Over a dozen disease-modifying therapies (DMT) are now available to treat relapsing forms of multiple sclerosis (MS), providing clinicians with a wealth of choices for initiating and maintaining therapy. But what is the rational approach to selecting the most appropriate therapy in an individual patient throughout the treatment course?
DMTs modulate the dysfunctional immune response in RRMS via different mechanisms, with T cell function as the primary target. Several lines of evidence support the central role of T cells in the pathophysiology of RRMS. The earliest was the observation of inflammatory infiltrates in demyelinating lesions as far back as a century ago, which were later determined to be largely composed of T cells and macrophages (reviewed in Rae-Grant et al. Mult Scler Relat Disord 2014;3:156-162). Adoptive transfer experiments in experimental autoimmune encephalomyelitis (EAE) have also shown that CD4+ cells alone are sufficient to induce autoimmunity (Pettinelli & McFarlin. J Immunol 1981;127:1420-1423); whereas T cell removal will prevent the development of EAE (Bernard et al. Eur J Immunol 1976;6:655-660).
B cells have always been implicated due to the frequent finding of oligoclonal bands in cerebrospinal fluid; antibodies were subsequently detected in MS lesions (Obermeier et al. J Neuroimmunol 2011;233:245-8). A number of antibody targets, including MBP, MOG and other CNS components, have now been identified (reviewed in Fraussen et al. Autoimmun Rev 2014;13:1126–37).
In recent years the relative importance of antibody production in the pathogenesis of MS has shifted toward a greater emphasis on B cells as antigen-presenting cells, and on their interactions with T cells. This is in part due to the observation that rituximab depletes B and T cells in CSF without altering total Ig levels (Cross et al. J Neuroimmunol 2006;180:63-70). B cell antigen presentation alone was not believed to be sufficient to initiate or sustain disease (Archambault et al. J Immunol 2013; 191:545-50). But more recent studies have reported that B cells, acting with dendritic cells, can regulate T cell autoreactivity in the CNS, thereby increasing disease severity (Harp et al. J Immunol 2015;194:5077-84). It has also been suggested that B cells modulate T cell function by providing costimulatory molecules (e.g. CD40) required for T cell activation; and through the production of pro-inflammatory cytokines, such as tumour necrosis factor (TNF)-alpha and lymphotoxin (LT)-alpha. Levels of these cytokines are elevated in MS lesions and have been shown in vitro to exacerbate demyelination (Plant et al. Glia 2005;49:1-14).
A number of B-cell-directed therapies (e.g. ocrelizumab, ofatumumab) are now in development. Ocrelizumab is a humanized monoclonal antibody infusion agent that targets CD20+ B cells; it increases antibody-dependent cell-mediated cytotoxicity (ADCC) and reduces complement-dependent cytotoxicity (CDC) relative to rituximab (Lycke J. Ther Adv Neurol Disord 2015;8:274-293), which may lower the risk of serious infections seen with the older agent. (In comparison, alemtuzumab acts via both ADCC and CDC, although ADCC is its primary mechanism.) Ofatumumab binds a CD20 epitope distinct from other B cell-directed therapies (Zhang B. MAbs 2009;1:326-331); it is a fully human MAb administered by subcutaneous injection, which reduces the risk of systemic reactions.
However, it should be noted that “B cell-directed” is a bit of a misnomer since part of the efficacy of most of the current DMTs has been attributed to their effects on B cell populations. For example, interferon-beta decreases the number of pathogenic memory B cells (Dooley et al. Neurol Neuroimmunol Neuroinflamm 2016;3:e240; Rizzo et al. Immunol Cell Biol 2016; epublished July, 2016). Teriflunomide inhibits the mitochondrial enzyme dihydro-orotate dehydrogenase, which reduces the proliferation of activated T and B cells (Bar-Or et al. Drugs 2014;74:659-74). Dimethyl fumarate has variable effects on B cell subsets at different time points in the treatment course (Lundy et al. Neurol Neuroimmunol Neuroinflamm 2016;3:e211). Fingolimod reduces the population of pathogenic memory B cells, increases the proportion of regulatory B cell subsets and promotes the production of anti-inflammatory cytokines (e.g. IL-10) by B cells (Claes et al. PLoS One 2014;9:e111115; Blumenfeld et al. J Autoimmun 2016;70:40-51). Natalizumab blocks autoreactive B cells from entering the CNS, thereby reducing the number of B cells available for antigen presentation. Alemtuzumab depletes both T and B cells through its mechanism of action of targeting CD52 on T and B lymphocytes (Havrdova et al. Ther Adv Neurol Disord 2015;8:31-45). Cladribine reduces the number of active and resting T and B lymphocytes by inhibiting DNA synthesis and promoting apoptosis (Leist & Weissert. Clin Neuropharmacol 2011;34:28-35).
