Re Ellsworth

Abundance and distribution of polychlorinated biphenyls (PCBs) in breast tissue.

CTRC-AACR San Antonio Breast Cancer Symposium: 2008 AbstractsnnAbstract #5039 nnBackground: Breast cancer is a complex disease whose causative factors remain ambiguous and poorly defined. Many environmental chemicals accumulate in human tissues and may contribute to breast cancer risk. Polychlorinated biphenyls (PCBs), which constitute a class of chlorinated compounds that are ubiquitous in the environment, are believed to be associated with adverse health effects. Because the rising incidence of breast cancer may result from exposure to increasing levels of exogenous chemicals, there is keen interest in determining relationships between exposure to environmental chemicals and breast cancer risk. Material and Methods: Frozen sections of breast tissue were collected from each quadrant and the central region of the breast from 47 patients enrolled in the Clinical Breast Care Project at Walter Reed Army Medical Center. Pathological diagnoses ranged from disease free prophylactic mastectomy to metastatic breast cancer. From a portion of each sample, 98 PCB congeners were assayed by pressurized liquid extraction followed by high resolution gas chromatography. Results: PCBs were abundant in breast tissues. Only one congener was not observed in any patient and four (153, 138+163, 180, 206) were seen in all patients in at least one quadrant. Mean total PCB concentration in 202 breast quadrants was 602 ng/g lipid (range 0-4030), while for individual congeners, PCB 153 had an average level of 107 ng/g lipid (range 8-773 ng/g). Total PCB concentration (P<<0.001) and presence of 26 congeners (P<0.05) were highly correlated with patient age, particularly in the UIQ and LOQ, but not in the central region, suggesting that chemicals may accumulate at different rates in different regions of the breast. Similarly, no significant differences were observed for PCB concentrations among the four traditional quadrants, but 23 PCBs were present at significantly lower levels (P<0.05) in the central region. When examining PCB levels by disease status, seven individual congeners and median overall levels were significantly higher (P<0.05) in patients with invasive cancer compared to disease free patients. Discussion: Population monitoring is beginning to assess exposures of the general public to environmental chemicals to evaluate long-term health effects. This study showed that many PCB residues are present in human breast tissue, these chemicals are not distributed evenly, and several are correlated with age and presence of invasive disease. Improved understanding of the true carcinogenic potential of environmental chemicals may be useful in efforts to reduce individual risk for breast cancer.nnCitation Information: Cancer Res 2009;69(2 Suppl):Abstract nr 5039.

Abstract P4-10-08: Can a diagnosis of invasive breast cancer effectively motivate patients to follow healthy lifestyles?

Background: Survival rates for patients diagnosed with invasive breast cancer have increased dramatically, yet survivors often face a host of adverse health effects. Factors such as obesity, physical inactivity and tobacco use may contribute to decreased survival and quality of life. Behavioral risk factors in patients with and without breast cancer were evaluated to determine whether a diagnosis of invasive disease was sufficient motivation to modify lifestyle choices. Methods: The dataset included female patients diagnosed between 2001-2011with malignant (n=421) or benign (n=230) breast disease and who had baseline and >1-year follow-up information available. Changes in body mass index (BMI), fat intake, exercise frequency, alcohol and tobacco use, caffeine consumption, hormone replacement therapy (HRT) use and frequency of breast self-exam (BSE) were assessed. Random coefficients models were used to examine longitudinal effects of an invasive diagnosis on healthy behaviors. P Results: At diagnosis, patients with invasive cancer were significantly older (59 years), more likely to consume >7 glasses of alcohol/week (7%) but less likely to be using HRT (3%) than those with benign disease (50 years, 3% and 11%, respectively). At baseline, a majority of both invasive and benign patients were overweight, non (current) smokers, and consumed a high fat, highly caffeinated diet and exercised 50% of both groups performed BSE at least once/month. Exercise, BMI, and caffeine, alcohol and fat intake did not change over time in either invasive or benign groups. Smoking decreased to a similar extent in both invasive and benign patients. In contrast, compliance with monthly BSE increased and HRT use decreased significantly at each time point in the invasive patient group, with no corresponding changes in patients with benign disease. Conclusions: These data support the critical importance of providing education and recommendations about engaging in healthy behaviors by the clinical staff. The two behaviors that improved significantly in patients with invasive breast cancer are both addressed by the clinical staff at Walter Reed National Military Medical Center at the time of diagnosis: with nurse navigators providing education about proper BSE and physicians recommending discontinuation of HRT. Failure to significantly change other behaviors suggests that a diagnosis of breast cancer is not a sufficient motivating factor for the patient to adopt healthier lifestyle choices without provision of education and resources. Our data suggest a need for increased health-related behavioral counseling and support systems to successfully modify personal behaviors, thus improving the health and quality of life of breast cancer survivors. Citation Format: Ellsworth RE, Costantino N, Toro AL, Shriver CD, Ellsworth DL. Can a diagnosis of invasive breast cancer effectively motivate patients to follow healthy lifestyles?. [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 P4-10-08.

