Outcome of post-mastectomy radiotherapy after primary systemic treatment in patients with different clinical tumor and nodal stages of breast cancer: a cohort study

Abstract

To evaluate the effect of post-mastectomy radiation therapy (PMRT) stratified by clinical tumor (T) or nodal (N) staging and determine predictors of overall survival (OS), locoregional recurrence (LRR), distant metastasis, and disease-free survival (DFS) in patients with breast cancer who received neoadjuvant chemotherapy (NACT) and total mastectomy (TM), we enrolled patients who received a diagnosis of breast invasive ductal carcinoma who received NACT followed by TM. Cox regression analysis was employed to calculate hazard ratios (HRs) and confidence intervals (CIs). Univariate and multivariate Cox regression analyses indicated that non-PMRT, Charlson comorbidity index ≥ 2, advanced clinical T or N stage, pathologic partial response, pathologic stationary disease, or pathologic progression disease were poor prognostic factors for OS. Well-differentiated tumor grade, pathologic complete response, and positive hormone receptors were better independent prognostic factors for OS. Adjusted HRs derived from PMRT for breast cancer after NACT and TM were 0.69 (0.53-0.89) and 0.74 (0.59-0.93) in clinical T3 and T4, respectively. aHRs derived from PMRT for breast cancer after NACT and TM were 0.67 (0.45-0.99), 0.75 (0.62-0.92), and 0.77 (0.60-0.98) in clinical N0, N1, N2-3, respectively. The aHRs (95% CI) of the PMRT group to the non-PMRT group for LRR-free survival and DFS were improved significantly. Our study indicated that PMRT significantly improved OS in clinical T3N0-T4N3 and for LRR-free survival and DFS in clinical T2N0-T4N3 from those of non-PMRT patients regardless of pathologic response and other predictors.

Introduction

Neoadjuvant chemotherapy (NACT) is the systemic treatment of breast invasive intraductal carcinoma (IDC) before definitive surgical therapy [1]. Although all systemic therapy for nonmetastatic invasive breast cancer is intended to reduce the risk of distant recurrence, it is administered neoadjuvantly to assess the response to treatment, to downstage the cancer, and to reduce postoperative complications such as lymphedema [1-4]. For most patients who receive NACT, indications for post-mastectomy radiation therapy (PMRT) depend on various multifactors, such as the pretreatment stage and type of surgery (total mastectomy [TM] or breast conservative surgery), and consider pathologic responses to NACT [5-11]. In Taiwan, breast cancer is the most common cancer among women [12,13]. Despite its increased incidence, its 5-year mortality rate has not significantly improved over the past decade (4.5% in 1997 and 4.4% in 2008) [13]. Over the last 10 years, NACT has become increasingly popular in Taiwan and worldwide for patients with breast cancer [13,14]. Therefore, an effective and optimal adjuvant therapy, including PMRT after NACT and TM, is crucial for such patients given the increasing incidence of breast cancer [5-11,15] because no improvement in survival or local control has occurred over the past decade [13].

The incidence of breast cancer has decreased in North America but not in Asia, where it continues to increase [15]. A notable manifestation of the bimodal age distribution of breast cancer was observed in women [16]. The occurrence of early-onset breast cancer in the Asian or Taiwanese population is earlier than that in the Western population, which results in a higher incidence of breast cancer in young Taiwanese women [17-21]. Moreover, the late onset age distribution of patients with breast cancer in Asia or Taiwan (40-50 years) is earlier than that in Western countries (60-70 years) and peaks at an age of 45-50 years in most women [17-21]. The age-specific incidence rates of breast cancer increase sharply until the menopausal stage [22]. Younger patients with breast cancer typically have more aggressive tumors that are more likely to recur both locoregionally and distantly, and older patients with breast cancer tend to have less aggressive disease [23]. Determining suitable indications that PMRT improves survival in younger patients with breast cancer with aggressive tumors who receive aggressive treatments, such as NACT and TM, is crucial.

The benefits of PMRT remain unclear and controversial with variable indications including prechemotherapy staging, tumor size, nodal status, post-chemotherapy cancer staging, tumor size, nodal status, and pathologic response [5-11]. In addition, the benefits of PMRT are undetailed regarding outcomes of locoregional recurrence (LRR), overall survival (OS), distant metastasis (DM), and disease-free survival (DFS). Few studies have verified the effect of PMRT based on clinical tumor (T) or clinical nodal (N) staging, and we do not prefer to rely on pathologic staging or pathologic response because pathologic response depends on many complex factors [24-26]. In this study, we evaluated the effect of PMRT stratified by clinical T or N staging and discovered the predictors of OS, LRR, DM, and DFS in patients with breast cancer who received NACT and TM.

