Predict Breast Cancer Recurrence Research Articles

66P - Prognostic and clinical values for prediction of breast cancer recurrence in HER2 positive early breast cancer after surgery in King Chulalongkorn Memorial Hospital 2005-2016

Background: Human epidermal growth factor receptor 2 (HER-2) overexpressed breast cancer occurs in 15-20% and associated with diminished disease-free and overall survival. After surgery, the standard treatment in HER-2 positive early breast cancer is systemic chemotherapy combined with trastuzumab. However, one fifth of these patients had disease recurrence in five years after treatment. So, the study was aimed to identify prognostic and clinical values that associated with survival outcomes especially in term of recurrence free survival. Methods: One hundred and sixty-nine patients with early stage HER-2 positive breast cancer who were diagnosed and treated at King Chulalongkorn Memorial Hospital from January 2005 to December 2016 were reviewed retrospectively. Clinico-pathologic features and survival outcomes were analyzed. Results: In initial analysis, median age was 50 years (26-83), 49.1% had premenopausal status, 93.4% had invasive ductal carcinoma, 52.6% had T2 tumor, 58.6% had node negative, 46.1% had lymphovascular invasion, 79.2% had free margin and 46.7% had positive estrogen and progesterone receptor. For the treatment, 69.8% underwent modified radical mastectomy, 96.4% received neoadjuvant/adjuvant chemotherapy, 55.6% received adjuvant trastuzumab, 48.5% received adjuvant hormonal therapy and 68% received post-operative radiotherapy. 35 out of 169 (20.7%) patients had recurrence disease. Distant metastasis was identified in 62.8%. From univariate and multivariate analysis, trastuzumab used (HR 0.44 [95%CI 0.21-0.9]; p = 0.007 from multivariate) and surgical margin (HR 1.02 [95%CI 1.0-1.03]; p = 0.012 from multivariate) were associated with increase risk of recurrences. After median follow up 4.9 years, the 5-years recurrence free survival was 81.2%, 89.2% and 70.7% in overall populations, trastuzumab used subgroup and non-trastuzumab used subgroup, respectively. Overall survival will be further analyzed. Conclusions: Trastuzumab used and surgical margin status were statistically significant prognostic values for predicting disease recurrence in HER-2 positive early breast cancer. Legal entity responsible for the study: Division of Medical Oncology, Department of Medicine, Faculty of Medicine, Chulalongkorn University and the King Chulalongkorn Memorial Hospital. Funding: None Disclosure: All authors have declared no conflicts of interest.

Designing Extreme Learning Machine Network Structure Based on Tolerance Rough Set

In this paper, we present a new extreme learning machine network structure on the basis of tolerance rough set. The purpose of this paper is to realize the high-efficiency and multi-dimensional ELM network structure. Various published algorithms have been applied to breast cancer datasets, but rough set is a fairly new intelligent technique that applies to predict breast cancer recurrence. We analyze Ljubljana Breast Cancer Dataset, firstly, obtain lower and upper approximations and calculate the accuracy and quality of the classification. The high values of the quality of classification and accuracy prove that the attributes selected can well approximate the classification. Rough sets approach is established to solve the prolem of tolerance.

Prognostic value of routine laboratory variables in prediction of breast cancer recurrence

The prognostic value of routine laboratory variables in breast cancer has been largely overlooked. Based on laboratory tests commonly performed in clinical practice, we aimed to develop a new model to predict disease free survival (DFS) after surgical removal of primary breast cancer. In a cohort of 1,596 breast cancer patients, we analyzed the associations of 33 laboratory variables with patient DFS. Based on 3 significant laboratory variables (hemoglobin, alkaline phosphatase, and international normalized ratio), together with important demographic and clinical variables, we developed a prognostic model, achieving the area under the curve of 0.79. We categorized patients into 3 risk groups according to the prognostic index developed from the final model. Compared with the patients in the low-risk group, those in the medium- and high-risk group had a significantly increased risk of recurrence with a hazard ratio (HR) of 1.75 (95% confidence interval [CI] 1.30–2.38) and 4.66 (95% CI 3.54–6.14), respectively. The results from the training set were validated in the testing set. Overall, our prognostic model incorporating readily available routine laboratory tests is powerful in identifying breast cancer patients who are at high risk of recurrence. Further study is warranted to validate its clinical application.

Differential expression of exosomal miRNAs between breast cancer patients with and without recurrence.

