Approximately 7 million Americans have chronic migraine, meaning they experience ≥15 headache days per month, including ≥8 days per month that meet the diagnostic criteria for migraine with or without aura. More than half of people with chronic migraine use medications intended to relieve the symptoms of a migraine headache too frequently, a condition known as . Chronic migraine with medication overuse is associated with a high frequency of headaches, substantial functional impairment, high rates of comorbid medical disorders, and the potential for medication-related toxicity and side effects. There has been clinical equipoise regarding the optimal treatment strategy for patients who have chronic migraine with medication overuse.
To compare 2 patient-centered management strategies for patients who have chronic migraine with medication overuse: migraine-preventive medication and (1) switching from the overused medication to an alternative used on a limited number of days with a limited frequency or (2) continuation of the overused medication with no maximum limit on the number of days used.
This was a prospective, randomized, longitudinal, open-label, pragmatic clinical trial. A total of 720 adult patients who had chronic migraine with medication overuse according to International Classification of Headache Disorders Third Edition beta (ICHD-3β) diagnostic criteria were enrolled from headache specialty, general neurology, and primary care clinics across the United States between February 20, 2017, and December 22, 2020. Participants were randomized 1:1 to 1 of 2 treatment strategies, each designed according to the advice of clinician and patient stakeholders, as follows: (1) migraine-preventive medication and immediate switching from the overused medication to an alternative medication in a different class used ≤2 days per week vs (2) migraine-preventive medication and continuation of the overused medication with no maximum limit on the number of days used. The primary aim was, with both study arms receiving migraine-preventive medications, to see if continuing the overused medication was noninferior to stopping it and switching to a drug from another class. The primary outcome was the number of moderate to severe headache days during weeks 9 to 12 postrandomization. If noninferiority was demonstrated (ie, mean difference of less than 1.5 moderate to severe headache days per 4 weeks), we would conduct a superiority analysis of number of moderate to severe headache days during the first 2 weeks postrandomization. A heterogeneity of treatment effect analysis for type of overused medication was performed.
A total of 720 patients were randomized. The average (SD) patient age was 44 (13) years, 87.5% of the patients were female, the average (SD) duration of migraine was 23 (14) years, and the average (SD) duration in a medication overuse pattern was 4.7 (5.9) years. Patients had averaged (SD) 22.5 (5.1) headache days per 4 weeks, including 12.8 (6.7) moderate to severe headache days, and used symptomatic medications on 21.4 (5.8) days per 4 weeks. The most commonly overused medications were simple analgesics (64% of patients), combination analgesics (39% of patients), and triptans (21% of patients). At weeks 9 to 12 postrandomization, migraine-preventive medication without switching or limiting the overused medication was noninferior to migraine-preventive medication with switching off the overused medication to an alternative drug used ≤2 days per week. The switching group averaged (SD) 9.3 (7.2) moderate to severe headache days per 28 days vs 9.1 (6.8) days in the group without switching ( = 0.75; difference = 0.2; 95% CI, −1.0 to 1.3). Since no-switching was noninferior to switching during weeks 9 to 12, a comparison was made during weeks 1 to 2. There were no differences in the efficacies of the treatment strategies during the first 2 weeks postrandomization, with an average (SD) of 6.6 (3.7) moderate to severe headache days per 14 days in the switching group vs 6.4 (3.6) in the no-switching group (Δ 0.2; minimal clinically important difference, −1.5; = 0.57; difference = 0.2; 95% CI, −0.4 to 0.7). Patients who were randomized to switch off the overused medication were less likely to be in a medication overuse pattern between weeks 9 and 12 (53% vs 73%; < .001; difference = −0.19, 95% CI, −0.27 to −0.11). The effect of switching did not differ between those overusing opioids or butalbital vs those not overusing these medications, but the CI was wide (difference = 1.2 days; 95% CI, −5.0 to 2.7).
When treating patients who have chronic migraine with medication overuse and using an outcome of moderate to severe headache frequency during the first 12 weeks, migraine-preventive medication without switching off or limiting the frequency of taking the overused medication is noninferior to migraine-preventive medication with switching off the overused medication to an alternative used ≤2 days per week.
Baseline headache and medication intake frequencies were determined according to patient recall. Although possibly impacted by recall bias, this methodology is consistent with how information is collected in clinical practice. The clinical trial did not include a treatment strategy that recommended discontinuation of the overused symptomatic medication without switching to an alternate symptomatic therapy and not using migraine-preventive therapy. Although this treatment strategy might be recommended by some clinicians, patient partners and other stakeholders considered this strategy to lack patient centricity and strongly recommended against its inclusion in this clinical trial. The most common reason that potentially eligible patients were not randomized was an unwillingness to be randomized to the treatment strategy that required switching off the overused symptomatic medication (n = 77). Although this could be considered a study limitation, it reflects actual clinical practice in which treatment strategies are advisable only when the patient agrees with the treatment recommendations.
