Verapamil for Cluster Headache. Clinical Pharmacology and Possible Mode of Action
Conflict of Interest: None
P. Tfelt-Hansen, Danish Headache Centre, Department of Neurology, Glostrup Hospital, DK-2600 Glostrup, Denmark.
Verapamil is used mainly in cardiovascular diseases. High-dose verapamil (360-720 mg) is, however, currently the mainstay in the prophylactic treatment of cluster headache. The oral pharmacokinetics are variable. The pharmacodynamic effect of verapamil, the effect on blood pressure, also varies considerably among subjects. The dose of verapamil used for cluster headache is approximately double the dose used in cardiovascular disease, most likely because verapamil is a substrate for the efflux transporter P-glycoprotein in the blood–brain barrier. The access of verapamil to the central nervous system is therefore limited. The clinical use of verapamil in cluster headache is reviewed and several relevant drug interactions are mentioned. Finally, its possible mode of action in cluster headache is discussed. The effect of verapamil in cluster headache most likely takes place in the hypothalamus.Verapamil is an L-type calcium channel blocker but it is also a blocker of other calcium channels (T-, P-, and possibly N- and Q-type Ca2+ channels) and the human ether-a-go-go-related gene potassium channel. With so many different actions of verapamil, it is impossible at the present time to single out a certain mode of action of the drug in cluster headache.
It was recently stated that current therapeutic actions of the existing available Ca2+ channel blockers are essentially confined to disorders of the cardiovascular system.1 Cluster headache is, however, an exemption. Thus, high-dose verapamil is an increasingly common preventive treatment in cluster headache, and verapamil is currently the mainstay in the preventive treatment of cluster headache.2,3 Verapamil in a dose of 360 mg daily is the only drug that has been shown in a double-blind randomized clinical trial to be superior to placebo in the preventive treatment of cluster headache.4
The dose of verapamil needed in the preventive treatment of cluster headache patients varies considerably. Thus, doses from 240 mg to 960 mg are recommended depending on effect and adverse events.3 In a few patients, even higher doses (1200 mg) are needed.2 This could be due to either pharmacokinetic or pharmacodynamic variability, or both, as suggested for acute migraine treatment.5
In the following commentary, we will shortly review the clinical pharmacology of verapamil. Then, the differences in doses of verapamil used in cardiology and neurology will be described. The clinical use of verapamil in cluster headache will be commented on; and finally, the possible mode of action will be discussed.
PHARMACOKINETICS OF VERAPAMIL
Verapamil is available as a racemate that contains equal amounts of (R)-verapamil and (S)-verapamil.6 Verapamil is lipophilic and easily absorbed from the gut.7 The drug is, however, the subject of extensive first-pass metabolism in the gut wall and the liver and the total oral bioavailability is only 10% to 30%.8 After multiple dosing of controlled-release verapamil, the oral bioavailabilty was 27% for (S)-verapamil and 55% for (R)-verapamil.9
Verapamil has a volume of distribution of 300 L for (R)-verapamil and 500 L for (S)-verapamil.9 It is cleared by the liver by the cytochrome CYP3A4 enzyme with a half-life of 6-7 for both enantiomers.10 The metabolism of verapamil is stereospecific with higher plasma concentrations of (R)-verapamil than (S)-verapamil6 with a ratio of 5 : 1 after oral administration.11 Verapamil is metabolized to active metabolites, for example norverapamil, which has 20% of the potency of verapamil to block atrioventricular conduction.12
These data are, however, mean values and as with other extensively metabolized drugs, pharmacokinetics vary considerably between subjects.13 After 240 mg as a controlled-release tablet, the Cmax varied 9-fold (30-278 ng/mL).13 After multiple doses of 240 mg verapamil, the area under the curve (AUC) from intake to 24 hours varied considerably. The mean AUC (hour × ng/mL) (±SD) was thus 3635 (+1733) for (R)-verapamil and 853 (+386) for (S)-verapamil.9 This high variability is probably the main reason for the variable doses needed in cluster headache.
As a lipophilic substance, verapamil easily crosses the blood–brain barrier. Verapamil is, however, a substrate for the efflux transporter P-glycoprotein (P-gp) in the blood–brain barrier.14 The P-pg restricts net brain uptake of verapamil by immediately transporting it out of the brain. The effect of the P-gp on verapamil uptake in the brain showed only minor variability.15 Thus, the mean ratio of AUCbrain/AUCblood of 11C-radioactivity was 0.42 (range 0.30 to 0.55) when 11C-verapamil was used in a PET investigation.15 Similarly, the function of the P-gp as measured by the cerebral volume of distribution (Vd) of (R)-[11C]verapamil was apparently relative invariant (0.6 to 0.85).14 These investigations14,15 were extensive in small numbers of subjects and do not exclude some variability because of genetic disposition in the general population.
