An important advance in the treatment of migraine has been the development of agents that target calcitonin gene-related peptide (CGRP), a neuropeptide that has emerged as a key player in the neurobiology of migraine.
The CGRP family comprises calcitonin, CGRP, adrenomedullin and amylin (Walker et al. Br J Pharmacol 2013;170:1293-1307). The receptors consist of one G protein-coupled receptor, either calcitonin receptor (CTR) or the calcitonin-like receptor (CLR), linked to a receptor activity-modifying protein (RAMP). Different combinations of these subunits result in functional receptors for CGRP (CLR + RAMP1), adrenomedullin (CLR + RAMP2 or RAMP3) and amylin (CTR + RAMP1, RAMP2 or RAMP3) (Hay et al. Br J Pharmacol 2018;175:3-17). There are two isoforms of CGRP: the alpha form, which is primarily found in the central and peripheral nervous systems, most notably the trigeminal ganglion and the dorsal root ganglion; and the beta form, which is mostly present in the enteric nervous system (Juaneda et al. Trends Pharmacol Sci 2000;21:432-438).
The CGRP pathway was first hypothesized to have a role in migraine in 1985 (Edvinsson L. Trends Neurosci 1985;8:126-131). CGRP has two key neurovascular effects of relevance to migraine. CGRP fibres originating in the trigeminal ganglion have a potent vasodilatory effect on cerebral arterioles (Edvinsson et al. Neurosci Lett 1985;58:213-217). This effect has been described as a protective vasodilatory reflex that occurs in response to vasoconstriction (McCulloch et al. PNAS USA 1986;83:5731-5735). A theoretical concern of CGRP blockade is impairment of cardioprotective mechanisms, but this has not been demonstrated. CGRP antagonists do not appear to cause constriction of coronary arteries in vitro (Chan et al. J Pharmacol Exp Ther 2010;334:746-752), nor is there an effect on coronary arteries during periods of ischemia in animal models (Lynch et al. Eur J Pharmacol 2009;623:96-102). During exercise testing in non-migraine subjects with stable angina, administration of the CGRP receptor blocker erenumab did not adversely affect exercise time (Cady et al. Cephalalgia 2017;37(1S):340-341; abstract PO-01-199). Thus, it has been suggested that CGRP antagonists restore normal vascular tone but do not cause abnormal vasoconstriction (Bigal et al. Headache 2013;53:1230-1244; Verheggen et al. Br J Pharmacol 2002;136:120-126).
CGRP was initially investigated when migraine was viewed as a primarily vascular disorder. More recent research has focused on the role of CGRP in the transmission and modulation of pain signals. CGRP receptors are widely distributed throughout the brain, but CGRP antagonists do not readily cross the blood-brain barrier (BBB). Thus, it is believed that the principal sites of action of CGRP antagonists are at structures lying outside of the BBB, notably the trigeminal ganglion and the dura (Edvinsson L. Headache 2017;57:47-55). CGRP is present in over 50% of trigeminal neurons, and is often expressed with substance P and 5-HT1B/D receptors (Lennerz et al. J Comp Neurol 2008;507:1277-1299). CGRP neurons project to the trigeminal nucleus caudalis and are involved in pain signalling from the brainstem to the thalamus (Eftekhari & Edvinsson. BMC Neurosci 2011;12:112). In addition, CGRP release at trigeminal nerve endings produces vasodilation, edema, and dural mast cell degranulation, which result in neurogenic inflammation and the release of proinflammatory mediators.
Thus, in the trigeminovascular system, CGRP appears to activate feed-forward circuits that promote and sustain pain signalling. This provides the rationale for disrupting pain signalling in migraine by blocking CGRP (Iyengar et al. Pain 2017;158:543-559).
