.
Discussion
Long-term breast cancer survivors had a higher risk of diseases of the circulatory system, diseases of the arteries, arterioles, and capillaries, diseases of the veins and lymphatics, and diseases of the heart when compared with the general population cohort in the follow-up period of >10 to 15 years. However, the risk did not differ by diagnosis year, tumor stage, tumor grade, and breast cancer treatments among this cohort of long-term breast cancer survivors. With a longer follow-up of >15 years, breast cancer survivors did not have a higher risk of CVD outcomes in comparison to the general population. Long-term breast cancer survivors were more likely to have a higher CCI score and family history of breast cancer than women from the general population.
In terms of overall incident CVDs, we observed a 33% higher risk of diseases of the circulatory system and 29% higher risk of diseases of the heart among long-term breast cancer survivors compared with the general population in the follow-up period of >10 to 15 years. Large-scale population-based studies of breast cancer survivors comparing CVD risks in breast cancer survivors to general population cohorts have been inconsistent. Some reported an increased or moderate CVD risk,7–9,12 while others reported no increased CVD risk for breast cancer patients compared to the general population.10,11 However, most of these studies did not examine the period beyond 10 years of follow-up. Khan et al. included breast cancer survivors who survived at least 5 years (10.2 mean years of follow up from cancer diagnosis) and reported an elevated incidence of congestive heart failure and coronary artery disease compared with the general population.8 In contrast, a study from the Netherlands by Schoormans et al. observed no increased CVD risk within 8–13 years following diagnosis among 6,762 breast cancer survivors compared to 6,762 cancer-free women from the general population.10 In a study of 2,535 breast cancer patients most of whom were identified in the Cancer Genetics Network, Hill et al. observed no increased risk for heart disease for >10 years after cancer diagnosis compared with the general population.11 A possible explanation for the difference in risk estimations from the Schoormans and Hill studies may be that our results, as well as those from the Khan study, are based on medical records rather than patient self-report data. Previous studies have shown that patients may confuse various CVD diagnoses, and 30% of self-reported cardiovascular outcomes were misclassified by patients.28
In the follow-up period of more than 15 years, we did not observe a higher risk for CVD outcomes in breast cancer survivors compared with the general population. There were 2,929 breast cancer survivors and 17,168 women from the general population within the >15 years of follow-up period (163 heart disease diagnoses in breast cancer survivors vs. 293 heart disease diagnoses in the general population). Although it is possible that we did not detect any associations in the 15-year follow-up because of lower statistical power, our results also support the possibility that CVD risk becomes similar over time in breast cancer survivors and women from the general population. To our knowledge, our study is the first to report on CVD risk estimates for long-term breast cancer survivors for 15 or more years after cancer diagnosis compared to cancer-free women from the general population.
In terms of CVD risk factors among long-term breast cancer survivors in our study, the predictors for both heart diseases and circulatory system diseases were older age, obesity, family history of CVDs and family history of breast cancer, and low education. Baseline CCI was a risk factor for diseases of the circulatory system. Older age and higher BMI have also been shown in other studies to be risk factors for heart disease or hospitalization after early-stage breast cancer,7 or congestive heart failure following trastuzumab among older breast cancer patients treated with radiotherapy, but were assessed at 10 years of follow-up.21 Another long-term study among breast cancer survivors diagnosed between 1958 and 2001 reported that patients with higher BMI, diabetes, or presence of other CVDs had higher rates of major coronary events following radiotherapy in right- vs. left- treated tumors after 10 years of follow-up.14
In our study, we did not observe an association between radiotherapy, chemotherapy, or endocrine therapy with the late-onset CVDs among this cohort of long-term breast cancer survivors. For radiotherapy treatment, we did not observe an increased risk of heart and circulatory system diseases. Previously, one long-term study among breast cancer patients reported an increased rate of major coronary events within 10 to 19 and >20 years,14 and another study reported an increase in incidence of all heart disease after 10 to 14 and >15 years29 following radiation in women with left- vs right- breast cancers. Our risk estimate for heart disease among patients receiving radiation therapy was not statistically significant but suggestive of an association (HR = 1.18, 95% CI = 0.95 to 1.47); we may have lacked statistical power to detect the risk since we focused on the long-term survivors who had smaller numbers of heart disease events. Similarly, the lack of association maybe explained by improvements in radiotherapy techniques over time that have contributed to reducing radiation exposure to the heart.
