Lymphoplasmapheresis for Refractory Anti–N‐Methyl‐D‐Aspartate Receptor Encephalitis: A Case Report

ABSTRACT

Anti–N‐methyl‐D‐aspartate receptor (NMDAR) encephalitis is generally responsive to immunotherapy, but some patients remain refractory despite treatment escalation. Lymphoplasmapheresis (LPE), a modified apheresis technique that removes both plasma components and circulating lymphocytes, may offer a salvage option in selected cases. We report a 61‐year‐old woman with refractory anti‐NMDAR encephalitis who showed limited improvement after corticosteroids, intravenous immunoglobulin, and ofatumumab. After a prolonged interval following ofatumumab induction, she remained severely impaired with persistent disease activity and elevated cerebrospinal fluid and serum anti‐NMDAR IgG titers. Three sessions of LPE were performed over 5 days. Clinical improvement was observed after LPE, accompanied by a decline in antibody titers. At 5‐month follow‐up after LPE, she maintained sustained recovery, with a Clinical Assessment Scale in Autoimmune Encephalitis score of 6 and a modified Rankin Scale score of 2. This case supports a close temporal association between LPE and clinical and immunological improvement, although causality cannot be definitively established from a single case.

Anti–N‐methyl‐D‐aspartate receptor (NMDAR) encephalitis is the most common form of autoimmune encephalitis and is characterized by a combination of psychiatric symptoms, seizures, movement disorders, and impaired consciousness [1]. Current treatment strategies mainly rely on immunotherapy, with first‐line approaches including corticosteroids, intravenous immunoglobulin (IVIg), and therapeutic plasma exchange (TPE), followed by second‐line agents such as rituximab or other B‐cell–depleting therapies in refractory cases [2]. Although most patients respond to these treatments, a subset remains refractory despite treatment escalation, resulting in prolonged disease courses, increased morbidity, and poor functional outcomes [3]. In such patients, effective salvage therapies are limited, and optimal management remains challenging.

Lymphoplasmapheresis (LPE) is a modified therapeutic apheresis modality derived from TPE. Unlike TPE, which primarily removes soluble plasma components such as autoantibodies, complement components, and inflammatory mediators, LPE is designed to remove both plasma components and part of the circulating leukocyte‐rich layer [4]. This approach may allow simultaneous reduction of circulating pathogenic antibodies and immunoreactive lymphocytes, providing a theoretical rationale for broader immunomodulation in selected refractory autoimmune neurological disorders. However, evidence supporting the use of LPE in anti‐NMDAR encephalitis remains limited. Key conceptual differences between conventional TPE and LPE are summarized in Table 1.

TABLE 1.

Comparison between therapeutic plasma exchange and lymphoplasmapheresis.

Feature	Therapeutic plasma exchange	Lymphoplasmapheresis
Main mechanism	Removes soluble pathogenic plasma components and mainly modulates humoral immunity	Removes soluble plasma components and part of the circulating leukocyte‐rich layer, potentially modulating both humoral and cellular immunity
Main removed components	Autoantibodies, immune complexes, complement components, cytokines, and plasma proteins	Autoantibodies, immune complexes, complement components, cytokines, plasma proteins, and circulating lymphocytes
Cellular removal	Not designed to remove immune cells	Designed to collect the leukocyte‐rich layer through interface adjustment
Exchange volume	Protocol‐dependent; commonly 1.0–1.5 plasma volumes per session	Protocol‐dependent; approximately 0.7 plasma volume per session in the present case
Potential clinical implication	Rapid short‐term reduction of circulating antibodies and inflammatory mediators	Potential combined reduction of circulating antibody burden and immune‐cell load
Key limitation	Does not directly remove immune cells	Lymphocyte contribution cannot be distinguished from antibody removal without controlled studies

A 61‐year‐old woman with no prior neurological, autoimmune, or psychiatric history presented with acute psychiatric symptoms, generalized tonic seizures, and progressive impairment of consciousness. The date of diagnosis was defined as month 0. Her Clinical Assessment Scale in Autoimmune Encephalitis (CASE) score was 23. Cerebrospinal fluid (CSF) showed pleocytosis of 55 × 10⁶/L and anti‐NMDAR IgG positivity at a titer of 1:1. She received high‐dose intravenous methylprednisolone, IVIg, acyclovir, and symptomatic treatment at a local hospital, but showed no meaningful improvement.

At month 1, she was transferred to our hospital with coma and preserved pupillary reflexes. The CASE score remained 23. Peripheral blood showed an elevated B‐cell proportion of 24.36%, and brain magnetic resonance imaging demonstrated bilateral occipital subcortical hyperintensities. Repeat antibody testing showed increased anti‐NMDAR IgG titers in CSF (1:3.2) and serum (1:10). Other autoimmune antibodies and infectious studies were negative. Weekly subcutaneous ofatumumab 20 mg was initiated. After four weekly doses, circulating B‐cell counts declined and consciousness improved slightly to a shallow coma, with a CASE score of 19. Given this partial response, monthly maintenance ofatumumab was not initiated after discharge; instead, she was monitored clinically while receiving tapering oral prednisone.

At month 5, she was readmitted with persistent severe neurological impairment despite prior first‐line therapy and four doses ofatumumab induction. She had persistent delirium, required tracheostomy and nasogastric feeding, and had a CASE score of 19. Anti‐NMDAR IgG titers were elevated in both CSF and serum at 1:10, and malignancy screening was negative. Because of the refractory course, persistently elevated antibody titers, and limited response to previous therapies, LPE was initiated as a salvage intervention.

