Multiple sclerosis: use of MRI in evaluating new therapies.

Although definitive evaluation of new therapies should be based on clinically meaningful outcomes, there are major difficulties to overcome when conducting treatment trials with clinical endpoints such as relapse rate or progression in disability. The variable and unpredictable natural history of MS necessitates large studies (hundreds of patients) of long duration (2-3 years), with an active treatment group being compared to a control group. It is thus not surprising that there has been much effort to identify alternative laboratory markers of disease activity to monitor treatment efficacy. To be an effective replacement (or surrogate, as it is often called) of clinical outcomes the laboratory measure of disease activity needs to be objective, sensitive and cost effective, accurate, reproducible, and most importantly, predictive of clinical outcome. Magnetic resonance imaging (MRI) is currently the only widely used surrogate outcome. This article reviews its status as a surrogate, and proposes MRI trial designs to address specific therapeutic questions.