Gail Score Research Articles

Breast MRI Surveillance in Women with Prior Curative-Intent Therapy for Breast Cancer

We sought to determine factors that predict the use of breast magnetic resonance imaging (MRI) surveillance in women previously treated for breast cancer and the incidence of in-breast recurrences and/or new cancers identified by MRI. We reviewed 141 patients who were treated between 2005 and 2008 who also underwent surveillance breast MRI following their treatment. Patient, tumor, treatment characteristics, and MRI findings were collected. Gail scores were calculated based on the patient's personal and family history prior to the breast cancer diagnosis. Data were compared using chi(2) and Fisher's exact test. The average age of the study population was 51 (range 24-73). One hundred forty-one women underwent 202 surveillance breast MRIs during the study period. Sixteen of 141 (11%) required second look imaging, and six of 141 (4%) required biopsy of suspicious lesions. Two of the six were invasive breast cancers, while four were benign. Overall, the rate of new cancer detection on surveillance MRI during the study period was 0.9% (two of 202 imaging studies). Of the 71 women with evaluable Gail scores, the average lifetime risk score was 16.7%. Eight patients had BRCA mutations and three previously underwent irradiation for Hodgkin's lymphoma. Patient age, Gail score, tumor stage, grade, histology, receptor status, and surgical treatment were not predictive of MRI surveillance use. Prospective studies are needed to determine which patients may potentially benefit from breast MRI surveillance following curative-intent treatment. The lack of standardized guidelines may result in excessive or inappropriate use, unnecessary follow-up procedures, and a concomitant low yield.

Mammographic density does not differ between unaffected BRCA1/2 mutation carriers and women at low-to-average risk of breast cancer

Elevated mammographic density (MD) is one of the strongest risk factors for sporadic breast cancer. Epidemiologic evidence suggests that MD is, in part, genetically determined; however, the relationship between MD and BRCA1/2 mutation status is equivocal. We compared MD in unaffected BRCA1/2 mutation carriers enrolled in the U.S. National Cancer Institute's Clinical Genetics Branch's Breast Imaging Study (n = 143) with women at low-to-average breast cancer risk enrolled in the same study (n = 29) or the NCI/National Naval Medical Center's Susceptibility to Breast Cancer Study (n = 90). The latter were BRCA mutation-negative members of mutation-positive families or women with no prior breast cancer, a Pedigree Assessment Tool score <8 (i.e., low risk of a hereditary breast cancer syndrome) and a Gail score <1.67. A single experienced mammographer measured MD using a computer-assisted thresholding method. We collected standard breast cancer risk factor information in both studies. Unadjusted mean percent MD was higher in women with BRCA1/2 mutations compared with women at low-to-average breast cancer risk (37.3% vs. 33.4%; P = 0.04), but these differences disappeared after adjusting for age and body mass index (34.9% vs. 36.3%; P = 0.43). We explored age at menarche, nulliparity, age at first birth, menopausal status, number of breast biopsies, and exposure to exogenous hormonal agents as potential confounders of the MD and BRCA1/2 association. Taking these factors into account did not significantly alter the results of the age/body mass index-adjusted analysis. Our results do not provide support for an independent effect of BRCA1/2 mutation status on mammographic density.

Preventive Health Behaviors in Unaffected High-Risk Women: The Impact of Perceived vs Actual Risk and Preferred Involvement in Decision-Making.

Abstract Background: While major advances have been made in the diagnosis and treatment of breast cancer, more limited progress has been made in the prevention of the disease. Although several clinical trials have demonstrated the advantages of lifestyle alteration, weight loss and use of anti-estrogens, uptake of such strategies is generally sporadic in women at high risk. It has been particularly challenging to engage even women at very high risk in clinical trials aimed at primary prevention. These decisions may be influenced by several factors, among which perceived risk, actual risk and preferences for participation in preventive decision making may be particularly important. In this project, we seek to describe how perceived risk, actual risk and the preferred level of involvement in the decision jointly impact the risk management intentions and subsequent decisions.Methods: As a regular component of risk assessment process at the High Risk Breast Assessment Clinic (HRBAC) of the Ottawa Regional Women's Breast Health Centre, a detailed questionnaire is administered to women referred to the clinic before their first risk assessment consultation. The questionnaire includes several items required for calculating actual risk (using Gail score) as well as questions concerning health practices (breast screening, clinical breast examination, breast self-examination) and lifestyle practices (weight, height, smoking, alcohol and physical activity). Women's perceived risk is assessed by the following question: “What do you think the likelihood is of you developing breast cancer in your lifetime?”, and women are required to provide a percentage to express their risk. Women are also asked to indicate their intentions about breast cancer prevention and to identify which prevention strategy is the most important to them. Preferred role involvement in decision making is also assessed by the following question: “What role would you like to take in making your decision?”and women are classified as active or passive. Patient's charts are used to obtain information about the prevention decisions that women made 1 year after receiving risk counselling.We hypothesize that risk reduction counselling will weaken the association between perceived risk and prevention decisions and conversely will strengthen the correlation between actual risk and risk reduction option considered by women.Results: Correlational and chi-square analyses of the demographic data, actual(calculated0 and perceived risk, as well as prevention behaviour uptake obtained from 300 patients will be presented. A low correlation between perceived and actual risks is expected before patients receive risk counselling at the clinic. Statistically reliable association between perceived risk and precounselling prevention intentions and a low level association between actual risk and precounselling prevention intentions are also expected. Citation Information: Cancer Res 2009;69(24 Suppl):Abstract nr 1040.

