Efficacy and Safety of Plasma Exchange and Immunoadsorption for Acute Optic Neuritis in Patients with Demyelinating Diseases: A Systematic Review and Meta-Analysis

Summary

Background:

Optic neuritis (ON), a common manifestation of demyelinating disorders such as multiple sclerosis (MS), neuromyelitis optica spectrum disorder (NMOSD), and myelin oligodendrocyte glycoprotein antibody-associated disease (MOGAD), often leads to significant visual impairment. Although high-dose corticosteroids are a first-line therapy, many patients experience incomplete recovery. Apheresis therapies, including plasma exchange (PLEX) and immunoadsorption (IA), are being increasingly used as rescue treatments, but their efficacy and safety are debated.

Methods:

We conducted a systematic review and meta-analysis of studies evaluating PLEX and IA for acute ON in patients with demyelinating diseases. Four databases were searched up to January 2025. The included studies reported visual or neurological outcomes in patients with MS, NMOSD, MOGAD, or idiopathic ON. The primary outcomes were treatment response and visual acuity changes (logMAR). The study quality was assessed using RoB 2, ROBINS-I, or JBI tools. Random effects models were employed for the meta-analysis.

Results:

Twelve studies (800+ ON attacks) were analyzed. Both PLEX and IA demonstrated significant visual improvement, with a pooled mean logMAR reduction of -1.11 (95% CI: -1.70 to -0.53). Early intervention correlated with better outcomes. MOGAD-ON patients showed the highest response rates, while delayed treatment in NMOSD-ON patients was linked to poorer vision. Adverse events were rare and mild.

Conclusion:

PLEX and IA are effective and safe second-line therapies for steroid-refractory ON, particularly when initiated early. These findings support the prompt use of apheresis in eligible patients to optimize visual recovery.

Introduction

Optic neuritis (ON) is a condition characterized by a sudden or gradual decline in vision [1]. It is frequently associated with central nervous system (CNS) demyelinating disorders, particularly multiple sclerosis (MS), neuromyelitis optica spectrum disorder (NMOSD), and myelin oligodendrocyte glycoprotein antibody-associated disease (MOGAD) [1,2]. These disorders have similar clinical features, but they are distinguished by distinct immunological markers and pathophysiologies [3]. ON is frequently the first sign of MS, and in some cases, it manifests as a clinically isolated syndrome that can progress to confirmed MS—particularly in white people [4].

The MS is a chronic inflammatory condition triggered by autoimmune activity that causes demyelination, axonal injury, and progressive degeneration of the CNS [5]. The disease has a wide geographic distribution and a clear female predominance [5,6]. Its cause is complex immune interactions, including dysregulation of T cells, B cells, and autoantibodies targeting myelin proteins [6,7]. NMOSD and MOGAD are caused by similar immune-mediated mechanisms, but these conditions are distinguished by specific antibodies—anti-aquaporin-4 (AQP4) in NMOSD and anti-MOG in MOGAD—that are required for the accurate diagnosis and prediction of disease progression [1,3].

The ON associated with NMOSD is more severe, often resulting in profound visual loss and poor recovery, especially in patients who test positive for AQP4 antibodies [3]. In contrast, MOGAD-related ON is characterized by recurring episodes but responds better to treatment [8]. A subset of ON patients remain seronegative for both AQP4 and MOG antibodies, and these cases are typically referred to as idiopathic or double-seronegative ON [9]. Regardless of the cause, ON can be extremely disabling if not treated promptly and effectively.

The standard treatment for acute ON is high-dose intravenous methylprednisolone (IVMP), which reduces inflammation and accelerates recovery [1,2]. However, a significant proportion of patients remain unresponsive to corticosteroids, necessitating second-line treatments [10]. In such cases, apheresis therapies such as plasma exchange (PLEX) and immunoadsorption (IA) are common therapeutic options [4,10,11]. These methods work by removing circulating pathogenic antibodies and immune complexes, which may reduce inflammation and restore neurological function [4]. Although both PLEX and IA are used in clinical practice, international guidelines differ in their recommendations for use. For example, the American Society for Apheresis recommends PLEX as a secondary treatment option for acute demyelinating attacks, whereas the role of IA is less clear [11]. Some studies suggest that early initiation of apheresis, particularly in severe ON or steroid-resistant cases, may result in better outcomes, but the timing, modality, and patient selection criteria are still debated [4,12,13]. Despite growing clinical interest, research on the efficacy and safety of PLEX and immunoadsorption (IA) in treating ON across various demyelinating diseases is fragmented. This systematic review and meta-analysis aims to summarize current research on the use of these treatments in acute ON, with a focus on visual and neurological outcomes, safety profiles, and prognostic variables.

