Comparison of IVIg and TPE efficacy in the treatment of neurological disorders: a systematic literature review

Abstract

Background:

Intravenous immunoglobulin (IVIg) and therapeutic plasma exchange (TPE) are among the main immunotherapies for neurological disorders. Their benefit is greatest in immune-mediated conditions, but their distinct efficacy cannot be simply explained.

Objectives:

This review aimed to systematically identify studies comparing the efficacy of TPE and IVIg treatments for selected autoimmune neurological disorders and identify optimal therapies for each condition.

Data Sources and Methods:

PubMed, MEDLINE and Embase databases were searched for original publications from 1990 to 2021. Additional publications were identified via expert recommendations. Conference abstracts older than 2017, review articles and articles without information on TPE and IVIg comparison in title and abstract were excluded. Risks of bias were descriptively addressed, without a meta-analysis.

Results:

Forty-four studies were included on Guillain–Barré syndrome (20 studies – 12 adult, 5 paediatric, 3 all ages), myasthenia gravis (11 studies –8 adult, 3 paediatric), chronic immune–mediated polyradiculoneuropathy (3 studies –1 adult, 2 paediatric), encephalitis (1 study in adults), neuromyelitis optica spectrum disorders (5 studies –2 adult, 3 all ages) and other conditions (4 studies – all ages). TPE and IVIg were mostly similarly efficacious, measured by clinical outcomes and disease severity scores. Some studies recommended IVIg as easy to administer. TPE procedures, however, have been simplified and the safety has been improved. TPE is currently recommended for management of neuromyelitis optica spectrum disorder relapses and some myasthenia gravis subtypes, in which rapid removal of autoantibodies is crucial.

Conclusion:

Despite some limitations (e.g. the low evidence levels), this review provides an extensive 30-year-long overview of treatments for various conditions. Both IVIg and TPE are usually comparably efficacious options for autoimmune neurological disorders, with few exceptions. Treatment choices should be patient-tailored and based on available clinical resources. Better designed studies are needed to provide higher-level quality of evidence regarding clinical efficacy of TPE and IVIg treatments.

Keywords

autoantibodies autoimmune neurological disorders intravenous immunoglobulin IVIg therapeutic plasma exchange TPE treatment efficacy

Introduction

Intravenous immunoglobulin (IVIg) and therapeutic plasma exchange (TPE) have been used to treat several immune-mediated neurological conditions.^1,2 While these treatments are often reported to have similar efficacies, there are some nuances and differences in their utility in clinical practice.

IVIg contains purified human immunoglobulins, manufactured from large pools of plasma from thousands of healthy donors.^1,3 IVIg is thought to achieve its therapeutic effects by inactivating the immune mediators (e.g. immune complexes and antibodies targeting nervous system components) and modulating immune responses [e.g. fragment crystallisable (Fc) receptor modulation, impeded activation of the complement system, altered B- and T-cell activation, decreased inflammatory cell adhesion and migration].^1,4

In contrast to IVIg, TPE aims to eliminate the immune mediators.^1,3 During TPE, the patient’s blood is separated into plasma and cellular components. The plasma fraction is replaced with fresh frozen plasma from a donor or a colloid solution (e.g. albumin), which is re-infused into the patient together with autologous blood cells.^1,3,5 Between 66% and 85% of the original noncellular intravascular constituents are removed after exchange of 1.0–1.5 plasma volumes.³ Removal of antibodies that are directly pathogenic may be particularly important in some conditions, such as myasthenia gravis (MG)^6,7 and neuromyelitis optica spectrum disorders (NMOSD).⁸

Both IVIg and TPE decrease the levels of circulating autoantibodies and immune mediators, although via different mechanisms and to varying degrees. As each approach brings specific benefits but also risks,^1,3 the therapy of choice must be tailored to each patient and their condition. For instance, neurologists may prefer IVIg over TPE in patients with severe infections. Furthermore, it may not always be feasible to administer each of these therapies. TPE requires planning and provision (e.g. trained medical personnel); TPE-associated complications may be catheter-related (e.g. suboptimal placement of central venous lines, line sepsis, hematoma) or exchange-related (e.g. electrolyte imbalance or hypotension).^1,3 Similarly, IVIg may cause adverse events (AEs), which are rarely severe (e.g. renal impairment, haemolytic anaemia, arterial or venous thrombosis) but may occur more frequently, particularly if patients had already experienced adverse effects or if IVIg is administered for the first time.⁹ The choice between TPE and IVIg often depends on several factors: availability, ease of administration and cost in local settings.¹⁰ Specific circumstances such as the ongoing COVID-19 pandemic may further impact availability and usage.¹⁰

The recommendations for IVIg or TPE treatment of neurological disorders vary locally, and many were published more than 5 years ago.^2,3,5,11 Establishing robust recommendations for IVIg and TPE use in autoimmune neurological disorders is limited by low disease incidence leading to underpowered therapeutic study designs, often enrolling fewer than 100 patients. This systematic literature review (SLR) aims to summarise data on IVIg and TPE treatment efficacies in several autoimmune diseases over a 30-year period.

Material and methods

Review objectives

To conduct an SLR of studies comparing the efficacy of TPE versus IVIg to treat autoimmune neurological diseases.

To identify those conditions in which both treatments show similar efficacies and those where one is preferred due to superior clinical benefit.

To recommend how to apply available evidence for treatment of rare autoimmune neurological diseases and recommend strategies for future studies.

Search strategy and selection criteria

This SLR focused on neurological disorders treated with IVIg or TPE. The PubMed, MEDLINE and Embase databases were screened over two rounds: the first in 2020 (period: 1990–2020) and the second, as an update of the first one, in 2021 (period: January 2020–June 2021), each with the same set of four search strings (Supplementary Table 1). Three searches (one search of PubMed, Embase and Medline each; Supplementary Table 1) were initially run more ‘globally’, that is, including all references describing IVIg and TPE treatments in all neurological conditions. This strategy, however, yielded few records directly comparing IVIg and TPE, which prompted us to run a fourth search (in PubMed) specifically designed to retrieve papers comparing both therapies (Supplementary Table 1). Immunoadsorption was not included as a search term, even though it is an alternative to TPE in some pathophysiological backgrounds. Immunoadsorption is a distinct technique, used to a significant extent only in some countries (e.g. Japan and Germany).¹² As the present review focuses on the widely available TPE and IVIg comparison, studies on immunoadsorption were not within its scope. The authors also provided input on relevant publications.

The following studies were included: original publications (from 1990 to 2021) or congress/conference abstracts (from 2017 to 2021), comparing efficacy of IVIg and TPE therapies for neurological disorders; studies in humans; no restrictions for language or geographical region. Studies in which patients were switched from one treatment to another were also included.

Publications on organ transplantation or non-neurological conditions, review papers, noncomparative TPE or IVIg efficacy studies (with the exception of NMOSD studies, see below) and studies with combined TPE and IVIg treatments were excluded. While review articles (narrative, systematic and meta-analyses) were excluded, additional studies of interest were identified by checking their reference lists.

During the first search round, four reviewers screened the identified records for inclusion in the review, and conflicts were resolved by consensus. Three reviewers were involved during the second round, in which conflicts were resolved by consensus and expert opinion. Selected records included papers and abstracts meeting all inclusion criteria. An exception was made for studies on NMOSD treatment, because of the relevance of this group of disorders, for which the search identified no comparative studies. Data from selected studies were extracted by two independent reviewers in the first and by one reviewer in the second search round. Oxford Centre for Evidence-Based Medicine criteria were used to determine study robustness [level of evidence (LoE)] and grade of recommendation.^13,14

The following characteristics were tabulated from each included study: country of study, study design, LoE, treatment protocols, objectives, main outcomes and conclusions. A meta-analysis of extracted data could not be done, due to the overall low LoE and differences in protocols of the included studies. Risks of bias were thus descriptively assessed following expert input.

Results

Search results

After conducting both search rounds and applying the selection criteria, a total of 2262 records (2067 publications and 195 conference abstracts) were identified in PubMed, MEDLINE and Embase databases; 1189 duplicates were removed yielding 1073 unique records. After excluding an additional 933 records, 140 publications and abstracts were assessed for eligibility. Of those, 33 were analysed in this review. Expert input provided 11 additional records of interest (Figure 1).

Figure 1.

PRISMA flow diagram.

Efficacy of IVIg and TPE therapies per category

Guillain–Barré syndrome

Guillain–Barré syndrome (GBS) is an autoimmune disease of the peripheral nervous system, often preceded by a bacterial or viral infection and diagnosed more often in men.¹⁵ It is characterised by rapidly progressing, ascending flaccid limb paresis and areflexia, usually starting in the lower limbs, and worsening over 1–4 weeks.¹⁵

GBS is usually classified into four subtypes: (1) acute inflammatory demyelinating polyradiculoneuropathy (AIDP), in which immune targeting of Schwann cells or myelin results in segmental demyelination with secondary axonal loss; (2) acute motor axonal neuropathy (AMAN), in which axonal degeneration occurs as a consequence of an autoimmune reaction against ganglioside epitopes; (3) acute motor and sensory axonal neuropathy (AMSAN); (4) Miller-Fisher syndrome, characterised by ataxia, ophthalmoplegia and areflexia but usually not limb weakness. The subtype prevalence differs between continents.¹⁶

The systematic search identified 20 studies comparing TPE versus IVIg (Table 1), in which treatment efficacy was measured by improvement on established disability scales, mostly the Hughes and GBS disability scale; one study also measured compound Functional Independence Measure and Montebello Rehabilitation Factor Score.³⁰

Table 1.

TPE and IVIg in treatment of Guillain–Barré syndrome.

