Clinical Therapeutics/Volume 31, Number 3, 2009

Transcription

1 Clinical Therapeutics/Volume 31, Number 3, 2009 Topiramate Versus Amitriptyline in Migraine Prevention: A 26-Week, Multicenter, Randomized, Double-Blind, Double-Dummy, Parallel-Group Noninferiority Trial in Adult Migraineurs David W. Dodick, MD1; Fred Freitag, D02;James Banks, MD3;Joel Saper, MD4; Jim Xiang, PhDs; Marcia Rupnow, PhDs; David Biondi, DOs; Steven J. Greenberg, MD6; and Joseph Hulihan, MDs; for the CAPSS-277 Investigator Group 1Mayo Clinic Hospital) Phoenix) Arizona; 2Diamond Headache Clinic) Chicago) l!iinois; 3Mercy Health Research) Ryan Headache Center, St. Louis) Missouri; 4Michigan Head Pain & Neurological Institute) Ann Arbor, Michigan; SOrtho-McNeiljanssen Scientific Affairs) LLC) Titusville) NewJersey; and 6EMD Serono) Rockland) Massachusetts ABSTRACT Objective: The primary objective of this study was to compare the efficacy and tolerability of topiramate and amitriptyline in the prophylaxis of episodic migraine headache. Methods: This was a 26-week, multicenter, randomized, double-blind, double-dummy, parallel-group noninferiority study. Adults with 3 to 12 migraines per month were randomized in a 1: 1 ratio to receive an initial dose of 25 mg/d of either topiramate or amitriptyline, subsequently titrated to a maximum of 100 mg/d (or the maximum tolerated dose). The primary efficacy outcome was the change from prospective baseline in the mean monthly number of migraine episodes. Secondary efficacy variables included changes from the prospective baseline phase to the end of the double-blind phase in the mean monthly (28-day) rate of days with migraine, mean monthly rate of days with headache (migraine and nonmigraine), mean monthly rate of acute abortive medication use, mean monthly migraine duration, and mean monthly migraine severity. Additional secondary efficacy variables included changes in the mean monthly severity of migraine-associated symptoms (photophobia, phonophobia, and nausea), change in the mean monthly frequency of migraine-associated vomiting, and response rates (based on monthly migraine days and total headache days). The Migraine-Specific Quality of Life Questionnaire (MSQ) and the Weight Satisfaction Scale Questionnaire, which measures subjective satisfaction with current weight, were administered. Treatment-emergent adverse events (TEAEs) were monitored through the end of double-blind treatment. Results: The intent-to-treat population included 331 subjects (172 topiramate, 159 amitriptyline; 84.9% female; 84.6% white; mean [SD] age, 38.8 [11.0] years; mean weight, 77.1 [20.1] kg) who provided at least 1 efficacy assessment. The least squares mean (LSM) change from baseline in the mean monthly number of migraine episodes was not significantly different between the topiramate and amitriptyline groups (-2.6 and -2.7, respectively; 95% CI, -0.6 to 0.7). There were no significant differences between treatment groups in any of the prespecified secondary outcome measures. Subjects receiving topiramate had a significantly greater improvement in mean functional disability scores during migraine attacks compared with amitriptyline (LSM change: vs -0.19; 95% CI, -0.3 to 0.0; P = 0.040) and in the role function-restrictive, role function-preventive, and emotional function domains of the MSQ (P = 0.012, P = 0.014, and P = 0.029, respectively). Subjects receiving topiramate had a mean weight loss of 2.4 kg, compared with a mean weight gain of 2.4 kg in subjects receiving amitriptyline. Subjects in the topiramate group reported an overall improvement from baseline in weight satisfaction, whereas the amitriptyline group reported an overall deterioration in weight satisfaction (P < 0.001, topiramate vs amitriptyline). TEAEs of mild or moderate severity were reported in 118 subjects (66.7%) in the topira- Accepted for publication December 17, doi: /j.c1 inthera /$ - see front matter 2009 Excerpta Medica Inc. All rights reserved. 542 Volume 31 Number 3

2 D.W. Dodick et al. mate group and 112 subjects (66.3%) in the amitriptyline group. Among the most common TEAEs (reported in ::::5% of subjects during the double-blind phase) in the topiramate group were paresthesia (29.9%), fatigue (16.9%), somnolence (11.9%), hypoesthesia (10.7%), and nausea (10.2%). The most commonly reported TEAEs in the amitriptyline group were dry mouth (35.5%), fatigue (24.3%), somnolence (17.8%), weight increase (13.6%), dizziness (10.7%), and sinusitis (10.7%). Conclusions: In this noninferiority study, topiramate was at least as effective as amitriptyline in terms of reducing the rate of mean monthly migraine episodes and all prespecified secondary efficacy end points. Topiramate was associated with improvement in some quality-of-life indicators compared with amitriptyline and was associated with weight loss and improved weight satisfaction. (Clin Ther. 2009;31: ) 2009 Excerpta Medica Inc. Key words: headache, migraine prophylaxis, topiramate, amitriptyline, quality of life, weight loss. INTRODUCTION Migraine is a primary headache disorder that affects 11.7% of the general population (17.1 % of women and 5.6% of men)l and is one of the leading causes of disability worldwide. 2 Migraine accounts for an annual expenditure of approximately $13 to $17 billion in the United States, including direct costs associated with migraine treatment and indirect costs related to disability, lost productivity, and work absenteeism. 3-5 In randomized, double-blind, placebo-controlled clinical trials in adults with episodic migraine, treatment with topiramate 100 or 200 mg/d was associated with significant reductions in mean monthly migraine frequency compared with placebo (P :s; 0.011).6-8 In randomized, double-blind, placebo-controlled clinical trials in migraineurs who had ::::15 headache days per month, treatment with topiramate 100 mg/d was associated with significant mean reductions in migraine days compared with placebo (P :s; 0.02).9,10 In 2004, the US Food and Drug Administration (FDA) approved topiramate for migraine prophylaxis in adults. Topiramate has been associated with adverse events (AEs) such as paresthesia, fatigue, diarrhea, infrequent word-finding difficulties, and weight 10ss.6,7 Amitriptyline is also used for migraine prevention in the United States,ll although it is not approved by the FDA for this indication. The efficacy of amitriptyline in the prophylaxis of migraine has been reported in active- and placebo-controlled clinical trials,12,13 and the American Academy of Neurology guidelines list it among the most effective migraine-preventive medications. 12 Amitriptyline has been associated with AEs such as dry mouth, constipation, orthostatic hypotension, somnolence, and weight gain The objective of this study was to compare the efficacy and tolerability of topiramate and amitriptyline, both given at a maximum daily dose of 100 mg, for the prevention of migraine. The topiramate dose was chosen based on the results of earlier studies suggesting that 100 mg/d had the best efficacy and tolerability profile for migraine prophylaxis. 6,7 The amitriptyline dose was selected based on the usual maintenance dose of 50 to 100 mg/d used in clinical practice and in clinical trials. ll - 13 The primary efficacy outcome was the change from prospective baseline in the mean monthly number of migraine episodes. Various prespecified secondary outcomes were evaluated, including responder rates; other prespecified objectives included a comparison of weight change between treatment groups. SUBJECTS AND METHODS Inclusion and Exclusion Criteria Eligible subjects were adults (age ::::18 years) with a history of migraine without or with aura (International Headache Society class 1.1 and 1.2, respectively17) for at least 6 months before the screening/washout period, along with 3 to 12 migraines per month in the 3 months before the screening/washout period, from 3 to 12 migraine episodes during the 28-day prospective baseline period, and no more than 15 headache days (migraine and nonmigraine) during the prospective baseline period, based on headache records. Subjects were excluded from study participation if they had previously failed >2 adequate trials (a trial of at least 3 months' duration at doses recommended for headache relief) of migraine-preventive medications or had failed an adequate trial of topiramate or amitriptyline because of lack of efficacy or AEs. Other exclusion criteria were acute abortive medication use on >15 treatment days per month; onset of migraine after the age of 50 years; migraine aura only (without headache); history of cluster headache, a progressive neurologic disorder other than migraine, or a condition more painful than headache; history of a medical condition in which use of amitriptyline is contraindi- March

