Predict Breast Cancer Recurrence Research Articles

Development and Validation of a New Multiparametric Random Survival Forest Predictive Model for Breast Cancer Recurrence with a Potential Benefit to Individual Outcomes.

PurposeBreast cancer (BC) is a multi-factorial disease. Its individual prognosis varies; thus, individualized patient profiling is instrumental to improving BC management and individual outcomes. An economical, multiparametric, and practical model to predict BC recurrence is needed.Patients and MethodsWe retrospectively investigated the clinical data of BC patients treated at the Third Affiliated Hospital of Sun Yat-sen University and Liuzhou Women and Children’s Medical Center from January 2013 to December 2020. Random forest-recursive feature elimination (run by R caret package) was used to determine the best variable set, and the random survival forest method was used to develop a predictive model for BC recurrence.ResultsThe training and validations sets included 623 and 151 patients, respectively. We selected 14 variables, the pathological (TNM) stage, gamma-glutamyl transpeptidase, total cholesterol, Ki-67, lymphocyte count, low-density lipoprotein, age, apolipoprotein B, high-density lipoprotein, globulin, neutrophil count to lymphocyte count ratio, alanine aminotransferase, triglyceride, and albumin to globulin ratio, using random survival forest (RSF)-recursive feature elimination. We developed a recurrence prediction model using RSF. Using area under the receiver operating characteristic curve and Kaplan–Meier survival analyses, the model performance was determined to be accurate. C-indexes were 0.997 and 0.936 for the training and validation sets, respectively.ConclusionThe model could accurately predict BC recurrence. It aids clinicians in identifying high-risk patients and making treatment decisions for Breast cancer patients in China. This new multiparametric RSF model is instrumental for breast cancer recurrence prediction and potentially improves individual outcomes.

Abstract P1-02-03: Tumor-infiltrating lymphocytes but not HER2 copy number or ratio show prognostic value in trastuzumab-treated HER2-positive breast cancer

Abstract Background: Around 15% of all breast cancers are HER2+, a subtype that prior to HER2-targeted therapy was associated with a poor prognosis. With advances in treatment the survival has dramatically improved, and previous studies suggest that the level of HER2 copies are predictive of the effect of anti-HER2 therapies. However, some patients progress despite HER2-targeted therapy and routine HER2 testing is hampered by inaccuracy to discriminate patients with therapy-resistant disease. Furthermore, tumor-infiltrating lymphocytes (TILs) in the tumor stroma shows prognostic value not only in triple negative but also in HER2+ breast cancer. However, the clinical utility in the adjuvant setting for HER2+ disease needs further validation. The aim of this study was to investigate the prognostic significance of HER2/CEP17 ratio, HER2 copy numbers and TILs among early breast cancer patients treated with HER2-targeted therapy. Materials and methods: A retrospective study cohort comprising a total of 584 patients with HER2+ early breast cancer, all treated with the monoclonal antibody trastuzumab targeting HER2 during 2006-2014 in the Stockholm Region, Sweden, was identified and HER2 re-testing (silver in situ hybridization and immunohistochemistry (IHC)) was performed on 474 whole tissue sections prior to neoadjuvant (18.1%) or adjuvant therapy. Stromal TILs were manually assessed on all hematoxylin-eosin stained slides. Registry-based information on patient- and tumor characteristics including treatment were completed from medical records with a median follow-up of 8.0 years. Results: Among the 474 analyzed tumors, 11.4% were HER2 IHC score 2+ and 86.7% were 3+. In addition, 61.6% were ER+ and the median TIL score was 15%. The median HER2 copy number was 9.8 signals/cell and only 12.9% had <6.0 average HER2 signals/cell and 2.5% had a HER2/CEP17 ratio <2.0. Survival analysis showed no significant association between HER2 IHC score (recurrence: p=0.58; breast cancer specific death: p=0.69), HER2 signals (p=0.77; p=0.73), HER2/CEP17 ratio (p=0.73; p=0.85), ER status (p=0.55; p=0.11), PR status (p=0.97; p=0.46), Ki67 (p=0.14; p=0.77) or tumor grade (p=0.71; p=0.16) with risk of recurrence or breast cancer specific death. The hazard ratio (HR) for recurrence did not significantly differ by HER2 signals/cell (HR 1.12, CI 0.53-2.37, p=0.77) or ER status (HR 1.18, CI 0.68-2.05, p=0.55). Tumors with low TILs were independently associated with recurrence (cutoff <40%: HR 4.5, CI 1.4-14.9, p=0.012) and breast cancer specific survival (cutoff <30%: HR 3.5, CI 1.2-10.1, p=0.02). Among the 56 recurrences (11.8%), 37.5% had ER- primary tumors and 82.4% of the 34 HER2-analyzed recurrent tumors remained HER2+. Conclusions: This study demonstrates that routine clinicopathological data, including HER2 status, is insufficient to predict breast cancer recurrence among patients treated with HER2-targeted therapy. Neither HER2 levels, HER2/CEP17 ratio or ER status correlated to risk of recurrence. Importantly, TILs showed robust prognostic potential in trastuzumab-treated patients. Our results emphasize the need for detailed molecular analyses to understand the biology behind resistance to HER2-targeted therapy and to further tailor therapy for HER2+ breast cancer patients. Citation Format: Stephanie Robertson, Caroline Rönnlund, Irma Fredriksson, Theodoros Foukakis, Johan Hartman. Tumor-infiltrating lymphocytes but not HER2 copy number or ratio show prognostic value in trastuzumab-treated HER2-positive breast cancer [abstract]. In: Proceedings of the 2021 San Antonio Breast Cancer Symposium; 2021 Dec 7-10; San Antonio, TX. Philadelphia (PA): AACR; Cancer Res 2022;82(4 Suppl):Abstract nr P1-02-03.