Study	Design, country, CEBM LoE	Patient population	Treatment	Measure of treatment efficacy	Clinical outcome	Author conclusion
van der Meché and Schmitz¹⁷	Multicentre randomised trial, The Netherlands 2	Adults, n = 147	• TPE (continuous): 200–250 ml/kg in five sessions within 7–14 days; n = 73 (ultrafiltration: n = 42, centrifugation: n = 31) • IVIg: 0.4 g/kg BW/day, 5 days; n = 74	• Clinical improvement by ⩾1 degree (motor functional scale) at 4 weeks after randomisation • Time to clinical improvement (days) • Time to unaided locomotion (days)	• Twelve patients discontinued TPE (different problems). • Clinical improvement: lower % in the TPE (34%) versus the IVIg group (53%). • The median time to improvement: longer in the TPE (41) versus the IVIg group (27). • Recovery of independent locomotion: longer in the TPE (69) versus the IVIg group (55).	• IVIg was comparably or more efficacious compared with TPE. • TPE was associated with more complications than IVIg treatment.
Bril et al.¹⁸	Randomised prospective clinical trial, Canada 2	Adults, n = 44	• TPE (continuous, with centrifugal or membrane-cell separation): 200–250 ml/kg over five sessions, over 7–10 days; n = 18 • IVIg: 0.5 g/kg/day, over consecutive days, for a total dose of 2.0 g/kg; n = 26	• Disability grade (Hughes scale) • Time to improvement (by ⩾1 grade on Hughes scale) (days) • Improvement at 1 month after treatment	• Clinical improvement: comparable between the TPE (61%) and the IVIg (69%) groups at 1 month after treatment. • Median time to improvement by one disability grade: comparable between TPE (90 days) and IVIg (65 days). • Complications more frequent in the TPE (19) versus the IVIg (5) group.	• The efficacy of the two treatments was comparable.
Plasma Exchange/Sandoglobulin Guillain–Barré Syndrome Trial Group¹⁹	Randomised controlled trial, multicentre 2	Adults and children (⩾16 years), n = 379	• TPE: 50 ml/kg, five sessions over 8–13 days; n = 121 • IVIg: 0.4 g/kg/day, for 5 days; n = 130 • TPE + IVIg: n = 128	• Disability grade • Arm grade • Vital capacity	• TPE versus IVIg: 0.09 grade improvement. • No difference in: time to recovery of unaided walking, duration of MV and recovery trend from disability up to 48 weeks. • Complications associated with treatment: eight patients in TPE, six patients in IVIg group. In TPE + IVIg group, six patients had complications associated with TPE and nine patients with IVIg treatment.	• TPE and IVIg treatment had equivalent efficacy.
Diener et al.¹²	Phase III clinical trial–open RCT, Germany 2	Adults with AIDP, n = 67	• TPE: five sessions; n = 26 • IVIg: 0.4 g/kg/day, for 5 days; n = 23 • Immune adsorption: five sessions; n = 18	• Clinical improvement (⩾1 grade on physical disability scale) by day 28, month 6 and month 12 • Interval between onset of symptoms and clinical improvement • Length of intubation, hospitalisation and rehabilitation (days)	• Clinical improvement by day 28: comparable in the TPE (71%) and IVIg groups (80%). • Length of intubation, hospitalisation and rehabilitation: comparable in the TPE and IVIg groups.	• Treatment efficacy was similar in the TPE and IVIg groups.
Kuwabara et al.²⁰	Clinical trial, Japan 4	Adults positive for anti-GM1 antibody, n = 34 • AMAN: 21 • AIDP: 2 • Unclassified: 1 • Mild symptoms: 10	• TPE (double-filtration): 50 ml/kg plasma, average 5.1 sessions; n = 14 • IVIg: 0.4 g/kg/day, five successive days; n = 10 • Untreated: n = 10	• Outcomes during 6 months of follow-up: • Physical disability (Hughes scale) • CMAP sum scores	• Hughes grade scores: significantly lower in IVIg (1.0–2.0) than TPE group (2.5) during the 6 months of follow-up. • Rapid recovery (improvement by ⩾2 grades in Hughes scale within 1 month): higher in the IVIg (60%) versus the TPE group (21%). • CMAP sum scores: trend to higher score in the IVIg than the TPE group, peaking at 6 months after treatment. • Return to normal activities: 90% in the IVIg versus 71% in the TPE group.	• IVIg may be more efficacious than TPE for treating anti-GM1-antibody positive GBS.
Ye et al.²¹	Clinical trial, China 3	Adults, n = 64	• TPE: 50 ml/kg; five sessions within 8–13 days; sixth session to achieve 250 ml/kg if first five inadequate; n = 32 • IVIg: one course of 0.4 g/kg daily for 5 days; second course for 3 days only if breathing difficulty persists or condition unstable; n = 32	• Clinical status and improvement at admission and 1 or 2 weeks after treatment (Hughes and MRC scores) • Efficacy: improvement of Hughes scale (⩾1 point) or MRC score (⩾1 point for ⩾2 limbs) • Laboratory indicators: immunoglobulin (IgG, IgA, IgM, C3, C4), fibrinogen concentrations and monocyte percentage	• Hughes and MRC scores: comparable between TPE and IVIg groups at 1 week (3.12 versus 3.37); significantly better in the TPE group (2.03 versus 2.56) at 2 weeks after treatment completion. • Curative effects: comparable between TPE and IVIg groups at 1 week (75% versus 53%); significantly higher for the TPE group at 2 weeks after treatment (96% versus 79%). • Laboratory indicators: significantly lower in the TPE group at both 1 and 2 weeks after treatment; more pronounced decrease between these timepoints in the TPE versus the IVIg group.	• Both TPE and IVIg have a high response as therapy and are reasonable therapeutic options for GBS. TPE, however, may have a better response than IVIg.
Ravasio et al.²²	Multicentre prospective study, Italy 4	Adults, n = 50	• TPE (centrifugation): average 1900 ml/session, average 5.3 sessions; n = 16 (five patients treated with second-line IVIg) • IVIg: 0.4 g/kg/day, 5 days; n = 20 (four patients treated with second-line TPE) • Steroids; n = 5 • Untreated; n = 9	• Functional grade (Hughes scale) and clinical improvement (change by ⩾1 grade) before and 1 month after treatment	• TPE first-line treatment: 56.3% improved, 18.8% recovered and 25% unchanged 1 month after treatment. • IVIg first-line treatment: 45% improved, 20% recovered and 35% unchanged at 1 month. • Disease progression: 0% in the TPE and 20% in the IVIg group (during the first few days of treatment).	• No significant differences in clinical outcome between treatment groups.
França et al.²³	Retrospective case review, Brazil 4	Adults >60 years, n = 18 AIDP: 8 (44%) • AMAN/AMSAN: 6 (33%) • MFS: 1 (5.6%) • Unclassified: 3 (16.7%)	• TPE (automated with continuous flow; central line): approximately 30 ml/kg of plasma, one to five sessions on alternate days; n = 12 • IVIg: a total dose of 2 g/kg for five consecutive days; n = 2 • TPE + IVIg; n = 1 • Untreated; n = 3	• Disability score at: admission, nadir and regular timepoints (7, 60 and 365 days) after therapy • Length of hospitalisation • Adverse event occurrence	• Disability scores at the different moments: clear and comparable improvement for both TPE (4.5 ± 0.5 at nadir, 2.1 ± 2.3 at 1 year after therapy) and IVIg (4.5 ± 0.7 at nadir, 2.0 ± 1.4 at 1 year after therapy). • Length of hospitalisation: longer in IVIg-treated individuals. • Adverse events were reported only in TPE-treated cases (central line–related issues, hypotension).	• Both IVIg and TPE were equally effective in elderly patients with GBS.
El-Bayoumi et al.²⁴	Prospective randomised study, Egypt 2	Children requiring MV, n = 41	• TPE: one volume daily, 5 days; n = 21 • IVIg: 0.4 g/kg/day, 5 days; n = 20	• Duration of MV (days) • Length of PICU stay (days) • Ability to walk unaided within 4 weeks after PICU discharge	• Shorter mean length of MV in the TPE (11.0) versus the IVIg group (13.0). • Length of PICU stay: shorter (no statistical significance) in the TPE versus the IVIg group. • 95.2% of children in the TPE versus 90% in the IVIg group could walk unaided within 4 weeks of discharge.	• Favourable response to both TPE and IVIg. • Using a weaning and extubation protocol, shorter MV period with TPE than IVIg.
Chaudhuri et al.²⁵	Case series–consecutive patients, India 4	Children and adults, n = 37 • AIDP: 21 • AMAN/AMSAN: 16	• TPE: 200–250 ml/kg plasma over five to eight sessions, daily; n = 18 • IVIg: 0.4 g/kg/day, 5 days; n = 19	• Functional motor deficits (Hughes scale) until 1 year after treatment • Muscle power (MRC sum score) • Length of hospital stay (days)	• Length of hospital stay significantly longer in the TPE compared with the IVIg group. • Clinical improvement at 30 days after treatment: comparable between the TPE and IVIg groups. • Complication rate comparable between the TPE and IVIg groups.	• TPE and IVIg therapy were equally effective. • TPE may be preferred in India (and other developing countries) because of low cost.
Kishore et al.²⁶	Retrospective analysis of consecutive patients, India 3	Adults, n = 90	• TPE (membrane-based apheresis): ~200–250 ml/kg plasma, five sessions (40–50 ml/kg per session) within 7–14 days; n = 61 • IVIg: 0.4 g/kg/day for 5 days; n = 29	• Disability grade score, immediately and 4 weeks after therapy • Duration of MV (days)	• TPE marginally superior to IVIg in improving the disability grading, including in MV patients.	• TPE and IVIg were effective in reducing the disability grade immediately post-therapy and after 4 weeks.
Cea et al.²⁷	Retrospective analysis of patient files, Chile 2	Adults, n = 29 (enrolled: n = 41) • AIDP: 27 • AMAN/AMSAN: 9 • ASAN: 2 • MFS: 3	• TPE: 40 ml/kg plasma, three to five sessions; n = 17 • IVIg: 2 g/kg/day, 4 or 5 days; n = 9 • TPE + IVIg: n = 3 • Untreated: n = 9	• Disability grade and clinical improvement (Hughes scale)	• Clinical improvement: more frequent in TPE (68.4%) than IVIg group (44.4%), but differences not statistically significant.	• Both therapies were efficacious in improving disability score.
Siddiqi et al.²⁸	Retrospective observational study (medical chart review), Canada 3	Children and adults, n = 82	• TPE: four or five to six sessions; n = 13 (adults) • IVIg: average 1.94 g/kg, average 3.9 days; n = 69 (children: n = 17; adults: n = 52)	• Proportion of patients with successful treatment (GDS or MV status) • Treatment relapse and failure	• 84.6% of TPE treatments were successful (MV in three cases) versus 91.3% of IVIg treatments were successful (MV in 13 cases).	• The efficacy of IVIg was comparable with that of TPE.
Bazaraa et al.²⁹	Prospective observational study, Egypt 3	Children with severe GBS and hospitalised at PICU, n = 40 • AIDP: 10 AMAN/ AMSAN: 30	• TPE: five single-volume sessions on alternate days; n = 33 • IVIg: 0.4 g/kg/day, for 5 days; n = 7	• Response to treatment (improvement in Hughes disability score) • Failure to respond to treatment: failure to extubate a patient on MV/residual bulbar manifestations/severe residual motor affection (Hughes score of 4 or 5)	• Initial disability score of 5 (MV required): 82% in the TPE versus 57% in the IVIg group. • Clinical improvement: lower in the TPE (33%) versus the IVIg group (71%). • Treatment failure: more frequent in the TPE (64%) versus the IVIg group (14%). • Early deaths: less frequent in the TPE (3%) than the IVIg group (14%).	• Critically ill children with severe GBS had guarded response to initial therapy with TPE or IVIg. • No adverse events were observed after TPE treatment. • Study was insufficiently powered to conclude if one treatment is better than the other.
Bondi et al.³⁰	Retrospective cohort study, Israel 3	Adults, n = 70 AIDP: 44	• TPE (treatment procedure not specified); n = 39 • IVIg (treatment procedure not specified); n = 31	• Clinical improvement at admission and discharge, evaluated by several scores: composite FIM (cognitive and motor), GDS and MRFS • Length of stay in ICU and length of rehabilitation (days)	• FIM scores: no significant differences between TPE and IVIg treatment. • All patients: significant improvement in functioning between admission and discharge.	• Similar efficacy of TPE and IVIg, regardless of the scoring system used.
Urtasun et al.³¹	Retrospective case review, Spain 4	Adults, n = 6	• TPE: three to six sessions; n = 4 • IVIg: 0.4 g/kg/day over 5 days; n = 2	• Length of hospitalisation (days) • Neurological disability score (muscle strength and vital capacity) during 90 days of follow-up	• Hospitalisation: longer in TPE (40–210) than IVIg group (28–30). • Neurological disability score: higher in TPE than IVIg at 30 (3–5 versus 2), 60 (2–5 versus 1) and 90 days (1–4 versus 0–1) of follow-up.	• IVIg may be an alternative to TPE for treatment of severe GBS.
Vajsar et al.³²	Case series, Canada 4	Children, n = 15	• TPE (continuous flow): 200–250 ml/kg/day, four or five sessions; n = 5 • IVIg: 1 g/kg/day, 2 days; n = 10	• Functional status (physical disability score) • Duration of severe disability (improvement by 2 grades) (days) • Duration of MV (days)	• Duration of severe disability: longer in the TPE (48.0 ± 23.4) versus the IVIg group (17.0 ± 12.4). • MV duration: longer in the TPE (25) than in the IVIg group (15).	• IVIg was more efficacious than TPE in speeding up GBS recovery. • IVIg is not recommended for patients with renal dysfunction.
El-Beleidy et al.³³	Prospective cohort study, Egypt 3	Children, n = 47: • AIDP: 20 (43%) (5 AB+) • AMAN: 23 (49%) (16 AB+) • Unclassified: 4 (8%) (1 AB+)	• TPE: 40–60 ml/kg per session, five sessions on five consecutive days, if failed to respond to IVIg; n = 31 • IVIg: 0.4 g/kg/day for five consecutive days; n = 47	• Overall disability sum score: motor disability grading in extremities • MV weaning/ discontinuation	• Unresponsive to IVIg: 95% of AB+ versus 41% AB–. • Both AB+ and AB– responded well to subsequent TPE.	• The patients with AMAN subtype, of whom most were AB+, responded well to TPE after IVIg failure.
Charra et al.³⁴	Prospective, single centre, nonrandomised study, Morocco 3	Adults requiring MV, n = 41 • AIDP: 33 • AMAN: 8	• TPE: four sessions, over 10–14 days; n = 20 • IVIg: 0.4 g/kg/day for 5 days; n = 21	• Length of ICU stay (days) • Start of mobility rehabilitation (days) • Weaning off MV (days)	• Length of stay in ICU: longer in the TPE (52.4 ± 5.3) than in the IVIg group (38.2 ± 7.6). • Weaning of MV: longer in the TPE (38.5) than in the IVIg group (18.7). • Start mobility rehabilitation: later in the TPE (18.7) than in the IVIg group (10.4).	• IVIg may be more beneficial than TPE for treating GBS patients requiring MV.
Saad et al.³⁵	Retrospective study of medical records, Egypt 3	Children, n = 62 • AIDP: 40 (65%) • AMAN/ AMSAN: 14 (23%) • MFS: 4 (6%) • Unclassified: 4 (6%)	• TPE (central line): 200–250 ml/kg plasma, over 7–10 days; n = 32 • IVIg: 0.4 g/kg/day for five consecutive days; n = 30	• Length of hospitalisation (days) • Required MV • Outcome at 6 months: complete recovery/residual deficit/death	• Mean duration of hospitalisation (days) was shorter in the TPE (15.7) than in the IVIg group (29.4). • MV: 4 (13%) in the TPE versus 20 (67%) in the IVIg group. • Complete recovery: 19 (59%) in the TPE versus 7 (23%) in the IVIg group. • Residual deficit was lower in the TPE (22%) than in the IVIg group (57%). • Death: 6 (19%) in the TPE and 6 (20%) in the IVIg group.	• TPE may reduce the duration of hospital stay and hasten the recovery of children with GBS.