Abstract P5-06-03: Effect of obesity on molecular characteristics of invasive breast tumors: Gene expression analysis of 405 tumors by BMI

Background: Obesity is a risk factor for breast cancer in postmenopausal women and weight gain after diagnosis is associated with decreased survival and less favorable clinical characteristics such as greater tumor burden, higher grade, and poor prognosis, regardless of menopausal status. Despite the negative impact of obesity on clinical outcome, the molecular mechanisms by which excess adiposity influences breast cancer are not well-understood. Methods: Affymetrix U133 2.0 gene expression data was available for 405 primary breast tumors and 20 tumor-adjacent adipose tissues; RNA was isolated from tumor epithelia or adipose using laser microdissection. Patients were classified as lean (BMI 30). Statistical analysis was performed by ANOVA using Partek Genomics Suite version 6.6. Results: Obese patients were significantly more likely to be diagnosed >50 years and to be of African American ancestry compared to lean or overweight women, while pathological characteristics including tumor stage, size or grade, lymph node status and intrinsic subtype did not differ significantly between groups. Principal component analysis could not effectively cluster the tumors by weight and no genes were differentially expressed using a false discovery rate Conclusions: Although tumor epithelial cells from obese women do not differ significantly from those of lean and/or overweight women at the gene expression level, tumor-adjacent adipose did differ by weight. These data demonstrate that less favorable outcomes in obese patients are not be attributable to the tumor itself but to influences from the microenvironment and suggest that decreasing breast adiposity may be an effective strategy to reducing risk and improving outcomes in obese women. Citation Format: Ellsworth RE, Toro AL, Costantino NS, Shriver CD, Ellsworth DL. Effect of obesity on molecular characteristics of invasive breast tumors: Gene expression analysis of 405 tumors by BMI. [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 P5-06-03.

Abstract P4-06-04: Effect of genomic heterogeneity on breast cancer progression and metastatic spread

Purpose : DNA fingerprinting has revealed discordant patterns of chromosomal alterations in primary breast tumors (PBT) compared to matched metastatic axillary lymph node tumors (MLNT); however, whether these genetic differences reflect the timing of dissemination of cells with metastatic potential from the primary tumor, intratumoral heterogeneity of the primary breast tumor, or independent seeding of lymph nodes with metastatic cells remains unclear. Patients and Methods : From the 30 node-positive patients evaluated in this study, allelic imbalance (AI) data was generated using 52 microsatellite markers from 5-19 areas of each PBT as well as from available MLNT. Data were analyzed using pairwise correlations, ANOVA and PHYLIP. Results : The frequency of genomic changes was significantly higher ( P 0.001) in PBT areas (13%) than MLNT (9%). No two PBT areas had identical patterns of AI and the percent concordance (PC) of AI events between tumor regions ranged from 0-65% (average 33%). Of the 196 MLNT from 28 patients, PC between MLNT from the same patient ranged from 0-88% (average = 32%). Neither the overall frequency of AI nor the PC differed significantly between sentinel (SLN) and non-SLNs. Phylogenetic analysis revealed that within patients, many MLNT appeared to be descended from different areas of the PBT, but patterns of descent were complex. Conclusions : Both PBT and MLNT are characterized by extensive molecular heterogeneity, MLNT appear to originate from different areas of the PBT, and SLN metastases are not genomically more advanced than non-SLN metastases. These data suggest that metastatic dissemination may be influenced by both spatial and temporal factors, with cells with metastatic potential colonizing lymph nodes throughout the development of the PBT, and that SLN metastases do not appear to be a source of metastatic cells for non-SLN but rather MLNT arise by independent colonization. Citation Information: Cancer Res 2013;73(24 Suppl): Abstract nr P4-06-04.

