Proteogenomic characterization of difficult-to-treat breast cancer with tumor cells enriched through laser microdissection

BACKGROUND

Breast cancer (BC) is the most commonly diagnosed cancer and the leading cause of cancer death among women globally. Despite advances, there is considerable variation in clinical outcomes for patients with non-luminal A tumors, classified as difficult-to-treat breast cancers (DTBC). This study aims to delineate the proteogenomic landscape of DTBC tumors compared to luminal A (LumA) tumors.

METHODS

We retrospectively collected a total of 117 untreated primary breast tumor specimens, focusing on DTBC subtypes. Breast tumors were processed by laser microdissection (LMD) to enrich tumor cells. DNA, RNA, and protein were simultaneously extracted from each tumor preparation, followed by whole genome sequencing, paired-end RNA sequencing, global proteomics and phosphoproteomics. Differential feature analysis, pathway analysis and survival analysis were performed to better understand DTBC and investigate biomarkers.

RESULTS

We observed distinct variations in gene mutations, structural variations, and chromosomal alterations between DTBC and LumA breast tumors. DTBC tumors predominantly had more mutations in TP53, PLXNB3, Zinc finger genes, and fewer mutations in SDC2, CDH1, PIK3CA, SVIL, and PTEN. Notably, Cytoband 1q21, which contains numerous cell proliferation-related genes, was significantly amplified in the DTBC tumors. LMD successfully minimized stromal components and increased RNA-protein concordance, as evidenced by stromal score comparisons and proteomic analysis. Distinct DTBC and LumA-enriched clusters were observed by proteomic and phosphoproteomic clustering analysis, some with survival differences. Phosphoproteomics identified two distinct phosphoproteomic profiles for high relapse-risk and low relapse-risk basal-like tumors, involving several genes known to be associated with breast cancer oncogenesis and progression, including KIAA1522, DCK, FOXO3, MYO9B, ARID1A, EPRS, ZC3HAV1, and RBM14. Lastly, an integrated pathway analysis of multi-omics data highlighted a robust enrichment of proliferation pathways in DTBC tumors.

CONCLUSIONS

This study provides an integrated proteogenomic characterization of DTBC vs LumA with tumor cells enriched through laser microdissection. We identified many common features of DTBC tumors and the phosphopeptides that could serve as potential biomarkers for high/low relapse-risk basal-like BC and possibly guide treatment selections.

Quantile Index Biomarkers Based on Single-Cell Expression Data

Current histocytometry methods enable single-cell quantification of biomolecules in tumor tissue sections by multiple detection technologies, including multiplex fluorescence-based immunohistochemistry or in situ hybridization. Quantitative pathology platforms can provide distributions of cellular signal intensity (CSI) levels of biomolecules across the entire cell populations of interest within the sampled tumor tissue. However, the heterogeneity of CSI levels is usually ignored, and the simple mean signal intensity value is considered a cancer biomarker. Here we consider the entire distribution of CSI expression levels of a given biomolecule in the cancer cell population as a predictor of clinical outcome. The proposed quantile index (QI) biomarker is defined as the weighted average of CSI distribution quantiles in individual tumors. The weight for each quantile is determined by fitting a functional regression model for a clinical outcome. That is, the weights are optimized so that the resulting QI has the highest power to predict a relevant clinical outcome. The proposed QI biomarkers were derived for proteins expressed in cancer cells of malignant breast tumors and demonstrated improved prognostic value compared with the standard mean signal intensity predictors. The R package Qindex implementing QI biomarkers has been developed. The proposed approach is not limited to immunohistochemistry data and can be based on any cell-level expressions of proteins or nucleic acids.

Selection of optimal quantile protein biomarkers based on cell-level immunohistochemistry data

BACKGROUND

Protein biomarkers of cancer progression and response to therapy are increasingly important for improving personalized medicine. Advanced quantitative pathology platforms enable measurement of protein expression in tissues at the single-cell level. However, this rich quantitative cell-by-cell biomarker information is most often not exploited. Instead, it is reduced to a single mean across the cells of interest or converted into a simple proportion of binary biomarker-positive or -negative cells.

RESULTS

We investigated the utility of retaining all quantitative information at the single-cell level by considering the values of the quantile function (inverse of the cumulative distribution function) estimated from a sample of cell signal intensity levels in a tumor tissue. An algorithm was developed for selecting optimal cutoffs for dichotomizing cell signal intensity distribution quantiles as predictors of continuous, categorical or survival outcomes. The proposed algorithm was used to select optimal quantile biomarkers of breast cancer progression based on cancer cells' cell signal intensity levels of nuclear protein Ki-67, Proliferating cell nuclear antigen, Programmed cell death 1 ligand 2, and Progesterone receptor. The performance of the resulting optimal quantile biomarkers was validated and compared to the standard cancer compartment mean signal intensity markers using an independent external validation cohort. For Ki-67, the optimal quantile biomarker was also compared to established biomarkers based on percentages of Ki67-positive cells. For proteins significantly associated with PFS in the external validation cohort, the optimal quantile biomarkers yielded either larger or similar effect size (hazard ratio for progression-free survival) as compared to cancer compartment mean signal intensity biomarkers.

CONCLUSION

The optimal quantile protein biomarkers yield generally improved prognostic value as compared to the standard protein expression markers. The proposed methodology has a broad application to single-cell data from genomics, transcriptomics, proteomics, or metabolomics studies at the single cell level.

Abstract P2-11-13: High PD-L2 Protein Expression in Cancer Cells is an Independent Marker of Unfavorable Prognosis in Luminal Breast Tumors

Background PD-1 inhibitors have shown significant efficacy in triple negative breast cancer (BC), however, durable responses are less common in estrogen receptor-positive (ER+) BC. Better markers are therefore needed that will identify likely responders to PD-1 inhibitors among patients with luminal BC. While most efforts have focused on the immune checkpoint protein PD-L1, the alternative PD-1 ligand, PD-L2, has been largely overlooked. We aimed to determine if PD-L2 is associated with unfavorable prognosis in ER+ BC. Methods PD-L2 protein levels in cancer cells and stromal cells were measured retrospectively by quantitative immunofluorescence histocytometry in tissue microarrays of therapy-naïve, localized or locoregional ER+ BC and correlated with progression-free survival (PFS). Evaluable tumor PD-L2 data were derived from a main study cohort A diagnosed between 1988-2005 (n=684) with extensive clinical and outcome data and from an independent validation cohort B diagnosed between 1992-2012 (n=273). Patients received standard-of-care adjuvant therapy without immune checkpoint inhibitors after tumor resection. Results Univariate analysis of the main cohort A revealed that high PD-L2 expression in cancer cells was associated with shorter PFS (HR=1.8; 95%CI:1.3-2.6; p=0.001), an observation that was validated in an independent cohort B (HR=2.3, 95%CI:1.1-4.8; p=0.026). Approximately one third of ER+ BC cases were classified as high PD-L2. After multivariable adjustment for common clinicopathological variables, high cancer cell levels of PD-L2 remained independently predictive of early recurrence (HR=2.0; 95%CI:1.4-2.9; p< 0.001). Sub-analysis of ER+ BC cases treated with adjuvant chemotherapy (n=197) suggested that high PD-L2 levels in cancer cells was associated with particularly increased risk of progression (multivariable HR=3.4; 95%CI:1.9-6.2; p< 0.001). The observed frequent expression of PD-L2 protein in BC provided scientific rationale for the design of our ongoing phase II clinical trial (NCT04243616) of neoadjuvant combined PD-1 inhibitor (cemiplimab; Regeneron Pharmaceuticals Inc) and chemotherapy in patients diagnosed with PD-L1+ and/or PD-L2+ BC. The primary objective of this trial is to assess pathologic responses to neoadjuvant treatment with secondary objective of assessing the correlation between PD-L1/PD-L2 status and tumor responses. Pathologist review of PD-L1 and PD-L2 expression in an initial set of ER+ tumors (n=15) screened for trial eligibility revealed frequent discordance between cancer cell positivity for PD-L1 and PD-L2 protein (PD-L1: median=0%; range 0-2% vs. PD-L2: median=18%; range < 1-60%) as well as immune cell positivity (PD-L1: median=5%; range 0-50% vs. PD-L2: median=0; range 0-50%). Conclusions In treatment-naïve ER+ breast tumors, high cancer cell expression of PD-L2 protein was an independent predictor of poor clinical outcome, with evidence of further elevated risk of progression in patients who received adjuvant chemotherapy. Preliminary analyses of ER+ tumors from our ongoing clinical trial showed frequent discordance between baseline PD-L1 and PD-L2 protein expression in both cancer cells and immune cells. Collectively, our analyses indicate that PD-L2 has prognostic value for ER+ BC, and our progress justify further studies to determine whether PD-L2, alone or in combination with PD-L1, may serve as a predictive marker of response to PD-1 inhibitors.

Citation Format: Lubna N. Chaudhary, Inna Chervoneva, Amy R. Peck, Yunguang Sun, Misung Yi, John F. Langenheim, Julie M. Jorns, Sailaja Kamaraju, Yee Chung Cheng, John Burfeind, Christopher R. Chitambar, Jeffrey A. Hooke, Albert J. Kovatich, Craig Shriver, Hai Hu, Juan P. Palazzo, Marluce Bibbo, Terry Hyslop, Richard Pestell, Edith P. Mitchell, Hallgeir Rui. High PD-L2 Protein Expression in Cancer Cells is an Independent Marker of Unfavorable Prognosis in Luminal Breast Tumors [abstract]. In: Proceedings of the 2022 San Antonio Breast Cancer Symposium; 2022 Dec 6-10; San Antonio, TX. Philadelphia (PA): AACR; Cancer Res 2023;83(5 Suppl):Abstract nr P2-11-13.

High PD-L2 Predicts Early Recurrence of ER-Positive Breast Cancer

PURPOSE

T-cell-mediated cytotoxicity is suppressed when programmed cell death-1 (PD-1) is bound by PD-1 ligand-1 (PD-L1) or PD-L2. Although PD-1 inhibitors have been approved for triple-negative breast cancer, the lower response rates of 25%-30% in estrogen receptor-positive (ER+) breast cancer will require markers to identify likely responders. The focus of this study was to evaluate whether PD-L2, which has higher affinity than PD-L1 for PD-1, is a predictor of early recurrence in ER+ breast cancer.

METHODS

PD-L2 protein levels in cancer cells and stromal cells of therapy-naive, localized or locoregional ER+ breast cancers were measured retrospectively by quantitative immunofluorescence histocytometry and correlated with progression-free survival (PFS) in the main study cohort (n = 684) and in an independent validation cohort (n = 273). All patients subsequently received standard-of-care adjuvant therapy without immune checkpoint inhibitors.

RESULTS

Univariate analysis of the main cohort revealed that high PD-L2 expression in cancer cells was associated with shorter PFS (hazard ratio [HR], 1.8; 95% CI, 1.3 to 2.6; P = .001), which was validated in an independent cohort (HR, 2.3; 95% CI, 1.1 to 4.8; P = .026) and remained independently predictive after multivariable adjustment for common clinicopathological variables (HR, 2.0; 95% CI, 1.4 to 2.9; P < .001). Subanalysis of the ER+ breast cancer patients treated with adjuvant chemotherapy (n = 197) revealed that high PD-L2 levels in cancer cells associated with short PFS in univariate (HR, 2.5; 95% CI, 1.4 to 4.4; P = .003) and multivariable analyses (HR, 3.4; 95% CI, 1.9 to 6.2; P < .001).

CONCLUSION

Up to one third of treatment-naive ER+ breast tumors expressed high PD-L2 levels, which independently predicted poor clinical outcome, with evidence of further elevated risk of progression in patients who received adjuvant chemotherapy. Collectively, these data warrant studies to gain a deeper understanding of PD-L2 in the progression of ER+ breast cancer and may provide rationale for immune checkpoint blockade for this patient group.

Abstract P1-05-04: Novel breast cancer proteomic subtyping with connection to cell of origin

Introduction: In clinical practice, immunohistochemistry (IHC) is combined with clinicopathologic information to stratify patients into subtypes and guide treatment decisions. In our mass spectrometry study of 116 breast cancers (BrCAs), we identified 34 significant proteins which provide additional prognostic information. This novel proteomic subtyping, combining IHC with proteomics, separates HER2 negative (Her2-) BrCA patients into Luminal-like and TN-like subtypes. A novel subtype, luminal by IHC and TN-like by proteomics (L/T), is associated with unfavorable outcome compared with the luminal subtype (L/L) by both IHC and proteomics. These IHC-based Her2- subtypes are described as L/L, L/T, T/L and T/T subtypes. We investigated the cell-of-origin of our enhanced subtypes using gene set enrichment analysis with gene sets consisting of breast cancer cell-of-origin signatures. We further identified genes representing our 34 proteins in the significant gene sets and investigated their association with survival outcomes across The Cancer Genome Atlas (TCGA) pan-cancers. Method: The single-cell RNA-sequencing data analyses of health breast samples identified 23 breast cell-of-origin signatures (Bhat-Nakshatri et al., 2021) with 46 up and down-regulated gene sets based on the provided fold changes. Using our cohort, we performed differential analysis for each enhanced subtype (L/L, L/T, T/T) versus the other subtypes separately. Pre-ranked gene set enrichment analyses were used to identify significantly enriched gene sets. The significance of enriched gene sets was reported at FDR&lt;0.05. We further identified genes of 34 biomarkers that overlapped in the identified significant gene sets and investigated the impact of the genes on survival across TCGA pan-cancers. Univariate overall survival (OS) and progression-free interval (PFI) analyses were performed on each cancer cohort and the significance of the association with survival outcome was reported by log-rank p-value &lt;0.05. Results: Our signature was compared to the clusters from healthy tissues. We found up-regulated genes in L/L subtype were enriched in gene sets elevated in mature luminal annotated clusters (N8 and N12). Down-regulated genes in L/L subtype were enriched in the elevated genes in the N3 (luminal progenitor) and N9 (luminal progenitor/basal) clusters. Up-regulated genes in L/T subtype were enriched in the elevated genes in the N19 (luminal progenitor) and N11 (basal) clusters. Down-regulated genes in L/T were enriched in the elevated genes in the N8 (mature luminal) cluster and down-regulated genes in the N9 (luminal progenitor/basal) cluster. Up-regulated genes in T/T subtype were enriched in genes highly expressed in the luminal progenitor (N3) and luminal progenitor/basal (N9) clusters. Down-regulated genes in T/T subtype were enriched in the elevated genes in the mature luminal annotated clusters (N8, N12). One gene, KPNA2, was up-regulated in both the N9 (luminal progenitor/basal) cluster and the Basal-like (TN-like) subtype. The higher expression of this gene had significant association with worse OS outcomes in most of the cancers. Conclusions: T/T and L/T subtypes likely arise from progenitor and basal cells. The L/L subtype may arise from mature luminal cells. The novel L/T subtype appears to have characteristics that result in poorer survival. Disclaimers The contents of this publication are the sole responsibility of the author(s) and do not necessarily reflect the views, opinions or policies of USUHS, HJF, the DOD or the Departments of the Army, Navy, or Air Force. Mention of trade names, commercial products, or organizations does not imply endorsement by the U.S. Government.

Citation Format: Guisong Wang, Punit Shah, Rick Searfoss, J. Leigh Fantacone-Campbell, Jeffrey A. Hooke, Brenda Deyarmin, Rebecca N. Zingmark, Stella Somiari, Jianfang Liu, Leonid Kvecher, Lori A. Sturtz, Praveen-Kumar Raj-Kumar, Elder Granger, Linda Vahdat, Niven R. Narain, Mary L. Cutler, Rangaprasad Sarangarajan, Hai Hu, Michael A. Kiebish, Albert J. Kovatich, Craig D. Shriver. Novel breast cancer proteomic subtyping with connection to cell of origin [abstract]. In: Proceedings of the 2021 San Antonio Breast Cancer Symposium; 2021 Dec 7-10; San Antonio, TX. Philadelphia (PA): AACR; Cancer Res 2022;82(4 Suppl):Abstract nr P1-05-04.