A major strength of our study is that we were able to assess a comprehensive list of cardiovascular disease outcomes over long-term follow-up in a large sample size of 6,641 breast cancer patients. We also had a large comparison group of 36,612 women without cancer from the general population who were matched on birth year and birth state. The population-based design and inclusion of all women diagnosed with breast cancer in Utah identified by the Utah Cancer Registry during the study period minimizes potential selection bias. Further, our assessment of outcomes is based on electronic medical records from two of the largest medical care providers in Utah and the statewide hospital ambulatory surgery and discharge records and are not subject to recall bias. While our study may miss less severe cardiovascular diagnoses, the available ICD diagnosis codes in our study allow us to capture severe cases. The high proportion of breast cancer patients followed up over the study minimizes survival bias which is an issue for survivorship studies relying on self-reported outcomes. Within five years following a cancer diagnosis, increased medical surveillance may result in a bias away from the null. However, our study is based on breast cancer patients starting at 10 years after cancer diagnosis, thus surveillance bias should be minimized.
This study had some limitations. Utah is becoming more diverse, and 9% of the breast cancer patients in our sample were Hispanic. However, our cohort of breast cancer survivors and the general population did not have a strong representation of other race and ethnicity groups. Thus, our study results may not be generalizable to other more diverse states. More detailed cancer treatment information, especially chemotherapy agents and related dosage, was not available in our study and could help further explain differences in risk across specific cardiovascular diseases among breast cancer patients. Similarly, we may not have captured the majority of breast cancer patients receiving extended endocrine therapy beyond 5 years to detect late-cardiac effects. However, endocrine therapy use in this cohort may be underrepresented, and therefore is likely to underestimate or bias towards to the null the long-term cardiac risk in this cohort of breast cancer survivors. In addition, we were unable to identify patients with cancer recurrence who would have initiated additional treatments that could influence development of additional CVD outcomes. However, we evaluated whether breast cancer patients with second or multiple primaries had higher risks of CVDs (Table 5). Further, developing methods to obtain more detailed treatment information and recurrence from the EMR and claims databases would be a future direction of research. Data on the expression of human epidermal growth factor receptor 2 (HER2) was unavailable until 2010, although estrogen and progesterone receptor expression status was available for this cohort of long-term breast cancer survivors.
Although we were able to assess cardiovascular disease outcomes for >15 years of follow-up, there were fewer breast cancer survivors in the >15-year follow-up period, which may explain why some possible associations were not detected for the later time period. Similarly, we may have lacked statistical power to detect an association for specific diseases of the heart such as congestive heart failure or myocardial infarction (<100 observations). While there is a potential for poorly ascertained cardiovascular diagnoses codes by ICD coding, administrative ICD-9 codes for heart failure and related comorbidities had 95% specificity and 93% positive predictive value.30 Any coding errors may result in non-differential misclassification or underestimation of the risk estimate. However, we observed higher risks in breast cancer survivors than in the general population cohort.
In summary, increased risk of cardiovascular diseases, namely heart diseases, diseases of the arteries, arterioles, capillaries, and diseases of veins and lymphatics, was observed among long-term breast cancer survivors compared with the matched general population cohort. However, for patients followed beyond 15 years after diagnosis, increased risks of any cardiovascular diseases were not observed for breast cancer survivors compared to the general population. Our findings highlight the importance of survivorship care, with monitoring and intervention for a broader spectrum of cardiovascular diseases, following breast cancer for survivors at increased risk. Further, factors affecting cardiovascular diseases in this study, such as higher BMI, may suggest opportunities for cardiovascular disease prevention in breast cancer patients during and years following treatment.
