Sleep Medicine 12 (2011) 351–360

Contents lists available at ScienceDirect

Sleep Medicine
journal homepage: www.elsevier.com/locate/sleep

Original Article

Efficacy and augmentation during 6 months of double-blind pramipexole for

restless legs syndrome
Birgit Högl a,⇑, Diego Garcia-Borreguero b,1, Claudia Trenkwalder c,2, Luigi Ferini-Strambi d,3,
Wayne Hening e, , Werner Poewe a,4, Stefanie S. Brenner f,5, Mandy Fraessdorf g,6, Michael Busse h,7,
Stefan Albrecht i,8, Richard P. Allen j,9
a
Department of Neurology, Innsbruck Medical University, Anichstrasse 35, A-6020 Innsbruck, Austria
b
Sleep Research Institute, Alberto Alcocer 19, 28036 Madrid, Spain
c
Center of Parkinsonism and Movement Disorders, Kassel, Dept. Clinical Neurophysiology, University of Goettingen, Klinikstrasse 16, 34128 Kassel, Germany
d
Sleep Disorders Center, Università Vita-Salute San Raffaele, Via Stamira d’Ancona 20, 20127 Milan, Italy
e
UMDNJ-RW Johnson Medical Center, New Brunswick, NJ, USA
f
Department of Clinical Research, Boehringer Ingelheim Pharma GmbH & Co., KG, Birkendorferstrasse 65, 88397 Biberach an der Riss, Germany
g
Medical Division, Boehringer Ingelheim Pharma GmbH & Co., KG, Binger Strasse 173, D-55216 Ingelheim am Rhein, Germany
h
CDept Medical Affairs, Boehringer Ingelheim GmbH, Binger Strasse 173, D-55216 Ingelheim am Rhein, Germany
i
Merz Pharmaceuticals GmbH, Eckenheimer Landstraße 100, 60318 Frankfurt am Main, Germany
j
Neurology and Sleep Medicine, Johns Hopkins University, Asthma and Allergy Bldg 1B76b, 5501 Hopkins Bayview Circle, Baltimore, MD 21224, USA

a r t i c l e i n f o a b s t r a c t

Article history: Background: Pramipexole is an effective treatment for restless legs syndrome (RLS), but no controlled
Received 11 June 2010 studies have lasted >12 weeks.
Received in revised form 7 December 2010 Methods: RLS patients (N = 331) with pretreatment serum ferritin >30 ng/mL were randomly assigned to take
Accepted 10 December 2010
double-blind optimized pramipexole (0.125–0.75 mg/d) or placebo for 26 weeks. The primary efficacy
Available online 26 February 2011
endpoint was change in International RLS Study Group Rating Scale (IRLS) score. Other endpoints assessed
global change, symptoms, and QoL. Patients maintained symptom diaries. Cases meeting predefined criteria
Keywords:
for suspected augmentation were reviewed by a blinded expert panel, which used a predefined algorithm.
Restless legs syndrome
Pramipexole
Results: Among 321 patients providing post-baseline data, of whom 234 completed 26 weeks, pramipexole
Dopamine agonists was more effective than placebo by multiple endpoints, including an adjusted mean IRLS score change of
Augmentation 13.7 vs. 11.1 (p = 0.0077) and an IRLS responder rate (P50% score reduction) of 58.6% vs. 42.8%
Therapy (p = 0.0044). Efficacy showed considerable country-to-country variability. Six-Month incidence of confirmed
Long-term treatment augmentation was 9.2% for pramipexole and 6.0% for placebo. The rate increased with treatment duration for
Withdrawal pramipexole but not placebo. Treatment-related adverse events (AEs) were more likely for pramipexole than
for placebo, but discontinuation due to AEs was less likely.
Conclusions: During a 6-month period, pramipexole was effective, safe, and generally well tolerated. Because
risk of augmentation may have increased over 6 months, it should be studied in longer trials. Beginning or mild
augmentation is difficult to distinguish from natural RLS fluctuation, at least in a non-iron-deficient population.
Ó 2011 Elsevier B.V. All rights reserved.

⇑ Corresponding author. Tel.: +43 512 504 23811; fax: +43 512 504 23842. 1. Introduction
E-mail addresses: birgit.ho@i-med.ac.at (B. Högl), dgb@iis.es (D. Garcia-
Borreguero), trenkwalder@pk-mx.de (C. Trenkwalder), ferinistrambi.luigi@hsr.it
Dopaminergic treatment is currently the first-line pharmaco-
(L. Ferini-Strambi), werner.poewe@i-med.ac.at (W. Poewe), stefanie.brenner@
boehringer-ingelheim.com (S.S. Brenner), mandy.fraessdorf@boehringeringelheim. logic intervention in restless legs syndrome (RLS) [1,2]. For moder-
com (M. Fraessdorf), michael.busse@boehringer-ingelheim.com (M. Busse), Stefan. ate to severe cases, the approved agents include pramipexole, a
Albrecht@merz.de (S. Albrecht), richardjhu@mac.com (R.P. Allen). nonergotamine dopamine agonist with high affinity for the D2-like
1
Tel.: +34 91345429; fax: +34 913509593. subfamily of dopamine receptors. In 4 randomized, double-blind
2
Tel.: +49 (561) 6009 200; fax: +49 (561) 6009 126.
3 phase 3 trials [3–6], a total of approximately 1000 patients re-
Tel.: +39 (02) 26433358; fax: +39 (02) 26433394.
4
Tel.: +43 512 504 3850; fax: +43 512 504 3852.
ceived the active drug or placebo for up to 12 weeks. Three of
5
Tel.: +49 7351 54 97894; fax: +49 7351 54 5726. the trials [3–5] had run-in or extension phases lasting up to
6
Tel.: +49 (6132) 77 92914; fax: +49 (6132) 72 92914. 46 weeks, all of which corroborated the benefits documented dur-
7
Tel.: +49 (6132) 77 90693; fax: +49 (6132) 72 90693. ing the double-blind phases. The added phases, however, were
8
Tel.: +49 (0) 69 15 03 1364; fax: +49 (0) 69 15 03 1445. partly or wholly open label, and hence could not definitively assess
9
Tel.: +1 (410) 550 1044; fax: +1 (410) 550 3364.
  the treatment’s long-term effects. The trials also were not designed
Deceased.

