Skip to main content

Table 3 Overview of included studies. For each recording modality, more recent articles appear first in the table

From: Headache-related circuits and high frequencies evaluated by EEG, MRI, PET as potential biomarkers to differentiate chronic and episodic migraine: Evidence from a systematic review

Source (DOI) Recording modality Population Study design Potential biomarker Major findings
Pan L et al. (2022) [14] EEG and MEG NA Non-systematic review article Beta connectivity (node degree) in anterior cingulate cortex The resting state electrophysiology (power and connectivity) can be used to detect pathological alterations in patients with migraine
Ong J et al. (2012) [15] EEG (from PSG) NA Non-systematic review article Slow-wave sleep activity Biobehavioral model to describe the interaction between headache, its impact on insomnia and sleep physiology, and the downstream propensity for future headache attacks
Gomez-Pilar J et al. (2020) [16] EEG HC: 39
CM: 42
EM: 45 (interictal)
Spectral and nonlinear analysis during resting state recordings (eyes closed) Relative power in high beta frequency band Significant differences in relative power between CM and EM around high beta frequency band. Significant differences in alpha between HC and migraine patients
Lisicki M et al. (2020) [17] EEG HC: 20
CM: 20
EM: 50 (30 interictal and 20 ictal)
Spectral analysis of Visual Evoked Potentials (VEPs) Spectral power in gamma band in occipital areas during VEPs recordings Gamma/alpha activity ratio can distinguish between CM and EM in the interictal stage. Non-significant differences were found between CM and EM during migraine attack
Fogang Y et al. (2015) [18] EEG HC: 24
CM: 48
EM: 232 (61 with aura and 171 without aura)
Spectral analysis of Steady-State Visual Evoked Potentials (SSVEPs) Photic driving power at 20 Hz during SSVEPs in occipital areas Increased photic driving power is related to attack duration and identifies subgroups of migraine patients with different habituation of cortical visual responses
Hsiao F et al. (2021a) [19] MEG HC: 65
CM: 80
EM: 70
Connectivity analysis of resting state activity (eyes closed) Node strength in the beta band based on imaginary coherence in the ACC Reduced beta connectivity in the anterior cingulate cortex linked to migraine chronification
Hsiao F et al. (2021b) [20] MEG CM: 37
EM: 30
Temporal analysis of SEF recordings during interictal stage Amplitude of SEF after treatment Somatosensory gating responses are similarly associated with treatment outcomes in patients with CM and EM
Wu T et al. (2018) [21] MEG HC: 35
CM: 15
EM: 20 (interictal)
Time–frequency analysis of emotional stimulation responses focused on M100 and M170 components NA (authors did not find significant differences between EM and CM groups) CM and EM have abnormal brain activity in the gamma band in response to negative emotional stimuli during interictal stage
Chen W et al. (2012) [22] MEG CM: 15
EM: 10 (remitted CM)
Temporal analysis of visual evoked fields (VEF) P100m amplitude Visual cortex excitability is dynamically modulated (reduced) in remission from CM to EM
Chen W et al. (2011a) [23] MEG HC: 24
CM: 18
EM: 39
Persistent visual aura patients: 6
Temporal analysis of visual evoked fields (VEF) NA (authors did not find significant differences between EM and CM groups) Patients with persistent visual aura maintains a steady-state hyperexcitability without significant dynamic modulation
Chen W et al. (2011b) 24[] MEG HC: 32
CM: 25
EM: 38 (29 interictal and 9 ictal)
Temporal analysis of visual evoked fields (VEF) P100m amplitude Different underlying mechanisms for interictal excitability of CM (hyperexcitability with habituation) and EM (hypo-excitability with potentiation)
Dai W et al. (2021) [25] fMRI HC: 30
CM: 17 (with medication-overuse)
EM: 18
Connectivity analysis during resting state Functional connectivity between bilateral habenula and salience network Increased functional connectivity between bilateral habenula and salience network (correlated with medication overuse duration) in patients with CM compared with patients with EM and HC, respectively
Chen C et al. (2019) [26] fMRI HC: 32
CM: 17
EM: 39 (19 infrequent EM and 20 frequent EM)
ReHo applied to resting state fMRI recordings ReHo in resting state at bilateral precentral gyri The regions affected by migraine change with the chronification of the disease
Chen Z et al. (2019) [27] fMRI HC: 21
CM: 16
EM: 18
