A placebo-controlled crossover study of rizatriptan in the treatment of multiple migraine attacks

Migraine affects 13 to 18% of women and 3 to 6% of men, with peak prevalence between 35 and 45 years of age.^1-4 Although there is considerable variation in the severity and frequency of migraine attacks both between patients and within individuals, more than half of all migraineurs have restricted significantly their work or social life.⁵ The exact pathophysiology of migraine remains poorly understood, but numerous observations have implicated the neurotransmitter serotonin (5-HT).^6-8 The serotonin (5-HT_1B/1D) receptor agonist is associated with antimigraine activity in humans.⁹ Sumatriptan, a 5HT_1B/1D receptor agonist, appears to maintain its efficacy when it is used in the acute treatment of a series of migraine attacks.^10,11

Rizatriptan is a potent 5HT_1B/1D receptor agonist that is absorbed rapidly in both healthy volunteers¹² and during an acute migraine attack. The efficacy and tolerability of rizatriptan are dose related,^13,14 with rizatriptan 10 mg considered to be a highly effective dose and 5 mg also effective but somewhat less active than 10 mg. Rizatriptan 2.5 mg was determined to be a no-effect dose. An additional study confirms the consistent efficacy and safety of rizatriptan 5 and 10 mg in the treatment of migraine and its recurrences within single attacks.¹⁵

The efficacy and safety of new migraine treatments are usually evaluated in patients undergoing a single attack.^13-17 Because many migraine patients have on average one or more attacks per month, effective medications should be evaluated rigorously during repeated use. Despite this, few studies have reported on the safety and efficacy of new antimigraine medications over a series of attacks.^11,18,19 A previously reported placebo-controlled study of sumatriptan in the treatment of three discrete migraine attacks¹¹ suggested that a study in multiple attacks was feasible. We report the results of a study on the safety and efficacy of rizatriptan 10 mg in the acute treatment of four migraine attacks and their recurrences. The hypotheses of this study were 1) rizatriptan 10 mg PO would be superior to placebo with regard to the proportion of responders at 2 hours posttreatment for the first attack, 2) rizatriptan 10 mg PO would be superior to placebo with regard to the proportion of responders at 2 hours posttreatment for multiple migraine attacks, and 3) rizatriptan 10 mg PO would be safe and well tolerated after multiple treatments in patients with migraine.

Methods. Study procedures. All patients gave written informed consent before entering the study, and the protocol was approved by local investigational review boards of the 23 investigative sites that participated in the study. A total of 473 male and female patients 18 to 65 years of age, who have from one to eight migraines per month according to International Headache Society criteria²⁰ with at least a 6-month history of migraine but otherwise in good health, were enrolled at 23 study sites in the United States. Before patients were randomized, a medical history was elicited and a physical examination, electrocardiogram(ECG), and laboratory evaluations (blood chemistry, hematology, urinalysis, and pregnancy test in fertile women) were performed.

All clinical supplies consisted of identical tablets of rizatriptan 10 mg and placebo. Patients were randomized equally and individually to one of five sequence groups as follows using a computer-generated allocation schedule(table 1). Patients in the first group received placebo for the first migraine attack and rizatriptan 10 mg for the remaining three attacks. Patients in the second group received rizatriptan 10 mg for the first migraine attack, placebo for the second attack, and rizatriptan 10 mg for the third and fourth attacks. Patients in the third group received rizatriptan 10 mg for the first and second attacks, placebo for the third attack, and rizatriptan 10 mg for the fourth attack. Patients in the fourth group received rizatriptan 10 mg for the first three attacks and placebo for the fourth attack. Patients in the fifth group received rizatriptan 10 mg for all four attacks. The group that received rizatriptan for all four attacks was added to assess better the carryover effects and to evaluate the safety of the compound on repeated use.

Patients could treat up to two headache recurrences within 24 hours for each attack. The medication for recurrences was the same as that for the initial attack. Before receiving their test medication, patients were instructed on drug administration and diary completion.