Abstract PD02-02: The effect of HER2 expression on luminal A breast tumors

Background: Luminal A (LumA) tumors are the most common subtype of breast cancer and have been associated with favorable prognosis. Although tumors can be considered LumA if they have >1% expression of ER, negative HER2 status ranges from 0+ (no expression) to 2+ (weak to moderate expression in >10% of cells) without concurrent genomic amplification. It is not clear how moderate levels of HER2 expression influence tumor biology in LumA tumors. Methods: The Clinical Breast Care Project database was queried to identify all hormone receptor positive cases with HER2 status defined as 0+ (n = 89) or 2+/not amplified (n = 154). Clinicopathological characteristics were compared between groups using chi-square analysis. Twenty-five 0+ and 25 2+ tumors were subjected to laser microdissection and hybridized to U133 2.0 microarrays (Affymetrix). Gene expression data was analyzed using Partek Genomics Suite. Results: Age at diagnosis did not differ between patients with no (58.5 years) or moderate (59.3 years) HER2 expression. Patient groups did not differ by ethnicity, tumor stage, grade or size, or lymph node status. Although breast cancer mortality was low, all patients who died of disease (n = 7) had moderate levels of HER2 expression (p = 0.016). At the molecular level, four genes were differentially expressed (p 2x-fold change): ESRRG expressed at significantly higher levels and DHRS2, SCUBE2 and SERPINA6 expressed at significantly lower levels in tumors with moderate HER2 expression. Discussion: LumA (2+) tumors are dissimilar from LumA (0+) tumors. Increased expression of ESRRG has been associated with increased cellular proliferation and tamoxifen resistance, while decreased expression of SCUBE2 and SERPINA6 have been associated with inhibition of growth and proliferation. Together, these molecular differences suggest that LumA tumors with moderate HER2 expression, although classified as HER2 negative, have more aggressive characteristics than those without HER2 expression which may contribute to the significantly higher mortality rate seen in patients with LumA tumors expressing HER2. Citation Information: Cancer Res 2012;72(24 Suppl):Abstract nr PD02-02.

Abstract P3-04-07: Comparison of breast tumors with HER2 amplification and polysomy 17

Background: Approximately 20% of invasive breast cancers overexpress the human epidermal growth factor receptor, Her2, and are associated with poor prognosis including decreased relapse-free and overall survival. Overexpression of the Her2 protein is primarily due to amplification of the HER2 oncogene on chromosome 17, and these patients are typically treated with targeted therapies such as trastuzumab or lapatinib. However, increased HER2 copies may also result from chromosome 17 polysomy in which the entire chromosome has been duplicated one or more times. The clinical benefit of treating patients with polysomy 17 in the absence of HER2 amplification with targeted therapy remains unclear. Herein, we compared the clinical characteristics as well as the gene expression profiles of breast tumors with increased HER2 copy numbers due to HER2 amplification or chromosome 17 polysomy. Methods: Patients for this study were enrolled in the Clinical Breast Care Project. HER2 and CEP17 copy numbers were determined using the PathVysion HER-2 DNA Probe Kit (Abbott Laboratories). Only those patients with at least 4 HER2 copies per cell were included in the analysis; those with a HER2/CEP17 ratio 2.2 (N = 44) were considered amplified. Microarray data were generated on HG U133A 2.0 arrays (Affymetrix) for those samples with tissue available (14 polysomy and 18 amplified cases) and analyzed using Partek Genomics Suite. Results: Tumor grade and size did not differ between polysomic and amplified samples. Although amplified cases tended to be younger at diagnosis and less likely to be diagnosed with Stage I tumors, this difference did not reach statistical significance. HER2 amplified tumors were more likely to be ER negative ( P = 0.027) and have lymph node metastases ( P = 0.028) than polysomic tumors. Comparison of polysomic and amplified cases by expression microarray identified 67 genes that were differentially expressed (P 2) between the two groups. Thirty-nine genes were more highly expressed and 28 genes more lowly expressed in tumors with HER2 gene amplification when compared to tumors from patients with chromosome 17 polysomy. Discussion: ER and lymph node status of breast tumors from women with increased HER2 copy number differed depending on whether the increased copy number was due to HER2 gene amplification or conversely, polysomy of chromosome 17. Likewise, gene expression profiles also differed depending on whether increased HER2 was due to gene amplification or chromosomal gain. Those genes with significant levels of differential expression between these two groups are involved in cell cycle control, cell adhesion and migration, cell differentiation, cytoskeleton organization, signal transduction, protein synthesis and degradation, and ion transport. Interestingly, the top 4 most significant genes with higher expression in the polysomic samples, including TUBG1, CPD, BLMH, and JUP, are all located on 17q. Currently, it is not clear whether treatment with targeted Her2 therapy is beneficial in patients with chromosome 17 polysomy. However, the genes identified here may represent novel targets for developing new treatments that are effective in these women and require further study. Citation Information: Cancer Res 2012;72(24 Suppl):Abstract nr P3-04-07.

