**3. Genetic aspects of cluster‐type headache**

Cluster headache (CH), the most severe primary headache, is characterized by recurrent, uni‐ lateral attacks of headache of great intensity and brief duration, accompanied by local signs and symptoms of autonomic dysfunction. In about 80% of the patients, the attacks occur in series lasting weeks or months, so‐called cluster periods [3]. The disease has an estimated prevalence of 1/500 and displays marked sex bias (female:male ratio 1:2.5 to 1:3.5) [43, 44].

Twin studies represent one of the simplest ways to unravel the relative importance of genetic and environmental effects. Much of the available literature on CH is reported in numerous concordant monozygotic twin pairs [45–48]. However, other studies demonstrated discordant twin pairs showed the importance of both genetic and individual specific factors and environ‐ mental factors in CH [49]. Beside this, epidemiological surveys indicate that compared with the general population, the first‐degree relatives of CH patients had a 14 fold increase in the disease risk in affected Danish probands and 39 fold increase in affected Italian probands. Also second‐degree relatives have two times higher risk in Danish probands and eight times higher risk in Italian probands [50–52]. The different results can be partly explained by meth‐ odological differences or selection bias. The increased familial risk of CH strongly suggests a genetic cause for the disease. However, the pattern of inheritance does not appear to be uniform. The familial clustering supports a model of autosomal dominant inheritance with reduced penetrance in some families but autosomal recessive model in others [53, 54] Also a study by Sjöstrand demonstrated a significantly lower mean age of onset in the second/third generation of families with CH than in the first generation. This can be explained by anticipa‐ tion or selection bias, since individuals with late age at onset from the second/third generation may not yet have symptoms [55, 56].

was significantly higher in CTTH patients with analgesic overuse. And it was suggested that serotonergic activity may be involved in the development of analgesic overuse in CTTH and that 5‐HTTLPR might be one of the genetically contributing factors [34]. But no significant differences were noticed between the 5‐HTTLPR and VNTR haplotype groups and success in treatment. 5HT2c‐receptor (5HTR2c) is another subtypes of 5HT2 families. The relationship between 5HTR2c Cys23Ser polymorphism and TTH was also investigated. However, there

Increased homocysteine levels are associated with various pathological conditions in humans, including stroke and cardiovascular disorders. So vasodilation of cerebral blood vessels may result in headache, or high levels of homocysteine may cause temporary thrombosis of cere‐ bral blood vessels, allowing less oxygen into the brain thus possibly causing headache. Frosst et al. reported an association between the homozygous C677T mutation in the 5,10‐methy‐ lenetetrahydrofolate reductase (MTHFR) gene and serum homocysteine levels [37]. In a case‐ control study, the prevalence of two common MTHFR polymorphisms, C677T and A1298C, in tension‐type headache patients and healthy controls was compared. And it was suggested that patients with C1298C and C677C/C1298C genotypes may predispose to tension‐type headache [38]. On the contrary, the study of Kowa et al. could not reach the same results. It was reported that MTHFR gene polymorphisms was not a genetic risk factor for TTH in their

Also it was investigated the relationship between TTH and tumor necrosis factor (TNF) gene polymorphisms (TNFA 308G > A and TNFB 252G > A) [40] and also estrogen receptor [ESR1 PvuII (rs2234693), ESR1 325 C→G (rs1801132)] and progesterone receptor [PROGINS (rs1042838)] polymorphisms [41]. But no risk was observed when TTH patients were com‐ pared with HC. Similarly, in a study that evaluates the relationship between GSTM1, T1 and P1 gene polymorphism and TTH, no difference was found between two groups in the geno‐

Cluster headache (CH), the most severe primary headache, is characterized by recurrent, uni‐ lateral attacks of headache of great intensity and brief duration, accompanied by local signs and symptoms of autonomic dysfunction. In about 80% of the patients, the attacks occur in series lasting weeks or months, so‐called cluster periods [3]. The disease has an estimated prevalence of 1/500 and displays marked sex bias (female:male ratio 1:2.5 to 1:3.5) [43, 44].

