The effect of postmastectomy radiotherapy in node-positive triple-negative breast cancer

Breast cancer is the most common malignant cancer and the primary cause of cancer-related mortality in women. Every year, almost 2,76,000 women are newly diagnosed with breast cancer and about 42,000 women die from it in the United States alone [1]. Enhancing survival outcome by multidisciplinary comprehensive treatment remains the global focus of breast cancer. In current clinical practice, the treatment of breast cancer is based not only on the T and N stages of patients but also on the status of estrogen receptor (ER), progesterone receptor, and human epidermal growth factor receptor-2 (HER-2). Breast cancer subtypes based on ER, PR, and HER-2 have been widely used. Triple-negative breast cancer (TNBC) is a subtype characterized by an absence of the ER, PR, and HER-2 status, and accounts for 15% of all breast cancers [2]. TNBC has a higher risk of early metastasis, local recurrence, and poorer prognosis than other types of breast cancer [3].

Mastectomy is one of the local operations for breast cancer. About one-third of all stage I or II and 68% of stage III breast cancer patients undergo mastectomy [3]. The National Comprehensive Cancer Network (NCCN) guidelines recommend radiotherapy for patients with more than four positive lymph nodes, while patients with 1–3 lymph nodes are recommended to receive radiotherapy [4]. However, the recommendation of PMRT does not refer to molecular subtypes of breast cancer. A meta-analysis of 22 randomized trials from the Early Breast Cancer Triallists’ Collaboration Group (EBCTCG) and other studies showed that radiotherapy could obviously decrease recurrence and mortality risk of TNBC and improve overall survival (OS) of patients with positive lymph nodes [5‐7]. The value of PMRT for TNBC patients in different pathological nodal stages has not yet been established. The Danish Breast Cancer Cooperative Group (DBCG) and some other studies have shown that there was no significant survival benefit for TNBC patients with positive lymph nodes from receiving PMRT [8‐10]. However, insufficient chemotherapy and axillary lymphadenectomy may have likely limited the power of the analysis. A previously conducted retrospective analysis reported no survival improvement with PMRT in patients with T1–2N1 disease [11]. Collectively, previous studies have shown contradictory results in the investigation of survival according to the status of PMRT receipt in different pathological stages of TNBC.

At present, the recommendation of PMRT is mainly based on the pathological T and N stage; the value of PMRT on TNBC patients needs to be further clarified. Therefore, we aimed to identify the effect of PMRT on breast cancer-specific survival (BCSS) in TNBC patients with different pathological nodal stages using the data from the Surveillance, Epidemiology, and End Results (SEER) database.

Although the recommendation for TNBC radiotherapy is similar to that for other breast cancer subtypes, the value of PMRT in TNBC patients remains unclear. In this study, we used a population-based study to retrospectively explore the effect of PMRT on 3-year BCSS among TNBC patients with different pathological stages. Our results showed that PMRT could provide significant benefits to 3-year BCSS for TNBC patients in T1-4N2 and T1-4N3 stages; TNBC patients with pathological T3–4N1 stage disease are more suitable for PMRT considering the results of multivariate analysis have more accuracy than the Kaplan–Meier analysis. However, those with T1–2 stage TNBC showed no benefit from PMRT.

The value of PMRT in patients with pathological T1-T2N1 stage TNBC disease is still controversial. Previous studies have evaluated the efficacy of PMRT on survival outcomes of TNBC patients. The analysis of Kindts et al. showed that TNBC patients that did not receive PMRT had a worse locoregional recurrence rate (LRR) outcome than those that did (HR = 4.45, 95% CI = 1.26–15.69, P = 0.020) [13]. Adjuvant radiotherapy after mastectomy has been shown to reduce the 5-year LRR by 17% (6% vs. 23%) and the 15-year breast cancer mortality by 5.4% (54.7% vs. 60.1%, P < 0.001) when compared to those without radiotherapy for node-positive cancer [14]. The American Society of Clinical Oncology (ASCO) recommends that PMRT can be used as a standard essential treatment for patients with ≥4 positive axillary lymph nodes [15]; other researchers have accordingly shown that TNBC patients with positive lymph nodes are also equally eligible for PMRT [16, 17]. However, the classification criteria of the number of lymph nodes is difficult to implement for some patients. The beneficial impact of PMRT on disease-free survival (DFS) as reported by Chen et al. supports the use of PMRT in pathological T1-T2N1 stage patients despite the absence of a superior locoregional recurrence-free survival (LRFS) outcome [18]. Recent studies showed that pathological T1–2N1M0 TNBC patients benefit from PMRT with an improved BCSS (P = 0.010) [19], while another study in 2004 with 152 TNBC patients showed that pathological T1–2N0–1 stage TNBC without radiotherapy carries a significantly higher risk of LRR (79.6% vs. 57.9%, P = 0.049) [20]. The results of the above-mentioned studies were contradictory compared with our outcome, in that pathological T1–2N1M0 stage TNBC patients showed no benefit from PMRT and the 3-year BCSS rates were comparable between the PMRT and non-PMRT subgroups (P = 0.191). The inherent selection biases in patient selection and treatment, lack of information about some clinical characteristics, small sample size, and the long interval with respect to the present study are some of the limitations that likely contributed to the inconsistent results.

