Medical and Surgical Management of Advanced Parkinson’s Disease
ABSTRACT: Advanced Parkinson’s disease is char- acterized by the presence of motor fluctuations, various degree of dyskinesia, and disability with functional impact on activities of daily living and independence. Therapeutic management aims to extend levodopa ben- efit while minimizing motor complications and includes, in selected cases, the implementation of drug infusion and surgical techniques. In milder forms of motor complications, these can often be controlled with manipulation of levodopa dose and the introduction of supplemental therapies such as catechol-O-methyl transferase inhibitors, monoamine oxidase B inhibitors, and dopamine agonists including apomorphine. Clinical experience and evidence from published studies indi- cate that when these agents cannot satisfactorily control motor complications, patients should be assessed and considered for device-aided therapies. This review article summarizes some of the newer avail- able therapeutic opportunities such as use of enzyme inhibitors like opicapone and safinamide, adenosine A2A receptor antagonists, apomorphine and levodopa/carbi- dopa intestinal gel infusion, deep brain stimulation including the role of closed-loop and adaptive stimula- tion, and MRI-guided focused ultrasound.
Key Words: apomorphine; deep brain stimulation; dyskinesia; inhaled levodopa; levodopa infusion gel; MR focused ultrasound; motor complications; opicapone; wearing-off; safinamide
Conventional Pharmacological Strategies Levodopa
Although the precise cause of motor complications is not known, they are associated with variability in plasma levodopa concentration and nonphysiologic restoration of brain dopamine levels despite multiple daily administrations.8,9 Moreover, depending on age and levodopa dose, patients receiving higher doses of levodopa are more prone to develop both wearing-off and dyskinesia.10 Early use of levodopa doses greater than 600 mg per day or more then 5-6 mg/kg body weight has been associated with significant risk of motor complication.11-13
There have been attempts to develop a long-acting formulation of levodopa that might provide continu- ously stable plasma levels and reduce the risk of inducing motor complications. This has proven diffi- cult to achieve. Preliminary positive results have been observed with the “accordion pill,” a gastro-retentive formulation composed of pharmaceutical biodegrad- able polymeric films folded into an accordion shape in a standard-size regular capsule. A phase 3 study to determine whether the gastric-retentive accordion pill carbidopa/levodopa is more effective than the com- mercially available immediate-release carbidopa/levo- dopa in reducing motor fluctuations is currently ongoing.
A dry powder aerosol containing levodopa has been developed to treat predictable motor and refractory “off” periods in APD. Pulmonary absorption allows rapid presentation of levodopa to the absorptive mem- brane, which has a large surface area and low meta- bolic activity, thus avoiding the variability in gastrointestinal absorption. A recent study demon- strated rapid elevation of levodopa plasma levels along with a significant reduction in motor scores after 10 minutes, up to 60 minutes relative to placebo. The inhalation therapy was well tolerated, although the study duration was only 4 weeks.13
Dopamine Agonists
Several dopamine agonists are currently widely used (ropinirole, pramipexole, rotigotine). In general, dopa- mine agonists are less prone to induce dyskinesia than levodopa but have fewer anti-parkinsonian effects.14,15 They are used as monotherapy in early disease and as an adjunct to levodopa in more advanced cases.16 They permit a reduction in the dose of levodopa that is required to provide satisfactory antiparkinsonian effects but do not prevent the development of motor complications or the need for surgical interventions.
The value of dopamine agonists as adjunct treatment has been recently emphasized in patients who were on oral levodopa but required better motor control.17 These patients were randomized either to receive more levodopa or to add ropinirole extended release. The trial showed similar improvement in both groups but significant differences in dyskinesia risk (17% in those who added additional levodopa vs 3% in those who received the dopamine agonist). Long-acting oral for- mulations of pramipexole and transdermal ropinirole are also available.
