Objectives Undertake a systematic review to assess what potential surrogate markers for disease progression in Parkinson's disease (PD) exist, whether any meet the criteria for use in clinical trials, and if not which looks most promising.

Methods Both MEDLINE and EMBASE (1950–2010) were searched using five different search strategies (four based on identified keywords and one on MeSH headings): (1) PD and blood; (2) PD and urine or Cerebrospinal fluid (CSF); (3) PD and imaging; (4) PD and neurophysiology; (5) PD and MeSH and biomarkers. A single researcher assessed identified abstracts to select papers which merited review in full. Reference lists of review articles and included articles were reviewed to identify important articles which may have been missed by the search strategy. We validated the electronic search by hand-searching the two journals from which most articles came. We included studies of participants with a clinical diagnosis of Idiopathic PD diagnosed by formal criteria, or by clearly described clinical means. We made no restriction on participant's age, disease duration, drug treatment or study design. We included studies of any test used to investigate disease progression in Parkinson's disease, including (but not restricted to) imaging, blood tests, tests of CSF and neurophysiological tests. We looked for associations of the test result with any clinical measure of disease progression—impairment, global cognitive function, disability, handicap, quality of life and survival. Associations to individual symptoms, parts of scoring systems, mood, disease duration, complications related to therapy and treatment status were excluded. Only papers available in full and in English were included. Study quality was assessed using a modified version of the assay methods and study design section of the Reporting recommendations for tumour MARKer prognostic studies (REMARK) reporting recommendations.

Results 183 studies were finally included: 163 (89%) cross-sectional, 20 (11%) longitudinal. Hand searching to validate the electronic search strategy revealed a sensitivity of 71.4% and a specificity of 97.2%. The median follow-up time in the longitudinal studies was only 2.0 years (IQR 1.1–3.5). Included studies were generally of poor quality, being cross-sectional with small numbers of participants, applying excessive inclusion/exclusion criteria, with flawed methodologies and inappropriate statistical analyses.

Conclusions We found insufficient evidence to recommend the use of any biomarker for disease progression in PD clinical trials at present. Future biomarker studies should be larger longitudinal studies, with follow-up at several time points over several years, apply minimal entry criteria and use appropriate statistical methods.