Lleixà et al. J Neuroinflammation (2021) 18:251 https://doi.org/10.1186/s12974-021-02301-0 RESEARCH Open Access Autoantibody screening in Guillain–Barré syndrome Cinta Lleixà1†, Lorena Martín‑Aguilar1†, Elba Pascual‑Goñi1, Teresa Franco2, Marta Caballero1, Noemí de Luna1,3, Eduard Gallardo1,3, Xavier Suárez‑Calvet1,3, Laura Martínez‑Martínez2, Jordi Diaz‑Manera1,3, Ricard Rojas‑García1,3, Elena Cortés‑Vicente1,3, Joana Turón1,3, Carlos Casasnovas3,5,6, Christian Homedes5, Gerardo Gutiérrez‑Gutiérrez7, María Concepción Jimeno‑Montero7, José Berciano4,8, Maria José Sedano‑Tous8, Tania García‑Sobrino9, Julio Pardo‑Fernández9, Celedonio Márquez‑Infante10, Iñigo Rojas‑Marcos11, Ivonne Jericó‑Pascual12, Eugenia Martínez‑Hernández3,13, Germán Morís de la Tassa14, Cristina Domínguez‑Go nzález3,15, Cándido Juárez2, Isabel Illa1,3 and Luis Querol1,3* Abstract Background: Guillain–Barré syndrome (GBS) is an acute inflammatory neuropathy with a heterogeneous presenta‑ tion. Although some evidences support the role of autoantibodies in its pathogenesis, the target antigens remain unknown in a substantial proportion of GBS patients. The objective of this study is to screen for autoantibodies target‑ ing peripheral nerve components in Guillain–Barré syndrome. Methods: Autoantibody screening was performed in serum samples from all GBS patients included in the Interna‑ tional GBS Outcome study by 11 different Spanish centres. The screening included testing for anti-ganglioside anti‑ bodies, anti-nodo/paranodal antibodies, immunocytochemistry on neuroblastoma-derived human motor neurons and murine dorsal root ganglia (DRG) neurons, and immunohistochemistry on monkey peripheral nerve sections. We analysed the staining patterns of patients and controls. The prognostic value of anti-ganglioside antibodies was also analysed. Results: None of the GBS patients (n = 100) reacted against the nodo/paranodal proteins tested, and 61 (61%) were positive for, at least, one anti-ganglioside antibody. GBS sera reacted strongly against DRG neurons more frequently than controls both with IgG (6% vs 0%; p = 0.03) and IgM (11% vs 2.2%; p = 0.02) immunodetection. No differences were observed in the proportion of patients reacting against neuroblastoma-derived human motor neurons. Reactiv‑ ity against monkey nerve tissue was frequently detected both in patients and controls, but specific patterns were only detected in GBS patients: IgG from 13 (13%) patients reacted strongly against Schwann cells. Finally, we confirmed that IgG anti-GM1 antibodies are associated with poorer outcomes independently of other known prognostic factors. *Correspondence: lquerol@santpau.cat † Cinta Lleixà and Lorena Martín-Aguilar have contributed equally to this work 1 Neuromuscular Diseases Unit, Department of Neurology, Hospital de la Santa Creu i Sant Pau, Universitat Autònoma de Barcelona, Barcelona, Sp ain Full list of author information is available at the end of the article © The Author(s) 2021. Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://​creat​iveco​mmons.​org/​licen​ses/​by/4.​0/. The Creative Commons Public Domain Dedication waiver (http://​creat​iveco​ mmons.​org/​publi​cdoma​in/​zero/1.​0/) applies to the data made available in this article, unless otherwise stated in a credit line to the data. Lleixà et al. J Neuroinflammation (2021) 18:251 Page 2 of 13 Conclusion: Our study confirms that (1) GBS patients display a heterogeneous repertoire of autoantibodies target‑ ing nerve cells and structures; (2) gangliosides are the most frequent antigens in GBS patients and have a prognostic value; (3) further antigen-discovery experiments may elucidate other potential antigens in GBS. Keywords: Guillain–Barré syndrome (GBS), Autoantibodies, Anti-ganglioside, Neurons, Prognosis Background Materials and methods Guillain–Barré syndrome (GBS) is an acute inflamma- Patients and controls tory neuropathy with a heterogeneous presentation Serum samples from 100 GBS patients included in the that includes diverse clinical variants [1–3]. Diagno- Spanish cohort of the International GBS Outcome study sis is based on clinical criteria; diagnostic biomark- (IGOS) [21] were used in this screening. The IGOS is a ers are not available for most patients [4]. The exact multicentre, prospective, observational cohort study that immunopathogenic mechanisms of GBS are rela- investigates factors that determine and predict the clini- tively unknown, but it is considered a paradigmatic cal course, subtype and outcome of GBS [22]; including post-infectious autoimmune disease [5]. Diverse patients fulfilling diagnostic criteria of the National Insti- mechanisms, including humoral and cellular immune tute of Neurological Disorders and Stroke (NINDS) [1] or responses, autoantibodies and complement, activated Miller Fisher syndrome (MFS) and other variants of GBS macrophages and lymphocytes, have been implicated [3, 23]. Patients from the Spanish cohort were enrolled in GBS pathogenesis [6, 7]. between February 2013 and January 2020. All patients Anti-ganglioside antibodies are detected in up to fulfilled diagnostic criteria for GBS and were included half of GBS patients. These autoantibodies arise via within 2 weeks from onset of weakness. Serum samples microbial molecular mimicry [8] and the associa- were aliquoted and stored at − 80 °C until needed. In tion of specific anti-ganglioside antibody reactivities this study, we used serum samples extracted at baseline. and specific disease variants is well-established in the Sixty-two (62%) of the baseline samples analysed were literature [9, 10], particularly the association of anti- collected before starting treatment. GM1 [11] and GQ1b [12] antibodies with the pure Clinical variants were defined as sensorimotor, pure motor and Miller Fisher syndrome (MFS) variants of motor, pure sensory, Miller Fisher syndrome (MFS) and GBS, respectively. In addition, the presence of anti- ataxic. Nerve conduction studies results were classified bodies targeting the GM1 [11] or GD1a [13, 14] gan- as acute inflammatory demyelinating polyneuropathy gliosides has also been associated with GBS prognosis. (AIDP), acute motor axonal neuropathy (AMAN), acute Antibodies against nodal and paranodal proteins (neu- motor-sensory axonal neuropathy (AMSAN), equivo- rofascin 140/186 [15] -NF140/186-, neurofascin 155 cal or normal. The outcome of all patients with GBS at [16, 17] -NF155-, contactin-associated protein 1 [18, 6 months and 1 year from disease onset were assessed 19] -CASPR1-, and