- Email: [email protected]
Treatment Outcomes in Male Breast Cancer: A Retrospective Analysis of 161 Patients
Clinical Oncology xxx (2018) 1e12 Contents lists available at ScienceDirect
Clinical Oncology journal homepage: www.clinicaloncologyonline.net
Original Article
Treatment Outcomes in Male Breast Cancer: A Retrospective Analysis of 161 Patients B.A. Wan, V. Ganesh, L. Zhang, P. Sousa, L. Drost, J. Lorentz, D. Vesprini, J. Lee, E. Rakovitch, F.-I. Lu, A. Eisen, C. Yee, H. Lam, E. Chow Odette Cancer Centre, Sunnybrook Health Sciences Centre, University of Toronto, Toronto, Ontario, Canada Received 10 October 2017; received in revised form 28 December 2017; accepted 11 January 2018
Abstract Aims: Male breast cancer is a rare disease with limited evidence-based guidelines for treatment. This study aimed to identify demographic, pathological and clinical factors associated with its prognosis. Materials and methods: A retrospective review of 161 male breast cancer patients diagnosed at a single institution from 1987 to June 2017 was conducted. Patient demographics, disease characteristics, treatment and outcome were extracted and included in competing-risk analysis and the univariate Cox proportional hazard model for univariate analysis. Factors with P < 0.10 were included in multivariable analysis. Results: The mean age at diagnosis was 67 years (standard deviation ¼ 11.2) and the median follow-up duration was 5.3 years (range 0e25 years). There were 48 deaths, including 23 cancer-specific deaths. The actuarial median survival was 19.9 years. In multivariable analysis, factors associated with overall survival were size of tumours (hazard ratio 2.0; 95% confidence interval 1.4e2.7, P < 0.0001) and diagnosis of metastatic disease (hazard ratio 8.7; 95% confidence interval 1.9e40.6; P ¼ 0.006). Of 138 patients without metastases at diagnoses, 11 had local-regional recurrence and 26 had distant metastases. In the multivariable model for local-regional recurrence, a more recent year of diagnosis was associated with reduced risk (hazard ratio 0.9, 95% confidence interval 0.8e1.0, P ¼ 0.008), whereas more positive lymph nodes was associated with higher risk (hazard ratio 2.2, 95% confidence interval 1.2e4.0, P ¼ 0.01). A higher risk of metastases was associated with more positive lymph nodes (hazard ratio 1.9; 95% confidence interval 1.1e3.3; P ¼ 0.03) and tumour size (hazard ratio 1.8; 95% confidence interval 1.1e2.9; P ¼ 0.01). A higher risk of any recurrence or metastases was associated with the number of positive nodes (hazard ratio 1.9; 95% confidence interval 1.2e3.0; P ¼ 0.005) and tumour size (hazard ratio 1.6; 95% confidence interval 1.1e2.2; P ¼ 0.01). Conclusion: In general, tumour size and more positive lymph nodes were associated with worse prognosis. Larger powered studies are needed to identify prognostic factors with smaller effect sizes. Ó 2018 The Royal College of Radiologists. Published by Elsevier Ltd. All rights reserved.
Key words: Breast cancer; male; treatment outcomes
Introduction Male breast cancer is rare and accounts for 0.6% of all breast cancer diagnoses [1]. This has resulted in a lack of evidence-based treatment guidelines due to difficulties in conducting large-scale randomised controlled trials. The current understanding of its biology and treatment guidelines has been based on epidemiological and retrospective
Author for correspondence: E. Chow, Department of Radiation Oncology, Odette Cancer Centre, Sunnybrook Health Sciences Centre, 2075 Bayview Avenue, Toronto, Ontario M4N 3M5, Canada. Tel: þ1-416-480-4974. E-mail address: [email protected] (E. Chow).
studies, and extrapolation from studies on female breast cancer. Factors associated with the risk of male breast cancer include family history, increased oestrogen exposure or hypoandrogenism, radiation/occupational exposure and heritable elements such as BRCA1 and BRCA2 gene mutations [2]. The characteristics of male breast cancer have been found to resemble postmenopausal female breast cancer in several studies [3,4]. However, there is increasing evidence that male breast cancer differs in aetiology, clinical-pathological presentation and outcomes when compared with female breast cancer. For example, lobular carcinomas are the second most common subtype in female breast cancer (11.8%), but present rarely in males (1%) [5,6].
https://doi.org/10.1016/j.clon.2018.02.026 0936-6555/Ó 2018 The Royal College of Radiologists. Published by Elsevier Ltd. All rights reserved.
Please cite this article in press as: Wan BA, et al., Treatment Outcomes in Male Breast Cancer: A Retrospective Analysis of 161 Patients, Clinical Oncology (2018), https://doi.org/10.1016/j.clon.2018.02.026
2
B.A. Wan et al. / Clinical Oncology xxx (2018) 1e12
In addition, male breast cancer is almost always oestrogen hormone receptor positive and when compared with female breast cancer there is an under-representation of BRCA1 mutations and an over-representation of BRCA2 mutations [3]. Moreover, an analysis of the immunehistopathological characteristics of male breast cancer by Abreu et al. [7] suggested that male breast cancer may be divided into multiple prognostic subgroups that highlight differences in tumourigenesis. Management of male breast cancer includes surgical excision, radiotherapy and systemic therapy in the form of chemotherapy, hormonal therapy for oestrogen receptor (ER) positive or progesterone receptor (PR) positive patients, or trastuzumab for patients with human epidermal growth factor receptor (HER2/neu) overexpression [8]. Although the overall survival of male breast cancer has improved in the last decade, there remains a lack of evidence-based data for its management [9]. The objectives of this present study were to evaluate the demographic, clinical, pathological characteristics and treatments associated with outcomes in male breast cancer.
Materials and Methods A retrospective review of patients who were diagnosed from 1987 to June 2017 at the Odette Cancer Centre was conducted. Ethics approval from Sunnybrook Health Sciences Centre was obtained before the start of the study. Patient demographic, treatment, pathology, biomarkers, years of cancer diagnosis (1987 to <2000, 2000 to <2010, 2010 to 2017) and follow-up data were extracted. Biomarker data on ER, PR and HER2/neu status were used to divide cancer into subtypes of luminal A-like (ER/PRþ, HER2/ neue), luminal B-like (ER/PRþ, HER2/neuþ) or triple negative (ER/PRe, HER2/neue) based on definitions by Ge et al. [10]. Clinical-pathological features were individually correlated to overall survival, risk of local-regional recurrence and risk of metastases. Overall survival in years was defined as the time from diagnosis to the date of death or to the last follow-up. Risk of recurrence or metastases was calculated for patients without metastases at diagnosis (M0) and was defined as the time from breast cancer diagnosis to breast cancer recurrence/metastases. Patients, disease and treatment characteristics were summarised as mean, standard deviation and range for continuous variables, and proportions for categorical variables. The univariate Cox proportional hazard model was used to identify significant covariates related to overall survival. KaplaneMeier overall survival curves, hazard ratios, 95% confidence intervals and P-values were generated. The generalised R2 statistic (between 0 and 1) was calculated based on the likelihood ratio statistic for testing the global null hypothesis; the larger the R2, the stronger the association with the outcome [10]. A competing risk analysis was conducted for local-regional recurrence, distant metastases or any recurrence/distant metastases in nonmetastatic patients. The cumulative incidence function
was estimated and plotted as well. To search for significant predictive factors of local-regional recurrence, distant metastases or any recurrence/distant metastases, univariate Cox proportional subdistribution hazard models were carried out using Fine and Gray’s method, considering death as the competing risk [11]. Subdistribution hazard ratios with 95% confidence intervals and P-value were estimated for each factor. In the multivariable analysis, all variables with P < 0.10 from the univariate analysis were selected for inclusion in the backward stepwise selection procedure. The final model would only keep the significant predictive factors with P < 0.05. All analyses were conducted using Statistical Analysis Software (SAS version 9.4 for Windows, Cary, NC) and R package (version 3.2.0).
Results Baseline Patient, Pathological and Treatment Characteristics Patient characteristics are summarised in Table 1. The median age of 161 patients at diagnosis was 67 years (range 34e92). The median duration of follow-up since diagnosis was 5.3 years (range 0e25 years). Only two patients (1.2%) had bilateral breast cancer. Twenty-five (15.5%) patients had gynaecomastia. Seventy-three (45.3%) patients had a family history of cancer. Of the 34 patients with known BRCA status, three had BRCA1 mutation (8.8%) and eight had BRCA2 mutation (23.5%). This included one patient who had a BRCA1 mutation of uncertain significance. There were 54 patients with previous, secondary or subsequent malignancies (33.5%), the most common of which was prostate cancer (n ¼ 23) followed by basal cell carcinoma (n ¼ 7). Disease characteristics and pathology are summarised in Table 2. Most patients presented with stage 2 disease (n ¼ 63, 39.1%) and no nodal involvement (n ¼ 61, 37.9%), while 11 patients had metastatic disease (6.8%). The most common histological type was invasive ductal carcinoma, no specific type (n ¼ 146, 90.7%) although there were 15 (9.3%) patients with other histological types. Overall, most patients were luminal A-like (n ¼ 86, 53.4%), followed by luminal B-like (n ¼ 15, 9.3%), with one triple-negative patient. The most common Nottingham histologic score was 2 (n ¼ 75, 63.0%). Treatments are summarised in Table 3. Most patients underwent surgery (n ¼ 143, 88.8%), the most common of which was mastectomy (n ¼ 133, 93.0%). Other treatment modalities included radiotherapy to the breast or chest wall (n ¼ 95), chemotherapy (n ¼ 69), systemic hormonal therapy (for 122 of 142 patients with ER/PR overexpression) or trastuzumab (for nine of 16 patients with HER2/neu overexpression). Outcome Of 161 patients there were 48 deaths (median time to death: 5.4 years; interquartile range [IQR] 2.1e9.0), of which 23 (47.9%) were breast cancer specific. Figure 1A shows the
Please cite this article in press as: Wan BA, et al., Treatment Outcomes in Male Breast Cancer: A Retrospective Analysis of 161 Patients, Clinical Oncology (2018), https://doi.org/10.1016/j.clon.2018.02.026
B.A. Wan et al. / Clinical Oncology xxx (2018) 1e12 Table 1 Patient characteristics Patient characteristics
3
Table 2 Disease characteristics and pathology Total patients (N ¼ 161) n
Disease characteristics and pathology
%
Age at diagnosis (years) Mean standard deviation 66.9 11.2 Minimum, maximum 34.0, 92.0 Duration of follow-up since diagnosis (years) Mean standard deviation 6.6 5.5 Diagnosis year categories 1987 to <2000 24 14.9% 2000 to <2010 77 47.8% 2010 to 2017 60 37.3% Gynaecomastia Absence 136 84.5% Presence 25 15.5% Family history of breast cancer No 32 19.9% Yes 73 45.3% Unknown 56 34.8% Genetic counselling No 105 65.2% Yes 53 32.9% Declined 3 1.9% BRCA status among tested (n ¼ 34) BRCA1 3 8.8% BRCA2 8 23.5% Negative 23 14.3% Other malignancy No 107 66.5% Yes 54 33.5% Type of second primary cancer from 54 patients who had ‘other malignancy’ Prostate 23 38.3% Basal cell carcinoma 7 11.7% Bladder 5 8.3% Renal cell carcinoma 4 6.7% Lung 3 5.0% Lymphoma 3 5.0% Melanoma 3 5.0% Other malignancy 12 20.0%
KaplaneMeier overall survival curve, with the dashed lines indicating the 95% confidence interval. The actuarial median survival time from KaplaneMeier estimations for all patients was 19.9 years (95% confidence interval 11.2e25.4). Among the 138 patients without metastases at diagnosis, 11 patients had local-regional recurrence (8.0%, median time to recurrence 5.0 years; IQR 1.7e8.3) and 26 patients had distant metastasis (18.8%, median time to metastases 4.7 years; IQR 1.3e8.3). There were 34 patients who had any recurrence or metastases, with the median time to event of 4.38 years (IQR 1.2e7.8). Overall Survival Analysis The univariate Cox proportional hazard model for overall survival of all patients using demographics,
Total patients (N ¼ 161) n
Breast cancer laterality Left Right Bilateral Unknown Size of invasive component (cm) Mean standard deviation T-stage 1 2 3 4 Tx N-stage 0 1 2 3 Nx M-stage 0 1 Mx Histologic type IDC, NOS Papillary Inflammatory Mucinous DCIS Invasive apocrine Micropapillary Unknown Nottingham grade 1 2 3 Unknown Skin involvement No Yes Unknown Muscle involvement No Yes Unknown LVI No Yes Unknown Total nodes removed Mean standard deviation Total nodes positive Mean standard deviation ER/PR status Negative Positive Unknown
%
79 49.1% 78 48.5% 2 1.2% 2 1.2% N ¼ 119 2.47 1.35 48 63 7 11 31
29.8% 39.1% 4.4% 6.8% 19.3%
61 36 10 14 40
37.9% 22.4% 6.2% 8.7% 24.8%
138 11 12
85.7% 6.8% 7.5%
89 5 3 3 2 1 1 57
55.3% 3.1% 1.9% 1.9% 1.2% 0.6% 0.6% 35.4%
9 75 35 42
5.6% 46.6% 21.7% 26.1%
63 34 64
39.1% 21.1% 39.8%
150 5 6
93.2% 3.1% 3.7%
59 54 48
36.7% 33.5% 29.8%
10.9 7.5 2.4 4.1 3 142 16
1.9% 88.2% 9.9%
(continued on next page)
Please cite this article in press as: Wan BA, et al., Treatment Outcomes in Male Breast Cancer: A Retrospective Analysis of 161 Patients, Clinical Oncology (2018), https://doi.org/10.1016/j.clon.2018.02.026
4
B.A. Wan et al. / Clinical Oncology xxx (2018) 1e12 Table 2 (continued ) Disease characteristics and pathology
Table 3 (continued ) Total patients (N ¼ 161) n
%
ER status Negative 3 Positive 126 Unknown 32 PR status Negative 15 Positive 113 Unknown 33 HER2/Neu status Negative 88 Positive 16 Unknown 57 Cancer subtype Luminal A-like 86 Luminal B-like 15 Triple negative 1 Unknown 59 Nuclear grade 1 1 2 27 3 15 Unknown 118 Presence of concurrent DCIS with invasive disease No 19 Yes 58 Unknown 84
1.9% 78.3% 19.9% 9.3% 70.2% 20.5% 54.7% 9.9% 35.4% 53.4% 9.3% 0.6% 36.7% 0.6% 16.8% 9.3% 73.3% 11.8% 36.0% 52.2%
IDC, NOS: Invasive ductal carcinoma, not otherwise specified; DCIS: Ductal carcinoma in situ; LVI: Lympho-vascular invasion; ER: Estrogen receptor; PR: Progesterone receptor; HER2: Human epidermal growth factor receptor 2.