Patients and methods

In this study, we established a cohort by using data from the Taiwan Cancer Registry database (TCRD). We enrolled patients who received a diagnosis of breast IDC between January 1, 2007, and December 31, 2015. The follow-up duration was from the index date to December 31, 2016. The Cancer Registry database of the Collaboration Center of Health Information Application contains detailed cancer-related information of patients including clinical stage, treatment modalities, pathological data, radiation techniques, irradiation doses, and chemotherapy regimens used [27-35]. In the study, we included PMRT to both the chest wall and regional nodes with at least 50 Gy. Our protocols were reviewed and approved by the Institutional Review Board of Taipei Medical University. The diagnoses of enrolled patients were confirmed using their pathological data, and patients who received a new diagnosis of breast IDC were confirmed to have no other cancers. Patients diagnosed with breast IDC who received NACT followed by TM, were aged ≥ 20 years, and had American Joint Committee on Cancer (AJCC) clinical cancer stage I-IV were included. AJCC clinical tumor, node, and metastasis staging information was recorded in the TCRD. The breast cancer stages were all based on the seventh edition of the AJCC. Patients with metastasis, missing sex data, aged <20 years, undergoing nonstandard PMRT, with unclear differentiation of tumor grade, having unclear pathologic response, missing estrogen receptor (ER) or progesterone receptor (PR) status, missing human epidermal growth factor receptor 2 (HER2) status, and unclear staging were excluded. We also excluded patients with unclear regimens of NACT, who underwent fewer than four cycles of NACT, with ill-defined nodal surgery, and with nonrecorded hospital levels (medical center or nonmedical center hospitals) or hospital regions in our cohort. Being ER/PR-positive was defined as ≥ 1% of tumor cells exhibiting positive nuclear staining through immunohistochemistry [36], and being HER2 positive was defined as fluorescence in situ hybridization with a ≥ 2 ratio [37]. Finally, we enrolled patients with breast IDC who received NACT followed by TM and categorized them into the following groups according to their treatment modality to compare their outcomes: group 1 (control group), consisting of patients who did not receive PMRT, and group 2 (case group), consisting of patients who received PMRT. The index date was the diagnosis date of breast cancer. The median total dose and fraction size of PMRT were 50 and 2 Gy per fraction in group 2. Comorbidities were scored using the Charlson Comorbidity Index (CCI) [38,39]. Only comorbidities observed 6 months before the index date were included. Comorbidities were identified and included according to the main International Classification of Diseases, Ninth Revision, Clinical Modification (ICD-9-CM) diagnosis codes for the first admission or more than two repeated main diagnosis codes for visits to the outpatient department.

After adjustment for confounders, the time dependent Cox proportional method was used to model the time from the index date to all-cause death, LRR, DM, and DFS among patients who underwent PMRT or non-PMRT. Through multivariate analysis, hazard ratios (HRs) were adjusted for age, diagnosis year, CCI scores, tumor grade differentiation, clinical T stage, clinical N stage, pathologic response, NACT regimen, nodal surgery, ER or PR status, HER2 status, hospital level, and hospital region. The effect of PMRT on OS, LRR-free survival, DM-free survival, and DFS was determined through multivariable Cox regression analysis for patients who received NACT and TM with or without PMRT and stratified by clinical T or clinical N stage. Stratified analyses in different clinical T or N stages were performed to evaluate mortality, recurrence, metastatic risk associated with PMRT or non-PMRT, age, diagnosis year, CCI score, tumor grade differentiation, pathologic response, NACT regimen, nodal surgery, ER or PR status, HER2 status, hospital level, and hospital region used in the multivariate analysis. All analyses were performed using SAS (v9.3; SAS, Cary, NC, USA). A two-tailed value of P<0.05 was considered statistically significant.

Results

A final cohort of 4236 patients (2917 and 1319 patients in groups 1 and 2, respectively) were eligible for further analysis. Patient characteristics are summarized in Table 1. No statistical differences appeared in age, CCI score, tumor grade, and ER or PR status between PMRT and non-PMRT groups (Table 1). More patients had received PMRT in 2011-2015 than in 2007-2014. This indicated that more patients had received PMRT in recent years. The PMRT group had more patients with advanced breast cancer in clinical T or N stages. Fewer patients with pathological complete response (pCR) received PMRT. Moreover, more patients with pathological stationary disease (pSD) or pathological progression of disease (pPD) received PMRT. Fewer patients receiving NACT with anthracycline-based regimens received PMRT. More patients receiving NACT with taxane-based regimens received PMRT. More patients received axillary lymph node dissection as nodal surgery in the PMRT group. The PMRT group had more HER2-positive patients. More patients received PMRT in nonmedical centers, and more patients were in the north of Taiwan (Table 1).

