Guillain-Barré syndrome following intracranial hemorrhage: a systematic review of case reports Authors: Sai Krishna Vallamchetla, Sai Kumar Reddy Pasya, Gary Gronseth, Long Davalos Published: 2026 ABSTRACT IntroductionGuillain-Barré syndrome (GBS) is an autoimmune polyradiculopathy linked to various triggers. Recently, several case reports described GBS development following intracranial hemorrhage (ICH), posing diagnostic and therapeutic challenges.MethodsPubMed and EMBASE were searched from inception till February 2023 for case reports. Two reviewers independently screened the articles and extracted data using a standardized form. The extracted data were then analyzed narratively. Quality was assessed with the JBI Checklist for Case Reports.ResultsTwenty-three cases from 11 countries were identified (14 males, mean age of 60.1 years). Onset of GBS occurred after a median of 9 days (range 2–24) post-ICH, with intracerebral hemorrhage being the most common antecedent (39.1%), followed by subarachnoid (26.1%) and subdural hemorrhage (21.7%). Electrophysiology showed predominantly axonal variants, while CSF consistently demonstrated albuminocytologic dissociation with positive antiganglioside antibodies in five cases. The clinical course was severe, marked by rapid progression to nadir (median 2.5 days), profound functional impairment (average Hughes Functional Grading Scale: 4.57 ± 0.66), and a high rate of respiratory failure requiring mechanical ventilation (78.3%). Axonal variants were associated with higher rates of mechanical ventilation and poorer outcomes than demyelinating forms. Among the 21 patients who received immunomodulatory treatment, 3 (14.3%) showed complete recovery, 13 (61.9%) had partial recovery, 3 (14.3%) had a poor recovery, and 2(9.5%) died at last follow up.ConclusionGBS may complicate recovery after ICH and appears to present with an aggressive course, particularly in axonal variants. Clinicians should consider GBS when patients with ICH develop progressive, symmetrical weakness unexplained by the initial brain injury, as prompt initiation of immunotherapy may improve outcomes. Introduction Guillain-Barré syndrome (GBS) is an acute autoimmune disorder of the peripheral nervous system characterized by progressive, bilateral, and symmetrical limb weakness, with generalized hyporeflexia or areflexia, which maybe further complicated by respiratory failure and autonomic dysfunction (1). Although the etiology of GBS is not fully understood, it is widely acknowledged that various infectious agents, such as Campylobacter jejuni, Mycoplasma pneumoniae, Cytomegalovirus, Epstein–Barr virus and Zika Virus, can elicit an immune response that ultimately leads to the development of GBS (2, 3). Other less common, non-infectious antecedent events, such as surgery, pregnancy, immune checkpoint inhibitor therapy, trauma, bone marrow transplantation, and systemic lupus erythematosus, have also been reported (2, 4–8). Several recent case reports have described the development of GBS following intracranial hemorrhage (ICH), though the relationship between these conditions remains poorly understood (9). Despite their distinct pathophysiological mechanisms, both conditions are neurological emergencies, and their coexistence poses a unique diagnostic and therapeutic challenge to healthcare providers. To date, the literature has been limited to individual case reports, resulting in a lack of comprehensive data on the clinical presentation, diagnosis, and management of this rare association. To address this gap, we conducted a systematic review of the literature to analyze and synthesize the existing data on GBS occurring after ICH, with the goal of providing a clear understanding of its clinical features, diagnosis, and management. Methods Study design This systematic review was conducted following the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines (10). Search strategy Two electronic databases (EMBASE and PubMed) were searched from inception until February 2023. The search strategy included a combination of MeSH terms and keywords related to GBS and Intracranial hemorrhage. Study selection Two independent reviewers screened the abstracts of all articles identified by the search strategy (11). Full-text articles were retrieved for all abstracts that met the inclusion criteria or where there was any uncertainty. Two independent reviewers then assessed the full-text articles for inclusion in the systematic review. Disagreements were resolved by consensus or by a third reviewer. We also manually searched the reference lists of included studies and relevant review articles to identify additional studies. The Joanna Briggs Institute (JBI) Critical Appraisal Checklist for Case Reports was used to assess the quality of the included studies (12). This systematic review included case reports of patients with a diagnosis of GBS following intracranial hemorrhage, published in peer-reviewed journals were included in the review. To be eligible, cases had to meet the National Institute of Neurological Disorders and Stroke (NINDS) diagnostic criteria for Guillain–Barré syndrome. For each eligible case report, two authors independently reviewed the published clinical, cerebrospinal fluid, and electrophysiological data to verify that the NINDS criteria were fulfilled. Excluded from the consideration were cases of GBS not related to intracranial hemorrhage, review articles, meeting abstracts as well as studies that are not published in peer-reviewed journals. Studies that are not in English are translated. Data extraction Two independent reviewers extracted data using a standardized