Guillain–Barré syndrome associated with immune checkpoint inhibitors exposure: a pharmacovigilance study Authors: Jean Cabon, Mathilde Ducloie, Basile Chretien, Charlotte Arnaud, Joachim Alexandre, Charles Dolladille, Angélique Da Silva Published: 2025 ABSTRACT AbstractBackgroundGuillain-Barré syndrome (GBS) is a life-threatening condition that has been associated with exposure to immune checkpoint inhibitors (ICIs); however, available data remain limited.MethodsWe conducted a retrospective, worldwide, observational analysis of individual case safety reports in VigiBase, the World Health Organization’s pharmacovigilance database. To minimize competition bias, we excluded reports of vaccines and infectious associated with a known or potential risk of GBS. Subsequently, we searched for reports of GBS linked to ICI regimens, whether as monotherapy or dual immunotherapy, from the Food and Drug Administration (FDA) approval date of each agent until 12 February 2024. The primary endpoint of this study was to assess the association between GBS reporting and exposure to ICI regimens (either monotherapy or dual immunotherapy) using disproportionality analysis. This analysis was performed utilizing the Information Component (IC) and its 95% credibility interval lower boundary (IC025).ResultsA total of 412 cases of GBS associated with ICIs were reported in VigiBase. The disproportionality analysis revealed a significant reporting signal between GBS and the anti-CTLA-4 and anti-PD-1 combination therapy (n = 102, IC025 = 4.6), specifically with nivolumab and ipilimumab (n = 100, IC025 = 4.6); anti-CTLA-4 monotherapy (n = 39, IC025 = 3.6) with ipilimumab monotherapy (n = 39, IC025 = 3.6); anti-PD-1 monotherapy (n = 217, IC025 = 3.4), including pembrolizumab (n = 124, IC025 = 3.5), nivolumab (n = 88, IC025 = 3), and cemiplimab (n = 5, IC025 = 1.2); and anti-PD-L1 monotherapy (n = 53, IC025 = 3) with atezolizumab (n = 36, IC025 = 3), durvalumab (n = 11, IC025 = 1.8), and avelumab (n = 6, IC025 = 1.1) in monotherapy. Among cases with available data (n = 123), the median time to onset was 68 days (interquartile range [IQR]: 24.5–119.5), with a shorter delay observed in patients receiving dual immunotherapy compared to those treated with monotherapy. Among the cases for which data was available (n = 242), data recovery or partial recovery was reported in 58.3% (n = 141/242), while a fatal outcome was reported in 9% (n = 21).ConclusionA significant reporting signal of GBS exists with the majority of ICI regimens employed in both monotherapy and dual immunotherapy.Supplementary InformationThe online version contains supplementary material available at 10.1007/s00415-025-13507-0.Keywords: Guillain–Barré Syndrome, Immune Checkpoint Inhibitors, Neurological immune-related adverse events, Peripheral nervous system, Pharmacogivilance database Introduction Immune checkpoint inhibitors (ICIs) have changed the landscape of cancer treatment over the past decade [1–3]. The indications for these therapies continue to expand, encompassing various stages of cancer management (neoadjuvant, adjuvant, and metastatic) [4–6]. A considerable proportion of patients exhibit a response to treatment, notably in advanced melanoma, where the objective response rate among patients who received the study treatment was 58% in the nivolumab plus ipilimumab group, 45% in the nivolumab group, and 19% in the ipilimumab group, following a minimum follow-up period of 60 months. [1] This response has been maintained over time, with five-year progression-free survival rates ranging from 8 to 36%, dependent on the treatment arm. Complete response rates varied from 6 to 22% across different treatment arms. [1] Despite the substantial benefits of ICIs, these treatments may induce immune-related adverse events (irAEs) that can affect all organ systems, particularly the nervous system. A wide spectrum of neurological toxicity has been documented, encompassing both central and peripheral nervous system damage [7]. Neurologic irAEs (n-irAEs) occur in 1–5% of patients treated with ICIs, with the most prevalent toxicities including meningo-encephalitis, myasthenia