Abstract
Objective: The study sought to evaluate whether topiramate prevents development of chronic daily headache (CDH, ≥15 headache days per month) in adult subjects with high-frequency episodic migraine (HFEM, 9–14 migraine headache days/month). A secondary objective was to assess the efficacy of topiramate as preventive migraine treatment in this population.
Methods: This was a multicenter, randomized, double-blind, placebo-controlled study comparing topiramate 100 mg/day and placebo for 26 weeks. The primary efficacy variable was new-onset CDH at month 6. Secondary efficacy measures included migraine and headache days. Adverse events (AEs) were evaluated.
Results: A total of 159 topiramate subjects and 171 placebo subjects were efficacy-evaluable. At month 6, 1.4% of topiramate subjects versus 2.3% of placebo subjects had CDH (p = .589). Compared with placebo, topiramate treatment was associated with statistically significant reductions in mean number of migraine days (6.6 vs. 5.3/28 days; p = .001) and headache days (6.6 vs 5.3/28 days; p = .001). Most commonly reported AEs in the topiramate versus placebo group included paresthesia (32.4% vs. 7.0%), fatigue (14.8% vs. 8.6%), dizziness (11.4% vs. 7.6%) and nausea (10.8% vs. 9.2%).
Conclusion: Topiramate 100 mg/day did not prevent the development of CDH at six months in subjects with HFEM. Topiramate was effective in reducing headache days and migraine headache days and generally well tolerated.
Introduction
Migraine is a common, disabling, costly primary headache disorder that affects approximately 17% of women and 5.6% of men worldwide (1–7). Migraine is increasingly recognized as a chronic neurological disorder with episodic attacks and a variable clinical course (8,9). A subgroup of migraineurs experiences a gradual escalation of headache frequency over a period of months to years (2,3,10,11). This process may lead to the development of a form of chronic daily headache (CDH) variously called transformed migraine or chronic migraine (12,13). Chronic migraine is associated with higher levels of disability, resource use and comorbidity than is episodic migraine (14).
Topiramate is approved for use in adults for migraine prevention. Several large, randomized, placebo-controlled clinical trials have shown that topiramate 100 mg/day significantly reduced the number of migraine headache days in patients with episodic migraine who experienced between three and 12 migraine episodes per month (15,16) and in patients with chronic migraine who experienced ≥15 headache days per month (17,18). In these clinical trials, topiramate treatment was safe and generally well-tolerated.
High monthly headache frequency is a risk factor for the progression of episodic migraine to chronic migraine (10,19–21). The prospect of preventing new-onset CDH in high-risk individuals with episodic migraine has been discussed, and clinical trial designs have been proposed (22). In a post-hoc analysis of pooled data from three clinical trials involving patients with episodic migraine treated with topiramate 100 mg/day, Limmroth and colleagues provided indirect evidence that preventive treatment with topiramate may be associated with a reduced risk of migraine progression (23).
The aggregate of epidemiological information related to the risk of migraine progression and exploratory analysis of pooled clinical trial data provided the rationale for this study: to evaluate whether topiramate treatment would prevent the transformation of episodic migraine headache to CDH in patients with a history of experiencing a high monthly frequency of migraine headaches. A secondary objective was to evaluate the efficacy of topiramate treatment in reducing the frequency of headache days and migraine days in this population.
Methods
Study design
This was a multicenter (87 sites), randomized, double-blind, placebo-controlled, parallel-group study [study ID CR002854] of adults with episodic migraine headache who were experiencing frequent migraine episodes. The study consisted of a pretreatment phase, a double-blind phase, and a taper/exit phase (Figure 1). It was conducted between September 2005 and August 2007.
Study design.
Pretreatment phase
The pretreatment phase of the study lasted up to 70 days and consisted of two study periods: a screening/washout period (day −70 to day −28) and a baseline period (day −28 to day −1).
Screening/washout period
Subjects with a history of migraine as defined by diagnostic criteria in the International Classification of Headache Disorders, second edition (ICHD-II) (24), for at least one year prior to screening and who met all other entry criteria for the study were identified. Eligible subjects were required to be considered at risk for progression of episodic migraine to chronic migraine based on a prior history of experiencing migraines at a high monthly frequency, which, for the purpose of this study, was defined as occurrence of migraine headache on at least 9 but <15 days and <15 total headache days over the 28 days before the screening visit. The migraine headaches must have met all ICHD-II diagnostic criteria for migraine except for duration. In this study, headache episodes with a duration of at least 30 minutes were included in the analysis.
