Torres-Ferrus et al. The Journal of Headache and Pain (2021) 22:75 https://doi.org/10.1186/s10194-021-01289-5 The Journal of Headache and Pain RESEARCH ARTICLE Open Access Cortical metabolic and structural differences in patients with chronic migraine. An exploratory 18FDG-PET and MRI study Marta Torres-Ferrus1,2, Deborah Pareto3, Victor J Gallardo2, Gemma Cuberas-Borrós4,5, Alicia Alpuente1,2, Edoardo Caronna2, Adrià Vila-Balló2, Carles Lorenzo-Bosquet5, Joan Castell-Conesa5, Alex Rovira3 and Patricia Pozo-Rosich1,2* Abstract Background: To describe interictal brain structural and metabolic differences between patients with episodic migraine (EM), chronic migraine (CM) and healthy controls (HC). Methods: This is an exploratory study including right-handed age-matched women with EM, CM and HC. On the same day, a sequential interictal scan was performed with 18FDG-PET and MRI. 3D T1-weighted images were segmented with FreeSurfer, normalized to a reference atlas and the mean values of metabolism, cortical thickness (CTh) and local gyrification index (IGI) were determined. Groups were compared using age-adjusted linear models, corrected for multiple comparisons. 18FDG-PET measurements between groups were also analysed adjusting by patient’s age, CTh and lGI. The variables independently associated with diagnosis were obtained using a logistic regression analysis. Results: Fifteen patients (8 EM, 7 CM) and 11 HC were included. Morphometric data showed an increased CTh in 6 frontal areas (L/R-Caudal Middle Frontal, L/R-Rostral Middle Frontal, L-Medial Orbitofrontal and L-Superior Frontal) in CM patients compared to HC without differences for IGI. The structural adjusted analysis in CM showed a statistically significantly hypometabolism in 9 frontal areas (L-Lateral Orbitofrontal, L/R-Medial Orbitofrontal, L-Frontal Superior, R- Frontal pole, R-Parts Triangularis, L/R-Paracentral and R-Precentral) and 7 temporal areas (L/R-Insula, L/R-Inferior temporal, L/R-Temporal pole and R-Banks superior temporal sulcus) compared to HC. EM patients presented intermediate metabolic values between EM and HC (non-significant). Conclusions: CM patients showed frontotemporal hypometabolism and increased frontal cortical thickness when compared to HC that may explain some cognitive and behavioural pain-processing and sensory integration alterations in CM patients. Combined information from sequential or simultaneous PET and MRI could optimize the study of complex functional neurological disorders such as migraine. Keywords: Chronic migraine, Neuroimaging, Positron emission tomography, Cortical thickness, Metabolism, Structural * Correspondence: ppozo@vhebron.net 1 Headache and Craniofacial Pain Unit, Neurology Department, Vall d’Hebron University Hospital, Barcelona, Spain 2 Headache and Neurological Pain Research Group, Vall d’Hebron Research Institute (VHIR), Universitat Autònoma de Barcelona, Barcelona, Spain Full list of author information is available at the end of the article © The Author(s). 2021 Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/. The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated in a credit line to the data. Torres-Ferrus et al. The Journal of Headache and Pain (2021) 22:75 Page 2 of 11 Background healthy women as control group. We excluded patients Migraine is a complex chronic neurological disorder that with neurological or psychiatric comorbidities as well as is characterized by the presence of recurrent attacks of other chronic pain syndromes, or patients with inciden- headache and other neurological symptoms. Migraine is tal relevant structural or metabolic abnormalities on im- thought to conform a disease spectrum with symptoms aging studies. Patients who were taking medications and headache frequency gradually increasing in some with central nervous effect or drugs used as migraine patients, evolving from an episodic to a chronic form preventive for other indications (including antidepressants, [1]. This transformation, is dynamic and characterized neuroleptics benzodiazepines, antiepileptic, betablockers, by several progressive neurophysiological and neuroana- antihypertensive medications and OnabotulinumtoxinA) tomical changes in pain, sensory or affective-related and patients with analgesic overuse were also excluded. brain areas [2, 3]. Inclusion period went from October 2016 to January 2018. Advanced neuroimaging could help us to better under- All participants were evaluated and diagnosed by a stand which brain structures are involved in migraine headache specialist using headache diaries. HC were also pathophysiology as well as to find potentially useful evaluated in depth, in order to rule out presence of any migraine imaging biomarkers [3]. Although there is an type of recurrent headaches or history of migraine in a increasing number of neuroimaging studies in migraine, first degree relative. We collected clinical and demo- only a few of them have specifically included chronic graphical data. All participants were screened for anxiety migraine (CM) patients. Previous MRI studies have and depression with Spanish version of State-Trait highlighted that brain regions with a key role in mi- Anxiety Inventory (STAI) [12] and Beck Depression graine attack generation, like the pons and hypothal- Inventory II (BDI-II) [13] and completed a health status amus might also be involved in migraine chronification evaluation with 36-items Short Form Health Survey (SF- [4–8]. Besides, non-controlled presence of comorbidities 36) [14] and Perceived Stress Scale [15]. Migraine partic- or medication as well as comparisons with true non- ipants also completed an evaluation of headache-related headache controls, make it difficult to draw accurate disability with Spanish version of Migraine Disability conclusions. Assessment (MIDAS) [16], Headache Impact Test 6 18 F-fluorodeoxyglucose positron emission tomog- (HIT-6) [17] and Migraine-Specific Quality of Life raphy (18 F-FDG-PET) is used to produce images of Questionnaire version 2.1 (MSQ) [18] and were asked to metabolic activity of the brain. Increased glucose uptake complete a daily headache diary for, at least, a month may be due to the presence of a hyperactive or dysfunc- before the imaging protocol. tional brain state as reduced 18 F-FDG uptake has been observed in abnormally functioning regions of brain in Neuroimaging protocol numerous disorders including dementia, stroke or Interictal brain 18 F-FDG-PET and MRI were performed interictally epilepsy [9]. Specifically in migraine, 18 F- on the same day. Patients who had a migraine attack the FDG-PET studies have identified disease or attack-specific day of the exam were rescheduled. Since CM patients alterations in brain activity within the migraine brain [10]. may have not had headache-free days, we did not ex- However, 18 F-FDG-PET imaging can only offer relatively clude patients with mild headache. Patients were asked poor anatomical information while MRI is considered to to rate headache pain (0–10 Numerical Scale) before the be a precise structural imaging modality. scan. 18 F-FDG-PET was performed with 6 h fasting and The aim of this study was to describe interictal brain blood glucose levels < 150 mg/dL, and MRI was metabolic and structural differences in a controlled and performed 6 h after the 18 F-FDG-PET to minimize radi- homogeneous population of patients with episodic ation exposure during the procedure. 18 migraine (EM), CM and healthy controls (HC). Moreover, F-FDG-PET acquisition. 150 MBq of 18 F-FDG we analyse how structural brain differences between was injected intravenously, and imaging was performed participants can influence their cerebral activity in an using a fusion PET-computed tomography (CT) scanner exploratory analysis. (Siemens Biograph mCT64S) after 30 min. The CT scan was used for attenuation correction. The scanning pa- Methods rameters may vary according to the type of CT scanner. Subjects Usually, the tube voltage is set around 120-140 kV. Brain This was a case-control study. We included right- PET acquisition started 30 min in 3-D mode. Brain PET handed adult women, 18 to 60 years of age, with EM image reconstruction utilized an iterative algorithm (4 it- and CM fulfilling International Headache Disorders erations, 12 subsets) with a 128 × 128 matrix size, Classification – 3beta [11] that were not taking any pre- 3.18mm pixel size and a post-filter Gaussian of 4 mm, ventive treatment for migraine and had never previously TOF, ultra HD-PET, decay and attenuation correction. been on one. We included a group of age-matched Prior to the administration of 18 F-FDG, a blood glucose Torres-Ferrus et al. The Journal of Headache and Pain (2021) 22:75 Page 3 of 11 determination was performed, in case of hyperglycemia Statistical significance between HC, EM and CM was (> 160 mg/dl), due to competitiveness with plasma levels assessed by Pearson’s chi-square when comparing cat- and FDG, it was postponed until the blood glucose levels egorical variables. In the case of having an expected were normal. count less than 5 in more than 20 % of cells in the MRI acquisition. MRI images were acquired in a 3.0T contingency table, Fisher’s exact test was used. Clinical magnet (Tim Trio, Siemens, Erlangen, Germany) with a differences between EM and CM were assessed by 12-channel phased-array head coil. The protocol included, independent t-test for continuous variables that among other sequences, a sagittal 3D T1-weighted followed a normal distribution (headache and analgesic MPRAGE (Magnetization Prepared – Rapid Gradient frequencies, migraine evolution, HIT-6 and MSQ) and, Echo) (TR = 2300ms, TE = 2.98ms, FOV = 256 × 256mm, Mann-Whitney U test was used for the rest of continuous 192 sections, voxel size = 1 × 1 × 1mm) and a sagittal 3D variables (MIDAS). Pre-planned one-way ANOVA with T2-weighted FLAIR (Fluid attenuation inversion recovery) Bonferroni correction in the post-hoc analysis was per- (TR = 9000ms, TE = 93ms, TI = 2500ms, flip angle 120º, formed in order to find clinical differences between HC, FOV = 256 × 200mm, voxel size = 1 × 1 × 1mm). EM and CM in the rest of continuous variables (Age, SF- Image analysis. 3D T1-weighted images were seg- 36, STAI, BDI-II and PSS). mented with FreeSurfer (version 6.2) and the corre- Metabolic (18 F-FDG-PET) and structural (CTh and sponding cortical thickness (CTh) and local gyrification lGI) values were analyzed using a pre-planned one-way index (lGI) of the parcellated regions was determined. ANCOVA, adjusted for the effect of age, for the study Briefly, white matter points are chosen based on their lo- groups (EM, CM and HC). Following significant group cations in Talairach space as well as on their intensity main effects, post-hoc analysis was conducted using the and the local neighborhood intensities. Voxels were then Bonferroni correction in order to find differences in the classified as white matter or something other than white neuroimaging variables (18 F-FDG-PET, CTh and lGI) matter based on intensity and neighbor constraints. between EM, CM and HC. Then, 18 F-FDG-PET mea- Cutting planes are chosen to separate the hemispheres surements between groups were also analyzed adjusting from each other. An initial surface was then generated by patient’s age and their CTh and lGI values in order to for each hemisphere by tiling the outside of the white study if metabolic activity was influenced by brain struc- matter mass for that hemisphere. This initial surface was ture. Finally, the degree of association between neuroim- then refined to follow the intensity gradients between aging