Share

March 05, 2019; 92 (10) Article

Iron deposition in periaqueductal gray matter as a potential biomarker for chronic migraine

Clara Domínguez, Ana López, Pedro Ramos-Cabrer, Alba Vieites-Prado, Maria Pérez-Mato, Carmen Villalba, Tomás Sobrino, Xiana Rodriguez-Osorio, Francisco Campos, José Castillo, Rogelio Leira
First published February 1, 2019, DOI: https://doi.org/10.1212/WNL.0000000000007047
Full PDF
Short Form
Citation
Permissions

Make Comment

See Comments

Downloads
347

Share

This article has a correction. Please see:

Abstract

Objective To study iron deposition in red nucleus (RN), globus pallidus (GP), and periaqueductal gray matter (PAG) as a potential biomarker of chronic migraine (CM) and its association with levels of biomarkers related to migraine pathophysiology.

Methods This case-control study included 112 patients with migraine (55 CM, 57 episodic migraine [EM]) and 25 headache-free controls. We analyzed iron deposition using 3T MRI and the NIH software platform ImageJ; we analyzed serum levels of markers of inflammation, endothelial dysfunction, and blood-brain barrier (BBB) disruption by ELISA in peripheral blood during interictal periods.

Results Patients with CM showed larger iron grounds volume in RN compared to patients with EM (70.2 ± 6.8 vs 25.5 ± 7.3 μL, p < 0.001) and controls (70.2 ± 6.8 vs 15.1 ± 10.8 μL, p < 0.001), as well as larger iron deposits in PAG compared to patients with EM (360.3 ± 6.5 vs 249.7 ± 6.9 μL, p < 0.001) and controls (360.3 ± 6.5 vs 168.6 ± 10.3 μL, p < 0.001). In PAG, differences were also significant between patients with EM and controls. No significant differences were obtained for GP. Receiver operating characteristic curves showed that the optimal threshold for iron volume was 15 μL in RN (80% sensitivity, 71% specificity) and 240 μL in PAG (93% sensitivity, 97% specificity). Iron grounds volume in PAG was correlated with higher plasma levels of soluble tumor necrosis factor–like WEAK (r = 0.395, p = 0.005) and cellular fibronectin (r = 0.294, p = 0.040).

Conclusions Patients with CM showed increased iron deposition in RN and PAG compared to patients with EM and controls. Iron grounds volume in PAG identified correctly patients with CM and was associated with elevated biomarkers of endothelial dysfunction and BBB disruption.

Glossary

BBB=
blood-brain barrier;
BG=
basal ganglia;
CAMERA=
Cerebral Abnormalities in Migraine, an Epidemiological Risk Analysis;
cFn=
cellular fibronectin;
CGRP=
calcitonin gene-related peptide;
EM=
episodic migraine;
GP=
globus pallidus;
hs-CRP=
high-sensitivity C-reactive protein;
IL=
interleukin;
PAG=
periaqueductal gray matter;
PTX3=
pentraxin 3;
RN=
red nucleus;
ROI=
region of interest;
sTWEAK=
soluble tumor necrosis factor–like WEAK;
TGV=
trigeminovascular;
TNF-α=
tumor necrosis factor-α

The pathophysiology of migraine includes both vascular and neural mechanisms.1 It involves nociceptive inputs from the raphe and locus coeruleus nuclei,2 the cortical spreading depression phenomenon,3 and the trigeminovascular (TGV) system activation. Inflammatory vasoactive peptides promote dilatation of the meningeal vessels and modulate endothelial function.4 The inflammatory response also contributes to endothelial dysfunction5 and could induce blood-brain barrier (BBB) disruption6 during migraine attacks.

Red nucleus (RN), the periaqueductal gray matter (PAG), and some basal ganglia (BG) such as globus pallidus (GP) have been implicated in the pathophysiology of migraine.2 PAG is the center of a descending antinociceptive neuronal network7 considered a major nodal point in the CNS. It regulates autonomic adjustments to antinociceptive, autonomic, and behavioral responses to threat.8 RN has been also implicated in the modulation of pain.9

Several studies10,–,12 have reported iron deposition in the putamen, GP, RN, and PAG in patients with episodic migraine (EM) and chronic daily headache, even correlating iron deposits and duration of the disorder.10 However, a follow-up study did not support the existence of a progression of iron accumulation in relation to duration of migraine.13

This study aims to compare differences in iron deposition in RN, GP, and PAG measured by MRI between patients with EM, patients with CM, and healthy control volunteers. In addition, iron deposition was correlated with clinical and molecular markers involved in migraine pathophysiology to evaluate the potential role of these mechanisms in iron accumulation.

Methods

Study population

Participants were recruited prospectively from the outpatient Headache Clinic of the Department of Neurology at the Hospital Clínico Universitario de Santiago de Compostela. Participation in the study was offered to all patients fulfilling inclusion and exclusion criteria during a 2-month period. One hundred twenty individuals diagnosed with EM or CM according to International Classification of Headache Disorders, 3rd edition criteria14 were selected. Preventive treatment use was allowed. Eight patients did not complete the study due to rejection to perform the MRI or the blood collection, or because they did not show up. Twenty-five healthy participants without a history of headache were enrolled as a control group. Controls were recruited from the hospital and university staff, students, and general population visiting our hospital. Controls were contacted directly or through public advertisements. All controls completed the study. All participants were >18 years old. Clinical variables were recorded; imaging studies were performed at 3T MRI; and selected molecular makers were determined in peripheral blood.