AB + and AB–, presence or absence of antiganglioside antibodies; AIDP, acute inflammatory demyelinating polyradiculoneuropathy; AMAN, acute motor axonal neuropathy; AMSAN, acute motor and sensory axonal neuropathy; ASAN, acute sensory axonal neuropathy; BW: body weight; C3 and C4, complement factors 3 and 4; CEBM, Centre for Evidence-Based Medicine; CMAP, compound muscle action potentials; FIM, Functional Independence Measure; GBS, Guillain–Barré syndrome; GDS, GBS disability score; GM1, ganglioside; IgA, IgG, IgM, immunoglobulin A, G, M; IVIg, intravenous immunoglobulin; LoE, level of evidence; MFS, Miller-Fisher syndrome; MRC, Medical Research Council; MRFS, Montebello Rehabilitation Factor Score; MV, mechanical ventilation; n, number of patients in a given category; (P)ICU, (paediatric) intensive care unit; RCT, randomised controlled trial; TPE, therapeutic plasma exchange.

Six records were (randomised) clinical trials mostly conducted in the 1990s, with numbers of randomised and treated patients ranging from 24 to 379.^12,17–21 Four of these trials concluded that IVIg and TPE were comparably efficacious in treating GBS (LoE: 2–4).^12,18,19,21 In addition, van der Meché et al.¹⁷ concluded that IVIg is at least comparably effective or may even be superior to TPE for treating GBS in patients aged 4 or more. Notably, the median treatment delay (TD; time elapsed between symptom onset and start of treatment) in the same study was longer in the TPE than IVIg group [1.0 (0–4) versus 0.0 (0–3) days]. This study also reported significantly lower complication rates in IVIg- versus TPE-treated patients,¹⁷ which was observed as a nonsignificant trend in two other trials.^12,18 Specifically, Diener et al.¹² observed comparable rates of clinical AEs, but lower rates of treatment-related AEs with IVIg than TPE treatment. This is the only trial that distinguished between disease- versus treatment-specific AEs in GBS patients.¹² While the above-mentioned four trials^12,17–19 included unselected patients or those with moderate or severe GBS, Kuwabara et al.²⁰ examined treatment efficacies in patients positive for antiganglioside antibodies and reported significantly faster clinical recovery after IVIg than TPE treatment. Of note, this trial included 24 patients, and the authors stated that efficacy of TPE may have been decreased by double plasma filtration.²⁰

Nine other studies (four prospective, five retrospective) concluded that TPE and IVIg were both viable options or that TPE was slightly more efficacious than IVIg for GBS treatment (LoE: 2–4).^22–30 Similar to the clinical trials above, most of these studies agreed that the safety of TPE and IVIg was acceptable. One study indicated that TPE may be more frequently associated with complications.²³ Of note, the route of TPE administration (central versus peripheral), the TPE device and the IVIg producer were often insufficiently detailed. Ravasio et al.²² reported having used IVIg from six different producers in their study, but did not state which product was used in which patients. Also, only three studies specified the TD duration.^22,26,29 The median TD was comparable between IVIg and TPE groups, except in Ravasio et al.²² [9.0 (4–65) versus 5.5 (2–40) days].²²

The remaining five studies on GBS came to different conclusions on how IVIg and TPE compare with each other in treating GBS. Two studies in adults reported shorter hospital stay periods and higher recovery rates with IVIg,^31,34 although the study of Urtasun et al.³¹ included only six patients.

Two studies in Egyptian children favoured TPE: 95% of IVIg-refractory patients positive for antiganglioside antibodies (AB+) responded to TPE treatment,³³ and the second reported shorter hospitalisation and higher recovery rates in patients treated with TPE versus IVIg.³⁵ Of note, the predominant GBS subtype differed: 49% of patients had AMAN³³ and 65% had AIDP.³⁵ A Canadian study, however, reported longer mechanical ventilation and duration of two-grade clinical improvement after TPE treatment.³² The paediatric studies differed in their sizes (e.g. 15 patients in the Canadian study compared with the 47 and 62 in the Egyptian studies), geographical regions and selection of patient populations (e.g. AB-status stratification, time elapsed between symptom onset and enrolment).

The grade of recommendation for superiority of TPE over IVIg and vice versa was low (grade C). The grade of recommendation for TPE being equally efficacious to IVIg was moderate (grade B).

MG

MG is an autoimmune disorder in which antibodies, mainly against the nicotinic acetylcholine receptor (AChR) or muscle-specific tyrosine kinase (MuSK), cause defective neuromuscular signal transduction.^7,36,37 The suboptimal signal transduction results in fluctuating and transient muscle weakness and fatiguability. MuSK-antibody-associated disease (MuSK MG) occurs in about 5% of MG patients and is phenotypically distinct from MG with AChR antibodies (AChR MG).⁷ While AChR MG is often accompanied by asymmetry of ptosis, divergent gaze and distal limb weakness, the dominant feature of MuSK MG is bulbar weakness.⁷ Thymic comorbidities (e.g. hyperplasia and thymoma) are common in patients with MG and result in referral of selected patients for thymectomy.^36–38

Some patients experience a myasthenia crisis – an acute episode of respiratory muscle weakness ending in respiratory failure.^6,38 Myasthenia crisis may be precipitated by systemic physiological changes (e.g. medication, pregnancy or illness).^6,38

Out of 11 comparative studies (Table 2), 8 concluded that TPE and IVIg were equally efficacious (LoE: 2–4)^{40,42–47,49} and 2 concluded that TPE was more efficacious than IVIg (LoE: 2 and 3).^39,48 One study favoured IVIg for the shorter patient intubation times necessary during treatment of MG (LoE: 2).⁴¹ The grade of recommendation for either therapy was moderate (grade B).

Table 2.

TPE and IVIg in treatment of myasthenia gravis.