contactin 1 [19] -CNTN1-) have using the GBS disability score (GDS), a widely accepted also been described in patients diagnosed of GBS. system for evaluating the functional ability of patients However, the target antigens remain unknown in a [24]. Patients unable to walk independently (≥ 3) at substantial proportion of GBS patients, particularly 6 months were defined as having a poor outcome in this of the acute inflammatory demyelinating polyradicu- study. loneuropathy (AIDP) variant, the most frequent in Additionally, serum samples from a control group patients of European ancestry. (n = 90) including 45 healthy controls and 45 patients Considering the broad clinical and epidemiological with other neuromuscular disorders [23 amyotrophic spectrum of GBS, the diverse infectious triggers and lateral sclerosis (ALS), 22 Charcot–Marie–Tooth (CMT)] the T-cell independent nature of the immune reaction were included. leading to the appearance of autoantibodies [20], we hypothesized that a broad repertoire of autoantibod- ies targeting diverse nerve components may be causing Autoantibody screening protocol nerve pathology in GBS. This study aims to (1) screen Autoantibody screening experiments included anti- for autoantibodies against known antigens; (2) screen ganglioside antibody detection with ELISA, nodo/para- for antibodies against human and rodent nerve cells nodal (NF155, NF140, NF186, CNTN1 and CNTN1/ and monkey nerve tissue; (3) describe the diversity of CASPR1 complex) antibody detection by ELISA and staining patterns and (4) perform clinical–immunolog- cell-based assays, immunocytochemistry using patient ical correlations, in a well-characterized GBS cohort. sera on murine dorsal root ganglia (DRG) neurons and Lleixà et al. J Neuroinflammation (2021) 18:251 Page 3 of 13 neuroblastoma-derived human motor neurons (IgG and (5-FU) (Sigma) were added to the medium to remove IgM) and reactivity pattern assessment by immunohis- fibroblasts. Then, medium was replaced every other day tochemistry on monkey sciatic nerve sections (IgG and until reaching complete growth and differentiation of IgM). DRG neurons. Live DRG neurons were incubated for 1 h with patients’ Testing for nodo/paranodal antibodies sera diluted 1/100 (for IgG experiments) or 1/40 (for IgM Autoantibodies against NF140, NF186, NF155, CNTN1 experiments) in culture medium at 37 °C. Cells were then and CASPR1 were tested by ELISA. fixed for 10 min with 4% PFA and incubated with sec- Maxisorb 96-well ELISA plates (Thermo Fisher Scien- ondary antibodies. Goat anti-human IgG or IgM AF488 tific, NUNC, Denmark) were coated with 1 μg/ml human (Molecular probes, Oregon, USA) were used as second- recombinant CNTN1 protein (Sino Biological Inc., Geor- ary antibodies at 1/1000 concentration. gia, USA), 1 μg/ml NF155 protein (Origene, Maryland, Coverslips were mounted with Vectashield with DAPI USA), 1 μg/ml NF140 protein (Sino Biological), 1 μg/ and fluorescence signal intensity was scored in a 0–3 ml NF186 protein (Origene) or 5 μg/ml CASPR1 pro- scale by two independent researchers. Images were tein (R&Dsystems, MI, USA) overnight at 4 °C. Wells obtained with an Olympus BX51 Fluorescence Micro- were blocked with 5% non-fat milk in PBS 0.1% Tween scope (Olympus Corporation, Tokyo, Japan). 20 for 1 h, incubated with sera diluted 1/100 in blocking buffer for 1 h, and then incubated with peroxidase con- Human neuroblastoma‑derived neurons jugated rabbit anti-human IgG secondary antibody (Inv- immunocytochemistry itrogen, CA, USA) for 1 h at room temperature. ELISA SH-SY5Y cells were plated in glass coverslips coated was developed with tetramethylbenzidine solution (Bio- with laminin at 2.5 µg/ml (Invitrogen). Cells were grown legend, California, USA), and the reaction was stopped in proliferation medium containing DMEM/F12 (1:1), with 25% sulfuric acid. Optical density (OD) was meas- fetal bovine serum (10%), l-glutamine (1%) and Sodium ured at 450 nm in a Multiscan ELISA reader. Samples pyruvate (1%). After 24 h, proliferation medium was were considered positive by ELISA when they had a ΔOD replaced by differentiation medium containing Neuroba- higher than mean healthy control ΔOD plus two stand- sal (Gibco) supplemented with B27 (Gibco), Glutamax ard deviations. (Gibco), nerve growth factor (Invitrogen) and retinoic Cell-based assays were used as previously described acid at 10 µM (Sigma). Then, medium was replaced every [25] as a second confirmatory technique for questionable other day until full differentiation was achieved. On days cases. Briefly, mammalian expression vectors encoding 5 or 6 of differentiation, cells were fixed for 15 min with human NF140, NF186, NF155, CNTN1 or CASPR1 were paraformaldehyde 4%; and blocked with 5% normal goat transfected into HEK293 cells using Lipofectamine 2000 serum in PBS; followed by incubation with patients’ sera (Invitrogen). Cells were then fixed with 4% paraformal- at 1/40 (for IgM) or 1/100 (for IgG). To observe the cor- dehyde and blocked. ICC experiments were performed rect differentiation of the cells we also incubated them using patient’s sera and appropriate primary and second- with chicken anti-panNeurofascin mAb (R&Dsystems) ary antibodies. at 1/200. Goat anti-chicken IgG AF594 and goat anti- human IgG AF488 or goat anti-human IgM AF488 Testing for anti‑ganglioside antibodies (Molecular Probes) were used as secondary antibodies at Patients’ sera were screened for the presence of anti- 1/1000 concentration. ganglioside antibodies using a previously validated ELISA Coverslips were mounted with Vectashield with DAPI protocol [26] as the general detection method, and thin and fluorescence signal intensity was scored in a 0–3 scale layer chromatography [27] for confirmatory experiments. by two independent researchers. Images were obtained Anti-ganglioside antibodies were considered positive at a with an Olympus BX51 Fluorescence Microscope. 