ISSN 0017-8748 Headache doi: 10.1111/head.12499 © 2015 American Headache Society Published by Wiley Periodicals, Inc. American Headache Society Evidence Assessment

The Acute Treatment of Migraine in Adults: The American Headache Society Evidence Assessment of Migraine Pharmacotherapies

Michael J. Marmura, MD; Stephen D. Silberstein, MD, FACP; Todd J. Schwedt, MD, MSCI

The study aims to provide an updated assessment of the evidence for individual pharmacological therapies for acute migraine treatment. Pharmacological therapy is frequently required for acutely treating migraine attacks. The American Academy of Neurology Guidelines published in 2000 summarized the available evidence relating to the efficacy of acute migraine medications. This review, conducted by the members of the Guidelines Section of the American Headache Society, is an updated assessment of evidence for the migraine acute medications. A standardized literature search was performed to identify articles related to acute migraine treatment that were published between 1998 and 2013. The American Academy of Neurology Guidelines Development procedures were followed. Two authors reviewed each abstract resulting from the search and determined whether the full manuscript qualified for review. Two reviewers studied each qualifying full manuscript for its level of evidence. Level A evidence requires at least 2 Class I studies, and Level B evidence requires 1 Class I or 2 Class II studies. The specific medications – triptans (almotriptan, eletriptan, frovatriptan, naratriptan, rizatriptan, sumatriptan [oral, nasal spray,injectable, transcutaneous patch], zolmitriptan [oral and nasal spray]) and dihydroergotamine (nasal spray,inhaler) are effective (Level A). Ergotamine and other forms of dihydroergotamine are probably effective (Level B). Effective nonspecific medications include acetaminophen, nonsteroidal anti-inflammatory drugs (aspirin, diclofenac, ibuprofen, and naproxen), opioids (butorphanol nasal spray), sumatriptan/naproxen, and the combination of acetaminophen/aspirin/caffeine (Level A). Ketoprofen, intravenous and intra- muscular ketorolac, flurbiprofen, intravenous magnesium (in migraine with aura), and the combination of isometheptene compounds, codeine/acetaminophen and tramadol/acetaminophen are probably effective (Level B). The antiemetics prochlorperazine, droperidol, chlorpromazine, and metoclopramide are probably effective (Level B). There is inadequate evidence for butalbital and butalbital combinations, phenazone, intravenous tramadol, methadone, butorphanol or meperidine injections, intranasal lidocaine, and corticosteroids, including dexamethasone (Level C). Octreotide is probably not effective (Level B). There is inadequate evidence to refute the efficacy of ketorolac nasal spray, intravenous acetaminophen, chlorproma- zine injection, and intravenous granisetron (Level C). There are many acute migraine treatments for which evidence supports efficacy.Clinicians must consider medication efficacy,potential side effects,and potential medication-related adverse events when prescribing acute medications for migraine. Although opioids, such as butorphanol, codeine/acetaminophen, and tramadol/ acetaminophen, are probably effective, they are not recommended for regular use.

Key words: migraine, acute treatment, pharmacology, episodic migraine, clinical trial

(Headache 2015;55:3-20)

Migraine is a highly prevalent disorder character- From the Department of Neurology, Jefferson Headache ized by attacks of moderate to severe throbbing head- Center, Thomas Jefferson University, Philadelphia, PA, USA aches that are often unilateral in location, worsened (M.J. Marmura and S.D. Silberstein); Department of Neurol- by physical activity, and associated with nausea ogy, Mayo Clinic Arizona, Scottsdale, AZ, USA (T.J. Schwedt). and/or vomiting, photophobia, and phonophobia.1,2 Address all correspondence to Michael J. Marmura, 900 Migraine treatment can include preventive therapy Walnut Street Suite 200, Philadelphia, PA 19107, USA. aimed at reducing the frequency and severity of Accepted for publication November 26, 2014. migraine attacks, as well as acute therapy used to

3 4 January 2015

Figure.—Clinical guidelines process. abort a migraine attack. In association with the the following question:Which pharmacological thera- American Headache Society, the American Academy pies are effective in treating acute migraine? of Neurology (AAN) has recently published guide- lines for preventive treatment.3 The last AAN Guide- DESCRIPTION OF THE ANALYTIC lines for acute treatment were published in 2000.4 PROCESS Herein we report the results of an updated sys- The American Headache Society Guidelines tematic review of the published literature addressing Committee performed this project using the AAN the efficacy of medications used for acute treatment protocol for systematic development of clinical guide- of migraine. These studies of acute migraine medica- lines (Figure).6 The author panel comprised headache tions use 1 or more of multiple possible end-points of experts. The members of the guidelines group dis- efficacy, including headache relief (ie, reduction from closed any conflict of interest prior to involvement. severe or moderate intensity to mild or none), head- Persons with a substantial conflict of interest, based ache freedom, decreased disability, the absence of on a portion of their incomes pertinent to the study, nausea or vomiting, and the absence of photophobia were excluded. Although this project began before or phonophobia. Outcomes have been measured at the publication of the new Institute of Medicine varying intervals following medication administra- (IOM) Clinical Practice Guidelines, our methods tion. However, the International Headache Society were largely consistent with the IOM methods, for we Clinical Trial Guidelines published in 2012 suggests attempted to follow these principles of guideline that the percentage of study participants headache- development, rating of evidence, rating strength, and free at 2 hours should usually be used as the primary external review.7 The clinical question was to deter- outcome in acute therapy trials.5 Sustained pain mine which acute treatments are effective for freedom with the absence of other migraine symp- migraine. Migraine was defined using the Interna- toms at 24 or 48 hours is also an important patient- tional Classification of Headache Disorders, 2nd centered outcome and is considered the “ideal” edition diagnostic criteria. We performed formal migraine treatment response.5 Since the last AAN EMBASE and Medline searches up to December Guidelines for acute treatment in 2000, multiple 2013 using predefined search terms to capture all fully large, randomized acute pharmacological migraine published clinical trials relevant to the subject. Search treatment clinical trials have been conducted.4 This terms included “migraine,” “unilateral headache,” updated assessment of the evidence seeks to answer “hemicranias,” or “cephalalgia,” combined with

Conflict of Interest: Dr. Marmura has received royalty income from Demos Medical, Cambridge University Press, and MedLink Neurology. Dr. Schwedt has received consulting fees and/or honoraria from Allergan, Inc., Supernus, and Zogenix. He has received royalty income from UpToDate. Dr. Silberstein has received consulting fees and/or honoraria from Allergan, Inc., Amgen, Avanir Pharmaceuticals, Inc., eNeura Inc, ElectroCore Medical LLC, Medscape, LLC, Medtronic, Inc, Mitsubishi Tanabe Pharma America, Inc., Neuralieve, NINDS, Pfizer, Inc, Supernus Pharmaceuticals, Inc., and Teva Pharmaceuticals. His employer, Jefferson University Hospitals, receives research support from Allergan, Inc, Amgen, Cumberland Pharmaceuticals, Inc, ElectroCore Medical, LLC, Labrys Biologics, Eli Lilly and Company, Merz Pharmaceuticals, and Troy Healthcare. Financial Support: Supported by the American Headache Society. Headache 5

