Abstract
Tonabersat is a novel benzopyran derivative that blocks the cortical spreading depression proposed to be associated with migraine attacks. The ability of single oral doses of 15, 25, 40 and 80 mg of tonabersat to relieve the symptoms of moderate to severe migraine was evaluated in 859 migraineurs enrolled in two dose-ranging, double-blind, randomized, placebo-controlled, parallel-group trials, one international and the other North American. In the international study, significantly more patients given tonabersat than given placebo experienced relief of headache pain at 2 h (15 mg, 36.8%; 40 mg, 40.7%), the principal efficacy variable, and at 4 h (40 mg, 63.0%) and complete abolition of headache at 4 h (40 mg, 34.3%). None of the primary or secondary efficacy variables indicated significant differences between tonabersat and placebo in the North American study. Tonabersat was generally well tolerated, with dizziness and nausea the most common side-effects. Serious adverse events were uncommon, and no patient withdrew from either study because of adverse events. These results suggest a possible interplay between tonabersat pharmacokinetics (the relatively long time required to reach maximum plasma concentrations) and patient characteristics (previous triptan exposure) in the management of acute migraine attacks. Based on the pharmacokinetics and actions on cortical spreading depression, tonabersat may have potential value in migraine prophylaxis.
Introduction
Migraine is a common, recurrent, often disabling, primary headache disorder. The International Headache Society (IHS) classifies migraine as with aura and without aura (1). In migraine without aura (MoA), the more common subtype, attacks last 4–72 h and the headaches are typically unilateral and pulsating, moderately to severely intense, aggravated by routine physical activity and associated with nausea and/or photophobia and phonophobia (1). Migraine with aura (MA) affects 15–20% of sufferers; the attacks of reversible, focal neurological symptoms generally occur gradually over 5–20 min and persist for < 60 min (1, 2). Typical aura may include flickering lights, spots or lines or loss of vision, a pins-and-needles sensation or numbness and dysphasic speech disturbances (1).
The debilitation resulting from migraine encompasses personal suffering, absence from the workplace, restrictions on daily activities, and an overall decreased quality of life. The present treatment options for acute migraine attacks include simple or combination analgesics and specific treatments, such as ergots or triptans (3, 4). Due to the limitations of currently available agents, there is a need for additional and alternative treatment options. Although triptans can effectively relieve migraine symptoms, not all patients benefit from treatment (5) and therefore drugs with a different mode of action are needed. Furthermore, triptans and ergots have vasoconstrictive properties and are contraindicated for migraine patients with vascular comorbidities, such as coronary heart disease, transient ischaemic attack or stroke or uncontrolled hypertension (6). For these patients, an effective drug without vasoconstrictive properties is needed.
Present acute therapies were developed with the aim of blocking pain transmission, inhibiting inflammation and eliciting vasoconstriction. At the time of their development, the interaction between cortical spreading depression (CSD), the physiological equivalent of the human aura, and meningeal vasodilation and increased blood flow was not yet known (7, 8). However, evidence from neuroimaging studies shows that CSD is responsible for migraine aura, and experimental studies have demonstrated that CSD is also responsible for the activation of the trigeminovascular system and possibly migraine headache (9, 10). Tonabersat, a novel benzopyran derivative, inhibits CSD (11) and therefore might be able to inhibit the early migraine mechanisms. In addition, tonabersat causes no vasoconstriction in isolated human blood vessels (12).
Relief of the symptoms of acute migraine with single oral doses of tonabersat was evaluated in two randomized controlled studies—one conducted in Europe, Australia and South Africa and one in Canada and the USA. Based on similarities in design, the results of both studies are presented here.
Patients and methods
Patients
In the two studies, subjects were eligible men and women aged 18–65 years, with migraine attacks for at least 1 year before screening and one to six attacks per month, with at least 48 h between attacks, during the previous 6 months, and were < 50 years old when migraine attacks began. The IHS criteria for MA and MoA were used to establish the diagnosis. Women of childbearing potential were enrolled only if they had been using a reliable method of contraception; oral contraceptive users had to have been taking the pill for a minimum of 3 months before screening.