1389-9457/$ - see front matter Ó 2011 Elsevier B.V. All rights reserved.
doi:10.1016/j.sleep.2010.12.007
352 B. Högl et al. / Sleep Medicine 12 (2011) 351–360
to study augmentation (RLS exacerbation attributed to an ongoing,
initially beneficial RLS treatment) or withdrawal (RLS exacerbation
if treatment stops).
Here we report the findings of a phase 4, randomized, double-
blind, placebo-controlled, dose-titration trial with a multiplicity
of long-term (26-week) efficacy endpoints. Safety was judged by
profiling adverse events (AEs) and tolerability by rates of prema-
ture discontinuation. The safety testing also included assessments
for incidence of augmentation, by submission of all suspected cases
to a blinded expert panel, and for incidence of withdrawal, by eval-
uation of RLS severity after each patient’s end of treatment.
2. Methods
At 42 sites in 9 European countries, adults (aged 18–85 years)
with idiopathic RLS were randomly assigned, at a 1:1 ratio, to receive
double-blind pramipexole or placebo. For entry, all patients were re-
quired to meet all diagnostic criteria of the International RLS Study
Group (IRLSSG) [7], to have a baseline total score >15 on the Study
Group’s International RLS Rating Scale (IRLS) [8], and to have expe-
rienced RLS symptoms at least 2–3 days per week throughout the
prior 3 months. Patients were excluded for serum ferritin 630 ng/
mL, known hypersensitivity to pramipexole, augmentation during
previous RLS treatment, unsuccessful previous treatment with non-
ergotamine dopamine agonists (e.g., pramipexole, ropinirole), any
non-RLS sleep disorder, any major Diagnostic and Statistical Manual
of Mental Disorders, Fourth Edition, Text Revision psychiatric disorder
within the prior 2 years, change in any antidepressant regimen
within the prior 4 weeks (or any anticipated change), and any use,
within the prior 2 weeks, of dopamine agonists, levodopa, or any
medication or dietary supplement capable of altering RLS
symptoms. Women with childbearing potential were excluded for
pregnancy, inadequate contraception, or current breastfeeding of a
child. Patients with daytime RLS symptoms were not excluded.
The study was approved by the independent ethics committees of
the participating countries. Informed, written consent was obtained
from all patients prior to randomization.
During the first 4 weeks, treatment was optimized. For pramip-
exole recipients, treatment commenced with a week at 0.125 mg/
d. After weeks 1, 2, 3, and 4, stepwise dosage adjustment was per-
mitted to 0.125, 0.25, 0.50, or 0.75 mg/d (pramipexole dihydro-
chloride monohydrate), based in each patient on clinically
sufficient response (on the Patient Global Impression [PGI] scale
[9]) and tolerability. Treatment then continued at the patient’s
optimized dosage until the end of week 26. All patients were asked
to take their treatment once daily, 2–3 h before expected bedtime.
After weeks 1, 4, 6, 12, 18, and 26, all patients underwent on-
site evaluation of treatment efficacy, treatment compliance, AEs,
and augmentation. At each of these visits, patients assessed them-
selves by IRLS (the trial’s primary outcome measure) and PGI, and
were assessed on the Clinical Global Impressions-Global Improve-
ment (CGI-I) scale [9]. For each IRLS assessment, each patient was
asked to read each of the instruments 10 items and provide the
required self-rating on a patient worksheet while the examiner re-
mained available to clarify any misunderstandings a patient may
have had and to record the patient’s responses on an investigator
worksheet. Self-assessment tools also included the Johns Hopkins
Restless Legs Syndrome Quality of Life Questionnaire (RLS-QoL)
[10] and the RLS-6 set of scales [11]. All investigators received
training on study procedures including all scales and worksheets
and the patient diary (described below).
The trial’s sample size was based on detection of treatment effi-
cacy. For 26-week change in mean IRLS total score, documentation
of a 4.5-point difference between pramipexole and placebo at a 2-
sided 5% significance level with 90% power would require that each
group have 151 patients. Recruitment of 160 per group was
therefore planned. Testing of mean IRLS change was, by analysis of
covariance (ANCOVA), adjusted for baseline, country, and treatment
group. For IRLS, CGI-I, and PGI, responder rates were defined, respec-
tively, as the proportion of patients with at least a 50% reduction
from their baseline total IRLS score, the proportion classified as at
least ‘‘much improved,’’ and the proportion classifying themselves
as at least ‘‘much better.’’ Among these endpoints, IRLS and CGI-I re-
sponder rates were the trial’s key secondary outcomes. All respon-
der-rate findings were subjected to Cochran–Mantel–Haenszel
test, and all other efficacy findings to van Elteren test (stratified by
country). In all analyses, p < 0.05 was considered statistically significant.
Adverse events were identified by asking each patient at each
trial visit, ‘‘How have you felt since your last visit?’’ Mild AEs were
defined as being easily tolerated, moderate AEs as interfering with
usual activities, and severe AEs as preventing usual activities. Seri-
ous AEs were defined as being life-threatening or resulting in death
or significant incapacity, requiring hospitalization, or requiring
medical or surgical intervention to prevent such outcomes. All
AEs occurring within 48 h after a patient’s last intake of trial med-
ication were classified as having happened under treatment. After
weeks 4, 12, and 26, supine and standing pulse rate and diastolic
and systolic blood pressure were measured.
Because augmentation would be identifiable only in reference
to initial response to treatment, all patients were asked to maintain
an RLS symptom diary for the week preceding randomization and
the week preceding each scheduled evaluation. For each hour of
each of those weeks, the diary documented the severity of a pa-
tient’s RLS symptoms (none; mild/nonbothersome; bothersome),
their time of onset, and the patient’s status (sleeping, awake and
active, awake but at rest). At all scheduled evaluations, investiga-
tors administered an augmentation questionnaire enabling them
to screen for suspected augmentation by four standards: IRLSSG
criteria [7], developed at a National Institutes of Health workshop
and published in 2003; current international criteria (Max-
Planck-Institute criteria) [12], developed at a European RLS Study
Group consensus conference and published in 2007; a score P5
on the 3-item Augmentation Severity Rating Scale (ASRS) [13];
and an investigator’s own judgment, including reliance on diary
data. A positive finding on any of these was cause for case referral
to an independent, 4-member expert panel. In accordance with
current criteria [12], only patients treated for P4 weeks could be
considered. To ensure that case referrals were based on thorough,
uniform knowledge of augmentation, all investigators had under-
gone training by the expert panel’s chairperson.
For each referred case, the expert panel, blinded to the subject’s
treatment group, reviewed all available data using a predefined
algorithm (Table 1), so as to achieve a consensus in confirming or
excluding augmentation or in judging the data to be insufficient
for a decision. In the absence of consensus, the panel’s chairperson
made the final decision. The counts referred to below as ‘‘classified
augmentation’’ include not only confirmed augmentation but also
all referred cases judged to have insufficient data. The counts re-
ferred to as ‘‘confirmed augmentation’’ include only cases in which
the data were judged to be sufficient.
Withdrawal was defined as a posttreatment increase (i.e., wors-
ening) of P4 points on a patient’s IRLS score, as measured
7 ± 1 days after cessation of treatment. The comparison was with
the higher of the patient’s scores at screening and randomization.
3. Results
3.1. Study subjects
Of 331 randomized patients, 329 received treatment (starting,
for the first subjects, in May 2007, and ending, for the last subjects,
 B. Högl et al. / Sleep Medicine 12 (2011) 351–360 353

Table 1
The expert panel’s algorithm for confirming or excluding augmentation.

