Home > Discussion - Migraine in Children:...

PubMed Health. A service of the National Library of Medicine, National Institutes of Health.

Shamliyan TA, Kane RL, Ramakrishnan R, et al. Migraine in Children: Preventive Pharmacologic Treatments [Internet]. Rockville (MD): Agency for Healthcare Research and Quality (US); 2013 Jun. (Comparative Effectiveness Reviews, No. 108.)

  • This publication is provided for historical reference only and the information may be out of date.

This publication is provided for historical reference only and the information may be out of date.


Our comprehensive review identified limited evidence about benefits and harms with migraine preventive drugs in children. Migraine prevention in children was examined in 24 publications of 22 RCTs that enrolled 1,578 children. Only one drug, the beta blocker propranolol, prevented migraine more effectively than placebo (Table 14). Propranolol (60 to 120 mg/day) would result in complete cessation of migraine attacks in 713 per 1,000 treated children (95% CI, 452 to 974) (low-strength evidence from a single RCT).79 Topiramate, divalproex, clonidine, trazodone, and magnesium oxide failed to prevent migraine in children. Divalproex sodium, 1000 mg/day, resulted in greater rates of treatment discontinuation due to adverse effects (Table 12). Topiramate significantly increased risk of weight loss, paresthesia, and upper respiratory tract infection (Table 13).

Table 14. Evidence of migraine prevention in children.

Table 14

Evidence of migraine prevention in children.

No studies examined whether specific characteristics of children modify the effectiveness or safety of preventive drugs. Treatment effects may differ between children and adolescents. Published trials did not provide treatment effects among age subgroups.

Clinical decisions about drugs and nonpharmacologic treatment for migraine prevention in children should include a careful estimation of the balance of benefits and harms. Our review confirmed previously published conclusions about the efficacy of propranolol for migraine prevention in children.98 However, nonpharmacologic treatments demonstrated better benefit-to-harm ratios than drugs in head-to-head RCTs.75,77,99 Individualized multimodal drug management showed promising results.89 Other complex disease management interventions including school-based psychological interventions and drug management programs have both demonstrated positive results in treating acute headache attacks, but neither has been examined for migraine prevention.100,101 RCTs have not yet examined other drug management interventions, including integrated care, coordinated care, patient education, drug surveillance, and interactive drug monitoring.

The off-label antiepileptic drugs, clonidine and trazodone failed to demonstrate efficacy for migraine prevention but resulted in bothersome adverse effects. Previously published reviews reported bothersome adverse effects with antiepileptic drugs in children with migraine102,103 or epilepsy.104 Off-label use of the antidepressant trazodone in children with migraine was not effective. We could not conclude the effectiveness of other antidepressants for preventing migraine in children, nor could we determine whether adverse effects of antidepressants are similar when used for children with migraine compared to children with depression. We do know that antidepressants may increase risk of suicidal behavior in children and adolescents.105 Use of off-label psychotropic drugs for migraine prevention could be justified in children with psychiatric comorbidity;106 however, trials available for review did not report presence of comorbid illnesses in enrolled patients.

In fact, few available trials examined the seriousness or bothersomeness of harms with drugs. Clinicians considering off-label drugs for children with migraine have very limited evidence about balance between benefits and harms for informed decisionmaking. Few clinical trials followed the recommendations from the Task Force on Adverse Events in Migraine Trials of the International Headache Society107 when testing safety of the drugs in children. Future fully powered trials involving children with migraine should examine long-term safety with preventive drugs, regardless of the investigators' perceptions about the causal association between the drugs and the detected harms.

Strength of evidence of drug benefits and harms was low in most cases due to risk of bias and imprecise estimates from underpowered RCTs. Reporting quality of trials was poor with few trials providing detailed information about prior or concomitant treatments, comorbidities, family history, socioeconomic status, drug overuse, and other important characteristics of children. On average, the trials lasted 20 weeks, and therefore did not provide sufficiently long-term evidence for benefits and harms with drugs that could be recommended for preventive use over very long time periods. The optimal duration of preventive treatment and sustained benefits and harms with preventive drugs in children with migraine remain unclear.28,108

Our review has limitations. We did not conclude strength of evidence for flunarizine, which has been reported to be effective in preventing migraines in children, because this drug has not been approved by the FDA. One low-risk-of-bias Italian RCT suggested that flunarizine resulted in ≥50 percent reduction in migraine attacks in 500 children per 1,000 treated (95% CI, 260 to 740).109 We do not know why flunarizine was never approved in the United States. We requested the FDA review of this drug and received a response that stated: “Any information on an application if submitted by a firm to the FDA that did not yet receive approval, belongs to the manufacturer/sponsor developing the drug (21 CFR 314.430).” We did not contact the sponsors directly to inquire about products under development. Comprehensive review of nonpharmacologic treatments was beyond our scope.