Exclusion criteria were the following: (1) high blood pressure (known high blood pressure or >2 measurements >140/90 mm Hg); (2) coronary disease; (3) diabetes mellitus; (4) hypercholesterolemia (pharmacologically treated or fasting serum cholesterol >200 mg/dL); (5) infectious diseases; (6) chronic inflammatory conditions such as rheumatoid arthritis, inflammatory bowel disease, systemic lupus, and other autoimmune conditions; (7) severe systemic diseases; (8) oligomenorrhea, polymenorrhea, or polycystic ovarian syndrome; (9) pregnancy or lactation; (10) obesity (body mass index >30 kg/m2); (11) smoking habit (within the previous 12 months); and (12) recent consumption of vasoactive drugs (>4 times the medium half-life of the active substance). Healthy controls fulfilled all inclusion and exclusion criteria and were free of any headache or psychiatric disorder.

Standard protocol approvals, registrations, and patient consents

The Research Ethics Committee of the Hospital Clínico Universitario de Santiago de Compostela approved the study. Written informed consent was obtained from all participants.

Study protocol

After a screening visit, eligible participants were invited to perform a MRI scan and blood sample collection. Migraineurs were headache-free from 24 hours before the visit to 24 hours after the blood sample collection. If a migraine occurred within the first 24 hours, measurements were repeated in another headache-free period. Participants had not consumed anti-inflammatory or analgesic medication in the previous 24 hours.

Clinical variables

All participants underwent a medical interview that included demographic data (age, sex) and personal and family histories. Physical examination was performed, and neuroimaging results were registered. For migraineurs, type of migraine (with or without aura), frequency of headaches (number of days of pain per month), time of evolution of migraine (measured in years from first symptoms), duration of CM (measured in months since diagnosis of CM), intensity of headaches (measured by average score in the visual analog scale in the last month), and duration of attacks (quantified in hours, based on the average duration of attacks in the last month) were registered.

Laboratory tests

Participants underwent a blood draw to measure serum levels of markers of inflammation (interleukin [IL]-6, IL-10, tumor necrosis factor-α [TNF-α], and high-sensitivity C-reactive protein [hs-CRP]), TGV activation (calcitonin gene-related peptide [CGRP]), and endothelial dysfunction (pentraxin 3 [PTX3], soluble tumor necrosis factor–like WEAK [sTWEAK], and BBB disruption [cellular fibronectin [cFn]).

Chemistry test tubes were used to collect 4.5 mL blood from the antecubital vein after an overnight fast. Samples were centrifuged at 3000g for 15 minutes, immediately frozen, and stored at −80°C. Serum levels of PTX3 and sTWEAK (Assay Biotech, Sunnyvale, CA), cFn (Cusabio Life Science, Wuhan, China), and CGRP (Phoenix Pharmaceuticals Inc, Burlingame, CA) were measured with commercial ELISA kits following manufacturer instructions. IL-6, IL-10, TNF-α, and hs-CRP were measured with an immunodiagnostic IMMULITE 1,000 System (Siemens Healthcare Global, Los Angeles, CA). Determinations were performed in a laboratory blinded to clinical data.

MRI acquisition

Images were acquired with a birdcage volumetric head coil in a 3T Philips (Best, the Netherlands) Achieva system. A turbo spin echo T2-weighted image was acquired with a train of 15 echoes with an effective echo time of 120 milliseconds. Repetition time was set to 3 seconds, with an excitation pulse of 90°. Two averages were acquired covering the whole brain with 28 axial slices of 4-mm thickness with an interslice gap of 1 mm (5-mm distance from center to center of consecutive slices). A field of view of 230 × 230 mm2 was covered with a matrix of 576 × 576 points, giving an in-plane resolution of 0.4 × 0.4 mm2.

Image analysis

Images were analyzed by a neuroradiologist blinded to clinical data with self-developed routines for the NIH software platform ImageJ15 and following a modified version of the methodology described by Jurgens et al.,16 which is as follows:

Step 1, image normalization

For each participant, a region of interest (ROI) covering a large portion of the corpus callosum (typically an area of 500–600 pixels containing only white matter) was selected in a central slice of the brain, and the mean signal intensity of such area was measured. Then, the signal intensity of each pixel in all 28 slices was multiplied by a factor f (100/mean signal intensity in ROI). Mean correction factor was f = 0.918, with values ranging from 0.432 to 1.656. Finally, the gray scale window was set to minimum of 0 and maximum of 255, and images were transformed to 8-bit depth.

Step 2, tissue segmentation

Pixels were segmented into 4 different groups: (1) pixels with values ranging from 65 to 140 (assigned to white matter), (2) pixels with intensities ranging from 140 to 210 (assigned to gray matter), (3) pixels with intensities ranging from 210 to 255 (assigned to CSF), and (4) pixels with intensities ranging from 0 to 65 (designated as hypointense areas) (figure 1). These thresholds were established after building up a histogram of pixel intensity values using the total set of 3,836 MRIs (28 slices for 137 participants) and fitting the corresponding plot to 4 independent gaussian functions. Once normalized, to take into account the differences in areas, the crossings between adjacent gaussian plots (65, 140, and 210) were established as threshold levels for image segmentation. Once created, segmented regions were transformed into a region mask and used for segmentation of the original 16-bit acquired images for further data analysis.

