Key points

Radiation therapy has been used for decades to eradicate occult microscopic disease in the postmastectomy chest wall and draining regional nodal basins, thereby decreasing locoregional failure (LRF) and even improving survival end points in select patients.

Many of the historic data clinicians use to guide the use of postmastectomy radiation therapy (PMRT) includes patients treated in the 1970s and 1980s, an era in which LRF was substantially higher than it is now. Since the publication of the landmark randomized studies showing the benefits of PMRT, there has been a marked decrease in LRF among patients with breast cancer after mastectomy because of multiple factors. These factors include improved diagnostic and staging tools, earlier stage at diagnosis, improved surgical techniques, and increasingly effective systemic therapies. In light of this, this article provides updates on recommendations regarding PMRT based on contemporary data.

PMRT for 1 to 3 positive nodes

The use of PMRT in the setting of 4 or more positive lymph nodes has been broadly accepted because of the high risk of LRF in this population; however, controversy remains regarding the utility of PMRT in patients with 1 to 3 positive nodes. The Early Breast Cancer Trialists’ Collaborative Group (EBCTCG) updated their meta-analysis on the role of PMRT in 2014. The meta-analysis included 22 trials and 8135 patients with breast cancer between 1964 and 1986 who were randomly assigned to receive chest wall and regional nodal radiotherapy after mastectomy and axillary dissection. A subset analysis of 1133 patients with 1 to 3 positive nodes who had received systemic therapy showed a LRF rate of 21% without irradiation and 4.3% with PMRT at 10 years ( P <.001). The 10-year rate for any recurrence, either local or distant, was 45.5% without radiation and 33.8% with radiation ( P <.001). Moreover, breast cancer mortalities were 49.4% and 41.5% ( P = .01) without radiation and with radiation, respectively. These data support the benefit of PMRT in preventing both LRF and overall recurrence, as well as improving breast cancer mortality. Although these numbers are clearly supportive of the benefits of PMRT, most patients included in the EBCTCG analysis were included in randomized trials conducted in the 1970s and 1980s and broadly grouped patients by nodal stage but not other more recently elucidated risk factors for recurrence such as receptor status, age, lymphovascular invasion (LVI), grade, and other pathologic features. In the current era, rates of LRF seem to be considerably lower than the rates mentioned earlier, and are understood to vary substantially with risk factors for recurrence. The trend toward lower rates of LRF in more recent years is highlighted in a study from the MD Anderson Cancer Center in which a cohort of 1027 patients with T1 to T2 breast cancer with 1 to 3 positive lymph nodes treated with mastectomy and systemic therapy with or without PMRT were analyzed by treatment era. Specifically, those treated in an early era (1978–1997) were compared with those treated in a later era (2000–2007). In the early cohort, PMRT was observed to decrease 5-year rates of LRF from 9.5% to 3.4% ( P = .028). However in the later cohort, PMRT did not seem to significantly decrease rates of LRF, and 5-year rates of LRF without PMRT were only 2.8%. Overall in modern series, rates of LRF in patients treated with mastectomy and systemic therapy without radiation generally range from 4% to 23% (with most studies ranging from 4% to 10%) depending on risk factors.

Several contemporary retrospective studies have highlighted the significance of specific risk factors for recurrence in determining the utility of PMRT in patients with 1 to 3 positive nodes. Moo and colleagues published a retrospective analysis of 1331 patients with T1 to T2 tumors with 1 to 3 positive nodes who underwent mastectomy with or without PMRT. The overall rate of LRF in the no-PMRT group was 4.3%. On a multivariate analysis of patients in the no-PMRT group, both age less than or equal to 50 years and lymphovascular invasion (LVI) were significantly associated with increased risk of LRF, suggesting that these factors warrant consideration of PMRT in this cohort. Similarly, Yildirim and Berberoglu published a study of patients with T1 to T2 tumors with 1 to 3 positive nodes who were observed without PMRT, and the overall rate of LRF was 4.3% at a median follow-up time of 70 months. On multivariate analysis, age less than or equal to 35 years, LVI, and ratio of positive nodes greater than 15% were the most important prognostic factors for LRF. Moreover, patients with 2 or 3 of the risk factors mentioned earlier had a LRF rate of 23%, compared with 2.7% among those who had only 1 risk factor. Both these studies highlight the importance of individual clinical-pathologic risk factors when evaluating the benefit of PMRT in patients with 1 to 3 positive nodes.

