Effectiveness and safety of efgartigimod in patients with anti‐NMDAR or anti‐LGI1 encephalitis and its 18F‐DPA714 PET imaging evaluation: A multicenter, single‐arm, prospective, observational real‐world study Authors: Huanyu Meng, Siqi Tu, Yifan Zhou, Ruinan Shen, Xiaoyu Chen, Zhaozhao Cheng, Limei Wang, Yan Wu, Jie Lu, Shengjun Wang, Yan Xu, Wanhui Lin, Sheng Chen Published: 2026 ABSTRACT AbstractBackgroundEfgartigimod disrupts the recycling of pathogenic IgG autoantibodies and may represent a candidate therapy for autoimmune encephalitis. This real‐world study evaluated the clinical efficacy, safety, and impact of efgartigimod on brain inflammation.MethodsThis multicenter, single‐arm, prospective, observational real‐world study enrolled 18 patients, including 10 with anti‐N‐methyl‐d‐aspartate receptor (NMDAR) encephalitis and 8 with anti‐leucine‐rich glioma‐inactivated 1 (LGI1) encephalitis. Participants received intravenous efgartigimod at a dose of 10 mg/kg weekly for a total of two to four doses, with or without concomitant immunosuppressive therapy. Anti‐NMDAR‐IgG/LGI1‐IgG antibody titers, modified Rankin scale (mRS), Clinical Assessment Scale in Autoimmune Encephalitis (CASE), Montreal Cognitive Assessment, Mini‐Mental State Examination, and Glasgow Coma Scale scores were assessed at baseline and at 1, 2, 3, 8, and 12 weeks after treatment. 18F‐DPA714 positron emission tomography/magnetic resonance imaging was used to monitor brain inflammation in a subset of patients. Analyses were performed using SPSS 26.0 and Prism 10.0. Non‐parametric tests, including Wilcoxon signed‐rank, Mann–Whitney U, and Friedman tests with Bonferroni post hoc correction, were applied. Two‐tailed p < 0.05 indicated statistical significance; corrected p < 0.01 was applied for five longitudinal pairwise comparisons.ResultsAt the final follow‐up, median mRS scores in patients with anti‐NMDAR encephalitis decreased significantly from 3 (range, 1–5) before treatment to 1 (range, 0–3) after treatment (Z = −2.588, p = 0.007). Median CASE scores decreased from 7.5 (range, 1–27) to 1.5 (range, 1–7) post‐treatment (Z = −2.714, p = 0.005). In patients with anti‐LGI1 encephalitis, CASE scores decreased markedly from 3 (range, 1–7) to 1.5 (range, 0–2) at the last visit (Z = −2.154, p = 0.046). All 18 patients underwent baseline imaging, and 7 patients underwent follow‐up PET/MRI examinations. A subset of patients (n = 4; three with anti‐NMDAR encephalitis and one with anti‐LGI1 encephalitis) showed reduced uptake in specific brain regions accompanied by symptom improvement. Both initial and follow‐up imaging results were negative for the remaining three patients. Comparative analysis of efgartigimod monotherapy versus combination therapy with glucocorticoids and immunosuppressants revealed no significant differences in key outcomes, indicating that the therapeutic effect was independent of concomitant medication use. Efgartigimod was well tolerated. One patient (1/18) developed mild pneumonia, but no other serious adverse events were reported.ConclusionEfgartigimod may represent a novel therapeutic option for anti‐NMDAR and anti‐LGI1 encephalitis, potentially improving clinical outcomes in affected patients.Registration InfoThis study was registered at the National Clinical Research Center (https://www.medicalresearch.org.cn; No. MR‐31‐24‐053225).Keywords: anti‐N‐methyl‐d‐aspartate receptor encephalitis, autoimmune, encephalitis, Efgartigimod, Fc, Immunoglobulin G, LGI1 protein, neonatal, N‐Methyl‐d‐Aspartate, receptors, treatment outcome INTRODUCTION Encephalitis associated with anti‐N‐methyl‐d‐aspartate receptor (NMDAR) and anti‐leucine‐rich glioma‐inactivated 1 (LGI1) antibodies represents one of the most prevalent subtypes of autoimmune encephalitis (AIE), accounting for 16.67% and 21.23% of the total incidence, respectively. 1 Anti‐NMDAR encephalitis is primarily characterized by a complex neuropsychiatric profile, including psychosis, seizures, and cognitive decline, 2 whereas anti‐LGI1 encephalitis is typically associated with faciobrachial dystonic seizures and cognitive impairment. 3 About 20–50% of patients diagnosed with anti‐NMDAR encephalitis and anti‐LGI1 encephalitis respond poorly to first‐line therapies. 4 , 5 Given the large proportions of these AIE subtypes, future therapeutic approaches may involve combining established treatments with new immunomodulatory agents to enhance clinical outcomes. Emerging AIE therapeutic strategies are designed to target specific immune system components. Developing novel immunotherapies for anti‐NMDAR and anti‐LGI1 encephalitis depends on a comprehensive understanding of the underlying pathogenic mechanisms. Both conditions involve IgG‐mediated mechanisms, and the neonatal Fc receptor (FcRn) has emerged as a promising therapeutic target, making it an ideal candidate for further investigation. FcRn plays a crucial role in the recycling and stabilization of IgG by binding to it in acidic environments within lysosomes, thereby preventing its degradation and facilitating its return to the circulation. 6 This process prolongs the half‐life of IgG and sustains elevated systemic levels. 7 Efgartigimod, a human IgG1 antibody Fc‐fragment and natural ligand of FcRn, competes with IgG for FcRn binding, with better affinity, thereby effectively inhibiting IgG recycling. 8 Efgartigimod has demonstrated therapeutic efficacy in patients with myasthenia gravis (MG), 9 systemic lupus erythematosus, 10 and chronic inflammatory demyelinating polyneuropathy (CIDP). 11 Therefore, the potential of efgartigimod to disrupt the recycling of pathogenic autoantibodies may represent a therapeutic strategy for IgG‐mediated anti‐NMDAR and anti‐LGI1 encephalitis. Efgartigimod is a new immunomodulator with reported effectiveness in limited AIE cases. 12 , 13 However, comprehensive, large‐scale, multicenter data confirming the efficacy and safety of efgartigimod in treating AIE are currently lacking. Assessing intracerebral inflammation in AIE remains challenging. The advent of 18F‐DPA714 positron emission tomography (PET)/magnetic resonance imaging (MRI), which traces translocator protein (TSPO) levels on microglial mitochondria, has enabled the visualization of microglial activation and aggregation in the brain, offering a potential method for assessing the severity of intracerebral inflammation. 14 Studies employing TSPO PET have demonstrated its ability to visualize the extent and severity of neuroinflammation in AIE subtypes, including anti‐NMDAR and anti‐LGI1 encephalitis, thereby complementing conventional imaging (e.g., MRI) and providing new insight into disease activity and treatment responses. 