Table 3 Treatment summary Treatment summary
Surgery No Yes Unknown Surgery type (n ¼ 143) Lumpectomy Mastectomy Unknown Use of radiotherapy Breast/chest wall radiotherapy No Yes Unknown Nodes/axilla radiotherapy No Yes Unknown
Treatment summary
Total patients (N ¼ 161) n
%
13 143 5
8.1% 88.8% 3.1%
8 133 2
5.6% 93.0% 1.4%
57 95 9
35.4% 59.0% 5.6%
106 43 12
65.8% 26.7% 7.5%
Boost radiotherapy No Yes Unknown Palliative radiotherapy No Yes Use of chemotherapy Chemotherapy No Yes Adjuvant Neoadjuvant Palliative Unknown Number of chemotherapy regimens 1 2 3 4 Chemotherapy regimens from 69 patients AC FEC-D Paclitaxel Docetaxel Capecitabine CEF CMF AC paclitaxel Other chemotherapy regimens For patients who were ER/PRþ (n ¼ 142) Use of hormonal therapy No Yes Name of hormonal therapy Tamoxifen Letrozole Anastrazole Exemestane Other/unknown For patients who were HER2/Neuþ (n ¼ 16) Use of trastuzumab No Yes
Total patients (N ¼ 161) n
%
152 7 2
94.4% 4.4% 1.2%
131 30
81.4% 18.6%
78 69 43 10 19 14
48.5% 42.9% 26.7% 6.2% 11.8% 8.7%
55 5 4 5
34.2% 3.1% 2.5% 3.1%
12 10 9 8 7 6 6 5 21
14.3% 11.9% 10.7% 9.5% 8.3% 7.1% 7.1% 6.0% 25.0%
20 122
14.1% 85.9%
103 17 16 9 9
62.8% 10.4% 9.8% 5.5% 5.5%
7 9
43.8% 56.3%
AC: Adriamycin and cyclophosphamide; CEF: Cyclophosphamide, epirubicin, fluorouracil; CMF: Cyclophosphamide, methotrexate, fluorouracil; ER/PR: Estrogen receptor/Progesterone receptor; HER2: human epidermal growth factor receptor 2.
pathological characteristics and treatment received as variables identified 15 significant predictive factors (Table A1). Older age, more recent year of diagnosis, stage T4, N3 or M1 disease, larger sizes of invasive tumours, presence of LVI, more total nodes removed or positive nodes, more chemotherapy regimens and palliative chemotherapy or radiotherapy treatment and palliative radiotherapy treatment were associated with a higher risk of death.
Please cite this article in press as: Wan BA, et al., Treatment Outcomes in Male Breast Cancer: A Retrospective Analysis of 161 Patients, Clinical Oncology (2018), https://doi.org/10.1016/j.clon.2018.02.026
B.A. Wan et al. / Clinical Oncology xxx (2018) 1e12
5
Fig 1. KaplaneMeier overall survival curves. (A) Overall survival for all patients, with dashed lines indicating 95% confidence intervals. (B) Metastatic (solid) versus non-metastatic (dashed) disease at diagnosis.
Multivariable analysis is presented in Table 4. Two significant factors remained as independent predictors of overall survival: invasive tumour size (hazard ratio 2.0; 95% confidence interval 1.4e2.7, P < 0.0001) and presence of metastatic disease at diagnosis (hazard ratio 8.7; 95% confidence interval 1.9e40.6; P ¼ 0.006). Figure 1B shows the KaplaneMeier overall survival curves comparing stage M0 versus M1 disease at diagnosis. The actuarial median overall survival was 19.9 years (95% confidence interval 12.8e25.4) or 2.6 years (95% confidence interval 1.0e5.5) in patients with M0 or M1 disease, respectively.
Competing Risks Analysis in Non-metastatic Patients For the 138 non-metastatic patients at diagnosis, 11 had local-regional recurrence and 26 had the competing event of death, with the cumulative incidence being 5.0% at 5 years and 11.2% at 10 years. Figure 2A shows the cumulative incidence of local-regional recurrence with the 95% confidence intervals (dashed). Table A2 shows the univariate Cox proportional subdistribution hazards model of local-regional recurrence, which identified that an earlier year of diagnosis, stage 3 disease, higher Nottingham
Table 4 Multivariable models for overall survival, local-regional recurrence, metastases or any recurrence or metastases. Values in bold are statistically significant (P < 0.05) Multivariable model for overall survival Predictive factor
Coefficient
Standard error
P-value
Invasive tumour size (cm) M-stage (1 versus 0)
0.67694 2.16019
0.16846 0.78789
<0.0001 0.0061
Standard error
P-value
Multivariable model for local-regional recurrence Predictive factor Coefficient
Year of diagnosis (continuous) Total positive nodes (log)
e0.12568 0.77912
0.04769 0.30167
0.0084 0.0098
Multivariable model for distant metastases Predictive factor
Coefficient
Standard error
P-value
Total positive nodes (log) Palliative chemotherapy (yes versus no) Invasive tumour size (cm)
0.63437 2.23426 0.60104
0.28371 0.78341 0.24057
0.0254 0.0043 0.0125
Multivariable model for any recurrence or metastases Predictive factor Coefficient
Standard error
P-value
Total positive nodes (log) Neoadjuvant chemotherapy (yes versus no) Palliative chemotherapy (yes versus no) Invasive tumour size (cm)
0.2319 0.69498 0.69773 0.18145
0.005 0.0451 0.012 0.0141
0.65068 1.39239 1.75181 0.44522
Hazard ratio (95% confidence interval) 1.968 1.414 2.738 8.673 1.851 40.626 Subdistribution hazard ratio and 95% confidence interval 0.882 0.803 0.968 2.18 1.207 3.937 Subdistribution hazard ratio and 95% confidence interval 1.886 1.081 3.288 9.34 2.011 43.368 1.824 1.138 2.923 Subdistribution hazard ratio and 95% confidence interval 1.917 1.217 3.02 4.024 1.031 15.713 5.765 1.469 22.631 1.561 1.094 2.227
R2 16.40%
R2 (%)
8.14%
R2 (%)
25.36%
R2 (%)
25.70%
Please cite this article in press as: Wan BA, et al., Treatment Outcomes in Male Breast Cancer: A Retrospective Analysis of 161 Patients, Clinical Oncology (2018), https://doi.org/10.1016/j.clon.2018.02.026
6
B.A. Wan et al. / Clinical Oncology xxx (2018) 1e12
Fig 2. Competing risk analysis in non-metastatic patients. (A) Cumulative incidence of local/regional recurrence. (B) Cumulative incidence of distant metastases. (C) Cumulative incidence of any recurrence or distance metastases.
grade, more total positive nodes removed or radiotherapy treatment had a higher risk of local-regional recurrence. In multivariable analysis, two variables remained significant (Table 4). Later year of diagnosis was associated with a reduced risk (hazard ratio 0.9, 95% confidence interval 0.8e1.0, P ¼ 0.008) and total number of positive lymph nodes was associated with a higher risk of local-regional recurrence (hazard ratio 2.2, 95% confidence interval 1.2e4.0, P ¼ 0.01). Among 138 patients diagnosed with non-metastatic disease, 26 developed distant metastases and 20 died. The cumulative incidence of distant metastases is shown in Figure 2B and was 17.7% at 5 years and 26.1% at 10 years. Univariate analysis identified that significant predictors of distant metastases included higher T or N stage, presence of other malignancies, larger invasive tumour size, greater number of dissected lymph nodes and positive lymph nodes, HER2/neu overexpression, luminal B-like tumour and treatment with chemotherapy or radiotherapy (Table A3). In particular, nodal staging was highly significant, with the risk of metastases highest for N3, followed by N2, N1 and N0. In multivariable analysis, three variables remained significantly associated with outcome (Table 4). More total positive nodes (hazard ratio 1.9; 95% confidence interval 1.1e3.3; P ¼ 0.03) or larger invasive tumour size (hazard ratio 1.8; 95% confidence interval 1.1e2.9; P ¼ 0.01) were associated with a higher risk of developing distant metastasis. When the presence of any local-regional recurrence or distant metastases was evaluated as one variable, the
cumulative incidence was 19.9% at 5 years and 33.9% at 10 years. The cumulative incidence of any recurrence or metastases is shown in Figure 2C. Significant predictive factors with the largest hazard ratios included stage N3 versus N0 (hazard ratio 12.2; 95% confidence interval 4.2e35.7; P < 0.0001) and stage T4 versus T1 (hazard ratio 5.5; 95% confidence interval 1.6e18.5; P ¼ 0.006) (Table A4). In the multivariable model (Table 4), a higher risk of recurrence/metastases was associated with the number of total positive nodes (hazard ratio 1.9; 95% confidence interval 1.2e3.0; P ¼ 0.005), invasive tumour size (hazard ratio 1.6; 95% confidence interval 1.1e2.2; P ¼ 0.01) or with palliative or neoadjuvant chemotherapy (hazard ratio 5.8; 95% confidence interval 1.5e22.6; P ¼ 0.01 or hazard ratio 4.0; 95% confidence interval 1.0e15.7; P ¼ 0.045).
Discussion Our study identified that the size of the invasive tumour, diagnosis of metastatic disease, year of diagnosis, number of positive lymph nodes and receipt of neoadjuvant or palliative chemotherapy were associated with a poorer prognosis. Patient characteristics were not identified as independent predictors of outcome. Previous studies have identified demographic predictors of male breast cancer outcome. Goss et al. [12] and Adami et al. [13] both found that men diagnosed at younger ages
Please cite this article in press as: Wan BA, et al., Treatment Outcomes in Male Breast Cancer: A Retrospective Analysis of 161 Patients, Clinical Oncology (2018), https://doi.org/10.1016/j.clon.2018.02.026
B.A. Wan et al. / Clinical Oncology xxx (2018) 1e12
had survival advantages when compared with older men. Results from our univariate analysis suggested that age was a significant factor predictive of overall survival, but not of recurrence or metastases. Goss et al. [12] also found that the median age of diagnosis was similar in male patients with and without a family history, whereas females with a family history presented with breast cancer at a younger age. Similarly, our study supported their findings, with the average age of diagnosis at 67 years for men with and without a family history of cancer. Epidemiological studies have identified that male carriers of the BRCA1 and BRCA2 mutations have higher risks of breast cancer and prostate cancer [4]. Our study found that the incidence of BRCA1 and BRCA2 mutations among patients who underwent genetic testing was consistent with published frequencies (8.8% and 23.5%, respectively). In a study by Hemminki et al. [14] of secondary primary malignancies in 3409 male breast cancer patients, 12.5% developed a second neoplasia; similarly, Masci et al. [15] found this incidence to be 17.6% in 97 patients. Our study found that the incidence of more than one primary carcinoma was high among BRCA1 and BRCA2 carriers. Among the three BRCA1 carriers, two had prostate cancer; among eight BRCA2 carriers, one had prostate cancer. Moreover, the overall incidence of second neoplasia was 33.5%, which is also higher than that found by Hemminki et al. [14]. Despite the high incidence of BRCA1/2 mutations, only one third of patients underwent genetic counselling, with even fewer opting to undergo genetic testing. Especially with the development of poly ADP ribose polymerase inhibitors targeted at BRCA1/2 breast cancers, genetic testing will have greater implications for male breast cancer patients and therefore should become an integral part of disease management [16]. A review of retrospective studies suggested that the 5 year local-regional recurrence rates in men who received radiotherapy for breast cancer ranged from 3 to 20% [17]. The result from our series was similar, with a 5 year localregional recurrence of 5.0%. During our study period there was a change in practice with regards to the use of radiotherapy to the chest wall, with its use associated with more recent treatment, which may be reflected in our finding that a more recent diagnosis was associated with a reduced risk. Studies have found that treatment options greatly influenced outcome in male breast cancer. Tamoxifen has been suggested to be effective in the management of male breast cancer, with both improved DFS and overall survival [2]. In a 2013 study by Goss et al. [12], breast-conserving surgery was associated with a higher recurrence rate compared with mastectomy. However, our study has not identified the use of tamoxifen or type of surgery to be significantly associated with outcome. This is further complicated by the fact that breast cancer management has changed with the development of modern systemic treatments, particularly within the past 10 years, which falls
7
within our study period. Moreover, our findings may be underpowered due to few non-censored patients with those treatments. Several recent studies have attempted to evaluate the histopathological characteristics of male breast cancer in order to stratify male breast cancer into prognostic subgroups [7,18]. Vermeulen et al. [18] conducted a central pathology review of 1483 male breast cancer cases and identified that histological grade was not significantly correlated with clinical outcome, unlike with female breast cancer patients. Similarly, our study found that histological grade was not an independent predictor of outcome. There are several limitations to this study. The first is that it was a retrospective study conducted at a single institution. Other limitations include the small sample size of metastatic patients and patients with recurrence, and incomprehensive data about tumour pathology, which limits the conclusions that can be drawn from this population. In addition, considering the wide range from 0 to 25 years of observation time, we had a short overall follow-up of 5.3 years. The strengths of our study include the comprehensive nature of assessed patient and disease characteristics, and treatment and disease outcomes. In addition, our study had one of the largest study populations from single-institution retrospective studies on male breast cancer [4]. Although the prognosis of male breast cancer has improved, there are still gaps in our knowledge in areas of treatment, prognosis and understanding of the biological mechanism. As male breast cancer treatment is still being modelled after treatment in females, validation of the efficacy of treatment options is required, particularly with regards to metastatic disease. These include the use of aromatase inhibitors and the efficacy of trastuzumab for HER2 enriched male breast cancers [4]. On the other hand, several recent collaborative international efforts have enabled large-scale characterisation of male breast cancer incidences, histopathological characteristics and outcomes [1,18]. Results from such initiatives may enable a better understanding of the aetiology and biology of male breast cancer, which can further inform treatment. Our results contribute to the current body of knowledge surrounding male breast cancer. In particular, this large retrospective series suggests that demographic factors have limited prognostic capacity, while total positive lymph nodes and invasive tumour size are independent predictors of local-regional recurrence and metastases.