BackgroundExosomal microRNAs (miRNAs) are promising candidate biomarkers for diagnosis or prognosis for breast cancer. We investigated the prognostic role of exosomal miRNAs in serum samples derived from patients with breast cancer and compared miRNA expression between serum and tumor tissues.MethodsThe miRNA profile derived from exosome between breast cancer patients with recurrence (n = 16) and without recurrence (n = 16) were compared by miRNA PCR array. Further, we examined the expression of miRNAs derived from tissues in the patients with breast cancer with (n = 35) and without recurrence (n = 39) by qRT-PCR.ResultsOf 384 miRNAs, three miRNAs (miR-338-3p, miR-340-5p, and miR-124-3p) were significantly upregulated and eight (miR-29b-3p, miR-20b-5p, miR-17-5p, miR-130a-3p, miR-18a-5p, miR-195-5p, miR-486-5p, and miR-93-5p) were significantly downregulated in the patients with recurrence. We evaluated the expression of the miRNAs in tumor tissues. The patients with recurrence had higher levels of miR-340 at their primary site as well as in the serum. In contrast, miR-195-5p, miR-17-5p, miR-93-5p, and miR-130a-3p, derived from tumor tissues that were downregulated in the serum from patients with recurrence, were higher in the patients with recurrence than in those with no recurrence. In logistic regression analysis, miR-340-5p, miR-17-5p, miR-130a-3p, and miR-93-5p were significantly associated with recurrence.ConclusionsSeveral exosomal miRNAs may be useful biomarkers to predict breast cancer recurrence. We show the different expression patterns of miRNAs between tumor tissues and serum. These findings may suggest selective mechanism of release of exosomal miRNAs by cancer cells to regulate their progression.

Abstract 669: Quantification of estrogen (ER) and progestin receptor (PR) as well as HER2/neu proteins and gene expression improves discrimination of clinical behavior by triple positive breast carcinomas

Abstract IHC analyses of ER, PR and HER2 proteins are used as clinical biomarkers for breast cancer management. However, IHC provides semi-quantitative results sometimes complicated by variation in methods and interpretation. Our goal was to ascertain clinical use of these three analytes to predict breast cancer recurrence when analyses for gene expression and protein products were quantified. Procedures: We examined ER and PR protein in 1059 de-identified carcinoma biopsies previously quantified by radio-ligand binding and/or enzyme immunoassay (EIA) using FDA approved reagents and HER2 oncoprotein determined by EIA to assess relationships between biomarker profiles and disease-free (DFS) and overall survival (OS) for 123 patients. Our comprehensive, IRB-approved Database also contained de-identified Microarray results obtained for ~ 22,000 genes from LCM-procured breast carcinoma cells of 247 primary breast biopsies. ESR1, PGR and ERBB2 gene expression levels were validated previously in 278 cancers by qPCR. Results: Generalized Pair Plots of the three protein biomarkers indicated a relationship only between ER and PR supporting evidence that ER complexed with estrogenic ligands promotes synthesis of PR. Of the 8 possible combinations of the 3 protein biomarkers, only breast cancers exhibiting ER+PR+HER2+ (triple positive breast cancer, TPBCa) exhibited increased OS, whereas those with triple negative breast cancers (TNBCa) had decreased OS. TPBCa was exhibited by 32.5 % of specimens compared to TNBCa, which represented 7.3% of biopsies when biomarker proteins were quantified. Number of TPBCa was less when expression of the 3 genes (ESR1+/PGR+/ERBB2+) was measured by microarray or by qPCR while those of TNBCa increased. Furthermore, when TPBCa were identified by either microarray or by qPCR measurements, collectively increased ESR1, PGR and ERBB2 expression was associated with increased PFS and OS of patients. In contrast, biopsies that were ESR1-/PGR-/ERBB+ (14.1%) correlated with poor prognosis and overall survival using qPCR data, while those exhibiting ESR1-PGR+ERBB2- (14.1%) profiles by microarray results exhibited decreased PFS and OS. Influence of patient menopausal status on biomarker profiles was examined. Conclusions: Quantitative measurements of ER, PR and HER2 proteins as well as microarray and qPCR assessment of ESR1, PGR and ERBB2 gene expression were correlated with prediction of clinical outcomes of breast carcinomas exhibiting Triple Positive Breast Cancers (TPBCa). Results reinforce importance of assessing levels of the three biomarkers in a quantified fashion to enhance their use in breast cancer management and prediction of risk of recurrence. Citation Format: Zohair R. Hameed, Michael W. Daniels, D. Alan Kerr, James L. Wittliff. Quantification of estrogen (ER) and progestin receptor (PR) as well as HER2/neu proteins and gene expression improves discrimination of clinical behavior by triple positive breast carcinomas [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 669. doi:10.1158/1538-7445.AM2017-669