variables and clinical characteristics was computed the white and grey matter (this is referred to as the by Pearson’s correlation coefficient and summarized by white surface). The white surface was then nudged to Pearson’s rho and related p-values. follow the intensity gradients between the gray matter A statistical power calculation was not previously con- and cerebrospinal fluid (this is the pial surface). The dis- ducted because this is an exploratory analysis and sam- tance between the white and the pia gives us the thick- ple size was based on the available data. No missing data ness at each location of cortex [19, 20]. Once the PET was obtained and p-values presented are for a two-tailed images have been reconstructed, they were co-registered test and p-values < 0.05 were considered statistically to the 3D T1-weighted image and the FreeSurfer parcel- significant. lation was applied in order to obtain the mean value in each region. Brain PET data was normalized by the cere- Results bellum white matter. Demographical and clinical data We included 17 right-handed women with migraine and Statistical analysis 11 HC. After performing the neuroimaging protocol, we Descriptive and frequency statistics were obtained and had to reject data from 3 subjects, as one showed abnor- comparisons made using the SPSS, version 21.0 for Win- mal 18 F-FDG-PET uptake (suggestive of another dows. Nominal variables (aura, pain side and allodynia) pathology), and 2 showed orthodontic appliance-derived were reported as frequencies (percentages) while mean ± artifacts that severely distorted the structural brain MRI. standard deviation (age, headache frequency, migraine So, we finally included in the data analysis 15 migraine evolution, SF-36, HIT-6, MSQ, STAI, BDI-II, PSS and patients (8 EM and 7 CM) and 10 HC. The day of the neuroimaging values) or median and interquartile range scan, 3 CM patients reported mild headache (visual nu- (IQR) (chronification, analgesic frequency and MIDAS) merical scale 1–3). were reported for continuous variables. Normality as- Demographical and clinical characteristics of the sumption of quantitative variables was checked through sample are summarized in Tables 1 and 2. We did not visual methods (Q-Q plots) and normality tests (Sha- find statistically significant differences regarding age, piro-Wilk and Kolmogorov-Smirnov tests). The Levene presence of aura, pain localization or migraine evolu- test was used to test variance homogeneity. tion. Mean headache frequency was 8.8 ± 1.8 days/ Torres-Ferrus et al. The Journal of Headache and Pain (2021) 22:75 Page 4 of 11 Table 1 Demographic and clinical characteristics of patients and controls HC EM CM HC-EM HC-CM EM-CM N = 10 N=8 N=7 Adj.Pvalue Adj.Pvalue Adj.Pvalue Agea, mean (SD), years 38.0 (14.3) 43.1 (12.2) 35.0 (17.0) 1.000 1.000 0.867 a STAI , mean (SD) State 36.2 (7.8) 36.6 (5.1) 47.3 (12.8) 1.000 0.066 0.083 Trait 33.9 (6.5) 39.4 (6.4) 44.6 (12.1) 0.515 0.053 0.722 BDI-IIa, mean (SD) 2.2 (2.2) 10.1 (8.6) 15.7 (11.1) 0.099 0.003 0.495 SF-36a, mean (SD) Physical functioning 81.7 (25.9) 88.8 (9.5) 74.3 (16.7) 1.000 1.000 0.514 Role physical 90.2 (38.9) 56.3 (54.7) 32.1 (23.8) 0.274 0.024 0.818 Role emotional 96.7 (10.0) 91.7 (15.4) 61.9 (40.5) 1.000 0.016 0.064 Vitality 68.5 (18.2) 55.6 (19.4) 51.4 (12.8) 0.371 0.160 1.000 General mental health 52.8 (31.2) 41.5 (19.4) 61.7 (16.8) 1.000 1.000 0.383 Social functioning 79.8 (21.0) 64.1 (26.3) 73.2 (19.7) 0.433 1.000 1.000 Bodily pain 80.3 (23.4) 56.3 (20.1) 46.4 (17.3) 0.064 0.008 1.000 General health 65.0 (21.0) 52.5 (16.5) 50.7 (24.2) 0.618 0.499 1.000 Health change 52.5 (17.5) 34.4 (18.6) 57.1 (12.2) 0.085 1.000 0.044 PSSa, mean (SD) 14.5 (4.7) 20.1 (10.0) 22.0 (14.2) 0.675 0.369 1.000 Bold font in the P values column indicates statistically significant a Significance assessed with one-way ANOVA with Bonferroni correction adjustment for multiple comparisons HC healthy controls; EM episodic migraine; CM chronic migraine; SD standard deviation; STAI state-trait anxiety inventory; BDI-II Beck depression inventory II; SF-36 36-Item short form survey; PSS Perceived stress scale month for the EM group and 24.7 ± 6.3 days/month Image findings for CM. CM patients showed higher scores in Brain metabolic activity (18 F-FDG-PET findings) headache-disability MIDAS scale (EM: 9.5 [29.0] vs. CM patients showed temporal hypometabolic areas CM: 32.0 [89.0], p = 0.039) although results from head- compared with HC. Figure 1; Table 3 and Supplemen- ache impact (HIT6) and quality of life (MSQ) were tary Table 1 show statistically significant group differ- comparable between patients. We did no find differ- ences on 18 F-FDG-PET values, adjusted for patient’s ences regarding, stress, anxiety or depression scores age. CM patients showed a lower glucose metabolic ac- between EM and CM patients. Statistically significant tivity compared to HC for a cluster of the left temporal differences were found in BDI-II scores and SF-36 lobe, which includes the left fusiform, left middle tem- physical role, emotional role, bodily pain and health poral, left inferior temporal, left temporal pole, bilateral change domains between HC and CM. insula and superior temporal sulcus. Frontal regions that Table 2 Migraine characteristics EM CM P value N=8 N=7 Headache frequencyb, mean (SD), days/month 8.8 (1.8) 24.7 (6.3) < 0.0001 Disease durationb, mean (SD), years 28.4 (13.2) 20.0 (19.5) 0.343 Chronification, median (IQR), years -- 4.0 (11.0) -- Auraa, n (%) 3 (37.5) 4 (57.1) 0.619 Unilateral Paina, n (%) 5 (62.5) 7 (100.0) 0.200 Analgesic Usec, median (IQR), days/month 0.5 (8.8) 5.0 (8.0) 0.613 c MIDAS , median (IQR) 9.5 (29.0) 32.0 (89.0) 0.029 HIT-6b, mean (SD) 60.5 (3.9) 61.3 (4.9) 0.735 MSQb, mean (SD) Role Function preventive 53.1 (23.8) 51.8 (15.8) 0.800 Role