J. O. Dostrovsky

Pain-related neurons in the human cingulate cortex.

Although it is widely accepted that the cortex participates in pain perception, there is no direct evidence for the existence of cortical neurons that respond to noxious or painful stimuli in humans. Anatomical and neurophysiological studies in animals as well as brain imaging and evoked potential studies in humans suggest that the anterior cingulate cortex (ACC) is an important area for processing sensory information related to pain. We have now identified single neurons in ACC that respond selectively to painful thermal and mechanical stimuli, supporting a role for the ACC in pain perception.

Identification and characterization of neurons with tremor-frequency activity in human globus pallidus.

Abstract Many previous studies have demonstrated the existence of neurons with tremor-frequency activity (”tremor cells”) in the thalamus of Parkinson’s disease (PD) patients and these neurons are presumed to play a role in the pathogenesis of tremor. Since a major input to motor thalamus (Voa and Vop) is from the internal segment of the globus pallidus (GPi), neurons with tremor-frequency activity in motor thalamus may receive input from neurons in GPi. The aim of this study was to quantify the characteristics of tremor cells in human globus pallidus. In three PD patients with tremor undergoing microelectrode exploration of the globus pallidus prior to pallidotomy, 228 neurons were sampled, and 28 (12.3%) were identified to fire at the same frequency as the tremor. These ”tremor cells” were located in the ventral portion of GPi. Autocorrelogram analysis of the sampled spike trains of these 28 tremor cells was carried out over sequential 10-s time segments, and autocorrelograms showing maximal oscillatory activity were graded from 0 to 10. Average tremor cell oscillation grades ranged from 6.8 to 7.8, similar to those reported in the MPTP-induced primate model of parkinsonism. The average tremor cell oscillation grade varied between patients, as did the clinical measures of tremor severity. Tremor cells had oscillations in spike discharges at the same average frequency (4.2–5.2 Hz) as the patient’s tremor determined from the electromyogram and accelerometry records of one or more limbs (4.0–5.4 Hz), and the individual values were correlated (r2=0.73) over the total range (3.7–5.6 Hz). The results of this study demonstrate the presence of neurons with 4–6 Hz tremor-frequency activity in GPi, supporting a role of the globus pallidus in the production of rest tremor in PD patients.

Long-term outcome of bilateral pallidal deep brain stimulation for primary cervical dystonia.

Ten patients with severe cervical dystonia (CD) unresponsive to medical treatment underwent bilateral globus pallidus internus (GPi) deep brain stimulation (DBS) and were followed for 31.9 ± 20.9 months. At last follow-up, the Toronto Western Spasmodic Torticollis Rating Scale (TWSTRS) severity score improved by 54.8%, the TWSTRS disability score improved by 59.1%, and the TWSTRS pain score improved by 50.4%. Bilateral GPi DBS is an effective long-term therapy in patients with CD.

Does stimulation of the GPi control dyskinesia by activating inhibitory axons

A 69‐year‐old woman with Parkinsons disease and levodopa‐induced dyskinesias had a deep brain stimulation (DBS) electrode inserted into the right globus pallidus internus (GPi). During the operation, the GPi was mapped with dual microelectrode recordings. Stimulation through one microelectrode in GPi inhibited the firing of GPi neurons recorded with another microelectrode 600–1,000 μm distant. The inhibition could be obtained with pulse widths of 150 μs and intensities as low as 10 μA. Single stimuli inhibited GPi neurons for ∼50 ms. Trains of 300 Hz stimuli inhibited GPi neuron firing almost completely. Postoperatively, stimulation through macroelectrode contacts located in the posterior ventral pallidum controlled the patients dyskinesias. The effect could be obtained with pulse widths of 50 μs and frequencies as low as 70–80 Hz. We postulate stimulation of the ventral pallidum controls dyskinesias by activating large axons which inhibit GPi neurons.

