Degnim, AC

Abstract Background: Despite a paucity of data on the benign breast transcriptome, reduction mammoplasty tissue is commonly used as a control for RNA biomarker discovery in breast cancer (BC). We evaluated the transcriptome of benign breast to characterize pre-menopausal and post-menopausal coding and non-coding RNA profiles. These profiles will provide insight into the metabolic landscape of benign breast and inform transcriptomic studies of BC and peri-tumoral microenvironments. Methods: With institutional approval, cryobanked fresh breast tissues from reduction mammoplasties (age 15-62y; follow-up: 4-10 y with no cancer events), with histologically-confirmed benign tissues with similar epithelial: stromal tissues were grouped into 4 age sets (< 30y, N=10; 30-39y, N=11; 40-49y, N=10; ≥50y (clinically post-menopausal), N=10). Following RNA extraction, library preparation (Illumina TruSeq Stranded mRNA kit), and sequencing (Illumina HiSeq 4000), reads were processed (MAP-RSeq v3.0.0) and aligned (STAR aligner; hg38). Differential expression (DE) analysis (edgeR 2.6.2) identified DE genes from normalized RPKM reads (absolute log2 fold change (FC) > 1 and false discovery rate (FDR) < 0.10), corrected for intra-group biases using medians (absolute FC cut-off of >1.5). Over-representation analysis [Ingenuity pathway analysis (IPA), Ingenuity® Systems] and gene set enrichment analysis [(GSEA), GeneTrail 2.0] identified significantly-enriched pathways. Results: Across the 4 age sets, 561 DE genes were identified. Compared to the post-menopausal (PM) set, the number of DE genes was highest in <30 y set (N= 372) and decreased with increasing age (N= 170, age 30-39 set and N=20, age 40-49 set), generating up-regulated (PMup) and down-regulated (PMdown) PM transcriptomic profiles. The top PMdown DE genes included RANKL, WNT4, MKI67, extracellular matrix, and lactation-related genes (lactoferrin ,MUC4, MUC16, p < 0.01). Top PMdown canonical pathways were cell cycle-related (CDK1, CCNA2, CCNB2, ESPL1, TOP2A)(p< 0.001). Top PMup genes included those involved in adipogenesis, NPY1R, NPY2R, unsaturated fatty acid synthesis and eicosanoid signaling (P< 0.001). Top PMup canonical pathways included acyl co-A hydrolysis, stearate biosynthesis, acute phase response and RXR signaling (P < 0.001). GSEA of the entire gene set (N=15,466; ranked in order of 561 DE genes) identified 16 significant pathways, functionally grouped as mitochondrial energy metabolism, proteasome, and unsaturated fatty acid biosynthesis and signaling (PPAR pathway). While coding RNAs comprised 85% of DE transcripts, lncRNA comprised ˜5%. Top DE lncRNAs and precursor miRNAs included MEG3, a tumor suppressor, LIN00092, miR22 and miR1182 (p < 0.0001). Conclusions: Pre and post-menopausal benign breast tissues have distinctive transcriptomic profiles. PM benign breast has down-regulated proliferation/cell-cycle-related pathways, and up-regulated genes in mitochondrial energy pathways, lipogenesis and inflammatory pathways. Notably, many of the top DE coding and non-coding RNAs in PM benign breast have been implicated in BC progression, highlighting their value in better understanding breast carcinogenesis and the need to characterize their functional roles in normal aging and menopause. Citation Format: Carter JM, Nair AA, Davila JI, Heinzen EP, Hoskin TL, Winham SJ, Radisky DC, Degnim AC, Visscher DW. Distinctive coding and non-coding RNA profiles of pre-menopausal and post-menopausal benign breast [abstract]. In: Proceedings of the 2018 San Antonio Breast Cancer Symposium; 2018 Dec 4-8; San Antonio, TX. Philadelphia (PA): AACR; Cancer Res 2019;79(4 Suppl):Abstract nr P5-06-01.