Acknowledgements We thank the generous support of Bratty Family Fund, Michael and Karyn Goldstein Cancer Research Fund, Joey and Mary Furfari Cancer Research Fund, Pulenzas Cancer Research Fund, Joseph and Silvana Melara Cancer Research Fund and Ofelia Cancer Research Fund.
Please cite this article in press as: Wan BA, et al., Treatment Outcomes in Male Breast Cancer: A Retrospective Analysis of 161 Patients, Clinical Oncology (2018), https://doi.org/10.1016/j.clon.2018.02.026
8
B.A. Wan et al. / Clinical Oncology xxx (2018) 1e12
Appendix Table A1 Univariate analysis for factors predictive of overall survival. Values in bold are statistically significant (P < 0.05) Variable
Demographics Age at diagnosis (years) Year of diagnosis (continuous) Year of diagnosis categories (overall effect) 2010 to 2017 versus <2000 2000 to <2010 versus <2000 Gynaecomastia (presence versus absence) Family history (yes versus no) Other malignancy (yes versus no) T-stage (overall effect) 2 versus 1 3 versus 1 4 versus 1 T-stage (3/4 versus 1/2) N-stage (overall effect) 1 versus 0 2 versus 0 3 versus 0 2 versus 1 3 versus 1 3 versus 2 M-stage (1 versus 0) BRCA status (overall effect) BRCA1 versus negative BRCA2 versus negative Surgery type (mastectomy versus lumpectomy) Pathological characteristics Invasive tumour size (cm) Closest margin (log) Histological type (overall effect) DCIS versus other IDC/NOS versus other Inflammatory versus other Nottingham grade Skin involvement (yes versus no) Muscle involvement (yes versus no) LVI (yes versus no) Total positive nodes (log) Total removed nodes (log) ER/PRþ (yes versus no) Individual ERþ (yes versus no) Individual PRþ (yes versus no) HER2þ (yes versus no) Presence of DCIS in tumour (yes versus no) Cancer subtype (luminal A-like versus luminal B-like) Treatment received Use of chemotherapy (yes versus no) Number of chemotherapy regimens Adjuvant chemotherapy (yes versus no) Neoadjuvant chemotherapy (yes versus no) Palliative chemotherapy (yes versus no) Use of hormonal therapy in patients with ER/PRþ (yes versus no) Use of trastuzumab in patients with HER2þ (yes versus no)
R2
Independent covariate P-value
Hazard ratio
95% confidence interval of hazard ratio
0.0037 0.027 0.0695 0.0519 0.0244 0.7542 0.9558 0.4528 0.0589 0.0542 0.365 0.0082 0.1024 0.0282 0.1068 0.0708 0.0037 0.5319 0.202 0.756 <0.0001 0.5022 0.4526 0.2914 0.405
1.044 1.066
1.014 1.007
1.074 1.128
2.897 2.714 1.125 1.023 0.798
0.991 1.138 0.538 0.464 0.442
8.464 6.472 2.354 2.252 1.439
2.187 2.28 4.068 2.005
0.986 0.383 1.438 0.87
4.85 13.566 11.504 4.62
2.157 3.255 3.988 1.509 1.849 1.225 11.612
0.848 0.905 1.568 0.415 0.719 0.34 5.176
5.49 11.711 10.147 5.482 4.754 4.417 26.05
4.011 2.438 3.356
0.107 0.466 0.194
150.443 12.768 58.003
5.05% 3.22% 4.14%
0.06% 0.00% 0.36% 5.97%
1.76% 7.89
15.02% 4.85%
0.77%
0.0001 0.0948 0.4054 0.653 0.8048 0.1904 0.5674 0.1411 0.2889 0.009 0.0015 0.0056 0.9079 0.8742 0.3124 0.0954 0.4956 0.0686
1.834 2.218
1.343 0.871
2.505 5.646
11.03% 2.69% 1.56%
0.49 1.179 3.324 1.181 1.862 2.015 2.986 1.727 2.719 0.889 0.851 0.614 2.17 1.542 0.435
0.022 0.32 0.551 0.668 0.814 0.552 1.315 1.232 1.341 0.122 0.116 0.238 0.873 0.444 0.177
10.981 4.338 20.066 2.088 4.259 7.351 6.78 2.421 5.517 6.502 6.252 1.582 5.392 5.362 1.066
0.28% 2.12% 0.62% 6.60% 7.60% 7.75% 0.01% 0.02% 0.71% 2.27% 0.66% 2.87%
0.0827 0.0036 0.2885 0.2344 0.0001 0.3093
1.729 1.417 0.685 2.052 3.549 1.712
0.932 1.121 0.34 0.627 1.853 0.607
3.209 1.791 1.378 6.715 6.796 4.823
2.05% 4.29% 0.75% 0.72% 7.12% 0.84%
0.1683
0.299
0.054
1.666
12.02%
Please cite this article in press as: Wan BA, et al., Treatment Outcomes in Male Breast Cancer: A Retrospective Analysis of 161 Patients, Clinical Oncology (2018), https://doi.org/10.1016/j.clon.2018.02.026
B.A. Wan et al. / Clinical Oncology xxx (2018) 1e12
9
Table A1 (continued ) Variable
Breast/chest wall radiation (yes versus no) Nodes/axilla radiation (yes versus no) Boost radiation (yes versus no) Palliative radiation (yes versus no)
R2
Independent covariate P-value
Hazard ratio
95% confidence interval of hazard ratio
0.0091 0.1976 0.8748 <0.0001
2.371 1.511 1.121 3.909
1.24 0.806 0.271 2.126
4.535 2.832 4.638 7.19
4.82% 1.06% 0.02% 9.92%
DCIS: Ductal carcinoma in situ; IDC/NOS: Invasive ductal carcinoma/Not otherwise specified; LVI: Lymphovascular invasion; ER: Estrogen receptor; PR: Progesterone receptor.
Table A2 Univariate analysis for factors predictive of local-regional recurrence. Values in bold are statistically significant (P < 0.05) Variable
Demographics Age at diagnosis (years) Year of diagnosis (continuous) Year of diagnosis categories (2000 versus <2000) Gynaecomastia (presence versus absence) Family history (yes versus no) Other malignancy (yes versus no) T-stage (overall effect) 2 versus 1 3 versus 1 4 versus 1 T-stage (3/4 versus 1/2) N-stage (overall effect) 1 versus 0 2 versus 0 3 versus 0 Pathological characteristics Invasive tumour size (cm) Closest margin (log) Historical type (IDC/NOS versus other) Nottingham grade Skin involvement (yes versus no) Muscle involvement (yes versus no) LVI (yes versus no) Total positive nodes (log) Total removed nodes (log) Individual PRþ (yes versus no) HER2þ (yes versus no) Presence of DCIS in tumour (yes versus no) Cancer subtype (luminal A-like versus luminal B-like) Treatment received Use of chemotherapy (yes versus no) Number of chemotherapy regimens Adjuvant chemotherapy (yes versus no) Neoadjuvant chemotherapy (yes versus no) Palliative chemotherapy (yes versus no) Use of hormonal therapy in patients with ER/PRþ (yes versus no) Breast/chest wall radiation (yes versus no) Nodes/axilla radiation (yes versus no) Boost radiation (yes versus no) Palliative radiation (yes versus no)
R2
Independent covariate P-value
Subdistribution hazard ratio with 95% confidence interval
0.1007 0.003 0.2871 0.9329 0.7034 0.7574 0.5255 0.2161 0.1764 0.4273 0.46 0.1363 0.0632 0.2012 0.0195
1.051 0.9 0.524 1.064 1.367 0.823
0.99 0.839 0.159 0.251 0.273 0.24
1.115 0.965 1.723 4.505 6.845 2.826
2.692 4.807 2.363 1.681
0.561 0.494 0.283 0.424
12.923 46.815 19.757 6.666
7.78 5.943 14.116
0.893 0.386 1.530
67.786 91.381 130.257
0.2909 0.087 0.7724 0.0483 0.4688 0.2528 0.3892 0.009 0.0907 0.6179 0.7634 0.81 0.7432
1.189 0.126 0.761 3.187 1.695 2.514 1.816 2.096 2.183 0.581 1.389 0.762 0.701
0.862 0.012 0.119 1.009 0.407 0.518 0.467 1.202 0.884 0.069 0.164 0.083 0.083
1.64 1.351 4.857 10.067 7.062 12.2 7.065 3.653 5.395 4.917 11.79 6.989 5.888
0.38% 2.66% 0.05% 3.26% 0.50% 0.44% 0.64% 4.44% 1.87% 0.20% 0.09% 0.07% 0.11%
0.0553 0.0581 0.1177 0.4929 0.8677 0.4898 0.0317 0.0078 0.03 0.4455
3.651 1.389 2.565 2.067 1.195 2.021 8.907 8.268 5.794 1.814
0.971 0.989 0.788 0.259 0.147 0.274 1.212 1.746 1.185 0.393
13.727 1.95 8.351 16.462 9.711 14.901 65.477 39.16 28.328 8.372
3.20% 1.22% 1.72% 0.28% 0.02% 0.44% 5.54% 6.67% 2.49% 0.37%
1.96% 3.30% 0.68% 0.00% 0.16% 0.07% 1.88%
0.32% 6.57%
IDC/NOS: Invasive ductal carcinoma/Not otherwise specified; LVI: Lymphovascular invasion; ER: Estrogen receptor; PR: Progesterone receptor; HER2: Human epidermal growth factor receptor 2; DCIS: Ductal carcinoma in situ. Please cite this article in press as: Wan BA, et al., Treatment Outcomes in Male Breast Cancer: A Retrospective Analysis of 161 Patients, Clinical Oncology (2018), https://doi.org/10.1016/j.clon.2018.02.026
10
B.A. Wan et al. / Clinical Oncology xxx (2018) 1e12
Table A3 Univariate analysis for factors predictive of distant metastases. Values in bold are statistically significant (P < 0.05) Variable
Demographics Age at diagnosis (years) Year of diagnosis (continuous) Year of diagnosis categories (overall effect) 2010 to 2017 versus <2000 2000 to <2010 versus <2000 Gynaecomastia (presence versus absence) Family history (yes versus no) Other malignancy (yes versus no) T-stage (overall effect) 2 versus 1 3 versus 1 4 versus 1 T-stage (3/4 versus 1/2) N-stage (overall effect) 1 versus 0 2 versus 0 3 versus 0 2 versus 1 3 versus 1 3 versus 2 BRCA status (BRCA2 versus negative) Pathological characteristics Invasive tumour size (cm) Closest margin (log) Historical type (IDC/NOS versus other) IDC/NOS versus other Inflammatory versus other Nottingham grade Skin involvement (yes versus no) Muscle involvement (yes versus no) LVI (yes versus no) Total positive nodes (log) Total removed nodes (log) ER/PRþ (yes versus no) Individual ERþ (yes versus no) Individual PRþ (yes versus no) HER2D (yes versus no) Presence of DCIS in tumour (yes versus no) Cancer subtype (luminal B-like versus luminal A-like) Treatment received Use of chemotherapy (yes versus no) Number of chemotherapy regimens Adjuvant chemotherapy (yes versus no) Neoadjuvant chemotherapy (yes versus no) Palliative chemotherapy (yes versus no) Use of hormonal therapy in patients with ER/PRþ (yes versus no) Use of trastuzumab in patients with HER2þ (yes versus no) Breast/chest wall radiation (yes versus no) Nodes/axilla radiation (yes versus no) Boost radiation (yes versus no) Palliative radiation (yes versus no)
R2
Independent covariate P-value
Subdistribution hazard ratio with 95% confidence interval
0.3323 0.0791 0.1843 0.0663 0.1308 0.9396 0.3249 0.0277 0.1089 0.0481 0.4457 0.018 0.1601 <0.0001 0.6086 0.045 <0.0001 0.1543 0.0013 0.1822 0.463
0.981 1.053
0.942 0.994
1.02 1.116
4.193 3.104 0.96 0.639 0.322
0.908 0.714 0.335 0.262 0.118
19.356 13.489 2.749 1.559 0.883
3.081 2.345 5.612 2.049
1.009 0.262 1.344 0.753
9.403 20.962 23.431 5.574
1.467 4.835 12.527 3.296 8.54 2.591 1.856
0.338 1.036 3.863 0.639 2.311 0.64 0.356
6.358 22.571 40.618 17.012 31.555 10.491 9.685
0.0007 0.6418 0.6754 0.8243 0.52 0.9223 0.359 0.2766 0.0859 <0.0001 0.0157 0.5086 0.5043 0.4396 0.0238 0.9905 0.0208
1.688 1.257
1.249 0.479
2.281 3.299
0.853 2.43 0.973 1.552 2.525 2.327 2.883 3.226 0.505 0.495 2.099 3.06 0.992 3.133
0.209 0.162 0.559 0.607 0.476 0.888 1.84 1.247 0.067 0.063 0.32 1.16 0.276 1.189
3.475 36.359 1.693 3.969 13.397 6.1 4.515 8.348 3.823 3.898 13.765 8.07 3.563 8.251
0.01% 0.84% 0.90% 2.91% 15.78% 7.42% 0.29% 0.34% 0.58% 4.05% 0.01% 4.40%
0.0002 <0.0001 0.0021 0.008 <0.0001 0.2388 0.7969 0.0011 0.0785 0.993 <0.0001
6.614 2.129 3.383 3.698 12.071 2.231 0.795 10.535 2.04 0.992 19.809
2.468 1.705 1.555 1.408 6.13 0.587 0.139 2.548 0.922 0.164 9.568
17.729 2.657 7.36 9.716 23.769 8.485 4.556 43.555 4.515 6.004 41.008
13.70% 16.25% 6.59% 3.01% 16.21% 1.19% 0.40% 13.35% 2.24% 0.01% 28.45%
0.76% 1.69% 3.10%
0.00% 0.94% 4.45% 5.87%
1.45% 15.73%
1.44% 8.27% 0.18% 0.59%
IDC/NOS: Invasive ductal carcinoma/Not otherwise specified; LVI: Lymphovascular invasion; ER: Estrogen receptor; PR: Progesterone receptor; HER2: Human epidermal growth factor receptor 2; DCIS: Ductal carcinoma in situ.
