- Email: [email protected]
Reduced afferent-induced facilitation of primary motor cortex excitability in restless legs syndrome
Accepted Manuscript Title: Reduced afferent-induced facilitation of primary motor cortex excitability in restless legs syndrome Author: P. Bocquillon, C Charley-Monaca, E. Houdayer, A. Marques, A. Kwiatkowski, P. Derambure, H. Devanne PII: DOI: Reference:
S1389-9457(16)30003-X http://dx.doi.org/doi: 10.1016/j.sleep.2016.03.007 SLEEP 3033
To appear in:
Sleep Medicine
Received date: Revised date: Accepted date:
2-12-2015 3-2-2016 13-3-2016
Please cite this article as: P. Bocquillon, C Charley-Monaca, E. Houdayer, A. Marques, A. Kwiatkowski, P. Derambure, H. Devanne, Reduced afferent-induced facilitation of primary motor cortex excitability in restless legs syndrome, Sleep Medicine (2016), http://dx.doi.org/doi: 10.1016/j.sleep.2016.03.007. This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
Reduced afferent-induced facilitation of primary motor cortex excitability in restless legs syndrome P. Bocquillona,*, C Charley-Monacaa,b, E. Houdayera, A. Marquesa, A. Kwiatkowskia, P. Deramburea,b, H. Devannea,c
Comment [JM1]: Authors: please list all authors names in full
a
Centre Hospitalier Régional Universitaire de Lille, Neurophysiologie Clinique, Lille, France
b
Université de Lille, Troubles cognitifs vasculaires et dégénératifs, INSERM U1171, Lille,
Comment [JM2]: Affiliations only need to include department, institution, city and country
France c
Université du Littoral Côte d’Opale, Unité de Recherche Pluridisciplinaire Sport, Santé, Comment [JM3]: As before
Société (URePSSS, EA 7369), Calais, France
*Corresponding author: Centre Hospitalier Régional Universitaire de Lille, Hôpital Roger Salengro, Neurophysiologie Clinique, F-59037 Lille cedex, France. Tel.:
Comment [JM4]: Please include a phone number here (with country code)
E-mail address:
Comment [JM5]: Please provide an e-mail address here
(Perrine Bocquillon)
Highlights Restless legs syndrome (RLS) features both motor and sensory symptoms (primarily in the evening). The perturbation of sensorimotor integration in RLS does not follow a circadian rhythm. Patients with RLS show low levels of afferent-induced facilitation. RLS might result indirectly from a disorder of cortical sensorimotor integration.
1
Page 1 of 21
Comment [JM6]: Auhtors: this journal favours structured absrtacts for original articles, with the following subheadings: Objective, Patients/methods, Results and Conclusion
Abstract Restless legs syndrome (RLS) is characterized by the association of an urge to move, and
Comment [JM7]: Authors: please check this word, as I’m not sure that is is English
vesperal or nocturnal sensory symptoms; it is frequently associated with periodic limb movements. Evidence from imaging and electrophysiological studies suggests that RLS is linked to changes in sensorimotor integration. Nevertheless, the underlying mechanisms have not been characterized, and the cortical origin has yet to be confirmed. The objective of the present study was to establish whether or not sensorimotor integration in RLS patients is impaired in the evening. The time-dependent modulation of motor cortex excitability following peripheral electric nerve stimulation was studied in 14 idiopathic RLS patients, and 14 paired healthy controls. Different inter-stimulus intervals were used to measure short-latency and long-latency afferent inhibition (SAI and LAI) and afferent-induced facilitation (AIF). Motor evoked potentials were recorded from the first dorsal interosseous muscle in two experimental sessions (one in the morning and one in the evening). With the exception of LAI (which was present in the morning but absent in the evening in both healthy controls and RLS patients), no circadian variations were observed in sensorimotor integration. Although SAI was present in patients with RLS, AIF was disrupted (relative to controls) – suggesting the presence of an indirect sensorimotor integration disorder affecting the long corticocortical pathways in patients with RLS. The lack of circadian modulation in sensorimotor integration suggests that clinical circadian variations have other causes.
Keywords: Restless legs syndrome Circadian rhythm
2
Page 2 of 21
Transcranial magnetic stimulation Sensorimotor integration Afferent-induced facilitation
Idiopathic restless legs syndrome (RLS) is clinically defined as an urge to move the limbs, and an unpleasant sensation in the legs. These symptoms are most apparent when the muscles are at rest during the evening and at night. In most cases, the symptoms disappear when the person starts to move. Moreover, RLS is often associated with periodic limb movements when sleeping and/or awake. The close relationship between sensory and motor complaints suggests that RLS may involve dysfunctions in sensorimotor loops. However, results from early studies have suggested that the cortex is not impaired in RLS, since both somatosensory evoked potentials [1] and the Bereitschaftspotential prior to the periodic RLS-associated limb movements [2] were normal. However, several other studies have provided indirect evidence to suggest that cortical sensorimotor networks may be altered. Firstly, structural [3] and functional imaging studies [4] have thrown up evidence of cortical impairment in patients with RLS. Secondly, analyses of electrocortical rhythm reactivity have shown that patients with RLS display beta and/or mu event-related desynchronization before periodic limb movements during wakefulness [5,6] – suggesting the presence of dysfunctional sensorimotor integration processes in the cortex. Transcranial magnetic stimulation (TMS) studies have also revealed the presence of altered cortical excitability in patients with RLS. Although motor evoked potential (MEP) amplitude was not modified [7], a decrease in short-interval intracortical inhibition and silent period duration in patients with RLS has been reported by several investigators (for review, see Lanza et al. [8]). Transcranial magnetic stimulation can also be used to study sensorimotor integration by pairing electrical stimulation of a peripheral nerve with a TMS pulse over the
3
Page 3 of 21
primary motor cortex; this probes the time course and the extent of afferent input influence on the primary motor cortex. It is believed that only one study has explored sensorimotor integration in RLS by measuring short-latency and long-latency afferent inhibition (SAI and LAI, respectively) in the abductor pollicis brevis (APB) muscle in response to peripheral median nerve stimulation (MNS). The SAI was found to be less intense in untreated patients with RLS than in healthy controls (HCs), but normalized upon treatment [9]. Moreover, paired-associative stimulation protocols combining MNS and TMS with an inter-stimulus interval (ISI) of 25 ms failed to induce significant plastic changes in patients with RLS (in contrast to HCs); this suggested that dopaminergic transmission is impaired in RLS [10]. Comment [JM8]: As my previous comment
Given the vesperal and nocturnal predominance of RLS symptoms, the present study was designed to examine the circadian variations in sensorimotor integration in untreated patients with RLS. The objective was to determine whether or not circadian modulation of sensorimotor symptoms is correlated with alterations in sensorimotor integration. Hence, the influence of MNS on APB motor responses evoked by TMS at ISIs evoking SAI, LAI, and afferent-induced facilitation (AIF, corresponding to the net increase in motor excitability that can be observed 45-60 ms after a peripheral stimulus) was probed. In humans, AIF has been observed in the muscles of the lower limbs [11] and the upper limbs [12]. It is thought to reflect sensorimotor integration processes in long corticocortical pathways. To the best of the authors’ knowledge, AIF has never been studied in patients with RLS.