The use of dopamine agonists has been somewhat tempered by their adverse event profile. In addition to dopaminergic side effects, there is also an increased risk of psychiatric problems, impulse control disor- ders, and sleep disturbances.There has been recent interest in the role of D1 dopamine agonists in treating PD based on laboratory studies suggesting that these agents may have more effective antiparkinsonian effects, may be beneficial for the nonmotor features of PD such as cognition and bladder function, and are less prone to cause dyskine- sia and possibly impulse control disorders. Clinical tri- als testing these agents is currently under way.
COMT Inhibitors
Catechol-O-methyltransferase (COMT) inhibitors have been employed based on their capacity to inter- fere with the peripheral metabolism of levodopa and thereby increase brain bioavailability of the drug. Tol- capone was the first COMT inhibitor to be introduced in the market and demonstrated prominent antiparkin- sonian effects.18 However, it is presently a drug of sec- ond choice because of reports of several cases of fatal liver toxicity.19 Entacapone is currently the most widely prescribed COMT inhibitor and has been dem- onstrated to superiority to placebo in extending on time and reducing off time, although less pronounced than with tolcapone but without evidence of liver tox- icity.20 The major adverse effects are related to increased dyskinesia (which can usually be controlled by downtitration of the levodopa dose), occasional cases of severe diarrhea, and discoloration of urine, which is of no clinical concern.21,22 The addition of entacapone increases the elimination half-life of levo- dopa, and it was hypothesized that frequent adminis- tration might provide more stable plasma levodopa levels and thereby reduce the risk of motor complica- tions.23 However, the drug was tested with a regimen that did not provide continuous plasma levels, and as a consequence patients randomized to levodopa/carbi- dopa/entacapone treatment actually developed more dyskinesia than those treated with levodopa/carbidopa alone. A new COMT inhibitor, opicapone, with a lon- ger half-life than entacapone has recently been approved and has the advantage of once-daily administration at a dose of 50 mg in the evening. Compared with both entacapone and tolcapone, opi- capone shows a much higher degree of COMT inhibi- tion in erythrocytes soon after intake and at 9 hours.24 Two pivotal double-blind studies have been performed: BIPARK I and II. BIPARK I25 compared opicapone in doses of 5, 25, and 50 mg daily with placebo and with 200 mg entacapone administered with each levodopa intake. Both entacapone and 50 mg opicapone were significantly more efficacious than placebo, with a trend for an even higher level of bene- fit with opicapone 50 mg. The most common adverse events were mild dyskinesia (16%), insomnia (6%), and constipation (6%), and in contrast to entacapone, patients did not experience urine discoloration or diar- rhea. In BIPARK II opicapone 25 and 50 mg/day were compared with placebo as an adjunct to levodopa26 and were shown to provide significant reduction in off-time. Once again, the most common adverse events were dyskinesia, constipation, and dry mouth. It remains to be determined if opicapone administration permits more stable levodopa plasma concentrations than other strategies and if this in turn can reduce the risk of motor complications.
MAO-B Inhibitors
Monamine oxidase (MAO)-B inhibitors have been employed because they inhibit the oxidative metabo- lism of synaptic dopamine and thereby extend its half- life and effect. MAO-B is primarily found in astrocytes in the striatum and degrades dopamine, whereas MAO-A is located within dopaminergic neurons. Although these agents primarily affect the MAO-B enzyme, they are not completely specific, and it cannot be excluded that some of the benefits relate to inhibi- tion of MAO-A and as well as to the -B isoform. There has also been interest in the potential of these agents to provide disease-modifying effects based on their incorporation of a propargylamine ring within their molecular structure. Experimentally, these agents have been shown to provide neuroprotective benefits against a wide variety of toxins.
