Summary

No available therapy has been shown to modify the course of Parkinson's disease. The 2018 International Parkinson and Movement Disorder Society evidence-based review found no clinically useful interventions to prevent or delay progression, and in the LEAP delayed-start trial (445 participants) early levodopa had no disease-modifying effect over 80 weeks. Treatment is therefore symptomatic, and each symptomatic class carries a trade-off.

Initial levodopa gives greater motor benefit than initial dopamine agonists, but long-term oral use is complicated by motor and nonmotor fluctuations in up to 80% of patients. In STRIDE-PD (747 participants), dyskinesia was present at week 134 in 42% of participants who started levodopa/carbidopa/entacapone and 32% of those who started levodopa/carbidopa. In the 5-year follow-up of LEAP, 46 of 160 early-start and 62 of 161 delayed-start participants had dyskinesia.

Adjunctive oral therapy for motor fluctuations yields off-time reductions near the 1-hour minimal clinically important difference. In a network meta-analysis of 74 randomized controlled trials (18,693 participants), the weighted mean reduction in daily off-time relative to placebo was 1.1 hours for dopamine agonists, 0.9 hours for MAO-B inhibitors, and 0.8 hours for COMT inhibitors, and all three classes increased the odds of dyskinesia (odds ratios 3.0, 1.6, and 3.3). Head-to-head evidence is limited. In the PD MED trial of initial therapy (1,620 participants), 179 of 632 participants (28%) assigned dopamine agonists and 104 of 460 (23%) assigned MAO-B inhibitors discontinued their allocated treatment because of side effects, compared with 11 of 528 (2%) assigned levodopa.

The currently available dopamine agonists act mainly on D2/D3 receptors and carry neuropsychiatric and sleep-related risks. In a cross-sectional study of 3,090 participants, impulse control disorders were present in 17.1% of those taking a dopamine agonist and 6.9% of those not taking one (odds ratio 2.72). In a longitudinal cohort of 306 participants without impulse control disorders at baseline, the 5-year cumulative incidence was 46.1% overall and 51.5% among dopamine agonist users. The pooled relative risk of somnolence with pramipexole or ropinirole compared with placebo was 4.98. Reducing or stopping a dopamine agonist can cause dopamine agonist withdrawal syndrome, which occurred in 5 of 26 participants (19%) in a retrospective cohort and 12 of 51 (24%) in a prospective cohort.

Device-aided therapies add procedural and tolerability burdens. In EARLYSTIM, serious adverse events related to surgical implantation or the device occurred in 17.7% of participants. Discontinuation with subcutaneous foslevodopa/foscarbidopa was 35.1% in the 12-week randomized trial and 44.8% in the 12-month open-label study.

No intervention prevents or delays disease progression

“There are no clinically useful interventions to prevent/delay disease progression.” (opens the source at this quote in a new tab)

Motor complications with long-term levodopa

“However, together with disease progression, long-term treatment with oral levodopa is complicated by motor and nonmotor fluctuations in up to 80% of patients.” (opens the source at this quote in a new tab)
“Survey data have shown that motor fluctuations are the most bothersome symptom of advanced PD.” (opens the source at this quote in a new tab)

Response oscillations and dyskinesia as drivers of continued development

“Levodopa’s long-term efficacy is, on the other hand, compromised by the development of response oscillations and drug-induced dyskinesias in a majority of patients due to the progressive degeneration of dopaminergic neurons and pulsatile administration of levodopa, and these limitations are the main reasons for continued efforts towards improvements of current dopaminergic strategies.” (opens the source at this quote in a new tab)

Dyskinesia and wearing-off prevalence after early or delayed levodopa initiation

“At 5 years, 46 of 160 patients in the early-start group and 62 of 161 patients in the delayed-start group experienced dyskinesia (P = 0.06).” (opens the source at this quote in a new tab)
“The prevalence of wearing off at 5 years did not differ between the early-start (58/160 patients, 36%) and delayed-start group (58/161 patients, 36%) (Supporting Information Table S3).” (opens the source at this quote in a new tab)
“However, the prevalence of any wearing off during the waking day at 5 years was 36% for both treatment groups and mainly consisted of patients with 1% to 25% of the time during the waking day spent in off drug state.” (opens the source at this quote in a new tab)
“In addition, there was no between-group difference at week 80 in the prevalence of dyskinesia, but there was a higher prevalence of early signs of motor response fluctuations in the delayed-start group.” (opens the source at this quote in a new tab)
“The results showed that motor complications are not associated with the duration of levodopa use, but that they are associated with disease duration and a higher levodopa daily dose.” (opens the source at this quote in a new tab)

