Oct 5, 1993 - Palhdal Neurons in Palanson's Disease. Djordje Sterio, MD, DSc,* Aleksandar BeriC, MD, DSc,* Michael Dogali, MD,? Enrico Fazzini, DO, PhD ...
Neurophysiological Properties of Palhdal Neurons in Palanson’s Disease Djordje Sterio, MD, DSc,* Aleksandar BeriC, MD, DSc,* Michael Dogali, MD,? Enrico Fazzini, DO, PhD,* George Alfaro, PhDJ and Orrin Devinsky, MD”
Neuronal properties of the human globus pallidus (GP) are not known. Since G P is the major output of the basal ganglia, it may be involved in the pathophysiology of Parkinson’s disease. We studied 12 patients with medically resistant Parkinson’s disease by using single cell recording of rhe GP during stereotaxic pallidotomy to define neuronal firing rate and its modulation during active and passive movements. Different frequency and pattern of single cell activity was found in globus pallidus externus compared with globus pallidus internus. Discharge rates of 19% of GP cells were modulated by passive contralateral movements. Pallidal units were most often related solely to single joint movement. Different patterns of activity in relation to the two different movements of the same joint were often observed. We identified somatotopically arranged cell clusters that alter discharge rate with related movements. These findings suggest at least a partial somatotopic organization of the human GP and similarity with experimental results in both healthy and MFTP monkeys, providing a rationale for surgical or pharmacological targeting of G P for treating Parkinson’s disease. Sterio D, BeriC A, Dogali M, Fazzini E, Alfaro G, Devinsky 0. Neurophysiological properties of pallidal neurons in Parkinson’s disease. Ann Neurol 1994;35:586-591
Approximately 25 years ago the treatment of Parkinson’s disease (PD) was revolutionized by the introduction of L-dopa. However, in most patients, L-dopa therapy becomes less effective over a period of 2 to 10 years. Those patients require increasing doses and suffer a higher frequency of disabling side effects such as hallucinations, agitation, and dyskinesia. Despite the introduction of better-tolerated preparations of [.-dopa and new medications, many Parkinson’s patients, especially those affected for more than 5 to 10 years, are severely disabled with poor quality of life. In selected patients with medically refractory PD, surgical therapy may be beneficial. Although renewed interest in surgical therapy has focused on fetal-adrenal transplants [ 1-41 interest in pallidotomy has also been rekindled. h t i n e n and colleagues [ 5 } reported 66 patients with medically refractory PD in whom pallidotomy significantly improved bradykinesia, akinesia, and rigidity. Preliminary experience with a pallidotomy procedure at our center in more than 25 patients supports these findings [6]. Recent advances in imaging techniques, such as highresolution magnetic resonance imaging (MRI) with stereotaxic computer systems for preoperative planning, enable increased accuracy for target localization and lesion placement. Other technical advances in microelectrode design and neurophysiological recordings
provide a unique opportunity to srudy electrical discharges of single units during stereotaxic surgery. These methods have been used mainly during stereotaxic thalamic surgery where destructive lesions are made to treat parkinsonian tremor or to place stirnulating electrodes for management of central pain disorders { 7- lo}. These recordings provide neurophysiological confirmation of anatomical localization and provide important insights into physiology and pathophysiology of these subcortical areas. The neurophysiology of the human globus pallidus (GP) remains unstudied. The issues that r e m i n unidentified in healthy subjects and in PD patients include cellular relationships and correlations, physiological and pathophysiological cellular properties, such as basic neuronal firing rate and changes in relation to the thought of movement, initiation of movement, and passive and active joint movements. Currently, no data can define precisely the optimal pallidal surgical site that yields maximal efficacy and minimal complications [I 1). Results of single cell recording may provide necessary insight into possible lesioning sites and the type of abnormality, enabling choice of the most appropriate size and shape of the lesion. We recorded and analyzed single unit activity in the external and internal segments of the human GP in
From rhc ’Deparrment of Neurology, New York University School of Medicine. Hospital for Joint Diseases, +Department of Neurosurgery, Division of Functional arid Stereoractic Neurosurgery, New York University Medical Center, a i d $Department of Neurology, Hospital for Joint Diseases, New York, NY.
