Adjuvant Breast Radiotherapy Improve Survival in Women With Heart Failure

BACKGROUND To date, no data on the effect of adjuvant whole breast radiotherapy (WBRT) on oncologic outcomes, such as all-cause death, locoregional recurrence (LRR), and distant metastasis (DM), are available in women with left-side breast intraductal carcinoma (IDC) and heart failure with reduced ejection fraction (HFrEF). Abstract This study is the rst to examine the effect of adjuvant whole breast radiotherapy (WBRT) on oncologic outcomes such as all-cause death, breast cancer death, locoregional recurrence (LRR), and distant metastasis (DM) in women with left-side breast intraductal carcinoma (IDC) and heart failure with reduced ejection fraction (HFrEF) receiving breast-conserving surgery (BCS). In the IPTW-adjusted models, adjuvant WBRT was associated with a decrease in all-cause death, breast cancer death, LRR, and DM in women with left IDC and HFrEF compared with no adjuvant WBRT. We suggest adjuvant WBRT for women with left-side IDC receiving BCS, even when they have HFrEF. treated clinical target volumes, smoking habits, hypertension, ischemic heart disease, heart valvular disease, cardiomyopathy, arrhythmias and conduction disorders, diabetes, adjuvant chemotherapy with anthracycline-based regimen, hormone receptor status, trastuzumab use, nodal surgery, hospital level (academic or nonacademic), and income, were analyzed using a multivariate analysis of the propensity score–weighted population to determine hazard ratios (HRs). of treatment cohort; weighted cohort no predictors in multivariable IPTW were stratied in the The endpoint all-cause death the left-side IDC and HFrEF BCS by adjuvant WBRT (Group 1, case group) in the women with left IDC and HFrEF who BCS had no adjuvant WBRT 2, control group). the overall survival cancer locoregional recurrence (LRR)–free survival, distant metastasis (DM)–free survival left IDC and HFrEF adjuvant no adjuvant log-rank all-cause mortality IDC adjuvant no adjuvant for age, diagnosis year, CCI pN, heart-sparing RT techniques, treated clinical target volumes, smoking habits, hypertension, ischemic heart disease, heart valvular disease, cardiomyopathy, arrhythmias and conduction disorders, diabetes, adjuvant chemotherapy with anthracycline, hormone receptor status, trastuzumab use, nodal surgery, hospital levels, and incomes.


Inclusion and exclusion criteria
The diagnoses of the enrolled patients with HFrEF were con rmed after their pathological data were reviewed, and the women with newly diagnosed left-side IDC were con rmed to have no other cancers or distant metastases. The women with HFrEF were included if they had received a left-side IDC diagnosis, were 20 years old or older, and had clinical stage IA-IIIC (American Joint Committee on Cancer [AJCC], 8th edition) without metastasis. Patients with HFrEF were excluded if they had a history of cancer before the IDC diagnosis date, unknown pathologic types, missing sex data, unclear staging, or non-IDC histology. In addition, patients undergoing neoadjuvant chemotherapy or with unclear differentiation of tumor grade, missing HR status, missing data on trastuzumab or anthracycline use, or unclear staging were excluded. Other adjuvant treatments such as adjuvant chemotherapy, hormone therapy, or the human epidermal growth factor receptor 2 inhibitors did not constitute exclusion criteria based on the National Comprehensive Cancer Network (NCCN) guidelines [20]. We also excluded patients with HFrEF with unclear data on surgical procedures such as BCS or TM, ill-de ned nodal surgery, unclear Charlson comorbidity index (CCI)., or unclear differentiation from our cohort. Hormone receptor positivity was de ned as ≥1% of tumor cells demonstrating positive nuclear staining through immunohistochemistry [21].
