METHODS

Participants

Men and women aged 18 to 79 years and diagnosed as having restless legs syndrome (using the International Restless Legs Syndrome Study Group (IRLSSG) diagnostic criteria¹⁹) were included in the study. Patients had a score of at least 15 on the international restless legs scale (IRLS)²⁰ and had either experienced at least 15 nights with symptoms of restless legs syndrome in the previous month or, if receiving treatment, reported they had had symptoms of this frequency before treatment.

Patients were excluded if they were suffering from other movement or primary sleep disorders, if they required treatment for restless legs syndrome during the daytime (defined as 10.00 to 18.00 hours), if they were experiencing augmentation or end of dose rebound, or if they had restless legs syndrome associated with end stage renal disease, iron deficiency anaemia, or pregnancy. Patients were also excluded if they had a history of alcohol or drug abuse, had previous intolerance to dopamine agonists, or were suffering from other clinically relevant conditions affecting assessments.

Patients gave written, informed consent before entering the study, which was done according to the principles of the 1996 amendment of the Declaration of Helsinki and approved by local ethics committees.

Study design

This was a double blind, randomised, placebo controlled trial of 12 weeks’ duration. It was conducted in 43 hospitals, sleep centres, and neurology clinics in 10 European countries (Austria, Belgium, France, Germany, Italy, Netherlands, Norway, Spain, Sweden, and the United Kingdom).

Patients receiving treatment known to affect restless legs syndrome or sleep, or to cause drowsiness, entered a washout phase equal to either five half lives of the drug or seven consecutive nights, whichever was the greater. A small proportion of patients may have had a washout period shorter than seven days before the implementation of the protocol amendment.

Patients were randomly assigned in a 1:1 ratio to receive ropinirole or placebo for 12 weeks. To ensure balance in the number of patients assigned to each treatment group, the allocation schedule was generated in blocks using the sponsor’s coding memo system. Investigators phoned into the randomisation and medication ordering system (RAMOS) to register and randomise each patient. Investigators, patients, and study monitors were blinded to the treatment status of the patients at all times. Investigators and the study sponsor held sealed envelopes containing patient code breaks (comprising the container number and the treatment allocation). In an emergency, an investigator could have scratched the disclosure panel with a coin to reveal details of a patient’s treatment. All envelopes were returned unopened to the sponsor at the end of the study.

Interventions

Ropinirole and matched placebo tablets were used to maintain study blinding. Patients received treatment once daily between one and three hours before bedtime and started ropinirole treatment at 0.25 mg/day. The dose was then titrated upwards during weeks 1 to 7, through seven predetermined dose levels, until patients were receiving the maximum dose (4.0 mg/day) or they were judged to have reached their optimal dose. A maximum of two dose reductions because of adverse events (by one dose level in each case) was permitted during the titration period. The dose could be increased again if adverse events ameliorated. Dose changes were not permitted after week 7.

Assessments and outcome measures

Patients visited the clinic at baseline, on day 2, weekly for the first two months, and then at week 12.

The primary end point was the mean change from baseline to week 12 in the IRLS total score. This scale was developed and validated by the IRLSSG²⁰ and comprises 10 questions about restless legs syndrome symptoms and their impact on daily activities and mood. All the responses are graded in the range 0 to 4 (0 = absence of a problem, 4 = very severe problem), giving a maximum score of 40 (table 1).

The clinical global impression–global improvement (CGI-I) scale was used to assess general improvements. Changes in the proportions of patients with scores of “much improved” or “very much improved” on this scale at weeks 1 and 12, and the mean change in the IRLS total score between baseline and week 1, were identified as the three key secondary end points.

Other secondary end points, concerning the impact of treatment on sleep (sleep adequacy, quantity, disturbance, and daytime somnolence), health related quality of life, work, and other activities, were measured using four patient reported questionnaires: the medical outcomes study (MOS) sleep scale, the MOS 36 item short form health survey (SF-36), the RLS QoL questionnaire, and the work productivity and activity impairment (WPAI) questionnaire.

Safety was assessed by collecting information on adverse events, vital signs, and laboratory parameters.

