Pathways to the Best Fit of Triptans for Migraine Patients

Abstract

Variability in drug response is a major barrier to the successful treatment of migraine, and most treatments are only optimal in a subset of patients. Although triptans provide the best therapeutic option for the treatment of acute migraine, it has not previously been possible to predict how well patients will respond to a specific triptan or whether they will experience unpleasant adverse events. Hence, it has been difficult for physicians to match individual patients with the most suitable agent to treat their migraine pain. Intrapatient variability has been associated with polymorphisms in genes encoding drug-metabolizing enzymes, drug transporters and drug targets. Pharmacogenetics provides the possibility of tailoring the therapeutic approach to individual patients, in order to maximize treatment efficacy while minimizing the potential for unwanted side-effects. This review demonstrates how almotriptan may overcome genetically determined responses by utilizing diverse metabolic pathways to provide therapeutic benefit to many migraineurs.

Introduction

The diagnosis and management of migraine is complicated because patients exhibit a broad range of clinical characteristics and often have other comorbidities, the most frequent being anxiety, mood disorder and hypertension. Variability in drug response is a major problem in clinical practice: individual patients may respond differently to the same medication, both in terms of therapeutic effect and adverse events (AEs), even when given comparable doses of the drug. Some of this variability is due to genetic heterogeneity between patients, although non-genetic factors—including patient age, sex, race, comorbidities, concomitant medications, organ function (e.g. hepatic or renal impairment), lifestyle (diet, smoking and/or alcohol consumption) and drug–drug interactions—can also have marked effects (1).

Patients who experience migraine take drugs in order to relieve pain and/or prevent recurrent migraine attacks. However, most drugs provide optimal therapy in only a subset of patients. It is estimated that approximately half of the patients treated with migraine prophylactics will have a ≥ 50% reduction in attack frequency (2). With respect to acute treatment, it has been estimated that up to 25% of all migraine sufferers and up to 40% of all migraine attacks do not respond to a triptan, according to the criterion of pain freedom or pain relief after 2 h (3). A certain number of non-responders or poor responders to ergots or non-steroidal anti-inflammatory drugs are also present, although there are no specific reports on these issues.

An ideal acute migraine treatment is one that treats attacks rapidly and consistently, without recurrence, and has minimal or no side-effects. These are the attributes that migraineurs consider most important to their satisfaction with treatment. Greater understanding of the pathophysiology of migraine attacks has made it possible to use experimental models to determine the best pharmacological profile for treatment (4), but when making treatment decisions, physicians must rely on the large volume of clinical data in the literature for guidance, and reading numerous publications describing the results of original clinical studies is very time-consuming. However, systematic reviews and meta-analyses, where data are pooled from several similarly designed studies and compared using statistical techniques, provide physicians with more accessible clinical data and help inform treatment decisions (5). This approach has been used to compare the efficacy and tolerability of seven different triptans at 13 doses in 53 randomized, controlled clinical trials (6). The triptans currently represent the best treatment option for acute migraine, although their efficacy and the severity of AEs vary, both between agents and from patient to patient (6). This meta-analysis showed that for sumatriptan 100 mg, an average of 59% [95% confidence interval (CI) 57, 61] of patients achieved a headache response, whereas 29% (95% CI 27, 31) were pain free at 2 h, 20% (95% CI 18, 21) achieved sustained pain free, and 39% (95% CI 37, 41) experienced an AE (6). Success rates were 10–38% higher with some of the newer triptans compared with sumatriptan 100 mg, whereas the risk of AEs was less with almotriptan 12.5 mg and greater with eletriptan 80 mg (6). Studies included in the meta-analysis that evaluated intrapatient consistency (i.e. response rates to triptan treatment of three consecutive attacks) showed that 16–47% of patients responded to treatment in all three attacks, whereas 11–21% did not respond in any attack (6). Other studies have suggested that up to 15% of patients never respond to sumatriptan, whereas up to 40% of responders experience headache recurrence within 24 h (7).

The ultimate aim for providers and payers of healthcare is to provide the right drug to the right individual, at the right dose and at the right time. Until recently, it has not been possible to predict how an individual will respond to a particular triptan. Data now available regarding the use of triptans in clinical practice are increasingly enabling physicians to select the agent with the greatest likelihood of success for individual patients. Such individualized treatment stems from a better understanding of the clinical characteristics of migraine, coupled with clearer pharmacokinetic and pharmacodynamic profiles of the triptans. However, a more recent approach to provide patients with individualized treatment involves the use of pharmacogenomics.

Pharmacogenomics is the application of whole genomic technology to analyse gene and protein expression in order to identify the genetic polymorphisms and phenotypic traits that govern individual responses and AEs to drug treatment. Although still in its infancy, pharmacogenomics provides the possibility of providing a tailored therapeutic approach in individual patients in order to treat migraine effectively and prevent unwanted side-effects (5, 8–11).

