You are here

Article Text

Article menu
Movement disorders
Research paper
Long-term subthalamic nucleus stimulation improves sensorimotor integration and proprioception
  1. Aparna Wagle Shukla1,
  2. Elena Moro2,3,
  3. Carolyn Gunraj4,
  4. Andres Lozano5,
  5. Mojgan Hodaie5,
  6. Anthony Lang2,3,
  7. Robert Chen2,3,4
  1. 1Center for Movement Disorders and Neurorestoration, University of Florida, Gainesville, Florida, USA
  2. 2Edmond J. Safra Program, Parkinson's Disease and National Parkinson's Disease Centre of Excellence, Toronto, Ontario, Canada
  3. 3Division of Neurology, Department of Medicine, University Health Network, University of Toronto, Toronto, Ontario, Canada
  4. 4Division of Brain Imaging and Behavior Systems—Neuroscience, Toronto Western Research Institute, University Health Network, Toronto, Ontario, Canada
  5. 5Division of Neurosurgery, Department of Surgery, University Health Network, University of Toronto, Toronto, Ontario, Canada
  1. Correspondence to Dr Robert Chen, Toronto Western Hospital, 7MC411, 399 Bathurst Street, Toronto, Ontario, Canada M5T 2S8; robert.chen{at}uhn.ca

Abstract

Objective Sensorimotor integration is impaired in patients with Parkinson's disease (PD). Short latency afferent inhibition (SAI) and long latency afferent inhibition (LAI) measured with transcranial magnetic stimulation (TMS) can be used to measure sensorimotor integration. Subthalamic nucleus (STN) deep brain stimulation (DBS) has been found to restore these abnormalities, but the time course of these changes is not known. We prospectively evaluated the short-term and long-term effects of STN DBS on SAI, LAI and proprioception. We hypothesised plasticity changes induced by chronic stimulation are necessary to normalise sensorimotor integration and proprioception.

Methods Patients with PD were studied preoperatively, at 1 month and more than 6 months postoperatively. SAI was tested with median nerve stimulation to the wrist preceding TMS pulse to motor cortex by ∼20 ms and LAI by 200 ms. Proprioception (distance and spatial errors) in the arm was quantitatively assessed. For postoperative assessments, patients were studied in the medication-off/stimulator-off, medication-off/stimulator-on, medication-on/stimulator-off and medication-on/stimulator-on conditions.

Results 11 patients with PD and 10 controls were enrolled. Preoperatively, SAI and proprioception was abnormal during the medication-on conditions and LAI was reduced regardless of the medication status. STN DBS had no significant effect on SAI, LAI and proprioception at 1 month. However, at 6 months SAI, LAI and distance errors were normalised in the medication-on/stimulator-on condition. Spatial error was normalised with DBS on and off.

Conclusions Chronic STN DBS in PD normalises sensorimotor integration and proprioception, likely through long-term plastic changes in the basal ganglia thalamocortical circuit.

  • PARKINSON'S DISEASE
  • NEUROPHYSIOLOGY

https://doi.org/10.1136/jnnp-2012-304102

Statistics from Altmetric.com

    Request Permissions

    If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.

    Introduction

    Sensory perception and sensorimotor integration are known to be impaired in Parkinson's disease (PD).1 A positron emission tomography study found reduced activation of the basal ganglia and the cortex in response to vibration2 and a somatosensory evoked potentials (SEPs) study3 reported decreased amplitude for the N30 component. Sensorimotor integration can be studied by examining the effects of sensory input on the motor cortical output. Median nerve stimulation (MNS) applied at the wrist about 20 or 200 ms before transcranial magnetic stimulation (TMS) of the motor cortex inhibits the motor evoked potentials, referred to as short (∼20 ms, SAI) and long latency (∼200 ms, LAI) afferent inhibition.4 ,5 We previously reported reduced SAI in patients with PD on dopaminergic medications, and LAI was reduced in off and on medication states. Thus, reduced SAI may represent an adverse effect of dopaminergic medications and LAI is possibly related to non-dopaminergic manifestations of PD.5 Dopaminergic medications also worsen proprioception in PD.6–8

    Subthalamic nucleus (STN) deep brain stimulation (DBS) has been reported to alleviate motor and non-motor manifestations of PD, with sensory fluctuations being one of the most improved features.9 We previously reported reduced SAI and LAI that occurred when DBS was acutely turned off were normalised with the stimulators turned on.10 However, it was not known whether the effects of STN DBS on SAI and LAI required only acute or acute on chronic stimulation, and whether STN DBS improves proprioceptive abnormalities in PD. We therefore embarked on a prospective, longitudinal study to evaluate the effects of STN DBS on SAI, LAI and proprioception in PD. We examined subjects undergoing bilateral STN DBS preoperatively and postoperatively at about 1 and 6 months after starting DBS. We hypothesised that plasticity effects due to chronic STN DBS are required to normalise SAI, LAI and proprioceptive deficits in PD.

    Methods

    Subjects

    We studied 11 patients with advanced PD with bilateral STN DBS (nine men, aged 58±8.3 years; mean±SD, table 1) and 10 healthy age-matched controls (eight men, aged 56±7.1 years). The more affected side was studied except in four patients in whom severe tremor or dystonia in the medication-off state precluded satisfactory recordings and the less affected side was studied.