Study	Design, country, CEBM LoE	Patient population	Treatment	Measure of treatment efficacy	Clinical outcome	Author conclusion
Liu et al.³⁹	Randomised clinical trial, China 2	Adults with late-onset MG, n = 40	• TPE (double-filtration): 40 ml/kg, three sessions, every 1–2 days; n = 15 • IVIg: 0.4 g/kg/day, 5 days; n = 15 • IA; n = 10	• Clinical efficacy evaluated by QMGS (⩽25% ineffective, >25% clinical remission and ⩾50% clinically effective) • Time to clinical remission (days) • Antibody titres of Titin-ab, AChR-ab and PrsmR-ab • Length of hospital stay (days)	• Serum anti-titin and anti-AChR antibody concentrations decreased more in the TPE and IA groups than in the IVIg group (titin decreased by: 62.2%±6.3% in TPE, 58.3% ± 2.9% in IA, 19.9% ± 3.1% in IVIg; AChR by: 62.6% ± 0.9% in TPE, 55.1% ± 3.2% in IA, 28.9% ± 3.8% in IVIg). • Serum IgG concentration significantly decreased in the TPE (by 55%) and IA groups, but remained unchanged in the IVIg group; serum albumin concentration decreased in the TPE but not the IA or IVIg groups. • Improvement in MG severity score was significantly higher in the TPE and IA groups (66.8% and 42.4%) versus the IVIg group (23.8%). • QMGS decreased more in the TPE and IA groups than in the IVIg group (60.8% ± 3.5%, 42.4% ± 4.2% versus 23.8% ± 3.7%). • Fourteen days after the beginning of treatment, the clinical efficacy of the double-filtration TPE and IA groups was higher than that of the IVIg group (80% versus 70% versus 40%). • Mean time to clinical remission was significantly shorter in the TPE and IA groups (6.7 and 5.4) versus the IVIg group (8.4). • Mean length of hospital stay: shorter in the TPE and IA groups (12.8 and 13.5) versus the IVIg group (16.0).	• TPE and IA showed better short-term clinical effectiveness than IVIg in late-onset MG patients.
Katzberg et al.⁴⁰	Randomised controlled trial, Canada 2	Adults with severe MG, n = 105	• TPE: not specified; n = 42 • IVIg: 2 g/kg split over 2 days; n = 63	• Disease severity evaluated by QMGS • Response to treatment: improvement by ⩾3.5 points in QMGS	• QMGS was higher in the IVIg responders versus nonresponders at baseline, but equivalent in the TPE group. • Responders and nonresponders were similar in terms of baseline characteristics. • Baseline QMGS was found to be the only independent predictor of response to immunomodulation.	• Compared with patients with a milder MG form, patients with more severe disease as measured by the QMGS appear to respond better to treatment with IVIg and TPE. • MG disease severity is a predictive factor of response to IVIg but not TPE.
Alipour-Faz et al.⁴¹	Single-centre randomised clinical trial, Iran 2	Adults with generalised MG referred for thymectomy, n = 24	• TPE: 1 litre plasma, five sessions, every other day; n = 12 • IVIg: 1 g/kg/day, 2 days; n = 12	• Duration of surgery (h) • Duration of intubation (h) • Duration of ICU and hospital stay (days)	• Rate of intubation: higher in the TPE (7/12) versus the IVIg group (2/12). • Intubation period significantly longer in the TPE (range: 2–216) versus the IVIg group (2–22). • ICU length of stay and hospital length of stay: not significantly different between the TPE and IVIg groups.	• TPE and IVIg both effective for optimisation patients prethymectomy. • IVIg may be more appropriate than TPE as intubation time is shortened.
Rønager et al.⁴²	Randomised controlled crossover study, Denmark 3	Adults with moderate to severe generalised MG, n = 12	• TPE: five sessions, every other day; n = 12 (first-line: n = 4; second-line: n = 8) • IVIg: 0.4 g/kg/day, 5 days; n = 12 (first-line: n = 8; second-line: n = 4) • Crossover treatment regimens (no details available)	• MG severity evaluated by QMGS • Treatment efficacy: change in QMGS. • Anti-AChR titre, IgG and IgM antibody concentrations.	• Improvement in QMGS: significantly greater for TPE (0.23) versus IVIg (0.10) at 1 week after treatment; comparable at 16 weeks after treatment. • Significant improvement in QMGS 4 weeks after TPE and IVIg treatment. • Antibody concentrations: significantly lower in the TPE (decrease by 79%) versus the IVIg group (no change) at 1 week, but comparable at 16 weeks after treatment. • IgG concentration: decreased at week 1, 4 and 8 after TPE (by 83%, 18% and 11%) and increased at week 1 and 4 after IVIg (by 159% and 43%).	• TPE and IVIg have comparable clinically efficacy in patients with chronic moderate to severe MG. • The onset of improvement was more rapid after TPE (within 1 week) than IVIg (benefit for MG crisis).
Jensen and Bril⁴³	Retrospective chart review, Canada 4	Adults with generalised MG referred for thymectomy, n = 47	• TPE (no details available); n = 9 • IVIg (no details available); n = 9	• MG grade based on the Osserman classification. • Duration of hospital stay (days).	• Postoperative (post-thymectomy) mean Osserman grades were similar between the TPE and IVIg groups. • 100% patients in the TPE group reported some improvement. • A nonsignificant trend of shorter hospital stay with IVIg (2.2 days) compared with TPE (3.3 days). • 44.5% patients in the IVIg group reported no subjective benefit.	• IVIg was a reasonable alternative to TPE in patients with MG requiring prethymectomy optimisation of strength.
Mandawat et al.⁴⁴	Cross-sectional retrospective cohort analysis of hospital discharge information, the United States 2	Adults with generalised MG or MG crisis, n = 1606	• TPE (no details available); n = 1266 • IVIg (no details available); n = 340 • TPE + IVIg; n = 32 (not analysed)	• Hospital mortality. • Inpatient complications. • Hospital length of stay (days).	• Mortality and complication rate: comparable between the TPE and IVIg groups for generalised MG; higher in the TPE versus the IVIg group in MG crisis (5.7% versus 0.6% and 30.1% versus 14.8%). • Length of hospital stay: significantly longer in the TPE versus the IVIg group for generalised MG (6 versus 4) and MG crisis (10 versus 5).	• IVIg was comparable with TPE for clinical outcome. • IVIg was superior to TPE for economical outcome.
Barth et al.⁴⁵	Randomised evaluator-masked study, Canada 2	Adults with moderate to severe generalised MG, n = 84	• TPE: one plasma volume, five sessions, every second day; n = 43 • IVIg: 1 g/kg/day, 2 days; n = 41	• MG severity evaluated by QMGS, MGFA clinical classification of MG score and subjective patient VAS before and up to 28 days after treatment. • Response to treatment: ⩾3.5 units change in QMGS at different timepoints. • AChR antibody status.	• TPE and IVIg resulted in similar change in MG severity score from baseline to days 14, 21 and 28. • QMGS unit decrease at day 14: 3.2 ± 4.1 in the IVIg and 4.7 ± 4.9 in the TPE group. QMGS reduction persisted until day 28. • Responders: 51% in the IVIg versus 57% in the TPE group. • Similar proportion of patients improved in TPE versus the IVIg groups (65% versus 69%) at day 14; 2% worsened on TPE versus 18% on IVIg at the same timepoint.	• TPE and IVIg had comparable efficacy in treatment of moderate to severe MG.
Liew et al.⁴⁶	Retrospective review of medical records, the United States 3	Children and adolescents with generalised MG, n = 27	• TPE: three to five sessions, typically over 1–2 weeks; n = 7 • IVIg: 1–2 g/kg/day, every 2–3 weeks; n = 10 • TPE + IVIg; n = 10	• Response to therapy, evaluated by motor examination and subjective patient opinion	• 100% of patients treated with TPE only responded to therapy versus 50% of patients treated with IVIg only. • No additional benefit in terms of response in children receiving both treatments.	• TPE and IVIg have high response rates in juvenile MG as maintenance therapy. • TPE may have a more consistent response rate than IVIg.
Gajdos et al.⁴⁷	Randomised clinical trial, France 2	Adults with MG exacerbation, n = 87	• TPE: 1.5 volumes [1 volume = (1 − haematocrit value) × 70 kg; cell separation or membrane filtration], three sessions; n = 41 • IVIg: 0.4 g/kg/day for 3 (n = 23) or 5 days (n = 23); n = 46	• Variation in MG muscular score within the first 15 days • Treatment response: improvement by ⩾20 points on the MG muscular scale • Time to treatment response (days)	• Responders: 26/41 (63.4%) in the TPE versus 22/46 (47.8%) in the IVIg group. • Similar improvement in MG muscular score 15 days after treatment with TPE (66%) and IVIg for 3 days (61%). • Anti-AChR antibody levels decreased in 63% (in the TPE group) and 65% (in the IVIg group) of seropositive patients (mean anti-AChR titre variations: –13.8% and –16.8%). • Fewer participants with ⩾1 adverse event in the IVIg (2%; 1 patient) versus the TPE group (20%; 14 events in 8 patients).	• Equal efficacy after TPE or IVIg for treatment of MG exacerbation. • Shorter IVIg infusion (3 days) was more beneficial than longer infusion (5 days).
Qureshi et al.⁴⁸	Retrospective multicentre chart review, the United States 3	Adults with MG crisis, n = 51	• TPE: 25–45 ml/kg, five or six sessions, on alternate days; n = 27 (28 episodes) • IVIg: 0.4 g/kg/day, 5 days; n = 24 (26 episodes)	• Disease severity, evaluated by QMGS. • MV status. • Functional outcome (death/MV requirement/tracheostomy without MV/spontaneous breathing with functional impairment/ spontaneous breathing without functional impairment). • Requirement for second treatment. • Length of ICU and hospital stay (days).	• Improvement in mean QMGS from baseline to 1 week after treatment initiation: slightly better in the TPE (6.9–11.1) versus the IVIg group (7.5–10.3). • TPE more efficacious than IVIg at 2 weeks for extubation and at 1 month for functional improvement. • Mean length of ICU and hospital stay: longer in the TPE (17.4 and 25.7) than in the IVIg group (14.0 and 17.7). • Complication rate: higher in the TPE [13/28 (46.4%) episodes] versus the IVIg [5/26 (19.2%)] group.	• TPE is superior to IVIg for treatment of MG involving respiratory failure. • TPE is associated with more complications. • IVIg is a good alternative to TPE for at-risk patients.
Pittayanon et al.⁴⁹	Retrospective single centre database review, Thailand 4	Adults with MG crisis, n = 26 (33 episodes)	• TPE: three to five sessions, alternate days; n = 21 episodes • IVIg: 0.4 g/kg/day, 5 days; n = 9 episodes	• Duration of intubation (days). • Length of hospital stay (days). • Complications during hospitalisation. • Discharge status.	• No significant difference in duration of intubation or length of hospital stay in the TPE versus the IVIg group.	• TPE and IVIg were equally effective and safe to rescue MG crisis. • A trend towards fewer complications with IVIg.

ab, antibody; AChR, acetylcholine receptor; CEBM, Centre for Evidence-Based Medicine; IA, immune adsorption; ICU, intensive care unit; IgG and IgM, immunoglobulin G and M; IVIg, intravenous immunoglobulin; LoE, level of evidence; MG, myasthenia gravis; MGFA, Myasthenia Gravis Foundation of America; MV, mechanical ventilation; n, number of participants in each category; PrsmR, presynaptic membrane receptor; QMGS, quantitative MG score; TPE, therapeutic plasma exchange; VAS, visual analogue scale.

