Skip to main content
NCBI home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2020 Oct 1.
Published in final edited form as: Endocr Relat Cancer. 2019 Oct;26(10):779–794. doi: 10.1530/ERC-19-0278

The molecular genetic make-up of male breast cancer

Cathy B Moelans 1, Joep de Ligt 3, Petra van der Groep 1,2, Pjotr Prins 3, Nicolle JM Besselink 3,12, Marlous Hoogstraat 3, Natalie D ter Hoeve 1, Miangela M Lacle 1, Robert Kornegoor 5, Carmen C van der Pol 4, Wendy WJ de Leng 1, Ellis Barbé 6, Bert van der Vegt 7, John Martens 8, Peter Bult 9, Vincent THBM Smit 10, Marco J Koudijs 3,12, Isaac J Nijman 3,12, Emile E Voest 11,12, Pier Selenica 13, Britta Weigelt 13, Jorge S Reis-Filho 13, Elsken van der Wall 4, Edwin Cuppen 3,14, Paul J van Diest 1
PMCID: PMC6938562  NIHMSID: NIHMS1536645  PMID: 31340200

Abstract

Male breast cancer (MBC) is extremely rare and accounts for less than 1% of all breast malignancies. Therefore, clinical management of MBC is currently guided by research on the disease in females. In this study, DNA obtained from 45 formalin-fixed paraffin-embedded (FFPE) MBCs with and 90 MBCs (52 FFPE and 38 fresh-frozen) without matched normal tissues was subjected to massively parallel sequencing targeting all exons of 1943 cancer-related genes. The landscape of mutations and copy number alterations was compared to that of publicly available estrogen receptor (ER)-positive female breast cancers (smFBCs) and correlated to prognosis. From the 135 MBCs, 90% showed ductal histology, 96% were ER-positive, 66% were progesterone receptor (PR)-positive, and 2% HER2-positive, resulting in 50%, 46% and 4% luminal A-like, luminal B-like and basal-like cases, respectively. Five patients had Klinefelter syndrome (4%) and 11% of patients harbored pathogenic BRCA2 germline mutations. The genomic landscape of MBC to some extent recapitulated that of smFBC, with recurrent PIK3CA (36%) and GATA3 (15%) somatic mutations, and with 40% of the most frequently amplified genes overlapping between both sexes. TP53 (3%) somatic mutations were significantly less frequent in MBC compared to smFBC whereas somatic mutations in genes regulating chromatin function and homologous recombination deficiency-related signatures were more prevalent. MDM2 amplifications were frequent (13%), correlated with protein overexpression (p=0.001) and predicted poor outcome (p=0.007). In conclusion, despite similarities in the genomic landscape between MBC and smFBC, MBC is a molecularly unique and heterogeneous disease requiring its own clinical trials and treatment guidelines.

Keywords: breast cancer, male, mutation, copy number, amplification, genomic

INTRODUCTION

Male breast cancer (MBC) is a rare disease accounting for about 1% of all breast cancers (Jemal et al., 2010, Giordano, 2018). Five-year overall survival is lower compared to female patients, related to the older age at diagnosis and more advanced stage at presentation (Stang and Thomssen, 2008, Giordano et al., 2004). Major genetic factors associated with an increased risk of MBC include BRCA2 and BRCA1 germline mutations, Klinefelter syndrome (KS), and family history. Additional suspected genetic factors include CHEK2 and PALB2 mutations, with less supporting evidence for AR and FANCM gene mutations, CYP17 polymorphism and Cowden syndrome(Weiss et al., 2005, Silvestri et al., 2017, Silvestri et al., 2018). Epidemiologic risk factors include hormonal imbalance and radiation exposure (Thomas et al., 1994, Abdelwahab Yousef, 2017). The majority of MBCs are invasive ductal carcinomas with high estrogen (ER) and progesterone (PR) receptor levels without overexpression of HER2 (Kornegoor et al., 2012c).

Because of its low incidence, MBC has not been studied as extensively as female breast cancer (FBC). Few retrospective studies have included more than 100 cases. Therefore, despite possible differences in pathogenesis, biology and genetics between both sexes, treatment strategies for MBC have largely been extrapolated from FBC. This, in combination with the older age at diagnosis, hampers improvement in outcome as has been seen FBC over the last decades.

Recent large scale cross-platform projects have provided a detailed characterization of the genomic landscape of FBC (Nik-Zainal et al., 2016, Bailey et al., 2018, Berger et al., 2018, Cancer Genome Atlas Network, 2012, Pereira et al., 2016, Curtis et al., 2012). The Cancer Genome Atlas (TCGA)(Cancer Genome Atlas Network, 2012) reported the enrichment of specific mutations in PIK3CA (47%), GATA3 (14%), MAP3K1 (13%), TP53 (12%), CDH1 (10%) and MAP2K4 (7%) with the luminal A subtype, and in TP53 (31%), PIK3CA (32%) and MAP3K1 (5%) within the luminal B subtype (Gao et al., 2015). Integrated genomic/transcriptomic analysis of breast cancers from METABRIC (Molecular Taxonomy of Breast Cancer International Consortium) identified 40 putative mutation-driver genes (Curtis et al., 2012, Pereira et al., 2016). Subsequent studies identified overlapping as well as new probable driver mutations (Bailey et al., 2018, Berger et al., 2018, Nik-Zainal et al., 2016).

Genomic data on MBC are scarce. On a molecular level, like FBC, MBC is a heterogeneous disease with differences in transcriptional (Johansson et al., 2012, Callari et al., 2011), copy number (Johansson et al., 2011, Tommasi et al., 2011, Biesma et al., 2015, Piscuoglio et al., 2016), microRNA (Fassan et al., 2009, Lehmann et al., 2010) and methylation (Johansson et al., 2015, Kornegoor et al., 2012a, Rizzolo et al., 2018) patterns. Genomic gains appear to be more common in MBC than in FBC and often involve whole chromosome arms, while genomic losses and high-level amplifications are less frequent (Johansson et al., 2011, Lacle et al., 2015, Kornegoor et al., 2012b, Lacle et al., 2013, Biesma et al., 2015, Tommasi et al., 2011). Luminal A-like and luminal B-like MBC subtypes seem to exhibit remarkably similar copy number aberration profiles, in contrast to their distinct landscape of somatic mutations (Piscuoglio et al., 2016). Piscuoglio et al recently studied mutations of 59 MBCs by massively parallel sequencing the exons of 241 genes, and demonstrated that MBCs less frequently harbored PIK3CA (20%) and TP53 mutations (8.5%) than subtype-matched (sm)FBC (Piscuoglio et al., 2016). The most frequently mutated genes in luminal A-like MBC included PIK3CA and MAP3K1 (both at 12%) whereas the most frequently mutated genes in luminal B-like MBC were PIK3CA (24%) and GATA3 (21%).

Thus, MBC share many features with FBC, but there is evidence for distinct differences with potential implications for clinical management. Here we i) used targeted capture massively parallel sequencing of 1943 cancer-related genes to study the mutational and copy number landscape in a cohort of 135 MBCs, including 5 patients with KS, and ii) compared these profiles to those of smFBCs.

MATERIALS AND METHODS

Patients and samples

MBC samples were collected from 10 pathology laboratories throughout The Netherlands. Patients were diagnosed between 1986 and 2011. Pathology reports were used to extract age, tumor size, and lymph node status. Tumor slides were reviewed by three expert breast pathologists (PJvD, RK, MML) to confirm diagnosis and demarcate normal and tumor areas. Histological type (WHO) and grade were assessed as described previously (van Diest et al., 1992, Elston and Ellis, 1991). Patient and tumor characteristics are summarized in Table 1. All samples and databases were anonymized before use. The Dutch national ethical guidelines (www.federa.org) state that no ethical approval is required for the use of anonymous leftover tissue, and this is also part of the standard treatment agreement(van Diest, 2002).

Table 1.

Clinico-pathologic features and immunohistochemical subtypes of the 135 male breast cancers included in this study

Clinico-pathologic features N=135 % actual %
Median age at diagnosis 67 (range 32-89)
Median mitotic activity index 10 (range 0-56)
Median tumor size (cm) 2,3 (range 0.8-11)
Type
Ductal 121 89,6% 90,3%
Lobular 2 1,5% 1,5%
Other 11 8,1% 8,2%
Missing 1 0,7%
Histologic Grade*
1 27 20,0% 22,7%
2 49 36,3% 41,2%
3 43 31,9% 36,1%
Missing 16 11,9%
Mitotic activity index/2 mm2
< 8 mitoses 50 37,0% 40,7%
8-14 mitoses 35 25,9% 28,5%
>14 mitoses 38 28,1% 30,9%
Missing 12 8,9%
ER status
Negative 6 4,4% 4,5%
Positive 128 94,8% 95,5%
Missing 1 0,7%
PR status
Negative 45 33,3% 33,8%
Positive 88 65,2% 66,2%
Missing 2 1,5%
HER2 status
Negative 129 95,6% 97,7%
Positive 3 2,2% 2,3%
Missing 3 2,2%
LN status
Negative 48 35,6% 41,4%
Positive 68 50,4% 58,6%
Missing 19 14,1%
Ki-67 labeling index
Low (<14%) 94 69,6% 71,8%
High (≥14%) 37 27,4% 28,2%
Missing 4 3,0%
Surrogate intrinsic subtype**
Luminal A-like 66 48,9% 50,4%
Luminal B-like 60 44,4% 45,8%
HER2-driven 0 0,0% 0,0%
Basal-like 5 3,7% 3,8%
Missing 4 3,0%
Open in a new tab

ER: estrogen receptor; PR: progesterone receptor; HER2: human epidermal growth factor receptor 2; LN: lymph node

*

according to the Nottingham grading system (ref Elston-Ellis 1991)

**

defined by immunohistochemistry according to St. Gallen criteria (Goldhirsch et al, 2013)

Immunohistochemistry was repeated to classify MBCs into the four different intrinsic subtypes according to the St. Gallen International criteria for FBC (Goldhirsch et al., 2013). Luminal A-like tumors were defined as ER-positive (1%) and PR-positive (≥20% positive cells), HER2-negative and Ki67 low (<14% positive cells), whereas the remaining ER-positive tumors were classified as Luminal B-like. HER2 driven tumors were defined as ER/PR negative and HER2 overexpressed or amplified (ASCO/CAP guidelines (Wolff et al., 2013)), and Basal-like tumors as ER/PR/HER2 negative. The DNA extraction procedure is detailed in Supplementary Methods. All tumor samples consisted of 60% or more neoplastic cells, as assessed by pathology review.

