This review concluded that the adjunctive use of dopamine agonists with levodopa, particularly non-ergots, could provide significant improvements in patient functioning and motor symptoms in patients with Parkinson's disease not controlled by levodopa therapy alone, but there was an increased risk of some adverse events. These conclusions appear to be reliable.

To evaluate the addition of a dopamine agonist to levodopa therapy for the treatment of advanced Parkinson's disease.

MEDLINE, EMBASE, CINAHL, Web of Science and the Cochrane Library (2007, Issue 2) were searched between 1990 and July 2007. Search terms were reported. Reference lists of reviews and primary studies were also screened.

Randomised controlled trials (RCTs) comparing a dopamine agonist and levodopa with a placebo and levodopa in patients with advanced Parkinson's disease (Hoehn and Yahr stage II to IV) already being treated with levodopa were eligible for inclusion. Eligible trials had to measure efficacy using the Unified Parkinson's Disease Rating Scale (UPDRS): activities of daily living (ADL) and motor scores; they also had to report change in 'off' time, withdrawals, and rates of dyskinesia, hallucinations and mortality. Trials of multiple drug doses not reporting cumulative results were excluded.

Included trials assessed ergot dopamine agonists (bromocriptine, cabergoline and pergolide) and non-ergot dopamine agonists (piribedil, pramipexole, ropinirole, rotigotine and sumanirole). Treatment duration ranged from 12 to 40 weeks. In included trials, mean patient age ranged from 59.7 to 66.2 years, duration of symptoms ranged from 4.3 to 12.6 years, and between 44 and 72% of participants were men.

The authors did not state how many reviewers performed the study selection.

The Jadad scale was used to assess trial validity, covering randomisation method, double-blinding and description of withdrawals/drop-outs. Scores range from zero to 5 points; trials scoring less than 3 points were considered low quality.

Validity was assessed by two reviewers independently, with disagreement resolved by consensus or discussion with a third reviewer.

Data were extracted by two reviewers independently, with disagreements resolved by consensus; a third reviewer verified all data before any analyses were conducted. Mean changes from baseline for continuous outcomes were extracted and used to calculate mean differences with 95% confidence intervals (CI). If a study did not report the variance of the change from baseline, it was estimated assuming a correlation coefficient of 0.5 between baseline and follow-up measurements. Dichotomous outcomes were used to calculate odds ratios (OR) with 95% confidence intervals.

Trials were pooled in meta-analyses using the DerSimonian-Laird random-effects model. Analyses were grouped by treatment comparisons: all dopamine agonists, ergot dopamine agonists and non-ergot dopamine agonists. Statistical heterogeneity was assessed using the Q statistic (p<0.10). Publication bias was assessed using funnel plots and Egger's regression method (p<0.10). Sensitivity analyses repeated the meta-analyses using a fixed-effect model and also by excluding trials considered to be poor quality.

Fifteen RCTs were included in the review (n=4,391). All trials scored 3 or more points on the Jadad scale.

All dopamine agonists: Dopamine agonists with levodopa resulted in a greater reduction in activities of daily living (ADL) scores compared with levodopa alone (placebo), with a weighted mean difference (WMD) of -2.20 (95% CI -2.64 to -1.76; 10 RCTs) and also a greater reduction in UPDRS motor scores (WMD -5.56, 95% CI -6.82 to -4.31; nine RCTs). There was no evidence of publication bias or heterogeneity for ADL scores, but there was evidence of both for motor scores (heterogeneity p=0.02; publication bias p=0.055). A greater reduction in 'off' time was seen for dopamine agonist treatment (WMD -1.20, 95% CI -1.78 to -0.62, four RCTs) with significant heterogeneity (p<0.001), although all trials showed a positive treatment effect. Dopamine agonist treatment resulted in significant reductions in levodopa dose compared with placebo (WMD -128.5, 95% CI -175.0 to -82.1, five RCTs), greater incidence of dyskinesia than placebo (OR 3.27, 95% CI 2.65 to 4.03, 12 RCTs) and a higher rate of hallucinations than placebo (OR 3.34, 95% CI 2.44 to 4.58, 13 RCTs), but no evidence of a difference for patient withdrawal or mortality.

Non-ergot dopamine agonists: Non-ergot dopamine agonists with levodopa resulted in a greater reduction in ADL scores compared with levodopa alone (placebo) (WMD -2.29, 95% CI -2.76 to -1.82; nine RCTs) and also a greater reduction in UPDRS motor scores (WMD -6.06, 95% CI -7.20 to -4.92; eight RCTs). Non-ergot dopamine agonists also resulted in a greater reduction in 'off' time (WMD -1.19, 95% CI -1.82 to -0.57, three RCTs), a greater than 120mg reduction in daily levodopa dose (WMD -121.8, 95% CI -177.2 to -66.4; three RCTs). Non-ergot dopamine agonist treatment resulted in greater incidence of dyskinesia compared with placebo (OR 2.94, 95% CI 2.31 to 3.75; 11 RCTs) and a significantly higher rate of hallucinations than placebo (OR 3.40, 95% CI 2.41 to 4.82; 11 RCTs), but no evidence of a difference for patient withdrawal or mortality. Significant heterogeneity was seen for 'off' time and change in dose analyses.

Ergot dopamine agonists: Ergot dopamine agonists with levodopa resulted in a greater reduction in ADL scores compared with levodopa alone (placebo) (WMD -1.45, 95% CI -2.35 to -0.56, two RCTs); UPDRS motor scores (WMD -3.06, 95% CI -5.80 to -0.32; two RCTs); and a greater than 150mg reduction in daily levodopa dose (WMD -151.1, 95% CI -284.5 to -17.6; two RCTs); but an increase in dyskinesia compared with placebo (OR 2.94, 95% CI 1.45 to 5.95, four RCTs). There was no evidence of a difference for withdrawals or rate of hallucinations.

Sensitivity analyses did not alter any conclusions.

The adjunctive use of dopamine agonists with levodopa, particularly non-ergots, provided significant improvements in patient functioning and motor symptoms in patients with Parkinson's disease who were not controlled by levodopa therapy alone. Dopamine agonists seemed useful in patients with weaning-off phenomenon from levodopa therapy, but there was an increased risk of some adverse events.

This review had a clearly stated question and inclusion criteria. The search covered relevant databases, but it was not clear if there were any language restrictions (which may have excluded some studies). However, there was a thorough assessment of publication bias, which suggested this was not a problem for all except one outcome. Data extraction and validity assessment were performed in duplicate, but it was not reported if trials were selected the same way.

The methods of meta-analysis appeared appropriate, but not all results were presented in plots or tables, so it was not possible to judge the reliability of the pooled results. Some statistical heterogeneity was observed, but this was mostly where there were few trials, and the test used was not very reliable in this situation.

Overall, the authors' conclusions appear to be reliable.

One author disclosed financial links with Boehringer Ingelheim Pharmaceutical, Inc (manufacturers of dopamine agonists).

The authors did not state any recommendations for practice or research.

Boehringer Ingelheim Pharmaceuticals, Inc.

This is a critical abstract of a systematic review that meets the criteria for inclusion on DARE. Each critical abstract contains a brief summary of the review methods, results and conclusions followed by a detailed critical assessment on the reliability of the review and the conclusions drawn.