According to the American Cancer Society’s most recent estimates, about 207,090 new cases of invasive breast cancer will be diagnosed in women in the United States in 2010; about 39,840 women will die from breast cancer.1 The American Joint Commission on Cancer (6th edition, 2002 and 7th edition, 2009) categorizes breast tumors that are >5 cm (T3) without regional lymph node disease (N0) or distant metastasis (M0) as stage IIB. This stage of breast cancer is rare, and the incidence is <4% from published literatures.2–4 After mastectomy, pT3N0 stage breast cancer historically has been treated with postmastectomy radiation therapy (PMRT) in a select group of patients. There are no clear recommendations for adjuvant PMRT in stage pT3N0 patients.
Several reports have shown no benefit of adjuvant PMRT, whereas others have tried to identify its benefit with regard to locoregional control, cause-specific survival (CSS), and overall survival (OS) in stage pT3N0 tumors with certain high-risk features like presence of lymphovascular invasion (LVI), large tumor size, and in premenopausal patients.5–11 More recently, neoadjuvant chemotherapy is supported by the National Comprehensive Cancer Network guidelines and is being prescribed in clinical stage T3N0 patients before surgery to facilitate breast conservation.12 This can further complicate the true pathologic lymph node status and adjuvant PMRT recommendations.
In the Early Breast Cancer Trialist Collaborative Group study, the large reductions in the local recurrence corresponded to significant reductions in the cumulative incidence of any recurrence at year 10 (45.2% radiotherapy vs. 53.8% control; absolute difference 8.7%; P<0.00001) and at year 20 (54.4% vs. 61.5%; absolute difference 7.1%; P<0.00001). However, in the subset of node-negative patients after mastectomy, there was the smallest absolute reduction in local recurrence rates with addition of PMRT as compared with without addition of adjuvant PMRT (2.7% PMRT vs. 9.2% no PMRT, respectively; P<0.0001). They interpreted this finding to show that addition of radiation therapy prevented the locoregional recurrence (LRR) that, in turn, led to a less rapid occurrence of distant metastases. In the meta-analysis, for a 20% absolute reduction in 5-year local recurrence, there was a corresponding 5% reduction in 15-year breast cancer mortality.5
In an analysis of 313 patients with stage pT3N0 selected from various National Surgical Adjuvant Breast and Bowel Project studies, the authors challenge the routine use of PMRT. They found an isolated LRR of only 7% after mastectomy with or without adjuvant systemic therapy and without radiotherapy. The authors recommended that PMRT should not be used routinely in these patients; however, in tumors with high-risk features, such as large size and 4 or more lymph nodes with metastases, there is an associated increased chance of local recurrence that could benefit from adjuvant PMRT.6 In the subset analysis of the Danish randomized studies (82b, 82c) involving stage II or III node-negative postmastectomy patients, PMRT decreased the incidence of locoregional failure (LRF) and, at 10 years, there was an OS benefit seen in the premenopausal group (82% for PMRT and chemotherapy vs. 70% for chemotherapy alone).7,8
However, the recommendation for adjuvant PMRT for postmastectomy patients in stage pT3N0 disease is still controversial. Given the rarity of stage T3N0 breast cancer, we used data from Surveillance, Epidemiology, and End Results (SEER). A prior SEER database (1988 to 2003) analysis of all postmastectomy patients treated with PMRT did not show an increase in OS. However, they reported an increased mortality in older patients, patients with estrogen receptor (ER)− disease, and in patients with high-grade tumors.13
In this study, we analyzed the impact of adjuvant PMRT on CSS and OS in postmastectomy stage pT3N0 breast cancer among women younger than 50 years of age. Local recurrence data are not available in SEER, and CSS and OS were used as the primary endpoints in the analysis. A subsequent stratified analysis was made for patients younger than 40 years of age and those between 40 and 50 years of age, to assess the impact of PMRT on CSS specifically in these 2 defined age groups. The association between PMRT and CSS was also stratified by ER status, progesterone receptor (PR) status, and ethnicity.