    View this table:
    Table 1

    Clinical information for the patients studied

    Study design

    Patients were studied on three separate occasions. The first study was performed 1–4 weeks before surgery in on and off medication states. DBS programming at our centre begins about a month after implantation of the programmable stimulator. The second study was conducted about 1 month after surgery to study the short-term effects of STN DBS and to minimise transient improvements related to a ‘subthalamotomy’ effect. Patients were studied in four conditions: medication off/stimulator off (MED-OFF/STIM-OFF), medication off/stimulator on (MED-OFF/STIM-ON), medication on/stimulator off (MED-ON/STIM-OFF) and medication on/stimulator on (MED-ON/STIM-ON). Both sides were simultaneously turned on and off except in patient 1, who could not tolerate both sides turned off, and therefore the side contralateral to the tested side remained on. The third study was performed more than 6 months after the surgery to examine the acute and chronic effects of STN DBS. For all three studies, the same side was examined and all patients for each study visit received the same (preoperative) levodopa equivalent dose. The DBS parameters used for ON stimulation testing remained unchanged for the second and third study.

    Experimental procedures

    At each study visit, the same protocol that required an entire day of recording was used. Subjects were studied in the morning at least 12 h off medications in the practically defined off state (MED-OFF).11 The N20 peak latency of median nerve SEP was established with both stimulators switched off. For the postoperative studies (figure 1), the STIM-ON (previously determined optimal parameter) and STIM-OFF conditions were studied in random order. Twenty minutes after a change in the stimulator condition, Unified Parkinson’s Disease Rating Scale (UPDRS) scores (items 18–31), quantitative testing of proprioception and TMS studies were performed. The experimenters and the patients were blinded to the stimulator condition. At the end of the MED-OFF sessions (about noon), the patients took their usual first morning dose of medications. At least 40 min after medication intake, when the subjects reported improvement in symptoms and documented by motor examination, the MED-ON recording began. UPDRS motor scoring, proprioceptive testing and TMS studies were performed. If severe dyskinesias were present, we waited until the dyskinesias improved before proceeding with the study.12 ,13 UPDRS assessments were performed by the same rater.

    Figure 1
    Figure 1

    Experimental procedure for the postoperative studies. Each postoperative study was conducted on the same day under four conditions. Patients arrived in medication off (MED-OFF) condition, stimulators were at first turned off to record somatosensory evoked potential (SEP), then the stimulation status was randomised to either OFF or ON. The patients and the transcranial magnetic stimulation (TMS) investigator were blinded to the stimulation status. TMS study, proprioception recordings and Unified Parkinson's Disease Rating Scale (UPDRS) scores were performed. Dopaminergic medications were administered after the MED-OFF recordings and in the afternoon MED-ON recordings with stimulation turned off and on were performed.

    Patient characteristics, experimental setup and data analysis are reported in the online supplementary material.

    Statistical analysis

    First, a paired t test was used to determine whether SAI and LAI were present in each group. In each condition, the motor-evoked potential amplitudes for the test stimulus preceded by MNS were compared with those of the test stimulus alone. For the preoperative study, the unpaired t test was used to assess the differences between controls versus MED-OFF or MED ON, and the paired t test was used to examine the differences between MED-OFF and MED-ON for SAI, LAI and proprioceptive tests. For the postoperative studies, MED-ON and MED-OFF conditions were analysed separately as MED-OFF always preceded MED-ON sessions. The unpaired t test was used to assess the differences between controls versus STIM-OFF or STIM-ON for SAI, LAI and proprioceptive tests. In addition, we determined the effects of time (1 vs 6 months) and stimulation status (ON vs OFF) using two-way repeated measures analysis of variance (ANOVA). If there was a significant main effect, Fisher's protected least significance difference post hoc testing was performed. Correlations between UPDRS, proprioceptive testing and SAI and LAI were determined using Spearman's correlation test. p≤0.05 was considered significant.

    Results

    SAI and LAI in control subjects

    We used the data for the dominant side of the control group since a previous study found no side-to-side differences for SAI and LAI.5 The control subjects demonstrated SAI and LAI based on paired t tests (figures 2, 3 and 4).

    Figure 2
    Figure 2

    Examples of afferent inhibition in a control subject and a patient with Parkinson's disease (PD) on medication. Motor evoked potentials (MEPs) (average of 10 trials) for the test stimulus (TS) alone and with preceding median nerve stimulation (MNS) for a control subject (A) and a patient with PD (patient 9, table 1) in the MED-ON/STIM-OFF (B) and the MED-ON/STIM-ON condition (C) during the second postoperative assessment are shown. For short latency afferent inhibition (ISI N20+3 ms), the MEPs for the control subject and for the patient with PD in the MED-ON/STIM-ON condition were inhibited compared with TS alone. However, the MED-ON/STIM-OFF condition did not show any inhibition compared with TS alone. For long latency afferent inhibition (ISI 200 ms), there was MEP inhibition in the patient with PD in both MED-ON conditions, but the inhibition was more prominent in the STIM-ON than in the STIM-OFF condition. MED-ON, medication on; STIM-OFF, stimulator off; STIM-ON, stimulator on.