Spatial Metrics of Interaction between CD163-Positive Macrophages and Cancer Cells and Progression-Free Survival in Chemo-Treated Breast Cancer

Tumor-associated macrophages (TAMs) promote progression of breast cancer and other solid malignancies via immunosuppressive, pro-angiogenic and pro-metastatic effects. Tumor-promoting TAMs tend to express M2-like macrophage markers, including CD163. Histopathological assessments suggest that the density of CD163-positive TAMs within the tumor microenvironment is associated with reduced efficacy of chemotherapy and unfavorable prognosis. However, previous analyses have required research-oriented pathologists to visually enumerate CD163+ TAMs, which is both laborious and subjective and hampers clinical implementation. Objective, operator-independent image analysis methods to quantify TAM-associated information are needed. In addition, since M2-like TAMs exert local effects on cancer cells through direct juxtacrine cell-to-cell interactions, paracrine signaling, and metabolic factors, we hypothesized that spatial metrics of adjacency of M2-like TAMs to breast cancer cells will have further information value. Immunofluorescence histo-cytometry of CD163+ TAMs was performed retrospectively on tumor microarrays of 443 cases of invasive breast cancer from patients who subsequently received adjuvant chemotherapy. An objective and automated algorithm was developed to phenotype CD163+ TAMs and calculate their density within the tumor stroma and derive several spatial metrics of interaction with cancer cells. Shorter progression-free survival was associated with a high density of CD163+ TAMs, shorter median cancer-to-CD163+ nearest neighbor distance, and a high number of either directly adjacent CD163+ TAMs (within juxtacrine proximity <12 μm to cancer cells) or communicating CD163+ TAMs (within paracrine communication distance <250 μm to cancer cells) after multivariable adjustment for clinical and pathological risk factors and correction for optimistic bias due to dichotomization.

Abstract 2342: Multidimensional metabolomic stratification of ER+/HER2- compared to ER-/HER2- breast tumors

Introduction: In the United States, breast cancer represents the leading cancer diagnosis among women and can readily be classified as a metabolic disease based on its distinct metabolic activity within the tumor microenvironment. Compared to other omics technologies, extensive lipidomic and metabolomic studies are lacking. Here in, we evaluated a cohort of 109 tumors characterized as ER+/HER2- and ER-/HER2- based on immunohistochemistry (IHC) and performed comprehensive structural lipidomic, signaling lipidomic, and global metabolomic analyses for an extensive characterization of the biophysical, signaling, and metabolic interplay between these tumors.

Methods: Clinical IHC subtyping of core biopsies was used to select a cohort of patients with ER+/HER2- or ER-/HER2- primary tumors from flash-frozen surgical samples. The positive/negative status of ER/PR/HER2 was defined using updated ASCO 2020 guidelines. Ki-67 status was determined using the 2011 St. Gallen's International Expert Consensus recommendations. ER low (1-10%) cases were excluded from this analysis. Structural lipidomic analysis was employed through the use of MS/MSALL high resolution shotgun lipidomics using a SCIEX 5600+ TripleTOF micro LC approach characterizing 23 lipid classes and over 1200 molecular species. Signaling lipids were analyzed using a SCIEX 6600 TripleTOF microLC platform characterizing 106 lipid analytes across octadecanoid, eicosanoid and docosanoid species. Metabolomics analysis was performed using LECO PEGASUS GC TOF, SCIEX 5500 HILIC LC MS/MS analysis, and SCIEX 6600 High resolution RP-LC-MS analysis detecting 450 metabolite Metabolomics data was further interpreted using MetaboAnalyst software.

Results/Conclusions: Compared to their ER+ counterparts, ER-/HER2- tumors exhibited a significant decrease in triacylglycerides, and a corresponding increase in cholesterol ester, phosphatidylcholine, phosphatidylethanolamine, and phosphatidylglycerol species demonstrating a signature of biophysical and metabolic rewiring with alterations in Kennedy pathway lipid shuttling. One signaling lipid was decreased and six were increased (predominantly arachidonic species) in ER-/HER2- tumors compared to ER+/HER2- ones. Metabolomic analysis revealed distinct alterations in cysteine/methionine, arginine/proline, purine, butanoate, and tryptophan metabolism. Utilizing a multidimensional metabolic integration approach, we identified distinct biophysical, signaling, and biochemical alterations in ER+/HER2- compared to ER-/HER2- breast tumors, which may impact selection of therapy and outcome in the future.

Citation Format: Valerie Bussberg, Vladimir Tolstikov, Guisong Wang, Punit Shah, Rick Searfoss, Leigh Fantacone-Campbell, Jeffrey A. Hooke, Brenda Deyarmin, Rebecca N. Zingmark, Stella Somiari, Jianfang Liu, Leonid Kvecher, Bradley Mostoller, Lori Sturtz, Praveen-Kumar Raj-Kumar, Elder Granger, Linda Vahdat, Mary L. Cutler, Chas Bountra, Rangaprasad Sarangarajan, Hai Hu, Albert J. Kovatich, Michael A. Kiebish, Niven R. Narain, Craig D. Shriver. Multidimensional metabolomic stratification of ER+/HER2- compared to ER-/HER2- breast tumors [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2021; 2021 Apr 10-15 and May 17-21. Philadelphia (PA): AACR; Cancer Res 2021;81(13_Suppl):Abstract nr 2342.

Abstract 1188: Reclassification of ER+ (luminal A/luminal B1 minus ER low)-like and ER- like breast tumors based on proteomic/gene and clinical outcome signatures

Introduction: Classification of breast cancer can incorporate immunohistochemical (IHC) detection of ER/PR/HER2/KI67 to stratify the subtypes. High throughput proteomics analysis allows for the expansion of biomarker discovery within the subtypes. We evaluated a cohort of 109 tumors characterized as ER+ (Luminal A and Luminal B1; HER2+ and ER low (1-10%) cases were excluded) compared to ER-/HER2- tumors. Utilizing an integrated bioinformatics approach, we developed a proteomic marker signature to reclassify tumors into ER+(like) and ER-(like) tumors. CPTAC (Proteomic)/TCGA (RNAseq) datasets and larger METBRIC and GSE96058 cohorts were used to validate this marker signature. The selected biomarkers demonstrated significant differences impacting survival outcome.

Methods: Clinical IHC subtyping of core biopsies was used to select a cohort of patients with ER+/HER2- and ER-/HER2- primary tumors from flash-frozen surgical samples. The positive/negative status of ER/PR/HER2 was defined using updated ASCO 2020 guidelines. Ki-67 status was determined using the 2011 St. Gallen's International Expert Consensus recommendations. Proteomic analysis was performed using Thermo Q-Exactive+ LC MS/MS analysis. Differential analysis was applied to select the significantly altered proteins between ER+ and ER- cases, Univariate survival analysis was engaged to filter informative protein/genes using TCGA RNA-Seq data. Nearest centroid analysis was deployed to define the classifier to predict novel molecular subtypes.

Results/Conclusions: We selected 34 proteins/genes from 164 significantly differentially expressed proteins for further analysis. The centroid model constructed with the 34 proteins defined 2 groups: ER+(like) and ER-(like). An additional 4 groups were defined across subtypes: luminal tumors classified both by IHC and marker signature (LL), luminal tumors classified by IHC but marker signature more like triple negative (LT), triple negative tumors classified by IHC but marker signature more like luminal (TL), and triple negative classified by both IHC and marker signature (TT). This marker signature segregated close to 5000 tumors across CPTAC, TCGA, METABRIC and GSE96058 cohorts. Survival analysis in these groups of patients revealed differences in radiation, hormone/radiation, hormone therapy, and hormone/radiation/chemotherapy treatments. In summary using proteomics data we identified a 34 gene/protein marker signature, validated in large external cohorts and exhibited impact on survival and response to therapy. Further, this signature was enriched in metabolism and microenvironmental associated factors that could represent novel targets or development combination strategies based on this signature.

Citation Format: Guisong Wang, Punit Shah, Rick Searfoss, Leigh Fantacone-Campbell, Jeffrey A. Hooke, Brenda Deyarmin, Rebecca N. Zingmark, Stella Somiari, Jianfang Liu, Leonid Kvecher, Bradley Mostoller, Lori A. Sturtz, Praven-Kumar Raj-Kumar, Elder Granger, Linda Vahdat, Mary L. Cutler, Chas Bountra, Rangaprasad Sarangarajan, Hai Hu, Michael A. Kiebish, Albert J. Kovatich, Niven R. Narain, Craig D. Shriver. Reclassification of ER+ (luminal A/luminal B1 minus ER low)-like and ER- like breast tumors based on proteomic/gene and clinical outcome signatures [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2021; 2021 Apr 10-15 and May 17-21. Philadelphia (PA): AACR; Cancer Res 2021;81(13_Suppl):Abstract nr 1188.

Abstract PS5-34: Identification of proteomics-based biomarkers for ER+/HER2- breast cancer stratification: Implications on clinical outcome

Introduction: Improving stratification of breast cancer (BC) patients based on molecular signatures for treatment responses and clinical outcomes is a critical unmet need. Currently, endocrine therapy is first-line for hormone receptor positive (HR+) BC. Chemotherapy is added to patients with high-risk luminal BC. In practice, levels of Ki-67 are used to distinguish luminal tumors as low-risk luminal A (LA) and high-risk luminal B (LB) for adjuvant therapy decisions. Herein, proteomic tumor assessment from ER-positive HER2-negative (ER+/HER2-) BC patients was utilized to define molecular subtyping, estimate congruency between proteomic subtyping and traditionally used Ki67 marker, and define a new set of potential predictive and prognostic therapeutic biomarkers for ER+/HER2- BC patients.

Method/Result: Clinical immunohistochemistry (IHC) subtyping of core biopsies was used to select a cohort of 86 BC patients with ER+/HER2- primary tumors from flash-frozen surgical samples. The positive/negative status of ER/PR/HER2 was defined using updated ASCO 2020 guidelines. Ki-67 status was determined using the 2011 St. Gallen’s International Expert Consensus recommendations. The cohort includes 28 LA (Ki67 &lt; 14%) cases and 58 LB1 (Ki67 &gt;= 14%) cases. Integrated consensus clustering algorithms with the most varying proteins in our cohort were applied to identify proteomic subtypes. Two distinct separations were observed from the analysis, resulting in one cluster enriched with LA (40 cases) and the other enriched with LB1 (46 cases) called by Fisher’s exact test. These clusters matched 100% with the clusters generated using 900+ proteins common to the 1500+ proteins used in the CPTAC-BC proteomics-based subtyping analysis (Mertins et al. Nature 2016). The differential analyses demonstrated that there is no significant difference between Ki67-defined subtypes and proteomics-defined subtypes (LA-enriched vs. LA cases, LB1-enriched vs. LB1 cases),indicating they are consistent in the molecular profile. Differential analysis was performed to compare LB1-enriched versus LA-enriched cases, resulting in 672 significantly differentially expressed proteins defined at false discovery rate (FDR) &lt; 0.05 and |log2(fold change)|&gt;1. 353 of the 672 proteins were correlated with mRNA at Pearson correlation &gt; 0.39 as reported in the CPTAC-BC study or cBioPortal for Cancer Genome, and their coding genes were used for progression free interval (PFI) analysis based on TCGA RNA-seq data in the TCGA ER+/HER2- cases (662 cases, c.f. Huo et al. JAMA Oncology 2017). 90 of the 353 coding genes significantly associated with PFI were detected at p-value&lt;0.05. Unsupervised hierarchical clustering method and principal component analysis (PCA) of the 90 genes were applied to our cohort to investigate the clustering performance and 94.2% of the cases were clustered correctly using support vector machine (SVM) method after PCA analysis. Biological process and molecular function GO term over-representation analyses of the 90 coding genes were performed separately. Some significant and biologically meaningful GO terms were identified at FDR&lt;0.05.

Conclusions: We identified a set of biomarkers that can be potentially employed as proteomic or gene signatures to stratify ER+/HER2- BC into low risk and high-risk groups.

Disclaimers The contents of this publication are the sole responsibility of the author(s) and do not necessarily reflect the views, opinions, or policies of Uniformed Services University of the Health Sciences, The Henry M Jackson Foundation for the Advancement of Military Medicine Inc., the Department of Defense or the Departments of the Army, Navy or Air Force. Mention of trade names, commercial products, or organizations does not imply endorsement by the U.S. Government.

Citation Format: Guisong Wang, Punit Shah, Rick Searfoss, J. Leigh Fantacone Campbell, Jeffrey A. Hooke, Brenda Deyarmin, Rebecca N. Zingmark, Stella Somiari, Jianfang Liu, Leonid Kvecher, Lori A. Sturtz, Praveen-Kumar Raj-Kumar, Elder Granger, Linda Vahdat, Mary L. Cutler, Rangaprasad Sarangarajan, Hai Hu, Michael A. Kiebish, Albert J. Kovatich, Niven R. Narain, Craig D. Shriver. Identification of proteomics-based biomarkers for ER+/HER2- breast cancer stratification: Implications on clinical outcome [abstract]. In: Proceedings of the 2020 San Antonio Breast Cancer Virtual Symposium; 2020 Dec 8-11; San Antonio, TX. Philadelphia (PA): AACR; Cancer Res 2021;81(4 Suppl):Abstract nr PS5-34.

Abstract PS18-38: Comparative analysis of differentially abundant proteins quantified by LC-MS/MS between flash frozen and laser microdissected OCT-embedded breast tumor samples

Background: Proteomic studies are typically conducted using flash-frozen (FF) samples utilizing tandem mass spectrometry. However, FF samples are comprised of multiple cell types, making it difficult to ascertain the proteomic profiles of specific cells. Conversely, OCT-embedded (Optimal Cutting Temperature compound) specimens can undergo laser microdissection (LMD) to capture and study specific cell types separately from the cell mixture. In the current study, we compared proteomic data obtained from FF and OCT samples to determine if samples that are stored and processed differently produce comparable results. Methods: Proteins were extracted from FF and OCT-embedded invasive breast tumors from 5 female patients. FF samples were lysed via homogenization (FF/HOM) while OCT-embedded specimens underwent LMD to collect only tumor cells (OCT/LMD-T) or both tumor and stromal cells (OCT/LMD-TS) followed by incubation at 37°C. Proteins were extracted using the illustra triplePrep kit and then trypsin-digested, TMT-labeled, and processed by two-dimensional liquid chromatography-tandem mass spectrometry (2D LC-MS/MS). Proteins were identified and quantified with Proteome Discoverer v1.4 and comparative analyses performed to identify proteins that were significantly differentially expressed amongst the different processing methods. Results: Among 4,950 proteins consistently quantified across all samples, 216 and 171 proteins were significantly differentially expressed (adjusted p-value &lt; 0.05; |log2 FC| &gt; 1) between FF/HOM vs. OCT/LMD-T and FF/HOM vs. OCT/LMD-TS, respectively, with most proteins being more highly abundant in the FF/HOM samples. PCA and unsupervised hierarchical clustering analysis with these 216 and 171 proteins were able to distinguish FF/HOM from OCT/LMD-T and OCT/LMD-TS samples, respectively. Likewise, PCA analysis and unsupervised clustering analysis using the 402 and 60 significantly differentially enriched GO terms (adjusted p-value (BH) &lt; 0.2) in the FF/HOM vs. OCT/LMD-T and FF/HOM vs OCT/LMD-TS comparisons, respectively, not only distinguished OCT/LMD from FF/HOM samples but also separated LA and LB1 breast cancer subtypes within each storage/preparation method from one another. Although FF/HOM appears to be more similar to OCT/LMD-TS than OCT/LMD-T based on the number of differentially enriched proteins (216 vs. 171; p=0.022) and GO terms (402 vs. 60; p &lt; 2.2 x 10-16), FF/HOM shows no greater similarity to OCT/LMD-TS than OCT/LMD-T based on PCA analysis with either proteins or GO terms ( based on weighted distance for pairwise samples, p = 0.97 from paired t-test). No significantly differentially enriched proteins or GO terms were detected between the OCT/LMD-T and OCT/LMD-TS samples but trended differences were detected. Conclusions: The proteomic profiles of the OCT/LMD-TS samples were more similar to those from OCT/LMD-T samples than FF/HOM samples, suggesting a strong influence from the sample processing methods. These results indicate that in LC-MS/MS proteomic studies, FF/HOM samples exhibit different protein profiles from OCT/LMD samples and thus, results from these two different methods cannot be directly compared. Our study also provides preliminary data for designing new studies to explore why OCT/LMD-TS samples are more similar to OCT/LMD-T than to FF/HOM samples, and to separate LA from LB1 samples. Disclaimer: The contents of this publication are the sole responsibility of the author(s) and do not necessarily reflect the views, opinions or policies of USUHS, HJF, the DOD or the Departments of the Army, Navy or Air Force. Mention of trade names, commercial products, or organizations does not imply endorsement by the U.S. Government.