1/1000 titre. Peripheral nerve immunohistochemistry Rat dorsal root ganglia neurons immunocytochemistry Macaque peripheral nerve tissue slides (Inova Diagnos- DRG were dissected from E16 rat embryos, dissociated tics, Inc., San Diego, CA) were blocked with 5% normal and plated in glass coverslips coated with laminin (Inv- goat serum in PBS; followed by incubation with patients’ itrogen) and poly-d-lysine (Sigma, MO, USA). Cells were sera at 1:10 (for IgM) or 1:20 (for IgG). Monkey-adsorbed grown in neurobasal medium (Gibco BRL, NY, USA) goat anti-human IgG AF488 (Southern Biotech, Alabama, supplemented with B27 (Gibco), Glutamax (Gibco) and US) or goat anti-human IgM AF488 (Molecular Probes) nerve growth factor (NGF) (Invitrogen). After 24 h, were used as secondary antibodies at 1/500 concentra- cytosine arabinoside (ARA-C) (Sigma) and fluorouracil tion. Finally, slides were mounted with Fluoromount Lleixà et al. J Neuroinflammation (2021) 18:251 Page 4 of 13 medium (Sigma) and examined by two independent evaluated in our previously reported multivariable logis- observers. Immunostaining patterns were analysed scor- tic regression analysis [29]. ing fluorescence signal intensity of each nerve struc- Odds-ratios (OR) for the logistic regression analy- ture in a 0–3 scale. The nerve structures analysed were: sis were reported with 95% confidence intervals and p nodes or paranodes, myelin from small myelinated fibres, values. myelin from large myelinated fibres, Schwann cells from unmyelinated fibres, large-fibre axons, and small-fibre Results axons. Reactivity against other non-nerve structures Baseline characteristics (fibroblasts, connective tissue, vessels) was not consid- We included 100 participants from 11 Spanish cen- ered in the analysis. tres participating in the IGOS study. GBS patients had To further study the staining patterns, peripheral nerve an average of 57.4 years and were predominantly men tissue slides were coated with mouse anti-human CD56 (57%). 65% of patients presented with the sensorimotor antibody (Becton Dickinson, New Jersey, USA) at 1:50 to variant, 19% presented with the pure motor GBS variant, stain non-myelinating Schwann cells (Remak bundles); 10% with MFS, 5% with the pure sensory variant and 1 or with rabbit anti-human S100 antibody (Abcam, Cam- patient with the ataxic variant. Regarding nerve con- bridge, UK) at 1:50 to stain myelinating Schwann cells. duction studies, 59% of patients were classified as AIDP, Goat anti-mouse IgG AF594 (for CD56), and goat anti- 12% as AMAN, 7% as AMSAN, 8% as normal, and 14% rabbit IgG AF594 (for S100) were used as secondary anti- as equivocal. Detailed epidemiological features of the bodies at 1/500 concentration. cohort were described elsewhere [29]. Images were acquired using Leica TSC SP5 confocal microscope. Screening for known autoantibodies None of the GBS patients included in the study reacted against the paranodal and nodal proteins tested (NF155, Statistical analysis NF140, NF186, CNTN1 and CASPR1). Results were analysed by GraphPad Prism v8.0 (Graph- Sixty-one patients tested positive for, at least, one anti- Pad Software). Statistical comparison of proportions ganglioside antibody (GM1, GM2, GM3, GD1b, GD3, among groups was performed using contingency analy- aGM1, GT1a, GT1b and GQ1b). Of these, 40 had IgG sis with the application of Chi-square and a two-tailed antibodies, 3 had IgM antibodies, and 18 had antibodies Fisher’s exact test, accepting an alpha-level < 0.05 for sta- from both isotypes. Detailed anti-ganglioside reactivities tistical significance. To represent the results and perform are shown in Additional file 1. hierarchical clustering of the results heatmap diagrams Most frequent anti-ganglioside antibodies in our using the Clustvis web tool were performed [28]. cohort were aGM1, GM1, GD1b and GQ1b. Overall, IgG To investigate the association between anti-gangli- anti-aGM1 antibodies were detected in 40% of patients; oside antibodies and prognosis a multivariable logis- IgG and IgM anti-GM1 antibodies were detected in 27% tic regression analysis to predict the inability to walk at and 15% of patients, respectively, IgG anti-GD1b anti- 6 months and at 1 year of follow-up (GDS ≥ 3) was per- bodies in 30%, and IgG anti-GQ1b antibodies in 21% formed using the STATA software. A stepwise backward patients. regression modelling to select variables independently associated with the outcome was performed first. The Antibodies targeting peripheral nerve neurons variables introduced in our initial multivariable models ICC experiments with primary cultures of rat DRG neu- were selected based on known prognostic factors: age, rons and human motor neurons derived from a neuro- initial GDS, diarrhoea, AMAN, serum NfL levels (ana- blastoma cell line were used to identify novel IgG and lysed in a previous study with the same cohort) [29], IgM reactivities against peripheral nerve neurons. The serum anti-GM1 IgG antibodies and serum anti-GD1a screening was performed in 100 serum samples from IgG antibodies [29–32]. To perform the multivariable GBS patients and 90 serum samples from a control analysis patients with MFS were excluded, because our group (including healthy controls and patients with other aim was to predict GBS prognosis and MFS is considered neuromuscular diseases). ICC results were grouped in a different disease, including different pathophysiology, three separate categories: moderate-to-strong positives clinical presentation (it does not present with weakness), (including scores 2 and 3), all positives (including scores treatment (often untreated) and outcome (considered 1, 2 and 3), and negatives (score 0). Detailed results are self-limiting and benign). Finally, the ability of the vari- shown in Additional file 2, Fig. 1 and Table 1. able “presence of anti-GM1 IgG antibodies” to predict the inability to run at 1 year of follow-up (GDS ≥ 2) was Lleixà et al. J Neuroinflammation (2021) 18:251 Page 5 of 13 Fig. 1 Reactivity against rat DRG neurons and human neuroblastoma-derived neurons. DRG neurons (A, B) stained with a GBS patient’s serum reacting moderately (score 2) in IgG (A), and a GBS patient’s serum reacting strongly (score 3) in IgM (B). Human neuroblastoma-derived neurons (C, D) stained in red with anti-panNeurofascin mAb, and in green with a GBS patient’s serum reacting moderately (score 2) in IgG (C) and a GBS patient’s serum reacting strongly (score 3) in IgM (D) Table 1 Statistical analysis of DRG and neuroblastoma neurons ICC, and monkey peripheral nerve IHC GBS patients (n = 100) Controls (n = 90/56) p value Any reactivity Strong reactivity Any reactivity Strong reactivity Any reactivity Strong reactivity Neuroblastoma neurons IgG 11 (11%) 2 (2%) 10/90 (11.1%) 0/90 (0.0%) > 0.999 (ns) 0.499 (ns) Neuroblastoma neurons IgM 28 (28%) 8 (8%) 11/90 (12.2%) 2/90 (2.2%) 0.011 (*) 0.105 (ns) DRG neurons IgG 32 (32%) 6 (6%) 6/90 (6.7%) 0/90 (0.0%) < 0.0001 (***) 0.030 (*) DRG neurons IgM 34 (34%) 11 (11%) 22/90 (24.4%) 2/90 (2.2%) 0.156 (ns) 0.020 (*) Monkey peripheral nerve IgG 56 (56%) 17 (17%) 30/56 (53.6%) 3/56 (5.4%) 0.8669 (ns) 0.0455 (*) Monkey