After applying the inclusion and exclusion criteria, we enrolled 294 women with HFrEF and AJCC clinical stage IA-IIIC and left-side IDC who had received a BCS followed by sentinel lymph node biopsy (SLNB) or axillary lymph node dissection (ALND) and divided them into two groups based on their adjuvant WBRT status to compare all-cause mortality: Group 1 (women with left-side IDC and HFrEF who received BCS followed by adjuvant WBRT) and Group 2 (women with left-side IDC and HFrEF who received BCS and no adjuvant WBRT). We also excluded women in Group 1 receiving nonstandard adjuvant WBRT (contrast with standard adjuvant radiotherapy consisting of irradiation to the whole breast with a minimum of 50 Gy). Contemporary RT techniques were included in our study and the conventional two-dimensional or three-dimensional RT technique was excluded. The included contemporary RT techniques were intensity-modulated radiation therapy (IMRT) or volumetric modulated arc therapy (VMAT) combined with heart-sparing RT techniques using deep inspiration breath-hold (DIBH) and/or gating [22]. The incidence of comorbidities was scored using the CCI [23]. Ischemic heart disease, heart valvular disease, cardiomyopathy, hypertension, diabetes, and arrhythmias and conduction disorders were excluded from the CCI scores to avoid repetitive adjustment in multivariate analysis. Only comorbidities observed within 6 months before the index date were included; they were coded and classi ed according to the International Classi cation of Diseases, 10th Revision, Clinical Modi cation (ICD-10-CM) codes at the rst admission or based on more than two repetitions of a code issued at outpatient department visits.

Study covariates and statistical analysis
Signi cant independent predictors, namely age, diagnosis year, CCI score, differentiation, pT, pN, heart-sparing RT techniques, treated clinical target volumes, smoking habits, hypertension, ischemic heart disease, heart valvular disease, cardiomyopathy, arrhythmias and conduction disorders, diabetes, adjuvant chemotherapy with anthracycline-based regimen, hormone receptor status, trastuzumab use, nodal surgery, hospital level (academic or nonacademic), and income, were analyzed using a multivariate analysis of the propensity score-weighted population to determine hazard ratios (HRs). We calculated the propensity score and applied inverse probability of treatment weighting (IPTW) to create a pseudo-study cohort; the weighted cohort avoids covariate bias and mimics randomized adjuvant WBRT or no adjuvant WBRT assignment: IPTW for patients with WBRT = 1/p(WBRT); IPTW for patients without WBRT = 1/(1 − p[WBRT]) [24]. The independent predictors were examined in multivariable analyses after IPTW adjustment. Moreover, they were controlled for and were strati ed in the analysis. The endpoint was all-cause death in the women with left-side IDC and HFrEF who received BCS followed by adjuvant WBRT (Group 1, case group) and in the women with left IDC and HFrEF who received BCS and had no adjuvant WBRT (Group 2, control group).
The cumulative incidence of death was estimated using the Kaplan-Meier method, and differences in the overall survival (OS), breast cancer death, locoregional recurrence (LRR)-free survival, and distant metastasis (DM)-free survival between women with left IDC and HFrEF receiving BCS followed by adjuvant WBRT versus no adjuvant WBRT were determined using a log-rank test. After confounders were adjusted for, IPTW-adjusted models were used to determine the time from the index date to all-cause mortality in the women with left IDC and HFrEF who received BCS followed by adjuvant WBRT or no adjuvant WBRT. Subsequently, in a multivariate analysis, HRs were adjusted for age, diagnosis year, CCI scores, differentiation, pT, pN, heart-sparing RT techniques, treated clinical target volumes, smoking habits, hypertension, ischemic heart disease, heart valvular disease, cardiomyopathy, arrhythmias and conduction disorders, diabetes, adjuvant chemotherapy with anthracycline, hormone receptor status, trastuzumab use, nodal surgery, hospital levels, and incomes. All analyses were conducted using SAS (Version 9.3; SAS, Cary, NC, USA), and a two-tailed P value <0.05 was considered statistically signi cant.

Study cohort
We enrolled 294 women with left-breast IDC at clinical stages IA-IIIC and HFrEF who received BCS followed by adjuvant WBRT or no adjuvant WBRT (Table 1). Among these women, 223 with left IDC and HFrEF received BCS followed by adjuvant WBRT (Group 1) and 71 with left IDC and HFrEF received BCS with no adjuvant WBRT ( Group 2). After IPTW was executed using the propensity score, the covariates between Groups 1 and 2 were found to be homogenous. The median follow-up durations after the index date were 6.96 and 5.09 years for women with left IDC and HFrEF who received BCS followed by adjuvant WBRT or no adjuvant WBRT, respectively. All standardized differences in covariates were smaller than 0.1 (Table 1) and were homogenous between the two groups.
