British Journal of Cancer最新文献

{"title":"Lipidomic profiling of human bile distinguishes cholangiocarcinoma from benign bile duct diseases with high specificity and sensitivity: a prospective descriptive study.","authors":"Fu-Sheng Liu, Ying-Yi Liu, Shi-Kun Zhang, Jun-Yu Zhou, Jing-Hua Li, Xiao-Mian Li, Ming-He Zhang, Xiao-Yu Pan, Yi-Bo Chai, Wei-Xian Fang, Tao Yuan, Xu-Yun Yan, Xi Chen, Tian-Gen Wu, Wei-Jie Ma, Bo Liao, Ping Jiang, Wei-Hua Huang, Song-Mei Liu, Shan Guo, Yu-Feng Yuan","doi":"10.1038/s41416-025-03144-9","DOIUrl":"10.1038/s41416-025-03144-9","url":null,"abstract":"<p><strong>Background: </strong>Cholangiocarcinoma (CCA) is a rare and highly aggressive malignancy originating in the bile ducts. Owing to limitations involving pathological sampling, the clinical differentiation of CCA from benign biliary diseases remains challenging. This study aimed to evaluate the differences between the bile lipidomes of CCA patients and those of patients with benign disease to develop a bile lipid classifier that can help to differentiate CCA from benign conditions.</p><p><strong>Methods: </strong>Bile samples were collected by endoscopic retrograde cholangiography (ERCP) from patients with CCA or benign disease. The participants were divided into three cohorts: the first two cohorts underwent untargeted lipidomic analysis, whereas the third cohort was subjected to targeted lipid quantification. Untargeted lipidomic analysis was performed via ultrahigh-performance liquid chromatography coupled with ion mobility quadrupole time-of-flight mass spectrometry (UHPLC/IM-QTOF-MS). Targeted lipid quantification was conducted via UHPLC‒MS/MS in multiple reaction monitoring (MRM) mode. Lipid features were screened to construct a bile lipid classifier using the machine learning algorithm, least absolute shrinkage and selection operator (LASSO) regression, followed by cross-validation in two cohorts. The selected lipid features were further validated by targeted quantification in the third cohort. The functions of the significantly differentially abundant lipids in proliferation were validated in CCA cell lines.</p><p><strong>Results: </strong>In total, 241 bile samples were collected and divided into three cohorts for independent lipidomic analysis: Cohort 1 included 32 CCA samples and 68 benign controls; Cohort 2 included 30 CCA samples and 30 benign controls; and Cohort 3 included 32 CCA samples and 49 benign controls. There were significant differences in the lipid profiles of the bile samples obtained from patients with CCA and individuals with benign disease, with multiple lipid classes, particularly lysophosphatidylcholine (LPC), significantly downregulated in the CCA group. Multimodule correlation networks constructed via weighted lipid coexpression network analysis (WLCNA) revealed significant associations between lipid modules and clinical traits. A machine learning-based bile lipid classifier, termed BileLipid, was developed for CCA diagnosis; this classifier incorporates six lipid features. This classifier achieved areas under the receiver operating characteristic curve (AUCs) of 0.943, 0.956, and 0.828 in Cohorts 1, 2, and 3, respectively. Additionally, the significantly downregulated lipid LPC in CCA bile was found to significantly inhibit the proliferation of CCA cell lines, suggesting its potential role as a protective factor in CCA.</p><p><strong>Conclusions: </strong>This study not only identified lipidomic alterations in CCA using bile samples but also established and validated a sex-related bile lipid classifier wi","PeriodicalId":9243,"journal":{"name":"British Journal of Cancer","volume":" ","pages":""},"PeriodicalIF":6.8,"publicationDate":"2025-09-02","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144943660","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Correction: Wnt/ERK/CDK4/6 activation in the partial EMT state coordinates mammary cancer stemness with self-renewal and inhibition of differentiation","authors":"Huizhi Liang, Outhiriaradjou Benard, Viney Kumar, Anthony Griffen, Zuen Ren, Kalaiselvi Sivalingam, Jingli Wang, Elena de Simone Benito, Xusheng Zhang, Jinghang Zhang, Kimita Suyama, Lindsay M. LaFave, Larry Norton, Rachel