Inter-hemispheric inhibition is impaired in mirror dystonia

Subjects

Thirteen patients with FHD (age 44–73 years, mean 56.3 ± 2.2 years; 11 males) and 12 age-matched healthy subjects (age 39–69 years, mean 56.4 ± 2.1 years; 10 males) participated in the study. In the FHD group, seven patients showed mirror dystonia (FHD-D, age 44–73 years, mean 57.6 ± 3.4 years; all males), whereas six patients did not (FHD-ND, age 46–61 years, mean 55 ± 2.2 years; four males). The movement part of the Fahn Scale (see Appendix S1 in the Supporting information) was assessed for the affected movement (writing in nine patients and the affected finger movement in the four musicians: three pianists and one guitarist). There was no significant difference between groups with respect to the severity on the Fahn Scale (range 0–28; 6.7 ± 3 in the FHD-D group and 8.5 ± 2 in the FHD-ND group). Healthy volunteers did not have mirror movements. All subjects were right-handed according to the Edinburgh handedness inventory (Oldfield, 1971). Patients with FHD had only unilateral symptoms in their right, dominant hand. Participants had never been exposed to neuroleptic drugs and had no history of neuropsychiatric disorders, neurosurgery, or metal or electronic implants. Most of the patients had been treated with local injections of botulinum toxin type A in the affected muscles. The last injection had been given at least 3 months before the experiments (Table 1). All participants gave their informed consent prior to the experiments, which were approved by the Institutional Review Board (IRB) of the National Institute of Neurological Disorders and Stroke (NINDS), and were in accordance with the NINDS and Declaration of Helsinki Guidelines.

Recording

Subjects were seated in a comfortable chair with their right arm resting on a side table, which was individually adjusted. In some subjects, the wrist was supported by a towel to help the subject keep the hand muscles as relaxed as possible. Disposable surface silver–silver chloride electromyogram (EMG) electrodes were placed on the APB and the FDI bilaterally in a bipolar montage. Impedance was reduced below 5 kΩ. The EMG signal was amplified using a conventional EMG machine (Viking IV, Nicolet Biomedical, Madison, WI, USA) and bandpass filtered (20–2000 Hz). The signal was digitized at a frequency of 4 kHz and fed into a computer for off-line analysis. Individual motor-evoked potential (MEP) amplitudes were measured in four phases (rest, premotor, phasic and tonic; see motor task). Background EMG was calculated by assessing the root mean square over 50 ms prior to MEP onset in the same four phases.

Motor task

With their right hand lying flat on a table beside them, subjects pushed down on a small force transducer (Strain Measurement Devices, Meriden, CT, USA; model S215 load cell) with the tip of their index finger in response to an acoustic signal. This led to flexion in the meta-carpo-phalangeal joint of the index finger. In preliminary tests, even healthy volunteers were unable to completely suppress EMG activity in thumb muscles when the FDI was abducted (primary movement), as the thumb is then the only finger to oppose the movement under this condition to stabilize the hand. To minimize concomitant EMG activity in APB, index finger flexion was used. FDI participates as a synergist rather than as a prime mover in this motion, but it has been shown that modulation of cortical excitability is similar in synergists and agonists (Sohn & Hallett, 2004a).

In the task, subjects generated 10% of their maximum force (F_max) during isolated flexion of the index finger as a reaction time task after the onset of an acoustic signal. The force level was individually adjusted and displayed as a line on an oscilloscope in front of them. The output of the force transducer was also displayed on the oscilloscope as feedback. The acoustic signal was present for 2 s and subjects maintained contraction for the duration of the tone. Subjects practiced the task at the beginning of the experiment to attain a consistent motor performance. The task used has previously been described in detail (Beck et al., 2008).

In four different phases of the movement, IHI was assessed in four different tests for each muscle (FDI and APB): rest (100 ms before the onset of the tone); premotor (50 ms before the onset of the EMG in FDI, movement initiation); phasic (the first peak of EMG in FDI, movement initiation); and tonic (1600 ms after onset of the acoustic signal, maintenance phase).