    Figure 3
    Figure 3

    Short latency afferent inhibition (SAI) in controls and in patients with Parkinson's disease. The y axis shows SAI as ratios of the conditioned (test stimulus with preceding median nerve stimulation) to the unconditioned (test stimulus alone) motor evoked potential amplitude. Values <1 indicate inhibition. (A) Preoperative assessment, (B) first postoperative assessment and (C) second postoperative assessment. For the preoperative assessments, white bars represent controls, black bars patients in the MED-OFF condition and grey bars patients in the MED-ON condition. For the postoperative assessments, white bars represent controls, black bars patients in the MED-OFF/STIM-OFF condition (seven patients, in first postoperative and nine patients in second postoperative assessment), white hatched bars patients in the MED-OFF/STIM-ON condition (eight patients in first postoperative and nine patients in second postoperative assessment), grey bars patients in the MED-ON/STIM-OFF condition and grey hatched bars patients in the MED-ON/STIM-ON condition (nine patients during first and second postoperative MED-ON). Error bars represent SEs. Rings above the bars indicate significant inhibition compared with test pulse alone (paired t test). Asterisks between the bars indicate significant differences between the groups (p≤0.05). MED-ON, medication on; MED-OFF, medication off; STIM-OFF, stimulator off; STIM-ON, stimulator on.

    Figure 4
    Figure 4

    Long latency afferent inhibition (LAI) in controls and in patients with Parkinson's disease. The y axis shows LAI as ratios of the conditioned (test stimulus with preceding median nerve stimulation) to the unconditioned (test stimulus alone) motor evoked potential amplitude. Values <1 indicate inhibition. (A) Preoperative assessment, (B) first postoperative assessment and (C) second postoperative assessment. For the preoperative assessments, white bars represent controls, black bars patients in the MED-OFF condition and grey bars patients in the MED-ON condition. For the postoperative assessments, white bars represent controls, black bars patients in the MED-OFF/STIM-OFF condition (seven patients in first postoperative and nine patients in second postoperative assessment), white hatched bars patients in the MED-OFF/STIM-ON condition (eight patients in first postoperative and nine patients in second postoperative assessment), grey bars patients in the MED-ON/STIM-OFF condition and grey hatched bars patients in the MED-ON/STIM-ON condition (nine patients during first and second postoperative MED-ON). Error bars represent SEs. Rings above the bars indicate significant inhibition compared with test pulse alone (paired t test). Asterisks between the bars indicate significant differences between the groups (p≤0.05). MED-ON, medication on; MED-OFF, medication off; STIM-OFF, stimulator off; STIM-ON, stimulator on.

    SAI and the effects of STN DBS in PD

    Preoperative study

    Paired t tests revealed SAI in MED-OFF but not in MED-ON assessment (p=0.0002) (figure 3A). Further, t test comparisons showed significantly reduced SAI for MED-ON compared with controls (p=0.004) and MED-OFF (p=0.05).

    First postoperative study

    There was a significant SAI for MED-OFF/STIM-ON (p=0.0007), MED-ON/STIM-OFF (p=0.02) and MED-ON/STIM-ON (p=0.001), but not for the MED-OFF/STIM-OFF condition (figure 3B). A t test revealed significant differences between MED-ON/STIM-OFF and the controls (p=0.01). However, there were no differences between the MED-ON/STIM-ON and MED-ON/STIM-OFF, indicating that STN DBS did not improve SAI. There were no significant results seen for MED-OFF comparisons.

    Second postoperative study

    There was significant SAI in the MED-OFF/STIM-OFF (p=0.02), MED-OFF/STIM-ON (p=0.002) and MED-ON/STIM-ON conditions (p=0.001), but not in MED-ON/STIM-OFF condition (figure 3C). The results of the t test showed greater SAI in the MED-ON/STIM-ON compared with MED-ON/STIM-OFF (p=0.004) and a greater SAI in the control group than in the MED-ON/STIM-OFF (p=0.001) condition. There were no significant results seen for MED-OFF session comparisons, indicating there was no effect of STN DBS. However, for the MED-ON session, STN DBS restored the SAI that was reduced in the presence of dopaminergic medications (figure 3C).

    LAI and the effects of STN DBS in PD

    Preoperative study

    LAI was absent for MED-ON and MED-OFF conditions (figure 4A). The t test results showed a reduced LAI for the MED-OFF (p=0.001) and MED-ON (p=0.001) conditions compared with the controls.

    First postoperative study

    Paired t tests showed LAI in the MED-OFF/STIM-OFF (p=0.03), MED-OFF/STIM-ON (p=0.03) and MED-ON/STIM-ON conditions (p=0.022) but not in the MEDON/STIM-OFF condition (figure 4B). The t test comparisons were not significant for MED-OFF and MED-ON sessions.

    Second postoperative study

    LAI was present only in the MED-ON/STIM-ON condition (p=0.015) (figure 4C). Based on t-test comparisons, we found for the MED-OFF session, there was a greater LAI in the control group than in MED-OFF/STIM-OFF (p=0.04) and MED-OFF/STIM-ON (p=0.03) conditions. There was no significant difference between STIM-OFF and STIM-ON conditions. In the MED-ON session, there was a greater LAI in the control group compared with MED-ON/STIM-OFF (p=0.02) but there was no significant difference between control and MED-ON/STIM-ON conditions.

    Proprioceptive testing

    Preoperative study

    A t-test analysis showed a significantly higher distance error for MED-ON compared with controls (p=0.04) and MED-OFF (p=0.03) (figure 5A). Spatial error was significantly higher for MED-ON compared with MED-OFF (p=0.05) (figure 6A).