4
Page 4 of 21
Methods Subjects Fourteen patients with severe to very severe RLS, according to the International Restless Legs Syndrome Study Group (IRLSSG) severity score [13] (nine females, five males; age range 24-74; median [Q1; Q3] age 46.5 [30; 57.7]), were included in the study. During an in-depth patient interview, RLS was diagnosed by an experienced physician (CCM), according to the 2003 NIH/IRLSSG clinical criteria [13]. Restless legs syndrome ‘mimics’ and patients who reported alcohol abuse were excluded. The study recorded: each patient’s personal and family medical history, any previous and current use of prescription drugs, the characteristics of the RLS (including disease duration and the symptom time profile), and any history of sleep disorders (sleep apnea, sleepwalking, etc). A neurological examination and (if necessary) an electroneuromyographic assessment were also performed in order to exclude patients with a neurological disorder. Blood samples were also analyzed to screen for iron status and biochemical disorders. After a short interview and neurological examination, 14 gender-matched HCs (age range 23-75; median [Q1; Q3] age 45.5 [30.7; 53.5]) were also included in the study. The two groups were matched in terms of age, sex, and handedness. All participants completed the Epworth Sleepiness Scale (ESS) [14]. All gave informed, written consent to participate. The local investigational review board (Comité Consultatif de Protection des Personnes NordOuest IV, Lille, France), in accordance with the Declaration of Helsinki, approved the study. In order to study possible circadian variations in sensorimotor integration, all examinations were performed twice: once in the morning (starting at between 09:20 and 11:00) and once in the evening (starting between 18:00 and 19:30). The order of the experiments was counterbalanced for each group: half the participants in each group started with the morning block, and the other half started with the evening block. Participants were 5
Page 5 of 21
asked to avoid caffeine consumption on the study days. Before the recording, patients were asked to state whether RLS symptoms were present or not.
Electromyographic recordings Subjects were comfortably seated with their forearms placed on an armrest in a semiprone position. Ag–AgCl surface electrodes were used to record the electromyographic (EMG) activity of APB, with an active electrode on the muscle belly and the reference electrode on the interphalangeal joint of the thumb. The electrodes were connected to isolated preamplifiers (Digitimer, Welwyn Garden City, UK). A ground electrode that was connected to the preamplifiers’ shared input was placed around the wrist. The raw signals were amplified (1000 times), filtered (10-1000 Hz), digitized at 2 kHz (1401 Micro MKII; Cambridge Electronic Design, Cambridge, UK), and captured on a computer for further off-line analysis with customized Signal® software (Cambridge Electronic Design). Skin impedances were kept under 10 kΩ.
Transcranial magnetic stimulation Magnetic stimuli were delivered using a 9.5 cm external diameter figure-of-eight focal coil connected to two Magstim 200 stimulators via a BiStim module (The Magstim Company Ltd, Whitland, UK). The coil was positioned over the primary motor area of the hemisphere, contralateral to the most symptomatic side (for patients) or over the matched hemisphere in the HCs (since each patient was matched with an HC). Fine adjustments of the coil position were made at the beginning of the experiments, in order to identify the optimal locations (hotspots) for evoking MEPs in the APB. Once defined, this position was maintained during the whole recording, with the help of markers on a swimming cap worn by the subject. The coil was held tangentially to the scalp with the handle pointing backwards and sideways (at a 6
Page 6 of 21
45° angle from the nasion-inion line). Once the hotspot had been determined, the resting motor threshold (RMT, defined as the lowest stimulus intensity that could evoke five MEPs greater than 50 µV out of ten consecutive stimulations) and the active motor threshold (AMT, defined as the lowest stimulus level at which five MEPs of at least 200 µV were evoked by ten consecutive stimuli, during a weak active contraction of the APB controlled by auditory feedback played to the subjects through a loudspeaker) were measured. The mean peak-topeak MEP amplitude evoked by eight consecutive stimuli delivered with an intensity of 1.2xRMT was also measured.
Motor evoked potential conditioning with median nerve stimulation The influence of MNS on MEP amplitude was studied at rest. Square-wave single pulses of electrical stimulation (duration 0.2 ms) were delivered through standard bipolar electrodes to the median nerve at the wrist, using a constant current stimulator (DS7A; Digitimer Ltd). The stimulus intensity was set to the lowest value able to evoke a slight thumb twitch in the APB. Seven ISIs were explored: 21 and 23 ms (for studying the SAI), 50 and 55 ms (for studying the AIF), and 100, 200, and 300 ms (for studying the LAI). Within the series, eight paired stimuli were delivered for each ISI in a pseudorandom order. Eight unconditioned test stimuli were also delivered pseudorandomly throughout the whole block. The intensity of the test magnetic stimulus was set to 1.2xRMT. At a given ISI, the eight conditioned responses were averaged and measured offline from peak to peak. The mean conditioned MEP amplitude was expressed as a percentage of the mean unconditioned MEP, and will be expressed below as the meanstandard error (SE).
7
Page 7 of 21
Statistical analysis Statistical analysis was performed with SPSS v.22 (IBM Corp., Armonk, NY). Inter-group comparisons of demographic features were performed with Mann-Whitney tests. Motor threshold and test MEP amplitudes were compared using a repeated-measures analysis of variance (RM-ANOVA), with group as the between-subjects factor, and time as the repeated measure (two levels). In order to improve the protocol’s statistical power, data from 21-23 ms, 50-55 ms, and 100-200-300 ms ISIs were pooled and defined as ‘SAI’, ‘AIF’, and ‘LAI’ data, respectively. A three-factor RM-ANOVA was used to compare conditioned MEP amplitudes, with group (patients with RLS vs controls) as a between-subjects factor, and time (morning or evening) and ISI (SAI, AIF or LAI) as within-subjects factors. A Mauchly’s sphericity test was used to assess the homogeneity of covariance between levels of the repeated-measures factor. When the sphericity condition was not met, Greenhouse-Geisser correction was applied. If required, post-hoc analyses were performed with Mann-Whitney tests. If the effect of ISI was significant, a one-sample t-test was used to assess significant differences between conditioned and unconditioned MEP amplitudes for each ISI pool (SAI, AIF, and LAI). The threshold for statistical significance was set to p<0.05 for all tests.
Results Clinical and demographic characteristics None of the participants reported previous or ongoing psychiatric, cognitive, or neurological disorders (other than RLS in the patient group), and none were taking psychoactive drugs, including alcohol. One patient had already taken a dopaminergic agonist (for 2 months, starting 18 months before the recording). Two patients had taken clonazepam in the past, one had taken pregabalin for 1 year, and one had taken amitriptyline for 1 week. As shown in
8
Page 8 of 21
Table 1, the two groups did not differ significantly in terms of age (Z= –1.05, p=0.91) or ESS score (Z= –1.5, p=0.134). In RLS patients, the median [Q1; Q3] IRLSSG severity score was 29.5 [24.2; 30.7]. The median disease duration was 2.5 years [1; 6.75]. Four patients reported headaches, nine reported nocturia, and five reported snoring. Nevertheless, none of the patients reported interrupted breathing during sleep or had been diagnosed with sleep apnea syndrome. One patient reported sleepwalking in childhood. The results of the neurological examination were perfectly normal for all but two patients (who displayed reduced osteotendinous reflexes, a slight reduction in pallesthesia but normal electroneuromyographic data). Blood assays revealed low ferritin levels in two patients; these patients were not excluded from the study because their symptoms persisted for several months after the study (despite iron supplementation), which was suggestive of idiopathic RLS associated with (but not resulting from) iron depletion. No other biochemical disorders were found. During the evening recording session, 12 of the 14 patients reported RLS symptoms.
Transcranial magnetic stimulation features Mean±SE motor threshold values are given in Table 1. The ANOVA performed on the RMT data did not reveal any main effect of group (F(1,26)=0.407, p=0.529) or time (F(1,26)=0.043, p=0.837) or interaction (F(1,26)=0.22, p=0.643). Likewise, there were significant main effects of group (F(1,24)=0.206, p=0.554), time (F(1,24)=2.827, p=0.106) or interaction (F(1,24)=0.756, p=0.393) on the AMT data. The ANOVA performed on the test MEP amplitude did not reveal a main effect of group (F(1,26)=2.684, p=0.113), time (F (1,26)=3.745, p=0.064) or interaction (F(1,26)=0.562, p=0.46).
