The goal of DCIS treatment is to avoid a future invasive breast cancer. Invasive breast cancer risk is heterogeneous across the population of DCIS patients. More than 40% of women with minimal, low grade DCIS will develop invasive breast cancer if left untreated and followed for more than 10 years. Risk of invasive recurrence is about 2% per year for DCIS treated by simple excision (lower for small low grade lesions and higher for larger high grade lesions). A reliable approach for risk-stratifying individual DCIS lesions could permit tailored approaches to therapy or avoidance of therapy all together in the lowest risk women. Screen-detected, grade 1 and 2 micropapillary DCIS in older women is likely the lowest risk lesion. Progression to invasive breast cancer is more likely related to focal myoepithelial cell dysfunction than molecular alterations in the bulk luminal cell populations. Nevertheless, p16, ER, COX2, HER-2/neu, and Ki67 staining patterns in the bulk luminal cell population have been linked to recurrence risk. A 12 gene recurrence score has been shown to predict the risk of recurrence for DCIS treated by excision without radiation. This assay was adapted from the 21-gene OncoTypeDx test and is weighted heavily towards proliferation markers. Accurate prediction of invasive recurrence risk will likely require detection of focal alterations in myopeithelial cells and the basement membrane. Citation Format: Euhus DM. Prognostic indicators in treatment of DCIS. [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 ES8-2.

Purpose: To evaluate the tolerability of a dose-escalated 5-fraction stereotactic body radiation therapy for partial-breast irradiation (S-PBI) in treating early-stage breast cancer after partial mastectomy; the primary objective was to escalate dose utilizing a robotic stereotactic radiation system treating the lumpectomy cavity without exceeding the maximum tolerated dose.Methods and Materials: Eligible patients included those with ductal carcinoma in situ or invasive nonlobular epithelial histologies and stage 0, I, or II, with tumor size <3 cm. Patients and physicians completed baseline and subsequent cosmesis outcome questionnaires. Starting dose was 30 Gy in 5 fractions and was escalated by 2.5 Gy total for each cohort to 40 Gy.Results: In all, 75 patients were enrolled, with a median age of 62 years. Median follow-up for 5 cohorts was 49.9, 42.5, 25.7, 20.3, and 13.5 months, respectively. Only 3 grade 3 toxicities were experienced. There was 1 dose-limiting toxicity in the overall cohort. Ten patients experienced palpable fat necrosis (4 of which were symptomatic). Physicians scored cosmesis as excellent or good in 95.9%, 100%, 96.7%, and 100% at baseline and 6, 12, and 24 months after S-PBI, whereas patients scored the same periods as 86.5%, 97.1%, 95.1%, and 95.3%, respectively. The disagreement rates between MDs and patients during those periods were 9.4%, 2.9%, 1.6%, and 4.7%, respectively. There have been no recurrences or distant metastases.Conclusion: Dose was escalated to the target dose of 40 Gy in 5 fractions, with the occurrence of only 1 dose-limiting toxicity. Patients felt cosmetic results improved within the first year after surgery and stereotactic body radiation therapy. Our results show minimal toxicity with excellent cosmesis; however, further follow-up is warranted in future studies. This study is the first to show the safety, tolerability, feasibility, and cosmesis results of a 5-fraction dose-escalated S-PBI treatment for earlystage breast cancer in the adjuvant setting. Published by Elsevier Inc.

BackgroundWomen with BRCA1 and BRCA2 mutations have substantially elevated risk of developing breast cancer. The aim of this study was to clarify the role of bilateral risk-reducing mastectomy (BRRM) in reducing breast cancer risk in women carriers of BRCA1 and BRCA2 mutations.

Purpose.-Patients with Cowden syndrome (CS) with underlying germline PTEN mutations are at increased risk of breast, thyroid, endometrial, and renal cancers. To our knowledge, risk of subsequent cancers in these patients has not been previously explored or quantified.Patients and Methods.-We conducted a 7-year multicenter prospective study (2005 to 2012) of patients with CS or CS-like disease, all of whom underwent comprehensive PTEN mutational analysis. Second malignant neoplasms (SMNs) were ascertained by medical records and confirmed by pathology reports. Standardized incidence ratios (SIRs) for all SMNs combined and for breast, thyroid, endometrial, and renal cancers were calculated.Results.-Of the 2,912 adult patients included in our analysis, 2,024 had an invasive cancer history. Germline pathogenic PTEN mutations (PTEN mutation positive) were identified in 114 patients (5.6%). Of these 114 patients, 46 (40%) had an SMN. Median age of SMN diagnosis was 50 years (range, 21 to 71 years). Median interval between primary cancer and SMN was 5 years (range, < 1 to 35 years). Of the 51 PTEN mutation-positive patients who presented with primary breast cancer, 11 (22%) had a subsequent new primary breast cancer and 10-year second breast cancer cumulative risk of 29% (95% CI, 15.3 to 43.7). Risk of SMNs compared with that of the general population was significantly elevated for all cancers (SIR, 7.74; 95% CI, 5.84 to 10.07), specifically for breast (SIR, 8.92; 95% CI, 5.85 to 13.07), thyroid (SIR, 5.83; 95% CI, 3.01 to 10.18), and endometrial SMNs (SIR, 14.08.07; 95% CI, 7.10 to 27.21).Conclusion.-Patients with CS with germline PTEN mutations are at higher risk for SMNs compared with the general population. Prophylactic mastectomy should be considered on an individual basis given the significant risk of subsequent breast cancer.