3 Clinical Therapeutics cated; history of an unstable medical condition within the past 2 years or of a major psychiatric disorder within the past 6 months that could impair reliable participation in the study or necessitate the use of medications not permitted in the study; history of drug or alcohol abuse within the past 2 years; or history of nephrolithiasis, active liver disease, or liver function tests ::::2 times the upper limit of normal. Pregnant or nursing women and those who were not practicing a medically accepted method of birth control were also excluded. Study Design This was a noninferiority clinical trial. A noninferiority design is used to compare a new medication/new use of a medication with an active medication that is an established treatment for a particular disorder. 18 Noninferiority trials use statistical methods to investigate whether the treatment under evaluation is at least as effective (by a predefined amount or margin) as the active comparator in achieving a clinically meaningful outcome. 18,19 This design also may be used to compare treatments having similar efficacy to examine whether the new treatment has potential advantages (ie, safety/ tolerability) relative to the standard treatment. 19 At the time this trial was designed, neither topiramate nor amitriptyline was approved by the FDA for migraine prophylaxis, although topiramate was subsequently approved for this use in adults. Given the lack of established treatments for migraine prophylaxis at the time the study was designed, it was felt that the published evidence for amitriptyline's effectiveness relative to placebo in the prevention of migraine ll - 13 was sufficient to support use of a noninferiority design to test the hypothesis that topiramate would be at least as effective as amitriptyline for migraine prevention. However, the study was designed to differentiate the treatments in terms of weight change and the potential impact of weight change on quality-of-life (QoL) indicators and patient-reported outcomes. Key elements of a well-designed noninferiority trial were met, including historical efficacy of at least one of the tested medications compared with placebo, an adequate sample size, and appropriate statistical analyses. 18,20,21 This randomized, double-blind, double-dummy, parallel-group study was conducted at 32 sites in the United States from February 2004 to October It was performed in accordance with the principles of the Declaration of Helsinki, and written informed consent was obtained from all participants before any study-related procedures were performed. An independent ethics committee or institutional review board approved the study protocol at each site. The study consisted of 3 phases: a pretreatment phase, a doubleblind phase, and a taper/exit phase (Figure 1). Pretreatment Phase The pretreatment phase lasted up to 56 days and consisted of 2 periods: a screening/washout period and a prospective baseline period. The screening/washout period took place within 28 days of the start of the prospective baseline period. At the screening visit (visit 1), subjects were evaluated for eligibility and were instructed to discontinue all migraine-preventive medications 14 to 28 days before the start of the prospective baseline period and throughout the study. The 28-day prospective baseline period began on study day -28 (visit 2), once subjects had completed the screening/washout period. Subjects were required to maintain a headache diary during this period, and those who met the prespecified entry criteria were considered eligible to enter the double-blind phase. Double-Blind Phase The randomization visit (visit 3) took place at least 26 days after the start of the prospective baseline period. Subjects who completed the prospective baseline period and met the entrance criteria were randomized in a 1: 1 ratio to receive either topiramate or amitriptyline. Five-digit subject code numbers and 4-digit medication code numbers were computer generated, preprinted on labels, and assigned as subjects qualified for participation. Randomization was performed using permuted blocks of 4 by site, and the assigned medication code numbers were retained for the duration of the study. The study drugs were provided in capsules of identical appearance and were shipped to each site by blocks. Blinding was maintained by administering the same number of capsules in the morning and evening. The double-blind phase consisted of a 4-week titration period and a 22-week maintenance period. During the titration period, visits occurred on day 1 (visit 3) and day 28 (visit 4). The titration schedule was the same in the 2 treatment groups: the starting dosage was 25 mg/d for 7 days (Table I). Weekly dose titrations of 25 mg/d were made at the investigator's dis- 544 Volume 31 Number 3