Targeted capture massively parallel sequencing

Sequencing libraries were prepared from 600 ng of sheared DNA as previously described with a few modifications (Harakalova et al., 2011). After end repair samples were A-tailed, followed by T-tailed adaptor ligation. Samples were pooled and subjected to hybrid capture enrichment using a custom SureSelect assay (Agilent) containing baits targeting all exons of 1943 genes including known oncogenes, tumor suppressor genes, all kinases and genes involved in important pathways related to tumorigenesis and anti-cancer treatment (e.g., angiogenesis; apoptosis; and EGFR, PIK3CA, TGFβ, mTOR, and VEGF pathways were included) (Supplementary Table 1). Pre-barcoded individual libraries were pooled, enriched, and subsequently sequenced on the SOLiD 5500 system (Thermo Fisher). Supplementary Table 2 summarizes sequencing statistics. All sequencing data have been deposited in EGA under accession number EGAS00001002683.

SOLiD data were processed with our in-house developed pipeline v1.2.1 (https://github.com/UMCUGenetics/IAP) by Genome Analysis Toolkit (GATK) v3.2.2 calling for all samples according to best practices guidelines, including somatic mutation analysis for matched pairs (McKenna et al., 2010, Van der Auwera et al., 2013). Filtering was applied to select high quality variants (coverage>=10, allele fraction>=0.15) and to exclude germline variants (population frequency <5%), full details available in Supplementary Methods.

Copy number alteration (CNA) analysis was performed using Codecz v1.0.1 with the appropriate design file kinome_design_SS_V2_110811. For FFPE samples all (normal) controls were merged into a control pool against which the tumors were analyzed. For FF samples, each sample was compared to the complete FF tumor data set as a means of normalization. Z-scores were calculated as described by Hoogstraat et al. (Hoogstraat et al., 2015).

Mutational signature analysis was performed on MBCs with available ER and HER2 status (n=129) and matched 1:2 with FBCs from TCGA with the same receptor status (n=258). Mutational signatures were defined by deconstructSigs using all SNVs for samples with ≥20 somatic SNVs, as previously described (Mueller et al., 2018, Rosenthal et al., 2016).

In addition, targeted Next Generation Sequencing (NGS) of the complete coding sequences of BRCA1 and BRCA2 was performed on normal tissue (germline) using the Oncomine BRCA FFPE gene panel (ThermoFisher) according to the manufacturer’s instructions. In short, 20 ng FFPE DNA was used for library preparation with 24 PCR cycles. Emulsion PCR and enrichment were performed using the Ion PGM Template OT2 200 Template Kit and the Ion One Touch 2 instrument (ThermoFisher). Sequencing was performed using the Ion PGM Sequencing 200 kit v2 using the Ion 318 chip (ThermoFisher). Samples were run on the Ion Torrent PGM System (ThermoFisher). Sequencing data were analyzed using SeqNext (JSI medical systems).

Sanger sequencing

A selection of NGS variants was validated by Sanger sequencing. After Exonuclease I - Shrimp Alkaline Phosphatase enzymatic cleanup of amplified PCR products, BigDye Terminator v1.1 Cycle Sequencing (ThermoFisher) of 1 μl PCR product, and Sephadex G-50 based sequencing reaction cleanup, sequencing products were analyzed on a 3730 DNA Analyzer (ThermoFisher). Primers are listed in Supplementary Table 3. Sanger sequencing validation is described in the Supplementary Results section.

MDM2 immunohistochemistry

Tissue microarray (triplicate cores) or full 4-μm slides were deparaffinized, pretreated for 24 minutes in CC1 (EDTA pH8.0) and stained on a Ventana Benchmark XT autostainer using a mouse monoclonal anti-MDM2 antibody (ThermoFisher Scientific, clone IF2, dilution 1/25). All samples with sufficient tumor material (n=106) were scored as positive (nuclear staining in at least 5% positive tumor cells) or negative by consensus of two observers.

Statistics

Associations between gene copy number or mutation status and clinicopathological features were calculated using the Pearson Chi-square test (or Fisher’s exact test when appropriate) for categorical variables (IBM SPSS Statistics version 21). Associations were evaluated for all genes and only significant associations were reported. Mitotic activity index (MAI) was dichotomized as ≤14 or >14 mitoses/2 mm2, tumor size as ≤2 or >2 cm, age as ≤60 or >60, and histological grade as grade 1/2 versus grade 3(Elston and Ellis, 1991). Mutation load and CNA load were dichotomized at the median (7 and 62, respectively). Associations of mutational load with clinicopathological characteristics are described in the Supplementary Results section. Cluster analysis for CNAs was performed in R using hclust (default) and heatmap.2. Five-year overall survival (OS) was analyzed with Kaplan-Meier plots/log-rank test. Backward LR Cox proportional hazards models were used to estimate risk of death adjusted for parameters significant or with trend in univariate analyses. Hazard ratios (HR) were calculated with 95% confidence intervals (CI). For multivariable analysis, p-values >0.1 were removed from the model. MedCalc statistical software was used for comparison of SNV and CNA proportions to publicly available METABRIC and male breast cancer data. Corrections for multiple comparisons were applied according to Holm-Bonferroni. For pathway analysis, PANTHER version 11.1 was used (released 2016-10-24)(Mi et al., 2017).

RESULTS

Characteristics of the cohort

From the 135 MBCs, 90% showed ductal histology, 96% were ER-positive, 66% were PR-positive, and 2% HER2-positive, resulting in 50%, 46% and 4% luminal A-like, luminal B-like and basal-like cases, respectively. None of the cases was HER2-driven (Table 1). Five patients had Klinefelter syndrome (KS; 4%) and 11% (5/44 with paired normal tissue) of patients harbored pathogenic BRCA2 germline mutations. The median age at diagnosis was 67 years. Follow-up was available for 111/135 patients. Median follow-up was 52 months (range 1-243), and 59 patients were alive at the time of last follow-up. Tumor size and lymph node status were significant predictors of poor 5-year OS (p=0.037 and p=0.019, respectively). Luminal B-like (p=0.122), PR negative (p=0.128) and poorly differentiated (p=0.094) tumors tended to be associated with worse 5-year OS compared to Luminal A-like, PR-positive and well differentiated tumors, respectively. Luminal B-like MBCs were associated with high grade (p=0.00004), high mitotic activity index (p=0.018) and PR-negativity (p<0.00001) when compared to luminal A-like MBCs.

The majority of BRCA2 germline mutation carriers presented with poorly differentiated tumors (4/5 were high-grade; p=0.149) at a similar age as non-BRCA2-mutation carriers (median 71 (range 44-78) vs. 69 (range 32-87) years, respectively). KS patients tended to be younger (3 of 5 were aged ≤60; p=0.120), with generally well-differentiated (5/5; p=0.158), PR-positive tumors (5/5; p=0.166).

Somatic mutations

Somatic mutations are listed in Supplementary Table 4. Overlapping FBC mutation-driver genes from 4 studies (Bailey et al., 2018, Berger et al., 2018, Nik-Zainal et al., 2016, Pereira et al., 2016, Curtis et al., 2012) were specifically compared to MBC. Of the 89 overlapping genes, 21 were not targeted and 1 (HRAS) was insufficiently covered by our panel, resulting in 67 interrogated genes. The remaining genes were mutated in 0-36% of MBC. Table 2 compares somatic mutation frequencies with FBC and MBC (Piscuoglio et al., 2016). Somatic mutations in TP53 were rare in MBC (3%). In fact we detected only four (2 in matched normal, 4 in the entire cohort) somatic mutations in TP53, two of them known from FBC and being likely oncogenic (OncoKB (Chakravarty et al., 2017)), the other two unknown. None of the hotspots from FBC (R175, R248 and R273) were affected.

Table 2.

Comparison of somatic mutation frequency with female breast cancer and male breast cancer

