Abstract P4-10-01: High dose omega-3 fatty acid supplementation modulates breast tissue biomarkers in post-menopausal women at high risk for development of breast cancer

Background: We conducted a pilot study of high dose omega-3 fatty acid (FA) supplementation in post-menopausal women to determine if risk biomarkers for breast cancer in benign breast tissue sampled by random peri-areolar aspiration (RPFNA) could be favorably modulated and to acquire preliminary data on possible mechanism of action.

Methods: 35 post-menopausal women at increased risk for breast cancer were accrued to a trial of 6-month intervention with 4 g daily of omega-3-acid ethyl esters [1.86 g eicosapentaenoic acid (EPA), 1.5 g docosahexaenoic acid (DHA)]. Subjects had RPFNA performed pre- and post-intervention and specimens evaluated for cytomorphology and proliferation (Ki-67). FA composition was determined in plasma, red blood cells, and RPFNA specimens. Additional specimens were frozen for assessment of hormones, a panel of 11 adipokines and cytokines by Luminex, and gene expression.

Results: 34 subjects completed study with specimens evaluable for change in biomarkers. The ratio of (EPA+DHA):Arachidonic Acid (AA) levels in erythrocyte phospholipid increased significantly by a median of 2.7-fold. Although there was a significant decrease in blood EPA+DHA between discontinuation at 6 months and 2 weeks later when RPFNA was performed, all ratios were above the baseline value (median 1.6-fold). There was favorable but not statistically significant modulation for cytologic evidence of atypia (53% at baseline to 41% at off-study). However, favorable modulation was exhibited for Masood score (medians of 15 to 14; p = 0.014), number of epithelial cells recovered (p = 0.019) and Ki-67 expression (medians of 1.7% to 0.75%, p = 0.036, despite 8 subjects having no Ki-67 expression at baseline). Luminex assay of serum indicated a statistically significant increase (p = 0.003) for adiponectin and decrease (p = 0.016) for TNF-alpha between baseline and off-study. For RPFNA specimens, there was a significant decrease (P = 0.001) in MCP-1 levels adjusted for protein content. By ELISA, serum high molecular weight adiponection increased (p = 0.046) and molar ratio of IGF-1:IGFBP3 decreased (p = 0.006). Note that all analyses were exploratory and without correction for multiple analyses.

Conclusion: Favorable modulation of a variety of blood and tissue risk biomarkers, including cytomorphology and proliferation, along with good tolerability suggests that high dose omega-3 FA esters should be tested further in a placebo-controlled trial.

Citation Information: Cancer Res 2013;73(24 Suppl): Abstract nr P4-10-01.

Relationship of recent pregnancy and cytologic atypia in the contralateral breast of patients with breast cancer

149

Background: Although early parity protects against breast cancer later in life, it is a risk factor in the years following pregnancy, particularly when child-bearing is delayed. In a separate study, we have reported that these pregnancy-associated breast cancers (PABC) are more likely to be hormone receptor (HR) negative; we have previously found that random fine needle aspiration (rFNA) of the contralateral breast (CB) frequently have atypia in women with HR negative breast cancer. We now report an analysis of the cytological features of contralateral rFNA in relation to recent parity, with the goal of assessing cytological features which may be risk factors for PABC.

Methods: Women with breast cancer undergoing surgery between 2006 and 2008 were enrolled in a prospective study of rFNA of the CB. Cytological analysis was performed on all the samples, using two scoring systems, Masood and Zalles (M and Z). Linear regression analyses were performed relating each score to interval from last pregnancy, stratified for parity, and adjusted for age.

Results: Eighty-two patients had rFNA and cytologic analysis. The parous group’s (n=52) interval since last pregnancy was inversely correlated with the overall M score (p=0.034); pleomorphism and (p=0.047 and p=0.013 respectively). M score was also related to age at last pregnancy (p=0.026) as were pleomorphism and chromatin pattern (p=0.044 and 0.035 respectively). In the nulliparous group (n=30), there were no significant relationships between any of the cytological parameters or age.

Conclusions: The CB of women with recent parity display specific cytological abnormalities that are related to risk of breast cancer, particularly HR negative breast cancer. This observation affords the opportunity to further evaluate rFNA as a tool for breast cancer risk assessment following pregnancy, and to identify molecular correlates of nuclear cytological abnormalities that may serve as targets for prevention of ER negative breast cancer.

