Cluster analysis showed long-term cognition can be predicted by looking at regional grey matter atrophy in the first two years of multiple sclerosis course.
Stefano Ziccardi, Maddalena Guandalini, Francesco Crescenzo, Luigi Martinelli, Agnese Tamanti, Gian Marco Schiavi, Albulena Bajrami, Valentina Camera, Damiano Marastoni, Massimiliano Calabrese
{"title":"Cluster analysis showed long-term cognition can be predicted by looking at regional grey matter atrophy in the first two years of multiple sclerosis course.","authors":"Stefano Ziccardi, Maddalena Guandalini, Francesco Crescenzo, Luigi Martinelli, Agnese Tamanti, Gian Marco Schiavi, Albulena Bajrami, Valentina Camera, Damiano Marastoni, Massimiliano Calabrese","doi":"10.1136/jnnp-2025-335925","DOIUrl":null,"url":null,"abstract":"<p><strong>Background: </strong>Grey matter (GM) atrophy is associated with cognitive impairment (CI) in multiple sclerosis (MS). We aimed to investigate the predictive role of early regional GM damage for long-term CI.</p><p><strong>Methods: </strong>A post-hoc cluster analysis was conducted on 175 patients with MS followed for 20 years from onset. Participants underwent a 1.5T-MRI scanning at diagnosis and after 2 years, and a comprehensive neuropsychological assessment after 20 years.</p><p><strong>Results: </strong>Three clusters have been identified: cluster 1 (primarily patients with long-term normal cognition), cluster 2 (primarily patients with long-term mild CI) and cluster 3 (primarily patients with long-term severe CI). Five brain regions have been identified showing a significant difference in early atrophy from cluster 1 and both clusters 2 and 3: precuneus (1 vs 2: p<0.001, relative risk ratio (RRR)=4.9, 95% confidence intervals (95% CIs) =2.4-10.1; 1 vs 3: p<0.001, RRR=5.5, 95% CIs=3.1-9.7), insula (1 vs 2: p<0.001, RRR=4.1, 95% CIs=1.9-8.6; 1 vs 3: p<0.001, RRR=4.3, 95% CIs=2.5-7.5), parahippocampal gyrus (1 vs 2: p<0.001, RRR=3.2, 95% CIs=1.8-5.7; 1 vs 3: p<0.001, RRR=3.1, 95% CIs=2.1-4.6), cingulate gyrus (1 vs 2: p<0.001, RRR=3.0, 95% CIs=1.7-5.3; 1 vs 3: p<0.001, RRR=2.2, 95% CIs=1.6-5.3) and cerebellum (1 vs 2: p=0.027, RRR=2.6, 95% CIs=1.5-4.6; 1 vs 3: p<0.001, RRR=2.1, 95% CIs=1.5-2.9). Four additional brain regions showed a significant difference in terms of early atrophy between cluster 1 and cluster 3: precentral gyrus (p<0.001, RRR=7.3, 95% CIs=3.1-17.3), postcentral gyrus (p<0.001, RRR=4.6, 95% CIs=2.2-9.8), superior frontal gyrus (p<0.001, RRR=4.0, 95% CIs=2.0-8.0) and hippocampus (p<0.001, RRR=2.4, 95% CIs=1.6-3.6).</p><p><strong>Conclusions: </strong>Cluster analysis identified the most specific brain regions whose early atrophy best distinguished future patients with CI. Long-term CI accumulation in MS can be predicted by early GM volume loss of specific cortical/deep GM regions.</p>","PeriodicalId":16418,"journal":{"name":"Journal of Neurology, Neurosurgery, and Psychiatry","volume":" ","pages":""},"PeriodicalIF":8.7000,"publicationDate":"2025-04-23","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"0","resultStr":null,"platform":"Semanticscholar","paperid":null,"PeriodicalName":"Journal of Neurology, Neurosurgery, and Psychiatry","FirstCategoryId":"3","ListUrlMain":"https://doi.org/10.1136/jnnp-2025-335925","RegionNum":1,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"Q1","JCRName":"CLINICAL NEUROLOGY","Score":null,"Total":0}
引用次数: 0
Abstract
Background: Grey matter (GM) atrophy is associated with cognitive impairment (CI) in multiple sclerosis (MS). We aimed to investigate the predictive role of early regional GM damage for long-term CI.
