Abstract
Migraine is a common, sometimes severe, episodic disorder of the central nervous system affecting an estimated 10–15% of the population (1, 2) and is associated with mostly unilateral headache, sensitivity to movement and autonomic symptoms such as nausea and vomiting, as well as photo- and phonophobia. Of these symptoms, pain is best correlated with the extent of disability. The second revised classification criteria of the International Headache Society (ICHD-II) defines chronic migraine (≥15 days) as a separate entity existing with and without association with medication overuse in a particular patient (3). Medication overuse has been considered to be the main risk factor for the chronification of migraine, underestimating the considerable number of chronic migraine patients without history of medication overuse. The true pathophysiological mechanisms underlying chronification remain unclear.
Previous studies have been able to show that antinociceptive brainstem structures such as the periaqueductal grey (PAG), the locus coeruleus (LC) and the nuclei raphe (RN) play a fundamental role in the pathophysiology of migraine (4–6). Dysfunction of these antinociceptive structures could involve a disinhibition of the trigeminal nociceptive system and thus might lead to the initiation of a migraine attack. Moreover, abnormal modulation of this migraine generator might be responsible for shifting the primarily transient and episodic nature of this disorder to something permanent with continuing illness (7, 8).
We present the follow-up findings on a previously reported case of symptomatic migraine due to a central pontine vascular malformation. The patient showed chronification of migraine symptoms over the previous 5 years and a shift from strictly right-sided migraine attacks to a bilateral manifestation. The slow increase in attack frequency over time lead to the assumption that the presented symptoms may be classified as chronic migraine. The pathophysiological mechanisms behind the development of chronicity, as well as the laterality of brainstem activation in migraine, will be subjects of further discussion.
Case report
A 38-year-old caucasian female was referred to our headache clinic with a 5-year history of bilateral, frontotemporal, severe throbbing/pulsating headache [score 6–7 on a visual analogue scale (VAS)] that showed aggravation on physical activity and usually lasted for about 24–48 h. The headache attack was accompanied by extensive nausea and vomiting, as well as photo- and phonophobia. Attacks were sporadic initially but frequency increased to about four to six times per month and were almost always preceded by a visual aura which the patient described as flickering and flashing of light only of her right visual field. There was no family history of migraine or of other primary headaches. There were no cranial autonomic features during the attack. Neurological examination revealed a slight cutaneous hyperalgesia and allodynia of the right forehead interictally, but no other neurological deficits. Migraine attacks responded well to subcutaneous application of triptans (sumatriptan 6 mg s.c.).
In the follow-up examination the patient reported that in the last 1.5 years she had experienced a gradual increase in attack frequency, reaching almost daily frequency (>15 per month). The clinical presentation of her migraine attacks changed: she experienced bilateral headache with lower pain intensity, shorter duration (6–12 h) and no visual aura. Neurological examination revealed only a slight remaining hyperalgesia interictally, while allodynia had completely vanished. Due to extensive use of analgesic drugs, this was first interpreted as a possible medication-overuse headache; the symptoms persisted after withdrawal, however, so that we reclassified the headache as possible chronic migraine with over 15 headache days per month lasting over 3 months. Migraine prophylactic medication was administered (flunarizine 10 mg daily) for approximately 4 months with only minor pain relief initially. Amitriptylin 75 mg for a treatment period of 10 weeks daily did not have a major impact. Triptan medication was still able to give some relief in the acute attack, but could not attain its former effectiveness.
Magnetic resonance imaging (MRI) showed a hyperintense lesion on T2 weighted sequence, reaching 5 × 10 × 12 mm in the central pons, slightly lateralized to the left, with no significant change after administration of contrast agent [Gadopentate dimeglumine (Gd-DTPA)](Fig. 1A,B). The haemosequence confirmed this finding, raising suspicion of capillary teleangiectasias with signs of residual haemorrhage (Fig. 1C). There was only a minor change of size in comparison with the previously performed MRI, but new evidence for haemorrhage not previously detected. This was the main reason why the vascular malformation was classified as cavernoma on the first neuroradiological MRI report (9).
Magnetic resonance imaging showing axial (A) and sagittal (B) T1 weighted image demonstrating capillary teleangiectasias in the central pons slightly lateralized to the left confirmed by an axial (C) haemosequence showing 5 × 10 × 12 mm hypointensity in the pons.
Electrophysiological nociceptive blink reflex (nBR) examination was not repeated. Since the patient developed bilateral headache, a side-to-side comparison was not possible. The currently existing findings on the nBR in migraineurs are based solely on differences in group and not single case effects. Therefore a novel nociceptive blink reflex examination seemed unreasonable.
Discussion
The presented case emphasizes the role of chronic impairment of antinociceptive brainstem structures in the pathophysiology of migraine chronification. MRI revealed a central pontine vascular malformation slightly lateralized to the left brainstem, including the raphe nuclei bilaterally, showing signs of subsided microhaemorrhage. This lesion is located in the same region reported to be associated with migraine on positron emission tomography (PET) (5, 6, 10). It seems this lesion may have caused small, asymptomatic bleeds over a prolonged period of several months to years, causing slowly progressing damage to adjacent regions of the brainstem. Thus, it can be assumed that the patient has developed symptomatic migraine with secondary chronification due to the continuing alteration of the central pons and the raphe nuclei. This case has been published previously (9) and presented itself at that time with strictly right-sided migraine with visual aura and slowly progressing frequency of attacks. This headache was already linked to the vascular malformation in the left paramedian pons, causing symptoms contralaterally.