15 , 16 , 17 This research represents the multicenter, single‐arm, prospective, observational, real‐world study conducted in central and eastern China. The study excluded patients who had received plasma exchange or immunoadsorption within 4 weeks before enrollment, or immunosuppressants or biologics at any time before enrollment (whereas a short course of corticosteroids or IVIg was permitted), and separately analyzed the study data of patients with two antibody types (anti‐NMDAR and anti‐LGI1) to avoid the confounding effect of prior medications and different antibody types on the therapeutic efficacy of efgartigimod. The study aimed to evaluate the therapeutic potential, efficacy, and safety of efgartigimod in patients with anti‐NMDAR and anti‐LGI1 antibody‐associated AIE. 18F‐DPA714 PET was used to evaluate changes in the intracranial inflammation levels of patients before and after efgartigimod treatment. The findings will provide a foundation for future research and significantly advance the clinical application of efgartigimod in IgG‐mediated AIE. METHODS Study design and participants This multicenter, single‐arm, prospective, observational real‐world study evaluated patients receiving efgartigimod for anti‐NMDAR and anti‐LGI1 encephalitis. Patients were recruited from the Ruijin Hospital of Shanghai Jiao Tong University; the First Affiliated Hospital of University of Science and Technology of China; the First Affiliated Hospital of Zhengzhou University; the Union Hospital, Tongji Medical College, Huazhong University of Science and Technology; the Nanjing Brain Hospital, the Qilu Hospital of Shandong University; the Union Hospital Affiliated with Fujian Medical University; and the Zhongnan Hospital of Wuhan University. The study was approved by the Medical Ethics Committee of Ruijin Hospital, Shanghai Jiao Tong University School of Medicine (approval No. 2024[350]) and registered at the National Clinical Research Center (No. MR‐31‐24‐053225) (Supplementary Material 1). All procedures involving human participants were conducted in accordance with the Declaration of Helsinki and the ethical standards of the Institutional Research Committee. All participants or their guardians provided written informed consent. The inclusion criteria were: (1) age between 18 and 80 years and (2) fulfillment of the diagnostic criteria for anti‐NMDAR encephalitis or anti‐LGI1 encephalitis. 18 The exclusion criteria were: (1) prior treatment with immunosuppressive agents or biologics such as plasma exchange (PE) or immunoadsorption within 4 weeks before enrollment, azathioprine, mycophenolate mofetil, cyclophosphamide, cyclosporine, tacrolimus, rituximab, or tocilizumab at any time before enrollment; a short course of corticosteroids or intravenous immunoglobulin (IVIg) as acute‐phase emergency treatment before referral was not an exclusion criterion and was documented separately; (2) the presence of untreated malignancies; (3) pregnancy, lactation, or a recent planned pregnancy; (4) allergy to efgartigimod; and (5) serious complications, including but not limited to infections, immunodeficiency, renal failure, hematological disorders (e.g., neutropenia, hypogammaglobulinemia, and hematologic malignancies), and acute cardiovascular or cerebrovascular diseases. As a real‐world observational study of rare diseases (anti‐NMDAR and anti‐LGI1 encephalitis), a formal a priori sample size calculation was not applicable. Instead, we planned a consecutive enrollment of all eligible patients from the eight participating centers during the predefined recruitment period (January 2024 to March 2025). A cohort of 18 patients, comprising 10 individuals diagnosed with anti‐NMDAR encephalitis and eight with anti‐LGI1 encephalitis from eight centers, was enrolled in this study. The final sample of 18 patients was considered adequate for a pilot evaluation of the efficacy, safety, and imaging changes of efgartigimod in this setting. No interim analysis was conducted. The flowchart of patient selection is shown in Figure 1. The diagnosis of anti‐NMDAR or anti‐LGI1 encephalitis was confirmed in all enrolled patients by detecting specific antibodies in serum and/or cerebrospinal fluid (CSF) using a cell‐based assay (CBA). Antibody positivity was determined by meeting both of the following criteria 2 , 18 , 19 , 20 , 21 : (1) a specific immunofluorescence staining pattern, characterized by distinctive staining on the cell membrane surface, observable under a fluorescence microscope. (2) Titers meeting the threshold for positivity, defined as a dilution titer of ≥1:10 in serum and ≥1:1 in CSF for anti‐NMDAR encephalitis, and ≥ 1:32 in serum and ≥1:1 in CSF for anti‐LGI1 encephalitis. While results from both serum and CSF were considered, priority was given to CSF‐positivity due to its higher specificity for autoimmune reactions within the central nervous system (CNS). The final antibody assignment was determined only if the CBA result met the specific staining pattern and titer requirements, allowing the patient to be included in the study. Supplementary Tables 1 and 2 provide detailed information on the detection methods, antibody titers, and sample types assayed for patients with anti‐NMDAR and anti‐LGI1 encephalitis. Anti‐NMDAR‐IgG and anti‐LGI1‐IgG were assessed by two independent investigators using a CBA at a single center (Nanjing Simcere Medical Laboratory Co. Ltd., Nanjing, China). Flowchart of patient screening, enrollment, and group allocation. This flowchart shows the patient disposition in this multicenter, single‐arm, prospective, real‐world study evaluating efgartigimod in anti‐NMDAR and anti‐LGI1 encephalitis. A total of 53 patients were initially screened across eight centers in China. Of these, 35 were excluded: 27 did not meet the inclusion criteria, six patients had previously received immunosuppressive or immunotargeted agents, and two patients had untreated malignant tumors. The remaining 18 eligible patients received efgartigimod and were divided into two groups: 10 with anti‐NMDAR encephalitis and 8 with anti‐LGI1 encephalitis. anti‐LGI1, anti‐leucine‐rich glioma‐inactivated 1; anti‐NMDAR, anti‐N‐methyl‐d‐aspartate receptor; IVIg, intravenous immunoglobulin; PE, plasma exchange. Treatment and follow‐up All patients received a standard dose of 10 mg/kg efgartigimod weekly via intravenous injection, totaling two to four doses (the number of doses was based on symptom improvement). Patients in the acute phase who had contraindications to or were unable to utilize first‐line treatments for various reasons were administered efgartigimod either alone or in combination with less than 60 mg prednisone per day. IVIg, PE, and other monoclonal antibody agents administered after enrollment were prohibited throughout the entire study follow‐up period (12 weeks). 22 Dose adjustments were not routinely performed. Efgartigimod dosing was primarily guided by clinical improvement, with financial considerations and complications influencing the decisions. However, treatment was temporarily suspended in patients who experienced significant adverse events, particularly infections. The decision to resume or discontinue efgartigimod was determined based on clinical judgment and patient tolerance. No pre‐specified dose modifications were established. All patients were assessed at baseline (Week 0) and 1, 2, 3, 8, and 12 weeks following the initiation of efgartigimod treatment. Of the 18 enrolled patients, four patients (three with anti‐NMDAR encephalitis and one with anti‐LGI1 encephalitis) completed all scheduled clinical assessments through the 12‐week follow‐up, primarily attributable to substantial clinical improvement leading to early hospital discharge, geographical inaccessibility, or financial constraints. No patient withdrew due to adverse events or perceived lack of efficacy. The number of patients assessed at each time point is detailed in Supplementary Table 3. Hybrid 18F‐DPA714 PET/MRI acquisition All AIE patients underwent 18F‐DPA714 PET/MRI scanning either before immunotherapy or 3 months after treatment. Before scanning, the patients were screened for rs6971 TSPO gene polymorphism (Tsingke Biotechnology Co. Ltd., Beijing, China). Only those with a high‐affinity binding variant were included in the study. Hybrid PET/MRI was performed using a Biograph mMR system (Siemens Healthineers, Erlangen, Germany). PET acquisition was performed in the list mode, 60–90 min after the intravenous administration of 3.7 MBq/kg of 18F‐DPA714. The precursor DPA714 was obtained from Beijing PET Technology Co. Ltd. (Beijing, China) and radiolabeled as previously described. 