4 D.W. Dodick et al. Pretreatment Phase Double-Blind Phase Taper/Exit Phase (up to 2 weeks) Maintenance Period (22 weeks) Screening! Washout (up to 28 days) Prospective Baseline (28 days) Titration, 25 mg/wk (4 weeks) Visit 1 (up to day -56) Visit 2 (day -28) Randomization, Visit 3 (day 1) Visit 4 (day 28) Visit 5 (day 56) Visit 6 (day 84) Visit 7 (day 133) Visit 8 (day 182) Visit 9 (day 196) Figure 1. Study design. cretion up to 50 mg BID (or the maximum tolerated dose). A stable dose of at least 50 mg/d was required for continuation into the maintenance period. During the maintenance period, visits occurred on day 56 (visit 5), day 84 (visit 6), and day 133 (visit 7), with the final visit on day 182 (visit 8). In the event that a subject discontinued study medication before completing the full course of treatment according to protocol, final visit procedures were performed at the time of discontinuation. Taper/Exit Phase All subjects exiting the study underwent tapering of study medication. The length of the taper was determined at the investigator's discretion and could vary according to the dose of study medication. The exit visit took place on day 196 (visit 9), 2 weeks after the completion of the taper. Concomitant Headache Medications All medication use was documented during the study. Any change in concomitant medications or new medication use was reviewed from visit 3 through the end of the study. Use of acute headache medications, including over-the-counter analgesics, NSAIDs, triptans, ergot derivatives, and dihydroergotamine mesy- late, was permitted for symptomatic relief of headaches throughout the study, but was not to exceed 4 days per week. Subjects were instructed to adhere as closely as possible to the type and timing of acute therapy used before enrollment. Use of preventive migraine treatments other than the study medication was not allowed for the duration of the trial. Subjects were to continue their usual medical therapies, with no unnecessary changes in diet, pattern of physical activity, or overall lifestyle. Efficacy Measures Headache records were provided to subjects at visit 3. At each subsequent visit (visits 4-8) during the double-blind phase, headache records were collected and reviewed for headache occurrence, severity, and duration, as well as for associated migraine symptoms (photophobia, phonophobia, nausea, and vomiting). These data were used for analysis of the primary efficacy variable and the prespecified secondary efficacy variables. New headache records were distributed at each visit. The primary efficacy measure was the change from prospective baseline in the mean monthly (28-day) rate of migraine episodes. A migraine episode was defined as the period from the onset to the cessation of painful March

5 Clinical Therapeutics Table I. Topiramate and amitriptyline dosing schedules. Study Week (Days) Morning Evening Total Daily Dose Topiramate Week 1 (1-7) 25 mg (1 capsule) 25 mg Week 2 (8-14) 50 mg (two 25-mg capsules) 50 mg Week 3 (15-21) 25 mg (1 capsule) 50 mg (two 25-mg capsules) 75 mg Placebo (1 capsule) Week 4 (22-28) 50 mg (two 25-mg capsules) 50 mg (two 25-mg capsules) 100 mg Placebo (2 capsules) Amitriptyline Week 1 (1-7) 25 mg (1 capsule) 25 mg Week 2 (8-14) 50 mg (two 25-mg capsules) 50 mg Week 3 (15-21) Placebo (1 capsule) 75 mg (three 25-mg capsules) 75 mg Week 4 (22-28) Placebo (2 capsules) 100 mg (four 25-mg capsules) 100 mg migraine symptoms, not to exceed 24 hours. If painful symptoms lasted >24 hours, this was considered a new and distinct migraine episode. If painful symptoms recurred within 24 hours of initial onset, this was considered part of the initial migraine episode. Prespecified secondary efficacy variables included changes from the prospective baseline phase to the end of the double-blind phase in the following variables: mean monthly (28-day) rate of days with migraine headache; mean monthly rate of days with headache (migraine and nonmigraine); mean monthly rate of acute abortive medication use; mean monthly migraine duration; and mean monthly migraine severity. Additional secondary efficacy variables were changes in the mean monthly severity of migraine-associated symptoms (photophobia, phonophobia, and nausea) and in the mean monthly frequency of migraineassociated vomiting, and response rates (using categorical response rates of ::::25%, ::::50%, ::::75%, or 100% reduction from baseline, based on monthly migraine days and total headache days). Measures of Disability, Quality of Life, and Weight Satisfaction Subjects rated their worst level of functional disability during migraine episodes using a single-question survey completed 24 hours after the onset of each episode. The response options were as follows: 1 = normal activity allowed; 2 = disturbance of normal activity but no interruption or bed rest necessary; 3 = dis- continuation of normal activity with bed rest required; and 4 = emergency department visit or hospitalization. The mean change in the severity of functional disability was assessed in each treatment group. The Migraine Disability Assessment Score (MIDAS)22 was also used to evaluate headache-related disability at visits 1 and 8. Subjects answered 5 questions regarding time lost from school or paid work, household work and family, and sociallieisure activities over the past 3 months because of headache. The levels of disability were categorized as little or none (score 0-5), mild (6-10), moderate (11-20), and severe (::::21). Two additional questions were used to assess the frequency of headaches and the intensity of head pain, but the responses to these questions did not contribute to the MIDAS total score. The Migraine-Specific Quality of Life Questionnaire (MSQ) version 2.1 was administered at visits 1 through 8. It comprises 14 items measuring 3 domains: role function-restrictive (RR), role function-preventive (RP), and emotional function (EF).23 RR measures the degree to which performance is limited by migraines; RP measures the degree to which performance of daily activities is interrupted by migraines; and EF measures feelings of frustration and helplessness caused by migraines. Subjects evaluated how migraines affected their performance of daily activities and emotional status over the previous 4 weeks. Scores for each domain were normalized to a range from 0 to 100, with higher scores indicating better health-related QoL. 546 Volume 31 Number 3

