ESMO Real World Data and Digital Oncology最新文献_第3页

Efficacy of cabozantinib and sunitinib for the treatment of intermediate/poor risk renal cell carcinoma based upon UK real-world data 基于英国真实世界数据的卡博替尼和舒尼替尼治疗中危/低危肾细胞癌的疗效

ESMO Real World Data and Digital Oncology

Pub Date : 2024-10-18 DOI: 10.1016/j.esmorw.2024.100087

D. Lee , G.J. Melendez-Torres , A. Challapalli , R. Frazer , J. McGrane , A. Bahl

Background

The purpose of this study was to explore the effectiveness of cabozantinib versus sunitinib for the treatment of first-line metastatic renal cell carcinoma in intermediate/poor risk patients.

Materials and methods

Retrospective review of cases between 1 January 2018 and 30 June 2021 across 17 UK centres. Univariable and multivariable Cox proportional hazards modelling to identify prognostic factors. Inverse probability of treatment weighting, to estimate the causal effect of first-line treatment type.

Results

Cabozantinib patients (n = 106) had poorer risk status, less prior nephrectomy, shorter time to therapy, and more clear cell histology than sunitinib patients (n = 218). More sunitinib patients received a second or third line of subsequent treatment (56% and 23% versus 43% and 13%). Though there was no significant difference between treatments in overall survival (OS) or progression-free survival (PFS) across models, the difference in PFS bordered on significant in a multipredictor analysis (benefit in favour of cabozantinib; P = 0.06). When the Kaplan–Meier curves were stratified by risk status (intermediate versus poor), patients had similar OS within the risk groups. PFS appeared to differ with poor risk patients performing better on cabozantinib. Inverse probability of treatment weighting analysis showed little difference from the unadjusted results: OS hazard ratio = 1.119 (95% confidence interval (CI) 0.823-1.521); PFS hazard ratio 0.825 (95% CI 0.636-1.070) for cabozantinib versus sunitinib.

Conclusions

Our results showed no significant difference in either OS or PFS between treatments. Cabozantinib trended towards improved PFS and reduced OS. Decision-making for tyrosine kinase inhibitor monotherapy should consider later-line treatment options. This analysis is of particular relevance as sunitinib is now off-patent meaning that the cost of a course of treatment has considerably reduced.

背景本研究旨在探讨卡博替尼对舒尼替尼治疗中危/低危患者一线转移性肾细胞癌的有效性。采用单变量和多变量 Cox 比例危险度模型确定预后因素。结果与舒尼替尼患者（n = 218）相比，卡博赞替尼患者（n = 106）的风险状况更差、既往肾切除术更少、治疗时间更短、透明细胞组织学更多。接受二线或三线后续治疗的舒尼替尼患者更多（56% 和 23% 对 43% 和 13%）。虽然不同治疗方法在总生存期（OS）或无进展生存期（PFS）方面没有显著差异，但在多预测因子分析中，PFS的差异接近显著（卡博赞替尼获益；P = 0.06）。当 Kaplan-Meier 曲线按风险状态（中度风险与重度风险）分层时，各风险组患者的 OS 相似。PFS似乎存在差异，风险较低的患者使用卡博替尼后表现更好。治疗的逆概率加权分析显示，与未调整的结果差别不大：卡博替尼对舒尼替尼的OS危险比=1.119（95%置信区间（CI）0.823-1.521）；PFS危险比0.825（95% CI 0.636-1.070）。卡博替尼有改善 PFS 和降低 OS 的趋势。酪氨酸激酶抑制剂单药治疗的决策应考虑晚期治疗方案。由于舒尼替尼目前已过专利期，这意味着一个疗程的费用已大大降低，因此这项分析具有特别重要的意义。

{"title":"Efficacy of cabozantinib and sunitinib for the treatment of intermediate/poor risk renal cell carcinoma based upon UK real-world data","authors":"D. Lee , G.J. Melendez-Torres , A. Challapalli , R. Frazer , J. McGrane , A. Bahl","doi":"10.1016/j.esmorw.2024.100087","DOIUrl":"10.1016/j.esmorw.2024.100087","url":null,"abstract":"<div><h3>Background</h3><div>The purpose of this study was to explore the effectiveness of cabozantinib versus sunitinib for the treatment of first-line metastatic renal cell carcinoma in intermediate/poor risk patients.</div></div><div><h3>Materials and methods</h3><div>Retrospective review of cases between 1 January 2018 and 30 June 2021 across 17 UK centres. Univariable and multivariable Cox proportional hazards modelling to identify prognostic factors. Inverse probability of treatment weighting, to estimate the causal effect of first-line treatment type.</div></div><div><h3>Results</h3><div>Cabozantinib patients (<em>n</em> = 106) had poorer risk status, less prior nephrectomy, shorter time to therapy, and more clear cell histology than sunitinib patients (<em>n</em> = 218). More sunitinib patients received a second or third line of subsequent treatment (56% and 23% versus 43% and 13%). Though there was no significant difference between treatments in overall survival (OS) or progression-free survival (PFS) across models, the difference in PFS bordered on significant in a multipredictor analysis (benefit in favour of cabozantinib; <em>P</em> = 0.06). When the Kaplan–Meier curves were stratified by risk status (intermediate versus poor), patients had similar OS within the risk groups. PFS appeared to differ with poor risk patients performing better on cabozantinib. Inverse probability of treatment weighting analysis showed little difference from the unadjusted results: OS hazard ratio = 1.119 (95% confidence interval (CI) 0.823-1.521); PFS hazard ratio 0.825 (95% CI 0.636-1.070) for cabozantinib versus sunitinib.</div></div><div><h3>Conclusions</h3><div>Our results showed no significant difference in either OS or PFS between treatments. Cabozantinib trended towards improved PFS and reduced OS. Decision-making for tyrosine kinase inhibitor monotherapy should consider later-line treatment options. This analysis is of particular relevance as sunitinib is now off-patent meaning that the cost of a course of treatment has considerably reduced.</div></div>","PeriodicalId":100491,"journal":{"name":"ESMO Real World Data and Digital Oncology","volume":"6 ","pages":"Article 100087"},"PeriodicalIF":0.0,"publicationDate":"2024-10-18","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142529798","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Human-centered AI as a framework guiding the development of image-based diagnostic tools in oncology: a systematic review 以人为本的人工智能作为肿瘤学图像诊断工具开发的指导框架：系统综述

ESMO Real World Data and Digital Oncology

Pub Date : 2024-10-07 DOI: 10.1016/j.esmorw.2024.100077

K. Allen , A.K. Yawson , S. Haggenmüller , J.N. Kather , T.J. Brinker

Background

Artificial intelligence diagnostic tools (AIDTs) in oncology show high image classification accuracy but limited clinical adoption. Their adoption could be enhanced by (i) using user feedback during the software design, (ii) demonstrating that AIDTs improve the user’s decisions, and (iii) providing explanations of AI decisions tailored to the user, three aspects central to human-centered AI (HCAI). This review assesses these three aspects in AIDTs for oncology in general, exemplifying its concepts in the established field of skin cancer diagnostics as a specific use case.

