Intra-arterial lidocaine therapy via the middle meningeal artery for migraine headache: Theory,current practice and future directions

Introduction

Recent developments in neurointerventional research have seen the emergence of the middle meningeal artery (MMA) as an effective conduit for intra-arterial (IA) therapy for a number of pathologies. These include embolisation of dural arteriovenous fistulas (dAVFs), ¹ chronic subdural hemorrhage (cSDH) ² and dural tumours such as meningiomas. ³ More recently, the use of IA lidocaine to treat refractory trigeminal neuralgia has demonstrated efficacy likely through the induction of a trigeminal nerve ganglion blockade. ⁴ This raises the possibility that IA lidocaine therapy via the MMA may have a role in treating migraine.

Migraine is an extremely common and debilitating neurological phenomenon with an estimated prevalence of 14–15%. ⁵ The headache is typically unilateral and often associated with sensitivity to light, sound, smells and head movement. However, the spectrum of migraine disorder may range from episodic headache with or without sensory aura to status migrainosus, a migraine attack which does not terminate after 72 h despite maximal medical therapy. ⁶

The pathophysiology of migraine is an ongoing area of debate which is subject to a large body of literature, the scope of which is beyond this paper. Current thinking may be summarised – with the aid of Khan et al.'s (2021) thorough review ⁷ – as an interaction of trigeminovascular activation, ⁸ neuronal modulation^9,10 and metabolic triggers, ¹¹ underwritten by predisposing genetic susceptibility factors. ¹² Whilst the precise origin of the headache is unknown, the widely accepted key role of dural nociception in relation to trigeminovascular nerve sensitisation implicates the MMA, which supplies the dura, as a potential avenue for treatment. In this review, we present the literature which supports the role of MMA in IA lidocaine treatment of migraine.

The MMA and the trigeminal nerve ganglion

The MMA is a branch of the internal maxillary artery (IMA), which in turn arises from the external carotid artery. After entering the foramen spinosum, the intracranial component of the MMA gives off the anterior and posterior dural branches. Prior to entering the foramen spinosum, the extracranial component of the MMA gives off the artery to the trigeminal nerve ganglion into the foramen ovale. ¹³

Experimental support for the MMA providing a conduit to the trigeminal nerve ganglion has been demonstrated by the trigeminal-cardiac reflex being triggered by dimethyl sulfoxide/ethylene vinyl alcohol copolymer injection into the MMA.^14,15 This effect is also a well-documented consequence of operative manipulation of the trigeminal nerve at the cerebellopontine angle. ¹⁶ Moreover, resolution of trigeminal neuralgia has been reported following embolisation of MMA-related arteriovenous malformations. ¹⁷

More recently, direct attempts to treat refractory trigeminal neuralgia in 3 patients have been made with MMA IA lidocaine injections of up to 50 mg adjacent to the artery to the trigeminal nerve ganglion. ⁴ In all patients, suppression of trigeminal nerve ganglion function was observed, manifesting as dose-dependent latency prolongation and amplitude reduction of direct and consensual blink reflex responses. Clinical improvement was concurrent with electrophysiological suppression of trigeminal nerve ganglion function in two of the three patients, with the third patient being under general anesthetic during the procedure. All three patients reported reduction in hyperalgesia in all sensory dermatomes of the trigeminal nerve following IA lidocaine injection. For two of these patients, the effects were protracted (for 3 and 5 days, respectively).

It is worth noting that temporary sensory loss in part or all of the trigeminal nerve distributions has also been reported with IA injection of lidocaine in the IMA when performed as provocative test prior to embolisation of dAVFs, reinforcing that the IMA may be implicated in trigeminal ganglion supply in some cases. ¹⁸

Overall, early evidence supports trigeminal nerve ganglion blockade as a rationale for an IA treatment option via the MMA in selected cases of trigeminal neuralgia. In a similar way, trigeminal nerve modulation may provide an avenue for treating migraine.

