Introduction
Postmastectomy radiotherapy (PMRT) plays an important role in the management of locoregionally advanced breast cancer. Multiple clinical trials have shown treatment benefits of PMRT in locoregional control, disease-free survival, and overall survival (OS) in patients with four or more positive axillary lymph nodes (LN) (N2 or higher) or T3–4 disease [1,2]. Despite the publication of the American Society of Clinical Oncology (ASCO) guideline and other similar guidelines over two decades ago, the benefit of PMRT in patients with a tumor size of 5 cm or smaller and metastasis to 1–3 axillary LN(s) (T1–2N1) remains unclear [3-6]. It is very challenging to determine the usefulness of PMRT since no randomized study has been conducted to assess the therapeutic value of PMRT for T1–2N1 breast cancer patients, and the patient characteristics between the PMRT and no-PMRT groups were not comparable in the retrospective studies. Most guidelines and consensus statements published in the 2000s indicate that there is insufficient evidence to support the routine use of PMRT for treating T1–2N1 disease and recommend its use exclusively for patients with poor prognostic factors [7-9]. Therefore, in the previously published studies, PMRT was performed according to each physician’s preference and institutional policy [10-13]. In Korea, there is no consensus of PMRT for T1–2N1 breast cancer. Nationwide surveys conducted by the Korean Radiation Oncology Group (KROG) have shown considerable variations in patterns of PMRT [14,15].
This study was to investigate practice patterns of PMRT for T1–2N1 breast cancer and longitudinal changes over 2 decades (1990, 2000, and 2010) in Korea, enabling us to establish the future treatment guidelines for T1–2N1 breast cancer.
Materials and Methods
1. Patients
The KROG authorized this study protocol (protocol No. KROG 14-21). Using three independent patient cohorts from the years of 1990, 2000, and 2010, we collected the data of T1–2N1 breast cancer patients who were treated with mastectomy at 15 institutions in Korea. The institutional review board of each institution approved this study. Patients who were pathologically diagnosed with T1–2N1 stage breast cancer were considered eligible for this study. Patients were excluded from the study if they met the following criteria: (1) had received neoadjuvant chemotherapy, (2) had a past history of other cancers except for thyroid/skin/uterine cervix cancer and distant metastasis, or (3) were male. Finally, a total of 700 selected patients were reviewed retrospectively.
2. Data collection
Patients’ data related to the following pathologic features were obtained: pathological type, nuclear grade, histological grade, tumor size, resection margin, lympho-vascular invasion, status of hormone receptors, HER2/neu status, Ki-67 expression, number and size of positive axillary LN, and extracapsular extension of axillary LN, the recurrence pattern, and the date and cause of death. Treatment information including surgical technique used for axillary LN, radiotherapy (RT), adjuvant chemotherapy, and hormonal therapy was carefully reviewed.
3. Statistical analysis
The medians and distributions were compared using the chi-square and Student’s t-tests. The OS and recurrence-free survival (RFS) were estimated using the Kaplan-Meier method. Multivariate logistic regression analyses were used to assess the correlation between the variables and survival. A p-value <0.05 was considered statistically significant. All statistical analyses were performed using SPSS version 25 (IBM Corp., Armonk, NY, USA).
Results
1. Patients’ characteristics
The patients and their tumor characteristics stratified by the study year are presented in Table 1. The numbers of patients receiving mastectomy in the years 1990, 2000, and 2010 were 54, 251, and 395, respectively. The median age was 48 (interquartile range [IQR], 42 to 56) of all patients. In 1990, the proportions of premenopausal and postmenopausal patients were similar, whereas the proportion of premenopausal patients increased in 2000 and 2010. The proportion of patients with more advanced features, such as T2 disease, three positive LNs, or LN size of ≥10 mm, significantly decreased in recent years. While the molecular subtypes in 2000 and 2010 were comparable, 1990 showed a notable difference due to a higher proportion of cases categorized as not available.