gravis-like syndrome, and peripheral neuropathies [8]. A systematic review of the literature, incorporating data from 59 trials with a total of 9,208 patients, reported an overall incidence of n-irAEs at 3.8% for those receiving anti-cytotoxic T-lymphocyte antigen-4 (anti-CTLA-4) antibodies, 6.1% for anti-programmed cell death 1 (PD-1) antibodies, and 12.0% for those receiving the combination of both, indicating a potentially higher incidence of n-irAEs. Nonetheless, the incidence of grade 3 and 4 irAEs was found to be less than 1% across all ICIs. [7] Guillain-Barré syndrome (GBS) represents the most common cause of acute flaccid paralysis worldwide and is classified as an immune-mediated polyradiculoneuropathy [9]. GBS and its variant, Miller Fisher syndrome (MFS), present in various clinical subtypes with distinct neurological features and clinical manifestations [10]. This syndrome constitutes both a diagnostic and therapeutic emergency, associated with a mortality rate of approximately 4% and a risk of long-term disability [11–13]. In the majority of cases, the acute onset of neurological symptoms follows an infectious illness [14]. Other potential triggers, including ICI therapy, have been discussed, with several cases documented in the literature [13, 14]. However, the association between GBS and ICI therapy remains unclear, and current data on this topic are limited. Thus, we performed a disproportionality analysis to investigate the association between GBS reporting and ICI exposure. Methods Study design and data sources We conducted a retrospective, worldwide, observational analysis of individual case safety reports (ICSRs) using VigiBase, the World Health Organization’s pharmacovigilance database, which contains over 36 million ICSRs from at least 130 member countries of the World Health Organization's International Drug Monitoring Program. Cases are reported by various sources including healthcare professionals, patients, and pharmaceutical companies, typically following the marketing of ICIs. This study was conducted in accordance with the reporting guidelines for disproportionality analyses in drug safety signal detection utilizing ICSRs in pharmacovigilance (READUS-PV) guidelines (Suppl. Table 1 for READUS-PV recommendations) [17]. Table 1Descriptive analysis of GBS associated with ICIsAnti-PD-1 monotherapyAnti-CTLA-4 monotherapyAnti-PD-L1 monotherapyAnti-CTLA-4 + anti-PD-1Anti-PD-L1 + anti-CTLA-4Overall casesNumber of cases21739531021412Sex, n (%) Available data2003349981381 Female/Male66(33)/134 (67)9(27.3)/24 (72.7)24 (49)/25 (51)39(39.8)/59(60.2)0/1 (100)138(36.2)/243 (63.8)Age, years Available data1512441690285 Median, IQR67 (60–74)69.5 (54–75.25)65 (60–72)60 (50–72)NA67 (57-73)TTO, days Available data69616320123 Median, IQR84 (27–192)94.5(61–117)59(12.25–109.5)53.5 (21–82.25)NA68 (24.5-119.5)Cancer type (%)Available dataN = 184N = 25N = 45N = 92N = 1N = 347 Melanoma55 (29.9)25 (100)1 (2.2)66 (71.7)0 (0)147 (42.4) Non-melanoma skin cancer3 (1.6)0 (0)1 (2.2)2 (2.2)0 (0)6 (1.7) Lung cancer48 (26.1)0 (0)21 (46.7)6 (6.6)0 (0)75 (21.6) Mesothelioma2 (1.1)0 (0)0 (0)1 (1.1)0 (0)3 (0.9) Genitourinary cancer38 (20.7)0 (0)11 (24.4)14 (15.2)1 (100)64 (18.4) Kidney cancer22 (12)0 (0)0 (0)14 (15.2)0 (0)36 (10.4) Bladder cancer13 (7.1)0 (0)11 (24.4)0 (0)1(100)25 (7.2) Prostate cancer3 (1.6)0 (0)0 (0)0 (0)0 (0)3 (0.9) GI cancer4 (2.2)0 (0)6 (13.3)1 (1.1)0 (0)11 (3.2) Gynecologic cancer7 (3.8)0 (0)1 (2.2)0 (0)0 (0)8 (2.3) Hematological cancer9 (4.9)0 (0)1 (2.2)2 (2.2)0 0)12 (3.5) Head and neck cancer8 (4.3)0 (0)0 (0)0 (0)0 (0)8 (2.3) Breast cancer9 (4.9)0 (0)2 (4.4)0 (0)0 (0)11 (3.2) Sarcoma1 (0.5)0 (0)0 (0)0 (0)0 (0)1 (0.3)Location (%)Available dataN = 217N = 39N = 53N = 102N = 1N = 412 North America79 (36.4)28 (71.8)21 (39.6)48 (47.1)1 (100)177 (43) Europe87 (40.1)4 (10.3)15 (28.3)40 (39.2)0 (0)146 (35.4) Asia30 (13.8)2 (5.1)12 (22.6)5 (4.9)0 (0)49 (11.9) Other21 (9.7)5 (12.3)5 (9.4)9 (8.8)0 (0)40 (9.7)Concomitant cancer therapy (%) Chemotherapy16 (7.4)0 (0)9 (17)3 (2.9)0 (0)28 (6.8) Anti-VEGF therapy10 (4.6)0 (0)6 (11.3)0 (0)0 (0)16 (3.9) Anti-MEK and/or anti-BRAF