TMS

For TMS, two high-power Magstim 200 machines (Magstim, Whitland, Dyfed, UK) were connected to two custom-made figure-of-eight coils with an outer diameter of 8 cm and an inner loop diameter of 3.5 cm. The experiment consisted of two parts. In each part, one of the two target muscles, FDI or APB, was tested. The order of the two parts was randomized.

At the beginning of each part (for FDI and APB separately), the ‘motor hot spot’ for eliciting MEPs in FDI or APB, respectively, was determined over the right and left M1 by moving the coil over the M1 area using a slightly supra-threshold stimulation intensity. Coil orientation was tangential to the scalp with the handle pointing backwards and laterally at a 45° angle away from the midline. The motor hot spot was on average 3 cm lateral and 1 cm posterior to Cz over the hand knob of M1. This position was marked on a tight-fitting cap to ensure proper coil placement throughout the four experiments (IHI at rest, during premotor, phasic and tonic phase) performed for each target muscle. In each part, the resting motor threshold (MT) was determined (for FDI or APB), to the nearest 1% of maximal stimulator output. MT was defined as the minimal stimulus intensity required to evoke MEPs of at least 50 μV in five out of 10 consecutive trials. MEP size was determined by averaging peak-to-peak amplitudes. Trials with a background EMG of more than 0.02 mV in APB (assessed as root mean square) over 50 ms before MEP onset were rejected.

For IHI, a 10-ms interstimulus interval was used, which was previously shown to be most effective (Ferbert et al., 1992). The four different phases (rest, premotor, phasic and tonic) were tested in four separate experiments for each muscle (FDI or APB). The test pulse was applied to the motor hotspot of FDI or APB, respectively, over the dystonic or dominant hemisphere. Both muscles were chosen as a target muscle in separate sessions and assessed in all four phases of the movement. In the beginning of each experiment, test pulse intensity was adjusted to evoke an MEP of 1.5 mV. IHI was first performed at rest. The intensity of the conditioning stimulus was adjusted to reduce the size of the test pulse to approximately 60% of the MEP induced by the test pulse (= 900 μV), resulting in an IHI of 40%. This adjustment was performed to avoid a floor or ceiling effect of IHI regulation. The same intensity of conditioning pulse was used for all phases.

Statistics

Outcome measures for EMG (root mean square), MT (percentage of stimulator output; see Table 2) and MEP (peak-to-peak amplitudes) were analysed using a two-way repeated-measures analysis of variance (anova) to compare the effect of ‘phase’ as a within-subject factor (four levels: rest, premotor, phasic and tonic) and ‘group’ as a between-subject variable (two levels: FHD and controls). MEP data were not always distributed normally. Therefore, Conover’s free distribution method, a non-parametric anova based on ranks (Conover & Iman, 1982) was used to calculate a two-way non-parametric anova in order to compare the four levels of the within-subject factor PHASE (rest; premotor, phasic and tonic) and two levels of the between-subject factor GROUP (FHD and CON). If significant in the two-way analysis, mean differences were calculated between the levels using a one-way design and confidence intervals were given after Bonferroni correction for repeated comparisons.

As a secondary analysis, the patients with FHD were split into two groups (with and without mirror dystonia), and a similar analysis as described above was performed for MEP (peak-to-peak amplitudes) using a two-way repeated-measures anova. The effect of ‘phase’ as a within-subject factor (four levels: rest, premotor, phasic and tonic) and ‘group’ as a between-subject variable (three levels: FHD-D, FHD-ND and controls) was compared. Again, MEP data were not always distributed normally. Therefore, Conover’s free distribution method, a non-parametric anova based on ranks (Conover & Iman, 1982) was used to calculate a two-way non-parametric anova in order to compare the four levels of the within-subject factor PHASE (rest, premotor, phasic and tonic) and three levels of the between-subject factor GROUP (FHD-D, FHD-ND and CON). If significant in the two-way analysis, mean differences were calculated between the levels using a one-way design and confidence intervals were given after Bonferroni correction for repeated comparisons.

