Current and experimental treatments of Parkinson disease: A guide for neuroscientists
Abstract
Over a period of more than 50 years, the symptomatic treatment of the motor symptoms of Parkinson disease (PD) has been optimized using pharmacotherapy, deep brain stimulation, and physiotherapy. The arsenal of pharmacotherapies includes L-Dopa, several dopamine agonists, inhibitors of monoamine oxidase (MAO)-B and catechol-o-methyltransferase (COMT), and amantadine. In the later course of the disease, motor complications occur, at which stage different oral formulations of L-Dopa or dopamine agonists with long half-life, a transdermal application or parenteral pumps for continuous drug supply can be subscribed. Alternatively, the patient is offered deep brain stimulation of the subthalamic nucleus (STN) or the internal part of the globus pallidus (GPi). For a more efficacious treatment of motor complications, new formulations of L-Dopa, dopamine agonists, and amantadine as well as new MAO-B and COMT inhibitors are currently tested in clinical trials, and some of them already yielding positive results in phase 3 trials. In addition, non-dopaminergic agents have been tested in the early clinical phase for the treatment of motor fluctuations and dyskinesia, including adenosine A2A antagonists (istradefylline, preladenant, and tozadenant) and modulators of the metabolic glutamate receptor 5 (mGluR5 - mavoglurant) and serotonin (eltoprazine) receptors. Recent clinical trials testing coenzyme Q10, the dopamine agonist pramipexole, creatine monohydrate, pioglitazone, or AAV-mediated gene therapy aimed at increasing expression of neurturin, did not prove efficacious. Treatment with nicotine, caffeine, inosine (a precursor of urate), and isradipine (a dihydropyridine calcium channel blocker), as well as active and passive immunization against α-synuclein and inhibitors or modulators of α-synuclein-aggregation are currently studied in clinical trials. However, to date, no disease-modifying treatment is available. We here review the current status of treatment options for motor and non-motor symptoms, and discuss current investigative strategies for disease modification. This review provides basic insights, mainly addressing basic scientists and non-specialists. It stresses the need to intensify therapeutic PD research and points out reasons why the translation of basic research to disease-modifying therapies has been unsuccessful so far.
This article is part of a special issue on Parkinson disease.
Abbreviations used
-
- DBS
-
- deep brain stimulation
-
- STN
-
- subthalamic nucleus
-
- VIM
-
- ventral intermediate nucleus
Parkinson disease (PD) is considered the most frequent movement disorder. Clinically, ‘movement disorder’ is defined by its cardinal motor symptoms: akinesia, rigidity, and tremor at rest. The major cause of these symptoms is the loss of dopaminergic neurons in the substantia nigra pars compacta, leading to a severe deficiency of dopamine in the putamen and the caudate nucleus. The neuropathological hallmark of PD is cytosolic Lewy bodies, which are characterized by aggregated α-synuclein. There likely is a pre-symptomatic period of several years, if not decades, in which neuropathological changes and subsequentially a depletion of dopamine are already present without the manifestation of clinical symptoms. Heiko Braak's work has revealed that aggregated α-synuclein is not only present in the substantia nigra but throughout the brain including the autonomic nervous system (Goedert et al. 2013). In fact, α-synuclein aggregates occur early in the autonomic nervous system, which is connected to the brain via the vagal nerve. According to Braak's staging hypothesis, the α-synucleinopathy of the brain spreads in a caudorostrally ascending, orchestrated fashion. The first α-synuclein-positive aggregates usually occur in the dorsal motor nucleus of the vagal and glossopharyngeal nerves or the olfactory bulb (stage 1). Stage 2 is characterized by the extension of the α-synucleinopathy to the medulla oblongata, the pontine tegmentum, and the locus coeruleus. By stage 3, the substantia nigra starts – albeit subclinically – to be affected and pathology also involves the amygdala. In stage 4, marked degeneration is observed in the dopaminergic neuronal population of the substantia nigra, Lewy pathology has extended into the temporal cortex and – latest at that stage – the cardinal motor symptoms become manifest. During stages 5 and 6, Lewy bodies and neurites occur in the neocortex, likely causing many of the cognitive symptoms associated with advanced PD. Braak's work has definitely changed the traditional view that PD was a disorder restricted to the well-known movement disorder characteristics. In fact, there most likely is a ‘prodromal’ symptomatic phase that precedes the movement disorder symptoms, including a disturbance of smell function, constipation, a rapid-eye-movement sleep behavior disorder and symptoms of depression (Berg et al. 2015). In later stages, many patients exhibit cognitive symptoms.
So far, the treatment of PD remains symptomatic. Although the treatment of the movement disorder symptoms offers various very effective medical and non-medical approaches, the treatment of the non-movement disorder symptoms is challenging and much less established. No registered treatment has proven beyond doubt to provide a disease-modifying effect. Because the term PD is defined neuropathologically, it comprises all pre-symptomatic and symptomatic stages of the disease whereas the term Parkinsonian syndrome refers to the movement disorder.
We here will review the state of the art in the treatment of PD patients and will give an outlook on treatments that are currently tested in clinical trials. Our major aim is to explain the treatment and challenges to neuroscientists studying neurodegenerative disorders rather than to the clinical expert.
Symptomatic treatment of the dopaminergic deficit
Aims of the therapy
Therapy of PD should take into account the time of onset and the duration of the disease; age, social situation and status of the patient; the efficacy of the therapies selected and the side-effect profile in the short- and long-term treatment. The therapy should target motor function, autonomic function, cognitive and communicative skills and psychiatric symptoms. It should aim to support self-sustainment in the activities of daily living. It should reduce or prevent disability and the need of support.
