Double-blind comparison of pramipexole and bromocriptine treatment with placebo in advanced Parkinson's disease

Ergoline-type dopamine agonists, such as bromocriptine,¹ pergolide,² lisuride,³ and cabergoline^4,5 have been used with levodopa therapy in Parkinson's disease (PD) treatment. Dopamine agonists, compared with levodopa, have lower symptomatic efficacy and more adverse effects.⁶ The discovery of new compounds that could enhance efficacy and reduce side effects in important to improve the therapy of parkinsonian patients.

Pramipexole is a synthetic aminobenzathiazol derivative and is a potent agonist of D₂ and D₃ dopamine receptors.^7,8 Preclinical trials have shown that pramipexole induced contralateral turning in 6-hydroxydopamine (6-OHDA)-lesioned rats and reversed parkinsonism in 1-methyl-4-phenyl-1,2,3,6-tetrahydropyridine(MPTP)-treated rhesus monkeys.⁷ The rank order of receptor affinity of pramipexole as D₃ > DF₂ > D₄ has been confirmed by saturation binding experiments.⁸ Currently it is not clear which dopamine receptor binding affinity profile is optimal for dopamine agonists developed for the treatment of PD. Pramipexole treatment suppresses prolactin secretion, confirming pharmacodynamic effects of this dopamine agonist in humans.⁹ Early clinical trials have been performed to assess the safety and efficacy of pramipexole treatment in early¹⁰ and advanced PD.¹¹

This trial compared the safety, tolerance, and efficacy of pramipexole versus placebo in advanced PD as an adjunct to levodopa therapy in patients with response fluctuations. To validate the study design, an active treatment group with bromocriptine was included to enable comparisons between bromocriptine and placebo.

Methods. This study design was a multicenter, controlled, prospective, double-blind, randomized, parallel-group trial in patients with advanced PD. A total of 34 sites were involved in six European countries and Canada. The duration of the trial was up to 9 months and 11 days. The three treatment groups were pramipexole, up to 4.5 mg per day; bromocriptine, up to 30 mg per day; and placebo.

Patient selection. Subjects enrolled in study were men or women who were at least 30 years old and had idiopathic PD with Hoehn and Yahr stages II to IV during an "on" period. Subjects had received an individual, optimized dosage of levodopa (and decarboxylase inhibitor) and were stable for at least 30 days prior to the initial administration of study medication. Immediate and controlled release preparations of levodopa were permitted in this study. All subjects experienced motor fluctuations characterized as end-of-dose phenomena or "wearing off" effects with their levodopa therapy. Subjects were required to keep an accurate daily record by keeping written patient diary forms of the "on" and "off" periods during waking hours twice a week for 2 weeks between each visit. Before enrollment all subjects gave written informed consent to participate.

Exclusion criteria included subjects with atypical parkinsonism; dementia that could impair participation in the study; psychosis except what was elicited by treatment with levodopa or dopamine agonists; history of a seizure within 2 years; clinically significant heart, liver, or kidney disease; and/or an elevation in either total bilirubin, alkaline phosphatase, lactic dehydrogenase, AST, or serum creatinine of more than 1.5 times the laboratory normal; retinopathia pigmentosa; presence of active neoplastic disease; surgery within 180 days of the baseline visit that would negatively impact the subject's participation; and stereotactic brain surgery. If at the screening visit subjects had a supine systolic BP of less than 100 mm Hg or evidence of a symptomatic drop of 20 mm Hg or more, measured at 1 minute after standing, they were not permitted to participate. Treatment with the following drugs during the month prior to administration of the trial medication was not permitted: alpha methyldopa, flunarizine, cinnarizine, parenteral ergot preparations, bromocriptine, pergolide, lisuride, monoamine oxidase inhibitors except deprenyl, methylphenidate hydrochloride, amphetamine derivatives, and beta blockers (unless used to treat parkinsonian symptoms) or treatment with neuroleptics or metoclopramide during the 2 months prior to administration of the trial medication. Females of childbearing potential not using a medically recognized means of contraception were not permitted to participate. Subjects who had electroconvulsive therapy within 90 days prior to administration of study medication were not included, as well as subjects participating in other studies of other investigational drugs within 30 days of baseline.

