A Stromal Immune Module Correlated with the Response to Neoadjuvant Chemotherapy, Prognosis and Lymphocyte Infiltration in HER2-Positive Breast Carcinoma Is Inversely Correlated with Hormonal Pathways

Anne-Sophie Hamy; Hélène Bonsang-Kitzis; Marick Lae; Matahi Moarii; Benjamin Sadacca; Alice Pinheiro; Marion Galliot; Judith Abecassis; Cecile Laurent; Fabien Reyal

doi:10.1371/journal.pone.0167397

Abstract

Introduction

HER2-positive breast cancer (BC) is a heterogeneous group of aggressive breast cancers, the prognosis of which has greatly improved since the introduction of treatments targeting HER2. However, these tumors may display intrinsic or acquired resistance to treatment, and classifiers of HER2-positive tumors are required to improve the prediction of prognosis and to develop novel therapeutic interventions.

Methods

We analyzed 2893 primary human breast cancer samples from 21 publicly available datasets and developed a six-metagene signature on a training set of 448 HER2-positive BC. We then used external public datasets to assess the ability of these metagenes to predict the response to chemotherapy (Ignatiadis dataset), and prognosis (METABRIC dataset).

Results

We identified a six-metagene signature (138 genes) containing metagenes enriched in different gene ontologies. The gene clusters were named as follows: Immunity, Tumor suppressors/proliferation, Interferon, Signal transduction, Hormone/survival and Matrix clusters. In all datasets, the Immunity metagene was less strongly expressed in ER-positive than in ER-negative tumors, and was inversely correlated with the Hormonal/survival metagene. Within the signature, multivariate analyses showed that strong expression of the “Immunity” metagene was associated with higher pCR rates after NAC (OR = 3.71[1.28–11.91], p = 0.019) than weak expression, and with a better prognosis in HER2-positive/ER-negative breast cancers (HR = 0.58 [0.36–0.94], p = 0.026). Immunity metagene expression was associated with the presence of tumor-infiltrating lymphocytes (TILs).

Conclusion

The identification of a predictive and prognostic immune module in HER2-positive BC confirms the need for clinical testing for immune checkpoint modulators and vaccines for this specific subtype. The inverse correlation between Immunity and hormone pathways opens research perspectives and deserves further investigation.

Citation: Hamy A-S, Bonsang-Kitzis H, Lae M, Moarii M, Sadacca B, Pinheiro A, et al. (2016) A Stromal Immune Module Correlated with the Response to Neoadjuvant Chemotherapy, Prognosis and Lymphocyte Infiltration in HER2-Positive Breast Carcinoma Is Inversely Correlated with Hormonal Pathways. PLoS ONE 11(12): e0167397. https://doi.org/10.1371/journal.pone.0167397

Editor: William B. Coleman, University of North Carolina at Chapel Hill School of Medicine, UNITED STATES

Received: September 19, 2016; Accepted: November 14, 2016; Published: December 22, 2016

Copyright: © 2016 Hamy et al. This is an open access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.

Data Availability: All relevant data are within the paper and its Supporting Information files.

Funding: A-S Hamy-Petit was supported by an ITMO-INSERM-AVIESAN cancer translational research grant. Funding was also obtained from the Site de Recherche Intégrée en Cancérologie/INstitut National du Cancer (InCa-DGOS-4654). The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.

Competing interests: The authors have declared that no competing interests exist.

Abbreviations: BC, breast cancer; BCSS, breast cancer-specific survival; BC_CL, breast cancer cell lines; CCLE, Cancer Cell Line Encyclopedia; CGP, Cancer Genome Project; ER, estrogen receptor; GE, gene expression; HER2-positive BC, HER2-positive breast cancer; HER2, human epidermal growth factor receptor 2; PR, progesterone receptor; pCR, pathological complete response; RMA, robust multichip average; TILs, tumor-infiltrating lymphocytes

Introduction

HER2-positive breast carcinomas (BCs) are defined by amplification and overexpression of the HER2 tyrosine kinase receptor gene (17q12). The tumors of this subgroup have aggressive pathological features and a high rate of early distant metastatic events. They are routinely treated with a combination of docetaxel plus a monoclonal antibody targeting the HER2 receptor (trastuzumab). Other drugs also appear to be of major interest and will probably be made available for routine treatment in the near future (lapatinib, pertuzumab and T-DM1).

HER2-positive BCs constitute a heterogeneous group of tumors differing in histological features, gene expression profiles, clinical behavior, overall prognosis, and response to conventional systemic cytotoxic therapy. Trastuzumab-based treatments have been used for the last decade and have substantially improved outcomes in patients with early or metastatic HER2-positive BC. However, some HER2-positive tumors display intrinsic or acquired resistance to trastuzumab. Robust classifiers are required, both to improve our understanding of the molecular basis of HER2-positive BC and to develop novel therapeutic interventions.

We developed a two-step biological network-driven gene selection process: 1) identification of the most variable genes displaying highly correlated patterns of expression, 2) direct connection of these genes within known biological networks. This method has been shown to construct molecular signatures efficiently [1–3]. We defined a HER2-positive molecular subtype classification and identified a stromal immune module gene expression profile strongly correlated with predicted response to chemotherapy, prognosis and lymphocytic infiltration. This classification provides considerable biological insight, and has potential for use in the development of therapeutic interventions, such as novel immunotherapies in particular.

Material and methods

Data normalization and quality control

Training, validation and Ignatiadis datasets.

We collected 21 publicly available datasets (described in the S1 File) containing raw gene expression data for 2893 primary human breast cancer samples. The data were normalized by the robust multichip average (RMA) procedure from the EMA R package [4]. The datasets were split into training (HGU-133A Affymetrix* arrays, 12 datasets, n = 1921) and validation (HGU-133Plus2 Affymetrix* arrays (9 datasets, n = 972) sets. Batch effects were eliminated by the median centering of each probe-set across arrays and by an independent quantile normalization of all arrays for each dataset. We controlled for outliers with the Array Quality Metrics R package. We also collected two large datasets to validate our classification: The Ignatiadis dataset (Affymetrix data n = 996) [5] and the METABRIC dataset (Illumina data n = 1992) published by Curtis et al. [6].

Determination and preprocessing of HER2-positive breast cancer samples

We identified the HER2-positive samples in the training and validation datasets, on the basis of transformed ERBB2 mRNA expression, as described by Gong et al. [7], and using the bimodal distribution of ERBB2 expression for the Ignatiadis and the METABRIC dataset.

Gene selection process

Consensus clustering with the ConsensusClusterPlus R Package was applied to the training set with a ward inner, final linkage and Pearson distance, to determine the optimal number of robust gene clusters for the most variable genes (standard deviation>0.8). We investigated the enrichment of each gene cluster in particular types of genes, and categorized and labeled genes clusters according to the different gene ontologies. We then identified known biological networks, for each gene cluster separately, using String* database software version 9.1 (http://string-db.org/) [8]. We then applied a two-step selection process: 1) we selected strong biological networks by retaining only genes for which connection scores of at least 0.7 were obtained with String* database software, 2) within each biological network, we selected groups of genes with correlated expression patterns and a correlation coefficient of at least 0.5.

For each dataset (the training, validation, Ignatiadis and METABRIC sets), we applied a hierarchical clustering procedure with a ward inner, final linkage and Pearson distance to the HER2-positive gene expression (GE) profiles, using the selected genes to visualize the optimal number of stable HER2-positive subtypes.

Metagene construction

We defined a metagene as an aggregate patterns of gene expression. Metagene expression was assessed by calculating the median normalized expression values of all probe sets in the respective gene clusters for each sample. The metagene value for each sample was then discretized on the basis of the median value, as “high” or “low”.

Association between expression of the Immunity metagene and that of ESR1, PGR, and AR

All the analyses were performed on all four datasets (training, validation, Ignatiadis, METABRIC). The levels of expression of ESR1, PGR and AR were compared between “Immunity low” and “Immunity high” samples, by ANOVA. Levels of Immunity metagene expression were compared between samples positive and negative for ER, PR, and AR, by ANOVA. We also performed ANOVA for each gene of the Immunity metagene as a function of ER status.

Analysis of the predicted response to NAC

We analyzed the predicted response to chemotherapy in the datasets published by Ignatiadis et al. [5]. Expression data were summarized by defining a metagene for each gene cluster. The clinical and pathological variables available for each dataset are described in S1 File. Qualitative variables were compared with logistic regression models.

Prognostic analysis

Prognostic analysis was performed on the METABRIC set. Expression data were summarized by defining a metagene for each gene cluster. The clinical and pathological variables available for each dataset are described in S1 File. Survival analyses were performed for the whole population, and separately for ER-positive and ER-negative patients, by calculating Kaplan-Meier estimates of the survival function. The endpoint of these analyses was breast cancer-specific survival (BCSS). Survival curves were compared in log-rank tests. Hazard ratios were estimated with Cox’s proportional hazard model. Predictive and prognostic analyses were performed with the R survival package. Variables associated with pCR or BCSS with a P-value <0.10 in univariate analysis were included in the multivariate model. Variables with P-values <0.05 in multivariate analysis were considered statistically significant.

Correlation with tumor-infiltrating lymphocyte levels

We downloaded the gene expression data from the REMAGUS 02 trial [9] and retrieved 27 samples for which paraffin-embedded tissue sections were available at our institution. All patients enrolled in this study gave their informed written consent. Histologic microbiopsy specimens were evaluated independently for the presence of a lymphocytic infiltrate (intratumoral TILs and stromal TILs by one BC pathologist (ML) and one breast physician (ASH) unaware of the gene expression classification. Percentages of TLs and StrL were compared, as a function of Immunity metagene status, in ANOVA. The correlations between Immunity metagene expression and the percentages of TLs and StrL were assessed by calculating Pearson’s correlation coefficient.

Expression of the gene signature in human breast cancer cell lines

We downloaded the gene expression profiles of the human cancer cell lines from the Cancer Cell Line Encyclopedia (CCLE) [10] of Novartis/the Broad Institute and the Cancer Genome Project (CGP) [11] of the Sanger Institute. We normalized the data for all the cell lines from different tissues together.