Received Jul 21, 1993. Accepred for publication Oct 5 , 1993.
586
Address corrcspondence to Dr Berit, Department of Neurology, Box 65, Hospital for Joint Diseases, SO1 East 17th Street, New York, NY 10003.
Copyright 0 1994 by the American Neurological Association
conscious, nonanesthetized Parkinson’s patients during the relaxed state and during voluntary and involuntary movements. We studied t h e neuronal events during preparation and performance of voluntary movements and during passive movements, t o map t h e neuronal topography of t h e human GP based on firing patterns of single units during different m o t o r functions.
Methods Twelve patients with PD, initially responsive to L-dopa, were studied. Patient selection was based on the presence of bradykinesia and rigidity, rather than tremor, as predominant symptoms. All patients had poor control of their disease with dopaminergic drugs either because of decreased efficacy or severe fluctuations of odoff symptoms (chorea-dystonidbradykinesia-rigidity-tremor-impaired ambulation). In addition, patients with an initial poor response to L-dopa, who have severe lower body bradykinesia despite upper extremity dyskinesia when in the “on” state were included. Patients were assigned for the pallidotomy by an advisory board consisting of two neurologists and one neurosurgeon. Protocol approval was obtained from the Institutional Review Board at the Hospital for Joint Diseases-New York University Medical Center. Informed consent for surgical and experimental procedures was obtained. The pallidal target was defined with high-resolution nuclear MRI technique. The frame was positioned with reference to the midpoint of the anterior commissure (ACjiposterior commissure (PC) line. Acquired data were digitized onto nine-track tape and transferred to an independent workstation (CASSiMidco), which allows correlations of fiducials in space, application of the most appropriate Schaltenbrand brain map, and preoperative planning of the target trajectories and target sites. The general anatomical target lies between 18 and 25 mm lateral to the midline, 4 to 6 mm below the AC/PC line, and 2 to 3 mm anterior to midcommissural point. In all patients unilateral single cell recording and pallidotomy were performed, under local anesthesia, requiring patient’s cooperation during the surgical procedure. A protective cannula, fixed to the stereotaxic frame, guided the microelectrode through a twist drill hole, along the distance to the point 10 mm above the target point. A microdrive, providing micrometer-graded extrusion of the microelectrode tip from the cannula, was employed. One to three electrode trajectories were performed for each patient. Single unit activity was recorded extracellularly with specially designed, tungsten-tip, disposable microelectrodes (1 mm diameter at the tip, 1-2 Mi1 impedance at 1,000 Hz). The guiding cannula was the reference electrode. Extracellular action potentials were amplified with an AC amplifier (DAM-80, WPI) and simultaneously recorded using standard recording techniques ( - 6 dB at 300 and 10,000 Hz), together with a descriptive voice channel on the magnetic tape. Rough computer “on-line” analysis was done as well. When the microelectrode tip was in the GP and cell firings identified, a battery of functional tests was employed. The patient was asked to expose the teeth, fledextend the elbow, fledextend the wrist, move the fingers, fledextend the knee and hip, and fledextend the ankle. Passive limb movements of elbow, wrist, fingers, ankle, hip, and knee were also stud-
ied. For each pallidal unit recorded, the same combination of tests have been used. The data obtained before, during, and after performance of the tests were analyzed. Along the microelectrode trajectories, the localization of neuronal populations responding to test stimuli was determined. These zones were placed in a stereotaxic triangular system, determined on the basis of MRI technique. After completing the procedure, single unit data from the tape were analyzed. Units were classified according to their relation to various movements. Spontaneous and evoked spike activities were analyzed using a computer data processing system (MacLab 4 ; Macintosh IIcx; dual-window discriminator, SA Instruments, CA). The following functions were estimated: firing rateitime histograms, interspike and interburst interval histograms. Pearson’s x2 statistics were used to compare the presence of finger movement-related cells in left versus right GP.