Levodopa motor complications: pooled risk estimate at 4 to 6 years and its clinical significance

“Hence, it seems likely that the approximate 40% frequency of motor fluctuations and slightly less than 40% frequency of dyskinesias is a very reasonable risk estimate relating to 4–6 years of levodopa treatment.” (opens the source at this quote in a new tab)
“For example, in one series, dyskinesias and motor fluctuations were reported in over 90% of young-onset PD patients within 5 years.” (opens the source at this quote in a new tab)
“However, these findings should not be interpreted as estimates of clinically important levodopa motor complication morbidity.” (opens the source at this quote in a new tab)
“By 4 to 6 years, slightly more than one-third of these modern era patients had experienced dyskinesias (the median of 21 clinical series).” (opens the source at this quote in a new tab)
“As shown in Table 1, only three modern era series were identified that reported the cumulative dyskinesia frequency at very long durations of therapy (≥9 years); the median dyskinesia frequency in this group was nearly 90%.” (opens the source at this quote in a new tab)
“Hence, the clinically important morbidity may be substantially less than the frequency values tabulated herein.” (opens the source at this quote in a new tab)

Duration of levodopa therapy, modern era series	Dyskinesia, median %, unweighted	Dyskinesia, median %, weighted by N	Dyskinesia, range	Dyskinesia, total N	Motor fluctuations, median %, unweighted	Motor fluctuations, median %, weighted by N	Motor fluctuations, range	Motor fluctuations, total N
7–12 months	“7.5”	“7.0”	“0–38.9%”	“432”	“1.5”	“3.0”	“0–16%”	“112”
13–24 months	“25.9”	“28.7”	“10.3–35.4%”	“575”	“45.7”	“47.9”	“0–50.4%”	“700”
2.5–3.5 years	“28.0”	“26.9”	“13.2–40%”	“747”	“23.3”	“31.0”	“5–35.1%”	“849”
4–6 years	“38.5”	“36.2”	“8–64%”	“1,599”	“42.1”	“40.8”	“11.8–60%”	“1,817”
9–15+ years	“87.8”	“87.8”	“60.7–95.5%”	“514”	“69.6”	“69.6”	“40.9–82.9%”	“514”

Levodopa equivalent daily dose 3 and 5 years after early or delayed levodopa initiation

Visit	Early-start group, LEDD change from baseline, median (IQR)	Delayed-start group, LEDD change from baseline, median (IQR)	Estimated difference, square root transformed (SE)	P value
3 years	“400 (300–600)”	“450 (300–600)”	“−0.61 (0.69)”	“0.38”
5 years	“600 (400–800)”	“600 (450–800)”	“−0.16 (0.78)”	“0.84”

Dyskinesia risk with levodopa and dopamine agonist initiation strategies

“Long-acting dopamine agonists are associated with a reduced risk of inducing motor complications” (opens the source at this quote in a new tab)
“but have less efficacy than L-dopa and are associated with other side effects such as psychosis and impulse control disorders.” (opens the source at this quote in a new tab)
“Furthermore, nearly all dopamine agonist-treated patients eventually require L-dopa and develop dyskinesias at a rate similar to what is observed when L-dopa is initiated without dopamine agonists.” (opens the source at this quote in a new tab)
“Thus, the development of a treatment strategy that provides the benefits of L-dopa with reduced motor complications remains among the major unmet medical needs in PD.” (opens the source at this quote in a new tab)
“Further, higher doses of L-dopa in both LC and LCE groups were associated with increased dyskinesias (26.4% of 375 patients receiving ≤400mg/day of L-dopa compared with 45.4% of 370 patients receiving >400mg/day; p = 0.01).” (opens the source at this quote in a new tab)
“The study confirms that higher doses of L-dopa are associated with an increased risk of dyskinesia, and supports careful L-dopa dose titration, using the lowest dose of L-dopa that provides satisfactory clinical control.” (opens the source at this quote in a new tab)