A number of clinical studies provide further support for the role of CGRP in migraine. CGRP has been shown to be released during acute migraine attacks in humans (Goadsby et al. Ann Neurol 1990;28:183-187). Indeed, CGRP was the only neuropeptide in the trigeminovascular system to be correlated with acute attacks (Edvinsson et al. Cephalalgia 2010;30:761-766). Moreover, administration of CGRP has been shown to trigger migraine attacks in patients with a history of migraine (Hansen et al. Cephalalgia 2010;30:1179-1186). Interestingly, many of the current therapies used to treat migraine have CGRP antagonist effects, including sumatriptan (Juhasz et al. Cephalalgia 2005;25:179-183; Asghar et al. Neurology 2010;75:1520-1526), ketoprofen (Vellani et al. Mediators Inflamm 2017;2017:9547056), and coenzyme Q (Nutr Neurosci 2018; Jan 3:1-9). Indeed, a reduction in saliva CGRP levels, which become elevated during migraine attacks, has been proposed as a biomarker of treatment response to triptans (Cady et al. Headache 2009;49:1258-1266).
These findings have led to the development of two classes of drug that specifically target CGRP, and both have been found to be effective in migraine. The first-generation small-molecule ‘gepants’ were shown in Phase II and III testing to be as effective as triptans in acute migraine, however, their development was stopped due to problems of bioavailability (olcegepant) or hepatotoxicity (telcagepant, MK-3207). Other gepant compounds are now in development for acute migraine treatment and prophylaxis.
In addition, two types of monoclonal antibody (MAb) are in development for migraine. There are three humanized MAbs that target the CGRP molecule (eptinezumab [ALD403], fremanezumab [TEV-48215], galcanezumab [LY2951742]), and one fully-human MAb targeting the CGRP receptor (erenumab [AMG-334]). Eptinezumab is administered intravenously, whereas the other three are administered by subcutaneous injection. Galcanezumab is administered every two weeks; the others are administered every four weeks.
There are theoretical differences between targeting the CGRP peptide and targeting its receptor. CGRP antagonists act on both the alpha and beta isoforms; effects on the enteric nervous system, such as alterations in gut mucosal integrity or gastrointestinal motility, have been suggested as potential side effects (Deen et al. J Headache Pain 2017;18:96). CGRP blockade may also permit peptides other than CGRP to bind to the CGRP receptor (MaassenVanDenBrink et al. Genome Medicine 2018;10:10). Conversely, targeting the CGRP receptor may result in CGRP interactions at other receptors, in effect producing ‘functional CGRP receptors’. However, at present these are speculative and it has not been determined if there are clinically meaningful differences with these two modes of action.
All of the CGRP MAbs are intended for migraine prophylaxis due to their parenteral administration, long time to achieve maximal drug concentration and prolonged treatment effect. Since they are large molecules, there is minimal penetration of the BBB, and central side effects would not be expected. In contrast to the gepant agents, there are few hepatotoxicity concerns with MAbs since they produce no hepatoxic metabolites and they are not eliminated via the hepatic route.
All of the CGRP MAbs appear to be very well tolerated. For example, in the phase III STRIVE trial in episodic migraine, the adverse event profile was similar with erenumab 140 mg and placebo. The most frequent adverse events were upper respiratory tract infection (erenumab 4.7% vs. placebo 5.6%), injection site pain (0.3% vs. 0.3%), and nasopharyngitis (11.0% vs. 10.0%) (Goadsby et al. N Engl J Med 2017;377:2123-32). The incidence of serious adverse events (1.9% vs. 2.2%) and adverse events leading to discontinuation (2.2% vs. 2.5%) was similar in the two groups. A similar safety profile has been reported in phase III testing of the MAbs targeting the CGRP molecule.
A number of phase III trials are completed or ongoing for erenumab (STRIVE, ARISE, LIBERTY), fremanezumab (HALO EM, HALO CM), eptinezumab (PROMISE 1, PROMISE 2) and galcanezumab (EVOLVE-1, EVOLVE-2, REGAIN). Results to date indicate that targeting CGRP is an effective, well-tolerated approach for the prevention of episodic and chronic migraine.