Are delayed-start design trials to show neuroprotection in Parkinson's disease fundamentally flawed?

Abstract

Considerable effort has gone into preclinical neuroprotection research in Parkinson's disease (PD) and several large clinical trials have been mounted, but no agent has been conclusively shown to be protective. The latest innovation in PD neuroprotection trial design is the delayed-start design trial. If patients with early PD do better after 12 to 18 months of immediate drug therapy compared to those in whom it is delayed for 6 to 9 months, this is attributed to a neuroprotective effect. However, delayed-start design trials may be fundamentally flawed. It has been suggested that physiological mechanisms compensating for the loss of dopaminergic neurones in early PD may be deleterious and that immediate treatment may prevent such mechanisms and thereby be neuroprotective. If this hypothesis is correct, any drug with a symptomatic effect will be neuroprotective in early PD and delayed-start design trials will show this generic effect, not a neuroprotective effect specific to the drug. Delayed-start design trials require patients to potentially stay untreated for 6 to 9 months. This may lead to the selection of slowly progressive types of PD, such as that in younger patients and those with tremor-dominant disease, so the results may not be generalizable to the majority of patients. Delayed-start design trials are powered to find small differences in total UPDRS score which may not be clinically significant; larger and longer placebo-controlled trials are required to confirm the clinical significance of their findings. These arguments add to the growing tide of opinion for a fundamental rethink of our policy toward neuroprotection research in PD.