Materials and methods

We carried out three Preferred Reporting Items for Systematic reviews and Meta-Analyses (PRISMA) searches using PubMed and ScienceDirect, limiting the results to articles published from 2019 to 2024. The first search focused on articles that used user feedback to develop AIDTs. The second search addressed whether AIDT improves dermatologists’ decisions. The third search targeted explainable AI in skin cancer.

Results

Five studies incorporated user feedback in AIDT design for cancer. Zooming in on AIDT for skin cancer, nine studies (3/37 in 2019, 3/93 in 2023) indicated that AIDTs improve dermatologists’ decisions in experimental (n = 5) and clinical settings (n = 1). Explainable AI was common in skin cancer diagnostics (n = 26), with papers assessing the user’s preference for explainable AI (XAI) methods or the impact of XAI on the user’s trust in AI diagnosis.

Conclusions

User feedback has been used to develop AIDTs tailored to clinicians’ needs. Evidence shows that AIDTs can improve clinicians’ decisions. This, combined with XAI, increases clinicians’ trust in AIDTs, potentially favoring their widespread usage.

背景肿瘤学领域的人工智能诊断工具（AIDT）显示出很高的图像分类准确性，但临床应用却很有限。可以通过以下方式提高其采用率：(i) 在软件设计过程中使用用户反馈；(ii) 证明人工智能诊断工具能改善用户的决策；(iii) 提供针对用户的人工智能决策解释，这三个方面是以人为本的人工智能（HCAI）的核心。本综述从总体上评估了肿瘤学 AIDTs 的这三个方面，并以皮肤癌诊断这一成熟领域的具体使用案例为例说明了其概念。材料与方法我们使用 PubMed 和 ScienceDirect 进行了三次系统综述和 Meta 分析首选报告项目（PRISMA）检索，检索结果仅限于 2019 年至 2024 年发表的文章。第一项检索侧重于使用用户反馈开发 AIDT 的文章。第二项搜索涉及 AIDT 是否改善了皮肤科医生的决策。第三次搜索针对皮肤癌中的可解释人工智能。结果五项研究将用户反馈纳入了癌症AIDT的设计中。在皮肤癌 AIDT 方面，9 项研究（2019 年 3/37 项，2023 年 3/93 项）表明，AIDT 在实验（5 项）和临床（1 项）环境中改善了皮肤科医生的决策。可解释人工智能在皮肤癌诊断中很常见（n = 26），有论文评估了用户对可解释人工智能（XAI）方法的偏好或 XAI 对用户对人工智能诊断信任度的影响。有证据表明，人工智能诊断可改善临床医生的决策。这一点与 XAI 相结合，增加了临床医生对 AIDTs 的信任，从而有可能促进 AIDTs 的广泛使用。

{"title":"Human-centered AI as a framework guiding the development of image-based diagnostic tools in oncology: a systematic review","authors":"K. Allen , A.K. Yawson , S. Haggenmüller , J.N. Kather , T.J. Brinker","doi":"10.1016/j.esmorw.2024.100077","DOIUrl":"10.1016/j.esmorw.2024.100077","url":null,"abstract":"<div><h3>Background</h3><div>Artificial intelligence diagnostic tools (AIDTs) in oncology show high image classification accuracy but limited clinical adoption. Their adoption could be enhanced by (i) using user feedback during the software design, (ii) demonstrating that AIDTs improve the user’s decisions, and (iii) providing explanations of AI decisions tailored to the user, three aspects central to human-centered AI (HCAI). This review assesses these three aspects in AIDTs for oncology in general, exemplifying its concepts in the established field of skin cancer diagnostics as a specific use case.</div></div><div><h3>Materials and methods</h3><div>We carried out three Preferred Reporting Items for Systematic reviews and Meta-Analyses (PRISMA) searches using PubMed and ScienceDirect, limiting the results to articles published from 2019 to 2024. The first search focused on articles that used user feedback to develop AIDTs. The second search addressed whether AIDT improves dermatologists’ decisions. The third search targeted explainable AI in skin cancer.</div></div><div><h3>Results</h3><div>Five studies incorporated user feedback in AIDT design for cancer. Zooming in on AIDT for skin cancer, nine studies (3/37 in 2019, 3/93 in 2023) indicated that AIDTs improve dermatologists’ decisions in experimental (<em>n</em> = 5) and clinical settings (<em>n</em> = 1). Explainable AI was common in skin cancer diagnostics (<em>n</em> = 26), with papers assessing the user’s preference for explainable AI (XAI) methods or the impact of XAI on the user’s trust in AI diagnosis.</div></div><div><h3>Conclusions</h3><div>User feedback has been used to develop AIDTs tailored to clinicians’ needs. Evidence shows that AIDTs can improve clinicians’ decisions. This, combined with XAI, increases clinicians’ trust in AIDTs, potentially favoring their widespread usage.</div></div>","PeriodicalId":100491,"journal":{"name":"ESMO Real World Data and Digital Oncology","volume":"6 ","pages":"Article 100077"},"PeriodicalIF":0.0,"publicationDate":"2024-10-07","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142420554","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Using real world data to bridge the evidence gap left by prostate cancer screening trials 利用真实世界的数据弥补前列腺癌筛查试验留下的证据缺口

ESMO Real World Data and Digital Oncology

Pub Date : 2024-10-07 DOI: 10.1016/j.esmorw.2024.100073

N. Norori , N. Burns-Cox , L. Blaney , N. Mayor , A. Rylance , T.T. Shah , A. Naranjo , M.D. Hobbs

Background and purpose

Population-based prostate cancer screening is currently not recommended in the UK because harms may outweigh benefits. Recent changes to the diagnostic pathway have improved safety and accuracy, but uncertainty remains as to how much they have shifted the screening harm to benefit ratio. Our work uses modelling and real-world data (RWD) to bridge this evidence gap.

Materials and methods

We analysed outcomes of men entering the current prostate cancer diagnostic pathway using RWD from two NHS registries covering 16 hospitals. To assess improvements, we compared current UK clinical practice outcomes with those reported in the Cluster Randomised Trial of PSA Testing for Prostate Cancer (CAP)/ProtecT trial and past UK clinical practice, and to a model built to represent expected outcomes from the UK current diagnostic pathway.

Results

Out of 10 000 men who underwent a prostate specific antigen (PSA) test, we estimated that the proportion of men with no cancer after a biopsy following a PSA test decreased from 9.46% in the pre-magnetic resonance imaging CAP/ProtecT pathway to 2.33% and 1.52% in the Rapid Assessment for Prostate Imaging and Diagnosis (RAPID) and South West of England RWD pathways, respectively. Clinically insignificant prostate cancer diagnoses decreased to 0.73% in RAPID and 0.83% in South West, while the proportion of men experiencing sepsis reduced from a historic 0.10% to 0.02% in RAPID. We estimated an increase in clinically significant prostate cancer diagnoses in the RWD pathways.