Subjects signed an informed consent and had a complete medical and medication history taken at visit 1 (day −70 to day −28). All medications used for migraine prevention were discontinued at least six weeks prior to entering the baseline period on day −28 (visit 2). Once screening procedures were completed, subjects returned for their baseline visit within one week.
Baseline period
The baseline period began on day −28 (visit 2), when the subject completed the screening/washout period. This phase was a period of between 28 and 35 days, during which subjects who continued to have 9–14 migraine headache days/28 days and <15 total headache days/28 days could be randomized into the double-blind phase of the study. Subjects were given headache records and appropriate instructions for recording comprehensive assessments.
Double-blind phase
Subjects who completed the baseline period, had the required baseline frequency of migraine headache days and met all other eligibility criteria were considered eligible for participation and could be randomized 1 : 1 in double-blind fashion on day 1 (visit 3) to daily treatment with either topiramate or matching placebo. The double-blind phase lasted 26 weeks and consisted of two study periods: a six-week titration period and a 20-week maintenance period. Throughout the double-blind phase, subjects recorded assessments in their headache records and were permitted to continue acute treatment of migraine headaches as necessary.
Titration period. All subjects who entered the titration period initiated treatment with a single 25-mg tablet of study medication (topiramate or matching placebo) administered in the evening of day 1 (visit 3) and continuing through day 7. The dosage of study medication was then increased each week by a single 25-mg tablet per day until a total daily dosage of 100 mg/day (two 25-mg tablets [50 mg] twice daily [BID]) was achieved. The titration of study medication was adjusted at the discretion of the investigator on the basis of subject tolerability. Subjects must have maintained a dose of at least 75 mg/day of study medication (topiramate or matching placebo) beginning at day 42 and throughout the remainder of the study.
Maintenance period. Following successful completion of the titration period, subjects entered the maintenance period. The dose of study medication taken at the end of the titration period was maintained during the maintenance period. Clinic visits occurred on day 42 (visit 4), day 70 (visit 5), day 98 (visit 6), day 126 (visit 7), day 154 (visit 8) and the final maintenance visit on day 182 (visit 9). If a subject discontinued the study prematurely, final visit procedures (day 182 [visit 9]) were performed at the time of discontinuation.
Taper/exit phase
For subjects who exited the study, either those who discontinued prematurely or those who completed the double-blind phase, taper from the study medication was recommended. For subjects entering the taper/exit phase, the daily dose of study medication was reduced by 50% for seven days and then discontinued. Subjects returned for the final taper/exit phase visit on visit 10 (day 196) approximately one week after completing the tapered discontinuation of study medication.
Ethics
The study was performed in accordance with International Conference on Harmonization Guidelines on Good Clinical Practice and applicable regulatory requirements consistent with the principles originated in the Declaration of Helsinki. Before any subjects were enrolled in the study at any site, the protocol, amendments, and informed consent forms were approved by the appropriate local Institutional Review Board, and each subject, or a legally acceptable representative, provided written informed consent.
Subjects
Inclusion criteria. Eligible subjects included adults from 18–65 years of age with an established history of migraine headache (ICHD-II 1.1, 1.2) for at least 12 months before entering the screening period. Subjects were required to have at least 9 but <15 migraine headache days and <15 total headache days over the 28 days before the screening visit and during the 28-day baseline period. Subjects had to be in generally good health, as confirmed by medical history, baseline physical examination, baseline neurological exam, vital signs and clinical laboratory evaluations and to be capable of taking oral medication. Females had to be postmenopausal for at least one year, surgically sterile or otherwise incapable of pregnancy, or using an acceptable method of birth control. In addition, female subjects of childbearing potential had to have a negative result on a urine pregnancy test before beginning study medication.