therapy3 (1.4)1 (2.6)1 (1.9)6 (5.9)0 (0)11(2.7) ADC3 (1.4)0 (0)1 (1.9)0 (0)1 (100)5 (1.2)Malignant neoplasm progression at the time of GBS reporting (%) Ntotal10 (4.6)1 (2.6)0 (0)1 (1)0 (0)12 (2.9)Concomitant non-neurological irAEs (%) Ntotal28 (12.9)4 (10.3)4 (7.5)27 (26.5)0 (0)63 (15.3)OutcomesSerious (%)Available dateN = 214N = 37N = 52N = 102N = 1N = 406 Yes 208 (97.2)37 (100)51 (98.1)102 (100)1 (100)399 (98.3) No6 (2.8)0 (0)1 (1.9)0 (0)0 (0)7 (1.7)Seriousness (%)Available dataN = 208N = 37N = 51N = 102N = 1N = 399 Other medically important condition151 (72.6)34 (91.9)18 (35.3)85 (83.3)0 (0)288 (72.2) Caused/prolonged hospitalization120 (57.7)17 (45.9)26 (51)66 (64.7)1 (100)230 (57.6) Life threatening20 (9.6)4 (10.8)7 (13.7)16 (15.7)1 (100)48 (12)GBS evolution (%)Available dataN = 131N = 19N = 32N = 59N = 1N = 242 Recovering + partial recovery74 (56.5)11 (57.9)18 (56.25)38 (64.4)0 (0)141 (58.3) Recovering67 (51.1)10 (52.6)13 (40.6)36 (61)0 (0)126 (52.1) Partial recovery7 (5.3)1 (5.3)5 (15.6)2 (3.4)0 (0)15 (6.2) No recovered47 (35.9)5 (26.3)11 (34.4)17 (28.8)0 (0)80 (33.1) Died10 (7.6)3 (15.8)3 (9.4)4 (6.8)1 (100)21 (8.7)GBS Guillain–Barré syndrome, PD-1 programmed death- 1, CTLA-4 cytotoxic T-lymphocyte antigen 4, PD-L1 programmed death-ligand 1, NA not available, TTO time to onset, N number; IQR interquartile range, VEGF vascular endothelial growth factor, ADC antibody–drug conjugate, MEK mitogen-activated protein kinase kinase, BRAF v-Raf murine sarcoma viral oncogene homolog B, GI gastrointestinal Descriptive analysis of GBS associated with ICIs GBS Guillain–Barré syndrome, PD-1 programmed death- 1, CTLA-4 cytotoxic T-lymphocyte antigen 4, PD-L1 programmed death-ligand 1, NA not available, TTO time to onset, N number; IQR interquartile range, VEGF vascular endothelial growth factor, ADC antibody–drug conjugate, MEK mitogen-activated protein kinase kinase, BRAF v-Raf murine sarcoma viral oncogene homolog B, GI gastrointestinal Selection of cases Initially, we identified and extracted all ICSRs from VigiBase dating from its inception (1967) to 12 February 2024 (ClinicalTrials.gov identifier: NCT04696250), excluding reports related to vaccines and infectious diseases with a known or potential risk of GBS [9] (Suppl. Table 2 for the list of infectious diseases). Reports of vaccines were excluded to limit competition bias, as the majority of these reports were associated with COVID-19 and influenza vaccines, which are potentially linked to a risk of GBS, albeit with controversial data [18–22]. Secondly, to identify GBS reporting associated with ICIs, we applied the narrow scope of the Standardized MedDRA Query (SMQ) "Guillain-Barré syndrome" (Suppl. Table 3 for definition of narrow SMQ), linked to ICIs, from the FDA approval of each agent until December 2023. The ICI regimens included anti-PD-1 antibodies (encompassing cemiplimab, dostarlimab, nivolumab, pembrolizumab, retifanlimab, and toripalimab), anti-programmed death-ligand 1 (PD-L1) antibodies (including atezolizumab, avelumab, and durvalumab), anti-CTLA-4 antibodies (including ipilimumab and tremelimumab), and anti-lymphocyte activation gene 3 (LAG-3) antibodies (limited to relatlimab). ICI regimen monotherapy was defined as treatment with a single ICI. Monotherapy with ICI regimens was defined as treatment with a single ICI, while dual immunotherapy was characterized as the combination of an anti-PD(L)−1 drug with an anti-CTLA-4 drug or with an anti-PD-1 drug plus an anti-LAG-3 agent. [23] (Fig. 1 for study flow chart). Study flow chart. ICSRs individual case safety reports, GBS Guillain–Barré syndrome, PD-1 programmed death- 1, CTLA-4 cytotoxic T-lymphocyte antigen 4, PD-L1 programmed death-ligand 1 Case characteristics We collected administrative data (country, type of reporter, registration date in VigiBase), demographic data (age, sex), and information pertaining to ICI drugs and GBS (type of ICI, ICI indication, time to onset [TTO] of GBS corresponding to the duration between the initiation of ICI treatment and the onset of GBS symptoms, nature and severity of GBS, course of the syndrome, mortality associated with