Targeting CGRP pathways and aura: A peripheral site with a central effect

Abstract

Targeting CGRP-pathways has substantially expanded our options for treating individuals with migraine. Although the efficacy of these drugs on migraine aura is yet to be fully revealed, it seems from existing studies that CGRP antagonism reduces the number of migraine auras. The present perspective summarizes the evidence linking CGRP to the migraine aura and proposes a model by which targeting the CGRP-pathways and, thus, inhibition the interaction between C- and Aδ-trigeminal fibers might reverse a possible high cortical glutamate level leading to a reduced number of migraine auras.

Introduction

During the past three decades, studies of the molecular underpinnings of migraine identified various novel drug targets (1,2). A major therapeutic breakthrough emerged from targeting CGRP or its receptor complex. Three monoclonal antibodies neutralizing the CGRP molecule (anti-CGRP mAbs: eptinezumab, fremanezumab, and galcanezumab) and one targeting the CGRP receptor (anti-CGRP-R mAb: erenumab) have proven effective for the prevention of episodic and chronic migraine. Additionally, four small-molecule drugs (gepants) targeting the CGRP receptor are beneficial for acute migraine treatment (rimegepant, ubrogepant and zavegepant) or for the prevention of episodic migraine (rimegepant and atogepant). These drugs substantially expanded our options for treating individuals with migraine (3). No placebo-controlled trial has assessed the efficacy of targeting CGRP-pathways in individuals with migraine with aura exclusively or has analyzed the effect on aura attacks separate from attacks without aura, and no trial directly compared drug effects in individuals with aura to individuals without aura. It seems from existing studies, however, that CGRP antagonism reduces the number of migraine auras (4–6). How and by what mechanism CGRP antagonism prevents migraine aura is yet to be elucidated. The present perspective summarizes the evidence linking CGRP to the migraine aura and proposes a mechanism by which targeting the CGRP-pathways might prevent migraine aura.

Cortical spreading depolarization

Cortical spreading depolarization (CSD), a wave of positively charged currents in neurons and glia cells slowly moving over the cortex, is widely accepted to be the cause of migraine aura (7). High extracellular glutamate and potassium (K⁺) levels are implicated in CSD initiation and propagation. Glutamate is the principal excitatory neurotransmitter in the central nervous system and glutamate receptor N-methyl-D-aspartate receptor (NMDAR), which conducts sodium (Na⁺) and calcium (Ca²⁺) ions into the cell and K⁺ ions outside of the cell, is thought to be involved in CSD propagation. Glutamate involvement in CSD is further supported by the findings that NMDAR antagonism inhibited in vitro generated CSD (8) and led to a significant reduction in the frequency of aura and headache in patients with migraine with aura (9). Although, the nature of the relation between the migraine aura and the headache phase of migraine is not fully understood, evidence from animal models of migraine convincingly showed that CSD is a trigger of trigeminovascular activation and headache (10).

Calcitonin gene-related peptide (CGRP)

CGRP is a vasoactive neuropeptide widely expressed in the central and peripheral nervous systems (11). Basic and clinical evidence (12–14) supporting the role of CGRP in migraine pathophysiology led to the development of CGRP receptor antagonists (gepants) and antibodies against CGRP or its receptor. Targeting the CGRP-pathways has proven efficacious for migraine treatment but despite this consensus, there is still much to learn about the location of pertinent CGRP receptor sites implicated in migraine.

Upon systemic administrations of gepants (orally or intranasally) and anti-CGRP(-R) mAbs (intravenously or subcutaneously) only a very small percentage of these drugs penetrates the blood-brain barrier (BBB) and the PNS is therefore the major site of action (15,16). Human studies showed that BBB remains intact during migraine attacks (17–19). Ziegeler and colleagues (20) reported that administration of erenumab modulates the activation of a specific brain network after trigeminal nociceptive input. The authors provided several explanations for this observation including a peripheral site of action that indirectly modulates central transmission of trigeminal nociceptive input (20). This modulation in the brain including in areas like the hypothalamus could also explain the effect of these drugs on migraine-associated symptoms (21,22). According to the Biopharmaceutical Drug Disposition and Classification System (BDDCS), gepants are classified as class 2 drugs owing to their low solubility and high permeability. This means that gepants, particularly zavegepant, could lead to effects on the central nervous system (CNS) despite being a good substrate for P-glycoprotein (23). However, a recent phase III trial showed that individuals using zavegepant to treat a single migraine attack did not report any CNS-related adverse events (24). Thus, CGRP antagonism might exert their therapeutic effect in the trigeminal ganglion and in trigeminal nerve fibers innervating dura mater and vascular and cutaneous tissues (25,26). This modulation is thought to inhibit CGRP-induced hyperalgesia and allodynia by reversing a possible hypersensitivity state in migraine patients (27,28).