Statistical analysis

Data were processed and statistical analyses were carried out with R software version 3.1.2 [12] (www.cran.r-project.org).

Results

HER2-positive gene expression profiles identify six main gene clusters

HER2-positive BC samples were selected from 21 publicly available datasets (n = 3,247 breast cancer samples) and separated into a training set and a validation set (S1 File and S1 Fig). In the training set, we applied a gene selection process based on biological networks (Fig 1A to 1C), to decrease the instability intrinsic to molecular classification methods (see S1 File), as previously described for triple-negative breast cancers (TNBCs) [3]. We selected a final set of 138 genes (S1 Table), composed of six gene clusters enriched in different gene ontologies: Immunity (n = 28), Interferon (n = 11), Signal transduction (n = 20), Hormonal/survival (n = 22), Tumor suppressors/Proliferation (n = 36), Matrix (n = 21) (Fig 1D). We defined a metagene for each of the six gene clusters identified in this way (S1 File). The Immunity and Interferon metagenes displayed similar patterns of expression. The Immunity and Hormonal/survival metagenes displayed the strongest inverse correlation for expression (coefficient of -0.46) (Fig 1E and 1F). The correlations between the 138 genes and the metagenes are described in more detail in S1 File. For validation, we applied hierarchical clustering methods to three additional independent HER2-positive datasets; a validation set (n = 194), the Ignatiadis dataset (n = 82) and the METABRIC dataset (n = 248) (S1 File and S2 Fig).

Download:

Fig 1. Gene selection process.

A Heatmap showing the 616 most variable genes in the 448 HER2-positive samples (training set). B String database software confidence view of the Matrix genes cluster. Stronger associations between genes are represented by thicker lines. C Cytoscape View for the Immunity gene cluster. GE correlations between genes are indicated by edges (edge color varies from green to red and edge size increases with increasing correlation) and gene expression variance is represented by node color (node color varies from green to red and node size increases with increasing variance). D Heatmap showing the relative expression of 138 selected genes in 448 HER2-positive samples from the training set. E Table of Pearson’s correlation coefficient values for the correlations between the 6 metagenes. F Heatmap showing the anticorrelation between the Immunity and the Hormone/Survival metagene.

https://doi.org/10.1371/journal.pone.0167397.g001

The expression of the Immunity metagene is strongly associated with ER status, PR, and AR status

Given the inverse correlation between Immunity metagene and the Hormonal/survival metagene expression (Fig 2A) and with the strong correlation of Hormonal/survival metagene expression with ESR1 expression (Pearson correlation coefficient = 0.77), we compared levels of ESR1, PGR and AR expression as a function of Immunity metagene status (Fig 2B). These three genes were consistently more strongly expressed in the “Immunity low” subgroup than in the “Immunity high” subgroup (p< 10⁻¹⁶, p< 10⁻⁸, p = 0.002 respectively). Similar results were obtained with the other three datasets, although less consistently for PR and AR (S1 File).

Download:

Fig 2. Association between Hormone genes expression and Immunity genes expression.

A Correlation of Immunity metagene and Hormone/Survival metagene expression (training set). Pearson’s correlation coefficient is -0.46 (95% CI [-52.7–38.0], p<10⁻¹⁶). B Boxplots of global gene expression and ESR1, PGR and AR expression by Immunity metagene status, “low” versus “high” in the training set (A). P-values for ANOVA are p = 10⁻¹⁶, p = 10⁻⁶ and p = 0.0002, respectively. C Boxplots of Immunity metagene and immune gene (CTLA4 and PD1) expression levels by ER, PR and AR status in the training set (A). The p values for ANOVA were p<10⁻¹⁶, p = 0.002 and p = 0.008 for the Immunity metagene, CTLA4 and PD1 by ER status, respectively; p = 0.0001, p = 0.05 and p = 0.001 by PR status, respectively; and p<10^−6, p = 0.006 and p = 0.23 by AR status, respectively. The statistical significance (p-value) of the difference between gene expression values is indicated by black stars (p-value ≤ 0.05: *; p-value≤ 0.01: **; p-value≤ 0.001: ***).

https://doi.org/10.1371/journal.pone.0167397.g002

We then compared the levels of expression of our Immunity metagene with those of two other immune genes (CTLA4 and PD1; PDL1 was not available on the HGU133a Chip) as a function of ER, PR, and AR status. The Immunity metagene and CTLA4 were significantly more strongly expressed in the ER-negative, PR-negative, and AR-negative subgroups (Fig 2C). PD1 was significantly more strongly expressed in ER-negative and PR-negative tumors, but the difference in expression levels according to AR status was not significant for this gene. Similar findings were obtained when we compared each of the genes of the Immunity metagene separately as a function of ER status, and across the three other datasets. The results were less consistent for PR and AR (see S1 File). The proportions of tumors in the Immunity metagene “low” and “high” subgroups as a function of ER status differed significantly in three of the four datasets. ER-positive samples were more likely to be in the Immunity metagene “low” group, whereas ER-negative samples were more likely to be in the Immunity metagene “high” group (S1 File).

These findings suggest that there are strong inverse interactions between immune pathways that are captured by the Immunity metagene and ER, PR, and AR hormonal pathways in HER2-positive breast cancer tumors.

Predictive value of the Immunity metagene in HER2-positive breast cancers

We assessed the value of the six metagenes for predicting the response to neoadjuvant chemotherapy (NAC) on 82 HER2-positive samples from the Ignatiadis dataset. Univariate analysis identified four factors (ER status, tumor grade, and Immunity and Hormone/survival metagene expression) correlated with pathological complete response (pCR) (Table 1). In multivariate analysis, both ER status and the Immunity metagene were significantly associated with pCR (ER-positive: OR = 0.29 [0.09–0.82] versus ER-negative (reference class), p = 0.02; Immunity metagene “high” expression: OR = 3.71, 95% CI [1.28–11.91], versus “low” expression (reference class), p = 0.02) (Fig 3A). Analyses in the subset of patients that did not receive trastuzumab (n = 75) yielded similar results (S1 File).

Download:

Table 1. Association of clinical factors and gene cluster expression with pathological response rates after neoadjuvant chemotherapy in the Ignatiadis dataset, univariate and multivariate analysis.

https://doi.org/10.1371/journal.pone.0167397.t001

Download:

Fig 3. pCR and DSS outcomes in the Ignatiadis and the METABRIC dataset.

A: pCR rates by ER and Immunity metagene status (low versus high in the Ignatiadis dataset). B: Kaplan-Meier plots. Disease-specific survival of the ER-negative population (n = 138) according to Immunity metagene expression (low/high) and nodal status in the METABRIC dataset.

https://doi.org/10.1371/journal.pone.0167397.g003

We compared the predictive value of the Immunity metagene with that of nine immune signatures or metagenes already validated as predictors of the response to chemotherapy for breast cancer, notably in HER2-positive BCs [13–18]. In multivariate analysis, the Immunity metagene and six of the other signatures or metagenes tested were identified as predictive of the response to chemotherapy. The smallest p-value obtained was that for our Immunity metagene (p = 0.019), OR = 3.71, 95% CI [1.28–11.91] (S2 Table).

We then investigated the reasons for which the Immunity metagene (28 genes) was predictive of pCR in HER2-positive BCs, whereas the Immunity2 metagene (47 genes) published by Bonsang et al. [3] was not in a TNBC population [3], despite the strong correlation between these two signatures in three independent datasets (correlation coefficients: 0.96; 0.94 and 0.96 in the training set, METABRIC and Ignatiadis dataset, respectively). We applied both signatures to the whole population for the Ignatiadis dataset, and analyzed pCR as a function of breast cancer subtype and Immunity metagene status. We found that pCR rates were significantly higher in the “Immunity high” subgroup in HER2-negative/ER-positive (16.7% versus 8.4%, OR = 2.17, p = 0.05), HER2-positive (43.6% versus 16.7%, OR = 3.84, p = 0.01), and TNBC breast cancers (37.3 versus 22.6%, OR = 2.08, p = 0.03) (S3A Fig). A similar pattern was observed for the Immunity2 metagene (HER2-negative-ER positive: 16.5% versus 8.1%, OR = 2.22, p = 0.05), HER2-positive (45.5% versus 18.7%, OR = 3.57, p = 0.01), and TNBC breast cancers (36.3 versus 24.6%, OR = 1.75, p = 0.08; S3B Fig), but the difference was not statistically significant (p = 0.08) in the TNBC subgroup. Interestingly, Immunity metagene status appeared to have a larger effect on pCR rates in the HER2-positive subgroup (OR = 3.84 and 3.57, respectively) than in the ER-positive (OR = 2.17 and 2.22, respectively) and TNBC (OR = 2.08 and 1.75, respectively) subgroups. The Immunity metagene therefore seems to be associated with the response to NAC in all breast cancer subtypes, with a marked effect in terms of both the strength and magnitude of the association in the HER2-positive subgroup.

Prognostic value of the Immunity metagene in HER2-positive breast cancers

The prognostic value of the 138-gene HER2-positive signature was assessed with 248 HER2-positive samples from the METABRIC dataset. Univariate analysis identified five factors (menopausal status, tumor size, nodal status, Immunity and Signal transduction metagene expression) significantly correlated with a poor outcome (disease-specific survival) (Table 2).

Download:

Table 2. Survival analysis (disease-specific survival) in the METABRIC dataset (univariate and multivariate analysis); whole population and ER-negative population.

https://doi.org/10.1371/journal.pone.0167397.t002

In multivariate analysis, nodal status (node-negative versus node-positive) was significantly associated with a poor outcome (HR = 3.29 [2.14–5.06], p<0.001), and there was a trend towards association between high levels of Immunity metagene expression and better disease-free survival (DFS; HR = 0.70 [0.48–1.01], p = 0.054). In the ER-negative population, the Immunity metagene was found to be of significant prognostic value in multivariate analysis (n = 138) (HR = 0.58 [0.36–0.94], p = 0.026; Fig 3B), but was not associated with DFS in the ER-positive population (n = 110) (p = 0.43). We compared the prognostic value of the Immunity metagene with that of nine previously published immune signatures or metagenes known to predict survival in several breast cancer subtypes [14,17–22]. None of the signatures or metagenes described above was significantly associated with prognosis (S2F Table).