23 3D T1‐weighted magnetization‐prepared rapid gradient echo, 3D T2‐weighted fluid‐attenuated inversion recovery (FLAIR), and diffusion‐weighted imaging (DWI) brain MRI images were acquired during PET acquisition. PET images were reconstructed with a matrix of 344 × 344 pixels, four iterations, 21 subsets, and a filter with a full width at half maximum of 2 mm, using a point spread function algorithm. Two radiologists specialized in neuroradiology and blinded to the patient's medical history independently interpreted and assessed the MRI scans. Two nuclear medicine physicians with expertise in nuclear neurology analyzed the 18F‐DPA714 PET images. Any disagreements were resolved by consensus with the assistance of a third senior radiologist and a nuclear medicine physician. The conventional MRI features of antibody‐negative AIE included hyperintense signals on the T2‐FLAIR sequence, with or without corresponding high‐signal regions in DWI or enhancement in postcontrast T1‐weighted imaging. 14 , 18 , 24 , 25 Data collection and outcomes Data collected from medical records included demographics, past medical history, age at disease onset, disease duration, clinical symptoms, laboratory findings, and treatments. Two neurologists assessed AIE severity using standardized scales: the Clinical Assessment Scale for Autoimmune Encephalitis (CASE), Glasgow Coma Scale (GCS), Modified Rankin Scale (mRS), Mini‐Mental State Examination (MMSE), and Montreal Cognitive Assessment (MoCA) at 0, 1, 2, 3, 8, and 12 weeks. Evaluating physicians from different centers received standardized training to ensure consistency in the clinical scale assessments. Anti‐NMDAR‐IgG and anti‐LGI1‐IgG antibodies in the patient's serum and/or CSF were detected using commercial CBA‐based panels. Patient samples before and after efgartigimod treatment were analyzed in the same laboratory. Total serum IgG levels were also measured. The primary clinical outcome was mRS scores measured before and at the final follow‐up visit after efgartigimod treatment. The secondary outcomes included improvements in disability evaluated using CASE, GCS, MMSE, and MoCA scores. Additional secondary outcomes included changes in laboratory indicators, such as anti‐NMDAR‐IgG and anti‐LGI1‐IgG titers and serum total IgG levels. A reduction of ≥4‐fold in anti‐NMDAR or anti‐LGI1 titers, or conversion to a negative result, was used as the criterion for a significant decrease. 26 A significant improvement in patient symptoms was defined as an mRS score improvement of ≥1 point, 27 a CASE score improvement of ≥ 1 point, 27 a GCS score improvement from 3 to 12 to 13–15, 28 and an MMSE or MoCA score improvement of ≥3 points 29 , 30 at 1, 2, 3, 8, and 12 weeks. Safety assessments were performed throughout the study period. All adverse events were systematically monitored, recorded, and classified according to the Medical Dictionary for Regulatory Activities. The treating physicians assessed the potential association between efgartigimod and each adverse events using clinical judgment. The following adverse events were specifically assessed based on the known safety profile of efgartigimod from previous clinical trials 17 , 31 , 32 , 33 , 34 : respiratory tract infection, urinary tract infection, headache, nasopharyngitis, paresthesia, myalgia, infusion‐related reactions, and hypersensitivity reactions. Detailed information was collected for each adverse events, including the type, onset time, severity, and treatment administered. Statistical analysis Statistical analyses were performed using SPSS version 26.0 (SPSS, Chicago, IL, USA) and GraphPad Prism version 10.0 (GraphPad Software, La Jolla, CA, USA). Continuous variables were tested for normality using the Shapiro–Wilk test. Normally distributed data are presented as the mean ± standard deviation (SD), while non‐normally distributed data are presented as medians (range). Categorical variables are expressed as frequencies (n/N) without percentages, as the total sample size was less than 100. For longitudinal comparisons involving multiple time points (baseline vs. Weeks 1, 2, 3, 8, and 12), the Friedman test for non‐parametric repeated measures data was applied, followed by post‐hoc pairwise comparisons using the Wilcoxon signed‐rank test with Bonferroni correction. The comparison between baseline and the last follow‐up was the principal paired analysis within this framework; a two‐tailed adjusted p‐value < 0.01 (0.05/5) was considered statistically significant for the five pairwise comparisons. For independent between‐group comparisons (e.g., efgartigimod monotherapy vs. combination therapy), the Mann–Whitney U test was employed. The standardized test statistic (Z value) is reported alongside the corresponding p value. For all other analyses, a two‐tailed p‐value < 0.05 was considered statistically significant. The statistical analysis plan, including all non‐parametric testing procedures, multiple‐comparison corrections, and the final results, was independently reviewed by a qualified biostatistician from the Department of Biostatistics, Ruijin Hospital, Shanghai Jiao Tong University School of Medicine. A certificate confirming the statistical review has been provided as a supplementary file (Supplementary Material 2). RESULTS Baseline characteristics Eighteen patients were enrolled from eight centers, comprising 10 with anti‐NMDAR encephalitis and eight with anti‐LGI1 encephalitis. The demographic, clinical, and laboratory features are presented in Tables 1 and 2. The mean age of the patients with anti‐NMDAR encephalitis was 34.6 ± 16.9 years (range: 18–67 years), with a female‐to‐male ratio of 4:1. The median time from disease onset to efgartigimod treatment was 14.5 days (range: 5–29). Seizures were the most frequent symptom (8/10), followed by behavioral and personality changes (5/10), decreased consciousness (4/10), and cognitive deficits (2/10). At baseline, the median scores were 3 for the mRS (range: 1–5), 7.5 for the CASE (range: 1–27), and 12 for the GCS (range: 3–15). Anti‐NMDAR‐IgG was detected in the serum of 8 of 10 patients and in all 10 CSF samples, with median titers of 1:100 (range: 1:10–1:320) and 1:10 (range: 1:1–1:100), respectively. Table 1Baseline demographic, clinical, and treatment characteristics of enrolled patients with anti‐NMDAR or anti‐LGI1 encephalitis.VariableAnti‐NMDAR encephalitis (n = 10)Anti‐LGI1 encephalitis (n = 8)Age (years)34.6 ± 16.9 (18–67)51.8 ± 18.0 (18–69)Gender, n (n/N)Female8 (8/10)3 (3/8)Male2 (2/10)5 (5/8)Time from onset to efgartigimod treatment (days), median (range)14.5 (5–29)15.5 (3–30)Infectious prodrome, n (n/N)3 (3/10)0 (0)Co‐existing autoimmune disease, n (n/N)0 (0)0 (0)Behavioral or personality changes, n (n/N)5 (5/10)1 (1/8)Cognitive deficits, n (n/N)2 (2/10)4 (4/8)Seizures, n (n/N)8 (8/10)6 (6/8)Generalized seizures6 (6/10)3 (3/8)Focal seizures2 (2/10)1 (1/8)FBDS0 (0)2 (2/8)Decreased level of consciousness, n (n/N)4 (4/10)1 (1/8)Movement disorders, n (n/N)1 (1/10)2 (2/8)Dystonia1 (1/10)0 (0)Tremor0 (0)0 (0)Tics0 (0)1 (1/8)Myoclonus0 (0)1 (1/8)Other presentations, n (n/N)Aphasia1 (1/10)0 (0)Paresthesia0 (0)1 (1/8)Efgartigimod treatment, n (n/N)Four doses3 (3/10)2 (2/8)Three doses4 (4/10)4 (4/8)Two doses3 (3/10)2 (2/8)Additional immunotherapy, n (n/N)8 (8/10)4 (4/8)Additional immunotherapy agent*, n (n/N)Oral prednisone (<60 mg qd)7 (7/10)4 (4/8)Tacrolimus (1 mg bid)1 (1/10)0 (0)Mycophenolate mofetil (0.5 g bid)2 (2/10)1 (1/8)Abbreviations: anti‐LGI1, anti‐leucine‐rich glioma‐inactivated 1; anti‐NMDAR, anti‐N‐methyl‐d‐aspartate receptor; FBDS, faciobrachial dystonic seizures.