Patients could not receive any other investigational compounds within 30 days before study entry. The following medications were restricted during treatment: lithium, methysergide, monoamine oxidase inhibitors, or propranolol and metoprolol (i.e., drug interactions between the latter two agents and rizatriptan were still under investigation at the time of the trial) within 2 weeks; ergot derivative, isometheptene, or sumatriptan within 48 hours; any opiate within 24 hours; or any other form of analgesia within 6 hours before treatment with any dose of study medication.

Outpatients were instructed to ingest one tablet of test medication when they developed a moderate or severe migraine headache that did not start to resolve spontaneously. Patients rated symptom severity in their diaries at the following time points: predose, and at 0.5, 1, 1.5, 2, 3, and 4 hours postdose. At the 2-hour point, patients completed a Global Question of Satisfaction with Medication questionnaire (rated on a seven-point scale). Any patient who still had a moderate or severe headache at 2 hours after taking the initial dose of study medication could take optional escape analgesia (i.e., opiates, acetaminophen, or nonsteroidal anti-inflammatory drugs). The procedure was repeated for each of the next three distinct attacks. A distinct attack was defined as a grade 2/3 migraine headache that was separated by more than 24 hours from the previous attack.

Headache recurrence was defined as the return of headache severity to grade 2 or 3 within 24 hours in a patient who had responded to rizatriptan or placebo at 2 hours. The details of any first and second recurrences (headache severity, date and time of recurrence, and response to treatment at 2 hours postdose) and adverse experiences that occurred within 24 hours of test medication were recorded in the diaries.

At the end of the 24-hour treatment period for each treated attack, patients completed a 24-Hour Migraine Quality of Life Questionnaire^21,22 and a Migraine Work Loss and Productivity Questionnaire. These data are the subject of a separate report.

Headache severity and functional disability were rated by patients at baseline and at each of the previously mentioned time points using four-point anchored scales (headache severity: 0, no headache; 1, mild headache; 2, moderate headache; and 3, severe headache; functional disability: 0, normal; 1, daily activities mildly impaired; 2, daily activities severely impaired; and 3, unable to do activities, requires bed rest). The primary end point was percentage of patients with pain relief (also termed response) at 2 hours postdose, defined as improvement of grade 3 (severe) or 2 (moderate) headache to grade 1 (mild) or 0 (pain free). The presence or absence of nausea, vomiting, photophobia, and phonophobia were also noted at each assessment time. Patients' overall satisfaction with medication was rated at 2 hours postdose on a scale of 1 to 7 as follows: 1, completely satisfied, couldn't be better; 2, very satisfied; 3, somewhat satisfied; 4, neither satisfied nor dissatisfied; 5, somewhat dissatisfied; 6, very dissatisfied; and 7, completely dissatisfied, couldn't be worse.

Patients were instructed to visit the clinic as soon as possible (usually within a few days) after treatment of the first attack to ensure that they understood the study procedures. Patients were to return to the clinic once each month for laboratory tests and monitoring of vital signs. Patients who had experienced any migraine attacks between clinic visits were asked to bring their completed diary and questionnaires, and any unopened tablet containers. Patients were instructed to report to the clinic within 7 days following treatment of the fourth (final) attack, at which time they were to return the diaries and any unused test drug. Laboratory examinations, vital signs, physical examination, and ECG were performed at the final visit.

Allocation schedule procedures. In this study, the patients, investigators, and sponsor were all blinded to treatment. The allocation schedule was generated electronically by a statistician not connected with the study, and the schedule was then assigned a unique number and was maintained by the clinical packaging department in secured electronic and paper files. The blind was maintained for all patients during this trial. The database was unblinded after the trial was completed and the following conditions were met: review of all label pages, indicating that no unblinding occurred; completion of data cleanup; and identification of protocol violators.