Figure 1
Figure 1 MRI analysis with tissue segmentation

Pixels with intensities ranging from 0 to 65 (designated as hypointense areas) corresponding to Fe accumulation. (A) Grayscale image; (B) Hypointensities (Fe). Pixels with gray scale value = < 65); (C) White substance. Pixels with gray scale value = 65–140; (D) Gray substance. Pixels with gray scale value = 140–210; (E) Cerebrospinal fluid. Pixels with gray scale value > 210; (F) Colored image marking all the segmented areas. Green = gray substance; Blue = white substance; Yellow = LCR; Magenta = hypointensities.

Step 3, analysis of hypointense regions

After segmentation of pixels designated as hypointense regions, those hypointense pixels located outside the BG were manually removed from each set of images. Then, the total number of hypointense pixels and the mean and SD of the signal intensities in those areas were measured for each participant. Total number of pixels was later translated in millimeters squared. Hypointense areas were studied for 2 different ROIs within BG, RN and GP, and PAG. PAG was delineated as a circular ROI of 4-mm diameter, manually located around the Silvius aqueduct.

All MRI studies were performed and analyzed by a radiologist (C.V.) and a physicist (P.R.) blinded to clinical data.

Data and statistical analysis

Statistical analysis was performed with IBM SPSS Statistics 16.0 software for Mac (SPSS Inc, Chicago, IL). Continuous normally distributed variables analyzed with the Kolmogorov-Smirnov test were reported as mean ± SD; continuous nonnormally distributed variables were expressed as median (range). Categorical variables were reported as percentages. Differences between 2 groups were assessed by independent t test (continuous normally distributed variables), Mann-Whitney test (continuous nonnormally distributed variables), and both parametric and nonparametric χ2 tests (categorical variables). We applied multivariate analysis of covariance to make comparisons between continuous variables among the 3 study groups. Parametric Pearson correlation coefficient was used to correlate iron deposits with clinical parameters of migraine and biomarkers in patients with CM. All analyses were adjusted for age.

To assess the validity of iron deposit in RN and PGA as a neuroimaging marker of CM, we calculated the best discriminant cutoff point of the mean iron volume in RN and PAG for CM using a receiver operating characteristic analysis. A significant difference was set as p < 0.05.

Data availability

Any data not published within the article are available and will be shared by request from any qualified investigator.

Results

Clinical characteristics of participants

One hundred twelve patients (55 with CM and 57 with EM) and 25 controls were included in the study. Mean age of patients with CM was 44.4 ± 11.1 years, and 89.5% were female. Mean age of patients with EM was 35.1 ± 11.7 years, and 98.2% were female. Among healthy controls, 24 were female (92.0%) and mean age was 36.1 ± 7.9 years. There was a statistically significant difference in age between patients with CM and those with EM (p = 0.000) and between patients with CM and controls (p = 0.006), while the age difference between patients with EM and controls was not significant (p = 1.000).

Clinical characteristics of migraineurs are displayed in table 1.

View this table:
Table 1

Comparisons of demographic and clinical data between migraineurs

Comparative analysis of iron volume in RN, GP, and PAG between groups

Patients with CM showed larger iron grounds volumes in RN compared to patients with EM (70.2 ± 6.8 vs 25.5 ± 7.3 mL, p < 0.001) and controls (70.2 ± 6.8 vs 15.1 ± 10.8 mL, p < 0.001) (figure 2A). Patients with CM showed larger iron grounds volumes in PAG compared to EM (360.3 ± 6.5 vs 249.7 ± 6.9 μL, p < 0.001) and controls (360.3 ± 6.5 vs 168.6 ± 10.3 μL, p < 0.001) (figure 2B). Iron grounds volumes in patients with EM were significantly larger in PAG compared to controls (249.7 ± 6.9 vs 168.6 ± 10.3 μL, p = 0.000), while differences were not relevant for RN (p = 1.000). No significant differences were obtained for the mean value of signal intensities in GP (1721.8 ± 161.7 vs 1700.0 ± 172.0 vs 1597.2 ± 254.4 μL) (figure 2C).

Figure 2
Figure 2 Levels of iron volume in RN, PAG, and GP

(A) Red nucleus (RN) (*p < 0.001 patients with chronic migraine [CM] vs patients with episodic migraine [EM], **p < 0.001 patients with CM vs controls), (B) periaqueductal gray matter (PAG) (*p < 0.001 patients with CM vs patients with EM, **p < 0.001 patients with CM vs controls, ***p < 0.001 patients with EM vs controls), and (C) globus pallidus (GP).

After calculation of receiver operating characteristic curves and area under the curve for iron deposits, the optimal threshold for CM regarding iron volume in RN in our sample was 15 μL, which would correctly identify 80% of patients with CM (sensitivity) and 71% of patients without CM (specificity). Similarly, the optimal CM threshold for iron in PAG according to this study was 240 μL, which would correctly identify 93% of patients with CM (sensitivity) and 97% of patients without CM (specificity) (figure 3).

Figure 3
Figure 3 Receiver operating characteristic curve analysis for iron volume in RN and PAG

Area under the curve (AUC) for iron volume in red nucleus (RN): 0.881 (95% confidence interval [CI] 0.800–0.962, p < 0.0001). Iron volume in RN ≥15 mL identifies patients with chronic migraine (CM) with a sensitivity of 80% and a specificity of 71%. AUC for iron volume in PAG: 0.888 (95% CI 0.887–1.000, p < 0.0001). (B) Iron volume in periaqueductal gray matter (PAG) ≥240 mL identifies patients with CM with a sensitivity of 93% and a specificity of 97%.

