OBJECTIVE. The goal was to evaluate the efficacy and tolerability of zolmitriptan nasal spray in the treatment of adolescent migraine.
METHODS. The “Double-Diamond” study used a novel, single-blind, “placebo challenge” in a multicenter, randomized, double-blind, placebo-controlled, 2-way, 2-attack, crossover design. A total of 248 US adolescent patients (12–17 years of age) with an established diagnosis of migraine, with or without aura, were enrolled. A single-blind placebo challenge was used for each migraine attack. No additional medications were taken if a headache response to the initial placebo treatment was achieved at 15 minutes; if migraine intensity remained moderate or severe, then patients treated the attack with zolmitriptan (5 mg) nasal spray or placebo according to a randomized, crossover schedule (double-blind). The primary efficacy variable was headache response at 1 hour after treatment. A comprehensive range of secondary end points included sustained headache response at 2 hours.
RESULTS. A total of 171 patients (mean age: 14.2 years; 57.3% female) treated ≥1 attack with study medication (intention-to-treat population). The onset of significant pain relief was apparent 15 minutes after treatment with zolmitriptan nasal spray. At 1 hour after the dose, zolmitriptan nasal spray produced a higher headache response rate than did placebo (58.1% vs 43.3%). Zolmitriptan nasal spray was also significantly superior to placebo in improvement in pain intensity, pain-free rates, sustained resolution of headache, and resolution of associated migraine symptoms. Return to normal activities was also consistently faster with zolmitriptan nasal spray than with placebo, with less use of any escape medication. Treatment with zolmitriptan nasal spray was well tolerated.
CONCLUSIONS. This novel, placebo-challenge study demonstrated that zolmitriptan nasal spray was well tolerated and provided fast and significantly effective relief of migraine symptoms in the acute treatment of adolescent migraine.
Migraine is common among adolescents, with a prevalence rate of 15% among 12- to 15-year-old subjects.1 In this age group, migraine is generally characterized by recurrent attacks of intense headache that tend to be shorter in duration than in adult migraine.2 In conjunction with autonomic symptoms, including nausea, vomiting, phonophobia, and/or photophobia, migraine contributes to decreased functional ability and a significant burden on quality of life among adolescents.3 However, migraine attacks among adolescents often go unrecognized by health care providers or are misattributed to causes such as sinus disease or emotional disorders.
The management of adolescent migraine requires an individually tailored approach that encompasses the balance of biobehavioral measures with agents for acute treatment and, if the frequency of attacks or associated disability so mandate, daily preventive medicines. It has proved extremely difficult, however, to demonstrate significant efficacy in the adolescent population with migraine-specific agents such as triptans.4–6 This has been attributed to the high placebo response rate in clinical trials of acute treatment for adolescent migraine. Placebo response rates of ∼35% for headache response at 2 hours have been observed for adults, whereas the corresponding placebo response rates in adolescent migraine studies have been reported as ≥50%.7 Such problems prompted exploration of novel study designs to determine the efficacy of migraine-specific therapies, relative to placebo, in this population.
Zolmitriptan nasal spray (NS) has been demonstrated to be fast-acting, efficacious, and well tolerated in the acute treatment of migraine attacks in adult patients,8,9 with comparable pharmacokinetic profiles for adolescents and adults.10 Therefore, the objective of the present study was to evaluate the efficacy and tolerability of zolmitriptan NS versus placebo in the acute treatment of adolescent migraine. An expert panel of pediatric neurologists, including the study centers performing this study, recommended that a novel study design be adopted to address the anticipated high placebo response rates among adolescent patients.
The present study (study code D1221C00005; “Double-Diamond”) was conducted between September 2003 and March 2004 at 17 pediatric neurology centers in the United States, according to the principles of the Declaration of Helsinki and good clinical practice. Written informed consent was obtained from each patient's parent or legal guardian, with assent from the patient, before the start of the study; the protocol received local ethics committee or institutional review board approval at each study center.