1. Review investigator’s basis for suspicion. Which standard(s) did the patient meet, and was the suspicion plausible?
2. Review initial treatment efficacy. By current augmentation criteria [12], there must initially have been benefit
3. Review relation to dosage. Did worsening emerge during dosage increase? (However, by current augmentation criteria [12],
only worsening occurring in patients treated P4 weeks could be considered.)
4. Review IRLS data. Was there a change consistent with augmentation?
5. Review diary data. Were there changes consistent with augmentation?
6. Review QoL data. Was QoL consistent with augmentation?

IRLS, International Study Group RLS Rating Scale; QoL, quality-of-life.

Table 2
in July 2008). In their baseline characteristics (Table 2, top), the
Patients’ baseline characteristics.
166 pramipexole recipients resembled the 163 placebo recipients,
including a mean IRLS score in the range (21–30) signifying severe Pramipexole Placebo
RLS. Within these groups, 162 and 159 patients contributed post- Number of patients treated 166 163
baseline IRLS/PGI/CGI-I/RLS-6 data, and 157 and 153 contributed Age, y; mean (SD) 57.9 (12.7) 55.8 (11.4)
Females, % 61.4 57.7
postbaseline RLS-QoL data. On all scales, the pramipexole group
RLS duration, y; median (P25%, P75%) 1.45 (0.0, 8.2) 1.20 (0.0, 7.2)
resembled the placebo group at baseline (Table 2, bottom). Prema- RLS duration, y; mean (SD) 6.0 (9.6) 5.4 (8.5)
ture discontinuation was less frequent for pramipexole than for Naïve to RLS treatment, % 80.1 76.1
placebo, at 21.1% (35/166) vs. 36.8% (60/163). Trial completion IRLS total score, mean (SD) 23.9 (5.3) 23.5 (5.4)
rates were therefore 78.9% (131/166) for pramipexole and 63.2% Number of patients providing postbaseline 157–162 153–159
(103/163) for placebo. data
RLS-6
Sleep dissatisfaction,a median (P25%, P75%) 6.0 (4.0, 8.0) 6.0 (4.0, 8.0)
3.2. Study-drug exposure Severity while falling asleep, median (P25%, 5.0 (3.0, 7.0) 5.0 (2.0, 7.0)
P75%)
Nighttime severity, median (P25%, P75%) 5.0 (2.0, 7.0) 5.0 (2.0, 7.0)
Among pramipexole recipients, 27 (16.3%) were maintained on Daytime severity while resting, median 4.0 (2.0, 6.0) 4.0 (2.0, 6.0)
0.125 mg/d, 53 (31.9%) on 0.25 mg/d, 44 (26.5%) on 0.5 mg/d, and (P25%, P75%)
42 (25.3%) on 0.75 mg/d. In the placebo group, maintenance ‘‘dos- Daytime severity while active, median 1.0 (0.0, 2.0) 1.0 (0.0, 2.0)
age’’ tended to be higher, with 17 patients (10.4%) on ‘‘0.125 mg/d,’’ (P25%, P75%)
Daytime tiredness/sleepiness, median (P25%, 4.0 (2.0, 6.0) 4.0 (2.0, 6.0)
29 (17.8%) on ‘‘0.25 mg/d,’’ 57 (35.0%) on ‘‘0.5 mg/d,’’ and 60
P75%)
(36.8%) on ‘‘0.75 mg/d.’’ RLS-QoL, median (P25%, P75%) 72.5 (60.0, 70.0 (57.5,
82.5) 82.5)

3.3. IRLS/CGI-I/PGI IRLS, International Study Group RLS Rating Scale; QoL, quality of life; RLS, restless
legs syndrome; SD, standard deviation.
a
Termed dissatisfaction here, so as to emphasize that on all RLS-6 scales, high
Among pramipexole recipients, IRLS total score (SE) decreased
scores represent severe detriment.
by an adjusted mean 13.7 (0.8), vs. 11.1 (0.8) for placebo
(p = 0.0077). The decrease was significant after week 1 and re-
mained so throughout the trial (Fig. 1), with a maximum difference
from placebo of 3.7 points at week 6. Country by country, however,
the 26-week mean score change showed a treatment-group differ- Weeks After Randomization
ence (Table 3) ranging from 7.4 points favoring pramipexole in 1 4 6 12 18 26
0
Germany to 1.5 points favoring placebo in the Netherlands. In con-
Adjusted Mean Change (SE)

sonance with this disparity, the CGI-I responder rate to placebo, -2

country by country (Table 4), was highest in the Netherlands, at
-4
66.7% of placebo recipients, and lowest in Germany, at 36.6%, while
CGI-I responder rates to pramipexole were almost identical in -6 -4.6
those countries, at 73.9% in the Netherlands and 73.8% in Germany.
Overall, the IRLS responder rate was 58.6% for pramipexole vs. -8
-7.2a
42.8% for placebo (p = 0.0044), the CGI-I responder rate was
68.5% vs. 50.3% (p = 0.0010), and the PGI responder rate was -10 -8.8
62.3% vs. 44.0% (p = 0.0011). -9.9
-12 -10.3 -10.3
-11.1
a
-12.0
-14
3.4. RLS-6 -13.2b -13.2b
-13.6a -13.7b
-16
On the scales for nighttime problems, baseline medians of 5.0–
6.0 improved by a median 2.5 vs. 2.0 (p = 0.0489) for sleep sat-
Fig. 1. Adjusted mean change in IRLS total score. IRLS, International RLS Study
isfaction, 3.0 vs. 1.0 (p = 0.0315) for severity while falling
Group Rating Scale; SE, standard error. ap < 0.001. bp < 0.01.
asleep, and 3.0 vs. 2.0 (p = 0.0735) for severity during the night.
For daytime problems, baseline medians of 4.0 for tiredness and for
symptoms while resting were insignificantly improved by a med- 3.5. RLS-QoL
ian 1.0 in both groups (p = 0.8093 and 0.8410) and a baseline
median 1.0 for symptoms, while active showed no median change Baseline medians of 72.5 and 70.0 improved by a median +15.0
(p = 0.9241). vs. +12.5 (p = 0.5905).
354 B. Högl et al. / Sleep Medicine 12 (2011) 351–360

Table 3
26-Week IRLS score change, by country.