Conclusions

This analysis of high-quality RWD estimates a 79% decrease in harm from the diagnostic process when comparing the previous pathway to actual outcomes from current UK clinical practice. This confirms and quantifies the harm reduction delivered by new diagnostic techniques.

背景和目的由于前列腺癌筛查弊大于利，英国目前并不推荐基于人群的前列腺癌筛查。最近对诊断路径的改变提高了安全性和准确性，但这些改变在多大程度上改变了筛查的危害与获益比仍存在不确定性。我们利用建模和真实世界数据（RWD）来弥补这一证据差距。材料和方法我们利用覆盖 16 家医院的两个英国国家医疗服务系统登记处的 RWD 分析了进入当前前列腺癌诊断途径的男性的结果。为了评估改进情况，我们将英国目前的临床实践结果与前列腺癌 PSA 检测集群随机试验 (CAP)/ProtecT 试验报告的结果和英国过去的临床实践结果进行了比较，并与根据英国目前的诊断途径建立的预期结果模型进行了比较。结果在1万名接受前列腺特异性抗原（PSA）检测的男性中，我们估计PSA检测后活检未发现癌症的男性比例从磁共振成像前CAP/ProtecT路径的9.46%降至前列腺成像和诊断快速评估（RAPID）路径的2.33%和英格兰西南部RWD路径的1.52%。临床症状不明显的前列腺癌诊断率在 RAPID 和西南地区分别降至 0.73% 和 0.83%，而出现败血症的男性比例则从历史上的 0.10% 降至 RAPID 的 0.02%。我们估计，在 RWD 诊断路径中，有临床意义的前列腺癌诊断有所增加。这证实并量化了新诊断技术所带来的危害减少效果。

{"title":"Using real world data to bridge the evidence gap left by prostate cancer screening trials","authors":"N. Norori , N. Burns-Cox , L. Blaney , N. Mayor , A. Rylance , T.T. Shah , A. Naranjo , M.D. Hobbs","doi":"10.1016/j.esmorw.2024.100073","DOIUrl":"10.1016/j.esmorw.2024.100073","url":null,"abstract":"<div><h3>Background and purpose</h3><div>Population-based prostate cancer screening is currently not recommended in the UK because harms may outweigh benefits. Recent changes to the diagnostic pathway have improved safety and accuracy, but uncertainty remains as to how much they have shifted the screening harm to benefit ratio. Our work uses modelling and real-world data (RWD) to bridge this evidence gap.</div></div><div><h3>Materials and methods</h3><div>We analysed outcomes of men entering the current prostate cancer diagnostic pathway using RWD from two NHS registries covering 16 hospitals. To assess improvements, we compared current UK clinical practice outcomes with those reported in the Cluster Randomised Trial of PSA Testing for Prostate Cancer (CAP)/ProtecT trial and past UK clinical practice, and to a model built to represent expected outcomes from the UK current diagnostic pathway.</div></div><div><h3>Results</h3><div>Out of 10 000 men who underwent a prostate specific antigen (PSA) test, we estimated that the proportion of men with no cancer after a biopsy following a PSA test decreased from 9.46% in the pre-magnetic resonance imaging CAP/ProtecT pathway to 2.33% and 1.52% in the Rapid Assessment for Prostate Imaging and Diagnosis (RAPID) and South West of England RWD pathways, respectively. Clinically insignificant prostate cancer diagnoses decreased to 0.73% in RAPID and 0.83% in South West, while the proportion of men experiencing sepsis reduced from a historic 0.10% to 0.02% in RAPID. We estimated an increase in clinically significant prostate cancer diagnoses in the RWD pathways.</div></div><div><h3>Conclusions</h3><div>This analysis of high-quality RWD estimates a 79% decrease in harm from the diagnostic process when comparing the previous pathway to actual outcomes from current UK clinical practice. This confirms and quantifies the harm reduction delivered by new diagnostic techniques.</div></div>","PeriodicalId":100491,"journal":{"name":"ESMO Real World Data and Digital Oncology","volume":"6 ","pages":"Article 100073"},"PeriodicalIF":0.0,"publicationDate":"2024-10-07","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142420606","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Oncology education in the age of artificial intelligence 人工智能时代的肿瘤学教育

ESMO Real World Data and Digital Oncology

Pub Date : 2024-10-07 DOI: 10.1016/j.esmorw.2024.100079

A. Prelaj , G. Scoazec , D. Ferber , J.N. Kather

The rapid advancements in artificial intelligence (AI) technology have broad implications for the clinical practice of oncology. AI methods enable us to analyze unstructured patient data in a quantitative way, at scale. AI can also help to manage the exponentially growing amount of specialist knowledge, for example, by parsing information from medical guidelines. As oncologists, we will increasingly interact with AI systems in clinical routine and research. These technical developments will require a reshaping of oncology education, which needs to include the development of ‘AI literacy’ as a core aim. Henceforth, oncologists require an understanding of the principles of AI, and the capabilities to adapt and use AI for clinical and research tasks. They should also be able to interpret AI-generated data effectively, especially when communicating with patients who will increasingly use AI tools. Oncologists who understand the fundamental concepts of AI, its limitations, and capabilities will be able to come up with new ideas and use cases, for example, in designing clinical trials. Currently, postgraduate oncology education lacks comprehensive curricula addressing AI integration. We propose that the scientific community, in particular academic institutions, professional societies, and policymakers, should prioritize implementing AI literacy within oncology curricula.

人工智能（AI）技术的飞速发展对肿瘤学的临床实践产生了广泛的影响。人工智能方法使我们能够以定量的方式大规模分析非结构化的患者数据。人工智能还能帮助管理呈指数级增长的专业知识，例如，通过解析医疗指南中的信息。作为肿瘤学家，我们将在临床日常工作和研究中越来越多地与人工智能系统互动。这些技术发展将要求重塑肿瘤学教育，其中需要将培养 "人工智能素养 "作为核心目标。因此，肿瘤学家需要了解人工智能的原理，并具备在临床和研究任务中适应和使用人工智能的能力。他们还应该能够有效解释人工智能生成的数据，尤其是在与越来越多地使用人工智能工具的患者交流时。了解人工智能基本概念、局限性和能力的肿瘤学家将能够提出新的想法和用例，例如在设计临床试验时。目前，肿瘤学研究生教育缺乏涉及人工智能整合的全面课程。我们建议科学界，特别是学术机构、专业协会和政策制定者，应优先在肿瘤学课程中落实人工智能扫盲。

