A multicenter trial of ropinirole as adjunct treatment for Parkinson's disease

Dopamine agonists have been used as antiparkinsonian agents since 1974.¹ They offer several theoretical advantages over L-dopa treatment. First, they stimulate postsynaptic striatal dopaminergic receptors directly and do not have to be converted to dopamine by a degenerating pool of nigrostriatal neurons or regulated by a reduced number of striatal terminals. Second, they can be designed to stimulate preferentially a particular subset of dopamine receptors. Third, they have a longer half-life than L-dopa and do not compete with dietary amino acids for entry into the circulation and brain.^2,3

In addition, dopamine agonists do not undergo oxidative metabolism and thereby avoid the potentially toxic metabolites of L-dopa. Accordingly, many clinicians are concerned that L-dopa treatment in PD might promote oxidative stress and accelerate neuronal degeneration. In contrast, dopamine agonists have been shown to protect against oxidative stress in in vitro and in vivo model systems.⁴ These findings suggest that the introduction of a dopamine agonist combined with a reduction in L-dopa dosage might be advantageous.

L-dopa therapy is also associated with the development of motor fluctuations and dyskinesia that are a major source of disability to the majority of PD patients.^5,6 Preclinical studies have demonstrated that dopamine agonists are less likely to induce this adverse effect.⁷ Clinically, the addition of dopamine agonists to L-dopa therapy has been shown both to increase the effectiveness of dopamine replacement and decrease L-dopa-induced motor complications.^8-10

Dopamine agonists, as adjuncts to L-dopa, have achieved an established role in the treatment of PD.^11-14 They are, however, not as widely used as might be expected from their pharmacologic profile, which may be related to the difficulty of managing patients on combined therapy.

Ropinirole is a nonergoline dopamine agonist that is selective for D2 and D3 dopamine receptors. It has negligible affinity for a wide range of central nondopaminergic receptors, including α- and β-adrenoreceptors, 5-hydroxytryptamine receptor type 1, 5-hydroxytryptamine receptor type 2, benzodiazepine, and γ-aminobutyric acid receptors,¹⁵ which may contribute to a reduced adverse effect profile of ropinirole in comparison with other agonists.¹⁶ Ropinirole, as an adjunct to L-dopa in PD patients with motor fluctuations, has been shown in earlier trials to improve parkinsonism and to decrease time spent in the "off" state.¹⁷ In these studies, the L-dopa dosage was also reduced significantly with ropinirole therapy. To test whether we could substitute a dopamine agonist for L-dopa without inducing deterioration in PD or clinically meaningful adverse effects, we conducted a prospective, double-blind, multicenter, placebo-controlled trial of ropinirole as an adjunct to L-dopa in PD patients with motor fluctuations. This study is the first to utilize a planned reduction in L-dopa dosage to evaluate the efficacy of a dopamine agonist.

Methods. Design. The study was designed as a prospective, randomized, double-blind, placebo-controlled, 6-month trial of ropinirole as an adjunct to L-dopa/carbidopa (Sinemet; Merck and Co., West Point, PA, DuPont Pharma, Wilmington, DE) in PD patients with motor fluctuations. Patients who met entry criteria signed an institutional review board-approved informed consent form, underwent baseline evaluations and a placebo run-in period, and were randomized to either ropinirole or an identical-appearing placebo tablet. Sixteen medical centers in the United States participated in this study.

Entry criteria. Patients included men and women with PD who were Hoehn and Yahr stage II through IV in the "off" state and who had evidence of a good response to L-dopa complicated by predictable motor fluctuations with or without dyskinesia. All patients had to have been receiving stable doses of immediate-release Sinemet (Sinemet), controlled-release Sinemet (Sinemet CR), or a combination of the two for a minimum of 4 weeks before entry into the trial.

Patients were excluded from the study if they suffered complex "on-off" phenomena or "yo-yoing," an abrupt and unpredictable loss of efficacy unrelated to the timing of L-dopa administration.⁵ Anticholinergic, amantadine, or selegiline treatment was permitted if the dose was stable for at least 4 weeks before entry, and was not increased during the study. Other dopamine agonists (bromocriptine or pergolide) were stopped at least 4 weeks before initiation of the trial. Also excluded were women of childbearing age; patients with a diastolic blood pressure of more than 110 mm Hg; patients taking antiarrhythmic medications, vasodilators, calcium channel blockers, beta blockers, or other antihypertensive agents (except diuretics); and patients with syncopal episodes, psychosis, dementia, or uncompensated heart, lung, liver, kidney, or endocrine disease. Patients with clinically significant medical or laboratory dysfunction were also excluded.

