Radiotherapy and Oncology 132 (2019) 48–54

Contents lists available at ScienceDirect

Radiotherapy and Oncology

journal homepage: www.thegreenjournal.com

Original article

Postmastectomy radiation therapy for triple negative, node-negative

breast cancer
Waqar Haque a,⇑, Vivek Verma b, Andrew Farach b, E. Brian Butler a, Bin S. Teh a
a
Department of Radiation Oncology, Houston Methodist Hospital; and b Department of Radiation Oncology, Allegheny General Hospital, Pittsburgh, USA

a r t i c l e i n f o a b s t r a c t

Article history: Purpose: The use of post-mastectomy radiation therapy (PMRT) for patients with node-negative, triple
Received 2 October 2018 negative breast cancer (TNBC) is controversial. This study of a large, contemporary US database described
Received in revised form 9 November 2018 national practice patterns and addressed the impact of PMRT on survival for patients with node-negative
Accepted 18 November 2018
TNBC.
Methods: The National Cancer Data Base was queried (2004–2014) for women with non-metastatic TNBC
with pT1-4N0M0 disease undergoing mastectomy. Use of PMRT was assessed. Multivariable logistic
Keywords:
regression ascertained factors associated with PMRT use. The Kaplan–Meier analysis evaluated overall
Breast cancer
Triple negative breast cancer
survival (OS) between patients managed with either PMRT or observation following mastectomy when
Radiation therapy stratifying by pT stage. Cox proportional hazards modeling determined variables associated with OS.
Chemotherapy Results: A total of 14,464 patients met the selection criteria; of these, 1,569 (10.8%) received PMRT,
Lumpectomy whereas 12,895 (89.2%) did not receive PMRT. Use of PMRT varied significantly with pT stage, with only
Mastectomy 5.7% of T1 patients undergoing PMRT, while 51.6% of patients with T3 disease underwent PMRT. Use of
PMRT was associated with superior OS for patients with pT3 disease but not for patients with other T
stages. Greater age was associated with decreased likelihood of PMRT use, while increased T stage and
positive surgical margins were associated with use of PMRT. On multivariate analysis, increased age, T
stage, and positive surgical margins were associated with worse OS.
Conclusions: In the largest study to date evaluating the use of PMRT in patients with node-negative TNBC,
the use of PMRT was low in patients with T1 and T2 disease. Additionally, while an OS benefit was
observed with the use of PMRT in patients with T3 disease, there was no benefit with the use of PMRT
in other T stage groups. Further prospective studies are recommended to further elucidate the benefit
on PMRT in patients with node-negative TNBC.
Ó 2018 Elsevier B.V. All rights reserved. Radiotherapy and Oncology xxx (2018) xxx–xxx

The National Comprehensive Cancer Network (NCCN) currently for patients with breast cancer today. Namely, women in the Dan-
recommends post-mastectomy radiation therapy (PMRT) for inva- ish 82b trial conducted in premenopausal women, received
sive breast cancer with 4 positive axillary nodes, with strong con- chemotherapy with Cyclophosphamide, Methotrexate, and Fluo-
sideration for 1–3 positive nodes, and a consideration for negative rouracil (CMF) (2); women in the Danish 82c trial, conducted in
axillary nodes but >5 cm primary tumor [1]. This recommendation postmenopausal women, received Tamoxifen alone (3); and
is based on three randomized trials that have shown an overall sur- women in the British Columbia trial also received CMF chemother-
vival (OS) benefit with the use of PMRT in patients with breast can- apy (4).
cer with positive surgical margins, tumor size >5 cm, or positive The aforementioned trials describing the efficacy of PMRT were
axillary lymph nodes, as well as a meta-analysis demonstrating performed before the era in which tumor markers were used to
higher OS with PMRT in women with breast cancer and positive guide treatment management. The current era of cancer manage-
axillary nodal disease [2–5]. It should be noted that while these tri- ment, however, is increasingly shifting toward personalized medi-
als guide current management regarding indications for PMRT, the cine. To this extent, numerous studies have examined the
systemic treatment used was different than what is typically used predictive and/or prognostic impacts of gene panels on various end-
points [6,7]. In some instances, gene profiling has demonstrated a
greater degree of accuracy than clinicopathologic factors in predict-
⇑ Corresponding author at: Department of Radiation Oncology, Houston Metho- ing the risk of distant recurrence following surgery [8–10].
dist Hospital, Cancer Center and Research Institute, Weil Cornell Medical College, As such, a major focus of ongoing investigation is to determine
Houston, TX 77030, USA.
whether there may be subsets of patients with more biologically
E-mail address: [email protected] (W. Haque).

