The purpose of this multicenter retrospective study was to evaluate radiomics analysis coupled with machine learning (ML) of dynamic contrast-enhanced (DCE) and diffusion-weighted imaging (DWI) radiomics models separately and combined as multiparametric MRI for improved breast cancer detection. Consecutive patients (Memorial Sloan Kettering Cancer Center, January 2018–March 2020; Medical University Vienna, from January 2011–August 2014) with a suspicious enhancing breast tumor on breast MRI categorized as BI-RADS 4 and who subsequently underwent image-guided biopsy were included. In 93 patients (mean age: 49 years ± 12 years; 100% women), there were 104 lesions (mean size: 22.8 mm; range: 7–99 mm), 46 malignant and 58 benign. Radiomics features were calculated. Subsequently, the five most significant features were fitted into multivariable modeling to produce a robust ML model for discriminating between benign and malignant lesions. A medium Gaussian support vector machine (SVM) model with five-fold cross validation was developed for each modality. A model based on DWI-extracted features achieved an AUC of 0.79 (95% CI: 0.70–0.88), whereas a model based on DCE-extracted features yielded an AUC of 0.83 (95% CI: 0.75–0.91). A multiparametric radiomics model combining DCE- and DWI-extracted features showed the best AUC (0.85; 95% CI: 0.77–0.92) and diagnostic accuracy (81.7%; 95% CI: 73.0–88.6). In conclusion, radiomics analysis coupled with ML of multiparametric MRI allows an improved evaluation of suspicious enhancing breast tumors recommended for biopsy on clinical breast MRI, facilitating accurate breast cancer diagnosis while reducing unnecessary benign breast biopsies.

Natural language processing can extract data from structured radiology reports to identify varying patterns of metastatic disease spread across different primary cancers. Background Patterns of metastasis in cancer are increasingly relevant to prognostication and treatment planning but have historically been documented by means of autopsy series. Purpose To show the...

To investigate the value of contrast-enhanced mammography (CEM) compared to full-field digital mammography (FFDM) in screening breast cancer patients after breast-conserving surgery (BCS), this Health Insurance Portability and Accountability Act-compliant, institutional review board-approved retrospective, single-institution study included 971 CEM exams in 541 asymptomatic patients treated with BCS who underwent screening CEM between January 2013 and November 2018. Histopathology, or at least a one-year follow-up, was used as the standard of reference. Twenty-one of 541 patients (3.9%) were diagnosed with ipsi- or contralateral breast cancer: six (28.6%) cancers were seen with low-energy images (equivalent to FFDM), an additional nine (42.9%) cancers were detected only on iodine (contrast-enhanced) images, and six interval cancers were identified within 365 days of a negative screening CEM. Of the 10 ipsilateral cancers detected on CEM, four were detected on low-energy images (40%). Of the five contralateral cancers detected on CEM, two were detected on low-energy images (40%). Overall, the cancer detection rate (CDR) for CEM was 15.4/1000 (15/971), and the positive predictive value (PPV3) of the biopsies performed was 42.9% (15/35). For findings seen on low-energy images, with or without contrast, the CDR was 6.2/1000 (6/971), and the PPV3 of the biopsies performed was 37.5% (6/16). In the post-BCS screening setting, CEM has a higher CDR than FFDM.

Compared with low-energy imaging, contrast-enhanced mammography depicted 11 additional cancers in asymptomatic women with extremely dense breasts and showed higher sensitivity but lower specificity, with specificity improving at follow-up examinations.

The purpose of this retrospective study was to assess whether radiomics analysis coupled with machine learning (ML) based on standard-of-care dynamic contrast-enhanced magnetic resonance imaging (DCE-MRI) can predict PD-L1 expression status in patients with triple negative breast cancer, and to compare the performance of this approach with radiologist review. Patients with biopsy-proven triple negative breast cancer who underwent pre-treatment breast MRI and whose PD-L1 status was available were included. Following 3D tumor segmentation and extraction of radiomic features, radiomic features with significant differences between PD-L1+ and PD-L1− patients were determined, and a final predictive model to predict PD-L1 status was developed using a coarse decision tree and five-fold cross-validation. Separately, all lesions were qualitatively assessed by two radiologists independently according to the BI-RADS lexicon. Of 62 women (mean age 47, range 31–81), 27 had PD-L1− tumors and 35 had PD-L1+ tumors. The final radiomics model to predict PD-L1 status utilized three MRI parameters, i.e., variance (FO), run length variance (RLM), and large zone low grey level emphasis (LZLGLE), for a sensitivity of 90.7%, specificity of 85.1%, and diagnostic accuracy of 88.2%. There were no significant associations between qualitative assessed DCE-MRI imaging features and PD-L1 status. Thus, radiomics analysis coupled with ML based on standard-of-care DCE-MRI is a promising approach to derive prognostic and predictive information and to select patients who could benefit from anti-PD-1/PD-L1 treatment.

BRCA1 and BRCA2 pathogenic variant carriers develop breast cancers with distinct pathological characteristics and mutational signatures that may result in differential response to chemotherapy. We compared rates of pathologic complete response (pCR) after NAC between BRCA1/2 variant carriers and noncarriers in a cohort of 1426 women (92 [6.5%] BRCA1 and 73 [5.1%] BRCA2) with clinical stage I–III breast cancer treated with NAC followed by surgery from 11/2013 to 01/2022 at Memorial Sloan Kettering Cancer Center. The majority received doxorubicin/cyclophosphamide/paclitaxel therapy (93%); BRCA1/2 carriers were more likely to receive carboplatin (p < 0.001). Overall, pCR was achieved in 42% of BRCA1 carriers, 21% of BRCA2 carriers, and 26% of noncarriers (p = 0.001). Among clinically node-positive (cN+) patients, nodal pCR was more frequent in BRCA1/2 carriers compared to noncarriers (53/96 [55%] vs. 371/856 [43%], p = 0.015). This difference was seen in HR+/HER2− (36% vs. 20% of noncarriers; p = 0.027) and TN subtypes (79% vs. 45% of noncarriers; p < 0.001). In a multivariable analysis of the overall cohort, BRCA1 status, and TN and HER2+ subtypes were independently associated with pCR. These data indicate that BRCA1 carriers may be more likely to achieve overall and nodal pCR in response to NAC compared with BRCA2 carriers and patients with sporadic disease. Further studies with a larger cohort of BRCA1/2 mutation carriers are needed, as a small sample size may have a restricted ability to detect a significant association between mutational status and pCR in sensitivity analyses stratified by subtype and adjusted for clinically relevant factors.