Please cite this article in press as: Wan BA, et al., Treatment Outcomes in Male Breast Cancer: A Retrospective Analysis of 161 Patients, Clinical Oncology (2018), https://doi.org/10.1016/j.clon.2018.02.026
B.A. Wan et al. / Clinical Oncology xxx (2018) 1e12
11
Table A4 Univariate analysis for factors predictive of any recurrence/metastases. Values in bold are statistically significant (P < 0.05) Variable
Demographics Age at diagnosis (years) Year of diagnosis (continuous) Year of diagnosis categories (overall effect) 2010 to 2017 versus <2000 2000 to <2010 versus <2000 Gynaecomastia (presence versus absence) Family history (yes versus no) Other malignancy (yes versus no) T-stage (overall effect) 2 versus 1 3 versus 1 4 versus 1 3 versus 2 4 versus 2 4 versus 3 T-stage (3/4 versus 1/2) N-stage (overall effect) 1 versus 0 2 versus 0 3 versus 0 2 versus 1 3 versus 1 3 versus 2 BRCA status (BRCA2 versus negative) Pathological characteristics Invasive tumour size (cm) Closest margin (log) Historical type (IDC/NOS versus other) IDC/NOS versus other Inflammatory versus other Nottingham grade Skin involvement (yes versus no) Muscle involvement (yes versus no) LVI (yes versus no) Total positive nodes (log) Total removed nodes (log) ER/PRþ (yes versus no) Individual ERþ (yes versus no) Individual PRþ (yes versus no) HER2þ (yes versus no) Presence of DCIS in tumour (yes versus no) Cancer subtype (luminal B-like versus luminal A-like) Treatment received Use of chemotherapy (yes versus no) Number of chemotherapy regimens Adjuvant chemotherapy (yes versus no) Neoadjuvant chemotherapy (yes versus no) Palliative chemotherapy (yes versus no) Use of hormonal therapy in ER/PRþ patients (yes versus no) Use of trastuzumab in patients with HER2þ (yes versus no) Breast/chest wall radiation (yes versus no) Nodes/axilla radiation (yes versus no) Boost radiation (yes versus no) Palliative radiation (yes versus no)
R2
Independent covariate P-value
Subdistribution hazard ratio with 95% confidence interval
0.7448 0.8658 0.765 0.4828 0.5386 0.7687 0.4458 0.0306 0.035 0.0128 0.073 0.0062 0.803 0.3703 0.73 0.054 <0.0001 0.0692 0.0119 <0.0001 0.2937 0.0034 0.1898 0.463
0.994 1.004
0.961 0.954
1.029 1.057
1.44 1.321 1.135 0.726 0.407
0.52 0.544 0.488 0.319 0.18
3.984 3.205 2.637 1.653 0.919
3.513 4.169 5.47 1.187 1.557 1.312 2.185
1.306 0.875 1.618 0.309 0.591 0.281 0.987
9.453 19.855 18.489 4.553 4.101 6.133 4.839
2.823 5.452 12.204 1.931 4.323 2.238 1.856
0.922 1.454 4.171 0.565 1.622 0.671 0.356
8.648 20.448 35.712 6.596 11.522 7.466 9.685
0.0004 0.6274 0.7497 0.5703 0.794 0.1173 0.3152 0.0643 0.0315 <0.0001 0.0078 0.7992 0.7697 0.9125 0.061 0.8443 0.0525
1.573 0.779
1.226 0.285
2.02 2.132
0.741 1.407 1.551 1.519 2.918 2.48 2.604 2.588 0.759 0.728 1.083 2.415 0.898 2.486
0.264 0.109 0.896 0.672 0.938 1.084 1.83 1.285 0.091 0.087 0.261 0.96 0.308 0.99
2.084 18.24 2.687 3.437 9.076 5.674 3.704 5.211 6.358 6.09 4.497 6.074 2.619 6.24
1.58% 0.98% 1.72% 4.41% 17.47% 7.26% 0.05% 0.08% 0.01% 2.72% 0.05% 2.95%
<0.0001 <0.0001 0.0002 0.0333 <0.0001 0.203 0.7969 <0.0001 0.0011 0.1038 <0.0001
5.363 2.042 3.636 3.1 8.338 2.02 0.795 9.982 3.24 2.552 11.923
2.402 1.655 1.828 1.093 4.469 0.684 0.139 3.136 1.604 0.825 6.46
11.971 2.519 7.232 8.788 15.56 5.965 4.556 31.775 6.547 7.894 22.008
14.91% 17.29% 9.45% 2.40% 13.26% 1.31% 0.40% 17.52% 7.76% 1.36% 23.63%
0.08% 0.02% 0.37%
0.06% 0.59% 3.98% 8.38%
2.34% 17.16%
1.44% 7.84% 0.24% 48.61%
IDC/NOS: Invasive ductal carcinoma/Not otherwise specified; LVI: Lymphovascular invasion; ER: Estrogen receptor; PR: Progesterone receptor; HER2: Human epidermal growth factor receptor 2; DCIS: Ductal carcinoma in situ.
Please cite this article in press as: Wan BA, et al., Treatment Outcomes in Male Breast Cancer: A Retrospective Analysis of 161 Patients, Clinical Oncology (2018), https://doi.org/10.1016/j.clon.2018.02.026
12
B.A. Wan et al. / Clinical Oncology xxx (2018) 1e12
References [1] Miao H, Verkooijen HM, Chia K, Bouchardy C, Pukkala E, Larønningen S, et al. Incidence and outcome of male breast cancer: an international population-based study. J Clin Oncol 2017;29:4381e4386. https://doi.org/10.1200/JCO.2011.36.8902. [2] Ruddy KJ, Winer EP. Male breast cancer: risk factors, biology, diagnosis, treatment, and survivorship. Ann Oncol 2013;24: 1434e1443. https://doi.org/10.1093/annonc/mdt025. [3] Fentiman IS. Male breast cancer is not congruent with the female disease. Crit Rev Oncol Hematol 2016;101:119e124. https://doi.org/10.1016/j.critrevonc.2016.02.017. [4] Losurdo A, Rota S, Gullo G, Masci G, Torrisi R, Bottai G, et al. Controversies in clinicopathological characteristics and treatment strategies of male breast cancer: a review of the literature. Crit Rev Oncol Hematol 2017;113:283e291. https:// doi.org/10.1016/j.critrevonc.2017.03.013. [5] Hill TD, Khamis HJ, Tyczynski JE, Berkel HJ. Comparison of male and female breast cancer incidence trends, tumor characteristics, and survival. Ann Epidemiol 2005;15:773e780. https://doi.org/10.1016/j.annepidem.2005.01.001. [6] Giordano SH, Cohen DS, Buzdar AU, Perkins G, Hortobagyi GN. Breast carcinoma in men: a population-based study. Cancer 2004;101:51e57. https://doi.org/10.1002/cncr.20312. [7] Abreu MH, Afonso N, Abreu PH, Menezes F, Lopes P, Henrique R, et al. Male breast cancer: looking for better prognostic subgroups. Breast 2016;26:18e24. https://doi.org/ 10.1016/j.breast.2015.12.001. [8] Patten DK, Sharifi LK, Fazel M. New approaches in the management of male breast cancer. Clin Breast Cancer 2013;13: 309e314. https://doi.org/10.1016/j.clbc.2013.04.003. [9] Korde LA, Zujewski JA, Kamin L, Giordano S, Domchek S, Anderson WF, et al. Multidisciplinary meeting on male breast cancer: summary and research recommendations. J Clin Oncol 2010;28:2114e2122. https://doi.org/10.1200/JCO.2009.25.5729.
[10] Ge Y, Sneige N, Eltorky MA, Wang Z, Lin E, Gong Y, et al. Immunohistochemical characterization of subtypes of male breast carcinoma. Breast Cancer Res 2009;11:R28. https://doi. org/10.1186/bcr2258. [11] Fine JP, Gray RJ. A proportional hazards model for the subdistribution of a competing risk. J Am Stat Assoc 1999;94: 496e509. [12] Goss PE, Reid C, Pintilie M, Lim R, Miller N. Male breast carcinoma: a review of 229 patients who presented to the Princess Margaret Hospital during 40 years: 1955e1996. Cancer 1999;85:629e639. https://doi.org/10.1002/(SICI)10970142(19990201)85:3<629::AID-CNCR13>3.0.CO;2-V. [13] Adami HO, Malker B, Holmberg L, Persson I, Stone B. The relation between survival and age at diagnosis in breast cancer. N Engl J Med 1986;315:559e563. lo G, Boffetta P, Mellemjkaer L, Tracey E, [14] Hemminki K, Sce Andersen A, et al. Second primary malignancies in patients with male breast cancer. Br J Cancer 2005;92:1288e1292. https://doi.org/10.1038/sj.bjc.6602505. [15] Masci G, Caruso M, Caruso F, Salvini P, Carnaghi C, Giordano L, et al. Clinicopathological and immunohistochemical characteristics in male breast cancer: a retrospective case series. Oncologist 2015;20:586e592. [16] Dziadkowiec KN, Ga˛ siorowska E, Nowak-Markwitz E, Jankowska A. PARP inhibitors: review of mechanisms of action and BRCA1/2 mutation targeting. Menopausal Rev 2016;4: 215e219. https://doi.org/10.5114/pm.2016.65667. [17] Fentiman I, Fourquet A, Hortobagyi G. Male breast cancer. Lancet 2006;367:595e604. [18] Vermeulen MA, Slaets L, Cardoso F, Giordano SH, Tryfonidis K, van Diest PJ, et al. Pathological characterisation of male breast cancer: results of the EORTC 10085/TBCRC/BIG/NABCG International Male Breast Cancer Program. Eur J Cancer 2017;82: 219e227. https://doi.org/10.1016/j.ejca.2017.01.034.