In 2008, Kyndi and colleagues published an analysis of a cohort of patients from the Danish Breast Cancer Cooperative Group (DBCG) protocol trials 82b and 82c in whom receptor status had been retrospectively analyzed, examining the impact of receptor status on benefit from PMRT. Although a locoregional control benefit was seen in all subgroups with the addition of PMRT, the greatest benefit was seen in patients with estrogen-receptor (ER)–positive disease, and the least in ER-negative disease. Improvement in overall survival with the addition of PMRT was seen in ER-positive and progesterone receptor (PR)–positive disease but not ER-negative and PR-negative disease. This study predated the era of HER2 targeted therapy and is thus not useful in assessing the impact of PMRT in HER2-positive disease in the current era. However, a recent analysis of the National Comprehensive Cancer Center Breast Outcomes Database again showed the highest risk of LRF and the least reduction in LRF with PMRT in the setting of triple-negative disease, as well as a very low risk of LRF in HER2-positive disease in the setting of HER2 targeted therapy with or without PMRT. Triple-negative receptor status in particular has emerged as a significant risk factor for LRF, and although the risk reduction with PMRT is smallest in this group, the benefit remains significant and triple-negative receptor status is generally considered an indication for PMRT in the setting of 1 to 3 positive nodes. With increasing data showing a locoregional control benefit with HER2 targeted therapy, there is no clear consensus on HER2 status as an indication for PMRT.

The most recent randomized data informing our PMRT decisions stems from the recent MA.20 and European Organisation for Research and Treatment of Cancer (EORTC) studies, which assessed the benefit of regional nodal irradiation (RNI) in addition to standard whole-breast irradiation. Although both studies included primarily women undergoing breast-conserving surgery, the benefit of RNI can be extrapolated to PMRT in node-positive patients, given that PMRT has historically included RNI. In MA.20, patients with either node-positive or high-risk node-negative disease were randomized to whole-breast irradiation (WBI) alone or with RNI (including internal mammary nodes, supraclavicular and level III axillary nodes). Most patients (85%) had 1 to 3 positive nodes. Although no difference in overall survival was observed, RNI conferred a significant improvement in both locoregional and distant disease-free survival. At 10 years, the incidence of LRF and distant metastasis (DM), respectively, was 4.5% and 12.9% for the WBI plus RNI group, compared with 7.2% and 16.5% for the WBI-alone group. Similarly, in the EORTC 22922 study reported by Poortmans and colleagues 4004 women with either medially located tumors (regardless of nodal involvement) or tumors in any part of the breast with axillary node positivity were randomized to WBI alone or with RNI. As in MA.20, no overall survival benefit was seen; however, LRF, distant disease-free survival, and breast cancer mortality were all significantly improved. Rates of LRF and DM in the WBI group, respectively, were 9.5% and 19.6%, compared with 8.3% and 15.9% in the WBI plus RNI group. Although these studies primarily included patients treated with breast conservation, and LRF outcomes may therefore not correlate with the PMRT setting, they are relevant because they allow clinicians to extrapolate a potential DM and disease-free survival benefit with PMRT/RNI even in the modern era. Most editorials discussing the impact of these key studies suggest consideration of risk factors and individual patient characteristics in estimating the benefit of PMRT/RNI, but generally support the conclusion that overall recurrence and disease-free survival benefits may be observed in addition to locoregional control in selected patients.

Current National Comprehensive Cancer Network guidelines suggest that, in patients who have undergone total mastectomy with axillary staging and have 1 to 3 positive nodes, PMRT to the chest wall and comprehensive nodal regions should be “strongly considered.” New American Society of Clinical Oncology (ASCO), American Society for Radiation Oncology (ASTRO), Society of Surgical Oncology (SSO) guidelines for PMRT recently addressed this cohort and issued consensus recommendations to evaluate the need for PMRT in a multidisciplinary setting and to consider a multitude of factors, including life expectancy, tumor size, biological characteristics, and plans for systemic therapy, in weighing the risks and benefits of radiation. In addition, smaller institutional series support strong consideration of PMRT in patients with T1, T2 disease and 1 to 3 positive nodes who are also younger (≤50 years), have LVI, triple-negative disease, skin or nipple invasion, or several of these factors.

Another challenge for clinicians that has become more common after the publication of American College of Surgeons Oncology Group (ACOSOG) Z-11 and the AMAROS trial (After mapping of the axilla: radiotherapy or surgery) is determining the role of PMRT in patients who have had a positive sentinel node biopsy without axillary dissection (ALND). In patients who undergo mastectomy and have a positive sentinel node biopsy and do not proceed to ALND, the joint ASCO-ASTRO-SSO panel has recommended to proceed with PMRT only if there is sufficient evidence to recommend adjuvant radiation with the available clinical and pathologic information, and without further information from an ALND. Otherwise, ALND is recommended. There remains substantial debate about which sentinel node–positive patients, if any, can forego both ALND and PMRT, because there are few data to guide clinicians in this area. Data from the International Breast Cancer Study Group 23-01 trial suggest no increase in LRF with the omission of PMRT in the setting of micrometastatic sentinel node–positive disease only, and thus it is reasonable to consider omission of PMRT in this setting, weighing additional risk factors for recurrence in the decision.