Clinical procedures. A screening assessment was performed that included a medical history and physical examination, modified Hoehn and Yahr Scale, supine and standing BP determinations, laboratory tests, chest radiograph, ECG, and a pregnancy test for fertile females. A baseline assessment was performed within 2 weeks of the screening assessment. The Unified Parkinson's Disease Rating Scale (UPDRS, parts I to IV), timed walking test, Parkinson Dyskinesia Scale, modified Schwab and England Scale, and modified Hoehn and Yahr Scales were performed. The UPDRS part III (motor examination) and Parkinson Dyskinesia Scale were performed approximately 2 to 3 hours after a dose of levodopa. Evaluation of the patient diary records, concomitant medication, adverse events, and laboratory test was also performed. Subjects had repeat assessments of their BP before medication was dispensed, and a Mini-Mental State Examination and quality-of-life assessments were performed with the Functional Status Questionnaire(FSQ)¹² and EuroQoL (European Quality of Life).¹³ The subject was then randomly assigned to one of the three treatment groups of pramipexole, bromocriptine, or placebo. After this visit, subjects were asked to take their study medication with food three times a day with a 6- to 8-hour interval between doses.

Ascending dose titration was accomplished by assessing subjects approximately every 2 weeks up to a maximum of 12 weeks after baseline(table 1). During these visits subjects had repeated motor assessments as performed at baseline, blood pressure evaluation, review of patient diary forms, and recording of any adverse events, concomitant medication, and compliance. Motor ratings were performed 2 to 3 hours after levodopa dose administration and 1 to 4 hours after the study medication was administered. Subjects were titrated to the maximal tolerated dose of study medication. If an adverse event occurred that did not respond to reduction of levodopa dosage, the subject entered the maintenance phase at the highest previously tolerated dose.

The duration of the maintenance dose period was 24 weeks. Visits occurred every 4 weeks and included the assessments performed during the ascending dose phase. In addition, laboratory tests and ECGs were performed at the beginning and end of the maintenance phase. The quality-of-life scales were administered at 3 months and at the end of the maintenance phase. A dose-reduction phase of up to 2 weeks was performed and the subjects returned for a final visit that included laboratory tests, ECG, physical examination, and parkinsonian evaluations.

During the study concomitant medication for the treatment of PD, such as anticholinergics, amantadine, and deprenyl, remained fixed. The times of levodopa preparations were also fixed, except that the dosage could be reduced to alleviate adverse events with the study medication. This included the development of dyskinesias, hallucinations, or psychiatric side effects. Levodopa dosage could be subsequently increased during the study if previously reduced, but could not exceed the amount at baseline or the subject would be dropped from the study. Domperidone was allowed to alleviate adverse effects of the study medication.

Efficacy end points. The primary end points were the UPDRS part II Activities of Daily Living (ADL) Scale (average of "on" and "off" scores) and the UPDRS part III Motor Examination Scale comparing baseline with 6 months of maintenance medication. The motor examinations were performed during an "on" time with the last dose of levodopa and study medication given within the specified time window stated.

Secondary end points included the UPDRS I, UPDRS IV, modified Hoehn and Yahr staging both "on" and "off," modified Schwab and England Disability Scale, Parkinson Dyskinesia Scale, timed walking test, Global Clinical Assessment of Efficacy, and patient diary records to evaluate the frequency of "on" and "off" times.

Quality-of-life measurements were also assessed with FSQ with several additional questions regarding occupation, normal work hours, lost work time, or employment changes due to disease. The EuroQoL, which is an instrument for evaluating subject health state utility, was also employed.

Safety. Safety data were summarized by the number of subjects who experienced adverse events and laboratory or physical examination abnormality. Orthostatic hypotension was recorded as either asymptomatic or symptomatic.

Statistical methods. The aim of the study was to determine if the treatment effect of the pramipexole group was greater than the placebo group. Data were analyzed by two different methods. A primary intent-to-treat(ITT) analysis using the last observation carried forward (LOCF) method and a secondary method was used concentrating on the observed cases, which comprised all patients with complete data for analysis. All patients with at least one dose of study medication and completion of at least one postbase-line assessment were suitable for ITT analysis. A nonparametric approach with the Wilcoxon-Mann-Whitney test was used.

