The Value of the 8th Edition of American Joint Committee on Cancer Pathological Prognostic Staging on the Selection of Postmastectomy Radiotherapy for T1–2N1 Breast Cancer

Methods and Materials Patients diagnosed with pT1-2N1M0 breast cancer between 2008 and 2018 in West China Hospital, Sichuan University were included. Locoregional-free survival (LRFS), distant metastasis-free survival (DMFS), disease-free survival (DFS), breast cancer-specific survival (BCSS), and overall survival (OS) were defined as endpoints. The propensity score matching (PSM), receiver operating characteristic (ROC) curve, the Kaplan-Meier analysis, and the Cox multivariable model were used for data analysis. Results We identified 1,615 patients with T1-2N1M0 breast cancer, and 44.9% (n = 744) of them were treated with PMRT. With a median follow-up of 76 months, 46 (2.8%) recurrences, 96 (5.9%) deaths, and 80 (5.0%) breast cancer-related deaths occurred. The 5-year LRFS, DMFS, DFS, BCSS, and OS were 98.6%, 95.3%, 93.7%, 96.5%, and 96.0%, respectively. PMRT could not improve 5-year LRFS, DMFS, DFS, BCSS, and OS compared with non-PMRT neither before nor after PSM in the era of contemporary systemic treatment. ROC curve showed that the 8th pathological prognostic staging had better discriminative ability compared with the 7th anatomical staging [the area under the curve (AUC) 0.653 vs. 0.546, P < 0.001]. In the anatomical staging system, PMRT had comparable 5-year BCSS in comparison with non-PMRT both in stages IIA (97.4% vs. 96.8%, P = 0.799) and IIB (95.3% vs. 97.0%, P = 0.071). When stratified according to the pathological staging, PMRT was associated with better 5-year BCSS in stage IIB (97.1% vs. 90.7%, P = 0.039), while not in stages IA, IB, IIA, and IIIA. Multivariate analysis demonstrated that PMRT was a significantly protective factor for BCSS in stage IIB (HR 0.331, 95% CI: 0.100-0.967, P = 0.044). Conclusion The new staging could better select high-risk patients with T1-2N1 breast cancer for radiotherapy compared with the 7th staging, and PMRT might be exempted except the 8th staging of IIB in the era of contemporary systemic therapy in this disease.


Introduction
Breast cancer ranks first in the prevalence of cancer among female patients and is the secondary cause of cancer-related mortality in the basis of 2021 prediction [1]. Patients with tumor size less than 5 cm and one to three involved axillary lymph nodes (LNs) (T1-2N1) are of the most interest to researchers among all cancer stages [2][3][4][5][6]. Surgery and chemotherapy are the cornerstone of this subtype; however, the effect of radiotherapy (RT) has not been well elaborated. Current clinical guidelines, such as the National Comprehensive Cancer Network guideline, strongly recommend postmastectomy radiotherapy (PMRT) for this patient subset [3]. The basis of their recommendation is derived from the result of the 2014 metaanalysis published in the Lancet that PMRT had significantly decreased locoregional recurrence (LRR) rate and cancer-related death in patients with 1-3 involved LNs [4]. However, most of the trials enrolled in this metaanalysis were completed before 1980s, when the RT technique and chemotherapy regimens were far from what it is now, and the LRR rate for the patients without PMRT was up to 30% at that time [5,6]. Therefore, the value of PMRT in T1-2N1 breast cancer should be revalidated in the era of contemporary systematic treatment.
The traditional anatomical staging system has been extensively utilized for predicting prognosis and making treatment decision in breast cancer [7,8]. However, the survival significance caused by molecular biomarker, such as estrogen receptor (ER), progesterone receptor (PR), and human epidermal growth factor-2 (HER2) make the anatomical staging unable to meet the trend of individualized treatment nowadays [9,10]. Therefore, the 8th edition of American Joint Committee on Cancer (AJCC) pathological staging integrates ER, PR, HER2, and pathological grade into the anatomical staging, which exhibited better value of prognostic prediction in breast cancer [11,12]. In the 7th AJCC staging, T1-2N1 breast cancer is divided into two substages (IIA and IIB), while the 8th AJCC staging classifies this group into five subtypes (IA-IIIA) [8,13]. However, there were fewer studies exploring the effect of PMRT for T1-2N1 breast cancer patients in the basis of new pathological staging.
Therefore, we aimed to evaluate the effect of PMRT in the era of contemporary systematic treatment in T1-2N1 breast cancer. In addition, we further explored the value of prognostic prediction and RT administration decision of the 8th AJCC staging system in T1-2N1 breast cancer.