Abstract Background: Accurate assessment of HER2 status is critical for selecting patients who will benefit from trastuzumab therapy. There is still no consensus regarding the optimal method to assess HER2 status. Computerized image analysis has been shown to provide a more accurate and objective way for quantification of HER2 expression by IHC than manual evaluation. It has been suggested that the use of image analysis may help to resolve some of the discrepancies between IHC and FISH assay. We compared the results of HER2 expression by IHC using automated image analysis with fluorescent in situ hybridization (FISH) assay. Design: Testing for HER2 expression by IHC and FISH was performed on 2853 specimens at UT Southwestern Medical Center between the years 2002 to 2011. Quantification of IHC HER2 expression was done by image analysis and scored as positive (>2.0), borderline (1.5 to 2.0) and negative. (<1.5). The PathVysion kit was used for FISH assay to evaluate HER2 amplification. Ratios >2.0, 1.8 to 2.0 and <1.8 were defined positive, borderline and negative amplification respectively. Results: IHC compared to FISH Conclusion: Despite improvements in IHC testing, the FISH assay may be a better method for determining HER2 status. Factors such as tissue fixation, scoring methods and choice of antibodies may contribute to the lower specificity of IHC. In the amplified group, the gene amplification ratio correlated with protein expression, being highest in the IHC positive cases and lowest in those that were negative. Citation Information: Cancer Res 2012;72(24 Suppl):Abstract nr P1-07-04.

Abstract Background: DNA ploidy has been shown to have prognostic significance in patients with breast cancer. Studies in the past have mainly utilized flow cytometry (FCM) for measuring DNA ploidy. However FCM has several disadvantages, the instrument cannot distinguish benign from malignant cells and it cannot be performed on small tumor samples. The relationship between DNA ploidy and biomarker expression in breast cancer has not been well studied. Recently, gene expression analysis has demonstrated distinct subtypes of breast cancer. Expression of ER, PR and Her2 by IHC has been used as a surrogate tool for the molecular classification of breast cancer. Aim: To determine the relationship between DNA ploidy, biomarkers (ER, PR, HER2, Ki67 and p53) expression and molecular subtypes of invasive breast cancer (IBCA) using image analysis. Design: DNA analysis was performed on Feulgen stained sections from the same tumor block used for biomarker analysis. DNA indices and ploidy were analyzed using the Autocyte Pathology Workstation (Tripath, Burlington, NC). Briefly, a total of 200–300 nuclei were collected and mean DNA index reported. DNA index was obtained by measuring the optical density of tumor cells in comparison with those of the non-neoplastic stromal cells in the sample using the latter as the diploid reference (value of 1.0). Tumors were classified as diploid (DNA indices of 0.90 to 1.1), aneuploid (DNA indices of <0.89 and > 1.1) and multiploid tumors; multiple indices. Patient's age, tumor size, histologic grade, stage, biomarker status and breast cancer subtypes were correlated with ploidy status. Results: Of the 248 cases of IBCA, 176 had aneuploid DNA and 72 diploid. Aneuploid tumors were predominantly grade 3, 72.1% versus 27.9% of diploid (p<0.0001). Tumor ploidy had no significant association with age (<40 years) versus older age (>50 years), p=0.118. The mean tumor size in aneuploid and diploid tumors was 4.2 cm and 3.2 cm respectively (p=0.009). Aneuploid tumors were frequently ER and PR negative compared to diploid (p<0.0001) with a mean ER level of 31.7% compared to 72.2% in diploid tumors (p<0.0001). The mean PR level was 17.2% in aneuploid versus 34.0% in diploid tumors (p=0.0009). The ki67 index was 52.4% in aneuploid compared to 29.1% in diploid (p<0.0001). The p53 expression was 30.3% in aneuploid versus 14.4% in diploid tumors. Aneuploid tumors were frequently of advanced T stage compared to diploid tumors (p=0.0048). There was no significant association with N stage (p=0.734). By multivariate analysis after adjusting for age, grade, T and N stage, ER (p=0.0021), PR (p=0.0003), HER2 over-expression (p= 0.0028), Ki67 (p=0.0383), T stage (p= 0.0172 and grade (<0.0001) were significantly associated with aneuploidy. Of breast cancer subtypes, Her2 (p<0.0001), triple negative (p<0.0001), luminal B subtype (p=0.006) were frequently aneuploid and luminal A was frequently diploid (p=0.051). Conclusion: DNA ploidy measured by image analysis has significant association with biomarker expression. Increased DNA content is associated with poor prognostic features and aggressive subtypes of breast cancer. Citation Information: Cancer Res 2011;71(24 Suppl):Abstract nr P4-09-30.

