A systematic review and critical appraisal of gene polymorphism association studies in medication-overuse headache

Abstract

Purpose of review

Medication-overuse headache is a secondary chronic headache disorder, evolving from an episodic primary headache type, caused by the frequent and excessive use of headache symptomatic drugs. While gene polymorphisms have been deeply investigated as susceptibility factors for migraine, little attention has been paid to medication-overuse headache genetics. In the present study we conducted a systematic review to identify, appraise and summarize the current findings of gene polymorphism association studies in medication-overuse headache.

Methods

A comprehensive literature search was conducted on PubMed and Web of Knowledge databases of primary studies that met the diagnostic criteria for medication-overuse headache according to the temporally-relevant Classification of Headache Disorder of the International Headache Society.

Results

A total of 17 candidate gene association studies focusing on medication-overuse headache were finally included in the qualitative review. Among these, 12 studies investigated the role of common gene polymorphisms as risk factors for medication-overuse headache susceptibility, six studies focused on the relationship with clinical features of medication-overuse headache patients, and four studies evaluated their role as determinants of clinical outcomes in medication-overuse headache patients.

Conclusion

Results of single studies show a potential role of polymorphic variants of the dopaminergic gene system or of other genes related to drug-dependence pathways as susceptibility factors for disease or as determinants of monthly drug consumption, respectively. In this systematic review, we summarize the findings of gene polymorphism association studies in medication-overuse headache and discuss the methodological issues that need to be addressed in the design of future studies.

Keywords

Medication-overuse headache genetics polymorphism systematic review

Introduction

Medication-overuse headache (MOH) is a daily or almost daily headache that results from the chronicization of a pre-existing primary headache as a consequence of symptomatic drug overuse (1). It is a paradoxical headache disorder, in which the cure for the headache becomes its leading cause in a susceptible patient (2). All acute headache medications have the potential to cause MOH. These include both migraine-specific medications, such as triptans and ergot alkaloids, and other pain-relief medications nonspecific for headache treatment, such as NSAIDs, paracetamol, combination analgesics, opioids and barbiturates (1,3). Epidemiological data indicate that MOH affects approximately 1–2% of the general population, with a similar prevalence in Europe, North America and Asian countries (1,4). It is noteworthy that MOH places a high economic burden on society, the per-person annual costs being estimated to be three times higher than those of migraine and 10 times higher than those of tension-type headache (5). Despite drug withdrawal therapy being performed very differently within and across countries, it represents the first-choice treatment for MOH (6). In most MOH patients, chronic headache improves significantly or is even reverted to its previous episodic pattern within two months of drug withdrawal (7,8). However, 30–45% of MOH patients with successful withdrawal therapy relapse into medication overuse, the majority of them within the first year of follow-up (9).

Various clinical and epidemiological studies suggest that MOH has a probable genetic basis. A hereditary susceptibility to MOH has been suggested, as the risk of developing MOH is greater in individuals with a family history of MOH or other substance abuse (10). In addition, patients overusing analgesics for pain conditions other than headache showed a higher risk of developing MOH only when they had a previous history of migraine (11,12). Furthermore, MOH appears to share some pathogenic mechanisms with other kinds of addictive disorders (13), and genetic factors are established factors that contribute to drug addiction (14). While gene polymorphisms have been deeply investigated as risk factors for migraine (15), less attention has been paid to the role of polymorphic gene variants in MOH susceptibility or as determinants of clinical features and outcomes of MOH patients. In order to assess the strength and quality of the evidence so far accumulated, in the present study we conducted a systematic review to identify, appraise and summarize the current findings of gene polymorphism association studies in medication-overuse headache.

Methods

Search strategy

The present systematic review was conducted in accordance with the PRISMA Statement principles (16) and the protocol published in the International Prospective Register of Systematic Reviews “PROSPERO” (registration number: CRD42017055206). A computerized literature search of PubMed and Web of Science databases was performed up to 13 January 2017 by using the Boolean combination of the following key terms: ((“Medication overuse headache” OR “MOH” OR “chronic headache with medication overuse” OR “chronic daily headache with medication overuse” OR “chronic daily headache with substantial medication overuse” OR “chronic medication overuse headache” OR “chronic daily headache associated with long-term misuse of headache medication” OR “drug abuse headache” OR “drug dependence headache” OR “drug-associated headache” OR “drug-induced headache” OR “drug-induced daily headache” OR “drug-induced chronic daily headache” OR “drug-induced refractory headache” OR “drug-rebound headache” OR “drug-related headache” OR “headache associated with chronic use of substances” OR “headache induced by chronic substance use” OR “headache induced by chronic substance exposure” OR “medication abuse-related headache” OR “medication-induced headache” OR “medication-misuse headache” OR “medication rebound headache” OR “painkiller headache” OR “primary headache diagnosis with medication overuse” OR “rebound headache” OR “rebound-withdrawal headache” OR “withdrawal headache”) AND (polymorphism OR polymorphisms OR SNP OR SNPs OR “single nucleotide polymorphism” OR “single nucleotide polymorphisms” OR “variant” OR “variants”)).

Relevant studies were required to meet the following inclusion criteria:

Human studies, including patients who met the diagnostic criteria for MOH according to the temporally relevant Classification of Headache Disorder of the International Headache Society (IHS) (17 –20)

Prospective or retrospective original studies investigating the association of common genetic variants with any of the following endpoints: i) susceptibility to MOH; ii) clinical features of MOH patients; iii) clinical outcomes of MOH patients.