{"title":"Oncology education in the age of artificial intelligence","authors":"A. Prelaj , G. Scoazec , D. Ferber , J.N. Kather","doi":"10.1016/j.esmorw.2024.100079","DOIUrl":"10.1016/j.esmorw.2024.100079","url":null,"abstract":"<div><div>The rapid advancements in artificial intelligence (AI) technology have broad implications for the clinical practice of oncology. AI methods enable us to analyze unstructured patient data in a quantitative way, at scale. AI can also help to manage the exponentially growing amount of specialist knowledge, for example, by parsing information from medical guidelines. As oncologists, we will increasingly interact with AI systems in clinical routine and research. These technical developments will require a reshaping of oncology education, which needs to include the development of ‘AI literacy’ as a core aim. Henceforth, oncologists require an understanding of the principles of AI, and the capabilities to adapt and use AI for clinical and research tasks. They should also be able to interpret AI-generated data effectively, especially when communicating with patients who will increasingly use AI tools. Oncologists who understand the fundamental concepts of AI, its limitations, and capabilities will be able to come up with new ideas and use cases, for example, in designing clinical trials. Currently, postgraduate oncology education lacks comprehensive curricula addressing AI integration. We propose that the scientific community, in particular academic institutions, professional societies, and policymakers, should prioritize implementing AI literacy within oncology curricula.</div></div>","PeriodicalId":100491,"journal":{"name":"ESMO Real World Data and Digital Oncology","volume":"6 ","pages":"Article 100079"},"PeriodicalIF":0.0,"publicationDate":"2024-10-07","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142420607","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Superhuman performance on urology board questions using an explainable language model enhanced with European Association of Urology guidelines 使用欧洲泌尿外科协会指南增强的可解释语言模型，在泌尿外科委员会问题上取得超人成绩

ESMO Real World Data and Digital Oncology

Pub Date : 2024-10-04 DOI: 10.1016/j.esmorw.2024.100078

M.J. Hetz , N. Carl , S. Haggenmüller , C. Wies , J.N. Kather , M.S. Michel , F. Wessels , T.J. Brinker

Background

Large language models encode clinical knowledge and can answer medical expert questions out-of-the-box without further training. However, this zero-shot performance is limited by outdated training data and lack of explainability impeding clinical translation. We aimed to develop a urology-specialized chatbot (UroBot) and evaluate it against state-of-the-art models as well as historical urologists’ performance in answering urological board questions in a fully clinician-verifiable manner.

Materials and methods

We developed UroBot, a software pipeline based on the GPT-3.5, GPT-4, and GPT-4o models by OpenAI, utilizing retrieval augmented generation and the 2023 European Association of Urology guidelines. UroBot was benchmarked against the zero-shot performance of GPT-3.5, GPT-4, GPT-4o, and Uro_Chat. The evaluation involved 10 runs with 200 European Board of Urology in-service assessment questions, with the performance measured by the mean rate of correct answers (RoCA).

Results

UroBot-4o achieved the highest RoCA, with an average of 88.4%, outperforming GPT-4o (77.6%) by 10.8%. Besides, it is clinician-verifiable and demonstrated the highest level of agreement between runs as measured by Fleiss’ kappa (κ = 0.979). In comparison, the average performance of urologists on urological board questions is 68.7% as reported by the literature.

Conclusions

UroBot is a clinician-verifiable and accurate software pipeline and outperforms published models and urologists in answering urology board questions. We provide code and instructions to use and extend UroBot for further development.

背景大语言模型对临床知识进行编码，无需进一步训练即可立即回答医学专家的问题。然而，由于训练数据过时且缺乏可解释性，这种 "零 "反馈的性能受到了限制，阻碍了临床翻译。我们的目标是开发一个泌尿科专业聊天机器人（UroBot），并以完全可由临床医生验证的方式，对照最先进的模型以及历史上泌尿科医生在回答泌尿科委员会问题时的表现对其进行评估。材料与方法我们开发了 UroBot，这是一个基于 OpenAI 的 GPT-3.5、GPT-4 和 GPT-4o 模型的软件管道，利用了检索增强生成和 2023 年欧洲泌尿外科协会指南。UroBot 以 GPT-3.5、GPT-4、GPT-4o 和 Uro_Chat 的零点性能为基准进行了评估。结果UroBot-4o的RoCA最高，平均为88.4%，比GPT-4o（77.6%）高出10.8%。此外，它还可由临床医生验证，并通过 Fleiss' kappa（κ = 0.979）测量显示出运行之间的最高一致性。相比之下，根据文献报道，泌尿科医生在回答泌尿科委员会问题时的平均成绩为 68.7%。结论UroBot 是一款可由临床医生验证的准确软件管道，在回答泌尿科委员会问题方面优于已发表的模型和泌尿科医生。我们提供了使用和扩展 UroBot 的代码和说明，以便进一步开发。

{"title":"Superhuman performance on urology board questions using an explainable language model enhanced with European Association of Urology guidelines","authors":"M.J. Hetz , N. Carl , S. Haggenmüller , C. Wies , J.N. Kather , M.S. Michel , F. Wessels , T.J. Brinker","doi":"10.1016/j.esmorw.2024.100078","DOIUrl":"10.1016/j.esmorw.2024.100078","url":null,"abstract":"<div><h3>Background</h3><div>Large language models encode clinical knowledge and can answer medical expert questions out-of-the-box without further training. However, this zero-shot performance is limited by outdated training data and lack of explainability impeding clinical translation. We aimed to develop a urology-specialized chatbot (UroBot) and evaluate it against state-of-the-art models as well as historical urologists’ performance in answering urological board questions in a fully clinician-verifiable manner.</div></div><div><h3>Materials and methods</h3><div>We developed UroBot, a software pipeline based on the GPT-3.5, GPT-4, and GPT-4o models by OpenAI, utilizing retrieval augmented generation and the 2023 European Association of Urology guidelines. UroBot was benchmarked against the zero-shot performance of GPT-3.5, GPT-4, GPT-4o, and Uro_Chat. The evaluation involved 10 runs with 200 European Board of Urology in-service assessment questions, with the performance measured by the mean rate of correct answers (RoCA).</div></div><div><h3>Results</h3><div>UroBot-4o achieved the highest RoCA, with an average of 88.4%, outperforming GPT-4o (77.6%) by 10.8%. Besides, it is clinician-verifiable and demonstrated the highest level of agreement between runs as measured by Fleiss’ kappa (κ = 0.979). In comparison, the average performance of urologists on urological board questions is 68.7% as reported by the literature.</div></div><div><h3>Conclusions</h3><div>UroBot is a clinician-verifiable and accurate software pipeline and outperforms published models and urologists in answering urology board questions. We provide code and instructions to use and extend UroBot for further development.</div></div>","PeriodicalId":100491,"journal":{"name":"ESMO Real World Data and Digital Oncology","volume":"6 ","pages":"Article 100078"},"PeriodicalIF":0.0,"publicationDate":"2024-10-04","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142420605","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Real-world occurrence, therapy, and outcome of patients with class 2 or 3 BRAF compared with class 1 BRAF-mutated cancers 与 1 类 BRAF 基因突变癌症相比，2 类或 3 类 BRAF 基因突变癌症患者的实际发病情况、治疗方法和疗效

ESMO Real World Data and Digital Oncology

Pub Date : 2024-09-25 DOI: 10.1016/j.esmorw.2024.100075

S. Pradervand , N. Freundler , B. Gosztonyi , L. Roncoroni , R. Achermann , T. Schwenk , G. de Fraipont , J. Garessus , S. Haefliger , A.B. Leichtle , M.K. Kiessling , T. Mueller-Focke , F.S. Krebs , V. Zoete , P. Tsantoulis , O. Michielin , C. Britschgi , A. Wicki

Background

BRAF V600 mutations are the epitome of targeted therapy. However, not much is known about non-V600 mutations. Using the new data infrastructure of the Swiss Personalized Oncology project of the Swiss Personalized Health Network (SPHN), we evaluated the fate of patients with cancer with non-V600 BRAF mutations in comparison to patients with class 1 mutations.