Data analysis. The primary efficacy end point was pain relief at 2 hours during the first attack period. The percentage of patients on rizatriptan 10 mg who had pain relief at 2 hours was compared with the percentage of those on placebo with relief at 2 hours using logistic regression. The statistical null hypothesis tested was that the odds ratio in terms of pain relief for the rizatriptan group relative to the placebo group was 1.0. Logistic regression was also performed to compare the effects of rizatriptan with placebo in terms of the percentages of patients who were pain free, who reported associated symptoms, who took escape medication at 2 hours, and who reported recurrences (secondary outcome measures). Cumulative logistic regression was used to compare treatment groups in terms of their degree of functional disability (secondary outcome measure). ANOVA was performed to compare treatment groups in terms of patients' overall satisfaction with medication (secondary outcome measure). Based on a planned sample size of 350 patients (280 on rizatriptan and 70 on placebo for the first attack) and an anticipated response rate of 50% on rizatriptan and 20% on placebo, the study had a ≥99% power to detect a difference between groups, with α = 0.05 (two-tailed test). The percentage of patients with pain relief at 2 hours over all attacks was compared between rizatriptan 10 mg and placebo using generalized estimating equation (GEE) procedure to account for intrapatient correlation resulting from the repeated migraine attack assessments. In this analysis, the logistic model included terms for treatment and attack period along with terms to evaluate a carryover effect.^23,24 The incidence of clinical adverse experiences reported during the first attack period was compared between rizatriptan and placebo using Fisher's exact test. To take into account patients' repeated exposure to rizatriptan in as many as four attacks(relative to the exposure to placebo in a single attack), the incidence of adverse experiences was adjusted. The adjusted incidence was determined by first calculating the proportion of headaches treated with rizatriptan in which an adverse experience occurred for each patient. Then, the arithmetic mean of those proportions was calculated, yielding the adjusted incidence over all attacks. A GEE modeling procedure was used to compare the total number of patients on rizatriptan and placebo who reported adverse experiences over all attacks.^23,24 The logistic model used for this analysis included the same terms as those for the analysis of pain relief. Additionally, the modeling procedure evaluated carryover effects before the assessment of treatment differences.

This approach to data analysis was to compare initially rizatriptan 10 mg to placebo for the first migraine attack, and then compare the two treatments over multiple attacks. The results have been organized to reflect this approach, with descriptions of the results for the first attack, followed by those for all attacks.

The primary efficacy analysis was completed on an "all-patients-treated" approach. This analysis included all patients who had at least one record of an efficacy measure after the initial dose (N = 402). For safety and tolerability, all patients who took study medication were included in the analysis (N = 407).

Results. Patient accounting. A total of 473 patients were enrolled at 23 study sites. Sixty-six patients did not take any test medication. Most of these patients withdrew from the study or did not have any migraine headaches within 2 months after randomization. Of the 407 patients who took study medication, 341 (84%) were women and 66 (16%) were men, with a mean age of 40.6 years. A total of 367 patients treated at least two migraine attacks, 339 treated at least three attacks, and 316 treated all four attacks. Of the 407 patients who treated at least one attack, very few patients (8) dropped out for lack of efficacy (three were on placebo at the time). In addition, only eight patients dropped out because of adverse effects (six were on placebo at the time). These patients were distributed similarly over the various treatment sequences. See thefigure for the study flow diagram.

Protocol violators were identified before the database was unblinded using the following criteria for the initial headache (each attack): baseline headache severity was not equal to 2 or 3, patient took initial tablet from wrong bottle and the bottles that were switched were rizatriptan 10 mg and placebo, patient treated attacks >24 hours apart, and patient took certain prohibited prior and concomitant medications within a certain cutoff period. Of the 402 patients included in the efficacy analysis using the all-patients-treated approach, 28 (7%) were considered protocol violators.

Efficacy: first attack. In the rizatriptan group, 76.9% of patients reported pain relief at 2 hours compared with 36.6% for the placebo group (p < 0.001). By 1 hour, patients on rizatriptan 10 mg already had a significantly better response than those who took placebo (55% versus 31%; p < 0.001). The response is apparent at 30 minutes, based on the clear numeric advantage of rizatriptan over placebo at this time point.