Pooled country Pramipexole Placebo Treatment-group difference in mean

26-week change [95%-CI]
n Mean at base- Mean 26-week n Mean at base- Mean 26-week
line (SD) change (SD) line (SD) change (SD)
Austria + Slowakia 8 22.8 (5.1) 12.4 (14.4) 7 25.3 (3.4) 9.6 (8.2) 2.8 [ 16.1, +10.5]
Belgium + Spain 10 25.1 (4.9) 16.4 (7.1) 10 23.0 (4.4) 11.3 (9.8) 5.1 [ 13.6, +2.9]
Finland 26 22.9 (4.8) 12.1 (8.8) 26 21.6 (4.0) 10.3 (10.8) 1.8 [ 7.3, +3.7]
Germany 42 25.8 (5.5) 15.2 (9.4) 41 26.4 (4.7) 7.7 (10.4) 7.4 [ 11.7, 3.1]
Ireland 11 25.2 (5.3) 18.2 (11.7) 13 21.9 (6.4) 12.5 (11.3) 5.7 [ 15.5, +4.0]
Netherlands 23 21.4 (4.2) 12.3 (8.3) 21 22.6 (6.6) 13.8 (10.7) +1.5 [ 4.4, +7.3]
United Kingdom 42 23.5 (5.5) 10.2 (6.3) 41 22.7 (5.5) 10.2 (8.8) 0.0 [ 3.3, +3.3]

CI, confidence interval; IRLS, International RLS Rating Scale; SD, standard deviation.

Table 4
recipients to experience AEs and were also more likely to experi-
CGI-I responder ratea at 26 weeks, by country.
ence treatment-related AEs. Rates of premature discontinuation
Pooled country Pramipexole Placebo due to AEs, however, were lower for pramipexole than for placebo.
n Responder rate (%) n Responder rate (%) AE frequencies are summarized by medical type in Table 6. Among
Austria + Slowakia 8 75.0 7 57.1 the three types with a frequency >10% in either treatment group,
Belgium + Spain 10 80.0 10 60.0 nausea was almost four times more common and fatigue was
Finland 26 69.2 26 50.0 slightly more common for pramipexole than for placebo, while
Germany 42 73.8 41 36.6
headache was less common for pramipexole than for placebo.
Ireland 11 63.6 13 46.2
Netherlands 23 73.9 21 66.7 The study’s 11 serious AEs were hospitalizations (of 4.8% of pram-
United Kingdom 42 57.1 41 53.4 ipexole and 1.8% of placebo recipients), of which only 1 (orthostatic
hypotension and vomiting in a placebo recipient) was deemed to
CGI-I, Clinical Global Impressions-Global Improvement scale.
a
Proportion of group classified as at least ‘‘much improved.’’
be treatment-related. Overall, vital signs (including supine and
standing systolic and diastolic blood pressure) showed no notable
changes from baseline in either treatment group.
Table 5
Frequencies of adverse events: summary.
3.7. Augmentation
Category, n (% of group) Pramipexole Placebo
group group
Among 152 pramipexole and 149 placebo recipients treated for
Patients treated 166 (100.0) 163 (100.0)
P4 weeks, and therefore eligible for augmentation diagnosis, 47
Proportion with any AE 120 (72.3) 106 (65.0)
Proportion with treatment-related AEs 64 (38.6) 50 (30.7) (30.9%) and 40 (26.8%) were referred for expert-panel assessment
Proportion with severea AEs 17 (10.2) 15 (9.2) of suspected augmentation. Of these referrals, 18 (11.8% of eligible
Proportion with seriousb AEs 8 (4.8) 3 (1.8) pramipexole recipients) and 14 (9.4% of eligible placebo recipients)
Proportion with AEs leading to 19 (11.4) 23 (14.1)
were classified as augmentation cases, encompassing either con-
withdrawal
firmed augmentation or insufficient data for a definitive decision
AEs, adverse events. (Fig. 2). Among the 32 classified cases, mean (SD) exposure to
a
Defined as blocking the patient from work or other usual activities. study drug was 170.6 (29.8) days for pramipexole, compared with
b
Defined as fatal, life-threatening, significantly disabling, or requiring hospi-
talization.
127.2 (70.3) for placebo.

Randomized:
Table 6 331
Adverse events reported by P5% of either treatment groupa. Randomized but not treated:
2
MedDRA preferred term, n (% of group) Pramipexole Placebo group Treated:
group 329

Fatigue 18 (10.8) 15 (9.2) Placebo: Pramipexole:

Nausea 24 (14.5) 6 (3.7) 329 166
Headache 13 (7.8) 17 (10.4)
RLS 11 (6.6) 11 (6.7) Treated ≥4 wks: Treated ≥4 wks:
Somnolence 11 (6.6) 8 (4.9) 149 (100%) 152 (100%)
Augmentation Augmentation
Nasopharyngitis 7 (4.2) 10 (6.1) not suspected: not suspected:
Muscle spasms 10 (6.0) 4 (2.5) 109 (73.2%) 105 (69.1%)
Augmentation Augmentation
Arthralgia 9 (5.4) 2 (1.2) suspected: suspected:
40 (26.8%) 47 (30.9%)
MedDRA, medical dictionary for drug regulatory activities; RLS, restless legs Augmentation Augmentation
syndrome. excluded: excluded:
a 26 (17.4%) 29 (19.1%)
Listed in descending overall frequency. Augmentation Augmentation
confirmed: confirmed:
9 (6.0%) 14 (9.2%)
Insufficient Insufficient
data: data:
3.6. AE profile 5 (3.4%) 4 (2.6%)

AE frequencies are summarized by general category in Table 5.

Overall, pramipexole recipients were more likely than placebo Fig. 2. Patient disposition, in reference to identification of augmentation.
 B. Högl et al. / Sleep Medicine 12 (2011) 351–360 355

5 before the study’s final visit, only 5 stopped treatment—2 (of 10)
on pramipexole and 3 (of 8) on placebo.