Effects of Arzoxifene on Breast Cancer Incidence in Postmenopausal Women with Osteoporosis or with Low Bone Mass.

Abstract Precinical and early clinical data indicate that arzoxifene, a third generation SERM, is more potent and bioavailable than raloxifene. Here we report the results of the phase 3, multicenter, placebo-controlled, double-blind GENERATIONS trial of 9354 postmenopausal women with osteoporosis (N=5252) or low bone mineral density (BMD) (N=4102), randomly assigned to arzoxifene 20mg/d (N=4676) or placebo (N=4678). The primary endpoints were the incidences of radiographic vertebral fracture in the osteoporotic population (at 36 months) and invasive breast cancer in all study participants (after all participants had completed 48 months of treatment). Breast cancers were detected by annual mammograms and/or clinical breast examination and adjudicated centrally by an independent committee. The primary analysis compared the incidence of adjudicated invasive breast cancer between treatment groups using a log-rank test. Both treatment groups were evenly balanced for risk of breast cancer including age, previous benign breast biopsies, family history of breast cancer, estimated Gail risk and BMD. The trial met both primary endpoints with a 41% reduction in the incidence of vertebral fractures (p&amp;lt;0.001) and 56% percent reduction in incidence of invasive breast cancer (p=0.002). Table 1 summarizes the breast cancer findings after the 48 month follow-up. The incidence of invasive breast cancer in the placebo group was higher in the women who had a Gail score &amp;gt; 1.66 compared to ≤ 1.66 (1.5% vs. 0.6%) and in those with low bone mass versus osteoporosis (1.2% vs. 0.7%), but the risk reduction with arzoxifene was similar between Gail risk groups (HR 0.32 vs 0.57. p-value for interaction=0.31) and between low bone mass and osteoporosis groups (HR 0.34 vs 0.57. p-value for interaction=0.35). Other findings included no significant reduction in non-vertebral fractures or cardiovascular events. Generally, arzoxifene was well tolerated, although there was a significant increase in venous thromboembolism, gall bladder disease, pulmonary obstructive/infective disorders, hot flushes, muscle cramps and gynecological-related events in the arzoxifene group. We conclude that arzoxifene reduces the incidence of invasive breast cancer as well as vertebral fracture. The results of the GENERATIONS trial provide further support for a significant risk reduction of invasive breast cancer by SERMs in post menopausal women. Citation Information: Cancer Res 2009;69(24 Suppl):Abstract nr 51.

Multiple Breast Cancer Risk Factors Do Not Improve Patient Acceptance of Chemoprevention.

Abstract Background: Patients may be at increased risk of developing breast cancer due to hereditary or familial risk or a pathological diagnosis of breast atypia or lobular carcinoma in situ (LCIS). Although tamoxifen and raloxifene are effective and approved chemopreventive agents to reduce risk of developing breast cancer in high risk individuals, acceptance of chemoprevention as a risk reduction strategy is low. Our objective was to assess whether chemoprevention is more readily accepted by those at genetic or familial risk versus pathologic risk, and if the presence of multiple breast cancer risk factors improves chemoprevention acceptance.Methods: All patients seen at the Loyola University Cancer Risk Assessment and Prevention clinic completed information including demographics, health, family and social history. Patients with a family history of either a first or second degree family member with breast cancer (FH) were compared to those who presented due to atypia or LCIS on breast biopsy as well as those with an elevated Gail score. Acceptance of risk reducing strategies between these groups was compared.Results: The initial 115 patients presenting to the clinic were analyzed. Thirty-six (31%) women had FH only, 26 (23%) had a biopsy with atypia or LCIS only, 10 (9%) had an elevated Gail score, 17 (15%) had a biopsy indicating atypia or LCIS plus FH, and 26 (23%) had an elevated Gail plus FH. Seventy-three patients were candidates for chemoprevention; 16 (22%) accepted chemoprevention. Patients with a FH plus atypia/LCIS did not accept chemoprevention more than patients with FH alone (p=1.00). Patients with a FH plus atypia/LCIS did not accept chemoprevention more than than patients with atypia/LCIS alone (p=0.451). Pts with elevated Gail plus FH did not accept chemoprevention more than patients with elevated Gail alone (p=0.144). Four women in the FH group were known BRCA mutation carriers; all elected to proceed with risk reducing surgery, none accepted chemoprevention.Conclusions: Acceptance of chemoprevention is low in women at genetic or familial risk, those at risk due to atypia/LCIS, and those with an elevated Gail score. Having both familial and pathologic risk factors for breast cancer development did not increase acceptance of chemoprevention. Better understanding of women's comprehension of breast cancer risk, and barriers to patient acceptance of chemoprevention are needed. Citation Information: Cancer Res 2009;69(24 Suppl):Abstract nr 1039.

