Molecular triggers of migraine aura: A systematic review of human pharmacological provocation studies

Abstract

Background

Migraine aura reflects transient neurological disturbances attributed to cortical spreading depolarization (CSD), yet the upstream molecular events that promote or precipitate this phenomenon remain uncertain. Human pharmacological provocation models provide a controlled approach to identifying endogenous signaling pathways capable of triggering aura and thereby clarifying their mechanistic role in migraine pathogenesis.

Methods

This systematic review was pre-registered in PROSPERO (ID: CRD420250636266) and conducted in accordance with the Synthesis Without Meta-Analysis (SWiM) guideline. PubMed and Embase were searched from database inception through January 1, 2026, without language restrictions, to identify experimental studies administering pharmacological agents to individuals with migraine with aura under controlled lab conditions. Two reviewers independently performed study selection, data extraction, and methodological assessment. Extracted variables included participant characteristics, study design, pharmacological trigger, dosing protocol, and the incidence, timing, and phenotype of provoked aura and headache. Because of substantial heterogeneity, findings were synthesized qualitatively.

Results

Fourteen studies met inclusion criteria, examining seven pharmacological agents: calcitonin gene-related peptide (CGRP), levcromakalim (ATP-sensitive potassium (K_ATP) channels opener), glyceryl trinitrate (GTN), sildenafil, cilostazol, endothelin-1, and histamine. Across studies, aura induction occurred far less frequently than headache induction. CGRP provoked aura in 17 (32%) of 53 participants across three studies, with latencies ranging from 10 to 360 min. Pharmacological opening of vascular K_ATP channels by levcromakalim elicited aura in 14 (27%) of 52 participants in randomized crossover trials, with onset between 20 and 120 min. Other agents produced minimal or no aura responses.

Conclusions

Human pharmacological provocation provides a reproducible framework for dissecting molecular pathways that can trigger migraine aura. Current evidence implicates CGRP signaling and vascular ATP-sensitive potassium channel activation as the most plausible associated candidate pathways. However, standardized protocols and larger controlled studies are required to confirm these mechanisms and refine experimental models of aura biology.

Trial Registration

PROSPERO (ID: CRD420250636266).

Graphical abstract

This is a visual representation of the abstract.

Keywords

migraine aura cortical spreading depolarization trigger provocation pathophysiology

Introduction

Migraine with aura is characterized by transient neurological disturbances that develop over minutes and reflect the propagation of cortical spreading depolarization (CSD).^1,2 This slow wave of neuronal and glial depolarization travels at 2 to 5 mm per minute, presumably creating the spatial gradient that shapes aura progression.³ As the wavefront advances, it disrupts cortical signaling, and its passage is followed by transient suppression of neural activity, which are thought to correspond to the nature of transient positive and negative neurological disturbances reported during aura.⁴ These linked processes produce the characteristic temporal and spatial evolution of spontaneous aura.

CSD also initiates downstream events relevant to migraine pathogenesis. Preclinical studies show that it activates the trigeminovascular system, promotes the release of pro-inflammatory mediators, and induces meningeal vasodilation, suggesting a biological bridge between aura and migraine headache.^5–8 However, the physiological conditions that permit or trigger CSD in humans remain uncertain. This uncertainty limits understanding of why aura precedes, follows, or overlaps with headache in individual attacks. As a result, the temporal relationship between these phenomena remains incompletely defined, with evidence supporting sequential, bidirectional, or parallel interactions.^9–12

Progress in clarifying these interactions requires direct observation of aura physiology, yet spontaneous aura rarely occurs under conditions that allow standardized study. Its brief duration and unpredictable onset restrict opportunities for systematic measurement in clinical or research settings.¹³ Human provocation models were developed to address this challenge by enabling safe and reproducible induction of migraine aura in controlled laboratory settings.^14,15 Early protocols focused on provoking migraine headache alone, but recent approaches now test whether pharmacological agents can also trigger aura in participants with migraine.^14,16–18 This shift reflects increasing interest in identifying molecular mechanisms that initiate CSD or reduce the threshold for its occurrence.

Pharmacological provocation supports this effort by providing a controlled platform for mechanistic investigation.¹⁵ It enables precise characterization of temporal relationships between agent exposure and symptom onset, offering insights unavailable in spontaneous attacks. It also allows structured comparison across distinct molecular triggers, thereby revealing pathway-specific effects on cortical excitability. Interestingly, non-pharmacological stimuli such as hypoxia, visual stimulation, or carotid puncture can induce aura, but their mechanisms are diffuse and difficult to isolate.^19,20 Pharmacological agents, in contrast, engage defined biological pathways with reproducible dosing.²¹ Nonetheless, consistent aura induction remains uncommon, and existing studies vary in design, participant selection, and outcome reporting, complicating interpretation.

These limitations underscore the need for a comprehensive synthesis of existing evidence. This systematic review integrates all published human provocation studies assessing aura induction in participants with migraine with aura, summarizes their principal findings, and evaluates their implications for understanding migraine aura biology.

Methods

This systematic review was pre-registered in the International Prospective Register of Systematic Reviews (PROSPERO, CRD 420250636266) and conducted in accordance with the Synthesis Without Meta-analysis (SWiM) guideline.²²

Overview of the human provocation model

The human provocation model provides a reproducible and controlled framework for investigating migraine mechanisms under lab conditions.²¹ It is founded on the principle that administering signaling molecules capable of inducing migraine attacks in persons with migraine can reveal molecular mechanisms central to disease pathogenesis.¹⁶ The reversible and treatable nature of migraine attacks allow these responses to be studied safely in lab settings. Initially developed to examine migraine headache, the model has been refined to determine whether pharmacological agents can also trigger migraine aura in individuals diagnosed with migraine with aura.^14,17 This adaptation enables systematic evaluation of aura attacks that are otherwise sporadic and short-lasting, thereby facilitating direct observation of its biological correlates.