form, including study and patient characteristics, clinical characteristics, treatment, and outcomes. Data synthesis Because the available evidence consisted solely of heterogeneous case reports, we did not perform a formal quantitative meta-analysis. Instead, we used a narrative synthesis approach. Data from each case were entered into a standardized extraction sheet and collated into summary tables. We grouped variables a priori into the following domains: (1) study and patient characteristics, (2) type and severity of hemorrhage, (3) clinical features of GBS, (4) investigations, and (5) treatment and outcomes. Within each domain, we summarized data descriptively (frequencies, proportions, means or medians with ranges, as appropriate) and described patterns and differences between subgroups (e.g., traumatic vs. spontaneous hemorrhage, axonal vs. demyelinating variants) in the text. Quality assessment using the JBI Critical Appraisal Checklist for Case Reports was summarized narratively to contextualize the certainty of the evidence but was not used as an exclusion criterion. The JBI checklist comprises eight items that assess the completeness and clarity of reporting of patient demographics, clinical history and presentation, diagnostic work-up, therapeutic interventions, post-intervention course, adverse events, and key take-home messages. Each item is rated as yes, no, unclear, or not applicable; for each case report we counted the number of items rated ‘yes’ (range 0–8) to provide an overall indicator of reporting quality. Results This Systematic review includes 23 case reports (Figure 1), and the data extracted from these studies are presented in Tables 1–3. Flowchart of study selection. Table 1Study characteristics.S. noFirst author & yearCountryAge (Yr)GenderType of hemorrhageLocation of hemorrhageTraumatic/spontaneous1Liu et al. (2022) (40)China62MIntracerebralRight temporal and insular lobesSpontaneous2Abdulmana et al. (2021) (14)Saudi Arabia74FIntracerebralHemorrhagic transformation in right MCA zoneSpontaneous3Abdulmana et al. (2021) (14)Saudi Arabia61MIntracerebralRight frontoparietal (midline shift +)Spontaneous4Yonekura et al. (2018) (18)Japan74MSubdural & SubarachnoidBilateral parietooccipital areasTraumatic5Duncan et al. (1987) (41)UK61MSubduralRight subduralTraumatic6Yilmaz et al. (2020) (42)Turkey41MSubduralLeft temporo-occipitalTraumatic7Connor et al. (2021) (27)Australia64MIntracerebralRight temporo-occipitalSpontaneous8Joshi et al. (2019) (43)Australia70FIntracerebralLeft basal gangliaSpontaneous9Wu et al. (2016) (44)China51FSubarachnoidAmbient and posterior interhemispheric cisternsSpontaneous10Wu et al. (2016) (44)China68MIntracerebralHemorrhagic transformation in left MCA zoneSpontaneous11Carr et al. (2015) (15)USA58MSubarachnoidn/a(mass effect +)Traumatic12Mantero et al. (2013) (16)Italy79FIntracerebralLeft cerebellum(mass effect +)Spontaneous13Riancho et al. (2013) (45)Spain63FSubarachnoidRight frontoparietal regionSpontaneous14Battaglia et al. (2013) (46)France73FIntracerebralBilateral temporal and right frontalTraumatic15Yardimci et al. (2009) (47)Turkey75FSubduralBilateral frontalTraumatic16Tan et al. (2010) (48)Australia44MSubarachnoidn/aSpontaneous17Rivas et al. (2008) (17)USA55MSubduralLeft posterior (no mass effect)Traumatic18Freitas et al. (1997) (49)Brazil29MSubarachnoidLeft cerebral hemisphereTraumatic19Hu et al. (2020) (50)China73FIntraventricularLateral, third, and fourth ventriclesSpontaneous20Jia et al. (2017) (52)China33MIntracerebralRight basal gangliaSpontaneous21Jia et al. (2017) (52)China41MSubduralLeft frontalTraumatic22Song et al. (2012) (51)China52FPontinePontine tegmentum of the brain stemSpontaneous23Hantson et al. (2013) (13)UK81MSubarachnoidn/aSpontaneousYr, year; M, male; F, female; UK, United Kingdom; USA, United States of America; MCA, middle cerebral artery; n/a, not available. Hemorrhage location and radiographic features, including mass effect and midline shift, are reported as described in the original case reports when available. Study characteristics. Yr, year; M, male; F, female; UK, United Kingdom; USA, United States of America; MCA, middle cerebral artery; n/a, not available. Hemorrhage location and radiographic features, including mass effect and midline shift, are reported as described in the original case reports when available. Table 2Clinical characteristics.S. NoTime interval between hemorrhage and GBS onset (days)Pre-GBS deficitsSensory involvementAreflexiaAutonomic involvementRespiratory involvementHughes Score at onset of diseaseDiagnosis of GBS (level of certainty)*18Left-sided hemiparesisYesYesNAYes4Clinical features+ CSF analysis (Level 1)25Left hemiparesis and dysarthriaYesYesYesYes5Clinical features+ CSF analysis+ Electrophysiology (Level 1)37Left-sided weakness and impaired consciousness.YesYesNAYes5Clinical features + CSF analysis + Electrophysiology (Level 1)42NoneUnclear: due to sedationYesNAYes5Clinical features+ CSF analysis (Level 1)59NoneYesYesNAYes5Clinical features+ Electrophysiology (Level 2)614NoneNAYesNAYes5Clinical features+ CSF analysis+ Electrophysiology (Level 1)711DiplopiaYesNANAYes3Clinical features+ CSF analysis+ Electrophysiology (Level 1)814Right-sided hemiplegiaYesYesYesYes5Clinical features+ Electrophysiology (Level 2)910NoneNAYesNAYes5Clinical features+ CSF analysis+ Electrophysiology (Level 1)107Right-sided hemiplegiaYesYesNAYes5Clinical features+ CSF analysis+ Electrophysiology (Level 1)1116NoneYesYesNANo5Clinical features+ CSF analysis+ Electrophysiology (Level 1)1211NoneNAYesNAYes4Clinical features+ CSF analysis+ Electrophysiology (Level 1)133Proximal tetra paresis, areflexiaNAYesYesYes4Clinical