The progressive phase of the disease is less than 4 weeks long and usually starts with ascending distal paresthesia of the lower extremities. Symmetric motor weakness is always present in the plateau phase, and deep reflexes are lost in 90 % of the cases. Involvement of the cranial nerves, especially bilateral peripheral facial nerve palsy, can be characteristic of AIDP. Autonomic symptoms and respiratory failure are the major concerns of life-threatening complications. Electroneurography (ENG) indicates distal and proximal multifocal demyelination: the motor nerve conduction is decreased, the distal motor latency and F wave latency are long, and temporal dispersion and conduction blocks can be present (Kuwabara et al. 2013; Rinaldi 2013; Yuki and Hartung 2012; Chavada and Willison 2012; Hadden et al. 2001; Rajabally and Uncini 2012; Uncini and Kuwabara 2012; Verma et al. 2013).

The CSF white cell count is usually below 50 cells/μl, and in a case series of 134 patients, only 15 patients (11 %) had more than 3 cells/μl (Van der Meché et al. 2001). One study with a limited number of patients showed an increased number of CD123⁺ dendritic cells in the CSF (Press et al. 2005).

Several different antibodies have been found in the CSF of patients with GBS, and a matched pattern of oligoclonal bands may be seen in up to 65 % of patients (Krüger et al. 1981; Segurado et al. 1986). A key feature of GBS is myelin disruption; IgG and IgM antibodies against myelin basic protein (MBP) have been found in the CSF of 56.2 and 53 % of patients, respectively (Marchiori et al. 1990). These antibodies could, however, not distinguish GBS from other inflammatory diseases such as MS and are therefore not suitable as biomarkers (Marchiori et al. 1990). IgG and IgM antibodies against cerebrosides (38.5 and 23 %, respectively) and IgG antibodies against cardiolipin (50 %) were also found (Marchiori et al. 1990). One of the major targets of antibodies detected in the CSF is gangliosides (Marchiori et al. 1990). These include anti-GM1 IgG and IgM in 48 % of GBS cases (Simone et al. 1993); GM1, GD1a, GD1b, and GM2 IgG/IgM (Mata et al. 2006); and GM1, GM2, GM3, AGM1, GD1a, and GD1b IgM or IgG (Brettschneider et al. 2009). Antibodies reacting with heat shock proteins (HSP) have also been found. These include HSP27, HSP60, HSP70, and HSP90 (Yonekura et al. 2004). Antibodies (IgG) targeting another stress protein, αβ-crystallin (αBC), was found to be elevated in the CSF. When using a cutoff of αBC-IgG index >0.8, the specificity was 85 % and the sensitivity was 76 % for GBS (Wanschitz et al. 2008).

Evidence also points to an activation of the complement system in GBS. Elevated levels of complement fragments C3a, C5a (Hartung et al. 1987), and C4d (Koguchi et al. 1995) and the fluid phase complement complex C5b-9 (Sanders et al. 1986) in the CSF have been described. The membrane attack complex C5b-9 was associated with a more severe disease course, being higher in patients dependent on respiratory support (Sanders et al. 1986).

Not surprisingly, levels of cytokines and chemokines are also altered in CSF of patients with GBS. Thus, studies have found increased levels of CXCL10 (IP-10) (Kieseier et al. 2002), MCP-1 (Press et al. 2003), and CX3CL1 (fractalkine) (Kastenbauer et al. 2003). Elevated concentrations of IL-18 (Jander and Stoll 2001), IL-17, IL-22 (Li et al. 2012, 2013), and IL-37 (Li et al. 2013) have been detected. Concentrations of IL-17 and IL-37 correlated with functional disability (Li et al. 2013). A recent study found increased IL-17 levels in the acute but not in the stable phase of both AMAN and AIDP (Han et al. 2014). The same study found an increased CSF osteopontin level in the CSF of both AIDP and AMAN patients (Han et al. 2014). Tumor necrosis factor (TNF)-α mRNA and its soluble 60 kDa receptor (sTNF-R p60) were also upregulated in the CSF (Petzold et al. 1998). Using a multiplex bead-based ELISA assay investigating 32 inflammatory mediators, increased levels of CCL2, CCL7, CCL27, CXCL9, CXCL10, CXCL12, ICAM-1, VCAM-1, and VEGF were detected similar to CIDP (Sainaghi et al. 2010). Moreover, concentrations of IL-8 and IL-1ra were higher in GBS compared to CIDP or controls (Sainaghi et al. 2010).

