15 Efficacy and Complications of Deep Brain Stimulation for Movement Disorders

Chronic electrical stimulation has now become a mainstay of treatment for patients with movement disorders. In fact, due to its effectiveness and potential reversibility, deep brain stimulation (DBS) has gradually replaced lesioning procedures for the treatment of Parkinson disease (PD),1 dystonia,² and essential tremor (ET)³ in many centers around the world. The three primary targets for DBS in movement disorders surgery are the thalamus, globus pallidus, and subthalamic nucleus (STN). As for lesions, different outcomes might be expected according to the disease and chosen target. We discuss the efficacy and adverse effects of DBS according to target in each of the three major disorders most commonly treated (PD, ET, and dystonia).

Efficacy

The motor thalamus is composed of several nuclei that receive afferents from the cerebellum and basal ganglia and send projections to the motor and premotor cortices. Several units in the motor thalamus in humans respond passively or actively to movement,4,5 and stimulation of the motor thalamus in nonhuman primates elicits motor responses.⁶ Stimulation of the thalamus may be efficacious through activation of the cerebellothalamocortical pathway rather than inhibition,^7,8 although the precise method of action remains controversial.

Diverse classifications have been proposed to subdivide the motor thalamus. Due to its extensive use among surgeons, the one created by Hassler is the most commonly employed in clinical practice.⁹ According to Hassler, the motor thalamus may be subdivided into oral, caudal, intermediate, and lateropolar segments. The ventral intermediate (Vim) nucleus of the thalamus is the most effective thalamic target for the treatment of tremor in various conditions.¹⁰ The most common disorder that presents with tremor is ET.³ Tremor is also a cardinal manifestation of PD and frequently presents as a disabling component of conditions such as multiple sclerosis (MS) or cerebellar disorders.

The mere introduction of an electrode to the Vim may result in a decrease in tremor. The duration of this “microthalamotomy effect” is variable, but typically ranges from days to weeks. Nevertheless, in a few patients the effect can persist for years, making stimulation unnecessary.¹¹ Different components of a patient’s tremor may respond differently to stimulation, though this has not been consistently demonstrated in all studies. As a rule, it appears that distal tremor is better controlled than proximal, and rest tremor better than kinetic tremor.

Essential Tremor

Surgical therapy for ET is considered when standard medications (primidone, β-blockers, gabapentin) fail and the patients continue to be disabled. Contralateral arm tremor control is observed in 68 to 79% of the patients treated with thalamic stimulation at 1 year.^12–17 The effects on postural tremor are less dramatic, with a 46 to 56% benefit at 3 months.^12,16 The benefits achieved with thalamic stimulation for ET are still significant at 6 years, although slightly lower than the ones observed at 1 year.¹⁷

Patients with ET can show the phenomenon of tolerance; that is, patients may require higher stimulation settings to capture tremor benefit.^17,18 At 6 years, the mean stimulation amplitude increased from 2.0 to 2.6 V, mean rate from 156 Hz to 173 Hz, and pulse width from 103 μs to 89 μs.¹⁷ This gradual increase in stimulation may be problematic because it may lead to side effects, such as speech difficulty or paresthesias, as well as premature battery failure. Some have recommended turning the stimulation off at night and attempting to minimize parameters in the hopes of avoiding this. Alternatively, some patients have gone on to have thalamotomy lesions made through their DBS electrodes with good results.¹⁹ The discontinuation of stimulation may induce a rebound effect in which the severity of tremor becomes worse than it was before stimulation commenced. Response of ET to stimulation is graded and diminishes with increasing frequency from 45 to 100 Hz. The optimal stimulation frequency is 100 to 130 Hz for most patients.²⁰

Tremor of Parkinson Disease

Contralateral arm tremor improves in 71 to 92% of patients with PD treated with thalamic stimulation at 3 months.¹² This appears to be well sustained, with other series reporting 74% of patients well controlled at 1 year.¹³ Contralateral foot tremor can be improved in 55 to 90% of the patients at 3 months.21

Although tremor is very well treated with Vim stimulation,²² this procedure does not treat akinesia, rigidity, gait disturbance, and postural instability.^12,23 For this reason the Vim is rarely the preferred target for patients with PD.¹ Certain patients who have predominantly unilateral tremor may benefit from thalamic surgery,²⁴ but most patients will progress with time and eventually be disabled by other symptoms. Therefore, thalamic surgery for PD is rarely performed in most centers.

