Thalamic physiology of intentional essential tremor is more like cerebellar tremor than postural essential tremor

Abstract

The neuronal physiological correlates of clinical heterogeneity in human essential tremor are unknown. We now test the hypothesis that thalamic neuronal and EMG activities during intention essential tremor are similar to those of the intention tremor which is characteristic of cerebellar lesions. Thalamic neuronal firing was studied in a cerebellar relay nucleus (ventral intermediate, Vim) and in a pallidal relay nucleus (ventral oral posterior, Vop) during stereotactic surgery for the treatment of tremor.

Nine patients with essential tremor were divided clinically into two categories: one with a substantial component of tremor with intention (termed intention ET) and the other without (postural ET). These types of essential tremor were compared with patients having intention tremor plus other clinical signs of cerebellar disease (cerebellar tremor). Neurons in patients with either intention ET or cerebellar tremor had lower firing rates and lower spike × EMG coherence than those in patients with postural ET. Patients with intention ET had a lower spike × EMG phase lead than those with postural ET. Overall, thalamic activity measures of intention ET were different from postural ET but not apparently different from those of cerebellar tremor.

One patient with the intention ET (number 4) had a good response to a left thalamotomy and then suffered a right cerebellar hemispheric infarct five years later. After the stroke the intention ET recurred, which is consistent with our hypothesis that intention ET is similar to that of the intention tremor which is characteristic of cerebellar lesions.

1 INTRODUCTION

Essential tremor is one of the most common adult movement disorders (Brin and Koller, 1998; Louis et al., 1998), and can be characterized as tremor which is related to movements or postures of the limbs (Deuschl et al., 1998; Elble, 2006; Elble and Koller, 1990). Recent studies have demonstrated substantial phenotypic variability in essential tremor, which may be a postural tremor or may include a substantial component of intention tremor (Deuschl et al., 2000; Elble and Deuschl, 2011). This intentional component is poorly understood and has not been consistently associated with measures of pathology, imaging, or central nervous system electrophysiology (Elble and Deuschl, 2011). Human and animal studies suggest that the cerebellum plus related structures, and the thalamus, and the cortex are all involved in the mechanism of essential tremor.

Essential tremor is reduced by surgical lesions or stimulation of a cerebellar and a pallidal receiving nucleus of the thalamus, which are termed ventral intermediate - Vim and ventral oral posterior - Vop, respectively (Figure 1A)(Hirai and Jones, 1989; Jankovic et al., 1995; Krack et al., 2002; Schuurman et al., 2000). Imaging studies show increased metabolic activation of the cerebellum, thalamus and sensorimotor cortex during essential tremor (Boecker and Brooks, 1998; Jenkins et al., 1993; Perlmutter et al., 2002). Deficits of cerebellar function in patients with essential tremor also suggest that cerebellar inputs to the thalamus and cortex are involved in the mechanism of essential tremor (Deuschl et al., 2000; Helmchen et al., 2003; Stolze et al., 2001).

Figure 1.

Thalamic map generated during stereotactic thalamotomy in Patient 4. A. Sagittal map of thalamic neuronal location relative to the AC-PC line (horizontal line, PC as indicated) and nuclear boundaries (Voa - ventral oral anterior, Vc - ventral caudal). Electrode trajectories (P2 and P1) are shown by the oblique lines. The oblong shape overlapping trajectories P1 and P2 is the estimated extent of the lesion (Lenz et al., 1994b) B. Locations of recording sites are indicated by ticks to the right of each trajectory (P1 and P2). Long tics indicate location of a site at which the noise level changed, or at which a neuron had tremor related activity. Long tics with associated symbols indicate responses to cutaneous stimulation (circle) or joint movement (triangle). Short ticks indicate unresponsive neurons. Locations of stimulation sites are indicated by ticks to the left of each trajectory. Triangles indicate the location of a site at which stimulation led to a decrease in tremor, and circles indicate sites at which stimulation led to sensations referred to the skin. Sites are numbered sequentially and numbers are indicated for every fifth tick. Scale as indicated. C. P1and P2 show the physiological activity below the site number, which is listed along the vertical line and which corresponds to that in B. To the right of the trajectories are the receptive fields (RF). Shaded figurines indicate cutaneous RFs, and joint movement is indicated by arrows on the figurine. To the left of the line, shaded areas on the figurine indicate the area of stimulation evoked cutaneous sensation, while stimulation evoked decrease in tremor as indicated by the phrase ‘dec. in tremor’. The threshold for these effects is indicated in A.

Intention tremor is defined as tremor which increases in amplitude as the target is approached during visually guided movements. Intention tremor is seen in human subjects with cerebellar pathology or injury to cerebellar pathways, and in monkeys with transient disruption of the deep cerebellar nuclei by cooling through an implanted probe (Flament and Hore, 1988; Vilis and Hore, 1980). These tremors have been termed cerebellar tremor, and it has been proposed that cerebellar injury leads to changes in the timing of outputs from the cerebellum (Lenz et al., 2002; Vilis and Hore, 1980). Similar changes have been found in thalamic neuronal activity, which is consistent with the thalamus being a relay for cerebellar connections to cortex (Lenz et al., 2002).