“acute” or “immediate” and “drug therapy” or “phar- • Level A: Established as effective (or ineffective) for macotherapy.” We excluded animal studies, non- acute migraine (supported by at least 2 Class I English-language articles, and comments, letters, or studies) editorials, and we removed duplicate articles. We did • Level B: Probably effective (or ineffective) for not consider unpublished data, gray literature, or con- acute migraine (supported by 1 Class I study or 2 ference abstracts for the purpose of this review. The Class II studies) search strategy may be found in Appendix 1. Two • Level C: Possibly effective (or ineffective) for acute study team members reviewed each of the abstracts migraine (supported by 1 Class II study or 2 Class that resulted from the search, and then independently III studies) determined if the study should be excluded from • Level U: Evidence is conflicting or inadequate to further review or if the full manuscript should be support or refute the use of the medication(s) for reviewed. Two reviewers then independently evalu- acute migraine ated each of the studies selected for full-text review and determined whether the study met criteria for TheAAN review published in 2000 did not use the inclusion. In the event of disagreement between the 2 current AAN therapeutic classification. Group 1 con- reviewers, a third study team member reviewed the sisted of medications with at least 2 double-blind, abstract and/or full manuscript to determine inclus- placebo-controlled studies supporting their use, while ion or exclusion. We rated each study based on the group 2 only required 1 study.The concept of Class I or quality of study design, considering the study size and Class II studies was not considered. Because a single length of treatment, treatment technique and doses, Class II study only supports a Level C recommenda- methods of data collection (prospective or retrospec- tion in the new AAN classification, many of the previ- tive), presence or absence of placebo, primary and ous group 2 medications are Level C in this review. secondary outcomes, and adverse events. Although we did not exclude open-label or observational ANALYSIS OF EVIDENCE studies, Class I and II studies required a placebo The original search produced 805 articles, of control arm. We considered dropout rates to be less which 132 were selected for review. Studies were relevant in these studies since many investigated a excluded from this report if they met any of the fol- single migraine attack. Both reviewers completed a lowing criteria: standardized data extraction form for each study (Appendix 2). The rating of each study followed rec- • Did not include adult subjects ommendations according to the AAN therapeutic • Did not evaluate medication (ie, were medical classification of evidence scheme, ranging from Class devices or procedures) I to Class IV. Class I studies, well-designed double- • Assessed treatment of migraine aura or preventive blind, randomized, placebo-controlled trials, were treatment (ie, prevention of menstrual migraine) considered best, while Class IV studies were often • Evaluated medications that are neither currently retrospective studies or case reports with unclear out- commercially available in the United States nor comes data. For a detailed description of the evidence pending approval scheme, see Appendix 3. • Used nonstandardized primary outcome measures, When we found no new studies for a particular such as patient satisfaction or disability drug, ratings were based on previous AAN Guide- • Focused on comparisons between 2 or more medi- lines. In the event of conflicting evidence, meaning at cations, rather than placebo least 1 negative and 1 positive study for a particular medication, we considered the Class I studies to be more important in assigning a level of evidence. RESULTS Based on the quality of studies, a level of evi- We found no new Class I or II studies published dence was assigned for each drug, as follows: since the most recent guidelines for butorphanol (intra- 6 January 2015 muscular or nasal spray), combinations of butalbital/ looking at sumatriptan nonresponders, subjects aspirin/caffeine or butalbital/aspirin/caffeine/codeine, receiving almotriptan 12.5 mg, compared with acetaminophen/codeine, dihydroergotamine (DHE) placebo, were more likely to be headache-free at 2 (nasal spray, intramuscular, or intravenous), hours (47.5% vs 23.2% placebo; P < .05) and experi- flurbiprofen, hydrocortisone, isometheptene, intranasal ence headache relief at 2 hours (33% vs 14%; lidocaine, and meperidine.Thus, no new review of previ- P < .05).12 ous studies was done, and we assigned levels of evidence Eletriptan.—Two parallel-group, Class I placebo- for these agents based on the 2000 AAN Guidelines.4 controlled trials of 1 migraine attack treated with Currently, no Class I or II studies exist for the use of eletriptan demonstrated superiority against placebo intravenous diphenhydramine, intravenous valproate, for 20 mg, 40 mg, and 80 mg doses. Subjects in the first intravenous verapamil,oral tolfenamic acid,or celecoxib. study who received 20 mg, 40 mg, and 80 mg had Rofecoxib 25 mg and the combination of ergotamine/ 2-hour headache relief of 64%, 67%, and 76%, caffeine/pentobarbital/Bellafoline were included in the respectively.13 All studied doses were superior to the previous version of the guidelines but deleted from this placebo response of 51% (P < .05). The second study review as they are no longer available. revealed 2-hour headache relief rates of 47%, 62%, Almotriptan.—In a Class I study, subjects taking and 59% for the 20 mg, 40 mg, and 80 mg doses, almotriptan 2 mg, 6.25 mg, 12.5 mg, and 25 mg had respectively, compared with 22% for placebo 2-hour headache relief of 30%, 56.3%, 58.5%, and (P < .01).14 Headache freedom rates at 2 hours were 66.5%, respectively, compared with 32.5% with 14%, 27%, and 27% for the 20 mg, 40 mg, and 80 mg placebo (P < .05 for 6.25 mg, 12.5 mg, and 25 mg).8 doses, which were superior to the 4% rate of head- Adverse events were reported by 17.1%, 16.2%, ache freedom among subjects receiving placebo 18.3%, and 25.5% of the patients in the almotriptan (P < .01). In a Class I, 3-attack study comparing 40-mg 2 mg, 6.25 mg, 12.5 mg, and 25 mg groups compared and 80-mg eletriptan against placebo, both eletriptan with 15.0% in the placebo group. Another Class I doses were superior to placebo for headache relief at study examined the treatment of 3 migraine attacks.9 2 hours and headache freedom at 2 hours (40 mg: Subjects reported headache relief at 2 hours for at 62% headache relief, 32% headache freedom; 80 mg: least 2 of 3 attacks in 63.8% and 74.5% of subjects 65% headache relief, 34% headache freedom; taking 6.25 mg and 12.5 mg, respectively, in 3 separate placebo: 19% headache relief, 3% headache freedom; migraine attacks, compared with 35.7% of those using P < .001).15 placebo (P < .001). A Class I study of early treatment with eletriptan In another 4-arm, randomized, controlled Class I 20 mg, eletriptan 40 mg, or placebo for 1 migraine trial, rates of headache freedom at 2 hours when attack found headache-free rates at 2 hours of 22% treating early (within 1 hour) with 12.5-mg for placebo, 35% for 20 mg (P < .01 compared with almotriptan were 53% at 2 hours, and 38% when placebo), and 47% for 40 mg (P < .001 compared treating moderate to severe headache, compared with with placebo).16 Although subjects were encouraged 25% treating early and 17% treating late with to treat migraine early, early treatment was not placebo (P = .0004 for early treatment; P = .0002 for required. In those who were treated when headache late treatment).10 Adverse events were low (<5%), was mild, the 40-mg 2-hour headache-free rate was with no difference between active drug and placebo. 68%, compared with 25% for placebo (P < .001). In a separate Class I trial, subjects received Another double-blind, parallel-group, placebo- almotriptan 12.5 mg or placebo within 1 hour of head- controlled Class I study specifically studied eletriptan ache onset.11 At 2 hours after headache onset, in subjects with poor response or tolerability to almotriptan-treated patients were more likely to be sumatriptan.17 Subjects with up to 3 attacks were headache-free (37.0% vs 23.9% placebo; P = .010) treated with either eletriptan 40 mg, eletriptan 80 mg, and experience headache relief (72.3% vs 48.4% or placebo. Both eletriptan 40 mg and eletriptan placebo; P < .001). In a Class I study specifically 80 mg demonstrated consistency of response that was Headache 7 superior to placebo; headache relief rates at 2 hours the 10 mg, 5 mg, and placebo groups, respectively on at least 2 of 3 attacks in subjects taking eletriptan (P < .01). The 10 mg dose was more effective than 40 mg (66%), 80 mg (72%), or placebo (15%; 5mg(P < .05). Headache-free rates at 2 hours were P < .001). 42%, 35%, and 10% (P < .01). A Class I study that asked subjects to treat Two other Class I studies evaluated rizatriptan migraine with 80-mg eletriptan during the aura phase 10 mg vs placebo for the treatment of migraine of before headache onset failed to demonstrate superi- mild intensity treated within 1 hour of onset.24 Sub- ority against placebo. Sixty-one percent of subjects jects using rizatriptan were more likely to achieve who treated with eletriptan developed moderate-to- headache freedom at 2 hours (57.3% and 58.9% vs severe headache within 6 hours, vs 46% subjects who 31.1% and 31.1%; P < .001) and had 24-hour sus- treated with placebo (P = ns).18 tained headache freedom (42.6% and 48.0% vs Frovatriptan.—Two Class I combined parallel 23.2% and 24.6%; P < .001). dose-titration studies reported that frovatriptan Two Class I studies compared rizatriptan 10 mg 2.5 mg was more effective than placebo at 2 hours for and placebo (randomized 2:1) for the treatment of headache relief (40% vs 23%; P < .001).19 Doses 1 menstrually related migraine attack.25 Rizatriptan higher than 2.5 mg were no more effective at 4 hours was significantly more effective in terms of both (P = ns). 2-hour headache relief (70% and 73% vs 53% and A Class I dose-titration study comparing 50% for placebo; P < .001) and 24-hour sustained frovatriptan 0.5 mg, 1 mg, 2.5 mg, and 5 mg vs placebo headache freedom (46% and 46% vs 33% and 33%; found that 2.5 mg was more effective than placebo at P = .016 study 1; P = .024 study 2). 2 hours for headache relief (38% vs 25%; P < .05), A Class I study compared rizatriptan oral dissolv- while the other doses of frovatriptan were not supe- able tablets 10 mg with placebo in patients taking rior to placebo.20 All doses of frovatriptan were supe- topiramate for migraine prophylaxis.26 Subjects with 3 rior to placebo at 4 hours for headache relief (48% for attacks were treated in this randomized, placebo- 0.5 mg, 68% for 2.5 mg vs 33% for placebo; P < .02). controlled, double-blind, multiple-attack study: 2 with In a Class II crossover study of 2 migraine attacks, rizatriptan and 1 with placebo. Rizatriptan was supe- 1 with frovatriptan 2.5 mg, followed by placebo in 2 rior to placebo in terms of 2-hour headache relief hours if headache increased to moderate or severe, (55.0% vs 17.4%; P < .001), sustained headache relief and the other with placebo, followed by frovatriptan from 2 to 24 hours (32.6% vs 11.1%; P < .001), and 2.5 mg in 2 hours for moderate or severe headache, 2-hour headache freedom (36.0% vs 6.5%; P < .001). sustained headache freedom at 24 hours was more In a Class II study of sumatriptan nonresponders, common in subjects taking frovatriptan early (40%) rizatriptan 10 mg orally disintegrating tablet was compared with those taking it late (31%; P < .05).21 compared with placebo for the treatment of a single Naratriptan.—The 2000 AAN Guidelines con- migraine attack.27 To be included in the study, subjects cluded that naratriptan was effective for acute had to fail treatment with open-label generic migraine. Since that guideline, a Class I study evalu- sumatriptan 100 mg in the baseline phase. Subjects ated naratriptan 2.5 mg vs placebo for the treatment with 3 attacks were treated: 2 with rizatriptan and 1 of menstrually related migraine.22 Subjects taking with placebo in random order. In this population, naratriptan were more likely to be headache-free at 4 rizatriptan was superior for 2-hour headache freedom hours (58% vs 30%; P = .004). (22% vs 12%; P = .013), 2-hour headache relief (51% Rizatriptan.—Rizatriptan was established as effec- vs 20%; P < .001), and sustained headache freedom tive for the treatment of acute migraine in the 2000 from 2 to 24 hours (20% vs 11%; P = .036). AAN Guidelines. Since that guideline, a Class I trial Sumatriptan.—The 2000 AAN Guidelines estab- compared rizatriptan oral dissolvable tablets, 5 mg lished sumatriptan tablets 25 mg, 50 mg, and 100 mg, and 10 mg, vs placebo for the treatment of 1 attack.23 sumatriptan nasal spray, and sumatriptan injection Headache relief at 2 hours was 74%, 59%, and 28% in 4 mg and 6 mg as effective. 8 January 2015

