Controversies in Migraine Management A Technology Assessment Final Report August 19, 2014 Completed by: Institute for Clinical and Economic Review ©Institute for Clinical and Economic Review, 2014 AUTHORS: Jeffrey A. Tice, MD Associate Professor of Medicine Division of General Internal Medicine Department of Medicine University of California San Francisco Daniel A. Ollendorf, PhD Chief Review Officer, Institute for Clinical and Economic Review Jed Weissberg, MD, FACP Senior Fellow, Institute for Clinical and Economic Review Karen K. Shore, PhD Program Director, Institute for Clinical and Economic Review Steven D. Pearson, MD, MSc President, Institute for Clinical and Economic Review DATE OF PUBLICATION: August 19, 2014 ©Institute for Clinical and Economic Review, 2014 ii About ICER The Institute for Clinical and Economic Review (ICER) is an independent non-profit research organization that evaluates medical evidence and convenes public deliberative bodies to help stakeholders interpret and apply evidence to improve patient outcomes and control costs. ICER receives funding from government grants, non-profit foundations, health plans, provider groups, and health industry manufacturers. Through all its work, ICER seeks to help create a future in which collaborative efforts to move evidence into action provide the foundation for a more effective, efficient, and just health care system. More information about ICER is available at www.icer-review.org About CTAF The California Technology Assessment Forum (CTAF) – a core program of ICER – reviews evidence reports and provides a public venue in which the evidence on the effectiveness and value of health care services can be discussed with the input of all stakeholders. CTAF seeks to help patients, clinicians, insurers, and policymakers interpret and use evidence to improve the quality and value of health care. The CTAF Panel is an independent committee of medical evidence experts from across California, with a mix of practicing clinicians, methodologists, and leaders in patient engagement and advocacy, all of whom meet strict conflict of interest guidelines, who are convened to evaluate evidence and vote on the comparative clinical effectiveness and value of medical interventions. More information about CTAF is available at www.ctaf.org ©Institute for Clinical and Economic Review, 2014 iii Table of Contents Abbreviations…………………………………………………………………………………………………………….……….….v Executive Summary......................................................................................................................ES1 Introduction .................................................................................................................................... 1 1. Background ................................................................................................................................. 2 2. Clinical Guidelines ....................................................................................................................... 9 3. Coverage Policies ...................................................................................................................... 11 4. Previous Systematic Reviews and Technology Assessments.................................................... 16 5. Ongoing Studies ........................................................................................................................ 18 6. Evidence Review (Methods & Results) ..................................................................................... 22 7. Model of Clinical and Economic Outcomes of Treatment Strategies for Migraine.................. 35 8. CTAF Voting Questions and Discussion..................................................................................... 53 References .................................................................................................................................... 62 APPENDIX ...................................................................................................................................... 73 ©Institute for Clinical and Economic Review, 2014 iv Abbreviations used in this report AEs: Adverse events AHRQ: Agency for Healthcare Research and Quality AMPP: American Migraine Prevalence and Prevention BOTOX®: Brand name for onabotulinumtoxinA (Allergan) CDC: Centers for Disease Control and Prevention CI: Confidence interval CMS: Centers for Medicare & Medicaid Services CTAF: California Technology Assessment Forum DARE: Database of Abstracts of Reviews of Effects DHE: Dihydroergotamine ED: Emergency department EMR: Electronic medical record FDA: US Food and Drug Administration HIT-6: Headache Impact Test HR: Hazard ratio IHS: International Headache Society MIDAS: Migraine Disability Assessment MOH: Medication overuse headache NR: Not reported NS: Not significant NSAID: Non-steroidal anti-inflammatory drug OR: Odds ratio PBO: Placebo QALY: Quality-adjusted life year RCT: Randomized controlled trial RR: Relative risk TENS: Transcutaneous electrical nerve stimulation TMS: Transcranial magnetic stimulation UK: United Kingdom US: United States VAS: Visual analog scale ©Institute for Clinical and Economic Review, 2014 v Executive Summary Abstract On July 11, 2014, the California Technology Assessment Forum (CTAF) held a meeting in Los Angeles on “Controversies in Migraine Management.” The CTAF Panel discussed the clinical effectiveness and reviewed economic analyses of four migraine treatments. Invited policy roundtable experts, including clinical leaders, health plan medical directors, and a patient advocate provided context and interpretation of the review findings. Two devices were considered. First, for the treatment of acute migraine headache accompanied by aura, one well-designed, moderate size study of a single-pulse transcranial magnetic stimulation device (SpringTMS™ by eNeura) showed superior pain relief compared with a sham device, but no benefit was found in several other common outcome measures. Economic modeling comparing the device with a commonly-used generic triptan found a high relative cost for its potential benefit. Second, only one small trial has been reported of a TENS device (Cefaly) for the prevention of frequent migraine headaches. This trial, further limited by concerns about unblinding and incomplete reporting of adverse effects, showed improvement in some commonly measured headache outcomes. At current pricing and with the best estimate of Cefaly’s clinical effectiveness compared with a commonly used generic, oral medication, modeling suggested lower overall benefit and higher cost. For both devices, the CTAF panel voted that the evidence is inadequate to demonstrate that they are as effective as other currently available care options. The CTAF Panel next considered the evidence on BOTOX, the only FDA-approved treatment for the prevention of chronic migraine. Two large, well-designed studies demonstrated improvements in a variety of headache outcomes versus sham injection, but the CTAF Panel voted that inadequate evidence exists to evaluate whether BOTOX is equivalent or superior to other lower-cost oral agents frequently used in this patient population. Subsequent discussion suggested that coverage policies requiring attempts with other treatments before using BOTOX are reasonable and reflect current practice. The CTAF Panel urged that research be performed to allow identification of patients who are better candidates for first-line treatment with BOTOX. Lastly, CTAF examined care options for migraine patients in emergency departments (EDs). These patients receive parenteral opioids over 50% of the time, even though the CTAF Panel confirmed that strong evidence shows that opioids offer no short term benefits compared to other treatment options and raise the long-term risk of exacerbating migraines and of contributing to opioid dependence. The downstream economic impact of such widespread opioid use is also substantial. The CTAF Panel recommended that 1) relevant specialty societies come together to educate patients and providers about the significant harms of opioid use in the ED, and 2) EDs and clinicians develop order sets and use data analysis and feedback to reduce opioid prescribing. ©Institute for Clinical and Economic Review, 2014 Page ES1 Background Headaches approach death and taxes as a near universal human experience. Migraine is a common and severe form of headache that causes throbbing or pulsating pain, usually on only one side of the head. These headaches are also often associated with nausea, vomiting, and extreme sensitivity to light and sound. They can start suddenly, worsen quickly, and can last between four and 72 hours. Migraine headaches, which are often misdiagnosed as other types of headaches, can interfere with sleep, work, and other everyday activities. They may occur as often as several times per week or as rarely as once or twice a year. Each year, one in seven people suffering from migraines goes to the emergency department (ED) seeking treatment for a severe migraine. Migraines affect about 12% of the population, and they are three times more common in women than in men. The exact cause of migraines is not known for certain, but there are many known triggers. These include stress, hormones in women, hunger (missed or delayed meals), too little or too much sleep, lack of regular exercise, certain foods and food additives, and odors (perfumes, cigarette smoke). Clinical understanding of the basis for migraine headaches is still at an early stage. Initially thought to be caused by dilation and constriction of blood vessels in the head, experts now believe that electrical activity in the surface layer of the brain (characterized by waves of excitation and inhibition), chronic stimulation of pain receptors related to the facial nerve, and aberrant neurotransmitter activity all play primary causative roles. Epidemiological studies have demonstrated a role for genetics, and ongoing animal and basic science models are studying this and other mechanisms. Although migraines cannot be cured, there are many options to prevent migraines and to treat symptoms once a migraine headache starts. Treatments include caffeine, over-the-counter pain medications, acupuncture, relaxation techniques, and, for patients with more severe and/or numerous headaches, prescription medications. Several classes of drugs are approved for the relief of the acute symptoms of migraine including analgesics, triptans, anti-emetic agents, and ergotamine derivatives. Other drugs are commonly used for the prevention of frequent migraines, from the classes of beta blockers, calcium channel blockers, anti-depressants, angiotensin- converting enzyme inhibitors, angiotensin receptor blockers, and anti-convulsant drugs. One injectable agent, onabotulinumtoxinA (BOTOX), has been FDA-approved for the prevention of chronic migraines and is used on an every three month basis. In this report for the California Technology Assessment Forum (CTAF), we examined two newly approved devices for the management of migraines, as well as two more established drug therapies. For the prevention of migraines, we reviewed the evidence for a transcutaneous ©Institute for Clinical and Economic Review, 2014 Page ES2 electrical nerve stimulation (TENS) device called Cefaly (available by prescription and used at home) and the evidence for BOTOX injections. For the treatment of migraine headaches with aura, we reviewed studies on single pulse transcranial magnetic stimulation (sTMS) and a newly approved device called SpringTMS (also available by prescription and used at home). For the acute treatment of migraines that have failed self-management and home medication treatment, we looked at the emergency department setting and critically reviewed evidence of comparative effectiveness as well as short- and longer- term harms of opioid medications, which are commonly used in this setting. Evidence Review Methods and Process As described fully in the evidence review (Section 6 of this report), we reviewed published meta- analyses and conducted a full search and review of the relevant literature with an emphasis on the highest quality trials. There are several challenges in the design, implementation, and interpretation of intervention trials for migraine headache management. The most important of these challenges is the difficulty of achieving adequate blinding of patients as to whether they are receiving an active agent or placebo, an issue particular to trials of devices or injections (with placebo represented by a sham device or injection). In many trials, patients in both the active treatment and placebo arms reported significant improvements in pain, number of headache days, and use of additional medications. However, the difference between the outcomes of the intervention and control study arms was typically much smaller than the improvement from baseline observed within the placebo arm itself. The evidence review provides more detail on the difficulties in interpreting data from these trials. In addition, trials done in the past often were underpowered, were not designed per the current recommendations of the International Headache Society (IHS), and did not directly compare commonly-used treatment approaches. During the CTAF meeting, the results of the evidence review as well as our simulation models of the costs and cost-effectiveness of these management options were presented. Clinical experts and a patient advocate provided context and deeper understanding before the CTAF Panel was asked to weigh the evidence of efficacy and harms to determine net benefit and then to pair net benefit with the economic analyses to arrive at judgments of overall value. ©Institute for Clinical and Economic Review, 2014 Page ES3 The key questions addressed in the clinical assessment included the following: 1. Among patients suffering from acute migraines with aura, what is the comparative effectiveness of acute treatment with the sTMS device (SpringTMS™ by eNeura) versus other acute therapies? 2. Among patients suffering from episodic migraine headaches, what is the comparative effectiveness of preventive treatment with the TENS device (Cefaly) versus other preventive therapies? 3. Among patients suffering from chronic migraine headaches, what is the comparative effectiveness of preventive treatment with BOTOX injections versus other preventive therapies? 4. Among patients presenting to the emergency department with acute migraine headaches, what is the comparative effectiveness of opioid analgesics versus other acute therapies? The economic analysis examined the following issues: • Potential costs and cost-effectiveness of Cefaly and SpringTMS devices versus relevant pharmacologic comparators • Potential costs and cost-effectiveness of BOTOX injections at two levels of baseline migraine headache frequency • Economic burden of opioids for treatment of migraines in California and potential cost savings from reductions in ED use of opioids Single-Pulse Transcranial Magnetic Stimulation for Treatment of Acute Migraines with Aura The portable, handheld SpringTMS device delivers 0.9 Tesla (a measure of magnetic field force) to the back of the head seeking to counter the “cortical inhibitory wave” that occurs at the early stages of a migraine headache. We reviewed the results of the only available comparative study of the single-pulse device, a high-quality study46 with data from 164 patients performed with Cerena, a predecessor of the SpringTMS device approved by the FDA in May 2014. There are no published studies with the SpringTMS device itself. The study of the Cerena device used a sham device arm and evaluated several of the established migraine headache outcomes. Pain relief was superior with the active device, although there were no differences for several other outcomes (e.g., response rate at 2 hours, disability rated with a standard and validated scale, use of “rescue” medications, or ©Institute for Clinical and Economic Review, 2014 Page ES4 the specific symptoms of sensitivity to light and sound). Major study results are shown in Table ES1 below. Table ES1. Primary Outcomes in the Randomized Trials of Single-Pulse TMS Study Group N Pain- Mild or Pain- Pain- Use of Use of Change Adverse free 2 no pain 2 free 24 free 48 rescue rescue MIDAS Events hours hours hours hours medication medication 0-2 hours 0-48 hours Lipton sTMS 82 39% 72% 29% 27% 18% 48% -4.6 14% 53 2010 Sham 82 22% 67% 16% 13% 16% 46% -4.7 9% USA 18 centers Our cost-effectiveness model compared use of SpringTMS with sumatriptan (a commonly used generic form of the triptan class of medications). Treatment response was modeled based on the proportion of patients reported to be pain-free 24 hours after treatment. Other costs of migraine management (e.g., doctor’s visits, additional medications) were assumed to be reduced by 25% among patients with a treatment response. Based on data from the Cerena trial as well as a systematic review of multiple sumatriptan trials,31 290 and 188 patients per 1,000 treated would be expected to respond to SpringTMS and sumatriptan, respectively. This greater response would lead to a reduction of $140,000 in other costs of migraine management. However, sumatriptan is a generic medication (estimated to cost $112 per patient annually), and the price of the SpringTMS unit has not yet been released in the US. Correspondence from the manufacturer suggests the unit will be leased at $250 per month. Based on the data from the single high quality published trial, use of the SpringTMS device would cost approximately $27,000 per additional treatment response and would be cost-saving relative to sumatriptan only if the monthly lease price was approximately $21. Transcutaneous Electrical Nerve Stimulation (TENS) – Cefaly for Prevention of Migraines There is very little controlled data for the Cefaly device. One 67-person trial54 demonstrated that some headache outcomes statistically improved with the use of the active device, but issues with potential unblinding, apparent lack of reporting of adverse events, and the small number of participants make the evidence on this device promising but inconclusive. Study results are shown in Table ES2 on the next page. ©Institute for Clinical and Economic Review, 2014 Page ES5 Table ES2. Primary Outcomes in the Randomized Trials of the Supraorbital Transcutaneous Stimulator (Cefaly) Study Group N Headache Headache Change Change Reduction Change in Patient AEs days per days per in in by at least use of satisfaction month, month, migraine migraine 50% in rescue (moderate run-in month 3 days at 3 attacks migraine medication or very) months at 3 days months Schoenen Cefaly 34 6.9 4.9 -2.1 0.1 38% -4.2 71% 0 54 2013 Sham 33 6.5 6.2 -0.3 0.5 12% 0 39% 0 Belgium 5 centers In our cost-effectiveness model, we compared the results derived from this single trial with data on metoprolol, a commonly used medication for prevention of episodic migraines. Treatment response in this instance was defined as the proportion of patients with a 50% or greater reduction in monthly headache frequency, a primary measure in both the Cefaly trial as well as a recent meta- analysis of metoprolol and other preventive agents for frequent migraine.37 Based on these data, totals of 382 and 395 patients per 1,000 treated would respond to Cefaly and metoprolol respectively. In addition, the purchase of Cefaly and accompanying electrode kits is estimated to be nearly 10 times the cost of annual treatment with generic metoprolol ($449 vs. $49 respectively). Based on these findings, a cost-effectiveness ratio for Cefaly could not be generated, as it is both less effective and more expensive than metoprolol. Given the uncertainty surrounding the Cefaly trial results, we also examined results at different levels of treatment response and cost for the device. For example, if Cefaly were 5% more effective than metoprolol, the cost per additional treatment responder would be approximately $99,000. Cefaly would only become cost-saving relative to metoprolol if its effectiveness nearly doubled to approximately 730 per 1,000 treated, or at an assumed 5% improvement in effectiveness relative to metoprolol along with an 85% reduction in price (from $449 to $76). OnabotulinumtoxinA (BOTOX) Injections for the Prevention of Chronic Migraines Since the original chance observation of potential impact on the frequency of chronic migraines, many studies have been performed to test the initial observation and to develop a standardized approach to BOTOX injections. Currently, the process is well defined in terms of the multiple (31) sites of injection in the head and neck, as well as the dose and additional injections in sites of particular muscle tenderness. Current therapy recommendations call for injections every 12 weeks, although the approach has only been studied and reported in cycles of one to two injections (24 weeks) in blinded studies with subsequent open label treatment. Ongoing observational registries and studies may provide information on longer-term efficacy and tolerability. ©Institute for Clinical and Economic Review, 2014 Page ES6 A recent meta-analysis80 concluded that BOTOX injections were ineffective for prevention of episodic migraines, so our analysis is restricted to use in chronic migraines, defined as severe migraine headache for 15 or more days a month (see evidence review for IHS criteria). This high- quality systematic review and meta-analysis reported on seven trials of BOTOX compared to a saline injection placebo arm or an active medication comparator. There was a small clinical improvement in headache outcomes with BOTOX compared with placebo injections; the number of monthly migraine headache days declined by 2.3 more headache days per month with BOTOX, and more patients had at least a 50% reduction in headache frequency. A few trials comparing BOTOX to commonly utilized preventive medications showed similar levels of effectiveness, although small numbers of patients prevented statistical conclusions regarding potential benefit over other medications. Adverse effects associated with BOTOX injections included drooping eyelids, neck pain, muscle weakness, neck stiffness, paresthesias, and skin tightness, although BOTOX recipients were generally not more likely to drop out of the studies than patients receiving placebo. The two largest and most sophisticated BOTOX-placebo studies in the prevention of chronic migraine are the PREEMPT 1 and 2 trials, and these were included in the meta-analysis. These trials were supported by the manufacturer of BOTOX, were conducted nearly simultaneously, and used similar outcome measures. PREEMPT 1 was a multicenter trial in the US, and PREEMPT 2 was a multicenter trial with US and global centers. They are discussed in detail in the evidence review (section 6.3), and primary outcomes are described in table ES3 on the next page. Both studies demonstrated superior efficacy of BOTOX compared with placebo, though the magnitude of improvement was small compared with the notable degree of improvement seen in the placebo treatment arm alone. A number of outcomes such as number of headache days, headache hours per month, and several quality of life measures for headache sufferers also statistically improved. Adverse events were more common with BOTOX and led to 3.5% of active arm participants withdrawing from the study compared with 1.4% of participants in the placebo arm. ©Institute for Clinical and Economic Review, 2014 Page ES7 Table ES3. Primary Outcomes in the PREEMPT Trials of Botulinum Toxin for the Prevention of Migraine Headaches Study Group* N Headache days Reduction by at least Use of rescue per month 50% in migraine days medication doses Chronic Migraine – Placebo control 74 Aurora 2010 Botulinum 341 7.1 Not reported -10.3 North America 56 Centers (PREEMPT 1) Placebo 338 8.1 -10.4 75 Diener 2010 Botulinum 347 11.2 Not reported -9.9 Europe, North America 66 Placebo 358 13.4 (though in pooled -8.4 Centers (PREEMPT 2) analysis of these two trials, 47% (active) vs 35% (placebo) was reported) Our cost-effectiveness model examined BOTOX compared with placebo injections and no treatment. Effectiveness was estimated based on reductions in the number of headache days from an assumed level of 20 per month (based on the reported frequency in the PREEMPT 1 and 2 trials). Reductions in the frequency of headache were applied to a daily cost of chronic migraine management (~$40) to estimate potential cost savings. We estimated that BOTOX would reduce the frequency of headache by nine days per month relative to no treatment, with the corresponding cost per headache day averted estimated to be $4 (at an assumed baseline frequency of 20 headache days per month). At this headache frequency, a reduction of one additional headache day relative to no therapy (i.e., from 9 fewer headache days per month to 10 fewer headache days) would represent the threshold at which BOTOX would be cost-saving. Complicating the interpretation of the trials is the strong placebo effect. BOTOX would need to reduce headache frequency from 20 to 4 headache days per month (i.e., an additional 7 days above the primary estimate) to achieve cost neutrality vs. placebo. Use of Opioid Medications to Treat Migraine Headaches in the Emergency Department Patients whose migraine headaches are not adequately treated at home frequently appear in urgent care or emergency department settings. National data suggest that approximately half of all patients with migraines seen in the ED setting are treated with opioid medication prior to discharge. A thorough systematic review of medication options to treat migraines in the ED was commissioned by the Agency for Healthcare Research and Quality (AHRQ) and published in 2008.34 This review found that patient outcomes achieved with the opioid meperidine were no better than ketorolac (a non-steroidal anti-inflammatory drug, or NSAID) and inferior to dihydroergotamine (one of a group ©Institute for Clinical and Economic Review, 2014 Page ES8 of drugs called ergot alkaloids that works by narrowing the blood vessels around the brain and affecting blood flow patterns) and anti-emetics (drugs used to reduce vomiting and nausea). Despite societal attention to the overuse of opioids and downstream consequences of prescription opioid use, there is a dearth of high quality studies and a complete lack of studies on the more commonly used opioid agents. The AHRQ systematic review found that the most effective parenteral combinations did not include opioids and that other second line agents provided analgesic