Abstract PD09-04: Weight Loss in Postmenopausal Women Is Associated with Modulation of Serum and Tissue Based Risk Biomarkers

Background

In a pilot study, a structured program of reduced energy diet, physical activity, and weekly group behavioral intervention with other high risk women was successful at producing a median 11% weight loss with at least a 5% weight loss in 88% of subjects. We evaluated the association of weight loss with changes in serum and breast tissue risk and mechanisms of action biomarkers. Methods

High risk postmenopausal women with BMI >25 kg/m2 had breast tissue harvested by random periareolar fine needle aspiration (RPFNA) before and after a 6-month energy balance intervention. Specimens were evaluated for biomarkers including cytomorphology, proliferation (immunocytochemical Ki-67), gene expression by RT-qPCR, and expression of cytokines and adpokines by Luminex assay. Fasting serum was assayed for insulin, glucose, adiponectin, leptin, high sensitivity CRP, IL-6, prolactin, SHBG, estradiol and testosterone using ELISA or Luminex.

Results

For 24 biomarker evaluable subjects, 21 had >5% weight loss (median = 11%). Cytologic atypia was present in 10/24 at baseline and 4/24 at 6 months (p=0.034). For 20 subjects with sufficient cells for assessment of Ki-67 at both times, median baseline Ki-67 was 0.7% and off study 0.3%, with a median change of -0.2% (p=0.19). Statistically significant changes (≥0.003; Wilcoxon) were observed for serum levels of adiponectin, adiponectin:leptin ratio, and SHBG (increases); and leptin, bioavailable estradiol and hsCRP (decreases). Reduction was also observed for insulin (p=0.018) and bioavailable testosterone (p=0.033). These results were duplicated (p≥0.014) by Luminex for adiponectin, leptin, adiponectin: leptin ratio, and insulin; plus hepatocyte growth factor (HGF, decrease). Also, an increase in the adiponectin:leptin ratio was observed for the RPFNA specimens (p=0.012). Gene expression (RT-qPCR) of pS2 was significantly modified (decrease, p=0.035). Further, the weight loss (expressed as relative change) was highly statistically correlated with change (relative) in serum leptin, adiponectin:leptin ratio, SHGB, and free estradiol; as well as with relative change in adiponectin:leptin ratio in RPFNA specimens. Conclusion

Weight loss in high risk postmenopausal women is accompanied by significant modulation of numerous serum and breast tissue-based biomarkers. For several risk and response biomarkers there is a significant correlation between change in the biomarker and the weight loss achieved. This suggests the possibility of identifying mechanisms of action and signaling pathways for dietary/energy balance interventions that may reduce risk for development of breast cancer.

Citation Information: Cancer Res 2010;70(24 Suppl):Abstract nr PD09-04.

Evaluation of Ki-67 measured in benign breast tissue acquired from premenopausal women treated with a flaxseed derivative

1507

Background: The lignans enterolactone and enterodiol are derived from the action of gut bacteria on ingested secoisolariciresinol diglycoside (SDG) which is commonly found in flaxseed. Enterolactone and enterodiol are thought to impair mammary carcinogenesis via reduction in aromatase activity and the mid-cycle surge of luteinizing hormone. We assessed the modulatory activity of 1 year of SDG on a number of risk biomarkers for breast cancer in a prospective phase II pilot study. We report the effect of SDG on the primary endpoint, proliferation in benign breast tissue as measured by Ki-67 immunocytochemistry, in the first 35 women completing study. Methods: Premenopausal women age 21 to 55 at increased risk for breast cancer underwent a baseline random periareolar fine needle aspiration (RPFNA) between the first and tenth days of their menstrual cycle. Those with RPFNA evidence of hyperplasia and Ki-67 greater than or equal to 2% were invited to participate. Women taking flaxseed or oral contraceptives were ineligible. All women took one Brevail (lignan research) capsule containing 50 mg of SDG daily. Ki-67 staining was performed with DAKO M7240 antibody on hematoxylin counterstained slides and the number of cells staining positive in 500 cells within hyperplastic clusters was counted. Results: Forty-nine women were enrolled on study between February 2006 and June 2008. Of these, four stopped prematurely, 10 women have not completed, and 35 have completed study and undergone follow-up RPFNA. Baseline characteristics of the 35 women completing study are as follows: median age 44 (range 29–50), median baseline 5-year Gail model risk 1.6% (range 0.1%-5.7%), median Ki-67 4.2% (range 2.0%-16.8%). Thirty seven percent had hyperplasia without atypia, and 63% had atypia. At repeat RPFNA, Ki-67 expression was reduced (median value of 2.0%, range 0%-15.2%); with 29 of the 35 women demonstrating a decrease (median relative reduction of 0.70). Conclusions: Based upon reduction in Ki-67 expression in hyperplastic benign breast tissue after 12 months, 50 mg of SDG administered daily as Brevail appears promising as a preventive. Supported in part by grant R21 CA117847 from the National Cancer Institute.