In a Class I study of sumatriptan 50 mg and first showed superiority over placebo at 10 minutes 100 mg for 1 migraine attack compared with placebo, for headache relief (11% vs 6%; P = .039). 51.1% and 66.2% of subjects were headache-free at 2 A Class I study compared a new iontophoretic hours in the 50 mg and 100 mg groups, compared with patch formulation of sumatriptan (6.5 mg transder- 19.6% of subjects treating with placebo (P < .001).28 mal with delivery over 4 hours) with placebo for acute Two large Class I pooled studies that comprised migraine treatment.31 Sumatriptan patch was superior 2696 patients found sumatriptan was superior to to placebo patch for the primary end-point of head- placebo as early as 20 minutes with the 100 mg dose ache freedom at 2 hours (18% vs 9%; P = .0092). For for headache relief (6% vs 4% placebo; P ≤ .05) and secondary measures, active drug was superior at 2 at 30 minutes with the 50 mg dose (19% vs 14% hours for freedom from photophobia (51% vs 36%; placebo; P ≤ .05).29 Two-hour headache-free rates, P = .0028), freedom from phonophobia (55% vs 39%; considered a secondary outcome measure in this P = .00021), freedom from nausea (84% vs 63%; P = study, were 40% with 50 mg, 47% with 100 mg, and .0001), and headache relief (53% vs 29%; P = .0001). 15% with placebo (P < .01). Similar results were A Class I randomized, double-blind, parallel- observed for the individual studies. In study 1, group, placebo-controlled study compared the effec- sumatriptan tablets were significantly more effective tiveness of intranasal sumatriptan 10 mg, intranasal than placebo at 25 minutes with the 100 mg dose and sumatriptan 20 mg, with placebo for a single migraine at 50 minutes with the 50 mg dose. In study 2, attack.32 Patients were instructed to use a new bidi- sumatriptan tablets were significantly more effective rectional powder delivery device for a moderate to than placebo at 17 minutes for the 100 mg dose and at severe attack. Intranasal sumatriptan 10 mg (54% vs 30 minutes for the 50 mg dose (P ≤ .05). In the pooled 25%; P < .05) and 20 mg (57% vs 25%; P < .05) were data, the cumulative percentages of patients with pain both superior to placebo for headache freedom and relief by 2 hours after dosing were 72% for the headache relief at 2 hours (10 mg 84% vs 44%, 100 mg dose and 67% for the 50 mg dose, compared P < .001; 20 mg 80% vs 44%, P < .01). with 42% for placebo (P ≤ .001; both sumatriptan Sumatriptan/Naproxen Sodium.—Two large doses vs placebo). The cumulative percentages of studies compared a fixed combination of sumatriptan patients with a pain-free response by 2 hours were 85 mg and naproxen sodium 500 mg against 47% for the 100 mg dose, 40% for the 50 mg dose, and sumatriptan 85 mg alone, naproxen sodium 500 mg 15% for placebo (P ≤ .001; both sumatriptan doses vs alone, and against placebo.33 In study 1, 2-hour placebo). In the individual studies, significantly more headache-free rates were 34% for the sumatriptan patients receiving either sumatriptan dose were 85 mg/naproxen sodium 500 mg combination, migraine-free 2 hours after dosing, and had sustained compared with 25% for sumatriptan 100 mg, 15% pain relief and a sustained pain-free response over 24 for naproxen 500 mg, and 9% for placebo. The hours, compared with placebo (P ≤ .001; both sumatriptan/naproxen sodium combination was supe- sumatriptan doses vs placebo). The only drug-related rior to sumatriptan alone (P = .009), naproxen, and adverse events reported in less than 2% of patients in placebo (P < .001). The incidence of headache any treatment group in either study were nausea relief 2 hours after dosing was 65%, 55%, 44%, and (both studies: 3% sumatriptan 100 mg, 2% suma- 28% with sumatriptan/naproxen sodium, sumatriptan triptan 50 mg, 1% placebo) and paresthesia (study 1: monotherapy, naproxen sodium monotherapy, and <1% sumatriptan 100 mg, <1% sumatriptan 50 mg, placebo, respectively, in study 1 (P < .001 for 0% placebo; study 2: 3% sumatriptan 100 mg, 1% sumatriptan/naproxen sodium, sumatriptan, and sumatriptan 50 mg, <1% placebo). naproxen sodium vs placebo). In study 2, 2-hour A Class I study of sumatriptan injection 4 mg headache-free rates were 30% for the sumatriptan reported 2-hour headache relief in 70% of 85 mg/naproxen sodium 500 mg combination, com- sumatriptan-treated subjects and 22% of those pared with 23% for sumatriptan 100 mg, 16% for receiving placebo (P < .001).30 Sumatriptan injections naproxen sodium 500 mg, and 10% for placebo. The Headache 9 sumatriptan/naproxen sodium combination was supe- nasal spray and placebo groups, respectively, were rior to sumatriptan alone (P = .02), naproxen, and 18.3% and 11.4% at 15 minutes, 39.2% and 24.1% at placebo (P < .001). The headache relief percentages 30 minutes, and 56.9% and 34.2% at 1 hour (P < .001). in study 2 were 57%, 50%, 43%, and 29% (P < .001 The second study compared zolmitriptan nasal for sumatriptan/naproxen sodium, sumatriptan, and spray 5 mg with placebo for the treatment of 1 attack naproxen sodium vs placebo; P = .03 for sumatriptan/ and used a primary outcome of total symptom naproxen sodium vs sumatriptan). freedom (freedom from headache, nausea, photopho- In a Class I 2-attack, crossover study of bia, and phonophobia) at 1 hour.38 Zolmitriptan nasal sumatriptan 85 mg/naproxen sodium 500 mg against spray was superior to placebo for total symptom placebo in persons with poor response to short- freedom (14.5% vs 5.1%; P < .0001) at 1 hour. Sec- acting triptans (almotriptan, eletriptan, rizatriptan, ondary outcomes showed zolmitriptan to be superior sumatriptan, or zolmitriptan), sumatriptan/naproxen to placebo for headache relief as early as 10 minutes sodium treatment was superior to placebo for 2- to (15.1% vs 9.1%; P = .0079) and for headache freedom 24-hour sustained headache-free response (study 1: as early as 30 minutes (7.7% vs 3.2%; P = .0039). The 26% vs 8%; study 2: 31% vs 8%; P < .001 for both headache relief rate at 2 hours post-dose was 66.2% comparisons).34 Headache-free rates at 2 hours were for the zolmitriptan group, compared with 35.0% for also superior in the sumatriptan/naproxen sodium the placebo group (P < .001). Zolmitriptan nasal group (study 1: 40% vs 17%; study 2: 44% vs 14%; spray also produced significantly higher headache P < .001). relief rates than placebo at all earlier time points Zolmitriptan.—Oral zolmitriptan was rated as assessed, starting as early as 15 minutes post-dose effective for acute migraine treatment in the 2000 (P < .001). Similar results were obtained for the AAN Guidelines. analysis of the first attack. Significantly higher pain- In a Class I study of zolmitriptan 2.5 mg oral free rates were obtained with zolmitriptan nasal dissolvable tablets against placebo for 2 migraine spray, compared with placebo, from 15 minutes post- attacks, 2-hour headache-free rates were higher in dose onward (P < .005). those taking zolmitriptan (40% vs 20%; P < .001).35 DHE.—A Class I double-blind, placebo- Headache-free rates were also higher at 1 hour controlled, proof-of-concept efficacy trial analyzed (13% vs 8%, P = .004) and 1.5 hours (25% vs 15%; DHE administration with a breath-synchronized P < .001). inhaler for the treatment of acute migraine.39 Treat- Zolmitriptan 5 mg oral dissolvable tablets were ment was randomized to 0.5 mg, 1.0 mg, and placebo compared against placebo in a Class I study for the in a 2:2:1 ratio. Rates of headache relief at 2 hours treatment of 1 migraine attack.36 For the primary end- were higher in those subjects receiving 0.5 mg (72%) point, headache relief at 30 minutes, zolmitriptan was compared with placebo (33%; P = .019), but not in superior to placebo (16.5% vs 12.5%; P = .048). Sus- those subjects receiving 1.0 mg (65%; P = .071). Sub- tained headache freedom for 24 hours with jects receiving 0.5 mg (44%) and 1.0 mg (35%) were zolmitriptan 5 mg was also superior to placebo more likely to be headache-free at 2 hours compared (42.5% vs 16.4%; P < .0001). with placebo (7%; P = .015 and .050, respectively). Two Class I studies compared zolmitriptan nasal A larger Class I phase 3, double-blind, placebo- spray 5 mg against placebo. In the first study, subjects controlled, parallel-group, single-attack study com- with 1 or 2 attacks were treated with either pared inhaled DHE 1.0 mg with placebo for the zolmitriptan nasal spray or placebo.37 Headache relief treatment of acute migraine, with the primary end- at 2 hours was greater in the zolmitriptan nasal spray points of headache relief, and absence of photopho- group compared with placebo (66.2% vs 35.0%; bia, phonophobia, and nausea at 2 hours.40 Patients P < .001), with rates of headache relief from were treated with DHE at the time of moderate or zolmitriptan being superior to placebo as early as 15 severe headache. Subjects treating with DHE were minutes.Actual headache relief rates for zolmitriptan more likely to experience headache relief at 2 hours 10 January 2015