In some patients, essential tremor has a substantial intentional component in the absence of cerebellar pathology. In other patients, tremor with intention is absent but there is a postural component, with or without a kinetic component. We arbitrarily term these two categories as intention ET and postural ET (cf (Deuschl et al., 1998; Elble and Deuschl, 2011; Marsden et al., 1983). One hypothesis is that essential tremor results from the increased activity of an olivary pacemaker, which transmits tremor related signals to the cerebellum and from there to the thalamus, cortex and periphery (Lamarre, 1995; Llinas, 1984). This is consistent with the finding that neurons in Vim and Vop of these patients show increased firing rates and tremor-related activity that are enabled by active movement (Hua and Lenz, 2005).

We now propose to test an alternate hypothesis that thalamic neuronal and EMG activities during intention ET are similar to those of the intention tremor which is characteristic of cerebellar lesions (cerebellar tremor). In this model, injury to the cerebellum or related structures leads to deafferentation of thalamic nuclei (monkey ventral lateral posterior and human Vim (Hirai and Jones, 1989)), and their cortical targets (Flament and Hore, 1988; Vilis and Hore, 1980). The cerebellar input to these nuclei is excitatory so that deafferentation results in decreased firing rates, and a phase lag in the thalamic spike train × EMG spectrum (Lenz et al., 2002; Vilis and Hore, 1980). We now test this hypothesis by examining thalamic neuronal activity in Vim and Vop during stereotactic thalamotomy in patients with postural ET, intention ET, and with intention tremor plus other signs of cerebellar disease (cerebellar tremor). As a critical test of these two possibilities, we examined the result of a cerebellar lesion in a patient with intention ET that would be predicted to increase tremor due to cerebellar disruption but decrease tremor due to a pacemaker in the cerebellum and related structures.

2 RESULTS

In total, 192 neurons along 57 trajectories were recorded in 13 patients undergoing thalamotomy or thalamic deep brain stimulation for the treatment of tremor. Five patients (54 neurons) with essential tremor were classified as having a substantial intentional component to their tremor, termed intention ET. Four essential tremor patients (40 neurons) were found to have an absent intention component, termed postural ET. Four patients (112 neurons) had intention tremor and signs of cerebellar disease and were classified as cerebellar tremor.

Most patients with essential tremor had a family history or an effect of alcohol upon their tremor or both, which is consistent with a diagnosis of essential tremor (Koller and Busenbark, 1997). The variability in the present population of patients with essential tremor is consistent with the known phenotypical variability of essential tremor including: the nature of the tremor itself (postural and intention ET), the presence of dystonic features and imbalance, plus the association with Parkinson’s disease (Elble and Deuschl, 2011). In this setting, other movement disorders occurring with essential tremor, such as non-tremulous cervical dystonia, may be viewed as co-morbidities of essential tremor (Hedera et al., 2010; Schiebler et al., 2011), which do not necessarily effect the ongoing essential tremor.

The control group consisted of recordings from three patients (61 neurons) who underwent surgery for chronic pain in the lower extremities. Some of the present results have been previously reported in separate studies of subjects with essential tremor, or cerebellar tremor, or chronic pain (Hua and Lenz, 2005; Lenz et al., 2002).

2.1 Thalamic Signal

2.1.1 Firing rates

The mean spontaneous firing overall varied significantly with the type of tremor (1-way ANOVA, F(3,247)=3.75, p=0.01). The mean rate was highest in the postural ET group (22.5±3 Hz) followed by controls with pain (20.9±1 Hz), then intention ET (17.7±3 Hz), Patient 4 (15.9 +2.8 Hz), and cerebellar tremor (12.4±1 Hz). Post hoc testing demonstrated that the firing rate postural ET was significantly greater than that for cerebellar tremor (P<0.05, section 4.4). The difference in firing rates was not significantly different between patient 4 and cerebellar tremor. In order to describe the thalamic data more accurately we next classified neurons by location into those in Vim versus Vop.

The Vim mean rates were significantly greater in postural ET than in cerebellar tremor (p<0.01, shown in Figure 2A), but not different from intention ET or from controls with pain (not shown). The mean Vim firing rate for intention ET was not significantly different from cerebellar tremor (not shown). The Vop mean rates of subjects with postural ET were higher than in cerebellar tremor (P=0.002, not shown). Therefore, firing rates in postural ET were consistently higher than those in cerebellar tremor.

Figure 2.

A. Mean spontaneous firing rates in each nucleus. B. Mean spontaneous firing rates among cells with sensory RFs. The number of neurons for which firing rates were measured is given by n. Bars indicate mean value ± standard error of the mean.

2.1.2 Effect of cell type (sensory vs non-sensory)

The activity of all neurons included was studied for a response to joint movements during mapping of the thalamus, as described in the Methods. These cells were located in the region anterior (Figure 1C: P2) and dorsal to the region in which cells respond to cutaneous stimuli (Figure 1C: P1, and (Lenz et al., 1988)). The proportions of cells responding to deep sensory stimuli and those not responding to such stimuli are shown by tremor type in Table 2. The proportion of neurons in Vim responding was greater for postural ET than cerebellar tremor (P=0.00012, Chi square with Bonferroni correction) and controls with pain (P=0.048). The number of sensory cells in Vop was different only between intention ET and cerebellar tremor (P=0.02, Fisher with Bonferroni).