Table 1.

Characteristics of breast invasive intraductal carcinoma after NACT and TM between patients who received PMRT and those who did not (non-PMRT)

Variable		Total Mastectomy
		PMRT (N = 2917)	Non-PMRT (N = 1319)	P
Age	Mean (SD)	51.3 (10.3)	52.0 (10.9)	0.1108
	Median (IQR; Q1, Q3)	51 (44.58)	51 (44.59)
	20-49 years	1301 (44.6%)	562 (42.6%)	0.2263
	50+ years	1616 (55.4%)	757 (57.4%)
Diagnosis year	2007-2010	956 (32.8%)	556 (42.2%)	<0.0001
	2011-2015	1961 (67.2%)	763 (57.8%)
CCI score	0	2423 (83.1%)	1042 (79.0%)	0.0065
	1	350 (12.0%)	196 (14.9%)
	2+	144 (4.9%)	81 (6.1%)
Differentiation	Well	185 (6.3%)	86 (6.5%)	0.9504
	Moderate	1505 (51.6%)	690 (52.3%)
	Poor	1227 (42.1%)	543 (41.2%)
Clinical T stage	0-1	66 (2.3%)	48 (3.6%)	<0.0001
	2	956 (32.8%)	688 (52.2%)
	3	998 (34.2%)	287 (21.8%)
	4	897 (30.8%)	296 (22.4%)
Clinical N stage	0	391 (13.4%)	436 (33.1%)	<0.0001
	1	1515 (51.9%)	636 (48.2%)
	2	647 (22.2%)	186 (14.1%)
	3	364 (12.5%)	61 (4.6%)
Pathologic response	pCR	804 (27.6%)	758 (57.5%)	<0.0001
	pPR	822 (28.2%)	326 (24.7%)
	pSD	738 (25.3%)	128 (9.7%)
	pPD	553 (19.0%)	107 (8.1%)
NACT regimen	Taxane-based	1176 (40.3%)	331 (25.1%)	<0.0001
	Anthracycline-based	772 (26.5%)	533 (40.4%)
	Both	833 (28.6%)	306 (23.2%)
	Neither	136 (4.7%)	149 (11.3%)
Nodal surgery	ALND	2104 (72.1%)	890 (67.5%)	<0.0001
	SLND	813 (27.9%)	429 (32.5%)
ER/PR status	Negative	1401 (48.0%)	653 (49.5%)	0.3726
	Positive	1516 (52.0%)	666 (50.5%)
HER2 status	Negative	1876 (64.3%)	915 (69.4%)	0.0013
	Positive	1041 (35.7%)	404 (30.6%)
Hospital level	Medical center	1595 (54.7%)	946 (71.7%)	<0.0001
	Nonmedical Center	1322 (45.3%)	373 (28.3%)
Hospital area	North	1633 (56.0%)	644 (48.8%)	<0.0001
	Middle	477 (16.4%)	250 (19.0%)
	South/East	807 (27.7%)	425 (32.2%)
Mean follow-up time, months (SD)		59.6 (30.1)	65.5 (32.8)
Death		754 (25.8%)	318 (24.1%)	0.2279
Distant metastasis		224 (7.7%)	106 (8.0%)	0.6879
Locoregional recurrence		706 (24.2%)	417 (31.6%)	<0.0001
DFS		578 (19.8%)	268 (20.3%)	0.7042

PMRT, post-mastectomy radiation therapy; T, tumor; N, nodal; DFS, disease-free survival; NACT, neoadjuvant chemotherapy; TM, total mastectomy; ER, estrogen receptor; PR, progesterone receptor; HER2, human epidermal growth factor receptor 2; pCR, pathological complete response; pPR, pathologic partial response; pSD, pathological stationary disease; pPD, pathological progression of disease; ALND, axillary lymph node dissection; SNLD, Sentinel Lymph Node Dissection; IQR, interquartile range; SD, Standard Deviation.

