Evaluation of the impact of neutralizing antibodies on IFNβ response

Highlights

• NAbs quantification improves management of IFNβ treated patients.

• High titer Nab-positive patients should change treatment.

• MxA and NAbs quantification identify a subset of IFNβ non-responders.

Abstract

IFNβ therapeutic action depends on a sequence of biological steps: i) the interaction between interferon beta (IFNβ) and its receptor (IFNAR) located at the cell surface of peripheral blood mononuclear cells; ii) activation of second messengers; iii) transcription of several genes containing specific ISRE regions (Interferon Stimulated Response Elements); and iv) synthesis of specific proteins.

Although IFNβ therapy has improved treatment options of patients with multiple sclerosis (MS), the long-term efficacy of IFNβs can be compromised due to the development of neutralizing antibodies (NAbs).

High titer NAbs develop in about 15% of patients; they abolish IFNβ biological activity and consequently the therapeutic action of IFNβ. Different IFNβ preparations carry different risks of developing NAbs, ranging from 3 to 28%. The risk of inducing NAbs must be considered in the selection of treatment.

Guidelines for NAbs testing and the therapeutic decision in case of NAbs positivity have been established. NAbs positivity predicts MRI and clinical activity. Precocious identification of Nabs-positive patients and switch to alternative treatments can improve the percentage of responders and allow a better allocation of relevant economical resources.

Introduction

All bio-pharmaceutics are potentially able to induce an immunogenic response characterized by anti-drug antibodies (ADA); examples are insulin, growth factors such as erythropoietin and granulocyte-macrophage colony-stimulating factor, monoclonal antibodies such as natalizumab, infliximab and adalimumab, botulin toxin, interferons alfa and beta, enzymes for replacement therapy such as glucosidase alfa.

Not all factors contributing to immunogenicity are known or can be controlled; they are both product-specific and host-specific. The products can present species-specific epitopes, form aggregates, contain impurities and contaminants, be modified by oxidation, deamidation or glycosylation; in addition different pH and excipients can influence drug immunogenicity. Host-specific factors include dose, frequency and duration of treatment, route of administration, immune-competence and immune-tolerance and genetic background. ADA may have clinical effects, but the relationship between ADA and clinical effects is not always direct and quantitative and possible scenarios range from no demonstrable effect to severe adverse events (SAE). ADA can alter the pharmaco-kinetics either increasing or decreasing the exposure to the drug; the immune-complex formed by ADA and drug can decrease the efficacy by reducing the half-life of the drug or directly neutralizing the drug; ADA can also trigger adverse drug reactions. When the drug is identical or very similar to an endogenous protein, ADA can cross-react with it leading to significant loss-of-function and SAE.

IFNβ became available for multiple sclerosis (MS) patients in 1993, as the first drug able to modify the natural history of MS. The importance of that event was clearly underlined in an editorial in Neurology by B.G. Arnason “The natural history of MS has been altered favorably, substantially and, above all, safely. Whether it is also the beginning of the end, time alone will tell. This is, I believe, the end of the beginning” [1].

The biochemical characteristics of IFNβ influence its clinical efficacy, its adverse events and the risk of losing therapeutic action. IFNβ is a recombinant cytokine available in 3 formulations, two as IFNβ-1a and one as IFNβ-1b that is produced in Escherichia coli and it differs from the natural human product by methionin-1 deletion, cystein-17 to serine mutation, and lack of glycosylation. There is a 10-fold increase in weight of protein present in a single IFNβ-1b dose compared with the IFNβ-1a versions in order to reach a suitable specific activity level. This is likely to lead to increased aggregation [2]), which may enhance its antigenicity. The IFNβ-1a preparations, in contrast, have primary and secondary structure identical to the native form, they are produced in mammalian cells, and the molecules are glycosylated, but not necessarily in the same way as human natural IFNβb.

IFNβ uses the same metabolic pathways as natural IFNβ, binding the receptor IFNAR that is shared with IFNα. IFNβ-1a and -1b bind to the same two-subunit cell-surface receptor IFNAR and both activate the same Janus kinase/signal transducer and activator of transcription (Jak/STAT) signaling pathway [3]. The stimulation of the receptor results in induction or reduction of expression of a large number of genes [4]. Those changes constitute the biological activity of IFNβ.

The biological activity of IFNβ can therefore be studied by measuring a number of IFNβ-induced gene products including Myxovirus resistance protein A (MxA) [5], β2-microglobulin [6] and neopterin [7]. In order to clearly reflect a response to IFNβ the chosen biomarker needs to be specific for IFNβ and the induction needs to be of a certain magnitude. Amongst all tested IFNβ-induced genes, MxA has proven to be one of the most reliable markers of the in vivo bioactivity of IFNβ [8], [9], [10].