Purpose To investigate the association between a validated, gene-expression-based, aggressiveness assay, Oncotype Dx RS, and morphological and texture-based image features extracted from magnetic resonance imaging (MRI). Materials and Methods This retrospective study received Internal Review Board approval and need for informed consent was waived. Between 2006–2012, we identified breast cancer patients with: 1) ER+, PR+, and HER2− invasive ductal carcinoma (IDC); 2) preoperative breast MRI; and 3) Oncotype Dx RS test results. Extracted features included morphological, histogram, and gray-scale correlation matrix (GLCM)-based texture features computed from tumors contoured on pre- and three postcontrast MR images. Linear regression analysis was performed to investigate the association between Oncotype Dx RS and different clinical, pathologic, and imaging features. P < 0.05 was considered statistically significant. Results Ninety-five patients with IDC were included with a median Oncotype Dx RS of 16 (range: 0–45). Using stepwise multiple linear regression modeling, two MR-derived image features, kurtosis in the first and third postcontrast images and histologic nuclear grade, were found to be significantly correlated with the Oncotype Dx RS with P = 0.0056, 0.0005, and 0.0105, respectively. The overall model resulted in statistically significant correlation with Oncotype Dx RS with an R-squared value of 0.23 (adjusted R-squared = 0.20; P = 0.0002) and a Spearman’s rank correlation coefficient of 0.49 (P < 0.0001). Conclusion A model for IDC using imaging and pathology information correlates with Oncotype Dx RS scores, suggesting that image-based features could also predict the likelihood of recurrence and magnitude of chemotherapy benefit.

Purpose To use features extracted from magnetic resonance (MR) images and a machine-learning method to assist in differentiating breast cancer molecular subtypes. Materials and Methods This retrospective Health Insurance Portability and Accountability Act (HIPAA)-compliant study received Institutional Review Board (IRB) approval. We identified 178 breast cancer patients between 2006–2011 with: 1) ERPR + (n = 95, 53.4%), ERPR−/HER2 + (n = 35, 19.6%), or triple negative (TN, n = 48, 27.0%) invasive ductal carcinoma (IDC), and 2) preoperative breast MRI at 1.5T or 3.0T. Shape, texture, and histogram-based features were extracted from each tumor contoured on pre- and three postcontrast MR images using in-house software. Clinical and pathologic features were also collected. Machine-learning-based (support vector machines) models were used to identify significant imaging features and to build models that predict IDC subtype. Leave-one-out cross-validation (LOOCV) was used to avoid model overfitting. Statistical significance was determined using the Kruskal–Wallis test. Results Each support vector machine fit in the LOOCV process generated a model with varying features. Eleven out of the top 20 ranked features were significantly different between IDC subtypes with P < 0.05. When the top nine pathologic and imaging features were incorporated, the predictive model distinguished IDC subtypes with an overall accuracy on LOOCV of 83.4%. The combined pathologic and imaging model's accuracy for each subtype was 89.2% (ERPR+), 63.6% (ERPR−/HER2+), and 82.5% (TN). When only the top nine imaging features were incorporated, the predictive model distinguished IDC subtypes with an overall accuracy on LOOCV of 71.2%. The combined pathologic and imaging model's accuracy for each subtype was 69.9% (ERPR+), 62.9% (ERPR−/HER2+), and 81.0% (TN). Conclusion We developed a machine-learning-based predictive model using features extracted from MRI that can distinguish IDC subtypes with significant predictive power.