https://doi.org/10.1016/j.radonc.2018.11.012
0167-8140/Ó 2018 Elsevier B.V. All rights reserved.
 W. Haque et al. / Radiotherapy and Oncology 132 (2019) 48–54 49

aggressive disease, such as patients with triple negative breast can- patients and consequently this variable was not factored into the
cer (TNBC), who may warrant PMRT even in the setting of early analysis. In accordance with the variables in NCDB files, informa-
stage, node-negative disease. One large retrospective review from tion collected on each patient broadly included demographic, clin-
Canada demonstrated that women with T1-2N0 TNBC treated with ical, and treatment data.
modified radical mastectomy without PMRT had worse local con- All statistical tests were two-sided, with a threshold of p < 0.05
trol than patients treated with breast conserving surgery and radi- for statistical significance, and were performed using STATA (ver-
ation therapy [11]. Furthermore, a randomized Chinese trial has sion 14, College Station, TX). Multivariable logistic regression mod-
demonstrated that the addition of PMRT to patients with stage I- eling determined characteristics predictive for PMRT. Survival
II TNBC led to improvements in 5 year OS [12]. analysis was per the Kaplan–Meier method, with group compar-
Given the current dearth of information, the NCCN does not isons done with the log-rank test. When conducting the Kaplan–
incorporate tumor markers into recommendations regarding Meier test to compare OS between patients either receiving or
TNBC. The goal of this large, national database investigation was not receiving PMRT, patients were stratified by T stage. OS referred
to evaluate trends and practice patterns of PMRT in patients with to the interval between the date of diagnosis and the date of death,
node-negative TNBC, along with assessing the clinical benefit from or censored at last contact. Univariate analysis determined factors
PMRT in this patient population. associated with overall survival; subsequently, Cox multivariate
analysis was performed and included variables that were signifi-
cant on univariate analysis with a p value < 0.05.
Materials & methods

The National Cancer Data Base (NCDB) is a joint project of the Results
Commission on Cancer (CoC) of the American College of Surgeons
and the American Cancer Society, which consists of de-identified A complete patient selection diagram is presented in Fig. 1. A
information regarding tumor characteristics, patient demograph- total of 14,464 patients met the selection criteria (Table 1). A total
ics, and patient survival for approximately 70% of the US
population [13–15]. All pertinent cases are reported regularly from
CoC-accredited centers and compiled into a unified dataset, which Table 1
Demographic and clinical characteristics for all patients.
is then validated. The NCDB contains information not included in
the Surveillance, Epidemiology, and End Results (SEER) database, Characteristic No radiation n = 12895 Radiation n = 1569 P
including details regarding use of systemic therapy. The data used (%) (%) value
in the study were derived from a de-identified NCDB file (2004– Age
2013). The American College of Surgeons and the CoC have not ver- 50 3882 (30.1%) 628 (40.0%) <0.001
51–60 3259 (25.3%) 436 (27.8%)
ified and are neither responsible for the analytic or statistical
61–70 2972 (23.1%) 284 (18.1%)
methodology employed nor the conclusions drawn from these data 71+ 2782 (21.6%) 221 (14.1%)
by the investigators. As all patient information in the NCDB data-
Race
base is de-identified, this study was exempt from institutional White 10,465 (81.2%) 1188 (75.7%) <0.001
review board evaluation. African 1807 (14.0%) 307 (19.6%)
Inclusion criteria for this study were women with newly diag- American
nosed, non-metastatic TNBC with clinically and pathologically N0 Other 623 (4.8%) 74 (4.7%)

