Pub Date : 2025-12-16eCollection Date: 2026-01-01DOI: 10.1016/j.eclinm.2025.103686
Vasileios Fragoulakis, Jesse J Swen, Margarita-Ioanna Koufaki, Kathrin Blagec, Tanja Blagus, Stefan Böhringer, Anne Cambon-Thomsen, Erika Cecchin, Ka-Chun Cheung, Vera H M Deneer, Mathilde Dupui, Siv Jonsson, Candace Joefield-Roka, Katja S Just, Mats O Karlsson, Lidija Konta, Rudolf Koopmann, Marjolein Kriek, Thorsten Lehr, Lisanne E N Manson, Emmanuelle Rial-Sebbag, Victoria Rollinson, Rossana Roncato, Matthias Samwald, Elke Schaeffeler, Maria Skokou, Matthias Schwab, Daniela Steinberger, Julia C Stingl, Roman Tremmel, Richard M Turner, Mandy H van Rhenen, Cathelijne H van der Wouden, Cristina Lucía Dávila-Fajardo, Vita Dolžan, Munir Pirmohamed, Gere Sunder-Plassmann, Giuseppe Toffoli, Henk-Jan Guchelaar, George P Patrinos, Christina Mitropoulou
Background: Pharmacogenetics (PGx) aims to revolutionize healthcare by individualizing drug doses and medication choices. However, clinical uptake will require positive evaluation evidence of both clinical utility and cost-effectiveness. We have recently demonstrated the clinical utility of this approach, using a panel-based PGx-guided treatment of patients from various indications recruited in seven countries (PREPARE study).
Methods: Here, we provide economic evidence from a multinational cost-utility analysis of PGx-guided treatment in 6930 patients participating in the PREPARE study. The study was conducted from March 2017 to June 2020. We used the national healthcare system's perspective in each participating country, including only direct medical costs that budget holders cover. A Visual Analog Scale was used to measure utility and the quality of life was estimated by averaging the Visual Analog Scale scores of participants over four specific time points in the study, namely baseline visit (day 1), week 4, week 12, and 18 months from the baseline visit.
Findings: Our analysis showed that the PGx-guided treatment is marginally cost-effective at the threshold of €11,000 QALYs. Cost drivers were hospitalization and ADRs costs, accounting for most of the resources used in both groups (46% and 37.5% in the PGx-guided group versus 49% and 48% in the control group, respectively), as a result of the average duration of hospitalization [1.51 days (95% CI: 1.23-1.82) for the PGx-guided group and 2.37 days (95% CI: 1.95-2.89) for the control group, resulting in a mean difference of 0.86 days (95% CI: 0.37-1.44). The difference in QALYs gained was 0.00178 (95% CI: 0.00176-0.00180). The ICER was €12,020 (95% CI: €10,957-€13,356) per QALY on average (SD: €116). When comparing cost and effectiveness of actionable PGx-guided versus actionable control patients, the total cost for the PGx-guided group was €491 (95% CI: €384-€613), versus €767 (95% CI: €583-€982) in the control group, with an incremental cost difference of €276 (95% CI: €62-€511), favoring the PGx-guided group. Also, the difference in effectiveness was 0.007 QALYs (95% CI: -0.021 to 0.033). Lastly, the difference in the mean total cost was estimated to be €21.4 (95% CI: €19.5-€23.8), while without considering the PGx test cost, indicative of a pre-emptive genetic testing approach, the PGx-guided treatment becomes a cost-saving option, with an estimated savings of approximately €103.6 (€124-€21.4) per patient.
Interpretation: These data suggest that panel-based PGx testing is cost-effective, which, together with the clinically beneficial outcomes already demonstrated in the PREPARE study, provides additional evidence of the need to implement PGx into clinical practice.
{"title":"Multinational cost-utility analysis of panel-based pharmacogenetics-guided treatment of patients enrolled in the U-PGx PREPARE study.","authors":"Vasileios Fragoulakis, Jesse J Swen, Margarita-Ioanna Koufaki, Kathrin Blagec, Tanja Blagus, Stefan Böhringer, Anne Cambon-Thomsen, Erika Cecchin, Ka-Chun Cheung, Vera H M Deneer, Mathilde Dupui, Siv Jonsson, Candace Joefield-Roka, Katja S Just, Mats O Karlsson, Lidija Konta, Rudolf Koopmann, Marjolein Kriek, Thorsten Lehr, Lisanne E N Manson, Emmanuelle Rial-Sebbag, Victoria Rollinson, Rossana Roncato, Matthias Samwald, Elke Schaeffeler, Maria Skokou, Matthias Schwab, Daniela Steinberger, Julia C Stingl, Roman Tremmel, Richard M Turner, Mandy H van Rhenen, Cathelijne H van der Wouden, Cristina Lucía Dávila-Fajardo, Vita Dolžan, Munir Pirmohamed, Gere Sunder-Plassmann, Giuseppe Toffoli, Henk-Jan Guchelaar, George P Patrinos, Christina Mitropoulou","doi":"10.1016/j.eclinm.2025.103686","DOIUrl":"10.1016/j.eclinm.2025.103686","url":null,"abstract":"<p><strong>Background: </strong>Pharmacogenetics (PGx) aims to revolutionize healthcare by individualizing drug doses and medication choices. However, clinical uptake will require positive evaluation evidence of both clinical utility and cost-effectiveness. We have recently demonstrated the clinical utility of this approach, using a panel-based PGx-guided treatment of patients from various indications recruited in seven countries (PREPARE study).</p><p><strong>Methods: </strong>Here, we provide economic evidence from a multinational cost-utility analysis of PGx-guided treatment in 6930 patients participating in the PREPARE study. The study was conducted from March 2017 to June 2020. We used the national healthcare system's perspective in each participating country, including only direct medical costs that budget holders cover. A Visual Analog Scale was used to measure utility and the quality of life was estimated by averaging the Visual Analog Scale scores of participants over four specific time points in the study, namely baseline visit (day 1), week 4, week 12, and 18 months from the baseline visit.</p><p><strong>Findings: </strong>Our analysis showed that the PGx-guided treatment is marginally cost-effective at the threshold of €11,000 QALYs. Cost drivers were hospitalization and ADRs costs, accounting for most of the resources used in both groups (46% and 37.5% in the PGx-guided group versus 49% and 48% in the control group, respectively), as a result of the average duration of hospitalization [1.51 days (95% CI: 1.23-1.82) for the PGx-guided group and 2.37 days (95% CI: 1.95-2.89) for the control group, resulting in a mean difference of 0.86 days (95% CI: 0.37-1.44). The difference in QALYs gained was 0.00178 (95% CI: 0.00176-0.00180). The ICER was €12,020 (95% CI: €10,957-€13,356) per QALY on average (SD: €116). When comparing cost and effectiveness of actionable PGx-guided versus actionable control patients, the total cost for the PGx-guided group was €491 (95% CI: €384-€613), versus €767 (95% CI: €583-€982) in the control group, with an incremental cost difference of €276 (95% CI: €62-€511), favoring the PGx-guided group. Also, the difference in effectiveness was 0.007 QALYs (95% CI: -0.021 to 0.033). Lastly, the difference in the mean total cost was estimated to be €21.4 (95% CI: €19.5-€23.8), while without considering the PGx test cost, indicative of a pre-emptive genetic testing approach, the PGx-guided treatment becomes a cost-saving option, with an estimated savings of approximately €103.6 (€124-€21.4) per patient.</p><p><strong>Interpretation: </strong>These data suggest that panel-based PGx testing is cost-effective, which, together with the clinically beneficial outcomes already demonstrated in the PREPARE study, provides additional evidence of the need to implement PGx into clinical practice.</p><p><strong>Funding: </strong>European Union Horizon 2020.</p>","PeriodicalId":11393,"journal":{"name":"EClinicalMedicine","volume":"91 ","pages":"103686"},"PeriodicalIF":10.0,"publicationDate":"2025-12-16","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC12768869/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145917330","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}
Pub Date : 2025-12-10eCollection Date: 2025-12-01DOI: 10.1016/j.eclinm.2025.103669
Tao Ge, Hongyi Zhang, Xinsheng Zhu, Meixin Teng, Yifei Zhou, Suyu Wang, Dongliang Bian, Lele Zhang, Haiyang Hu, Qiji Guo, Jincheng Su, Shiqi Hu, Huansha Yu, Yirui Zhou, Jie Huang, Jie Yang, Yuming Zhu, Yi Bao, Minwei Bao, Xuefei Hu, Wenxin He, Peng Zhang