No significant financial relationships to disclose.

Phase II breast cancer chemoprevention trial of the third generation selective estrogen receptor modulator arzoxifene

1001

Background: Arzoxifene (ARZ) is a third generation SERM with efficacy in metastatic breast cancer but lacking uterine agonist activity. Methods: We conducted a randomized, double-blind, placebo-controlled Phase II prevention trial in 199 high risk women assessing the effects of ARZ 20 mg/day on several risk biomarkers. Biomarkers, including cytomorphology of breast epithelial cells obtained by random periareolar FNA (RPFNA) were assessed at baseline and following 6 months of placebo or ARZ. Subjects were stratified by presence or absence of atypia, ER expression, BRCA1/2 mutation, as well as menopause status. Results: At entry, mean age was 47, 52% were premenopausal and 47% of postmenopausal women were taking HRT. Mean 10 year Gail was 6.8% and mean Masood cytology index score was 14.3. The ARZ and placebo groups were well matched. The protocol defined primary endpoint was a decrease in RPFNA cytology Masood index score by ≥3 points at 6 months and required 160 evaluable subjects for 81% power to detect a change from 30% to 52% of subjects showing improvement. For the 181 evaluable subjects, there was no significant difference in the proportion of women achieving ≥3 point improvement (19% placebo vs. 24% ARZ, p=0.46); or in change in mean index score (0.6 placebo vs. 0.9 ARZ, p=0.53). There was also no difference in grade 3 or 4 side effects or dropout prior to 6 months. However, comparing ARZ to placebo, there was favorable modulation of the two risk biomarkers, mammographic breast density (p=0.001) and IGF-1:IGFBP-3 ratio (p=0.001), and reduction in bone turnover biomarker osteocalcin (p= 0.002), but without an increase in endometrial thickness. Conclusions: Although improvement in cytomorphology after 6 months of ARZ was not shown, the acceptable side effect profile and favorable modulation of other biomarkers (breast density, IGF-1:IGFBP-3, osteocalcin) provides support for continued evaluation of ARZ as a breast cancer prevention agent.

No significant financial relationships to disclose.

Short-term breast cancer prediction by random periareolar fine-needle aspiration cytology and the Gail risk model

BACKGROUND

: Biomarkers are needed to refine short-term breast cancer risk estimates from epidemiologic models and to measure response to prevention interventions. The purpose of our study was to determine whether the cytologic appearance of epithelial cells obtained from breast random periareolar fine-needle aspirates or molecular marker expression in these cells was associated with later breast cancer development.

METHODS

: Four hundred eighty women who were eligible on the basis of a family history of breast cancer, prior precancerous biopsy, and/or prior invasive cancer were enrolled in a single-institution, prospective trial. Their risk of breast cancer according to the Gail model was calculated, and random periareolar fine-needle aspiration was performed at study entry. Cells were characterized morphologically and analyzed for DNA aneuploidy by image analysis and for the expression of epidermal growth factor receptor, estrogen receptor, p53 protein, and HER2/NEU protein by immunocytochemistry. All statistical tests are two-sided.

RESULTS

: At a median follow-up time of 45 months after initial aspiration, 20 women have developed breast cancer (invasive disease in 13 and ductal carcinoma in situ in seven). With the use of multiple logistic regression and Cox proportional hazards analysis, subsequent cancer was predicted by evidence of hyperplasia with atypia in the initial fine-needle aspirate and a 10-year Gail projected probability of developing breast cancer. Although expression of epidermal growth factor receptor, estrogen receptor, p53, and HER2/NEU was statistically significantly associated with hyperplasia with atypia, it did not predict the development of breast cancer in multivariable analysis.

CONCLUSION

: Cytomorphology from breast random periareolar fine-needle aspirates can be used with the Gail risk model to identify a cohort of women at very high short-term risk for developing breast cancer. We recommend that cytomorphology be studied for use as a potential surrogate end point in prevention trials.