Using Classification and K-means Methods to Predict Breast Cancer Recurrence in Gene Expression Data.

Background:Breast cancer is a type of cancer that starts in the breast tissue and affects about 10% of women at different stages of their lives. In this study, we applied a new method to predict recurrence in biological networks made from gene expression data.Method:The method includes the steps such as data collection, clustering, determining differentiating genes, and classification. The eight techniques consist of random forest, support vector machine and neural network, randomforest + k-means, hidden markov model, joint mutual information, neural network + k-means and suportvector machine + k-menas were implemented on 12172 genes and 200 samples.Results:Thirty genes were considered as differentiating genes which used for the classification. The results showed that random forest + k-means get better performance than other techniques. The two techniques including neural network + k-means and random forest + k-means performed better than other techniques in identifying high risk cases.Conclusion:Thirty of 12,172 genes are considered for classification that the use of clustering has improved the classification techniques performance.

Predictive value of DNA repair gene expression for response to neoadjuvant chemotherapy in breast cancer.

Genome-wide analysis using microarrays has revolutionized breast cancer (BC) research. A substantial body of evidence supports the clinical utility of the 21-gene assay (Oncotype DX) and 70-gene assay (MammaPrint) to predict BC recurrence and the magnitude of benefit from chemotherapy. However, there is currently no genetic tool able to predict chemosensitivity and chemoresistance to neoadjuvant chemotherapy (NACT) during BC treatment. In this study, we explored the predictive value of DNA repair gene expression in the neoadjuvant setting. We selected 98 patients with BC treated with NACT. We assessed DNA repair expression in 98 formalin-fixed, paraffin-embedded core biopsy fragments used at diagnosis and in 32 formalin-fixed, paraffin-embedded post-NACT residual tumors using quantitative reverse transcription-polymerase chain reaction. The following genes were selected: BRCA1, PALB2, RAD51C, BRCA2, ATM, FANCA, MSH2, XPA, ERCC1, PARP1, and SNM1. Of 98 patients, 33 (33.7%) achieved pathologic complete response (pCR). The DNA expression of 2 genes assessed in pre-NACT biopsies (PALB2 and ERCC1) was lower in pCR than in non-pCR patients (P=0.005 and P=0.009, respectively). There was no correlation between molecular subtype and expression of DNA repair genes. The genes BRCA2 (P=0.009), ATM (P=0.004), FANCA (P=0.001), and PARP1 (P=0.011) showed a lower expression in post-NACT residual tumor samples (n=32) than in pre-NACT biopsy samples (n=98). The expression of 2 genes (PALB2 and ERCC1) was lower in pCR patients. These alterations in DNA repair could be considered suitable targets for cancer therapy.

AN EFFICIENT BREAST CANCER RECURRENCE PREDICTION USING SIMULATED ANNEALING INERTIA WEIGHT-BASED CHICKEN SWARM OPTIMIZATION ALGORITHM WITH WEIGHTED MINKOWSKI RADIAL BASIS FUNCTION-BASED SUPPORT VECTOR MACHINE