Patients and methods

In this retrospective observational multicenter study, we have assembled a cohort of 392 patients with class 1 and 154 patients with nonclass 1 BRAF mutations (76 colorectal cancers, 96 lung cancers, 297 melanomas, and 77 other cancers). We carried out outcome analyses between mutational classes and therapeutic subgroups.

Results

Overall survival (OS) did not differ significantly between patients with class 1 and nonclass 1 mutations. Upon treatment with BRAF/MEK inhibitors, patients with class 1 mutant melanoma showed numerically longer progression-free survival (PFS; 217 days) than patients with nonclass 1 mutant disease (73 days). Overall, in patients with class 2 or 3 mutations, BRAF and MEK inhibitors showed no benefit over other systemic therapies. However, specific class 2 mutations such as K601E may confer sensitivity to BRAF/MEK inhibitors, with two out of five patients achieving a PFS >400 days.

Conclusions

The diversity of BRAF mutations presents significant treatment challenges. Despite similar OS, nonclass 1 mutant tumors showed a trend toward lower PFS with BRAF/MEK blockade. Selected class 2 mutations may confer sensitivity to BRAF/MEK inhibitors. This highlights the rationale for a mutation, rather than class-specific, clinical approach against nonclass 1 BRAF-mutant tumors.

背景BRAF V600突变是靶向治疗的缩影。然而，人们对非V600突变却知之甚少。利用瑞士个性化健康网络（SPHN）的瑞士个性化肿瘤学项目的新数据基础设施，我们评估了非V600 BRAF突变癌症患者与1类突变患者的命运对比。患者和方法在这项回顾性多中心观察研究中，我们收集了392名1类患者和154名非1类BRAF突变患者（76名结直肠癌患者、96名肺癌患者、297名黑色素瘤患者和77名其他癌症患者）。我们在突变类别和治疗亚组之间进行了结果分析。结果1类和非1类突变患者的总生存期（OS）没有显著差异。在接受BRAF/MEK抑制剂治疗后，1类突变黑色素瘤患者的无进展生存期（PFS；217天）比非1类突变患者（73天）要长。总体而言，对于 2 类或 3 类突变患者，BRAF 和 MEK 抑制剂与其他系统疗法相比并无优势。然而，特定的2类突变（如K601E）可能会使患者对BRAF/MEK抑制剂敏感，5名患者中有2名的PFS达到400天。尽管OS相似，但非1类突变肿瘤在使用BRAF/MEK阻断剂后显示出较低的PFS趋势。选定的2类突变可能会赋予BRAF/MEK抑制剂敏感性。这凸显了针对非1类BRAF突变肿瘤采用突变而非特异性临床方法的合理性。

{"title":"Real-world occurrence, therapy, and outcome of patients with class 2 or 3 BRAF compared with class 1 BRAF-mutated cancers","authors":"S. Pradervand , N. Freundler , B. Gosztonyi , L. Roncoroni , R. Achermann , T. Schwenk , G. de Fraipont , J. Garessus , S. Haefliger , A.B. Leichtle , M.K. Kiessling , T. Mueller-Focke , F.S. Krebs , V. Zoete , P. Tsantoulis , O. Michielin , C. Britschgi , A. Wicki","doi":"10.1016/j.esmorw.2024.100075","DOIUrl":"10.1016/j.esmorw.2024.100075","url":null,"abstract":"<div><h3>Background</h3><div><em>BRAF</em> V600 mutations are the epitome of targeted therapy. However, not much is known about non-V600 mutations. Using the new data infrastructure of the Swiss Personalized Oncology project of the Swiss Personalized Health Network (SPHN), we evaluated the fate of patients with cancer with non-V600 <em>BRAF</em> mutations in comparison to patients with class 1 mutations.</div></div><div><h3>Patients and methods</h3><div>In this retrospective observational multicenter study, we have assembled a cohort of 392 patients with class 1 and 154 patients with nonclass 1 <em>BRAF</em> mutations (76 colorectal cancers, 96 lung cancers, 297 melanomas, and 77 other cancers). We carried out outcome analyses between mutational classes and therapeutic subgroups.</div></div><div><h3>Results</h3><div>Overall survival (OS) did not differ significantly between patients with class 1 and nonclass 1 mutations. Upon treatment with BRAF/MEK inhibitors, patients with class 1 mutant melanoma showed numerically longer progression-free survival (PFS; 217 days) than patients with nonclass 1 mutant disease (73 days). Overall, in patients with class 2 or 3 mutations, BRAF and MEK inhibitors showed no benefit over other systemic therapies. However, specific class 2 mutations such as K601E may confer sensitivity to BRAF/MEK inhibitors, with two out of five patients achieving a PFS >400 days.</div></div><div><h3>Conclusions</h3><div>The diversity of <em>BRAF</em> mutations presents significant treatment challenges. Despite similar OS, nonclass 1 mutant tumors showed a trend toward lower PFS with BRAF/MEK blockade. Selected class 2 mutations may confer sensitivity to BRAF/MEK inhibitors. This highlights the rationale for a mutation, rather than class-specific, clinical approach against nonclass 1 BRAF-mutant tumors.</div></div>","PeriodicalId":100491,"journal":{"name":"ESMO Real World Data and Digital Oncology","volume":"6 ","pages":"Article 100075"},"PeriodicalIF":0.0,"publicationDate":"2024-09-25","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142317828","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Clinical trial and real-world evidence of circulating tumor DNA monitoring to predict recurrence in patients with resected colorectal cancer 监测循环肿瘤 DNA 以预测切除大肠癌患者复发的临床试验和实际证据

ESMO Real World Data and Digital Oncology

Pub Date : 2024-09-24 DOI: 10.1016/j.esmorw.2024.100076

H.T. Nguyen , V.-A. Nguyen Hoang , T.V. Nguyen , N.A.L. Trinh , T.H. Pham , D.N. Le , H.H. Ho , T.D. Nguyen , H.D. Nguyen , T.S.L. Thi , N. Nguyen , D.H. Tran , M.T. Le , T.C. Dinh , T.S. Nguyen , K.C.N. The , H. Mai , M.T. Chu , D.H. Pham , N.H.T. Phuc , L.N. Tu

Background

Circulating tumor DNA (ctDNA) is a novel biomarker to predict recurrence in colorectal cancer (CRC). However, clinical validation data in underrepresented patient populations like Southeast Asians are lacking.

