Presented in part at the annual meetings of the American Academy of Neurology, Toronto, ON, Canada, April 17-24, 1999; the American Association for the Study of Headache, Boston, Mass, June 11-13, 1999; and the International Headache Congress, Barcelona, Spain, June 22-26, 1999.
Sumatriptan for the Range of Headaches in Migraine Sufferers: Results of the Spectrum Study
Version of Record online: 7 JUL 2008
Headache: The Journal of Head and Face Pain
Volume 40, Issue 10, pages 783–791, December 2000
How to Cite
Lipton, R. B., Stewart, W. F., Cady, R., Hall, C., O'Quinn, S., Kuhn, T. and Gutterman, D. (2000), Sumatriptan for the Range of Headaches in Migraine Sufferers: Results of the Spectrum Study. Headache: The Journal of Head and Face Pain, 40: 783–791. doi: 10.1111/j.1526-4610.2000.00143.x
- Issue online: 7 JUL 2008
- Version of Record online: 7 JUL 2008
Background.—Migraineurs experience a spectrum of headaches: migraine, migrainous, and episodic tension-type as defined by the International Headache Society (IHS).
Objective.—To evaluate the effectiveness of sumatriptan, 50-mg tablets, in treating the spectrum of headaches in IHS-diagnosed migraineurs.
Design/Methods.—Migraineurs with severe disability (Headache Impact Questionnaire score 250 or greater) were enrolled in a randomized, double-blind, placebo-controlled, crossover study. Patients treated up to 10 headaches with sumatriptan, 50 mg, or placebo (4:1). Headache features, recorded prior to treatment, were used to classify each headache using IHS criteria. Headache response (moderate or severe pain reduced to mild or no pain) and pain-free response were recorded at 2 and 4 hours postdose (primary endpoint). Because patients treated multiple attacks, statistical methods controlling for within-subject correlation were used.
Results.—Two hundred forty-nine migraineurs treated 1576 moderate or severe headaches: migraine (n = 1110), migrainous (n = 103), and tension-type (n = 363). Sumatriptan was superior to placebo for headache response 4 hours postdose (primary endpoint) across all headache types (migraine, 66% versus 48%; P<.001; migrainous, 71% versus 39%; P<.01; tension-type, 78% versus 50%, P<.001). Sumatriptan was also superior to placebo for pain-free response 4 hours postdose for migraine (41% versus 24%, P<.001) and tension-type headaches (56% versus 36%, P = .001). Sumatriptan provided superior pain-free response 2 hours postdose for migraine (18% versus 7%, P<.0001) and tension-type headache (28% versus 14%, P = .0005) compared with placebo.
Conclusion.—Sumatriptan, 50-mg tablets, are effective for the full spectrum of headaches experienced by patients with disabling migraine due to a sumatriptan-responsive mechanism.
episodic tension-type headache
Headache Impact Questionnaire
Migraine sufferers experience a spectrum of headaches that includes migraine and nonmigraine headaches,1 specifically: migraine with (IHS 1.2) and without aura (IHS 1.1), migrainous (IHS 1.7), and episodic tension-type headache (ETTH) (IHS 2.1).2-7 In one population-based study of migraineurs, 49% of the reported headaches were migraine, 31% were migrainous, and 20% were tension-type headache.6 Some investigators hypothesize that the variety of headaches experienced by migraine sufferers are manifestations of the same underlying pathophysiologic substrate and that these headache types differ in degree, rather than kind.4,5,8 This is sometimes called the continuum model. The alternate dichotomous view postulates that these headache types are distinct pathophysiologic entities.1
One strategy of testing these competing models is to determine if migraine, migrainous, and ETTH headaches differ in treatment response to migraine-specific drugs. The continuum model would predict that these three pathophysiologically related headache types should all respond to triptans based on their mechanistic similarities. The dichotomous model predicts that ETTH should not respond to triptans, as it is mechanistically distinct. A triptan response in ETTH would arise under the dichotomous model only as the result of a coincidental overlap of mechanisms in the two distinct disorders. Available controlled clinical trials do not adequately address these issues.