Supplementary Material
Funding:
This work was supported by grants from the NIH (R21 CA185811, R03 CA159357, M.Hashibe, PI), the Huntsman Cancer Institute, and the Cancer Control and Population Sciences Program (HCI Cancer Center Support Grant P30CA042014). This research was supported by the Utah Cancer Registry, which is funded by the National Cancer Institute’s SEER Program, Contract No. HHSN261201800016I, the U.S. Center for Disease Control and Prevention’s National Program of Cancer Registries, Cooperative Agreement No. NU58DP0063200-01, with additional support from the University of Utah and Huntsman Cancer Foundation. Partial support for all datasets within the Utah Population Database is provided by the University of Utah, Huntsman Cancer Institute and the Huntsman Cancer Institute Cancer Center Support grant, P30 CA42014 from the National Cancer Institute.
Footnotes
Conflict of Interest: The authors have no conflict of interest to disclose.
References
- 1.Cancer Stat Facts: Female Breast Cancer. National Cancer Institute, Surveillance, Epidemiology, and End Results Program. Accessed April 2, 2021. https://seer.cancer.gov/statfacts/html/breast.html
- 2.American Cancer Society: Cancer Facts and Figures 2019. Accessed April 2, 2021. https://www.cancer.org/latest-news/report-number-of-cancer-survivors-continues-to-grow.html
- 3.Women and Heart Disease. Centers for Disease Control and Prevention. Accessed April 2, 2021. https://www.cdc.gov/heartdisease/women
- 4.Abdel-Qadir H, Thavendiranathan P, Austin PC, et al. The risk of heart failure and other cardiovascular hospitalizations after early-stage breast cancer: a matched cohort study. J Natl Cancer Inst. 2019;111(8):854–862. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Khan NF, Mant D, Carpenter L, Forman D, Rose PW. Long-term health outcomes in a British cohort of breast, colorectal and prostate cancer survivors: a database study. Br J Cancer. 2011;105(1):S29–S37. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Strongman H, Gadd S, Matthews A, et al. Medium and long-term risks of specific cardiovascular diseases in survivors of 20 adult cancers: a population-based cohort study using multiple linked UK electronic health records databases. The Lancet. 2019;394(10203):1041–1054. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Schoormans D, Vissers PAJ, van Herk-Sukel MPP, et al. Incidence of cardiovascular disease up to 13 year after cancer diagnosis: A matched cohort study among 32 757 cancer survivors. Cancer Med. 2018;7(10):4952–4963. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Hill DA, Horick NK, Isaacs C, et al. Long-term risk of medical conditions associated with breast cancer treatment. Breast Cancer Res Treat. 2014;145(1):233–243. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.Park NJ, Chang Y, Bender C, et al. Cardiovascular disease and mortality after breast cancer in postmenopausal women: Results from the Women’s Health Initiative. PloS One. 2017;12(9):e0184174. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.Boekel NB, Schaapveld M, Gietema JA, et al. Cardiovascular Disease Risk in a Large, Population-Based Cohort of Breast Cancer Survivors. Int J Radiat Oncol Biol Phys. 2016;94(5):1061–1072. [DOI] [PubMed] [Google Scholar]
- 11.Darby SC, Ewertz M, McGale P, et al. Risk of ischemic heart disease in women after radiotherapy for breast cancer. N Engl J Med. 2013;368(11):987–998. [DOI] [PubMed] [Google Scholar]
- 12.Tan CH, Chao TT, Liu JC, et al. Breast cancer therapy and age difference in cardiovascular disease risks: A population-based cohort study in Taiwan. Taiwan J Obstet Gynecol. 2016;55(1):98–103. [DOI] [PubMed] [Google Scholar]
- 13.Boero IJ, Paravati AJ, Triplett DP, et al. Modern Radiation Therapy and Cardiac Outcomes in Breast Cancer. Int J Radiat Oncol Biol Phys. 2016;94(4):700–708. [DOI] [PubMed] [Google Scholar]
