Abstract
We showed reduced motor intracortical inhibition (ICI) and paradoxical increase of intracortical facilitation (ICF) to 1 Hz repetitive transcranial magnetic stimulation (rTMS) in patients affected by migraine with aura (MA). In conditions of enhanced excitability due to a reduced inhibition, high-frequency rTMS was found to potentiate intracortical inhibition. Here we explored the conditioning effects of high-frequency priming stimulation of motor cortex with the aim of normalizing excitability reverting paradoxical facilitation by 1 Hz rTMS in MA. Nine patients with MA and nine healthy controls underwent a paired-pulse TMS paradigm to evaluate motor intracortical excitability (ICI and ICF) before and after the following rTMS conditions: 1 Hz alone or preceded by a real or sham conditioning high-frequency (10 Hz) rTMS. Sham was used to control for rTMS specificity. In baseline, ICI was significantly lower in migraineurs with respect to controls. One hertz stimulation reduced motor evoked potential amplitude and ICF in healthy controls, while it caused a significant paradoxical ICF increase in migraineurs. High-frequency rTMS conditioning normalized excitability in migraine, increasing short ICI and so reversing the paradoxical effects of 1 Hz rTMS. These findings raise the possibility that the interictal reduced intracortical inhibition in migraine could be normalized by high-frequency rTMS. This would open perspectives for new treatment strategies in migraine prevention.
Introduction
Abnormal cortical excitability has been described in migraine, especially in migraine with aura (MA), and it has been suggested to play an important role as a possible predisposing factor to cortical spreading depression. Indeed, this phenomenon has been hypothesized to represent the pathophysiological basis of the migraine aura (1). The relevance of abnormal cortical excitability has received further support in recent years by the finding that gene alterations underlying different forms of familial hemiplegic migraine (FHM1 (2), FHM2 (3) and FHM3 (4)) all affect membrane structures (calcium channels in FHM1, ATPasi in FHM2, sodium channel in FHM3) that are critical for neuronal excitability.
We recently found that patients affected by MA present abnormalities of primary (5) and secondary visual cortex (6) with paradoxical facilitatory response to 1 Hz repetitive transcranial magnetic stimulation (rTMS), a stimulation frequency that in normal subjects reduces cortical excitability. According to Ziemann et al. (7, 8), we interpreted these results as following reduced efficiency of inhibitory circuits, unable to be up-regulated by low-frequency rTMS. Then, in a further study, by using the technique of paired pulse magnetic stimulation that allows selective evaluation of intracortical inhibitory and facilitatory circuits in the motor cortex, we were able to confirm previous results showing reduced activity of motor intracortical inhibition in patients affected by MA (9). Also in this study, the effects of low-frequency rTMS consisted of a paradoxical increase of the activity of facilitatory circuits. More recently, we have found that following light deprivation, a condition known to increase cortical excitability through reduction of inhibitory efficiency, the visual cortex of normal subjects behaves like that of migraineurs, showing a paradoxical response to low-frequency rTMS (10). In this experimental set we found that low-frequency rTMS delivered in the last 15 min of a 1-h period of light deprivation increased visual cortical excitability greatly delaying recovery to normal excitability level (phoshene threshold values) after light re-exposure. More interestingly, we found that high-frequency 10-Hz rTMS reduced cortical excitability favouring a more rapid and rapid return to normal excitability levels immediately or shortly after light re-exposure. So it appeared that high-frequency rTMS was able to restore normal levels of cortical excitability, probably through the resetting of cortical inhibitory circuits. According to this finding, some authors have shown that rTMS effects critically depend on the basal state of cortical excitability (11,12). In particular, in the study by Daskalakis et al., subjects with basally reduced short intracortical inhibition (SICI) showed a potentiation of inhibitory circuits by high-frequency rTMS (reset of SICI), whereas no effects were observed in subjects with higher SICI levels at baseline. Moreover, Lefaucheur et al. in a recent study demonstrated that high-frequency rTMS (hf-rTMS) can restore normal SICI levels in patients with chronic pain who show reduced cortical inhibition in baseline (13).
Evidence of the potential therapeutic effects of hf-rTMS has been recently provided by Fumal et al., who showed the ability of hf-rTMS to restore the normal pattern of visual cortical excitability in migraine, normalizing visual evoked potential habituation to repeated stimulation (14).
On this basis, we planned to explore if the paradoxical effects observed of 1 Hz rTMS in migraine, and attributed to reduced cortical inhibition, could be reversed by a priming effect of hf-rTMS. The hypothesis was that hf-rTMS could reset cortical inhibitory interneuron activity allowing a normal response to the following low-frequency (lf)-rTMS train.
Methods
Demographic and clinical features of the patients
All patients experienced visual aura in at least 50% of their attacks. Somestesic aura was also present in three patients. Patients and controls were not taking any medication and, to avoid non-specific effects on cortical excitability, female subjects (patients and controls) were not examined during the menstrual phase (16). Patients were examined interictally at least 48 h before or after an attack (absence of attacks after the recording was checked by means of a telephone call 2 days after examination).