Correlation analysis of iron volume in RN, GP, and PAG and clinical parameters of migraine

Our results showed a positive correlation between iron volume in RN and intensity of headache (r = 0.303, p = 0.031). No correlation was observed between iron deposits in RN, GP, or PAG and other clinical variables (duration of attacks, analgesics overuse, comorbid conditions, or allodynia) in either CM or EM.

Correlation analysis of iron volume in RN, GP, and PAG and molecular markers of inflammation, endothelial dysfunction, and BBB disruption in patients with CM

Patients with CM showed higher plasma levels of IL-6 compared to patients with EM (9.1 ± 0.5 vs 5.9 ± 0.5 pg/mL, p < 0.001). Patients with CM compared to healthy controls showed higher plasma levels of CGRP (103.6 ± 10.9 vs 23.27 ± 17.1 ng/mL, p < 0.001), PTX3 (1273.4 ± 66.0 vs 442.1 ± 104.1 pg/mL, p < 0.001), cFn (13.5 ± 0.6 vs 7.9 ± 1.0 μg/mL, p < 0.001), IL-6 (9.2 ± 0.5 vs 3.6 ± 0.8 pg/mL, p < 0.001), IL-10 (1.3 ± 0.8 vs 7.0 ± 1.2 pg/mL, p < 0.001), and sTWEAK (215.6 ± 26.9 vs 16.6 ± 42.4 pg/mL, p < 0.001). Table 2 shows biomarker values in patients with CM, patients with EM, and controls.

View this table:
Table 2

Comparison of levels of molecular markers of inflammation, endothelial dysfunction, BBB disruption, neuronal damage, and TGV activation in healthy controls and patients with CM and EM (adjusted to age)

Table 3 shows correlations between iron volume in RN, GP, and PAG and plasma levels of biomarkers. Iron grounds volume in PAG showed a significant correlation with higher plasma levels of sTWEAK (r = 0.395, p = 0.005) and cFn (r = 0.294, p = 0.040) (figure 4). Iron grounds volume in RN and GP did not show a significant correlation with plasma levels of any molecular marker.

View this table:
Table 3

Correlation between iron deposits in BG and PAG and plasma levels of several molecular markers in CM (adjusted to age)

Figure 4
Figure 4 Correlation between iron volume in periaqueductal gray matter and plasma levels of sTWEAK and cFn in patients with chronic migraine

cFn = cellular fibronectin;sTWEAK = soluble tumor necrosis factor–like WEAK.

Discussion

Our results suggest that migraineurs show larger volume of iron deposits than healthy controls in RN and PAG. Patients with CM show greater iron accumulation in RN and in PAG than patients with EM, and patients with EM show greater iron deposits in PAG compared to healthy controls. Volume of iron in RN ≥15 μL can identify patients with CM with a sensitivity of 80% and a specificity of 71%; volume of iron in PAG ≥240 μL can identify patients with CM with a sensitivity of 93% and specificity of 97%. Moreover, we found a significant correlation between larger volume of iron deposits in RN and PAG and increased plasma levels of biomarkers of endothelial dysfunction and BBB disruption.

To date, several imaging studies in patients with migraine have revealed larger iron deposits in migraineurs, located in BG and other deep nuclei, including PAG. Welch et al.10 found increased iron levels in PAG and RN of a small group of patients attending a headache clinic for migraine or chronic daily headache. Those findings were hypothesized to be related to cell damage due to repeated migraine attacks but were considered cautiously because that study was performed in a selected group of patients with severe headache. Later, Kruit et al.11 confirmed these findings in the Cerebral Abnormalities in Migraine, an Epidemiological Risk Analysis (CAMERA) I study, observing a decreased T2 signal (consistent with increased iron levels), particularly in putamen, GP, and RN, in patients <50 years of age who had a migraine history of at least 23 years. Although they did not measure values in PAG and could not confirm the results of Welch et al. in this area, their analysis is controlled by age, and their results suggest a relation between longer duration of migraine history and T2 values. Moreover, Tepper et al.12 in 2011 confirmed increased iron levels in BG, particularly in GP in migraineurs. Conversely, the results obtained in the CAMERA follow-up13 did not provide data supporting the hypothesis of increased iron accumulation in deep brain nuclei in patients with migraine. In both CAMERA studies,11,13 iron grounds volume in PAG was not evaluated.

Iron deposits in GB and PGA have been associated with frequency of migraine attacks and time of evolution of migraine, suggesting a causal relation between recurring attacks and accumulation of iron.10,–,12 In our study, we found a significant correlation between iron accumulation in RN and intensity of headache in patients with CM (r = 0.303, p = 0.031) but not with frequency of attacks or duration of disease in patients with either EM or CM. However, in contrast to previous reports,10 in which PAG iron levels were equally abnormally high in patients with EM and those with chronic daily headache, we observed that patients with CM had larger volume of iron deposits in RN and PAG. Moreover, volume of iron in PAG ≥240 μL could identify patients with CM with a sensitivity of 93% and specificity of 97%, a finding that suggests, although indirectly, a role of disease severity and duration in iron deposition.

