Two recombinant human interferon-beta 1a pharmaceutical preparations produce a similar transcriptional response determined using whole genome microarray analysis

Abstract

Objectives: Recombinant human interferon-beta (IFN-b) is a well-established treatment for multiple sclerosis (MS). The regulatory process for marketing authorization of biosimilars is currently under debate in certain countries. In the EU, EMEA has clearly defined the process including overarching and product-specific guidelines, which includes clinical testing. Biosimilarity needs to be based on comparability criteria, including at least molecular characterization, biological activity relevant for the therapeutic effect and relative bioavailability (“bioequivalence”). In the case of such complex diseases as MS, where the effect of treatment is not so directly measurable, in vitro tools can provide additional data to support comparability. Genomic microarrays assays might be useful to compare multisource biopharmaceuticals. The aim of the present study was to compare the pharmacodynamic genomic effects (in terms of transcriptional regulation) of two recombinant human IFN-β1a preparations on lymphocytes of multiple sclerosis patients using a whole genome microarray assay. Methods: We performed an ex vivo whole genome expression profiling of the effect of two preparations of IFN-β1a on non-adherent mononuclears from five relapsing-remitting MS patients analyzing microarrays (CodeLink™ Human Whole Genome). Patients blood was drawn, PBMCs isolated and cultured in three different conditions: culture medium (control), 1,000 U/ml of IFN-β1a (BLA- (STOFERON™, Bio Sidus) and 1,000 U/ml of IFN-β1a (REBIF™, Serono) RNA was purified from non-adherent cells (mostly lymphocytes), amplified and hybridized. Raw data were generated by CodeLink™ proprietary software. Data normalization, quality control and analysis of differential gene expression between treatments were done using linear model for microarray data. Functional annotation analysis of IFN-β1a MS treatment transcription was done using DAVID. Results: Out of the approximately 45,000 human sequences examined, no evidence of differential regulation was found when both treatments were compared (minimum adjusted p-value > 0.999). The IFN-β1a effect differentially regulated the expression of 868 genes. The expression of standard markers such as GTP cyclohidrolase, MxA, and OAS isoenzymes A and B changed as a consequence of the action of IFN-β1a. Conclusions: This exhaustive and highly sensitive assay did not show differences in the genomic expression profile of these two products under the assayed experimental conditions. These results suggest that this technology might be useful for the initial comparison of biosimilars, being part of a comprehensive comparability program that includes clinical testing.