Please cite this article in press as: Wan BA, et al., Treatment Outcomes in Male Breast Cancer: A Retrospective Analysis of 161 Patients, Clinical Oncology (2018), https://doi.org/10.1016/j.clon.2018.02.026
Clinical Oncology journal homepage: www.clinicaloncologyonline.net
Original Article
Treatment Outcomes in Male Breast Cancer: A Retrospective Analysis of 161 Patients B.A. Wan, V. Ganesh, L. Zhang, P. Sousa, L. Drost, J. Lorentz, D. Vesprini, J. Lee, E. Rakovitch, F.-I. Lu, A. Eisen, C. Yee, H. Lam, E. Chow Odette Cancer Centre, Sunnybrook Health Sciences Centre, University of Toronto, Toronto, Ontario, Canada Received 10 October 2017; received in revised form 28 December 2017; accepted 11 January 2018
Abstract Aims: Male breast cancer is a rare disease with limited evidence-based guidelines for treatment. This study aimed to identify demographic, pathological and clinical factors associated with its prognosis. Materials and methods: A retrospective review of 161 male breast cancer patients diagnosed at a single institution from 1987 to June 2017 was conducted. Patient demographics, disease characteristics, treatment and outcome were extracted and included in competing-risk analysis and the univariate Cox proportional hazard model for univariate analysis. Factors with P < 0.10 were included in multivariable analysis. Results: The mean age at diagnosis was 67 years (standard deviation ¼ 11.2) and the median follow-up duration was 5.3 years (range 0e25 years). There were 48 deaths, including 23 cancer-specific deaths. The actuarial median survival was 19.9 years. In multivariable analysis, factors associated with overall survival were size of tumours (hazard ratio 2.0; 95% confidence interval 1.4e2.7, P < 0.0001) and diagnosis of metastatic disease (hazard ratio 8.7; 95% confidence interval 1.9e40.6; P ¼ 0.006). Of 138 patients without metastases at diagnoses, 11 had local-regional recurrence and 26 had distant metastases. In the multivariable model for local-regional recurrence, a more recent year of diagnosis was associated with reduced risk (hazard ratio 0.9, 95% confidence interval 0.8e1.0, P ¼ 0.008), whereas more positive lymph nodes was associated with higher risk (hazard ratio 2.2, 95% confidence interval 1.2e4.0, P ¼ 0.01). A higher risk of metastases was associated with more positive lymph nodes (hazard ratio 1.9; 95% confidence interval 1.1e3.3; P ¼ 0.03) and tumour size (hazard ratio 1.8; 95% confidence interval 1.1e2.9; P ¼ 0.01). A higher risk of any recurrence or metastases was associated with the number of positive nodes (hazard ratio 1.9; 95% confidence interval 1.2e3.0; P ¼ 0.005) and tumour size (hazard ratio 1.6; 95% confidence interval 1.1e2.2; P ¼ 0.01). Conclusion: In general, tumour size and more positive lymph nodes were associated with worse prognosis. Larger powered studies are needed to identify prognostic factors with smaller effect sizes. Ó 2018 The Royal College of Radiologists. Published by Elsevier Ltd. All rights reserved.
Key words: Breast cancer; male; treatment outcomes
Introduction Male breast cancer is rare and accounts for 0.6% of all breast cancer diagnoses [1]. This has resulted in a lack of evidence-based treatment guidelines due to difficulties in conducting large-scale randomised controlled trials. The current understanding of its biology and treatment guidelines has been based on epidemiological and retrospective
Author for correspondence: E. Chow, Department of Radiation Oncology, Odette Cancer Centre, Sunnybrook Health Sciences Centre, 2075 Bayview Avenue, Toronto, Ontario M4N 3M5, Canada. Tel: þ1-416-480-4974. E-mail address: [email protected] (E. Chow).
studies, and extrapolation from studies on female breast cancer. Factors associated with the risk of male breast cancer include family history, increased oestrogen exposure or hypoandrogenism, radiation/occupational exposure and heritable elements such as BRCA1 and BRCA2 gene mutations [2]. The characteristics of male breast cancer have been found to resemble postmenopausal female breast cancer in several studies [3,4]. However, there is increasing evidence that male breast cancer differs in aetiology, clinical-pathological presentation and outcomes when compared with female breast cancer. For example, lobular carcinomas are the second most common subtype in female breast cancer (11.8%), but present rarely in males (1%) [5,6].
https://doi.org/10.1016/j.clon.2018.02.026 0936-6555/Ó 2018 The Royal College of Radiologists. Published by Elsevier Ltd. All rights reserved.
Please cite this article in press as: Wan BA, et al., Treatment Outcomes in Male Breast Cancer: A Retrospective Analysis of 161 Patients, Clinical Oncology (2018), https://doi.org/10.1016/j.clon.2018.02.026
2
B.A. Wan et al. / Clinical Oncology xxx (2018) 1e12
In addition, male breast cancer is almost always oestrogen hormone receptor positive and when compared with female breast cancer there is an under-representation of BRCA1 mutations and an over-representation of BRCA2 mutations [3]. Moreover, an analysis of the immunehistopathological characteristics of male breast cancer by Abreu et al. [7] suggested that male breast cancer may be divided into multiple prognostic subgroups that highlight differences in tumourigenesis. Management of male breast cancer includes surgical excision, radiotherapy and systemic therapy in the form of chemotherapy, hormonal therapy for oestrogen receptor (ER) positive or progesterone receptor (PR) positive patients, or trastuzumab for patients with human epidermal growth factor receptor (HER2/neu) overexpression [8]. Although the overall survival of male breast cancer has improved in the last decade, there remains a lack of evidence-based data for its management [9]. The objectives of this present study were to evaluate the demographic, clinical, pathological characteristics and treatments associated with outcomes in male breast cancer.
Materials and Methods A retrospective review of patients who were diagnosed from 1987 to June 2017 at the Odette Cancer Centre was conducted. Ethics approval from Sunnybrook Health Sciences Centre was obtained before the start of the study. Patient demographic, treatment, pathology, biomarkers, years of cancer diagnosis (1987 to <2000, 2000 to <2010, 2010 to 2017) and follow-up data were extracted. Biomarker data on ER, PR and HER2/neu status were used to divide cancer into subtypes of luminal A-like (ER/PRþ, HER2/ neue), luminal B-like (ER/PRþ, HER2/neuþ) or triple negative (ER/PRe, HER2/neue) based on definitions by Ge et al. [10]. Clinical-pathological features were individually correlated to overall survival, risk of local-regional recurrence and risk of metastases. Overall survival in years was defined as the time from diagnosis to the date of death or to the last follow-up. Risk of recurrence or metastases was calculated for patients without metastases at diagnosis (M0) and was defined as the time from breast cancer diagnosis to breast cancer recurrence/metastases. Patients, disease and treatment characteristics were summarised as mean, standard deviation and range for continuous variables, and proportions for categorical variables. The univariate Cox proportional hazard model was used to identify significant covariates related to overall survival. KaplaneMeier overall survival curves, hazard ratios, 95% confidence intervals and P-values were generated. The generalised R2 statistic (between 0 and 1) was calculated based on the likelihood ratio statistic for testing the global null hypothesis; the larger the R2, the stronger the association with the outcome [10]. A competing risk analysis was conducted for local-regional recurrence, distant metastases or any recurrence/distant metastases in nonmetastatic patients. The cumulative incidence function
was estimated and plotted as well. To search for significant predictive factors of local-regional recurrence, distant metastases or any recurrence/distant metastases, univariate Cox proportional subdistribution hazard models were carried out using Fine and Gray’s method, considering death as the competing risk [11]. Subdistribution hazard ratios with 95% confidence intervals and P-value were estimated for each factor. In the multivariable analysis, all variables with P < 0.10 from the univariate analysis were selected for inclusion in the backward stepwise selection procedure. The final model would only keep the significant predictive factors with P < 0.05. All analyses were conducted using Statistical Analysis Software (SAS version 9.4 for Windows, Cary, NC) and R package (version 3.2.0).
Results Baseline Patient, Pathological and Treatment Characteristics Patient characteristics are summarised in Table 1. The median age of 161 patients at diagnosis was 67 years (range 34e92). The median duration of follow-up since diagnosis was 5.3 years (range 0e25 years). Only two patients (1.2%) had bilateral breast cancer. Twenty-five (15.5%) patients had gynaecomastia. Seventy-three (45.3%) patients had a family history of cancer. Of the 34 patients with known BRCA status, three had BRCA1 mutation (8.8%) and eight had BRCA2 mutation (23.5%). This included one patient who had a BRCA1 mutation of uncertain significance. There were 54 patients with previous, secondary or subsequent malignancies (33.5%), the most common of which was prostate cancer (n ¼ 23) followed by basal cell carcinoma (n ¼ 7). Disease characteristics and pathology are summarised in Table 2. Most patients presented with stage 2 disease (n ¼ 63, 39.1%) and no nodal involvement (n ¼ 61, 37.9%), while 11 patients had metastatic disease (6.8%). The most common histological type was invasive ductal carcinoma, no specific type (n ¼ 146, 90.7%) although there were 15 (9.3%) patients with other histological types. Overall, most patients were luminal A-like (n ¼ 86, 53.4%), followed by luminal B-like (n ¼ 15, 9.3%), with one triple-negative patient. The most common Nottingham histologic score was 2 (n ¼ 75, 63.0%). Treatments are summarised in Table 3. Most patients underwent surgery (n ¼ 143, 88.8%), the most common of which was mastectomy (n ¼ 133, 93.0%). Other treatment modalities included radiotherapy to the breast or chest wall (n ¼ 95), chemotherapy (n ¼ 69), systemic hormonal therapy (for 122 of 142 patients with ER/PR overexpression) or trastuzumab (for nine of 16 patients with HER2/neu overexpression). Outcome Of 161 patients there were 48 deaths (median time to death: 5.4 years; interquartile range [IQR] 2.1e9.0), of which 23 (47.9%) were breast cancer specific. Figure 1A shows the
Please cite this article in press as: Wan BA, et al., Treatment Outcomes in Male Breast Cancer: A Retrospective Analysis of 161 Patients, Clinical Oncology (2018), https://doi.org/10.1016/j.clon.2018.02.026
B.A. Wan et al. / Clinical Oncology xxx (2018) 1e12 Table 1 Patient characteristics Patient characteristics
3
Table 2 Disease characteristics and pathology Total patients (N ¼ 161) n
Disease characteristics and pathology
%
Age at diagnosis (years) Mean standard deviation 66.9 11.2 Minimum, maximum 34.0, 92.0 Duration of follow-up since diagnosis (years) Mean standard deviation 6.6 5.5 Diagnosis year categories 1987 to <2000 24 14.9% 2000 to <2010 77 47.8% 2010 to 2017 60 37.3% Gynaecomastia Absence 136 84.5% Presence 25 15.5% Family history of breast cancer No 32 19.9% Yes 73 45.3% Unknown 56 34.8% Genetic counselling No 105 65.2% Yes 53 32.9% Declined 3 1.9% BRCA status among tested (n ¼ 34) BRCA1 3 8.8% BRCA2 8 23.5% Negative 23 14.3% Other malignancy No 107 66.5% Yes 54 33.5% Type of second primary cancer from 54 patients who had ‘other malignancy’ Prostate 23 38.3% Basal cell carcinoma 7 11.7% Bladder 5 8.3% Renal cell carcinoma 4 6.7% Lung 3 5.0% Lymphoma 3 5.0% Melanoma 3 5.0% Other malignancy 12 20.0%
KaplaneMeier overall survival curve, with the dashed lines indicating the 95% confidence interval. The actuarial median survival time from KaplaneMeier estimations for all patients was 19.9 years (95% confidence interval 11.2e25.4). Among the 138 patients without metastases at diagnosis, 11 patients had local-regional recurrence (8.0%, median time to recurrence 5.0 years; IQR 1.7e8.3) and 26 patients had distant metastasis (18.8%, median time to metastases 4.7 years; IQR 1.3e8.3). There were 34 patients who had any recurrence or metastases, with the median time to event of 4.38 years (IQR 1.2e7.8). Overall Survival Analysis The univariate Cox proportional hazard model for overall survival of all patients using demographics,
Total patients (N ¼ 161) n
Breast cancer laterality Left Right Bilateral Unknown Size of invasive component (cm) Mean standard deviation T-stage 1 2 3 4 Tx N-stage 0 1 2 3 Nx M-stage 0 1 Mx Histologic type IDC, NOS Papillary Inflammatory Mucinous DCIS Invasive apocrine Micropapillary Unknown Nottingham grade 1 2 3 Unknown Skin involvement No Yes Unknown Muscle involvement No Yes Unknown LVI No Yes Unknown Total nodes removed Mean standard deviation Total nodes positive Mean standard deviation ER/PR status Negative Positive Unknown
%
79 49.1% 78 48.5% 2 1.2% 2 1.2% N ¼ 119 2.47 1.35 48 63 7 11 31
29.8% 39.1% 4.4% 6.8% 19.3%
61 36 10 14 40
37.9% 22.4% 6.2% 8.7% 24.8%
138 11 12
85.7% 6.8% 7.5%
89 5 3 3 2 1 1 57
55.3% 3.1% 1.9% 1.9% 1.2% 0.6% 0.6% 35.4%
9 75 35 42
5.6% 46.6% 21.7% 26.1%
63 34 64
39.1% 21.1% 39.8%
150 5 6
93.2% 3.1% 3.7%
59 54 48
36.7% 33.5% 29.8%
10.9 7.5 2.4 4.1 3 142 16
1.9% 88.2% 9.9%
(continued on next page)
Please cite this article in press as: Wan BA, et al., Treatment Outcomes in Male Breast Cancer: A Retrospective Analysis of 161 Patients, Clinical Oncology (2018), https://doi.org/10.1016/j.clon.2018.02.026
4
B.A. Wan et al. / Clinical Oncology xxx (2018) 1e12 Table 2 (continued ) Disease characteristics and pathology
Table 3 (continued ) Total patients (N ¼ 161) n
%
ER status Negative 3 Positive 126 Unknown 32 PR status Negative 15 Positive 113 Unknown 33 HER2/Neu status Negative 88 Positive 16 Unknown 57 Cancer subtype Luminal A-like 86 Luminal B-like 15 Triple negative 1 Unknown 59 Nuclear grade 1 1 2 27 3 15 Unknown 118 Presence of concurrent DCIS with invasive disease No 19 Yes 58 Unknown 84
1.9% 78.3% 19.9% 9.3% 70.2% 20.5% 54.7% 9.9% 35.4% 53.4% 9.3% 0.6% 36.7% 0.6% 16.8% 9.3% 73.3% 11.8% 36.0% 52.2%
IDC, NOS: Invasive ductal carcinoma, not otherwise specified; DCIS: Ductal carcinoma in situ; LVI: Lympho-vascular invasion; ER: Estrogen receptor; PR: Progesterone receptor; HER2: Human epidermal growth factor receptor 2.