MATERIALS AND METHODS
SEER 17 Registries Database
The SEER research data include SEER incidence and population data stratified by age, sex, race, year of diagnosis, and geographic areas (including SEER registry and county). This database contains data from the SEER 13, plus Greater California, Kentucky, Louisiana, and New Jersey. Cases are associated with the population data using 5 racial groups: white, black, American Indian/Alaska Native, Asian/Pacific Islander, and Hispanic. This data set includes cases from 2000 to 2008, but because of the impact of Hurricane Katrina from July to December 2005, Louisiana cases diagnosed for that 6-month time period have been excluded from the research database.14 SEER*Stat software (version 6.6.2) was used to perform all queries. Data were queried from the database as follows: Incidence−SEER 17 Regs Research Data+Hurricane Katrina Impacted Louisiana Cases, November 2009 Sub (1973 to 2007 varying)−Linked To County. Patients younger than the age of 50 with malignant breast cancer who underwent mastectomy and were staged as pT3N0M0, diagnosed during 1998 to 2007, were analyzed in the study. We only included patients who had a clear record of receiving PMRT or no PMRT. Patient age, race, tumor size and grade, ER/PR status, and survival time were recorded for all identified patients. Survival time was recorded at 1-month intervals up to 120 months. Exclusion criteria included male breast cancer, ductal carcinoma in situ, lobular carcinoma in situ, all histology other than invasive ductal and lobular carcinoma, all stages except pT3N0, all neoadjuvant treatment, all second primary cancers, or unknown causes of death (N=6) for the purpose of calculating CSS. Inclusion criteria included female breast cancer, invasive ductal and lobular carcinoma, stage pT3N0, mastectomy and axillary node-negative status, and age up to 50 years.
The primary endpoints of the analysis were OS and CSS. OS was defined as the time from diagnosis until death from any cause (or until date of last follow-up if no death). CSS was ascertained by selecting breast cancer as the cause of death in the SEER database search. Deaths due to causes other than breast cancer were censored when estimating CSS. CSS was defined as the time from diagnosis until death from breast cancer (or until date of last follow-up). Kaplan-Meier survival analysis was performed to evaluate OS and CSS, and the log-rank test was used to compare OS/CSS between patients treated with and without PMRT. Univariate Cox proportional hazards regression analysis was subsequently performed to estimate hazard ratios (HR) for the association between PMRT treatment and CSS, stratified by the following factors: age (younger than 40 vs. older than or equal to 40), race (white vs. black/other), ER/PR status, tumor size (<7 vs. ≥7 cm), and tumor grade (1 vs. 2/3). All HRs reflect the increased (or decreased) risk of breast cancer death (ie, CSS) associated with PMRT treatment relative to no-PMRT treatment (referent group). Multivariate Cox proportional hazards regression analysis was performed to evaluate the independent effect of PMRT treatment on CSS after adjustment for age, ethnicity, tumor size, tumor grade, and ER/PR status. All P-values are 2 sided with statistical significance evaluated at the 0.05 α level. Ninety-five percent confidence intervals (95% CIs) for the HRs were calculated to assess the precision of the obtained estimates. All analyses were performed using SPSS version 19.0 (SPSS Inc., Chicago, IL).
From the SEER database, there were 1104 patients eligible for this study; all were available for the OS analysis but only 1098 (99.5%) were available for CSS analysis. Six (0.5%) patients were excluded from the CSS analysis due to information missing from the SEER system pertaining to cause of death. The descriptive counts and frequencies of relevant variables are shown in Table 1.