*Three patients received two additional immunotherapy agents; therefore, the total number of agent administrations exceeds the number of patients. Baseline demographic, clinical, and treatment characteristics of enrolled patients with anti‐NMDAR or anti‐LGI1 encephalitis. Abbreviations: anti‐LGI1, anti‐leucine‐rich glioma‐inactivated 1; anti‐NMDAR, anti‐N‐methyl‐d‐aspartate receptor; FBDS, faciobrachial dystonic seizures. Three patients received two additional immunotherapy agents; therefore, the total number of agent administrations exceeds the number of patients. Table 2Clinical scores and antibody titers at baseline and last follow‐up in patients with anti‐NMDAR or anti‐LGI1 encephalitis.VariablesAnti‐NMDAR encephalitis (n = 10)Anti‐LGI1 encephalitis (n = 8) Serum anti‐NMDAR/LGI1‐IgG. n (n/N) Baseline10 (10/10)8 (8/8)Last visit9 (9/10)7 (7/8) CSF anti‐NMDAR/LGI1‐IgG. n (n/N) Baseline10 (10/10)8 (8/8)Last visit3 (3/10)0* Serum anti‐NMDAR/LGI1‐IgG positivity. n (n/N) Baseline8 (8/10)8 (8/8)Last visit6 (6/9)5 (5/7) CSF anti‐NMDAR/LGI1‐IgG positivity. n (n/N) Baseline10 (10/10)8 (8/8)Last visit2 (2/3)‐ Serum anti‐NMDAR/LGI1‐IgG titer Baseline1:100 (1:10–1:320)1:66 (1:10–1:1000)Last visit1:10 (1:10–1:100)1:10 (1:10–1:100) CSF anti‐NMDAR/LGI1‐IgG titer Baseline1:10 (1:1–1:100)1:6.6 (1:1–1:30) mRS Baseline3 (1–5)1 (1–5)Last visit1 (0–3)1 (0–1) CASE Baseline7.5 (1–27)3 (1–7)Last visit1.5 (1–7)1.5 (0–2) GCS Baseline12 (3–15)15 (6–15)Last visit15 (12–15)15 (13–15)Abbreviations: anti‐LGI1, anti‐leucine‐rich glioma‐inactivated 1; anti‐NMDAR, anti‐N‐methyl‐d‐aspartate receptor; CASE, Clinical Assessment Scale in Autoimmune Encephalitis; CSF, cerebrospinal fluid; GCS, Glasgow Coma Scale; mRS, modified Rankin Scale.*Missing data indicate that the assessment was not performed at that follow‐up visit. Clinical scores and antibody titers at baseline and last follow‐up in patients with anti‐NMDAR or anti‐LGI1 encephalitis. Abbreviations: anti‐LGI1, anti‐leucine‐rich glioma‐inactivated 1; anti‐NMDAR, anti‐N‐methyl‐d‐aspartate receptor; CASE, Clinical Assessment Scale in Autoimmune Encephalitis; CSF, cerebrospinal fluid; GCS, Glasgow Coma Scale; mRS, modified Rankin Scale. Missing data indicate that the assessment was not performed at that follow‐up visit. The mean age of patients diagnosed with anti‐LGI1 encephalitis was 51.8 ± 18.0 years (range: 18–69 years), with a female‐to‐male ratio of 3:5. The median time from disease onset to efgartigimod treatment was 15.5 days (range: 3–30). Seizures (6/8) and cognitive decline (4/8) were commonly observed. At baseline, the median scores were 1 for the mRS (range: 1–5), 3 for the CASE (range: 1–7), and 15 for the GCS (range: 6–15). Anti‐LGI1‐IgG antibodies were detected in the serum and CSF of all patients, with median titers of 1:66 (range: 1:10–1:1000) and 1:6.6 (range: 1:1–1:30), respectively. Treatments Five patients (5/18; three with anti‐NMDAR encephalitis and two with anti‐LGI1 encephalitis) received four doses of efgartigimod, eight patients (8/18; four with anti‐NMDAR encephalitis and four with anti‐LGI1 encephalitis) received three doses of efgartigimod, and five patients (5/18; three with anti‐NMDAR and two with anti‐LGI1 encephalitis) received two doses. Twelve patients (12/18, eight with anti‐NMDAR encephalitis and four with anti‐LGI1 encephalitis) received additional immunotherapy after efgartigimod treatment, including oral prednisone (<60 mg per day; n = 11, seven with anti‐NMDAR encephalitis and four with anti‐LGI1 encephalitis), tacrolimus (1 mg twice per day; n = 1, one with anti‐NMDAR encephalitis), and mycophenolate mofetil (0.5 g twice per day; n = 3, two with anti‐NMDAR encephalitis and one with anti‐LGI1 encephalitis; some patients received more than one agent.). The median follow‐up duration for patients with anti‐NMDAR encephalitis was 8 weeks (range: 2–12 weeks), and 2 weeks (range: 1–12 weeks) for patients with anti‐LGI1 encephalitis. The number of patients assessed at each follow‐up time point is detailed in Supplementary Table 3. Notably, the follow‐up duration was shorter in the anti‐LGI1 group than in the anti‐NMDAR group, which should be considered when interpreting the longitudinal outcomes. Efficacy assessments Clinical and imaging improvement Among patients with anti‐NMDAR encephalitis, the median mRS score decreased from 3 before efgartigimod treatment to 1 at the last follow‐up after treatment (Z = −2.588, p = 0.007). A trend toward improvement was observed after two efgartigimod doses (Z = −2.410, p = 0.015, no longer significant after correction for multiple comparisons). The median CASE score decreased from 7.5 to 1.5 post‐treatment (Z = −2.714, p = 0.005). The median GCS score increased from 12 to 15 at the last visit (Z = 2.371, p = 0.015). In patients with anti‐LGI1 encephalitis, the CASE scores decreased from 3 to 1.5 at the last visit (Z = −2.154, p = 0.046). The differences in the mRS and GCS scores were not statistically significant. However, mRS scores either decreased (4/8) or stabilized (4/8) in all eight cases, and the patient with a reduced GCS score (n = 1) showed improvement after efgartigimod treatment. Cognitive assessments were conducted using the MMSE and MoCA in two patients with anti‐NMDAR encephalitis and seven with anti‐LGI1 encephalitis. MMSE and MoCA scores remained unchanged in the anti‐NMDAR encephalitis group after efgartigimod treatment. Among patients with anti‐LGI1 encephalitis, five of seven (71%) demonstrated improved MMSE or MoCA scores; however, these improvements were not statistically significant. In the LGI1 encephalitis group, the median MMSE score increased from 26 (range: 14–29) to 27 (range: 20–30) (Z = −1.590, p = 0.156), and the median MoCA score increased from 26 (range: 17–30) to 28 (range: 23–30) (Z = −1.782, p = 0.093) (Figures 2 and 3). Changes in clinical scores (mRS, CASE, GCS scores) and serum antibody titers from baseline to the last follow‐up in patients with anti‐NMDAR and anti‐LGI1 encephalitis. Paired comparisons of mRS, CASE, GCS, and serum anti‐NMDAR‐IgG or anti‐LGI1‐IgG titers before efgartigimod treatment (baseline) and at the final follow‐up visit. Each line connects individual patient data points. The median values are indicated by horizontal bars. Statistical significance was assessed using the Wilcoxon signed‐rank test. Sample sizes: anti‐NMDAR group, n = 10; anti‐LGI1 group, n = 8. anti‐LGI1, anti‐leucine‐rich glioma‐inactivated 1; anti‐NMDAR, anti‐N‐methyl‐d‐aspartate receptor; CASE, Clinical Assessment Scale in Autoimmune Encephalitis; GCS, Glasgow Coma Scale; IgG, immunoglobulin G; mRS, modified Rankin Scale; NS, not significant. *p < 0.05, **p < 0.01, ***p < 0.001. Longitudinal changes in mRS, CASE, and GCS scores over the 12‐week follow‐up period. Patients received intravenous efgartigimod at 10 mg/kg weekly for 2–4 doses. Clinical assessments were performed at baseline (Week 0) and at weeks 1, 2, 3, 8, and 12 after the first dose. Lines represent median values with maximum and minimum ranges; individual patient data points are overlaid. The number of patients assessed at each time point is detailed in Supplementary Table 3. CASE, Clinical Assessment Scale in Autoimmune Encephalitis; GCS, Glasgow Coma Scale; mRS, modified Rankin Scale. For the LGI1 group (panels D–F), the lower number of data points at week 3 compared with week 8 reflects that some patients missed the earlier visit but returned for later follow‑up. MRI and 18F‐DPA714 PET imaging revealed changes in four patients with different clinical manifestations, before and after treatment. T2‐FLAIR signals improved or disappeared in patients with pre‐existing hyperintensities 3 months