Treatment of stiff-person syndrome with chronic plasmapheresis.

Stiff-person syndrome (SPS) is a disabling central nervous system disorder characterized by muscle stiffness, rigidity, and superimposed painful spasms affecting axial and proximal limb muscles. Electromyography shows continuous motor unit activity in axial muscles. There is strong evidence of an autoimmune pathogenesis for SPS. High levels of antibodies against the 65-kD isoform of glutamic acid decarboxylase (anti-GAD65) are found in the serum and/or cerebrospinal fluid of 85% of patients. Paraneoplastic SPS commonly occurs in association with antiamphiphysin antibodies. Different immunomodulatory treatments have been tried in SPS with variable success. Plasmapheresis, usually performed in 1 cycle of 5 plasma exchange (PE) sessions over 1 or 2 weeks, has been tried in a few SPS patients with conflicting results. We report the clinical improvement induced by chronic plasmapheresis in 2 patients with SPS and previous poor response to symptomatic and other immunomodulatory treatments. Clinical characteristics of the patients are summarized in Table 1. Plasmapheresis was performed using plasmaFlux PSu 2S filters (Fresenius Medical Care, Waltham, MA). A central venous catheter (subclavian) was inserted for the first sessions. Once it was confirmed that plasmapheresis was effective, a radiocephalic fistula was created. Patients underwent 1 PE per week, with 2 L of 5% albumin as replacement fluid during each procedure. In patient 1, clinical improvement was observed after the first 4 sessions of PE and increased during the following 2 months. The spasms disappeared, face and limb stiffness subsided, and trunk stiffness was considerably reduced. The patient was able to bend at the waist and could place her fingertips at a distance of 20 cm from the floor. Lumbar pain disappeared. The patient’s clinical situation remains stable after 3 years of weekly PE. Serum anti-GAD65 antibody titers (enzyme-linked immunosorbent assay), determined 1 and 2 years after the onset of chronic plasmapheresis, remained >2000 U/mL (normal values < 10 U/mL). In patient 2, clinical improvement was observed after the first month of therapy and continued during the second month before becoming stable. After 14 months of weekly PE, the frequency of axial spasms was reduced by more than 50%, and there was marked improvement in axial stiffness. The patient is now able to bend at the waist and can place his fingertips a distance of 10 cm from the floor. He did not report falls. Anti-GAD65 antibody titers in serum after 4 months of chronic plasmapheresis remained >2000 U/mL. We found that chronic plasmapheresis can improve spasms and axial stiffness in patients with severe SPS who did not obtain sufficient benefit from symptomatic or other immunomodulatory treatments. The improvement in the cardinal symptoms of SPS was sustained for up to 3 years of weekly PE, which, to our knowledge, is the longest reported period of effective chronic plasmapheresis in SPS. The modulatory effect of plasmapheresis on the immune system does not persist over time, and the frequency of PE needs to be determined individually according to the underlying disease, the patient’s improvement, and the rebound of pathogenic factors. Therefore, it seems plausible that patients with SPS—a chronic condition of probable autoimmune etiology—may experience sustained benefit from chronic plasmapheresis. In addition, long-term treatment with 1 PE per week was well tolerated, and no adverse effects were observed in our patients. It is noteworthy that serum anti-GAD65 antibody titers remained high (>200-fold the normal values) despite the frequency of PE. This finding has also been observed after short cycles of PE. The discrepancy between clinical improvement and high levels of anti-GAD65 antibodies calls into question the pathogenic role of these antibodies in SPS. Although clinical symptoms may be more related to intrathecal synthesis of anti-GAD65 antibodies, it is unclear whether these antibodies cause SPS directly, are markers of autoimmunity, or are an epiphenomenon of neural destruction. The therapeutic effect of plasmapheresis in our cases may be related to the elimination of other pathogenic autoantibodies, complement, or cytokines or to the modulation of other components of the immune system.