Abstract Background: Differences in ER+ and ER- breast cancer tumor biology are well-documented, but little is known about the field of background benign breast in which those cancers arise. We evaluated the transcriptome of benign breast tissues from women with concurrent ipsilateral ER+ and ER- breast cancers (BC) to characterize coding and non-coding RNA profiles. This pilot study provides insight into the transcriptomic landscapes of benign breast tissues in patients with BC, and the microenvironment of the at-risk benign tissue. Methods: With institutional approval, cryobanked breast tissues from patients with concurrent ipsilateral ER+ BC (benign ER+BC, N=14) or ER- BC (benign ER-BC, N=10), were selected for benign tissues with similar epithelial:stromal ratios and grouped into pre (preM) and post-menopausal (PM) groups. Following RNA sequencing (Illumina TruSeq Stranded mRNA kit & Illumina HiSeq 4000), reads were processed (MAP-RSeq v3.0.0) and aligned (STAR aligner; hg38). Differential expression (DE) analysis (edgeR 2.6.2) identified DE genes from normalized RPKM reads (absolute log2 fold change (FC) > 1 and false discovery rate (FDR) < 0.10), corrected for intra-group biases. Over-representation analysis [Ingenuity pathway analysis (IPA), Ingenuity® Systems] and gene set enrichment analysis [(GSEA), GeneTrail 2.0] identified significantly-enriched pathways. Results: In the PM group, there were 144 DE transcripts between benign ER+BC and benign ER-BC, including coding RNAs (40%), antisense RNAs (35%) and lncRNAs (7%). In contrast, the preM group had no significantly DE genes between benign ER+BC and benign ER-BC. In the PM DE coding gene set, the top DE transcripts in benign ER+BC included many genes implicated in BC development# or ER+ BC progression (*) (e.g. up-regulated: KAAG1*, DNAJB7/HSP40*, TMEM151B, ZBTB32# p < 0.001 ; down-regulated: CPB1*, FOS, TPPP3#, CLEC3B#p < 0.001). Top canonical pathways altered in benign ER+BC included MAPK, PI3K, and acute phase response pathways (p<0.05). GSEA of the entire gene set (N=15,223; ranked in order of 144 DE genes) identified 72 altered pathways (P < 0.005); those with the highest normalization enrichment scores (NES) (> 0.4) functionally grouped as immune function-related (T cell function and antigen presentation). Depleted pathways with NES > 0.4, (N=6) functionally grouped into proteasome-related, fatty acid biosynthesis and mitochondrial energy metabolism. Among the non-coding DE gene set, notably, the entire DE antisense RNA gene set (N=51 transcripts) was up-regulated in benign ER+ BC compared to benign ER-BC (P< 0.001) with a subset (N=11) showing marked up-regulation (> 4 log2FC). Among the DE antisense RNAs, 70% have reported roles in carcinogenesis or BC progression (e.g. KRT7-AS, NAV2-AS2, CCDC144NL-AS1, RP11-66B24.4, HSA-MIR4454). Conclusion: The benign breast transcriptome differs between postmenopausal women with ER+ vs ER- BC, with distinctive coding and non-coding RNA signatures. In postmenopausal women with ER+ BC, benign breast expresses a unique antisense RNA set and is enriched in genes implicated in BC development or progression. These data provide insight into at-risk benign breast and facilitate identification of potential biomarkers of carcinogenesis. Citation Format: Carter JM, Nair AA, Davila JI, Heinzen EP, Hoskin TL, Winham SJ, Radisky DC, Visscher DW, Degnim AC. A unique coding and non-coding benign breast transcriptome in post-menopausal ER+ breast cancer [abstract]. In: Proceedings of the 2018 San Antonio Breast Cancer Symposium; 2018 Dec 4-8; San Antonio, TX. Philadelphia (PA): AACR; Cancer Res 2019;79(4 Suppl):Abstract nr P3-08-10.