As a secondary hypothesis the trial was designed to show that pramipexole and bromocriptine are "equally effective" under the condition that either drug is superior to placebo. For this analysis, a two-step procedure was performed. A global, two-sided Kruskal-Wallis Test on alpha 0.05 was performed. The mean treatment differences for the groups were then tested pairwise using the Wilcoxon-Mann-Whitney test each on the level alpha = 0.05(nested test). The study was not planned to show statistical differences between active treatment groups.

Results. A total of 247 subjects were randomized into the protocol and 246 were included in the ITT analysis. A summary of the baseline demographic information is listed in table 2. There were no statistical differences between the three groups based on sex, age, and duration of PD. The average dose of pramipexole used in the maintenance phase of the study was 3.36 mg per day, and the average dose of bromocriptine was 22.64 mg per day. The primary hypothesis primary outcome measures are shown in table 3 and figure 1. Pramipexole treatment was significantly better than placebo in alleviating PD symptoms as measured by the UPDRS II and III subscales. Bromocriptine treatment was also significantly better than placebo, but the magnitude of the response was less than that observed with pramipexole. Analysis with the observed patient method was not different compared with the ITT.

For the secondary end points, there were no significant differences between the active treatment groups compared with placebo treatment with analysis of the UPDRS I, UPDRS IV, modified Hoehn and Yahr staging both "on" and "off," modified Schwab and England Disability Scale, Parkinson Dyskinesia Scale, and the timed walking test. There was a significant difference in the average percentage of "off" time in the pramipexole treatment group compared with controls. The pramipexole treatment group had a 15% reduction of the amount of awake hours "off" time (p = 0.007), which resulted in approximately 2.5 more hours of "on" time each day. The bromocriptine group, however, did not experience a significant change in its average percentage of"off" time (p = 0.2). Improvements as expressed by relative percentages were equivalent to 45.6% for pramipexole, 29.7% for bromocriptine, and 5.8% for placebo. Figure 2 shows the treatment benefit in the pramipexole group by week 4 with persisting benefit throughout the maintenance period, whereas the bromocriptine treatment group had a slower onset of benefit that was not seen until week 8 and was not maintained. The calculation of p values (post hoc) for each visit with Wilcoxon's rank sum test for pramipexole compared to bromocriptine revealed that there were significant differences with respect to the early aspect of the ascending-dose phase. Pramipexole showed a larger improvement compared with bromocriptine at week 4 (p = 0.0007) and week 6 (p = 0.027). There were no differences in the use of deprenyl in the three treatment groups.

Results of the secondary hypothesis testing did not reveal any differences between the pramipexole and bromocriptine treatment groups for the primary end points. The study was not powered to show a difference between the two active treatment groups. There was a trend to significance with the Global Clinical Assessment of Efficacy showing more improvement with pramipexole compared with bromocriptine (p = 0.056).

Quality-of-life assessments showed significant differences in both of the active treatment groups compared with controls with respect to the FSQ Basic Activities of Daily Living, Intermediate Activities of Daily Living, and Mental Health Scales. These scales revealed comparable improvements of pramipexole and bromocriptine treatment compared with placebo. Other measurements that were part of the FSQ, such as days in bed due to illness, approached statistical significance (p = 0.054) but did not show any differences between bromocriptine and pramipexole treatment. On the other hand, EuroQoL testing approached statistical significance with ANOVA(p = 0.065), with subsequent analysis showing that pramipexole was different than controls (p = 0.02) but bromocriptine was not(p = 0.26). This trend was the only quality-of-life measurement that was different between the two active treatments.

In the pairwise comparisons of secondary end points, bromocriptine did not show a trend of superiority over pramipexole in any of the tests, while pramipexole showed a trend of superiority compared with bromocriptine as stated earlier in the average percentage of "off" time (p = 0.1) and the Global Clinical Assessment of Efficacy (p = 0.056). Also, statistically significant superiority over placebo was found in the Global Clinical Assessment of Efficacy (p < 0.001) and average percentage "off" time (p = 0.0068) for pramipexole, while this applied only to the Global Clinical Assessment of Efficacy (p = 0.002) for bromocriptine.

