This study aimed to investigate the impact of postoperative radiotherapy (PORT) in
This study enrolled 426 patients with dnMBC administered PTR alone or with PORT. The primary and secondary outcomes were overall and progression-free survival (OS and PFS), respectively.
The median follow-up time was 53.7 months (range, 3.1 to 194.4). The 5-year OS and PFS rates were 73.2% and 32.0%, respectively. For OS, clinical T3/4 category, triple-negative breast cancer (TNBC), postoperative chemotherapy alone were significantly poor prognostic factors, and administration of PORT failed to show its significance. Regarding PFS, PORT was a favorable prognostic factor (hazard ratio, 0.64; 95% confidence interval, 0.50 to 0.82; p < 0.001), in addition to T1/2 category, ≤ 5 metastases, and non-TNBC. According to the multivariate analyses of OS in the PORT group, we divided the patients into three groups (group 1, T1/2 and non-TNBC [n=193]; group 2, T3/4 and non-TNBC [n=171]; and group 3, TNBC [n=49]), and evaluated the effect of PORT. Although PORT had no significance for OS in all subgroups, it was a significant factor for good prognosis regarding PFS in groups 1 and 2, not in group 3.
PORT was associated with a significantly better PFS in patients with dnMBC who underwent PTR. Patients with clinical T1/2 category and non-TNBC benefited most from PORT, while those with TNBC showed little benefit.
Breast cancer is the most common malignancy in women worldwide. Approximately 5%–8% of patients with breast cancer are initially diagnosed with distant metastases, known as
Locoregional treatment (LRT) including primary tumor resection (PTR) and/or radiotherapy (RT) has arisen not only for local control but also for improving survival. However, this use remains controversial, and there are no clear guidelines. Three prospective randomized controlled trials (RCTs) have compared PTR to systemic therapy alone [
Despite conflicting results from RCTs, 35%–80% of patients with stage IV breast cancer undergo PTR [
To reveal the effects of PORT, this retrospective multicenter study investigated a cohort of patients with dnMBC who underwent PTR. We also aimed to identify the subset of patients that would most benefit from PORT.
Data from patients who were diagnosed with dnMBC and underwent PTR between October 2000 and December 2015 at 15 cancer centers were retrospectively analyzed. Patients with (1) palliative surgery due to bleeding, necrosis, or infection; (2) progression after preoperative chemotherapy or within 2 months after PTR; (3) bilateral breast cancer; (4) no information on RT; and (5) double primary cancer other than thyroid or skin cancer were excluded. Finally, a total of 426 patients were enrolled.
Staging was performed based on the American Joint Committee on Cancer 7th edition. The initial workup modalities included physical examination, mammography, ultrasound, pathologic confirmation, chest computed tomography (CT), and abdominopelvic CT. Brain imaging was additionally performed in patients with symptoms suggestive of central nerve system involvement. 18-Fluoro-deoxyglucose positron emission tomography (FDG-PET) was optional, but most patients underwent FDG-PET CT (50.2%) or FDG-PET alone (35.9%). Biopsy confirmation for metastatic sites was performed in 19.0% of patients. Metastatic burden was categorized according to (1) number (≤ 5 or > 5) and (2) single, oligo, and disseminated categories according to Kobayashi et al. [
Multidisciplinary teams including a surgeon, a medical oncologist, a radiation oncologist, and a radiologist developed the management plans. The surgeon determined the types of PTR and nodal evaluation (axillary LN dissection, sentinel LN biopsy, or none) based on individual tumor and patient characteristics. Preoperative and/or postoperative chemotherapy was administered to selected patients. Based on immunohistochemistry (IHC) markers of estrogen receptor (ER), progesterone receptor (PR), and human epidermal growth factor 2 (HER2), hormonal therapy and/or anti-HER2 therapy (trastuzumab and/or pertuzumab) were administered. IHC types were defined as follows; (1) luminal A (ER and/or PR positive, but HER2 negative), (2) luminal B (ER and/or PR positive, and HER2 positive), (3) HER2-enriched (ER and PR negative, but HER2 positive) and (4) triple-negative. Treatment responses were assessed according to the Response Evaluation Criteria in Solid Tumors ver. 1.1 [
The primary outcome of the present study was 5-year OS. The secondary endpoint was 5-year PFS. Progression included locoregional recurrence (LRR), a failure in the breast/chest wall or regional nodal areas, and distant progression (DP), the appearance of any new lesions, and/or a ≥ 20% increase in the sum of the diameters of the target metastatic lesions compared to the last preoperative measurements. Follow-up durations and survivals were calculated from the date of PTR.
