INTRODUCTION — Motor fluctuations and dyskinesia (collectively referred to as motor complications) affect approximately 40 percent of patients with Parkinson disease (PD) after five or more years of levodopa therapy . Intolerable motor complications that interfere with quality of life often prompt referral to a multidisciplinary movement disorder clinic for consideration of device-assisted and lesioning therapies. Patients with refractory tremor may also be candidates for these therapies.
Eligibility for specific procedures is individualized based on a multidisciplinary assessment of patient and disease characteristics, treatment availability, and patient values and preferences. All procedures offer potential symptomatic benefit and do not alter progression of the underlying neurodegenerative process. With appropriate selection, patients with PD can derive substantial gains in function and quality of life from these therapies.
This topic will review device-assisted and surgical therapies for patients with PD who are experiencing motor complications or refractory tremor. The initial medical management of motor complications of PD and levodopa therapy is reviewed separately. (See "Medical management of motor fluctuations and dyskinesia in Parkinson disease".)
Additional reviews on the management of patients with PD include:
Eligibility — Patients with PD who maintain a clear response to levodopa are candidates for device-assisted therapies when motor fluctuations and/or dyskinesia cause disability or reduced quality of life despite optimal oral levodopa or adjunctive therapies [2-5].
In nearly all cases, the outcome of device-assisted therapy will never be better than the best preoperative levodopa response. The exception to this rule is levodopa-resistant tremor, which can be improved with deep brain stimulation (DBS), but not the infusion therapies.
Generally, device-assisted therapies do not help cognition or levodopa-unresponsive axial symptoms in PD, which include postural instability, speech and swallowing problems, and freezing of gait in the "on" state . Other nonmotor symptoms of PD tend to respond more with DBS than the infusion therapies. In general, disabling nonmotor fluctuations that are present mainly when levodopa wears off (eg, pain, panic attacks, akathisia) may respond to all of the device-assisted therapies. For patients undergoing DBS, nonmotor symptoms that result from high doses of dopaminergic medications (eg, somnolence, impulse control behaviors, orthostatic hypotension) may improve if doses are able to be reduced or discontinued postoperatively. More limited evidence also suggests that DBS may improve sleep, pain, gastrointestinal symptoms, and urinary symptoms in PD [6-8].
Exclusions and precautions — Patients with secondary causes of parkinsonism (eg, drug-induced, vascular) or atypical parkinsonian disorders (eg, multiple system atrophy) are not candidates for device-assisted therapies.
Patients with PD who have dementia or significant cognitive impairment are also not suitable for these therapies . However, PD patients with mild to moderate cognitive impairment may be considered, especially for the infusion therapies .
For patients with active psychiatric issues, such as anxiety, depression, hallucinations, and delusions, device-assisted therapies should be delayed until psychiatric symptoms are stabilized. There is concern that DBS, particularly of the subthalamic nucleus (STN), may increase the risk for suicide [9-13]. While this increased risk is not firmly established , clinicians should screen potential candidates for depression using standardized screening instruments and exclude those at high risk for suicide from consideration of DBS [11,15]. (See "Screening for depression in adults", section on 'Screening options' and "Suicidal ideation and behavior in adults", section on 'Patient evaluation'.)
Multidisciplinary evaluation — Patients with PD who are potential candidates for device-assisted therapies should be referred to a specialized center for a multidisciplinary team evaluation . The multidisciplinary team typically includes a movement disorders neurologist, a functional neurosurgeon, and a neuropsychologist. Many centers also include representation from psychiatry, speech therapy, physical therapy, occupational therapy, nursing, and social work.
The goals of the multidisciplinary evaluation are to:
●Confirm the diagnosis of typical PD
●Make sure the patient has undergone adequate medication trials
●Make sure that the patient has at least some high-quality "on" time from levodopa and that there is a significant improvement from "off" to "on" time by observing the patient in both conditions
●Assess surgical risk (see "Preoperative medical evaluation of the healthy adult patient")
●Screen for cognitive impairment/dementia and psychiatric problems
●Educate the patient and family on the available therapies, their benefits and risks, as well as realistic expectations
After potential candidates complete the full evaluation, the team convenes for a discussion about each candidate. If the patient is thought to be appropriate for device-assisted therapies, the team would then discuss the most appropriate therapy before sharing the consensus opinion with the patient and family.
PROCEDURE SELECTION — Three device-assisted therapies are available for management of medically refractory motor complications in patients with PD: deep brain stimulation (DBS), continuous levodopa-carbidopa intestinal gel (LCIG), and continuous subcutaneous apomorphine infusion (CSAI) (table 1). However, CSAI is not universally available, including in the United States. Invasive lesioning procedures (thalamotomy, pallidotomy) are primarily reserved for resource-poor settings and patients with medication-refractory tremor.
Patients with disabling motor complications — For eligible patients with disabling motor complications, the choice among the three device-assisted therapies for PD (DBS, LCIG, and CSAI) is individualized based on patient and procedure characteristics, availability of surgical expertise, and patient values and preferences (table 1).
There are no randomized, controlled trials directly comparing these treatments. The available observational evidence and clinical experience suggest that efficacy for reducing "off" time is comparable among the three . Some factors that may guide the choice among device-assisted therapies include the following observations :
●DBS has the longest track record and most efficacy data from randomized, controlled trials for reducing "off" time as well as improving dyskinesia compared with best medical care (see 'Efficacy' below). DBS is nondestructive, can be performed bilaterally with low neurologic morbidity, and can be modified over time to deal with changing or progressing patient symptoms. Disadvantages of DBS include the requirement for stereotactic brain surgery for electrode placement; surgical pulse generator implantation; the introduction of hardware with risk of infection, hemorrhage, or mechanical breakdown; and the need for periodic reprogramming. High cost is another disadvantage compared with conventional lesioning procedures such as thalamotomy or pallidotomy.