effects equivalent to those of opioids. Furthermore, opioids are associated with a variety of adverse effects, including transformation of episodic migraines to a more chronic and severe disease state. Clinical guidance from the American Academy of Neurology, the IHS, and primary care societies discourage the use of opioids for migraines, yet opioids are still commonly administered and prescribed. Our population-based cost-effectiveness model looked broadly at the use of opioids for treatment of migraines in all settings, the economic burden of such use, and consequences of changes in utilization in the ED. Since migraines are so common, large numbers of people are potentially prescribed opioids. In California, about 550,000 people with migraines might be expected to be given opioids for migraine. Over one year, our model results suggest that such use will result in over 20,000 patients developing chronic migraines because of opioid overuse, with 3,000 patients becoming addicted to opioids. Including patients already dependent on opioids, the total estimated economic burden of opioid use among migraineurs in California is $2.8 billion, based on the costs of additional health care services as well as lost productivity. Given the importance of the ED setting in the management of treatment-resistant migraines, our model also examined potential savings associated with reduced use of opioids in the ED. In this case, we used estimates of health care costs among patients receiving and not receiving opioids in the ED to estimate the incremental direct medical care costs associated with such use. These data suggest that annual costs of medical care are 2.5-fold higher among opioid users vs. nonusers (approximately $6,500 vs. $2,600, respectively), principally due to increased use of inpatient and ED services.117 In California, if the percentage of patients with migraines who are treated with opioids in the ED dropped to 25% from the current national average of 53%, and the per-patient medical costs reverted to the opioid nonuser level of ~$2,600, annual health care costs could potentially be reduced by $126 million (see Figure ES1 on the next page). ©Institute for Clinical and Economic Review, 2014 Page ES9 Figure ES1. Annual Health Care Costs at Multiple Levels of Opioid Use in the ED. Total costs decline with each incremental decrease in the percent of patients who receive opioids in the ED. Baseline (53% Opioid Use) 25% Opioid Use 10% Opioid Use 5% Opioid Use $- $50 $100 $150 $200 $250 $300 $350 $400 Millions CTAF Votes on Comparative Clinical Effectiveness and Value Following review of the clinical evidence and economic models, the CTAF panel was asked to vote on five questions from the perspective of Medi-Cal or a provider organization that must make resource decisions within a relatively fixed budget for care. To aid in the classification of “low”, “reasonable/comparable”, or “high” value votes, the matrix shown in Figure ES2 on the next page was used. The voting questions and overall results are presented below the figure. ©Institute for Clinical and Economic Review, 2014 Page ES10 Figure ES2. Evidence Categories for Ratings of Low, Reasonable/Comparable, and High Value Low Value Reasonable/ Comparable High Value Value* Worse outcomes; Worse outcomes; Comparable outcomes; Higher or equivalent cost Lower cost Lower cost Comparable outcomes; Comparable outcomes; Promising but inconclusive Higher cost Comparable cost evidence of better outcomes; Lower cost Promising but inconclusive Promising but inconclusive Better outcomes; evidence of better outcomes; evidence of better outcomes; Lower or comparable cost Higher cost Comparable cost Better outcomes; Better outcomes; Better outcomes; Too high a cost Reasonable higher cost Slightly higher cost * For comparisons of one drug or a set of drugs to another drug or set of drugs, the term “comparable” is used in the value assessment; for comparisons of one drug or a set of drugs to no treatment, the term “reasonable” is used in the value assessment. 1. For the acute treatment of migraine with aura, is the evidence adequate to demonstrate that the net health benefits of transcranial magnetic stimulation (SpringTMS) are equivalent to or better than those of other standard acute treatment medications? CTAF Panel Vote: 2 yes 9 no 2. For the prevention of episodic migraine, is the evidence adequate to demonstrate that the net health benefits of the Cefaly device are equivalent to or better than those of usual care with preventive medications? CTAF Panel Vote: 1 yes 10 no 3. For patients who have inadequate relief with other preventive therapies for chronic migraine, is the evidence adequate to demonstrate that the net health benefits of BOTOX injections used on an every 12-week schedule are better than no treatment? CTAF Panel Vote: 11 yes ©Institute for Clinical and Economic Review, 2014 Page ES11 3a. If yes, what is the comparative value of BOTOX injections vs. no treatment? CTAF Panel Vote: 7 low 3 reasonable 1 high Low value votes: “Promising but inconclusive evidence of better outcomes and higher cost”: 5 “Better outcomes at too high a cost”: 2 Reasonable value votes: “Better outcomes at reasonable higher cost”: 3 High value votes: “Better outcomes at slightly higher cost”: 1 4. For patients who are considering multiple therapeutic options for chronic migraine, is the evidence adequate to demonstrate that the net health benefits of BOTOX injections used on an every 12-week schedule are equivalent to or better than those of other preventive therapies? CTAF Panel Vote: 1 yes 10 no 5. For the acute treatment of migraine in the emergency department, is the evidence adequate to demonstrate that the net health benefits of parenteral opioids as first-line therapy are inferior to those of non-opioid alternatives? CTAF Panel Vote: 9 yes 2 no Policy Roundtable Discussion and Key Policy Recommendations Following its deliberation on the evidence and subsequent voting, the CTAF Panel engaged in a moderated discussion with a policy roundtable composed of clinical experts, a patient advocate, and payer representatives. A list of the participants on the policy roundtable is shown below. • Sylvia Carlisle, MD, MBA, Managing Medical Director, Anthem Blue Cross • Robert Cowan, MD, FAAN, Professor of Neurology and Director of the Headache Program, Stanford University • Lynne McCullough, MD, Medical Director, Ronald Reagan UCLA Medical Center Emergency Department, and Clinical Professor of Medicine and Emergency Medicine, UCLA • Sonja Potrebic, MD, PhD, Headache Specialist, Kaiser Permanente Los Angeles Medical Center • Ellen Schnakenberg, patient educator and advocate • Sam Torbati, MD, Co-chair and Medical Director, Department of Emergency Medicine, Cedars-Sinai Medical Center • John Yao, MD, MBA, MPH, Staff Vice President of Medical Policy Development, WellPoint ©Institute for Clinical and Economic Review, 2014 Page ES12 The policy roundtable discussion explored the implications of CTAF Panel votes for clinical practice and medical policy, considered real life issues critical for developing best practice recommendations in this area, and identified potential avenues for applying the evidence to improve patient care within a value context. The main recommendations from the discussion are summarized below, and the rationale for these recommendations is presented in the body of the report beginning on page 56. The policy roundtable discussion with the CTAF Panel reflected multiple perspectives and opinions, and therefore none of the recommendations below should be taken as a consensus view held by all participants. 1) The CTAF Panel voted that the evidence is currently inadequate to demonstrate the effectiveness of the Cefaly and SpringTMS devices to prevent and treat migraine, respectively. 2) For patients who have inadequate relief with other preventive therapies for chronic migraine, the CTAF Panel voted unanimously that the evidence demonstrates that the net health benefits of BOTOX injections are better than no treatment. 3) Given that patients with migraine can respond so differently to a specific drug and that many factors can affect a clinician’s drug choice for an individual patient, insurers may wish to consider adding more choices within a particular drug class such as triptans and to allow clinicians to try and retry a variety of preventive and abortive agents. 4) In contrast to current practice, opioids should very rarely be used to treat migraine in the ED. Multiple initiatives are needed to reduce the use of opioids for migraine in the ED. a. Specialty societies should work collaboratively to make a strong statement about the harms of using opioids for migraine pain relief in the ED and ensure that clinicians are supported in efforts to reduce opioid use. b. Data on the use of opioids to treat migraines in the ED should be analyzed and feedback provided to emergency departments, in combination with education about evidence- based practice. c. Electronic systems with order sets that encourage ED physicians to select non-opioid alternatives to treat migraine symptoms should be implemented. d. Migraine patients and their clinicians should develop a written treatment plan for rescue medications that can be used by patients at home and also made readily available to guide care in urgent care and emergency care settings. e. Purchasers, insurers, and other policymakers should consider identifying reduction of opioid use in the ED as a specific target for quality improvement initiatives in collaboration with clinicians and patient groups. ©Institute for Clinical and Economic Review, 2014 Page ES13 5) Patients and providers should partner to better understand migraine triggers and potential therapies to ensure that patients with migraines get the right treatment in the right place as quickly as possible. 6) Guidelines that encourage the use of opioids to treat pain should be revised to reflect current evidence. 7) Support for basic science research is needed to improve clinical understanding of the biological underpinnings of migraine, which will ultimately improve clinicians’ ability to effectively diagnose patients and identify appropriate treatments. 8) There is a significant need for higher-quality, larger research studies to show the effectiveness of various migraine therapies, and to compare these agents head-to-head in a comparative effectiveness context. As a follow-up to the public meeting and as a complement to this report, an action guide for each of three groups (patients, clinicians, and payers/policymakers) will be developed and distributed to interested parties and available on the CTAF website. In addition, ICER on behalf of CTAF will contact specialty societies (e.g., International Headache Society, American Academy of Neurology, American College of Emergency Physicians) to seek their leadership in making a strong statement about the harms of using opioids for relief of migraine pain in the ED. ©Institute for Clinical and Economic Review, 2014 Page ES14 Introduction This assessment for the California Technology Assessment Forum (CTAF) evaluates the evidence on the comparative clinical effectiveness and value of four new or controversial therapies for migraine headaches. These include two medical devices recently approved by the Food and Drug Administration (FDA): a single-pulse transcranial magnetic stimulation (sTMS) device (SpringTMS by eNeura) for treating acute pain and a transcutaneous electrical nerve stimulator (TENS) device (Cefaly) for the prevention of migraine attacks. Botulinum toxin (BOTOX) is increasingly being used to treat patients suffering from at least 15 days of headache pain each month. Finally, opioids are commonly used to treat headache pain in the emergency department (ED) despite statements from professional neurology and headache societies that opioids should rarely or never be used to treat migraines. The key questions addressed in this assessment include the following: 1. Among patients suffering from acute migraines with aura, what is the comparative effectiveness of acute treatment with the sTMS device (SpringTMS™ by eNeura) versus other acute therapies? 2. Among patients suffering from episodic migraine headaches, what is the comparative effectiveness of preventive treatment with the TENS device (Cefaly) versus other preventive therapies? 3. Among patients suffering from chronic migraine headaches, what is the comparative effectiveness of preventive treatment with botulinum toxin injections versus other preventive therapies? 4. Among patients presenting to the emergency department with acute migraine headaches, what is the comparative effectiveness of opioid analgesics versus other acute therapies? The appearance of sham controls in the studies of the two devices and BOTOX is noteworthy. Investigators have been aware of the powerful effect of placebo interventions on pain for more than four decades.1 As the evidence review will demonstrate, the magnitude of the placebo effect on pain is often greater than the additional improvement gained through the use of the active agent under study. It will be essential for investigators to ensure that participants in future studies remain blinded to the intervention. Given the large number of therapies currently used for both acute treatment and prevention of migraine headaches, comparisons with placebo or sham therapy are only the first step in determining the comparative effectiveness of new therapies. Direct comparisons with active controls are needed to fully evaluate the comparative effectiveness of new therapies with the standard of care. ©Institute for Clinical and Economic Review, 2014 Page 1 1. Background 1.1 Migraine Headaches Migraine headaches are typically described as episodic, severe, unilateral headaches associated with nausea and light and/or sound sensitivity. They are common, affecting approximately 16% of women and 6% of men annually in the US.2-4 In 2010, they were the third most common disorder and seventh leading cause of disability worldwide.5 Although migraines often first appear in the second decade of life, the peak prevalence of migraines is at approximately 40 years of age with the prevalence declining after that.2-4 There appears to be a genetic component to migraines. If one parent suffers from migraines, there is a 40% chance that their children will have migraines; if both parents suffer from migraines, there is a 75% chance that their children will have migraines.6-8 The pathophysiology is not fully understood. It used to be thought that migraines were caused by the dilation of blood vessels and the aura from vasoconstriction. That is no longer thought to be true.9,10 The aura appears to be due to cortical spreading depression, a wave of excitation followed by a wave of inhibition that spreads across the cerebral cortex. This is thought to cause the aura, activate the trigeminal nerve, and alter the blood-brain barrier.11-16 Activated trigeminal nerve pain receptors in the meninges are thought to be responsible for the pain associated with the headache.11,17 The aura represents focal neurologic symptoms that typically develop over five minutes and last up to 60 minutes.18 The most common aura is visual with flashing bright lights or an enlarging bright spot with jagged edges. Non-visual auras include numbness and tingling spreading across one arm or the side of the face, trouble speaking, or cognitive difficulties.18 Pain occurs within an hour of the aura, but individuals can experience the aura alone. Approximately 25% to 30% of patients with migraines experience an aura. The five symptoms characteristic of migraines are (1) a pounding or throbbing headache, (2) one day duration, though the range is four to 72 hours, (3) unilateral location, (4) nausea or vomiting, and (5) disabling intensity, with alteration in usual activities.19 ©Institute for Clinical and Economic Review, 2014 Page 2 1.2 Definitions International Headache Society 3rd Edition Criteria for Migraine Diagnosis19 Without aura A. At least five attacks lifetime meeting criteria B-D B. Headache lasting 4-72 hours untreated C. Headache with at least two of the following • Unilateral location • Pulsating quality • Moderate or severe pain intensity • Aggravated by or causing avoidance of routine physical activity (e.g., walking or climbing stairs) D. Occurrence of at least one of the following symptoms: • Nausea and/or vomiting • Photophobia and phonophobia E. Not better accounted for by another headache diagnosis With typical aura A. At least two attacks lifetime meeting criteria B-C B. Aura consisting of visual, sensory, and/or speech/language symptoms, each fully reversible, but no motor weakness or brainstem symptoms C. At least two of the following characteristics: • At least one aura symptom spreads gradually over five or more minutes and/or two or more symptoms occur in succession • Each individual aura symptom lasts 5-60 minutes • At least one aura symptom is unilateral • The aura is accompanied by or followed within 60 minutes by headache D. Not better accounted for by another headache diagnosis and transient ischemic attacks have been excluded Episodic migraine: Headaches occurring less than 15 times a month Chronic migraine: A headache occurring 15 or more days a month for three months with migraine features on at least eight days per month. One headache episode can last more than one day. Chronic daily headache: Headaches that occur for > 4 hours on ≥ 15 days per month. ©Institute for Clinical and Economic Review, 2014 Page 3 Medication overuse headache (MOH): A headache present on ≥ 15 days per month that has developed or worsened following the regular use of symptomatic headache medications. 1.3 Treatment of Migraine Headaches General principles An important first step in the management of migraines is the identification and avoidance of triggers.20 Common triggers include stress, hormones in women, hunger (missed or delayed meals), too little or too much sleep, lack of regular exercise, dietary elements (wine, caffeine, MSG, artificial sweeteners, nitrates), and odors (perfumes, cigarette smoke).20-25 A headache diary can help to identify potential triggers. To evaluate potential triggers, they should be avoided for at least four weeks before re-introducing them one at a time.20 The response rate of acute medical therapy for migraines is greatest if therapy is delivered at the full dose required for therapy as early as possible. A single large dose is more effective than repetitive small doses. Ideally, it should be taken at the onset of the aura, prior to the development of pain. A non-oral route of administration should be used for patients with early nausea and vomiting. Once treatment has been initiated, patients should continue to keep a headache diary recording the headache severity, disability, response to therapy, and potential triggers. Abortive therapy: The acute treatment of migraine headaches Simple analgesics are the first-line therapy for mild to moderate migraine headaches. These include acetaminophen, aspirin, and non-steroidal anti-inflammatory drugs (NSAIDs) such as ibuprofen and naproxen.26-29 A Cochrane systematic review and meta-analysis found that more than 50% of patients with migraines responded to ibuprofen.29 Simple analgesics are preferred because of their relative safety, ready availability, and low cost. Migraine-specific agents are used in patients who fail simple analgesics or have more severe migraine symptoms. These include the triptans (almotriptan, eletriptan, frovatriptan, naratriptan, rizatriptan, sumatriptan, zolmitriptan) and the ergots (ergotamine, dihydroergotamine). The triptans are first-line therapy as they are more effective than the ergots and cause less nausea.30 Intranasal and subcutaneous preparations of the triptans are available for use in patients who are unable to tolerate oral therapy because of early nausea and vomiting. The typical outcomes for sumatriptan in randomized, placebo-controlled trials are summarized in Table 1 on the next page.31 ©Institute for Clinical and Economic Review, 2014 Page 4 Sumatriptan is the most studied migraine-specific agent, but some of the newer triptans appear to have even larger response rates.31 Table 1: Therapeutic Response to Oral Sumatriptan in Randomized Trials Triptans Placebo Headache response at 2 hours 60% 30% Pain-free at 2 hours 30% 10% Sustained pain-free at 24 hours 20% - Patients presenting to the ED usually have severe headaches and have already tried their usual abortive therapy.32 A number of parenteral therapies are effective in this setting including sumatriptan, dihydroergotamine, ketorolac, antiemetics (chlorpromazine, droperidol, metoclopramide, prochlorperazine, promethazine), dexamethasone, and opioids (meperidine, tramadol, nalbuphine).32-35 Ketorolac is the preferred first-line agent because of its minimal side effects and at least eight randomized trials demonstrating equivalence or superiority to parenteral sumatriptan and the antiemetics. The evidence about opioids will be discussed in detail in the evidence review. Antiemetic therapy, such as metoclopramide, can facilitate the use of oral agents to treat patients who suffer from nausea and vomiting with their headaches. Frequent use of acute therapies for migraines should be discouraged. The use of acute therapy more frequently than 10 days a month is associated with the development of medication overuse headaches and chronic daily headaches.36 Opioids and barbiturates are thought to be the highest risk medications, although frequent use of NSAIDS and triptans can also lead to chronic migraines and medication overuse headaches.36 Patients with frequent headaches should be treated with preventive therapy. Preventive treatment to reduce the frequency of migraine headaches Guidelines recommend that physicians discuss preventive therapy with patients who suffer from two or more headaches per month that interfere with daily activities. The goal is to reduce the frequency, intensity, and length of headache attacks. Effective preventive therapy reduces headache frequency by 35% to 55%, but it usually does not completely prevent migraines. A recent systematic review reported that in randomized trials, the FDA-approved drugs reduced headache frequency by at least 50% in 40-50% of participants with episodic migraines compared with 23-25% of participants randomized to placebo.37 Many classes of medications are effective at reducing the frequency and intensity of migraine headaches. These include beta-blockers (propranolol, metoprolol), anti-convulsants (valproate, ©Institute for Clinical and Economic Review, 2014 Page 5 topiramate), anti-depressants (amitriptyline, venlafaxine), angiotensin-converting enzyme inhibitors (lisinopril), calcium channel blockers (nicardipine), and angiotensin receptor blockers (candesartan). The medication choice is usually based on an indication for a particular drug class because of concomitant conditions and tolerability. There are fewer data on preventive therapies for chronic migraines. Botulinum toxin will be discussed below. Three randomized trials found that topiramate was significantly more effective than placebo in reducing headache days in patients with chronic migraines.124-127 There is one published placebo-controlled randomized trial supporting the use of each of the following drugs for chronic migraine: sodium valproate, gabapentin, and tizanidine.128-130 1.4 New or Controversial Therapies for Migraine Treatment Single-pulse Transcranial Magnetic Stimulation (sTMS): Cerena/SpringTMS TMS is a non-invasive therapy that applies a magnetic field to the scalp and underlying cortex, which induces electric current. It has been studied with single pulses, paired pulses, and repeated trains of pulses. A reduction in migraines was observed in patients with resistant depression treated with repeated train TMS. In addition, animal models of cortical spreading depression suggested that TMS could abort the spreading wave.38 In December 2013, the FDA approved the Cerena Transcranial Magnetic Stimulator (eNeura) for the acute treatment of pain in patients who have migraines with aura. The FDA considered it a class II device (one with moderate risk to health). It is a portable, handheld device that delivers a brief pulse of magnetic energy at 0.9 Tesla to the back of the head in order to generate an electric current in the occipital cortex. Complete treatment consists of two pulses delivered within two minutes of each other. The primary safety concern is triggering a seizure. Contraindications to the device include metal in the head or upper body that is attracted to magnets, pacemakers, implantable defibrillators, and other active implanted devices. On May 23, 2014, the FDA approved the SpringTMS (eNeura) device. SpringTMS is a new version of the Cerena device that is smaller, lighter, and uses a rechargeable battery but delivers the same therapy. The SpringTMS device will be marketed in the US following a 600 patient pilot study. The Cerena device will not be marketed in the US. Figure 1 on the next page shows the SpringTMS device. ©Institute for Clinical and Economic Review, 2014 Page 6 Figure 1. eNeura’s SpringTMS Device Copyright 2014 eNeura Inc. Transcutaneous Electrical Nerve Stimulation (TENS): Cefaly Cefaly is a battery-powered headband device with a transcutaneous electrical nerve stimulator (TENS) centered above the eyes. The device delivers steady current at 14 mA to the supraorbital transcutaneous nerves (branches of the trigeminal nerve) through a set of reusable electrodes. It is supposed to be worn for 20 minutes each day to reduce the frequency of migraine headaches. TENS has been used for many years to treat chronic pain at pain centers, although its efficacy remains controversial.39-42 Figure 2 below shows the Cefaly device. Figure 2. Cefaly Device ©Institute for Clinical and Economic Review, 2014 Page 7 Botulinum Toxin (BOTOX) Botulinum toxin type A inhibits the release of acetylcholine at the motor nerve terminals. During cosmetic use to prevent wrinkle formation, patients reported a decline in episodic migraines. The standard protocol used in the Phase III trials of botulinum toxin for patients with chronic migraines is the injection of 155 units of botulinum toxin in 31 specific sites in seven muscle groups of the head and neck. An additional 40 units can be used in up to eight additional sites based on patients’ perception of the location of their worst pain. The procedure is repeated every 12 weeks. Figure 3 below shows a depiction of the recommended injection