Citation Format: Lori A. Sturtz, Guisong Wang, Punit Shah, Richard Searfoss, Praveen-Kumar Raj-Kumar, Jeffrey A. Hooke, J. Leigh Fantacone-Campbell, Brenda Deyarmin, Mary Lou Cutler, Rangaprasad Sarangarajan, Niven R. Narain, Hai Hu, Michael A. Kiebish, Albert J. Kovatich, Craig D. Shriver. Comparative analysis of differentially abundant proteins quantified by LC-MS/MS between flash frozen and laser microdissected OCT-embedded breast tumor samples [abstract]. In: Proceedings of the 2020 San Antonio Breast Cancer Virtual Symposium; 2020 Dec 8-11; San Antonio, TX. Philadelphia (PA): AACR; Cancer Res 2021;81(4 Suppl):Abstract nr PS18-38.

Comparative analysis of differentially abundant proteins quantified by LC-MS/MS between flash frozen and laser microdissected OCT-embedded breast tumor samples

Background

Proteomic studies are typically conducted using flash-frozen (FF) samples utilizing tandem mass spectrometry (MS). However, FF specimens are comprised of multiple cell types, making it difficult to ascertain the proteomic profiles of specific cells. Conversely, OCT-embedded (Optimal Cutting Temperature compound) specimens can undergo laser microdissection (LMD) to capture and study specific cell types separately from the cell mixture. In the current study, we compared proteomic data obtained from FF and OCT samples to determine if samples that are stored and processed differently produce comparable results.

Methods

Proteins were extracted from FF and OCT-embedded invasive breast tumors from 5 female patients. FF specimens were lysed via homogenization (FF/HOM) while OCT-embedded specimens underwent LMD to collect only tumor cells (OCT/LMD-T) or both tumor and stromal cells (OCT/LMD-TS) followed by incubation at 37 °C. Proteins were extracted using the illustra triplePrep kit and then trypsin-digested, TMT-labeled, and processed by two-dimensional liquid chromatography-tandem mass spectrometry (2D LC–MS/MS). Proteins were identified and quantified with Proteome Discoverer v1.4 and comparative analyses performed to identify proteins that were significantly differentially expressed amongst the different processing methods.

Results

Among the 4,950 proteins consistently quantified across all samples, 216 and 171 proteins were significantly differentially expressed (adjusted p-value < 0.05; |log₂ FC|> 1) between FF/HOM vs. OCT/LMD-T and FF/HOM vs. OCT/LMD-TS, respectively, with most proteins being more highly abundant in the FF/HOM samples. PCA and unsupervised hierarchical clustering analysis with these 216 and 171 proteins were able to distinguish FF/HOM from OCT/LMD-T and OCT/LMD-TS samples, respectively. Similar analyses using significantly differentially enriched GO terms also discriminated FF/HOM from OCT/LMD samples. No significantly differentially expressed proteins were detected between the OCT/LMD-T and OCT/LMD-TS samples but trended differences were detected.

Conclusions

The proteomic profiles of the OCT/LMD-TS samples were more similar to those from OCT/LMD-T samples than FF/HOM samples, suggesting a strong influence from the sample processing methods. These results indicate that in LC–MS/MS proteomic studies, FF/HOM samples exhibit different protein expression profiles from OCT/LMD samples and thus, results from these two different methods cannot be directly compared.

Abstract 5311: Integrated proteomic and informatic assessment of ER+/HER2- and ER-/HER2- breast tumors

Introduction

Emerging OMIC technologies enable the comprehensive molecular characterization of breast cancer to continually evolve. Published gene expression-based studies have differentiated ER-positive and ER-negative breast cancer. Herein, we have expanded these investigations utilizing a proteomics approach.

Methods

Clinical IHC subtyping on core biopsies was used to select available flash-frozen surgical samples for a cohort of 121 primary breast cancer patients. The cohort includes 103 ER+/HER2- and 18 ER-/PR-/HER2- cases. Mass spectrometry-based analyses were performed on these samples for proteomic, metabolomic, and lipidomic features. Comparative analysis was performed on proteomic data between the two groups of samples using Limma with stratified subsampling method and significance is reported at FDR&lt;0.05 and absolute FC&gt;1.5. PCA and hierarchical clustering analysis were applied on our training dataset and independent testing datasets using the identified significantly differentially expressed proteins as a signature to distinguish ER+/HER2- samples from ER-/PR-/HER2- samples. Euclidian distance was used as distance matrix and Ward was used as a linkage method in hierarchical clustering algorithm.

Results

The cohort was split into a training set (81 cases) and a testing set (40 cases) using stratification randomization. There were 171 differentially expressed proteins identified from the training dataset. 95% cases in our training dataset and 92.5% cases in our testing dataset were clustered correctly using hierarchical clustering method with the 171 proteins. Publicly available datasets from TCGA (RNA-seq, 846 cases) and CPTAC (proteomics, 57 cases) further validated the proteomic results from our cohort. 94.7% cases in CPTAC cohort and 93.2% cases in TCGA RNA-seq cohort were clustered correctly using the 171 proteins. KEGG pathway analysis using the differentially expressed proteins identified two significant pathways at FDR &lt; 0.05: DNA replication and cysteine and methionine metabolism. Independent metabolomic and lipidomic analysis identified profiles supporting our proteomic analysis. These results confirm that ER+/HER2- and ER-/PR-/HER2- tumors are molecularly distinct which can be separated by a panel of expressed proteins.

Disclaimers

The contents of this publication are the sole responsibility of the author(s) and do not necessarily reflect the views, opinions or policies of Uniformed Services University of the Health Sciences, The Henry M. Jackson Foundation for the Advancement of Military Medicine, Inc., the Department of Defense, the Departments of the Army, Navy, or Air Force. Mention of trade names, commercial products, or organizations does not imply endorsement by the U.S. Government.

Citation Format: Guisong Wang, Punit Shah, Vladimir Tolstikov, Praveen-Kumar Raj-Kumar, Jianfang Liu, Lori A. Sturtz, J. Leigh Fantacone-Campbell, Rebecca Zingmark, Jeffrey A. Hooke, Mary L. Cutler, Kris Richardson, Leonardo Rodrigues, Valerie Bussberg, Rangaprasad Sarangarajan, Niven R. Narain, Hai Hu, Michael A. Kiebish, Albert J. Kovatich, Craig D. Shriver. Integrated proteomic and informatic assessment of ER+/HER2- and ER-/HER2- breast tumors [abstract]. In: Proceedings of the Annual Meeting of the American Association for Cancer Research 2020; 2020 Apr 27-28 and Jun 22-24. Philadelphia (PA): AACR; Cancer Res 2020;80(16 Suppl):Abstract nr 5311.

Malignant cell-specific pro-tumorigenic role of type I interferon receptor in breast cancers

Within the microenvironment of solid tumors, stress associated with deficit of nutrients and oxygen as well as tumor-derived factors triggers the phosphorylation-dependent degradation of the IFNAR1 chain of type I interferon (IFN1) receptor and ensuing suppression of the IFN1 pathway. Here we sought to examine the importance of these events in malignant mammary cells. Expression of non-degradable IFNAR1^S526A mutant in mouse mammary adenocarcinoma cells stimulated the IFN1 pathway yet did not affect growth of these cells in vitro or ability to form subcutaneous tumors in the syngeneic mice. Remarkably, these cells exhibited a notably accelerated growth when transplanted orthotopically into mammary glands. Importantly, in human patients with either ER+ or ER- breast cancers, high levels of IFNAR1 were associated with poor prognosis. We discuss the putative mechanisms underlying the pro-tumorigenic role of IFNAR1 in malignant breast cells.

Abstract P2-18-03: Breast cancer treatment and association with clinicopathologic features in TCGA

Background: For a decade, The Cancer Genome Atlas (TCGA) program collected clinicopathologic annotation data along with multi-platform molecular profiles of more than 11,000 human tumors across 33 different cancer types. TCGA clinical data were analyzed and a standardized dataset named the TCGA Pan-Cancer Clinical Data Resource was created. However, TCGA treatment data are not systematically analyzed. Here we focus on TCGA primary breast cancer (TCGA-BC) treatment data to assess their completeness, regimen patterns, and association with clinicopathologic features.

Method: 814 TCGA-BC patients with treatment data diagnosed from 2001 to 2013 were selected for this study. The treatment data were prepared and classified to be adjuvant Chemotherapy (CT), adjuvant Radiation Therapy (RT), and adjuvant Hormone Therapy (HT). An in-house Clinical Breast Care Project (CBCP) treatment dataset (n=1051 from 2001 to 2017), which were relatively complete, were used as a reference for data completeness. Multinomial logistic regression was used to analyze the associations between different therapies and clinical features.

Results: There is no consistent treatment data difference between TCGA-BC and CBCP. In TCGA-BC, 68.7%, 64.4%, and 64.5% of patients received CT, RT, and HT respectively, whereas in CBCP, the corresponding percentages were 59.0%, 71.2%, and 77.1%. The percentages of patients receiving combined therapies are even more comparable between the two cohorts (data not shown due to many combinations).

The associations between treatment and clinicopathologic features were analyzed using multinomial logistic regressions with age, race, menopausal status, AJCC stage, and PAM50 subtype as covariates. Several of these covariates significantly associate with the use of therapies. Compared to patients with age &lt; 50, patients with age ≥ 65 were significantly more likely to receive HT, HT+RT, or RT than CT only (OR=11.34, 9.25, 4.82; p=0.001, 0.002, and 0.024 respectively); Compared to pre-menopausal patients, post-menopausal patients were significantly more likely to receive HT+RT than CT only (OR=4.12, p=0.034); Compared to patients with early stages (I, II), patients with advanced stages (III, IV) were more likely to receive CT+HT+RT (OR=1.92, p=0.038) or CT+RT (OR=2.38, p=0.010) than CT only, and less likely to receive CT+HT (OR=0.39, p=0.033) or HT (OR=0.30, p=0.008) than CT only. PAM50 subtype also significantly associates with the use of therapies. Compared to Luminal A patients, patients with Basal subtype were significantly less likely to receive CT+HT (OR=0.14, p=1.3e-05), CT+HT+RT (OR=0.06, p=3.0e-10), HT+RT (OR=0.01, p=9.2e-06), or HT (OR=0.02, p=3.8e-07) than CT alone. Patients with HER2+ subtype showed similar patterns (note targeted therapies were classified as CT). In addition, patients with Basal subtype were significantly more likely to receive CT+RT (OR=2.47, p=0.01) than CT only.

Conclusion: TCGA-BC patient treatment data are relatively as complete as our in-house CBCP patient treatment data, enabling us to perform a preliminary analysis of the former for association with clinicopathologic features of the patients. The significant associations of age, menopausal status, AJCC stage, and PAM50 subtype with treatment regimens are consistent with clinical knowledge, suggesting potential validity of TCGA BC treatment data for research use.

Disclaimer: The contents of this publication are the sole responsibility of the author(s) and do not necessarily reflect the views, opinions or policies of Uniformed Services University of the Health Sciences (USUHS), The Henry M. Jackson Foundation for the Advancement of Military Medicine, Inc., the Department of Defense (DoD), the Departments of the Army, Navy, or Air Force. Mention of trade names, commercial products, or organizations does not imply endorsement by the U.S. Government.

Citation Format: Jianfang Liu, J. Leigh Fantacone-Campbell, Albert J. Kovatich, Brenda Deyarmin, Bradley J. Mostoller, Jeffrey A. Hooke, Hallgeir Rui, Craig D. Shriver, Hai Hu. Breast cancer treatment and association with clinicopathologic features in TCGA [abstract]. In: Proceedings of the 2019 San Antonio Breast Cancer Symposium; 2019 Dec 10-14; San Antonio, TX. Philadelphia (PA): AACR; Cancer Res 2020;80(4 Suppl):Abstract nr P2-18-03.

Abstract P6-06-09: Evaluation of laser microdissected primary breast tumors for RNA Seq over bulk processing and validated with cohort control

Introduction: Laser microdissection (LMD) is a valuable method to isolate target populations of cells for molecular analysis. LMD of breast tumor samples can isolate breast tumor cells whereas bulk processing of tumor tissue will incorporate surrounding non-cancerous cells and bias tumor expression profiling. Here, we evaluated the advantage of using LMD breast tumors for RNA-Seq over bulk processing.

Methods: Tissue samples for the in-house dataset were from breast cancer patients consented by a HIPAA-compliant, IRB-approved protocol of the Clinical Breast Care Project. A total of 118 primary breast tumors embedded in OCT (Optimum Cutting Temperature) were selected and processed by LMD. Total RNA and protein were extracted using the illustra triplePrep kit. Paired-end RNA sequencing of 118 cases was performed using the Illumina HiSeq platform and the reads were preprocessed using a PERL-based pipeline involving PRINSEQ, GSNAP and HTSeq. The Cancer Genome Atlas (TCGA) primary breast cancer RNA-Seq data for 1097 tumors, bulk processed was downloaded. Differential expression of genes (DEG) was assessed using DESeq2. Significance was described for DEG with fold change &gt;2 and p-adjusted value of 0.05.

Results: A total of 24,518 genes with a mean expression of ≥ 10 (~9%) raw counts across 118 tumor samples were identified in the in-house LMD dataset. In TCGA breast cancer RNA-Seq, 14,281 genes with a mean expression of ≥ 100 (~9%) raw counts across 1097 tumor samples were identified. The conventional PAM50 classifier was used for intrinsic subtyping of in-house data, yielding 36 Basal-like, 14 HER2-enriched, 43 Luminal A, 22 Luminal B and 3 Normal-like calls. The provided PAM50 calls for TCGA were 192 Basal-like, 82 HER2-enriched, 566 Luminal A, 217 Luminal B and 40 Normal-like calls. Within commonly expressed 13,165 genes, LMD (in-house) and bulk (TCGA) processing exhibited approximately 40-78% non-overlap in significantly differentially expressed genes (SDEG) among the conventional intrinsic subtypes. 21 unique stromal genes were present in SDEG unique to TCGA whereas there were only 5 SDEG unique to in-house dataset. We validated the results with 34 patients that had both LMD and bulk processing RNA-Seq data and found the non-overlap genes percentage to be even greater from 46-85%. The observed percentages of non-overlapping genes in the whole datasets were also validated in the 34 overlapping cases when using IHC subtypes. Overall high positive correlation is observed among the stromal genes present in SDEG unique to TCGA suggesting strong stromal contribution in bulk processing. Pathway analysis of SDEG unique to LMD data suggested alterations in known cancer pathways (B-cell immune response, RNA metabolism and splicing, phagocytosis, and signaling components).