Chemoprevention Indication Score: A user-friendly tool for prevention of breast cancer – Pilot analysis

BackgroundDespite the results of prospective randomized placebo controlled studies, endorsement from various professional societies, and approval by the FDA, the chemoprevention of breast cancer is limited. This is attributable to the perceived risks of complications with tamoxifen. Individualized risk–benefit calculation regarding the use of tamoxifen is burdensome for practical clinical use. We propose a Chemoprevention Indication Score (CIS) that is easy to compute and reliable to identify women suitable for chemoprevention. Material and methodsChart-review of all patients attending a university-based high-risk breast clinic identified the women offered chemoprevention and those who accepted it. Age, Gail risk score, past medical history and physician's reasons for not offering tamoxifen were recorded. CIS was developed weighing risks and benefits of tamoxifen; high, moderate and low indication score categories were defined for recommendation of tamoxifen. CIS was validated by cross-tabulating with the rate of physician recommendation of tamoxifen and by agreement with Gail's risk–benefit index. ResultsBetween 2004 and 2006,105 women attended the high-risk clinic. Median (interquartile) age was 47 (42–53) yrs; 142(90%) were Caucasian. Median (interquartile) Gail score was 2.3 (1.0–3.2). Forty-eight (46%) women were offered tamoxifen while 15(14%) complied. Tamoxifen was offered to 1 of 5(20%) women with low; 32 of 82(39%) women with moderate and 15 of 18(83%) women with high CIS [p=0.0008]. The McNemar's test for agreement between CIS (<6 vs. ≥6) and Gail's risk–benefit index was significant at p<0.0001 and Kruskal–Wallis test comparing median Gail's risk–benefit index across CIS was significant at p<0.0001. ConclusionInability to identify appropriate candidates has been a great barrier towards acceptance of chemoprevention for breast cancer. The CIS can be used for individual risk–benefit analysis for recommendation of breast cancer chemoprevention. However, CIS needs to be validated on a larger scale with methodologically more rigorous studies before proposing generalized use in the community.

Underestimation of Breast Cancer Risk: Influence on Screening Behavior

To describe perceived breast cancer risk, identify the percentage of women with inaccurate risk perceptions, and examine the influence of perceived and objective risk on screening behavior. Descriptive, correlational, cross-sectional. Community settings in a metropolitan area on the western coast of the United States. Multicultural sample of 184 English-speaking women (57% non-Caucasian, X age = 47 +/- 12 years) who have never been diagnosed with cancer. Two perceived risk scales (verbal and comparative) and the Gail model were used to assess perceived and objective breast cancer risk, respectively. Perceived breast cancer risk, objective breast cancer risk, screening behavior. Participants reported that they "probably will not" get breast cancer and that their risk was "somewhat lower" than average. Family history of breast cancer was a significant predictor of perceived risk. Demographic characteristics and objective risk factors were not associated with perceived risk. Most women at high risk for breast cancer (89%) underestimated their actual risk; fewer women with low to average risk for breast cancer (9%) overestimated their risk. Age, Gail scores, and health insurance status promoted breast cancer screening; underestimation of risk had the opposite effect. Inaccurate perceptions of risk do not promote optimal breast cancer screening. The finding has implications for most women at high risk for developing breast cancer who underestimate their risk. Oncology nurses can use risk assessment tools to provide individualized counseling regarding breast cancer risk factors and screening. Women at high risk who underestimate their risk could benefit from additional screening and from advances in cancer chemoprevention.

Differences in reproductive risk factors for breast cancer in middle-aged women in Marin County, California and a sociodemographically similar area of Northern California