Less than half (N=519; 47%) of the 1104 patients analyzed had received PMRT, whereas 585 (53%) did not receive PMRT. At diagnosis, 347 (31.4%) patients were younger than the age of 40 and 757 (68.6%) were 40 years of age or older. In 742 (67.2%) patients, tumor size was ≤7 cm, and in 342 (31%), it was >7 cm. Both ER and PR status was negative in 420 (38%) patients. Kaplan-Meier survival analysis was performed to compare OS and CSS between patients with and without PMRT (Figs. 1A, B). No difference in OS or CSS was detected, and neither median OS nor CSS were reached for either group. The 5-year OS was 87.6% and 88.3%, for both the no-PMRT and PMRT groups, respectively (P=0.34 by log-rank test). The 5-year CSS was 89.4% and 91.5%, for the no-PMRT and PMRT groups, respectively (P=0.44 by log-rank test). Subsequently, univariate Cox proportional hazards analysis was performed to compare CSS between the PMRT and no-PMRT groups, stratified by prognostic factors of interest (Table 2). Among ER+ and PR+ patients, the HRs for breast cancer-related death were increased with PMRT treatment compared with without (P=0.10 and 0.14, among ER+ and PR+ patients, respectively). Among white patients, PMRT treatment was associated with an increased risk of breast cancer-related death (P=0.10). Similarly, among patients 40 years of age or older, PMRT treatment was associated with an increased risk of breast cancer-related death (P=0.09). However, for patients younger than the age of 40, PMRT treatment was associated with a trend toward a decreased risk of breast cancer-related death (HR=0.65; P=0.25). A trend was detected for improved CSS for women younger than 40 years of age with PMRT. The 5-year CSS was 85.6% for no-PMRT and 90.8% for PMRT groups (P=0.25 by log-rank test). Median CSS was not reached for either group (Fig. 1C). The detailed data for all prognostic factor stratifications are listed in Table 2. Multivariate Cox proportional hazards analysis was performed to evaluate predictors of CSS (Table 3). Factors such as black/other race (HR=1.71; 95% CI, 0.99-2.96; P=0.05), ER− (HR=2.24; 95% CI, 0.82-6.12; P=0.12), and PR− (HR=1.95; 95% CI, 0.70-5.47; P=0.20), all suggested a trend toward decreased CSS. PMRT was not associated with increased CSS in the multivariate analysis (HR=1.09; 95% CI, 0.64-1.86; P=0.76).
In the current study, we analyzed the impact of PMRT on CSS in patients younger than the age of 50. We also stratified the impact of PMRT on CSS by age; specifically for patients younger than the age of 40 and for patients between 40 and 50 years of age. We also focused on the effect of PMRT on CSS for ER− and PR− patients, because triple-negative (ER, PR, and Her2/neu) patients have a worse outcome and the use of PMRT in this subset of patients is not clearly understood. Because Her2/neu receptor information is currently not available in the SEER database, we used only ER and PR information in our study. Our analysis showed that the 5-year CSS was 89.4% and 91.5% for the no-PMRT and PMRT groups, respectively; the 5-year OS was 87.6% and 88.3% for the no-PMRT and PMRT groups, respectively. PMRT adversely affected CSS after approximately 90 months; however, this relationship between radiation treatment and CSS seemed to be dependent on age (older than or equal to 40 vs. younger than 40). After stratifying by age group, this adverse radiation treatment effect was restricted to the 40 years of age or older group (ie, 40 to 50 y). In patients younger than age 40, PMRT treatment seemed to be beneficial. The phenomena could be explained by the assumption that the potential advantage of improved LC with PMRT did not translate into a statistically significant CSS benefit in patients younger than age 40.