Data are presented as means and standard error of means. P-values less than 0.05 are considered significant. For analysis, spss 16.0.1 was used.

IHI in the agonist muscle (FDI)

IHI was calculated, and is shown as the percentage of the conditioned MEP (MEP_cond) with reference to the test MEP (MEP_test) [IHI = MEP_cond/MEP_test × 100 (%)]. The target size of MEP size was 60% for the rest condition. IHI at rest was 73 ± 7% in the control group, 63 ± 7% in FHD-D and 62 ± 10% in FHD-ND.

Comparing FDI IHI in the three active levels to rest, there was a trend for a significant main effect for PHASE (F_3,69 = 2.3, P = 0.08), without any subsequent significant contrasts (all P > 0.1), and no effect for GROUP (F_1,23 = 0.06, P = 0.8) or for the GROUP-by-PHASE interaction (F_3,69 = 1.6, P = 0.2). When FDI IHI was compared among all three groups (controls, FHD-D and FHD-ND), the analysis showed again a trend for a main effect of PHASE (F_3,66 = 2.2, P = 0.09), indicating that IHI in the premotor phase was reduced compared with rest (contrast between rest and premotor F_1,22 = 4.1, P = 0.05), while there were no other significant main effects, interactions or contrasts (all P > 0.1; Fig. 1). Thus, the comparison of IHI in FDI showed no main effect for GROUP (F_2,22 = 0.1, P = 0.9), or GROUP-by-PHASE interaction (F_6,66 = 0.9, P = 0.5; Fig. 1).

The analysis of the FDI IHI test MEP size did not reveal any significant main effects or interactions (all P > 0.1; rest: CON 1.8 ± 0.2 mV, FHD-D 1.7 ± 0.1 mV, FHD-ND 1.5 ± 0.2 mV; premotor: CON 1.8 ± 0.2 mV, FHD-D 1.6 ± 0.3 mV, FHD-ND 1.5 ± 0.2 mV; phasic: CON 1.7 ± 0.1 mV, FHD-D 1.5 ± 0.4 mV, FHD-ND 1.4 ± 0.1 mV; tonic: CON 1.5 ± 0.1 mV, FHD-D 1.2 ± 0.1 mV, FHD-ND 1.4 ± 0.2 mV).

IHI in the surrounding muscle (APB)

The target size of MEP size was 60% for the rest condition. In fact, the mean of the induced IHI in APB was 65 ± 4% for CON, 63 ± 8% for FHD-D and 54 ± 11% for FHD-ND.

Comparing APB IHI in all patients with FHD with controls, there was a significant main effect for PHASE (F_3,69 = 24.6, P < 0.001), but not for GROUP (F_1,23 = 0.025, P = 0.9). The GROUP-by-PHASE interaction showed a trend for significance, indicating that IHI may be reduced in the premotor phase in the patients with FHD (F_3,69 = 2, P = 0.09; see Fig. 2).

When APB IHI was then compared among controls, FHD-ND and FHD-D during the four different phases, there was no significant main effect for GROUP (F_1,22 = 1.4, P = 0.27). However, there was a highly significant main effect for PHASE (F_3,66 = 23.8, P < 0.001) and for the GROUP-by-PHASE interaction (F_6,66 = 2.5, P = 0.03; Fig. 3). Contrasts between phases for the GROUP-by-PHASE interaction showed that IHI was modulated differently in the three groups during the premotor phase (F_1,22 = 4.2, P = 0.03; Fig. 3), in that APB IHI was abolished in the FHD-D group, as revealed by pairwise comparisons of means from each phase (rest vs. premotor: −62%; 95% CI ±−91%, −34%; P = 0.001; phasic vs. premotor: −27%; 95% CI ±−55%, 1%; P = 0.06; all other comparisons not significant; see 1, 2). During the phasic phase, IHI was abolished without a difference between groups (contrast between rest and phasic F_1,22 = 72.7, P < 0.001; means for CON 105 ± 6%; FHD-D 90 ± 8%; FHD-ND 103 ± 8%; see Fig. 2). During the tonic phase, IHI was reduced as compared with rest without a difference between groups (contrast between rest and tonic F_1,22 = 40, P < 0.001; means for CON 86 ± 6%; FHD-D 89 ± 8%; FHD-ND 81 ± 9%; see Fig. 2).

