A "cure" for Parkinson's disease: can neuroprotection be proven with current trial designs?Mov Disord. 2004 May; 19(5):491-8.MD
Current medical and surgical therapies for Parkinson's disease provide symptomatic control of motor impairments rather than slowing or halting the progression of the disease. Previous clinical trials examining drugs such as dopamine agonists and selegiline for neuroprotective effects used "surrogate" outcomes, including clinical measures (rating scales, time to require levodopa), neuroimaging techniques (beta-CIT single photon emission computed tomography; fluorodopa positron emission tomography), and mortality tracking. These studies failed to provide conclusive results because of design faults such as failing to control for symptomatic effects, small sample size, and not accounting for the possible effects of drugs on radionuclide tracer handling. Lessons must be learned from these failed neuroprotection trials. This review summarises the problems with previous neuroprotection studies and makes recommendations for future trial design. It is concluded that the primary outcome of explanatory trials should continue to be clinical measures such as the Unified Parkinson's Disease Rating Scale (UPDRS). It should be assumed that all agents have a symptomatic effect, which necessitates evaluation after a prolonged drug washout period. To achieve the evaluation after a prolonged drug washout period more effectively, trials must be performed in early disease and over a short period (6-12 months) so that symptomatic therapy is not required. To achieve adequate statistical power, these trials will need to include thousands of patients. Radionuclide imaging can only be used in such trials after considerable methodological work has been performed to establish its validity and reliability. To be affordable, such large explanatory trials need more streamlined designs with fewer hospital visits, fewer outcome measures, and rationalised safety monitoring. The clinical effectiveness of promising compounds from explanatory trials will need to be established in large long-term pragmatic trials using outcome measures such as quality of life, cost-effectiveness, and mortality. Such pragmatic trials could be continuations of the explanatory trials: after the primary outcome of the explanatory study (e.g., UPDRS) has been reported in an interim analysis, the trial could be continued for a further 5 to 10 years to report on quality of life and health economics outcomes.