Table 3 Treatment summary Treatment summary
Surgery No Yes Unknown Surgery type (n ¼ 143) Lumpectomy Mastectomy Unknown Use of radiotherapy Breast/chest wall radiotherapy No Yes Unknown Nodes/axilla radiotherapy No Yes Unknown
Treatment summary
Total patients (N ¼ 161) n
%
13 143 5
8.1% 88.8% 3.1%
8 133 2
5.6% 93.0% 1.4%
57 95 9
35.4% 59.0% 5.6%
106 43 12
65.8% 26.7% 7.5%
Boost radiotherapy No Yes Unknown Palliative radiotherapy No Yes Use of chemotherapy Chemotherapy No Yes Adjuvant Neoadjuvant Palliative Unknown Number of chemotherapy regimens 1 2 3 4 Chemotherapy regimens from 69 patients AC FEC-D Paclitaxel Docetaxel Capecitabine CEF CMF AC paclitaxel Other chemotherapy regimens For patients who were ER/PRþ (n ¼ 142) Use of hormonal therapy No Yes Name of hormonal therapy Tamoxifen Letrozole Anastrazole Exemestane Other/unknown For patients who were HER2/Neuþ (n ¼ 16) Use of trastuzumab No Yes
Total patients (N ¼ 161) n
%
152 7 2
94.4% 4.4% 1.2%
131 30
81.4% 18.6%
78 69 43 10 19 14
48.5% 42.9% 26.7% 6.2% 11.8% 8.7%
55 5 4 5
34.2% 3.1% 2.5% 3.1%
12 10 9 8 7 6 6 5 21
14.3% 11.9% 10.7% 9.5% 8.3% 7.1% 7.1% 6.0% 25.0%
20 122
14.1% 85.9%
103 17 16 9 9
62.8% 10.4% 9.8% 5.5% 5.5%
7 9
43.8% 56.3%
AC: Adriamycin and cyclophosphamide; CEF: Cyclophosphamide, epirubicin, fluorouracil; CMF: Cyclophosphamide, methotrexate, fluorouracil; ER/PR: Estrogen receptor/Progesterone receptor; HER2: human epidermal growth factor receptor 2.
pathological characteristics and treatment received as variables identified 15 significant predictive factors (Table A1). Older age, more recent year of diagnosis, stage T4, N3 or M1 disease, larger sizes of invasive tumours, presence of LVI, more total nodes removed or positive nodes, more chemotherapy regimens and palliative chemotherapy or radiotherapy treatment and palliative radiotherapy treatment were associated with a higher risk of death.
Please cite this article in press as: Wan BA, et al., Treatment Outcomes in Male Breast Cancer: A Retrospective Analysis of 161 Patients, Clinical Oncology (2018), https://doi.org/10.1016/j.clon.2018.02.026
B.A. Wan et al. / Clinical Oncology xxx (2018) 1e12
5
Fig 1. KaplaneMeier overall survival curves. (A) Overall survival for all patients, with dashed lines indicating 95% confidence intervals. (B) Metastatic (solid) versus non-metastatic (dashed) disease at diagnosis.
Multivariable analysis is presented in Table 4. Two significant factors remained as independent predictors of overall survival: invasive tumour size (hazard ratio 2.0; 95% confidence interval 1.4e2.7, P < 0.0001) and presence of metastatic disease at diagnosis (hazard ratio 8.7; 95% confidence interval 1.9e40.6; P ¼ 0.006). Figure 1B shows the KaplaneMeier overall survival curves comparing stage M0 versus M1 disease at diagnosis. The actuarial median overall survival was 19.9 years (95% confidence interval 12.8e25.4) or 2.6 years (95% confidence interval 1.0e5.5) in patients with M0 or M1 disease, respectively.
Competing Risks Analysis in Non-metastatic Patients For the 138 non-metastatic patients at diagnosis, 11 had local-regional recurrence and 26 had the competing event of death, with the cumulative incidence being 5.0% at 5 years and 11.2% at 10 years. Figure 2A shows the cumulative incidence of local-regional recurrence with the 95% confidence intervals (dashed). Table A2 shows the univariate Cox proportional subdistribution hazards model of local-regional recurrence, which identified that an earlier year of diagnosis, stage 3 disease, higher Nottingham
Table 4 Multivariable models for overall survival, local-regional recurrence, metastases or any recurrence or metastases. Values in bold are statistically significant (P < 0.05) Multivariable model for overall survival Predictive factor
Coefficient
Standard error
P-value
Invasive tumour size (cm) M-stage (1 versus 0)
0.67694 2.16019
0.16846 0.78789
<0.0001 0.0061
Standard error
P-value
Multivariable model for local-regional recurrence Predictive factor Coefficient
Year of diagnosis (continuous) Total positive nodes (log)
e0.12568 0.77912
0.04769 0.30167
0.0084 0.0098
Multivariable model for distant metastases Predictive factor
Coefficient
Standard error
P-value
Total positive nodes (log) Palliative chemotherapy (yes versus no) Invasive tumour size (cm)
0.63437 2.23426 0.60104
0.28371 0.78341 0.24057
0.0254 0.0043 0.0125
Multivariable model for any recurrence or metastases Predictive factor Coefficient
Standard error
P-value
Total positive nodes (log) Neoadjuvant chemotherapy (yes versus no) Palliative chemotherapy (yes versus no) Invasive tumour size (cm)
0.2319 0.69498 0.69773 0.18145
0.005 0.0451 0.012 0.0141
0.65068 1.39239 1.75181 0.44522
Hazard ratio (95% confidence interval) 1.968 1.414 2.738 8.673 1.851 40.626 Subdistribution hazard ratio and 95% confidence interval 0.882 0.803 0.968 2.18 1.207 3.937 Subdistribution hazard ratio and 95% confidence interval 1.886 1.081 3.288 9.34 2.011 43.368 1.824 1.138 2.923 Subdistribution hazard ratio and 95% confidence interval 1.917 1.217 3.02 4.024 1.031 15.713 5.765 1.469 22.631 1.561 1.094 2.227
R2 16.40%
R2 (%)
8.14%
R2 (%)
25.36%
R2 (%)
25.70%
Please cite this article in press as: Wan BA, et al., Treatment Outcomes in Male Breast Cancer: A Retrospective Analysis of 161 Patients, Clinical Oncology (2018), https://doi.org/10.1016/j.clon.2018.02.026
6
B.A. Wan et al. / Clinical Oncology xxx (2018) 1e12
Fig 2. Competing risk analysis in non-metastatic patients. (A) Cumulative incidence of local/regional recurrence. (B) Cumulative incidence of distant metastases. (C) Cumulative incidence of any recurrence or distance metastases.
grade, more total positive nodes removed or radiotherapy treatment had a higher risk of local-regional recurrence. In multivariable analysis, two variables remained significant (Table 4). Later year of diagnosis was associated with a reduced risk (hazard ratio 0.9, 95% confidence interval 0.8e1.0, P ¼ 0.008) and total number of positive lymph nodes was associated with a higher risk of local-regional recurrence (hazard ratio 2.2, 95% confidence interval 1.2e4.0, P ¼ 0.01). Among 138 patients diagnosed with non-metastatic disease, 26 developed distant metastases and 20 died. The cumulative incidence of distant metastases is shown in Figure 2B and was 17.7% at 5 years and 26.1% at 10 years. Univariate analysis identified that significant predictors of distant metastases included higher T or N stage, presence of other malignancies, larger invasive tumour size, greater number of dissected lymph nodes and positive lymph nodes, HER2/neu overexpression, luminal B-like tumour and treatment with chemotherapy or radiotherapy (Table A3). In particular, nodal staging was highly significant, with the risk of metastases highest for N3, followed by N2, N1 and N0. In multivariable analysis, three variables remained significantly associated with outcome (Table 4). More total positive nodes (hazard ratio 1.9; 95% confidence interval 1.1e3.3; P ¼ 0.03) or larger invasive tumour size (hazard ratio 1.8; 95% confidence interval 1.1e2.9; P ¼ 0.01) were associated with a higher risk of developing distant metastasis. When the presence of any local-regional recurrence or distant metastases was evaluated as one variable, the
cumulative incidence was 19.9% at 5 years and 33.9% at 10 years. The cumulative incidence of any recurrence or metastases is shown in Figure 2C. Significant predictive factors with the largest hazard ratios included stage N3 versus N0 (hazard ratio 12.2; 95% confidence interval 4.2e35.7; P < 0.0001) and stage T4 versus T1 (hazard ratio 5.5; 95% confidence interval 1.6e18.5; P ¼ 0.006) (Table A4). In the multivariable model (Table 4), a higher risk of recurrence/metastases was associated with the number of total positive nodes (hazard ratio 1.9; 95% confidence interval 1.2e3.0; P ¼ 0.005), invasive tumour size (hazard ratio 1.6; 95% confidence interval 1.1e2.2; P ¼ 0.01) or with palliative or neoadjuvant chemotherapy (hazard ratio 5.8; 95% confidence interval 1.5e22.6; P ¼ 0.01 or hazard ratio 4.0; 95% confidence interval 1.0e15.7; P ¼ 0.045).
Discussion Our study identified that the size of the invasive tumour, diagnosis of metastatic disease, year of diagnosis, number of positive lymph nodes and receipt of neoadjuvant or palliative chemotherapy were associated with a poorer prognosis. Patient characteristics were not identified as independent predictors of outcome. Previous studies have identified demographic predictors of male breast cancer outcome. Goss et al. [12] and Adami et al. [13] both found that men diagnosed at younger ages
Please cite this article in press as: Wan BA, et al., Treatment Outcomes in Male Breast Cancer: A Retrospective Analysis of 161 Patients, Clinical Oncology (2018), https://doi.org/10.1016/j.clon.2018.02.026
B.A. Wan et al. / Clinical Oncology xxx (2018) 1e12
had survival advantages when compared with older men. Results from our univariate analysis suggested that age was a significant factor predictive of overall survival, but not of recurrence or metastases. Goss et al. [12] also found that the median age of diagnosis was similar in male patients with and without a family history, whereas females with a family history presented with breast cancer at a younger age. Similarly, our study supported their findings, with the average age of diagnosis at 67 years for men with and without a family history of cancer. Epidemiological studies have identified that male carriers of the BRCA1 and BRCA2 mutations have higher risks of breast cancer and prostate cancer [4]. Our study found that the incidence of BRCA1 and BRCA2 mutations among patients who underwent genetic testing was consistent with published frequencies (8.8% and 23.5%, respectively). In a study by Hemminki et al. [14] of secondary primary malignancies in 3409 male breast cancer patients, 12.5% developed a second neoplasia; similarly, Masci et al. [15] found this incidence to be 17.6% in 97 patients. Our study found that the incidence of more than one primary carcinoma was high among BRCA1 and BRCA2 carriers. Among the three BRCA1 carriers, two had prostate cancer; among eight BRCA2 carriers, one had prostate cancer. Moreover, the overall incidence of second neoplasia was 33.5%, which is also higher than that found by Hemminki et al. [14]. Despite the high incidence of BRCA1/2 mutations, only one third of patients underwent genetic counselling, with even fewer opting to undergo genetic testing. Especially with the development of poly ADP ribose polymerase inhibitors targeted at BRCA1/2 breast cancers, genetic testing will have greater implications for male breast cancer patients and therefore should become an integral part of disease management [16]. A review of retrospective studies suggested that the 5 year local-regional recurrence rates in men who received radiotherapy for breast cancer ranged from 3 to 20% [17]. The result from our series was similar, with a 5 year localregional recurrence of 5.0%. During our study period there was a change in practice with regards to the use of radiotherapy to the chest wall, with its use associated with more recent treatment, which may be reflected in our finding that a more recent diagnosis was associated with a reduced risk. Studies have found that treatment options greatly influenced outcome in male breast cancer. Tamoxifen has been suggested to be effective in the management of male breast cancer, with both improved DFS and overall survival [2]. In a 2013 study by Goss et al. [12], breast-conserving surgery was associated with a higher recurrence rate compared with mastectomy. However, our study has not identified the use of tamoxifen or type of surgery to be significantly associated with outcome. This is further complicated by the fact that breast cancer management has changed with the development of modern systemic treatments, particularly within the past 10 years, which falls
7
within our study period. Moreover, our findings may be underpowered due to few non-censored patients with those treatments. Several recent studies have attempted to evaluate the histopathological characteristics of male breast cancer in order to stratify male breast cancer into prognostic subgroups [7,18]. Vermeulen et al. [18] conducted a central pathology review of 1483 male breast cancer cases and identified that histological grade was not significantly correlated with clinical outcome, unlike with female breast cancer patients. Similarly, our study found that histological grade was not an independent predictor of outcome. There are several limitations to this study. The first is that it was a retrospective study conducted at a single institution. Other limitations include the small sample size of metastatic patients and patients with recurrence, and incomprehensive data about tumour pathology, which limits the conclusions that can be drawn from this population. In addition, considering the wide range from 0 to 25 years of observation time, we had a short overall follow-up of 5.3 years. The strengths of our study include the comprehensive nature of assessed patient and disease characteristics, and treatment and disease outcomes. In addition, our study had one of the largest study populations from single-institution retrospective studies on male breast cancer [4]. Although the prognosis of male breast cancer has improved, there are still gaps in our knowledge in areas of treatment, prognosis and understanding of the biological mechanism. As male breast cancer treatment is still being modelled after treatment in females, validation of the efficacy of treatment options is required, particularly with regards to metastatic disease. These include the use of aromatase inhibitors and the efficacy of trastuzumab for HER2 enriched male breast cancers [4]. On the other hand, several recent collaborative international efforts have enabled large-scale characterisation of male breast cancer incidences, histopathological characteristics and outcomes [1,18]. Results from such initiatives may enable a better understanding of the aetiology and biology of male breast cancer, which can further inform treatment. Our results contribute to the current body of knowledge surrounding male breast cancer. In particular, this large retrospective series suggests that demographic factors have limited prognostic capacity, while total positive lymph nodes and invasive tumour size are independent predictors of local-regional recurrence and metastases.