Gene Studies where driver N MBC (n=135) METABRIC ER+ FBC (n=1825)* p-value TCGA ER+ FBC (n=600)*** p-value Berger BRCA (n=982)*** p-value Bailey BRCA (N=779) p-value Piscuoglio et al. (n=59) p-value
PIK3CA 1,2,3,4 108 36,1% 44,5% 0,0876 39,8% 0,4688 34,7% 0,7720 34,4% 0,7280 20,3% 0,0346
KMT2C 1,3,4 122 20,5% 12,3% 0,0087 8,8% 0,0001 NA NA 9,5% 0,0003 5,1% 0,0074
PBRM1 1,2,3 105 20,0% 1,5% <0,0001 0,7% <0,0001 NA NA NA NA NA NA
GATA3 1,2,3,4 132 15,2% 14,4% 0,8008 14,3% 0,7903 12,8% 0,4429 12,6% 0,4113 15,3% 0,9859
NSD1 3,4 126 15,1% NA NA 1,0% <0,0001 NA NA NA NA NA NA
NOTCH1 2,3 27 14,8% 4,4% 0,0101 0,3% <0,0001 NA NA NA NA NA NA
SETD2 2,3 125 13,6% 2,4% <0,0001 1,7% <0,0001 NA NA NA NA NA NA
NF1 1,2,3,4 99 13,1% 3,5% <0,0001 2,3% <0,0001 4,6% 0,0004 3,9% 0,0001 0,0% 0,0038
ARID1A 1,2,3,4 123 13,0% 5,1% 0,0002 4,7% 0,0005 4,3% <0,0001 3,3% <0,0001 1,7% 0,0145
ATRX 2,3 72 12,5% NA NA 1,8% <0,0001 NA NA NA NA 0,0% 0,0051
APC 2,3 108 12,0% 1,9% <0,0001 1,2% <0,0001 NA NA NA NA 1,7% 0,0219
NOTCH2 2,3 116 11,2% NA NA 2,3% <0,0001 NA NA NA NA NA NA
NCOR1 1,2,3 120 10,8% 4,0% 0,0004 5,0% 0,0144 NA NA 4,7% 0,0065 0,0% 0,0090
CREBBP 2,3 103 10,7% NA NA 1,2% <0,0001 NA NA NA NA NA NA
KDM6A 1,2,3 51 9,8% 1,7% <0,0001 1,7% 0,0002 NA NA NA NA NA NA
ATM 2,3,4 85 9,4% NA NA 2,3% 0,0005 3,3% 0,0047 NA NA 0.0% 0,0157
USP9X 1,2,3 77 9,1% 2,6% 0,0008 1,0% <0,0001 NA NA NA NA NA NA
ATR 2,3 99 8,1% 3,9% 0,0405 1,0% <0,0001 NA NA NA NA 1,7% 0,0946
AKT1 1,2,3,4 101 7,9% 4,7% 0,1505 3,2% 0,0237 3,0% 0,0103 3,2% 0,0192 0,0% 0,0272
KMT2D 2,3 115 7,8% 6,8% 0,6808 1,8% 0,0003 NA NA NA NA 3,4% 0,2598
BRAF 2,3 65 7,7% 0,6% <0,0001 0,5% <0,0001 NA NA NA NA 3,4% 0,3024
MED12 3,4 92 7,6% NA NA 2,2% 0,0042 NA NA NA NA 3,4% 0,2890
EGFR 2,3 108 6,5% 1,4% 0,0001 0,7% <0,0001 NA NA NA NA 1,7% 0,1669
STAG2 2,3 55 5,5% NA NA 0,8% 0,0022 NA NA NA NA NA NA
TAF1 3,4 75 5,3% 1,5% 0,0114 1,7% 0,0405 2,8% 0,2197 NA NA NA NA
RASA1 3,4 77 5,2% NA NA 0,3% <0,0001 0,8% 0,0004 NA NA NA NA
BRCA2 1,2,3 78 5,1% 1,5% 0,0145 2,2% 0,1255 NA NA NA NA 0,0% 0,0794
MAP3K1 1,2,3,4 123 4,9% 11,5% 0,0241 10,7% 0,0480 9,1% 0,1181 8,3% 0,1925 8,5% 0,3423
SMAD4 1,2,3 105 4,8% 1,2% 0,0022 1,0% 0,0038 NA NA NA NA NA NA
CIC 2,3 89 4,5% NA NA 0,8% 0,0038 NA NA NA NA NA NA
MSH2 2,3 92 4,3% NA NA 1,3% 0,0383 NA NA NA NA 0,0% 0,1077
BRCA1 1,2,3,4 117 4,3% 1,2% 0,0055 1,8% 0,0925 3,0% 0,4460 3,1% 0,4956 1,7% 0,3722
ERBB3 2,3,4 117 4,3% 2,6% 0,2718 2,2% 0,1871 NA NA NA NA 1,7% 0,3722
SMARCA4 2,3 74 4,1% NA NA 0,8% 0,0125 NA NA NA NA NA NA
CASP8 2,3,4 112 3,6% 1,1% 0,0206 1,3% 0,0807 2,1% 0,3111 2,1% 0,3212 NA NA
CDH1 1,2,3,4 118 3,4% 11,4% 0,0069 16,8% 0,0002 13,3% 0,0019 13,1% 0,0024 1,7% 0,5217
TP53 1,2,3,4 125 3,2% 20,6% <0,0001 20,8% <0,0001 34,7% <0,0001 33,8% <0,0001 8,5% 0,1211
MEN1 1,2,3 102 2,9% 1,1% 0,1036 0,8% 0,0609 NA NA NA NA NA NA
PMS2 2,3 120 2,5% NA NA 0,7% 0,0706 NA NA NA NA 0,0% 0,2219
RB1 1,2,3,4 44 2,3% 1,7% 0,7620 1,5% 0,6791 2,8% 0,8436 2,7% 0,8731 0,0% 0,2440
AXIN1 2,3 91 2,2% NA NA 0,5% 0,0749 NA NA NA NA NA NA
EP300 3,4 92 2,2% 2,6% 0,8134 1,0% 0,3171 2,2% 1,0000 NA NA NA NA
GNAS 2,3 96 2,1% NA NA 1,2% 0,4741 NA NA NA NA NA NA
PTEN 1,2,3,4 96 2,1% 4,0% 0,3490 5,3% 0,1761 6,1% 0,1081 5,8% 0,1302 1,7% 0,8615
CHEK2 1,3 53 1,9% 0,8% 0,3888 0,8% 0,4139 NA NA NA NA 0,0% 0,2897
TBL1XR1 1,3 115 1,7% 0,7% 0,2310 1,7% 1,0000 NA NA NA NA NA NA
MLH1 2,3 117 1,7% NA NA 1,0% 0,5096 NA NA NA NA 1,7% 1,0000
FBXW7 2,3,4 118 1,7% 0,9% 0,3920 0,8% 0,3567 1,9% 0,8799 1,8% 0,9391 NA NA
KRAS 1,2,3,4 118 1,7% 0,5% 0,0942 0,8% 0,3567 0,8% 0,3273 0,9% 0,4172 0,0% 0,3152
ASXL1 2,3 121 1,7% 2,2% 0,7147 0,2% 0,0249 NA NA NA NA NA NA
TBX3 1,2,3,4 70 1,4% 5,8% 0,1174 4,0% 0,2776 3,6% 0,3303 3,1% 0,4217 1,7% 0,8907
BAP1 1,3 76 1,3% 1,0% 0,7994 0,7% 0,5726 NA NA NA NA NA NA
ERBB2 1,2,3 91 1,1% 2,6% 0,3739 2,2% 0,4910 NA NA 2,6% 0,2815 1,7% 0,7553
CTNNB1 3,4 104 1,0% NA NA 0,3% 0,2988 0,7% 0,7324 NA NA 0,0% 0,4424
RUNX1 1,2,3,4 112 0,9% 3,5% 0,1378 4,7% 0,0629 4,8% 0,0567 5,0% 0,0501 0,0% 0,4662
FGFR2 2,3 114 0,9% NA NA 1,3% 0,7233 NA NA NA NA 3,4% 0,2362
PIK3R1 1,2,3,4 117 0,9% 2,3% 0,3187 1,2% 0,7809 3,5% 0,1324 2,7% 0,2419 0,0% 0,4661
CDKN1B 1,2,3,4 117 0,9% 1,3% 0,7103 1,7% 0,5245 1,4% 0,6575 1,2% 0,7777 1,7% 0.6420
NRAS 2,3 127 0,8% 0,1% 0,0563 0,2% 0,2655 NA NA NA NA 0,0% 0,4920
CCND1 3,4 116 0,0% NA NA 0,2% 0,6300 0,3% 0,5549 NA NA NA NA
CDKN2A 1,2,3 119 0,0% 0,6% 0,3970 0,3% 0,5500 NA NA NA NA 1,7% 0,1549
ESR1 2,3 85 0,0% NA NA 0,8% 0,4083 NA NA NA NA 0,0% 1,0000
MAP2K4 1,2,3,4 91 0,0% 3,9% 4,7% 0,0348 4,4% 0,0412 4,0% 0,0522 1,7% 0,1549
NF2 2,3 102 0,0% 0,7% 0,3967 0,3% 0,5799 NA NA NA NA 0,0% 1,0000
PPP2R1A 3,4 115 0,0% NA NA 0,2% 0,1353 0,8% 0,3360 NA NA NA NA
RHOA 2,3 118 0,0% NA NA 0,8% 0,3300 NA NA NA NA NA NA
STK11 2,3 55 0,0% 0,8% 0,5056 0,0% 1,0000 NA NA NA NA NA NA
HRAS 2,3 unsufficient coverage 0,2% - 0,0% - NA NA NA NA 0,0% -
ARID5B 3,4 not targeted 1,3% - 0,3% - 1,0% - NA NA NA NA
B2M 3,4 not targeted NA NA 0,3% - 0,3% - NA NA NA NA
BCOR 2,3 not targeted NA NA 0,8% - NA NA NA NA NA NA
CBFB 1,2,3 not targeted 5,7% - 3,3% - NA NA 2,3% - 0,0% -
CHD4 3,4 not targeted NA NA 3,0% - 3,3% - 2,6% - 1,7% -
CTCF 1,2,3,4 not targeted 2,7% - 2,7% - 2,6% - 2,6% - 1,7% -
DNMT3A 2,3 not targeted NA NA 0,8% - NA NA NA NA NA NA
FOXA1 2,3 not targeted NA NA 4,0% - NA NA 3,5% - 1,7% -
FOXP1 1,2 not targeted 0,9% - 1,0% - NA NA NA NA NA NA
GPS2 1,3 not targeted 1,3% - 1,3% - NA NA 1,0% - NA NA
LATS1 3,4 not targeted NA NA 0,7% - 1,1% - NA NA NA NA
MLLT4 1,2 not targeted 2,6% - 1,8% - NA NA NA NA NA NA
NIPBL 3,4 not targeted NA NA 1,7% - 3,0% - NA NA NA NA
PHF6 2,3 not targeted NA NA 0,5% - NA NA NA NA NA NA
RNF43 3,4 not targeted NA NA 0,8% - 1,0% - NA NA NA NA
RPL22 3,4 not targeted NA NA 0,5% - 0,4% - NA NA NA NA
SF3B1 1,2,3 not targeted 3,5% - 2,3% - NA NA 2,1% - NA NA
SPOP 3,4 not targeted NA NA 0,5% - 0,9% - NA NA NA NA
TET2 2,3 not targeted NA NA 0,8% - NA NA NA NA NA NA
ZFHX3 3,4 not targeted NA NA 2,3% - NA NA NA NA 5,1% -
ZFP36L1 1,2,3 not targeted 1,0% - 1,2% - NA NA NA NA NA NA
PIK3CA 1,2,3,4 108 36,1% 44,5% 0,0876 39,8% 0,4688 34,7% 0,7720 34,4% 0,7280 20,3% 0,0346
Open in a new tab
*

METABRIC: Nature 2012 + Nature 2016; data accessed through cBioPortal dd March 15, 2019

**

TCGA: Cell 2015; data accessed through cBioPortal dd March 15, 2019

NA = not available

PIK3CA (36%), KMT2C (21%), PBRM1 (20%) and GATA3 (15%) were identified as most frequently mutated genes in this study (Figure 1). The pattern of somatic mutations found in PIK3CA resembled that of FBC: all mutations found in MBC were missense mutations and 75-80% (12/15 in matched normal; 30/40 in the entire cohort) affected the FBC hotspots (H1047, E542 and E545; Supplementary Figure 1). Besides PIK3CA, other recurrent mutations in MBC included a K358 frameshift GATA3 mutation (4/25 GATA3 mutations), the E17K AKT1 mutation (3/8 AKT1 mutations) and the Q26K NCOR1 mutation (4/21 NCOR1 mutations).

Figure 1.

Figure 1.

Open in a new tab

Somatic non-synonymous mutations and recurrent amplifications in 135 male breast cancers. Individual genes are represented as rows, and individual patients are represented as columns. Patients with pathogenic germline BRCA2 mutations or Klinefelter Syndrome are indicated, as well as tumor histology and the distribution of surrogate IHC subtype.

PIK3CA mutations were significantly more abundant in poorly differentiated (p=0.013; n=96) tumors with high mitotic activity (p=0.014; n=99), and in lymph node positive (p=0.006; n=94) and younger patients (p=0.046; n=108). Somatic mutations in PIK3CA were associated with worse 5-year OS (p=0.026; n=90; Supplementary Figure 2). In multivariable analysis, however, PIK3CA somatic mutations did not independently predict survival alongside tumor grade (p=0.034; HR 2.627) and size (p=0.001; HR 1.410).