B. Hazan","doi":"10.1038/s41416-025-03178-z","DOIUrl":"10.1038/s41416-025-03178-z","url":null,"abstract":"","PeriodicalId":9243,"journal":{"name":"British Journal of Cancer","volume":"133 7","pages":"1066-1066"},"PeriodicalIF":6.8,"publicationDate":"2025-09-02","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.nature.comhttps://www.nature.com/articles/s41416-025-03178-z.pdf","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144943601","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Systemic activation of NRF2 contributes to the therapeutic efficacy of clinically-approved KRAS-G12C anti-cancer drugs.","authors":"Liam Baird, Lin Zhang, Takanori Hidaka, Lyu Xi, Ke Wang, Keiko Tateno, Tatsuro Iso, Takafumi Suzuki, Kazuki Kumada, Fumiki Katsuoka, Kengo Kinoshita, Masayuki Yamamoto","doi":"10.1038/s41416-025-03162-7","DOIUrl":"10.1038/s41416-025-03162-7","url":null,"abstract":"<p><strong>Background: </strong>The development and clinical success of KRAS^G12C inhibitors was a landmark achievement in anti-cancer drug development, as oncogenic KRAS had long been considered an intractable therapeutic target. Patients with KRAS mutant lung cancers frequently present with co-mutations in the KEAP1-NRF2 pathway, and because genetic activation of NRF2 results in resistance to all current anti-cancer therapies, we were motivated to explore how aberrant activation of NRF2 impacts the clinical response to KRAS^G12C inhibitors.</p><p><strong>Methods: </strong>A broad range of techniques, including genetic knockouts, scRNA-seq and surface plasmon resonance, were used to determine the effect of KRAS^G12C drugs on NRF2.</p><p><strong>Results: </strong>At physiologically-relevant concentrations, both of the clinically-approved KRAS^G12C inhibitors Sotorasib and Adagrasib also function as inducers of NRF2. Mechanistically, the same cysteine-targeting functionality which allows these electrophilic drugs to inhibit the mutant KRAS^G12C protein also facilitates their binding to cysteine-based sensors in KEAP1, resulting in the upregulation of the NRF2-dependent gene expression program.</p><p><strong>Conclusions: </strong>The activation of NRF2 by KRAS-G12C inhibitors represents a unique example of anti-cancer drugs which positively regulate the activity of a protein which is normally considered to be an oncogene. In both the malignant cells of the tumour and immune cells within the microenvironment, activation of NRF2 by electrophilic KRAS inhibitors positively contributes to the clinical efficacy of these drugs by promoting anti-cancer immunity. This unprecedented situation, in which the NRF2-dependent oxidative stress response is induced globally within cancer patients, has a number of important clinical implications, particularly in relation to ongoing combination chemotherapy clinical trials, as well as for selecting patient populations which may derive the most benefit from G12Ci anti-cancer drugs.</p>","PeriodicalId":9243,"journal":{"name":"British Journal of Cancer","volume":" ","pages":""},"PeriodicalIF":6.8,"publicationDate":"2025-09-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144943451","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Evaluating CAR-T cells from neurofibromatosis type 1 (NF1) patients for targeting AXL in malignant peripheral nerve sheath tumors associated with NF1.","authors":"Po-Yuan Huang, I-An Shih, Ying-Chih Liao, Chien-Ting Lin, Huey-Ling You, Ming-Jen Lee","doi":"10.1038/s41416-025-03151-w","DOIUrl":"10.1038/s41416-025-03151-w","url":null,"abstract":"<p><strong>Background: </strong>Neurofibromatosis type 1 (NF1) is an autosomal dominant disorder characterized by neurofibromas, with 5-13% of patients risk developing malignant peripheral nerve sheath tumors (MPNST). Current treatments for MPNST are largely ineffective. AXL, overexpressed in MPNST, is a potential target for Chimeric Antigen Receptor T (CAR-T) cell therapy. This study evaluates the immunophenotypes, efficacy, and safety of NF1-derived AXL-CAR-T cells in treating MPNST.