Deep brain stimulation for movement disorders
Correspondence Address: Source of Support: None, Conflict of Interest: None DOI: 10.4103/0028-3886.226438
Source of Support: None, Conflict of Interest: None
Keywords: Deep Brain Stimulation, dystonia, movement disorders, Parkinson's disease
The modern era of neurostimulation for movement disorders was ushered in by Brice and McLellan, who reported the efficacy of midbrain and basal ganglia stimulation to control tremor in multiple sclerosis patients. This was followed by Benabid's report of control of Parkinsonian tremor by thalamic stimulation., Over the past four decades since its inception, therapeutic neurostimulation has witnessed several advances, including newer indications and targets, and refinement of technology, enabling better control of symptoms and lesser adverse effects. Parkinson's disease (PD) continues to be the commonest indication for deep brain stimulation (DBS); indications outside movement disorders (like treatment-resistant depression, obsessive compulsive disorder, refractory epilepsy and Alzheimer's disease) are also emerging. This review focuses on the possible mechanisms by which DBS act, the indications and surgical targets, techniques and long-term outcome in patients with movement disorders.
The movement disorders which benefit from DBS arise from dysfunction in neuronal circuits constituted by cortical and subcortical components - motor basal ganglia circuits in the case of dystonia and PD, and the cerebello-thalamo-cortical circuits in Parkinsonian tremor and essential tremor (ET). The motor circuits of the basal ganglia are involved in the motivation and scaling of movements, selection of appropriate actions and simultaneous inhibition of unwanted ones.,, The cortical inputs to the basal ganglia (from the motor and premotor cortices as well as the supplementary motor and cingulate motor areas) reach the striatum and the output is funneled through the internal globus pallidus (GPi) and the functionally related pars reticulata of the substantia nigra (SNr) to the cortex; basal ganglia output also projects to brainstem centers like the pedunculopontine nucleus (PPN). The most popular one among the basal ganglia circuit models consist of a monosynaptic 'direct' striatum to GPi/SNr pathway; as well as the indirect pathway which channels through the external globus pallidus (Gpe) and the subthalamic nucleus (STN) to the GPi/SNr. In this model, the GPi exerts an inhibitory control over the thalamocortical excitation. The direct pathway, by inhibiting the GPi, facilitates thalamocortical excitation and the indirect pathway has an opposite effect. The final effect is selection and scaling of appropriate motor programs for the desired actions and inhibition of the unwanted ones. A third, 'hyperdirect' pathway from the cortex to the STN also exists and its influence on thalamo-cortical relay is similar to that of the indirect pathway except that it is faster as it bypasses the striatum and is involved in the preparatory step for the selection of appropriate movement.
Striatal dopamine released from nigro-striatal terminals acts as a neuromodulator, and by facilitating the cortico-striatal inputs in the direct pathway and inhibiting the indirect pathway, has an inhibitory effect on Gpi/SNr, thereby ultimately facilitating the thalamic excitation of the cortex. In this model, nigrostriatal dopaminergic cell loss in PD results in underactivity of the direct, and overactivity of the indirect pathways, thereby resulting in GPi/SNr overactivity and excessive inhibition of the thalamic excitation of cortex.
Concordant with this model, high frequency stimulation of the STN or the GPi was believed to act similar to a surgically created lesion in PD (pallidotomy/subthalamotomy) by inhibiting the output from the target nuclei. Several mechanisms were hypothesized to explain the inhibitory effects of high frequency stimulation, like inactivation of voltage gated channels, depolarization block and stimulation of the afferent inhibitory (GABAergic) terminals in the target nucleus resulting in release of gama aminobutyric acid (GABA) and inhibition.,, Though the direct and indirect pathway model could explain the pathophysiology underlying some of the movement disorders and was to a certain extent supported by experimental evidence, some paradoxes observed in clinical practice indicated that the model was an oversimplification. For example, the model predicts that pallidal lesioning should relieve Parkinsonism and result in a hyperkinetic state, while in reality, it does not result in hyperkinesia; on the contrary, pallidotomy/pallidal DBS is a very effective treatment to relieve the dyskinesias in PD. Secondly, thalamic lesions are expected to cause Parkinsonism, which is not seen in clinical practice or following experimental lesions made in animals., Functional imaging studies also indicated that the neural output from the target nuclei are actually increased, contrary to what is expected if DBS were to act by inhibiting the target nuclei.