Materials and Methods
2.1. Patients. The patients were enrolled from Breast Cancer Information Management System (BCIMS) of West China Hospital, Sichuan University, which prospectively collects information on medical history, patient demographic, tumor characteristics, accessory examination, treatment, and follow-up, dating back to 1989. Patients meeting the following criterions were identified: (1) diagnosed between Jan 2008 and Dec 2018; (2) receiving radical or modified radical mastectomy; (3) pathologically diagnosed with pT1-2N1M0 invasive breast cancer; (4) receiving adjuvant chemotherapy; and (5) having detailed information on age, menopausal status, histological subtype, tumor stage, number of involved axillary LNs, ER status, PR status, HER2 status, pathological grade, the 7th AJCC anatomical stage, the 8th AJCC pathological stage, RT administration, endocrine therapy, anti-HER2 therapy, and follow-up. Male patients and patients aged ≤18 years, with bilateral breast cancer, and treated with neoadjuvant chemotherapy were excluded. The selection flowchart of the study population is presented in Figure 1.
The study was approved by the Biomedical Ethics Committee (Approval number: 2020427) of West China Hospital, Sichuan University. All patients in our database signed informed consent at initial diagnosis.
2.2. Treatments, Follow-Up, and Endpoints. All patients received radical mastectomy or modified radical mastectomy and adjuvant chemotherapy. Adjuvant chemotherapy protocols included anthracycline-based or taxane-based regimens and other regimens, such as cyclophosphamide, methotrexate, and fluorouracil (CMF) regimen, and oral chemotherapeutic drugs (capecitabine, S-1). RT administration was negotiated by the radiotherapy physicians and patients using three-dimensional conformal radiotherapy or intensity modulated radiotherapy techniques, with a prescribed dose   Journal of Oncology of 46-50 grays in 25 fractions to the ipsilateral chest wall and infra-and supraclavicular regions. When tumors located in the inner side or central area of the breast, internal mammary LN region was included. Patients with hormone positive and HER2 overexpression received adjuvant endocrine therapy and anti-HER2 therapy, respectively, if economic conditions permitted. Follow-up information was collected by telephone calls, office visits, and emails. Follow-up protocols were as follows: once every 4 months in the first 3 years, once every six months during 4-5 years, and every year after 5 years. The primary endpoint was breast cancer-specific survival (BCSS), calculated as the periods from surgery to the death of breast cancer. The secondary endpoints were locoregional-free survival (LRFS), calculated as the months from the start of surgery to the recurrence of ipsilateral chest wall, axillary, infra-or supraclavicular regions, or internal mammary LNs; distant metastasis-free survival (DMFS), from surgery to the metastasis of distant organs; diseasefree survival (DFS), from surgical treatment to recurrence or metastasis or second primary breast cancer or any cause of death; and overall survival (OS), from surgery to any cause of death.

Statistical Analysis.
We used the Chi-squared test to compare the difference of patient clinicopathological characteristics between the PMRT and non-PMRT groups. Propensity score matching (PSM) (1 : 1 nearest neighbor matching) was used to balance the patient characteristics including age at diagnosis, menopausal status, histological subtype, T stage, number of involved axillary LNs, ER, PR, HER2, grade, the 7th stage, and the 8th stage between the two groups. Kaplan-Meier analysis was utilized to calculate the differences and draw the survival curves of LRFS, BCSS, DMFS, DFS, and OS in patients receiving and not receiving PMRT. Cox multivariate hazards regression model was applied to identify the predictors of LRFS, BCSS, DMFS, DFS, and OS. The receiver operating characteristics (ROC) curve was utilized to discriminate the ability of predicting survival between the 7th and 8th AJCC staging. IBM SPSS 22.0 was used for analyzing and mapping. A P value <0.05 (two-tail) was considered to be statistically significant.

