Abstract
Background: The effectiveness of postmastectomy radiation (PMRT) in breast cancer patients with N2 and N3 nodal status is often evaluated in combination with other postmastectomy adjuvant treatments. To account for interaction effects and determine the individual impact of PMRT alone, the National Cancer Data Base (NCDB) was analyzed from 2004-2011. Patients and Methods: We evaluated a cohort of 38,442 women diagnosed with pathological stage N2/N3 breast cancer who underwent mastectomy between 2004 and 2011 from the NCDB. Overall survival was the outcome variable; primary predictor variable was treatment (PMRT, adjuvant anti-hormonal therapy, and adjuvant chemotherapy). Additional variables addressed and adjusted for included: age, race, Charlson Comorbidity index, education, income, payer status, distance traveled, facility type, diagnosing/treating facility, treatment delay, grade of tumor, tumor size, and stage at diagnosis, tumor histology, ER/PR status, and lymph node invasion as well as PMRT boost and dosage. Multivariate Cox regression was used to investigate the effect of PMRT on overall survival while adjusting for secondary predictive factors. Results: The majority of patients received one or more postmastectomy procedures such as radiation (69.31%), chemotherapy (88.79%), and/or hormone therapy (62.66%). The median overall survival for all patients was 8.41 years. In multivariate analysis, effects of treatment on survival were significant for chemotherapy alone, hormonal therapy alone, and a combination of PMRT with either chemotherapy, hormonal therapy, or both. Compared to patients without treatment, patients who received PMRT alone were not significantly associated with an increased risk of death; patients who received hormone therapy alone or chemotherapy alone had a reduced risk of death by 15% and 31%, respectively. With the combination of all three treatments, risk of death was reduced by 64%. Conclusion: PMRT was not found to be a significant predictor of risk of death for pN2/N3 breast cancer patients when adjusting for socioeconomic factors, disease characteristics, and interaction effects of chemotherapy and hormonal therapy. The benefit of PMRT in addition to chemotherapy, hormonal therapy or both on overall survival seems to be marginal and not statistically significant.
Abbreviations: PMRT, Postmastectomy radiation therapy; LRR, prevent locoregional recurrence; AJCC, American Joint Committee on Cancer; NCDB, National Cancer Data Base; OS, overall survival; DAS, direct adjusted survival; NCCN, National Comprehensive Cancer Network; ASCO, American Society of Clinical Oncology; ER, estrogen receptor; PR, progesterone receptor; HR, Hazard Ratio; CI, Confidence Interval.
Postmastectomy radiation therapy (PMRT) is commonly used, often in conjunction with other treatments such as chemotherapy and hormone therapy, in order to prevent locoregional recurrence (LRR) and increase survival in breast cancer patients after definite surgery. Increased risk for LRR can be attributed to factors such as lymph node involvement, large tumor size, and positive or close surgical margins (1). The American Joint Committee on Cancer (AJCC) classifies pathologic regional lymph node (pN) involvement: no lymph node involvement (pN0), 1-3 positive nodes (pN1), 4-9 positive modes (pN2), >9 positive nodes (pN3).
The use of PMRT for breast cancer patients with low (pN0/1) risk for LRR is contentious. Earlier guidelines for breast cancer treatment did not suggest PMRT for patients with less than 4 positive lymph nodes (pN0/1). Some studies show that PMRT is beneficial for pN1 breast cancer patients (2-5) while others suggest that PMRT does not significantly increase survival for lower-risk groups (6-8).
PMRT in high-risk breast cancer has demonstrated not only reduction in locoregional recurrences but also prolonged overall survival. The significance of PMRT in conjunction with adjuvant chemotherapy has been established by three randomized trials and two meta-analyses (9-13). A phase III randomized trial by the Danish breast cancer cooperative groups (protocol 82b and 82c) evaluated the role of postoperative radiation in addition to adjuvant systemic therapy in 3,083 pre- and post-menopausal women with high-risk breast cancer (14). In the univariate analysis of patients with the 4+ positive lymph node group, PMRT reduced 15-year locoregional failure rates from 51% to 10% and reduced overall 15-year survival rate from 21% to 12% in radiation therapy (RT) and RT omitted group, respectively. A sub-group (n=1,000) univariate analysis of the Danish Study (15) indicated that the large local recurrence reduction in the poor-prognosis group did not translate into any reduction in 15-year breast cancer mortality for women with PMRT.