Dyskinesia risk with adjunctive oral therapy

“All drug classes were significantly associated with dyskinesias when compared with placebo: COMT-Is OR 3.3 (95% CI 2.7 to 4.0), NNEH 5; DRAs OR 3.0 (95% CI 2.5 to 3.5), NNEH 7; and MAOB-Is OR 1.6 (95% CI 1.2 to 2.2), NNEH 24 (table 3).” (opens the source at this quote in a new tab)
“DRAs and MAOB-Is were significantly associated with increased occurrence of hallucinations: DRAs OR 3.2 (95% CI 2.4 to 4.1), NNEH 19; MAOB-Is OR 2.3 (95% CI 1.2 to 4.5), NNEH 78.” (opens the source at this quote in a new tab)
“First, the treatment duration was relatively short in most RCTs: in 43 publications, the treatment duration was less than 24 weeks.” (opens the source at this quote in a new tab)

Discontinuation of levodopa-sparing therapy because of side effects

“179 (28%) of 632 patients allocated dopamine agonists and 104 (23%) of 460 patients allocated MAOBI discontinued allocated treatment because of side-effects compared with 11 (2%) of 528 patients allocated levodopa (p<0·0001).” (opens the source at this quote in a new tab)
“For compliance, participants allocated MAOBI or dopamine agonists were significantly more likely to discontinue their allocated drug class than those allocated levodopa: 7-year probabilities were 72% for MAOBI, 50% for dopamine agonists, and 7% for levodopa (p<0·0001, figure 2).” (opens the source at this quote in a new tab)
“The side-effects (mainly psychological, sleep disturbance, and gastrointestinal) were usually mild, only 16 patients (nine given dopamine agonists, four given MAOBI, and three given levodopa) had serious adverse events believed to be possibly related to trial treatment.” (opens the source at this quote in a new tab)
“Participants allocated and still taking MAOBI or dopamine agonists were also significantly more likely than those allocated levodopa to need a drug from another class added to their treatment: 2-year probabilities were 64% for MAOBI, 40% for dopamine agonists, and 20% for levodopa (p<0·0001, appendix).” (opens the source at this quote in a new tab)

Limited comparative evidence to guide adjunct selection

“There are few comparative studies with long‐term follow‐up to support a particular choice.” (opens the source at this quote in a new tab)
“Over a median of 4.5 years of follow‐up, participants in the dopamine agonist group had a mean PDQ‐39 mobility score that was 2.4 points (95% CI −1.3 to 6.0 points) better than that of the combined MAO‐B and COMT groups; however, this difference was not statistically significant.” (opens the source at this quote in a new tab)
“It would be helpful in making a choice between the various options if more studies became available that compare treatments head‐to‐head.” (opens the source at this quote in a new tab)

Side-effect profile of non-ergot dopamine agonists and gaps in comparative safety evidence

“The major issue with DAs (at all disease stages) remains side effects.” (opens the source at this quote in a new tab)
“Overall, nonergot DAs have similar profile of side effects (sleepiness, postural hypotension, peripheral edema, and neuropsychiatric issues).” (opens the source at this quote in a new tab)
“In clinical practice, a significant side effect is the high risk of impulse control disorders (ICDs) with DAs compared to levodopa.” (opens the source at this quote in a new tab)
“Although lower rates of ICDs associated with long-acting or transdermal Das have been reported, to date there has been no interventional study evaluating the relative risk of ICDs between the DAs, and this remains an important area of research.” (opens the source at this quote in a new tab)
“Long-term comparative side effects between nonergot DAs, including risk of ICDs, is as yet unclear.” (opens the source at this quote in a new tab)

Impulse control disorders with dopamine agonists: cross-sectional study

“An ICD was identified in 13.6% of patients (gambling in 5.0%, compulsive sexual behavior in 3.5%, compulsive buying in 5.7%, and binge-eating disorder in 4.3%), and 3.9% had 2 or more ICDs.” (opens the source at this quote in a new tab)
“Impulse control disorders were more common in patients treated with a dopamine agonist than in patients not taking a dopamine agonist (17.1% vs 6.9%; odds ratio [OR], 2.72; 95% confidence interval [CI], 2.08-3.54; P < .001).” (opens the source at this quote in a new tab)
“Impulse control disorder frequency was similar for pramipexole and ropinirole (17.7% vs 15.5%; OR, 1.22; 95% CI, 0.94-1.57; P = .14).” (opens the source at this quote in a new tab)