Exclusion criteria were:

Studies not conducted in humans

Reviews, case studies, editorials, and meeting abstracts.

There were no language or time of publication restrictions. All potentially relevant studies retrieved in the initial screening step were then read in their entirety to assess their appropriateness for inclusion in the systematic review. The reference lists of all included studies and relevant reviews were checked to identify additional studies missed from the original electronic searches. Eligibility assessment of manuscripts was independently performed by two investigators (S.C. and S.T.) and any discrepancies were resolved through consensus.

Data extraction and quality assessment

A standardized form was used for each primary study included in the qualitative analysis, in which the following information was collected: first author’s last name, year of publication, study location, number of patients enrolled for each patients’ cohort, medication(s) overused, gene polymorphism analysed, primary (and, if any) secondary endpoint(s) investigated, statistics used to analyse genotype data, application or not of multiple testing correction, p-value used as cut-off for determining statistical significance and, for each endpoint investigated, the polymorphic variant found to be significantly associated. The scientific quality of studies included in the systematic review was evaluated according to the Strengthening the Reporting of Genetic Association (STREGA) recommendations for reports on genetic association studies (21). Briefly, the quality of the studies was assessed based on the clear definition or evaluation of the following 11 points: 1) Objectives and hypothesis, 2) eligibility criteria for study participants, 3) outcome(s) of interest, 4) power analysis or sample size calculations, 5) quality control of genotyping method, 6) assessment of Hardy-Weinberg equilibrium (HWE), 7) use of adjusted or multivariate analysis, 8) descriptive clinical data (e.g. age, gender), 9) statement of outcome results, 10) multiple testing correction, and 11) consideration of study limitation and potential bias. Studies with a total score of at least eight out of 11 points were considered of high quality. Data extraction and quality assessment were independently performed by two investigators (S.C. and S.T.). Any discrepancies were resolved through consensus.

Results

Characteristics of identified studies

The detailed flowchart of literature review process is reported in Figure 1. The search strategy on PubMed and Web of Knowledge identified a total of 128 studies, of which 16 were duplicates. After the removal of 95 hits not fulfilling the inclusion criteria, 17 studies were finally included in the qualitative review of gene association studies focusing on MOH (22 –38). The main characteristics of the identified studies are summarized in Table 1. Briefly, these studies were published between 2006 and 2015 by five research groups, four of which were located in Italy – including our group – and one in Japan, with sample sizes of MOH patients ranging from 22 to 227. Twelve studies had a case-control design (22,25 –29,31 –34,36,37), while the remaining five were case-only studies (23,24,30,35,38). Thirteen studies were conducted on MOH patients with Caucasian ancestry (22 –27,29,30,34 –38), while four studies recruited MOH patients with Asian ethnicity (28,31 –33). Most of the studies enrolled MOH patients overusing different classes of antimigraine medications, except one study which included MOH subjects overusing exclusively triptans (26). The median total score of study quality was eight (range: 3–11, online supplementary Table S1), indicating that the overall methodological quality was moderate to high. Seven studies (41%) conducted an a priori statistical power analysis (24,27,29,34 –37), while six out of 13 studies (46%) applied a method (Bonferroni’s) to correct for testing multiple gene variants (25,27,29,34 –36).

Figure 1.

Flowchart of the literature review process.

Table 1.

Characteristics and findings of studies included in the systematic review.