Electrophysiological studies and local field potential measurements in patients and experimental animals have proven that Parkinsonian state is associated with abnormal bursting and hypersynchronous activity in the nodes of the motor circuit (like the GPi and the STN). Oscillations in 10-25Hz (Beta band) occur in an uncontrolled Parkinsonian state and are replaced by the more desynchronized Gamma band (60-80 Hz) oscillations on treatment with levodopa, accompanied by relief of symptoms., Based on such electrophysiological and clinical observations and computational models, the firing pattern of basal ganglia neurons is currently believed to be as important or more important than the firing rate, in shaping the final motor output. The beta band oscillations correlate with rigidity and bradykinesia but not tremor; the latter symptom in PD does not correlate with the severity of nigrostriatal dopaminergic denervation unlike the other two. Abnormalities of the cerebellothalamocortical circuit are implicated in tremor; tremor responds well to thalamic DBS which modulates this circuit. The altered patterns of firing and abnormal levels of synchrony in various basal ganglia disorders lead to abnormal information flow in the circuits, finally leading to a pathological alteration of functions in the thalamocortical and brainstem networks to which the basal ganglia output project. The high frequency stimulation of the nodes in the circuit achieved by DBS, rather than inhibiting the output from the target, disrupts the pathological synchrony and thereby the outflow of abnormal information from the basal ganglia to the brainstem and thalamocortical networks. Thus, DBS acts as an “information lesion” and filters out the abnormal information flow; this could explain its efficacy in hypo- as well as hyperkinetic disorders.
The United States Food and Drug Administration (FDA) approved indications for DBS in movement disorders include PD, dystonia and essential tremor (ET). DBS is currently the standard of care for patients experiencing the motor complications of levodopa treatment in Parkinson's disease. Apart from this, levodopa unresponsive tremor of PD also responds well to DBS. Careful patient selection is pivotal to a good outcome for DBS. A good levodopa response of motor signs has to be documented before taking up patients with PD for DBS, as only levodopa responsive symptoms are likely to respond to DBS (with the exception of tremor). PD-dementia is an absolute contra-indication; active depression, psychosis and other neuropsychiatric symptoms should be corrected and stable remission achieved before considering surgery. An advanced age is a relative contra-indication; an operational age cut-off of 70 years is followed by most centers as the criterion for deciding the 'advanced age.' An age more than 70 years is often associated with more rapid progression of PD, steeper decline of cognition, higher burden of co-morbidities and more brain atrophy, posing technical difficulties in targeting and planning a safe trajectory., A minimum of 4-year duration of motor symptoms is recommended to confidently establish the clinical diagnosis of PD after excluding atypical Parkinsonian syndromes, and the documentation of a good levodopa response may not be possible before this.
DBS is typically done in PD when the disease duration is of 11-13 years; in the majority of patients, motor complications have a major deleterious impact on the quality of life (QOL) by this time. Recent trials have shown that DBS has a convincing QOL benefit over best medical treatment, at least in a subset of patients who are younger and with low surgical risks, if these patients are taken up for the procedure earlier than this disease duration. The operational criteria for patient selection for DBS in PD are provided in [Table 1].