Table 2.

Proportions of sensory/non-sensory cells in Vim and Vop by diagnosis. See text.

Sensory/non-sensory	Vim	Vop
Pain	9/42	0/9
Postural ET	21/24	3/6
Intention ET	10/26	10/9
Cerebellar tremor	11/75	01/16

Since sensory inputs may be an important factor in the relationship of cerebellar tremor and cortical activity (Flament et al., 1984; Hore and Flament, 1986; Vilis and Hore, 1977), we next examined the mean spontaneous rates for sensory cells across the four groups (Figure 2B). There was a clear and significant change in the firing rate of sensory cells according to patient groups (1-way ANOVA, F=3.47, p<0.05). Post-hoc testing showed that the firing rate of sensory cells in the postural ET group was significantly higher than that of cerebellar tremor and controls with pain. The rate for intention ET was not different from postural ET.

2.1.3 Spectral analysis

We next examined how the thalamic signal qualitatively differed between groups of patients. The frequency at which peak spike activity occurred was found for each neuron within the tremor range (1.9 Hz to 7 Hz) (Lenz et al., 2002). The mean “frequency of peak spike power” occurred at a different frequency for each group of tremor patients (1-way ANOVA, F(3,259)=8.75, p <0.0005). The mean frequency of this peak is significantly higher in postural ET patients (4.8Hz + 0.25, mean + SEM) as compared to cerebellar tremor (3.4Hz + 0.2, post-hoc Newman-Keuls test, p=0.0057) and intention ET (3.7Hz + 0.4, p=0.032). The frequency was not significantly different between intention ET and cerebellar tremor (Neuman-Keuls test p=0.34).

The signal-to-noise ratio (SNR) for a particular frequency was defined as the power at that frequency divided by the mean power throughout the spectrum; a threshold of 2 was used to identify a significant concentration of power at a frequency in the SNR (Jenkins and Watts, 1968). The spike SNR at the peak in the tremor frequency range varied significantly by patient group (1-way ANOVA, F(3,256)=9.64, p<0.0001). Post-hoc testing found that the mean SNR was significantly greater for postural ET (5.3 + 0.48) than for cerebellar tremor (2.0 + 0.27) or intention ET patients (2.54 + 0.32, Tukey HSD tests p<0.005 for both). The SNR in the tremor frequency range indicates the maximum concentration of power, which may reflect the ability of a cell to influence tremor.

2.2 Relation of Thalamic Activity to EMG Signal

The cross-correlation function for spike trains × simultaneously recorded EMG signals were estimated from the coherence and phase between these two signals (see Supplementary Figures 1 and 2 which are copied from (Lenz et al., 2002) and (Hua and Lenz, 2005)). The calculation of coherence and phase have been described in Section 4.4 (Experimental Procedures, Analytic Techniques) and tremor-related neuronal activity was defined by a SNR > 2 AND coherence > 0.42. Phase is only interpretable where the two signals are linearly related, i.e. spike channel × EMG coherence > 0.42 (Lenz et al., 2002).

Overall, there was no apparent difference between sensory versus non-sensory neurons in the proportion of neurons with tremor-related activity, as identified in spike trains with SNR > 2 AND spike × EMG Coherence > 0.042 (12/35 vs 43/91, 2-tailed Chi square p>0.05). There was no difference in the proportion of cells with tremor-related activity between Vim versus Vop (44/101 vs 10/17, P=0.30, Chi square). Significant differences were not found in the proportion of cells with tremor-related activity between the sensory cells in the postural ET (10/23) versus the intention ET (6/13) group (Chi square tests, p>0.05).

2.2.1 Spike × EMG coherence and phase

The mean coherence of the spike × EMG channel with the highest coherence was determined for each neuron at the frequency of the auto-power peak in the tremor frequency range. This measure of cross-correlation is shown in Figure 3 for each group of patients by neuronal nuclear location. The mean coherence of neurons in Vim was significantly higher in postural ET patients than either intention ET patients or cerebellar tremor patients (1-way ANOVA, post-hoc Neuman-Keuls tests p<0.05). Intention ET and cerebellar tremor patients did not differ in the mean coherence of the neuronal spike trains in either nucleus (post-hoc Neuman-Keuls tests Vim: p=0.145 and Vop: p=0.491). The mean coherence in Vop was significantly higher in postural ET than in intention ET patients (post-hoc Neuman-Keuls test p<0.05).

Figure 3.

Coherence of spike × EMG signal for patients with postural ET, intention ET and cerebellar intention tremor (see 4.4 Experimental Procedures: Analytic Techniques). The number n indicates the segments analyzed, which is not necessarily the same as the number of cells.

The lower thalamic SNR and coherence in cerebellar tremor may seem inconsistent with the amplitude of this tremor. However, the thalamic SNR and coherence are greater in tremor characterized by regularity, while cerebellar tremor is irregular (Hua and Lenz, 2005; Lenz et al., 2002).