sites of BOTOX for chronic migraine. Figure 3. Recommended Injection Sites for BOTOX Source: Allergan, prescribing information – BOTOX (onabotulinumtoxinA) for injection, January 2013 Opioid Analgesics Parenteral opioid analgesics have been used for the acute treatment of pain in the emergency room for many years, and ED physicians are comfortable with their use. A recent analysis of national data in the US found that opioids were used 53% of the time to treat migraine pain in the ED in 2010 versus 50% in 1998.43 In a prospective cohort study of 8,219 patients with episodic migraine headaches, opioids had twice the risk of transforming migraines from episodic to chronic compared to acetaminophen; neither triptans nor NSAIDs had an increased risk.44 Thus, there are concerns that harms associated with the use of opioid analgesics outweigh the benefits. ©Institute for Clinical and Economic Review, 2014 Page 8 2. Clinical Guidelines American Academy of Neurology (AAN) https://www.aan.com/Guidelines/Home/ByTopic?topicId=16 http://www.choosingwisely.org/doctor-patient-lists/american-academy-of-neurology/ Their 2012 guidelines do not address the use of the home TMS device for the acute treatment of migraines with aura nor the Cefaly device for prevention of episodic migraines. They note that botulinum toxin is probably ineffective for prevention of episodic migraines and that they have not addressed the evidence for its use in chronic migraine. In their 2013 Choosing Wisely list, the AAN states that opioid medications should not be used to treat migraines except as a last resort because opioids can make headaches worse and they are not as effective as other migraine drugs. American Headache Society (AHS) http://www.americanheadachesociety.org/assets/1/7/How_I_Do_It_Acute_Treatment.pdf http://www.americanheadachesociety.org/new_guidelines_treatments_can_help_prevent_migrain e/ http://www.choosingwisely.org/doctor-patient-lists/american-headache-society/ Their 2012 guidelines do not address the use of the home TMS device for the acute treatment of migraines with aura nor the Cefaly device for prevention of episodic migraines. They do not address botulinum toxin. The AHS states that the evidence for opioid use in migraines is generally poor or negative and that the best clinical advice is that opioids should not be prescribed. In their 2013 Choosing Wisely list, the AHS states that opioid medication use for migraines should be limited due to the risk for addiction and transformation of migraines to chronic headaches. American Academy of Family Physicians (AAFP) / American College of Physicians (ACP) /American Society of Internal Medicine (ASIM) Joint Guideline http://annals.org/data/Journals/AIM/20020/0000605-200211190-00014.pdf The 2012 AAFP/ACP-ASIM guideline does not address the use of the home TMS device for the acute treatment of migraines with aura nor the Cefaly device for prevention of episodic migraines. They do not address botulinum toxin. Opioids are not listed as first or second line therapy because of limited evidence of efficacy and the potential for harm. ©Institute for Clinical and Economic Review, 2014 Page 9 National Institute for Health and Care Excellence (NICE) http://pathways.nice.org.uk/pathways/headaches/management-of-migraine-with-or-without-aura NICE considers the evidence on the efficacy of TMS for the treatment of migraines to be limited in quantity and requires special arrangements for its use. They offer no guidance on the Cefaly device. NICE’s current guidelines recommend treatment with botulinum toxin as an option for patients with chronic migraines who have failed at least three prior preventive drugs; it is not recommended for the treatment of episodic migraines. NICE recommends that opioids NOT be offered for the acute treatment of migraines because of the lack of evidence, their addictive properties, and the risk for medication overuse headache. European Headache Federation http://www.thejournalofheadacheandpain.com/content/pdf/1129-2377-14-86.pdf In 2013, the European Headache Federation evaluated all forms of neuromodulation for chronic headaches including transcranial magnetic stimulation and supraorbital transcutaneous nerve stimulation. They found the evidence to be promising but insufficient and often of suboptimal quality. They recommend further studies of these approaches at tertiary headache centers as part of valid studies. American College of Emergency Physicians (ACEP) http://www.acep.org/policystatements/ The ACEP has no policy directly addressing the management of migraine headache pain. Searching their website brings up a 2007 article entitled “When Migraine Comes to the ED, Fluids First, Opiates Last,” which begins “The use of opiates to treat primary headache in the emergency department is often a knee-jerk response and should be avoided when possible.” ©Institute for Clinical and Economic Review, 2014 Page 10 3. Coverage Policies Coverage policies of a variety of public and private payers for migraine prevention and treatment relevant to this report were reviewed in May 2014 and are described below. 3.1 Single-pulse Transcranial Magnetic Stimulation (sTMS): Cerena/SpringTMS Medicare & Medicaid No publicly available coverage policies or prior authorization protocols for sTMS were available from the Centers for Medicare & Medicaid Services (CMS) or Medi-Cal, California’s Medicaid agency. Regional Private Payers No publicly available coverage policies or prior authorization protocols for sTMS were available from regional private payers. National Private Payers Aetna: http://www.aetna.com/cpb/medical/data/700_799/0707.html http://www.aetna.com/cpb/medical/data/400_499/0462.html TMS is considered experimental and investigational including for migraine. Anthem: http://www.anthem.com/medicalpolicies/policies/mp_pw_a047769.htm TMS of the brain is considered investigational and not medically necessary for migraine. Humana: http://apps.humana.com/tad/Tad_New/Search.aspx?criteria=migraine&searchtype=freetext&polic yType=both The sTMS device is considered investigational. ©Institute for Clinical and Economic Review, 2014 Page 11 United Healthcare: https://www.unitedhealthcareonline.com/ccmcontent/ProviderII/UHC/en- US/Assets/ProviderStaticFiles/ProviderStaticFilesPdf/Tools%20and%20Resources/Policies%20and% 20Protocols/Medical%20Policies/Medical%20Policies/Transcranial_Magnetic_Stimulation.pdf TMS is unproven for treating all conditions including headaches. 3.2 Transcutaneous Electrical Nerve Stimulation (TENS): Cefaly Medicare & Medicaid No publicly-available coverage policies or prior authorization protocols for Cefaly or TENS for migraine were available from CMS or Medi-Cal, California’s Medicaid agency. Regional Private Payers No publicly available coverage policies or prior authorization protocols for Cefaly or TENS for migraine were available from regional private payers. National Private Payers Aetna: http://www.aetna.com/cpb/medical/data/700_799/0707.html http://www.aetna.com/cpb/medical/data/400_499/0462.html Aetna considers nerve stimulation investigational at this time. 3.3 Botulinum Toxin (BOTOX) Medicare & Medicaid Medicare/Noridian: http://www.cms.gov/medicare-coverage-database/details/lcd- details.aspx?LCDId=33513&ContrId=280&ver=9&ContrVer=2&CoverageSelection=Both&ArticleType =All&PolicyType=Final&s=California+Entire+State&KeyWord=botulinum+toxin&KeyWordLookUp=Ti tle&KeyWordSearchType=And&bc=gAAAABAAAAAAAA%3d%3d& A Medicare local coverage decision that covers California states that botulinum toxin type A (BOTOX-onabotulinumtoxinA) is covered for prevention of headaches in adult patients with chronic migraines (≥15 days per month with headache lasting four hours a day or longer). ©Institute for Clinical and Economic Review, 2014 Page 12 Medi-Cal: http://files.medi-cal.ca.gov/pubsdoco/bulletins/artfull/ph201105.asp#a10 OnabotulinumtoxinA (BOTOX) is reimbursable for the prevention of headaches in adult patients with chronic migraines (15 or more days per month with headache lasting four hours a day or longer). Regional Private Payers Health Net: https://www.healthnet.com/static/general/unprotected/html/national/pa_guidelines/xeomin_natl. html Health Net provides coverage for botulinum toxin for patients with chronic migraines for at least three months who have failed trials of at least three classes of migraine prevention medications. The patient must be evaluated by a neurologist and have documented significant disability from the migraine headaches. Blue Shield of California: Blue Shield of California provides coverage for botulinum toxin for patients with chronic migraines who are being treated by a neurologist and have failed at least two classes of migraine prevention medications, and have significant frequency of migraine headaches. National Private Payers Aetna: http://www.aetna.com/cpb/medical/data/100_199/0113.html http://www.aetna.com/products/rxnonmedicare/data/2014/MISC/botulinum_toxin.html Aetna provides coverage for botulinum toxin for patients with chronic migraines who have failed two-month trials of at least three classes of migraine prevention medications. Anthem/WellPoint: http://www.anthem.com/medicalpolicies/policies/mp_pw_a049843.htm Anthem provides coverage for botulinum toxin for patients with chronic migraines for at least six months who have failed trials of at least two classes of migraine prevention medications. ©Institute for Clinical and Economic Review, 2014 Page 13 Cigna: https://cignaforhcp.cigna.com/public/content/pdf/coveragePolicies/pharmacy/ph_1106_coveragep ositioncriteria_botulinum_therapy.pdf Cigna provides coverage for botulinum toxin for patients who meet both requirements: diagnosed with chronic migraines and have failure, contraindication, or intolerance to at least two different prescription migraine prevention therapies. Humana: http://apps.humana.com/tad/tad_new/Search.aspx?criteria=botox&searchtype=freetext&policyTy pe=both Humana provides coverage for botulinum toxin for patients with chronic migraine. Their coverage policy notes that triptans differ in effectiveness so more than one triptan is recommended for treatment of acute migraines before a trial of BOTOX for chronic migraines. United Healthcare: https://www.unitedhealthcareonline.com/ccmcontent/ProviderII/UHC/en- US/Assets/ProviderStaticFiles/ProviderStaticFilesPdf/Tools%20and%20Resources/Policies%20and% 20Protocols/Medical%20Policies/Drug%20Policies/Botulinum_toxin_policy.pdf United Healthcare provides coverage for botulinum toxin for patients with chronic migraines who have failed trials of at least three classes of migraine prevention medications. *Note, as of 9/1/14, the coverage policy will require history of failure (after a trial of at least two months), contraindication, or intolerance to preventive therapy with one agent from two therapeutic classes. https://www.unitedhealthcareonline.com/ccmcontent/ProviderII/UHC/en- US/Assets/ProviderStaticFiles/ProviderStaticFilesPdf/Tools%20and%20Resources/Policies%20and% 20Protocols/Medical%20Policies/Medical%20Policies/MPUB_Policies/Botulinum_toxin_policy_911 4.pdf 3.4 Opioid Analgesics Medicare & Medicaid No publicly-available coverage policies, prior authorization protocols, or formulary designations on opioid use for the treatment of migraines were available from CMS or Medi-Cal, California’s Medicaid agency. ©Institute for Clinical and Economic Review, 2014 Page 14 Regional Private Payers No publicly available coverage policies, prior authorization protocols, or formulary designations on opioid use for the treatment of migraines were available from regional private payers. National Private Payers No publicly available coverage policies, prior authorization protocols, or formulary designations on opioid use for the treatment of migraines were available from national private payers. ©Institute for Clinical and Economic Review, 2014 Page 15 4. Previous Systematic Reviews and Technology Assessments The Agency for Healthcare Research and Quality (AHRQ) recently assessed the comparative effectiveness of preventive therapies for migraines including botulinum toxin. They also assessed the comparative effectiveness of parenteral therapies in the emergency department including opioids. Our search identified two systematic reviews of botulinum toxin for the prevention of migraines and one assessing meperidine for the treatment of acute migraine. 4.1 Formal Health Technology Assessments Agency for Healthcare Research and Quality (AHRQ) http://effectivehealthcare.ahrq.gov/ehc/products/313/1453/migraine-report-130408.pdf http://effectivehealthcare.ahrq.gov/ehc/products/289/1323/CER84_Migraine_FinalReport_201211 19.pdf AHRQ did not assess the use of the home TMS device for the acute treatment of migraines with aura nor the Cefaly device for prevention of episodic migraines. AHRQ found that botulinum toxin reduced headaches in patients with chronic migraine, though with frequent bothersome side effects. AHRQ found that opioids were inferior to antiemetic monotherapy or dihydroergotamine (DHE) plus an antiemetic for the acute treatment of migraine pain in the emergency room; opioids were effective compared to placebo, but less effective than other therapies. They used network meta-analysis to estimate the magnitude of pain reduction for parenteral therapies studied in randomized trials in the ED. The combination of DHE plus metoclopramide or prochlorperazine or neuroleptic antiemetics alone produced the largest reductions in pain (about 40/100 points more than placebo). Parenteral NSAIDS, opioids, and metoclopramide reduced pain by about 25 points. National Institute of Health and Clinical Excellence (NICE) http://www.nice.org.uk/nicemedia/live/13776/59836/59836.pdf There were no formal technology assessments of the use of the home TMS device for the acute treatment of migraines with aura, the Cefaly device for prevention of episodic migraines, or opioids in the ED setting. Botulinum toxin is recommended as an option for the prevention of chronic migraine headaches that have not responded to at least three prior pharmacological preventive therapies. They recommend that treatment with botulinum toxin should be stopped in people ©Institute for Clinical and Economic Review, 2014 Page 16 whose headaches do not adequately respond to treatment (defined as less than a 30% reduction in headache days per month after two treatment cycles). 4.2 Systematic Reviews Botulinum Toxin Shuhendler 2009 Shuhendler AJ, Lee S, Siu M, et al. Efficacy of botulinum toxin type A for the prophylaxis of episodic migraine headaches: a meta-analysis of randomized, double-blind, placebo-controlled trials. Pharmacotherapy. Jul 2009;29(7):784-791. This systematic review and meta-analysis of eight randomized trials with 1,601 participants concluded that botulinum toxin A for the preventive treatment of episodic migraine headaches was not significantly different from placebo, both statistically and clinically. Jackson 2012 Jackson JL, Kuriyama A, Hayashino Y. Botulinum toxin A for prophylactic treatment of migraine and tension headaches in adults: a meta-analysis. JAMA: the Journal of the American Medical Association. Apr 25 2012;307(16):1736-1745. This systematic review and meta-analysis found that botulinum toxin is not effective for the prevention of episodic migraines (nine studies, 1,838 participants) but is more effective than sham injections for chronic migraines (five studies, 1,508 participants). In head-to-head trials, botulinum toxin was not associated with a reduction in headache frequency compared with topiramate, amitriptyline, or valproate. Opioids Friedman 2008 Friedman BW, Kapoor A, Friedman MS, Hochberg ML, Rowe BH. The relative efficacy of meperidine for the treatment of acute migraine: a meta-analysis of randomized controlled trials. Annals of Emergency Medicine. Dec 2008;52(6):705-713. This systematic review and meta-analysis of 11 trials with 625 participants found that meperidine was less effective than DHE and the antiemetics at providing headache relief and that it provided similar relief to ketorolac. It caused more sedation and dizziness than other therapies but less akathisia than the antiemetics. The authors concluded that clinicians should consider alternatives to meperidine when treating acute migraines with injectable agents. ©Institute for Clinical and Economic Review, 2014 Page 17 5. Ongoing Studies There are no ongoing studies reported in ClinicalTrials.gov for transcranial magnetic stimulation for the treatment of migraines and only two very small studies of the Cefaly TENS device for the prevention of episodic migraine. There are a large number of ongoing trials of botulinum toxin, the largest of which are observational cohorts. Finally, there are no registered trials of opioid medications for treatment of migraines. The details of the individual trials are listed on the next three pages. ©Institute for Clinical and Economic Review, 2014 Page 18 Title/Trial Sponsor Study Design Comparators Patient Population Primary Estimated Outcomes Completion Date Single-pulse Transcranial Magnetic Stimulation (sTMS): Cerena/SpringTMS No additional studies identified Title/Trial Sponsor Study Design Comparators Patient Population Primary Estimated Outcomes Completion Date Transcutaneous Electrical Nerve Stimulation (TENS): Cefaly Cathodal tDCS in Chronic Migraine: Case series None • 18-65 years old Migraine September 2014 Neurophysiological Study and Pilot Therapeutic • Chronic migraine frequency Trial (CATCHROMIG) Open label NCT02122237 N= 14 Sponsor: University Hospital of Liege Nodal Transcranial Direct Current Stimulation of RCT Sham Cefaly • 18-65 years old Migraine September 2014 the Visual Cortex Versus Sham Stimulation for frequency • Chronic migraine the Prevention of Episodic Migraine (ANODEM) Double blind NCT02122757 N = 30 Sponsor: University Hospital of Liege Neurophysiological Study of tDCS Effects in Case series None • 18-65 years old Neurophysio- April 2014 Healthy Volunteers logical • Healthy volunteers Open label measures NCT02125422 N=18 Sponsor: University Hospital of Liege ©Institute for Clinical and Economic Review, 2014 Page 19 Title/Trial Sponsor Study Design Comparators Patient Population Primary Estimated Outcomes Completion Date Botulinum toxin Investigation of Efficacy and Safety of RCT Botulinum toxin • Age ≥ 18 years Pain intensity December 2012 Botulinum Toxin A (BOTOX-Allergan Inc) in • Chronic migraine ≥ 3 months (VAS) Migraine Headaches Placebo controlled Saline placebo NCT00660192 N= 50 Sponsor: Yale University A Study Using Botulinum Toxin Type A as RCT Botulinum tox dose 1 • Age 12 – 17 years Pain intensity September 2016 Headache Prophylaxis in Adolescents With • Chronic migraine ≥ 6 months (VAS) Chronic Migraine Placebo controlled Botulinum tox dose 2 NCT01662492 N= 126 Saline placebo Sponsor: Allergan Safety and Efficacy of Botulinum Toxin Type A Cohort Botulinum toxin • Chronic migraine Adverse events May 2015 (BOTOX®) to Treat Chronic Migraine in Korea Quality of life N=600 NCT01976611 Sponsor: Allergan BOTOX® Prophylaxis in Patients With Chronic Cohort Botulinum toxin • Age ≥ 18 years Dysphagia May 2015 Migraine • Chronic migraine Worsening N=1,160 headache NCT01432379 Sponsor: Allergan Use of a Treatment Benefit Questionnaire in RCT Botulinum toxin • Age ≥ 18 years Quality of life December 2014 Patients With Chronic Migraine Treated With • Chronic migraine ≥ 6 months OnabotulinumtoxinA (BOTOX®) Placebo controlled Saline placebo NCT01833130 N= 80 Sponsor: Allergan ©Institute for Clinical and Economic Review, 2014 Page 20 Title/Trial Sponsor Study Design Comparators Patient Population Primary Estimated Outcomes Completion Date A Long-term Efficacy, Safety, and Tolerability Cohort Botulinum toxin • Age ≥ 18 years Headache March 2016 Study of BOTOX® in Patients With Chronic • Chronic migraine frequency Migraine 96 weeks of therapy NCT01516892 N=551 Quality of life Sponsor: Allergan An Observational Study of BOTOX® as Headache Cohort Botulinum toxin • Age ≥ 18 years Health care February 2017 Prophylaxis for Chronic Migraine • Chronic migraine resource N=1,400 utilization NCT01686581 Sponsor: Allergan Title/Trial Sponsor Study Design Comparators Patient Population Primary Estimated Outcomes Completion Date Opioid analgesics for the acute treatment of migraine No additional studies identified ©Institute for Clinical and Economic Review, 2014 Page 21 6. Evidence Review (Methods & Results) The goal of this technology assessment was to evaluate the comparative effectiveness and value of the two acute treatments for migraine headache pain (Cerena sTMS device, opioids in the ED) and two preventive therapies (Cefaly TENS device, botulinum toxin). There is a large body of randomized trial evidence for both acute treatments and prevention of migraine. Therefore, this review focused primarily on the randomized trial evidence for these therapies. Large observational studies were used to supplement the data on harms available from clinical trials. The International Headache Society has published standards for the design and execution of clinical trials studying migraine therapies.45 Their recommended primary outcome for trials of acute treatment is the proportion of patients who are pain-free two hours after treatment. Important secondary outcomes include headache response (reduction in pain from “moderate to severe” at baseline to “mild or none“ at two hours), the incidence of relapse (the proportion of patients pain- free at two hours who experience recurrent headache pain over 48 hours), sustained pain-free status (pain-free without the use of rescue medications from two to 48 hours after treatment), and adverse events (AEs). The recommended primary outcome for trials of preventive treatment is either the total number of headaches or headache days per treatment period (four weeks or one month). Secondary outcomes include the responder rate (the proportion of patients who have a 50% or greater reduction in headache frequency), the use of drugs for acute treatment, and adverse events.45 Because of the large placebo effect observed in randomized trials of therapies for headaches, the International Headache Society recommends that at the end of any trial, all participants be asked to give their best guess as to which treatment they received.45 The Medline database, Embase, Cochrane clinical trials database, Cochrane reviews database, and the Database of Abstracts of Reviews of Effects (DARE) were searched using the key words “transcranial magnetic stimulation” OR “nerve stimulator” OR “botulinum toxin” OR “opioid” AND the keyword “migraine.” Studies of TMS for prevention were excluded. The search was limited to clinical trials published in English. The search was performed for the period from 1945 through May 21, 2014. Full details of the search are in the Appendix. The bibliographies of systematic reviews and key articles were manually searched for additional references. The abstracts of citations were reviewed for relevance, and all potentially relevant articles were reviewed in full. We adopted the approach of the ICER Evidence Rating Matrix* to evaluate the overall evidence for each therapy. The quality of individual studies was assessed by considering the domains listed on the next page, which are adapted from AHRQ’s methods guide: * http://www.icer-review.org/wp-content/uploads/2013/04/Rating-Matrix-User-Guide-Exec-Summ-FINAL.pdf ©Institute for Clinical and Economic Review, 2014 Page 22 • Similarity of baseline characteristics and prognostic factors between comparison groups • Well-described methods for randomization and concealment of treatment assignment • Use of valid, well-described primary outcomes • Blinding of subjects, providers, and outcome assessors • Intent-to-treat analysis (all randomized subjects included) • Limited and non-differential loss to follow-up • Disclosure of any conflicts of interest Fundamentally, the evidence rating reflects a joint judgment of two critical components: a) The magnitude of the difference between a therapeutic agent and its comparator in “net health benefit” – the balance between clinical benefits and risks and/or adverse effects AND b) The level of certainty in the best point estimate of net health benefit. 