PMRT for 1 to 3 positive nodes

The use of PMRT in the setting of 4 or more positive lymph nodes has been broadly accepted because of the high risk of LRF in this population; however, controversy remains regarding the utility of PMRT in patients with 1 to 3 positive nodes. The Early Breast Cancer Trialists’ Collaborative Group (EBCTCG) updated their meta-analysis on the role of PMRT in 2014. The meta-analysis included 22 trials and 8135 patients with breast cancer between 1964 and 1986 who were randomly assigned to receive chest wall and regional nodal radiotherapy after mastectomy and axillary dissection. A subset analysis of 1133 patients with 1 to 3 positive nodes who had received systemic therapy showed a LRF rate of 21% without irradiation and 4.3% with PMRT at 10 years ( P <.001). The 10-year rate for any recurrence, either local or distant, was 45.5% without radiation and 33.8% with radiation ( P <.001). Moreover, breast cancer mortalities were 49.4% and 41.5% ( P = .01) without radiation and with radiation, respectively. These data support the benefit of PMRT in preventing both LRF and overall recurrence, as well as improving breast cancer mortality. Although these numbers are clearly supportive of the benefits of PMRT, most patients included in the EBCTCG analysis were included in randomized trials conducted in the 1970s and 1980s and broadly grouped patients by nodal stage but not other more recently elucidated risk factors for recurrence such as receptor status, age, lymphovascular invasion (LVI), grade, and other pathologic features. In the current era, rates of LRF seem to be considerably lower than the rates mentioned earlier, and are understood to vary substantially with risk factors for recurrence. The trend toward lower rates of LRF in more recent years is highlighted in a study from the MD Anderson Cancer Center in which a cohort of 1027 patients with T1 to T2 breast cancer with 1 to 3 positive lymph nodes treated with mastectomy and systemic therapy with or without PMRT were analyzed by treatment era. Specifically, those treated in an early era (1978–1997) were compared with those treated in a later era (2000–2007). In the early cohort, PMRT was observed to decrease 5-year rates of LRF from 9.5% to 3.4% ( P = .028). However in the later cohort, PMRT did not seem to significantly decrease rates of LRF, and 5-year rates of LRF without PMRT were only 2.8%. Overall in modern series, rates of LRF in patients treated with mastectomy and systemic therapy without radiation generally range from 4% to 23% (with most studies ranging from 4% to 10%) depending on risk factors.

Several contemporary retrospective studies have highlighted the significance of specific risk factors for recurrence in determining the utility of PMRT in patients with 1 to 3 positive nodes. Moo and colleagues published a retrospective analysis of 1331 patients with T1 to T2 tumors with 1 to 3 positive nodes who underwent mastectomy with or without PMRT. The overall rate of LRF in the no-PMRT group was 4.3%. On a multivariate analysis of patients in the no-PMRT group, both age less than or equal to 50 years and lymphovascular invasion (LVI) were significantly associated with increased risk of LRF, suggesting that these factors warrant consideration of PMRT in this cohort. Similarly, Yildirim and Berberoglu published a study of patients with T1 to T2 tumors with 1 to 3 positive nodes who were observed without PMRT, and the overall rate of LRF was 4.3% at a median follow-up time of 70 months. On multivariate analysis, age less than or equal to 35 years, LVI, and ratio of positive nodes greater than 15% were the most important prognostic factors for LRF. Moreover, patients with 2 or 3 of the risk factors mentioned earlier had a LRF rate of 23%, compared with 2.7% among those who had only 1 risk factor. Both these studies highlight the importance of individual clinical-pathologic risk factors when evaluating the benefit of PMRT in patients with 1 to 3 positive nodes.