Experimental protocols most often apply randomized, double-blind, placebo-controlled crossover designs, although open-label studies have been conducted in exploratory contexts. Participants must be free from both aura and headache and abstain from acute medications for at least 24 to 48 h before provocation.^17,23

Baseline evaluations include vital parameters, aura and headache status, and accompanying symptoms.¹⁷ Following intravenous or oral administration of the provoking agent, participants are observed at regular intervals—typically every ten minutes—for 1 to 2 h and continue symptom recording on an hourly basis for up to 12 h after exposure.

An experimentally induced migraine headache is defined as an attack fulfilling International Classification of Headache Disorders (ICHD-3) criteria C and D for migraine without aura, or one replicating the participant's usual spontaneous migraine headache.²⁴ An experimentally induced aura is defined as the occurrence of neurological symptoms meeting ICHD criteria B and C for migraine with aura.²⁵

Search strategy

A systematic search of PubMed and Embase was conducted to identify clinical investigations using the human provocation model to study the induction of migraine aura. The search encompassed all available records from database inception through January 1, 2026, without language restrictions. No publication date restrictions were applied. Both free-text keywords and Medical Subject Headings (MeSH) terms were combined to capture studies addressing migraine with aura and pharmacological provocation paradigms. The complete PubMed and Embase search strings are provided in Supplemental Table 1. In addition, the reference lists of all included studies and relevant systematic reviews were examined manually to ensure coverage of the available evidence.

Selection criteria

Studies were eligible for inclusion only if they enrolled participants with a confirmed diagnosis of migraine with aura according to ICHD criteria and prospectively evaluated aura occurrence following administration of a defined pharmacological trigger under controlled lab conditions.

Two independent reviewers (ZH and BAC) systematically screened all titles and abstracts to determine potential eligibility. Full-text articles were retrieved for studies meeting inclusion criteria or for which eligibility could not be confirmed from the abstract. Disagreements between reviewers were resolved through discussion or, when necessary, consultation with a third reviewer (HMA). The overall selection process is illustrated in the Preferred Reporting Items for Systematic reviews and Meta-Analyses (PRISMA) flow diagram (Figure 1), and detailed eligibility criteria are provided in Supplemental Table 2.

Figure 1.

Study flow diagram.

Data extraction

Data were independently extracted by two reviewers (ZH and BAC) using a pre-specified standardized form. Extracted variables encompassed study-level characteristics, including authorship, publication year, and design, as well as participant demographics such as sample size, age, and sex. Information regarding the molecular triggers used was recorded in detail, including compound type, administered dose, and route of administration. For each provocation study, data on headache induction were collected, encompassing attack intensity, quality, localization, associated symptoms, duration, and time to onset. Aura-related variables were similarly extracted, including subtype, timing, anatomical distribution, duration, and detailed symptom descriptions. Discrepancies between reviewers were resolved through discussion and, when necessary, adjudicated by a third reviewer (HMA).

Data synthesis

Because of substantial heterogeneity across study designs, participant populations, and intervention protocols, a quantitative meta-analysis was not feasible. Instead, a structured qualitative synthesis was performed, integrating study characteristics, methodological parameters, and primary findings into both tabular and narrative formats. To facilitate interpretability, studies were organized according to the pharmacological trigger investigated.

Risk of bias was independently assessed by two reviewers in accordance with the Cochrane Handbook for Systematic Reviews of Interventions, evaluating random sequence generation, allocation concealment, blinding, incomplete outcome data, selective reporting, and other sources of bias.²⁶

Results

A total of 2007 records were identified across PubMed and Embase. After removing 463 duplicates, 1544 titles and abstracts were screened. Following full-text review and manual reference searches, 14 studies fulfilled the eligibility criteria and were included in the qualitative synthesis^{18,25,27–38} (Figure 1). The included studies were published between 1999 and 2024.

Sample sizes for participants diagnosed with migraine with aura ranged from 3 to 34 (Figure 2). Thirteen studies examined aura induction in individuals with migraine with aura accompanied by headache,^{18,25,27–33}^,35–38 and one enrolled participant who experienced aura without headache.³⁴ Six studies applied randomized, double-blind, placebo-controlled crossover designs,^32–37 whereas six were open-label,^{18,25,27,28,30,31} one was double-blind,³⁸ and one was single-blind.²⁹ Overall, double-blind randomized studies were judged at low risk of selection, performance, and detection bias, whereas open-label or uncontrolled studies were judged at higher risk (Supplementary Table 3).

Figure 2.

Aura induction rates across pharmacological triggers. GTN: glyceryl trinitrate, CGRP: calcitonin gene related peptide, ET-1: endothelin-1, n: number of participants, %: percentage.

Across these 14 studies, seven pharmacological agents were investigated for their ability to trigger migraine aura: glyceryl trinitrate (GTN) (n = 4),^18,25,27,28 calcitonin gene-related peptide (CGRP) (n = 3),^29–31 levcromakalim (n = 3),^32–34 cilostazol (n = 1),³⁵ sildenafil (n = 1),³⁶ endothelin-1 (n = 1),³⁷ and histamine (n = 1).³⁸ Intravenous infusion represented the most frequent route of administration (n = 10),^18,27–34^,37 followed by oral (n = 2),^35,36 sublingual (n = 1),²⁵ and inhalational (n = 1) delivery.³⁸ All studies used diagnostic criteria from the ICHD: three applied ICHD-1,^18,25,38 two ICHD-2,^27,29 four ICHD-3β,^28,35–37 and five ICHD-3 criteria.^30–34

Medication policies varied across studies. Two allowed continuation of stable preventive therapy,^30,31 seven required no preventive medication use at the time of participation,^{18,25,27,32,33,35,36} and five did not specify medication status.^{28,29,34,37,38} General study characteristics and principal findings are summarized in Table 1 –3.