The capacity of the P-pg efflux system is demonstrated by the results with [11C]verapamil in mdr1a/b double-knockout mice with complete absence of the P-pg protein.16 In these mice, 7% of the dose of [11C]verapamil was found in the brain after 60 minutes whereas in the wild-type mice, it was only 0.4% showing the key role of the P-pg transport system in verapamil brain clearance.17 There are no data on the ability of norverapamil to cross the blood–brain barrier.
High dose of verapamil has been reported to result in nonlinear pharmacokinetics because of partial saturation of presystemic metabolism of the drug.17-19 In contrast, the P-gp efflux function in brain is most likely not saturable with even the higher doses used in cluster headache. There are no primary central nervous system (CNS) symptoms indicating an action on CNS with verapamil overdose (up to 4 g). The toxic effects are cardiovascular with severe bradycardia and hypotension.20
All Ca2+ blockers are metabolized to less active metabolites in the liver by oxidative pathways, predominantly by cytochrome P-450 CYP3A, a subgroup of the cytochrome P-450 enzyme family, and lesser extent by other members of this enzyme family. And it was recently shown that the pharmacokinetics and pharmacodynamics of verapamil in healthy volunteers depend on the cytochrome P450, and the CYP3A5 genotype had a higher oral clearance of the drug.21 However, there was no association between the genotypes of CYP3A5 and the antihypertensive effect of verapamil.22
An important feature of the Ca2+ blockers is that age decreases the clearance of the drugs.23 This should be considered by the clinician as AV block and sick sinus-node syndrome likewise increases with age.
PHARMACODYNAMICS OF VERAPAMIL
Verapamil is an L-type calcium channel antagonist with anti-arrhythmic (PR-interval prolongation),24 angina, and antihypertensive effects. Its mode of action in cardiovascular diseases is somewhat well elucidated.7 In contrast, its mode of action in cluster headache is unknown.3
The enantiomers, (R)- and (S)-verapamil, have different pharmacodynamic properties. Thus, (S)-verapamil was 20-fold more potent than (R)-verapamil for prolongation of the PR interval in healthy volunteers.6 In contrast, (R)-verapamil caused a decrease in mean blood pressure.6
It is not fully elucidated why there is a tissue-selectivity to Ca2+ channels blockers, but there are experimental data that may explain this phenomenon. Verapamil and the other calcium antagonists are all known to block calcium channels other than the L-type calcium channel.25-27 The potency of differential effect on other Ca2+ channels may in part explain the tissue selectivity of the drugs. Dihydropyridine Ca2+ channel blockers also differ in their selectivity toward L- and P/Q-type Ca2+ channels. Thus, amlodipine and cilnidipine inhibit both L- and P/Q-type Ca2+ channels, whereas nifedipine and nitrendipine are selective toward L-type calcium channels.28 Furthermore, in vitro, several calcium antagonists (eg, nifedipine) bind with some selectivity to the L-type calcium channel in blood vessels, whereas verapamil binds equally well to cardiac and vascular L-type calcium channels.29In vitro, all Ca2+ channel blockers depress sinus-node activity and slow atrioventricular conduction;1 yet, only verapamil and diltiazem delay atrioventricular conduction or cause sinus-node depression at doses used clinically. This may explain their use in supraventricular arrhythmia.
Not only is there a large variability in the effect of the different drugs on the same person, but likewise, the response with the same drug in a cohort is widely variable. Thus, a study found that in individuals with hypertension systolic blood pressure response to verapamil ranged from a 33 mmHg decline to a 4 mmHg increase, with an average 12 mmHg decrease.30 Therefore, a genetic component to this variability likely exists. In accordance with this hypothesis, it was observed that the large-conductance and voltage-dependent potassium (BK) channel beta 1 subunit gene, KCNMB1, influences the responsiveness to verapamil SR and risk of adverse cardiovascular outcomes.31
COMPARISON OF DOSES OF VERAPAMIL USED IN CARDIOVASCULAR DISEASES AND CLUSTER HEADACHE
In the very large (n = 22,576) INVEST study on hypertension in patients with coronary artery disease, the starting dose of verapamil was 240 mg as a slow release formulation and the maximum dose was 180 mg twice daily.32 In cardiology, the usually used dose is 240 mg and the maximum dose of verapamil is 480 mg per day, whereas the maximum dose of verapamil in the treatment of cluster headache is 720-960 mg.2 A few cluster headache patients need up to 1200 mg daily.2,3 In one series of 108 cluster headache patients, age between 17 and 69 years, the mean daily dose was 587 mg (range 240 to 1200 mg).2
The “mean neurological dose” is thus approximately twice “the mean cardiovascular dose.”