peripheral nerve IgM 44 (44%) 10 (10%) 20/56 (35.7%) 6/56 (10.7%) 0.3964 (ns) > 0.999 (ns) Comparison between GBS patients and controls. Strong reactivity includes scores 2 and 3, and any reactivity includes scores 1, 2 and 3. Fluorescence intensity scores were analysed using contingency analysis with the application of a Fisher’s exact test, accepting an alpha-level of < 0.05 to determine significance Overall, 22 (22%) GBS patients reacted moder- were considered, both in IgG (6% vs 0%, p = 0.03) and ately or strongly against DRG or neuroblastoma IgM experiments (11% vs 2.2%, p = 0.02). neurons, whereas 4 (4.4%) controls reacted only mod- In neuroblastoma-derived neuron ICC experiments 28 erately. These differences were statistically significant (28%) samples from the GBS group showed IgM autoan- (p = 0.0005). tibodies; of these 8 (8%) showed moderate or strong Antibodies against DRG neurons appeared significantly reactivity. These proportions were significantly higher more frequently in GBS patients than in controls (32% vs than in the control group (12.2% and 2.2%, respectively; 6.7%, p < 0.0001) taking all positive tests in account; the p = 0.011). Differences in autoantibody proportions same happened if only moderate and strong positives between GBS patients and controls were not observed Lleixà et al. J Neuroinflammation (2021) 18:251 Page 6 of 13 in neuroblastoma-derived neuron experiments when (13%) GBS patients’ IgG had a strong reactivity against assessing IgG antibodies. Schwann cells (myelinating and non-myelinating) while only one of the controls (1.8%) showed strong reactivity Antibodies targeting peripheral nerve tissue against Schwann cells (this difference is statistically sig- We analysed the full GBS cohort and 56 controls. nificant; p = 0.0192). We analysed the staining intensity of six different struc- tures within the nerve, including nodes or paranodes, Combined autoantibody screening analysis myelin from small myelinated fibres, myelin from large We also analysed if GBS patients with or without anti- myelinated fibres, Schwann cells from unmyelinated ganglioside antibodies differed in the reactivity patterns fibres, large-fibre axons, and small-fibre axons. Staining in the peripheral nerve cell and tissue autoantibody patterns can be found in Additional file 3. screening experiments. No differences were found IgG and IgM reactivity against nerve tissue was fre- between those two groups (Additional file 4), suggest- quently detected in GBS patients and controls. Overall, ing that the heterogeneity of the autoantibody repertoire about 70% of GBS patients and controls sera bound to appears even when a specific antigen is found. nerve structures. IgG and IgM from GBS patients reacted We used a heatmap graph to represent all the autoan- moderately in 17 (17%) and strongly in 10 (10%) against tibody screening results performed in our GBS cohort monkey nerve structures. In the control group IgG and (Fig. 3) [28]. This graph provides visual representation of IgM reacted moderately in 8 (14.3%) and strongly in 1 the heterogeneity of the autoantibody repertoire in GBS (1.8%) against monkey nerve structures. The difference sera. between the amount of GBS patients and controls react- ing moderately or strongly against monkey peripheral Clinical correlations nerve was statistically significant (p = 0.0455) only for Among patients with Miller Fisher syndrome, 8/10 (80%) IgG autoantibodies (Table 1). had IgG anti-GQ1b antibodies, whereas in the rest of Differences in IHC patterns of reactivity from GBS GBS patients only 14.4% (13/90) had these antibod- patients and controls were not statistically significant ies, usually in combination with other reactivities. IgG for any of the structures analysed (Additional file 4). anti-GM1 antibodies were more frequently detected in Nonetheless, some specific reactivity patterns were only patients with the pure motor variant than in those with detected in GBS patients and not in controls (Fig. 2). other clinical variants [13/19 (68.4%) vs 14/81 (17.3%)]; Eight (8%) GBS patients’ IgG reacted strongly against and in patients classified electrophysiologically as AMAN myelin, whereas only 2 controls showed weak reactiv- than in the rest of GBS patients (83.3% vs 19.3%). All ity against this structure. Moreover, we observed that 13 these differences were statistically significant (p < 0.0001). Fig. 2 Reactivity against Schwann cells in peripheral nerve sections. Macaque peripheral nerve sections stained in red with S100 (A–D) or CD56 (NCAM) monoclonal antibody (E–H), and in green with GBS patient’s sera reacting against myelin from small myelinated fibres (A, E), myelin from large myelinated fibres (B, F), and non-myelinating Schwann cells (C, G). D and H are stained with sera from negative controls Lleixà et al. J Neuroinflammation (2021) 18:251 Page 7 of 13 Fig. 3 Heatmap showing all the screening performed in GBS patients. Patients and reactivities against anti-ganglioside antibodies, neuroblastoma-derived human motor neurons, murine dorsal root ganglia neurons and monkey peripheral nerve tissue, are ordered according to Euclidean clustering. Each row represents one GBS patient. The score of the anti-ganglioside titre is indicated by the colour of the square (0 =  < 1/1000, 1 = 1/1000–1/2500, 2 = 1/2500–1/12500, 3 =  > 1/12500). The score of staining intensity in the other structures is indicated by the colour of the square (0 = negative, 1 = mild positive, 2 = moderate positive, 3 = strong positive). Columns in the left contain information related to the clinical variant and the outcome at 6 months and at 1 year of follow-up When we analysed the general clinical characteris- outcome at 6 months and 1 year are similar in the two tics of the subgroup of patients with strong IgG reactiv- groups (about 75%). ity against Schwann cells (n = 13), we did not observe In the subgroup of patients with IgG or IgM reactivity any specific pattern that could distinguish them from against DRG neurons (n = 14), we did not find clinical the rest of the cohort. Ten (76.9%) of these patients pre- differences with the whole GBS cohort. Briefly, 71.4% of sented with the sensorimotor clinical variant, whereas patients staining strongly DRG neurons were classified in the general cohort 65% presented this variant; and as AIDP, and 64.3% presented with the sensorimotor the proportions of nerve conduction studies subgroups clinical variant. were similar to those found in the general cohort (53.8% We did not detect any