Conclusion: Analysis of The Cancer Genome Atlas breast cancer RNA-Seq data set (based on bulk tissue processing) suggested contribution of stromal signature genes and important differences from LMD specimens. Thus, tumor selection via LMD may allow us to unveil signals that are more specific to cancer cells.

Disclaimer: The contents of this publication are the sole responsibility of the author(s) and do not necessarily reflect the views, opinions or policies of Uniformed Services University of the Health Sciences (USUHS), The Henry M. Jackson Foundation for the Advancement of Military Medicine, Inc., the Department of Defense (DoD), the Departments of the Army, Navy, or Air Force. Mention of trade names, commercial products, or organizations does not imply endorsement by the U.S. Government.

Citation Format: Praveen-Kumar Raj-Kumar, Lori A. Sturtz, Albert J. Kovatich, Brenda Deyarmin, Jeffrey A. Hooke, Leigh Fantacone-Campbell, Anupama Praveen-Kumar, Jianfang Liu, James Craig, Leonid Kvecher, Jennifer Kane, Jennifer Melley, Stella Somiari, Stephen C. Benz, Justin Golovato, Shahrooz Rabizadeh, Patrick Soon-Shiong, Richard Mural, Craig D. Shriver, Hai Hu. Evaluation of laser microdissected primary breast tumors for RNA Seq over bulk processing and validated with cohort control [abstract]. In: Proceedings of the 2019 San Antonio Breast Cancer Symposium; 2019 Dec 10-14; San Antonio, TX. Philadelphia (PA): AACR; Cancer Res 2020;80(4 Suppl):Abstract nr P6-06-09.

Abstract 3402: Evaluation of laser microdissected primary breast tumors for RNA-Seq over bulk processing

Introduction: RNA-Seq based gene expression profiling of breast tumor samples is widely used to subgroup patients and to identify gene signatures of prognostic value. However, tumor samples are highly heterogeneous, and so bulk processing of tumor tissue will consist of several different cell types. Here, we evaluated the advantage of using laser microdissected (LMD) breast tumors for RNA-Seq over bulk processing.

Methods: Patients for the in-house dataset were duly consented under an IRB-approved protocol of the Clinical Breast Care Project. A total of 118 primary breast tumors embedded in OCT (Optimum Cutting Temperature) were selected and processed by LMD. Total RNA and protein were extracted using the Illustra triplePrep kit. Paired-end RNA sequencing of 118 cases was performed using the Illumina HiSeq platform and the reads were preprocessed using a PERL-based pipeline involving PRINSEQ, GSNAP and HTSeq. The Cancer Genome Atlas (TCGA) primary breast cancer RNA-Seq data for 1097 samples was downloaded. Differential expression of genes (DEG) was assessed using DESeq2. Significance was described for DEG with fold change &gt;2 and p-adjusted value of 0.05.

Results: A total of 24,518 genes with a mean expression of ≥ 10 raw counts across 118 tumor samples were identified in the in-house LMD dataset. In TCGA breast cancer RNA-Seq, 14,281 genes with a mean expression of ≥ 100 raw counts across 1097 tumor samples were identified. The conventional PAM50 classifier was used for intrinsic subtyping of in-house data, yielding 36 Basal-like, 14 HER2-enriched, 43 Luminal A, 22 Luminal B and 3 Normal-like calls. The provided PAM50 calls were used for TCGA which are 192 Basal-like, 82 HER2-enriched, 566 Luminal A, 217 Luminal B and 40 Normal-like calls. Within commonly expressed 13,165 genes, LMD and bulk processing exhibited approximately 40-78% non-overlap in significantly differentially expressed genes (SDEG) among the intrinsic subtypes. 21 unique stromal genes were present in SDEG unique to TCGA whereas there were only 5 SDEG unique to in-house dataset. Overall high positive correlation is observed among the stromal genes present in SDEG unique to TCGA suggesting strong stromal contribution in bulk processing. Pathway analysis of SDEG unique to LMD data suggested alterations in known cancer pathways (B-cell immune response, RNA metabolism and splicing, phagocytosis, and signaling components).

Conclusion: Analysis of The Cancer Genome Atlas breast cancer RNA-Seq data set (based on bulk processing tissue) suggested contribution of stromal signature genes and important differences from LMD specimens. Thus, tumor selection via LMD can result in better expression profiling by RNA-Seq which has the potential to uncover many cancer genes and pathways. The views expressed in this abstract are those of the author and do not reflect the official policy of the Department of Army/Navy/Air Force, Department of Defense, or U.S. Government.

Citation Format: Praveen-Kumar Raj-Kumar, Lori A. Sturtz, Albert J. Kovatich, Brenda Deyarmin, Jeffrey A. Hooke, Leigh Fantacone-Campbell, Anupama Praveen-Kumar, Jianfang Liu, James Craig, Leonid Kvecher, Jennifer Kane, Jennifer Melley, Stella Somiari, Stephen C. Benz, Justin Golovato, Shahrooz Rabizadeh, Patrick Soon-Shiong, Richard J. Mural, Craig D. Shriver, Hai Hu. Evaluation of laser microdissected primary breast tumors for RNA-Seq over bulk processing [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2019; 2019 Mar 29-Apr 3; Atlanta, GA. Philadelphia (PA): AACR; Cancer Res 2019;79(13 Suppl):Abstract nr 3402.

Loss of Nuclear Localized Parathyroid Hormone-Related Protein in Primary Breast Cancer Predicts Poor Clinical Outcome and Correlates with Suppressed Stat5 Signaling

PURPOSE

Parathyroid hormone-related protein (PTHrP) is required for normal mammary gland development and biology. A PTHLH gene polymorphism is associated with breast cancer risk, and PTHrP promotes growth of osteolytic breast cancer bone metastases. Accordingly, current dogma holds that PTHrP is upregulated in malignant primary breast tumors, but solid evidence for this assumption is missing.

EXPERIMENTAL DESIGN

We used quantitative IHC to measure PTHrP in normal and malignant breast epithelia, and correlated PTHrP levels in primary breast cancer with clinical outcome.

RESULTS

PTHrP levels were markedly downregulated in malignant compared with normal breast epithelia. Moreover, low levels of nuclear localized PTHrP in cancer cells correlated with unfavorable clinical outcome in a test and a validation cohort of breast cancer treated at different institutions totaling nearly 800 cases. PTHrP mRNA levels in tumors of a third cohort of 737 patients corroborated this association, also after multivariable adjustment for standard clinicopathologic parameters. Breast cancer PTHrP levels correlated strongly with transcription factors Stat5a/b, which are established markers of favorable prognosis and key mediators of prolactin signaling. Prolactin stimulated PTHrP transcript and protein in breast cancer cell lines in vitro and in vivo, effects mediated by Stat5 through the P2 gene promoter, producing transcript AT6 encoding the PTHrP 1-173 isoform. Low levels of AT6, but not two alternative transcripts, correlated with poor clinical outcome.

CONCLUSIONS

This study overturns the prevailing view that PTHrP is upregulated in primary breast cancers and identifies a direct prolactin-Stat5-PTHrP axis that is progressively lost in more aggressive tumors.

Abstract 284: Integrated proteogenomic analysis of laser microdissected primary breast tumors define proteome clusters

Introduction: Breast tumors have 4 well-established intrinsic subtypes based on transcriptome profiling. However, clusters defined by proteomics are often in disagreement with those defined by transcriptomics. Here, we report the findings of proteogenomic profiling of 118 laser microdissected (LMD) breast tumors using RNA-Seq and mass-spectrometry (MS)-based proteomic technologies.

Methods: Cases used in this study were drawn from the Clinical Breast Care Project, with patients consented using an IRB-approved protocol. A total of 118 primary breast tumors embedded in OCT were selected and processed by LMD. Total RNA and protein were extracted using the Illustra triplePrep kit. Paired-end RNA sequencing of 118 cases was performed using the Illumina HiSeq platform, and the reads were preprocessed using a PERL-based pipeline involving the preprocessing tool PRINSEQ, splice-aligner GSNAP and HTSeq for quantifying expression. Quantitative global proteomics analyses were performed on 113 cases using isobaric TMT 6-plex labeling with the “universal reference” strategy. MS data were acquired using a Q-Exactive instrument and analyzed using Proteome Discoverer with Byonic node. Sample-to-sample normalization was conducted to remove pipetting errors and ComBat was used to remove batch effect. K-means clustering was done using Bioconductor package Consensusclustering.

Results: The number of preprocessed RNA sequencing reads for the 118 cases ranged from over 43 to 295 million. An average of 83% of reads was mapped, and 24,518 genes with a mean expression of ≥ 10 counts across 118 tumor samples were identified. The PAM50 algorithm was used for intrinsic subtyping, yielding 37 Basal-like, 16 HER2-enriched, 39 Luminal A and 26 Luminal B calls. Unsupervised clustering of 3,000 highly varying genes reflected 4 intrinsic subtypes. In the global proteomics data, 840 proteins were identified across all 113 cases. Unsupervised K-means consensus clustering on all 840 or just using the top 210 highly varying proteins indicated the optimal number of clusters to be 3. These 3 clusters were identified as Basal-enriched, Luminal A-enriched and Luminal B-enriched. HER2-enriched cases were distributed among these clusters.

We did not observe a stromal-enriched cluster in this analysis of LMD-prepared samples that selected against stromal components of the tumor.

Conclusion: Analysis of LMD breast tumors using proteogenomic technologies resulted in 3 clusters for proteome data: basal-enriched, luminal A-enriched and luminal B-enriched. Unlike a recent report on proteomics clustering using bulk processing of tumors, a stromal-enriched cluster was not observed in this analysis which excluded stromal components of the samples.

The views expressed in this abstract are those of the author and do not reflect the official policy of the Department of Army/Navy/Air Force, Department of Defense, or U.S. Government.

Citation Format: Praveen-Kumar Raj-Kumar, Tao Liu, Lori A. Sturtz, Albert J. Kovatich, Marina A. Gritsenko, Vladislav A. Petyuk, Brenda Deyarmin, Viswanadham Sridhara, James Craig, Jason E. McDermott, Anil K. Shukla, Ronald J. Moore, Matthew E. Monroe, Bobbie-Jo M. Webb-Robertson, Jeffrey A. Hooke, J.Leigh Fantacone-Campbell, Leonid Kvecher, Jianfang Liu, Jennifer Kane, Jennifer Melley, Stella Somiari, Stephen C. Benz, Justin Golovato, Shahrooz Rabizadeh, Patrick Soon-Shiong, Richard D. Smith, Richard J. Mural, Karin D. Rodland, Craig D. Shriver, Hai Hu. Integrated proteogenomic analysis of laser microdissected primary breast tumors define proteome clusters [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2018; 2018 Apr 14-18; Chicago, IL. Philadelphia (PA): AACR; Cancer Res 2018;78(13 Suppl):Abstract nr 284.

Abstract P4-09-14: Analysis of breast cancer recurrence using gene set enrichment analysis

Background: Even after successful treatment of primary breast tumors, there is a continued risk of recurrence. The risk varies between subtypes and there are ongoing efforts that aim to improve prediction of such risks for individual patients. Detection of subclinical metastases might be achieved by biomarkers in blood. In this study, we profiled protein expression in blood plasma from patients with known clinical outcome (recurrence vs no recurrence) to identify prognostic markers of breast cancer recurrence.

Methods: The subjects and specimens were made available through the Clinical Breast Care Project using IRB-approved protocols. We analyzed blood plasma samples taken at the time of diagnosis from consented patients who subsequently relapsed (33 cases) as well as those with no disease recurrence (31 controls). Based on hormone receptor and lymph node status the samples were grouped as: ER-/HER2- (17 cases/15 controls), ER+/LN+ (10/10) and ER+/LN- (6/6). We used aptamer-based SOMAscan assay platform to study the expression of 1252 proteins. We analyzed the protein expression data by using their coding genes in order to apply the Gene Set Enrichment Analysis method (GSEA v.2, Broad Institute). Pathway databases of KEGG, REACTOME, BIOCARTA and C4 collection were used. Significant gene sets were called at 5% FDR, and overlaps and low coverage gene sets (Tags &lt;70%) were removed. Statistical analysis and clustering were done using R.

Results: Unsupervised clustering showed some difference in signal in the ER+/LN- group. Even though there was a lack of significantly differentiated proteins between the cases and controls of this group, many significant gene sets were identified. After applying the cutoff filters and removing the overlaps, there were 5 gene sets enriched with the pathway collection, involved in B-cell receptor signaling, mRNA metabolism, tight junction and SCF-KIT signaling. Similarly, 9 gene sets from the MORF compendium were differentially expressed with the C4 collection and included neighborhood genes of NME2, ACTG1, EIF3S2, AP2M1, DAP3, UBE2I, NPM1, AATF and NPM1. In contrast, neither differentially expressed proteins nor gene sets were identified from the ER+/LN+ and ER-/HER2- groups. Since the sample size of the ER+/LN- group was small, we conducted a similar analysis by randomly choosing 6 case and control samples in the other two groups respectively. There were still no differentially expressed proteins or gene sets identified above the specified cutoff parameters.

Conclusion: Using plasma protein expression data we identified underlying gene sets differentially expressed between ER+/LN- patients who had cancer recurrence and no recurrence. Many genes in these sets were already known biomarkers (e.g. PTEN, AKT1, STAT3, SET etc.). These results can be used for understanding patterns of recurrence in different cancer subtypes. Further research is needed to estimate the clinical significance of these gene products.

The views expressed in this article are those of the author and do not reflect the official policy of the Department of Army/Navy/Air Force, the Department of Defense, or U.S. Government.

Citation Format: Praveen Kumar A, Kovatich AJ, Biancotto A, Cheung F, Davidson-Moncada JK, Kvecher L, Liu J, Ru Y, Kovatich AW, Deyarmin B, Fantacone-Campbell JL, Hooke JA, Raj Kumar PK, Rui H, Hu H, Shriver CD. Analysis of breast cancer recurrence using gene set enrichment analysis [abstract]. In: Proceedings of the 2017 San Antonio Breast Cancer Symposium; 2017 Dec 5-9; San Antonio, TX. Philadelphia (PA): AACR; Cancer Res 2018;78(4 Suppl):Abstract nr P4-09-14.

Abstract 213: Integrated proteogenomic analysis of laser capture microdissected breast tumors

Introduction: Molecular characteristics of breast tumors play an important role in determining patients’ survival outcome. Here, we report preliminary findings of proteogenomic profiling of 50 breast tumors using RNA-Seq and mass-spectrometry (MS) based proteomic technologies. An additional 60 tumors are being analyzed, including WGS for all samples. We are also collecting patient survival outcome data.

Methods: Cases used in this study were drawn from the Clinical Breast Care Project, where patients were consented using an IRB-approved protocol. A total of 50 breast tumors were selected and processed by laser capture microdissection (LCM). This cohort includes 36 Caucasian Americans (CA) and 8 African Americans (AA), and the age of the patients is 57 ± 13 years. Protein and RNA were extracted using the Illustra triplePrep kit, which isolates DNA from the same cells as well. Quantitative global proteomics and phosphoproteomics analyses were performed using isobaric TMT 6-plex labeling with the “universal reference” strategy and IMAC enrichment of phosphopeptides. Mass spectrometry data were acquired using a Q-Exactive instrument and analyzed using Proteome Discoverer with Byonic node. Phosphopeptide abundance was normalized to abundance measurements of the parent protein for all of the phosphorylation analyses. Phosphoproteomic data was also searched for the presence of O-GlcNAc modifications. RNA-Seq analyses were done on Illumina HiSeq and the data were analyzed using GSNAP.