Levels of the proteolytic enzyme cystatin C were decreased, whereas cathepsin B concentration was increased in the CSF of patients with GBS, similar to CIDP (Nagai et al. 2000). The concentration of another abundant brain protein, prostaglandin D2 synthetase (PGDS), was elevated in the CSF of patients with AIDP compared to controls despite of a decreased intrathecal synthesis, in contrast to Miller Fisher syndrome or CIDP (Huang et al. 2009). However, data regarding PGDS are conflicting, since it was found to be downregulated in a recent proteome analysis (Chang et al. 2007).

Markers of axonal damage have also been investigated. The level of heavy chain subunit of neurofilament (NfH) was 12.5-fold higher in the CSF of patients with electrophysiological evidence of axonal involvement, and high NfH concentration seemed to be associated with worse outcome (Merkies et al. 2002; Petzold et al. 2006, 2009; Dujmovic et al. 2013). Other axonal proteins, such as tau, have not been found to be elevated in GBS, when compared to controls (Süssmutt Süssmuth et al. 2001). However, a small study has suggested that CSF levels might be higher in patients with worse outcome (Jin et al. 2006).

Other neuronal and glial markers investigated include neuron-specific enolase (NSE), which was higher in the CSF of patients with GBS compared to controls, and S100B (Vermuyten et al. 1990; Mokuno et al. 1994). Both have been described to correlate with the number of months to recovery (Mokuno et al. 1994). A more recent multicenter study, however, found that only NfH but not tau, GFAP, or S100B correlated with outcome (Petzold et al. 2009).

One study also reported detectable level of protein 14-3-3 in 29 of 38 patients with GBS (Bersano et al. 2006), in contrast to others (Satoh et al. 1999); this controversy may be explained by methodological differences.

Several recent studies have focused on potential biomarkers by analyzing the CSF proteome in GBS. Almost all studies have found an upregulation of haptoglobin and a downregulation of transthyretin, besides other potentially interesting molecular alterations. These include upregulation of apolipoprotein A-IV, PRO2044, serine/threonine kinase 10, alpha-1-antitrypsin, SNC73, alpha II spectrin, IgG kappa chain, and cathepsin D preprotein (Lehmensiek et al. 2007); orosomucoid and apolipoprotein A-IV (Chang et al. 2007); vitamin D-binding protein, beta-2 glycoprotein I (ApoH), and a complement component C3 isoform (D’Aguanno et al. 2010); and alpha-1-antitrypsin, apolipoprotein A-IV, and neurofilaments (Yang et al. 2009).

Downregulation was reported for fibrinogen (Jin et al. 2007); transferrin, caldesmon, galactose-1-phosphate uridylyltransferase (GALT), human heat shock protein 70, amyloidosis patient HL-heart-peptide 127aa (Lehmensiek et al. 2007); prostaglandin D2 synthase (Chang et al. 2007); apolipoprotein E, albumin, and five of its fragments (D’Aguanno et al. 2010); and cystatin C, apolipoprotein E, and heat shock protein 70 (Yang et al. 2009).