function restrictive 47.9 (17.1) 50.0 (13.3) 0.901 Emotional function 51.4 (27.2) 59.5 (16.3) 0.503 Bold font in the P values column indicates statistically significant a Significance assessed with Fisher’s exact test b Significance assessed with independent t-test c Significance assessed with Mann-Whitney U test HC healthy controls; EM episodic Migraine; CM chronic migraine; SD standard deviation; IQR interquartile range; y years; d/mo days per month; MIDAS migraine disability assessment; HIT6 headache impact test 6; MSQ migraine-specific quality-of-life Torres-Ferrus et al. The Journal of Headache and Pain (2021) 22:75 Page 5 of 11 Fig. 1 Statistically significant brain metabolic activity differences in 18FDG-PET (green), 18FDG-PET adjusted by brain structure (blue) or both analysis models (red) between healthy controls and chronic migraine. A Right hemisphere lateral view, (B) Left hemisphere lateral view, (C) Right hemisphere medial view, (D) Left hemisphere medial view showed differences between CM and HC were the left frontal and right superior frontal showed a statistically paracentral and the left medial orbitofrontal areas. significant thinning in EM compared to CM. Regarding the occipital lobe, only the left lateral occipital region showed an hypometabolism in CM versus HC. EM patients presented higher metabolic activity than CM but Local Gyrification Indexes (lGI) lower than HC for these regions (See Supplementary No statistically significant differences (See Supplementary Table 1). However, only the metabolic activity for the bi- Table 3) were found in the local gyrification ratings (ad- lateral temporal pole in EM showed statistically significant justed by patient’s age) between study groups (HC vs. EM differences compared to HC. We did not find any differ- vs. CM). ences in cerebral glucose metabolism when we compared both migraine subgroups together (CM and EM) with HC. Cerebral metabolic activity adjusted by brain structure We studied the influence of brain structure in the meta- Cortical Thickness (CTh) bolic activity. Hence, we analyzed again brain metabolic CTh age-adjusted analysis showed frontal CTh increase activity adjusted by age, CTh and lGI (See Table 3). In in CM compared with HC (See Table 4 and Supplemen- comparison with HC, we found that CM presented and tary Table 2): bilateral caudal middle frontal, left medial hypometabolism in several frontal regions: left lateral orbitofrontal, bilateral rostral middle frontal and left su- orbitofrontal, bilateral medial orbitofrontal, bilateral perior frontal. We also found a statistically significant paracentral, left superior frontal and right frontal pole. higher CTh in EM respect to HC in the right parahippo- For the temporal lobe, statistically significant differences campal region and right inferior parietal region. EM pa- were also found in right superior temporal sulcus, tients presented intermediate CTh values between CM bilateral inferior temporal, bilateral temporal pole and and HC for the previous areas; only CTh values for right bilateral insula between HC and CM. The right lateral pars opercularis, right transverse temporal and right occipital presented also statistically significant lower temporal pole were statistically significant thinner in EM metabolism in CM. Regarding to EM, we found a statis- in comparison to HC. Finally, between EM and CM, we tically significant hypometabolism in comparison to HC found that left rostral middle frontal, left caudal middle in the bilateral temporal pole. Table 3 Statistically significant brain metabolic activity (FDG-PET) differences adjusted by brain structure between healthy controls and migraine subgroups (CM and EM) Anatomical Areas, HC EM CM Adj. P Value HC EM CM Adj. P Value mean (SD) (N = 10) (N = 8) (N = 7) (N = 10) (N = 8) (N = 7) HC-EM HC-CM EM-CM HC-EM HC-CM EM-CM LEFT side RIGHT side Frontal Areas Lateral 1.605 (0.029) 1.552 (0.036) 1.503 (0.036) 0.296 0.038 0.366 1.598 (0.033) 1.527 (0.040) 1.492 (0.041) 0.201 0.057 0.562 Orbitofrontal Torres-Ferrus et al. The Journal of Headache and Pain Medial Orbitofrontal 1.564 (0.034) 1.522 (0.043) 1.413 (0.047) 0.453 0.018 0.132 1.590 (0.033) 1.493 (0.038) 1.469 (0.041) 0.067 0.033 0.686 Paracentral 1.443 (0.033) 1.438 (0.040) 1.309 (0.042) 0.919 0.019 0.042 1.473 (0.034) 1.438 (0.038) 1.345 (0.041) 0.503 0.030 0.116 Pars Triangularis 1.724 (0.033) 1.689 (0.040) 1.626 (0.043) 0.506 0.085 0.313 1.749 (0.036) 1.644 (0.041) 1.625 (0.044) 0.074 0.043 0.763 Precentral 1.559 (0.033) 1.538 (0.040) 1.462 (0.043) 0.703 0.095 0.217 1.560 (0.030) 1.550 (0.035) 1.446 (0.037) 0.821 0.029 0.070 (2021) 22:75 Superior Frontal 1.623 (0.037) 1.591 (0.043) 1.490 (0.050) 0.572 0.050 0.187 1.607 (0.034) 1.591 (0.040) 1.504 (0.046) 0.766 0.090 0.201 Frontal Pole 1.702 (0.040) 1.645 (0.049) 1.602 (0.051) 0.385 0.135 0.551 1.778 (0.042) 1.696 (0.047) 1.632 (0.050) 0.220 0.039 0.354 Temporal areas Banks Superior 1.548 (0.034) 1.470 (0.040) 1.444 (0.044) 0.152 0.083 0.673 1.584 (0.032) 1.507 (0.035) 1.459 (0.038) 0.127 0.026 0.372 Temporal Sulcus Inferior Temporal 1.449 (0.026) 1.392 (0.031) 1.354 (0.034) 0.183 0.039 0.426 1.447 (0.023) 1.382 (0.026) 1.362 (0.029) 0.076 0.035 0.631 Temporal Pole 1.137 (0.021) 1.045 (0.024) 1.038 (0.026) 0.011 0.008 0.840 1.098 (0.021) 0.994 (0.025) 1.005 (0.025) 0.007 0.031 0.357 Insula 1.373 (0.027) 1.311 (0.032) 1.281 (0.035) 0.162 0.050 0.541 1.388 (0.028) 1.309 (0.032) 1.287 (0.033) 0.080 0.031 0.649 Occipital Areas Lateral Occipital 1.550 (0.036) 1.536 (0.042) 1.436 (0.045) 0.815 0.069 0.122 1.630 (0.044) 1.582 (0.048) 1.476 (0.051) 0.489 0.036 0.143 Bold font in the P values column indicates statistically significant Significance assessed with ANCOVA modelling adjusted by patients’ age and brain structural variables (cortical thickness and