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


Brain. 2018;141(10):2866-2877. 

In This Article


Our study demonstrates the marked worldwide variation of GBS with respect to clinical variants, severity, electrophysiological subtypes, and outcome. This variation is influenced by regional differences in demography, preceding events, and treatment.

In all three regions, the frequency of GBS increased with age, for both males and females. Similar age distributions for GBS have been found previously (McGrogan et al., 2009; Sejvar et al., 2011a). Patients from Bangladesh were younger than patients from the other two regions, which corresponds to results from a previous study in Bangladesh, where the median age was 21 years (range 2–65) (Islam et al., 2010). The regional differences in age distribution may be explained by the variation in demography of the general populations and merely reflect the relative number of persons at risk in each age category per region (UN, Males were more frequently affected than females in a ratio of 1.5:1, in all age categories and regions. Similar male to female ratios have been reported previously (Hughes and Cornblath, 2005; van den Berg et al., 2014). Therefore, male gender and higher age are independent risk factors for developing GBS worldwide.

The full clinical spectrum of GBS was observed in patients from all countries participating in IGOS, but the frequency of variants differed considerably between regions. The predominant variant in Europe/Americas was sensorimotor, whereas in Bangladesh pure motor GBS predominated. The proportion of patients with MFS or MFS-GBS overlap syndrome was higher in Asia than in the other two regions. A similar distribution of clinical variants per region has been suggested in previous reports from single countries. In these studies, the frequency of pure motor GBS ranged from 10–18% in Europe (Visser et al., 1995) to as high as 92% in Bangladesh (Islam et al., 2010). The frequency of MFS varied from 3% in Europe (Lo, 2007) to 34% in Eastern Asia (Mori et al., 2001; Mitsui et al., 2015). The clinical presentation of the patients in the IGOS cohort was similar to previous studies from single countries in Europe/Americas (Fokke et al., 2014), Asia (Matsui et al., 2018) and Bangladesh (Islam et al., 2016; Ishaque et al., 2017).

Almost all patients reached nadir within 4 weeks after study entry (99.8%), and 96% of patients even within 2 weeks. In another study, 3% of the patients reached nadir between 4 and 6 weeks (Fokke et al., 2014). While a progressive phase of more than 4 weeks could be regarded as an exception, subacute inflammatory demyelinating polyradiculoneuropathy should be considered in these patients, a previously described intermediate form between GBS and chronic inflammatory demyelinating polyradiculoneuropathy (Hughes et al., 1992). At the other end of the GBS spectrum, patients reached clinical nadir within days. Some patients already had inexcitable nerves at first NCS. The mechanism of nerve inexcitability is unknown but may be mediated by early loss of axonal or myelin structural integrity or by functional block at the nodes of Ranvier or nerve terminals, caused by anti-nerve antibodies, ionic imbalance, or other inflammatory mediators.

Demyelinating and axonal subtypes of GBS were seen in all participating countries but the frequencies varied between regions. The demyelinating subtype was the predominant subtype in all regions. However, in Bangladesh a substantial proportion of patients had axonal neuropathy. These findings are in line with results from previous studies, where demyelinating GBS was found in 60–80% of North American and European patients (Hadden et al., 1998; van den Berg et al., 2014). Axonal GBS was reported in 3–17% in Europe (Hadden et al., 1998; Sekiguchi et al., 2012; Kuwabara and Yuki, 2013), in 23–65% in Asia (Kuwabara and Yuki, 2013; Mitsui et al., 2015), and up to 67% in Bangladesh (Islam et al., 2010). Interestingly, in all three regions patients with axonal GBS were younger than patients with demyelinating GBS. The influence of electrophysiological subtype on prognosis is under debate, as recovery in axonal GBS can be slow and incomplete due to axonal degeneration, or faster due to resolving transient conduction blocks, and may depend upon the subtype criteria (Kuwabara and Yuki, 2013; van den Berg et al., 2014). The current study showed that the axonal subtype was significantly associated with poor recovery in the full cohort and a similar trend was observed in the subgroup analysis per region (Supplementary Table 2). The association between axonal GBS and younger age may reduce the effect of axonal involvement on poor recovery. Further analysis of NCS and other prognostic factors is required to determine the association between GBS subtype and outcome.

The regional differences in frequencies of clinical and electrophysiological subforms of GBS may be explained in part by the variation in local exposure to infections. The frequency of patient-reported gastroenteritis in our cohort ranged from 25% in Europe/Americas to 36% in Bangladesh. Previous studies have shown an association between preceding gastroenteritis and pure motor and axonal GBS (Islam et al., 2010; Kuwabara and Yuki, 2013). Campylobacter jejuni is the predominant cause of gastroenteritis preceding GBS worldwide, but previous reports suggest that the frequency of this infection may differ substantially among regions. The association between preceding C. jejuni infection and axonal GBS is related to the induction of cross-reactive antibodies to gangliosides (Willison et al., 2016). A recent retrospective study indicated a relatively high frequency of the demyelinating subtype (49%) and lower frequency of the axonal subtype (19%) in Southern China (Liu et al., 2018), while previous studies from Northern China from the 1990s reported the axonal subtype in 65% of GBS patients (Ho et al., 1995). It is unknown whether this variation represents a regional difference within China or a change in GBS spectrum over time in parallel to changes in exposure to infections, especially with C. jejuni (Baker et al., 2012; Liu et al., 2018). Future serological studies will investigate the role of preceding infections, and immune responses to these infections, to explain the regional differences.

The clinical course and outcome varied substantially among the three regions. The best outcome was observed in Asia, in part related to the higher frequency of MFS in that region (Mori et al., 2001; Mitsui et al., 2015). The worst outcome was found in Bangladesh, despite the younger age of these patients. Several factors previously associated with poor prognosis were more frequent in Bangladesh, such as the frequency of preceding gastroenteritis, axonal subtype, and more severe disease in the acute stage. Most importantly, only 13% of the patients in Bangladesh received plasma exchange or IVIg and the facilities for supportive care were limited.

Although this study is the largest prospective study on GBS so far, there are several limitations. First, IGOS aimed to include the full spectrum of GBS, irrespective of age, disease severity, and treatment, but referral bias probably favoured inclusion of patients with more severe disease that required hospitalization and treatment. Participating centres were mostly tertiary care hospitals with specific neuromuscular expertise. It is unknown whether referral bias differed among countries and if this might have influenced the observed regional differences. Second, the number of inclusions varied per country and several areas, especially Asia, Africa, and Australia, were under-represented. The centre in Dhaka, Bangladesh, in contrast, is the national and public tertiary care hospital for GBS, which explains the high number of inclusions and the high proportion of patients receiving supportive care only (Islam et al., 2010, 2016; Ishaque et al., 2017). Third, although IGOS included 1000 patients, the numbers in some subgroups were small and their analyses had limited power. Enrolment of patients in IGOS is continuing to overcome this problem. Lastly, patients were classified according to only one set of electrophysiological criteria using a single NCS, while the assigned GBS subtype depends on the criteria used and may change during follow-up. The electrophysiology of GBS and performance of different sets for classification will be evaluated in future dedicated studies.

The standardized collection of data in IGOS has enabled us to identify differences in the preceding factors, clinical presentation, neurophysiological classification and course of GBS between regions. In combination with the biosamples collected at the same time, this information will improve understanding of pathogenesis—involving identification of risk factors for GBS, including preceding infections of which some may be preventable—and allow better prognostic modelling, adapted to different parts of the world.