<p><strong>Background: </strong>Approximately 30% of patients with non-small-cell lung cancer (NSCLC) have initially resectable disease. This trial aimed to evaluate the safety and feasibility of administering sintilimab-based induction therapy for potentially resectable stage III NSCLC to offer a more advantageous therapeutic strategy.</p><p><strong>Methods: </strong>This investigator-initiated, open-label, phase 2 trial (NCT04728724) aimed to explore the efficacy and safety of sintilimab-based induction therapy in patients with potentially resectable stage III NSCLC on the basis of tumour PD-L1 expression. Eligible patients with PD-L1 expression in at least 50% of tumour cells received four cycles of sintilimab (200 mg every 3 weeks) as monotherapy, and patients with PD-L1 expression in less than 50% of tumour cells or of unknown status received four cycles of sintilimab combined with carboplatin-based chemotherapy before surgical resection. Evaluation of efficacy was conducted by a multidisciplinary team every two cycles. Surgical resection was performed within 21-28 days of the last dose. The primary endpoint was the major pathological response rate. Additionally, bulk RNA sequencing of baseline and surgical samples was conducted to investigate the tumour microenvironment and identify potential biomarkers.</p><p><strong>Findings: </strong>100 patients were enrolled between November 16, 2022, and September 11, 2023, and 97 patients were included in analyses. 75 (77%) of 97 patients completed four cycles of neoadjuvant treatment (range 1-6 cycles). An overall response was observed in 80 patients (82%; 95% CI 75-90), and all had a partial response. Surgery was performed in 58 patients (60%), and all patients had R0 resection. 38 (66%, 95% CI 53-78) of 58 patients had a major pathological response. As of September 24, 2025, the median follow-up duration was 31.0 months (IQR 27.5-32.6), and median progression-free survival and overall survival were not reached. 24-month progression-free survival was 78% (95% CI 71-87), and 24-month overall survival was 88% (81-94). Grade 3 or higher treatment-related adverse events occurred in 40 (41%) of 97 patients, and the most common grade 3-4 treatment-related adverse events were neutropenia (n = 35 [36%]), leukopenia (n = 18 [19%]). Immune-related adverse events occurred in 51 (53%) of 97 patients receiving study treatment. Differential gene expression analysis in the combination of baseline and surgical tumour samples revealed the upregulation of immune-related genes in responders to neoadjuvant immune checkpoint blockade, and the upregulation of metabolism-related genes in non-responders.</p><p><strong>Interpretation: </strong>Sintilimab-based induction treatment strategy could be a feasible option for potentially resectable NSCLC patients with stage III disease. Notably, we found that increased plasma cell signatures were predictive of response, while elevated AKR1C family gene expression substantially co
背景：大约30%的非小细胞肺癌（NSCLC）患者最初是可切除的。本试验旨在评估以辛替利单抗为基础的诱导治疗潜在可切除的III期NSCLC的安全性和可行性，以提供更有利的治疗策略。方法：这项由研究者发起的、开放标签的2期试验（NCT04728724）旨在探讨基于肿瘤PD-L1表达的西替利单抗诱导治疗在可能可切除的III期NSCLC患者中的有效性和安全性。在至少50%的肿瘤细胞中PD-L1表达的符合条件的患者接受4个周期的辛替单抗单药治疗（每3周200 mg），而PD-L1表达低于50%的肿瘤细胞或状态未知的患者在手术切除前接受4个周期的辛替单抗联合卡铂化疗。疗效评估由多学科小组每两个周期进行一次。最后一次给药后21 ~ 28天内进行手术切除。主要终点为主要病理反应率。此外，对基线和手术样本进行了大量RNA测序，以研究肿瘤微环境并识别潜在的生物标志物。研究结果：100名患者在2022年11月16日至2023年9月11日期间入组，其中97名患者纳入分析。97例患者中有75例（77%）完成了4个周期的新辅助治疗（范围1-6个周期）。80例患者（82%;95% CI 75-90）出现总体缓解，所有患者均出现部分缓解。58例（60%）患者接受手术，所有患者均行R0切除。58例患者中有38例（66%,95% CI 53-78）出现主要病理反应。截至2025年9月24日，中位随访时间为31.0个月（IQR为27.5-32.6），中位无进展生存期和总生存期均未达到。24个月无进展生存率为78% (95% CI 71-87)， 24个月总生存率为88%（81-94）。97例患者中有40例（41%）发生3级及以上治疗相关不良事件，最常见的3-4级治疗相关不良事件为中性粒细胞减少症（n = 35[36%]）、白细胞减少症（n = 18[19%]）。接受研究治疗的97例患者中有51例（53%）发生免疫相关不良事件。基线和手术肿瘤样本的差异基因表达分析显示，免疫相关基因在新辅助免疫检查点阻断的应答者中上调，而代谢相关基因在无应答者中上调。结论：以辛替利单抗为基础的诱导治疗策略可能是可切除的III期非小细胞肺癌患者的可行选择。值得注意的是，我们发现增加的浆细胞特征可以预测应答，而升高的AKR1C家族基因表达与免疫检查点阻断的抵抗有很大的相关性。基金资助：国家自然科学基金项目（82430053）；上海市科学技术委员会（资助号：24SF1904500）；上海市教委创新计划项目（批准号：2023ZKZD33）；AI for Science)；同济大学“医学- x”跨学科研究计划项目（批准号：2025-0554-ZD-03）；上海市肺科医院基金资助(批准号：；LYRC202402 FKLY20004)。
{"title":"Efficiency and safety of sintilimab-based induction therapy in potentially resectable stage III non-small cell lung cancer (LungMate-009): an open-label, phase 2 trial.","authors":"Tao Ge, Hongyi Zhang, Xinsheng Zhu, Meixin Teng, Yifei Zhou, Suyu Wang, Dongliang Bian, Lele Zhang, Haiyang Hu, Qiji Guo, Jincheng Su, Shiqi Hu, Huansha Yu, Yirui Zhou, Jie Huang, Jie Yang, Yuming Zhu, Yi Bao, Minwei Bao, Xuefei Hu, Wenxin He, Peng Zhang","doi":"10.1016/j.eclinm.2025.103669","DOIUrl":"10.1016/j.eclinm.2025.103669","url":null,"abstract":"<p><strong>Background: </strong>Approximately 30% of patients with non-small-cell lung cancer (NSCLC) have initially resectable disease. This trial aimed to evaluate the safety and feasibility of administering sintilimab-based induction therapy for potentially resectable stage III NSCLC to offer a more advantageous therapeutic strategy.</p><p><strong>Methods: </strong>This investigator-initiated, open-label, phase 2 trial (NCT04728724) aimed to explore the efficacy and safety of sintilimab-based induction therapy in patients with potentially resectable stage III NSCLC on the basis of tumour PD-L1 expression. Eligible patients with PD-L1 expression in at least 50% of tumour cells received four cycles of sintilimab (200 mg every 3 weeks) as monotherapy, and patients with PD-L1 expression in less than 50% of tumour cells or of unknown status received four cycles of sintilimab combined with carboplatin-based chemotherapy before surgical resection. Evaluation of efficacy was conducted by a multidisciplinary team every two cycles. Surgical resection was performed within 21-28 days of the last dose. The primary endpoint was the major pathological response rate. Additionally, bulk RNA sequencing of baseline and surgical samples was conducted to investigate the tumour microenvironment and identify potential biomarkers.</p><p><strong>Findings: </strong>100 patients were enrolled between November 16, 2022, and September 11, 2023, and 97 patients were included in analyses. 75 (77%) of 97 patients completed four cycles of neoadjuvant treatment (range 1-6 cycles). An overall response was observed in 80 patients (82%; 95% CI 75-90), and all had a partial response. Surgery was performed in 58 patients (60%), and all patients had R0 resection. 38 (66%, 95% CI 53-78) of 58 patients had a major pathological response. As of September 24, 2025, the median follow-up duration was 31.0 months (IQR 27.5-32.6), and median progression-free survival and overall survival were not reached. 24-month progression-free survival was 78% (95% CI 71-87), and 24-month overall survival was 88% (81-94). Grade 3 or higher treatment-related adverse events occurred in 40 (41%) of 97 patients, and the most common grade 3-4 treatment-related adverse events were neutropenia (n = 35 [36%]), leukopenia (n = 18 [19%]). Immune-related adverse events occurred in 51 (53%) of 97 patients receiving study treatment. Differential gene expression analysis in the combination of baseline and surgical tumour samples revealed the upregulation of immune-related genes in responders to neoadjuvant immune checkpoint blockade, and the upregulation of metabolism-related genes in non-responders.</p><p><strong>Interpretation: </strong>Sintilimab-based induction treatment strategy could be a feasible option for potentially resectable NSCLC patients with stage III disease. Notably, we found that increased plasma cell signatures were predictive of response, while elevated AKR1C family gene expression substantially co","PeriodicalId":11393,"journal":{"name":"EClinicalMedicine","volume":"90 ","pages":"103669"},"PeriodicalIF":10.0,"publicationDate":"2025-12-10","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC12752746/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145877803","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}
Pub Date : 2025-12-09eCollection Date: 2025-12-01DOI: 10.1016/j.eclinm.2025.103684
Shiyao Cheng, Yuandan Wei, Huaguang Zheng, Qingrong Zhang, Xuanyan Yang, Zhen Lu, Zehua Huang, Zijun Wan, Yanfeng Shi, Jie Zhang, Jay Chol Choi, Chul-Hoo Kang, Hong Jun Kim, Joong-Goo Kim, Hee-Joon Bae, Juneyoung Lee, Jeong-Yoon Lee, Hannah Jung, Cang Guo, Xiaoling Liao, Jianping Wu, Xia Meng, Zixiao Li, Stéphanie Debette, Hao Li, Yongjun Wang, Si Cheng, Siyang Liu