The first MAO-B inhibitor to be registered was sele- giline. It is an irreversible inhibitor of the enzyme and was tested in the expectation not only of extending dopamine half-life but also of slow neurodegeneration by reducing oxidative stress in dopamine neurons.27 Broad use of selegiline has been challenged by con- cerns about cardiac safety possibly because of its metabolite methamphetamine, but this has never been confirmed.28 Rasagiline is a second-generation MAO- B inhibitor with higher affinity for the enzyme and few adverse effects.29 Rasagiline is indicated for the treatment of PD as monotherapy or as adjunct therapy (with levodopa) in patients with end-of-dose fluctua- tions. There are 2 pivotal studies in patients with motor complications, the PRESTO and LARGO stud- ies, both showing a reduction in off-time of about 1 hour.30,31 More recently, safinamide, a reversible MAO-B inhibitor with 1000 times higher affinity for MAO-B compared with MAO-A has been introduced to the market in doses of 50 or 100 mg once daily.32 Safinamide, also reduces glutamate release by blocking activity-dependent sodium channels and thus promot- ing calcium channel modulation. These findings sug- gest that safinamide may have antiparkinsonian and antidyskinesia effects. Pivotal studies testing safina- mide as an adjunct to levodopa demonstrated an increase in on-time and a decrease in off-time with associated improvement in quality of life.33,34 Impor- tantly, there was no increase in on time with dyskine- sia, and in 1 long-term prospective study there was a trend toward reduction in dyskinesia that was nomi- nally significant, with the 10-mg dose in the subgroup that had dyskinesia at baseline.35
Other Drugs
Amantadine, an N-methyl-D-aspartate receptor antagonist, is the only drug shown to improve dyski- nesia without worsening parkinsonism in double-blind trials.36 These results have led to the development of a new extended-release formulation of the drug (ADS- 5102), which is administered once a day at a dose of 274 mg (equivalent to 340 mg of amantadine HCl). Its efficacy in improving dyskinesia and reducing off- time has been reported in 2 prospective double-blind studies.37,38 Amantadine was generally well tolerated, although up to 20% of patients discontinued treat- ment (vs 7%-8% with placebo). Adverse events were mainly dry mouth, orthostatic hypotension and dizzi- ness, hallucinations, and peripheral edema.
Finally, adenosine A2A receptor antagonists have been developed to modulate the effect of A2A receptor activation on D2-bearing striatal neurons that is seen in PD and is thought to contribute to the development of motor complications.39 Their localization in the basal ganglia is restricted to GABAergic neurons of the indirect pathway, projecting from the caudate putamen to the external globus pallidus, which also selectively expresses the D2 dopamine receptor. Because it could be demonstrated that patients with PD who were treated with dopaminergic drugs and presented with motor complications developed more adenosine A2A receptors,40 it became appealing to test adenosine A2A receptor antagonists. Several clinical studies have been performed based on preclinical data suggesting that adenosine A2A receptor antagonist-like istradefylline could potentiate the benefit of dopamine agonists or levodopa and prevent the development of levodopa-induced dyskinesias. However, results have shown very weak and inconsistent benefit. Istradefyl- line is currently registered in Japan, and an application has also been filed in the United States, but results have been inconsistent, including those of the last double-blind study, which failed to meet the primary end point.41,42 Tozadenant had demonstrated signifi- cant antiparkinsonian benefits with no increase in troublesome dyskinesia in a phase 2b study, but the phase 3 study was recently interrupted because of the high rate of adverse events.43
Device-Aided Therapies
The term device-aided therapies is used to identify treatments requiring technical implementation and various levels of management complexity. Apomor- phine infusion and deep brain stimulation (DBS) have been available for more than 20 years, whereas levo- dopa duodenal infusion gel (LCIG) has been approved more recently. New subcutaneous levodopa delivery systems have recently been tested, whereas incisionless ablative techniques such as MRI-guided focused ultra- sound are now being used to treat specific symptoms like tremor in patients who are not candidates for DBS or other surgical procedures.
