Department of Neurology, Samsung Medical Center, Sungkyunkwan University School of Medicine, Seoul, Korea.
Neuroscience Center, Samsung Medical Center, Seoul, Korea.
Pain. 2019 Dec;160(12):2776-2786. doi: 10.1097/j.pain.0000000000001676.
Migraine headache is an episodic phenomenon, and patients with episodic migraine have ictal (headache), peri-ictal (premonitory, aura, and postdrome), and interictal (asymptomatic) phases. We aimed to find the functional characteristics of the migraine brain regardless of headache phase using dynamic functional connectivity analysis. We prospectively recruited 50 patients with migraine and 50 age- and sex-matched controls. All subjects underwent a resting-state functional magnetic resonance imaging. Significant networks were defined in a data-driven fashion from the interictal (>48 hours apart from headache phases) patients and matched controls (interictal data set) and tested to ictal or peri-ictal patients and controls (ictal/peri-ictal data set). Both static and dynamic analyses were used for the between-group comparison. A false discovery rate correction was performed. As a result, the static analysis did not reveal a network which was significant in both interictal and ictal/peri-ictal data sets. Dynamic analysis revealed significant between-group differences in 7 brain networks in the interictal data set, among which a frontoparietal network (controls > patients, P = 0.0467), 2 brainstem networks (patients > controls, P = 0.0467 and <0.001), and a cerebellar network (controls > patients, P = 0.0408 and <0.001 in 2 states) remained significant in the ictal/peri-ictal data set. Using these networks, migraine was classified with a sensitivity of 0.70 and specificity of 0.76 in the ictal/peri-ictal data set. In conclusion, the dynamic connectivity analysis revealed more functional networks related to migraine than the conventional static analysis, suggesting a substantial temporal fluctuation in functional characteristics. Our data also revealed migraine-related networks which show significant difference regardless of headache phases between patients and controls.
偏头痛是一种发作性现象，发作性偏头痛患者有发作期（头痛）、发作间期（先兆、前驱期和后驱期）和发作间期（无症状期）。我们旨在使用动态功能连接分析来寻找无论头痛阶段如何的偏头痛大脑的功能特征。我们前瞻性招募了 50 名偏头痛患者和 50 名年龄和性别匹配的对照者。所有受试者均接受了静息态功能磁共振成像检查。从发作间期（与头痛期相隔>48 小时）的患者和匹配的对照者（发作间期数据集中）中以数据驱动的方式定义显著网络，并对发作期或发作前期的患者和对照者进行测试（发作期/发作前期数据集中）。使用静态和动态分析进行组间比较。进行了假发现率校正。结果，静态分析未显示在发作间期和发作期/发作前期数据集中均显著的网络。动态分析显示，在发作间期数据集中，7 个脑网络存在显著的组间差异，其中一个额顶网络（对照组>患者，P=0.0467）、2 个脑干网络（患者>对照组，P=0.0467 和<0.001）和一个小脑网络（对照组>患者，P=0.0408 和<0.001 在 2 种状态下）在发作期/发作前期数据集中仍然显著。使用这些网络，偏头痛在发作期/发作前期数据集中的分类敏感性为 0.70，特异性为 0.76。总之，与传统的静态分析相比，动态连接分析揭示了更多与偏头痛相关的功能网络，提示功能特征存在实质性的时间波动。我们的数据还揭示了无论患者和对照者之间的头痛阶段如何，都存在显著差异的偏头痛相关网络。