Pharmacogenomics and drug-response profiling

Pharmacogenomics has already been used extensively in oncology with drugs such as trastuzumab, demonstrating that molecular data can facilitate assessment of disease heterogeneity and aid the identification of molecular markers that predict response to drugs (12). Similarly, using molecular diagnostics to identify the genetic variants associated with migraine susceptibility and polymorphisms that influence drug disposition and determine a patient's response to drugs, should make it increasingly possible to select the medications and doses that are optimal for individual migraineurs.

Polymorphisms in genes encoding drug-metabolizing enzymes, drug transporters and drug targets, as well as disease-modifying genes, have all been associated with individual variability in drug response and AE reactions (1, 13–15). Cytochrome P450—an ancient enzyme superfamily expressed mainly in the liver—is the most important system for metabolizing xenobiotics as well as many endogenous substrates. From an archeobiology point of view, the antiquity of this enzyme superfamily indicates that the system is absolutely necessary for survival and has gone through many changes over the millennia, which may explain the large number of existing polymorphisms. The human genome has 58 different cytochrome P450 genes, but only six isoforms have a significant role in drug metabolism. Together, the isoforms CYP2D6, CYP2C9 and CYP3A4 account for 60–70% of all phase I-dependent metabolism (i.e. oxidation reduction and hydrolysis reactions) of clinically important drugs (16). Some of these isoforms are genetically polymorphic and their activity can, therefore, vary tremendously between individuals. For example, the activity of the CYP2D6 isoform is genetically determined, but is also modified by environmental factors. CYP2D6 is responsible for the metabolism of 20–25% of the known drugs, including several that could be taken by migraineurs (i.e. antidepressants and β-blockers). Over 100 different allele forms of CYP2D6 have been described (17). These variant alleles can result in various mutations (including promoter or enhancer polymorphisms, gene duplication, defective splicing or gene deletion), which have consequent effects on activity or substrate specificity and result in > 100-fold variability in the rate of drug metabolism between individuals. These polymorphisms correlate with phenotypic subgroups exhibiting differing rates of drug metabolism, ranging from poor metabolizers to intermediate metabolizers, efficient metabolizers and ultra-rapid metabolizers of CYP2D6 drugs (18, 19). People with multiple copies of CYP2D6 metabolize drugs faster than others are able to do, and therapeutic plasma levels are, consequently, not achieved at typical drug doses; dose adjustments may therefore be required in order to achieve a therapeutic effect in this population. However, people lacking functional CYP2D6 genes and those with inactive enzymes metabolize drug substrates more slowly (or not at all) and consequently have a higher risk of AEs because of drug accumulation. In cases where pro-drugs require activation by CYP2D6 in order to become effective (e.g. codeine), lack of CYP2D6 results in reduced drug effectiveness. In contrast, ultra-rapid metabolism of codeine can lead to the accumulation of active metabolites and result in life-threatening opioid intoxication and acute renal failure (20). The frequency of these polymorphic alleles varies significantly between different ethnic populations (19), although CYP2D6 polymorphisms seem to be equally distributed within a given geographical area (21), suggesting that polymorphisms may be related mostly to environment rather than to ethnicity.

Genetic variation in drug targets (receptors) can also predispose individuals to different drug responses, both in terms of efficacy and AEs. For example, polymorphism in the β₂-adrenergic receptor has been associated with altered susceptibility to bronchodilator desensitization in asthma patients (22), whereas polymorphism within the coding region of the 5-HT_2A receptor affects the response to clozapine in patients with schizophrenia (23). Furthermore, mutations in the ryanodine receptor gene have been associated with susceptibility to the development of malignant hyperthermia following administration of certain anaesthetics (24).

Another feature that can be determined by pharmacogenomics is gene–gene interference, i.e. the interplay between genes that express for enzymes acting on the same substrate and that may or may not affect the therapeutic response of a drug. Knowledge of the factors influencing drug-metabolizing activity helps physicians to predict possible interactions between concurrently administered drugs. Drug interactions where the effect of one drug results in the inhibition or induction of metabolism of another can result in drug toxicity or therapeutic failure, and may necessitate dose adjustments. Such drug interactions may contribute to the variability in clinical response in migraine patients, as most patients receive concomitant drugs in addition to triptans.

Pharmacogenomics and migraine therapy

Pharmacogenetics has undoubtedly contributed to recent advances in identifying genetic polymorphisms associated with the different forms of migraine (8, 16). Using pharmacogenomics to profile drug use in migraineurs should assist physicians in predicting whether a patient will respond to a triptan or experience intolerable AEs, and thereby enable patients with differing genotypes to be matched to the most suitable drug for their metabolism. In addition, it may be possible to deduce to what extent an individual's triptan metabolism may interfere with any concomitant prophylactic treatment, which may affect the efficacy and tolerability of that drug.

In order to investigate if genetic receptor diversity may lead to the observed variable clinical responses of migraine patients to triptans, the allele frequencies of two common polymorphisms in the 5-HT_1B receptor gene (G861C and T261G) were compared in patients showing differing therapeutic outcomes to sumatriptan (consistently good response, headache recurrence, or no response) and differing tolerability outcomes (expressed as the presence or absence of chest symptoms). Differences in allele frequency were, however, insufficient to account for differences in outcome, indicating that genetic diversity of the 5-HT_1B receptor does not appear to influence either efficacy or tolerability outcomes, at least for sumatriptan (25).