4
New Confirmed Cases

3.8. Withdrawal

3 Of 135 pramipexole recipients with no RLS therapy after their

drug-stop date, 14 (10.4%) exhibited a posttreatment increase of
P4 points on IRLS total score (compared with the highest of the
2
patient’s scores at screening and randomization). Among them, 2
(10.5% of 19) had been on 0.125 mg/d, 5 (11.4% of 44) on
1 0.25 mg/d, 5 (14.3% of 35) on 0.5 mg/d, and 2 (5.4% of 37) on
0.75 mg/d. Of 133 placebo recipients with no RLS therapy after
their drug-stop date, 2 (1.5%) had a posttreatment IRLS increase
0 of P4 points.
5&6 7&8 9&10 11&12 13&14 15&16 17&18 19&20 21&22 23&24 25&26
Weeks
4. Discussion
Fig. 3. Time to augmentation (for cases confirmed by expert panel).
During 6 months of double-blind treatment, pramipexole was
an effective, well-tolerated RLS intervention. On global ratings by
IRLS, CGI-I, and PGI, pramipexole-related improvement was signif-
For confirmed augmentation, the expert panel’s findings were icant vs. placebo, although the magnitudes of the recorded differ-
14 cases (9.2%) among eligible pramipexole recipients, compared ences between pramipexole and placebo tended to be smaller
with 9 cases (6.0%) among eligible placebo recipients (Fig. 2). For than in previous trials. For patient self-rating by IRLS, the differ-
each of the confirmed cases, the time to augmentation is displayed ence was 2.6 points, compared with 9.2, 6.6, and 4.3 points in pre-
in Fig. 3. The mean time to augmentation was 110.0 ± 54.7 days for vious 3-, 6-, and 12-week trials [3,4,6]. Nevertheless, the difference
confirmed cases on pramipexole, compared with 80.6 ± 52.3 days varied markedly among countries (see Table 3). Because this vari-
for confirmed cases on placebo. Moreover, the rate of confirmed ability was mirrored by country-to-country variation in blinded
augmentation after week 12 (the study’s approximate midpoint) clinicians’ ratings of their patients’ response to placebo (see
was 11 cases (79% of the total of 14) among eligible pramipexole Table 4), the variability might represent different countries’ per-
recipients, compared with 4 cases (44% of the total of 9) among eli- ceptions or expectations, shared by patients and their physicians,
gible placebo recipients. Unlike treatment duration, treatment dos- of RLS detriment and its response to intervention. Conceivably, var-
age showed no relation to confirmed-augmentation incidence in iability in country-to-country experience with RLS diagnosis and
the pramipexole group (Fig. 4). assessment may also have had relevant magnitude. On ratings of
Among the total of 23 patients with confirmed augmentation, specific RLS symptoms or of overall RLS severity at specific times
initial improvement in IRLS total score was greater on pramipexole of day, improvement was greatest for nighttime problems and
than on placebo, at a mean 16.1 vs. 8.0 points. Subsequent score maximally significant for efforts to fall asleep but insignificant
deterioration was also greater, at a mean 11.6 vs. 5.7 points. But for daytime problems, unlike some outcomes in previous trials
only 6 of these patients deteriorated to IRLS scores worse than at [5,6].
baseline—4 (of 14) on pramipexole and 2 (of 9) on placebo. More- A further, although probably minor, contribution to narrowed
over, of 18 patients whose confirmed augmentation emerged difference between response to pramipexole and to placebo might
be the present trial’s duration. In a recent meta-analysis of 14
dopamine-agonist trials in RLS, an inverse relationship between
study length and IRLS score improvement was identified, in which
60 adjusted mean IRLS score change from baseline decreased by
approximately 0.66 points for each additional study week [14]. In
50
Proportion of Pramipexole Recipients (%)