BackgroundThe Northern California county of Marin (MC) has historically had high breast cancer incidence rates. Because of MC's high socioeconomic status (SES) and racial homogeneity (non-Hispanic White), it has been difficult to assess whether these elevated rates result from a combination of established risk factors or other behavioral or environmental factors. This survey was designed to compare potential breast cancer risks and incidence rates for a sample of middle-aged MC women with those of a demographically similar population.MethodsA random sample of 1500 middle-aged female members of a large Northern California health plan, half from Marin County (MC) and half from a comparison area in East/Central Contra Costa County (ECCC), were mailed a survey covering family history, reproductive history, use of oral contraceptives (OC) and hormone replacement therapy (HRT), behavioral health risks, recency of breast screening, and demographic characteristics. Weighted data were used to compare prevalence of individual breast cancer risk factors and Gail scores. Age-adjusted cumulative breast cancer incidence rates (2000–2004) were also calculated for female health plan members aged 40–64 residing in the two geographic areas.ResultsSurvey response was 57.1% (n = 427) and 47.9% (n = 359) for MC and ECCC samples, respectively. Women in the two areas were similar in SES, race, obesity, exercise frequency, current smoking, ever use of OCs and HRT, age at onset of menarche, high mammography rates, family history of breast cancer, and Gail scores. However, MC women were significantly more likely than ECCC women to be former smokers (43.6% vs. 31.2%), have Ashkenazi Jewish heritage (12.8% vs. 7.1%), have no live births before age 30 (52.7% vs. 40.8%), and be nulliparous (29.2% vs. 15.4%), and less likely to never or rarely consume alcohol (34.4% vs. 41.9%). MC and ECCC women had comparable 2000–2004 invasive breast cancer incidence rates.ConclusionThe effects of reproductive risks factors, Ashkenazi Jewish heritage, smoking history, and alcohol consumption with regard to breast cancer risk in Marin County should be further evaluated. When possible, future comparisons of breast cancer incidence rates between regions should adjust for differences in income and education in addition to age and race/ethnicity, preferably by using a sociodemographically similar comparison group.

Mammographic density does not differ between unaffected BRCA1/2 mutation carriers and women at low-to-average risk of breast cancer.

Abstract Abstract #5002 Background: Mammographic density (MD) is one of the strongest risk factors for sporadic breast cancer. Since MD is a non-invasive, reproducible and quantitative measure, it has been proposed as a compelling intermediate marker in studies aimed at understanding breast cancer etiology and prevention. The mechanism by which MD influences breast cancer risk is unknown. However, epidemiologic evidence suggests that dense breast tissue may be, in part, genetically regulated. Increased MD has been positively associated with family history of breast cancer. Family studies of sisters, including twins, provide further support for a genetic influence on MD. The relationship between MD and BRCA1/2 mutation status is equivocal.&amp;#x2028; Methods: We examined whether MD differed between unaffected BRCA1/2 mutation carriers ages 22-55 enrolled in the Clinical Genetics Branch's Breast Imaging Study (n=143) and women at low-to-average risk of breast cancer ages 25-79 enrolled in the same study (n=29) or the National Naval Medical Center's Susceptibility to Breast Cancer Study (n=90). Low-to-average risk women were defined as being BRCA mutation-negative members of mutation-positive families; or women with no personal history of breast cancer, a Pedigree Assessment Tool score &amp;lt;8 (i.e., low risk of a hereditary breast cancer syndrome) and a Gail score &amp;lt;1.7. Participants from both studies had standardized, quantitative calculations of MD measured by the same experienced radiologist (CKC). Percent MD was calculated from digitized mammograms using a computerized method developed at NCI. Standard breast cancer risk factor information was collected from all subjects.&amp;#x2028; Results: MD values were normally distributed. The intraclass correlation coefficient for MD assessed blindly in 100 paired sets was 0.889. The mean age of BRCA mutation carriers was 38.0 vs. 48.4 in low-to-average risk women. The unadjusted mean percent MD was higher in women with BRCA1/2 mutations compared with women at low-to-average breast cancer risk (37.3% vs. 33.4%; p=0.04). There was no difference in MD after adjusting for age and body mass index (35.1% vs. 36.3%; p=0.48). We explored age at menarche, nulliparity, age at first birth, menopausal status, number of breast biopsies, and exposure to exogenous hormonal agents as covariates of potential interest with regard to modulating MD. Taking these factors into account did not substantially alter the results of the age-/BMI-adjusted analysis.&amp;#x2028; Conclusion: Our results do not provide support for an independent effect of BRCA1/2 mutation status on mammographic density. We are currently investigating whether novel quantitative characteristics imbedded within the breast parenchyma in mammographic images (e.g., texture and contrast) may have a stronger correlation with mutation status. Our goal is to identify imaging features which might more accurately classify women with regard to their breast cancer risk status. Citation Information: Cancer Res 2009;69(2 Suppl):Abstract nr 5002.

Assessment of breast cancer risk prediction models in a high-risk population.