<p><strong>Background: </strong>Stroke is a leading cause of mortality and disability globally, with high rates of adverse outcomes. We aimed to develop and test a clinically accessible and explainable predictive model for post-stroke composite outcomes that could identify patients at high-risk for targeted interventions.</p><p><strong>Methods: </strong>This modelling study used artificial intelligence to develop and validate a prognostic predictive model for post-stroke outcomes at 3 months and over 5 years, using comprehensive data from the third China National Stroke Registry (CNSR-III; 2015-2018). The development cohort comprised 11,313 patients from CNSR-III. Validation cohort 1 (n = 2627 from CNSR-III; internal validation) were selected at random on the basis of hospital assignment. Validation cohort 2 (external validation) included 5158 patients with a history of ischaemic stroke, excluding those diagnosed with TIA, from the CHANCE-2 trial (2019-2021) done at 202 centres in China. Validation cohort 3 (external validation: disability and mortality) involved 3855 patients with a history of acute ischaemic stroke or TIA from a single centre in the Clinical Research Center for Stroke-Fifth Division (CRCS-5) cohort (2011-2024) in South Korea. We evaluated 309 hospitalisation variables, including baseline characteristics, medical history, hospitalisation data, biomarkers, geographical factors, National Institutes of Health Stroke Scale (NIHSS) scores, modified Rankin Scale (mRS), and stroke polygenic risk scores, using an extreme gradient boosting tree model. Feature importance was assessed via Shapley values. Primary outcomes were 3-month stroke recurrence, disability (mRS >2), and mortality. Secondary outcomes were assessed at six additional time points over 5 years. A nested cross-validation scheme was employed for feature selection and internal validation in 80% of patients (n = 11,313) from CNSR-III cohort. Testing was performed in the remaining 20% of patients (n = 2627) from CNSR-III (validation cohort 1). External validation was performed in validation cohorts 2 and 3.</p><p><strong>Findings: </strong>Global and domain-specific delta-NIHSS<sub>(admission-discharge)</sub> emerged as the strongest predictor of stroke recurrence, disability, and mortality. The Delta-NIHSS-based Predictor for Post-Stroke Composite Outcomes (DISCO) model, integrating 16 delta-NIHSS<sub>(admission-to-discharge)</sub> measures, age, gender, TOAST subtypes, history of TIA, stroke, diabetes and mRS at discharge, achieved the area under the receiver operating curve (AUCs) of 0.805 and 0.815 for recurrence and mortality and of 0.852 for disability at 3 months in validation cohort 2. The 1% patients with highest-risk exhibited a 19.27-fold relative risk (RR) for recurrence, 36.59-fold RR for disability, and 225.5-fold RR for mortality at 3 months. The DISCO model is available at https://www.discosysu.cn.</p><p><strong>Interpretation: </strong>The DISCO model, incorp
{"title":"Artifical intelligence-powered delta-NIHSS-based model for predicting recurrence, disability and mortality after acute ischaemic strokes (DISCO): a modelling study.","authors":"Shiyao Cheng, Yuandan Wei, Huaguang Zheng, Qingrong Zhang, Xuanyan Yang, Zhen Lu, Zehua Huang, Zijun Wan, Yanfeng Shi, Jie Zhang, Jay Chol Choi, Chul-Hoo Kang, Hong Jun Kim, Joong-Goo Kim, Hee-Joon Bae, Juneyoung Lee, Jeong-Yoon Lee, Hannah Jung, Cang Guo, Xiaoling Liao, Jianping Wu, Xia Meng, Zixiao Li, Stéphanie Debette, Hao Li, Yongjun Wang, Si Cheng, Siyang Liu","doi":"10.1016/j.eclinm.2025.103684","DOIUrl":"10.1016/j.eclinm.2025.103684","url":null,"abstract":"<p><strong>Background: </strong>Stroke is a leading cause of mortality and disability globally, with high rates of adverse outcomes. We aimed to develop and test a clinically accessible and explainable predictive model for post-stroke composite outcomes that could identify patients at high-risk for targeted interventions.</p><p><strong>Methods: </strong>This modelling study used artificial intelligence to develop and validate a prognostic predictive model for post-stroke outcomes at 3 months and over 5 years, using comprehensive data from the third China National Stroke Registry (CNSR-III; 2015-2018). The development cohort comprised 11,313 patients from CNSR-III. Validation cohort 1 (n = 2627 from CNSR-III; internal validation) were selected at random on the basis of hospital assignment. Validation cohort 2 (external validation) included 5158 patients with a history of ischaemic stroke, excluding those diagnosed with TIA, from the CHANCE-2 trial (2019-2021) done at 202 centres in China. Validation cohort 3 (external validation: disability and mortality) involved 3855 patients with a history of acute ischaemic stroke or TIA from a single centre in the Clinical Research Center for Stroke-Fifth Division (CRCS-5) cohort (2011-2024) in South Korea. We evaluated 309 hospitalisation variables, including baseline characteristics, medical history, hospitalisation data, biomarkers, geographical factors, National Institutes of Health Stroke Scale (NIHSS) scores, modified Rankin Scale (mRS), and stroke polygenic risk scores, using an extreme gradient boosting tree model. Feature importance was assessed via Shapley values. Primary outcomes were 3-month stroke recurrence, disability (mRS >2), and mortality. Secondary outcomes were assessed at six additional time points over 5 years. A nested cross-validation scheme was employed for feature selection and internal validation in 80% of patients (n = 11,313) from CNSR-III cohort. Testing was performed in the remaining 20% of patients (n = 2627) from CNSR-III (validation cohort 1). External validation was performed in validation cohorts 2 and 3.</p><p><strong>Findings: </strong>Global and domain-specific delta-NIHSS<sub>(admission-discharge)</sub> emerged as the strongest predictor of stroke recurrence, disability, and mortality. The Delta-NIHSS-based Predictor for Post-Stroke Composite Outcomes (DISCO) model, integrating 16 delta-NIHSS<sub>(admission-to-discharge)</sub> measures, age, gender, TOAST subtypes, history of TIA, stroke, diabetes and mRS at discharge, achieved the area under the receiver operating curve (AUCs) of 0.805 and 0.815 for recurrence and mortality and of 0.852 for disability at 3 months in validation cohort 2. The 1% patients with highest-risk exhibited a 19.27-fold relative risk (RR) for recurrence, 36.59-fold RR for disability, and 225.5-fold RR for mortality at 3 months. The DISCO model is available at https://www.discosysu.cn.</p><p><strong>Interpretation: </strong>The DISCO model, incorp","PeriodicalId":11393,"journal":{"name":"EClinicalMedicine","volume":"90 ","pages":"103684"},"PeriodicalIF":10.0,"publicationDate":"2025-12-09","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC12753274/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145877779","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}
Pub Date : 2025-12-08eCollection Date: 2025-12-01DOI: 10.1016/j.eclinm.2025.103671
Dongmei Tang, Dantong Gu, Jiamin Gong, Guangyu Liu, Lei Zhou, Aqiang Dai, Yan Huo, Pengfei Guan, Jianning Zhang, Xinsheng Huang, Yunfeng Wang, Shan Sun, Huawei Li
<p><strong>Background: </strong>Tinnitus is a common and often debilitating auditory condition with limited treatment options. While sound therapy is widely used, robust evidence from long-term randomised trials is scarce. We aimed to evaluate the 9-month efficacy and 3-month posttreatment durability of four sound therapies for adults with chronic subjective tinnitus and identify predictors of response.</p><p><strong>Methods: </strong>In this multicentre, double-blind, randomised controlled clinical trial, participants (aged 18-80 years) with chronic subjective tinnitus from three academic hospitals in China were included. Participants were randomly 1:1:1:1 assigned to receive one of four daily 2-h interventions: unmodified music (UM), UM plus pitch-centered narrowband noise (UM + NBN), high-frequency-enhanced music (HFEM), or digital frequency-customised relieving sound (DFCRS). Primary outcome was tinnitus severity assessed by Tinnitus Handicap Inventory (THI). Assessments occurred at baseline, 1, 2, 3, 6, and 9 months, with a 3-month posttreatment follow-up. Two prespecified primary endpoints were defined: (a) complete remission, operationalised as a THI score of 0 at any follow-up visit within the 9-month period. Participants achieving this endpoint were considered clinically cured, and sound therapy was discontinued; or (b) if complete remission was not achieved by the end of 9 months, the magnitude of improvement was defined as the change in THI score from baseline to the 9-month endpoint. The primary analysis followed the intention-to-treat (ITT) principle, This trial is registered with the Chinese Clinical Trial Registry, ChiCTR2000039007.</p><p><strong>Findings: </strong>Between May 14, 2021, and November 30, 2022, 440 participants (median age 45 years [IQR, 35-56]; 222/440 [50·5%] male; median tinnitus duration 13 months [IQR, 7-36]) were enrolled and randomly assigned (UM, n = 111; UM + NBN, n = 110; HFEM, n = 108; DFCRS, n = 111). Baseline characteristics were balanced between the groups. Only one participant in the HFEM group achieved complete remission, with a THI score of 0 at the 6-month follow-up. THI scores significantly decreased over time in all groups (median 50·00 [IQR 36·00-62·00]) at baseline to 9-month follow-up (35·00 [24·00-48·00]; <i>p</i> < 0·0001), with effects sustained posttreatment. Significant group × time interactions occurred (UM: F<sub>(5, 618)</sub> = 11·45; UM + NBN: F<sub>(5, 605)</sub> = 7·17; HFEM: F<sub>(5, 599)</sub> = 8·3; DFCRS: F<sub>(5, 619)</sub> = 12·65; all <i>p</i> < 0·0001) in all arms. DFCRS demonstrated superior efficacy (parameter estimate -4·37, 95% CI -6·25 to -2·48; <i>p</i> < 0·0001), when compared to UM as reference. No adverse events were reported in any group.</p><p><strong>Interpretation: </strong>In this exploratory trial, personalised acoustic therapy may provide promising efficacy for chronic tinnitus. Although interpretation is tempered by the absence of a blank control arm and o