Twin studies represent one of the simplest ways to unravel the relative importance of genetic and environmental effects. Much of the available literature on CH is reported in numerous concordant monozygotic twin pairs [45–48]. However, other studies demonstrated discordant twin pairs showed the importance of both genetic and individual specific factors and environ‐ mental factors in CH [49]. Beside this, epidemiological surveys indicate that compared with the general population, the first‐degree relatives of CH patients had a 14 fold increase in the disease risk in affected Danish probands and 39 fold increase in affected Italian probands.

were no differences found among TTH and control groups [36].

8 Current Perspectives on Less-known Aspects of Headache

study [39].

type and allele distribution [42].

**3. Genetic aspects of cluster‐type headache**

To date, no clear molecular genetic evidence has been shown for CH. A point mutation was reported in mitochondrial transfer RNAleu(UUR) gene at nucleotide pair 3243 in a Japanese man with sporadic CH [56]. However, this mutation was not detected in Italian and German patients with CH [57, 58] and the involvement of mitochondrial genes in CH remains unproven.

Neuroimaging studies have identified the posterolateral hypothalamic grey matter as the key area for the basic defect in CH [59]. Hypocretin‐1 and ‐2 (also called orexin‐A and ‐B) are newly discovered neuropeptides [60, 61]. Hypocretin‐containing cells are located exclusively in the posterolateral hypothalamus, with widespread projections to the entire neuroaxis. Hypocretin‐1 and ‐2 bind to 2 G protein‐coupled receptors, termed HCRTR1 and HCRTR2. The peptides of the hypocretin/orexin system influence a wide range of physiologic and behavioral processes in mammals [62, 63]. Some of these, such as sleep, neuroendocrine, loco‐ motor, autonomic regulation, feeding behavior and energy homeostasis, may be of relevance for the pathogenesis of CH. Also a striking feature of CH is its diurnal and seasonal period‐ icity, suggesting that circadian and infradian rhythms regulate CH attacks. The hypocretin system plays a pivotal role in generating such rhythms and hypocretin‐containing neurons originate almost exclusively from the posterolateral hypothalamus [60, 61, 64]. Recent studies suggest a contribution of hypocretin to the pathogenesis of CH. A study among 109 Italian CH patients showed a strong association between the hypocretin type 2 receptor (HCRTR2) G1246A polymorphism and CH [65]. This association was confirmed in a major study from Germany, showing that homozygous carriers of the G‐allele had a twofold increase in disease risk [66]. Also in another study among Italian patients five additional intronic polymorphisms was genotyped, covering more than 75% of the entire 108.35 kb sequence of the HCRTR2 gene. And the carriage of the GTAAGG haplotype was shown to be associated with the dis‐ ease and resulted in a 3.7‐fold increased risk for CH [67]. On the contrary, the association was not replicated in a dataset of CH patients of Danish, Swedish and British origin [68]. In addition there are two published meta‐analysis studies investigating the association between polymorphisms of the HCRTR2 gene and CH. However, there are conflicting results between two studies, Rainero et al. suggested that the G1246A polymorphism of the HCRTR2 gene may modulate the genetic risk for CH [69] but Weller et al. did not find evidence for associa‐ tion of G1246A polymorphism (rs2653349) [70].

Besides this, the association between CH and a variable number tandem repeat (VNTR) polymorphism of the PER3 clock gene that has been associated to preferred daily rhythm (chronotype) in several studies was investigated. The hypothalamic biological clock may thus be involved in the pathophysiology and 149 patients were genotyped, but no difference in PER3 VNTR polymorphisms between patients and controls was found. And no association between CH, PER3 VNTR polymorphism and chronotype was found in the study [71].

