Luminal A Research Articles

The Influence of Breast Cancer Subtype on Survival after Palliative Radiation for Osseous Metastases

BACKGROUND: Among patients with osseous metastases, breast cancer (BC) patients typically have the best prognosis. In the palliative setting, BC is often considered a single disease, but based on receptor status there are four distinct subtypes: luminal A (LA), luminal B (LB), triple negative (TN), and HER2-enriched (HER2). We hypothesize that survival and palliative outcomes following palliative RT for osseous metastases correlate with breast cancer subtype (BCS). METHODS: We identified 3,895 BC patients with known receptor status who received palliative RT for osseous metastases from 2004-2013 in the National Cancer Database. Kaplan-Meier method with log-rank testing and univariate/multivariate Cox-regression was used to identify survival factors. Incomplete radiation courses, 30-day mortality rate, and percentage remaining life spent receiving RT (PRLSRT) were calculated. RESULTS: Subtypes were 54% LA, 33% LB, 8% TN, and 5% HER2 with median survival of 34.1, 28.2, 5.3, and 15.7 months, respectively (p < 0.001). Overall 82% of patients received ≥10 fractions. Although BCS had limited effect on radiation regimens, TN received nearly twice as many single or hypofractionated (≤5 fractions) treatments, but the overall rate of these fraction schemes was low at 3.7 and 13.7%, respectively. Compared to LA and LB, TN and HER2 patients had worse palliative outcomes; higher rates of incomplete courses at 18.8% and 18.3% versus 12.7%-14.4%; higher 30-day mortality post-radiotherapy at 21.5% and 16.0% versus 6.3%-7.9%, and higher median PRLSRT of 7.7% and 3.7% versus 2.2%-2.4% for LA and LB. On multivariate analysis, BCS was associated with overall survival with TN (HR 3.7), HER2 (HR 1.75), and LB (HR 1.28) fairing worse than LA (p < 0.001). CONCLUSIONS: BCS correlated with survival and palliative outcome following radiation to osseous metastases. BCS should be considered by physicians when planning palliative RT to maximize quality-of-life, avoid unnecessary treatment, and ensure palliative benefits.

Abstract 5392: Exosomal miRNA expression profiling in triple-negative breast cancer

Abstract Exosomes are extracellular vesicles that are present in body fluids and known to play key roles in intercellular signaling communication. Exosomes carry several types of molecules of their tumor of origin, including miRNAs, and therefore present potential as biomarkers for cancer prognosis and treatment. Triple-negative breast cancer (TNBC) is a clinically aggressive subtype of breast cancer, associated with treatment resistance, recurrence and high mortality rates. These tumors are molecularly heterogeneous, which impairs the identification of effective prognostic molecular markers and target therapies. The main objective of this study was to characterize the miRNA expression profile of exosomes derived from TNBC cells in comparison to normal controls and to the luminal A (LA) breast cancer subtype. The exosomal miRNAs were isolated from 19 serum samples of breast cancer patients (10 with the LA and 9 with the TNBC subtype) and 10 serum samples of healthy women (control group) using the exosome precipitation method. Exosome characterization and quantification was assessed by Western blot using the exosomal surface markers, CD6 and CD9, and Nanotracking analysis. The tumor and normal exosomal samples were combined into 6 different pools (of at least 4 samples in each group) and sequenced in the Ion Torrent Proton platform (Applied Biosystems). RNA-seq analysis revealed differentially expressed (DE) levels of exosomal miRNAs, with log2FC ranging from 1.2 to 3.6, among the four groups studied (control, cancer (TNBC+LA), TNBC and LA groups): 22 DE miRNAs were observed in the cancer vs. control, 19 in the LA vs control, 30 in the TNBC vs. control and 7 in the LA vs. TNBC groups. Eleven miRNAs were commonly DE in the cancer, LA and TNBC groups when compared to the control group, which included miRNAs, such as miR-103a-3p, miR-107 and miR-423-5p, involved in the biosynthesis of unsaturated fatty acids and neutrophin signaling pathways. Nine miRNAs were only observed DE in the TNBC group (vs. control), with the majority of them being downregulated. The top 10 significant functional enriched pathways affected by these miRNAs included proteoglycans, pathways in cancer and TGF-beta signaling pathways. Only three miRNAs were observed DE in the TNBC vs. LA groups, all with downregulated expression in the TNBC group. These miRNAs were observed mostly involved in fatty acid biosynthesis, proteoglycans in cancer and TGF-beta signaling pathways. In conclusion, our results indicate that expression alterations of exosomal miRNAs affect critical cancer-associated signaling pathways, suggesting their potential use as noninvasive cancer biomarkers. Additional studies should be performed to determine whether the expression patterns of these exosomal miRNAs reflect the ones observed in the corresponding patients' TNBC tissue. Funding: CAPES, PPSUS-Fundação Araucária, Brazil. Citation Format: Patricia M. Ozawa, Débora S. Lemos, Evelyn Vieira, Ingrid L. Souza, Silvio M. Zanata, Vânia C. Pankievicz, Thalita R. Tuleski, Emanuel M. Souza, Rosiane V. Silva, Pryscilla F. Wowk, Rodrigo C. Almeida, Iglenir J. Cavalli, Danielle M. Ferreira, Luciane R. Cavalli, Enilze M. Ribeiro. Exosomal miRNA expression profiling in triple-negative breast cancer [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2018; 2018 Apr 14-18; Chicago, IL. Philadelphia (PA): AACR; Cancer Res 2018;78(13 Suppl):Abstract nr 5392.

Abstract 5660: Proteomic comparative assessment of flash-frozen and OCT embedded breast cancer tissues for utilization in precision oncology discovery studies