背景：耳鸣是一种常见且常使人衰弱的听觉疾病，治疗方法有限。虽然声音疗法被广泛使用，但来自长期随机试验的有力证据很少。我们旨在评估四种声音疗法对成人慢性主观性耳鸣的9个月疗效和治疗后3个月的持久性，并确定反应的预测因素。方法：在这项多中心、双盲、随机对照临床试验中，来自中国三所学术医院的慢性主观性耳鸣患者（年龄18-80岁）被纳入研究。参与者被随机1:1:1:1分配接受四种每日2小时干预之一：未经修改的音乐（UM）， UM加以音高为中心的窄带噪声（UM + NBN），高频增强音乐（HFEM）或数字频率定制缓解声音（DFCRS）。主要终点是耳鸣障碍量表（THI）评估耳鸣严重程度。评估发生在基线、1、2、3、6和9个月，治疗后随访3个月。定义了两个预先指定的主要终点：(a)完全缓解，在9个月内的任何随访中，THI评分为0。达到这个终点的参与者被认为是临床治愈，声音治疗停止；或(b)如果在9个月结束时未达到完全缓解，则改善的幅度定义为THI评分从基线到9个月终点的变化。初步分析遵循意向治疗（ITT）原则，该试验已在中国临床试验注册中心注册，注册号为ChiCTR2000039007。研究结果：在2021年5月14日至2022年11月30日期间，440名参与者（中位年龄45岁[IQR， 35-56]; 222/440[50.5%]男性；中位耳鸣持续时间13个月[IQR， 7-36]）被纳入并随机分配（UM, n = 111; UM + NBN, n = 110； HFEM, n = 108； DFCRS, n = 111）。各组之间的基线特征平衡。在6个月的随访中，HFEM组中只有一名参与者达到了完全缓解，THI评分为0。在基线至9个月随访期间（35·00[24·00-48·00];p < 0.0001），各组THI评分均随时间显著下降（中位数为50·00 [IQR 36.00 -62·00]），治疗后效果持续。显著组×时间交互作用发生(UM: F(5,618) = 11·45；Um + nbn: f(5,605) = 7·17；Hfem: f(5,599) = 8·3；Dfcrs: f(5,619) = 12·65；所有实验组均p < 0.0001)。与作为参考的UM相比，DFCRS显示出更好的疗效（参数估计- 4.37,95% CI - 6.25至- 2.48;p < 0.0001）。两组均无不良事件报告。解释：在这项探索性试验中，个性化声学治疗可能为慢性耳鸣提供有希望的疗效。虽然由于缺乏空白对照组和客观依从性监测，解释受到了限制，但这些局限性突出了未来研究改进方法和更有力地验证治疗益处的机会。资助：国家科技部、上海市申康发展中心、上海市科学技术委员会、国家自然科学基金。
{"title":"Efficacy and safety of a modified sound therapy for patients with subjective tinnitus (MOST): a multicentre, double-blind, randomised controlled trial.","authors":"Dongmei Tang, Dantong Gu, Jiamin Gong, Guangyu Liu, Lei Zhou, Aqiang Dai, Yan Huo, Pengfei Guan, Jianning Zhang, Xinsheng Huang, Yunfeng Wang, Shan Sun, Huawei Li","doi":"10.1016/j.eclinm.2025.103671","DOIUrl":"10.1016/j.eclinm.2025.103671","url":null,"abstract":"<p><strong>Background: </strong>Tinnitus is a common and often debilitating auditory condition with limited treatment options. While sound therapy is widely used, robust evidence from long-term randomised trials is scarce. We aimed to evaluate the 9-month efficacy and 3-month posttreatment durability of four sound therapies for adults with chronic subjective tinnitus and identify predictors of response.</p><p><strong>Methods: </strong>In this multicentre, double-blind, randomised controlled clinical trial, participants (aged 18-80 years) with chronic subjective tinnitus from three academic hospitals in China were included. Participants were randomly 1:1:1:1 assigned to receive one of four daily 2-h interventions: unmodified music (UM), UM plus pitch-centered narrowband noise (UM + NBN), high-frequency-enhanced music (HFEM), or digital frequency-customised relieving sound (DFCRS). Primary outcome was tinnitus severity assessed by Tinnitus Handicap Inventory (THI). Assessments occurred at baseline, 1, 2, 3, 6, and 9 months, with a 3-month posttreatment follow-up. Two prespecified primary endpoints were defined: (a) complete remission, operationalised as a THI score of 0 at any follow-up visit within the 9-month period. Participants achieving this endpoint were considered clinically cured, and sound therapy was discontinued; or (b) if complete remission was not achieved by the end of 9 months, the magnitude of improvement was defined as the change in THI score from baseline to the 9-month endpoint. The primary analysis followed the intention-to-treat (ITT) principle, This trial is registered with the Chinese Clinical Trial Registry, ChiCTR2000039007.</p><p><strong>Findings: </strong>Between May 14, 2021, and November 30, 2022, 440 participants (median age 45 years [IQR, 35-56]; 222/440 [50·5%] male; median tinnitus duration 13 months [IQR, 7-36]) were enrolled and randomly assigned (UM, n = 111; UM + NBN, n = 110; HFEM, n = 108; DFCRS, n = 111). Baseline characteristics were balanced between the groups. Only one participant in the HFEM group achieved complete remission, with a THI score of 0 at the 6-month follow-up. THI scores significantly decreased over time in all groups (median 50·00 [IQR 36·00-62·00]) at baseline to 9-month follow-up (35·00 [24·00-48·00]; <i>p</i> < 0·0001), with effects sustained posttreatment. Significant group × time interactions occurred (UM: F<sub>(5, 618)</sub> = 11·45; UM + NBN: F<sub>(5, 605)</sub> = 7·17; HFEM: F<sub>(5, 599)</sub> = 8·3; DFCRS: F<sub>(5, 619)</sub> = 12·65; all <i>p</i> < 0·0001) in all arms. DFCRS demonstrated superior efficacy (parameter estimate -4·37, 95% CI -6·25 to -2·48; <i>p</i> < 0·0001), when compared to UM as reference. No adverse events were reported in any group.</p><p><strong>Interpretation: </strong>In this exploratory trial, personalised acoustic therapy may provide promising efficacy for chronic tinnitus. Although interpretation is tempered by the absence of a blank control arm and o","PeriodicalId":11393,"journal":{"name":"EClinicalMedicine","volume":"90 ","pages":"103671"},"PeriodicalIF":10.0,"publicationDate":"2025-12-08","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC12766481/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145910957","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}
Pub Date : 2025-12-05eCollection Date: 2025-12-01DOI: 10.1016/j.eclinm.2025.103688
Tingyu Wang, Yuting Yan, Hui Wang, Ying Sun, Xiantao Liu, Rui Lyu, Wenjie Xiong, Gang An, Wei Liu, Yan Xu, Shuhui Deng, Qi Wang, Chenxing Du, Liang Huang, Dehui Zou, Yaozhong Zhao, Lugui Qiu, Zengjun Li, Shuhua Yi
Background: Effective time-limited therapies remain an unmet need in chronic lymphocytic leukaemia (CLL). We evaluated a novel regimen combining ibrutinib with intermittent fludarabine, cyclophosphamide, and rituximab (FCR) in patients with CLL.
Methods: This single-arm, multicentre, phase 2 trial was conducted at three centres in China. Eligible participants were adults (aged 18-65 years) with previously untreated CLL requiring therapy according to the International Workshop on Chronic Lymphocytic Leukaemia criteria. The regimen consisted of ibrutinib (420 mg daily) combined with three cycles of FCR administered on days 1-3 of cycles 1, 5, and 9. After the completion of induction therapy, among patients who had achieved complete remission (CR) with undetectable minimal residual disease (uMRD): those without TP53 deletion or mutation could discontinue treatment; those with TP53 deletion or mutation continued maintenance therapy for 6 months before discontinuing treatment; all other patients continued ibrutinib until they achieved CR-uMRD, followed by an additional 6 months of treatment. The primary endpoint was CR rate at best response. This trial is registered with ClinicalTrials.gov, NCT03980002.
Findings: Between June 1, 2019, and July 30, 2023, 50 patients were enrolled with a median follow-up of 55 months (IQR 40-65). The median age of participants was 57 years (IQR 48-62), and 36 (72%) were male. Among 47 patients with available IGHV data, 18 (38%) had unmutated IGHV. TP53 deletion/mutation were present in four (8%) patients. The CR rate, the primary outcome, was 70% (35/50) at best response. 18 patients (36%) achieved CR with undetectable MRD in both bone marrow and peripheral blood, discontinuing treatment post-induction. Four deaths occurred; one due to Richter's transformation, two from COVID-19, and one from cerebral infarction. Grade 3/4 neutropenia occurred in 13 (26%) patients, leucopenia in 11 (22%) patients, and lymphocytopenia in 12 (24%) patients. No therapy-related cases of myelodysplastic syndrome or acute myeloid leukaemia were observed.