Abstract The objective of this lecture is to review clinical management of women with the following histologic lesions associated with increased breast cancer risk: flat epithelial atypia (FEA), atypical ductal hyperplasia (ADH), atypical lobular hyperplasia (ALH), and lobular carcinoma in situ (LCIS). Women found to have one or more of these histologic findings in a benign breast biopsy have an increased risk of developing a future breast cancer. Published data will be reviewed that provides estimates of annual and longterm risk associated with these lesions. In addition, management of these lesions after core biopsy will be discussed, specifically the circumstances when surgical excision is recommended. Medical and surgical options for breast cancer risk reduction will be reviewed, as well as their appropriate clinical indications. Citation Format: Degnim AC. Clinical management of women with increased breast cancer risk based on histologic lesions [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 ES2-2.

Abstract Background Women with benign breast disease (BBD) have an increased risk for breast cancer (BC). Almost 30% of all BCs develop in women with prior BBD. Information about the features of the expected BCs after BBD would enable individualized surveillance and prevention strategies for these women. We sought to characterize BCs developing in a large cohort of women with BBD. Methods Our cohort includes 13,485 women who underwent breast biopsy for mammographic or palpable concern between 1967-2001. Biopsy slides were reviewed and classified as nonproliferative (NP), proliferative disease without atypia (PDWA), or atypical hyperplasia (AH). BCs were identified by follow-up questionnaires, medical records, and Tumor Registry data. BC tissues were obtained and reviewed. Results With a median follow-up of 15.8 years, 1273 women developed BC. Most BCs were invasive (81%), among which 61% were ductal, 13% mixed ductal/lobular, and 14% lobular. Two thirds were intermediate or high-grade, and 29% were node positive. Cancer characteristics were similar across the three histologic categories of BBD, with similar frequency of DCIS and invasive disease, tumor size, time to invasive BC, histologic type of BC, nodal positivity and HER2 positivity. Women with AH had a higher frequency of ER+ BC (91%) vs women with PD (80%) or NP (85%), p=0.02. Conclusion A substantial proportion of all BCs develop in women with prior BBD. We show that the majority of BCs after BBD are invasive tumors of ductal type, with a substantial proportion node positive. 84% of cancers were ER-positive. Chemoprevention should be strongly encouraged in higher-risk women with BBD. Citation Format: Visscher DW, Frost MH, Hartmanan LC, Frank RD, Vierkant RA, McCullough AE, Winham SJ, Vachon C, Ghosh K, Brandt KR, Farrell AM, Tarabishy Y, Hieken TJ, Haddad TC, Kraft RA, Radisky DC, Degnim AC. Clinicopathologic features of breast cancers that develop in women with previous benign breast disease. [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 P6-10-19.

Abstract Background Women with atypical hyperplasia (AH) are at an approximately four-fold increased risk of subsequent breast cancer (BC). Mammographic breast density (MBD) is a well-established risk factor for BC, but its contribution to BC risk in women with AH remains an open question. We previously reported no association between MBD [measured by Wolfe's parenchymal pattern (PP)] and BC risk in a cohort of 147 women with AH. Here, we present results in an expanded cohort of 459 women diagnosed with AH between 1985 and 2001. Methods The Mayo Clinic Benign Breast Disease Cohort includes 13,485 women who had benign core and/or excisional biopsy 1967-2001. Biopsy tissues were reviewed by our study pathologist to determine presence of AH. MBD was available from clinical records starting in 1985, coded as PP (the standard for 1985-1996) or BI-RADS (1997-2001) density criteria. The original four-level PP (N1-fatty, P1-ductal prominence <25% of breast, P2-ductal prominence >25%, DY-dysplasia) and BI-RADS (fatty, scattered densities, heterogeneously dense, extremely dense) measures were re-categorized as low, moderate or high MBD by combining the middle two categories for each. BC events and clinical information were obtained by questionnaires, medical records and the Mayo Clinic Tumor Registry. Women were followed from benign biopsy to date of BC, death or last contact. Standardized incidence ratios (SIRs) were generated overall and