Although guidelines such as those of the National Comprehensive Cancer Network consider 18F-FDG PET/CT for systemic staging of newly diagnosed stage III breast cancer patients, factors in addition to stage may influence the utility of PET/CT. Because invasive lobular carcinoma (ILC) is less conspicuous than invasive ductal carcinoma (IDC) on 18F-FDG PET, we hypothesized that tumor histology may be one such factor. We evaluated PET/CT systemic staging of patients newly diagnosed with ILC compared with IDC. Methods In this Institutional Review Board–approved retrospective study, our Hospital Information System was screened for ILC patients who underwent PET/CT in 2006–2013 before systemic or radiation therapy. Initial stage was determined from examination, mammography, ultrasound, MR, or surgery. PET/CT was performed to identify unsuspected distant metastases. A sequential cohort of stage III IDC patients was evaluated for comparison. Upstaging rates were compared using the Pearson χ2 test. Results The study criteria were fulfilled by 146 ILC patients. PET/CT revealed unsuspected distant metastases in 12 (8%): 0 of 8 with initial stage I, 2 of 50 (4%) stage II, and 10 of 88 (11%) stage III. Upstaging to IV by PET/CT was confirmed by biopsy in all cases. Three of 12 upstaged patients were upstaged only by the CT component of the PET/CT, as the metastases were not 18F-FDG–avid. In the comparison stage III IDC cohort, 22% (20/89) of patients were upstaged to IV by PET/CT. All 20 demonstrated 18F-FDG–avid metastases. The relative risk of PET/CT revealing unsuspected distant metastases in stage III IDC patients was 1.98 times (95% confidence interval, 0.98–3.98) that of stage III ILC patients (P = 0.049). For 18F-FDG–avid metastases, the relative risk of PET/CT revealing unsuspected 18F-FDG–avid distant metastases in stage III IDC patients was 2.82 times (95% confidence interval, 1.26–6.34) that of stage III ILC patients (P = 0.007). Conclusion 18F-FDG PET/CT was more likely to reveal unsuspected distant metastases in stage III IDC patients than in stage III ILC patients. In addition, some ILC patients were upstaged by non–18F-FDG-avid lesions visible only on the CT images. Overall, the impact of PET/CT on systemic staging may be lower for ILC patients than for IDC patients.

Purpose Determine if the histology of a breast malignancy influences the appearance of untreated osseous metastases on FDG PET/CT. Methods This retrospective study was performed under IRB waiver. Our Hospital Information System was screened for breast cancer patients who presented with osseous metastases, who underwent FDG PET/CT prior to systemic therapy or radiation from 2009–2012. Patients with invasive ductal carcinoma (IDC), invasive lobular carcinoma (ILC), or mixed ductal/lobular (MDL) histology were included. Patients with history of other malignancies were excluded. PET/CT was evaluated, blinded to histology, to classify osseous metastases on a per patient basis as sclerotic, lytic, mixed lytic/sclerotic, or occult on CT, and record SUVmax for osseous metastases on PET. Results 95 patients met inclusion criteria (74 IDC, 13 ILC, and 8 MDL). ILC osseous metastases were more commonly sclerotic and demonstrated lower SUVmax than IDC metastases. For all IDC and MDL patients with osseous metastases, at least one was FDG-avid. For ILC, all patients with lytic or mixed osseous metastases demonstrated at least one FDG-avid metastasis; however, only 3 of 7 patients with sclerotic osseous metastases were apparent on FDG PET. Conclusions The histologic subtype of breast cancer affects the appearance of untreated osseous metastases on FDG PET/CT. In particular, non-FDG-avid sclerotic osseous metastases were more common in patients with ILC, than in those with IDC. Breast cancer histology should be considered when interpreting non-FDG-avid sclerotic osseous lesions on PET/CT, which may be more suspicious for metastases (rather than benign lesions) in patients with ILC.

Background: For breast cancer patients undergoing neoadjuvant chemotherapy (NAC), pathologic complete response (pCR; no invasive or in situ) cannot be assessed non-invasively so all patients undergo surgery. The aim of our study was to develop and validate a radiomics classifier that classifies breast cancer pCR post-NAC on MRI prior to surgery. Methods: This retrospective study included women treated with NAC for breast cancer from 2014 to 2016 with (1) pre-and post-NAC breast MRI and (2) post-NAC surgical pathology report assessing response. Automated radiomics analysis of pre-and post-NAC breast MRI involved image segmentation, radiomics feature extraction, feature prefiltering, and classifier building through recursive feature elimination random forest (RFE-RF) machine learning. The RFE-RF classifier was trained with nested five-fold cross-validation using (a) radiomics only (model 1) and (b) radiomics and molecular subtype (model 2). Class imbalance was addressed using the synthetic minority oversampling technique. Results: Two hundred seventy-three women with 278 invasive breast cancers were included; the training set consisted of 222 cancers (61 pCR, 161 no-pCR; mean age 51.8 years, SD 11.8), and the independent test set consisted of 56 cancers (13 pCR, 43 no-pCR; mean age 51.3 years, SD 11.8). There was no significant difference in pCR or molecular subtype between the training and test sets. Model 1 achieved a cross-validation AUROC of 0.72 (95% CI 0.64, 0.79) and a similarly accurate (P = 0.1) AUROC of 0.83 (95% CI 0.71, 0.94) in both the training and test sets. Model 2 achieved a cross-validation AUROC of 0.80 (95% CI 0.72, 0.87) and a similar (P = 0.9) AUROC of 0.78 (95% CI 0.62, 0.94) in both the training and test sets. Conclusions: This study validated a radiomics classifier combining radiomics with molecular subtypes that accurately classifies pCR on MRI post-NAC.

