Plasma exchange is a well-established therapeutic procedure commonly used in many neurological disorders of autoimmune etiology. It is thought that the beneficial effects of plasma exchange occur through the elimination of pathognomonic inflammatory mediators, including autoantibodies, complement components, and cytokines. In various neurological disorders, randomized controlled studies have demonstrated the efficacy of plasma exchange (eg, in Guillain-Barré syndrome and other forms of immune neuropathies). Although widely used, the potential benefit of plasma exchange in the treatment of multiple sclerosis, myasthenia gravis, and Lambert-Eaton syndrome is less clear.
This part of our review series outlines the rationale and technical aspects of plasma exchange. We also define its current role in the treatment of inflammatory disorders of the central nervous system (CNS).
Plasma exchange has become an established therapeutic procedure in neurological practice for numerous pathologic conditions. In fact, the latest review of plasma exchange use by the Canadian Apheresis Group indicates that 3 neurological disorders (myasthenia gravis, Guillain-Barré syndrome, and chronic inflammatory demyelinating polyneuropathy) are among the 5 most frequent indications for this therapy.1 Most neurological disorders that are treated with plasma exchange are associated with presumed aberrant humoral immune responses, including myasthenia gravis, Guillain-Barré syndrome, and chronic inflammatory demyelinating polyneuropathy.2 In some of these disorders, the efficacy of plasma exchange has already been demonstrated in randomized controlled clinical trials, whereas its role in the treatment of other diseases remains less clear.
This review will assess and analyze information on relevant randomized controlled studies that provide the basis for the use of plasma exchange in neurological disorders. In the first part of this series, we will review the potential mechanisms of action of plasma exchange and the technical aspects of the procedure and will focus on its current role in the treatment of inflammatory disorders of the CNS. To rate the quality of evidence of specific clinical studies, a rating system was adopted from the report of assessment of disease-modifying therapies in multiple sclerosis (MS) of the Therapeutics and Technology Assessment Subcommittee of the American Academy of Neurology and the MS Council for Clinical Practice Guidelines3,4 (Table 1).
Therapeutic plasma exchange constitutes an extracorporeal blood purification technique designed to remove large molecular weight particles from plasma. The removal of circulating autoantibodies, immune complexes, cytokines, and other inflammatory mediators is thought to be the principal mechanism of action. Antibodies against self have been identified in various neurological disorders, including antibodies against nicotinic acetylcholine receptor in myasthenia gravis,5 antibodies against P/Q-type voltage-gated calcium channels in Lambert-Eaton syndrome,6 and antimyelin oligodendrocyte glycoprotein antibodies in MS.7 Cytokines, including chemokines8 and complement,9 are other potentially injurious molecules that may be removed by plasmapheresis.
Clinical benefit from plasma exchange is primarily observed in diseases with a self-limited course, whereas a long-term effect in chronic disorders is less frequently achieved. In antibody-mediated diseases, this could be due to the removal of an insufficient number of pathogenic autoantibodies and their continued synthesis with repeated antigenic stimulation. A sudden decline in autoantibody levels may also result in an increased antibody production after treatment. Furthermore, the intravascular and extravascular distribution of substances that are desired to be removed by plasma exchange has to be considered. Most large molecular weight substances have considerable concentrations in the extravascular space. After removal of the substance in the intravascular space, there may be a rapid substance redistribution into the intravascular space, which usually requires repeated treatments with plasma exchange.10
