Multiple sclerosis, the measurement of disability and access to clinical trial data

Background: Inferences about long-term effects of therapies in
multiple sclerosis (MS) have been based on surrogate markers
studied in short-term trials. Nevertheless, MS trials have been
getting steadily shorter despite the lack of a consensus definition
for the most important clinical outcome - unremitting progression
of disability. Methods: We have examined widely used surrogate
markers of disability progression in MS within a unique database of
individual patient data from the placebo arms of 31 randomised
clinical trials. Findings: Definitions of treatment failure used in
secondary progressive MS trials include much change unrelated to
the target of unremitting disability. In relapsing-remitting MS,
disability progression by treatment failure definitions was no more
likely than similarly defined improvement for these disability
surrogates. Existing definitions of disease progression in
relapsing-remitting trials encompass random variation, measurement
error and remitting relapses and appear not to measure unremitting
disability. Interpretation: Clinical surrogates of unremitting
disability used in relapsing -remitting trials cannot be validated.
Trials have been too short and/or degrees of disability change too
small to evaluate unremitting disability outcomes. Important
implications for trial design and reinterpretation of existing
trial results have emerged long after regulatory approval and
widespread use of therapies in MS, highlighting the necessity of
having primary trial data in the public domain.