LPE was performed using a Fresenius COM.TEC blood cell separator through peripheral venous access. Lymphocyte removal was achieved by manual adjustment of the separation interface to the leukocyte‐rich layer, allowing lymphocytes to be removed together with plasma. Three sessions were performed over 5 days, with approximately 48‐h intervals. Sodium citrate was used for anticoagulation, and approximately 0.7 plasma volume was exchanged per session using fresh frozen plasma as the replacement fluid. The cellular components in the removed fluid were assessed by complete blood count analysis, and each session removed an estimated 1–3 × 10⁹ lymphocytes. Hemoglobin and platelet counts remained within safe ranges. No hemodynamic instability, clinically relevant coagulation abnormality, bleeding, infection, allergic reaction, or other procedure‐related adverse event was observed.

After two LPE sessions, the patient could follow simple commands, including eye opening and limb movement, and the CASE score decreased to 15. After the third session, she could orient toward sound, with a CASE score of 14. Anti‐NMDAR IgG titers declined to 1:3.2 in both CSF and serum. At 2‐month follow‐up after LPE, corresponding to month 8 after diagnosis, she was fully oriented, able to communicate appropriately, and showed improved motor function. The CASE score decreased to 10, CSF antibody titers declined to 1:1, and serum titers remained 1:3.2. At 5‐month follow‐up after LPE, corresponding to month 11 after diagnosis, she maintained sustained recovery, with independent ambulation, partial self‐care ability, no residual psychiatric or brainstem symptoms, a CASE score of 6, and a modified Rankin Scale score of 2. The integrated treatment timeline, CASE score trajectory, and serial CSF and serum antibody titers are shown in Figure 1.

FIGURE 1.

Clinical course after diagnosis, including treatment stages, CASE score, and anti‐NMDAR IgG titers. The upper bar shows the major treatment stages from diagnosis, including first‐line therapy with intravenous methylprednisolone and intravenous immunoglobulin, ofatumumab induction, LPE, and follow‐up. Panel (A) shows serial CASE scores. Panel (B) shows serial anti‐NMDAR IgG titers in cerebrospinal fluid and serum. The LPE interval is visually expanded to improve readability. Clinical improvement after LPE was accompanied by a decline in both CSF and serum antibody titers. CASE, Clinical Assessment Scale in Autoimmune Encephalitis; CSF, cerebrospinal fluid; IVIg, intravenous immunoglobulin; IVMP, intravenous methylprednisolone; LPE, lymphoplasmapheresis; NMDAR, N‐methyl‐D‐aspartate receptor.

1. Discussion

This case illustrates a close temporal association between LPE and clinical and immunological improvement in refractory anti‐NMDAR encephalitis. Previous studies have suggested that TPE may provide early clinical improvement and reduce anti‐NMDAR antibody titers in severe or refractory cases, supporting the relevance of extracorporeal antibody removal in this disease [5]. In addition, studies in other immune‐mediated disorders, including myasthenia gravis and thrombotic thrombocytopenic purpura, have suggested that LPE may achieve comparable clinical efficacy to conventional TPE while reducing the number of treatment sessions or exchange volume in selected protocols [4, 6, 7]. These findings provided a procedural rationale for considering LPE in this refractory case, although disease‐specific evidence in anti‐NMDAR encephalitis remains limited. LPE was selected because the patient had persistent disease activity and rising antibody titers despite prior ofatumumab induction and a decline in circulating B‐cell counts, suggesting that immediate removal of circulating pathogenic factors remained clinically relevant. In addition, LPE may remove part of the circulating lymphocyte compartment in addition to soluble plasma components, providing a theoretical rationale for broader immunomodulation than TPE alone. Notably, nearly 3 months had elapsed between completion of ofatumumab induction and LPE initiation, during which she remained severely impaired and antibody titers increased, making a delayed ofatumumab effect less likely as the sole explanation for subsequent improvement.

Several limitations should be acknowledged. This single case cannot establish causality or demonstrate superiority of LPE over TPE. The independent contribution of lymphocyte removal cannot be distinguished from plasma‐mediated antibody clearance, and detailed lymphocyte immunophenotyping by flow cytometry was not performed. Moreover, anti‐NMDAR antibodies may be produced both peripherally and intrathecally, and LPE primarily acts on the peripheral circulation; therefore, changes in serum or even CSF antibody titers may not fully capture CNS antibody dynamics, which are also influenced by blood–brain barrier function [8, 9]. Finally, fresh frozen plasma was used according to institutional practice at that time, although albumin may be preferable in patients without coagulopathy to reduce unnecessary plasma exposure. Further studies are needed to clarify the efficacy, mechanisms, safety profile, and optimal procedural parameters of LPE in refractory autoimmune encephalitis.

Funding

The authors have nothing to report.

Ethics Statement

Ethical approval for this case report was waived by the Institutional Review Board of West China Hospital, Sichuan University, as this study describes a single patient and does not involve experimental interventions.

Consent

Written informed consent was obtained from the patient for publication of this case report and any accompanying images.

Conflicts of Interest

The authors declare no conflicts of interest.

Data Availability Statement

The data that support the findings of this study are available on request from the corresponding author. The data are not publicly available due to privacy or ethical restrictions.