Interpretation: This regimen achieved deep remissions with manageable toxicity. Whilst acknowledging study limitations and that pathway inhibitors are the current standard of care, these findings provide proof-of-principle that this regimen could be a feasible, time-limited treatment option for CLL. Particularly in resource-limited healthcare systems where continuous novel agents are less accessible. Future studies should explore long-term durability beyond 5 years and validate MRD thresholds for treatment cessation.
Funding: National Nature Science Foundation of China, the Chinese Academy of Medical Sciences Innovation Fund for Medical Sciences, and Beijing Xisike Clinical Oncology Research Foundation.
{"title":"Ibrutinib alternating with three cycles of interval fludarabine, cyclophosphamide, and rituximab (FCR) in adults with untreated chronic lymphocytic leukaemia as time-limited regimen: a single-arm, multicentre phase 2 trial in China.","authors":"Tingyu Wang, Yuting Yan, Hui Wang, Ying Sun, Xiantao Liu, Rui Lyu, Wenjie Xiong, Gang An, Wei Liu, Yan Xu, Shuhui Deng, Qi Wang, Chenxing Du, Liang Huang, Dehui Zou, Yaozhong Zhao, Lugui Qiu, Zengjun Li, Shuhua Yi","doi":"10.1016/j.eclinm.2025.103688","DOIUrl":"10.1016/j.eclinm.2025.103688","url":null,"abstract":"<p><strong>Background: </strong>Effective time-limited therapies remain an unmet need in chronic lymphocytic leukaemia (CLL). We evaluated a novel regimen combining ibrutinib with intermittent fludarabine, cyclophosphamide, and rituximab (FCR) in patients with CLL.</p><p><strong>Methods: </strong>This single-arm, multicentre, phase 2 trial was conducted at three centres in China. Eligible participants were adults (aged 18-65 years) with previously untreated CLL requiring therapy according to the International Workshop on Chronic Lymphocytic Leukaemia criteria. The regimen consisted of ibrutinib (420 mg daily) combined with three cycles of FCR administered on days 1-3 of cycles 1, 5, and 9. After the completion of induction therapy, among patients who had achieved complete remission (CR) with undetectable minimal residual disease (uMRD): those without TP53 deletion or mutation could discontinue treatment; those with TP53 deletion or mutation continued maintenance therapy for 6 months before discontinuing treatment; all other patients continued ibrutinib until they achieved CR-uMRD, followed by an additional 6 months of treatment. The primary endpoint was CR rate at best response. This trial is registered with ClinicalTrials.gov, NCT03980002.</p><p><strong>Findings: </strong>Between June 1, 2019, and July 30, 2023, 50 patients were enrolled with a median follow-up of 55 months (IQR 40-65). The median age of participants was 57 years (IQR 48-62), and 36 (72%) were male. Among 47 patients with available IGHV data, 18 (38%) had unmutated IGHV. TP53 deletion/mutation were present in four (8%) patients. The CR rate, the primary outcome, was 70% (35/50) at best response. 18 patients (36%) achieved CR with undetectable MRD in both bone marrow and peripheral blood, discontinuing treatment post-induction. Four deaths occurred; one due to Richter's transformation, two from COVID-19, and one from cerebral infarction. Grade 3/4 neutropenia occurred in 13 (26%) patients, leucopenia in 11 (22%) patients, and lymphocytopenia in 12 (24%) patients. No therapy-related cases of myelodysplastic syndrome or acute myeloid leukaemia were observed.</p><p><strong>Interpretation: </strong>This regimen achieved deep remissions with manageable toxicity. Whilst acknowledging study limitations and that pathway inhibitors are the current standard of care, these findings provide proof-of-principle that this regimen could be a feasible, time-limited treatment option for CLL. Particularly in resource-limited healthcare systems where continuous novel agents are less accessible. Future studies should explore long-term durability beyond 5 years and validate MRD thresholds for treatment cessation.</p><p><strong>Funding: </strong>National Nature Science Foundation of China, the Chinese Academy of Medical Sciences Innovation Fund for Medical Sciences, and Beijing Xisike Clinical Oncology Research Foundation.</p>","PeriodicalId":11393,"journal":{"name":"EClinicalMedicine","volume":"90 ","pages":"103688"},"PeriodicalIF":10.0,"publicationDate":"2025-12-05","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC12766422/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145911012","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}
Pub Date : 2025-12-05eCollection Date: 2025-12-01DOI: 10.1016/j.eclinm.2025.103690
Hyolim Kang, Ahyoung Lim, Andrew Clark, Felipe J Colón González, Hannah Eleanor Clapham, Jean-Paul Carrera, Jong-Hoon Kim, Megan Auzenbergs, Preethi Lakshminarayanan, Sandra López-Vergès, So Yoon Sim, Su Myat Han, Thiago Cerqueira-Silva, Timothy Endy, Zulma M Cucunubá, W John Edmunds, Sushant Sahastrabuddhe, Oliver J Brady, Kaja Abbas
Background: Two chikungunya vaccines, Ixchiq and Vimkunya are licensed. In April 2025, Brazil is the first endemic country to license Ixchiq, but optimal age groups for vaccination remain unclear. Our aim is to model the public health impact of age-specific chikungunya outbreak response immunisation strategies in Brazil and infer broader implications for vaccine use case scenarios in outbreak prone regions.
Methods: We developed an age-structured transmission dynamic model calibrated with state-level Brazilian surveillance data for 2022 and long-term average annual force of infections. We simulated outbreak response immunisation strategies targeting ages 1-11, 12-17, 18-59, and ≥60 years for Ixchiq and Vimkunya across 11 out of 27 states in Brazil. We assessed vaccine impact by symptomatic cases, deaths, and disability-adjusted life years (DALYs) averted and number needed to vaccinate (NNV) based on vaccine protection against disease only and against both disease and infection.
Findings: Ixchiq and Vimkunya showed similar vaccine impact. Across strategies, vaccinating children 1-11 years yielded the lowest NNV for both vaccines, whereas vaccinating adults 18-59 years achieved the greatest absolute reduction in symptomatic cases, averting 62.5% (95% Uncertainty Intervals [UI]: 54.2-84.1) of total symptomatic cases with Vimkunya and 66.2% (58.2-86.0) with Ixchiq, under disease and infection blocking mechanism. Vaccinating adults 18-59 years with Ixchiq or Vimkunya yielded similar efficiency, with NNVs to avert a DALY of 339 (39-3412) and 361 (40-3777) respectively, under disease and infection-blocking mechanism.
Interpretation: Under current licensure, vaccinating adolescents aged 12-17 years first, followed by 18-59 years are efficient strategies, with similar NNVs for both Ixchiq and Vimkunya. If eligibility expands to younger populations, vaccinating 1-11-year age group will have relatively higher efficiency.
Funding: International Vaccine Institute and Japan Agency for Medical Research and Development.
{"title":"Age-specific chikungunya outbreak response immunisation strategies in Brazil: a modelling study.","authors":"Hyolim Kang, Ahyoung Lim, Andrew Clark, Felipe J Colón González, Hannah Eleanor Clapham, Jean-Paul Carrera, Jong-Hoon Kim, Megan Auzenbergs, Preethi Lakshminarayanan, Sandra López-Vergès, So Yoon Sim, Su Myat Han, Thiago Cerqueira-Silva, Timothy Endy, Zulma M Cucunubá, W John Edmunds, Sushant Sahastrabuddhe, Oliver J Brady, Kaja Abbas","doi":"10.1016/j.eclinm.2025.103690","DOIUrl":"10.1016/j.eclinm.2025.103690","url":null,"abstract":"<p><strong>Background: </strong>Two chikungunya vaccines, Ixchiq and Vimkunya are licensed. In April 2025, Brazil is the first endemic country to license Ixchiq, but optimal age groups for vaccination remain unclear. Our aim is to model the public health impact of age-specific chikungunya outbreak response immunisation strategies in Brazil and infer broader implications for vaccine use case scenarios in outbreak prone regions.</p><p><strong>Methods: </strong>We developed an age-structured transmission dynamic model calibrated with state-level Brazilian surveillance data for 2022 and long-term average annual force of infections. We simulated outbreak response immunisation strategies targeting ages 1-11, 12-17, 18-59, and ≥60 years for Ixchiq and Vimkunya across 11 out of 27 states in Brazil. We assessed vaccine impact by symptomatic cases, deaths, and disability-adjusted life years (DALYs) averted and number needed to vaccinate (NNV) based on vaccine protection against disease only and against both disease and infection.</p><p><strong>Findings: </strong>Ixchiq and Vimkunya showed similar vaccine impact. Across strategies, vaccinating children 1-11 years yielded the lowest NNV for both vaccines, whereas vaccinating adults 18-59 years achieved the greatest absolute reduction in symptomatic cases, averting 62.5% (95% Uncertainty Intervals [UI]: 54.2-84.1) of total symptomatic cases with Vimkunya and 66.2% (58.2-86.0) with Ixchiq, under disease and infection blocking mechanism. Vaccinating adults 18-59 years with Ixchiq or Vimkunya yielded similar efficiency, with NNVs to avert a DALY of 339 (39-3412) and 361 (40-3777) respectively, under disease and infection-blocking mechanism.</p><p><strong>Interpretation: </strong>Under current licensure, vaccinating adolescents aged 12-17 years first, followed by 18-59 years are efficient strategies, with similar NNVs for both Ixchiq and Vimkunya. If eligibility expands to younger populations, vaccinating 1-11-year age group will have relatively higher efficiency.</p><p><strong>Funding: </strong>International Vaccine Institute and Japan Agency for Medical Research and Development.</p>","PeriodicalId":11393,"journal":{"name":"EClinicalMedicine","volume":"90 ","pages":"103690"},"PeriodicalIF":10.0,"publicationDate":"2025-12-05","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC12766480/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145910946","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}
Pub Date : 2025-12-05eCollection Date: 2025-12-01DOI: 10.1016/j.eclinm.2025.103674
Erik Alonso, Raúl López-Izquierdo, Emma Bourke-Matas, Michael Eichinger, Carlos Del Pozo Vegas, Bas de Groot, Isabel de la Torre, Begoña Polonio-López, José Luis Martín-Conty, Ancor Sanz-García, Francisco Martín-Rodríguez
Background: Scoring systems have demonstrated their usefulness in predicting short-term mortality when applied by emergency medical services (EMS). However, their implementation should be supported by validation studies using real-world data. This work aims to validate a previously developed modified Sequential Organ Failure Assessment (mSOFA) score and conduct external revalidation for its use in prehospital critical care to predict short-term mortality.