ATM somatic mutations were associated with high tumor grade (p=0.040; n=74) and high mitotic counts (p=0.014; n=77). NOTCH2 mutations were more frequent in poorly differentiated tumors (p=0.017; n=100) without lymph node metastases (p=0.020; n=100). ATRX and CREBBP somatic mutations were more frequently observed in KS patients (p=0.001 (n=72) and p=0.030 (n=103), respectively). PTPRD mutations were more abundant in Ki67-high tumors (p=0.005; n=96) and MED12 mutations in luminal A-like tumors (p=0.026; n=86).

Compared to mutation frequencies in ER-positive FBC from METABRIC and TCGA, TP53 (both p<0.0001), MAP3K1 (p=0.02 and p=0.05, respectively), and CDH1 (p=0.007 and p=0.0002, respectively) somatic mutations were significantly less frequent in MBC (Table 2), whereas particularly PBRM1 (both p<0.0001), NSD1 (p<0.0001 TCGA) and SETD2 (both p<0.0001) mutations were more frequent. Overall, genes regulating chromatin function appeared more often affected in MBC (Figure 2).

Figure 2.

Figure 2.

Open in a new tab

Comparison of MBC somatic mutation frequencies to METABRIC and TCGA ER-positive FBC. Genes have been sorted per function and 95% confidence intervals (http://vassarstats.net/prop1.html) have been indicated. * indicates a significant difference with METABRIC and/or TCGA

Mutational signature analysis revealed that 24% of MBCs have a dominant signature 3 associated with defective homologous recombination DNA repair, whereas only 13% of TCGA FBCs demonstrated a dominant signature 3 (Figure 3). In addition, another 24% of MBCs have a dominant signature 8, as compared to 0.4% of FBCs. In a recent review by Nik-Zainal et al., this signature 8 was reported to be increased in homologous recombination deficiency and late in cancer evolution (Nik-Zainal and Morganella, 2017).

Figure 3.

Figure 3.

Open in a new tab

Mutational signature analysis in MBCs compared to ER/HER2- matched FBCs (TCGA). HRD = homologous recombination defect; MMR = mismatch repair.

Copy number aberrations

Copy number aberration (CNA) load

On average, each tumor harbored 138 gene CNAs (range 6-1230), including amplifications (Z>2.80) and homozygous deletions (Z<−2.80). We observed no correlation between CNA load and ER/PR status, tumor histology, lymph node status, Ki67 status or age (data not shown). CNA load was not associated with mutational load (p=0.286) or 5-year survival (p=0.549), and was similar between luminal A-like and luminal B-like tumors (p=0.285). CNA load was however higher in poorly differentiated (p=0.012) and larger tumors (p=0.045) with high mitotic activity index (p=0.011)(Supplementary Figure 4). CNA load was significantly higher in tumors harboring mutations in KDM6A (p=0.006) and NOTCH2 (p=0.028).

Figure 4 depicts a heatmap of all CNAs in the MBC cohort after unsupervised hierarchical cluster analysis. MBCs of the same surrogate intrinsic molecular subtype did not cluster on the basis of their CNA profile, whereas KS patients seemed to be subgrouped as a result of their CNA-naïve profile with chromosome X gains. In addition, a cluster of 17 mainly luminal A-like MBCs with high genomic instability was apparent. No correlation was observed with clinicopathological variables or survival when compared to the other MBCs but somatic mutation load tended to be slightly higher (p=0.075).

Figure 4.

Figure 4.

Open in a new tab

Heatmap of copy number alterations in the male breast cancers following unsupervised hierarchical cluster analysis. Two clusters seem to stand out: the top cluster with many events dominated by Luminal A-like tumors and the bottom cluster dominated by Luminal B-like cancers. Klinefelter Syndrome patients (grey) show few copy number alterations.

Amplifications

Analysis of CNAs revealed that MBCs harbored recurrent amplifications on 8p/8q, 16p, 20q and 1q. Supplementary Table 6 summarizes the associations of a set of 220 genes amplified in at least 10% of MBCs with intrinsic subtype, clinicopathological characteristics, SNV load and PIK3CA mutation status. Some of the most recurrently amplified genes in both cohorts (FF and FFPE) included UBR5 (24%; 8q), TSC2 (22%; 16p), ANK1 (22%; 8p), OBSCN (21%; 1q), and WHSC1L1 (21%; 8p). A comparison between amplification frequencies in MBC and smFBC is presented in the same table. Merely 86/220 (39%) frequently amplified genes in MBC were also amplified in more than 10% of ER-positive FBC. PANTHER reactome pathway analysis of these 220 genes demonstrated an overrepresentation of genes involved in extracellular matrix organization (3-fold enrichment; p=0.0004), chromatin organization (2-fold enrichment; p=0.0012) and STAT3 signaling (6-fold enrichment; p=0.046) whereas cell cycle related genes were relatively underrepresented (0.4-fold enrichment; p=0.0012)(Mi et al., 2013). Based on The Drug Gene Interaction Database, 19% of these genes (41/220) are clinically actionable and for 26% (57/220), drug interactions have been described (Supplementary Table 6).

Individually, 7/220 genes were indicative of poor 5-year OS when amplified: MDM2, PAK1, SCYL3, TGFB2, CLTC, SMYD3 and ASH1L (Supplementary Table 6). In multivariable analysis including clinical and pathological characteristics, the amplification status of PAK1, SCYL3, MDM2, ASH1L, E2F7, CLTC or TGFB2 remained in the model. In multivariable analysis including all 7 genes, PAK1 (HR 3.65; p=0.011), TGFB2 (HR 3.14; p=0.011) and E2F7 (HR 2.98; p=0.017) remained. When all 7 gene amplifications as well as PIK3CA mutations were included, only ASH1L (HR 2.44; p=0.057) and lymph node status (HR 2.44; p=0.087) remained in the model as independent predictors of survival (Supplementary Table 7).

Of particular interest was MDM2, amplified in as much as 13% of the cohort, and predictor of worse 5-year OS (p=0.007; Figure 5) even after correction for other established prognostic variables (p=0.024; Supplementary Table 7). MDM2 amplifications were more frequent in PR negative (p=0.004) and poorly differentiated (p=0.040) tumors (Supplementary Figure 5). Co-amplifications with COL1A1 (p=0.007), HER2 (p=0.009), DNAH11 (p=0.027), PIK3C2B (p=0.027) and PPM1D (p=0.043) were frequent. Tumors harboring PIK3CA mutations (p=0.005) or NF1 mutations (p=0.026) were more likely to harbor MDM2 amplifications (Supplementary Figure 5 and Figure 1). MDM2 amplifications were significantly correlated with MDM2 protein overexpression (p=0.001) but protein overexpression did not predict poor outcome (p=0,337).

Figure 5.

Figure 5.

Open in a new tab

MDM2 amplification in male breast cancer is associated with poor 5-year overall survival, PR negativity and high tumor grade.

Germline mutations

With on average 1723 reads per normal tissue sample, we detected pathogenic BRCA2 germline mutations in 5/44 (11%) matched normal samples using the IonTorrent Oncomine BRCA FFPE gene panel (library preparation failed for 1 sample; Supplementary Table 5). There was no indication for loss of heterozygosity of the wild-type BRCA2 allele based on variant allele frequencies. No pathogenic BRCA1 germline mutations were identified. Tumors from BRCA2 germline mutation carriers harbored a similar amount of somatic mutations and total CNAs compared to other MBCs. BRCA2-related tumors appear to have increased CNA counts on chromosome 8q (p=0,056) accompanied by decreased counts on 8p (p=0,004). All 5 BRCA2-carriers demonstrated UBR5 (8q; p=0,002), ASAP1 (8q; p=0,001), CSMD3 (8q; p=0,001) and CCNE2 (8q; p=0,0002) amplification, but DNAH11 (7p; p=0,003), RRM2B (8q; p=0,007), FZD6 (8q; p=0,010), RUNX1T1 (8q; p=0,023) and SGK3 (8q; p=0,023) amplifications were also more frequently seen. We observed no enrichment for particular somatic mutations. None of the BRCA2 germline mutation carriers harbored a PIK3CA or GATA3 mutation (Figure 1). Within AR, CHEK2 and PALB2 (FANCM was not targeted), we found no likely pathogenic germline mutations. Germline mutations in these genes predicted to have medium or high impact are described in the Supplementary Results section.

DISCUSSION

Compared to FBC, MBC has been relatively poorly characterized on the molecular level. Here we have characterized a cohort of 135 MBC, 46% of which were deemed luminal B-like tumors. This is consistent with a recent pathology review of the EORTC 10085/TBCRC/BIG/NABCG International Male Breast Cancer Program where 49.3% of 1328 MBC were luminal B-like tumors (Doebar et al., 2017). In our MBC series, 46% of patients received radiotherapy, 16% chemotherapy and 42% endocrine therapy. Compared to the EORTC cohort, fewer men received chemotherapy (EORTC 27.7%; 220/794 patients with available data) and endocrine therapy (EORTC 82.6%; 583/706 patients with available data). These discrepancies are likely mainly due to uncertainties in treatment decision-making. Studies investigating MBCs at the molecular level and correlating their findings with survival are therefore of the utmost importance.

TP53, CDH1, MAP3K1 and to a lesser extent PIK3CA, TBX3, MAP2K4, RUNX1 and PTEN somatic mutations appear to be less frequent in MBC compared to smFBC whereas PBRM1, NSD1 and SETD2 somatic mutations seem to be more frequent.

After correction for multiple comparisons, there were no significant differences between our study and the MBC study of Piscuoglio et al. Similar to their MBC study, somatic mutations in TP53 were very rare (Piscuoglio et al., 2016). Their TP53 mutation frequency was however higher (8.5%) than ours (3%), possibly due to a different cohort composition. In FBC, luminal B-like tumors show a higher TP53 mutation frequency than luminal A-like tumors and the MBC set of Piscuoglio et al. was likely enriched for luminal B-like cancers (71% of their cohort)(Cancer Genome Atlas Network, 2012, Piscuoglio et al., 2016). In addition, a study using a highly similar predecessor “Cancer mini-genome” gene panel and approach in triple negative FBC identified TP53 mutations in 55% of the specimens, thereby excluding technical bias(Lips et al., 2015). In contrast to TP53, the pattern of somatic mutations found in PIK3CA resembled that of FBC. Its somatic mutation frequency in our MBC cohort (36%) was in between that reported by Piscuoglio et al. (MBC; 20%) and ER-positive FBC (n=1431; 48%). PIK3CA mutations were associated with more aggressive tumor characteristics and reduced survival, in line with findings in ER-positive FBC subgroups (Pereira et al., 2016, Ishida et al., 2018).