Conversely, B-cell therapies have also been shown to reduce T cell proliferation and inflammatory cytokine production by Th17 cells (Monson et al. PLoS One 2011;6:e17103). Indeed, it is likely that the beneficial effects of most immunomodulatory therapies, with the possible exception of glatiramer acetate and daclizumab (Bielekova et al. Proc Natl Acad Sci USA 2006;103:5941-5946), are attributable to their effects on both T and B cells. Thus, a recent suggestion has been to view therapies not as a means of selectively targeting immune components, but rather as different methods of shaping the immune response to a less inflammatory profile either by a non-selective immune modulation (e.g. interferons), or by sequestering (e.g. fingolimod, natalizumab) or depleting (e.g. alemtuzumab, cladribine) immune populations (Bittner & Wiendl. Neurotherapeutics 2016;13:4-19).
The effect of a given therapy on absolute lymphocyte counts is probably less important – both in reference to its safety as well as efficacy – than its effect on T and B cell subsets. These effects, which have not been well characterized, would likely be a better measure of therapeutic benefit and risk than the relative potency of agents as reported in clinical trials. In addition, the site of action may be clinically relevant: most DMTs act in the peripheral circulation whereas most immune cells are found in tissue. Small-molecule agents may be more likely to cross the blood-brain barrier and interact with immune cells within the CNS. These issues will be addressed in more detail later in the series.
However, there are a number of limitations to using a drug’s mechanism of action as a guide to treatment selection. A targeted therapy requires a known target, and the immune target(s) in MS is largely unknown. Moreover, modulating the immune response to a less inflammatory profile would be a more achievable goal if the optimal balance of immune cell subsets (e.g. CD4+/CD8+ ratio, Tregs) were established. Indeed, it is likely that immune dysregulation in MS is as individual as a person’s “immune biography” (Krone et al. J Neurol 2009;256:1052-60). However, the lack of established biomarkers of disease activity and progression has thus far barred entry into the era of truly personalized medicine.
MS therapeutics has been likened to that of rheumatology, but a more apt simile might be infectious disease. The approach to treatment is empiric; if a given agent, regardless of its mode of action, is efficacious, there is no need to identify the underlying cause. Similarly, an individual agent’s mechanism of action does not provide a clear guide on the optimal DMT with which to initiate treatment in a given patient. Other, more pragmatic factors need to be considered for treatment selection. These factors will be discussed in Part 2 of our series.
Dr. Mark Freedman: If only one could know what immune mechanism might be operating to cause damage in anyone’s CNS at any time in a patient with MS, it might be possible to specifically target that mechanism to maximize effective therapy while not compromising other immune mechanisms that serve to protect or repair the body. Unfortunately, that is not possible. In fact, not only may different damaging mechanisms be operating in different patients, they can change within the same patient. That may explain at least some of the variability in treatment responses observed in MS populations, and possibly why some patients’ disease stops responding well to one agent but does respond to another.
Despite the approach to MS treatment being somewhat empiric, knowledge of the various purported mechanisms of action of the different DMTs may aid in choosing a medication for a patient, keeping in mind: the level of disease one perceives to be present; the risk of early disease progression due to unchecked inflammatory activity; previous response to medication; and co-morbidities, which may increase safety risks. By carefully considering the different mechanisms of DMTs, it may be possible to use them in a logical sequence that complements their capabilities of controlling immune responses. Similarly, there may be greater safety risks if DMTs are used in a different sequence.
It is important to consider what one is trying to accomplish with a therapy. Immunomodulation implies the ability to alter damaging immune responses by dissuading cells from attacking the CNS either by getting them to alter their cytokine release pattern (e.g. interferon-beta) or distracting from their CNS target (e.g. glatiramer acetate). A DMT that interrupts cell trafficking to the CNS (e.g. natalizumab or fingolimod) may offer immediate disease control, but will eventually compromise the immune system by impeding immunosurveillance. Cell -depleting therapies are non-specific, and though rapidly dividing autoimmune cells are more likely to be sensitive to the effects of these agents (presumably the ones causing the disease), normal immune functions are also affected. The ultimate therapy for patients may well be a combination of certain DMTs, but current pharmacoeconomics precludes that possibility. Therefore, we must rely on careful consideration of treatment sequencing to optimize treatment.
Back to top
NeuroSens survey on sequencing – Part 1This poll is currently disabled.