Immunoadsorption therapy in patients with multiple sclerosis with steroid-refractory optical neuritis

Previous studies have described four immunopathological patterns of demyelination in early multiple sclerosis (MS) lesions, with pattern II being characterized by antibody and/or complement-associated demyelination [1]. Several specific antibodies have been described and discussed to contribute to the humoral autoimmune response in MS [2]. Immunoglobulins are synthesized intrathecally; however, at least part of the humoral response in MS is derived systemically from the blood [3].

Therapeutic plasma exchange (PE) is based on the separation of plasma from cellular blood components, allowing the removal of substances up to a molecular weight of 3 × 10³ kDa. As shown in a randomized placebo controlled cross-over study, PE was efficient for steroid - refractory relapses in about 40% to 50% of cases of acute central nervous system inflammatory demyelinating diseases [4]. The usefulness of PE has also been extended to severe optic neuritis in patients with MS [5, 6]. Thus, the use of PE in steroid-refractory relapses has become an integral part of European guidelines for the treatment of MS [7]. Clinical-pathological correlation analyses have shown that all patients with pattern II pathology but none with pattern I or pattern III experienced improvement in neurological deficits after being treated with PE [8]. This selective response suggests a removal of pathogenic humoral and plasma factors by PE.

Immunoadsorption (IA) provides a more selective approach and the potential for technical innovations in therapeutic apheresis techniques, allowing the elimination of pathogenic antibodies while sparing other plasma proteins. With IA, relevant side effects of PE resulting from protein substitution can be avoided [9]. We hypothesized that IA is at least equally efficient compared to PE as an escalation therapy for steroid-unresponsive relapses of MS. Therefore, we performed a prospective trial to compare IA treatment in 11 patients with MS with our earlier patient population treated with PE for their MS [10]. In addition, proteomic analyses of column-bound proteins were performed as well as measurements of corresponding changes in patients’ plasma samples.

IA treatments were started after a mean time of 26.6 days after the initial symptoms and 10.8 days after the start of corticosteroid therapy. Each patient received at least two courses of corticosteroid therapy prior to IA therapy, with a mean cumulative dose of 10.9 g prednisolone equivalent but without a significant improvement, thus fulfilling the indication of therapeutic apheresis as adjunct treatment. An improvement of visual acuity up to 0.6 or more was achieved in eight of eleven patients (72.7%) undergoing IA treatment (responder group). Two patients did not respond to therapy at any time point and one patient improved during IA therapy but deteriorated shortly after the end of IA therapy, associated with the incidence of a jugular venous thrombosis on the same side (non-responder group). The relevant data of patients’ characteristics are shown in Table 1.

PE is an efficient treatment in acute central nervous system inflammatory demyelinating diseases [4], including severe optic neuritis, motor impairment or ataxia [5, 6] after steroid-refractory relapses, successful in about 40% to 50% of cases. Here, we report on the first prospective investigation of tryptophan-IA in 11 relapsing MS patients with optic neuritis refractory to corticosteroid pulses in an open prospective study. Overall, eight of eleven patients (72.7%) achieved a remission. One patient gradually improved but deteriorated again along with the development of jugular venous thrombosis, and two patients did not respond at all. The response to IA seems to be comparable to the best results achieved in PE series [5, 6] and to two very recent retrospective analyses of the effect of IA in steroid-refractory MS cases [18, 19]. As in our own PE study [10], significant clinical improvement was seen after the third extracorporeal treatment session with a trend in favor of early IA initiation compared to delayed IA initiation. Our treatment protocol was limited to a total of five IA sessions. Additional experience showed that increasing the number of apheresis sessions did not correlate with further improvement of outcome (data not shown). This observation is in accord with results of IA in acute autoimmune neuropathies like Guillain-Barré Syndrome [20].

Most side effects were typical of any apheresis procedure using central venous lines as vascular access, but not characteristic of IA. Compared with the safety data from previously published PE studies [15, 21], the incidence of mild adverse events was higher in our study, and moderate side effects were slightly more frequent. However, moderate side effects were almost level with our own study of neurological patients treated with PE [15]. Potential side effects of PE known to be related to the substitution of human plasma products were completely avoided [12].

Apart from clinical data, we analyzed possible therapeutic effects of IA with the help of proteomic analyses. Several relevant proteins, particularly fibrinogen and the immunoglobulins, were monitored. The decrease of fibrinogen is one limiting factor in the use of tryptophan-IA that makes regular controls necessary. Our protocol with five IA sessions on alternate days did not decrease fibrinogen to critical levels. Moreover, immunoglobulin depletion along with prior corticosteroid pulse therapy reflects a strong immunosuppression, which makes close controls of clinical and laboratory infection signs necessary. Previous investigations reported that the restoration of serum IgG levels until day 5 after IA does not result from increased antibody synthesis, but is probably related to changes of catabolism and immunoglobulin backflow [7]. Interestingly enough, we found a significant immunoglobulin increase beyond baseline values until day 180 ± 10 after the start of IA, suggesting additional mechanisms other than backflow alone.

Several mechanisms of PE action in neuroimmunological disorders have been described, such as a removal of pathogenic autoantibodies, a redistribution of pathogens from the extravascular to the intravascular compartment, increased proliferation of immune cells, an enhanced production of immunoglobulins, a promotion of suppressor T-cell function, and a deviation of cytokine patterns redressing a disturbed T-helper type 1 and T-helper type 2 balance [3]. Although IA has been termed specific, several studies have demonstrated additional binding properties of ligands other than immunoglobulins alone [22, 23]. According to our proteomics data, several proteins that are possibly involved in MS pathogenesis are removed from the plasma by IA, for example, transthyretin [24], serum amyloid P [24], complement factors [24], clusterin [24], gelsolin [24], kininogen-1 [24], MBP [25, 26], CD5L [27] and immunoglobulins [1, 3, 8]. We confirmed a decrease of serum levels by IA in two of them, MBP and sCD5L. MBP-like material has been detected in several body fluids including cerebral spinal fluid and the urine of patients with MS [28]. Since MBP and other myelin proteins have been shown to be encephalitogenic in animal models of MS, they could drive the systemic autoimmune response in patients with MS. Other investigations have demonstrated the prevalence of MBP-specific memory B-cells in the peripheral blood of relapsing patients with remitting MS that might prime T-cells in lymphoid organs to migrate into the central nervous system and to elicit IFN-γ secretion [25]. These data were further corroborated with the evidence of MBP-reactive T-cells among IL-2 expanded lymphocytes in patients with MS [26]. Thus, removal of MBP from the plasma by IA might interrupt these autoimmune mechanisms, although more research in this hypothesis is definitely needed.

sIL-2R, a marker of T_H1 cell activation, is increased in the serum of patients with relapsing MS [14]. We induced a significant decrease of sIL-2R after IA in responders, but not in non-responders. Decreased sIL-2R levels might reflect the silencing of cellular autoimmune responses effective only in responders.

Our clinical results show a high clinical efficacy of tryptophan-IA comparable to PE in the treatment of MS relapses refractory to corticosteroids. Furthermore, our experimental data suggest several possible effects on MS pathogenesis: not only removal of immunoglobulins and complement from plasma, but also reduced levels of circulating autoantigens and regulatory proteins. We suggest that more prospective studies are needed to confirm and extend our results, to give new insights into this treatment approach, to optimize therapeutic IA protocols and, lastly, to investigate further therapeutic principles, for example, T-cell reactivity to MBP before and after IA.