The British Columbia randomized trial demonstrated better survival outcomes in high-risk breast cancer patients treated with PMRT and adjuvant chemotherapy (16). In that trial, 318 pre-menopausal women with positive lymph nodes who underwent a modified radical mastectomy and axillary lymph node dissection, in addition to chemotherapy, were randomized to either PMRT (37.5 Gy in 16 fractions) or no PMRT. PMRT and chemotherapy compared to chemotherapy alone, were associated with 27% reduced risk of death (HR=0.73, 95% CI=0.55-0.98). However, the effect of PMRT on OS is not statistical significant for ≥4 lymph nodes patients (HR=0.7, 95% CI=0.46-1.06).
A 2015 analysis of the National Cancer Data Base (NCDB) and Surveillance, Epidemiology, and End Results by Huo et al. concluded that PMRT reduced cause-specific mortality for patients with 3 positive lymph nodes by 14% but was not significantly beneficial for those with 1 positive lymph node (17). As patients with 4 or more positive nodes (pN2/3) are considered high risk for LRR and both National Comprehensive Cancer Network (NCCN) and American Society of Clinical Oncology (ASCO) guidelines recommend these patients to receive PMRT as part of their standard treatment (18, 19). A recent study using NCDB showed that one-third of high-risk patients did not receive PMRT as suggested by treatment guidelines (20). Several studies (15-17, 20) have not addressed survival outcomes of pN2/3 breast cancer patients who received PMRT while accounting for other treatments (chemotherapy and hormone therapy) received.
A 2015 study by Whelan et al. concluded that regional irradiation in early-stage breast cancer did not improve survival (7). Another large randomized study by Poortmans et al. showed that PMRT had no effect on survival of breast cancer patients (8).
Studies have shown that survival outcomes of breast cancer patients were also affected by stage at diagnosis, age, race, and socioeconomic factors (education, income, payer status, etc.) (21-25). The effect of socioeconomic factors such as payer status in particular are not limited to breast cancer (24, 25). The present study aimed to evaluate the overall survival of high-risk breast cancer patients based on treatment regimen with particular focus on the impact of PMRT while adjusting for other treatment factors, demographics, social economic status and disease characteristics.
Patients and Methods
The dataset used in this study derived from a de-identified NCDB file. The NCDB captures approximately 70% of all newly-diagnosed cases of cancer in the United States at the institutional level (26). The International Classification of Disease for Oncology, third edition codes (C500-C506, and C508, C509) associated with a diagnosis of breast cancer were used to select patients. A total of 38,442 women with American Joint Committee on Cancer (AJCC) pathological stage N2 and N3 (pN2/N3) breast cancer who underwent mastectomy between 2004 and 2011 were analyzed. Patients were followed until December 31, 2012. Patients who were identified as pN2/3, but diagnosed at any AJCC anatomic stage other than III were excluded.
Survival time of breast cancer patients was calculated from date of diagnosis to date of death, date of loss of follow-up, or date of study end (December 31, 2012). Primary predictor variable was treatment (PMRT, anti-hormonal therapy, and chemotherapy). Additional variables were addressed and adjusted for age, race, Charlson Comorbidity index, payer status, income, education, distance travelled, facility type, diagnosing/treating facility, and diagnosis to treatment interval, tumor grade, tumor histology, tumor size, estrogen receptor (ER) status or progesterone receptor (PR) status, and lymph node invasion as well as PMRT boost and dosage.
Payer status was categorized as uninsured, private, Medicaid, Medicare (or other government insurance plan). Hormone receptor status was categorized as positive (either ER or PR or both positive) or negative (both ER and PR negative) from the database. Histology was grouped as Invasive Ductal and others. Lymph node status was grouped as 4-9 positives and 10+ positives. Radiation regional dose was grouped as 0 CGY, 1-5,000 CGY, >5,000 CGY. Radiation boost dose was grouped as received and not received.
Age was grouped as 18-49 years, 50-64 years, 65-74 years, ≥75 years. Patients' race was categorized as white, black, or Asian. The Charlson Comorbidity (27) index was categorized as 0, 1 and ≥2. Income, or median household income at the Zip Code level, was grouped as <$30,000, $30,001-34,999, $35,000-45,999 or ≥$46,000. Education, a measure of the percent of adults in the patient's zip code who did not graduate from high school, was grouped as <14%, 14-19%, 20-28% and ≥29%. Income and education were determined using 2000 census data. Distance travelled from the patient's residential Zip Code to his or her medical center's Zip Code was grouped as ≤30 miles and >30 miles.