●Continuous LCIG infusion is effective for reducing "off" time and increasing "on" time compared with best medical care, although there are fewer data from randomized, controlled trials compared with DBS (see 'Efficacy' below). Drawbacks to LCIG include the need for surgical placement of the percutaneous gastrojejunostomy tube, daily tube flushing, diligent stoma care, the requirement to wear a portable pump, and adverse effects including infection.
●CSAI is likewise effective for reducing "off" time and increasing "on" time. It does not require surgery and is the least invasive of the device-assisted therapies. Like LCIG, there is less supporting evidence from randomized, controlled trials compared with the evidence for DBS (see 'Efficacy' below). Drawbacks to CSAI include the need for subcutaneous needle placement, skin care at injection sites, and the requirement to wear a portable pump. Long-term use may be limited by adverse effects, particularly the development of skin nodules. CSAI is also not an available option in the United States at this time.
Patients with refractory tremor — DBS and lesioning procedures are the main considerations for patients with refractory tremor who are seeking relief from a surgical therapy (table 1). The infusion therapies (LCIG and CSAI) are not helpful for levodopa-resistant tremor. As for other indications, DBS has largely replaced conventional lesioning.
In terms of the target, DBS of the ventral intermediate nucleus (VIM), subthalamic nucleus (STN), and internal globus pallidus (GPi) are all effective for tremor. The choice of target depends on the patient's individual symptoms. VIM DBS helps only tremor, while STN and GPi DBS will also treat contralateral rigidity and bradykinesia. While GPi and STN DBS appear to be equally effective for the motor symptoms and complications of PD (see 'Deep brain stimulation' below), STN DBS is thought to be more effective for tremor than GPi DBS.
Among lesioning procedures, conventional unilateral thalamotomy, made by an electrode, and focused ultrasound (FUS) thalamotomy are the main options for refractory tremor. Neither procedure is effective for motor complications or bradykinesia. (See 'Lesioning procedures' below.)
Patients in resource-poor areas — Device-assisted therapies require expertise and ongoing maintenance that may not be available in geographic locations remote from major medical centers where long-term follow-up is essential. For patients without access to these therapies, conventional one-time lesioning procedures (unilateral pallidotomy for motor complications, unilateral thalamotomy for refractory tremor) may be an option. (See 'Lesioning procedures' below.)
DEEP BRAIN STIMULATION — Deep brain stimulation (DBS) is the most frequently performed surgical procedure for the treatment of PD . There are two main targets for alleviating motor fluctuations and dyskinesia associated with advanced PD: the subthalamic nucleus (STN) or the internal globus pallidus (GPi).
Evidence from randomized, controlled trials shows that both DBS targets are comparable in terms of motor outcomes. DBS of the ventral intermedius nucleus (VIM) of the thalamus is effective only for PD tremor, not for bradykinesia or motor fluctuations. Since bradykinesia and motor fluctuations typically become more disabling over time, GPi and STN DBS are performed more often, even if tremor is the main symptom.
Procedure and mechanism of action — The procedure for DBS involves stereotactic brain surgery for unilateral or bilateral electrode placement. The implanted leads are then connected via a wire anchored to the burr hole site and tunneled under the skin to a pulse generator implanted in the chest wall below the clavicle (similar to a permanent pacemaker). When turned on, the battery-powered pulse generator delivers high-frequency electrical stimulation to the particular target (GPi or STN) chosen by the surgical team. Since DBS produces a safe and entirely reversible physiologic effect without destroying brain tissue, DBS has largely replaced lesioning procedures (pallidotomy and thalamotomy) for treating patients with advanced PD at major medical centers. The surgical procedure for DBS is reviewed in more detail separately. (See "Anesthesia for deep brain stimulator implantation", section on 'Surgical procedure'.)
DBS for motor complications of PD works in the following hypothetical way: two downstream physiologic effects of the hallmark cellular degeneration of the substantia nigra and corresponding loss of dopamine production are excessive STN excitation of the GPi and excessive GPi inhibition of the thalamus. These, in turn, cause reduced thalamocortical activity, which is believed to mediate akinesia and rigidity. Pallidotomy reverses the excess pallidal inhibitory effect and improves parkinsonism by targeted ablation of the GPi. High-frequency DBS suppresses neuronal activity and also activates efferent fiber pathways leaving the targeted nucleus. This explanation is a widely accepted model for the physiologic circuitry of the basal ganglia, but the actual mechanism is largely unknown.
PD with motor complications — Randomized, controlled trials comparing DBS with best medical therapy have shown that bilateral DBS improves motor function in selected patients with PD and motor fluctuations [18-22]. Results are illustrated by the following studies:
●In the first fully blinded trial performed with a sham control design (INTREPID), 191 patients with PD were implanted with bilateral STN DBS and randomly assigned for the first three months to active therapeutic stimulation or subtherapeutic (control) stimulation in a 3:1 ratio . All patients received active, open-label stimulation after three months and were followed for up to five years. The device itself used novel multiple independent contact current-controlled (MICC) technology. In an interim analysis of the first 160 randomized patients, STN DBS improved mean "on" time without troubling dyskinesia from baseline to three months by three hours compared with inactive control stimulation. Most secondary outcomes were also superior in the active group, including motor scores off medication, quality of life, and treatment satisfaction. The rate of serious adverse events was 13 percent in the first three months.
●The multicenter Veterans Affairs-National Institutes of Health (VA-NIH) trial randomly assigned 255 adults with advanced PD and motor complications to receive either DBS (with further randomization to STN versus GPi) or best medical therapy [18,23]. Outcome assessment was blinded. For the primary outcome measure of "on" time without troubling dyskinesia at six months, patients in the DBS group (pooled STN and GPi stimulation) improved by a mean of 4.6 hours per day versus zero hours per day for the best medical therapy group (mean difference, 4.5 hours per day, 95% CI 3.7-5.4) . Patients in the DBS group also had a higher rate of clinically meaningful motor improvement (71 versus 32 percent) and improved quality of life, as assessed by the Parkinson Disease Questionnaire (PDQ39). Results were similar for STN versus GPi stimulation. (See 'Subthalamic nucleus versus globus pallidus DBS' below.)
