Abstract
This randomized, double-blind, two-attack, placebo-controlled, crossover study explored the efficacy and tolerability of rizatriptan 10 mg compared with sumatriptan 50 mg as well as rizatriptan 5 mg compared with sumatriptan 25 mg in the acute treatment of migraine. Following randomization to one of six possible treatment sequences, patients (n = 1447) treated two sequential attacks, of moderate or severe intensity, separated by at least 5 days. Patients assessed pain severity, migraine-associated symptoms, and functional disability at 0.5, 1, 1.5, and 2 h post treatment. Compared with placebo, all treatments were effective. On the primary endpoint of time to pain relief, rizatriptan 10 mg was not statistically different from sumatriptan 50 mg [odds ratio (OR) 1.10, P = 0.161], and rizatriptan 5 mg was statistically superior to sumatriptan 25 mg (OR 1.22, P = 0.007). In general, rizatriptan 10 mg and 5 mg treatment resulted in improvement compared with the corresponding doses of sumatriptan on measures of pain severity, migraine symptoms, and functional disability and the 5-mg dose reached statistical significance on almost all measures. All treatments were generally well tolerated.
Introduction
The 5-HT1B/1D receptor agonists, rizatriptan and sumatriptan, have been shown to be effective and generally well tolerated for the acute treatment of migraine. Various oral dosage strengths of these medications have been compared in several head-to-head studies; however, only one of these studies involved a comparison over several dosage strengths (1). In that study, Goldstein et al. compared rizatriptan 5 mg and 10 mg with sumatriptan 25 mg and 50 mg, respectively. At any point in the first 2 h after treatment, rizatriptan 5 mg and 10 mg were more likely to provide headache relief, and provided greater relief of migraine-associated symptoms, than the corresponding doses of sumatriptan. The present study was conducted to confirm the results of the Goldstein study (1). The primary results have been previously reported (2) and other data from the study have been included in published meta-analyses (3–6). Here, we provide a full report of the study.
Methods
This study was similar in design to that reported by Goldstein et al. (1). The following is a summary of key aspects of the methodology.
Patients
Men and women in good health aged ≥ 18 years with at least a 6-month history of migraine with or without aura, as defined by the International Headache Society diagnostic criteria for migraine (7), were enrolled at 67 study sites in the USA. All patients gave written informed consent, and the study protocol was approved by the Institutional Review Board at each site. Patients were excluded from study participation if they used monoamine oxidase inhibitors, methysergide, or propranolol; however, standard antimigraine prophylactic medications (with the exception of non-steroidal anti-inflammatory drugs, daily analgesics, or propranolol) were permitted. Women who were pregnant or nursing were not eligible for the study. Patients also were excluded if they had participated in the previous comparison study (1).
Study design and procedure
This was a randomized, incomplete-block, double-blind, placebo-controlled, two-attack crossover study. Patients were randomized according to a computer-generated allocation schedule to one of six two-period treatment sequences: rizatriptan 10 mg followed by sumatriptan 50 mg, sumatriptan 50 mg followed by rizatriptan 10 mg, rizatriptan 5 mg followed by sumatriptan 25 mg, sumatriptan 25 mg followed by rizatriptan 5 mg, or placebo followed by placebo. (There were two placebo/placebo groups: one group matched the rizatriptan 5 mg/sumatriptan 25 mg treatments, and the other group matched the rizatriptan 10 mg/sumatriptan 50 mg treatments. No significant differences in results were observed between the two placebo groups, so their data were combined.)
Patients were asked to treat two migraine headache attacks of moderate or severe intensity, as assessed using the standard 4-point headache pain scale (0 = no pain, 1 = mild pain, 2 = moderate pain, 3 = severe pain). The two treatment periods were separated by at least 5 days. Rescue use of analgesics and anti-emetics was permitted from 2 h onwards in the case of treatment failure or headache recurrence (return of a moderate or severe headache within 24 h in a patient who had experienced initial pain relief at 2 h).
Outcome measures
For each migraine treated, patients completed a separate diary. Assessments were made at baseline (immediately before the administration of study medication) and at 0.5, 1, 1.5, 2, 3, and 4 h after dosing. Patients were administered timers to facilitate recording of their symptoms at the prespecified time points. Patients assessed multiple features of each attack: headache severity, presence of migraine-associated symptoms (photophobia, phonophobia, nausea, or vomiting) and the level of functional disability according to a 4-point scale (0 = normal, 1 = daily activities mildly impaired, 2 = daily activities severely impaired, 3 = unable to carry out daily activities/required bed rest). Patients were asked about adverse experiences in an open-ended fashion and reported these throughout the study.