Previous studies demonstrate that PAG is activated during migraine attacks.10 These findings allow us to hypothesize that repeated migraine attacks and activation of PAG could increase free radical cell damage associated with hyperemia and may lead to iron deposition. Iron deposition could reflect progressive dysfunction of PAG8 and other brainstem9 structures related to normal antinociceptive function, contributing to migraine chronification. On the other hand, the minor incidence of migraine in older age, when iron deposits increase, argues against this hypothesis.

Several molecular and physiopathologic mechanisms may explain the high PAG iron levels in patients with migraine10: higher concentration of transferrin receptors in PAG,17 high iron content in glial cells,18 impaired iron homeostasis possibly associated with neuronal dysfunction or neuronal damage in repeatedly activated networks involved in nociception,11 and finally hyperoxia, which could cause free radical cellular damage with cumulative iron sequestration in brain tissue.10 Iron also might be stored in microglia regardless of the mechanism of cell death.18 Our study supports the idea that iron accumulation in PAG and RN could be related to migraine chronification because our findings show larger deposits in patients with CM than in those with EM. A causative role of iron in this process remains speculative, but we know on one hand that iron deposits are present in several neurodegenerative diseases19 and on the other that structural damage to certain brainstem regions can cause migraine.20

Several molecular mechanisms, including inflammation, TGV system activation, endothelial dysfunction, and BBB disruption, have been implicated in migraine pathophysiology. Among these, inflammation has been proposed to have a key role in cellular death and destruction mediated by iron accumulation.21 The repetition of neural and vascular inflammatory mechanisms during migraine attacks, in a manner similar to that observed in other neuroinflammatory diseases of the brain,22,–,24 may lead to cellular destruction and iron deposit enlargement. Observational studies have found a relationship between the risk of single-nucleotide polymorphisms in the IL-1β genetic family, the shrinkage in both gray and white matter of the brain parenchyma,25 and the increase in MRI estimates of iron.26 In our study, however, we did not find a significant correlation between levels of biomarkers of inflammation previously related to migraine such as IL-6 and TNF-α and iron deposits.

On the other hand, our results show a correlation between levels of sTWEAK and cFn and iron deposits. These findings suggest that mechanisms implicated in the pathophysiology of migraine other than inflammation such as endothelial dysfunction and BBB disruption might have a role in iron deposition.

sTWEAK levels are surrogate markers of endothelial dysfunction and atherosclerosis.27 Among the main effects induced by TWEAK interactions are inflammation and cell death or cell proliferation, depending on the particular cell type and cytokine context.28 High TWEAK levels can contribute to permeabilization of the blood-CSF barrier and BBB, which may contribute to the inflammatory cascade in the CNS and therefore to iron accumulation.28

The integrity of the BBB is known to be compromised during migraine attacks and plays a significant role in many different neurologic disorders. During inflammation of the CNS, there is an increase in membrane permeability that allows local extravasation of macromolecules in regions surrounding affected neurons.29 Brain capillary endothelial cells with intact BBB properties are able to downregulate their expression of transferrin receptors in response to increasing availability.30 Brain endothelial cells also contain relatively high amounts of ferritin and hence serve as active reservoirs of iron.31 Iron overload and iron-mediated free radical production after focal transient ischemia32 cause loss of tight junction proteins and degeneration of endothelial cells, followed by opening of the BBB. Iron is essential for normal cell function, but it also generates toxic reactive oxygen species that adversely affect vascular endothelium and the BBB.33 Vascular endothelial cells are direct targets for free hemoglobin and its oxidative derivative methemoglobin. Methemoglobin readily releases heme, an abundant source of redox-active ferric iron, which produces ferric-iron brain accumulation on the endothelium and causes brain capillary defects.32 Our results underline the importance of endothelial dysfunction and BBB disruption in migraine pathophysiology, relating volume of iron deposits to higher levels of sTWEAK and cFn. However, it is difficult to determine whether the long-term dysfunction of endothelium and BBB leads to an increase in iron deposits or whether, on the contrary, the iron deposition damages the endothelium and raises the levels of these biomarkers. It may even be possible that both molecular and imaging changes are just a normal physiologic marker of increased cellular function. A long-term prospective study is necessary to determine which event comes first.

This study has several limitations, related mainly to demographic and clinical characteristics of participants, First, the patients with CM are significantly older than patients with EM and controls. In normal aging, iron concentrations are greater in the subcortical nuclei than in cerebral white matter and cortex,34,35 especially in striatum and GP.36,–,38 However, in other deep brain nuclei such as subcortical nuclei and RN, thalamus, and the cerebellar dentate nucleus, the age dependence is more inconsistent.38,39 In our sample, age was correlated with iron accumulation in GP (r = 0.446; p = 0.001) but not with iron deposition in RN and PAG (r = 0.162, p = 0.252; and r = 0.096; p = 0.485 respectively); therefore, we have adjusted our results by age. Sex has also been related to iron deposition, and our sample is composed mainly of women (>90%), who have lower levels of peripheral iron levels.36,40 Men have higher iron concentrations according to changes in MRI contrast in the cortical white matter and subcortical nuclei,36,40 while women have lower total subcortical brain iron levels beginning in midlife compared to men and young women.40 However, the prevalence of migraine is higher in women, and our sample reflects this predominance. Regarding data collection, some clinical aspects of our sample such as frequency, duration, or intensity of attacks are based on patient recall or headache diaries and lack accuracy. Finally, molecular markers were measured during interictal periods, without taking into consideration the half-value period of some molecules such as CGRP.41 However, values should be similar between participants because they have been collected in homogeneous conditions.