Abstract Proteomic analysis is an important tool in classifying molecular architecture of tissue samples for subtyping and precision oncology applications. Tissues are commonly either flash frozen (FF) or embedded in optimal cutting temperature (OCT) compound for long-term storage. OCT provides a convenient specimen matrix for cryostat sectioning but contains glycols and resins which are incompatible with mass spectrometric analysis of proteins. In contrast, flash frozen preservation of tissue is historically considered to be more suitable and compatible with mass spectrometry applications. As a pilot study, we evaluated different tissue preservation methods, sample processing methods, and protein extraction methods, and their impact on quantitation and coverage of the proteome. In this pilot study, 30 breast cancer cases with matching flash frozen and OCT-embedded samples were used for proteomic analysis. Tissues were procured from the Clinical Breast Care Project (CBCP) in partnership with the Chan Soon-Shiong Institute of Molecular Medicine at Windber, and the Walter Reed National Military Medical Center (WRNMMC). Breast cancer subtypes were determined by clinical and expanded immunohistochemical panels. These subtypes were Her2, Luminal A (LA), Luminal B (LB) and Triple Negative (TN). Flash frozen tissues were cryosectioned to isolate 60-100 mg total tissue per case, and proteins were extracted using an 8M urea lysis buffer. The OCT specimens were laser microdissected to isolate primarily tumor cells from the samples. Proteins were extracted using the illustra TriplePrep kit. Digested proteins were then multiplexed using Tandem-Mass-Tag (TMT) 10plex isobaric labeling reagents, and chromatographically resolved using a Waters 2D nanoAcquity liquid chromatographer (LC) coupled to a Thermo Q Exactive Plus mass spectrometer (MS). Data was then analyzed using Proteome Discoverer. In total, 6130 proteins were identified using a 1% peptide FDR confidence filter in the 30 samples analyzed. Unsupervised hierarchical clustering of proteins across the samples resulted in two primary clusters, OCT and FF. Within the primary clustering, there were sub-clusters of Luminal, Her2, and Triple Negatives. At 2-fold threshold, 246 proteins were differentially expressed between OCT and FF tissues. However, the differential proteins between ER+ (LA and LB) and ER- (Her2 and TN) tumors correlate irrespective of the cryopreservation and processing methods. In conclusion, proteome analysis of samples stored under two different conditions provide subtyping information; however, there is differential expression of proteins between the two types of sample handling techniques which can impact biological interpretation of breast cancer precision oncology studies. Citation Format: Kiki Panagopoulos, Punit Shah, Albert Kovatich, Michael A. Kiebish, Jeffrey Hooke, Leigh Campbell, Mary Lou Cutler, Audrey Kovatich, Hai Hu, Brenda Deyarmin, Lori Sturtz, Praveen Kumar Raj Kumar, Stella Somiari, Emily Y. Chen, Elder Granger, Viatcheslav Akmaev, Rangaprasad Sarangarajan, Niven Narain, Craig D. Shriver. Proteomic comparative assessment of flash-frozen and OCT embedded breast cancer tissues for utilization in precision oncology discovery studies [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2018; 2018 Apr 14-18; Chicago, IL. Philadelphia (PA): AACR; Cancer Res 2018;78(13 Suppl):Abstract nr 5660.

Abstract 4222: The prognostic impact of human cytomegalovirus infection in breast cancer

Abstract Emerging evidence demonstrates a frequent presence of Human Cytomegalovirus (HCMV) proteins and nucleic acids in several primary malignant tumours including breast cancer (BC). HCMV maintains latency in monocytes and reactivates in an inflammatory or immune-suppressive state. The virus infection expresses more than 750 proteins, many has revealed both oncogenic and onco-modulatory properties which fulfil all the criteria for the hallmarks of cancer. HCMV is not believed to transform the host cell itself, but HCMV-encoded proteins are involved in cell cycle progression, interact with key proteins in the extrinsic and intrinsic apoptotic signalling cascade, and can induce drug resistance to chemotherapeutic agents. Earlier studies from our lab detected HCMV-IE (immediate early) and HCMV-LA (late antigen) proteins in &amp;gt;90% of analysed primary breast cancer tumours, sentinel lymph node and brain metastasis, and that their prevalence is higher in advanced breast cancer. We have reported a significant correlation between HCMV and hormone receptor expression in a certain subgroup of breast cancer patients. The purpose of this study is to retrospectively investigate the prognosis and overall survival of 120 patients diagnosed and treated for breast cancer at the Akershus University Hospital, Norway, between 1996 and 2010. Material/methods: Formalin-fixed, paraffin-embedded, consecutively collected breast tumour samples and metastases, from clinically well-characterized patients were examined for HCMV-IE and HCMV-LA, using immuno-histochemical techniques. HCMV values was then correlated with BC subtypes: HER-2 positive, luminal A (LUM-A), luminal B (LUM-B) and triple negative breast cancer (TNBC). As well as time to: distant metastasis, loco-regional relapse and overall relapse. All data were defined in Kaplan-Meier curves for survival and a Cox regression model was fitted to each subgroup. Results: Both HCMV-IE and HCMV-LA was present in &amp;gt;99% of the tissue samples analysed. We found a significant association between HCMV-LA and overall survival in TNBC patients (Hazard ratio 2.502; P-value 0.025). In addition, HCMV-LA was significantly correlated to the risk of local relapses in TNBC patients (Hazard ratio 4.149; P-value: 0.026). Finally, we also found the risk for local relapses and distant metastasis to be significantly tied to HCMV-LA levels in TNBC patients (Hazard ratio 4.126; P-value: 0.032). HCMV-LA and HCMV-IE levels were not found to be associated with either survival or risk for relapses in any other BC subtype. Conclusion: Our findings reveal a significant correlation between HCMV-LA expression and TNBC subtype concerning both risk of recurrence and overall survival. While no other BC subtype was found to be significantly influenced by the presence of HCMV-related proteins, the presence of HCMV-LA that imply an active HCMV infection seems to indicate a more aggressive disease when detected in TNBC patients. Citation Format: Joel Touma, Afsar Rahbar, Belghis Davoudi, Jonas Christoffer Lindstrom, Torill Sauer, Ida Rashida Bukholm, Katja Vetvik, Cecilia Soderberg-Naucler, Jurgen Geisler. The prognostic impact of human cytomegalovirus infection in breast cancer [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2018; 2018 Apr 14-18; Chicago, IL. Philadelphia (PA): AACR; Cancer Res 2018;78(13 Suppl):Abstract nr 4222.

Luminal and basal subtyping of metastatic castration-resistant prostate cancer (mCRPC) and its clinical implications.

197 Background: The PAM50 gene expression classifier (PAM50) identifies luminal and basal subtypes and predicts response to androgen deprivation therapy in localized prostate cancer. The clinical utility of using PAM50 to molecularly subtype mCRPC was evaluated. Methods: PAM50 was applied to RNA expression data from 86 metastatic tumor biopsies from the SU2C-AACR-PCF West Coast Prostate Cancer Dream Team (WCDT; NCT02432001). Overall survival (OS) differences between luminal A (LuA), luminal B (LuB), or basal patients (pts) were determined using Kaplan-Meier analyses. PAM50 was also applied to 15,136 prospectively collected radical prostatectomy (RP) samples from the Decipher GRiD database (NCT02609269). Drug response signatures (DRS) for 89 drugs were derived using publicly available data from the NCI60 cell line panel, and applied to gene expression data from the RP samples to predict patient-specific drug sensitivities. Differences in DRS as a function of PAM50 subtype were assessed using the Pearson’s chi-squared test. Results: The application of PAM50 to mCRPC transcriptional data segregated pts into LuA, LuB, and basal populations (43%, 14%, and 43%, respectively). The median OS for LuA, LuB, and basal pts was 20.6 months, 9.5 months, and 10.4 months, respectively (p=0.04), which was consistent with localized prostate cancer where LuB pts have the worst prognosis. DRS analyses revealed statistically significant differences in drug sensitivities, with LuA and LuB pts predicted to be markedly more sensitive to docetaxel than basal pts (p&lt;0.00001), and basal pts predicted to be markedly more sensitive to platinums and etoposide than LuA and LuB pts (p&lt;0.00001). Conclusions: PAM50 subtyping is prognostic in mCRPC, with LuA pts demonstrating the longest OS. Luminal and basal subtypes have distinct in silico drug response profiles that may be associated with response to mCRPC therapies. Prospective testing of DRS as a biomarker to guide treatment in mCRPC is warranted. Clinical trial information: NCT02432001, NCT02609269.