The Immunity metagene is correlated with tumor-infiltrating lymphocytes (TILs) in HER2-positive breast cancer

We then investigated the correlation between Immunity metagene expression and lymphocyte infiltration. We analyzed an independent set of HER2-positive tumors for which both histology and gene expression data were available (n = 27). Intratumoral TILs (TLs) and stromal TILs (StrL) were evaluated separately. Intratumoral TIL percentages were significantly higher in patients with strong Immunity metagene expression than in those with weak Immunity metagene expression (24% and 9%, respectively, p = 0.001) (Fig 4A). The same pattern was observed for the percentage of stromal TILs (36% versus 16.6%, p = 0.009) (Fig 4B). The coefficients of correlation between Immunity metagene expression level on the one hand and the percentage of intratumoral TILs (Fig 4C) or stromal TILs (Fig 4D) on the other hand were high (r = 0.60, p<0.001 and r = 0.69, p<0.00001 respectively). Lymphocyte infiltration is shown for two specimens, one with weak (Fig 5A and 5B), and the other with strong lymphocyte infiltration (Fig 5C and 5D). The Immunity metagene was therefore strongly correlated with the amount of lymphocyte infiltration in both the stromal compartment and the tumor bed.

Download:

Fig 4. Association between tumor-infiltrating lymphocyte levels and Immunity metagene expression in the REMAGUS dataset.

A: Percentage of intratumoral TILs according to Immunity metagene status (low versus high). B Percentage of stromal TILs according to Immunity metagene status (low versus high). C: Correlation between metagene expression and the percentages of intratumoral TILs. D: Correlation between metagene expression and the percentage of stromal TILs.

https://doi.org/10.1371/journal.pone.0167397.g004

Download:

Fig 5. Lymphocytic infiltration in breast tumors.

A and B: Tumor specimen with weak lymphocytic infiltration (A: zoom x10 B: zoom x 40). Abbreviations: S = stroma, T = tumor, L = lymphocytes. Intratumoral TILs are indicated by a black star. C and D: Tumor specimen with prominent lymphocytic infiltration. (C: zoom x10 D: zoom x 40). Abbreviations: S = stroma, T = tumor, L = lymphocytes. Intratumoral TILs are indicated by a black star; stromal TILs are indicated by a blue star.

https://doi.org/10.1371/journal.pone.0167397.g005

The Immunity metagene corresponds to the B-cell, T-cell and CD8 cell pathways

The Immunity metagene was strongly correlated with several published immune signatures (S4 Fig and S1 File), suggesting the use of similar immune pathways (see S1 File). We analyzed the correlation between expression of the Immunity and Interferon metagenes and expression of the metagenes defined by Gatza et al. [23] (IFN-alpha, IFN-gamma, STAT3, TGF-beta, TNF-alpha) and Palmer et al. [24] (LB, LT, CD8, GRANS, LYMPHS). This analysis was performed on the METABRIC dataset. The Immunity metagene was highly correlated with the B-cell, T-cell and CD8 cell metagenes (Pearson correlation coefficients: 0.89, 0.86, and 0.90, respectively; S5 Fig). We also assessed the correlations between the expression of PD1, PDL1, CTLA4, and that of their respective metagenes. The PD1 and CTLA-4 metagenes were constructed from the genes most strongly correlated with the PD1 and CTLA-4 genes, respectively (Pearson’s correlation coefficient > 0.8). The PDL1 metagene was defined by Sabatier et al. [25]. Pearson’s correlation coefficients for the relationships between the Immunity metagene and each individual gene were strong for PD1 and CTLA-4 (Pearson’s correlation coefficient: 0.75 and 0.84, respectively), and weaker for PDL1 (0.36), but the expression of all three metagenes was strongly correlated with that of the Immunity metagene (Pearson’s correlation coefficient: PD1: 0.89, PDL1: 0.95, CTLA-4: 0.93), opening up new possibilities for therapeutic intervention.

The Immunity metagene is probably expressed by stromal cells

In breast cancer cell lines (CCLE and CGP datasets), the Immunity metagene displayed very low levels of expression, similar to those of the CD8 metagene (S6A and S6B Fig), consistent with expression only in the tumor stromal compartment. This pattern was observed for all cell lines and breast cancer cell lines tested. The Interferon module genes had higher median expression levels and a broader range of expression than those of the Immunity metagene in breast cancer cell lines, consistent with their expression by tumor cells. We also explored the contributions of stromal and cancer cells to the expression of the Immunity and Interferon metagenes in detail, by comparing our gene lists with the “stromal contribution to global gene expression evaluated in PDX RNAseq data”, as defined by Isella et al. [26]. The stromal fraction of the Immunity metagene was high, although lower than those of the Matrix and the Tumor suppressor/proliferation metagenes. The Interferon metagene had a low stromal fraction, like the Hormone/survival and Signal transduction metagenes (S6C Fig). Although these data relate to the colon cancer PDX model, they provide support for the stromal expression of the Immunity metagene.

Discussion

By analyzing the gene expression profiles of 448 HER2-positive breast cancers, we identified a six-metagene signature (138 genes) in which each of the various metagenes was enriched in a different gene ontology. Within these metagenes, we identified an immune stromal module inversely correlated with the ER and hormonal pathways and strongly associated with the predicted response to chemotherapy, prognosis, and tumor lymphocyte infiltration. We report here one of the first immune signatures identified as both predictive and prognostic, reflecting histological immune infiltration in HER2-positive breast cancers. We also provide a relevant analysis by HR status.

We previously developed a strategy for defining gene expression signatures based on the analysis of biological networks for the most variable genes [3]. Since the early 2000s, a molecular classification of breast cancers has emerged that is continually being refined. Several authors have proposed TNBC subclassifications [3,27,28] but, to our knowledge, only one classifier has been published, but was not subsequently validated in HER2-positive BC [18]. The various metagenes in our signature were enriched in different gene ontologies: two clusters were enriched in immunity genes, one in signal transduction genes, one in hormonal/survival genes, one in tumor suppressor/proliferation genes and one in matrix genes. Unlike several other teams [29–31], we did not identify a subgroup to tumors overexpressing androgen receptor pathways in HER2-positive BCs by our biology-driven approach. The expression of the Immunity and Hormone/survival metagenes accurately predicted the response to NAC, but the expression of the Hormone/survival metagene had no significant effect in multivariate analysis, because the information it provided largely overlapped with ER status. Moreover, only the Immunity metagene was found to be of significant prognostic value.

Several authors have previously identified immunity patterns in HER2-positive BC. The Immunity module identified in our study had many biological connections with other predictive or prognostic immune signatures published for HER2-positive breast cancers [13–21], but it outperformed previous classifiers. This module includes genes encoding chemokines for T cells (CXCL10, CXCL9, CCL5), B cells (CXCL13), both B and T cells (CCL19) or other immune cells (CXCL13, CCL5); chemokine receptors (CCR7); cytokines (LTB); adhesion molecule-associated genes (SELL), and genes encoding proteins involved antigen processing and presentation (HLA-DRA), B-lymphocyte cell surface molecules (PTPRC, HLA-DRA), complement pathway proteins (C1QB), and proteins involved in CTL-mediated immune responses to target cells (CD3D), dendritic cell regulation of Th1 and Th2 development (CD2, IL7R), granzyme-mediated apoptosis (GZMA), IL12-mediated signaling events (CD3D, HLA-DRA, GZMA, LCK), the IL2 signaling pathway (LCK), T-cell surface molecules (PTPRC, CD3D, CD2), and molecules of the T-cell receptor signaling pathway (PTPRC, CD3D, HLA-DRA, LCK). It was also strongly correlated with the B-cell, T-cell and CD8 cell pathways.

There was a marked significant inverse association between ESR1 expression and that of the Immunity metagene. Similar inverse associations were found between PGR, AR and immunity, but these associations were weaker and less consistent. There is growing evidence for sex-based differences in the innate and adaptive immune responses underlying susceptibility to infectious diseases and the prevalence of autoimmune diseases. A higher proportion of men than of women display infectious diseases and their severity is also greater in men than in women [32]. By contrast, many autoimmune diseases predominantly affect women [33]. There are also difference between men in terms of humoral and cellular responses to infection and vaccination, with women often displaying higher response rates and mounting stronger humoral responses [34]. Estrogen receptors are expressed in most of the cells of the innate and adaptive immune system, including T cells, B cells, neutrophils, macrophages, dendritic cells (DC), and natural killer (NK) cells [35]. The effects of major sex steroid hormones were reviewed by Giefing-Kröll [36]. Estradiol and testosterone have opposite effects on the cells of the adaptive and innate immune systems, with estradiol having mostly enhancing and testosterone mostly suppressive effects. Estrogens affect the expression of some chemokine receptors (CCR1 and CCR5) by T cells [37]. They also affect B-cell development [38], decrease the cytotoxicity of NK cells [39] and regulate DC development [40]. TReg-cell frequencies within the CD4⁺ population change considerably during the ovarian cycle, with potential effects on immunoregulation [41]. Unlike the differences between the sexes in terms of infection and auto-immunity, the relationships between tumor immunology, sex and steroid hormones have remained largely unexplored. In two phase III trials, immunotherapy had a significant beneficial effect on survival only in male patients [42,43]. However, it remains unclear whether there is a true “sex” effect on the efficacy of immunotherapy or whether these findings are purely incidental.

The interaction between the ER, immunity and HER2 pathways is complex. There is increasing evidence to suggest that interactions between HER2 and hormone-receptor pathways play an important role in disease progression and that there is extensive, complex, bidirectional, crosstalk between the HER2 and ER pathways [44]. Immune signatures have been reported to have a predictive or prognostic role mostly in ER-negative breast cancers [45–48]. In HER2-positive breast cancer subtypes, Rody found that an immune T-cell metagene was of predictive value in both ER-positive and ER-negative HER2-positive BC [49]. The prognostic value of HDDP was demonstrated in both subgroups (11), but its value for predicting the response to NAC was not evaluated as a function of ER status. Conversely, the IRSN-23 [15] was not predictive in the ER-positive subpopulation. However, few authors determined the predictive [18] or prognostic value of their metagene or signature as a function of ER status within HER2-positive breast cancers [5,13,14,19–21]. The inverse association observed between ESR1 expression and immunity genes may be an important piece of the puzzle, and merits further investigation.