GBS, concomitant oncological treatments, and concomitant non-neurological irAEs, as defined in previous studies) [24]. A serious reaction was defined as an adverse drug reaction leading to death, life-threatening situations, hospitalization or prolonged hospitalization, persistent incapacity or disability, or judged clinically serious by the physician reporting the case. We reported median values and interquartile ranges (IQR) for quantitative variables, and percentages for qualitative variables. Outcomes The primary endpoint of this study was to analyze the association between GBS reporting and ICI regimen exposure, whether monotherapy or dual immunotherapy, utilizing disproportionality analysis. Secondary objectives included performing a disproportionality analysis to examine the association between GBS reporting and each individual ICI drug in cases of significant disproportionality signals identified in our primary analysis; a descriptive analysis of the included cases; an exploratory drug-drug interaction analysis for ICI combination therapy with a significant reporting signal; and comparisons between significant reporting ICI monotherapy regimens and dual immunotherapy versus ICI monotherapy regimens using reporting Odds Ratios (RORs). Statistical analysis The association between GBS and ICIs was evaluated through disproportionality analysis (also referred to as case/non-case analysis), which is the most prevalent method in VigiBase for assessing a suspected association between a drug and an adverse even. The IC is a disproportionality reporting indicator that uses a Bayesian approach. This approach compares the proportion of an adverse event associated with a given drug to the proportion of the same adverse event with all other drugs in VigiBase [25–27] (Suppl.Table 4 for IC formula). A disproportionality analysis was first performed for ICI regimens in monotherapy or dual immunotherapy. A positive IC025 value (IC025 > 0) is the conventional threshold employed in the detection of statistical signals. When an association between GBS and an ICI regimen was identified (with IC025 > 0), a supplementary analysis was conducted on the ICI drugs within this regimen. In instances of a significant reporting signal, a disproportionality analysis was performed to compare the incidence of GBS reporting across different ICI regimens. Disproportionality can also be calculated using ROR when the full database is used as a comparator, and using ROR only when different ICI regimens are used as comparators (Suppl. Table 5 for ROR formula). The lower end of the ROR 95% confidence interval (CI) ≥ 1 is the threshold used to detect a significant statistical signal [23, 28]. A disproportionality analysis was used to assess the impact of the combination of two ICI drugs on GBS reporting. We specifically focused on combinations demonstrating significant statistical signals as identified in the disproportionality analysis. The IC is not calculated in the same manner as used in the primary analysis [29] (Suppl. Table 6 for formula). The statistical interpretation of the IC may be similar to the interpretation of disproportionality outlined previously. This IC may indicate a synergistic reporting effect between the two drugs regarding the adverse event of interest. A synergistic effect refers to the circumstance in which the combined impact of the drugs on adverse event reporting exceeds the cumulative effects observed when administered separately (Suppl. Table 7 for formula). Results The disproportionality analysis demonstrated a significant association between GBS reporting and the anti-CTLA-4 and anti-PD-1 combination therapy (n = 102, IC025 = 4.6), anti-CTLA-4 monotherapy (n = 39, IC025 = 3.6), anti-PD-1 monotherapy (n = 217, IC025 = 3.4), and anti-PD-L1 monotherapy (n = 53, IC025 = 3.0). (Fig. 2 and details in Suppl. Table 8). Forest plot of ICI regimen-related GBS signal detection. IC information component, IC025 and IC975 are the 95% lower- and upper-ends of the credibility interval for the IC, respectively; A positive IC025 value (> 0) denotes an association between a drug and an adverse event For the