The analysis of the APB IHI test MEP size did not reveal any significant main effects or interactions (all P > 0.1; rest CON 1.5 ± 0.1 mV, FHD-D 1.7 ± 0.2 mV, FHD-ND 1.6 ± 0.1 mV; premotor CON 1.3 ± 0.1 mV, FHD-D 1.2 ± 0.2 mV, FHD-ND 1.6 ± 0.1 mV; phasic CON 1.4 ± 0.1 mV, FHD-D 1.5 ± 0.1 mV, FHD-ND 1.6 ± 0.2 mV; tonic CON 1.2 ± 0.1 mV, FHD-D 1.4 ± 0.1 mV, FHD-ND 1.5 ± 0.2 mV).

Background EMG

Background EMG was not different between groups as indicated by a non-significant main effect for GROUP, as well as the GROUP-by-PHASE interaction and the GROUP-by-PHASE-by-MUSCLE interaction (all P > 0.1; Table 3). The main effect for PHASE was not significant. A significant main effect for MUSCLE (F = 7.36, P = 0.013) and the PHASE-by-MUSCLE interaction (F = 5.4, P = 0.012) reflected differential modulation between FDI and APB. While simple contrasts did not reveal differences between the three different phases and rest in APB, background EMG increased in FDI for the phasic phase (F = 7.1, P = 0.01) and showed a trend to be enhanced during the tonic phase (F = 3.4, P = 0.08).

Timing of inter-hemispheric interactions

For single-pulse TMS over the contralateral M1, inhibition of MEP can be observed starting from 80 to 100 ms before movement in the homologous muscle, which is more pronounced when directed toward the dominant M1 (Leocani et al., 2000; Liepert et al., 2001). This time interval before EMG onset seems to be relevant for the interaction between the two M1 (Duque et al., 2005). In that study, MEP size in the contralateral, relaxed homologous muscle was inhibited by mirroring its primary movement and was facilitated when the index finger was moved in the opposite direction.

In the current study, two different target muscles were assessed: a synergist muscle (FDI) and a surrounding muscle (APB). IHI was slightly reduced when testing FDI in the premotor phase compared with rest in all groups. Concerning the time interval, this finding is consistent with the previous reports in healthy volunteers (Duque et al., 2005).

For APB, IHI decreased from rest to the premotor phase and then completely disappeared during the phasic phase in the control group and patients without mirror dystonia. This finding for the timing is again consistent with previous reports in healthy volunteers using single-pulse TMS to assess the interaction between the two M1 areas (Duque et al., 2005). However, in patients with mirror dystonia, IHI changed from inhibition to facilitation in the premotor phase instead. During the tonic phase, IHI was restored without difference between groups.

These results reflect a difference in the modulation of the premotor phase between two types of muscles (synergist vs. surrounding muscle) as well as between the two patient groups. In the current results, lack of IHI in patients with mirror dystonia occurred in the same (premotor) phase before EMG onset in which interactions between the two M1 areas have been described previously (Duque et al., 2005), suggesting that similar pathways were involved. While mirror dystonia is clinically observable throughout all active phases, it has already been shown that surround inhibition on the M1 level is only present during movement initiation and not during the maintenance of a contraction (Beck et al., 2008). Taken together with the reported interhemispheric interaction starting from 80 to 100 ms before EMG onset (Leocani et al., 2000; Liepert et al., 2001; Duque et al., 2005), the current results further support the notion that the crucial, deficient phase on the motor cortical level is movement selection during movement initiation. Dystonic overflow occurring in this phase may then secondarily activate subcortical circuits projecting to non-synergistic muscles.

