Overdiagnosis in breast cancer screening may be reduced by identifying lesions that, although detected on screening mammograms, are unlikely to progress to poor outcomes and may not require recall. As a proof-of-concept, we evaluated prognostic models for 10-year distant recurrence-free survival (DRFS) using radiomics features from invasive breast cancers (IBCs) presenting as masses on screening mammograms. In a cohort of screened women, 1466 IBC patients with clinical, pathological, and follow-up data were identified through national registries. Using radiomics features, tumor volume, and specific growth rate, proportional hazards models were developed to predict 10-year distant recurrence risk. Models were trained using positive screening mammograms of patients with screen-detected IBC (n = 1063) and diagnostic mammograms of patients with interval cancer (n = 406). Performance was evaluated only in screen-detected IBCs using repeated nested cross-validation. Median follow-up was 5.1 years (10th-90th percentile: 2.1-10.1), with 111 distant recurrences within 10 years. Model performance was moderate (C-index 0.70 [SD 0.01], calibration slope 1.22 [SD 0.13]), with predicted 10-year recurrence risks ranging from 4.9% to 18.0% (10th-90th percentile). The 42 IBCs with the lowest predicted risk had a 10-year recurrence probability of 2.6% (95% CI 0-7.5%), of whom only five received adjuvant systemic therapy. These findings suggest that mammography-based radiomics features may help identify low-risk IBCs and potentially reduce overdiagnosis by reconsidering recall for selected cases.

Understanding how lifestyle habits influence breast cancer survival is crucial for improving long-term outcomes and guiding individualized care. The purpose of this study is to comprehensively understand how dietary habits, exercise frequency, sleep quality and the frequency of tobacco and alcohol use affect the survival of breast cancer patients by analyzing the synergistic effects of multiple lifestyle habits. We analyzed data from 21,219 female breast cancer patients in the UK Biobank, a large-scale population-based cohort. Three feature selection methods were applied to identify key prognostic variables. Five survival risk models were compared to investigate the relationship between breast cancer survival and lifestyle habits. Kaplan-Meier survival analysis and SHapley Additive exPlanations (SHAP) interpretability analysis were performed on risk scores calculated from the best-performing model. Through feature selection, we identified that age, Body Mass Index (BMI), smoking status, and treatment pill intake were pinpointed as critical characteristics influencing the survival of breast cancer patients. Among all models, the eXtreme Gradient Boosting (XGBoost) model demonstrated the highest predictive performance (3 years: AUC = 0.748, 6 years: AUC = 0.749, 9 years: AUC = 0.765). Kaplan-Meier analysis showed that patients classified as high-risk based on the median risk score had significantly worse survival outcomes (P < 0.0001). SHAP analysis further confirmed the dominant influence of key characteristics on mortality risk. This study highlights the prognostic value of lifestyle habits in breast cancer survival. By integrating routine health indicators into interpretable machine learning models, our findings provide a practical foundation for individualized risk assessment and lifestyle-based intervention strategies.

To investigate the influence of gadolinium-based contrast agent (GBCA) administration on breast lesions’ signal intensities on diffusion-weighted images and its derived ADC maps acquired 2 min after GBCA administration. An intra-patient analysis was performed. 110 lesions > 5 mm in 93 patients, visible on screening and pre-operative MRI scans at 1.5 T and 3.0 T, were included. Pre-contrast and post-contrast DWI were acquired with single-shot spin-echo echo-planar imaging (SSh-SE-EPI) with b0, b50, b150, and b800 diffusion weightings. 2D regions of interest were drawn. Each b value was separately analyzed regarding signal intensities on DWI. Furthermore, ADC values were compared based on the pairs b0–b800 and b150–b800. The comparison between pre- and post-contrast DWIs was investigated within and between different groups (overall, malignant/benign, and at 1.5/3.0 T). The Wilcoxon-signed rank test and independent Student T test were used when appropriate. P < 0.01 was considered as statistically significant. Overall, and for malignant and benign lesions separately, a significant change after contrast administration in signal intensity was observed at each specific b value, irrespective of the use of 1.5 T or 3.0 T. The effect of contrast administration was overall notably different when comparing b0 (+ 9.4