Validity of Danish Breast Cancer Group (DBCG) registry data used in the predictors of breast cancer recurrence (ProBeCaRe) premenopausal breast cancer cohort study

Background: Validation studies of the Danish Breast Cancer Group (DBCG) registry show good agreement with medical records for adjuvant treatment data, but inconsistent recurrence information. No studies have validated changes in menopausal status or endocrine therapy during follow-up. In a longitudinal study, we validated DBCG data using medical records as the gold standard.Material and methods: From a cohort of 5959 premenopausal women diagnosed during 2002–2010 with stage I–III breast cancer, we selected 151 patients – 77 estrogen-receptor-positive and 74 estrogen-receptor-negative – from three hospitals. We assessed the validity of DBCG registry data on patient, tumor, and treatment factors, and follow-up information on menopausal transition, changes in endocrine therapy, and recurrence. We computed positive predictive values (PPVs) with 95% confidence intervals (95%CI).Results: Agreement was near perfect for tumor size, lymph node involvement, receptor status, surgery type, and receipt of radiotherapy, chemotherapy, or tamoxifen treatment. The PPV for a change in endocrine therapy in the DBCG was 96% (95%CI = 83, 100). The PPV for menopausal transition was 61% (95%CI = 42, 77). The PPV for DBCG-recorded recurrence was 100%. However, of 19 patients who had a recurrence documented in their medical record, 13 had the recurrence registered in DBCG.Conclusions: DBCG data are valid for most epidemiological studies of breast cancer treatment. Data on menopausal transition may be less valid, though this interpretation depends on the suitability of medical records for making this assessment. Although recurrence is missing for some, this would not bias most ratio measures of association.

Cell-free circulating DNA integrity is an independent predictor of impending breast cancer recurrence.

Non-invasive blood-based molecule markers are evaluated as promising biomarkers these days. Here we investigated the potential of cell-free circulating DNA Integrity (cfDI) as blood-based marker for the prediction of recurrence during the follow-up of breast cancer patients within a prospective study cohort. cfDI was determined in plasma of 212 individuals, by measuring ALU and LINE1 repetitive DNA elements using quantitative PCR. A significant decrease of cfDI in recurrent breast cancer patients was observed. The group of patients who had impending recurrence during the follow-up had significant lower cfDI compared to the group of non-recurrent patients (P < 0.001 for ALU and LINE1 cfDI). cfDI could differentiate recurrent breast cancer patients from non-recurrent breast cancer subjects (area under the curve, AUC = 0.710 for ALU and 0.704 for LINE1). Univariate and multivariate analysis confirmed a significant association of recurrence and cfDI. Breast cancer patients with a lower cfDI had a much higher risk to develop recurrence than the patients with a higher cfDI (P = 0.020 for ALU cfDI and P = 0.019 for LINE1 cfDI, respectively). Further we show that cfDI is an independent predictor of breast cancer recurrence. In combination with other molecular markers, cfDI might be a useful biomarker for the prediction for breast cancer recurrence in clinic utility. We propose that cfDI might also be useful for the prediction of recurrence during the follow-up of other cancers.

Apolipoprotein D expression does not predict breast cancer recurrence among tamoxifen-treated patients.

BackgroundApolipoprotein D (ApoD) has been proposed as a predictor of breast cancer recurrence among estrogen receptor-positive (ER+), tamoxifen-treated patients.MethodsWe conducted a population-based case-control study nested in a population of 11,251 women aged 35–69 years at diagnosis with Stage I–III breast cancer between 1985 and 2001 on Denmark’s Jutland Peninsula and registered with the Danish Breast Cancer Cooperative Group. We identified 541 recurrent or contralateral breast cancers cases among women with ER+ disease treated with tamoxifen for at least 1 year and 300 cases in women with ER– disease never treated with tamoxifen. We matched one control subject per case and assessed ApoD expression in the tumor cell nucleus and cytoplasm using tissue microarray immunohistochemistry. We computed the odds ratio (OR) associating ApoD expression with recurrence and adjusted for potential confounding using logistic regression.ResultsCytoplasmic ApoD expression was seen in 68% of ER+ tumors, in 66% of ER– tumors, and in 66% of controls across both groups. In women with ER+ tumors, the associations of cytoplasmic ApoD expression with recurrence (OR = 1.0; 95% CI = 0.7 to 1.4) and increasing cytoplasmic expression with recurrence (OR = 1.0; 95% CI = 0.996 to 1.003) were null, as were those for women with ER– tumors. Associations for nuclear ApoD expression and combined nuclear and cytoplasmic expression were similarly near-null.ConclusionApoD expression is likely not a predictor of recurrence in tamoxifen-treated patients.ImpactThis study eliminates the previously suggested marker ApoD as a predictor of recurrence among tamoxifen-treated women.