We next examined the phase spectrum in which a negative phase indicated that neuronal activity led EMG. Figure 4 shows that neuronal activity leads EMG activity in all three groups, and that the postural ET group has the largest lead (mean ± SEM: −107 ± 9°) followed by cerebellar tremor (−96 ± 12°) then intention ET (−65 ± 7°). This difference was significant between intention ET and postural ET groups (Kruskall-Wallis test H=2.84, p<0.05) but not between cerebellar tremor and either essential tremor group (p>0.05). In combination with the analysis of coherence, these results demonstrate that postural ET is different from intention ET at the level of spike × EMG interaction.

Figure 4.

Phase of spike × EMG signal for patients with postural ET, intention ET and cerebellar intention tremor (see 4.4 Experimental Procedures: Analytic Techniques).

The results show that the physiology of postural ET is different from that of cerebellar tremor, as demonstrated in 7/9 physiological variables which were significantly different from cerebellar tremor. These 7 variables included: peak frequency in the tremor frequency range, firing rates (in Vim and Vop separately), incidence of sensory cells, firing rates of sensory cells, SNR, and coherence (in Vim only). Vop coherence, and phase lead were not different. Intention ET was not different from cerebellar tremor in any of these variables. Based on these results postural ET had more physiological differences from cerebellar tremor than intention ET had from cerebellar tremor (7/9 vs 0/9, P<0.05, Fisher). These results suggest that postural ET is different from cerebellar tremor while intention ET is not. Postural ET had similar numbers of physiological differences from cerebellar tremor and from intention ET (7/9 vs 5/9, P=1, Fisher), which demonstrates again that intention ET is not apparently different from cerebellar tremor.

2.3 Ipsilateral cerebellar stroke worsens intention ET

2.3.1 Description of Patient 4

We next examined the result of a cerebellar lesion in a patient with intention ET since the lesion should increase tremor due to a cerebellar disruption but decrease tremor due to a pacemaker in the cerebellum and related structures. Patient 4 (Table 1) with intention ET began to have tremor in the right upper extremity which then spread to the left. The patient had no family history of tremor and did not know the effect of alcohol upon the tremor. Propranolol and primidone had been tried without benefit. Tremor was demonstrated with posture and intention, both graded at 4/4 bilaterally on the Fahn clinical rating scale (Fahn et al., 1988). There was no head tremor and the remainder of the neurological examination was within normal limits. The preoperative MRI scan was within normal limits. Fifteen years after the onset of tremor, the patient underwent an uncomplicated left thalamotomy which resulted in a small lesion, as shown in Figure 1A (Lenz et al., 1994b). At follow-up 2 months later, the patient had a substantial improvement in activities of daily living, with a tremor rating of 1/4 in the right upper extremity with posture only.

Table 1.

Patient information. Only pertinent neurological examination findings are indicated. Family history of tremor is accompanied by number of first degree relatives affected. Where no formal tremor rating score is available, the arrows indicate an improvement (down) or increase (up) in tremor as assessed by the referring neurologist at follow up. MS: multiple sclerosis, BG: Basal Ganglia.

Case		Sex/Age/Han dedness	Tremor	Relief by Etoh/ Famil y Hx	Abnor mal finding s besides tremor on neurolo gical exam	Normal brain imagin g	Fahn Score Pre Post		Recorded units/trajector y	Procedur e site
Intention ET	1	F/36/Right	Right upper extremity, 6 Hz,	+/+2	none	+	2.1	1.9	1/6	Right
	2	M/46/Right	Right upper extremity worse, postural & kinetic, 3 cm amplitude.	+/+	none	+	3.4	1.3	15/4	Left
	3	M/69/Right	Bilateral upper extremity, posture & kinetic, 7 Hz.	+/+3	none	+	3.7	1.3	12/4	Left
	4	M / 77 / Right	Right upper extremity & head.	?/−	none	+	4.0	1.0	20/2	Left
	5	F / 54 / Right	Kinetic 4 cm amplitude.	+/+4	none	+	2.5	↓	8/3	Bilateral
Postural ET	6	M / 51 / Right	Hand and head, posture3 cm amplitude, posture.	+/+2	none	+	3.1	1.4	14/5	Left
	7	F / 41 / Left	Bilateral postural.	+/+	Mild non- tremulou s cervical dystonia	+	3.4	1.1	19/4	Right
	8	F / 73 / Right	Voice & hands. Right worse. Posture.	+/+1	none	+	1.4	1.0	12/3	Left
	9	F / 49 / Right	Left worse, postural, 2 cm amplitude.	?/?	Mild non- tremulou s cervical dystonia	+	0.9	1.1	9/6	Right
Intention tremor plus	1 0	M / 41 / Right and Left	Severe bilateral intention tremor and Head bob, nil at rest..	−/−	MS: internucl ear ophthalm oplegia, plantar upgoing, gait, scanning speech	Mild cerebral atrophy		↓	1 6/5	Bilateral
	1 1	M / 27 / Right	Left sided Postural, 6 cm amplitude.	−/−	Dysdiado chokinesi s, spasticity , gait, upgoing plantar	Intensit y Right frontal lobe, thalamu s & BG cyst?	3	↓	22/5	Right
	1 2	F / 45 / Right	Left arm intention & titubation.	−/−	MS: clonus, nystagm us, right plantar upgoing, gait	Perivent ricular white matter lesions		↓	31/5	Right
	1 3	F / 43 / Right	Head & trunk., right upper arm, Intention, 4 cm	−/−	MS: gait, diplopia, scanning speech	Not typical for MS		↓	43/5	Left