Adolescent patients (12–17 years of age for the duration of the study) with an established diagnosis of migraine, with or without aura, for ≥1 year, according to International Headache Society criteria11,12 or International Headache Society revised criteria,13 were eligible for enrollment. Baseline migraine frequency was required to be ≥2 attacks per month during the school year and <14 days/month without migraine headache for the 3 months before screening. The usual duration of untreated migraine was to be >2 hours in the 3 months before screening. The protocol allowed for inclusion of patients taking medication for migraine prophylaxis, provided this had been stabilized within 2 months before randomization. Standard exclusion criteria for triptan therapy trials applied, including a history of basilar, ophthalmoplegic, or hemiplegic migraine; evidence of or significant risk factors for ischemic heart disease, arrhythmia, or accessory conduction pathway disorders; and previous unacceptable adverse events with any triptan agent.
This was a multicenter, randomized, double-blind, placebo-controlled, 2-way, 2-attack, crossover study, with a single-blind “placebo challenge” for each migraine attack. Each attack was treated initially with placebo NS within 30 minutes after the headache reached moderate or severe intensity (single-blind). If a headache response was achieved at 15 minutes (ie, reduction in headache pain from moderate/severe to mild or none), no additional medication was to be taken. If migraine intensity remained moderate or severe, then patients treated the attack with zolmitriptan NS (5 mg) or matching placebo NS according to the double-blind randomization schedule (prepared by the study sponsor). Patients were randomly assigned strictly sequentially. A second dose of randomized treatment or approved escape medications (nonsteroidal antiinflammatory drugs, antiemetics, analgesics, or sedatives) was permitted 2 hours after the dose if moderate or severe headache pain persisted. The second migraine attack within the 12-week study period was to be treated with the crossover medication.
The primary efficacy variable was headache response (defined as 2-point reduction in headache intensity on a 4-point scale, from 1 [no headache] to 4 [severe headache]) at 1 hour after randomized treatment. Secondary efficacy variables included headache response at 15, 30, and 45 minutes and 1.5 and 2 hours after randomized treatment; sustained headache response at 2 hours (defined as patients who responded at 1 hour after the dose, remained responders through 2 hours, and did not use escape medications); pain-free rates, 1-point decreases in pain intensity (evaluated on a 4-point scale: none, mild, moderate and severe), and resolution of associated migraine symptoms at all time points from 15 minutes to 2 hours after the dose; time to return to normal activities; and use of escape medications. Patients recorded migraine pain severity, symptoms, and other responses by using an electronic hand-held device (LogPad system; PHT Corp, Charlestown, MA).
At the request of the US Food and Drug Administration (FDA), which had concerns regarding dropping patients from the intention-to-treat (ITT) analysis after randomization, headache response data were imputed for placebo responders. In this “worst-case scenario” analysis, all responders to placebo challenge who would have subsequently received randomized placebo were deemed “responders to placebo” but those who would have subsequently received randomized zolmitriptan NS were deemed “nonresponders to zolmitriptan NS.”
The ITT population was defined as patients who did not respond to placebo challenge in the first 15 minutes and subsequently treated the migraine attack with randomized medication (zolmitriptan NS or placebo for that attack). The primary efficacy variable, that is, 1-hour headache response, was analyzed with the generalized estimating equations method by using alternating logistic regression analyses14 that simultaneously modeled the dependence of the response on the explanatory variables and the association of the responses from the same patients. The same model was used for analyses of headache responses at other time points (including sustained headache response at 2 hours), pain-free rates, 1-point decreases in pain intensity, return to normal activities, resolution of associated symptoms, and incidence of use of escape medications. Results of the analyses are presented in terms of the odds ratio (OR) and 95% confidence interval (CI) for the treatment effect. Comparisons were defined as significant at P < .05.
