Complementing the therapeutic armamentarium for Miller Fisher Syndrome and related immune neuropathies

In 1956 Miller Fisher described in his seminal paper three patients with an unusual variant of acute onset polyneuritis characterized by the clinical triad of ophthalmoplegia, ataxia and areflexia (Fisher, 1956). Although there was only minor limb involvement in his cases, Miller Fisher recognized that the syndrome he delineated has a significant overlap with other inflammatory neuropathies such as Guillain–Barré syndrome (GBS). Nowadays Miller Fisher syndrome (MFS) is widely regarded a part of the GBS spectrum (Hughes and Cornblath, 2005; Lo, 2007). MFS is commonly preceded by an infectious illness such as Campylobacter jejuni enteritis (Koga et al., 2005). Specific C. jejuni strains associated with MFS contain in their lipooligosaccharides ganglioside mimics that elicit an antibody response to shared neural epitopes (Yuki, 2007). In the vast majority of MFS patients, serum antibodies to the tetrasialoganglioside GQ1b can be detected (Chiba et al., 1993; Willison and Yuki, 2002). These antibodies are usually IgG antibodies of complement fixing isotypes. Furthermore, anti-GQ1b antibodies can bind to the nodes of Ranvier and to the presynaptic terminal of neuromuscular junctions in extraocular and somatic muscles. Immunolocalization studies demonstrated that GQ1b is much more widely expressed in the oculomotor cranial nerves than in other cranial and peripheral nerves. It is also expressed on sensory neurons. This differential distribution may account for part of the clinical spectrum typically seen in Miller Fisher syndrome (Chiba et al., 1997). It is worth noticing that antibodies of other related specificities, such as for gangliosides GT1a, GD1b and GD3, and more recently for ganglioside complexes containing GQ1b have also been identified in MFS, and for that matter, GBS (Kaida et al., 2006; Nagashima et al., 2007; Yuki, 2007; Kanzaki et al., 2008). Insight into the molecular targets of an aberrant immune response in MFS and related neuropathies has advanced significantly, even to the atomic level (Houliston et al., 2007) but this knowledge, unfortunately, is not paralleled by the availability of an equally sophisticated therapeutic armamentarium. While data from randomized controlled trials in MFS are lacking, treatment of this GBS variant follows the evidence-based guidelines established for GBS (Overell et al., 2007): intravenous immunoglobulins (IVIg) or plasma exchange hasten recovery from this almost invariably monophasic disease (Hughes and Cornblath, 2005; Lehmann et al., 2006). Notably, however, no therapeutic advance for this group of acute inflammatory neuropathies has been achieved since 1992 when van der Meche and his colleagues reported positive outcomes of the IVIG trial in GBS (van der Meche and Schmitz, 1992). This was later confirmed by the largest ever conducted trial in GBS, the PSGBS trial that also established equal efficacy of plasma exchange and IVIG (Plasma Exchange/Sandoglobulin Guillain–Barré Syndrome Trial Group, 1997). With these treatments the majority of patients have a favourable outcome, however, a considerable proportion of patients show a poor recovery, especially if severe axonal damage occurs during the disease course (Hughes and Cornblath, 2005). This emphasizes the need for more efficacious treatments. Over the last decade Professor Willison's laboratory in Glasgow, in close cooperation with the Leiden group headed by Dr Plomp, has extensively studied the pathological effects of human and murine anti-GQ1b antibodies on the neuromuscular junction (Goodyear et al., 1999; Jacobs et al., 2003; Halstead et al., 2005; O'Hanlon et al., 2001). By demonstrating that these antibodies impair neuromuscular transmission and induce degeneration of motor axons and perisynaptic Schwann cells through complement-dependent cytotoxicity they provided important mechanistic insights into the pathogenesis of this disorder. In the current issue Halstead and colleagues have expanded previous work and now present the experimental proof for a new therapeutic approach in antibody-mediated neuropathies (Halstead et al., 2008). Using an in vitro model of mouse hemi-diaphragm phrenic nerve preparations exposed to an IgM anti-GQ1b antibody they show that eculizumab, a humanized monoclonal antibody which prevents the formation of the terminal membrane attack complex C5b-9 (Hillmen et al., 2006; Mollnes and Kirschfink, 2006), protects from anti-GQ1b antibody-mediated, complement-dependent injury at the neuromuscular junction. The study is even more valuable: Halstead and colleagues established a new in vivo model, which is based on complement-dependent pathogenic effects of anti-GQ1b antibodies injected intraperitoneally into Balb/c mice. Respiratory failure can be assessed by whole-body plethysmography and neuromuscular transmission analysed by intracellularly recording miniature endplate potentials. This experimental paradigm shares some clinical features with MFS. In this model eculizumab shields mice from respiratory paralysis and terminal motor neuropathy as evidenced by functional and histological assessments. Translating these results into a clinical perspective requires further consideration of the role of complement in inflammatory neuropathies as well as the therapeutic potential of eculizumab. During inflammation different pathways may activate the complement system, although the classical pathway usually requires the presence of complement-fixing antibodies. Regardless of the