Table 1.

Characteristics and outcomes of pharmacological provocation studies in migraine with aura using nitric oxide donors, phosphodiesterase inhibitors, endothelin-1, and histamine.

Author (Year)	Study design	Participants, n	Female, n (%)	Trigger	Exposure	Route	Duration	Provoked aura, n (%)	Applied criteria*
Christiansen et al. (1999)¹⁸	Non-Randomized Trial	12	8 (67)	GTN	0.5 μg/kg/min	IV	20 min	0 (0)	NA
Afridi et al. (2004)²⁷	Non-Randomized Trial	21	17 (81)	GTN	0.5 μg/kg/min	IV	20 min	2 (9)	NA
Karsan et al. (2020)²⁸	Non-Randomized Trial	32	NA	GTN	0.5 μg/kg/min	IV	20 min	7 (22)	NA
Sances et al. (2004)²⁵	Non-Randomized Trail	22	NA	GTN	0.9 mg	Sublingual	NA	3 (14)	NA
Butt et al. (2018)³⁵	Double-Blind RCT	16	14 (88)	Cilostazol	200 mg	Oral	NA	1 (6)	Diagnostic aura diary
Butt et al. (2022)³⁶	Double-Blind RCT	16	14 (88)	Sildenafil	100 mg	Oral	NA	3 (19)	Diagnostic aura diary
Hougaard et al. (2020)³⁷	Double-Blind RCT	14	11 (79)	ET-1	8 ng/kg/min	IV	20 min	0 (0)	ICHD-3 beta (criteria B and C
Lassen et al. (1996)³⁸	Double-Blind Trial	3	NA	Histamine	2, 4, 8, 16, 32, 64 mg/ml	Inhalation	2 min	0 (0)	NA

GTN: glyceryl trinitrate, ET-1: endothelin-1, NA: not applicable, IV: intravenous, n: number, %: percentage, min: minutes, RCT: randomized placebo-controlled trial, ICHD: international classification of headache disorders, μg: microgram, mg: milligram, kg: kilogram, ml: milliliter.

*Defined criteria applied to induced aura episodes.

Table 2.

Characteristics and outcomes of pharmacological provocation studies in migraine with aura using calcitonin gene-related peptide.

Author (Year)	Study design	Participants, n	Female, n (%)	Trigger	Exposure	Route	Duration	Provoked aura, n (%)	Applied criteria*
Hansen et al. (2010)²⁹	Non-Randomized Trial	14	8 (57)	CGRP	1.5 µg/min	IV	20 min	4 (29)	NA
Al-Khazali et al. (2023)³⁰	Non-Randomized Trial	34	30 (88)	CGRP	1.5 µg/min	IV	20 min	13 (38)	ICHD-3 (criteria B and C)
Zupan et al. (2021)³¹	Non-Randomized Trial	5	5 (100)	CGRP	1.5 µg/min	IV	20 min	0 (0)	NA

CGRP: calcitonin gene related peptide, IV: intravenous, ICHD: international classification of headache disorders, NA: not applicable, n: number, %: percentage, min: minutes, μg: microgram.

*Defined criteria applied to induced aura episodes.

Table 3.

Characteristics and outcomes of pharmacological provocation studies in migraine with aura using levcromakalim.

Author (Year)	Study design	Participants, n	Female, n (%)	Trigger	Exposure	Route	Duration	Provoked aura, n (%)	Applied criteria*
Al-Karagholi et al. (2021)³²	Double-Blind RCT	17	13 (76)	Levcromakalim	0.05 mg/min	IV	20 min	10 (59)	ICHD-3 (criteria B and C)
Thomsen et al. (2024)³³	Double-Blind RCT	27	17 (63)	Levcromakalim	0.05 mg/min	IV	20 min	4 (15)	ICHD-3 (criteria B and C)
Thomsen et al. (2024)³⁴	Double-Blind RCT	8	6 (75)	Levcromakalim	0.05 mg/min	IV	20 min	0 (0)	ICHD-3 (criteria B and C)

IV: intravenous, NA: not applicable, n: number, %: percentage, min: minutes, RCT: randomized placebo-controlled trial, ICHD: international classification of headache disorders, mg: milligram.

*Defined criteria applied to induced aura episodes.

GTN

Four studies investigated GTN as a pharmacological provocation agent,^18,25,27,28 three of which included healthy controls (HCs).^18,25,27 Intravenous infusion administered over 20 min was used in three studies,^18,27,28 whereas one used sublingual delivery.²⁵ Across these investigations, aura induction was uncommon, whereas migraine headache was consistently provoked (Table 1).

In the first non-randomized, open-label study, none of the 12 participants with migraine with aura developed aura following intravenous infusion, although migraine headache occurred in six (50%).¹⁸ None of the 14 HCs experienced either aura or headache. A subsequent non-randomized, open-label used the same infusion protocol in 21 participants with migraine with aura and 11 HCs to examine reproducibility across two sessions.²⁷ Migraine headache developed in 14 (67%) participants with migraine with aura during the first session. A single participant experienced a visual aura after the first infusion and again during the second, with similar onset times of 150 and 160 min. One additional participant developed visual aura during the second session, while no aura or headache occurred among HCs. Among 23 participants with migraine without aura, 19 (83%) developed a migraine headache following the infusion.