The characteristics of patients in each treatment arm were compared using Student’s t tests. Univariate and multivariate analyses were performed using Cox proportional hazards models. Backward elimination Cox regression was used for multivariate analysis. Factors with p ≤ 0.2 in univariate analysis were included in the multivariate analysis. The Kaplan-Meier method was used to estimate survival. To identify the subgroup of patients who most benefited after PORT, univariate and multivariate analyses were also performed among patients who received PORT. Using the statistically significant factors, we defined subgroups and compared the primary and secondary endpoints between patients who did and did not receive PORT. All statistical analyses were performed using Web-R ver. 3.4.1 (
Among the 426 patients, 176 and 250 patients received PTR alone and PORT, respectively. The characteristics of the patients according to PORT are listed in
The median follow-up time was 53.7 months (range, 3.1 to 194.4 months). The 5-year OS and PFS of the overall patient population were 73.2% (
Univariate and multivariate analyses of patients who had undergone PORT were performed to identify the subgroup of patients that had most benefited from PORT (
Univariate and multivariate analyses in group 1, the most favorable group, showed that no factor was significantly associated with OS (
LRR occurred in 65 patients, with a 5-year LRR incidence rate of 15.7%. Thirty cases of LRR (46.2%) coincided with DP and eight failures occurred after DP. LRR preceded DP in 17 patients with a median interval of 9.7 months (range, 1.9 to 33.7 months). The breast/chest wall was the most common site of LRR (34 patients). Among these patients, 10 also had the regional nodal failure. Axilla and supraclavicular recurrences occurred in 23 and 21 patients each, while four patients had internal mammary node recurrence. Among the 65 patients with LRR, 30 received PORT, and all but two cases were in-field recurrences. Multivariate analyses of LRR in the overall population showed significant associations between lower LRR rates and lower N category (N0/1), bone metastases, and PORT (HR, 0.43; 95% CI, 0.26 to 0.71; p=0.001). Among the subgroups, PORT was a significant factor for LRR only in group 1 (HR, 0.36; 95% CI, 0.15 to 0.84; p=0.019) with a metastatic burden. However, PORT had no impact on the LRR rate in other groups, and neither did other factors.
A total of 290 patients experienced DP, with a 5-year incidence rate of 65.9%. PORT was a meaningful prognostic factor for DP in the multivariate analyses (HR, 0.71; 95% CI, 0.55 to 0.91; p=0.008). Following factors were also significant factors regarding DP; T3/4 category, number of metastases over 5, TNBC, and PTR type of mastectomy. Among subgroups, PORT was significantly associated with lower DP in group 1 (HR, 0.58; 95% CI, 0.40 to 0.86; p=0.006) with the number of metastases ≤ 5 and breast-conserving operation. PORT also had significant impact in group 2 (HR, 0.67; 95% CI, 0.46 to 0.96; p=0.029) with single metastatic burden. However, PORT did not impact DP in group 3, while no administration of chemotherapy (HR, 6.61; 95% CI, 1.28 to 34.1; p=0.024) and T3/4 (HR, 3.02; 95% CI, 1.13 to 8.04; p=0.027) disease status were poor prognostic factors.