(59% vs 35%; P < .0001), and freedom from phono- Droperidol.—In a Class I study, droperidol pro- phobia (53% vs 34%; P < .0001), photophobia (47% vided superior rates of 2-hour headache relief com- vs 27%; P < .0001), and nausea or vomiting (67% vs pared with placebo.46 Subjects in this study were 59%; P = .0210). DHE was also significantly more randomized to receive injections of 0.1 mg, 2.75 mg, effective for headache freedom at 2 hours (28% vs 5.5 mg, and 8.25 mg or matching placebo for treat- 10%; P < .001) and for 2- to 24-hour sustained head- ment of moderate to severe migraine.Two-hour head- ache freedom (23% vs 7%; P < .001). ache relief rates were superior for subjects receiving Acetaminophen.—Based on the 2000 AAN Guide- 2.75 mg (87%), 5.5 mg (81%), and 8.25 mg (85%) lines, oral acetaminophen was considered probably compared with placebo (57%; P < .002). Headache effective for acute migraine based on available evi- relief from droperidol was superior to placebo as dence (Level B).41,42 Since that guideline, a Class I early as 1 hour for the 2.75 mg dose (P < .01), 90 study comparing oral acetaminophen 1000 mg with minutes for the 5.5 mg dose (P < .001), and 30 placebo for non-incapacitating migraine, ie, vomiting minutes for the 8.25 mg dose (P < .001). less than 20% attacks and no need for bed rest, found Phenazone.—A Class II study of oral phenazone 2-hour headache relief in 57.8% of those taking acet- 1000 mg vs placebo for 1 migraine attack treated aminophen vs 38.7% taking placebo (P = .002).43 within 4 hours of migraine onset determined that sub- Acetaminophen was superior in terms of 2-hour jects receiving phenazone were more likely to have headache relief in those with severe headache (50.9% 2-hour headache relief compared with placebo vs 27% placebo; P = .008). There was not a signifi- (48.6% vs 27.2% placebo; P = .002).47 Phenazone was cantly better response to acetaminophen in the sub- superior to placebo for those subjects with moderate group of subjects with moderate headache (62% vs intensity headache (53.6% vs 32.4% placebo; 48.1% placebo; P = .07). Two-hour headache-free P = .016) and for those subjects with severe headache rates were higher in subjects taking acetaminophen (38.9% vs 17.1% placebo; P = .042). Two-hour (22.4%) than in those treating with placebo (11.3%; headache-free rates were also higher in the phena- P = .01). zone group vs placebo (27.6% vs 13.6%; P = .016). A Class II study of intravenous acetaminophen Aspirin.—The 2000 AAN Guidelines established 1000 mg for the treatment of 1 acute migraine attack, oral aspirin as effective for the acute treatment of with 30 clinic patients in each group, failed to demon- migraine. In a Class I crossover study of 2 migraine strate significant differences between acetaminophen attacks, using aspirin mouth-dispersible formulation and placebo in terms of headache freedom at 2 hours 900 mg vs placebo, 2-hour headache relief was (10% vs placebo 13%; P = ns), headache relief at 2 achieved in 48% of the aspirin group compared with hours (30% vs 20% placebo; P = ns), and headache 19% of the placebo group (P < .0005).48 Aspirin freedom after 24 hours (31% vs 33% placebo; began to show superiority over placebo at 30 minutes P = ns).44 in terms of headache intensity and at 3 hours for Chlorpromazine.—A Class I study compared intra- headache freedom. venous chlorpromazine 0.1 mg/kilogram vs placebo in In a separate double-blind, placebo-controlled the acute treatment of migraine with or without aura Class I study of 1 migraine attack, aspirin 1000 mg in the emergency department.45 Compared with was compared with placebo.49 Aspirin was superior in placebo, chlorpromazine-treated subjects had higher terms of headache relief at 2 hours (52% vs 34%; rates of headache relief at 30 minutes (46% vs 7%; P < .001), and 20% of patients receiving aspirin were P < .05) and at 60 minutes (82% vs 15%; P < .05). headache-free from 1 to 6 hours after treatment, com- Rates of headache freedom at 1 hour were greater pared with 6% of those receiving placebo (P < .05). among chlorpromazine-treated subjects (65% vs 8%; Diclofenac.—In the 2000 AAN Guidelines, P < .05). Benefits were similar in subjects who had diclofenac was considered probably effective for the migraine with aura and in those who had migraine acute treatment of migraine.50,51 Since that guideline, without aura. a Class I study compared treatment with diclofenac Headache 11