6.1 Single-pulse Transcranial Magnetic Stimulation (sTMS): Cerena/SpringTMS There are data available from one randomized trial using the Cerena device.46 There is also one dose finding series using a tabletop device.47 Finally, there is a literature review of the safety data from prior trials of TMS, although most of the data came from repetitive TMS used for indications other than migraine.48 There are no published data on the SpringTMS device. Several studies used TMS as preventive therapy in patients with chronic migraines.49-52 It is instructive to note that in the most recent trial, the number of headache days decreased more in the sham group than in the TMS group (p<0.001), as did disability as measured by the Migraine Disability Assessment (MIDAS) score (p<0.001).49 This highlights the potential for significant placebo effects in studies of migraine treatment and the importance of using a believable sham when studying devices. An early, uncontrolled case series47 used a tabletop TMS device for the acute treatment of migraine headaches. The investigators randomized 42 participants with headaches (10 with aura) to low- versus high-dose TMS.47 There was no control group. There were no observed differences in pain or headache recurrence between groups. The proportion of participants pain-free two hours after treatment was not reported. The investigators reported that pain decreased by 75% using a 5-point Likert scale at minutes 5, 10, 15, and 20. At 24 hours following treatment, 32% of patients were pain-free. One patient reported transient dizziness and a second reported fatigue. The manufacturers for the Cerena sTMS device sponsored a high-quality randomized trial comparing sTMS to sham.46 The characteristics of the trial and its participants are summarized in Appendix Table A1. The elements involved in the assessment of the risk of bias in the trial are ©Institute for Clinical and Economic Review, 2014 Page 23 summarized in Appendix Table A2. The primary results of the trial are summarized in Table 2 below. The sham device appeared identical to the active device, and when asked after treatment, equal proportions of the active and sham arms thought they received the active device. The study was appropriately randomized and the blind was not broken until the data was locked, so patients and study staff were blinded throughout the study. A modified intention-to-treat analysis was used, including only those participants who had used the device at least once. The study participants were between 18 and 70 years old and experienced between one and eight migraines per month with aura preceding the migraine in at least 30% of the episodes. The headaches were required to be moderate to severe 90% of the time. Potential participants were excluded if they had auras lasting more than 60 minutes or had contraindications to magnetic stimulation, such as metal implants in their heads or pacemakers. The study initially randomized 201 participants but only analyzed 164 (82 in each group) because the others did not use their device during the study period. The primary outcome was being pain- free at two hours after therapy for the first treated attack during follow-up. Participants were followed for three months. The pain-free response was significantly greater in the active group compared to the sham group (39% versus 22%, p=0.018). The sustained pain-free response at 24 hours (29% versus 16%, p=0.04) and 48 hours (27% versus 13%, p=0.03) also favored the active treatment. However, there were no significant differences between the two groups in the response rate at two hours, relief of photophobia and phonophobia at two hours, the use of rescue medications at two and 48 hours, or disability measured using the MIDAS score. Adverse event rates were similar in both groups (14% active, 9% sham). The most common AEs were dizziness, increased nausea, paresthesias, and increased headache pain. No single AE was clearly greater in the active treatment group, perhaps because of the relatively small size of the trial. Table 2: Primary Outcomes in the Randomized Trials of Single Pulse TMS Study Group N Pain- Mild or Pain- Pain- Use of rescue Use of rescue Change AEs free 2 no pain 2 free 24 free 48 medication medication MIDAS hours hours hours hours 0-2 hours 0-48 hours Lipton sTMS 82 39% 72% 29% 27% 18% 48% -4.6 14% 53 2010 Sham 82 22% 67% 16% 13% 16% 46% -4.7 9% USA 18 centers A review of the literature on the safety of TMS was published in 2010.48 Most of the data come from studies of repetitive TMS treatments. The most common AEs were headache, neck ache, and scalp irritation. There is a theoretical risk of inducing seizures, but this has not been observed in clinical trials to date. ©Institute for Clinical and Economic Review, 2014 Page 24 In summary, the data on the use of sTMS for the acute treatment of migraines with aura is inconclusive. Although there was an increase in the proportion of patients who were pain-free at two hours, there was no difference in the proportion of patients with a response to treatment and no reduction in the use of rescue medications for the headache or in disability scores. Thus, the net benefit is at best small, although there do not appear to be significant harms associated with use of the device. The degree of certainty about the effect is low because only one relatively small study has been published, and that study was both funded and conducted by the manufacturer. 6.2 Transcutaneous Electrical Nerve Stimulation (TENS): Cefaly There are data available from one small, randomized trial using the Cefaly device to treat migraines54 and one randomized trial in healthy volunteers evaluating side effects of the Cefaly device.55 There are also safety data from a prospective cohort.56 The characteristics of the Cefaly trial and its participants are summarized in Appendix Table A3. The elements involved in the assessment of the risk of bias in the trial are summarized in Appendix Table A4. The primary results of the trial are summarized in Table 3 on the next page. The PREvention of MIgraine using the STS Cefaly (PREMICE) study was a prospective, multicenter, sham-controlled trial funded by the Walloon region of Belgium.54 The sham device was identical to the active device, but used 1 Hz 1 mA biphasic impulses rather than the 60 Hz 16 mA impulses used in the active device. The investigators note that it was “not possible to distinguish a sham from a verum simulator without testing both devices in parallel.” The study did not ask participants about their perceptions on receiving a placebo or active device, unlike the study of the sTMS device described above. The study was appropriately randomized and the blind was not broken until the data was locked, so patients and study staff were blinded throughout the study. The likelihood of bias was judged to be high because it is likely that participants could tell if they were randomized to the active device group of the trial and because the groups were not comparable at baseline. The study participants were between 18 and 65 years old and experienced at least two migraines per month. Potential participants were excluded if they used migraine preventive therapies in the past three months or had failed more than three prior trials of preventive therapies. The study randomized 67 participants. The primary outcome measures were the reduction in the number of migraine days comparing the one month run-in period to the third month of use of the device and the proportion of patients with at least a 50% reduction in monthly migraine days. Participants were followed for three months. The participants used the device on approximately 58% of the 90 potential treatment days. ©Institute for Clinical and Economic Review, 2014 Page 25 There was a non-significant greater decrease in migraine days per month (active -2.06, sham -0.32, p= 0.054), but the difference in those with at least a 50% reduction in migraine days was significant (active 38%, sham 12%, p= 0.023). The change in use of rescue medications for migraines was greater in the active treatment group (-4.2 versus 0, p=0.007). In addition, a greater proportion of patients in the active treatment group were moderately or very satisfied with the device (71% versus 39%, p NR). No adverse events were reported in either group, which likely reflects no systematic gathering of adverse events, as it would be unusual for none to occur over three months in a group of 67 participants. Table 3: Primary Outcomes in the Randomized Trials of the Supraorbital Transcutaneous Stimulator (Cefaly) Study Group N Headache Headache Change Change Reduction Change in Patient AEs days per days per in in by at least use of satisfaction month, month, migraine migraine 50% in rescue (moderate run-in month 3 days at 3 attacks migraine medication or very) months at 3 days months Schoenen Cefaly 34 6.9 4.9 -2.1 0.1 38% -4.2 71% 0 54 2013 Sham 33 6.5 6.2 -0.3 0.5 12% 0 39% 0 Belgium 5 centers In an accompanying editorial, the AAN considered this only Class III evidence (Class I being the highest level of evidence) because the trial was small, the confidence interval for the primary outcome was wide, there were some apparent baseline differences between the two groups, and because of potential unblinding.57 The potential for unblinding was highlighted as the most important potential problem with the trial. The authors note that the stimulation electrodes of the active device could be painful to the touch with fingers while the sham device electrodes would not be painful. They recommend that future sham devices be designed with stimuli strong enough to be perceived.57 A second randomized trial using the Cefaly device was done in healthy volunteers and thus had no migraine-specific outcomes.55 The investigators used a cross-over design to evaluate potential sedative side effects of supraorbital nerve stimulation based on prior studies. They found that high- frequency stimulation caused a decrease in vigilance and attention compared with low frequency stimulation (p<0.001). The high frequency stimulation (120 Hz) was twice that of the typical frequency recommended clinically for the Cefaly device (60 Hz). Finally, investigators described the experience of 2,313 patients with migraines from France, Belgium, and Switzerland who rented the device for 40 days.56 The rate of adverse events was 4.3%. The most common AE was intolerance to the paresthesias felt during electrical stimulation ©Institute for Clinical and Economic Review, 2014 Page 26 (1.3% of patients). Two percent (n=46) stopped using the device due to an AE. Other common AEs included sleepiness (0.5%), headache following treatment (0.5%), and forehead skin irritation (0.2%). A total of 53% of patients elected to purchase the device after the trial period; the rest returned it. Interestingly, those who returned the device used it 59% of the recommended length of time during the rental period, almost identical to the utilization duration observed in the randomized trial. In summary, the data on the use of the Cefaly TENS device for the prevention of migraine headaches is inconclusive. The net benefit appears to be small. There was an increase in the proportion of patients who had a 50% or greater reduction in migraine days, and there were no reported adverse events, though this is likely due to underreporting given the rate of AEs reported in the observational study. There is also a low degree of certainty about the effect because only one relatively small study has been published, and differential unblinding of the intervention may have biased the results of that study. 6.3 Botulinum Toxin for Prevention of Migraines In contrast to the two devices, there is a wealth of randomized trial evidence evaluating the efficacy of botulinum toxin. The literature search identified 22 trials that randomized 4,920 patients to botulinum toxin injections or to placebo injections.58-79 Prior meta-analyses and systematic reviews reported that botulinum toxin was effective in reducing migraine frequency in patients with chronic migraine headaches but not in patients with episodic migraine headaches.80,81 No new trials have been published since those systematic reviews, so we adopted their framework to examine the net health benefits of botulinum toxin. The characteristics of the randomized trials and their participants are summarized in Appendix Table A5. The elements involved in the assessment of the risk of bias in the trials are summarized in Appendix Table A6. The primary results of the trials are summarized in Table 4 on the next two pages. The principal adverse events in the trials are summarized in Table 5 on page 29. The mean age of participants in the trials was the early 40s, and more than 85% were women. The average number of headaches in the participants ranged from 4 to 9 in the trials among patients with episodic migraines and from 13 to 25 among patients with chronic migraines. The dose of botulinum toxin ranged widely across the trials from a low of 7.5 units to a high of 300 units. The two Phase III trials (PREEMPT 174 and PREEMPT 275) are the only trials that evaluated the current standard treatment approach of 155 units in 5-unit increments at 31 specific sites plus up to 40 units of additional 5-unit injections at sites of maximal pain. Those two trials will be described in more detail below. Among patients with episodic migraine headaches, the meta-analysis showed no benefit of botulinum toxin injections in the reduction of migraines compared to placebo (difference in ©Institute for Clinical and Economic Review, 2014 Page 27 headache days of 0.05 per month, 95% CI -0.26 to +0.36).80 There was also no difference in the proportion of patients with at least a 50% reduction in headache frequency (RR 1.0, 95% CI 0.85 to 1.18)80 in this population. Table 4: Primary Outcomes in the Randomized Trials of Botulinum Toxin for the Prevention of Migraine Headaches Study Group* N Headache days Reduction by at least Use of rescue per month 50% in migraine days medication doses Episodic Migraine 58 Silberstein 2000 Botulinum 82 2.7 45% -2.4 US 12 Centers Placebo 41 3.4 24% -0.8 59 Barrientos 2003 Botulinum 15 2.6 Not reported 1.7 Chile Placebo 15 4.0 5.6 60 Evers 2004 Botulinum 40 2.8 30% 4.4 Germany 1 Center Placebo 20 3.2 33% 5.2 64 Anand 2006 Botulinum 16 3.0 Not reported Not reported India 1 Center Placebo 16 5.7 65 Elkind 2006 Botulinum 312 4.0 Not reported Not reported North America 7 Centers Placebo 106 3.7 66 Aurora 2007 Botulinum 187 2.9 59% Not reported North America 20 Centers Placebo 182 2.4 60% 67 Relja 2007 Botulinum 377 2.0 55% Not reported Europe 37 Centers Placebo 118 1.9 44% 68 Saper 2007 Botulinum 187 4.2 Not reported Not reported North America 7 Centers Placebo 45 4.1 70 Cady 2008 Botulinum 40 7.4 Not reported Not reported US 1 Center Placebo 19 8.4 73 Petri 2009 Botulinum 60 3.0 Not reported Not reported Germany 16 Centers Placebo 62 4.0 78 Chankrachang 2011 Botulinum 86 3.2 Not reported Not reported Thailand 6 Centers Placebo 42 2.6 Chronic Migraine – Placebo control 61 Ondo 2004 Botulinum 30 20.0 Not reported 3.7 US 1 Center Placebo 30 24.8 4.7 62 Mathew 2005 Botulinum 173 3.5 46% -6.0 North America 13 Centers Placebo 182 4.9 35% -5.0 ©Institute for Clinical and Economic Review, 2014 Page 28 Study Group* N Headache days Reduction by at least Use of rescue per month 50% in migraine days medication doses 63 Silberstein 2005 Botulinum 524 5.4 Not reported Not reported North America 28 Centers Placebo 178 7.4 69 Vo 2007 Botulinum 15 20.7 Not reported Not reported US 1 Center Placebo 17 20.9 71 Freitag 2008 Botulinum 30 10.1 33% 18 US 1 Center Placebo 30 15.4 17% 21 74 Aurora 2010 Botulinum 341 7.1 Not reported -10.3 North America 56 Centers (PREEMPT 1) Placebo 338 8.1 -10.4 75 Diener 2010 Botulinum 347 11.2 Not reported -9.9 Europe, North America 66 Centers (PREEMPT 2) Placebo 358 13.4 -8.4 79 Sandrini 2011 Botulinum 33 12.0 70% 10.7 Italy Placebo 35 15.9 31% 14.3 Chronic Migraine – Active control 72 Mathew 2009 Botulinum 19 9.2 41% Not reported US 1 Center Topiramate 17 8.7 43% 76 Magalhaes 2010 Botulinum** 35 11.8 68% 8.3 Brazil 1 Center Amitriptyline 37 9.7 72% 7.0 77 Cady 2011 Botulinum 29 13.8 38% Not reported North America 3 Centers Topiramate 30 12.4 50% Note: *placebo = sham injection with saline; **Botulinum in this study was a different form of botulinum toxin A marketed as Dysport by Ipsen Biopharm Ltd. Table 5: Adverse Events in the Randomized Trials of Botulinum Toxin for the Prevention of Migraine Headaches Likely Associated With Active Treatment Adverse event Botulinum toxin Placebo Relative risk Absolute risk Number difference needed to harm Any adverse event 57% 46% 1.2 11% 9 Withdrawal, any 40% 32% 1.3 8% 12 cause Muscle weakness 21% 2% 9.0 19% 5 Neck pain 19% 4% 5.3 15% 6 Neck stiffness 14% 4% 3.2 10% 10 Drooping eyelid 8% 1% 7.6 7% 15 Parasthesia 3% 1% 2.2 2% 61 Among patients with chronic migraine headaches, the meta-analysis showed a benefit of botulinum toxin injections in the reduction of migraines compared to placebo (difference in headache days of ©Institute for Clinical and Economic Review, 2014 Page 29 -2.30 per month, 95% CI -0.3.66 to -0.94).80 There was also a significant difference in the proportion of patients with at least a 50% reduction in headache frequency (RR 2.2, 95% CI 1.3 to 3.8)80 among patients with chronic migraine headaches. However, the absolute benefit is relatively small compared to the placebo effect. This is best seen in the Phase III trials. In the PREEMPT 1 trial, 341 participants were randomized to botulinum toxin and 338 patients to placebo injections and followed for 24 weeks.74 Eligible patients were between the ages of 18 and 65 years old with chronic migraine headaches without hemiplegic or basilar-type migraines. They were not required to have tried and failed any prior preventive therapies. The participants had an average of 20 headache days during the 28-day run-in period, and 38% had never been treated with any preventive therapy. For the primary outcome measure, headache episodes, there was no difference between the two groups at 24 weeks (-5.2 days per month botulinum toxin, -5.3 days per month placebo, p=0.344). An important secondary outcome was the change in total headache days (a headache episode can last for more than one day): botulinum toxin was more effective than placebo for this outcome (-7.8 days per month botulinum toxin, -6.4 days per month placebo, p=0.006). While statistically significant, the 1.4 day difference between the botulinum toxin group and the placebo group is relatively small compared to the 6.4 day decrease in headaches observed in the placebo group. There were no differences in the reduction in use of all acute headache medications (-10.3 botulinum toxin, -10.4 placebo), but there was a significant reduction in the use of triptans (-3.3 botulinum toxin, -2.2 placebo, p=0.023). In addition, there was a significantly greater improvement in the Headache Impact Test (HIT-6) score, a six-item questionnaire assessing the impact of headaches on functional status (-4.7 versus -2.4, p<0.001). The score ranges between 36 and 78 points with the minimally important change estimated to be 2.5 points.82 Thus, the 2.3 point difference between groups, while statistically significant, is only minimally significant from a clinical perspective. The primary outcome of the PREEMPT 2 trial was amended to be the change in the baseline frequency of headache days based in part on the results of the PREEMPT 1 trial.75 As in PREEMPT 1, eligible patients were between the ages of 18 and 65 years old with chronic migraine headaches without hemiplegic or basilar-type migraines. They were not required to have tried and failed any prior preventive therapies. The participants had an average of 20 headache days during the 28 day run-in period, and 35% had never been treated with any preventive therapy. For the primary outcome measure, headache days, there was a significant difference between the two groups at 24 weeks (-9.0 days per month botulinum toxin, -6.7 days per month placebo, p<0.001). Again, the between group difference (2.3 days) was much smaller than the absolute reduction in the placebo group (6.7 days). In PREEMPT 2, there was also a small but significantly greater reduction in headache episodes (-5.3 days per month botulinum toxin, -4.6 days per month placebo, p=0.003). As in PREEMPT 1, there was no difference in the reduction of all acute medications used (-9.9 versus -8.4, p=0.13), but there was a significant reduction in triptan use (-1.7 versus-1.3, p<0.001) and a greater improvement on the HIT-6 score (-4.9 versus -2.4, p<0.001). ©Institute for Clinical and Economic Review, 2014 Page 30 Other benefits of botulinum toxin in the two PREEMPT trials were described in a pooled analysis.131 The number of headache days per month decreased by 8.4 in the botulinum toxin group and by 6.6 days in the placebo group (difference -1.8 days, 95% CI -2.5 to -1.1, p<0.001). A greater proportion of patients in the botulinum toxin group had at least a 50% reduction in headache days (47% versus 35%, p<0.001). The number of migraine days per month decreased by 8.2 in the botulinum toxin group and by 6.2 days in the placebo group (difference -2.0 days, 95% CI -2.7 to -1.3, p<0.001). The number of headache hours per month decreased by 120 in the botulinum toxin group and by 81 in the placebo group (difference -39 hours, 95% CI -48 to -21, p<0.001). Similarly, the number of headache episodes per month decreased by 5.2 in the botulinum toxin group and by 4.9 in the placebo group (difference -0.3 episodes, 95% CI -1.2 to -0.2, p=0.009), and the number of migraine episodes per month decreased by 4.9 in the botulinum toxin group and by 4.5 in the placebo group (difference -0.4 episodes, 95% CI -1.2 to -0.2, p=0.004). As noted above, the between group differences were much smaller than the improvement observed with placebo alone. The quality of life measures improved more in the botulinum toxin group. The total HIT-6 scores declined by 4.8 points in the botulinum toxin group and by 2.4 points in the placebo group (difference -2.4 points, 95% CI -3.1 to -1.7, p<0.001). A similar doubling in the improvement in scores was seen in the role restrictive, role preventive, and emotional function sub-scales of the Migraine-Specific Quality of Life Questionnaire (p<0.001 for all 3 comparisons). There are consistently more adverse events in the summary of all randomized trials of botulinum toxin (see Table 5).80,81,83 Muscle weakness, neck pain, neck stiffness, and drooping eyelids occurred between 3 and 9 times more often in the botulinum toxin groups than in the placebo groups. The number needed to treat to cause one adverse event (number needed to harm) ranged from 5 to 10 for four of the common events listed in Table 5. Despite this, the discontinuation rate due to AEs from botulinum toxin was low (4%). In a pooled analysis of the two PREEMPT trials, the adverse event rates were consistently lower than those reported in the summary of all trials.83 This may reflect the evolution of the protocol for treating chronic migraines with botulinum toxin. For example, in the PREEMPT trials, neck pain was reported in 14% of patients in the botulinum toxin group and 2% in the placebo group, which is lower than the prevalence of neck pain in Table 5. Muscle weakness was even less frequently reported (9% versus 0.3%). The large difference in the absolute event rates for adverse events clearly related to botulinum toxin as shown in Table 5, such as muscle weakness (21% versus 2%), raises the possibility of differential unblinding of participants in the trial. Because the placebo effect is so large in the trials of botulinum toxin and because the between group differences in the outcomes are much smaller than the placebo effect, unblinding by side effects likely explains at least some of the between group differences. The International Headache Society guidelines45 recommend that both patients ©Institute for Clinical and Economic Review, 2014 Page 31 and providers be asked about group assignment at the end of the study. Those results were not reported in the PREEMPT studies.74,75 In prior trials of botulinum toxin, 70% of participants correctly identified whether they were in the botulinum toxin or placebo group after their first treatment.62 One author suggests that subjects cannot be adequately blinded to botulinum toxin because it rapidly paralyzes the forehead muscles and thus prevents the usual wrinkling, which is readily visible in the mirror.84,85 He argues that the placebo effect plus the nocebo effect (disappointment when not in the active group leading to a reduced placebo effect) may explain all of the differences between groups in the PREEMPT trials. In summary, there is consistent direct evidence from multiple randomized trials that botulinum toxin offers no clinically significant benefits in the prevention of episodic migraines and that it causes frequent adverse events. There is also consistent, direct evidence from multiple randomized trials that botulinum toxin offers small but statistically significant benefits in the prevention of chronic migraines compared to sham therapy. Given the demonstrated adverse events caused by botulinum toxin and the uncertainty about whether the observed clinical benefits are due to unblinding, there is moderate certainty that the net health benefits are small, at best. There are three small, randomized trials in patients with chronic migraines that directly compare botulinum toxin to established preventive therapies for episodic migraines (two versus topiramate72,77 and one versus amitriptyline76). The size of the trials (n = 59, 60, and 79) was too small to demonstrate equivalence or superiority (for example, the sample sizes in the PREEMPT trials were 679 and 705), and the amitriptyline trial was against a formulation of botulinum toxin that is not approved for the prevention of chronic migraines (Dysport).76 In these three underpowered studies, the investigators concluded that botulinum toxin and oral therapies had equivalent efficacy. Given this limited evidence base and the at best small net health benefits for botulinum toxin compared to sham therapy, the evidence is insufficient to assess the net health benefits for botulinum toxin compared to oral therapies for the prevention of chronic migraines. 