The value of PTR for dnMBC is still controversial despite three RCTs comparing systemic therapy plus PTR with systemic therapy alone [
The main goal of this study was to evaluate the impact of PORT. Although PORT failed to show a significant improvement in OS, it was one of the significant factors associated with PFS. PORT decreased not only LRR but also DP. However, there was a fundamental limitation in this study that the no-PORT group included more patients with the disseminated metastatic burden. Also, the PORT group might be more amenable to receive other systemic therapy or RT for metastatic lesions. However, some studies have also reported improved PFS after PORT [
Among the three subgroups, patients with TNBC showed no PFS benefit from PORT. This might be attributed to the fact that no effective systemic therapy is available for TNBC compared with the other subtypes. Before the advent of anti-HER2 therapy, HER2 positivity in breast cancer was associated with rapid tumor proliferation, shorter disease-free survival, and poorer OS [
Group 1 (T1/2 and non-TNBC) most benefited from PORT. In this group, the OS showed a trend favoring PORT and the LRR and DP were significantly decreased by PORT. However, PORT did not affect the LRR in group 2 (T3/4 and non-TNBC). Patients with clinical T1/2 disease more often received breast-conserving operations compared to patients with T3/4 disease; thus, the decrease in LRR might be more dramatic in group 1 than that in group 2. A recent review proposed a decision algorithm to select the optimal candidates for LRT in dnMBC and reported that LRT could be an option for ER/PR or HER2-positive patients [
The present study had several limitations. First, the characteristics differed significantly between patients in the PORT and no-PORT groups. Statistical analysis to compensate for this limitation such as propensity score matching was not feasible due to the small number of patients. Moreover, we did not collect data on toxicities, which is important clinical data for certain treatments. PTR may increase the risks of infection, hematoma, and lymphedema, while PORT may cause radiation dermatitis, esophagitis, fibrosis, etc. [
PORT was associated with a significantly better PFS in patients with dnMBC who received PTR. Patients with clinical T1/2 and non-TNBC disease benefited most from PORT, while PORT had little effect on those with TNBC. The retrospective design of the study prevented us from obtaining direct evidence to support PORT. However, our results suggest that further studies on LRT for dnMBC should focus on diseases with favorable prognoses such as ER/PR or HER2-positive tumors.
Supplementary materials are available at Cancer Research and Treatment website (
The study protocol was approved by the institutional review boards of 15 centers, which waived the requirement for informed consent.
Conceived and designed the analysis: Kim SS.
Collected the data: Kim YJ (Yeon Joo Kim), Kim YJ (Yeon-Joo Kim), Kim YB, Lee IJ, Kwon J, Kim K, Cha J, Kim M, Jo IY, Kim JH, Park J, Kim JH, Kim J, Shin KH.
Contributed data or analysis tools: Kim YJ (Yeon Joo Kim), Kim YJ (Yeon-Joo Kim), Kim YB, Lee IJ, Kwon J, Kim K, Cha J, Kim M, Jo IY, Kim JH, Park J, Kim JH, Kim J, Shin KH.
Performed the analysis: Kim YJ (Yeon Joo Kim).
Wrote the paper: Kim YJ (Yeon Joo Kim).
Provided revisions of the paper: Shin KH, Kim SS.
Conflict of interest relevant to this article was not reported.
This work was supported by a National Research Foundation (NRF) grant funded by the Korean government (NRF-2018R1D1A1B07049970).
We would like to thank Editage (
Overall survival and progression-free survival according to the administration of postoperative radiotherapy (PORT) in group 1 (A, B), group 2 (C, D) and group 3 (E, F).