Exclusion criteria. Subjects who met any of the following criteria were excluded from participation in the study:
Previously failed more than two adequate trials of medications from different drug classes used for migraine prophylaxis because of a lack of efficacy, or used a medication generally considered to be effective for migraine prevention in the six weeks before visit 2 (initiation of baseline period) Previously discontinued topiramate therapy because of a lack of efficacy (defined as a trial of at least three months’ duration at an adequate dose of medication that in the investigator’s opinion did not result in clinical benefit) or discontinued topiramate therapy because of an adverse event (AE) Had onset of migraine after age 50, had exclusively migraine aura without headache or, at the time of screening, had an equally painful or more painful condition than their headache pain (e.g. osteoarthritis, herniated nucleus pulposus) or had cluster headache or basilar or hemiplegic migraine Had used a combination of acute headache medications for any reason for >4 days/week on a regular basis during the three months before visit 2 Had a progressive neurological disorder other than migraine; a malignancy or a history of malignancy within the past five years, except for a basal cell carcinoma that was treated with local excision and was no longer present; a significant medical history or medical condition of neurological, cardiovascular, hepatic or renal disease (e.g. neurodegenerative or cerebrovascular disorder or stroke; angina, uncontrolled hypertension or myocardial infarction; chronic active hepatitis or chronic liver or renal disease); nephrolithiasis or any unstable medical condition that may have impaired a subject’s reliable participation in the study or necessitate the use of medications not permitted in the study Renal or liver function tests at least two times the upper limit of the normal (ULN) range or abnormal screening laboratory tests exceeding any of the following limits: alanine transaminase or aspartate transaminase >2x ULN); total white blood cell count <2300/mm3 or >2x ULN; platelet count <80,000/mm3; serum creatinine >2x ULN Any history of suicide attempt or suicidal ideation or of a major psychiatric disorder A history of drug or alcohol abuse within the past two years and a positive urine drug screen for amphetamines, cocaine metabolite, marijuana metabolite, methadone, methaqualone, phencyclidine, propoxyphene or alcohol
Randomization and blinding
During the screening period, subjects who met the inclusion criteria and entered the study were given a five-digit subject number. The first two digits represented the investigator number, and the last three digits represented the subject in ascending sequential order beginning with 001 (e.g. investigator 01, first subject = 01001). This subject number was retained for the duration of the study. At the time of randomization (visit 3), subjects were assigned to either of the two treatment groups based on a computer-generated predetermined randomization schedule prepared by the sponsor before the study. Randomization sequences were generated for each site. Stratification was not done for other covariates or potential confounders such as age, gender and attack frequency (within the eligible range). Medication code numbers were preprinted on study medication labels and assigned as subjects qualified for the study and were randomized to treatment. Sealed envelopes containing the study medication identification (e.g. active or placebo) were provided to the investigator and kept in a limited access area accessible 24 hours/day. The double-blind study medication tablets were identical in appearance and packaged in identically appearing bottles.
Concomitant medication use
All medication taken from the time of visit 1 until visit 10 was documented. Subjects were permitted to take acute headache medication as indicated throughout the study. The type and method of acute headache medication use was as consistent as possible with that used by the subject prior to enrollment.
Subjects with concurrent illnesses who required prescription or over-the-counter medication within one month of entering the study were permitted, if necessary, to continue their medication or to initiate such therapy at the start of the study or any time during the study. Subjects’ preexisting treatment regimens were modified only when deemed clinically appropriate by the investigator and not for the explicit purpose of entering the trial.
All migraine preventive medications were discontinued at least six weeks before visit 2 and for the duration of the trial. The list of pharmacological agents considered potentially effective preventive migraine therapy was derived from data prepared for the Agency for Health Care Policy and Research (25) and by the US Headache Consortium (26).
Definitions used in this study
Baseline value. For headache-related variables, the baseline value was defined based on observations made during the 28-day baseline period. The final date during the double-blind phase was defined as the date the subject stopped study medication prior to taper. For all other efficacy and safety variables, the baseline value was defined as the last non-missing observation before the first administration of study medication during the titration period. The final value during the double-blind phase was defined as the last non-missing observation before or on final visit 9 during the double-blind phase.
Migraine headache. This was defined as headache pain that met the ICHD-II definition (1.1, 1.2) for migraine headache, with the exception of headache duration.
Headache pain duration. For the purposes of this study, the minimum required duration of headache pain to qualify as a migraine headache or non-migraine headache was at least 30 minutes.
Calendar day and headache day. A calendar day extended from 12:00 midnight to 11:59
Migraine headache day. A migraine headache day was defined as a calendar day (see definition above) in which the subject experienced headache pain of at least 30 minutes that otherwise met the ICHD-II definition for migraine headache with or without aura. Once a headache met the definition of migraine headache, that calendar day was designated a migraine headache day independent of whether non-migraine headaches occurred within the same calendar day. Headache pain that met the criteria for migraine headache and began at for example at 11:45
Efficacy measures
Primary efficacy measure. The primary efficacy measure was whether a subject reported ≥15 headache days (migraine or non-migraine) per 28-day period at month 6.