A third non-randomized, single arm, open-label investigation by the same research group evaluated reproducibility in 32 participants with migraine with aura and 21 with migraine without aura.²⁸ During the first session, seven (22%) participants with migraine with aura experienced aura. In the second session, which included 15 participants, aura recurred in four participants (27%). The report did not specify whether aura occurred exclusively among those with a confirmed history of aura, nor were migraine headache rates or aura characteristics detailed.

The final non-randomized, open-label study administered sublingual GTN to 22 participants with migraine with aura, 168 with migraine without aura, and 53 HCs.²⁵ Aura developed in three (14%) participants with migraine with aura, each describing transient visual disturbances (e.g., phosphenes, scotomas) resembling their typical spontaneous aura. Migraine headache followed in seven (32%) participants with migraine with aura, and 138 (82%) participants with migraine without aura.

CGRP

Three studies investigated human α-CGRP's potential to induce migraine aura, all using intravenous infusion at 1.5 µg per minute for 20 min.^29–31 One included a HCs,²⁹ whereas the remaining two were open label.^30,31 (Table 2)

In the first non-randomized, single-blind study, 14 participants diagnosed with migraine with aura and 11 HCs received intravenous CGRP.²⁹ Migraine aura developed in four participants (29%), including one prolonged episode lasting 10 h. The median onset of aura symptoms was 70 min (range, 10–360 min). Reported phenomena included fortification spectra, bilateral flickering lights, left-sided homonymous scotoma, and tingling paresthesia. Migraine headache followed in eight participants (57%), whereas none of the HCs experienced migraine aura.

A second single-arm, open-label study involving 34 participants with both migraine with and without aura found that 13 (38%) reported aura within 12 h after infusion.³⁰ The median onset was 30 min (range, 10–360 min). Visual aura occurred in 10 participants and somatosensory aura in 7, several experiencing both. In 11 of 13 cases, migraine aura appeared after headache onset, and all described the provoked symptoms as closely resembling their spontaneous aura. Migraine headache occurred in 24 participants (71%), with a median onset of 40 min (range, 10–240 min). The authors found that those who experienced migraine aura after CGRP infusion had a higher frequency of migraine aura in the past year, compared with those who did not.

A third single-arm, open-label study examined cerebral hemodynamic responses to CGRP in 5 participants with migraine with aura, and 15 with migraine without aura.³¹ None of participants with migraine with aura experienced aura, although all developed migraine headache. Among participants with migraine without aura, 11 (73%) developed a migraine headache following the infusion.

Levcromakalim

Three randomized, placebo-controlled crossover trials assessed levcromakalim's ability to induce migraine aura using a standardized 20-min intravenous infusion^32–34 (Table 3).

In the first study, 17 participants with migraine with aura received both levcromakalim and placebo on separate days.³² Aura occurred in 10 participants (59%) after levcromakalim, with a median onset of 44 min (range, 20–120 min). All aura attacks involved visual disturbances such as flickering lights, blurred vision, dark spots, or scotomas, while 3 participants also experienced somatosensory symptoms including paresthesia or numbness. Migraine headache developed in 14 participants (82%) after levcromakalim, with a median onset of 168 min (range, 60–240 min). No aura or headache occurred following placebo.

A second trial enrolled 27 participants who had a diagnosis of both migraine with and without aura.³³ Following intravenous infusion of levcromakalim, four (15%) participants reported experiencing migraine aura, whereas two (7%) did so after placebo (a non-significant difference). The median onset of aura symptoms was 27.5 min (range, 30–323 min). Reported aura symptoms included visual phenomena (n = 3), somatosensory disturbances (n = 2), and subjective motor symptoms (n = 1). Migraine headache developed in 14 (52%) following levcromakalim compared with three (11%) after placebo, with a median latency of 180 min (range, 30–360 min).

A third examined eight individuals who experienced aura without headache.³⁴ None developed aura after levcromakalim, whereas two (25%) reported aura after placebo. Migraine-like headache occurred in five (63%) following levcromakalim.

Cilostazol

A randomized, placebo-controlled crossover study evaluated the effects of oral cilostazol in 16 participants diagnosed with migraine with and without aura.³⁵ Migraine aura occurred in one (6%) participant after cilostazol and in another (6%) after placebo. The cilostazol-induced aura manifested as a left-sided homonymous scotoma beginning 4.5 h post-ingestion, whereas the placebo-related aura appeared 15 min after administration. Migraine headache developed in 12 (75%) participants after cilostazol and in one (6%) after placebo, with a median onset of 330 min (range, 120–600 min).

Sildenafil

A randomized, placebo-controlled crossover study evaluated the ability of oral sildenafil to provoke migraine aura in 16 participants diagnosed with both migraine with and without aura.³⁶ Migraine aura developed in three (19%) participants after sildenafil administration but in none after placebo. The median time to aura onset was 150 min (range, 60–180 min). Two participants reported unilateral flickering light spots, and one described bilateral scintillating phenomena, all resembling their typical spontaneous aura attacks. Migraine headache occurred in nine participants (56%) after sildenafil, with a median onset of 300 min (range, 120–360 min), whereas no headaches were reported after placebo.

Endothelin-1

A randomized, placebo-controlled crossover study evaluated the ability of endothelin-1 to induce migraine aura through a 20-min intravenous infusion administered to 14 participants diagnosed with migraine with and without aura.³⁷ None of the participants developed aura on either experimental day, and migraine headache occurred in only one individual (7%) following endothelin-1 infusion, with onset approximately 120 min post-administration.