Univariate and multivariate analyses of factors predictive of OS and PFS in the overall patient population
Characteristic | OS | PFS | ||||||
---|---|---|---|---|---|---|---|---|
|
| |||||||
Univariate | Multivariate | Univariate | Multivariate | |||||
|
|
|
| |||||
HR (95% CI) | p-value | HR (95% CI) | p-value | HR (95% CI) | p-value | HR (95% CI) | p-value | |
| ||||||||
≥ 47 (vs. <47) | 1.08 (0.76–1.53) | 0.677 | - | - | 0.86 (0.69–1.08) | 0.206 | - | - |
| ||||||||
| ||||||||
T3/4 (vs. T1/2) | 1.80 (1.26–2.57) | 0.001 | 1.83 (1.23–2.71) | 0.003 | 1.76 (1.40–2.21) | < 0.001 | 1.62 (1.26–2.09) | < 0.001 |
| ||||||||
| ||||||||
N0/1 (vs. N2/3) | 1.03 (0.72–1.47) | 0.859 | - | - | 0.83 (0.65–1.04) | 0.106 | - | - |
| ||||||||
| ||||||||
Bone and others (vs. bone only) | 1.83 (1.12–3.00) | 0.016 | - | - | 1.82 (1.34–2.47) | < 0.001 | - | - |
| ||||||||
Others (vs. bone only) | 1.35 (0.90–2.03) | 0.142 | - | - | 1.02 (0.79–1.31) | 0.907 | - | - |
| ||||||||
| ||||||||
Single (vs. disseminated) | 0.55 (0.35–0.88) | 0.013 | - | - | 0.46 (0.34–0.61) | < 0.001 | - | - |
| ||||||||
Oligo (vs. disseminated) | 0.90 (0.59–1.35) | 0.598 | - | - | 0.60 (0.46–0.78) | < 0.001 | - | - |
| ||||||||
| ||||||||
>5 (vs. ≤ 5) | 1.49 (1.04–2.14) | 0.032 | 1.53 (1.06–2.23) | 0.025 | 2.00 (1.58–2.52) | < 0.001 | 1.81 (1.41–2.32) | < 0.001 |
| ||||||||
| ||||||||
Luminal A (vs. triple-negative) | 0.36 (0.22–0.59) | < 0.001 | 0.25 (0.15–0.41) | < 0.001 | 0.65 (0.46–0.97) | 0.021 | 0.55 (0.38–0.79) | 0.001 |
| ||||||||
Luminal B (vs. triple-negative) | 0.37 (0.21–0.67) | < 0.001 | 0.32 (0.18–0.59) | < 0.001 | 0.76 (0.51–1.14) | 0.186 | 0.65 (0.43–0.99) | 0.045 |
| ||||||||
HER2-enriched (vs. triple-negative) | 0.36 (0.19–0.69) | 0.002 | 0.27 (0.13–0.53) | < 0.001 | 0.63 (0.40–0.99) | 0.045 | 0.52 (0.33–0.82) | 0.005 |
| ||||||||
| ||||||||
Preoperative (vs. both) | 0.80 (0.52–1.25) | 0.333 | 1.03 (0.65–1.63) | 0.899 | 0.73 (0.56–0.96) | 0.022 | - | - |
| ||||||||
Postoperative (vs. both) | 1.45 (0.92–2.28) | 0.112 | 2.44 (1.47–4.04) | 0.001 | 0.80 (0.58–1.10) | 0.164 | - | - |
| ||||||||
None (vs. both) | 1.02 (0.45–2.30) | 0.962 | 1.59 (0.69–3.63) | 0.277 | 1.01 (0.62–1.63) | 0.977 | - | - |
| ||||||||
| ||||||||
Mastectomy (vs. breast-conserving) | 1.72 (1.14–2.61) | 0.011 | 1.52 (0.97–2.39) | 0.685 | 1.66 (1.28–2.15) | < 0.001 | 1.24 (0.92–1.67) | 0.159 |
| ||||||||
| ||||||||
No PORT (vs. PORT) | 0.67 (0.47–0.95) | 0.024 | - | - | 0.52 (0.41–0.65) | < 0.001 | 0.64 (0.50–0.82) | < 0.001 |
| ||||||||
| ||||||||
Yes (vs. no) | 0.86 (0.58–1.27) | 0.442 | - | - | 0.68 (0.52–0.88) | 0.004 | - | - |
CI, confidence interval; HER2, human epidermal growth factor 2; HR, hazard ratio; IHC, immunohistochemistry; OS, overall survival; PFS, progression-free survival; PORT, postoperative radiotherapy; PTR, primary tumor resection.