Data analysis
Efficacy analyses were based on all patients who treated at least one headache and recorded at least one efficacy assessment after the baseline severity assessment.
Missing values in the treatment phase were imputed by carrying forward the last recorded efficacy value. Comparisons between active treatments were made by using data from both headaches. All comparisons of active treatments with placebo were made using data from the first attack only, in order to eliminate any bias introduced when a patient treated both migraines with placebo. The analysis focused on time points up to 2 h, the last time before rescue medications were allowed.
The primary objective of the study was to compare rizatriptan 10 mg and sumatriptan 50 mg in terms of time-to-pain relief during the 2 h after taking study drug. Time-to-pain relief was the first protocol-specified diary assessment time when a patient reported no headache or a headache of mild severity. This endpoint is discussed in further detail elsewhere (1, 8). A common odds ratio (OR) was used to approximate a hazard ratio. To provide at least 80% power to detect a common OR of at least 1.2 between two treatments, the study required 240 patients in each of the rizatriptan 10 mg/sumatriptan 50 mg and sumatriptan 50 mg/rizatriptan 10 mg treatment sequence groups, based on a two-sided test with type I error rate α = 0.05. The time-to-pain relief analysis was performed using regression models for grouped survival data with a generalized estimating equation approach. The effects due to treatment, treatment period, treatment sequence, and study-site region were assessed, as well as the underlying model assumption of proportional hazards.
Additional secondary endpoints were 2-h pain-relief status, 2-h pain-free status, and the presence of associated symptoms at 2 h. For each of the secondary endpoints, two analyses were performed: (i) comparisons of the rizatriptan doses with the sumatriptan doses (i.e. rizatriptan 5 mg vs. sumatriptan 25 mg and rizatriptan 10 mg vs. sumatriptan 50 mg); and (ii) comparisons of the rizatriptan doses and the sumatriptan doses with placebo. All pairwise comparisons of active treatments were based on logistic regression models with a generalized estimating equation method; all pairwise comparisons with placebo were based on simple logistic regression models. Functional disability measures were compared between active treatments using a generalized ordinal categorical regression model and between active treatments and placebo using a cumulative logistic regression model.
All patients who took study medication were included in the safety analysis, for which the primary endpoint was the overall incidence of adverse events. Comparisons of active treatments with placebo were performed using data from the first attack only. An overall χ2 test was performed first for each of the adverse events at α = 0.10 level. If the test was significant, then pairwise comparisons were performed at α = 0.05 level to investigate further the source of differences. Fisher's exact test was used for the pairwise comparisons. All comparisons between active treatments were made using combined data from both attacks, and all combinations of pairwise comparisons were performed. These data were analysed using a generalized estimating equation method that took into account the crossover nature of the study design.
Results
Study population
A total of 1622 patients were randomized to six different treatment sequences in this study. Of the 1622 randomized patients, 1447 treated at least one attack, 1288 treated two attacks, and 1287 patients completed the study and all study-related procedures. Of the 1447 patients who treated the first attack, 160 (11%) did not complete the study for the following reasons: three discontinued due to clinical adverse experiences; 94 did not have a second headache during the enrolement period; 18 were lost to follow-up; 13 withdrew from the study; 14 were uncooperative; and 18 discontinued due to other reasons such as lack of therapeutic response, need for concomitant medication, or inclusion/exclusion criteria not met. The numbers of patients discontinuing, and reasons for discontinuation, were similar between the treatment groups. For the first attack (total patients treated = 1447), 288 patients treated with rizatriptan 5 mg, 296 with rizatriptan 10 mg, 290 with sumatriptan 25 mg, 285 with sumatriptan 50 mg, and 288 with placebo. For the second attack (total patients treated = 1288), 248 patients treated with rizatriptan 5 mg, 251 with rizatriptan 10 mg, 264 with sumatriptan 25 mg, 265 with sumatriptan 50 mg, and 260 with placebo.
Patients included in the study were predominantly white (87%), and female (86%), with a mean age of 40 years. The initial treated migraine attack was moderate in 62% of patients and severe in 38% of patients. The treatment groups were similar with respect to these characteristics.
