BMC Cancer最新文献

{"title":"Impacts of omega-3 fatty acids intervention on immune functions and clinical outcomes in surgical gastric cancer patients: a systematic review and trial sequential meta-analysis.","authors":"Han Yang Yue, Xue Ting Yang, Jian Cheng Zhang, Ping Zhou, Jue Xia He, Jun Zeng, Jun Ying Li, Hua Jiang","doi":"10.1186/s12885-025-14788-4","DOIUrl":"https://doi.org/10.1186/s12885-025-14788-4","url":null,"abstract":"","PeriodicalId":9131,"journal":{"name":"BMC Cancer","volume":"25 1","pages":"1366"},"PeriodicalIF":3.4,"publicationDate":"2025-08-23","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC12374278/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144942724","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"MRD4U: A path to development for personalized liquid biopsy for children with central nervous system tumors.","authors":"Anthony R Miller, Tithi Shah, Corinne N Strawser, Adithe Rivaldi, Sarah A Wilson, Huachun Zhong, Jocelyn M Lucyshyn, Elizabeth Ar Garfinkle, Makenna McCoy, Seth Cummins, Alejandro Otero Bravo, Peter Chang, Gregory L Wheeler, David M Gordon, Benjamin J Kelly, Caitlyn Evanovich, Samara L Potter, Elizabeth A Varga, Diana P Rodriguez, Richard K Wilson, Catherine E Cottrell, Ralph Salloum, Sara Khan, Maryam Fouladi, Elaine R Mardis, Margot A Lazow, Katherine E Miller","doi":"10.1186/s12885-025-14711-x","DOIUrl":"https://doi.org/10.1186/s12885-025-14711-x","url":null,"abstract":"<p><strong>Background: </strong>Liquid biopsy assays using cerebrospinal fluid (CSF) can revolutionize care for children with central nervous system (CNS) tumors by enabling precise monitoring of therapeutic responses and detecting recurrence or measurable residual disease (MRD). These assays can detect cell-free, circulating tumor DNA (ctDNA) via somatic alterations, though accurately measuring low-abundance ctDNA in CSF is challenging.</p><p><strong>Methods: </strong>Our research focused on the optimization of next-generation sequencing library preparation from cell-free DNA (cfDNA), evaluating four commercial kits to address the low nucleic acid yield in CSF-derived cfDNA. The selected kit minimized false positives and detected somatic variants at 5% variant allele frequency using 0.1 ng input of synthetic cfDNA, suitable for low-volume CSF samples.</p><p><strong>Results: </strong>We then applied our optimized workflow to six children with CNS tumors using a personalized hybrid-capture sequencing strategy (\"MRD4U\"), in which individualized panels were designed based on each patient's tumor sequencing. Using MRD4U, we identified ctDNA in two samples, even though neither patient had radiographic or clinical evidence of disease at the time of liquid biopsy. Notably, one ctDNA-positive patient developed radiographic recurrence four months later, demonstrating the assay's potential to detect molecular relapse ahead of conventional clinical measures.</p><p><strong>Conclusions: </strong>These findings demonstrate applicability of our personalized MRD4U assay in early detection of disease recurrence. Unlike non-targeted or tumor-agnostic CSF liquid biopsy approaches, MRD4U leverages patient-specific genomic information to enable sensitive, tumor-informed monitoring that can be deployed across a wide range of pediatric CNS tumors. Our approach is broadly applicable to any tumor type with existing genomic data, enabling ctDNA detection across diverse diagnoses. Ultimately, this strategy may inform clinical decision-making and enable earlier therapeutic intervention.</p>","PeriodicalId":9131,"journal":{"name":"BMC Cancer","volume":"25 1","pages":"1365"},"PeriodicalIF":3.4,"publicationDate":"2025-08-23","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC12374395/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144942812","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Osimertinib-related myotoxicity: a disproportionality analysis of the FDA adverse event reporting system.","authors":"Yaqian Tan, Qi Song","doi":"10.1186/s12885-025-14743-3","DOIUrl":"https://doi.org/10.1186/s12885-025-14743-3","url":null,"abstract":"<p><strong>Background: </strong>Osimertinib is a third-generation epidermal growth factor receptor tyrosine kinase inhibitor that has been widely applied as a standard first-line treatment in advanced non-small cell lung cancer. However, the risk signals of osimertinib-related myotoxicity have not yet been fully examined. This study aimed to explore osimertinib-related myotoxicity by conducting a real-world disproportionality analysis.