another meta-analysis encompassing 24 RLS clinical trials (includ-

50 ing 8 trials of nondopaminergic agents) that reported response
rates—most often (in 17 of the trials) a CGI-I responder rate—
40 placebo response was greater for longer trials by 2.75% per week
[15]. Along with the substantial placebo response that may be ex-
32
29
pected when a CNS disorder and its treatment are both dopaminer-
30 27 gic, so that a placebo recipient’s benefit expectation may stimulate
25
endogenous increase of the very neurotransmitter for which an ac-
21
20 tive agent attempts to substitute [16], it has been proposed that in
16 episodic disorders or those with natural fluctuation in symptom
intensity, relief due only to spontaneous, temporary improvement
10 may reinforce a patient’s belief in placebo efficacy as a trial
proceeds [17].
0 It is also possible that the IRLS may be unduly sensitive to pla-
0
0.125 0.25 0.50 0.75 cebo response. In the aforementioned meta-analysis encompassing
Pramipexole Dosage (mg/d) 24 RLS treatment trials [15], placebo response indeed was mark-
edly larger for IRLS than for all other RLS scales studied (and also
than for all sleep parameters, QoL and tiredness scales, and a peri-
Fig. 4. Pramipexole dosage levels at confirmed augmentation (vs. final dosage odic-leg-movement index). The explanation may lie in the multidi-
among all pramipexole recipients). mensionality of the 10 IRLS items, which ask a patient to rate not
356 B. Högl et al. / Sleep Medicine 12 (2011) 351–360
only RLS symptom frequency and intensity but also RLS impact on
daily life, mood, tiredness, and sleep.
The present trial was designed to investigate withdrawal as a
safety parameter. In general, withdrawal is considered most likely
to develop after high dosage. Here, however, the frequency of post-
treatment IRLS score deterioration by P4 points was independent
of dosage. The deterioration could therefore be interpreted as ver-
ifying loss of benefit if an effective treatment stops [5]. Neverthe-
less, the treatment-group difference (of 10.5% among
pramipexole recipients vs. 1.5% among placebo recipients) sug-
gests that pharmacodynamic mechanisms may be involved.
The present trial was also designed to investigate augmentation
as a facet of safety, utilizing placebo control (making this the first
pramipexole trial to have done so). Augmentation is perhaps the
most clinically relevant complication of long-term dopaminergic
treatment of RLS [18]. In the present trial, its 6-month incidence
on pramipexole resembled the rate for placebo, both for classified
augmentation, a highly conservative analysis in which all cases
involving inadequate information were counted as augmentation
(11.8% vs. 9.4%), and for confirmed augmentation (9.2% vs. 6.0%).
This resemblance might reflect an RLS sample too small to calcu-
late augmentation rates, especially given the length of the study.
As noted above, the sample size was based on statistical testing
for efficacy, not augmentation (which was not a primary endpoint,
and for which there was no previous experience to guide in design-
ing controlled trials).
It should be added that in a clinical trial, assessment of augmen-
tation may be hampered by fixed maximum dosage (at which level
a patient’s fulfillment of augmentation criteria might signify only
insufficient RLS treatment). It may likewise be hampered by the
artificiality of an expert panel not only blinded to treatment (for
the sake of placebo control) but also unable to interview patients.
In the present trial, the recorded rates might also reflect the use of
a serum ferritin level 630 ng/mL as an exclusion criterion. The
present trial is alone among published trials in employing this va-
lue. In a large 30-week study of levodopa vs. cabergoline for RLS
[19], lower ferritin at baseline was a risk factor for occurrence of
augmentation during the trial, and in a recently published series
of 302 consecutive RLS patients [20], those with current augmen-
tation had lower ferritin values (although the mean was within
the normal range) than those without augmentation.
Even so, the present trial’s 6-month rates of augmentation
resembled those in previous, retrospective studies of pramipexole
for RLS, among which Ferini-Strambi [21] reported a rate of 8.3% in
patients treated at least 6 months and Silber et al. [22] a rate of
approximately 9% at 6 months. For its part, the similarity between
rates of augmentation for pramipexole and for placebo suggests
that beginning or mild augmentation can be difficult to distinguish
from natural symptom fluctuation occurring in untreated RLS.
The critical question is whether or not the likelihood of random
fluctuation, which can be mistaken for augmentation, remains sta-
ble over time, while rates of augmentation on medication increase
with duration of treatment. The present study’s augmentation-
incidence findings show this general pattern. Moreover, a relation-
ship between augmentation and duration of treatment is
consistent with prior studies of open-label RLS therapy. For
open-label pramipexole, Silber et al. [22] showed (in a Kaplan–
Meier plot) that augmentation occurred in about 6–8% of patients
over the first 6 months of treatment, and in another 12–14% over
the next 6 months. The study reported overall rates of 20% for
1 year and 33% for a mean 27.2 months. Similarly, Winkelman
et al. [23] reported a rate of 32% for a mean 21.2 months. In a re-
cent 6-month study of levodopa for RLS [24], augmentation also
occurred progressively, with a rate of 60% at 6 months (and no
indication that the rate would not have continued to increase with
duration of treatment).
Together, the available data suggest some guidance for evalua-
tion of suspected augmentation. In general, for dopamine agonists
such as pramipexole, augmentation should not be expected within
the first 3 months of treatment, at least in patients with a serum
ferritin level >30 ng/mL and no history of augmentation on dopa-
mine agonists. Hence, events reported during the first 3 months
should be carefully assessed as possible random RLS symptom fluc-
tuation. With increasing duration of treatment, augmentation rates
can be expected to increase (while RLS fluctuation may not in-
crease). Indeed, within pramipexole’s approved dosage range, and
for serum ferritin >30 ng/mL, treatment duration appears to be
more important than treatment dosage as an augmentation risk
factor, at least during the first 6 months.
Lastly, cases of augmentation developing during the first
6 months appear likely to be mild, at least in patients with serum
ferritin >30 ng/mL. Mildness would make early cases all the harder
to differentiate from ‘‘noise,’’ e.g., RLS fluctuation, or, in clinical tri-
als, the varying diagnostic accuracy in different countries or at dif-
ferent clinical centers. This does not preclude the possibility that
augmentation severity as well as prevalence increases with dura-
tion of treatment, even at dosage not exceeding an approved max-
imum. For researchers, the present trial’s most salient implication
may be that a fuller understanding of the clinical impact of aug-
mentation will have to be based on comparative RLS-intervention
trials lasting substantially longer than 6 months.
Disclosures
Birgit Högl has been a consultant for Boehringer Ingelheim (BI),
GlaxoSmithKline (GSK), and UCB (for RLS); has served on advisory
boards for BI, UCB, Cephalon, Merz, Jazz, Pfizer, and Lundbeck; and
received speaker honoraria from BI, UCB, Cephalon, GSK, Pfizer,
and Novartis. Diego Garcia-Borreguero received honoraria for advi-
sory boards or lectures from Boehringer Ingelheim, GlaxoSmithK-
line, UCB Pharma, Pfizer, Xenoport, Sanofi-Aventis, Jazz, and MSD.
Stefanie Brenner, Michael Busse, and Mandy Fraessdorf are employ-
ees of Boehringer Ingelheim. Stefan Albrecht was an employee of
Boehringer Inghelheim at the time of this study and approved the fi-
nal draft of this paper. Claudia Trenkwalder is a consultant for Nov-
artis and Mundipharma. She is on the advisory board of UCB and
Boehringer Ingelheim (BI), and has received honoraria from UCB,
BI, GlaxoSmithKline, Pfizer, and Teva. She has also provided expert
testimony for Mundipharma and Axxonis. Richard P. Allen has
received research support from GlaxoSmithKline, NIH, and Pharma-
cosmos, and consultation fees, honoraria, or travel support from
GlaxoSmithKline, Boehringer Ingelheim, Jazz, UCB, Xenoport, Luit-
pold, Pharmacosmos, Pfizer, EMD-Serono, Neurogen, Novartis, and
Orion Pharma. Luigi Ferini-Strambi received honoraria from serving
on the scientific advisory board of UCB Pharma, Boehringer Ingel-
heim, GlaxoSmithKline, Sanofi-Aventis, and Transept Pharmaceuti-
cals. Werner Poewe has served on the advisory board of and has
received consultancy and lecture fees (including honoraria) from
Eisai, Teva, Novartis, GlaxoSmithKline, Boehringer Ingelheim, UCB,
Orion Pharma, Merck Serono, and Schering Plough in relation to clin-
ical drug development programs for Parkinson’s disease.
Conflicts of Interest
The ICMJE Uniform Disclosure Form for Potential Conflicts of
Interest associated with this article can be viewed by clicking on
the following link: doi:10.1016/j.sleep.2010.12.007.
Acknowledgement
This study was funded by Boehringer Ingelheim GmbH.
 B. Högl et al. / Sleep Medicine 12 (2011) 351–360 357

Appendix Investigatorsand other important participants at each study site

Total number of centres: 42

Centre Investigator name and Affiliation and address Number of
number role patients enrolled
per centre
AP-Slovakia Number of patients enrolled in country: 13
4201 Tormasiova – PI Fakultná nemocnica Louisa 2
Feketeova – Investigator Pasteura, Rastislavova 43,Košice
Workplace: Neurologická klinika,
Trieda SNP 1
040 66 Košice
4202 Droppa – PI Nemocnica s poliklinikou Brezno, 3
Zlevsky – Investigator n.o., Neurologické oddelenie,
Banisko 1
977 01 Brezno
4203 Kurca – PI Martinská fakultná nemocnica 2
Grofik – Investigator Neurologická klinika, Kollárova 2
036 59 Martin
4204 Saling – PI Fakultná nemocnica s 3
poliklinikou Bratislava
Workplace: Dérerova nemocnica
II. Neurologická klinika, Limbova 5
833 05 Bratislava
4205 Turcani – PI Neurologická ambulancia, 3
Goldenberg – Investigator Gorkého 1
811 01 Bratislava