Author (Ref)	Year of publication	Country	Patients enrolled (N)	Medication(s) overused	Gene polymorphism(s) analyzed	Endpoint	Statistics used	Multiple correction (cut-off P_value)	Associated gene variant
Cevoli et al., (22)	2006	Italy	MOH: 103 EM: 101 HC: 117	Triptans (n = 30); NSAIDs (n = 37); combination analgesics (n = 73); overuse of more than one drug (n = 43)	DRD4 (48-bp VNTR) SLC6A3 (3' UTR 40-bp VNTR) COMT (rs4680) MAOA (T1460C)	Disease susceptibility (MOH vs. EM; MOH vs. HC)	Exact test for r x k contingency table based on Markov chain procedure and Fischer exact test	Not performed (0.05)	MOH vs. EM: DRD4 48-bp VNTR; SLC6A3 3' UTR 40-bp VNTR
Di Lorenzo et al., (23)	2007	Italy	MOH: 82	NR	WFS1 (His611Arg)	MOH clinical features	Multiple linear regression analysis	Not performed (0.05)	WFS1 His611Arg and monthly drug number; WFS1 His611Arg and BDI
Di Lorenzo et al., (24)	2009	Italy	MOH: 107	Triptans (n = 29); NSAIDs (n = 18); combination drugs (n = 27); overuse of more than one drug (n = 33)	BDNF (rs6265)	MOH clinical features	Stepwise multiple linear regression analysis	Not performed (0.05)	BDNF rs6265 and monthly drug number; BDNF rs6265 and MOH duration
Cevoli et al., (25)	2010	Italy	MOH: 138 EM: 101 HC: 117	Triptans (n = 49); NSAIDs (n = 40); acetaminophen (n = 5); metamizole (n = 2); combination analgesics (n = 91); overuse of more than one drug (n = 66)	SLC6A4 (STin2 VNTR) 5HT1A (A82G) 5HT2A (T102C) 5HT1B (T261G, G861C) 5HT6 (C267T)	Disease susceptibility (MOH vs. EM; MOH vs. HC)	Log-linear analysis for categorical data in two dimensional tables or Fisher’s exact test	Bonferroni’s correction (not mentioned)	No association found
Gentile et al., (26)	2010	Italy	MOH: 75 CM: 51	Triptans (n = 75)	MAOA (uVNTR) CYP1A2 (A163C, G3860A) GNB3 (C825T)	Primary endpoint: Disease susceptibility (MOH vs. CM) Secondary endpoint: Triptan response in abusers	Primary endpoint: chi-square test Secondary endpoint: chi-square test	Not performed (0.05)	Primary endpoint: CYP1A2 A163C Secondary endpoint: CYP1A2 A163C
Terrazzino et al., (27)	2010	Italy	MOH: 227 HC: 312	Triptans (n = 32); NSAIDs (n = 79); combination drugs (n = 37); ergot (n = 2); opiates (n = 1); two or more classes of drugs (n = 76)	5HT2A (C516T, A1438G)	Primary endpoint: Disease susceptibility (MOH vs. HC) Secondary endpoint: MOH clinical features	Primary endpoint: logistic regression analysis Secondary endpoint: Kruskal–Wallis or Mann–Whitney U-tests followed by stepwise multiple linear regression analysis	Bonferroni’s correction (0.025)	Primary endpoint: No association found. Secondary endpoint: 5HT2A C516T and monthly drug number
Ishii et al., (28)	2012	Japan	MOH: 22 EM: 47	Triptans (n = 2); simple analgesics (n = 9); combination analgesics (n = 14)	TNF-β (G252A)	Disease susceptibility (MOH vs. EM)	Chi-square or Fisher’s exact test	Not performed (0.05)	TNF-β G252A
Terrazzino et al., (29)	2012	Italy	MOH: 227 HC: 312	Triptans (n = 32); NSAIDs (n = 79); combination drugs (n = 37); ergot (n = 2); opiates (n = 1); two or more classes of drugs (n = 76)	SLC6A4 (5HTTLPR, STin2 VNTR, rs1042173)	Primary endpoint: Disease susceptibility (MOH vs. HC) Secondary endpoint: Monthly headache days	Primary endpoint: multivariate logistic regression analysis Secondary endpoint: multivariate linear regression analysis	Bonferroni’s correction (0.00625)	Primary endpoint: No association found Secondary endpoint: SLC6A4 haplotype (STin2 VNTR- rs1042173)
Di Lorenzo et al., (30)	2012	Italy	MOH: 43	Triptans (n = 8); NSAIDs (n = 17); triptans + NSAIDs (n = 11); combination medications (n = 7)	ACE (Ins/Del)	MOH clinical features	ANOVA	Not performed (0.05)	ACE Ins/Del and cortical response to somatosensory stimulation
Ishii et al., (31)	2013	Japan	MOH: 22 EM: 47	Triptans (n = 2); simple analgesics (n = 9); combination analgesics (n = 14)	TPH2 (rs4565946, rs4570625, rs4341581)	Disease susceptibility (MOH vs. EM)	Chi-square or Fisher’s exact test	Not performed (0.05)	No association found
Onaya et al., (32)	2013	Japan	MOH: 22 EM: 47	Triptans (n = 2); simple analgesics (n = 9); combination analgesics (n = 14)	SLC6A4 (5HTTLPR, STin2 VNTR); 5HT2A (T102C); 5HT1B (G861C); MAOA (uVNTR, T941G); MTHFR (C677T); ACE (Ins/Del); ESR1 (G325C, G594A); DRD2 (rs6275); BDNF (rs6265)	Disease susceptibility (MOH vs. EM)	Multivariate stepwise logistic regression analysis	Not performed (0.05)	MTHFR C677T; DRD2 rs6275
Ishii et al., (33)	2014	Japan	MOH: 22 EM: 47	Triptans (n = 2); simple analgesics (n = 9); combination analgesics (n = 14)	eNOS (rs1799983) nNOS (rs2682826) GTPCH (rs841)	Disease susceptibility (MOH vs. EM)	Chi-square or Fisher’s exact test	Not performed (0.05)	No association found
Cargnin et al., (34)	2014	Italy	MOH: 227 HC: 312	Triptans (n = 32); NSAIDs (n = 79); combination drugs (n = 37); ergotamine (n = 2); opiates (n = 1); two or more classes of drugs (n = 76)	DRD2 (rs1799732, rs1800497, rs6275) DBH (rs1611115) COMT (rs6269, rs4680) BDNF (rs6265) OPRM1 (rs1799971) EEAT2 (rs4354668) 5HT2A (C516T, A1438G) SLC6A4 (5-HTTLPR, STin2 VNTR, rs1042173)	Primary endpoint: Response to drug withdrawal therapy Secondary endpoint: Disease susceptibility (MOH vs. HC)	Primary endpoint: Multivariate logistic regression analysis Secondary endpoint: Logistic regression analysis adjusted by sex and age	Bonferroni’s correction (0.00357)	Primary endpoint: None. Note that carriers of at least four of the six top-ranked gene variants (rs1799971, rs4680, rs6275, rs6265, rs6305, rs1611115) were at higher odds for unsuccessful detoxification than patients with ≤3 high-risk genotypes Secondary endpoint: No association found
Cargnin et al., (35)	2014	Italy	MOH: 83	Triptans (n = 17); NSAIDs (n = 23); combination drugs (n = 12); two or more classes of drugs (n = 30)	SLC6A4 (STin2 VNTR) COMT (rs6269, rs4680)	Relapse within one year after successful withdrawal therapy	Multivariate logistic regression analysis	Bonferroni’s correction (0.012)	COMT rs4680; COMT rs6269; COMT rs6269-rs4680 haplotype
Cargnin et al., (36)	2015	Italy	MOH: 130 EM: 219 HC: 209	Triptans (n = 20); NSAIDs (n = 39); combination drugs (n = 22); ergotamine (n = 1); opiates (n = 1); two or more classes of drugs (n = 47)	CALCA (rs3781719, rs3754701) RAMP1 (rs7590387)	Disease susceptibility (MOH vs. EM; MOH vs. HC)	Multivariate logistic regression analysis	Bonferroni’s correction (0.0035)	MOH vs. EM: RAMP1 rs7590387
Gentile et al., (37)	2015	Italy	MOH: 96 HC: 45	NR	GSTP1 (rs1695) GSTM1 (Ins/Del) GSTT1 (Ins/Del) SOD2 (rs4880) CAT (rs1001179) eNOS (rs1799983) PON1 (rs662, rs854560) CYBA (rs4673, rs9932581)	Disease susceptibility (MOH vs. HC)	Chi-square test	Not performed (0.05)	No association found
Pisanu et al., (38)	2015	Greece, Italy	MOH: 23	NR	HDAC3 (rs2530223, rs1421896, rs32954, rs2547547, rs149330805, rs41290601)	Primary endpoint: MOH clinical features Secondary endpoint: Response to valproate treatment	Primary endpoint: Multivariate analysis Secondary endpoint: Fisher’s exact test	Not performed (0.05)	Primary endpoint: HDAC3 rs2530223 and monthly drug number; HDAC3 rs2530223 and days/month of drug abuse Secondary endpoint: No association found

