Evidence That

Cluster headache (CH) is a primary headache disorder with a complex genetic background. Several studies indicate a potential link between iron homeostasis and the pathophysiology of primary headaches. The HFE gene encodes for a protein involved in iron metabolism, while genetic variants in HFE have been associated with hereditary hemochromatosis (HH), an iron overload disorder. The objective of the current study was to examine the association of the more common HFE H63D variant, with the susceptibility to develop CH and diverse clinical phenotypes in a population of Southeastern European Caucasian (SEC) origin. Genomic DNA samples from 128 CH patients and 294 neurologically healthy controls were genotyped for the HFE rs1799945 (H63D) variant. H63D genotypic and allelic frequency distribution did not differ significantly between patients and controls (p > 0.05). Subgroup analysis revealed a significantly more frequent occurrence of the variant G allele in chronic compared to episodic CH patients, indicative for a possible correlation of the HFE gene with the susceptibility for disease chronification. Although homozygosity for the less prevalent H63D variant G allele was minimal in the CH cohort, the results of the present study are in accordance with previous studies in CH and migraine patients, suggesting that HFE H63D variant modifies the disease clinical characteristics. Hence, despite the absence of a per se association with CH susceptibility in the current SEC cohort, variability in HFE gene may be potentially regarded as a disease modifier genetic factor in CH.