features+ CSF analysis+ Electrophysiology (Level 1)147NoneYesYesNANo5Clinical features+ CSF analysis+ Electrophysiology (Level 1)1510NoneNAYesNANA5Clinical features+ CSF analysis+ Electrophysiology (Level 1)167NoneYesYesNAYes5Clinical features+ CSF analysis+ Electrophysiology (Level 1)177NoneNAYesNAYes5Clinical features+ CSF analysis+ Electrophysiology (Level 1)185NoneNAYesYesyes5Clinical features+ CSF analysis+ Electrophysiology (Level 1)193NoneNAYesNAYes4Clinical features+ CSF analysis+ Electrophysiology (Level 1)2014Left hemiparesisNAYesNAYes4Clinical features+ CSF analysis+ Electrophysiology (Level 1)2114NoneNAYesNAYes5Clinical features+ CSF analysis+ Electrophysiology (Level 1)2221Quadriparesis, bilateral facial weakness, severe dysarthriaNAYesYesNA4Clinical features+ CSF analysis+ Electrophysiology (Level 1)2321Comatose at admissionNAYesNAYes3Clinical features+ CSF analysis+ Electrophysiology (Level 1)GBS, Guillain-Barré Syndrome; CSF, cerebrospinal fluid; EP, electrophysiology; NA, Not Available or Not Assessed. *Level of diagnostic certainty according to the Brighton criteria. Clinical characteristics. GBS, Guillain-Barré Syndrome; CSF, cerebrospinal fluid; EP, electrophysiology; NA, Not Available or Not Assessed. *Level of diagnostic certainty according to the Brighton criteria. Table 3Investigations, treatment, and outcomes.S. noCSF cells level (<5 cells/ul)CSF protein level (15–45 mg/dl)Albumino-cytological dissociationTested for C. jejuni or other virusesAntiganglioside antibodiesNeurophysiological studiesVariantImmunomodulatory treatmentClinical Outcome125102NoNoYesNAAMSANIVIG+ methylprednisolonePoor recovery, LAMA21370YesNoYesNCS: Absent motor and sensory responses from the sural, superficial peroneal, median, and ulnar nerves with stimulation of the four limbsAMSANIVIGPoor recovery, full ventilator support on tracheostomy for 2 months, Death due to severe dysautonomia310161.1NoNoYesNCS: demonstrating absent sensory, motor, and F responses from the upper and lower limbs’ nervesAMSANIVIG + PPPoor recovery in 1 month follow up4892NoNoNoNANAIVIGPartial recovery in 8 weeks follow up5NANANANoNANCS: Severe generalized predominantly motor demyelinating peripheral neuropathy entirely consistent with the diagnosis of GBSAIDPPPComplete recovery in 5 weeks follow up6NANAYesYesNoNCS: showed prolonged F-wave latencies and evidence of conduction block in both upper and lower extremity motor nerves.AIDPIVIGComplete recovery in 1 month follow up718148NoYesNoNCS, EMG: no spontaneous activity or activation in responseNAIVIGPartial recovery in 3 months follow up8NANANANoNANCS: Diffuse, predominantly demyelinating polyneuropathy with superimposed proximal axonal degeneration, absent F waves in both upper and lower limbs, and evidence of denervation in the proximal upper limb and chest musclesAMSANIVIGPartial recovery912390.2NoNoYesNCS: Decreased occurrence rate of F-waves, spontaneous potential observed in left gastrocnemius, femoral, and anterior tibialis muscle, decreased left tibial motor nerve conduction velocityAMANIVIGPartial recovery in 1 year follow up108290.2NoNoNANCS: Motor nerve conduction velocity was decreased, debased amplitude, prolonged latency and decreased motor and sensory nerve conduction velocity.AIDPIVIGPoor recovery in 1 month follow up115117YesNoYesNCS: were significant for prolonged F wave latencies and evidence of conduction block in both upper and lower extremity motor nerves.AMANIVIGPartial recovery12NANAYesNoNAEMG: proximal partial conduction blocks in the lower limbs and significant increase of distal latencies and F wave in the lower limbs.AIDPIVIG + PPComplete recovery in 6 months follow up13NA65YesNoNANCS: Demyelinating polyradiculoneuropathy with a frank increase in distal latencies and virtually generalized absence in F waves.NAIVIGPartial recovery14NANAYesYesNoNCS: increase of motor distal latencies and temporal dispersion of motor action potential in four limbs, with F-waves latencies increased in lower limbs, and all sensory responses were abolished.AMSANIVIGPartial recovery in 3-month follow-up15NANAYesYesNoNCS: Severe generalized, predominantly motor-demyelinating peripheral neuropathyAIDPPPPartial recovery in 3 months follow up16NA182YesYesNoNCS: Absent motor and sensory responses, absent blink reflexes, and an absence of spontaneous and voluntary activity on electromyographyAxonal (NOS)IVIGPartial recovery in 10 months follow up17Xanthochromia+, Normal cell countNAYesNoNoNCS: Inexcitability of all nerves and fibrillation potentials of all muscles sampled, severe loss of myelinated axons without significant demyelinationNAPPPartial recovery in 6 months follow up183200YesNoNAEMG: demyelinating patternAIDPDeath before treatment was givenNo recovery, Death198107.3NoNoNoNANAIVIGNo recovery, Death at 1 month follow up204560YesNoNoNCS: CMAP of the right ulnar nerve significantly decreased; motor nerve conduction velocity (NCV) was close to normal.AMANIVIGPartial recovery in a 4-month follow-up212187YesNoNANCS: CMAP of the common peroneal nerve on both sides were significantly decreased; F waves were not evoked in either upper or lower limbs; NCV was close to normal.AMANIVIGPartial recovery in 6 months follow-up22Normal cell count458YesNoNoNCS: markedly reduced compound motor action potential amplitudes of the median (0.1 mV), ulnar (0.2 mV) and peroneal nerves (0.5 mV), whereas distal latencies and motor conduction velocities were normal. The results of sensory conduction studies were completely normal. Fibrillation potentials and positive sharp waves were found in all examined muscles.AMANIVIGPartial recovery in 1 year follow-up.23Normal