Response rates of patients on placebo and rizatriptan increased from hour 2 to hour 3 (3.6 percentage points for placebo and 6.0 percentage points for rizatriptan). By 4 hours postdose, 84% of patients on rizatriptan 10 mg were responders compared with 46% of patients on placebo. The increase in response rates at 3 and 4 hours for patients on placebo was influenced by the use of escape medication taken by 42% of these patients at 2 hours after the initial dose (compared with 15% of patients on rizatriptan 10 mg). Seetable 2 for the percentage of patients with pain relief through 4 hours.

Significantly more patients given rizatriptan 10 mg reported complete relief (headache severity of 0) at 2 hours compared with those given placebo during the first attack (44% versus 7%; p < 0.001). Beginning at 1 hour postdose for the first attack, there was a significantly higher percentage of patients on rizatriptan 10 mg who showed complete relief compared with those on placebo (13% versus 2%, p = 0.018). By 4 hours, 54% of patients on rizatriptan demonstrated complete relief compared with 13% for patients on placebo. See table 3 for the percentage of patients with complete relief through 4 hours.

Efficacy: attacks 2 through 4. Similar response rates over time were observed during attacks 2, 3, and 4 over the 4-hour period as observed during attack 1 (see table 2). During each attack, rizatriptan 10 mg showed consistently high response rates of approximately 77% by 2 hours, which were markedly higher than placebo (28 to 54%). Placebo response increased somewhat during attack 4 (54%), which included the fewest number of patients. These patients received rizatriptan 10 mg for their three previous attacks.

For subsequent attacks there was a modest increase in response rates at 4 hours (relative to those at 2 hours) for patients on rizatriptan, ranging from 82 to 84% compared with 49 to 61% for patients on placebo (seetable 2). As with earlier time points, response rates at 3 and 4 hours for patients on rizatriptan were consistent across all four attacks. Again, the increase in response rates for patients on placebo was influenced by the use of escape medication taken by these patients at 2 hours after the initial dose.

For attacks 2, 3, and 4, proportions of rizatriptan-treated patients who demonstrated complete relief at 2 hours (44 to 49%) were consistent with those of the first attack and were substantially higher than those of placebo-treated patients (11 to 21%). Similar proportions of patients on rizatriptan 10 mg reported complete relief within 2 hours for attacks 2, 3, and 4 (see table 3). By 4 hours, 56 to 62% of patients on rizatriptan had complete relief compared with 21 to 32% of patients on placebo (see table 3).

Overall pain relief over four attacks. Over all four attacks, the average percentage of patients experiencing pain relief was 77% for patients on rizatriptan 10 mg and 39% for patients on placebo. Analysis based on the GEE method showed that rizatriptan was significantly better than placebo over the four attack periods (p < 0.001). Importantly, no significant carryover by treatment interaction for the overall analysis of pain relief at 2 hours was observed. This indicates that the efficacy of rizatriptan was consistent over all four attacks. Since the carryover by treatment interaction was not significant and the rizatriptan 10-mg response rate for any headache was greater than placebo for each attack, the treatment comparison between rizatriptan and placebo was made according to the main effects of treatment and attack period.

Response in individual patients. Rizatriptan 10 mg provided consistent relief of migraine headaches over four migraine attacks(table 4). Of the 356 patients who treated at least two migraine attacks with rizatriptan 10 mg, 341 (96%) responded in at least one of the attacks. Of the 315 patients who treated at least three migraine attacks with rizatriptan 10 mg, 272 (86%) had relief from at least two of the attacks. Of the 63 patients who treated four attacks, 46 (73%) achieved relief at least three times.

Patients who did not respond during attack 1 were examined for response during attack 2. Among the 220 patients who treated attack 1 and attack 2 with rizatriptan 10 mg, 170 patients (77%) obtained relief during the first attack and 50 (23%) did not. When the 50 patients who did not obtain relief were treated during attack 2, 35 of them (70%) responded to rizatriptan 10 mg. Therefore, 205 (i.e, 170 plus 35) of the 220 patients (93%) responded during at least one of the first two attacks.