100 mg, diclofenac 50 mg, sumatriptan 100 mg, and those taking ibuprofen 200 mg, in 40.8% of those placebo.52 Subjects were asked to treat 4 separate taking ibuprofen 400 mg, and in 28.1% of the placebo migraine attacks, and they received a different treat- group (P = .004 for 200 mg; P = .006 for 400 mg). In ment for each attack (1 of each). One hundred forty- those with severe migraine, 400 mg was superior to four patients received at least 1 treatment, and 115 placebo (36.9% vs 21.6% placebo; P = .048), but there patients (80%) completed the study. The primary was no significant difference for the 200 mg group. outcome was 2-hour headache intensity using a visual Ketorolac.—A Class II study evaluated an intrana- analog scale where 100 mm was maximal headache. sal formulation of ketorolac tromethamine, contain- Both diclofenac 50 mg and 100 mg were superior to ing 6% lidocaine for the acute treatment of migraine placebo (22 mm average with 100 mg, 26 mm with with and without aura, treated within the first 4 hours 50 mg, and 46 mm with placebo; P < .001). of a migraine attack.57 There was no significant differ- In another Class I trial, subjects treated 4 ence for headache freedom at 2 hours in the ketorolac migraine attacks, 2 with diclofenac 65 mg, and 2 with group compared with placebo (18% vs 10%; P = .17). placebo, in random order.53 Subjects with migraine Ketorolac was superior to placebo for several second- with aura were excluded. Subjects treating migraine ary outcome measures, including lack of disability at 2 with diclofenac were more likely to be headache-free hours (24% vs 10%; P = .009) and 2-hour headache at 2 hours (45.8% vs 25.1%; P < .0001). relief (51.5% vs 31.9%; P = .02). Another Class I study evaluated the use of 50-mg Tramadol/Acetaminophen.—A Class I study com- diclofenac potassium sachets and diclofenac 50 mg pared tramadol 75 mg/acetaminophen 650 mg (given tablets in comparison to placebo in a crossover trial.54 as 2 tablets) with 2 placebo tablets for the treatment Patients using sachets were more likely to be of a single migraine attack.58 Headache relief at 2 headache-free at 2 hours than those using placebo hours was superior for subjects taking tramadol/ (24.7% vs 11.7%; P < .0001) and those using 50 mg acetaminophen (55.8 vs 33.8% placebo; P < .001.) tablets (24.7% vs 18.5%; P = .0035). Subjects using Headache freedom at 2 hours was also superior in tablets were more often headache-free at 2 hours those taking tramadol/acetaminophen (22.1% vs than those using placebo (18.5% vs 11.7%; P = .0040). 9.3% placebo; P < .001). Tramadol/acetaminophen For 2-hour headache relief, both sachets (46.0% vs was superior to placebo for photophobia at 2 hours 24.1%; P < .0001) and tablets (41.6% vs 24.1%; (34.6% vs 52.2%; P = .003) and phonophobia (34.3% P < .0001) outperformed placebo. vs 44.9%; P = .008), but not for migraine-related Another Class I study compared 50-mg nausea (38.5% vs 29.4% placebo; P = .681). diclofenac potassium oral solution with placebo in a Tramadol.—Intravenous tramadol 100 mg was double-blind, randomized, placebo-controlled, single- compared with placebo for acute emergency treat- attack trial.55 Diclofenac was superior for 2-hour ment of a single attack in a single-blind Class II headache freedom (25% vs 10%; P < .001), freedom study.59 Tramadol was superior for the primary end- from nausea (65% vs 53%; P = .002), freedom from point of 50% headache relief at 1 hour (76% vs photophobia (41% vs 27%; P < .001), and phonopho- 35.6%; P = .04), but not for the secondary end-point bia (44% vs 27%; P < .001) compared with placebo. of headache freedom (29% vs 11% placebo; P = ns). Ibuprofen.—The 2000 AAN Guidelines estab- Octreotide.—In a Class I placebo-controlled cross- lished ibuprofen as effective for acute migraine treat- over study, patients treated acute migraine with at ment. Since that guideline, a Class II study compared least moderate intensity with either 100-μg subcuta- ibuprofen 200 mg and ibuprofen 400 mg with neous octreotide or placebo. Octreotide was not supe- placebo.56 Subjects were excluded if they experienced rior for headache relief at 2 hours (14% vs 20% either incapacitating migraines requiring bed rest placebo; P = ns) or for pain freedom at 2 hours (2% more than 50% of the time or vomiting more than vs 7% placebo; P = ns).60 20% of the time. For mild to moderate intensity head- Magnesium.—In a Class II placebo-controlled aches, 2-hour headache relief occurred in 41.7% of trial of intravenous magnesium sulfate (1 g for the 12 January 2015 treatment of 1 migraine attack), magnesium provided knowing the strength of the evidence supporting benefit superior to placebo at 60 minutes for the that superiority. According to this evidence assess- treatment of migraine with aura attacks in regard to ment, specific medications within the following headache relief (50% vs 13%; P < .05) and headache classes are deemed effective for migraine acute freedom (37% vs 7%; P < .05).61 In subjects without therapy: triptans, ergotamine derivatives, nonsteroi- aura, there was no significant difference between dal anti-inflammatory drugs (NSAIDs), opioids, and magnesium and placebo (headache relief: 33% vs combination medications. Several other medications 17%; headache freedom: 23% vs 10%; P = ns for both are “probably effective” or “possibly effective” as comparisons). listed above. Another Class II study of intravenous magne- This systematic assessment of the literature does sium compared treatment with 2-g magnesium sulfate not provide guidance regarding which medications with 10-mg metoclopramide and placebo in a 1:1:1 should be used for the acute therapy of migraine for a ratio for the treatment of 1 migraine attack in the specific patient. Although a clinician would want to emergency department.62 The primary end-point was prescribe an acute migraine medication that has headache intensity on a visual analog scale at 30 strong evidence in support of its efficacy, potential minutes. Average headache intensity at 30 minutes side-effects, potential adverse events, patient-specific was 3.9 mm in those receiving magnesium, 3.7 mm contraindications to certain medications, and drug– after metoclopramide, and 4.8 after placebo. There drug interactions all need to be considered when were no significant differences between treatment choosing a migraine acute medication. In clinical arms, but there was a benefit from magnesium com- practice, acute treatment can be associated with pared with placebo in the subgroup experiencing serious adverse events, such as tolerance and depen- migraine with aura compared with placebo (P = .04) dence with barbiturates or opioids, peptic ulcer or and metoclopramide (P = .03). renal disease with NSAIDs, or worsening migraine Intravenous Valproate.—In a Class IV open-label from medication-overuse headache. Thus, categoriza- study, patients with various primary headache disor- tion of a medication as having Level A evidence of ders received 1 treatment of intravenous valproate benefit does not necessarily mean that the medication for headache of moderate or severe intensity using should be considered a first-line drug for the acute doses ranging from 300 mg to 1200 mg. The majority treatment of migraine. For example, although of patients (63.1%) reported improvement, although butorphanol nasal spray has strong evidence for its those with episodic headache were more likely to do superiority over placebo, this medication is com- so.63 Another Class IV study reported the use of intra- monly avoided due to concerns about dependence, venous valproate 500 mg via slow intravenous bolus addiction, and development of medication-overuse injection for acute migraine. In this observational headache. study, 32 of 36 patients, including those with or This review also does not address acute migraine without ongoing valproate prophylaxis, reported treatment in children or the elderly. There is limited improvement after treatment (Table).64 evidence for the treatment of acute migraine in chil- dren, and both high placebo responses and the DISCUSSION shorter attack length in pediatric migraine influence This systematic assessment of the literature is a clinical trial design.2,65,66 Clinical trials usually exclude comprehensive evaluation of the evidence for efficacy patients aged 65 and over, meaning there is little evi- of individual migraine acute medications. The dence for the acute treatment of migraine in elderly strength of the evidence for each medication has been patients. graded. Thus, this American Headache Society evi- This systematic assessment of the literature dence assessment can be used as a guide for knowing cannot be used to compare the efficacy of individual which medications have been shown to be superior to migraine acute therapies. This assessment reports placebo for the acute therapy of migraine and for the strength of evidence supporting superiority of Headache 13