Materials and methods

In our prospective clinical trial, 92 patients with stage I-IVA CRC and eligible for curative-intent surgery were recruited. Blood samples were collected before surgery, after surgery and during surveillance. ctDNA analysis was carried out using a personalized tumor-informed approach. Performance of ctDNA monitoring was further evaluated in a retrospective analysis of 32 patients who had commercial ctDNA testing.

Results

Before surgery, the ctDNA detection rate was 91.2%, significantly higher than the carcinoembryonic antigen (CEA) elevation rate at 42.9%. At 2-4 weeks after surgery, ctDNA detection, not CEA, was significantly associated with shorter disease-free survival (DFS) [hazard ratio (HR) = 30.3, 95% confidence interval (CI) 3.3-278.9, P = 0.002]. Longitudinally, ctDNA positivity showed the strongest prognostic value with 100% sensitivity and specificity to detect recurrence (HR = 173.6, 95% CI 19.7 to >1000, P = 0.001); the median lead time was 8.0 months. In the real-world data analysis, post-operative ctDNA remained the only significant prognostic biomarker for DFS (HR = 44.2, 95% CI 9.2-212.0, P < 0.0001). Among relapsed patients, the most common mutations found in ctDNA were in TP53 (58.8%), APC (52.9%) and KRAS (41.2%), with high frequency of co-mutations.

Conclusion

Our analysis from both clinical trial and real-world data indicated that ctDNA was an independent and strong prognostic biomarker to predict recurrence in CRC. ctDNA testing could be helpful for clinical decision making by enabling personalized intervention and surveillance strategies.

背景循环肿瘤 DNA（ctDNA）是预测结直肠癌（CRC）复发的新型生物标记物。材料与方法在我们的前瞻性临床试验中，共招募了 92 例 I-IVA 期 CRC 患者，他们都符合治愈性手术的条件。在手术前、手术后和监测期间采集血样。ctDNA分析采用个性化肿瘤信息方法。结果手术前，ctDNA检测率为91.2%，明显高于癌胚抗原（CEA）升高率42.9%。术后2-4周，ctDNA检测（而非癌胚抗原）与无病生存期（DFS）明显相关[危险比（HR）=30.3，95%置信区间（CI）3.3-278.9，P=0.002]。纵向来看，ctDNA阳性显示出最强的预后价值，其检测复发的敏感性和特异性均为100%（HR = 173.6，95% CI 19.7 to >1000，P = 0.001）；中位前导时间为8.0个月。在真实世界数据分析中，术后ctDNA仍是DFS唯一显著的预后生物标志物（HR = 44.2，95% CI 9.2-212.0，P <0.0001）。在复发患者中，ctDNA中最常见的突变是TP53（58.8%）、APC（52.9%）和KRAS（41.2%），而且共突变的频率很高。

{"title":"Clinical trial and real-world evidence of circulating tumor DNA monitoring to predict recurrence in patients with resected colorectal cancer","authors":"H.T. Nguyen , V.-A. Nguyen Hoang , T.V. Nguyen , N.A.L. Trinh , T.H. Pham , D.N. Le , H.H. Ho , T.D. Nguyen , H.D. Nguyen , T.S.L. Thi , N. Nguyen , D.H. Tran , M.T. Le , T.C. Dinh , T.S. Nguyen , K.C.N. The , H. Mai , M.T. Chu , D.H. Pham , N.H.T. Phuc , L.N. Tu","doi":"10.1016/j.esmorw.2024.100076","DOIUrl":"10.1016/j.esmorw.2024.100076","url":null,"abstract":"<div><h3>Background</h3><div>Circulating tumor DNA (ctDNA) is a novel biomarker to predict recurrence in colorectal cancer (CRC). However, clinical validation data in underrepresented patient populations like Southeast Asians are lacking.</div></div><div><h3>Materials and methods</h3><div>In our prospective clinical trial, 92 patients with stage I-IVA CRC and eligible for curative-intent surgery were recruited. Blood samples were collected before surgery, after surgery and during surveillance. ctDNA analysis was carried out using a personalized tumor-informed approach. Performance of ctDNA monitoring was further evaluated in a retrospective analysis of 32 patients who had commercial ctDNA testing.</div></div><div><h3>Results</h3><div>Before surgery, the ctDNA detection rate was 91.2%, significantly higher than the carcinoembryonic antigen (CEA) elevation rate at 42.9%. At 2-4 weeks after surgery, ctDNA detection, not CEA, was significantly associated with shorter disease-free survival (DFS) [hazard ratio (HR) = 30.3, 95% confidence interval (CI) 3.3-278.9, <em>P</em> = 0.002]. Longitudinally, ctDNA positivity showed the strongest prognostic value with 100% sensitivity and specificity to detect recurrence (HR = 173.6, 95% CI 19.7 to >1000, <em>P</em> = 0.001); the median lead time was 8.0 months. In the real-world data analysis, post-operative ctDNA remained the only significant prognostic biomarker for DFS (HR = 44.2, 95% CI 9.2-212.0, <em>P</em> < 0.0001). Among relapsed patients, the most common mutations found in ctDNA were in <em>TP53</em> (58.8%), <em>APC</em> (52.9%) and <em>KRAS</em> (41.2%), with high frequency of co-mutations.</div></div><div><h3>Conclusion</h3><div>Our analysis from both clinical trial and real-world data indicated that ctDNA was an independent and strong prognostic biomarker to predict recurrence in CRC. ctDNA testing could be helpful for clinical decision making by enabling personalized intervention and surveillance strategies.</div></div>","PeriodicalId":100491,"journal":{"name":"ESMO Real World Data and Digital Oncology","volume":"6 ","pages":"Article 100076"},"PeriodicalIF":0.0,"publicationDate":"2024-09-24","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.sciencedirect.com/science/article/pii/S2949820124000547/pdfft?md5=669e201a99dd4cf1572edc157b3e2fb7&pid=1-s2.0-S2949820124000547-main.pdf","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142316063","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Sequencing the anti-EGFR-TKIs; route to the longest survival 抗EGFR-TKIs的排序；获得最长生存期的途径

ESMO Real World Data and Digital Oncology

Pub Date : 2024-09-18 DOI: 10.1016/j.esmorw.2024.100074

H.G. Güzel , Y. İlhan , A.H. Önder

引用次数: 0

EBANO study: real-world data from patients with locally advanced/metastatic urothelial carcinoma (la/mUC) in Northern Spain EBANO 研究：西班牙北部局部晚期/转移性尿路上皮癌 (la/mUC) 患者的真实数据

ESMO Real World Data and Digital Oncology

Pub Date : 2024-09-18 DOI: 10.1016/j.esmorw.2024.100063

R. Fernández Rodríguez , N. Sagastibeltza , E. Pujol Obis , N. Lainez Milagro , R. Sánchez-Escribano , M. Martínez Kareaga , J.A. Verdún Aguilar , M. Arruti Ibarbia , M. Pumares González , T. de Portugal Fernández del Rivero , A. Lacalle Emborujo , I. Gil Arnaiz , A. Pereira-Elorrieta , C. Álvarez Fernández , I. Duran , GONORTE Collaborative Group

Background

Limited real-world data exist on managing locally advanced/metastatic urothelial carcinoma (la/mUC), crucial for predicting outcomes and integrating new treatments. This study explores la/mUC in Northern Spain.