6 D.W. Dodick et al. The Quality of Life Enjoyment and Satisfaction Questionnaire-Short Form (Q-LES-Q-SF) was administered at visit 1 and visits 3 through 8. It is a generic patient-reported measure of satisfaction with various aspects of life (eg, physical health, mood, work, household activities, social/family relationships, sexual drivel interest/performance, economic status, and living! housing situation) in the past week. 24,25 The overall Q-LES-Q-SF score was the sum of scores on items 1 to 14, normalized to a range from 1 to 100. Items 15 and 16 of the Q-LES-Q-SF were analyzed separately. The Weight Satisfaction Scale Questionnaire (WSSQ) was administered at visits 1 and 8. It is a self-reported measure containing a single question on subjective satisfaction with current weight. The response was scored on a scale from -3 (extremely dissatisfied) to +3 (extremely satisfied), with a higher score indicating greater satisfaction. Mean changes in weight and body mass index 26 (BMI) were evaluated from baseline to the last available measurement during the double-blind phase. Weight increase was evaluated from visit 2 through visit 9, and the proportions of subjects with a change in BMI category (<18.5, 18.5-<25,25-<30, and ::::30 kg!m 2 ) from baseline to the last available measurement during the double-blind phase were also evaluated. Safety Measures Brief physical examinations were performed at visits 1 and 8, and brief neurologic examinations were performed at visit 1. Vital signs were measured at visits 1 through 9; body weight and height were recorded at visit 1; and an electrocardiogram was performed at visit 1 and visits 4 through 8. Clinical laboratory tests were performed (visit 1 and visits 4-8) by a licensed, certified central laboratory and included liver function (aspartate aminotransferase, alanine aminotransferase, total bilirubin, alkaline phosphatase), serum chemistry (creatinine, blood urea nitrogen, sodium, potassium, chloride, bicarbonate, high-sensitivity C reactive protein [hs-crp], glycosylated hemoglobin [HbA 1c ]' interleukin-6 [IL-6], total cholesterol, triglycerides, low-density lipoprotein cholesterol), and a urine pregnancy test for women of childbearing potential. Written instructions were provided for the collection, preparation, and shipment of blood, plasma, and urine samples from the investigative site to the central laboratory. The clinical laboratory data were included with the source documents and sub- jects' case-report forms, and were transmitted directly to the sponsor's database. Analytes were compared against laboratory reference ranges at baseline and at each scheduled time point. The laboratory provided a list of subjects whose laboratory results were markedly abnormal, and the investigators followed subjects with any clinically significant laboratory abnormality that persisted at the end of the study until resolution of the abnormality or until the subject was clinically stable. AE data included both spontaneously reported AEs and those elicited by indirect questioning from visit 2 through the end of the study. The investigator recorded the date of onset, severity, action taken, and outcome for each AE and evaluated the potential relationship of each AE to treatment. An AE was defined as any untoward event or medical event that occurred during the study. Treatment-emergent AEs (TEAEs) were defined as those that were new in onset or aggravated in severity or frequency, or as abnormal results on diagnostic procedures occurring between the start and the end of double-blind treatment. The severity of TEAEs was rated as mild, moderate, or marked. Limiting AEs were defined as those that resulted in premature discontinuation of study medication. Serious AEs (SAEs) were defined as any untoward medical event that was fatal or immediately life threatening, that was permanently or significantly disabling, that resulted in or prolonged an existing hospitalization, or that was a congenital anomaly/birth defect. Statistical Analysis Efficacy analyses were performed on the intent-totreat (ITT) population, which consisted of all randomized subjects who received at least 1 dose of study medication and provided at least 1 postrandomization efficacy evaluation. Efficacy evaluations were based on information in subjects' headache records, as were disability assessments (MIDAS and functional disability), QoL assessments (MSQ and Q-LES-Q-SF), and the WSSQ. Safety analyses were performed on all randomized subjects who received at least 1 dose of study medication and for whom at least 1 posttreatment safety measure was available. The primary efficacy outcome was determined by using a testing procedure that was applied under the context of switching from noninferiority to superiority and controlled the overall type I error rate at 0.05,27 The null hypothesis was that the change in the mean monthly rate of migraine episodes (topiramate - ami- March

7 Clinical Therapeutics triptyline) was less than The alternative hypothesis of therapeutic noninferiority in efficacy was that the difference (topiramate - amitriptyline) in the primary efficacy variable was -1 or greater. If the lower bound of the 2-sided 95% CI was greater than -1, then noninferiority was assumed. Next, if the interval not only lay entirely above -1 but was also above 0, then there was evidence of the superiority of topiramate to amitriptyline for the primary efficacy variable at a significance level of 5%. After this, a testing procedure was applied in which a 2-sided 95% CI about the difference (topiramate - amitriptyline) in least squares mean (LSM) for the change in the mean monthly rate of migraine episodes was calculated to evaluate noninferiority of efficacy. The LSM was obtained from an analysis of covariance (ANCOVA) model using treatment and center as qualitative independent factors and the baseline mean monthly rate of migraine episodes as a covariate. The actual P value for testing superiority was obtained from the ANCOVA model. Assuming a common SD of 3, the required sample size of 165 subjects in each treatment group would have 85% power. Secondary efficacy variables, in addition to scores on the MSQ, MIDAS, Q-LES-Q-SF, and WSSQ, were analyzed separately by ANCOVA, with treatment and center as qualitative independent factors and baseline value as a covariate. The responder rate was analyzed using the Cochran-Mantel-Haenszel 29 (CMH) test. No adjustment was made for multiple comparisons. Two data summaries were performed, one based on the lastobservation-carried-forward approach and the other based on the observed data. RESULTS Demographic and Baseline Headache Characteristics Demographic characteristics (Table II) and baseline headache characteristics (Table III) were comparable in the 2 treatment groups. The mean (SD) age at migraine onset was 20.3 (10.3) years. The mean monthly rate of migraine episodes at baseline was 6.2 (2.7), with a mean migraine attack duration of 18.3 (18.4) hours. Patients had a mean of 8.5 (3.5) total headache days per month. Patient Disposition Of the 614 subjects screened, 347 were enrolled and randomized to treatment with topiramate (n = 178) or amitriptyline (n = 169) (Figure 2). The reasons for the exclusion of 267 subjects were as follows: 135 did not meet the inclusion criteria, 41 declined to participate, 23 were lost to follow-up, 16 had no reason recorded, 7 were recorded as having AEs (AE reports were collected before the start of study drug), and 45 were not enrolled for other reasons. The ITT population consisted of331 subjects (172 topiramate, 159 amitriptyline; mean [SD] age, 38.8 [11.0] years; age range, years). The safety population consisted of 346 subjects (177 topiramate, 169 amitriptyline). One hundred two of 178 topiramate-treated subjects (57.3%) and 95 of 169 amitriptyline-treated subjects (56.2%) completed double-blind treatment (Figure 2). Seventy-six topiramate-treated subjects (42.7%) and 74 amitriptyline-treated subjects (43.8 %) discontinued study medication before completion of the double-blind treatment phase according to protocol. Forty-seven of the 76 discontinuations in the topiramate group (61.8%) and 40 of the 74 discontinuations in the amitriptyline group (54.1%) occurred during the first 56 days of treatment. The most common reason for discontinuing treatment was limiting AEs (35/178 [19.7%] in the topiramate group, 38/169 [22.5%] in the amitriptyline group). Treatment Exposure The mean (SD) daily dose of study medication during the double-blind phase was 74.8 (23.9) mg in the topiramate treatment group and 76.5 (21.1) mg in the amitriptyline treatment group. The mean daily doses of study medication during the maintenance period were 90.6 (16.0) and 88.9 (18.7) mg, respectively. The majority of subjects (111/172 [64.5%] in the topiramate group, 104/159 [65.4%] in the amitriptyline group) received study medication for at least 134 days. The mean duration of treatment in the respective treatment groups was (72.3) and (67.0) days. Primary Efficacy Measure During the prospective baseline phase, subjects in the topiramate and amitriptyline treatment groups had a mean (SD) monthly rate of 6.3 (2.5) and 6.0 (2.3) migraine episodes, respectively. During doubleblind treatment, the LSM change from baseline in the monthly rate of migraine episodes was -2.6 in the topiramate treatment group and -2.7 in the amitriptyline treatment group (95% CI, -0.6 to 0.7; P = NS) 548 Volume 31 Number 3