Table.—Strength of the Evidence

Level A Level B Level C Level U Others

Analgesic Antiemetics Antiepileptic NSAIDs Level B negative Acetaminophen 1000 mg *Chlorpromazine IV Valproate IV 400-1000 mg Celecoxib Other (for non-incapacitating 12.5 mg 400 mg Octreotide SC 100 attacks) Droperidol IV 2.75 mg μg *Metoclopramide IV 10 mg *Prochlorperazine IV/IM 10 mg; PR 25 mg Ergots Ergots Ergot Others Level C negative DHE DHE * IV, IM, SC 1 mg *Ergotamine 1-2 mg *Lidocaine IV Antiemetics *Nasal spray 2 mg *Ergotamine/caffeine *Hydrocortisone *Chlorpromazine Pulmonary inhaler 1 mg 1/100 mg IV 50 mg IM 1 mg/kg *Granisetron IV 40-80 μg/kg NSAIDs NSAIDs NSAIDs NSAIDs *Aspirin 500 mg *Flurbiprofen 100 mg Phenazone 1000 mg Ketorolac Diclofenac 50, 100 mg Ketoprofen 100 mg tromethamine Ibuprofen 200, 400 mg Ketorolac IV/IM 30-60 mg nasal spray *Naproxen 500, 550 mg Opioids Opioid Analgesic *Butorphanol nasal spray *Butorphanol IM 2 mg Acetaminophen 1mg *Codeine 30 mg PO IV 1000 mg *Meperidine IM 75 mg *Methadone IM 10 mg *Tramadol IV 100 mg Triptans Others Steroid

Almotriptan 12.5 mg MgSO4 IV (migraine with Dexamethasone IV 4-16 mg Eletriptan 20, 40, 80 mg aura) 1-2 g Frovatriptan 2.5 mg *Isometheptene 65 mg *Naratriptan 1, 2.5 mg *Rizatriptan 5, 10 mg Sumatriptan *Oral 25, 50, 100 mg *Nasal spray 10, 20 mg Patch 6.5 mg *SC 4, 6 mg Zolmitriptan nasal spray 2.5, 5 mg *Oral 2.5, 5 mg Combinations Combinations Others *Acetaminophen/aspirin/ *Codeine/acetaminophen *Butalbital 50 mg caffeine 500/500/130 mg 25/400 mg *Lidocaine intranasal Sumatriptan/naproxen Tramadol/acetaminophen 85/500 mg 75/650 mg Combinations *Butalbital/acetaminophen/ caffeine/codeine 50/325/ 40/30 mg *Butalbital/acetaminophen/ caffeine 50/325/40 mg

*Based on 2000 American Academy of Neurology evidence review. Level A: Medications are established as effective for acute migraine treatment based on available evidence. Level B: Medications are probably effective for acute migraine treatment based on available evidence. Level C: Medications are possibly effective for acute migraine treatment based on available evidence. Level U: Evidence is conflicting or inadequate to support or refute the efficacy of the following medications for acute migraine. Level B negative: Medication is probably ineffective for acute migraine. Level C negative: Medication is possibly ineffective for acute migraine. 14 January 2015 individual drugs relative to placebo, but it does not that process might not have been identified. This compare the relative efficacy of migraine acute medi- project was undertaken before the US Institute of cations to one another. Such comparisons are best Medicine published its recommendations for guide- made through well-conducted, head-to-head studies lines processes.Thus, these guidelines do not conform or carefully conducted network meta-analyses. Fur- entirely to those recommendations. For example, we thermore, there is substantial variability in the design did not formally incorporate assessments of side of the studies included in this assessment with regard effects and harms into our assessment process. to subject inclusion and exclusion criteria (eg, migraine phenotype, frequency, and severity; exclu- CONCLUSIONS sion of patients with specific medical conditions; According to this systematic review of the litera- allowance for use of migraine prophylactic therapy; ture and structured grading of the evidence strength, inclusion of patients who had failed prior migraine specific medications within the following classes are acute therapies), number of attacks treated, timing considered “effective” for the acute therapy of for administering the acute therapy (eg, treat when migraine: triptans, ergotamine derivatives, NSAIDs, mild vs treat when moderate to severe; early after opioids, and combination medications. Several other onset of migraine vs later after onset of migraine), medications are considered “probably effective” or and outcome measures. Primary outcome measures “possibly effective.” This evidence base for medica- varied, but the most common were 2-hour headache tion efficacy should be considered along with poten- relief, followed by 2-hour headache freedom. Other tial medication side effects, potential adverse events, primary outcomes included change in headache patient-specific contraindications to use of a particu- intensity before and after treatment based on visual lar medication, and drug-to-drug interactions when analog scales; time to headache freedom, 24-hour sus- deciding which medication to prescribe for acute tained relief; and 4-hour headache relief. Secondary therapy of a migraine attack. measures also differed, but usually included relief of Acknowledgments: Dr. Christina Szperka (Division nausea, photophobia, phonophobia, and disability, as of Neurology, Children’s Hospital of Philadelphia), Dr. well as measures of headache relief. Headache Eric Hastriter (Department of Neurology, Mayo Clinic freedom and sustained headache relief are harder to Scottsdale), Dr. Laura McGowan (Buffalo Medical achieve than 2-hour headache relief, so studies that Group), and Dr. Shatabdi Patel (Department of Neurol- use headache relief instead of headache freedom as a ogy, Thomas Jefferson University) all reviewed articles primary outcome are more likely to have a higher for this study. The authors wish to thank the American percentage of responders. Headache Society and Linda McGillicuddy for their The studies included in this assessment of the assistance in organizing this study. literature evaluated the efficacy of migraine acute medications in adults with episodic migraine with or REFERENCES without aura. However, they did not evaluate the effi- cacy of these medications for the treatment of status 1. Stewart WF, Lipton RB, Celentano DD, Reed ML. migraine, in patients with chronic migraine, or for the Prevalence of migraine headache in the United treatment of menstrual migraine. Furthermore, in States. Relation to age, income, race and other sociodemographic factors. JAMA. 1992;267:64-69. some studies, subjects with especially severe migraine 2. Headache Classification of the International Head- were excluded. Thus, the specific characteristics of an ache Society. The International Classification of individual patient and the migraine attack that is to Headache Disorders, 3rd edition (beta version). be treated must be considered before applying the Cephalalgia. 2013;33:627-808. findings of this evidence assessment. Some of the evi- 3. Holland S, Silberstein SD, Freitag F, Dodick DW, dence assessments in this analysis rely on work done Argoff C, Ashman E. Evidence-based guideline for the original 2000 AAN Guidelines. It is possible update: NSAIDs and other complementary treat- that errors or misclassifications that occurred during ments for episodic migraine prevention in adults: Headache 15