Therapy should preserve the patient's independence and social competence (family, society) and should try to maintain the capability to work and to maintain or regain health-related quality of life. Secondary complications such as orthopedic or internal medical comorbidities should be reduced to a minimum. In addition, therapy-induced motor and non-motor complications including dopaminergic adverse events should be avoided wherever possible.
Types of therapy
Treatment for PD includes pharmacotherapy, functional stereotaxic neurosurgery (deep brain stimulation), and supportive therapy such as physiotherapy, speech therapy, and dietary measures. All treatments available until 2016 are of symptomatic nature. No therapy is currently available that slows down the progression of PD or even to prevent its manifestation.
Established therapies
Oral L-Dopa therapy
The first symptomatic and the most effective treatment of the dopaminergic deficit that was established is L-Dopa, always in fixed combination with a decarboxylase inhibitor. This statement is based on decade long clinical experience and several comparative studies between L-Dopa and dopamine agonists. According to present knowledge, L-Dopa does not influence (delay) the progression of the disease. The effects of L-Dopa are dose-dependent. Based on in vitro data, there was an intense debate in the 1990s that despite its unequivocal symptomatic effects, L-Dopa might cause toxicity because of its oxidative metabolism and the potential to produce reactive oxygen species. However, well-designed animal studies and a study comparing patients who were treated with three different dosages of L-Dopa for 40 weeks followed by a wash-out period of 2 weeks clearly showed that L-Dopa dose-dependently mitigates the motor symptoms of PD and that these effects are still observed at 2 weeks after wash-out (Fahn et al. 2004). If L-Dopa induced toxicity, one would have expected that the placebo-treated group should have performed better after the 2-week wash-out period than the L-Dopa treated group. However, the opposite was the case, arguing for increased plasticity of the brain triggered by the patients’ enhanced mobility during the L-Dopa-treatment phase (Jung Kang and Auinger 2012).
Treatment with L-Dopa is recommended in all stages of the disease (early monotherapy, in non-fluctuating and fluctuating patients with and without dyskinesia (see Figure 1), in patients with motor complications and non-motor symptoms). This statement holds true for otherwise healthy as well as multimorbid PD patients. As uptake of L-Dopa from the duodenum into the blood and from the blood to the brain competes with the uptake mechanism for neutral amino acids, it is recommended to take the L-Dopa preparation about 1 h before or after a meal.
L-Dopa is in its effect and short-term side effect profiles superior as monotherapy to all other PD medications (amantadine hydrochloride, anticholinergics, dopamine agonists, MAO-B inhibitors). This is also expressed in a much better adherence to medication as compared with dopamine agonists and MAO-B inhibitors (Gray et al. 2014). The introduction of L-Dopa therapy has substantially increased the quality of life of patients with PD, but it has recently been questioned whether it also increases life expectancy (Macleod et al. 2014) by avoiding disease-related medical complications (pneumonia, fractures as a result of falls etc.).
Slow release L-Dopa preparations comprise a special galenic composition and therefore maximum plasma concentrations are achieved 1 or 2 h later than with oral L-Dopa standard preparations. Plasma half-time is very similar for all three preparations (standard, dispersible, and slow release). Oral standard preparations have a bioavailability of about 90%, while slow release preparations provide a bioavailability of about 70%. Therefore, slow release preparations are preferentially administered at night time – particularly to avoid potential interaction of the slowly released L-Dopa preparation with protein-rich food – in order to reduce nocturnal akinesia.
L-Dopa-Infusion therapy (Duodopa pump)
Intrajejunal infusion therapy with L-Dopa is registered for the treatment of PD in advanced stages with marked motor fluctuations (Olanow et al. 2014), which often coincide with marked dyskinesia. L-Dopa bound to a particular matrix is continuously delivered directly into the jejunum via a percutaneous tube fixated to the abdominal wall and connected to an externally carried pump (the distal end of the tube is located beyond the ligament of Treitz). This continuous application results in a constant L-Dopa plasma level, thus not only avoids pulsatile receptor stimulation, but also does not depend on the resorption from a normal gastric emptying. The jejunal L-Dopa therapy is generally recommended as a mono therapy. This treatment is very expensive and requires technical skills especially in the department of gastroenterology responsible for the precise placing of the tube into the jejunal cavity.
- Motor fluctuations and dyskinesia in advanced states of PD that cannot be controlled with oral pharmacotherapy,
- Sufficient social support for the nursing attention to the percutaneous endoscopic jejunal tubing.
A sudden loss of efficacy under jejunal L-Dopa infusion is most likely - a result of dislocation, disconnection, or obstruction of the jejunal tubing. This situation is a potential emergency as akinetic crises may result if L-Dopa therapy is not rapidly reinstituted. As a transient help, oral therapy during the rapidly initiated control of the application device should be performed in expert centers. If not available, a rapid switch to oral therapy should be initialized.
Dopamine agonists
Ten dopamine agonists (five ergot- and five non-ergot-derivates) are available for the treatment of PD. Ergot dopamine agonists subsume bromocriptine, cabergoline, α-dihydroergocriptine, lisuride, and pergolide. The non-ergot derivates include the oral substances piribedil, pramipexole (standard and slow release preparation), ropinirole (standard and slow release preparation), the parenterally applied apomorphine and the transdermally applied rotigotine (silicone-matrix-patch). Dopamine agonists are recommended to be used (1) in early (de novo) PD patients as mono therapy, (2) as adjunct therapy (2a) in the non-fluctuating and fluctuating PD patients together with L-Dopa therapy (intermediate state of PD) and (2b) in the advanced states of the disease. Particular caution is warranted if dopamine agonists are applied in the advanced PD stages at which the likelihood of cognitive impairment or even development of dementia increases. In patients with PD and cognitive impairment, dopamine agonists clearly have the potential to induce hallucinations or enhance existing hallucinations (Poewe 2008), thus should not be used or else be withdrawn slowly but completely.
The efficacy of dopamine agonists in the symptomatic monotherapy has been demonstrated in methodically sound, placebo controlled trials at least for the more recently registered preparations (Adler et al. 1997; Shannon et al. 1997). Also, efficacy in the early combination of L-Dopa with an L-Dopa-sparing effect, and the improvement of L-Dopa-associated fluctuations in advanced combination therapy of L-Dopa plus dopamine agonists, has been shown in studies with some of the preparations (Rinne et al. 1997; Parkinson Study Group 2000). In addition, initial treatment with dopamine agonists led to a lower occurrence of dyskinesia in the course of the following three to 5 years compared to an L-Dopa monotherapy (Parkinson Study Group 2000, Rascol et al. 2000; Oertel et al. 2006). However, a reduced severity of dyskinesia when initiating monotherapy with dopamine agonist and later adding L-Dopa has not been demonstrated in long-term follow-up studies (Parkinson Study Group CALM Cohort Investigators 2009; Gray et al. 2014). One trial with initial ropinirole monotherapy implied a reduced risk of developing dyskinesias in comparison to initial monotherapy with L-Dopa over a period of 10 years (Hauser et al. 2007). However, this is most likely because of a selection process since only a small minority of patients who tolerate monotherapy will remain in the trial (retainment). As per current knowledge, dyskinesia is not relevant to the quality of life of most PD patients in the first 4 years. Long-term trials covering 6–10 years comparing initial non-ergot dopamine agonist monotherapy with pramipexole versus initial L-Dopa monotherapy clearly shows that the group initial receiving the dopamine agonist is more likely to require supplementation of L-Dopa in order to achieve a sufficiently symptomatic therapy. Secondly, the occurrence of dyskinesia increases with time (Parkinson Study Group CALM Cohort Investigators 2009).
Thus, in summary, initial monotherapy with dopamine agonists is not superior to initial monotherapy with L-Dopa with respect to incidence of dyskinesia after adjustment to disease duration and daily L-Dopa doses of patients. To date, the claim that dopamine agonist therapy may slow down the progression of PD has not been confirmed (Schapira et al. 2013). Thus, agonist therapy is considered entirely symptomatic. The present long-term data clearly show that the initial advantage of reduced occurrence of dyskinesia under initial dopamine agonist therapy – when compared to L-Dopa – is lost over time because L-Dopa provides a better symptomatic efficacy and a lower potential of non-motor adverse effects at the same quality of life.
All dopamine agonists act on peripheral dopamine receptors and therefore can induce multiple and more severe side effects than L-Dopa. Adverse effects include nausea and orthostatic dysregulation. Leg edema can be the limitation for dosing under both ergot and non-ergot derivatives. Switching from one dopamine agonist to another is an accepted strategy as PD patients may react to a different dopamine agonist with a different side effect profile. Adverse effects of neuropsychiatric nature such as dopamimetic-induced psychosis, impulse control disorder, the dopaminergic dysregulation syndrome, or punding (complex stereotypically repetitive activities without directed goals) have to be specifically enquired about at each visit since patients tend to not report these most severe forms of adverse effects under dopamine agonist therapy, and even less so under L-Dopa therapy. Under safety aspects, two main side effects are most important: excessive daytime sleepiness for the non-ergot dopamine agonists, and pulmonary or cardiac valvular fibroses for the ergot dopamine agonists. The adverse effect profile of the latter condition has led to the consensus that ergot dopamine agonist should be used as a secondary therapy if therapy with a non-ergot dopamine agonist is not tolerated.
Although L-Dopa remains the ‘gold standard’ in the treatment for motor symptoms of PD, its use needs to take into account motor complications that develop over the course of the disease (Fig. 1). After an initial period with a continuous clinical response (‘honeymoon period’), motor fluctuations, wearing-off (the transition from ‘ON’- to ‘OFF’-state) and dyskinesia frequently occur. As documented in the above-mentioned placebo-controlled L-Dopa study in de novo PD patients, the 600-mg daily dose may induce wearing-off in up to 30% and dyskinesia in up to 17% of patients after as little as 40 weeks of treatment (Fahn et al. 2004). In a recent study comparing Italian PD patients treated early in the disease course with L-Dopa, and PD patients from Ghana who only received L-Dopa treatment after several years into the disease, the occurrence of motor fluctuations and dyskinesias depended on the disease duration but not on the duration of L-Dopa treatment (Cilia et al. 2014). In a multivariate analysis, disease duration and daily L-Dopa dose (in mg/kg body weight) were associated with motor complications, whereas the total cumulative dose of L-Dopa or the duration of L-Dopa treatment and the disease duration at the initiation of L-Dopa treatment were not (Cilia et al. 2014). Therefore, motor fluctuations and dyskinesia are not associated with the duration of L-Dopa therapy, but rather with longer disease duration and higher daily L-Dopa dose.
Over the last decades, several strategies to manage motor complications have been developed. The common approach has been to prolong the duration of striatal dopamine receptor stimulation that can be achieved with standard L-Dopa therapy. L-Dopa action at the, dopamine receptor is limited by the short plasma elimination half-life and the reduced capacity to produce (Fig. 2) and store dopamine in the synaptic vesicles of the degenerating dopaminergic nigrostriatal fibers. The latter leads to a continuous reduction of the long-term response during the disease course and also explains why the total disease duration and not the time of L-Dopa treatment correlates with the occurrence of motor complications.
Combination pharmacotherapy for motor complications
Pharmacological approaches to reduce 'OFF''-time are adjustments of L-Dopa dose and frequency, novel formulations of L-Dopa (continuous intrajejunal infusion of L-Dopa-carbidopa intestinal gel [Duodopa], and extended-release L-Dopa IPX066), and/or the add-on administration of a dopamine agonist (pramipexole, ropinirole, rotigotine, apomorphine). Furthermore the duration of the effect of a single L-Dopa dose can be prolonged (and the effect itself slightly increased) by blocking degrading enzymes of L-Dopa and/or dopamine, i.e. by adding a centrally active MAO-B-inhibitor (rasagiline, safinamide, selegiline) or a peripherally active COMT-inhibitor (entacapone, opicapone or tolcapone). The dopamine receptor agonists pramipexole, ropinirole, pergolide, and cabergoline delay the onset of motor complications when applied before L-Dopa treatment is initiated (Fox et al. 2011). However, this effect is lost after several years (Gray et al. 2014) and once L-Dopa is initiated (see above). Bilateral subthalamic nucleus (STN) stimulation at early (Deuschl et al. 2013) or delayed (Deuschl et al. 2006) time points of the disease as well as stimulation of the globus pallidus pars internus (GPi) (Odekerken et al. 2013) are alternatives to pharmacological approaches to treat motor fluctuations (see below). Treatment with amantadine, STN and GPi stimulation are efficacious in managing L-Dopa-induced dyskinesia (Fox et al. 2011). In this respect, the use of amantadine may increase, as a slow release formulation (twice a day) of amantadine (ADS-5102) has provided a marked decrease in dyskinesia in advanced L-Dopa-treated PD patients (Pahwa et al. 2016).
Autonomic side effects of L-Dopa and dopamine agonists comprise orthostatic hypotension, gastrointestinal disturbances such as nausea and vomiting. Neuropsychiatric adverse effects include excessive daytime sleepiness, dopamine-induced psychosis, impulse control disorder, punding, and the so-called dopaminergic dysregulation syndrome.
Deep brain stimulation
Deep brain stimulation (DBS) of the ventral intermediate nucleus (VIM), STN, and GPi has been developed for the treatment of the motor symptoms of PD particularly after the occurrence of motor fluctuations (STN, GPi) and for tremor at rest that is resistant to pharmacotherapy (VIM, STN). Motor fluctuations are characterized by rapid changes between good to excellent response (ON state) to dopamimetic therapies and poor (waning) response (OFF state – lack of sufficient level of dopamimetic medication in the CNS) resulting in recurrence of motor symptoms and – if excessive – in immobility. It was initially hypothesized that the electrical stimulation during DBS inhibited the output nuclei of the basal ganglia. This hypothesis was later revised toward increasing agreement that DBS overrides, close to the electrode, the disturbed neural network activity and thus disrupts pathological network oscillations. In this issue of the Journal of Neurochemistry, McIntyre and Anderson (McIntyre and Anderson 2016) excellently review the electrophysiological, cellular, and neurochemical effects of DBS and also discuss non-neuronal DBS effects.
Targets in the basal ganglia are the VIM, STN, and GPi. The VIM was the first target area for stereotactic lesion neurosurgery (thalamotomy) and later for DBS, and still is selected for the treatment of essential tremor. However, today, for monosymptomatic PD tremor or tremor-dominant PD, the STN is the target of choice because bradykinesia and rigidity will develop and increase during the course of the disease. Although in Europe, STN is mostly targeted, in other countries, GPi stimulation is considered equally adequate. GPi stimulation mainly abates dyskinesia but does not allow reducing the amount of medication (L-Dopa and dopamine agonists). By contrast, STN stimulation frequently leads to a reduction of medication of more than 50%. Therefore, it also helps to reduce some of the other, i.e. neuropsychiatric, side effects of pharmacotherapy. Furthermore, in patients with motor fluctuations, DBS is superior to the best available medical treatment early and late in the disease course (Deuschl et al. 2006, 2013 if Parkinson patients are carefully selected for DBS). Even though DBS is effective over many years as per the difference in symptoms between DBS ‘stimulation on’ and ‘stimulation off’, no protective efficacy against the progressive deterioration of the disease has been proven so far.
Treatment of non-motor symptoms
Today PD is not considered a pure movement disorder anymore, but a multisystem/multisymptom disorder. This is justified, as (i) REM sleep behavior disorder, hyposmia, depression, or constipation can occur years before the onset of the cardinal motor symptoms, (ii) these symptoms can later be followed by orthostatic hypotension and dementia, and (iii) all these non-motor symptoms occur in combination with symptoms that are directly related to the dopaminergic treatment, including psychosis, punding, and the dopaminergic dysregulation syndrome.
For all the substantial body of literature on therapeutic studies of motor symptoms and motor complications, there is hardly any controlled pivotal trial which addresses the various non-motor symptoms or addresses a non-motor symptom as the primary endpoint, i.e. indication for therapy in PD. Thus the treatment of non-motor symptoms is a field of unmet needs for PD patients and their families.
The poor development of this field is demonstrated by the following fact: It is generally accepted and common clinical practice that treatment of PD patients with acetylcholinesterase inhibitors (Burn et al. 2006), particularly the rivastigmine patch (Mamikonyan et al. 2015) and donepezil inhibitors (Aarsland et al. 2002; Ravina et al. 2005) leads to an improvement in cognitive performance and attention but also of neuropsychiatric symptoms such as to a reduction of hallucinations and psychotic episodes. However, only rivastigmine dispensed as a capsule – and therefore associated with a much higher side effect profile than its patch preparation – is officially registered for the indication dementia in PD in Europe and is reimbursed by the health insurances.
Failed attempts of disease modifying therapy
If a patient is asked, what is the most important unmet need in the diagnosis, therapy and care of PD patients, nearly every PD patient will respond: ‘Can you, the physician, slow down or even stop the progression of the disease?’
Some – being aware of the genetic contribution to the risk of developing PD – will even ask: ‘Can you prevent the disease in my children or siblings?’.
The first randomized, placebo-controlled trial that was powered to detect disease modification was the DATATOP trial, testing the MAO-B inhibitor selegiline and α-tocopherol in untreated de novo PD patients (for detailed review see (Schulz 2012)). Although α-tocopherol had no effect, patients treated with selegiline required symptomatic L-Dopa treatment later on (Parkinson Study Group DATATOP Investigators 1989). However, this effect was only temporary and disappeared after the wash-out of selegiline. In general, this was interpreted as a symptomatic and not a disease-modifying effect. Open extension studies supported this view (Parkinson Study Group DATATOP Investigators 1996a,b).
Rasagiline was developed as a new MAO-B inhibitor, and in addition to its MAO-B-inhibiting activity showed promising anti-oxidative, anti-apoptotic activities in vitro and led to the induction of neurotrophic factors (Schulz 2012). In clinical trials, it provided symptomatic effects (Parkinson Study Group 2005, Rascol et al. 2005). In addition to its symptomatic effects, it was studied with a delayed-start design for its potential disease-modifying effects in the ADAGIO trial (Olanow et al. 2009). Newly diagnosed patients with PD were randomized to 1 mg rasagiline, 2 mg rasagiline, or placebo once a day. After 9 months, patients who had received placebo were switched to 1 mg rasagiline or 2 mg rasagiline and patients who already were in a rasagiline treatment arm stayed on their treatment for another 9 months under double blind conditions. After 18 months, the 1 mg rasagiline treatment arm fulfilled all three pre-specified endpoints considered to reflect a disease-modifying effect, which are based on the change of the total Unified Parkinson Disease Rating Scale score; however, the 2 mg rasagiline dose did not. Thereby, only the results of the 1 mg rasagiline treatment arm implicated disease-modifying effects. In addition, results from the so-called TEMPO trial (Parkinson Study Group, 2004) could not be confirmed for the 2 mg arm. The specific reasons for this discrepancy are unknown. Possible explanations have been discussed in detail (Rascol et al. 2011; Schulz 2012). The Food and Drug Administration in the US did not approve for rasagiline a label for a disease-modifying therapy.
All other attempts to establish a disease-modifying therapy failed in phase 3 trials. Because early treatment with dopamine agonists reduces the amount of L-Dopa as well as motor complications during the disease course (as already discussed above), it was hypothesized that dopamine agonists might also have a truly protective effect and slow the progression of the disease. For pramipexole, the same delayed start study design as for rasagiline was used in a large phase 3 study (Schapira et al. 2013). However, the results did not imply any disease-modifying potential of the substance.
All recent phase 3 trials testing for disease-modification, including the use of coenzyme Q10 (Parkinson Study Group et al. 2014), creatine monohydrate (Kieburtz et al. 2015), pioglitazone (NINDS Exploratory Trials in Parkinson Disease FS-ZONE Investigators 2015), and gene therapy with neurturin (Olanow et al. 2015), failed to provide any evidence for disease modification.
New approaches in clinical trials
Despite an intense research focus on disease-modifying therapies, the traditional pharmacotherapeutic concepts are still worthwhile, as symptomatic therapy of motor and non-motor symptoms with dopaminergic and non-dopaminergic substances is far from optimal, especially in the intermediate to advanced stages of PD.
New symptomatic treatments
The following chapter will review compounds under way for symptomatic therapy of motor and non-motor symptoms. Details on the developmental stage of the compounds are provided in Tables 1-4. (see also above for the testing of slow release amantadine - Pahwa et al 2016).
| Compound | Company Sponsor | Indication | mode of action | Phase of Development | Commentary |
|---|---|---|---|---|---|
Melevodopa/Carbidopa Soluble tablet |
Sirio Chiesi | Motor | Modified form of l-Dopa | Approved | Only in Italy |
| Safinamide | Zambon | Motor wearing off | MAO-B-inhibitor, glutamate modulator add-on to l-Dopa | Approved | Reimbursed in EU no active comparator study to other MAO-B- inhibitors available |
| Opicapone | BIAL | Motor wearing-off | COMT-inhibitor add-on to l-Dopa | Approved | Long acting |
| Compound | Company sponsor | Indication | Mode of action | Phase of development | Commentary |
|---|---|---|---|---|---|
| l-Threo-Dops | Northera Lundbeck | Motor freezing | Noradrenaline precursor | 3 ongoing | Approved in USA |
| Istradefylline, KW-6002 | Kyowa Hakko Kirin | Motor wearing off | Adenosine 2A receptor antagonist | 3 positive | Approved in Japan Phase III ongoing in EU |
| Tozadenant | Biotie | Motor dyskinesia wearing off | Adenosine 2A receptor antagonist | 3 ongoing | |
Pimavanserin ACP-Nuplazid |
Acadia | Non-motor psychosis | 5HT2A inverse agonist | 3 positive | Approved in USA |
| Pitolisant | Bioprojet | Non-motor daytime sleepiness | Histamine autoreceptor antagonist | 3 completed | Data not released |
| Compound | Company Sponsor | Indication | Mode of action | Phase of development | Commentary |
|---|---|---|---|---|---|
| Aripiprazol | Otsuka | ICD and dyskinesia | D2-antagonist antidopaminergic | 4 positive | Not useful, as it induces marked akinesia with a latency of weeks |
| Donepezil | Eisai | Non-motor dementia in PD | Acetylcholinesterase-inhibitor | 3b ongoing | |
| Donepezil | Eisai | Non-motor gait, dementia in PD | Acetylcholinesterase-inhibitor | 3 ongoing | |
| Duloxetine | Univ. Toulouse | Non-motor pain | SSNRI | 3 ongoing | |
| Oxycodon/Naloxon | Mundi Pharma | Pain syndrome in PD | Opioid | 3 positive | |
| Zonisamid | Eisai | Motor UPDRS III wearing off | Antiepileptic drug | 3 positive |
| Compound | Company sponsor | Indication | Mode of action | Phase of development |
|---|---|---|---|---|
| Caffeine | Univ. Montreal, Canada | Motor Early PD |
Adenosine-receptor antagonist | Phase 3b ongoing |
| Isradipine | Novartis, NIH, Univ. Chicago | Motor Early PD |
Dihydropyridine calcium channel blocker | Phase 3b ongoing |
| Nicotine | German Parkinson Study Group/Parkinson Study Group US IPF, CNP, DPG, MJFF |
Motor de novo PD |
Cholinergic, modulation of α-synuclein-aggregation? | Phase 3b ongoing |
| Immunotherapeutic compounds | ||||
| Active immunization | Asfiris | Motor | Phase 2 ongoing | |
| Passive immunization | Biogen, Roche | Motor | Phase 2 in planning | |
| α-synuclein aggregation modulators | ||||
| NPT200-11 | UCB/Neuropore | Motor ? | Phase I in planning | |
| ANLE 138b | MODAG | Motor ? | Phase I in planning | |
In Table 1, new compounds are listed for which the mode of action is well known, as substances with identical or very similar modes of action are already approved (for example a further MAO-B-inhibitor or COMT-inhibitor). These developments aim at (1) symptomatic pharmacotherapy of (1a) motor or (1b) non-motor symptoms with predominantly dopaminergic mode of action.
In Table 2, compounds are listed which address a new therapeutic target, thus provide a novel mode of action and for which similar substances are not approved or only to a limited extent (for example adenosine A2 receptor antagonists). These new developments aim for symptomatic pharmacotherapy of motor or non-motor symptoms with a non-dopaminergic mode of action.
Table 3 lists compounds which are approved for other indications and are now tested for their efficacy and safety in the therapy of motor complications and of non-motor symptoms in PD by means of a non-dopaminergic mode of action.
New, potentially disease-modifying therapies
- There is no perfect animal model. Although the MPTP model and the 6-hydroxydopamine model are accepted and established to induce selective death of dopaminergic neurons in vivo and faithfully reflect the biochemical changes of neurotransmitters, they are quite acute models since they do not develop their toxic effects over weeks or months, nor show α-synuclein aggregates, nor reflect the caudorostral progression of pathology mirroring the ‘prodromal’ and ‘manifest’ pathological stages as defined by Braak. By contrast, transgenic α-synuclein mice show α-synuclein aggregates but almost no death of dopaminergic neurons. This is only and reliably achievable by virus-mediated over-expression of different forms of α-synuclein (for example, wild-type α-synuclein or A53T mutated α-synuclein) or by inoculation of either synthetic α-synuclein fibrils (Luk et al. 2012) or α-synuclein fibrils obtained from postmortem human tissue.
- Mainly cellular cell death mechanisms like excitotoxicity, necrosis and apoptosis, mitochondrial dysfunction and inflammatory processes have been addressed. These targets likely represent late stages of the disease and their treatment will not result in sufficient preservation of healthy neurons and neuronal networks, or will rescue only mildly affected neurons.
- Our increasing knowledge on the pathology of PD, α-synuclein aggregates, their progressive spreading in pre-specified stages and on pre-motor ‘prodromal’ symptoms that can precede the onset of motor symptoms for years (if not decades) clearly points out that for protective treatments to be successful, they need to be applied sufficiently early, probably in phases without motor symptoms on non-dopaminergic neuronal populations or even glia cells – a topic which has hardly been addressed so far.
- PD is unlikely to be a homogenous disorder. Although ‘idiopathic’ PD is defined by the occurrence of Lewy bodies and α-synuclein aggregates, there is clear evidence for a mitochondrial dysfunction. The autosomal-recessive mutations in the Parkin, Pink1, and DJ-1 genes point to a direct involvement of mitochondria and the electron transport chain. In fact, Pink1 was recently identified to phosphorylate serine 250 of NdufA10, a component of the complex I of the electron transport chain (Morais et al. 2014). This phosphorylation is essential for the ubiquinone reduction by complex I, and its absence leads to a reduction of complex I activity and of ATP generation. A complex I deficiency has frequently been reported in PD (Schulz and Beal 1994) and complex I inhibitors are used to model PD in cellular systems and animals alike. It is unlikely that all patients will benefit from the same treatment, rather some patients will benefit from mitochondrial enhancers and antioxidants, like coenzyme Q10, whereas others will benefit from therapies that interfere with the aggregation of α-synuclein. This heterogeneity of patients may be the reason for the failure of the phase III coenzyme Q10 trial (Beal et al. 2014). To test this hypothesis, a clinical trial with coenzyme Q10 in PD patients with a proven mutation in PINK1 should be performed.
In 1997, Polymeropoulus and coworkers described the first autosomal-dominant gene PARK1 that causes a mutation in the protein α-synuclein (Polymeropoulos et al. 1997). This finding was backed up by the subsequent discoveries of other genes that caused or represented a risk factor for PD. Based on this rapidly increasing knowledge, scientists and pharmaceutical industry slowly, but steadily moved away from classical targets for pharmacotherapy such as precursors of transmitters, transmitter receptor agonists, transmitter uptake blockers, or transmitter degrading enzyme inhibitors. Instead, research for new therapies started to focus on the basic biology and metabolism of α-synuclein and related proteins. This shift required a change in mindset and build-up of a new field of PD research including the establishment of new preclinical models and the search for new primary endpoints for progression during the different prodromal stages and in the very early motor manifestation of PD in clinical trials. This change and search is still ongoing. After 20 years, the first compounds have entered the early clinical phase of testing for PD modification (Table 4).
On the other hand, several compounds such as caffeine or nicotine are being tested which may offer the promise of modifying the course of PD, mainly based on epidemiological studies. The following chapter will deal with these developments (see Table 4).
Caffeine
Drinking coffee is associated with a reduced risk to develop PD (Costa et al. 2010). In addition, symptomatic benefits have been described in PD under caffeine intake (Wills et al. 2013). Controlled data on caffeine therapy in PD are limited. As of yet, a 6 week, placebo-controlled randomized study investigated the symptomatic effect of caffeine (200–400 mg/d) in PD (Postuma et al. 2012). Improvement in motor symptoms was observed in the Unified Parkinson Disease Rating Scale motor score (3.15 points compared to placebo), whereas the compound did not improve excessive daytime somnolence, the primary endpoint of the study. A long-term randomized controlled phase III study with a delayed start design (TEDA) has been initialized to reinvestigate the effect of caffeine on motor symptoms (primary endpoint) and several non-motor symptoms, also addressing the potentially disease modifying component of caffeine (clinicaltrials.gov/ct2/show/NCT01738178).
Basic science data support an effect on α-synuclein aggregation and an influence of caffeine on inflammatory processes in the gut.
Inosine
Inosine is a precursor of urate and its oral application leads to increase of serum urate levels. Urate possesses antioxidant properties in vitro and substantia nigra pars compacta dopaminergic neurons in rodents can be protected against 6-OHDA toxicity by elevated levels of urate (Gong et al. 2012; Zhang et al. 2014). Higher serum urate levels are associated with a decreased risk to develop PD (de Lau et al. 2005; Weisskopf et al. 2007). In addition, early PD patients with a higher plasma urate level present with a slower progression of disease (Ascherio et al. 2009). Based on this evidence, SURE-PD, a multicentre double-blind randomized controlled safety trial on inosine enrolled 75 early PD patients. These received either inosine (0.5 – 3 g/d) or placebo for 8–24 months (clinicaltrials.gov/ct2/show/NCT00833690). Inosine appeared to be safe and triggered increased urate levels in serum and CSF, previously associated with decelerated disease progression (Bhattacharyya et al. 2016). It remains to be demonstrated whether inosine (urate) possesses a disease-modifying potential in PD.
Nicotine
Nicotine as a potential treatment for PD has been reviewed in the past (Quik et al. 2008). Numerous epidemiological studies over the last decades have consistently shown an inverse relation between tobacco consumption and susceptibility to PD is less prevalent among smokers than among never-smokers (Morens et al. 1995; Gorell et al. 1999; Quik 2004). Tanner et al. (2002) even found a dose dependence of this effect: within twin pairs, the risk of PD was inversely correlated with the amount of cigarette smoking. Although some authors argue that this link between the environmental factor ‘smoking’ and PD may be – at least in part – because of an epiphenomenon caused by specific personality traits in persons susceptible to PD (Allam et al. 2004; Evans et al. 2006), the existing data do not sustain the reverse conclusion, i.e. that PD was a protective factor against smoking. In experimental neuroscience, nicotine has been shown to up-regulate anti-apoptotic proteins that can prevent or slow down neurodegeneration (Dasgupta et al. 2006), to induce enzymes of the cytochrome P450 family that can detoxify neurotoxins (Miksys and Tyndale 2006), and to protect against toxin-induced nigrostriatal degeneration in a non-human primate model of PD (Quik et al. 2006). An alpha-7-nicotine receptor agonist reduces dopaminergic neuronal nigral cell loss in the MPTP toxin-induced PD mouse model associated with a marked reduction in inflammatory parameters in midbrain. Nicotine is able to prevent aggregation of wild-type α-synuclein and of the A53T α-synuclein mutation in the test tube (Hong et al. 2009). In a culture system of pure human mesencephalic dopaminergic neurons (LUHMES), nicotine is able to reduce cell death induced by over-expression of human wild-type α-synuclein. (Höllerhage et al. 2013). These data support a disease-modifying potential. In addition, nicotine has also been shown to stimulate dopamine release (Tsukada et al. 2005), thus a symptomatic effect cannot be ruled out entirely.
Previous trials conducted in this field have merely focused on the symptomatic effect of transdermal nicotine in moderately affected PD subjects over an observation period of several weeks (Vieregge et al. 2001; Lemay et al. 2004). These clinical trials could not detect a clear symptomatic effect during nicotine treatment. Likewise, no effect on motor function was reported after short-term use of a nicotine chewing gum (Clemens et al. 1995). In contrast to these trials, one single published study using extremely high doses of nicotine over a period of 17 weeks reported motor improvement that allowed a reduction of the dopaminergic treatment (Villafane et al. 2007). Whether or not nicotine has any disease-modifying effect has not yet been investigated clinically. Based on the above described evidence, the study NIC-PD has been initiated, a randomized, placebo-controlled, double-blind multicenter trial in Germany and USA that tests for a potential disease-modifying effect of transdermal nicotine treatment. The primary endpoint is to demonstrate that transdermal nicotine treatment slows disease progression as measured by the change in total (part I, II, III) UPDRS score between baseline and after 52 weeks of study treatment plus a wash-out period of two more months, compared between nicotine and placebo treatment, respectively. Based on previous experiences in therapeutic trials in early PD patients, a deterioration of 4 points should be detected with a power of 0.8. The study enrolled 160 early PD subjects within 18 months of diagnosis and not expected to require dopamine agonist or levodopa therapy for 1 year, with a Hoehn and Yahr stage ≤ 2, and either with or without stable (> 2 months) MAO-B inhibitor therapy. Treatment was executed to 12 months, followed by a wash-out period of 2 months (3 weeks down-titration and 5 weeks run-out). Recruitment of NIC-PD is completed and first results are expected in 2017 (EudraCT No.: 2010-020299-42).
On the other hand, a recent large case–control study reported that PD patients have less difficulties to stop smoking than controls (Ritz et al. 2014). Thus, the negative association of PD and smoking may be explained by an alternative hypothesis, that the dependency of or addiction to nicotine may be lower (different) in manifest and possibly in prodromal PD than in controls.
Immunotherapy for α-synucleinopathy
α-synuclein is considered a key molecule in the neuronal death in PD (Karpinar et al. 2009). In addition, evidence increases that α-synuclein is involved in the spread and propagation of the neurodegenerative progression of PD by means of a prion-like mechanism (Angot et al. 2010; Luk et al. 2012). It is assumed that α-synuclein misfolds intracellularly and aggregates into toxic forms, enters the extracellular space and is subsequently taken up by an adjacent neuronal structure – leading to propagation of the toxic α-synuclein seeds. There is some evidence that the CSF content of α-synuclein in exosomes correlates with the disease phenotype and progression (Stuendl et al. 2016). Thus, active or passive immunotherapeutic strategies aim to reduce the level of extracellular toxic α-synuclein aggregates and thus to limit limit its propagation.
Active immunotherapy
In the active approach, short immunogenic peptides are administered subcutaneously. These peptides mimic the C-terminus of α-synuclein (Affiris, study code PD01A – clinicaltrials.gov/ct2/show/NCT02216188). Data accessible from a press release indicate that the vaccine was tested in two doses over a 12 months period in advanced PD in an open label design. In addition, half of the patients vaccinated developed anti-α-synuclein-peptide antibodies in serum and some even contained the antibodies in CSF. The vaccine was reported to be safe and well tolerated, and the patients of this trial now undergo long-term follow-up (clinicaltrials.gov/ct2/show/NCT01885494). In addition, a phase II trial is ongoing. A second group of peptides (Afiris, PD03 - clinicaltrials.gov/ct2/show/NCT02267434) is under assessment in a randomized placebo-controlled trial with early PD patients (Kalia et al. 2015).
Passive immunotherapy
Several companies actively pursue this strategy at present. One agent, a monoclonal antibody against α-synuclein (PRX002), has entered the clinical phase: a test in 40 healthy volunteers has been completed. Side effects were announced (although not formally published) to have occurred in < 10% of the healthy volunteers and no drug-related serious adverse reactions were observed. Interestingly, the treatment with the antibody led to a dose-dependent and rapid decrease in α-synuclein levels in serum. The sponsor has subsequently initiated a randomized clinical trial in PD subjects (clinicaltrials.gov/ct2/show/NCT021557714) (Kalia et al. 2015).
α-Synuclein aggregation modulators
Small molecules with the capacity to block, or at least modulate, the aggregation of α-synuclein to, e.g. toxic oligomers appear to be a promising approach in the search for disease-modifying therapy (Table 4). Two compounds are close to enter clinical testing: NPT200-11 (Szoke et al. 2014) and ANLE138b (Wagner et al. 2013; Levin et al. 2014). Both substances have been reported to modify and thus reduce the aggregation of α-synuclein in cell culture and in protein aggregation models in mice. In addition, both compounds are able to pass the blood–brain barrier. No data are so far in the public domain on results of their testing for toxicity in primates or any data on safety in humans. In addition, compounds to enhance autophagy to remove α-synuclein aggregates are under development.
The above list of compounds reflects the major shift in focus onto strategies for PD modification. The lack of primary endpoints which reflect the progression of PD in the prodromal stages, however, remains the crucial bottleneck in clinical therapeutic studies in a phase when the substantia nigra is not or only very mildly affected.
Thus, similar efforts – as presently placed on drug development – should be made in the methodological field for the search for new endpoints/parameters which are therapy-sensitive and which reflect the progression of the PD pathology both inside and outside the brain.
These combined efforts – in drug development and endpoint definition – may well lead the field away from the substantia nigra and its downstream connections. For example, evidence constantly increases that α-synuclein pathology is present in various parts of the gastrointestinal system (Stokholm et al. 2016), the submandibular gland (Vilas et al. 2016), or even in (mainly autonomic) fibers in the skin of PD patients (Donadio et al. 2015, Doppler et al. 2014).
Conclusion
We have to ‘Rethink Parkinson disease’ and have already entered a new area of diagnostic and therapeutic research, at the end of which hopefully stands the prevention of manifest (i.e. motor) PD. In the next decade, we will also see a rise in the design of personalized (precision) medicine for the treatment of PD, including symptomatic and disease-modifying approaches.
Acknowledgments
JBS is the current Editor-in-Chief of the Journal of Neurochemistry.