8 D.W. Dodick et al. Table II. Demographic and baseline characteristics (intent-to-treat population). Topiramate Amitriptyline Total Variable (n = 172) (n = 159) (N=331) Sex, no. (%) Female 149 (86.6) 132 (83.0) 281 (84.9) Male 23 (13.4) 27 (17.0) 50 (15.1) Age, y Mean (SO) 39.7 (10.7) 37.9(11.3) 38.8 (11.0) Median (range) 40.0 (18-63) 37.0 (18-70) 40.0 (18-70) Race, no. (%) White 143 (83.1) 137 (86.2) 280 (84.6) Black 23 (13.4) 15 (9.4) 38(11.5) Asian 2 (1.2) 2 (1.3) 4 (1.2) Other 4 (2.3) 5 (3.1) 9 (2.7) Weight, kg Mean (SO) 76.8 (19.4) 77.3 (20.7) 77.1 (20.1) Range BMI, kg/m h Mean (SO) 27.8 (6.7) 28.5 (7.6) 28.2 (7.2) Range BMI bodymass index. *No data were available for 1 subject in the topiramate group. Table III. Baseline headache characteristics (intentto-treat population). Values are mean (SO). Topiramate Amitriptyline Variable (n = 172) (n = 159) Age at migraine onset, y 20.8 (10.7) 19.7 (9.8) Migraine attack duration, h 18.6 (16.6) 18.1 (20.1) Monthly rate of migraine episodes 6.1 (2.7) 6.2 (2.7) Monthly rate of total headache days 8.7 (3.1) 8.4 (2.9) No. of days per month ofacute abortive medication use 6.50 (3.0) 6.1 (3.1) (Figure 3). Because the 2-sided 95% CI was entirely above -1 and contained 0, topiramate was assumed to be noninferior to amitriptyline in terms of the primary efficacy measure. Prespecified Secondary Efficacy Measures During the prospective baseline phase, subjects in the topiramate and amitriptyline treatment groups had a mean (SD) monthly rate of 7.4 (2.9) and 7.1 (2.6) migraine days, respectively (Table IV). During double-blind treatment, the LSM change from baseline in the monthly rate of migraine days was -3.2 in the topiramate treatment group and -3.1 in the amitriptyline treatment group (95% CI, -0.9 to 0.7; P = NS). During the prospective baseline phase, subjects in the topiramate and amitriptyline treatment groups had a mean (SD) monthly rate of total headache days of 8.7 (3.1) and 8.4 (2.9), respectively (Table IV). During double-blind treatment, the LSM change from baseline in the monthly rate of total headache days was March

9 Clinical Therapeutics Randomization (N - 347) I TPM (n=178), I Discontinued Subject's choice (1) AMI (n=169), Discontinued Lost to follow-up (3) Limiting AEs (1) Other(1) TPM Evaluable AMI (n = 177) for safety (n=169) t Discontinued Subject's choice (2) Lost to follow-up (3) Limiting AEs (4) Other(1) TPM ITI AMI (n=172) (n=159) Discontinued Discontinued Subject's choice (13) Subject's choice (15) Significant protocol Significant protocol violation (2) violation (8) Lost to follow-up (9) Lost to follow-up (9) Limiting AEs (34) Limiting AEs (34) Other (6) Lack of efficacy (2) TPM AMI Com pieters Other (2) (n=102) (n = 95) Figure 2. Subject disposition. TPM = topiramate; AMI = amitriptyline; AEs = adverse events; ITT = intent to treat in both treatment groups (95% CI, -0.9 to 0.8; P = NS). For the analysis of response rates based on headache days, the proportion of patients experiencing a ::::50% reduction in mean monthly migraine days did not differ significantly between the topiramate and amitriptyline treatment groups (55.6% and 45.9%, respectively). For the analysis of response rates based on total headache days, the proportion of patients with a ::::50% reduction in mean monthly total headache days did not differ significantly between the topiramate and amitriptyline treatment groups (54.4% and 43.9%). There were no significant differences between groups in the mean monthly changes from prospective baseline to the end of the double-blind treatment phase in any of the other prespecified secondary efficacy measures (Table IV). Measures of Disability, Quality of Life, and Weight Satisfaction During the prospective baseline phase, subjects in the topiramate and amitriptyline treatment groups had mean (SD) scores on the functional disability question of 2.4 (0.4) and 2.3 (0.5), respectively. The LSM change from baseline in scores on the functional disability question was significantly greater in the topiramate group compared with the amitriptyline group (-0.33 [0.6] vs [0.7], respectively; 95% CI, -0.3 to 0.0; P = 0.040, ANCOVA) (Figure 4). During the prospective baseline phase, subjects in the topiramate treatment group had mean (SD) scores on the RR, RP, and EF domains of the MSQ of 55.8 (16.3),68.8 (20.1), and 55.9 (26.6), respectively (Figure 5). The corresponding scores in the amitriptyline treatment group were 55.7 (15.2), 72.2 (17.8), and 550 Volume 31 Number 3

