A robust type I interferon gene signature from blood RNA defines quantitative but not qualitative differences between three major IFNβ drugs in the treatment of multiple sclerosis

We analysed gene expression microarray data from whole blood samples from 228 multiple sclerosis (MS) patients either untreated or treated with one of three alternative commonly used interferon beta (IFNβ) disease modifying drugs: Avonex (×1 weekly), Betaseron (every second day) or Rebif (×3 weekly). Patient injections were not timed to coordinate sample collections, thus providing a global transcriptomic profile for each population of patients studied. Three hundred and fifty one genes were significantly differentially expressed by at least one of the IFNβ drugs. Despite the different drug sources with distinct injection and dosage protocols, a striking similarity was found in the identity and functional classes of the differentially expressed genes induced. Using the 25 most-upregulated genes, we defined a robust IFNβ gene expression signature that quantifies the IFN activation state per blood sample collected irrespective of the type of IFNβ therapy. This 25-gene signature also defined basal IFN activation states among untreated MS patients, which differed among individuals but remained relatively constant per patient with time. The maximum drug-induced IFN-activation state was similar for all three drugs despite a 1.7-2.0-fold diminished average effect for Avonex. This and a more erratic effect of Avonex per patient across longitudinal measurements is likely a result of its reduced injection frequency. In summary, we have defined a robust blood-derived type I IFN gene signature from MS patients. This signature could potentially serve to generically quantify the systemic Type I IFN activation status for any other clinical manifestation, inclusive of other autoimmune diseases.