MOH NMA Liu 2025
(2025)Objective
To evaluate the comparative efficacy of different strategies for managing medication-overuse headache (MOH), focusing on reducing monthly headache days, via a network meta-analysis of randomized controlled trials.
Study Summary
• Restriction of overused medication + oral prevention + anti-CGRP(R) therapy (R+P+A) reduced monthly headache days by -8.47 (95% CI: -12.78 to -4.15)
• Headache prevention strategies—oral prevention (P), anti-CGRP(R) therapies (A), and botulinum toxin (B)—each significantly reduced monthly headache days, with no single prevention strategy superior to the others
• Abrupt withdrawal alone (W) was NOT effective: mean difference -2.77 (95% CI: -5.74 to 0.20)
Intervention
Network meta-analysis comparing withdrawal strategies, headache prevention (oral preventives, anti-CGRP(R) therapies, botulinum toxin), additional education, and bridging therapies (e.g., greater occipital nerve block) — alone or in combination — for medication-overuse headache.
Inclusion Criteria
Adults (≥18 years) meeting ICHD-2, ICHD-3, or ICHD-3 beta criteria for MOH, or with a primary headache disorder plus medication overuse, enrolled in randomized controlled trials comparing withdrawal, prevention, education, or neurostimulation strategies (alone or in combination), with reduction in monthly headache/migraine days as outcome.
Study Design
Arms: 16 RCTs with 3,000 participants total, comparing 12 strategies: control (C), abrupt withdrawal (W), oral prevention (P), botulinum toxin (B), anti-CGRP(R) therapy (A), W+P, W+B, W+Nb (greater occipital nerve block), W+E (education), R+P (restriction + prevention), R+E, W+P+Nb, W+P+E, R+P+A.
Patients per Arm: Sample sizes per included RCT ranged from 46 to 904; total N = 3,000 across 16 trials.
Outcome
• R+P+A: MD -8.47 (95% CI: -12.78 to -4.15)
• Oral prevention (P), anti-CGRP(R) (A), and botulinum toxin (B) monotherapies each significantly reduced headache days, with no clear winner among them
• Abrupt withdrawal alone (W): no significant efficacy, MD -2.77 (95% CI: -5.74 to 0.20)
• Combination therapies including anti-CGRP(R) and nerve blocks ranked highest by p-scores
Bottom Line
Combination therapies—particularly abrupt withdrawal + oral prevention + greater occipital nerve block, and restriction + oral prevention + anti-CGRP(R) therapy—are the most effective initial strategies for medication-overuse headache. Oral prevention, anti-CGRP(R) therapies, and botulinum toxin all significantly reduce monthly headache days with similar efficacy as monotherapies. Abrupt withdrawal alone is insufficient and should not be used in isolation.
Major Points
- Combination of abrupt withdrawal + oral prevention + greater occipital nerve block (W+P+Nb) produced the largest reduction in monthly headache days: MD -10.6 (95% CI: -15.03 to -6.16) vs control
- Restriction + oral prevention + anti-CGRP(R) therapy (R+P+A) also highly effective: MD -8.47 (95% CI: -12.78 to -4.15)
- Headache prevention strategies—oral prevention, anti-CGRP(R) therapies, and botulinum toxin—each significantly reduced monthly headache days, with no single agent superior to the others
- Abrupt withdrawal alone (W) was NOT effective: MD -2.77 (95% CI: -5.74 to 0.20)—no significant difference from control
- Greater reduction in headache frequency may lower MOH relapse risk, supporting use of more effective combination regimens as first-line therapy
- Findings support combination therapies—especially those incorporating anti-CGRP(R) or nerve blocks—as initial treatment options for MOH
Study Design
- Study Type
- Systematic review and network meta-analysis of randomized controlled trials
- Randomization
- No
- Sample Size
- 3000
- Follow-up
- Most studies (81.25%) reported outcomes at 8–16 weeks; three extended to 24 weeks; outcomes standardized to 12 weeks where possible
- Countries
- Iran, Denmark, Multicenter, Norway, Turkey, Brazil, Netherlands, Italy, United States
Primary Outcome
Definition: Reduction in monthly headache days (MHDs) compared to control, derived from network meta-analysis combining direct and indirect evidence
| Control | Intervention | HR/OR | P-value |
|---|---|---|---|
| Reference (MD = 0) | W+P+Nb: MD -10.6; R+P+A: MD -8.47; W alone: MD -2.77 (NS) | - (W+P+Nb: -15.03 to -6.16; R+P+A: -12.78 to -4.15; W alone: -5.74 to 0.20) |
Limitations & Criticisms
- Significant clinical and methodological heterogeneity across included studies (variability in patient populations, intervention types, and follow-up duration)
- Diagnostic criteria for MOH varied (ICHD-2 vs ICHD-3 vs ICHD-3 beta), evolving over the study period
- Two studies included participants with primary headache + medication overuse rather than strict MOH
- Outcome measurement timepoints varied (8–24 weeks); standardization to 12 weeks not always possible
- Limited number of studies per node in network may reduce precision of indirect comparisons
- Only English-language publications included, raising potential publication/language bias
- Anti-CGRP(R) and botulinum toxin had limited number of contributing trials
Citation
Koonalintip et al. The Journal of Headache and Pain (2025) 26:43