Twenty years after the onset of tremor, the patient had a total knee replacement and one week thereafter developed an acute onset of true vertigo and imbalance leading to falls, but no other symptoms or signs. These symptoms resolved and physical therapy was completed as planned. Five weeks after discharge from hospital, there was worsening of the patient’s right sided tremor, and by six weeks the right tremor was graded at 4/4 with posture and action. There was a substantial component of tremor with intention. By self-report, the tremor was similar to that prior to thalamotomy, being worse on moving the arm to eat and drink. An MRI within a month of the ictus (shown in Figure 5) demonstrated a complete stroke involving the right cerebellar hemisphere.

Figure 5.

Ipsilateral cerebellar stroke that worsened previously treated intention ET in patient 4 with intention ET (see Section 2.3: Ipsilateral cerebellar stroke worsens intention ET). A. Axial MRI showing a right cerebellar infarct. B. Comparison of mean spontaneous firing rate of thalamic neurons in both Vim and Vop in patient 4 at the time of the left thalamotomy, in pain controls and in patients with cerebellar tremor. Error bars represent standard error of the mean, n the number of neurons, and cross bars show significant differences.

2.3.2 Comparison of this subject with stroke to the subject groups

Firing rates in thalamic nuclei Vim and Vop for patient 4 are compared to patients with cerebellar tremor, postural ET and controls with pain in Section 2.1.1. There was no difference in the firing rates, or spike × EMG coherence or phase from this patient and the rest of the intention ET group (Mann-Whitney U test, z=1.22, p>0.2 for all comparisons).

3 DISCUSSION

We have now tested the hypothesis that thalamic neuronal and EMG activities during intention ET are similar to those of cerebellar tremor. The results show that intention ET was similar to cerebellar tremor in multiple measures of tremor related activity while intention ET was apparently different from postural ET in multiple measures. Overall, the characteristics of intention ET are consistent with a mechanism similar to that of cerebellar tremor but different from that of postural ET (Hua and Lenz, 2005; Lenz et al., 2002; Vilis and Hore, 1980). This mechanism may be based upon disruption of cerebellar function, as in cerebellar tremor. Specifically, intention ET versus postural ET demonstrated lower firing rates, lower SNR, and smaller phase lead of spike × EMG, all of which are consistent with the deafferentation of the thalamus by a cerebellar lesion, as shown in monkey studies (Lenz et al., 2002; Vilis and Hore, 1977; Vilis and Hore, 1980).

Postural ET had as many differences from intention ET as from cerebellar tremor, which suggests that postural ET is not due to cerebellar disruption. In addition, the higher firing rates, SNR, and phase lead of postural ET may result from excitatory oscillatory input to the thalamus, consistent with a pacemaker in the olive (Lamarre, 1995; Llinas, 1984). The cerebellar lesion occurring in patient 4 with intention ET is a critical test of whether intention ET is the result of cerebellar disruption or a cerebellar pacemaker. The lesion should increase tremor due to a cerebellar disruption but decrease tremor due to a pacemaker in the cerebellum and related structures. Patient 4 with intention ET had a cerebellar stroke (Table 1, Figure 5), which increased his intention tremor. In light of this case, the physiological differences described above strongly suggest that intention ET is the result of disruption of the cerebellum.

3.1 Methodological Considerations

The frequency of thalamic activity during cerebellar tremor in this series is consistent with the accepted frequency range for cerebellar tremor in the literature (Deuschl et al., 1998; Elble and Deuschl, 2011). All patients with cerebellar tremor had MS and signs indicating lesions of systems other than the cerebellum and its pathways. Nevertheless, the morphology of the tremor, the frequency of tremor related activity, and the associated cerebellar signs suggest that tremor in these MS patients are similar to that in patients with lesions restricted to the cerebellum and its pathways.

The actual location of cells in the present study is uncertain because stereotactic localization on the basis of the AC-PC line is subject to significant random errors (Lenz et al., 1990; Lenz et al., 1994a). The uncertainty in location was minimized in the present case by aligning the anterior border of Vc with the most anterior cell along any trajectory where the majority of cells located posteriorly are sensory cells (as in Figure 1). This procedure minimizes the random radiologic errors that effect the estimate of nuclear location.

Spike × EMG phase is an important estimator of relative latency, but does not give an unambiguous measure of the relationship between the two signals. For example, a phase angle of X may also indicate a phase angle either of X or of X-360. Therefore, the phase cannot be expressed in terms of latency of spike versus EMG signals, and the interpretation of phase results in terms of tremor mechanisms must be approached cautiously.

3.2 Thalamic Activity by Tremor Diagnosis

This study is in agreement with the finding that tonic firing rates in the human thalamus, and in nucleus Vim particularly, are lower in patients with cerebellar intention tremor than controls (Figure 2). Mean spontaneous firing rates in patients with postural ET have previously been shown to be elevated compared to controls (Hua and Lenz, 2005; Molnar et al., 2005) at between 18-25 spikes/sec. By comparing postural ET and intention ET, this study demonstrates a higher firing rate in postural ET. In addition, spontaneous firing rates in intention ET were not different from the tonic rate seen in patients with cerebellar tremor.