within subgroups defined by density measure (PP vs. BI-RADS), number of atypical foci, and BMI by dividing the observed number of BCs by population-based expected values. Cox regression was used to estimate MBD hazard ratios after adjustment for demographic and clinical variables. Results Of the 551 women diagnosed with AH between 1985 and 2001, 459 (83%) had MBD data within 1 year prior to biopsy. Of these, 68 (15%) had low, 221 (48%) had moderate, and 170 (37%) had high MBD, respectively. Over a median follow-up of 11.7 years, 80 BCs were observed. SIRs for breast cancer did not differ significantly across density categories, overall or within any subgroups examined (see Table). Cox regression adjusting for age, BMI and density measure (PP vs. BI-RADS) also failed to identify an association with MBD (p=0.55). Low MBDModerate MBDHigh MBD N / BCsSIR (95% CI)N / BCsSIR (95% CI)N / BCsSIR (95% CI)P-valueOverall68/123.5 (1.8,6.1)221/393.6 (2.5,4.9)170/293.4 (2.3,4.8)0.97MBD Measure PP59/113.6 (1.8,6.5)85/153.0 (1.7,5.0)130/243.3 (2.2,5.0)0.87BI-RADS9/12.7 (0.1,14.7)136/244.0 (2.6,6.0)40/53.4 (1.1,7.9)0.90No. Atypical Foci 147/62.3 (0.9,5.1)123/182.8 (1.7,4.4)96/163.5 (2.0,5.8)0.63214/47.5 (2.0,19.1)58/134.8 (2.6,8.3)41/62.5 (0.9,5.5)0.213+7/26.7 (0.8,24.0)40/84.4 (1.9,8.6)33/74.0 (1.0,8.2)0.83BMI <2525/21.5 (0.2,5.4)75/184.9 (2.9,7.7)101/173.4 (2.0,5.5)0.1625-2919/54.9 (1.6,11.3)68/92.8 (1.3,5.2)36/42.1 (0.6,5.4)0.4530+23/54.8 (1.6,11.2)76/123.0 (1.6,5.3)32/74.0 (1.6,8.2)0.67SIRs compare observed numer of BCs to expected using Iowa SEER data. Analyses account for the effects of age and calendar period. P-value is test of heterogeneity in SIRs across columns. Conclusions We found no evidence of an association between MBD and subsequent BC in women with AH. Citation Format: Vierkant RA, Degnim AC, Hartmann LC, Frank RD, Radisky DC, Visscher DW, Frost MH, Winham SJ, Ghosh K, Vachon CM. No evidence of association between mammographic breast density and risk of breast cancer in women with atypical hyperplasia. [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 P6-09-05.

Abstract Introduction: CD4+ and CD8+ T cells are important effector cells in the adaptive immune response against cancers. In premalignant proliferative breast tissues, the role of the immune system remains relatively undefined. In this study, we sought to investigate and quantify the presence of CD4 and CD8 expressing immune cells in benign breast tissues and their association with breast cancer risk. Method: Archived breast tissue samples from women with benign breast disease (BBD) or no clinical breast disease [Komen Tissue Bank (KTB)] were obtained for this age-matched case-control study. BBD samples included BBD cases (subsequently developed breast cancer) and BBD controls). Tissue sections were characterized for non-proliferative or proliferative disease and degree of lobular involution. CD4+ and CD8+ cell densities (cells/mm2) were obtained by digital image quantitation for up to 10 individual lobules per sample and a per-sample estimate of cell density was calculated as the median cell density across lobules within a sample. Statistical analysis was performed using Wilcoxon signed-rank and Kruskal-Wallis tests. Results: Among 267 women (median age 52, range 35-75) comprising 89 age-matched triplets, 2417 lobules were evaluated [783 KTB, 804 BBD case, 830 BBD control]. 1372 (57%) lobules were normal and 1045 (43%) were fibrocystic. CD4+ cells were present in 1903 (79%) lobules while CD8+ cells were present in 2214 (92%) lobules. CD4 and CD8 densities showed a moderate positive correlation (r = 0.39) with each other, but overall, lobules showed a lower density of CD4+ than CD8+ cells (mean difference -54 cells/mm2, 95% CI: -68 to -39 cells/mm2). At the per sample level (see Table), BBD controls showed significantly higher levels of both CD4+ (p=0.009) and CD8+ cells (p=0.0001) compared to KTB samples. In contrast, the BBD cases only showed elevated CD8 (p=0.007) and not CD4 (p=0.21) cell infiltrates compared to KTB samples. Comparing BBD cases to BBD controls, BBD cases showed significantly lower CD8 cell density as compared to BBD controls (p=0.002) and cases had lower CD4 cell density but the difference was not significant (p=0.07). Despite these differences in the cell densities, the ratio of CD4:CD8+ cells did not differ significantly across groups (p=0.83). The CD4:CD8+ ratio also did not differ significantly by histologic impression (p=0.74), degree of involution (p=0.27), or