difference in peripheral vs 59% of AIDP electrophysiological variant). Regarding nerve cell and tissue reactivity patterns or frequen- the outcome, the percentage of patients having a good cies between the samples collected before starting the treatment (62%) and those collected after the treat- ment (38%). We neither observed any correlation of Lleixà et al. J Neuroinflammation (2021) 18:251 Page 8 of 13 any of the reactivity patterns with other clinical char- nerves, this may imply that autoantibodies targeting acteristics as pain, ataxia, or the presence of antecedent nerve structures are present in normal human repertoire infection. at lower titers, that they arise as a natural epiphenome- non of a T-cell mediated damage and are not pathogenic, Prognostic value of anti‑ganglioside antibodies or that other autoantibodies, targeting different types of First, we conducted a univariate analysis to select varia- molecules (such as lipids or glycans), for which our tech- bles that were associated with the outcome. Patients with niques are not optimized, are yet to be discovered. serum IgG anti-GM1 antibodies presented poorer out- Whether these autoantibodies arise from a process of comes than patients without the antibodies at 6 months molecular mimicry, or from an unspecific and polyclonal [38.1% vs 16.1% (p = 0.04)], and 1 year [35.3% vs 9.7% activation of pre-existing B cells, remains unclear. The (p = 0.014)]. Anti-GD1a IgG antibodies were not associ- general absence of common patterns suggests the lat- ated with prognosis (Table 2). For the multivariate analy- ter, but the well-established molecular mimicry process sis, we included GM1 IgG, serum NfL levels, diarrhoea, described in anti-ganglioside-associated GBS supports age, and initial GDS. the former. In anti-GM1-associated GBS the sequence We observed that having anti-GM1 IgG antibodies at of pathogenic events includes an immune response baseline was independently associated with the inability to an infection leading to the appearance of antibod- to walk at 1 year of follow-up, after a backward stepwise ies cross-reacting with peripheral nerve and nerve root selection modelling (OR 6.98, 95% CI 1.6–30.36; p = 0.01). gangliosides and triggering post-infectious inflamma- However, the presence of anti-GM1 IgG antibodies was tion [33]. Interestingly, in this screening we did not find not independently associated with having a poor out- clear differences in the reactivity patterns between GBS come at 6 months (Table 2). patients with or without anti-ganglioside antibodies, but To analyse if anti-GM1 titres were associated with the we observed in both groups a higher amount of patients GBS disability score, we performed a linear regression. staining nerve structures than in controls. These find- We did not observe a positive correlation between anti- ings suggest that the immune response in GBS is not body titres and disability at 6 months and at 1 year. restricted to the production of anti-ganglioside anti- Finally, when we included the presence of anti-GM1 bodies, but it is also targeting other peripheral nerve antibodies in our previously reported prognostic study structures. This observation may either reflect the pres- [29], we observed that having anti-GM1 IgG antibodies at ence of a polyclonal, not antigen-driven, reactivation of baseline was associated with the inability to run at 1 year, a pre-existing repertoire that, in some patients, includes but this association was not independent from the other gangliosides, or the concomitant activation (by epitope known prognostic factors and sNfL, age and AMAN spreading or bystander activation) of unspecific B cells remained in the model as independent factors associated in addition to the ganglioside-driven antigen-specific with residual disability at 1 year. response. Previous studies in other inflammatory neuropathies Discussion such as chronic inflammatory demyelinating polyneurop- Our work describes a comprehensive autoantibody athy (CIDP), showed that frequencies of reactivity against screening that provides experimental evidence of the het- DRG neurons in CIDP patients did not differ from erogeneity of the autoantibody repertoire in patients ful- healthy controls [34], in contrast with our results (shown filling GBS diagnostic criteria. in Table 2). GBS and CIDP are similar diseases both Our study shows that GBS patients have a heteroge- clinically and electrophysiologically, so this difference neous repertoire of autoantibodies targeting nerve cells supports the idea that a heterogeneous autoantibody and structures. Except for patients with anti-ganglioside response against multiple nerve antigens arises in GBS antibodies and a minor subset of patients with antibodies while this does not happen in chronic inflammatory and targeting Schwann cells and the myelin sheath, this rep- demyelinating neuropathies in which a specific, antigen- ertoire varies in frequency and intensity of staining, but driven autoantibody response arises, as the recent dis- it is not qualitatively different from controls. Antibodies covery of the nodo-paranodal antibodies supports [35]. targeting peripheral nerve cells of both IgG and IgM iso- Differences in severity of these two diseases may also types are significantly more frequent in patients than in account for this observation. controls, but no clear differences are seen when antibod- We observed that 13% of GBS patients showed ies are tested using immunohistochemistry on monkey strong IgG reactivity against Schwann cells of mon- nerve preparations. Considering that whole nerve mon- key peripheral nerve. This observation is in agree- key preparations likely display protein antigens in a con- ment with previous findings: Kwa et al. observed formation that is phylogenetically closer to that of human that 24% of GBS patients had IgG antibodies against Table 2 Association between baseline anti-GM1 and anti-GD1a antibodies and prognostic Univariate and Multivariable logistic analysis for inability to walk independently at 6 months Variable OR 95% CI p Univariate analysis GM1 IgG 3.2 1.05–9.71 0.04 GD1a IgG 0.48 0.05–4.24 0.515 Lleixà et al. J Neuroinflammation logNFL 3.22 1.46–7.14 0.004 AMAN 2.97 0.73–12.07 0.127 Diarrhoea 3.64 1.21–10.91 0.021 Age 1.05 1.01–1.10 0.005 Initial GDS 3.15 1.38–7.18 0.006 (2021) 18:251 Multivariable logistic analysis logNfL 3.13 1.27–7.67 0.012 Age 1.05 1.01–1.10 0.013 Univariate and Multivariable logistic analysis for inability to walk independently at 1 year Variable OR 95% CI