after treatment. Moreover, the abnormal 18F‐DPA714 uptake markedly decreased following treatment, regardless of T2‐FLAIR hyperintensities. Patient No. 3, who was diagnosed with anti‐NMDAR encephalitis, exhibited behavioral or personality changes and seizures, with elevated 18F‐DPA714 uptake in the left frontal lobe (standardized uptake value ratio (SUVR) = 1.31). After efgartigimod treatment, the uptake in this region decreased (SUVR = 1.22). Patient No. 5, who also had anti‐NMDAR encephalitis, demonstrated high 18F‐DPA714 uptake in the right mesial temporal lobe (SUVR = 1.46) after epilepsy onset. This value decreased to 1.09 post‐treatment, accompanied by the relief of symptoms. Patient No. 10, who had anti‐LGI1 encephalitis, presented with cognitive impairment and seizures, and showed increased uptake in the right mesial temporal lobe (SUVR = 1.55). The uptake in this region significantly decreased (SUVR = 0.67) 3 months after efgartigimod initiation. Patient No. 13, who also had anti‐NMDAR encephalitis, exhibited impaired consciousness and elevated baseline 18F‐DPA714 uptake in the brainstem (SUVR = 1.96). At the 3‐month follow‐up,18F‐DPA714 uptake had markedly decreased (SUVR = 1.49) (Table 3, Figure 4). Table 3Regional brain 18F‐DPA714 standardized uptake value ratios (SUVRs) at baseline in patients with anti‐NMDAR or anti‐LGI1 encephalitis.Patient ID numberType of encephalitisFrontal lobe (L)Frontal lobe (R)Parietal lobe (L)Parietal lobe (R)Temporal lobe (L)Temporal lobe (R)Mesial temporal lobe (L)Mesial temporal lobe (R)Occipital lobe (L)Occipital lobe (R)Cerebellum (L)Cerebellum (R)BrainstemNo.1Anti‐NMDAR1.381.401.381.401.171.291.601.791.311.261.311.401.69No.2Anti‐NMDAR1.111.141.211.161.070.931.421.231.231.161.041.121.42No.3(1st)Anti‐NMDAR1.311.081.261.311.401.691.291.601.791.311.081.111.48No.3(2nd)1.221.251.161.041.121.420.931.421.231.231.131.151.57No.4Anti‐LGI11.151.171.171.171.171.111.091.261.401.321.421.231.55No.5(1st)Anti‐NMDAR1.181.171.341.201.401.691.121.461.291.410.900.971.24No.5(2nd)0.971.011.321.421.231.551.111.091.261.400.850.901.19No.6(1st)Anti‐LGI10.760.810.921.030.780.950.890.810.910.900.800.861.00No.6(2nd)1.081.131.041.120.890.960.890.881.061.021.010.980.97No.7Anti‐NMDAR1.271.331.331.321.121.121.461.291.411.341.201.401.69No.8Anti‐NMDAR0.910.941.111.131.071.070.840.841.181.170.880.790.67No.9Anti‐LGI11.201.231.251.231.111.181.251.231.371.301.391.271.40No.10(1st)Anti‐LGI11.291.331.321.421.111.091.231.551.261.401.251.371.57No.10(2nd)1.151.181.170.881.070.840.790.670.841.181.301.381.46No.11Anti‐LGI11.041.061.191.081.000.981.081.011.141.060.940.951.24No.12(1st)Anti‐NMDAR1.091.121.121.060.920.891.211.211.211.091.231.211.45No.12(2nd)1.011.041.091.030.890.971.121.121.141.071.071.031.38No.13(1st)Anti‐NMDAR1.101.141.060.940.951.240.981.081.011.141.101.161.96No.13(2nd)1.061.131.261.171.261.611.021.331.281.410.940.991.49No.14Anti‐LGI11.331.391.461.281.131.021.331.281.411.261.171.261.61No.15Anti‐LGI11.191.221.21.21.041.071.281.281.281.21.0411.46No.16(1st)Anti‐LGI10.860.830.970.990.730.770.930.900.920.950.730.751.03No.16(2nd)1.081.091.091.090.820.891.151.111.111.081.000.971.34No.17Anti‐NMDAR1.031.031.071.070.860.891.081.051.091.070.840.821.09No.18Anti‐NMDAR1.291.321.341.531.091.511.561.381.431.271.181.221.49Abbreviations: anti‐LGI1, anti‐leucine‐rich glioma‐inactivated 1; anti‐NMDAR, anti‐N‐methyl‐d‐aspartate receptor; L, left; R, right; SUVR, standardized uptake value ratio. Regional brain 18F‐DPA714 standardized uptake value ratios (SUVRs) at baseline in patients with anti‐NMDAR or anti‐LGI1 encephalitis. Abbreviations: anti‐LGI1, anti‐leucine‐rich glioma‐inactivated 1; anti‐NMDAR, anti‐N‐methyl‐d‐aspartate receptor; L, left; R, right; SUVR, standardized uptake value ratio. Brain MRI and 18F‐DPA714 PET before and after efgartigimod treatment in four representative patients with anti‐NMDAR or anti‐LGI1 encephalitis. Paired brain MRI sequences (T1WI, T2 FLAIR, and DWI) and 18F‐DPA714 PET scans acquired at baseline (pre‐treatment) and 3 months after efgartigimod initiation. (A) Patient No. 3 (anti‐NMDAR encephalitis) presented with behavioral or personality changes and seizures. Abnormal signals were observed in the left frontal cortex on pre‐treatment T2 FLAIR and DWI sequences, with increased uptake in the same region on 18F‐DPA714 PET (SUVR[left frontal lobe] = 1.31). At the 3‐month follow‐up, the abnormal MRI signals and elevated 18F‐DPA714 uptake (SUVR[left frontal lobe] = 1.22) in the left frontal cortex had resolved. (B) Patient No. 5 (anti‐NMDAR encephalitis) presented with seizures. Pre‐treatment MRI showed hyperintensity in the right mesial temporal lobe on T2 FLAIR, with increased uptake in the same area on 18F‐DPA714 PET (SUVR[right mesial temporal lobe] = 1.46). Three months after treatment, the T2 FLAIR hyperintensity had disappeared, and 18F‐DPA714 uptake in the right mesial temporal lobe had decreased (SUVR[right mesial temporal lobe] = 1.09). (C) Patient No. 10 (anti‐LGI1 encephalitis) presented with cognitive impairment and seizures. No apparent abnormalities were detected on any MRI sequences, but 18F‐DPA714 PET showed increased uptake in the right mesial temporal lobe (SUVR[right mesial temporal lobe] = 1.55). Three months after initiating efgartigimod treatment, uptake in this region had significantly decreased (SUVR[right mesial temporal lobe] = 0.67). (D) Patient No. 13 (anti‐ NMDAR encephalitis) presented with impaired consciousness. Pre‐treatment imaging showed abnormal signals in the right pons on T2 FLAIR and DWI sequences, with elevated 18F‐DPA714 uptake in the same region (SUVR[brainstem] = 1.96). At the 3‐month follow‐up,18F‐DPA714 uptake in the right pons had markedly decreased (SUVR[brainstem] = 1.49), accompanied by attenuation of T2 FLAIR abnormalities and resolution of the DWI signal changes. anti‐LGI1, anti‐leucine‐rich glioma‐inactivated 1; anti‐NMDAR, anti‐N‐methyl‐d‐aspartate receptor; DWI, diffusion‐weighted imaging; FLAIR, fluid‐attenuated inversion recovery; MRI, magnetic resonance imaging; PET, positron emission tomography; SUVR, standardized uptake value ratio; T1WI, T1‐weighted imaging. We compared patients receiving efgartigimod with oral glucocorticoids or other oral immunosuppressants to those receiving efgartigimod monotherapy to distinguish the therapeutic effects of efgartigimod from those of concomitant medications. This analysis demonstrated no significant difference in the changes (Δvalues) of mRS, CASE, or GCS scores from baseline to the final follow‐up between the two groups (Figure 5). Comparison of clinical improvement between efgartigimod monotherapy and combination therapy with glucocorticoids or immunosuppressants. Delta (Δ) values represent the change from baseline to the last follow‐up, defined as improvement: for mRS and CASE, Δ = baseline score ‐ follow‐up score (positive values indicate improvement); for GCS, Δ = follow‐up score ‐ baseline score (positive values indicate improvement). Each dot represents an individual patient. Comparisons between monotherapy and combination therapy groups were performed using the Mann–Whitney U test. Sample sizes: anti‐NMDAR group‐monotherapy n = 2, combination therapy n = 8; anti‐LGI1 group‐monotherapy n = 4, combination therapy n = 4. No significant differences were observed between the two treatment strategies for any of the three clinical scales, indicating that the therapeutic effect of efgartigimod was independent of concomitant medication use. anti‐LGI1, anti‐leucine‐rich glioma‐inactivated 1; anti‐NMDAR, anti‐N‐methyl‐d‐aspartate receptor; CASE, Clinical Assessment Scale in Autoimmune Encephalitis; GCS, Glasgow Coma Scale; mRS, modified