A, Cumulative incidence function for cause-specific death in patients younger than 66 years at breast cancer diagnosis. B, Cumulative incidence function for cause-specific death in patients 66 years or older at breast cancer diagnosis. CV indicates cardiovascular.
A, Cumulative incidence function for patients without CV disease. B, Cumulative incidence function for patients with a CV diagnosis documented before breast cancer diagnosis.
eAppendix 1. Key Findings from Prior Peer-Reviewed Publications Assessing Risk of Death from Breast Cancer Relative to Cardiovascular Disease
eAppendix 2. Sources Used for Collection of Data on the Cohort
eTable 1. List of Codes Used for the Determination of Outcomes
eTable 2. Baseline Characteristics of Patients in the Full Cohort and Those Who Died During Available Follow-up, Classified by Cause of Death
eTable 3. Cause-Specific Hazard Ratios From Regression Models of Time to Cardiovascular Death
eFigure 1. Cumulative Incidence of Cause-Specific Death in Women With Early-Stage Breast Cancer, and Age-matched Women Without Cancer
eFigure 2. Landmark Analysis of the Cumulative Incidence of Cause-Specific Death Among Women Who Survived 5 Years Beyond Breast Cancer Diagnosis
Customize your JAMA Network experience by selecting one or more topics from the list below.
Abdel-Qadir H, Austin PC, Lee DS, et al. A Population-Based Study of Cardiovascular Mortality Following Early-Stage Breast Cancer. JAMA Cardiol. 2017;2(1):88–93. doi:10.1001/jamacardio.2016.3841
How likely are women diagnosed with early-stage breast cancer to die of breast cancer compared with cardiovascular disease?
Overall, 21 123 of the 98 999 women in this population-based cohort study who were diagnosed with breast cancer died during follow-up: 49.9% from breast cancer and 16.3% from cardiovascular disease. However, women with prior cardiovascular disease or older women who had survived at least 5 years after diagnosis were more likely to die from cardiovascular disease than breast cancer by 10 years of follow-up.
Cardiovascular death is an important competing risk in women with early-stage breast cancer, particularly for older women and those with a history of cardiovascular disease.
There is increasing interest in the effect of cardiovascular disease on cancer survivors. However, there are limited contemporary population-based data on the risk of cardiovascular death after early-stage breast cancer.
To describe the incidence of cardiovascular death in a contemporary population of women with early-stage breast cancer while accounting for competing risks.
Design, Setting, and Participants
A population-based cohort study was conducted among 98 999 women diagnosed with early-stage breast cancer between April 1, 1998, and March 31, 2012. Patients were followed up until death or were censored on December 31, 2013. Baseline characteristics were determined from administrative databases and the Ontario Cancer registry. Vital statistics data were used to determine the cause of death. Cumulative incidence functions were used to estimate the incidence of cause-specific mortality. We studied the association between baseline characteristics and rates of cardiovascular death using cause-specific hazard functions. The analyses accounted for competing risks of noncardiovascular death. Statistical analysis was performed from July 16, 2015, to August 4, 2016.
Early-stage breast cancer, age, cardiovascular disease, hypertension, and diabetes.
Main Outcomes and Measures
Cause of death, which was classified as breast cancer, cardiovascular disease, other cancers, or other noncancer causes.
Of the 98 999 women (median age, 60 years [interquartile range, 50-71 years]) in the study, 21 123 (21.3%) died during follow-up. The median time to death was 4.2 years (IQR, 2.2-7.1 years). Breast cancer was the most common cause of death (10 550 deaths [49.9%]); 3444 deaths [16.3%] were from cardiovascular causes. Cardiovascular death was infrequent in women younger than 66 years without prior cardiovascular disease, diabetes, or hypertension. Among women 66 years or older, the risks of breast cancer death and cardiovascular death at 10 years were 11.9% (95% CI, 11.6%-12.3%) and 7.6% (95% CI, 7.3%-7.9%), respectively. Among patients with prior cardiovascular disease, the risk of death from breast cancer and cardiovascular disease were equivalent for the first 5 years, after which death from cardiovascular causes was more frequent (10-year cumulative incidence, 14.6% [95% CI, 13.7%-15.4%] for breast cancer vs 16.9% [95% CI, 16.0%-17.8%] for cardiovascular disease). For women 66 years or older who survived 5 years or more after diagnosis of breast cancer, cardiovascular disease exceeded breast cancer as the leading cause of death at 10 years after diagnosis, when the cumulative incidence of each was 5%.