The sample size for this study was calculated by using McNemar's test (χ2 test) of equality of paired proportions. It was assumed that, with the exclusion of early placebo responders, a headache response at 1 hour would be achieved by 50% of zolmitriptan NS recipients, compared with 35% of placebo recipients, and the proportion of discordant pairs would be 45%. With a 2-sided, type I error of .05 and power of 90%, a sample size of 203 patients providing data for both attacks was required. However, with estimates that 25% of randomly assigned patients would not provide efficacy data for both attacks, the total calculated sample size was 272 patients (ie, 136 patients per sequence).
Two hundred forty-eight adolescent patients entered the study and were assigned randomly either to sequence A (placebo then zolmitriptan; n = 120) or to sequence B (zolmitriptan then placebo; n = 128) (Fig 1). A total of 171 patients treated ≥1 attack with study medication (ITT population). The mean age was 14.2 years, and 98 (57.3%) of 171 subjects were female; baseline demographic and migraine characteristics are shown in Table 1. In total, 275 migraine attacks were treated during the study (148 with zolmitriptan NS and 127 with placebo). Overall, demographic characteristics were similar between those who responded and those who did not respond to the placebo challenge (data not shown).
Study completion rates for treatment sequences A and B were 80% (96 of 120 patients) and 80.5% (103 of 128 patients), respectively. The most common reasons for discontinuation were protocol noncompliance (13 and 11 patients, respectively) and loss to follow-up monitoring (9 and 10 patients, respectively).
At 1 hour, zolmitriptan NS produced a higher headache response rate (58.1%; 86 of 148 attacks) than did placebo (43.3%; 55 of 127 attacks; OR: 1.827; 95% CI: 1.137–2.936; P < .05). The onset of action was as early as 15 minutes after the dose (37.2% [55 of 148 attacks], compared with 22.8% [29 of 127 attacks] for placebo; OR: 2.020; 95% CI: 1.162–3.510; P < .05). Zolmitriptan NS also produced higher headache responses than placebo at 1.5 and 2 hours after the dose, although the difference was not statistically significant (Fig 2). For 2-hour sustained headache response, zolmitriptan NS was superior to placebo (51.4% [76 of 148 attacks] vs 33.1% [42 of 127 attacks]; OR: 2.243; 95% CI: 1.325–3.796; P < .01).
Pain-free rates were higher with zolmitriptan NS than with placebo at all time points assessed, and these differences between treatment groups were significant at 1 hour (27.7% [41 of 148 attacks] vs 10.2% [13 of 127 attacks]), 1.5 hours (32.4% [48 of 148 attacks] vs 15.7% [20 of 127 attacks]), and 2 hours (39.2% [58 of 148 attacks] vs 18.9% [24 of 127 attacks]; all P < .01) after the dose. At the earlier time points, the differences between the treatment groups were of similar magnitudes; however, a valid statistical analysis could not be performed because of the small number of responders (Fig 3). The early onset of action of zolmitriptan NS was also evident in assessments of a 1-point decrease in pain intensity, with a greater proportion of patients reporting a reduction in pain intensity at the early time points of 15 and 30 minutes after the dose, compared with placebo (50.0% [74 of 148 attacks] vs 35.4% [45 of 127 attacks] and 54.1% [80 of 148 attacks] vs 40.2% [51 of 127 attacks], respectively; both P < .05).
The improvement in headache pain with zolmitriptan NS was accompanied by increased resolution of associated migraine symptoms, compared with placebo. At 30 minutes after the dose, photophobia and phonophobia had resolved in a greater number of zolmitriptan NS-treated attacks, compared with placebo-treated attacks (32.5% [37 of 114 attacks] vs 19.6% [19 of 97 attacks] and 36.0% [32 of 89 attacks] vs 21.4% [18 of 84 attacks], respectively; both P < .05). Rates of return to normal activities (for patients with pretreatment impairment) were also consistently higher with zolmitriptan NS than with placebo at all time points, with significance at 45 minutes and 1 hour after the dose (48.0% [61 of 127 attacks] vs 33.9% [37 of 109 attacks] and 55.1% [70 of 127 attacks] vs 41.3% [45 of 109 attacks], respectively; both P < .05). The use of any escape medication was also lower for zolmitriptan NS-treated attacks, compared with placebo-treated attacks (38.3% [62 of 162 attacks] vs 50.0% [74 of 148 attacks]; P < .05).
