Duchenne型肌营养不良（DMD）患者从仰卧到丧失站起能力时间的175PA预后评分

IF 2.8 4区医学 Q2 CLINICAL NEUROLOGY

Neuromuscular Disorders Pub Date : 2025-09-01 DOI:10.1016/j.nmd.2025.105528

F. Muntoni , J. Signorovitch , J. Marden , N. Done , S. Wang , H. Akbarnejad , H. Kang , T. Li , S. Ward , A. Manzur , N. Goemans , V. Straub , E. Mercuri , L. Servais

{"title":"Duchenne型肌营养不良（DMD）患者从仰卧到丧失站起能力时间的175PA预后评分","authors":"F. Muntoni , J. Signorovitch , J. Marden , N. Done , S. Wang , H. Akbarnejad , H. Kang , T. Li , S. Ward , A. Manzur , N. Goemans , V. Straub , E. Mercuri , L. Servais","doi":"10.1016/j.nmd.2025.105528","DOIUrl":null,"url":null,"abstract":"<div><div>Loss of ability to rise from supine (LoR) is an early milestone in DMD that reduces independence in daily living and marks the progression towards loss of ambulation. Previous research has identified single thresholds of rise from supine (RFS) time as prognostic for LoR (e.g., >5s). We used machine learning to develop a prognostic score for time to LoR based on multiple data-driven thresholds and patient characteristics that are easily assessed in clinical practice and routinely included in clinical trials. Longitudinal data from 595 boys with DMD were drawn from UZ Leuven, PRO-DMD-01, North Star UK, iMDEX, and placebo arms from trials of tadalafil, ataluren, and drisapersen. LoR was defined as the first assessment with timed RFS ≥30s, graded RFS of 1, or NSAA rise score of 0. A Weibull survival classification and regression tree model was used to identify risk groups. Candidate predictors included age, height, weight, BMI, steroid regimen, type, and duration, RFS, 10-meter walk/run (10MWR), and NSAA items. Patients had a mean ±standard deviation baseline age of 8.3±2.1 years, with RFS time of 6.1±4.2s, and NSAA total score of 24.9±5.8. The fitted model identified 5 risk groups showing excellent risk stratification, with median times to LoR of 0.3, 0.7, 1.5, 2.7, and 4.5 years (log-rank p<0.001). Risk groups differed by baseline RFS with thresholds at <5s, <10s, and <21s, with further stratification by NSAA sit-to-stand score (0/1 vs 2). The model explained 42.2% of variation in time to LoR, comparable to more complex survival models with both continuous measures and age and 10MWR included as additional predictors (pseudo-R² 44.8%). This prognostic score, based on readily assessable functional measures, provides clinically meaningful differentiation of times to LoR. Upon validation in independent data, this model could inform the design of clinical trials seeking to enrich for levels of LoR risk and benchmark LoR outcomes for patients receiving novel therapies.</div></div>","PeriodicalId":19135,"journal":{"name":"Neuromuscular Disorders","volume":"53 ","pages":"Article 105528"},"PeriodicalIF":2.8000,"publicationDate":"2025-09-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"0","resultStr":"{\"title\":\"175PA prognostic score for time to loss of ability to rise from supine in Duchenne muscular dystrophy (DMD)\",\"authors\":\"F. Muntoni , J. Signorovitch , J. Marden , N. Done , S. Wang , H. Akbarnejad , H. Kang , T. Li , S. Ward , A. Manzur , N. Goemans , V. Straub , E. Mercuri , L. Servais\",\"doi\":\"10.1016/j.nmd.2025.105528\",\"DOIUrl\":null,\"url\":null,\"abstract\":\"<div><div>Loss of ability to rise from supine (LoR) is an early milestone in DMD that reduces independence in daily living and marks the progression towards loss of ambulation. Previous research has identified single thresholds of rise from supine (RFS) time as prognostic for LoR (e.g., >5s). We used machine learning to develop a prognostic score for time to LoR based on multiple data-driven thresholds and patient characteristics that are easily assessed in clinical practice and routinely included in clinical trials. Longitudinal data from 595 boys with DMD were drawn from UZ Leuven, PRO-DMD-01, North Star UK, iMDEX, and placebo arms from trials of tadalafil, ataluren, and drisapersen. LoR was defined as the first assessment with timed RFS ≥30s, graded RFS of 1, or NSAA rise score of 0. A Weibull survival classification and regression tree model was used to identify risk groups. Candidate predictors included age, height, weight, BMI, steroid regimen, type, and duration, RFS, 10-meter walk/run (10MWR), and NSAA items. Patients had a mean ±standard deviation baseline age of 8.3±2.1 years, with RFS time of 6.1±4.2s, and NSAA total score of 24.9±5.8. The fitted model identified 5 risk groups showing excellent risk stratification, with median times to LoR of 0.3, 0.7, 1.5, 2.7, and 4.5 years (log-rank p<0.001). Risk groups differed by baseline RFS with thresholds at <5s, <10s, and <21s, with further stratification by NSAA sit-to-stand score (0/1 vs 2). The model explained 42.2% of variation in time to LoR, comparable to more complex survival models with both continuous measures and age and 10MWR included as additional predictors (pseudo-R² 44.8%). This prognostic score, based on readily assessable functional measures, provides clinically meaningful differentiation of times to LoR. Upon validation in independent data, this model could inform the design of clinical trials seeking to enrich for levels of LoR risk and benchmark LoR outcomes for patients receiving novel therapies.</div></div>\",\"PeriodicalId\":19135,\"journal\":{\"name\":\"Neuromuscular Disorders\",\"volume\":\"53 \",\"pages\":\"Article 105528\"},\"PeriodicalIF\":2.8000,\"publicationDate\":\"2025-09-01\",\"publicationTypes\":\"Journal Article\",\"fieldsOfStudy\":null,\"isOpenAccess\":false,\"openAccessPdf\":\"\",\"citationCount\":\"0\",\"resultStr\":null,\"platform\":\"Semanticscholar\",\"paperid\":null,\"PeriodicalName\":\"Neuromuscular Disorders\",\"FirstCategoryId\":\"3\",\"ListUrlMain\":\"https://www.sciencedirect.com/science/article/pii/S096089662500255X\",\"RegionNum\":4,\"RegionCategory\":\"医学\",\"ArticlePicture\":[],\"TitleCN\":null,\"AbstractTextCN\":null,\"PMCID\":null,\"EPubDate\":\"\",\"PubModel\":\"\",\"JCR\":\"Q2\",\"JCRName\":\"CLINICAL NEUROLOGY\",\"Score\":null,\"Total\":0}","platform":"Semanticscholar","paperid":null,"PeriodicalName":"Neuromuscular Disorders","FirstCategoryId":"3","ListUrlMain":"https://www.sciencedirect.com/science/article/pii/S096089662500255X","RegionNum":4,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"Q2","JCRName":"CLINICAL NEUROLOGY","Score":null,"Total":0}

