Acute episodic migraine: the 2025 ACP network meta-analysis and guideline

In March 2025, the American College of Physicians (ACP) published a network meta-analysis of 21 head-to-head and 165 placebo-controlled trials of acute treatments for episodic migraine ^[1]. The companion ACP guideline applies to adults with 1 to 14 headache days per month who are managed in outpatient settings ^[2]. This summary sets out what the guideline recommended, what the review found, and where the evidence remains thin.

What did the 2025 ACP guideline recommend?

The ACP guideline drew on evidence from the network meta-analysis, systematic reviews of patients' values and preferences, and comparative cost-effectiveness analyses ^[2]. The Clinical Guidelines Committee used the Grading of Recommendations, Assessment, Development and Evaluation (GRADE) approach to rate the certainty of evidence for pain freedom and pain relief at 2 hours, sustained outcomes up to 48 hours, rescue medication within 24 hours, and overall and serious adverse events ^[2]. Its first recommendation, graded strong on moderate-certainty evidence, is a triptan added to an NSAID for nonpregnant adults with moderate to severe attacks who do not respond adequately to an NSAID alone ^[2]. Its second recommendation, graded conditional on low-certainty evidence, is a triptan added to acetaminophen for nonpregnant adults with moderate to severe attacks who do not respond adequately to acetaminophen alone ^[2].

How do the drug classes compare on pain freedom?

In the ACP network meta-analysis, triptans had larger effects than NSAIDs (high certainty) or acetaminophen (low certainty) on pain outcomes at 2 hours and on pain freedom up to 48 hours ^[1]. The ACP review also found larger effects for triptan-NSAID combinations than for acetaminophen (low certainty), gepants (low certainty), NSAIDs (high certainty) and triptan monotherapy (moderate certainty) ^[1]. For triptan-acetaminophen combinations, the ACP review found larger effects than acetaminophen alone (moderate certainty) but not than triptans alone (low certainty) ^[1].

A 2024 BMJ network meta-analysis of 137 trials and 89,445 participants found that all 17 active oral drugs exceeded placebo for pain freedom at 2 hours, with odds ratios from 1.73 (95% CI 1.27 to 2.34) to 5.19 (95% CI 4.25 to 6.33) ^[3]. The BMJ authors reported that four triptans (eletriptan, rizatriptan, sumatriptan and zolmitriptan) were more efficacious than the newer ditan and gepant agents ^[3]. In a 2024 network meta-analysis of five trials in 3,004 people with insufficient triptan response, ditans and gepants raised 2-hour pain freedom compared with placebo (relative risk 1.93, 95% CI 1.52 to 2.46) without significant differences between agents ^[4].

What about adverse events and cost?

The ACP review reported that triptan regimens often had a higher risk of adverse events ^[1]. In the same review, one economic study found triptans more cost-effective than ditans and gepants ^[1]. The BMJ authors advised careful consideration of triptans in patients with a high risk cardiovascular profile ^[3].

Where is the evidence uncertain?

The ACP review limited its harms assessment to randomized trials ^[1]. The guideline authors captured additional harms data from U.S. Food and Drug Administration labels ^[2]. Many comparisons in the review lacked enough evidence for conclusions ^[1]. A 2026 network meta-analysis of 25 trials and 22,100 participants compared over-the-counter analgesics with a ditan and two gepants ^[5]. Three authors of that review were employed by one company, and two further authors received financial support from that company ^[5]. Its authors called for larger trials and patient-level analyses to refine efficacy and dosing ^[5].

At the point of care: what to check

• Attack frequency: the ACP guideline population is 1 to 14 headache days per month ^[2].
• Attack severity: both ACP recommendations apply to moderate to severe attacks ^[2].
• Pregnancy status: both ACP recommendations are limited to nonpregnant adults ^[2].
• Prior response: both ACP recommendations hinge on an inadequate response to an NSAID or acetaminophen alone ^[2].
• Cardiovascular profile: the BMJ authors flagged a high risk cardiovascular profile as a consideration for triptans ^[3].