cell count122YesNoNoNCS: consistent with a severe axonopathy (absent distal compound muscle action potentials) and mild reduction of conduction velocities.Axonal form of GBSNo immunomodulatory treatmentNo recovery, death at a 7 month follow up due to infectious complicationsAIDP, Acute Inflammatory Demyelinating Polyneuropathy; AMAN, Acute Motor Axonal Neuropathy; AMSAN, Acute Motor and Sensory Axonal Neuropathy; C. jejuni, Campylobacter jejuni; CMAP, Compound Muscle Action Potential; CSF, Cerebrospinal Fluid; EMG, Electromyography; IVIG, Intravenous Immunoglobulin; LAMA, Left Against Medical Advice; NA, Not Available or Not Assessed; NCS, Nerve Conduction Studies; NCV, Nerve Conduction Velocity; PP, Plasmapheresis. Investigations, treatment, and outcomes. AIDP, Acute Inflammatory Demyelinating Polyneuropathy; AMAN, Acute Motor Axonal Neuropathy; AMSAN, Acute Motor and Sensory Axonal Neuropathy; C. jejuni, Campylobacter jejuni; CMAP, Compound Muscle Action Potential; CSF, Cerebrospinal Fluid; EMG, Electromyography; IVIG, Intravenous Immunoglobulin; LAMA, Left Against Medical Advice; NA, Not Available or Not Assessed; NCS, Nerve Conduction Studies; NCV, Nerve Conduction Velocity; PP, Plasmapheresis. Demographic characteristics BS Cases (n = 23) were from the China (n = 7), Turkey (n = 2), USA (n = 2), UK (n = 2), Australia (n = 3), Japan (n = 1), France (n = 1), Saudi Arabia (n = 2), Brazil (n = 1), Italy (n = 1) and Spain (n = 1) (Table 1). Of the 23 patients, 14 were male (60.9%) and 9 were female (39.1%), with a male-to-female ratio of 1.56. The average age was 60.1 ± 14.8 years (median: 62 years; range: 29–81 years). There was a statistically significant difference in age at onset between males and females (mean: 55.1 ± 15.4 vs. 67.8 ± 10.2 years, respectively; p = 0.028). Type and severity of hemorrhage Intracerebral hemorrhage was the most common type (39.1%), followed by subarachnoid hemorrhage (26.1%) and subdural hemorrhage (21.7%). Epidural hematoma was not reported, whereas intraventricular hemorrhage and a single case of combined subdural and subarachnoid hemorrhage were relatively rare, each representing 4.3% of cases. There were 10 traumatic and 13 spontaneous hemorrhages. In the traumatic group, subdural hematoma and subarachnoid hemorrhage were most common (4 and 3 cases, respectively), while intracerebral hemorrhage predominated in the spontaneous group (7 cases). Hemorrhage locations were heterogeneous (Table 1), with a clear predominance of supratentorial involvement, including the frontoparietal, temporal, basal ganglia, and middle cerebral artery territories, whereas infratentorial hemorrhages were uncommon, observed in only two cases involving the pontine tegmentum and cerebellar hemisphere. Details regarding the severity of the initial hemorrhage were heterogeneously reported. Standardized severity grading scales were rarely used, with only one case reporting Fisher Grade IV (13). The initial Glasgow Coma Scale (GCS) was available for 6 of the 23 cases. Among these, 3 patients (50%) presented with a mild brain injury (GCS 13–15), 1 (16.7%) with a moderate injury (GCS 9–12), and 2 (33.3%) with a severe injury (GCS < 9). Radiographic features such as midline shift and mass effect were explicitly documented in only four cases, being present in three (14–16) and absent in one (17), while in several other reports, repeat imaging at the time of GBS symptom onset showed stability or regression of the hemorrhage, indirectly suggesting the absence of progressive mass effect. Surgical intervention for the hemorrhage was required in 6 cases (26.1%). Furthermore, 11 patients (47.8%) had pre-existing neurological deficits from the initial hemorrhage prior to the onset of GBS. This inconsistency in radiographic reporting limits the ability to systematically assess the contribution of hemorrhage severity and anatomical burden to the neurological outcomes reported in this cohort. Clinical features of GBS Among the 23 patients, the median time to symptom onset following ICH was 9 days (interquartile range, 7) (Table 2). The minimum interval was 2 days (18), and the maximum was 21 days (13, 51), suggesting a considerable variation in the time between ICH and GBS onset. The most common manifestation of GBS associated with ICH was paresis (23/23), consistently accompanied by hyporeflexia or areflexia (23/23), sensory loss or paresthesia (9/23), peripheral facial nerve palsy (7/23), bulbar palsy (4/23-dysphagia &dysarthria, 2/23-not specified), autonomic dysfunction (6/23), and extraocular involvement including ptosis, diplopia, or ophthalmoplegia (9/23). The median time to nadir from GBS onset was 2.5 days (range 0.5–12 days). Clinical severity was assessed using the Hughes Functional Grading Scale with an average score of 4.57 ± 0.66 (range: 3–6) (19, 20). Investigations Electrophysiological examinations were performed in 21 patients, with detailed reports available for eight (Table 2). Of the 23 cases, 26.1% (6/23) were consistent with acute inflammatory demyelinating polyneuropathy (AIDP), while axonal variants accounted for 52.2% (12/23), including five cases of acute motor and sensory axonal neuropathy (AMSAN), five cases of acute motor axonal neuropathy (AMAN), and two unspecified axonal subtypes. In the remaining five cases, the electrophysiological subtype was not reported. Cerebrospinal fluid (CSF) analysis was conducted in 21 of the 23 cases. Excluding missing data, the average CSF protein level was 222.0 ± 147.7 mg/dL (range: 65–560 mg/dL). Albuminocytologic dissociation (cell count ≤ 5/μL with protein > 45 mg/dL) was reported in 14 of the 22 cases. Mild pleocytosis (cell count > 5/μL) was observed in 7/21 (33.3%) cases, with a maximum cell count of 25/μL. Xanthochromia was reported in one case. Most cases did not have a documented infection based on the initial workup. Among the five cases tested for specific antibodies, none demonstrated evidence of C. jejuni or viral serology. In other cases, an infection was ruled out based on the absence of relevant history, clinical symptoms (fever, diarrhea, respiratory issues) and normal lab results. Out of the fifteen patients who underwent antiganglioside antibody testing, five tested positive including anti-GM1 (3 cases), GD1a (3 cases), and GD1b (4 cases). Additionally, MRI findings suggestive of GBS such as diffuse contrast enhancement, hyperintense signal changes, and patchy enhancement of the spinal roots were observed in three patients. Treatment and prognosis of GBS The clinical course was severe, with 18 of 23 patients (78.3%) requiring mechanical ventilation for respiratory failure (Tables 3, 4). A total of 21 patients received immunomodulatory treatment: 15(71.4%) were treated with intravenous immunoglobulin (IVIG) alone, 3(14.3%) were treated with plasmapheresis alone, 2(9.5%) with a combination of IVIG and plasmapheresis, and 1(4.8%) with IVIG and methylprednisolone. Two patients did not receive any form of immunomodulatory therapy, and both died. Follow-up durations ranged from 5 weeks to 1 year. At final follow-up, three patients (13%) had fully recovered, thirteen (56.5%) had partial recovery, three (13.0%) had poor recovery, and four (17.4%) had died. Reported causes of death included severe dysautonomia, cardiorespiratory arrest, and infectious complications. Three of the four patients who died had documented sepsis and one met SIRS criteria without a confirmed pathogen, whereas none of the three patients with poor recovery had documented sepsis or SIRS. Table 4Treatment and prognosis of GBS.Immunomodulatory treatmentComplete recovery (n)Partial recovery (n)Poor recovery (n)Death (n)IVIG (n = 15)11112PP (n = 3)1200IVIG + PP (n = 2)1010IVIG + methylprednisolone (n = 1)0010No Immunomodulatory treatment (n = 2)0002GBS – Guillain-Barré Syndrome; IVIG – Intravenous Immunoglobulin; PP – Plasmapheresis. Definitions: 1. Complete recovery: The patient has returned to normal function and has no residual deficits. 2. Partial recovery: The patient has recovered some function but still has some residual deficits. 3. Poor recovery: The patient has not recovered much function and has significant residual deficits. This can include being bedridden, requiring assistance with activities of daily living, or being unable to return to work. No recovery: The patient has not recovered any function and remains completely dependent on others for care. Treatment and prognosis of GBS. GBS – Guillain-Barré Syndrome; IVIG – Intravenous Immunoglobulin; PP – Plasmapheresis. Definitions: 1. Complete recovery: The patient has returned to normal function and has no residual deficits. 2. Partial recovery: The patient has recovered some function but still has some residual deficits. 3. Poor recovery: The patient has not recovered much function and has significant residual deficits. This can include being bedridden, requiring assistance with activities of daily living, or being unable to return to work. No recovery: The patient has not recovered any function and remains completely dependent on others for care. Ganglioside antibody–positive patients (n = 5) had markedly worse outcomes than antibody-negative patients (n = 10). None of the antibody-positive patients achieved full recovery (2 had poor recovery, 2 partial recovery, and 1 had no recovery and died), whereas 80% of the antibody-negative group showed favorable outcomes, with 1 achieving full recovery and 7 achieving partial recovery (2 died). A subgroup analysis based on electrophysiological variant (Supplementary Table 1) showed that axonal GBS (AMAN/AMSAN/Axonal NOS, n = 12) was associated with a higher need for mechanical ventilation (10/12) compared to demyelinating GBS (n = 6; 4/6). Outcomes were less favorable in the axonal group, with only one complete recovery, seven partial recoveries, one poor recoveries, and two deaths. In contrast, demyelinating GBS demonstrated relatively better outcomes, with three complete recoveries, two partial recoveries, one poor recovery, and one death. Within axonal subtypes, AMSAN (n = 5) had the poorest prognosis, with no complete recoveries and three cases of poor recovery, whereas AMAN (n = 5) showed more favorable outcomes, with one complete recovery and the remainder achieving partial recovery without deaths. Quality assessment To evaluate the methodological quality of studies reporting GBS following ICH, we conducted a JBI quality assessment of case reports (Table 5). Of the 23 case reports identified, 21 were determined to be of high-quality reporting, scoring 7 or 8. However, two studies scored 5, indicating minor issues with reporting. Overall, the assessment findings suggest that the case reports included in our systematic review are of high methodological quality, which enhances the credibility and dependability of our results. Table 5Quality assessment of included case reports using Joanna Briggs Institute (JBI) critical appraisal checklist for case reports.S. noReference (author, year)A1A2A3A4A5A6A7A8Total1Liu et al. (2022) (40)YesYesYesYesYesYesNAYes72Abdulmana et al. (2021) (14)YesYesYesYesYesYesNAYes73Abdulmana et al. (2021) (14)YesYesYesYesYesYesNAYes74Yonekura et al. (2018) (18)YesYesYesYesYesYesNAYes75Duncan et al. (1987) (41)YesYesYesNoNoYesNAYes56Yilmaz et al. (2020) (42)YesYesYesYesYesYesNAYes77Connor