Nowadays the more often diagnosed type of cancer in women is breast cancer. Approximately 12 percent of females worldwide are afflicted by it. Recurrent breast cancer refers to breast cancer that recurs despite having been successfully treated. Breast cancer recurrence is one of the most feared outcomes for cancer patients. As a result, their quality of life is impacted. Even though early-stage breast cancer prediction has always been a difficult research challenge, data mining algorithms can be of considerable assistance in addressing this issue. The prior method used naive Bayes, REPTree, and K-nearest neighbor, Particle Swarm Optimization (PSO) based feature selection for breast cancer recurrence prediction. It is simple for the Particle Swarm Optimization (PSO) method to fall into local optimum in high-dimensional space and has a poor rate of convergence in the iterative process. Furthermore, because of the Nave Bayes’ class conditional independence and the resulting loss of accuracy, it is not recommended. To deal with this problem, the proposed system devised a Simulated Annealing Inertia Weight-based Chicken Swarm Optimization (SAIWCSO) algorithm with a Weighted Minkowski Radial Basis Function-based Support Vector Machine (WMRBF-SVM) for the diagnosis of the reputation of breast cancer. The WBCD is used as an initial source of data. Z-score normalization is then used to ensure that the data is in a consistent format. An algorithm called SAIWCS selects the best features. It enhances the categorization accuracy. To improve the high detection rate, a Weighted Minkowski Radial Basis Function-based Support Vector Machine (WMRBF-SVM) is used to diagnose breast cancer. Python is being used to model the experiments. The experimental findings reveal that the suggested system outperforms the present system with regard to the accuracy, precision, recall, specificity, and f-measure.

Privacy-preserving breast cancer recurrence prediction based on homomorphic encryption and secure two party computation.

Protecting patients’ privacy is one of the most important tasks when developing medical artificial intelligence models since medical data is the most sensitive personal data. To overcome this privacy protection issue, diverse privacy-preserving methods have been proposed. We proposed a novel method for privacy-preserving Gated Recurrent Unit (GRU) inference model using privacy enhancing technologies including homomorphic encryption and secure two party computation. The proposed privacy-preserving GRU inference model validated on breast cancer recurrence prediction with 13,117 patients’ medical data. Our method gives reliable prediction result (0.893 accuracy) compared to the normal GRU model (0.895 accuracy). Unlike other previous works, the experiment on real breast cancer data yields almost identical results for privacy-preserving and conventional cases. We also implement our algorithm to shows the realistic end-to-end encrypted breast cancer recurrence prediction.

Breast cancer recurrence prediction system using k-nearest neighbor, naïve-bayes, and support vector machine algorithm

Breast cancer is a serious disease and one of the most fatal diseases in the world. Statistics show that breast cancer is the second common cancer worldwide with around two million new cases per year. Some research has been done related to breast cancer, and with the advancements of technology, breast cancer can be detected earlier by using artificial intelligence or machine learning. There are popular machine learning algorithms that can be used to predict the existence or recurrence of breast disease, for example, k-Nearest Neighbor (kNN), Naïve Bayes, and Support Vector Machine (SVM). This study aims to check the prediction of breast cancer recurrence using those three algorithms using the dataset available at the University of California, Irvine (UCI). The result shows that the kNN algorithm gives the best result in terms of accuracy to predict breast cancer recurrence.

Socioeconomic position and prognosis in premenopausal breast cancer: a population-based cohort study in Denmark

BackgroundTo investigate how socioeconomic position (SEP) influences the effectiveness of cancer-directed treatment in premenopausal breast cancer patients in terms of breast cancer recurrence and mortality.MethodsWe conducted a cohort study nested in the ProBeCaRe (Predictors of Breast Cancer Recurrence) cohort (n = 5959). We identified all premenopausal women aged 18–55 years diagnosed with non-metastatic breast cancer and prescribed docetaxel-based chemotherapy in Denmark during 2007–2011. Population-based administrative registries provided data on SEP: marital status (married including registered partnership or single including divorced or widowed), cohabitation (cohabiting or living alone), education (low, intermediate, or high), income (low, medium, or high), and employment status (employed, unemployed, or health-related absenteeism). For each SEP measure, we computed incidence rates, cumulative incidence proportions (CIPs), and used Poisson regression to compute incidence rate ratios (IRRs) and 95% confidence intervals (CIs) of recurrence and death. We stratified on estrogen receptor (ER) status/tamoxifen to evaluate interaction.ResultsOur study cohort included 2616 women; 286 (CIP 13%) experienced recurrence and 223 (CIP 11%) died during follow-up (median 6.6 and 7.2 years, respectively). Single women had both increased 5-year risks of recurrence (IRR 1.45, 95% CI 1.11–1.89) and mortality (IRR 1.83, 95% CI 1.32–2.52). Furthermore, we observed increased 5-year mortality in women with low education (IRR 1.49, 95% CI 0.95–2.33), low income (IRR 1.37, 95% CI 0.83–2.28), unemployment (IRR 1.61, 95% CI 0.83–3.13), or health-related work absenteeism (IRR 1.80, 95% CI 1.14–2.82), but smaller or no increased risk of recurrence. These findings were especially evident among women with ER+ tumors prescribed tamoxifen. Overall analyses (follow-up max. 10 years) provided similar results.ConclusionsLow SEP in premenopausal women with non-metastatic breast cancer was associated with increased mortality, but not always recurrence. This suggests underdetection of recurrences in certain groups. Poor prognosis in women with low SEP, especially single women, may partly be explained by tamoxifen adherence.