Abstract P2-06-04: Use of principal component analyses to select ER-balanced subset for gene centering in PAM50 subtyping

Abstract Background: PAM (Prediction Analysis of Microarray) 50 is an established gene expression-based algorithm to classify breast tumors into basal-like, HER2-enriched, luminal A (LA), and luminal B (LB) subtypes. Clinical subtyping is mainly based on immunohistochemistry (IHC) assays of estrogen receptor (ER), progesterone receptor (PR), human epidermal growth factor receptor 2 (Her2) and Ki67 classifying tumors into triple-negative (ER-/PR-/Her2-), Her2+ (ER-/PR-/Her2+), LA (ER+/Her2-/Ki67-), LB1 (ER+/Her2-/Ki67+) and LB2 (ER+/Her2+). These two subtyping methods do not completely match even on comparable subtypes. Nevertheless, the ER-balanced subset for gene-centering in PAM50 subtyping was selected based on clinical status. Here we explored the possibility of using principal component analyses and iterative PAM50 call to refine the selection of an ER balance subset to improve consistency between these methods focusing on LB calls which is more aggressive than LA tumors. Methods: Normalized gene expression data was obtained from TCGA research network for 712 primary tumors which had IHC status available for ER, PR and Her2. Since Ki67 status was not available LA and LB was discriminated for ER+ cases with Her2- and Her2+ respectively. In house RNA-Seq dataset had 118 primary tumors and were drawn from the Clinical Breast Care Project where breast cancer patients were consented using an IRB-approved protocol. Tumors were selected and processed by laser microdissection. RNA was extracted from tissues using the Illustra triplePrep kit (GE Healthcare). Paired-end mRNA sequencing was performed using the Illumina HiSeq platform. Sequenced reads were processed using PERL based pipeline utilizing PRINSEQ, GSNAP and HTSeq. Principal component analysis (PCA) was done using R. Wilcoxon rank sum test was used for statistical significance (p&amp;lt;0.05). Results: In both datasets, the PCA map grouping of cases does not perfectly reflect the clinical subtypes. This motivated us to select ER balance subset based on the PC1 separation and IHC subtype. The resulting PAM50 subtypes on PCA map distinguished Basal and LA as two well separated components. Using all of Basal and equal number of LA cases for ER balance subset for PAM50 resulted in increased LB call and a better consistency with IHC LB calls. Among 712 cases in TCGA LB numbers increased from 142 in initial PAM50 call to 203 in ER balanced refined PAM50 call. We noticed that there was significantly higher (p-value = 4.414e-11) MKI67 expression for the 39 cases switch from LA to LB between PAM50 calls. Similar trend was observed in our in-house dataset where majority of the IHC-LB1 cases was called as LB in PAM50. The new method increased LB call from 22 to 27 which in-turn increased consistency between molecular and clinical subtypes from 73 to 79 out of the total of 118 cases. Conclusion: We show that an iterative PAM50 call coupled with PCA for selection of ER balance set potentially enhanced the consistency of the LB calls with clinical subtyping and that the tumors switched from LA to LB have high MKI67 expression. The views expressed in this article are those of the author and do not reflect the official policy of the Department of Army/Navy/Air Force, the Department of Defense, or U.S. Government. Citation Format: Raj-Kumar P-K, Liu J, Kovatich AJ, Kvecher L, Shriver CD, Hu H. Use of principal component analyses to select ER-balanced subset for gene centering in PAM50 subtyping [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 P2-06-04.

Abstract P3-10-17: Predictors of overall survival and disease-free survival for women diagnosed with invasive breast cancer: A large single-center study

Abstract Background/Objective: Biomarkers such as estrogen receptor (ER), progesterone receptor (PR) and human epidermal growth factor 2 (HER2neu) are associated with the survival and recurrence of invasive breast cancer (IBC). Presence of these markers often helps to dictate treatment plans. The purpose of this study is to identify the predictors of overall survival (OS) and disease-free survival (DFS) for women diagnosed with IBC. In this study, we compared OS and DSF in the following five categories based on biomarker expression: triple-negative (TN), triple-positive (TP), HER2neu-positive, Luminal A (LA) and Luminal B (LB). Method: We retrospectively examined 6,231 records of women diagnosed with IBC from 2006 through 2015 within the Aurora Health Care system. The majority of the women were white (90.8%) and non-Hispanic (97.4%). The women in this study were categorized in the age groups: 20-39 (4.9%), 40-59 (42.8%), 50-69 (40.7%) and 70+ years (11.6%). We used descriptive statistics for all category and numeric variables. Kaplan-Meier curve was used to compare OS and DFS in the five categories. Cox proportional hazards regression was used to identify the significant predictors of OS and DFS. We used alpha of 0.05 for all statistical tests and all statistical analysis was done using SAS 9.4. Results: OS was 86.39% and the DFS rate was 84.43%. In multivariate analysis significant predictors of OS included age (HR=1.94 for 60-69 vs 20-39 p=0.001 and HR=5.3 for 70+ vs 20-39 p&amp;lt; 0.001), race (HR=0.57, p=0.043 for white vs other), ethnicity (HR=0.34, p=0.032), HER2neu expression (HR=0.58, p &amp;lt;0.001), size of the tumor (HR=1.001 p=0.011), grade differentiation (HR=0.62 p &amp;lt; 0.001) and cancer stages (HR of 1.17, 4.8, 15.2 for II, III and IV vs I, p &amp;lt;0.001). Furthermore treatments such as surgery (HR=0.64, p&amp;lt; 0.001), radiation (HR=0.64, p &amp;lt;0.001), chemotherapy (HR=0.63, p &amp;lt;0.001) and hormonal therapy (HR=0.48, p &amp;lt;0.001) showed significantly improved OS in the patients. Similarly significant predictors of DFS included, age (HR=1.8, p &amp;lt;0.001 for 70+ vs 20-39); PR (HR=0.70, p=0.003), HER2neu (HR=0.60, p &amp;lt;0.001) tumor size (HR=1.001, p=0.002), cancer stage (HR=2.7, 9.2, 25.5 for II, III, IV vs I, p&amp;lt;0.001); differentiation of the tumor (HR=0.75, p=0.002, moderate differentiation; well differentiated, HR=0.50, p &amp;lt;0.001). Moreover, the treatments such as surgery (HR=0.42, p &amp;lt;001) radiation, chemotherapy and hormonal therapy (HR=0.58, 0.62, 0.51 p &amp;lt;0.001) significantly increased DFS. Furthermore OS was better for women diagnosed with TP, LA, HER2, LB types compared to women with TN (p&amp;lt;0.001). Similarly the probability for DFS was higher for TP also was higher compared to the probabilities all the other categories (LA, LB, HER2 and TN p &amp;lt;0.05). LA had better DFS then LB, HER2 and TN. LB and HER2 had better DFS than TN (p&amp;lt;0.05). Conclusion: This large study suggests that treatment based on biomarker expression (ER, PR and HER2/Neu) has improved OS and DFS for women with IBC. Increase in tumor size, cancer stage and poor differentiation of the tumor had adverse effect on both OS and DFS. Treatment modalities such as surgery, radiation, chemotherapy and hormonal therapies significantly improved both OS and DFS. Citation Format: Singh M, Konduri SD, Bobustuc GC, Rovin RA. Predictors of overall survival and disease-free survival for women diagnosed with invasive breast cancer: A large single-center study [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 P3-10-17.