</p><p><strong>Methods: </strong>AXL-CAR-T cells, containing an anti-AXL single-chain variable fragment, were derived from NF1 patients (n = 27) and healthy donors (n = 15). Immunophenotypes were characterized using CCR7, CD45RA, CD4, and CD8 markers. The cytotoxicity of CAR-T cells was tested in vitro against MPNST cells, and efficacy and safety were evaluated in an MPNST xenograft mouse model. Multiplex immunoassays and ELISA measured cytokines and granzyme b release.</p><p><strong>Results: </strong>AXL-CAR-T cells from both NF1 patients and healthy donors had a similar partition in stem cell-like memory T cell composition and comparable ex vivo expansion. AXL-CAR-T cells from both groups effectively lysed MPNST cells in vitro. In vivo, tumor volumes in xenograft mice were significantly reduced with no on-target off-tumor toxicity.</p><p><strong>Conclusions: </strong>NF1 patient-derived AXL-CAR-T cells demonstrated similar quality and efficacy to those from healthy donors, supporting their potential autologous therapy for MPNST.</p>","PeriodicalId":9243,"journal":{"name":"British Journal of Cancer","volume":" ","pages":""},"PeriodicalIF":6.8,"publicationDate":"2025-09-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144943657","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Correction: RON signalling promotes therapeutic resistance in ESR1 mutant breast cancer","authors":"Derek Dustin,&nbsp;Guowei Gu,&nbsp;Amanda R. Beyer,&nbsp;Sarah K. Herzog,&nbsp;David G. Edwards,&nbsp;Hangqing Lin,&nbsp;Thomas L. Gonzalez,&nbsp;Sandra L. Grimm,&nbsp;Cristian Coarfa,&nbsp;Doug W. Chan,&nbsp;Beom-Jun Kim,&nbsp;Jean-Paul De La O,&nbsp;Matthew J. Ellis,&nbsp;Dan Liu,&nbsp;Shunqiang Li,&nbsp;Alana L. Welm,&nbsp;Suzanne A. W. Fuqua","doi":"10.1038/s41416-025-03096-0","DOIUrl":"10.1038/s41416-025-03096-0","url":null,"abstract":"","PeriodicalId":9243,"journal":{"name":"British Journal of Cancer","volume":"133 6","pages":"913-915"},"PeriodicalIF":6.8,"publicationDate":"2025-09-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.nature.comhttps://www.nature.com/articles/s41416-025-03096-0.pdf","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144943604","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Identifying health conditions associated with an increased risk of pancreatic ductal adenocarcinoma at medium term in nationwide electronic health records of primary care physicians.","authors":"Elie Rassy, Suzette Delaloge, Yannis Slaouti, Thomas Pudlarz, Béranger Lekens, Alice Boilève, Stefan Michiels, Mariam Karimi","doi":"10.1038/s41416-025-03172-5","DOIUrl":"10.1038/s41416-025-03172-5","url":null,"abstract":"<p><strong>Background: </strong>Pancreatic ductal adenocarcinoma (PDAC) presents an urgent challenge that necessitates improved early risk identification. We investigated the association between predictors of PDAC present at least two years before diagnosis and PDAC occurence.</p><p><strong>Methods: </strong>This case-control study used electronic health records from The Health Improvement Network Database (UK). Cases (10,575) were matched with controls (105,750) in a 1:10 ratio by gender, age, follow-up duration, and year of inclusion in the database. Variables included clinical features, comorbidities and blood result abnormalities, reported at least 2 years before PDAC diagnosis or equivalent timeframe for controls. Conditional logistic regression model with backward estimated odds ratios (OR) and 95% confidence interval (CI).</p><p><strong>Results: </strong>Electronic health records of cases reported higher prevalence of alcohol intake, cigarette smoking, dyslipidemia, increased blood pressure, and diabetes of more than four years' duration. Independent risk factors included pancreatic cysts (OR = 4.39), pancreatitis (OR = 2.16), dyslipidemia (OR = 1.50), smoking (OR = 1.09), and alcohol intake (OR = 1.09). Laboratory markers associated with higher risk included elevated alkaline phosphatase (OR = 3.21), bilirubin (OR = 2.48), alanine aminotransferase (OR = 1.76), erythrocyte sedimentation rate (OR = 1.27), and decreased albumin (OR = 1.54).</p><p><strong>Conclusions: </strong>Primary care electronic records can identify individuals at medium-term increased risk of PDAC, thus raises the opportunity to develop early detection models and address modifiable factors.