The benefit in motor symptoms seen in all of these trials comes at the cost of an increased risk of complications related to surgery, as serious adverse events were significantly more common with DBS than with medical treatment alone (see 'Complications and adverse effects' below). Limitations of some of the trials included a lack of a sham surgery control group.
PD with early motor complications — STN DBS is a viable treatment option for PD patients who develop troubling motor complications early in the natural history of the illness that cannot be managed with best medical therapy alone [15,24-26].
The potential benefit of DBS in earlier stages of PD is illustrated by results of the two-year EARLYSTIM trial, which evaluated 251 patients (mean age 53 years, mean duration of disease eight years, mean duration of levodopa use five years) with levodopa-responsive PD, early motor complications (present for less than three years), and no dementia [15,27]. Patients were randomly assigned to either bilateral STN DBS plus medical therapy or to medical therapy alone. Compared with medical therapy, STN DBS led to clinically meaningful improvement in self-assessed quality of life (the primary endpoint) as measured by the PDQ39 as well as improvements in the motor disability and activities of daily living (ADL) subscales of the UPDRS, levodopa-induced motor complications, and time with good mobility and no dyskinesia .
Secondary analyses of the EARLYSTIM trial reported that neuropsychiatric fluctuations decreased with bilateral subthalamic DBS plus medical therapy and did not change with medical therapy alone . Moreover, impaired quality-of-life scores at baseline were an important predictor of benefit in quality of life at two years .
Early-stage PD without motor complications — A pilot trial investigated the use of bilateral STN DBS in patients with less than four years of dopaminergic therapy and no motor complications, with the purpose of showing safety and feasibility of conducting a DBS trial in such an early stage of PD . Patients were randomly assigned to either bilateral STN DBS (15 patients) or optimal drug therapy (15 patients). At 24 months, there was no difference in mean total or part 3 UPDRS scores. Two patients in the DBS group had serious adverse effects, but the trial demonstrated the willingness of early-stage PD patients to participate in a trial testing an invasive surgical therapy. At five years of follow-up, patients in the STN DBS plus optimal drug therapy group had lower levodopa dose requirements, less requirement for polypharmacy, and lower odds of having rest tremor compared with patients receiving optimal drug therapy alone .
Subthalamic nucleus versus globus pallidus DBS — STN and GPi DBS are fairly equivalent for improving motor symptoms and quality of life in patients with PD complicated by medically refractory fluctuations or dyskinesias, and they have similar safety profiles. Supporting evidence comes from the multicenter, randomized VA-NIH trial (discussed above) and the Netherlands subthalamic and pallidal stimulation (NSTAPS) trial, which both directly compared STN DBS with GPi DBS [23,32].
However, there are some subtle differences in the effects of stimulating the two targets. STN DBS is generally associated with a greater postoperative reduction in antiparkinson medication , while GPi DBS provides a greater reduction in dyskinesia compared with STN DBS . There is some literature suggesting that cognitive, mood, and behavioral disturbances may be more common following STN than GPi DBS [33,34], but the results of the direct comparison trials have to some extent alleviated these concerns. (See 'Complications and adverse effects' below.)
Factors predictive of benefit — Preoperative levodopa responsiveness is a robust predictor of postsurgical improvement after STN DBS [35-37]. Symptoms that do not respond to levodopa are unlikely to improve following surgery (with the exception of tremor). Levodopa responsiveness is assessed by comparing the UPDRS or revised Movement Disorder Society-UPDRS (MDS-UPDRS) motor score between the "off" and "on" medication states, with a threshold of around 30 percent improvement needed to qualify for DBS . While there are no studies of predictive factors for benefit after GPi DBS, preoperative levodopa responsiveness is typically used at most DBS centers to evaluate candidacy for both STN and GPi DBS.
Duration of benefit — Long-term observational studies have been conducted for both GPi stimulation (with at least three to five years of follow-up) [39-41] and STN stimulation (with at least 10 years of follow-up) [42-45]. All of these studies demonstrate that DBS provides sustained antiparkinsonian benefit in many patients, although some patients will show further deterioration due to the unmodified, ongoing process of neurodegeneration.
In general, improvements in rigidity and tremor during "off" periods are maintained in the long term compared with the pre-DBS baseline. Control of dyskinesia and motor fluctuations is also generally maintained long term. By contrast, early improvements in bradykinesia and axial signs begin to decline by five years out, in line with progressive neurodegeneration [41-44]. Other levodopa-nonresponsive symptoms such as dysarthria, postural instability, freezing of gait, and cognitive function are not improved by DBS and continue to deteriorate over time [41,46]. In an observational study that included 104 patients who completed 10-year follow-up after STN DBS, the cumulative dementia incidence at 1, 5, and 10 years was 2, 11, and 26 percent, respectively, reflecting similar rates compared with the general PD population .
Complications and adverse effects — Potential complications and adverse effects of DBS can include the following:
●Surgical complications – A practice parameter from the American Academy of Neurology (AAN) reviewed adverse events from four studies [47-50] involving 360 patients (288 with PD) who had DBS . Adverse surgical events within one month of the procedure included death in 0.6 percent and permanent neurologic sequelae in 2.8 percent. Other surgical complications that did not result in permanent neurologic sequelae included infection (5.6 percent), intracerebral hemorrhage (3.1 percent), confusion/disorientation (2.8 percent), seizures (1.1 percent), pulmonary embolism (0.6 percent), cerebrospinal fluid leak (0.6 percent), peripheral nerve injury (0.6 percent), and venous infarction (0.3 percent) .
A separate systematic review of 29 studies of STN DBS (778 patients) found a higher reported rate of transient confusion (16 percent) and similar rates of intracerebral hemorrhage (3.9 percent), infection (1.7 percent), seizures (1.5 percent), and pulmonary embolism (0.3 percent) .
