• opioid;
  • triptan;
  • migraine;
  • placebo;
  • rizatriptan;
  • pooled analysis

Background.— Limited evidence suggests that, in individuals with migraine, prior use of opioids reduces responsiveness to treatment with subsequent acute migraine therapies. The evidence is more robust with regard to opioids as a risk factor for chronic migraine.

Objectives.— To explore whether recent prior opioid use influenced treatment response, in a post hoc pooled analysis of rizatriptan clinical trials.

Methods.— We included rizatriptan 10 mg and placebo data from phase 3 moderate/severe migraine studies as well as from the rizatriptan “Treat a Migraine Early” (TAME) studies. As part of the clinical assessment, medication usage for migraine and other reasons in the 30 days prior to a screening visit and up to the time of taking study drug was captured via patient's self-report. The influence of recent prior opioid use on the endpoint of pain freedom at 2 hours was assessed via logistic regression. We further explored the influence of gender and disability on treatment response.

Results.— In the moderate/severe migraine studies of 2068 individuals treated with rizatriptan, 284 (13.7%) reported recent prior use of opioids. Of those treated with placebo (1258), 12.5% recently used opioids. In the TAME studies, the proportions were lower, 3.9% and 5.3%, respectively. The pretreatment and demographic characteristics were similar across the study groups. In the moderate/severe studies, recent prior opioid use was associated with reduced 2-hour pain freedom. Although the influence of recent prior opioid use was assessed independently of treatment, the finding was driven primarily by rizatriptan (recent prior use vs no use: 34% vs 42% for rizatriptan and 10% vs 10% for placebo; P = .013). In the TAME studies, recent prior opioid use was also predictive of reduced efficacy (recent prior use vs no use: 41% vs 59% for rizatriptan and 13% vs 32% for placebo; P = .007).

Conclusion.— Recent prior opioid use was associated with lower triptan response. Because of the post hoc nature of the analysis and limitations in capturing amount of opioid used, as well as to adjust for disability levels, these findings require replication in prospective studies.