Abstract P5-02-04: Opto-acoustic imaging of breast masses: Correlation with breast biopsy prognostic indicators

Abstract Purpose: The Imagio™ OA/US breast imaging system, a diagnostic opto-acoustic (OA) imaging device bearing the CE Mark, is in the U.S. FDA Premarket Approval process. OA/US provides both functional (relative oxygenation/de-oxygenation) and anatomic (angiogenesis) information that is co-registered and temporally interleaved in real time with gray-scale ultrasound that may improve discrimination between benign and malignant masses. We recently reported correlation studies demonstrating tumor-zone specific OA attributes in histopathologic grade I versus grade III malignancies. The relationship between OA attributes (individual feature scores or summed feature results) and pathologically-determined prognostic markers (PDPM) in malignant lesions is the subject of this report. Materials and Methods: In this HIPAA-compliant, IRB-approved prospective multi-center trial across 16 U.S clinical sites; 1,808 masses in 1,739 subjects assessed as BI-RADS 3, 4 or 5 were imaged with OA/US. Of these, 655 were invasive malignancies and the subject of this analysis. Each mass was scored by 8 blinded readers on 3 internal zone features of the tumor nidus and 2 external features (0-5, 6) of the tumor boundary and peripheral zones (OA attributes). Pathologic diagnoses were confirmed by an experienced central breast pathologist blinded to the OA assessment. Tumor histologic classification and grading was performed in all subjects.Evaluation of tumor estrogen receptor (ER) and progesterone receptor (PR) were performed at each site by immunohistochemistry (IHC) and was reported as percent of tumor cells expressing the receptor or, as positive if greater than 1%. Tumor HER2-neu expression was reported by IHC as 0, 1+ (negative, not over-expressed), 2+ (indeterminate) and 3+ (over-expressed). All 2+ results reflexed to fluorescence in-situ hybridization (FISH) and reported as over-expressed or not over-expressed. Tumor Ki-67 expression was evaluated with IHC and reported as percent of tumor cells positive for the antigen. Statistical analysis of categorical measures of PDPM is in process and will be performed using a two-way Analysis of Variance (ANOVA) and Tukey HSD (honest significant difference) test for pairwise comparisons. This ANOVA will be repeated for each PDPM to test which specific PDPM sub-categories are related to OA attributes. Correlation coefficients will be generated for PDPM that are continuous, not categorical. All statistical testing will be done at a 5% significance level. Results: A total of 655 invasive and 22 DCIS were scored for internal (nidus) and external (boundary and periphery) OA attributes and compared with PSBC as defined by ER, PR, Her2 and Ki-67 expression. Of these, 108 were Luminal-A (LA), 153 Luminal-B (LB), 80 Triple-negative (TN), 23 Her2-enriched (HER2) and 314 unclassified (including 22 DCIS). OA attributes differentiated LA (99%CI 2.8,3.1) from TN (99%CI 3.1,3.4), p=0.027 and HER2 (99%CI 3.1,3.6), p=0.036. OA features strongly suggested LA vs. LB (99%CI 3.1,3.3) subtype, p=0.060. LB vs.TN(p=0.59) and HER2(p=0.41) were non-significant. TNBC vs. HER2 was p=0.62. Citation Format: Grobmyer SR, Butler R, Neuschler EI, Stavros AT, Aitchison RD, Lavin PT, Tucker FL. Opto-acoustic imaging of breast masses: Correlation with breast biopsy prognostic indicators [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 P5-02-04.

18F-Fluorodeoxyglucose Positron Emission Tomography/Computed Tomography Features in Locally Advanced Breast Cancer and Their Correlation with Molecular Subtypes.

Purpose:18F-Fluorodeoxyglucose positron emission tomography/computed tomography (18F-FDG PET/CT) is now recognized as a staging investigation for locally advanced breast cancer. This retrospective review of data was performed to correlate the maximum standardized uptake value (SUVmax) of the primary tumor with the molecular subtype of breast cancer.Materials and Methods:Patients with biopsy-proven, treatment naïve, Stage III breast cancer, for whom 18F-FDG PET/CT data and immunohistochemistry 4 was available were included in the study. Correlations were deduced between the SUVmax of primary tumor to the molecular subtypes.Results:Three hundred and two patients were included in the study. Fifty-two (17.2%) tumors were Luminal A (LA), 131 (43.4%) Luminal B (LB), 42 (13.9%) human epidermal growth factor receptor-2 enriched (HE), and 77 (25.5%) basal-like (BL). SUVmax of the primary tumor differed significantly between LA and other subtypes (SUVmax: LA Median 7.4, LB 11.65, HE 13.5, BL 15.35, P < 0.001). Estrogen receptor (ER) and progesterone receptor (PR) positivity were inversely correlated to the SUVmax of the primary (SUVmax: ER + Median 10.4, ER - 14.2, P < 0.001, PR + 9.65, PR − 13.9, P < 0.001). There was a strong positive correlation observed between Ki67 and SUVmax (Pearson Coefficient 0.408, P < 0.001). A SUVmax value of 9.65 was determined as a cutoff on receiver operating characteristic curve to differentiate between LA and other subtypes with a sensitivity of 92.3% and specificity of 70.6%.Conclusions:SUVmax of primary showed a statistically significant difference between LA subtypes when compared to other subtypes. However, there was overlap of values in each subgroup and thus 18F-FDG PET/CT cannot be used to accurately assess the molecular characteristics of the tumor.

