MxA mRNA decrease preceding NAb detection in IFNβ‐treated MS patients

Abstract Background Multiple sclerosis (MS) patients treated with interferon beta (IFNβ) are at risk of a declining response to treatment because of the production of IFNβ‐neutralizing antibodies (NAbs). The expression of Myxovirus resistance protein A (MxA) mRNA is regarded as a marker of IFNβ bioactivity. Aims The aim of this study was to analyze the kinetics of MxA mRNA expression during long‐term IFNβ treatment and assess its relationship to NAb production. Methods A prospective, observational, open‐label, non‐randomized study was designed in multiple sclerosis patients starting IFNβ treatment. NAbs and MxA mRNA were monitored every six months. Results 119 patients were consecutively enrolled and 107 were included in the final analysis. Both the presence of NAbs and a decrease in MxA mRNA below the cut off were revealed in 15 patients, however, in six patients (40%) positivity for NAbs was preceded by the decrease in MxA mRNA. In addition, a further six patients showing a decline in MxA mRNA did not have detectable NAbs. Conclusion Our data indicate that quantification of MxA mRNA is a more sensitive identifier of loss of IFNβ efficacy than the NAb positivity.


| INTRODUCTION
Interferon β (IFNβ) is one of the first-line treatments for patients with clinically isolated syndrome or relapsing remitting multiple sclerosis (MS). The responsiveness to this treatment may be lost because of the production of neutralizing antibodies (NAbs), which prevent the interaction between IFNβ and its receptor. Myxovirus resistance protein A (MxA) is a downstream gene product of IFNβ and is used as a marker of the biological activity of IFNβ. The level of MxA mRNA quantifies the IFNβ biological response (Bertolotto et al., 2001) and there is clear evidence that lack of MxA mRNA indicates a completely attenuated response to IFNβ (Hesse, Sellebjerg, & Sørensen, 2009).
The aim of our study was to assess the long-term kinetics of these IFNβ bioactivity markers (NAbs, MxA mRNA), specifically focusing on the temporal appearance of NAbs and the decline in MxA mRNA ( Figure 1).  (Bertolotto et al., 2001). Reactions were performed on Rotor-Gene (Corbett Research) and Rotor-Gene Q (Quiagen) in duplicate. We set up a comparative quantification assay (Livak & Schmittgen, 2001)  NAbs were determined using the antiviral cytopathic effect assay (Kawade, Finter, & Grossberg, 2003). Patients with positivity for NAbs and/or a decrease in MxA mRNA below the cut off were identified in the cohort and the time of appearance of both markers was compared.

| METHODS
The study protocol was approved by the Ethics Committee of Motol University Hospital and each participating patient gave informed consent. The cut off value for MxA mRNA was set to reflect an optimal reaction to IFNβ using the mean MxA value of control samples plus two standard deviations. An MxA mRNA value lower than this cut off in IFNβ-treated patients was interpreted as inefficacy of the IFNβ treatment.

| RESULTS
In 40% of patients (six out of 15) demonstrating both NAb positivity and an MxA mRNA decrease below the cut off, the decrease in MxA mRNA was detected either six or 12 months earlier than the NAb positivity. This was mainly observed in patients treated with IFNβ-1b s.c. (five patients), but also, in one patient treated with IFNβ-1a s.c.
( but no NAb positivity over the whole follow-up period. In two patients, NAbs and the decrease in MxA mRNA were not detected until month 24 ( Figure 1).

| DISCUSSION
Our results indicate that the decrease in MxA mRNA is a more reliable marker of IFNβ treatment efficacy than the NAb positivity alone.
The decline in MxA mRNA can precede the detection of NAbs by several months. Our observations support the hypothesis that once the patient has started IFNβ treatment, recurrently low MxA mRNA expression is predictive of NAb development. It is feasible that the early decrease in MxA mRNA could be caused by IFNβ binding antibodies (BAbs) that are detectable in the first months of IFNβ therapy.
In some cases, BAbs have been shown to cause a loss of bioactivity, but not neutralizing activity in the CPE assay, thus these antibodies are not classed as NAbs (Gilli et al., 2007). However, it has been proposed that BAb titres may be a predictive tool for future NAb development (Hegen et al., 2014).
Another explanation for the decrease in MxA mRNA preceding NAb detection could be insufficient sensitivity of the CPE assay.
Providing there is optimal methodological standardization, the CPE has a defined sensitivity limit of 5 TRU/ml and a titre of 20 TRU/ ml is considered positive for NAbs, although mostly with a low clinical significance (Kawade et al., 2003). However, the effect of NAbs on the bioactivity of IFNβ is variable (Bertolotto, 2015) and there is a theoretical possibility that the biological activity could also be reduced in patients with very low NAb titres (lower than 5 TRU/ml).
In several patients with a drop in MxA mRNA, NAbs were not detected at all during the follow-up period. It is possible that our follow-up period was too short to detect NAbs in all cases, which is supported by two cases not presenting with NAb positivity until month 24. Other explanations for non-antibody mediated abolished biological activity have been reported, for example, patient non-compliance and saturation of the IFN receptor (Gilli et al., 2007).
Nevertheless, in our opinion, it is the level of MxA mRNA expression and not NAb positivity that should be used as a primary tool for monitoring IFNβ treatment efficacy in MS patients. A lack of MxA mRNA expression strongly correlates with a loss of IFNβ therapy efficacy, regardless of NAb positivity. The current clinical recommendation is that IFNβ treatment is continued until NAbs are detected (Polman et al., 2010;Sørensen et al., 2005)

ACKNOWLEDGMENTS
The authors would like to thank Eva Hyncicova, Petra Liskova and Eva Houzvickova for their cooperation and care of the patients included in the study.

CONFLICTS OF INTEREST
Tomek