Abstract Abstract #4074 Background: A variety of risk assessment models have been developed to quantify a woman's risk for developing breast cancer. Although the Gail model (GM) is the most widely utilized model, there are limitations. The Tyrer-Cuzick model (TCM), which has been validated in the United Kingdom (UK), addresses these limitations but its performance in other populations is unclear. The purpose of this study was to evaluate the predictive ability of the TCM in a cohort of high-risk women from the New York metropolitan area and compare the results to the GM.&amp;#x2028; Methods: The Women At Risk (WAR) Registry provided the study population. Due to the age limitation of the Gail model, we excluded women who were under 35 and over 80 years of age. Calculation of lifetime Gail scores included the following: age, race, age at first menses, age at first live birth, number of first degree relatives with breast cancer, number of previous breast biopsies, and atypical hyperplasia. These variables were also included in the lifetime Tyrer-Cuzick scores as well as information on menopausal status, use of hormone replacement therapy, LCIS, and age at diagnosis for relatives with ovarian and breast cancer. The Receiver Operating Characteristic (ROC) and estimated area under the ROC curve (AUC) with 95% confidence intervals (CI) were used to assess prediction accuracy of both models.&amp;#x2028; Results: Out of a total study population of 1523 women, 82 (5%) developed breast cancer during a median follow-up of 56 months. The median age of women without breast cancer was 49 years, and the median age of women who developed breast cancer was 54 years. The AUC and 95% CI for the Gail model was 0.547 (0.479 to 0.615) and 0.501 (0.433 to 0.569) for the Tyrer-Cuzick model. Our results indicate poor discriminatory accuracy for both models, which performed no better than pure chance (0.5).&amp;#x2028; &amp;#x2028; Discussion: Quantitative breast cancer risk assessment models are critically important in developing effective risk management strategies. Although the TCM addresses some limitations of the GM, it is unclear how well it performs in populations outside the UK due to the wide variation in international breast cancer rates. In addition, the TCM excludes more recently identified risk factors such as breast density. Bayesian techniques may ultimately lead to a better risk assessment tool as they allow continual modification of the model as new risk factors are identified. Accurate breast cancer risk prediction is a complex undertaking and further studies are warranted in order to achieve truly comprehensive and widely applicable models. Citation Information: Cancer Res 2009;69(2 Suppl):Abstract nr 4074.

Breast Cancer Risk Factors in Younger and Older Women

Information is lacking regarding the interaction of established breast cancer risk factors and patient age. We attempted to study this interaction in high-risk women at the extremes of age in our population. The Women-At-Risk Registry was queried for women who were < or =35 and > or =70 years of age. Enrollment criteria included: strong family history of breast cancer (FHBC), and/or biopsy-proven history of atypical ductal hyperplasia (ADH), atypical lobular hyperplasia (ALH), or lobular carcinoma in situ (LCIS). Descriptive analyses and Fisher's exact tests were used to analyze these factors and to assess their influence on breast cancer development. Our population included 1,412 high-risk women with median follow-up of 4 years. Of 195 women < or =35 years, 3 (1.5%) developed breast cancer. All three had strong FHBC and none had a prior high-risk lesion. Of 82 women > or =70 years, 6 (7.3%) developed breast cancer. Mean Gail score for women > or =70 years was 4.3, as compared with 4.7 in the subset of older women diagnosed with cancer. Fisher's tests demonstrated that ADH (p = 0.15), ALH (p = 1.0), LCIS (p = 1.0), and FHBC (p = 1.0) were not associated with breast cancer development in older women. We conclude that, for women < or =35 years, a significant FHBC may be a stronger predictor for breast cancer development than high-risk lesions. For women > or =70 years, FHBC and history of ADH, ALH, and LCIS were not predictors of breast cancer. This study emphasizes the importance of defining age-appropriate recommendations for breast cancer risk management, including surveillance and chemoprevention.

Vascular endothelial growth factor and breast cancer risk

Vascular endothelial growth factor (VEGF) is a key factor in angiogenesis and is important to carcinogenesis. Previous studies relating circulating levels of VEGF to breast cancer have been limited by small numbers of participants and lack of adjustment for confounders. We studied the association between serum VEGF and breast cancer in an unmatched case-control study of 407 pre- and postmenopausal women (n = 203 cases, n = 204 controls). Logistic regression was used to model the breast cancer risk as a function of natural log transformed VEGF levels adjusted for age, Gail score, education, physical activity, history of breastfeeding, serum testosterone, and hormone therapy (HT) use. The majority of the population was postmenopausal (67.6%) and the average age was 56 years; age and menopausal status were similar among cases and controls. Geometric mean VEGF levels were non-significantly higher in cases (321.4 pg/ml) than controls (291.4 pg/ml; p = 0.21). In a multivariable model, the odds of breast cancer was 37% higher for women with VEGF levels > or =314.2 pg/ml compared to those with levels below 314.2 pg/ml, albeit not significantly (p = 0.16). There was no interaction between VEGF and menopausal status (p = 0.52). In this case-control study, VEGF was not significantly associated with breast cancer risk in pre- and postmenopausal women.

Abstract B65: Atypia in ductal lavage: A reproducible marker of risk?