Exclusion criteria included pregnancy or risk of pregnancy and breastfeeding; non-migrainous headaches requiring analgesic medication on > 10 days/month during the 6 months before screening; diagnosis of both migraines and another headache that the patient could not reliably distinguish from migraine; history or current evidence of alcohol or drug abuse; misuse or overuse of analgesic, opioid and/or antimigraine treatments within the past 3 months (treatment for > 3 days/week or, in the case of ergotamine, on more than two occasions per week); clinically significant cerebrovascular or cardiovascular disease; clinical diagnosis of a major depressive disorder or schizophrenia; consistent failure to respond to conventional acute-migraine therapies; vertebrobasilar or hemiplegic migraine; renal dysfunction (serum creatinine level > 125% of the upper limit of normal for a patient's age group); hepatic dysfunction (liver function test result more than twice the upper limit of normal for a patient's age group); participation in more than two clinical studies of investigational drugs within the past year; use of an investigational drug within 30 days or five half-lives before screening; and other clinically significant conditions that, in the investigator's opinion, rendered a patient unsuitable for the study. All patients provided written informed consent for participation. The protocol and statement of informed consent were approved by an institutional review board or ethics committee at each site. The studies were not submitted to clinical trial registries, as the US registry was not established when the trials were conducted in 1997.
Study design
Both dose-ranging studies used a randomized, double-blind, placebo-controlled, parallel-group design. At the initial screening visit, a medical history was obtained; a physical examination, including full neurological assessment and measurement of vital signs, was conducted; and a 12-lead electrocardiogram (ECG) was performed. Samples were collected for routine haematology, blood chemistry and urinalysis. Women of childbearing potential underwent a urine dipstick pregnancy test.
Eligible patients were randomized to treatment using a computer-generated list. Investigator-allocated treatment packs were given to patients in strict sequential order. The master randomization list was held by the sponsor (SmithKline Beecham), and individual sealed code envelopes were held by the investigator or pharmacist. Each study required the enrolment of 420 patients. Assuming a 30% attrition rate, resulting in 288 evaluable patients per study, each study would be projected to have at least 90% power to detect a difference of 30% between groups in the proportion of patients with headache relief 2 h post treatment.
Patients took a single oral dose of 15 mg (international trial only), 25 mg (North American trial only), 40 mg or 80 mg of tonabersat or matching placebo at the onset of moderate (grade 2) or severe (grade 3) headache symptoms associated with a migraine attack. The time between onset of the migraine or waking with a headache and treatment with study medication was not to exceed 6 h, and the medication was to be used within 8 weeks of screening. Patients who had taken ergotamine, dihydroergotamine or sumatriptan within 48 h of the evaluable attack; a simple analgesic medication, such as paracetamol or a non-steroidal anti-inflammatory drug, within 6 h prior to the attack; or medication specifically for migraine prophylaxis within 1 month prior to the attack were not to initiate treatment with the study medication.
Patients recorded the following on diary cards: details of migraine onset, headache severity and the presence of nausea, vomiting, photophobia and phonophobia before treatment and at 1, 2 and 4 h afterwards; time of treatment with the study medication; use of rescue medications, time when meaningful relief was obtained; and occurrence of adverse events (AEs). Twenty-four hours after treatment, patients noted any recurrence of headache symptoms, assessed the effectiveness of the study medication and indicated whether they would take it again. The diary cards were returned at a follow-up visit 7–10 days post treatment.
The primary efficacy measure was headache relief 2 h after treatment, defined as the proportion of patients whose headache intensity changed from grade 2 or 3 to grade 0 (none) or 1 (mild). Secondary efficacy measures included headache relief at 1 and 4 h, complete abolition of headache at 1, 2 and 4 h, headache recurrence at 2–24 and 4–24 h post treatment, time to headache recurrence, use of rescue medication within 2 h post treatment and time to meaningful headache relief. Additional efficacy measures were the incidence of nausea, vomiting, photophobia and phonophobia within 1, 2 and 4 h post treatment, use of rescue medication during the 24 h post treatment, patient evaluation of study medication and the proportion of patients willing to take study medication again. AEs and results of vital signs, ECGs and laboratory tests were used to assess safety at the post-treatment follow-up visit. The relationship of the study drugs to AEs was judged by the investigators as not related, unlikely, suspected (reasonable possibility) or probable. Those AEs judged as suspected or probable were considered treatment related.
Efficacy variables were analysed by means of logistic regression. Covariates included sex, headache severity before treatment, duration of headache before treatment, age, type of migraine (with or without aura) and typical number of attacks per month. Two-tailed statistical tests were used to compare each dose of tonabersat with placebo, and the modified Bonferroni procedure for multiple comparisons was used to determine whether comparisons of individual dose levels of tonabersat with placebo were statistically significant at P < 0.05. For the primary efficacy variable, a dose–response relationship was investigated by means of the Cochran–Mantel–Haenszel test and was considered statistically significant at P < 0.05.
