Neurophysiology of the ventral intermediate nucleus

The VIM in general has tonic cells with a frequency range of 20–50 Hz. Many of the cells in the lower two-thirds of the nucleus respond to joint movements. Important to the placement of the DBS lead are cells that have a bursting pattern that are highly correlated to the tremor activity [18].

The close temporal relationship between tremors and bursts of thalamic neuronal activity has led to the suggestion that these thalamic tremor cells could act as tremorigenic pacemakers. Intraoperative electrical stimulation in areas of the thalamus that are populated by tremor-synchronous cells momentarily arrests tremor, an observation that is important in the selection of the most effective target for tremor control [19]. It is thought that VIM stimulation desynchronizes neuronal overactivity within the VIM and leads to “functional ablation” [20, 21]. The VIM, as part of the nucleus ventralis lateralis (VL), seems to be the last relay of a complex of neuronal loops that finally projects to the motor and premotor cortex [22]. Moreover, the complete ventral thalamic nuclei are considered to act not only as relays but also as convergence points for pallidal and cerebellar afferent pathways [23].

The ventral intermediate nucleus procedure

The VIM-DBS procedure is divided into the following stages: stereotactic imaging, thalamic mapping, electrode implantation, implantation of the pulse generator and programming. Physiological mapping is carried out to identify the tremogenic region of the VIM and the location of the VC nucleus (somatosensory relay nucleus) that lies immediately posterior to the VIM. Stimulation of this nucleus produces paresthesias in the corresponding part of the body. Historically, prior to microelectrode recording, the VC nucleus was targeted, and once the area that responded to brushing of the thumb was located, the electrode was moved anteriorly. The VIM nucleus of the thalamus is located anterior to the VC and has characteristic features of VIM neurons, which include response of neurons to kinesthetic inputs and the spontaneous tremor-synchronous activity. Sites within the VIM in which tremor is arrested with stimulation are potential surgical targets.

Results of ventral intermediate nucleus surgery

The results of VIM-DBS show that this procedure is highly beneficial for tremor control in patients with PD but ineffective for the other disabling features of PD, including bradykinesia, rigidity, gait and postural disturbances. Unilateral or bilateral stimulation of the VIM is therefore considered only for tremor-dominant PD with a long duration of disease and in the absence of pronounced rigidity or bradykinesia. It has largely been replaced by subthalamic nucleus (STN) and globus pallidus interna (GPi) stimulation [24].

There have been several studies showing lack of efficacy of the thalamic target on the other symptoms of PD, namely rigidity, bradykinesia and dyskinesias [1, 11, 13, 20, 25–32].

As seen in Table 20.1, symptom control with DBS tends to be stable over time, although as many as 20% of patients with VIM stimulation experience reduced efficacy over time, and between 22 and 57% may develop rebound tremor upon discontinuation of stimulation [40], which is in contradiction to lesioning procedures [38].

The effectiveness of VIM-DBS for tremor has been established; however, the long-term studies pay little attention to progression of other signs of parkinsonism. All of the studies included a relatively small number of patients with PD. Benabid et al. [20] studied 26 patients with PD and six patients with essential tremor; seven of these patients had already undergone thalamotomy contralateral to the stimulated side, and 11 others had bilateral VIM stimulation at the same time. In patients with PD, rigidity was moderately improved but akinesia was not. Levodopa (l-DOPA) treatment had to be continued in all but two patients, but doses were reduced by more than 30% in ten patients.

Blond et al. [27] studied ten patients with parkinsonian tremor. Three of the ten parkinsonian patients had previously undergone contralateral thalamotomy. The follow-up period ranged from 6 to 26 months (mean 19.4 months). The intensity that effectively suppressed tremor was low (1–1.5 V) during the first 2 months but was increased in seven patients to maintain tremor alleviation. During the follow-up period, tremor reappeared transiently in three patients. Akinesia and rigidity were not changed by stimulation. Dyskinesias affecting limbs disappeared in all cases. Dystonic dyskinesias disappeared in two out of three patients after surgery. Neuropsychological evaluation failed to reveal memory, speech or praxis skill disorders.

Benabid et al. [26] followed 80 patients with PD for up to 8 years. Tremor was essentially the only symptom that was substantially influenced by VIM stimulation. Rigidity was slightly affected but was most likely due to a reduction in cogwheeling when the tremor was suppressed. There was almost no change in bradykinesia or any other symptom of PD.