disease and pT1-4 disease treated with mastectomy. Patients with Charlson–Deyo’s Score
either clinically or pathologically node-positive disease were 0 10,431 (80.9%) 1312 (83.6%) 0.032
1 1944 (15.1%) 295 (13.1%)
excluded from the present analysis. A record of radiation therapy
2 520 (4.0%) 52 (3.3%)
use was required for inclusion in the study, and patients included
Insurance status
in the PMRT category were treated with conventionally fraction-
Medicaid 896 (7.0%) 213 (13.6%) <0.001
ated radiation. Information regarding whether or not the regional Private 7058 (54.7%) 873 (55.6%)
nodes were included in the field in addition to the chest wall in Medicare 4412 (34.2%) 378 (24.1%)
the radiation therapy field was not available for a majority of Not insured 260 (2.0%) 71 (4.5%)
Other 269 (2.1%) 34 (2.2%)

National Cancer Data Base Median Income

Triple negative breast cancer $62999 8655 (67.1%) 1127 (71.8%) 0.001
Diagnosed 2004-2013 $63000 4196 (32.5%) 438 (27.9%)
(N=106,982) Not recorded 44 (0.3%) 4 (0.3%)
Facility type
Excluded (N=92,518) Non academic 8180 (63.4%) 950 (60.6%) <0.001
M+ disease (N=12,260) Academic 3537 (27.4%) 376 (24.0%)
cN+ and/or pN+ (N=55,618) Not recorded 1178 (9.1%) 243 (15.5%)
No recorded pT stage (N=3,347) T stage
Did not undergo mastectomy (N=21,125)
1 7560 (58.6%) 457 (29.1%) <0.001
Radiation use not recorded (N=127)
2 4828 (37.4%) 611 (38.9%)
Male gender (N=41)
3 388 (3.0%) 413 (26.3%)
4 119 (0.9%) 88 (5.6%)
Systemic chemotherapy
Yes 10,488 (81.3%) 1480 (94.3%) <0.001
Study population N= 14,464 No 2344 (18.2%) 89 (5.7%)
Not recorded 63 (0.5%) 0 (0.0%)
Observation Radiation Surgical margins
N=12,895 N=1569 Positive 225 (1.7%) 94 (6.0%) <0.001
Negative 12,604 (97.7%) 1452 (92.5%)
Not reported 66 (0.5%) 23 (1.5%)
Fig. 1. Patient selection diagram.
50 PMRT for TNBC

A) 100.00%
90.00%
80.00%
70.00%

Percent ulizaiton
60.00%
50.00%
40.00%
30.00%
20.00%
10.00%
0.00%
T1 T2 T3 T4
pT stage

Radiaon Observaon

B) 90.00%

80.00%

70.00%
Percent ulizaiton

60.00%

50.00%

40.00%

30.00%

20.00%

10.00%

0.00%
2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014
Year of diagnosis

All paents T1 T2 T3 T4

Fig. 2. (A) Post-mastectomy radiation therapy use by T stage. (B) Post-mastectomy radiation therapy utilization by year of diagnosis.