Consistent with previous reports [13,15,16], we found that the Immunity metagene was predictive of the response to NAC in HER2-positive BC. However, despite the similar gene module identification methods used and the strong correlation between the Immunity metagene and the Immunity2 metagene previously described by our team for TNBC [3], the Immunity metagene was predictive of the response to chemotherapy in HER2-positive BC, whereas the Immunity2 metagene was not predictive of the response to chemotherapy in TNBC. This finding was reported in the princeps report by Ignatiadis, in which high immune module scores were strongly and independently associated with a higher probability of pCR probability in HER2-positive tumors, whereas this association, although still significant, was weaker in TNBC [5]. ER-positive tumors have long been described as chemoresistant, with low pCR rates after NAC. Taking Immunity metagene expression into account, pCR rates ranged from 7.4 to 29.4%, with the highest rates close to those of ER-negative tumors.

The Immunity metagene was also prognostic in HER2-positive ER-negative breast cancer. The impact of immunity on prognosis has been reported before [21](Alexe et al., 2007)[21]¹⁴[18,20,21]. Together with our work, these findings suggest that immunity gene expression is highly predictive and of prognostic value in HER2-positive breast cancer. Nevertheless, the HER2-positive patients of the METABRIC dataset did not receive targeted anti-HER2 therapies, and our results would probably be influenced by adjuvant trastuzumab treatment.

We also demonstrated a correlation between Immunity metagene expression and stromal and intratumoral lymphocyte infiltration. The significance of TILs has recently become apparent, with advances in tumor immunology and the availability of cancer immunotherapies. TIL levels are strongly correlated with breast cancer subtype, and are higher in HER2-positive BCs than in ER-positive BCs, but lower than in TNBCs [50]. TIL levels are consistently higher in ER-negative tumors than in ER-positive tumors [51]. This was also found to be the case when the analysis was limited to HER2-positive BC only [52], [50]. The value of TIL levels for predicting pCR after NAC is less clear in HER2-positive BC than in TNBC. Stromal TILs and the lymphocyte-predominant breast cancer phenotype (LPBC) were strongly associated with treatment response in the GeparSixto trial [13]. However, this effect was found to be nonlinear in the NeoALTTO trial, and the optimal cutoff value remains unclear [52]. Two large studies in the adjuvant setting gave conflicting results. A positive association between higher levels of TILs and greater benefit from trastuzumab in HER2-positive disease was found in a retrospective analysis of the FinHER trial [50], whereas the opposite result was reported in the ALLIANCE N9831 study [53]. No difference in DFS between chemotherapy and chemotherapy plus trastuzumab was found in LPBC, whereas benefits of trastuzumab in addition to chemotherapy were observed only in non-LPBC. Thus, the prognostic impact of TILs on survival remains a matter of debate in HER2-positive BC. A few authors have reported a correlation between TIL and stromal lymphocyte levels and gene expression in HER2-positive breast cancers [13,15,21]. If this correlation is further validated, TIL levels could be used as a surrogate marker for the Immunity metagene, as TIL assessment is carried out in routine practice and is currently undergoing standardization [54].

Conclusion

Our work opens up a number of exciting therapeutic perspectives in HER2-positive breast cancers. Due to the high immunogenicity of HER2-positive breast cancers and the considerable predictive and prognostic impact of immunity in this subtype, immunotherapies may soon become part of the therapeutic arsenal for such cancers. Preclinical models have suggested that there is synergy between anti-HER2 monoclonal antibody and anti-PD-1 [55] or anti-CTLA4 antibodies [56]. The PANACEA phase Ib/II trial is currently investigating the use of pembrolizumab (KEYTRUDA^®) in combination with trastuzumab, to determine whether the addition of an anti-PD-1 treatment can overcome trastuzumab resistance in patients with HER2-positive breast cancer whose cancer spread whilst they were on trastuzumab. Future challenges in the field of immunity and HER2-positive breast cancers include:

The public accessibility of large sets of gene expression data for tumors from patients treated with HER2-targeting treatments. As treatments are constantly changing for this breast cancer subtype, it is important for expression data to be shared promptly, to facilitate comprehensive research and the identification of predictive and prognostic markers in patients treated with cutting edge care.
Improvements in our understanding of hormone and immunity pathways in HER2-positive breast cancers. In particular, it would be very useful to determine whether a subset of patients with HER2-positive ER-positive cancers could be effectively treated by a combination of endocrine therapy/immune checkpoint blockade/ targeted therapy, without the need for chemotherapy.
Drug positioning strategies in HER2-positive BC, because, by contrast to other breast cancer subtypes, the HER2-targeting drug pipeline contains many candidates despite the comparative rarity of this particular disease.
The selection criteria for the candidates most likely to benefit from immune checkpoint blockade is a key point. The use of PD-L1 as a surrogate marker of anti-PD-1 efficacy remains controversial, even in cancers for which immunotherapy treatments have proved effective, and few data are available for breast cancer. The standardization and demonstrations of the reproducibility of published immune signatures would be useful, as would improvements in our understanding of the prognostic value of TILs in HER2-positive breast cancers. Moreover, it remains to be determined whether and how the immunogenic power of tumors with low expression of immunity genes could be enhanced.

Once these challenges have been overcome, given the outstanding results of immunotherapy for other cancers (e.g. melanoma, lung cancer) and the expected efficacy of such treatment for HER2-positive disease, such therapies could revolutionize the course of HER2-positive breast cancer in the near future.

Supporting Information

S1 Fig. Methodology flow chart.

https://doi.org/10.1371/journal.pone.0167397.s001

(PDF)

S2 Fig. Heatmaps of the selected genes in the HER2-positive datasets.

Training set (upper left); validation set (upper right), Ignatiadis (lower left), METABRIC (lower right).

https://doi.org/10.1371/journal.pone.0167397.s002

(JPG)

S3 Fig. pCR rates by breast cancer subtype and Immunity metagene.

A: pCR rates by breast cancer subtype by Immunity metagene status (low versus high). B: pCR rates by breast cancer subtype by Immunity2 metagene status (low versus high) as previously published by Bonsang et al [3].

https://doi.org/10.1371/journal.pone.0167397.s003

(PDF)

S4 Fig. Heatmaps of the gene expression profiles of published immune signatures and connections between all immune genes.

Fig A. Heatmap of the gene expression profiles of the nine immune predictive signatures or metagenes previously published, applied to the Ignatiadis dataset. The samples were ordered according to our classification of Low/High ‘Immunity’ metagene expression. B: Heatmap of the gene expression profiles of the immune prognostic signatures or metagenes previously published, applied to the METABRIC dataset. The samples were ordered according to our classification of Low/High ‘Immunity’ metagene expression. C: String Software connections between genes of our Immunity metagenes and the genes of previously published predictive or prognostic immune signatures or metagenes. Stronger associations between genes are represented by thicker lines. Associations between genes with a coefficient < 0.9 are shown in green. Associations between genes with a coefficient ≥ 0.9 are shown in red. Associations between genes with a coefficient between 0.4 to 0.7 are not shown.

https://doi.org/10.1371/journal.pone.0167397.s004

(PDF)

S5 Fig. Distribution histograms for our Immune metagenes and immune pathways.

Distribution histograms for our Immune metagenes (Immunity and Interferon) and the immune pathway metagenes published by Gatza et al. (Interferon alpha, Interferon gamma, STAT3, TGF beta, TNF alpha) and Palmer et al. (B Cell, T Cell, CD8 T Cells, Granulocytes, Lymphocytes), Pearson correlation coefficient values and pairwise scatter plots.

https://doi.org/10.1371/journal.pone.0167397.s005

(PDF)

S6 Fig. Gene expression for the Immune metagenes and pathway, in cell lines and xenografts.

A. Boxplots of gene expression for the Immune metagenes, the immune pathway metagenes (published by Gatza et al. and Palmer et al.) and the PD1, PDL1, CTLA4 metagenes in breast cancer cell lines from the CCLE (A) and the CGP (B). C: Boxplots of the stromal contribution to global gene expression evaluated with PDX RNAseq data (Isella et al.), for each of the gene clusters for our signature.

https://doi.org/10.1371/journal.pone.0167397.s006

(PDF)

S1 File. Supplementary methods and results.

https://doi.org/10.1371/journal.pone.0167397.s007

(PDF)

S1 Table. 138-gene signature.

https://doi.org/10.1371/journal.pone.0167397.s008

(XLS)

S2 Table. Association of published immune signatures or metagenes with response to chemotherapy and prognosis.

Response to chemotherapy is assessed in the Ignatiadis dataset (univariate and multivariate analysis) (S2A to S2E Table). The association of published immune signatures or metagenes with prognosis is assessed in the METABRIC dataset (univariate analysis) (S2F Table).

https://doi.org/10.1371/journal.pone.0167397.s009

(XLS)

Acknowledgments

The authors thank Vassili Soumelis and Sergio Roman-Roman for reviewing of the study and the manuscript.

Author Contributions

Conceptualization: FR AP MG.
Data curation: BS.
Formal analysis: BS JA.
Investigation: BS CL.
Methodology: CL MM.
Project administration: FR CL.
Resources: ML.
Supervision: FR.
Validation: ML.
Writing – original draft: ASH HBK.
Writing – review & editing: ASH.