Acknowledgements We thank the generous support of Bratty Family Fund, Michael and Karyn Goldstein Cancer Research Fund, Joey and Mary Furfari Cancer Research Fund, Pulenzas Cancer Research Fund, Joseph and Silvana Melara Cancer Research Fund and Ofelia Cancer Research Fund.
Please cite this article in press as: Wan BA, et al., Treatment Outcomes in Male Breast Cancer: A Retrospective Analysis of 161 Patients, Clinical Oncology (2018), https://doi.org/10.1016/j.clon.2018.02.026
8
B.A. Wan et al. / Clinical Oncology xxx (2018) 1e12
Appendix Table A1 Univariate analysis for factors predictive of overall survival. Values in bold are statistically significant (P < 0.05) Variable
Demographics Age at diagnosis (years) Year of diagnosis (continuous) Year of diagnosis categories (overall effect) 2010 to 2017 versus <2000 2000 to <2010 versus <2000 Gynaecomastia (presence versus absence) Family history (yes versus no) Other malignancy (yes versus no) T-stage (overall effect) 2 versus 1 3 versus 1 4 versus 1 T-stage (3/4 versus 1/2) N-stage (overall effect) 1 versus 0 2 versus 0 3 versus 0 2 versus 1 3 versus 1 3 versus 2 M-stage (1 versus 0) BRCA status (overall effect) BRCA1 versus negative BRCA2 versus negative Surgery type (mastectomy versus lumpectomy) Pathological characteristics Invasive tumour size (cm) Closest margin (log) Histological type (overall effect) DCIS versus other IDC/NOS versus other Inflammatory versus other Nottingham grade Skin involvement (yes versus no) Muscle involvement (yes versus no) LVI (yes versus no) Total positive nodes (log) Total removed nodes (log) ER/PRþ (yes versus no) Individual ERþ (yes versus no) Individual PRþ (yes versus no) HER2þ (yes versus no) Presence of DCIS in tumour (yes versus no) Cancer subtype (luminal A-like versus luminal B-like) Treatment received Use of chemotherapy (yes versus no) Number of chemotherapy regimens Adjuvant chemotherapy (yes versus no) Neoadjuvant chemotherapy (yes versus no) Palliative chemotherapy (yes versus no) Use of hormonal therapy in patients with ER/PRþ (yes versus no) Use of trastuzumab in patients with HER2þ (yes versus no)
R2
Independent covariate P-value
Hazard ratio
95% confidence interval of hazard ratio
0.0037 0.027 0.0695 0.0519 0.0244 0.7542 0.9558 0.4528 0.0589 0.0542 0.365 0.0082 0.1024 0.0282 0.1068 0.0708 0.0037 0.5319 0.202 0.756 <0.0001 0.5022 0.4526 0.2914 0.405
1.044 1.066
1.014 1.007
1.074 1.128
2.897 2.714 1.125 1.023 0.798
0.991 1.138 0.538 0.464 0.442
8.464 6.472 2.354 2.252 1.439
2.187 2.28 4.068 2.005
0.986 0.383 1.438 0.87
4.85 13.566 11.504 4.62
2.157 3.255 3.988 1.509 1.849 1.225 11.612
0.848 0.905 1.568 0.415 0.719 0.34 5.176
5.49 11.711 10.147 5.482 4.754 4.417 26.05
4.011 2.438 3.356
0.107 0.466 0.194
150.443 12.768 58.003
5.05% 3.22% 4.14%
0.06% 0.00% 0.36% 5.97%
1.76% 7.89
15.02% 4.85%
0.77%
0.0001 0.0948 0.4054 0.653 0.8048 0.1904 0.5674 0.1411 0.2889 0.009 0.0015 0.0056 0.9079 0.8742 0.3124 0.0954 0.4956 0.0686
1.834 2.218
1.343 0.871
2.505 5.646
11.03% 2.69% 1.56%
0.49 1.179 3.324 1.181 1.862 2.015 2.986 1.727 2.719 0.889 0.851 0.614 2.17 1.542 0.435
0.022 0.32 0.551 0.668 0.814 0.552 1.315 1.232 1.341 0.122 0.116 0.238 0.873 0.444 0.177
10.981 4.338 20.066 2.088 4.259 7.351 6.78 2.421 5.517 6.502 6.252 1.582 5.392 5.362 1.066
0.28% 2.12% 0.62% 6.60% 7.60% 7.75% 0.01% 0.02% 0.71% 2.27% 0.66% 2.87%
0.0827 0.0036 0.2885 0.2344 0.0001 0.3093
1.729 1.417 0.685 2.052 3.549 1.712
0.932 1.121 0.34 0.627 1.853 0.607
3.209 1.791 1.378 6.715 6.796 4.823
2.05% 4.29% 0.75% 0.72% 7.12% 0.84%
0.1683
0.299
0.054
1.666
12.02%
Please cite this article in press as: Wan BA, et al., Treatment Outcomes in Male Breast Cancer: A Retrospective Analysis of 161 Patients, Clinical Oncology (2018), https://doi.org/10.1016/j.clon.2018.02.026
B.A. Wan et al. / Clinical Oncology xxx (2018) 1e12
9
Table A1 (continued ) Variable
Breast/chest wall radiation (yes versus no) Nodes/axilla radiation (yes versus no) Boost radiation (yes versus no) Palliative radiation (yes versus no)
R2
Independent covariate P-value
Hazard ratio
95% confidence interval of hazard ratio
0.0091 0.1976 0.8748 <0.0001
2.371 1.511 1.121 3.909
1.24 0.806 0.271 2.126
4.535 2.832 4.638 7.19
4.82% 1.06% 0.02% 9.92%
DCIS: Ductal carcinoma in situ; IDC/NOS: Invasive ductal carcinoma/Not otherwise specified; LVI: Lymphovascular invasion; ER: Estrogen receptor; PR: Progesterone receptor.
Table A2 Univariate analysis for factors predictive of local-regional recurrence. Values in bold are statistically significant (P < 0.05) Variable
Demographics Age at diagnosis (years) Year of diagnosis (continuous) Year of diagnosis categories (2000 versus <2000) Gynaecomastia (presence versus absence) Family history (yes versus no) Other malignancy (yes versus no) T-stage (overall effect) 2 versus 1 3 versus 1 4 versus 1 T-stage (3/4 versus 1/2) N-stage (overall effect) 1 versus 0 2 versus 0 3 versus 0 Pathological characteristics Invasive tumour size (cm) Closest margin (log) Historical type (IDC/NOS versus other) Nottingham grade Skin involvement (yes versus no) Muscle involvement (yes versus no) LVI (yes versus no) Total positive nodes (log) Total removed nodes (log) Individual PRþ (yes versus no) HER2þ (yes versus no) Presence of DCIS in tumour (yes versus no) Cancer subtype (luminal A-like versus luminal B-like) Treatment received Use of chemotherapy (yes versus no) Number of chemotherapy regimens Adjuvant chemotherapy (yes versus no) Neoadjuvant chemotherapy (yes versus no) Palliative chemotherapy (yes versus no) Use of hormonal therapy in patients with ER/PRþ (yes versus no) Breast/chest wall radiation (yes versus no) Nodes/axilla radiation (yes versus no) Boost radiation (yes versus no) Palliative radiation (yes versus no)
R2
Independent covariate P-value
Subdistribution hazard ratio with 95% confidence interval
0.1007 0.003 0.2871 0.9329 0.7034 0.7574 0.5255 0.2161 0.1764 0.4273 0.46 0.1363 0.0632 0.2012 0.0195
1.051 0.9 0.524 1.064 1.367 0.823
0.99 0.839 0.159 0.251 0.273 0.24
1.115 0.965 1.723 4.505 6.845 2.826
2.692 4.807 2.363 1.681
0.561 0.494 0.283 0.424
12.923 46.815 19.757 6.666
7.78 5.943 14.116
0.893 0.386 1.530
67.786 91.381 130.257
0.2909 0.087 0.7724 0.0483 0.4688 0.2528 0.3892 0.009 0.0907 0.6179 0.7634 0.81 0.7432
1.189 0.126 0.761 3.187 1.695 2.514 1.816 2.096 2.183 0.581 1.389 0.762 0.701
0.862 0.012 0.119 1.009 0.407 0.518 0.467 1.202 0.884 0.069 0.164 0.083 0.083
1.64 1.351 4.857 10.067 7.062 12.2 7.065 3.653 5.395 4.917 11.79 6.989 5.888
0.38% 2.66% 0.05% 3.26% 0.50% 0.44% 0.64% 4.44% 1.87% 0.20% 0.09% 0.07% 0.11%
0.0553 0.0581 0.1177 0.4929 0.8677 0.4898 0.0317 0.0078 0.03 0.4455
3.651 1.389 2.565 2.067 1.195 2.021 8.907 8.268 5.794 1.814
0.971 0.989 0.788 0.259 0.147 0.274 1.212 1.746 1.185 0.393
13.727 1.95 8.351 16.462 9.711 14.901 65.477 39.16 28.328 8.372
3.20% 1.22% 1.72% 0.28% 0.02% 0.44% 5.54% 6.67% 2.49% 0.37%
1.96% 3.30% 0.68% 0.00% 0.16% 0.07% 1.88%
0.32% 6.57%
IDC/NOS: Invasive ductal carcinoma/Not otherwise specified; LVI: Lymphovascular invasion; ER: Estrogen receptor; PR: Progesterone receptor; HER2: Human epidermal growth factor receptor 2; DCIS: Ductal carcinoma in situ. Please cite this article in press as: Wan BA, et al., Treatment Outcomes in Male Breast Cancer: A Retrospective Analysis of 161 Patients, Clinical Oncology (2018), https://doi.org/10.1016/j.clon.2018.02.026
10
B.A. Wan et al. / Clinical Oncology xxx (2018) 1e12
Table A3 Univariate analysis for factors predictive of distant metastases. Values in bold are statistically significant (P < 0.05) Variable
Demographics Age at diagnosis (years) Year of diagnosis (continuous) Year of diagnosis categories (overall effect) 2010 to 2017 versus <2000 2000 to <2010 versus <2000 Gynaecomastia (presence versus absence) Family history (yes versus no) Other malignancy (yes versus no) T-stage (overall effect) 2 versus 1 3 versus 1 4 versus 1 T-stage (3/4 versus 1/2) N-stage (overall effect) 1 versus 0 2 versus 0 3 versus 0 2 versus 1 3 versus 1 3 versus 2 BRCA status (BRCA2 versus negative) Pathological characteristics Invasive tumour size (cm) Closest margin (log) Historical type (IDC/NOS versus other) IDC/NOS versus other Inflammatory versus other Nottingham grade Skin involvement (yes versus no) Muscle involvement (yes versus no) LVI (yes versus no) Total positive nodes (log) Total removed nodes (log) ER/PRþ (yes versus no) Individual ERþ (yes versus no) Individual PRþ (yes versus no) HER2D (yes versus no) Presence of DCIS in tumour (yes versus no) Cancer subtype (luminal B-like versus luminal A-like) Treatment received Use of chemotherapy (yes versus no) Number of chemotherapy regimens Adjuvant chemotherapy (yes versus no) Neoadjuvant chemotherapy (yes versus no) Palliative chemotherapy (yes versus no) Use of hormonal therapy in patients with ER/PRþ (yes versus no) Use of trastuzumab in patients with HER2þ (yes versus no) Breast/chest wall radiation (yes versus no) Nodes/axilla radiation (yes versus no) Boost radiation (yes versus no) Palliative radiation (yes versus no)
R2
Independent covariate P-value
Subdistribution hazard ratio with 95% confidence interval
0.3323 0.0791 0.1843 0.0663 0.1308 0.9396 0.3249 0.0277 0.1089 0.0481 0.4457 0.018 0.1601 <0.0001 0.6086 0.045 <0.0001 0.1543 0.0013 0.1822 0.463
0.981 1.053
0.942 0.994
1.02 1.116
4.193 3.104 0.96 0.639 0.322
0.908 0.714 0.335 0.262 0.118
19.356 13.489 2.749 1.559 0.883
3.081 2.345 5.612 2.049
1.009 0.262 1.344 0.753
9.403 20.962 23.431 5.574
1.467 4.835 12.527 3.296 8.54 2.591 1.856
0.338 1.036 3.863 0.639 2.311 0.64 0.356
6.358 22.571 40.618 17.012 31.555 10.491 9.685
0.0007 0.6418 0.6754 0.8243 0.52 0.9223 0.359 0.2766 0.0859 <0.0001 0.0157 0.5086 0.5043 0.4396 0.0238 0.9905 0.0208
1.688 1.257
1.249 0.479
2.281 3.299
0.853 2.43 0.973 1.552 2.525 2.327 2.883 3.226 0.505 0.495 2.099 3.06 0.992 3.133
0.209 0.162 0.559 0.607 0.476 0.888 1.84 1.247 0.067 0.063 0.32 1.16 0.276 1.189
3.475 36.359 1.693 3.969 13.397 6.1 4.515 8.348 3.823 3.898 13.765 8.07 3.563 8.251
0.01% 0.84% 0.90% 2.91% 15.78% 7.42% 0.29% 0.34% 0.58% 4.05% 0.01% 4.40%
0.0002 <0.0001 0.0021 0.008 <0.0001 0.2388 0.7969 0.0011 0.0785 0.993 <0.0001
6.614 2.129 3.383 3.698 12.071 2.231 0.795 10.535 2.04 0.992 19.809
2.468 1.705 1.555 1.408 6.13 0.587 0.139 2.548 0.922 0.164 9.568
17.729 2.657 7.36 9.716 23.769 8.485 4.556 43.555 4.515 6.004 41.008
13.70% 16.25% 6.59% 3.01% 16.21% 1.19% 0.40% 13.35% 2.24% 0.01% 28.45%
0.76% 1.69% 3.10%
0.00% 0.94% 4.45% 5.87%
1.45% 15.73%
1.44% 8.27% 0.18% 0.59%
IDC/NOS: Invasive ductal carcinoma/Not otherwise specified; LVI: Lymphovascular invasion; ER: Estrogen receptor; PR: Progesterone receptor; HER2: Human epidermal growth factor receptor 2; DCIS: Ductal carcinoma in situ.