Methods

Database search strategy

We carried out a structured search across electronic databases such as MEDLINE (via PubMed), Scopus, Google Scholar, and Web of Science, covering all records up to January 2025.

Inclusion criteria for optic neuritis

Individuals diagnosed with acute optic neuritis (ON) owing to demyelinating conditions, such as multiple sclerosis (MS), neuromyelitis optica spectrum disorders (NMOSD), myelin oligodendrocyte glycoprotein antibody disease (MOGAD), or idiopathic ON.

Secondary outcomes

Additional outcomes considered were changes in the Expanded Disability Status Scale (EDSS), improvement, and any side effects.

Study types included

The analyses included both randomized controlled trials (RCTs) and observational cohort studies, whether prospective or retrospective.

Exclusion criteria

Articles were not included if they were case reports, case series, or literature reviews.

Title and abstract screening

The titles and abstracts of the retrieved articles were independently reviewed by two researchers to assess their potential eligibility.

Full-text review

The same reviewers then independently evaluated the full texts of studies deemed potentially suitable.

Discrepancy resolution

Any disagreements were resolved by mutual agreement or with input from a third reviewer.

Data extraction

Two reviewers separately used a predesigned Excel template to extract the study data.

Study characteristics extracted

The collected data covered study details such as the author, year of publication, study type, location, and sample size.

RCT quality assessment

The quality of randomized trials was appraised using the updated Cochrane RoB 2 tool [14].

Nonrandomized study bias tool

The ROBINS-I tool was utilized to assess bias in nonrandomized intervention studies [15].

Study evaluation process

Each study was independently evaluated by two reviewers, with differences settled by agreement or a third-party adjudicator.

Appraisal checklist

The JBI Critical Appraisal Checklist was used for evaluation [16].

Majority study review

Most of the included studies underwent analysis.

Meta-analysis method

A random-effects model meta-analysis was conducted to compute the mean difference in logMAR visual acuity before and after plasma exchange.

Bias visualization

The Robvis package in R software was employed to create visual graphs showing the risk of bias assessments across the studies.

Outcome data limitations

Only a few studies reported similar outcomes for the other variables.

Results

Study selection

A total of 665 records were discovered through database searches, and no additional records were found in the trial registries. After 110 duplicates were removed, 555 unique records remained, which were screened on the basis of titles and abstracts. Of these, 489 were rejected because they did not meet the inclusion requirements. Sixty-one full-text articles were retrieved and reviewed for eligibility. Eighteen studies were excluded because they were editorials or commentaries, while 31 were excluded because of irrelevant or insufficient outcome data. Finally, 12 studies met the inclusion criteria and were used in the qualitative and quantitative synthesis (see Supplementary Fig. 1) [17,18,19,20,21,22,23,24,25,26,27,28].

Study characteristics

The included studies were published between 2011 and 2023 and comprised a variety of randomized controlled trials (RCTs), retrospective cohort studies, and prospective studies. These studies included more than 800 patients with acute ON linked to demyelinating diseases such as MS, neuromyelitis optica spectrum disorder (NMOSD), MOGAD, and idiopathic ON. The sample size ranged from 11 to 395 ON attacks, and the majority of participants were middle-aged adults, with a greater proportion of female patients. PLEX, therapeutic plasma exchange (TPE), and immunoadsorption (IA) were investigated as interventions, either alone or in combination with intravenous methylprednisolone (IVMP). The comparisons varied across studies and included standard care, sham apheresis, and other active apheresis modalities (Table 1).

Table 1.

Characteristics of the included studies.