Results
Patients
Patient enrolment began in June 1997, and the last study visit was in October 1997. Of 549 patients screened at 35 centres in Australia, South Africa, Belgium, France, Germany and the UK in the international study, 525 were randomized and 441 received study medication and formed the intention-to-treat (ITT) population (Table 1). In the North American study, 534 patients were screened at 29 centres in Canada and the USA and 506 were randomized; the ITT population comprised 418 patients who took study medication (Table 2). The most common reason for exclusion from the ITT population in both studies was failure to take study medication either because a migraine meeting the protocol-defined criteria did not occur within the specified time or did occur but the patient did not initiate treatment. All but three patients taking tonabersat in the international study and four taking tonabersat and one taking placebo in the North American study were included in the ITT analysis.
Disposition of patients in the international study
Disposition of patients in the North American study
∗Did not meet migraine criteria for treatment (n = 42); did not take study medication (n = 17); various reasons (n = 14).
Baseline demographic characteristics and migraine history were generally similar among treatment groups within each study and, with the exception of migraine type and previous sumatriptan use, across studies (Table 3). In both studies, most patients were white and female; the mean age of patients ranged from 39.4 to 40.5 years. The mean age when migraine attacks began ranged from 19.6 to 21.6 years, and the mean duration of migraine history ranged from 18.3 to 20.3 years. Most patients in both studies had MoA. The mean number of migraine attacks per month in the two studies ranged from 2.6 to 3.2. Approximately two-thirds of North American study participants had previously used sumatriptan, compared with about one-quarter of international study patients. Overall, no marked differences were evident among treatment groups in the demographic profile or migraine history in either study; however, the proportion of North American patients with aura was approximately twofold higher in each of the tonabersat groups than in the placebo group.
Demographic characteristics and migraine history
∗Data are presented as mean (
Before treatment, headaches lasted a mean of 2–3 h. Intensity was moderate for approximately 55% of patients in the international study and severe in about 45%; respective estimated values in the North American study were 66% and 36%. The most common associated symptoms before treatment were phonophobia and photophobia; < 10% of each study population reported vomiting. In both studies, characteristics of the evaluable headache generally were balanced among treatment groups.
Efficacy of tonabersat
Primary efficacy variable
In the international study, 36.8, 40.7 and 29.9% of patients who received tonabersat 15, 40 and 80 mg, respectively, experienced headache relief at 2 h, compared with 21.3% of patients who received placebo (Fig. 1). Tonabersat doses of 15 mg (P = 0.013) and 40 mg (P = 0.002) provided significant relief compared with placebo in this evaluation. The odds ratio of obtaining relief with these two doses was more than twice that with placebo. The proportion of patients with relief at 2 h did not differ significantly between tonabersat 80 mg and placebo (P = 0.150). No significant association was apparent between the tonabersat dose and headache relief at 2 h (P = 0.130).
Proportion of patients treated with tonabersat and placebo who experienced headache relief at 2 h in the international and North American studies. ∗P = 0.013 vs. vehicle (modified Bonferroni procedure). †P = 0.002 vs. vehicle (modified Bonferroni procedure).
Treatment centre and pretreatment headache intensity significantly affected the likelihood of relief at 2 h in the international study. Patients enrolled outside France were less likely to have relief than those enrolled at French centres, regardless of treatment (P = 0.025), and patients with severe headaches were less likely to obtain relief than those with moderate headaches (P = 0.009).
In the North American study, similar percentages of patients reported headache relief at 2 h with tonabersat 25 mg (29.3%) and 40 mg (26.9%) and with placebo (24.0%). The incidence was higher with tonabersat 80 mg (33.6%) than with placebo, but the between-group difference did not reach statistical significance (P = 0.11).
In both studies, exposure to sumatriptan appeared to affect results for the primary efficacy variable. In the international study, the proportions of the placebo and tonabersat 15- and 40-mg groups with headache relief at 2 h were higher among sumatriptan-naive patients (placebo, 29.7%; tonabersat 15 mg, 39.7%; tonabersat 40 mg, 45.1%; tonabersat 80 mg, 31.1%) than among those with prior exposure (placebo, 0%; tonabersat 15 mg, 25.0%; tonabersat 40 mg, 33.3%; tonabersat 80 mg, 33.3%). This association was especially marked in the placebo group: no previous sumatriptan users reported relief, compared with 29.7% of those naive to sumatriptan. In the North American study, response rates with tonabersat 25 and 80 mg were approximately twofold higher in patients without previous sumatriptan use (placebo, 24.2%; tonabersat 25 mg, 43.8%; tonabersat 40 mg, 20.5%; tonabersat 80 mg, 47.6%). Patients in the 40-mg tonabersat group, however, were more likely to have their headaches relieved if they had a history of sumatriptan exposure (placebo, 24.2%; tonabersat 25 mg, 22.4%; tonabersat 40 mg, 31.2%; tonabersat 80 mg, 24.6%). Response rates with placebo were identical regardless of previous sumatriptan use.