Secondary efficacy measures. Secondary efficacy variables included:
Whether a subject reported ≥15 headache days during the last 28 days of the double-blind phase for those subjects who completed at least 28 days of the double-blind phase Time to the first reporting of ≥15 headache days per 28-day period, which was defined as the last day of the first 28-day period with ≥15 headaches Whether a subject reported ≥15 headache days, of which at least half were migraine headaches during the last 28 days in the double-blind phase Time to the first reporting of ≥15 headache days, of which at least half were migraine headaches per 28-day period, which was defined as the last day of the first 28-day period with the aforementioned characteristics Change from baseline through the double-blind phase in 28-day rate of headache days Change from baseline through the double-blind phase in 28-day rate of migraine days Change from baseline in acute medication use (days) per 28 days Categorical response: ≥50% and ≥75% reduction in headache days and migraine headache days Changes from baseline in the 28-day frequency of nausea, phonophobia and photophobia. Subjects recorded the severity of migraine-associated symptoms (nausea, photophobia, phonophobia) using the following four-point scale: 0 = none, 1 = mild, 2 = moderate, and 3 = severe Health-related quality of life (HRQoL) end points and measures of migraine disability were also evaluated and included Migraine-Specific Quality-of-Life Questionnaire (MSQ) (27) and Migraine Disability Assessment (MIDAS) (28)
Safety evaluations
AEs were reported by the subject or when appropriate by a subject’s caregiver, surrogate or legally acceptable representative for the duration of the study. An AE was defined as any untoward medical occurrence or any unfavorable and unintended sign (including an abnormal finding), symptom, or disease in a subject temporally associated with the use of a medicinal or investigational product, whether or not related to the medicinal or investigational product. This included any occurrence that was new in onset or aggravated in severity or frequency from the baseline condition, or abnormal results of diagnostic procedures including laboratory test abnormalities.
AEs were documented by the investigator, who assessed the level of severity (mild, moderate or severe), whether the AE was serious (SAE) and whether there was any relationship to the treatment. Brief physical and neurological exams, vital signs, clinical laboratory tests (hematology, chemistry and urinalysis) and liver function tests were performed. Urine pregnancy exams were done on women of childbearing potential. Any clinically significant abnormalities persisting at the end of the study were followed up by the investigator until resolution or until reaching a clinically stable endpoint.
Statistical methods
Analysis sets. The intent-to-treat (ITT) analysis set comprised randomized subjects who received at least one dose of study drug and had at least one post-dose efficacy assessment. The efficacy-evaluable (EE) analysis set comprised ITT subjects who completed at least 28 days of the double-blind phase. The safety analysis set included randomized subjects who took at least one dose of study drug and had at least one post-dose safety assessment.
Methods of analysis. Demographics and baseline information were summarized by treatment group. Continuous variables such as age, weight, height, age at migraine onset and number of headache days were summarized by descriptive statistics. Categorical variables such as gender and race are presented with frequency distributions (N, %) by treatment group. The number of subjects included in the ITT, EE and safety analysis sets is presented by treatment group. The primary analysis of the primary efficacy measure, whether a subject reported ≥15 headache days per 28-day period at month 6, was analyzed based on the EE analysis set. Six 28-day periods during the double-blind phase were designated as months 1 through 6. For each subject, a binary outcome of whether ≥15 headache days/28 days was experienced or not experienced was determined for each month. A generalized linear mixed model (GLMM) using a logit link function was used to analyze this repeated binary outcome data. The linear predictor was a cell means model that contains a random factor for subject and fixed terms for treatment, month and treatment by month interaction. The standard assumption of local independence of repeated measures within a subject given the subject effect was made. Baseline monthly headache day rate was included as a covariate in the model. The null hypothesis tested was that the difference between treatment groups at month 6 as measured by the log odds ratio was 0. The marginal probability of reporting ≥15 headache days at each month was estimated by generating random normal deviates from the estimated normal distribution of the subject effect. A plot of the observed monthly probabilities of reporting ≥15 headache days against those predicted by the GLMM was generated.