Rankin Scale; NS, not significant. Change in antibody titers All patients with anti‐NMDAR encephalitis underwent baseline serum and CSF testing for anti‐NMDAR‐IgG. Serum anti‐NMDAR‐IgG antibodies were detected in 8 of 10 patients, with a median titer of 1:100 (range: 1:10–1:320). CSF anti‐NMDAR‐IgG antibodies were present in all cases, with a median titer of 1:10 (range: 1:1–1:100). Following efgartigimod treatment, anti‐NMDAR‐IgG was analyzed in the serum of nine patients and in the CSF of three patients. Of these, 6 of 9 remained serum‐positive, and 2 of 3 remained CSF‐positive. The median serum anti‐NMDAR‐IgG titer after efgartigimod treatment was 1:10 (range: 1:10–1:100), which was significantly lower compared to baseline (Z = −2.375, p < 0.01). Median serum total IgG levels decreased from 10.34 g/L (range: 4.68–18.05 g/L) before efgartigimod to 5.70 g/L (range: 2.97–9.88 g/L) (Z = −1.992, p = 0.062). Anti‐LGI1‐IgG antibody levels were measured in the serum and CSF samples of all patients diagnosed with anti‐LGI1 encephalitis. All patients tested positive for serum anti‐LGI1‐IgG antibodies, with a median titer of 1:66 (range: 1:10–1:1000), and antibodies in the CSF, with a median titer of 1:6.6 (range: 1:1–1:30). After efgartigimod treatment, anti‐LGI1‐IgG levels were assessed in the serum of seven patients; however, CSF samples were not available. At the last follow‐up visits, the seropositive rate of anti‐LGI1‐IgG decreased to 71.4% from 100% at baseline, and the median titer declined to 1:10 (range: 1:10–1:100). However, this decrease was not statistically significant. Additionally, median serum total IgG levels significantly decreased from 10.91 g/L (range: 5.72–14.88 g/L) to 5.21 g/L (range: 3.45–8.76 g/L) (Z = −2.197, p = 0.031). Safety assessment During the study period, one patient (1/18) with anti‐LGI1 encephalitis developed mild pneumonia 1 week after receiving two doses of efgartigimod, which resolved with standard antibiotic treatment. No other adverse events, including respiratory tract infection, urinary tract infection, headache, nasopharyngitis, paresthesia, myalgia, infusion‐related reactions, or hypersensitivity reactions, were reported in the entire cohort. No treatment discontinuation due to adverse events occurred except for the aforementioned case, and no serious adverse events or deaths were observed. DISCUSSION This study was the multicenter, single‐arm, prospective, observational real‐world study of the clinical effects, safety, and 18F‐DPA714 PET evaluation of efgartigimod in patients with anti‐NMDAR and anti‐LGI1 encephalitis, enrolling patients from multiple regions across the country. The study excluded patients who had previously received plasma exchange or immunoadsorption within 4 weeks before enrollment, or immunosuppressants or biologics at any time before enrollment (except a short course of corticosteroids or IVIg), and separately analyzed the study data of patients with two antibody types, anti‐NMDAR and anti‐LGI1, to avoid the confounding effect of prior medications and different antibody types on the therapeutic efficacy of efgartigimod. Efgartigimod administration significantly improved GCS, mRS, and CASE scores. Additionally, efgartigimod increased MMSE and MoCA scores in patients with encephalitis and cognitive impairments. Treatment with efgartigimod led to decreases in anti‐NMDAR‐IgG and anti‐LGI1‐IgG serum and CSF levels and seizure frequency. These therapeutic effects were observed after just two doses, highlighting the efficacy of efgartigimod. Some patients discontinued efgartigimod treatment after the second or third dose, primarily due to substantial clinical improvement, complications such as pulmonary infection (in one case), or financial constraints that limited access to all four planned treatments. Dosing decisions were based on clinical response, financial considerations, and adverse events. The comparison of clinical indicators before and after efgartigimod administration demonstrated favorable outcomes, with improvements in mRS (13/18) and CASE scores (16/18). Therapeutic effects of efgartigimod in anti‐NMDAR and anti‐LGI1 encephalitis Previous studies on treating AIE with biologic agents predominantly concentrated on CD20‐targeting therapies, such as rituximab, with multiple studies demonstrating their efficacy. Thus, various guidelines recommend using these agents as a second‐line treatment for AIE. 31 However, their clinical use is limited by adverse events, including an increased risk of allergic reactions or infections. 32 Although other CD20‐targeted immunotargeting drugs, such as ofatumumab, have been used in some patients with AIE, these B‐cell depletion therapies require a longer time to achieve effects. 33 Consequently, they cannot alleviate clinical symptoms or quickly reduce pathogenic antibody levels in the early stages. Thus, these limitations restrict the broader application of immunotargeting drugs in treating AIE. Anti‐NMDAR‐IgG and anti‐LGI1‐IgG are neuronal surface IgG antibodies involved in the pathophysiology of AIE. These antibodies are directly pathogenic, as demonstrated by Qiu et al., who established a humanized mouse model of anti‐NMDAR encephalitis by intraperitoneally injecting patient‐derived peripheral blood mononuclear cells into BALB/c mice. 34 Bauer et al. also identified a feline model of limbic encephalitis with anti‐LGI1 antibodies, called feline complex partial seizures with orofacial involvement, which closely resembles human anti‐LGI1 encephalitis. 35 These findings highlight the importance of reducing antibody levels to manage AIE effectively. Traditional clinical approaches have primarily focused on IVIg, PE, or immunoadsorption to decrease serum antibody titers. 36 , 37 , 38 Approximately half of the patients receiving immunotherapy, including high‐dose IVMP and IVIg, experience improvement, while those unresponsive to these interventions require additional treatments. 39 Furthermore, the limited availability of plasma often impedes the ability to meet clinical demands, which is particularly critical for patients with severe symptoms who require early PE. 40 Therefore, therapies that can rapidly and effectively reduce serum antibody levels during the early stages of the disease to prevent progression and improve clinical outcomes are urgently needed. Efgartigimod functions as an FcRn antagonist, thereby reducing IgG levels by inhibiting antibody recycling. 41 The FcRn receptor protects IgG from lysosomal degradation in an acidic environment by binding to it, thus extending the IgG half‐life. 41 As an FcRn antagonist, efgartigimod decreases the binding affinity of FcRn to IgG, consequently enhancing IgG degradation. 42 Theoretically, this mechanism offers therapeutic potential for IgG‐dependent autoimmune disorders, such as MG, CIDP, stiff‐person syndrome, and immune thrombocytopenia. 43 The efficacy and safety of efgartigimod have been demonstrated in clinical trials involving patients with MG 9 and CIDP. 11 Efgartigimod produced very favorable outcomes among patients with MG who exhibit resistance to corticosteroids and IVIg therapy. 44 Thus, efgartigimod may exhibit therapeutic effects against anti‐NMDAR and anti‐LGI1 encephalitis. At present, efgartigimod has been utilized in only four case reports and case series related to AIE, primarily involving antibody types such as anti‐LGI1, anti‐NMDAR, and anti‐glutamic acid decarboxylase (GAD) antibodies. 