age (p=0.32). KTBBBD ControlsBBD CasesMedian (IQR)(n=89)(n=89)(n=89)CD4+ cells/mm240 (17-95)60 (25-124)49 (21-92)CD8+ cells/mm285 (49-131)157 (84-225)134 (80-168)Ratio (CD4:CD8)*0.45 (0.14-0.96)0.41 (0.15-0.81)0.40 (0.19-0.70)*Ratio is missing for n=7 samples (3 KTB, 2 BBD control, 2 BBD case) with undefined ratio due to zero CD8+ cells for all lobules in the sample. Conclusion: Our findings suggest that BBD is associated with increased infiltration of T cells into the breast tissue. Furthermore, these immune responses appear more robust in BBD patients at lower risk of developing breast cancer. Citation Format: Pena Jimenez A, Hoskin TL, Arshad MA, Visscher DW, Brahmbhatt RD, Miller EE, Stallings Mann ML, Carter JM, Murphy LM, Winham SJ, Radisky DC, Denison LA, Knutson KA, Degnim AC. CD4 and CD8 T cell densities are increased in benign breast disease compared to normal breast tissue. [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 P4-04-09.

Abstract Background: Recent studies have indicated that vaccination can protect against cancer development. One key aspect of developing vaccines is circumventing peripheral tolerance by identifying subdominant epitopes that are unique to the deregulated tumor microenvironment. While existing subdominant epitope vaccines are showing efficacy in preventing cancer, these vaccines are applicable only for subsets of patients with specific HLA subtypes. Therefore, we recently developed a degenerate HER2 subdominant epitope-based vaccine that should be useful in approximately 85% of all patients. The vaccine consists of a pool of four HLA-DR-restricted 15-amino acid epitopes (p59, p88, p422, and p885) that are naturally processed and are designed to elicit helper T cell immunity, the cornerstone of immune surveillance. Here we present Phase I trial results of this multi-peptide HER2 vaccine. Methods: Eligible women had HER2+ breast cancer (Stages II-III) and had completed standard treatment (i.e. surgery, chemotherapy, and trastuzumab) at least 90 days prior to enrollment and were rendered disease free. Vaccine included the above epitope pool along with adjuvant GM-CSF. Patients were vaccinated six times over six months and were monitored for toxicity at each visit. Peripheral blood samples were collected for immune responses evaluating for T cell and antibody immunity. Endpoints were safety and immunogenicity leading to the development CD4 helper T cells that recognized naturally-processed HER2. Results: Twenty-two subjects (age 33 to 69 years) were enrolled. At the present analysis, 21 have completed all 6 vaccination cycles; one patient withdrew after developing a grade 1 injection site reaction during the first vaccination cycle. Twenty patients have had LVEF measured after vaccination; only 2 patients had an LVEF drop of 10% or more but remained in the normal LEVF range. One severe toxicity was reported: a grade 3 INR increase considered unrelated to treatment. Mild to moderate (grade 1-2) toxicities included injection site reactions, fatigue, and white blood cell count decreases. All patients were alive at analysis and only one experienced a recurrence (median follow-up 507 days, range 22 – 844). Twenty patients have had immune response assessments. Vaccine induced T cell immunity was observed in 94% of patients to p59, in 94% of patients to p88, in 82% of patients to p422, and in 74% of patients to p885. Importantly, T cell immunity to naturally processed HER2 proteins occurred in 94% of patients. The mean number of T cells specific for each peptide, post vaccination, ranged from 349–528 T cells per million peripheral blood mononuclear cells (PBMCs). The mean number of T cells specific for whole HER2 protein was 783 T cells per million PBMCs compared to a mean of 898 T cells/million PBMCs specific for the foreign tetanus toxin. In contrast to T cell responses, modestly increased antibody immunity to HER2 occurred in 35% of patients, consistent with the T cell-inducing design of the vaccine. Conclusion: Our results show that it is possible to develop vaccines with broad HLA coverage that circumvent natural tolerance and induce tumor antigen-specific immunity in the vast majority of patients. Citation Format: Knutson KL, Kalli KR, Block MS, Hobday TJ, Padley DJ, Erskine CL, Dockter T, Suman VJ, Wilson G, Degnim AC. Robust generation of T cell immunity to HER2 in HER2+ breast cancer patients with a degenerate subdominant HLA-DR epitope vaccine. [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 P2-11-02.