9
Page 9 of 21
Modulation of the MEP amplitude by peripheral nerve stimulation When the median nerve was stimulated before the test pulse delivered over the primary motor cortex, the MEP obtained was modulated in a time-dependent manner. The ANOVA showed a main effect of ISI (F(2,52)=20.66, p<0.001) and group (F(2,52)=7.186, p=0.013) and showed a group x ISI interaction (F(2,52)=4.01, p=0.026). The post-hoc analysis revealed a significant difference between patients with RLS and HCs at the AIF ISI (Z= –3.589, p<0.001). In contrast, no difference was found for SAI (Z= –1.065, p=0.287) and LAI (Z= –1.432, p=0.145) ISIs. Unexpectedly, no main effect of time (F(1,26)=0.492, p=0.489) or interaction between time and group or between time and ISI was found (F(1,26)=0.59, p=0.45, F(2,52)=2.36, p=0.123, respectively). As shown in Fig. 1 and Fig. 2, significant inhibition at the SAI ISI was observed in all recording sessions, whether for patients with RLS (morning 51.9%±8.5, t13= –4.062, p=0.001; evening 56.3%±13.3, t13= –3.198, p=0.007) or HCs (morning 60.5%±9.7, t13= –6.098, p<0.001; evening 59.6%±6.6, t13= –5.655, p<0.001). In contrast, significant inhibition was observed at the LAI ISI in the morning only in the patient group (54.4%±9.1, t13= –4.998, p<0.001) (ie, no modulation of the conditioned MEPs was found when applying the LAI ISI in the evening (90.3%±15.7, t13= –0.616, p=0.548). Similarly, LAI was observed in the morning session (79.8%±8.3, t13= –2.426, p=0.031) but not in the evening session (84.5%±11.0, t13= –1.407, p=0.183) in the HCs. Lastly, only HCs displayed AIF (141.2%±18.0, t13=2.289, p=0.039 and 135.6%±13.9, t13=2.568, p=0.023 for the morning and evening session, respectively). Afferent-induced facilitation was never observed in the patients with RLS – regardless of the session.
10
Page 10 of 21
Discussion The present results evidenced the impairment of sensorimotor integration (more specifically, decreased AIF) of the upper limbs in patients with RLS – even though most patients reported clinical symptoms in the lower limbs. This finding suggests the presence of diffuse cortical sensorimotor integration dysfunction, and alterations in other cortical circuits (as attested to by the shorter silent period [15,16,17,18], and shorter latency of intracortical inhibition [19] in patients with RLS, relative to HCs). The study deliberately chose to record the upper limbs because: (1) MEPs are usually more intense and easier to obtain in upper limb muscles, and (2) SAI in the upper limbs is reportedly disrupted in patients with RLS [9]. Rather unexpectedly, the sensorimotor integration processes were not modulated by a circadian rhythm: SAI was present in the morning and evening sessions, and AIF was not observed in patients with RLS at either time of day. Although the possibility that the evening recording session was performed too early in the evening for the RLS symptoms to be pronounced cannot be ruled out, the literature data suggest that the circadian appearance of RLS symptoms does not result from circadian variation of cortical dysfunction within sensorimotor networks. Firstly, the silent period duration and the AMT (both measures reflecting the excitability of the primary motor cortex) are the same at 08:00 as they are at 20:00 [16]. Moreover, a positron emission tomography (PET) scan study (using [11C] raclopride and [11C] FLB 457 as radioligands of D2 dopamine receptors) has shown that hypoactive dopaminergic transmission was present (at similar levels) in both the morning and evening in the striatal and non-striatal regions most likely to be involved in sensorimotor integration. In line with this hypothesis, the appearance of sensory symptoms and the disruption of somatosensory integration disruption may be two aspects of the set of symptoms in RLS [20]. Nevertheless, this hypothesis needs to be confirmed by recording patients later at night.
11
Page 11 of 21
The most striking finding in the present study was the lack of AIF in the patients with RLS. Afferent-induced facilitation occurs in the primary motor cortex when large diameter afferent fibers are stimulated with a delay between 45-65 ms for both the distal upper limb muscles [12] and ankle muscles [11]. Several arguments suggest that AIF probably involves cortical circuits in general, and the long corticocortical loops in particular. Firstly, AIF is absent when the corticospinal tract is stimulated at the brainstem level [11] or by transcranial electric stimulation [12] that directly activates the cortical motoneurons. Secondly, the spinal motoneuron excitability of the tibialis anterior muscle (as measured by the H-reflex) is not modulated by stimulation of the tibial nerve afferents at ISIs of around 50 ms, whereas cortical excitability is increased at this ISI [11]. Secondly, during the AIF-related period of increased corticospinal excitability, MNS also induces a decrease in short-latency intracortical inhibition and an increase in intracortical facilitation – indicating the presence of an interaction between AIF and M1 intrinsic interneurons [11,12]. Thus, in accordance with the recent observation that functional activity in the sensorimotor regions is abnormally low in patients with RLS [21], the data clearly indicate that cortical sensorimotor integration is altered in RLS. Duration of 50-55 ms is indeed long enough to allow sensitive peripheral inputs to reach the somatosensory cortex, and indirectly modulate the motor cortex via secondary sensorial areas such as the posterior parietal cortex (PPC). Although this hypothesis is speculative, it is supported by two recent findings. First, inactivating the PPC with a burst of high-frequency TMS pulses disrupts the normal activity of the contralateral PPC, and decreases the threshold for electrical stimulation of the median nerve at the wrist [22]. Second, patients with RLS show a change in the amplitude of resting state spontaneous brain activity in various brain regions – including the left posterior parietal area; this change is also observable during the asymptomatic periods [21]. The lack of AIF in patients with RLS might therefore be due to impairment in the long corticocortical pathways, which strongly supports
12
Page 12 of 21
the concept of RLS as a network disorder (as suggested by Trenkwalder and Paulus [23]). However, further investigations are needed to accurately determine the pathways and neurotransmitters involved in AIF. It will also be important to consider sensorimotor disruption in the lower limbs, since AIF is known to occur in the ankle muscles [11]. It is also noteworthy that no abnormally low levels of afferent inhibition in patients with RLS were observed, even though it was expected to see decreased inhibition of cortical excitability (given the positive clinical symptoms). Hence, the report by Rizzo et al. [9], who showed that SAI was abnormally low in untreated patients with RLS but was normalized by dopaminergic treatment, was not replicated. Short-latency afferent inhibition is observed in the first few milliseconds after the N20-evoked potential; by assuming that the afferent conduction time from the median nerve at the wrist to the sensorimotor cortex is 20 ms, most researchers use an ISI of 20-22 ms [24]. One could therefore suppose that the SAI had already started to fall at the ISI of 25 ms chosen by Rizzo et al.; this may explain why the SAI was weak in their patient population [9]. One other possible explanation for the discrepancy between the study of Rizzo et al. and the present work relates to the statistical method employed. Rizzo et al. applied a one-factor ANOVA for each parameter, whereas in the present study, repeated-measures ANOVA was chosen to simultaneously evaluate the circadian effect, the ISI, and group effects. This approach reduced the statistical power, which then may not have been high enough to detect small but significant differences in SAI between groups, ISIs or recording sessions. These controversial results should now be investigated in multicenter studies of larger patient populations.
Conclusion and perspectives
13
Page 13 of 21
In conclusion, abnormally low levels of AIF were observed in patients with RLS; this suggests the presence of impaired sensorimotor integration in the long corticocortical loops and the absence of circadian variation. It is tempting to link the lack of AIF in RLS (suggesting impaired somatosensory integration) and the sensory symptoms. However, this association has not been experimentally confirmed. Better knowledge of the intrinsic mechanisms underlying AIF would probably help to understand the cause of the impairment in sensorimotor integration; the latter might appear after (or in parallel with) dysfunction in dopaminergic transmission in striatal and non-striatal brain regions. This could be investigated in multicenter studies of larger patient populations and better control of confounding factors; these would enable patients with even mild iron depletion or poor– quality sleep (with a sleep recording and completion of a sleep diary before the TMS session, for example). In fact, it has been shown that sleep fragmentation might influence the motor threshold or the ESS [25]. Nevertheless, it is believed that there are no literature data concerning the influence of sleep quality on sensorimotor integration. In contrast to Scalise et Comment [JM9]: Authors: should this not be in the discussion?
al. [25], an intergroup difference in the motor threshold was not observed. This observation further suggests that the AIF impairment was disease-specific and did not result from sleep fragmentation.
