Table 1.

A summary of the main findings of neuroimaging studies in chronic migraine patients.

Study	Study cohorts	Main findings	Potential confounders
RESTING STATE fMRI STUDIES
Androulakis et al. (14, 15)	13 CM patients without MOH (all females), 16 CM patients with MOH (all females) vs. 19 controls (all females)	Compared to controls, CM patients, regardless of MOH status, showed: - ↓ overall network connectivity of the DMN, SN and ECN Frequency of headache attacks was negatively correlated with the strength of the SN and ECN intrinsic connectivity Severity of cutaneous allodynia was positively correlated with the strength of the SN intrinsic connectivity	15 patients were taking migraine prophylaxis
Chen et al. (16)	16 CM patients without MOH, 18 EM patients vs. 21 controls	Compared to controls and EM patients, CM patients had: - ↑ RS FC between the anterior hypothalamus and the right orbital gyrus	14 CM patients had headache without migraine features during MRI scanning
Coppola et al. (17)	20 CM patients without MOH vs. 20 controls	Compared to controls, CM patients showed: - ↓ RS FC between the DMN and ECN - ↑ RS FC between the DAS and DMN - ↓ RS FC between the DAS and ECN. The severity of headache was positively correlated with the strength of the DAS intrinsic connectivity The severity of headache was negatively correlated with the strength of the ECN intrinsic connectivity
Lee et al. (18)	19 CM patients without MOH vs. 45 EM patients	Compared to EM patients, CM patients showed: - ↑ RS FC of pain processing areas, including the anterior cingulate cortex - ↓ RS FC between pain processing brain areas and the hypothalamus - ↑ RS FC between pain processing brain areas and the dorsal raphe nucleus
Lerebours et al. (19)	25 CM patients with MOH vs. 22 EM patients	Compared to EM patients, CM patients had: - ↑ RS FC between the anterior hypothalamus and the spinal trigeminal nucleus	16 CM patients had mild headache during the MRI 6 CM patients were taking migraine prophylaxis
Schwedt et al. (20)	20 CM patients without MOH vs. 20 controls	Atypical RS FC between: - Left anterior insula and left pulvinar, parieto-temporal areas, right precuneus, cingulate cortex and bilateral thalamus - Right anterior insula and left pulvinar, right periaqueductal gray, middle temporal cortex, and bilateral thalamus - Left amygdala and right superior frontal gyrus - Right amygdala and occipital cortex Disease duration was positively correlated with: - RS FC between bilateral anterior insula and right thalamus - RS FC between right anterior insula and right periaqueductal gray Anxiety scores were negatively correlated with: - RS FC between right anterior insula and right periaqueductal gray	8 patients were taking migraine prophylaxis
fMRI STUDIES OF NOCICEPTIVE STIMULATION
Ferraro et al. (21)	9 CM patients with MOH (all females) vs. 9 controls (all females)	Compared to controls, CM patients with MOH showed: - ↓ pain-related activity of bilateral inferior parietal lobule, somatosensory cortex and right supramarginal gyrus In CM patients, the activity of pain processing regions normalized at 6 months after withdrawal	During the MRI exam, all patients had a moderate headache
Schulte et al. (22)	17 CM patients without MOH, 18 EM patients vs. 19 controls	Compared to controls, CM patients showed: - ↑ activation of the anterior right hypothalamus Compared to controls and migraineurs (EM and CM) without headache, migraineurs with headache showed: - ↑ activation of the posterior hypothalamus bilaterally	4 CM patients were taking migraine prophylaxis 19 patients (7 EM and 12 CM patients) had headache during the MRI
fMRI STUDIES DURING A DECISION-MAKING TASK
Ferraro et al. (23)	8 CM patients with MOH (all females), 8 detoxified CM patients with MOH (all females), 8 CM patients without MOH (all females) vs. 8 controls (all females)	Compared to controls, CM patients with MOH showed: - ↓ task-related activity in the substantia nigra/ventral tegmental area complex - ↑ task-related activity in the ventromedial prefrontal cortex Compared to CM without MOH, CM patients with MOH showed: - ↓ task-related activity in the substantia nigra/ventral tegmental area complex Compared to detoxified MOH patients, CM patients with MOH showed: - ↑ task-related activity in the ventromedial prefrontal cortex	During the MRI exam, all patients had a moderate headache
fMRI STUDIES OF VISUAL STIMULATION
Schulte et al. (24)	17 CM patients without MOH, 18 EM patients vs. 19 controls	Compared to controls, CM patients showed: - ↑ activation of the spinal trigeminal nucleus and superior colliculi	4 CM patients were taking migraine prophylaxis 19 patients (7 EM and 12 CM patients) had headache during the MRI
POSITRON EMISSION TOMOGRAPHY STUDIES
Aurora et al. (25)	10 CM patients with or without MOH	CM patients had: - ↑ metabolism in the pons and right temporal cortex compared to the global cerebral metabolism - ↓ metabolism in the bilateral caudate nuclei, frontal and parietal cortex compared to the global cerebral metabolism
Fumal et al. (26)	16 CM patients with MOH vs. 68 controls	Before withdrawal, compared to controls, CM patients with MOH showed: - ↓ metabolism of the bilateral thalamus, orbitofrontal cortex, anterior cingulate gyrus, insula, ventral striatum, and right inferior parietal lobule - ↑ cerebellar metabolism In CM patients with MOH, all dysmetabolic areas recovered to almost normal glucose uptake 3 weeks after withdrawal, except the orbitofrontal cortex where a further metabolic decrease was found
Matharu et al. (27)	8 CM patients	No significant differences in the activity of the dorsal rostral pons in CM patients during pain and in pain-free patients during bilateral suboccipital stimulation ↓ activation of the anterior cingulate cortex in pain-free CM patients during bilateral suboccipital stimulation ↑ activation of the anterior cingulate cortex and cuneus in CM patients during pain