Recurrence Prediction Score of the Localy Advance Breast Cancer in West Sumatera Province, Indonesia

Introduction: Breast cancer is the most common malignancy that occurs in women in the world, which is about 18% of cancer cases that occur in women. Every year there are 1 million new cases of breast cancer worldwide and nearly 60,000 patients die from this disease. Therefore, the authors want to find out more about the prognosis factors that play a role in influencing the survival of patients with locally advanced breast cancer in Padang city, West Sumatera Province Indonesia.
 Methods: This study is a retrospective case-control study of locally advanced breast cancer for patients in Padang city. The study was conducted at the Surgical Oncology Division of Dr. M Djamil Padang Hospital, Ropana Suri Hospital, Ibnu Sina Hospital and Siti Rahmah Padang Hospital from September 2018 to December 2018. The total number of samples for the two groups was 242 people. The sampling technique used was convenience sampling.
 Results: In this study 4 prognostic factors were found that had a significant effect on predicting breast cancer recurrence, namely histopathological type, stage, regional lymph and lymphovascular invasion with p <0.05.
 Conclusion: Prognosis factors that influence breast cancer recurrence are histopathological type, stage, regional lymph node status and lymphovascular invasion. A scoring system is obtained that is good enough to predict the possibility of breast cancer recurrence.

Recurrence Prediction Score of the Localy Advance Breast Cancer in West Sumatera Province, Indonesia

Introduction: Breast cancer is the most common malignancy that occurs in women in the world, which is about 18% of cancer cases that occur in women. Every year there are 1 million new cases of breast cancer worldwide and nearly 60,000 patients die from this disease. Therefore, the authors want to find out more about the prognosis factors that play a role in influencing the survival of patients with locally advanced breast cancer in Padang city, West Sumatera Province Indonesia.
 Methods: This study is a retrospective case-control study of locally advanced breast cancer for patients in Padang city. The study was conducted at the Surgical Oncology Division of Dr. M Djamil Padang Hospital, Ropana Suri Hospital, Ibnu Sina Hospital and Siti Rahmah Padang Hospital from September 2018 to December 2018. The total number of samples for the two groups was 242 people. The sampling technique used was convenience sampling.
 Results: In this study 4 prognostic factors were found that had a significant effect on predicting breast cancer recurrence, namely histopathological type, stage, regional lymph and lymphovascular invasion with p <0.05.
 Conclusion: Prognosis factors that influence breast cancer recurrence are histopathological type, stage, regional lymph node status and lymphovascular invasion. A scoring system is obtained that is good enough to predict the possibility of breast cancer recurrence.

A hybrid multi-stage learning technique based on brain storming optimization algorithm for breast cancer recurrence prediction

Breast cancer disease is considered to be the second leading reason for death among women. Unfortunately, even if the treatment of cancer started soon after diagnosis, the cancer cells may remain in the body, and cancer may recur. Various Machine Learning (ML) methods to predict breast cancer recurrence were applied recently, and the ML methods’ performance needs to be examined to determine the proper method for prediction. Usually, the datasets contain many features which may sometimes mislead the prediction process; as some features may lead to confusion or inaccurate prediction. Thus, in this study, two breast cancer recurrence datasets were statistically analyzed and further refined by Brain Storming Optimization algorithm (BSO). The proposed multi-stages technique consists of three main stages; first, the statistical feature selection methods (SFM) which statistically select the discriminative features based on importance ranking and features correlations and statistical hypothesis testing, to be passed to the second stage. The features are ranked based on their correlation with the class variable. The second stage namely the multi classifier (MC), which evaluates each method based on three classifiers and produces a combination of features performed by two SFM and three classifiers. In the third stage, the best combination of the selected features was recognized by the BSO algorithm to search for an optimal solution that produces the highest accuracy. In addition, the BSO algorithm has been modified to deal with the feature selection problem. The performance of the proposed technique was evaluated by stratified 10-fold cross-validation. As a result, the multi-stage learning technique showed to be effective in ranking the features and improved the classification accuracy for breast cancer recurrence.

Breast cancer recurrence prediction with ensemble methods and cost-sensitive learning.