Results T h e activity of 419 cells, b o t h from GP externus (GPe) and GP internus (GPi), was recorded (298 cells from GPi and 1 2 1 cells from GPe). Only the spikes that were clearly separable from t h e background noise at a ratio greater than 2 : 1 were analyzed. Average spontaneous activity for GPe was 4 3 . 5 0 (SD 9.12) spikes per second and for G P i neurons 58.52 (SD 6.31) spikes p e r second. The majority of units showed spikes with a mainly negative waveform and with short duration (0.3-0.6 msec). The patterns of ongoing discharge of pallidal neurons were stable over relatively long periods. N e u r o n s in GPe exhibited mainly two patterns, i.e., (1) high-frequency, long-lasting bursts interrupted by pauses and (2) low frequency of discharge with occasional, brief, high-frequency bursts. The long-lasting bursts occurred with marked regularity in s o m e neurons. More often, however, t h e rhythm of these bursts varied. Up to 25 spikes were grouped within t h e bursts at firing rates reaching 500 t o 700/sec. The duration of the silent period between t h e bursts varied from 30 t o 250 msec. The slow rhythm of bursts (at 3-7/sec) occurred in very few situations with patients showing tremor. Cells in G P i had mostly a characteristic, sustained, but irregular, fluctuating high-discharge rate. Figure 1 shows the typical discharge patterns observed in both GPe and G P i . The discharge of 19% of cells i n both pallidal segments was clearly modulated during passive movements of individual body parts. Among these cells, 2 0 9 w e r e nonspecifically activated during different limb movements. Ninety-two percent of movement-related cells increased their activity and 85% decreased the activity during limb movements. The periods of 500 to 1,000 msec were analyzed during separate limb movements. T h e cell was considered as reacting if the change of activity was greater than 40%; and if the change was present during all three movements of t h e same limb. During movements of contralateral limbs, the units Sterio ct al: Pallidal Neurons in Parkinson’s Disease
587
a
b
C
d
e
f
200 ms Fig I . Dzj'firent &barge
patterns of relli registered in both
globin pallidus externus (GPei and globus pallidus internus
IGPil. la. bi Tuv tyj%s of cel1 dhcharges characteristicfor GPe neurons (a: cell S- 15 18 = bursting actiiity; b: cell B-225 7 = Lou frequency. iwcgular actizdy). (c, di High-frequeny, irreguIas diicbarges L-baracteriJticfor GPi neurons. (e) Artici6y of mo?~ement-wlatedcell (F-71.34) during hip and Knee movementJ. actioation during fiexion and inhibition during extension. i f ) Activity of movement-related cell !F-8868. same patient, same trajectory, l .7mm below the cel.! F-7134) during hip and knee movements; inhibition during fiexion and activation during extrmion. Thick horizontal line = flexion; thin horizontal line
=
extevzsion.
consistently showed changes in discharge frequency in temporal relation to movement cycles. During ipsilateral limb movements, however, the discharge of the same unit showed no consistent relation to movements. A small number of units ( 2 % ) discharged nearly equally in reiation to ipsilateral and contralateral limb movements. The relationship of unit discharge to the different phases of the movement varied from one unit to another; i.e., some discharged most intensely during 588 Annals of Neurology Vol 35
No 5
May 1994
flexion and others during extension. Typically, both increases and decreases in discharge rate over resting levels were observed in reiation to particuiar phases of the movement. Changes in discharge pattern of most reacting cells were related to direction and amplitude of movement. The relations between movement direction and the frequency of cell discharge were studied in 82 cells. More pallidal units responded to joint movements of the lower ( 5 8 % ) than upper ( 4 2 % ) limbs. Figure l e and f display changes in both pattern and frequency of cell discharges during hip and knee flexion and extension. The cell (e) had irregular, almost bursting, highfrequency discharge pattern during flexion. With extension this activity was markedly altered into irregular, single spike, low-frequency discharges. The second cell (f), located 1.5 mm below the first cell, in the same patient, showed a different pattern of activity. During flexion, activity was mainly reduced, while during extension, irregular, high-frequency discharges occurred. Movement-related cells were somatotopically organized in the G P (Fig 2A,B). In both GPi and