Methods: This prospective, observational, multicentre, EMS-based validation and external validation study, was conducted across three EMS systems in Spain (one for validation and two for revalidation). Adults with undifferentiated acute conditions who were transported with high priority to the emergency department (ED), excluding pregnancy, cardiac arrest, and palliative patients. The primary outcome was 2-day all-cause in-hospital mortality. Prehospital and in-hospital demographic data, vital signs, and point-of-care testing variables were collected to calculate mSOFA and SOFA scores. Score performance was evaluated through validation/revalidation and a random quasi-stratified K-fold cross-validation scheme.
Findings: Between January 1, 2021, and March 30, 2025, a total of 12,212 patients were enrolled (validation cohort#1 (n = 9063), revalidation cohort#2 (n = 1816), and revalidation cohort#3 (n = 1333)). The median age was 67 years (IQR: 51-80), and 41.1% (5040 patients) were female. The overall 2-day mortality rate was 5% (609 cases). The mSOFA score showed an AUC of 0.949 (95% CI: 0.939-0.958) in the validation cohort, and 0.939 (95% CI: 0.925-0.954) and 0.944 (95% CI: 0.921-0.967) in the two revalidation cohorts. Comparison between mSOFA and SOFA scores for 2-day mortality revealed statistically significant differences (p < 0.0001), with AUCs of 0.947 (95% CI: 0.939-0.954) for mSOFA and 0.927 (95% CI: 0.917-0.937) for SOFA.
Interpretation: Our findings suggest that the mSOFA score shows good discriminatory power for predicting short-term mortality, consistent across different cohorts. Study limitations included not blinded extractors, time limit of primary outcome, and patients' heterogeneity. This validates score performance in other health systems and demonstrates a better mSOFA performance over the SOFA score. Future work will pursue clinical validation of the score via a randomized clinical trial.
Funding: This work was supported by the Institute of Health Carlos III, co-financed by the European Union, by the Basque Government, and by the University of the Basque Country.
{"title":"Performance of a modified Sequential Organ Failure Assessment score in pre-hospital critical care to predict short-term mortality: a prospective, multicentre, validation cohort study.","authors":"Erik Alonso, Raúl López-Izquierdo, Emma Bourke-Matas, Michael Eichinger, Carlos Del Pozo Vegas, Bas de Groot, Isabel de la Torre, Begoña Polonio-López, José Luis Martín-Conty, Ancor Sanz-García, Francisco Martín-Rodríguez","doi":"10.1016/j.eclinm.2025.103674","DOIUrl":"10.1016/j.eclinm.2025.103674","url":null,"abstract":"<p><strong>Background: </strong>Scoring systems have demonstrated their usefulness in predicting short-term mortality when applied by emergency medical services (EMS). However, their implementation should be supported by validation studies using real-world data. This work aims to validate a previously developed modified Sequential Organ Failure Assessment (mSOFA) score and conduct external revalidation for its use in prehospital critical care to predict short-term mortality.</p><p><strong>Methods: </strong>This prospective, observational, multicentre, EMS-based validation and external validation study, was conducted across three EMS systems in Spain (one for validation and two for revalidation). Adults with undifferentiated acute conditions who were transported with high priority to the emergency department (ED), excluding pregnancy, cardiac arrest, and palliative patients. The primary outcome was 2-day all-cause in-hospital mortality. Prehospital and in-hospital demographic data, vital signs, and point-of-care testing variables were collected to calculate mSOFA and SOFA scores. Score performance was evaluated through validation/revalidation and a random quasi-stratified K-fold cross-validation scheme.</p><p><strong>Findings: </strong>Between January 1, 2021, and March 30, 2025, a total of 12,212 patients were enrolled (validation cohort#1 (n = 9063), revalidation cohort#2 (n = 1816), and revalidation cohort#3 (n = 1333)). The median age was 67 years (IQR: 51-80), and 41.1% (5040 patients) were female. The overall 2-day mortality rate was 5% (609 cases). The mSOFA score showed an AUC of 0.949 (95% CI: 0.939-0.958) in the validation cohort, and 0.939 (95% CI: 0.925-0.954) and 0.944 (95% CI: 0.921-0.967) in the two revalidation cohorts. Comparison between mSOFA and SOFA scores for 2-day mortality revealed statistically significant differences (p < 0.0001), with AUCs of 0.947 (95% CI: 0.939-0.954) for mSOFA and 0.927 (95% CI: 0.917-0.937) for SOFA.</p><p><strong>Interpretation: </strong>Our findings suggest that the mSOFA score shows good discriminatory power for predicting short-term mortality, consistent across different cohorts. Study limitations included not blinded extractors, time limit of primary outcome, and patients' heterogeneity. This validates score performance in other health systems and demonstrates a better mSOFA performance over the SOFA score. Future work will pursue clinical validation of the score via a randomized clinical trial.</p><p><strong>Funding: </strong>This work was supported by the Institute of Health Carlos III, co-financed by the European Union, by the Basque Government, and by the University of the Basque Country.</p>","PeriodicalId":11393,"journal":{"name":"EClinicalMedicine","volume":"90 ","pages":"103674"},"PeriodicalIF":10.0,"publicationDate":"2025-12-05","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC12721286/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145818664","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}
Pub Date : 2025-12-05eCollection Date: 2025-12-01DOI: 10.1016/j.eclinm.2025.103692
Nadav Liam Modlin, Victoria Williamson, Guy M Goodwin, Ekaterina Malievskaia, Merve Atli, Zsofia Elek, Alice Gaillard, Don Koelpin, Anthony Cleare, Manish Agrawal, Rachel Yehuda, Namik Kirlic, James Rucker
<p><strong>Background: </strong>Post-traumatic stress disorder (PTSD) is a debilitating condition leading to significant personal and societal burden. Standard treatments frequently demonstrate limited efficacy, leading to persistent symptoms and high dropout rates. Psilocybin has shown promise in treating depression, a condition that is often comorbid with PTSD. We aimed to explore participant experiences of psilocybin treatment for PTSD, emphasising the role of monitoring and support for safety, direct and indirect engagement with trauma-related material during psilocybin treatment, and differences between psilocybin and standard treatments.</p><p><strong>Methods: </strong>This qualitative study was nested within a quantitative, open-label phase 2 trial assessing the safety and tolerability of COMP360 psilocybin in adults with PTSD. Eligible participants were adults (18 years or older) who met Diagnostic and Statistical Manual of Mental Disorders, Fifth Edition (DSM-5) diagnostic criteria for PTSD secondary to a traumatic event experienced in adulthood. Recruitment was conducted at three sites across two countries: two in the United States and one in the United Kingdom. Enrolled participants underwent standardised preparation, a single psilocybin administration session, and follow-up integration sessions. Semi-structured interviews were conducted before, the day after, and 12 weeks post-treatment. Data were analysed using reflexive thematic analysis, which is a distinct and theoretically grounded approach to co-construction of recurring themes pertaining to participants' preparedness for treatment, how participants' index trauma presented during treatment, and how psilocybin compared to standard treatments. The quantitative phase 2 trial, which the present qualitative study is nested within, is registered with ClinicalTrials.gov, NCT05312151.</p><p><strong>Findings: </strong>Between June 10, 2022, and Feb 12, 2024, 21 participants were enrolled and participated in this qualitative sub-study and completed the in-person qualitative interviews. The analysis revealed four core themes: (1) non-pharmacological factors for psychological safety and trust, (2) the experiential nature of psilocybin treatment, (3) engagement with trauma-related material during psilocybin treatment, and (4) comparative reflections on prior therapies and psilocybin treatment. Emphasising safety, treatment education, and informed consent, the treatment facilitated an experience of both direct and indirect engagement with trauma-related material during psilocybin treatment. Unlike standard treatments requiring direct confrontation with trauma memories, psilocybin appears to enable a broader, indirect engagement with traumatic material via a range of affective, somatic and self-transcendent experiences (e.g., moments of perceived unity, dissolution of self, or felt connection with a larger whole).</p><p><strong>Interpretation: </strong>Our qualitative findings suggests that ps