Among the conditions causing dystonia, DBS is indicated in 'primary' (isolated dystonia with a genetic etiology and without evidence of neurodegeneration or structural brain damage) generalized or segmental dystonia in which an adequate trial of medical therapy has failed to provide relief. Those with a shorter duration of dystonia and positive genetic test, particularly the DYT-1 mutation, may respond better. DBS is currently considered as the second line treatment in focal, particularly cervical, dystonia when botulinum toxin therapy fails. The disability from dystonia must be severe enough to justify the surgical risks. There are reports of a good response of cranial dystonia, writer's cramp and myoclonus-dystonia, to DBS. DBS has also been tried in cases of secondary dystonia (cerebral palsy, heredodegenerative conditions like neurodegeneration with brain iron accumulation, and drug induced dystonia) with less robust and poorly sustained responses, except in case of tardive dystonia where the effect sizes with DBS were larger.,
Majority of the patients with ET experience limitation of daily activities from tremor; however, only less than 50% of them have any functionally meaningful responses to pharmacotherapy., Though thalamotomy offers an effective symptomatic relief, thalamic DBS is the preferred management in patients with medically refractory ET as DBS can be done bilaterally and is associated with fewer adverse effects. Apart from these established indications, DBS has been tried for symptomatic relief in a variety of other movement disorders like Tourette syndrome, rubral tremor in multiple sclerosis, traumatic brain injury and Huntington's disease, with variable results.,,, The newer indications for DBS are discussed separately in another article in this issue of the journal.
Subthalamic nucleus (STN) and the internal globus pallidus (GPi) are the most widely used surgical targets for DBS in PD. Though the debate on which among the two is better is still continuing, majority of the centers prefer STN target unless there are contra-indications for the same., Better cognitive, neuropsychiatric and quality of life outcomes with GPi target as compared to STN target shown in earlier reports and meta-analyses ,, have not been reproduced in recent prospective head-to-head comparisons with long term follow-up. Recently published 3-year follow-up observations have suggested better motor outcomes and no significant difference in cognitive outcomes with the STN target., Robust control of Parkinsonian motor symptoms enabling a significant reduction in the dosage of dopaminergic medications is an advantage of the STN target, compared to its counterpart., The GPi target does not allow such remarkable medication reductions; GPi DBS acts predominantly by a direct anti-dyskinetic effect. The improvement of dyskinesias following STN DBS is primarily attributable to reduction in dopaminergic medication doses, though a direct antidyskinetic effect also has been shown by some studies.
The pedunculopontine nucleus (PPN) has been targeted in patients with PD for the control of axial symptoms, particularly the freezing of gait, which are often refractory to stimulation of GPi or STN. The initial results have not been very encouraging and there is insufficient evidence at present to recommend the use of this target in clinical practice., The caudal zona incerta/posterior subthalamic area region (which is close to the STN and could be stimulated using electrode contacts of leads implanted in STN), as well as the VIM (Ventral intermediate) nucleus of thalamus can be targeted for tremor control; however, these are not preferred sites as the other Parkinsonian symptoms may not be adequately controlled. The latter is the preferred target for neurostimulation in ET., The recommended target for DBS for generalized dystonia as well as segmental or focal (cervical) dystonia which does not respond satisfactorily to botulinum toxin, is the GPi. STN stimulation has also shown promising results in smaller series of patients, and could be an alternative target. The medial thalamic target (centromedian-parafascicular complex and nucleus ventralis oralis internus) DBS provides around 50% relief of tics in selected patients with severe, medication refractory Tourette syndrome.,
The deep brain stimulation leads are traditionally implanted using stereotactic frame-based techniques. The intracranial end of the DBS lead with quadripolar (leads with eight electrode contacts are also available) electrode contacts is placed in the target. The extracranial end projecting out through the burr hole to the subcutaneous plane of the scalp is connected using extension wires, to the pulse generator. The pulse generator is implanted in an infra-clavicular pocket (conventionally on the left side by most surgeons) on the chest wall. The target can be planned in multiple ways, like using standard stereotactic atlas based co-ordinates, direct visualization of the target based on high-resolution magnetic resonance imaging (MRI) or using indirect methods (when the target boundaries are not clear on MRI) based on other clearly visible anatomical landmarks like the red nucleus and the floor of third ventricle. A combination of these three methods is used by most surgeons to ensure accuracy. The STN is typically located 1-4 mm posterior, 9-12 mm lateral and 4-5 mm inferior to the midpoint of the line joining the anterior and posterior commissures (mid commissural point; [Figure 1]).,, The GPi is approximately 2-3 mm anterior and 2-5 mm inferior to the mid-commissural point and 17-21 mm lateral to the third ventricular wall.,,,
Implantation of DBS leads is performed under local anaesthesia. The stereotactic frame is fixed on the day of surgery and the patient undergoes imaging (the computed tomographic [CT]scan is preferred at our centre, considering the patient comfort). The images, with the stereotactic frame in place firmly fixed on the patient's head, are fused with the pre-operative MR images and the coordinates are deduced (X, Y Z, ring angle and arc angle). The ring and arc calculations allow for a specific burr-hole site selection which is suitable for planning the safest trajectory to reach the target. The trajectory is important and has to be carefully planned with a sufficient margin of safety from the ventricular wall, vessels and deep sulci. After prepping and draping, the burr-hole, with its site pre-determined by frame coordinates, is made and micro-electrode recording (MER) electrodes (usually an array of five MER electrodes are used) are introduced to the target area, usually 1 cm above the radiological target. The MER electrodes are then made to descend in a co-ordinated fashion using an electrode carrier and the target is electrophysiologically mapped based on the micro-electrode recording obtained.