Aside from PIK3CA, the serine/threonine protein kinase AKT1 was the only other mutated gene with hotspots corresponding to FBC. The majority of AKT1 missense mutations were E17K (4/9) or L52R (2/9) substitutions. Both mutations are known to be oncogenic and there is promising clinical data in FBC patients with AKT1 E17K mutant ER+ ductal breast cancer treated with the pan-AKT targeted inhibitor AZD5363 (level 3A evidence; oncoKB)(Davies et al., 2012, Addie et al., 2013, Davies et al., 2015, Hyman et al., 2017).

As expected from FBC, most GATA3 mutations were likely oncogenic frameshift mutations including K358fs, H435fs, A332fs, R330fs and M294K/R (Ellis et al., 2012, Cerami et al., 2012, Gao et al., 2013). In contrast to Piscuoglio et al., GATA3 somatic mutations were not exclusive to luminal B-like MBC tumors (Piscuoglio et al., 2016). Overall, genes regulating chromatin function appear more often mutated in MBC, and mutational signatures 3 and 8, both associated with homologous recombination deficiency, are more prevalent than in ER/HER2 matched FBC. Of particular clinical importance, HR deficiency may point to a selective sensitivity to PARP (poly-ADP ribose polymerase) inhibitors (Nik-Zainal and Morganella, 2017).

Like FBC, MBC is CNA driven, and despite many similarities in the CNA landscape between MBC and FBC (Curtis et al., 2012, Piscuoglio et al., 2016), only 40% of frequently amplified genes in MBC were also frequently amplified in FBC, underlining differences in biology and genetics. MDM2 amplifications were observed in 13% of the MBC cohort while in FBC, these amplifications occur in merely 3.7% of cases. Since MDM2 mediates ubiquitination of p53, its increased amplification frequency may be related to the decreased TP53 mutation frequency observed in MBC, the 2 events being mutually exclusive in the same pathway. Of note, besides MDM2, many of the aberrant genes in MBC seem to be involved in the inactivation of the p53 pathway (e.g. amplification of PAK1 and SMYD2, and inactivating mutations in ARID1A, PBRM1 and KMT2C)(Wu et al., 2014, Huang et al., 2006, Woo et al., 2017). MDM2 amplifications predicted more aggressive tumor behavior and correlated with protein overexpression. The high rate of MDM2 overexpression in prostate and ER-positive FBC, and the ability of MDM2 inhibitors to ubiquitinate steroid hormone receptors, has led to the evaluation of this class of drugs in combination with endocrine therapies (CRUKE/12/032). Many MDM2 small molecule inhibitors have only recently progressed from preclinical development into early clinical trials (such as RG7112, also known as RO5045337, and AMG-232)(Burgess et al., 2016). Thus, MDM2 is a putative therapeutic target in MBC.

Four genes independently predicted poor OS in MBC when amplified: PAK1 (11q13.5-q14.1), ASH1L (1q22), E2F7 (12q21.2) and TGFB2 (1q41). PAK1 (p21 activated kinase 1) amplifications were present at a similar frequency compared to ER-positive FBC (12% vs. 10%)(Radu et al., 2014, Ye and Field, 2012, Ong et al., 2011, Curtis et al., 2012) and amplification and overexpression of PAK1 is associated with poor outcome in luminal FBC as well. I In vitro testing of a small molecule inhibitor, FRAX1036, in combination with docetaxel, further supported PAK1 as a potential target in breast cancer (Ong et al., 2015). PAK1 has also been implicated in tamoxifen resistance (Holm et al., 2006, Bostner et al., 2007, Bostner et al., 2010), and in upregulating genes involved in the Fanconi Anemia (FA)/BRCA pathway, paving the way for combined PAK1 and PARP inhibition in FA/BRCA-proficient cancers (Villamar Cruz et al., 2016). ASH1L, encoding a histone methyltransferase (HMT), showed a similar amplification frequency in MBC compared to ER-positive FBC (18% vs. 21%). This HMT is classified as dysregulated by genetic alterations and as candidate therapeutic target in FBC (Liu et al., 2015). ASH1L inhibitors are currently being developed (Rogawski et al., 2015). Atypical E2F transcription factor 7 (E2F7) amplification appears to be more frequent in MBC than ER-positive FBC (13% vs 1.4%). E2F7 participates in various processes such as angiogenesis, polyploidization and DNA damage response. E2F7 overexpression leads to tamoxifen resistance in breast cancer cells (Chu et al., 2015), and anthracycline resistance in squamous cell carcinoma (Hazar-Rethinam et al., 2015). A small molecule pan-E2F inhibitor (HLM006474) has been developed and was proven to be effective in a melanoma and lung cancer cell culture model (Ma et al., 2008, Kurtyka et al., 2014). TGFB2 encodes a secreted ligand of the transforming growth factor-beta superfamily and its mRNA levels predict tamoxifen response in breast cancer cells (Buck et al., 2008). Neutralizing antibodies to TGFB2 have been shown to reverse tamoxifen resistance of human breast carcinomas in vivo (Arteaga et al., 1999). Several pharmacological strategies have been developed to block TGF-beta signalling (Colak and Ten Dijke, 2017). Overall, 19% of the most frequently amplified genes in MBC are clinically actionable and for 26%, drug interactions have been described, suggesting many potential druggable targets.

We detected pathogenic BRCA2 germline mutations in 11% of the paired tumor/normal subcohort (5/44). Other larger studies have described similar frequencies (8-14%)(Pritzlaff et al., 2017, Basham et al., 2002, Couch et al., 1996). In contrast to Silvestri et al., who reported a median age at diagnosis for male BRCA2 mutation carriers of 62 years and an inverse relationship between age at diagnosis and tumor grade in 375 male BRCA2 mutation carriers, our BRCA2 cohort consisted of older patients with poorly differentiated tumors (Silvestri et al., 2016). In our study, tumors from BRCA2 germline mutation carriers harbored a similar amount of somatic mutations and total CNAs compared to other MBCs, but appeared to have increased CNA counts on 8q and decreased counts on 8p. In light of the small BRCA2 mutation carrier cohort reported here, results based on this subgroup should however be interpreted with caution.

KS patients tended to be younger at presentation, with generally well-differentiated tumors with few CNAs. The increased risk of developing breast cancer for these KS patients can be attributed to a direct effect of the supernumerary X chromosome(s), or the combined action of abnormal chromosome dosage and hormonal imbalances (Kawakami et al., 2004)(Bonomi et al., 2017).

It should be noted that this study used FFPE as well as fresh frozen material. It is well recognized that the quality of FFPE samples due to fixation and tissue processing is inferior to that of frozen material. DNA extracted from FFPE tissues is fragmented which can lead to sequence artifacts (Do and Dobrovic, 2015). Sequence artifacts can be difficult to distinguish from true mutations, especially in the context of tumor heterogeneity, and are an increasing interpretive problem in this era of massively parallel sequencing. We therefore included as much fresh frozen material as possible. Nevertheless, in light of the rarity of male breast cancer, we have included archived FFPE material as well. This might have led to an overestimation of mutational load (Yost et al., 2012).

In conclusion, this comprehensive molecular analysis of MBC suggests that, while MBCs show remarkable similarities to FBC, clear differences are also found, subscribing the amounting evidence that MBC should be seen as an entity of its own. This should also be reflected in the clinical approach that should refrain from merely imitating FBC management. As we have observed some genetic alterations to be clearly more prevalent in MBC, e.g. MDM2 amplifications and homologous recombination deficient mutational signatures, focus should be on associated relevant drugs. As such, honoring MBC for its specific molecular portrait may result in improvements in outcome that parallel those we have observed over the last years in FBC.

Supplementary Material

01

Supplementary Table 1. List of 1943 genes included in the targeted capture massively parallel sequencing platform.

10

Supplementary Figure 3. Sanger sequencing confirmation of a selected set of SNVs in HER2 (ERBB2), GATA3, BRCA2 and PIK3CA. Positions and samples are indicated. Not all samples with SNVs at these positions are shown.

11

Supplementary Figure 4. Association of copy number alteration (CNA) load with clinical, pathological and genomic characteristics.

12

Supplementary Figure 5. Association of MDM2 amplifications with clinical, pathological and genomic characteristics.

13

Supplementary Methods. Detailed protocol for DNA isolation and mutation calling.

14

Supplementary Results. Association of mutational load with clinicopathological characteristics, Sanger sequencing validation, and germline mutations in AR, CHEK2 and PALB2 with medium to high predicted impact

02

Supplementary Table 2. Sequencing statistics (SOLID 5500).

03

Supplementary Table 3. Sanger sequencing variants and primer sequences.

04

Supplementary Table 4. List of somatic mutations.

05

Supplementary Table 5. BRCA2 germline mutations within the paired tumor/normal subcohort.

06

Supplementary Table 6. List of the 220 most frequently amplified genes in this study. The table includes their chromosomal location, the amplification frequency in ER-positive female breast cancer with associated p-value for difference in proportions, the preference for surrogate intrinsic molecular subtype, and associations with clinical, pathological and genetic characteristics. Potentially druggable gene categories, clinical actionability and known drug interactions are indicated per gene.

07

Supplementary Table 7. Kaplan Meier and multivariable Cox Regression 5-year overall survival (OS) analysis of 10 frequently amplified genes in MBC.

08

Supplementary Figure 1. Pattern of PIK3CA somatic mutations in male (n=107) versus female breast cancer (n=5284; cBioPortal)(Cerami et al., 2012, Gao et al., 2013).

09

Supplementary Figure 2. PIK3CA mutations predict worse 5-year overall survival in male breast cancer.

Acknowledgements

We are grateful to the Borstkanker Onderzoek Groep (BOOG; study nr 2009-04; NCT01101425) for providing us with clinical and biological data of MBC patients.

Funding

This study was funded by the Dutch Cancer Society (UU 2012-5637), and in part by the National Institutes of Health/National Cancer Institute MSK Cancer Center Support Grant P30CA008748. JSR-F is funded in part by the Breast Cancer Research Foundation. The funding agencies had no role in the design of the study, the collection, analysis, and interpretation of data, and in writing the manuscript.

Footnotes

Competing interests

The authors declare that they have no competing interests.

Ethics approval and consent to participate

The Dutch national ethical guidelines (www.federa.org) state that no ethical approval is required for the use of anonymous leftover tissue, and this is also part of the standard treatment agreement (van Diest, 2002).

Availability of data and materials

All sequencing data have been deposited in EGA under accession number EGAS00001002683. All data generated or analysed during this study are included in this published article and its supplementary information files.