Impulse control disorders with dopamine agonists: longitudinal cohort

“In 306 patients without ICDs at baseline, the 5-year cumulative incidence of ICDs was 46.1% (95% confidence interval [CI] 37.4–55.7, DA ever users 51.5% [95% CI 41.8–62.1], DA never users 12.4% [95% CI 4.8–30.0]).” (opens the source at this quote in a new tab)
“Lifetime average daily dose and duration of treatment were independently associated with ICDs with significant dose-effect relationships.” (opens the source at this quote in a new tab)
“ICDs progressively resolved after DA discontinuation.” (opens the source at this quote in a new tab)

Somnolence with dopamine agonists

“The pooled relative risk of somnolence in this analysis was 4.98 [95% confidence interval (CI) 1.79 to 13.89].” (opens the source at this quote in a new tab)
“Seven trials were included in the analysis of patients taking levodopa and pramipexole or ropinirole compared with those taking levodopa alone. The pooled relative risk was 2.06 (95% CI 1.47 to 2.88).” (opens the source at this quote in a new tab)

Dopamine agonist withdrawal syndrome: retrospective cohort

“Of 40 subjects treated with a DA, 26 underwent subsequent DA taper. Of these 26 subjects, 5 (19%) developed DAWS and 21 (81%) did not.” (opens the source at this quote in a new tab)
“All subjects with DAWS had baseline DA-related impulse control disorders.” (opens the source at this quote in a new tab)
“Dopamine agonists have a stereotyped withdrawal syndrome that can lead to profound disability in a subset of patients.” (opens the source at this quote in a new tab)

Dopamine agonist withdrawal syndrome: prospective cohort

“Twelve out of fifty-one patients (24%) met clinical criteria for DAWS, the most common symptoms of which were anxiety (91.7%), pain (50%), sweating (41.7%), and anhedonia (16.7%), after the withdrawal of a DA (ropinirole, pramipexole, or cabergoline).” (opens the source at this quote in a new tab)

Dopamine agonist withdrawal syndrome: review

“Dopamine agonist withdrawal syndrome affects even up to 24% of PD patients treated with DAs in whom the medication was tapered.” (opens the source at this quote in a new tab)
“Dopamine agonist withdrawal syndrome remains a therapeutic challenge as it is refractory to levodopa (LD), other PD therapies, anti-depressants, and anti-psychotics.” (opens the source at this quote in a new tab)
“Based on a thorough literature review, it may develop in around 8–24% of PD patients during DAs withdrawal” (opens the source at this quote in a new tab)
“The majority of the literature agrees that the major risk factors for the development of DAWS are as follows: high cumulative DAs exposure, high baseline dose (the higher the dose and the longer the duration of use, the greater the risk of developing DAWS), high peak DAs dose (≥ 150 mg LEDD), diagnosis of ICD in the past, increased vulnerability to addictive behaviors, and neurobiological causes such as mesocorticolimbic dysfunction” (opens the source at this quote in a new tab)
“Symptoms of DAWS can last for several months or even years or resolve spontaneously without management” (opens the source at this quote in a new tab)
“There is no specific treatment of DAWS with proven efficacy. Treatment in this syndrome should always be individualized, symptomatic, and long-term. This syndrome is refractory to LD. Reintroduction of DAs seems to be the only procedure alleviating the symptoms of DAWS” (opens the source at this quote in a new tab)

Unmet need for sustained, predictable relief with fewer adverse events

“Currently, available therapeutics for the treatment of Parkinson’s disease (PD) fail to provide sustained and predictable relief from motor symptoms without significant risk of adverse events (AEs).” (opens the source at this quote in a new tab)
“Even with the broad range of symptomatic treatment options that are currently available, patients with PD report significantly less treatment satisfaction than patients with other chronic diseases” (opens the source at this quote in a new tab)

Juvmo (tavapadon)

Limitations of current therapies