MER is used by most centers to perform an on-table electrophysiological marking of the target boundaries [Figure 2] and to ensure accuracy. This procedure is supplemented by an intra-operative clinical assessment to determine the improvement of clinical signs and to detect any adverse effects related to stimulation., This needs the patient to be awake during the procedure and has disadvantages related to it. MER, especially if a multichannel recording is used, is also considered to increase the risk of hemorrhagic complications and adverse effects related to a more profound lesioning of the target, particularly in case of the STN target., Moreover, it has not been clearly demonstrated so far that the use of MER improves accuracy and outcome. Anatomical monitoring using an intra-operative imaging is now being evaluated as an alternative technique for improving the accuracy of electrode placement on the target, and its non-inferiority, and even superiority, in certain aspects of outcome have been recently demonstrated., Robot assisted and “frameless” techniques are also emerging and preliminary reports suggest that these could have an accuracy comparable to the conventional frame based techniques. However, these techniques need more expensive infrastructure, and evidence for accuracy and outcome compared to the conventional techniques is currently insufficient to recommend their use in routine clinical practice. The implantation of the pulse generator is done under general anesthesia, once the leads have been successfully placed.
The implanted pulse generator needs to be programmed optimally so that the patient gets the best possible relief of symptoms with no/minimum possible adverse effects related to stimulation. Both, the rechargeable (the stimulator needs to be charged periodically, once in three or four days or so, using an externally placed charger) and the regular (non-rechargeable) pulse generators are commercially available. The rechargeable stimulators are more expensive but have a longer and somewhat fixed service life while the service life of a non-rechargeable stimulator depends largely on the energy consumption. Therefore, the use of stimulation parameters that consume minimum electrical energy is a consideration for getting a longer service life in such stimulators. The fine-tuning of neuro-stimulation, making alterations in the active electrode contact in the multi-polar DBS lead and the frequency, width and amplitude of the electrical pulses delivered is called 'programming'. Impulse durations ('pulse width') of 60-90 microseconds and a frequency of 130Hz are usually used in STN stimulation for PD., GPi stimulation, particularly in dystonia, may require higher impulse durations and frequencies in view of the larger tissue volume to be stimulated. First generation DBS devices delivered stimulation in the voltage-controlled mode; instabilities in stimulation fields secondary to changes in tissue impedance form a disadvantage of this mode. Newer generation devices are capable of a constant-current mode that is devoid of this problem.