Study	Study design	Country	Sample size	Age Mean (SD)	Sex (N)	Outcomes measured	Intervention
Roesner et al., 2011	Retrospective case series study	Germany	23	Median age was 42 years (range: 24–73 years).	Male: 8 Female: 15	Visual acuity	Plasma exchange (PLEX)
Koziolek et al., 2012	Prospective study (single arm)	Germany	11	33.6 ± 10.5 years	Male: 4 Female: 7	Visual acuity, EDSS, visual evoked potentials, incapacity status scale, side effects.	Immunoadsorption (IA) using tryptophan-linked polyvinyl alcohol adsorber
Deschamps et al., 2016	Retrospective study	France	34 13 (MS-related ON), 11 (NMOSD-related ON), 10 (idiopathic ON)	Median age was 32 years (range: 17–63 years)	Male: 7 Female: 27	Visual acuity (VA) recovery, categorized as final VA ≥ 0.5 (good outcome) or < 0.5 (poor outcome)	Plasma exchange (PLEX)
Dorst et al., 2019	RCT	Germany	61	IA group: 40.1 (33.7–50.8) years PE group: 37.5 (30.1–41.0) years	Male: 20 Female: 41	Visual acuity, EDSS, response rates	Immunoadsorption (IA) using protein A adsorber vs. plasma exchange (PE)
Lipphardt et al., 2019	Retrospective cohort study	Germany	140 (127 with multiple sclerosis [MS], 13 with neuromyelitis optica [NMO])	Median age: 38 years (range: 11–72)	Male: 48 Female: 92	EDSS, visual acuity (VA), side effects	Immunoadsorption (IA) using protein A adsorber vs. plasma exchange (PE)
Brochet et al., 2020	Double-blind RCT	France	38 7 patients excluded from analysis	TPE group: 33.5 ± 9.2 years Sham-TPE group: 41.9 ± 9.6 years	Male: 10 Female: 28	EDSS, FS, CFS, VAS, proportion of patients with improvement.	Therapeutic plasma exchange (TPE) vs. Sham-TPE (sham plasma exchange)
Songthammawat et al., 2020	RCT	Thailand	11 1 patient excluded from analysis	37.9 ± 14.7	Male: 0 Female: 11	EDSS, MRC Scale, logMAR, mRS, adverse events.	IVMP + PLEX (simultaneous IVMP and plasma exchange)
Tan et al., 2020	Prospective cohort study	China	37	38.7 ± 14.8	Male: 11 Female: 26	Visual acuity (VA) recovery, categorized as final VA > 20/40 (good outcome) or ≤ 20/40 (poor outcome)	Plasma exchange (PLEX)
Li et al., 2021	Retrospective cohort study	China	61	IVMP+IA group: 41.45 ± 16.81 years. IVMP+PE group: 36.86 ± 13.27 years. IVMP alone: 46.82 ± 13.06 years.	Male: 15 Female: 46	Significant improvement in attack-related neurological deficits (3 M-SI). Changes in EDSS score, visual FSS, pyramidal FSS, Hauser ambulation index, and visual outcome scale score.	Immunoadsorption (IA) using protein A adsorber vs. Plasma Exchange (PE)
Zhang et al., 2021	Retrospective cohort study	China	112 patients	Median age was 33 years (range: 22–48 years)	Male: 13 Female: 99	Visual acuity (VOS), Expanded Disability Status Scale (EDSS), Annualized Relapse Rate (ARR)	Plasma exchange (PLEX)
Gonzalez et al., 2022	Retrospective cohort study	Colombia	83	40.9 (±13.7)	Male: 22 Female: 61	Visual acuity, EDSS, Keegan's criteria, side effects	Therapeutic plasma exchange (TPE) as rescue therapy after steroid treatment
Chen et al., 2023	Retrospective study	International study involving 11 centers across 6 countries (France, Israel, New Zealand, Thailand, UK, USA)	395 optic neuritis (ON) attacks from 317 patients	Median age was 37 years (range: 9–75 years)	Male: 92 Female: 225	Visual acuity (VA) recovery, stratified into complete (20/20), good (20/40 or better), medium (20/40 to 20/200), and poor (20/200 or worse) recovery	Plasma exchange (PLEX)