References

1. Fröhlich H. Network based consensus gene signatures for biomarker discovery in breast cancer. PLoS One. 2011;6: e25364. pmid:22046239
- View Article
- PubMed/NCBI
- Google Scholar
2. Cun Y, Fröhlich HF. Prognostic gene signatures for patient stratification in breast cancer: accuracy, stability and interpretability of gene selection approaches using prior knowledge on protein-protein interactions. BMC Bioinformatics. 2012;13: 69. pmid:22548963
- View Article
- PubMed/NCBI
- Google Scholar
3. Bonsang-Kitzis H, Sadacca B, Hamy-Petit AS, Moarii M, Pinheiro A, Laurent C, et al. Biological network-driven gene selection identifies a stromal immune module as a key determinant of triple-negative breast carcinoma prognosis. Oncoimmunology. 2015;
- View Article
- Google Scholar
4. Servant N, Gravier E, Gestraud P, Laurent C, Paccard C, Biton A, et al. EMA—A R package for Easy Microarray data analysis. BMC Res Notes. BioMed Central Ltd; 2010;3: 277.
- View Article
- Google Scholar
5. Ignatiadis M, Singhal SK, Desmedt C, Haibe-Kains B, Criscitiello C, Andre F, et al. Gene modules and response to neoadjuvant chemotherapy in breast cancer subtypes: a pooled analysis. J Clin Oncol. 2012;30: 1996–2004. pmid:22508827
- View Article
- PubMed/NCBI
- Google Scholar
6. Curtis C, Shah SP, Chin S-F, Turashvili G, Rueda OM, Dunning MJ, et al. The genomic and transcriptomic architecture of 2,000 breast tumours reveals novel subgroups. Nature. 2012;486: 346–352. pmid:22522925
- View Article
- PubMed/NCBI
- Google Scholar
7. Gong Y, Yan K, Lin F, Anderson K, Sotiriou C, Andre F, et al. Determination of oestrogen-receptor status and ERBB2 status of breast carcinoma: a gene-expression profiling study. Lancet Oncol. 2007;8: 203–211. pmid:17329190
- View Article
- PubMed/NCBI
- Google Scholar
8. Jensen LJ, Kuhn M, Stark M, Chaffron S, Creevey C, Muller J, et al. STRING 8—a global view on proteins and their functional interactions in 630 organisms. Nucleic Acids Res. 2009;37: D412–6. pmid:18940858
- View Article
- PubMed/NCBI
- Google Scholar
9. De Cremoux P, Valet F, Gentien D, Lehmann-Che J, Scott V, Tran-Perennou C, et al. Importance of pre-analytical steps for transcriptome and RT-qPCR analyses in the context of the phase II randomised multicentre trial REMAGUS02 of neoadjuvant chemotherapy in breast cancer patients. BMC Cancer. 2011;11: 215. pmid:21631949
- View Article
- PubMed/NCBI
- Google Scholar
10. Barretina J, Caponigro G, Stransky N, Venkatesan K, Margolin AA, Kim S, et al. NIH Public Access of anticancer drug sensitivity. 2012;483: 603–607.
- View Article
- Google Scholar
11. Garnett MJ, Edelman EJ, Heidorn SJ, Greenman CD, Dastur A, Lau KW, et al. Europe PMC Funders Group Systematic identification of genomic markers of drug sensitivity in cancer cells. Nature. 2012;483: 570–575.
- View Article
- Google Scholar
12. R Development Core Team. R: A Language and Environment for Statistical Computing. R Foundation for Sta- tistical Computing. Vienna, Austria 2009. [Internet]. http://www.R-project.org
13. Denkert C, von Minckwitz G, Brase JC, Sinn BV, Gade S, Kronenwett R, et al. Tumor-infiltrating lymphocytes and response to neoadjuvant chemotherapy with or without Carboplatin in human epidermal growth factor receptor 2-positive and triple-negative primary breast cancers. J Clin Oncol Off J Am Soc Clin Oncol. 2015;33: 983–991.
- View Article
- Google Scholar
14. Gu-Trantien C, Loi S, Garaud S, Equeter C, Libin M, Wind De A, et al. CD4+ follicular helper T cell infiltration predicts breast cancer survival. J Clin Invest. 2013;123: 1–20.
- View Article
- Google Scholar
15. Sota Y, Naoi Y, Tsunashima R, Kagara N, Shimazu K, Maruyama N, et al. Construction of novel immune-related signature for prediction of pathological complete response to neoadjuvant chemotherapy in human breast cancer. Ann Oncol Off J Eur Soc Med Oncol ESMO. 2014;25: 100–106.
- View Article
- Google Scholar
16. Stoll G, Enot D, Mlecnik B, Galon J, Zitvogel L, Kroemer G. Immune-related gene signatures predict the outcome of neoadjuvant chemotherapy. Oncoimmunology. 2014;3: e27884. pmid:24790795
- View Article
- PubMed/NCBI
- Google Scholar
17. Rody A, Karn T, Liedtke C, Pusztai L, Ruckhaeberle E, Hanker L, et al. A clinically relevant gene signature in triple negative and basal-like breast cancer. Breast Cancer Res BCR. 2011;13: R97. pmid:21978456
- View Article
- PubMed/NCBI
- Google Scholar
18. Staaf J, Ringnér M, Vallon-Christersson J, Jönsson G, Bendahl P-O, Holm K, et al. Identification of subtypes in human epidermal growth factor receptor 2—positive breast cancer reveals a gene signature prognostic of outcome. J Clin Oncol Off J Am Soc Clin Oncol. 2010;28: 1813–1820.
- View Article
- Google Scholar
19. Desmedt C, Haibe-Kains B, Wirapati P, Buyse M, Larsimont D, Bontempi G, et al. Biological processes associated with breast cancer clinical outcome depend on the molecular subtypes. Clin Cancer Res. 2008;14: 5158–5165. pmid:18698033
- View Article
- PubMed/NCBI
- Google Scholar
20. Finak G, Bertos N, Pepin F, Sadekova S, Souleimanova M, Zhao H, et al. Stromal gene expression predicts clinical outcome in breast cancer. Nat Med. 2008;14: 518–527. pmid:18438415
- View Article
- PubMed/NCBI
- Google Scholar
21. Alexe G, Dalgin GS, Scanfeld D, Tamayo P, Mesirov JP, DeLisi C, et al. High expression of lymphocyte-associated genes in node-negative HER2+ breast cancers correlates with lower recurrence rates. Cancer Res. 2007;67: 10669–10676. pmid:18006808
- View Article
- PubMed/NCBI
- Google Scholar
22. Perez EA, Thompson EA, Ballman KV, Anderson SK, Asmann YW, Kalari KR, et al. Genomic Analysis Reveals That Immune Function Genes Are Strongly Linked to Clinical Outcome in the North Central Cancer Treatment Group N9831 Adjuvant Trastuzumab Trial. J Clin Oncol Off J Am Soc Clin Oncol. 2015;
- View Article
- Google Scholar
23. Gatza ML, Lucas JE, Barry WT, Kim JW, Wang Q, Crawford MD, et al. A pathway-based classification of human breast cancer. Proc Natl Acad Sci U S A. 2010;107: 6994–6999. pmid:20335537
- View Article
- PubMed/NCBI
- Google Scholar
24. Palmer C, Diehn M, Alizadeh A a, Brown PO. Cell-type specific gene expression profiles of leukocytes in human peripheral blood. BMC Genomics. 2006;7: 115. pmid:16704732
- View Article
- PubMed/NCBI
- Google Scholar
25. Sabatier R, Finetti P, Mamessier E, Adelaide J, Chaffanet M, Ali HR, et al. Prognostic and predictive value of PDL1 expression in breast cancer. Oncotarget. 2015;6: 5449–5464. pmid:25669979
- View Article
- PubMed/NCBI
- Google Scholar
26. Isella C, Terrasi A, Bellomo SE, Petti C, Galatola G, Muratore A, et al. Stromal contribution to the colorectal cancer transcriptome. Nat Genet. Nature Publishing Group; 2015; 1–11.
- View Article
- Google Scholar
27. Lehmann BD, Bauer JA, Chen X, Sanders ME, Chakravarthy AB, Shyr Y, et al. Identification of human triple-negative breast cancer subtypes and preclinical models for selection of targeted therapies. J Clin Invest. 2011;121: 2750–2767. pmid:21633166
- View Article
- PubMed/NCBI
- Google Scholar
28. Burstein MD, Tsimelzon A, Poage GM, Covington KR, Contreras A, Fuqua S, et al. Comprehensive Genomic Analysis Identifies Novel Subtypes and Targets of Triple-negative Breast Cancer. Clin Cancer Res Off J Am Assoc Cancer Res. 2014;
- View Article
- Google Scholar
29. Doane AS, Danso M, Lal P, Donaton M, Zhang L, Hudis C, et al. An estrogen receptor-negative breast cancer subset characterized by a hormonally regulated transcriptional program and response to androgen. Oncogene. 2006;25: 3994–4008. pmid:16491124
- View Article
- PubMed/NCBI
- Google Scholar
30. Farmer P, Bonnefoi H, Becette V, Tubiana-Hulin M, Fumoleau P, Larsimont D, et al. Identification of molecular apocrine breast tumours by microarray analysis. Oncogene. 2005;24: 4660–4671. pmid:15897907
- View Article
- PubMed/NCBI
- Google Scholar
31. Lehmann-Che J, Hamy A-S, Porcher R, Barritault M, Bouhidel F, Habuellelah H, et al. Molecular apocrine breast cancers are aggressive estrogen receptor negative tumors overexpressing either HER2 or GCDFP15. Breast Cancer Res. 2013;15: R37. pmid:23663520
- View Article
- PubMed/NCBI
- Google Scholar
32. Klein SL. The effects of hormones on sex differences in infection: from genes to behavior. Neurosci Biobehav Rev. 2000;24: 627–638. pmid:10940438
- View Article
- PubMed/NCBI
- Google Scholar
33. Lockshin MD. Sex differences in autoimmune disease. Lupus. 2006;15: 753–756. pmid:17153846
- View Article
- PubMed/NCBI
- Google Scholar
34. Cook IF. Sexual dimorphism of humoral immunity with human vaccines. Vaccine. 2008;26: 3551–3555. pmid:18524433
- View Article
- PubMed/NCBI
- Google Scholar
35. Fish EN. The X-files in immunity: sex-based differences predispose immune responses. Nat Rev Immunol. 2008;8: 737–744. pmid:18728636
- View Article
- PubMed/NCBI
- Google Scholar
36. Giefing-Kröll C, Berger P, Lepperdinger G, Grubeck-Loebenstein B. How sex and age affect immune responses, susceptibility to infections, and response to vaccination. Aging Cell. 2015;14: 309–321. pmid:25720438
- View Article
- PubMed/NCBI
- Google Scholar
37. Mo R, Chen J, Grolleau-Julius A, Murphy HS, Richardson BC, Yung RL. Estrogen regulates CCR gene expression and function in T lymphocytes. J Immunol Baltim Md 1950. 2005;174: 6023–6029.
- View Article
- Google Scholar
38. Sakiani S, Olsen NJ, Kovacs WJ. Gonadal steroids and humoral immunity. Nat Rev Endocrinol. 2013;9: 56–62. pmid:23183675
- View Article
- PubMed/NCBI
- Google Scholar
39. Hao S, Zhao J, Zhou J, Zhao S, Hu Y, Hou Y. Modulation of 17beta-estradiol on the number and cytotoxicity of NK cells in vivo related to MCM and activating receptors. Int Immunopharmacol. 2007;7: 1765–1775. pmid:17996687
- View Article
- PubMed/NCBI
- Google Scholar
40. Siracusa MC, Overstreet MG, Housseau F, Scott AL, Klein SL. 17beta-estradiol alters the activity of conventional and IFN-producing killer dendritic cells. J Immunol Baltim Md 1950. 2008;180: 1423–1431.
- View Article
- Google Scholar
41. Arruvito L, Sanz M, Banham AH, Fainboim L. Expansion of CD4+CD25+and FOXP3+ regulatory T cells during the follicular phase of the menstrual cycle: implications for human reproduction. J Immunol Baltim Md 1950. 2007;178: 2572–2578.
- View Article
- Google Scholar
42. Robert C, Thomas L, Bondarenko I, O’Day S, Weber J, Garbe C, et al. Ipilimumab plus dacarbazine for previously untreated metastatic melanoma. N Engl J Med. 2011;364: 2517–2526. pmid:21639810
- View Article
- PubMed/NCBI
- Google Scholar
43. Brahmer J, Reckamp KL, Baas P, Crinò L, Eberhardt WEE, Poddubskaya E, et al. Nivolumab versus Docetaxel in Advanced Squamous-Cell Non-Small-Cell Lung Cancer. N Engl J Med. 2015;373: 123–135. pmid:26028407
- View Article
- PubMed/NCBI
- Google Scholar
44. Giuliano M, Trivedi MV, Schiff R. Bidirectional Crosstalk between the Estrogen Receptor and Human Epidermal Growth Factor Receptor 2 Signaling Pathways in Breast Cancer: Molecular Basis and Clinical Implications. Breast Care Basel Switz. 2013;8: 256–262.
- View Article
- Google Scholar
45. Sabatier R, Finetti P, Mamessier E, Raynaud S, Cervera N, Lambaudie E, et al. Kinome expression profiling and prognosis of basal breast cancers. Mol Cancer. 2011;10: 86. pmid:21777462
- View Article
- PubMed/NCBI
- Google Scholar
46. Teschendorff AE, Miremadi A, Pinder SE, Ellis IO, Caldas C. An immune response gene expression module identifies a good prognosis subtype in estrogen receptor negative breast cancer. Genome Biol. 2007;8: R157. pmid:17683518
- View Article
- PubMed/NCBI
- Google Scholar
47. Karn T, Pusztai L, Holtrich U, Iwamoto T, Shiang CY, Schmidt M, et al. Homogeneous datasets of triple negative breast cancers enable the identification of novel prognostic and predictive signatures. PloS One. 2011;6: e28403. pmid:22220191
- View Article
- PubMed/NCBI
- Google Scholar
48. West NR, Milne K, Truong PT, Macpherson N, Nelson BH, Watson PH. Tumor-infiltrating lymphocytes predict response to anthracycline-based chemotherapy in estrogen receptor-negative breast cancer. Breast Cancer Res BCR. BioMed Central Ltd; 2011;13: R126.
- View Article
- Google Scholar
49. Rody A, Holtrich U, Pusztai L, Liedtke C, Gaetje R, Ruckhaeberle E, et al. T-cell metagene predicts a favorable prognosis in estrogen receptor-negative and HER2-positive breast cancers. Breast Cancer Res BCR. 2009;11: R15. pmid:19272155
- View Article
- PubMed/NCBI
- Google Scholar
50. Loi S, Michiels S, Salgado R, Sirtaine N, Jose V, Fumagalli D, et al. Tumor infiltrating lymphocytes are prognostic in triple negative breast cancer and predictive for trastuzumab benefit in early breast cancer: results from the FinHER trial. Ann Oncol Off J Eur Soc Med Oncol ESMO. 2014;25: 1544–1550.
- View Article
- Google Scholar
51. Mahmoud SMA, Paish EC, Powe DG, Macmillan RD, Grainge MJ, Lee AHS, et al. Tumor-infiltrating CD8+ lymphocytes predict clinical outcome in breast cancer. J Clin Oncol Off J Am Soc Clin Oncol. 2011;29: 1949–1955.
- View Article
- Google Scholar
52. Salgado R, Denkert C, Campbell C, Savas P, Nucifero P, Aura C, et al. Tumor-Infiltrating Lymphocytes and Associations With Pathological Complete Response and Event-Free Survival in HER2-Positive Early-Stage Breast Cancer Treated With Lapatinib and Trastuzumab: A Secondary Analysis of the NeoALTTO Trial. JAMA Oncol. 2015;1: 448–454. pmid:26181252
- View Article
- PubMed/NCBI
- Google Scholar
53. Perez E. Stromal tumor-infiltrating lymphocytes(S-TILs): In the alliance N9831 trial S-TILs are associated with chemotherapy benefit but not associated with trastuzumab benefit [Internet]. 2014. http://www.abstracts2view.com/sabcs14/view.php?nu=SABCS13L_1455
54. Salgado R, Denkert C, Demaria S, Sirtaine N, Klauschen F, Pruneri G, et al. The evaluation of tumor-infiltrating lymphocytes (TILs) in breast cancer: recommendations by an International TILs Working Group 2014. Ann Oncol Off J Eur Soc Med Oncol ESMO. 2015;26: 259–271.
- View Article
- Google Scholar
55. Stagg J, Loi S, Divisekera U, Ngiow SF, Duret H, Yagita H, et al. Anti-ErbB-2 mAb therapy requires type I and II interferons and synergizes with anti-PD-1 or anti-CD137 mAb therapy. Proc Natl Acad Sci U S A. 2011;108: 7142–7147. pmid:21482773
- View Article
- PubMed/NCBI
- Google Scholar
56. Wang Q, Li S-H, Wang H, Xiao Y, Sahin O, Brady SW, et al. Concomitant targeting of tumor cells and induction of T-cell response synergizes to effectively inhibit trastuzumab-resistant breast cancer. Cancer Res. 2012;72: 4417–4428. pmid:22773664
- View Article
- PubMed/NCBI
- Google Scholar