Also some researchers performed a genetic association study to evaluate the relationship between CH and polymorphisms in the Clock gene, another highly conserved circadian gene, that influence the circadian phase in humans [72]. But they found that phenotype and allele frequencies were similarly distributed in CH patients and controls. Also it was determined that the clinical features of the disease were not significantly influenced by different geno‐ types. In conclusion, studies reported that the 3092 T‐‐>C polymorphism of the Clock gene is unlikely to play an important role in CH [73, 74].

Recent studies suggested that iron metabolism may be involved in the pathophysiology of primary headaches. The genetic association studies are shown in **Table 2**. In patients with migraine and chronic daily headache, Welch and colleagues [75] reported elevated iron con‐ centrations in the periaqueductal gray matter, one of the pain‐modulating centers of the brainstem. To evaluate whether mutations of the HFE gene would modify the occurrence and the clinical features of CH, an association study was performed in a cohort of Italian CH patients and healthy controls. They did not find C282Y mutation in both controls and cases. The prevalence of the H63D mutation was nearly similar in controls and cases so it was sug‐ gested that genetic variations within the HFE gene are associated with CH. But the HFE D63D genotype determined showed the onset of the disease at a significantly later age in compari‐ son with both H63H and H63D patients. So they recommended the HFE gene may influence the disease phenotype and may be regarded as a disease modifier gene [76].

Nitric oxide (NO) plays a critical role in the regulation of vasodilation, neurotransmission, inflammation and many other events throughout the body. NO also appears to be an impor‐ tant mediator of vascular headache pathophysiology [77, 78]. And an association analysis of five polymorphic microsatellite markers in the three different NO synthase (NOS) genes; nNOS (NOS1), iNOS (NOS2A) and eNOS (NOS3) was performed. However, it is unlikely that genetic variations within the NOS genes contribute greatly to CH susceptibility [79].

Methylenetetrahydrofolate reductase (MTHFR) is a key enzyme in the remethylation path‐ way converting homocysteine to methionine. The TT genotype of the common MTHFR 677C>T polymorphism (rs1801133) has been shown to impair enzyme activity and increase homocysteine levels [37]. It was shown that this variant has been linked to migraine in recent meta‐analyses [80, 81] Homocysteine and oxidized metabolites like homocysteic acid exert excitatory effects on neurons and homocysteic acid has been shown to increase cell firing of trigeminal neurons [82]. Hence, a link between the MTHFR 677C>T polymorphism and CH is plausible. But in an investigation on the association between the MTHFR 677C>T polymor‐ phism and CH among German patients and controls do not indicate an association between genotypes of the MTHFR 677C>T polymorphism and CH overall [83].

Alcohol is a well‐known trigger factor for CH attacks during the active phases of the disease. The alcohol dehydrogenase (ADH) pathway, which converts alcohol to the toxic substance


**Table 2.** The genetic association studies in CH.

(chronotype) in several studies was investigated. The hypothalamic biological clock may thus be involved in the pathophysiology and 149 patients were genotyped, but no difference in PER3 VNTR polymorphisms between patients and controls was found. And no association between CH, PER3 VNTR polymorphism and chronotype was found in the study [71].

Also some researchers performed a genetic association study to evaluate the relationship between CH and polymorphisms in the Clock gene, another highly conserved circadian gene, that influence the circadian phase in humans [72]. But they found that phenotype and allele frequencies were similarly distributed in CH patients and controls. Also it was determined that the clinical features of the disease were not significantly influenced by different geno‐ types. In conclusion, studies reported that the 3092 T‐‐>C polymorphism of the Clock gene is

Recent studies suggested that iron metabolism may be involved in the pathophysiology of primary headaches. The genetic association studies are shown in **Table 2**. In patients with migraine and chronic daily headache, Welch and colleagues [75] reported elevated iron con‐ centrations in the periaqueductal gray matter, one of the pain‐modulating centers of the brainstem. To evaluate whether mutations of the HFE gene would modify the occurrence and the clinical features of CH, an association study was performed in a cohort of Italian CH patients and healthy controls. They did not find C282Y mutation in both controls and cases. The prevalence of the H63D mutation was nearly similar in controls and cases so it was sug‐ gested that genetic variations within the HFE gene are associated with CH. But the HFE D63D genotype determined showed the onset of the disease at a significantly later age in compari‐ son with both H63H and H63D patients. So they recommended the HFE gene may influence