The aforementioned analysis was also performed based on the ITT analysis set, with the following rules: for subjects who were in the double-blind phase for ≤14 days, the month 1 rate was computed by combining baseline information with the ≤14 days in the double-blind phase and computing a normalized 28-day rate. For subjects in the double-blind phase for ≤28 days but >14, the month 1 rate was, to be consistent with above structure, computed as a normalized 28-day rate on available data. The primary efficacy data variable was also analyzed using another statistical approach, the generalized estimating equation model for the EE analysis set.
Secondary efficacy variables involving change from baseline and percent change in the mean 28-day rate during the double-blind phase were analyzed using analysis of covariance (ANCOVA) methodology with treatment and center as independent factors and baseline value (of the dependent) variable as a covariate. Categorical secondary variables were analyzed using the Cochran-Mantel-Haenszel test with modified ridit score, stratified by center. Analysis of time to the first reporting of ≥15 headache days per 28-day period, and time to the first reporting of ≥15 or more headache days, of which at least half were migraine, were analyzed using Kaplan-Meier (with a log rang test for treatment group difference) methodology and Cox’s proportional hazards model, with baseline headache days or migraine headache days as a covariate.
The percentage of subjects with treatment-emergent adverse events (TEAEs) is presented for each treatment group. Descriptive statistics were calculated for each laboratory analyte, vital signs, and the results of physical and neurological examinations at baseline and at each scheduled time point.
Sample size determination
Assuming a conversion rate (i.e. percent achieving ≥15 migraine headache days during the last 28 days in the double-blind phase) of 10.4% for placebo subjects and 2.6% for topiramate subjects, 236 subjects per group were necessary, assuming a type I error of 5% and power of 90%. It was assumed that 7% of randomized subjects would either not have taken study drug or would have the primary outcome missing. Therefore, it was determined that 254 subjects per treatment group needed to be randomized.
Because of slower-than-expected enrollment, the size of the study population needed to assess the primary efficacy variable was recalculated using an interim summary distribution of baseline headache frequency from the enrolled population. At that time, the distribution of baseline frequency headache days per 28 days was obtained and assessed in a blinded fashion. As a result, the sample size requirements were modified. Power was reduced to 80%, and 174 subjects per group were deemed necessary.
Results
Demographic and baseline headache characteristics
Subject demographics and baseline headache characteristics (efficacy-evaluable analysis set)
SD, standard deviation; BMI, body mass index; BMI (kg/m2) = weight (kg)/height (m)2.
Patient disposition
Based on the all-randomized analysis, 119/188 (63.3%) of subjects treated with topiramate 100 mg/day and 111/197 (56.3%) of placebo-treated subjects completed double-blind treatment (Figure 2). The most common reason for discontinuation of treatment was lost to follow-up in both treatment groups (25 [13.3%] of topiramate 100 mg/day– and 29 [14.7%] of placebo-treated subjects). Limiting AEs was the reason for discontinuation in 21 (11.2%) topiramate- and 18 (9.1%) placebo-treated subjects.
Subject disposition. TPM = topiramate. ITT = intent-to-treat.
Treatment exposure (EE analysis set)
The mean (± standard deviation [SD]) daily dose of study medication used during the maintenance period was 89.5 (±14.2) mg in the topiramate 100 mg/day group and 90.5 mg equivalent (±14.9) in the placebo group. The majority of subjects, 87 (54.7%), treated with topiramate and 87 (50.9%) treated with placebo, received study medication for at least 183 days. The mean (±SD) duration of treatment was 163.6 (±45.8) and 155.6 (±53.0) days for the topiramate and placebo treatment groups, respectively.
Efficacy measures
Primary efficacy measure. The proportion of subjects reporting ≥15 headache days/month at month 6 during the double-blind phase for the EE analysis was not statistically significantly different between treatment groups using the GLMM (topiramate 100 mg/day group 1.4% vs. placebo group 2.3%; p = .589) (Figure 3). No statistically significant differences between the topiramate and placebo treatment groups were observed using the generalized estimating equation model analysis.
Proportion of subjects reporting ≥15 headache days/month during the double-blind phase (generalized linear mixed model) (efficacy-evaluable analysis set).
Secondary efficacy measures. The percentage of EE subjects with ≥15 or more headache days per 28-day period during the last 28 days in the double-blind phase was not statistically significantly different between the topiramate and placebo treatment groups (3.1% vs 3.5%; p = .728). There was no statistically significant difference between treatment groups in the time to the first reporting of ≥15 headache days per 28-day period for the EE analysis set (data not shown). There were no statistically significant differences between treatment groups observed in the percentage of EE subjects with ≥15 or more headache days (of which at least half were migraine) per 28-day period or the time to the first reporting of ≥15 headache days (of which at least half were migraine) per 28-day period for the EE analysis set (data not shown).