[ref1] 1. Fröhlich H. Network based consensus gene signatures for biomarker discovery in breast cancer. PLoS One. 2011;6: e25364. pmid:22046239
View Article
PubMed/NCBI
Google Scholar

[2] View Article

[3] PubMed/NCBI

[4] Google Scholar

[ref2] 2. Cun Y, Fröhlich HF. Prognostic gene signatures for patient stratification in breast cancer: accuracy, stability and interpretability of gene selection approaches using prior knowledge on protein-protein interactions. BMC Bioinformatics. 2012;13: 69. pmid:22548963
View Article
PubMed/NCBI
Google Scholar

[6] View Article

[7] PubMed/NCBI

[8] Google Scholar

[ref3] 3. Bonsang-Kitzis H, Sadacca B, Hamy-Petit AS, Moarii M, Pinheiro A, Laurent C, et al. Biological network-driven gene selection identifies a stromal immune module as a key determinant of triple-negative breast carcinoma prognosis. Oncoimmunology. 2015;
View Article
Google Scholar

[10] View Article

[11] Google Scholar

[ref4] 4. Servant N, Gravier E, Gestraud P, Laurent C, Paccard C, Biton A, et al. EMA—A R package for Easy Microarray data analysis. BMC Res Notes. BioMed Central Ltd; 2010;3: 277.
View Article
Google Scholar

[13] View Article

[14] Google Scholar

[ref5] 5. Ignatiadis M, Singhal SK, Desmedt C, Haibe-Kains B, Criscitiello C, Andre F, et al. Gene modules and response to neoadjuvant chemotherapy in breast cancer subtypes: a pooled analysis. J Clin Oncol. 2012;30: 1996–2004. pmid:22508827
View Article
PubMed/NCBI
Google Scholar

[16] View Article

[17] PubMed/NCBI

[18] Google Scholar

[ref6] 6. Curtis C, Shah SP, Chin S-F, Turashvili G, Rueda OM, Dunning MJ, et al. The genomic and transcriptomic architecture of 2,000 breast tumours reveals novel subgroups. Nature. 2012;486: 346–352. pmid:22522925
View Article
PubMed/NCBI
Google Scholar

[20] View Article

[21] PubMed/NCBI

[22] Google Scholar

[ref7] 7. Gong Y, Yan K, Lin F, Anderson K, Sotiriou C, Andre F, et al. Determination of oestrogen-receptor status and ERBB2 status of breast carcinoma: a gene-expression profiling study. Lancet Oncol. 2007;8: 203–211. pmid:17329190
View Article
PubMed/NCBI
Google Scholar

[24] View Article

[25] PubMed/NCBI

[26] Google Scholar

[ref8] 8. Jensen LJ, Kuhn M, Stark M, Chaffron S, Creevey C, Muller J, et al. STRING 8—a global view on proteins and their functional interactions in 630 organisms. Nucleic Acids Res. 2009;37: D412–6. pmid:18940858
View Article
PubMed/NCBI
Google Scholar