    Figure 5
    Figure 5

    Distance errors in controls and patients with Parkinson's disease. (A) Preoperative assessment, (B) first postoperative assessment and (C) second postoperative assessment. For the preoperative assessments, white bars represent controls, black bars patients in the MED-OFF condition and grey bars patients in the MED-ON condition. For the postoperative assessments, white bars represent controls, black bars patients in the MED-OFF/STIM-OFF condition (seven patients in first postoperative and nine patients in second postoperative assessment), white hatched bars patients in the MED-OFF/STIM-ON condition (eight patients in first postoperative and nine patients in second postoperative assessment), grey bars patients in the MED-ON/STIM-OFF condition and grey hatched bars patients in the MED-ON/STIM-ON condition (nine patients during first and second postoperative MED-ON). Error bars represent SEs. Rings above the bars indicate significant inhibition compared with test pulse alone (paired t test). Asterisks between the bars indicate significant differences between the groups (p≤0.05). MED-ON, medication on; MED-OFF, medication off; STIM-OFF, stimulator off; STIM-ON, stimulator on.

    Figure 6
    Figure 6

    Spatial errors in controls and patients with Parkinson's disease. (A) Preoperative assessment, (B) first postoperative assessment and (C) second postoperative assessment. For the preoperative assessments, white bars represent controls, black bars patients in the MED-OFF condition and grey bars patients in the MED-ON condition. For the postoperative assessments, white bars represent controls, black bars patients in the MED-OFF STIM-OFF condition (seven patients in first postoperative and nine patients in second postoperative assessment), white hatched bars patients in the MED-OFF STIM-ON condition (eight patients in first postoperative and nine patients in second postoperative assessment), grey bars patients in the MED-ON STIM-OFF condition and grey hatched bars patients in the MED-ON STIM-ON condition (nine patients during first and second postoperative MED-ON). Error bars represent SEs. Rings above the bars indicate significant inhibition compared with test pulse alone (paired t test). Asterisks between the bars indicate significant differences between the groups (p<0.05). MED-ON, medication on; MED-OFF, medication off; STIM-OFF, stimulator off; STIM-ON, stimulator on.

    First postoperative study

    In the MED-ON session, distance errors were significantly increased for MED-ON/STIM-OFF (p=0.02) and MED-ON/STIM-ON (p=0.02) compared with controls (figure 5B). There were no significant results for the MED-OFF comparisons. The MED-ON/STIM-OFF (p=0.03) and MED-ON/STIM-ON (p=0.008) conditions also had greater spatial errors compared with controls (figure 6B). However, there were no significant differences between MED-ON/STIM-OFF and MED-ON/STIM-ON for distance and spatial errors, indicating that DBS had no effect on spatial and distance errors. The distance and spatial errors in the MED-ON conditions were similar to those seen in the preoperative studies (figures 5A and 6A).

    Second postoperative study

    The MED-OFF session comparisons for spatial or distance errors were not significant. In the MED-ON session, the MED-ON/STIM-OFF condition had higher distance errors compared with controls (p=0.02) and the MED-ON/STIM-ON (p=0.04) condition (figure 5C), however there were no significant results for spatial errors (figure 6C) and the values for both MED-ON conditions (STIM-OFF and STIM-ON) were similar to controls. Therefore, long term, DBS normalised spatial errors even with DBS transiently turned off. There was no correlation between distance or spatial errors and SAI or LAI in any of the assessments.

    Effects of duration of DBS on SAI, LAI and proprioception

    We tested the effects of duration of DBS (factor time of 1 and 6 months) and stimulation status (STIM-OFF and STIM-ON) on SAI, LAI and proprioception findings using two-way repeated measures ANOVA. We analysed the MED-OFF and MED-ON sessions separately.

    For SAI (figure 3) in MED-OFF sessions, the effects of time, stimulation status and their interactions were not significant. For the MED-ON sessions, we found a significant effect of stimulation status (p=0.01, greater SAI with stimulator on), no significant effect of time but the interaction between time and stimulation status was significant (p=0.01), confirming that the effects of DBS on SAI were different at 1 and 6 months of DBS. Figure 3 shows that this was due to an increase in SAI with DBS-ON at 6 months but not at 1 month.

    For LAI (figure 4), there was no significant effect of time, stimulation status and their interactions in the MED-OFF state. In the MED-ON state, the effects of time and stimulation status on LAI were not significant but the interaction between time and stimulation status was significant (p=0.05), indicating that the DBS had different effects on LAI at 1 and 6 months. Figure 4 shows that this was due to increase in LAI with DBS-ON at 6 months but not at 1 month.

    For distance error (figure 5), in the MED-OFF state there was no significant effect of time or stimulation but the interaction between time and stimulation status was significant (p=0.006). In the MED-ON state, there were no significant effects of time, stimulation or their interaction.

    For spatial error (figure 6), in the MED-OFF state there were no significant effects of time, stimulation or their interaction. In the MED-ON state, there was a significant effect of time (p=0.02) with lower spatial errors at 6 months than at 1 month. There was no significant effect of stimulation status and the interaction between the time and stimulation status.

    Further details of the results are provided in the online supplementary material.

    Discussion

    The novel findings of this prospective, longitudinal study are that STN DBS restored SAI and LAI to normal levels and improved distance and spatial errors at 6 months but not at 1 month after surgery, despite using the same DBS parameters with similar degrees of motor improvement as measured by UPDRS. These findings underscore the importance of chronic stimulation in modulation of sensorimotor integration and proprioception. Plasticity induced by long-term STN DBS likely mediates these changes.