An electrophysiological association with the nBR was established, showing that the lesion in the left nucleus raphe magnus accounted for a facilitation of the right trigeminal nociception and right-sided migraine attacks (9, 11). This case provides strong evidence that midbrain dysfunction may play an essential role in the basic pathophysiology of migraine and, by considering the current progress, may present a plausible biological explanation for the developing mechanisms of chronic migraine. Since the raphe nuclei have intensive and complex connections to other antinociceptive brainstem nuclei (7), including bilateral projections to PAG and spinal trigeminal nuclei (12), they may also have a key role in the dysfunction of this antinociceptive brainstem circuitry either through aberrant activation or modulation of impulse flow in the trigeminal system. This is particularly so considering the narrow alignment of central structures in the brainstem, where the raphe nuclei are located right beside the midline and bilateral involvement caused by a paramedian lesion is easily conceivable.
How this central sensitization and neuromodulation leads to the actual migraine attack is not entirely clear. Vasodilator peptides such as calcitonin gene-related peptide (CGRP), substance P, neurokinin A, neuropeptide Y (NPY) and galanin are found in trigeminal neurons and could be involved in vascular nociception, supporting the neurogenic inflammation theory that proposes CGRP release from trigeminal sensory afferents causing vasodilation and plasma extravasation from dural vessels (13–15). That is controversial, however, because this effect has not been demonstated in humans during migraine (16, 17). The importance of CGRP release due to neuronal hyperexitability and the up-regulation of counteracting antinociceptive peptides in central sensitization is indisputable, but the underlying specific biochemical mechanisms are not entirely understood. A channelopathy of the P/Q-calcium channel has been suggested, since microinjection of agatoxin in the PAG led to a facilitation of trigeminal nociceptive activity (18). Other mechanisms such as the sensitization of the trigeminal nucleus by magnesium gating of n-Methyl-D-Aspartat (NMDA) receptors have also been discussed (19).
Abnormally high iron levels (a marker of disturbed neuronal function) have been discovered recently in the migraine generator structures (PAG, RN) of patients with episodic and chronic migraine (8). Alteration of these structures or failure of the generator to switch on or its inappropriate activation might lead to the initiation of a migraine attack. Permanent dysfunction may lead to chronicity and might explain the vulnerability of developing pain in other body systems. Iron accumulation could be caused by repeated hyperoxia due to activation with subsequent free radical-mediated cell damage in metabolically highly active structures with a high iron turnover or by iron accumulating from other sources such as haemoglobin following microhaemorrhage. Dysfunction of normally tightly maintained iron homeostasis may lead to disturbed brain function and could explain the common transition of an episodic illness to becoming chronic over time (20). Cellular dysfunction or damage may be detected by iron deposition. Cortical antioxidant gene up-regulation was shown in a mouse model to support the neuronal protective response to free-radical cell damage during a migraine attack (21). Oxidative DNA injury in cerebral neurons after experimental intracerebral haemorrhage has been demonstrated in mice, suggesting that oxidative stress and free radical formation leads to 8-hydroxyl-2′-deoxyguanosine (8-OHdG) accumulation and may lead to neuron damage and apoptotic cell death (22). The neurotoxic nature of iron has been discussed in correlation with many other neurodegenerative processes such as Alzheimer’s disease, Parkinson’s disease and Huntington’s disease (23). Epidemiological and genotyping studies have linked haemochromatosis (an autosomal recessive disease associated with inappropriately increased iron absorption) with an increased risk for headache and possibly migraine (24).
The case reviewed here as well as the two reported by Goadsby (2002) and Afridi et al. (2003) involved vascular malformations in the migraine generator region that were suspicious of past microhaemorrhage which might have led to the accumulation of iron initiating the mechanisms discussed above (6, 7). Actual iron measurements were able to provide some but not entirely conclusive evidence that levels in the migraine generating structures might be abnormally high in migraineurs (8). A different reported symptomatic migraine case had a brainstem glioma in the right dorsal pons causing left frontal head pain, also leading to neuronal decline (25).
In summary, a follow-up case of chronic symptomatic migraine due to a pontine vascular malformation with signs of residual haemorrhage is presented here for comparison with the previously reported condition from our headache clinic and two similar cases from the recent literature. An important role of the RN and other antinociceptive brainstem structures is emphasized in the pathophysiology of migraine. A connection of a possible iron overload in these structures and disturbed brainstem function are established, offering a reasonable explanation for a possible link of residual haemorrhage and the development of chronicity in migraine, which has yet to be demonstrated with certainty. This is the first reported case where a change in the apparent origin of migraine leads to a plausible explanation for a shift in symptomatology and the development of chronic features of this disease, possibly taking us one step further to an understanding of the mechanisms underlying chronicity in migraine.