PROTON MAGNETIC RESONANCE SPECTROSCOPY STUDIES
Niddam et al. (28)	25 CM patients without MOH, 24 EM patients vs. 25 controls	Compared to controls, CM patients had: - ↓ N-acetyl-aspartate concentration of the right thalamus and anterior cingulate cortex - Altered interregional N-acetyl-aspartate correlations between the thalamus and anterior cingulate cortex and between the thalamus and occipital cortex in the right hemisphere Compared to controls and EM patients, CM patients had: - ↓ N-acetyl-aspartate concentration of the left thalamus In CM patients, the right thalamic N-acetyl-aspartate concentrations was negatively correlated with patients' disease duration In CM patients, there was a positive correlation between the N-acetyl-aspartate concentration and gray matter volume of the right anterior cingulate cortex	21 patients (3 EM and 18 CM patients) had headache the day of the MRI
MORPHOMETRIC STUDIES
Bilgic et al. (29)	17 CM patients without MOH (all females), 7 CM patients with MOH (all females) vs. 24 controls (all females)	Compared to controls, CM had: ↓ cerebellar and brainstem volume	7 patients were taking migraine prophylaxis
Chen et al. (16)	16 CM patients without MOH, 18 EM patients vs. 21 controls	Compared to controls and EM patients, patients with CM showed: - ↓ volume of the anterior hypothalamus In CM patients, the anterior hypothalamic volume was positively correlated with headache frequency Cut-off volume of the hypothalamus as 1.429 ml had a good diagnostic accuracy for CM with sensitivity of 81% and specificity of 100%	14 CM patients had headache without migraine features during MRI scanning
Coppola et al. (30)	20 CM patients without MOH vs. 20 controls	Compared to controls, CM patients had: - ↓ gray matter volume of the right cerebellum, left pallidum, amygdala, orbitofrontal, temporal, and occipital cortex In CM patients, the cerebellar gray matter volume was: - Negatively correlated with patients' disease duration - Positively correlated with the number of acute medications taken per month	4 patients had mild headache without migrainous features during the MRI exam
Hubbard et al. (31)	23 CM patients (11 responders and 12 non-responders to prophylactic treatment with onabotulinumtoxinA)	Compared to non-responders, patients who responded to onabotulinumtoxinA showed: - ↑ cortical thickness of the right primary somatosensory cortex, anterior insula, left superior temporal gyrus and pars opercularis In responders patients, disease duration was: - Negatively correlated with cortical thickness of fronto-parietal and temporo-occipital regions - Positively associated to the cortical thickness of the left primary motor cortex In non-responders patients, disease duration was: - Negatively associated to the cortical thickness of the left primary motor cortex - Positively associated to the cortical thickness of the left inferior temporal gyrus and lateral occipital cortex	Some patients in the non-responder group may have had mild headache the day of the MRI
Lai et al. (32)	33 CM patients with MOH (19 responders to common preventive treatments), 33 CM patients without MOH vs. 33 controls	Compared to CM patients without MOH, patients with MOH showed: - ↓ gray matter volume of the bilateral orbitofrontal cortex and left middle occipital gyrus - ↑ gray matter volume of the left temporal pole/parahippocampus In CM patients with MOH, clinical improvement after 12 months of preventive treatment was significantly associated to the gray matter volume of the orbitofrontal cortex In CM patients, gray matter volume changes could predict the frequency of analgesics use	33 patients had migraine the day of the MRI exam (13 CM patients without MOH and 20 CM patient with MOH)
Liu et al. (33)	39 CM patients, 83 EM patients (15 patients with MOH) vs. 31 controls	In CM and EM patients, the volume of the bilateral hippocampus and left amygdala varied as a function of headache frequency At 2-year follow-up, the volume of the right hippocampus was positively associated with a good migraine outcome
Neeb et al. (34)	6 CM patients without MOH, 15 CM patients with MOH, 21 EM patients vs. 21 controls	Compared to controls, CM patients had: - ↑ gray matter volume of the right amygdala, superior parietal lobule, hippocampus, parahippocampus, left insula, and bilateral basal ganglia Compared to EM patients, CM patients showed: - ↑ gray matter volume of bilateral temporal areas - ↓ gray matter volume of the left cuneus In CM and EM patients, gray matter volume alterations were influenced by headache frequency	13 patients (9 CM and 4 EM patients) were taking migraine prophylaxis
Niddam et al. (28)	25 CM patients without MOH, 24 EM patients vs. 25 controls	Compared to controls, CM patients had: - ↓ gray matter volume of the right anterior cingulate cortex
Riederer et al. (35)	31 CM patients with MOH (10 responders and 8 non-responders to medication withdrawal)	At baseline, compared to responder patients, non-responders had: - ↓ gray matter volume of the right orbitofrontal cortex Only responders patients showed ↓ gray matter volume of the midbrain after medication withdrawal Treatment response correlated positively with: - Baseline gray matter volume of the orbitofrontal cortex - Gray matter volume change in the midbrain after medication withdrawal	20 patients were taking migraine prophylaxis
Schwedt et al. (36)	15 CM patients without MOH, 51 EM patients vs. 54 controls	Average accuracy of classifiers consisting of cortical surface area, cortical thickness, and regional volumes of fronto-temporal areas was: -86.3% for CM patients vs. controls - 84.2% for CM vs. EM patients - 67.2% for EM patients vs. controls

CM, chronic migraine; DAS, dorsal attention system; DMN, default mode network; EM, episodic migraine; ECN, executive control network; fMRI, functional magnetic resonance imaging; MOH, medication overuse headache; RS FC, resting state functional connectivity; SN, salience network.