A comparison of the burst activity of lateral thalamic neurons in chronic pain and non-pain patients

Thalamic neurons are known to switch their firing from a tonic pattern during wakefulness to a bursting pattern during sleep. Several studies have described the existence of bursting activity in awake chronic pain patients and have suggested that this activity is abnormal and may be related to their pain. However, we have frequently observed bursting activity in awake non-pain patients suggesting that there may not be a causal relationship between thalamic bursting activity and chronic pain. To examine this issue more rigorously we compared the incidence and pattern of bursting activity of lateral thalamic neurons of both pain and non-pain patients in a state of wakefulness. Recordings were obtained from lateral thalamic areas of different groups of patients (n = 91) suffering from pain disorders (e.g. anaesthesia dolorosa, phantom limb pain, trigeminal neuralgia, post-stroke pain) and motor disorders (e.g. Parkinsons disease, essential tremor) during stereotactic surgical procedures for the treatment of pain and movement disorders. Burst indices (the number of bursting cells per electrode track) were computed for all the explorations in the two groups. The burst indices in the pain and non-pain groups (1.73 +/- 0.28 and 1.14 +/- 0.16, respectively) were not significantly different from each other. The bursts were analyzed to see if they fulfilled the criteria of low-threshold calcium spike (LTS)-evoked bursts characterized by (i) a shortening of the first interspike interval with an increase in the number of interspike intervals in the burst and also (ii) a progressive prolongation of successive interspike intervals. LTS-evoked bursts were identified in 27/47 (57%) bursting cells in pain patients and 15/32 (47%) cells in non-pain patients. These data demonstrate that the occurrence of bursting activity and of LTS-evoked bursts in the human thalamus is prevalent in both pain and non-pain patients. This suggests that the bursting activity of thalamic neurons in pain patients is not necessarily related to the occurrence of their pain.

Increased Gamma Oscillatory Activity in the Subthalamic Nucleus During Tremor in Parkinson's Disease Patients

Rest tremor is one of the main symptoms in Parkinsons disease (PD), although in contrast to rigidity and akinesia, the severity of the tremor does not correlate well with the degree of dopamine deficiency or the progression of the disease. Studies suggest that akinesia in PD patients is related to abnormal increased beta (15-30 Hz) and decreased gamma (35-80 Hz) synchronous oscillatory activity in the basal ganglia. Here we investigated the dynamics of oscillatory activity in the subthalamic nucleus (STN) during tremor. We used two adjacent microelectrodes to simultaneously record neuronal firing and local field potential (LFP) activity in nine PD patients who exhibited resting tremor during functional neurosurgery. We found that neurons exhibiting oscillatory activity at tremor frequency are located in the dorsal region of STN, where neurons with beta oscillatory activity are observed, and that their activity is coherent with LFP oscillations in the beta frequency range. Interestingly, in 85% of the 58 sites examined, the LFP exhibited increased oscillatory activity in the low gamma frequency range (35-55 Hz) during periods with stronger tremor. Furthermore, in 17 of 26 cases where two LFPs were recorded simultaneously, their coherence in the gamma range increased with increased tremor. When averaged across subjects, the ratio of the beta to gamma coherence was significantly lower in periods with stronger tremor compared with periods of no or weak tremor. These results suggest that resting tremor in PD is associated with an altered balance between beta and gamma oscillations in the motor circuits of STN.

New developments in understanding the etiology of Parkinson's disease and in its treatment

Important recent advances have been made in understanding the etiology and pathogenesis of Parkinsons disease, as well as in developing novel treatments. Two newly identified genes, alpha-synuclein and parkin, have been linked to parkinsonism. In addition, disturbances to the normal basal ganglia circuits in Parkinsons patients are being described at both anatomical and physiological levels. These developments provide a strong scientific basis for novel medical and surgical strategies to treat the profound motor disturbances in patients with Parkinsons disease.