The findings of this study support the hypothesis that there is an increased iron accumulation in BG nuclei, especially RN, and in PAG in patients with CM. This could be due to disruptive iron homeostasis in dysfunctional glial cells or neurons or to repeated activation and hyperemia (and hyperoxia) of nuclei associated with pain processing during migraine attacks. It is necessary to establish whether these lesions (iron deposits in RN and PAG) are chronic and irreversible (in which case they could be associated with the irreversibility of the migraine process and could be considered a biomarker of CM) or reversible.

Study funding

Funded by Spanish Ministry of Economy and Competitiveness–Instituto de Salud Carlos III, grants/awards PI12/00532, PI13/00292, PI14/01879, SAF2014-5342-R, and PI15/01578; Spanish Research Network on Cerebrovascular Diseases RETICS-INVICTUS, grant/award RD12/0014; Xunta de Galicia, grant/award Conselleria Educacion GRC2014/027; European Union program FEDER, grant/award CP14/00154 and CP12/03121; Miguel Servet Program of Instituto de Salud Carlos III; fellowship (FPI) from the Spanish Ministry of Economy and Competitiveness, grant/award BES-2012-056027; Secretaria de Estado de Investigacion; Desarrollo e Innovacion, grant/award BES-2012-056027; and Conselleria de Cultura, Educacion e Ordenacion Universitaria, Xunta de Galicia, grant/award GRC2014/027.

Disclosure

The authors report no disclosures relevant to the manuscript. Go to Neurology.org/N for full disclosures.

Appendix Authors

Table
Table

Footnotes

  • Go to Neurology.org/N for full disclosures. Funding information and disclosures deemed relevant by the authors, if any, are provided at the end of the article.

  • CME Course: NPub.org/cmelist

  • Received August 27, 2018.
  • Accepted in final form October 31, 2018.

