Pharmacotherapeutics of migraine and the blood-brain barrier: serendipity, empiricism, hope, and hype.

Bigal and Krymchantowski[1] review the role of new and potential pharmacologic agents for the preventive treatment of migraine.[1] Whereas cortical spreading depression and brainstem involvement are presumed pathogenetic concepts, a substantial body of clinical, experimental, and pharmacologic evidence challenges the traditional view that migraine is the outcome of a brain neuronal aberration.[2–5] In direct contrast to attack-aborting agents, migraine-preventing agents unambiguously affect the primary pathogenetic processes (afferent limb of migraine) rather than influencing events of the trigeminovascular system (efferent limb of migraine).[3] To critically influence brain neuronal function, it cannot be argued that migraine prophylactic agents must readily or freely cross the intact blood-brain barrier (BBB).[2–4] First, atenolol, a first-line migraine preventive agent that – like propranolol – prevents both migraine with aura and migraine without aura but – unlike propranolol – does not readily cross the BBB or critically influence brain neuronal function.[2,3] Second, drugs that abort the aura of migraine, such as nifedipine, nitrates, and isoproterenol, do not significantly cross the BBB and, therefore, cannot directly influence the visual or other cortical functions.[3,4] Third, as reviewed, the migraine preventive effect of topiramate is unpredictable, not dose-related; does not prevent migraine aura; and is not superior to propranolol while being found effective for the prevention of cluster headache – a primary headache variant for which pathogenetic brain cortical neuronal dysfunction has not been envisaged.[4] Fourth, although lamotrigine does not prevent migraine under controlled conditions, acetazolamide has no direct action on the P/Q-type neuronal calcium ion channel.[4] Fifth, magnesium supplementation has evoked much interest in migraine prevention, but magnesium depletion is not specific to migraine,[5] and exogenous magnesium does not readily cross the BBB.[4] Sixth, amitriptyline, a migraine preventive agent in the “well-accepted” or “proven” category,[6] is an established combined activator of both noradrenergic and serotonergic brain neuronal systems and a “proepileptic” agent.[3] If both brain noradrenergic suppression (induced by propranolol) and excitation (induced by amitriptyline) can prevent migraine, a reorganization of concepts is essential.[3] Seventh, the combination of propranolol and amitriptyline that has been found to be therapeutically useful for migraine or mixed headaches[7] appears to be pharmacologically invalid insofar as alteration of brain noradrenergic function is concerned, but is nevertheless clinically effective and therefore conceptually challenging.[3] Finally, the use of botulinum toxin for migraine prevention is a serendipitous exercise that is not supported by available clinical and basic science evidence.[8] Botulinum toxin has not been found to have a genuine or direct analgesic action in humans, manifests short-lasting modulation of analgesia-related behavior in rats, does not cross the BBB, and can precipitate headaches.[8] Suppression of release of substance P, a suggested mechanism of action for botulinum toxin,[1] does not contribute to analgesia in humans.[9] Randomized controlled trial (RCTs) are a useful clinical tool, the limitations of which must be comprehended.[5,10] RCTs mask biological idiosyncracies, and their results can remain ambiguous; clinical trials are not designed to answer questions about pathogenesis and cannot supplant the need for balanced conceptual groundwork.[5,10] The biology of migraine is in its infancy, and use of mathematical logic in RCT to justify the use of newer agents at this stage can worsen the extant pathophysiologic confusion. Being entirely subjective, endpoints in migraine research-related RCTs are relatively much softer. Besides, it seems to be becoming increasingly difficult to maintain perspective that even the advent of propranolol for migraine prevention was a serendipitous event or that suggested efficacy for migraine prevention for any pharmacologic agent in RCTs does not render such therapeutic efforts any less empiric.[4] Approval of topiramate for migraine prophylaxis by the US Food and Drug Administration[1] does not address any of the limitations of the drug discussed herein, and it is premature to currently consider it (or any other new anticonvulsant agent) as a first-line migraine preventive agent. Additionally, a search of PubMed and of Medline did not indicate that carvedilol crosses the BBB or influences brain neuronal function. Finally, in attributing an important role to nitric oxide in migraine,[1] it is essential to underscore that nitroglycerin, a nitric oxide donor, induces migraine attacks as well as instantaneously aborts the migraine aura.[11] There is no central idea in migraine to elaborate a general theory that could ultimately lead to the creation of a unifying hypothesis to collect the various strands of evidence into a coherent and logically defensible, intelligible synthesis. Migraine is a classic example of a disease entity in which RCTs are increasing conceptual confusion by pushing issues that really matter into the background.[12] The suggestion that the ideal migraine drug would be effective both for prophylaxis and for management of attacks[1] merits reflection because it basically assumes a significant overlap between the physiologic basis of the “afferent” and the “efferent” limbs of migraine. The phenomenology, and underlying physiologic basis, of migraine attacks is radically different from the interictal state (in-between attacks), and a drug that might manage both states is highly unlikely to emerge. Readers are encouraged to respond to George Lundberg, MD, Editor of MedGenMed, for the editor's eyes only or for possible publication via email: ten.epacsdem@grebdnulg