gyrification index) in order to find metabolic activity differences between healthy controls and migraine patients (EM and CM) HC healthy controls; EM episodic Migraine; CM chronic migraine SD standard deviation; Adj. P value adjusted P value (Bonferroni corrected) Page 6 of 11 Table 4 Statistically significant cortical thickness differences between healthy controls and migraine subgroups (CM and EM) Anatomical Areas, HC EM CM Adj. P Value HC EM CM Adj. P Value mean (SD) (N = 10) (N = 8) (N = 7) (N = 10) (N = 8) (N = 7) HC-EM HC-CM EM-CM HC-EM HC-CM EM-CM LEFT side RIGHT side Frontal Areas Torres-Ferrus et al. The Journal of Headache and Pain Caudal Middle Frontal 2.480 (0.054) 2.481 (0.063) 2.608 (0.069) 0.986 0.028 0.043 2.414 (0.067) 2.473 (0.080) 2.578 (0.080) 0.389 0.040 0.138 Medial Orbitofrontal 2.317 (0.067) 2.378 (0.057) 2.472 (0.072) 0.377 0.025 0.093 2.330 (0.049) 2.305 (0.059) 2.367 (0.062) 0.748 0.642 0.480 Pars Opercularis 2.540 (0.040) 2.596 (0.048) 2.609 (0.041) 0.386 0.302 0.857 2.638 (0.039) 2.469 (0.047) 2.604 (0.050) 0.011 0.587 0.064 Rostral Middle 2.305 (0.043) 2.287 (0.050) 2.415 (0.054) 0.688 0.022 0.016 2.186 (0.044) 2.217 (0.054) 2.316 (0.058) 0.508 0.022 0.052 (2021) 22:75 Frontal Superior Frontal 2.673 (0.052) 2.684 (0.062) 2.793 (0.067) 0.836 0.028 0.062 2.630 (0.031) 2.577 (0.037) 2.707 (0.039) 0.286 0.136 0.026 Temporal areas Transverse Temporal 2.451 (0.067) 2.291 (0.080) 2.469 (0.085) 0.141 0.868 0.146 2.506 (0.058) 2.318 (0.069) 2.435 (0.073) 0.049 0.460 0.260 Parahippocampal 2.892 (0.087) 2.776 (0.104) 2.693 (0.110) 0.405 0.172 0.595 2.944 (0.100) 2.777 (0.115) 2.705 (0.122) 0.080 0.018 0.488 Temporal Pole 3.717 (0.082) 3.620 (0.098) 3.720 (0.104) 0.459 0.983 0.498 4.021 (0.086) 3.677 (0.103) 3.997 (0.100) 0.018 0.863 0.057 Parietal Areas Inferior Parietal 2.414 (0.034) 2.441 (0.040) 2.478 (0.043) 0.625 0.252 0.530 2.452 (0.050) 2.465 (0.060) 2.559 (0.052) 0.787 0.048 0.108 Bold font in the P values column indicates statistically significant Significance assessed with ANCOVA modelling adjusted by patients’ age in order to find cortical thickness differences between healthy controls and migraine patients (EM and CM) HC healthy controls; EM episodic migraine; CM chronic migraine; SD standard deviation; Adj. P value adjusted P value (Bonferroni correction) Page 7 of 11 Torres-Ferrus et al. The Journal of Headache and Pain (2021) 22:75 Page 8 of 11 Clinical correlation with imaging measurements hypometabolism has been found in CM with medication In order to further investigate the association between overuse patients after withdrawal [22], as well as, with metabolism and structural measurements with clinical other substance addictions [23] so, it has been suggested variables, we performed the following correlation that this structure may be linked to the dependence analysis. Regarding metabolism activity, we observed that drive [24, 25]. However, this finding has been more re- a longer migraine duration was correlated with hypome- cently replicated also in a group of non-medication- tabolism of the left lateral orbitofrontal (L: r= -0.572, overuse migraineurs compared to HC [26], suggesting p = 0.026), bilateral medial orbitofrontal (L: r= -0.642, that this finding is genuinely associated with migraine p = 0.010 and R: r= -0.616, p = 0.014), left inferior tem- rather than medication overuse. Our group of partici- poral (L: r= -0.518, p = 0.048), bilateral temporal pole (L: pants was thoroughly screened for current or past his- r= -0.558, p = 0.031 and R: r= -0.516, p = 0.049), bilateral tory of medication overuse which makes us think that insula (L: r= -0.656, p = 0.008 and R: r= -0.716, p = this finding would be linked to affective aspects or mi- 0.003) and bilateral lateral occipital (L: r= -0.414, graine but not exclusively with medication use. Apart p = 0.036 and R: r= -0.391, p = 0.048). We also found from prefrontal areas, involved in executive and behav- a negative correlation between headache frequency iour control functions, we have also found hypometabo- and brain metabolism in the left paracentral region lism of paracentral and precentral structures, that are (L: r= -0.417, p = 0.034). The left banks of superior likely to mediate part of the cognitive dimension of temporal sulcus (L: r=-0.638, p = 0.011), left middle pain-processing associated with localization and ‘motor’ temporal (L: r=-0.570, p = 0.027) and left inferior response to pain [21]. temporal (L: r=-0.589, p = 0.021) metabolism were Besides frontal hypometabolism, our results suggest negatively correlated to analgesic intake. In regards that CM patients have a hypometabolism in the bilateral to CTh, a positive correlation was found with disease inferior temporal lobe, temporal pole and bilateral insula duration and left medial orbitofrontal (L: r = 0.405, p = compared to HC. Previous studies also found a temporal 0.040), left rostral middle frontal (L: r = 0.424, p = 0.031), dysfunction in patients with EM during the interictal left superior frontal (L: r = 0.443, p = 0.023) and right period. Kim et al. [27] found that migraine patients had inferior parietal (R: r = 0.443, p = 0.024). A statistically significant hypometabolism in the bilateral insula and significant negative correlation was also found between superior temporal cortex. Although this study does not medial orbitofrontal area and headache frequency (L: r= include CM patients specifically, they found a significant -0.554, p = 0.036). We did not find any statistically signifi- negative correlation with estimated lifetime headache cant correlation with lGI and clinical variables. frequency and right insula metabolism, which could in- directly point to chronification. Insular hypometabolism Discussion was also found in a group of patients with CM and This is the first study