P5-05-03: The Effect of Breastfeeding on Molecular Characteristics of Invasive Breast Cancer.

INTRODUCTION : Breastfeeding has been associated with an overall decreased risk of developing breast cancer and the protective effect of breastfeeding has been associated with decreased risk of triple negative and estrogen-responsive breast cancer. Although the mechanism by which breastfeeding provides a protective effect is not well-defined, physical changes in the mammary epithelium reflecting maximal differentiation, may be involved. METHODS : The database of the Clinical Breast Care Project was queried to identify all patients born between 1946 and 1964 (“baby boomers”) with invasive breast cancer who had given birth to at least one living child. Clinicopathologic characteristics were compared between patients who breastfed for at least six cumulative months and those who never breastfed using chi-square and Fisher9s exact tests. RNA was isolated after laser-microdissection of 24 pairs of breastfeeding and non-breastfeeding tumors matched by stage, subtype and grade. Gene expression data was generated using HG U133A 2.0 microarrays and analyzed using the Partek Genomics Suite using a FDR RESULTS : Of the 1,008 parous women with invasive breast cancer in the database, 325 (33%) breastfed at least 6 months, 165 (16%) breastfeed less than 6 months, and 518 (51%) never breastfed. Women who breastfed CONCLUSIONS : Despite the matching of tumor samples, and lack of differences in pathological characteristics between women who breastfed ≥6 months and those who did not breastfeed, differentially expressed genes were identified between tumors from the two groups. Many of the differentially expressed genes have been associated with prognosis, especially in ER negative breast tumors. These data suggest that breastfeeding does alter the underlying molecular characteristics of invasive breast tumors, and may reflect alterations in exposure to estrogen levels during breastfeeding. Citation Information: Cancer Res 2011;71(24 Suppl):Abstract nr P5-05-03.

P1-09-01: Effect of Obesity on Gene Expression in Invasive Breast Tumors.

Background: Obesity is a risk factor for breast cancer in postmenopausal women and weight gain after diagnosis is associated with decreased survival and less favorable clinical characteristics such as greater tumor burden, higher grade, and poor prognosis, regardless of menopausal status. Despite the negative impact of obesity on clinical outcome, molecular mechanisms influencing breast cancer prognosis remain elusive. Methods: Postmenopausal women with invasive breast cancer (n=20) in the Clinical Breast Care Project who were obese at diagnosis (BMI ≥30) were matched by age, stage, ER status, HER2 status, and tumor grade to women who maintained a healthy weight (BMI = 18.5−24.9). Sections of pure invasive carcinoma were obtained by laser microdissection from flash-frozen or OCT embedded sections. Gene expression data using U133 2.0 microarrays was generated to compare genome-wide patterns of expression in breast tumors between obese and normal-weight patients. Data were analyzed using Partek Genomics Suite. Results: Principal Component Analysis accurately clustered the specimens into two groups, healthy weight or obese. Using a false-discovery rate (FDR) p-value Conclusions: Invasive breast tumors from obese women are genetically different from those of healthy-weight women. Altered gene expression may affect tumor cell proliferation and survival, contributing to aggressive phenotypes and altered immune response. These underlying molecular differences may contribute to the less favorable prognosis observed in obese women with breast cancer. Citation Information: Cancer Res 2011;71(24 Suppl):Abstract nr P1-09-01.