GPe, clusters of the neurons reacting clearly during passive joint movements were found. Cell clusters showing altered discharge patterns during passive movements were demonstrated for the finger, hip, knee, and ankle. In contrast elbow and lip movements were more widely represented. One to three different trajectories were used in the same patient. The trajectories differed in anterior-posterior (AP) angle of penetration. AP angle varied from 28 to 60 degrees. In most of the patients, a larger number of movement-related cells in trajectories ranging from 40 to 5 0 degrees were found. More movement-related cells were found in GPi than in GPe. In 5 of 12 right-handed patients we found finger movement-responsive cells. Those cells were identified in 4 of 6 left GP (13 finger-responsive cells in 6 left G P ) and 1 of 6 right G P (only 1 finger-responsive cell in 6 right GP). The difference is statistically significant (xz = 10.26, p = 0.0035). Discussion Our recordings of more than 400 single pallidal cells in 12 P D patients, the first such study, reveal differences between the two pallidal segments and evidence of a somatotopic organization. GPe single cell activity had lower frequency and was more irregular than in GPi. This finding is consistent with the primate studies of DeLong [12), which first demonstrated distinct discharge patterns between cells in the two pallidal segments. This finding has been more recently confirmed [13, 141. In primates, this difference is so distinct that it was possible to determine when the microelectrode entered and when it left each segment of GP. The observed differences in discharge patterns between
A
POSTERIOR
ANTERIOR
Fig 2. Location of globus pallidus (GPI cells responding t o contralateral dzyerent body parts movement in coronal (3 mm anterior fmm midconzmissural line) and sagittal (22 mm lateral from interhemispheric line) planes. (A)Responses from ldt GP in 6 patients. (Bj Responses from right GP in another 6 patients. GPe = globus pallidus externus; GPi = globus pallidus internus; ICL = anterior-posterior intercommis.sural line; (J) = fingers; ( +) = wrist; (#) = elbow; (@) = lips; (0)= ankle; (a) = hip und knee.
GPe and GPi cells are consistent with different anatomical connections and different functional roles of those two segments [ljl. We observed differences between the pallidal segments in PD patients, but the differences were not as great as those reported in primates. This could reflect interspecies differences but is most likely because OUT recordings were obtained in PD patients rather than healthy subjects. Support for this explanation comes from primate studies comparing GPi neuronal activ-
B
POSTERIOR
I
ANTERIOR
ity in normal animals and in animals with l-methyl4-phenyl-l,2,3,6-tetrahydropyridine(MPTP)-induced parkinsonism 113, 161. Firing rates of GPi neurons in MPTP-treated monkeys were up to 50% higher than those of GPi neurons in healthy monkeys. Also, the tonic discharge characteristic of GPi neurons in intact animals was transformed after MPTP treatment into a burst-silence pattern. These findings suggest that GPi activity in METP monkeys became similar to GPe activity found in normal animals. We did not find any significant correlation between the discharge pattern of either GPi or GPe units and patient tremor. Similarly, in primate experiments, no correlation between discharge pattern and tremor was found [17]. In contrast to the GP, in PD patients, Lenz and collaborators { 181 demonstrated that the spike trains of many thalamic cells in the ventral nuclear group had a major power component at tremor
Sterio et d: Pallidal Neurons in Parkinson's Disease
589
frequency. Other investigators reported tremorfrequency activity of thalamic cells, too [ 7 , 19, 20). The response of neurons in the human central nervous system to joint movement is of particular interest because some of these cells may generate or transmit information critical for movement and position of various body parts. Nineteen percent of G P cells in our patients were movement related. In primate experiments a wide range, 20 to 50%, of movement-related pallidal cells were identified 112, 21, 22). Generally, more responding cells were found in pathological experimental situations such as in MPTP-treated animals 1131. Nambu and colleagues [23} categorized primate pallidal cells into the following three groups based on electrophysiological responsiveness: ( 1) light-related neurons, (2) delay-related neurons, and (3) movementrelated neurons. In