Note: EDSS: Expanded Disability Status Scale, MRC Scale: Medical Research Council Scale (for muscle strength), logMAR: Logarithm of the Minimum Angle of Resolution (visual acuity measurement), mRS: Modified Rankin Scale (neurological disability measure), IVMP: Intravenous Methylprednisolone, PLEX/PE: Plasma Exchange, TPE: Therapeutic Plasma Exchange, IA: Immunoadsorption, FS: Functional Score, CFS: Combined Functional Score, VAS: Visual Analog Scale, FSS: Functional Systems Score, ARR: Annualized Relapse Rate, VOS: Visual Outcome Scale, MS: Multiple Sclerosis, NMOSD: Neuromyelitis Optica Spectrum Disorder.

Visual outcomes

Most studies revealed that PLEX or immunoadsorption therapy resulted in significant improvements in visual acuity, especially when therapy was initiated early in the disease course (Table 2). For example, Roesner et al. (2011) reported that 70% of patients improved significantly after receiving PLEX therapy, whereas Koziolek et al. (2012) reported a significant increase in visual acuity after IA therapy. Similarly, Tan et al. (2020) reported that PLEX therapy caused significant visual recovery, particularly when initiated early in the disease course. In the largest cohort, Chen et al. (2023) reported that most patients experiencing ON attacks responded well to PLEX therapy, with a final median visual acuity of 20/25, and that patients with MOGAD-associated ON experienced the best results. Older age, severe baseline vision loss, and delayed treatment initiation were poor prognostic indicators. Randomized trials revealed that compared with PLEX therapy, IA therapy provided more sustained improvement, whereas TPE therapy outperformed the sham in terms of functional recovery. However, some studies, such as Songthammawat et al. (2020), reported no significant difference in logMAR improvements across intervention groups.

Table 2.

Summary of key findings and adverse events from the included studies.