Results: There were 19 Luminal A, 7 Luminal B, 8 HER2-enriched, and 16 basal-like subtypes based on the PAM50 algorithm. In the global proteomics data, we were able to quantitate &gt;8600 proteins. Unsupervised clustering on the highly varying proteins across the samples resulted in two primary clusters, with one being luminal-enriched. The other cluster contains a basal-like tumor sub-cluster and a sub-cluster of mixed subtypes. Differential protein expression analyses between the two primary clusters confirmed known markers (e.g., overexpression of KRT8/KRT18 in luminal-enriched cluster). The luminal-enriched cluster is primarily CA with post-menopausal status.

A similar search of the phosphoproteomic data yielded quantitation of &gt;12500 phosphopeptides. Unsupervised clustering of the phosphoproteins resulted in four primary groups, with one being basal-enriched and another being luminal-enriched. We also observed &gt;50 overexpressed phosphopeptides. While some of these phosphosites have been previously reported (e.g., on RANBP2), other phosphosites appeared to be novel (e.g., on IRF2BP2).

Conclusion: Analysis of LCM breast tumors using proteogenomic technologies resulted in basal- and luminal-enriched clusters, thus enabling us to study protein and phosphopeptide markers across multiple platforms.

The views expressed in this article are those of the author and do not reflect the official policy of the Department of Defense, or U.S. Government.

Citation Format: Viswanadham Sridhara, Tao Liu, Marina A. Gritsenko, Lori A. Sturtz, Albert J. Kovatich, Vladislav A. Petyuk, Brenda Deyarmin, Jason E. McDermott, Anil K. Shukla, Ronald J. Moore, Matthew E. Monroe, Bobbie-Jo M. Webb-Robertson, Jeffrey A. Hooke, Leigh Fantacone-Campbell, Praveen Kumar Raj Kumar, Leonid Kvecher, Jianfang Liu, Jennifer Kane, Jennifer Melley, Stella Somiari, Joji Iida, Stephen C. Benz, Justin Golovato, Shahrooz Rabizadeh, Patrick Soon-Shiong, Richard D. Smith, Richard J. Mural, Craig D. Shriver, Hai Hu, Karin D. Rodland. Integrated proteogenomic analysis of laser capture microdissected breast tumors [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2017; 2017 Apr 1-5; Washington, DC. Philadelphia (PA): AACR; Cancer Res 2017;77(13 Suppl):Abstract nr 213. doi:10.1158/1538-7445.AM2017-213

Abstract 5403: Reproducible elevation of RNA versus DNA mutation signal in low purity breast tumors

Background: Accurate detection of somatic mutations is critical for informing targeted therapy options. Prevalent non-cancer cell admixture complicates this detection in breast cancer. Conventional mutation detection relies on DNA sequencing; however in prior work, we demonstrated that combining RNA and DNA sequencing increases mutation signal strength, or mutant allele fraction (MAF). The ratio of RNA MAF versus DNA MAF (RNA:DNA MAF) was greatest in low purity breast tumors. We hypothesized that this elevation is biologically driven and would be conserved in a second, distinct tissue specimen of the same tumors. Here, we compare mutation characteristics between two tissue blocks in a cohort of breast tumors (n = 8) to evaluate possible preservation of RNA versus DNA mutation signal throughout the tumor.

Methods: We selected four high purity and four low purity breast tumors (“Block1”) from The Cancer Genome Atlas (TCGA) cohort and associated ABSOLUTE purity analysis. For these tumors, we acquired a second tissue block (“Block2”) not analyzed by TCGA, cut sections, analyzed sections by H&E stains, and extracted nucleic acids. Whole genome DNA sequencing and mRNA sequencing was performed for Block2 specimens using Illumina X and NextSeq 500 sequencers, respectively. Somatic mutations in Block2 were detected using UNCeqR and compared to published UNCeqR somatic mutations from TCGA. We then evaluate MAF characteristics in the entire TCGA breast tumor cohort (n = 695).

Results: Tumor purity estimates, determined by histology and by sequencing, were reduced in Block2 of the low purity tumor set versus the high purity tumor set, consistent with Block1 analysis. Molecular properties of genome-wide gene expression and somatic DNA copy number were highly similar between block-mated specimens (p &lt; 0.01). We then identified expressed mutations present in Block1 and Block2 of the same tumor and compared the MAFs on these common mutations. DNA MAF and RNA MAF were each significantly correlated between Block1 and Block2 (p &lt; 1e-12 in both cases). The average RNA:DNA MAF was 2.5 for the cohort, indicating that RNA mutation signal is greater than DNA in general. In Block2 specimens, the RNA:DNA MAFs were significantly greater in the low purity tumor set than the high purity tumor set (mean 2.7 versus 2.1, p &lt; 6e-5), reflecting the same trend observed in Block1 specimens. Analyzing the entire TCGA cohort, RNA:DNA MAF was positively correlated with proliferation pathway gene expression (p &lt; 3e-16 ) and was greatest in the Basal subtype versus other subtypes (p &lt; 2e-9).

Conclusion: Mutant allele fraction both of DNA and of RNA was conserved across breast tumor subsections. Low purity and basal subtype breast tumors had elevated RNA:DNA MAF supporting a relationship to underlying biology and identifying classes of tumors with pronounced benefit for DNA and RNA integrated mutation analysis.

Citation Format: Jerez Te, Coralie Viollet, Xijun Zhang, Jatinder Singh, Jeffrey A. Hooke, Harvey B. Pollard, Hai Hu, Craig D. Shriver, Clifton L. Dalgard, Matthew D. Wilkerson. Reproducible elevation of RNA versus DNA mutation signal in low purity breast tumors [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2017; 2017 Apr 1-5; Washington, DC. Philadelphia (PA): AACR; Cancer Res 2017;77(13 Suppl):Abstract nr 5403. doi:10.1158/1538-7445.AM2017-5403

Validation of tumor protein marker quantification by two independent automated immunofluorescence image analysis platforms

Protein marker levels in formalin-fixed, paraffin-embedded tissue sections traditionally have been assayed by chromogenic immunohistochemistry and evaluated visually by pathologists. Pathologist scoring of chromogen staining intensity is subjective and generates low-resolution ordinal or nominal data rather than continuous data. Emerging digital pathology platforms now allow quantification of chromogen or fluorescence signals by computer-assisted image analysis, providing continuous immunohistochemistry values. Fluorescence immunohistochemistry offers greater dynamic signal range than chromogen immunohistochemistry, and combined with image analysis holds the promise of enhanced sensitivity and analytic resolution, and consequently more robust quantification. However, commercial fluorescence scanners and image analysis software differ in features and capabilities, and claims of objective quantitative immunohistochemistry are difficult to validate as pathologist scoring is subjective and there is no accepted gold standard. Here we provide the first side-by-side validation of two technologically distinct commercial fluorescence immunohistochemistry analysis platforms. We document highly consistent results by (1) concordance analysis of fluorescence immunohistochemistry values and (2) agreement in outcome predictions both for objective, data-driven cutpoint dichotomization with Kaplan-Meier analyses or employment of continuous marker values to compute receiver-operating curves. The two platforms examined rely on distinct fluorescence immunohistochemistry imaging hardware, microscopy vs line scanning, and functionally distinct image analysis software. Fluorescence immunohistochemistry values for nuclear-localized and tyrosine-phosphorylated Stat5a/b computed by each platform on a cohort of 323 breast cancer cases revealed high concordance after linear calibration, a finding confirmed on an independent 382 case cohort, with concordance correlation coefficients >0.98. Data-driven optimal cutpoints for outcome prediction by either platform were reciprocally applicable to the data derived by the alternate platform, identifying patients with low Nuc-pYStat5 at ~3.5-fold increased risk of disease progression. Our analyses identified two highly concordant fluorescence immunohistochemistry platforms that may serve as benchmarks for testing of other platforms, and low interoperator variability supports the implementation of objective tumor marker quantification in pathology laboratories.

Abstract P4-09-14: PhosphohistoneH3 as a prognostic marker in breast cancer: High expression is associated with younger age, triple negative subtype, and disease specific survival

BACKGROUND PhosphohistoneH3 (PPH3) is an emerging marker in breast cancer and has been linked to both patient survival and age. Phosphorylation of HistoneH3 is an important step during the cell cycle leading to proper compaction of the chromatin during late G2 and early mitosis. Here we assessed the use of PPH3 as a prognostic marker within a group of invasive breast cancers in the Clinical Breast Care Project (CBCP).

METHODS CBCP participants and their samples were collected following IRB-approved, HIPAA-compliant protocols. Samples from 157 CBCP patients were selected for tissue whole section immunohistochemistry (IHC), using antibodies to PPH3, ER, PR, Ki67, and Her2. For each sample, staining of PPH3 was assessed across 5 high powered microscope fields and was considered positive if there was on average &gt;2 stained cells per field. ER and PR were considered positive when there was &gt;5% nuclear staining, and Ki67 was positive when there was &gt;15% nuclear staining. Her2 was considered positive with an IHC score of 3+ or 2+ with a FISH score above 2.2. The samples were subtyped as Luminal A (LA: ER+/HER2-/Ki67-), Luminal B1 (LB1: ER+/HER2-/Ki67+), Luminal B2 (LB2: ER+/HER2+), Her2+ (ER-/PR-/HER2+), and Triple Negative (TN: ER-/PR-/HER2-). PPH3 was tested for associations with age and subtype using a stratified univariate Wilcoxon rank-sum analysis and a multivariate analysis controlling for subtype. To test the efficacy of PPH3 as a prognostic marker, Kaplan-Meier curves for disease specific survival were analyzed and the cox proportional hazard regression model was calculated. Further analysis addressing population demographics and additional cancer characteristics is ongoing.

RESULTS Wilcoxon analysis revealed an association between higher PPH3 levels and younger age (P=.0038). Subtype was also found to be associated with PPH3, with the TN subtype 6.26 times more likely to have higher PPH3 expression than LA (P=.005). The association with age was confirmed by repeating the analysis and stratifying into non-TN subtypes (P=.05) and TN only subtype (P=.017). Non-TN subtypes positive for PPH3 expression had median age of 53.18 at diagnosis and 63.29 for negative PPH3 expression; TN subtypes that were positive for PPH3 had a median age of 50.44 and 72.9 for negative PPH3. Multivariate analysis with age and subtype as the variables also supported these results (age P=.017; TN vs LA P=.022). Disease specific survival analysis showed that a shorter survival time was associated with positive PPH3 protein levels (P=0.03; hazard ratio=6.97).

CONCLUSIONS High expression of PPH3 is associated with a younger age, poorer survival rate, and the TN subtype. These results corroborate the use of PPH3 as a prognostic marker for breast cancer patients.

The views expressed in this article are those of the author and do not reflect the official policy of the Department of Defense, or U.S. Government.

Citation Format: Craig J, Kovatich AJ, Hooke JA, Kvecher L, Liu J, Fantacone-Campbell JL, Rui H, Shriver CD, Hu H. PhosphohistoneH3 as a prognostic marker in breast cancer: High expression is associated with younger age, triple negative subtype, and disease specific survival. [abstract]. In: Proceedings of the Thirty-Eighth Annual CTRC-AACR San Antonio Breast Cancer Symposium: 2015 Dec 8-12; San Antonio, TX. Philadelphia (PA): AACR; Cancer Res 2016;76(4 Suppl):Abstract nr P4-09-14.

Molecular Heterogeneity in Primary Breast Carcinomas and Axillary Lymph Node Metastases Assessed by Genomic Fingerprinting Analysis

Molecular heterogeneity within primary breast carcinomas and among axillary lymph node (LN) metastases may impact diagnosis and confound treatment. In this study, we used short tandem repeated sequences to assess genomic heterogeneity and to determine hereditary relationships among primary tumor areas and regional metastases from 30 breast cancer patients. We found that primary carcinomas were genetically heterogeneous and sampling multiple areas was necessary to adequately assess genomic variability. LN metastases appeared to originate at different time periods during disease progression from different sites of the primary tumor and the extent of genomic divergence among regional metastases was associated with a less favorable patient outcome (P = 0.009). In conclusion, metastasis is a complex process influenced by primary tumor heterogeneity and variability in the timing of dissemination. Genomic variation in primary breast tumors and regional metastases may negatively impact clinical diagnostics and contribute to therapeutic resistance.

Steroid induction of therapy-resistant cytokeratin-5-positive cells in estrogen receptor-positive breast cancer through a BCL6-dependent mechanism

Therapy resistance remains a major problem in estrogen receptor-α (ERα)-positive breast cancer. A subgroup of ERα-positive breast cancer is characterized by mosaic presence of a minor population of ERα-negative cancer cells expressing the basal cytokeratin-5 (CK5). These CK5-positive cells are therapy resistant and have increased tumor-initiating potential. Although a series of reports document induction of the CK5-positive cells by progestins, it is unknown if other 3-ketosteroids share this ability. We now report that glucocorticoids and mineralocorticoids effectively expand the CK5-positive cell population. CK5-positive cells induced by 3-ketosteroids lacked ERα and progesterone receptors, expressed stem cell marker, CD44, and displayed increased clonogenicity in soft agar and broad drug-resistance in vitro and in vivo. Upregulation of CK5-positive cells by 3-ketosteroids required induction of the transcriptional repressor BCL6 based on suppression of BCL6 by two independent BCL6 small hairpin RNAs or by prolactin. Prolactin also suppressed 3-ketosteroid induction of CK5+ cells in T47D xenografts in vivo. Survival analysis with recursive partitioning in node-negative ERα-positive breast cancer using quantitative CK5 and BCL6 mRNA or protein expression data identified patients at high or low risk for tumor recurrence in two independent patient cohorts. The data provide a mechanism by which common pathophysiological or pharmacologic elevations in glucocorticoids or other 3-ketosteroids may adversely affect patients with mixed ERα+/CK5+ breast cancer. The observations further suggest a cooperative diagnostic utility of CK5 and BCL6 expression levels and justify exploring efficacy of inhibitors of BCL6 and 3-ketosteroid receptors for a subset of ERα-positive breast cancers.

Abstract P3-07-20: Survival comparative analysis of patients with invasive breast cancer treated by a military medical center and matched patients of the US general population

BACKGROUND

U.S. military beneficiaries differ from the U.S. general population with regards to access to health care as care is provided at no or much lower cost in the military health system. Other differences also exist. Many of these differences are known factors affecting invasive breast cancer outcomes. Thus it is desirable to conduct a comparative analysis of breast cancer patient outcomes between these two populations to find out whether there is any outcome difference, and if yes what the contributing factors are.

METHODS

We compared overall survival (OS), disease-specific survival (DSS), and 5-year OS and DSS rates in breast cancers between 399 patients from the Clinical Breast Care Project at the Walter Reed National Military Medical Center (CBCP-WR) and 1,000 sets of 1596 matched patients from the Surveillance, Epidemiology, and End Results (SEER) Program of the National Cancer Institute. All patients were diagnosed between 2001 and 2010. Each CBCP-WR patient was randomly matched to four SEER patients on six demographic and clinicopathologic variables (age at diagnosis in 5-year groups, race, diagnosis year, estrogen receptor (ER), progesterone receptor, and AJCC stage).