According to multivariate Cox regression analysis, PMRT was a significant independent predictor of OS, LRR, and DFS (Table 2) but was a nonsignificant predictor of DM. Both univariate and multivariate Cox regression analyses indicated that non-PMRT, CCI ≥ 2, advanced clinical T or N stages, pathologic partial response (pPR), pSD, and pPD were poor prognostic factors for OS. Well-differentiated tumor grade, pCR, and being ER/PR-positive were better independent prognostic factors for OS. Advanced clinical N stage, poorly differentiated tumor grade, pPR, pSD, and pPD were poor prognostic factors for DM. In addition, poor prognostic factors after multivariate analysis for LRR comprised non-PMRT, poorly differentiated tumor grade, advanced T stage, advanced N stage, pPR, pSD, pPD, and being HER2 positive. Poor prognostic factors for DFS in patients with breast IDC status post-NACT and TM were non-PMRT, poor differentiation, advanced clinical T or N stages, poor pathologic response (including pPR, pSD, and pPD compared to CR), being ER/PR negative, and being HER2 positive. According to both univariate and multivariate Cox regression analyses, the aHRs (95% confidence interval [CI]) of PMRT to non-PMRT were 0.75 (0.65-0.87), 0.86 (0.66-1.11), 0.42 (0.37-0.48), and 0.68 (0.58-0.80) in all-cause death, DM, LRR, and DFS, respectively.

Table 2.

Multivariate Cox proportional hazards mode analysis of breast invasive intraductal carcinoma after NACT and TM

		OS		DM		LRR		DFS
		HR (95% CI)	p-value	HR (95% CI)	p-value	HR (95% CI)	p-value	HR (95% CI)	p-value
PMRT	No	ref	0.0001	ref	0.24	ref	<0.0001	ref	<0.0001
	Yes	0.75 (0.65-0.87)		0.86 (0.66-1.11)		0.42 (0.37-0.48)		0.68 (0.58-0.80)
Age	20-49 years	ref	0.61	ref	0.34	ref	0.18	ref	0.43
	50+ years	1.04 (0.91-1.18)		0.79(0.56-1.89)		0.92 (0.81-1.04)		0.94 (0.81-1.09)
Diagnosis year	2007-2010	ref	0.90	ref	0.38	ref	0.40	ref	0.41
	2011-2015	0.99 (0.87-1.13)		0.90 (0.71-1.14)		1.06 (0.93-1.20)		1.06 (0.92-1.24)
CCI score	0	ref	0.0008	ref	0.1393	ref	0.47	ref	0.66
	1	0.94 (0.77-1.14)		1.23 (0.88-1.71)		1.06 (0.88-1.27)		0.98 (0.79-1.22)
	2+	1.53(1.21-1.93)		1.44 (0.21-1.93)		1.16 (0.90-1.50)		1.15 (0.84-1.55)
Differentiation	Poor	ref	<0.0001	ref	0.0059	ref	0.0106	ref	0.0162
	Moderate	0.74 (0.64-0.85)		0.90 (0.98-1.72)		0.90 (0.78-1.03)		0.98 (0.83-1.15)
	Well	0.44 (0.30-0.63)		0.83 (0.81-0.91)		0.65 (0.47-0.89)		0.62 (0.43-0.89)
Clinical T stage	0-1	ref	<0.0001	ref	0.43	ref	<0.0001	ref	0.0334
	2	1.88 (1.03-3.45)		0.77 (0.38-1.54)		1.26 (0.79-1.98)		1.56 (0.89-2.73)
	3	2.42 (1.32-4.45)		0.67 (0.33-1.35)		1.58 (1.00-2.51)		1.77 (1.00-3.12)
	4	2.80 (1.52-5.15)		0.82 (0.41-1.67)		1.87 (1.18-2.98)		1.93 (1.09-3.41)
Clinical N stage	0	ref	0.0001	ref	0.0001	ref	<0.0001	ref	<0.0001
	1	1.36 (1.11-1.66)		1.08 (1.01-1.35)		1.51 (1.24-1.83)		1.44 (1.16-1.80)
	2	1.45 (1.15-1.82)		1.17 (1.03-1.57)		1.43 (1.14-1.79)		1.36 (1.05-1.77)
	3	1.82 (1.40-2.35)		1.45 (1.09-2.26)		1.75 (1.35-2.26)		1.95 (1.46-2.60)
Pathologic response	pCR	ref	<0.0001	ref	0.26	ref	<0.0001	ref	<0.0001
	pPR	1.52 (1.26-1.84)		1.29 (0.95-1.74)		1.41 (1.18-1.68)		1.32 (1.08-1.62)
	pSD-pPD	2.42 (2.02-2.90)		1.16 (0.85-1.59)		2.34 (1.97-2.79)		1.90 (1.56-2.32)
NACT regimen	Anthracycline-based	ref	0.47	ref	0.16	ref	0.99	ref	0.24
	Taxane-based	1.12 (0.95-1.31)		1.34 (0.85-1.56)		1.07 (0.73-1.30)		1.28 (0.87-1.53)
	Both	1.03 (0.88-1.21)		1.14 (0.98-1.34)		1.03 (0.77-1.36)		1.09 (0.91-1.30)
	Neither	1.14 (0.88-1.47)		1.32 (0.83-1.70)		1.08 (0.62-1.56)		1.19 (0.66-1.25)
Nodal surgery	SLND	ref	0.70	ref	0.1109	ref	0.37	ref	0.44
	ALND	1.17 (0.79-1.73)		1.10 (0.92-1.55)		1.28 (0.91-1.82)		1.31 (0.86-1.98)
ER/PR positive		0.67 (0.59-0.76)	<0.0001	0.94 (0.89-1.15)	0.30	0.82 (0.69-1.01)	0.08	0.44 (0.24-1.38)	<0.0001
HER-2 positive		1.07 (1.03-1.15)	0.0087	1.79 (1.42-2.26)	<0.0001	1.54 (1.36-1.75)	<0.0001	1.72 (1.49-1.99)	<0.0001
Hospital level	Medical center	ref	0.21	ref	0.20	ref	0.61	ref	0.0385
	Nonmedical Center	0.92 (0.81-1.05)		0.86 (0.67-1.09)		1.03 (0.91-1.17)		0.86 (0.74-0.99)