12 , 13 , 45 , 46 , 47 However, these are individual case reports, and multicenter studies investigating the use of efgartigimod in these subtypes are lacking. Although intrathecal antibody production is a hallmark of these conditions, circulating antibodies also play a critical role in disease pathogenesis through multiple mechanisms such as translocation across the blood‐brain barrier (BBB), peripheral immune activation, and replenishment of the CNS antibody compartment. Efgartigimod has shown therapeutic efficacy, even in significant intrathecal antibody synthesis conditions. Efgartigimod can decrease peripheral and CNS antibody loads by promoting IgG catabolism, thereby facilitating clinical improvement, despite ongoing intrathecal synthesis. Continued intrathecal antibody production may account for the need for repeated or adjunctive therapies in certain patients, highlighting the need for future studies to evaluate how FcRn antagonism interacts with intrathecal antibody production dynamics. 48 , 49 , 50 , 51 In the present study, patients with anti‐NMDAR and anti‐LGI1 encephalitis exhibited significant decreases in serum antibody levels, which correlated with marked clinical improvements, including reductions in MMSE and CASE scores and elevations in GCS scores among individuals with consciousness disorders. Additionally, MMSE and MoCA scores improved in patients with cognitive impairments. The significant improvement in serum antibody levels further supports the efficacy of efgartigimod in treating these conditions. However, MoCA and MMSE scores in some patients with anti‐NMDAR encephalitis decreased, reflecting the inherent variability of cognitive assessments due to multiple factors such as disease progression, therapeutic response, and intercurrent clinical events. Since this was a real‐world study, the mRS, CASE, and GCS scores were compared at baseline and at the final follow‐up between patients who received efgartigimod in combination with oral glucocorticoids or immunosuppressants and those treated with efgartigimod alone to minimize the impact of concomitant medications on the clinical efficacy of efgartigimod. No statistically significant differences were observed between the two groups, further confirming the efficacy of efgartigimod in treating anti‐NMDAR encephalitis and anti‐LGI1 encephalitis. Examining the correlation between fluctuations in autoantibody titers and clinical symptom improvement in individual patients is crucial for understanding the pathophysiology and therapeutic response in AIE. Evidence indicates that CSF antibody titers may more accurately reflect disease severity and clinical progression than serum titers, in specific AIE subtypes, particularly anti‐NMDAR encephalitis. Gresa‐Arribas et al. demonstrated that CSF antibody titers correlated with clinical outcomes and served as more consistent markers than serum levels during follow‐up. 26 However, this relationship is not invariably linear. Clinical improvement can occur even when antibody titers remain detectable, suggesting that immune modulation may mitigate pathogenic effects without complete antibody clearance. 52 , 53 , 54 The possible explanations include reduced antibody access to the CNS due to the restoration of BBB integrity, alterations in receptor internalization dynamics, immune regulation downstream of antibody binding, and the development of epitope‐specific tolerance. 55 , 56 , 57 , 58 The imaging evaluation of patients showed that individuals who tested positive for anti‐NMDAR and anti‐LGI1 antibodies, among the most detected neuronal surface antibodies in clinical practice, also exhibited negative brain MRI results. This highlights the importance of developing alternative imaging modalities that accurately reflect intracerebral inflammation in AIE. Previous studies suggested that 18F‐DPA714 PET may possess the capability to assess brain inflammation in both antibody‐positive and antibody‐negative AIE patients. 24 In our study, one patient among the four representative cases presented with negative brain MRI results. Nevertheless, regardless of the MRI findings, all patients demonstrated a significant decrease in 18F‐DPA714 uptake 3 months after efgartigimod treatment. These findings suggest that efgartigimod may potentially reduce intracerebral inflammation in AIE, with such improvements effectively monitored through 18F‐DPA714 PET imaging. Safety of efgartigimod in patients with anti‐NMDAR and anti‐LGI1 encephalitis The primary adverse events associated with efgartigimod include an increased risk of infections, 59 headache, and nasopharyngitis. 9 Previous studies reported that the most common events were headache, nasopharyngitis, nausea, diarrhea, upper respiratory tract infections, and urinary tract infections, with the majority being mild or moderate in severity. 9 In the efgartigimod group, 39 (46%) patients experienced adverse events related to infections. 9 In our study, only one patient (1/18) developed mild pneumonia following two doses of efgartigimod, which resolved with standard treatment. No additional adverse events were reported. This preliminary finding suggested the safety of this approach for treating AIE. Limitations The preliminary evidence in this study demonstrated the efficacy and safety of efgartigimod in treating anti‐NMDAR encephalitis and anti‐LGI1 encephalitis. However, this study has several limitations. First, this study was an uncontrolled clinical case series with a limited sample size (n = 18) and no comparator group, such as patients receiving standard first‐line therapy alone or a placebo. The absence of a control group limits our ability to exclude the influence of confounding variables, including the natural progression of anti‐NMDAR and anti‐LGI1 encephalitis, and the impact of concurrent immunosuppressants on the observed improvements in clinical outcomes and antibody levels. We restricted the use of other immunosuppressive therapies prior to treatment to minimize the bias caused by other treatment regimens on the results of this study. Given clinical and humanitarian considerations, we allowed patients to receive oral corticosteroids, tacrolimus, and mycophenolate mofetil following treatment with efgartigimod. Admittedly, this treatment regimen may introduce potential bias. However, we analyzed the differences in scores between patients with and without concomitant oral medication and found no significant differences between the two groups, suggesting that oral medications did not introduce significant bias to the results in this study. Furthermore, since some patients received efgartigimod in combination with oral glucocorticoids, we cannot completely exclude the potential bias introduced by these combination regimens, which makes it challenging to isolate the independent contribution of efgartigimod to the overall treatment effect. Larger randomized controlled trials with standardized protocols are needed to confirm our findings and establish the precise role of efgartigimod in the treatment algorithm for AIE. Second, given the inherent limitations of uncontrolled observational designs, further randomized, controlled, double‐blind, and prospective clinical trials are necessary to validate our findings. Such trials would facilitate robust comparisons of efgartigimod with standard‐of‐care therapies (or placebo) in larger, more diverse cohorts, enabling the quantification of efgartigimod's specific efficacy, optimal dosing, and long‐term safety. Due to the small sample size in this study, no adverse events other than pulmonary infection were observed, as reported in previous studies. 