Run-in period and randomization. Patients who completed screen evaluations underwent a 7-day placebo run-in period before randomization. During the run-in period, patients were trained to recognize their "on" and "off" states and to record them in a home diary. Patients who were unable to complete the diary were excluded from participating in the study. Patients who completed the run-in period successfully were randomized according to a computer-generated schedule to treatment with either ropinirole or placebo in a 2:1 ratio. Patients were stratified as to whether they were taking selegiline because of the potential for selegiline to decrease the need for L-dopa-one of the primary outcomes in this study.

Study medication (ropinirole or placebo) and L-dopa. Ropinirole or an identical-appearing placebo was initiated at a total daily dose of 0.75 mg in three divided doses. Study medication was increased gradually in 0.75-mg/day increments until a dose of 3.0 mg/day was reached over approximately 2 weeks. Thereafter, the daily dose could be increased by 1.5 mg each week to a total dose of 9.0 mg/day and by 3.0 mg/day each week to a maximal dose of 24 mg/day. All patients had to be titrated to a minimum dose of 7.5 mg/day. Further increases were at the discretion of the blinded physician.

The study protocol distinguished between planned and unplanned reductions in doses of L-dopa. When patients were titrated to a dose of 7.5 mg/day, the planned decreases in L-dopa were begun and the L-dopa dose was decreased by one half to one tablet of Sinemet or Sinemet CR. In patients receiving both Sinemet and Sinemet CR, the reduction in Sinemet CR preceded the reduction of Sinemet. For each increase in study medication, there was a corresponding decrease in L-dopa. If PD symptoms increased after the decrease in L-dopa, the study medication could be increased. If PD symptoms did not resolve after two increases in study medication, the L-dopa dose could be increased. If a patient had adverse effects related to excess dopaminergic stimulation, the L-dopa could be decreased as an unplanned dose reduction. If the adverse effects did not resolve, then the study medication could be decreased. All changes in medication were performed by a blinded investigator.

Evaluations. Patient evaluations were performed at baseline and at each study visit. The planned and unplanned L-dopa dose was recorded, and patients were evaluated according to the Clinical Global Impression (CGI) and home diary reflecting hours "on" and "off" during 2 consecutive days before the visit. All evaluations were performed by the same blinded evaluator.

Efficacy. The primary end point measure of efficacy was the number of patients who achieved a 20% or greater decrease in L-dopa dose and a 20% or greater reduction in the percent time spent "off" between the baseline and final visits. The dose of L-dopa was noted at each visit, and distinctions were made between a planned reduction, as per the protocol, and an unplanned reduction due to adverse effects such as dyskinesias. The planned and the total (planned and unplanned) reductions in L-dopa, as well as the change in the total daily dose of L-dopa, were analyzed separately as a secondary efficacy measure. Other secondary measures of efficacy were change from baseline to final visit in the percent of the waking day in the "off" state as determined by the home diary, as well as the proportion of patients rated as improved (score 1, 2, or 3) on the CGI.

Safety. All patients had a physical examination, blood chemistries, hematology, urinalysis, chest radiograph, and a 12-lead EKG before entry into the trial. At each study visit, blood pressure and pulse rate were measured with the patient supine for 5 minutes and after standing for 2 minutes. Chemistries, hematology, and urinalysis were repeated at weeks 4, 12, and 24, and the physical examination was repeated at week 24. All adverse experiences, reported spontaneously on general questioning or observed directly by the investigator, were recorded.

Statistics. All patients were included in the analysis (an intent-to-treat analysis). If patients withdrew from the study, the last observation was carried forward to the 24-week end point. The primary efficacy measure (namely, the percentage of patients with a 20% or greater reduction in L-dopa dose and a 20% or greater reduction in percent time spent "off") was analyzed by logistic regression using PROC LOGISTIC (version 6.08; Statistical Analysis System [SAS] Institute Inc., Cary, NC). Estimates of treatment odds ratios with 95% CIs were calculated. Logistic regression, with estimates of treatment odds ratios and 95% CIs, was also used to analyze the percentage of patients rated as improved on the CGI and the percentage of patients requiring reinstatement of L-dopa. The reduction from baseline in the total daily dose of L-dopa and the change from baseline in the percent time "off" were analyzed by ANCOVA using PROC GLM and PROC MIXED in SAS (version 6.08). Separate analyses were performed for the decrease in L-dopa dose related to the titration schedule (the planned decrease) and for all decreases in L-dopa dose (planned and unplanned decreases). Adjusted means and the difference between treatment means and the corresponding 95% CIs were calculated. For all analyses, specific demographic and baseline disease severity parameters were included in the statistical model.