It is often supposed that an olivary pacemaker drives essential tremor (see Section 1: Introduction) which predicts increased firing rates in Vim, as the major recipient nucleus of excitatory cerebellar drive (Anderson and Turner, 1991). Indeed, Vim rates during postural ET were higher than both controls with cerebellar tremor and controls with pain. This is strong evidence that postural ET is associated with increased excitatory drive from the cerebellum.

Firing rates in postural ET were higher than those in cerebellar tremor for neurons both in Vim and Vop, which is unexpected given the inhibitory projection from the internal pallidum to Vop (Anderson and Turner, 1991). Excitatory cortico-thalamic connections rather than inhibitory input from the pallidum may be the most important determinant of firing rates in the monkey pallidal receiving nucleus of the thalamus (Monkey Ventral Lateral anterior corresponding to human Vop)(Anderson and Turner, 1991; Hirai and Jones, 1989). Vop is reciprocally connected with the supplementary motor area (Holsapple et al., 1991; Sakai et al., 1999; Schell and Strick, 1984), which is densely interconnected with the motor cortex (Tanji, 1994). Vim is reciprocally connected with motor cortex and Vop receives considerable input from these cortical motor structures through the cortico-thalamic projection. These connections may explain why increased firing rates in postural tremor are found in both Vim and Vop (Hirai and Jones, 1989; Hua and Lenz, 2005; Lenz et al., 2002).

3.3 Spike × EMG by Tremor Group

Spectral analysis showed that coherence and phase for intention ET were more similar to cerebellar tremor than to postural ET (Figures 4 and 5). The phase lead was significantly greater for postural ET than for intention ET. Intention ET and cerebellar tremor patients had much lower coherence and SNR than postural ET subjects. Overall, this physiology seems to confirm the clinical observation that intention ET is similar to cerebellar tremor (Brennan et al., 2002; Elble and Koller, 1990).

The frequency of peak spike power was higher for the postural ET group than for the intention ET or cerebellar tremor groups, which is consistent with the similarities noted above. Clinically, cerebellar tremor is of lower frequency than essential tremor (Elble, 2006; Findley and Koller, 1987). The lower frequency of intention ET and cerebellar tremor versus postural ET again suggests that intention ET is more like cerebellar tremor than postural ET.

3.4 Effect of Cerebellar Stroke upon Intention ET

Patient 4 was not an exceptional case since there was no bias in the sampling of cell types or predominance of a particular nuclear location. Patient 4 was indistinguishable from other patients in the intention ET group in terms of spontaneous firing rates, and frequency of peak spike power in the tremor range.

If a pacemaker of the cerebellum and related systems drives intention ET then a lesion of the cerebellum might further reduce this patient’s tremor and would not increase tremor. In postural ET, lesions of the cerebellum or pontine cerebellar connections decrease tremor (Dupuis et al., 1989; Nagaratnam and Kalasabail, 1997). Therefore, the increase in intention tremor following the cerebellar stroke in patient 4 is consistent with a mechanism of intention ET related to disruption of the cerebellum rather than to a pacemaker (Destexhe and Sejnowski, 2001; Lenz et al., 1994b; Stein and Oguztoreli, 1976). This difference could be tested by imaging studies of basal- and tremor-evoked activity in patients with either postural ET or intention ET.

The analysis of thalamic neuronal activity and of the spike × EMG cross-correlation demonstrates that intention ET is more like cerebellar tremor than like postural ET. Of course, cerebellar tremor is often associated with lesions of the cerebellum or its output pathways (Carrea and Mettler, 1947; Gilman et al., 1976). Therefore, the present results suggest the intention ET is associated with disruption of the cerebellum, which may be consistent with the histologic changes in Essential Tremor (Louis et al., 2012).

In healthy monkeys, active oscillations of the wrist are associated with a substantial phase lead of thalamic activity upon tremor (Butler et al., 1992). The present results show a lag in thalamic activity during intention ET relative to other types of tremor (Figure 4). This lag may be congruent to delays in motor cortical activity during tremulous isotonic movements that occur with cooling of the cerebellar nuclei in monkeys (Vilis and Hore, 1977; Vilis and Hore, 1980). By analogy, the spike × EMG phase in intention ET and cerebellar tremor may contribute to the tremor, which is observed in these groups. In turn, the resulting delay in motor cortical activity may reflect the influence of sensory feedback on cerebellar feed-forward activity in tremulous movements associated with cerebellar cooling (Hore and Flament, 1988), and possibly with intention ET. The similarity of intention ET to cerebellar tremor suggests that it may result from disruption of the cerebellum, and not from the cerebellar pacemaker which is often associated with postural ET.

4 EXPERIMENTAL PROCEDURES

This study was carried out during the physiological exploration of the thalamus, which preceded implantation of deep brain stimulation electrodes or thalamotomy, either for the treatment of tremor or chronic pain. The descriptions of all techniques used in this manuscript have previously been published in detail (Hua and Lenz, 2005; Lenz et al., 2002).