1529 Background: Increased body mass index (BMI) has traditionally been associated with an increased risk for hormone receptor positive breast cancer in postmenopausal women, but recent studies have also linked it to aggressive phenotype breast cancers in premenopausal women. A better understanding of the relationship between BMI and specific intrinsic subtypes of breast cancer can contribute directly to the design and development of cancer prevention and treatment trials that address energy balance and obesity.PURPOSETo evaluate the relationship between BMI and the intrinsic subtypes of breast cancer.METHODSFrom 740 patients with primary invasive breast cancer diagnosed during 2008 and 2009, BMI at time of diagnosis was ascertainable for 693 patients. BMI was categorized according to the World Health Organization classification and intrinsic subtypes were approximated by immunohistochemistry using ER, PR, HER2/neu, CK 5/6, EGFR, and Ki67 as previously described and validated (Clin Cancer Res, 2008 and J Natl Cancer Inst, 2009).RESULTSDistribution of the sample set by BMI included 15 (2.2%) underweight, 190 (27.4%) ideal weight, 202 (29.1%) overweight, and 286 (41.3%) obese patients. Increasing BMI was associated with an increasing proportion of luminal B cases (28.6%, 27.9%, 33.3% and 38.9% for underweight, ideal weight, overweight and obese women respectively, p= 0.02). This was significant for premenopausal (p=0.004) but not postmenopausal women (p=0.4). Increasing BMI was also found to be associated with higher tumor grades (p= 0.03), but not with triple negative or HER2 subtypes.CONCLUSIONSIncreased BMI is associated with a unique breast cancer pathobiology favoring the development of high grade and luminal B breast cancers, especially in premenopausal women. The success of clinical trials targeting overweight and obese women to improve treatment response and outcome may be improved by focusing on this subtype.

Abstract Background: Molecular subtypes of breast cancer have been characterized by gene expression analysis. Luminal subtypes are hormone receptor (HR) positive and Her2/neu negative. Immunohistochemistry (IHC) has been used as a surrogate test for gene expression. Ki67 is a proliferation marker that identifies high-risk subtype of luminal breast cancer. Recently, the proposed Ki67 index (KI) of 14% was suggested as a cut-off to distinguish between luminal A and luminal B tumors (JNCI. 2009;101:736-750). The oncotype dx (ODx) is a 21- gene test that provides prognostic and predictive information in early stage HR-positive breast cancer patients. The test is reported as low (<18), intermediate (18-30) and high (>30) risk recurrence scores (RS) Design: We investigated the relationship between ODx RS and luminal subtypes of breast cancers using the KI of 14% as the cut-off for distinguishing luminal A and B tumors. Biomarker analysis (ER, PR, Her2/neu, Ki67 and p53) was performed as part of the diagnostic work-up using standard IHC procedures. Scoring was done by automated image analysis. Her2/neu FISH was performed on all IHC 2+ and 3 + results. Pathologic parameters such as, tumor size, grade, and presence of LVI were evaluated. Ploidy was performed by the Autocyte system (Tripath) on Feulgen stained paraffin sections. Results: We identified 106 patients with HR positive breast cancer who were tested for ODx from February 2006 to May 2010. 85/106 had Ki67 data available for analysis. 46/85 (54%) were luminal A and 39/85 (46%) luminal B tumors. The mean KI in luminal A was 6.93% versus 31.1% in luminal B (P<0.0001). The mean tumor size was 1.94 and 1.92 cm in luminal A and B respectively. Grade 1, 2 and 3 comprised 28/85 (32.9%), 49 (57.6%) and 8 (9.41%) of all tumors respectively. Of the grade1 tumors, 75% were luminal A, and grade 3 tumors were predominantly luminal B (p=0.013). LVI was present in 16 cases, 11 (68.8%) in luminal B and 5/16 (31.2%) in luminal A tumors (p=0.0416). Luminal A tumors were predominantly diploid 30/45 (66.6%) and luminal B were mostly aneuploid 21/32 (65.6%) (p=0.019). The overall mean RS in luminal A and B tumors was 14.67 and 20.15 respectively (P<0.0004). Luminal A tumors had low RS in 32/46 (66.6%). and luminal B tumors had predominantly intermediate/high RS in 23/39 (62.1%) (p=0.0082). ER Allred Scores were 7.1 and 7.3 and percent positivity was 88.1% and 91% respectively for luminal A and B subtypes. PR Allred scores were 5.6 and 5.8, and percent positivity was 62.1 and 52.9% for A and B tumors respectively. Information regarding treatment was available in 72 cases. 19 (26.3%) were treated by a combination of chemo and anti-hormonal therapy and 53 (73.6%) were treated by anti hormonal therapy alone, 8 of the 19 (42.1%) luminal A patients received combination therapy versus 11 (57.8%) in the luminal B category (p=0.277). Conclusion: Ki67 is a useful marker that showed significant correlation with RS by ODx. Luminal A tumors are more likely to be low grade, diploid with low RS compared to luminal B tumors. Ki67 in conjunction with other pathologic parameters may serve as a surrogate marker for the ODx RS. Citation Information: Cancer Res 2010;70(24 Suppl):Abstract nr P4-08-07.