Topiramate treatment resulted in statistically significantly greater mean (±SD) decreases per 28 days from baseline compared with placebo in the number of headache days (6.6 [±3.8] vs. 5.3 [±3.6]; p = .001) (Figure 4); the number of migraine days (6.6 [±3.5] days vs. 5.3 [±3.6]; p = .001) (Figure 4); and the number of days of acute medication use (4.8 [±3.5] vs. 3.8 [±3.7]; p = .001) (Figure 4). Headache and migraine response rates (≥50% and ≥75% reduction/28 days) were higher in the topiramate compared with the placebo treatment group. The distribution of categorical responses (the percentage of subjects achieving a response across the range of responses from 25–100%) was statistically significantly different for the topiramate treatment group than for the placebo group for headache days and migraine days (p ≤ .001). The mean decrease from baseline in the frequency of nausea, photophobia and phonophobia per 28 days was statistically significantly greater in the topiramate than in the placebo treatment group (nausea: 4.4 vs. 3.8 events; p = .002; photophobia: 6.6 vs. 5.2 events; p = .001; phonophobia: 6.5 vs. 5.1; p = .001).
Mean change from baseline in number of migraine days, total headache days and acute headache medication use days per 28-day period (efficacy-evaluable analysis set).
Results of HRQoL end points and measures of migraine disability also showed several statistically significant improvements versus placebo but were not the focus of this report and are reported elsewhere (29).
Safety and tolerability measures
Treatment-emergent adverse events (safety analysis set) reported by ≥5% of subjects in any treatment group
TEAE, treatment-emergent adverse event. *Subjects with multiple occurrences of the same TEAEs were counted only once for that particular preferred term.
Eight subjects (three in the topiramate group and five in the placebo group) reported a total of nine SAEs. In the topiramate treatment group, SAEs included spontaneous abortion (one subject) and bradycardia (one subject); both resolved and were reported as of doubtful relationship to topiramate treatment. One topiramate subject was hospitalized on day 8 of the study during the titration phase and diagnosed with bipolar disorder and suicidal thoughts. The subject had received 25 mg of topiramate for seven days. Treatment was discontinued on day 8. The subject’s suicidal thoughts resolved by day 15. The investigator considered both of these events to be possibly related to the study drug.
In the placebo treatment group, five subjects each reported one SAE (neuropathy [cauda equina syndrome], spontaneous abortion, fractured pelvis secondary to motor vehicle accident, chest pain and worsening of migraine). All SAEs in the placebo treatment group resolved and were reported as not related to treatment except for worsening migraine, which was reported as being of doubtful relationship to treatment. Study medication was discontinued for the placebo-treated subjects with neuropathy and with fractured pelvis. No clinically relevant differences were observed between treatment groups in mean change from baseline to final visit for vital signs (systolic and diastolic blood pressure, pulse rate) or the clinical laboratory test values analyzed in this study. Fourteen subjects in the topiramate group and no subjects in the placebo group had reports of low bicarbonate values that met the study-defined criteria for markedly abnormal laboratory values. This laboratory finding is likely explained by the known pharmacologic effect of topiramate to inhibit the enzyme carbonic anhydrase. None of the low bicarbonate values were reported as an AE. No clinically meaningful patterns of abnormal changes in physical examination were observed in either treatment group.
Discussion
To the best of our knowledge, this is the first large, randomized, double-blind, placebo-controlled study designed to test the hypothesis that migraine preventive treatment would prevent progression to CDH in a population with a high frequency of episodic migraine at baseline. The study was exploratory with respect to the primary outcome measure. Treatment with topiramate 100 mg/day did not result in a statistically significantly lower proportion of subjects reporting ≥15 headache days per 28-day period at month 6 compared with placebo treatment. However, topiramate treatment resulted in statistically significant improvements compared with placebo in several prespecified secondary outcome measures of efficacy including: the 28-day rate of the numbers of headache days and migraine headache days; the proportions of subjects who responded to treatment; the mean number of days subjects used acute headache medications; and the number of episodes of photophobia, phonophobia and nausea. Improvements in several prespecified secondary measures of efficacy observed in this study are important and extend the results of previous studies and pooled analyses of topiramate preventive migraine treatment (30,31).