Introduction
Cluster headache (CH) is a debilitating primary headache disorder, classified in the group of trigeminal autonomic cephalalgias (TACs) and affecting around 0.1-0.4% of the general population (Snoer et al. 2018). The pain attacks in CH are described as excruciating, strictly unilateral, lasting between 15 and 180 min and usually occurring in bouts up to 8 times daily with a circadian and circannual periodicity (Láinez and Guillamón 2017; Liampas et al. 2020). Ipsilateral, prominent, and consistent cranial autonomic symptoms are accompanying CH attacks (Vikelis and Rapoport 2016;Goadsby 2018). According to the criteria developed by International Classification of Headache Disorders 3rd edition (ICHD-3), CH can be sub-diagnosed as episodic (eCH) or chronic (cCH). In a small proportion of patients (about 10%), CH attacks are characterized as chronic, occurring in bouts lasting one year or more without remission or with pain-free periods lasting less than 3 months (Headache Classification Committee of the International Headache Society (IHS) 2018).
Although CH is considered a neurovascular disorder with hypothalamic activation and trigeminal-autonomic reflex playing a predominant role in disease manifestation, its complex pathophysiology is still partially understood. Population, family, and twin studies support a heritable component for this primary headache, with increased risk for developing CH in first-and second-degree relatives compared to the general population and higher concordance in monozygotic twins (Russell 2004;Gibson et al. 2019); thus, potential genetic risk factors seem to predispose to CH development (O'connor et al. 2020;Waung et al. 2020). Consequently, many studies endeavored to discover the genetic components implicated in the etiopathology of CH. Most genetic association studies investigated variants in candidate genes inferred to be involved in the observed CH pathophysiology-related molecular pathways and/or associated with other primary headache disorders, particularly migraine (Sjöstrand et al. 2002;Pinessi et al. 2005;Schürks 2010;Zarrilli et al. 2015;Ran et al. 2017Ran et al. , 2020Fan et al. 2018;Fourier et al. 2019;Gibson et al. 2019;Papasavva et al. 2020).
The most frequent type of hereditary hemochromatosis (HH), an endocrine disorder of iron overload, is associated with mutations in the HFE gene. The HFE (high Fe) gene, located in chromosome 6 (6p21.3), encodes for an atypical major histocompatibility complex (MHC) class Itype glycoprotein (Nixon et al. 2018;Pantopoulos 2018). HFE protein is implicated in iron homeostasis by interacting with transferrin receptor 1 (TfR1) on cellular membranes and thereby negatively regulates transferrin-mediated iron uptake. HFE interacts also with TfR2 and probably additional protein molecules, such as hemojuvelin (HJV) and bone morphogenetic proteins (BMPs) and their receptor (BMPR) via the BMP/small mothers against decapentaplegic (SMAD) signaling pathway, shaping a putative iron-sensing complex that upregulates hepcidin expression, the major negative regulatory hormone of systemic iron homeostasis (Kent et al. 2015;Wu et al. 2015;Jiang et al. 2017;Traeger et al. 2018;Katsarou et al. 2019). Excessive iron stores, elevated plasma iron levels, and inflammation induce hepcidin synthesis whereas is suppressed by hypoxia and erythropoietic drive such as iron deficiency anemia (IDA) (Stoffel et al. 2019). Two missense variants in the coding-region of the HFE gene are responsible for the majority of HH cases: the rs1800562 (C282Y) in exon 4, a G to A transition at nucleotide position 845 resulting in a cysteine to tyrosine substitution at amino acid 282, which abrogates HFE protein from reaching the cell surface and interacting with TFRs and hepcidin, and the rs1799945 (H63D) in exon 2, a C to G transition at nucleotide 187 resulting in histidine to aspartic acid substitution at amino acid 63, which reduces the affinity of the altered protein for TfR1 (Barton et al. 2015;Katsarou et al. 2016;Wang et al. 2017). Hence, due to malfunctioning in BMP/SMAD pathway of patients with HFE HH and Hfe − / − mice, the expression of hepcidin is inappropriately decreased and the response of hepcidin to iron intake is diminished (Pantopoulos 2008(Pantopoulos , 2018. Various studies investigated the potential role of iron homeostasis in brain disorders and specifically primary headaches, with some reports indicating a link between elevated iron levels in brainstem pain-processing structures and primary headaches. Welch et al. reported enhanced iron levels in the periaqueductal gray matter (PAG), a descending antinociceptive-modulating center, of episodic migraine patients with and without aura and in patients with chronic daily headache, that further elevate with the duration of the disease (Welch et al. 2001). In line with these findings, Kruit et al. showed enhanced iron deposition in multiple central pain mediating-nuclei, predominantly in putamen, globus pallidus, and red nucleus, with higher iron accumulation in migraineurs with longest disease history (Kruit et al. 2009). Conversely, the 9-year follow-up study failed to provide supportive evidence for progressive iron accumulation in deep brain nuclei related to migraine; nevertheless, the authors explained that these findings may be attributed to age-related iron increase, thus age can serve as conflicting factor (Palm-Meinders et al. 2017). Tepper et al. confirmed the association between increased iron deposition in basal ganglia, particularly in globus pallidus, and migraine risk as well as attack frequency and disease duration (Tepper et al. 2012). A more recent study supported the hypothesis of an increased iron deposition in basal ganglia nuclei, especially red nucleus and PAG, in chronic migraine patients (Domínguez et al. 2019). Few studies indicated an association between IDA and migraine incidence, particularly in females (Gür-Özmen and Karahan-Özcan 2015; Tayyebi et al. 2019). Furthermore, a large study in a Norwegian population indicated that phenotypic hemochromatosis, a progressive iron overload disease, and/or HFE C282Y/C282Y genotype carriers were both associated with increased headache prevalence among female participants . Coexistence of migraine and CH has been documented in some patients and a genetic link for these primary headache disorders has been suggested (Kudrow and Kudrow 1994;D'Amico et al. 1997). Lastly, hereditary haemochromatosis was reported in two cousins with CH, while normalization of iron deposits improved temporarily CH attacks . Consequently, if elevated iron levels in the brain painmodulating areas are implicated in headache pathophysiology, the co-existence of hemochromatosis could potentially have an impact on disease progression or development. Although the more common HFE H63D variant rarely causes HH, it is suggested to be a disease risk and progression modifier for several neurological disorders (Kim and Connor 2020). Thus, the HFE H63D variant serves as an interesting candidate genetic variant for CH susceptibility.
The aim of the current association study was to examine whether the rs1799945 (H63D) HFE variant might confer susceptibility to develop CH and/or diverse disease phenotypes in a cohort of Southeastern European Caucasian (SEC) CH patients and neurologically healthy control subjects.

Subjects and Methods
Study Population A total of 128 unrelated SEC CH patients (95 males and 33 females), aged between 22 to 68 years (mean ± standard deviation: 41.7 ± 10.3 years; 75.8% male) served as CH group. The diagnosis of CH was carried out by experienced neurologists according to the International Classification of Headache Disorders criteria (ICHD-3). The main clinical characteristics of CH cohort and demographic data for the study population are summarized in Table 1. Case subjects were recruited from specialized headache clinics located in Glyfada and Thessaloniki, Greece. Two hundred ninety-four unrelated neurologically healthy SEC control subjects (149 males, 145 females) with no personal and family history of migraine or any other headache disorder, aged between 21 to 85 years (mean ± standard deviation: 57.3 ± 13.0 years; 50.7% male) were recruited from the Neurology Department, University Hospital of Larissa, Greece. Data collected from control subjects included only age and gender. All study subjects were South-eastern European Caucasians. A written informed consent was provided by each study subject. The research was performed in accordance with the principles outlined in the Declaration of Helsinki. The study protocol was reviewed and approved by the appropriate Ethics Committees (Mediterraneo Hospital, Glyfada, Greece, and University Hospital of Larissa).