Breast cancer is one of the most common cancers in women all over the world. Due to the improvement of medical treatments, most of the breast cancer patients would be in remission. However, the patients have to face the next challenge, the recurrence of breast cancer which may cause more severe effects, and even death. The prediction of breast cancer recurrence is crucial for reducing mortality. This paper proposes a prediction model for the recurrence of breast cancer based on clinical nominal and numeric features. In this study, our data consist of 1,061 patients from Breast Cancer Registry from Shin Kong Wu Ho-Su Memorial Hospital between 2011 and 2016, in which 37 records are denoted as breast cancer recurrence. Each record has 85 features. Our approach consists of three stages. First, we perform data preprocessing and feature selection techniques to consolidate the dataset. Among all features, six features are identified for further processing in the following stages. Next, we apply resampling techniques to resolve the issue of class imbalance. Finally, we construct two classifiers, AdaBoost and cost-sensitive learning, to predict the risk of recurrence and carry out the performance evaluation in three-fold cross-validation. By applying the AdaBoost method, we achieve accuracy of 0.973 and sensitivity of 0.675. By combining the AdaBoost and cost-sensitive method of our model, we achieve a reasonable accuracy of 0.468 and substantially high sensitivity of 0.947 which guarantee almost no false dismissal. Our model can be used as a supporting tool in the setting and evaluation of the follow-up visit for early intervention and more advanced treatments to lower cancer mortality.

Early Prediction of Breast Cancer Recurrence for Patients Treated with Neoadjuvant Chemotherapy: A Transfer Learning Approach on DCE-MRIs.

Simple SummaryAn early prediction of Breast Cancer Recurrence (BCR) for patients undergoing neoadjuvant chemotherapy (NACT) could better guide clinicians in the identification of the most suitable combination treatments for individual patient scenarios. We proposed a transfer learning approach to give an early prediction of three-year BCR for patients undergoing NACT, using DCE-MRI exams from I-SPY1 TRIAL and BREAST-MRI-NACT-Pilot public databases. Because no technical expertise is required in the extraction of meaningful features from images, the predictive model qualifies as a user-friendly tool for any medical expert in support of therapeutic choices. Only pre-treatment and early-treatment MRI examinations were analyzed to allow for potential therapy changes at a very early stage of treatment. We tested the strength of the model on an independent test. The best predictive performances (accuracy of 85.2%, sensitivity of 84.6%, and AUC of 0.83) were achieved by combining the extracted features with some clinical factors: age, ER, PgR, HER2+.Cancer treatment planning benefits from an accurate early prediction of the treatment efficacy. The goal of this study is to give an early prediction of three-year Breast Cancer Recurrence (BCR) for patients who underwent neoadjuvant chemotherapy. We addressed the task from a new perspective based on transfer learning applied to pre-treatment and early-treatment DCE-MRI scans. Firstly, low-level features were automatically extracted from MR images using a pre-trained Convolutional Neural Network (CNN) architecture without human intervention. Subsequently, the prediction model was built with an optimal subset of CNN features and evaluated on two sets of patients from I-SPY1 TRIAL and BREAST-MRI-NACT-Pilot public databases: a fine-tuning dataset (70 not recurrent and 26 recurrent cases), which was primarily used to find the optimal subset of CNN features, and an independent test (45 not recurrent and 17 recurrent cases), whose patients had not been involved in the feature selection process. The best results were achieved when the optimal CNN features were augmented by four clinical variables (age, ER, PgR, HER2+), reaching an accuracy of 91.7% and 85.2%, a sensitivity of 80.8% and 84.6%, a specificity of 95.7% and 85.4%, and an AUC value of 0.93 and 0.83 on the fine-tuning dataset and the independent test, respectively. Finally, the CNN features extracted from pre-treatment and early-treatment exams were revealed to be strong predictors of BCR.

Weakly supervised temporal model for prediction of breast cancer distant recurrence

Efficient prediction of cancer recurrence in advance may help to recruit high risk breast cancer patients for clinical trial on-time and can guide a proper treatment plan. Several machine learning approaches have been developed for recurrence prediction in previous studies, but most of them use only structured electronic health records and only a small training dataset, with limited success in clinical application. While free-text clinic notes may offer the greatest nuance and detail about a patient’s clinical status, they are largely excluded in previous predictive models due to the increase in processing complexity and need for a complex modeling framework. In this study, we developed a weak-supervision framework for breast cancer recurrence prediction in which we trained a deep learning model on a large sample of free-text clinic notes by utilizing a combination of manually curated labels and NLP-generated non-perfect recurrence labels. The model was trained jointly on manually curated data from 670 patients and NLP-curated data of 8062 patients. It was validated on manually annotated data from 224 patients with recurrence and achieved 0.94 AUROC. This weak supervision approach allowed us to learn from a larger dataset using imperfect labels and ultimately provided greater accuracy compared to a smaller hand-curated dataset, with less manual effort invested in curation.