Methods

Retrospective observational analysis from 16 hospitals across 12 provinces. The study population included adult patients diagnosed with la/mUC (1 January 2007-31 December 2019). Median overall survival and progression-free survival were determined using the Kaplan–Meier method.

Results

A total of 1230 patients were included, 83% were males, median age was 68, and 70% had a smoking history. The most common primary tumor was bladder (89.2%) followed by renal pelvis (6.4%), ureter (4.3%), and urethra (0.1%) and most of them (91%) were pure urothelial tumors. Cystectomy was the predominant treatment of localized disease (69%) and 30% received perioperative chemotherapy. Twenty-four percent of la/mUC patients never received systemic therapy and out of 934 patients treated with first-line (1L) therapy, 55% were fit for cisplatin. Progression to further lines of treatment was poor, only 53% (n = 492) progressed to receive a second line and 22% (n = 209) to a third line. Median overall survival (95% confidence interval) was 12.2 (11.3-12.9) months for the entire cohort and 14.5 and 10.8 months for patients who received first-line cisplatin or carboplatin-based chemotherapy, respectively. No unexpected toxicity was reported.

Conclusions

This Spanish real-world data analysis echoes previous findings and highlights unmet needs, including the proportion of patients not receiving systemic treatment and the limited progression to subsequent lines of therapy, hampering access to new effective treatments.

背景关于管理局部晚期/转移性尿路上皮癌（la/mUC）的真实世界数据有限，而这些数据对于预测预后和整合新疗法至关重要。本研究探讨了西班牙北部的局部晚期/转移性尿路上皮癌（la/mUC）情况。研究对象包括确诊为la/mUC的成年患者（2007年1月1日至2019年12月31日）。结果共纳入1230名患者，83%为男性，中位年龄为68岁，70%有吸烟史。最常见的原发肿瘤是膀胱（89.2%），其次是肾盂（6.4%）、输尿管（4.3%）和尿道（0.1%），其中大多数（91%）为纯泌尿道上皮肿瘤。膀胱切除术是治疗局部疾病的主要方法（69%），30%的患者接受了围手术期化疗。24%的la/mUC患者从未接受过系统治疗，在接受一线（1L）治疗的934名患者中，55%适合顺铂治疗。进展到进一步治疗的情况很差，只有53%（n = 492）的患者进展到接受二线治疗，22%（n = 209）的患者进展到接受三线治疗。整个队列的中位总生存期（95%置信区间）为12.2（11.3-12.9）个月，接受顺铂或卡铂一线化疗的患者的中位总生存期分别为14.5个月和10.8个月。结论这项西班牙真实世界数据分析与之前的研究结果一致，并强调了尚未满足的需求，包括未接受系统治疗的患者比例以及后续治疗线进展有限，这阻碍了患者获得新的有效治疗。

{"title":"EBANO study: real-world data from patients with locally advanced/metastatic urothelial carcinoma (la/mUC) in Northern Spain","authors":"R. Fernández Rodríguez , N. Sagastibeltza , E. Pujol Obis , N. Lainez Milagro , R. Sánchez-Escribano , M. Martínez Kareaga , J.A. Verdún Aguilar , M. Arruti Ibarbia , M. Pumares González , T. de Portugal Fernández del Rivero , A. Lacalle Emborujo , I. Gil Arnaiz , A. Pereira-Elorrieta , C. Álvarez Fernández , I. Duran , GONORTE Collaborative Group","doi":"10.1016/j.esmorw.2024.100063","DOIUrl":"10.1016/j.esmorw.2024.100063","url":null,"abstract":"<div><h3>Background</h3><p>Limited real-world data exist on managing locally advanced/metastatic urothelial carcinoma (la/mUC), crucial for predicting outcomes and integrating new treatments. This study explores la/mUC in Northern Spain.</p></div><div><h3>Methods</h3><p>Retrospective observational analysis from 16 hospitals across 12 provinces. The study population included adult patients diagnosed with la/mUC (1 January 2007-31 December 2019). Median overall survival and progression-free survival were determined using the Kaplan–Meier method.</p></div><div><h3>Results</h3><p>A total of 1230 patients were included, 83% were males, median age was 68, and 70% had a smoking history. The most common primary tumor was bladder (89.2%) followed by renal pelvis (6.4%), ureter (4.3%), and urethra (0.1%) and most of them (91%) were pure urothelial tumors. Cystectomy was the predominant treatment of localized disease (69%) and 30% received perioperative chemotherapy. Twenty-four percent of la/mUC patients never received systemic therapy and out of 934 patients treated with first-line (1L) therapy, 55% were fit for cisplatin. Progression to further lines of treatment was poor, only 53% (<em>n</em> = 492) progressed to receive a second line and 22% (<em>n</em> = 209) to a third line. Median overall survival (95% confidence interval) was 12.2 (11.3-12.9) months for the entire cohort and 14.5 and 10.8 months for patients who received first-line cisplatin or carboplatin-based chemotherapy, respectively. No unexpected toxicity was reported.</p></div><div><h3>Conclusions</h3><p>This Spanish real-world data analysis echoes previous findings and highlights unmet needs, including the proportion of patients not receiving systemic treatment and the limited progression to subsequent lines of therapy, hampering access to new effective treatments.</p></div>","PeriodicalId":100491,"journal":{"name":"ESMO Real World Data and Digital Oncology","volume":"6 ","pages":"Article 100063"},"PeriodicalIF":0.0,"publicationDate":"2024-09-18","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.sciencedirect.com/science/article/pii/S2949820124000419/pdfft?md5=6518ef0344086f222997c2c513b5b982&pid=1-s2.0-S2949820124000419-main.pdf","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142242323","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Comparative effectiveness among BRAF plus MEK inhibitors for patients with BRAF V600-mutant melanoma☆ BRAF和MEK抑制剂对BRAF V600突变黑色素瘤患者的疗效比较☆。

ESMO Real World Data and Digital Oncology

Pub Date : 2024-09-18 DOI: 10.1016/j.esmorw.2024.100071

G.K. In , K. Chen , G. Sajeev , R. Simpson , S. Kalia , D. Christensen , D. Liu , N. Rezai , A. di Pietro , J. Signorovitch

Background

Understanding of comparative efficacy of treatments can inform clinical decision-making. This study compared overall survival (OS) and progression-free survival (PFS) across patients with metastatic BRAFV600-mutant melanoma initiating encorafenib + binimetinib (ENCO + BINI), dabrafenib + trametinib (DAB + TRAM), and vemurafenib + cobimetinib (VEM + COBI).