</p><p><strong>Methods: </strong>Data from January 1st 2015 to March 31st 2024 were retrieved from the U.S. Food and Drug Administration Adverse Event Reporting System database. Reporting odds ratio (ROR), proportional reporting ratio (PRR), and information component (IC) were employed to perform disproportionality analysis.</p><p><strong>Results: </strong>A total of 121 cases with osimertinib-related myotoxicity were identified and analyzed. The adverse events were mostly reported in females (n = 74, 61.2%) and patients aged over 65 years old (n = 56, 46.3%). The median value of adverse event onset time was 40 (13.3, 164.5). Disproportionality analysis revealed that blood creatine phosphokinase increased (ROR₀₂₅ = 5.00, IC₀₂₅ = 2.07, PRR = 6.18), myositis (ROR₀₂₅ = 2.72, IC₀₂₅ = 1.26, PRR = 4.22), and myopathy (ROR₀₂₅ = 1.28, IC₀₂₅ = 0.63, PRR = 2.32) carried significant risk signals of osimertinib-related myotoxicity.</p><p><strong>Conclusions: </strong>Our study comprehensively revealed the safety characteristics of osimertinib-associated myotoxicity. The results would offer referable evidence on the safety and prognosis of osimertinib.</p>","PeriodicalId":9131,"journal":{"name":"BMC Cancer","volume":"25 1","pages":"1360"},"PeriodicalIF":3.4,"publicationDate":"2025-08-22","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC12374422/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144942796","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Anti-PD-1 therapy achieves favorable outcome in gastric cancer combined with chronic hepatitis B.","authors":"Lingyan Wang, Fei Yang, Qinhui Dong, Tao-Hsin Tung, Junyan Ye, Kaili Cen, Yupei Yang, Jingjing Xie, Qunyi Guo","doi":"10.1186/s12885-025-14776-8","DOIUrl":"https://doi.org/10.1186/s12885-025-14776-8","url":null,"abstract":"<p><strong>Purpose: </strong>To investigate the efficacy and safety of immune checkpoint inhibitors in the treatment of gastric cancer patients with different hepatitis B virus (HBV) infection statuses.</p><p><strong>Patients and methods: </strong>The clinical data of 89 gastric cancer patients treated at our centre were retrospectively analysed. The patients were divided into chronic hepatitis B (CHB)-infected patients (13 patients), resolved hepatitis B (RHB)-infected patients (49 patients), and HBV-uninfected (HBV-) patients (27 patients) according to their HBV infection status. The efficacy and safety of antitumour treatment in patients in the three groups were analysed.</p><p><strong>Results: </strong>During anti-programmed death (ligand) 1 (PD-(L)1) therapy, no significant differences in the overall response rate (69.2% vs. 53.1% vs. 51.9%; P = 0.536) or disease control rate (92.3% vs. 79.6% vs. 81.5%; P = 0.567) were observed among the three groups. Progression-free survival benefited more from anti-PD-(L)1 antibody therapy in CHB patients than in HBV- patients (Not reached vs. 7 months; P = 0.024; hazard ratio (HR) = 0.38(95%CI 0.17-0.88)) and RHB patients (Not reached vs. 6 months; P = 0.001; HR = 0.34(95%CI 0.17-0.65)). Overall survival was also better in CHB patients than in HBV- patients (Not reached vs. 15 months; P = 0.014; HR = 0.31(95%CI 0.12-0.79)) and RHB patients (Not reached vs. 16 months; P = 0.005; HR = 0.32(95%CI 0.14-0.71)).</p><p><strong>Conclusion: </strong>Anti-PD-(L)1 immunotherapy may be safe and effective in gastric cancer patients with HBV infection. CHB patients are more likely to have lasting benefits than are HBV- patients and RHB patients. This may be attributed to alterations in the immune microenvironment in patients, and future studies should further explore the relevant mechanisms involved. CHB patients may have a more inflamed immune microenvironment, leading to enhanced ICI efficacy.</p>","PeriodicalId":9131,"journal":{"name":"BMC Cancer","volume":"25 1","pages":"1354"},"PeriodicalIF":3.4,"publicationDate":"2025-08-22","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC12372377/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144942862","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Metabolic landscape of clear cell renal cell carcinoma and search for metabolites predictive of drug response.","authors":"Michinobu Ozawa, Sei Naito, Hideki Makinoshima, Hiromi Ito, Atsushi Tsuya, Takafumi Narisawa, Hiroki Fukuhara, Yuki Takai, Masaki Ushijima, Mayu Yagi, Hayato Nishida, Norihiko Tsuchiya","doi":"10.1186/s12885-025-14661-4","DOIUrl":"https://doi.org/10.1186/s12885-025-14661-4","url":null,"abstract":"<p><strong>Background: </strong>Clear cell renal cell carcinoma (ccRCC) commonly exhibits biallelic inactivation of VHL genes, profoundly impacting intracellular metabolic pathways and utilization of metabolic substrates. The aims of this study were to validate the metabolomic profile previously identified in ccRCC surgical specimens, to construct a metabolic classification in ccRCC, and to exploratorily investigate metabolic biomarkers of systemic therapy response.