Duodenal Infusion of LD/CD Intestinal Gel (LCIG)
Since its initial development in 1991 as liquid solu- tion, levodopa intestinal infusion has undergone exten- sive research aimed at reducing the injected volume and making it more stable for commercial use. This eventually led to studies proving the safety and effi- cacy of the current LCIG formulation in the treatment of patients with advanced PD.44-46 The superiority of LCIG versus the best oral medical treatment in reduc- ing off time and extending on time without signifi- cantly worsening dyskinesia has been proven in a double-blind, double-dummy, double-titration trial.47 At the end of the 12-week observation period, there was a decrease in off time and an increase in on time in patients treated with LCIG of -1.91 and 11.86, respectively, compared with optimized oral levodopa.
Moreover, troublesome dyskinesia also improved in those patients presenting with >1 hour of involuntary movements at baseline.48 Overall, the study showed that motor changes were associated with better quality of life and reduced caregiver burden. Nonetheless, a
significant number of subjects experienced procedure- related complications, particularly in the first week of treatment, as well as the inconvenience of wearing an infusion system.49 Results of the double-blind trial were further strengthened by the additional open-label observation of improvements in nonmotor symptoms, particularly sleep, mood, and fatigue in the GLORIA study, in which motor and nonmotor benefits and improvements in dyskinesia were observed in more than 250 patients followed prospectively for 2 years.50
Some concerns have been raised by the observation of polyneuropathy in patients treated with LCIG. These reports have shown 2 general profiles of poly- neuropathy in LCIG patients: a less severe sensory axonal subtype that is slowly progressive and a less common subtype that clinically resembles acute inflammatory polyneuropathy (Guillain-Barr´e like syn- drome) and causes severe deficits.51,52 The overall inci- dence is estimated to affect approximately 5% of treated patients.
Apomorphine
Subcutaneous administration of apomorphine ensures rapid bioavailability, avoiding issues associ- ated with gastrointestinal transit time and first-pass liver metabolism. This approach has proven to be par- ticularly suitable for acute subcutaneous injection and continuous delivery using a pump system.
Acute subcutaneous apomorphine injections are highly effective for the treatment of acute off periods but are largely underused because of difficulties in assembling and administering the therapy in the pres- ence of severe disability during the off period. By con- trast, effects on motor fluctuations with continuous subcutaneous infusion are well established.53,54 Multi- ple open-label studies have shown a reduction in off time, extension of on time, and improvement in dis- ability and nonmotor symptoms. In particular, benefit for depression and, more recently, impulse control dis- orders has been reported in some patients.55,56 The effect on dyskinesia is more variable and depends on the possibility of increasing the infusion dose and reducing concomitant oral levodopa.57,58 The major adverse events with apomorphine infusion or injection relate to subcutaneous nodules, which could progress, causing ulcerations and thickening of abdominal skin. These are common but can be minimized with correct administration into the subcutaneous tissue, appropri- ate skin cleaning, and changing sites of injection, allowing treatment continuation for more than a decade.59
The first double-blind, placebo-controlled studies with apomorphine have only recently been completed (the TOLEDO study).60 The apomorphine infusion dose was individually titrated at a dose of 3-8 mg/hour and administered for 16 6 2 hours of the waking day. The study demonstrated the benefits of continuous sub- cutaneous apomorphine infusion in providing a signifi- cant and clinically meaningful reduction in off time of approximately 2 hours greater than placebo, with an equal extension of on time without troublesome dyski- nesia. The study also confirmed well-known reported complications, particularly skin nodules at the infusion site, which were present in 44% of treated patients, fol- lowed by nausea somnolence and skin erythema. Less common was the occurrence of hemolytic anemia, which required regular hematologic monitoring.
Although efficacious, subcutaneous apomorphine is not widely used, possibly because the parenteral admin- istration is inconvenient to use by many patients.APL-130277 is a sublingually administered apomor- phine film strip in clinical development that was recently tested for the treatment of both predictable and unpredictable off episodes. The sublingual formu- lation is rapidly absorbed through the oral cavity mucosa, allowing for rapid delivery. A recent phase 2 trial showed that 15 of 19 patients turned fully “on” within 30 minutes, with a mean duration of 50 minutes, suggesting that this new route of delivery may be more practical than subcutaneous injection.61 However, randomized studies are not needed to prove equivalent efficacy and better safety.