Report of the Quality Standards Subcommittee of controlled study conducted in the United States. the American Academy of Neurology and the Headache. 2003;43:202-213. American Headache Society. Neurology. 2012;78: 15. Stark R, Dahlof C, Haughie S, Hettiarachchi J. Effi- 1346-1353. cacy, safety and tolerability of oral eletriptan in the 4. Silberstein SD. Practice parameter – Evidence- acute treatment of migraine: Results of a phase III, based guidelines for migraine headache (an multicentre, placebo-controlled study across three evidence-based review): Report of the quality stan- attacks. Cephalalgia. 2002;22:23-32. dards Subcommittee of the American Academy of 16. Brandes JL, Kudrow D, Cady R, Tiseo PJ, Sun W, Neurology for the United States Headache Consor- Sikes CR. Eletriptan in the early treatment of acute tium. Neurology. 2000;55:754-762. migraine: Influence of pain intensity and time of 5. Tfelt-Hansen P, Pascual J, Ramadan N, et al. Guide- dosing. Cephalalgia. 2005;25:735-742. lines for controlled trials of drugs in migraine: Third 17. Farkkila M, Olesen J, Dahlof C, et al. Eletriptan for edition. A guide for investigators. Cephalalgia. the treatment of migraine in patients with previous 2012;32:6-38. poor response or tolerance to oral sumatriptan. 6. American Academy of Neurology. Clinical Practice Cephalalgia. 2003;23:463-471. Guideline Process Manual. St. Paul, MN: The Ameri- 18. Olesen J, Diener HC, Schoenen J, Hettiarachchi J. can Academy of Neurology; 2011. No effect of eletriptan administration during the 7. Institute of Medicine (US) Committee on Standards aura phase of migraine. Eur J Neurol. 2004;11:671- for Developing Trustworthy Clinical Practice Guide- 677. lines. Clinical Practice Guidelines We Can Trust. 19. Rapoport A, Ryan R, Goldstein J, Keywood C. Dose Washington, DC: National Academies Press; 2011. range-finding studies with frovatriptan in the acute 8. Dahlof C, Tfelt-Hansen P, Massiou H, Fazekas A. treatment of migraine. Headache. 2002;42(Suppl. 2): Dose finding, placebo-controlled study of oral S74-S83. almotriptan in the acute treatment of migraine. 20. Goldstein J, Keywood C. Frovatriptan for the acute Neurology. 2001;57:1811-1817. treatment of migraine: A dose-finding study. Head- 9. Pascual J, Falk RM, Piessens F, et al. Consistent effi- ache. 2002;42:41-48. cacy and tolerability of almotriptan in the acute 21. Cady R, Elkind A, Goldstein J, Keywood C. Ran- treatment of multiple migraine attacks: Results of a domized, placebo-controlled comparison of early large, randomized, double-blind, placebo-controlled use of frovatriptan in a migraine attack versus study. Cephalalgia. 2000;20:488-496. dosing after the headache has become moderate or 10. Goadsby PJ, Zanchin G, Geraud G, et al. Early severe. Curr Med Res Opin. 2004;20:1465-1472. vs non-early intervention in acute migraine- 22. Massiou H, Jamin C, Hinzelin G, Bidaut-Mazel C. “Act when Mild (AwM).” A double-blind, placebo Efficacy of oral naratriptan in the treatment of controlled trial of almotriptan. Cephalalgia. 2008; menstrually related migraine. Eur J Neurol. 2005; 28:383-391. 12:774-781. 11. Mathew NT, Finlayson G, Smith TR, et al. Early 23. Ahrens SP, Farmer MV, Williams DL, et al. Efficacy intervention with almotriptan: Results of the and safety of rizatriptan wafer for the acute treat- AEGIS trial (AXERT Early Migraine Intervention ment of migraine. Rizatriptan Wafer Protocol 049 Study). Headache. 2007;47:189-198. Study Group. Cephalalgia. 1999;19:525-530. 12. Diener HC, Gendolla A, Gebert I, Beneke M. 24. Cady R, Martin V, Mauskop A, et al. Efficacy of Almotriptan in migraine patients who respond rizatriptan 10 mg administered early in a migraine poorly to oral sumatriptan:A double-blind, random- attack. Headache. 2006;46:914-924. ized trial. Headache. 2005;45:874-882. 25. Mannix LK, Loder E, Nett R, et al. Rizatriptan for 13. Eletriptan Steering Committee. Efficacy and safety the acute treatment of ICHD-II proposed menstrual of eletriptan 20 mg, 40 mg and 80 mg in Japanese migraine: Two prospective, randomized, placebo- migraineurs. Cephalalgia. 2002;22:416-423. controlled, double-blind studies. Cephalalgia. 2007; 14. Sheftell F, Ryan R, Pitman V. Efficacy, safety, and 27:414-421. tolerability of oral eletriptan for treatment of acute 26. Seeburger JL, Cady RK, Winner P, et al. Rizatriptan migraine: A multicenter, double-blind, placebo- for treatment of acute migraine in patients taking 16 January 2015

topiramate for migraine prophylaxis. Headache. 37. Dodick DW, Brandes J, Elkind A, Mathew N, 2012;52:57-67. Rodichok L. Speed of onset, efficacy and tolerability 27. Seeburger JL, Taylor FR, Friedman D, et al. Efficacy of zolmitriptan nasal spray in the acute treatment of and tolerability of rizatriptan for the treatment of migraine: A randomised, double-blind, placebo- acute migraine in sumatriptan non-responders. controlled study. CNS Drugs. 2005;19:125-136. Cephalalgia. 2011;31:786-796. 38. Gawel M, Aschoff J, May A, Charlesworth BR. 28. Carpay J, Schoenen J, Ahmad F, Kinrade F, Boswell Zolmitriptan 5 mg nasal spray: Efficacy and onset of D. Efficacy and tolerability of sumatriptan tablets in action in the acute treatment of migraine – Results a fast-disintegrating, rapid-release formulation for from phase 1 of the REALIZE study. Headache. the acute treatment of migraine: Results of a multi- 2005;45:7-16. center, randomized, placebo-controlled study. Clin 39. Aurora SK, Rozen TD, Kori SH, Shrewsbury SB. A Ther. 2004;26:214-223. randomized, double blind, placebo-controlled study 29. Sheftell FD, Dahlof CG, Brandes JL, Agosti R, of MAP0004 in adult patients with migraine. Head- Jones MW, Barrett PS. Two replicate randomized, ache. 2009;49:826-837. double-blind, placebo-controlled trials of the time 40. Aurora SK, Silberstein SD, Kori SH, et al. to onset of pain relief in the acute treatment of MAP0004, orally inhaled DHE: A randomized, con- migraine with a fast-disintegrating/rapid-release trolled study in the acute treatment of migraine. formulation of sumatriptan tablets. Clin Ther. Headache. 2011;51:507-517. 2005;27:407-417. 41. Peatfield RC, Petty RG, Rose FC. Double blind 30. Wendt J, Cady R, Singer R, et al. A randomized, comparison of mefenamic acid and acetaminophen double-blind, placebo-controlled trial of the efficacy (paracetamol) in migraine. Cephalalgia. 1983;3:129- and tolerability of a 4-mg dose of subcutaneous 134. sumatriptan for the treatment of acute migraine 42. Karabetsos A, Karachalios G, Bourlinou P, Reppa attacks in adults. Clin Ther. 2006;28:517-526. A, Koutri R, Fotiadou A. Ketoprofen versus 31. Goldstein J, Smith TR, Pugach N, Griesser J, Sebree paracetamol in the treatment of acute migraine. T, Pierce M. A sumatriptan iontophoretic transder- Headache. 1997;37:12-14. mal system for the acute treatment of migraine. 43. Lipton RB, Baggish JS, Stewart WF, Codispoti M. Headache. 2012;52:1402-1410. Efficacy and safety of acetaminophen in the treat- 32. Djupesland PG, Docekal P. Intranasal sumatriptan ment of migraine: Results of a randomized, double- powder delivered by a novel breath-actuated blind, placebo-controlled, population-based study. bi-directional device for the acute treatment of Arch Intern Med. 2000;160:3486-3492. migraine: A randomised, placebo-controlled study. 44. Leinisch E, Evers S, Kaempfe N, et al. Evaluation of Cephalalgia. 2010;30:933-942. the efficacy of intravenous acetaminophen in the 33. Brandes JL, Kudrow D, Stark SR, et al. Sumatriptan- treatment of acute migraine attacks:A double-blind, naproxen for acute treatment of migraine: A ran- placebo-controlled parallel group multicenter study. domized trial. JAMA. 2007;297:1443-1454. Pain. 2005;117:396-400. 34. Mathew NT, Landy S, Stark S, et al. Fixed-dose 45. Bigal ME, Bordini CA, Speciali JG. Intravenous sumatriptan and naproxen in poor responders to chlorpromazine in the emergency department treat- triptans with a short half-life. Headache. 2009;49: ment of migraines: A randomized controlled trial. J 971-982. Emerg Med. 2002;23:141-148. 35. Loder E, Freitag FG, Adelman J, Pearlmand S, 46. Silberstein SD, Young WB, Mendizabal JE, bu-Shakra S. Pain-free rates with zolmitriptan Rothrock JF, Alam AS. Acute migraine treat- 2.5 mg ODT in the acute treatment of migraine: ment with droperidol: A randomized, double-blind, Results of a large double-blind placebo-controlled placebo-controlled trial.Neurology.2003;60:315-321. trial. Curr Med Res Opin. 2005;21:381-389. 47. Gobel H, Heinze A, Niederberger U, Witt T, 36. Spierings EL, Rapoport AM, Dodick DW, Zumbroich V. Efficacy of phenazone in the treat- Charlesworth B. Acute treatment of migraine with ment of acute migraine attacks: A double-blind, zolmitriptan 5 mg orally disintegrating tablet. CNS placebo-controlled, randomized study. Cephalalgia. Drugs. 2004;18:1133-1141. 2004;24:888-893. Headache 17