anti-CTLA-4 and anti-PD-1 combination, we discovered a significant association with nivolumab and ipilimumab (n = 100, IC025 = 4.6). In the context of anti-CTLA-4 monotherapy, we reported a significant association solely with ipilimumab monotherapy (n = 39, IC025 = 3.6). For anti-PD-1 monotherapy, significant associations were noted for pembrolizumab monotherapy (n = 124, IC025 = 3.5), nivolumab monotherapy (n = 88, IC025 = 3), and cemiplimab monotherapy (n = 5, IC025 = 1.2). In the case of anti-PD-L1 monotherapy, significant associations were found with atezolizumab monotherapy (n = 36, IC025 = 3.0), durvalumab monotherapy (n = 11, IC025 = 1.8), and avelumab monotherapy (n = 6, IC025 = 1.1). (Fig. 3 and details in Suppl. Table 9). Forest plot of ICI drug-related GBS signal detection. IC information component, IC025 and IC975 are the 95% lower- and upper-ends of the credibility interval for the IC, respectively; A positive IC025 value (> 0) denotes an association between a drug and an adverse event GBS cases were reported significantly more frequently with anti-CTLA-4 and anti-PD-1 combination therapy compared to anti-PD-1 monotherapy (ROR 2.7, 95% CI 2.1–3.4) or anti-CTLA-4 monotherapy (ROR 1.6, 95% CI 1.1–2.3). Furthermore, GBS cases were reported less frequently with anti-PD-1 monotherapy in comparison to anti-CTLA-4 (ROR 0.6, 95% CI 0.4–0.8); a similar result was observed when comparing anti-PD-L1 to anti-CTLA-4 (ROR 0.6, 95% CI 0.4–0.8). Comparable RORs were noted when contrasting anti-PD-1 monotherapy with anti-PD-L1 monotherapy (ROR 1.1, 95% CI 0.8–1.4). (Fig. 4 and details in Suppl. Table 10). Forest plot of Reporting Odds-Ratios and its 95% confidence interval (CI) for comparison ICI class. ROR Reporting Odds-Ratio, PD-1 programmed death-1, CTLA-4 cytotoxic T-lymphocyte antigen 4, PD-L1 programmed death-ligand 1; The lower end of the ROR 95% confidence interval (CI) ≥ 1 is the threshold used to detect a statistically significant signal In conducting a drug-drug interaction analysis, no significant association was found for the combination of nivolumab and ipilimumab (IC025 = − 3.1). (Suppl. Table 11 for details). We identified 412 cases of GBS associated with ICI therapy (Fig. 5), with the majority corresponding to the MedDRA preferred term ("Guillain-Barré syndrome") (n = 292, 70.9%) (n = 292, 70.9%) (Table 1 and Suppl. Table 12 for case details). The most common locations of reporting were Europe (n = 146, 35.4%) and North America (n = 177, 43%). Most cases with available data (n = 381) were male (n = 243, 63.8%), and the median age of patients with available data (n = 285) was 67 years (IQR: 57–73). Among patients with available data (n = 123), the median TTO was 68 days (IQR: 24.5–119.5), with a shorter median TTO observed in patients treated with a combination of anti-PD-1 and anti-CTLA-4 (53.5 days, IQR: 21–82.25) compared to those treated with anti-PD-1 (84 days, IQR: 27–192), anti-CTLA-4 (94.5 days, IQR: 61–117), and anti-PD-L1 (59 days, IQR: 12.25–109.5) when utilized as monotherapy. The three most prevalent cancers for which data were available (n = 347) included melanoma (n = 147, 42.4%), lung cancer (n = 75, 21.6%), and genitourinary cancer (n = 64, 18.4%). Details of GBS cases. AMAN acute motor axonal neuropathy, AMSAN acute motor-sensory axonal neuropathy, CIDP chronic inflammatory demyelinating polyradiculoneuropathy, GBS Guillain–Barré syndrome, MFS Miller-Fischer syndrome, SIDP subacute inflammatory demyelinating polyneuropathy; preferred terms (PTs). There were no cases corresponding to the following PTs: ascending flaccid paralysis or zika virus associated GBS. There was only 1 case of PT "Bickerstaff encephalitis" and 2 cases of PT "SIDP" More than 50% of ICI-associated GBS reporting was linked to anti-PD-1 monotherapy, primarily pembrolizumab (n = 124) and nivolumab (n = 88). For the remaining cases, 102 were reported with anti-PD-1 and anti-CTLA-4 combination therapy, predominantly involving nivolumab and ipilimumab (n = 100), 53 cases with anti-PD-L1 therapy, with a predominance of atezolizumab monotherapy (n = 36), and 39 cases with anti-CTLA-4, exclusively