DNA Purification and Genotyping
Epithelial cells from the oral cavity of the study subjects were collected by sterile buccal swabs. Genomic DNA was extracted from the epithelial cells with the Nucleospin Tissue DNA Mini Kit (Macherey-Nagel GmbH & Co., KG, Düren, Germany) according to the manufacturer's protocol. DNA concentration in the purified samples was quantified by NanoDrop One Spectrophotometer (Thermo Fischer Scientific), and the samples were stored in −20 °C until screening of the selected HFE variant. H63D (rs1799945) genotypes of each participant were determined by the SNP TaqMan qPCR method using the Applied Biosystems QuantStudio 12 K Flex Real-Time PCR System.

Statistical Analysis
Continuous variables are shown as mean ± standard deviation (SD). Categorical variables are presented as frequencies (n) and percentages (%). Chisquare (χ 2 ) (Pearson or Fischer's exact when appropriate) tests were used to compare H63D variant frequency distribution between groups and CH subgroups under codominant, dominant, over-dominant, recessive genotypic, and allelic inheritance models. Multivariable logistic regression analysis was used to adjust for potential confounding factors including age and gender as covariates, to exclude any bias due to the skewed age distribution and gender ratio between the groups. The crude and adjusted odds ratios (OR) with the corresponding 95% confidence intervals (CI) were calculated to estimate the risk derived from statistically significant differences. Statistical analysis of the data was carried out using version 26.0 of the IBM SPSS Statistics software for Windows. All p-values were two-sided, and values less than 0.05 were considered statistically significant. The web-based PS: power and sample size calculation program (v3.1.6), was used for power analysis (https:// vbios tatps. app. vumc. org/ ps/). With the available sample size and a minor allele frequency (MAF) of 0.14, true odds ratios (OR) below 0.32 or above 2.13 can be detected for CH cohort with 80% power (Dupont and Plummer 1990). Deviation from the Hardy-Weinberg Equilibrium (HWE) in control subjects was assessed with the web-based Online Encyclopedia for Genetic Epidemiology studies software (Rodriguez et al. 2009).

Genotype and Allele Frequency Distribution analysis Between Cases and Controls
The genotypic call rate for HFE H63D (rs1799945) was 98.8%. Consistency with HWE in the control group was verified for the HFE variant (p > 0.05). The genotype and allele frequencies of the investigated genetic variant in CH patients and controls are presented in Table 2. Chi-square test and logistic regression analysis with age and gender as covariates showed no statistically significant difference in genotypic and allelic frequency distribution of the HFE H63D variant between case and control subjects, in any of the genetic inheritance model tested (p > 0.05), possibly indicating an absence of association between H63D variant and susceptibility to CH in the current SEC population. (Table 3).

Subgroup Analysis According to CH Clinical Phenotypes
In order to investigate the association between HFE H63D variant and diverse CH phenotypes, patients were stratified into subgroups according to the form of CH (episodic/chronic) and the frequency of attacks. The CH group included 85 (69.1%) patients diagnosed with episodic CH form and 38 (30.9%) chronic CH patients. A statistically significant correlation between the HFE H63D and CH type was revealed.
As listed in