Effects of adjuvant WBRT on oncologic outcomes in women with left-side IDC and HFrEF receiving BCS IPTW-adjusted models indicated that adjuvant WBRT was a signi cantly better independent prognostic factor for OS, breast cancer death, LRR, and DM in the women with left IDC and HFrEF receiving BCS ( Other independent predictors of all-cause death, breast cancer death, LRR, and DM in the women with left IDC and HFrEF receiving BCS Old age (>65 years), CCI ≥ 1, advanced pT stages (pT2-4), advanced pN stages (pN1-3), treated clinical target volumes including breast, axillary lymph nodes and internal mammary chain, hormone receptor negative status, and differentiation Grade II and III were identi ed as crucial independent poor prognostic factors for OS (Tables 2). IPTW-adjusted models were adjusted for age, diagnosis year, CCI score, differentiation, pT, pN, heart-sparing RT techniques, treated clinical target volumes, smoking habits, hypertension, ischemic heart disease, heart valvular disease, cardiomyopathy, arrhythmias and conduction disorders, diabetes, adjuvant chemotherapy with anthracycline-based regimen, hormone receptor status, trastuzumab use, nodal surgery, hospital level, and income; the aHRs (95% CIs) of all-cause death for age > 65 years, CCI ≥ 1, pT2, pT3, pT4, pN1, pN2, PN3, heart-sparing technique, treated clinical target volumes including breast, axillary lymph nodes and internal mammary chain, differentiation Grades II and III, and hormone receptor positive status were 1. 31  Survival curves of adjuvant WBRT or no adjuvant WBRT in women with left IDC and HFrEF receiving BCS Figure 1 presents Kaplan-Meier curves that illustrate the primary endpoint of OS of the women with left IDC and HFrEF receiving BCS with adjuvant WBRT or no adjuvant WBRT. The 5-year overall survival rates were 86.47% and 75.92% in the adjuvant WBRT and no adjuvant WBRT groups, respectively ( Figure 1); the OS rate was associated with an increasing trend in the adjuvant WBRT group (log-rank test, P = 0.0618) compared with the non-WBRT group. Additionally, the 5-year LRR-free survival in women with left IDC and HFrEF receiving BCS was 95.78% and 86.11% in the adjuvant WBRT group and no adjuvant WBRT group, respectively ( Figure 2; log-rank test, P = 0.0083). The 5-year DM-free survival in women with left IDC and HFrEF receiving BCS was 96.23% and 78.33% in the adjuvant WBRT group and no adjuvant WBRT group, respectively ( Figure 3; log-rank test, P = 0.0027).

Discussion
The use of RT has contributed to signi cant improvements in disease-speci c survival among patients with early stage breast cancer [25]. The success of RT, used either alone or in combination with other modalities, has resulted in large cohorts of breast cancer survivors who are vulnerable to late complications such as RICT from RT [10,[26][27][28][29][30][31]. Awareness regarding the potential cardiotoxicity of RT has led to the application of improved RT techniques that minimize irradiation to the heart [27,28]. In patients receiving RT for left-sided breast IDC, reducing the heart from the primary radiation beams is the preferred strategy, and DIBH is a particularly useful approach in this setting [22,[32][33][34][35]. Systematic cardiac blocking without DIBH often results in underdosage of portions of the breast, internal mammary nodes and thus may not be prudent. With DIBH, the heart is away from the left breast, and thus when the beams are shaped to exclude the heart, the amount of potential targets that are underdosed is reduced. Thus, DIBH can improve the therapeutic ratio of RT, and enable cardiac sparing with lesser impacts on target coverage [22,[32][33][34][35]. In our study, over 90% patients receiving adjuvant WBRT combined with DIBH techniques might be contributed to better survival bene ts for breast cancer women with heart failure receiving adjuvant WBRT compared with those have no adjuvant WBRT (Table 1 and 2). Because dose constrains, RT techniques for cardiac protection and data on irradiation of the lymphatic drainage pathways were important factors of RICT [36] [22,[32][33][34][35], we added these covariates like heart-sparing RT techniques and treated clinical target volumes (breast only, breast and axillary lymph nodes, or breast, axillary and internal mammary chain) in Table 1. In our study, heart-sparing RT technique was associated with reduction of all-cause death and extensive RT eld (including breast, axillary lymph nodes, and internal mammary chain) was associated with an increase if all-cause death ( Table 2). The potential reasons of survival bene ts in heart-sparing RT techniques and disadvantages of extensive RT might be contributed the higher heart irradiation dose-volume with more cardiac death in the women with HFrEF.