The amplitude of stimulation is fine-tuned to maximize benefits and minimize side effects. Alteration in frequencies and pulse widths are tried in patients with inadequately controlled symptoms. Patients with freezing of gait and dysarthria sometimes benefit from stimulation at lower (60-80 Hz) frequencies and higher amplitudes. Initiation of programming is generally deferred till the electrode insertional effects (microlesional effects of the electrode insertion on the target nuclei) wane off. Medication adjustments are needed simultaneously with the programming sessions. Multiple programming sessions will be needed for most patients; they may also require repeat programming sessions later to manage new/worsening symptoms secondary to progression of the disease. Programming of DBS in conditions like dystonia may require prolonged periods of observation, as the benefits may not be evident immediately. New generation devices allow shaping of the stimulation field using paradigms like interleaving of stimulation using two adjacent contacts and these could be useful in selected cases like those who develop adverse effects or insufficient symptom control with conventional settings. General guidelines have been published to aid in programming of DBS.,,
DBS has emerged as a time-tested standard of care for patients with PD experiencing motor complications, supported by evidence from several randomized controlled trials ,,,,, and numerous systematically conducted prospective observational studies.,,,, These have demonstrated its superiority over medical management and its ability to provide sustained improvement in functioning, cardinal motor symptoms of PD, motor fluctuations, dyskinesias and QOL, up to a period of five years. Long term observations beyond this point have shown continued benefits (compared to baseline) on cardinal motor symptoms of PD like tremor, bradykinesia and rigidity even 10 years after surgery.,,,, However, the axial motor dysfunction (postural instability, gait dysfunction including freezing of gait) and speech have been shown to worsen beyond the initial five years. This contributes to the declining of the QOL. The effect of DBS on the QOL has been shown to wane off by 5-7 years when DBS is done according to the traditional indications. Around a quarter to a third of patients in such long-term follow-up studies were also found to develop significant cognitive dysfunction.,, The worsening axial motor and cognitive dysfunction are attributable to the natural progression of PD which is unaffected by DBS.,, Though DBS has been found to provide a survival advantage, it has no proven neuro-protective effects or effects on progression of PD pathology. These studies have stimulated considerable interest in considering intervention with DBS early, so that patients could enjoy the QOL benefits of the procedure for a longer period. Most of the information available on long-term outcome is for the STN target; systematic follow-up information on outcome is currently available for GPi DBS in PD, for up to 5-6 years only.
The efficacy of DBS in controlling dystonia has been clearly demonstrated by randomized controlled trials (DBS versus sham stimulation)., Sustained improvement of dystonia scores by around 50 -60% have been demonstrated with GPi DBS in generalized and segmental dystonia.,,,, The reported effect sizes tend to be larger for DYT-1 positive cases compared to non-DYT-1 dystonia., Smaller series of cases where STN was targeted, have also shown sustained improvement of dystonia scores for many years. Persisting improvement with thalamic stimulation (around 40% improvement in stimulation ON state) has been demonstrated in patients with ET more than 10 years after surgical implantation. Patients with PD also get sustained improvement of tremor with thalamic DBS, but the other parkinsonian signs like rigidity and bradykinesia do not improve., [Table 2] shows some of the major studies which demonstrated the efficacy of DBS for common movement disorders.
Intracranial hemorrhage could occur in 1-4% of the patients and could be asymptomatic, or symptomatic intracerebral hemorrhage or intraventricular hemorrhage.,, An advanced age, the presence of hypertension, a higher number of MER penetrations, and closeness of the trajectory to ventricular wall or sulci have been identified as risk factors for this complication [Figure 3]. The incidence of infective complications, including hardware infections, is 1.5-5.0% in most centers. Early detection and aggressive management is important and this could be helpful in salvaging whole or part of the hardware in around half of the cases; rest would require total explantation of the DBS hardware.,,,,, The other surgical and hardware related complications of DBS are listed in [Table 3] and their incidence in some of the large reported series is presented in [Table 4].
DBS is a relatively safe procedure on a long term basis. Stimulation related adverse effects like dysarthria, diplopia, sensory symptoms and dyskinesias in PD can be managed by programming, which is an advantage when compared to lesioning surgeries., Stimulation related gait disturbances and dysarthria following GPi DBS have been reported in patients with dystonia; and paraesthesias, dysarthria and ataxia following thalamic DBS may occur in patients with ET.,, Patients who have undergone DBS need careful follow-up evaluations for programming and medication adjustments aimed at controlling new or worsening symptoms, which could be stimulation related or arising from progression of the underlying disease.