involving ipilimumab monotherapy. Chemotherapy was implicated in approximately 7% of cases (n = 28), while treatment with anti-vascular endothelial growth factor (VEGF) therapy, anti-mitogen-activated protein kinase kinase (MEK), and/or anti-v-Raf murine sarcoma viral oncogene homolog B (BRAF) therapy, along with antibody–drug conjugates (ADCs), was associated with 3.9%, 2.7%, and 1.2% of cases, respectively. Serious reactions were reported in 98.3% of cases (data available for 406 cases). Focusing on the seriousness criteria (available for 399 cases), we noted hospitalization caused/prolonged in 57.6% (n = 230), other medically significant conditions in 72.2% (n = 288), and life-threatening conditions in 12% (n = 48%). Among cases with available data (n = 242), recovery and partial recovery were reported in 58.3% (n = 141), complete recovery in 52.1% (n = 126), partial recovery in 6.2% (n = 15), no recovery in 33.1% (n = 80), and fatal outcomes related to GBS in approximately 9% (n = 21). Malignant neoplasm progression at the time of GBS reporting was observed in approximately 3%, while concomitant non-neurological irAEs were reported in 15% (n = 63), predominantly thyroiditis (n = 23), colitis (n = 13), hepatitis (n = 10), and myocarditis (n = 8). Discussion To our knowledge, this study comprises the largest cohort of GBS associated with ICIs, encompassing a total of 412 cases. We identified a significant reporting signal for nearly all ICI regimens or dual immunotherapy utilized in clinical practice. These findings are consistent with two pharmacovigilance studies from the Food and Drug Administration Adverse Event Reporting System (FAERS) which, despite including fewer cases, also reported a significant safety signal for pembrolizumab, nivolumab, ipilimumab, atezolizumab, durvalumab, avelumab, and the combination of nivolumab and ipilimumab [30, 31]. However, these pharmacovigilance studies did not assess the association between GBS reporting and ICI regimens, and we identified a novel significant reporting signal for cemiplimab in monotherapy (IC025 = 1.2). Disproportionality analysis was not feasible for dostarlimab, retifanlimab, toripalimab, anti-PD-1 and anti-CTLA-4 combinations, or anti-PD-1 and anti-LAG-3 as these are emerging treatments with limited indications at present. Moreover, no cases were reported for tremelimumab monotherapy and retifanlimab monotherapy, as these are currently utilized in clinical settings solely in conjunction with another ICI [32, 33]. Furthermore, to our knowledge, this is the inaugural study to conduct a drug-drug interaction analysis, wherein we did not observe a potential synergistic association with the combination of nivolumab and ipilimumab (IC025 = −3.1) concerning GBS reporting. Our findings indicate a worldwide male predominance, consistent with various studies on the subject. Melanoma was the cancer most frequently associated with GBS cases, potentially aligning with studies highlighting the phenomenon of molecular mimicry. Indeed, several studies suggest that patients with melanoma may be at an elevated risk for ICI-associated demyelinating polyneuropathy due to shared epitopes between melanocytes and Schwann cells [8]. However, contradictory data exist; for instance, Chen Yan et al. identified a significantly higher odds ratio (OR) for developing n-irAEs in patients with melanoma in univariate analysis (OR 95%: 4.8 [1.6–14.5], p = 0.006), although this finding was not confirmed in multivariate models. [34] In our study, GBS cases were overreported with anti-CTLA-4 and anti-PD-1 combination therapy in comparison to anti-PD-1 monotherapy (ROR 2.7, 95% CI 2.1–3.4) or anti-CTLA-4 monotherapy (ROR 1.6, 95% CI 1.1–2.3). A similar trend was noted by Johnson et al. (ROR 2.0, 95% CI 1.2–3.1 for the combination of PD-1/PD-L1 and CTLA-4 blockade compared to monotherapy) [35]. Nonetheless, a meta-analysis and systematic review involving 11,687 patients reported no significant difference in neurological toxicities between anti-CTLA-4 and anti-PD-1 combination therapy versus anti-CTLA-4 or anti-PD-1 monotherapy, particularly concerning ICI-induced