Austria Number of patients enrolled in country: 6

4302 Högl – PI Univ.-Prof. Dr. Birgit Högl, 3
Medizinische Universität
Innsbruck
Universitätsklinik für
Neurologie, Spezialambulanz für
Schlafmedizin, Anichstraße 35
A-6020 Innsbruck
4304 Ransmayr – PI Prim. Univ.-Doz. Dr. Gerhard 3
Ransmayr, AKH der Stadt Linz
Abt. für Neurologie u. Psychiatrie,
Krankenhausstraße 9
A-4021 Linz

Belgium Number of patients enrolled in country: 6

3201 De Volder – PI Universitair Ziekenhuis 6
Antwerpen, Centrum voor
klinisch slaap-en
waakonderzoek, Wilrijkstraat 10
2650 Edegem, Belgium

Finland Number of patients enrolled in country: 99

35801 Partinen – PI Skogby Sleep Clinic Kumputie 3 25
Karjalainen – Investigator 02980 Espoo, Finland
Mikkola – Investigator
Ruotsalainen – Investigator
Sved – Investigator
35802 Myllylä – PI Oulu University Hospital 9
Kärppä – Investigator Neurology Clinic, Kajaanintie 50
Korpelainen – Investigator 90220 Oulu, Finland
35804 Alhainen – PI Muistikeskus-Joensuu, 12
Lääkärikeskus Suinuu, Torikatu
17, 80100 Joensuu, Finland
35805 Jama – PI Vitalmed Oy, Sitratori 3, 00420 42
Helsinki, Finland
35806 Pietilä – PI Neurokeskus Oy, Hatanpään 11
valtatie 1, Koskikeskus
33100 Tampere, Finland

Germany Number of patients enrolled in country: 111

4901 Albrecht – PI Dr. med. Walter Albrecht, 5
Butz – Investigator Neurologie Psychiatrie
Psychotherapie
Naturheilverfahren, Schöner
Graben 19, 73479 Ellwangen
(continued on next page)
358 B. Högl et al. / Sleep Medicine 12 (2011) 351–360

Appendix (continued)
Total number of centres: 42
Centre Investigator name and Affiliation and address Number of
number role patients enrolled
per centre
4902 Bergtholdt – PI Emovis GmbH, Bereich Klinische 34
Filler – Investigator Forschung, Dr. med. Bettina
Novy – Investigator Bergtholdt, Wilmersdorfer
Straße 79, 10629, Berlin
4903 Ehret – PI Dr. med. Reinhard Ehret, 4
Friedrich – Investigator Schloßstr. 29
2163 Berlin-Steglitz
4904 Gussmann – PI ClinPharm Intern. GmbH 17
Effenberg – Investigator Studienzentrum, Berlin
Stössel – Investigator Dr. med. Manfred Gussmann
Janusz-Korczak-Str. 8
12627 Berlin (Hellersdorf)
4905 Stahl – PI ClinPharm International GmbH 15
Degtyareva – Investigator Dr. med. Hans-Detlev Stahl
Sigal – Investigator Johannisplatz 1, 04103 Leipzig
Streck – Investigator
4906 Simonow - PI Alexander Simonow 2
Neurologische Praxis
Hauptstrasse 106
35745 Herborn
4907 Klein – PI Medizinisches Studienzentrum 5
Helbig – Investigator Dr. med. Martin Klein
Oehler – Investigator Augustinerstr. 15
97070 Würzburg
4908 Schumann – PI Neurologie, Psychiatrie, 9
Bitter – Investigator Psychotherapie, Dr. med. Günther
Schumann, Castroper Hellweg
422, 44805 Bochum
4909 Benes – PI Somni bene, Institut für 20
Meissner – Investigator Medizinische Forschung und
Schlafmedizin Schwerin GmbH
Dr. med. Heike Benes
Arsenalstr. 10, 19053 Schwerin

Ireland Number of patients enrolled in country: 26

35301 Mc Laughlin – PI Main Street, Carrigtwohill 8
Co., Cork, Ireland
35302 Dillon – PI Carlow Gate, Castledermot 5
Co., Kildare, Ireland
35303 Kelly – PI Western House Medical Centre 13
Western Street, Clonmel
Co. Tipperary, Ireland

Netherlands Number of patients enrolled in country: 56

31001 Rol – PI Huisartsenpraktijk H. Rol 6
Dr. H. Rol, Zwarteweg 65
2121 BB, BENNEBROEK
31002 Jonker – PI Huisartsenpraktijk Dr. Jonker 21
Dr. F. Jonker, F. Clockstraat 167
9665 BJ OUDE PEKELA
31003 Feis – PI Huisartspraktijk Dr. Feis 11
Dr. W.L. Feis, H. Westerstraat 100
9665 AR OUDE PEKELA
31004 De Backer – PI Huisartspraktijk W.A. de Backer 5
Hoekstra – Investigator Dr. W.A. de Backer, Haagweg 191
2281 AK RIJSWIJK
31005 De Jong – PI Huisartsenpraktijk Dr. de Jong 8
Assema-de Jong – Dr. A. de Jong, Burgermeester
Investigator Hoogenboomlaan 9
Montauban – Investigator 1718 BG HOOGWOUD
31006 Prak – PI Huisartsenpraktijk Prak 5
Dr. H. Prak, Rembrandtlaan 3
9581 AJ Musselkanaal

Spain Number of patients enrolled in country: 27

3401 Egatz – PI Instituto de Investigaciones del 2
Calvo – Investigator Sueño, Alberto Alcocer, 19
Larrosa – Investigator 28036 Madrid
3402 Estivill – PI Clínica del Son Estivill-Dexeus 11
Albares – Investigator Rosales, 9, 08017 Barcelona
Pascual – Investigator
 B. Högl et al. / Sleep Medicine 12 (2011) 351–360 359

Appendix (continued)
Total number of centres: 42
Centre Investigator name and Affiliation and address Number of
number role patients enrolled
per centre
3403 Poza – PI Hospital de Donostia, Neurology 6
Department, Paseo Dr.
Beguiristain, s/n
20014 San Sebastián
3405 Paniagua – PI Hospital Virgen de las Nieves 2
Sanchez – Investigator Neurophysiology Department
Pabellón de Traumatología
Carretera de Jaén, s/n
18014 Granada
3406 Ordoño – PI Hospital Arnau de Vilanova 6
Chornet – Investigator Neurophysiology Department
San Clemente, 12
46015 Valencia