{"title":"Investigational psilocybin treatment for post-traumatic stress disorder: a qualitative study of participant experience, trauma engagement, and differences from standard treatment.","authors":"Nadav Liam Modlin, Victoria Williamson, Guy M Goodwin, Ekaterina Malievskaia, Merve Atli, Zsofia Elek, Alice Gaillard, Don Koelpin, Anthony Cleare, Manish Agrawal, Rachel Yehuda, Namik Kirlic, James Rucker","doi":"10.1016/j.eclinm.2025.103692","DOIUrl":"10.1016/j.eclinm.2025.103692","url":null,"abstract":"<p><strong>Background: </strong>Post-traumatic stress disorder (PTSD) is a debilitating condition leading to significant personal and societal burden. Standard treatments frequently demonstrate limited efficacy, leading to persistent symptoms and high dropout rates. Psilocybin has shown promise in treating depression, a condition that is often comorbid with PTSD. We aimed to explore participant experiences of psilocybin treatment for PTSD, emphasising the role of monitoring and support for safety, direct and indirect engagement with trauma-related material during psilocybin treatment, and differences between psilocybin and standard treatments.</p><p><strong>Methods: </strong>This qualitative study was nested within a quantitative, open-label phase 2 trial assessing the safety and tolerability of COMP360 psilocybin in adults with PTSD. Eligible participants were adults (18 years or older) who met Diagnostic and Statistical Manual of Mental Disorders, Fifth Edition (DSM-5) diagnostic criteria for PTSD secondary to a traumatic event experienced in adulthood. Recruitment was conducted at three sites across two countries: two in the United States and one in the United Kingdom. Enrolled participants underwent standardised preparation, a single psilocybin administration session, and follow-up integration sessions. Semi-structured interviews were conducted before, the day after, and 12 weeks post-treatment. Data were analysed using reflexive thematic analysis, which is a distinct and theoretically grounded approach to co-construction of recurring themes pertaining to participants' preparedness for treatment, how participants' index trauma presented during treatment, and how psilocybin compared to standard treatments. The quantitative phase 2 trial, which the present qualitative study is nested within, is registered with ClinicalTrials.gov, NCT05312151.</p><p><strong>Findings: </strong>Between June 10, 2022, and Feb 12, 2024, 21 participants were enrolled and participated in this qualitative sub-study and completed the in-person qualitative interviews. The analysis revealed four core themes: (1) non-pharmacological factors for psychological safety and trust, (2) the experiential nature of psilocybin treatment, (3) engagement with trauma-related material during psilocybin treatment, and (4) comparative reflections on prior therapies and psilocybin treatment. Emphasising safety, treatment education, and informed consent, the treatment facilitated an experience of both direct and indirect engagement with trauma-related material during psilocybin treatment. Unlike standard treatments requiring direct confrontation with trauma memories, psilocybin appears to enable a broader, indirect engagement with traumatic material via a range of affective, somatic and self-transcendent experiences (e.g., moments of perceived unity, dissolution of self, or felt connection with a larger whole).</p><p><strong>Interpretation: </strong>Our qualitative findings suggests that ps","PeriodicalId":11393,"journal":{"name":"EClinicalMedicine","volume":"90 ","pages":"103692"},"PeriodicalIF":10.0,"publicationDate":"2025-12-05","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC12766423/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145910834","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}
Pub Date : 2025-12-05eCollection Date: 2025-12-01DOI: 10.1016/j.eclinm.2025.103691
Xue Feng, Lei Sun, Jintao Zhu, Xuan Yao, Zhongheng Zhang, Qing Pan, Luping Fang, Gangmin Ning
<p><strong>Background: </strong>Sepsis is a heterogeneous syndrome with varying degrees of multi-organ dysfunction. Identifying dynamic inter-organ interactions is critical for accurate sepsis subphenotyping and targeted therapy, yet remains unexplored. In this study, we aimed to quantify the dynamic trajectories of organ interactions to define sepsis phenotypes, supporting personalized treatment and clinical decision-making.</p><p><strong>Methods: </strong>We proposed a novel deep temporal graph clustering model to identify sepsis phenotypes by quantifying dynamic multi-organ interactions within 48 h post-diagnosis. The model was trained and validated on the Medical Information Mart for Intensive Care III (MIMIC-III) dataset (admissions from 2001 to 2012) and externally validated on the eICU Collaborative Research (eICU) dataset (admissions from 2014 to 2015). Its effectiveness was benchmarked against state-of-the-art clustering algorithms. Patient characteristics, multi-organ system states coupling patterns, and prognostic outcomes were compared across the identified phenotypes. Extreme gradient boosting (XGBoost) was used for early phenotype classification at 4 h post-diagnosis. To enhance clinical applicability, a user-friendly web interface was developed. Propensity score matching and weighted logistic regression were employed to evaluate the effects of the fluid management strategies on in-hospital mortality of patients with various phenotypes.</p><p><strong>Findings: </strong>A total of 10,181 and 6208 unique sepsis patients were employed as the cohorts for the model development and external validation, respectively. Three distinct phenotypes were identified and labeled as Phenotype A, B, and C, exhibiting significant differences in baseline characteristics, organ system states coupling patterns, and outcomes (P-value < 0.05). Phenotype A had the lowest mortality (5.59%) and accounted for the largest proportion of patients (46.34%). In contrast, Phenotype C represented the highest mortality (38.27%) and comprised the smallest proportion (22.78%). Phenotype A was characterized by sustained synchronous improvement across organ system states. Phenotype B showed persistent decoupling of organ system states. Phenotype C exhibited a rapid transition from early asynchrony to synchronization. The model demonstrated robust clustering performance in external validation. The simplified classifier showed high predictive performance, achieving an area under the receiver operating characteristic curve (AUROC) of 0.84 (95% CI [0.83, 0.86]) for phenotype prediction at 4 h post-diagnosis. The beneficial fluid management strategies varied across different phenotypes, highlighting the need for targeted fluid intervals.</p><p><strong>Interpretation: </strong>This study characterizes sepsis phenotypes using organ interaction trajectories and identifies three heterogeneous patterns of disease progression. These patterns offer new insights into the underlying pa
{"title":"Subphenotyping sepsis based on organ interaction trajectory using a deep temporal graph clustering model: a retrospective cohort study.","authors":"Xue Feng, Lei Sun, Jintao Zhu, Xuan Yao, Zhongheng Zhang, Qing Pan, Luping Fang, Gangmin Ning","doi":"10.1016/j.eclinm.2025.103691","DOIUrl":"10.1016/j.eclinm.2025.103691","url":null,"abstract":"<p><strong>Background: </strong>Sepsis is a heterogeneous syndrome with varying degrees of multi-organ dysfunction. Identifying dynamic inter-organ interactions is critical for accurate sepsis subphenotyping and targeted therapy, yet remains unexplored. In this study, we aimed to quantify the dynamic trajectories of organ interactions to define sepsis phenotypes, supporting personalized treatment and clinical decision-making.</p><p><strong>Methods: </strong>We proposed a novel deep temporal graph clustering model to identify sepsis phenotypes by quantifying dynamic multi-organ interactions within 48 h post-diagnosis. The model was trained and validated on the Medical Information Mart for Intensive Care III (MIMIC-III) dataset (admissions from 2001 to 2012) and externally validated on the eICU Collaborative Research (eICU) dataset (admissions from 2014 to 2015). Its effectiveness was benchmarked against state-of-the-art clustering algorithms. Patient characteristics, multi-organ system states coupling patterns, and prognostic outcomes were compared across the identified phenotypes. Extreme gradient boosting (XGBoost) was used for early phenotype classification at 4 h post-diagnosis. To enhance clinical applicability, a user-friendly web interface was developed. Propensity score matching and weighted logistic regression were employed to evaluate the effects of the fluid management strategies on in-hospital mortality of patients with various phenotypes.</p><p><strong>Findings: </strong>A total of 10,181 and 6208 unique sepsis patients were employed as the cohorts for the model development and external validation, respectively. Three distinct phenotypes were identified and labeled as Phenotype A, B, and C, exhibiting significant differences in baseline characteristics, organ system states coupling patterns, and outcomes (P-value < 0.05). Phenotype A had the lowest mortality (5.59%) and accounted for the largest proportion of patients (46.34%). In contrast, Phenotype C represented the highest mortality (38.27%) and comprised the smallest proportion (22.78%). Phenotype A was characterized by sustained synchronous improvement across organ system states. Phenotype B showed persistent decoupling of organ system states. Phenotype C exhibited a rapid transition from early asynchrony to synchronization. The model demonstrated robust clustering performance in external validation. The simplified classifier showed high predictive performance, achieving an area under the receiver operating characteristic curve (AUROC) of 0.84 (95% CI [0.83, 0.86]) for phenotype prediction at 4 h post-diagnosis. The beneficial fluid management strategies varied across different phenotypes, highlighting the need for targeted fluid intervals.</p><p><strong>Interpretation: </strong>This study characterizes sepsis phenotypes using organ interaction trajectories and identifies three heterogeneous patterns of disease progression. These patterns offer new insights into the underlying pa","PeriodicalId":11393,"journal":{"name":"EClinicalMedicine","volume":"90 ","pages":"103691"},"PeriodicalIF":10.0,"publicationDate":"2025-12-05","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC12766501/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145910837","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}