References

  1. 1.↵
    1. May A,
    2. Goadsby PJ
    . The trigeminovascular system in humans: pathophysiologic implications for primary headache syndromes of the neural influences on the cerebral circulation. J Cereb Blood Flow Metab 1999;19:115–127.
    OpenUrlCrossRefPubMed
  2. 2.↵
    1. Akerman S,
    2. Holland PR,
    3. Goadsby PJ
    . Diencephalic and brainstem mechanisms in migraine. Nat Rev Neurosci 2011;12:570–584.
    OpenUrlCrossRefPubMed
  3. 3.↵
    1. Zhang X,
    2. Levy D,
    3. Kainz V,
    4. Noseda R,
    5. Jakubowski M,
    6. Burstein R
    . Activation of central trigeminovascular neurons by cortical spreading depression. Ann Neurol 2011;69:855–865.
    OpenUrlCrossRefPubMed
  4. 4.↵
    1. Durham P,
    2. Papapetropoulos S
    . Biomarkers associated with migraine and their potential role in migraine management. Headache 2013;53:1262–1277.
    OpenUrlPubMed
  5. 5.↵
    1. Kurth T,
    2. Gaziano JM,
    3. Cook NR,
    4. Logroscino G,
    5. Diener HC,
    6. Buring JE
    . Migraine and risk of cardiovascular disease in women. JAMA 2006;296:283–291.
    OpenUrlCrossRefPubMed
  6. 6.↵
    1. Gursoy-Ozdemir Y,
    2. Qiu J,
    3. Matsuoka N, et al
    . Cortical spreading depression activates and upregulates MMP-9. J Clin Invest 2004;113:1447–1455.
    OpenUrlCrossRefPubMed
  7. 7.↵
    1. Smith GS,
    2. Savery D,
    3. Marden C, et al
    . Distribution of messenger RNAs encoding enkephalin, substance P, somatostatin, galanin, vasoactive intestinal polypeptide, neuropeptide Y, and calcitonin gene-related peptide in the midbrain periaqueductal grey in the rat. J Comp Neurol 1994;350:23–40.
    OpenUrlCrossRefPubMed
  8. 8.↵
    1. Bandler R,
    2. Carrive P
    . Integrated defence reaction elicited by excitatory amino acid microinjection in the midbrain periaqueductal grey region of the unrestrained cat. Brain Res 1988;439:95–106.
    OpenUrlCrossRefPubMed
  9. 9.↵
    1. Jaggi AS,
    2. Singh N
    . Role of the different brain areas in peripheral nerve injury-induced neurophatic brain. Brain Res 2011;1381:187–201.
    OpenUrlCrossRefPubMed
  10. 10.↵
    1. Welch KM,
    2. Nagesh V,
    3. Aurora SK,
    4. Gelman N
    . Periaqueductal gray matter dysfunction in migraine: cause or the burden of illness? Headache 2001;41:629–637.
    OpenUrlCrossRefPubMed
  11. 11.↵
    1. Kruit MC,
    2. Launer LJ,
    3. Overbosch J,
    4. van Buchem MA,
    5. Ferrari MD
    . Iron accumulation in deep brain nuclei in migraine: a population-based magnetic resonance imaging study. Cephalalgia 2009;29:351–359.
    OpenUrlCrossRefPubMed
  12. 12.↵
    1. Tepper SJ,
    2. Lowe MJ,
    3. Beall E, et al
    . Iron deposition in pain-regulatory nuclei in episodic migraine and chronic daily headache by MRI. Headache 2012;52:236–243.
    OpenUrlCrossRefPubMed
  13. 13.↵
    1. Palm-Meinders IH,
    2. Koppen H,
    3. Terwindt GM, et al
    . Iron in deep brain nuclei in migraine? CAMERA follow-up MRI findings. Cephalalgia 2017;37:795–800.
    OpenUrl
  14. 14.↵
    Headache Classification Committee of the International Headache Society. The International Classification of Headache Disorders, 3rd edition (beta version). Cephalalgia 2013;33:629–808.
    OpenUrlCrossRefPubMed
  15. 15.↵
    1. Rasband WS
    . Image J. Bethesda: US NIH; 1997–2014. Available at: imagej.nih.gov/ij/. Accessed March 2016.
  16. 16.↵
    1. Jurgens CK,
    2. Jasinschi R,
    3. Ekin A, et al
    . MRI T2 hypointensities in basal ganglia of premanifest Huntington's disease. PLoS Curr 2010;2:RRN1173.
    OpenUrlCrossRefPubMed
  17. 17.↵
    1. Morris CM,
    2. Candy JM,
    3. Omar S,
    4. Bloxham CA,
    5. Edwardson JA
    . Transferrin receptors in the parkinsonian midbrain. Neuropathol Appl Neurobiol 1994;20:468–472.
    OpenUrlCrossRefPubMed
  18. 18.↵
    1. Benkovic SA,
    2. Connor JR
    . Ferritin, transferrin, and iron in selected regions of the adult and aged rat brain. J Comp Neurol 1993;338:97–113.
    OpenUrlCrossRefPubMed
  19. 19.↵
    1. Vyzamal J,
    2. Brokks RA,
    3. Patronas N,
    4. Hajek M,
    5. Bulte JW,
    6. Di Chiro G
    . Magnetic resonance imaging of brain iron in health and disease. J Neurol Sci 1995;134:19–26.
    OpenUrlCrossRefPubMed
  20. 20.↵
    1. Gee JR,
    2. Chang J,
    3. Dublin AB,
    4. Vijayan N
    . The association of brainstem lesions with migraine-like headache: an imaging study of multiple sclerosis. Headache 2005;45:670–677.
    OpenUrlCrossRefPubMed
  21. 21.↵
    1. Swaiman KF,
    2. Machen VL
    . Iron uptake by glial cells. Neurochem Res 1985;10:1635–1644.
    OpenUrlCrossRefPubMed
  22. 22.↵
    1. Kell DB
    . Iron behaving badly: inappropriate iron chelation as a major contributor to the aetiology of vascular and other progressive inflammatory and degenerative diseases. BMC Med Genomics 2009;2:2.
    OpenUrlCrossRefPubMed
  23. 23.↵
    1. Linker RA,
    2. Kroner A,
    3. Horn T,
    4. Mäurer M,
    5. Bendszus M
    . Iron particle-enhanced visualization of inflammatory central nervous system lesions by high resolution: preliminary data in an animal model. AJNR Am J Neuroradiol 2006;27:1225–1229.
    OpenUrlAbstract/FREE Full Text
  24. 24.↵
    1. Williams S,
    2. Rohr AM,
    3. Wang WT, et al
    . Iron deposition is independent of cellular inflammation in a cerebral model of multiple sclerosis. BMC Neurosci 2011;12:59.
    OpenUrlCrossRefPubMed
  25. 25.↵
    1. Zivadinov R,
    2. Weinstock-Gutman B,
    3. Pirko I
    . Iron deposition and inflammation in multiple sclerosis: which one comes first? BMC Neurosci 2011;12:60.
    OpenUrlCrossRefPubMed
  26. 26.↵
    1. Persson N,
    2. Ghisletta P,
    3. Dahle CL, et al
    . Regional brain shrinkage over two years: individual differences and effects of pro-inflammatory genetic polymorphisms. Neuroimage 2014;28:334–348.
    OpenUrl
  27. 27.↵
    1. Przybylowski P,
    2. Malysko JS,
    3. Malysko J
    . Soluble tumor necrosis factor-like weak inducer of apoptosis plasma levels as a novel biomarker of endothelial function in prevalent orthotopic heart transplant recipients. Transplant Proc 2011;43:1900–1903.
    OpenUrlPubMed
  28. 28.↵
    1. Campbell S,
    2. Michaelson J,
    3. Burkly L,
    4. Putterman C
    . The role of TWEAK/Fn14 in the pathogenesis of inflammation and systemic autoimmunity. Front Biosci 2004;9:2273–2284.
    OpenUrlCrossRefPubMed
  29. 29.↵
    1. Carvey PM,
    2. Zhao CH,
    3. Hendey B, et al
    . 6-Hydroxydopamine-induced alterations in blood–brain barrier permeability. Eur J Neurosci 2005;22:1158–1168.
    OpenUrlCrossRefPubMed
  30. 30.↵
    1. Moos T,
    2. Morgan EH
    . Transferrin and transferrin receptor function in brain barrier systems. Cell Mol Neurobiol 2000;20:77–95.
    OpenUrlCrossRefPubMed
  31. 31.↵
    1. Burdo JR,
    2. Simpson IA,
    3. Menzies S,
    4. Beard J,
    5. Connor JR
    . Regulation of the profile of iron-management proteins in brain microvasculature. J Cereb Blood Flow Metab 2004;24:67–74.
    OpenUrlCrossRefPubMed
  32. 32.↵
    1. Won SM,
    2. Lee JH,
    3. Park UJ,
    4. Gwag J,
    5. Gwag BJ,
    6. Lee YB
    . Iron mediates endothelial cell damage and blood-brain barrier opening in the hippocampus after transient forebrain ischemia in rats. Exp Mol Med 2011;43:121–128.
    OpenUrlCrossRefPubMed
  33. 33.↵
    1. Li DJ,
    2. Luo H,
    3. Wang LL,
    4. Zou GL
    . Potential of peroxynitrite to promote the conversion of oxyhemoglobin to methemoglobin. Acta Biochim Biophys Sin 2004;36:87–92.
    OpenUrl
  34. 34.↵
    1. Li W,
    2. Wu B,
    3. Batrachenko A, et al
    . Differential developmental trajectories of magnetic susceptibility in human brain gray and white matter over the lifespan. Hum Brain Mapp 2014;35:2698–2713.
    OpenUrlCrossRefPubMed
  35. 35.↵
    1. Haacke EM,
    2. Cheng NY,
    3. House MJ, et al
    . Imaging iron stores in the brain using magnetic resonance imaging. Magn Reson Imaging 2005;23:1–25.
    OpenUrlCrossRefPubMed
  36. 36.↵
    1. Bartzokis G,
    2. Tishler TA,
    3. Lu PH, et al
    . Brain ferritin iron may influence age- and gender-related risks of neurodegeneration. Neurobiol Aging 2007;28:414–423.
    OpenUrlCrossRefPubMed
  37. 37.↵
    1. Bilgic B,
    2. Pfefferbaum A,
    3. Rohlfing T,
    4. Sullivan EV,
    5. Adalsteinsson E
    . MRI estimates of brain iron concentration in normal aging using quantitative susceptibility mapping. Neuroimage 2012;59:2625–2635.
    OpenUrlCrossRefPubMed
  38. 38.↵
    1. Cherubini A,
    2. Peran P,
    3. Caltagirone C,
    4. Sabatini U,
    5. Spalletta G
    . Aging of subcortical nuclei: microstructural, mineralization and atrophy modifications measured in vivo using MRI. Neuroimage 2009;48:29–36.
    OpenUrlCrossRefPubMed
  39. 39.↵
    1. Pfefferbaum A,
    2. Adalsteinsson E,
    3. Rohlfing T,
    4. Sullivan EV
    . MRI estimates of brain iron concentration in normal aging: comparison of field-dependent (FDRI) and phase (SWI) methods. Neuroimage 2009;47:493–500.
    OpenUrlCrossRefPubMed
  40. 40.↵
    1. Persson N,
    2. Wu J,
    3. Zhang Q, et al
    . Age and sex related differences in subcortical brain iron concentrations among healthy adults. Neuroimage 2015;122:385–398.
    OpenUrlCrossRefPubMed
  41. 41.↵
    1. Kraenzlin ME,
    2. Ch’ng JL,
    3. Mulderry PK,
    4. Ghatei MA,
    5. Bloom SR
    . Infusion of a novel peptide, calcitonin gene-related peptide (CGRP) in man. Pharmacokinetics and effects on gastric acid secretion and on gastrointestinal hormones. Regul Pept 1985;10:189–197.
    OpenUrlCrossRefPubMed