184P - Distribution of genomically defined recurrence risk in luminal A and B breast tumors defined by inmunohistochemistry: A retrospective study in Spanish population

Background: Semiquantitative inmunohistochemical (IHC) expression of progesterone receptor (PR) adds prognostic value to the current IHC-based luminal A (LA) definition, such that patients (pts) with LA (Ki67 <14% and PR > 20%) tumors can be spared from adjuvant chemotherapy (CT). Oncotype Dx® (ODX) and MammaPrint® (MP) assays have been validated as predictors of CT benefit. This study assessed the distribution of recurrence risk in LA and LB breast tumors as defined by Ki67 and PR. Methods: A retrospective analysis was performed in 889 T1-2, N0-Nmic, M0 tumors for which ODX or MP results and local pathology data were available. Ki67 was assessed by IHC (high ≥14% and low <14%). PR was assessed by semiquantitative IHC (PR low < or = 20%, high >20%). Histological grade was defined using the Nottingham grading system. Results: Median age 54 years (18-77). All pts had HER2 negative tumors. Median tumor size 15 mm (2-88). Three hundred (33.7%) tumors were classified as LA and 589 (66.3%) as LB. Grade 1 tumors were higher in LA (27%) than in LB (19%) pts (p < 0,001). CT was first recommended in 137 pts (45.7%) with LA vs. 361 pts (61.3%) with LB tumors. ODX was performed in 432 (48.6%) pts and MP in 457 (51.4%). Recurrence risk distribution varied significantly between groups (p < 0.001). ODX: among LA pts, 71.4%, 25.7% and 2.9% had low, intermediate and high recurrence risk respectively; among LB pts respective values were 46.2%, 44.2% and 9.6%. MP: among LA pts, 81.2% and 18.8% had low and high recurrence risk, respectively; among LB pts respective values were 54.2% and 45.8% (Table). After test results CT was recommended to 61 pts (20.3%) with LA vs. 268 pts (45.5%) with LB tumors.Table184PRecurrence Risk (%)LA (n = 300)LB (n = 589)ODX (n = 432)Low71.446.2Intermediate25.744.2High2.99.6MP (n = 457)Low81.254.2High18.845.8 Open table in a new tab Conclusions: There is a wide distribution of recurrence risk results between LA and LB tumors defined by Ki67 and PR which confirms the important role of gene-expression assays in adjuvant decision making. Of note about half of pts with LB tumors had low recurrence risk indicating minimal benefit from adjuvant CT. Legal entity responsible for the study: Ministry of Health of the Community of Madrid (Spain) Funding: Ministry of Health of the Community of Madrid (Spain) Disclosure: S. López-Tarruella: Advisory role: Novartis, Pfizer, Celgene. Honoraria/lecturer: Novartis, Pfizer, AstraZeneca, Celgene, Roche. I. Márquez-Rodas: Advisory role: Bristol-Myers Squibb, MSD, Novartis, Roche, Pierre Fabre, Amgen and Bioncotech. M. Martin Jimenez: Holds a patent on the predictive value of the PAM50 gene profile assay. All other authors have declared no conflicts of interest.

Abstract 3372: Differential exon usage in the HER2 subtype of breast cancer identified with RNA-seq and proteomic data

Abstract Background: Heterogeneity between different breast cancer tumors plays an important role in patients’ survival outcomes, and such tumor heterogeneity unveiled by transcriptome and proteome are often in disagreement. We hypothesize that alternative splicing of mRNA could explain heterogeneity within intrinsic breast cancer subtypes. Methods: Cases used in this study were drawn from the Clinical Breast Care Project where breast cancer patients were consented using an IRB-approved protocol. Fifty breast tumors were selected and processed by laser microdissection. RNA and protein were extracted from tissues using the Illustra triplePrep kit (GE Healthcare). Paired-end mRNA sequencing was performed using the Illumina HiSeq platform. Paired-end reads were preprocessed using PRINSEQ and splice-aligned to the genome using GSNAP software. Gene counts were measured using HTSeq while Exon counts used the DEXSeq. Differential expression was called at 10% false discovery rate. Proteome Discoverer with Byonic node was used for analyzing the quantitative global proteomics dataset, and we were able to quantitate 8600 proteins. All other analyses were performed using Perl and R. Results: The number of sequencing reads for the 50 cases ranged from 60 million to 410 million reads. After preprocessing, an average of 93% of reads was mapped, and 36,700 genes were identified among 50 tumor samples. PAM50 algorithm was used for intrinsic subtype calls, yielding 16 Basal, 8 Her2+, 19 Luminal A (LA) and 7 Luminal B tumors. Unsupervised clustering of the 8 Her2+ cases using the PAM50 genes and highly varying proteins resulted in different clustering patterns, with the latter clustering 6 of the 8 Her2+ cases with two distinct 3-case sub-clusters. Between these two sub-clusters, there were 10 differentially expressed (DEX) genes and 25 DEX proteins, but none of the 25 proteins were mapped to the 10 DEX genes. This motivated us to investigate the DEX exons, and we found 7,076 DEX exons between the Her2+ sub-clusters, and 9 of the 25 DEX proteins were matched to genes bearing DEX exons. For comparison, we performed the same analyses between similarly clustered Basal and LA sub-clusters. Even though the number of DEX genes between Basal (8) and LA (18) sub-clusters were comparable to that between the Her2+ sub-clusters, DEX exons were much lower for both subtypes (616 &amp; 157), and none of the DEX proteins (3 &amp; 7) mapped to DEX exons or genes for either subtype. Conclusions: Our findings imply that there is more proteomic-level heterogeneity in the Her2+ subtype than in Basal and LA subtypes, which could be due to alternative usage of exons (alternative splicing). If such heterogeneity is associated with patients’ survival outcomes then our results will further stress the importance of alternative splicing in breast cancer. The views expressed in this article are those of the author and do not reflect the official policy of the Department of Defense, or U.S. Government. Citation Format: Praveen Kumar Raj Kumar, Tao Liu, Lori A. Sturtz, Albert Kovatich, Marina A. Gritsenko, Vladislav A. Petyuk, Brenda Deyarmin, Viswanadham Sridhara, James Craig, Jason E. McDermott, Anil K. Shukla, Ronald J. Moore, Matthew E. Monroe, Bobbie-Jo M. Webb-Robertson, Jeffrey A. Hooke, Leigh Fantacone-Campbell, Leonid Kvecher, Jianfang Liu, Jennifer Kane, Jennifer Melley, Stella Somiari, Joji Iida, Stephen C. Benz, Justin Golovato, Shahrooz Rabizadeh, Patrick Soon-Shiong, Richard D. Smith, Richard J. Mural, Karin D. Rodland, Craig D. Shriver, Hai Hu. Differential exon usage in the HER2 subtype of breast cancer identified with RNA-seq and proteomic data [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2017; 2017 Apr 1-5; Washington, DC. Philadelphia (PA): AACR; Cancer Res 2017;77(13 Suppl):Abstract nr 3372. doi:10.1158/1538-7445.AM2017-3372

