**4. Discussion**

In this population-based study of men diagnosed with breast cancer in the United States over a 20-year period, treatment with chemotherapy was associated with elevated risk of CVD mortality, while no significant association was observed between radiation therapy and deaths due to CVD. Racial and ethnic disparities in the association of chemotherapy and CVD mortality were observed, with Hispanic men having higher risk of CVD deaths compared to non-Hispanic Black and non-Hispanic White men. To our knowledge, this is the first study to comprehensively characterize CVD mortality due to cancer treatment among men diagnosed with breast cancer.

Some breast cancer therapeutics have been reported to result in early or delayed cardiotoxicity comprising of hypertension, arrhythmias, pericarditis, thromboembolism, valvular disease, left ventricular dysfunction, heart failure, and myocardial infarction [14,15]. Accordingly, it has been estimated that the cumulative incidence of treatment-related cardiotoxic outcomes among breast cancer patients may be as high as 33% [25]. There are limited prospective investigations of the relation of neoadjuvant or adjuvant chemotherapy on CVD morality in MBC patients. Results from the current study of elevated risk of CVD mortality among MBC patients who received chemotherapy is supported by several pieces of evidence of the cardiotoxic effects of chemotherapy in murine models and studies conducted among FBC patients [14,26–30]. The most widely reported cardiotoxic effect of chemotherapy is left ventricular dysfunction that manifests as overt heart failure over time [14,25], although other cardiac events such as thrombosis, arrhythmias, myocarditis, pericarditis, and myocardial infarction have also been reported [31]. For instance, Yang et al. [32] reported a 74% elevated risk of heart failure among breast cancer patients who received chemotherapy. Conversely, as seen in some studies among women [33], a few studies conducted mostly among small samples of men with breast cancer have also reported lower mortality in men who received adjuvant chemotherapy [34–36]. However, these studies did not specifically evaluate cardiovascular-related mortality.

There are several mechanisms by which chemotherapy may influence cardiovascular health in breast cancer patients. Anthracyclines, such as doxorubicin interacts with deoxyribonucleic acids, intercalating and inhibiting macromolecular biosynthesis of cardiac myocytes that eventually leads to apoptosis of myocytes and permanent damage to the myocardium [14,31]. Additionally, chemotherapeutics fosters the generation of reactive oxygen species which damage deoxyribonucleic acids, proteins, and mitochondrial membrane of myocytes [14,31]. In light of this, finding avenues to reduce the risk of CVD events among MBC patients is of grea<sup>t</sup> importance. With adjuvant chemotherapy not improving overall or breast cancer-specific survival among MBC patients with stage I and IIA cancer, the risk of CVD mortality may be reduced in this population by perhaps skipping chemotherapy for MBC patients with early-stage disease [37]. In addition, more consideration may be given to administering adjuvant trastuzumab which often, but not always, results in reversible LV dysfunction together with chemotherapy for patients with HER2-positive early-stage breast due to the reported marked improvement in survival and reoccurrence of cancer with this treatment regimen [14,38]. Finally, the risk-benefit profile of each MBC patient should be taken into consideration when choosing chemotherapy especially for those who are at high risk for CVD [14]. For those who have a risk-benefit profile in favor of chemotherapy, early detection and interception of cardiotoxicity remains important for clinicians.

Emerging evidence suggests that there are declining CVD mortality trends by radiation therapy among breast cancer patients [16,39]. Vo et al. [16] evaluating trends in heart disease mortality in the United States among women with invasive breast cancer from 1975 to 2017 observed significant declines in heart disease mortality for breast cancer survivors treated with radiotherapy alone compared to the general population, while an increasing trend in heart disease mortality was seen for regional stage patients treated with chemotherapy alone. From 1975–1984 to 2005–2016, the 10-year cumulative heart disease mortality declined from 6.35% to 2.94% among breast cancer survivors treated with radiotherapy alone while the 10-year cumulative heart disease mortality reduced from 1.78% to 1.21% [16]. Similarly, Hooning et al. [39] studying 7425 patients in the Netherlands treated for early breast cancer from 1970 to 1986 and followed through to 2000 found

no increased CVD mortality for post-lumpectomy radiation, with the risk estimates for CVD mortality highest for post-lumpectomy radiation administered before 1979. Studies conducted in the modern era of breast cancer therapy have largely found no association between radiation therapy and CVD outcomes [31,39]. Similar to the current study where no association between radiotherapy (with or without chemotherapy) and CVD mortality, regardless of tumor laterality, was observed among MBC patients, Onwudiwe et al. using data from women aged 66 years and older with stage 0–III breast cancer diagnosed between 2000 and 2005 in the SEER-Medicare database also observed no association between radiation therapy and combined endpoints of death or cardiovascular disease [40]. Another register-based matched cohort study of Swedish breast cancer patients diagnosed from 2001 to 2008 and followed up until 2017 also observed no elevated risk of heart disease following locoregional radiotherapy [32].

