Defining meaningful outcome measures in trials of disease-modifying therapies in Parkinson's disease.

INTRODUCTION: Although a range of treatments is available for Parkinson's disease (PD), conventional therapies merely improve the symptoms of the disorder and are not considered to have any effect on underlying neurodegeneration. Neuronal loss continues inexorably and leads to increasing handicap and disability as the disease advances. A fundamental goal of PD research has been to develop drugs that delay or halt progression - so-called disease-modifying therapies (DMTs). Although this effect has been claimed for a number of different agents over the past decades - most recently rasagiline - these effects have not proved to be reproducible. AREAS COVERED: The first part of this article reviews the results of the major neuroprotective trials conducted in PD and we argue that flaws in the design of these studies may account for their equivocal results. To establish whether a therapy has an effect on the prevention or postponement of disability, relevant endpoints must be identified for clinical trials. We go on to examine the results of natural history studies in PD, identified using a MEDLINE search with the appropriate MesH headings, can tell us about the evolution of the disorder, and how these insights can in turn be used to identify meaningful clinical endpoints for DMT trials. We frame our discussion by proposing two alternative models of PD progression and illustrate how the use of such models can inform the design of future DMT trials in PD, including the next generation of studies using novel interventional therapies. EXPERT OPINION: The conventional design of DMT trials in PD is flawed. Insights from natural history studies allow the identification of outcome measures which reflect true clinical progression - for example, the failure of postural reflexes - and must inform future research in this important area.