Acetazolamide Acts on Neuromuscular Transmission Abnormalities Found in Some Migraineurs

Abstract

Mild subclinical impairment of neuromuscular transmission can be detected with single-fibre electromyography (SFEMG) in subgroups of patients suffering from migraine and could be due to dysfunctioning Ca²⁺-channels on motor axons controlling stimulation-induced acetylcholine release. Acetazolamide, which is thought to ameliorate ion channel function, was shown effective in familial hemiplegic migraine and episodic ataxia type 2, both of which are associated with mutations of the neuronal Ca²⁺-channel gene CACNA1A, as well as in aura status. We treated therefore in an open pilot study five non-hemiplegic migraineurs showing mild SFEMG abnormalities with acetazolamide for several weeks. This was followed by a normalization of SFEMG recordings in all patients and by clinical improvement in four. These results support the assumption that the subclinical impairment of neuromuscular transmission found in certain migraineurs might be due to dysfunctioning Ca²⁺-channels.

Introduction

Using single-fibre electromyography (SFEMG), we have recently found mild abnormalities of neuromuscular transmission in some migraineurs with prolonged and/or complex auras (sensorimotor, language or balance impairment) (1–3). In view of the genetics of migraine (4) and the electrophysiological data in tottering mice (5), these abnormalities were interpreted as the possible reflection of dysfunctioning P/Q-type Ca²⁺-channels, which are responsible for stimulation-induced acetylcholine release at the neuromuscular junction (6).

Familial hemiplegic migraine (FHM), a rare form of migraine with aura, and episodic ataxia type 2 (EA-2) are associated with mutations in the CACNA1A gene (chromosome 19p13) coding for the α_1A-subunit of voltage-dependent P/Q-type Ca²⁺-channels (7). Acetazolamide is able to modify ion channel function (8) and was found effective in EA-2 (9) and in some cases of FHM associated (10) or not (11) with cerebellar ataxia. In one report, acetazolamide was able to interrupt migraine aura status (12), a complication of migraine with aura in which no CACNA1A mutations have been found up to now.

We examined therefore the possible effect of acetazolamide on clinical symptoms and neuromuscular transmission in a heterogeneous group of five migraineurs selected because of abnormal SFEMG findings.

Case reports

Among the five enrolled migraineurs (four women, one man; median age 33 years, range 21–51 years; median attack frequency 4/month, range 2–20), two suffered from migraine without aura (MO; IHS code 1.1) (13) associated with vertigo, interictally in one (MOa), both ictally and interictally in the other (MOb). Three patients had migraine with aura (IHS code 1.2.1) (13): two had attacks both without and with typical aura (MO +MTAa and MO +MTAb) and the last one attacks with prolonged aura (MPA; IHS code 1.2.2 (13)). All these migraine with aura patients had regularly sensorimotor, language and/or balance impairment during the aura. No patient had any other detectable medical condition, nor was taking drugs on a regular basis or within 3 days before the first recording. The clinical course was monitored using headache diaries. All patients were allowed to treat any migraine attack occurring between the first and the second SFEMG session with non-steroidal anti-inflammatory drugs.

They underwent a stimulation SFEMG study (14) at least 3 days after the last headache attack with a Viking IV device (Nicolet® Biomedical, Madison, WI, USA). One of us (A.A.) stimulated suprathreshold the motor branch of the radial nerve and assessed the variability in latency, i.e. the jitter, of single-fibre action potentials in m. extensor digitorum communis (EDC) of the right arm. Stimulation acquisitions were 100 per endplate and stimulation rate was 10 Hz. Results were expressed as the ‘mean value of consecutive differences’ (MCD) of successive stimulus–response intervals. We recorded 25 endplates per patient and on average 18 of them, artefact-free, were selected off-line to assess the patient's mean MCD. In our previous studies, we collected our own normative SFEMG values from a group of 16 healthy volunteers with no personal or familial history of migraine, and thus with a low risk to develop migraine. In that group mean MCD was 17.09±2.65 µs and none of the subjects had ‘single endplate abnormalities’, i.e. an MCD value >40 µs in individual endplates (2). In the present study, a mean MCD >23 µs was therefore considered abnormal as well as the presence of abnormalities in single endplates expressed as a percentage of those used to calculate the mean MCD. All patients enrolled had ‘single endplate abnormalities’ on the first SFEMG recording.

They received thereafter acetazolamide (250 mg b.i.d. orally). In one patient (MPA) the dosage was reduced to 125 b.i.d. after a few days because of intolerable side-effects. All patients underwent a second SFEMG study between 4 and 16 weeks later (median 10 weeks). This second recording was performed by the same examiner, who was blinded for treatment modality and clinical course. In one patient (MOb) the follow-up was prolonged up to 11 months, during which acetazolamide treatment was interrupted for several weeks in order to verify the reversibility of both clinical and electrophysiological effects.

Ordinal values, expressed as medians and ranges, were compared with the Wilcoxon matched pairs test, nominal values by the Yates' corrected χ² test.

The study was conducted after approval of our institution's ethics committee and with the understanding and consent of each involved subject.

Results

The median mean MCD value was 23.1 µs (range 18.4–29.9 µs) at the first and 16.9 µs (range 13.7–19.0 µs) at the second recording. This reduction was significant (P=0.043). Three out of five patients (MOa, MOb and MPA) had mean MCD values above our normative limits (2) before treatment, compared with none during treatment with acetazolamide.