HRs, hazard ratios; CI, confidence interval; PMRT, post-mastectomy radiation therapy; T, tumor; N, nodal; NACT, neoadjuvant chemotherapy; TM, total mastectomy; ER, estrogen receptor; PR, progesterone receptor; HER2, human epidermal growth factor receptor 2; pCR, pathological complete response; pPR, pathologic partial response; pSD, pathological stationary disease; pPD, pathological progression of disease; ALND, axillary lymph node dissection; SNLD, sentinel lymph node dissection; OS, overall survival; LRR, locoregional recurrence; DM, distant metastasis; DFS, disease-free survival.

For stratified different clinical T or N stages, multivariate Cox regression analyses revealed that PMRT in patients with breast cancer who received NACT and TM was a significant independent predictor of better OS, except for clinical T1 and T2 (Figure 1A). Regardless of the pathologic response and other predictors adjusted in the model, PMRT resulted in better OS in clinical T3, T4, N0, N1, N2, and N3. aHRs derived for PMRT for breast cancer after NACT and TM were 0.69 (0.53-0.89) and 0.74 (0.59-0.93) in clinical T3 and T4, respectively (Figure 1A). Moreover, aHRs derived for PMRT for breast cancer after NACT and TM were 0.67 (0.45-0.99), 0.75 (0.62-0.92), and 0.77 (0.60-0.98) in clinical N0, N1, N2, and N3, respectively (Figure 1A). The aHRs (95% CI) of the PMRT group to the non-PMRT group for LRR-free survival were 0.48 (0.37-0.61), 0.37 (0.29-0.48), 0.40 (0.32-0.50), 0.41 (0.27-0.61), 0.41 (0.34-0.49), and 0.44 (0.35-0.56) in clinical T2, T3, T4, N0, N1, N2, and N3, respectively (Figure 1C). The aHRs of DFS derived for PMRT for breast cancer after NACT and TM were 0.67 (0.50-0.88), 0.71 (0.52-0.97), and 0.68 (0.52-0.89) in clinical T2, T3, and T4, respectively (Figure 1D). Moreover, the aHRs of DFS derived for PMRT for breast cancer after NACT and TM were 0.58 (0.36-0.92), 0.72 (0.58-0.89), and 0.70 (0.52-0.93) in clinical N0, N1, N2, and N3, respectively (Figure 1D). The aHRs of DM-free survival derived for PMRT for breast cancer after NACT and TM were 0.55 (0.34-0.89) and 0.51 (0.28-0.95) in patients with clinical T3 and N0, respectively (Figure 1B). There were no statistical differences between the PMRT and non-PMRT groups for DM-free survival after multivariate analysis of clinical T0, T1, T2, T4, N1, N2, and N3 (Figure 1B).

Figure 1.

