Towards a neuroimaging biomarker for amyotrophic lateral sclerosis.

400 www.thelancet.com/neurology Vol 10 May 2011 might benefi t from early initiation of therapy to prevent or delay progression to multiple sclerosis. Data from this study could also provide clues as to the causes and development of multiple sclerosis, because the interactions between genetic and environmental factors were analysed in children during the period of fi rst exposure to environmental factors. A major limitation inherent to this type of study is the likelihood of outcome misclassifi cation because of insuffi cient follow-up. Because conversion to multiple sclerosis is dependent on time, the group with monophasic demyelination might include some children who would develop multiple sclerosis given extended follow-up. However, in Banwell and colleagues’ study, the probability of misclassifi cation was reduced by the fact that conversion to multiple sclerosis is most likely to happen within the fi rst year after the acute demyelinating event. The Article by Banwell and colleagues will set a precedent for future prospective studies, in both paediatric and adult populations, through its integration of genetic, environmental, clinical, and radiological data, and its production of a refi ned assessment of the risk of multiple sclerosis at the time of a fi rst event that is suggestive of CNS demyelination. As more knowledge is gained about the genetic and environmental factors that determine the risk of multiple sclerosis, predictive models will be able to incorporate new predictors and analyse, in a multivariable manner, their individual or synergistic contributions to outcomes.