</p><p><strong>Materials and methods: </strong>We first examined the metabolome in 52 paired tumor/normal surgical ccRCC samples, and then compared the metabolites using paired t-test. Unsupervised clustering analysis was done, and the patients were divided into four subgroups. Then tumor grade and transcriptome analyses were compared among the four groups. Finally, to compare the metabolome according to the effect of systemic therapy, we analyzed 9 patients with vascular endothelial growth factor (VEGF)-tyrosine kinase inhibitor (VEGF-TKI) and 11 patients with immune checkpoint blockade (ICB), respectively.</p><p><strong>Results: </strong>The upper stream of glycolysis and the pentose phosphate pathway were commonly activated, along with elevated glutamine levels and reduced proteinogenic amino acids (AAs) other than glutamine in ccRCC tissues, consistent with prior findings. Lactate levels, previously reported as elevated, varied across metabolic subgroups. he Metabolic subgroups enriched with high-grade tumors demonstrated lower expression of VEGF pathway-related genes. VEGF-TKI responders showed decreased levels of some fatty acids. ICI responders exhibited reduced levels of tryptophan and hydroquinone. alongside increased levels of pyruvic acid-oxime, 3-hydroxypropinoic acid, and hydroxylamine.</p><p><strong>Conclusions: </strong>We validated the landscape of the ccRCC metabolome and identified some metabolites as potential biomarkers for predicting therapeutic response in RCC.</p>","PeriodicalId":9131,"journal":{"name":"BMC Cancer","volume":"25 1","pages":"1357"},"PeriodicalIF":3.4,"publicationDate":"2025-08-22","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC12372187/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144942833","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Development and validation of a nomogram for predicting immunotherapy outcomes in lung cancer patients using clinical and blood biomarkers.","authors":"Ting Ouyang, Fei Zhang, Yan Yang, Tianwei Luo, Ao Chen, Ning Han, Jun Qian, Xiaoyuan Chu, Chao Chen, Mi Yang","doi":"10.1186/s12885-025-14559-1","DOIUrl":"https://doi.org/10.1186/s12885-025-14559-1","url":null,"abstract":"<p><strong>Background: </strong>Lung cancer remains the leading cause of cancer-related mortality worldwide. While immune checkpoint inhibitors (ICIs) have improved survival outcomes for some patients, their efficacy and adverse effects vary significantly. Thus, developing accurate and practical prognostic tools is essential to optimize treatment decision-making.</p><p><strong>Methods: </strong>This retrospective study analyzed 436 lung cancer patients treated with ICIs, who were randomly divided into training (70%) and validation (30%) cohorts. Independent prognostic factors for overall survival (OS) and progression-free survival (PFS) were identified using LASSO regression and multivariate Cox regression. Nomograms were constructed based on clinical and blood biomarkers. Model performance was assessed using the concordance index (C-index), ROC curve, calibration curve, and decision curve analysis (DCA). Kaplan-Meier analysis validated patient stratification.</p><p><strong>Results: </strong>The key independent predictive factors for OS and PFS included neutrophil-to-lymphocyte ratio (NLR), previous surgery, liver metastasis, clinical stage, treatment lines, and treatment response evaluation. The nomograms achieved C-index values of 0.709 (OS) and 0.730 (PFS) in the training cohort, with validation C-indexes of 0.655 (OS) and 0.694 (PFS). The ROC curves demonstrated good predictive accuracy for 12-, 24-, and 36-month outcomes. High-risk patients exhibited significantly shorter median OS and PFS (P < 0.001).</p><p><strong>Conclusion: </strong>The nomograms developed in this study, integrating clinical and blood biomarkers, provide a cost-effective, simple, and accurate tool for predicting the prognosis of lung cancer patients receiving ICIs treatment, to facilitate personalized clinical decision-making.