Compared with placebo, no significant benefit for zolmitriptan NS was apparent in the 1-hour headache response rate according to the worst-case scenario analysis (data not shown). Similar findings were apparent for the worst-case scenario analysis of the 2-hour sustained headache response, and additional analyses of secondary variables with these criteria were not performed.
Tolerability and Safety
Zolmitriptan NS was well tolerated in this adolescent population. No patient withdrew from the study because of adverse events, and no serious adverse events were reported. The most commonly reported adverse events were taste disturbance, nasal discomfort, and nasal congestion (Table 2).
This uniquely designed study, which used a novel strategy to address the problem of early placebo responders, demonstrated that zolmitriptan NS (5 mg) was effective and well tolerated for the acute treatment of migraine in adolescents. The onset of action was rapid, with significant differences between zolmitriptan and placebo being shown as early as 15 minutes after intranasal administration, and was sustained for up to 2 hours.
The objective of migraine attack management is complete relief of headache and accompanying symptoms, with minimal risk of adverse effects. Compared with placebo, zolmitriptan NS produced numerically higher rates of symptom relief at all time points assessed in the present study, reaching statistical significance for headache responses at 15 minutes, 30 minutes, and 1 hour after the dose. In addition, functional ability was improved, with a faster return to usual activities and a reduced need for escape medications. Zolmitriptan NS was well tolerated among adolescent migraineurs, with a pattern and low incidence of adverse effects similar to those seen in adults. Taken together, these findings point toward the favorable utility of zolmitriptan NS for relief of migraine in the adolescent population.
Several reviews have evaluated current pharmacologic treatment of migraine in adolescents.4–6 The consensus is that evidence supports the use of ibuprofen, acetaminophen (paracetamol), and sumatriptan NS for the acute treatment of migraine in such patients (although sumatriptan has not received FDA approval for this indication). Sumatriptan NS is the most extensively studied triptan in adolescent migraine, and the results from controlled clinical trials with this agent illustrate the difficulty of proving the efficacy of triptans in the presence of high placebo response rates. In 2 of the largest placebo-controlled studies (both with several hundred patients enrolled), sumatriptan (20 mg) NS failed to achieve statistical significance for the primary end points of 2-hour headache response15 and 1-hour headache response.16 In the present study, however, the use of a novel trial design served to eliminate early placebo responders. Consequently, the observed placebo response rate was ∼43%, and a significant difference between zolmitriptan NS and placebo was achieved for the primary end point.
With respect to the placebo challenge, a 15-minute response time was adopted to provide patients with fast onset of relief from their migraine. If a 30-minute response time had been used, then patients would have had to wait longer (ie, 30 vs 15 minutes) before receiving therapy that might relieve their headache, which is an important consideration, given the briefer duration of migraine attacks in adolescents. The choice of a 15-minute response time for the placebo challenge was a compromise that took into account the design of the study and the study population.
Selection of the most clinically relevant end points for the adolescent migraine population is problematic, as demonstrated by the previous sumatriptan studies. Most clinical studies in this setting, for example, used the 2-hour headache response as the primary end point; however, given the generally shorter duration of adolescent migraine, compared with adult migraine, it may be more relevant to examine responses at earlier time points. This study suggests that the 1-hour headache response, coupled with the 2-hour sustained response, is a valuable end point. Success with this end point was associated with other highly desirable outcomes, such as reduction of associated symptoms, return to normal activities, less use of rescue medications, and freedom from pain.