Secondary efficacy variables
Among patients in the international study, higher percentages treated with tonabersat 40 mg (23.9%) and 80 mg (21.3%) reported headache relief at 1 h, compared with patients receiving tonabersat 15 mg or placebo (∼17% in each group). At 4 h, rates were 56.2–63.0% with tonabersat and 49.0% with placebo; the difference between the tonabersat 40-mg group (63.0%) and placebo was statistically significant (P = 0.022) (Table 4). The analysis of headache relief at 1 and 4 h demonstrated a treatment-by-centre interaction. The 1-h incidence with placebo was higher at French centres than at non-French centres (23.7% vs. 8.2%), although response rates in the tonabersat groups were similar at French and non-French centres. Compared with placebo, tonabersat 40 mg (P = 0.005) and 80 mg (P = 0.004) relieved headache at 1 h at non-French centres, but rates did not differ significantly between any of the tonabersat doses and placebo at French centres. Four hours after treatment, the placebo response rate was slightly higher at French than at non-French centres (52.7% vs. 44.7%). Compared with placebo, tonabersat 80 mg relieved headache at 4 h at non-French centres (44.7% placebo, 69.4% tonabersat; P = 0.009), but no tonabersat dose was significantly different from placebo at any French centre.
Secondary efficacy measures—international study
∗ P = 0.022 (modified Bonferroni procedure).
† P = 0.009 (modified Bonferroni procedure).
In the international study, generally few patients reported complete abolition of headache at 1 and 2 h, with no significant difference between placebo and any of the tonabersat doses. Four hours after treatment, the proportions were slightly higher in the tonabersat groups (23.8–34.3%) than in the placebo group (20.6%), and the difference between placebo and tonabersat 40 mg reached statistical significance (P = 0.009) (Table 4).
Other secondary efficacy variables did not differ significantly between tonabersat and placebo in the international study (Table 4). None of the secondary efficacy variables demonstrated significantly greater improvement with any tonabersat dose compared with placebo in the North American study (Table 5).
Secondary efficacy measures—North American study
Safety
In the international study, 23.9, 37.4 and 33.9% of patients who received tonabersat 15, 40 and 80 mg, respectively, reported AEs, compared with 25.0% of patients given placebo. AEs occurring in at least 5% of any group were dizziness, nausea, vertigo, vomiting, somnolence and abdominal pain. Overall, 15.6, 27.0 and 20.2% of patients who received tonabersat 15, 40 and 80 mg, respectively, compared with 13.0% of patients given placebo reported AEs that were suspected to be, or probably were, related to study medication. Treatment-related AEs included dizziness, nausea, vertigo, somnolence, abdominal pain and palpitation (Table 6). The incidences of dizziness and nausea were similar among patients who received tonabersat 40 and 80 mg but were higher than placebo. The incidence of somnolence was noticeably higher in patients who received 40 mg of tonabersat compared with placebo.
Adverse events (AEs) in > 3% of patients in any treatment group suspected or considered to be probably related to treatment—international study
In the North American study, 21.6, 29.2 and 45.0% of patients who received tonabersat 25, 40 and 80 mg, respectively, reported AEs, compared with 25.7% of patients given placebo. AEs occurring in at least 5% of any group were dizziness, nausea and somnolence. Overall, 12.7, 22.6 and 33.0% of patients who received tonabersat 25, 40 and 80 mg, respectively, compared with 22.8% of patients given placebo reported AEs that were suspected to be, or probably were, related to study medication. Treatment-related AEs included dizziness and nausea (Table 7). The incidences of dizziness and nausea were notably higher in patients who received 80 mg of tonabersat compared with placebo. No withdrawals from either study were attributable to AEs.
Adverse events (AEs) in > 5% of patients in any treatment group suspected or considered to be probably related to treatment—North American study
Serious AEs—defined as fatal, life-threatening, disabling or incapacitating, or that resulted in hospitalization, prolonged hospital stay or were associated with congenital abnormality, cancer or overdose—occurred in two patients in the two studies combined. A North American patient was hospitalized for muscle spasms before receiving study medication, and one international patient required surgery for a prolapsed disc after an accident. No events were judged to be related to the study medication. No patients died during the study or within 30 days of its completion.
Laboratory values and vital signs generally remained within the normal range and did not differ in the tonabersat and placebo groups. No significant ECG changes were detected internationally. Each of two North American patients had a significant ECG change 9 days after taking tonabersat 25 mg: (i) a long QT interval, considered unlikely to be related to the study drug, and (ii) a T-wave inversion suspected to be due to tonabersat. In a third patient treated with tonabersat 80 mg, T-wave abnormalities identified at screening showed no significant change at follow-up.