</p>","PeriodicalId":9243,"journal":{"name":"British Journal of Cancer","volume":" ","pages":""},"PeriodicalIF":6.8,"publicationDate":"2025-08-30","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144943668","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Protein markers of ovarian cancer and its subtypes: insights from proteome-wide Mendelian randomisation analysis.","authors":"Anwar Mulugeta, David Stacey, Amanda L Lumsden, Iqbal Madakkatel, S Hong Lee, Johanna Mäenpää, Martin K Oehler, Elina Hyppönen","doi":"10.1038/s41416-025-03143-w","DOIUrl":"https://doi.org/10.1038/s41416-025-03143-w","url":null,"abstract":"<p><strong>Background: </strong>Ovarian cancer (OC) is often diagnosed at an advanced stage when prognosis is poor. We aimed to identify blood plasma proteins predictive of OC risk.</p><p><strong>Methods: </strong>We conducted proteome-wide Mendelian randomisation (MR) analyses using summary-level protein quantitative trait locus data covering 2337 plasma proteins, and genome-wide association data on OC and its subtypes (up to 25,509 cases) from the Ovarian Cancer Association Consortium. Wald ratio or inverse-variance weighted MR analysis was used as the primary method, depending on the number of instruments. We evaluated pleiotropy using MR-Egger intercept test and leave-one-out analysis.</p><p><strong>Results: </strong>From 2337 plasma proteins, 12 were associated (p < 7.4 × 10^-5) with OC or its subtypes. Robust evidence linked follitropin subunit beta (FSHB) with endometrioid OC (per SD higher, OR 2.41, 95% CI 1.56, 3.71). Associations for the other 11 proteins could be explained by pleiotropy from ABO or MAPT-AS1 loci. We identified 12 suggestive associations with OC or its subtypes at nominal threshold (p < 0.05), involving 11 plasma proteins, with no evidence of pleiotropy from leave-one-out and MR-Egger intercept tests (P_intercept > 0.17). Potential drug targets were identified for follitropin receptor and eight other proteins.</p><p><strong>Conclusion: </strong>Our study suggests FSHB and 11 additional plasma proteins as of potential interest in OC (or subtypes) prognosis, mostly representing potentially druggable targets.</p>","PeriodicalId":9243,"journal":{"name":"British Journal of Cancer","volume":" ","pages":""},"PeriodicalIF":6.8,"publicationDate":"2025-08-28","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144943248","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Primary tumor sidedness and negative hyperselection to modulate anti-EGFR-based maintenance strategies in patients with RAS wild-type metastatic colorectal cancer: individual patient data pooled analysis of two randomized clinical trials.","authors":"Alexej Ballhausen, Federica Morano, Arndt Stahler, Sara Lonardi, Andreas Jay Kind, Chiara Cremolini, Susanna Swoboda, Giovanni Randon, David Horst, Michele Prisciandaro, Annabel Helga Sophie Alig, Chiara Carlotta Pircher, Armin Jarosch, Paola Andena, Annika Kurreck, Anna Alessandra Chiaramonte, Sebastian Stintzing, Filippo Pietrantonio, Dominik Paul Modest, Alessandra Raimondi","doi":"10.1038/s41416-025-03164-5","DOIUrl":"https://doi.org/10.1038/s41416-025-03164-5","url":null,"abstract":"<p><strong>Background: </strong>Patients with RAS wild-type (WT), left-sided metastatic colorectal cancer (mCRC), negatively hyperselected for anti-EGFR resistance alterations, benefit most from anti-EGFR-based first-line treatment. The predictive impact of these stratification parameters on maintenance strategy efficacy is unclear.</p><p><strong>Methods: </strong>This pooled analysis included individual patient data from the PanaMa (NCT01991873) and Valentino (NCT02476045) phase 2 trials. Patients with RAS WT mCRC received FOLFOX plus Panitumumab induction therapy followed by maintenance with 5-fluorouracil/leucovorin (5-FU/LV) plus Panitumumab vs. 5-FU/LV monotherapy (PanaMa) or Panitumumab monotherapy (Valentino). Outcomes included progression-free survival (PFS) and overall survival (OS). Subgroup analyses examined primary tumor sidedness (left vs. right) and hyperselection status (negative vs. altered).</p><p><strong>Results: </strong>Among 607 patients receiving induction, sidedness and hyperselection status were available for 589 and 511 patients, respectively. Left-sided and negative hyperselected tumors were observed in 80.2% and 63.9% of patients, respectively. Panitumumab-based maintenance improved PFS in left-sided, negative hyperselected patients compared to 5-FU/LV alone, with no OS differences. PFS and OS were comparable for Panitumumab alone vs. Panitumumab plus 5-FU/LV.