Combining capture ELISA and in vivo MxA test to detect neutralizing antibodies to interferon-beta in a large cohort of multiple sclerosis patients

Objective: To investigate the use of a capture ELISA and an in vivo MxA assay to determine the occurrence of antibodies and biological responses in a large number of multiple sclerosis (MS) patients treated with interferon-beta.

Background: Beta-interferons (IFN-beta) are widely used in the treatment of MS. A significant number of patients develop neutralizing antibodies (NAbs) to the drug, which may eventually interfere with treatment efficacy. However, therapy fails in a significant number of patients, who do not develop NAbs.

Design/methods: More than 1500 MS patients on interferon-beta therapy were included in the study. Serum and RNA were obtained from all patients 12 hours after injection of IFN-beta. Antibodies to IFN-beta were quantified in serum by capture ELISA and the expression of the MxA and GAPDH by real time PCR. Standard samples were used in all experiments to allow direct comparison of antibody titers and MxA induction among different experiments.

Results: We detected significant titers of binding antibodies to IFN-beta in more than 30% of all patients. The MxA response was reduced in 30% of patients. We found a strong correlation between the level of binding antibody titers and the reduction of the MxA response in vivo. While the majority of patients, who showed a reduced MxA response had binding antibodies, a subgroup of patients had impaired MxA responses in the absence of antibodies. Antibodies with significant reduction of the MxA response persisted in more than 80% of patients when retested 3 months later. The best predictor for a persistent NAb response was the extent of MxA reduction in vivo at the first time point.

Conclusions/relevance: The combination of capture ELISA and MxA allows identifying patients with reduced IFN-beta response in vivo. In the majority of patients reduced MxA responses are due to the development of antibodies to IFN-beta. Strong reduction of the MxA response in vivo was a good predictor of persistent NAbs. However, in a subgroup of patients the reduced MxA response might result from non compliance or genetic factors that interfere with the biological activity of IFN-beta. This subgroup of patients may also not fully benefit from IFN-beta therapy.