- 14.Rehammar JC, Jensen MB, McGale P, et al. Risk of heart disease in relation to radiotherapy and chemotherapy with anthracyclines among 19,464 breast cancer patients in Denmark, 1977–2005. Radiother Oncol. 2017;123(2):299–305. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.Doyle JJ, Neugut AI, Jacobson JS, et al. Radiation therapy, cardiac risk factors, and cardiac toxicity in early-stage breast cancer patients. Int J Radiat Oncol Biol Phys. 2007;68(1):82–93. [DOI] [PubMed] [Google Scholar]
- 16.Chen J, Long JB, Hurria A, Owusu C, Steingart RM, Gross CP. Incidence of heart failure or cardiomyopathy after adjuvant trastuzumab therapy for breast cancer. J Am Coll Cardiol. 2012;60(24):2504–2512. [DOI] [PubMed] [Google Scholar]
- 17.Pinder MC, Duan Z, Goodwin JS, Hortobagyi GN, Giordano SH. Congestive heart failure in older women treated with adjuvant anthracycline chemotherapy for breast cancer. J Clin Oncol. 2007;25(25):3808–3815. [DOI] [PubMed] [Google Scholar]
- 18.Gong IY, Verma S, Yan AT, et al. Long-term cardiovascular outcomes and overall survival of early-stage breast cancer patients with early discontinuation of trastuzumab: a population-based study. Breast Cancer Res Treat. 2016;157(3):535–544. [DOI] [PubMed] [Google Scholar]
- 19.Chavez-MacGregor M, Zhang N, Buchholz TA, et al. Trastuzumab-related cardiotoxicity among older patients with breast cancer. J Clin Oncol. 2013;31(33):4222–4228. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20.Elixhauser A, Steiner C, Palmer L. Clinical Classifications Software (CCS), 2015. U.S. Agency for Healthcare Research and Quality. Accessed April 2, 2021. http://www.hcup-us.ahrq.gov/toolssoftware/ccs/ccs.jsp
- 21.American Academy of Family Physicians. Coding Reference. Tobacco Use Prevention and Cessation Counseling. CPT Copyright 2016 American Medical Association. Accessed April 2, 2021. https://www.aafp.org/dam/AAFP/documents/patient_care/tobacco/codes-tobacco-cessation-counseling.pdf
- 22.United States Census Bureau. 2013. Census Bureau Region and Division Codes and State FIPS codes. Accessed April 2, 2021. https://www.census.gov/geographies/reference-files/2013/demo/popest/2013-geocodes-all.html
- 23.Next Steps for the Inter-State Exchange of Non-resident Data Between State Health Data Organizations. NAHDO Summary of Findings and Proposed Next Steps. By: The National Association of Health Data Organizations. September 2009. Accessed August 16, 2021. https://www.nahdo.org/sites/default/files/Resources/Publications/nex%20steps%20in%20x%20border.pdf [Google Scholar]
- 24.US Census Bureau. State-to-State Migration Flow. 2020. Accessed April 28, 2021. https://www.census.gov/data/tables/time-series/demo/geographic-mobility/state-to-state-migration.html
- 25.Park HA, Neumeyer S, Michailidou K, Bolla MK, Wang Q, Dennis J et al. Mendelian randomization study of smoking exposure in relation to breast cancer risk. Br J Cancer. 2021;125(8):1135–45. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26.Gelman A, Hill J, Yajima M. Why we (usually) don’t have to worry about multiple comparisons. J R Educ Eff. 2012;5(2):189–211. [Google Scholar]
- 27.Fine JP, Gray RJ. A proportional hazards model for the subdistribution of a competing risk. J Am Stat Assoc. 1999;94(446):496–509. [Google Scholar]
- 28.Barr EL, Tonkin AM, Welborn TA, Shaw JE. Validity of self‐reported cardiovascular disease events in comparison to medical record adjudication and a statewide hospital morbidity database: the AusDiab study. Intern Med J. 2009;39(1):49–53. [DOI] [PubMed] [Google Scholar]
- 29.McGale P, Darby SC, Hall P, et al. Incidence of heart disease in 35,000 women treated with radiotherapy for breast cancer in Denmark and Sweden. Radiother Oncol. 2011;100(2):167–175. [DOI] [PubMed] [Google Scholar]
- 30.Birman-Deych E, Waterman AD, Yan Y, Nilasena DS, Radford MJ, Gage BF. Accuracy of ICD-9-CM codes for identifying cardiovascular and stroke risk factors. Med Care. 2005;43(5):480–485. [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.