Study	Key findings	Adverse Events (AEs)
Roesner et al., 2011	-70% of patients improved after PE therapy; 69% showed good/very good response. -Mean visual acuity improved from 16% → 45% (post-PE therapy) → 60% (follow-up). -No benefit was observed in 30% of patients.	- No serious AEs; mild hypofibrinogenemia and transient hypotension were reported.
Koziolek et al., 2012	-72.7% of patients responded (visual acuity improved from 0.15 → 0.89 at 6 months). -IA therapy removed MBP fragments and immunoglobulins. -sIL-2R decreased in responders.	- Mild AEs (hypotension, catheter issues); one case of jugular vein thrombosis (nonresponder) was reported.
Deschamps et al., 2016	-56% of patients achieved final VA ≥ 0.5. -Poor outcome was associated with prior ipsilateral ON. -No difference in response was observed between patients with MS and those with NMOSD.	- Well tolerated; minor AEs (hypotension, allergic reactions to albumin).
Dorst et al., 2019	-IA therapy was superior to PE therapy: MSFC improvement at 4 weeks was 0.385 (IA therapy) vs. 0.265 (PE therapy). -IA therapy produced delayed but sustained response; PE therapy produced an earlier but less durable improvement. -IgG reduction was greater after IA therapy.	- One case of DVT was reported per group. PE therapy resulted in more cases of infection; IA therapy resulted in more cases of thrombocytopenia/allergies.
Lipphardt et al., 2019	-Response rates were similar between IA (58.8%) and PE (59.4%, P = 0.65) therapies. -Predictors of response were younger age (P = 0.02) and earlier apheresis initiation (P = 0.01). -EDSS and VA improvements were sustained at 6 months (for patients with RRMS or CIS: EDSS 5 → 2.5; VA 0.15 → 0.65). -There was difference in efficacy between IA and PE therapies.	-Overall AE rate was 3.7% (29/784 sessions). -Mild AEs (rate of 2.2%; transient hypotension, nausea, shivering); moderate AEs (rate of 1.4%; catheter thrombosis/infection, arrhythmia); severe AEs (rate of 0.1%; one case of suspected air embolism (no sequelae) was reported in a patient receiving PE therapy). -AEs did not differ between IA and PE therapies (P = 0.96).
Brochet et al., 2020	-There was no difference in primary endpoint after TPE therapy (moderate improvement of 64.3% vs. 52.9% in sham, P = 0.72). -In the optic neuritis subgroup, TPE therapy resulted in significant improvement (50% of patients experienced full recovery vs. 0% in sham). -Combined functional score (CFS) improved at 1, 3, and 6 months (P<0.05). -EDSS trended toward improvement at 1 month (P = 0.06).	-Serious AEs (1 case of local hematoma (not severe) was reported in a patient receiving TPE therapy; 2 cases of hospitalizations and 2 of MS relapses were reported in patients receiving the sham); mild AEs (hypotension, nausea, shivering) -AEs did not differ between TPE therapy and the sham.
Songthammawat et al., 2020	-EDSS improvement was significant in both groups at 6 months (P = 0.05 for IVMP+PLEX therapy; P = 0.07 for IVMP add-on PLEX therapy). -IVMP+PLEX therapy resulted in a faster/better recovery (80% of patients experienced improvement after IVMP+PLEX therapy vs. 60% after add-on PLEX therapy). -Motor recovery (MRC) was greater in patients receiving IVMP+PLEX therapy (P = 0.05). -No significant visual acuity improvement was observed (logMAR P = 0.39).	- 15 AEs were reported in 8 patients: 1 case of arteriovenous fistula (surgery required), 6 of hypofibrinogenemia (asymptomatic), 3 of hypocalcemia (2 were symptomatic), 1 of DVT, 1 of steroid-induced hyperglycemia, 1 of transient hypotension.
Tan et al., 2020	-PE therapy improved VA (logMAR: 2.61 → 1.66; P < 0.001). -13.5% of patients achieved 20/20 vision; 45.9% improved from NLP to >20/800. -Early PE therapy initiation (shorter onset-to-PE window) predicted better outcomes.	- Mild AEs (hypotension, citrate toxicity); no severe reactions reported.
Li et al., 2021	-Rescue IA and PE therapies showed comparable efficacies (45.45% vs. 47.62% Improvement, respectively, in 3 patients with M-SI). -IA therapy tended to result in better outcomes in patients experiencing myelitis attacks (OR = 9.587, P = 0.055). -There was no correlation between IgG/AQP4-IgG reduction therapy and clinical improvement.	-AEs in patients receiving IA therapy included 1 case of hemolysis and 1 case of chills/fever, as well as transient hypoimmunoglobulinemia. -AEs in patients receiving PE therapy included allergic reactions (6 cases of rash, 1 of fever, 1 of eyelid edema) and 1 case of catheter infection.
Zhang et al., 2021	-LPE and TPE therapies had similar efficacies (76.5% vs. 83.3% disability Improvement, respectively; P = 0.392). -ARR decreased (LPE therapy: 1.5 → 0.6; TPE therapy: 1.6 → 0.2). -No difference in VA/EDSS improvement was observed between groups.	-AEs in patients receiving LPE therapy included allergic reactions (rate of 7.3%) and citrate toxicity (rate of 1.6%). -AEs in patients receiving TPE therapy included allergic reactions (rate of 6.4%) and catheter thrombosis (rate of 1.0%).
Gonzalez et al., 2022	-78.3% of patients experienced improved VOS by ≥1 point in VOS; 47% by ≥2 points. -Age and pre-TPE VOS predicted the response. -No association was observed between AQP4+ status and outcomes.	-Hypofibrinogenemia occurred in 26% of sessions. -Minor AEs: hypocalcemia (rate of 21.5%), catheter dysfunction (rate of 1.96%).
Chen et al., 2023	-Most patients experiencing ON attacks experienced improvement after PLEX therapy (final median VA: 20/25). -Patients with MOGAD-ON experienced better outcomes (median VA: 20/20). -Poor outcomes were predicted by older age, severe vision loss at nadir, and delayed PLEX therapy. -Outcomes in patients receiving PLEX therapy were superior to those receiving ONTT (12% vs. 33%, respectively, with VA < 20/40).	-1% of patients experienced infection (central line-related). -Rare AEs: DVT (rate of 0.3%), electrolyte imbalances (rate of 0.5%).