Anthracycline is among the chemotherapeutic agents implicated in cardiotoxicity, and irradiation combined with anthracycline-based chemotherapy was associated with substantially increased cardiotoxicity [37]. Thus, in a patient with normal baseline cardiac function, the bene ts of anthracycline likely outweigh the risk, especially in patients higher risk breast cancers [38,39]. In addition, trastuzumab-related cardiotoxicity is often manifested by an asymptomatic decrease in LVEF [40][41][42]. Thus, numerous treatment-related factors are responsible for cardiotoxicity in women with breast cancer [6, 38-46]. Therefore, decision-making regarding the use of adjuvant WBRT for women with left-side IDC and HFrEF receiving BCS is often a concern for physicians and patients irrespective of whether chemotherapy with antracycline-based regimens or trastuzumab is conducted. Thus, we conducted the study to determine the survival bene ts offered by adjuvant WBRT in women with left-side IDC and HFrEF receiving BCS.
The recent decline in mortality in women with HF have been improving a lot [47,48]. Therefore, adjuvant WBRT will be more worthy for the women with breast cancer and HF. Long-term mortality rates in patients with HF have improved over time [47,48]. A decline in HF mortality was noted in an analysis by the Mayo Clinic that included 4537 patients, with the majority hospitalized around the time of diagnosis [47]. The two main causes of death in patients with HF are sudden death and progressive pump failure [49,50]. However, patients with breast cancer might experience adverse effects from many cardiotoxic treatments such as adjuvant RT, anthracycline-based chemotherapy, or trastuzumab 6, [6, 10, 26-31, 38-46]. Although cardiovascular diseases such as HF, heart attacks, and stroke remain the leading cause of death in women, many believe breast cancer to be more deadly [51]. In fact, the risk of RICT should be weighed against the potential bene ts of adjuvant WBRT with respect to the patients' prognosis and likely clinical bene t [10,[26][27][28][29][30][31]. Until now, no data have been available for the evaluation of oncologic outcomes of adjuvant WBRT in women with left-side breast IDC and HFrEF receiving BCS. This is the rst study to explore the value of adjuvant WBRT for women with left-side breast IDC and HFrEF receiving BCS. As shown in Table 2, adjuvant WBRT resulted in better OS, LRR-free status, and DM-free status compared with no adjuvant WBRT in women with left-side breast IDC and HFrEF receiving BCS. The potential reasons might be the recent decline in mortality in women with HF [47,48] and the advances in contemporary RT techniques with reduced irradiation volumes to the heart [2,22,27,28].
The propensity score is de ned as a subject's probability of adjuvant WBRT selection, conditional on the observed baseline covariates [24]. Weighting subjects through IPTW creates a synthetic sample in which treatment assignment is independent of the measured baseline covariates [24]. IPTW with the propensity score helped us obtain unbiased estimates of average treatment effects such as OS, breast cancer death, LRR, and DM (Tables 1-3). The use of IPTW has increased rapidly in recent years, but a majority of recent studies have not formally examined whether the weighting balanced the measured covariates between treatment groups [24]. In addition, extreme weights at the tails of the propensity score distribution increase the variance and decrease the balance between covariates [52]. Thus, further trim patients with extreme scores (i.e., remove them from the weighted analysis in our study). and multivariate analysis of propensity score-weighted population could be still necessary [52]. Finally, the standardized differences for all covariates in our study were smaller than 0.1, which means the weighted covariates were balanced between the adjuvant WBRT and no adjuvant WBRT groups (Table 1).