Deep Brain Stimulation
DBS of the subthalamic nucleus (STN) and the globus pallidus internus (GPi) is currently the most common and effective surgical procedure for advanced PD, with about 150,000 patients treated worldwide. Lesion surgeries have also been performed, such as tha- lamotomy, pallidotomy, subthalamotomy, and gamma- knife thalamotomy, but their indication in the present time is much more limited because of the risks associ- ated with a surgical lesion, particularly with bilateral procedures.62,63 Other surgical techniques, such as motor cortex and spinal cord stimulation, gene therapy (GAD-65/GAD-67, neurturin/GDNF), and cell trans- plants (embryonic dopamine neurons, retinal pigmental cells, carotid body glomus cells, and stem cells) are still under investigation, but none has demonstrated clinical benefits in double-blind trials.64,65 The use of other tar- gets for DBS (pedunculopontine areas, zona incerta, substantia nigra reticulata, other thalamic nuclei) is also currently being investigated.66
There is consistent evidence indicating that DBS of both the STN and GPi improve motor fluctuations and dyskinesia in advanced PD (Table 2).67-71 Benefits endure in long-term studies, but DBS does not prevent the development of nondopaminergic features such as gait/balance impairment and dementia.72 Adverse effects are observed with respect to the procedure, the stimulation system, and the stimulation itself.73 DBS has also been evaluated in patients with less advanced PD and been shown to provide benefits compared with a natural history control group receiving best medical treatment. Specifically, the EARLYSTIM trial was conducted in PD patients with relatively short dis- ease duration (mean, 7.5 years), recent onset of levodopa-induced motor complications ( 3 years), and preserved social and occupational functioning.74 Patients were randomized to receive bilateral STN DBS or best medical treatment. At the 2-year point,quality of life, motor disability, and levodopa-induced motor complications (time with good mobility and no dyskinesia) were significantly improved in the neuro- stimulation group compared with standard optimized medical therapy. Similar to previous studies, serious adverse events related to surgical implantation or the neurostimulation device occurred in 17.7% of patients.
Both STN and GPi DBS significantly improved motor fluctuations and dyskinesia.69-72 Advantages of the STN target may include greater medication reduc- tion, less frequent battery changes, and a more favor- able economic profile. Advantages of GPi DBS may include more robust dyskinesia suppression, less depression, easier programming, and greater flexibility in adjusting medications. Patients with STN stimula- tion could present with more mood, behavioral, and cognition changes after surgery.75,76
The most predictive factor of a good outcome with surgery is the preoperative response of motor signs to levodopa. Indeed, motor symptoms that do not respond to a preoperative levodopa challenge are unlikely to improve with surgical lesions or DBS.77-79 However, recent evidence suggests that the PD pheno- type could also predict DBS outcome.80 Tremor- dominant-type PD seemed to respond better than rigid-akinetic type to DBS and to GPi compared with STN.81 The relationship between genotype, pheno- type, levodopa response, and DBS outcome was stud- ied in a series of 94 PD patients who underwent DBS (STN, GPi, or ventral intermediate nucleus). A good number of patients (29%) tested positive for at least 1 of the currently known PD genes. Patients with Parkin mutations needed surgery later within the course of the disease, whereas glucocerebrosidase (GBA) muta- tion carriers needed DBS earlier and developed earlier cognitive impairment after DBS.80,82
There are currently several Food and Drug Adminis- tration- and European Union-approved DBS devices. These devices have similar technical characteristics, although there are differences regarding the duration of rechargeable implantable pulse generators (IPGs), IPG design, and other features. Some new IPGs allow reduction of the pulse width below 60 microseconds or the use of different frequencies at the same time, feature that could reduce stimulation-related side effects.83 Computed models have suggested that reduc- tion in electrode size and the ability to provide direc- tional stimulation could increase the efficacy of DBS. Electrodes with the possibility to steer current to avoid stimulating outside the DBS target and directional DBS leads could widen the therapeutic window com- pared with omnidirectional stimulation.71,84,85
Closed-Loop and Adaptive Stimulation
DBS systems operate through an open-loop mechanism, in which stimulation is delivered in a uni- directional manner using specific preprogrammed parameters. The same amount of energy is delivered constantly independent of the patient’s clinical state. As a consequence, overstimulation during some parts of the day may result in side effects such as dyskinesia, speech and gait worsening, and unnecessary battery drain, whereas understimulation at other times may result in inadequate control of PD features.84 Closed- loop stimulation systems are being developed to receive feedback regarding the clinical status of the patient and automatically adapt the output. The potential benefits of such adaptive DBS devices include reduction in stimulation-related side effects, more effi- cacious stimulation, and extended battery life.86-89
The concept of adaptive DBS (aDBS) is based on the possibility of recording local field potentials (LFPs) from the implantable DBS lead and correlating these LFPs with the patient’s clinical status (off and on, on dyskinesia). When compared with continuous DBS (cDBS), aDBS was 30% more effective in controlling PD features while using 50% less energy than cDBS. cDBS was compared with aDBS in 1 mobile PD patient.90 The aDBS portable device was equipped with an ad hoc algorithm to detect a patient’s LFPs beta-band power (13-17 Hz) and adapt voltage stimu- lation linearly. aDBS showed a more stable condition than cDBS, with better control of symptoms and dys- kinesias over time. The same group has recently shown that aDBS can reduce levodopa-induced dyski- nesia,88 aDBS was recently tested in a PD patient with chronically implanted DBS, and it was at least as effective as cDBS when objectively assessed.91 Double- blind studies to confirm these results in appropriately sized trials are required.
MR-Guided Focused Ultrasound
Noninvasive MR-guided focused ultrasound (FUS) thalamotomy provides anatomical localization and physiological verification for target localization before the final ablation procedure. Lesioning can be moni- tored in real time by single-section dimensional mag- netic resonance thermometry. Two pilot studies92 and one randomised control trial93 targeting the ventral intermediate thalamus in patients with essential tremor demonstrated consistent tremor improvements in the contralateral limb. A randomized study of active FUS versus a sham procedure was conducted in tremor- dominant PD patients showing greater improvement in the active surgical arm at 3 months.94 However, the number of treated patients was relatively small, the placebo effect in the sham group was not negligible, and side effects such as mild hemiparesis and ataxia were observed in a few patients. More recently, the first pilot study applying focal ultrasound to target the subthalamic nucleus in 10 PD patients has been reported.95 These initial results essentially mimic a previous report with radiofrequency lesion of the sub- thalamic nucleus, which was not associated with any relevant permanent side effects. Clearly, longer-term follow-up and double-blind studies are needed to con- firm efficacy and safety of FUS and to determine its potential role in the treatment of PD.
Conclusion
Motor complications can be disabling for PD patients, and their management is complex and requires experienced assessment to provide the best treatment choice. Currently available therapies include drug manipulation, infusion, and surgery. New approaches are being investigated, and the number of treatment options is increasing. In general, the best results are achieved with timely referral to tertiary centers that provide appropriate patient screening and selection and a multidisciplinary approach. This means that patients should be properly evaluated by a multi- disciplinary team of specialists that would ideally include movement disorders neurologists, neurosur- geons, neurophysiologists, psychiatrists, neuropsychol- ogists, nurse practitioners/coordinators, and physical therapists.96 This would also require the ability to pro- vide monitoring of efficacy to assess the benefit of the various therapies, and specifically to make the decision to move from conventional pharmacologic treatment to more invasive infusion or surgical treatments.
In this respect, the presence of APD features that are poorly responsive to dopaminergic therapy like gait and postural instability as well as cognitive dysfunc- tion deserves particular relevance not only as exclu- sion criteria for invasive treatments but also because nonpharmacological approaches like dedicated rehabilitative programs and cognitive training have demonstrated some benefit.97,98