Note: AE: Adverse event; ARR: Annualized relapse rate; CFS: Combined Functional Score; MS: Clinically isolated syndrome; DVT: Deep vein thrombosis; EDSS: Expanded Disability Status Scale; IA: Immunoadsorption; IgG: Immunoglobulin G; IVMP: Intravenous methylprednisolone; LPE: Low-volume plasma exchange; MBP: Myelin basic protein; MRC: Medical Research Council scale; MS: Multiple sclerosis; MSFC: Multiple Sclerosis Functional Composite; NLP: No light perception; NMOSD: Neuromyelitis optica spectrum disorder; ON: Optic neuritis; PE/PLEX/TPE: Plasma exchange/Plasmapheresis/Therapeutic plasma exchange; RRMS: Relapsing-remitting multiple sclerosis; sIL-2R: Soluble interleukin-2 receptor; VA: Visual acuity; VOS: Visual Outcome Scale.

Functional and neurological outcomes

Several studies have also investigated secondary outcomes, such as improvements in Expanded Disability Status Scale (EDSS) scores and overall disability (Table 2). Apheresis therapy was associated with long-term functional gains and reduced disability. Notably, IA was linked to higher Multiple Sclerosis Functional Composite scores, indicating broader neurological benefits.

Safety profile

Studies have shown that apheresis therapies are generally well tolerated. Most adverse events were mild to moderate, including hypotension, hypocalcemia, and allergic reactions. Serious side effects were uncommon. One study described a single case in which surgical intervention was needed due to an arteriovenous fistula, while another reported less than 1% deep vein thrombosis or infections associated with central line placement. There was no reported mortality due to the intervention (Table 2).

Effect of plasma exchange on visual acuity

A meta-analysis of four studies that investigated the effect of PLEX therapy on visual acuity, as measured by the mean difference in logMAR values before and after treatment (see Supplementary Fig. 2), is shown in Fig. 2. All included studies demonstrated significant postintervention improvements, with mean differences ranging from -0.80 to -1.60, indicating a shift from poor to good vision. The greatest effect was found in the Chen (2023) study, with all confidence intervals excluding zero, confirming statistical significance. The pooled analysis revealed a significant improvement in logMAR visual acuity, with a mean difference of -1.18 (95% CI: -1.29 to -1.08) under the common effects model and -1.11 (95% CI: -1.70 to -0.53) under the random effects model (P < 0.0001 for both). However, there was significant heterogeneity among the studies (I² = 93.6%), reflecting variability in patient populations, timing of therapy, or apheresis protocols.

Risk of bias in the included studies

The risk of bias in the 12 included studies varied according to study design (see Supplementary Fig. 3). The majority of RCTs were rated as low to moderate risk in most domains, with some concerns primarily about deviations from intended interventions and outcome measurement. The ROBINS-I assessment of nonrandomized studies revealed a moderate risk of bias, particularly in retrospective cohorts, owing to confounding and participant selection. One study (Roesner et al., 2011) was found to have a high risk of bias because of methodological limitations inherent in the retrospective case series design. Overall, no study showed a high risk in all domains, and the methodological quality was deemed acceptable. Nonetheless, variability in reporting standards and study design heterogeneity are potential concerns that could affect pooled estimates.

Discussion

This systematic review and meta-analysis brings together the most recent evidence on the efficacy and safety of PLEX and immunoadsorption (IA) therapies for the treatment of acute ON associated with demyelinating diseases such as MS, NMOSD, and MOGAD. These findings support the clinical utility of apheresis therapies, especially in patients who do not respond to standard corticosteroid therapy. Both PLEX and IA therapies were consistently linked to significant improvements in visual acuity and neurological function, with a generally positive safety profile.

Consistent with prior research, such as the meta-analysis by Yu et al. (2020) that focused on patients with NMOSD [4], our review confirms that PLEX therapy can produce significant increases in visual acuity, particularly when it is administered promptly. The pooled results of our meta-analysis revealed an average improvement in logMAR visual acuity of -1.11 under a random-effects model, with all included studies showing significant gains over baseline. This finding is consistent with the findings of Yu et al., who reported an average improvement in the EDSS score of 1.83 points after NMOSD attacks and a 74% overall response rate to PLEX therapy [4].