引用次数: 0

摘要

从仰卧位起身能力的丧失是DMD的早期里程碑，它降低了日常生活的独立性，标志着逐渐丧失行走能力。先前的研究已经确定了仰卧起跳（RFS）时间的单一阈值作为LoR的预后（例如，>5）。基于多个数据驱动的阈值和患者特征，我们使用机器学习开发了一个预后评分，该评分可以在临床实践中轻松评估，并常规纳入临床试验。595名患有DMD的男孩的纵向数据来自UZ Leuven， PRO-DMD-01, North Star UK， iMDEX和安慰剂组，来自他达拉非，阿塔鲁伦和德利哌森的试验。LoR定义为第一次评估时RFS≥30s，分级RFS为1，或NSAA上升评分为0。使用威布尔生存分类和回归树模型来确定危险组。候选预测因子包括年龄、身高、体重、BMI、类固醇治疗方案、类型和持续时间、RFS、10米步行/跑步（10MWR）和NSAA项目。患者平均±标准差基线年龄8.3±2.1岁，RFS时间6.1±4.2s， NSAA总分24.9±5.8分。拟合模型确定了5个风险组，显示出极好的风险分层，到LoR的中位数时间为0.3、0.7、1.5、2.7和4.5年（log-rank p<0.001）。风险组的基线RFS差异，阈值分别为5分、10分和21分，并根据NSAA坐站评分进一步分层（0/1 vs 2）。该模型解释了42.2%的时间变化，与更复杂的生存模型相比，包括连续测量、年龄和10MWR作为额外的预测因子（伪r²44.8%）。这种预后评分基于易于评估的功能指标，提供了临床有意义的LoR时间区分。在独立数据验证后，该模型可以为临床试验的设计提供信息，以丰富接受新疗法的患者的LoR风险水平和基准LoR结果。

本文章由计算机程序翻译，如有差异，请以英文原文为准。

查看原文本刊更多论文

175PA prognostic score for time to loss of ability to rise from supine in Duchenne muscular dystrophy (DMD)

Loss of ability to rise from supine (LoR) is an early milestone in DMD that reduces independence in daily living and marks the progression towards loss of ambulation. Previous research has identified single thresholds of rise from supine (RFS) time as prognostic for LoR (e.g., >5s). We used machine learning to develop a prognostic score for time to LoR based on multiple data-driven thresholds and patient characteristics that are easily assessed in clinical practice and routinely included in clinical trials. Longitudinal data from 595 boys with DMD were drawn from UZ Leuven, PRO-DMD-01, North Star UK, iMDEX, and placebo arms from trials of tadalafil, ataluren, and drisapersen. LoR was defined as the first assessment with timed RFS ≥30s, graded RFS of 1, or NSAA rise score of 0. A Weibull survival classification and regression tree model was used to identify risk groups. Candidate predictors included age, height, weight, BMI, steroid regimen, type, and duration, RFS, 10-meter walk/run (10MWR), and NSAA items. Patients had a mean ±standard deviation baseline age of 8.3±2.1 years, with RFS time of 6.1±4.2s, and NSAA total score of 24.9±5.8. The fitted model identified 5 risk groups showing excellent risk stratification, with median times to LoR of 0.3, 0.7, 1.5, 2.7, and 4.5 years (log-rank p<0.001). Risk groups differed by baseline RFS with thresholds at <5s, <10s, and <21s, with further stratification by NSAA sit-to-stand score (0/1 vs 2). The model explained 42.2% of variation in time to LoR, comparable to more complex survival models with both continuous measures and age and 10MWR included as additional predictors (pseudo-R² 44.8%). This prognostic score, based on readily assessable functional measures, provides clinically meaningful differentiation of times to LoR. Upon validation in independent data, this model could inform the design of clinical trials seeking to enrich for levels of LoR risk and benchmark LoR outcomes for patients receiving novel therapies.

求助全文

通过发布文献求助，成功后即可免费获取论文全文。去求助

来源期刊

Neuromuscular Disorders 医学-临床神经学

CiteScore

4.60

自引率

3.60%

发文量

543

审稿时长

53 days

期刊介绍： This international, multidisciplinary journal covers all aspects of neuromuscular disorders in childhood and adult life (including the muscular dystrophies, spinal muscular atrophies, hereditary neuropathies, congenital myopathies, myasthenias, myotonic syndromes, metabolic myopathies and inflammatory myopathies). The Editors welcome original articles from all areas of the field: • Clinical aspects, such as new clinical entities, case studies of interest, treatment, management and rehabilitation (including biomechanics, orthotic design and surgery). • Basic scientific studies of relevance to the clinical syndromes, including advances in the fields of molecular biology and genetics. • Studies of animal models relevant to the human diseases. The journal is aimed at a wide range of clinicians, pathologists, associated paramedical professionals and clinical and basic scientists with an interest in the study of neuromuscular disorders.