Regional Variation of Guillain-Barré Syndrome

Alex Y. Doets; Christine Verboon; Bianca van den Berg; Thomas Harbo; David R. Cornblath; Hugh J. Willison; Zhahirul Islam; Shahram Attarian; Fabio A. Barroso; Kathleen Bateman; Luana Benedetti; Peter van den Bergh; Carlos Casasnovas; Guido Cavaletti; Govindsinh Chavada; Kristl G. Claeys; Efthimios Dardiotis; Amy Davidson; Pieter A. van Doorn; Tom E. Feasby; Giuliana Galassi; Kenneth C. Gorson; Hans-Peter Hartung; Sung-Tsang Hsieh; Richard A.C. Hughes; Isabel Illa; Badrul Islam; Susumu Kusunoki; Satoshi Kuwabara; Helmar C. Lehmann; James A.L. Miller; Quazi Deen Mohammad; Soledad Monges; Eduardo Nobile Orazio; Julio Pardo; Yann Pereon; Simon Rinaldi; Luis Querol; Stephen W. Reddel; Ricardo C. Reisin; Nortina Shahrizaila; Soren H. Sindrup; Waheed Waqar; Bart C. Jacobs; the IGOS Consortium

Disclosures

Brain. 2018;141(10):2866-2877. 

In This Article

Results

We excluded 62 (6%) patients from analysis because of alternative diagnosis: acute onset chronic inflammatory demyelinating polyneuropathy (n = 37), other peripheral neuropathy (n = 8), CNS disorder (n = 12), functional disorder (n = 2), or disorder not specified (n = 3). We excluded five patients because of protocol violations, and eight patients because of insufficient data. The remaining cohort of 925 patients originated from Argentina (n = 43), Australia (n = 4), Bangladesh (n = 125), Belgium (n = 16), Canada (n = 25), Denmark (n = 76), France (n = 27), Germany (n = 45), Greece (n = 4), Italy (n = 82), Japan (n = 36), Malaysia (n = 28), The Netherlands (n = 67), South Africa (n = 11), Spain (n = 76), Taiwan (n = 5), UK (n = 129), and USA (n = 126). At 1 year, 143 (16%) patients were lost to follow-up.

Cohort Description and Heterogeneity of GBS

GBS occurred in all age categories with an overall median age of 51 years (IQR 33–64, range 6 months to 88 years) (Figure 1). The number of patients increased with age and reached its peak at the age categories of 50–59 and 60–69 years. Males predominated in all age categories with an overall male to female ratio of 1.5.

Figure 1.

Age and gender distribution of IGOS cohort. *P < 0.05 for difference in number of males and females per age category. n = 919.

An antecedent event in the 4 weeks before neurological onset was reported in 649 (76%) patients, mainly upper respiratory tract infections (35%) and gastroenteritis (27%). At study entry, 677 (73%) patients had tetraparesis, 105 (11%) had paraparesis, and 19 (2%) had upper limb weakness only. During follow-up, 22 (21%) patients who presented with paraparesis and three (16%) patients who presented with sole weakness of upper limbs also developed tetraparesis. Only five patients had asymmetrical limb weakness.

The median time from onset of symptoms to study entry was 6 days (IQR 3–9). Nadir was reached within 2 weeks in 824 (96%) patients, and within 4 weeks in 858 (99.8%) patients. One patient continued to deteriorate until Week 8 and another until Week 13. At nadir, the median MRC sum score was 44 (IQR 25–53), which was 2 points lower than at entry (46, IQR 33–54) (Wilcoxon signed ranks test P < 0.001).

The clinical course defined by the GBS disability score was highly variable (Figure 2). For those unable to walk independently at nadir, 439 (77%) regained the ability to walk independently at 6 months, and 445 (81%) at 12 months. Overall, 19% required mechanical ventilation during the disease course. Seven per cent died during follow-up, and the median time from onset of weakness to death was 33 days (IQR 16–88, range 6–280) (Table 1).

Figure 2.

Clinical course during 1-year follow-up.

CSF was examined in 823 (89%) patients within a median time of 4 days (IQR 2–8) from onset of neurological symptoms. Elevated CSF protein level was detected in 561 (68%) of these patients. The CSF protein level was strongly influenced by the timing of the lumbar puncture: only 50% had an elevated CSF protein level when tested within 3 days from onset of neurological symptoms, compared to 84% when tested after 7 days. Median CSF protein level in the early group was 0.45 g/l (IQR 0.33–0.73), and in the late group 0.98 g/l (IQR 0.59–1.84) (P < 0.001). Most patients had a normal CSF leucocyte count (<5 cells/μl) (n = 641, 80%). A mildly elevated cell count (5–50/μl) was found in 149 (19%) patients, but 14 (2%) patients had more than 50 leucocytes/μl (range 53–232). No alternative diagnosis was found during follow-up in these patients with CSF pleiocytosis (>50 μl) despite extensive diagnostic work-up. Six (43%) of these patients required mechanical ventilation, compared to 148 of 790 (19%) patients without pleiocytosis (P = 0.035), but the clinical course and outcome were similar between the two groups. Cytoalbuminological dissociation was present in 538 (67%) patients.