All subjects gave their informed consent and the study was conducted according to the Declaration of Helsinki (17).
Patients and controls underwent an experimental paradigm in which excitability of corticospinal tracts and of intracortical circuits was explored in baseline and after rTMS in three different conditions: lf-rTMS alone and lf-rTMS preceded by a real or by a sham (as control) priming train of hf-rTMS.
Subjects were comfortably seated in a chair and instructed to maintain relaxation throughout the experiment. They wore a tight-fitting plastic swimmer's cap in order to mark the optimum site of stimulation and coil placement. Electromyography (EMG) signals were recorded from the right abductor pollicis brevis (APB) muscle using 0.9 cm diameter Ag–AgCl surface electrodes placed 3 cm apart over the belly and tendon of the muscle. EMG activity was recorded with a bandpass between 10 and 1000 Hz and a display gain ranging from 50 to 200 µV/cm. EMG signals were collected, averaged and analysed off-line.
Focal TMS was applied over the hand motor cortex of the left hemisphere by using a figure-of-eight coil connected to two Magstim 200 stimulators through a Bistim module (Magstim Co., Whitland, UK).
The stimulating coil was placed over the optimal site for eliciting responses in the contralateral target muscle, with care to keep orientation and position constant. The handle of the coil pointed backward and the direction of the induced current was from posterior to anterior and was optimal to activate the motor cortex trans-synaptically (18).
The resting motor threshold (MT) for eliciting responses in the relaxed APB muscle was defined as the intensity of stimulation needed to produce responses of 50 µV in at least 50% of trials. Subjects were given audiovisual feedback of EMG activity to assist in maintaining complete relaxation.
Intracortical facilitation (ICF) and SICI of motor cortex were assessed by means of a paired-pulse paradigm with a subthreshold conditioning stimulus (CS) set to 80% of the MT followed by a testing stimulus (TS) at 120% of MT intensity. We did not re-evaluate MT after rTMS train because of: (i) its potential interference with the following intracortical excitability measures; (ii) evidence that MT is not affected by lf-rTMS on the motor cortex, applied either alone (19) or after priming stimulation (20).
Two different interstimulus intervals (ISIs) were used: 2 ms and 10 ms. These ISIs were chosen as they represent, respectively, the inhibitory and facilitatory limbs of the paired-pulse paradigm (21).
rTMS was delivered through a water-cooled figure-of-eight coil powered by a Cadwell High Speed Magnetic Stimulator (Cadwell Laboratories, Kennewick, WA, USA). Low-frequency repetitive stimulation was applied as a single 1-Hz 15 min long train (900 stimuli) delivered at 90% of MT over the hot spot of the right APB muscle. In the priming stimulation the 1-Hz train was preceded by 900 stimuli delivered at 10 Hz frequency, 90% MT intensity delivered in 18 trains (45 stimuli each) separated by 10-s intervals. For sham stimulation the coil was tilted 90° perpendicular to the scalp.
All subjects underwent three experimental sessions (1 Hz, real priming and sham priming rTMS) performed in separated days at least 1 week apart. Each session consisted of three conditions: TS alone and paired stimulation at ISIs of 2 and 10 ms explored before (baseline) and after rTMS. Conditions were applied 10 times each, intermixed in a pseudo-randomized order based on single trials. The intertrial interval was 10 s for all measurements. Constant coil position was continuously monitored during the experiment.
The peak-to-peak amplitude of motor evoked potentials (MEPs) was measured in the single trial and averages were calculated for each condition before and after rTMS in each subject. Approximately 3% of the recordings were discarded because the amplitude was too small (< 50 µV) or contaminated by voluntary muscle activity. MEPs elicited by the TS alone in all sessions before rTMS were averaged and used as baseline. The amplitude of MEPs recorded at 2 and 10 ms ISIs before and after the different rTMS sessions (1 Hz alone, real priming, sham priming) was averaged in each subject and expressed as a percentage of change from the respective mean TS alone within each session.
Statistical analysis
Comparisons of mean MT between patients and controls were made using the unpaired
t-test. Changes of unconditioned TS MEP and conditioned MEP
for SICI and ICF before and after rTMS conditions were examined through three
separate repeated measures analysis of variance (
Results
The TMS procedure throughout the whole experiment was generally well tolerated and no side-effects were reported.
Before rTMS no significant differences in mean MT (patients 50.3 ± 7.03% vs. controls 52.5 ± 7.6% of the maximum stimulator output) were found between subject groups.
Since baseline values of TS MEP, SICI and ICF showed no significant differences
across the tree rTMS conditions, the mean value of each measure was used to perform
each
TS MEP
Repeated measure Mean testing stimulus motor evoked potential amplitude values in the
different experimental conditions in patients and controls.
SICI
Mean values of short intracortical inhibition and intracortical
facilitation in the different experimental conditions in patients
and controls.