All patients were assessed by a neurologist specializing in movement disorders and underwent a full clinical assessment (Table 1). The severity of tremor was graded using the validated Fahn rating scale (Fahn et al., 1988), which includes objective evaluation of tremor amplitude.

For this analysis, nine essential tremor patients were divided into sub categories: one with a substantial component of tremor with intention (termed intention ET) and the other without (termed postural ET), based on the visual inspection. All patients were operated more than five years after the diagnosis. None of the patients with essential tremor had a preexisting cerebellar injury on the basis of history, neurological examination, and brain imaging as reviewed with a neuroradiologist. None of the patients with cerebellar tremor had any family or personal history of essential tremor or any symptoms of tremor before their cerebellar injury.

The control group consisted of patients with neuropathic lower extremity pain. None of them had a personal or family history of tremor. They had no cerebral or cerebellar pathology based on detailed neurological examination and MRI imaging. Their electrophysiological recordings were therefore suitable to use a control for comparison of firing rates and other parameters with the tremor patients.

The protocol was reviewed and approved annually by the Institutional Review Board of the Johns Hopkins University. All patients signed an informed consent for these studies. Details of the methods used in this study have been previously described (Hua and Lenz, 2005).

4.1 Operative Procedure

Thalamic exploration was performed as a stereotactic procedure using the Leksell frame in patients who were off tremor medications for at least 18 hours. First, the frame coordinates of the anterior (AC) and posterior commissures (PC, Figure 1A) were measured by magnetic resonance imaging (MRI) or computed tomography. These coordinates were used to estimate the nuclear locations. Physiological corroboration of nuclear location was then performed under local anesthesia without sedation (i.e. subject fully conscious) by single unit recording and microstimulation through a microelectrode.

We used a platinum-iridium electrode etched to a tip of 3-4 mm and coated with solder glass to give an impedance of approximately 2.5M Ohms, which was reduced to approximately 0.5M Ohms by microstimulation (50 microA) in the brain. The electrode was advanced toward the target as localized by pre-operative imaging. The signals recorded on magnetic tape (Model 4000, Vetter Corp., Rebersberg, PA, USA) or electronically (Cambridge Electronic Design, CED 1401 interface) during the procedure included: the foot pedal indicating events during the examination, the microelectrode signal, electromyogram (EMG) for wrist flexors and extensors plus elbow flexors and extensors in the contralateral upper limb, the audio channel describing instructions to the patient as well as technical details of the procedure.

4.2 Physiological Techniques

The physiological exploration with the microelectrode involved both the recording of neuronal activity and stimulating at microampere current levels. When a neuron was isolated, spontaneous activity was recorded. The activity of the isolated neuron was then studied to identify neurons responding to cutaneous stimuli such as light touch, tapping or pressure to skin. Neurons responding to deep stimuli were identified by a response to manipulation of muscles or ligaments and passive joint movement in the absence of a response to cutaneous structures deformed by these stimuli. These neurons are termed cutaneous or deep sensory cells, respectively.

The activity of neurons was also examined as patients carried out movements such as making a fist, flexing or extending the wrist and elbow. Tremor was studied at the arm position achieved at the end of a movement in which the subject pointed to the corner of the room. The patient was seated in a reclining position with the head of the bed at 20° elevation to the horizontal. In this position, the shoulder was flexed to about 45° with the elbow, wrist, metacarpophalangeal and interphalangeal joints all extended to a little less than 180°. The tremor was provoked by this maneuver and the neuronal activity related to tremor was recorded for a period of between 20 and 60 sec.

Microstimulation was carried out along each trajectory, delivered through the microelectrode in trains of approximately 1 sec duration at 300 Hz using a biphasic pulse consisting of a 0.2 ms anodal pulse followed in 0.1 ms by a 0.2 ms cathodal pulse of the same magnitude. At each stimulation site, patients were asked during stimulation if they felt anything. If any effect was observed then the current was lowered in a series then raised in a series until a threshold for the effect was established. This technique, called threshold microstimulation (Lenz et al., 2004), allows the nature of the effect and the location of the projected field to be determined at threshold. The locations of sites at which neurons were recorded or microstimulation was carried out are shown in Figure 1B.

4.3 Estimate of Nuclear Location

In human studies, the borders of thalamic nuclei must be defined physiologically since radiological estimates are not reliable. Microstimulation evokes changes in the ongoing movement disorder, which are occasionally associated with brief muscle twitches. These latter changes are not common enough to reliably define the borders of Vim. The borders of Vc (ventral caudal nucleus of the thalamus) can be defined physiologically and used to extrapolate locations of other nuclei by registration with atlas maps to the borders of Vc.

Previous studies in humans indicate that sensory cells account for the majority of cells in Vc but are in the minority in Vim and Vop (Figure 1). Therefore, the anterior border of Vc was defined by the most anterior cell in the region where the majority of cells were deep or cutaneous sensory cells. The physiological map of each patient was shifted along the AC-PC line so that the most anterior cell in this region was at the anterior border of Vc, as in previous studies (Hua and Lenz, 2005), and as illustrated in Figure 1. The borders of presumed Vim and Vop were determined from this transformed map.

