Background: Migraine headache is a common, disabling condition that impacts 1 in 6 Americans. Although many interventions for migraine prevention have been shown to be effective, decision making for patients and physicians can be complex. Although many newer interventions have received United States (US) Food and Drug Administration (FDA) clearance, no systematic reviews have synthesized evidence for efficacy or harms across old and newer interventions.
Purpose: To summarize evidence from randomized controlled trials (RCTs) for pharmacologic drugs and devices for migraine prevention in visual web-based evidence maps. Specifically, these evidence maps were intended to do the following:
Visualize all existing evidence from randomized clinical trials on drugs and non-invasive devices for migraine prevention.
Assess effectiveness of guideline recommended drugs and noninvasive devices for migraine reduction, tolerability, and reported harms.
Present findings in an easy-to-use interactive visual format.
Methods: The ECRI Evidence-based Practice Center performed a rapid review of the literature to identify existing RCTs for 44 drugs and 2 devices for migraine prevention. We searched PubMed and EMBASE from inception to June 24, 2020. We included English-language RCTs enrolling adults with episodic or chronic migraine, a study duration ≥ 8 weeks, and >10 patients per arm.
For a subset of interventions (15 drugs, 2 devices) we performed meta-analyses of inactive controlled RCTs to assess efficacy and harms. We used the Cochrane risk of bias tool to assess individual studies and Grading and Recommendations Assessment, Development and Evaluation (GRADE) rating system to assess the quality of evidence.
We summarized findings using 3 web-based evidence maps, accessible here:
Results: Overall, 203 RCTs were included: 78 trials in Map 1, 123 trials in Map 2, and 133 trials in Map 3. Two visualizations (Maps 1 and 2) presented findings from placebo-/sham-controlled RCTs, while Map 3 displayed comparisons from head-to-head RCTs.
Key Findings: Placebo-/Sham-Controlled RCTs:
Episodic migraine: Aside from onabotulinumtoxinA (no effect), interventions improved headaches by 0.5 to 2.4 migraine days per month. Efficacy for common first-line interventions (amitriptyline, propranolol, topiramate) was underwhelming (0.73 to 0.95 fewer migraine days per month) and based on low/very-low quality evidence. Calcitonin gene–related peptide (CGRP) antagonists generally provided larger efficacy with fewer side effects and higher quality evidence. Further research is needed to confirm efficacy of devices.
Chronic migraine: Aside from valproate, interventions reduced migraines by 0.9 to 4.2 migraine days per month. Based on 2 small trials, valproate offered a large reduction of 13.2 migraine days per month, although this evidence was rated very low quality.
Sparse evidence for tricyclic antidepressants: Although commonly used as first-line therapy, placebo-controlled RCTs supporting efficacy are sparse (amitriptyline) or nonexistent (nortriptyline).
Key Findings: Head-to-Head RCTs:
Key evidence gaps: No direct comparisons of CGRP antagonists or devices to standard migraine prevention therapies exist.
Conclusions: Many interventions are effective for reducing migraine, although to what extent these reductions are clinical significant remains unclear. While the largest migraine reduction was seen for valproate (an older drug), most newer therapies appeared to have comparable efficacy and favorable tolerability. Head-to-head comparisons are needed to assess comparative effectiveness to support policy and treatment decisions.
©2021 Patient-Centered Outcomes Research Institute. For more information see www.pcori.org.