[28] View Article

[29] PubMed/NCBI

[30] Google Scholar

[ref9] 9. De Cremoux P, Valet F, Gentien D, Lehmann-Che J, Scott V, Tran-Perennou C, et al. Importance of pre-analytical steps for transcriptome and RT-qPCR analyses in the context of the phase II randomised multicentre trial REMAGUS02 of neoadjuvant chemotherapy in breast cancer patients. BMC Cancer. 2011;11: 215. pmid:21631949
View Article
PubMed/NCBI
Google Scholar

[32] View Article

[33] PubMed/NCBI

[34] Google Scholar

[ref10] 10. Barretina J, Caponigro G, Stransky N, Venkatesan K, Margolin AA, Kim S, et al. NIH Public Access of anticancer drug sensitivity. 2012;483: 603–607.
View Article
Google Scholar

[36] View Article

[37] Google Scholar

[ref11] 11. Garnett MJ, Edelman EJ, Heidorn SJ, Greenman CD, Dastur A, Lau KW, et al. Europe PMC Funders Group Systematic identification of genomic markers of drug sensitivity in cancer cells. Nature. 2012;483: 570–575.
View Article
Google Scholar

[39] View Article

[40] Google Scholar

[ref12] 12. R Development Core Team. R: A Language and Environment for Statistical Computing. R Foundation for Sta- tistical Computing. Vienna, Austria 2009. [Internet]. http://www.R-project.org

[ref13] 13. Denkert C, von Minckwitz G, Brase JC, Sinn BV, Gade S, Kronenwett R, et al. Tumor-infiltrating lymphocytes and response to neoadjuvant chemotherapy with or without Carboplatin in human epidermal growth factor receptor 2-positive and triple-negative primary breast cancers. J Clin Oncol Off J Am Soc Clin Oncol. 2015;33: 983–991.
View Article
Google Scholar

[43] View Article

[44] Google Scholar

[ref14] 14. Gu-Trantien C, Loi S, Garaud S, Equeter C, Libin M, Wind De A, et al. CD4+ follicular helper T cell infiltration predicts breast cancer survival. J Clin Invest. 2013;123: 1–20.
View Article
Google Scholar

[46] View Article

[47] Google Scholar

[ref15] 15. Sota Y, Naoi Y, Tsunashima R, Kagara N, Shimazu K, Maruyama N, et al. Construction of novel immune-related signature for prediction of pathological complete response to neoadjuvant chemotherapy in human breast cancer. Ann Oncol Off J Eur Soc Med Oncol ESMO. 2014;25: 100–106.
View Article
Google Scholar

[49] View Article

[50] Google Scholar

[ref16] 16. Stoll G, Enot D, Mlecnik B, Galon J, Zitvogel L, Kroemer G. Immune-related gene signatures predict the outcome of neoadjuvant chemotherapy. Oncoimmunology. 2014;3: e27884. pmid:24790795
View Article
PubMed/NCBI
Google Scholar

[52] View Article

[53] PubMed/NCBI

[54] Google Scholar

[ref17] 17. Rody A, Karn T, Liedtke C, Pusztai L, Ruckhaeberle E, Hanker L, et al. A clinically relevant gene signature in triple negative and basal-like breast cancer. Breast Cancer Res BCR. 2011;13: R97. pmid:21978456
View Article
PubMed/NCBI
Google Scholar

[56] View Article

[57] PubMed/NCBI

[58] Google Scholar

[ref18] 18. Staaf J, Ringnér M, Vallon-Christersson J, Jönsson G, Bendahl P-O, Holm K, et al. Identification of subtypes in human epidermal growth factor receptor 2—positive breast cancer reveals a gene signature prognostic of outcome. J Clin Oncol Off J Am Soc Clin Oncol. 2010;28: 1813–1820.
View Article
Google Scholar

[60] View Article

[61] Google Scholar

[ref19] 19. Desmedt C, Haibe-Kains B, Wirapati P, Buyse M, Larsimont D, Bontempi G, et al. Biological processes associated with breast cancer clinical outcome depend on the molecular subtypes. Clin Cancer Res. 2008;14: 5158–5165. pmid:18698033
View Article
PubMed/NCBI
Google Scholar

[63] View Article

[64] PubMed/NCBI

[65] Google Scholar

[ref20] 20. Finak G, Bertos N, Pepin F, Sadekova S, Souleimanova M, Zhao H, et al. Stromal gene expression predicts clinical outcome in breast cancer. Nat Med. 2008;14: 518–527. pmid:18438415
View Article
PubMed/NCBI
Google Scholar

[67] View Article

[68] PubMed/NCBI

[69] Google Scholar

[ref21] 21. Alexe G, Dalgin GS, Scanfeld D, Tamayo P, Mesirov JP, DeLisi C, et al. High expression of lymphocyte-associated genes in node-negative HER2+ breast cancers correlates with lower recurrence rates. Cancer Res. 2007;67: 10669–10676. pmid:18006808
View Article
PubMed/NCBI
Google Scholar

[71] View Article

[72] PubMed/NCBI

[73] Google Scholar

[ref22] 22. Perez EA, Thompson EA, Ballman KV, Anderson SK, Asmann YW, Kalari KR, et al. Genomic Analysis Reveals That Immune Function Genes Are Strongly Linked to Clinical Outcome in the North Central Cancer Treatment Group N9831 Adjuvant Trastuzumab Trial. J Clin Oncol Off J Am Soc Clin Oncol. 2015;
View Article
Google Scholar

[75] View Article

[76] Google Scholar

[ref23] 23. Gatza ML, Lucas JE, Barry WT, Kim JW, Wang Q, Crawford MD, et al. A pathway-based classification of human breast cancer. Proc Natl Acad Sci U S A. 2010;107: 6994–6999. pmid:20335537
View Article
PubMed/NCBI
Google Scholar

[78] View Article

[79] PubMed/NCBI

[80] Google Scholar

[ref24] 24. Palmer C, Diehn M, Alizadeh A a, Brown PO. Cell-type specific gene expression profiles of leukocytes in human peripheral blood. BMC Genomics. 2006;7: 115. pmid:16704732
View Article
PubMed/NCBI
Google Scholar

[82] View Article

[83] PubMed/NCBI

[84] Google Scholar

[ref25] 25. Sabatier R, Finetti P, Mamessier E, Adelaide J, Chaffanet M, Ali HR, et al. Prognostic and predictive value of PDL1 expression in breast cancer. Oncotarget. 2015;6: 5449–5464. pmid:25669979
View Article
PubMed/NCBI
Google Scholar

[86] View Article

[87] PubMed/NCBI

[88] Google Scholar

[ref26] 26. Isella C, Terrasi A, Bellomo SE, Petti C, Galatola G, Muratore A, et al. Stromal contribution to the colorectal cancer transcriptome. Nat Genet. Nature Publishing Group; 2015; 1–11.
View Article
Google Scholar

[90] View Article

[91] Google Scholar

[ref27] 27. Lehmann BD, Bauer JA, Chen X, Sanders ME, Chakravarthy AB, Shyr Y, et al. Identification of human triple-negative breast cancer subtypes and preclinical models for selection of targeted therapies. J Clin Invest. 2011;121: 2750–2767. pmid:21633166
View Article
PubMed/NCBI
Google Scholar

[93] View Article

[94] PubMed/NCBI

[95] Google Scholar

[ref28] 28. Burstein MD, Tsimelzon A, Poage GM, Covington KR, Contreras A, Fuqua S, et al. Comprehensive Genomic Analysis Identifies Novel Subtypes and Targets of Triple-negative Breast Cancer. Clin Cancer Res Off J Am Assoc Cancer Res. 2014;
View Article
Google Scholar

[97] View Article

[98] Google Scholar

[ref29] 29. Doane AS, Danso M, Lal P, Donaton M, Zhang L, Hudis C, et al. An estrogen receptor-negative breast cancer subset characterized by a hormonally regulated transcriptional program and response to androgen. Oncogene. 2006;25: 3994–4008. pmid:16491124
View Article
PubMed/NCBI
Google Scholar

[100] View Article

[101] PubMed/NCBI

[102] Google Scholar

[ref30] 30. Farmer P, Bonnefoi H, Becette V, Tubiana-Hulin M, Fumoleau P, Larsimont D, et al. Identification of molecular apocrine breast tumours by microarray analysis. Oncogene. 2005;24: 4660–4671. pmid:15897907
View Article
PubMed/NCBI
Google Scholar

[104] View Article

[105] PubMed/NCBI

[106] Google Scholar

[ref31] 31. Lehmann-Che J, Hamy A-S, Porcher R, Barritault M, Bouhidel F, Habuellelah H, et al. Molecular apocrine breast cancers are aggressive estrogen receptor negative tumors overexpressing either HER2 or GCDFP15. Breast Cancer Res. 2013;15: R37. pmid:23663520
View Article
PubMed/NCBI
Google Scholar

[108] View Article

[109] PubMed/NCBI

[110] Google Scholar

[ref32] 32. Klein SL. The effects of hormones on sex differences in infection: from genes to behavior. Neurosci Biobehav Rev. 2000;24: 627–638. pmid:10940438
View Article
PubMed/NCBI
Google Scholar

[112] View Article

[113] PubMed/NCBI

[114] Google Scholar

[ref33] 33. Lockshin MD. Sex differences in autoimmune disease. Lupus. 2006;15: 753–756. pmid:17153846
View Article
PubMed/NCBI
Google Scholar

[116] View Article

[117] PubMed/NCBI

[118] Google Scholar

[ref34] 34. Cook IF. Sexual dimorphism of humoral immunity with human vaccines. Vaccine. 2008;26: 3551–3555. pmid:18524433
View Article
PubMed/NCBI
Google Scholar

[120] View Article

[121] PubMed/NCBI

[122] Google Scholar

[ref35] 35. Fish EN. The X-files in immunity: sex-based differences predispose immune responses. Nat Rev Immunol. 2008;8: 737–744. pmid:18728636
View Article
PubMed/NCBI
Google Scholar