The transient effects of plasma exchange and the possibility of a rebound phenomenon would call for the combination of short-term active plasma exchange with long-term immunosuppression. The sudden abolition of feedback inhibition of B cells may synchronize their activity and render them particularly sensitive to a pulse of immunosuppression, a concept that has been tried in the treatment of severe lupus erythematosus.11
Further proposed modes of action of plasma exchange relate to potential disturbing effects on the cellular network of T cells, autoantibody-producing plasma cells, and macrophages. Experimental findings suggest that treatment with plasma exchange can lead to an increased proliferation of immune cells,12 an enhanced production of immunoglobulins,13 a promotion of suppressor T-cell function,14 and a deviation of cytokine patterns redressing a disturbed T-helper type 1 and T-helper type 2 balance.15- 17
Complete removal of pathogenic antibodies is impossible to achieve. Because of a slow equilibration of large macromolecules between the vascular space and the interstitium, the rate of removal can be expressed as a first-order kinetic. With the use of replacement solutions, the exchange of a single volume of plasma will lower the level of a specific macromolecule by 50% to 60%. Similarly, an increase to 1.4 plasma volumes will lower plasma levels by 75%.18
Therapeutic plasma exchange is based on the separation of plasma from the blood's cellular elements. This can be achieved with centrifugation devices or with permeable blood filters. During continuous or intermittent centrifugation, blood components separate because of differences in density. In membrane ultrafiltration, the separation is according to molecular size. Plasma filter membrane pores are up to 0.2 μm in diameter (approximately 30 times the diameter of pores in conventional high-flux hemofilter membranes), allowing the removal of substances up to a molecular weight of 3 × 106 Da, which includes immunoglobulins, immune complexes, complement factors, lipoproteins, and endotoxin. At large filtrate production rates, an extent of hemoconcentration is achieved that may predispose to premature clotting of the filter. Therefore, blood flow rates requiring a central venous access are usually necessary. Furthermore, the degree of anticoagulation required to prevent clotting is somewhat higher than that needed in hemodialysis patients. Often, systemic heparinization is used for anticoagulation. When centrifugation techniques are applied, the minimum blood flow is generally lower, allowing the use of a peripheral venous access. Citrate anticoagulation instead of a heparin pump is an integrated part of many technical devices. Centrifugal plasmapheresis removes all nonsolid elements from the blood. However, compared with membrane filtration, the loss of cellular elements and especially platelets is unavoidable. Net clearance of a substance by centrifugation is equal to the volume of plasma removed in a given time. When filtration techniques are used, a given volume has to be multiplied by the membrane's sieving coefficient for a distinct protein (the ratio between the protein's filtrate concentration and plasma concentration). Sieving coefficients for several inflammatory mediators (eg, in patients with sepsis) have been published.9
In most pathologic conditions in which plasma exchange is used, 1 to 1.5 plasma volumes are exchanged per procedure per day. Routine treatment durations of 3 to 5 days may be prolonged, depending on the diagnosis and the individual patient's condition. Clinical data referring to the number of plasma exchanges necessary to achieve a maximum effect without excessive risks and costs will be discussed later. There is no consensus on the ideal replacement solution for plasma discarded during plasmapheresis. Except for distinct diseases like hemolytic uremic syndrome or thrombotic thrombocytopenic purpura in which substitution is clearly by fresh-frozen plasma, colloid replacement can be achieved with the use of fresh-frozen plasma, albumin, albumin and saline, or albumin and plasma expander solutions.19 In a routine setting, the oncotic pressure in a patient is calculated using a slide normogram, and an iso-oncotic replacement fluid is used that is composed of human albumin and a basic electrolyte solution.