Discussion
In the international study, tonabersat was significantly more efficacious than placebo in relieving moderate to severe migraine pain 2 h after treatment (15 and 40 mg), in relieving headache pain at 4 h (40 mg) and in completely abolishing headache at 4 h (40 mg). Efficacy was not demonstrated for any of these variables in patients who received tonabersat 80 mg; the reason for the lack of efficacy with this higher dose is unknown. A treatment-by-centre interaction for headache relief was evident: significant differences between placebo and tonabersat 40 and 80 mg at 1 h and tonabersat 80 mg at 4 h at non-French but not French centres. These results could be due to a higher placebo response in patients at French centres than in those at non-French centres.
Headache relief at 2 h demonstrated no statistically significant differences between any tonabersat dose and placebo in the North American study. Similarly, none of the tonabersat doses improved any of the secondary efficacy variables significantly better than did placebo.
Prior exposure to sumatriptan may have affected the response to tonabersat in both studies. Approximately two-thirds of North American enrollees had taken sumatriptan in the past. Headache relief at 2 h with tonabersat 25 and 80 mg was approximately twofold higher in sumatriptan-naive patients than in previous sumatriptan users. Conclusions about the 40-mg dose are difficult to draw because of the small number of patients. Approximately one-quarter of the international enrolment had used sumatriptan, and these patients posted lower rates of headache relief at 2 h with placebo and with tonabersat 15 and 40 mg. The expectation of a triptan-like effect in patients with a sumatriptan history may have masked a therapeutic response to a novel agent that did not produce such an effect.
Pharmacokinetic properties of tonabersat also may have affected efficacy. The median time to the peak plasma concentration (T max) reported in healthy individuals after single oral doses of 10–80 mg of tonabersat ranged from 2–3 h. The median T max during a migraine attack was 3 h (unpublished data). Delayed absorption in patients with migraine, presumably due to slowed gastric emptying, has been reported in migraineurs during an attack (13). If higher plasma tonabersat concentrations are efficacious, the relatively long T max of tonabersat may explain the low rate of headache relief in these studies.
Tonabersat was well tolerated. The incidence of AEs in patients who received tonabersat 80 mg, including AEs suspected or considered to be related to treatment, was higher in the North American study than in the international study. Specifically, dizziness and nausea occurred with greater frequency among patients who received tonabersat 80 mg in the North American study. The reason for this increased incidence is unknown. With the exception of the reports of dizziness and nausea in the North American study and somnolence in the international study, no single AE was reported by > 6.6% of patients in either study. Only two serious AEs occurred, neither of which was judged to be related to the study medication. Just two ECG changes of questionable significance were reported in the two trials combined.
The findings of these two studies may indicate a complex interplay between patient characteristics (previous triptan exposure) and the pharmacokinetics and dose of tonabersat that could account for the mixed results with tonabersat in the acute treatment of migraine. The slow absorption of tonabersat will probably rule out its use in acute treatment; however, this would not be a limitation for prophylactic use. Tonabersat inhibition of CSD, an event proposed to be associated with the initiation of migraine, also suggests that tonabersat may be effective in migraine prophylaxis (11).
The efficacy of tonabersat for the prevention of migraine was examined in a 3-month Phase 2a study (14). For the primary end-point, patients reported a non-significant 1-day reduction in the number of migraine days between the third month of treatment and the baseline period. Significant differences between tonabersat and placebo were demonstrated for two out of 10 secondary efficacy end-points. In month 3, more patients receiving tonabersat (62%) than receiving placebo (45%) reported a > 50% reduction in migraine attacks, and rescue medication use was reduced by 1.8 days in patients who received tonabersat when compared with placebo. No significant between-group differences were found for the other eight secondary end-points. The utilization of tonabersat for the prevention of migraine is being further examined in two ongoing studies.
Study investigators
International study
North American study
Footnotes
Competing interests
S.D.S. is a paid consultant to Minster Pharmaceuticals and owns stock in the company; J.S. has received an honorarium from Minster Pharmaceuticals; H.G. has received funding for research carried out in this work; A.H.E. has conducted clinical trials with tonabersat. H.C.D., J.A.K. and R.A.H. declare no conflicts of interest.
Acknowledgements
K. Skolnik, P. Kontur, PhD and M. Kersting, PhD (JL Shapiro Associates Inc., Edison, NJ, USA) assisted with manuscript preparation with financial support from Minster Pharmaceuticals.