The trigeminovascular hypothesis of migraine

The idea that a pathophysiological link exists between migraine and the innervation of the meninges dates back to 1979. ¹⁹ The term ‘trigeminovascular system’ was coined in the subsequent decade to describe the immunohistochemical and neurochemical interaction between the trigeminal nerve, the central nervous system and both dural and pial arteries.^20–23 In more recent iterations of our understanding, migraine-specific triggers ¹¹ cause primary brain dysfunction, ²⁴ which causes dilation of pial and dural blood vessels. The dilated blood vessels mechanically activate perivascular trigeminal sensory nerve fibers which cause a pain response to be conveyed to the brainstem and from there to higher brain centres. Of vital importance, trigeminal nerve activation also evokes the release of potent vasodilatory peptides from the afferent nerve endings specifically around the dural vessels. Examples of these peptides are substance P, calcitonin gene-related peptide and pituitary adenylate cyclase-activating polypeptide. These peptides exacerbate vasodilation and cause neurogenic inflammation characterised by leakage of blood vessels and degranulation of mast cells. ²⁵ The vasodilation and neurogenic inflammation further increase activation of the sensory trigeminal fibers, perpetuate more vasoactive peptide release and modulate transmission of pain impulses to the brain. As migraine progresses, the brainstem and spinal cord centres that are the first to receive the pain impulses from the trigeminal nerve are hypothesised to become sensitised with a resultant worsening of headache pain and increased sensitivity to environmental and other stimuli. ²⁶ The complex role that the trigeminovascular system plays within the overall pathophysiology of migraine has provided rationale for the IA treatment of migraine headache with lidocaine via the MMA.

Current evidence for the treatment of headache with IA lidocaine via the MMA

In-human evidence for IA headache therapy remains sparse. A retrospective analysis of 76 patients undergoing MMA embolisation for cSDH discovered that, of the nine patients who concurrently suffered from chronic headache prior to treatment, seven reported complete post-embolisation headache resolution. ²⁷ Another series reported the treatment of two patients with intractable headaches (i.e., failing conventional medical management) following intraparenchymal haemorrhage. ²⁸ Both patients received IA injection of 40 mg lidocaine and 20 mg methylprednisolone into the MMA. An initial improvement of headache pain manifested within 5 min of injection (10/10 to 5/10 and 7/10 to 4/10, respectively) and lasted between 5 and 8 h after injection. Both patients then reported reduced headache intensity 24 h post-procedure compared to pre-procedural levels. These types of chronic headaches, secondary to intracranial haemorrhage, are examples of pain which can be readily attributed to dural nociception.

A more recent series applied the same treatment principles to manage four patients with treatment-resistant headaches of different aetiologies. ²⁹ The first patient (53 years of age, female) complained of persistent throbbing headache (average severity of 6/10) following diagnosis of subarachnoid haemorrhage 3 weeks earlier with no offending vascular lesion demonstrated on cerebral angiography. After discharge, the patient's headache continued despite the use of hydrocodone–acetaminophen, naloxone and hydromorphone. A subsequent twice daily regimen of topiramate 25mg yielded minimal benefit. The patient was scheduled for a follow-up six-vessel cerebral angiogram and agreed to undergo IA lidocaine injection in middle meningeal arteries as part of cerebral angiogram. The patient reported improvement in headache (self-reported intensity of 8/10 to 0/10 after lidocaine injection) and was discharged. The patient was completely pain-free on day 0 and day 1 and pain recurred on day 2 (self-reported intensity of 6/10). By day 7 post-procedure, the severity was still lower than pre-procedure severity (5/10 vs. 8/10).

The second patient (52 years old, male) presented to the emergency department with severe intractable headaches in the absence of a migraine history. During the admission, the patient had computed tomography scan and magnetic resonance imaging of the brain and cerebrospinal fluid analysis by lumbar puncture, all of which did not identify any abnormalities. For headache management, the patient had received multiple doses of intravenous magnesium sulphate, ketorolac, valproate sodium and hydromorphone hydrochloride. The patient was scheduled for a six-vessel cerebral angiogram to identify any evidence of reversible cerebral vasoconstriction syndrome or vasculitis. The patient agreed to undergo IA lidocaine injection in middle meningeal arteries as part of the cerebral angiogram. The patient reported improvement in headache (self-reported intensity of 10/10 to 0/10 after lidocaine injection) and was discharged following the procedure. The patient was completely pain-free on day 0 and day 1 and pain recurred on day 2 (self-reported intensity of 7/10). At day 7 post-procedure, the severity was still lower than pre-procedure severity (8/10 vs. 10/10).

The third and fourth patients were known migraineurs who presented with status migrainosus. ⁶ One patient (55-year-old, female) received IA lidocaine to the left MMA. The patient reported improvement in headache from a self-reported intensity of 5/10 (on numeric rating scale) to 0/10 after lidocaine injection. The patient's post-treatment three-month Migraine Disability Assessment Test ³⁰ score was lower compared with the pre-treatment three-month score (3 vs. 30). This conforms to a post-procedure improvement from severe disability to little or no disability. The other patient (45 years old, female) received bilateral MMA lidocaine injection with a self-reported post-injection headache improvement from 9/10 to 2/10 in intensity. The patient's post-treatment score was lower compared with the pre-treatment three-month score (90 vs. 55); however, both the pre- and post-procedure scores met the definition of severe disability. This initial evidence demonstrates feasibility for the treatment of selected cases of headache, including migraine, with MMA lidocaine injection.