Postmastectomy Radiotherapy After Neoadjuvant Chemotherapy in cT1-2N+ Breast Cancer Patients: A Single Center Experience and Review of Current Literature

Purpose Postmastectomy radiotherapy (PMRT) after neoadjuvant chemotherapy (NAC) in breast cancer patients with initial clinical stage cT1-2N+, especially for those who achieved ypT1-2N0, is still controversial. This study was to evaluate the survival prognosis of cT1-2N+ patients after NAC with or without PMRT, and to discuss the selection of patients who may omit PMRT. Patients and Methods From January 2005 to December 2017, 3055 female breast cancer patients underwent mastectomy in our medical center, among whom 215 patients of cT1-2N+ stage, receiving NAC with or without PMRT were finally analyzed. The median follow-up duration was 72.6 months. The primary endpoint was disease-free survival (DFS), and secondary endpoint was overall survival (OS). Comparison was conducted between PMRT and non-PMRT subgroups. Results Of the 215 eligible patients, 35.8% (77/215) cT1-2N+ patients achieved ypT0-2N0 after NAC while 64.2% (138/215) of the patients remained nodal positive (ypT0-2N+). The 5-year DFS of ypT0-2N0 non-PMRT was 79.5% (95% confidence interval [CI] 63.4-95.6%). No statistically significant difference was observed between the ypT0-2N0 PMRT and non-PMRT subgroups for the 5-year DFS (78.5% vs 79.5%, p = 0.673) and OS (88.8% vs 90.8%, p = 0.721). The 5-years DFS didn’t obviously differ between the ypT0-2N0 non-PMRT subgroup and cT1-2N0 subgroup (79.5% vs 93.3%, p = 0.070). By using Cox regression model in multivariate analyses of prognosis in ypT0-2N+ PMRT subgroup, HER2 overexpression and triple-negative breast cancer were significantly poor predictors of DFS and OS, while ypN stage was significant independent predictors of OS. Conclusion An effective response to NAC (ypT0-2N0) indicates a sufficiently favorable prognosis, and PMRT might be omitted for cT1-2N+ breast cancer patients with ypT0-2N0 after NAC.


INTRODUCTION
Neoadjuvant chemotherapy (NAC) is increasingly used to treat patients with breast cancer, especially those with locally advanced disease (1). NAC can facilitate surgery by converting an inoperable tumor to an operable tumor or by converting a patient requiring mastectomy to one who can be treated with breast-conserving surgery (2). NAC usually alters tumor stage and has been found to downsize primary tumors in 70-80% of patients (3,4). In addition, NAC was reported to downstage axillary lymph nodes status in 20-40% of patients (4,5).
Postmastectomy radiotherapy (PMRT) has been shown to reduce the risk of locoregional recurrence and enhance overall survival in patients with breast cancer. PMRT is recommended for patients with tumors ≥ 5 cm size or those with at least four positive lymph nodes (6,7). In contrast, the role of PMRT in T 1-2 tumors with 1-3 positive lymph nodes remain unclear, with this treatment usually recommended for patients with T 1-2 N 1 tumors and high-risk features. These guidelines for PMRT guidance, however, were formulated for patients in the absence of NAC, making the indications for PMRT unclear for patients who receive NAC. Moreover, the potential downstaging associated with NAC can alter the standard indications for PMRT.
Indications for PMRT following NAC remain unclear, particularly in patients initially diagnosed with stage cT 1-2 N + breast tumors. Because increasing numbers of patients undergo breast reconstruction after surgery, determining indications for PMRT has become more important, as PMRT can adversely affect the incidence of complications and aesthetic outcomes of immediate breast reconstructions (8). Moreover, it has not yet been determined whether an effective response to NAC (e.g., achieve ypT 0-2 N 0 ) can allow the omission of PMRT.
The present study evaluated the efficacy of PMRT after NAC and surgery in breast cancer patients with initial clinical stage cT 1-2 N + M 0 . Three questions were addressed: 1) whether patients presenting with cT 1-2 N + disease who achieve ypT 0-2 N 0 after NAC require PMRT; 2) the comparative prognosis in the absence of PMRT of patients with cT 1-2 N + and cT 1-2 N 0 who achieve ypT 0-2 N 0 after NAC; and 3) the correlations between clinical variables and prognosis in patients with residual nodal disease after NAC. These findings may help determine the indications for PMRT after NAC.