Acknowledgments The authors thank Dr David Fraser (Biotech Communication, Damery, France) for helpful comments on the manuscript’s English. Financial disclosure: This study was funded by a grant from the Association France Ekbom charity and GlaxoSmithKline. The funders had no role in the study design, data collection and analysis, decision to publish or preparation of the manuscript.
14
Page 14 of 21
References 1. Mosko SS, Nudleman KL. Somatosensory and brainstem auditory evoked responses in sleep-related periodic leg movements. Sleep 1986;9:399-404. 2. Trenkwalder C, Bucher SF, Oertel WH, Proeckl D, Plendl H, et al. Bereitschaftspotential in idiopathic and symptomatic restless legs syndrome. Electroencephalogr Clin Neurophysiol 1993;89:95-103. 3. Unrath A, Juengling FD, Schork M, Kassubek J. Cortical grey matter alterations in idiopathic restless legs syndrome: An optimized voxel-based morphometry study. Mov Disord 2007;22:1751-6. 4. Ruottinen HM, Partinen M, Hublin C, Bergman J, Haaparanta M, et al. An FDOPA PET study in patients with periodic limb movement disorder and restless legs syndrome. Neurology 2000;54:502-4. 5. Rau C, Hummel F, Gerloff C. Cortical involvement in the generation of "involuntary" movements in restless legs syndrome. Neurology 2004;62:998-1000. 6. Tyvaert L, Houdayer E, Devanne H, Bourriez JL, Derambure P, et al. Cortical involvement in the sensory and motor symptoms of primary restless legs syndrome. Sleep Med 2009;10: 1090-6. 7. Tergau F, Wanschura V, Canelo M, Wischer S, Wassermann EM, et al. Complete suppression of voluntary motor drive during the silent period after transcranial magnetic stimulation. Exp Brain Res 1999;124:447-54.
15
Page 15 of 21
8. Lanza G, Cantone M, Lanuzza B, Pennisi M, Bella R, et al. Distinctive patterns of cortical excitability to transcranial magnetic stimulation in obstructive sleep apnea syndrome, restless legs syndrome, insomnia, and sleep deprivation. Sleep Med Rev 2014;19:39-50. 9. Rizzo V, Arico I, Liotta G, Ricciardi L, Mastroeni C, et al. Impairment of sensory-motor integration in patients affected by RLS. J Neurol 2010;257:1979-85. 10. Rizzo V, Arico I, Mastroeni C, Morgante F, Liotta G, et al. Dopamine agonists restore cortical plasticity in patients with idiopathic restless legs syndrome. Mov Disord 2009;24: 710-5. 11. Roy FD, Gorassini MA. Peripheral sensory activation of cortical circuits in the leg motor cortex of man. J Physiol 2008;586:4091-105. 12. Devanne H, Degardin A, Tyvaert L, Bocquillon P, Houdayer E, et al. Afferent-induced facilitation of primary motor cortex excitability in the region controlling hand muscles in humans. Eur J Neurosci 2009;30:439-48. 13. Allen RP, Picchietti D, Hening WA, Trenkwalder C, Walters AS, et al. Restless legs syndrome: diagnostic criteria, special considerations, and epidemiology. A report from the restless legs syndrome diagnosis and epidemiology workshop at the National Institutes of Health. Sleep Med 2003;4:101-19. 14. Johns MW. A new method for measuring daytime sleepiness: the Epworth sleepiness scale. Sleep 1991;14:540-5. 15. Kutukcu Y, Dogruer E, Yetkin S, Ozgen F, Vural O, et al. Evaluation of periodic leg movements and associated transcranial magnetic stimulation parameters in restless legs syndrome. Muscle Nerve 2006;33:133-7.
16
Page 16 of 21
16. Entezari-Taher M, Singleton JR, Jones CR, Meekins G, Petajan JH, et al. Changes in excitability of motor cortical circuitry in primary restless legs syndrome. Neurology 1999;53: 1201-5. 17. Stiasny-Kolster K, Haeske H, Tergau F, Muller HH, Braune HJ, et al. Cortical silent period is shortened in restless legs syndrome independently from circadian rhythm. Suppl Clin Neurophysiol 2003;56:381-9. 18. Ahlgren-Rimpilainen A, Lauerma H, Kahkonen S, Markkula J, Rimpilainen I. Recurrent CSPs after Transcranial Magnetic Stimulation of Motor Cortex in Restless Legs Syndrome. Neurol Res Int 2012:628949. 19. Quatrale R, Manconi M, Gastaldo E, Eleopra R, Tugnoli V, et al. Neurophysiological study of corticomotor pathways in restless legs syndrome. Clin Neurophysiol 2003;114:163845. 20. Cervenka S, Palhagen SE, Comley RA, Panagiotidis G, Cselenyi Z, et al. Support for dopaminergic hypoactivity in restless legs syndrome: a PET study on D2-receptor binding. Brain 2006;129:2017-28. 21. Liu C, Dai Z, Zhang R, Zhang M, Hou Y, et al. Mapping intrinsic functional brain changes and repetitive transcranial magnetic stimulation neuromodulation in idiopathic restless legs syndrome: a resting-state functional magnetic resonance imaging study. Sleep Med 2015;16:785-91. 22. Blankenburg F, Ruff CC, Bestmann S, Bjoertomt O, Eshel N, et al. Interhemispheric effect of parietal TMS on somatosensory response confirmed directly with concurrent TMSfMRI. J Neurosci 2008;28:13202-8.
17
Page 17 of 21
23. Trenkwalder C, Paulus W. Restless legs syndrome: pathophysiology, clinical presentation and management. Nat Rev Neurol;6:337-46. 24. Cash RF, Isayama R, Gunraj CA, Ni Z, Chen R. The influence of sensory afferent input on local motor cortical excitatory circuitry in humans. J Physiol 2015;593:1667-84. 25. Scalise A, Pittaro-Cadore I, Serafini A, Simeoni S, Fratticci L, et al. Transcranial magnetic stimulation in sleep fragmentation: a model to better understand sleep disorders. Sleep Med;15:1386-91.
Comment [JM10]: Is there a year for references 23 and 25? Comment [JM11]: In references 2, 4,, 6, 7, 12, 13, 15, 25 please list the first 6 authors before using et al
18
Page 18 of 21
Fig. 1. Modulation of the motor evoked potential amplitude by median nerve stimulation for each interstimulus interval: representative examples of individual raw electromyogram waveform datasets for a healthy subject (on the left) and a restless legs syndrome patient.
19
Page 19 of 21
Fig. 2. Modulation of the motor evoked potential amplitude by median nerve stimulation in each group. Values are the mean±SE; *p<0.001
20
Page 20 of 21
Tables Table 1. Demographic and transcranial magnetic stimulation features in healthy controls and patients with restless legs syndrome. Demographic characteristic Age ESS score TMS feature RMT (%) AMT (%) MEP (mV)
Healthy controls
Patients with RLS
45.5 [30.7; 53.56] 12.6 [5.2; 14] morning evening
46.5 [30; 57.7] 6 [4.2; 8.7] morning evening
48.4 (2.3) 38 (1.3) 0.70 (0.2)
46.5 (2.6) 37.8 (1.9) 0.46 (0.1)
48.8 (1.9) 39.8 (1.4) 1.1 (0.3)
47.6 (2.3) 39.1 (1.9) 0.62 (0.1)
Demographic features are quoted as the median [Q1; Q3]. TMS features are quoted as the mean±SE.