Cortical (inter)neurons involved, inhibitory pathways

IHI is thought to be mediated via excitatory transcallosal projections, which then activate the local inhibitory interneurons in M1 (Ferbert et al., 1992). Therefore, three structures, the corpus callosum as well as the dystonic and the contralateral M1, are possible candidates to cause deficient IHI. Because lack of IHI was limited to one specific phase during the movement, it seems unlikely that the corpus callosum, as the connecting fiber tract, is deficient.

Previous studies have already demonstrated a functional impairment of the intra-cortical inhibitory networks in the affected M1 in FHD (Molloy et al., 2002; Stinear & Byblow, 2004; Beck et al., 2008). Other studies indicated bilateral abnormalities (Meunier et al., 2001; Merello et al., 2006; Tamura et al., 2008). The inhibitory interneurons in M1, which are activated by TMS, play an important role in shaping the output from M1. In healthy people, SICI is reduced in active muscles and enhanced in surrounding muscles (Sohn & Hallett, 2004a), may be increased volitionally (Liepert et al., 2001), and contributes to surround inhibition before and during movement onset (Stinear & Byblow, 2004; Beck et al., 2008). Thus, SICI as well as IHI are context-dependent and modulated in the phase preceding the movement.

While it is unclear if there is a difference in SICI between patients with and without mirror dystonia, the current study showed a time-specific lack of IHI in the FHD-D group only. This may indicate that the interaction between dystonic and contralateral M1 may not be generally abnormal in patients with FHD. Because there was no significant difference in symptom severity, the results suggest incomplete malfunction of local inhibitory interneurons in the dystonic M1 on which the transcallosal fibers project. Depending on the subtypes of affected interneurons, SICI or IHI would be impaired. In view of the previous studies, it is more likely that the pathology lies in the dystonic M1, although pathology in the contralateral M1 can not be completely excluded. However, IHI was not upregulated in the surrounding muscle, as one would expect, if it contributed to the generation of surround inhibition, putting into question its contribution to surround inhibition and thus also the possibility that IHI plays a causal role in the pathophysiology of FHD.

Cortical excitability in mirror dystonia

Lateralization of voluntary movements requires a mature motor system (Cincotta & Ziemann, 2008). In unilateral FHD, there are only limited electrophysiological data about mirror dystonia, while mirror movements in Parkinson’s disease (PD) are well described. Mirror movements in PD are more common on the less affected side (Vidal et al., 2003; Espay et al., 2005) and are thought to be due to impaired basal ganglia output to energize the cortical mechanisms that are relevant in movement preparation and execution (Berardelli et al., 2001; Cincotta & Ziemann, 2008). In contrast, mirror dystonia specifically describes the induction of a dystonic movement or posture by the task-specific activation of the homologous contralateral (non-affected) muscle and mostly occurs on the pathologic side, thereby revealing the affected limb (Sitburana & Jankovic, 2008). Similar to the pathophysiology in PD, this phenomenon may correspond with increased cortical excitability or deficient cortical inhibition in the dystonic M1 that is well-known in FHD (Ikoma et al., 1996; Hallett, 2004; Sohn & Hallett, 2004b; Stinear & Byblow, 2004; Butefisch et al., 2005). As shown in our study, IHI toward the dystonic M1 was specifically diminished in surrounding muscles in the patients with mirror dystonia, whereas it was preserved in patients with FHD without mirror dystonia and in the control group.

In conclusion, our findings indicate that mirror dystonia in FHD may emerge from a deficient restriction of excitatory input from the contralateral hemisphere. This impairment is most likely due to deficient intramotor-cortical inhibition in the dystonic hemisphere, but further studies are needed to better characterize which specific interneurons are involved. Another investigation that should be done is to evaluate IHI in the dystonic hand in the time period prior to movement of the non-dystonic hand. Because patients with FHD without mirror dystonia did not show the same time-specific loss of IHI, this pathway does not seem to play a major role in the general pathophysiology of FHD.