48. MacGregor EA, Dowson A, Davies PT. Mouth- treatment of migraine: Safety and efficacy data from dispersible aspirin in the treatment of migraine: A a phase II clinical trial. Cephalalgia. 2012;32:766-777. placebo-controlled study. Headache. 2002;42:249- 58. Silberstein SD, Freitag FG, Rozen TD, et al. 255. Tramadol/acetaminophen for the treatment of acute 49. Lipton RB, Goldstein J, Baggish JS, Yataco AR, migraine pain: Findings of a randomized, placebo- Sorrentino JV, Quiring JN. Aspirin is efficacious for controlled trial. Headache. 2005;45:1317-1327. the treatment of acute migraine. Headache. 2005;45: 59. Alemdar M, Pekdemir M, Selekler HM. Single-dose 283-292. intravenous tramadol for acute migraine pain in 50. Dahlof C, Bjorkman R. Diclofenac-K (50 and adults: A single-blind, prospective, randomized, 100 mg) and placebo in the acute treatment of placebo-controlled clinical trial. Clin Ther. 2007; migraine. Cephalalgia. 1993;13:117-123. 29:1441-1447. 51. Massiou H, Serrurier D, Lasserre O, Bousser MG. 60. Levy MJ, Matharu MS, Bhola R, Meeran K, Effectiveness of oral diclofenac in the acute treat- Goadsby PJ. Octreotide is not effective in the acute ment of common migraine attacks: A double-blind treatment of migraine. Cephalalgia. 2005;25:48-55. study versus placebo. Cephalalgia. 1991;11:59-63. 61. Bigal ME, Bordini CA, Tepper SJ, Speciali JG. Intra- 52. The Diclofenac-K-Sumatriptan Migraine Study venous magnesium sulphate in the acute treatment Group. Acute treatment of migraine attacks: Effi- of migraine without aura and migraine with aura. A cacy and safety of a nonsteroidal anti-inflammatory randomized, double-blind, placebo-controlled study. drug, diclofenac/potassium, in comparison to oral Cephalalgia. 2002;22:345-353. sumatriptan and placebo. Cephalalgia. 1999;19:232- 62. Cete Y, Dora B, Ertan C, Ozdemir C, Oktay C. A 240. randomized prospective placebo-controlled study 53. Vecsei L, Gallacchi G, Sagi I, et al. Diclofenac of intravenous magnesium sulphate vs meto- epolamine is effective in the treatment of acute clopramide in the management of acute migraine migraine attacks. A randomized, crossover, double attacks in the emergency department. Cephalalgia. blind, placebo-controlled, clinical study. Cephalal- 2005;25:199-204. gia. 2007;27:29-34. 63. Stillman MJ, Zajac D, Rybicki LA. Treatment of 54. Diener HC, Montagna P, Gacs G, et al. Efficacy and primary headache disorders with intravenous tolerability of diclofenac potassium sachets in valproate: Initial outpatient experience. Headache. migraine: A randomized, double-blind, cross-over 2004;44:65-69. study in comparison with diclofenac potassium 64. Waberzinek G, Markova J, Mastik J. Safety and effi- tablets and placebo. Cephalalgia. 2006;26:537-547. cacy of intravenous sodium valproate in the treat- 55. Lipton RB, Grosberg B, Singer RP, et al. Efficacy ment of acute migraine. Neuro Endocrinol Lett. and tolerability of a new powdered formulation of 2007;28:59-64. diclofenac potassium for oral solution for the acute 65. Lewis D, Ashwal S, Hershey A, Hirtz D, Yonker M, treatment of migraine: Results from the Interna- Silberstein S. Practice parameter: Pharmacological tional Migraine Pain Assessment Clinical Trial treatment of migraine headache in children and ado- (IMPACT). Cephalalgia. 2010;30:1336-1345. lescents: Report of the American Academy of Neu- 56. Codispoti JR, Prior MJ, Fu M, Harte CM, Nelson rology Quality Standards Subcommittee and the EB. Efficacy of nonprescription doses of ibuprofen Practice Committee of the Child Neurology Society. for treating migraine headache. A randomized con- Neurology. 2004;63:2215-2224. trolled trial. Headache. 2001;41:665-679. 66. Rothner AD, Wasiewski W, Winner P, Lewis D, 57. Pfaffenrath V, Fenzl E, Bregman D, Farkkila M. Stankowski J. Zolmitriptan oral tablet in migraine Intranasal ketorolac tromethamine (SPRIX(R)) treatment: High placebo responses in adolescents. containing 6% of lidocaine (ROX-828) for acute Headache. 2006;46:101-109. 18 January 2015

APPENDIX 1: ADVANCED SEARCH STRATEGY FOR RELEVANT ARTICLES

Searches Results

Migraine and related terms component 1 Migraine Disorders 2 migraine*.mp. 3 (unilateral adj2 headach*).mp. 4 cephalalgi*.mp. 5 hemicrania?.mp. 6 or/1-5

Acute component 7 acute*.mp. 8 immediate.mp. 9 or/7-8

Drug therapy and related terms component 10 exp Drug Therapy/ 11 (drug adj2 therap*).mp. 12 pharmacotherap*.mp. 13 ae.fs. 14 ad.fs. 15 dt.fs. 16 pd.fs. 17 th.fs. 18 tu.fs. 19 or/10-18

Combined results, limited to human, English, 2004-current, and excluded comments, editorial, and letter 20 6 and 9 and 19 21 limit 20 to (comment or editorial or letter) 22 20 not 21 23 exp animals/ not (exp animals/ and exp humans/) 24 22 not 23 25 limit 24 to (English language and yr = “2004 -Current”) 26 remove duplicates from 25 Headache 19

APPENDIX 2: DATA EXTRACTION FORM 20 January 2015

APPENDIX 3: STUDY CLASSIFICATION patients on the standard treatment are com- FOR THERAPEUTIC INTERVENTIONS parable to those of previous studies estab- FROM THE AMERICAN ACADEMY OF lishing efficacy of the standard treatment. NEUROLOGY 4. The interpretation of the study results is based on a per-protocol analysis that Class I accounts for dropouts or crossovers. – Randomized, controlled clinical trial in a represen- tative population Class II – Masked or objective outcome assessment – Cohort study meeting criteria a–e (see Class I) or a – Relevant baseline characteristics are presented and randomized, controlled clinical trial that lacks 1 or substantially equivalent between treatment groups, 2 criteria b–e (see Class I) or there is appropriate statistical adjustment for – All relevant baseline characteristics are presented differences and substantially equivalent among treatment – Also required: groups or there is appropriate statistical adjust- a. Concealed allocation ment for differences b. Primary outcome(s) clearly defined – Masked or objective outcome assessment c. Exclusion/inclusion criteria clearly defined d. Adequate accounting for dropouts (with at least Class III 80% of enrolled subjects completing the study) – Controlled studies (including well-defined natural and crossovers with numbers sufficiently low to history controls or patients serving as their own have minimal potential for bias controls) e. For non-inferiority or equivalence trials claim- – A description of major confounding differences ing to prove efficacy for 1 or both drugs, the between treatment groups that could affect following are also required*: outcome** 1. The authors explicitly state the clinically – Outcome assessment masked, objective, or per- meaningful difference to be excluded by formed by someone who is not a member of the defining the threshold for equivalence or treatment team non-inferiority. 2. The standard treatment used in the study is substantially similar to that used in previous Class IV studies establishing efficacy of the standard – Did not include patients with the disease treatment (eg, for a drug, the mode of – Did not include patients receiving different inter- administration, dose, and dosage adjust- ventions ments are similar to those previously shown – Undefined or unaccepted interventions or outcome to be effective). measures 3. The inclusion and exclusion criteria for – No measures of effectiveness or statistical precision patient selection and the outcomes of presented or calculable