The battery life of the implanted pulse generator (IPG) generally lasts three to five years with chronic stimulation, depending on the stimulation parameters. While generally safe, IPG replacement is a surgical procedure, and complications, such as infection, are possible. Rechargeable pulse generators with a guaranteed life of at least 15 years are now available, minimizing the number of future IPG replacement procedures for patients.
●Hardware complications – Hardware problems following DBS are common. Electrode/wire replacement is required in approximately 5 percent of patients due to fracture, migration, or malfunction [37,51]. Other complications include lead misplacement requiring repositioning (2 to 3 percent), extension wire or implantable pulse generator malfunction (4 percent), hardware infection (2 percent), and allergic reaction to the hardware (<1 percent).
●Stimulation-related complications – Common stimulation-related complications of DBS include paresthesia, dysarthria, eyelid-opening apraxia, hemiballismus, dizziness, gait imbalance, dyskinesia, and facial contractions [37,51]. In general, these are mild and can be minimized or resolved with adjustment of stimulation parameters. Weight gain is reported in approximately 8 percent of patients undergoing STN DBS .
●Cognitive and behavioral complications – Most studies have reported only mild or no significant adverse effects of STN DBS on long-term neurocognitive function and mood, with the exception of declines in tests of verbal fluency [28,52-58]. However, a few studies suggest STN stimulation may be associated with more significant adverse cognitive, behavioral, or psychiatric problems, especially apathy [59-64]. Increased apathy in this setting may be triggered by a postoperative decrease in dopaminergic medication and may be alleviated by careful reintroduction or increased dosing of dopaminergic drugs [63,65].
Some studies suggest that cognitive and psychiatric disturbances may be more common with STN than GPi stimulation [33,34], although the evidence is inconsistent. One of these small trials found evidence of persistent cognitive changes in 2 of 10 patients assigned to STN DBS at 12 months and none of 10 patients assigned to GPi DBS . By contrast, the NSTAPS trial of 128 patients found no differences between STN DBS and GPi DBS at one year in cognitive, mood, and behavioral side effects [32,58].
The occasional psychiatric reactions seen in patients after STN stimulation have included transient psychosis, hypomania, pathologic gambling, and sexual disinhibition [59,60]. These reactions are similar to those seen in the dopamine dysregulation syndrome associated with dopamine agonist medications . (See "Initial pharmacologic treatment of Parkinson disease", section on 'Dopamine dysregulation syndrome'.)
●Risk of suicide – There is a concern that DBS, particularly of the STN, may increase the risk for suicide [9-13]. Although the risk is not firmly established , clinicians should screen DBS candidates for suicidality and depression and exclude those at high risk for suicide [11,15]. In addition, postsurgical follow-up should include continued attention to symptoms of depression and risk factors for suicide. Suicide risk assessment is discussed separately. (See "Suicidal ideation and behavior in adults".)
The largest report was an international survey of 55 centers with data for 5311 patients who had DBS targeting the STN . The suicide rate in the first postoperative year after DBS was 0.26 percent per year, 13 times higher than expected when compared with the suicide rate of general population controls matched for age, gender, and country of origin. Independent risk factors associated with suicide were postoperative depression, single marital status, and a previous history of impulse control disorders or compulsive medication use. These accounted for 51 percent of the variance for attempted suicide risk. Suicides were also associated with younger age, younger PD onset, and a previous suicide attempt. A smaller retrospective study identified similar preoperative risk factors among patients with a completed or attempted suicide attempt after DBS, including suicidality, psychotic symptoms, psychiatric mediation use, family psychiatric history, and higher frontal and depression scores on neurocognitive testing .
●Other considerations – At least one study has reported the safe use of DBS in a small number of patients with cardiac pacemakers , although concomitant use of the two systems is felt by some to be contraindicated because of possible electrical interference between the DBS pulse generator and the cardiac pacemaker. (See "Pacing system malfunction: Evaluation and management".)
DBS patients should avoid diathermy, a form of treatment that delivers heat to soft tissue using electrical energy. Physical therapists or chiropractors may use this modality to increase blood circulation, decrease pain, and improve healing. Diathermy can induce a radiofrequency current that causes heating of the DBS electrodes, resulting in permanent brain injury .
DBS patients may safely undergo magnetic resonance imaging (MRI) if certain requirements are met (eg, magnet strength no stronger than 1.5 Tesla, no impedance abnormalities with the DBS system).
CONTINUOUS LEVODOPA-CARBIDOPA INTESTINAL GEL INFUSION — Carbidopa-levodopa enteral suspension, commonly known as levodopa-carbidopa intestinal gel (LCIG), is approved by the US Food and Drug Administration (FDA) for use in patients with PD as a strategy to reduce clinically problematic fluctuations in dopaminergic drug concentrations and thereby minimize motor fluctuations. Patients who are too frail to undergo or who have no interest in DBS may choose LCIG as a potentially safer alternative.
LCIG is delivered as a continuous infusion (up to 16 hours per day) through a percutaneous gastrojejunostomy tube by battery-powered pump and may be used in place of oral carbidopa-levodopa to decrease "off" time. LCIG comes in the form of a cassette containing carbidopa-levodopa suspension. The infusion pump is carried from the neck or waist in a small bag.
Efficacy — Supporting evidence for the effectiveness of LCIG includes a 12-week randomized placebo-controlled trial of 71 patients with advanced PD in which all patients underwent tube and pump placement prior to treatment assignment . Patients and investigators were blinded to treatment allocation, and simultaneous titration of active and placebo therapy was done for patients in both groups to maintain the integrity of the blinding. Compared with intermittent dosing of immediate-release oral carbidopa-levodopa, LCIG led to a significantly greater reduction in mean "off" time (4.0 versus 2.1 hours) and an improvement in mean "on" time without troublesome dyskinesia (4.1 versus 2.2 hours). In addition, many smaller observational studies have reported that LCIG treatment is associated with a mean 40 to 80 percent reduction in "off" time [70-72].