Histamine

A single non-randomized, double-blind study investigated the effects of histamine inhalation in three participants diagnosed with migraine with aura, 12 with migraine without aura, and 15 HCs.³⁸ All participants received escalating doses of inhaled histamine under controlled lab. None of the participants with migraine with aura or the HCs developed aura following exposure. Headache outcomes were not reported for the migraine group.

Discussion

This systematic review synthesized all available human pharmacological provocation studies evaluating whether defined molecular pathways can induce migraine aura under controlled laboratory conditions. Fourteen studies met inclusion criteria and collectively examined seven endogenous or vasoactive compounds. Across these investigations, aura induction occurred far less often than migraine headache induction, underscoring that the two phenomena, although clinically linked, are not mechanistically identical. Importantly, negative provocation findings should not be interpreted as evidence against biological involvement of a given pathway in aura pathogenesis. They might instead reflect limitations in model sensitivity, including dosing paradigms, timing, observation windows, outcome definitions, and study design.

Of all tested triggers, CGRP and the vascular K_ATP channel opener levcromakalim were associated with the induction of aura. However, aura induction rates remained modest, and these findings should be interpreted as evidence of association rather than definitive mechanistic proof. About one-third of participants receiving CGRP and one quarter receiving levcromakalim developed aura, whereas no aura occurred in HCs or in participants with migraine without aura. Across studies, provoked aura symptoms closely resembled spontaneous aura, lending support to the validity of the provocation paradigm. Together, these findings identify CGRP and vascular K_ATP activation as the most promising candidate pathways for experimental induction of migraine aura.

Existing evidence and added value

The findings of this review build on several decades of work defining the biological correlates of aura and its relationship to migraine headache. Early experimental studies demonstrated that CSD produces a wave of oligemia and altered neurovascular coupling that mirrors the clinical evolution of aura.³⁹ Subsequent research showed that CSD releases potassium, hydrogen ions, and vasoactive peptides capable of activating perivascular trigeminal afferents.^6,39,40 More recent work suggests that CSD might also influence trigeminovascular activation through glymphatic-mediated cerebrospinal fluid signaling from the cortex to the trigeminal ganglion.⁴¹ These observations established a physiological bridge between cortical events and trigeminovascular activation, yet they did not clarify which certain signaling pathways promote CSD initiation in humans.

Human provocation models help address this gap by testing whether well-defined molecular triggers can elicit migraine aura under controlled experimental conditions. The present synthesis advances the field by assessing such triggers across studies, demonstrating that intravenous CGRP administration and vascular K_ATP channel activation show the highest reproducibility in eliciting aura. These findings across independent investigations suggests that certain signaling pathways exert direct or indirect modulatory effects on cortical excitability, even though their primary sites of action are located outside the central nervous system (CNS).⁴² Notably, the potential of other neuropeptides such as pituitary adenylate cyclase-activating polypeptide (PACAP) to elicit aura remains unexplored, despite their established ability to trigger migraine headache, representing an important gap for future research.⁴³

This review further underscores the importance of individual susceptibility, particularly baseline aura frequency, as a potential key determinant of responsiveness to experimental provocation.³⁰ In addition, the experimental evidence highlights important limitations in using headache induction alone as a surrogate marker for CSD susceptibility, as molecular pathways that reliably provoke headache do not necessarily trigger aura. This dissociation might, at least in part, reflect dose–response effects, whereby higher stimulus intensities or doses are required to exceed the threshold for aura generation, whereas lower levels of activation can be sufficient to elicit migraine headache without aura. Taken together, these insights emphasize the need for more targeted and mechanistically informed provocation paradigms capable of testing hypotheses about migraine aura with greater precision.

Hypothesized mechanisms

The mechanistic implications of these findings can be interpreted within two hypotheses describing the relationship between aura and headache: the sequential or build-up model and the bidirectional neurovascular model. Evidence from this review, together with findings from preventive migraine treatment, places these models in clearer perspective.

The sequential model proposes that CSD is the primary initiating event in migraine with aura.^5,7 In this model, CSD disrupts cortical homeostasis and releases potassium, glutamate, hydrogen ions, and pro-inflammatory mediators into the cerebrospinal fluid.⁵ These mediators then activate meningeal nociceptors and initiate trigeminovascular signaling, producing headache once a critical threshold is reached. The build-up hypothesis refines this model by linking headache timing to the cumulative concentration of these mediators, thereby explaining variability in the latency between aura and headache.⁴⁴

Extensive preclinical work supports this directionality. CSD consistently activates meningeal afferents, increases cellular FOS (c-Fos) expression in trigeminovascular nuclei, and produces headache-like behaviors in rodents.⁴⁵ These findings make a strong case that CSD leads totrigeminovascular activation is a robust biological pathway.

However, the sequential model alone cannot explain key clinical observations. Foremost among them is the consistent preventive effect of peripherally restricted anti-CGRP monoclonal antibodies on both migraine headache and seemingly also migraine aura.^46–48 These antibodies do not cross the blood–brain barrier in meaningful quantities, yet some interventional studies report reductions in aura frequency. A purely sequential model would require that aura prevention occur through direct cortical effects of treatment, which is incompatible with the pharmacology of these antibodies. This discrepancy suggests that a peripheral pathway capable of modulating CSD susceptibility must exist, and that this pathway is not captured by the traditional sequential view.