Statistical analyses

The intention to treat (ITT) dataset was used for all efficacy analyses; this comprised all patients who received at least one dose of double blind drug treatment and had at least one valid post-baseline efficacy assessment. Safety and tolerability data were reported for all patients who received at least one dose of the study drug. For this study, the ITT and safety populations were identical. For all end points, when data were missing, the value from the previous visit was carried forwards (except for the IRLS total score at day 2, which represented an inadequate recall period from baseline and was not carried forward to estimate week 1 or week 12 results).

Continuous efficacy variables were assessed using analyses of covariance, dichotomous variables using logistic regression, and “time to event” data with Cox’s regression model. Baseline score (with the exception of CGI-I), country group, and treatment were fitted as terms in the model. Statistical modelling assumptions were assessed and not found to be violated.

For the primary end point, a sample size of 116 patients per group was necessary to ensure 90% power to detect a difference of six points between groups with an SD of 14 (α = 0.05, two tailed test).

RESULTS

Patient characteristics and flow through the study

The study was carried out between 8 November 2001 and 19 August 2002. In all, 286 patients were randomised and 284 received treatment (146 ropinirole, 138 placebo; the ITT population). Similar numbers of patients completed the study in the two groups (fig 1). Patient characteristics at entry were also similar in the two groups (table 2).

Treatment

The mean (SD) daily dose of ropinirole at 12 weeks was 1.90 (1.13) mg. The corresponding placebo “dose” was 2.80 (1.25) mg.

Efficacy results

Primary end point

The mean (SD) IRLS total score at week 12 was lower in the ropinirole group (13.5 (9.3) points) than with the placebo group (17.1 (9.4) points). The adjusted mean (SE) improvement in the IRLS total score at week 12 was significantly greater for ropinirole (−11.04 (0.719) points) than for placebo (−8.03 (0.738) points; adjusted treatment difference = −3.01 (95% confidence interval (CI), −5.03 to −0.99)) (p = 0.0036).

Key secondary endpoints

Significantly more patients in the ropinirole group (53.4%, 78/146) showed a “much improved” or “very much improved” score on the CGI-I scale at week 12 compared with the placebo group (40.9%, 56/137; adjusted odds ratio = 1.7 (1.02 to 2.69)) (p = 0.0416).

Treatment differences in favour of ropinirole were apparent by the first assessment—that is, at week 1. The mean (SE) reduction in the IRLS total score from baseline was significantly greater for ropinirole (−8.19 (0.59) points) than for placebo at week 1 (−5.14 (0.62) points; adjusted treatment difference = −3.05 (−4.72 to −1.38)) (p = 0.0004). Similarly, significantly more patients in the ropinirole group (34.2%; 50/146) had a score of “much improved” or “very much improved” on the CGI-I scale compared with the placebo group at week 1 (13.1%; 18/137; adjusted odds ratio = 3.7 (2.02 to 6.94)) (p<0.0001).

Other secondary end points

There were significant differences in favour of ropinirole for the median time taken to show a decrease of at least six points on the IRLS total score and the median time taken to report a score of “much improved” or “very much improved” on the CGI-I scale (figs 2 and 3).

The ropinirole group experienced significantly greater mean improvements in sleep adequacy and quantity (p = 0.0015 and p = 0.0331, respectively), and greater mean reductions in daytime somnolence and sleep disturbance (p = 0.0064 and p = 0.0245) compared with the placebo group (fig 4). Furthermore, ropinirole was also associated with a greater mean improvement in quality of life as assessed with the disease specific measure, the RLS QoL questionnaire (17.1 v 12.6, respectively) (p = 0.0314). There were no differences between groups, however, in the changes on the generic measures, the SF-36 and the WPAI questionnaire.

Safety results

The most commonly reported adverse events were nausea and headache; nausea was more common with ropinirole than with placebo (table 3). Most events were mild to moderate in intensity (table 3). The frequency of adverse events declined over time in both groups: after day 70, only 9.6% (14 patients) in the ropinirole group and 5.8% (eight patients) in the placebo group reported new adverse events. There were no reports of augmentation.