10 D.W. Dodick et al <l) c: v III -1.0 E 2 l.l -1.5 <l) 0() c: rn Topiramate Amitriptyline (n-l72) (n - 159) o+-_,...-_,...-_..l.-_ -2.6 p Figure 3. Least squares mean (LSM) change from baseline in the monthly (28-day) rate of migraine episodes (intent to treat) (24.9). The LSM changes from baseline in the 3 domains were significantly greater in the topiramate group compared with the amitriptyline group. For the RR domain, the LSM change was 23.7 for topiramate and 18.4 for amitriptyline (95% CI, ; P = 0.012, ANCOVA); for the RP domain, the respective values were 16.7 and 12.5 (95% CI, ; P = 0.014, ANCOVA), and for the EF domain, 25.6 and 20.5 (95% CI, ; P = 0.029, ANCOVA). MIDAS scores at both baseline and the final visit were available for 152 of 172 subjects in the topiramate group and 143 of 159 subjects in the amitriptyline group. Mean (SD) overall MIDAS scores during the prospective baseline phase were 26.4 (19.6) in the topiramate group and 25.5 (20.4) in the amitriptyline group. Mean changes in MIDAS scores did not differ significantly between the topiramate and amitriptyline groups (-12.1 [23.4] and [20.7], respectively). Mean (SD) Q-LES-Q-SF scores at baseline were 65.9 (15.7) and 65.3 (13.4) in the topiramate and amitriptyline treatment groups, respectively. The LSM change from baseline did not differ significantly between the topiramate and amitriptyline groups (4.6 and 4.9, respectively; 95% CI, -3.1 to 2.6). Mean (SD) weight satisfaction scores during the prospective baseline phase were -0.7 (1.8) and -1.0 (1. 7) in the topiramate and amitriptyline groups, respectively. At the end of the study, subjects in the topiramate group reported an overall increase in weight satisfaction from baseline, based on an LSM improvement in WSSQ score of 0.8, whereas the amitriptyline group reported an overall deterioration in weight satisfaction, based on an LSM decrease in WSSQ score of 0.3. The between-group difference was statistically significant (95% CI, ; P < 0.001, CMH test). Body Weight and Body Mass Index Mean (SD) baseline body weight was 76.8 (19.4) kg in the topiramate group and 77.3 (20.7) kg in the amitriptyline group (Table II). The mean weight change from baseline was -2.4 (4.2) kg in the topiramate group and +2.4 (3.9) kg in the amitriptyline group. Nearly two thirds (64.6%) of amitriptyline-treated subjects gained ::::1 % of body weight during the study and 28.5% gained ::::5% of body weight, compared with a respective 17.0% and 4.1 % of topiramatetreated subjects (Figure 6). In contrast, nearly two thirds (64.4%) of topiramate-treated subjects lost ::::1 % of body weight during the study and 29.9% lost ::::5% of body weight, compared with 19.0% and 3.2% of amitriptyline-treated subjects. The mean (SD) baseline BMI was 27.8 (6.7) kg/m 2 in the topiramate group and 28.5 (7.6) kg/m 2 in the amitriptyline group (Table II). The mean changes in BMI from baseline were small in both the topiramate and amitriptyline groups (-0.9 [1.5] kg/m 2 and +0.9 [1.5] kg/m 2, respectively). Analysis of the categorical changes in BMI from baseline to the final visit suggested that 29 of 177topiramate-treated subjects (16.4%) and 5 of 169 amitriptyline-treated subjects (3.0%) moved to a lower BMI category, and 1 topiramatetreated subject (0.6%) and 18 amitriptyline-treated subjects (10.7%) moved to a higher BMI category (Table V). Safety Profile No substantial changes in mean laboratory test values were observed in either treatment group (Tables VI and VII), although amitriptyline-treated subjects had a numerically greater LSM change from baseline in hs-crp, a significant LSM reduction in HbA 1c (P = 0.005), and a numerically lower LSM change in IL-6 over 26 weeks of treatment (Table VI). March

11 til til N Q :::l ;:;. -l ::r I'D P:l "'0 I'D... l: ;:;. III Table IV. Least squares mean (LSM) monthly changes from prospective baseline in prespecified secondary efficacy measures in the topiramate (n = 172) and amitriptyline (n = 159) groups. Prospective Baseline, Mean LSM Monthly Change From Baseline Secondary Efficacy Measure Topiramate Amitriptyline Topiramate Amitriptyline P (95% CI) Total migraine days (-0.9 to 0.7) Total headache days (-0.9 to 0.8) Rate of acute abortive medication use, d (-0.4 to 0.9) Migraine attack duration, d (-0.4 to 0.4) Migraine severity (-0.2 to 0.1) Severity of migraine-related nausea (-0.2 to 0.1) Severity of migraine-related photophobia (-0.2 to 0.1) Severity of migraine-related phonophobia (-0.2 to 0.1) Frequency of migraine-related nausea (-0.5 to 0.5) Frequency of migraine-related photophobia (-0.5 to 0.7) Frequency of migraine-related phonophobia (-0.5 to 0.7) Frequency of migraine-related vomiting (-0.3 to 0.4) i: 3I'D (JJ... Z l: 3 c:r I'D... (JJ

12 D.W. Dodick et al <li c v Vl oj CO E LL Topiramate Amitriptyline (n=l72) (n = 159) <li b.o c oj..c U ::?: P Figure 4. Least squares mean (LSM) change from baseline in mean migraine severity on the functional disability question (intent to treat). In terms of vital signs, topiramate- and amitriptylinetreated subjects had small mean (SD) changes from baseline to the final visit in systolic blood pressure (-2.7 [12.5] and 0.9 [12.7] mm Hg, respectively) and diastolic blood pressure (-2.1 [8.8] and 1.4 [9.8] mm Hg). Mean changes in heart rate from baseline to the final visit were -1.4 (10.7) beats/min in the topiramate group and +6.0 (10.9) beats/min in the amitriptylinetreated group. No deaths were reported during the study or within 30 days after study completion. TEAEs were reported in 152 of 177 subjects (85.9%) in the topiramate group and 150 of 169 subjects (88.8%) in the amitriptyline group (Table VIII). The most commonly occurring TEAEs (reported in ::::5% of subjects during the double-blind phase) in the topiramate group were paresthesia (29.9%), fatigue (16.9% ),somnolence (11.9%), hypoesthesia (10.7%), and nausea (10.2%). The most commonly reported TEAEs in the amitriptyline group were dry mouth (35.5%), fatigue (24.3%), somnolence (17.8%), weight increase (13.6%), dizziness (10.7%), and sinusitis (10.7%). TEAEs of mild or moderate severity were reported in 118 subjects (66.7%) in the topiramate group and 112 subjects (66.3 %) in the amitriptyline group; marked TEAEs were reported in 34 (19.2%) and 38 (22.5%) subjects, respectively. Treatment-related AEs occurred in 121 subjects (68.4%) in the topiramate group and 128 subjects (75.7%) in the amitriptyline group. In the topiramate treatment group, the most frequent TEAEs leading to study withdrawal (accounting for >1.1 % of subjects) were fatigue (3.4%), dizziness (1.7%), hypoesthesia (1. 7%), anxiety (1. 7%), and confusion (1. 7%). In the amitriptyline treatment group, the most frequent TEAEs leading to study withdrawal (accounting for >1.2% of subjects) were weight increase (4.7%), somnolence (4.1%), fatigue (2.4%), aggravation of migraine (1.8%), and dry mouth (1.8%). Twelve subjects-4 in the topiramate treatment group and 8 in the amitriptyline treatment grouphad a total of 14 SAEs. The SAEs were treatment emergent in 11 subjects (4 in the topiramate group, 7 in the amitriptyline group). The 4 SAEs in the topiramate group were menorrhagia (1), menstrual disorder and vaginal hemorrhage (1), injury and purpura (1), and ovarian disorder (1). With the exception of injury, which was classified as moderate, all SAEs in the topiramate group were classified as marked in severity; all SAEs in this group were considered unrelated or of doubtful relationship to study treatment. The 8 SAEs in the amitriptyline group were renal calculus, migraine aggravated, breast neoplasm (malignant), gastroenteritis, esophagitis, menorrhagia, cholecystitis, and brain neoplasm (benign). All SAEs in the amitriptyline treatment group were classified as marked in severity, with the exception of menorrhagia and brain neoplasm (benign), which were classified as moderate. The renal calculus, which had an onset on day 39 and had resolved by day 40, was considered by the investigator to be very likely related to amitriptyline treatment. DISCUSSION In this noninferiority analysis, topiramate was at least as effective as amitriptyline in terms of the LSM reduction in the monthly rate of migraine episodes from prospective baseline to the end of the double-blind treatment phase. There were no statistically significant differences between groups in any of the prespecified secondary efficacy variables. March