Patient Population
Of the 3055 women diagnosed with breast cancer who underwent mastectomy at the Second Affiliated Hospital, Zhejiang University School of Medicine from January 2005 to December 2017, 456 (14.9%) received NAC before mastectomy. Patients who received fewer than two cycles of NAC (n=5), those with ypT 3-4 stage after NAC (n=6), those with undetermined ypT stage (n=19) and ypN stage (n=1), and those with unknown radiotherapy (n=62) were excluded. This retrospective analysis therefore included 215 patients diagnosed with cT 1-2 N + stage breast before NAC followed by mastectomy. The study protocol was approved by the Ethics Committee of our hospital (Approval No: 2020-363).
The medical records of all included patients were extracted from the computerized database of the Second Affiliated Hospital, Zhejiang University School of Medicine. Follow-up information on all patients was obtained by clinic visits and telephone contact. Clinical tumor size (cT) was determined by ultrasound, mammography or magnetic resonance imaging. Patients classified as cN + were defined as those with clinically diagnosed metastatic lymph nodes, including palpable lymph nodes that were fixed or matted, metastatic lymph nodes diagnosed by imaging, or lymph node metastases pathologically confirmed after biopsy. Of the 215 cT 1-2 N + patients in the study cohort, 138 (64.2%) were pathologically confirmed as having lymph node metastases, either by biopsy before NAC or by intraoperative sentinel lymph node biopsy or axillary lymph node dissection. TNM staging was performed in accordance with American Joint Committee on Cancer guidelines (version 8,2017). Estrogen receptor (ER) and progesterone receptor (PR) status were evaluated by immunohistochemistry (IHC), with cutoffs of 1% to dichotomize tumors as positive or negative (9). Human epidermal receptor 2 (HER2) status was evaluated by IHC, with tumors having IHC scores of 2+, defined as indeterminate, further analyzed by IHC and fluorescence in situ hybridization (FISH) (10).

Treatment
The patients with cT 1-2 N + breast cancer were divided into two groups, the ypT 0-2 N + and the ypT 0-2 N 0 groups, based on the pathological lymph node status of surgical specimens removed after NAC. Each of these groups was divided into two subgroups, those who did and did not receive PMRT ( Figure 1). Of the patients in this study, 97.2% underwent axillary lymph node dissection. All hormone receptor-positive patients received adjuvant endocrine therapy, and 47 (49.5%) of the 95 patients with HER2-positive breast cancer were treated with Trastuzumab.
In patients undergoing PMRT, 50 Gy (range 45-50 Gy) radiation in 25 fractions (range 25-28) of 2 Gy each was delivered to the chest wall and/or regional lymph nodes (i.e., the supraclavicular/infraclavicular and/or internal mammary lymph node regions). Of the patients who received PMRT, 86.4% received intensity-modulated radiation therapy and 13.6% received 3D conformal radiation therapy. The chest wall and supraclavicular/infraclavicular nodal regions were generally irradiated with photons, and the internal mammary lymph node regions were irradiated with electrons. locoregional recurrence (LR), distant metastasis (DM), new primary tumors in the breast, death from any cause, or last follow-up (11). The secondary endpoint was overall survival (OS), defined as the time from the date of diagnosis to the date of death due from any cause or last follow-up (11). LR included clinical, radiographic, or pathological evidence of recurrence in the ipsilateral chest wall and/ or regional lymph nodes, whereas DM included recurrences at all other sites (i.e., contralateral breast, bone, lung, liver, or brain).