Most clinical studies of acute migraine therapies require that patients selectively treat attacks that are clearly migraine with moderate or severe pain intensity. Within this context, the efficacy and tolerability of sumatriptan tablets is well-established for the treatment of moderate to severe migraine pain and migraine-associated symptoms.9-13 Little is known about the efficacy of sumatriptan in nonmigraine headaches experienced by migraine sufferers, although these headaches are common and often produce severe disability.
A retrospective analysis from an open-label study examined the responsiveness of nonmigraine headaches to sumatriptan, 6-mg injection, in 43 patients with IHS-diagnosed migraine of at least moderate pain intensity.14 Pretreatment headache symptoms were captured in diaries and an independent headache specialist classified attacks into headache types. Headache response rates were over 95% 2 hours postdose for migraine, migrainous, and ETTH and did not differ by headache type. While not placebo controlled, these results suggested that the full spectrum of headaches experienced by IHS-diagnosed migraineurs might be responsive to treatment with sumatriptan.
Given the frequency of nonmigraine headaches in migraine sufferers and the efficacy of sumatriptan for these headaches in an open-label study, we conducted a randomized, double-blind, placebo-controlled, clinical study to determine the efficacy of sumatriptan, 50-mg tablets, for the treatment of the spectrum of headaches experienced by patients with disabling IHS-diagnosed migraine. The treatment response to sumatriptan across headache types has implications for the continuum hypothesis and for clinical practice. Because migraine sufferers with disabling headache have the greatest documented need for treatment with 5-HT1 agonists,15-19 this study focused on IHS-diagnosed migraineurs with severe disabling headache as defined by a Headache Impact Questionnaire (HIQ) score of 250 or greater.20,21
The study was conducted at three centers in the United States between January 1997 and February 1998. Adults 18 to 65 years of age who met IHS criteria for migraine with (1.2) or without aura (1.1)1 were eligible for study participation. Patients completed written informed consent prior to study enrollment. Patients were required to have had a 1-year history of migraine with onset before the age of 50 years and to have experienced at least one migraine, but no more than 10 migraine attacks per month in the 6 months prior to study enrollment. This study also included patients with clinical diagnoses of migrainous headache and ETTH and will be the subject of an independent publication (work in progress).
Only patients who were disabled by their migraines, as defined by an HIQ score of 250 or greater at screening, were eligible for study enrollment. The HIQ provides a summary measure of pain and activity limitations due to headache in three domains (work, chores, and social, family, or leisure activity) over a 3-month period. The instrument has demonstrated reliability and validity in population-based studies.20,21 A score of 250 or greater (grades III and IV) targets the most severely affected upper half of migraine sufferers.
Patients were excluded from the study if they had a history of confirmed or suspected ischemic heart disease; uncontrolled hypertension; organic heart or central nervous system disease; seizure disorder requiring anticonvulsant medication; headache as a result of traumatic head or neck injury; impaired hepatic or renal function; history of migraine variants; evidence of alcohol, drug, or substance abuse within the previous year; or current monoamine oxidase inhibitor (MAOI) therapy. Women were excluded if they were pregnant, breast-feeding, or of childbearing potential and not using adequate contraceptive measures.
Patients were primarily identified from the general population and from specialty and primary care practices. Screening was conducted in person or by telephone while patients were migraine-free. Patients were evaluated clinically to determine final study eligibility. They provided medical and migraine histories and received a complete physical examination, laboratory tests (hematology, chemistry), electrocardiogram if clinically indicated, and measurement of vital signs (visit 1). Women of childbearing potential had received a urine pregnancy test. Eligible patients were then assigned to treatment groups using a computer-generated randomization such that sumatriptan, 50 mg, or identical-appearing placebo was given in a series of five headaches in a ratio of 4:1. Half of the patients were randomly assigned to receive placebo as the first dose. The remaining half was randomly assigned to receive placebo for treated attacks 2, 3, 4, or 5. Randomization schedules and patient assignments were sealed, maintained by the sponsor, and remained intact throughout the study.