Moreover, the factors affecting the mortality due to HF such as age, sex (all enrolled patients were women with breast IDC), race (all enrolled patients were Asian), smoking habits, ischemic heart disease, heart valvular disease, cardiomyopathy, hypertension, diabetes, and arrhythmias and conduction disorders were weighted through IPTW by using the propensity score (Table 1). In addition, all risk factors for all-cause death for patients with breast cancer receiving BCS, such as age, diagnosis year, CCI score, differentiation, pT, pN, heart-sparing RT techniques, treated clinical target volumes, smoking habits, hormone receptor status, nodal surgery, hospital level, and income, were also weighted using IPTW using the propensity score (Table 1).. Preexisting cardiovascular disease [4], age at the time of radiation and presence of other cardiac risk factors such as hypertension, diabetes, coronary artery disease, smoking habits also increase the risk of RICT [53]. The aforementioned risk factors for cardiotoxicity in the women with breast cancer receiving BCS related to treatments use were also weighted through IPTW by using the propensity score ( Table 1). The homogenous covariates between adjuvant WBRT and no adjuvant WBRT after IPTW by using the propensity score may have been crucial in estimating causal treatment effects in our observational study. Thus, we believe that adjuvant WBRT is bene cial for favorable oncologic outcomes compared with no adjuvant WBRT after IPTW by using the propensity score (Tables 1 and 2). Our follow-up time (median follow-up durations were 6.96 and 5.09 years for adjuvant WBRT and no adjuvant WBRT, respectively) was also su cient to estimate the signi cant differences in OS, breast cancer death, LRR, and DM between adjuvant WBRT and no adjuvant WBRT in women with left-side breast IDC and HFrEF receiving BCS.
According to our literature review, this is the rst study to estimate the oncologic outcomes of adjuvant WBRT among women with left-side breast IDC and HFrEF receiving BCS. No consensus or evidence for the use of adjuvant WBRT in women with left-side breast IDC and HFrEF receiving BCS is present. In the IPTW-adjusted models, adjuvant WBRT was associated with a decrease in the risk of all-cause death, breast cancer death, LRR, and DM among women with left-side breast IDC and HFrEF receiving BCS (Table 2 and 3). The improvement in contemporary RT techniques with decreased irradiation doses and decreased volumes to the heart and the long-term improvement in mortality rates in patients with HFrEF over time might have caused the signi cant bene cial oncologic outcomes of adjuvant WBRT in women with leftside breast IDC and HFrEF receiving BCS [2,27,28]. Our study is the rst to demonstrate that the potential bene ts of adjuvant WBRT with contemporary RT techniques outweigh the risk of RICT given the patients' prognosis and likely longterm OS, breast cancer death, LRR, and DM bene ts (Table 2 and 3). According to our ndings, we strongly suggested that women with left-side breast IDC and HFrEF receiving BCS should also receive adjuvant WBRT to decrease the risk of allcause death, breast cancer death, LRR, and DM. Table 2 and 3, adjuvant WBRT was a signi cant prognostic factor for OS, breast cancer death, LRR, and DM compared with no adjuvant WBRT in women with left-side IDC and HFrEF receiving BCS; moreover, old age (>65 years), CCI ≥ 1, advanced pT stages (pT2-4), advanced pN stages (pN1-3), extensive RT eld [54], hormone receptor negative status, and differentiation Grade II-III were signi cant prognostic factors for OS, compatible with ndings of previous studies [16,17,[55][56][57][58][59]. Moreover, advanced pN stages (pN1-3)., hormone receptor negative status, no heart-sparing RT technique [54], and differentiation Grade II-III were signi cant poor prognostic factors for LRR and DM in women with left-side breast IDC and HFrEF receiving BCS, which is also compatible with ndings of previous studies [16,17,[55][56][57][58][59]. Our ndings of prognostic factors for OS, breast cancer death, LRR, and DM in women with IDC and HFrEF receiving BCS are similar with those of previous studies [16,17,[55][56][57][58][59], and no additional prognostic factor has been identi ed in previous studies other than the ones determined in the current study irrespective of whether underlying HFrEF was present.