Univariate and multivariate analyses of factors predictive of OS and PFS among patients with clinical T1/2 category disease excluding those with triple-negative molecular subtypes (n=193)
Characteristic | OS | PFS | ||||||
---|---|---|---|---|---|---|---|---|
|
| |||||||
Univariate | Multivariate | Univariate | Multivariate | |||||
|
|
|
| |||||
HR (95% CI) | p-value | HR (95% CI) | p-value | HR (95% CI) | p-value | HR (95% CI) | p-value | |
| ||||||||
≥ 47 (vs. < 47) | 0.83 (0.45–1.52) | 0.540 | - | - | 0.71 (0.50–1.02) | 0.637 | 0.72 (0.50–1.04) | 0.078 |
| ||||||||
| ||||||||
N0/1 (vs. N2/3) | 1.23 (0.67–2.28) | 0.504 | - | - | 1.00 (0.69–1.43) | 0.988 | - | - |
| ||||||||
| ||||||||
Bone and others (vs. bone only) | 0.86 (0.29–2.52) | 0.780 | - | - | 1.37 (0.79–2.40) | 0.266 | - | - |
| ||||||||
Others (vs. bone only) | 0.97 (0.51–1.83) | 0.920 | - | - | 0.92 (0.63–1.36) | 0.690 | - | - |
| ||||||||
| ||||||||
Single (vs. disseminated) | 0.48 (0.21–1.08) | 0.076 | 0.54 (0.24–1.24) | 0.148 | 0.51 (0.33–0.81) | 0.004 | - | - |
| ||||||||
Oligo (vs. disseminated) | 0.99 (0.48–2.04) | 0.974 | 1.16 (0.55–2.47) | 0.691 | 0.65 (0.41–1.01) | 0.053 | - | - |
| ||||||||
| ||||||||
> 5 (vs. ≤ 5) | 1.25 (0.66–2.39) | 0.493 | - | - | 1.75 (1.20–2.54) | 0.004 | 1.63 (1.11–2.40) | 0.013 |
| ||||||||
| ||||||||
Luminal A (vs. HER2-enriched) | 0.95 (0.41–2.19) | 0.896 | 1.11 (0.53–2.30) | 0.790 | 1.07 (0.63–1.81) | 0.810 | - | - |
| ||||||||
Luminal B (vs. HER2-enriched) | 1.04 (0.39–2.73) | 0.945 | 1.06 (0.44–2.51) | 0.903 | 1.18 (0.65–2.15) | 0.590 | - | - |
| ||||||||
| ||||||||
Preoperative (vs. both) | 1.36 (0.49–3.74) | 0.555 | - | - | 0.71 (0.44–1.14) | 0.158 | - | - |
| ||||||||
Postoperative (vs. both) | 2.43 (0.90–6.54) | 0.080 | - | - | 0.71 (0.42–1.18) | 0.898 | - | - |
| ||||||||
None (vs. both) | 2.03 (0.54–7.57) | 0.292 | - | - | 0.96 (0.48–1.90) | 0.182 | - | - |
| ||||||||
| ||||||||
Mastectomy (vs. breast-conserving) | 1.65 (0.88–3.10) | 0.121 | - | - | 1.64 (1.13–2.36) | 0.009 | 1.46 (1.00–2.14) | 0.053 |
| ||||||||
| ||||||||
No PORT (vs. PORT) | 0.59 (0.32–1.07) | 0.084 | 0.59 (0.31–1.10) | 0.096 | 0.46 (0.32–0.67) < | 0.001 | 0.57 (0.39–0.83) | 0.004 |
| ||||||||
| ||||||||
Yes (vs. no) | 0.76 (0.39–1.49) | 0.422 | - | - | 0.76 (0.52–1.11) | 0.156 | - | - |
CI, confidence interval; HER2, human epidermal growth factor 2; HR, hazard ratio; IHC, immunohistochemistry; OS, overall survival; PFS, progression-free survival; PORT, postoperative radiotherapy; PTR, primary tumor resection.