P2-01-09: Gene Expression Alterations in the Lymph Node Microenvironment in Response to Successful Metastatic Colonization.

Background: Breast stroma is known to play an active role in tumorigenesis, undergoing both phenotypic and molecular changes to facilitate and promote tumor development and growth. The metastatic microenvironment also plays a role in successful colonization; however, the genetic changes in these secondary microenvironments associated with metastasis are not well described. Methods: Women with invasive breast cancer with at least one lymph node with macrometastases and one lymph node with no detectable metastases were identified from the Clinical Breast Care Project. Lymph node tissue was microdissected from both the metastatic lymph node microenvironment and negative nodes and hybridized to U133A 2.0 gene expression arrays. Differential expression was detected using Partek® Genomics Suite™6.5 using a cutoff of P 2-fold change. Results: Nineteen genes were differentially expressed between negative lymph nodes and lymph node tissue microdissected from lymph nodes with metastatic tumors. Eleven genes, including EPCAM, KRT19 and MUC1 were expressed at significantly higher levels in lymph node tissue from metastatic lymph nodes while eight genes, such as CXCL2 and CXCL5, were expressed at significantly higher levels in negative lymph nodes. Results have been validated in external sample sets for AZGP1, CLEC4M, CXCL2, EPCAM, MUC1, PIP and TFPI. Conclusions: Lymph node tissue differs in gene expression between those harboring metastatic tumors and those without metastasis. Genes expressed at higher levels in lymph nodes with macrometastases are involved in tumorigenesis, suggesting that like the breast stromal microenvironment, the metastatic microenvironment undergoes crosstalk with the tumor cells. In addition, a cluster of genes involved in immune function are expressed at lower levels in metastatic lymph nodes, suggesting that suppression of proper immune response may be required for successful metastatic colonization. Citation Information: Cancer Res 2011;71(24 Suppl):Abstract nr P2-01-09.

Identification of breast cancer metastasis initiation and virulence genes.

CTRC-AACR San Antonio Breast Cancer Symposium: 2008 AbstractsnnAbstract #2049 nnBackground: Metastasis involves processes such as invasion, angiogenesis, intravasation, extravasation and eventual proliferation in the target organ. Metastasis initiation genes are those involved in dissemination of cells from the primary tumor while metastasis virulence genes contribute to metastatic colonization, specifically at the secondary site. Identification of metastasis initiation and virulence genes is critical to improve our understanding of and the ability to effectively treat metastatic breast cancer. Methods: A 51-gene signature was previously identified using RNA from 20 primary breast and matched metastatic lymph node tumors using the U133 2.0 microarrays (Affymetrix). Real-time PCR was performed for each assay using TaqMan assays (Applied Biosystems) using RNA from 25 pairs of primary breast and matched metastatic lymph node tumors. Data was evaluated using Mann-Whitney testing. Results: Gene expression levels were significantly different for 40 genes including 25 with P 50-fold difference in expression including WNT2, KRT14, TAC1, COL11A1, MMP13, GRP, and KERA with higher expression in primary tumors and EPHA3, PAX5 and NTS with higher expression in the metastasis. Discussion: A breast cancer metastasis signature involving 40 genes has been identified and validated. Genes expressed at higher levels in primary breast tumors are largely involved in degradation of the extracellular matrix (ECM), likely enabling cells with metastatic potential to disseminate. Decreased expression of these genes in the metastatic tumors suggests that once tumor cells have disseminated, invasion into foreign tissues does not require active tissue remodeling. Rather, cells that have successfully metastasized are characterized by the expression of genes involved in transcription, metabolism and immune response, potentially blocking cellular differentiation and providing cells with proliferation and survival advantages. These data not only improve our understanding of the biological processes involved in successful metastatic but provide new targets to arrest tumor cells dissemination and metastatic colonization.nnCitation Information: Cancer Res 2009;69(2 Suppl):Abstract nr 2049.