Patients were instructed to treat up to 10 moderate or severe headaches over a 6-month period, whether or not the headaches were migraine. A 24-hour headache-free interval was required between treated headaches. Patients were instructed not to use rescue medication for 4 hours postdose, if possible. Patients recorded pretreatment characteristics of each treated headache in a diary. Investigators used IHS criteria to retrospectively classify each headache as either migraine, migrainous, or ETTH using pretreatment headache characteristics recorded in the migraine diaries, without access to information regarding treatment response.
Patients returned to the clinic 12 weeks after study enrollment (visit 2). Those who treated at least one headache during this period were followed for an additional 12 weeks or until 10 headaches were treated. All patients completed an exit visit (visit 3) after 6 months or the 10th treated headache. All previously unretrieved diaries were collected at visit 3 and patients were queried for adverse events. A final physical examination and urine pregnancy test, if applicable, were also performed.
Patients rated head pain using a 4-point pain scale (0 = no pain, 1 = mild pain, 2 = moderate pain, 3 = severe pain) and recorded head pain predose and 2, 4, and 24 hours postdose. Headache response was defined as a reduction in predose head pain from moderate or severe to mild or no pain (pain scale of 2 or 3 reduced to 1 or 0). Pain-free response was defined as a reduction in predose head pain from moderate or severe to no pain (pain scale of 2 or 3 reduced to 0). The primary efficacy outcome was headache response 4 hours postdose. Headache response 2 hours postdose and pain-free response 2 and 4 hours postdose were also recorded and analyzed as secondary endpoints.
Headache recurrence, defined as the worsening of head pain (moderate to severe pain) following a period of headache response (mild or no pain) within 4 to 24 hours posttreatment, was evaluated by headache type and treatment group. The use of rescue medication was also recorded. Associated symptoms were recorded at baseline to assist in the classification of headache type, but were not examined postdose due to their established responsiveness to sumatriptan.10-12
Safety and Tolerability.
Safety and tolerability were assessed by treatment-emergent adverse events (AEs) recorded in the treatment diary. An AE was defined as any untoward medical occurrence following the use of study drug. Serious AEs were defined as any medical occurrence that was fatal, life-threatening, disabling, required hospitalization, or causing congenital anomaly in the offspring of a patient.
Diary data from a population-based sample of migraineurs were used to estimate the proportion of each headache type that could occur in the study.6 Given these estimates, we predicted that 49% of treated headaches would be migraine, 31% migrainous, and 20% ETTH. Assuming that one of five headaches would be treated with placebo and a 30% placebo response would occur, approximately 200 patients were required to detect a statistically significant difference (80% power, a two-tailed α = .05).
Data were analyzed using generalized estimating equations (GEE)22 to adjust for within-patient correlation of treatment response across attacks and derive adjusted estimates of treatment response. Statistical efficacy analyses were performed on the intent-to-treat population that consisted of all patients who treated at least one headache and returned evaluable data. Last observation carried forward (LOCF) was used for missing values in otherwise evaluable patients. Patients who took rescue medication within 4 hours of using study medication were considered treatment failures. Odds ratios (ORs) contrasting sumatriptan and placebo by headache type were computed for each endpoint; a test of the hypothesis that these ORs were equal was performed using the Wald chi-square test assuming asymptotic normality of the estimates.22 Descriptive statistics are presented for patient demographics, headache recurrence, and safety and tolerability; no tests of statistical significance were performed on these parameters.
Three hundred eleven migraineurs were randomized and treated 1714 headaches and comprise the safety population. Unexpectedly, 47 headaches were treated while the pain intensity was mild; these attacks were excluded from analysis. In addition, 91 headaches were treated with medication prohibited by the protocol and were also excluded from the analysis. Two hundred forty-nine patients treated a total of 1576 moderate or severe headaches and comprise the intent-to-treat population. Figure 1 summarizes the disposition of patients and headaches throughout the study.