LCIG is also effective in reducing dyskinesia. In a 12-week open-label randomized controlled trial of LCIG in 61 patients with advanced PD (mean age 69 years), LCIG improved dyskinesia rating scale scores by 15 points compared with optimal medical care, which was an approximately 30 percent improvement from baseline . Improvements in "on" time, activities of daily living, and quality of life were also observed, and the safety profile was similar to that seen in other studies.
Administration — Measurement of baseline homocysteine, methylmalonic acid, and B12 levels, with subsequent B12 supplementation as needed prior to starting LCIG, is suggested given the possible risk of neuropathy associated with LCIG. (See 'Adverse effects' below.)
An initiation visit, preferably with the patient in the "off" medication state, is typically scheduled a few days to two weeks after gastrojejunostomy tube placement. LCIG infusion is then started, replacing the patient's total daily oral levodopa dose. In most cases, LCIG is given as a morning bolus dose followed by a continuous maintenance dose for 16 hours. Extra bolus doses can be used as needed to manage motor symptoms not controlled by the maintenance dose, which can be increased by the clinician at the next clinic visit if patients require multiple extra bolus doses.
Many patients find adequate 24-hour control of symptoms with the daytime infusion only. Patients who experience "wearing off" overnight may benefit from a bedtime dose of oral levodopa, either immediate-release carbidopa-levodopa or a longer-acting formulation (controlled-release tablet or extended-release capsule). Rare patients (eg, those with nocturnal akinesia) may benefit from a 24-hour infusion regimen, in which case the continuous dose is usually reduced during sleep by approximately 30 percent.
Use of LCIG requires daily flushing of the tube at the end of administration. Sudden cessation of the smooth response to levodopa may occur if the J-tube extension is displaced into the stomach.
Adverse effects — Adverse events related to the procedure or device include abdominal pain, skin infection, peritonitis, pneumoperitoneum, gastric reflux, and aspiration [73,74]. In one report of 62 patients, postoperative wound infection occurred in 18 percent .
Approximately 5 to 10 percent of patients may develop a generalized polyneuropathy, which may develop acutely or subacutely following initiation of LCIG . Higher total daily doses of levodopa and higher serum homocysteine levels appear to correlate with the presence of peripheral neuropathy . While evidence is not conclusive, LCIG in particular might precipitate peripheral neuropathy more often than oral levodopa due to increased availability/absorption of levodopa and decreased intestinal absorption of vitamins due to the enteral suspension, both leading to higher homocysteine levels and lower folate, B6, and B12 levels .
CONTINUOUS SUBCUTANEOUS APOMORPHINE — Apomorphine, a dopamine agonist, can be administered subcutaneously by intermittent injection or continuous infusion for the treatment of motor fluctuations in PD. Apomorphine is more potent and effective than the oral dopamine agonists, with antiparkinsonian benefit that is comparable to that of oral levodopa. Intermittent injection as a rescue strategy for sudden "wearing off" is discussed separately. (See "Medical management of motor fluctuations and dyskinesia in Parkinson disease", section on 'On-demand rescue strategies'.)
Continuous subcutaneous apomorphine infusion (CSAI) is available in many European countries but not currently in the United States. Continuous apomorphine is administered by a small portable infusion pump through a fine-caliber infusion line and needle that is inserted subcutaneously in the fatty tissue of the abdomen, upper arm, or upper leg. Since no surgery is involved, CSAI is the least invasive of the three main device-assisted therapies for advanced PD. The pump can be worn attached to the waist or strapped around the neck.
Efficacy — Results from the double-blind TOLEDO trial indicate that CSAI is effective for treating motor complications of advanced PD . The TOLEDO trial enrolled 107 adults with advanced PD and randomly assigned them to treatment with continuous apomorphine infusion or placebo (saline) infusion. At 12 weeks, the reduction in "off" time was greater for the apomorphine group compared with the placebo group (2.5 versus 0.6 hours), and the difference (1.9 hours, 95% CI 0.6-3.2) was considered clinically meaningful. In addition, the apomorphine group experienced an increase in "on" time without troublesome dyskinesia of approximately two hours. Active treatment was generally safe and well tolerated; the most common side effects were skin nodules at the infusion site, nausea, and somnolence. (See 'Adverse effects' below.)
Administration — Prior to starting CSAI, some experts advise obtaining an electrocardiogram (ECG) to exclude prolonged QT syndrome and arrhythmias, which are potential complications of domperidone premedication as well as apomorphine itself. In addition, a direct antiglobulin (Coombs) test and a complete blood count (at baseline and repeated every 6 to 12 months) are suggested to monitor for hemolytic anemia and hypereosinophilia syndrome, which are potential uncommon complications of CSAI [79,80].
Premedication with domperidone 10 mg three times per day (where available) can be used three days before the start of CSAI and continued for one to several weeks to reduce the occurrence of nausea and postural hypotension until tolerance is achieved . The European Medicines Agency recommends that domperidone ordinarily should not be used for longer than one week, because of a possible risk of QT interval prolongation and arrhythmias . Trimethobenzamide 300 mg three times daily is an alternative in countries where domperidone is not available. It is usually continued for no longer than two months, since trimethobenzamide may increase the risk of somnolence, dizziness, and falls in patients treated with apomorphine .
The use of ondansetron and other serotonin receptor antagonists is contraindicated with apomorphine, as the combination may cause severe hypotension and loss of consciousness . In addition, dopamine antagonists used to treat nausea and vomiting such as prochlorperazine and metoclopramide should be avoided, as they may reduce the effectiveness of apomorphine.
A challenge test dose of apomorphine must precede routine use of CSAI due to risk of hypotension. This is usually done with a 2 mg subcutaneous injection under medical supervision and monitoring of standing and supine blood pressure before the injection, and repeated at 20, 40, and 60 minutes after.