Per NCDB's PUF dictionary, cancer programs were categorized as community cancer program, comprehensive community cancer program or academic research program. Other services and clinics of cancer programs were excluded due to small numbers. Diagnosing /treating facility was categorized as patients diagnosed and treated at same facility or different facility. Time from diagnosis to any treatment was grouped as 0-30 days, and >30 days. Post-mastectomy treatment status was categorized as received or not received for each of the treatment options: chemotherapy (CT), radiation therapy (RT), and hormone therapy (HT). This categorization resulted in eight postmastectomy treatment groups: no treatment, RT alone, HT alone, CT alone, RT and HT, RT and CT, HT and CT, and a combination of all three treatments.
Characteristics of N2/N3 breast cancer patients with mastectomy.
Kaplan-Meier methods were used to estimate survival curves. Log rank tests were used to compare the survival distributions in univariate analysis. Multivariate Cox regression was used to simultaneously estimate the hazard of death (Hazard Ratio) for treatment received and adjust for other factors. Direct Adjusted Overall Survival was calculated by using Multivariate Cox regression. Statistical Software SAS 9.4 (SAS Inc. Gary, NC) was used for data management, statistical analysis, and modeling. All p-values <0.05 or 95% confidence interval of hazard ratio that did not include 1 were considered statistically significant.
Results
This study used the 2004-2011 NCDB to determine survival outcomes of 38,442 female pN2/N3 breast cancer patients after mastectomy (Table I). Out of these patients, 63.76% had N2 and 36.24% had N3 involvement and 94.71%, 3.90%, 0.48%, 0.91% received a modified radical mastectomy, radical mastectomy, extended radical mastectomy, and a mastectomy that was not otherwise specified. In addition, a majority of these patients received one or more postmastectomy procedures such as RT (69.31%), CT (88.79%), and/or HT (62.66%). There were 20.96% patients treated with RT and CT and 47.46% patients treated with RT and CT and HT.
Among these patients, 63.15%, and 36.85% presented with 4-9 positive and 10+ positive lymph nodes status, respectively. The distribution of tumor grade was 54.58%, 38.67%, and 6.76% for poorly/un-differentiated, moderately differentiated, and well differentiated, respectively. Seventy-four percent of patients were ER/PR positive. Among ER/PR positive patients, 71.99%, 87.66% and 84.64% of patients received RT, CT and HT, respectively (data not shown). The distribution of patient insurance status was 54.01%, 31.82%, 10.59%, and 3.58% for private, Medicare, Medicaid, and uninsured, respectively. The median overall survival (OS) for all patients was 8.41 years. Univariate analysis was conducted and all variables were identified to be significantly associated with OS except distance travelled and PMRT dosage (data not shown).
Multivariate Cox regression of overall survival of predictor effect.
Hazard of death in multivariate Cox regression analysis of postmastectomy radiation effect.
Multivariate cox regression analysis evaluated the effect of treatment on risk of death (Table II). Compared to patients who did not receive any postmastectomy treatment, patients who received PMRT alone did not experience any significant changes in risk of death (HR= 1.03, 95% CI=0.64-1.66). Patients who received CT alone (HR=0.69), HT alone (HR=0.85), or both CT and HT (HR=0.45) demonstrated significantly reduced risk of death compared to peers who did not receive postmastectomy treatments. Patients who received PMRT in addition to CT alone or HT alone or both CT and HT demonstrated further reduced risk of death with HRs of 0.51, 0.56, and 0.36 respectively.
While hazard ratio estimates comparative risk of death overall, direct adjusted survival (DAS) probabilities present the absolute percentage of patient survival at a certain time. Multivariate analysis demonstrated that the postmastectomy treatment with the highest DAS probabilities (Figure 1) was a combination consisting of all three treatments (RT, CT and HT). PMRT as a single modality of treatment alone showed a 0.9% lower 5-year DAS probability compared to patients receiving no postmastectomy treatments. However, compared to patients who received no treatments, patients who received CT alone and HT alone demonstrated improved 5-year DAS by 4.7% and 10.9% respectively. Treatments consisting of hormone therapy and chemotherapy improved DAS by 21.4%. Combination of RT, HT and CT was more effective, improving DAS by 26.5% (data not shown).