Abstract S6-05: Comprehensive comparison of prognostic signatures for breast cancer in TransATAC

Abstract Background: A number of prognostic signatures have been developed for the prediction of breast cancer recurrence in the past decade. We have developed two signatures (Clinical Treatment Score (CTS), four immunohistochemical markers (IHC4)) and validated four prognostic signatures (Oncotype Dx Recurrence Score (RS), PAM50-based Prosigna (ROR), Breast Cancer Index (BCI), and EndoPredict (EPclin)) in the TransATAC cohort. Here, we compare the prognostic performance of these six signatures for distant recurrence (DR) in years 0-10, and specifically in years 5-10 after treatment cessation. Methods: 1231 postmenopausal women with hormone receptor positive and HER2-negative breast cancer had at least one test performed. Of these, 818 women had data on all six signatures available. IHC4, RS and BCI (linear) are molecular only signatures whereas CTS, ROR and EPclin include clinicopathological factors. The primary endpoint was DR and the primary objective was to compare the prognostic value of the six signatures in terms of DR for years 0-10, 0-5, and 5-10. Secondary objectives included the comparison of the prognostic performance for node-negative and node-positive patients separately and the additional prognostic performance of each signature to the others. Likelihood ratio statistics (LR-χ2) were used to assess the prognostic information of each signature alone or in combination with other signatures. Results: Median follow-up for this analysis was 9.94 years (IQR 8.01-10.09) and a total of 126 DR were recorded. 818 women with HER2-negative disease for whom data of all six signatures were available were included in this analysis. For all patients, CTS and EPclin were the most prognostic signatures in years 0-10 (CTS: LR-χ2=124.9; EPclin: LR-χ2=116.2) and years 5-10 (CTS: LR-χ2=59.6; EPclin: LR-χ2=56.8) in the univariate analysis. The other four signatures performed similarly well in years 0-5, but of those only BCI and ROR provided substantial prognostic information in years 5-10 (BCI: LR-χ2=25.3; ROR: LR-χ2=43.8). In multivariate analyses comparing the added information of the molecular signatures over CTS, IHC4 and BCI provided the most information (IHC4: ΔLR-χ2=19.0; BCI: ΔLR-χ2=19.8). In node-negative patients (72.3%), the ROR showed the most prognostic value in years 0-10 (LR-χ2=48.6) and years 5-10 (LR-χ2=31.3) whereas the RS was least prognostic in this patient group. For patients with node-positive disease (27.7%), the CTS and EPclin were the most prognostic and the other four signatures provided much less prognostic information for this patient population (data not shown). Conclusion: Overall, the CTS and EPclin were the most prognostic signatures for DR and also added significant prognostic value to the other scores in women with HER2-negative disease, primarily due to the incorporation of nodal status in these signatures. For women with node-negative disease, the ROR, BCI, and EPclin signatures provided most prognostic value whereas for those with positive nodes CTS and EPclin were most prognostic. Our analyses showed that the inclusion of clinic-pathological factors into gene signatures is highly important for deriving an accurate prognostic assessment, particularly in node-positive patients. Citation Format: Sestak I, Buus R, Cuzick J, Dubsky P, Kronenwett R, Ferree S, Sgroi D, Schnabel C, Baehner R, Mallon E, Dowsett M. Comprehensive comparison of prognostic signatures for breast cancer in TransATAC [abstract]. In: Proceedings of the 2016 San Antonio Breast Cancer Symposium; 2016 Dec 6-10; San Antonio, TX. Philadelphia (PA): AACR; Cancer Res 2017;77(4 Suppl):Abstract nr S6-05.

Abstract P6-09-45: Long-term follow-up of early stage breast cancer patients with results of MammaPrint®, Oncotype DX® and MammoStrat® risk classification assays