AMT, active motor threshold; ESS, Epworth Sleepiness Scale; MEP, motor evoked potential; RLS, restless legs syndrome; RMT, resting motor threshold; TMS, transcranial magnetic stimulation
21
Page 21 of 21
S1389-9457(16)30003-X http://dx.doi.org/doi: 10.1016/j.sleep.2016.03.007 SLEEP 3033
To appear in:
Sleep Medicine
Received date: Revised date: Accepted date:
2-12-2015 3-2-2016 13-3-2016
Please cite this article as: P. Bocquillon, C Charley-Monaca, E. Houdayer, A. Marques, A. Kwiatkowski, P. Derambure, H. Devanne, Reduced afferent-induced facilitation of primary motor cortex excitability in restless legs syndrome, Sleep Medicine (2016), http://dx.doi.org/doi: 10.1016/j.sleep.2016.03.007. This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
Reduced afferent-induced facilitation of primary motor cortex excitability in restless legs syndrome P. Bocquillona,*, C Charley-Monacaa,b, E. Houdayera, A. Marquesa, A. Kwiatkowskia, P. Deramburea,b, H. Devannea,c
Comment [JM1]: Authors: please list all authors names in full
a
Centre Hospitalier Régional Universitaire de Lille, Neurophysiologie Clinique, Lille, France
b
Université de Lille, Troubles cognitifs vasculaires et dégénératifs, INSERM U1171, Lille,
Comment [JM2]: Affiliations only need to include department, institution, city and country
France c
Université du Littoral Côte d’Opale, Unité de Recherche Pluridisciplinaire Sport, Santé, Comment [JM3]: As before
Société (URePSSS, EA 7369), Calais, France
*Corresponding author: Centre Hospitalier Régional Universitaire de Lille, Hôpital Roger Salengro, Neurophysiologie Clinique, F-59037 Lille cedex, France. Tel.:
Comment [JM4]: Please include a phone number here (with country code)
E-mail address:
Comment [JM5]: Please provide an e-mail address here
(Perrine Bocquillon)
Highlights Restless legs syndrome (RLS) features both motor and sensory symptoms (primarily in the evening). The perturbation of sensorimotor integration in RLS does not follow a circadian rhythm. Patients with RLS show low levels of afferent-induced facilitation. RLS might result indirectly from a disorder of cortical sensorimotor integration.
1
Page 1 of 21
Comment [JM6]: Auhtors: this journal favours structured absrtacts for original articles, with the following subheadings: Objective, Patients/methods, Results and Conclusion
Abstract Restless legs syndrome (RLS) is characterized by the association of an urge to move, and
Comment [JM7]: Authors: please check this word, as I’m not sure that is is English
vesperal or nocturnal sensory symptoms; it is frequently associated with periodic limb movements. Evidence from imaging and electrophysiological studies suggests that RLS is linked to changes in sensorimotor integration. Nevertheless, the underlying mechanisms have not been characterized, and the cortical origin has yet to be confirmed. The objective of the present study was to establish whether or not sensorimotor integration in RLS patients is impaired in the evening. The time-dependent modulation of motor cortex excitability following peripheral electric nerve stimulation was studied in 14 idiopathic RLS patients, and 14 paired healthy controls. Different inter-stimulus intervals were used to measure short-latency and long-latency afferent inhibition (SAI and LAI) and afferent-induced facilitation (AIF). Motor evoked potentials were recorded from the first dorsal interosseous muscle in two experimental sessions (one in the morning and one in the evening). With the exception of LAI (which was present in the morning but absent in the evening in both healthy controls and RLS patients), no circadian variations were observed in sensorimotor integration. Although SAI was present in patients with RLS, AIF was disrupted (relative to controls) – suggesting the presence of an indirect sensorimotor integration disorder affecting the long corticocortical pathways in patients with RLS. The lack of circadian modulation in sensorimotor integration suggests that clinical circadian variations have other causes.
Keywords: Restless legs syndrome Circadian rhythm
2
Page 2 of 21
Transcranial magnetic stimulation Sensorimotor integration Afferent-induced facilitation
Idiopathic restless legs syndrome (RLS) is clinically defined as an urge to move the limbs, and an unpleasant sensation in the legs. These symptoms are most apparent when the muscles are at rest during the evening and at night. In most cases, the symptoms disappear when the person starts to move. Moreover, RLS is often associated with periodic limb movements when sleeping and/or awake. The close relationship between sensory and motor complaints suggests that RLS may involve dysfunctions in sensorimotor loops. However, results from early studies have suggested that the cortex is not impaired in RLS, since both somatosensory evoked potentials [1] and the Bereitschaftspotential prior to the periodic RLS-associated limb movements [2] were normal. However, several other studies have provided indirect evidence to suggest that cortical sensorimotor networks may be altered. Firstly, structural [3] and functional imaging studies [4] have thrown up evidence of cortical impairment in patients with RLS. Secondly, analyses of electrocortical rhythm reactivity have shown that patients with RLS display beta and/or mu event-related desynchronization before periodic limb movements during wakefulness [5,6] – suggesting the presence of dysfunctional sensorimotor integration processes in the cortex. Transcranial magnetic stimulation (TMS) studies have also revealed the presence of altered cortical excitability in patients with RLS. Although motor evoked potential (MEP) amplitude was not modified [7], a decrease in short-interval intracortical inhibition and silent period duration in patients with RLS has been reported by several investigators (for review, see Lanza et al. [8]). Transcranial magnetic stimulation can also be used to study sensorimotor integration by pairing electrical stimulation of a peripheral nerve with a TMS pulse over the
3
Page 3 of 21
primary motor cortex; this probes the time course and the extent of afferent input influence on the primary motor cortex. It is believed that only one study has explored sensorimotor integration in RLS by measuring short-latency and long-latency afferent inhibition (SAI and LAI, respectively) in the abductor pollicis brevis (APB) muscle in response to peripheral median nerve stimulation (MNS). The SAI was found to be less intense in untreated patients with RLS than in healthy controls (HCs), but normalized upon treatment [9]. Moreover, paired-associative stimulation protocols combining MNS and TMS with an inter-stimulus interval (ISI) of 25 ms failed to induce significant plastic changes in patients with RLS (in contrast to HCs); this suggested that dopaminergic transmission is impaired in RLS [10]. Comment [JM8]: As my previous comment
Given the vesperal and nocturnal predominance of RLS symptoms, the present study was designed to examine the circadian variations in sensorimotor integration in untreated patients with RLS. The objective was to determine whether or not circadian modulation of sensorimotor symptoms is correlated with alterations in sensorimotor integration. Hence, the influence of MNS on APB motor responses evoked by TMS at ISIs evoking SAI, LAI, and afferent-induced facilitation (AIF, corresponding to the net increase in motor excitability that can be observed 45-60 ms after a peripheral stimulus) was probed. In humans, AIF has been observed in the muscles of the lower limbs [11] and the upper limbs [12]. It is thought to reflect sensorimotor integration processes in long corticocortical pathways. To the best of the authors’ knowledge, AIF has never been studied in patients with RLS.