We present a segmentation approach that combines GrowCut (GC) with cancer-specific multiparametric Gaussian Mixture Model (GCGMM) to produce accurate and reproducible segmentations. We evaluated GCGMM using a retrospectively collected 75 invasive ductal carcinoma with ERPR+ HER2− (n = 15), triple negative (TN) (n = 9), and ER-HER2+ (n = 57) cancers with variable presentation (mass and non-mass enhancement) and background parenchymal enhancement (mild and marked). Expert delineated manual contours were used to assess the segmentation performance using Dice coefficient (DSC), mean surface distance (mSD), Hausdorff distance, and volume ratio (VR). GCGMM segmentations were significantly more accurate than GrowCut (GC) and fuzzy c-means clustering (FCM). Breast cancer is one of the most commonly diagnosed cancers in women and the second most common cause of cancer-related deaths 1 . Although the increasing availability of novel treatment options has helped to improve survival among patients, robust tools are critically needed to effectively monitor treatment response 2 . Miranikova et al. 3 have shown that tumour volumes measured on magnetic resonance imaging (MRI) predict treatment response in neoadjuvant settings. However, accurate and reproducible tumour segmentation is crucial for evaluating breast cancer response to treatments 4 and to improve surgical outcomes 5 . Accurate and reasonably fast segmentation is critical for radiomics analysis 6 which consists of extracting image features from large datasets with the purpose of identifying non-invasive image-based surrogates for diagnosis (differentiating disease aggressiveness) and for predicting treatment response. Radiomics analysis of breast cancers have been used for predicting cancer treatment outcomes 7-9 and for differentiating between breast cancers by molecular subytpe 10-13 or for classifying cancers by their aggressiveness 14,15 . The first and crucial step in extracting the various texture measures is segmentation of the cancer. With the exception of 11,15 , the vast majority of works have employed manual tumour segmentation for radiomics analysis due to the difficultly in ensuring accurate computer segmentations. However, manual delineation is time consuming. Therefore, majority of works 12-14 including ours 10,16 have used manual segmentation of one or a few representative slices. Recently, semi-automatic segmentations including GrowCut (GC) 17 have been reported to produce more reproducible texture features compared with features computed from manually delineated lung tumors 18 , thereby, underscoring the importance and utility of computer-generated segmentations for high-throughput radiomics.

Terms of use and reuse: academic research for non-commercial purposes, see here for full terms. http://www.springer.com/gb/openaccess/authors-rights/aam-terms-v1

Pathologic evaluation of breast specimens requires a fixation and staining procedure of at least 12 hours duration, delaying diagnosis and post-operative planning. Here we introduce an MRI technique with a custom-designed radiofrequency resonator for imaging breast and lymph tissue with sufficient spatial resolution and speed to guide pathologic interpretation and offer value in clinical decision making. In this study, we demonstrate the ability to image breast and lymphatic tissue using 7.0 Tesla MRI, achieving a spatial resolution of 59 × 59 × 94 μm3 with a signal-to-noise ratio of 15–20, in an imaging time of 56 to 70 minutes. These are the first MR images to reveal characteristic pathologic features of both benign and malignant breast and lymph tissue, some of which were discernible by blinded pathologists who had no prior training in high resolution MRI interpretation.