Benabid et al. [41] followed 91 patients with PD, with a follow-up period of up to 17 months; 11 patients had contralateral thalamotomy and 43 patients had bilateral implants. Tremor was the only parkinsonian symptom that was greatly influenced by VIM stimulation. Parkinsonian resting tremor had the best benefit after surgery. Similarly, rigidity was only slightly affected, most likely due to a reduction in cogwheeling when the tremor was suppressed. There was no change in bradykinesia or in other parkinsonian symptoms. The effect on tremor was scored independently by the neurologist on a 5-point scale. At 3 months after the procedure, 86% of patients had a global score of 3 + 4 (permanent suppression of tremor or only slight recurrence on rare occasions) in the upper limbs and lower limbs. Resting tremor was better controlled than action tremor, distal-limb tremor was better controlled than proximal or axial tremor, and upper-limb tremor was better controlled than lower-limb tremor. In all patients, the effect was strictly coincident with stimulation, and there was no significant delay at the onset or after the effect at the cessation of stimulation. Thirty-nine of 80 patients (48.7%) had their levodopa dosage decreased by 20% at their 3-month follow-up. Because of disease progression, there were only 12 patients (15%) at the last follow-up. Rigidity and pain, as well as levodopa-induced dyskinesias, were partly reduced. Akinesia was not modified. The effect of VIM stimulation remained stable in 96%.

Hariz et al. [42] followed 22 patients with PD for up to 21 months (range 3–52 months). Seventeen patients were followed for more than 1 year. Seventeen patients had stimulation of the left thalamus. Five patients (23%) had previous contralateral thalamotomy, nine patients (41%) had a contralateral pallidotomy and three patients (14%) had a pallidotomy ipsilateral to the DBS side. All patients needed incremental increases in their stimulation parameters during the first 6–12 months after surgery. Ninety percent of patients with PD initially had good tremor control after surgery. At 1 year, 70% of patients with PD still exhibited good tremor control. Beyond 1 year, stimulation parameters and the effect on tremor seemed to stabilize. There was also an overall improvement of the Unified Parkinson’s Disease Rating Scale (UPDRS) scores at 1 year, with the greatest improvement on the tremor items of the contralateral limb. Mean UPDRS-III motor scores improved from 37.2 preoperatively to 26.6 postoperatively while the stimulation was on (P < 0.01). However, the contralateral tremor scores showed a tendency to increase after the stimulation was switched off, compared with the preoperative values (P = 0.007).

Kumar et al. [36] followed seven patients; four had a bilateral procedure and three had a unilateral procedure. The mean follow-up time was 16.2 ± 7.0 months. The Schwab and England score and the total UDPRS score showed significant improvement at 1 week postoperatively (P < 0.01). However, long-term follow-up at 16.2 ± 7.0 months failed to show any significant change in these scores. Contralateral arm and leg resting tremor and ipsilateral resting leg tremor were found to be significantly decreased 1 week postoperatively (P < 0.04) and at follow-up (P < 0.02). Ipsilateral arm tremor did not show any statistically significant change. No significant improvement was seen in rigidity, bradykinesia, gait, speech or postural stability at immediate or long-term follow-up. Dopamine dosage was decreased by about 20% over early follow-up, but there was a tendency for this dose to increase with later follow-up. There was no significant difference in medication usage at long-term follow-up.

Lyons et al. [37] followed 12 patients with PD. Nine patients had follow-up evaluation for longer than 24 months; three patients were lost to follow-up. The last postsurgical evaluation occurred on average at 40 months after surgery. Motor UPDRS scores did not change significantly from baseline to long-term follow-up (40.7 at baseline to 35.6 at long-term follow-up; P value not significant). However, tremor scores for the targeted side were significantly improved when stimulation was on at long-term follow-up, compared with baseline (7.2 at baseline to 0.9 at long-term follow-up; P = 0.007). There were no significant changes in any stimulation parameters from follow-up at 3 months to the last follow-up at 40 months.

Krauss et al. [35] prospectively assessed 45 patients with PD with a mean follow-up of 11.9 months (range 3–24 months). Symptomatic improvement of tremor was rated as excellent in 51% of the patients. There was no description of other symptoms of PD.

Pollak et al. [43] followed 81 patients. Use of VIM surgery alleviated contralateral tremor in approximately 85% of these patients. “Off”-period dystonia related to chronic levodopa treatment was not alleviated, and VIM stimulation was ineffective on akinesia.

Kumar et al. [44] followed eight patients with a mean follow-up of 49 months (range 44–62 months). Contralateral arm tremor mean scores were improved with stimulation compared with baseline by 82% at 1 year (n = 8, P < 0.05) and by 86% at the final follow-up (n = 7). Total body tremor scores improved by 62% with stimulation at 1 year (n = 8, P < 0.05) and by 44% at the final follow-up (n = 4). Patients showed varied long-term responses. Three patients had excellent relief of disabling tremor. Two patients had some improvement in tremor with stimulation; however, tremor was no longer the major source of disability since levodopa-induced dyskinesias, motor fluctuations and bradykinesia had become most problematic, despite stimulation-off tremor ratings, which were worse than baseline in one patient. Three patients did not use stimulation as their tremor progressively declined.