Immediate Motor Effects of Stimulation through Electrodes Implanted in the Human Globus Pallidus

The immediate motor effects of stimulation through electrodes chronically implanted in the globus pallidus internus (GPI) were studied in 9 subjects with Parkinson’s disease. Single stimuli (at ≫0.4 Hz) produced short latency facilitation of voluntarily activated contralateral muscles in all subjects. The latency and distribution of the facilitation, its probably monosynaptic nature, and the short chronaxie and refractory period of the activated neural elements suggest that the facilitation results from the direct excitation of the fast conducting corticospinal pathway. The facilitation of motoneurons followed high frequency (e.g. 200 Hz) stimulation without decrement and occurred at stimulus intensities well below those required to produce a visible muscle contraction. We conclude that, while there may be other effects, GPI stimulation through electrodes may activate the corticospinal tract, even when the stimuli are below the threshold for a visible muscle contraction, and that continuous stimulation may do so continuously. This may be an unwanted side effect, but possible therapeutic actions are considered. The reproducible short latency facilitation enabled us to estimate current spread from the quadripolar electrodes used for deep brain stimulation. When the current is sufficient to excite large myelinated fibers near one of the quadripolar electrodes, an additional 1-mA current will activate similar fibers at an additional distance of 1.8 mm with bipolar stimulation and at a distance of 5.7 mm with monopolar stimulation.

Computerized Tomography (CT) Is Just as Accurate as Ventriculography for Functional Stereotactic Thalamotomy

Eighty-three consecutive functional stereotactic procedures (56 thalamotomies, 1 medial thalamotomy, and 26 chronic brain stimulatory electrode introductions) were done using CT to identify the three-dimensional coordinates of anterior and posterior commissures. The three-dimensional locations of the tactile relay in ventrobasal complex for manual digits were then determined as the first step in physiological corroboration of target site using single-cell recordings and microstimulation. The measured location of this structure was then compared with that predicted by the Schaltenbrand and Bailey atlas. There was no discrepancy in the mediolateral plane in 62.7%, in the dorsoventral plane in 63.9%, and in the anteroposterior plane in 44.6% of the cases. Over 2 mm deviation occurred in 10.8, 12.0, and 19.2% of the cases in these three planes, respectively. This precision of localization is better than that reported with ventriculography. Many of the larger discrepancies occurred in patients who had suffered from stroke, multiple sclerosis, severe head injury, or after craniotomy.

Plasticity in Human Somatosensory Thalamus as a Result of Deafferentation

Experimental studies indicate that deafferentation results in reorganization of the somatosensory map at various levels of the CNS, such that the representation of a body part adjacent to a region that is denervated expands into the deafferented area. Recent data suggest that in the human this occurs at the cortical level, but subcortical structures have not been systematically investigated. To test the hypothesis that the human thalamus is capable of significant reorganization as a result of changes in afferent input, microelectrode recording and stimulating techniques were used to define thalamic somatotopy in 61 patients undergoing stereotactic procedures. Five groups were compared: those with pain in the deafferented body part, face (n = 9), arm/hand (n = 4), leg/foot (n = 8) and hemibody (n = 5) and those with neither pain nor deafferentation, i.e., movement disorder (n = 24). Trunk representation, as determined from receptive fields, was significantly larger in patients with leg/foot deafferentation than in patients without deafferentation (1.8 +/- 0.7 vs. 0.5 +/- 0.2 mm; p < 0.01). Also, microstimulation induced paraesthesiae in the face from a significantly larger region of thalamus in the facially denervated group compared to the movement disorder group (13.8 +/- 2.8 vs. 3.7 +/- 0.6 mm; p < 0.001). There were no significant differences in the representation of other body parts in the five groups. The results in the leg-deafferented group agree with conclusions reached from animal studies; however, the human situation is more complex. There appear to be different patterns and degrees of somatotopic reorganization in the human, all of which may be associated with pain syndromes.