The difference in mean doses could be explained by the fact that the cardiovascular effects are directly related to blood levels of verapamil, whereas the “neurological” effect (effect on cluster headache) takes place across the blood–brain barrier where P-gp restricts the access of verapamil to the brain.
CLINICAL USE OF VERAPAMIL IN CLUSTER HEADACHE
In patients without previous experience with verapamil, it is not possible to predict the optimal dose of verapamil. In this situation, the clinician should start low and increase the dose every second week until the optimal dose is reached. For use of verapamil in cluster headache, see the Table.
Table Table.—. Clinical Use of Verapamil in Cluster Headache
|Increase of dose depending on effect and side effect||Increments with 80 mg every 2 weeks with ECG control|
|Usual maximum dose||720 mg† daily-960 mg† (increments from 720 mg with 40 mg weekly with ECG control)|
|Unusual supra-maximum doses||1200 mg† daily (increments from 960 mg with 40 mg weekly with ECG control)|
|Some clinically relevant drug interactions (see text)||Midazolam‡,33 buspirone‡,34 simvastatin‡,35,36 carbamazepine‡,37, 38 atorvastatin§,39 erythromycin¶,40 rifampin††41|
|Side effects6,29||Constipation, dizziness, crural edema, impotence, PR-interval prolongation, and syncope|
The treatment of cluster headache should be evidence-based and the starting dose should be the 360 mg effective in a randomized clinical trial.4 The dose can then be increased with 80 mg every second week with electrocardiograph (ECG) control2 until 720 mg. The recommended clinical dose is 480 mg and if this dose is exceeded, this fact needs to be explained to the patient and informed consent obtained. A few patients may need 960 or 1200 mg. Because of the possible partial saturation presystemic metabolism of verapamil17-19 with possible nonlinear pharmacokinetics, weekly increments with 40 mg with ECG control could be a cautious option. In patients with previous experience with verapamil in previous cluster periods, the start dose could be the previously titrated dose.
In clinical practice, drug interactions with verapamil may be a problem, especially in chronic cluster headache patients, who can be on verapamil for many years. The possible interactions of verapamil, a CYP3A4 inhibitor, are legion.42 In addition, the transport of verapamil by the P-gp at the blood–brain barrier may theoretically be the site of drug interactions .43-45 In addition, verapamil is a P-gp inhibitor.45,46
Among the many possible interactions, the following interactions are clinically relevant: verapamil administration can inhibit the metabolism of midazolam,33 buspirone,34 simvastatin,35,36 atorvastatin,39 and carbamazepine.37,38 The doses of these drugs should be adjusted or alternative drugs should be used. For example, pravastatin, which is not the subject of cytochrome P450 3A4 enzyme metabolism, can be used instead of simvastatin as an lipid-lowering agent when the patients are treated with verapamil.35 Atorvastatin inhibits the metabolism of verapamil resulting in higher blood levels of verapamil,47 and with the high doses of verapamil used in cluster headache, this combination should be avoided.
A pharmacodynamic interaction between erythromycin and verapamil with prolongation of the PR interval has been described in rats.48 One case of total AV block caused by concomitant use of verapamil and erythromycin has been described.40 Erythromycin has been associated with increased risk of sudden death from cardiac causes especially in those who concurrently used CYP3A inhibitors.49 Erythromycin should therefore not be used together with verapamil.
A drug interaction between verapamil and lithium with profound bradycardia, followed by a fatal myocardial infarction in one case, has been described.50 These 2 drugs are often used concomitantly in chronic cluster headache and these patients should be controlled carefully with control of plasma levels of lithium and with ECG. Neurotoxicity of the combination of lithium and verapamil with normal plasma levels of lithium has, however, been described in one case.51
Rifampin is a potent inducer of the cytochrome P450 oxidative enzymes and of the P-gp transport system.41 Concomitant use results in decreased biovailability and plasma levels of verapamil.41
Finally, grapefruit juice should be avoided during verapamil treatment because it increases plasma level of verapamil.52 In contrast, smoking decreases plasma level.52
WHERE AND HOW DOES VERAPAMIL WORK IN CLUSTER HEADACHE?
