Sequencing CGRP Therapies
With multiple CGRP-targeting agents now available — four monoclonal antibodies and two preventive gepants — clinicians face practical questions about ordering, switching, and combining these therapies. No head-to-head trials exist between any two CGRP agents, so sequencing decisions rely on consensus guidance, pharmacologic reasoning, and emerging real-world data.
Bottom Line
- Failure of one CGRP mAb does not mean failure of the class — ~30-40% of non-responders to one agent respond to another
- Switch mechanism when possible: If a receptor-targeted agent (erenumab) fails, try a ligand-targeted agent (fremanezumab, galcanezumab, eptinezumab), and vice versa
- Assess response at 3 months before switching (AHS consensus)
- CGRP mAb + onabotulinumtoxinA can be combined for refractory chronic migraine — different mechanisms, growing evidence of additive benefit
- Do not combine a CGRP mAb with a preventive gepant — same target, no evidence of benefit, potential for increased side effects
Defining CGRP Therapy Failure
Before switching, ensure the current agent has been given an adequate trial:
- Minimum duration: 3 monthly doses (mAbs) or 3 months of daily/EOD gepant use (AHS consensus)
- Adequate dose: Some patients respond to dose escalation — erenumab 70 mg → 140 mg; eptinezumab 100 mg → 300 mg; atogepant 30 mg → 60 mg
- Partial response (some improvement but insufficient): Consider adding Botox or optimizing acute treatment before switching
- No response at 3 months at maximum tolerated dose: Switch agents
What Counts as a Meaningful Response?
- ≥50% reduction in monthly migraine days is the traditional clinical trial threshold
- In practice, a 30-50% reduction may be clinically meaningful, especially in chronic migraine or highly refractory patients
- Also assess: reduction in attack severity, fewer acute medication days, improved function (MIDAS/HIT-6), and patient satisfaction
- Some patients are "slow responders" — benefit increases through months 3-6. If there is a trend toward improvement at 3 months, continuing another 3 months is reasonable.
Switching Within the CGRP mAb Class
Evidence for Switching
Multiple real-world studies demonstrate that switching between CGRP mAbs produces meaningful responses in a substantial proportion of initial non-responders:
- Overeem et al. (2022): 78 patients who failed erenumab switched to galcanezumab or fremanezumab — 41% achieved ≥50% response
- Souza et al. (2023): Retrospective analysis of 154 patients switching between mAbs — 38% became responders after switch
- Ziegeler et al. (2023): Patients switching from erenumab to fremanezumab showed improvement even when erenumab failed completely
Switching Strategy
| Current Agent | Suggested Switch | Rationale |
|---|---|---|
| Erenumab (receptor) | Fremanezumab, galcanezumab, or eptinezumab (ligand) | Different target — ligand-targeted mAbs may capture CGRP that erenumab misses at non-canonical receptors |
| Ligand-targeted mAb | Erenumab (receptor) or another ligand-targeted mAb | Receptor blockade may be more effective for some patients; also, different antibody pharmacokinetics |
| SC mAb (any) | Eptinezumab (IV) | 100% bioavailability, fastest onset; useful if adherence or absorption is a concern |
| Monthly dosing | Fremanezumab 675 mg quarterly or eptinezumab IV quarterly | Reduces injection burden; may improve adherence |
Timing of Switch
- No washout period needed between CGRP mAbs — start the new agent when the next dose of the previous agent would have been due
- If switching from a monthly mAb to eptinezumab IV, schedule the infusion at the next monthly dose date
- Some clinicians administer a loading dose of the new mAb (e.g., galcanezumab 240 mg loading) at the switch
Switching Between mAbs and Gepants
| Scenario | Approach | Notes |
|---|---|---|
| mAb failed → try gepant | Start atogepant or rimegepant after mAb washout (~1 month after last SC dose) | Gepants block the receptor directly; may work when ligand-targeted mAbs fail. Small molecule penetrates CNS. |
| Gepant failed → try mAb | Start mAb immediately (no washout needed due to gepant's short half-life) | mAbs provide continuous CGRP blockade vs intermittent gepant coverage; sustained levels may overcome partial gepant failure |
| Patient prefers oral → injectable | Switch from gepant to mAb for adherence reasons | Monthly/quarterly injection vs daily pill; same pathway |
| Pregnancy planning | Switch from mAb to gepant (rapid washout) or stop both | Gepant clears in hours vs months for mAbs |
Combining CGRP Therapy With OnabotulinumtoxinA
This is the most well-supported combination strategy for refractory chronic migraine. OnabotulinumtoxinA and CGRP mAbs act through different mechanisms — Botox inhibits release of CGRP and other neuropeptides from trigeminal afferents, while mAbs block CGRP in the extracellular space.