Effect of post-mastectomy radiation therapy (PMRT) on overall survival, locoregional recurrence-free survival, distant disease-free survival, and disease-free survival from multivariable Cox regression analysis in patients who received neoadjuvant chemotherapy and total mastectomy with or without PMRT stratified by clinical tumor and nodal stage. A. Effect of post-mastectomy radiation therapy on overall survival. HR Ratio: All variables presented in Table 2 were used in the multivariate analysis. HRs, hazard ratios; CI, confidence interval; PMRT, post-mastectomy radiation therapy; T, tumor; N, nodal; OS, overall survival; LRR, locoregional recurrence; DM, distant metastasis; DFS, disease-free survival. B. Effect of post-mastectomy radiation therapy on distant metastasis-free survival. Adjusted HR: All variables presented in Table 2 were used in the multivariate analysis. HRs, hazard ratios; CI, confidence interval; PMRT, post-mastectomy radiation therapy; T, tumor; N, nodal; OS, overall survival; LRR, locoregional recurrence; DM, distant metastasis; DFS, disease-free survival. C. Effect of post-mastectomy radiation therapy on locoregional recurrence-free survival. Adjusted HR: All variables presented in Table 2 were used in the multivariate analysis. HRs, hazard ratios; CI, confidence interval; PMRT, post-mastectomy radiation therapy; T, tumor; N, nodal; OS, overall survival; LRR, locoregional recurrence; DM, distant metastasis; DFS, disease-free survival. D. Effect of post-mastectomy radiation therapy on disease-free survival. Adjusted HR: All variables presented in Table 2 were used in the multivariate analysis. HRs, hazard ratios; CI, confidence interval; PMRT, post-mastectomy radiation therapy; T, tumor; N, nodal; OS, overall survival; LRR, locoregional recurrence; DM, distant metastasis; DFS, disease-free survival.

Discussion

Indications of PMRT in patients with breast IDC who received NACT followed by TM remain controversial [5,6,8-10,40-43]. Indications of PMRT after NACT and TM in patients with breast cancer have deepened by clinical tumor size, lymph node status, and pathologic response in tumor or lymph nodes in previous studies but no solid conclusions on indications of PMRT have produced clear outcome benefits regarding PMRT on OS, DFS, LRR, or DM [5,6,8-10,40-43]. In some studies, PMRT has been considered for patients with any residual pathologic N stages after NACT, based on retrospective evidence that suggests a higher rate of recurrence in such patients [40]. PMRT has also been considered in patients with residual pathologic breast T stages, although the threshold to omit PMRT in such patients is lower than that for patients with residual pathologic N stages [6,8,10,40-43]. Without prospective data to guide the approach for patients with a pCR to NACT, some retrospective studies have suggested that patients who presented with clinical stage III disease based on AJCC stages be treated with PMRT, regardless of pathologic response [6,8,10,40-43]. The clinical AJCC stages appear to be valuable markers for indications of adjuvant PMRT in patients with breast cancer who receive NACT and TM [6,8,10,40-43]. Thus, most retrospective data in women with clinical stage III disease suggest that PMRT improves LRR, even in patients who have a pCR to NACT; however, the endpoints of previous studies have rarely been OS or DM [6,8,10,40-43]. For example, in one retrospective study that included over 670 women treated with NACT followed by TM, PMRT was associated with a significantly lower rate of LRR over 10 years (22% vs 11%) [6]. Among 46 patients who presented with stage III or IV disease and achieved a pCR with NACT, PMRT was associated with a decreased 10-year rate of LRR (3% vs 33% among patients not receiving PMRT) [6]. By contrast, other retrospective data suggested that certain patients who achieve a pCR with NACT have low rates of LRR following TM without using PMRT [5,9]. A large retrospective study of 3000 women treated with TM with or without PMRT discovered that PMRT was associated with only a modest reduction in 10-year LRR (10.3% vs 12.6% among patients who did not receive PMRT) [5]. Thus, PMRT may not be considered for such patients [5]. Therefore, pretreatment factors and patients’ pathologic response to NACT were evaluated, with a lower threshold of omission from PMRT [5,9]. In addition, another study reported that post-chemotherapy pathologic nodal status was not predictive of relative survival benefits obtained from PMRT [7]. Therefore, clear evidence and outcome benefits of PMRT with pretreatment factors is needed to better inform recommendations for PMRT after NACT and TM. Until now, the effects of PMRT on patients after NACT and TM in pretreatment clinical T and N stages after pathologic response is adjusted have been controversial. In this study, we estimated the effects of PMRT by determining the effects of pretreatment clinical T and N stages on patients receiving NACT and TM regardless of pathologic response because complex pathologic responses after NACT depend on various physiologic or molecular factors [44-47].