primate experiments, when active limb movements were applied with conditioning and other behavioral tasks, a higher percentage of movement-related cells were found 124). The movement-related cells that we identified probably represent only a fraction of potentially reacting units. More movement-related cells would likely be identified if more sophisticated recording and averaging techniques or behavioral paradigms were employed. We were unable to employ extensive protocol with techniques such as conditioning because of time limitations in the operating room and potential patient discomfort. The relationship of G P unit activity to movement appears specific, since units were most often related solely to single joint movement and different activity patterns occurred in relation to the two different movements of the same joint (i.e., flexion and extension). Twenty percent of human movement-related cells were nonspecifically activated, i.e., their discharge patterns changed in relation with movement of two and more joints. The percentage of nonspecifically activated cells is higher in MFTP monkeys than in normal animals 1131. Some nonspecifically activated cells that we recorded probably generalized their activity pattern due to pathophysiological changes associated with PD. We found clusters of neurons reacting clearly during hip and knee, ankle and finger movements in both GPi and GPe. More movement-related cells were found in GPi than in GPe; however, this finding might be artifactual since lesions were made in GPi, where most of our mapping was directed. In primates, Iansek and Porter [21) found that movement-related neurons are regionally organized and located mainly in posterior parts of the GP. Movement-related cells were clustered, separated by areas of nonresponding cells. Dehng 1121 found that in primates most movementreacting cells were localized in ventral and lateral portions of the GP. Filion and associates [13] found that in both normal and parkinsonian monkeys, movement-
590
Annals of Neurology Vol 35
No 5
May 1994
related neurons were located mostly in central parts of the GP. In primates movement-related cells were found in different GP areas possibly reflecting differences in the methodologies used. The identification of finger movement-responsive neurons in the left but not right G P suggests possible hemispheric specialization, with greater representation of the dominant hand. In our right-handed patients, finger movement responsive cells were identified in 4 of 6 left G P and 1 of 6 right GP. The difference may reflect small sample size, but a hemispheric specialization is a plausible explanation. Additional studies are, however, needed to determine if there is a difference between the representation of finger movements in the left and right human GP. Subcortical specialization has been observed for human language and attentional functions. Left thalamic and striatal lesions cause aphasia 125-271, while right thalamic and striatal lesions cause left-sided neglect and dysprosody 128-3 1). In conclusion, (1) both different frequency and different pattern of single cell activity were found in two pallidal segments, GPe and GPi; (2) the discharge of 19% of cells in both pallidal segments was clearly modulated during passive movements of individual body parts; and (3) the cell clusters that alter discharge rates, a result of related movements, were identified. These findings suggest at least a partial somatotopic organization of the human G P and similarity with experimental results in both normal and MFTP monkeys. If the concept that the GP is hyperactive in PD proves valid, identification of cells and their abnormalities would help in providing the optimal target for very localized lesions of the GP and result in predictable improvement of PD symptoms. We thank Jonathan 0. Dostrovsky, PhD, for initial support and technical advice, Fred Kummer, PhD, for technical assistance, and Dale Samelson RN, MA, for target analysis support.