Methods: A post-hoc cluster analysis was conducted on 175 patients with MS followed for 20 years from onset. Participants underwent a 1.5T-MRI scanning at diagnosis and after 2 years, and a comprehensive neuropsychological assessment after 20 years.
Results: Three clusters have been identified: cluster 1 (primarily patients with long-term normal cognition), cluster 2 (primarily patients with long-term mild CI) and cluster 3 (primarily patients with long-term severe CI). Five brain regions have been identified showing a significant difference in early atrophy from cluster 1 and both clusters 2 and 3: precuneus (1 vs 2: p<0.001, relative risk ratio (RRR)=4.9, 95% confidence intervals (95% CIs) =2.4-10.1; 1 vs 3: p<0.001, RRR=5.5, 95% CIs=3.1-9.7), insula (1 vs 2: p<0.001, RRR=4.1, 95% CIs=1.9-8.6; 1 vs 3: p<0.001, RRR=4.3, 95% CIs=2.5-7.5), parahippocampal gyrus (1 vs 2: p<0.001, RRR=3.2, 95% CIs=1.8-5.7; 1 vs 3: p<0.001, RRR=3.1, 95% CIs=2.1-4.6), cingulate gyrus (1 vs 2: p<0.001, RRR=3.0, 95% CIs=1.7-5.3; 1 vs 3: p<0.001, RRR=2.2, 95% CIs=1.6-5.3) and cerebellum (1 vs 2: p=0.027, RRR=2.6, 95% CIs=1.5-4.6; 1 vs 3: p<0.001, RRR=2.1, 95% CIs=1.5-2.9). Four additional brain regions showed a significant difference in terms of early atrophy between cluster 1 and cluster 3: precentral gyrus (p<0.001, RRR=7.3, 95% CIs=3.1-17.3), postcentral gyrus (p<0.001, RRR=4.6, 95% CIs=2.2-9.8), superior frontal gyrus (p<0.001, RRR=4.0, 95% CIs=2.0-8.0) and hippocampus (p<0.001, RRR=2.4, 95% CIs=1.6-3.6).
Conclusions: Cluster analysis identified the most specific brain regions whose early atrophy best distinguished future patients with CI. Long-term CI accumulation in MS can be predicted by early GM volume loss of specific cortical/deep GM regions.
背景：灰质（GM）萎缩与多发性硬化症（MS）的认知障碍（CI）有关。我们的目的是研究早期区域性GM损伤对长期CI的预测作用。方法：对175例多发性硬化症患者进行回顾性聚类分析。参与者在诊断时和2年后进行1.5T-MRI扫描，并在20年后进行全面的神经心理学评估。结果：已经确定了三个集群：集群1（主要是长期认知正常的患者），集群2（主要是长期轻度CI患者）和集群3（主要是长期重度CI患者）。已经确定了5个脑区域，在第1类和第2类和第3类的早期萎缩中表现出显著差异：楔前叶（1 vs 2）。结论：聚类分析确定了最具体的脑区域，其早期萎缩最能区分未来CI患者。MS的长期CI积累可以通过特定皮质/ GM深部区域的早期GM体积损失来预测。
期刊介绍：
The Journal of Neurology, Neurosurgery & Psychiatry (JNNP) aspires to publish groundbreaking and cutting-edge research worldwide. Covering the entire spectrum of neurological sciences, the journal focuses on common disorders like stroke, multiple sclerosis, Parkinson’s disease, epilepsy, peripheral neuropathy, subarachnoid haemorrhage, and neuropsychiatry, while also addressing complex challenges such as ALS. With early online publication, regular podcasts, and an extensive archive collection boasting the longest half-life in clinical neuroscience journals, JNNP aims to be a trailblazer in the field.
求助内容：
应助结果提醒方式：
京公网安备 11010802042870号
