The powerful pre-treatment effect: placebo responses in restless legs syndrome trials

Objective To investigate whether dopaminergic pre-treatment alters placebo and dopamine agonist responses in restless legs syndrome (RLS). Methods Two large, multi-centre trials (SP790 and SP792; registration numbers NCT00136045 and NCT00135993) on the efficacy of rotigotine in RLS reported supplemental International RLS (IRLS) sum score data for pre-treated and drug-naive patients, allowing for the estimation of the regression slope of the clinical response (change in the IRLS sum score) on baseline IRLS sum score. Results In both trials, patients pre-treated with dopaminergic medications tended to have blunted responses after placebo administration compared with drug-naive patients. In the SP790 study, the pre-treated group had a negative slope (i.e. the response observed after placebo administration decreased as the baseline IRLS sum score increased), whereas the slope was positive in drug-naive patients (slope, −0.43 vs. 0.28; P = 0.027). In the SP792 study, the two slopes were parallel (P = 0.84), but the magnitude of the response after placebo administration was smaller in the pre-treated group (6.31 vs. 10.49; P = 0.0089). Pre-treatment had no significant effect on rotigotine-group responses in either of the two studies. Conclusions In RLS trials, dopaminergic pre-treatment tends to increase the apparent effect of new dopaminergic drugs by decreasing the placebo effect in the placebo arm without substantially modifying the placebo effect in the active treatment arm. This observation highlights that placebo-controlled trials are not necessarily placebo-effect controlled trials.