Abstract Introduction: The use of genomic tests for the prediction of breast cancer recurrence is becoming more common. MammaPrint® (MP, Agendia Inc.) is a 70-gene microarray assay designed to assess the 10-year risk of recurrence in an untreated population that was not selected for ER/HER2 results. The Oncotype DX® Recurrence Score® (RS, Genomic Health, Inc.) is a 21-gene RT-PCR assay that is clinically validated to predict the 10-year risk of distant recurrence in ER+ patients treated with Tamoxifen. MammoStrat® (MS, Clarient, Inc.) is an IHC assay that uses 5 antibodies and has been validated in a similar population as RS. Several recent reports show that these assays classify patients differently with significant discordances for all risk groups (Shivers, et al., SABCS 2013; Denduluri, et al., ASCO Breast 2011; Poulet, et al., SABCS 2012; Schneider, et al., ASCO 2013). The present study is an analysis of long-term follow-up in a cohort of patients who have results for all three of these risk-stratifying assays side by side in the same samples. Methods: Patients with ER+ N0-N1 early-stage breast cancer with an MP result obtained as part of their routine clinical care were identified at the University of South Florida (USF, N=65) and Morton Plant Hospital (N=83). After local IRB approval, slides and/or blocks were cut and de-identified at USF and sent to Genomic Health and Clarient for blinded testing. Clinicopathological features were also reviewed by 3 breast pathologists. Results: 148 patients with an MP result had tissue available to send for RS and MS assays. These patients had a median age of 62 years; median tumor size 1.8 cm; 9% low grade, 59% intermediate grade and 32% high grade. In our previous analysis of this study, of 148 patients with MP results, 53% were low risk and 47% were high risk. Of 135 samples that yielded enough RNA to produce an RS result, 53% were low risk, 26% were intermediate risk and 21% were high risk. Of 129 samples that yielded an MS result, 44% were low risk, 28% were moderate risk and 28% were high risk. Of 121 patients with results for all 3 assays, only 22% were concordant for low risk and 9% were concordant for high risk across all 3 assays. Overall, 30% of cases showed a major discordance such as low risk for one assay and high risk for another. After median follow-up of 54 months, 9 patients have had a distant metastasis and/or 8 patients have died (11 patients total). One patient who had bone metastasis and died had been classified as low risk by all 3 assays. Three patients with distant metastases had a major discordance between assays, with two high risk and one low risk result. Seven patients were classified as high or intermediate/moderate risk by all 3 assays. Conclusions: This direct comparison demonstrates that although the assays classify a large proportion of patients differently, the patients who ended up with a distant metastasis and/or died of breast cancer had been classified as high risk by at least two of the three assays. This study has important clinical implications since these assays are used to help make treatment decisions regarding which patients might benefit from chemotherapy. Citation Format: Shivers SC, Russell S, Blumencrancz L, Mehindru A, Acs G, Ellis D, Vrcelj V, Zanchi A, Blumencrancz PW, Carter E, King J, Cox CE. Long-term follow-up of early stage breast cancer patients with results of MammaPrint®, Oncotype DX® and MammoStrat® risk classification assays [abstract]. In: Proceedings of the 2016 San Antonio Breast Cancer Symposium; 2016 Dec 6-10; San Antonio, TX. Philadelphia (PA): AACR; Cancer Res 2017;77(4 Suppl):Abstract nr P6-09-45.

Abstract P3-05-11: Clinical outcomes of estrogen-receptor negative breast carcinoma are associated with protein hormone-cognate receptor gene expression

Abstract Background: Since certain protein hormones appear to be associated with progression of breast cancer cells in vitro, we employed a global approach by evaluating relationships between genes for these hormones and their cognate receptors in human breast cancer biopsies as independent predictors of risk of recurrence. Our goal is to derive clinically relevant molecular signatures correlating various ER/PR subtypes. Methods: Microarray analyses of LCM-procured carcinoma cells from 247 de-identified biopsies determined expression of genes for 61 protein hormones (the ligands) and 81 cognate receptors. Total RNA was extracted, purified and amplified from cells to determine expression of 22,000 genes. Univariable and multivariable Cox regressions were determined using expression levels of each hormone/receptor gene. Kaplan-Meier plots were constructed with an adjusted p-value &amp;lt;0.30 (discovery cutoff) for each gene candidates with/without ER status of the lesion. Cox regression analyses with/without interaction models deciphered candidates of hormone-receptor complexes that predicted risk of recurrence as well as overall survival using clinical follow-up that extended as much as 12 years. LASSO was used to derive gene subsets to predict overall survival (OS) and progression free survival (PFS) within ER/PR subtypes. Results: Expression levels of 7 genes for protein hormones and 10 receptor genes were significant for OS at an adjusted p-value of &amp;lt;0.30 while expression of 15 hormone and 19 receptor genes were significant for PFS without regard to ER/PR subtypes. When expression levels were considered with ER status only, 3 genes for ER+ lesions and 6 genes for ER- lesions were identified that predicted OS/PFS. Categorization of cancers according to combined ER and PR status revealed one gene for ER+/PR+, 3 genes for ER+/PR- and 2 genes for ER-/PR- associated with PFS/OS. After stratifying gene expression into above/below median for the linear predictor, Kaplan-Meier plots revealed patient groups with each signature for ER+/PR+, ER-/PR- and ER- lesions that exhibited significantly different survival. Molecular signatures comprising 16 genes for ER+/PR+ carcinomas, 10 genes for ER-/PR- lesions and 6 genes for ER- carcinomas predicted clinical outcomes with C-indices of 0.79, 0.76 and 0.71, respectively. Conclusion: We revealed many breast carcinomas synthesize mRNA species for various protein hormones and their cognate receptors by determining gene expression directly on pure populations of these cells procured by LCM. We report a novel ten-gene ER-/PR- signature containing four genes in common with that of a six-gene ER- only signature that predicts breast cancer recurrence. Collectively, results of mRNA expression suggest that often breast carcinomas exhibit substantial elements of endocrine autonomy for regulating progression, warranting investigation of protein products of gene candidates in isolated populations of breast carcinoma cells. Citation Format: Daniels MW, Wittliff JL, Brock GN. Clinical outcomes of estrogen-receptor negative breast carcinoma are associated with protein hormone-cognate receptor gene expression [abstract]. In: Proceedings of the 2016 San Antonio Breast Cancer Symposium; 2016 Dec 6-10; San Antonio, TX. Philadelphia (PA): AACR; Cancer Res 2017;77(4 Suppl):Abstract nr P3-05-11.

