PACAP in migraine research — and the antibody drugs aimed at blocking it

Migraine is not a vascular headache in the way that phrase was once used — the mechanism is more complex and more interesting than simple blood vessel dilation. But vascular biology is part of it, peptide signaling is central to it, and the last decade of migraine pharmacology has been defined by a breakthrough understanding of which peptides drive the cascade and how to block them.

The breakthrough was CGRP. Calcitonin gene-related peptide is a thirty-seven amino acid neuropeptide expressed in high concentrations in the trigeminal nerve — the cranial nerve responsible for sensation in the face and head, and the nerve whose activation is central to migraine pain. During a migraine attack, CGRP is released from the peripheral terminals of trigeminal neurons that innervate the meningeal blood vessels: the dura mater, the pia, the arachnoid, and their associated vasculature. CGRP causes vasodilation of those vessels, activates trigeminal pain fibers, and drives the sensitization cascade that produces the pulsating, position-sensitive, activity-worsening headache characteristic of migraine. Plasma CGRP levels rise during migraine attacks and fall when triptans — which cause vasoconstriction and inhibit trigeminal activation — abort the attack. The causal logic was compelling.

The drug development story that followed took years and multiple failed attempts before succeeding. Gepants — small-molecule CGRP receptor antagonists — were developed in the 2000s, ran into liver toxicity concerns with early compounds, and eventually yielded a new generation: ubrogepant and rimegepant for acute treatment, atogepant and rimegepant for prevention. Simultaneously, monoclonal antibodies against CGRP itself or its receptor advanced through clinical trials: erenumab targets the CGRP receptor, fremanezumab and galcanezumab and eptinezumab target the CGRP ligand. All four are now FDA-approved for migraine prevention, the first mechanistically targeted preventive therapies migraine has had. For patients who respond — roughly half of people with chronic migraine see a fifty percent or greater reduction in monthly migraine days — they represent a genuine shift in what migraine management looks like.

But roughly half don't respond adequately, and the question of what to do for them is where PACAP enters the story.

Pituitary Adenylate Cyclase-Activating Polypeptide — PACAP-38 — emerged as a parallel migraine trigger through a different research pathway. Researchers studying the trigeminovascular system noticed that PACAP, like CGRP, is expressed in trigeminal neurons and released from meningeal terminals. The key experiment was direct: intravenous infusion of PACAP-38 in people with a history of migraine. The results were unambiguous. Most participants developed a delayed headache, typically appearing one to six hours after infusion, that met diagnostic criteria for migraine — throbbing character, unilateral or bilateral, accompanied by photophobia and phonophobia, lasting hours and responding to triptans. Control subjects without migraine history had less frequent and less severe responses. PACAP was not just associated with migraine in indirect ways — it could directly provoke the syndrome in susceptible individuals, just as CGRP infusion could.

The receptor biology that underlies PACAP's migraine-triggering effect focuses on PAC1. PACAP binds three G-protein coupled receptors: VPAC1, VPAC2, and PAC1. PAC1 is highly selective for PACAP over VIP — the structurally related neuropeptide that shares VPAC1 and VPAC2 — and PAC1 is heavily expressed in the trigeminal ganglion, the sphenopalatine ganglion (the parasympathetic ganglion that contributes to migraine activation), and the trigeminal nucleus caudalis in the brainstem, which processes nociceptive signals from the head. When PACAP activates PAC1 on these structures, it causes vasodilation of dural vessels through mast cell degranulation and direct vascular effects, activates trigeminal nociceptors, promotes the release of CGRP itself from trigeminal terminals, and sensitizes the second-order neurons in the trigeminal nucleus that transmit pain signals centrally. PACAP's migraine-triggering effect is not independent of CGRP — it appears to work partly by stimulating CGRP release — but it also has CGRP-independent components, which is why PACAP blockade might help patients who don't respond to CGRP-targeting agents.

Lundbeck's development of Lu AG09222 followed the logic that if PACAP drives migraine in a mechanistically distinct way, neutralizing it should reduce migraine frequency. Lu AG09222 is a humanized monoclonal antibody that binds PACAP-38 directly, preventing it from activating any of its receptors. It was designed to intercept circulating PACAP before it reaches trigeminovascular targets, analogous to how fremanezumab and galcanezumab intercept circulating CGRP. The antibody has a long half-life appropriate for monthly subcutaneous dosing, similar to the approved CGRP antibodies.