Conclusions and Relevance
Cardiovascular death is an important competing risk for older women with early-stage breast cancer. This finding mandates adequate attention to cardiovascular preventive therapy after diagnosis of breast cancer.
Disease-specific survival for women with early-stage breast cancer (EBC) has substantially improved.1 Breast cancer shares risk factors with cardiovascular disease,2 and its treatments have adverse cardiovascular effects.3 Most data on cardiovascular mortality after EBC are dated or based on studies with small sample sizes (eAppendix 1 in the Supplement). Competing risk methods provide the appropriate framework to analyze the interplay between EBC and cardiovascular disease. Failure to appropriately account for competing risks can lead to biased conclusions, with incidence estimates that are biased upwards.4
In this study, we used a population-based cohort of women diagnosed with EBC to assess the balance between the competing risks of death from breast cancer and from cardiovascular disease. We focused on the effect of age, medical history, and time since diagnosis on the likelihood of death from different causes.
The Ontario Health Insurance Plan provides universal health insurance coverage for the province’s residents. Patients’ data were linked across multiple administrative databases (eAppendix 2 in the Supplement) and analyzed at the Institute for Clinical Evaluative Sciences, which is a prescribed entity for the purposes of s. 45(1) of Ontario’s Personal Health Information Protection Act, and as such, research ethics board approval was not legally required. Nevertheless, this study underwent retrospective review and approval by the Research Ethics Board of Sunnybrook Health Sciences Centre. Patient consent was not required.
We identified all adult women diagnosed with EBC between April 1, 1998, and March 31, 2012, using the Ontario Cancer Registry. The index date was that of pathologic diagnosis of EBC. Exclusion criteria included ineligibility for the Ontario Health Insurance Plan, missing key data (age, sex, or diagnosis date), age older than 105 years, and advanced disease. We also created an age-matched cohort of women with no record in the Ontario Cancer Registry as of the index date with similar exclusion criteria. Patients were followed up until death or were censored on December 31, 2013. The outcome was death, with its cause categorized into breast cancer, cardiovascular disease, other cancers, or other noncancer causes. For women 66 years or older, the fourth category of cause of death was subdivided into dementia or cerebral degeneration; sepsis, pneumonia, and urinary tract infections; diabetes; and other. The list of International Classification of Diseases, Ninth Revision codes used to define outcomes is provided in eTable 1 in the Supplement.
Statistical analysis was performed from July 16, 2015, to August 4, 2016. Age was summarized using the median (with interquartile range [IQR]), while counts (with percentages) were used for other variables. We generated cumulative incidence function curves to describe the incidence over time of death from cardiovascular disease, breast cancer, other cancers, and other noncancer causes. For the EBC cohort, we conducted stratified analyses based on the following: (1) age (<66 vs ≥66 years), (2) prior cardiovascular disease (myocardial infarction, congestive heart failure, arterial revascularization, valvular disease, peripheral vascular disease, or cerebrovascular disease), and (3) age 66 years or older or presence of cardiovascular disease, diabetes, or hypertension. We also performed a landmark analysis of women who survived 5 years or more after diagnosis.
We used cause-specific hazard models to estimate the association between baseline characteristics and the rate of cardiovascular death.4 Age was used as a categorical variable with either 2 (<66 vs ≥66 years) or 4 (≤55, 56-65, 66-75, or >75 years) levels. The regression models incorporated diagnosis year as a stratification variable. All analyses were performed using SAS, version 9.4 (SAS Institute Inc). Statistical significance was defined by a 2-tailed P < .05.
We identified 98 999 eligible women with EBC. As shown in eTable 2 in the Supplement, the median age was 60 years (IQR, 50-71 years), and 37 425 women (37.8%) were 66 years or older at diagnosis. A total of 6559 women (6.6%) had prior cardiovascular disease; the median age of these women was 77 years (IQR, 68-84 years). A total of 42 724 women (43.2%) were younger than 66 years and had no prior cardiovascular disease, diabetes, or hypertension. Among 54 587 (55.1%) women with data on cancer laterality, 27 673 (50.7%) had left-sided disease. A total of 44 675 women (45.1%) received chemotherapy, 45 592 (46.1%) received radiotherapy, and 7851 (7.9%) received trastuzumab therapy. Data on endocrine therapy were available uniformly for women 66 years or older, of whom 11 094 (29.6%) received aromatase inhibitors and 13 331 (35.6%) received tamoxifen.