Statistical Analysis
Categorical variables in two groups of patients were compared using the Pearson's c 2 test or Fisher's exact test, as appropriate, whereas continuous variables were compared by t-test. Survival, including DFS and OS, was analyzed by the Kaplan-Meier method, with differences compared by log-rank tests. Patients who were lost to follow-up were censored at the last contact. Factors associated with survival in NAC patients with residual nodal disease were assessed by univariate and multivariate Cox proportional hazards analyses, with variables found to be significant (P<0.10) on univariate analysis were included in the multivariate model. ER status and triple-negative breast cancer were excluded from the multivariate model due to their collinearity with molecular subtypes. All tests were two-sided, with p values < 0.05 considered statistically significant. Statistical analyses were performed using SPSS version 20.0 software (IBM Institute, Chicago, IL, USA) and GraphPad Prism version 8 software (GraphPad Inc, San Diego, CA, USA).
Eleven (22%) of the 50 patients in the ypT 0-2 N 0 PMRT subgroup and seven (25.9%) of the 27 patients in the ypT 0-2 N 0 FIGURE 1 | Study design. NAC, neoadjuvant chemotherapy; PMRT, postmastectomy radiotherapy; cT, clinical tumor size; cN, clinical lymph node; ypT, pathologic tumor size after neoadjuvant therapy; ypN, pathologic lymph node after neoadjuvant therapy. non-PMRT subgroup had LR or DM. Analysis of the recurrence patterns in these subgroups showed that the LR rate was significantly lower in the ypT 0-2 N 0 PMRT than in the ypT 0-2 N 0 non-PMRT subgroup (2% vs. 14.8%, p=0.048), whereas the DM rate did not differ in these two subgroups ( Table 3).
Because the prognosis of patients in the ypT 0-2 N 0 non-PMRT subgroup was non-inferior to that of patients in the ypT 0-2 N 0 PMRT subgroup, the survival of ypT 0-2 N 0 non-PMRT patients was compared with that of patients with cT 1-2 N 0 stage tumors before NAC. The clinical and treatment characteristics of these two groups did not differ significantly (Supplementary Table 1), nor did their 5-year DFS rates (79.5% vs. 93.3%, p = 0.070) ( Figure 3A). By the date of the last follow-up, no patient in the cT 1-2 N 0 group had died. The 5-year OS rates in the ypT 0-2 N 0 non-PMRT and cT 1-2 N 0 group also did not differ significantly (p = 0.063) ( Figure 3B).

Correlations Between Clinical Variables and Prognosis in NAC Patients With Residual Nodal Disease
Based on National Cancer Institute guidelines (12), 128 (92.8%) of the 138 patients with residual nodal disease after NAC and surgery (ypT 0-2 N + ) received PMRT. Analysis of the correlations between clinical variables and prognosis in these patients showed that various prognostic factors correlated with DFS and OS ( Table 4). Univariate analyses showed that ypN stage, ER status and molecular subtypes, including HER2 overexpression and TNBC were significantly associated with both DFS and OS. Distant recurrence and all-cause mortality rates were was 1.98 and 4.19 times higher, respectively, in patients with ypN 2-3 than in those with ypN 1 . Multivariate analyses showed that HER2 overexpression and TNBC were significant predictors of poorer DFS and OS, whereas ypN stage was a significant independent predictor of OS in the ypT 0-2 N + PMRT subgroup ( Table 5).