Directed adjusted survival of patients with pN2/3 after mastectomy.
Multivariate analysis of hazard ratios associated with disease characterization and progression, confirmed their role as significant predictors of risk of death. Lymph node invasion and ER/PR status were also found to be significant predictors of risk of death. Patients with 10 or more positive lymph nodes had a 51% increased risk of death compared to 4-9 positive lymph nodes. The effects of treatments on DAS probabilities were further stratified by 4-9 and ≥10 positive lymph nodes (Figure 2). For all treatments, improvement in 5-year DAS probabilities was about 10% higher for patients with 4-9 positive lymph nodes than 10+ positive lymph nodes. Patients who were ER/PR-positive were at a 42% reduced risk of death compared to ER/PR-negative patients. Similar 5-year DAS probabilities advantage of minimum 10% were also presented for ER/PR-positive status (Figure 3). For all treatment modalities, improvement in 5-year DAS probabilities was higher for patients with ER/PR positive status. Additionally, patients with poorly or undifferentiated tumors were at 69% higher risk of death compared to patients with well-differentiated tumors. Patients with tumor size of 50 mm were at 69% of higher risk of death compared to peers with tumor size of (≤20 mm).
Table II also demonstrates that, beyond treatment, several socioeconomic factors also significantly affect risk of death. In particular, payer status was found to be a significant predictor of risk of death from breast cancer. Compared to peers with private insurance; patients with no insurance, Medicaid, and Medicare had significantly higher risk of death by 19%, 28%, and 32%, respectively. Further, patients treated at a Community Cancer Program had a 14% higher risk of death compared to peers treated at an Academic/Research Program
Older age at diagnosis was found to be a significant predictor of increased risk of death. Compared to younger peers (ages 18-49 years), patients who were diagnosed at ages 75+ were 1.73-times more likely to die. Black patients had a 29% increased risk of death compared to white peers while Asian patients had a 22% decreased risk of death. Comorbidities also significantly correlated with an increased risk of death; Charlson indexes of 1 and 2 were associated with a 23% and 73% increased risk of death, respectively.
Direct adjusted 5-year survival rate by treatment and lymph nodes status.
Discussion
The commonest site of locoregional recurrence after mastectomy and axillary lymph nodal dissection is the chest wall (50-75%), supraclavicular fossa and infraclavicular regions (20-40%) and axillary (<5%) at 10 years (1). These locoregional recurrences of breast cancer in patients treated with mastectomy and chemotherapy without radiation were reported to be 21% and 22% for patients with 4-9 and ≥10 involved lymph nodes. In the present study, there were only 10.7% and 12.6% patients treated with mastectomy and chemotherapy without radiation for patients with 4-9 and ≥10 involved lymph nodes, repetitively (data not shown). National guidelines (1, 18, 19) specify that breast cancer patients who are at a high risk for locoregional recurrence should receive PMRT.
Accounting for the compounding effects of postmastectomy chemotherapy and hormone therapy, multivariate analysis revealed that PMRT when used alone was not significantly associated with changes in risk of death (HR=1.03, 95%CI=0.64-1.66). This was consistent with existing literature (4, 8, 15, 17). This could suggest minimal influence of PMRT by itself even for high-risk breast cancer patients with more than 4 positive lymph nodes. It is possible that without systemic treatment such as chemotherapy or hormone therapy, prevention of locoregional recurrence does not significantly alter overall survival. In clinical practice, the fraction of patients receiving RT alone is very low (0.77% in present study) and may represent patient population with either poor performance status or significant comorbidities or advanced age.
Analysis also demonstrated that postmastectomy chemotherapy alone and hormone therapy alone significantly reduced risk of death by 31% and 15%, respectively. When PMRT was added to either chemotherapy alone, hormone therapy alone, or both chemotherapy and hormone therapy, risk of death was reduced. While all these hazard ratios were significantly lower compared to the no-treatment group, the marginal benefit of additional PMRT was not statistically significant (Table III). For example, when compared to patients who received chemotherapy alone, patients who received both PMRT and chemotherapy had a hazard ratio of 0.74 (95% CI=0.47-1.15). When compared to patients who received hormone therapy alone, those who received both PMRT and hormonal therapy had a hazard ratio of 0.65 (95% CI=0.41-1.04). When compared to patients who received hormone and chemotherapy, those who received additional PMRT had a hazard ratio of 0.79 (95% CI=0.51-1.24). However, this retrospective study, consistent with existing literature (4, 8, 15, 17), concludes that the addition of PMRT to a treatment regime of chemotherapy alone, hormone alone, or both chemotherapy and hormone therapy has no improvement in overall survival. A randomized controlled clinical trial (prospective study) could further confirm the effect of addition of PMRT to postmastectomy treatment of this group of breast cancer patients.