of 1569 (10.8%) received PMRT, whereas 12,895 (89.2%) did not use of PMRT was associated with improved OS (5 year OS 62.6%
receive PMRT. An analysis of the rates of PMRT by stage (Fig. 2A) vs. 74.3%, p < 0.001, Fig. 3C).
demonstrated significant variation of PMRT utilization by T stage In the overall cohort, there were several predictors of improved
(p < 0.001). PMRT use was 5.7% for patients with pT1 disease, OS on Cox multivariate analysis (Table 3). These included younger
11.2% for patients with pT2 disease, 51.6% for patients with pT3 age, fewer comorbidities, private insurance, higher socioeconomic
disease, and 42.5% for patients with pT4 disease. PMRT use also status, lower pT stage, negative surgical margins, and use of
varied by surgical margin status, with 29.5% of patients with pos- chemotherapy. While use of PMRT was associated with worse OS
itive surgical margins undergoing PMRT, whereas 10.3% of patients on univariate analysis (hazard ratio [HR] 0.622 for observation
with negative surgical margins received PMRT (p < 0.001). No sig- compared to use of PMRT, 95% confidence interval [CI] 0.541–
nificant differences were found when analyzing use of PMRT 0.715, p < 0.001), this association did not persist on multivariate
among all patients (Fig. 2B) over time. analysis (HR 0.883, 95% CI 0.754–1.033, p = 0.119).
On multivariable logistic regression analysis (Table 2), increas-
ing age, private and Medicare insurance, treatment at an academic
facility, negative surgical margins, and omission of systemic Discussion
chemotherapy were associated with omission of PMRT. Increasing
pT stage was associated with an increased likelihood of PMRT use. The present investigation using U.S. hospital-based data is the
Median follow-up using the reverse Kaplan–Meier method was largest study to date of present-day practice patterns regarding
38.2 months (interquartile range, 25.9–51.7 months). When sub- the utilization of PMRT for patients with node-negative TNBC.
stratifying for pT stage, there was no difference in OS between The results presented herein demonstrate that among this patient
PMRT or observation for patients with pT1 disease (5 year OS cohorts, PMRT is used in a minority of patients, though its utiliza-
86.9% vs. 83.4%, p = 0.083, Fig. 3A), pT2 disease (5 year OS 77.8% tion varies significantly with pT stage. Additionally, younger age,
vs. 72.2%, p = 0.250, Fig. 3B), or pT4 disease (5 year OS 51.7% vs. positive surgical margins, and use of chemotherapy were associ-
49.3%, p = 0.362, Fig. 3D). However, for patients with pT3 disease, ated with PMRT use in this patient population. Importantly, PMRT
 W. Haque et al. / Radiotherapy and Oncology 132 (2019) 48–54 51

Table 2
Multivariable logistic regression analysis for factors predictive of post-mastectomy radiation therapy.

Characteristic Odds ratio 95% Confidence interval P value

Age
50 1 (reference)
51–60 0.976 0.832–1.145 0.765
61–70 0.768 0.631–0.934 0.008
71+ 0.661 0.515–0.845 0.001
Race
White 1 (reference)
African American 1.162 0.998–1.355 0.054
Other 1.084 0.832–1.414 0.549
Charlson–Deyo’s Score
0 1 (reference)
1 0.864 0.728–1.027 0.097
2 0.816 0.591–1.126 0.215
Insurance status
Medicaid 1 (reference)
Private 0.723 0.600–0.873 0.001
Medicare 0.638 0.501–0.814 <0.001
Not insured 0.914 0.647–1.292 0.612
Other 0.665 0.435–1.016 0.059
Median Income
$62999 1 (reference)
$63000 0.923 0.812–1.050 0.223
Not recorded 0.716 0.238–2.152 0.552
Facility type
Non academic 1 (reference)
Academic 0.842 0.733–0.966 0.014
Not recorded 1.265 1.040–1.538 0.019
T stage
1 1 (reference)
2 1.902 1.671–2.164 <0.001
3 17.978 5.052–21.473 <0.001
4 12.519 9.178–17.077 <0.001
Systemic chemotherapy
Yes 1 (reference)
No 0.343 0.270–0.436 <0.001
Not recorded – – –
Surgical margins
Positive 1 (reference)
Negative 0.271 0.206–0.358 <0.001
Not reported 0.75 0.418–1.344 0.334