However, some studies showed that pathological N1 stage TNBC patients do not require PMRT as a necessary treatment [9, 21].. PMRT was associated with an increase in radiation toxicities such as lymphedema, cardiotoxicity, and pneumonitis [22‐24]. The consensus of the St. Gallen Breast Cancer Conference showed that more than 64% experts opposed the recommendation of introducing PMRT as a routine treatment for patients with pathological T1–2N1M0 stage breast cancer, and instead recommended considering the omission of PMRT in pathological T1–2 patients with 1–3 positive lymph nodes [25]. Although there are no exact statistics about the efficacy of PMRT in patients with pathological T1–2N1M0 stage TNBC, most clinicians seemed to indicate that PMRT provided no significant benefit for patients with pathological T1–2N1 stage disease in overall breast cancer types. In a retrospective multicenter analysis, Kim et al. found that T1–2N1 stage patients who received PMRT had no obvious improvement of DFS and OS because of the malignant biological features and radio-resistance of TNBC cells [26]. Bhoo-Pathy et al’s study showed that PMRT was not associated with improved survival in T1–2N0–1 TNBC patients, but showed better survival outcome in T3–4N2–3 TNBC patients [27]. The results of these studies are consistent with our outcome. Mastectomy and chemotherapy have been associated with low risk of LRR in TNBC patients with 1–3 positive lymph nodes, and TNBC patients benefit more from chemotherapy when PMRT alone does not provide better curative efficacy [28]. Neo-adjuvant chemotherapy is an important treatment modality in TNBC; however, its effect on PMRT is still controversial and more prospective trials are needed to validate the results [29, 30]. Patients with 1–3 positive lymph nodes who received modern taxane-based chemotherapy had excellent locoregional control; thus, the use of PMRT in patients with 1–3 positive nodes should be tailored to individual patient risks [31]. As seen in our study, PMRT provided no benefits for patients with T1–2N1 stage TNBC. We reasonably assumed that TNBC patients with pathological T1–2N1 stage show relatively good disease control by undergoing mastectomy plus chemotherapy. PMRT is not universally administered in pathological T1–2N1 stage TNBC and should be chosen as a personal requirement especially for patients with higher nodal stages or high-risk biology [32], such as those aged ≥50 years and of black ethnicity as analyzed in our multivariate Cox hazard analysis. Therefore, the risk factors in TNBC patients with pathological T1–2N1 stage who consider receiving PMRT should be fully considered by clinicians.

According to the results of our study, an increasing trend of 3-year BCSS benefit was reflected based on the Kaplan–Meier plots. The patients were stratified for survival analysis by pathological T and N stages for more solid elucidation. Our results validated the data that PMRT is a strong predictive factor of better BCSS for patients with pathological T3–4N1, T1-4N2, and T1-4N3 stage TNBC. Both Kaplan–Meier and multivariate analysis showed that patients with T1-4N2 and T1-4N3 stages who received PMRT had better 3-year BCSS, which was in line with previous studies that reported that additional PMRT for patients with high-risk (stage T3–4 and/or N2–3) TNBC showed superior outcomes in the LRFS and DFS [18, 33]. Patients with pathological T3–4N1 stage TNBC showed contradictory results of 3-year BCSS in both the Kaplan–Meier and multivariate analyses. However, considering that the Kaplan–Meier analysis is a univariate analysis and that the effect of some other factors on PMRT may be omitted, we believe that the results of multivariate analysis are more reliable in order to improve the long-term BCSS. The ASCO recommends considering the need for PMRT for patients with T1–2N1 stage disease by evaluating the risk factors [15]. The NCCN breast cancer guidelines recommend that patients with pathological T3–4 primary tumors or N2–3 axillary lymph nodes after mastectomy should undergo PMRT as a standard adjuvant therapy [4].

Our study has some limitations. First, this is a retrospective analysis from the SEER database rather than a prospective, randomized controlled trial. Therefore, we cannot clarify the reasons for patients choosing PMRT; moreover, the inherent selection biases could undermine the validity of our analysis. Second, some information about patients’ complications, systemic chemotherapy regimens (adjuvant and neo-adjuvant), and the situation of nodal irradiation were absent from the SEER database, which may have led to unconvincing results. At the same time, some of the selected patients in this study had a short follow-up period, and the data might be skewed by the number of patients with relatively short follow-up. Therefore, we did not consider overall survival as an analytical indicator. Thus, we were unable to evaluate the influence of PMRT on LRR of TNBC patients and balance the different possible effects of chemotherapy with PMRT. LRR is an important indicator to evaluate the efficacy of radiotherapy; hence, missing information on LRR, distant metastasis, dose of radiotherapy, and target area are also limitations of the study. Last, SEER may underestimate the reporting rate of radiotherapy. Hence, a randomized controlled clinical study is needed to provide more evidence and a scientific basis to guide clinical treatment. We are currently looking forward to the results of the Selective Use of Postoperative Radiotherapy after Mastectomy trial, where patients with pathological T1–2N1 stage disease were separated randomly into radiotherapy or non-radiotherapy subgroups [34].

Radiotherapy is an important treatment modality for TNBC patients. The study has provided evidence that patients with pathological T3–4N1 and T1-4N2–3 stage TNBC benefit from PMRT with a better 3-year BCSS; however, PMRT is not beneficial for those with pathological T1–2N1 stage TNBC. These patients should decide whether to undergo PMRT in light of other high-risk factors. Further prospective studies are recommended to elucidate the benefit of PMRT in TNBC patients and provide a strong evidence base for patient selection.