mode of activation and pathway initially triggered, the cleavage of C3 in C3a and C3b and the subsequent split of C5 into C5a and C5b leads to the generation of proinflammatory peptides and the assembly of the cytotoxic pore forming membrane attack complex (MAC), consisting of C5b, C6, C7, C8 and C9 (Markiewski and Lambris, 2007). Targeting the complement system by specific antibodies has been shown to be an effective treatment in animal models of autoimmune diseases (Mollnes and Kirschfink, 2006). For example, the systemic administration of a murine anti-C5 antibody prevents inflammation in an animal model of rheumatoid arthritis (Wang et al., 1995). Complement activation is also thought to be a crucial step in the pathogenesis of all variants of GBS and therefore represents an appealing candidate for selective immunotherapy. Post-mortem studies show that in axonal and demyelinating variants of GBS deposits of IgG, complement activation product C3d and MAC are prominent pathological features (Koski et al., 1987; Hafer-Macko et al., 1996a, b). In an animal model of AMAN, the acute motor axonal variant of GBS, anti-GM1 antibodies were demonstrated to engage the complement system disrupting sodium channel clusters in peripheral motor nerve fibers (Susuki et al., 2007). Moreover, the complement factors C3a and C5a are elevated in the cerebrospinal fluid in GBS (Hartung et al., 1987). Consequently the therapeutic potential of complement depletion was assessed in experimental models of GBS, such as the experimental allergic neuritis. A major drawback of this animal model is that it replicates a predominantly cellular-mediated autoimmune response and may therefore not be ideally suited to study potential pathogenic effects of autoantibodies and complement. Nonetheless, in these studies, the non-specific depletion of complement attenuated the disease course or even completely suppressed disease (Feasby et al., 1987; Vriesendorp et al., 1995). Using a recombinant soluble complement receptor type 1 (sCR1), Jung and colleagues were similarly able to diminish disease severity in EAN (Jung et al., 1995). Although these data strongly supported the hypothesis that complement inhibition can effectively prevent peripheral nerve injury in vivo, the therapeutic potential of complement inhibition has not been followed systematically over the last years due to the unavailability of (i) animal models in which pathology is purely mediated by complement dependent effects of specific autoantibodies and (ii) potential drugs that allow to specifically intervene in the complement cascade to prevent MAC activation without causing major side effects. Fortunately, recent progress in the development of therapeutic monoclonal antibodies has provided tools to overcome the latter obstacle (Ricklin and Lambris, 2007). In 2007 the United States Food and Drug Administration approved eculizumab for the treatment of paroxysmal nocturnal haemoglobinuria. Eculizumab is a humanized monoclonal antibody directed against the C5 convertase enzyme. Binding of eculizumab reduces the esterase activity. Consequently, hydrolysis of C5 into the C5b subunit and C5a is diminished and formation of MAC is curtailed. The pivotal trial of eculizumab in paroxysmal nocturnal haemoglobinuria demonstrated that intravenous administration of 600–900 mg eculizumab at weekly intervals appears to be safe and is associated with only minor adverse effect (headache, nasopharyngitis, back pain, nausea) (Hillmen et al., 2006). The data presented in the paper from Halstead and colleagues provide strong evidence that eculizumab works effectively in an in vivo model of anti-ganglioside-mediated nerve injury. Since antibodies against gangliosides GM1 and GD1a have also been shown to exert complement dependent pathogenic effects in different experimental paradigms (Zhang et al., 2004; Goodfellow et al., 2005; Susuki et al., 2007) it seems indeed worthwhile to consider a clinical trial of eculizumab in MFS and other autoantibody mediated neuropathies. In this context it needs to be mentioned that established treatment options such as plasma exchange and IVIg have also been demonstrated to target the complement system, amongst protean other pathogenetically relevant molecular pathways, by reducing serum complement levels and preventing complement activation. It would, however, appear timely now to embark on seeking proof for the hypothesis that, given the key importance of complement-mediated nerve injury, selective blockade of complement activation via eculizumab, other C5- or C5a-directed monoclonal antibodies, small molecule C5a antagonists (Allegretti et al., 2005; Mollnes and Kirschfink, 2006), or targeted regulators such as APT070 or Crry-Ig (Atkinson et al., 2008; Halstead et al., 2005) may yield greater clinical benefit. Clearly, with the current data provided by Halstead and colleagues and approval of the drug in another complement-dependent autoimmune disease, it seems prudent to start with eculizumab in MFS and GBS. In view of its high costs, the presumably greater pathogenetic heterogeneity and more protracted temporal evolution of immunological changes (Kieseier et al., 2006), other novel immunointerventional strategies are likely to be preferred as attractive candidate agents for evaluation in chronic B cell mediated neuropathies (Meyer zu Hörste et al., 2007). At any rate, intensified efforts are warranted to establish improved treatments for MFS, GBS and other disabling immune-mediated neuropathies. The foundations are laid by the work from the Willison group.