Results. A total of 149 patients met the entry criteria and completed the run-in successfully. Ninety-five of these patients were randomized to ropinirole and 54 to placebo. There were no significant baseline differences between the groups with regard to sex, age, disease duration, duration of L-dopa therapy, L-dopa dose, or disease severity (table 1). In the ropinirole group, 50.5% took selegiline, 25.3% took anticholinergics, and 15.8% took amantadine. Among patients on placebo, 55.6% took selegiline, 13.0% took anticholinergics, and 11.1% took amantadine.

Of the 149 patients who entered the trial, 109 (73%) completed it; 74 patients took ropinirole (77.9%) and 35 took placebo (64.8%). Fifteen ropinirole-treated patients (15.8%) and nine placebo patients (16.7%) withdrew because of adverse effects. Four ropinirole (4.2%) and eight placebo patients (14.8%) withdrew because of lack of efficacy. Two patients in each group withdrew for social or family reasons unrelated to efficacy or adverse effects.

Overall, 35% of patients treated with ropinirole and 13% of placebo-treated patients achieved a 20% or greater reduction in L-dopa dose and a 20% or greater reduction in the percentage of time spent "off" (p = 0.002; figure 1A). If the unplanned reduction in L-dopa dose in response to adverse effects was excluded, significantly more patients on ropinirole versus placebo still achieved a 20% or greater reduction in L-dopa dose and a 20% or greater reduction in the percentage of time spent "off" (27.7% versus 11.1%; p = 0.003). Patients randomized to ropinirole achieved a significantly greater reduction in total daily L-dopa dose (242 mg/day versus 51 mg/day; p < 0.001). The percent reduction in L-dopa dose was also significantly larger in the ropinirole group compared with the placebo group (31% versus 6%; p < 0.001; figure 1B). Additionally, significantly fewer ropinirole-treated patients required reinstatement of L-dopa back to or even beyond the baseline level in response to worsening parkinsonism (23.2% versus 42.6%; p < 0.001).

Selegiline had no significant effect on the number of ropinirole-treated patients who achieved a 20% or greater reduction in L-dopa dose and a 20% or greater reduction in the percent time spent "off" (17 patients who took selegiline versus 16 who did not take selegiline). In the placebo group, the number who achieved a 20% or greater reduction in L-dopa dose and a 20% or greater reduction in the percent time spent "off" was greater in the selegiline subgroup than in the nonselegiline subgroup (six patients versus one). In comparison with placebo, a greater percentage of ropinirole-treated patients improved on the CGI scale (58.5% versus 32.1%; p = 0.002; figure 2). A significant interaction between selegiline strata and treatment was observed for this variable. For patients receiving selegiline, the percentage of patients rated as improved was not significantly different from placebo (45.8% versus 43.3%), whereas in patients not receiving selegiline, there was a significantly larger percentage of patients in the ropinirole group designated as improved (71.7% versus 17.4%; p = 0.002; figure 2). In the ropinirole-treated group, patients had a significantly greater reduction from baseline in the percentage of hours spent in the "off" state during the waking day (11.7% versus 5.1%; p = 0.039).

Adverse effects reported by at least 15% of patients in either group are listed in table 2. Dyskinesia was the only adverse effect that increased significantly in the ropinirole group. However, of the 32 ropinirole-treated patients who experienced dyskinesia, 14 (43%) experienced the new onset or worsening of dyskinesia only during the fixed-dose phase of the trial, when it was not permitted to reduce the L-dopa dose. When the L-dopa dose was reduced, the dyskinesia was ameliorated. There were no clinically relevant changes on physical examination, vital signs, laboratory chemistries, or EKG in either group.

Discussion. This trial shows that ropinirole, as an adjunct to L-dopa/carbidopa, permits a reduction in L-dopa dose and provides enhanced antiparkinsonian benefits in PD patients with motor fluctuations. The study also confirms the safety of ropinirole as an adjunct to L-dopa in this population of patients. In comparison with placebo, a significantly greater number of ropinirole-treated patients achieved a 20% or greater reduction in L-dopa dose and a 20% or greater reduction in the percent time spent "off". A significant benefit in favor of ropinirole was still detected if unplanned dose reductions in response to adverse effects were excluded. Ropinirole patients as a group were also able to achieve a significant reduction in their total daily L-dopa dose. The presence or absence of selegiline did not influence the capacity of ropinirole to facilitate an L-dopa dose reduction. In addition, ropinirole-treated patients had significant improvement on the CGI and had a significant decrease in the percent time "off" during the waking day.

In this study, unlike in previous trials, a planned attempt was made to decrease the dose of L-dopa in conjunction with administration of ropinirole and before the development of adverse events. In only one previous trial⁸ was an attempt made to reduce the dose of L-dopa on the addition of an agonist and before the occurrence of adverse effects. Our trial differs in that a reduction in L-dopa dose combined with a reduction in time spent "off" was the primary end point.