Dystonia

Due to the effectiveness of pallidal DBS in dystonia, thalamic stimulation has not been thoroughly explored in recent years. In a series of 12 patients with primary and secondary dystonia treated with thalamic stimulation in the ventrolateral posterior nucleus, improvements in global functional outcome were noted in 67% of the patients. Yet no improvements in dystonia scores were reported.²⁵ Individual reports of patients responding well to chronic stimulation of thalamic targets (e.g., ventralis oralis anterior nucleus) have also been published.^26,27

Other Tremor Disorders

The benefits of thalamic procedures for patients with other disorders, such as posttraumatic tremor, tremor associated with MS, and cerebellar disorders are less predictable, being of lesser magnitude or transient.^28,29 Patients with MS may expect a 60% reduction in their tremor scores with thalamic DBS at 1 year.³⁰ Nevertheless, this may have a low impact on the quality of life of these patients in the long term due to the associated pyramidal, cerebellar, or sensory symptoms that often develop in these patients postoperatively.³⁰

Globus Pallidus

The globus pallidus internus (GPi) is one of the main signal outflow channels of the basal ganglia. It receives afferents from the striatum, globus pallidus externus (GPe), STN, and substantia nigra compacta. It sends efferent projections to the thalamus, habenula, and brain stem tegmental structures, such as the pedunculopontine nucleus. Motor, associative, and limbic territories have been identified in the GPi, with the former comprising the ventrolateral two thirds of the nucleus. In fact, the posteroventral portion of the GPi is considered the preferred target for stereotactic lesioning.

The GPi is a large structure, which provides an opportunity for variations in the site of implantation of DBS electrodes, leading to great variability in surgical outcomes. In PD, this was one of the reasons for the recent inclination toward the use of the STN as a target. For dystonia, the GPi continues to be the most frequently chosen surgical target.

Parkinson Disease

In contrast to Vim surgery, the GPi and STN have emerged as effective targets not only to control tremor but also to ameliorate rigidity, bradykinesia, and gait disturbances.^2,31–49 In addition, motor side effects of dopa-replacement therapy such as dyskinesias, freezing, and on–off fluctuations are effectively reduced with stimulation of these targets.

In general, surgery improves parkinsonism to the level achieved with L-dopa, and the response to surgery can be predicted by the improvement obtained after L-dopa administration. In fact, symptoms that are resistant to L-dopa, such as bladder dysfunction, constipation, speech difficulties, sexual dysfunction, psychological difficulties, seborrhea, and cognitive dysfunction, are also resistant to surgery.^38,50 To date there is little effective medical or surgical therapy for these problems.

The range of reported motor outcomes with chronic stimulation of the GPi for PD is considerable,^47,51 with the best series reporting a 67% improvement⁴⁷ in the Unified Parkinson’s Disease Rating Scale (UPDRS) motor scores in the off-medication state. Most studies, however, have shown improvements in the range of 30 to 55% with bilateral GPi stimulation^32,52–57 in the off-medication state. Benefits from unilateral stimulation are more modest (around 30 to 40%) and predominantly on the contralateral side, although mild ipsilateral improvements may also be noted.^56,58,59 The reduction in tremor with GPi surgery approximates 80%. The rigidity and akinesia scores improve approximately by 60%, whereas improvement in gait and posture is on the order of 40%. Involuntary movements induced by L-dopa improve on the order of 80 to 90%.^{32,52–56,60}

Dystonia

The clinical response to surgery is dependent on the etiology of the dystonia. Although data remain preliminary, it is becoming clear that the primary generalized forms respond better to GPi DBS than do the secondary forms.^25,61–64

Although the effects of pallidal DBS in PD can be immediate, patients with dystonia may not realize benefit for several days, weeks, or even longer.^65–67 The reasons why the benefits of pallidal stimulation are delayed and often progressive with ongoing stimulation for dystonia patients are not fully understood.