In accordance with instructions from the US FDA, the present study also analyzed the efficacy of zolmitriptan NS according to a worst-case scenario in which all responders to the placebo challenge were treated as responders for the randomized placebo group and as nonresponders for the zolmitriptan arm. A worst-case imputation, however, unfairly devalues active drug at the expense of placebo, negating the objective of the placebo challenge. Furthermore, imbalances, which were prestudy concerns of the FDA in view of the crossover design, were not observed (Table 3). Not surprisingly, this analysis failed to demonstrate a significant benefit for zolmitriptan NS. Another potential problem with the worst-case procedure is that placebo responders (ie, patients with early resolution of the attack) are not the population of interest. In contrast, the analysis used in the present study seems to provide a more-appropriate assessment of the treatment effect of zolmitriptan NS in this challenging setting.
Zolmitriptan NS was well tolerated in the adolescent migraine population, and the profile of adverse events was similar to that seen for adult patients.8,9 The most frequently reported adverse event, namely, taste disturbance, was experienced by 6.5% of zolmitriptan NS recipients. This incidence is much lower than that reported previously for sumatriptan NS (20 mg), which is typically associated with taste disturbance in ∼25% of adolescent migraineurs.17 This uniquely designed, placebo-controlled study demonstrated that acute treatment with zolmitriptan NS provided fast and significantly effective relief of migraine symptoms for adolescent migraineurs.
This study was supported by AstraZeneca.
We thank Andree Rose, Steve Winter, and Vicki Oldfield, from Wolters Kluwer Health, who provided medical editing assistance and other writing support, funded by AstraZeneca.
- Accepted March 26, 2007.
- Address correspondence to Donald W. Lewis, MD, FAAN, FAAP, Division of Pediatric Neurology, Children's Hospital of the King's Daughters, Eastern Virginia Medical School, Norfolk, VA 23510. E-mail:
Financial Disclosure: Dr Lewis received research support from AstraZeneca, Ortho-McNeil, Abbott Laboratories, GlaxoSmithKline, Merck, and American Home Products. Dr Winner received research support from AstraZeneca, GlaxoSmithKline, Merck, Ortho-McNeil, and Allergan; he also participated in the speaker's bureaus for GlaxoSmithKline, Merck, Allergan, and Ortho-McNeil and acted as a consultant/advisory board member for GlaxoSmithKline, Merck, Allergan, Ortho-McNeil, and Eisai. Dr Hershey received research support from AstraZeneca, USB Pharma, Tishcon, and Ortho-McNeil. Dr Wasiewski was an employee of AstraZeneca at the time the study was conducted.
- ↵Abu-Arefeh I, Russell G. Prevalence of headache and migraine in school children. BMJ.1994;309 :765– 769
- ↵Lewis D, Ashwal S, Hershey A, et al. Practice parameter: pharmacological treatment of migraine headache in children and adolescents: report of the American Academy of Neurology Quality Standards Subcommittee and the Practice Committee of the Child Neurology Society. Neurology.2004;63 :2215– 2224
- Damen L, Bruijn JK, Verhagen AP, Berger MY, Passchier J, Koes BW. Symptomatic treatment of migraine in children: a systematic review of medication trials. Pediatrics.2005;116 :295– 302
- ↵Charlesworth B, Dowson A, Purdy A, Becker WJ, Boes-Hansen S, Farkkila M. Speed of onset and efficacy of zolmitriptan nasal spray in the acute treatment of migraine: a randomised, double-blind, placebo-controlled, dose-ranging study versus zolmitriptan tablet. CNS Drugs.2003;17 :653– 667
- ↵Linder S, Winner P, Wasiewski W, Ryan S, Kågedal M. Pharmacokinetics of zolmitriptan nasal spray in adult and adolescent migraineurs between migraine attacks [abstract]. J Head Pain.2004;5 :56
- ↵Carey VC, Zeger SL, Diggle PJ. Modelling multivariate binary data with alternating logistic regressions. Biometrika.1993;80 :517– 526
- ↵Winner P, Rothner AD, Saper J, et al. A randomized, double-blind, placebo-controlled study of sumatriptan nasal spray in the treatment of acute migraine in adolescents. Pediatrics.2000;106 :989– 997
- Copyright © 2007 by the American Academy of Pediatrics