In addition to plasmapheresis, the immunoadsorption technique represents a newer approach that allows a more selective removal of circulating antibodies. By binding to a ligand, removal of immunoglobulin fractions can be achieved.20 After adsorption, the cellular blood components and plasma are combined and reinfused. This technique has already been used in different neurological diseases, including MS21 and myasthenia gravis.22
Possible complications related to the use of central venous access, anticoagulation, or replacement fluids have to be considered, especially when treatment alternatives exist. Citrate infused for anticoagulation or as part of fresh-frozen plasma may lead to hypocalcemia or alterations in acid-base homeostasis (a metabolic acidosis resulting from the breakdown of bicarbonate).23 Symptoms of hypocalcemia include paresthesiae, muscle cramps, and, in severe cases, cardiac arrhythmias. The incidence of hypocalcemia-related symptoms ranges from 1.5% to 9%.24,25 Repeated apheresis treatments with albumin replacement may lead to depletion of coagulation factors and immunoglobulins, increasing the risk of bleeding and infections. Adverse effects that are related to the central venous access are infection and septicemia, thrombosis, and pneumothorax.24,26 If filtration techniques are used, hemolysis and hypotension may occur.27 Although extremely rare, there is a risk of viral transmission by replacement of fresh-frozen plasma.28
Multiple sclerosis is a multifocal inflammatory disease of the CNS, characterized by chronic inflammation, demyelination, axonal damage, and subsequent gliosis.29- 31 Current concepts of its pathogenesis assume that in genetically susceptible individuals potentially self-reactive T cells are activated in the immune system, home onto the CNS, and may initiate tissue damage via release of inflammatory cytokines, stimulation of B cells and macrophages, and activation of the complement system. Recently, there has been a revival of interest in the role of humoral factors in the pathologic process (Figure). Antibodies against myelin basic protein and myelin oligodendrocyte glycoprotein have been detected in subgroups of patients with MS.32 These antibodies may mediate injury by complement fixation or linking with innate immune effector cells such as macrophages.33 Berger et al7 demonstrated that antibodies in serum samples of patients with a clinically isolated syndrome are predictive of early conversion to clinical definite MS. While clearly immunoglobulins are synthesized intrathecally, these findings along with others suggest that at least part of the humoral response in MS is derived systemically from the blood. However, no major therapeutic effect of plasma exchange can be expected once antibodies are deposited in situ in CNS lesions. The role of plasma exchange in the management of patients with progressive forms of MS remains unclear.
Hypothetical scheme of the humoral immune response in multiple sclerosis. Antibodies against myelin proteins may cause demyelination and subsequent axonal damage. Other humoral factors that may contribute to the pathogenesis include cytokines, metalloproteinases (MMPs), and complement (Compl). Abs indicates antibodies; AM, adhesion molecules; APC, antigen presenting cell; B, B-cell; BBB, blood-brain barrier; C5b-9, complement 5b-9; CNS, central nervous system; IL, interleukin; LN, lymph node; Mφ, φ macrophage; and T, T cell.
The first randomized, controlled, double-blind trial of plasmapheresis, by Khatri and coworkers34 among 54 patients with a progressive form of MS, showed a clinical benefit after 5 months in the plasma exchange group compared with the sham treatment group. A larger 3-armed trial of the Canadian Cooperative Multiple Sclerosis Study Group35 randomized 168 patients with chronic progressive MS to receive daily cyclophosphamide and oral prednisone (n = 55), daily cyclophosphamide and prednisone every other day for 22 weeks and weekly plasma exchange for 20 weeks (n = 57), or placebo medications and sham plasma exchange (n = 56). All patients were followed up for at least 12 months. The study showed no efficacy in the outcome measure, defined as Expanded Disability Status Scale progression by at least 1 point (treatment failure) at 6 months and mean change in Expanded Disability Status Scale at 12 months. The long-term benefit of plasma exchange in the treatment of chronic progressive forms of MS remains elusive, and plasma exchange cannot be recommended (type A recommendation [Table 1]).3
A weakness of most studies of plasma exchange in chronic progressive forms of MS is the lack of evaluation of magnetic resonance imaging activity, nowadays the most important surrogate marker in the assessment of potential therapeutic drug effects in MS. A small randomized, single-blind, crossover study36 of 11 patients with secondary progressive MS investigated the effect of plasma exchange in combination with azathioprine on metrics of magnetic resonance imaging activity (gadolinium-enhancing lesions and new enhancing lesions). No significant differences were found with regard to this outcome measure, perhaps confirming the disappointing results in the earlier larger trials. The Therapeutics and Technology Assessment Subcommittee of the American Academy of Neurology and the MS Council for Clinical Practice Guidelines concluded that plasma exchange “is of little or no value in the treatment of progressive MS.”3(p117)
In contrast to chronic progressive MS, plasma exchange may be useful in acute exacerbations of MS.2 The largest trial to study the effectiveness of plasma exchange in acute exacerbations in MS dates back to 1989.37 In this randomized, controlled, double-blind trial, 116 patients were assigned to receive 11 treatment cycles of plasma exchange or sham exchange during 8 weeks. Corticotropin (ACTH) and cyclophosphamide were given to patients in both groups as indicated. During the treatment period, a greater clinical improvement was observed in the plasma exchange group. Unfortunately, the long-term effect at 12 months was less clear (class I evidence [Table 1]).