The dosing of CSAI can be divided into a morning dose (if needed), a maintenance dose, and extra bolus doses given as needed . In most cases, the optimal maintenance dose of apomorphine ranges from 4 to 7 mg/hour and is delivered for 16 hours each day . Bolus doses of apomorphine (eg, 4 mg) can be given by injection pen or through the pump as needed in the morning and during the day.
One goal of CSAI therapy is to substantially reduce or stop levodopa, particularly if troublesome dyskinesia are present, although this process may take weeks to months . Other antiparkinsonian medications (beginning with oral dopamine agonists) should be tapered gradually and stopped over five to seven days while the dose of CSAI is being titrated up.
Adverse effects — The most common adverse effect of CSAI is the development of skin nodules at injection sites, which may affect approximately 50 percent of patients [78,79]. In some cases, nodules may undergo necrotic ulceration or lead to eosinophilic panniculitis. Injection sites may also develop pain, bruising, and itching. Based upon expert consensus, possible strategies to reduce the occurrence of skin complications include rotating the location of infusion sites, maintaining good skin hygiene, applying emollients at the infusion site, using Teflon needles, massaging the infusion site, and using silicone gel dressings .
Other potential adverse effects of CSAI include somnolence, nausea, vomiting, confusion, visual hallucinations, orthostatic hypotension, autoimmune hemolytic anemia, hypereosinophilia syndrome, yawning, rhinorrhea, and impulse control disorder. Dopamine agonist treatment has been associated with rare sleep attacks , and patients should be cautioned about driving or operating machinery. Dose-related QT prolongation has been demonstrated at therapeutic doses of apomorphine .
LESIONING PROCEDURES — Several lesioning procedures have been studied in advanced PD, including thalamotomy, pallidotomy, and subthalamotomy. These procedures can be performed conventionally or using magnetic resonance imaging (MRI)-guided focused ultrasound (FUS).
Conventional lesioning procedures use stereotactic surgical and electrophysiologic techniques to locate the brain target before using radiofrequency to create a lesion. This may be an option for patients with advanced PD in resource-poor regions where device-assisted therapies are not affordable or available. MRI-guided FUS is a newer technique that creates a permanent lesion but does not require a surgical incision.
Conventional thalamotomy and pallidotomy — Unilateral thalamotomy can safely and effectively control medication-resistant tremor in selected cases of PD. However, it does not help bradykinesia, decrease dyskinesia, or improve motor fluctuations. As discussed above, it has largely been replaced by deep brain stimulation (DBS) but remains an option for patients who do not have access to DBS. (See 'Patients with refractory tremor' above.)
Unilateral pallidotomy is effective and relatively safe compared with medical therapy in patients with severe dyskinesia and motor fluctuations [85,86]. A randomized trial in 32 patients found that those treated with unilateral pallidotomy had improvements in tremor, rigidity, bradykinesia, gait, and balance compared with those treated medically, and subsequent follow-up found that the benefits of unilateral pallidotomy were sustained after two years . The greatest improvement occurred on the side contralateral to the lesion, but significant ipsilateral improvement was also observed for bradykinesia, rigidity, and drug-induced dyskinesia. Observational studies have shown that the effect of unilateral pallidotomy in suppressing contralateral dyskinesia can be maintained for many years [88,89].
A systematic review of unilateral pallidotomy found that the risk of permanent adverse effects was 14 percent, symptomatic infarction or hemorrhage occurred in 4 percent, and the associated mortality was 1 percent . Bilateral pallidotomy carries a high risk of permanent pseudobulbar speech and swallowing impairment in patients already compromised in these functions .
Conventional subthalamotomy — Unilateral subthalamotomy can be effective for the treatment of advanced PD, but the existing evidence comes from several uncontrolled and largely unblinded studies . Compared with bilateral subthalamotomy, unilateral subthalamotomy appears to be as effective and has been associated with fewer adverse effects but produces shorter-lasting ipsilateral improvement . A prospective study comparing bilateral subthalamotomy, bilateral subthalamic nucleus (STN) DBS, and unilateral subthalamotomy with contralateral STN DBS using unblinded ratings showed no significant differences among treatment groups in efficacy or adverse effects when assessed 12 months after surgery .
The main concern regarding subthalamotomy has been adverse neurologic effects. Whereas persistent dyskinesia is rare after STN DBS, it is clearly common after subthalamotomy. In a case series of 89 patients who underwent unilateral subthalamotomy, 52 (58 percent) experienced early contralateral dyskinesia that resolved over 4 to 12 weeks in 38 cases . However, dyskinesia remained unchanged for 14 patients (16 percent) and was severe in eight of these, causing a combination of hemiballism and hemichorea that required a second lesioning procedure (pallidotomy) within four weeks to two years after subthalamotomy.
Accessibility, reduced cost, and lack of hardware problems may make subthalamotomy advantageous over DBS in some parts of the world, but subthalamotomy remains an experimental treatment for advanced PD and should be carried out only at experienced centers.
MRI-guided focused ultrasound — MRI-guided FUS is a newer, minimally invasive method that uses high-energy ultrasound beams to generate brain lesions . Although MRI-guided FUS has been marketed as "noninvasive," meaning it does not require surgical incisions, it should be considered an invasive procedure, as brain tissue is destroyed.
Like conventional lesioning procedures as well as DBS, the potential targets of MRI-guided FUS in patients with PD are the thalamus, subthalamus, and globus pallidus (GPi), and predominant symptoms determine the relevant target. All targets for lesioning procedures are unilateral. (See 'Procedure selection' above.)
●FUS thalamotomy – MRI-guided FUS thalamotomy received regulatory approval in the United States in 2018 to treat PD tremor based on the results of a randomized, sham-controlled trial involving 27 tremor-predominant PD patients . Long-term outcomes for this procedure are not yet established. Importantly, FUS thalamotomy has not been shown to be safer than other ablative methods; it is an invasive procedure that may cause adverse neurologic effects similar to those caused by conventional surgical stereotactic thalamotomy.