RESULTS

The CBCP-WR cohort had better survival than the SEER population. At the whole cohort level, the mean hazard ratios (HRs) from 1,000 matched comparisons for OS and DSS were 0.774 and 0.708, with mean log-rank P-values of 0.124 and 0.125. The numbers of 175 and 141 comparisons showing a log-rank P-value &lt;0.05 out of the 1,000 tests were significantly higher than what would be expected from a random distribution of these P-values (P&lt;0.00001, exact binomial test). By stratifying the cohorts we identified that this survival disparity was mainly contributed by patients with a diagnosis age ≥50 years (for DSS, mean HR=0.550, mean P=0.049, and 642 of 1,000 tests showed a P&lt;0.05; for OS, mean HR=0.713, mean P=0.081, and 377 of 1,000 tests showed a P&lt;0.05), but not by patients with a diagnosis age &lt;50 years. The absolute differences in 5-year DSS rates were 4.4% (94.6% in CBCP-WR vs. 90.2% in SEER; mean P=0.010) for all matched patients and 4.8% (95.2% vs. 90.4%; mean P=0.015) for patients diagnosed at an age ≥50 years. Again there was no significant difference for patients diagnosed at an age &lt;50 years. When stratified by race, ER, stage or grade, most of the patient subpopulations showed favorable 5-year OS and DSS rates in the CBCP-WR cohorts.

CONCLUSION

Overall, these results suggested that breast cancer patients, especially older patients seen in the CBCP-WR, carried more favorable outcomes than those from the general population. The findings warrant further analyses of the contributing factors, such as health care access, treatments, population characteristics, additional pathologic characteristics, and socioeconomic statuses, to this outcome disparity.

The views expressed in this article are those of the author and do not reflect the official policy of the Department of Defense, or U.S. Government.

Citation Format: Yuanbin Ru, Jianfang Liu, Jamie Leigh Campbell, Kangmin Zhu, Albert J Kovatich, Jeffrey A Hooke, Leonid Kvecher, Brenda Deyarmin, Audrey W Kovatich, Frank Cammarata, Hallgeir Rui, Richard J Mural, Craig D Shriver, Hai Hu. Survival comparative analysis of patients with invasive breast cancer treated by a military medical center and matched patients of the US general population [abstract]. In: Proceedings of the Thirty-Seventh Annual CTRC-AACR San Antonio Breast Cancer Symposium: 2014 Dec 9-13; San Antonio, TX. Philadelphia (PA): AACR; Cancer Res 2015;75(9 Suppl):Abstract nr P3-07-20.

Abstract B49: Molecular characterization of breast tumor-associated adipose

Background: Long thought to function only as an inert energy storage depot, the role of adipose tissue in breast tumorigenesis has been largely ignored. In light of increasing rates of obesity and use of breast conserving therapy and autologous fat grafting, improved understanding of the role of adipose in tumor etiology is crucial.

Methods: Adipose adjacent to and distant from invasive breast tumors (n=20), or adjacent to non-malignant diagnoses (n=20) was laser microdissected from post-menopausal women. Gene expression data were generated using microarrays and data analyzed to identify significant patterns of differential expression between adipose groups, at the individual gene and molecular pathway level.

Results: Pathway analysis revealed significant differences in immune response between non-malignant, distant and tumor adjacent adipose, with the highest response in tumor-adjacent and lowest in non-malignant adipose. Adipose from invasive breasts exhibits increased expression in anti-inflammatory genes, such as MARCO and VSIG4, while genes differentially expressed between tumor-adjacent and distant adipose such as SPP1, RRM2 and MMP9, are associated with increased cellular proliferation, invasion, and angiogenesis.

Conclusions: Gene expression levels differ in breast adipose, depending on presence of or proximity to tumor cells. Heightened immunotolerance in adipose from invasive breasts provides a microenvironment favorable to tumorigenesis. In addition, tumor-adjacent adipose demonstrates expression of genes associated with tumor growth and progression. Together, these data suggest that adipose is not an inert component of the breast microenvironment but plays an active role in tumorigenesis.

Citation Format: William C. Yarina, Lori A. Field, Brenda Deyarmin, Ryan van Laar, Jeffrey A. Hooke, Craig D. Shriver, Rachel E. Ellsworth. Molecular characterization of breast tumor-associated adipose. [abstract]. In: Abstracts: AACR Special Conference on Cellular Heterogeneity in the Tumor Microenvironment; 2014 Feb 26-Mar 1; San Diego, CA. Philadelphia (PA): AACR; Cancer Res 2015;75(1 Suppl):Abstract nr B49. doi:10.1158/1538-7445.CHTME14-B49

Prolactin-Stat5 signaling in breast cancer is potently disrupted by acidosis within the tumor microenvironment

INTRODUCTION

Emerging evidence in estrogen receptor-positive breast cancer supports the notion that prolactin-Stat5 signaling promotes survival and maintenance of differentiated luminal cells, and loss of nuclear tyrosine phosphorylated Stat5 (Nuc-pYStat5) in clinical breast cancer is associated with increased risk of antiestrogen therapy failure. However, the molecular mechanisms underlying loss of Nuc-pYStat5 in breast cancer remain poorly defined.

METHODS

We investigated whether moderate extracellular acidosis of pH 6.5 to 6.9 frequently observed in breast cancer inhibits prolactin-Stat5 signaling, using in vitro and in vivo experimental approaches combined with quantitative immunofluorescence protein analyses to interrogate archival breast cancer specimens.

RESULTS

Moderate acidosis at pH 6.8 potently disrupted signaling by receptors for prolactin but not epidermal growth factor, oncostatin M, IGF1, FGF or growth hormone. In breast cancer specimens there was mutually exclusive expression of Nuc-pYStat5 and GLUT1, a glucose transporter upregulated in glycolysis-dependent carcinoma cells and an indirect marker of lactacidosis. Mutually exclusive expression of GLUT1 and Nuc-pYStat5 occurred globally or regionally within tumors, consistent with global or regional acidosis. All prolactin-induced signals and transcripts were suppressed by acidosis, and the acidosis effect was rapid and immediately reversible, supporting a mechanism of acidosis disruption of prolactin binding to receptor. T47D breast cancer xenotransplants in mice displayed variable acidosis (pH 6.5 to 6.9) and tumor regions with elevated GLUT1 displayed resistance to exogenous prolactin despite unaltered levels of prolactin receptors and Stat5.

CONCLUSIONS

Moderate extracellular acidosis effectively blocks prolactin signaling in breast cancer. We propose that acidosis-induced prolactin resistance represents a previously unrecognized mechanism by which breast cancer cells may escape homeostatic control.

Abstract P4-06-03: Assays on core biopsies and surgically resected tumors may result in different subtyping of the invasive breast cancer from the same patient

Background Core biopsies (CBs) are often used for biomarker expression assays to determine the treatment regimen. However, a number of other clinically important analyses (e.g. OncoType Dx), are performed on surgically resected tumors (SRTs). A previous study has shown that biomarkers ER, PR, and Ki67 expressed higher in CBs than in SRTs. Here we analyze how this difference impacts the subtyping of ER+ breast tumors.

Methods Female patients enrolled in the Clinical Breast Care Project (CBCP) from a civilian site were selected for this study, where expression of ER, PR, HER2, and Ki67 were assayed by IHC in a reference lab on CBs; the same 4 assays were performed on SRTs by a CBCP central lab. Both labs are CLIA-certified. Patients treated with neoadjuvant chemotherapy and those with multiple tumors were excluded. 167 cases were identified for this study to compare assays performed on CBs and SRTs from the same patients. ER and PR were positive if &gt;1% nuclear staining, HER2 was negative if IHC = 0 or 1+, positive if IHC = 3+, and for IHC = 2+ FISH was used for the final call. Ki67 was positive if &gt; = 15% nuclear staining. LA was ER+/HER2-/Ki67-, LB1 was ER+/HER2-/Ki67+, and LB2 was ER+/HER2+. For histologic grades, only readings from the central lab on SRTs were used. Statistical analyses were performed using SAS.

Results This analysis confirmed that Ki67, ER, and PR showed higher percent nuclear staining in CBs than in SRTs from the same patients. The difference for Ki67 was more striking and unidirectional. ER and PR cases clustered at the upper percent levels. Histograms with a bin-width of 15% show a peak at 15% for Ki67 difference between CBs and SRTs, whereas the peaks for ER and PR differences were at 0%. McNemar's (or Exact McNemar’s) test showed significant differences between the binary status calls for Ki67 (p = 3.2E-15) and ER (p = 0.012), but not for PR (p = 0.65). Assays on CBs and SRTs resulted in different subtype calls for the cases (Table 1). Grade distributions were different between LA and LB (p&lt;0.001 for both CB- and SRT-based subtypes, Chi-Square or Fisher's Exact test), but not so between LB1 and LB2 (p = 0.23 for CB, 0.31 for SRT). However, SRT-based LB1 cases concentrate more on higher grades compared to CB-based cases (p = 0.048).

Table 1. ER+ subtypes based on IHC assays (from CBs and SRTs) and corresponding grades (from SRTs) CBSRTSubtypeG1G2G3G1G2G3LA2126034518LB11435342820LB2036032

Discussion On IHC assays, Ki67 expression is strikingly higher in CBs than in SRTs, and ER expression is also higher in CBs than in SRTs. This directly resulted in more LB than LA subtypes based on CBs. SRT-based LB1 cases concentrate more on higher grades compared to CB-based cases, which is more consistent with the observation that LB subtypes have worse outcomes. A limitation of this study is that technical differences between the labs may contribute to the observed differences between CBs and SRTs. Further studies need to be performed to determine whether SRT should also be assayed in addition to CB for treatment regimen decision-making.

The views expressed in this abstract are those of the authors and do not reflect the official policy of the Department of Defense, or US Government.

Citation Information: Cancer Res 2013;73(24 Suppl): Abstract nr P4-06-03.

Abstract P1-08-20: Increased risk of hormone therapy failure in breast cancers expressing low phospho-Stat5: Validation of quantitative immunofluorescence assay parameters

Previous analyses of three breast cancer cohorts revealed that loss of phospho-Stat5 in breast cancer is associated with significantly elevated risk of hormone therapy failure (1, 2). Nuclear localized tyrosine phosphorylated Stat5 (Nuc-pYStat5) may therefore have clinical value as a predictive marker. Analysis of two of the three previously reported anti-estrogen treated patient cohorts used pathologist scoring of diaminobenzidine (DAB) chromogen-stained Stat5. However the third cohort, analyzed by quantitative immunofluorescence analysis (QIF) on the Genoptix/HistoRx AQUA platform, revealed a greater hazard ratio than the cohorts analyzed by pathologist DAB-scoring. To extend and validate these observations, we applied the Nuc-pYStat5 cutpoint derived in our previous study (2) to an independent cohort of anti-estrogen-treated breast cancer patients using two distinct QIF software platforms, AQUA and Definiens Tissue Studio. Tissue Studio relies on supervised machine learning and multiparametric features of a high-resolution whole slide image to identify cancer cell regions, while AQUA software relies on costaining of a tumor marker to identify cancer cell regions. The two QIF platforms produced highly concordant Nuc-pYStat5 levels (R2 linear = 0.96, P&lt;0.001, N = 344) and confirmed a significant elevated risk of failing antiestrogen therapy in patients whose tumors had lost Nuc-pYStat5 (Hazard ratio 3.6; 95% CI 1.8-7.4; P&lt;0.02; N = 98). On both QIF platforms, Nuc-pYStat5 remained an independent marker after multivariate adjustment for standard pathology parameters, including ER/PR, HER2, age, node status and grade, with a hazard ratio of 5.8 (95% CI 1.3-22.2; P = 0.02; N = 52). High concordance between Nuc-pYStat5 levels produced by the two QIF platforms held up in a second independent dataset of more than 300 breast cancer specimens (R2 linear = 0.97, P&lt;0.001, N = 382). Nuc-pYStat5 levels by the two QIF methods remained highly concordant across the entire dynamic range in both patient cohorts. Furthermore, high concordance was also observed between replicate QIF analyses of Nuc-pYStat5 on serial tumor microarray sections stained in the same run on an automated immunostainer (Concordance Correlation Coefficient (CCC) = 0.96; 95% CI 0.96-0.97). Modest inter-assay staining variation (CCC = 0.84; 95% CI 0.82-0.87) for Nuc-pYStat5 when serial tumor microarrays were stained on different runs several days apart could be corrected for by normalization procedures (CCC = 0.94; 95% CI 0.92-0.95). This progress supports the utility of QIF analysis of Nuc-pYStat5 levels in human breast cancer and further documents the potential value of Nuc-pYStat5 as a predictive marker of response to antiestrogen therapy. The study confirms that further retrospective and prospective validation studies are warranted.

References:

1) Yamashita et al. Stat5 expression predicts response to endocrine therapy and improves survival in estrogen receptor-positive breast cancer. Endocr Relat Cancer. 2006;13:885-93.

2) Peck et al. Loss of nuclear localized and tyrosine phosphorylated Stat5 in breast cancer predicts poor clinical outcome and increased risk of antiestrogen therapy failure. J Clin Oncol. 2011;29:2448-58.

Citation Information: Cancer Res 2013;73(24 Suppl): Abstract nr P1-08-20.

Abstract P4-06-09: HER2+ and HER2- luminal B subtypes have similar overall survival and histologic grade distributions

Background There are multiple subtypes in invasive breast cancers (IBCs). Immunohistochemistry (IHC)-based assays using ER, PR, HER2, and Ki67 for subtyping has been developed. However, association between such subtypes and treatment outcomes and histology is not completely known, and are impacted by dataset-to-dataset and pathologist-to-pathologist variations. We report an analysis on these problems, as a pilot study of a project involving 5,000 patients and 250 protein biomarkers.

Methods Patients were enrolled for the Clinical Breast Care Project from a military site with data collected per IRB-approved protocols, from 2000 to 2010. Total of 215 female IBC cases were included in this study, with surgically resected tumors (SRT) assayed for ER, PR, HER2, and Ki67 by IHC in a central CLIA-certified lab following clinical guidelines where applicable. All slides were reviewed by a single experienced breast pathologist. ER and PR was positive if nuclear staining was &gt;5%. HER2 was negative if IHC = 0 or 1+ and positive if IHC = 3+; For IHC = 2+, the FISH result determined the final call. Ki67 was positive if nuclear staining was &gt; = 15%. For IBC subtypes, LA was ER+/HER2-/Ki67-; Two LB subtypes were defined, with LB1 being ER+/HER2-/Ki67+ and LB2 being ER+/HER2+; Her2+ was ER-/PR-/HER2+; TN was ER-/PR-/HER2-. Statistical analyses were performed using SAS, Kaplan-Meier estimate and log-rank test were used for survival analysis and the follow-up period was 10 years with a median of 4.6 years. Chi-Square test was used for categorical data analysis supplemented by Fisher's Exact test as appropriate.

Results 204 of the 215 cases were classified into subtypes of LA (n = 74, 7 deceased), LB1 (n = 53, 4 deceased), LB2 (n = 14, 1 deceased), Her2+ (n = 14, 1 deceased), and TN (n = 49, 16 deceased). Despite the low number of events in some subtypes, there was a significant difference in overall survival between the 5 subtypes of IBCs defined here (p = 0.0023), with TN cases showing the least favorable outcome. No difference was observed in outcome between LB1 and LB2 (p = 0.86). Overall, Ki67+ cases trended toward worse outcomes (p = 0.08), which was also observed in TN (p = 0.17) but not other subtypes. Histologic grades were significantly different among the 5 subtypes (p = 6.25E-20); 96% of LA cases were G1 or G2, over 80% of LB1 and LB2 cases were G2 or G3, and all Her2+ and 93% of TN cases were G2 or G3. Within the luminal subtypes, grade distribution for LA cases was significantly different from that for LB cases (p&lt;0.0001) but there was no difference between LB1 and LB2 cases (p = 0.95).