Safety data are presented in table 3. There were no major differences in the active treatment groups. There were a number of dropouts in each treatment group due to adverse events. There were 16 (20%) in the pramipexole group, 17 (20%) in the bromocriptine group, and 33 (40%) in the placebo group. The majority of the placebo-treated group dropouts were due to worsening of PD during the study. Table 4

Discussion. The results of this multicenter, controlled, double-blind trial show that pramipexole is significantly better than placebo in the treatment of advanced PD patients experiencing end-of-dose deterioration with levodopa. The present study shows robust results with significant differences in both primary end points of the UPDRS ADL and motor scores. The treatment benefit was prolonged, lasting the duration of the 6-month maintenance phase. These results are similar to those reported by Lieberman and the Pramipexole Study Group¹⁴ in another large, double-blind, placebo-controlled trial of pramipexole in advanced PD subjects. Both studies showed a significant change compared with placebo, whereas the present study reveals a larger pramipexole-related improvement in the UPDRS ADL (27% versus 22%) and UPDRS motor scores (35% versus 25%) over a similar time period.

Treatment with bromocriptine also showed significant differences compared with placebo in both primary end points measured. The magnitude of the observed changes in the UPDRS ADL and motor scores were less in the bromocriptine treatment group than the pramipexole group. These results validate the study design, which included an active treatment group to confirm that dopamine agonist treatment was better than placebo treatment in this controlled study. The secondary hypothesis testing did not show a difference between pramipexole and bromocriptine treatments. There was a trend toward statistical significance in the Global Clinical Assessment of Efficacy, which favored pramipexole over bromocriptine. Further studies designed to test this hypothesis will be necessary to show if pramipexole is different than bromocriptine in the treatment of advanced PD.

Patients treated with pramipexole experienced a significant reduction in the number of "off" hours during the waking day compared with placebo. This was assessed by patient diary forms when baseline measurements were compared with the end of the maintenance period. This resulted in approximately 2.5 hours more "on" time each day. This reduction was achieved rapidly with maximal effects within the first 4 weeks. This improvement was sustained throughout the 6-month maintenance period. Bromocriptine therapy, however, did not significantly reduce "off" time compared with placebo. The changes in"off" time were delayed compared with pramipexole and occurred after approximately 8 weeks. This benefit with bromocriptine therapy was not sustained throughout the study. The faster onset of benefit and its sustained effect in reduction of "off" time suggest that pramipexole may provide advantages over bromocriptine in improving the functional disability in advanced PD. This earlier result with pramipexole may be due to the ascending dose schedule that was employed in our study. The bromocriptine titration that was used is the recommended clinical schedule, which allows for a slow dose escalation to avoid adverse events. Previous experiences with pramipexole have shown that therapeutic benefit is achieved rapidly within 3 weeks without significant adverse events,¹⁴ and this study has confirmed this.

Quality-of-life measurements were included in this study. Both active treatment groups showed statistically significant improvements in the FSQ-derived Basic Activities of Daily Living, Intermediate Activities of Daily Living, and Mental Health Scales. The EuroQoL testing approached significance for pramipexole but not for bromocriptine.

There were no unexpected adverse events observed in this study. Both pramipexole and bromocriptine were well tolerated. Nausea and dyskinesias were more common in both treatment groups than in the placebo group. These adverse events are common with other dopamine agonist therapies.⁶ These effects may have been more prominent and differences between study drugs could not be elucidated since the protocol design involved titrating the dose of the agonists to the maximally tolerated dose.

Pramipexole is an effective treatment for advanced PD patients with end-of-dose deterioration with levodopa therapy. Pramipexole improved motor function and ADLs, reduced the amount of "off" time, and was not associated with any unexpected, serious adverse events. Pramipexole therapy is a promising addition to the treatment of patients with advanced PD. Further studies are needed to compare this new dopamine agonist directly with existing therapies to elucidate further its role in the PD treatment paradigm.

Appendix.