BDI: Beck Depression Inventory; CM: chronic migraine; EM: episodic migraine; HC: healthy controls. MOH: medication-overuse headache; NR: not reported.

Gene polymorphism associations with susceptibility to MOH

Twelve case-control studies investigated the role of common genetic polymorphisms as risk factors for MOH susceptibility (22,25 –29,31 –34,36,37). Overall, 50 polymorphisms were analysed in 33 genes related to serotoninergic transmission (SLC6A4, 5HT1A, 5HT2A, 5HT1B, GNB3, 5HT6, TPH2), dopaminergic transmission (SLC6A3, COMT, DBH, DRD2 and DRD4), drug dependence (ACE, OPRM1, BDNF, EEAT2), metabolic pathways (CYP1A2, MAOA, MTHFR), oxidative stress (eNOS, nNOS, GTPCH, GSTP1, GSTM1, GSTT1, SOD, CAT, PON, CYBA), CGRP pathway (CALCA, RAMP1), and genes postulated to be relevant for migraine pathophysiology (ESR1, TNF-β). Among the identified studies, seven studies compared genotype or allelic frequencies of MOH patients with healthy controls (HCs) (22,25,27,29,34,36,37), seven considered the comparison of MOH patients with episodic migraineurs (EMs) (22,25,28,31 –33,36), and one study compared genotype or allelic frequencies of MOH patients with chronic migraineurs without drug overuse (26).

Among the seven studies assessing genotypic or allelic frequencies of MOH patients in comparison to healthy controls (HCs), three gene variants including 5HT6 C267T (25), SLC6A4 rs1042173 (29) and RAMP1 rs7590387 (36) emerged as nominally significant (p < 0.05) in at least one study. Specifically, an overrepresentation of the C allele of the 5HT6 C267T polymorphism was identified in MOH patients compared to HCs (p = 0.033) (25). In addition, results obtained from our group provided evidence of association between carriers of the minor allele for SLC6A4 rs1042173 (GT + GG) and higher risk of being affected by MOH (GT + GG vs. TT: OR 1.58, 95% CI: 1.05–2.37, p = 0.028) (29), and of association between RAMP1 rs7590387GG genotype and lower risk of MOH (GG vs. CC + CG: OR 0.43, 95% CI: 0.22–0.85, p = 0.011) (36). However, none of these three gene variants remains significant after Bonferroni correction or would be significant after its application.

Among the seven studies comparing genotype distributions between episodic migraineurs (EMs) and MOH patients, nominal associations emerged in five studies (22,25,28,32,36) of which two applied correction for multiple testing (25,36). Specifically, allele 10 of SLC6A3 3’ UTR 40-bp VNTR (p = 0.016) and allele 4 of DRD4 48-bp VNTR (p = 0.0053) were reported to be underrepresented in MOH patients compared to EMs (22). The same research group subsequently reported an association of the G allele of the 5HT1A A82G polymorphism with a lower risk of MOH (p = 0.039) (25). In a recent study (36), we found RAMP1 rs7590387 GG carriers at lower risk of episodic migraine transformation into MOH compared to rs7590387C allele carriers (vs. non-GG, OR: 0.27, 95% CI: 0.13–0.57, p = 0.0002; threshold p-value for Bonferroni significance = 0.0029). Based on results obtained from 22 MOH and 47 EM, one single research group from Japan reported in two studies (28,32) association of a total of three polymorphic gene variants including TNF-β G252A (p = 0.013), MTHFR C677T (p = 0.017) and DRD2 rs6275 (p = 0.015). Among all these nominally significant polymorphisms, only RAMP1 rs7590387 and DRD4 48-bp VNTR remained significant after Bonferroni correction or would be significant after its application. In the only study comparing MOH patients to chronic migraineurs without drug overuse (26), the 163A allele of CYP1A2 A163G was found to be overrepresented in MOH patients (p = 0.029), but statistical significance would be lost if Bonferroni correction were to have been applied.