Evidence
- Blumenfeld et al. (2021): Retrospective analysis of 257 CM patients on Botox who added erenumab — additional -4.4 migraine days/month reduction beyond Botox alone
- Ailani et al. (2022): Prospective cohort of Botox partial responders who added fremanezumab — 54% achieved ≥50% response with combination vs 12% with Botox alone
- Real-world data consistently shows additive benefit: patients who have partial response to Botox gain an additional 3-5 migraine day reduction when adding a CGRP mAb
Practical Combination Approach
- Combination is most appropriate for chronic migraine with partial response to either agent alone
- Start one agent first, assess response at 3 months, then add the second if partial response
- No dose adjustment needed for either agent when used together
- Insurance may require documentation of partial response to each agent individually before approving combination
- Gepants can also be combined with Botox (different mechanism); however, do not combine a gepant with a CGRP mAb
Combinations to Avoid
Do Not Combine These
- CGRP mAb + preventive gepant: Both block the same pathway. No evidence of additive benefit. Potential for increased constipation, hypertension, and cost without clinical gain.
- Two CGRP mAbs simultaneously: No rationale; increased immunogenicity risk without expected benefit
- Acute gepant use on days when CGRP mAb is active: Technically safe but may have diminished acute gepant efficacy since CGRP is already being blocked. Most experts allow acute rimegepant or ubrogepant for breakthrough attacks on mAb therapy, as clinical data supports this.
Suggested Sequencing Algorithm
| Step | Action | Assess At |
|---|---|---|
| 1 | First CGRP agent (mAb or gepant based on patient preference, insurance, comorbidities) | 3 months |
| 2 | If inadequate: dose optimize (e.g., erenumab 70 → 140 mg, atogepant 30 → 60 mg) | 3 months |
| 3 | If still inadequate: switch to different CGRP agent (different mechanism if possible) | 3 months |
| 4 | If partial response to mAb: add onabotulinumtoxinA (CM only) | 2-3 Botox cycles (6-9 months) |
| 5 | If all CGRP agents fail: reassess diagnosis, address MOH, consider traditional preventives or clinical trials | — |
Insurance and Access Considerations
- Most US payers require failure of 2-3 oral preventives (from different classes) before approving CGRP mAbs
- Gepants (atogepant, rimegepant) sometimes have fewer step therapy requirements as oral agents
- Switching within the mAb class often requires a new prior authorization
- Manufacturer patient assistance programs can bridge gaps in coverage
- Document each failed preventive with specific dates, doses reached, reason for discontinuation (side effects vs lack of efficacy), and duration of adequate trial
References
- American Headache Society. Consensus statement: integrating new migraine treatments into clinical practice. Headache. 2019;59(1):1-18.
- Overeem LH, et al. Switching between CGRP monoclonal antibodies: a real-world analysis. Cephalalgia. 2022;42(11-12):1153-1162.
- Souza MNP, et al. Effectiveness of switching CGRP monoclonal antibodies: a systematic review and meta-analysis. Headache. 2023;63(7):890-901.
- Blumenfeld AM, et al. Real-world evidence for the addition of CGRP monoclonal antibodies to onabotulinumtoxinA treatment for migraine. Headache. 2021;61(8):1246-1256.
- Ailani J, et al. Combination onabotulinumtoxinA and CGRP monoclonal antibody therapy for chronic migraine. Headache. 2022;62(1):42-53.