The bidirectional model addresses this limitation by proposing that peripheral trigeminovascular activation can influence cortical excitability and promote CSD initiation.^30,49 This 2023 hypothesis is supported by human provocation studies summarized in this review. In this model, activation of meningeal nociceptors produces ascending signals through the trigeminal pain pathways. These inputs modulate thalamocortical circuits projecting to cortical regions, including the visual cortex, where CSD is most commonly initiated.⁵⁰ In people with migraine with aura, this modulation might be sufficient to lower the CSD threshold.

The bidirectional model explains several observations that remain puzzling under the sequential model alone. CGRP, which does not cross the blood–brain barrier, induces aura in a substantial proportion of participants with migraine with aura. The absence of aura induction in participants with migraine without aura and healthy volunteers further supports a threshold-based mechanism in which peripheral input must converge on a cortex already predisposed to CSD.⁴⁴

These findings indicate that peripheral activation is sufficient to precipitate CSD when cortical susceptibility is high, supporting a trigeminovascular → cortical direction of causality. The model also accommodates the preventive effects of anti-CGRP monoclonal antibodies, because blocking peripheral CGRP signaling reduces the likelihood of trigeminovascular activation and therefore reduces the chance of triggering CSD via thalamocortical modulation.^51,52

Importantly, the bidirectional model does not negate the well-established ability of CSD to activate trigeminovascular pathways. Instead, it integrates both directions into a unified framework in which cortical and peripheral processes can each initiate attacks and amplify one another. CSD can activate meningeal nociceptors, and peripheral nociceptive activation can, under some conditions, facilitate CSD. This bidirectionality might explain the complex temporal relationships between aura and headache seen clinically, including aura during headache, aura without headache, and headache without aura.⁵³

Taken together, the available evidence positions the bidirectional neurovascular model as the most coherent framework for both provoked and spontaneous aura. It incorporates extensive preclinical data showing that CSD activates meningeal nociceptors. It also explains CGRP-induced aura through meningeal effects that can modulate thalamocortical excitability. This model further resolves the paradox created by the preventive effects of peripherally acting anti-CGRP monoclonal antibodies, which reduce aura despite minimal CNS penetration. Lastly, it offers a biologically plausible account of the inter-individual variability in aura susceptibility observed across provocation studies.

Strengths and limitations

This systematic review has several strengths. It provides the most comprehensive synthesis of human pharmacological provocation studies evaluating migraine aura under controlled lab conditions. By integrating findings across seven mechanistically distinct molecular triggers, the review offers a cohesive overview of how different molecular mechanisms might influence cortical susceptibility to aura.

Several limitations warrant consideration. First, the included studies were small, somewhat heterogeneous, and often open-label, increasing susceptibility to expectancy effects. Second, aura characterization was sometimes incomplete, with inconsistent reporting of symptom evolution, anatomical distribution, and duration. Third, provocation models might not fully recapitulate spontaneous attack dynamics, and mechanistic interpretations must therefore remain cautious. Finally, aura induction rates could depend on dose, timing, observation window, outcome definitions, and statistical power. Hence, negative provocation findings should not be interpreted as evidence against biological involvement of a pathway, particularly for agents studied in small or methodologically limited trials.

Conclusions

Human provocation studies show that migraine aura can be experimentally induced, with somewhat consistent responses arising from activation of CGRP signaling and vascular K_ATP channels. These potential aura triggers identify plausible mechanistic routes through which peripheral (meningeal) events might facilitate CSD. Larger studies are now needed to clarify these mechanisms, strengthen causal inference, and guide future research directions.

Article highlights

Human pharmacological provocation studies show that migraine aura can be experimentally induced, but far less frequently than migraine headache, highlighting distinct underlying mechanisms.

Among tested agents, calcitonin gene-related peptide (CGRP) and pharmacological opening of vascular ATP-sensitive potassium (KATP) channels were seemingly able to induce migraine aura, implicating these pathways as the most plausible triggers.

The findings support a bidirectional neurovascular model, in which meningeal trigeminovascular signaling might lower cortical spreading depolarization thresholds in people with migraine with aura, refining current concepts of migraine aura pathogenesis.

Supplemental Material

sj-docx-1-cep-10.1177_03331024261429112 - Supplemental material for Molecular triggers of migraine aura: A systematic review of human pharmacological provocation studies

Supplemental material, sj-docx-1-cep-10.1177_03331024261429112 for Molecular triggers of migraine aura: A systematic review of human pharmacological provocation studies by Zahra Hakimzadeh, Basit Ali Chaudhry, Haidar M. Al-Khazali, Rune Häckert Christensen, Messoud Ashina and Håkan Ashina in Cephalalgia

Footnotes

Abbreviations

Acknowledgements

Not applicable.

Author contributions

ZH: Data curation; Investigation; Methodology; Validation; Writing – original draft; Writing – review & editing.

BAC: Data curation; Investigation; Validation; Writing – review & editing.

HMA: Methodology; Supervision; Writing – review & editing.

RHC: Methodology; Writing – review & editing.

MA: Conceptualization; Funding acquisition; Methodology; Supervision; Writing – review & editing.

HA: Conceptualization; Funding acquisition; Methodology; Project administration; Supervision; Writing – review & editing.

All authors meet ICMJE authorship criteria and approved the final manuscript.

Consent to participate

Not applicable.

Consent for publishing

All authors have read and approved the final version of this manuscript and consent to its submission and publication. The corresponding author accepts responsibility for ensuring that all authors provide informed consent for publication.

Data availability statement

Upon reasonable request, the corresponding author will provide the necessary data and materials to interested researchers for the purpose of academic scrutiny, reproducibility, and further scientific investigation.