Similar to conventional thalamotomy, MRI-guided FUS thalamotomy treats only tremor, not bradykinesia or motor fluctuations. Patients who have contraindications to MRI, previous brain surgery, or high skull thickness are not appropriate candidates for this procedure. Patients who need bilateral control of their tremors are also not candidates because bilateral thalamotomy causes an unacceptably high rate of speech impairment. (See 'Patients with refractory tremor' above.)
●FUS subthalamotomy – MRI-guided FUS subthalamotomy is under investigation for the treatment of PD with markedly asymmetric motor features [96,97], but it does not have regulatory approval for this target. Further optimization of the target may be required to achieve an acceptable risk-to-benefit ratio compared with established options, such as DBS .
Data include results of a sham-controlled randomized trial of unilateral FUS subthalamotomy in 40 patients with markedly asymmetric PD and prominent tremor who were felt to be poor candidates for DBS or declined to undergo the procedure . Motor scores in the subthalamotomy group (n = 27) improved by approximately 50 percent over baseline at four months post-procedure, compared with an approximately 10 percent improvement in the sham group. Common adverse events in the active treatment group included speech disturbance (56 percent), gait disturbance (48 percent), dyskinesia (22 percent), and contralateral weakness (19 percent). Although most of these effects were transient, a small number of patients had persistent deficits at 12 months.
●FUS pallidotomy – Preliminary data describe results of pallidotomy for the treatment of levodopa-induced dyskinesia and cardinal motor symptoms of PD , but further trials are required to establish the safety and effectiveness of this target in patients with PD.
Gene therapy — There are two different approaches to gene therapy treatment for PD: symptomatic and restorative.
Symptomatic gene therapy approaches have included delivery of rate-limiting dopamine biosynthetic enzymes, such as tyrosine hydroxylase, amino acid decarboxylase (AADC), and GTP-cyclohydrolase 1, into the striatum, thus improving the synthesis and availability of dopamine to help PD symptoms. Delivery of glutamic acid decarboxylase (GAD) into the subthalamic nucleus (STN) is another approach. GAD is the enzyme that catalyzes the synthesis of gamma-aminobutyric acid (GABA), an inhibitory neurotransmitter, which could reduce hyperactivity of the STN resulting from the degeneration of dopaminergic nigral neurons in PD.
●One phase I/II open-label European trial reported the results of delivering AADC, tyrosine hydroxylase, and GTP-cyclohydrolase 1 into the putamen using a lentiviral vector . Three different dose levels of the viral vector were assessed in four patient cohorts. Seventeen patients were enrolled, but two withdrew before receiving treatment, leaving 15 patients for analysis. Off-medication Unified Parkinson Disease Rating Scale (UPDRS) motor scores were significantly reduced at 6 months (-33 percent) and 12 months (-31 percent) compared with baseline, but there was no difference seen between the different dose cohorts. Although this continued improvement in the UPDRS off-medication score was seen in six patients at the five-year follow-up , eight patients from the original cohort underwent deep brain stimulation (DBS) at different time points over the follow-up period, suggesting inadequate motor symptom control. The most common adverse events over the long-term follow-up period were dyskinesia and "on-off" phenomena.
●Several open-label trials using an adeno-associated viral vector (AAV) to deliver AADC to the putamen have been reported [102-104]. The two earlier trials used smaller infusion volumes and did not use real-time imaging, but demonstrated the therapy was safe and well tolerated [102,103]. The third phase I open-label trial used magnetic resonance imaging (MRI)-guided administration to achieve greater coverage of the putamen . Fifteen patients were enrolled in three cohorts of five. Each cohort had a different dose concentration or total volume. In addition to demonstrating greater coverage of the putamen, dose-related increases of AADC activity were observed using 18-F fluorodopa positron emission tomography (18F-FDOPA PET) imaging. Additionally, there was an increase in "on" time with nontroublesome dyskinesia of between 1.5 and 3.3 hours at 12 months based on patient diary, depending on the patient cohort, with corresponding decreases in "off" time and "on" time with troublesome dyskinesia. A phase II sham-controlled, randomized trial is underway.
●Based on potential clinical benefit seen in an open-label safety and tolerability trial using AAV to deliver GAD to the STN in 12 patients with PD , a randomized, sham-surgery controlled trial of AAV-GAD was conducted in 45 patients with advanced PD . Data from 37 patients were evaluated, 16 assigned to AAV-GAD and 21 assigned to the sham arm. Eight patients were excluded from analysis because of infusion failure or catheter tip misplacement. Compared with the sham surgery group at six months, the AAV-GAD group showed statistically significant improvement in UPDRS motor scores.
Restorative gene therapy approaches have delivered neurotrophic growth factors such as glial-derived neurotrophic growth factor (GDNF) or neurturin (a natural analogue of GDNF).
●Two trials of neurturin delivered via AAV have shown no difference in effects compared with sham surgery [107,108].
●GDNF is a more potent neurotrophic factor than neurturin, and a phase I open-label trial of AAV2-GDNF evaluated 13 patients at three different dose cohorts, under MRI-guided administration . The procedure was well tolerated and UPDRS scores remained stable at 6- and 18-month follow-ups. The majority of patients had enhanced 18F-FDOPA PET uptake at 6 and 18 months postinfusion, suggesting increased GDNF expression.
Neural transplantation — The rationale of neural transplantation is to place dopamine-producing cells in or near the sites of the brain where dopamine production is lacking.