4
Page 4 of 21
Methods Subjects Fourteen patients with severe to very severe RLS, according to the International Restless Legs Syndrome Study Group (IRLSSG) severity score [13] (nine females, five males; age range 24-74; median [Q1; Q3] age 46.5 [30; 57.7]), were included in the study. During an in-depth patient interview, RLS was diagnosed by an experienced physician (CCM), according to the 2003 NIH/IRLSSG clinical criteria [13]. Restless legs syndrome ‘mimics’ and patients who reported alcohol abuse were excluded. The study recorded: each patient’s personal and family medical history, any previous and current use of prescription drugs, the characteristics of the RLS (including disease duration and the symptom time profile), and any history of sleep disorders (sleep apnea, sleepwalking, etc). A neurological examination and (if necessary) an electroneuromyographic assessment were also performed in order to exclude patients with a neurological disorder. Blood samples were also analyzed to screen for iron status and biochemical disorders. After a short interview and neurological examination, 14 gender-matched HCs (age range 23-75; median [Q1; Q3] age 45.5 [30.7; 53.5]) were also included in the study. The two groups were matched in terms of age, sex, and handedness. All participants completed the Epworth Sleepiness Scale (ESS) [14]. All gave informed, written consent to participate. The local investigational review board (Comité Consultatif de Protection des Personnes NordOuest IV, Lille, France), in accordance with the Declaration of Helsinki, approved the study. In order to study possible circadian variations in sensorimotor integration, all examinations were performed twice: once in the morning (starting at between 09:20 and 11:00) and once in the evening (starting between 18:00 and 19:30). The order of the experiments was counterbalanced for each group: half the participants in each group started with the morning block, and the other half started with the evening block. Participants were 5
Page 5 of 21
asked to avoid caffeine consumption on the study days. Before the recording, patients were asked to state whether RLS symptoms were present or not.
Electromyographic recordings Subjects were comfortably seated with their forearms placed on an armrest in a semiprone position. Ag–AgCl surface electrodes were used to record the electromyographic (EMG) activity of APB, with an active electrode on the muscle belly and the reference electrode on the interphalangeal joint of the thumb. The electrodes were connected to isolated preamplifiers (Digitimer, Welwyn Garden City, UK). A ground electrode that was connected to the preamplifiers’ shared input was placed around the wrist. The raw signals were amplified (1000 times), filtered (10-1000 Hz), digitized at 2 kHz (1401 Micro MKII; Cambridge Electronic Design, Cambridge, UK), and captured on a computer for further off-line analysis with customized Signal® software (Cambridge Electronic Design). Skin impedances were kept under 10 kΩ.
Transcranial magnetic stimulation Magnetic stimuli were delivered using a 9.5 cm external diameter figure-of-eight focal coil connected to two Magstim 200 stimulators via a BiStim module (The Magstim Company Ltd, Whitland, UK). The coil was positioned over the primary motor area of the hemisphere, contralateral to the most symptomatic side (for patients) or over the matched hemisphere in the HCs (since each patient was matched with an HC). Fine adjustments of the coil position were made at the beginning of the experiments, in order to identify the optimal locations (hotspots) for evoking MEPs in the APB. Once defined, this position was maintained during the whole recording, with the help of markers on a swimming cap worn by the subject. The coil was held tangentially to the scalp with the handle pointing backwards and sideways (at a 6
Page 6 of 21
45° angle from the nasion-inion line). Once the hotspot had been determined, the resting motor threshold (RMT, defined as the lowest stimulus intensity that could evoke five MEPs greater than 50 µV out of ten consecutive stimulations) and the active motor threshold (AMT, defined as the lowest stimulus level at which five MEPs of at least 200 µV were evoked by ten consecutive stimuli, during a weak active contraction of the APB controlled by auditory feedback played to the subjects through a loudspeaker) were measured. The mean peak-topeak MEP amplitude evoked by eight consecutive stimuli delivered with an intensity of 1.2xRMT was also measured.
Motor evoked potential conditioning with median nerve stimulation The influence of MNS on MEP amplitude was studied at rest. Square-wave single pulses of electrical stimulation (duration 0.2 ms) were delivered through standard bipolar electrodes to the median nerve at the wrist, using a constant current stimulator (DS7A; Digitimer Ltd). The stimulus intensity was set to the lowest value able to evoke a slight thumb twitch in the APB. Seven ISIs were explored: 21 and 23 ms (for studying the SAI), 50 and 55 ms (for studying the AIF), and 100, 200, and 300 ms (for studying the LAI). Within the series, eight paired stimuli were delivered for each ISI in a pseudorandom order. Eight unconditioned test stimuli were also delivered pseudorandomly throughout the whole block. The intensity of the test magnetic stimulus was set to 1.2xRMT. At a given ISI, the eight conditioned responses were averaged and measured offline from peak to peak. The mean conditioned MEP amplitude was expressed as a percentage of the mean unconditioned MEP, and will be expressed below as the meanstandard error (SE).
7
Page 7 of 21
Statistical analysis Statistical analysis was performed with SPSS v.22 (IBM Corp., Armonk, NY). Inter-group comparisons of demographic features were performed with Mann-Whitney tests. Motor threshold and test MEP amplitudes were compared using a repeated-measures analysis of variance (RM-ANOVA), with group as the between-subjects factor, and time as the repeated measure (two levels). In order to improve the protocol’s statistical power, data from 21-23 ms, 50-55 ms, and 100-200-300 ms ISIs were pooled and defined as ‘SAI’, ‘AIF’, and ‘LAI’ data, respectively. A three-factor RM-ANOVA was used to compare conditioned MEP amplitudes, with group (patients with RLS vs controls) as a between-subjects factor, and time (morning or evening) and ISI (SAI, AIF or LAI) as within-subjects factors. A Mauchly’s sphericity test was used to assess the homogeneity of covariance between levels of the repeated-measures factor. When the sphericity condition was not met, Greenhouse-Geisser correction was applied. If required, post-hoc analyses were performed with Mann-Whitney tests. If the effect of ISI was significant, a one-sample t-test was used to assess significant differences between conditioned and unconditioned MEP amplitudes for each ISI pool (SAI, AIF, and LAI). The threshold for statistical significance was set to p<0.05 for all tests.
Results Clinical and demographic characteristics None of the participants reported previous or ongoing psychiatric, cognitive, or neurological disorders (other than RLS in the patient group), and none were taking psychoactive drugs, including alcohol. One patient had already taken a dopaminergic agonist (for 2 months, starting 18 months before the recording). Two patients had taken clonazepam in the past, one had taken pregabalin for 1 year, and one had taken amitriptyline for 1 week. As shown in
8
Page 8 of 21
Table 1, the two groups did not differ significantly in terms of age (Z= –1.05, p=0.91) or ESS score (Z= –1.5, p=0.134). In RLS patients, the median [Q1; Q3] IRLSSG severity score was 29.5 [24.2; 30.7]. The median disease duration was 2.5 years [1; 6.75]. Four patients reported headaches, nine reported nocturia, and five reported snoring. Nevertheless, none of the patients reported interrupted breathing during sleep or had been diagnosed with sleep apnea syndrome. One patient reported sleepwalking in childhood. The results of the neurological examination were perfectly normal for all but two patients (who displayed reduced osteotendinous reflexes, a slight reduction in pallesthesia but normal electroneuromyographic data). Blood assays revealed low ferritin levels in two patients; these patients were not excluded from the study because their symptoms persisted for several months after the study (despite iron supplementation), which was suggestive of idiopathic RLS associated with (but not resulting from) iron depletion. No other biochemical disorders were found. During the evening recording session, 12 of the 14 patients reported RLS symptoms.
Transcranial magnetic stimulation features Mean±SE motor threshold values are given in Table 1. The ANOVA performed on the RMT data did not reveal any main effect of group (F(1,26)=0.407, p=0.529) or time (F(1,26)=0.043, p=0.837) or interaction (F(1,26)=0.22, p=0.643). Likewise, there were significant main effects of group (F(1,24)=0.206, p=0.554), time (F(1,24)=2.827, p=0.106) or interaction (F(1,24)=0.756, p=0.393) on the AMT data. The ANOVA performed on the test MEP amplitude did not reveal a main effect of group (F(1,26)=2.684, p=0.113), time (F (1,26)=3.745, p=0.064) or interaction (F(1,26)=0.562, p=0.46).