ICF
Discussion
The main results of our study was that 1 Hz rTMS increased motor intracortical facilitation in patients affected by MA; this effect was reversed delivering a high-frequency ‘priming’ train just before 1 Hz stimulation.
Since the first work by Chen et al. (22), 1 Hz rTMS has been known to decrease cortical excitability in the motor (23) or visual cortex (24) of normal subjects.
The suppressive effect of 1 Hz rTMS on cortical excitability has been used to induce a temporary cortical dysfunction in healthy subjects (25), or, presumably by normalizing increased levels of excitability, to induce beneficial effects in patients with neuropsychiatric disorders (26,27).
Some have speculated that the depressant effect is related to long-term depression (LTD) of cortical synapses (28). An important feature of in vivo LTD is the phenomenon of priming: brief pretreatment with stimulation in the 5–6 Hz range greatly increases the ability of subsequent 1 Hz stimulation to produce a decrease in synaptic efficiency (29).
This phenomenon, during which previous neuronal activity modulates the capacity for subsequent plastic change, has been termed ‘metaplasticity’ (29,30). More recently the effect of priming stimulation on the depressant effect of 1 Hz rTMS has also been demonstrated in the human motor cortex of healthy subjects (20,31). In line with these data, we found a significant increase in the amount of cortical depression (i.e. reduced TS MEP and ICF) both in migraineurs and in controls when 1 Hz rTMS was preceded by a hf-TMS train. Indeed, in migraineurs the effect was more evident on TS MEP amplitude, but importantly, the decrease of ICF, even if it did not reach statistical significance, showed a trend towards normal behaviour.
More recently it has been shown that the plastic changes induced by 1 Hz rTMS critically depend on the functional state of the stimulated cortex: preconditioning with cathodal transcranial direct current stimulation can flip the normal suppressive effect of 1 Hz rTMS and cause an apparently paradoxical facilitation (31).
Therefore in neuropsychiatric diseases, systematic changes in cortical excitability before rTMS are likely to alter the conditioning effects of 1 Hz rTMS. In migraine, the loss of balance between facilitatory and inhibitory circuits might invert the stimulation frequency-dependent effects of rTMS on cortical excitability. Whether this is due to decreased inhibition or to an abnormal cortical responsivity following a decreased preactivation level recently attributed to a thalamo-cortical dysrhythmia by Coppola et al. (32) remains to be established.
If the paradoxical increasing of ICF to 1 Hz rTMS was consequent upon the deficiency of intracortical inhibitory circuits in migraine, we suggest that the effect of priming stimulation could be due to a recovery of activity of these circuits by hf-train.
Baseline SICI was significantly lower in migraineurs than in controls, as previously reported (9), but significantly increased after real priming rTMS. In our results, this effect can be reasonably attributed to high-frequency stimulation and seems to be specific, as no changes in SICI were induced by 1 Hz, either when given alone or when preceded by sham stimulation.
It is known that SICI is a relatively complex measure in which the entity of the synaptic inhibition critically depends on the intensity of the first and second pulses.
In particular, evidence has been provided that the increased intensity the first stimulus may recruit facilitatory processes that superimpose with inhibition [short interval intracortical facilitation (SICF)] (33). This possibility, however, may hardly account for lower SICI values in our patients, because the risk of SICF contamination would have occurred also in controls.
Other evidence of up-regulation of inhibitory circuits by hf-rTMS in pathological conditions comes from studies in Parkinson's disease (34,35), where reduced cortical inhibition has been repeatedly reported (36).
Here, silent period duration (34,35,37) or SICI (38) are increased during (35) or after (34,37,38) hf-rTMS trains.
Indeed, in the present study we did not test the effects of hf-train alone for the following reasons: (i) to reduce the global number of magnetic pulses delivered to each subject; (ii) because the hf stimulation parameters were shown able to restore inhibition in previous works (10,37,38); and (iii) because the first aim of the present study was to investigate if normalization of cortical excitability state might invert the response to 1 Hz rTMS in the motor cortex of migraine patients.
In our findings, SICI did not significantly change after all rTMS conditions in control subjects. This to some degree paralleled results of previous reports (23,38,39).
In conclusion, our results show that hf-rTMS, probably by restoring cortical inhibitory circuits in migraine, normalizes the excitability changes in M1 following motor cortex 1 Hz rTMS.
If the reduced interictal intracortical inhibition is widely accepted so far, its role in migraine pathophysiology is a matter of debate (32,40). In pathological conditions with reduced intracortical inhibition, such as Parkinson's disease or chronic pain, normalization of cortical excitability by hf-rTMS has been associated with beneficial effects also on motor (35,37) or pain symptoms (13).
On this basis it would be worth treating patients with hf-rTMS for prophylaxis of migraine in order to normalize cortical excitability and consequently to avoid abnormal responsiveness to incoming (external or internal) stimuli. Further studies are needed to explore the therapeutic potential of hf-rTMS in migraine, defining also the best stimulation parameters and the duration of the long-term effects.