6.4 Opioids for Acute Treatment of Migraines in the Emergency Department As with botulinum toxin, there are a large number of randomized trials of parenteral opioids for the acute treatment of migraine pain in the ED setting. However, the majority are small, of poor quality, and do not follow the International Headache Society recommendations for the design of randomized trials for the treatment of migraines. The literature search identified 17 trials that randomized 1,203 patients to parenteral opioids and either other classes of parenteral medications or placebo.86-102 Prior systematic reviews and meta-analyses concluded that opioids were effective in reducing headache pain compared with placebo but had equivalent or worse efficacy compared ©Institute for Clinical and Economic Review, 2014 Page 32 with other parenteral interventions. No new trials have been published since those systematic reviews.32,34,103 The characteristics of the randomized trials and their participants are summarized in Appendix Table A7. The elements involved in the assessment of the risk of bias in the trials are summarized in Appendix Table A8. The primary results of the trials are summarized in Appendix Table A9. The mean age of participants in the trials was the early 30s, and approximately 80% were women. The average headache intensity in the participants was about 8 on a 10-point visual analog scale (VAS). Meperidine (Demerol) was far and away the most commonly studied opioid. Hydromorphone (Dilaudid), the most commonly used opioid medication for migraine,43 has not been studied in randomized trials for acute migraine therapy. Four randomized trials compared an opioid medication to placebo.87,89,92,101 In three of the four, the opioids nalbuphine, meperidine, and tramadol all reduced headache pain more effectively than placebo, though with increased adverse events such as sedation, nausea, and dizziness.87,89,101 In the fourth trial, the reduction in pain with the combination of meperidine plus promethazine was equivalent to the pain reduction achieved with the placebo saline injection.92 Fifteen randomized trials compared an opioid medication to a non-opioid active comparator.86,88- 100,102 This includes two trials that had both active and placebo control groups.89,92 The results in these trials were mixed. Only one trial found an opioid more effective than an active comparator.97 In this trial of 31 participants, with baseline differences between the two groups and poor reporting of the trial methodology, meperidine was more effective than ketorolac at reducing headache pain and disability. Nine trials found equivalence between opioid therapy and other parenteral therapies, including three trials comparing meperidine to ketorolac.92,94,96 Finally, five trials reported that the active controls were more effective than opioid therapy.89,95,99,100,102 Few trials compared opioids to parenteral ergotamines, and no trials compared opioids to parenteral triptan therapy. As noted above, the quality of these trials was generally poor. There was inadequate description of the methods of randomization and allocation concealment, no primary outcome was specified, no intent-to-treat analysis was described, no power calculations were performed, and there was inadequate or no assessment of adverse events. The current standard outcomes of complete pain relief at two hours, pain response at two hours, use of rescue medications, and relapse at 24 and 48 hours were rarely reported. The trials were also quite small, with large baseline differences between the randomization groups or no comparison between groups reported. This field would benefit from large, high-quality randomized trials comparing the efficacy and adverse effects of the commonly used therapies for severe migraine headaches in the ED. ©Institute for Clinical and Economic Review, 2014 Page 33 In the absence of high-quality head-to-head randomized trials, AHRQ performed a network meta- analysis as part of its review of acute migraine treatment in emergency settings.103 Their analysis showed that the most effective treatments were combination therapy with dihydroergotamine added to either neuroleptics or metoclopramide or neuroleptic monotherapy with a pain reduction of approximately 40 mm on a visual analog scale compared to placebo therapy. Metoclopramide monotherapy, opioids, and NSAIDs were the next most effective treatments, with a pain reduction of approximately 24 mm but low strength of evidence. Other agents (DHE alone, triptans alone) were less effective, with a pain reduction of approximately 12-16 mm.103 In summary, there is fair evidence that opioid analgesics are more effective than placebo at relieving severe migraine headache pain in the emergency setting. However, there is strong evidence that alternative therapies are more effective. There are significant concerns that opioids may convert episodic migraines to chronic migraines and that opioids have the potential for dependence and misuse. Thus, there is consistent direct, but poor quality, evidence from multiple randomized trials that parenteral opioid therapy offers no clinically significant benefits in the acute treatment of migraines compared to established alternatives and that it is associated with adverse events. ©Institute for Clinical and Economic Review, 2014 Page 34 7. Model of Clinical and Economic Outcomes of Treatment Strategies for Migraine 7.1 Overview To further understand the clinical and economic implications of the controversies in migraine management highlighted in this review, we developed separate models of the clinical and economic outcomes of migraine management. Models were developed to support the major review topics, as listed below: • Potential costs and cost-effectiveness of Cefaly and SpringTMS devices versus relevant pharmacologic comparators • Potential costs and cost-effectiveness of BOTOX injections at two levels of migraine frequency • Economic burden of opioids for treatment of migraines in California and potential cost savings from reductions in ED use of opioids The Cefaly, SpringTMS, and BOTOX analyses focused on hypothetical cohorts of 1,000 patients with episodic (Cefaly, SpringTMS) or chronic (BOTOX) migraine. The opioid model was developed as a population-based analysis to document the breadth and impact of opioid use in the statewide migraine population. All costs were expressed in 2013 US dollars and were updated as necessary using the medical care component of the US Consumer Price Index.104 All analyses were conducted using Microsoft Excel® 2010. Detailed methods and primary findings for each model are discussed in the sections that follow. 7.2 Potential Costs and Cost-effectiveness of SpringTMS for Acute Treatment of Migraine As noted in the evidence review, there are limitations in the evidence base for SpringTMS, but there is also clinical interest in its potential use as an acute treatment for episodic migraine. For our analysis of the potential costs and cost-effectiveness of SpringTMS, we considered a hypothetical cohort of 1,000 migraineurs with episodic migraine and compared potential outcomes and costs to abortive treatment with triptans. Model parameters are presented in Table 6 on the following page. We chose sumatriptan 100 mg as the comparator given its longstanding use and availability in generic form. The outcome of interest was treatment response, which was defined as the proportion of patients who were pain-free 24 hours after treatment, a major outcome of a comprehensive meta-analysis of 53 triptan RCTs31 and a major endpoint in the Cerena vs. sham ©Institute for Clinical and Economic Review, 2014 Page 35 treatment RCT.46 The percentage of patients with pain-free response in the Cerena RCT was 29%; we further assumed that SpringTMS’ effectiveness would be identical to that of the Cerena device used in the trial. The pain-free proportion for sumatriptan was estimated to be 20% based on data from the triptan meta-analysis.31 Table 6. Key Parameter Estimates for SpringTMS Model Parameter Estimate Source(s) Pain-free treatment response (%) SpringTMS 29.0 Lipton, 2010 Sumatriptan 20.0 Ferrari, 2001 Discontinuation due to adverse effects (%) SpringTMS 0.0 Lipton, 2010 Sumatriptan 6.0 Ferrari, 2001 Frequency of episodic migraine, monthly (n) For SpringTMS and sumatriptan 4.4 Lipton, 2010 Intervention cost, one year ($) SpringTMS $3,000 Assumption (estimated rental fee of $250/month) Sumatriptan 100 mg per attack / $2 / $112 Redbook, 2014 annual † Costs of episodic migraine ($) Per year, non-responders $2,221 Munakata, 2009 Per year, responders $1,314 Assumption and Munakata, 2009 † Includes cost of medications and other health care services as well as lost productivity Patients were assumed to use the device or sumatriptan 4.4 times per month, which was derived based on the baseline migraine frequency reported in the Cerena trial. We also assumed that those not responding to treatment would receive rescue therapy with intramuscular ketorolac. No patient was assumed to discontinue the device due to adverse events, consistent with findings from the clinical trial and information from a TMS safety review.46,47 The rate of discontinuation from sumatriptan was assumed to be 6%, based on rates of serious central nervous system and cardiac events reported in the triptan meta-analysis.31 Patients discontinuing sumatriptan therapy were assumed to incur the costs of one month of therapy but to have none of the clinical benefit (i.e., 0% response rate). ©Institute for Clinical and Economic Review, 2014 Page 36 We estimated the cost per sumatriptan tablet to be approximately $2 based on published wholesale acquisition costs (Medi-Cal does not have a published contracted rate for the drug).105 The cost of generic ketorolac IM injection (60 mg) was estimated to total $38 based on published wholesale acquisition costs.105 The SpringTMS device does not yet have a published price in the US, but it is expected to be provided to patients through a monthly leasing arrangement. Correspondence from the manufacturer indicated that the monthly fee will be approximately $250 per month ($3,000 annually). We estimated the other costs of episodic migraine management (approximately $2,000 annually) using data from a study of the costs of health care services and productivity loss among patients with episodic and chronic migraine based on data obtained from the American Migraine Prevalence and Prevention (AMPP) study.106 Because this analysis focused on mitigation of pain, not reduction in other symptoms or prevention of future migraine, our estimates of the economic effects of treatment response were conservative. Specifically, we estimated that response to SpringTMS or sumatriptan would eliminate the use of other acute medications and would reduce other migraine management costs by 25%, resulting in annual costs of slightly more than $1,300. A summary of key assumptions for the SpringTMS model is below: • Patients discontinuing sumatriptan would incur costs of one month of drug therapy but receive no clinical benefit • No SpringTMS user would discontinue due to adverse events • SpringTMS’ effectiveness is assumed to be identical to that of the earlier-generation device • Patients responding to either treatment would eliminate the need for other acute medications and have 25% reductions in other costs of episodic migraine management • Non-responders require use of intramuscular ketorolac for rescue and full costs of episodic migraine management Model Results Findings from the SpringTMS model can be found in Table 7 on the following page. In a hypothetical cohort of 1,000 patients with episodic migraine, 290 and 188 patients would be expected to respond to SpringTMS and sumatriptan therapy respectively over one year. Based on the anticipated monthly rental fee of $250, SpringTMS would result in excess costs of therapy of approximately $2.9 million over one year. Other costs of migraine management would be reduced with SpringTMS by approximately $140,000, resulting in total excess costs of $2.8 million. The cost per treatment response based on this assumed device price is approximately $27,000. ©Institute for Clinical and Economic Review, 2014 Page 37 Table 7. One-year Outcomes and Costs of SpringTMS and Sumatriptan Acute Treatment among 1,000 Hypothetical Patients with Episodic Migraine Outcome/Cost SpringTMS Sumatriptan Difference (SpringTMS- Sumatriptan) Treatment response (n) Responders 290 188 102 Non-responders 710 812 Costs ($) Intervention $3,000,000 $106,278 2,893,722 Other migraine mgmt $2,283,405 $2,422,732 ($139,328) Total $3,033,405 $2,529,011 $504,394 Cost per treatment response ($) ~$27,000 Holding all other outcomes constant, reductions in the cost of the SpringTMS device would continue to reduce the cost-effectiveness ratio. For example, at a monthly rental fee of $150, the cost per treatment response would be approximately $15,000. At a monthly fee of $100, the resulting ratio would be reduced to approximately $9,400. Based on the clinical data, assumptions, and other reimbursement levels used in this analysis, the SpringTMS device has the potential to be cost-saving relative to sumatriptan therapy over one year at a monthly rental cost of approximately $21. We compared costs and outcomes over a two-year time period in an alternative analysis (both outcomes and costs were discounted by 3.5% in the second year). The difference in the number of patients responding would grow to 153 per 1,000 in favor of SpringTMS (vs. 102 over one year). However, despite an increase in the cost offsets due to treatment response to approximately $400,000 (vs. $140,000 at one year), the differential in intervention costs would continue to widen. At the assumed monthly rental cost of $250, the cost per additional treatment response would be approximately $34,000. Cost neutrality would be achieved over two years at a monthly rental fee of $26. 7.3 Potential Costs and Cost-effectiveness of Cefaly for Prevention of Migraine While the evidence for the Cefaly device is also limited to a single, small RCT with quality concerns that compared the device to a sham instrument,54 we nevertheless believed it would be worthwhile to explore the potential costs and cost-effectiveness of the device in comparison to a relevant comparator for a population of episodic migraineurs who are candidates for initial treatment. As with the SpringTMS device, we evaluated outcomes and costs over one year in 1,000 hypothetical ©Institute for Clinical and Economic Review, 2014 Page 38 patients, although in this instance, the focus was on prevention of episodic migraines rather than their treatment. Like the SpringTMS analysis, the outcome of primary interest in this evaluation was the proportion of treatment responders, but in this case, treatment response was defined as a reduction of 50% or more in migraine days per month. Key model parameters are available in Table 8 below. Table 8. Key Parameter Estimates for Cefaly Model Parameter Estimate Source(s) Treatment response (%) Cefaly 38.2 Schoenen, 2013 Metoprolol 39.9 Shamliyan, 2013 Discontinuation due to adverse effects (%) Cefaly 0.0 Schoenen, 2013 Metoprolol 1.0 Shamliyan, 2013 Intervention cost, one year ($) Cefaly $449 Manufacturer website Metoprolol 200 mg daily (per $4 / $50 Medi-Cal Contract Drug List month / per year) † Costs of episodic migraine ($) Per year, non-responders $2,221 Munakata, 2009 Per year, responders $1,040 Assumption and Munakata, 2009 † Includes cost of medications and other health care services as well as lost productivity The comparator agent chosen for this evaluation was the beta blocker metoprolol 200 mg daily, which is widely used off-label for migraine prevention and was found to have the most favorable side-effect profile in a recent systematic review of preventive therapies.81 We estimated the proportion of patients responding to Cefaly treatment to be 38.2%, based on data from the RCT,54 while a similar proportion (39.9%) for metoprolol was obtained from a meta- analysis of studies in the previously-mentioned systematic review. 81 No adverse events leading to discontinuation were reported in the Cefaly RCT or a safety study conducted in over 2,300 device recipients.54,56 The discontinuation rate due to side effects for metoprolol was assumed to be 1% based on supplemental data from the systematic review of preventive therapies. 81 We estimated the cost of the Cefaly device based on the published price for the device ($299) and six electrode kits ($150) to cover one year of therapy, resulting in a total price of $449.118 The cost ©Institute for Clinical and Economic Review, 2014 Page 39 of metoprolol was estimated to be approximately $50 annually, based on a published Medi-Cal price of $0.07 per 100 mg tablet. As with the SpringTMS analysis, patients discontinuing metoprolol were assumed to incur the cost of one month of therapy but receive no clinical benefit. The cost of care for episodic migraines was estimated to be slightly more than $2,000 annually, based on a previously-described analysis of resource utilization data from the AMPP.106 In comparison to the effects of SpringTMS, which are primarily on pain and not on other migraine symptoms, treatment response in the Cefaly model would mean a reduction in the number of migraine attacks. Accordingly, our estimates of reductions in the other costs of migraine management were more aggressive. Specifically, treatment response in this instance would eliminate the need for additional preventive therapies and would reduce all other costs by 50%, resulting in an annual cost estimate of approximately $1,000. A summary of key assumptions for the Cefaly model is below: • Patients discontinuing metoprolol would incur costs of one month of drug therapy but receive no clinical benefit • No Cefaly user would discontinue due to adverse events • Patients responding to either treatment would eliminate the need for other preventive medications and have 50% reductions in other costs of episodic migraine management Model Results Findings from the Cefaly model are presented in Table 9 on the following page. In a hypothetical cohort of 1,000 patients, 382 and 395 patients would be expected to respond to Cefaly and metoprolol treatment respectively after accounting for metoprolol discontinuation. Costs of intervention and migraine management are both higher with Cefaly given the slightly inferior performance of the device, yielding total excess costs of over $400,000 among Cefaly patients. No cost-effectiveness ratio for Cefaly could be calculated due to higher costs and lower effectiveness. Given that we conducted an indirect comparison based on limited data, we also evaluated several thresholds of potential performance for Cefaly. First, if effectiveness was assumed to be equivalent between the two therapies but the discontinuation rate remained higher for metoprolol, an additional four responders per 1,000 would be obtained with Cefaly, and the cost per treatment responder would be approximately $99,000. If Cefaly is assumed to be 5% more effective than metoprolol (41.9% response rate), the cost per treatment responder drops to about $15,500. Increasing levels of assumed incremental effectiveness would further reduce the cost-effectiveness ratios, and at current reimbursement levels, Cefaly would be cost-saving relative to metoprolol at a treatment response rate of 73%. We also examined Cefaly’s cost-effectiveness at the 41.9% response rate and different assumed prices for the device. In this scenario, Cefaly would become cost-saving at a drop in total device/electrode price from $449 to $76. ©Institute for Clinical and Economic Review, 2014 Page 40 Table 9. One-year Outcomes and Costs of Cefaly and Metoprolol Prevention among 1,000 Hypothetical Patients with Episodic Migraine Outcome/Cost Cefaly Metoprolol Difference (Cefaly- Metoprolol) Treatment response (n) Responders 382 395 (13) Non-responders 618 605 Costs ($) Intervention $449,000 $49,225 $399,775 Other migraine mgmt $1,770,053 $1,754,691 $15,363 Total $2,219,053 $1,804,371 $415,138 Cost per treatment response ($) Less effective, more expensive Finally, because the one-time purchase of the Cefaly device might provide benefits over longer periods of follow-up, we also examined the cost-effectiveness over a two-year time horizon. We assumed that initial response rates would persist throughout the time period, that Cefaly would have a 5% higher response rate than metoprolol, and that second-year costs for Cefaly would be limited to the electrode sets alone ($150). As with the SpringTMS analysis, both outcomes and costs were discounted by 3.5% in the second year. The cost per treatment responder in this analysis was approximately $9,000 (vs. $15,500 in the one-year model). 7.4 Costs and Cost-effectiveness of BOTOX for Prevention of Chronic Migraine We examined the potential costs and cost-effectiveness of BOTOX in a hypothetical cohort of 1,000 patients with chronic migraine. Major model inputs can be found in Table 10 on the following page. Clinical effectiveness was measured in terms of reductions in the number of headache days per month, consistent with one of the major endpoints of the Phase III trials. Because these RCTs involved comparisons to placebo (i.e., sham injection), we compared the costs and outcomes of BOTOX to those of sham injection as well as to no treatment since several payer policies cover BOTOX only after failure of multiple pharmacologic alternatives. As noted in the evidence review, individual studies directly comparing BOTOX to other drug treatments were few in number and underpowered for the outcomes of primary interest. In addition, we were alerted to the fact that the RCT comparing botulinum toxin A and amitriptyline76 that we used as a source for an analysis in the draft report actually utilized another form of the toxin (Dysport®, Ipsen Biopharm Ltd.) that is ©Institute for Clinical and Economic Review, 2014 Page 41 not approved for migraine prevention. As such, we have removed this comparison from the final report. In primary analyses, chronic migraine was associated with a headache frequency of 20 days per month, consistent with baseline characteristics from the two large Phase III trials of BOTOX.74,75 We estimated reductions in headache days relative to placebo (2.3 days per month) based on findings from a meta-analysis of these and other BOTOX RCTs in chronic migraine.80 Of note, however, placebo injections also reduce headache frequency; for example, reductions were 9 and 6.7 days for BOTOX and placebo in the Jackson meta-analysis. The percentage of headache days reduced at the base level of 20 (i.e., 2.3/20 or 11.5%) was also applied to the threshold headache frequency for chronic migraine (15 days per month) to obtain an estimate of the absolute reduction vs. placebo (1.7 days). Table 10. Key Parameter Estimates for BOTOX Model* Parameter Estimate Source(s) Number of headache days per 20.0 Diener, 2010; Aurora, 2010 month Headache days averted for BOTOX vs. placebo, per month @20 days/month 2.3 Jackson, 2012 @15 days/month 1.7 Derived D/C due to side effects (%) BOTOX 3.8 Diener, 2010; Aurora, 2010 Placebo 0.0 Assumption Intervention costs ($) BOTOX (per injection / per year) $1,198 / $4,793 Medi-Cal Injectable Drug List Placebo $0 Assumption (varied in sensitivity analyses) † Costs of chronic migraine ($) Per year $9,800 Munakata, 2009 Per headache day (assume 20/mo) $41 Derived D/C: Discontinue * Revised to include correct cost of Botox † Includes cost of medications and other health care services as well as lost productivity Discontinuation due to side effects was assumed to be 0% for placebo. We estimated a rate of 3.8% for BOTOX based on pooled data from the two Phase III trials.74,75 As with the device models ©Institute for Clinical and Economic Review, 2014 Page 42 described previously, patients who discontinued were assumed to do so after one BOTOX injection, and it was also assumed they did not achieve any reduction in headache frequency before discontinuing. The annual cost of migraine was calculated using an estimate of the costs of health care services and productivity loss for chronic migraine ($9,800) from the previously described AMPP-based study.106 We estimated a cost per headache day of $41 based on this amount and the assumed base frequency of 20 headaches per month. Any reductions in the number of days of headache were multiplied by $41 to estimate the potential cost offsets associated with therapy. Costs of therapy were obtained from publicly-available Medi-Cal documents for injectable medications.107,108 BOTOX is reimbursed at $9.91 per unit for the first unit, which includes a $4.46 administrative fee. The remaining units are reimbursed at a rate of $5.45 per unit. In addition to drug reimbursement, physicians are also paid a fee for the injection itself, which is estimated to be $104 per injection.107 Based on these inputs, use of two 100-unit vials for each injection and an injection frequency of once every three months yielded an annual cost estimate of $4,793 ($1,198 per injection). We assumed no cost for placebo in primary analyses but did address this in sensitivity analyses. The cost-effectiveness of BOTOX was expressed as the incremental cost per headache day averted and was calculated relative to both placebo injections and to no therapy. A summary of key assumptions for the BOTOX model is below: • Reductions in the number of headache days per month resulted in offsets to the cost of each headache day (see above); no other effects were assumed • Patients discontinuing BOTOX due to side effects were assumed to have one injection before discontinuing and to experience no reductions in headache frequency Model Results Findings for the comparison of BOTOX to placebo (sham) injections among 1,000 hypothetical patients with chronic migraines can be found in Table 11 on the following page. Because sham injection also results in reductions in headache frequency, the numbers of headache days without any intervention are also presented for reference. At a headache frequency of 20 days a month, use of no intervention would result in 240,000 headache days annually for the cohort. Use of placebo injections would reduce the number of days by one-third, to approximately 160,000. BOTOX injections would reduce the number of days by 43%, to approximately 136,000. As headache frequency further declines, however, the absolute number of days of headache averted also decline. For example, at a headache frequency of 20 days per month, BOTOX saves over 23,000 days vs. placebo. Savings at a frequency of 15 headache days per month are reduced to approximately 18,000. ©Institute for Clinical and Economic Review, 2014 Page 43 The costs of BOTOX over one year are estimated to total $4.7 million in the 1,000-person cohort, while the costs of managing chronic migraine without intervention would range between $7.3 million and $9.8 million depending on headache frequency. At the baseline frequency of 20 headaches per month, BOTOX and placebo injections would reduce chronic migraine costs by approximately $4.2 and $3.3 million respectively, resulting in savings of nearly $1 million for BOTOX vs. placebo. Savings decline to approximately $700,000 at a headache frequency of 15 days per month. Table 11. One-year Outcomes and Costs of BOTOX and Placebo Migraine Prevention among 1,000 Hypothetical Patients with Chronic Migraine* Outcome/Cost No Therapy BOTOX Placebo Difference (BOTOX-Placebo) Headache days/yr @20/mo 240,000 136,104 159,600 (23,496) @15/mo 180,000 102,078 119,700 (17,622) Intervention cost ($) --- $4,656,710 --- $4,656,710 Costs of other migraine mgmt ($) @20/mo $9,799,913 $5,557,531 $6,516,942 ($959,411) @15/mo $7,349,935 $4,168,148 $4,887,707 ($719,559) Total costs ($) @20/mo $9,799,913 $10,214,241 $6,516,942 $3,697,299 @15/mo $7,349,935 $8,824,858 $4,887,707 $3,937,152 * Revised to include correct cost of Botox Total costs for BOTOX are estimated to range from $9-$10 million depending on headache frequency, cost increases over no treatment range from $400,000-$1.5 million, and cost increases over placebo range from $3.7-$4 million. Figure 4 on the following page depicts our estimates of the cost-effectiveness of BOTOX vs. no therapy. At the baseline frequency of 20 headaches per month, the incremental cost per headache day averted would be only $4, as the intervention costs are nearly completely offset by reductions in the costs of chronic migraine management. At a headache frequency of 15 days per month, the cost per headache day averted rises to $19. In comparisons of BOTOX to placebo injection, cost- effectiveness ratios were higher ($157 and $223 per headache day averted for 20 and 15 headache ©Institute for Clinical and Economic Review, 2014 Page 44 days per month, respectively) due to the potential for reductions in migraine management costs from significant placebo effects. Figure 4. Cost per Headache Day Averted for BOTOX vs. Placebo Migraine Prevention and to No Therapy, by Monthly Headache Frequency* $250 $223 vs. Placebo $200 vs. No therapy $157 $150 $100 $50 $19 $4 $0 20 15 Number