The is the first study to estimate clinical T or N stages and the effect of PMRT on OS, DM, LRR, and DFS in patients with breast cancer who received NACT and TM. The multivariate Cox proportional hazards model analysis of breast IDC after NACT and TM revealed some significant predictors of OS, DFS, DM, and LRR (Table 2). PMRT, pathologic response, tumor grade, CCI score, clinical T or clinical N stage, and being ER/PR-positive were independent predictors of OS. PMRT resulted in better OS, LRR, and DFS after multivariate analysis in patients with breast cancer who received NACT and TM (Table 2). No significant benefits appeared from applying PMRT on DM in patients with breast cancer who received NACT and TM. According to our literature review, no sufficient data support the advantages of PMRT in OS, LRR, DM, or DFS for patients with breast cancer who received NACT and TM. This is the first data to show the benefits of PMRT on OS, DFS, and LRR in patients with breast cancer who received NACT and TM, but no association appeared between PMRT and the risk of DM in these patients.

Our findings are compatible with those of previous studies in which PMRT, pathologic response, and clinical T or N stage were independent predictors of OS in patients with breast cancer who received NACT and TM [7,10]. Moreover, tumor grade, CCI ≥ 2, and being ER/PR negative were the first reported independent poor prognostic factors of OS for patients with breast cancer who received NACT and TM (Table 2). In our study, the predictors of OS in such patients were PMRT, tumor grade, being ER/PR negative, and clinical T or N stage before chemotherapy. In the past, whether these predictors of OS were attributed to distant or locoregional failure was unclear. In the current study, we clarified the effects of these predictors on distant and locoregional failure in patients with breast cancer who received NACT and TM. Poorly differentiated appearance resulted in poor OS because of a higher risk of DM and LRR. These findings are compatible with those of previous studies that have shown poorly differentiated breast cancer to result in a high risk of DM and a high LRR rate [48-50]. CCI ≥ 2 was an independent poor prognostic factor for OS but not a significant predictor of DM and LRR (Table 2). Higher CCI scores were associated with poor OS, which is compatible with the findings of previous studies that present comorbidity at breast cancer diagnosis as an independent adverse prognostic factor for death [51]. Our finding was reasonable regarding poor OS in patients with breast cancer with multicomorbidities who received NACT and TM compared with relatively healthy patients with CCI = 0. No association appeared between the risk of DM and LRR in patients with CCI ≥ 2. Clinical T stage before NACT was a predictor of OS, LRR, and DFS but not a significant predictor of DM (Table 2). Clinical T stage was a predictor of OS, LRR, and DFS for patients with breast cancer who received NACT and TM. The outcomes are similar to those of previous studies [5,6,8,52]. Moreover, a higher advanced clinical T stage was associated with a higher risk of LRR in breast cancer, rather than the risk of DM (Table 2). Clinical N stage was an independent predictor of OS, DM, LRR, and DFS in patients with breast cancer who received NACT and TM, which is compatible with similar outcomes from previous studies [5,6,8,52]. The difference between clinical T and N stage was in the effect of DM on patients who received NACT and TM. Higher advanced clinical N stage was proportional to a higher risk of DM; however, higher advanced T stage was not associated with a higher risk of DM. Pathologic response after NACT was a key predictor of OS, DM, LRR, and DFS in patients who received NACT and TM. This is the first study to discover that pCR and pPR were predictors of OS, DM, LRR, and DFS in patients with breast cancer who received NACT and TM. In addition, being ER/PR positive was a good predictor of OS and DFS in such patients, and this result is similar to that of a study by Buchholz et al., who reported a lower LRR risk in ER/PR-positive patients with hormone therapy [6]. Our study is the first to indicate that being HER2 positive was a poor prognostic factor of OS, DM, DFS, and LRR in patients with breast cancer who received NACT and TM. This finding may have resulted because having a HER2-positive disease (along with a negative disease) is associated with a lower pathologic response rate after NACT [44,53,54].

In summary, this is the first study to show significant predictors of OS, LRR, DM, and DFS in patients with breast cancer who received NACT and TM. Moreover, PMRT is a crucial adjuvant treatment and improved OS, LRR, and DFS in such patients regardless of their pathologic response after multivariate analysis. We clarified the benefits of PMRT by using clinical stage instead of pathologic response (Figure 1) because different chemotherapy regimens, tumor differentiation, molecular appearance (HER2 and ER/PR-positivity), and body index mass are associated with different pathologic responses [24-26,45,55,56]. Indications of PMRT in patients can undergo interference from complicated factors in pathologic response [6,8,10,40-43]. Thus, simplifying indications of PMRT by using clinical T or N stages can be valuable and help make further decisions regarding PMRT. The clear benefits of PMRT in OS, DM, LRR, and DFS are clarified in Figure 1.