Failure patterns according to molecular subtype in patients with invasive breast cancer following postoperative adjuvant radiotherapy: long-term outcomes in contemporary clinical practice

Although gene expression profiling provides critical information, knowledge remains limited regarding the differential effects of molecular subtype on clinical course. This study evaluated the impact of molecular status on long-term patterns of failure in patients with non-metastatic breast cancer. We analyzed data from 1181 individuals with invasive breast cancer undergoing surgery plus PORT from 2003 to 2011. Molecular subtypes were defined as luminal A (LA), luminal B (LB)-HER2(-), LB-HER2(+), HER2, and triple-negative (TN) based on the 2013 St. Gallen Consensus criteria. Competing risks analysis and baseline hazard rate function plots were used to explore subtype-specific recurrence patterns. The 10-year overall survival rates of LA, LB-HER2(-), LB-HER2(+), HER2, and TN groups were 96, 93, 94, 84, and 85%, respectively (P<0.001). Distant metastatic events differed significantly according to molecular subtype (P<0.001). In competing risks regression analysis, initial development of distant metastasis was the highest with TN tumors, followed by HER2, LB-HER2(-), and LB-HER2(+) subtypes (P=0.005). Regarding preferential sites of distant metastasis, the risk of initial brain metastasis was significantly higher with HER2 tumors, followed by TN tumors (P=0.001). A low-level but sustained metastatic risk increment was observed in luminal tumors, whereas TN and HER2 subtypes showed a short-term risk surge within 5years. From the significant impact of molecular profile on distant metastasis, subtype-specific individualization of systemic treatment and close surveillance are suggested. The preferential and long-term risk of brain metastasis in the HER2 subtype underlines the importance of alternative anti-HER2 therapies.

Breast carcinoma subtypes show different patterns of metastatic behavior.

The aim of our retrospective study was to analyze patterns of subtype specific metastatic spread and to identify the time course of distant metastases. A consecutive series of 490 patients with breast cancer who underwent surgery and postoperative treatment at Semmelweis University, Hungary, and diagnosed between the years 2000 and 2007 was identified from the archives of the 2nd Department of Pathology, Hungary. Molecular subtypes were defined based on the 2011 St. Gallen recommendations. Statistical analysis was performed with SPSS Statistics for Windows, Version 22.0. Distant metastasis free survival (DMFS) was defined as the time elapsed between the first pathological diagnosis of the tumor and the first distant metastasis detection. Distant metastases were detected in 124 patients. Mean time to develop metastasis was 29months (range 0-127months). The longest DMFS was observed in the Luminal A (LUMA) subtype (mean 39months) whereas the shortest was seen in the HER2-positive (HER2+) subtype (mean 21months; p=0.012). We confirmed that HER2+ tumors carry a higher risk for distant metastases (42.1%). LUMA-associated metastases were found to be solitary in 59% of cases, whereas HER2+ tumors showed multiple metastases in 79.2% of cases. LUMA tumors showed a preference for bone-only metastasis as compared with HER2+ and triple negative breast cancer (TNBC) cases, which exhibited a higher rate of brain metastasis. The most frequent second metastatic sites of hormone receptor (HR) positive tumors were the lung and liver, whereas the brain was the most affected organ in HR-negative (HR-) cases. Tumor subtypes differ in DMFS and in pattern of distant metastases. HER2+ tumors featured the most aggressive clinical course. Further identification of subtype-specific factors influencing prognosis might have an impact on clinical care and decision-making.

Quantitative proteomics revealed novel proteins associated with molecular subtypes of breast cancer

The early diagnosis and successful treatment of breast cancer (BC) is still a challenging task due to the diverse origin and functional heterogeneity of cancer cells. The heterogeneity of BC may likely to explained by molecular BC subtypes, comprises Luminal-A (LA), Luminal-B (LB), Triple-negative (TN) and HER2-positive (HP), which are governed by a variety of cancer associated pathways. To identify protein signatures in different BC subtypes, we performed isobaric tag for absolute and relative quantitation (iTRAQ) of enriched blood plasma samples of BC subtypes (N=32) and healthy subjects (N=8). After analyses of data, 58 proteins were found to be modulated in BC subtypes from healthy subjects (p<0.05) and among these; Fibronectin (FN1), Alpha-2-macroglobulin (A2M), and Complement component-4-binding protein-alpha (C4BPA) and Complement factor-B (CFB) were selected for validation in BC subtypes and healthy subjects in the independent set of blood plasma (N=100) and tissue samples (N=25). Statistical analysis showed the significant modulation of FN1 and C4BPA in LB, and A2M in TN patients in both plasma as well as tissues comparatively control (p<0.05). Further, FN1 and C4BPA in LB subtype revealed a good diagnostic accuracy in plasma level validation. The receiver operating characteristic (ROC) curve and regression analysis demonstrated that these proteins with associated criterion of expression could act as discriminating signatures among BC subtypes with diagnostic and prognostic relevance. SignificanceThe heterogeneity of breast cancer (BC) has gained many challenges for successful management of BC, thus, the delineating proteomic alterations BC subtypes may provide great clinical values in diagnostic, prognostic and therapeutics of BC. The findings from the present quantitative proteomic study have deciphered the altered proteomic patterns and their possible molecular interactions in each BC subtype. The study showed a strong association of FN1, A2M, C4BPA and CFB in molecular subtypes of BC, in which, C4BPA and A2M demonstrated a potent signature in blood plasma and tissue samples of LB and TN subtypes in BC patients, respectively. The findings also revealed the altered level expressions of these selected proteins could classify BC subtypes through plasma and tissue based expression analysis in patients and control samples. Hence, these proteins could have clinical importance for the diagnosis and prognosis purposes among molecular BC subtypes.