Axillary LN dissection was performed in all patients. Sentinel lymph node biopsy (SLNB) was not adopted in the 1990 patient cohort. SLNB was conducted in 6.3% and 73.1% of the patients from the 2000 and 2010 cohorts, respectively. Number of dissected LNs has decreased over time. Adjuvant chemotherapy was given to 83.3%, 95.6%, and 91.9% of patients in 1990, 2000, and 2010, respectively. Majority of the patients (95.6%) received CMF (cyclophosphamide, methotrexate, and 5-fluorouracil) in 1990; however, the use of CMF reduced significantly thereafter. Anthracycline-based chemotherapy was performed in 38.8% and 93.4% of the patients in 2000 and 2010, respectively. Use of hormone therapy increased over the study years. The test for HER2/neu status was performed in only one patient from the patient cohort of 1990; however, its trend increased significantly during the subsequent study years. Seventy-eight of 175 patients (44.6%) with a positive HER2/neu status received treatment with trastuzumab.
2. Postmastectomy radiotherapy
The utilization of PMRT showed a slight increase over time. Of 700 patients, 70 patients (10.0%) received PMRT, which included 7.4%, 8.8%, and 11.1% of the patients from 1990, 2000, and 2010, respectively. However, this increase was not statistically significant. The use of PMRT varied substantially among the hospitals, ranging from 0%–87.5%. Three of the 16 institutions did not perform PMRT throughout the entire study period. Patients from these three institutions exhibit various risk factors including close resection margin, positive LN size of ≥2 cm, and high histologic grade, among others.
There were significant differences observed between the PMRT and no-PMRT subgroups. Patients who received PMRT were more likely to exhibit high-risk features such as three positive LNs (p = 0.015), positive LN size of ≥10 mm (p = 0.002), positive or close resection margin (p = 0.006), high grade (p = 0.036), extensive intraductal component (p = 0.005), and high levels of Ki-67 expression (p = 0.009) (Table 2).
RT technique stratified by the study year is listed in Table 3. For chest wall irradiation, the reverse hockey stick technique was exclusively used in 1990, but the adoption of tangential beam three-dimensional conformal RT increased in subsequent decades. Chest wall boost was performed in about 28% of the patients in the years 2000 and 2010, and the median boost dose was 9 Gy (range, 8 to 10.8 Gy). While axillary irradiation was administered to all patients in 1990, its use became selective in subsequent years, with 72.7% of patients receiving it in 2000 and 84.1% in 2010. Approximately one-quarter of the patients received a posterior axillary boost, which remained consistent throughout the study periods. The inclusion of axillary LN levels, internal mammary nodes, and supraclavicular nodes decreased over time. The most common radiation dose used was 50.4 Gy (range, 45 to 60 Gy), with a daily dose of 1.8 Gy. Hypofractionated radiation scheme (45 Gy in 15 fractions) was performed for only one patient in 2010.
3. Treatment outcomes
The median follow-up was 58.5 months (IQR, 52.3 to 121.2). The median follow-up for the 1990, 2000, and 2010 cohorts was 96.4 months (IQR, 60.8 to 161.6), 137.5 months (IQR, 83.7 to 151.1), and 54.4 months (IQR, 50.2 to 58.2), respectively. The 10-year OS, RFS, locoregional recurrence–free survival (LRRFS), and distant metastasis–free survival (DMFS) were 81.3%, 76.3%, 89.7%, and 80.0%, respectively. In the 1990 cohort, the pathological and molecular data were not available for a relatively large number of patients and most of patients received CMF chemotherapy, rather than anthracycline or taxane-based chemotherapy. In addition, in the 2000 cohort, all patients with HER2/neu (+) did not receive trastuzumab. Therefore, we conducted a survival analysis by era (Table 4). Except for the 10-year OS, all other treatment outcomes demonstrated improvement over the study period.
The available follow-up data for 700 patients revealed that 116 patients (16.6%) experienced recurrence. Recurrence at locoregional and distant sites decreased significantly over the study years: 24 (44.4%), 59 (23.5%), and 33 (8.4%) patients in 1990, 2000, and 2010, respectively (p < 0.001). The first recurrence site was locoregional recurrence (LRR) in 29 (25.0%), distant metastasis (DM) in 66 (56.9%), and LRR and DM in 21 (18.1%). Nine (12.9%) of 70 patients with PMRT and 107 (17.0%) of 630 patients without PMRT exhibited recurrences. Of nine patients with recurrences in PMRT group, LRR at their first recurrence was observed in one patient (1.4%). The LRR site was supraclavicular LN. In patients without PMRT, LRR at their first recurrence was present in 49 patients (7.8%). LRR was significantly lower in the PMRT group compared to the no-PMRT group (p = 0.043). In the PMRT group, 13 patients (18.6%) did not receive axillary irradiation. Among them, four patients had recurrence as DM (1 bone metastasis, 1 liver metastasis, and 2 lung metastases). There was no LRR. Detailed information on the recurrence sites in the PMRT group and the no-PMRT group is provided in Supplementary Table 1.