13 Clinical Therapeutics o Topiramate(n = 170) Amitriptyline (n = 157) p = <lj C <lj VI ro 20 dl E 0 '- LL <lj C ro..c U 30 P= ";l P = Role Function- Restrictive Role Funetion Preventive Emotional Function Mean baseline Figure S. Least squares mean (LSM) change from baseline in Migraine-Specific Quality of Life Questionnaire scores (intent to treat). o Topiramate (n = 172) Amitriptyline (n = 159) VI w u<lj 25 E ::J 20 Vl I+- 0?f? o 10% Decrease 5%to <10% Decrease 1%to<5% Decrease <1% Decrease to <1% Increase 1%to <5% Increase 5%to <10% Increase 10% Increase Weight Change Figure 6. Percent weight change from baseline (intent to treat). 554 Volume 31 Number 3

14 s: ri ::r N o o 10 Table V. Number (%) of patients with a change in body mass index (BMI) category from baseline* to the final visitt (safety population). Topiramate (n = 177)* Amitriptyline (n = 169) Final BMI Category, kg/m 2 Final BMI Category, kg/m 2 Baseline BMI Missing Missing Category, kg/m 2 <18.5 :?:18.5-<25 :?:25-<30 :?:30 Data <18.5 :?:18.5-<25 :?:25-<30 :?:30 Data < (2.8) (0.6) 1 (0.6) :?:18.5-<25 5 (2.8) 64 (36.2) 50 (29.6) 8 (4.7) 4 (2.4) :?:25-<30 10 (5.6) 29 (16.4) 1 (0.6) 3 (1.7) - 35 (20.7) 9 (5.3) 4 (2.4) :?: (7.9) 42 (23.7) 3 (1.7) (3.0) 49 (29.0) 3 (1.8) * Last available measurement before the first administration of study medication. t Last available measurement during double-blind treatment. *One subject in the topiramate group was evaluable for safety but did not have a baseline value and was not included in this table. til til til o o o0... ;:;. " I'D... e!...

15 Clinical Therapeutics Table VI. Changes from baseline to the final visit in high-sensitivity C-reactive protein (hs-crp), glycosylated hemoglobin (HbA,c)' and interleukin-6 (IL-6) (safety population). L5M Change From Baseline Topiramate Amitriptyli ne Analyte (n 177) (n 169) p* hs-crp, mg/l n 148 n HbA 1c,% n 147 n IL-6, pg/ml n 135 n LSM = least squares mean. *Analysis-of-covariance model including treatment and center as main effects, and baseline as a covariate. Table VII. Mean changes from baseline to the final visit in selected clinical laboratory parameters (safety population). Topiramate Amitriptyline Topiramate Amitriptyline Analyte* (n = 177) (n = 169) Analyte* (n 177) (n = 169) Chloride, mmol/l n = 169 n = 157 Aspartate Baseline aminotransferase, U/L n = 169 n = 157 Final visit Baseline Mean (SO) change 1.81 (3.020) (2.464) Final visit Mean (SO) change (6.695) 2.40 (12.589) Creatinine, mg/dl n = 146 n 143 Baseline Final visit Mean (SO) change 0.05 (0.124) 0.02 (0.116) Blood urea nitrogen, mg/dl n = 146 n 143 Baseline Final visit Mean (SO) change 0.34 (3.287) (3.335) Alanine aminotransferase, U/L n 169 n = 157 Baseline Final visit Mean (SO) change (13.678) 5.00 (27.326) Carbon dioxide,t Total bilirubin, mg/dl n = 169 n = 157 mmol/l n 169 n = 157 Baseline Baseline Final visit Final visit Mean (SO) change (0.201) (0.207) Mean (SO) change (2.725) 0.76 (2.464) *The number of subjects shown for each analyte represents the number with values at baseline and the final visit. t Bicarbonate (HCO-3) was reported as carbon dioxide. 556 Volume 31 Number 3