In female breast cancer, stage pT3N0 is rare with an incidence of 0.5 to 4% of all female breast cancers.4 Current National Comprehensive Cancer Network guideline supports adjuvant PMRT for patients with tumor size >5 cm with no axillary node metastases.12 Given the low incidence and paucity of literature, the role of PMRT is not clearly elucidated. Several publications have noted that indications of adjuvant PMRT should be individualized in stage pT3N0 patients based on risk factors for recurrence, for example, LVI, premenopausal status, ER status, age, size of tumor, and ethnicity.9,11,15,16 In a multi-institution study of 70 patients, Floyd and colleagues reported that node-negative breast cancer and tumors >5 cm treated with mastectomy and adjuvant systemic therapy but without PMRT had a low (7.6%) actuarial LRR. In their study, the LRF was 21% at 5 years for tumors with LVI. The disease-free survival and OS was 83% and 87%, respectively, at 5 years.9
Wallgren and colleagues reported on the risk factors for LRF after mastectomy among breast cancer patients from the International Breast Cancer Study Group Trials I through VII. Among the 1275 node-negative patients, the high-risk factors for premenopausal women were LVI and tumor size >2 cm, but only LVI for postmenopausal women. Patients received either 1 cycle of perioperative chemotherapy or observation after mastectomy. The 10-year cumulative incidence of LRF with or without distant metastases was 16% for the premenopausal group.10 Mignano and colleagues reported on a single-institution study from Johns Hopkins Hospital of 101 patients with stage T3N0 breast cancer. Adjuvant PMRT showed LRF and distant failure of 11% in each group. There was no effect of tumor size (<7 vs.>7 cm) on LRF; however, grade, LVI, and premenopausal status approached significance in predicting local recurrence.15
In the pooled analysis of 8878 breast cancer patients from National Surgical Adjuvant Breast and Bowel Project B-13, B-14, B-19, B-20, and B-23 node-negative trials, patients with tumors ≥5 cm had an overall 10-year cumulative isolated LRF incidence of only 7.1% after mastectomy and with or without systemic treatment but no PMRT. Chest wall recurrence was the most common site of failure that was noted in 24/28 patients. They concluded that PMRT would not be necessary as a routine adjuvant treatment in node-negative T3N0 patients. The authors, however, urge caution while interpreting these data, because factors such as LVI, grade, and margins were not included in this database, and pathologic tumor size alone with node-negative status should not be a basis for adjuvant PMRT.17
Some studies have reported a decrease in locoregional control only, whereas others have noted benefit with regard to OS and/or CSS or disease-specific survival. In the Danish randomized trials, PMRT in both premenopausal and postmenopausal women (82b, 82c) showed decreased LRF. Among node-negative premenopausal women, the LRF was 17% (no PMRT) versus 3% (PMRT) and in postmenopausal women, it was 23% (no PMRT) compared with 6% (PMRT). At 10 years, PMRT also showed an OS benefit in the premenopausal group (82% for PMRT and chemotherapy and 70% for chemotherapy alone).7,8 The total number of axillary lymph nodes removed in this study was lower than other trials, which could mean that some of the undissected nodes have occult metastasis and therefore caused concern regarding the percentage of true stage T3N0 patients. In the retrospective study conducted by Helinto and colleagues, they identified 81 patients with T3N0 breast cancer; upon review, only 38 patients were of true T3N0 status. LRF was noted in 3 of 5 (60%) patients without PMRT as compared with 3/33 (9%) patients with PMRT. The authors concluded that patients who received PMRT had a statistically better distant disease-free survival rate and OS rate compared with patients who did not receive PMRT.18
Although the SEER database does not contain information about local control, OS and CSS have been used as readouts for PMRT effect. Two SEER studies using different data sets examined the impact of PMRT on pT3N0 patients. Yu and colleagues used the SEER 1988 to 2003 database for a propensity score case-matched analysis to study the average treatment effect of PMRT on mortality. Certain high-risk subgroups like older age, presence of high-grade tumors, African American ethnicity, and ER− tumors were significant predictors for worsened survival overall and showed increased mortality. They were unable to find definite OS benefit with PMRT in all patients, including the high-risk subgroups.13 Another SEER study used the 1988 to 2002 databases and reported the impact of PMRT on CSS and OS in patients with stage T3N0. They noted increased PMRT rate, from 22% (1988 to 1997 database) to 41%. They noted a statistically significant increase in 10-year OS rate for PMRT, seen only in women older than 50 years; no benefit was seen in actuarial 10-year CSS. The authors concluded that the increased OS benefit with PMRT in the absence of CSS was likely related to patient selection in the nonrandomized data set.19
Although previous SEER studies showed no benefit in OS from PMRT in patient younger than 50, this could be due to the fact that the SEER database does not provide local control data and would not mean that PMRT does not have a local control benefit. As shown in meta-analysis performed by the Early Breast Cancer Trialists’ Cooperative Group, PMRT does contribute to better local control in node-negative patients but does not improve CSS.5 The use of PMRT is on the rise, from 22% during 1988 to 1997 to 41% during 1998 to 2002.19 Our study used the most current available data set from the SEER database (1998 to 2007), which showed a PMRT rate of 47% in pT3N0 patients aged younger than 50. In a survey study by Ceilley et al20 reported that, in real practice, 84.8% of European physicians and 88.3% of American physicians would offer chest wall PMRT to patients with pT3N0 lesions.