Materials and methods

In this hybrid study, we contextualized OS and PFS between patients with metastatic BRAF V600E/K-mutant melanoma receiving ENCO + BINI in the phase III COLUMBUS trial (enrollment: December 2013 to April 2015) versus real-world data (RWD) from a nationwide electronic health record-derived deidentified database (treatment initiation: 2014-2021). After observing consistent outcomes across trial and RWD, we compared OS and PFS for a pooled ENCO + BINI cohort across these settings versus comparable DAB + TRAM and VEM + COBI cohorts from the real-world database.

Results

Of 716 patients [ENCO + BINI (n = 275; n = 192 from COLUMBUS, n = 83 from RWD), DAB + TRAM (n = 387), VEM + COBI (n = 54)], mean age was 56-60 years. OS and PFS were similar for ENCO + BINI-treated patients in COLUMBUS and RWD [adjusted hazard ratios: 1.03 (95% CI 0.62-1.72) for OS, 1.10 (0.69-1.75) for PFS]. Relative to the pooled ENCO + BINI group, adjusted OS and PFS were significantly worse for DAB + TRAM [OS: 1.32 (1.05-1.65), PFS: 1.49 (1.20-1.87)] and comparable for VEM + COBI [OS: 1.17 (0.76-1.79), PFS: 1.20 (0.79-1.82)]. Results were similar in comparisons based on the RWD groups alone, when trial data were excluded.

Conclusions

OS and PFS were longer with ENCO + BINI relative to DAB + TRAM and comparable to VEM + COBI, after adjusting for differences in patient profiles. These findings add to evidence informing the use of combination v-Raf murine sarcoma viral oncogene homolog B protein (BRAF)/mitogen-activated protein kinase kinase (MEK) inhibitors in metastatic BRAF V600-mutant melanoma.

背景了解治疗方法的比较疗效可为临床决策提供依据。本研究比较了转移性BRAFV600突变黑色素瘤患者开始接受安戈非尼+比尼替尼（ENCO+BINI）、达拉非尼+曲美替尼（DAB+TRAM）和维莫非尼+氯贝替尼（VEM+COBI）治疗后的总生存期（OS）和无进展生存期（PFS）。材料与方法在这项混合研究中，我们将在 COLUMBUS III 期试验（入组时间：2013 年 12 月至 2015 年 4 月）中接受 ENCO + BINI 治疗的转移性 BRAF V600E/K 突变黑色素瘤患者的 OS 和 PFS 与来自全国范围内电子健康记录衍生的去识别数据库的真实世界数据（RWD）（治疗开始时间：2014-2021 年）进行了对比。在观察到试验和 RWD 的一致结果后，我们比较了在这些情况下汇总的 ENCO + BINI 队列与真实世界数据库中可比的 DAB + TRAM 和 VEM + COBI 队列的 OS 和 PFS。结果在 716 名患者中[ENCO + BINI（n = 275；n = 192 来自 COLUMBUS，n = 83 来自 RWD），DAB + TRAM（n = 387），VEM + COBI（n = 54）]，平均年龄为 56-60 岁。COLUMBUS 和 RWD 的 ENCO + BINI 治疗患者的 OS 和 PFS 相似[调整后危险比为 1.03（95% CI）]：OS 为 1.03（95% CI 0.62-1.72），PFS 为 1.10（0.69-1.75）]。相对于ENCO + BINI组，DAB + TRAM的调整后OS和PFS明显更差[OS：1.32（1.05-1.65），PFS：1.49（1.20-1.87）]，而VEM + COBI的调整后OS和PFS相当[OS：1.17（0.76-1.79），PFS：1.20（0.79-1.82）]。在排除试验数据后，仅基于RWD组的比较结果与之相似。结论在调整患者资料的差异后，ENCO + BINI的OS和PFS长于DAB + TRAM，与VEM + COBI相当。这些发现为在转移性BRAF V600突变黑色素瘤中联合使用v-Raf小鼠肉瘤病毒癌基因同源物B蛋白（BRAF）/介原激活蛋白激酶激酶（MEK）抑制剂提供了更多证据。

{"title":"Comparative effectiveness among BRAF plus MEK inhibitors for patients with BRAF V600-mutant melanoma☆","authors":"G.K. In , K. Chen , G. Sajeev , R. Simpson , S. Kalia , D. Christensen , D. Liu , N. Rezai , A. di Pietro , J. Signorovitch","doi":"10.1016/j.esmorw.2024.100071","DOIUrl":"10.1016/j.esmorw.2024.100071","url":null,"abstract":"<div><h3>Background</h3><p>Understanding of comparative efficacy of treatments can inform clinical decision-making. This study compared overall survival (OS) and progression-free survival (PFS) across patients with metastatic BRAFV600-mutant melanoma initiating encorafenib + binimetinib (ENCO + BINI), dabrafenib + trametinib (DAB + TRAM), and vemurafenib + cobimetinib (VEM + COBI).</p></div><div><h3>Materials and methods</h3><p>In this hybrid study, we contextualized OS and PFS between patients with metastatic BRAF V600E/K-mutant melanoma receiving ENCO + BINI in the phase III COLUMBUS trial (enrollment: December 2013 to April 2015) versus real-world data (RWD) from a nationwide electronic health record-derived deidentified database (treatment initiation: 2014-2021). After observing consistent outcomes across trial and RWD, we compared OS and PFS for a pooled ENCO + BINI cohort across these settings versus comparable DAB + TRAM and VEM + COBI cohorts from the real-world database.</p></div><div><h3>Results</h3><p>Of 716 patients [ENCO + BINI (<em>n</em> = 275; <em>n</em> = 192 from COLUMBUS, <em>n</em> = 83 from RWD), DAB + TRAM (<em>n</em> = 387), VEM + COBI (<em>n</em> = 54)], mean age was 56-60 years. OS and PFS were similar for ENCO + BINI-treated patients in COLUMBUS and RWD [adjusted hazard ratios: 1.03 (95% CI 0.62-1.72) for OS, 1.10 (0.69-1.75) for PFS]. Relative to the pooled ENCO + BINI group, adjusted OS and PFS were significantly worse for DAB + TRAM [OS: 1.32 (1.05-1.65), PFS: 1.49 (1.20-1.87)] and comparable for VEM + COBI [OS: 1.17 (0.76-1.79), PFS: 1.20 (0.79-1.82)]. Results were similar in comparisons based on the RWD groups alone, when trial data were excluded.</p></div><div><h3>Conclusions</h3><p>OS and PFS were longer with ENCO + BINI relative to DAB + TRAM and comparable to VEM + COBI, after adjusting for differences in patient profiles. These findings add to evidence informing the use of combination v-Raf murine sarcoma viral oncogene homolog B protein (BRAF)/mitogen-activated protein kinase kinase (MEK) inhibitors in metastatic BRAF V600-mutant melanoma.</p></div>","PeriodicalId":100491,"journal":{"name":"ESMO Real World Data and Digital Oncology","volume":"6 ","pages":"Article 100071"},"PeriodicalIF":0.0,"publicationDate":"2024-09-18","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.sciencedirect.com/science/article/pii/S2949820124000493/pdfft?md5=515da39c8107032de596189f1dab29b3&pid=1-s2.0-S2949820124000493-main.pdf","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142242276","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0