</p><p><strong>Conclusion: </strong>Tumor sidedness and hyperselection status significantly influence maintenance strategy efficacy in mCRC. For left-sided, negative hyperselected patients, Panitumumab monotherapy may optimize efficacy while minimizing toxicity. Further investigation into the relative contribution of individual hyperselection parameters in this setting is warranted.</p>","PeriodicalId":9243,"journal":{"name":"British Journal of Cancer","volume":" ","pages":""},"PeriodicalIF":6.8,"publicationDate":"2025-08-28","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144943648","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Correction: Biomarkers of response to camrelizumab combined with apatinib: an analysis from a phase II trial in recurrent/metastatic nasopharyngeal carcinoma","authors":"Kaiqi Lan,&nbsp;Shibing Li,&nbsp;Guodong Jia,&nbsp;Suchen Li,&nbsp;Siyi Xie,&nbsp;Linquan Tang,&nbsp;Haiqiang Mai,&nbsp;Li Yuan","doi":"10.1038/s41416-025-03077-3","DOIUrl":"10.1038/s41416-025-03077-3","url":null,"abstract":"","PeriodicalId":9243,"journal":{"name":"British Journal of Cancer","volume":"133 6","pages":"912-912"},"PeriodicalIF":6.8,"publicationDate":"2025-08-27","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.nature.comhttps://www.nature.com/articles/s41416-025-03077-3.pdf","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144943591","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Relation knowledge distillation 3D-ResNet-based deep learning for breast cancer molecular subtypes prediction on ultrasound videos: a multicenter study.","authors":"Yingnan Wu, Lei Zhou, Jing Zhao, Yanqing Peng, Xiaoying Li, Yaoting Wang, Sutian Zhu, Chunjie Hou, Pei Du, Lei Ling, Ying Wang, Jiawei Tian, Litao Sun","doi":"10.1038/s41416-025-03146-7","DOIUrl":"https://doi.org/10.1038/s41416-025-03146-7","url":null,"abstract":"<p><strong>Background: </strong>To develop and test a relation knowledge distillation three-dimensional residual network (RKD-R3D) model for predicting breast cancer molecular subtypes using ultrasound (US) videos to aid clinical personalized management.</p><p><strong>Methods: </strong>This multicentre study retrospectively included 882 breast cancer patients (2375 US videos and 9499 images) between January 2017 and December 2021, which was divided into training, validation, and internal test cohorts. Additionally, 86 patients was collected between May 2023 and November 2023 as the external test cohort. St. Gallen molecular subtypes (luminal A, luminal B, HER2-positive, and triple-negative) were confirmed via postoperative immunohistochemistry. The RKD-R3D based on US videos was developed and validated to predict four-classification molecular subtypes of breast cancer. The predictive performance of RKD-R3D was compared with RKD-R2D, traditional R3D, and preoperative core needle biopsy (CNB). The area under the receiver operating characteristic curve (AUC), sensitivity, specificity, accuracy, balanced accuracy, precision, recall, and F1-score were analyzed.</p><p><strong>Results: </strong>RKD-R3D (AUC: 0.88, 0.95) outperformed RKD-R2D (AUC: 0.72, 0.85) and traditional R3D (AUC: 0.65, 0.79) in predicting four-classification breast cancer molecular subtypes in the internal and external test cohorts. RKD-R3D outperformed CNB (Accuracy: 0.87 vs. 0.79) in the external test cohort, achieved good performance in predicting triple negative from non-triple negative breast cancers (AUC: 0.98), and obtained satisfactory prediction performance for both T1 and non-T1 lesions (AUC: 0.96, 0.90).</p><p><strong>Conclusions: </strong>RKD-R3D when used with US videos becomes a potential supplementary tool to non-invasively assess breast cancer molecular subtypes.</p>","PeriodicalId":9243,"journal":{"name":"British Journal of Cancer","volume":" ","pages":""},"PeriodicalIF":6.8,"publicationDate":"2025-08-26","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144943407","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}