Interaction of MRE11 and Clinicopathologic Characteristics in Recurrence of Breast Cancer: Individual and Cumulated Receiver Operating Characteristic Analyses

The interaction between the meiotic recombination 11 homolog A (MRE11) oncoprotein and breast cancer recurrence status remains unclear. The aim of this study was to assess the interaction between MRE11 and clinicopathologic variables in breast cancer. A dataset for 254 subjects with breast cancer (220 nonrecurrent and 34 recurrent) was used in individual and cumulated receiver operating characteristic (ROC) analyses of MRE11 and 12 clinicopathologic variables for predicting breast cancer recurrence. In individual ROC analysis, the area under curve (AUC) for each predictor of breast cancer recurrence was smaller than 0.7. In cumulated ROC analysis, however, the AUC value for each predictor improved. Ten relevant variables in breast cancer recurrence were used to find the optimal prognostic indicators. The presence of any six of the following ten variables had a high (79%) sensitivity and a high (70%) specificity for predicting breast cancer recurrence: tumor size ≥ 2.4 cm, tumor stage II/III, therapy other than hormone therapy, age ≥ 52 years, MRE11 positive cells > 50%, body mass index ≥ 24, lymph node metastasis, positivity for progesterone receptor, positivity for epidermal growth factor receptor, and negativity for estrogen receptor. In conclusion, this study revealed that these 10 clinicopathologic variables are the minimum discriminators needed for optimal discriminant effectiveness in predicting breast cancer recurrence.

A Hybrid Computer-aided-diagnosis System for Prediction of Breast Cancer Recurrence (HPBCR) Using Optimized Ensemble Learning

Cancer is a collection of diseases that involves growing abnormal cells with the potential to invade or spread to the body. Breast cancer is the second leading cause of cancer death among women. A method for 5-year breast cancer recurrence prediction is presented in this manuscript. Clinicopathologic characteristics of 579 breast cancer patients (recurrence prevalence of 19.3%) were analyzed and discriminative features were selected using statistical feature selection methods. They were further refined by Particle Swarm Optimization (PSO) as the inputs of the classification system with ensemble learning (Bagged Decision Tree: BDT). The proper combination of selected categorical features and also the weight (importance) of the selected interval-measurement-scale features were identified by the PSO algorithm. The performance of HPBCR (hybrid predictor of breast cancer recurrence) was assessed using the holdout and 4-fold cross-validation. Three other classifiers namely as supported vector machines, DT, and multilayer perceptron neural network were used for comparison. The selected features were diagnosis age, tumor size, lymph node involvement ratio, number of involved axillary lymph nodes, progesterone receptor expression, having hormone therapy and type of surgery. The minimum sensitivity, specificity, precision and accuracy of HPBCR were 77%, 93%, 95% and 85%, respectively in the entire cross-validation folds and the hold-out test fold. HPBCR outperformed the other tested classifiers. It showed excellent agreement with the gold standard (i.e. the oncologist opinion after blood tumor marker and imaging tests, and tissue biopsy). This algorithm is thus a promising online tool for the prediction of breast cancer recurrence.