Clinicopathological correlates, oncological impact, and validation of Oncotype DX™ in a European Tertiary Referral Centre.

Oncotype DX™ (ODX) score estimates prognosis and predicts breast cancer recurrence. It also individualizes patient adjuvant chemotherapy prescription in breast cancer. This assay relies on genetic and molecular markers; the clinicopathological phenotype of which are tested routinely. The aim of this study was determine whether clinicopathological and immunohistochemical information predicts ODX recurrence score (RS). Secondly, to assess the impact on adjuvant chemotherapy (AC) and oncological outcome of ODX testing in patients in a European tertiary referral center. Estrogen receptor positive (ER+), human epidermal growth factor receptor-2 negative (HER2-), lymph node negative (LN-), and female breast cancer patients with ODX testing performed between 2007 and 2015 were categorized into low- (<11), intermediate- (11-25), and high-risk (>25) groups. Clinicopathological and immunohistochemical correlates of RS were determined. Predictors of RS were assessed using binary logistic regression. Oncological outcome was assessed using Kaplan-Meier and Cox regression analyses. ODX was performed in 400 consecutive ER+LN- patients. Median follow-up was 74.1months (3.0-144.4). Low grade (odds ratio [OR]:2.39; 95% confidence interval [CI]:1.04-5.51, p=0.041) independently predicted low ODX, while high grade (OR:2.04; 95% CI: 1.19-3.49, p=0.009) and reduced progesterone receptor (PgR) expression (OR: 2.57, 95% CI: 1.42-4.65, p=0.002) independently predicted high ODX. Omission of AC in intermediate- (p=0.159) and high-risk (p=0.702) groups did not negatively impact survival. In conclusion, tumor grade independently predicts low and high RS, while PgR negativity predicts high RS. ODX reduced AC prescription without compromising oncological outcome.

Abstract PS6-06: The 70-gene signature (MammaPrint) accurately predicts distant breast cancer recurrence risk in patients aged ≥70 years from the population-based observational FOCUS cohort