Pub Date : 2025-12-05eCollection Date: 2025-12-01DOI: 10.1016/j.eclinm.2025.103694
Yulu Liu, Haoquan Chen, Jiaqi Zhao, Lieji Wang, Zimei Gao, Yuan Peng, Wenhui Ren, Kaining Shi, Sen Zhang, Xiufeng Chen, Chenxin Pang, Manle Yan, Caizhen Feng, Fan Chai, Jin Gao, Xiaoxuan Jia, Shu Wang, Nan Hong, Yi Wang
<p><strong>Background: </strong>Contrast-enhanced breast MRI, while highly sensitive, faces limitations including complexity, long acquisition times, and reliance on gadolinium-based contrast agents. Noncontrast diffusion-weighted imaging (DWI) offers an ultrafast alternative, but its diagnostic accuracy has been insufficient for standalone use. We investigated whether a deep learning (DL) model could enable accurate breast cancer diagnosis using only DWI.</p><p><strong>Methods: </strong>This four-centre study included 2493 patients with pathologically confirmed breast lesions. A DWI-based model (DWI-DL) was developed using data from 1286 patients at Peking University People's Hospital (PKUPH; January 2015 to July 2021), who were randomly divided into three cohorts at a ratio of 6:2:2 (train n = 774, validation n = 256, test n = 256). Independent testing used three external cohorts (n = 661) and a prospective cohort from PKUPH that was retrospectively analysed (n = 546; August 2021 to September 2022). The diagnostic performance of DWI-DL model was compared to that of an abbreviated enhanced-based (AE-DL) model and two expert radiologists. Subgroup analysis assessed performance across external and prospective cohorts. A multireader multicase validation, using DW-DL model and selective contrast-enhanced workflow, was conducted by 12 radiologists from ten institutions to evaluate the clinical utility of DWI-DL in assisting diagnosis. Performance was assessed using the area under the curve (AUC) of receiver operating characteristics. The multireader multicase study was registered with the China Clinical Trial Registry (ChiCTR2500095953).</p><p><strong>Findings: </strong>DWI-DL model demonstrated diagnostic performance across all cohorts that was comparable to the AE-DL model (AUC: 0.771-0.912 vs. 0.780-0.898, p = 0.36-0.98). DWI-DL outperformed two expert radiologists interpreting DWI alone (AUC: 0.781-0.858 vs. 0.714-0.770, p range <0.0001-0.023). In the multireader multicase validation, the AI-guided selective sequence protocol was non-inferior to the full protocol (Protocol E vs. C, AUC: 0.834 [95% CI: 0.785, 0.883] vs. 0.835 [95% CI: 0.789, 0.881], difference: -0.001 [95% CI: -0.029, 0.027], p = 0.94), while simultaneously reduced interpretation time by 55.5% (Protocol E vs. C, mean time: 59.6 [43.5] vs. 134.0 [72.4] seconds, p < 0.0001).</p><p><strong>Interpretation: </strong>A DL model using only non-contrast DWI can accurately diagnose breast cancer. The DWI-DL model demonstrated robust performance, comparable to the AE-DL model and surpassing human experts in DWI interpretation, while reducing interpretation time. However, its performance was lower than that of expert radiologists interpreting standard breast MRI. By improving diagnostic efficiency without affecting accuracy, our workflow cooperating the DWI-DL model and selective contrast-enhanced sequence presents a promising, rapid, and safe tool with the potential to streamline the clinic
背景：乳腺造影增强MRI虽然高度敏感，但也存在复杂性、采集时间长、依赖钆基造影剂等局限性。非对比弥散加权成像（DWI）提供了一种超快的替代方法，但其诊断准确性不足以单独使用。我们研究了深度学习（DL）模型是否可以仅使用DWI进行准确的乳腺癌诊断。方法：本研究纳入2493例经病理证实的乳腺病变患者。采用2015年1月至2021年7月北京大学人民医院1286例患者的数据，按6:2:2的比例随机分为3个队列（训练n = 774，验证n = 256，检验n = 256），建立基于dwi的模型（DWI-DL）。独立测试使用三个外部队列（n = 661）和一个来自PKUPH的前瞻性队列进行回顾性分析（n = 546； 2021年8月至2022年9月）。将DWI-DL模型的诊断性能与一种简化的增强模型（AE-DL）和两名放射科专家的诊断性能进行比较。亚组分析评估了外部和前瞻性队列的表现。来自10家机构的12名放射科医生使用DW-DL模型和选择性对比增强工作流程进行了多阅读器多病例验证，以评估DWI-DL在辅助诊断方面的临床应用。使用接收器工作特性的曲线下面积（AUC）评估性能。该多阅读器多病例研究已在中国临床试验注册中心注册（ChiCTR2500095953）。结果：DWI-DL模型在所有队列中的诊断性能与AE-DL模型相当（AUC: 0.771-0.912 vs. 0.780-0.898, p = 0.36-0.98）。DWI-DL优于单独使用DWI的两名放射科专家（AUC: 0.781-0.858 vs. 0.714-0.770， p范围）解释：仅使用非对比DWI的DL模型可以准确诊断乳腺癌。DWI- dl模型表现出强大的性能，与AE-DL模型相当，在DWI解释方面超过了人类专家，同时减少了解释时间。然而，它的表现低于专家放射科医生解释标准乳房MRI。通过提高诊断效率而不影响准确性，我们的工作流程与DWI-DL模型和选择性对比增强序列相结合，提供了一种有前途的、快速的、安全的工具，有可能简化乳腺癌的临床诊断。基金资助：国家自然科学基金（82471964）和北京大学人民医院科研发展基金（RDGS2022-10）资助。
{"title":"AI-driven diffusion weighted imaging-based non-contrast protocol for breast cancer diagnosis: a multicentre, multidimensional validation study.","authors":"Yulu Liu, Haoquan Chen, Jiaqi Zhao, Lieji Wang, Zimei Gao, Yuan Peng, Wenhui Ren, Kaining Shi, Sen Zhang, Xiufeng Chen, Chenxin Pang, Manle Yan, Caizhen Feng, Fan Chai, Jin Gao, Xiaoxuan Jia, Shu Wang, Nan Hong, Yi Wang","doi":"10.1016/j.eclinm.2025.103694","DOIUrl":"10.1016/j.eclinm.2025.103694","url":null,"abstract":"<p><strong>Background: </strong>Contrast-enhanced breast MRI, while highly sensitive, faces limitations including complexity, long acquisition times, and reliance on gadolinium-based contrast agents. Noncontrast diffusion-weighted imaging (DWI) offers an ultrafast alternative, but its diagnostic accuracy has been insufficient for standalone use. We investigated whether a deep learning (DL) model could enable accurate breast cancer diagnosis using only DWI.</p><p><strong>Methods: </strong>This four-centre study included 2493 patients with pathologically confirmed breast lesions. A DWI-based model (DWI-DL) was developed using data from 1286 patients at Peking University People's Hospital (PKUPH; January 2015 to July 2021), who were randomly divided into three cohorts at a ratio of 6:2:2 (train n = 774, validation n = 256, test n = 256). Independent testing used three external cohorts (n = 661) and a prospective cohort from PKUPH that was retrospectively analysed (n = 546; August 2021 to September 2022). The diagnostic performance of DWI-DL model was compared to that of an abbreviated enhanced-based (AE-DL) model and two expert radiologists. Subgroup analysis assessed performance across external and prospective cohorts. A multireader multicase validation, using DW-DL model and selective contrast-enhanced workflow, was conducted by 12 radiologists from ten institutions to evaluate the clinical utility of DWI-DL in assisting diagnosis. Performance was assessed using the area under the curve (AUC) of receiver operating characteristics. The multireader multicase study was registered with the China Clinical Trial Registry (ChiCTR2500095953).</p><p><strong>Findings: </strong>DWI-DL model demonstrated diagnostic performance across all cohorts that was comparable to the AE-DL model (AUC: 0.771-0.912 vs. 0.780-0.898, p = 0.36-0.98). DWI-DL outperformed two expert radiologists interpreting DWI alone (AUC: 0.781-0.858 vs. 0.714-0.770, p range <0.0001-0.023). In the multireader multicase validation, the AI-guided selective sequence protocol was non-inferior to the full protocol (Protocol E vs. C, AUC: 0.834 [95% CI: 0.785, 0.883] vs. 0.835 [95% CI: 0.789, 0.881], difference: -0.001 [95% CI: -0.029, 0.027], p = 0.94), while simultaneously reduced interpretation time by 55.5% (Protocol E vs. C, mean time: 59.6 [43.5] vs. 134.0 [72.4] seconds, p < 0.0001).</p><p><strong>Interpretation: </strong>A DL model using only non-contrast DWI can accurately diagnose breast cancer. The DWI-DL model demonstrated robust performance, comparable to the AE-DL model and surpassing human experts in DWI interpretation, while reducing interpretation time. However, its performance was lower than that of expert radiologists interpreting standard breast MRI. By improving diagnostic efficiency without affecting accuracy, our workflow cooperating the DWI-DL model and selective contrast-enhanced sequence presents a promising, rapid, and safe tool with the potential to streamline the clinic","PeriodicalId":11393,"journal":{"name":"EClinicalMedicine","volume":"90 ","pages":"103694"},"PeriodicalIF":10.0,"publicationDate":"2025-12-05","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC12766479/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145910938","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}
扫码关注我们
求助内容：
应助结果提醒方式：
京公网安备 11010802042870号