Biological insights from the premonitory symptoms of migraine

DOI: https://doi.org/10.1038/s41582- 018- 0098- 4

Abstract-Summary Premonitory (prodromal) symptoms can start hours to days before the onset of a migraine headache and can predict its onset in some individuals.

This earliest phase of the migraine attack provides valuable insights into the neurobiology of the disorder, furthering our understanding of how and why these phenotypically heterogeneous symptoms are mediated.

Improvements in our understanding of migraine could provide novel therapeutic opportunities, with the possibility of closing the therapeutic gap that remains owing to a lack of sufficiently effective and well-tolerated acute and preventive treatments. In this Review, we discuss the current evidence in the literature in relation to the phenotype and mediation of premonitory symptoms in migraine, and discuss the neurobiological insights gained from these studies.

Extended: In this Review, we focus on the prodromal symptoms, which we refer to as the premonitory phase in line with the current literature; the terms prodromal and premonitory are interchangeable.

In this Review, we have set out some aspects of the complex pathophysiological and neurobiological basis of migraine, in particular in relation to the emerging study of the premonitory stage.

Introduction Premonitory symptoms can start and be present during any phase of the migraine attack, but are perhaps less noticeable in the headache phase.

The broad clinical symptomatology can contribute to the disability associated with migraine attacks and indicates involvement of brain areas outside the well- accepted pain network, such as emotional processing areas and sensory areas [210], suggesting wider CNS involvement in migraine.

This Review focuses on the prevalence, clinical phenotypes and current under- standing of the neurobiological basis of premonitory symptomatology in migraine and considers how an improved understanding of this phase of the condition could lead to future research into therapeutics.

Without considering premonitory symptoms but acknowledging the presence of disease-defining canonical symptoms in migraine, this concept leads to an under- standing of migraine as a brain disorder that involves more of the brain than the pain matrix and at least some additional sensory processing areas.

Premonitory Symptoms in Migraine Consistent with this concept, one study showed that when individuals in a large cohort of patients with migraine were exposed to commonly perceived triggers,

272

2 Mechanisms

such as flickering light, specific foods and odours, patients who considered their migraines to be triggered by these factors were more likely to report the correspond- ing symptoms of photophobia, food cravings and osmophobia in the premonitory stage [211].

When assessing a patient, it is important to ask them what they consistently and reliably find triggers a migraine attack, the time between exposure to such a trigger and the onset of a migraine, and specific premonitory symptoms that are associated with their attacks.

In an important study in which prediction of attacks was tracked using a prospec- tive electronic diary design, 72% of 97 patients could correctly predict headache onset within 72 h on the basis of premonitory symptoms [212].

Brain Activity in the Premonitory Stage Evidence suggests that neurophysiological activity in frontal cortical areas and their limbic connections—probable mediators of premonitory symptoms such as diffi- culty concentrating, fatigue and emotional changes—is altered before headache onset in migraine [213].

Another study demonstrated that a component of the visual evoked potential is increased in people with migraine 72 h before the onset of headache, again suggest- ing an increase in visual cortex responsiveness in the preictal period compared with interictal responsiveness [214].

These studies suggest that cortical function changes in the lead up to a migraine headache and that this dysfunction could mediate some of the symptoms experi- enced during the premonitory stage.

These findings correlate well with those of neurophysiological studies of cortical activity in the lead up to a migraine attack and explain some of the sensory sensitiv- ity and the mood and attentional symptoms that patients experience during the pre- monitory stage.

Insights from Premonitory Symptoms These studies strongly suggest an early role for dopamine systems in migraine, particularly given the high reliability, predictive value and specificity of this symp- tom in migraine.

Further evidence for a role of dopamine in migraine comes from clinical use of domperidone as an anti-emetic and gastric motility agent in migraine and from early studies that suggested that domperidone was beneficial for headache prevention when used during the premonitory stage [215, 216].

A study in which patients with nausea were compared with those without dem- onstrated that nausea was associated with increased activity in the rostral dorsal medullary area, indicating that this area probably mediates nausea during migraine [217].

These studies not only provide evidence for probable central mechanisms that mediate nausea and photophobia in the premonitory stage of migraine, but also sup- port the theory that cortical and subcortical dysfunction occurs via various neuro- chemical systems before headache onset in migraine.

2.2 Biology

273

Emerging Relevant Neurochemical Systems Given the areas of the brain that are activated during the premonitory stage and the clinical phenotype of symptoms experienced, several other brain neurochemical systems are of interest in relation to migraine.

Somatostatin is a neuropeptide that is secreted by the posterior hypothalamus and is involved in several brain functions, including motor and sensory function, sleep and cognition [218, 219].

Basic research has demonstrated that somatostatin has a role in the descending modulation of trigeminovascular signalling, probably via GABAergic mechanisms, as augmentation of somatostatin receptors in the posterior hypothalamus with cycloso- matostatin reduced A-fibre and C-fibre activity in response to dural stimulation [220]. One study has shown that cholecystokinin expression in the hypothalamus increases in response to noxious trigeminal stimulation [221], suggesting that the peptide is involved in the mediation of appetite loss in migraine.

Thirst and polyuria can be premonitory symptoms of migraine and that the hypo- thalamus secretes ADH, we think it may well prove to be a substance that warrants further research attention in migraine.

Conclusions and Therapeutic Relevance The presence of symptoms, altered brain physiology and abnormal functional imag- ing findings during this stage confirm the involvement of several brain areas before the onset of pain, rather than in response to pain, and provide evidence for a central neural basis to symptoms that some people with migraine experience during an attack. Cortical and subcortical brain areas, the ascending and descending connections of these areas, and neurochemical systems, have been implicated in mediating pre- monitory symptoms and trigeminovascular nociception.

Functional correlations between the phenotype of the premonitory stage and the brain areas that are activated during this stage in functional imaging studies, as well as basic laboratory studies that have examined the neurochemical pathways that project from these regions in more detail, have helped to identify neurochemical systems that mediate these symptoms, and have consequently identified potential therapeutic targets.

Acknowledgement A machine generated summary based on the work of Karsan, Nazia; Goadsby, Peter J. 2018 in Nature Reviews Neurology.

Behavioral and cognitive animal models in headache research