Abstract B65 Background The majority of breast cancers originate in the ductal epithelium of the breast. Ductal lavage provides a minimally invasive tool for assessing epithelial cells and their microenvironment. Previous studies of pathological atypical hyperplasia on biopsy, and cytological aypia in nipple aspirate fluid (NAF) and random fine needle aspiration (FNA) have suggested that these cells are precancerous because of the subsequent increased risk of cancer. It has therefore been hypothesized that serial ductal lavage could be used to identify atypia and monitor the effects of chemoprevention over time. This report is part of a larger study examining the differences in hormone levels and cytology between ducts and their reproducibility. This report presents a cohort of women unselected for risk who voluntarily underwent ductal lavage of the same ducts on two occasions six months apart. Samples were analyzed for epithelial cells and macrophages. Methods A total of 107 women were recruited in this IRB-approved prospective study to undergo ductal lavage of three fluid or non-fluid-yielding ducts within a breast. Women were unselected for risk: the average Gail score was 1.67 and ten women had a history of contralateral breast cancer. The average age was 51 years. Fifty-four women were premenopausal, 4 were perimenopausal and 49 were postmenopausal. Samples were submitted in a blinded fashion to a central laboratory and were analyzed by duct for both macrophage and epithelial cell number on Papanicolaou-stained slides. If adequate epithelial cells were present (&amp;gt;10), a diagnosis of each duct (benign, mild atypia, or marked aytpia) was made by an experienced cytopathologist. Results Of the 107 women, 317 (99%) ducts were successfully cannulated without regard to nipple aspirate fluid (NAF) production. Thirty ducts in 23 women were diagnosed with mild or marked atypia on either of the two visits. Of these ducts, 16 were diagnosed at the initial visit and 13 at the second visit. Only one duct was diagnosed with mild atypia on both visits, in a woman with a Gail score of 1.4. Cellular atypia in ductal lavage was not reproducible at a six month interval (p=0.45). This absence of reproducibility was seen in women at all risk levels. Macrophage and epithelial cell numbers also showed a lack of reproducibility at the six month interval (p=0.31 and p=0.22, respectively). Conclusion Atypia on cytology in ductal lavage in women unselected for risk was not reproducible at a six month interval, suggesting that a single diagnosis of atypia on lavage should not be used as a marker for the identification of women at risk. Further studies are necessary to indicate whether persistent atypia over an extended interval could represent an elevated risk for breast cancer. In addition, ongoing studies are being performed to identify better cellular markers of premalignancy and subsequent risk. Citation Information: Cancer Prev Res 2008;1(7 Suppl):B65.

Hip bone density predicts breast cancer risk independently of Gail score

The Gail model has been commonly used to estimate a woman's risk of breast cancer within a certain time period. High bone mineral density (BMD) is also a significant risk factor for breast cancer, but it appears to play no role in the Gail model. The objective of the current study was to investigate whether hip BMD predicts postmenopausal breast cancer risk independently of the Gail score. In this prospective study, 9941 postmenopausal women who had a baseline hip BMD and Gail score from the Women's Health Initiative were included in the analysis. Their average age was 63.0 +/- 7.4 years at baseline. After an average of 8.43 years of follow-up, 327 incident breast cancer cases were reported and adjudicated. In a multivariate Cox proportional hazards model, the hazards ratios (95% confidence interval [95% CI]) for incident breast cancer were 1.35 (95% CI, 1.05-1.73) for high Gail score (>or=1.67%) and 1.25 (95% CI, 1.11-1.40) for each unit of increase in the total hip BMD T-score. Restricting the analysis to women with both BMD and a Gail score above the median, a sharp increase in incident breast cancer for women with the highest BMD and Gail scores was found (P < .05). The contribution of BMD to the prediction of incident postmenopausal breast cancer across the entire population was found to be independent of the Gail score. However, among women with both high BMD and a high Gail score, there appears to be an interaction between these 2 factors. These findings suggest that BMD and Gail score may be used together to better quantify the risk of breast cancer.

Limitation of the Gail score and mammographic density in recruiting candidate for prevention of breast cancer in high-risk women with precancerous lesions