Some of the most striking features of episodic cluster headache are the circadian rhythmicity of the attacks: 50% of attacks occurring during the night.53 There is also a seasonal variation,54 and a relapsing-remitting course. Cluster headache is associated with abnormal diurnal rhythms of hormones such as cortisol, testosterone, growth hormone, and endorphins.55,56 The clinical course and the hormonal alterations have pointed to hypothalamus as the center of origin in cluster headache.55,56 In addition, PET studies have shown activation in the posterior hypothalamic gray matter in patients with attacks of cluster headache.57 With voxel-based magnetic resonance imaging morphometry, increased neuronal density was found in the same hypothalamic region.58 Deep brain stimulation in this region has been used with success in chronic intractable cluster headache patients.59
The central biological clock of the mammalian brain is located in the suprachiasmatic nucleus (SCN) and in rats a day–night difference in the L-type Ca2+ current in SCN that was sensitive to nimodipine was observed.60 Could verapamil exert an effect of SCN in humans? The drug has, however, been in use in medicine for decades and it is unlikely that an effect on the biological clock would be unnoticed as an adverse event during this extensive use. In addition, the concentration of nimodipine used in this60 and in another study61 on SCN was 2 µM and therefore suprapharmacological. This corresponds to 836 ng/mL (MW = 418 g), whereas Cmax after nimodipine 60 mg was 10 ng/mL62 and moreover the protein binding in plasma is greater than 95%. The used dose in vitro of nimodpine60,61 is thus more than 100 times greater than the free drug concentrations in vivo. Furthermore, the permeability of nimodipine into the brain is restricted by P-gp.63,64 So, whether nimodipine can exert an effect on SCN in man remains an open question.
Synaptic transmission is dependent upon the entry of Ca2+ through presynaptic voltage-activated Ca2+ channels.26 In central mammalian neurons, one low (T-type) and at least 6 high voltage-activated Ca2+ channels (L-, N-, O-, P-, Q-, and R-type) have been identified.26,65
In one study in the rat striatum, the authors found that verapamil blocked P- and at higher concentrations possibly N- and Q-type Ca2+ channels, whereas diltiazem appeared to block only P-type Ca2+ channels.26 The human T-type calcium channel is also blocked by verapamil.66
Both nimodipine and diltiazem showed memory-enhancing effect in mice.67 In contrast, in this study68 and in several other studies, verapamil had no effect on learning of different tasks.68 These results do of course not explain the effect of verapamil but demonstrate merely in principle that among the L-type Ca2+ channel antagonists there are differences in their effect on the CNS in animals. The effect of verapamil could be due to its additional effect on other Ca2+ channels or to differences in binding to the L-type Ca2+ channels Cav1.2 and Cav1.3, which are the predominant brain L-type Ca2+ channels,69 including the anterior hypothalamus,70 paraventricular and ventromedial nuclei.71 T-type,28 P/Q-type,25,27 and N-type72 calcium channels are also present in the hypothalamus.
The 3 major classes of calcium blockers, dihydropyridines, diltiazem, and verapamil, have separate binding sites, but overlapping or allosterically linked, to the α1c subunit of the Cav1.2 Ca2+ channel.73 The dihydropyridines binding site is near the opening of the channel, the diltiazem site in the middle, and the verapamil site is near the central pore of the channel.74
Finally, it should be noted that the effect of verapamil is not restricted to calcium channels. Verapamil inhibits various potassium channels.75 The voltage-dependent K+ channels belonging to the ether-a-go-go family are widely distributed in the mammalian CNS.76,77 Verapamil is an antagonist at the human ether-a-go-go-related gene (HERG or KCNH2) potassium channel whereas nifedipine has no blocking effect78 and verapamil has been used in cardiovascular research as an HERG potassium channel blocker.75,79 It is possible that verapamil also acts on HERG potassium channels in the CNS.
In conclusion, the possible modes of action of verapamil in cluster headache are legion but most likely due to an effect on either the low (T-type) or the high voltage-activated Ca2+ channels (L-, N-, P, Q-type). Although generally described as an L-type calcium blocker, verapamil has possible effects on a range of other Ca2+ channels present at the most likely pathogenic center, the hypothalamus. At this point, we are thus not able to point to a specific action underlying the efficacy of verapamil in cluster headache.