September 30, 2003
Mark S. Freedman, MD, MSc, Professor of Medicine (Neurology), University of Ottawa, and Director, Multiple Sclerosis Research Clinic, The Ottawa Hospital, Ottawa, Canada
Medscape Neurology & Neurosurgery 5(2), 2003.
Both interferon-beta (IFN-beta) and glatiramer acetate have been established as effective disease-modifying agents for the treatment of patients with multiple sclerosis (MS). Both of these medications currently require continued administration by injection in order to maintain their modest effect at reducing disease activity and slowing disease progression. As part of the body's normal defense, the immune system effectively reacts to foreign agents, and particularly to agents introduced into the body via injection. Thus, in response to the injection of protein-based products such as IFN-beta or glatiramer acetate, the immune system typically reacts by producing antibodies.
Binding Abs and Neutralizing Abs
Although IFN-beta is manufactured by gene-recombinant technology that uses the human gene, the process itself employs either bacterial (Escherichia coli in the case of IFN-beta 1b) or mammalian (Chinese hamster ovarian cells in the case of both IFN-beta 1a preparations) cell lines to manufacture large quantities of drug. Thus, these cells are not human, and sufficient (if slight) differences remain and can induce the formation of Ab in some people. Although the protein sequence may be identical to (in the case of IFN-beta 1a) or only slightly different from (in the case of IFN-beta 1b) native IFN-beta, the way that sugars are added (IFN-beta 1a) or not (IFN-beta 1b) in the final processing of the protein may be important indicators to the immune system that the agent is indeed "foreign." In the case of glatiramer acetate, the random number and size of the polymers produced lead to a fairly widespread immune response that might in fact be part of its mechanism of action. If the immune system reacts to these injected agents, the response might well be to generate antibodies.
Antibodies can simply bind to IFN-beta or glatiramer acetate (binding Ab, or BAb) with no subsequent effect on function, or they can block or neutralize (neutralizing Ab, or NAb) their biological activity. BAbs can bind to the agent but do not sterically hinder its biological action; in some cases BAbs might even extend the half-life of the agent, by protecting it from elimination. NAbs will partially or completely abrogate the biological action of an agent by either blocking an important surface molecule needed for its activity or by interfering with the binding of the agent to its receptor on a target cell. The type of Ab and its affinity (stickiness) to either agent are important areas, but beyond the scope of discussion here.
The development of BAbs or NAbs depends on the dose, frequency, and mode of injection or route of entry (skin > muscle > intravenous > oral) of an agent. A low dose given less frequently will induce very few Abs, but higher doses given more often tend to induce greater Ab formation. Interestingly, if too little or too much of a given agent is administered, the production of Ab is actually inhibited -- a phenomenon known as low- or high-zone tolerance.
The skin is the best site for stimulating Ab formation; hence, vaccines are often administered subcutaneously. Not surprisingly, both IFN-betas that are injected subcutaneously (Betaseron, Rebif) induce greater Ab formation compared with the formulation that is injected into muscle (Avonex). The big question in the case of MS agents is whether the formation of NAbs compromises the efficacy of any of the treatments. Let's look at the information that is available today to help answer this question.
Before discussing the clinical data, several key issues concerning the detection of Abs should be clarified. The definitions of BAb and NAb have been derived from in vitro assays. Commonly, BAbs are found with a capture assay that simply binds the molecule, if it can be found in serum, and identifies it. The simplest and most widely used method is the ELISA together with a "capture Ab." Such an assay detects only the presence of an Ab, and does not provide any information about its biological activity.
Whether or not the identified Ab is biologically active, which in this case means "neutralizing," is an area of great controversy: the assays rely on examining known biological actions of IFN-beta, even though no single action has ever been clearly linked to its beneficial effect on MS. One of the 2 most popular assays looks for the induction of a protein (eg, Mx protein) known to respond stoichiometrically to IFN-beta; the other, a cytopathic effect (CPE)assay, detects a basic function of IFN-beta, that of protecting cells from being lysed by a set amount of introduced virus. These assays are sensitive to even very low titers of Abs.
In Vivo Effects of Antibodies
Two approaches have been used to elucidate the in vivo effects of Abs: the first is to assess some of the biological response markers known to be affected by IFN-beta; the second is to examine the results from clinical trials. In the first approach, scientists measure certain biological response markers in the serum of patients receiving an IFN-beta before and after starting the agent, or compared with a parallel group that is using only placebo. A normal response would show elevations in the amount of the biological response marker, which would remain elevated as long as the agent is being administered. The second method examines MS-related outcomes in placebo-controlled or comparative clinical studies. In both cases, it is possible to determine whether the presence of NAb is at all linked to either smaller increases in levels of a biological response marker, or a poorer clinical outcome.
