Abstract
This editorial prefaces two papers addressing the possibility that 5-hydroxyptryptamine 1F(5-HT1F) receptor agonism may be a valuable new treatment principle for acute migraine attacks. Nelson et al. (1) describe the preclinical pharmacological profile of the selective 5-HT1F receptor agonist Lasmiditan and Ferrari et al. (2) present the first proof-of-concept study. Several years ago, an indole-based 5-HT1F receptor agonist, LY334370, showed promise in the treatment of acute migraine attacks (3), but was not further developed due to compound-related toxicity. Subsequently, the Eli Lilly Company launched a search for an improved molecule. It was requested that this molecule be more selective for the 5-HT1F receptor than LY334370, that it should not have affinity to other 5-HT receptors, in particular 5-HT1B and 5-HT1D receptors, and that it should be based on a new non-indole skeleton. The result of this search was Lasmiditan. In vitro binding studies showed a high binding affinity to the 5-HT1F receptor compared with the binding to the 5-HT1B and 5-HT1D receptors. The molecule also had a relatively low binding affinity to the 5-HT1A receptor. It did not bind to a host of other receptors and ion channels. In agreement with this profile, Lasmiditan did not constrict rabbit saphenous vein rings. Lasmiditan potently inhibited dural plasma protein extravassation and also inhibited electrically stimulated induction of the immediate early gene c-fos in the trigeminal nucleus caudalis. Finally, Lasmiditan was based on a scaffold not found in any other antimigraine drug. After preclinical development, Lasmiditan was taken over by the CoLucid company for clinical development.
Ferrari et al. (2) present the first clinical study where Lasmiditan has been tested in the treatment of acute migraine attacks. A slow intravenous infusion of the drug was used, possibly in order to avoid variation in absorption and better to be able to do pharmacokinetic–pharmacodynamic modelling. This route of administration required patients to come to the hospital for treatment, a notoriously difficult approach as many patients do not come despite initial expression of interest. The difficulty in recruiting to in-patient treatment were overcome partly by using the relatively novel, highly effective group-sequential adaptive treatment assignment design. This resulted in the selection of an effective dose of 20 mg after testing a wide range of doses despite treating only 130 patients. Interestingly, this is very similar to the number needed to demonstrate efficacy of a calcitonin gene-related peptide (CGRP) receptor antagonist (4). The study by Ferrari et al. (2) proves beyond doubt that 5-HT1F receptor agonism has efficacy in the treatment of acute migraine attacks. While the last word has not been said about the most effective dosing of Lasmiditan, the study allows a very good guess at the doses to be tested in the future development of the compound.
After CGRP antagonists, 5-HT1F agonists are the second class of substance to prove that migraine attacks can be treated with a drug that has no vascular activity. Does this mean the end of the vascular hypothesis of migraine? As the only mechanism of migraine, vasodilatation has definitely been ruled out. However, could vasodilatation perhaps play a more limited role? Previous studies using transcranial Doppler demonstrated a slight dilatation on the painful side of a migraine attack compared to the non-painful side (5). It was stated that this amount of dilatation was not itself enough to be the cause of migraine pain but that, combined with sensitisation of peripheral sensory nerve-endings, vasodilatation could contribute to migraine pain. A similar result was found regarding the superficial temporal artery using high frequency ultrasound imaging (6). In contrast, a magnetic resonance angiography study was unable to find any dilatation of the external portion of the middle meningeal artery or of the internal carotid or middle cerebral artery during nitroglycerin-induced migraine attacks (7); therefore, the question of vasodilatation in migraine is still unresolved. Furthermore, all endogenous substances that can cause a migraine attack by infusion are vasodilators and all marketed specific treatments for acute migraine attacks, ergots and triptans, are vasoconstrictors. A clue to the possible role of vasoconstriction may come from the absolute effect size of CGRP antagonists and Lasmiditan. While there is no statistical proof that CGRP antagonists are inferior to the triptans, the absolute effect size has tended to be lower both in oral and intravenous use (4,8). The absolute efficacy of intravenous infusion of Lasmiditan was not as high as the efficacy in previous studies of subcutaneous triptans. Also, the placebo effect in the Lasmiditan study was very high; consequently, the therapeutic gain (the difference between active and placebo) was relatively low. However, the design did not allow exploration of the optimal dose. Larger head-to-head comparative trials of triptans versus Lasmiditan are needed, preferably using an intravenous administration in order to eliminate variability due to absorption, to evaluate a possible difference in efficacy. While the importance of vascular mechanisms in migraine have thus been reduced over the last couple of decades, it is still not possible to decide whether vasoconstriction may add to the efficacy of acute migraine treatments.
Regardless of these mechanistic speculations, and acknowledging that the present study does not have the power or design to predict absolute efficacy or most effective dose, it must again be emphasized that the two papers this editorial concerns are of landmark significance. It will be exciting to follow the further development of Lasmiditan and to see how its role in acute migraine treatment will gradually be more precisely defined. Further studies of Lasmiditan should not only focus on its efficacy but also on its mechanisms of action. Understanding Lasmiditan fully will undoubtedly shed light on basic migraine mechanisms. Does the compound, for example, cross the blood–brain barrier? No data have been given on this in the two papers. More data are also needed on the pharmacokinetics of the compound and its pharmacodynamic actions outside and inside of the blood–brain barrier. Hopefully, the two published papers are just an important introduction to a range of exciting new discoveries.