Outcomes According to Breast Cancer Subtype in Patients Treated With Accelerated Partial Breast Irradiation

BackgroundThe purpose of the study was to determine outcomes for patients treated with accelerated partial breast irradiation (APBI) on the basis of breast cancer subtype (BCST). Patients and MethodsOur single-institution, institutional review board-approved APBI database was queried for patients who had complete testing results for the estrogen (ER), progesterone (PR), and HER2/neu receptors to determine outcomes for each BCST. Women were assigned as luminal A (LA), luminal B (LB), HER2, and basal BCST using their ER, PR, and HER2/neu receptor status. Degree of ER expression supplemented the receptor-based luminal BCST assignment. Two hundred seventy-eight patients had results for all 3 receptors (LA = 164 [59%], LB = 81 [29%], HER2 = 5 [2%], basal = 28 [10%]), which were submitted for analysis (ipsilateral breast tumor recurrence [IBTR], regional nodal failure, distant metastasis [DM], disease-free survival [DFS], cause-specific survival [CSS], and overall survival [OS]). ResultsMedian follow-up was 5.4 years (range, 0.1-12.4 years). Basal and HER2 subtype patients had higher histologic grades (Grade 3 = 75% vs. 10% LA/LB; P < .001), larger tumors (13.0 mm basal vs. 10.7 mm LA/LB; P = .059), and were more likely to receive chemotherapy (68% vs. 15% LA/LB; P < .001). Margin and nodal status were similar among BCSTs. At 5 years, IBTR rates were similar (1.8%, 2.9%, 0%, and 4.8%) for LA, LB, HER2, and basal subtypes, respectively (P = .62). DM was only seen in LA (2.9%) and LB (1.3%) (P = .83). DFS (95%-100%), CSS (97%-100%), and OS (80%-100%) were not statistically different (P = .97, .87, .46, respectively). ConclusionFive-year local control rates after breast-conserving surgery, APBI, and appropriate systemic therapy are excellent for luminal, HER2, and basal phenotypes of early-stage breast cancer; however, further study of receptor subtype effect on risk stratification in early-stage breast cancer is needed.

Breast Cancer Subtypes in Patients Aged 70 Years and Older

ABSTRACTRecurrence and survival pattern in breast cancer (bc) patients (pts) ≥ 70 years subcategorized according to subtype and age are still an area of uncertainty. Tumor characteristics, patient demographics, therapies applied, and recurrence pattern were compared between luminal A (LA), luminal B (LB), Her2/neu overexpressing (Her+) and triple-negative (TN) bc subtypes and the age subcategories 70–74, 75–79, ≥80 years. Based on univariate Cox-regression-analyses distant-disease-free-survival (DDFS) differed significantly for bc subtypes (p = 0.0002), notably for Her+ vs. LA (p = 0.0014), TN vs. LA (p < 0.001), and TN vs. LB (p = 0.0086). Not age, but Her+ and TN represented prognostic factors for DDFS.

Discrepancies between immunohistochemistry (IHC) form core biopsy (CB) and surgical specimen (SP) in patients with early breast cancer (EBC): A clinical decision-making challenge.

e12072Background: Various studies analyze the differences found between the IHC from CB and the SP. The discordance in this information can bring serious problems when it comes to making clinical decisions. Methods: In this retrospective study, we evaluated female patients with invasive EBC that were admitted to IOHM between June of 2014 and December of 2015, but only those in which the CB and SP had been carried out by the same medical team and whose IHC (Ki67, ER, PR and Her2) had been carried out by the same certified pathology laboratory. All tumors were classified as: Luminal A (LA); Luminal B Her2 negative (LB); Luminal B Her positive (LBH+); Her2 and Triple negative (TN). The correlation between ICH of the CB and SP was studied according to the Kappa test. Results: Seventy-three patients were included. Median age 53 years (r = 26-89) Stages (I = 27;II = 26,III = 20). In 45 cases (62%) there was no discrepancy but in 28 cases (38%) there was a discrepancy between the IHC from CB and the SP, which ha...

Abstract P1-01-03: Variability of Ki67 labeling index depending on different measurements in luminal type breast cancers

Abstract Background: Hormone receptor-positive and HER2-negative breast cancers are divided into luminal A (LA) and luminal B (LB) types using a cutoff of 14% Ki67 labeling index (LI). LB type breast cancer is associated with poor prognosis compared with LA, and considered as an indication of adjuvant chemotherapy. The main obstacles in classification of luminal type breast cancers are non-standardized measurement and Ki67 heterogeneity. We aimed to assess Ki67 variability depending on different methods of measurement and spot selection, and to assess Ki67 heterogeneity in luminal type breast cancers. Design: We retrospectively retrieved 174 Ki67 stained slides of surgically excised invasive breast cancers with ER+ and/ or PR+, HER2-, and Ki67 LI of 10-20% range. On each slide, 3 spots of low, intermediate, and high Ki67 labeling densities were captured via camera snapshot. And then, Ki67 LI was measured by digital image analysis (DIA) and manual count (MC) on 3 captured spots. Each case was classified into LA and LB types based on Ki67 LI measured by 4 methods; hot spot index by DIA; average index by DIA; hot spot index by MC; average index by MC. Ki67 heterogeneity was defined by the tumor with Ki67 KI &amp;lt;5% in low labeled spot, and ≥14% in high labeled spot by DIA. Correlation between the Ki67 heterogeneity and pathological parameters were analyzed. Result: Ki67 LI measured by DIA was higher than by MC in 3 spots (low: 4.95 vs. 3.39, p&amp;lt;0.001; intermediate: 10.07 vs.9.26, p=0.058; high: 14.88 vs. 14.03, p=0.025) with the biggest difference in low labeled spot. The 4 methods showed poor agreement on classification of LA/LB types. The proportion of LB type ranged from 3.4% to 60.9% depending on methods. Classification of luminal A and luminal B types by different methods Luminal ALuminal BHot spot index-DIA68 (39.1%)106 (60.9%)Average index-DIA157 (90.2%)17 (9.8%)Hot spot index-MC82 (47.1%)92 (52.9%)Average index-MC168 (96.6%)6 (3.4%) Forty two of 174 cases (24.1%) showed Ki67 heterogeneity. Ki67 heterogeneity was most frequent in T2 (p=0.001) and histologic grade 3 (p=0.023) tumors. Comparison of pathological features according to Ki67 heterogeneityKi67 heterogeneityAbsentPresentp-valueNumber 132(75.9%)42 (24.1%) T stageT195(84.1%)18(15.9%)0.001 T232(58.2%)23(41.8%) T35(83.3%)1(16.7%) Histologic grade130(90.9%)3(9.1%)0.023 293(74.4%)32(25.6%) 39(56.2%)7(43.8%) SubtypeInvasive ductal carcinoma, NOS114(74%)40(26%)0.475 Invasive lobular carcinoma8(88.9%)1(11.1%) Mucinous carcinoma4(80%)1(20%) Others (invasive micropapillary, tubular, etc.)6(100%)0(0%) Conclusion: Ki67 LI is markedly variable depending on different measurements and spot selections. Ki67 heterogeneity exists in 24.1% of luminal breast cancers in this series. Therefore, whole section slides should be preferred for Ki67 LI measurement, rather than tissue microarray or biopsy specimen which may not entirely reflect the proliferation index of the tumor. Citation Format: Kim S, Moon B-I, Sung SH. Variability of Ki67 labeling index depending on different measurements in luminal type breast cancers. [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 P1-01-03.

