血清神经丝轻链对多发性硬化症增强病变的识别准确性。
Accuracy of serum neurofilament light to identify contrast-enhancing lesions in multiple sclerosis.
机构信息
Department of Neurology, Brigham and Women's Hospital, Boston, MA, USA; Harvard Medical School, Boston, MA, USA/Department of Medicine, The University of Ottawa and Ottawa Hospital Research Institute, Ottawa, ON, Canada.
Department of Neurology, Brigham and Women's Hospital, Boston, MA, USA/Harvard Medical School, Boston, MA, USA.
出版信息
Mult Scler. 2023 Oct;29(11-12):1418-1427. doi: 10.1177/13524585231198751. Epub 2023 Sep 15.
BACKGROUND
Contrast-enhancing magnetic resonance imaging (MRI) lesions (CELs) indicate acute multiple sclerosis inflammation. Serum biomarkers, neurofilament light (sNfL), and glial fibrillary acidic protein (sGFAP) may increase in the presence of CELs, and indicate a need to perform MRI.
OBJECTIVE
We assessed the accuracy of biomarkers to detect CELs.
METHODS
Patients with two gadolinium-enhanced MRIs and serum biomarkers tested within 3 months were included ( = 557, 66% female). Optimal cut-points from Bland-Altman analysis for spot biomarker level and Youden's index for delta-change from remission were evaluated.
RESULTS
A total of 116 patients (21%) had CELs. A spot sNfL measurement >23.0 pg/mL corresponded to 7.0 times higher odds of CEL presence (95% CI: 3.8, 12.8), with 25.9% sensitivity, 95.2% specificity, operating characteristic curve (AUC) 0.61; while sNfL delta-change >30.8% from remission corresponded to 5.0 times higher odds (95% CI: 3.2, 7.8), 52.6% sensitivity, 81.9% specificity, AUC 0.67. sGFAP had poor CEL detection. In patients > 50 years, neither cut-point remained significant. sNfL delta-change outperformed spot levels at identifying asymptomatic CELs (AUC 0.67 vs 0.59) and in patients without treatment escalation between samples (AUC 0.67 vs 0.57).
CONCLUSION
Spot sNfL >23.0 pg/mL or a 30.8% increase from remission provides modest prediction of CELs in patients <50 years; however, low sNfL does not obviate the need for MRI.
背景
对比增强磁共振成像（MRI）病变（CELs）表明急性多发性硬化症炎症。在存在 CELs 的情况下，血清生物标志物神经丝轻链（sNfL）和胶质纤维酸性蛋白（sGFAP）可能会增加，并表明需要进行 MRI。
目的
我们评估了生物标志物检测 CELs 的准确性。
方法
纳入了在 3 个月内进行了两次钆增强 MRI 和血清生物标志物检测的患者（ = 557 例，66%为女性）。通过 Bland-Altman 分析评估了斑点生物标志物水平的最佳切点和缓解期 delta 变化的 Youden 指数。
结果
共有 116 例患者（21%）出现 CELs。单点 sNfL 测量值>23.0 pg/mL 对应于 CEL 存在的可能性高 7.0 倍（95%CI：3.8，12.8），敏感性为 25.9%，特异性为 95.2%，曲线下面积（AUC）为 0.61；而 sNfL 缓解期 delta 变化>30.8%对应于可能性高 5.0 倍（95%CI：3.2，7.8），敏感性为 52.6%，特异性为 81.9%，AUC 为 0.67。sGFAP 对 CEL 的检测能力较差。在年龄>50 岁的患者中，这两个切点均不显著。sNfL 缓解期 delta 变化优于单点水平，可识别无症状 CELs（AUC 0.67 与 0.59）和两次采样之间无治疗升级的患者（AUC 0.67 与 0.57）。
结论
单点 sNfL >23.0 pg/mL 或缓解期增加 30.8%可适度预测年龄<50 岁患者的 CELs；然而，低 sNfL 并不能排除 MRI 的必要性。
Zotero 插件