[124] View Article

[125] PubMed/NCBI

[126] Google Scholar

[ref36] 36. Giefing-Kröll C, Berger P, Lepperdinger G, Grubeck-Loebenstein B. How sex and age affect immune responses, susceptibility to infections, and response to vaccination. Aging Cell. 2015;14: 309–321. pmid:25720438
View Article
PubMed/NCBI
Google Scholar

[128] View Article

[129] PubMed/NCBI

[130] Google Scholar

[ref37] 37. Mo R, Chen J, Grolleau-Julius A, Murphy HS, Richardson BC, Yung RL. Estrogen regulates CCR gene expression and function in T lymphocytes. J Immunol Baltim Md 1950. 2005;174: 6023–6029.
View Article
Google Scholar

[132] View Article

[133] Google Scholar

[ref38] 38. Sakiani S, Olsen NJ, Kovacs WJ. Gonadal steroids and humoral immunity. Nat Rev Endocrinol. 2013;9: 56–62. pmid:23183675
View Article
PubMed/NCBI
Google Scholar

[135] View Article

[136] PubMed/NCBI

[137] Google Scholar

[ref39] 39. Hao S, Zhao J, Zhou J, Zhao S, Hu Y, Hou Y. Modulation of 17beta-estradiol on the number and cytotoxicity of NK cells in vivo related to MCM and activating receptors. Int Immunopharmacol. 2007;7: 1765–1775. pmid:17996687
View Article
PubMed/NCBI
Google Scholar

[139] View Article

[140] PubMed/NCBI

[141] Google Scholar

[ref40] 40. Siracusa MC, Overstreet MG, Housseau F, Scott AL, Klein SL. 17beta-estradiol alters the activity of conventional and IFN-producing killer dendritic cells. J Immunol Baltim Md 1950. 2008;180: 1423–1431.
View Article
Google Scholar

[143] View Article

[144] Google Scholar

[ref41] 41. Arruvito L, Sanz M, Banham AH, Fainboim L. Expansion of CD4+CD25+and FOXP3+ regulatory T cells during the follicular phase of the menstrual cycle: implications for human reproduction. J Immunol Baltim Md 1950. 2007;178: 2572–2578.
View Article
Google Scholar

[146] View Article

[147] Google Scholar

[ref42] 42. Robert C, Thomas L, Bondarenko I, O’Day S, Weber J, Garbe C, et al. Ipilimumab plus dacarbazine for previously untreated metastatic melanoma. N Engl J Med. 2011;364: 2517–2526. pmid:21639810
View Article
PubMed/NCBI
Google Scholar

[149] View Article

[150] PubMed/NCBI

[151] Google Scholar

[ref43] 43. Brahmer J, Reckamp KL, Baas P, Crinò L, Eberhardt WEE, Poddubskaya E, et al. Nivolumab versus Docetaxel in Advanced Squamous-Cell Non-Small-Cell Lung Cancer. N Engl J Med. 2015;373: 123–135. pmid:26028407
View Article
PubMed/NCBI
Google Scholar

[153] View Article

[154] PubMed/NCBI

[155] Google Scholar

[ref44] 44. Giuliano M, Trivedi MV, Schiff R. Bidirectional Crosstalk between the Estrogen Receptor and Human Epidermal Growth Factor Receptor 2 Signaling Pathways in Breast Cancer: Molecular Basis and Clinical Implications. Breast Care Basel Switz. 2013;8: 256–262.
View Article
Google Scholar

[157] View Article

[158] Google Scholar

[ref45] 45. Sabatier R, Finetti P, Mamessier E, Raynaud S, Cervera N, Lambaudie E, et al. Kinome expression profiling and prognosis of basal breast cancers. Mol Cancer. 2011;10: 86. pmid:21777462
View Article
PubMed/NCBI
Google Scholar

[160] View Article

[161] PubMed/NCBI

[162] Google Scholar

[ref46] 46. Teschendorff AE, Miremadi A, Pinder SE, Ellis IO, Caldas C. An immune response gene expression module identifies a good prognosis subtype in estrogen receptor negative breast cancer. Genome Biol. 2007;8: R157. pmid:17683518
View Article
PubMed/NCBI
Google Scholar

[164] View Article

[165] PubMed/NCBI

[166] Google Scholar

[ref47] 47. Karn T, Pusztai L, Holtrich U, Iwamoto T, Shiang CY, Schmidt M, et al. Homogeneous datasets of triple negative breast cancers enable the identification of novel prognostic and predictive signatures. PloS One. 2011;6: e28403. pmid:22220191
View Article
PubMed/NCBI
Google Scholar

[168] View Article

[169] PubMed/NCBI

[170] Google Scholar

[ref48] 48. West NR, Milne K, Truong PT, Macpherson N, Nelson BH, Watson PH. Tumor-infiltrating lymphocytes predict response to anthracycline-based chemotherapy in estrogen receptor-negative breast cancer. Breast Cancer Res BCR. BioMed Central Ltd; 2011;13: R126.
View Article
Google Scholar

[172] View Article

[173] Google Scholar

[ref49] 49. Rody A, Holtrich U, Pusztai L, Liedtke C, Gaetje R, Ruckhaeberle E, et al. T-cell metagene predicts a favorable prognosis in estrogen receptor-negative and HER2-positive breast cancers. Breast Cancer Res BCR. 2009;11: R15. pmid:19272155
View Article
PubMed/NCBI
Google Scholar

[175] View Article

[176] PubMed/NCBI

[177] Google Scholar

[ref50] 50. Loi S, Michiels S, Salgado R, Sirtaine N, Jose V, Fumagalli D, et al. Tumor infiltrating lymphocytes are prognostic in triple negative breast cancer and predictive for trastuzumab benefit in early breast cancer: results from the FinHER trial. Ann Oncol Off J Eur Soc Med Oncol ESMO. 2014;25: 1544–1550.
View Article
Google Scholar

[179] View Article

[180] Google Scholar

[ref51] 51. Mahmoud SMA, Paish EC, Powe DG, Macmillan RD, Grainge MJ, Lee AHS, et al. Tumor-infiltrating CD8+ lymphocytes predict clinical outcome in breast cancer. J Clin Oncol Off J Am Soc Clin Oncol. 2011;29: 1949–1955.
View Article
Google Scholar

[182] View Article

[183] Google Scholar

[ref52] 52. Salgado R, Denkert C, Campbell C, Savas P, Nucifero P, Aura C, et al. Tumor-Infiltrating Lymphocytes and Associations With Pathological Complete Response and Event-Free Survival in HER2-Positive Early-Stage Breast Cancer Treated With Lapatinib and Trastuzumab: A Secondary Analysis of the NeoALTTO Trial. JAMA Oncol. 2015;1: 448–454. pmid:26181252
View Article
PubMed/NCBI
Google Scholar

[185] View Article

[186] PubMed/NCBI

[187] Google Scholar

[ref53] 53. Perez E. Stromal tumor-infiltrating lymphocytes(S-TILs): In the alliance N9831 trial S-TILs are associated with chemotherapy benefit but not associated with trastuzumab benefit [Internet]. 2014. http://www.abstracts2view.com/sabcs14/view.php?nu=SABCS13L_1455

[ref54] 54. Salgado R, Denkert C, Demaria S, Sirtaine N, Klauschen F, Pruneri G, et al. The evaluation of tumor-infiltrating lymphocytes (TILs) in breast cancer: recommendations by an International TILs Working Group 2014. Ann Oncol Off J Eur Soc Med Oncol ESMO. 2015;26: 259–271.
View Article
Google Scholar

[190] View Article

[191] Google Scholar

[ref55] 55. Stagg J, Loi S, Divisekera U, Ngiow SF, Duret H, Yagita H, et al. Anti-ErbB-2 mAb therapy requires type I and II interferons and synergizes with anti-PD-1 or anti-CD137 mAb therapy. Proc Natl Acad Sci U S A. 2011;108: 7142–7147. pmid:21482773
View Article
PubMed/NCBI
Google Scholar

[193] View Article

[194] PubMed/NCBI

[195] Google Scholar

[ref56] 56. Wang Q, Li S-H, Wang H, Xiao Y, Sahin O, Brady SW, et al. Concomitant targeting of tumor cells and induction of T-cell response synergizes to effectively inhibit trastuzumab-resistant breast cancer. Cancer Res. 2012;72: 4417–4428. pmid:22773664
View Article
PubMed/NCBI
Google Scholar

[197] View Article

[198] PubMed/NCBI

[199] Google Scholar

Figures

Abstract

Introduction

Methods

Results

Conclusion

Introduction

Material and methods

Data normalization and quality control

Training, validation and Ignatiadis datasets.

Determination and preprocessing of HER2-positive breast cancer samples

Gene selection process

Metagene construction

Association between expression of the Immunity metagene and that of ESR1, PGR, and AR

Analysis of the predicted response to NAC

Prognostic analysis

Correlation with tumor-infiltrating lymphocyte levels

Expression of the gene signature in human breast cancer cell lines

Statistical analysis

Results

HER2-positive gene expression profiles identify six main gene clusters

The expression of the Immunity metagene is strongly associated with ER status, PR, and AR status

Predictive value of the Immunity metagene in HER2-positive breast cancers

Prognostic value of the Immunity metagene in HER2-positive breast cancers

The Immunity metagene is correlated with tumor-infiltrating lymphocytes (TILs) in HER2-positive breast cancer

The Immunity metagene corresponds to the B-cell, T-cell and CD8 cell pathways

The Immunity metagene is probably expressed by stromal cells

Discussion

Conclusion

Supporting Information

S1 Fig. Methodology flow chart.

S2 Fig. Heatmaps of the selected genes in the HER2-positive datasets.

S3 Fig. pCR rates by breast cancer subtype and Immunity metagene.

S4 Fig. Heatmaps of the gene expression profiles of published immune signatures and connections between all immune genes.

S5 Fig. Distribution histograms for our Immune metagenes and immune pathways.

S6 Fig. Gene expression for the Immune metagenes and pathway, in cell lines and xenografts.

S1 File. Supplementary methods and results.

S1 Table. 138-gene signature.

S2 Table. Association of published immune signatures or metagenes with response to chemotherapy and prognosis.

Acknowledgments

Author Contributions

References