Not surprisingly, the clinical trials have yielded some interesting but controversial results. In the case of IFN-beta, although up to almost 80% of treated patients became BAb positive over time, NAbs were found in only 2% to 47% of patients. This implies that the majority of Abs to IFN-beta do not interfere with the biological action of the drug. The presence of BAbs and particularly NAbs was time dependent, with most positivity occurring after the first 6 months of treatment and fading with time. The lower, less frequent dose of intramuscular IFN-beta was associated with the lowest number of NAb-positive patients. Higher, more frequent doses of the subcutaneously administered IFN-betas were associated with a greater proportion of NAb-positive patients; however, in comparative studies, both higher-dose agents were shown to provide superior efficacy compared with the lower-dose formulation. In addition, increasing the dosage of subcutaneous IFN-beta 1a led to a reduction of NAb positivity, explained perhaps by the phenomenon of "high-zone tolerance." In studies examining biological response markers, the group of patients with particularly high NAb titers experienced changes in biological response markers that were no different from those of the placebo group, suggesting there was no longer any measurable (by biological response marker) effect of IFN-beta.
As a group, especially using an interval analysis approach that examines periods when patients are NAb positive or negative, the highest and most persistently positive titers of NAb were associated with a greater number of relapses, more progression, or a higher degree of MRI activity in patients with MS. However, many individual patients with high NAb titers had no measurable clinical activity, and others with undetectable NAbs experienced clinical deterioration. Thus, it was only possible to determine the effect of NAbs using the group data.
It would appear just as difficult to determine the negative effect of a NAb as it is to measure the benefit of any single treatment in an individual patient with MS. A final and curious finding is that in all studies to date, the individuals who could be identified as those who would eventually harbor NAbs achieved a greater reduction in relapse rates during the first year compared with those who would never develop a NAb titer of significance. Some interesting theories regarding this latter phenomenon suggest that these patients may actually have a different and perhaps less biologically potent native IFN-beta than NAb negative patients. The introduction of a highly potent recombinant product would therefore generate a greater clinical response when first injected, but would also be regarded as more "foreign" by the host's immune system, leading to a greater NAb response (see Opdenakker review). It is possible that such patients may be naturally predisposed to greater disease, explaining the association of NAb positivity with negative clinical outcomes. Such a theory warrants further investigation.
At this time, it is not possible to determine which patients will develop the more significant NAbs when they begin therapy, but scientists are searching for the clues that will identify such individuals. The incidence of NAbs is greatest (30% to 40%) for subcutaneous IFN-beta 1b (the most "foreign" of the IFN-betas), second highest for subcutaneous IFN-beta 1a (12% to 14%), and lowest for intramuscular IFN-beta 1a (2% to 5%). Although both IFN-beta 1a molecules are identical, the difference in NAb rates is probably due to the higher dose and frequency of administration of the subcutaneous formulation. However, in direct comparative studies, both subcutaneous formulations demonstrated superior clinical efficacy compared with the intramuscular formulation, an overall treatment effect that stood out even if you compared results for the former groups that were NAb positive with those in the latter group who were NAb negative. In addition, the majority of patients do not develop persistent NAbs and the clinical benefits of IFN-beta in these NAb individuals are even greater than those found for the group as a whole. It is therefore not at all clear that we should consider sacrificing the greater efficacy of the subcutaneous formulations in favor of intramuscular IFN-beta in order to minimize the possible development of NAbs. It would also not make sense to switch from a subcutaneous to the intramuscular formulation in case NAbs do develop, because NAbs cross-react and could just as well negatively affect treatment with the lower-dose formulation.
Glatiramer acetate treatment has been associated with nearly 100% BAb, and only recently has been associated with what appear to be NAbs in some 47% of treated patients. Unlike the studies of IFN-beta, NAbs associated with glatiramer acetate were determined on the basis of their ability to "neutralize" some of the known in vitro immunologic effects of glatiramer acetate that are presumably linked to their proposed mechanism of action in producing the clinical benefit in patients with MS.
No consensus has yet been reached regarding which Ab tests are the most
reliable, what titer is most important, when in the course of treatment
measurements should be taken, and most importantly, what, if anything,
should be done for patients who test positive. In addition, because NAbs
are not always associated with a negative clinical outcome, vary in titer,
and more than 50% will disappear completely within a few months to years
with continued IFN-beta treatment, it is not possible to understand the
true significance of being NAb positive. It is therefore not surprising
that experts worldwide have recommended that the response to IFN-beta or
glatiramer acetate treatment be judged clinically, and that decisions on
whether to continue on a given product not be based on the result of a
controversial NAb test.
Copyright © 2003, Medscape