Weinshenker and coworkers38 demonstrated in a randomized, controlled, double-blind trial the beneficial effects of plasma exchange in acute forms of CNS demyelinating diseases. Twenty-two patients with MS or other demyelinating diseases (transverse myelitis, Marburg variant of MS, acute disseminated encephalomyelitis, neuromyelitis optica, and focal cerebral demyelination) failing to respond to high-dose corticosteroids received 7 cycles of plasma exchange or sham treatment. Patients who showed no clinical improvement on day 14 were crossed over to the other trial arm. In this study, plasma exchange was superior to sham treatment in the primary outcome measure, which was defined as improvement of a target neurological deficit (coma, aphasia, hemiplegia, paraplegia, or quadriplegia) (class I evidence [Table 1]). Although the number of patients in this trial was small, plasma exchange is now regarded as a treatment option for patients with acute fulminant CNS demyelinating diseases who fail to improve with high-dose corticosteroid treatment (type A recommendation). However, further studies with a larger number of patients are required to examine the optimal number of treatment cycles.
Recently, Keegan and colleagues39 retrospectively reviewed the outcome after plasma exchange treatment in 19 patients who underwent brain biopsy before treatment initiation. This cohort was stratified according to the immunopathologic pattern of MS lesions, as described by Lucchinetti et al.40 Keegan and coworkers demonstrated that there was a strong correlation between the beneficial effects of plasma exchange therapy and the immunopathologic pattern II, which is characterized by antibody-mediated and complement-mediated demyelination. This study underlines the importance of identifying patient subgroups that are more likely to benefit from plasma exchange.
Favorable effects of plasma exchange were also seen in 2 recent retrospective case series of patients with severe optic neuritis and severe demyelinating events of the CNS.41,42 Despite limitations in the study design (no control group), these results encourage further evaluation of the role of plasma exchange in patients with severe demyelinating events.
Rasmussen encephalitis is a rare disorder of childhood onset characterized by intractable focal seizures, hemiplegia, dementia, and unilateral brain dysfunction.43 Cumulative evidence suggests that Rasmussen encephalitis has an autoimmune etiology.43 Humoral factors may be involved in the pathogenesis of the disease. Pathologic hallmarks are inflammation and gliosis in the affected cerebral hemisphere. Focal disruption of the blood-brain barrier, perhaps caused by focal seizures, may allow the access of pathologic humoral factors to brain tissue. The subsequent local inflammation perpetuates the vicious circle of neural injury, focal seizure, and transient disruption of the blood-brain barrier.44 Antibodies have been detected in serum samples of patients with Rasmussen encephalitis that are directed against the glutamate receptor GluR3. A major pathogenic role of anti-GluR3 antibodies has been challenged because they have also been identified in patients with focal epilepsy and (in lower frequency) in other neurological diseases,45,46 and their contribution remains unresolved. Cytotoxic CD8 T cells have been identified in the brains of affected individuals with Rasmussen encephalitis, and it has been suggested that their direct assault on neurons underlies disease pathogenesis.47
To our knowledge, no randomized controlled trial has examined the effectiveness of plasma exchange in Rasmussen encephalitis; therefore, evidence is categorized as class IV (Table 1). Single case reports have demonstrated marked clinical improvement and decreasing immunoreactivity against GluR3 in serum samples of treated patients; thus, the clinical value of plasma exchange is uncertain (type U recommendation).44,48 Andrews and coworkers44 reported in a small case series the rapid resolution of intractable seizures in some patients with plasma exchange (Table 2). Therefore, status epilepticus, a complication frequently observed in Rasmussen encephalitis, might be an additional clinical indication for plasma exchange.