Nitric oxide (NO) plays a critical role in the regulation of vasodilation, neurotransmission, inflammation and many other events throughout the body. NO also appears to be an impor‐ tant mediator of vascular headache pathophysiology [77, 78]. And an association analysis of five polymorphic microsatellite markers in the three different NO synthase (NOS) genes; nNOS (NOS1), iNOS (NOS2A) and eNOS (NOS3) was performed. However, it is unlikely that

Methylenetetrahydrofolate reductase (MTHFR) is a key enzyme in the remethylation path‐ way converting homocysteine to methionine. The TT genotype of the common MTHFR 677C>T polymorphism (rs1801133) has been shown to impair enzyme activity and increase homocysteine levels [37]. It was shown that this variant has been linked to migraine in recent meta‐analyses [80, 81] Homocysteine and oxidized metabolites like homocysteic acid exert excitatory effects on neurons and homocysteic acid has been shown to increase cell firing of trigeminal neurons [82]. Hence, a link between the MTHFR 677C>T polymorphism and CH is plausible. But in an investigation on the association between the MTHFR 677C>T polymor‐ phism and CH among German patients and controls do not indicate an association between

Alcohol is a well‐known trigger factor for CH attacks during the active phases of the disease. The alcohol dehydrogenase (ADH) pathway, which converts alcohol to the toxic substance

genetic variations within the NOS genes contribute greatly to CH susceptibility [79].

the disease phenotype and may be regarded as a disease modifier gene [76].

genotypes of the MTHFR 677C>T polymorphism and CH overall [83].

unlikely to play an important role in CH [73, 74].

10 Current Perspectives on Less-known Aspects of Headache

acetaldehyde, is responsible for most of the alcohol breakdown in the liver. And Rainero and colleagues investigated the association of genetic variants within the ADH4 gene with CH susceptibility and phenotype. They suggested that CH was associated with the ADH4 gene or a linked locus. For rs1126671 polymorphism, the carriage of the AA genotype, in comparison with remaining genotypes, was associated with a significantly increased disease risk of 2.33 times. [84] But the results were not confirmed in Swedish population. The data from this study did not support an association of the ADH4 SNPs rs1126671 and rs1800759 with CH [85].

Only about 70% of migraine and CH patients report significant treatment responses to triptans, which are agonists at 5‐HT1B/D receptors belonging to the family of G protein‐coupled recep‐ tors. A C825T polymorphism identified in the gene for the G protein β3 (Gβ3) subunit (GNB3) has been associated with an enhanced signal transduction via GPCR [86]. It was investigated whether a common polymorphism in the gene for the G protein β3 subunit (GNB3 C825T) modu‐ lates responder rates to triptans among a large cohort of Caucasian CH patients. It was suggested that pain relief by triptans is significantly modulated by a common genetic GNB3 variant [88].

Also mutations of the P/Q type calcium channel alpha 1 subunit (CACNA1A) gene on chro‐ mosome 19p13 have been shown to cause several neurological disorders with a wide clinical spectrum, mainly episodic diseases. Missense mutations of the gene cause familial hemiplegic migraine (FHM) and it is also likely to be involved in the more common forms of migraine. It was investigated whether the CACNA1A gene is also a candidate gene for CH. In this study an association analysis of an intragenic polymorphic (CA)n‐repeat with marker D19S1150 and a (CAG)n‐repeat in the 3'UTR region was performed, in 75 patients with CH in Swedish population. But it was found that genotypes and allele frequencies were similarly distributed in patients and controls. Also linkage disequilibrium between the two markers was similar in patients and controls. And it was suggested that any significance of the CACNA1A gene in CH is unlikely [88]. Similarly Haan J et al. suggested that there is no involvement of the calcium channel gene (CACNA1A) mutations in a Dutch family with CH [89].