The interval between attacks ranged from 1 day to about 18 weeks, with the average time between attacks about 2 to 3 weeks. The time between attacks was similar across treatment sequences and did not increase or decrease with each subsequent attack.

Migraine recurrence. Recurrence rates among responders who took rizatriptan were very consistent across all four attacks (41 to 47%). However, the percentages of patients on placebo who reported recurrences were quite variable (33 to 59%), perhaps due to the small number of patients on placebo who experienced initial relief. The response rate at 2 hours after treatment of recurrence (72% for patients on rizatriptan and 46% for patients on placebo) was similar to that seen during treatment of the initial headache. The response rates for the treatment of recurrence in patients treating attacks 2 through 4 were somewhat higher for both rizatriptan and placebo. Patients on rizatriptan had response rates of 86% for attack 2, 85% for attack 3, and 90% for attack 4. Patients on placebo (N ∼ 12 for each attack) had response rates of 77% for attack 2, 86% for attack 3, and 60% for attack 4.

Functional disability. For the first attack, 48% of patients on rizatriptan rated themselves in the "normal" category of functional disability at 2 hours compared with only 22% on placebo (p < 0.001). The distribution of functional disability ratings observed during attack 1 was similar to those observed during attacks 2, 3, and 4 (e.g., 54% during attack 2, 53% during attack 3, and 48% during attack 4 of patients on rizatriptan 10 mg with a "normal" rating at 2 hours).

Associated symptoms. A significantly smaller (p < 0.05) percentage of patients taking rizatriptan 10 mg reported photophobia(40%), phonophobia (32%), or nausea (23%) at 2 hours after treatment of the first attack than those on placebo (59% for photophobia, 50% for phonophobia, and 42% for nausea). A greater reduction in the incidence of these associated symptoms was observed across all attacks among patients who took rizatriptan 10 mg compared with placebo.

Overall satisfaction with medication at 2 hours. Patients who took rizatriptan 10 mg were significantly more satisfied with their medication than patients who took placebo (p < 0.001). At 2 hours, 69% of the patients on rizatriptan were "completely satisfied," "very satisfied," or "somewhat satisfied" with their medication compared with 20% on placebo. Similar results were observed across the remaining attacks.

Safety. First attack. Table 5 displays the incidences of the most common adverse experiences (≥2% in either treatment group) reported before a second dose of test medication and any time after any dose of study medication for the first attack. The most common adverse experiences for patients on rizatriptan were dizziness, somnolence, nausea, and asthenia/fatigue, with the incidences of these adverse experiences increasing slightly with repeated dosing. Adverse events were typically mild and transient in character. Although nausea was a component of migraine headache in many patients, it was also sometimes reported as an adverse experience. For example, nausea was reported as an adverse event by 8% of patients after taking rizatriptan 10 mg (versus 4% of patients taking placebo). However, the percentages of patients treated with rizatriptan who reported nausea as a migraine-associated symptom decreased dramatically. At 2 hours after dosing, 23% of patients on rizatriptan 10 mg reported nausea as an associated symptom compared with 55% at baseline (for comparison, 42% of patients on placebo reported nausea as an associated symptom at 2 hours compared with 58% at baseline).

All four attacks. Table 6 displays a similar summary of the most common adverse experiences (≥2% in either treatment group) over all four attacks. The most common adverse experiences over all attacks were dizziness, somnolence, nausea, and asthenia/fatigue. When the incidences of adverse experiences were adjusted to take into account patients' repeated exposure to rizatriptan in as many as four attacks(relative to the exposure to placebo in a single attack), adverse experiences actually occurred less frequently than those observed in the first attack. That is, the incidence of adverse experiences per attack decreased after the first attack.

Two serious adverse experiences occurred in two patients. One serious adverse experience, cellulitis (judged not to be drug-related by the investigator), occurred 10 days after rizatriptan 10 mg had been taken to treat a second attack. The patient also treated the first attack with rizatriptan 10 mg. The other serious adverse experience, anaphylaxis(attributed by the investigator to antibiotic therapy), occurred in a patient who received placebo in the attack just before the serious adverse event.