Abstract Background Predicting breast cancer recurrence in patients aged ≥70 years is challenging, as they generally have more indolent tumors and a higher chance of dying of competing causes than younger patients. The 70-gene signature test (MammaPrint) has been shown to accurately predict recurrence in women with early breast cancer and up to 3 positive lymph nodes. Aim To study outcome related to MammaPrint result in patients aged ≥70 years with breast cancer using a population-based cohort. Methods The population-based FOCUS cohort included all 2095 consecutive patients with any stage breast cancer, diagnosed between 1997 and 2004, aged ≥65 years, in the Comprehensive Cancer Center region West, the Netherlands. In the present exploratory sub-study, patients from FOCUS with the following criteria were included: ≥70 years old, T1-2N0-3M0, hormone receptor positive, HER2 negative, no neo-adjuvant treatment and available tumor specimens. MammaPrint is a genomic risk profile based on microarray gene expression analysis, classifying patients as ultralow risk (M-ULR), low (not UL) risk (M-LR) or high risk (M-HR) of developing a recurrence. Patients were considered clinically low risk (C-LR) with T1-2N0 grade 1-2 tumors and clinically high risk (C-HR) with N+ or T2/grade 3 tumors. Primary endpoint was 10-year distant recurrence free interval (DRFi) in relation to genomic risk, estimated from cumulative incidence and Fine and Gray analyses to take competing mortality into account. Results In this study, 422 patients were included. Median age was 78 years, 238 patients (56%) had node negative disease, 235 patients (56%) had T2 tumors and 227 patients (54%) were C-LR. Most patients were treated with endocrine therapy (ET), and 22 patients (5%) were treated with chemotherapy (CT; table 1). Overall, 50 (12%) patients were M-ULR, 226 (53%) were M-LR and 146 (35%) were M-HR. Discrepancies were found between C and M risk groups in 18/50 M-ULR patients with C-HR, and 56/146 M-HR patients with C-LR. Of the 59 patients that experienced a recurrence during 10 years of follow-up, 44 (75%) were distant recurrences. In the M-ULR group, DRFi was 2% (95%CI 0-6) after 10 years of follow-up, this was 8% (95%CI 5-12) in the M-LR group and 17% (95%CI 11-23) in the M-HR group (p&amp;lt;0.001). In the C-HR subgroup, none of the 18 M-ULR patients developed a recurrence, and DRFi was 10% (95%CI 3-16) in M-LR patients and 20% (95%CI 12-28) in M-HR patients (p=0.015). C risk alone was not able to predict distant recurrence risk (C-LR 8%, C-HR 14%, sHR 1.8 [95%CI 0.9-3.2); p=0.060; table 2). Conclusion MammaPrint accurately predicts 10-year DRFi in older patients with breast cancer. Patients classified as ultralow risk by MammaPrint had a very low chance of developing metastatic disease. Even in clinically high-risk patients who were M-ULR, recurrent disease did not occur 10 years after diagnosis. These findings are in line with published results of the STO-3 trial (JAMA Oncol, 2017) and provide foundation for de-escalation of treatment in older patients guided by genomic testing. Table 1: Baseline characteristics. BCS = breast conserving surgery. RT = radiotherapyM-ULRM-LRM-HRTotal patientsN (%)50 (11.8)226 (53.6)146 (34.6)AgeMedian (IQR)79 (74-85)79 (74-84)77 (74-84)Histological grade [N (%)]I11 (22.0)39 (17.3)5 (3.4)II21 (42.0)92 (40.7)42 (28.8)III1 (2.0)31 (13.7)54 (37.0)Missing17 (34.0)64 (28.3)45 (30.8)T-stage [N (%)]Tis0 (0.0)2 (0.9)0 (0.0)T117 (34.0)111 (49.1)54 (37.0)T233 (66.0)113 (50.0)89 (61.0)Missing0 (0.0)0 (0.0)3 (2.0)N-stage [N (%)]N032 (64.0)140 (61.9)66 (45.2)N117 (34.0)72 (31.9)67 (45.9)N20 (0.0)6 (2.7)6 (4.1)N30 (0.0)2 (0.9)3 (2.1)Missing1 (2.0)6 (2.7)4 (2.7)Clinical risk [N (%)]Low32 (64.0)139 (61.5)56 (38.4)High18 (36.0)87 (38.5)90 (61.6)Local treatment [N (%)]None4 (8.0)5 (2.2)6 (4.1)BCS only3 (6.0)16 (7.1)13 (8.9)BCS + RT9 (18.0)62 (27.4)29 (19.9)Mastectomy34 (68.0)143 (63.3)98 (67.1)Adjuvant ET [N (%)]None24 (48.0)103 (45.6)40 (27.4)Tamoxifen17 (34.0)87 (38.5)80 (54.8)Aromatase inhibitor3 (6.0)11 (4.9)9 (6.2)Unspecified ET6 (12.0)25 (11.1)17 (11.6)Adjuvant CT [N (%)]No46 (92.0)217 (96.0)137 (93.8)Yes4 (8.0)9 (4.0)9 (6.2) Table 2: Primary endpoint stratified by genomic and clinical risk. sHR=subdistribution hazard ratio.M-ULRM-LRM-HRTotalC-LRN=32 (14%). DRFi=3% (95%CI 0-9). sHR=1 (reference).N=139 (61%). DRFi=7% (95%CI 3-12). sHR=2.1 (95%CI 0.3-16.5).N=56 (25%). DRFi=13% (95%CI 4-21). sHR=4.3 (95%CI 0.5-34.7).N=227. DRFi=8% (95%CI 4-12).C-HRN=18 (9%).DRFi=0% (95%CI 0-0). sHR=N/A (n events=0).N=87 (45%). DRFi=10% (95%CI 3-16). sHR=1 (reference).N=90 (46%). DRFi=20% (95%CI 12-28). sHR=3.0 (95%CI 1.3-6.9).N=195. DRFi=14% (95%CI 9-19).TotalN=50 (12%). DRFi=2% (95%CI 0-6). sHR=1 (reference).N=226 (53%). DRFi=8% (95%CI 5-12). sHR=3.8 (95%CI 0.5-28.2).N=146 (35%). DRFi=17% (95%CI 11-23). sHR=9.8 (95%CI 1.3-72.6).N=422. Citation Format: Iris Noordhoek, Esther Bastiaannet, Ersan Lujinovic, Laura Esserman, Jelle Wesseling, Astrid Scholten, Carolien P Schröder, Sjoerd Elias, Nienke A de Glas, Judith R Kroep, Johanneke EA Portielje, Miranda Kleijn, Gerrit-Jan J Liefers. The 70-gene signature (MammaPrint) accurately predicts distant breast cancer recurrence risk in patients aged ≥70 years from the population-based observational FOCUS cohort [abstract]. In: Proceedings of the 2020 San Antonio Breast Cancer Virtual Symposium; 2020 Dec 8-11; San Antonio, TX. Philadelphia (PA): AACR; Cancer Res 2021;81(4 Suppl):Abstract nr PS6-06.