Declaration of conflicting interests

BAC has received personal fees from Pfizer, outside the submitted work. HMA has received personal speaker fees from AbbVie, Pfizer, and Lundbeck, outside of the submitted work. RHC has received personal speaker fees from AbbVie, Pfizer, Lundbeck, and Teva, outside of the submitted work. MA has also received institutional research grants from the Danish National Research Foundation, Novartis, Novo Nordisk Foundation, and Lundbeck Foundation. MA is also an Associate Editor of Brain and The Journal of Headache and Pain. HA has received personal fees from AbbVie, Lundbeck, Pfizer and Teva, outside of the submitted work. HA is also an Editorial Board Member of The Journal of Headache and Pain. The remaining authors declare no conflicts of interest.

Ethical considerations

Not applicable for reviews.

Funding

Funding was provided by the Danish National Research Foundation (DNRF188 to MA) and Lundbeck Foundation (R310-2018-3711 to MA; R403-2022-1352 and R481-2024-1392 to HA). The funding sources were not involved in the design or conduct of the study; in the collection, analysis, or interpretation of the data; or in the preparation, review, or approval of the manuscript.

Open practices

Not applicable.

ORCID iDs

Zahra Hakimzadeh

Basit Ali Chaudhry

Haidar M. Al-Khazali

Rune Häckert Christensen

Messoud Ashina

Håkan Ashina

Supplemental material

Supplemental material for this article is available online.

References

Eftekhari

Wang

, et al. Cortical spreading depression as a site of origin for migraine: role of CGRP. Cephalalgia 2019; 39: 428–434.

Leao

. Further observations on the spreading depression of activity in the cerebral cortex. J Neurophysiol 1947; 10: 409–414.

Vitale

Tottene

Zarin Zadeh

, et al. Mechanisms of initiation of cortical spreading depression. J Headache Pain 2023; 24: 105.

Kramer

Fujii

Ohiorhenuan

, et al. Cortical spreading depolarization: pathophysiology, implications, and future directions. J Clin Neurosci 2016; 24: 22–27.

Zhang

Levy

Kainz

, et al. Activation of central trigeminovascular neurons by cortical spreading depression. Ann Neurol 2011; 69: 855–865.

Karatas

Erdener

Gursoy-Ozdemir

, et al. Spreading depression triggers headache by activating neuronal Panx1 channels. Science 2013; 339: 1092–1095.

Bolay

Reuter

Dunn

, et al. Intrinsic brain activity triggers trigeminal meningeal afferents in a migraine model. Nat Med 2002; 8: 136–142.

Zhang

Levy

Noseda

, et al. Activation of meningeal nociceptors by cortical spreading depression: implications for migraine with aura. J Neurosci 2010; 30: 8807–8814.

Charles

Hansen

. Migraine aura: new ideas about cause, classification, and clinical significance. Curr Opin Neurol 2015; 28: 255–260.

10.

Grodzka

Dzagoevi

Rees

, et al. Migraine with and without aura-two distinct entities? A narrative review. J Headache Pain 2025; 26: 77.

11.

Spierings

. Parallel concept of migraine pathogenesis. Ann Neurol 2002; 51: 139–140.

12.

Moskowitz

. Defining a pathway to discovery from bench to bedside: the trigeminovascular system and sensitization. Headache 2008; 48: 688–690.

13.

Hadjikhani

Vincent

. Neuroimaging clues of migraine aura. J Headache Pain 2019; 20: 32.

14.

Lindblad

Hougaard

Amin

, et al. Can migraine aura be provoked experimentally? A systematic review of potential methods for the provocation of migraine aura. Cephalalgia 2017; 37: 74–88.

15.

Ashina

Hansen

. Pharmacological migraine provocation: a human model of migraine. Handb Clin Neurol 2010; 97: 773–779.

16.

Olesen

. Provocation of attacks to discover migraine signaling mechanisms and new drug targets: early history and future perspectives - a narrative review. J Headache Pain 2024; 25: 105.

17.

Ashina

Hansen

Á Dunga

, et al. Human models of migraine - short-term pain for long-term gain. Nat Rev Neurol 2017; 13: 713–724.

18.

Christiansen

Thomsen

Daugaard

, et al. Glyceryl trinitrate induces attacks of migraine without aura in sufferers of migraine with aura. Cephalalgia 1999; 19: 660–626.

19.

Frank

Faulhaber

Messlinger

, et al. Migraine and aura triggered by normobaric hypoxia. Cephalalgia 2020; 40: 1561–1573.

20.

Olesen

Friberg

Olsen

, et al. Timing and topography of cerebral blood flow, aura, and headache during migraine attacks. Ann Neurol 1990; 28: 791–798.

21.

Ashina

Hansen

Olesen

. Pearls and pitfalls in human pharmacological models of migraine: 30 years’ experience. Cephalalgia 2013; 33: 540–553.

22.

Campbell

McKenzie

Sowden

, et al. Synthesis without meta-analysis (SWiM) in systematic reviews: reporting guideline. Br Med J 2020; 368: l6890.

23.

Ghanizada

Iljazi

Ashina

, et al. Nocebo response in human models of migraine: a systematic review and meta-analysis of randomized, double-blind, placebo-controlled, two-way crossover trials in migraine without aura and healthy volunteers. Cephalalgia 2021; 41: 99–111.

24.

Schytz

Birk

Wienecke

, et al. PACAP38 Induces migraine-like attacks in patients with migraine without aura. Brain 2009; 132: 16–25.

25.

Sances

Tassorelli

Pucci

, et al. Reliability of the nitroglycerin provocative test in the diagnosis of neurovascular headaches. Cephalalgia 2004; 24: 110–119.

26.