Gene polymorphism associations with clinical features of MOH patients

Six studies investigated the correlation between candidate gene polymorphisms and clinical features of MOH patients (23,24,27,29,30,38). Overall, five gene variants (WFST His611Arg, BDNF rs6265, HDAC3 rs2530223, 5HT2A C516T, ACE I/D) and haplotype combination of SLC6A4 gene polymorphisms (STin2 VNTR-rs1042173) were reported to be associated to clinical features of MOH patients, in at least one study. Specifically, homozygous carriers of the minor allele of WFST His611Arg (R/R patients) showed significantly higher monthly drug consumption (p = 0.00075) and more severe depressive symptoms on the BDI questionnaire (p = 0.003) than non-R/R (23). The same authors subsequently reported that MOH patients harbouring the A allele (non-G/G) of BDNF rs6265 G > A showed a higher monthly consumption of acute drugs (p < 0.0001) and longer MOH duration (p = 0.004) when compared to G/G patients (24), while no significant differences between G/G and non-G/G patient groups were found in terms of other clinical and socio-demographic features. In a study from another group, the G allele of HDAC3 rs2530223 was significantly associated with the number of acute medications per month and with the number of days per month in which medications were used (p = 0.006 and p = 0.007, respectively), but neither with headache frequency nor with the intensity of pain (38). Results from our group revealed an association of the C516T polymorphism of 5HT2A with monthly drug consumption in MOH patients (p = 0.018; threshold p-value for Bonferroni significance = 0.025) (27), and association of the STin2 VNTR- rs1042173 haplotype of SLC6A4 with the number of headache days per month (global haplotype association p = 0.0056, threshold p-value for Bonferroni significance = 0.00625) (29). It is also noteworthy that angiotensin converting enzyme (ACE) I/D gene polymorphism significantly influenced the cortical response of MOH patients to somatosensory stimulation (30).

Gene polymorphism associations with clinical outcomes of MOH patients

Four studies investigated the role of common genetic variants as determinants of clinical outcomes in MOH patients (26,34,35,38), of which two stemmed from our group. In the first published study (34), we assessed 14 SNPs in eight genes (i.e. DRD2, DBH, COMT, BDNF, OPRM1, EAAT2, 5HT2A, SLC6A4) as predictors of response to drug withdrawal therapy at two months of follow-up. No polymorphism remained of statistical significance when the conservative Bonferroni correction was used for multiple testing. Nevertheless, in the multivariate analysis, carriers of at least four of the six top-ranked gene variants (rs1799971, rs4680, rs6275, rs6265, rs6305, rs1611115) were found to be at higher odds of unsuccessful detoxification than patients with ≤3 high-risk genotypes (OR: 3.40, 95 % CI 1.65–7.01, p = 0.001), an association which remained significant after correction for multiple testing (threshold p-value for Bonferroni significance = 0.00357). In the other study (35), we assessed the correlation between SLC6A4 STin2, COMT rs4680, COMT rs6269 and relapse risk at one-year of follow-up after successful detoxification (threshold p-value for Bonferroni significance = 0.012). COMT rs4680G allele carriers or the COMT rs6269G-rs4680G haplotype carriers were found to be at lower risk of relapsing into MOH when compared, respectively, to homozygous rs4680A allele carriers (OR: 0.17, 95% CI: 0.05–0.61, p = 0.007) or with the COMT rs6269A-rs4680A haplotype (OR: 0.19, 95% CI 0.06–0.54, p = 0.003). Regarding the two other studies investigating gene polymorphisms associations with clinical outcomes of MOH patients, no association was detected between HDAC3 gene polymorphisms and response to valproate treatment (38), while a nominal association was reported between the CYP1A2*1F allele and grade of response to triptan administration (26), which however would have not survived Bonferroni correction after its application.

Discussion

In the last decade, a number of gene association studies have attempted to assess the role of polymorphic gene variants as susceptibility factors for MOH or as determinants of clinical features and outcomes of MOH patients. In the present systematic review, we summarized the evidence accumulated so far by gene association studies focusing on MOH, with the aim of highlighting research clues and methodological issues useful for orienting towards a better design of future studies in the field. First, all the studies identified used a hypothesis-driven approach, based on the selection of polymorphisms in candidate genes postulated to be involved in MOH pathophysiology. Most of the studies so far conducted focused on polymorphic genes of serotonergic and dopaminergic systems, which have been selected on the basis of previous evidence showing abnormalities of serotoninergic and dopaminergic transmission in MOH patients (39 –41). In addition, polymorphic genes known to be implicated in drug addiction have been postulated in MOH patients to have a role in monthly drug consumption, which represents the clinical hallmark of the disease and the expression of the drug-seeking behavior (42). The lack of genome-wide association studies in MOH highlights an enormous distance between genetic research in MOH compared to migraine (15), and also suggests limited funds to this field, which is surprising given the high economic and social burden of MOH (5).