A nerve conduction study was performed in 829/862 (96%) patients, median 7 days (IQR 4–11) from onset of weakness. In 84 (10%) of these patients, the NCS could not be evaluated because of missing raw data or missing local reference values. NCS of the remaining 745 patients were classified as demyelinating (n = 390, 52%), axonal (n = 71, 10%), inexcitable (n = 20, 3%), equivocal (n = 215, 29%), or normal (n = 49, 7%). Compared to the demyelinating group, patients with axonal GBS were younger (31 years, IQR 20–56 versus 54 years, IQR 36–67; P < 0.001) and more often reported preceding diarrhoea (24/71, 34% versus 85/390, 22%; P = 0.03). Furthermore, patients with axonal GBS had more severe limb weakness at both study entry (MRC sum score 33, IQR 14–44 versus 46, IQR 34–54; P < 0.001) and nadir (19, IQR 5–41 versus 42, IQR 24–51; P < 0.001). At 6 months, 31/50 (62%) patients with axonal neuropathy were able to walk independently, versus 216/262 (82%) in the demyelinating group (P = 0.001). At 12 months, 34/47 (72%) with axonal GBS and 220/252 (87%) with demyelinating GBS were able to walk independently (P = 0.01).

Geographical Variation of GBS

The demography, antecedent events, clinical presentation, electrophysiological subtypes, diagnostic findings, treatment and outcome of GBS were compared between 'Europe/Americas' (n = 715), 'Asia' (n = 69), and 'Bangladesh' (n = 125) (Table 2, Table 3, Figs 3A, B, 4 and Supplementary Table 1).

Figure 3.

Clinical variants (Week 2) (A) and antecedent events (B) in different geographical areas. (A) MFS: Miller Fisher and Miller Fisher GBS overlap syndromes. Other: pharyngeal-cervical-brachial, pure sensory, ataxic and other clinical variants. (B) Other: urinary tract infection, vaccination, surgery and other antecedent events. URTI = upper respiratory tract infection.

Figure 4.

Kaplan-Meier analysis of time to walk unaided in different geographical areas. Kaplan-Meier analysis for patients unable to walk unaided (GBS disability score > 2) at disease nadir.

Patients from Bangladesh were significantly younger (age 28 years, IQR 16–40) than patients from Europe/Americas (55 years, IQR 37–67, P < 0.001) and Asia (50 years, IQR 34–60, P < 0.001). An upper respiratory tract infection was the most common reported antecedent event in Europe/Americas (38%) and Asia (51%), whereas in Bangladesh, gastroenteritis was predominant (36%). Patients from Bangladesh had more severe muscle weakness than patients from the other two regions at study entry and nadir. Sensory deficits were more frequent in patients from Europe/Americas than in patients from the other two regions. Cranial nerve involvement was more frequent in patients from Asia and Bangladesh than in patients from Europe/Americas. In Asia, more patients had oculomotor weakness, whereas in Bangladesh the proportion of patients with bulbar weakness was significantly higher than in the other regions.

Patients from Asia reported pain less frequently than patients from Europe/Americas and Bangladesh. Seventy-seven (62%) of 125 patients from Bangladesh reported pain at study entry, of whom 73 (95%) had either muscle or joint pain, also including patients with a pure motor variant. Patients from Europe/Americas were less frequently ventilated (17%) than patients from Asia (25%, P = 0.13) and Bangladesh (29%, P = 0.003).

The predominant clinical pattern of GBS in Europe/Americas and Asia was sensorimotor (Europe/Americas: n = 387, 69%; Asia n = 27, 43%), whereas in Bangladesh most patients had pure motor GBS (n = 74, 69%). MFS or MFS-GBS overlap occurred more frequently in Asia (n = 14, 22%) than in Europe/Americas (n = 57, 11%) and Bangladesh (n = 1, 1%) (P < 0.001).

Considerable variation was observed in treatment of GBS between regions. IVIg was the most common treatment for patients from Europe/Americas (n = 612, 86%) and Asia (n = 50, 73%), whereas in Bangladesh the majority of patients (n = 108, 86%) received no immunomodulating therapy.

The median time to regain the ability to walk independently was 63 days (IQR 28–186) in Europe/Americas, 39 days (IQR 17–94) in Asia, and 95 days (IQR 36–190) in Bangladesh (P = 0.002). The proportion of patients who regained the ability to walk independently after 12 months follow-up was 69% in Bangladesh, 83% in Europe/Americas, and 91% in Asia (P = 0.003; Table 2, Table 3 and Figure 4). Mortality was significantly higher in Bangladesh (n = 19, 17%) than in Europe/Americas (n = 23, 5%, P < 0.001) and Asia (n = 1, 2%, P = 0.02).

The predominant electrophysiological subtype was demyelinating for all regions (Europe/Americas: n = 312, 55%; Asia: n = 29, 45%; Bangladesh: n = 38, 40%). The axonal subtype occurred more often in Bangladesh (n = 34, 36%). Clinical differences among electrophysiological subtypes were compared for each region (Supplementary Table 2). In all three regions, patients with the axonal subtype were younger than patients with the demyelinating subtype. Sensory deficits at entry and nadir were less frequent in patients with axonal neuropathy. There was a trend towards a lower MRC sum score at study entry and nadir (only significant for Europe/Americas), and poorer outcome at 6 and 12 months in the axonal groups compared to the demyelinating groups (Supplementary Table 2).

Comments

3090D553-9492-4563-8681-AD288FA52ACE

processing....