Breast Cancer Tissue Sub-Region Classification from Second Harmonic Generation Imagery via Machine Learning

Cancer treatment involves complex decision-making processes. A better understanding of recurrence risk at diagnosis, as well as prediction of treatment response (e.g., to choose the most cost-effective treatment path in an informed manner) are needed to produce the best possible outcomes at the patient level. Prior work shows that the forward/backward (F/B) ratio calculated from Second Harmonic Generation (SHG) imagery can be indicative of risk of recurrence if relevant tissue sub-regions are selected. The choice of which sub-regions to image is currently made by human experts, which is subjective and labor intensive. In this paper, we investigate machine learning methods to automatically identify tissue sub-regions that are most relevant to the prediction of breast cancer recurrence. We formulate the task as a multi-class classification problem and use support vector machine (SVM) classifiers as the inference engine. Given the limited amount of data available, we focus on exploring the feature extraction stage. To that end, we evaluate methods leveraging handcrafted features, deep features extracted from pre-trained models, as well as features extracted via transfer learning. The results show a steady trend of improvement on the classification accuracy as the features become more data-driven and customized to the task at hand. This is an indication that having larger amounts of labeled data could be beneficial for improving automated methods of classification. The best results achieved, using features learned via transfer learning from ResNet-101, correspond to 85% accuracy in a 3-class problem and 94% accuracy in a binary classification problem.

Machine Learning Classifiers on Breast Cancer Recurrences

Breast cancer remains to be a leading cause of cancer-related deaths among women. Mortality is mainly attributed to metastasis and recurrence. Hence, early detection of breast cancer recurrence has become a real-world medical problem. Using data mining approaches, we compared four popular machine learning models (Logistic Regression, Naïve Bayes, K-Nearest Neighbors, and Support Vector Machines) on a high-dimensional but very small dataset, the Wisconsin Prognostic Breast Cancer Data Set for classifying breast cancer recurrences on four different configurations: a) only scaling applied, b) scaling with PCA, c) scaling with PCA and oversampling of minority class, and d) only with selected features (i.e. choose only one from each set of features that have high correlation with each other). Our results showed that Logistic Regression provided the best scores in almost all metrics (precision, recall, accuracy, f1 score (weighted), AUROC, AUPROC, and Cohen Kappa statistic in all four configurations, followed by Support Vector Machines, and then by K-Nearest Neighbors. Naive Bayes performed poorly especially in the scaling with PCA configuration, however, when we retained only one of many features that have high correlations with each other, Naïve Bayed performance improved. KNN improved its recall with oversampling while SVM’s accuracy score has been fairly constant in all four configurations. Based on this study, the Logistic Regression model can serve as a potential model for predicting breast cancer recurrence that would enable clinicians to propose treatment options based on whether patient’s features correspond to a good or bad prognosis (recurrence). This indicates the clinical utility of data mining methods for the early detection of breast cancer recurrence in post-surgical patients to save lives.

Metabolic Gene Signature for Predicting Breast Cancer Recurrence Using Transcriptome Analysis

Background: The study aimed at identifying a metabolic gene signature for stratifying the risk of recurrence in breast cancer. Materials & methods: The data of patients were obtained from the Cancer Genome Atlas (TCGA) and Gene Expression Omnibus (GEO) database. The limma package was used to identify differentially expressed metabolic genes, and a metabolic gene signature was constructed. Results: A five-gene metabolic signature was established that demonstrated satisfactory accuracy and predictive power in both training and validation cohorts. Also, a nomogram for predicting recurrence-free survival was established using a combination of the metabolism gene risk score and the clinicopathological features. Conclusions: The proposed metabolic gene signature and nomogram have a significant prognostic value and may improve the recurrence risk stratification for breast cancer patients.

Predicting breast cancer recurrence using principal component analysis as feature extraction: an unbiased comparative analysis

Breast cancer recurrence is among the most noteworthy fears faced by women. Nevertheless, with modern innovations in data mining technology, early recurrence prediction can help relieve these fears. Although medical information is typically complicated, and simplifying searches to the most relevant input is challenging, new sophisticated data mining techniques promise accurate predictions from high-dimensional data. In this study, the performances of three established data mining algorithms: Naïve Bayes (NB), k-nearest neighbor (KNN), and fast decision tree (REPTree), adopting the feature extraction algorithm, principal component analysis (PCA), for predicting breast cancer recurrence were contrasted. The comparison was conducted between models built in the absence and presence of PCA. The results showed that KNN produced better prediction without PCA (F-measure = 72.1%), whereas the other two techniques: NB and REPTree, improved when used with PCA (F-measure = 76.1% and 72.8%, respectively). This study can benefit the healthcare industry in assisting physicians in predicting breast cancer recurrence precisely.