Anatomical and functional connectivity analysis during fMRI resting state recordings Functional connectivity in anterior hypothalamus Volume of hypothalamus (HTH) was significantly decreased on CM and EM vs HC. Decreased volume of anterior HTH in CM vs HC and CM vs EM. Increased functional connectivity between anterior HTH and MOrG in CM vs EM
Lerebours F et al. (2019) [28] fMRI CM: 25 (with medication-overuse)
EM: 22 (interictal)
Connectivity analysis during fMRI resting state recordings Functional connectivity in hypothalamus Significant connectivity between anterior hypothalamus and trigeminal nucleus for CM vs. EM, no correlated with pain intensity. This connectivity is similar to that seen in the preictal phase of EM, suggesting that CM are locked in the preictal phase
Bogdanov V et al. (2019) [29] fMRI HC: 24
CM: 7 (with medication-overuse)
EM: 19 (14 interictal and 5 ictal)
fMRI during transitions between continuous noxious cold and innocuous warm thermal stimulations BOLD response to cold/warm transitions of thermal stimuli over motor cortex and superior temporal sulcus Migraine patients showed hyperactivation on “salience-matrix” areas compared to HC. CM and EM (ictal) showed increased unspecific transitional BOLD responses in motor cortex and superior temporal sulcus versus EM (interictal) and HC. CM overactivated also other “salience” matrix areas compared to EM-b
Imai N (2018) [30] fMRI CM: 31
EM: 31
Functional connectivity analysis during resting state Functional connectivity in ACC and ROcG Increased FC in ACC and ROcG for CM vs EM, suggesting that occipital pole plays a key role in migraine chronification
Schulte L et al. (2017) [31] fMRI HC:19
CM: 17
EM: 18
Analysis of fMRI during painful ammonia stimulation Functional activation of the hypothalamus during ammonia stimulation Hypothalamus plays a crucial role in the pathophysiology of migraine chronification and acute pain stage of migraine patients
Chen Z et al. (2017) [32] fMRI HC: 18
CM: 16
EM: 18
Anatomical and connectivity analysis during resting state (interictal) Functional connectivity (FC) in bilateral amygdala Amygdala volume showed no differences between groups. Increased FC between amygdala and inferior temporal gyrus and
orbitofrontal gyrus for CM vs EM. Enhanced FC in left amygdala for EM vs. HC. Decreased FC in right amygdala for CM vs HC
Hubbard C et al. (2016) [33] fMRI CM: 11 (responders to BoNT-A)
CM: 12 (non-responders to BoNT-A)
Longitudinal anatomical and functional connectivity analyses during resting state (before and after treatment) Functional connectivity between SI-DMPFC and SI-LOC Responders showed significant cortical thickening (SI, aINS, STG, ParsOp) and functional connectivity differences (SI-LOC, SI-DMPFC) in CM that reverted to EM compared to CM that did not respond to the treatment
Chen Z et al. (2016) [34] fMRI HC: 32
CM: 60 (44 with medication-overuse)
EM: 18
Connectivity analysis during resting state Functional connectivity between marginal division of neostriatum (MrD) and pain network MrD demonstrated different pain modulation patterns in different subtypes of headache. Functional connectivity between MrD and other regions yielded significant differences between all groups. Compared with EM, both CM and HC showed decreased functional connectivity of MrD
Torres-Ferrus M et al. (2021) [35] PET HC: 11
CM: 7
EM: 8 (interictal)
Resting F-FDG-PET in interictal stages Not available (no changes CM vs EM) CM patients showed frontotemporal hypometabolism and increased frontal cortical thickness (CTh) when compared to HC. EM presented intermediated values but not significant
Jassar H et al. (2019) [36] PET HC: 7
CM: 7
EM: 8 (ictal)
Measure μ-opioid in resting PET after injection of μOR radiotracer with and without thermal pain threshold challenge μOR availability measured with [11C]carfentanil nondisplaceable binding potential (BPND) CM had decreased μOR BPND relative to HC in thalamus and left caudate. Lower μOR BPND in right parahippocampal region and right amygdala compared to EM. Increased endogenous μOR receptor is seen in the limbic system of CM patients
Deen M et al., (2019) [37] PET HC: 16
CM: 16
EM: 15
Measure of serotonin 5-HT in resting PET scans after injection of 5-HT4 receptor radioligand Not available (no changes CM vs EM) CM had 9.1% lower binding than HC. Thus, cerebral levels of 5-HT are elevated in CM. No significant differences between CM and EM. No association between binding and no migraine days. Elevated 5-HT may not be a risk factor for conversion from EM to CM
  1. NA Not applicable
  2. HC Healthy control; EM Episodic migraine; CM Chronic migraine; PSG Polysomnography; BoNT-A onabotulinumtoxinA
  3. SEF Somatosensory Evoked Field; ReHo: Regional Homology Analysis Method