which explores metabolic and medication overuse. However, metabolism of this area structural differences in migraine patients combining normalized after analgesic withdrawal showing no statis- PET and MRI techniques. Our results suggest that CM tical differences with HC [22]. Similar to our study, MRI patients show cortical frontotemporal structural and studies in EM and CM patients have also demonstrated metabolic differences when compared to HC. It is im- enhanced stimulus-induced activation and structural dif- portant to point out that we have included structural ferences of the temporal pole [25–27], a region that inte- data to adjust cortical metabolism analysis, and we have grates auditory, olfactory, visual, and painful stimuli. found that the structural adjusted 18 F-FDG-PET values There are few brain PET-imaging reports on chronic may increase the sensitivity of the model, showing more nociceptive pain other than headache disorders. A num- areas with statistical differences than the non-adjusted ber of studies have investigated the cerebral response model. pattern to painful, usually allodynic, stimulation in Our data show that CM patients have a decreased me- chronic neuropathic pain models showing metabolic tabolism in prefrontal areas including frontal pole, su- changes in multiple brain areas including orbitofrontal perior and inferior frontal cortex as well as orbitofrontal cortex, somatosensory cortex, anterior cingulate cortex, areas. Prefrontal structures have been related to cogni- anterior insula, thalamus brainstem, and cerebellum [28] tive, affective and sensory processing functions. Previous which have been identified as pain processing areas. imaging studies have demonstrated that the prefrontal Cortical metabolism values in EM patients were in cortex, connected to limbic regions, can regulate the between HC and CM for all areas, except for right perception and behavioral expression of pain in humans temporal pole. EM showed statistically significant bi- [21]. So, frontal-limbic dysfunction could be involved in lateral temporal pole hypometabolism compared to chronic pain processing, as well as, in chronic migraine. HC and an increased metabolism for the left paracen- On the other hand, persistent orbitofrontal cortex tral compared to CM. All other comparations did not Torres-Ferrus et al. The Journal of Headache and Pain (2021) 22:75 Page 9 of 11 reach statistical significance. This might point to a re- Our study has several strengths. This is the first study lationship between migraine chronification and de- that explores metabolic and structural differences in mi- creased cortical metabolism. graine patients combining PET and MRI techniques, that In relation to structural data, to date, the number of were done on the same day, avoiding any possible biases morphometric studies that include CM patients has been in relation to natural migraine fluctuations. 18 F-FDG scarce and has not been able to establish which areas uptake values have been adjusted by structural measure- could be involved in the chronification process. Only ments in each corresponding region, providing a more two previously published studies have explored differ- accurate measure of cortical metabolic activity. Secondly, ences in CTh of CM patients compared to HC. One the selection of participants was strictly and meticu- showed significantly thinner cortices in the bilateral in- lously performed and HC were also recruited specifically sular cortex, caudal middle frontal gyrus, precentral for this study and were also screened for headache fam- gyrus, and parietal lobes in patients with CM [28], while ily history or recurrent headaches. These strict inclusion the other has not found significant differences between and exclusion criteria increase the internal validity of CM patients and controls [29]. Due to the inconsistency our findings and allow us to attribute the differences of previous morphometric data on CM and the small found to CM and not to other confounders. The main number of patients included in our study, we need to be limitation is the low number of patients included that cautious when interpreting our structural data. forces us to be cautious in data interpretation. Finally, In regards to the correlation between clinical character- although patients where scanned outside of a migraine istics and imaging measurements, we found that temporal attack, three patients had mild pain and, we cannot areas metabolism, including insula and orbitofrontal exclude interferences with mild pain, premonitory or cortex, was correlated with migraine disease duration. We postdromal changes in brain metabolism. In spite of the also found negative correlations between metabolism limitations, we believe that our findings are a reliable values in temporal areas and analgesic use as well as me- basis to develop future projects and increase our under- tabolism in left paracentral area and headache frequency. standing on migraine pathophysiology. Previous PET studies have not explored clinical correla- tions. MRI studies performed in CM patients have found Conclusions some correlations between migraine duration, headache CM is correlated with brain metabolic and structural frequency as well as other clinical variables and temporal differences. CM patients showed frontotemporal hypome- sulcus, insula CTh [29, 30] or cerebellum grey matter tabolism and increased frontal cortical thickness when volumes [31]. compared to HC. EM patients presented higher metabolic Finally, a point that needs further discussion is the activity than CM but lower than HC for these regions joint analysis of structural and metabolic data. Since it which follows a headache frequency-related spectrum of has been reported that migraine patients show regional change. Metabolic data