BACKGROUND:Neoadjuvant treatment comprising six to nine cycles of anti-HER2 based chemotherapy yields high pathological complete response rates and excellent survival outcomes in patients with stage II-III HER2-positive breast cancer but comes with noteworthy side-effects. The TRAIN-3 study investigated if patients treated with this regimen who have a rapid complete radiological response on MRI are candidates for early surgery. The study showed that one-third of patients with hormone receptor-negative tumours and one-sixth of patients with hormone receptor-positive tumours had a pathological complete response after only three cycles of chemotherapy. Here, we report the results of the primary endpoint, 3-year event-free survival. METHODS:TRAIN-3 is a multicentre, single-arm, phase 2 study conducted across 43 hospitals in the Netherlands. Patients with stage II-III HER2-positive breast cancer aged 18 years or older and with a WHO performance status of 0 or 1 were eligible for inclusion. Patients received neoadjuvant chemotherapy consisting of paclitaxel (80 mg/m2, day 1 and 8 of each 21-day cycle, intravenous), trastuzumab (6 mg/kg on day 1 of each cycle [loading dose 8 mg/kg on day 1 of cycle 1], intravenous), carboplatin (area under the curve 6 mg/mL per min on day 1 of each cycle, intravenous) and pertuzumab (420 mg on day 1 of each cycle [loading dose 840 mg on day 1 of cycle 1], intravenous) for up to nine cycles, and were referred for surgery once a complete radiological response was observed on MRI. Patients who had a pathological complete response post-surgery completed 1 year of adjuvant trastuzumab and pertuzumab. Patients with residual invasive disease in the resection specimen continued chemotherapy for a total of nine cycles, followed by 14 cycles of trastuzumab emtansine (3·6 kg/mg, day 1 of each 21-day cycle, intravenous). Adverse events of grade 3 or worse and grade 2 or worse left ventricular ejection fraction and neuropathy were registered until 30 days after the last adjuvant treatment cycle. The primary endpoint was 3-year event-free survival; analyses were done in hormone receptor-negative and hormone receptor-positive patients separately on an intention-to-treat basis. This trial is registered with ClinicalTrials.gov (NCT03820063); recruitment is closed, and follow-up for secondary endpoints is ongoing. FINDINGS:Between April 1, 2019, and May 12, 2021, 235 patients with hormone receptor-negative tumours and 232 with hormone receptor-positive tumours were enrolled. All participants, except one, were female. Median follow-up was 40·1 months (IQR 35·3-45·6). 3-year event-free survival was 92·2% (95% CI 88·7-95·9) in patients with hormone receptor-negative tumours and 92·0% (88·5-95·6) in patients with hormone receptor-positive tumours. In total, 40 (9%) patients had experienced an event (19 [8%] patients in the hormone receptor-negative group and 21 [9%] in the hormone receptor-positive group). 3-year event-free survival rates among patients treated with one to three cycles were 96·1% (95% CI 91·8-100) in the hormone receptor-negative group and 98·6% (95·8-100) in the hormone receptor-positive group. Event-free survival rates for patients receiving four to six cycles were 89·2% (82·4-96·6) and 94·2% (88·8-99·9), and for those receiving seven to nine cycles, 90·6% (83·8-98·1) and 85·4% (78·3-93·1%), respectively. The most common grade 3-4 adverse events were neutropenia (187 [40%] of 467), anaemia (82 [18%]), diarrhoea (60 [13%]), thrombocytopenia (46 [10%]), and hypokalaemia (35 [7%]), and treatment-related serious adverse events occurred in 56 (12%) patients. The frequency of grade 3-4 adverse events and grade 2 or worse neuropathy increased with the increasing number of neoadjuvant chemotherapy cycles. No treatment-related deaths were reported. INTERPRETATION:MRI-guided optimisation of neoadjuvant chemotherapy duration was associated with favourable 3-year event-free survival outcomes in patients with stage II-III HER2-positive breast cancer. This approach represents a novel strategy that reduces treatment burden, minimises toxicity, and preserves quality of life in a subset of patients with early HER2-positive breast cancer. FUNDING:Roche Netherlands.