</p>","PeriodicalId":9131,"journal":{"name":"BMC Cancer","volume":"25 1","pages":"1353"},"PeriodicalIF":3.4,"publicationDate":"2025-08-22","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC12372320/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144942713","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Epidemiological and clinicopathological characteristics of ovarian cancer in Syria during the war years: a national multicenter retrospective study (2017-2021).","authors":"Shant Apelian, Alaa Hamdan, Rita Mohammad, Aram Abbas, Saliha Rizvi, Gowhar Rashid, Rim M Harfouch, Shahd Maarrawi, Abdallah Aladna, Ibrahim Shammas, Mounzer Boubou, Michael Georgeos, Zuheir Alshehabi, Safa K Salman","doi":"10.1186/s12885-025-14784-8","DOIUrl":"https://doi.org/10.1186/s12885-025-14784-8","url":null,"abstract":"<p><strong>Background: </strong>Despite the high burden and regional variability of ovarian cancer (OC), data on its status in Syria, a country impacted by long-standing conflict, remain scarce. This study aimed to provide the first nationwide analysis of OC epidemiological and clinicopathological characteristics.</p><p><strong>Methods: </strong>In this national, multicenter, retrospective cross-sectional study, we collected data from three tertiary referral university hospitals in Syria: Tishreen University Hospital, Al-Bairouni University Hospital, and Ibn Rushd Hospital. The inclusion period spanned from 2017 to 2021. All included patients were 18 years or older and had a newly diagnosed, confirmed primary OC. Data collected included sociodemographic information, presenting symptoms, CA-125 levels, and treatment details.</p><p><strong>Results: </strong>A total of 531 patients with OC with a mean age of 53.1 years (range: 18-91) were included in the study, representing 1.2% (95% confidence interval: 1.1%-1.3%) of the 44,248 patients diagnosed with all types of cancer. Serous carcinoma (55.4%) was the most common OC subtype, followed by mucinous (13.8%) and endometrioid carcinoma (9.7%). Among epithelial OCs, having high levels of CA-125 was associated with higher odds of serous carcinoma (adjusted OR: 2.30, 95% CI: 1.08-4.91) and lower odds of endometrioid cancer adjusted OR: 0.21, 95% CI: 0.08-0.54). Abdominal bloating was the most frequently reported symptom, affecting 194 patients (36.5%).</p><p><strong>Conclusions: </strong>Our study revealed a lower proportion of OCs diagnoses among all cancer cases compared to global and regional data which may be attributed to underreporting, influenced by the disease's specific characteristics and presence of more barriers for women to healthcare access in Syria's conflict-affected setting. Given the study's limitations, further research is essential to provide a more comprehensive assessment of the OC burden in Syria.</p>","PeriodicalId":9131,"journal":{"name":"BMC Cancer","volume":"25 1","pages":"1359"},"PeriodicalIF":3.4,"publicationDate":"2025-08-22","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC12374289/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144942779","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"The effects of nutrition education with Pecha Kucha method on prevention of malnutrition in cancer patients undergoing radiotherapy: a randomised controlled study.","authors":"Gülcan Bahcecioglu Turan, Fatma Karaaslan, Zülfünaz Özer","doi":"10.1186/s12885-025-14626-7","DOIUrl":"https://doi.org/10.1186/s12885-025-14626-7","url":null,"abstract":"<p><strong>Aim: </strong>The present study aimed to evaluate the effects of nutrition education by using the Pecha Kucha (20*20) presentation method on preventing malnutrition in cancer patients undergoing radiotherapy.</p><p><strong>Method: </strong>This randomized controlled experimental study. A total of 113 patients were assessed for eligibility. Thirty-three patients declined to participate, and 20 did not meet the inclusion criteria. The final sample consisted of 60 volunteering cancer patients who met the inclusion criteria and agreed to participate. These patients were randomly assigned to either the experimental group (n = 30) or the control group (n = 30). All participants completed the study.The experimental group received nutrition education by using the Pecha Kucha method, while the control group received traditional education. Data were collected by using Personal Information Form, NRS-2002 (Nutritional Risk Screening) Assessment Form and Subjective Global Assessment (SGA).</p><p><strong>Results: </strong>The first and third month NRS-2002 scores of the experimental group were significantly lower than the control group(better nutritional status) (p < 0.001). After three months, the number of well-nourished individuals decreased in both groups; however, this decline was more pronounced in the control group (p = 0.001). In contrast, the intervention group largely maintained their nutritional status. NRS-2002 (β = -0.683, p < 0.001) and SGA (β = -3.324, p < 0.001) values were lower in the group that received pecha kucha application.