Interplay between CSD and CGRP

Understanding the reciprocal connections between CSD and CGRP might elucidate how targeting CGRP-pathways prevents migraine aura. CSD increases CGRP synthesis and release in the cerebral cortex. Repeated CSD events upregulated the CGRP gene and increased peptide levels in various brain regions (29), and in vitro generated CSD increased CGRP release from cortical tissue (8). Whether CSD increases CGRP synthesis in the trigeminal ganglia remains to be determined. Yisarakun and colleagues (30) demonstrated that CSD increased the number of CGRP-positive cells in rat trigeminal ganglia, which might indicate an increased synthesis. Although the exact cause underlying CSD-induced cortical CGRP gene expression is yet to be revealed, several mechanisms have been proposed including oxidative stress and generation of reactive oxygen species (29). Overall, these data provide evidence linking CSD and CGRP expression that might contribute to migraine pathogenesis. Close and colleagues (31) tried to explain how CSD affects CGRP by proposing a bidirectional communication model between CSD and CGRP. The postulated model consists of three essential elements. First, CSD leads to CGRP release in the cerebral cortex and from trigeminal afferents in the meninges. Second, CGRP dilates meningeal vessels and arterioles in the pia and brain parenchyma. Third, vasodilation and increased flow in parenchymal vessels maintain increased neural activity and CGRP release and create a self-sustaining positive feedback loop by vascular-neural coupling. Of note, this model did not include how CGRP contributes to CSD.

To date, there is no evidence that CGRP can trigger CSD events. Previous pharmacological provocation studies reported few cases of aura (four out of 14 participants with migraine with aura, 28%) after CGRP infusion (32). Studies with larger sample sizes consistently reported migraine attacks without aura but no aura attacks (33). In a recent open-label, non-randomized, single-arm study, 13 (38%) of 34 participants with migraine with aura (the median number of aura attacks was 25 in the past year) developed migraine aura after CGRP infusion (34). The fundamental question is whether CGRP antagonism can modulate CSD. Multiple in vitro cortical slice studies reported that blockage of CGRP receptors resulted in inhibition of CSD (8). In mice, systemic administration of olcegepant, a CGRP receptor antagonist, resulted in inhibition of repetitive CSD events and altered the vascular response to CSD (35). Moreover, in rats, systemic administration of fremanezumab prevented activation of central trigeminovascular neurons induced by CSD, braked the propagation velocity of CSD, and reduced the duration of cortical recovery that followed the spreading depolarization (36). Together, these data indicate that CGRP or activation of its receptor(s) may contribute to CSD initiation and propagation. This notion is further supported by some reports indicating that CGRP antagonism prevents migraine aura (4–6). However, systemic application of MK-8825, a highly selective and potent antagonist at rat CGRP receptors, failed to block CSD and did not alter changes in cerebral blood flow, but did attenuate pain behaviors (37).

Proposed model: Peripherally acting anti-CGRP drugs affect central glutamate

Binding of CGRP to their receptors leads to activation of protein kinase A (PKA) and thus upregulation of potassium and sodium channels essential for pain transmission and neural sensitization (38). Peripheral trigeminal fibers within the trigeminal pain pathway consist of fibers storing CGRP (C-fibers) and fibers expressing CGRP receptors (Aδ-fibers) (39,40). The novel drugs exert their antimigraine effect by preventing binding of CGRP ligand to Aδ-fibers (10). Targeting CGRP-pathways prevents, if efficacious, not just the migraine pain but also aura and other central threshold symptoms such as photophobia, phonophobia, and prodromes (41). The following model highlights how peripherally acting drugs can reduce the frequency of aura attacks that originate in the brain.

Binding of CGRP to their receptor at the peripheral terminal of the trigeminal Aδ-fiber leads to an activation, and action potentials from here reach the central terminal of the trigeminal Aδ-fiber and cause a release of glutamate leading to activation of neurons within the trigeminal cervical complex (TNC). Nociceptive signals from the trigeminal pain pathway reach and activate multiple cortical areas (42,43), partly through glutamatergic neurotransmission leading to elevated levels of extracellular glutamate (44). Thus, glutamate plays a key role in the sensitization of the trigeminal pain pathway and targeting CGRP-pathways might attenuate central glutamate release from Aδ-fiber. Since nociceptive signals from the trigeminal pain pathway reach multiple cortical areas, inhibition of the trigeminal pain pathway might decrease extracellular glutamate levels in brain cortex. Lower extracellular glutamate levels upon targeting the CGRP-pathways might reduce susceptibility to CSD (Figure 1). Collectively, modulation of the interaction between C- and Aδ-trigeminal fibers might reverse a possible high cortical glutamate level. This theoretical model emphasizes several outstanding questions: whether targeting CGRP-pathways affects glutamate level in the visual cortex?; whether a long-term treatment has a sustained effect?; and whether other neurotransmitters are involved in this modulation?