Background: Understanding how lifestyle habits influence breast cancer survival is crucial for improving long-term outcomes and guiding individualized care. The purpose of this study is to comprehensively understand how dietary habits, exercise frequency, sleep quality and the frequency of tobacco and alcohol use affect the survival of breast cancer patients by analyzing the synergistic effects of multiple lifestyle habits. Methods: We analyzed data from 14,901 female breast cancer patients in the UK Biobank, a large-scale population-based cohort. Three feature selection methods were applied to identify key prognostic variables. Five survival risk models were compared to investigate the relationship between breast cancer survival and lifestyle habits. Kaplan-Meier survival analysis and SHapley Additive exPlanations (SHAP) interpretability analysis were performed on risk scores calculated from the best-performing model. Results: Through feature selection, we identified that age, number of medications taken, Body Mass Index (BMI), alcohol intake, diastolic blood pressure and smoking status were pinpointed as critical characteristics influencing the survival of breast cancer patients. Among all models, the eXtreme Gradient Boosting (XGBoost) model demonstrated the highest predictive performance (3 years: AUC = 0.72, 6 years: AUC = 0.727, 9 years: AUC = 0.749). Kaplan-Meier analysis showed that patients classified as high-risk based on the median risk score had significantly worse survival outcomes ( P < 0.0001). SHAP analysis further confirmed the dominant influence of age and BMI on mortality risk. Conclusions: This study highlights the prognostic value of lifestyle habits in breast cancer survival. By integrating routine health indicators into interpretable machine learning models, our findings provide a practical foundation for individualized risk assessment and lifestyle-based intervention strategies.

BACKGROUND:Anti-HER2 based (neo)adjuvant chemotherapy has substantially improved outcomes for patients with stage II-III HER2-positive breast cancer but often leads to long-term toxicities that impair survivors' health-related quality of life (HRQOL). The TRAIN-3 study demonstrated that shortening neoadjuvant chemotherapy based on early complete response on breast MRI preserves efficacy. The impact of chemotherapy duration on HRQOL remains unclear. Here, we report the results from the pre-planned HRQOL analyses. METHODS:HRQOL was assessed at baseline, and at 6, 12 and 24 months using the EORTC QLQ-C30 and QLQ-BR45 questionnaires. Patients with baseline and at least one follow-up assessment were included for this analysis. RESULTS:Ninety patients were evaluable: 29 (32%) received 1-3 cycles, 26 (29%) received 4-6 cycles, and 35 (39%) received 7-9 cycles of neoadjuvant chemotherapy. Patients receiving 1-3 cycles reported consistently better preservation of physical and role functioning, less fatigue, and lower neuropathy scores compared to those receiving longer chemotherapy. At 24 months, clinically meaningful deterioration in global health status was observed in 26%, 55% and 46% of patients in the 1-3, 4-6 and 7-9 cycle groups, respectively. Greater neuropathy was associated with worse global HRQOL at both 12 and 24 months. CONCLUSION:MRI-guided shorter neoadjuvant chemotherapy duration is associated with less pronounced declines and better long-term preservation of HRQOL in patients with early HER2-positive breast cancer. These findings support MRI-guided treatment de-escalation as a strategy to reduce survivorship burden without compromising efficacy.

Abstract Objectives Gender equity in medicine remains a topic of increasing attention. The aim was to investigate if gender influences the radiology profession, with a focus on career progression, leadership roles, work-life balance, research activity and perceived barriers. Materials and methods An anonymous online survey consisting of 22 questions was distributed by the European Society of Radiology (ESR) to its members between October and December 2024. The survey covered demographics, work schedules, family responsibilities, career development, leadership roles, research involvement, and perceived personal experiences. Quantitative data were analyzed using descriptive statistics, chi-square test, and rate differences with confidence intervals. Open-ended responses were explored qualitatively using thematic analysis. Results Among 830 respondents, 657 completed the questionnaire (63.3% female, 35.3% male, 1.3% others). Women more frequently reported caregiving responsibilities beyond childcare (4.1% vs 3%), longer parental leave (46.2% vs 21.5%), and experiences of harassing behaviors at work. Men held a higher proportion of leadership roles (33.2% vs 25.2%). Respondents involved in research were more likely to work > 30% extra hours (47.2% vs 29.0%). Although research activity rates were similar across genders, women more often reported barriers to attending conferences and a lack of protected research time. Career fulfillment increased with age among men but decreased among women. Gender was considered a career disadvantage by 44.5% of women versus 9.5% of men. Conclusion The survey reveals perceived gender disparities in radiology, particularly in leadership access, work conditions, and career satisfaction. Addressing structural barriers and promoting supportive workplace policies are essential to achieving true gender equity in the field. Critical relevance statement Despite improvements in the last few decades, gender inequity remains present in radiology. Variability between geographical regions suggests that key critical areas can be addressed to promote improvement and support a more equitable professional environment. Key Points Perceived gender disparities in radiology are present across career progression, leadership roles, and work-life balance. Women were significantly more likely than men to perceive gender as a career disadvantage (44.5% vs 9.5%; p < 0.0001 35.0% [95% CI: 28.9 to 41.1]). They reported slightly higher caregiving responsibilities and longer parental leave. Structural inequalities impact gender equity in radiology, requiring targeted institutional and cultural changes. Graphical Abstract