p Univariate analysis GM1 IgG 5.09 1.38–18.73 0.014 GD1a IgG 0.92 0.10–8.44 0.944 logNFL 3.11 1.27–7.55 0.012 AMAN 1.475 0.27–7.97 0.652 Diarrhoea 2.7 0.74–9.82 0.131 Age 1.06 1.01–1.11 0.014 Initial GDS 2.79 1.12–6.93 0.027 Multivariable logistic analysis GM1 IgG 6.98 1.60–30.36 0.01 Age 1.07 1.01–1.12 0.01 Univariate and Multivariable logistic analysis for inability to run independently at 1 year Variable OR 95% CI p Univariate analysis GM1 IgG 4.16 1.39–12.49 0.011 GD1a IgG 0.72 0.13–3.99 0.710 logNFL 4.37 1.95–9.78 < 0.0001 AMAN 6.53 1.57–27.17 0.01 Diarrhoea 2.33 0.83–6.57 0.109 Page 9 of 13 Lleixà et al. J Neuroinflammation (2021) 18:251 Table 2 (continued) Univariate and Multivariable logistic analysis for inability to run independently at 1 year Variable OR 95% CI p Age 1.04 1.01–1.07 0.023 Initial GDS 1.81 1.08–3.04 0.025 Multivariable logistic analysis Age 1.04 1.01–1.08 0.022 AMAN 6.19 1.01–38.02 0.049 logNfL 3.17 1.34–7.50 0.009 logNfL log-transformed neurofilament light chain; AMAN acute motor axonal neuropathy; GDS Guillain–Barré Syndrome Disability Score Page 10 of 13 Lleixà et al. J Neuroinflammation (2021) 18:251 Page 11 of 13 non-myelinating human Schwann cells [36], and Val- the variable “presence of serum IgG anti-GM1 antibod- lat et al. also detected that a significant percentage of ies”, sNfL levels remained as an independent prognos- CIDP and GBS patients (about 25%) presented with IgG tic factor, whereas anti-GM1 antibodies did not. These or IgM reactivity against myelin and that the staining results confirm that sNfL levels are a prognostic factor patterns on Schwann cells were diverse, suggesting that that informs better on axon status and, consequently, on diverse myelin antigens are being recognized by the long-lasting disability. autoantibodies [37]. It is interesting to note that we did not find any Our study also confirms, in a well-characterized GBS patient with anti-nodal/paranodal antibodies (CNTN1, cohort, that gangliosides are the most frequent spe- NF140, NF186, NF155 and CASPR1) in our GBS cohort. cific antigens in GBS patients and that they associate Although previous studies from other authors have found to specific disease variants. The value of testing anti- some GBS positive patients in their cohorts [14–18], and ganglioside antibodies in the GBS routine clinical care case-reports and series describe the association of anti- is controversial, but it is clear that some antibodies are nodal/paranodal antibodies with aggressive inflamma- associated with specific clinical phenotypes [38]. IgG tory neuropathies frequently misdiagnosed as GBS, these anti-GQ1b antibody is a diagnostic marker and a patho- antibodies are rare and we cannot rule out the possibility genic antibody in MFS, and is often cross-reactive with that they are present in other selected patients that our GT1a [9]. Moreover, IgG anti-GM1 antibodies associate cohort failed to capture. with the pure motor (clinical) and AMAN (electrophysi- One of the limitations of our study is the number of ological) variants. Our results, with 80% of MFS patients patients and controls included. We have small groups having anti-GQ1b antibodies and 68.4% of pure motor of patients with similar staining patterns in which it is patients having anti-GM1 antibodies at baseline, con- difficult to establish clear clinical-immunological cor- firm these associations. However, our study lacks power relations. Nevertheless, this is the first large prospec- to find other potential associations previously described tive study assessing the autoantibody repertoire against (anti-GD1b with acute ataxic neuropathy, anti-GT1a and peripheral nerve structures in GBS patients and antigen- pharyngo-cervico-brachial variant) [39, 40] that will need identification experiments will follow in those patients to be confirmed in even larger cohorts. Likewise, the showing specific staining patterns that are absent in clinical relevance of antibodies targeting other structures controls. (neurons, peripheral nerve tissue…) is unclear, since their The existence of clear subgroups associated with anti- association with GBS is not completely specific or the ganglioside antibodies, in contrast with the diversity in number of patients with each particular reactivity is too the new reactivities analysed, suggests that this appar- low to draw any conclusions. ent heterogeneity may be also due to technical caveats, Some studies have reported a correlation between IgG because our study protocol is optimized for proteins and anti-GM1 and anti-GD1a antibodies with a poor out- not for lipids or glycans. Moreover, other, not properly come in GBS patients [13, 14, 33, 41]. In our cohort, IgG controlled factors, could have influenced heterogeneity in anti-GD1a antibodies did not associate to a poor out- staining patterns (treatment, comorbidities…), and will come of the disease [13]. However, our data confirm that need to be assessed in larger cohorts. IgG anti-GM1 antibody is an independent prognostic fac- tor that associates with poor prognosis at 1 year, support- ing that it may be a marker for long-term axonal damage. Conclusions Whether the presence of complement-fixing anti-GM1 In conclusion, our study highlights the heterogeneity of antibodies is the driver of this long-term disability, an the profile of autoantibodies targeting peripheral nerve important therapeutic question (that would enable the structures, confirms gangliosides as the most frequent use of complement inhibitors in these patients), remains target antigens in the GBS autoantibody repertoire and to be elucidated. their prognostic value in long-term GBS prognosis, and Although in this study we analysed the prognostic identifies small subsets of GBS patients with specific value of anti-ganglioside antibodies using the tradi- staining patterns in which further antigen-identification tional outcome measures: inability to walk (GDS ≥ 3) at experiments could demonstrate novel and clinically rel- 6 months and at 1 year, we have recently used the inabil- evant autoantibody reactivities in the future. ity to run (GDS ≥ 2) as a measure of the presence of long- term residual disability. In this recent study we showed Abbreviations that high baseline sNfL were independently associated GBS: Guillain–Barré syndrome; IGOS: International GBS Outcome Study; DRG: with inability to run at 1 year [29]. In agreement with Dorsal-root ganglia; MFS: Miller Fisher syndrome; AIDP: Acute inflammatory these findings, we observed that including in the