Discussion
In this study, we aimed to evaluate the value of PMRT in pT1-2N1M0 breast cancer patients with contemporary systematic treatment and further explored the guiding value of the anatomical staging and pathological prognostic staging systems on PMRT. Result demonstrated that PMRT could not improve 5-year LRFS, DMFS, DFS, BCSS, and OS compared with non-PMRT. The 8th AJCC staging could better predict the prognosis and guide the RT administration that patients with stage IIB could benefit from PMRT, while those with stages IA, IB, IIA, and IIIA could not.  Journal of Oncology The value of PMRT in T1-2N1 breast cancer has always been controversial. The large-scale meta-analysis in 2014 showed incremental benefit of PMRT on locoregional recurrence and breast cancer mortality, and the use of PMRT increased from 30.6% to 47.1% over years in T1-2N1 breast cancer patients [4,14]. However, recent studies found that PMRT could not improve survival outcomes with modern systematic therapy [14][15][16][17][18]. Our result was consistent with theirs that patients receiving PMRT had no benefit on LRFS, DMFS, DFS, BCSS, and OS compared with those not receiving PMRT. The LRR rate of this study in the absence of PMRT was 3.3%, which was significantly lower than the studies included in 2014 meta-analysis (30%) [4,19,20]. This probably attributed to the progress of surgical techniques and modern chemotherapy regimens and anti-HER2 drugs [16]. In our study, 99.3% of the patients received taxane-based and/or anthracycline-based chemotherapy regimens, and only four patients received CMF regimen. Previous studies had demonstrated taxane-based and/ or anthracycline-based regimens had better survival benefits compared with CMF regimen [16,17]. In addition, all the patients in this study had negative surgical margins, and nearly half of the patients with HER2 overexpression received targeted therapy. Therefore, the administration of PMRT for T1-2N1 patients should be taken a cautious approach with such a lower LRR rate of 3.3% in the epoch of excellent systematic treatment. Let us look forward to the results of contemporary SUPREMO trial, an ongoing large-scale randomized controlled trial exploring the availability of PMRT in T1-2N1 breast cancer [21].
Previous studies have proven that the 8th AJCC staging had better prognostic value compared with anatomical staging in breast cancer [11,12,22]. However, few studies have evaluated the value of the new staging in predicting prognosis in T1-2N1 breast cancer, a disease with significant het-erogeneity of clinical and histopathological behavior. In this study, the 8th AJCC pathological prognostic staging showed better discriminative value compared with the 7th staging in this disease (P < 0:001). However, patients with stage IIB had better survival than those with stage IIA in the 8th staging, which was different than the common knowledge and similar study from the Surveillance, Epidemiology, and End Results (SEER) database [23]. The possible reason was that racial disparities and the addition of molecular markers and pathological grade made the patients with stage IIB a better survival than those with stage IIA. Nonetheless, the new pathological staging indeed exhibited the heterogeneity of stage T1-2N1 disease and provided more precise prognostic information compared with anatomical staging.
Several recent studies, including our result, had demonstrated that PMRT showed no incremental survival benefit in T1-2N1 breast cancer patients receiving perfect contemporary systemic therapy [14][15][16][17][18]. Then, the selection of patients with high risk for PMRT in this highly heterogeneous disease appears to be essential. The 8th AJCC staging combining anatomical and pathological factors provides better choice of high-risk patients for PMRT [23,24]. In our study, PMRT was not associated with improved survival in all stages of the 7th staging compared with non-PMRT in T1-2N1 breast cancer, while PMRT could improve BCSS in the new staging of IIB in this disease. Therefore, the new pathological prognostic staging had better value of selecting high-risk patients for receiving PMRT than the 7th staging. It is noteworthy that the 8th staging of IIB could benefit from PMRT, while stage IIIA could not in this study, which is inconsistent with the result that the higher stage is more likely to benefit from PMRT [25]. The possible explanation was that patients with stage IIIA were all triple negative breast cancer (ER, PR, and HER2) in our study population,  Journal of Oncology which is highly invasive and more prone to have distant metastasis than other molecular types, and the value of locoregional RT cannot be reflected [26,27]. Therefore, the administration of PMRT should be evaluated according to tumor and biologic characteristics. Our study recommended that PMRT could be administrated in stage IIB, while discreetly carried out in stage IIIA in view of new staging.
This study has several limitations that should be acknowledged. Firstly, this study is a retrospective study, and the intrinsic defects, such as selection bias and inconsistency, should not be neglected. Secondly, the baseline characteristics and treatment information between the PMRT and non-PMRT groups were not balanced, and patients receiving PMRT were more likely to have worse baseline characteristics, which could not represent the whole T1-2N1 breast cancer patients. Thirdly, the follow-up time of our study was short (medium: 76 months; range: 4-164 months), and long period is needed to validate the effect of PMRT in this disease. Finally, there were fewer patients with the 8th staging of IIIA, and more patients should be enrolled to evaluate the value of PMRT in this stage. Although this study has some limitations, we utilized new pathological staging to select the patients who need to receive PMRT in T1-2N1 breast cancer, which is in line with the trend of personalized treatment and reduce the economic burden for patients.

Conclusion
In conclusion, the 8th AJCC pathological prognostic staging had better value of guiding RT administration than the 7th anatomical staging in T1-2N1 breast cancer. PMRT might be exempted in the era of contemporary systemic therapy, except the 8th staging of IIB in this disease. Further studies should be conducted to evaluate the value of PMRT in T1-2N1 breast cancer.

Data Availability
The raw data supporting the conclusions of this article is available from the corresponding author on reasonable request.