Multidisciplinary team meetings (MDTMs) in breast cancer care improve outcomes but are time-consuming and costly. This study investigates using data from the Dutch national cancer registry (NCR) and hospital electronic medical records (EMR) to efficiently calculate MDTM discussion durations, while complying with privacy laws. This retrospective study analyzed breast cancer MDTM discussion durations using NCR and EMR data from 2014 to 2018. Discussion times were estimated from EMR timestamps, excluding outliers, and analyzed statistically. Ethical approval was obtained, including only non-opt-out patients. The final dataset included 1,048 tumors, 996 patients, and 1,487 MDTM discussion times after exclusions for missing data, duplicates, and outliers. Discussion durations varied significantly, with pre-operative and no-surgery discussions being longer than post-operative ones (p = 0.000), and factors such as MRI availability, tumor differentiation, malignancy, menopausal status, and cancer stage influencing duration in pre-operative cases. In post-operative discussions, molecular subtype, cancer stage, and tumor size were significant, while age, tumor differentiation, and menopausal status had no impact. This study evaluates discussion durations in breast cancer MDTMs using retrospective data from the NCR and EMR, demonstrating a feasible approach to assess MDTM functioning. Differences in discussion duration based on patient and tumor characteristics may help optimize MDTM efficiency. Not applicable.

To investigate the surgical impact of preoperative breast MRI in patients diagnosed with invasive lobular breast cancer (ILC) in a prospective observational study. The prospective MIPA observational study database was queried for patients aged 18–80 with newly diagnosed unilateral ILC at needle biopsy referred for primary surgery. Patients who underwent preoperative MRI (MRI group) were matched (1:1) with those who did not (noMRI group) according to nine confounding covariates. Surgical outcomes were compared between the matched groups with nonparametric statistics after calculating odds ratios (ORs). A total of 547 women with unilateral needle biopsy-diagnosed ILC were identified (158 noMRI group, 389 MRI group). After patient matching, each group retained 103 patients, for a total of 206 matched patients. For the rate of women having a first-line mastectomy, there was no significant difference between the MRI group (21.4

Longitudinal medical imaging is crucial for monitoring neoadjuvant therapy (NAT) response in clinical practice. However, mainstream artificial intelligence (AI) methods for disease monitoring commonly rely on extensive segmentation labels to evaluate lesion progression. While self-supervised vision-language (VL) learning efficiently captures medical knowledge from radiology reports, existing methods focus on single time points, missing opportunities to leverage temporal self-supervision for disease progression tracking. In addition, extracting dynamic progression from longitudinal unannotated images with corresponding textual data poses challenges. In this work, we explicitly account for longitudinal NAT examinations and accompanying reports, encompassing scans before NAT and follow-up scans during mid-/post-NAT. We introduce the multi-modal longitudinal representation learning pipeline (MLRL), a temporal foundation model, that employs multi-scale self-supervision scheme, including single-time scale vision-text alignment (VTA) learning and multi-time scale visual/textual progress (TVP/TTP) learning to extract temporal representations from each modality, thereby facilitates the downstream evaluation of tumor progress. Our method is evaluated against several state-of-the-art self-supervised longitudinal learning and multi-modal VL methods. Results from internal and external datasets demonstrate that our approach not only enhances label efficiency across the zero-, few- and full-shot regime experiments but also significantly improves tumor response prediction in diverse treatment scenarios. Furthermore, MLRL enables interpretable visual tracking of progressive areas in temporal examinations, offering insights into longitudinal VL foundation tools and potentially facilitating the temporal clinical decision-making process.