Direct adjusted 5-year survival rate by treatment and ERPR status.
Beyond postmastectomy treatments, disease and socioeconomic factors were also found to be significant predictors of overall survival. Consistent with existing literature, multivariate analysis demonstrated that disease progression and characterization were significant prognostic factors for breast cancer survival (22, 23, 28). Poorly or undifferentiated tumors, negative ER/PR status, and ≥10 positive lymph nodes were correlated with higher risk of death. Similarly, demographic and socioeconomic characteristics such as older age at diagnosis, black race, uninsured/Medicaid insurance status, and increased comorbidities predicted increased risk of death (21-23, 29).
While there are several strengths to the multivariate analysis performed in this study, such as adjustment for compounding treatments, certain limitations also exist. The size of patient cohort analyzed in the study makes it one of the largest data set pertaining to subset of patients with pN2/pN3 breast cancer and this attributes to our narrow 95% confidence intervals, attesting to statistical significance. However, the reader should keep in mind that this statistical significance does not necessarily translate to clinical significance. In addition, analysis of the effect of PMRT alone may also be limited by the relatively small amount of patients who received PMRT alone (n=196) compared to patients who received other treatment regimens (e.g. chemotherapy alone, n=3,060). Furthermore, patient compliance with recommended treatment guidelines was not investigated in this study. Most importantly the detailed radiation dose received in this population was not investigated. Zip code education and income were collected in lieu of individual education and income. Bias in patient selection and variation in institution reporting is highly possible in a large retrospective national database such as NCDB.
Conclusion
NCDB data were analyzed to determine the impact of PMRT on survival outcomes of pN2/N3 female breast cancer patients who received a mastectomy. Despite guidelines recommending PMRT for breast cancer patients with high risk of locoregional recurrence, the present study found that PMRT alone was not a significant predictor of survival. The marginal benefit of adding PMRT to either chemotherapy alone, hormone therapy alone, or both chemotherapy and hormone therapy was not statistically significant. Recommendation of using PMRT alone to these patients should be deferred or avoided without additional evidence. Further prospective studies are necessary to investigate PMRT effect and additional factors on survival, especially patient compliance to treatment guidelines, associated with postmastectomy treatments for breast cancer with high risk of locoregional recurrence.
Acknowledgements
The Authors wish to acknowledge the Commission on Cancer of the American College of Surgeons and the American Cancer Society for making public data available through the NCDB. The data used in this study were derived from a de-identified NCDB file. The American College of Surgeons and the Commission on Cancer have not verified and are not responsible for the analytic or statistical methodology employed or the conclusions drawn from these data by the investigator.
Footnotes
Conflicts of Interest
The Authors declare that they have no competing interests.
Ethics Statement
With the support from the Chair of Louisiana State University Hospital in Shreveport (currently University Health Shreveport) Cancer program, the corresponding author has applied and has been awarded the National Cancer Data Base (NCDB) Participant Use Data File (PUF) for 1998 to 2012 from the Commission on Cancer (CoC). The PUF is a Health Insurance Portability and Accountability Act (HIPAA) compliant data file containing cases submitted to the Commission on Cancer's (CoC) National Cancer Data Base (NCDB). The PUF contains de-identified patient level data that do not identify hospitals, healthcare providers, or patients as agreed to in the Business Associate Agreement that each CoC-accredited program has signed with the American College of Surgeons. The PUFs are designed to provide investigators associated with CoC-accredited cancer programs with a data resource they can use to review and advance the quality of care delivered to cancer patients through analyses of cases reported to the NCDB. NCDB PUFs are only available through an application process to investigators associated with CoC-accredited cancer programs.
- Received November 11, 2015.
- Revision received December 7, 2015.
- Accepted December 10, 2015.
- Copyright© 2016 International Institute of Anticancer Research (Dr. John G. Delinassios), All rights reserved
References
In this issue
Jump to section
Related Articles
Cited By...
- No citing articles found.