4.4 Analytical Techniques

Cells analyzed in the present report were located in the region where cells exhibited activity related to tremor, deep sensory stimulation, or active movements of the upper extremity. The electrophysiological results were effectively blinded as data was analyzed and processed before the final assignment of patients to essential tremor diagnosis groups, i.e. postural ET versus intention ET (Hua and Lenz, 2005). Times of occurrence of action potentials were digitized at a clock rate of 1000 Hz and EMG signals were digitized at a rate of 200 Hz and processed, as described below. When data was recorded in digital form, spike train signals were analyzed post-operatively using Spike2 software (CED, Cambridge, UK) that allowed for template-matching of waveforms.

We had data epochs that were long enough to be analyzed by a standard spectral analysis technique, rather than by a multi-taper technique, which would have been appropriate for shorter epochs (Percival and Walden, 1993). The thalamic and EMG signals were analyzed in the frequency domain. The EMG signal was band-pass filtered to eliminate movement artifact then full wave rectified and filtered at 20 Hz to produce a signal known as the demodulated EMG (Hua and Lenz, 2005). The spike train was first converted into an equivalent analog signal; thereafter standard techniques were used to take the spectrum of the spike and EMG signals (Bendat and Piersol, 1976; French and Holden, 1971; Hua and Lenz, 2005).

The concentration of the power of the neuronal signal in the tremor frequency range, and the cross-correlation of the neuronal signal to EMG signals were measures of the extent to which the neuronal signal reflected the tremor signal.

The coherence function was used as a measure of the probability that any two signals were linearly related. The coherence has a value of zero if the two signals are completely unrelated and one if there is a perfect linear relationship between the signals at a particular frequency. By the technique used in the present study, a coherence of greater than 0.42 indicated that two signals were linearly related at the level of p<0.05 (Hua and Lenz, 2005). Phase was calculated by standard techniques for epochs of neuronal and EMG signals in which the coherence demonstrated a significant linear relationship. A negative phase in the spike × EMG cross-correlation function indicated that the spike train had a phase lead with respect to the EMG signal.

Statistical testing of parametric data was carried out using a 1-way ANOVA with a post-hoc Tukey HSD test (Honestly Significant Difference), or Neuman-Keuls test. Testing of non-parametric data was carried out using Chi square or Fisher exact test, as appropriate.

5 CONCLUSION

Postural ET and intention ET were identified and were compared with intention tremor plus other clinical signs of cerebellar disruption (cerebellar tremor). Thalamic neurons in patients with either intention ET or cerebellar tremor had lower firing rates and lower spike × EMG coherence than those in patients with postural ET. Patients with intention ET had a lower spike × EMG phase lead than those with postural ET. Overall, thalamic activity in intention ET was different from postural ET but not apparently different from cerebellar tremor.

One patient with the intention ET had a good response to a left thalamotomy and suffered a right cerebellar hemispheric infarct five years later. After the stroke the intention ET recurred, which is consistent with our hypothesis that intention ET is similar to cerebellar tremor. After such a stroke, intention ET would be predicted to increase if it were due to cerebellar disruption but decrease if it were due to a pacemaker in the cerebellum and related structures. This difference in mechanism suggests an explanation of cases in which postural ET progresses to intention tremor over time.

Supplementary Material

01

Supplementary Figure 1. Simultaneous recording of thalamic single neuron activity and peripheral EMG in a patient with postural ET. A, Digitized spike train (upper trace) and demodulated EMG channels (lower two traces). B, Smoothed spike autopower spectrum of the spike train illustrated in A. C,D, Smoothed autopower spectra for the two demodulated EMG channels. The dot indicates the frequency at which the maximum spectral component in the tremor frequency range (4-10Hz in this study) occurs in the EMG autopower spectrum—that is, tremor frequency (5.81 Hz). E, F, G, cross power spectra, H, I, J, coherence spectra, and K, L, M, phase spectra between spike X EMG1, spike X EMG2, and EMG1 X EMG2, respectively. The numbers to the right of the spectra indicate the frequency with the highest autopower (B, C, D), and the value at tremor frequency of cross power SNR (E, F, G), coherence (H, I, J), and phase (K, L, M). This figure is copied from {Hua, 2005 4392/id}, with permission.

02

Supplementary Figure 2. Thalamic spike train and EMG signals in a patient with cerebellar tremor. A, the times of occurrence of thalamic action potentials (upper trace) and the demodulated EMG signal in wrist flexors (lower trace). B and C, respectively, raw spike power spectrum (frequency given in Hz) and the spike power spectra after smoothing by frequency averaging in groups of 8 non-overlapping raw estimates. D, coherence and phase functions for the spike X EMG cross-correlation functions. This figure is copied from {Lenz, 2002 3736/id}, with permission.

• -Is intention essential tremor (ET) similar to cerebellar tremor (CT)?
• -Study of human EMG and thalamic neuronal firing in tremor in humans.
• -Intention ET and CT have similar firing rates and spike × EMG phase lead.
• -Intention ET has lower firing rates and phase lead than postural ET.
• -Intention ET may be due to disruption of the cerebellum, like CT.