Abstract Background: Age-related lobular involution (LI) is a physiological process, usually accelerated with menopause, in which the epithelial tissue of the breast gradually regresses. Our previous analysis of a cohort of more than 14,000 women who had a benign breast biopsy at the Mayo Clinic revealed that women who had undergone LI were at significantly reduced risk of subsequent breast cancer development. In our studies of postmenopausal women specifically, we found that more than 40% had not completed the process of LI, and that the increased breast cancer incidence in older women was strongly associated with incomplete LI. We sought to determine the potential protective mechanistic effects of lobular involution. Methods: We generated primary human mammary epithelial cell lines (HMECs) from breast biopsies of women who showed no LI (N = 6), partial LI (N = 7), or complete LI (N = 6). These cell lines were analyzed for proliferation in culture and profiled for genomic signatures using Affymetrix gene arrays. We also analyzed RNA isolated from a cohort of archived formalin-fixed, paraffin-embedded (FFPE) benign breast biopsy samples from postmenopausal women who had either completed the process of LI (N = 4) or who had not initiated LI (N = 8) for genomic signatures using Whole-Genome DASL Assays. Results: We found that LI was associated with decreased cellular proliferation and inhibition of processes associated with tumor development and progression. Specifically, we found that HMECs derived from benign breast biopsies in which LI was not evident showed increased proliferation and elevated expression of Twist (p = 0.002), a transcription factor associated with epithelial-mesenchymal transition (EMT) and breast cancer development. We are assessing differentially expressed genes for association with breast cancer risk using a nested case:control group of benign breast biopsies from women in our cohort who subsequently developed breast cancer (N = 85) vs those who did not (N = 142) and for whom LI status at the time of initial biopsy was known. Conclusions: LI may decrease cancer risk through inhibition of key tumorigenic processes. Identification of processes associated with the physiologic risk reduction seen with LI may provide insight into novel avenues for risk reduction. Citation Information: Cancer Res 2012;72(24 Suppl):Abstract nr P6-06-01.

