Abstract

It has been reported recently that Haemophilus influenzae can elicit an axonal form of Guillain–Barré syndrome. To investigate the incidence and features of H. influenzae-related Guillain–Barré syndrome, anti-H. influenzae antibody titres were measured by enzyme-linked immunosorbent assay (ELISA) in 46 consecutive Japanese patients with Guillain–Barré syndrome, 49 normal controls, 24 patients with multiple sclerosis and 27 patients with amyotrophic lateral sclerosis (ALS). Whole bacteria of non-encapsulated (non-typable) H. influenzae isolated from one of the Guillain–Barré syndrome patients was the antigen used. Elevated anti-H. influenzae antibodies for two or three classes of IgG, IgM and IgA were found in six (13%) Guillain–Barré syndrome patients, but not in the normal controls and patients with multiple sclerosis or ALS. The incidence was significantly higher in patients with Guillain–Barré syndrome than in the normal controls (P = 0.01) and patients with multiple sclerosis or ALS (P = 0.009). Western blot analysis confirmed that the H. influenzae-positive patients' IgG recognized the lipopolysaccharides of H. influenzae. Guillain–Barré syndrome patients with anti-H. influenzae antibodies showed relatively uniform clinical and laboratory features: prodromal respiratory infection, less frequent cranial and sensory nerve involvement, pure motor axonal degeneration on electrophysiology, and positivity for IgG anti-GM1 antibodies. Although the features were similar to those in Guillain–Barré syndrome patients infected by Campylobacter jejuni, the recoveries seemed to be better in patients with H. influenzae-related Guillain–Barré syndrome. It is concluded that a form of Guillain–Barré syndrome occurs after respiratory infection by H. influenzae in the Japanese population. A particular strain of non-typable H. influenzae has a ganglioside GM1-like structure and elicits axonal Guillain–Barré syndrome similar to C. jejuni-related Guillain–Barré syndrome.

Introduction

Guillain–Barré syndrome, an acute inflammatory, predominantly motor polyneuropathy (Asbury and Cornblath, 1990), is preceded by an acute infectious illness in approximately two-thirds of cases, the onset of neuropathic symptoms being preceded by respiratory infection in 30–45% of cases and by gastrointestinal infection in 15–20% of cases (Winer et al., 1988; Boucquey et al., 1991). Various infectious agents are reported as possibly having an association with Guillain–Barré syndrome (Winer et al., 1988; Ropper et al., 1991), but the causal relationship between the infectious agent for prodromal illness and peripheral nerve damage is not known. Campylobacter jejuni was recently identified as the most common cause of preceding infection in Guillain–Barré syndrome (Kaldor and Speed, 1984; Winer et al., 1988; Mishu and Blaser, 1993; Rees et al., 1995b), and C. jejuni-related Guillain–Barré syndrome is speculated to be associated with axonal degeneration of the motor nerves (Yuki et al., 1990; Rees et al., 1995a; Visser et al., 1995) or positivity for anti-ganglioside GM1 antibody (Rees et al., 1995a; Jacobs et al., 1996). Molecular mimicry between the GM1 in neural tissue and the GM1-like structure on the lipopolysaccharide (LPS) of C. jejuni has also been suggested (Yuki et al., 1992a, 1993; Oomes et al., 1995). The majority of Guillain–Barré syndrome patients, however, have had prodromal respiratory infection. Cytomegalovirus (CMV), Epstein–Barr virus (EBV) and Mycoplasma pneumoniae are also possible causes of respiratory tract infections that precede Guillain–Barré syndrome, but their frequencies are <15% (Winer et al., 1988; Boucquey et al., 1991; Hao et al., 1998; Jacobs et al., 1998).

We reported recently a patient with Guillain–Barré syndrome subsequent to respiratory infection by Haemophilus influenzae and the possible presence of a GM1-like structure on the surface of this bacterium (Mori et al., 1999). To investigate the incidence and the clinical and laboratory features of H. influenzae-related Guillain–Barré syndrome, we carried out a retrospective case–control study of 46 Guillain–Barré syndrome patients in whom anti-H. influenzae antibodies had been detected.

Methods

Subjects

Forty-six consecutive Guillain–Barré syndrome patients seen at Chiba University Hospital or its affiliated hospitals between January 1989 and June 1998 were studied. All fulfilled the clinical criteria for Guillain–Barré syndrome (Asbury and Cornblath, 1990). Clinical disabilities were evaluated by Hughes functional grading (Hughes et al., 1978), and patients were followed for up to 3 months after onset. Serum samples obtained within 4 weeks of the onset of neurological symptoms were stored at –80°C until used. Whenever possible, sputum or stool samples were collected from each patient. Serum samples from 49 normal controls and 51 patients with other neurological diseases seen during the same period, including 24 patients with multiple sclerosis and 27 with amyotrophic lateral sclerosis (ALS), who had not had any obvious infectious events during the 4 weeks before the collection of the blood sample, were the controls.