Two hundred fifteen patients completed the study. Among the 34 patients who did not complete the study, 15 patients were lost to follow-up, 6 patients withdrew consent, 7 patients withdrew due to AEs, 3 patients moved, and 3 patients had incomplete information. One hundred two patients (47%) treated 10 headaches, 65 patients (30%) treated 5 to 9 headaches, and 48 patients (22%) treated 4 or less headaches. Seventy-eight percent (n = 167) of all patients received at least one dose of placebo.
The intent-to-treat population was composed predominantly of women (86%) between the ages of 30 and 49 years (mean age, 38.1 years). Eighty-nine percent of the population was white, 10% black, and less than 1% multiracial. The breakdown of attack type differed from expectations. A higher proportion of headache attacks were migraine (70%, 1110/1576), while only 7% (103/1576) of headaches were migrainous, and 23% (363/1576) were ETTH. There was no association between headache type and patient demographic characteristics. Seventy-nine percent of the headaches were treated with sumatriptan, 50 mg.
Sumatriptan, 50 mg, provided superior headache response 4 hours postdose across all headache types compared with placebo. Specifically, adjusted ORs and 95% confidence intervals (CIs) for sumatriptan versus placebo were: migraine, 2.12 (CI, 1.59 to 2.82), P<.001; migrainous, 3.81 (CI, 1.43 to 10.12), P<.01; ETTH, 3.62 (CI, 2.13 to 6.13), P<.0001. Migraine treated with sumatriptan, 50 mg, also had superior headache response at 2 hours postdose—adjusted OR, 1.73 (CI, 1.27 to 2.34), P<.001, compared with placebo. Though the effect of sumatriptan was similar in magnitude for the migrainous group at 2 hours, the OR differences did not reach statistical significance (adjusted OR, 2.15; CI, 0.77 to 5.95; P = .14). Similarly for ETTH, at 2 hours sumatriptan did not have a statistically significant effect (adjusted OR, 1.45; CI, 0.86 to 2.43; P = .16). The percentage of headaches with headache response 2 and 4 hours postdose are presented in Figure 2 by headache type.
Significantly more sumatriptan-treated migraine and ETTH attacks achieved pain-free response compared with placebo 4 hours postdose—adjusted ORs: migraine, 2.18 (CI, 1.60 to 2.97), P<.0001; ETTH, 2.20 (CI, 1.36 to 3.56), P = .001. The effect size was similar in magnitude for the migrainous group, but differences at 4 hours were not statistically significant (adjusted OR, 2.17; CI, 0.68 to 6.92; P = .19). Similarly, 2 hours postdose, significantly more sumatriptan-treated migraine and ETTHs were pain-free compared with placebo—adjusted ORs: migraine, 2.92 (CI, 1.76 to 4.86), P<.0001; ETTH, 2.40 (CI, 1.31 to 4.40), P = .005. Differences were not statistically significant for the migrainous group (adjusted OR, 4.67; CI, 0.47 to 46.6; P = .19) at the 2-hour time point. The percentage of headaches with pain-free response 2 and 4 hours postdose are presented in Figure 3 by headache type.
Efficacy did not differ significantly by headache type (headache response at 4 and 2 hours, P = .51 and P = .74; pain-free response at 4 and 2 hours, P = .99 and P = .81, respectively), indicating that the efficacy of sumatriptan was independent of headache type among migraine sufferers.
Overall headache recurrence in the sumatriptan-treated headaches was low and varied from 9% to 13% across the headache types. Headache recurrence was lower in headaches with no pain (pain-free) 4 hours postdose than in headaches with mild pain (headache response) at the same time point (Table 1). Seven percent (112/1576) of all treated headaches were treated with rescue medication, the majority of which were analgesic medications (Table 2). Rescue medication use among sumatriptan-treated headaches did not differ by headache type.