Generalised MG (including moderate and severe forms)

Out of three clinical trials in adults with MG (LoE: 2),^39–41 two found that TPE and IVIg treatments comparably reduced disease severity [measured by, e.g., the quantitative MG score (QMGS)].^40,41 Katzberg et al.⁴⁰ described that baseline QMGS was predictive of IVIg but not TPE response, although they stated that neither disease duration nor underlying medical conditions were decisive factors for the use of either therapy. Alipour-Faz et al.⁴¹ suggested that IVIg may be advantageous to reduce the period of intubation in MG patients undergoing thymectomy. Disease duration, however, was significantly longer in the TPE- versus the IVIg-treated group in this study. Liu et al.³⁹ reported on higher short-term clinical effectiveness and more efficient removal of immune mediators (anti-titin, -AChR and serum IgG antibodies) by TPE than IVIg in adults with late-onset MG. TPE-treated patients required relatively less respiratory support and were hospitalised for a shorter time than IVIg-treated patients, but AEs were relatively more frequent in the TPE- than the IVIg-treated group.³⁹

In agreement with the mentioned clinical trials of Alipour-Faz et al.⁴¹ and Katzberg et al.,⁴⁰ four other studies in patients with generalised MG found TPE and IVIg to be comparably efficacious treatment options.^42–45 Only one study by Barth et al.⁴⁵ relied on the Osserman classification as a disease severity measure, but still concluded that IVIg was a similarly efficacious alternative to TPE in patients with MG who were referred for thymectomy. This study in 84 patients found that IVIg and TPE similarly improved electrophysiological characteristics 2 weeks after treatment, but this improvement persisted until day 28 only for IVIg.⁴⁵ The authors nevertheless concluded that both immunomodulatory therapies were viable options to treat moderate to severe MG.⁴⁵ Mandawat et al.⁴⁴ found that mortality and complication rates were comparable, but that length of hospital stay was longer with TPE. One retrospective study in children and adolescents reported on a 100% response rate to TPE (7/7 patients) versus 50% for IVIg (5/10 patients) treatment.⁴⁶ While response rates to TPE treatment were found to be more consistent, both TPE and IVIg were comparably efficacious as maintenance therapy in juvenile MG (LoE: 3).⁴⁶

MG exacerbation and myasthenia crisis

Treatment responses to TPE and IVIg were comparable with respect to severity score, intubation and hospital stay length in adults.^47,49

A single clinical trial by Gajdos et al.⁴⁷ reported on similar response rates and muscular score improvement with TPE and IVIg in adults with MG exacerbation. Similar to Liu et al.,³⁹ this trial found that AEs were relatively more frequent after TPE than IVIg treatment. Comparable percentages of TPE- and IVIg-treated patients, however, had a decrease in anti-AChR antibodies.⁴⁷ Of note, the study by Gajdos et al.⁴⁷ was limited due to unmasked analyses and biased towards IVIg benefits through failure to administer a full course of TPE.

In agreement with the above, two other studies found that, while TPE was efficacious in treating patients with MG exacerbations or myasthenia crises, it was also associated with relatively more AEs and complications.^48,49 Qureshi et al.⁴⁸ reported on more favourable ventilatory status (percent of extubated patients at 2 weeks) and higher proportion of patients without disability at 1 month after TPE versus IVIg treatment. The same study observed higher efficacy with respect to severity score improvement, but longer (although not statistically significant) intensive care unit and hospital stay times for patients treated with TPE versus IVIg (LoE: 3, grade: B).⁴⁸ While the studies of Pittayanon et al. and Qureshi et al. support the use of TPE as an effective treatment of MG exacerbation and myasthenic crisis, the low numbers of recruited patients limit data interpretation and are insufficient to support superiority of TPE. The efficacy of TPE and IVIg treatments was comparable for juvenile myasthenia crisis (LoE: 3).⁴⁶

Additional considerations regarding studies in MG

Of the 11 studies in patients with MG, five included information on autoantibody serological status.^{39,40,45–47} All five reported comparable seropositivity rates for anti-AChR antibodies (range: 61–80%)^{39,40,45–47} and three on anti-MuSK antibody seropositivity (range: 4–18%).^40,45,46 Aside from Katzberg et al.,⁴⁰ none of these studies stratified clinical outcomes and treatment efficacy by autoantibody serological status; effects of IVIg and TPE on MuSK MG were investigated only in noncomparative studies. The degree of MG-associated autoantibody reduction was similar or favoured TPE where this was studied in comparative trials on MG.^39,42,45,47 Patients with MuSK MG have been consistently shown to respond well to TPE,^50,51 but less so to IVIg. Anti-MuSK antibodies mainly belong to the immunoglobulin G4 (IgG4) subclass and are therefore not complement fixing⁵² and are directly pathogenic on a passive transfer mouse model.⁵³ Their pathogenicity probably underpins the benefit of direct removal by TPE. Acute treatment of MuSK MG crisis has TPE as the mainstay⁷ and transfer to a hospital unit with an available TPE device is warranted in this MG variant.

Chronic inflammatory demyelinating polyradiculoneuropathy

Chronic inflammatory demyelinating polyradiculoneuropathy (CIDP) is an acquired, mostly treatable, immune-mediated disorder of the peripheral nervous system with slow symptom progression, exceeding 8 weeks, followed by a progressive or relapsing course.⁵⁴ CIDP is usually treated with corticosteroids, TPE, IVIg or a combination thereof; chronic treatment with immunotherapies is a valuable alternative in patients with certain CIDP subtypes refractory to corticosteroid treatment.⁵⁴ CIDP has been classified into different subtypes falling under typical and atypical CIDP.⁵⁴

The present literature search identified three records, one clinical trial in adult patients (LoE: 3)⁵⁵ and two paediatric studies (LoE: 4)^56,57 comparing IVIg and TPE efficacy in treating CIDP (Table 3). Treatment responses and outcomes were measured by changes in functional disability: modified Rankin scale (MRS)^56,57 or a combination of several scores.⁵⁵

Table 3.

TPE and IVIg in treatment of chronic inflammatory demyelinating polyradiculoneuropathy.

Study	Study type, country, CEBM LoE	Patient population	Treatment	Measure of treatment efficacy	Clinical outcome	Author conclusion
Dyck et al.⁵⁵	Observer-blind clinical trial, the United States 3	Adults, n = 19	• TPE: two sessions/week for 3 weeks, then one session/week for 3 weeks; n = 9 in first-line • IVIg: 0.4 g/kg/week for 3 weeks, then 0.2 g/kg/week for 3 weeks; n = 10 in first-line. • Crossover design with washout in between (no details available); TPE + IVIg; n = 13.	• NDS and NDS weakness subset scores • CMAP of ulnar, median and peroneal nerves • SNAP sum score of median and sural nerves • VDT of the great toe	Significant improvement with TPE and IVIg in: • NDS • NDS weakness subset score • CMAP.	• TPE and IVIg were equally efficacious for improving neuropathic endpoints. • Effects were short-lived, worsened during washout period. • IVIg preferred due to convenience.
McMillan et al.⁵⁷	Cohort data combined with case series, the United States 4	Children, n = 135	• TPE (no details available); n = 7. • IVIg (no details available); n = 25.	• Clinical response (based on combined MRS): good (minimal or no functional impairment), partial (some degree of improvement, but therapy addition or modification necessary) and no response (no improvement or deterioration).	• First-line treatment response: 14% in the TPE versus 79% in the IVIg group.	• IVIg was superior to TPE for treatment of childhood CIDP.
Kim et al.⁵⁶	Retrospective review of patient records, Korea 4	Children (6 monophasic, 8 polyphasic), n = 14	• TPE: seven sessions over 2 weeks equals one cycle; n = 9. • IVIg: 2 g/kg per cycle over 2–5 consecutive days; n = 14.	• Clinical response to treatment and treatment outcomes (using the MRS). • Clinical response: good (improvement of ⩾2 points on MRS, no need to change or adjust treatment), partial (improvement of ⩾1 point on MRS, need to change or adjust treatment) and no response (no change or worsening of MRS). • Long-term treatment outcomes: complete remission (MRS 0–1, no treatment for ⩾6 months), partial remission (MRS 2–3, need for treatment within last 6 months for functionality) and no remission (MRS 4–5 and no improvement).	• TPE more effective than IVIg in cases with monophasic disease course. • TPE and IVIg comparable efficacy in cases with polyphasic disease course.	• TPE was mostly effective in childhood cases of CIDP refractory to IVIg or corticosteroids.

CEBM, Centre for Evidence-Based Medicine; CIDP, chronic inflammatory demyelinating polyradiculoneuropathy; CMAP, compound muscle action potential; IVIg, intravenous immunoglobulin; LoE, level of evidence; MRS, modified Rankin scale; n, number of patients in a given category; NDS, neuropathy disability scale; TPE, therapeutic plasma exchange; VDT, vibratory detection threshold; SNAP, summated sensory nerve action potential.

The single clinical trial in adult patients with CIDP found no differences between TPE and IVIg treatment in terms of neurological disability and weakness scores or summated muscle motor potentials.⁵⁵ The same trial reported that therapeutic effects were short-lived for both therapies and waned during the washout period.⁵⁵ The authors concluded that IVIg may be preferred due to its availability and ease of administration. The time of study conduct (1994), low number of enrolled participants and the complex and insufficiently clarified study design must be taken into account when interpreting the results of this study.⁵⁵

Data from the paediatric studies were inconsistent, with McMillan et al.⁵⁷ reporting a higher proportion of responders among IVIg- than TPE-treated patients; Kim et al.⁵⁶ found that TPE and IVIg were similarly efficacious, but TPE was likely a more favoured option for IVIg- and corticosteroid-refractory CIDP. These studies also reported different clinical experiences with TPE administration. McMillan et al.⁵⁷ claimed that TPE catheter placement is technically challenging and leads to complications (e.g. infection, hypotension); Kim et al.⁵⁶ described TPE administration without significant complications.

The studies in patients with CIDP found that TPE and IVIg were both efficacious.^55–57 IVIg may, however, be preferable for its availability and relatively fewer AEs.⁵⁵ The AE occurrence should be interpreted cautiously as centrally administered TPE is more frequently associated with complications, and only one⁵⁶ of the cited studies specified the route of TPE administration.

Owing to low-level evidence (grade of recommendation for equal efficacy of TPE and IVIg: C) and limited number of participants and studies, it is difficult to conclude and recommend whether TPE or IVIg was superior for the treatment of CIDP. Previously, several rare variants of CIDP associated with IgG4 subclass autoantibodies to peripheral nerve proteins at the node and paranode (e.g. neurofascin) were found to respond poorly to IVIg.⁵⁸ There are no published data comparing IVIg with TPE in this group of CIDP patients.

Encephalitis

Several encephalitis types have been described so far. Autoimmune encephalitis (AIE) is a noninfectious, immune-mediated encephalitis, often characterised by antibodies against neuronal antigens and accompanied by various symptoms (behavioural and psychiatric changes, movement and autonomic disorders and seizures).⁵⁹ To date, at least 10 autoantibodies have been identified in different AIE subclasses, but a significant proportion of cases may have an immune origin even when no antibodies can be detected.⁵⁹

The AIE Alliance Clinicians Network has recently proposed AIE classification into four subclasses based on serological parameters: antibodies to intracellular antigens, antibodies for clinically relevant surface and other antigens, antibodies with surface antigens of low clinical relevance and seronegative AIE.⁵⁹ The 68 experts from 17 countries included in this multinational survey identified IVIg and TPE as possible immunotherapeutic options for both acute AIE and long-term management of AIE-associated symptoms.^59,60 Of note, certain practicality issues may arise with the acute treatment of AIE in noncooperative (e.g. agitated) patients.⁵⁹ These patients will rarely tolerate a TPE procedure, making central venous access difficult unless intubated and sedated. They, however, may tolerate a peripheral venous line suitable for IVIg, as observed in clinical practice.

This review included one study on unspecified encephalitis,⁶¹ which reported on longer hospitalisation times and lower discharge possibility after treatment with TPE (Table 4).⁶¹ The mean age of patients treated with TPE was higher than of those treated with IVIg, however.

Table 4.

TPE and IVIg in treatment of encephalitis.