Paraneoplastic disorders of the CNS, including cerebellar degeneration, paraneoplastic encephalomyelitis, stiff-person syndrome, and others, are rare neurological disorders associated with cancer.49,50 The presence of circulating autoantibodies against different neural antigens implicates an autoimmune-driven pathogenesis in these disorders. Paraneoplastic cerebellar degeneration is clinically characterized by subacute cerebellar dysfunction and is mostly associated with carcinoma of the ovary and lung. In most patients, immunoreactivity against Purkinje cell antigens can be detected. Paraneoplastic encephalomyelitis is a multifocal inflammatory disease of the CNS, mainly associated with small cell lung cancer and the presence of antineuronal nucleus antibodies (anti-Hu).51 Because of the rare occurrence of both syndromes (paraneoplastic cerebellar degeneration and paraneoplastic encephalomyelitis), there are no randomized controlled studies of the use of plasma exchange, to our knowledge. However, retrospective studies and case series did not show any substantial benefit of clinical and serological parameters51- 53 (class IV, type U recommendation [Table 1]).
In patients with stiff-person syndrome, high titers of antibodies against glutamic acid decarboxylase have been demonstrated in serum samples and cerebrospinal fluid. Clinical symptoms of this rare disorder are muscle rigidity and spasm of axial and limb musculature. Although we are unaware of any randomized controlled trial of the utility of plasma exchange in stiff-person syndrome, there are some anecdotal reports with beneficial results54- 57 (class IV, type U recommendation [Table 1]).
Correspondence: Hans-Peter Hartung, MD, Department of Neurology, Heinrich Heine University of Düsseldorf, Moorenstrasse 5, 40225 Düsseldorf, Germany (firstname.lastname@example.org).
Accepted for Publication: October 17, 2005.
Author Contributions:Study concept and design: Hartung, Hetzel, and Stüve. Acquisition of data: Stueve. Analysis and interpretation of data: Lehmann, Hartung, Hetzel, Stüve, and Kieseier. Drafting of the manuscript: Lehmann, Hartung, and Stüve. Critical revision of the manuscript for important intellectual content: Hartung, Hetzel, and Kieseier. Obtained funding: Kieseier. Administrative, technical, and material support: Hartung and Stüve. Study supervision: Hetzel and Stüve.
Thank you for submitting a comment on this article. It will be reviewed by JAMA Neurology editors. You will be notified when your comment has been published. Comments should not exceed 500 words of text and 10 references.
Do not submit personal medical questions or information that could identify a specific patient, questions about a particular case, or general inquiries to an author. Only content that has not been published, posted, or submitted elsewhere should be submitted. By submitting this Comment, you and any coauthors transfer copyright to the journal if your Comment is posted.
* = Required Field
Disclosure of Any Conflicts of Interest*
Indicate all relevant conflicts of interest of each author below, including all relevant financial interests, activities, and relationships within the past 3 years including, but not limited to, employment, affiliation, grants or funding, consultancies, honoraria or payment, speakers’ bureaus, stock ownership or options, expert testimony, royalties, donation of medical equipment, or patents planned, pending, or issued. If all authors have none, check "No potential conflicts or relevant financial interests" in the box below. Please also indicate any funding received in support of this work. The information will be posted with your response.
Some tools below are only available to our subscribers or users with an online account.
Download citation file:
Web of Science® Times Cited: 48
Customize your page view by dragging & repositioning the boxes below.
Users' Guides to the Medical Literature
Table 9.2-3 Refuted Evidence From Observational Studiesa
All results at
Enter your username and email address. We'll send you a link to reset your password.
Enter your username and email address. We'll send instructions on how to reset your password to the email address we have on record.
Athens and Shibboleth are access management services that provide single sign-on to protected resources. They replace the multiple user names and passwords necessary to access subscription-based content with a single user name and password that can be entered once per session. It operates independently of a user's location or IP address. If your institution uses Athens or Shibboleth authentication, please contact your site administrator to receive your user name and password.