In 1951, Bickerstaff and Cloake described the first cases characterized by external ophthalmoparesis, ataxia, and alteration of consciousness; the syndrome was preceded by infection, and symptoms improved spontaneously (Bickerstaff and Cloake 1951). Bickerstaff described additional cases in 1957. One year earlier, in 1956, Miller Fisher published a similar syndrome with areflexia, ocular nerve palsies, ataxia, and spontaneous improvement. Both authors discussed the similarities to GBS. In 1992, the presence of anti-GQ1b antibodies was discovered in MFS (Chiba et al. 1992), and similar antibodies have later been found in BBE. In 2008, an analysis of 581 cases (53 BBE and 466 MFS) was published, which indicated the clinical and serological overlap of the two diseases (Ito et al. 2008). In MFS, ataxia, ophthalmoparesis, and areflexia are characteristic symptoms without major motor weakness. In BBE, alteration of consciousness is characteristic, deep reflexes can be increased, and Babinski reflex is present in about 10 % of the cases. Ptosis, mydriasis, facial nerve palsy, and peripheral sensory symptoms can be present in both diseases. The spectrum may contain additional rare variants: anti-GQ1b-seropositive acute isolated ophthalmoparesis (n. III and n. VI), acute ataxic neuropathy without ophthalmoparesis, and acute pharyngo-cervico-brachial palsy (PCB). MFS and BBE are most probably caused by antibodies cross-reacting with axonal GQ1b antigens in the paranodal region and at the neuromuscular junction. GQ1b is highly expressed in cranial nerves and Ia afferents of the muscle spindles, which may explain ataxia. Involvement of the brainstem reticular formation may be responsible for the alteration of consciousness in BBE, but experimental evidences are lacking. Both MFS and BBE are preceded by C. jejuni (21–23 %) and H. influenzae (6–8 %) infection, similar to AMAN and AMSAN (Fig. 18.1). In a small number of patients, an overlap syndrome of AMAN, MFS, and BBE may occur characterized by coexistence of anti-GM1, anti-GD1a, and anti-GQ1b antibodies. Despite the alarming symptoms, more than half of the patients completely recover within 6 months; the presence of anti-GQ1b antibodies suggests good prognosis with high sensitivity and specificity (Ito et al. 2008; Odaka et al. 2001; Shahrizaila and Yuki 2013; Chavada and Willison 2012; Yuki et al. 2004).

Reviewing CSF alterations in 375 cases of MFS and 44 cases of BBE indicated albuminocytological dissociation in 37 and 25 %, respectively, during the first week of illness. Pleocytosis was more characteristic for BBE (32 %, 0–668 cells/μl) and occurred only in 4 % (0–105 cells/μl) of patients with MFS. In the second week, albuminocytological dissociation became more frequent in MFS and occurred in 76 % of patients in contrast to 46 % in BBE. There was no change in the frequency of pleocytosis, which remained 5 % in MFS and 31 % in BBE. These data may indicate a more severe breakdown of the blood-CSF barrier in BBE, but examination of the CSF cannot discriminate between BBE and MFS (Ito et al. 2008).

Analyzing the anti-GQ1b syndrome, 58 % of patients had albuminocytological dissemination: it was the highest among patients with BBE/GBS (75 %) and occurred in 66 % of patients with MFS, in 43 % of those with MFS/GBS, and in 42 % of those with BBE (Odaka et al. 2001). Comparing CSF protein concentrations between MFS and GBS, protein content was higher in GBS in the first week (25 % vs 44 %) but increased continuously during the first 3 weeks in both diseases (84 % vs. 75 %) up to 1.5 g/l. CSF albuminocytological dissociation was present in 59 % of patients with MFS and in 62 % of those with GBS. One-third of patients with MFS had anti-GQ1b antibody in the sera but no albuminocytological dissociation in the CSF; in contrast, only 7 % presented albuminocytological dissociation without anti-GQ1b antibodies, indicating the higher sensitivity of antibody testing in MFS. In GBS, the frequency of albuminocytological dissociation was higher than the presence of anti-ganglioside antibodies (anti-GM1, anti-GD1a, and anti-GQ1b) without albuminocytological dissociation. These findings suggest that during the first 3 weeks, testing of such antibodies is inferior to the examination of the CSF for supporting a diagnosis of GBS, whereas anti-GQ1b antibody testing is superior to a CSF examination for the diagnosis of MFS. This is, however, true only in the first week of MFS (anti-GQ1b 48 % vs. albuminocytological dissociation only 4 %): after the second week, there is no difference in the frequency of anti-GQ1b antibodies and albuminocytological dissociation in MFS (Nishimoto et al. 2004).