Treatment with topiramate was safe and generally well-tolerated in this study. The proportion of subjects discontinuing the study because of AEs was generally low in both treatment groups. In this population of subjects with a high frequency of migraine headaches, the reported AEs were in accordance with expectations for the topiramate dose used in this study.
The study failed to demonstrate that topiramate treatment prevented the transformation of episodic migraine to CDH in contradistinction to clinical observations and a post hoc clinical trial analysis (23). There are several possible explanations for this failure. The study enrolled individuals with a high frequency of episodic migraine headache days (9–14 days/28 days) based on evidence that headache frequency is a risk factor for migraine progression (10,21,32–37).
Though incidence of CDH in topiramate treated patients was in line with expectations, the incidence of CDH in the placebo-treated group was much lower than expected (21,22). The lower-than-expected rate of transformation from episodic migraine to CDH over the course of this six-month study attenuated power. Emerging evidence suggests that headache frequency often varies from month to month, above and below the threshold of ≥15 headache days/month (21,32,37–39). Careful exclusion of persons with ≥15 headache days/month may have reduced CDH incidence in comparison with observational studies (10,19–21). In addition, the washout of acute headache medications before randomization and the administration of study medication may have reduced the headache frequency for some subjects after randomization, contributing to a reduced rate of transformation to CDH in placebo-treated subjects.
With these uncertainties and observations, the study’s eligibility criteria might have influenced study outcomes. Modifying eligibility criteria may help identify patients at greater risk for migraine transformation. For example, a history of prior CDH may enrich the sample for persons likely to progress while on placebo. Additionally, requiring subjects to keep headache records for at least two or three consecutive months before randomization may reduce the bias introduced through selective enrollment during unusually bad months (22).
Several other factors also may have contributed to the lack of a topiramate treatment effect in preventing migraine transformation. This study excluded subjects who used acute headache medications in excess of four days/week. By excluding overuse of acute headache medication, we may have diminished rates of progression from episodic headache to chronic forms (10,20,34–36). Similarly, excluding subjects who had failed more than two prior trials of preventive migraine medications may have resulted in the selection of a more placebo-responsive subgroup (39–41). Shifts in the amount of acute medication used per migraine attack as well as specific medications or classes of medications used were not evaluated and also may have influenced the outcome. Finally, population-based and headache clinic–based studies that have analyzed the progression of episodic migraine headache to CDH, and risk factors associated with transformation, have typically used a one-year period of observation (10,21,32,37,38); this study used a six-month period of treatment and observation.
Perhaps preventive migraine treatments, including topiramate, may not modify the transformation of episodic migraine to CDH in some patients. Patients with a high frequency of migraine headaches may experience structural or functional neurologic changes that limit the effectiveness of preventive migraine treatments. This seems unlikely, given the statistically significant effects of topiramate on headache frequency both in high-frequency episodic and chronic migraine patients (17,18).
The modification of migraine transformation is a novel and intriguing target of clinical study that requires a sophisticated design. The information gained from this study may help inform the design of future studies that endeavor to address this objective. For instance, stratification of subjects according to risk factors at the level of subject allocation or intervention, or subsequent to analysis, may provide insight regarding the influence of various risk factors for progression on the eventual results. In the future, genetic studies may facilitate identification of a clinical phenotype at risk for progression by linking associations with specified laboratory, radiographic or demographic biomarkers. Future studies also may benefit from analyses of the influences of comorbidities on the risk for progression and migraine treatment outcomes. Finally, future studies may need to address the probability that the migraine population is heterogeneous and therefore, simply using a diagnosis of migraine as the primary common eligibility requirement may not define a homogenous population. Enrichment of the subject population by enrolling known responders to preventive migraine treatment who meet the previously suggested broadened eligibility criteria, who are then assigned treatment and followed for at least one year to assess prevention of transformation to chronic migraine, may have much more clinical relevance.
Footnotes
Acknowledgements
This study was funded by Ortho-McNeil Janssen Scientific Affairs (ClinicalTrials.gov Identifier: NCT00212810).
The authors thank Li-Joy Wang for statistical programming support and Steve Greenberg for study design, data analysis and medical oversight of the trial. Editorial support was provided by George Rogan (Phase Five Communications, New York, NY, USA), with funding provided by Ortho-McNeil Janssen Scientific Affairs.