The 10-year LRRFS showed a trend toward improvement in the PMRT group compared to the no-PMRT group (98.6% vs. 88.8%, p = 0.055). However, there was no significant difference in 10-year OS, RFS, or DMFS between PMRT and no-PMRT groups (Table 5). Given the substantial differences in treatment approaches between patients in the 1990s and 2000s and those in more recent years, we further analyzed the impact of PMRT within the 2010 cohort. In 2010 cohort, PMRT was not associated with a significant survival benefit (Supplementary Table 2).
In the current study, chemotherapy regimen was related to survival outcomes of the participating patients. Anthracycline-based chemotherapy significantly reduced the rates of LRR and DM compared to CMF chemotherapy (p < 0.001 and p = 0.001) (Table 6). Additionally, it led to significant improvements in 10-year OS, RFS, and LRRFS (p < 0.001, p = 0.003, and p < 0.001). Of 175 patients with positive HER2/neu status, treatment with trastuzumab improved 10-year RFS (94.7% vs. 66.2%, p = 0.040). There was no significant difference in 10-year OS for trastuzumab and no-trastuzumab groups (93.3% vs. 86.5%, p = 0.609).
Treatment-related complications were analyzed separately in the PMRT group and the no-PMRT group. Among the 70 patients who received PMRT, PMRT-related complications were evaluable in 63 patients: three patients (75.0%) in 1990, 20 patients (90.9%) in 2000, and 40 patients (90.9%) in 2010. In the 1990 cohort, PMRT-related complications was not observed. In the 2000 cohort, one patient (4.5%) developed grade 3 lymphedema. In the 2010 cohort, five patients (11.4%) experienced lymphedema (three with grade 1, while the remaining two had no reported grade information), and two patients developed grade 1 radiation pneumonitis. The most common PMRT-related complication was lymphedema. Axillary LN dissection, the number of dissected LNs, axillary LN irradiation, and the irradiated axillary node level were not associated with the development of lymphedema (p = 0.999, p = 0.997, p = 0.999, and p = 0.995, respectively).
In the no-PMRT group, treatment-related complications were evaluable in 398 patients: 39 patients (78.0%) in 1990, 125 patients (54.6%) in 2000, and 234 patients (66.7%) in 2010. No treatment-related complications were reported in the 1990 cohort. In the 2000 cohort, nine patients (7.2%) experienced treatment-related complications, including lymphedema in six patients (one with grade 2 and five with grade 1), neutropenia in three patients (one with grade 3, one with grade 2, and one with grade 1), and grade 3 nausea in one patient. In the 2010 cohort, treatment-related complications occurred in 30 patients (12.8%), including lymphedema in 23 patients (three with grade 2, 18 with grade 1, and two with no reported grade information), neutropenia in two patients (one with grade 1 and one with no reported grade information), cardiac complications associated with herceptin in two patients (both grade 2), web syndrome (grade 1) in one patient, pulmonary fibrosis in one patient (no reported grade information), and chest wall swelling in one patient (no reported grade information). The most common treatment–related complication was lymphedema. Axillary LN dissection and the number of dissected LNs were not associated with the development of lymphedema (p = 0.306 and p = 0.998, respectively).