Abstract P4-09-24: CD163 expression is associated with young age, triple negative subtype, and poor outcome in breast cancer

Abstract BACKGROUND CD163 is a scavenger receptor specifically expressed by cells of monocyte/macrophage lineage. It is a biomarker in many clinical conditions, including coronary artery disease, anti-inflammatory response, and cancers. In breast cancer, high numbers of CD163-positive macrophages correlated with unfavorable outcome. Expression of CD163 in breast cancer was also related to early distant recurrence and poor survival. In this study we evaluated whether CD163 expression was associated with aggressive breast cancers from patients enrolled in the Clinical Breast Care Project (CBCP). METHODS Patients were enrolled into the CBCP following IRB-approved, HIPAA-compliant protocols. The study focused on 129 invasive breast cancer samples with CD163 immunohistochemistry (IHC) results. Expression of CD163, ER, PR, HER2, and Ki67 were assayed by IHC. CD163 was positive if IHC&amp;gt;0. ER and PR were positive if there was &amp;gt;5% nuclear staining. HER2 was negative if IHC=0 or 1+ and positive if IHC=3+. For HER2 IHC=2+, HER2 was negative if FISH was &amp;lt;1.8 and positive if FISH was &amp;gt;2.2. Ki67 was positive if there was ≥15% nuclear staining. For subtyping, Luminal A (LA) was ER+/HER2-/Ki67-, Luminal B1 (LB1) was ER+/HER2-/Ki67+, Luminal B2 (LB2) was ER+/HER2+, Her2+ was ER-/PR-/HER2+, and triple negative (TN) was ER-/PR-/HER2-. Associations of CD163 IHC score with age, race, and IHC subtype were examined by Wilcoxon rank-sum tests and/or Fisher's exact tests. Prognoses of CD163 in overall survival (OS), disease specific survival (DSS), disease free survival (DFS), and recurrence were studied using univariable and multivariable Cox proportional hazards regression models. CD163 score, age, race, AJCC stage, and subtype were included in the multivariable model. RESULTS CD163 IHC score displayed a significant negative correlation with age (R=-0.20, P=0.022). Patients with a CD163 score of 3+ were significantly younger than those with a score of 0 (P=0.019). CD163 score distributions were not statistically different between white and African American patients. CD163 scores of LA tumors were significantly lower than those of the tumors with all other subtypes except Her2+. Similarly, the CD163 scores of TN tumors were significantly higher than those of the tumors with all other subtypes but LB2. A higher CD163 score predicted worse DSS (HR=3.87 &amp; P=0.0020 in univariable model; HR=4.21 &amp; P=0.033 in multivariable model) and higher risk of recurrence (HR=2.85 &amp; P=0.00016 in univariable model; HR=2.81 &amp; P=0.012 in multivariable model). CONCLUSION Higher CD163 expression in breast cancer was significantly associated with younger age, the TN subtype, worse DSS, and higher risk of recurrence. These results highlight CD163 as a prognostic marker for breast cancer. The views expressed in this article are those of the author and do not reflect the official policy of the Department of Defense, or U.S. Government. Citation Format: Ru Y, Hu PT, Kovatich AJ, Hooke JA, Liu J, Kvecher L, Fantacone-Campbell JL, Deyarmin B, Kovatich AW, Cammarata F, Rui H, Davidson-Moncada J, Shriver CD, Hu H. CD163 expression is associated with young age, triple negative subtype, and poor outcome in breast cancer. [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 P4-09-24.

Constitutive expression of AhR and BRCA-1 promoter CpG hypermethylation as biomarkers of ERα-negative breast tumorigenesis.

BackgroundOnly 5–10 % of breast cancer cases is linked to germline mutations in the BRCA-1 gene and occurs early in life. Conversely, sporadic breast tumors, which represent 90-95 % of breast malignancies, have lower BRCA-1 expression, but not mutated BRCA-1 gene, and tend to occur later in life in combination with other genetic alterations and/or environmental exposures. The latter may include environmental and dietary factors that activate the aromatic hydrocarbon receptor (AhR). Therefore, understanding if changes in expression and/or activation of the AhR are associated with somatic inactivation of the BRCA-1 gene may provide clues for breast cancer therapy.MethodsWe evaluated Brca-1 CpG promoter methylation and expression in mammary tumors induced in Sprague–Dawley rats with the AhR agonist and mammary carcinogen 7,12-dimethyl-benzo(a)anthracene (DMBA). Also, we tested in human estrogen receptor (ER)α-negative sporadic UACC-3199 and ERα-positive MCF-7 breast cancer cells carrying respectively, hyper- and hypomethylated BRCA-1 gene, if the treatment with the AhR antagonist α-naphthoflavone (αNF) modulated BRCA-1 and ERα expression. Finally, we examined the association between expression of AhR and BRCA-1 promoter CpG methylation in human triple-negative (TNBC), luminal-A (LUM-A), LUM-B, and epidermal growth factor receptor-2 (HER-2)-positive breast tumor samples.ResultsMammary tumors induced with DMBA had reduced BRCA-1 and ERα expression; higher Brca-1 promoter CpG methylation; increased expression of Ahr and its downstream target Cyp1b1; and higher proliferation markers Ccnd1 (cyclin D1) and Cdk4. In human UACC-3199 cells, low BRCA-1 was paralleled by constitutive high AhR expression; the treatment with αNF rescued BRCA-1 and ERα, while enhancing preferential expression of CYP1A1 compared to CYP1B1. Conversely, in MCF-7 cells, αNF antagonized estradiol-dependent activation of BRCA-1 without effects on expression of ERα. TNBC exhibited increased basal AhR and BRCA-1 promoter CpG methylation compared to LUM-A, LUM-B, and HER-2-positive breast tumors.ConclusionsConstitutive AhR expression coupled to BRCA-1 promoter CpG hypermethylation may be predictive markers of ERα-negative breast tumor development. Regimens based on selected AhR modulators (SAhRMs) may be useful for therapy against ERα-negative tumors, and possibly, TNBC with increased AhR and hypermethylated BRCA-1 gene.