use was not associated with improved OS in all patients, though its in the early trials investigating the efficacy of PMRT, as the Early
use was found to be associated with improved OS among patients Breast Cancer Trialists Cooperative Group (EBCTCG) demonstrated
with pT3 disease. These results have implications on the appropri- that for patients with node-negative disease, the local recurrence
ate use of PMRT in the setting of node-negative TNBC, and suggest rate without PMRT was only 1.6% [5]. While there are factors that
that the current NCCN guidelines of not including TNBC status as a are predictive for local recurrence in patients with node-negative
criterion to include PMRT in patients with T1-2N0 disease is breast cancer having undergone mastectomy, including lympho-
appropriate. vascular space invasion (LVSI), premenopausal status, and tumor
The present results suggest that physicians are not routinely size >2 cm [17–20], TNBC has not been demonstrated to be a risk
using TNBC status alone as an indication for PMRT use for patients factor necessitating the use of PMRT, and the present results indi-
with node-negative disease. Only 5.7% of patients with pT1N0 dis- cate physicians in the U.S. have not been using TNBC status alone
ease and 11.2% of patients with pT2N0 disease received PMRT. as a factor to determine the need for PMRT. Moreover, certain clin-
However, the utilization of PMRT increased to 51.6% for patients ical and demographic characteristics were predictive of omission
with pT3N0 TNBC and to 42.5% for patients with pT4N0 disease, of PMRT. Increasing age and omission of systemic chemotherapy
again suggesting that physicians may be using the NCCN guidelines were both predictive of omission of PMRT, suggesting that patients
of considering PMRT use in patients with T3 or higher disease. who may have been less likely to benefit from PMRT due to either
Importantly, the utilization rates of PMRT remain low, even for comorbid conditions or an inability to tolerated the toxicities of
patients with pT3 or pT4 disease, and are significantly lower than treatment may have not received PMRT. The presence of negative
rates of PMRT for patients with ypN1 (57.9%), ypN2 (68.4%), or surgical margins was predictive of a decreased likelihood of PMRT
ypN3 (67.0%) disease [16], though it should be noted that these use as well, which is in line with data that demonstrates that
PMRT rates are from a study of patients undergoing neoadjuvant positive surgical margins lead to a higher chance of local
chemotherapy, and it may be that these higher PMRT rates are recurrence.
due to the residual nodal positive even after neoadjuvant treat- No difference in OS was observed with the use of PMRT in
ment. The lack of routine PMRT even for patients with T3 or higher patients with pT1, pT2, or pT4 disease. While local control is not
disease may be reasonable despite the inclusion of these patients recorded in the NCDB, the OS information is valuable and suggests
52 PMRT for TNBC

A) B)

T1N0M0 T2N0M0
Log-rank P-value =0.0834 Log-rank P-value =0.250

C) D)

T3N0M0 T4N0M0
Log-rank P-value <0.001 Log-rank P-value = 0.362

Fig. 3. Kaplan–Meier’s curves comparing overall survival for patients with (A) T1 disease; (B) T2 disease; (C) T3 disease; (D) T4 disease.