16 D.W. Dodick et al. Table VIII. Treatment-emergent adverse events reported in 2:5% of subjects (safety population). Adverse Event Topiramate (n = 177) Amitriptyline (n = 169) p No. (%) ofsubjects with any adverse event* 152 (85.9) 150 (88.8) Specific adverse events, no. (%) Paresthesia Fatigue Somnolence Hypoesthesia Nausea Dizziness Sinusitis Viral infection Upper respiratory tract infection Dry mouth Anorexia Difficulty with concentration/attention Taste perversion Dyspepsia Abnormal vision Headache Coughing Pharyngitis Consti pation Weight increase 53 (29.9) 30 (16.9) 21 (11.9) 19 (10.7) 18 (10.2) 15 (8.5) 14 (7.9) 14 (7.9) 14 (7.9) 12 (6.8) 12 (6.8) 12 (6.8) 10 (5.6) 9 (5.1) 9 (5.1) 9 (5.1) 9 (5.1) 8 (4.5) 6 (3.4) o 8 (4.7) 41 (24.3) 30 (17.8) 6 (3.6) 12 (7.1) 18 (10.7) 18 (10.7) 11 (6.5) 11 (6.5) 60 (35.5) 8 (4.7) 5 (3.0) 6 (3.6) 14 (8.3) 9 (5.3) o 7 (4.1) 11 (6.5) 14 (8.3) 23 (13.6) < < <0.001 *Subjects with>1 occurrence ofthe same adverse event were counted only once for that event. Topiramate-treated subjects in this study had a significant reduction compared with amitriptylinetreated subjects in the mean severity of self-reported functional disability associated with migraine episodes (P = 0.040) and improvements on all 3 domains of the MSQ (RR: P = 0.012; RP: P = 0.014; EF: P = 0.029). The observed improvements in MSQ scores with topiramate extend the results of previous trials in which treatment with topiramate 100 mg/d was associated with clinically significant improvements in all 3 MSQ domains compared with placebo. 30 Changes from baseline in scores on the Q-LES-Q-SF and MIDAS were comparable in the 2 treatment groups. The incidence of AEs was comparable in the 2 treatment groups, and the nature of the AEs was consistent with the established AE profiles of the 2 medications at the dosages used. The most commonly reported TEAEs in the topiramate group were paresthesia (29.9%), fatigue (16.9%), somnolence (11.9%), hypoesthesia (10.7%), and nausea (10.2%). The most commonly reported TEAEs in the amitriptyline group were dry mouth (35.5%), fatigue (24.3%), somnolence (17.8%), weight increase (13.6%), dizziness (10.7%), and sinusitis (10.7%). There was no significant difference between groups in the incidence of discontinuations due to AEs. Numerically more amitriptyline-treated subjects than topiramate-treated subjects gained ::::1 % and ::::5% of their baseline body weight during double-blind treatment. During the same period, numerically more topiramatetreated subjects than amitriptyline-treated subjects lost ::::1 % and ::::5% of their baseline body weight. Numerically more topiramate-treated subjects moved to a lower BMI category during the study, and numerically more amitriptyline-treated subjects moved to a higher BMI category. The impact of body weight change over the March ?

17 Clinical Therapeutics course of the study may have exerted an effect on diastolic and systolic blood pressure, heart rate, total cholesterol, low-density lipoprotein cholesterol, triglycerides, HbA 1c ' and hs-crp, but the significance and clinical relevance of such changes require further investigation, as does the clinical significance of body weight change associated with migraine preventive treatment. The study outcomes may have been limited by use of the Q-LES-Q-SF and MIDAS. The Q-LES-Q-SF uses a recall period of 1 week, which may not be adequate to fully capture the impact on QoL of a chronic disorder with an episodic impact such as migraine. In addition, this instrument measures various aspects of life that are not generally relevant to migraine, potentially rendering it less sensitive to change over time. MIDAS uses a 3-month recall period; some subjects provided an endpoint assessment within 3 months of their baseline assessment, resulting in an overlap of recall periods. The effect of these potential limitations on the outcomes of this trial cannot be determined. In addition, the exclusion criteria used may limit the ability to extrapolate the results to the general population of migraineurs, including those who are refractory to preventive migraine medications, those overusing acute headache medications, and those with comorbid conditions. CONCLUSIONS The results of this study suggest that topiramate 100 mg/d was at least as effective as amitriptyline 100 mg/d in reducing the mean monthly rate of migraine episodes. There were no significant differences between treatments in terms of improvements in any of the prespecified secondary efficacy end points. Topiramate treatment was associated with improvements in some QoL indicators, weight loss, and improved weight satisfaction compared with amitriptyline. Both treatments appeared to be well tolerated in this population with episodic migraine. ACKNOWLEDGMENTS The design, conduct, and analysis of this study were funded by Ortho-McNeil Janssen Scientific Affairs, LLC. Editorial support was provided by George Rogan, MSc, Phase Five Communications Inc., New York, New York, with funding provided by Ortho McNeil Janssen Scientific Affairs. REFERENCES 1. Lipton RB, Bigal ME, Diamond M, et ai, for the AMPP Advisory Group. Migraine prevalence, disease burden, and the need for preventive therapy. Neurology. 2007;68: Leonardi M, Steiner Tj, Scher AT, Lipton RB. The global burden of migraine: Measuring disability in headache disorders with WHO's Classification of Functioning, Disability and Health (ICF). J Headache Pain. 2005;6: Ferrari MD. The economic burden of migraine to society. Pharmacoeconomics. 1998;13: Hu XH, Markson LE, Lipton RB, et al. Burden of migraine in the United States: Disability and economic costs. Arch Intern Med. 1999;159: Goldberg LD. The cost of migraine and its treatment. Am JManag Care. 2005;11 (Suppl):S62-S Silberstein SD, Neto W, Schmitt j, jacobs D, for the MIGR-001 Study Group. Topiramate in migraine prevention: Results ofa large controlled trial. Arch Neurol. 2004; 61 : BrandesjL, Saper jr, Diamond M, et ai, for the MIGR 002 Study Group. Topiramate for migraine prevention: A randomized controlled trial.jama. 2004;291 : Diener HC, Tfelt-Hansen P, DahlbfC, et ai, for the MIGR 003 Study Group. Topiramate in migraine prophylaxisresults from a placebo-controlled trial with propranolol as an active control.j Neurol. 2004;251 : Silberstein SD, Lipton RB, Dodick DW, et ai, for the Topiramate Chronic Migraine Study Group. Efficacy and safety of topiramate for the treatment of chronic migraine: A randomized, double-blind, placebo-controlled trial. Headache. 2007;47: Diener HC, Bussone G, Van Oene jc, et ai, for the TOP MAT-MIG-201 (TOP-CHROME) Study Group. Topiramate reduces headache days in chronic migraine: A randomized, double-blind, placebo-controlled study [published correction appears in Cephalalgia. 2007;27:962]. Cephalalgia. 2007;27: Snow V, Weiss K, Wall EM, Mottur-Pilson C, for the American Academy of Family Physicians and the American College of Physicians-American Society of Internal Medicine. Pharmacologic management of acute attacks of migraine and prevention of migraine headache. Ann Intern Med. 2002;137: Ramadan NM, Silberstein SD, Freitag FG, et ai, for the US Headache Consortium. Evidence-based guidelines for migraine headache in the primary care setting: Pharmacological management for prevention of migraine. fsl gl pd f. Accessed October 30, Modi S, Lowder DM. Medications for migraine prophylaxis [published correction appears in Am Fam Physician. 2006;74:1685]. Am Fam Physician. 2006;73: Krymchantowski AV, Silva MT, BarbosajS, Alves LA. Amitriptyline versus amitriptyline combined with fluoxetine in SS8 Volume 31 Number 3