9
Page 9 of 21
Modulation of the MEP amplitude by peripheral nerve stimulation When the median nerve was stimulated before the test pulse delivered over the primary motor cortex, the MEP obtained was modulated in a time-dependent manner. The ANOVA showed a main effect of ISI (F(2,52)=20.66, p<0.001) and group (F(2,52)=7.186, p=0.013) and showed a group x ISI interaction (F(2,52)=4.01, p=0.026). The post-hoc analysis revealed a significant difference between patients with RLS and HCs at the AIF ISI (Z= –3.589, p<0.001). In contrast, no difference was found for SAI (Z= –1.065, p=0.287) and LAI (Z= –1.432, p=0.145) ISIs. Unexpectedly, no main effect of time (F(1,26)=0.492, p=0.489) or interaction between time and group or between time and ISI was found (F(1,26)=0.59, p=0.45, F(2,52)=2.36, p=0.123, respectively). As shown in Fig. 1 and Fig. 2, significant inhibition at the SAI ISI was observed in all recording sessions, whether for patients with RLS (morning 51.9%±8.5, t13= –4.062, p=0.001; evening 56.3%±13.3, t13= –3.198, p=0.007) or HCs (morning 60.5%±9.7, t13= –6.098, p<0.001; evening 59.6%±6.6, t13= –5.655, p<0.001). In contrast, significant inhibition was observed at the LAI ISI in the morning only in the patient group (54.4%±9.1, t13= –4.998, p<0.001) (ie, no modulation of the conditioned MEPs was found when applying the LAI ISI in the evening (90.3%±15.7, t13= –0.616, p=0.548). Similarly, LAI was observed in the morning session (79.8%±8.3, t13= –2.426, p=0.031) but not in the evening session (84.5%±11.0, t13= –1.407, p=0.183) in the HCs. Lastly, only HCs displayed AIF (141.2%±18.0, t13=2.289, p=0.039 and 135.6%±13.9, t13=2.568, p=0.023 for the morning and evening session, respectively). Afferent-induced facilitation was never observed in the patients with RLS – regardless of the session.
10
Page 10 of 21
Discussion The present results evidenced the impairment of sensorimotor integration (more specifically, decreased AIF) of the upper limbs in patients with RLS – even though most patients reported clinical symptoms in the lower limbs. This finding suggests the presence of diffuse cortical sensorimotor integration dysfunction, and alterations in other cortical circuits (as attested to by the shorter silent period [15,16,17,18], and shorter latency of intracortical inhibition [19] in patients with RLS, relative to HCs). The study deliberately chose to record the upper limbs because: (1) MEPs are usually more intense and easier to obtain in upper limb muscles, and (2) SAI in the upper limbs is reportedly disrupted in patients with RLS [9]. Rather unexpectedly, the sensorimotor integration processes were not modulated by a circadian rhythm: SAI was present in the morning and evening sessions, and AIF was not observed in patients with RLS at either time of day. Although the possibility that the evening recording session was performed too early in the evening for the RLS symptoms to be pronounced cannot be ruled out, the literature data suggest that the circadian appearance of RLS symptoms does not result from circadian variation of cortical dysfunction within sensorimotor networks. Firstly, the silent period duration and the AMT (both measures reflecting the excitability of the primary motor cortex) are the same at 08:00 as they are at 20:00 [16]. Moreover, a positron emission tomography (PET) scan study (using [11C] raclopride and [11C] FLB 457 as radioligands of D2 dopamine receptors) has shown that hypoactive dopaminergic transmission was present (at similar levels) in both the morning and evening in the striatal and non-striatal regions most likely to be involved in sensorimotor integration. In line with this hypothesis, the appearance of sensory symptoms and the disruption of somatosensory integration disruption may be two aspects of the set of symptoms in RLS [20]. Nevertheless, this hypothesis needs to be confirmed by recording patients later at night.
11
Page 11 of 21
The most striking finding in the present study was the lack of AIF in the patients with RLS. Afferent-induced facilitation occurs in the primary motor cortex when large diameter afferent fibers are stimulated with a delay between 45-65 ms for both the distal upper limb muscles [12] and ankle muscles [11]. Several arguments suggest that AIF probably involves cortical circuits in general, and the long corticocortical loops in particular. Firstly, AIF is absent when the corticospinal tract is stimulated at the brainstem level [11] or by transcranial electric stimulation [12] that directly activates the cortical motoneurons. Secondly, the spinal motoneuron excitability of the tibialis anterior muscle (as measured by the H-reflex) is not modulated by stimulation of the tibial nerve afferents at ISIs of around 50 ms, whereas cortical excitability is increased at this ISI [11]. Secondly, during the AIF-related period of increased corticospinal excitability, MNS also induces a decrease in short-latency intracortical inhibition and an increase in intracortical facilitation – indicating the presence of an interaction between AIF and M1 intrinsic interneurons [11,12]. Thus, in accordance with the recent observation that functional activity in the sensorimotor regions is abnormally low in patients with RLS [21], the data clearly indicate that cortical sensorimotor integration is altered in RLS. Duration of 50-55 ms is indeed long enough to allow sensitive peripheral inputs to reach the somatosensory cortex, and indirectly modulate the motor cortex via secondary sensorial areas such as the posterior parietal cortex (PPC). Although this hypothesis is speculative, it is supported by two recent findings. First, inactivating the PPC with a burst of high-frequency TMS pulses disrupts the normal activity of the contralateral PPC, and decreases the threshold for electrical stimulation of the median nerve at the wrist [22]. Second, patients with RLS show a change in the amplitude of resting state spontaneous brain activity in various brain regions – including the left posterior parietal area; this change is also observable during the asymptomatic periods [21]. The lack of AIF in patients with RLS might therefore be due to impairment in the long corticocortical pathways, which strongly supports
12
Page 12 of 21
the concept of RLS as a network disorder (as suggested by Trenkwalder and Paulus [23]). However, further investigations are needed to accurately determine the pathways and neurotransmitters involved in AIF. It will also be important to consider sensorimotor disruption in the lower limbs, since AIF is known to occur in the ankle muscles [11]. It is also noteworthy that no abnormally low levels of afferent inhibition in patients with RLS were observed, even though it was expected to see decreased inhibition of cortical excitability (given the positive clinical symptoms). Hence, the report by Rizzo et al. [9], who showed that SAI was abnormally low in untreated patients with RLS but was normalized by dopaminergic treatment, was not replicated. Short-latency afferent inhibition is observed in the first few milliseconds after the N20-evoked potential; by assuming that the afferent conduction time from the median nerve at the wrist to the sensorimotor cortex is 20 ms, most researchers use an ISI of 20-22 ms [24]. One could therefore suppose that the SAI had already started to fall at the ISI of 25 ms chosen by Rizzo et al.; this may explain why the SAI was weak in their patient population [9]. One other possible explanation for the discrepancy between the study of Rizzo et al. and the present work relates to the statistical method employed. Rizzo et al. applied a one-factor ANOVA for each parameter, whereas in the present study, repeated-measures ANOVA was chosen to simultaneously evaluate the circadian effect, the ISI, and group effects. This approach reduced the statistical power, which then may not have been high enough to detect small but significant differences in SAI between groups, ISIs or recording sessions. These controversial results should now be investigated in multicenter studies of larger patient populations.
Conclusion and perspectives
13
Page 13 of 21
In conclusion, abnormally low levels of AIF were observed in patients with RLS; this suggests the presence of impaired sensorimotor integration in the long corticocortical loops and the absence of circadian variation. It is tempting to link the lack of AIF in RLS (suggesting impaired somatosensory integration) and the sensory symptoms. However, this association has not been experimentally confirmed. Better knowledge of the intrinsic mechanisms underlying AIF would probably help to understand the cause of the impairment in sensorimotor integration; the latter might appear after (or in parallel with) dysfunction in dopaminergic transmission in striatal and non-striatal brain regions. This could be investigated in multicenter studies of larger patient populations and better control of confounding factors; these would enable patients with even mild iron depletion or poor– quality sleep (with a sleep recording and completion of a sleep diary before the TMS session, for example). In fact, it has been shown that sleep fragmentation might influence the motor threshold or the ESS [25]. Nevertheless, it is believed that there are no literature data concerning the influence of sleep quality on sensorimotor integration. In contrast to Scalise et Comment [JM9]: Authors: should this not be in the discussion?
al. [25], an intergroup difference in the motor threshold was not observed. This observation further suggests that the AIF impairment was disease-specific and did not result from sleep fragmentation.