In 2008, Kyndi and colleagues published an analysis of a cohort of patients from the Danish Breast Cancer Cooperative Group (DBCG) protocol trials 82b and 82c in whom receptor status had been retrospectively analyzed, examining the impact of receptor status on benefit from PMRT. Although a locoregional control benefit was seen in all subgroups with the addition of PMRT, the greatest benefit was seen in patients with estrogen-receptor (ER)–positive disease, and the least in ER-negative disease. Improvement in overall survival with the addition of PMRT was seen in ER-positive and progesterone receptor (PR)–positive disease but not ER-negative and PR-negative disease. This study predated the era of HER2 targeted therapy and is thus not useful in assessing the impact of PMRT in HER2-positive disease in the current era. However, a recent analysis of the National Comprehensive Cancer Center Breast Outcomes Database again showed the highest risk of LRF and the least reduction in LRF with PMRT in the setting of triple-negative disease, as well as a very low risk of LRF in HER2-positive disease in the setting of HER2 targeted therapy with or without PMRT. Triple-negative receptor status in particular has emerged as a significant risk factor for LRF, and although the risk reduction with PMRT is smallest in this group, the benefit remains significant and triple-negative receptor status is generally considered an indication for PMRT in the setting of 1 to 3 positive nodes. With increasing data showing a locoregional control benefit with HER2 targeted therapy, there is no clear consensus on HER2 status as an indication for PMRT.

The most recent randomized data informing our PMRT decisions stems from the recent MA.20 and European Organisation for Research and Treatment of Cancer (EORTC) studies, which assessed the benefit of regional nodal irradiation (RNI) in addition to standard whole-breast irradiation. Although both studies included primarily women undergoing breast-conserving surgery, the benefit of RNI can be extrapolated to PMRT in node-positive patients, given that PMRT has historically included RNI. In MA.20, patients with either node-positive or high-risk node-negative disease were randomized to whole-breast irradiation (WBI) alone or with RNI (including internal mammary nodes, supraclavicular and level III axillary nodes). Most patients (85%) had 1 to 3 positive nodes. Although no difference in overall survival was observed, RNI conferred a significant improvement in both locoregional and distant disease-free survival. At 10 years, the incidence of LRF and distant metastasis (DM), respectively, was 4.5% and 12.9% for the WBI plus RNI group, compared with 7.2% and 16.5% for the WBI-alone group. Similarly, in the EORTC 22922 study reported by Poortmans and colleagues 4004 women with either medially located tumors (regardless of nodal involvement) or tumors in any part of the breast with axillary node positivity were randomized to WBI alone or with RNI. As in MA.20, no overall survival benefit was seen; however, LRF, distant disease-free survival, and breast cancer mortality were all significantly improved. Rates of LRF and DM in the WBI group, respectively, were 9.5% and 19.6%, compared with 8.3% and 15.9% in the WBI plus RNI group. Although these studies primarily included patients treated with breast conservation, and LRF outcomes may therefore not correlate with the PMRT setting, they are relevant because they allow clinicians to extrapolate a potential DM and disease-free survival benefit with PMRT/RNI even in the modern era. Most editorials discussing the impact of these key studies suggest consideration of risk factors and individual patient characteristics in estimating the benefit of PMRT/RNI, but generally support the conclusion that overall recurrence and disease-free survival benefits may be observed in addition to locoregional control in selected patients.

Current National Comprehensive Cancer Network guidelines suggest that, in patients who have undergone total mastectomy with axillary staging and have 1 to 3 positive nodes, PMRT to the chest wall and comprehensive nodal regions should be “strongly considered.” New American Society of Clinical Oncology (ASCO), American Society for Radiation Oncology (ASTRO), Society of Surgical Oncology (SSO) guidelines for PMRT recently addressed this cohort and issued consensus recommendations to evaluate the need for PMRT in a multidisciplinary setting and to consider a multitude of factors, including life expectancy, tumor size, biological characteristics, and plans for systemic therapy, in weighing the risks and benefits of radiation. In addition, smaller institutional series support strong consideration of PMRT in patients with T1, T2 disease and 1 to 3 positive nodes who are also younger (≤50 years), have LVI, triple-negative disease, skin or nipple invasion, or several of these factors.

Another challenge for clinicians that has become more common after the publication of American College of Surgeons Oncology Group (ACOSOG) Z-11 and the AMAROS trial (After mapping of the axilla: radiotherapy or surgery) is determining the role of PMRT in patients who have had a positive sentinel node biopsy without axillary dissection (ALND). In patients who undergo mastectomy and have a positive sentinel node biopsy and do not proceed to ALND, the joint ASCO-ASTRO-SSO panel has recommended to proceed with PMRT only if there is sufficient evidence to recommend adjuvant radiation with the available clinical and pathologic information, and without further information from an ALND. Otherwise, ALND is recommended. There remains substantial debate about which sentinel node–positive patients, if any, can forego both ALND and PMRT, because there are few data to guide clinicians in this area. Data from the International Breast Cancer Study Group 23-01 trial suggest no increase in LRF with the omission of PMRT in the setting of micrometastatic sentinel node–positive disease only, and thus it is reasonable to consider omission of PMRT in this setting, weighing additional risk factors for recurrence in the decision.