Background: Many patients with multiple sclerosis do not respond to interferon beta, which is widely used as an immunomodulatory treatment in this disease. We aimed to assess the functional relevance of tumour necrosis factor (TNF)-related apoptosis inducing ligand (TRAIL), which is upregulated on incubation with interferon beta, for clinical treatment response.
Methods: We quantified gene expression longitudinally by realtime-PCR of the peripheral immune cells of 82 patients with multiple sclerosis. In a first cohort of 62 patients, 20 were classified as first-year responders since they did not have relapses during treatment with interferon beta 1a; 19 were classified as first-year non-responders; and 23 developed neutralising antibodies to interferon beta. A second cohort, also characterised by MRI, consisted of 11 patients on interferon beta 1a and nine patients who were not treated. Concentrations of soluble TRAIL were determined by ELISA in serum samples of nine non-treated patients, 49 patients before treatment (29 responders, 20 non-responders), as well as longitudinally in a subset of 23 patients.
Findings: In both patient cohorts, drug-responders could be distinguished from non-responders by early and sustained induction of TRAIL (p<0.0001, each). In the presence of neutralising antibodies, initial upregulation of TRAIL expression was subsequently abrogated. Raised concentrations of soluble TRAIL in patients' serum samples before the start of treatment allowed prediction of the treatment response in the first year (ROC analysis with area under the curve 0.879 [0.785-0.974]).
Interpretation: Our data suggest that TRAIL expression is a candidate for pretreatment assessment and might thus be used as a prognostic marker of treatment response to interferon beta in multiple sclerosis. Furthermore, our observations have implications for the development of future immunoregulatory strategies in multiple sclerosis therapy.