analysis using precise structural cortical volume and thickness changes [32], we can ex- anatomical information allowed us to obtain more accur- pect that the structural changes may cause differences in ate models. So, combined information from sequential or PET intensity due to anatomy rather than activity itself simultaneous PET and MRI could optimize the study of or viceversa, blurring the differences between groups. To functional neurological disorders such as migraine. our knowledge, no surface-based analyses using precise structural anatomical information have been applied to the statistical analysis of PET images in migraine Abbreviations MRI : Magnetic Resonance Imaging; CM : Chronic Migraine; 18F-FDG : 18F- research. Our 18FDG-PET data has been analysed adjust- Fluorodeoxyglucose Positron Emission Tomography; BDI-II : Beck Depression ing by patient’s age and their CTh and lGI values in Inventory II; CT : Computed Tomography; CTh : Cortical Thickness; EM order to minimize the influence of cortical structure on : Episodic Migraine; HC : Healthy Controls; HIT-6 : Headache Impact Test 6; IGI : Local Gyrification Index; IQR : Interquartile Range; MIDAS : Migraine metabolic activity. In the frontal lobe, where we have Disability Assessment; MSQ : Migraine-Specific Quality of Life Questionnaire; found remarkable structural differences between CM PSS : Perceived Stress Scale; SF-36 : 36-Items Short Form Health Survey; STAI and HC, a structural-adjusted approach has increased : State-Trait Anxiety Inventory the number of frontal areas that showed differences from 2 to 9. In contrast, this method confirms only 5 Supplementary information The online version contains supplementary material available at https://doi. over 9 temporal areas, specifically the right temporal org/10.1186/s10194-021-01289-5. pole and inferior temporal. As scientific data available up to now suggest that structural changes may underlie Additional file 1. migraine pathophysiology, we believe that structural data should be always taken into consideration when meta- Acknowledgements bolic imaging studies are performed. Not applicable. Torres-Ferrus et al. The Journal of Headache and Pain (2021) 22:75 Page 10 of 11 Authors’ contributions 2. Torres-Ferrús M, Ursitti F, Alpuente-Ruiz A, Brunello F, Chiappino D et al All authors made substantial contributions to conception and study design and (2020) From transformation to chronification of migraine: data collection. Pareto D, Cubera-Borrós G and Gallardo VJ contributed to pathophysiological and clinical aspects. J Headache Pain J Headache Pain analysis and interpretation of data. Torres-Ferrus M wrote first draft. Alpuente A, 21(1):42. Caronna C, Vila-Balló A, Lorenzo-Bosquet C, Castell J, Rovira A, and Pozo-Rosich 3. Pozo-Rosich P, Coppola G, Pascual J, Schwedt TJ. How does the brain R critically revised for important intellectual content. All authors finally approved change in chronic migraine? Developing disease biomarkers. Cephalalgia. the version to be published and agree to be accountable for all aspects of the 2020;333102420974359 work in ensuring that questions related to the accuracy or integrity of any part 4. Neeb L, Bastian K, Villringer K, Israel H, Reuter U, Fiebach JB. Structural Gray of the work are appropriately investigated and resolved. Matter Alterations in Chronic Migraine: Implications for a Progressive Disease? Headache. 2017;57(3):400–16 Funding 5. Chen Z, Chen X, Liu M, Liu S, Ma L, Yu S (2017) Volume expansion of The project leading to these results has received funding The Headache periaqueductal gray in episodic migraine: a pilot MRI structural imaging Study Group Grant 2016 of Spanish Neurological Society and Mutual Medica study. J Headache Pain 18(1):83 Grant 2016 and “La Caixa” Foundation under the project code LCF/PR/PR16/ 6. Liu H-Y, Chou K-H, Lee P-L, Fuh J-L, Niddam DM, Lai K-L et al (2017) 151110005. Hippocampus and amygdala volume in relation to migraine frequency and prognosis. Cephalalgia 37(14):1329–1336 Availability of data and materials 7. Dominguez C, Lopez A, Ramos-Cabrer P, Vieites-Prado A, Perez-Mato M, The datasets used and/or analysed during the current study are available Villalba C et al (2019) Iron deposition in periaqueductal gray matter as from the corresponding author on reasonable request. a potential biomarker for chronic migraine. Neurology 92(10):e1076– e1085 Declarations 8. Schulte LH, Allers A, May A (2017) Hypothalamus as a mediator of chronic migraine: Evidence from high-resolution fMRI. Neurology 88(21):2011–2016 Ethics approval and consent to participate 9. Schöll M, Damián A, Engler H (2014) Fluorodeoxyglucose PET in Neurology The study was approved by Vall d’Hebron Ethics Committee (PR(IR)40/2016). and Psychiatry. PET Clin 9(4):371–390 Subjects’ written informed consent was obtained according to the 10. Schulte LH, May A (2016) Functional Neuroimaging in Migraine: Chances Declaration of Helsinki. and Challenges. Headache 56(9):1474–1481 11. Headache Classification Committee of the International Headache Society Consent for publication (IHS). The International Classification of Headache Disorders, 3rd edition Subjects gave consent for the publication of data. (beta version). Cephalalgia. 