Efficacy
The primary efficacy variable, expressed as the hazard ratio of rizatriptan 10 mg vs. sumatriptan 50 mg, was 1.10 [95% confidence interval (CI) 0.96, 1.26; P = 0.161]. Rizatriptan 5 mg was statistically superior (P = 0.007) to sumatriptan 25 mg on this measure of time-to-pain relief; the hazard ratio of rizatriptan 5 mg vs. sumatriptan 25 mg was 1.22 (95% CI 1.06, 1.41). Thus, in this study patients taking rizatriptan 5 mg were 22% more likely to achieve pain relief sooner within the 2-h period after initial treatment than patients taking sumatriptan 25 mg.
Table 1 shows the comparison between active treatment groups in the percentage of patients reporting pain relief at time points up to 2 h. At all time intervals, the percentage of patients taking rizatriptan 10 mg who reported pain relief was greater than that of sumatriptan; at the 1-h time interval, this difference reached a significant level (P< 0.05). At no observed time interval did the percentage of patients who reported pain relief with sumatriptan exceed the percentage of patients taking rizatriptan. For patients taking rizatriptan 5 mg, the percentage of patients reporting pain relief was significantly greater at 0.5 (P< 0.05), 1.5 (P< 0.05), and 2 h (P< 0.005) after taking study medication compared with those taking sumatriptan 25 mg. There was no statistically significant difference between these treatment groups at 1 h.
Number (%) of patients reporting pain relief at time points within 2 h by treatments (both attacks, all patients treated)
N, Number of patients with non-missing (or carried forward) pain severity; n, %, number (percent) of patients with pain relief; odds ratio, the odds ratio is from the logistic regression model using GEE.
We found no significant effects due to sequence of the active treatments in this crossover study.
When compared with placebo (using data from the first attack only), rizatriptan 10 mg treatment resulted in statistically greater proportions of patients with pain relief at 0.5 h (P< 0.05) and at 1, 1.5, and 2 h (P< 0.01); rizatriptan 5 mg resulted in statistically greater proportions of patients with pain relief at 1.5 and 2 h (P< 0.05). Both doses of sumatriptan were significantly better than placebo at 1 h (P< 0.05), and 1.5 and 2 h (P< 0.01).
Table 2 shows the results for the additional secondary efficacy variables at the 2-h time point in terms of the paired active treatment comparisons. The rizatriptan 10 mg-treated patients had significantly less nausea (P = 0.004) compared with those treated with sumatriptan 50 mg. For all other secondary measures at 2 h, rizatriptan 10 mg treatment was numerically greater, but not statistically different than sumatriptan 50 mg. Rizatriptan 5 mg treatment resulted in significantly improved scores on the measures of pain-relief status, pain-free status, functional disability, and migraine-associated symptoms (except vomiting) as assessed at 2 h compared with sumatriptan 25 mg.
Rizatriptan 5 mg vs. sumatriptan 25 mg and rizatriptan 10 mg vs. sumatriptan 50 mg (both attacks), secondary efficacy endpoints at 2 h
N represents maximum number of patients in each group. This number varied slightly, but by no more than 11 patients, on each measure.
An odds ratio < 1 for the presence of any associated symptom means that this symptom is less likely to be present with rizatriptan compared with sumatriptan.
When the active treatments were each compared with placebo at 2 h (using data from the first attack only), all four treatment groups were significantly better than placebo for photophobia and phonophobia, but not for presence of nausea; for this associated symptom, only the rizatriptan 5 mg group experienced significantly less nausea (P< 0.01) when compared with the placebo group. The percentage of patients reporting nausea were 22% (rizatriptan 5 mg), 26% (rizatriptan 10 mg), 29% (sumatriptan 25 mg), 32% (sumatriptan 50 mg), and 32% (placebo). None of the active treatment groups differed from placebo with regard to vomiting.
In addition to efficacy observations up to the 2-h time point, we also examined the use of additional migraine medication from 2 to 4 h. There was no statistically significant difference between rizatriptan 10 mg and sumatriptan 50 mg (19.7% vs. 21.1%) or rizatriptan 5 mg and sumatriptan 25 mg (26.1% vs. 22.9%) with regard to percentage of patients needing additional analgesia. The percentage of patients requiring additional analgesia and/or anti-emetics was significantly less in each of the treatment groups when compared with placebo during the first attack (38.9% for placebo; 29.6% for rizatriptan 5 mg; 22.6% for rizatriptan 10 mg; 22.8% for sumatriptan 25 mg; 20.7% for sumatriptan 50 mg).