Discussion In this cohort where all IHC and pathology slides were reviewed by a single pathologist, we used cell proliferation marker Ki67 to help classify luminal IBCs into LA, LB1 (HER2-), and LB2 (HER2+). Overall survival analysis result for all cases was consistent with the literature, Ki67+ cases trended toward worse outcomes, and no outcome difference was identified between LB1 and LB2. Histologic grade distributions in different subtypes were consistent with the literature; we further found no difference between LB1 and LB2 subtypes.

The views expressed in this abstract are those of the authors and do not reflect the official policy of the Department of Defense, or US Government.

Citation Information: Cancer Res 2013;73(24 Suppl): Abstract nr P4-06-09.

Abstract P2-05-21: Molecular subtypes of invasive breast cancers show differential expression of the proliferation marker Aurora Kinase A (AURKA)

Background: Invasive breast cancer (IBC) has been classified into four major subtypes based on gene expression profiling. The luminal A subtype (LA) has the best prognosis, when compared to luminal B (LB), HER2+, and basal-like (Basal). Ki67 by gene expression or immunohistochemistry (IHC) is commonly used as a proliferation index. The function of Ki67 in proliferation remains unknown. AURKA (STK15) is known to play an important role in mitosis, and is a component of the 21-gene recurrence score of the Oncotype Dx. With multiple platforms of molecular data available from hundreds of IBC tissues in The Cancer Genome Atlas project (TCGA), we sought to study the association of AURKA with different IBC subtypes and explore its use as a proliferation marker in IBCs.

Methods: Gene expression (Agilent, log2 transformed), relative DNA copy number (CN, Affymetrix SNP 6.0), and exome sequence mutation (Illumina) data for 459 IBC cases were downloaded from the TCGA data portal. PAM50 classification results of all samples were obtained from the TCGA breast cancer AWG group and included 203 LA, 113 LB, 51 HER2+, 84 Basal-like, and 8 Normal-like which were not used in this study due to the low numbers. Kruskal-Wallis tests were used to evaluate the differences among four subtypes on AURKA expression and CN, followed by Wilcoxon Mann-Whitney test with Bonferroni adjustment for pairwise analyses. Pearson's Correlation Coefficient was used for correlation analyses. All statistical analyses were performed using SAS and R, and two-sided, p values &lt;.05 were considered statistically significant.

Results: There was a significant difference among IBC subtypes, in gene expression as well as in CN (p values &lt; 0.0001). AURKA mRNA levels were significantly lower in LA (mean±SD, −2.61±0.63) compared to LB (−1.45±0.78), HER2+ (−1.38±0.61), and Basal (−1.26±0.62) subtypes (p values all &lt; 0.0001). No significant difference was detected between other subtype pairs. In CN analysis, Basal (0.09±0.22) was lower than HER2+ (0.32±0.308, p &lt; 0.0002) and LB (0.33±0.41, p &lt; 0.0001), and LA (0.14±0.28) is lower than HER2 (p &lt; 0.0016) and LB (p &lt; 0.0001), but no other significant CN difference between the subtypes were found. The means and SDs are provided for reference only. No correlation of p53 mutation status and AURKA expression were observed. However, AURKA gene expression level is correlated with MKI67 gene expression (R = 0.69, p &lt; 2.2e−16), and its correlation with PAM50 proliferation score is even higher (R = 0.80, p &lt; 2.2e−16).

Discussion: Using the TCGA data we observed that the mean gene expression level of AURKA is significantly lower in LA than the other IBC subtypes, by more than 50% (note the log2 transformation). This differential expression is not completely due to CN changes (especially for the Basal subtype). There is a strong association with other tumor cell proliferation markers such as the MKI67 gene and the PAM50 proliferation score. We are using computational and laboratorial studies to better understand the role of AURKA in the etiology of invasive breast cancers.

The views expressed in this article are those of the author and do not reflect the official policy of the Department of Defense, or U.S. Government.

Citation Information: Cancer Res 2012;72(24 Suppl):Abstract nr P2-05-21.

Abstract P6-02-08: Molecular drivers of adipogenotoxicosis in breast tumor-associated adipose

Background: Having long been thought to function only as an inert energy storage depot, the role of adipose tissue in tumorigenesis has been largely ignored; however, adipose is an active endocrine organ that can directly influence tumor growth. Improved understanding of the role of adipose in tumorigenesis is crucial given the association between obesity and breast cancer risk in post-menopausal women, increasing rates of obesity and use of autologous fat transfer in breast reconstruction.

Methods: Adipose, adjacent to and distant from (&gt;3 cm from the closest tumor margin) invasive breast tumors, was laser microdissected from 20 post-menopausal women, and from 22 post-menopausal women with non-malignant breast disease. Gene expression data were generated using U133A 2.0 microarrays. Data were analyzed to identify significant patterns of differential expression between adipose classes at the individual gene and molecular pathway level. Gene expression differences were validated using qRT-PCR in an additional set of 29 specimens.

Results: SPP1, RRM2, MMP9 and PLA2G7 were expressed at &gt;3-fold (P &lt; 0.01) higher levels in adjacent adipose compared to distant adipose from the same breast. A number of immune response genes including MARCO, FABP7, ELF5, MYBPC1, MMP7, CLDN8, HLA-DQB1 and HLA-DQA1 were differentially expressed in distant adipose compared to adipose from non-malignant breasts. The most significant gene expression differences were detected between tumor-adjacent and non-malignant adipose with &gt;3-fold higher expression of EGFL6 and ITGB2 and &gt;3-fold lower levels of PIP, which are involved in growth, proliferation, and cellular adhesion in adjacent compared to non-malignant adipose. Pathway analysis revealed that immune response differs between non-malignant, distant and tumor-adjacent adipose with an enhanced B- and T-cell response detected in adjacent compared to distant or non-malignant adipose. Inflammatory response as well as DNA transcription and replication pathways were differentially expressed in distant compared to non-malignant adipose.

Conclusions: Gene expression levels differ in breast adipose depending on presence of and proximity to tumor cells. Adipose adjacent to the tumor demonstrated the largest immune response, supporting the idea of adipogenotoxicosis, which through pro-inflammatory and genotoxic responses, promotes tumor development. In addition, genes involved in cellular proliferation, degradation of the extracellular matrix and angiogenesis are differentially expressed in adjacent compared to distant or non-malignant adipose, thus tumor-adjacent adipose may be contributing to the growth and invasion of the primary tumor. These data thus suggest that adipose is not an inert component of the breast microenvironment but plays an active role in tumorigenesis.

Citation Information: Cancer Res 2012;72(24 Suppl):Abstract nr P6-02-08.

Abstract P5-01-07: Fibroadenomatoid changes are more prevalent in middle-aged women and have a positive association with invasive breast cancer

Background: The role of benign breast diseases (BBDs) in the development of invasive breast cancers (IBCs) has been studied for many years. Some BBDs have been studied comprehensively (e.g., fibrocystic changes (FCC)) while less is known about other BBDs (e.g., fiboadenomatoid changes (FAC)). FAC has been considered by some researchers as a precursor of fibroadenoma (FA). Conclusions from different studies vary, partially due to different interpretation methods and diagnostic criteria when multiple hospitals and pathologists were involved. In this study, we used subjects in the Clinical Breast Care Project (CBCP) from a military medical center where pathology slides were reviewed by a single breast pathologist to study FAC, FA, and FCC in comparison to the published literature.

Methods: Subjects were enrolled in the study following IRB-approved, HIPAA-compliant protocols. All the clinicopathologic data are available from the CBCP data warehouse (DW4TR). In the CBCP, FCC is composed of 4 components: stromal fibrosis, cysts, apocrine metaplasia, and sclerosing adenosis. Two modeling studies were performed. i) For the BBDs and IBC association study, two groups of subjects were identified: 1136 subjects diagnosed with “Benign” or “Atypical” diseases, and 619 cases diagnosed with IBCs. A logistic regression model was developed for the prediction of IBCs by the 3 BBDs and 2 well-established risk factors (RF): age (younger, &lt;=40; middle-aged, 41–60; older, &gt;60) and race (Caucasian, African American, Asian, and other). ii) For the RF association study with the BBDs, 6 additional RFs reported to be associated with these BBDs were identified from the literature: current use of oral contraceptives, number of live births, education, body mass index, hormonal replacement therapy, and IBC family history. These 8 RFs were used to develop a logistic regression model for each of the BBDs. The analyses were performed in SAS.

Results: In the first study, age and race were confirmed as RFs for IBCs. FAC was positively associated with IBC (OR = 3.04, 95%CI=2.06 to 4.50). FA was negatively associated with IBC, and the level of the association was stronger in women without FCC (OR = 0.15, 95%CI=0.08 to 0.28), compared to women with FCC (OR = 0.40, 95%CI=0.24 to 0.65). FCC was not significantly associated with IBC. Results from the second study indicated that, age was significantly associated with FAC (p = 0.015), specifically the middle-aged women were more likely to have FAC compared to younger women (OR = 2.03, 95%CI=1.23 to 3.34), while the older women were at a non-significantly increased risk. Trends of association with FAC were also noted for the number of live birth (p = 0.095), ethnicity (p = 0.096), and current oral contraceptive pill use (p = 0.077). The FCC model results were in general consistent with the literature, and we also confirmed that age was negatively associated with the diagnosis of FA.

Discussion: Our study was consistent with FCC findings in the literature. We observed that FAC was positively associated with IBC, whereas FA was negatively associated. Also, FAC occurred more often in middle-aged women while FAs occurrence was higher in younger women. Our results suggest that FAC and FA may be two different diseases.

Citation Information: Cancer Res 2012;72(24 Suppl):Abstract nr P5-01-07.

P1-06-24: Nuclear Localization of Stat5a Predicts Response to Antiestrogen Therapy and Prognosis of Clinical Breast Cancer Outcome

Nuclear-localized and tyrosine-phosphorylated Stat5 has been reported as a favorable prognostic marker and predictor of response to antiestrogen therapy in breast cancer. Phospho-Stat5 antibodies do not distinguish between phosphorylated Stat5a and the closely related Stat5b, but Stat5a is considered more critical for normal mammary development than Stat5b. The purpose of this study was to determine whether levels of nuclear-localized Stat5a protein (Nuc-Stat5a) were prognostic of clinical outcome or predictive of antiestrogen response. Stat5a was detected by traditional diaminobenzidine-chromogen immunohistochemistry (IHC) and pathologist scoring or by quantitative immunofluorescence in five archival cohorts of breast cancer. Levels of nuclear-localized Stat5a (Nuc-Stat5a) were evaluated by pathologist scoring of whole tissue sections detected by IHC or automated quantitative analysis (AQUA) of immunofluorescently-labeled tissue microarrays. Levels of Nuc-Stat5a were reduced in invasive breast cancer tissues and lymph node metastases compared to normal tissue and ductal carcinoma in situ when quantified by AQUA (Material I; n=180). Tissues from patients not treated with adjuvant therapy or treated with antiestrogen monotherapy were analyzed according to Nuc-Stat5a status for breast cancer-specific survival (CSS) and time to recurrence (TTR) using univariate and multivariate statistical models, adjusting for clinical features including tumor grade, size, lymph node and ER, PR and Her2 status. In two prognostic cohorts of node-negative breast cancer patients, low expression of Nuc-Stat5a, detected by standard IHC (Material II; n=223) or quantitative analysis (Material III; n=198), was prognostic of poor breast cancer outcome as measured by univariate and multivariate CSS (Material II/III) and TTR (Material II). CSS and TTR analysis of two independent materials of tumors from patients treated with antiestrogen monotherapy and analyzed by standard IHC (Material IV; n=73) or quantitative immunofluorescence (Material V; n=97) indicated that patients whose tumors expressed low levels of Nuc-Stat5a were at a greater than 4-fold risk of antiestrogen therapy failure when adjusted for hormone receptor status and clinical features (multivariate CSS: Material IV HR=4.3 (1.2,15.6), p=0.03; Material V HR=5.0 (1.87,13.06), p=0.001). In conclusion, loss of Nuc-Stat5a is a promising independent marker of poor breast cancer prognosis in node-negative, non-adjuvant treated breast cancer patients. Additionally, Nuc-Stat5a may be a useful clinical tool to predict tumor response to antiestrogen therapy.

Citation Information: Cancer Res 2011;71(24 Suppl):Abstract nr P1-06-24.

P1-03-06: Fibroadenomatoid Changes Have a Higher Occurrence Rate in Middle-Aged Benign Breast Disease Patients with the Trend Retained in Cancer Patients

Background: Fibroadenoma (FA) is a common benign breast lesion known to have a high incidence rate in younger women. There are controversial reports whether FA elevates the risk of developing breast cancers. In clinical practice, FA may be surgically removed due to multiple reasons making it complicated to study its impact on the development of breast cancers that have a higher incidence rate in older women. Fibroadenomatoid change (FAC), also known as fibroadenomatous hyperplasia, is an uncommon lesion with histologic features similar to that of FA but lacking well-defined borders and usually discovered incidentally on breast biopsy specimens. FAC is not surgically targeted. The Walter Reed Army Medical Center, through the Clinical Breast Care Project, has enrolled over 2000 subjects undergoing a biopsy; all the pathology was reviewed by a single pathologist. These subjects provide an opportunity to study the age-dependent pattern of FAC in different patient populations.

Methods: Subjects were enrolled following IRB-approved protocols with data collected through two comprehensive questionnaires, a Core Questionnaire and a Pathology Checklist. A total of 1964 female subjects were identified for this study, including 1135 benign/atypical, 192 in situ, and 637 invasive cancer patients. Patients were divided into three age groups: &lt;=45 years, 46–65 years, and &gt;=66 years. Chi-Square test in the SAS was used for statistical analysis.

Results: As shown in the table, FA occurrence rate decreases significantly with increasing age in benign disease patients. FAC, on the other hand, shows a significantly higher occurrence rate in middle-aged patients with benign findings, and this trend is retained in the invasive or in situ cancer populations. FAC rate is also significantly higher in patients with cancer (invasive, or invasive and in situ combined) compared to benign patients in each age group with p-values ranging from 0.0001 to 0.019 (not shown).

P3-03-03: Congruence between Patterns of microRNA Expression and Histologic Grading of Invasive Breast Carcinomas

Background

Histologic grading may be used as an indicator of prognosis in breast cancer; patients with low-grade carcinomas have ∼85% ten-year survival compared to just 45% survival in patients with high-grade disease. Although useful for risk stratification, assigning nuclear grade is subjective, and a large proportion of carcinomas are classified as intermediate-grade with uncertain prognosis, thus limiting clinical utility. MicroRNAs (miRNAs) regulate gene expression and serve an important role in breast cancer development. In this study we examined miRNA expression profiles in low-grade and high-grade breast carcinomas to determine if miRNA expression is associated with pathological classifications of tumor grade.

Methods: Breast tumors were obtained from 69 patients enrolled in the Clinical Breast Care Project. Samples were partitioned into low-grade (n=30) or high-grade (n=39) categories using the Nottingham Histologic Score. Following laser microdissection of frozen tissue sections, miRNA was isolated from pure populations of breast tumor cells and hybridized to Affymetrix GeneChip® miRNA arrays containing over 800 human miRNA probes. Expression profiles were analyzed with Partek Genomics Suite using the miRNA Expression Module.

Results: We identified 30 unique miRNAs that showed differential expression at a False Discovery Rate (FDR) p&lt;0.05 between low-grade and high-grade breast carcinomas. Gene targets for these miRNAs function primarily in metabolism and cell communication. Expression of hsa-miR-18a and hsa-miR-572 was significantly different between histologic grades at an FDR p&lt;1×10−8 and hierarchical clustering based on these miRNAs correctly classified 97% (29/30) of low-grade and 90% (35/39) of high-grade tumors. miR-18a has been shown to inhibit ER signaling and promote cellular differentiation, while the role of miR-572 in breast carcinogenesis is not well known.