model demyelinating polyneuropathy; AMAN: Acute motor axonal neuropathy; Lleixà et al. J Neuroinflammation (2021) 18:251 Page 12 of 13 AMSAN: Acute motor–sensory axonal neuropathy; GDS: GBS disability Declarations score; ALS: Amyotrophic lateral sclerosis; CMT: Charcot–Marie–Tooth; NF155: Neurofascin-155; NF140: Neurofascin-140; NF186: Neurofascin-186; CNTN1: Ethics approval and consent to participate Contactin-1; CASPR1: Contactin-associated protein 1; PFA: Paraformaldehyde; The study was approved by the Ethics Committee of the Hospital de la Santa NfL: Neurofilament light chain; OR: Odds-ratio; CIDP: Chronic inflammatory Creu i Sant Pau and the local institutional review boards of participating hos‑ demyelinating polyneuropathy. pitals or universities. All patients gave written informed consent to participate in the study. Animal procedures were performed according to a protocol approved by our Institution’s Animal Ethics’ Committee. Supplementary Information The online version contains supplementary material available at https://​doi.​ Consent for publication org/​10.​1186/​s12974-​021-​02301-0. Not applicable. Competing interests Additional file 1: Figure 1. Heatmap showing anti-ganglioside antibod‑ The authors declare that they have no competing interests. ies in the GBS cohort. Patients and reactivities against anti-ganglioside antibodies are ordered according to Euclidean clustering. Each row Author details represents one GBS patient. The score of the anti-ganglioside titre is 1 Neuromuscular Diseases Unit, Department of Neurology, Hospital de la indicated by the colour of the square (0 =  < 1/1000, 1 = 1/1000–1/2500, Santa Creu i Sant Pau, Universitat Autònoma de Barcelona, Barcelona, Spain. 2 = 1/2500–1/12500, 3 =  > 1/12500). Columns in the left contain informa‑ 2 Immunology Department, Hospital de la Santa Creu i Sant Pau, Universitat tion related to the clinical variant and the outcome at 6 months and at Autònoma de Barcelona, Barcelona, Spain. 3 Centro para la Investigación 1 year of follow-up. Biomédica en Red en Enfermedades Raras (CIBERER), Madrid, Spain. 4 Centro Additional file 2: Figure 2. Staining paterns analized in ICC over rat DRG de Investigación Biomédica en Red en Enfermedades Neurodegenerativas, neurons and human neuroblastoma-derived neurons. DRG neurons (A–H) CIBERNED, Madrid, Spain. 5 Neuromuscular Unit, Department of Neurology, or neuroblastoma neurons (I–P) were stained with control or patient’s sera Bellvitge University Hospital, Barcelona, Spain. 6 Neurometabolic Diseases in green. Signal intensity of IgG (A–D, I–L) or IgM (E–H, M–P) reactivity Group, Bellvitge Biomedical Research Institute (IDIBELL), Barcelona, Spain. 7 was scored in a 0–3 scale (0 = negative, 1 = mild positive, 2 = moderate Department of Neurology, Hospital Universitario Infanta Sofía, Madrid, Spain. 8 positive, 3 = strong positive). Neuroblastoma neurons were counter‑ Department of Neurology, Hospital Universitario Marqués de Valdecilla stained in red with anti-panNeurofascin mAb. (IDIVAL), University of Cantabria, Santander, Spain. 9 Department of Neurol‑ ogy, Hospital Clínico Universitario de Santiago, Santiago, Spain. 10 Depart‑ Additional file 3: Figure 3. Staining paterns analized in IHC over monkey ment of Neurology, Hospital Universitario Virgen del Rocío, Seville, Spain. peripheral nerve. Macaque peripheral nerve transverse sections stained 11 Department of Neurology, Hospital Universitario Reina Sofia, Córdoba, Spain. with CNTN1 positive CIDP patient’s serum reacting against paranodes 12 Department of Neurology, Complejo Hospitalario de Navarra, Pamplona, (A), small fiber axons (B), non-myelinating Schwann cells (C), myelin from Spain. 13 Department of Neurology, Hospital Clínic de Barcelona, Barcelona, small myelinated fibers (D), large fiber axons (E), and myelin from large Spain. 14 Department of Neurology, Hospital Universitario Central de Asturias, myelinated fibers (F). Oviedo, Spain. 15 Neuromuscular Diseases Unit, Department of Neurology, Additional file 4: Table 1. Statistical analysis of structures observed in IHC Research Institute imas12, Hospital Universitario 12 de Octubre, Madrid, Spain. over monkey peripheral nerve. Table 2. Statistical comparison between GBS patients with and without anti-ganglioside antibodies. Received: 14 July 2021 Accepted: 18 October 2021 Acknowledgements The authors would like to acknowledge the Department of Medicine at the Universitat Autònoma de Barcelona and the IGOS consortium for their References support. We also would like to thank all our patients for their patience and col‑ 1. Asbury AK, Cornblath DR. Assessment of current diagnostic criteria for laboration, and the animals that were killed to perform this study. Guillain–Barré syndrome. Ann Neurol. 1990;27:S21–4. 2. Feasby TE, et al. An acute axonal form of Guillain–Barré polyneuropathy. Author’s contributions Brain. 1986;109:1115–26. CL acquired the data, performed the experiments, analysed the data and 3. Wakerley BR, Uncini A, Yuki N. Guillain–Barré and Miller Fisher sín‑ drafted the manuscript; TF performed anti-ganglioside antibodies experi‑ dromes—new diagnostic classification. Nat Rev Neurol. 2014;10:537–44. ments; LMA, EPG, JDM, RRG, ECV, JT, CC, CH, GGG, MCJM, JB, MJST, TGS, JPF, 4. Shahrizaila N, Lehmann HC, Kuwabara S. Guillain–Barré syndrome. Lancet CMI, IRM, IJP, EMH, GMT and CDG acquired samples and data and revised the (London, England). 2021;397:6736. manuscript for intellectual content; MC, NL, EG, XSC, LMM, CJ and II revised 5. Van Den Berg B, et al. Guillain–Barré syndrome: pathogenesis, diagnosis, the manuscript for intellectual content; LQ designed and conceptualized the treatment and prognosis. Nat Rev Neurol. 2014;10:469–82. study, interpreted the data and revised the manuscript for intellectual content. 6. Martín-Aguilar L, Pascual-Goñi E, Querol L. Autoantibodies in All authors read and approved the final manuscript. immune-mediated inflammatory neuropathies. Med Clín (English Ed). 2019;153:360–7. Funding 7. Restrepo-Jiménez P, et al. The immunotherapy of Guillain–Barré syn‑ This work is supported by Fondo de Investigaciones Sanitarias (FIS), Instituto de drome. Expert Opin Biol Ther. 2018;18:619–31. Salud Carlos III, Spain and FEDER under Grant FIS19/01407 and INT20/00080; 8. Green DM. Advances in the management of Guillain–Barré syndrome. by the GBS-CIDP Foundation International; and by the Gofundme campaign Curr Neurol Neurosci Rep. 2002;2:541–8. number 86547262 organized by Adela Gómez. LMA was supported by a 9. Kaida K, Kusunoki S. Antibodies to gangliosides and ganglioside com‑ personal Rio Hortega grant CM19/00042. XSC was supported by a “Sara Bor‑ plexes in Guillain–Barré syndrome and Fisher syndrome: mini-review. J rell” post-doctoral fellowship project “CD18/00195”, funded by Instituto de Neuroimmunol. 