In the acute treatment of migraine, clinical data have shown that almotriptan has several benefits over other triptans in terms of efficacy, speed of action and tolerability (6, 26, 27). Pharmacogenomics has now provided further evidence to support almotriptan as the migraine treatment of choice for most patients. In vitro studies have shown that almotriptan can be metabolized by several different metabolic pathways involving monoamine oxidase (MAO)-A, cytochrome P450s and flavin monooxygenase-3 (Fig. 1) (28). Ensuring that a drug can be metabolized is important in terms of maintaining its pharmacokinetic and pharmacodynamic profiles. Reduced metabolism results in drug accumulation and, although this may prolong efficacy, it also increases the risk of AEs. Therefore, having multiple routes for metabolism is advantageous because it means that there are alternative metabolic pathways with in-built redundancy—so if one pathway is inactive, others can take over to ensure the effectiveness of the drug. In contrast to almotriptan, sumatriptan and rizatriptan are metabolized only by MAO-A, whereas eletriptan, naratriptan and frovatriptan are metabolized only by cytochrome P450 enzymes (Table 1) (29). Having a single metabolic pathway increases the likelihood of AEs if the enzyme(s) associated with the pathway is absent or defective. Variation in MAO metabolism is believed to contribute to differences in the therapeutic effect and AEs seen in patients taking similar doses of sumatriptan. Drugs with single metabolic pathways are also more likely to interact with other drugs metabolized by the same pathway. Consequently, as sumatriptan and rizatriptan are metabolized only by MAO-A, they cannot be administered to patients taking MAO inhibitors. Propranolol interferes with rizatriptan metabolism, resulting in increased bioavailability and risk of rizatriptan toxicity. Propranolol increases plasma concentrations of rizatriptan by inhibiting MAO-A (30). When prescribing rizatriptan to migraine patients receiving propranolol for prophylaxis, the 5-mg dose of rizatriptan is recommended. Administration with other β-adrenoceptor blockers does not require consideration of a dose adjustment.

Table 1

Metabolic pathways involved in the metabolism of triptans (29)

Figure 1

Metabolism of almotriptan (28). Reproduced with permission from the American Society for Pharmacology and Experimental Therapeutics.

Eletriptan is metabolized only by CYP3A4 and can interact with drugs that are also substrates for this enzyme; dose adjustments are usually necessary when eletriptan is co-administered with inhibitors or inducers of CYP3A4 (29, 31). Co-administration of eletriptan with CYP3A4 inhibitors is discouraged because its metabolism is inhibited and, consequently, there is increased exposure to high plasma concentrations of eletriptan. Some patients require higher doses of eletriptan because they possess an active P-glycoprotein efflux transporter at the blood–brain barrier, which limits the amount of drug entering the brain and reaching central sites of action (31). Studies in mice have shown that there was a 40-fold difference in brain exposure to eletriptan between those expressing the gene for the P-glycoprotein efflux pump and those lacking the pump (31). The presence or absence of this efflux pump adds to the difficulty in identifying the correct dose of eletriptan to give to patients with migraine, as the higher doses needed to achieve a therapeutic effect also result in more side-effects. The presence or absence of the P-glycoprotein efflux pump gene had a much smaller effect on brain exposure to other triptans, which is believed to be related to their lower lipophilicity compared with eletriptan. Almotriptan has the lowest lipophilicity of all the triptans; this property is thought to be the reason for almotriptan's low rate of central nervous system side-effects (−1.5% vs. placebo) (9).

As almotriptan has several metabolic pathways, there is less risk of drug–drug interactions and, consequently, no dose modification is required when almotriptan is co-administered with MAO inhibitors or cytochrome inhibitors or inducers (29).

A recent meta-analysis of 53 clinical trials involving 24 089 migraineurs compared sustained pain-free and AE rates associated with various triptans, and found that there was a significant relationship between efficacy and tolerability for most individual agents (32). However, almotriptan 12.5 mg had an AE rate approximately 30% lower than would be expected, whereas eletriptan had an AE rate approximately 20% higher than expected, based on their efficacy (32). The use of composite end-points measuring treatment success in terms of both efficacy and tolerability, such as sustained freedom from pain with no AEs, should aid in the selection of agents that can offer patients the highest likelihood of consistent treatment success and, therefore, improve patient satisfaction with acute migraine treatment.

Conclusions

In summary, pharmacogenomics is a powerful tool and represents the future of research in migraine treatment. Genetically determined responses to drugs may facilitate the personalization of individual treatment for patients and lead to improved migraine therapy. Almotriptan is an effective agent for the treatment of acute migraine; its diverse metabolic pathways may overcome different genetically determined responses in terms of efficacy and side-effects, thereby providing therapeutic benefit to many migraineurs.