et al. (2021) (27)YesYesYesYesYesYesYesYes88Joshi et al. (2019) (43)YesYesNoYesYesNoNAYes59Wu et al. (2016) (44)YesYesYesYesYesYesNAYes710Wu et al. (2016) (44)YesYesYesYesYesYesNAYes711Carr et al. (2015) (15)YesYesYesYesYesYesNAYes712Mantero et al. (2013) (16)YesYesYesYesYesYesNAYes713Riancho et al. (2013) (45)YesYesYesYesYesYesNAYes714Battaglia et al. (2013) (46)YesYesYesYesYesYesNAYes715Yardimci et al. (2009) (47)YesYesYesYesYesYesNAYes716Tan et al. (2010) (48)YesYesYesYesYesYesNAYes717Rivas et al. (2008) (17)YesYesYesYesYesYesNAYes718Freitas et al. (1997) (49)YesYesYesYesYesYesNAYes719Hu et al. (2020) (50)YesYesYesYesYesYesNAYes720Song et al. (2012) (51)YesYesYesYesYesYesNAYes721Hantson et al. (2013) (13)YesYesYesYesYesYesNAYes722Jia et al. (2017) (52)YesYesYesYesYesYesNAYes723Jia et al. (2017) (52)YesYesYesYesYesYesNAYes7Answer Options include—Yes, No, Unclear, Not Applicable (NA). A1—Were patient’s demographics characteristics clearly described? A2—Was the patient’s history clearly described and presented as a timeline? A3—Was the current clinical condition of the patient on presentation clearly described? A4—Were diagnostic tests or assessment methods and the results clearly described? A5—Was the intervention(s) or treatment procedure(s) clearly described? A6—Was the post-intervention clinical condition clearly described? A7—Were adverse events (harms) or unanticipated events identified and described? A8—Does the case report provide takeaway lessons? Quality assessment of included case reports using Joanna Briggs Institute (JBI) critical appraisal checklist for case reports. Answer Options include—Yes, No, Unclear, Not Applicable (NA). A1—Were patient’s demographics characteristics clearly described? A2—Was the patient’s history clearly described and presented as a timeline? A3—Was the current clinical condition of the patient on presentation clearly described? A4—Were diagnostic tests or assessment methods and the results clearly described? A5—Was the intervention(s) or treatment procedure(s) clearly described? A6—Was the post-intervention clinical condition clearly described? A7—Were adverse events (harms) or unanticipated events identified and described? A8—Does the case report provide takeaway lessons? Discussion This systematic review of 23 case reports describes a temporal relationship between antecedent ICH and the subsequent onset of GBS. While the evidence is limited to case-level data and is insufficient to prove causation, this association has been documented following all major types of ICH. The clinical course described in these reports was typically severe, characterized by the rapid onset of symmetrical weakness and a swift progression to maximum deficit. Symptom onset varied from 2 to 24 days post-ICH, a timeframe consistent with the typical presentation of GBS (21). In all cases, the diagnosis was subsequently confirmed through a combination of clinical, cerebrospinal fluid, and electrophysiological findings. Although based on a limited number of cases, these observations underscore the importance of considering GBS in the differential diagnosis when patients with ICH exhibit unexplained, progressive neurological decline—particularly in the absence of new radiographic findings. A striking finding of our review is the marked severity of GBS in the context of ICH. Approximately three-quarters of patients required mechanical ventilation, compared to the 25–35% typically reported in general GBS populations (22). This severity was further underscored by a rapid progression to nadir and a predominance of axonal GBS subtypes, which are known to be associated with more severe disease and poorer outcomes (23, 24). Our subgroup analysis further supports this, demonstrating that axonal variants were associated with higher rates of mechanical ventilation and worse functional outcomes compared to demyelinating forms, with AMSAN showing the poorest prognosis. This pattern is consistent with prior literature showing that axonal variants, particularly AMSAN, are characterized by rapid progression, increased risk of respiratory failure, and prolonged recovery with greater residual disability (23, 25, 26). The combination of this electrophysiological pattern and the elevated mortality rate, suggests that GBS following ICH may represent a particularly aggressive form of the disease. This likely reflects a multifactorial process involving the severity of the initial brain injury, the superimposed burden of GBS, and the challenges of managing two concurrent, life-threatening neurological conditions. Most patients received standard immunomodulatory therapy, predominantly intravenous immunoglobulin (IVIG). However, prognosis remained highly variable: around one-fifth of patients died and many survivors experiencing only partial recovery. Interpreting disease severity and treatment response in this population is challenging due to the confounding impact of the initial hemorrhage. Nearly half of the patients had substantial pre-existing neurological deficits from the hemorrhage, and over a quarter required surgical intervention. These factors likely contributed significantly to poor outcomes, independent of the GBS or its treatment. Furthermore, systemic infection was a uniform finding in all fatal cases; three of the four deaths had documented sepsis, while the fourth met SIRS criteria. In contrast, none of the three patients with poor recovery had documented sepsis or SIRS, implying that outcomes in this subgroup were more likely driven by the severity of the initial hemorrhage, the GBS disease burden itself, or both. While our review describes temporal association, the biological