a lack of benefit to PMRT in patients with node-negative, early have had an OS benefit due to more biologically aggressive disease,
stage TNBC. The significant OS gain observed in patients with whereas the randomized study design of the aforementioned trial
pT3 disease undergoing PMRT suggests that tumor size may be by Wang et al. was able to eliminate clinical differences between
an important indicator of biologic aggressiveness, and patients the two cohorts of patients and was thus able to demonstrate dif-
with tumors >5 cm may require PMRT to eradicate microscopic ferences in outcome with the use of PMRT. Finally, nearly 60% of
residual disease that persists following a mastectomy. The lack of patients included in the trial were 50 years of age, whereas only
observed benefit with use of PMRT among patients with pT4 dis- about 33% of patients in the present observational study were
ease is curious, but may be a result of the relatively small sample 50 years old. PMRT may have a greater benefit among younger,
size of the pT4 cohort. Alternatively, it may be due to the lack of premenopausal patients.
benefit of local treatment in patients who may already have distant Two recent retrospective reviews have demonstrated a benefit
micrometastatic disease. with the addition of PMRT in patients with TNBC. In a review of
The present results are discordant with a randomized trial that 104 patients with Stage II–III TNBC treated with neoadjuvant
demonstrated significant OS improvements with use of PMRT for chemotherapy and modified radical mastectomy, Chen et al.
patients with stage I–II TNBC [12]. There are a number of possible showed that patients treated with PMRT had improved local con-
reasons for the differences in outcomes observed in the aforemen- trol and distant disease control, regardless of whether or not these
tioned trial and those observed in the present study. First, the trial patients has positive lymph nodes [21]. In another retrospective
randomized patients diagnosed with TNBC between 2001 and review from a single institution, Shen et al. demonstrated that
2006, an earlier era in which systemic therapy may not have been out of 167 patients with TNBC and T1-2N1 disease, the addition
as comprehensively efficacious. PMRT may have thus compensated of PMRT resulted in decreased locoregional failure [22]. Due to
for inadequate systemic treatment. Second, while over 80% of the limitations of the NCDB, the present study was not able to
patients included in the trial had N0 disease, about 20% of patients report on potential local control benefits with PMRT, and this is
did have node-positive disease. It is possible that most of the OS an area that requires further exploration.
benefit with the use of PMRT was limited to the node-positive Though the NCDB offers a unique platform to study national
patients; since the present study did not include any patients with practice patterns of PMRT use among patients with node-negative
node-positive disease, no OS benefit was observed for T1-2 TNBC, there are several known limitations inherent to this dataset
patients. Third, it is possible that there was a selection bias present [14,23]. NCDB studies are characteristically retrospective and con-
in the present observational study, and patients at higher risk for sequently, can never eliminate selection biases, including but not
recurrence due to factors not present within the NCDB may have limited to judgment by individual providers, nature of follow-up
been preferentially selected for PMRT. Consequently, the patients management, and referral patterns. Therefore, the results of this
receiving PMRT with T1-2 disease in the present study may not study should be interpreted with caution, given that the specific
 W. Haque et al. / Radiotherapy and Oncology 132 (2019) 48–54 53

Table 3
Univariate and multivariate analysis for factors predictive of overall survival.

Univariate analysis Multivariate analysis

Characteristic Hazard ratio 95% Confidence interval P value Hazard ratio 95% Confidence interval P value
Group
Radiation 1 (reference) 1 (reference)
Observation 0.622 0.541–0.715 <0.001 0.883 0.754–1.033 0.119
Age
50 1 (reference) 1 (reference)
51–60 1.149 0.969–1.362 0.109 1.079 0.896–1.299 0.424
61–70 1.571 1.331–1.855 <0.001 1.241 1.014–1.519 0.036
71+ 3.625 3.140–4.184 <0.001 2.179 1.759–2.698 <0.001
Race
White 1 (reference) – – –
African American 1.150 0.999–1.323 0.051 – – –
Other 0.762 0.572–1.014 0.06 – – –
Charlson–Deyo’s Score
0 1 (reference) 1 (reference)
1 1.682 1.475–1.918 <0.001 1.425 1.248–1.628 <0.001
2 3.610 3.022–4.313 <0.001 2.525 2.107–2.038 <0.001
Insurance status
Medicaid 1 (reference) 1 (reference)
Private 0.509 0.419–0.619 <0.001 0.661 0.542–0.805 <0.001
Medicare 1.479 1.226–1.783 <0.001 0.957 0.766–1.196 0.698
Not insured 0.915 0.635–1.319 0.634 0.881 0.611–1.271 0.498
Other 0.614 0.391–0.965 0.034 0.668 0.424–1.052 0.082
Median Income
$62999 1 (reference) 1 (reference)
$63000 0.682 0.603–0.771 <0.001 0.835 0.737–0.946 0.005
Not recorded 2.451 1.271–4.724 0.007 2.817 1.459–5.438 0.002
Facility type
Non academic 1 (reference) 1 (reference)
Academic 0.847 0.751–0.956 0.007 0.986 0.873–1.113 0.817
Not recorded 0.535 0.423–0.675 <0.001 0.928 0.710–1.213 0.585
T stage
1 1 (reference) 1 (reference)
2 2.061 1.831–2.320 <0.001 2.144 1.900–2.420 <0.001
3 4.568 3.869–5.393 <0.001 4.162 3.468–5.00 <0.001
4 7.578 5.966–9.644 <0.001 6.202 4.827–7.967 <0.001
Systemic chemotherapy
Yes 1 (reference) 1 (reference)
No 1.779 1.578–2.006 <0.001 1.472 1.287–1.682 <0.001
Not recorded 2.098 1.125–3.911 0.020 1.620 0.865–3.034 0.132
Surgical margins
Positive 1 (reference) 1 (reference)
Negative 0.597 0.449–0.794 <0.001 0.708 0.531–0.944 0.019
Not reported 0.478 0.226–1.009 0.053 0.654 0.309–1.384 0.267