Abstract Background: Fibroadenoma (FA) is a relatively common benign breast tumor that can occur in women of any age, with a peak incidence during the second and third decades. The only previous study of this lesion reported that FAs are associated with a 2.2 times increased risk of developing invasive breast cancer (BC) compared to matched controls [1]. This relative risk may increase to 3.1 among patients with complex FA, and remains elevated for decades after diagnosis. However, this study did not thoroughly account for other forms of concomitant risk factors. Our investigative team sought to examine breast cancer risk among women with non-complex and complex FA, overall and stratified by other BC risk factors. Materials and Methods: The study cohort included women between ages 18 to 85 in the Mayo Benign Breast Disease (BBD) Cohort who underwent excisional breast biopsy between 1967and1991 and were found to have a FA. FA was defined histologically as a combination of epithelial and stromal proliferation. Complex FA was defined as FA associated with any of the following features: sclerosing adenosis, epithelial calcifications, papillary apocrine change, and microcysts greater than 3.0 mm. The primary endpoint was a diagnosis of BC, determined using the Mayo medical record and questionnaire information from study participants. A single breast pathologist, blinded to the initial diagnosis and clinical outcome, performed pathology review. Observed vs. expected BC risk across levels of FA was assessed via standardized incidence ratios (SIRs), using age-stratified incidence rates from the Iowa Surveillance, Epidemiology, and End Results (SEER) registry. Analyses were carried out overall and within subgroups of involution status (none, partial, complete) and overall histology (non-proliferative disease [NP], proliferative disease without atypia [PDWA] or atypical hyperplasia [AH]). Results: Of 9097 women in the Mayo BBD Cohort, FA were identified in 2139- non-complex in 1903 (20.9%) and complex in 236 (2.6%). The greatest proportion of FA occurred in the 40–69 age range. The mean ages for women with non-complex FA and complex FA were 45.7 and 50.2 years respectively. The SIR of breast cancer in the overall BBD cohort was 1.5 (95% CI [1.4–1.6]). The SIR among women with non-complex FA was 1.5 (95% CI [1.3–1.8]), and for complex FA increased to 2.41 (95% CI [1.7–3.4]). However, women with complex FA were more likely to have other concomitant high-risk histologic features such as PDWA and incomplete involution. In stratified analyses accounting for involution status and PDWA, complex FA did not demonstrate an independent increase in BC risk. Conclusion: Complex FA does not confer an increased risk for BC beyond other established histologic features. Therefore, women with complex FA should be managed based upon a risk level consistent with the major histologic category of PDWA and/or AH. Citation Information: Cancer Res 2012;72(24 Suppl):Abstract nr P5-01-08.

Abstract Background: Optimal early detection and prevention strategies for breast cancer are predicated on our ability to accurately identify women at increased risk for disease. Current breast cancer risk prediction models rely on clinical and epidemiologic features. As demonstrated with other cancers, such as those of the cervix and colon, cancer risk assessment is enhanced when the tissue at risk is examined for evidence of premalignant abnormalities. Here, we present a multivariate model for individualized breast cancer risk prediction that incorporates tissue features identified from biopsies from women with benign breast disease (BBD). Methods: We used our cohort of more than 9000 women diagnosed with BBD at the Mayo Clinic from 1967–1991, and identified a nested case:control series of 377 women who developed cancer (cases) matched with 734 women who did not (controls). We identified a set of variables that either alone or in two-way interactions contributed to a multivariable risk prediction model. We combined this relative risk model with the baseline incidence of breast cancer and the baseline hazard of death from our full BBD cohort to develop a risk prediction tool that computes absolute risk estimates for breast cancer. We calculated 5-year risk predictions for these cases and controls using our model and using the Gail model and compared c-statistics from these risk predictions. We validated our model using an independent set of 378 breast cancer cases and 728 matched controls drawn from our BBD cohort. Results: Cases and controls in the model development data set differed significantly on a number of features when examined individually. Those included in a multivariable relative risk model for breast cancer among women with BBD are shown below. After including interactions between lobular involution, and histologic impression, with age at first live birth/number of children, and degree of involution with sclerosing adenosis, we devised a model that predicts the absolute risk of breast cancer. Predicted five-year risks in the model development series, we observed a c-statistic of 0.66, higher than the c-statistic of 0.61 for the Gail model. With risk prediction estimates from our validation series of cases and matched controls, the c-statistic of our model decreased to 0.59 and the c-statistic for the Gail model was 0.55. Conclusion: We have identified an initial risk prediction model for breast cancer in women with BBD. The identified model includes basic demographic information as well as a number of histologic features of the tissue samples. Its initial performance is promising, providing improved discrimination over the Gail model in our replication data set. Further work is needed to identify features which might provide additional predictive information for breast cancer risk. Citation Information: Cancer Res 2012;72(24 Suppl):Abstract nr P4-12-03.