    Mechanisms underlying SAI and LAI

    SAI is due to cortical inhibition as it suppressed corticospinal waves induced by TMS14 and had no influence on F waves evoked through spinal mechanisms.4 LAI is likely mediated by mechanisms different from SAI and is thought to arise from more widespread activation. At longer latencies (>40 ms), there is widespread activation of the sensory cortex (S1) and the bilateral secondary sensory areas (S2) and the contralateral posterior parietal cortex.15 ,16 Our preoperative results that SAI was reduced in the presence of dopaminergic medications (MED-ON) and LAI was decreased in MED-OFF and MED-ON sessions are consistent with previous studies.5

    Effects of short-term and long-term STN DBS on SAI and LAI

    At 1 month after surgery, patients with PD had significant SAI in the MED-ON/STIM-OFF, MED-ON/STIM-ON conditions (figure 3B). This is different from the MED-ON condition in the preoperative study and from the MED-ON/STIM-OFF condition in the second postoperative study in which no SAI was observed. This may be related to a persistent subthalamotomy effect in some patients at 1 month after surgery.

    STN DBS had no significant effect on SAI whether in the MED-OFF or MED-ON conditions at 1 month. At 6-month follow-up, SAI was normal in MED-OFF/STIM-OFF and MED-OFF/STIM-ON conditions, was reduced in the MED-ON/STIM OFF condition and was restored to normal during the MED-ON/STIM-ON condition. STN DBS had no significant effect on LAI at 1 month regardless of medication status, but for the 6-month results, LAI was restored to normal during the MED-ON/STIM-ON condition. These findings for SAI and LAI during the second postoperative study are similar to our previous study.10 The presence of significant modulating effects of STN DBS on SAI and LAI at 6 months but not at 1 month after surgery indicates that DBS needs to be switched on and a background of chronic STN DBS is required for modulation of SAI and LAI.

    Effects of STN DBS on proprioception

    The higher distance and spatial errors in the preoperative MED-ON conditions compared with the MED-OFF conditions are similar to previous observations that dopaminergic medications worsen proprioception.7 A previous study found improved kinaesthesia sensation in patients with PD with STN DBS.17 We observed the effects of STN DBS on proprioception only at 6 months but not at 1 month after surgery. Distance error showed improvement in the second postoperative but not in the first postoperative MED-ON recordings (figure 4). Thus, the effects of medications and the time course of the effects of STN DBS on distance error are similar to those for SAI, and acute stimulation on a background of chronic stimulation is required for these effects.

    The increased spatial errors in the MED-ON session were not affected by DBS in the first postoperative session but spatial error was normal in the MED-ON session regardless of the stimulator being turned on or off at 6 months after surgery. These findings suggest that correction of spatial error requires chronic stimulation but the effects were retained even if the stimulator was acutely turned off. In a previous study, Conte et al18 found that increased somatosensory temporal discrimination thresholds in PD were corrected by dopaminergic medications but not by STN stimulation. Proprioception and temporal discrimination are different percepts mediated by different subcortical and cortical circuits and this may explain the different effects of STN DBS on these percepts.

    Limitations of the study

    The study had limitations. Previous studies had shown some patients with PD benefit from ipsilateral stimulation.19–21 In our study, all patients except one had stimulating electrodes switched off and on simultaneously. Therefore, to establish whether the effects that we observed were a result of contralateral or bilateral stimulation will require further study. However, the effects at 6 months after surgery were similar to the results of a previous study that used only contralateral stimulation,10 therefore it is likely that the current findings largely reflect the results of contralateral stimulation.

    The proprioception changes in the study did not correlate with SAI and LAI and could be related to several factors. The proprioceptive testing included integration of sensory information from multiple joints, especially proximal body parts, whereas the SAI and LAI are based purely on distal sensory input, namely median nerve innervation in the hand. Besides, spatial and distance error testing involved complex processing of sensory input from the arm and motor components which are required to position the arm in space.

    There were 11 patients enrolled, but not all patients could be studied and followed as planned. Two patients could not participate in off-medication assessments during the first postoperative study, due to intolerable off-medication symptoms. This is not unexpected because it is difficult to perform lengthy studies in patients with advanced PD in the MED-OFF and STIM-OFF conditions. Nevertheless, seven patients completed all 10 assessments and they provided clear results.

    None of the patients had prior DBS surgery. Two patients had prior unilateral pallidotomy but we studied the side that was opposite to the side that had pallidotomy. Two patients participated in a trial of intraputaminal injections of glial cell line derived neurotrophic factor (GDNF). The GDNF trial results had been published and were negative,22 and injections had been discontinued more than 6 months before the current study. We felt it was appropriate to include these patients.

    Although the patients and investigator were blinded to the stimulation status, the patients may have been aware whether DBS was ON or OFF because of its clinical effect. However, this was unlikely to affect the SAI and LAI because electrophysiological recordings do not require active patient participation. Even distance and spatial errors were not affected by the clinical improvement because the findings were worse in the MED-ON state preoperatively when the patients had fewer symptoms.

    Most patients on STN DBS had reduction in the doses of dopaminergic medications after surgery. Thus, the effects of chronic stimulation and reduction of medications occurred in parallel. Although we used the same levodopa equivalent dose of medications for all three visits and the same stimulation parameters for the two postoperative visits, the physiological effects of chronic stimulation or chronic medication reduction cannot be disentangled with certainty.