Median follow-up was 6.6 years (IQR, 3.6-10.4 years), with 26 752 women (27%) having more than 10 years’ follow-up. A total of 21 123 women (21.3%) died during follow-up. The median time to death was 4.2 years (IQR, 2.2-7.1 years). Breast cancer was the most common cause of death (10 550 [10.7% of cohort, 49.9% of deaths]). Cardiovascular disease was the cause of death in 3444 women (3.5% of cohort, 16.3% of deaths). Women 66 years or older at diagnosis accounted for 89.6% of cardiovascular deaths (3087 of 3444) and 43.9% of breast cancer deaths (4629 of 10 550).
The cumulative incidence of death from cardiovascular disease, breast cancer, and other causes in the EBC cohort and age-matched, cancer-free women is illustrated in eFigure 1 in the Supplement. The risk of death owing to breast cancer was negligible in the cancer-free cohort of age-matched controls but constituted the leading cause of death in the EBC cohort. The cumulative incidence of death from causes other than breast cancer was lower in the EBC cohort. In particular, the 10-year cumulative incidence of cardiovascular death in the EBC cohort was 3.2% compared with 4.4% in the cancer-free cohort.
The cumulative incidence of cause-specific death among younger and older women with EBC is illustrated in Figure 1. Figure 2 demonstrates these data among patients with and without prior cardiovascular disease. The cumulative incidence of death owing to cardiovascular disease and breast cancer in analyzed strata is also listed in the Table. Among patients with prior cardiovascular disease, the risks of death from breast cancer (12.4%) and cardiovascular disease (12.6%) were similar for the first 5 years, but cardiovascular death assumed greater prominence beyond that time point. In contrast, the 10-year cumulative incidence of cardiovascular death was negligible in women younger than 66 years without prior cardiovascular disease, diabetes, or hypertension (0.5%).
Quantifying these observations, the regression analysis (eTable 3 in the Supplement) highlights that age had the largest independent effect on cardiovascular death. Women 66 years or older at diagnosis had a greater than 10-fold increase in the cause-specific hazard for cardiovascular death compared with younger women (hazard ratio, 10.42; 95% CI, 9.29-11.70). With age categorized into 4 groups, (≤55, 56-65, 66-75, and >75 years), women older than 75 years had 42 times the risk of cardiovascular death relative to those 55 years or younger.
eFigure 2 in the Supplement illustrates the cumulative incidence of cause-specific death in women who survived 5 years or more after diagnosis of EBC. Breast cancer remained the most frequent cause of death among younger women. However, among women 66 years or older, cardiovascular disease passed breast cancer as the most common cause of death at 10 years after diagnosis, when the cumulative incidence of each was 5% (with follow-up beginning at year 5).
Cardiovascular disease was the leading competing risk within our population-based cohort of 98 999 women with EBC. It was the leading cause of death among women with established cardiovascular disease and among those 66 years or older at diagnosis who had survived 5 years or more after diagnosis of breast cancer. Conversely, we observed that younger women without prior cardiovascular disease, hypertension, or diabetes are at extremely low risk for cardiovascular death in the first 10 years after breast cancer. The general patterns were similar to those in age-matched women without cancer, although breast cancer deaths acted as a competing risk that lowered the absolute risks of alternative causes of death.
Our results add to prior research, summarized in eAppendix 1 in the Supplement. These studies mostly involved patients diagnosed before the turn of the century. Since then, the management and outcomes associated with breast cancer1,5-7 and cardiovascular disease8 have changed substantially. These changes limit the applicability of these studies to today’s patients. Our data suggest that the amount of time for cardiovascular disease to overtake breast cancer as the leading cause of death has increased compared with that in older studies. This finding may reflect population-wide trends, in which cardiovascular death rates have declined faster than those owing to cancer during the past 20 years.9
Although management of breast cancer should be the primary focus of women with EBC, our data highlight the importance of competing risks. Women with EBC are just as likely to die from causes other than breast cancer, chief among which is cardiovascular disease. Given the effectiveness of both primary and secondary prevention interventions for cardiovascular disease, appropriate care for women with EBC should involve counseling and preventive interventions that are proportionate to their expected cardiovascular risk. Our results highlight that age and previous history of cardiovascular disease are strong predictors of cardiovascular risk in women with EBC. These traditional risk factors can serve as an inexpensive initial screening tool to identify women who could most benefit from preventive interventions. They can also inform decisions about adjuvant endocrine therapy beyond 5 years. These agents have adverse cardiovascular effects10 and may not be of benefit among groups in which breast cancer carries a smaller effect on overall survival. Our research also highlights the importance of accounting for competing risks, particularly in the development of risk assessment tools. Moreover, our demonstration of the importance of cardiovascular disease as a competing risk among patients with cancer supports the growth of the new field of cardio-oncology.