Letters: Rapid online correspondence

Comment

REQUIREMENTS

If you are uploading a letter concerning an article:
You must have updated your disclosures within six months: http://submit.neurology.org

Your co-authors must send a completed Publishing Agreement Form to Neurology Staff (not necessary for the lead/corresponding author as the form below will suffice) before you upload your comment.

If you are responding to a comment that was written about an article you originally authored:
You (and co-authors) do not need to fill out forms or check disclosures as author forms are still valid
and apply to letter.

Submission specifications:

  • Submissions must be < 200 words with < 5 references. Reference 1 must be the article on which you are commenting.
  • Submissions should not have more than 5 authors. (Exception: original author replies can include all original authors of the article)
  • Submit only on articles published within 6 months of issue date.
  • Do not be redundant. Read any comments already posted on the article prior to submission.
  • Submitted comments are subject to editing and editor review prior to posting.

More guidelines and information on Disputes & Debates

Compose Comment

More information about text formats

Plain text

  • No HTML tags allowed.
  • Web page addresses and e-mail addresses turn into links automatically.
  • Lines and paragraphs break automatically.
Author Information
NOTE: The first author must also be the corresponding author of the comment.
First or given name, e.g. 'Peter'.
Your last, or family, name, e.g. 'MacMoody'.
Your email address, e.g. higgs-boson@gmail.com
Your role and/or occupation, e.g. 'Orthopedic Surgeon'.
Your organization or institution (if applicable), e.g. 'Royal Free Hospital'.
Publishing Agreement
NOTE: All authors, besides the first/corresponding author, must complete a separate Publishing Agreement Form and provide via email to the editorial office before comments can be posted.
CAPTCHA
This question is for testing whether or not you are a human visitor and to prevent automated spam submissions.

Vertical Tabs

You May Also be Interested in

Back to top

More Online

CME Course

Related Articles

Topics Discussed

Alert Me

Recommended articles