Designs to Detect Disease Modification

Designing a trial to determine whether or not an intervention has modified the underlying course of the disease is straightforward for certain conditions, such as cancer, in which it is possible to directly measure the disease course. For many other diseases, the disease course is latent, and one must rely on indirect measures such as clinical symptoms to quantify the effects of interventions. In this case, it is difficult with conventional trial designs to determine the extent to which the treatment is modifying the disease course as opposed to merely alleviating the symptoms of the disease. This distinction has become critically important in the study of treatments for neurodegenerative diseases such as Alzheimer’s disease and Parkinson’s disease, but it applies to many other diseases as well. This chapter discusses proposed strategies for trial design to attempt to distinguish between the disease-modifying and symptomatic effects of a treatment in diseases with a latent disease course. Two-period designs, such as the withdrawal design and the delayed start design, are being used for this purpose, most commonly in neurodegenerative disease. In these designs, the first period involves a standard randomization of participants to active and placebo treatments. In the second period, those in the active treatment group are switched to placebo (withdrawal design), or those in the placebo group are switched to active treatment (delayed start design). These designs are reviewed in detail in terms of their underlying assumptions, limitations, and strategies for statistical analysis.