A key limitation of our study is that we do not provide information on other risk factors, such as dyslipidemia and smoking. We relied on death certificate data, a source that has been used in other studies of cause-specific mortality8,11-13 but that may not always be completely accurate. However, a study of Ontario death certificate data reported a sensitivity of 95% and specificity of 88% in identifying death owing to breast cancer.14 We were also unable to detect death in women who left the province.
Cardiovascular death is a clinically important competing risk in women with EBC, especially among women with prior cardiovascular disease, and older women who survive 5 years beyond their cancer diagnosis date. Appropriate risk stratification can identify women with EBC who could benefit from cardiovascular disease prevention strategies. Further research on cardiovascular disease in women with breast cancer is warranted.
Corresponding Author: Geoffrey M. Anderson, MD, PhD, Institute of Health Policy, Management and Evaluation, University of Toronto, Health Sciences Building, 155 College St, Fourth Floor, Ste 425, Toronto, ON M5T 3M6, Canada (email@example.com).
Accepted for Publication: August 19, 2016.
Correction: This article was corrected on May 17, 2017, to fix a 95% CI in the Table.
Published Online: October 12, 2016. doi:10.1001/jamacardio.2016.3841
Author Contributions: Dr Abdel-Qadir had full access to all the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.
Study concept and design: Abdel-Qadir, Lee, Anderson.
Acquisition, analysis, or interpretation of data: All authors.
Drafting of the manuscript: Abdel-Qadir, Amir, Anderson.
Critical revision of the manuscript for important intellectual content: Abdel-Qadir, Austin, Lee, Amir, Tu, Thavendiranathan, Fung.
Statistical analysis: Abdel-Qadir, Fung.
Obtained funding: Anderson.
Administrative, technical, or material support: Abdel-Qadir, Lee, Thavendiranathan, Fung.
Study supervision: Lee, Amir, Tu, Thavendiranathan, Anderson.
Conflict of Interest Disclosures: All authors have completed and submitted the ICMJE Form for Disclosure of Potential Conflicts of Interest. Dr Abdel-Qadir reported being supported by a fellowship from the Canadian Institutes of Health Research. Dr Lee reported being supported by a mid-career research award from the Heart and Stroke Foundation and the Ted Rogers Chair in Heart Function Outcomes. Dr Austin reported being supported in part by a Career Investigator Award from the Heart and Stroke Foundation. Dr Tu reported being supported by a Tier 1 Canada Research Chair in Health Services Research and an Eaton Scholar award. Dr Thavendiranathan reported being supported by the New Investigator Award from the Canadian Institutes of Health Research. No other conflicts were reported.
Funding/Support: Funding for part of this study was provided by the Heart and Stroke Foundation of Canada. This study was supported by the Institute for Clinical Evaluative Sciences, which is funded by an annual grant from the Ontario Ministry of Health and Long-Term Care. This study was conducted with the support of the Ontario Institute for Cancer Research and Cancer Care Ontario through funding provided by the Government of Ontario.
Role of the Funder/Sponsor: The funding sources had no role in the design and conduct of the study; collection, management, analysis, and interpretation of the data; preparation, review, or approval of the manuscript; and decision to submit the manuscript for publication.
Disclaimer: The opinions, results, and conclusions reported in this article are those of the authors and are independent from the funding sources. No endorsement by the Institute for Clinical Evaluative Sciences, the Ontario Ministry of Health and Long-Term Care, or Cancer Care Ontario is intended or should be inferred. Parts of this article are based on data and information compiled and provided by the Canadian Institute for Health Information and by Cancer Care Ontario. However, the analyses, conclusions, opinions, and statements expressed herein are those of the authors and not necessarily those of the Canadian Institute for Health Information or Cancer Care Ontario. Parts of this material are based on data and information provided by Cancer Care Ontario. The opinions, results, views, and conclusions reported in this article are those of the authors and do not necessarily reflect those of Cancer Care Ontario. No endorsement by Cancer Care Ontario is intended or should be inferred.
Additional Contributions: The clinical registry data used in this publication are from participating hospitals through the Cardiac Care Network of Ontario, which serves as an advisory body to the Ministry of Health and Long-Term Care, is funded by the Ministry of Health and Long-Term Care, and is dedicated to improving the quality, efficiency, access and equity in the delivery of the continuum of adult cardiac services in Ontario, Canada.
Create a personal account or sign in to: