Neuroprotection in Parkinson's Disease: Are We Getting Close?
NEUROPROTECTIVE STRATEGIES
The primary objective of neuroprotective therapies is to combat underlying neurodegenerative mechanisms in order to halt or slow progression of PD and to promote the preservation, rescue, or restoration of lost neurons.45,95,96 Progress in the development of neuroprotective therapies has been hampered by several factors. Current animal models do not replicate the disease process of PD, it is difficult to translate medication dosing in animal models to patients with PD, there is no validated quantitative biomarker to assess disease progression in patients, and clinical trials do not assess mechanism of action and are often confounded by potential symptomatic effects of medications. To date, no therapies for PD have been proven to definitively exhibit neuroprotective activity in clinical trials. Several potential neuroprotective therapies, their possible mechanisms of action, and the clinical trials assessing them will be discussed (Tables 4 and 5).
MAO-B inhibitors
Monoamine oxidase metabolizes monoamine neurotransmitters and neuromodulators, including dopamine.95 The 2 isoforms of this enzyme, MAO-A and MAO-B, are distinguished by different substrate specificities and tissue distribution.97 Monoamine oxidase B is the primary form of the enzyme in the brain, accounting for greater than 80% of total MAO activity, and selective inhibition of this enzyme leads to elevated levels of dopamine in the brain.97 In vitro and animal studies have suggested that 2 irreversible MAO-B inhibitors, selegiline and rasagiline, have neuroprotective activity and have revealed several possible mechanisms by which this activity may be mediated (Table 4),95-97 including protection against neurotoxins, apoptosis, and free radicals and promotion of neuronal survival through the upregulation of neurotrophic factors. The protective actions of these inhibitors appear to be independent of their ability to inhibit MAO-B.98-100 Selegiline is metabolized to desmethylselegiline,101 which may be responsible for selegiline’s neuroprotective effects.102-105 However, selegiline is also metabolized to methamphetamines, which can interfere with its neuroprotective effects.106-108 In contrast, rasagiline does not have amphetamine metabolites,109 and its aminoindan metabolites may provide neuroprotection.107,110
Table 4. Potential Mechanisms of Neuroprotection
| MAO-B inhibitors | Dopamine receptor agonists |
|---|---|
|
|
MAO, monoamine oxidase.
Several clinical trials have attempted to assess the potential neuroprotective actions of selegiline in patients with PD (Table 5). The first reported study was a small double-blind, placebo-controlled trial in which treatment of patients with early PD with selegiline slowed clinical disease progression and delayed the need for levodopa.162 The DATATOP (Deprenyl and Tocopherol Antioxidative Therapy of Parkinsonism) study, originally designed as a randomized, placebo-controlled trial (Figure 4A), evaluated the efficacy of selegiline or tocopherol (vitamin E) or both in delaying clinical progression of early PD with the primary endpoint of time to need for levodopa.163,164 An early interim analysis demonstrated that selegiline significantly reduced the onset of disability and delayed the need for levodopa.163 However, subsequent analyses showed that selegiline had provided a small but significant symptomatic effect and raised the question of whether the results of the trial merely reflected this symptomatic relief rather than a neuroprotective benefit.164 The protocol was then modified such that all patients who had not reached the primary endpoint (requiring levodopa) discontinued study medication (selegiline or selegiline placebo) for 2 months and were then administered open-label selegiline. After the addition of open-label selegiline, patients from the original selegiline placebo group “caught up” to the selegiline group, as assessed by UPDRS scores.46 Thus, there was no sustained benefit observed for initial therapy with selegiline compared with placebo. Additionally, the findings that the initial advantages of selegiline were not sustained46 and mortality was not reduced165 further called into question the neuroprotective ability of this therapeutic agent.
The SIN-DEP-PAR (Sinemet® [carbidopa-levodopa; Bristol-Myers Squibb Company, Princeton, NJ]-Deprenyl-Parlodel® [bromocriptine; Novartis Pharmaceuticals Corporation, East Hanover, NJ]) study also attempted to examine the neuroprotective effects of selegiline but used a washout design (Figure 4B).166 In this study, patients with early PD were randomized to receive selegiline or placebo and were also treated with levodopa or bromocriptine. After a 2-month washout of selegiline study medication (selegiline or placebo) and a 1-week washout of levodopa and bromocriptine, patients treated with selegiline had significantly better UPDRS scores compared with placebo. This is consistent with a potentially neuroprotective effect of selegiline, but it is not clear how long selegiline must be discontinued to wash out all of its symptomatic effects. Because it is an irreversible inhibitor of MAO-B, the symptomatic effect may be prolonged (ie, weeks to months).
In a long-term, 7-year, double-blind study conducted by the Swedish Parkinson Study Group, patients with early PD were randomly assigned to treatment with selegiline or placebo, and levodopa was added as necessary.167 Results showed that patients treated with selegiline experienced significantly less clinical progression of disability, as assessed by UPDRS scores, and required significantly less levodopa.
In summary, the results of these studies are consistent with a potentially neuroprotective effect of selegiline. However, the DATATOP and SIN-DEP-PAR studies are confounded by potential symptomatic effects of selegiline. In the Swedish study, selegiline appeared to improve patients' long-term outcome, although the mechanism of action is not clearly defined.
Table 5. Clinical Trial Design and Outcomes for Assessment of Potential Neuroprotective Therapies
| Drug | Study | N | Trial design | Dose | Observations |
|---|---|---|---|---|---|
| MAO-B inhibitors | |||||
| Selegiline | Tetrud and Langston162 | 54 |
|
|
Slowed clinical disease progression on various scales and delayed need for levodopa |
| Selegiline | DATATOP46,163-165 | 800 |
|
|
Greater changes in UPDRS score improvements and slower rate of development of disability as assessed by UPDRS scores; differences not sustained; no decrease in mortality |
| Selegiline | SIN-DEP-PAR166 | 101 |
|
|
Smaller change in total UPDRS score with selegiline |
| Selegiline | Swedish Parkinson's Study Group167,168 | 157 |
|
|
Decrease in UPDRS scores, slower rate of progression of clinical disability as measured by UPDRS scores, delayed need for levodopa, lowered levodopa dose requirements, delayed progression of clinical disability in combination with levodopa |
| Rasagiline | TEMPO169,170 | 404 |
|
|
|
| Rasagiline | ADAGIO171,172 | 1176 |
|
Rasagiline 1 or 2 mg/d |
All 3 UPDRS-based endpoints met after 72 wk in 1 mg/d early-start group |
| Dopamine receptor agonists | |||||
| Pramipexole | CALM-PD86,91,92 | 82 |
|
|
|
| Ropinirole | REAL-PET90 | 162 |
|
|
|
| Ropinirole | Ropinirole 056173,174 | 268 |
|
|
|
| Dopamine replacement | |||||
| Levodopa | ELLDOPA47 | 361 |
|
|
Levodopa reduced severity of PD symptoms but accelerated reduction in uptake of [123I]β-CIT |
18F, fluoride 18; 123I, iodine 123; β-CIT, 2β-carbomethoxy-3-β-(4-iodophenyl) tropane; MAO, monoamine oxidase; mo, month; t.i.d., 3 times daily; UPDRS, Unified Parkinson's Disease Rating Scale; wk, week; y, year.
Figure 4. Various clinical trial strategies employed in the evaluation of neuroprotective therapies in the treatment of Parkinson’s disease. ITT, intent to treat.
In models of PD, rasagiline, a second-generation MAO-B inhibitor, has greater activity in both MAO-B inhibition and potential neuroprotection than selegiline.106,175-178 Early studies demonstrated the efficacy of rasagiline as monotherapy for symptomatic management in patients with early PD179 and as adjunctive therapy with levodopa for the treatment of motor fluctuations.180-182 The efficacy of rasagiline as a neuroprotective therapy has been tested in 2 delayed-start clinical trials (Figure 4C; Table 5). In contrast to standard randomized (Figure 4A) or washout (Figure 4B) trials, delayed-start trials are not necessarily confounded by short-term symptomatic benefits.169,171 In the TEMPO (Rasagiline Mesylate [TVP-1012] in Early Monotherapy for PD Outpatients) study, patients were randomly assigned to 3 groups—placebo for 6 months followed by rasagiline 2 mg/d for 6 months or rasagiline 1 or 2 mg/d for 12 months.169 At the end of the study, patients who received rasagiline for 12 months had less functional decline (smaller change from baseline in UPDRS scores) than those who received placebo for 6 months followed by rasagiline for 6 months (Figure 5). Results from long-term follow-up demonstrated that the early initiation of rasagiline provided significant benefit to 5.5 years of the 6.5 total study years.170 The ADAGIO (Attenuation of Disease Progression with Azilect® [rasagiline; Teva Neuroscience, Inc, Kansas City, MO] Once Daily) study employed a similar delayed-start design but was much larger and had longer treatment periods.171 The early-start groups were treated with rasagiline 1 or 2 mg/d for 72 weeks, while the delayed-start groups received placebo for 36 weeks followed by treatment with rasagiline 1 or 2 mg/d for 36 weeks. Recent results from this study indicate that patients treated with rasagiline 1 mg/d for 72 weeks experienced less clinical progression, as assessed by UPDRS scores, than patients who received placebo for 36 weeks followed by rasagiline 1 mg/d for 36 weeks.172 This indicates that rasagiline 1 mg/d provides benefits beyond a simple symptomatic effect and is potentially consistent with a neuroprotective action. Patients in the ADAGIO study are now being followed long term.
Figure 5. Mean changes in UPDRS scores for patients treated with 2 mg/d (pink), 1 mg/d (blue), or 2 mg/d delayed-start (purple) rasagiline. Patients in delayed-start group received placebo for 6 months followed by rasagiline for 6 months.169 UPDRS, Unified Parkinson’s Disease Rating Scale.
Click to enlarge the figure.
From Parkinson Study Group.169 With permission. Copyright ©2004 American Medical Association. All rights reserved.
Dopamine receptor agonists
Dopamine agonists directly activate striatal dopamine receptors. These include the ergot alkaloids (eg, pergolide, bromocriptine, cabergoline) and the nonergot alkaloids (eg, pramipexole, ropinirole, rotigotine). These agents are effective in the symptomatic treatment of motor features of PD52,96 and can delay the need for levodopa by 2 to 3 years.95 In vitro and animal studies suggest that they may also have a role as neuroprotective therapies with potential mechanisms for neuroprotection similar to those described for the MAO-B inhibitors (Table 4).95,96 Also similar to the MAO-B inhibitors, the neuroprotective actions of dopamine agonists may be independent of their action on dopamine receptors.131,133,136,142,148
Two clinical trials investigating the neuroprotective potential of dopamine receptor agonists have been reported (Table 5). In both the CALM-PD and the Ropinirole 056 studies,86,92,173,174 patients were randomly assigned to initial treatment with levodopa or a dopamine agonist (pramipexole or ropinirole, respectively) to which levodopa could be added as necessary. In both studies, UPDRS scores through 4 to 5 years were more improved in the levodopa groups. However, the dopamine agonist groups developed fewer motor complications, and in the CALM-PD study there was less decline on imaging assessments. Subsequent analyses demonstrated that the benefit of the dopamine agonists to delay motor complications appeared to be related to their ability to delay the need for levodopa, probably through their symptomatic effects. Further, as discussed above, there is concern that the imaging assessments in the CALM-PD study could have been affected by the treatments themselves, either directly or indirectly, and are, therefore, not considered reliable to assess disease progression in these trials. Ten-year follow-up of patients in the Ropinirole 056 trial found that UPDRS scores were not significantly different in patients who had initially been treated with ropinirole or levodopa.173 Thus, reliable evidence for neuroprotection from dopamine agonists in patients with PD is lacking.
Dopamine replacement
Levodopa has been the gold standard for symptomatic relief of PD. However, long-term treatment with levodopa is associated with motor complications, including wearing off and dyskinesias.51,52 Moreover, some in vitro and animal studies have suggested that levodopa may be toxic to neurons,183-188 and in vitro and animal studies have shown that this toxicity may be mediated by toxic reactive species and oxidative stress.155,184,188,189 In contrast, other studies have shown that levodopa is not likely toxic,187,190-198 may promote recovery of dopaminergic neurons,191,196 and may be neuroprotective.195 To determine whether levodopa is toxic, neuroprotective, or neither in patients with PD, the Parkinson Study Group conducted the ELLDOPA trial (Table 5).47 As discussed above, treatment with levodopa reduced the severity of PD symptoms compared with placebo, as measured by UPDRS scores (even after a 2-week washout), but was associated with a greater decline in [123I]β-CIT compared with placebo. Thus, the ELLDOPA study yielded conflicting results with regard to a possible neurotoxic versus neuroprotective effect for levodopa.
Other potential neuroprotective therapies
Other potential neuroprotective therapies have been described. N-methyl-D-aspartate (NMDA) receptor antagonists (eg, amantadine) have been used to provide improvement in PD symptoms.95 These agents appear to have a neuroprotective function, as they can protect neurons from hypoxia and excitotoxicity induced by NMDA and glutamate in vitro and in vivo,199-203 and exhibited a partial protective effect in a rat model of PD.204 To date, the only NMDA receptor antagonist to show clinical potential as a neuroprotective therapy is amantadine. A retrospective study identified increased survival in patients who received amantadine versus those who did not.205 Prospective studies are needed to explore possible neuroprotective effects of amantadine.
Creatine, a dietary supplement, is well tolerated and was shown to be nonfutile in a NET-PD futility study (Figure 4D).206 Further, a randomized, placebo-controlled trial reported that creatine treatment led to a reduced need for dose increase of dopaminergic therapies over 2 years; however, there was no improvement in UPDRS scores or results from imaging studies with creatine versus placebo.207 A phase 3 efficacy trial is currently under way to assess whether creatine improves global long-term clinical outcomes,208 whatever the mechanism of action.
The infusion or lentiviral expression of the neurotrophic factor glial-derived nerve growth factor (GDNF) in an experimental model of PD in rhesus monkeys revealed evidence for both protective and restorative actions of GDNF.209,210 Two small clinical trials evaluating putaminal infusion of GDNF yielded promising results for GDNF-mediated neuroprotection in patients with PD.211,212 However, a larger controlled trial found no clinical benefit with GDNF and was terminated early.213,214