The most frequent adverse effect observed with the addition of ropinirole to L-dopa was dyskinesia. However, for 43% of the patients with dyskinesias, new onset or worsening of dyskinesia occurred with a dose of ropinirole less than 7.5 mg-the dose in the protocol at which L-dopa could be reduced. With subsequent dose reduction, there was amelioration of dyskinesia. This observation is similar to that seen in other studies of agonists⁸ and presumably reflects the dopaminergic effect of the agonist in an L-dopa-treated patient.

Nausea, vomiting, and postural hypotension, the major early adverse effects of dopaminergic therapy, were not significantly different from placebo. This may reflect the low initial dose and the gradual titration schedule employed. Drowsiness, insomnia, and other adverse effects were also comparable in both groups. Confusion, hallucinations, and psychosis, which have been a problem in other trials of dopamine agonists,⁸ were infrequent in this study. This may partially reflect patient selection (dementia was an exclusion criterion) or the possibility that ropinirole is associated with fewer mental changes than other agonists because of its selective receptor stimulation profile.¹¹ There was no difference in the percentage of patients on ropinirole versus placebo who discontinued the trial prematurely because of adverse effects.

The capacity of ropinirole to reduce the dose of L-dopa was selected as one of the primary outcome measures of this study in fluctuating patients because of increasing evidence that both the dose and duration of L-dopa administration may contribute to motor complications. In the 1-methyl-4-phenyl-1,2,3,6-tetrahydropyridine (MPTP)-treated monkey, Bedard et al.⁷ showed that 14 of 14 levodopa-treated monkeys developed dyskinesia, whereas dyskinesia was seen in only 1 of 10 bromocriptine-treated animals, despite comparable levels of motor benefit. The development of dyskinesia has been attributed to postsynaptic alterations in dopamine receptor-regulated gene expression in striatal neurons that occur in response to pulsatile stimulation of dopamine receptors.^18,19 Because L-dopa has a relatively short half-life (60 to 90 minutes), it is thought to induce pulsatile stimulation of dopamine receptors, particularly in patients with advanced disease who have a reduced number of striatal dopamine terminals and a reduced capacity to buffer fluctuations in plasma L-dopa concentration. In contrast, dopamine agonists such as ropinirole have a longer half-life (approximately 6 to 7 hours) and are likely to provide more continuous stimulation of dopamine receptors. Motor complications are not seen with continuous infusion of L-dopa²⁰ and occur with short-acting dopamine agonists such as quinpirole or 4-propyl-9-hydroxynaphthoxazine.^21,22 Furthermore, the addition of long-acting dopamine agonists, coupled with a reduction in L-dopa dose, reduces motor complications in experimental animals.²³ Similar observations have been noted in clinical trials of PD patients. Several studies have now demonstrated that the incidence and severity of motor complications may be reduced in patients receiving combined therapy with low-dose L-dopa plus a dopamine agonist compared with higher doses of L-dopa alone.^8-10 Furthermore, high-dose L-dopa therapy may be associated with increased frequency and severity of motor complications.^24,25 Thus the combination of an L-dopa dose reduction coupled with the addition of an agonist is a rational approach to the treatment of patients with motor complications.

There is also concern that L-dopa may promote oxidative stress in PD patients. L-dopa can generate free radicals and other oxidative species through auto-oxidation or by way of its decarboxylation to dopamine.²⁶ In these reactions, L-dopa can promote the formation of the cytotoxic hydroxyl radical and consume glutathione, an essential antioxidative defense that is already deficient in PD.²⁷ L-dopa has been shown to be toxic to cultured and implanted embryonic dopaminergic neurons.^28,29 In addition, L-dopa has been shown to promote oxidative stress as indicated by an increase in oxidized glutathione, hydroxyl radical formation, lipid peroxidation, DNA damage, and inhibition of mitochondrial respiratory chain activity.^30-32 L-dopa, however, has not been shown to be toxic to nigral neurons when administered to normal rodents or humans.^33,34

In contrast, dopamine agonists have the potential to protect against L-dopa-mediated toxicity by permitting a decrease in the cumulative L-dopa dosage, by stimulating D2 autoreceptors and inducing a reduction in dopamine turnover,³⁵ and by recently demonstrated direct antioxidative effects.³⁶ This putative neuroprotective effect is illustrated by the ability of dopamine agonists to protect dopaminergic neurons in vitro and in vivo from the effects of aging and 6-hydroxydopamine toxicity.^37,38

Taken together, it is reasonable to consider therapeutic strategies in PD that involve the addition of a dopamine agonist coupled with a reduction in the dose of L-dopa. We confirm in our current study that the addition of the dopamine agonist ropinirole to L-dopa-treated PD patients with motor fluctuations permits a reduction in L-dopa dose and provides enhanced clinical benefits without causing untoward adverse effects.

Appendix