mechanism connecting ICH to GBS remains a subject of speculation. The prevailing theory, analogous to post-infectious GBS, involves molecular mimicry initiated by a breach in the blood–brain barrier (BBB) (24). Following an ICH, the physical disruption of the BBB and the direct cellular damage caused by iron-rich heme released from the lysing hematoma may allow previously sequestered neuronal antigens such as gangliosides and myelin basic protein to enter the systemic circulation (27). This exposure could trigger an autoimmune response that cross-reacts with structurally similar components of peripheral nerves. The presence of anti-ganglioside antibodies (anti-GM1, GD1a, and GD1b) in some patients supports this hypothesis. However, this explanation is not entirely sufficient, as many patients in the reviewed cases tested negative for these specific antibodies. While this does not disprove the molecular mimicry theory, it suggests that alternative or complementary mechanisms may be at play. An alternative or complementary hypothesis centers on the profound systemic inflammatory response triggered by the initial ICH. A severe brain hemorrhage is a potent inflammatory event, known to induce a “cytokine storm” that releases pro-inflammatory mediators like TNF-α and various interleukins into the circulation (28). This inflammatory cascade is likely initiated at the site of the bleed, where the breakdown of hemoglobin releases cytotoxic heme, a powerful pro-oxidant and pro-inflammatory molecule. This systemic inflammation preferentially targets the peripheral nervous system, as its blood-nerve barrier (BNB) is more susceptible to inflammatory breakdown than the resilient blood–brain barrier (BBB) (29). A compromised BNB allows for a direct T-cell mediated attack or inflammatory damage to the peripheral nerves (29), providing a plausible mechanism for GBS development, particularly in the antibody-negative cases observed in our review. It is therefore conceivable that these mechanisms are not mutually exclusive; the ICH could provide the antigenic trigger for molecular mimicry, while the resulting systemic inflammation breaks down the protective BNB, creating a “perfect storm” for autoimmune peripheral nerve injury. It is crucial to note that in most cases, a preceding infection was not identified, strengthening the hypothesis that the ICH itself served as the primary immunological trigger. Critical illness polyneuropathy (CIP) is an important differential to consider in this context (Table 6) (30). Although CIP is classically described as subacute in onset following prolonged ICU admission, early-onset cases have been documented in patients with systemic inflammatory response syndrome or sepsis, sometimes within the first days of critical illness (31, 32), meaning that temporal profile alone cannot reliably exclude CIP in this setting. From an electrodiagnostic standpoint, AIDP can be distinguished from CIP by demyelinating features such as prolonged compound muscle action potential (CMAP) latency and slowed conduction velocity, while AMAN is characterized by pure motor involvement. Differentiating CIP from AMSAN, however, is more challenging, as both conditions demonstrate reduced CMAP and sensory nerve action potential (SNAP) amplitudes. In such cases, clinical features and ancillary testing become essential. AMSAN more frequently involves cranial nerves, particularly the facial nerve, and is often associated with dysautonomia, findings not typical of CIP. It also commonly demonstrates albuminocytologic dissociation in CSF and may show nerve root enhancement on MRI, whereas CIP does not. Intravenous immunoglobulins (IVIG) and plasma exchange can be effective in AMSAN (25, 33), but not in CIP, aside from limited reports in severe COVID-19 (34). Table 6Differentiation between critical illness polyneuropathy and Guillain–Barré syndrome.Critical illness polyneuropathyGuillain–Barré syndromeAIDPAMAN and AMSANProdromal conditionsSepsis and multiple organ failureGastrointestinal or respiratory infectionClinical presentationOnset of the disorder usually after intensive care unit admissionOften characterized by fairly symmetric limb muscle weakness sparing cranial nerves; sensory deficits less prominentOnset of the disorder is usually before intensive care unit admission.Infections precede the onset of progressive weakness and sensory disturbances; frequent cranial nerve involvement and dysautonomia.Sensory symptoms are present in AIDP and AMSAN. Axonal variants can have a more rapid progression leading to paralysis and respiratory failure over a few days.Cerebrospinal fluidUsually normalAlbuminocytologic dissociationElectrophysiologyAxonal motor & sensory polyneuropathyDemyelinating polyneuropathyAxonal motor polyneuropathy in AMAN, axonal motor & sensory polyneuropathy in AMSAN. It can have unresponsive nerves, abundant spontaneous activity.Magnetic resonance imagingNo significant findingsOccasional enhancement of spinal nerve rootsBiopsyPrimarily axonal degeneration of distal peripheral nerves without inflammationPrimarily demyelinating process with inflammationMotor/sensory axonal degeneration, or motor axonal degeneration only.TreatmentNo specific therapy, usually anti-septic treatmentPlasmapheresis, intravenous immune globulinOutcomeRecovery may be spontaneous and of variable timing; 50% of patients with full recoveryUsually > 60% complete recoveryAMAN: Some recover rapidly, others slowly depending on axonal damage.AMSAN: prolonged course with greater residual disability.