reason for the selection patients to receive PMRT could not be Disclaimers
ascertained. Second, the NCDB does not record information regard-
ing the chemotherapy agents used, which may have been a source None. This has never been presented/published before in any
of confounding. Third, as mentioned above, important information form. All authors declare that conflicts of interest do not exist.
regarding clinical outcomes including local control or distant
metastasis free survival, all potential indicators for benefit of PMRT
Funding
use, are not included in the NCDB. Fourth, this investigation has a
limited follow up time. It is possible that with longer follow up
There was no research support for this study.
worse outcomes may have been recorded in patients not receiving
Waqar Haque was responsible for the statistical analysis.
PMRT. Finally, the NCDB does not contain detailed information
regarding the radiation fields, including whether or not the supra-
clavicular nodes and internal mammary nodes were included, for a References
majority of patients.
[1] National Comprehensive Cancer Network. Breast Cancer. Version 1.2018.
This study using a large, contemporary data base of national https://www.nccn.org/professionals/physician_gls/pdf/breast.pdf. Accessed
practice patterns, demonstrates that the use of PMRT for node- September 4, 2018.
[2] Overgaard M, Hansen PS, Overgaard J, et al. Postoperative radiotherapy in
negative TNBC is low for early stage patients, and the use of PMRT
high-risk premenopausal women with breast cancer who receive adjuvant
is greater for pT3-4 patients as well as patients with positive sur- chemotherapy. Danish Breast Cancer Cooperative Group 82b Trial. N Engl J
gical margins. Additionally, use of PMRT is associated with an OS Med 1997;37:949–55.
advantage in pT3 patients. These results suggest PMRT should [3] Overgaard M, Jensen MB, Overgaard J, et al. Postoperative radiotherapy in
high-risk postmenopausal breast-cancer patients given adjuvant tamoxifen:
not be used routinely in patients with pT1-2N0 disease, but may Danish Breast Cancer Cooperative Group DBCG 82c randomised trial. Lancet
be indicated in patients with pT3 disease. 1999;353:1641–8.
54 PMRT for TNBC