1548 Background: The purpose of our ongoing prospective Breast Cancer (BC) Prevention Trial using chemoprevention in high-risk women with precancerous lesions, hyperplasia usual type (HUT) or atypical hyperplasia (AH) was to refine the methodology in recruiting more accurately the candidates. Methods: A total of 282 postmenopausal women following conventional first risk assessment were entered in an ongoing prospective trial, in a single institution, N.N. Blokhin Cancer Research Center, Moscow, Russia from 10/2005 to 11/2007. Following prior biopsies for HUT or AH patients were measured for Gail Score (GS) and mammographic density (MD). Results: Following fine needle aspiration biopsy (FNAB) 149 patients were classified into five categories. The FNAB cell number count is described in terms of the median (range) and proportion of cells (POC) >3000 (95% confidence interval). The results follow: Category 1: HUT, ER(+), 29 patients, median=3,427 (2,200–5,626), POC=0.57 (0.42, 0.72); Category 2: AH, ER(+), 41 patients, median=5,790 (4,550–7,100), POC=0.81 (0.69, 0.90); Category 3: HUT, ER(-), 29 patients, median=1,500 (1,277–2,640), POC=0.28 (0.15, 0.45); Category 4: AH, ER(-), 26 patients, median=4,850 (3,160–6,160), POC=0.65 (0.50, 0.79); Category 5: Non-Proliferative, 24 patients, median=95 (45–299), POC=0.0 (0.0, 0.09). When the results of proliferative and non-proliferative cytology were analyzed with GS<1.67, 45% of the cases would be excluded from the therapy. Furthermore, 46% of cases with GS > 1.67 with non-proliferative cytology would be included in chemoprevention studies in absence of proliferative disease. What was important for us was to have accurate readings of breast density and recognize the presence of precancer (HUT or AH) and to arrest or reverse premalignant lesions. Conclusions: The presence of cytomorphology of HUT or AH obtained FNAB in high-risk patients for breast cancer would allow more accurate selection of candidates for chemoprevention than GS model and breast density, which can substantially under- or overestimate precancerous lesions. A series of patients with increased MD would receive unnecessary prophylactic therapy in the absence of precancerous lesions. Author Disclosure Employment or Leadership Consultant or Advisory Role Stock Ownership Honoraria Research Expert Testimony Other Remuneration Novartis

Breast cancer risk perception in a high-risk population

1542 Background: Women's perceptions of breast cancer risk and how they assess the value of prevention strategies are likely to determine their interest in preventive interventions and trials. The objectives of this study were to assess the effectiveness of an educational intervention on women's perception of breast cancer risk, understand their perceptions on the effectiveness of risk reduction interventions and assess which level of risk reduction would justify their participation in chemoprevention trials. Methods: An educational seminar on breast cancer risk was conducted. Questionnaires were administered before and after the educational intervention. These pre- and post-questionnaires provided data to calculate Gail risk scores and assess risk perception. The IRB of Nyack Hospital approved this study. Paired data analyses using Mantel-Haenszel odds ratios were conducted with statistical significance at p-value <0.05. Results: While 65 women completed both questionnaires, the total study population was restricted to the 38 participants who had no prior history of breast cancer and who provided information to compute a Gail score. Median age was 56 years (range: 28–72). 12 out of 38 (32%) women had a strong family history of breast cancer. The mean Gail score was 2.2. 17 out of 38 (45%) women had Gail scores ≥1.7%. 34% of the participants overestimated the average woman's lifetime risk of developing breast cancer. This perception decreased significantly to 8% after the educational intervention. Most women rated diet as the “most effective” risk reduction intervention, followed by physical activity. A majority of the participants (53%) said they would be willing to consider participating in chemoprevention trials if they promised a risk reduction of at least 35%. Conclusions: The results of this study demonstrate that educational efforts can reduce inaccuracies in women's perception of breast cancer risk. Participants expressed marked interest in participating in potential risk reduction interventions including vitamins, diet, and chemoprevention trials. Participants indicated that a potential risk reduction of 35% would be required for them to consider a chemoprevention trial. No significant financial relationships to disclose.

When the BRCA test is negative—Prevention strategies in women who have no detectable mutation (BRCA-) compared to women harboring positive BRCA mutations (BRCA+)

1534 Background: Much attention has been focused on screening and risk reduction, including prophylactic mastectomy (PM) or oophorectomy (PO) in BRCA+ mutation carriers (Rebbeck 1999, Eisen 2005, Kauff 2007). In families with a known BRCA mutation (BRCA+) or a history of hereditary breast cancer (BC) or ovarian cancer (OC), relatives in whom a mutation is not identified are considered ‘high risk’ and prevention strategies, including the need for PM or PO are undefined. We undertook a prospective evaluation of relatives from families wherein a BRCA mutation was present, with the intent to understand how BRCA- members view and manage their risks of BC or OC over time. Methods: Participants were 1st and 2nd degree unaffected relatives of 27 BRCA+ patients known to have BC or OC (20 BRCA1 and 7 BRCA2), referred to the high risk clinic for screening and prevention advice. Prior to and following genetic testing, participants were evaluated for individual risk factors for BC, self-perceived risk of BC, anxiety, prevention preferences and intentions. Genetic testing identified 62 BRCA+ (med. age 44) and 53 BRCA- (med. age 39) relatives, median Gail scores were 3.10 and 2.59 respectively. Results: Prior to genetic testing, 90% of all patients, regardless of BRCA status reported ‘extreme’ anxiety and a ‘very high’ perceived lifetime risk of BC. Two-thirds expressed interest in PM and/or PO. After genetic testing, 50% and 90% in the BRCA+ cohort, and 20% and 32% in the BRCA- cohort underwent PM and PO respectively. Less than 10% chose chemoprevention (CP) in both groups. At 6 months post PM or PO 80% and 65% of BRCA+ and BRCA- cohorts expressed decisional satisfaction. Significant factors determining PM in the BRCA- cohort were: age (>40), 1st degree relative with BC, family history of BRCA1 mutation, ‘high’ anxiety persisting >1 year. Conclusions: BRCA- relatives of BRCA+ families believe that they are highly susceptible to BC, experiencing high anxiety post-test disclosure and have an unexpectedly high rate of PM and PO. CP is unexpectedly low in both cohorts. Improved strategies for assessment and management of risk in these patients are required. No significant financial relationships to disclose.