Predicting Breast Cancer Recurrence Using Machine Learning Techniques

Background: Recurrence is an important cornerstone in breast cancer behavior, intrinsically related to mortality. In spite of its relevance, it is rarely recorded in the majority of breast cancer datasets, which makes research in its prediction more difficult. Objectives: To evaluate the performance of machine learning techniques applied to the prediction of breast cancer recurrence. Material and Methods: Revision of published works that used machine learning techniques in local and open source databases between 1997 and 2014. Results: The revision showed that it is difficult to obtain a representative dataset for breast cancer recurrence and there is no consensus on the best set of predictors for this disease. High accuracy results are often achieved, yet compromising sensitivity. The missing data and class imbalance problems are rarely addressed and most often the chosen performance metrics are inappropriate for the context. Discussion and Conclusions: Although different techniques have been used, prediction of breast cancer recurrence is still an open problem. The combination of different machine learning techniques, along with the definition of standard predictors for breast cancer recurrence seem to be the main future directions to obtain better results.

150PD - Breast cancer-specific survival in >4,600 patients with lymph node-positive (LN+) hormone receptor-positive (HR+) invasive breast cancer (BC) and 21-gene recurrence score® (RS) results in the SEER registries

The 21-gene RS assay has been shown to predict BC recurrence and adjuvant chemotherapy benefit in LN +, HR +, HER2-negative BC. We assessed 5-year BC-specific survival (BCSS) in LN+ patients with 21-gene RS results in the SEER registries, a cancer surveillance program that covers 30% of the US population. All SEER BC cases diagnosed 2004-2012 were linked to 21-gene assays performed by the Genomic Health Clinical Laboratory. The analysis was restricted to single primary invasive BC, LN +, no distant metastases, HR+ (per SEER), HER2-negative (per RT-PCR). Using the actuarial method in SEER*Stat, BCSS was assessed for those who were diagnosed 2004-2011 with survival follow-up through 2012, by RS category and by number of positive nodes. The proportion of women with LN +, HR+ BC who had RS results (n = 7695) increased over time between 2004 (0.3%, n = 33) and 2012 (13.8%, n = 2010), and was lower with increasing nodal involvement, from micrometastases only (36.0%, n = 792) to 4+ nodes (2.6%, n = 81) in 2012. BCSS differed significantly by both RS category (log-rank p < 0.001) and number of positive nodes (p < 0.001). Five-year BCSS outcomes for those with RS <18 ranged from 99.4% (95% CI, 97.4%-99.9%) in those with micrometastases to 85.7% (95% CI, 33.4%-97.9%) in those with 4+ nodes. Similar patterns were found for those with RS 18-30 and RS ≥31 (Table).Tabled 1RS <18RS 18-30RS ≥31# of positive LNN5-y BCSS (95% CI)N5-y BCSS (95% CI)N5-y BCSS (95% CI)Micromets only120799.4 (97.4-99.9)73299.3 (98.2-99.7)13584.2 (69.2-92.2)1108499.2 (97.2-99.8)67596.3 (91.8-98.4)13591.3 (79.5-96.4)230097.9 (92.7-99.4)19296.0 (84.6-99.0)3673.2 (20.4-94.1)310394.7 (77.4-98.9)7094.0 (75.9-98.6)2285.4 (51.6-96.3)4+8285.7 (33.4-97.9)9379.5 (58.0-90.8)3175.1 (39.2-91.6) Open table in a new tab Overall, 5-year BCSS is excellent in patients with RS <18 and few positive nodes and worsens with increasing number of involved lymph nodes and with higher RS. Updated data with longer follow-up is will be presented.

Big Data Analytics to Predict Breast Cancer Recurrence on SEER Dataset using MapReduce Approach

traditional data analytic might not have the capacity to handle enormous amount of data. Due to the rapid growth of information, solutions need to be contemplated and provided in order to handle and extract value and knowledge from these data sets. Moreover, decision makers should have the capacity to increase significant bits of knowledge from such fluctuated and quickly evolving information. Such esteem can be given utilizing big data analytic, which is the utilization of advanced analytic techniques on big data using MapReduce approach. This paper examines to develop a high performance platform to efficiently analyse big SEER (Surveillance, Epidemiology, and End Results) breast cancer data set using MapReduce to find the recurrence of breast cancer. Keywordscancer; Big data, Classification; Data analytics, MapReduce, SEER.