Few patients withdrew because of adverse events (16 with ropinirole v 6 with placebo). Nausea led to most withdrawals in the ropinirole group (6, v 0 with placebo), and restless legs symptoms in the placebo group (3, v 0 with ropinirole). Serious adverse events were also unusual (3 with ropinirole v 4 with placebo) and none led to withdrawal or was judged to be related to the study drug (menstrual disorder; injury; gastrointestinal disorder; urinary tract infection; fever, syncope and pharyngitis; fever; abdominal pain).

DISCUSSION

Ropinirole, at single doses of up to 4.0 mg/day, was significantly more effective than placebo in alleviating symptoms of restless legs syndrome, improving sleep quantity and adequacy, reducing sleep disturbance and daytime somnolence, and improving health related quality of life. Importantly, the superiority of ropinirole was apparent as early as week 1. Furthermore, ropinirole was generally well tolerated and there were no reports of augmentation or sudden onset of sleep.

This is the largest controlled trial in restless legs syndrome conducted to date and it employs a validated scale,²⁰ the IRLS, to determine the impact of treatment on the severity of the condition. As there has been limited experience with the IRLS as a tool for monitoring treatment response, the CGI-I was also employed, and these data confirm the clinical relevance of the IRLS results. Furthermore, in the key area of sleep disruption, which is troublesome to a very large proportion of patients with restless legs syndrome, ropinirole was significantly superior to placebo on all of the variables tested. The reduction in daytime somnolence has particular importance as it indicates that ropinirole is not acting as a sedative. This is further supported by the fact that similar proportions of patients in each group reported somnolence (not necessarily daytime somnolence) as an adverse event (12% of patients in the ropinirole arm and 7% in the placebo arm).

The positive results in this trial are consistent with the efficacy findings from small short term studies in which ropinirole doses of up to 6.0 mg/day were associated with significant improvements in symptoms¹⁵¹⁷ and benefits were apparent within one day.¹⁸²¹ An early onset of action is clearly an important factor influencing the choice of treatment.

The adverse event profile of ropinirole in the present study is typical of those associated with dopaminergic treatment generally. The rates of withdrawal because of nausea and vomiting were, however, rather low (5%).

Ropinirole was effective despite the presence of a placebo effect that was larger than expected. Such effects are common in trials in other disease areas in which subjective tools are used to assess efficacy (for example, migraine and depression). Furthermore, restless legs syndrome is generally poorly recognised,²² and the receipt of a long awaited diagnosis, coupled with individual attention from physicians, may have enhanced the patients’ expectations of treatment. In support of this, patients receiving placebo experienced a greater number of dose increments, with 45% reaching the maximum “dose” compared with 16% in the ropinirole group.

Ropinirole is an effective and well tolerated treatment for restless legs syndrome. Follow up studies are required, however, to confirm that efficacy and the lack of augmentation are maintained in the longer term and to explore whether improved sleep among patients treated with ropinirole will be associated with improved cognitive functioning. Data from additional work examining the efficacy of ropinirole on periodic leg movements and further sleep measures are expected soon.

MEMBERS OF THE TREAT RLS 1 STUDY GROUP

Austria: W O Poewe, G Saletu-Zyhlarz; Belgium: I De Volder, R Dom, M Gonce, A Lachman, O C A G Lebon; France: I Arnulf, J-D Guieu, J Krieger, E Lainey, J-P Macher, M Tiberge, J Touchon; Germany: H Benes, M Fickinger, J Glass, G Hajak, M Haug, B Jeschke, T Krohn, G Mayer, R Neuhaus, W Oertel, M Pellkofer, G Pfeffer, F Roeth, U Schreiber, R Steinberg, C Trenkwalder, E Wattig; Italy: G Gigli, P Montagna, L Murri, L Ferini-Strambi; Netherlands: J H M De Groen, A W de Weerd, M G Smits; Norway: M Lossius, W Telstad; Spain: A Beneto-Pascual, A Caminero-Rodriguez, F Cañellas-Dols, E Estivil-Sancho, D Garcia-Borreguero, J J Poza-Aldea, E Rodriguez-Saez, T Sagales-Sala, J Santamaría-Cano, J Vaamonde; Sweden: C-P Anderberg, J Hetta; United Kingdom: N Hyman, J Shneerson, P Tidswell, A Young