Recurrent MFS is a rare condition, altogether 28 cases are described. It is slightly more common in men (16 vs. 12 cases) with an average age at onset of 34 years and an average age at the last episode of 47 years. Cerebrospinal fluid in 41 such episodes displayed elevated protein in 52 % of the cases (0.2–2.1 g/l). Pleocytosis was present only in two cases. Repeated CSF examination in two cases indicated a slight increase in the protein content (Heckmann and Dütsch 2012).

Some special biomarkers have been also examined in the CSF of patients with MFS, mostly validating previous data obtained in GBS. Proteomic analysis of the CSF indicated several proteins to be up- and downregulated in patients with GBS: cathepsin D preprotein, haptoglobin, cystatin C, and prostaglandin D₂ synthase were validated by Western blot or ELISA (Chang et al. 2007; Jin et al. 2007; Lehmensiek et al. 2007; Yang et al. 2009). Transthyretin, considered to be neuroprotective in traumatic brain injury and Alzheimer disease (Zou et al. 1998; Long et al. 2003; Merched et al. 1998; Stein et al. 2004), was consistently downregulated; therefore, it was later examined by ELISA in the CSF of patients with neurological diseases including GBS and MFS (Chiang et al. 2009). The mean protein content (1.08 and 1.23 g/l) was elevated in both diseases, while the albumin level, the albumin ratio, the transthyretin concentration, and the ratio of transthyretin/total protein were elevated only in GBS. The authors discussed that such elevation in GBS in contrast to downregulation in other studies might be explained by methodological differences, measuring the target protein relative to the total protein content in the previous studies; the elevated transthyretin concentration may be caused by barrier dysfunction. In the four patients with MFS, transthyretin concentration in the CSF was not different compared to controls or GBS. Another downregulated protein, prostaglandin D₂ synthase (PGDS), the most abundant brain-synthetized protein in the CSF catalyzing the synthesis of the proinflammatory prostaglandin D₂ (PGD₂), was also investigated in 18 patients with AIDP and nine patients with MFS by using Western blot (Huang et al. 2009). In contrast to AIDP, where concentration of PGDS was increased in the CSF despite of a presumably decreased intrathecal synthesis, neither the concentration nor the PGDS/albumin ratio was different from controls.

In a single case, cytokines and chemokines were measured in the serum and CSF in the acute and recovery phase of MFS. Concentration and CSF/serum ratio of two chemokines, MCP-1 and IL-8, were elevated in the acute phase and decreased after IVIG treatment (Sato et al. 2009). MCP-1 has been suggested to be involved in the infiltration of spinal nerve roots by macrophages in GBS (Press et al. 2003). IL-8, a chemoattractant for neutrophils and monocytes, has not been implicated in MFS or GBS earlier.

CSF hypocretin-1 levels were reported to be reduced in MFS similar to GBS and CIDP (Nishino et al. 2003).

The first cases of GBS without demyelination were described in 1986. Thereafter, two Japanese cases were published in 1990, which were characterized by preceding infection with C jejuni and by association with anti-GM1 antibodies. The disease was termed AMAN in 1993 (Kuwabara and Yuki 2013; Rinaldi 2013). Based on experimental data, pathogenetic antibodies generated against the lipo-oligosaccharides of C. jejuni (mainly anti-GM1 and anti-GD1a) cross-react with the motor axolemma (Chavada and Willison 2012; Yuki et al. 2004). This results in the disappearance of the sodium channels at the Ranvier nodules and complement-mediated disruption of the paranodal myelin. The antibodies can also interfere with the function of the sodium channels, which may be responsible for the reversible conduction failure. In contrast to Asia, AMAN is rare in Europe and North America and is responsible for about 3–17 % of cases (Fig. 18.1). Characteristically, the progression of the disease is faster compared to AIDP: the ascending, symmetric peripheral motor weakness peaks within 5–9 days. Cranial nerve palsies and autonomic dysfunction are less frequent. Rapid atrophy of the muscles may evolve, and the prognosis is usually poorer (Kuwabara et al. 2013; Rinaldi 2013; Yuki and Hartung 2012).