Within the primary generalized dystonias, the genetically identified DYT-1 mutation type is perhaps the most responsive, with early reports demonstrating a 90% decrease in the Burke-Fahn-Marsden Dystonia Rating Scale (BFMDRS) after a 12-month follow-up.⁶⁸ This is in agreement with our observations of improvement greater than 80% in DYT-1 dystonia.⁶¹ Other forms of primary generalized dystonia may be slightly less responsive, on the order of 48 to 84%.^69–72

Cervical dystonia seems to benefit from pallidal DBS as well, with treated patients obtaining a reduction of ~60% ⁷³ in all three components (motor symptoms, pain, and disability)of the Toronto Western Spasmodic Torticollis Rating Scale (TWSTRS). It has been noted by several authors that the time course of response for each of these areas is different. Although pain often responds very quickly, motor symptoms and disability often respond more slowly and progressively.

For secondary dystonias, the average improvement on the Abnormal Involuntary Movement Scale (AIMS) is ~40%.73 However, the range of responses is quite variable (from 0% to 75%), indicating that this group of heterogeneous conditions needs further investigation to determine the appropriate indications and realistic outcomes. There is evidence that some of these patients may derive benefit from thalamic procedures; the choice of the most appropriate target remains unresolved.^73,74

Subthalamic Nucleus

The STN has been regarded as an important modulator of basal ganglia output. It receives its major afferents from the cerebral cortex, thalamus, GPe, and brain stem. It projects mainly to both segments of the globus pallidus, substantia nigra, striatum, and brain stem. The STN is primarily composed of projection glutamatergic neurons. Lesions of the STN can induce choreiform movements and ballism on the contralateral side of the body.⁷⁵

Due to the uniformity of the clinical results and the compact size of the nucleus, the STN is currently the most popular target for stimulation in patients with PD.^{1,2,32,33,37–47} Overall, STN DBS yields greater than 50% improvement in UPDRS motor scores at 12 months in the off-medication condition in well-selected patients. Tremor, rigidity, and bradykinesia improve 80, 60, and 55%, respectively, on average, at 12 months. With STN surgery, L-dopa–induced dyskinesias improve ~80 to 90%. This is, in part, due to the reduction in dopaminergic drugs that is often achieved after these procedures.^1,2,31,40,45

The Best Target for Parkinson Disease

The issue of whether the GPi or the STN is a better target for the treatment of advanced PD is still debated.^{1,31,32,38,52,57,76,77} The globus pallidus is a larger structure, and there is more heterogeneity in the response to surgery.⁷⁸ As previously stated, this may be due to the variation in the position of the electrodes within the pallidal complex.^79,80 By comparison, the STN is smaller and provides more consistent results.^{2,31–33, 36,38,39,41,42,44,46,47,49} However, there are regions within the STN with limbic and associative connections, which are closely apposed to the motor region of the nucleus. Spillover of electrical stimulation into these territories might explain the higher incidence of cognitive and emotional side effects seen after STN surgery compared with GPi surgery.⁷⁶ STN stimulation appears to be more likely to improve bra-dykinesia.⁷⁶ Stimulation of both targets reduces dyskine-sias.⁷⁶ Of note, STN DBS was also found to improve cervical dystonia and ET.^23,81,82

Complications of Deep Brain Stimulation Surgery

The most fearsome complications of stereotactic procedures are intracranial hemorrhages. With an incidence of ~2 to 3%, most hemorrhages are intraparenchymal, but subdural or intraventricular hemorrhages are occasionally seen as well. Most hemorrhages are asymptomatic, observed only on postoperative brain imaging.83,84 However, in some patients the effects of a bleed can be serious, leading to permanent sequelae.

There is some debate in the literature as to whether the use of multiple passes of microelectrodes for mapping, while presumably decreasing the adverse effects associated with targeting inaccuracy, may in fact increase the risk of hemorrhage. In a series of 481 lead implantations, 0.6% were associated with hematomas causing permanent deficit.⁸⁴ Patients who developed hematomas had a slightly greater, but not significant, number of microelectrode recording penetrations than patients who did not have hematomas.⁸⁴ In addition, some neurosurgeons feel that the increased operative time for microelectrode mapping may increase the infection risk. These questions are important but cannot be definitively addressed with the currently available data.

