In this issue Antonio Bertolotto and colleagues (see pp 1294–9),¹ use MxA mRNA expression in peripheral blood mononuclear cells (PBMCs) to assess the in vivo bioactivity of interferon-beta (IFN-β) treatment in patients with multiple slerosis (MS). MxA transcription and translation is relatively specific for the type 1 interferons, such as interferon-alpha (IFN-α) and IFN-β, and plays an important role in the anti-viral response. MxA protein can also be used as readout for IFN-β but with a longer temporal profile.² MxA is therefore a suitable marker to assess the bioactivity of IFNβ—i.e. receptor binding, second messenger activation, gene transcription, and, in the case of MxA protein, gene translation. It appears, however, that quantitative MxA mRNA expression has some advantages over the MxA protein translation. Firstly, it has a wider dynamic range and the RT-qPCR method used to quantify its expression should be easier to standardise.³ Unfortunately, commercial antibody pairs suitable for an MxA protein enzyme linked immunosorbent assay (ELISA) are not currently available.

Not surprisingly, the majority of subjects (11/13) who did not have a biological response to IFN-β were persistently positive for neutralising anti-IFN- β antibodies (NAbs). Could in vivo MxA induction supplant NAb testing? Probably not, as this assay does not provide any information on the NAb titre, which is important in predicting whether or not NAbs are likely to persist. In general patients with a titre greater than approximately 100 NU are likely to remain NAb positive and those with lower titres are more likely to revert to being NAb negative.⁴ NAb titres are therefore likely to be incorporated into future management strategies of NAb positive patients. MxA induction, however, could be a useful initial screen to select patients for NAb testing. Although the overall efficacy of IFN-β in MS is relatively modest, the efficacy in subjects who remain NAb negative is substantially better than in those who become NAb positive. In the pivotal IFN-β-1b trial the reduction in relapse rate in the NAb negative patients was > 50%.⁵ This is significantly higher than the oft quoted reduction in relapse rate of 30% for the class of IFN-β preparations.

Interestingly, two patients in the Bertolotto and colleagues study did not have a biological response to IFN-β and did not have NAbs. This indicates that these subjects may be “true” non-responders and it would be interesting to know the mechanism of this lack of response—whether it is due to either a qualitative or quantitative biological trait. Could a qualitative trait prove useful—possibly as part of a battery of markers—to predict who is likely to respond to IFN-β treatment? Could a quantitative trait have the potential to be used to optimise IFN-β dosing in individual patients? Several international studies are currently addressing these questions.

As expected subjects receiving once weekly IFN-β-1a (Avonex®) had lower baseline induction of MxA mRNA compared with subjects receiving either IFN-β-1a thrice weekly (Rebif®), or IFN-β-1b (Betaferon®) every other day. Interestingly, in subjects receiving the more frequently administered IFN-β preparations 18%—or almost 1 in 5—injections failed to induce a biological response. This may relate to the unpredictable bioavailability of subcutaneous IFN-β or is more likely to be due to the lack of functional interferon receptors on circulating PBMCs.⁶ Once saturated and internalised it takes time for new functional receptors to be regenerated. These data imply that the current dosing regimens of IFN-β in MS have not been optimised—once weekly injections are not frequent enough, and thrice weekly or every other day injections are possibly too frequent. Twice weekly administration would seem the logical regimen for further investigation.

In conclusion, the therapeutic efficacy of IFN-β and more importantly its cost effectiveness could be rationalised. Future strategies include preventing or treating NAbs, selecting potential responders, not treating patients in whom IFN-β is not bioactive, and optimising IFN-β dosing for individual patients. The in vivo induction of MxA appears to be a promising biomarker in the pursuit of these aims.