Our findings also suggest that, although less frequently studied, IA therapy may provide more long-term and potentially broader neurological benefits. For example, Dorst et al. (2019) reported that compared with PLEX therapy, IA therapy was associated with more durable neurological recovery owing to its targeted removal of IgG and immune complexes [20]. These findings are consistent with those of Lipphardt et al. (2019), who reported similar efficacy between IA and PLEX therapies, with IA therapy occasionally providing better results for specific patient subgroups [21].

According to multiple studies, early initiation of apheresis is among the strongest predictors of a positive treatment response. Tan et al. (2020) and Chen et al. (2023) reported that patients who received PLEX therapy earlier in the disease course had significantly better visual outcomes [22,23]. These findings support those of Yu et al., who reported that earlier PLEX therapy initiation, within 14 days, was associated with greater improvements in the EDSS score and higher response rates [4]. These findings support Bonnan et al.'s (2018) hypothesis that ON and other NMOSD relapses progress through distinct temporal phases, with early apheresis in the “inflammatory window” yielding optimal results, whereas delayed intervention may miss the opportunity to reverse active immune-mediated damage [29].

Our analysis also revealed a significant variation in treatment response by ON subtype. Patients with MOGAD-associated ON appeared to experience the best visual recovery, whereas those with NMOSD-associated ON—particularly those positive for AQP4—often experienced poorer outcomes unless treatment was prompt and aggressive. Notably, Chen et al. (2023) reported that patients with MOGAD had a median final visual acuity of 20/20, which is consistent with the broader literature indicating that patients with MOGAD respond well to both steroids and apheresis alike [23].

Interestingly, some studies have suggested that AQP4-IgG-seronegative patients may respond better to PLEX therapy, a finding supported by Yu et al.'s subgroup analysis [4]. This paradox may be because of differences in immune mechanisms between antibody-mediated injury and other types of injury, suggesting that PLEX therapy may act via broader immunomodulatory pathways, such as the removal of cytokines and complement factors [30]. Baseline prognostic factors—including older age, more severe vision loss at nadir, and delayed treatment—were consistently associated with poorer outcomes, emphasizing the importance of early emphasis and intervention [31].

The apheresis treatments were well tolerated. The majority of adverse events were mild and manageable, with serious complications such as deep vein thrombosis or infections occurring in less than 1% of cases, which is consistent with previous reports [32]. Our review did not find a significant difference in adverse event rates between IA and PLEX therapies, indicating that their safety profiles are comparable. These findings are consistent with those of Yu et al. (2021), who reported a favorable safety profile for these modalities, with adverse effects occurring in approximately one-third of sessions, most of which were related to catheter placement or transient metabolic disturbances [4].

Limitations

Several limitations should be recognized. First, studies showed significant heterogeneity, particularly in terms of outcome definitions, timing of intervention, and disease subtypes. This was reflected in our meta-analysis of visual acuity, in which the I² statistic was above 90%2. Many of the included studies were retrospective or lacked control groups, which limited causal inference. Second, the outcomes were based on visual measures, whereas neurological metrics such as the EDSS score are more commonly used in studies of NMOSD and may not fully capture the nuances of optic nerve involvement. Finally, differences in definitions of “response” between studies may have influenced the pooled estimates.

In conclusion, this review shows that both PLEX and immunoadsorption (IA) therapies are effective and safe treatments for acute ON associated with demyelinating diseases. Both interventions resulted in significant improvements in visual outcomes, especially when treatment began early. Among the subtypes, patients with MOGAD-associated ON experienced the best results, whereas those with NMOSD-associated ON required prompt treatment to avoid poor outcomes. IA therapy proved to be at least as effective as PLEX therapy and, in some cases, offered superior results. Overall, apheresis treatments were well tolerated, with few serious side effects reported. These findings support early apheresis as a viable second-line treatment option for patients with ON who do not respond to steroids.

Acknowledgments

We would like to thank enago for English language editing.

Disclosure

Authors have no conflict of interests and the work was not supported or funded by any drug company.