of Headaches per Month * Revised to include correct cost of Botox We conducted a number of alternative analyses to further explore the effects of BOTOX vs. no therapy. At a headache frequency of 20 per month, a reduction of one additional headache day relative to no therapy (i.e., from 9 fewer headache days per month to 10 fewer headache days) would represent the threshold at which BOTOX would be cost-saving. At a frequency of 15 headache days, however, a reduction of an additional three days (i.e., from 7 to 10 days) would be required for overall cost savings. As noted previously, we did not include an estimate of the cost of placebo injections in our primary analysis. If the cost were assumed to be equivalent to the injection fee ($104 per injection), BOTOX would need to reduce headache frequency from 20 to 4 headache days per month (i.e., an additional 7 days above the primary estimate) to achieve cost neutrality vs. placebo. ©Institute for Clinical and Economic Review, 2014 Page 45 7.5 Economic Burden of Opioids and Potential Savings from Reduced Use in the Emergency Department The purpose of the opioid model was twofold: (1) to document the overall breadth of opioid use and dependence among migraineurs in California as well as the medical and societal impacts of widespread use, and (2) to estimate the potential cost savings from a focused attempt to reduce the use of opioids in the ED. The timeframe of interest for this evaluation was one year. Major model inputs are provided in Table 12 on the following page. The population of interest was all persons aged 12 and older in California, based on information from the US Census Bureau.109 We estimated the prevalence of both episodic and chronic migraines based on a review of data from multiple national population-based surveys;110 prevalence was stratified both by sex and by age (12- 17 vs. 18+). The prevalence of opioid use among migraineurs was estimated based on findings from the AMPP Study, which was further stratified according to whether patients likely did or did not meet criteria for opioid dependence.111 Among opioid-using patients, primary outcomes of interest in the model included (a) the incidence and costs of transformation from episodic to chronic migraine, and (b) the costs of both incident and prevalent cases of opioid dependence. Data on the incidence of migraine transformation as well as the increased risk of this outcome among opioid users was obtained from a subset analysis of the AMPP.43 The difference in annual costs between episodic and chronic migraines, which included health care and productivity-loss costs, was estimated to be approximately $7,500, again based on data on from the AMPP-based study employed in the treatment-specific model.106 The annual incidence of new cases of opioid dependence among opioid users was estimated to be 0.6%, based on data from the National Epidemiologic Survey on Alcohol and Related Conditions (NESARC), a longitudinal and representative US-based survey.112 Both incident and prevalent cases of opioid dependence were assumed to incur the costs of opioid dependence, which were calculated to be approximately $28,000 annually, based on estimates of increased health care expenditures and productivity losses.113,114 ©Institute for Clinical and Economic Review, 2014 Page 46 Table 12. Key Parameter Estimates for Opioid-use Model Parameter Estimate Source(s) Migraine prevalence (%) Episodic Chronic Smitherman, 2013 Female Adolescent 7.1 0.2 Adult 15.8 1.3 Male Adolescent 4.8 0.1 Adult 5.1 0.5 Opioid use (%) Buse, 2011 Female Nondependent 13.9 Dependent 2.5 Male Nondependent 10.6 Dependent 3.2 Transformation (%) Bigal, 2008 Base rate 2.5 Rate with opioid use: men* 7.0 Rate with opioid use: women* 3.3 Opioid dependence (%) 0.6 Katz, 2013 Opioid use in ED (%) 53.0 West, 2014 Annual migraine costs ($) Episodic $2,221 Munakata, 2009 Chronic $9,800 Munakata, 2009 Combined, health care only $2,571 Hawkins, 2008 † Among opioid users $6,428 Xie, 2013 Opioid dependence costs ($) Health care $20,111 White, 2005 Productivity loss $7,878 Wall, 2000 ED: Emergency department *Based on sex-specific odds ratios for transformation among opioid users †Based on multiplier (2.5) of health care costs among patients receiving opioids in ED/inpatient settings vs. not Estimates of potential savings in the ED setting came from separate sources. The annual number of migraine-related ED visits was estimated to total approximately 117,000 based on data submitted to the California Office of Statewide Health Planning and Development.115 Of these, 53% were assumed to involve use of an opioid based on recently-presented data from the 2010 round of the National Hospital Ambulatory Medical Care Survey (NHAMCS).43 Importantly, we did not model ©Institute for Clinical and Economic Review, 2014 Page 47 potential savings from reduced use of opioids in the ED based on reductions in migraine transformation or opioid dependence, as the available ED data does not distinguish between dependent patients seeking opioids in the ED, those not dependent but regularly using opioids, and those receiving opioids for the first time. We chose to estimate potential savings more conservatively based on estimates from observational studies of the annual attributable health care costs of migraine (~$2,600)116 and the increase in health care expenditures (2.5-fold) among patients receiving opioids in an inpatient or ED setting vs. those without such receipt.117 A summary of key assumptions for the opioid model is below: • The incidence of transformation was calculated among patients with episodic migraine only • The incidence of opioid dependence was calculated among nondependent opioid users only • Both incident and prevalent cases of opioid dependence received full costs of opioid dependence • Other social costs of dependence (e.g., law enforcement, victimization) were not included, as opioids were assumed to be obtained through legal channels in this analysis • The reported number of ED encounters was assumed to be equivalent to the number of migraine patients visiting the ED (i.e., one encounter per patient on average) Model Results The results of the model depicting the burden of opioid use among migraineurs in California are shown in Table 13 on the next page. Of the approximately 32 million individuals in the state age 12 or older, we estimate that approximately 3.5 million would have migraine; of these, approximately 250,000 would be adolescents. Seventy-five percent of migraineurs would be female, and 93% would have the episodic form of the disease. Among patients with migraine, approximately 16% (~550,000) would be expected to be using opioids, over 90,000 of whom would meet criteria for dependence. We further estimate that opioid use among patients with episodic migraines would result in approximately 21,000 new cases of transformation to the chronic form, with associated costs of nearly $160 million. New cases of dependence among nondependent opioid users would total nearly 3,000. Taken together with prevalent cases, we estimate opioid dependence in the California migraine population to total approximately 96,000 persons over one year at a cost of almost $28,000 per person, yielding associated excess costs of nearly $2.7 billion. The total economic burden of opioid use in this population, including the costs of transformation and dependence, is estimated to be $2.8 billion. ©Institute for Clinical and Economic Review, 2014 Page 48 Table 13. Clinical and Economic Burden of Opioid Use in the California Migraine Population Female Male Total Grand Estimate (N or $) Adolescent Adult Adolescent Adult Adolescent Adult Total Population (CA) 2,063,701 14,231,394 2,171,510 13,810,868 4,235,211 28,042,262 32,277,473 Migraine Episodic 147,185 2,249,983 103,957 707,116 251,142 2,957,100 3,208,242 Chronic 3,465 183,585 2,447 66,292 5,912 249,877 255,789 Total 150,650 2,433,568 106,404 773,409 257,054 3,206,977 3,464,031 Opioid Use Nondependent 20,909 337,751 11,313 82,232 32,222 419,984 452,205 Dependent 3,778 61,033 3,414 24,811 7,192 85,845 93,036 Total 24,687 398,785 14,727 107,044 39,414 505,828 545,242 Among Episodic Pts 24,119 368,701 14,388 97,869 38,507 466,569 505,076 Transformation to Chronic Migraine Incident Cases 784 11,983 1,007 6,851 1,791 18,834 20,625 Excess Costs $ 5,940,755 $ 90,814,828 $ 7,633,119 $ 51,920,697 $ 13,573,875 $ 142,735,526 $ 156,309,400 Dependence Prevalent Cases 3,778 61,033 3,414 24,811 7,192 85,845 93,036 Incident Cases 132 2,128 71 518 203 2,646 2,849 Total 3,910 63,161 3,485 25,330 7,395 88,491 95,885 Excess Costs $ 109,439,175 $ 1,767,855,357 $ 97,537,946 $ 708,964,833 $ 206,977,121 $ 2,476,820,190 $ 2,683,797,311 Total Costs $ 115,379,930 $ 1,858,670,186 $ 105,171,065 $ 760,885,530 $ 220,550,996 $ 2,619,555,716 $ 2,840,106,712 Statistics for opioid use in the ED are presented in Figure 5 on the next page. We estimate that there would be a total of 116,696 ED encounters for migraine over one year in California; of these, 61,849 would involve receipt of opioid medications during the encounter. Based on estimates of annual health care costs among patients receiving opioids, the total costs of care for migraine patients receiving opioids in the ED are estimated to be nearly $400 million over one year. Reducing opioid use from the national average of 53% to 25% of encounters (from 61,849 to 29,174 annually) would potentially reduce health care costs by $126 million, a reduction in use to 10% (11,670 encounters) could cut costs by nearly half (to $204 million), and a reduction to 5% could reduce costs by nearly $250 million, down to $159 million. ©Institute for Clinical and Economic Review, 2014 Page 49 Figure 5. Annual Health Care Costs at Multiple Levels of Opioid Use in the ED Baseline (53% Opioid Use) 25% Opioid Use 10% Opioid Use 5% Opioid Use $- $50 $100 $150 $200 $250 $300 $350 $400 Millions 7.6 Summary While the economic analyses described in this section had different goals and methods, their implications are important to summarize. First and perhaps foremost, the prevalence and burden of opioid use is substantial in the California migraine population. We estimate that there are approximately 3.5 million current migraineurs in the state, of whom more than half a million are currently using opioids. This widespread use is estimated to result in nearly 21,000 instances of transformation from episodic to chronic migraines over one year, which would generate nearly $160 million in additional expenditures for health care services and lost productivity during that period. More importantly, over 90,000 migraineurs in the state may already be dependent on opioids, and an additional 3,000 may become dependent over one year. The costs of dependence in these individuals total $2.7 billion, even without considering the costs of criminal activity in which some of these dependent individuals might engage. Certainly, the use of parenteral opioids in the ED, which are employed in over half of all ED encounters for migraine nationally, is a major contributor to these costs. It is difficult to estimate the portion of opioid-associated costs that are generated in the ED as there are no data on the number of patients who are dependent, chronic users but not dependent, and first-time users in this setting. Nevertheless, the excess medical-care costs alone in patients receiving opioids in the ED are estimated to total nearly $400 million statewide; a reduction in opioid use in the ED to 25% of encounters could save over $125 million. ©Institute for Clinical and Economic Review, 2014 Page 50 Our findings with respect to BOTOX are also interesting. As noted in the evidence review, BOTOX provides modest incremental benefit for chronic migraines relative to sham injection (i.e., reductions of 2.3 headache days per month), but the absolute benefits seen with the sham injections themselves are almost three times this amount. Nevertheless, the cost per headache day averted compared to no treatment is relatively low, ranging from $4-$19 depending on headache frequency. Ratios compared to placebo are higher given the smaller incremental benefit, however, ranging from $157-$223 per headache day averted. As a point of reference, estimates of the cost per headache day averted for other preventive therapies in chronic migraine vs. placebo (i.e., topiramate, divalproex, gabapentin) have ranged from $48-$138 in other studies.119,120 Findings such as these may be important inputs into payer coverage policies that address appropriate candidates for BOTOX therapy. Our findings with respect to the Cefaly and SpringTMS devices are subject to greater uncertainty given the paucity of available evidence as well as study quality concerns with the Cefaly RCT. Our model results indicate that treatment response (defined as ≥50% reduction in monthly headache frequency) with Cefaly is inferior to that of metoprolol, a commonly-used and generically-available drug for prevention of episodic migraines. If Cefaly were to be 5% more effective than metoprolol, the cost per treatment response would be nearly $16,000. This stands in stark contrast to other comparisons of active treatments for migraine prevention as mentioned above ($48-$138). At the reimbursement levels used in this analysis, Cefaly would not be cost-neutral relative to metoprolol unless its response rate is essentially doubled relative to the RCT results (i.e., from 38% to 73%) or if the startup cost was reduced by 85% (from $449 to $76). The SpringTMS trial was of higher quality and showed a 24-hour pain-free response rate of 29%, which is higher than the response rate seen with sumatriptan (20%), a longstanding therapy that is available generically. However, even after accounting for greater discontinuation due to side effects with sumatriptan, our analysis suggests a cost per treatment response of over $27,000 based on an estimate of the likely monthly rental fee ($250). Again, prior comparisons of active treatments for episodic migraines suggest incremental costs per pain-free response of $38-$57 in a comparison of almotriptan to ergotamine,121 $7 in a comparison of almotriptan vs. rizatriptan,122 and $20-$38 in a comparison of aspirin/triptan treatment stratified by headache severity to stepped care for all patients.123 Our model further suggests that SpringTMS would only be cost-neutral at monthly rental rates of $20-$25. We note some limitations of our analyses. Wherever possible, we attempted to make our cost inputs relevant to the price Medi-Cal pays for treatment and other services, consistent with the perspective that we ask the CTAF Panel to adopt. This was not always possible, however, so we used available literature-based estimates to fill these gaps. ©Institute for Clinical and Economic Review, 2014 Page 51 In addition, we used a short-term time horizon for all models; estimates of outcome and cost may have differed with longer time horizons. For example, we note that the Cefaly device has the potential to be a “one-time” purchase versus the need for chronic drug therapy as a comparator. However, follow-on or maintenance costs would continue to produce excess costs over generic drug therapy. Finally, we note that some of our analyses were intended to promote discussion rather than produce definitive results. For example, our analysis of BOTOX’s economic performance at a lower level of migraine headache frequency (15 days per month) was not based on any RCT findings but was instead an extrapolation of data from the large Phase III trials. **** This is the first review of this technology by the California Technology Assessment Forum. ©Institute for Clinical and Economic Review, 2014 Page 52 8. CTAF Voting Questions and Discussion Introduction During CTAF public meetings, the CTAF Panel deliberates and votes on key questions related to the systematic review of the clinical evidence, a cost analysis of the applications of the medical technologies or treatments under examination, and the supplementary information presented. At the July 11, 2014 meeting, the CTAF Panel discussed issues regarding the application of the available evidence to help patients, providers, and payers address important questions related to controversies in migraine management. The key questions are developed by the research team for each evidence review, with input from the CTAF Advisory Board to ensure that the questions are framed to address the issues that are most important in applying the evidence to support clinical practice and medical policy decisions. Following the evidence presentation and public comments, the CTAF Panel voted on questions concerning the comparative clinical effectiveness and comparative value of four treatments for migraine headache. The voting results are presented below, along with comments reflecting considerations mentioned by CTAF Panel members during the voting process. When voting on comparative value, the CTAF Panel was asked to assume the perspective of Medi- Cal (the state Medicaid program) or a provider organization that must make resource decisions within a relatively fixed budget for care. The CTAF Panel is not given prescribed boundaries or thresholds for budget impact or incremental cost-effectiveness ratios to guide its judgment of low, reasonable/comparable, or high value. However, the CTAF Panel did make use of a value framework designed for the CTAF process with different categories of evidence on effectiveness and cost to assist the CTAF Panel in assigning an overall value rating of low, reasonable/comparable, or high value (see Figure 6 on the following page). Because all of the voting questions asked whether a particular drug or device was equivalent to or better than a comparator, the CTAF Panel did not have the option to vote for two of the categories shown in the value matrix in the Executive Summary of this report, as these categories refer to a drug or device that has “worse outcomes”. ©Institute for Clinical and Economic Review, 2014 Page 53 Figure 6. Evidence Categories for Ratings of Low, Reasonable/Comparable, and High Value Low Value Reasonable/Comparable Value High Value 7. Comparable outcomes; Lower cost 1. Comparable outcomes; 4. Comparable outcomes; 8. Promising but inconclusive Higher cost Comparable cost evidence of better outcomes; Lower cost 2. Promising but inconclusive 5. Promising but inconclusive 9. Better outcomes; evidence of better evidence of better Lower or comparable cost outcomes; outcomes; Higher cost Comparable cost 3. Better outcomes; 6. Better outcomes; 10. Better outcomes; Too high a cost Reasonable higher cost Slightly higher cost * For comparisons of one drug or a set of drugs to another drug or set of drugs, the term “comparable” is used in the value assessment; for comparisons of one drug or a set of drugs to no treatment, the term “reasonable” is used in the value assessment. 8.1 Summary of the Votes and Considerations for Policy 1. For the acute treatment of migraine with aura, is the evidence adequate to demonstrate that the net health benefits of transcranial magnetic stimulation (SpringTMS) are equivalent to or better than those of other standard acute treatment medications? CTAF Panel Vote: 2 yes 9 no The CTAF Panel noted that the study of SpringTMS was of moderate size and sponsored by industry. They also stated that the study did not show an impact on other important outcomes that may be of interest to patients (such as nausea and vomiting) and that there was no information provided on disability or the use of rescue medications. The CTAF Panel thought that a study showing effectiveness (real-world experience), including information on benefits and harms, in large, diverse populations would be particularly useful, noting that the currently available information is on efficacy only (effects observed under study conditions). ©Institute for Clinical and Economic Review, 2014 Page 54 2. For the prevention of episodic migraine, is the evidence adequate to demonstrate that the net health benefits of the Cefaly device are equivalent to or better than those of usual care with preventive medications? CTAF Panel Vote: 1 yes 10 no The CTAF Panel expressed a concern about potential unblinding of patients in randomized trials of the Cefaly device versus sham device and suggested that future studies should be larger and include a question about whether the study participant thought they had an active device. 3. For patients who have inadequate relief with other preventive therapies for chronic migraine, is the evidence adequate to demonstrate that the net health benefits of BOTOX injections used on an every 12-week schedule are better than no treatment? CTAF Panel Vote: 11 yes 3a. If yes, what is the comparative value of BOTOX injections vs. no treatment? CTAF Panel Vote: 7 low 3 reasonable 1 high Low value votes: “Promising but inconclusive evidence of better outcomes and higher cost”: 5 “Better outcomes at too high a cost”: 2 Reasonable value votes: “Better outcomes at reasonable higher cost”: 3 High value votes: “Better outcomes at slightly higher cost”: 1 When asked to comment on the distribution of value votes, one CTAF Panel member suggested that there is no standard way to interpret “too high a cost” and that this may account for some of the variation in response. Also discussed were concerns about potential unblinding in the PREEMPT trials, as it was not reported if study participants were asked whether they thought they had received active or placebo injections. 4. For patients who are considering multiple therapeutic options for chronic migraine, is the evidence adequate to demonstrate that the net health benefits of BOTOX injections used on an every 12-week schedule are equivalent to or better than those of other preventive therapies? CTAF Panel Vote: 1 yes 10 no ©Institute for Clinical and Economic Review, 2014 Page 55 Neurologists on the CTAF Panel and policy roundtable indicated that they typically have a conversation with migraine patients about therapeutic options (e.g., tolerability, safety, cost, co-morbidities, patient preferences) prior to injecting BOTOX and suggested that experts in the management of frequent migraines may wish to retry drugs, and classes of drugs, used previously. The patient on the policy roundtable said that there are some patients for whom other medications are not viable options (possibly due to allergies or intolerance), and so BOTOX should be considered. There was a call for larger RCTs involving the direct comparison of BOTOX using the appropriate protocol to other treatment options such as topiramate, placebo injections, placebo pills, and non-medicinal therapies to allow for assessment of effectiveness of the various options. It was also suggested that blood samples be collected from patients so that genotyping could be used to predict which patients would respond best to various therapies, allowing for tailored treatment. 5. For the acute treatment of migraine in the emergency department, is the evidence adequate to demonstrate that the net health benefits of parenteral opioids as first-line therapy are inferior to those of non-opioid alternatives? CTAF Panel Vote: 9 yes 2 no Comments were made that even though opioids generally offer more harm than benefit compared to other options, some patients will require opioids given prior histories of inadequate response to or adverse effects with other treatments. It was also noted that patients who present to the ED for pain may have access issues or disease that is more challenging. There was a call for more sophisticated research involving chart reviews and electronic medical records (EMRs) in order to determine the appropriateness of opioid use in individual patients. 8.2 Policy Roundtable Discussion and Key Policy Recommendations Following its deliberation on the evidence and subsequent voting, the CTAF Panel engaged in a moderated discussion with a policy roundtable composed of clinical experts, a patient advocate, and payer representatives. A list of the participants on the policy roundtable is shown below. • Sylvia Carlisle, MD, MBA, Managing Medical Director, Anthem Blue Cross • Robert Cowan, MD, FAAN, Professor of Neurology and Director of the Headache Program, Stanford University • Lynne McCullough, MD, Medical Director, Ronald Reagan UCLA Medical Center Emergency Department, and Clinical Professor of Medicine and Emergency Medicine, UCLA • Sonja Potrebic, MD, PhD, Headache Specialist, Kaiser Permanente Los Angeles Medical Center ©Institute for Clinical and Economic Review, 2014 Page 56 • Ellen Schnakenberg, patient educator and advocate • Sam Torbati, MD, Co-chair and Medical Director, Department of Emergency Medicine, Cedars-Sinai Medical Center • John Yao, MD, MBA, MPH, Staff Vice President of Medical Policy Development, WellPoint The roundtable discussion explored the implications of CTAF Panel votes for clinical practice and medical policy, considered real life issues critical for developing best practice recommendations in this area, and identified potential avenues for applying the evidence to improve patient care within a value context. The main themes and recommendations from the discussion are summarized below. 1) The CTAF Panel voted that the evidence is currently inadequate to demonstrate the effectiveness of the Cefaly and SpringTMS devices to prevent and treat migraine, respectively. Because there is strong interest in non-pharmaceutical migraine treatment alternatives, both on their own and as a complement to pharmaceutical management, further study of these two devices is needed to generate more evidence of the effectiveness of each device versus sham devices and the comparative effectiveness of each device to other active therapies. While noting that the Cefaly and SpringTMS devices need further study, the patient on the policy roundtable described strong patient interest in the devices as another potential tool to manage migraine symptoms. 