Figure 1.

Proposed glutamate model. Nociceptive fibers (C-fibers) originating from the trigeminal ganglion release calcitonin gene-related peptide (CGRP) and innervate the meninges and large cerebral arteries. The release of CGRP results in meningeal vasodilation and stimulation of Aδ-trigeminal fibers. Targeting CGRP-pathways (erenumab is shown, part 1) inhibits vasodilation and might attenuate central glutamate release from Aδ-fibers (part 2). Since nociceptive signals from the trigeminal pain pathway reach multiple cortical areas, inhibition of the trigeminal pain pathway might decrease extracellular glutamate levels in brain cortex (part 3). Lower extracellular glutamate levels might reduce susceptibility for CSD. NMDAR, N-methyl d-aspartate receptor; RCP, Receptor component protein; TG, trigeminal ganglion; TCC, trigeminal cervical complex; TNC, trigeminal nucleus caudalis.

Numerous observations implicated glutamate in migraine pathophysiology. NMDARs are expressed in cortical tissue and a sudden drop in membrane resistance via opening of these nonselective large-conductance cation channels might lead to CSD. Mouse models of familial hemiplegic migraine (FHM) showed a reduced rate of glutamate clearance, high extracellular glutamate levels and hence an increased susceptibility to CSD (45). Crivellaro and colleagues explained this observation by showing an increased and prolonged activation of NMDARs in cortical pyramidal cells (46). Evidence from animal models indicates that CSD activates meningeal nociceptors and central trigeminovascular neurons (47,48). Hadjikhani and colleagues (49) suggested that aura induces meningeal inflammation and central activation through microvessels linking brain-meninges-bone marrow. This crosstalk between brain and meninges is also a plausible explanation of how a peripheral site has a central effect.

Since NMDARs are expressed at several levels of the trigeminal pain pathway, glutamate and its NMDAR could be the missing link between migraine aura and migraine pain. Of note, a consumption of a single dose (150 mg/kg) of monosodium glutamate (MSG) caused headache, craniofacial sensitivity, and nausea in healthy participants (50,51). A repeated MSG intake (150 mg/kg) for five daily sessions for one-week reduced pressure pain thresholds and caused headache in healthy participants compared to placebo. Whether administration of glutamate induces migraine aura and migraine headache is yet to be elucidated. NMDAR antagonists such as Mg²⁺, ketamine, and memantine have been investigated for the treatment of migraine, and several studies reported that administration of NMDAR antagonists led to a significant reduction in (1) the monthly headache frequency, (2) the mean number of days with severe pain, and (3) the mean disability score (52–56). Furthermore, administration of ketamine reduced the severity and the duration of migraine aura. Other glutamate receptors might be implicated in migraine including metabotropic glutamate receptor (mGluR1 and mGluR5) (57).

Future perspectives

Large-scale randomized clinical trials are needed to assess the therapeutic gain of targeting CGRP pathway in individuals with migraine with aura. Measurement of glutamate levels in the trigeminal pain pathway upon targeting CGRP pathway will further elucidate whether the current model is tenable. Also, other central ion channels and neurotransmitters might be involved (58–60).

CSD activation of the trigeminovascular system is a debated topic. Animal studies showed that daily administration of migraine prophylactic drugs including lamotrigine, topiramate and valproate for at least one week suppressed CSD frequency (61). Also, a single dose of topiramate or gabapentin led to a suppression of CSD susceptibility (62,63). CSD suppression explains at least partially the efficacy of these drugs in migraine prophylaxis. These data and the reported activation of the trigeminovascular system upon repeated CSD induction indicate that CSD is a potential trigger of migraine attacks. Yet, clinical observations speak against a causal relation between CSD and headache. Premonitory symptoms can be presented long before any aura symptoms (64), some individuals with migraine with aura experience isolated auras without headache (65), and auras cannot predict the side of the headache (66). Moreover, experiencing several auras in succession is highly uncommon and would be a clinical warning sign (animal models applied repeated CSD induction). Of note, a recent case study showed that aura was not a trigger of headache attacks (67). Further studies are required to clarify this complex causal relation between aura and headache in migraine.