Univariate and multivariate analyses of factors predictive of OS and PFS among patients with clinical T3/4 category excluding those with triple-negative immunohistochemistry subtypes (n=171)
Characteristic | OS | PFS | ||||||
---|---|---|---|---|---|---|---|---|
|
| |||||||
Univariate | Multivariate | Univariate | Multivariate | |||||
|
|
|
| |||||
HR (95% CI) | p-value | HR (95% CI) | p-value | HR (95% CI) | p-value | HR (95% CI) | p-value | |
| ||||||||
≥ 47 (vs. < 47) | 1.21 (0.73–2.03) | 0.463 | - | - | 1.06 (0.76–1.49) | 0.733 | - | - |
| ||||||||
| ||||||||
N0/1 (vs. N2/3) | 1.28 (0.75–2.20) | 0.362 | - | - | 1.06 (0.73–1.54) | 0.762 | - | - |
| ||||||||
| ||||||||
Bone and others (vs. bone only) | 2.67 (1.29–5.52) | 0.008 | 2.67 (1.29–5.52) | 0.008 | 1.69 (1.09–2.62) | 0.020 | - | - |
| ||||||||
Others (vs. bone only) | 1.97 (1.00–3.88) | 0.049 | 1.97 (1.00–3.88) | 0.049 | 1.00 (0.66–1.50) | 0.988 | - | - |
| ||||||||
| ||||||||
Single (vs. disseminated) | 0.64 (0.32–1.28) | 0.210 | - | - | 0.43 (0.27–0.67) | < 0.001 | - | - |
| ||||||||
Oligo (vs. disseminated) | 0.89 (0.50–1.60) | 0.701 | - | - | 0.58 (0.40–0.86) | 0.006 | - | - |
| ||||||||
| ||||||||
> 5 (vs. ≤ 5) | 1.43 (0.85–2.41) | 0.179 | - | - | 2.06 (1.46–2.92) | < 0.001 | 1.76 (1.23–2.56) | 0.003 |
| ||||||||
| ||||||||
Luminal A (vs. HER2-enriched) | 1.13 (0.52–2.43) | 0.759 | - | - | 1.08 (0.67–1.76) | 0.746 | - | - |
| ||||||||
Luminal B (vs. HER2-enriched) | 1.15 (0.48–2.75) | 0.748 | - | - | 1.30 (0.76–2.24) | 0.337 | - | - |
| ||||||||
| ||||||||
Preoperative (vs. both) | 1.07 (0.59–1.94) | 0.824 | - | - | 0.91 (0.63–1.32) | 0.616 | - | - |
| ||||||||
Postoperative (vs. both) | 1.77 (0.86–3.63) | 0.120 | - | - | 1.14 (0.66–1.94) | 0.644 | - | - |
| ||||||||
None (vs. both) | 1.88 (0.56–6.37) | 0.309 | - | - | 1.44 (0.62–3.35) | 0.402 | - | - |
| ||||||||
| ||||||||
Mastectomy (vs. breast-conserving) | 1.25 (0.56–2.79) | 0.582 | - | - | 1.32 (0.76–2.30) | 0.329 | - | - |
| ||||||||
| ||||||||
No PORT (vs. PORT) | 0.95 (0.56–1.61) | 0.860 | - | - | 0.55 (0.39–0.77) | < 0.001 | 0.66 (0.46–0.97) | 0.032 |
| ||||||||
| ||||||||
Yes (vs. no) | 1.06 (0.58–1.94) | 0.839 | - | - | 0.61 (0.39–0.95) | 0.030 | - | - |
CI, confidence interval; HER2, human epidermal growth factor 2; HR, hazard ratio; IHC, immunohistochemistry; OS, overall survival; PFS, progression-free survival; PORT, postoperative radiotherapy; PTR, primary tumor resection.