2) For patients who have inadequate relief with other preventive therapies for chronic migraine, the CTAF Panel voted unanimously that the evidence demonstrates that the net health benefits of BOTOX injections are better than no treatment. In their prior authorization criteria, several commercial insurers currently require step therapy with one or more oral medications prior to the approval of BOTOX for chronic migraine prevention, and this seems reasonable given the large placebo effect reported in the clinical trials of BOTOX and its cost. However, no other drug is FDA approved specifically for the prevention of chronic migraine, and expert opinions on the CTAF Panel and policy roundtable were mixed on step therapy requirements. In addition, some of the headache specialists on the policy roundtable said that they are often referred patients who have used oral medications for an insufficient time period or at insufficient doses to accurately assess their effectiveness, so clinicians may wish to explore whether re-trying an oral medication is clinically appropriate before injecting patients with BOTOX. Clinicians on the policy roundtable noted that when discussing treatment options with a patient, they consider effectiveness to include overall side effects, tolerability, and invasiveness of different therapies. Policy roundtable participants also acknowledged that while BOTOX injection therapy clearly results in excellent results for some chronic migraineurs, it doesn’t for others, thus raising the option of using specific “stopping rules” (i.e., cessation of the drug after one to two injection ©Institute for Clinical and Economic Review, 2014 Page 57 cycles if no benefit is observed). Such approaches should be defined and tested in clinical studies, and the results of the long-term observational registries are eagerly anticipated. 3) Given that patients with migraine can respond so differently to a specific drug and that many factors can affect a clinician’s drug choice for an individual patient, insurers may wish to consider adding more choices within a particular drug class such as triptans and to allow clinicians to try and retry a variety of preventive and abortive agents. Clinicians on the policy roundtable highlighted that migraine still is not a well-understood disease and that current classifications – episodic vs. chronic, with or without aura, etc. – may not be sufficient to capture its diversity. In light of these challenges, they indicated that it is important to consider multiple factors that potentially affect a patient’s migraines (e.g., their work, family situation, diet, stress) when choosing a drug therapy, and that they typically try oral agents first when treating migraine. Covering more drugs within a specific class may increase the likelihood of identifying an effective treatment and lead to improved patient outcomes. 4) In contrast to current practice, opioids should very rarely be used to treat migraine in the ED. Multiple initiatives are needed to reduce the use of opioids for migraine in the ED. a. Specialty societies should work collaboratively to make a strong statement about the harms of using opioids for migraine pain relief in the ED and ensure that clinicians are supported in efforts to reduce opioid use. b. Data on the use of opioids to treat migraines in the ED should be analyzed and feedback provided to emergency departments, in combination with education about evidence-based practice. c. Electronic systems with order sets that encourage ED physicians to select non-opioid alternatives to treat migraine symptoms should be implemented. d. Migraine patients and their clinicians should develop a written treatment plan for rescue medications that can be used by patients at home and also made readily available to guide care in urgent care and emergency care settings. e. Purchasers, insurers, and other policymakers should consider identifying reduction of opioid use in the ED as a specific target for quality improvement initiatives in collaboration with clinicians and patient groups. National data show that about 50% of patients with migraine symptoms who present to an ED receive opioids, and there is no reason to expect that patients in California receive opioids at a higher or lower rate than reported nationally. The research literature and the participants on the policy roundtable agreed that opioids should be used sparingly to relieve migraine pain and noted ©Institute for Clinical and Economic Review, 2014 Page 58 that many changes would be needed to change the culture around narcotic prescribing for migraine symptoms in the ED. Specific recommendations include: a) The relevant professional societies (e.g., International Headache Society, American Academy of Neurology, American College of Emergency Physicians) should take the lead by making a strong statement about the harms of using opioids for relief of migraine pain in the ED. b) Health plans, delivery systems, and EDs should gather and analyze data to assess opioid prescribing practices for migraine pain in California EDs. To change prescribing behavior, feedback should be provided on the use of opioids to treat migraine symptoms at both the hospital and individual physician levels. This should be followed by academic detailing in which clinicians would be educated about changing practice to be consistent with medical evidence on the use of opioids to treat migraines. c) EDs should develop standard order sets that encourage ED physicians to select non- opioid alternatives and require explanation or justification before they permit an override to allow opioid use for migraine symptom relief. d) Migraine patients and their primary care and specialty clinicians should work together to develop a written treatment plan that patients can take with them when they need urgent or emergency care. Ideally, these plans would include a statement about avoiding opioids except as an acute treatment of last resort. e) In considering possible quality improvement initiatives, purchasers, insurers, and other policymakers may wish to focus on measuring and reducing the use of opioids in the ED. Congruence between the patient treatment plans and the actual treatments rendered might be used as a quality improvement metric for EDs. 5) Patients and providers should partner to better understand migraine triggers and potential therapies to ensure that patients with migraines get the right treatment in the right place as quickly as possible. Because they are more likely to see patients when they are not in acute distress, primary care and specialty clinicians have the best opportunities to educate patients about what role lifestyle (e.g., diet, sleep, stress) and pharmaceutical interventions may play in migraine treatment (not just pain treatment). Office visits are also an opportune time to discuss what drugs may be safest and most effective for managing migraine pain in urgent/emergency care settings, as well as the harms associated with use of opioids to treat migraine (e.g., conversion of episodic to chronic migraine, migraine overuse headache, potential dependence). Working in partnership with their clinicians, patients may find that lifestyle changes and use of effective pharmaceuticals may lead to fewer headaches, shorter headache duration, or reductions in other symptoms, as well as reduce their need for urgent and emergency care. ©Institute for Clinical and Economic Review, 2014 Page 59 To ensure the best patient care after a visit to urgent or emergency care to treat migraine symptoms, it is important for patients to have quick access through a follow-up visit to primary care and headache specialists/neurologists. As EMRs become more prevalent in clinical settings, there may be increased opportunities for care coordination, which may ultimately improve patient outcomes. 6) Guidelines that encourage the use of opioids to treat pain should be revised to reflect current evidence. In the policy roundtable discussion, it was noted that over the past 20 years, the Medical Board of California established guidelines to improve prescribing practices for effective pain management. The guidelines encouraged clinicians to “view effective pain management as a high priority in all patients” and were designed so that “physicians have a higher level of comfort in using controlled substances, including opioids, in the treatment of pain.” In the discussion with the CTAF Panel, it was suggested that a perception that providers were under-treating pain may have led to the development of guidelines that created an unintended consequence of over-treating pain with opioids, particularly in the ED. 7) Support for basic science research is needed to improve clinical understanding of the biological underpinnings of migraine, which will ultimately improve clinicians’ ability to effectively diagnose patients and identify appropriate treatments. While migraine affects about 12% of people, it is often under-diagnosed and under-treated. Clinical understanding of the basis for migraine headaches is evolving, and there have been recent advances in the understanding of causal factors and migraine triggers, as well as in the definitions of types of migraine (episodic and chronic). Yet, more research is needed to answer the many questions that remain about the exact causes and specific classification of migraines. Without these answers, it is particularly challenging for clinicians to make accurate and precise diagnoses and to identify effective treatments. Ideally, a biomarker specific for the condition will be identified. 8) There is a significant need for higher-quality, larger research studies to show the effectiveness of various migraine therapies, and to compare these agents head-to-head in a comparative effectiveness context. For a condition that affects 12% of the population, relatively little is known about how best to treat migraine. The CTAF panel and policy roundtable participants acknowledged that placebo and nocebo effects presented particular challenges in studying the Cefaly and SpringTMS devices and ©Institute for Clinical and Economic Review, 2014 Page 60 BOTOX, but stated the need for further high-quality, larger studies (ideally RCTs) to adequately assess the clinical effectiveness of these and various other treatments. For the two devices, it would be useful to follow patients and obtain data on the frequency and type of use, as well as whether patients thought they received the active device. For BOTOX, RCTs are needed that directly compare BOTOX to oral agents for the prevention of chronic migraine. In the policy roundtable discussion, it was stated that studies not sponsored by industry would be most compelling, and there was interest in exploring possible support for the design and execution of such studies from other sources such as PCORI. Another important component to future research studies is the use of outcome measures that are meaningful to migraine headache patients. The patient advocate on the policy roundtable pointed out that many of the usual outcome measures used in prevention and treatment trials of migraine agents do not represent what patients are seeking; namely, complete relief from migraines. Thus, even if a treatment meets the conventional bar of a 50% reduction in headache days, patients will continue to look for, or add, therapies to achieve even more relief of headaches. ©Institute for Clinical and Economic Review, 2014 Page 61 References 1. Turner JA, Deyo RA, Loeser JD, Von Korff M, Fordyce WE. The importance of placebo effects in pain treatment and research. JAMA: The Journal of the American Medical Association. May 25 1994;271(20):1609-1614. 2. Lipton RB, Bigal ME, Diamond M, Freitag F, Reed ML, Stewart WF. Migraine prevalence, disease burden, and the need for preventive therapy. Neurology. Jan 30 2007;68(5):343- 349. 3. Lipton RB, Stewart WF, Diamond S, Diamond ML, Reed M. Prevalence and burden of migraine in the United States: data from the American Migraine Study II. Headache. Jul-Aug 2001;41(7):646-657. 4. Stewart WF, Shechter A, Rasmussen BK. Migraine prevalence. A review of population-based studies. Neurology. Jun 1994;44(6 Suppl 4):S17-23. 5. Vos T, Flaxman AD, Naghavi M, et al. Years lived with disability (YLDs) for 1160 sequelae of 289 diseases and injuries 1990-2010: a systematic analysis for the Global Burden of Disease Study 2010. Lancet. Dec 15 2012;380(9859):2163-2196. 6. Devoto M, Lozito A, Staffa G, D'Alessandro R, Sacquegna T, Romeo G. Segregation analysis of migraine in 128 families. Cephalalgia: An International Journal of Headache. Jun 1986;6(2):101-105. 7. Gervil M, Ulrich V, Kyvik KO, Olesen J, Russell MB. Migraine without aura: a population- based twin study. Annals of Neurology. Oct 1999;46(4):606-611. 8. Merikangas KR, Risch NJ, Merikangas JR, Weissman MM, Kidd KK. Migraine and depression: association and familial transmission. Journal of Psychiatric Research. 1988;22(2):119-129. 9. Charles A. Advances in the basic and clinical science of migraine. Annals of Neurology. May 2009;65(5):491-498. 10. Charles A. Vasodilation out of the picture as a cause of migraine headache. Lancet Neurology. May 2013;12(5):419-420. 11. Bolay H, Reuter U, Dunn AK, Huang Z, Boas DA, Moskowitz MA. Intrinsic brain activity triggers trigeminal meningeal afferents in a migraine model. Nature Medicine. Feb 2002;8(2):136-142. 12. Gursoy-Ozdemir Y, Qiu J, Matsuoka N, et al. Cortical spreading depression activates and upregulates MMP-9. The Journal of Clinical Investigation. May 2004;113(10):1447-1455. ©Institute for Clinical and Economic Review, 2014 Page 62 13. Hadjikhani N, Sanchez Del Rio M, Wu O, et al. Mechanisms of migraine aura revealed by functional MRI in human visual cortex. Proceedings of the National Academy of Sciences of the United States of America. Apr 10 2001;98(8):4687-4692. 14. Karatas H, Erdener SE, Gursoy-Ozdemir Y, et al. Spreading depression triggers headache by activating neuronal Panx1 channels. Science (New York, N.Y.). Mar 1 2013;339(6123):1092- 1095. 15. Leao AA. Further observations on the spreading depression of activity in the cerebral cortex. Journal of Neurophysiology. Nov 1947;10(6):409-414. 16. Moskowitz MA, Nozaki K, Kraig RP. Neocortical spreading depression provokes the expression of c-fos protein-like immunoreactivity within trigeminal nucleus caudalis via trigeminovascular mechanisms. The Journal of Neuroscience: The Official Journal of the Society for Neuroscience. Mar 1993;13(3):1167-1177. 17. Olesen J, Burstein R, Ashina M, Tfelt-Hansen P. Origin of pain in migraine: evidence for peripheral sensitisation. Lancet Neurology. Jul 2009;8(7):679-690. 18. Kirchmann M. Migraine with aura: new understanding from clinical epidemiologic studies. Current Opinion in Neurology. Jun 2006;19(3):286-293. 19. Headache Classification Committee of the International Headache S. The International Classification of Headache Disorders, 3rd edition (beta version). Cephalalgia: An International Journal of Headache. Jul 2013;33(9):629-808. 20. Martin VT, Behbehani MM. Toward a rational understanding of migraine trigger factors. The Medical Clinics of North America. Jul 2001;85(4):911-941. 21. Fukui PT, Goncalves TR, Strabelli CG, et al. Trigger factors in migraine patients. Arquivos de Neuro-psiquiatria. Sep 2008;66(3a):494-499. 22. Mansfield LE, Vaughan TR, Waller SF, Haverly RW, Ting S. Food allergy and adult migraine: double-blind and mediator confirmation of an allergic etiology. Annals of Allergy. Aug 1985;55(2):126-129. 23. Robbins L. Precipitating factors in migraine: a retrospective review of 494 patients. Headache. Apr 1994;34(4):214-216. 24. Spierings EL, Ranke AH, Honkoop PC. Precipitating and aggravating factors of migraine versus tension-type headache. Headache. Jun 2001;41(6):554-558. 25. Van den Bergh V, Amery WK, Waelkens J. Trigger factors in migraine: a study conducted by the Belgian Migraine Society. Headache. Apr 1987;27(4):191-196. 26. Derry S, Moore RA. Paracetamol (acetaminophen) with or without an antiemetic for acute migraine headaches in adults. The Cochrane Database of Systematic Reviews. 2013;4:Cd008040. ©Institute for Clinical and Economic Review, 2014 Page 63 27. Kirthi V, Derry S, Moore RA. Aspirin with or without an antiemetic for acute migraine headaches in adults. The Cochrane Database of Systematic Reviews. 2013;4:Cd008041. 28. Law S, Derry S, Moore RA. Naproxen with or without an antiemetic for acute migraine headaches in adults. The Cochrane Database of Systematic Reviews. 2013;10:Cd009455. 29. Rabbie R, Derry S, Moore RA. Ibuprofen with or without an antiemetic for acute migraine headaches in adults. The Cochrane Database of Systematic Reviews. 2013;4:Cd008039. 30. Tfelt-Hansen P, Saxena PR, Dahlof C, et al. Ergotamine in the acute treatment of migraine: a review and European consensus. Brain: A Journal of Neurology. Jan 2000;123 ( Pt 1):9-18. 31. Ferrari MD, Roon KI, Lipton RB, Goadsby PJ. Oral triptans (serotonin 5-HT(1B/1D) agonists) in acute migraine treatment: a meta-analysis of 53 trials. Lancet. Nov 17 2001;358(9294):1668-1675. 32. Kelley NE, Tepper DE. Rescue therapy for acute migraine, part 3: opioids, NSAIDs, steroids, and post-discharge medications. Headache. Mar 2012;52(3):467-482. 33. Colman I, Brown MD, Innes GD, Grafstein E, Roberts TE, Rowe BH. Parenteral dihydroergotamine for acute migraine headache: a systematic review of the literature. Annals of Emergency Medicine. Apr 2005;45(4):393-401. 34. Friedman BW, Kapoor A, Friedman MS, Hochberg ML, Rowe BH. The relative efficacy of meperidine for the treatment of acute migraine: a meta-analysis of randomized controlled trials. Annals of Emergency Medicine. Dec 2008;52(6):705-713. 35. Kelly AM, Walcynski T, Gunn B. The relative efficacy of phenothiazines for the treatment of acute migraine: a meta-analysis. Headache. Oct 2009;49(9):1324-1332. 36. Bigal ME, Lipton RB. Excessive acute migraine medication use and migraine progression. Neurology. Nov 25 2008;71(22):1821-1828. 37. Shamliyan TA, Choi JY, Ramakrishnan R, et al. Preventive pharmacologic treatments for episodic migraine in adults. Journal of General Internal Medicine. Sep 2013;28(9):1225-1237. 38. Holland PR, Schembri CT, Frederick JP, Goadsby PJ. Transcranial magnetic stimulation for the treatment of migraine aura. Cephalalgia: An International Journal of Headache. 2009;72:A250. 39. Hurlow A, Bennett MI, Robb KA, Johnson MI, Simpson KH, Oxberry SG. Transcutaneous electric nerve stimulation (TENS) for cancer pain in adults. The Cochrane Database of Systematic Reviews. 2012;3:Cd006276. 40. Khadilkar A, Odebiyi DO, Brosseau L, Wells GA. Transcutaneous electrical nerve stimulation (TENS) versus placebo for chronic low-back pain. The Cochrane Database of Systematic Reviews. 2008(4):Cd003008. ©Institute for Clinical and Economic Review, 2014 Page 64 41. Fernandez-Del-Olmo M, Alvarez-Sauco M, Koch G, et al. How repeatable are the physiological effects of TENS? Clinical Neurophysiology: Official Journal of the International Federation of Clinical Neurophysiology. Aug 2008;119(8):1834-1839. 42. Nnoaham KE, Kumbang J. Transcutaneous electrical nerve stimulation (TENS) for chronic pain. The Cochrane Database of Systematic Reviews. 2008(3):Cd003222. 43. West J, Friedman BW. Current management of migraine in U.S. emergency departments: An analysis of NHAMCS, 33rd Annual Meeting of the American Pain Society; 2014; Tampa, Florida. 44. Bigal ME, Serrano D, Buse D, Scher A, Stewart WF, Lipton RB. Acute migraine medications and evolution from episodic to chronic migraine: a longitudinal population-based study. Headache. Sep 2008;48(8):1157-1168. 45. Tfelt-Hansen P, Pascual J, Ramadan N, et al. Guidelines for controlled trials of drugs in migraine: third edition. A guide for investigators. Cephalalgia: An International Journal of Headache. Jan 2012;32(1):6-38. 46. Lipton RB, Dodick DW, Silberstein SD, et al. Single-pulse transcranial magnetic stimulation for acute treatment of migraine with aura: a randomised, double-blind, parallel-group, sham-controlled trial. Lancet Neurology. Apr 2010;9(4):373-380. 47. Clarke BM, Upton AR, Kamath MV, Al-Harbi T, Castellanos CM. Transcranial magnetic stimulation for migraine: clinical effects. The Journal of Headache and Pain. Oct 2006;7(5):341-346. 48. Dodick DW, Schembri CT, Helmuth M, Aurora SK. Transcranial magnetic stimulation for migraine: a safety review. Headache. Jul 2010;50(7):1153-1163. 49. Conforto AB, Amaro E, Jr., Goncalves AL, et al. Randomized, proof-of-principle clinical trial of active transcranial magnetic stimulation in chronic migraine. Cephalalgia: An International Journal of Headache. May 2014;34(6):464-472. 50. Misra UK, Kalita J, Bhoi SK. High frequency repetitive transcranial magnetic stimulation (rTMS) is effective in migraine prophylaxis: an open labeled study. Neurological Research. Jul 2012;34(6):547-551. 51. Teepker M, Hotzel J, Timmesfeld N, et al. Low-frequency rTMS of the vertex in the prophylactic treatment of migraine. Cephalalgia: An International Journal of Headache. Feb 2010;30(2):137-144. 52. Brighina F, Piazza A, Vitello G, et al. rTMS of the prefrontal cortex in the treatment of chronic migraine: a pilot study. Journal of the Neurological Sciences. Dec 15 2004;227(1):67- 71. ©Institute for Clinical and Economic Review, 2014 Page 65 53. Lipton RB, Pearlman SH. Transcranial magnetic simulation in the treatment of migraine. Neurotherapeutics: The Journal of the American Society for Experimental NeuroTherapeutics. Apr 2010;7(2):204-212. 54. Schoenen J, Vandersmissen B, Jeangette S, et al. Migraine prevention with a supraorbital transcutaneous stimulator: a randomized controlled trial. Neurology. Feb 19 2013;80(8):697-704. 55. Piquet M, Balestra C, Sava SL, Schoenen JE. Supraorbital transcutaneous neurostimulation has sedative effects in healthy subjects. BMC Neurology. 2011;11:135. 56. Magis D, Sava S, d'Elia TS, Baschi R, Schoenen J. Safety and patients' satisfaction of transcutaneous supraorbital neurostimulation (tSNS) with the Cefaly(R) device in headache treatment: a survey of 2,313 headache sufferers in the general population. The Journal of Headache and Pain. 2013;14:95. 57. Asano E, Goadsby PJ. How do we fashion better trials for neurostimulator studies in migraine? Neurology. Feb 19 2013;80(8):694. 58. Silberstein S, Mathew N, Saper J, Jenkins S. Botulinum toxin type A as a migraine preventive treatment. For the BOTOX Migraine Clinical Research Group. Headache. Jun 2000;40(6):445- 450. 59. Barrientos N, Chana P. Botulinum toxin type A in prophylactic treatment of migraine headaches: a preliminary study. The Journal of Headache and Pain. 2003;4:146-151. 60. Evers S, Vollmer-Haase J, Schwaag S, Rahmann A, Husstedt IW, Frese A. Botulinum toxin A in the prophylactic treatment of migraine--a randomized, double-blind, placebo-controlled study. Cephalalgia: An International Journal of Headache. Oct 2004;24(10):838-843. 61. Ondo WG, Vuong KD, Derman HS. Botulinum toxin A for chronic daily headache: a randomized, placebo-controlled, parallel design study. Cephalalgia: An International Journal of Headache. Jan 2004;24(1):60-65. 62. Mathew NT, Frishberg BM, Gawel M, Dimitrova R, Gibson J, Turkel C. Botulinum toxin type A (BOTOX) for the prophylactic treatment of chronic daily headache: a randomized, double- blind, placebo-controlled trial. Headache. Apr 2005;45(4):293-307. 63. Silberstein SD, Stark SR, Lucas SM, Christie SN, Degryse RE, Turkel CC. Botulinum toxin type A for the prophylactic treatment of chronic daily headache: a randomized, double-blind, placebo-controlled trial. Mayo Clinic Proceedings. Sep 2005;80(9):1126-1137. 64. Anand KS, Prasad A, Singh MM, Sharma S, Bala K. Botulinum toxin type A in prophylactic treatment of migraine. American Journal of Therapeutics. May-Jun 2006;13(3):183-187. 65. Elkind AH, O'Carroll P, Blumenfeld A, DeGryse R, Dimitrova R. A series of three sequential, randomized, controlled studies of repeated treatments with botulinum toxin type A for ©Institute for Clinical and Economic Review, 2014 Page 66 migraine prophylaxis. The Journal of Pain: Official Journal of the American Pain Society. Oct 2006;7(10):688-696. 66. Aurora SK, Gawel M, Brandes JL, Pokta S, Vandenburgh AM. Botulinum toxin type a prophylactic treatment of episodic migraine: a randomized, double-blind, placebo- controlled exploratory study. Headache. Apr 2007;47(4):486-499. 67. Relja M, Poole AC, Schoenen J, Pascual J, Lei X, Thompson C. A multicentre, double-blind, randomized, placebo-controlled, parallel group study of multiple treatments of botulinum toxin type A (BoNTA) for the prophylaxis of episodic migraine headaches. Cephalalgia: An International Journal of Headache. Jun 2007;27(6):492-503. 68. Saper JR, Mathew NT, Loder EW, DeGryse R, VanDenburgh AM. A double-blind, randomized, placebo-controlled comparison of botulinum toxin type a injection sites and doses in the prevention of episodic migraine. Pain Medicine (Malden, Mass.). Sep 2007;8(6):478-485. 69. Vo AH, Satori R, Jabbari B, et al. Botulinum toxin type-a in the prevention of migraine: a double-blind controlled trial. Aviation, Space, and Environmental Medicine. May 2007;78(5 Suppl):B113-118. 70. Cady R, Schreiber C. Botulinum toxin type A as migraine preventive treatment in patients previously failing oral prophylactic treatment due to compliance issues. Headache. Jun 2008;48(6):900-913. 71. Freitag FG, Diamond S, Diamond M, Urban G. Botulinum Toxin Type A in the treatment of chronic migraine without medication overuse. Headache. Feb 2008;48(2):201-209. 72. Mathew NT, Jaffri SF. A double-blind comparison of onabotulinumtoxina (BOTOX) and topiramate (TOPAMAX) for the prophylactic treatment of chronic migraine: a pilot study. Headache. Nov-Dec 2009;49(10):1466-1478. 73. Petri S, Tolle T, Straube A, Pfaffenrath V, Stefenelli U, Ceballos-Baumann A. Botulinum toxin as preventive treatment for migraine: a randomized double-blind study. European Neurology. 2009;62(4):204-211. 74. Aurora SK, Dodick DW, Turkel CC, et al. OnabotulinumtoxinA for treatment of chronic migraine: results from the double-blind, randomized, placebo-controlled phase of the PREEMPT 1 trial. Cephalalgia: An International Journal of Headache. Jul 2010;30(7):793-803. 75. Diener HC, Dodick DW, Aurora SK, et al. OnabotulinumtoxinA for treatment of chronic migraine: results from the double-blind, randomized, placebo-controlled phase of the PREEMPT 2 trial. Cephalalgia: An International Journal of Headache. Jul 2010;30(7):804-814. 76. Magalhaes E, Menezes C, Cardeal M, Melo A. Botulinum toxin type A versus amitriptyline for the treatment of chronic daily migraine. Clinical Neurology and Neurosurgery. Jul 2010;112(6):463-466. ©Institute for Clinical and Economic Review, 2014 Page 67 77. Cady RK, Schreiber CP, Porter JA, Blumenfeld AM, Farmer KU. A multi-center double-blind pilot comparison of onabotulinumtoxinA and topiramate for the prophylactic treatment of chronic migraine. Headache. Jan 2011;51(1):21-32. 78. Chankrachang S, Arayawichanont A, Poungvarin N, et al. Prophylactic botulinum type A toxin complex (Dysport(R)) for migraine without aura. Headache. Jan 2011;51(1):52-63. 79. Sandrini G, Perrotta A, Tassorelli C, et al. Botulinum toxin type-A in the prophylactic treatment of medication-overuse headache: a multicenter, double-blind, randomized, placebo-controlled, parallel group study. The Journal of Headache and Pain. Aug 2011;12(4):427-433. 