|Overall Headache Recurrence||Headache Recurrence Post Pain-Free Response|
|Headache Type||Placebo||Sumatriptan, 50 mg||Placebo||Sumatriptan, 50 mg|
|Migraine||13/114 (11)||77/578 (13)||3/60 (5)||41/373 (11)|
|Migrainous||0/7 (0)||6/65 (9)||0/6 (0)||3/44 (7)|
|Episodic tension-type headache||4/38 (11)||24/219 (11)||2/26 (8)||14/152 (9)|
|Total||17/159 (11)||107/862 (12)||5/92 (5)||58/569 (10)|
|Headache Type||Sumatriptan 50 mg||Placebo|
|Migraine||61/870 (7)||20/240 (8)|
|Migrainous||6/86 (7)||3/17 (8)|
|Episodic tension-type headache||14/283 (5)||8/80 (10)|
Safety and Tolerability.
Overall, 6% (19/356) of the placebo-treated headaches and 7% (83/1358) of sumatriptan-treated headaches in the safety population were classified as having potentially drug-related AEs. Eleven of the drug-related AEs were judged as severe in intensity and included vomiting, pain in the side (body), lightheadedness, nausea and vomiting, ringing in the ears, chest tightness and difficulty swallowing (same patient), facial flushing, rapid breathing/pulse rate, heart palpitations and head pounding (same patient). All drug-related AEs, except vomiting, followed administration of active drug and all patients recovered without sequelae. The incidence of AEs did not increase across multiple attacks treated with sumatriptan. No clinically significant laboratory, vital sign, or ECG changes were reported.
Seven patients withdrew from the study due to drug-related AEs described as mild or moderate in intensity. These AEs included drowsiness (mild), nausea (mild), muscle aches (mild), tingling in hands/arms (mild), and irregular heart beat (moderate). One serious AE (leukemia), judged as unrelated to study drug, was reported during the study.
Headache sufferers with disabling IHS-diagnosed migraine experienced a spectrum of headaches during the present study that included migraine, migrainous, and ETTH. At the primary endpoint, headache response at 4 hours, sumatriptan, 50-mg tablets, were significantly more effective at treating all of these headache types than placebo. Sumatriptan was also significantly more effective than placebo for headache response at 2 hours in migraine and pain-free response for both migraine and ETTH 2 and 4 hours postdose. Response to sumatriptan in migrainous headaches did not reach statistical significance for the secondary endpoints, 2-hour headache response or 2 and 4 hours pain-free response, most likely due to the small sample size. Nonetheless, the magnitude of the treatment effect was similar to that observed for migraine and ETTH. In addition, headache response was found to be independent of headache type in individuals with disabling IHS-diagnosed migraine. Given the statistically significant findings for the primary endpoint and the supportive but nonsignificant trends in secondary endpoints, we conclude that sumatriptan was effective in migrainous headache, as well as migraine and ETTH.
Though there are no comparable randomized, placebo-controlled trials, these results are similar to a retrospective open-label study by Cady et al,14 that demonstrated high efficacy rates with sumatriptan injection for migraine, migrainous, and ETTH in diagnosed migraineurs. In comparison with other studies of sumatriptan where only migraines were studied, the efficacy rates reported in the present study fell at the lower end of the previously reported ranges.9-13,23 This finding may be related to our focus on severe disabling migraine and is consistent with the linear relationship reported for headache severity and efficacy for another 5-HT1 agonist.24
Sumatriptan, 50 mg, was well tolerated across all headache types in the present study, with safety results comparable to placebo. In general, the AEs were mild or moderate, transient, and similar to those reported in other sumatriptan clinical studies.9,11,13,23 The incidence of AEs did not increase across multiple attacks treated with sumatriptan, reflecting the long-term safety of the tablet formulation.