A number of the studies identified in the present systematic review reported positive findings on the role of polymorphic gene variants in MOH (Table 2). These are mainly located in genes of the dopaminergic system (DRD4, DRD2, COMT, SLC6A3) or in genes involved in drug dependence pathways (WSF1, BDNF, ACE, HDAC3), suggesting that the most promising area for further research in MOH genetics would be to focus on these pathways. However, the findings obtained from these studies must be cautiously interpreted in the light of the following methodological flaws. First, most of the studies did not apply correction for testing multiple gene variants, thus preventing ruling out with certainty the possibility of false-positive results (“type I errors”). This is surprising, given that methods including Bonferroni’s (43) or false discovery rate (44) are available in gene association studies to overcome type I errors. Second, the identified studies were performed mostly on a limited number of MOH patients, with 10 out of 17 studies having enrolled less than 100 MOH patients. While this observation may suggest potential challenges in the recruitment process of these patients, such a study-design shortcoming might have increased the chance of having incurred “type II errors”; that is, failure to detect an effect that, conversely, is actually present (i.e. “false negatives” results). Despite sample size and power calculations being strongly recommended to avoid type II errors (45), less than half of the studies (41%) conducted an a priori statistical power analysis to estimate the sample size required to achieve the expected effect size or, at least, the minimal detectable effect given the sample size. Third, no studies reporting significant associations attempted to validate their findings in an independent cohort of MOH patients. Future gene association studies in this field should use a two-stage discovery/replication design for corroboration of positive findings. To this end, major efforts should be made for the recruitment of larger cohorts of MOH patients, preferably in the context of multicenter and international collaborations. Fourth, the present systematic review also highlights that none of the positive findings were assessed in another study for replication. The lack of replication studies precluded the possibility of conducting a meta-analysis to pool estimates from different studies. Such an approach has proven to be useful for the identification of genuine genetic associations by overcoming the potential type II errors of primary studies (46). Lastly, it should be acknowledged that the identified studies, being conducted by few research groups, included MOH patients which partly overlapped between studies, making questionable the setting of a p-value threshold to be chosen for correction of multiple testing.

Table 2.

Summary of positive findings in gene polymorphism association studies in MOH.

		Classification of gene variants for their potential involvement in:
Author, year (Ref)	Patients’ ethnicity	Dopaminergic neurotransmission	Serotoninergic neurotransmission	Drug dependence pathways	Drug metabolism	Migraine pathophysiology
MOH susceptibility
Cevoli et al., 2006 (22)	Caucasian	MOH vs. EM: DRD4 48-bp VNTR SLC6A3 3’ UTR 40-bp VNTR
Gentile et al., 2010 (26)	Caucasian				MOH vs. CM: CYP1A2 A163C
Ishii et al., 2012 (28)	Asian					MOH vs. EM: TNF-β G252A
Onaya et al., 2013 (32)	Asian	MOH vs. EM: DRD2 rs6275			MOH vs. EM: MTHFR C677T
Cargnin et al., 2015 (36)	Caucasian					MOH vs. EM: RAMP1 rs7590387
MOH clinical features
Di Lorenzo et al., 2007 (23)	Caucasian			WSF1 His611Arg and monthly drug number or BDI
Di Lorenzo et al., 2009 (24)	Caucasian			BDNF rs6265 and monthly drug number or MOH duration
Terrazzino et al., 2010 (27)	Caucasian		5HT2A C516T and monthly drug number
Terrazzino et al., 2012 (29)	Caucasian		SLC6A4 haplotype (STIn2 VNTR-rs1042173) and monthly headache days
Di Lorenzo et al., 2012 (2012)	Caucasian			ACE Ins/Del and cortical response to somatosensory stimulation
Pisanu et al., 2015 (38)	Caucasian			HDAC3 rs2530223 and monthly drug number or days/month of drug abuse
Clinical outcomes of MOH patients
Gentile et al., 2010 (26)	Caucasian				CYP1A2 A163C and triptan response
Cargnin et al., 2014 (35)	Caucasian	COMT rs4680 COMT rs6269 COMT haplotype (rs6269-rs4680) and relapse risk

BDI: Beck Depression Inventory; CM: chronic migraine; EM: episodic migraine; MOH: medication-overuse headache.

MOH is a complex disorder that, in most cases, evolves from migraine, less frequently from tension-type headache (8,47) and sporadically from cluster headache (48) in parallel with the transformation of the primary headache from episodic into chronic. Since 2004, the ICHD classifications require the application of the double diagnosis: the primary headache that has chronified, and MOH (ICHD-II e ICHD-3beta). Yet, in many studies, only the MOH diagnosis is considered. This is the case for some of the papers considered in the present review (24 –26,30,37) that did not specify the type of primary headache from which MOH evolved, nor provided the double diagnosis. This misconception may of course disturb the comparability of the studies. It might also confound the interpretation of results also in the reverse condition, that is, when genetic factors for migraine chronification are searched without considering the presence or absence of medication overuse (49). In future studies, it is important that all existing diagnoses are specified a priori or, at the very least, that patients are stratified on the basis of the presence or absence of medication overuse.

In conclusion, the present systematic review highlights that genetic research on MOH is still in its infancy. Despite a relevant number of candidate gene association studies having reported positive findings, unequivocal and definitive conclusions cannot be drawn due to some methodological flaws, including weak study design and lack of internal or external replication studies. With the recent advances of high throughput genomic technologies, application of the genome-wide approach and of next-generation sequencing are expected to provide a significant breakthrough in unravelling MOH pathogenesis (50). It is noteworthy that, at the time of writing this article, a first genome-wide study of DNA methylation has been published, which reported association of several loci potentially involved in headache chronification (51). Nevertheless, candidate gene association studies with a better study design, including those with a discovery/replication design, could be still useful to validate previously reported genetic associations or for corroborating novel findings. In addition, GWAS studies are warranted to explore in a hypothesis-free manner the contribution of genetic background to MOH. In this frame, large cohorts of MOH patients appear to be an essential prerequisite to unequivocally identify genetic, epigenetic or pharmacogenetic determinants in MOH. This will undoubtedly benefit from multicenter and international collaborations as well as from a broad consensus on phenotype definition.