9 , 11 , 59 Additionally, randomized controlled trials would elucidate whether efgartigimod provides incremental benefit when added to existing immunosuppressive regimens and whether its effects vary across AIE subtypes, which are critical steps for integrating these real‐world data into clinical practice guidelines. Third, the assessment of clinical improvement, particularly for symptoms such as seizure reduction and cognitive impairment, relied on clinician‐reported scales and was not quantitatively evaluated using objective measures (e.g., long‐term video‐electroencephalographic monitoring for seizure frequency or standardized, detailed neuropsychological test batteries). The subjective nature of these assessments and the absence of quantitative benchmarks limit the ability to precisely characterize the extent of neurological recovery in these domains. Fourth, the follow‐up duration varied and was relatively short for many patients, with a median follow‐up of 8 weeks (range: 2–12 weeks) for anti‐NMDAR encephalitis and only 2 weeks (range: 1–12 weeks) for anti‐LGI1 encephalitis. As shown in Supplementary Table 3, the number of patients assessed at each time point decreased substantially over the 12‐week study period, with only three of ten anti‐NMDAR patients and one of eight anti‐LGI1 patients completing the full 12‐week follow‐up. This asymmetry, particularly the shorter follow‐up in the anti‐LGI1 group, may underestimate the long‐term efficacy of efgartigimod in anti‐LGI1 encephalitis and limits our ability to assess the durability of treatment response, long‐term safety, and potential late relapses. Future studies with larger sample sizes and standardized, complete follow‐up protocols are necessary to confirm the sustained efficacy and safety of efgartigimod in these patient populations. Fifth, the repeated assessment of clinical scores across multiple time points necessitated Bonferroni correction for pairwise comparisons at individual intermediate time points. Under this conservative threshold (p < 0.01), the nominally significant improvements observed at isolated time points did not remain statistically significant and are therefore reported as trends. These interim findings should be interpreted with caution and require confirmation in future studies with larger cohorts or pre‐specified time‐window analyses. CONCLUSIONS The FcRn antagonist efgartigimod demonstrated efficacy in improving disease severity in patients with anti‐NMDAR and anti‐LGI1 encephalitis. It effectively reduced consciousness impairment, cognitive dysfunction, and seizures while reducing both serum and CSF antibody levels. Notably, consistent improvement was observed, irrespective of concomitant oral prednisone or immunosuppressive therapy. These results suggest efgartigimod as a promising therapeutic option for ameliorating clinical symptoms during the acute phase of AIE. AUTHOR CONTRIBUTIONS Huanyu Meng: Conceptualization (equal); data curation (equal); formal analysis (equal); investigation (equal); methodology (equal); resources (equal); writing—original draft (equal); writing—review and editing (equal). Siqi Tu: Methodology (equal); resources (equal); software (equal); supervision (equal); validation (equal); writing—original draft (equal). Yifan Zhou: Conceptualization (equal); data curation (equal); investigation (equal); methodology (equal); software (equal); validation (equal); visualization (equal); writing—original draft (equal). Ruinan Shen: Methodology (equal); project administration (equal); resources (equal); software (equal); supervision (equal); validation (equal); visualization (equal). Xiaoyu Chen: Formal analysis (equal); methodology (equal); software (equal); supervision (equal); validation (equal); writing—original draft (equal). Zhaozhao Cheng: Supervision (equal); validation (equal); visualization (equal); writing—original draft (equal). Limei Wang: Software (equal); supervision (equal); visualization (equal); writing—original draft (equal). Yan Wu: Resources (equal); software (equal); validation (equal); visualization (equal). Jie Lu: Software (equal); validation (equal); visualization (equal); writing—original draft (equal). Shengjun Wang: Software (equal); supervision (equal); validation (equal); visualization (equal). Yan Xu: Resources (equal); supervision (equal); validation (equal); visualization (equal). Wanhui Lin: Resources (equal); software (equal); validation (equal); visualization (equal); writing—original draft (equal). Sheng Chen: Conceptualization (equal); data curation (equal); funding acquisition (equal); project administration (equal); supervision (equal); writing—review and editing (equal). CONFLICT OF INTEREST STATEMENT The authors declare no conflicts of interest. ETHICS STATEMENT This study was conducted in accordance with the Declaration of Helsinki and the ethical standards of the institutional research committee. The study protocol was reviewed and approved by the Medical Ethics Committee of Ruijin Hospital, Shanghai Jiao Tong University School of Medicine (Approval No. 2024(350)). The study was registered at the National Clinical Research Center (No. MR‐31‐24‐053225). Written informed consent was obtained from all participants or their legally authorized representatives prior to enrollment, and all participants were informed that their anonymized data would be used for research and publication purposes. No identifying information is included in this manuscript. Supporting information Supplementary Material. ACKNOWLEDGMENTS The authors thank all participating patients and their families for their cooperation. We also thank the clinical research coordinators and nursing staff at each participating center for their assistance with patient management and data collection. We would like to thank Jian Li, Ph.D., Epidemiologist and Statistician, Chief Doctor, Clinical Research Center, Ruijin Hospital, Shanghai Jiao Tong University School of Medicine, for his support regarding the statistical calculations in this study. We also acknowledge the technical support from Nanjing Simcere Medical Laboratory Science Co. Ltd. for antibody testing. Meng H, Tu S, Zhou Y, et al. Effectiveness and safety of efgartigimod in patients with anti‐NMDAR or anti‐LGI1 encephalitis and its 18F‐DPA714 PET imaging evaluation: a multicenter, single‐arm, prospective, observational real‐world study. Neuroprotection. 2026;1‐16. 10.1002/nep3.70063 42023270 Huanyu Meng and Siqi Tu contribute equally to this study. Managing Editor: Ningning Wang Contributor Information Wanhui Lin, Email: wanhuilin@fjmu.edu.cn. Sheng Chen, Email: mztcs@163.com. DATA AVAILABILITY STATEMENT The data supporting the findings of this study are available on reasonable request from the corresponding author. Due to ethical and privacy restrictions, including the protection of patient confidentiality, individual‐level data are not publicly shared or deposited in a public repository. Requests for de‐identified data may be directed to the corresponding authors, Professor Wanhui Lin (wanhuilin@fjmu.edu.cn) or Professor Sheng Chen (mztcs@163.com), and will be considered in accordance with institutional data‐sharing policies and applicable ethical regulations. REFERENCES Associated Data Supplementary Materials Supplementary Material. Data Availability Statement The data supporting the findings of this study are available on reasonable request from the corresponding author. Due to ethical and privacy restrictions, including the protection of patient confidentiality, individual‐level data are not publicly shared or deposited in a public repository. Requests for de‐identified data may be directed to the corresponding authors, Professor Wanhui Lin (wanhuilin@fjmu.edu.cn) or Professor Sheng Chen (mztcs@163.com), and will be considered in accordance with institutional data‐sharing policies and applicable ethical regulations.