Headache recurrence and use of rescue medications was lower in the present study than previously reported,10,23 especially given the severity disability of the patients. The reason for these findings is unknown. Interestingly, headache recurrence was lowest in the patients treated with sumatriptan who were pain-free 4 hours postdose, supporting observations that obtaining complete relief reduces headache recurrence following 5-HT1 agonist treatment.25
Several explanations may account for the robust response of ETTH to sumatriptan observed in the present study. Perhaps the attacks identified as ETTH were actually misclassified attacks of migraine; at the biological level, this may be true. But the IHS criteria classify primary headaches based on symptoms and the exclusion of other disorders. Our patients recorded their headache symptoms in treatment diaries and withheld treatment until their head pain was at least moderate in intensity. At that point in a migraine attack, one would expect the development of migraine-associated symptoms. When diaries were reviewed and investigators assigned each treated attack to one of three categories using rigorous IHS criteria, headaches classified as ETTH clearly met IHS criteria for this disorder. A second possibility is that sumatriptan is an effective treatment both for migraine and ETTH, even though these disorders are pathophysiologically distinct. This explanation seems unlikely based on the poor response of ETTH to sumatriptan in nonmigraine headache sufferers.26 In addition, in the ETTH arm of this study patients with IHS-diagnosed ETTH did not respond to sumatriptan (work in progress). A third and more likely explanation is that the attacks of ETTH and the attacks of migraine were pathophysiologically related, even though their symptom profiles varied. This explanation is consistent with the hypothesis that among migraine sufferers, ETTH and migraine are part of a single pathophysiologic continuum.8,27-30
The present study enhances our understanding of the pathogenesis of migraine as well as the mechanism of action of sumatriptan. The IHS classification system postulates that migraine and ETTH are biologically distinct disorders with differing pathophysiology. Given this view, ETTH would respond to 5-HT1 agonists only as a result of a coincidental convergence of biological mechanisms. The continuum model of headache8,27-30 places ETTH, migrainous, and migraine along a spectrum of severity, based upon the existence of a common biological mechanism. This model would predict that ETTH should respond to sumatriptan in a manner similar to migraine. Our results support the continuum model and suggest that among migraine sufferers, the full spectrum of headaches responds to sumatriptan. However, our findings do not address ETTH in individuals who do not have IHS-diagnosed migraine. The available evidence suggests that ETTH in individuals without IHS migraine does not have a robust response to sumatriptan.26 We conclude that migraine and ETTH are biologically distinct, but that the full spectrum of headaches in disabled migraine sufferers has a triptan-responsive mechanism.
These results have important implications for clinical practice. Patients with migraine are routinely counseled not to use sumatriptan until they are certain that their headache is migraine. Our data demonstrate that from an efficacy perspective, these instructions are unnecessary as sumatriptan effectively treated nonmigraine headaches in persons with IHS migraine. Additionally, a post hoc subgroup analysis from the present study demonstrated that in IHS migraine sufferers, sumatriptan was very efficacious for the treatment of mild head pain.31 Together, these results suggest that sumatriptan may be useful in the early treatment of disabling migraine. This issue should be studied in a specifically designed study. We predict that in a subgroup of migraineurs whose headaches are disabling, early intervention with a single dose of sumatriptan may be an efficacious treatment strategy.
Acknowledgments: This study was sponsored by a grant from Glaxo Wellcome, Inc, Research Triangle Park, NC. Appreciation is expressed to Barbara Wilson for preparing this manuscript.
- 15Impact of oral sumatriptan on workplace productivity, health-related quality of life, healthcare use, and patient satisfaction with medication in nurses with migraine. Am J Managed Care. 1996;2:1407-1417., & , et al.
- 17Work productivity and resource consumption among migraineurs under current treatment and during treatment with sumatriptan—an economic evaluation of acute treatment in moderate to severe migraineurs. Headache Q. 1997;8:237-246.& .
- 24The International 311C90 Long-term Study Group. The long-term tolerability and efficacy of oral zolmitriptan (Zomig, 311C90) in the acute treatment of migraine. An international study. Headache. 1998;38:173-183.
- 29Headache. In: BonicaJ, Loeser, Chapman, Fordyce, eds. The Management of Pain. Philadelphia, Pa: Lea & Febiger; 1990:687-726.& .
- 31Treatment of mild pain with sumatriptan in the spectrum of headache patients with disabling IHS migraine [abstract]. Neurology. 1999;52:A205-A206., , , & .