Clinical implications

This systematic review highlights the results of single studies showing the potential role of candidate polymorphic variants of the dopaminergic gene system (DRD4, DRD2, SLC6A3), or of other genes related to drug-dependence pathways (WSF1, BDNF, ACE, HDAC3), as risk factors for MOH susceptibility or as determinants of monthly drug consumption, respectively.

Nevertheless, unequivocal and definitive conclusions cannot be drawn due to some methodological flaws, including weaknesses of study designs and lack of replication studies.

Multicenter and international cooperative efforts are strongly warranted for the recruitment of larger cohorts of MOH patients in order to explore, in a hypothesis-free manner, the contribution of genetic factors to disease susceptibility as well as to variability of clinical features and outcomes of MOH patients.

Footnotes

Declaration of conflicting interests

The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The authors received no financial support for the research, authorship, and/or publication of this article.

References

Diener

Limmroth

. Medication-overuse headache: A worldwide problem. Lancet Neurol 2004; 3: 475–483.

Katsarava

Limmroth

Finke

et al.

Rates and predictors for relapse in medication overuse headache: A 1-year prospective study. Neurology 2003; 60: 1682–1683.

Dowson

Dodick

Limmroth

. Medication overuse headache in patients with primary headache disorders: Epidemiology, management and pathogenesis. CNS Drugs 2005; 19: 483–497.

Evers

Marziniak

. Clinical features, pathophysiology, and treatment of medication-overuse headache. Lancet Neurol 2010; 9: 391–401.

Linde

Gustavsson

Stovner

et al.

The cost of headache disorders in Europe: The Eurolight project. Eur J Neurol 2012; 19: 703–711.

Tassorelli

Jensen

Allena

et al.

A consensus protocol for the management of medication-overuse headache: Evaluation in a multicentric, multinational study. Cephalalgia 2014; 34: 645–655.

Zeeberg

Olesen

Jensen

. Probable medication-overuse headache: The effect of a 2-month drug-free period. Neurology 2006; 66: 1894–1898.

Tassorelli C, Jensen R, Allena M, et al. The added value of an electronic monitoring and alerting system in the management of medication-overuse headache: A controlled multicentre study. Cephalalgia 2017; 37: 1115–1125.

Katsarava

Muessig

Dzagnidze

et al.

Medication overuse headache: Rates and predictors for relapse in a 4-year prospective study. Cephalalgia 2005; 25: 12–15.

10.

Cevoli

Sancisi

Grimaldi

et al.

Family history for chronic headache and drug overuse as a risk factor for headache chronification. Headache 2009; 49: 412–418.

11.

Wilkinson

Becker

Heine

. Opiate use to control bowel motility may induce chronic daily headache in patients with migraine. Headache 2001; 41: 303–309.

12.

Bahara

Walsh

Menon

et al.

Does chronic daily headache arise de novo in association with regular use of analgesics?

Headache 2003; 43: 179–190.

13.

Calabresi

Cupini

. Medication-overuse headache: Similarities with drug addiction. Trends Pharmacol Sci 2005; 26: 62–68.

14.

Khokhar

Ferguson

Zhu

et al.

Pharmacogenetics of drug dependence: Role of gene variations in susceptibility and treatment. Annu Rev Pharmacol Toxicol 2010; 50: 39–61.

15.

Gormley

Anttila

Winsvold

et al.

Meta-analysis of 375,000 individuals identifies 38 susceptibility loci for migraine. Nat Genet 2016; 48: 856–866.

16.

Moher

Liberati

Tetzlaff

et al.

Preferred Reporting Items for Systematic Reviews and Meta-Analyses: The PRISMA Statement. PLoS Med 2009; 6(7): e1000097–e1000097.

17.

Headache Classification Committee of the International Headache Society. Classification and diagnostic criteria for headache disorders, cranial neuralgias and facial pain. Cephalalgia 1988; 8: S1–S96.

18.

Headache Classification Subcommittee of the International Headache Society. The International Classification of Headache Disorders, 2nd edn. Cephalalgia 2004; 24: S9–S160.

19.

Headache Classification Committee of the International Headache Society. New appendix criteria open for a broader concept of chronic migraine. Cephalalgia 2006; 26: 742–746.

20.

Headache Classification Committee of the International Headache Society (IHS). The International Classification of Headache Disorders, 3rd edition (beta version). Cephalalgia 2013; 33: 629–808.

21.

Little

Higgins

Ioannidis

et al.

STrengthening the REporting of Genetic Association studies (STREGA): an extension of the STROBE Statement. Ann Intern Med 2009; 150: 206–215.

22.

Cevoli

Mochi

Scapoli

et al.

A genetic association study of dopamine metabolism – related genes and chronic headache with drug abuse. Eur J Neurol 2006; 13: 1009–1013.

23.

Di Lorenzo

Sances

Di Lorenzo

et al.

