Abstract
Similarities between Leao’s cortical spreading depression (CSD) (1), the propagation rate of Lashley’s scotoma (2), and perfusion changes occurring during migraine aura (3,4) have led to the hypothesis that CSD is the animal equivalent and the probable mechanism that underlies migraine aura (5). These data helped change thinking from vascular theories postulated by Wolff (6) to a more neuronal basis of migraine. Perhaps as a consequence, the importance or role of migraine aura (or CSD) in migraine pathophysiology has been overstated so that it is sometimes considered essential even in patients without clinically manifest aura.
The relative importance of the brain in the generation of the nociceptive trigger for the pain component of the migraine attack differs greatly among different theories of the disorder, and the possible role of aura as a trigger is a key component of this controversy. Simply put, one view is that migraine is driven by a sequential phenomenon that generates, in necessary order, activation of the trigeminovascular system, brainstem and diencephalic structures (as demonstrated in imaging studies) and the full spectra of the migrainous phenotype. Within this theory the primary event to trigger an attack is at the level of the meningeal nociceptors or first-order trigeminal neurons causing trigeminovascular activation, and continuous activation will sequentially drive further activation of the system (7). In addition, it is hypothesized that aura/CSD may form the trigger to activate the meningeal nociceptors. This is based substantially on findings that induction of CSD in rodents results in neuronal activation in the caudal portion of the trigeminal nucleus and dural blood vessel dilation (8,9), via activation of the peripheral meningeal nociceptors to activate neurons in the TCC. This is supported by electrophysiological data demonstrating prolonged increases in spontaneous firing in the trigeminal ganglion in 50% of animals tested, after CSD induction (10).
In a recent issue of the Journal Lambert and colleagues (11) present insightful and carefully conducted studies that challenge this theory, favoring the view that migraine pain and trigeminovascular activation can be, and probably is, a consequence of an essentially central mechanism not requiring a primary sensory input. First, they demonstrate that lignocaine injection into the trigeminal ganglion reversibly inhibits spontaneous discharge and dural craniovascular and facial evoked firing in the TCC. This illustrates that lignocaine blocks peripheral meningeal nociceptor inputs to the trigeminal nucleus and also that there is a basal level of traffic in the primary neuron. They further demonstrate that lignocaine has no effect on the burst in spontaneous discharge when given after CSD induction, nor does it prevent CSD-induced spontaneous discharge, even when the basal spontaneous firing has been attenuated by the prior injection of lignocaine. This is the clearest demonstration that activation of the trigeminovascular system, and therefore potential pain in migraine, does not necessarily require activation of the peripheral projections to the dural vasculature, but can be generated within the brain, via connections between the cortex and brainstem. Any changes in the periphery are a likely consequence of this centrally driven activation (9,10).
A prerequisite of the aura/CSD hypothesis is that it activates peripheral meningeal nociceptors to produce trigeminovascular activation in a sequential manner. On the basis of Lambert and colleagues’ study in rodents it would appear that a peripheral input is not required and may not be the mechanism by which CSD activates the trigeminovascular system; this is probably via a purely central mechanism. Although this can still imply that CSD is a trigger for migraine pain, a closer look at the clinical picture and things become less clear. Migraine aura affects only 20–30% of migraine patients (12). Clinically silent aura seems unlikely, particularly for those that suffer migraine for many years without ever experiencing aura. It also seems highly unlikely that the different forms of migraine are a consequence of different pathophysiological mechanisms, particularly when some patients may suffer both migraine with and without aura, and even more extreme forms of migraine than these, such as familial hemiplegic migraine. A common pathophysiology seems likely. If CSD is the mechanism that triggers head pain in both forms of migraine, treatments should be at least as effective in migraine without aura as in migraine with aura (13). Tonabersat is an interesting test of this concept: although it is ineffective in migraine as such, with and without aura (13,14), it is reported to reduce aura (13). Similarly, lamotrigine was ineffective in a randomized controlled trial in migraine (15), while being considered useful for aura (16). Taken together these data cast doubt on the CSD hypothesis of migraine. Moreover, the sequence of events during a migraine attack tells us that aura is not even the first event. It is preceded by premonitory symptoms that include tiredness, reduced concentration, irritability, yawning, and other non-headache symptoms hours or even days before the onset of aura or headache (17). These are symptoms most likely attributed to alterations in the hypothalamus, in the brain (18). By looking at the clinical phenotype of migraine with and without aura it seems unlikely that aura or CSD can be the trigger for the pain in migraine.
If migraine is not a consequence of a peripheral sensory input, triggered by CSD that sequentially activates regions in the brain, what is responsible for the mechanism behind this multi-faceted disorder? It seems ironic, but studies on CSD have made the answer much clearer: it is the brain. We believe that migraine is best explained as a ‘brain state’ that contributes to changes in cellular and vascular function in many regions of the brain in parallel. Migraine may be described as a dysfunction in the subcortical structures, including the brainstem and diencephalic nuclei that normally modulate sensory inputs. These nuclei serve as a ‘migraine mediator’ (19) and dysfunction may lead to the abnormal perception of basal level of primary traffic and therefore perceived activation of the trigeminovascular system and pathways that activate central structures. Therefore aura is triggered by the same mechanism that is responsible for the pain and other symptoms, and this process resides and is determined by dysfunction in the brainstem and diencephalon. The data presented from Lambert et al. provide the clearest example that trigeminovascular activation can be generated within the brain without a peripheral sensory input. The study does not definitively provide the answers to this theory of migraine pathophysiology; a central connection between the cortex and trigeminovascular system is clear, but how the subcortical structures modulate processes to produce aura is not. A step in the right direction has been made and the study goes some way to balance the long held, and misleading, view that migraine requires a peripheral sensory input.