Discussion. The results of this double-blind, placebo-controlled, multicenter study confirm that rizatriptan 10 mg is highly effective, rapid acting, and well tolerated in the acute treatment of migraine and its recurrences. Rizatriptan 10 mg provides complete relief, and relieves accompanying symptoms and functional disability in a high percentage of patients. Importantly, these results show clearly that the tolerability and efficacy of rizatriptan 10 mg is maintained uniformly over a series of attacks.

Because migraine is a markedly painful condition, the use of placebo control to treat a series of attacks, although advantageous from a scientific design perspective, is impractical in terms of patient retention and recruitment and would also raise ethical concerns. This study, based on a variation of design, provided a placebo control but circumvented the need for any given patient to take placebo for more than one attack. The design permitted a principal examination of the efficacy and safety of rizatriptan, compared with placebo for the first attack (plus recurrences) that was analyzed as a parallel group design. Furthermore, an analysis based on a standard modification of the GEE model was used to examine the efficacy and safety of rizatriptan 10 mg compared with placebo over the course of all attacks. The potential weaknesses of the design (i.e., crossover effects and differential dropouts due to efficacy or safety) were tested and found not to influence the results.

During the first attack, 77% of patients on rizatriptan achieved relief at 2 hours. This response was maintained over four attacks and ranged from 75 to 80%. High response rates observed with rizatriptan were not due to an artifact of the current study design, because our results were neither influenced by differential dropouts due to inefficacy or adverse experiences nor by carryover effects. Therefore rizatriptan 10 mg appears to offer very high efficacy and rapid onset of symptom relief.

Recurrence rates in this study were relatively high but similar to those reported in studies with sumatriptan 100 mg.^17,19 In a long-term (6- to 12-month) study of rizatriptan, a lower recurrence rate(30%) was observed with rizatriptan 10 mg.²⁵

The novel design of this study permitted examination of the efficacy of rizatriptan 10 mg over a series of four attacks. The efficacy that was documented for the first attack was maintained through the remaining three attacks. Approximately 77% of patients receiving rizatriptan 10 mg responded at 2 hours in attacks 2 through 4. The response to placebo was significantly less than the response to rizatriptan 10 mg for each attack, but was more variable over the four attacks (28 to 55%). This variability was likely due in part to the small numbers of patients on placebo (approximately 50 to 80 patients), especially during the last attack. It should be noted that patients receiving placebo for their fourth attack received rizatriptan 10 mg for their previous three attacks, and their expectation of a response may have been greater. Most patients on rizatriptan 10 mg responded in the majority of treated attacks. Approximately 86% of patients who treated three attacks with rizatriptan 10 mg responded to at least two of the attacks.

Rizatriptan was very well tolerated in this study. The safety and tolerability of rizatriptan 10 mg after repeated treatment of multiple discrete migraine attacks was as favorable as that observed after its use in single attacks. The most common adverse experiences (dizziness, somnolence, asthenia/fatigue, and nausea) were typically mild and transient in nature and did not increase with cumulative exposure to rizatriptan over a series of discrete attacks. As expected, raw incidences of clinical adverse experiences(i.e., unadjusted) were somewhat greater over multiple attacks treated with rizatriptan 10 mg compared with the incidence observed during the initial attack. However, when adjusted (see Methods), adverse experiences actually occurred less frequently than those observed in the first attack. That is, the incidence of adverse experiences per attack decreased after the first attack. This indicates that adverse experiences tended to occur early in the series of attacks, not after cumulative exposure. For example, 33 of 55 patients (60%) reporting somnolence did so initially, after they treated their first attack with rizatriptan 10 mg, indicating that only 22 patients reported this side effect after treatment of subsequent attacks. These adverse experiences have been reported in other patients receiving drugs of the 5-HT_1B/1D receptor agonist class. There were no serious drug-related clinical adverse experiences associated with use of rizatriptan 10 mg in this study.

Appendix