Twenty questions about multiple sclerosis clinical trials methodologies.

Abstract

The heterogeneity of methods used in multiple sclerosis (MS) clinical trials prevents fair comparison of trials and reduces confidence in the validity of the therapeutic claims made. The validity of recent clinical trials is lessened by the following factors: MS shows variability in type and in rates of disease progression; primary progressive MS may not be the same disease as typical MS, and inclusion of subjects with this condition may have skewed results of trials to date. Using a new model, "relapsing-remitting" and "secondary progressive" MS are considered to represent earlier and later stages of the same disease. The variety of endpoints used in clinical trials impairs comparisons. The differences between EDSS stages vary at different levels and it is concluded that this is no longer the most appropriate tool, although it could be improved by modifying the scoring or scales to assess certain focused items. The clinical significance of a reduction in relapse rate is questioned, as are the inclusion criteria employed in recent trials. Drug doses based upon body mass differ from those based on surface area, making it hard to compare the effects of trial agents. The definitions of "sustained worsening" are not uniform and the concept is complicated by regression to the mean. Trials should continue for long enough to be sure that any beneficial effects noted are permanent. Although extensions provide better long-term data, they are usually statistically underpowered and clinically and demographically imbalanced. Ethically, if benefit is determined to be present, a trial should be stopped so that all subjects may be offered the beneficial agent, but determination of benefit may be imperfect. The "intention to treat" paradigm is a shibboleth, providing data on effectiveness rather than efficacy. The simple listing and addition of unwanted effects (UEs) is unproductive. Trivial UEs detract little from quality of life and are unimportant. The remainder should be separated between the notable and the serious, judging the net benefit of the agent causing them accordingly. Conventional MRI measures the density of hydrogen protons and thus is a map of water, which implies edema and thus the presence of a local inflammatory response, but conventional images do not inform on de- or re-myelination, axonal loss or gliosis and correlate poorly with clinical scores. It is concluded that recent MS therapeutic trials are subject to important criticism.