    How does STN DBS modulate afferent inhibition?

    STN neurons in humans respond to passive movement and electrical MNS.23 ,24 STN DBS may alter the response of the basal ganglia thalamocortical system to sensory input. For example, STN DBS may normalise aberrantly synchronous oscillatory patterns25 ,26 and restore abnormal processing of sensory signals in the basal ganglia. In addition, STN DBS may also modulate cortical inhibition and excitability in the sensory and motor cortices.13 ,27

    There was a stronger effect of STN DBS on LAI during MED-ON than MED-OFF conditions (figure 4C) and this suggests that there is synergism between STN DBS and dopaminergic medications. Several studies found synergistic effects of STN DBS and medications on non-dopaminergic features of PD, such as axial symptoms,28–30 postural control and gait.29 ,30 Since LAI represents a non-dopaminergic feature, it is not surprising a synergistic effect was seen for DBS and medications.

    Potential mechanism of plastic changes induced by STN DBS

    Our findings strongly suggest that normalisation of SAI, LAI, distance and spatial errors requires acute stimulation on a background of chronic stimulation. Long-term DBS and reduction of pulsatile dopaminergic stimulation resulted in reduction of the dyskinesia threshold over time.31 Since abnormalities in SAI5 and proprioception7 have been postulated to be related to dyskinesia, correction of SAI and proprioception may be related to improvement in dyskinesia with STN DBS.

    When DBS is acutely turned on, it evoked physiological changes in the cortex. STN DBS evoked cortical potentials and changes in cortical excitability through antidromic stimulation of the hyperdirect corticosubthalamic pathway.27 ,32 ,33 These effects were observed immediately after surgery.32 ,34 STN DBS also activated the cortex at longer latencies, likely via orthodromic transmission in the basal ganglia circuit through the internal globus pallidus and the thalamus.27 ,33 ,35 ,36 In addition, STN DBS restored deficient cortical inhibition.13 ,35 Since repeated stimulation is a well established way to induce long-term potentiation (LTP) and long-term depression plasticity in animals and humans, long-term DBS that changes the excitability in these pathways may lead to plastic changes in the sensory and motor cortices. LTP-like plasticity in the motor cortex is impaired in PD, particularly in patients with dyskinesias.37 In patients with dystonia treated with globus pallidus pars interna DBS, it has been suggested that changes in LTP-like plasticity in the motor cortex may drive the improvement in dystonia.38 ,39 Whether restoration of sensorimotor integration by long-term STN DBS is related to LTP-like plasticity mechanisms requires further study.