GBS, despite only 8 cases of grade 3–5 GBS documented among 3,131 patients treated with ICIs in seven studies (OR = 3, 95% CI 0.31–29.08; P > 0.05) [36]. In our study, GBS cases were reported less frequently with anti-PD-1 monotherapy compared to anti-CTLA-4 (ROR 0.6, 95% CI 0.4–0.8), and this trend was confirmed when comparing anti-PD-L1 to anti-CTLA-4 (ROR 0.6, 95% CI 0.4–0.8). These findings mirror the observations outlined in Johnson et al.'s study when contrasting anti-PD-1/PD-L1 with anti-CTLA-4 (ROR 0.48, 95% CI 0.32–0.72). [35] Neurologic irAEs typically manifest within the first three months following the initiation of ICIs. [37–40] The median TTO for GBS cases in our study was 68 days (range 0–1011), aligning with Johnson et al. but later than the 38-day (range 0–628) median TTO reported in a previous pharmacovigilance study using the FAERS database; notably, our study encompassed approximately twice as many GBS cases [30, 35]. Moreover, the median TTO of GBS cases was reduced in patients treated with dual immunotherapy compared to monotherapy (53.5 days for anti-PD-1 and CTLA-4 versus 84 days for anti-PD-1, 94.5 days for anti-CTLA-4, and 59 days for anti-PD-L1), suggesting possible simultaneous activation of multiple immune effector pathways. Additional risk factors require identification, including a history of autoimmune diseases, genetic predisposition, or pre-existing immune dysregulation (e.g., paraneoplastic syndrome), as these factors may correlate with an elevated risk of n-irAEs [41, 42]. All of these factors that are potentially involved in the onset of GBS reporting could explain why there is a longer delay compared to cases of GBS associated with infections. However, a better understanding of the mechanisms explaining this difference will require more in-depth studies of pathophysiology [43]. We identified approximately 9% of deaths related to GBS, consistent with findings from a systematic review and meta-analysis focusing on the fatal toxic effects associated with immune checkpoint inhibitors, which reported a fatality rate of 10–17% for n-irAEs [44]. In our study, out of cases (n = 242) with available data, recovery and partial recovery were reported in approximately 58%, yet no data were available regarding the management of GBS. In a systematic review encompassing 33 cases of GBS in patients treated with ICIs, 24 cases (72.7%) recovered or exhibited only mild residual dysfunction. Of these, twenty-seven cases (81.8%) received IVIG treatment, sixteen from the cohort of twenty-seven received a combination of steroid therapy, and one received PLEX [15]. A separate case series and literature review focusing on ICI-induced polyradiculoneuropathy, which included 36 cases, reported favorable treatment responses, principally with steroids or IVIG (administered to over two-thirds of patients) [16]. In our study, concomitant non-neurological irAEs were documented in 15% (n = 63) of cases, a figure likely an underestimation due to the possibility of missing data, with a predominance of thyroiditis (n = 23), colitis (n = 13), hepatitis (n = 10), and myocarditis (n = 8). The study by Okada K reported concomitant non-neurological irAEs in 25%, mainly involving gastrointestinal and endocrine irAEs.. In a recent analysis of 147 cases of n-irAEs, which included 87 cases of peripheral nervous system (PNS) involvement, 51 cases of central nervous system (CNS) involvement, and both systems in 9 cases, a concomitant non-neurological irAE was identified in 65 out of 147 (44.2%) cases, predominantly comprising gastrointestinal, endocrine, and cardiological irAEs [16, 45]. Limitations This study has several limitations common to pharmacovigilance database research. Pharmacovigilance databases are subject to biases, including underreporting and incomplete data. In our worldwide study, we were unable to reconstruct the clinical evaluations leading to the diagnosis of GBS due to insufficient available data on clinical presentation, cerebrospinal fluid (CSF) analysis, and electrophysiological findings to accurately characterize GBS cases. It is possible that the GBS phenotype was