Rehncrona et al. [49] studied 20 patients, with follow-up at 2 years and 6–7 years after DBS. The UPDRS evaluations were done in a double-blinded manner. Stimulation significantly (P < 0.025) suppressed tremor in both the upper and lower extremities at 2 years as well as at 6–7 years after surgery. Tremor suppression was also statistically significant compared with baseline (before operation) for both follow-up periods. In the majority of patients, resting tremor in the upper extremity improved two grades or more and in the lower extremity at least by one grade. In some patients, stimulation completely suppressed tremor. One of these patients had microthalamotomy effect at operation and only mild tremor (grade 1) without stimulation at the first follow-up. At the 6–7 year follow-up, this patient had no tremor, irrespective of whether the stimulator was activated or not. Stimulation also significantly (P < 0.025) suppressed kinetic tremor in PD patients. Use of DBS (stimulation on compared with off) improved the total motor score at 6–7 years, not only by suppressing tremor but also by decreasing akinesia (P < 0.025) in the side contralateral to the implanted electrode. Speech and postural stability deteriorated at the last visit and were not improved by stimulation. The mean daily intake of levodopa in the entire group increased at the last visit.

Putzke et al. [40] followed 23 patients for up to 3 years. Stimulation was associated with significant improvement in both subjective and objective measures of performance of activities of daily living, midline tremor, and contralateral upper- and lower-extremity tremor, including parkinsonian resting and action tremors, over the follow-up period. Ipsilateral tremor showed little or no effect of stimulation after the first 3 months. Use of antiparkinsonian medications and stimulation parameters showed little or no change over the course of follow-up.

Tarsy et al. [24] followed 17 patients for up to 4.5–7.3 years (mean 5.5 years). There was significant improvement in resting and postural/action tremor in the target upper extremity at both the 12-month and last visit with the DBS on compared with the DBS off. Tremor severity was not significantly increased at the last visit compared with the 12-month visit, with the DBS either on or off. However, there was no significant improvement in rigidity at either 12 months or the last visit, and rigidity was not significantly increased at the last visit compared with the 12-month visit with the DBS either on or off. Thalamic DBS produced significant improvement in upper-extremity akinesia in the target extremity (contralateral extremity) at the 12-month visit but not at the last visit. Three of four measures of upper-extremity akinesia were improved at the last visit compared with the 12-month visit with the DBS off but not with the DBS on. Leg agility was improved at the last visit compared with the 12-month visit only with the DBS on. The total UPDRS motor scores were significantly improved with the DBS on compared with the DBS off at 12 months and at the last visit, which was attributable to improved tremor and akinesia scores with the DBS on. There was no improvement in speech, facial expression, postural stability and gait disturbance at either the 12-month or the last visit.

Finally, Hariz et al. [34] described 38 patients with PD, with up to 6 years of follow-up. Tremor was still effectively controlled by DBS, and appendicular rigidity and akinesia remained stable compared with baseline. Axial scores (speech, gait and postural instability), however, worsened, and the initial improvement in activities of daily living scores at the 1-year follow-up had disappeared at 6 years, despite sustained improvement of tremor. Remarkably, neither daily doses of dopaminergic medication nor fluctuations and dyskinesias had changed at 6 years compared with baseline in this patient group.

Stimulation parameters

Stimulation parameters showed little or no change over time [37, 39, 40]. Voltage and pulse width showed an increase in stimulus frequency and only a slight increase in pulse width. Impedance increased over time.

The voltage intensity that effectively suppressed tremor was low (1–1.5 V) during the first 2 months but was increased over time [27]. Similar results were seen by Benabid et al. [41] when voltage increased from an average initial value of 1 ± 0.5 to approximately 4 ± 0.5 V, reaching a plateau after approximately 2 months, with a mean value of 3.06 V. Hariz et al. [34] found lower stimulation amplitudes of 1.9 V at 3 months and 2.14 V at 12 months. Pulse width was 73.7 μs at 1 year. Koller et al. [11] found slightly higher amplitudes and pulse widths (3.03 V at 12 months and 117.5 μs at 12 months). Kumar et al. [36] found the mean stimulus intensity was 1.86 ± 00.83 V at 7 days postoperatively and increased to 2.9 ± 01.0 V at 120 days postoperatively. The impedance also increased from a mean of 738 Ω postoperatively to 1032 Ω by day 120. This increase in impedance was thought to be responsible for the increase in voltage needed to control the tremors. Lyons et al. [37] found that amplitude did not change significantly from the third postoperative month to follow-up at more than 12 months (from 3.2 to 3.6 V). Pulse width changed from 76.7 μs at 3 months to 80.0 μs at more than 12 months. There was no significant change in rate from 3 months to long-term follow-up (from 155.0 to 158.3 Hz). Benabid et al. [26] found that amplitude reached a plateau after approximately 2 months, with a mean value of 3.06 V. They also found increased impedances from a mean of 794 Ω on postoperative day 14 to 1057 Ω at day 106. They also suggested that an increase in impedance could account for the increase in threshold voltage intensity. Limousin et al. [29], in their multicenter study, found that voltage was slightly but significantly increased over 1 year, and changes in pulse width and frequency were small. Monopolar stimulation was used in most of the patients in this study. Bipolar stimulation was used when tremor reduction was obtained, with fewer adverse effects than monopolar stimulation.