    References

    1. ↵
      1. Abbruzzese G,
      2. Berardelli A
      . Sensorimotor integration in movement disorders. Mov Disord 2003;18:231–40.
      OpenUrlCrossRefPubMedWeb of Science
    2. ↵
      1. Boecker H,
      2. Ceballos-Baumann A,
      3. Bartenstein P,
      4. et al
      . Sensory processing in Parkinson's and Huntington's disease: investigations with 3D H(2)(15)O-PET. Brain 1999;122(Pt 9):1651–65.
      OpenUrlAbstract/FREE Full Text
    3. ↵
      1. Ulivelli M,
      2. Rossi S,
      3. Pasqualetti P,
      4. et al
      . Time course of frontal somatosensory evoked potentials. relation to L-dopa plasma levels and motor performance in PD. Neurology 1999;53:1451–7.
      OpenUrlCrossRef
    4. ↵
      1. Chen R,
      2. Corwell B,
      3. Hallett M
      . Modulation of motor cortex excitability by median nerve and digit stimulation. Exp Brain Res 1999;129:77–86.
      OpenUrlCrossRefPubMedWeb of Science
    5. ↵
      1. Sailer A,
      2. Molnar GF,
      3. Paradiso G,
      4. et al
      . Short and long latency afferent inhibition in Parkinson's disease. Brain 2003;126:1883–94.
      OpenUrlAbstract/FREE Full Text
    6. ↵
      1. Bronte-Stewart HM,
      2. Minn AY,
      3. Rodrigues K,
      4. et al
      . Postural instability in idiopathic Parkinson's disease: the role of medication and unilateral pallidotomy. Brain 2002;125:2100–14.
      OpenUrlAbstract/FREE Full Text
    7. ↵
      1. O'Suilleabhain P,
      2. Bullard J,
      3. Dewey RB
      . Proprioception in Parkinson's disease is acutely depressed by dopaminergic medications. J Neurol Neurosurg Psychiatry 2001;71:607–10.
      OpenUrlAbstract/FREE Full Text
    8. ↵
      1. Rocchi L,
      2. Chiari L,
      3. Horak FB
      . Effects of deep brain stimulation and levodopa on postural sway in parkinson's disease. J Neurol Neurosurg Psychiatry 2002;73:267–74.
      OpenUrlAbstract/FREE Full Text
    9. ↵
      1. Witjas T,
      2. Kaphan E,
      3. Regis J,
      4. et al
      . Effects of chronic subthalamic stimulation on nonmotor fluctuations in parkinson's disease. Mov Disord 2007;22:1729–34.
      OpenUrlCrossRefPubMed
    10. ↵
      1. Sailer A,
      2. Cunic DI,
      3. Paradiso GO,
      4. et al
      . Subthalamic nucleus stimulation modulates afferent inhibition in Parkinson disease. Neurology 2007;68:356–63.
      OpenUrlCrossRef
    11. ↵
      1. Defer GL,
      2. Widner H,
      3. Marie RM,
      4. et al
      . Core assessment program for surgical interventional therapies in Parkinson's disease (CAPSIT-PD). Mov Disord 1999;14:572–84.
      OpenUrlCrossRefPubMedWeb of Science
    12. ↵
      1. Chen R,
      2. Garg RR,
      3. Lozano AM,
      4. et al
      . Effects of internal globus pallidus stimulation on motor cortex excitability. Neurology 2001;56:716–23.
      OpenUrlCrossRefPubMed
    13. ↵
      1. Cunic D,
      2. Roshan L,
      3. Khan FI,
      4. et al
      . Effects of subthalamic nucleus stimulation on motor cortex excitability in Parkinson's disease. Neurology 2002;58:1665–72.
      OpenUrlCrossRefPubMed
    14. ↵
      1. Tokimura H,
      2. Di Lazzaro V,
      3. Tokimura Y,
      4. et al
      . Short latency inhibition of human hand motor cortex by somatosensory input from the hand. J Physiol 2000;523(Pt 2):503–13.
      OpenUrlCrossRefPubMedWeb of Science
    15. ↵
      1. Allison T,
      2. McCarthy G,
      3. Wood CC,
      4. et al
      . Human cortical potentials evoked by stimulation of the median nerve. I. Cytoarchitectonic areas generating short-latency activity. J Neurophysiol 1989;62:694–710.
      OpenUrlAbstract/FREE Full Text
    16. ↵
      1. Forss N,
      2. Hari R,
      3. Salmelin R,
      4. et al
      . Activation of the human posterior parietal cortex by median nerve stimulation. Exp Brain Res 1994;99:309–15.
      OpenUrlPubMedWeb of Science
    17. ↵
      1. Maschke M,
      2. Tuite PJ,
      3. Pickett K,
      4. et al
      . The effect of subthalamic nucleus stimulation on kinaesthesia in Parkinson's disease. J Neurol Neurosurg Psychiatry 2005;76:569–71.
      OpenUrlAbstract/FREE Full Text
    18. ↵
      1. Conte A,
      2. Modugno N,
      3. Lena F,
      4. et al
      . Subthalamic nucleus stimulation and somatosensory temporal discrimination in Parkinson's disease. Brain 2010;133:2656–63.
      OpenUrlAbstract/FREE Full Text
    19. ↵
      1. Kumar R,
      2. Lozano AM,
      3. Sime E,
      4. et al
      . Comparative effects of unilateral and bilateral subthalamic nucleus deep brain stimulation. Neurology 1999; 53:561–6.
      OpenUrlCrossRef
    20. ↵
      1. Linazasoro G,
      2. Van Blercom N,
      3. Lasa A
      . Unilateral subthalamic deep brain stimulation in advanced Parkinson's disease. Mov Disord 2003;18:713–16.
      OpenUrlCrossRefPubMed
    21. ↵
      1. Castrioto A,
      2. Meaney C,
      3. Hamani C,
      4. et al
      . The dominant-STN phenomenon in bilateral STN DBS for Parkinson's disease. Neurobiol Dis 2011;41:131–7.
      OpenUrlCrossRefPubMed
    22. ↵
      1. Lang AE,
      2. Gill S,
      3. Patel NK,
      4. et al
      . Randomized controlled trial of intraputamenal glial cell line-derived neurotrophic factor infusion in Parkinson disease. Ann Neurol 2006;59:459–66.
      OpenUrlCrossRefPubMedWeb of Science
    23. ↵
      1. Pesenti A,
      2. Priori A,
      3. Locatelli M,
      4. et al
      . Subthalamic somatosensory evoked potentials in Parkinson's disease. Mov Disord 2003;18:1341–5.
      OpenUrlCrossRefPubMedWeb of Science
    24. ↵
      1. Theodosopoulos PV,
      2. Marks WJ Jr.,
      3. Christine C,
      4. et al
      . Locations of movement-related cells in the human subthalamic nucleus in Parkinson's disease. Mov Disord 2003;18:791–8.
      OpenUrlCrossRefPubMedWeb of Science
    25. ↵
      1. Rubin JE,
      2. Terman D
      . High frequency stimulation of the subthalamic nucleus eliminates pathological thalamic rhythmicity in a computational model. J Comput Neurosci 2004;16:211–35.
      OpenUrlCrossRefPubMedWeb of Science
    26. ↵
      1. Brown P,
      2. Oliviero A,
      3. Mazzone P,
      4. et al
      . Dopamine dependency of oscillations between subthalamic nucleus and pallidum in Parkinson's disease. J Neurosci 2001;21:1033–8.
      OpenUrlAbstract/FREE Full Text
    27. ↵
      1. Kuriakose R,
      2. Saha U,
      3. Castillo G,
      4. et al
      . The nature and time course of cortical activation following subthalamic stimulation in Parkinson's disease. Cereb Cortex 2010;20:1926–36.
      OpenUrlAbstract/FREE Full Text
    28. ↵
      1. Bejjani BP,
      2. Gervais D,
      3. Arnulf I,
      4. et al
      . Axial parkinsonian symptoms can be improved: the role of levodopa and bilateral subthalamic stimulation. J Neurol Neurosurg Psychiatry 2000;68:595–600.
      OpenUrlAbstract/FREE Full Text
    29. ↵
      1. Faist M,
      2. Xie J,
      3. Kurz D,
      4. et al
      . Effect of bilateral subthalamic nucleus stimulation on gait in Parkinson's disease. Brain 2001;124:1590–600.
      OpenUrlAbstract/FREE Full Text
    30. ↵
      1. Maurer C,
      2. Mergner T,
      3. Xie J,
      4. et al
      . Effect of chronic bilateral subthalamic nucleus (STN) stimulation on postural control in Parkinson's disease. Brain 2003;126:1146–63.
      OpenUrlAbstract/FREE Full Text
    31. ↵
      1. Moro E,
      2. Esselink RJ,
      3. Benabid AL,
      4. et al
      . Response to levodopa in parkinsonian patients with bilateral subthalamic nucleus stimulation. Brain 2002;125:2408–17.
      OpenUrlAbstract/FREE Full Text
    32. ↵
      1. Ashby P,
      2. Paradiso G,
      3. Saint-Cyr JA,
      4. et al
      . Potentials recorded at the scalp by stimulation near the human subthalamic nucleus. Clin Neurophysiol 2001;112:431–7.
      OpenUrlCrossRefPubMedWeb of Science
    33. ↵
      1. Baker KB,
      2. Montgomery EB Jr.,
      3. Rezai AR,
      4. et al
      . Subthalamic nucleus deep brain stimulus evoked potentials: physiological and therapeutic implications. Mov Disord 2002;17:969–83.
      OpenUrlCrossRefPubMedWeb of Science
    34. ↵
      1. Hanajima R,
      2. Ashby P,
      3. Lozano AM,
      4. et al
      . Single pulse stimulation of the human subthalamic nucleus facilitates the motor cortex at short intervals. J Neurophysiol 2004;92:1937–43.
      OpenUrlAbstract/FREE Full Text
    35. ↵
      1. Dauper J,
      2. Peschel T,
      3. Schrader C,
      4. et al
      . Effects of subthalamic nucleus (STN) stimulation on motor cortex excitability. Neurology 2002;59:700–6.
      OpenUrlCrossRef
    36. ↵
      1. MacKinnon CD,
      2. Webb RM,
      3. Silberstein P,
      4. et al
      . Stimulation through electrodes implanted near the subthalamic nucleus activates projections to motor areas of cerebral cortex in patients with Parkinson's disease. Eur J Neurosci 2005;21:1394–402.
      OpenUrlCrossRefPubMedWeb of Science
    37. ↵
      1. Morgante F,
      2. Espay AJ,
      3. Gunraj C,
      4. et al
      . Motor cortex plasticity in Parkinson's disease and levodopa-induced dyskinesias. Brain 2006;129:1059–69.
      OpenUrlAbstract/FREE Full Text
    38. ↵
      1. Ruge D,
      2. Cif L,
      3. Limousin P,
      4. et al
      . Shaping reversibility? Long-term deep brain stimulation in dystonia: the relationship between effects on electrophysiology and clinical symptoms. Brain 2011;134:2106–15.
      OpenUrlAbstract/FREE Full Text
    39. ↵
      1. Ruge D,
      2. Tisch S,
      3. Hariz MI,
      4. et al
      . Deep brain stimulation effects in dystonia: time course of electrophysiological changes in early treatment. Mov Disord 2011;26:1913–21.
      OpenUrlCrossRefPubMed