Chandler

Cumpston

, et al. Cochrane handbook for systematic reviews of interventions. Hoboken: Wiley 2019; 4: 14651858.

27.

Afridi

Kaube

Goadsby

. Glyceryl trinitrate triggers premonitory symptoms in migraineurs. Pain 2004; 110: 675–680.

28.

Karsan

Bose

Thompson

, et al. Headache and non-headache symptoms provoked by nitroglycerin in migraineurs: a human pharmacological triggering study. Cephalalgia 2020; 40: 828–841.

29.

Hansen

Hauge

Olesen

, et al. Calcitonin gene-related peptide triggers migraine-like attacks in patients with migraine with aura. Cephalalgia 2010; 30: 1179–1186.

30.

Al-Khazali

Ashina

Wiggers

, et al. Calcitonin gene-related peptide causes migraine aura. J Headache Pain 2023; 24: 124.

31.

Zupan

Zaletel

Visočnik

, et al. Calcitonin gene-related peptide-induced hemodynamic changes in migraine with and without aura. Acta Neurol Scand 2021; 144: 616–622.

32.

Al-Karagholi

Ghanizada

Nielsen

CAW

, et al. Opening of ATP sensitive potassium channels causes migraine attacks with aura. Brain 2021; 144: 2322–2332.

33.

Thomsen

Al-Karagholi

Hougaard

, et al. Investigations of the migraine-provoking effect of levcromakalim in patients with migraine with aura. Cephalalgia 2024; 44: 3331024241237247.

34.

Thomsen

Al-Karagholi

Hougaard

, et al. Effects of levcromakalim in patients with migraine aura without headache: an experimental study. Cephalalgia 2024; 44: 3331024241274366.

35.

Butt

Rostrup

Hansen

, et al. Induction of migraine-like headache, but not aura, by cilostazol in patients with migraine with aura. Brain 2018; 141: 2943–2951.

36.

Butt

S Eddelien

Kruuse

. The headache and aura-inducing effects of sildenafil in patients with migraine with aura. Cephalalgia 2022; 42: 984–992.

37.

Hougaard

Younis

Iljazi

, et al. Intravenous endothelin-1 infusion does not induce aura or headache in migraine patients with aura. Headache 2020; 60: 724–734.

38.

Lassen

Heinig

Oestergaard

, et al. Histamine inhalation is a specific but insensitive laboratory test for migraine. Cephalalgia 1996; 16: 550–553.

39.

Lauritzen

. Pathophysiology of the migraine aura. The spreading depression theory. Brain 1994; 117: 199–210.

40.

Lambert

Michalicek

Storer

, et al. Effect of cortical spreading depression on activity of trigeminovascular sensory neurons. Cephalalgia 1999; 19: 631–638.

41.

Kaag Rasmussen

Møllgård

Bork

PAR

, et al. Trigeminal ganglion neurons are directly activated by influx of CSF solutes in a migraine model. Science 2024; 385: 80–86.

42.

Ashina

Schytz

Ashina

. CGRP In human models of migraine. Handb Exp Pharmacol 2019; 255: 109–120.

43.

Ashina

Christensen

Hay

, et al. Pituitary adenylate cyclase-activating polypeptide signalling as a therapeutic target in migraine. Nat Rev Neurol 2024; 20: 660–670.

44.

Moskowitz

. Rethinking migraine with aura: why cortical spreading depolarization (depression), not aura, causes headaches. Cephalalgia 2025; 45: 3331024251370629.

45.

Moskowitz

Nozaki

Kraig

. Neocortical spreading depression provokes the expression of c-fos protein-like immunoreactivity within trigeminal nucleus caudalis via trigeminovascular mechanisms. J Neurosci 1993; 13: 1167–1177.

46.

Braca

Miele

Stornaiuolo

, et al. Are anti-calcitonin gene-related peptide monoclonal antibodies effective in treating migraine aura? A pilot prospective observational cohort study. Neurol Sci 2024; 45: 1655–1660.

47.

Romozzi

Burgalassi

Vollono

, et al. Prospective evaluation of aura during anti-calcitonin gene-related peptide monoclonal antibody therapy after 52 weeks of treatment. Confinia Cephalalgica 2024; 34: 15762.

48.

Ashina

McAllister

Cady

, et al. Efficacy and safety of eptinezumab in patients with migraine and self-reported aura: post hoc analysis of PROMISE-1 and PROMISE-2. Cephalalgia 2022; 42: 696–704.

49.

Al-Karagholi

. Targeting CGRP pathways and aura: a peripheral site with a central effect. Cephalalgia 2023; 43: 3331024231212895.

50.

Christensen

Ashina

. The vessel-to-neuron trigeminovascular hypothesis of migraine pathogenesis - the ‘pro’ argument. J Headache Pain 2025; 26: 248. Published 2025 Nov 10. doi:10.1186/s10194-025-02130-z.

51.

Noseda

Schain

Melo-Carrillo

, et al. Fluorescently-labeled fremanezumab is distributed to sensory and autonomic ganglia and the dura but not to the brain of rats with uncompromised blood brain barrier. Cephalalgia 2020; 40: 229–240.

52.

Melo-Carrillo

Strassman

Nir

, et al. Fremanezumab-A humanized monoclonal anti-CGRP antibody-inhibits thinly myelinated (aδ) but not unmyelinated (C). Meningeal Nociceptors. J Neurosci 2017; 37: 10587–10596.

53.

Viana

Linde

Sances

, et al. Migraine aura symptoms: duration, succession and temporal relationship to headache. Cephalalgia 2016; 36: 413–421.

Supplementary Material

Please find the following supplemental material available below.

For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.

For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.

0.05 MB

0.00 MB