[4] Ragaz J, Olivotto IA, Spinelli JJ, et al. Locoregional radiation therapy in patients [14] Haque W, Verma V, Butler EB, Teh BS. Patterns of care and outcomes of multi-
with high-risk breast cancer receiving adjuvant chemotherapy: 20-year agent versus single-agent chemotherapy as part of multimodal management
results of the British Columbia randomized trial. J Natl Cancer Inst of low grade glioma. J Neurooncol 2017;133:369–75.
2005;97:116–26. [15] Haque W, Verma V, Butler EB, Teh BS. National practice patterns and outcomes
[5] Early Breast Cancer Trialists’ Collaborative Group. Effect of radiotherapy after for T4b urothelial cancer of the bladder. Clin Genitourin Cancer 2017. https://
mastectomy and axillary surgery on 10-year recurrence and 20-year breast doi.org/10.1016/j.clgc.2017.08.013.
cancer mortality: meta-analysis of individual patient data for 8135 women in [16] Ohri N, Moshier E, Ho A, et al. Postmastectomy radiation in breast cancer
22 randomised trials. Lancet 2014;383:2127–35. patients with pathologically positive lymph nodes after neoadjuvant
[6] van de Vijver MJ, He YD, van’t Veer LJ, et al. A gene-expression signature as a chemotherapy: usage rates and survival trends. Int J Radiat Oncol Biol Phys
predictor of survival in breast cancer. N Engl J Med 2002;347:1999–2009. 2017;99:549–59.
[7] Sparano JA, Gray RJ, Makower DF, et al. Adjuvant chemotherapy guided [17] Jagsi R, Raad RA, Goldberg S, et al. Locoregional recurrence rates and
by a 21-gene expression assay in breast cancer. N Engl J Med 2018;379: prognostic factors for failure in node-negative patients treated with
111–21. mastectomy: implications for postmastectomy radiation. Int J Radiat Oncol
[8] Tang G, Shak S, Paik S, et al. Comparison of the prognostic and predictive Biol Phys 2005;62:1035–9.
utilities of the 21-gene Recurrence Score assay and Adjuvant! for women with [18] Abi-Raad R, Boutrus R, Wang R, et al. Patterns and risk factors of locoregional
node-negative, ER-positive breast cancer: results from NSABP B-14 and NSABP recurrence in T1–T2 node negative breast cancer patients treated with
B-20. Breast Cancer Res Treat 2011;127:133–42. mastectomy: implications for postmastectomy radiotherapy. Int J Radiat
[9] Zhang Y, Schnabel CA, Schroeder BE, et al. Breast cancer index identifies early- Oncol Biol Phys 2011;81:e151–7.
stage estrogen receptor-positive breast cancer patients at risk for early- and [19] Truong PT, Lesperance M, Culhaci A, Kader HA, Speers CH, Olivotto IA. Patient
late-distant recurrence. Clin Cancer Res 2013;19:4196–205. subsets with T1–T2, node-negative breast cancer at high locoregional
[10] Gnant M, Filipits M, Greil R, et al. Predicting distant recurrence in receptor- recurrence risk after mastectomy. Int J Radiat Oncol Biol Phys
positive breast cancer patients with limited clinicopathological risk: using the 2005;62:175–82.
PAM50 Risk of Recurrence score in 1478 postmenopausal patients of the [20] Rowell NP. Radiotherapy to the chest wall following mastectomy for node-
ABCSG-8 trial treated with adjuvant endocrine therapy alone. Ann Oncol negative breast cancer: a systematic review. Radiother Oncol 2009;91:23–32.
2014;25:339–45. [21] Chen X, Xia F, Luo J, et al. Postmastectomy radiotherapy reduces locoregional
[11] Abdulkarim BS, Cuartero J, Hanson J, et al. Increased risk of locoregional and disease recurrence in patients with stage II-III triple negative breast cancer
recurrence for women with T1-2N0 Triple-negative breast cancer treated with treated with neoadjuvant chemotherapy and mastectomy. Onco Targets Ther
modified radical mastectomy without adjuvant radiation therapy compared 2018;11:1973–80.
with breast-conserving therapy. J Clin Oncol 2011;29:2852–8. [22] Shen H, Zhao L, Wang L, et al. Postmastectomy radiotherapy benefit in Chinese
[12] Wang J, Shi M, Ling R, et al. Adjuvant chemotherapy and radiotherapy in triple- breast cancer patients with T1–2 tumor and 1–3 positive axillary lymph nodes
negative breast carcinoma: a prospective randomized controlled multi-center by molecular subtypes: an analysis of 1369 cases. Tumour Biol
trial. Radiother Oncol 2011;100:200–4. 2016;37:6465–75.
[13] Bilimoria KY, Stewart AK, Winchester DP, et al. The national cancer data base: [23] Haque W, Verma V, Fakhreddine M, et al. Addition of chemotherapy to
a powerful initiative to improve cancer care in the United States. Ann Surg definitive radiotherapy for IB1 and IIA1 cervical cancer: Analysis of the
Oncol 2008;15:683–90. National Cancer Data Base. Gynecol Oncol 2017;144:28–33.