PMRT benefitted African American patients younger than 50 (HR=0.70; P=0.33; slight nonsignificant trend indicated for favorable radiation effect) but not white patients younger than 50 (HR=1.58; P=0.10; nonsignificant trend indicated for unfavorable radiation effect). ER+ and/or PR+ patients did not benefit from PMRT (HR=2.87; P=0.05). Multivariate analysis did not show any effect of PMRT on CSS. African American patients, ER− and PR− tumors showed a nonsignificant trend, which indicated a higher HR for breast cancer-related death in our study. These findings correlated with prior studies indicating race, older age, and ER status being important for prognosis and outcome of patients with breast cancer.9,13,16,18 This also confirmed the validity of our SEER database data extraction and statistical analysis. The 2 prior SEER studies that reported on PMRT in stage T3N0 breast cancer were unable to show a clear CSS or OS benefit in patients younger than 50.13,19 In our study, among patients younger than age 40, PMRT treatment seemed to be beneficial with a nonsignificant trend (HR=0.65; P=0.25). Limitations of this study include the fact that the SEER database is limited in information regarding local control rate, the type of adjuvant chemotherapy, hormonal therapy, surgical margin status, LVI, and Her2 receptor status. The impact of these factors on local control, on CSS, and possibly on OS is difficult to ascertain. Multi-institutional collaborative prospective studies with genome information could further elucidate the role of PMRT in this subset of younger patients with pT3N0 stage breast cancer.
On the basis of our SEER database study, adjuvant PMRT could be beneficial in stage T3N0 patients younger than the age of 40. African American patients and ER−/PR− patients younger than age 50 were at a higher risk for breast cancer-related mortality and could potentially benefit from PMRT. Patients with positive ER and/or PR status did not benefit from PMRT. A significant difference in OS and CSS in patients younger than 50 with PMRT was not detected. Although local control data are lacking, it is prudent to postulate that PMRT can lead to better local control in patients younger than 40. Future studies on the impact of PMRT on patients younger than age 50 and association of ER/PR/Her2 triple-negative status might offer more insight in individualizing radiation treatment.
1. Jemal A, Siegel R, Xu J, et al .Cancer Statistics, 2010.CA Cancer J Clin. 2010; 60:277–300.
2. Edge SB, Compton CC .The American Joint Committee on Cancer: the 7th edition of the AJCC cancer staging manual and the future of TNM.Ann Surg Oncol. 2010; 17:1471–1474.
3. Greene FL .The American Joint Committee on Cancer: updating the strategies in cancer staging.Bull Am Coll Surg. 2002; 87:13–15.
4. Floyd SR, Taghian AG .Post-mastectomy radiation in large node-negative breast tumors: does size really matter? Radiother Oncol. 2009; 91:33–37.