As shown in
A strength of our study was that it was the rst long-term follow-up cohort study to estimate the survival outcomes of adjuvant WBRT or no adjuvant WBRT among women with left-side IDC and HFrEF receiving BCS. The covariates between the adjuvant WBRT and no adjuvant WBRT groups were homogenous for women with left-side IDC and HFrEF receiving BCS, with no selection bias (Table 1). No study has estimated the effect of adjuvant WBRT on women with left-side IDC and HFrEF receiving BCS. In our study, the poor prognostic factors for OS in women with left-side IDC and HFrEF receiving BCS were old age, CCI ≥ 1, advanced pT stages (pT2-4), advanced pN stages (pN1-3), extensive RT eld, no heart-sparing RT techniques, hormone receptor negative status, and differentiation Grade II-III of ( Table 2), which are consistent with factors in women with breast cancer without HFrEF reported in previous studies [54,58,59]. Furthermore, our study is the rst to demonstrate the bene ts of adjuvant WBRT with contemporary RT techniques (over 90% combined with DIBH) for OS, breast cancer death, LRR, and DM in women with left-side IDC and HFrEF receiving BCS. Our ndings should be considered in future clinical practice and prospective clinical trials. We suggest that adjuvant WBRT is valuable to achieving better outcomes of OS, breast cancer death, LRR, and DM in women with left-side IDC and HFrEF receiving BCS.
This study has some limitations. First, because all women with left-side breast IDC and HFrEF were enrolled from an Asian population, the corresponding ethnic susceptibility compared with the non-Asian population remains unclear; hence, our results should be cautiously extrapolated to non-Asian populations. However, no evidence exists as to the differences in oncologic outcomes in Asian versus non-Asian patients with breast IDC and HFrEF receiving BCS. Second, irradiation doesvolume is very important for heart and left anterior descending artery [60-63]. However, most observational studies from database have no detail dose-volume to the heart and left anterior descending artery [10,[26][27][28][29][30][31], not only our study. This is the major limitation in our study and aforementioned observational studies. Nevertheless, our study having more detail characteristics of TNM stages and molecular expression, endpoints including breast cancer-speci c survival and overall survival comrade with other observational studies [10,[26][27][28][29][30][31]. Our ndings is also the rst to examine the effect of adjuvant WBRT on detail oncologic outcomes such as all-cause death, breast cancer death, LRR, and DM in women with left-side breast IDC and HFrEF receiving BCS. Third, the diagnoses of all comorbid conditions were based on ICD-10-CM codes. However, the combination of Taiwanese TCRD and National Health Insurance Research Database appears to be a valid resource for population research on cardiovascular diseases, stroke, or chronic comorbidities [64-66]. Moreover, the Taiwan Cancer Registry Administration randomly reviews charts and interviews patients to verify the accuracy of the diagnoses, and hospitals with outlier chargers or practices may be audited and subsequently be heavily penalized if any malpractice or discrepancy is detected. Accordingly, to obtain crucial information on population speci city and disease occurrence, a large-scale randomized trial comparing carefully selected patients undergoing suitable treatments is essential. Finally, the TCRD does not contain information regarding dietary habits or body mass index, which may be risk factors for mortality. Nevertheless, considering the magnitude and statistical signi cance of the observed effects in this study, these limitations are unlikely to affect the conclusions.

Conclusions
Adjuvant WBRT was associated with a decrease in all-cause death, breast cancer death, LRR, and DM among women with left-side breast IDC and HFrEF compared with no adjuvant WBRT. We suggest adjuvant WBRT combined with DIBH for women with left-side IDC receiving BCS, even if they have HFrEF.

Consent for publication: Not applicable
Availability of data and material: The data sets supporting the study conclusions are included in this manuscript and its supplementary les.  10908, 10909, 11001, 11002, 11003, 11006, and 11013).   Table 2 were adjusted. Table 3. Multivariate analysis of breast cancer death for propensity score-weighted population with breast cancer and heart failure with reduced ejection fraction receiving breast conservative surgery Figure 1 Kaplan-Meier overall survival curves of propensity score-weighted population with breast cancer and heart failure with reduced ejection fraction receiving breast conservative surgery Kaplan-Meier locoregional recurrence-free survival curves of propensity score-weighted population with breast cancer and heart failure with reduced ejection fraction receiving breast conservative surgery Kaplan-Meier distant metastasis-free survival curves of propensity score-weighted population with breast cancer and heart failure with reduced ejection fraction receiving breast conservative surgery