REFERENCES

  1. Abdelwahab Yousef AJ 2017. Male Breast Cancer: Epidemiology and Risk Factors. Semin Oncol, 44, 267–272. [DOI] [PubMed] [Google Scholar]
  2. Addie M, Ballard P, Buttar D, Crafter C, Currie G, Davies BR, Debreczeni J, Dry H, Dudley P, Greenwood R, et al. 2013. Discovery of 4-amino-N-[(1S)-1-(4-chlorophenyl)-3-hydroxypropyl]-1-(7H-pyrrolo[2,3-d]pyrimidin −4-yl)piperidine-4-carboxamide (AZD5363), an orally bioavailable, potent inhibitor of Akt kinases. J Med Chem, 56, 2059–73. [DOI] [PubMed] [Google Scholar]
  3. Arteaga CL, Koli KM, Dugger TC & Clarke R 1999. Reversal of tamoxifen resistance of human breast carcinomas in vivo by neutralizing antibodies to transforming growth factor-beta. J Natl Cancer Inst, 91, 46–53. [DOI] [PubMed] [Google Scholar]
  4. Bailey MH, Tokheim C, Porta-Pardo E, Sengupta S, Bertrand D, Weerasinghe A, Colaprico A, Wendl MC, Kim J, Reardon B, et al. 2018. Comprehensive Characterization of Cancer Driver Genes and Mutations. Cell, 174, 1034–1035. [DOI] [PMC free article] [PubMed] [Google Scholar]
  5. Basham VM, Lipscombe JM, Ward JM, Gayther SA, Ponder BA, Easton DF & Pharoah PD 2002. BRCA1 and BRCA2 mutations in a population-based study of male breast cancer. Breast Cancer Res, 4, R2. [DOI] [PMC free article] [PubMed] [Google Scholar]
  6. Berger AC, Korkut A, Kanchi RS, Hegde AM, Lenoir W, Liu W, Liu Y, Fan H, Shen H, Ravikumar V, et al. 2018. A Comprehensive Pan-Cancer Molecular Study of Gynecologic and Breast Cancers. Cancer Cell, 33, 690–705 e9. [DOI] [PMC free article] [PubMed] [Google Scholar]
  7. Biesma HD, Schouten PC, Lacle MM, Sanders J, Brugman W, Kerkhoven R, Mandjes I, Van Der Groep P, Van Diest PJ & Linn SC 2015. Copy number profiling by array comparative genomic hybridization identifies frequently occurring BRCA2-like male breast cancer. Genes Chromosomes Cancer, 54, 734–44. [DOI] [PubMed] [Google Scholar]
  8. Bonomi M, Rochira V, Pasquali D, Balercia G, Jannini EA, Ferlin A & Klinefelter Italia NG 2017. Klinefelter syndrome (KS): genetics, clinical phenotype and hypogonadism. J Endocrinol Invest, 40, 123–134. [DOI] [PMC free article] [PubMed] [Google Scholar]
  9. Bostner J, Ahnstrom Waltersson M, Fornander T, Skoog L, Nordenskjold B & Stal O 2007. Amplification of CCND1 and PAK1 as predictors of recurrence and tamoxifen resistance in postmenopausal breast cancer. Oncogene, 26, 6997–7005. [DOI] [PubMed] [Google Scholar]
  10. Bostner J, Skoog L, Fornander T, Nordenskjold B & Stal O 2010. Estrogen receptor-alpha phosphorylation at serine 305, nuclear p21-activated kinase 1 expression, and response to tamoxifen in postmenopausal breast cancer. Clin Cancer Res, 16, 1624–33. [DOI] [PubMed] [Google Scholar]
  11. Buck MB, Coller JK, Murdter TE, Eichelbaum M & Knabbe C 2008. TGFbeta2 and TbetaRII are valid molecular biomarkers for the antiproliferative effects of tamoxifen and tamoxifen metabolites in breast cancer cells. Breast Cancer Res Treat, 107, 15–24. [DOI] [PubMed] [Google Scholar]
  12. Burgess A, Chia KM, Haupt S, Thomas D, Haupt Y & Lim E 2016. Clinical Overview of MDM2/X-Targeted Therapies. Front Oncol, 6, 7. [DOI] [PMC free article] [PubMed] [Google Scholar]
  13. Callari M, Cappelletti V, De Cecco L, Musella V, Miodini P, Veneroni S, Gariboldi M, Pierotti MA & Daidone MG 2011. Gene expression analysis reveals a different transcriptomic landscape in female and male breast cancer. Breast Cancer Res Treat, 127, 601–10. [DOI] [PubMed] [Google Scholar]
  14. Cancer Genome Atlas Network 2012. Comprehensive molecular portraits of human breast tumours. Nature, 490, 61–70. [DOI] [PMC free article] [PubMed] [Google Scholar]
  15. Cerami E, Gao J, Dogrusoz U, Gross BE, Sumer SO, Aksoy BA, Jacobsen A, Byrne CJ, Heuer ML, Larsson E, et al. 2012. The cBio cancer genomics portal: an open platform for exploring multidimensional cancer genomics data. Cancer Discov, 2, 401–4. [DOI] [PMC free article] [PubMed] [Google Scholar]
  16. Chakravarty D, Gao J, Phillips SM, Kundra R, Zhang H, Wang J, Rudolph JE, Yaeger R, Soumerai T, Nissan MH, et al. 2017. OncoKB: A Precision Oncology Knowledge Base. JCO Precis Oncol, 2017. [DOI] [PMC free article] [PubMed] [Google Scholar]
  17. Chu J, Zhu Y, Liu Y, Sun L, Lv X, Wu Y, Hu P, Su F, Gong C, Song E, et al. 2015. E2F7 overexpression leads to tamoxifen resistance in breast cancer cells by competing with E2F1 at miR-15a/16 promoter. Oncotarget, 6, 31944–57. [DOI] [PMC free article] [PubMed] [Google Scholar]
  18. Colak S & Ten Dijke P 2017. Targeting TGF-beta Signaling in Cancer. Trends Cancer, 3, 56–71. [DOI] [PubMed] [Google Scholar]
  19. Couch FJ, Farid LM, Deshano ML, Tavtigian SV, Calzone K, Campeau L, Peng Y, Bogden B, Chen Q, Neuhausen S, et al. 1996. BRCA2 germline mutations in male breast cancer cases and breast cancer families. Nat Genet, 13, 123–5. [DOI] [PubMed] [Google Scholar]
  20. Curtis C, Shah SP, Chin SF, Turashvili G, Rueda OM, Dunning MJ, Speed D, Lynch AG, Samarajiwa S, Yuan Y, et al. 2012. The genomic and transcriptomic architecture of 2,000 breast tumours reveals novel subgroups. Nature, 486, 346–52. [DOI] [PMC free article] [PubMed] [Google Scholar]
  21. Davies BR, Greenwood H, Dudley P, Crafter C, Yu DH, Zhang J, Li J, Gao B, Ji Q, Maynard J, et al. 2012. Preclinical pharmacology of AZD5363, an inhibitor of AKT: pharmacodynamics, antitumor activity, and correlation of monotherapy activity with genetic background. Mol Cancer Ther, 11, 873–87. [DOI] [PubMed] [Google Scholar]
  22. Davies BR, Guan N, Logie A, Crafter C, Hanson L, Jacobs V, James N, Dudley P, Jacques K, Ladd B, et al. 2015. Tumors with AKT1E17K Mutations Are Rational Targets for Single Agent or Combination Therapy with AKT Inhibitors. Mol Cancer Ther, 14, 2441–51. [DOI] [PubMed] [Google Scholar]
  23. Do H & Dobrovic A 2015. Sequence artifacts in DNA from formalin-fixed tissues: causes and strategies for minimization. Clin Chem, 61, 64–71. [DOI] [PubMed] [Google Scholar]
  24. Doebar SC, Slaets L, Cardoso F, Giordano SH, Bartlett JM, Tryfonidis K, Dijkstra NH, Schroder CP, Van Asperen CJ, Linderholm B, et al. 2017. Male breast cancer precursor lesions: analysis of the EORTC 10085/TBCRC/BIG/NABCG International Male Breast Cancer Program. Mod Pathol. [DOI] [PubMed] [Google Scholar]
  25. Ellis MJ, Ding L, Shen D, Luo J, Suman VJ, Wallis JW, Van Tine BA, Hoog J, Goiffon RJ, Goldstein TC, et al. 2012. Whole-genome analysis informs breast cancer response to aromatase inhibition. Nature, 486, 353–60. [DOI] [PMC free article] [PubMed] [Google Scholar]
  26. Elston CW & Ellis IO 1991. Pathological prognostic factors in breast cancer. I. The value of histological grade in breast cancer: experience from a large study with long-term follow-up. Histopathology, 19, 403–10. [DOI] [PubMed] [Google Scholar]
  27. Fassan M, Baffa R, Palazzo JP, Lloyd J, Crosariol M, Liu CG, Volinia S, Alder H, Rugge M, Croce CM, et al. 2009. MicroRNA expression profiling of male breast cancer. Breast Cancer Res, 11, R58. [DOI] [PMC free article] [PubMed] [Google Scholar]
  28. Gao J, Aksoy BA, Dogrusoz U, Dresdner G, Gross B, Sumer SO, Sun Y, Jacobsen A, Sinha R, Larsson E, et al. 2013. Integrative analysis of complex cancer genomics and clinical profiles using the cBioPortal. Sci Signal, 6, pl1. [DOI] [PMC free article] [PubMed] [Google Scholar]
  29. Gao Y, Jones A, Fasching PA, Ruebner M, Beckmann MW, Widschwendter M & Teschendorff AE 2015. The integrative epigenomic-transcriptomic landscape of ER positive breast cancer. Clin Epigenetics, 7, 126. [DOI] [PMC free article] [PubMed] [Google Scholar]
  30. Giordano SH 2018. Breast Cancer in Men. N Engl J Med, 378, 2311–2320. [DOI] [PubMed] [Google Scholar]
  31. Giordano SH, Cohen DS, Buzdar AU, Perkins G & Hortobagyi GN 2004. Breast carcinoma in men: a population-based study. Cancer, 101, 51–7. [DOI] [PubMed] [Google Scholar]
  32. Goldhirsch A, Winer EP, Coates AS, Gelber RD, Piccart-Gebhart M, Thurlimann B, Senn HJ & Panel M 2013. Personalizing the treatment of women with early breast cancer: highlights of the St Gallen International Expert Consensus on the Primary Therapy of Early Breast Cancer 2013. Ann Oncol, 24, 2206–23. [DOI] [PMC free article] [PubMed] [Google Scholar]
  33. Harakalova M, Mokry M, Hrdlickova B, Renkens I, Duran K, Van Roekel H, Lansu N, Van Roosmalen M, De Bruijn E, Nijman IJ, et al. 2011. Multiplexed array-based and in-solution genomic enrichment for flexible and cost-effective targeted next-generation sequencing. Nat Protoc, 6, 1870–86. [DOI] [PubMed] [Google Scholar]
  34. Hazar-Rethinam M, De Long LM, Gannon OM, Topkas E, Boros S, Vargas AC, Dzienis M, Mukhopadhyay P, Simpson F, Endo-Munoz L, et al. 2015. A novel E2F/sphingosine kinase 1 axis regulates anthracycline response in squamous cell carcinoma. Clin Cancer Res, 21, 417–27. [DOI] [PubMed] [Google Scholar]
  35. Holm C, Rayala S, Jirstrom K, Stal O, Kumar R & Landberg G 2006. Association between Pak1 expression and subcellular localization and tamoxifen resistance in breast cancer patients. J Natl Cancer Inst, 98, 671–80. [DOI] [PubMed] [Google Scholar]
  36. Hoogstraat M, Hinrichs JW, Besselink NJ, Radersma-Van Loon JH, De Voijs CM, Peeters T, Nijman IJ, De Weger RA, Voest EE, Willems SM, et al. 2015. Simultaneous detection of clinically relevant mutations and amplifications for routine cancer pathology. J Mol Diagn, 17, 10–8. [DOI] [PubMed] [Google Scholar]
  37. Huang J, Perez-Burgos L, Placek BJ, Sengupta R, Richter M, Dorsey JA, Kubicek S, Opravil S, Jenuwein T & Berger SL 2006. Repression of p53 activity by Smyd2-mediated methylation. Nature, 444, 629–32. [DOI] [PubMed] [Google Scholar]
  38. Hyman DM, Smyth LM, Donoghue MTA, Westin SN, Bedard PL, Dean EJ, Bando H, El-Khoueiry AB, Perez-Fidalgo JA, Mita A, et al. 2017. AKT Inhibition in Solid Tumors With AKT1 Mutations. J Clin Oncol, 35, 2251–2259. [DOI] [PMC free article] [PubMed] [Google Scholar]
  39. Ishida N, Baba M, Hatanaka Y, Hagio K, Okada H, Hatanaka KC, Togashi K, Matsuno Y & Yamashita H 2018. PIK3CA mutation, reduced AKT serine 473 phosphorylation, and increased ERalpha serine 167 phosphorylation are positive prognostic indicators in postmenopausal estrogen receptor-positive early breast cancer. Oncotarget, 9, 17711–17724. [DOI] [PMC free article] [PubMed] [Google Scholar]
  40. Jemal A, Siegel R, Xu J & Ward E 2010. Cancer statistics, 2010. CA Cancer J Clin, 60, 277–300. [DOI] [PubMed] [Google Scholar]
  41. Johansson I, Lauss M, Holm K, Staaf J, Nilsson C, Fjallskog ML, Ringner M & Hedenfalk I 2015. Genome methylation patterns in male breast cancer - Identification of an epitype with hypermethylation of polycomb target genes. Mol Oncol, 9, 1565–79. [DOI] [PMC free article] [PubMed] [Google Scholar]
  42. Johansson I, Nilsson C, Berglund P, Lauss M, Ringner M, Olsson H, Luts L, Sim E, Thorstensson S, Fjallskog ML, et al. 2012. Gene expression profiling of primary male breast cancers reveals two unique subgroups and identifies N-acetyltransferase-1 (NAT1) as a novel prognostic biomarker. Breast Cancer Res, 14, R31. [DOI] [PMC free article] [PubMed] [Google Scholar]
  43. Johansson I, Nilsson C, Berglund P, Strand C, Jonsson G, Staaf J, Ringner M, Nevanlinna H, Barkardottir RB, Borg A, et al. 2011. High-resolution genomic profiling of male breast cancer reveals differences hidden behind the similarities with female breast cancer. Breast Cancer Res Treat, 129, 747–60. [DOI] [PubMed] [Google Scholar]