The wolframin His611Arg polymorphism influences medication overuse headache. Neurosci Lett 2007; 424: 179–184.

24.

Di Lorenzo

Sances

et al.

Drug consumption in medication overuse headache is influenced by brain-derived neurotrophic factor Val66Met polymorphism. J Headache Pain 2009; 10: 349–355.

25.

Cevoli

Marzocchi

Capellari

et al.

Lack of association between five serotonin metabolism-related genes and medication overuse headache. J Headache Pain 2010; 11: 53–58.

26.

Gentile

Borro

Lala

et al.

Genetic polymorphisms related to efficacy and overuse of triptans in chronic migraine. J Headache Pain 2010; 11: 431–435.

27.

Terrazzino

Sances

Balsamo

et al.

Role of 2 common variants of 5HT2A gene in medication overuse headache. Headache 2010; 50: 1587–1596.

28.

Ishii

Onaya

Katoh

et al.

Association between the G252A tumor necrosis factor-β gene polymorphism and medication-overuse headache. J Clin Neurol 2012; 8: 301–304.

29.

Terrazzino

Tassorelli

Sances

et al.

Association of haplotype combination of serotonin transporter gene polymorphisms with monthly headache days in MOH patients. Eur J Neurol 2012; 19: 69–75.

30.

Di Lorenzo

Coppola

Currà

et al.

Cortical response to somatosensory stimulation in medication overuse headache patients is influenced by angiotensin converting enzyme (ACE) I/D genetic polymorphism. Cephalalgia 2012; 32: 1189–1197.

31.

Ishii

Katoh

Onaya

et al.

Tryptophan hydroxylase 2 gene polymorphisms in Japanese patients with medication overuse headaches. Acta Neurol Taiwan 2013; 22: 147–151.

32.

Onaya

Ishii

Katoh

et al.

Predictive index for the onset of medication overuse headache in migraine patients. Neurol Sci 2013; 34: 85–92.

33.

Ishii

Yahara

Katoh

et al.

Polymorphisms of nitric oxide synthase and GTP cyclohydrolase I genes in Japanese patients with medication overuse headaches. Neurology Asia 2014; 19: 277–281.

34.

Cargnin

Viana

Sances

et al.

Combined effect of common gene variants on response to drug withdrawal therapy in medication overuse headache. Eur J Clin Pharmacol 2014; 70: 1195–1202.

35.

Cargnin

Viana

Ghiotto

et al.

Functional polymorphisms in COMT and SLC6A4 genes influence the prognosis of patients with medication overuse headache after withdrawal therapy. Eur J Neurol 2014; 21: 989–995.

36.

Cargnin

Pautasso

Viana

et al.

Association of RAMP1 rs7590387 with the risk of migraine transformation into medication overuse headache. Headache 2015; 55: 658–668.

37.

Gentile

Negro

D'Alonzo

et al.

Lack of association between oxidative stress-related gene polymorphisms and chronic migraine in an Italian population. Expert Rev Neurother 2015; 15: 215–225.

38.

Pisanu

Caproni

Congiu

et al.

HDAC3 role in medication consumption in medication overuse headache patients: A pilot study. Hum Genomics 2015; 9: 30–30. DOI: 10.1186/s40246-015-0051-1.

39.

Srikiatkhachorn

Anthony

. Serotonin receptor adaptation in patients with analgesic-induced headache. Cephalalgia 1996; 16: 419–422.

40.

Srikiatkhachorn

Puangniyom

Govitrapong

. Plasticity of 5-HT serotonin receptor in patients with analgesic-induced transformed migraine. Headache 1998; 38: 534–539.

41.

Ferraro

Grazzi

Muffatti

et al.

In medication-overuse headache, fMRI shows long-lasting dysfunction in midbrain areas. Headache 2012; 52: 1520–1534.

42.

Radat

Lanteri-Minet

. What is the role of dependence-related behavior in medication-overuse headache? Headache 2010; 14: 1597–1611.

43.

Armstrong

. When to use the Bonferroni correction. Ophthalmic Physiol Opt 2014; 34: 502–508.

44.

Benjamini

Hochberg

. Controlling the false discovery rate: A practical and powerful approach to multiple testing. J Royal Stat Soc Ser B-Methodol 1995; 57: 289–300.

45.

Kim

Clark

Dionne

. Genetic contributions to clinical pain and analgesia: Avoiding pitfalls in genetic research. J Pain 2009; 10: 663–693.

46.

Ioannidis

. Genetic associations: False or true? Trends Mol Med 2003; 9: 135–138.

47.

Diener

Holle

Solbach

et al.

Medication-overuse headache: Risk factors, pathophysiology and management. Nat Rev Neurol 2016; 12: 575–583.

48.

Paemeleire

Bahra

Evers

et al.

Medication-overuse headache in patients with cluster headache. Neurology 2006; 67: 109–113.

49.

Louter

Fernandez-Morales

de Vries

et al.

Candidate-gene association study searching for genetic factors involved in migraine chronification. Cephalalgia 2015; 35: 500–507.

50.

Rudkjobing

Esserlind

Olesen

. Future possibilities in migraine genetics. J Headache Pain 2012; 13: 505–511.

51.

Winsvold BS, Palta P, Eising E, et al. Epigenetic DNA methylation changes associated with headache chronification: A retrospective case-control study. Cephalalgia. 2018; 38: 312–322.

Supplementary Material

Please find the following supplemental material available below.

For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.

For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.

0.00 MB

0.17 MB