2013;33(9):629–808 12. Spielberger CD, Gorsuch RL, Lushene R, Vagg PR, Jacobs GA (1983) Competing interests Manual for the State-Trait Anxiety Inventory. Consulting Psychologists Torres-Ferrus M has received honoraria as a consultant or speaker for Press, Palo Alto Allergan-Abbvie, Almirall, Chiesi, Eli Lilly, Novartis and Teva. Pareto D has re- 13. Beck AT, Steer RA, Brown GK (1996) Manual for the Beck Depression ceived speaking honoraria from Novartis and Sanofi-Genzyme. Alpuente A Inventory-II. Psychological Corporation, San Antonio has received honoraria as a consultant or speaker for Allergan-Abbvie, Chiesi 14. Cohen S, Kamarck T, Mermelstein R (1983) A Global Measure of Perceived and Novartis. Rovira A serves on scientific advisory boards for Novartis, Stress. J Health Soc Behav 24(4):385–396 Sanofi-Genzyme, SyntheticMR, Bayer, Roche, Biogen, and OLEA Medical, and 15. Ware JEJ, Sherbourne CD (1992) The MOS 36-item short-form health has received speaker honoraria from Sanofi-Genzyme, Bracco, Merck-Serono, survey (SF-36). I. Conceptual framework and item selection. Med Care Teva Pharmaceutical Industries Ltd, Novartis, Roche and Biogen. Lorenzo- 30(6):473–483 Bosquet C has received honoraria as speaker for GE Healthcare. Pozo-Rosich 16. Stewart WF, Lipton RB, Dowson AJ, Sawyer J (2001) Development and P has received honoraria as a consultant and speaker for: Allergan-Abbvie, testing of the Migraine Disability Assessment (MIDAS) Questionnaire Almirall, Amgen, Biohaven, Chiesi, Eli Lilly, Medscape, Neurodiem, Novartis to assess headache-related disability. Neurology. Mar 1;56(suppl 1):S20 and Teva. Her research group has received research grants from Allergan, LP-S28 and has received funding for clinical trials from Alder, Allergan-Abbvie, Elec- 17. Kosinski M, Bayliss MS, Bjorner JB, Ware JEJ, Garber WH, Batenhorst A et al trocore, Eli Lilly, Novartis and Teva. She is a trustee member of the board of (2003) A six-item short-form survey for measuring headache impact: the the International Headache Society, member of the Council of the European HIT-6. Qual Life Res 12(8):963–974 Headache Society. She is in the editorial board of Revista de Neurologia, as- 18. Martin BC, Pathak DS, Sharfman MI, Adelman JU, Taylor F, Kwong WJ et al sociate editor for Cephalalgia, Headache, Neurologia, Frontiers of Neurology (2000) Validity and reliability of the migraine-specific quality of life and The Journal of Headache and Pain. She is a member of the Clinical Trials questionnaire (MSQ Version 2.1). Headache 40(3):204–215 Guidelines Committee of the International Headache Society. She has edited 19. Dale AM, Fischl B, Sereno MI (1999) Cortical surface-based analysis. I. the Guidelines for the Diagnosis and Treatment of Headache of the Spanish Segmentation and surface reconstruction. Neuroimage 9(2):179–194 Neurological Society. She is the founder of www.midolordecabeza.org. Pozo- 20. Fischl B, Dale AM (2000) Measuring the thickness of the human cerebral Rosich P does not own stocks from any pharmaceutical company. Cubera- cortex from magnetic resonance images. Proc Natl Acad Sci U S A 97(20): Borrós G, Gallardo VJ and Vilà-Balló A report no disclosures. 11050–11055 21. Lee MC, Tracey I, Colvin L, Rowbotham DJ (2013) Imaging pain: A Author details potent means for investigating pain mechanisms in patients. Br J 1 Headache and Craniofacial Pain Unit, Neurology Department, Vall d’Hebron Anaesth 111(1):64–72 University Hospital, Barcelona, Spain. 2Headache and Neurological Pain 22. Fumal A, Laureys S, Di Clemente L, Boly M, Bohotin V, Vandenheede M et al Research Group, Vall d’Hebron Research Institute (VHIR), Universitat (2006) Orbitofrontal cortex involvement in chronic analgesic-overuse Autònoma de Barcelona, Barcelona, Spain. 3Section of Neuroradiology, headache evolving from episodic migraine. Brain 129(2):543–550 Department of Radiology, Vall d’Hebron University Hospital and Research 23. Volkow ND, Fowler JS (2000) Addiction, a disease of compulsion and drive: Institute (VHIR), Barcelona, Spain. 4Research and Innovation Unit , Althaia involvement of the orbitofrontal cortex. Cereb Cortex 10(3):318–325 Xarxa Assistencial Universitària de Manresa , Manresa, Spain. 5Nuclear 24. Fumal A, Laureys S, Di Clemente L, Boly M, Bohotin V, Vandenheede Medicine Department , Vall d’Hebron University Hospital, Barcelona, Spain. M et al (2006) Orbitofrontal cortex involvement in chronic analgesic- overuse headache evolving from episodic migraine. Brain 129(Pt 2): Received: 24 May 2021 Accepted: 6 July 2021 543–550 25. Riederer F, Gantenbein AR, Marti M, Luechinger R, Kollias S, Sándor PS (2013) Decrease of Gray Matter Volume in the Midbrain is Associated with References Treatment Response in Medication-Overuse Headache: Possible Influence of 1. Lipton RB (2009) Tracing transformation: chronic migraine classification, Orbitofrontal Cortex. J Neurosci 33(39):15343–15349 progression, and epidemiology. Neurology 72(5 Suppl):S3–S7 Torres-Ferrus et al. The Journal of Headache and Pain (2021) 22:75 Page 11 of 11 26. Magis D, D’Ostilio K, Thibaut A, De Pasqua V, Gerard P, Hustinx R et al (2017) Cerebral metabolism before and after external trigeminal nerve stimulation in episodic migraine. Cephalalgia 37(9):881–891 27. Kim JH, Kim S, Suh S-I, Koh S-B, Park K-W, Oh K (2010) Interictal metabolic changes in episodic migraine: a voxel-based FDG-PET study. Cephalalgia 30(1):53–61 28. Kupers R, Kehlet H (2006) Brain imaging of clinical pain states: a critical review and strategies for future studies. Lancet Neurol 5(12):1033–1044 29. Lai K-L, Niddam DM, Fuh J-L, Chen W-T, Wu J-C, Wang S-J (2016) Cortical morphological changes in chronic migraine in a Taiwanese cohort: Surface- and voxel-based analyses. Cephalalgia 36(14):1324–1333 30. Woldeamanuel YW, DeSouza DD, Sanjanwala BM, Cowan RP (2019) Clinical Features Contributing to Cortical Thickness Changes in Chronic Migraine - A Pilot Study. Headache 59(2):180–191 31. Coppola G, Petolicchio B, Di Renzo A, Tinelli E, Lorenzo C, Di, Parisi V et al. Cerebral gray matter volume in patients with chronic migraine: correlations with clinical features The Journal of Headache and Pain. J Headache Pain. 2017;18 32. Filippi M, Messina R (2019) The Chronic Migraine Brain: What Have We Learned From Neuroimaging? Front Neurol 10:1356 Publisher’s Note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.