Dopamine-producing cells can come from several different sources. Early randomized trials in PD used fetal-derived mesencephalic dopaminergic cells but found no clear benefit [110,111]. One of the main issues from these trials was that some patients developed "off" period dyskinesia that could not be controlled with medication adjustments, sometimes requiring DBS to reestablish motor control [110-112]. Despite these setbacks, these trials have shown proof of concept that fetal dopaminergic cells can be transplanted, grow robustly, and survive in vivo. Important information about the etiology and pathogenesis of midbrain dopamine degeneration in PD has also been generated. For example, postmortem analyses 11 to 16 years after transplantation in several patients with PD found Lewy body pathology in a small proportion of surviving transplanted neurons [105,112,113], suggesting that neural transplantation may be a treatment but not a cure.
Because fetal mesencephalic cells have limited availability, other sources for cells for neural transplantation are being explored. One alternative is the generation of dopamine neurons from human embryonic stem cells (hESCs). Grafts of dopamine neurons derived from hESCs have been demonstrated to survive in multiple animal models of PD , and human clinical trials of dopamine neurons from hESCs are in development . While tissue availability is less of a concern than for fetal-derived cells, hESCs still have ethical implications since they are generated from embryos. In addition, there is a need for immunosuppression in order for grafts to survive.
Reprogramming of human somatic tissue into autologous induced pluripotent stem cells (iPSCs) is also being investigated in PD. iPSCs can be generated from patients themselves and thus circumvent the limitations of fetal embryonic cells. Dopaminergic progenitor cells have already been generated from iPSCs and successfully implanted in nonhuman primate models of PD  as well as a first patient with PD via bilateral putaminal injections . Generation of patient-specific grafts poses high potential costs, and the generated cells would still retain the individual's genetic susceptibility to PD. Tumor formation from the transplanted grafts is also a possibility.
SOCIETY GUIDELINE LINKS — Links to society and government-sponsored guidelines from selected countries and regions around the world are provided separately. (See "Society guideline links: Parkinson disease".)
INFORMATION FOR PATIENTS — UpToDate offers two types of patient education materials, "The Basics" and "Beyond the Basics." The Basics patient education pieces are written in plain language, at the 5th to 6th grade reading level, and they answer the four or five key questions a patient might have about a given condition. These articles are best for patients who want a general overview and who prefer short, easy-to-read materials. Beyond the Basics patient education pieces are longer, more sophisticated, and more detailed. These articles are written at the 10th to 12th grade reading level and are best for patients who want in-depth information and are comfortable with some medical jargon.
Here are the patient education articles that are relevant to this topic. We encourage you to print or e-mail these topics to your patients. (You can also locate patient education articles on a variety of subjects by searching on "patient info" and the keyword(s) of interest.)
●Basics topic (see "Patient education: Parkinson disease (The Basics)")
PATIENT PERSPECTIVE TOPIC — Patient perspectives are provided for selected disorders to help clinicians better understand the patient experience and patient concerns. These narratives may offer insights into patient values and preferences not included in other UpToDate topics. (See "Patient perspective: Parkinson disease".)
SUMMARY AND RECOMMENDATIONS
●For patients with Parkinson disease (PD) who experience motor complications or refractory tremor interfering with quality of life, device-assisted treatment options include:
•Deep brain stimulation (DBS)
•Continuous levodopa-carbidopa intestinal gel (LCIG) infusion delivered through a percutaneous gastrojejunostomy tube by battery-powered pump
•Continuous subcutaneous apomorphine infusion (CSAI) administered by a battery-powered pump (available in some countries but not currently available in the United States)
These treatments are useful for reducing "off" time and increasing "on" time without troublesome dyskinesia. Patients with PD who may be appropriate candidates for these treatments should be referred for evaluation by a movement disorder specialist. (See 'Patient selection' above.)
●The choice among device-assisted therapies for PD is based mainly on individual patient characteristics, availability, and patient values and preferences (table 1). There are no randomized, controlled trials directly comparing these treatments. (See 'Procedure selection' above.)
●DBS has the best evidence of efficacy and is less burdensome (apart from surgical lead and device implantation) because there are no lines, needles, or portable pumps to attend to. DBS is nondestructive, can be performed with low neurologic morbidity, and can be modified over time to deal with changing or progressing symptoms. Disadvantages of DBS include the requirement for stereotactic brain surgery for electrode placement; surgical pulse generator implantation; the introduction of hardware with risks of infection, hemorrhage, or mechanical breakdown; and the need for periodic reprogramming. (See 'Procedure selection' above and 'Deep brain stimulation' above.)
●LCIG infusion is effective for reducing "off" time and increasing "on" time, although there are fewer data from randomized, controlled trials compared with DBS. Drawbacks to LCIG include the need for surgical placement of the percutaneous gastrojejunostomy tube, daily tube flushing, diligent stoma care, the requirement to wear a portable pump, and adverse effects including infection. (See 'Procedure selection' above and 'Continuous levodopa-carbidopa intestinal gel infusion' above.)
●CSAI is likewise effective for reducing "off" time and increasing "on" time, does not require surgery, and is the least invasive of the device-assisted therapies. However, there is less supporting evidence from randomized, controlled trials compared with DBS. Drawbacks to CSAI include the need for subcutaneous needle placement, skin care at injection sites, and the requirement to wear a portable pump. Long-term use may be limited by adverse effects, particularly the development of skin nodules. (See 'Procedure selection' above and 'Continuous subcutaneous apomorphine' above.)
●Traditional lesioning procedures such as thalamotomy, pallidotomy, and subthalamotomy are potential options for PD with disabling motor complications where device-assisted therapies are not affordable or available. Magnetic resonance imaging (MRI)-guided focused ultrasound (FUS) thalamotomy is a minimally invasive method for performing thalamotomy that may be considered in selected patients with refractory tremor.
●Several investigational therapies, including gene therapy and neural transplantation, are in various stages of development and may become options for treating advanced PD in the future. (See 'Investigational therapies' above.)
ACKNOWLEDGMENT — The UpToDate editorial staff acknowledges Galit Kleiner-Fisman, MD, FRCPC, and Ludy Shih, MD, who contributed to earlier versions of this topic review.