Parallel Computing to Predict Breast Cancer Recurrence on SEER Dataset using Map-Reduce Approach

to the late overpowering development rate of large scale data, the advancement of handling faster processing algorithms with optimal execution has turned into a critical need of the time. In this paper, parallel Map-Reduce algorithm is proposed, that encourages concurrent participation of various computing hubs to develop a classifier on SEER breast cancer data set. Our algorithm can prompt supported models whose speculation execution is near the respective baseline classifier. By exploiting their own parallel architecture the algorithm increases noteworthy speedup. In addition, the algorithm don't require singular processing hubs to communicate with each other, to share their data or to share the knowledge got from their data and consequently, they are powerful in safeguarding privacy of computation also. This paper utilized the Map-Reduce framework to implement the algorithms and experimented onSEER breast cancer data sets to exhibit the execution as far as classification accuracy and speedup. Keywordscancer; Big dataanalytics, Classification; Parallel Computing; MapReduce, SEER.

Identification of a circulating microRNA signature to distinguish recurrence in breast cancer patients.

There is an urgent need for novel noninvasive prognostic biomarkers for monitoring the recurrence of breast cancer. The purpose of this study is to identify circulating microRNAs that can predict breast cancer recurrence. We conducted a microRNA profiling experiment in serum samples from 48 breast cancer patients using Exiqon miRCURY microRNA RT-PCR panels. Significantly differentiated miRNAs for recurrence in the discovery profiling were further validated in an independent set of sera from 20 patients with breast cancer recurrences and 22 patients without recurrences. We identified seven miRNAs that were differentially expressed between breast cancer patients with and without recurrences, including four miRNAs upregulated (miR-21-5p, miR-375, miR-205-5p, and miR-194-5p) and three miRNAs downregulated (miR-382-5p, miR-376c-3p, and miR-411-5p) for recurrent patients. Using penalized logistic regression, we built a 7-miRNA signature for breast cancer recurrence, which had an excellent discriminating capacity (concordance index=0.914). This signature was significantly associated with recurrence after adjusting for known prognostic factors, and it was applicable to both hormone-receptor positive (concordance index=0.890) and triple-negative breast cancers (concordance index=0.942). We also found the 7-miRNA signature were reliably measured across different runs of PCR experiments (intra-class correlation coefficient=0.780) and the signature was significantly higher in breast cancer patients with recurrence than healthy controls (p=1.1×10−5). In conclusion, circulating miRNAs are promising biomarkers and the signature may be developed into a minimally invasive multi-marker blood test for continuously monitoring the recurrence of breast cancer. It should be further validated for different subtypes of breast cancers in longitudinal studies.

PCBP1/HNRNP E1 Protects Chromosomal Integrity by Translational Regulation of CDC27.

CDC27 is a core component of the anaphase-promoting complex/cyclosome (APC/C), a multisubunit E3 ubiquitin ligase, whose oscillatory activity is responsible for the metaphase-to-anaphase transition and mitotic exit. Here, in normal murine mammary gland epithelial cells (NMuMG), CDC27 expression is controlled posttranscriptionally through the RNA binding protein poly(rC) binding protein 1 (PCBP1)/heterogeneous nuclear ribonucleoprotein E1 (HNRNP E1). shRNA-mediated knockdown of HNRNP E1 abrogates translational silencing of the Cdc27 transcript, resulting in constitutive expression of CDC27. Dysregulated expression of CDC27 leads to premature activation of the G2-M-APC/C-CDC20 complex, resulting in the aberrant degradation of FZR1/CDH1, a cofactor of the G1 and late G2-M-APC/C and a substrate normally reserved for the SCF-βTRCP ligase. Loss of CDH1 expression and of APC/C-CDH1 activity, upon constitutive expression of CDC27, results in mitotic aberrations and aneuploidy in NMuMG cells. Furthermore, tissue microarray of breast cancer patient tumor samples reveals high CDC27 levels compared with nonneoplastic breast tissue and a significant correlation between disease recurrence and CDC27 expression. These results suggest that dysregulation of HNRNP E1-mediated translational regulation of Cdc27 leads to chromosomal instability and aneuploidy and that CDC27 expression represents a significant predictor of breast cancer recurrence. The RNA-binding protein HNRNP E1 mediates translational regulation of the cell-cycle regulator CDC27 and that dysregulation of CDC27 leads to aneuploidy. In addition, high CDC27 expression in breast cancer patient tumor specimens significantly predicts disease recurrence, suggesting a novel role for CDC27 as a predictor of relapse. Mol Cancer Res; 14(7); 634-46. ©2016 AACR.

Breast cancer specific survival in patients with node-positive hormone receptor positive invasive breast cancer and Oncotype DX recurrence score results in the SEER database.

6575Background: The 21-tumor gene expression assay has been shown to predict breast cancer recurrence and adjuvant chemotherapy benefit in lymph node positive (LN+), hormone receptor positive (HR+)...