In the first SFEMG session, all enrolled patients had mild ‘single endplate abnormalities’ and the median percentage of individual fibres with abnormal MCD values was 6.25% (range 5–13.3%). At the second recording performed after several weeks of acetazolamide treatment, ‘single endplate abnormalities’ had disappeared (P=0.011). These results are summarized in Table 1.

Table 1

Electrophysiological findings

Patients	Age	Gender	First SFEMG recordings		Second SFEMG recordings
			Mean MCD (µV)	% of abnormal single endplates	Mean MCD (µV)	% of abnormal single endplates
MOa	50	M	24.9	13.3	19.0	0
MOb	51	F	29.9	12.5	13.7	0
MO +MTAa	30	F	18.4	5.55	16.9	0
MO +MTAb	21	F	21.9	5	17.4	0
MPA	33	F	23.1	6.25	16.0	0

From a clinical point of view, four patients out of five had a ≥50% reduction of attack frequency with acetazolamide. The median monthly attack frequency decreased from four (range two to 20) before to two (range zero to three) after treatment (P=0.053). No clinical improvement occurred in patient MOa. During treatment with acetazolamide patient MOb had a marked reduction of both migraine attacks and interictal episodes of vertigo. Frequency of the latter and MCD values varied in parallel during intake or discontinuation of the drug throughout follow-up (Fig. 1). One patient (MO + MTAa) suffering from migraine without and with typical aura became attack-free. In the second patient suffering also from both types of attacks (MO +MTAb), the auras became strictly visual while they were associated with sensory symptoms before acetazolamide. In the third migraine with aura patient (MPA), attacks that persisted were characterized by shorter and milder aura symptoms.

Figure 1

Clinical and electrophysiological follow-up of patient MOb. Hatched blocks indicate treatment periods with acetazolamide.

Discussion

In this open pilot study, the mild single endplate abnormalities detected by SFEMG in five migraineurs disappeared during acetazolamide treatment. The latter also produced clinical improvement in four of them.

Acetazolamide has shown effectiveness in P/Q-type channelopathies such as EA-2 and FHM (9–11). Recently it was found efficacious in three EA-2 patients with loss-of-function CACNA1A mutations and pronounced SFEMG abnormalities (12). The mechanisms by which it is thought to improve Ca²⁺-channel functions in these disorders are not well understood. Acetazolamide is known to induce metabolic acidosis. An acidic pH, however, has a rather negative effect on the neuromuscular junction, as it reduces the number of acetylcholine quanta released by nerve impulses in frog muscle and the amplitude of endplate potentials (15, 16). Metabolic acidosis may modify free Ca²⁺ and Mg²⁺ ions which have an effect on neuromuscular transmission. Ionized Mg²⁺ is known to antagonize Ca²⁺-induced neurotransmitter release at the NMJ, which depresses neuromuscular transmission (17). It reduces the frequency of miniature endplate potentials both in mice carrying the tottering mutation on the CACNA1A gene and in wild-type mice (5). A decrease in Mg²⁺ which would have a favourable influence on neuromuscular transmission has been reported in migraine with and without aura in brain (18), peripheral blood cells (19, 20), and serum (21). It is thus unlikely that a reduced Mg²⁺ level plays a role in the mild impairment of neuromuscular transmission observed in migraineurs and that its reversal by acetazolamide-induced acidosis explains the normalization of SFEMG studies shown here. Acetazolamide might improve neurotransmission via increase of ionized Ca²⁺ or via a direct action on P/Q Ca²⁺-channel function similar to the one shown for the sarcolemma K_Ca²⁺ channels (8).

Our preliminary observation that acetazolamide reduces frequency as well as severity and complexity of migrainous auras is in line with the report showing that it is able to interrupt migraine aura status (12). These data suggest that acetazolamide may reduce both occurrence and cortical extension of spreading depression, as already suggested by studies in animals (22). Migraine with aura can also be ameliorated by acetazolamide treatment when it occurs as a part of a CADASIL spectrum (23). This effect was tentatively attributed to an increased perfusion in oligaemic cortical regions. Increased blood flow in muscles is unlikely to be important in the effects of acetazolamide on the neuromuscular junction, as increased muscle temperature does not improve neuromuscular transmission (24).

Although some studies do not support the hypothesis that CACNA1A gene is involved in common forms of migraine (25, 26), migraineurs who present clinical aura features similar to those commonly found in attacks of FHM may nonetheless share a similar genetic background. One patient who belonged to a FHM family but had attacks with non-hemiplegic auras had a CACNA1A gene mutation (27) and sib-pair analyses are in favour of the involvement of this gene in the common forms of migraine, especially in migraine with aura (28, 29).

Other drugs such as the calcium antagonist verapamil were found effective in interrupting hemiplegic aura (30) and are supposed to act by reversing cerebral vasospasm (29). Interestingly, verapamil seems to act at some extent also on P/Q Ca²⁺-channels (31).

Placebo-controlled studies are needed to confirm that acetazolamide has beneficial effects in subgroups of migraineurs. The latter may be selected on the basis of their aura features, but SFEMG recordings could be of some help in this process.