Study	Study type, country, CEBM LoE	Patient population	Treatment	Measure of treatment efficacy	Clinical outcome	Author conclusion
Gomez et al.⁶¹	Retrospective case control study, the United States 4	Adults with unspecified encephalitis; n = 347	• TPE; n = 188 • IVIg; n = 159	• Length of hospital stay (days) • Discharge or transfer likelihood	• Patients were on average older in the TPE (44 years) than in the IVIg group (34 years). • Mean length of hospital stay was longer in the TPE (31.1) than in the IVIg group (22.9). • The TPE group was 37.0% less likely to be discharged at home. • The TPE group was 79.6% more likely to be transferred to another facility for further treatment.	• IVIg patients were more likely to be discharged than TPE patients.

CEBM, Centre for Evidence-Based Medicine; IVIg, intravenous immunoglobulin; LoE, level of evidence; n, number of patients in each category; TPE, therapeutic plasma exchange.

NMOSDs

NMOSD is an autoimmune disease of the central nervous system, primarily affecting the spinal cord and optic nerve, characterised by the presence of antibodies against aquaporin 4 (AQP4). The clinical criteria for diagnosing NMOSD and distinguishing it from multiple sclerosis (MS) have been updated over the years.^62,63 Aside from AQP4-specific IgG antibodies, most patients with NMOSD are usually characterised by longitudinally extensive transverse myelitis, severe optic neuritis and area postrema syndrome, and often have recurrent disease.⁶² Although rare, NMOSD has been described across different ethnicities worldwide.⁶²

The present systematic search did not identify systematic studies that directly compare TPE and IVIg in NMOSD treatment (Table 5). Nevertheless, an exception was made for this group of disorders due to their relevance to the field and emerging clinical evidence. The selection of included publications in Table 5 was based on expert opinion.

Table 5.

TPE and IVIg in treatment of neuromyelitis optica spectrum disorders.

Study	Study type, country, CEBM LoE	Patient population	Treatment	Measure of treatment efficacy	Clinical outcome	Author conclusion
Songthammawat et al.⁶⁴	Prospective randomised controlled study, Thailand 3	Adult patients with severe acute NMO attack; n = 10	• IVMP add-on TPE: IVMP 1 g/day for seven consecutive days and then centrifugal TPE (on alternate days for five sessions, 1 litre of plasma exchanged with similar volume of 5% albumin) if patient nonresponsive; n = 5 • IVMP + TPE: IVMP treatment as above combined with centrifugal TPE from the start; n = 5	• Follow-up: until 6 months after discharge • Different scoring systems used: EDSS, MRC, mRS • Safety evaluation	• Baseline IVMP add-on TPE versus IVMP + TPE: younger age at onset and at attack, lower frequency of TM attacks and better EDSS score prior to attack. • EDSS score at nadir: 6.7 in IVMP add-on TPE versus 7.2 in IVMP + TPE. • EDSS scores similar at attack and at discharge in both groups, but significantly better at 6 months follow-up (no differences between groups). • Shift from severe to less severe disability: somewhat slower and less frequent in IVMP add-on TPE versus IVMP + TPE. • Results with mRS similar to EDSS. • Three relapses during 6-month follow-up: two in IVMP add-on TPE and one in IVMP + TPE. • Fifteen adverse events reported by eight patients.	• Both IVMP add-on TPE and IVMP + TPE are beneficial in decreasing disability outcomes of NMOSD patients with severe acute attacks. • An observed trend favours IVMP + TPE.
Elsone et al.⁶⁵	Retrospective case review, the United Kingdom 4	Adults and children with NMO(SD) relapses, n = 10 (11 relapses)	• IVIg: 2 g/kg daily over 5 days	• EDSS score • Occurrence of adverse events	• Clinical improvement in 5 out of 11 NMO relapses (45.5%). • Median EDSS change: 7.0 (4–9) pretreatment to 6.5 (3–9) at follow-up. • Adverse events: possibly related to IVIg in two patients.	• The benefit of IVIg in treatment of NMO relapses needs to be explored further. • It is currently not possible to distinguish effects of IVIg over preceding therapies.
Bonnan et al.⁸	Retrospective review of medical records, France 4	Adults and children with monophasic or relapsing NMOSD, n = 60	• TPE: initiated as soon as possible after diagnosis; central line on a continuous flow device (5% albumin as replacement) for 5 days	• Clinical status: EDSS and visual acuity (using the Snellen scale) scores • Δ clinical score = late clinical score (⩾6 months after treatment) – baseline clinical score • Clinical response: complete (baseline-level), good (66.0–100.0%) and poor (0.0–33.3%)	• Two-thirds of patients were IgG seropositive. • Treatment delay was short: 34.0% and 63.0% received TPE within 5 and 10 days. • Probability of complete clinical improvement: decreased with increased treatment delay (50.0% with 0–1 days to 5.0% with >20 days). • Adverse events: 8.6% of patients.	• TPE is effective and recommended for all NMOSD attacks, regardless of IgG status.
Ipe et al.⁶⁶	Retrospective chart review, the United States 4	Paediatric (⩽21 years) and adult patients, n = 114	• TPE: single treatment course with 10 consecutive procedures; 1.0–1.25 plasma volumes exchanged in 89.9% treatment series; 82.5% with central access, 4.9% with peripheral access and 3.9% with femoral line; n = 114	• Clinical status improvement, defined as a score 3–5	• All patients had an improved clinical status with TPE; maximal improvement by the fourth or fifth TPE treatment. • Score in paediatric patients on average 1.59 points lower than for adults. • Adverse events reported by 9.9% of treated patients.	• TPE improved clinical scores, especially in adults. • TPE was well tolerated.
Li et al.⁶⁷	Retrospective case review, China 4	Adults with NMOSD, presenting with acute attacks; n = 198 (243 attacks)	• HD-S: 500 mg of intravenous glucocorticoid/day for three consecutive days, then reduced to 60 mg oral glucocorticoid; n = 136 (153 attacks) • IVIg: 0.4 g/kg for five consecutive days; n = 13 (14 attacks) • HD-S + IVIg: co-administration as per individual regimens; n = 66 (69 attacks) • TPE: treatment not specified; n = 7 (7 attacks)	• EDSS score • Remission outcomes: good (66–100%), moderate (33–65%) and poor (0–32%) • Clinical outcomes	• 67.6% of patients positive for anti-AQP4 antibodies. • EDSS at onset was significantly higher in HD-S + IVIg (6.5) versus IVIg group (5.0). • Both in the overall patient cohort and the anti-AQP4 seropositive patients, HD-S and HD-S + IVIg led to fewer poor remissions (30.8–34.3% and 44.7–47.8%) than IVIg alone (63.3–66.7%). • For the HD-S + IVIg group: the proportion of attacks with good and moderate remission status was higher when IVIg was administered first (IVIg + HD-S; 33.0 and 42.0%) than vice versa (HD-S + IVIg: 25.0% and 22.0%). • In uni- and multi-variate analyses, lower EDSS scores at onset were predictive of better treatment outcomes; IVIg + HD-S combination was superior to HD-S alone.	• Treatment with IVIg alone may be partially effective in ameliorating acute episodes of NMOSD but not superior to HD-S alone. • Adding HD-S to IVIg is beneficial to relieve onset of NMOSD with high EDSS score.

Δ, difference; AQP4, aquaporin 4; CEBM, Centre for Evidence-Based Medicine; EDSS, Expanded Disability Status Scale; HD-S, high-dose intravenous steroids; Ig, immunoglobulin; IVIg, intravenous immunoglobulin; IVMP, intravenous methylprednisolone; LoE, level of evidence; n, number of patients included in each category; NMO(-SD), neuromyelitis optica (spectrum disorders); MRC, Medical Research Council; mRS, modified Rankin scale; TM, transverse myelitis; TPE, therapeutic plasma exchange.

It is widely accepted that TPE is a mainstay treatment for NMO relapses,⁶⁸ as corroborated by three studies included in this SLR.^8,64,66 The only included prospective controlled study, Songthammawat et al.⁶⁴ (LoE: 3), concluded that TPE was more efficacious when administered concurrently with intravenous steroids than as an add-on therapy in patients with NMOSD attacks. Bonnan et al.⁸ found that TPE was efficacious for treating NMOSD attacks. The authors highlighted the importance of initiating TPE early in the attack. Longer TDs were associated with worse outcomes; administering TPE on day 0 translated into 50% probability to regain complete improvement, which was significantly higher compared with TPE administered after day 20 (1–5% probability for complete improvement).

In contrast, only a few case series have reported the possible benefit of IVIg in NMOSD. For the sake of balance and discussion, we have included selected studies on IVIg (Table 5).

Elsone et al.⁶⁵ observed a modest clinical benefit of IVIg in adult and paediatric patients with NMO relapses. Approximately half (5/11; 45.5%) of events improved with IVIg, with improvement of the Expanded Disability Status Scale score nonsignificantly.⁶⁵ Another study found that IVIg therapy conveyed moderate benefit to patients with acute NMOSD attacks and that combining IVIg with high-dose intravenous steroids helped to relieve onset of severe NMOSD.⁶⁷

Other conditions

This group includes several conditions for which only one study comparing TPE and IVIg treatment efficacies (per condition) was identified in the literature search (Supplementary Table 2). The conditions included MS,⁶⁹ acute immune-mediated polyradiculoneuropathy (AIMP; general condition that also includes GBS),⁷⁰ postinfectious polyradiculoneuropathy⁷¹ and Sydenham’s chorea⁷² (Supplementary Table 2). IVIg and TPE showed comparable efficacies for all these conditions.

Discussion

This SLR aimed to analyse available evidence from 1990 to 2021 on TPE and IVIg therapeutic efficacy for both peripheral and central nervous system autoimmune neurological disorders. GBS was the best represented (20 studies), followed by MG (11), NMOSD (5), CIDP (3) and encephalitis (1). Most studies (19) were retrospective, while 13 were clinical trials. After screening 44 comparative studies, we conclude that TPE and IVIg are efficacious in treating neurological conditions of largely autoimmune origin; no clear superiority of one treatment over the other has been clearly demonstrated thus far.

Until 2001, most evidence on therapeutic efficacy of TPE and IVIg was derived from studies on GBS, CIDP and MG in which both treatments have been established as mainstay, first-line therapeutic modalities.^3,11,73 With some exceptions,^17,19 however, these studies were limited by their retrospective approach and designed to test noninferiority of IVIg (the newer treatment) versus TPE. The noninferiority design warrants caution as it is only moderately accurate, subject to bias and different from superiority designs that require greater statistical power and typically greater participant numbers.⁷⁴ Other limitations of the mentioned studies were insufficiently defined and variable study designs,^20,32,47 variable treatment outcome grading and small sample sizes.^{12,22,31,32,42,55}

More recent studies offered further evidence on rare autoimmune neurological disorders, including more recently recognised conditions, such as AIE, NMOSD and MuSK MG. Although no comparative studies were identified in this review, both TPE and IVIg have led to clinical improvement in patients with AIE^59,60 and, with a modest benefit of IVIg, NMOSD.^8,64–67 Of note, in absence of larger comparator studies, IVIg benefit in NMOSD may have been underestimated by the expert choice of the included studies. Some other relevant diseases of autoimmune origin, such as the stiff-person syndrome,⁷⁵ were not included in the present review due to the lack of eligible studies.