80. Jackson JL, Kuriyama A, Hayashino Y. Botulinum toxin A for prophylactic treatment of migraine and tension headaches in adults: a meta-analysis. JAMA: The Journal of the American Medical Association. Apr 25 2012;307(16):1736-1745. 81. Shamliyan TA, Kane RL, Taylor FR. Migraine in Adults: Preventive Pharmacologic Treatments. AHRQ Comparative Effectiveness Reviews. Rockville (MD)2013. 82. Smelt AF, Assendelft WJ, Terwee CB, Ferrari MD, Blom JW. What is a clinically relevant change on the HIT-6 questionnaire? An estimation in a primary-care population of migraine patients. Cephalalgia: An International Journal of Headache. Jan 2014;34(1):29-36. 83. Diener HC, Dodick DW, Turkel CC, et al. Pooled analysis of the safety and tolerability of onabotulinumtoxinA in the treatment of chronic migraine. European Journal of Neurology: The Official Journal of the European Federation of Neurological Societies. Jun 2014;21(6):851-859. 84. Solomon S. Botulinum toxin for the treatment of chronic migraine: the placebo effect. Headache. Jun 2011;51(6):980-984. 85. Solomon S. OnabotulinumtoxinA for treatment of chronic migraine: the unblinding problem. Headache. May 2013;53(5):824-826. 86. Taheraghdam AA, Amiri H, Shojaan H, Shamsvahdati S, Houshyar Y. Intravenous dexamethasone versus morphine in relieving of acute migraine headache. Pakistan Journal of Biological Sciences: PJBS. Jun 15 2011;14(12):682-687. 87. Alemdar M, Pekdemir M, Selekler HM. Single-dose intravenous tramadol for acute migraine pain in adults: a single-blind, prospective, randomized, placebo-controlled clinical trial. Clinical Therapeutics. Jul 2007;29(7):1441-1447. 88. Engindeniz Z, Demircan C, Karli N, et al. Intramuscular tramadol vs. diclofenac sodium for the treatment of acute migraine attacks in emergency department: a prospective, randomised, double-blind study. The Journal of Headache and Pain. Jun 2005;6(3):143-148. ©Institute for Clinical and Economic Review, 2014 Page 68 89. Cicek M, Karcioglu O, Parlak I, et al. Prospective, randomised, double blind, controlled comparison of metoclopramide and pethidine in the emergency treatment of acute primary vascular and tension type headache episodes. Emergency Medicine Journal: EMJ. May 2004;21(3):323-326. 90. Richman PB, Allegra J, Eskin B, et al. A randomized clinical trial to assess the efficacy of intramuscular droperidol for the treatment of acute migraine headache. The American Journal of Emergency Medicine. Jan 2002;20(1):39-42. 91. Carleton SC, Shesser RF, Pietrzak MP, et al. Double-blind, multicenter trial to compare the efficacy of intramuscular dihydroergotamine plus hydroxyzine versus intramuscular meperidine plus hydroxyzine for the emergency department treatment of acute migraine headache. Annals of Emergency Medicine. Aug 1998;32(2):129-138. 92. Harden RN, Gracely RH, Carter T, Warner G. The placebo effect in acute headache management: ketorolac, meperidine, and saline in the emergency department. Headache. Jun 1996;36(6):352-356. 93. Scherl ER, Wilson JF. Comparison of dihydroergotamine with metoclopramide versus meperidine with promethazine in the treatment of acute migraine. Headache. May 1995;35(5):256-259. 94. Davis CP, Torre PR, Williams C, et al. Ketorolac versus meperidine-plus-promethazine treatment of migraine headache: evaluations by patients. The American Journal of Emergency Medicine. Mar 1995;13(2):146-150. 95. Klapper JA, Stanton J. Current emergency treatment of severe migraine headaches. Headache. Nov-Dec 1993;33(10):560-562. 96. Duarte C, Dunaway F, Turner L, Aldag J, Frederick R. Ketorolac versus meperidine and hydroxyzine in the treatment of acute migraine headache: a randomized, prospective, double-blind trial. Annals of Emergency Medicine. Sep 1992;21(9):1116-1121. 97. Larkin GL, Prescott JE. A randomized, double-blind, comparative study of the efficacy of ketorolac tromethamine versus meperidine in the treatment of severe migraine. Annals of Emergency Medicine. Aug 1992;21(8):919-924. 98. Stiell IG, Dufour DG, Moher D, Yen M, Beilby WJ, Smith NA. Methotrimeprazine versus meperidine and dimenhydrinate in the treatment of severe migraine: a randomized, controlled trial. Annals of Emergency Medicine. Nov 1991;20(11):1201-1205. 99. Lane PL, McLellan BA, Baggoley CJ. Comparative efficacy of chlorpromazine and meperidine with dimenhydrinate in migraine headache. Annals of Emergency Medicine. Apr 1989;18(4):360-365. ©Institute for Clinical and Economic Review, 2014 Page 69 100. Belgrade MJ, Ling LJ, Schleevogt MB, Ettinger MG, Ruiz E. Comparison of single-dose meperidine, butorphanol, and dihydroergotamine in the treatment of vascular headache. Neurology. Apr 1989;39(4):590-592. 101. Tek D, Mellon M. The effectiveness of nalbuphine and hydroxyzine for the emergency treatment of severe headache. Annals of Emergency Medicine. Mar 1987;16(3):308-313. 102. Hoag R, Mortimer L. Methotrimeprazine compared to meperidine and dimenhydrinate in the treatment of migraine headache. CAEP Review. 1986:29-31. 103. Sumamo Schellenberg E, Dryden DM, Pasichnyk D, et al. AHRQ Comparative Effectiveness Reviews. Acute Migraine Treatment in Emergency Settings. Rockville (MD): Agency for Healthcare Research and Quality (US); 2012. 104. U.S. Bureau of Labor Statistics. Consumer Price Index. Accessed at: http://www.bls.gov/cpi/home.htm, June 10, 2014. 105. Redbook 2014: Truven Health Analytics, Inc. Micromedex 2.0: Redbook Online. Accessed at: http://www.micromedexsolutions.com/micromedex2/librarian/deeplinkaccess?SearchTerm =TEST, June 10, 2014. 106. Munakata J, Hazard E, Serrano D, Klingman D, Rupnow MF, Tierce J, Reed M, Lipton RB. Economic burden of transformed migraine: results from the American Migraine Prevalence and Prevention (AMPP) Study. Headache. 2009 Apr;49(4):498-508. 107. State of California, Department of Health Services (Medi-Cal). Medi-Cal Rates, by Procedure Code. Accessed at: http://files.medi- cal.ca.gov/pubsdoco/rates/rates_information.asp?num=23&first=A0030&last=L3253, June 10, 2014. 108. State of California, Department of Health Services (Medi-Cal). Contract Drugs List, Part 9. Accessed at: http://files.medi- cal.ca.gov/pubsdoco/manual/man_query.asp?wSearch=%28%23filename+drugscdl*.doc+O R+%23filename+drugscdl*.zip%29&wFLogo=Contract+Drugs+List&wFLogoH=52&wFLogoW= 516&wAlt=Contract+Drugs+List&wPath=N, June 10, 2014. 109. U.S. Census Bureau. Population Estimates: State Characteristics. Accessed at: http://www.census.gov/popest/data/state/asrh/2012/index.html, June 10, 2014. 110. Smitherman TA, Burch R, Sheikh H, Loder E. The prevalence, impact, and treatment of migraine and severe headaches in the United States: a review of statistics from national surveillance studies. Headache. 2013 Mar;53(3):427-36. 111. Buse DC, Pearlman SH, Reed ML, Serrano D, Ng-Mak DS, Lipton RB. Opioid use and dependence among persons with migraine: results of the AMPP study. Headache. 2012 Jan;52(1):18-36. ©Institute for Clinical and Economic Review, 2014 Page 70 112. Katz C, El-Gabalawy R, Keyes KM, Martins SS, Sareen J. Risk factors for incident nonmedical prescription opioid use and abuse and dependence: results from a longitudinal nationally representative sample. Drug Alcohol Depend. 2013 Sep 1;132(1-2):107-13. 113. White AG, Birnbaum HG, Mareva MN, Daher M, Vallow S, Schein J, Katz N. Direct costs of opioid abuse in an insured population in the United States. J Manag Care Pharm. 2005 Jul- Aug;11(6):469-79. 114. Wall R, Rehm J, Fischer B, Brands B, Gliksman L, Stewart J, Medved W, Blake J. Social costs of untreated opioid dependence. J Urban Health. 2000 Dec;77(4):688-722. 115. State of California, Office of Statewide Health Planning & Development. ICD-9-CM Code Frequencies, Emergency Department, 2013. Accessed at: http://www.oshpd.ca.gov/HID/Products/EmerDeptData/FreqTables/ICD9CodeFreq/default. asp, June 10, 2014. 116. Hawkins K, Wang S, Rupnow M. Direct cost burden among insured US employees with migraine. Headache. 2008 Apr;48(4):553-63. Epub 2007 Dec 7. 117. Xie L, Joshi AV, Schaaf D, Mardekian J, Harnett J, Shah ND, Baser O. Differences in Healthcare Utilization and Associated Costs Between Patients Prescribed vs. Nonprescribed Opioids During an Inpatient or Emergency Department Visit. Pain Pract. 2014 Jun;14(5):446- 56. 118. Cefaly Technology. Cefaly order form. Accessed at: http://www.cefaly.us/en/cefaly-shop, June 10, 2014. 119. Brown JS, Papadopoulos G, Neumann PJ, Friedman M, Miller JD, Menzin J. Cost- effectiveness of topiramate in migraine prevention: results from a pharmacoeconomic model of topiramate treatment. Headache. 2005 Sep;45(8):1012-22. 120. Adelman JU, Adelman LC, Von Seggern R. Cost-effectiveness of antiepileptic drugs in migraine prophylaxis. Headache. 2002 Nov-Dec;42(10):978-83. 121. Slof J, Láinez JM, Comas A, Heras J. [Almotriptan vs. ergotamine plus caffeine for acute migraine treatment. A cost-efficacy analysis]. Neurologia. 2009 Apr;24(3):147-53. Spanish. 122. Williams P, Reeder CE. Cost-effectiveness of almotriptan and rizatriptan in the treatment of acute migraine. Clin Ther. 2003 Nov;25(11):2903-19. 123. Williams P, Dowson AJ, Rapoport AM, Sawyer J. The cost effectiveness of stratified care in the management of migraine. Pharmacoeconomics. 2001;19(8):819-29. 124. Diener HC, Bussone G, Van Oene JC, Lahaye M, Schwalen S, Goadsby PJ. Topiramate reduces headache days in chronic migraine: a randomized, double-blind, placebo-controlled study. Cephalalgia: An International Journal of Headache. Jul 2007;27(7):814-823. ©Institute for Clinical and Economic Review, 2014 Page 71 125. Silberstein S, Lipton R, Dodick D, et al. Topiramate treatment of chronic migraine: a randomized, placebo-controlled trial of quality of life and other efficacy measures. Headache. Sep 2009;49(8):1153-1162. 126. Silberstein SD, Lipton RB, Dodick DW, et al. Efficacy and safety of topiramate for the treatment of chronic migraine: a randomized, double-blind, placebo-controlled trial. Headache. Feb 2007;47(2):170-180. 127. Silvestrini M, Bartolini M, Coccia M, Baruffaldi R, Taffi R, Provinciali L. Topiramate in the treatment of chronic migraine. Cephalalgia: An International Journal of Headache. Oct 2003;23(8):820-824.] 128. Yurekli VA, Akhan G, Kutluhan S, Uzar E, Koyuncuoglu HR, Gultekin F. The effect of sodium valproate on chronic daily headache and its subgroups. The Journal of Headache and Pain. Feb 2008;9(1):37-41. 129. Spira PJ, Beran RG. Gabapentin in the prophylaxis of chronic daily headache: a randomized, placebo-controlled study. Neurology. Dec 23 2003;61(12):1753-1759. 130. Saper JR, Lake AE, 3rd, Cantrell DT, Winner PK, White JR. Chronic daily headache prophylaxis with tizanidine: a double-blind, placebo-controlled, multicenter outcome study. Headache. Jun 2002;42(6):470-482. 131. Aurora SK, Winner P, Freeman MC, et al. OnabotulinumtoxinA for treatment of chronic migraine: pooled analyses of the 56-week PREEMPT clinical program. Headache. Oct 2011;51(9):1358-1373. ©Institute for Clinical and Economic Review, 2014 Page 72 APPENDIX ©Institute for Clinical and Economic Review, 2014 Page 73 1. Search Strategies The Medline database, Embase, Cochrane clinical trials database, Cochrane reviews database, and the Database of Abstracts of Reviews of Effects (DARE), were searched using the key words “transcranial magnetic stimulation” OR “nerve stimulator” OR “botulinum toxin” OR “opioid” AND the keyword “migraine.” Studies of TMS for prevention were excluded. The search was limited to clinical trials published in English. The search was performed for the period from 1945 through May 21, 2014. 2. Key Articles Single pulse Transcranial Magnetic Stimulation for the acute treatment of migraines 1. Lipton RB, Pearlman SH. Transcranial magnetic simulation in the treatment of migraine. Neurotherapeutics: The Journal of the American Society for Experimental NeuroTherapeutics. Apr 2010;7(2):204-212. Supraorbital Transcutaneous Stimulation (Cefaly) for the prevention of migraines 2. Schoenen J, Vandersmissen B, Jeangette S, et al. Migraine prevention with a supraorbital transcutaneous stimulator: a randomized controlled trial. Neurology. Feb 19 2013;80(8):697-704. Botulinum toxin for the prevention of migraines 3. Aurora SK, Dodick DW, Turkel CC, et al. OnabotulinumtoxinA for treatment of chronic migraine: results from the double-blind, randomized, placebo-controlled phase of the PREEMPT 1 trial. Cephalalgia: An International Journal of Headache. Jul 2010;30(7):793-803. 4. Diener HC, Dodick DW, Aurora SK, et al. OnabotulinumtoxinA for treatment of chronic migraine: results from the double-blind, randomized, placebo-controlled phase of the PREEMPT 2 trial. Cephalalgia: An International Journal of Headache. Jul 2010;30(7):804-814. ©Institute for Clinical and Economic Review, 2014 Page 74 Supplemental Migraine Evidence Tables Table A1: Characteristics of the Randomized Trials of Single Pulse Transcranial Magnetic Stimulation Study N TMS Inclusion criteria Exclusion Dates Follow-up Age, Sex, % Migraines Primary criteria (months) years Female per month outcome Lipton 201 Cerena • Migraine w/aura • Aura > 60 min 2006- 3 39 80 4.4 Pain-free 2 53 2010 • 1-8 headaches / • Metal 2008 hours after month implants treatment, st USA 18 • Overuse of 1 attack centers headache drugs Table A2: Methodological Quality of the Randomized Trials of Single Pulse Transcranial Magnetic Stimulation Study Randomization Allocation Groups Blinding of Limited and non- Intention-to- Conflicts Likelihood of concealment comparable Participants / Providers differential loss treat (ITT) of interest bias / Outcome assessor to follow-up analysis 53 Lipton 2010 Yes Yes Yes Yes/Yes/Yes Yes Modified ITT Yes Intermediate USA 18 centers ©Institute for Clinical and Economic Review, 2014 Page 75 Table A3: Characteristics of the Randomized Trials of the Cefaly/TENS Device Study N Device Inclusion Exclusion criteria Dates Follow-up Age, years Sex, % Migraines Primary outcome criteria (months) Female per month Schoenen 201 Cefaly • 18-65 years • Preventive therapy 2009-2011 3 37 91 6.7 Change in 54 2013 •≥2 in past 3 months migraine days migraines / • Failed ≥ 3 Belgium 5 month preventive therapies At least 50% centers • MOH reduction in migraine days Table A4: Methodological Quality of the Randomized Trials of the Cefaly/TENS Device Study Randomization Allocation Groups Blinding of Limited and non- Intention-to- Conflicts Likelihood concealment comparable Participants / Providers differential loss treat analysis of interest of bias / Outcome assessor to follow-up 54 Schoenen 2013 Yes Yes No, Active 4.5 Yes/Yes/Yes Yes Yes No High years younger, Belgium 5 centers 3.5 fewer years Unblinding not with migraines assessed; likely ©Institute for Clinical and Economic Review, 2014 Page 76 Table A5: Characteristics of the Randomized Trials of Botulinum Toxin for the Prevention of Migraine Headaches Study N Lost to Dose, Control* Follow- Age, Sex, % Migraine Primary Follow- units up years Female days per outcome up % (months) month Episodic Migraine 58 Silberstein 2000 123 1% 25, 75 Placebo 3 44 85 4.4 Not stated US 12 Centers 59 Barrientos 2003 30 0% 50 Placebo 3 41 80 5.1 Frequency of migraine Chile episodes 60 Evers 2004 60 0% 16, 100 Placebo 3 38 83 6.2 ≥ 50% reduction in migraine Germany 1 Center episodes 64 Anand 2006 32 - 50 Placebo 3 - 75 8.8 Frequency of migraine India 1 Center episodes 65 Elkind 2006 418 9% 7.5,25, Placebo 4 44 85 5.6 Frequency of North America 7 50 migraine Centers episodes 66 Aurora 2007 369 23% 260 Placebo 9 45 89 6.5 Frequency of North America 20 migraine Centers episodes 67 Relja 2007 495 19% 75, 150, Placebo 9 43 88 4.5 Frequency of Europe 37 Centers 225 migraine episodes 68 Saper 2007 232 3% 25 Placebo 3 44 86 5.7 Frequency of North America 7 migraine Centers episodes 70 Cady 2008 59 8% 139 Placebo 3 42 85 7.7 Not stated US 1 Center 73 Petri 2009 122 4% 80, 210 Placebo 3 46 84 4.8 Frequency of migraine Germany 16 Centers episodes ©Institute for Clinical and Economic Review, 2014 Page 77 Study N Lost to Dose, Control* Follow- Age, Sex, % Migraine Primary Follow- units up years Female days per outcome up % (months) month 78 Chankrachang 2011 128 7% 120, 3 39 95 5.1 Frequency of 240 migraine Thailand 6 Centers episodes Chronic Migraine – Placebo control 61 Ondo 2004 60 3% 200 Placebo 3 46 82 15.3 Frequency of headache-free US 1 Center days 62 Mathew 2005 355 23% 260 Placebo 9 44 85 13.1 Frequency of North America 13 headache-free Centers days 63 Silberstein 2005 702 6% 75, 150, Placebo 9 43 83 13.8 Frequency of North America 28 225 headache-free Centers days 69 Vo 2007 32 35% 205 Placebo 3 42 84 19.4 Frequency of headache US 1 Center episodes 71 Freitag 2008 60 8% 100 Placebo 4 42 75 14.2 Frequency of migraine US 1 Center episodes 74 Aurora 2010 679 13% 195 Placebo 6 42 88 19.1 Frequency of North America 56 headache Centers (PREEMPT 1) episodes 75 Diener 2010 705 9% 195 Placebo 6 41 85 19.8 Frequency of Europe, North headache-free America 66 Centers days (PREEMPT 2) 79 Sandrini 2011 68 18% 100 Placebo 3 34 80 24.9 Frequency of headache days Italy Chronic Migraine – Active control Mathew 2009 60 45% 200 Topira- 9 37 90 15.5 Physician global mate assessment US 1 Center ©Institute for Clinical and Economic Review, 2014 Page 78 Study N Lost to Dose, Control* Follow- Age, Sex, % Migraine Primary Follow- units up years Female days per outcome up % (months) month 76, ** Magalhaes 2010 72 - 250 Amitrip- 3 90 97 24 Not stated tyline Brazil 1 Center 77 Cady 2011 59 25% 300 Topira- 3 40 92 21.1 Physician global North America 3 mate assessment Centers *placebo = sham injection with saline; **Botulinum in this study was a different form of botulinum toxin A marketed as Dysport by Ipsen Biopharm Ltd. ©Institute for Clinical and Economic Review, 2014 Page 79 Table A6: Methodological Quality of the Randomized Trials of Botulinum Toxin for the Prevention of Migraine Headaches Study Randomization Allocation Groups Blinding of Limited and non- Intention-to- Industry Likelihood concealment comparable Participants / Providers differential loss treat analysis sponsored of bias / Outcome assessor to follow-up Episodic Migraine 58 Silberstein 2000 Unclear Unclear Yes Yes/Yes/Yes Yes Yes Yes Moderate US 12 Centers 59 Barrientos 2003 Yes Unclear Yes Yes/Yes/Yes Yes Yes Yes Low Chile 60 Evers 2004 Yes Yes Yes Yes/Yes/Yes Yes Yes Yes Low Germany 1 Center 64 Anand 2006 Unclear Unclear Unclear Unclear Unclear No Unclear High India 1 Center 65 Elkind 2006 Yes Unclear Yes Yes/Yes/Yes Yes Yes Yes Moderate North America 7 Centers 66 Aurora 2007 Yes Unclear Yes Yes/Yes/Yes No Yes Yes Moderate North America 20 Centers 67 Relja 2007 Yes Unclear Yes Yes/Yes/Yes Unclear No Yes Moderate Europe 37 Centers 68 Saper 2007 Yes Unclear Yes Yes/Yes/Yes Yes Yes Yes Low North America 7 Centers 70 Cady 2008 Yes Unclear Yes Unclear Yes No Yes High US 1 Center 73 Petri 2009 Yes Unclear No Yes/Yes/Yes Yes No Yes Moderate Germany 16 Centers ©Institute for Clinical and Economic Review, 2014 Page 80 Study Randomization Allocation Groups Blinding of Limited and non- Intention-to- Industry Likelihood concealment comparable Participants / Providers differential loss treat analysis sponsored of bias / Outcome assessor to follow-up Chankrachang Yes Yes Yes Yes/Yes/Yes Yes Yes Yes Low 78 2011 Thailand 6 Centers Chronic Migraine – Placebo control 61 Ondo 2004 Unclear Unclear Yes Yes/Yes/Yes Yes No Unclear Moderate US 1 Center 62 Mathew 2005 Yes Unclear Yes Yes/Yes/Yes No Yes Yes Moderate North America 13 Centers 63 Silberstein 2005 Yes Yes Yes Yes/Yes/Yes Yes Yes Yes Low North America 28 Centers 69 Vo 2007 Unclear Unclear Yes Yes/Yes/Yes No No No High US 1 Center 71 Freitag 2008 Yes No Yes Yes/Yes/Yes Yes Yes Yes Moderate US 1 Center 74 Aurora 2010 Yes Yes No Yes/Yes/Yes Yes Yes Yes Moderate North America 56 Centers (PREEMPT 1) 75 Diener 2010 Yes Yes Yes Yes/Yes/Yes Yes Yes Yes Low Europe, North America 66 Centers (PREEMPT 2) 79 Sandrini 2011 Yes Yes Yes Yes/Yes/Yes Yes Yes Yes Low Italy ©Institute for Clinical and Economic Review, 2014 Page 81 Chronic Migraine – Active control Mathew 2009 Unclear Yes No Yes/Yes/Yes No Not stated Yes Moderate US 1 Center 76,* Magalhaes 2010 Yes No No Unclear NR No No Low Brazil 1 Center 77 Cady 2011 Yes Unclear No Unclear No Yes No High North America 3 Centers * Botulinum in this study was a different form of botulinum toxin A marketed as Dysport by Ipsen Biopharm Ltd. ©Institute for Clinical and Economic Review, 2014 Page 82 Table A7: Characteristics of the Randomized Trials of Opioids for the Acute Treatment of Migraine in the Emergency Department Study N Opioid Control Dates Age, Sex, % Pain intensity Headache Primary outcome years Female (10 point Duration VAS) (hours) Versus placebo 101 Tek 1987 94 Nalbuphine Placebo 1985 29 67% NR >4 Pain response at one hour 92 Harden 1996 20 Meperidine Placebo Unclear 32 80% 8.0 Not reported Not stated 89 Cicek 2004 97 Meperidine Placebo 2000-01 39 88% 7.9 Not reported Not stated 87 Alemdar 2007 34 Tramadol Placebo NR 50 82% 6.3 16 Pain response at one hour Versus active control 102 Hoag 1986 40 Meperidine Methotrimeprazine Unclear NR NR 8.3 NR Pain severity 100 Belgrade 1989 64 Meperidine Dihydroergotamine Unclear 30 63% NR NR Pain improvement 99 Lane 1989 46 Meperidine Chloropramazine Unclear 31 85% 8.1 48 Pain response at 45 minutes 98 Stiell 1991 75 Meperidine Methotrimeprazine 1990 31 76% 7.9 26 Change in pain at one hour 96 Duarte 1992 50 Meperidine Ketorolac Unclear 35 80% 8.0 29 Not stated 97 Larkin 1992 31 Meperidine Ketorolac Unclear 33 77% NR NR Not stated 95 Klapper 1993 28 Meperidine DHE+Met Unclear NR NR NR NR Not stated 94 Davis 1995 42 Meperidine Ketorolac 1992-93 35 81% NR NR Not stated 93 Scherl 1995 27 Meperidine DHE+Met NR 31 70% NR NR Percent pain relief at one hour 92 Harden 1996 20 Meperidine Ketorolac Unclear 32 80% 8.0 Not reported Not stated 91 Carleton 1998 170 Meperidine DHE + H 1991-92 32 82% 7.8 29 Percent pain relief at one hour 90 Richman 2002 29 Meperidine Droperidol Unclear 32 72% 8.2 22 Change in pain at 30 minutes 89 Cicek 2004 99 Meperidine Metoclopramide 2000-01 39 88% 7.9 Not reported Not stated 88 Engindeniz 2005 47 Tramadol Diclofenac 2003-04 37 78% NR NR Pain response at two hours 86 Taheraghdam 2011 190 Morphine Dexamethasone 2008-09 44 38% 8.6 NR Not stated ©Institute for Clinical and Economic Review, 2014 Page 83 Table A8: Methodological Quality of the Randomized Trials of Opioids for the Acute Treatment of Migraine in the Emergency Department Study Randomization Allocation Groups Blinding of Limited and non- Intention-to- Industry Likelihood concealment comparable Participants / Providers differential loss treat analysis sponsored of bias / Outcome assessor to follow-up Versus placebo 101 Tek 1987 Yes Unclear Yes Yes/Yes/Yes Yes Not stated No Moderate 92 Harden 1996 Unclear Yes Unclear Yes/Yes/Yes Yes Not stated Yes High 89 Cicek 2004 Yes Yes No Yes/Yes/Yes Yes Not stated No High 87 Alemdar 2007 Unclear Unclear No Yes/No/Yes No Not stated NR High Versus active control 102 Hoag 1986 Unclear Unclear No Yes/Yes/Yes Yes Not stated No High 100 Belgrade 1989 Unclear Unclear Yes Yes/Yes/Yes Yes Not stated No High 99 Lane 1989 Unclear Yes Yes Yes/Yes/Yes Yes Not stated No Moderate 98 Stiell 1991 Yes Yes Yes Yes/Yes/Yes Yes Not stated No Moderate 96 Duarte 1992 Unclear Yes No Yes/Yes/Yes Yes Not stated NR Moderate 97 Larkin 1992 Unclear Unclear Yes Yes/Yes/Yes Yes Not stated NR Moderate 95 Klapper 1993 Unclear Unclear Unclear Yes/Yes/Yes Yes Not stated Yes High 94 Davis 1995 Unclear Yes Yes Yes/Yes/Yes Yes Not stated Yes Moderate 93 Scherl 1995 Unclear Unclear Unclear Yes/Yes/Yes Yes Not stated NR High 92 Harden 1996 Unclear Yes Unclear Yes/Yes/Yes Yes Not stated Yes High 91 Carleton 1998 Yes Yes Yes Yes/Yes/Yes Yes No Yes Moderate 90 Richman 2002 Unclear Unclear No Yes/Yes/Yes Yes Yes NR Moderate 89 Cicek 2004 Yes Yes No Yes/Yes/Yes Yes Not stated No High 88 Engindeniz 2005 Yes Yes Yes Yes/Yes/Yes Yes No NR Moderate 86 Taheraghdam 2011 Unclear Unclear Yes Yes/Yes/Yes Not reported Not stated NR High ©Institute for Clinical and Economic Review, 2014 Page 84 Table A9: Primary Outcomes in the Randomized Trials of Opioids for the Acute Treatment of Migraine in the Emergency Department Study Group N Pain-free Mild or no Pain-free Pain-free Use of rescue Use of rescue Change VAS AEs 1-2 hours pain 1-2 24 hours 48 hours medication medication 10 points hours 0-2 hours 0-48 hours Versus placebo 101 Tek 1987 Nalbuphine 22 NR NR NR NR NR NR NR Sedation 60% vs Placebo 24 17% 92 Harden 1996 Mep + Pro 10 NR NR NR NR NR NR No NR difference Placebo 11 between groups 89 Cicek 2004 Meperidine 49 NR NR NR NR 41% NR -6.1 57% Estimated Placebo 48 57% -4.0 12% 87 Alemdar 2007 Tramadol 17 29% 71% NR NR 29% NR -3.1 6% Placebo 17 12% 35% 65% -2.1 0% Versus active con 102 Hoag 1986 Mep + Dim 18 NR NR NR NR NR NR -2.2 Incom- plete Methotrim 22 -3.7 100 Belgrade 1989 Mep + H 22 NR NR NR NR NR NR -3.7 Incom- Butorphanol 19 -5.4 plete DHE + Met 23 -5.9 99 Lane 1989 Meperdine 22 NR 50% NR NR 50% NR -4.4 27% Chlorpromazine 24 92% 8% -7.1 46% 98 Stiell 1991 Meperidine 37 NR NR NR NR 27% NR -4.7 Overall NR Methotrim 37 30% -4.0 96 Duarte 1992 Meperidine 25 16% 56% NR NR 28% NR -4.9 48% Ketorolac 25 20% 60% 36% -4.4 28% ©Institute for Clinical and Economic Review, 2014 Page 85 Study Group N Pain-free Mild or no Pain-free Pain-free Use of rescue Use of rescue Change VAS AEs 1-2 hours pain 1-2 24 hours 48 hours medication medication 10 points hours 0-2 hours 0-48 hours 97 Larkin 1992 Meperidine 16 29% NR NR NR 38% NR NR 0% Ketorolac 15 6% 73% 0% 95 Klapper 1993 Meperidine 14 21% NR NR NR NR -0.9 0% DHE + Met 14 93% -2.1 0 94 Davis 1995 Mep + Pro 22 NR 64% NR NR NR NR NR NR Ketorolac 20 50% 93 Scherl 1995 Mep + Pro 13 77% NR NR NR NR 33 NR 7.2 DHE + Met 14 86% 54 3.9 92 Harden 1996 Mep + Pro 10 NR NR NR NR NR NR No NR difference Ketorolac 9 between groups 91 Carleton 1998 Mep + H 78 NR NR NR NR 18% NR -4.5 42% DHE + H 78 21% -4.1 41% 90 Richman 2002 Meperidine 14 NR NR NR NR NR NR -3.7 Incom- plete Droperidol 15 -4.7 89 Cicek 2004 Meperidine 49 NR NR NR NR 41% NR -6.1 57% Estimated Met 50 14% -6.7 38% 88 Engindeniz 2005 Tramadol 20 35% 80% NR 30% 20% NR NR 5% Diclofenac 20 45% 80% 40% 20% 10% Taheraghdam Morphine NR NR NR NR NR NR NR -6.4 NR 86 2011 Dexamethasone NR -5.6 Mep = Meperidine Pro = Promethazine Met = metoclopramide DHE = Dihydroergotamine H = hydroxyzine Dim = Dimenhydrinate Methotrim = Methotrimeprazine ©Institute for Clinical and Economic Review, 2014 Page 86