BACKGROUND:We studied changes in mammographic density (MD) among premenopausal women with a pathogenic germline variant (PGV) in the BRCA1 or BRCA2 gene, comparing those who did and did not undergo risk-reducing salpingo-oophorectomy (RRSO) in the interval between mammograms, accounting for changes in exogenous oral contraceptive or hormone replacement therapy (HRT) use. METHODS:From five studies of the International BRCA1/2 Carrier Cohort Study consortium, we included 691 participants who had two or more screening mammograms available, were less than 47 years at the time of RRSO (N = 208), or premenopausal at all mammograms without RRSO (N = 483). MD metrics [percent density (PD), dense area (DA), and non-DA] were quantified using STRATUS. Multivariable linear mixed models assessed changes in MD metrics between groups, adjusting for confounders. RESULTS:The mean PD at first mammogram was 26.8% ± 15.3 (RRSO) and 31.3% ± 18.1 (no RRSO). In a median 1.1 years between mammograms, PD decreased on average by 0.9% [95% confidence interval (CI), -1.6 to -0.2] among women who did not undergo RRSO in the interval between mammograms compared with 5.9% (95% CI, -7.4 to -4.5) among women who underwent RRSO in the interval (adjusted difference, -5.9%; 95% CI, -9.5 to -2.2; P = 0.002). Results were driven primarily by MD changes among BRCA2 PGV carriers. The use of HRT after RRSO attenuated the decline in PD. CONCLUSIONS:On average, PD and DA decrease following RRSO in premenopausal carriers, particularly among BRCA2 PGV carriers. HRT formulation affects MD changes. IMPACT:A decrease in MD may inform the potential protective effect of RRSO against breast cancer.

Breast magnetic resonance imaging (MRI) is widely recognized for its high sensitivity in detecting breast cancer. However, interpreting breast MRI scans remains a complex, time-consuming, and resource-intensive task, even for experienced radiologists. To address these challenges, artificial intelligence-based methods are increasingly being employed. In this study, we developed a multimodal breast MRI language-image pre-training (MLIP) approach as an initial exploration of a breast MRI foundation model to aid in the interpretation of scans. Two types of inferences were used to evaluate MLIP's performance. First, MLIP could retrieve corresponding MRI cases from a dataset based on a query, achieving an area under the receiver operating characteristic curve of 0.717 for suspicious and malignant cases, 0.640 for dense breasts, and 0.601 for low background parenchymal enhancement (BPE). Second, MLIP demonstrated the ability to predict the level of disease suspicion for a given MRI case. The results suggest that MLIP has the potential to serve as a foundation model for breast MRI interpretation. Future work will focus on expanding its capabilities through various downstream tasks and integrating additional models to enhance overall performance.

Abstract Objectives The breast ultrasound trial (BUST) demonstrates a high negative predictive value of ultrasound for women presenting with focal breast complaints, suggesting a potential shift from mammography/digital breast tomosynthesis (DBT) to ultrasound as a primary imaging modality. This BUST side-study explored women’s perspectives on adopting ultrasound as the primary diagnostic tool. Methods Twenty-nine female BUST participants (mean age = 48.4, SD = 8.3) with benign findings after diagnostic evaluation participated in one of six focus group interviews 18–24 months post-imaging. Discussions were transcribed and analyzed thematically. Results Four overarching themes were identified; personal health situation, organization of breast care, effectiveness of imaging, and professionals’ attitudes and roles. Participants considered their own health history and complaint type (personal health situation) and discussed eligibility for national screening programs and the costs of both exams (organization of breast care). Opinions varied on the effectiveness of imaging, particularly regarding the importance of detecting additional non-symptomatic findings with mammography/DBT that may be missed by ultrasound. Concerns were also raised about implementing research findings without conclusive scientific evidence. Health professionals’ attitudes and roles encompassed the influence of GPs’ and radiologists’ attitudes and the process of image interpretation. Conclusion These findings reveal diverse attitudes of women towards ultrasound as a primary modality, warranting caution when transitioning to new clinical standards. Providing comprehensive information about the evidence on the benefits and risks of different imaging modalities is essential, and fostering shared decision-making could enhance acceptance. Offering women the choice of additional imaging, such as mammography/DBT after an initial ultrasound, may balance clinical performance with patient autonomy. Critical relevance statement Patients’ perspectives on medical procedures are increasingly significant in modern healthcare. Women’s perceived barriers and facilitators to diagnostic imaging, shaped by numerous factors, offer healthcare professionals insights beyond pure biomedical approaches, fostering shared decision-making within radiology and other clinical contexts. Key Points This study explores women’s perspectives on breast ultrasound as a primary imaging modality. Attitudes towards an ultrasound-first approach are shaped by internal and external considerations. Considerations are highly informed by emotional responses and lack of knowledge. Insights into women’s perspectives inform healthcare professionals and foster shared decision-making within radiology. Graphical Abstract