  44. Kawakami T, Zhang C, Taniguchi T, Kim CJ, Okada Y, Sugihara H, Hattori T, Reeve AE, Ogawa O & Okamoto K 2004. Characterization of loss-of-inactive X in Klinefelter syndrome and female-derived cancer cells. Oncogene, 23, 6163–9. [DOI] [PubMed] [Google Scholar]
  45. Kornegoor R, Moelans CB, Verschuur-Maes AH, Hogenes M, De Bruin PC, Oudejans JJ & Van Diest PJ 2012a. Promoter hypermethylation in male breast cancer: analysis by multiplex ligation-dependent probe amplification. Breast Cancer Res, 14, R101. [DOI] [PMC free article] [PubMed] [Google Scholar]
  46. Kornegoor R, Moelans CB, Verschuur-Maes AH, Hogenes MC, De Bruin PC, Oudejans JJ, Marchionni L & Van Diest PJ 2012b. Oncogene amplification in male breast cancer: analysis by multiplex ligation-dependent probe amplification. Breast Cancer Res Treat, 135, 49–58. [DOI] [PMC free article] [PubMed] [Google Scholar]
  47. Kornegoor R, Verschuur-Maes AH, Buerger H, Hogenes MC, De Bruin PC, Oudejans JJ, Van Der Groep P, Hinrichs B & Van Diest PJ 2012c. Molecular subtyping of male breast cancer by immunohistochemistry. Mod Pathol, 25, 398–404. [DOI] [PubMed] [Google Scholar]
  48. Kurtyka CA, Chen L & Cress WD 2014. E2F inhibition synergizes with paclitaxel in lung cancer cell lines. PLoS One, 9, e96357. [DOI] [PMC free article] [PubMed] [Google Scholar]
  49. Lacle MM, Kornegoor R, Moelans CB, Maes-Verschuur AH, Van Der Pol C, Witkamp AJ, Van Der Wall E, Rueschoff J, Buerger H & Van Diest PJ 2013. Analysis of copy number changes on chromosome 16q in male breast cancer by multiplex ligation-dependent probe amplification. Mod Pathol, 26, 1461–7. [DOI] [PubMed] [Google Scholar]
  50. Lacle MM, Moelans CB, Kornegoor R, Van Der Pol C, Witkamp AJ, Van Der Wall E, Rueschoff J, Buerger H & Van Diest PJ 2015. Chromosome 17 copy number changes in male breast cancer. Cell Oncol (Dordr), 38, 237–45. [DOI] [PMC free article] [PubMed] [Google Scholar]
  51. Lehmann U, Streichert T, Otto B, Albat C, Hasemeier B, Christgen H, Schipper E, Hille U, Kreipe HH & Langer F 2010. Identification of differentially expressed microRNAs in human male breast cancer. BMC Cancer, 10, 109. [DOI] [PMC free article] [PubMed] [Google Scholar]
  52. Lips EH, Michaut M, Hoogstraat M, Mulder L, Besselink NJ, Koudijs MJ, Cuppen E, Voest EE, Bernards R, Nederlof PM, et al. 2015. Next generation sequencing of triple negative breast cancer to find predictors for chemotherapy response. Breast Cancer Res, 17, 134. [DOI] [PMC free article] [PubMed] [Google Scholar]
  53. Liu L, Kimball S, Liu H, Holowatyj A & Yang ZQ 2015. Genetic alterations of histone lysine methyltransferases and their significance in breast cancer. Oncotarget, 6, 2466–82. [DOI] [PMC free article] [PubMed] [Google Scholar]
  54. Ma Y, Kurtyka CA, Boyapalle S, Sung SS, Lawrence H, Guida W & Cress WD 2008. A small-molecule E2F inhibitor blocks growth in a melanoma culture model. Cancer Res, 68, 6292–9. [DOI] [PMC free article] [PubMed] [Google Scholar]
  55. Mckenna A, Hanna M, Banks E, Sivachenko A, Cibulskis K, Kernytsky A, Garimella K, Altshuler D, Gabriel S, Daly M, et al. 2010. The Genome Analysis Toolkit: a MapReduce framework for analyzing next-generation DNA sequencing data. Genome Res, 20, 1297–303. [DOI] [PMC free article] [PubMed] [Google Scholar]
  56. Mi H, Huang X, Muruganujan A, Tang H, Mills C, Kang D & Thomas PD 2017. PANTHER version 11: expanded annotation data from Gene Ontology and Reactome pathways, and data analysis tool enhancements. Nucleic Acids Res, 45, D183–D189. [DOI] [PMC free article] [PubMed] [Google Scholar]
  57. Mi H, Muruganujan A, Casagrande JT & Thomas PD 2013. Large-scale gene function analysis with the PANTHER classification system. Nat Protoc, 8, 1551–66. [DOI] [PMC free article] [PubMed] [Google Scholar]
  58. Mueller JJ, Schlappe BA, Kumar R, Olvera N, Dao F, Abu-Rustum N, Aghajanian C, Delair D, Hussein YR, Soslow RA, et al. 2018. Massively parallel sequencing analysis of mucinous ovarian carcinomas: genomic profiling and differential diagnoses. Gynecol Oncol, 150, 127–135. [DOI] [PMC free article] [PubMed] [Google Scholar]
  59. Nik-Zainal S, Davies H, Staaf J, Ramakrishna M, Glodzik D, Zou X, Martincorena I, Alexandrov LB, Martin S, Wedge DC, et al. 2016. Landscape of somatic mutations in 560 breast cancer whole-genome sequences. Nature, 534, 47–54. [DOI] [PMC free article] [PubMed] [Google Scholar]
  60. Nik-Zainal S & Morganella S 2017. Mutational Signatures in Breast Cancer: The Problem at the DNA Level. Clin Cancer Res, 23, 2617–2629. [DOI] [PMC free article] [PubMed] [Google Scholar]
  61. Ong CC, Gierke S, Pitt C, Sagolla M, Cheng CK, Zhou W, Jubb AM, Strickland L, Schmidt M, Duron SG, et al. 2015. Small molecule inhibition of group I p21-activated kinases in breast cancer induces apoptosis and potentiates the activity of microtubule stabilizing agents. Breast Cancer Res, 17, 59. [DOI] [PMC free article] [PubMed] [Google Scholar]
  62. Ong CC, Jubb AM, Haverty PM, Zhou W, Tran V, Truong T, Turley H, O’brien T, Vucic D, Harris AL, et al. 2011. Targeting p21-activated kinase 1 (PAK1) to induce apoptosis of tumor cells. Proc Natl Acad Sci U S A, 108, 7177–82. [DOI] [PMC free article] [PubMed] [Google Scholar]
  63. Pereira B, Chin SF, Rueda OM, Vollan HK, Provenzano E, Bardwell HA, Pugh M, Jones L, Russell R, Sammut SJ, et al. 2016. The somatic mutation profiles of 2,433 breast cancers refine their genomic and transcriptomic landscapes. Nat Commun, 7, 11479. [DOI] [PMC free article] [PubMed] [Google Scholar]
  64. Piscuoglio S, Ng CK, Murray MP, Guerini-Rocco E, Martelotto LG, Geyer FC, Bidard FC, Berman S, Fusco N, Sakr RA, et al. 2016. The Genomic Landscape of Male Breast Cancers. Clin Cancer Res, 22, 4045–56. [DOI] [PMC free article] [PubMed] [Google Scholar]
  65. Pritzlaff M, Summerour P, Mcfarland R, Li S, Reineke P, Dolinsky JS, Goldgar DE, Shimelis H, Couch FJ, Chao EC, et al. 2017. Male breast cancer in a multi-gene panel testing cohort: insights and unexpected results. Breast Cancer Res Treat, 161, 575–586. [DOI] [PMC free article] [PubMed] [Google Scholar]
  66. Radu M, Semenova G, Kosoff R & Chernoff J 2014. PAK signalling during the development and progression of cancer. Nat Rev Cancer, 14, 13–25. [DOI] [PMC free article] [PubMed] [Google Scholar]
  67. Rizzolo P, Silvestri V, Valentini V, Zelli V, Zanna I, Masala G, Bianchi S, Palli D & Ottini L 2018. Gene-specific methylation profiles in BRCA-mutation positive and BRCA-mutation negative male breast cancers. Oncotarget, 9, 19783–19792. [DOI] [PMC free article] [PubMed] [Google Scholar]
  68. Rogawski DS, Ndoj J, Cho HJ, Maillard I, Grembecka J & Cierpicki T 2015. Two Loops Undergoing Concerted Dynamics Regulate the Activity of the ASH1L Histone Methyltransferase. Biochemistry, 54, 5401–13. [DOI] [PMC free article] [PubMed] [Google Scholar]
  69. Rosenthal R, Mcgranahan N, Herrero J, Taylor BS & Swanton C 2016. DeconstructSigs: delineating mutational processes in single tumors distinguishes DNA repair deficiencies and patterns of carcinoma evolution. Genome Biol, 17, 31. [DOI] [PMC free article] [PubMed] [Google Scholar]
  70. Silvestri V, Barrowdale D, Mulligan AM, Neuhausen SL, Fox S, Karlan BY, Mitchell G, James P, Thull DL, Zorn KK, et al. 2016. Male breast cancer in BRCA1 and BRCA2 mutation carriers: pathology data from the Consortium of Investigators of Modifiers of BRCA1/2. Breast Cancer Res, 18, 15. [DOI] [PMC free article] [PubMed] [Google Scholar]
  71. Silvestri V, Rizzolo P, Zelli V, Valentini V, Zanna I, Bianchi S, Tibiletti MG, Varesco L, Russo A, Tommasi S, et al. 2018. A possible role of FANCM mutations in male breast cancer susceptibility: Results from a multicenter study in Italy. Breast, 38, 92–97. [DOI] [PubMed] [Google Scholar]
  72. Silvestri V, Zelli V, Valentini V, Rizzolo P, Navazio AS, Coppa A, Agata S, Oliani C, Barana D, Castrignano T, et al. 2017. Whole-exome sequencing and targeted gene sequencing provide insights into the role of PALB2 as a male breast cancer susceptibility gene. Cancer, 123, 210–218. [DOI] [PubMed] [Google Scholar]
  73. Stang A & Thomssen C 2008. Decline in breast cancer incidence in the United States: what about male breast cancer? Breast Cancer Res Treat, 112, 595–6. [DOI] [PubMed] [Google Scholar]
  74. Thomas DB, Rosenblatt K, Jimenez LM, Mctiernan A, Stalsberg H, Stemhagen A, Thompson WD, Curnen MG, Satariano W, Austin DF, et al. 1994. Ionizing radiation and breast cancer in men (United States). Cancer Causes Control, 5, 9–14. [DOI] [PubMed] [Google Scholar]
  75. Tommasi S, Mangia A, Iannelli G, Chiarappa P, Rossi E, Ottini L, Mottolese M, Zoli W, Zuffardi O & Paradiso A 2011. Gene copy number variation in male breast cancer by aCGH. Cell Oncol (Dordr), 34, 467–73. [DOI] [PubMed] [Google Scholar]
  76. Van Der Auwera GA, Carneiro MO, Hartl C, Poplin R, Del Angel G, Levy-Moonshine A, Jordan T, Shakir K, Roazen D, Thibault J, et al. 2013. From FastQ data to high confidence variant calls: the Genome Analysis Toolkit best practices pipeline. Curr Protoc Bioinformatics, 43, 11 10 1–33. [DOI] [PMC free article] [PubMed] [Google Scholar]
  77. Van Diest PJ 2002. No consent should be needed for using leftover body material for scientific purposes. For. BMJ, 325, 648–51. [PubMed] [Google Scholar]
  78. Van Diest PJ, Baak JP, Matze-Cok P, Wisse-Brekelmans EC, Van Galen CM, Kurver PH, Bellot SM, Fijnheer J, Van Gorp LH, Kwee WS, et al. 1992. Reproducibility of mitosis counting in 2,469 breast cancer specimens: results from the Multicenter Morphometric Mammary Carcinoma Project. Hum Pathol, 23, 603–7. [DOI] [PubMed] [Google Scholar]
  79. Villamar Cruz O, Prudnikova TY, Araiza-Olivera D, Perez-Plasencia C, Johnson N, Bernhardy AJ, Slifker M, Renner C, Chernoff J & Arias-Romero LE 2016. Reduced PAK1 activity sensitizes FA/BRCA-proficient breast cancer cells to PARP inhibition. Oncotarget, 7, 76590–76603. [DOI] [PMC free article] [PubMed] [Google Scholar]
  80. Weiss JR, Moysich KB & Swede H 2005. Epidemiology of male breast cancer. Cancer Epidemiol Biomarkers Prev, 14, 20–6. [PubMed] [Google Scholar]
  81. Wolff AC, Hammond ME, Hicks DG, Dowsett M, Mcshane LM, Allison KH, Allred DC, Bartlett JM, Bilous M, Fitzgibbons P, et al. 2013. Recommendations for human epidermal growth factor receptor 2 testing in breast cancer: American Society of Clinical Oncology/College of American Pathologists clinical practice guideline update. J Clin Oncol, 31, 3997–4013. [DOI] [PubMed] [Google Scholar]
  82. Woo TG, Yoon MH, Hong SD, Choi J, Ha NC, Sun H & Park BJ 2017. Anti-cancer effect of novel PAK1 inhibitor via induction of PUMA-mediated cell death and p21-mediated cell cycle arrest. Oncotarget, 8, 23690–23701. [DOI] [PMC free article] [PubMed] [Google Scholar]
  83. Wu RC, Wang TL & Shih Ie M 2014. The emerging roles of ARID1A in tumor suppression. Cancer Biol Ther, 15, 655–64. [DOI] [PMC free article] [PubMed] [Google Scholar]
  84. Ye DZ & Field J 2012. PAK signaling in cancer. Cell Logist, 2, 105–116. [DOI] [PMC free article] [PubMed] [Google Scholar]
  85. Yost SE, Smith EN, Schwab RB, Bao L, Jung H, Wang X, Voest E, Pierce JP, Messer K, Parker BA, et al. 2012. Identification of high-confidence somatic mutations in whole genome sequence of formalin-fixed breast cancer specimens. Nucleic Acids Res, 40, e107. [DOI] [PMC free article] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