DISCUSSION
Both prospective and retrospective studies have provided evidence for recommending PMRT after NAC for breast cancer patients with cT 3-4 , cN 2-3 or residual lymph node disease, as well as omitting PMRT for cT 1-2 N 0-1 patients who develop a pCR (22). However, the benefits of PMRT after NAC in breast cancer patients with initial clinical stage cT 1-2 N + , especially for those who achieved ypT 1-2 N 0 , remain largely unknown. The present study retrospectively analyzed outcomes of PMRT after NAC in cT 1-2 N + breast cancer patients in our medical center. These findings suggest that an effective response to NAC (ypT 0-2 N 0 ) is indicative of a sufficiently favorable prognosis and that PMRT may be unnecessary for cT 1-2 N + patients who achieve ypT 0-2 N 0 after NAC.
Most of the clinical and treatment characteristics of patients in the ypT 0-2 N 0 PMRT and ypT 0-2 N 0 non-PMRT groups did not differ significantly, except for pCR rate and the percentages of HER2+ patients treated with trastuzumab. The pCR rate was lower in patients in the ypT 0-2 N 0 non-PMRT group, which may have been due to the lower percentage of patients in this group who had completed NAC regimens (33.3% vs. 64%, p = 0.010) (Supplementary Table 2). The inclusion of patients who had received adjuvant chemotherapy in this analysis resulted in similar percentages of patients in the ypT 0-2 N 0 PMRT and ypT 0-2 N 0 PMRT non-PMRT groups who had completed both neoadjuvant and adjuvant chemotherapy regimens (96% vs. 92.6%, p = 0.609). The findings in this study suggest that the chemotherapy completion rate was more related to patient prognosis than the completion of NAC alone.
Of the HER2+ patients in the ypT 0-2 N 0 group, only 55.6% were treated with trastuzumab, perhaps because this agent was not approved for local medical insurance until September 2017. The percentage of HER2+ patients receiving trastuzumab was lower in the ypT 0-2 N 0 non-PMRT than in the ypT 0-2 N 0 PMRT group, resulting in inadequate treatment of the former. Even under all these circumstances, however, the 5-year DFS and OS rates did not differ significantly in these two subgroups, providing further evidence that ypT 0-2 N 0 patients have a favorable prognosis, even without PMRT.
Interestingly survival outcomes did not differ significantly in cT 1-2 N + patients who achieved ypT 0-2 N 0 without PMRT and cT 1-2 N 0 patients before NAC. Although there were concerns that the lymph node status of cT 1-2 N 0 might be downstaged, all the cT 1-2 N 0 patients remained ypT 1-2 N 0 after NAC. Based on PMRT guidelines, T 1-2 N 0 stage patients without NAC do not need PMRT (6,7). None of the patients in the present study with cT 1-2 N 0 received PMRT after NAC, while having favorable 5-year DFS and OS rates. Thus, the comparable survival of cT 1-2 N + patients who achieved ypT 0-2 N 0 without PMRT and cT 1-2 N 0 patients suggested the former can show favorable survival outcomes, even in the absence of PMRT. However, due to the limited sample size in ypT 0-2 N 0 and cT 0-2 N 0 subgroup, there is a possibility of increasing type I error caused by limited sample size.
Two ongoing prospective trials are addressing the need for PMRT in cT 1-2 N + patients who achieve nodal pCR after NAC. The NSABP51 trial is a randomized phase III clinical trial evaluating PMRT in pathologically proven cT 1-3 N 1 patients A B   who convert to pN 0 after NAC (www.nsabp.pitt.edu/B-51.asp). The RAPCHEM trial is prospective observational study evaluating the 5-year LRR rate in cT 1-2 N 0-1 patients after NAC, breast surgery and radiotherapy, with protocols based on ypTNM stage (https://clinicaltrials.gov/ct2/show/study/ NCT01279304). The protocol of the RAPCHEM trial resulted in the stratification of patients with ypN 0 to a low risk group, with these patients not receiving PMRT. Several previous retrospective studies have analyzed the need for PMRT in subjects who received NAC. For example, a study of 676   patients with locally advanced breast cancer, including 145 with cT 1-2 stage tumors, who were treated with NAC and mastectomy, found that PMRT did not reduce LRR or improve 10-years CSS in patients with cT 1-2 stage tumors (13). Although only 29.7% of patients achieved ypN 0 and 68.9% had residual nodal disease after NAC, PMRT did not benefit survival. Another study of 106 cT 1-4 N + breast cancer patients who achieved pCR after NAC also found that PMRT did not improve 10-year LRR in clinical stage I-II patients, but significantly improved 