Phase II trial data published in 2023 showed that monthly subcutaneous Lu AG09222 reduced monthly migraine days significantly compared to placebo in people with episodic migraine, with a responder rate — the proportion of patients achieving fifty percent or greater reduction — comparable to what has been observed with approved CGRP antibodies in equivalent trial designs. The tolerability profile appeared manageable. Phase III trials have been progressing. The compound is not yet FDA-approved, and the Phase III data will determine whether the Phase II signal holds in the larger populations and longer durations that regulatory approval requires.

The question of how PACAP blockade compares to CGRP blockade, and which patients might prefer one over the other, is still being worked out. The current hypothesis is that the mechanisms are complementary rather than overlapping: CGRP and PACAP are both released during migraine, both activate vasodilation and trigeminal sensitization, but through different receptor systems and with different upstream triggers. A patient whose migraine is driven primarily by CGRP release might respond well to CGRP antibodies. A patient in whom PACAP-driven pathways are more prominent — or in whom CGRP blockade is insufficient because PACAP-independent trigeminovascular activation continues — might respond to PACAP blockade where CGRP blockade fell short. This is speculative in the absence of head-to-head data and biomarker-guided patient selection, but it's the therapeutic logic that motivated the PACAP antibody program as a distinct development effort rather than a redundant one.

The paradox that anyone following PACAP biology encounters is the contrast between its migraine-triggering role and its extensively documented neuroprotective role in brain injury. In animal models of stroke, traumatic brain injury, and neurodegeneration, PACAP is protective — it reduces infarct volume, preserves neurological function, promotes cell survival and neuronal plasticity. Some researchers have explored intranasal PACAP delivery as a route to cognitive enhancement and neuroprotection, leveraging the olfactory pathway to deliver the peptide to the brain without systemic exposure. This research points toward PACAP as something valuable to the brain, something to support or supplement rather than block.

Both things are true, and the resolution lies in receptor distribution and tissue context. In the brain parenchyma — in neurons and astrocytes of the hippocampus, prefrontal cortex, cerebellum, and brainstem nuclei involved in cognition and sensory processing — PAC1 activation drives survival signaling, CREB activation, BDNF upregulation, and neuroplasticity. In the trigeminovascular system — in the sensory neurons of the trigeminal ganglion, the mast cells of the dura, the smooth muscle and endothelium of meningeal vessels — the same PAC1 activation drives vasodilation, neurogenic inflammation, and pain sensitization. The peptide carries the same message. The consequence depends entirely on who receives it.

A drug that neutralizes circulating PACAP-38, like Lu AG09222, would in principle reduce PACAP signaling throughout the body — including whatever beneficial signaling PACAP ordinarily does in brain tissue. The CGRP antibody programs raised a similar concern — CGRP is expressed in the heart and periphery, not just the trigeminovascular system, and blocking a pleiotropic peptide systemically always carries the theoretical risk of blocking its useful functions along with its harmful ones. The long-term cardiovascular monitoring data for the CGRP antibodies suggest the risk is manageable in the populations studied, but the monitoring has been ongoing and the field has been appropriately watchful. The same scrutiny will apply to PACAP antibody programs as they advance.

What the migraine pharmacology revolution of the last decade — from gepants to CGRP antibodies to the emerging PACAP antibody program — has established is a model for how complex peptide signaling in pain can be addressed with targeted molecular intervention. The model requires identifying the peptide, characterizing which receptors mediate the harmful versus the beneficial effects, and then designing drugs that intercept the molecule or its specific receptor in the tissue context where harm occurs. CGRP and PACAP were both discovered decades before their migraine relevance was understood. The science that translated that basic biology into clinical therapies took another generation.

The intranasal cognitive enhancement research and the migraine antibody research are, in a real sense, opposite bets on the same peptide: one tries to deliver more PACAP to the brain, the other tries to neutralize PACAP before it reaches trigeminovascular tissue. That both bets are being made simultaneously, by different research communities with different patient populations in mind, is less paradoxical than it first appears. The peptide doesn't have one job. The goal of the pharmacologist is to choose which job to support and which to interrupt — and to do so precisely enough that the intervention reaches its target without unacceptable effects on everything else the molecule does.