2010;223:5–12. Salud Carlos III and co-funded by European Union (ERDF/ESF, “A way to make 10. Emilien D, Hugh W. Diagnostic utility of auto antibodies in inflammatory Europe”/“Investing in your future”). RR-G, NDL, EC-V, JT-S, EG, II and LQ are nerve disorders. J Neuromuscul Dis. 2015;2:107–12. members of the ERN Euro-NMD. 11. Kuwabara S, et al. IgG anti-GM1 antibody is associated with reversible conduction failure and axonal degeneration in Guillain–Barre syndrome. Availability of data and materials Ann Neurol. 1998;44:202–8. The datasets used and/or analysed during the current study are available from the corresponding author on reasonable request. Lleixà et al. J Neuroinflammation (2021) 18:251 Page 13 of 13 12. Willison HJ, Veitch J, Paterson G, Kennedy PGE. Miller Fisher syndrome is 28. Metsalu T, Vilo J. ClustVis: a web tool for visualizing clustering of multivari‑ associated with serum antibodies to GQ1b ganglioside. J Neurol Neuro‑ ate data using Principal Component Analysis and heatmap. Nucleic Acids surg Psychiatry. 1993;56:204–6. Res. 2015;43:W566–70. 13. Yamagishi Y, et al. Serum IgG anti-GD1a antibody and mEGOS predict 29. Martín-Aguilar L, et al. Serum neurofilament light chain predicts long- outcome in Guillain–Barré syndrome. J Neurol Neurosurg Psychiatry. term prognosis in Guillain–Barré syndrome patients. J Neurol Neurosurg 2020. https://​doi.​org/​10.​1136/​jnnp-​2020-​323960. Psychiatry. 2020. https://​doi.​org/​10.​1136/​jnnp-​2020-​323899. 14. Rajabally YA, Uncini A. Outcome and its predictors in Guillain–Barré 30. Hiraga A, et al. Recovery patterns and long term prognosis for axonal syndrome. J Neurol Neurosurg Psychiatry. 2012;83:711–8. Guillain–Barré syndrome. J Neurol Neurosurg Psychiatry. 2005;76:719–22. 15. Devaux JJ, Odaka M, Yuki N. Nodal proteins are target antigens in Guil‑ 31. Kuwabara S, Mori M, Ogawara K, Hattori T, Yuki N. Indicators of rapid clini‑ lain–Barré syndrome. J Peripher Nerv Syst. 2012;17:62–71. cal recovery in Guillain–Barré syndrome. J Neurol Neurosurg Psychiatry. 16. Prüss H, Schwab JM, Derst C, Görtzen A, Veh RW. Neurofascin as target of 2001;70:560–2. autoantibodies in Guillain–Barré syndrome. Brain. 2011;134:1–2. 32. Hadden RDM, et al. Preceding infections, immune factors, and outcome 17. Stengel H, et al. Anti-pan-neurofascin IgG3 as a marker of fulminant auto‑ in Guillain–Barré syndrome. Neurology. 2001;56:758–65. immune neuropathy. Neurol Neuroimmunol NeuroInflamm. 2019;6:1–11. 33. Van Koningsveld R, et al. Infections and course of disease in mild forms of 18. Doppler K, et al. Auto-antibodies to contactin-associated protein 1 Guillain–Barré syndrome. Neurology. 2002;58:610–4. (Caspr) in two patients with painful inflammatory neuropathy. Brain. 34. Querol L, et al. Antibodies against peripheral nerve antigens in 2016;139:2617–30. chronic inflammatory demyelinating polyradiculoneuropathy. Sci Rep. 19. Appeltshauser L. Anti-paranodal antibodies and IgG subclasses in acute 2017;7:1–9. autoimmune neuropathy. Neurol Neuroimmunol Neuroinflamm. 2020. 35. Querol L, Devaux J, Rojas-Garcia R, Illa I. Autoantibodies in chronic inflam‑ https://​doi.​org/​10.​1212/​NXI.​00000​00000​000817. matory neuropathies: diagnostic and therapeutic implications. Nat Rev 20. Schafflick D, Kieseier BC, Wiendl H, Meyer zu Horste G. Novel pathomech‑ Neurol. 2017;13:533–47. anisms in inflammatory neuropathies. J Neuroinflamm. 2017;14:1–17. 36. Kwa MSG, et al. Autoimmunoreactivity to Schwann cells in patients with 21. Jacobs BC, et al. International Guillain–Barré Syndrome Outcome Study: inflammatory neuropathies. Brain. 2003;126:361–75. protocol of a prospective observational cohort study on clinical and 37. Vallat JM, et al. Antibody- and macrophage-mediated segmental demy‑ biological predictors of disease course and outcome in Guillain–Barré elination in chronic inflammatory demyelinating polyneuropathy: clinical, syndrome. J Peripher Nerv Syst. 2017;22:68–76. electrophysiological, immunological and pathological correlates. Eur J 22. Doets AY, et al. Regional variation of Guillain–Barré syndrome. Brain. Neurol. 2020;27:692–701. 2018;141:2866–77. 38. Leonhard SE, et al. Diagnosis and management of Guillain–Barré syn‑ 23. Sejvar JJ, et al. Guillain–Barré syndrome and Fisher syndrome: case drome in ten steps. Nat Rev Neurol. 2019;15:671–83. definitions and guidelines for collection, analysis, and presentation of 39. Pan CL, Yuki N, Koga M, Chiang MC, Hsieh ST. Acute sensory ataxic neu‑ immunization safety data. Vaccine. 2011;29:599–612. ropathy associated with monospecific anti-GD1B IgG antibody. Neurol‑ 24. Hughes RAC, Newsom-Davis J, Perkin G, Pierce J. Controlled trial of ogy. 2001;57:1316–8. prednisolone in acute polyneuropathy. Lancet (London, England). 40. Koga M, Yuki N, Ariga T, Morimatsu M, Hirata K. Is IgG anti-GT1a antibody 1978;312:750–3. associated with pharyngeal-cervical-brachial weakness or oropharyngeal 25. Siles AM, et al. Antibodies against cell adhesion molecules and neural palsy in Guillain–Barre syndrome? J Neuroimmunol. 1998;86:74–9. structures in paraneoplastic neuropathies. Ann Clin Transl Neurol. 41. Kuwabara S, et al. Intravenous immunoglobulin therapy for Guillain–Barré 2018;5:559–69. syndrome ith IgG anti-GM1 antibody. Muscle Nerve. 2001;24:54–8. 26. Willison HJ, et al. Inter-laboratory validation of an ELISA for the determina‑ tion of serum anti-ganglioside antibodies. Eur J Neurol. 1999;6:71–7. 27. Ohanlon GM, et al. Peripheral neuropathy associated with anti-GM2 gan‑ Publisher’s Note glioside antibodies: clinical and immunopathological studies. Autoim‑ Springer Nature remains neutral with regard to jurisdictional claims in pub‑ munity. 2000;32:133–44. lished maps and institutional affiliations. Ready to submit your research ? Choose BMC and benefit from: • fast, convenient online submission • thorough peer review by experienced researchers in your field • rapid publication on acceptance • support for research data, including large and complex data types • gold Open Access which fosters wider collaboration and increased citations • maximum visibility for your research: over 100M website views per year At BMC, research is always in progress. Learn more biomedcentral.com/submissions