Our study is the first to estimate the effect of PMRT on OS, LRR, DM, and DFS in patients with breast cancer who received NACT and TM who were stratified by clinical T and N stage after multivariable Cox regression analysis. Our study is a good reference for using clinical T or N stages for adjuvant PMRT in patients with breast cancer after receiving NACT and TM. Our study showed that PMRT for clinical T3N0 improved OS regardless of pathologic response (Figure 1A). The findings are compatible with those of previous studies [10,57]. However, previous studies have not focused on detailed outcomes of OS, LRR, DM, or DFS [10,57], and our study presented the clear benefits of PMRT in patients with breast IDC who received NACT and TM. In our study, PMRT was beneficial for OS for clinical T3N0 (Figure 1A), for LRR for clinical T2N0 (Figure 1C), and for DFS for clinical T2N0 (Figure 1D). No clear benefits appeared from PMRT for DM in patients with breast cancer who received NACT and TM (Figure 1B). Differences in PMRT in DM for patients with breast cancer who received NACT and TM may have been attributed to complex reactions between NACT and pathologic responses associated with various risks from DM.

Our study is the largest cohort study in Taiwan to estimate outcomes of PMRT for patients with breast cancer regarding OS, LRR, DM, and DFS depending on clinical T or N stages. The treatment of PMRT and regimens of NACT were homogenous in our study. Few studies have estimated the effects of PMRT for detailed outcomes of OS, LRR, DFS, and DM in patients with breast cancer who received NACT and TM, and all covariates including pathologic response were adjusted. In our study, poor prognostic factors for OS in such patients were non-PMRT, higher clinical T or N stage before NACT, poorly differentiated tumor grade, pPR, pSD, pPD, CCI ≥ 2, being ER/PR negative, and being HER2 positive (Table 2). Clinical T stage and CCI were not associated with DM, but clinical T stage was associated with LRR and DFS. From the multivariable Cox regression analysis for patients who received NACT and TM with or without PMRT, PMRT resulted in superior OS in clinical T3, T4, and N0-3 regardless of pathologic response and other predictors (Figure 1A). In our study, the effect of PMRT was significantly superior for OS in clinical T3N0-T4N3, for LRR-free survival in clinical T2N0-T4N3, and for DFS in clinical T2N0-T4N3 to non-PMRT patients regardless of pathologic response and other predictors (Figure 1A-D).

This study has some limitations. First, because all patients with breast IDC were enrolled from an Asian population, the corresponding ethnic susceptibility compared with the non-Asian population remains unclear; hence, our results should be cautiously extrapolated to non-Asian populations. However, no evidence indicates any differences in outcomes from PMRT among patients with breast cancer who received NACT and TM between Asian and non-Asian populations. Second, the diagnoses of all comorbid conditions were based on ICD-9-CM codes. Nevertheless, the Taiwan Cancer Registry Administration randomly reviews charts and interviews patients to verify the accuracy of these diagnoses, and hospitals with outlier chargers or practices may be audited and subsequently be heavily penalized if malpractice or discrepancies are identified. Third, to prevent the creation of several subgroups, various neoadjuvant treatments were not categorized separately during analysis. Thus, the effects of different neoadjuvant treatments remain unclear. Accordingly, to obtain crucial information on population specificity and disease occurrence, a large-scale randomized trial comparing carefully selected patients undergoing suitable treatments is essential. Finally, the TCRD does not contain information regarding dietary habits, socioeconomic status, or body mass index, all of which may be risk factors for mortality. However, given the magnitude and statistical significance of observed effects in this study, these limitations are unlikely to affect conclusions.

Conclusions

PMRT significantly improved OS in clinical T3N0-T4N3 and LRR-free survival and DFS in clinical T2N0-T4N3 from those of non-PMRT patients regardless of pathologic response and other predictors.

Disclosure of conflict of interest

None.

Abbreviations

PMRT

post-mastectomy radiation therapy

T

tumor

N

nodal

OS

overall survival

LRR

locoregional recurrence

DM

distant metastasis

DFS

disease-free survival

NACT

neoadjuvant chemotherapy

TM

total mastectomy

HRs

hazard ratios

CI

confidence intervals

IDC

invasive intraductal carcinoma

TCRD

Taiwan Cancer Registry database

AJCC

American Joint Committee on Cancer

ER

estrogen receptor

PR

progesterone receptor

HER2

human epidermal growth factor receptor 2

ICD-9-CM

International Classification of Diseases, Ninth Revision, Clinical Modification

pCR

pathological complete response

pPR

pathologic partial response

pSD

pathological stationary disease

pPD

pathological progression of disease

ALND

axillary lymph node dissection

SNLB

sentinel lymph node biopsy