5. . Early Breast Cancer Trialists’ Collaborative Group .Favorable and unfavorable effects on long-term survival of radiotherapy for early breast cancer: an overview of the randomized trials.Lancet. 2000; 355:1757–1770.
6. Taghian A, Jeong JH, Mamounas E, et al .Patterns of loco regional failure in patients with operable breast cancer treated by mastectomy and adjuvant chemotherapy with or without tamoxifen and without radiotherapy: results from five National Surgical Adjuvant Breast and Bowel Project randomized clinical trials.J Clin Oncol. 2004; 22:4247–4254.
7. Overgaard M, Hansen PS, Overgaard J, et al .Danish Breast Cancer Cooperative Group 82b Trial. Postoperative radiotherapy in high-risk premenopausal women with breast cancer who receive adjuvant chemotherapy.N Engl J Med. 1997; 337:949–955.
8. Overgaard M, Jensen MB, Overgaard J, et al .Postoperative radiotherapy in high-risk postmenopausal breast-cancer patients given adjuvant tamoxifen: Danish Breast Cancer Cooperative Group DBCG 82c randomized trial.Lancet. 1999; 353:1641–1648.
9. Floyd SR, Buchholz TA, Haffty BG, et al .Low local recurrence rate without post mastectomy radiation in node-negative breast cancer patients with tumors 5 cm and larger.Int J Radiat Oncol Biol Phys. 2006; 66:358–364.
10. Wallgren A, Bonetti M, Gellber RD, et al .Risk factors for loco regional recurrence among breast cancer patients: results from International Breast Cancer Study Group Trials I through VII.J Clin Oncol. 2003; 21:1205–1213.
11. Jagsi R, Raad AR, Goldberg S, et al .Loco regional recurrence rates and prognostic factors for failure in node-negative patients treated with mastectomy: implications for post mastectomy radiation.Int J Radiat Oncol Biol Phys. 2005; 62:1035–1039.
13. Yu JB, Wilson LD, Dasgupta T, et al .Post mastectomy radiation therapy for lymph node-negative, locally advanced breast cancer after modified radical mastectomy: analysis of the NCI Surveillance, Epidemiology, and End Results database.Cancer. 2008; 113:38–47.
Surveillance, Epidemiology, and End Results (SEER) Program, 1973–2008 (Including July-December 2005 Hurricane Katrina Impacted Louisiana Cases) National Cancer Institute, DCCPS, Surveillance Research Program, Cancer Statistics Branch, Accessed October 2, 2011
15. Mignano JE, Gage I, Piantadosi S, et al .Local recurrence after mastectomy in patients with T3N0 breast carcinoma treated without postoperative radiation therapy.Am J Clin Oncol. 2007; 30:466–472.
16. O’Brien KM, Cole SR, Tse CK, et al .Intrinsic breast tumor subtypes, race, and long-term survival in the Carolina Breast Cancer Study.Clin Cancer Res. 2010; 16:6100–6110.
17. Taghian AG, Jeong JH, Mamounas EP, et al .Low loco regional recurrence rate among node-negative breast cancer patients with tumors 5 cm or larger treated by mastectomy, with or without adjuvant systemic therapy and without radiotherapy: results from five national surgical adjuvant breast and bowel project randomized clinical trials.J Clin Oncol. 2006; 24:3927–3932.
18. Helinto M, Blomqvist C, Heikkila P, et al .Post-mastectomy radiotherapy in pT3N0M0 breast cancer: is it needed? Radiother Oncol. 1999; 52:213–217.
19. McCammon R, Finlayson C, Schwer A, et al .Impact of post mastectomy radiotherapy in T3N0 invasive carcinoma of the breast: a Surveillance, Epidemiology, and End Results database analysis.Cancer. 2008; 113:683–689.
20. Ceilley E, Jagsi R, Goldberg S, et al .Radiotherapy for invasive breast cancer in North America and Europe: results of a survey.Int J Radiat Oncol Biol Phys. 2005; 61:365–373.