*AIDP: Acute inflammatory demyelinating polyneuropathy, AMAN: Acute motor axonal neuropathy, AMSAN: Acute motor-sensory axonal neuropathy. Information compiled from Zhou et al. (30). Differentiation between critical illness polyneuropathy and Guillain–Barré syndrome. *AIDP: Acute inflammatory demyelinating polyneuropathy, AMAN: Acute motor axonal neuropathy, AMSAN: Acute motor-sensory axonal neuropathy. Information compiled from Zhou et al. (30). Taken together, several findings argue against CIP in the case reports of this systematic review. First, the clinical picture was typical of GBS with rapidly progressive symmetric ascending weakness and generalized areflexia, often with cranial or autonomic involvement, which is not characteristic of CIP (35). Second, the diagnostic studies supported GBS. CSF consistently showed albuminocytologic dissociation, and nerve conduction studies predominantly showed patterns consistent with AIDP. Third, the presence of antiganglioside antibodies in a subset of cases supports an immune mediated process instead of the metabolic and microvascular derangements of critical illness (35). We acknowledge that intracranial hemorrhage particularly subarachnoid and intracerebral subtypes along with neurosurgical procedures and traumatic lumbar punctures can introduce blood into the CSF (36, 37). Such contamination may spuriously raise protein, and cell counts. However, a substantial RBC burden is required to produce a clinically meaningful false elevation (approximately 1 mg/dL of protein and about 1 WBC per 500–1,000 RBCs) (38). Among the included cases, only one case reported xanthochromia, while CSF analysis was precluded in two others due to procedural deferral or grossly bloody fluid. Although the absence of precise RBC counts in multiple cases prevents the definitive exclusion of blood-related distortion, the consistent observation of albuminocytologic dissociation without significant pleocytosis serves as supportive evidence for GBS in this context. The quality assessment of the reports included revealed that most were of high methodological quality, which strengthens the reliability of our findings. However, our study has certain limitations that must be considered. First, our study only included case reports, which are subject to publication bias, potentially favoring the reporting of more severe or unusual cases, while milder or self-limited cases may remain unpublished. This may lead to an overestimation of true disease severity and mortality. Second, the lack of a control group precludes the establishment of causality or the calculation of true incidence, limiting our findings to a temporal association. Third, the largest proportion of cases in our review originated from China, where axonal GBS variants are more frequently reported than in many western populations (39). This geographic clustering may have contributed to the high proportion of axonal phenotypes observed and limits the generalizability of our electrophysiological findings to regions in which AIDP predominates. Fourth, the retrospective nature of the data resulted in heterogeneous and inconsistent reporting of key clinical and radiographic details, including the specifics of electrophysiological studies, the use of standardized electrodiagnostic criteria (e.g., EAN/PNS criteria), the inconsistent performance of antiganglioside antibodies testing, and the thoroughness of the workup to exclude antecedent infections. Furthermore, variations in the reporting of hemorrhage severity, anatomical characteristics, standardized grading scales, and radiographic features such as mass effect or midline shift may have independently influenced neurological recovery and mortality, confounding the interpretation of GBS related outcomes. Despite these limitations, our study offers several valuable insights. From a clinical standpoint, this review highlights the significant diagnostic challenge of “diagnostic overshadowing.” In a patient recovering from a severe brain injury, new-onset weakness can easily be misattributed to the primary injury, sedation, or critical illness myopathy/neuropathy. Recognizing the symmetrical, ascending pattern of weakness, associated hyporeflexia or areflexia, and confirming the diagnosis through CSF analysis (albuminocytologic dissociation) is therefore critical. Although both patients who did not receive immunotherapy died, this cannot be attributed to treatment alone given the concurrent confounders of hemorrhage severity, location, and systemic complications such as sepsis. Nonetheless, prompt initiation of IVIG or plasmapheresis remains reasonable once GBS is suspected, as these represent standard of care and diagnostic delay carries independent risk. Conclusion GBS is a rare but clinically significant complication that can occur following an ICH. Our analysis suggests that post-ICH GBS represents a particularly aggressive clinical phenotype, characterized by rapid progression to nadir and a high rate of respiratory failure. These findings underscore the critical need for clinicians to maintain a high index of suspicion for GBS in patients with ICH who develop progressive, symmetrical weakness unexplained by the initial brain injury. While prompt diagnosis and initiation of standard immunotherapy is appropriate once GBS is suspected, outcomes in this population are likely multifactorial, and the independent contribution of treatment cannot be established from case-level data alone. Future research should focus on prospective, multi-center registry-based studies to determine the true incidence of this association and to conduct more systematic immunological analyses to elucidate the precise pathophysiological link between CNS hemorrhage and peripheral nervous system autoimmunity.