Hip bone density predicts breast cancer risk independently of Gail score: Results From the Women's Health Initiative

Hip bone density predicts breast cancer risk independently of Gail score: Results From the Women's Health Initiative

Comparison of Ductal Lavage and Random Periareolar Fine Needle Aspiration as Tissue Acquisition Methods in Early Breast Cancer Prevention Trials

Short-term phase I and phase II breast cancer prevention trials require tissue acquisition at baseline and after intervention to evaluate modulation of potential biomarkers. Currently used tissue acquisition methods include ductal lavage (DL), random periareolar fine needle aspiration (RPFNA), and core needle biopsy. The optimum method to retrieve adequate samples and the most accepted method by study participants is not known. We compared RPFNA and DL as breast tissue acquisition methods for short-term breast cancer prevention trials by evaluating sample adequacy and tolerability in subjects who participated in two prospective phase II breast cancer prevention trials. Eighty-six women at increased risk for breast cancer were included in this study and underwent baseline DL and RPFNA. High risk was defined as having a 5-year Gail score of >1.67% or a history of atypical hyperplasia (AH), lobular carcinoma, or breast cancer. Median age was 54.5 years (range, 39-75 years); 75% of the women were postmenopausal. About 51% of the women yielded nipple aspiration fluid, and breast fluid samples via DL were retrieved in 73% of these subjects. Of these samples, 71% were adequate samples (greater than 10 epithelial cells). However, when the entire cohort was considered, only 31% of the subjects had adequate samples. RPFNA was also attempted in all subjects, and sample retrieval rate was 100%. Out of these, 96% of the subjects had adequate samples. In DL samples, AH rate was 3.7% was and hyperplasia (H) rate was 11.1%. In RPFNA samples, AH rate was 12.9%, and H rate was 24.7%. Cytology findings in RPFNA samples correlated with age, menopausal status, and breast cancer risk category (previous history of lobular carcinoma in situ). Both procedures were well tolerated, and no complications occurred among participants. Considering that the main end point for short-term prevention trials is the modulation of biomarkers, it is important to optimize adequate sample acquisition; therefore, RPFNA is a more practical option for future phase I and II breast cancer prevention trials compared with DL.

Participant characteristics on an international NCIC CTG breast cancer prevention trial

1531 Background: Antagonizing estrogen with tamoxifen has set the precedent for preventing breast cancer. Aromatase inhibitors reduce contralateral breast cancer in the adjuvant setting more than tamoxifen. The goal of MAP.3 is to determine whether exemestane reduces invasive breast cancers in post-menopausal women at risk for the disease. Methods: NCIC CTG MAP.3 is a randomized placebo controlled double blind trial of exemestane versus placebo. Recruitment began in September 2004 in North America and Spain. Criteria for entry on MAP.3 include: a Gail score of 1.66%; previous benign breast disease including LCIS; age over 60. The target sample size for the trial is 4560. Baseline characteristics of women enrolled thus far on MAP.3 were reviewed to determine whether their projected risk of breast cancer is consistent with the assumptions made in designing the trial. Results: Baseline characteristics are available on the 1784 women enrolled to MAP.3 to date. Over 70% of participants are over 60 years old (mean 64; SD=8.2 years) and the majority of the women are Caucasian (85%). The average 5-year risk of breast cancer is 2.7% (median, 2.2%). However, a greater proportion of women over 60 (30%) have a substantially lower Gail score (&lt;1.66) compared to women less than 60 years of age (8%). Few women in the MAP.3 cohort have been enrolled on a basis of a history of prior breast atypical hyperplasia or LCIS (8%), or DCIS treated with mastectomy (2%). Nearly half the MAP.3 participants (45%) have at least one first degree relative with malignant breast cancer, and 7.5% of these women have 2 or more first degree relatives with the disease. Conclusions: The risk profile of the women on MAP.3 will determine the event rate and therefore the time to trial unblinding and analysis. This initial description of the MAP.3 population enrolled thus far indicates a slightly lower baseline breast cancer risk in comparison to the risk profile of those enrolled in previous breast cancer prevention trials with tamoxifen or raloxifene (NSABP P1, IBIS-1 and STAR). Reasons for this difference are being explored. [Table: see text]