The basal ganglia circuits modulated by DBS have cognitive as well as affective functions and this could result in cognitive and neuro-psychiatric sequelae, following the performance of the DBS procedure. A systematic evaluation of the cognitive and psychiatric status is, therefore, mandatory during patient selection as well as follow-up. Cognitive changes attributable to DBS include reduction in verbal fluency and mild impairment in executive functions and memory, which are generally not severe enough to be clinically significant. Lesioning effects of the implant surgery as well as the direct effects of stimulation, compounded by disease progression are responsible for such changes. The changes in mood and behavior are often mixed. Improvement or worsening of depression could occur; presence of significant depression before surgery may predict post-operative worsening. Stimulation could result in worsening of mania or rarely, appearance in those who did not have overt symptoms pre-operatively. These underscore the need for a proper psychiatric evaluation during patient selection. An over-aggressive reduction in dopaminergic medications following the DBS procedure could contribute to apathy and worsening of depression in PD patients. Dopaminergic therapy (particularly dopamine agonists) as well as stimulation could contribute to impulse control disorders (ICDs). These ICDs could improve, worsen or occur de novo depending on the medication adjustments, the electrode position and the stimulation parameters used., The concerns regarding the risk of suicide following DBS have been largely resolved and the increased risks, if at all present, are low and present only in the initial years after surgery.,
Overall, DBS is safe from cognitive and psychiatric angles if patient selection is carefully done and the patients are followed up for neuropsychiatric complications systematically; the benefits of DBS outweigh these comparatively smaller risks. Recent studies suggest that the cognitive and psychiatric effects in the long term are not markedly different among the two targets commonly selected for PD, though earlier reports have shown an edge for GPi over STN in these aspects.
An abrupt interruption of long term stimulation may result in an akinetic state akin to the neuroleptic malignant syndrome, in patients with PD who have undergone DBS procedure. This is a medical emergency and patients remain unresponsive to high doses of dopaminergic medication; they often retain the therapeutic response to neurostimulation and improve with restoration of stimulation. This could occur when the neurostimulator has to be removed due to the presence of an infection or in the case of an inadvertent implantable pulse generator (IPG) damage. In the developing world where medical insurance coverage is not universal, it is, therefore, important to counsel the patients regarding recurring expenses for neurostimulator replacements when DBS is initially discussed with them.
The field of neurostimulation is witnessing many fascinating technological advances, including the use of intra-operative imaging and frameless techniques, current steering using directional electrode designs  and the concept of closed loop stimulation or “adaptive DBS”. Apart from the use of intra-operative neuroimaging, the other radiological advances in relation to DBS include the use of 7 Tesla MRI for targeting of sub-nuclei/specific areas within the target, and the use of tractography to delineate specific tracts for better anatomical localization. The conventional cylindrical/'ring shaped' electrode contacts deliver current circumferentially, resulting in a sphere-shaped field of stimulation. Current steering is done by dividing the 360 degrees of the ring-shaped electrode contact into three or four sectors, each of which could be separately activated. Current steering and modification of the shape of stimulation field could be helpful in reducing the adverse effects related to stimulation. Adaptive DBS is based on the principle of detection of a signal (for example, changes in the local field potential) from the target nuclei or from elsewhere that is indicative of a pathological neuronal activity; the signal, in turn, triggers or modifies the delivery of stimulation. Thus, the closed loop DBS delivers stimulation as and when required and does not modify or disrupt functionality when the neuronal activity is normal.
DBS has emerged as a successful symptomatic therapy for several movement disorders. A careful patient selection and follow-up by a dedicated multi-disciplinary team are crucial to achieving a good outcome and in avoiding the neuropsychiatric adverse effects. Technological advances in the field are expected to further improve the outcomes with neurostimulation and widen the scope of the therapy to areas outside movement disorders, like the managment of epilepsy as well as psychiatric and cognitive disorders.
Financial support and sponsorship
Conflicts of interest
There are no conflicts of interest.
[Figure 1], [Figure 2], [Figure 3]
[Table 1], [Table 2], [Table 3], [Table 4]