Conclusions: Dysregulation of miRNAs may accompany changes in cellular morphology typically used in histologic classification of breast carcinomas. Patterns of miRNA expression may improve reproducibility and clinical utility of tumor grading and may prove useful for prediction of recurrence and survival for patients with intermediate-grade carcinomas.

Citation Information: Cancer Res 2011;71(24 Suppl):Abstract nr P3-03-03.

P3-06-06: Comparison of Gene Expression Profiles of Lymph Node Positive and Lymph Node Negative ER Positive Breast Tumors in Pre- and Postmenopausal Women

Background

Breast cancer is the most common female cancer in US and is the second leading cause of cancer related death in women. Metastases are the primary cause of cancer morbidity and mortality. Axillary lymph node (LN) status has long been used as a prognostic factor for breast cancer. The molecular mechanisms that control LN metastasis remains poorly understood. To better understand the various genes and regulatory pathways that drive breast cancer LN metastasis, we compared the gene expression profiles between breast tumors that have metastasized to the LNs and those which have not in pre- and postmenopausal women.

Material and Methods: Tumor cells were isolated from the primary tumors (ER+) of postmenopausal node positive (PMNP; N=20), postmenopausal node negative (PMNN; N=19), premenopausal node positive (PRNP; N=18) and premenopausal node negative (PRNN; N=16) women using laser capture microdissection. RNA was isolated using the RNAqueous®-micro kit (Ambion, Austin, TX). Total RNA was converted to Biotin-labelled aRNA using two rounds of amplification with MessageAmp II aRNA amplification kit (Applied Biosystems, Foster City, CA). The aRNA concentration was determined by Nanodrop 1000 and the quality was assessed with a Bioanalyzer. The aRNA was fragmented and hybridized to Human Genome U133 Plus 2.0 GeneChip (Affymetrix, Santa Clara, CA). Microarray raw data were analyzed using a variety of R programming packages for probe density processing, background correction, normalization, quality control/quality assessment, and calculation of gene expression value, etc. To identify differentially expressed genes, Wilcoxon rank sum test with FDR (false discovery rate) control was performed for pair-wise comparison between different groups. Functional analyses were performed on the identified statistically significant differentially expressed genes to search for the functional categories and pathways in which they are involved and further understand their potential roles in breast cancer metastatic process.

Results: Multivariate data mining (hierarchical clustering analysis and principal component analysis, etc) revealed that in postmenopausal women, the node positive and node negative women are well separated while this was not the case in premenopausal women. Further analysis of the PMNN and PMNP groups to identify differentially expressed genes (with at least a 1.5 fold difference) at FDR =0.1 showed that 232 genes were upregulated and 470 genes were downregulated in PMNP vs PMNN groups. Gene function analysis revealed that genes down regulated in the PMNP group compared to PMNN are related to extracellular matrix, cell adhesion, EGF-like pathway, cytoskeleton etc, while the over-expressed genes are related to cell cycle and cell division, chromosome condensation, etc.

Discussion: The ability to differentiate lymph node positive cases from lymph node negative cases in ER+ breast cancer based on transcriptional profiling may have an impact on the clinical management of ER+ breast cancer cases. Having transcriptional profiles that identify ER+ tumors likely to have poor outcomes would suggest more aggressive treatment for such patients.

Citation Information: Cancer Res 2011;71(24 Suppl):Abstract nr P3-06-06.

P3-05-02: Subtype-Specific Co-Occurrence of Atypical Hyperplasia and In Situ Carcinoma with Invasive Breast Cancers

Background

Atypical ductal hyperplasia (ADH), lobular carcinoma in situ (LCIS), and ductal carcinoma in situ (DCIS) are considered risk factors for the development of invasive breast cancer (IBC). The co-occurrence of these lesions with IBC may provide insights into cancer initiation and development. IBC subtypes have distinct clinicopathological features. A clinically practical IHC-based subtyping classification has been developed based on the expression of ER, PR, HER2, and Ki67, defining Luminal A (LA), Luminal B (LB), HER2+, and Triple Negative (TN) subtypes. The Walter Reed Army Medical Center (WRAMC), through the Clinical Breast Care Project (CBCP), has enrolled over 500 IBC subjects with single pathologist review and central lab analysis. The co-occurrence of ADH, LCIS, and DCIS will be studied in relation to IBC subtypes.

Methods: Subjects were enrolled following IRB-approved protocols. IBC patients enrolled at WRAMC were selected and their clinical and pathology data were reviewed. ER and PR positivity is defined as &gt; 5% nuclear staining. The HER2 result is negative if the IHC=0 or 1+ and positive if IHC=3+. For IHC=2+, the FISH result determines the final HER2 status. Ki67 is considered positive if nuclear staining is &gt;= 15%. For IBC subtypes, LA is ER+/HER2−/Ki67-; LB is either ER+/HER2−/Ki67+, or ER+/HER2+; HER2+ is ER-/PR-/HER2+; TN is ER-/PR-/HER2−. Statistical analysis was performed using SAS, and the Chi-Square test was used for categorical data analysis supplemented by the Fisher's Exact test where appropriate. For age analysis, ANOVA was performed with Bonferroni adjustment for multi-pair t-test.

Results: A total of 459 IBC patients were identified and categorized into LA (41.6%), LB (27.7%), HER2+ (10.2%), and TN (20.5%). Many of the previously reported subtype-specific characteristics were confirmed. Age at diagnosis varied by subtype (p=0.0034) with LA being the oldest (Mean±SD=59.9+12.5 years) and TN the youngest (54±12.6 years, p=0.0048). Ethnicity distribution of African American (AA) relative to Caucasian American patients varied significantly in subtypes with AA=18% in LA, 31% in LB, 32% in Her2+, and 42% in TN (p=0.0008). The grade, the AJCC stage and its components T and N were all significantly different among the subtypes (p ranges from &lt;0.0001 to 0.0020). The grades and stages were consistently lowest for LA, highest for HER2+ and TN. We further found that the co-occurrence of ADH, DCIS, and LCIS with IBC were subtype-specific with the following distributions: ADH—LA (25.1%), LB (18.9%), HER2+ (0%), and TN (6.4%) (p&lt;0.0001, n=78); DCIS—LA (63.4%), LB (76.4%), HER2+ (80.9%), and TN (58.5%) (p=0.0039, n=311); LCIS—LA (36.7%), LB (19.7%), HER2+ (4.3%), and TN (6.4%) (p&lt;0.0001, n=103).

Discussion: By including Ki67 in IHC-based IBC subtyping we confirmed many subtype-specific clinico-pathological characteristics in the CBCP WRAMC population. We further report subtype-specific co-occurrences of ADH, DCIS, and LCIS. These co-occurrence patterns may reveal distinct developmental mechanisms between the different subtypes of IBC.

Citation Information: Cancer Res 2011;71(24 Suppl):Abstract nr P3-05-02.

DW4TR: A Data Warehouse for Translational Research

The linkage between the clinical and laboratory research domains is a key issue in translational research. Integration of clinicopathologic data alone is a major task given the number of data elements involved. For a translational research environment, it is critical to make these data usable at the point-of-need. Individual systems have been developed to meet the needs of particular projects though the need for a generalizable system has been recognized. Increased use of Electronic Medical Record data in translational research will demand generalizing the system for integrating clinical data to support the study of a broad range of human diseases. To ultimately satisfy these needs, we have developed a system to support multiple translational research projects. This system, the Data Warehouse for Translational Research (DW4TR), is based on a light-weight, patient-centric modularly-structured clinical data model and a specimen-centric molecular data model. The temporal relationships of the data are also part of the model. The data are accessed through an interface composed of an Aggregated Biomedical-Information Browser (ABB) and an Individual Subject Information Viewer (ISIV) which target general users. The system was developed to support a breast cancer translational research program and has been extended to support a gynecological disease program. Further extensions of the DW4TR are underway. We believe that the DW4TR will play an important role in translational research across multiple disease types.

Identification of differentially expressed genes in breast tumors from African American compared with Caucasian women

BACKGROUND

Breast tumors from African American women have less favorable pathological characteristics and higher mortality rates than those of Caucasian women. Although socioeconomic status may influence prognosis, biological factors are also likely to contribute to tumor behavior.

METHODS

Patients with invasive breast cancer were matched by age, grade, and estrogen receptor status; patients with benign disease were matched by age and diagnosis type. RNA from laser microdissected tumors and whole-sectioned nonmalignant breast tissues was hybridized to HG U133A 2.0 microarrays. Data were analyzed using Partek Genomics Suite using a cutoff of P < .001, >1.5-fold change, and results were validated by quantitative real-time polymerase chain reaction.

RESULTS

Clinicopathological factors did not differ significantly between groups for age at diagnosis, tumor size or stage, lymph node or human epidermal growth receptor 2 status, intrinsic subtype, or mortality. Two-way analysis of the tumor specimens revealed 25 probes representing 23 genes differentially expressed between populations; hierarchical clustering classified 24 of 26 African American women and 25 of 26 Caucasian women correctly. In the nonmalignant specimens, 15 probes representing 13 genes were differentially expressed, including 5 genes that also differed in the tumor specimens; these genes were able to correctly classify nonmalignant breast specimens from 20 of 22 of African American women and all of the Caucasian women.

CONCLUSIONS

Despite matching of tumors by pathological characteristics, molecular profiles differed between African American women and Caucasian women in both invasive tumors and benign breast tissues. These differentially expressed genes, including CRYBB2, PSPHL, and SOS1, are involved in cellular growth and differentiation, invasion, metastasis, and immune response and thus may contribute to the poor outcome in African American women.

Loss of nuclear localized and tyrosine phosphorylated Stat5 in breast cancer predicts poor clinical outcome and increased risk of antiestrogen therapy failure

PURPOSE

To investigate nuclear localized and tyrosine phosphorylated Stat5 (Nuc-pYStat5) as a marker of prognosis in node-negative breast cancer and as a predictor of response to antiestrogen therapy.

PATIENTS AND METHODS

Levels of Nuc-pYStat5 were analyzed in five archival cohorts of breast cancer by traditional diaminobenzidine-chromogen immunostaining and pathologist scoring of whole tissue sections or by immunofluorescence and automated quantitative analysis (AQUA) of tissue microarrays.

RESULTS

Nuc-pYStat5 was an independent prognostic marker as measured by cancer-specific survival (CSS) in patients with node-negative breast cancer who did not receive systemic adjuvant therapy, when adjusted for common pathology parameters in multivariate analyses both by standard chromogen detection with pathologist scoring of whole tissue sections (cohort I; n = 233) and quantitative immunofluorescence of a tissue microarray (cohort II; n = 291). Two distinct monoclonal antibodies gave concordant results. A progression array (cohort III; n = 180) revealed frequent loss of Nuc-pYStat5 in invasive carcinoma compared to normal breast epithelia or ductal carcinoma in situ, and general loss of Nuc-pYStat5 in lymph node metastases. In cohort IV (n = 221), loss of Nuc-pYStat5 was associated with increased risk of antiestrogen therapy failure as measured by univariate CSS and time to recurrence (TTR). More sensitive AQUA quantification of Nuc-pYStat5 in antiestrogen-treated patients (cohort V; n = 97) identified by multivariate analysis patients with low Nuc-pYStat5 at elevated risk for therapy failure (CSS hazard ratio [HR], 21.55; 95% CI, 5.61 to 82.77; P < .001; TTR HR, 7.30; 95% CI, 2.34 to 22.78; P = .001). CONCLUSION Nuc-pYStat5 is an independent prognostic marker in node-negative breast cancer. If confirmed in prospective studies, Nuc-pYStat5 may become a useful predictive marker of response to adjuvant hormone therapy.

Differential gene expression in primary breast tumors associated with lymph node metastasis

Lymph node status remains one of the most useful prognostic indicators in breast cancer; however, current methods to assess nodal status disrupt the lymphatic system and may lead to secondary complications. Identification of molecular signatures discriminating lymph node-positive from lymph node-negative primary tumors would allow for stratification of patients requiring surgical assesment of lymph nodes. Primary breast tumors from women with negative (n = 41) and positive (n = 35) lymph node status matched for possible confounding factors were subjected to laser microdissection and gene expression data generated. Although ANOVA analysis (P < .001, fold-change >1.5) revealed 13 differentially expressed genes, hierarchical clustering classified 90% of node-negative but only 66% of node-positive tumors correctly. The inability to derive molecular profiles of metastasis in primary tumors may reflect tumor heterogeneity, paucity of cells within the primary tumor with metastatic potential, influence of the microenvironment, or inherited host susceptibility to metastasis.

Abstract 2277: Loss of nuclear localized and tyrosine phosphorylated Stat5: A predictor of poor clinical outcome and increased risk of antiestrogen therapy failure in breast cancer

Stat5 transcription factor is activated by prolactin in the mammary gland and is necessary for mammary gland development, differentiation and lactation. Previous work based on tissue microarrays has suggested that nuclear localized and tyrosine phosphorylated Stat5 (Nuc-pYStat5) is a marker of prognosis in node-negative breast cancer. The purpose of this study was to validate the prognostic value of Nuc-pYStat5 in whole tissue sections and expand the analyses to a quantitative immunofluoresence-based assay. In addition, we explored Nuc-pYStat5 as a predictor of response to antiestrogen therapy. Levels of Nuc-pYStat5 were analyzed in five archival materials of breast cancer by traditional diaminobenzidine-chromogen immunohistochemistry (DAB-IHC) and pathologist scoring of whole tissue sections or by immunofluorescence and automated quantitative analysis (AQUA) of tissue microarrays. In two prognostic cohorts of node-negative breast cancer patients not receiving systemic adjuvant therapy, Nuc-pYStat5 was an independent marker of cancer-specific survival (CSS) when adjusted for common pathology parameters in multivariate analyses. Corresponding results were obtained both by standard DAB-IHC with pathologist scoring of whole tissue sections (Material I; n=233) and by quantitative immunofluorescence of a tissue microarray (Material II; n=291) and using two distinct monoclonal antibodies. A breast tissue progression array (Material III; n=130) revealed loss of Nuc-pYStat5 in invasive ductal carcinoma (IDC) compared to normal breast epithelia or ductal carcinoma in situ (DCIS), with greatest loss of Nuc-pYStat5 in lymph node metastases. Importantly, loss of Nuc-pYStat5, detected by DAB-IHC (Material IV; n=221), was associated with increased risk of antiestrogen therapy failure as measured by CSS and time to recurrence (TTR). More sensitive AQUA quantification of Nuc-pYStat5 in patients treated with antiestrogen monotherapy (Material V; n=97) identified a subset of patients with low Nuc-pYStat5 at a 7.4-fold elevated risk of dying from breast cancer (CSS univariate Cox regression HR=7.36 (2.94-18.42), p&lt;0.001, n=53). Importantly, low levels of Nuc-pYStat5 also predicted therapy failure in multivariate analyses (multivariate Cox regression, CSS HR=21.55 (5.61-82.77), p&lt;0.001; TTR HR=7.30 (2.34-22.78), p=0.001; n=53) independent of ER/PR status and lymph node status. We conclude that Nuc-pYStat5 is an independent prognostic marker in node-negative breast cancer. Furthermore, this study suggests that Nuc-pYStat5 may become useful as a predictive marker of response to systemic adjuvant hormone therapy if confirmed by prospective studies.

Citation Format: {Authors}. {Abstract title} [abstract]. In: Proceedings of the 102nd Annual Meeting of the American Association for Cancer Research; 2011 Apr 2-6; Orlando, FL. Philadelphia (PA): AACR; Cancer Res 2011;71(8 Suppl):Abstract nr 2277. doi:10.1158/1538-7445.AM2011-2277