References 1. Lindvall 0, Backlund EO, F a d e L, et al. Transplantation in Parkinson’s disease: m o cases of adrenal medullary grafts to the putamen. Ann Neurol 1987;22:475-486 2. Lindvall 0, Rehncrona S, Brundin P, et al. Human fetal dopamine neurons grafted into the striatum in two patients with severe Parkinson’s disease. Arch Neurol 1989;46:615-63 1 3. Bakay RAE, Watts U, Freeman A, et al. Preliminary report on adrenal-brain transplantation for parkinsonism in man. Srereotact Punct Neurasurg 199O;>4-55:3 12-323 4 . Kordower JH, Cochran E, Penn RD, Goetz CG. Putative chromaffin cell survival and enhanced host-derived TH-fiber innervation following a functional adrenal medulla autvgrafr for Parkinson’s disease. Ann Neurol 1991;29:405-412 5 . h t i n e n LV, Bergenheim AT, Hariz MI. Leksell’s posteroventral pallidotomy in the treatment of Parkinson’s disease. J Neurosurg 1992;76:53-61 6. Fazzini E, Dogali M, Eidelberg J, e t al. Long-term follow-up on patients with Parkinson’s disease receiving unilateral ventroposterior medial pallidotomy. Neurology 1993;43(suppl):S27I
7. Gaze RM, Gillingham FJ, Kalyanaraman S, et al. Microelectrode recordings from the human thalamus. Brain 1964;87:691-706 8. Jasper HH, Bertrand G. Thalamic units involved in somatic sensation and voluntary and involuntary movements in man. In: Purpura DP, Yahr MD, eds. The thalamus. New York: Columbia University Press, 1966:365-390 9. Albe-Fessard D, Afrel G, Derome P, Guilbaud G. Thalamic unit activity in man. Electroencephalogr Clin Neurophysiol 1967;25(suppl):132-142 10. Ohye C, Narabayashi H . Activity of thalamic neurons and their receptive fields in different functional states in man. In: Somjen G, ed. Neurophysiology studied in man. Amsterdam: Excerpta Medica, 1972:79-84 11. Bakay RAE, DeLong MR, Vitek JL. Posteroventral pallidotomy for Parkinson’s disease. J Neurosurg 1992;77:487-488 12. DeLong MR. Activity of pallidal neurons during movement. J Neurophysiol 1971;34:414-427 13. Filion M, Tremblay L, Bedard PJ. Abnormal influences of passive limb movement on the activity of globus pallidus neurons in parkinsonian monkeys. Brain Res 1988;444:165-176 14. Richardson RT, DeLong MR. Electrophysiologicalstudies of the functions of the nucleus basalis in primates. In: Napier TC, Kalivas PW, Hanin I, eds. The basal forebrain. New York: Plenum Press, 1991:233-252 15. DeLong MR, Alexander GE, Mitchel SJ, Richardson RT. The contribution of basal ganglia to limb control. Prog Brain Res 1986;64:161-174 16. Filion M. Tremblay L, Bedard PJ. Effects of dopamine agonists on the spontaneous activity of globus pallidus neurons in monkeys with MPTP-induced parkinsonism. Bran Res 1991;547: 152-161 17. Pare D, Curro’Dossi R, Steriade M. Neuronal basis of the parkinsonian resting tremor: a hypothesis and its implications for treatment. Neuroscience 1990;35:217-226 18. Lenz FA, Tasker RR, Kwan HC, et al. Single unit analysis of the human ventral thalamic nuclear group: correlation of thalamic “tremor cells” with the 3-6 Hz component of Parkinsonian tremor. J Neurosci 1988;8:754-764
19. Ohye C , Saito Y , Fukamachi A, Narabayashi H. An analysis of the spontaneous rhythmic and non-rhythmic burst dischargcs in the human thalamus. J Neurol Sci 1974;22:245-259 20. Ohye C, Albe-Fessard D. Rhythmic discharges related to tremor in humans and monkeys. In: Chalazonitis N, Boisson M, eds. Abnormal neuronal discharges. New York: Raven Press, 1978: 37-48 2 1. Iansek R, Porter R. The monkey globus pallidus: neuronal discharge properties in relation to movement. J Physiol 1980;301: 439-455 22. H m a d a I, DeLong MR, Mano NI. Activity of identified wristrelated pallidal neurons during step and ramp wrist movements in the monkey. J Neurophysiol 1990;64:1892-1906 23. Nambu A, Yoshida S, Jinnai K. Projection on the motor cortex of thalamic neuron with pallidal input in the monkey. Exp Brain Res 1988;71:658-662 24. Brotchie P, Iansek R, Horne MK. Motor function of the monkey globus pallidus. Brain 1991;114:1667-1681 25. Cappa SF, Vignolo LA. Transcortical features of aphasia following left thalamic hemorrhage. Cortex 1979;15:121-130 26. Damasio AR, Damasio H , Rizzo M, et al. Aphasia with nonhemorrhagic lesions in the basal ganglia and internal capsule. Arch Neurol 1482;39:501-506 27. Naeser MA, Alexander MP, Levinc HL, et al. Aphasia with predominantly subcortical lesion sites. Arch Neurol 1982;39: 2-14 28. Damasio AR, Damasio H, Chui HC. Neglect following damage to frontal lobe or basal ganglia. Neuropsychologia 1980;18: 123-112 29. Watson RT, Valenstein E, Heilrnan KM. Thalamic neglect: possible role of the medial thalamus and nucleus rrticularis in behavior. Arch Neurol 1981;38:501-506 30. Ferro J, Kertesz A, Black SE. Subcortical neglect: quantification, anatomy, and recovery. Neurology 1987;37:1487- 1492 31. Wolfe GI, Ross ED. Sensory aprosodia with left hemiparesis from subcortical infarction: right hemisphere analogue of sensory-type aphasia with right hemiparesis. Arch Neurol 1987;44: 668-67 1
Sterio et al: Pallidal Neurons in Parkinson’s Disease
591