</p><p><strong>Conclusion: </strong>The negative beta coefficient for NRS-2002 suggests that the Pecha Kucha training significantly reduced patients' nutritional risk. Similarly, the reduction in SGA scores reflects improved nutritional status, indicating a meaningful clinical improvement. Nutrition education provided by Pecha Kucha method is effective in preventing malnutrition in cancer patients undergoing radiotherapy. This educational method is considered to be an effective approach with high applicability in clinical settings. It is recommended that innovative and audiovisual presentation techniques such as Pecha Kucha should be utilized to strengthen responsibilities of nurses in patient education and to increase the effectiveness of education processes.</p><p><strong>Trial registration: </strong>The study has been registered on ClinicalTrials.gov Protocol Registration ( https://register.</p><p><strong>Clinicaltrials: </strong>gov/prs/app/action/ViewOrUnrelease?uid=U0004RLH&ts=1&sid=S000D85L&cx=bmoiql ) on 5 January 2023 (Trial Registration Number: NCT05852756).</p>","PeriodicalId":9131,"journal":{"name":"BMC Cancer","volume":"25 1","pages":"1355"},"PeriodicalIF":3.4,"publicationDate":"2025-08-22","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC12372213/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144942719","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Treatment patterns and survival outcomes of peripheral T-Cell lymphoma in Saudi Arabia: a 12-Year retrospective analysis.","authors":"Mubarak Al-Mansour, Hosam A Alghanmi, Duaa E Shareef, Esraa S Alshamat, Abdulaziz M Asseri, Mohammed O Bayamin, Syed Sameer Aga, Manal A Alsaggaf","doi":"10.1186/s12885-025-14793-7","DOIUrl":"https://doi.org/10.1186/s12885-025-14793-7","url":null,"abstract":"","PeriodicalId":9131,"journal":{"name":"BMC Cancer","volume":"25 1","pages":"1361"},"PeriodicalIF":3.4,"publicationDate":"2025-08-22","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC12374265/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144942876","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Deep learning radiomics based on MRI for differentiating tongue cancer T - staging.","authors":"Zhaoyi Lu, Bowen Zhu, Hang Ling, Xi Chen","doi":"10.1186/s12885-025-14627-6","DOIUrl":"https://doi.org/10.1186/s12885-025-14627-6","url":null,"abstract":"<p><strong>Objective: </strong>To develop a deep learning-based MRI model for predicting tongue cancer T-stage.</p><p><strong>Methods: </strong>This retrospective study analyzed clinical and MRI data from 579 tongue cancer patients (Xiangya Cancer Hospital and Jiangsu Province Hospital). T2-weighted (T2WI) and contrast-enhanced T1-weighted (CET1) sequences were preprocessed (anonymization/resampling/calibration). Regions of interest (ROI) were segmented by two radiologists (intraclass correlation coefficient (ICC) > 0.75), and using PyRadiomics, 2375 radiomics features were extracted. ResNet18 and ResNet50 algorithms were employed to build deep learning models (deep learning radiomics (DLR) resnet18 / DLRresnet50), compared with a radiomics model (Rad) based on 17 optimized features. Performance was evaluated via AUC, DCA, IDI, and NRI in different sets.</p><p><strong>Results: </strong>In training set, deep learning models outperformed Rad (AUC: DLRresnet18 = 0.837, DLRresnet50 = 0.847 vs. Rad = 0.828). Test set and and external validation set results were consistent (DLRresnet18, AUC = 0.805 / 0.857; DLRresnet50, AUC = 0.810 / 0.860). The decision curve analysis (DCA) demonstrated that both deep learning models performed better than the Rad model in the training set, test set, and external validation set. Furthermore, both NRI and IDI of the two deep learning models compared with the Rad model were greater than 0.</p><p><strong>Conclusion: </strong>DLRresnet18 and DLRresnet50 models significantly improve T-stage prediction accuracy over traditional radiomics, reducing subjective interpretation errors and supporting personalized treatment planning. This research achievement provides new ideas and tools for image-assisted diagnosis of tongue cancer T-stage.</p><p><strong>Level of evidence: </strong>III.</p>","PeriodicalId":9131,"journal":{"name":"BMC Cancer","volume":"25 1","pages":"1358"},"PeriodicalIF":3.4,"publicationDate":"2025-08-22","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC12372389/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144942688","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}