The lack of a positive association of radiotherapy with CVD mortality observed in the current study as well as other studies of cancer therapy administered in the 21st century reflects the impact of changes in radiotherapy procedures [5,14]. However, it should be noted that radiation-associated cardiotoxicity often appears about 10 to 30 years after treatment and most studies including the current study did not have any individuals with follow-up beyond 20 years [41]. Increasing clinical guidelines about the adverse cardiac effects of radiation therapy has advanced cardio-protection strategies to minimize radiation-related damage to the cardiovascular system [14]. For example, reduction in radiation doses to the left side of the chest during radiotherapy, positioning patients to displace the heart during radiotherapy administration, the use of more precise radiotherapy using imaging and brachytherapy, and alternative radiotherapy options have all gone a long way to reduce the effects of radiation therapy on cardiac damage during cancer treatment [16,32,42–45]. Alternatively, the null association between radiotherapy and CVD mortality in MBC patients may be due to patients with left-sided breast cancer being less likely to be selected for radiotherapy due to the proximity of the tumor to the heart [32,45]. Future studies evaluating dosages of radiation to the heart and CVD mortality will enhance our understanding of a safe threshold of radiation that enhances cancer treatment response and at the same time reduce the risk for CVD outcomes in breast cancer patients.

Another interesting observation from the current study is the racial and ethnic disparities in the relation of chemotherapy with CVD mortality in MBC patients. The risk of CVD mortality in Hispanic men was more than twice the risk in non-Hispanic White men with no association observed between chemotherapy and CVD mortality among non-Hispanic Black men with breast cancer. While reasons for these findings are largely unknown, it is possible that differences in sociodemographic, socioeconomic, behavioral, and biological factors as well as differences in access to cancer treatment may partly explain these findings. For instance, compared to non-Hispanic White individuals, Hispanic populations are less likely to partake in mammography screening and adhere to cancer screening recommendations [46–48]. Thus, they often experience longer times to diagnosis of cancer resulting in them being likely to be diagnosed with advanced staged cancer [46]. Furthermore, they often experience poor quality of life following diagnosis of cancer than non-Hispanic White individuals [46]. Due to language barriers among low-acculturated Hispanic individuals, they often receive limited communication about cancer diagnosis and treatment which hinders the decision-making processes concerning cancer treatments [46,49]. A few studies among men [50] and several studies among women with breast cancer consistently report longer delays in receipt of chemotherapy among Hispanic and non-Hispanic Black individuals [51–53]. Taken together, it is possible that all these factors may contribute to the high risk of CVD mortality due to chemotherapy among Hispanic population.

With non-Hispanic Black individuals also experiencing delays in chemotherapy treatment [54,55] despite rates of oncologic consultation being similar between Black and White cancer patients [56], it would have been expected that this population would also experience high CVD risk due to chemotherapy. However, this was not the case in the current study. We speculate that the greater proportion of early discontinuation of chemotherapy of

non-Hispanic Black patients mostly due to negative beliefs about efficacy of chemotherapy often due to concern about adverse effects [55,57,58], coupled with Black patients having lower pathologic complete response to neoadjuvant chemotherapy than Hispanic and other racial groups [59] may result in them having reduced exposure to the cardiotoxic effect of chemotherapeutics. With delays and interruptions in breast cancer treatment being positively related to breast cancer-specific mortality [60], this explanation is further supported by the observation that Black MBC patients have greater breast cancer-related mortality than CVD mortality compared to MBC patients of other racial and ethnic groups [12,50,61].

Currently, most treatment options for breast cancer in men are based on evidence from trials among women with breast cancer [62]. Although some reports show that treatment options in men produce comparable results to FBC patients [63], overall survival in MBC patients is lower than those for FBC patients [64] with some studies reporting excess mortality rates of about 60% in men when compared to women [10]. The lack of evidence-based treatment recommendations and screening guidelines for breast cancer in men, coupled with limited reports on treatment-associated complications continue to impact treatment choices and care for men with breast cancer [62]. Some studies report that screening mammography yields similar cancer detection rates between men and women at high risk for breast cancer [65]. Therefore, interventions focusing on increasing awareness and promoting breast cancer education in men, together with enhancing access to care among high-risk groups regardless of race and ethnicity will go a long way to increase earlystage cancer diagnosis and reduce racial and ethnic disparities in survival outcomes [66]. Furthermore, the few clinical trials among male breast cancer patients [18,67,68] currently underway will provide comprehensive data on the long-term managemen<sup>t</sup> of MBC to inform treatment recommendations and guidelines on regimens that optimize cancer therapy and at the same time limit the risk of CVD [15].

The strength of the current study includes the use of a large population-based sample of MBC patients selected within a modern timeframe of cancer treatment. Limitations of the study include the lack of detailed information on specific drugs or hormone therapy not being available in the SEER registry for most of the period of observation for this study. HER2 positivity status was not evaluated in the current study as such information was only available after 2010. Furthermore, information on CVD risk factors at the time of cancer diagnosis as well as information on other comorbid noncancer diseases were not collected by SEER program. Finally, the chance of misclassification bias influencing the results of the study due to the use of death certificates to identify deaths attributable to CVD cannot be entirely ruled out. However, cause-of-death information in the SEER registry have been reported to have good validity [69].