Safety
All 1447 patients who took study medication were included in the safety analysis. Table 3 provides a summary of the number of clinical adverse experiences by treatment. All active treatments were generally well tolerated. As expected, a greater number of patients reported adverse experiences in the higher dose active treatment groups. Using data from the first attack, the overall incidence of any adverse experiences in the rizatriptan 10 mg and sumatriptan 50 mg groups was each significantly higher than in the placebo group (P< 0.01). There was a statistical difference for these treatment groups compared with placebo with regards to several of the commonly reported adverse experiences: flushing, dizziness, headache, and paresthesia. There were no drug-related serious adverse events. The most common adverse experiences reported (≥ 3% of the patients in at least one treatment group) were asthenia/fatigue, chest pain, diarrhoea, dry mouth, nausea, dizziness, headache, paresthesia, somnolence, and flushing.
Adverse events (AEs), first attack
P < 0.05 when compared with placebo;
P < 0.01 when compared with placebo.
N, Number of patients exposed to the indicated treatment group during the first attack.
Discussion
The efficacy and tolerability of rizatriptan and sumatriptan in the treatment of migraine have been clearly demonstrated. Within the triptan class, it is valuable to understand the relative efficacy and tolerability of different therapeutic agents. A previous study directly comparing rizatriptan 10 mg with sumatriptan 100 mg suggested that rizatriptan 10 mg was faster acting and more effective than sumatriptan 100 mg, the highest recommended dose at the time of the study (9). A subsequent study was conducted by Goldstein et al. to compare the efficacy and tolerability of rizatriptan 5 mg vs. sumatriptan 25 mg, and rizatriptan 10 mg vs. sumatriptan 50 mg (1). In that study, both doses of rizatriptan were more likely to provide faster onset of pain relief than the corresponding doses of sumatriptan, consistent with the more rapid absorption and higher bioavailability of rizatriptan compared with oral sumatriptan (T max of approximately 1 h for rizatriptan compared with 2.5 h for sumatriptan) (10–12).
The current study is in general consistent with the Goldstein et al. study: the hazard ratio for time to pain relief in the present study again favoured rizatriptan 5 mg vs. sumatriptan 25 mg, indicating patients were more likely to achieve pain relief sooner within 2 h after taking rizatriptan than sumatriptan. The hazard ratio for the rizatriptan 10 mg vs. sumatriptan 50 mg comparison (1.10) was similar to that reported in the Goldstein et al. study (1.14), but did not reach statistical significance in the present study. The hazard ratio for rizatriptan 5 mg compared with sumatriptan 25 mg (1.22) was statistically significant. Significantly more patients treated with rizatriptan 10 mg reported pain relief at 1 h compared with those treated with sumatriptan 50 mg; however, the two treatments were not statistically different at 2 h. These findings suggest a faster onset of pain relief with rizatriptan 10 mg than with sumatriptan 50 mg.
Additional secondary efficacy measures such as headache pain relief, the presence of migraine-associated symptoms, functional disability, and the need for additional medications favoured rizatriptan 10 mg and 5 mg compared with sumatriptan 50 mg and 25 mg, respectively. At no endpoint or time did sumatriptan demonstrate greater efficacy than rizatriptan. Notably, although all active treatments were significantly better on the measure of pain relief than placebo at 1 h, rizatriptan 10 mg was the only treatment that provided significant pain relief, compared with placebo, at 0.5 h.
Both rizatriptan and sumatriptan were generally well tolerated. The adverse events following rizatriptan use were similar, both qualitatively and quantitatively, to those that were reported in previous studies. Flushing, dizziness, headache, and paresthesia occurred more often in the higher dose active treatment groups than in the placebo group, again consistent with previous reports.