Please cite this article in press as: Wan BA, et al., Treatment Outcomes in Male Breast Cancer: A Retrospective Analysis of 161 Patients, Clinical Oncology (2018), https://doi.org/10.1016/j.clon.2018.02.026
B.A. Wan et al. / Clinical Oncology xxx (2018) 1e12
11
Table A4 Univariate analysis for factors predictive of any recurrence/metastases. Values in bold are statistically significant (P < 0.05) Variable
Demographics Age at diagnosis (years) Year of diagnosis (continuous) Year of diagnosis categories (overall effect) 2010 to 2017 versus <2000 2000 to <2010 versus <2000 Gynaecomastia (presence versus absence) Family history (yes versus no) Other malignancy (yes versus no) T-stage (overall effect) 2 versus 1 3 versus 1 4 versus 1 3 versus 2 4 versus 2 4 versus 3 T-stage (3/4 versus 1/2) N-stage (overall effect) 1 versus 0 2 versus 0 3 versus 0 2 versus 1 3 versus 1 3 versus 2 BRCA status (BRCA2 versus negative) Pathological characteristics Invasive tumour size (cm) Closest margin (log) Historical type (IDC/NOS versus other) IDC/NOS versus other Inflammatory versus other Nottingham grade Skin involvement (yes versus no) Muscle involvement (yes versus no) LVI (yes versus no) Total positive nodes (log) Total removed nodes (log) ER/PRþ (yes versus no) Individual ERþ (yes versus no) Individual PRþ (yes versus no) HER2þ (yes versus no) Presence of DCIS in tumour (yes versus no) Cancer subtype (luminal B-like versus luminal A-like) Treatment received Use of chemotherapy (yes versus no) Number of chemotherapy regimens Adjuvant chemotherapy (yes versus no) Neoadjuvant chemotherapy (yes versus no) Palliative chemotherapy (yes versus no) Use of hormonal therapy in ER/PRþ patients (yes versus no) Use of trastuzumab in patients with HER2þ (yes versus no) Breast/chest wall radiation (yes versus no) Nodes/axilla radiation (yes versus no) Boost radiation (yes versus no) Palliative radiation (yes versus no)
R2
Independent covariate P-value
Subdistribution hazard ratio with 95% confidence interval
0.7448 0.8658 0.765 0.4828 0.5386 0.7687 0.4458 0.0306 0.035 0.0128 0.073 0.0062 0.803 0.3703 0.73 0.054 <0.0001 0.0692 0.0119 <0.0001 0.2937 0.0034 0.1898 0.463
0.994 1.004
0.961 0.954
1.029 1.057
1.44 1.321 1.135 0.726 0.407
0.52 0.544 0.488 0.319 0.18
3.984 3.205 2.637 1.653 0.919
3.513 4.169 5.47 1.187 1.557 1.312 2.185
1.306 0.875 1.618 0.309 0.591 0.281 0.987
9.453 19.855 18.489 4.553 4.101 6.133 4.839
2.823 5.452 12.204 1.931 4.323 2.238 1.856
0.922 1.454 4.171 0.565 1.622 0.671 0.356
8.648 20.448 35.712 6.596 11.522 7.466 9.685
0.0004 0.6274 0.7497 0.5703 0.794 0.1173 0.3152 0.0643 0.0315 <0.0001 0.0078 0.7992 0.7697 0.9125 0.061 0.8443 0.0525
1.573 0.779
1.226 0.285
2.02 2.132
0.741 1.407 1.551 1.519 2.918 2.48 2.604 2.588 0.759 0.728 1.083 2.415 0.898 2.486
0.264 0.109 0.896 0.672 0.938 1.084 1.83 1.285 0.091 0.087 0.261 0.96 0.308 0.99
2.084 18.24 2.687 3.437 9.076 5.674 3.704 5.211 6.358 6.09 4.497 6.074 2.619 6.24
1.58% 0.98% 1.72% 4.41% 17.47% 7.26% 0.05% 0.08% 0.01% 2.72% 0.05% 2.95%
<0.0001 <0.0001 0.0002 0.0333 <0.0001 0.203 0.7969 <0.0001 0.0011 0.1038 <0.0001
5.363 2.042 3.636 3.1 8.338 2.02 0.795 9.982 3.24 2.552 11.923
2.402 1.655 1.828 1.093 4.469 0.684 0.139 3.136 1.604 0.825 6.46
11.971 2.519 7.232 8.788 15.56 5.965 4.556 31.775 6.547 7.894 22.008
14.91% 17.29% 9.45% 2.40% 13.26% 1.31% 0.40% 17.52% 7.76% 1.36% 23.63%
0.08% 0.02% 0.37%
0.06% 0.59% 3.98% 8.38%
2.34% 17.16%
1.44% 7.84% 0.24% 48.61%
IDC/NOS: Invasive ductal carcinoma/Not otherwise specified; LVI: Lymphovascular invasion; ER: Estrogen receptor; PR: Progesterone receptor; HER2: Human epidermal growth factor receptor 2; DCIS: Ductal carcinoma in situ.
Please cite this article in press as: Wan BA, et al., Treatment Outcomes in Male Breast Cancer: A Retrospective Analysis of 161 Patients, Clinical Oncology (2018), https://doi.org/10.1016/j.clon.2018.02.026
12
B.A. Wan et al. / Clinical Oncology xxx (2018) 1e12
References [1] Miao H, Verkooijen HM, Chia K, Bouchardy C, Pukkala E, Larønningen S, et al. Incidence and outcome of male breast cancer: an international population-based study. J Clin Oncol 2017;29:4381e4386. https://doi.org/10.1200/JCO.2011.36.8902. [2] Ruddy KJ, Winer EP. Male breast cancer: risk factors, biology, diagnosis, treatment, and survivorship. Ann Oncol 2013;24: 1434e1443. https://doi.org/10.1093/annonc/mdt025. [3] Fentiman IS. Male breast cancer is not congruent with the female disease. Crit Rev Oncol Hematol 2016;101:119e124. https://doi.org/10.1016/j.critrevonc.2016.02.017. [4] Losurdo A, Rota S, Gullo G, Masci G, Torrisi R, Bottai G, et al. Controversies in clinicopathological characteristics and treatment strategies of male breast cancer: a review of the literature. Crit Rev Oncol Hematol 2017;113:283e291. https:// doi.org/10.1016/j.critrevonc.2017.03.013. [5] Hill TD, Khamis HJ, Tyczynski JE, Berkel HJ. Comparison of male and female breast cancer incidence trends, tumor characteristics, and survival. Ann Epidemiol 2005;15:773e780. https://doi.org/10.1016/j.annepidem.2005.01.001. [6] Giordano SH, Cohen DS, Buzdar AU, Perkins G, Hortobagyi GN. Breast carcinoma in men: a population-based study. Cancer 2004;101:51e57. https://doi.org/10.1002/cncr.20312. [7] Abreu MH, Afonso N, Abreu PH, Menezes F, Lopes P, Henrique R, et al. Male breast cancer: looking for better prognostic subgroups. Breast 2016;26:18e24. https://doi.org/ 10.1016/j.breast.2015.12.001. [8] Patten DK, Sharifi LK, Fazel M. New approaches in the management of male breast cancer. Clin Breast Cancer 2013;13: 309e314. https://doi.org/10.1016/j.clbc.2013.04.003. [9] Korde LA, Zujewski JA, Kamin L, Giordano S, Domchek S, Anderson WF, et al. Multidisciplinary meeting on male breast cancer: summary and research recommendations. J Clin Oncol 2010;28:2114e2122. https://doi.org/10.1200/JCO.2009.25.5729.
[10] Ge Y, Sneige N, Eltorky MA, Wang Z, Lin E, Gong Y, et al. Immunohistochemical characterization of subtypes of male breast carcinoma. Breast Cancer Res 2009;11:R28. https://doi. org/10.1186/bcr2258. [11] Fine JP, Gray RJ. A proportional hazards model for the subdistribution of a competing risk. J Am Stat Assoc 1999;94: 496e509. [12] Goss PE, Reid C, Pintilie M, Lim R, Miller N. Male breast carcinoma: a review of 229 patients who presented to the Princess Margaret Hospital during 40 years: 1955e1996. Cancer 1999;85:629e639. https://doi.org/10.1002/(SICI)10970142(19990201)85:3<629::AID-CNCR13>3.0.CO;2-V. [13] Adami HO, Malker B, Holmberg L, Persson I, Stone B. The relation between survival and age at diagnosis in breast cancer. N Engl J Med 1986;315:559e563. lo G, Boffetta P, Mellemjkaer L, Tracey E, [14] Hemminki K, Sce Andersen A, et al. Second primary malignancies in patients with male breast cancer. Br J Cancer 2005;92:1288e1292. https://doi.org/10.1038/sj.bjc.6602505. [15] Masci G, Caruso M, Caruso F, Salvini P, Carnaghi C, Giordano L, et al. Clinicopathological and immunohistochemical characteristics in male breast cancer: a retrospective case series. Oncologist 2015;20:586e592. [16] Dziadkowiec KN, Ga˛ siorowska E, Nowak-Markwitz E, Jankowska A. PARP inhibitors: review of mechanisms of action and BRCA1/2 mutation targeting. Menopausal Rev 2016;4: 215e219. https://doi.org/10.5114/pm.2016.65667. [17] Fentiman I, Fourquet A, Hortobagyi G. Male breast cancer. Lancet 2006;367:595e604. [18] Vermeulen MA, Slaets L, Cardoso F, Giordano SH, Tryfonidis K, van Diest PJ, et al. Pathological characterisation of male breast cancer: results of the EORTC 10085/TBCRC/BIG/NABCG International Male Breast Cancer Program. Eur J Cancer 2017;82: 219e227. https://doi.org/10.1016/j.ejca.2017.01.034.
Please cite this article in press as: Wan BA, et al., Treatment Outcomes in Male Breast Cancer: A Retrospective Analysis of 161 Patients, Clinical Oncology (2018), https://doi.org/10.1016/j.clon.2018.02.026