confused with ICI-related myositis, particularly in cases of concomitant ICI-related myocarditis (n = 8 in our study) due to the recognized overlap between myositis and myocarditis [46]. We were unable to evaluate the likelihood of a causal relationship between ICIs and GBS on an individual case basis. All cases in VigiBase are self-reported using MedDRA terms; additional clinical data regarding the severity of irAEs, particularly the Common Terminology Criteria for Adverse Events classification and management strategies for toxic effects, are not included. The analysis of VigiBase data is predicated on reported adverse events only; consequently, the incidence rate of ICI-related GBS cannot be derived. Pharmacovigilance analyses facilitate signal detection and formulate hypotheses that require replication, ideally in prospective clinical trials. This study delineates an association between GBS cases and ICIs, identifying a pattern concerning the time to onset (potentially shorter with dual immunotherapy), mortality rates, the progression of GBS, concomitant non-neurological irAEs, and other factors. A vital message from this study is to enhance awareness among physicians involved in the care of these patients, enabling them to manage care as effectively as possible and promptly refer patients to a neurologist, ideally within a network of specialists focused on managing n-irAEs, thereby optimizing care as swiftly as possible to mitigate morbidity and mortality associated with these adverse events. Conclusion A significant reporting signal of GBS exists across the majority of ICI regimens employed in both monotherapy and dual immunotherapy. Supplementary Information Below is the link to the electronic supplementary material. Supplementary file1 (PDF 424 KB) Acknowledgements The authors thank the custom searches team at the Uppsala Monitoring Centre (Uppsala, Sweden) research section for providing the VigiBase (the WHO global database of individual case safety reports), without whom this study would not have been possible. The information presented in this study does not represent the opinion of the Uppsala Monitoring Centre nor the World Health Organization. Abbreviations Antibody drug conjugate Acute motor axonal neuropathy Acute motor-sensory axonal neuropathy V-Raf murine sarcoma viral oncogene homolog B Chronic inflammatory demyelinating polyradiculoneuropathy Cerebrospinal fluid Cytotoxic T-lymphocyte antigen-4 Electromyography Food and drug administration Guillain–Barré syndrome High level term Immune checkpoint inhibitor Individual case safety reports Indoleamine 2,3-dioxygenase Interquartile range Immune-related adverse events Intravenous immunoglobulin Lymphocyte activation gene 3 Medical dictionary for regulatory activities Mitogen-activated protein kinase kinase Miller-Fischer syndrome Neurological immune-related adverse events Programmed death-1 Programmed death-ligand 1 Plasma exchange Reporting Odds-Ratio Subacute inflammatory demyelinating polyneuropathy Standardised MedDRA Queries Time to onset Preferred term Vascular endothelial growth factor Funding This work has not received any funding support. Data availability At this time, data from VigiBase (the WHO pharmacovigilance database) are only available for national pharmacovigilance centers and the Uppsala Monitoring Centre. Public access to overview statistics from VigiBase can be gained through the VigiAccess website, http://www.vigiaccess.org/. Declarations Conflicts of interest The authors declared no conflicts of interest. Ethical approval This study was conducted in accordance with the reporting guidelines for disproportionality analyses in drug safety signal detection utilizing ICSRs in pharmacovigilance (READUS-PV) guidelines References Associated Data Supplementary Materials Supplementary file1 (PDF 424 KB) Data Availability Statement At this time, data from VigiBase (the WHO pharmacovigilance database) are only available for national pharmacovigilance centers and the Uppsala Monitoring Centre. Public access to overview statistics from VigiBase can be gained through the VigiAccess website, http://www.vigiaccess.org/.