Identification of immune mediators involved in disease pathogenesis (e.g. AQP4 in NMOSD,⁶² N-methyl-D-aspartate receptor in encephalitis)⁷⁶ has improved diseases understanding and treatment efficacy. Nevertheless, many limitations from previous studies persist in the recent literature. The disease rarity precludes conduct of larger randomised controlled trials (RCTs). Treatment decisions are thus often based on anecdotal evidence, clinical experience and findings extrapolated from related disease areas. In addition, treatment protocols are nonstandardised and insufficiently described, limiting reproducibility. Finally, while the clinical benefit of IVIg and TPE treatments has been correlated to disease pathogenesis in some conditions (e.g. AQP4 in NMOSD, MuSK in MG), our understanding of the topic is limited for two reasons. First, the disease pathogenesis is incompletely understood at the molecular and physiological levels. Second, when immune mediators are known, the clinical benefit of a given treatment does not always correlate with seropositivity for specific immune mediators. Better designed studies are therefore needed.

Despite limitations, the studies included in this SLR discuss clinical efficacy of IVIg and TPE in light of the trade-offs between clinical benefit and complications. Many studies reported higher complication rates and longer hospital stays with TPE versus IVIg treatment.^{17,25,31,34,44,47,48,61} Comparing complications between TPE and IVIg is fraught with difficulties because of the different techniques of TPE administration (e.g. central versus peripheral venous access) which are often not specified in the studies. TPE procedure has advanced, and the use of peripheral line TPE has reduced the number of both catheter- (e.g. deep vein thrombosis) and exchange-related (e.g. acute renal failure) AEs.⁷⁷ Newer TPE devices can accurately track fluid exchange, thereby minimising the risk of AEs.⁷⁷ Unlike membrane-mediated TPE with a central line access associated with many AEs, centrifugal TPE with a peripheral line access was shown to be acceptably safe and well tolerated.⁷⁷ TPE efficacy may have also been lowered by the suboptimal choice of replacement fluid. For instance, Kuwabara et al.²⁰ admitted that using double-filtered plasma may have negatively affected their treated patients with GBS. Central line placement may also delay TPE, significantly reducing its therapeutic efficacy,⁸ so peripheral line TPE is advantageous as quicker to administer.

The main argument in favour of IVIg was its simplicity of use compared with the TPE protocols. Administering IVIg does not necessitate trained personnel and is preferred for patients in whom TPE is difficult to administer, such as psychotic or combative patients.⁵⁹ IVIg, however, does not directly remove immune mediators, although it may accelerate catabolism of autoantibodies or modulate downstream effects on complement and cellular antibody-dependent mechanisms. In IgG4-autoantibody-mediated diseases, such as MuSK MG^7,50,51 or neurofascin-antibody CIDP,⁵⁸ that may be refractory or poorly responsive to IVIg, TPE remains the therapy of choice. Despite lack of comparative studies, the benefit of TPE is most prominent in NMOSD, in which it continues to be a mainstay therapeutic approach.⁸ Also, some patients (e.g. patients with GBS seropositive for antiganglioside antibodies)³³ may be unresponsive or overly sensitive to IVIg or may have contraindications (e.g. renal failure).⁹ Finally, the rapid clinical improvement observed after TPE favoured this treatment for the management of myasthenia crisis.⁴²

The main strengths of this review are its broad literature scope spanning 30 years and the variety of analysed neurological disorders. One significant limitation is the low quality of available evidence in many conditions, due to low numbers of participants in studies of autoimmune neurological disorders, patient factors influencing the choice of therapeutic modality and economic considerations. In addition, not all studies directly compared IVIg and TPE, and the benefit of other treatments (e.g. corticosteroids) was not considered in this analysis. Many included publications called for further and larger studies to validate the clinical benefit of TPE and IVIg treatment for these diseases, in agreement with opinions of expert authors of this review. Other limitations include changes in disease classification (e.g. separation of NMOSD from MS)⁶² and strict exclusion criteria at title/abstract screening, possibly biasing the final selection of studies. Together with low LoE of the included studies, it was not possible to more thoroughly analyse the results discussed here.

The present review offers comprehensive and contemporary information on optimal TPE versus IVIg treatment for autoimmune neurological conditions. While the LoE may not be suitable for regulatory decision-making,¹⁴ it may be satisfactory to guide physicians who need to treat patients with these disorders in real-world practice.

Conclusion

TPE and IVIg are both viable treatment options for several autoimmune neurological disorders: GBS, MG (including myasthenic crisis), CIDP and encephalitis. For NMOSD, the evidence from the targeted literature review favoured TPE over IVIg, although some studies point to a potential benefit of IVIg. The choice of the best therapeutic modality should be based on the available clinical resources and tailored according to individual patient conditions and needs. Future studies on autoimmune neurological disorders should ideally be prospective, with well-defined and standardised protocols and consensus evaluation criteria. We further propose the inclusion of quality-of-life assessments as a more meaningful way to evaluate the long-term benefit of different treatments to the patients.

Supplemental Material

sj-docx-1-tan-10.1177_17562864231154306 – Supplemental material for Comparison of IVIg and TPE efficacy in the treatment of neurological disorders: a systematic literature review

Supplemental material, sj-docx-1-tan-10.1177_17562864231154306 for Comparison of IVIg and TPE efficacy in the treatment of neurological disorders: a systematic literature review by Ashwin A. Pinto, Jerome De Seze, Anu Jacob, Stephen Reddel, Anna Yudina and Kevin Tan in Therapeutic Advances in Neurological Disorders

Footnotes

Acknowledgements

The authors acknowledge Irena Zurnic Bönisch (Akkodis Belgium c/o Terumo Blood and Cell Technologies) for writing support and Sophie Timmery (Akkodis Belgium c/o Terumo Blood and Cell Technologies) for editing and coordination support. The authors also acknowledge the contribution of Rania Hassanein and Aicha Bah to the literature search and paper review. The authors would also like to acknowledge their colleagues in infusion and TPE teams and their commitment to clinical care. The listed authors have authorised the submission of their manuscript via a third party and approved any statements or declarations included in the manuscript.

Declarations

ORCID iD

Ashwin A. Pinto

Supplemental material

Supplemental material for this article is available online.

References

McDaneld

Fields

Bourdette

, et al. Immunomodulatory therapies in neurologic critical care. Neurocrit Care 2010; 12: 132–143.

Patwa

Chaudhry

Katzberg

, et al. Evidence-based guideline: intravenous immunoglobulin in the treatment of neuromuscular disorders: report of the Therapeutics and Technology Assessment Subcommittee of the American Academy of Neurology. Neurology 2012; 78: 1009–1015.

United Kingdom Blood Services. Handbook of transfusion medicine. 5th ed. TSO, UK, 2013.

Dalakas

MC.

Update on intravenous immunoglobulin in neurology: modulating neuro-autoimmunity, evolving factors on efficacy and dosing and challenges on stopping chronic IVIg therapy. Neurotherapeutics 2021; 18: 2397–2418.

Cortese

Chaudhry

, et al. Evidence-based guideline update: plasmapheresis in neurologic disorders: report of the Therapeutics and Technology Assessment Subcommittee of the American Academy of Neurology. Neurology 2011; 76: 294–300.

Juel

Massey

JM.

Autoimmune myasthenia gravis: recommendations for treatment and immunologic modulation. Curr Treat Options Neurol 2005; 7: 3–14.

Reddel

Morsch

Phillips

WD.

Clinical and scientific aspects of muscle-specific tyrosine kinase-related myasthenia gravis. Curr Opin Neurol 2014; 27: 558–565.

Bonnan

Valentino

Debeugny

, et al. Short delay to initiate plasma exchange is the strongest predictor of outcome in severe attacks of NMO spectrum disorders. J Neurol Neurosurg Psychiatry 2018; 89: 346–351.

Guo

Tian

Wang

, et al. Adverse effects of immunoglobulin therapy. Front Immunol 2018; 9: 1299.

10.

Hartmann

Klein

HG.

Supply and demand for plasma–derived medicinal products – a critical reassessment amid the COVID-19 pandemic. Transfusion 2020; 60: 2748–2752.

11.

Sanders

Wolfe

Benatar

, et al. International consensus guidance for management of myasthenia gravis: executive summary. Neurology 2016; 87: 419–425.

12.

Diener

Haupt

Kloss

, et al. A preliminary, randomized, multicenter study comparing intravenous immunoglobulin, plasma exchange, and immune adsorption in Guillain-Barré syndrome. Eur Neurol 2001; 46: 107–109.

13.

GRADE Working Group. GRADE handbook (eds Schünemann

Broz.ek

Guyatt

, et al.). 2013, https://gdt.gradepro.org/app/handbook/handbook.html (accessed 10 February 2023).

14.

OCEBM Levels of Evidence Working Group. The Oxford 2011 levels of evidence, https://www.cebm.ox.ac.uk/resources/levels-of-evidence/ocebm-levels-of-evidence (2011, accessed 1 November 2021).

15.

Kuwabara

Guillain-Barré syndrome: epidemiology, pathophysiology and management. Drugs 2004; 64: 597–610.

16.

Willison

Jacobs

van Doorn

PA.

Guillain-Barré syndrome. Lancet 2016; 388: 717–727.

17.

van der Meché

Schmitz

. A randomized trial comparing intravenous immune globulin and plasma exchange in Guillain-Barré syndrome. Dutch Guillain-Barré Study Group. N Engl J Med 1992; 326: 1123–1129.

18.

Bril

Ilse

Pearce

, et al. Pilot trial of immunoglobulin versus plasma exchange in patients with Guillain-Barré syndrome. Neurology 1996; 46: 100–103.

19.

Plasma Exchange/Sandoglobulin Guillain-Barré Syndrome Trial Group. Randomised trial of plasma exchange, intravenous immunoglobulin, and combined treatments in Guillain-Barré syndrome. Plasma Exchange/Sandoglobulin Guillain-Barré Syndrome Trial Group. Lancet 1997; 349: 225–230.

20.

Kuwabara

Mori

Ogawara

, et al. Intravenous immunoglobulin therapy for Guillain-Barré syndrome with IgG anti-GM1 antibody. Muscle Nerve 2001; 24: 54–58.

21.

YJ.

Comparison on therapeutic effect of plasma exchange and intravenous immunoglobulin for Guillain–Barré syndrome. Transfus Med 2015; 25: 79–84.

22.

Ravasio

Pasquinelli

Currò Dossi

, et al. High dose intravenous immune globulins and plasma exchange in Guillain-Barré syndrome. Ital J Neurol Sci 1995; 16: 487–492.

23.

França

MCJ

Deus-Silva

de Castro

, et al. Guillain-Barré syndrome in the elderly: clinical, electrophysiological, therapeutic and outcome features. Arq Neuropsiquiatr 2005; 63: 772–775.

24.

El-Bayoumi

El-Refaey

Abdelkader

, et al. Comparison of intravenous immunoglobulin and plasma exchange in treatment of mechanically ventilated children with Guillain Barré syndrome: a randomized study. Crit Care 2011; 15: R164.

25.