01

Supplementary Table 1. List of 1943 genes included in the targeted capture massively parallel sequencing platform.

NIHMS1536645-supplement-01.xlsx (45.1KB, xlsx)
10

Supplementary Figure 3. Sanger sequencing confirmation of a selected set of SNVs in HER2 (ERBB2), GATA3, BRCA2 and PIK3CA. Positions and samples are indicated. Not all samples with SNVs at these positions are shown.

NIHMS1536645-supplement-10.pdf (1MB, pdf)
11

Supplementary Figure 4. Association of copy number alteration (CNA) load with clinical, pathological and genomic characteristics.

NIHMS1536645-supplement-11.pdf (34.1KB, pdf)
12

Supplementary Figure 5. Association of MDM2 amplifications with clinical, pathological and genomic characteristics.

NIHMS1536645-supplement-12.pdf (38.9KB, pdf)
13

Supplementary Methods. Detailed protocol for DNA isolation and mutation calling.

NIHMS1536645-supplement-13.docx (28.4KB, docx)
14

Supplementary Results. Association of mutational load with clinicopathological characteristics, Sanger sequencing validation, and germline mutations in AR, CHEK2 and PALB2 with medium to high predicted impact

NIHMS1536645-supplement-14.docx (17.4KB, docx)
02

Supplementary Table 2. Sequencing statistics (SOLID 5500).

NIHMS1536645-supplement-02.xlsx (44.4KB, xlsx)
03

Supplementary Table 3. Sanger sequencing variants and primer sequences.

NIHMS1536645-supplement-03.xlsx (14.5KB, xlsx)
04

Supplementary Table 4. List of somatic mutations.

NIHMS1536645-supplement-04.xlsx (1MB, xlsx)
05

Supplementary Table 5. BRCA2 germline mutations within the paired tumor/normal subcohort.

NIHMS1536645-supplement-05.xlsx (12.3KB, xlsx)
06

Supplementary Table 6. List of the 220 most frequently amplified genes in this study. The table includes their chromosomal location, the amplification frequency in ER-positive female breast cancer with associated p-value for difference in proportions, the preference for surrogate intrinsic molecular subtype, and associations with clinical, pathological and genetic characteristics. Potentially druggable gene categories, clinical actionability and known drug interactions are indicated per gene.

NIHMS1536645-supplement-06.xlsx (77.5KB, xlsx)
07

Supplementary Table 7. Kaplan Meier and multivariable Cox Regression 5-year overall survival (OS) analysis of 10 frequently amplified genes in MBC.

NIHMS1536645-supplement-07.xlsx (14.4KB, xlsx)
08

Supplementary Figure 1. Pattern of PIK3CA somatic mutations in male (n=107) versus female breast cancer (n=5284; cBioPortal)(Cerami et al., 2012, Gao et al., 2013).

NIHMS1536645-supplement-08.pdf (365.8KB, pdf)
09

Supplementary Figure 2. PIK3CA mutations predict worse 5-year overall survival in male breast cancer.

NIHMS1536645-supplement-09.pdf (348.8KB, pdf)

RESOURCES