Other acute complications include postoperative nausea in ~3 to 5%, headaches in 1.9 to 5%, seizures in 1.6 to 5.5%, and perioperative confusion in up to 15% of patients.^{2,31– 33,36,38–45,47,49} Some unusual complications such as venous air embolisms have also been reported.^85,86

Stimulation-Related Complications

Thalamus

Specific adverse effects encountered with DBS in the Vim include speech problems, corticospinal symptoms, and ataxia, as well as paresthesias related to stimulation of the adjacent tactile ventrocaudal (Vc) nucleus and the medial lemniscus. Because these effects are related to the spread of current to adjacent structures, in many cases stimulation parameters can be adjusted to reduce their incidence.⁸⁷

Paresthesias have been reported in 9 to 100% of these patients, in whom symptoms ceased when stimulation was discontinued.^12,13,16,21 Cerebellar complaints have been reported in less than 10% of the patients. Weakness (from spread of current to the internal capsule) has been reported in less than 12% of the patients. Dysarthria has been reported in less than 8% of the patients following unilateral stimulation. It is significantly more common after bilateral procedures (higher than 45% in most studies).^12,13,16,21

Globus Pallidus

During globus pallidus surgery, it is important to identify the sensorimotor territory of the GPi (populated by neurons that respond to movements of the limbs), as well as the optic tract and the corticospinal tract. Intraoperative electrical stimulation in the optic tract produces phosphenes, whereas stimulation of the corticospinal tract produces motor contractions.80,88 During the programming of the patients, increasing stimulation current beyond a given patient’s threshold may result in transient paresthesias, tonic contractions of the contralateral side of the body, dysarthria, and photopsia. Decrease in stimulation parameters usually improves these adverse effects. Patients with dystonia may experience a rebound effect when stimulation is discontinued, which may be extremely severe and potentially life threatening.^73,89

Subthalamic Nucleus

The objectives during surgery on the STN are to identify the sensorimotor territory in the STN and to avoid adverse side affects related to important adjacent structures, such as the fibers of the third cranial nerve (medial to the STN), the corticospinal tract (anterior and lateral), and fibers of the medial lemniscus (posterior). Yet, dyskinesias, paresthesias, diplopia, dystonia, and motor contractions are relatively common side effects with STN stimulation.⁵⁷ In addition, hypophonia, eyelid apraxia, increased libido, sialorrhea, hypomania, and decreased memory have also been reported.⁹⁰ Depression and weight gain occur in ~5 to 15% of patients.^{2,31–33,36,38–40,43–47,49,57,64} There are probably multiple mechanisms underlying these events.

Hardware-Related Complications

There are long-term risks associated with the implantation of any device. These appear to be associated with all applications of DBS. A review of 124 electrodes implanted in 79 patients over a 6-year period⁹¹ showed that 20 patients (25.3%) had hardware-related complications. These involved 23 (18.5%) of the electrodes. Of the 23, there were four lead fractures, four lead migrations, three short or open circuits, 12 infection/erosions, two allergic reactions, and one cerebrospinal fluid (CSF) leak. Although these were not related to the selected target, lead fractures occurred more commonly in patients with prominent cervical dystonia or dyskinesia. The hardware-related complication rate was 8.4% per electrode-year. Importantly, 19.2% of the patients developed complications within the first month, 42.3% between the first and twelfth months, and 38.5% 12 months after the procedure. Thus the possibility of complications persists for the life of the device. Replacement of the implantable pulse generator (IPG) for battery depletion was needed in 12 of the patients ranging from 7 to 70 months after implantation, averaging 45 months.⁹²

Certain authors have also noted that lead fracture appears more common in patients with dystonia, likely due to extreme movements of the neck. This complication appears to be more frequent when the connector to the extension cable is low in the neck, below the mastoid process.⁷⁰ In another study² for PD, the overall rate of hardware infections was 2.9% (33% with removal of at least part of the system), and the incidence of lead problems was 2.9%. Incidence of lead infection appears to be increased by periods of externalization for lead testing.⁹³

Conclusion

Related posts:

History of Surgery for Movement Disorders Preparation for Movement Disorder Surgery Gamma Knife Deep Brain Stimulation for Tremor Stereotactic Surgery with Microelectrode Recordings Deep Brain Stimulation Programming

Stay updated, free articles. Join our Telegram channel

Join