Chaudhuri

Alladi

Mridula

, et al. Clinical outcome of Guillain-Barré syndrome with various treatment methods and cost effectiveness: a study from tertiary care center in South India: Yashoda GBS Registry. Neurol Asia 2014; 19: 263–270.

26.

Kishore

Vijayabhaskar

Vishnu Vardhan

, et al. Management of Guillain-Barré syndrome with plasmapheresis or immunoglobulin: our experience from a tertiary care institute in South India. Ren Fail 2014; 36: 732–736.

27.

Cea

Jara

Quevedo

[Clinical features of Guillain-Barré syndrome in 41 patients admitted to a public hospital]. Rev Med Chil 2015; 143: 183–189.

28.

Siddiqi

Courtney

Hanna

, et al. Gamunex(R) in Guillain-Barré syndrome: a postmarketing, retrospective, observational study. Can J Neurol Sci 2017; 44: 711–717.

29.

Bazaraa

Rady

Mohamed

, et al. Initial response and outcome of critically ill children with Guillain Barré Syndrome. Front Pediatr 2019; 7: 378.

30.

Bondi

Engel-Haber

Wolff

, et al. Functional outcomes following inpatient rehabilitation of Guillain-Barré syndrome patients: intravenous immunoglobulins versus plasma exchange. Neurorehabilitation 2021; 48: 543–551.

31.

Urtasun

Lopez

Munain

Carrera

, et al. High-dose intravenous immune globulin in the management of severe Guillain-Barré syndrome. Ann Pharmacother 1992; 26: 32–33.

32.

Vajsar

Sloane

Wood

, et al. Plasmapheresis vs intravenous immunoglobulin treatment in childhood Guillain-Barré syndrome. Arch Pediatr Adolesc Med 1994; 148: 1210–1212.

33.

El-Beleidy

Khattab

El-Sherbini

, et al. Antiganglioside antibodies determine the clinical severity and predict response to therapy in Egyptian children with Guillain-Barré syndrome. Pediatr Polska 2013; 88: 224–229.

34.

Charra

Hachimi

Benslama

, et al. Intravenous immunoglobulin vs plasma exchange in treatment of mechanically ventilated adults with Guillain-Barré syndrome. Pan Afr Med J 2014; 18: 35.

35.

Saad

Mohamad

Abd El-Hamed

, et al. A comparison between plasmapheresis and intravenous immunoglobulin in children with Guillain-Barré syndrome in Upper Egypt. Ther Adv Neurol Disord 2016; 9: 3–8.

36.

Conti-Fine

Milani

Kaminski

HJ.

Myasthenia gravis: past, present, and future. J Clin Invest 2006; 116: 2843–2854.

37.

Masi

O’Connor

KC.

Novel pathophysiological insights in autoimmune myasthenia gravis. Curr Opin Neurol 2022; 35: 586–596.

38.

Souto

Lima

Campos

, et al. Myasthenia gravis: state of the art and new therapeutic strategies. J Neuroimmunol 2019; 337: 577080.

39.

Liu

Wang

Xue

, et al. Comparing the autoantibody levels and clinical efficacy of double filtration plasmapheresis, immunoadsorption, and intravenous immunoglobulin for the treatment of late-onset myasthenia gravis. Ther Apher Dial 2010; 14: 153–160.

40.

Katzberg

Barnett

Bril

Predictors of response to immunomodulation in patients with myasthenia gravis. Muscle Nerve 2012; 45: 648–652.

41.

Alipour-Faz

Shojaei

Peyvandi

, et al. A comparison between IVIG and plasma exchange as preparations before thymectomy in myasthenia gravis patients. Acta Neurol Belg 2017; 117: 245–249.

42.

Rønager

Ravnborg

Hermansen

, et al. Immunoglobulin treatment versus plasma exchange in patients with chronic moderate to severe myasthenia gravis. Artif Organs 2001; 25: 967–973.

43.

Jensen

Bril

A comparison of the effectiveness of intravenous immunoglobulin and plasma exchange as preoperative therapy of myasthenia gravis. J Clin Neuromuscul Dis 2008; 9: 352–355.

44.

Mandawat

Kaminski

Cutter

, et al. Comparative analysis of therapeutic options used for myasthenia gravis. Ann Neurol 2010; 68: 797–805.

45.

Barth

Nabavi Nouri

, et al. Comparison of IVIg and PLEX in patients with myasthenia gravis. Neurology 2011; 76: 2017–2023.

46.

Liew

Powell

Sloan

, et al. Comparison of plasmapheresis and intravenous immunoglobulin as maintenance therapies for juvenile myasthenia gravis. JAMA Neurol 2014; 71: 575–580.

47.

Gajdos

Chevret

Clair

, et al. Clinical trial of plasma exchange and high-dose intravenous immunoglobulin in myasthenia gravis. Myasthenia Gravis Clinical Study Group. Ann Neurol 1997; 41: 789–796.

48.

Qureshi

Choudhry

Akbar

, et al. Plasma exchange versus intravenous immunoglobulin treatment in myasthenic crisis. Neurology 1999; 52: 629–632.

49.

Pittayanon

Treeprasertsuk

Phanthumchinda

Plasmapheresis or intravenous immunoglobulin for myasthenia gravis crisis in King Chulalongkorn Memorial Hospital. J Med Assoc Thai 2009; 92: 478–482.

50.

Evoli

Tonali

Padua

, et al. Clinical correlates with anti-MuSK antibodies in generalized seronegative myasthenia gravis. Brain 2003; 126: 2304–2311.

51.

Sanders

El-Salem

Massey

, et al. Clinical aspects of MuSK antibody positive seronegative MG. Neurology 2003; 60: 1978–1980.

52.

McConville

Farrugia

Beeson

, et al. Detection and characterization of MuSK antibodies in seronegative myasthenia gravis. Ann Neurol 2004; 55: 580–584.

53.

Cole

Reddel

Gervásio

, et al. Anti-MuSK patient antibodies disrupt the mouse neuromuscular junction. Ann Neurol 2008; 63: 782–789.

54.

Koike

Katsuno

Pathophysiology of chronic inflammatory demyelinating polyneuropathy: insights into classification and therapeutic strategy. Neurol Ther 2020; 9: 213–227.

55.

Dyck

Litchy

Kratz

, et al. A plasma exchange versus immune globulin infusion trial in chronic inflammatory demyelinating polyradiculoneuropathy. Ann Neurol 1994; 36: 838–845.

56.

Kim

Shim

Choi

, et al. Chronic inflammatory demyelinating polyneuropathy: plasmapheresis or cyclosporine can be good treatment options in refractory cases. Neuromuscul Disord 2019; 29: 684–692.

57.

McMillan

Kang

Jones

, et al. Childhood chronic inflammatory demyelinating polyradiculoneuropathy: combined analysis of a large cohort and eleven published series. Neuromuscul Disord 2013; 23: 103–111.

58.

Querol

Nogales-Gadea

Rojas-Garcia

, et al. Neurofascin IgG4 antibodies in CIDP associate with disabling tremor and poor response to IVIg. Neurology 2014; 82: 879–886.

59.

Abboud

Probasco

Irani

, et al. Autoimmune encephalitis: proposed best practice recommendations for diagnosis and acute management. J Neurol Neurosurg Psychiatry 2021; 92: 757–768.

60.

Abboud

Probasco

Irani

, et al. Autoimmune encephalitis: proposed recommendations for symptomatic and long-term management. J Neurol Neurosurg Psychiatry 2021; 92: 897–907.

61.

Gomez

Walker

Lacci

, et al. Need for ICD code for suspected autoimmune encephalitis as demonstrated by retrospective ten year analysis of unspecified autoimmune encephalitis treated with immunoglobulin or plasma exchange (P2.2-010). Neurology 2019; 92: P2.2-010.

62.

Pandit

Asgari

Apiwattanakul

, et al. Demographic and clinical features of neuromyelitis optica: a review. Mult Scler 2015; 21: 845–853.

63.

Wingerchuk

Banwell

Bennett

, et al. International consensus diagnostic criteria for neuromyelitis optica spectrum disorders. Neurology 2015; 85: 177–189.

64.

Songthammawat

Srisupa-Olan

Siritho

, et al. A pilot study comparing treatments for severe attacks of neuromyelitis optica spectrum disorders: intravenous methylprednisolone (IVMP) with add-on plasma exchange (PLEX) versus simultaneous ivmp and PLEX. Mult Scler Relat Disord 2020; 38: 101506.

65.

Elsone

Panicker

Mutch

, et al. Role of intravenous immunoglobulin in the treatment of acute relapses of neuromyelitis optica: experience in 10 patients. Mult Scler 2014; 20: 501–504.

66.

Ipe

Raval

Fernando

, et al. Therapeutic plasma exchange for neuromyelitis optica spectrum disorder: a multicenter retrospective study by the ASFA neurologic diseases subcommittee. J Clin Apher 2020; 35: 25–32.

67.

Tian

Fan

, et al. Intravenous immunoglobulin for acute attacks in neuromyelitis optica spectrum disorders (NMOSD). Mult Scler Relat Disord 2020; 44: 102325.

68.

Siritho

Nopsopon

Pongpirul

Therapeutic plasma exchange vs conventional treatment with intravenous high dose steroid for neuromyelitis optica spectrum disorders (NMOSD): a systematic review and meta-analysis. J Neurol 2021; 268: 4549–4562.

69.

Nazareth

Datar

TC.

Treatment effectiveness for resolution of multiple sclerosis relapse in a US health plan population. Neurol Ther 2019; 8: 383–395.

70.

Barnes

Herkes

. Acute immune-mediated polyneuropathies: the Royal North Shore Hospital experience (P1.155). Neurology 2018; 90: P1.155.

71.

Hosokawa

Hamaguchi

Tomioka

, et al. Comparative study of efficacy of plasma exchange versus intravenous gammaglobulin treatment on acute postinfectious polyradiculoneuropathy: a preliminary report. Ther Apher 1998; 2: 288–291.

72.

Garvey

Snider

Leitman

, et al. Treatment of Sydenham’s chorea with intravenous immunoglobulin, plasma exchange, or prednisone. J Child Neurol 2005; 20: 424–429.

73.

Van den Bergh

PYK

van Doorn

Hadden

RDM

, et al. European Academy of Neurology/Peripheral Nerve Society guideline on diagnosis and treatment of chronic inflammatory demyelinating polyradiculoneuropathy: report of a joint Task Force-Second revision. J Peripher Nerv Syst 2021; 26: 242–268.

74.

Head

Kaul

Bogers

, et al. Non-inferiority study design: lessons to be learned from cardiovascular trials. Eur Heart J 2012; 33: 1318–1324.

75.

Ortiz

Ghani

Morillo Cox

, et al. Stiff-person syndrome: a treatment update and new directions. Cureus 2020; 12: e11995.

76.

Shin

Lee

Park

, et al. Treatment strategies for autoimmune encephalitis. Ther Adv Neurol Disord 2017; 11: 1756285617722347.

77.

Guptill

Oakley

Kuchibhatla

, et al. A retrospective study of complications of therapeutic plasma exchange in myasthenia. Muscle Nerve 2013; 47: 170–176.

Supplementary Material

Please find the following supplemental material available below.

For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.

For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.

0.00 MB

0.03 MB