    Supplementary materials

    • Supplementary Data

      This web only file has been produced by the BMJ Publishing Group from an electronic file supplied by the author(s) and has not been edited for content.

      Files in this Data Supplement:

    Footnotes

    • Contributors AWS: data collection, analysis and interpretation, drafting and revising the manuscript. EM: data collection, manuscript revision. CG: data collection, manuscript revision. AML: data collection, manuscript revision. MH: data collection, manuscript revision. AEL: data collection, manuscript revision. RC: conceptualisation of the study, data interpretation, manuscript revision

    • Funding The study was funded by the Michael J. Fox Foundation for Parkinson Research and the Canadian Institutes of Health Research (CIHR, grant number MOP15128 to RC). AL is supported by the Canada Research Chair in Neurosciences, AL is supported by the Jack Clark Chair for Parkinson's Disease Research and the Edmond J. Safra Parkinson's Disease Research Program and RC is supported by a CIHR-Industry (Medtronic Inc) Partnered Investigator Award and the Catherine Manson Chair in Movement Disorders.

    • Competing interests AWS has received research grants from Dystonia Coalition, Dystonia Medical Research Foundation and MBI Institute at University of Florida. She also participates in research funded by Michael J Fox, Abbott, Merz, and Phytopharm, but has no owner interest in any of the pharmaceutical companies. She reports no conflicts pertaining to this study. EM received honoraria for serving on the educational advisory board for Medtronic. She was a consultant for and received honoraria for lecturing from Medtronic, and received research support from St Jude Medical. CG has nothing to disclose. AML received honoraria and a research grant from Medtronic Inc and St Jude Medical. He serves as consultant for Medtronic, St Jude Medical, Boston Scientific, Amgen, Ely Lilly, Bristol, Myers, Elekta, Bayer, Schering-Plough, QIG, and Functional Neuroscience. MH received speaker honoraria and a research grant from Medtronic Inc, and received research support from St Jude Medical. AEL has served as an advisor for Medtronic. RC received consulting fees from Medtronic Inc. This study has no industry sponsorship.

    • Ethics approval University of Toronto.

    • Provenance and peer review Not commissioned; externally peer reviewed.