United Kingdom Number of patients enrolled in country: 152

44001 Robinson – PI (Lead Synexus Manchester Clinical 18
Investigator) Research Centre, 1st Floor,
Williams House, Manchester
Science Park, Lloyd Street North
Manchester M15 6SX
44002 Shaw – PI Thames Valley Clinical Research 20
Centre, Whitley Glebe, 11 Glebe
Road, Off Christchurch Gardens
Reading, Berkshire RG2 7AG
44003 Pawa – PI Synexus Lancashire Clinical 33
Research Centre, 24 Eaton
Avenue, Matrix Park, Buckshaw
Village, Chorley PR7 7NA
44004 Donnachie – PI Synexus Scotland Clinical 26
Research Centre, 3rd Floor,
Erskine House, Clydebank
Business Park, Clydebank,
Glasgow G81 2DR
44005 Pavel-Knox – PI Synexus Merseyside Clinical 23
Research Centre, Burlington
House, Crosby, Waterloo,
Liverpool L22 0LG
44006 Sarimentio – PI Synexus Midlands Clinical 32
Research Centre, Birmingham
Research Park, Vincent Drive
Edgbaston, Birmingham B15 2SQ

References manual for psychopharmacology, revised ed. Rockville, MD: NIMH USA; 1976.
p. 216–22.
[10] Abetz L, Vallow SM, Kirsch J, Allen RP, Washburn T, Earley CJ. Validation of the
[1] Silber MH, Ehrenberg BL, Allen RP, et al. An algorithm for the management of
restless legs syndrome quality of life questionnaire. Value Health 2005;8:
restless legs syndrome. Mayo Clin Proc 2004;79:916–22.
157–67.
[2] Oertel WH, Trenkwalder C, Zucconi M, et al. State of the art in restless legs
[11] Kohnen R, Oertel WH, Stiasny-Kolster K, Benes H, Trenkwalder C. Severity
syndrome therapy: practice recommendations for treating restless legs
rating of restless legs syndrome: validation of the RLS-6 scales. Sleep
syndrome. Mov Disord 2007;22:S466–75.
2004;27(Suppl.):A304. Abstract 680.
[3] Partinen M, Hirvonen K, Jama L, et al. Efficacy and safety of pramipexole in
[12] Garcia-Borreguero D, Allen RP, Kohnen R, et al. Diagnostic standards for
idiopathic restless legs syndrome: a polysomnographic dose-finding study—
dopaminergic augmentation of restless legs syndrome: report from a World
the PRELUDE study. Sleep Med 2006;7:407–17.
Association of Sleep Medicine-International Restless Legs Syndrome Study Group
[4] Oertel WH, Stiasny-Kolster K, Bergtholdt B, et al. Efficacy of pramipexole in
consensus conference at the Max Planck Institute. Sleep Med 2007;8:520–30.
restless legs syndrome: a six-week, multicenter, randomized, double-blind
[13] Garcia-Borreguero D, Kohnen R, Högl B, et al. Validation of the augmentation
study (effect-RLS study). Mov Disord 2006;21:343–53.
severity rating scale (ASRS): a multicentric, prospective study with levodopa
[5] Trenkwalder C, Stiasny-Kolster K, Kupsch A, Oertel WH, Koester J, Reess J.
on restless legs syndrome. Sleep Med 2007;8:455–63.
Controlled withdrawal of pramipexole after 6 months of open-label
[14] Baker WL, White CM, Coleman CI. Effect of nonergot dopamine agonists on
treatment in patients with restless legs syndrome. Mov Disord 2006;21:
symptoms of restless legs syndrome. Ann Fam Med 2008;6:253–62.
1404–10.
[15] Fulda S, Wetter TC. Where dopamine meets opioids: a meta-analysis of the
[6] Winkelman JW, Sethi KD, Kushida CA, et al. Efficacy and safety of pramipexole
placebo effect in restless legs syndrome treatment studies. Brain 2008;131:
in restless legs syndrome. Neurology 2006;67:1034–9.
902–17.
[7] Allen RP, Picchietti D, Hening WA, et al. Restless legs syndrome: diagnostic
[16] de la Fuente-Fernandez R, Lidstone S, Stoessl AJ. Placebo effect and dopamine
criteria, special considerations, and epidemiology. A report from the restless
release. J Neural Trans Suppl 2006;70:415–8.
legs syndrome diagnosis and epidemiology workshop at the National Institutes
[17] Perlis ML, McCall WV, Jungquist CR, Pigeon WR, Matteson SE. Placebo effects in
of Health. Sleep Med 2003;4:101–19.
primary insomnia. Sleep Med Rev 2005;9:381–9.
[8] Walters AS, LeBrocq C, Dhar A, et al. Validation of the International Restless
[18] Allen RP, Earley CJ. Augmentation of the restless legs syndrome with
Legs Syndrome Study Group rating scale for restless legs syndrome. Sleep Med
carbidopa/levodopa. Sleep 1996;19:205–13.
2003;4:121–32.
[19] Trenkwalder C, Högl B, Benes H, Kohnen R. Augmentation in restless legs
[9] National Institute of Mental Health (NIMH). Early clinical drug evaluation unit
syndrome is associated with low ferritin. Sleep Med 2008;9:572–4.
(ECDEU). Clinical global impressions. In: Guy W, editor. ECDEU assessment
360 B. Högl et al. / Sleep Medicine 12 (2011) 351–360

[20] Frauscher B, Gschliesser V, Brandauer E, et al. The severity range of restless
legs syndrome (RLS) and augmentation in a prospective patient cohort:
association with ferritin levels. Sleep Med 2009;10:611–5.
[21] Ferini-Strambi L. Restless legs syndrome augmentation and pramipexole
treatment. Sleep Med 2002;3(Suppl.):S23–5.
[22] Silber MH, Girish M, Izurieta R. Pramipexole in the management of restless
legs syndrome: an extended study. Sleep 2003;26:819–21.
[23] Winkelman JW, Johnston L. Augmentation and tolerance with long-term
pramipexole treatment of restless legs syndrome (RLS). Sleep Med 2004;5:
9–14.
[24] Högl B, Garcia-Borreguero D, Kohnen R, et al. Progressive development of
augmentation during long-term treatment with levodopa in restless legs
syndrome: results of a prospective multi-center study. J Neurol 2010;257:
230–7.