10-year LRR, DMFS and OS in clinical stage III patients (14), suggesting that PMRT may be more likely to improve survival in patients with more advanced stage tumors. Similarly, two other studies found that PMRT did not improve 10year LRFS and OS in clinical stage II-III patients who achieved pN 0 after NAC (15,16). A large study in 10,283 cN + patients found that PMRT significantly improved 5-year OS in cT 1-3 N 1 patients after NAC, whether they achieved ypN 0 or remained ypN + (17). However, 40.1% of these patients had stage cT 3 at diagnosis, a class of patients shown to benefit from PMRT, with clinical consensus indicating that PMRT after NAC can improve survival in patients with clinical stage III breast cancer (i.e., T 3 N 1 ) (12). A study of 88 cT 1-2 N + patients who developed ypN 0 after NAC found that PMRT significantly improved 5-year LRFS, but had no effect on DMFS and DFS (18). Because 39.8% of these patients remained ypN 1 , a class that can gain survival benefits from PMRT, the effect of PMRT on cT 1-2 N + patients who achieve ypN 0 after NAC has not yet been clarified. Interestingly, a study of 217 cT 1-2 N 0-1 patients found that PMRT significantly reduced 5-year LRR in high risk, but not in low risk, patients (19). Risk factors in that study included ypN stage, histologic grade and lymphatic vessel invasion. The cT 1-2 N + patients who achieved ypN 0 after NAC in our study may comparable to the low risk cT 1-2 N 0-1 population in the previous study (19), in which PMRT did not decrease 5-year LRR. An analysis of 142 cT 1-2 N 1 breast cancer patients found that PMRT did not improve OS in cT 1-2 N 1 patients who achieved ypT 1-2 N 0 after NAC, results consistent with the findings of the present study (20). That study, however, found that PMRT significantly improved RFS but not LRFS, whereas rates of locoregional recurrence and distant metastases did not difference in patients who did and did not receive PMRT. A recent study of 554 clinical stage II-III patients found that PMRT after NAC significantly reduced 5-year LRR in clinical stage II-III patients, but had no local control or survival benefit in patients with ypN 0 after NAC (21). To date, no Grade 1 evidence has shown that PMRT can be omitted in cT 1-2 N + patients who achieve ypT 0-2 N 0 after NAC. Although the present study suggests that PMRT can be omitted, this study had several limitations. First, this was a retrospective study, with selection bias between the PMRT and non-PMRT groups being the most important inherent shortcoming. Secondly, the sample size and number of events in this study are limited, which may lead to the statistical type I error. With interest, we will continue to follow up the outcomes of ypT 0-2 N 0 PMRT and non-PMRT patients in our hospital, and may conduct a prospective study to investigate the possibility of exempting radiotherapy in cT 1- chemotherapy in ypT 0-2 N 0 non-PMRT subgroup may cause lower lymph node complete response rate (ypN 0 ), which might lead to fewer patients enrolled in ypT 1-2 N 0 group. In addition, initial axillary lymph node status in a small proportion of patients (16.3%) was not determined by pathology.
In conclusion, this study found that PMRT did not affect survival in cT 1-2 N + breast cancer patients who achieved ypT 0-2 N 0 after NAC. Without PMRT, the prognosis of cT 1-2 N + patients who achieved ypT 0-2 N 0 after NAC did not differ significantly from that of cT 1-2 N 0 patients, suggesting that PMRT may be safely omitted for cT 1-2 N + breast cancer patients who achieve ypT 0-2 N 0 after NAC. Prospective studies in larger patient populations are warranted.

DATA AVAILABILITY STATEMENT
In this study, all the patients' data were from the Second Affiliated Hospital, Zhejiang University School of Medicine, which are not publicly available in order to protect patient privacy, but can be accessed from the corresponding author, Dr. Jiaojiao Zhou (zhoujj@zju.edu.cn), on reasonable request.

ETHICS STATEMENT
This study was approved by the Ethics Committee of our hospital (Approval No: 2020-363) with waiver of informed consent.

AUTHOR CONTRIBUTIONS
JZ, YC and SZ planned and designed this study. JZ, YC drew the outline of this study. JZ, ML and HC wrote the manuscript. ML, HC, HD and YJ collected the all the clinical data. ML and HC performed the data analysis. HD, YJ, KZ and HM did the followup of patients. ML, HC and GW participated in searching relevant literatures. All authors have read and approved the final manuscript.