Some of the data presented here were previously included as part of an extensive meta-analysis of 53 trials of oral triptans involving over 24 000 patients (3). The authors of that meta-analysis also performed a separate analysis of all direct comparative studies. Both approaches gave similar results. The authors noted that rizatriptan 10 mg achieved greater efficacy, with similar tolerability, than treatment with sumatriptan 100 mg, which is double the dose stud-ied here. (Sumatriptan 100 mg has similar efficacy to the 50-mg dose, however (13)). The authors also noted that, based on their analyses, the efficacy of rizatriptan 10 mg is among the highest in the triptan class, particularly when rapid pain relief is desired. The present study achieved results that are consistent with both the previous study by Goldstein et al. and the triptan-class meta-analysis (3). These observations of the relative efficacy of rizatriptan and sumatriptan, as measured by the endpoint of time to pain relief and supported by secondary efficacy measures, can be used to guide therapeutic choice within the triptan class.
Footnotes
Acknowledgements
The authors would like to especially acknowledge and thank Dr Carolyn Hustad and Dr Christopher Lines for their review and contributions to this manuscript. We also thank the following investigators who participated in this study (Protocol 052): Michael Andriola MD, Clearwater, FL; Jeffrey Baggish MD, Catonsville, MD; Harvey Blumenthal MD, Tulsa, OK; Robert E. Broker MD, Greer, SC; Richard Buckler MD, Sellersville, PA; Elias Chalhub MD, Mobile, AL; John B. Chawluk MD, Pottsville, PA; Angel R. Chinea MD, Santurce, PR; Louis Cohen MD, Sarasota, FL; Stephen E. Daniels DO, Austin, TX; Naomi de Sola Pool MD, Hackensack, NJ; Seymour Diamond MD, Chicago, IL; Eugene A. Duboff MD, Denver, CO; Arthur H. Elkind MD, Mount Vernon, NY; Debra Elliott MD, New Orleans, LA; Charles Fogarty MD, Spartanburg, SC; Carol A. Foster MD, Phoenix, AZ; Gordon L. Gibson MD, Little Rock, AR; Norman M. Gordon MD, East Providence, RI; Leland Greenwald MD, Sunnyvale, CA; Daniel Gremillion MD, Nashville, TN; Anne Hawes MD, Charlotte, NC; Scott A. Heatley MD, PhD, Redwood City, CA; Scott Heller MD, CCCR, Chicago, IL; Lynne Hopkins MD, Winter Park, FL; Raymond E. Jackson MD, Southfield, MI; Gary W. Jay MD, Northglenn, CO; Robert G. Kaniecki MD, Pittsburgh, PA; Thomas C. Klein MD, Wichita, KS; David Kudrow MD, Encino, CA; Stephen H. Landy MD, Germantown, TN; Burton W. Lazar MD, Portland, OR; Robert S. Lipetz DO, Spring Valley, CA; Thomas W. Littlejohn III MD, Winston-Salem, NC; Elizabeth Loder MD, Boston, MA; Sylvia Lucas MD, PhD, Seattle, WA; Donald W. Middleton Jr MD, North Dartmouth, MA; Carmen Montoya MD, San Antonio, TX; Robert B. Nett MD, San Antonio, TX; Francis J. O’Donnell DO, Westerville, OH; John E. Pappas MD, Lexington, KY; William M. Patterson MD, Birmingham, AL; Kenneth S. Peters MD, Mountain View, CA; John Porter MD, Winston-Salem, NC; Nabih Ramadan MD, Cincinnati, OH; Marc Raphaelson MD, Frederick, MD; George J. Rederich MD, MS, Torrance, CA; Jeffrey B. Rosen MD, Coral Gables, FL; John Rubino MD, Raleigh, NC; Elliott A. Schulman MD, Upland, PA; Lawrence Seiden MD, Towson, MD; Fred D. Sheftell MD, Stamford, CT; Richard J. Sievers DO, Kettering, OH; Richard Singer MD, Pembroke Pines, FL; David A. Smith MD, Atlanta, GA; Malcolm Sperling MD, Fountain Valley, CA; Egilius L.H. Spierings MD, PhD, Wellesley Hills, MA; R. Malcom Stewart MD, Dallas, TX; John Stoukides MD, East Providence, RI; Gus Stratton MD, Barrington, RI; Anita Vaughn MD, New York, NY; Russell W. Walker MD, Phoenix, AZ; Susan Wehle MD, Brandon, FL; KMA Welch MD, Detroit, MI; Jeannette Wendt MD, Tucson, AZ; Nina Zasorin MD, Long Beach, CA; Steven Zeig MD, Ft Lauderdale, FL.