Acknowledgments The authors thank Dr David Fraser (Biotech Communication, Damery, France) for helpful comments on the manuscript’s English. Financial disclosure: This study was funded by a grant from the Association France Ekbom charity and GlaxoSmithKline. The funders had no role in the study design, data collection and analysis, decision to publish or preparation of the manuscript.
14
Page 14 of 21
References 1. Mosko SS, Nudleman KL. Somatosensory and brainstem auditory evoked responses in sleep-related periodic leg movements. Sleep 1986;9:399-404. 2. Trenkwalder C, Bucher SF, Oertel WH, Proeckl D, Plendl H, et al. Bereitschaftspotential in idiopathic and symptomatic restless legs syndrome. Electroencephalogr Clin Neurophysiol 1993;89:95-103. 3. Unrath A, Juengling FD, Schork M, Kassubek J. Cortical grey matter alterations in idiopathic restless legs syndrome: An optimized voxel-based morphometry study. Mov Disord 2007;22:1751-6. 4. Ruottinen HM, Partinen M, Hublin C, Bergman J, Haaparanta M, et al. An FDOPA PET study in patients with periodic limb movement disorder and restless legs syndrome. Neurology 2000;54:502-4. 5. Rau C, Hummel F, Gerloff C. Cortical involvement in the generation of "involuntary" movements in restless legs syndrome. Neurology 2004;62:998-1000. 6. Tyvaert L, Houdayer E, Devanne H, Bourriez JL, Derambure P, et al. Cortical involvement in the sensory and motor symptoms of primary restless legs syndrome. Sleep Med 2009;10: 1090-6. 7. Tergau F, Wanschura V, Canelo M, Wischer S, Wassermann EM, et al. Complete suppression of voluntary motor drive during the silent period after transcranial magnetic stimulation. Exp Brain Res 1999;124:447-54.
15
Page 15 of 21
8. Lanza G, Cantone M, Lanuzza B, Pennisi M, Bella R, et al. Distinctive patterns of cortical excitability to transcranial magnetic stimulation in obstructive sleep apnea syndrome, restless legs syndrome, insomnia, and sleep deprivation. Sleep Med Rev 2014;19:39-50. 9. Rizzo V, Arico I, Liotta G, Ricciardi L, Mastroeni C, et al. Impairment of sensory-motor integration in patients affected by RLS. J Neurol 2010;257:1979-85. 10. Rizzo V, Arico I, Mastroeni C, Morgante F, Liotta G, et al. Dopamine agonists restore cortical plasticity in patients with idiopathic restless legs syndrome. Mov Disord 2009;24: 710-5. 11. Roy FD, Gorassini MA. Peripheral sensory activation of cortical circuits in the leg motor cortex of man. J Physiol 2008;586:4091-105. 12. Devanne H, Degardin A, Tyvaert L, Bocquillon P, Houdayer E, et al. Afferent-induced facilitation of primary motor cortex excitability in the region controlling hand muscles in humans. Eur J Neurosci 2009;30:439-48. 13. Allen RP, Picchietti D, Hening WA, Trenkwalder C, Walters AS, et al. Restless legs syndrome: diagnostic criteria, special considerations, and epidemiology. A report from the restless legs syndrome diagnosis and epidemiology workshop at the National Institutes of Health. Sleep Med 2003;4:101-19. 14. Johns MW. A new method for measuring daytime sleepiness: the Epworth sleepiness scale. Sleep 1991;14:540-5. 15. Kutukcu Y, Dogruer E, Yetkin S, Ozgen F, Vural O, et al. Evaluation of periodic leg movements and associated transcranial magnetic stimulation parameters in restless legs syndrome. Muscle Nerve 2006;33:133-7.
16
Page 16 of 21
16. Entezari-Taher M, Singleton JR, Jones CR, Meekins G, Petajan JH, et al. Changes in excitability of motor cortical circuitry in primary restless legs syndrome. Neurology 1999;53: 1201-5. 17. Stiasny-Kolster K, Haeske H, Tergau F, Muller HH, Braune HJ, et al. Cortical silent period is shortened in restless legs syndrome independently from circadian rhythm. Suppl Clin Neurophysiol 2003;56:381-9. 18. Ahlgren-Rimpilainen A, Lauerma H, Kahkonen S, Markkula J, Rimpilainen I. Recurrent CSPs after Transcranial Magnetic Stimulation of Motor Cortex in Restless Legs Syndrome. Neurol Res Int 2012:628949. 19. Quatrale R, Manconi M, Gastaldo E, Eleopra R, Tugnoli V, et al. Neurophysiological study of corticomotor pathways in restless legs syndrome. Clin Neurophysiol 2003;114:163845. 20. Cervenka S, Palhagen SE, Comley RA, Panagiotidis G, Cselenyi Z, et al. Support for dopaminergic hypoactivity in restless legs syndrome: a PET study on D2-receptor binding. Brain 2006;129:2017-28. 21. Liu C, Dai Z, Zhang R, Zhang M, Hou Y, et al. Mapping intrinsic functional brain changes and repetitive transcranial magnetic stimulation neuromodulation in idiopathic restless legs syndrome: a resting-state functional magnetic resonance imaging study. Sleep Med 2015;16:785-91. 22. Blankenburg F, Ruff CC, Bestmann S, Bjoertomt O, Eshel N, et al. Interhemispheric effect of parietal TMS on somatosensory response confirmed directly with concurrent TMSfMRI. J Neurosci 2008;28:13202-8.
17
Page 17 of 21
23. Trenkwalder C, Paulus W. Restless legs syndrome: pathophysiology, clinical presentation and management. Nat Rev Neurol;6:337-46. 24. Cash RF, Isayama R, Gunraj CA, Ni Z, Chen R. The influence of sensory afferent input on local motor cortical excitatory circuitry in humans. J Physiol 2015;593:1667-84. 25. Scalise A, Pittaro-Cadore I, Serafini A, Simeoni S, Fratticci L, et al. Transcranial magnetic stimulation in sleep fragmentation: a model to better understand sleep disorders. Sleep Med;15:1386-91.
Comment [JM10]: Is there a year for references 23 and 25? Comment [JM11]: In references 2, 4,, 6, 7, 12, 13, 15, 25 please list the first 6 authors before using et al
18
Page 18 of 21
Fig. 1. Modulation of the motor evoked potential amplitude by median nerve stimulation for each interstimulus interval: representative examples of individual raw electromyogram waveform datasets for a healthy subject (on the left) and a restless legs syndrome patient.
19
Page 19 of 21
Fig. 2. Modulation of the motor evoked potential amplitude by median nerve stimulation in each group. Values are the mean±SE; *p<0.001
20
Page 20 of 21
Tables Table 1. Demographic and transcranial magnetic stimulation features in healthy controls and patients with restless legs syndrome. Demographic characteristic Age ESS score TMS feature RMT (%) AMT (%) MEP (mV)
Healthy controls
Patients with RLS
45.5 [30.7; 53.56] 12.6 [5.2; 14] morning evening
46.5 [30; 57.7] 6 [4.2; 8.7] morning evening
48.4 (2.3) 38 (1.3) 0.70 (0.2)
46.5 (2.6) 37.8 (1.9) 0.46 (0.1)
48.8 (1.9) 39.8 (1.4) 1.1 (0.3)
47.6 (2.3) 39.1 (1.9) 0.62 (0.1)
Demographic features are quoted as the median [Q1; Q3]. TMS features are quoted as the mean±SE.
AMT, active motor threshold; ESS, Epworth Sleepiness Scale; MEP, motor evoked potential; RLS, restless legs syndrome; RMT, resting motor threshold; TMS, transcranial magnetic stimulation
21
Page 21 of 21