Discussion
The current study examined the association of H63D (rs1799945) variant in the HFE gene with the susceptibility to develop CH and diverse clinical phenotypes, in a SEC case-control population. Genotypic and allelic frequency distributions of the H63D genetic variant were similar between patients and control subjects, indicating no association between this HFE variant and CH risk susceptibility in SECs. Nonetheless, the stratified genotypic and allelic analyses according to CH subgroups (episodic and chronic CH) showed that the variant G allele was significantly more prevalent in chronic compared to episodic CH patients, conferring an increased risk for disease chronification for the patients carrying the variant G allele. Thus, although it is unlikely that HFE genetic variation contributes considerably to CH genetic susceptibility, a potential role for HFE as a disease modifier gene may be regarded. The possible molecular mechanisms interpreting the findings of the current study are not clear. As mentioned, various studies suggested a correlation between increased iron deposits in brainstem pain-processing structures and primary headaches. The HFE protein is expressed in the brain along with TfR; thus, variants in the HFE gene may influence brain iron uptake. HFE variants are related with enhanced iron accumulation (Bell et al. 2021); therefore, H63D variant might regulate iron deposition in brain pain-modulating centers and contribute to brain iron overload and increased oxidative stress states. Iron overload may alter the threshold for headache triggering, rendering carriers of HFE variants more susceptible to environmental challenges and/or other genetic modifiers (Nandar and Connor 2011;Nandar et al. 2013). Moreover, the presence of HFE variant genotypes may confer susceptibility for CH even in the absence of enhanced iron deposits. Lastly, a correlation between irondeficiency anemia (IDA) and the incidence of migraine was suggested (Gür-Özmen and Karahan-Özcan 2015; Tayyebi et al. 2019). To maximize iron absorption in IDA, hepcidin levels are usually repressed (Girelli et al. 2016;Dewan et al. 2019). Since the HFE protein is involved in the regulation of hepcidin expression, the underlying biological mechanism may be associated with defective hepcidin pathway.
Only one previously reported study by Rainero et al. investigated the impact of C282Y and H63D HFE missense polymorphisms on CH occurrence and clinical features, in an Italian cohort of 109 CH patients and 211 age-matched and geographically matched healthy controls. They observed no C282Y variant in the Northern Italian population of the cohort, whereas, similarly to the current study, they concluded that the examined HFE variants are not significantly associated with CH. On the contrary, they found no significant difference between episodic and chronic CH patients. The lack of statistically significant difference may be attributed to the small number of chronic CH patients included in the Italian CH cohort (14 chronic CH patients) and the absence of D63D genotype carriers in the chronic CH subgroup. Lastly, they observed a later age of disease onset in the four D63D homozygous patients, indicative of an involvement of HFE gene in disease modification ). An association study in an Italian population of 256 migraine patients and 237 healthy age-, sex-, and ethnicitymatched controls showed that patients with the H63D GG (D63D) genotype presented significantly later age of disease onset and increased frequency of migraine attacks, concluding that the HFE H63D variant may be a modifying genetic factor in migraine (Rainero et al. 2007).
Certain limitations of the current study should be acknowledged. Firstly, the relatively small sample size may not be sufficient to reveal small effects of the examined genetic variants in CH. Secondly, the study population was enrolled from a particular geographical area and does not consist of ethnically or racially different populations to avoid

Conclusion
Even though CH is regarded as a multifactorial primary headache disorder with both genetic and environmental inputs, the phenotypic variability defined by genetic variation is still uncertain. The results of the current study did not provide supportive evidence for significant association between the investigated HFE gene variant and CH susceptibility in the SEC population; hence, HFE gene might not be considered as a genetic risk factor for CH. Nevertheless, a correlation between H63D HFE variant and disease progression could be perceived; thus, in accordance with previous studies in primary headaches, HFE may serve as a candidate disease modifier gene. More, larger scaled studies in different populations, considering gene-gene and gene-environment interactions as well as additional iron metabolism regulating genes and biochemical parameters should be followed to elucidate the precise role of HFE and iron homeostasis in CH pathophysiology and their possible involvement in CH genetic susceptibility and diverse phenotypic characteristics. Detecting the underlying genetic components may potentially unravel the molecular pathways implicated in disease pathophysiology and facilitate the development of novel therapeutic targets.
Author Contribution MP, MSK, and ND conceptualized and designed the study. MV, VS, ED, and ED contributed to clinical data and sample collection. MP prepared the samples and performed the research. KK genotyped samples. MP and CP performed the statistical analyses. MP wrote the manuscript. GL contributed to the final editing. All authors read and approved the final manuscript.

Availability of Data and Material Available upon reasonable request.
Code Availability Not applicable.

Declarations
Ethics Approval Approval was obtained from the ethics committee of Mediterraneo Hospital, Glyfada, Greece, and from the University Hospital of Larissa Ethics Committee. The study was performed in line with the principles of the Declaration of Helsinki.

Consent to Participate
All individual participants included in the study gave prior written informed consent.

Consent for Publication
All participants provided written informed consent for publication of any associated data.

Conflict of Interest
The authors declare no competing interests.
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:// creat iveco mmons. org/ licen ses/ by/4. 0/.