Discussion and Conclusion
We investigated the use of PMRT in patients with T1–2N1 breast cancer treated with mastectomy at 16 institutions in Korea over 2 decades. The number of patients for this study markedly increased in recent years, which represented the rise of incidence of breast cancer in Korea. PMRT was performed in about 10% of patients in the entire study population. The proportion of patients who received PMRT did not change significantly over 2 decades; however, PMRT utilization was profoundly varied (ranging from 0%–87.5%) among the participating institutions. PMRT was preferred for the patients with high-risk factors according to each institution’s criteria. Radiation field and technique were different among the radiation oncologists. The majority of patients received anthracycline-based chemotherapy (66.8%) or CMF (27.9%). Despite the presence of high-risk features in patients receiving PMRT, 10-year LRRFS was higher in the PMRT group.
The benefits of PMRT in patients with T1–2N1 breast cancer following mastectomy remain controversial because of undetermined significance on the patient’s survival and potential concerns of morbidity [4,6,16]. Several challenges complicate the assessment of PMRT outcomes. (1) Including this study, the patients who received PMRT were more likely to exhibit high-risk features; however, the definition of high-risk features varies across institutions, and the institutional criteria for administering PMRT in patients with a limited number of involved LNs remain unstandardized. (2) RT field and technique vary significantly institutions. While some institutions irradiate the chest wall and comprehensive LN area (axillary, supraclavicular, and internal mammary LN), others limit the treatment field to the chest wall alone or the chest wall with select LNs area. (3) Breast cancer is a disease with a long natural history. Based on Early Breast Cancer Trialists' Collaborative Group review on the effect of PMRT in the patients with N1 disease, who were followed up for 20 years, LRR in patients with N1 disease was shown to increase steadily for 10 years after mastectomy and RT reduced both recurrence and breast cancer mortality [17]. They reported that approximately one breast cancer-related death was avoided in the 20 years after RT for every 1.5 recurrences of any type (either locoregional or distant) avoided during the first 10 years after RT. Therefore, it is necessary to long-term follow-up for more than 10 years to determine the beneficial effects of PMRT on the survival outcomes in the patients with T1–2N1 breast cancer. Furthermore, a large number of subjects is required to draw a meaningful conclusion in the era of modern multimodality therapy.
Lately, the proportion of patients with more advanced features, such as T2 disease, three positive LNs, or LN size of ≥10 mm, have decreased significantly, and increased use of anthracycline-based chemotherapy over CMF has yielded favorable results. It has been shown that the emerging treatment strategies for breast cancer, such as taxane-based chemotherapy, endocrine therapy, and immune-oncologic agents tailored to subtype are associated with superior and more effective results [10,18]; therefore, the effect of PMRT on survival might be expected to decrease in the modern multimodal treatment era. Consistent with this, PMRT did not provide any survival benefit in the 2010 cohort (Supplementary Table 2).
In order to determine the patient groups that can truly benefit from RT, several investigators have attempted to identify the risk factors associated with the recurrence after mastectomy in the patients who did not receive PMRT. Young age and three positive LNs were associated with a significantly high risk of LRR, implying that PMRT may also be favorable to this group [11,19-21]. Furthermore, there might be patients with molecular features that are at high risks of LRR after mastectomy, and they may benefit from PMRT. The 21-gene recurrence score (RS) assay has been validated as an independent prognosticator of LRR, distant recurrence, and OS in women with node-positive estrogen receptor (ER)–positive breast cancer [22-24]. The 21-gene RS assay may therefore be useful as a predictive marker for potential OS benefit from PMRT in women with T1–2N1 ER-positive breast cancer. Now, the Canadian Tailor RT trial (CCTG MA.39, NCT03488693) is underway on women with T1–2N1 disease and documented low RS who underwent breast-conserving surgery or mastectomy randomizing to either regional radiation or no regional radiation, with breast cancer recurrence-free interval as the primary objective. This trial can provide clarity regarding the value of RS as a predictor of the value of RT in women with T1–2 N1 breast cancer.
In summary, PMRT utilization in patients with T1–2N1 breast cancer following mastectomy did not change significantly over two decades in Korea. Despite the presence of high-risk features in those receiving PMRT, there was an improvement in the 10-year LRRFS of the PMRT group. Prospective randomized study with a large number of patients and long-term follow-up is necessary to determine the impact of PMRT on the survival outcomes and for the selection of the patients who may benefit from PMRT. Further research is needed to reflect the latest patterns of practice.
PDF Links
PubReader
ePub Link
XML Download
Full text via DOI
Download Citation 
Print 