Our comprehensive literature search in several databases, trial registries, and the FDA reviews detected a very low publication rate of registered completed clinical trials involving children. We do not know why the studies were not published. We assume publication bias but did not contact the investigators of completed trials for unpublished data. We requested additional data from the sponsors of completed trials but received few responses. Thus, we know neither the results from unpublished trials nor how many unregistered studies have been conducted and never published. We relied on reported information and did not contact study authors for additional details about the trials, including design, execution, or poorly reported results that we could not reproduce.

Key Messages

  • Propranolol was more effective than placebo for preventing migraine in children, with no bothersome adverse effects that could lead to treatment discontinuation.
  • Antiepileptics were no more effective than placebo in preventing migraine but resulted in increased risk of adverse effects.
  • Internet-based self-management with multimodal CBT was better than education in preventing migraine in children and adolescents at 6 weeks but not at 6 months of followup.
  • Reporting quality of studies involving children is poor.

Our report offers insights for future research on preventive treatments for childhood migraine, all of which should be conducted according to the recently published Standards for Research in Child Health group (Table 15).110-115 Future randomized trials should examine the comparative effectiveness of multimodal drug and disease management; long-term benefits, safety, and adherence with preventive treatments; and the role of specific characteristics of children that could modify benefits and harms with preventive drugs.

Table 15. Future research needs.

Table 15

Future research needs.

Future studies should also specifically examine the effects and risks of off-label drug use for migraine prevention in children. Randomized trials have examined only a few pharmacologic agents, but practicing clinicians use many off-label drugs to treat children, and little is known about the comparative effectiveness or safely of the drug classes used. Large observational studies, including the American Migraine Prevalence and Prevention study, relied on self-reported use of preventive medications and did not assess exact drug use or effectiveness.1 The few available studies of off-label drug use in children show that 5 percent of all antiepileptic drug prescriptions were for migraine.116 The National Ambulatory Medical Care Surveys from 2001 to 2004 demonstrated that 62 percent of outpatient pediatric visits included off-label prescribing; 86 percent of those prescriptions were for pain.117 European studies demonstrated that about 30 percent of hospitalized children118 and 40 percent of children in outpatient settings received off-label drug prescriptions.119 European observational studies found a significantly higher risk of adverse effects with off-label drugs, concluding an improper balance of benefits and risks with off-label drugs in pediatric patients.119,120

As a first step, the comparative effectiveness and safety of off-label drugs used for migraine prevention in children should be examined by analyzing administrative databases. Such analyses could, for example, shed light on practice patterns in migraine prevention and provide insight into the comparative effectiveness of preventive drugs for reducing visits to emergency rooms. Based on these analyses, RCTs could be designed to examine the drugs that demonstrate the most favorable ratios of benefits to harms. Future studies should also assess the efficacy and comparative effectiveness of multidisciplinary migraine drug management interventions based in the community, family, or school.

Trials should examine treatment effectiveness based on patient-centered outcomes, including complete cessation of migraine attack or reduction in monthly migraine attacks by ≥50 percent, quality of life, migraine-related disability, and success with schooling. Studies should examine subgroups of children identified by age, sex, race, ethnicity, puberty, socioeconomic status, family history, and comorbid psychiatric conditions and other diseases. Studies should also examine the effectiveness and safety of preventive drugs and multidisciplinary interventions in children with chronic migraine, migraine with versus without aura, and those for whom previous preventive treatments have failed.

Existing clinical research policy does not guarantee availability of the results from all studies involving children. Results are unavailable for more than half of the studies involving children, revealing a substantial publication bias.121 Registration and posting of results on ClinicalTrials.gov should be mandatory for all studies involving children.121

Cover of Migraine in Children: Preventive Pharmacologic Treatments
Migraine in Children: Preventive Pharmacologic Treatments [Internet].
Comparative Effectiveness Reviews, No. 108.
Shamliyan TA, Kane RL, Ramakrishnan R, et al.


AHRQ (US Agency for Healthcare Research and Quality)

PubMed Health Blog...

read all...

Recent Activity

Your browsing activity is empty.

Activity recording is turned off.

Turn recording back on

See more...