Laboratory investigations

Anti-H. influenzae antibody assay

Serum anti-H. influenzae antibody titre was measured by the enzyme-linked immunosorbent assay (ELISA), as described previously (Kerttula et al., 1987; Yuki et al., 1992b; Burman et al., 1994; Kurtti et al., 1997), but with the following minor modifications. Strains of non-typable H. influenzae isolated from the sputum of one of the patients with Guillain–Barré syndrome (Mori et al., 1999) enrolled in this study were used to prepare the H. influenzae antigen. They were grown on chocolate agar for 48 h at 37°C in 5% CO2, inactivated in 1% formaldehyde, then washed. The product was used as the H. influenzae antigen. Wells of polystyrene plates were coated with the antigen in 100 μl carbonate buffer (50 mmol/l NaHCO3, 50 mmol/l Na2CO3, pH 9.6) then washed with solution A (20 mmol/l Tris buffer, pH 7.4, 100 mmol/l NaCl, 0.5% skim milk, 0.05% Tween-20), after which they were blocked with 200 μl solution B (20 mmol/l Tris buffer, pH 7.4, 100 mmol/l NaCl, 3% skim milk, 0.05% Tween-20) overnight. The plates were then incubated overnight with 100 μl serum diluted 1 : 1000 in solution B, after which they were washed and incubated for 90 min at room temperature with peroxidase-conjugated goat anti-human IgG, IgM and IgA diluted 1 : 2000 in solution A. After another wash, the plates were developed with the substrate solution [containing 2,2′-azino-bis-(3-ethylbenzothiazoline sulphonate) (ABTS; Zymed, San Francisco, Calif., USA) and hydrogen peroxide in 0.1 M citrate buffer]. Optical density (OD) values at 405 nm were corrected by subtracting the OD obtained for each well without antigen. Serum was considered anti-H. influenzae-positive if its optical density was more than three standard deviations above the mean value for the 49 normal control samples. Each positive sample was diluted serially, starting at 1 : 1000. Antibody titre was defined as the highest serum dilution at which the OD was ≥0.1. The same assay was carried out twice to confirm the reproducibility of the results. Subjects were considered H. influenzae-positive if they had elevated anti-H. influenzae antibody for two or three classes of IgG, IgM or IgA.

Western blot analysis

Purified H. influenzae LPS (Westphal et al., 1965) was prepared as described previously, from the non-typable strain used in the ELISA described above. Escherichia coli LPS was purchased from Difco Laboratories (Mich., USA). Sodium dodecyl sulphate–polyacrylamide gel electrophoresis was performed according to Laemmli (Laemmli, 1970); the separating gels contained 15% acrylamide and 2 M urea. All the preparations were transferred electrophoretically onto polyvinylidene difluoride sheets. Unreactive binding sites were blocked with blocking solution containing 3% skim milk in phosphate-buffered saline. The blots were incubated for 2 h at room temperature with patients' sera diluted 1 : 200 in washing solution containing 0.2 M NaCl, 0.2% Triton X-100, and 0.05 M Tris–HCl, pH 7.5. The strips were washed, then incubated for 2 h at room temperature with peroxidase-conjugated antibodies to human IgG diluted 1 : 1000 in washing solution. After another wash, the reaction products were developed with enhanced chemiluminescence reagents (Amersham, San Francisco, Calif., USA) and exposed to film.

Anti-ganglioside antibody assay

IgG and IgM class antibodies to gangliosides GM2, GM1, GM1b, GD1a, GalNac-GD1a, GD1b, GQ1b and GT1b were measured by an ELISA described elsewhere (Koga et al., 1998b) by one person (M. Koga, School of Medicine, Dokkyo University). Antibody titre was the highest serum dilution at which the optical density at 492 nm was ≥0.1. Serum was considered positive when the titre was ≥1 : 500.

Assays for anti-C. jejuni, anti-CMV and anti-M. pneumoniae antibodies

IgG anti-C. jejuni antibody also was measured by an ELISA described elsewhere (Koga et al., 1998b). IgM antibodies against CMV were measured with the ELISA, IgM antibodies against EBV-specific capsid antigen with the indirect fluorescent antibody test, and M. pneumoniae antibodies with a commercial particle agglutination and complement fixation test (Special Reference Laboratory, Tokyo, Japan).

Electrodiagnostic studies

Nerve conduction studies were done by the conventional procedure within 3 weeks of onset of neurological signs. Patients were classified as having acute motor axonal neuropathy (AMAN) (motor axonal Guillain–Barré syndrome) or acute inflammatory demyelinating polyneuropathy (AIDP) (demyelinating Guillain–Barré syndrome) using the electrodiagnostic criteria of Ho and colleagues (Ho et al., 1995).

Statistical analysis

Differences in proportions were tested with Fisher's exact test using StatView version 4.0 software.

Results

Anti-H. influenzae antibody assay

ELISA results of the IgG, IgM and IgA anti-H. influenzae antibody assays are shown in Fig. 1. Elevated anti-H. influenzae antibodies for two or three classes of IgG, IgM or IgA were present in six (13%) of the patients with Guillain–Barré syndrome, but not in the normal controls or patients with other neurological diseases (multiple sclerosis or ALS). In the Guillain–Barré syndrome patients, antibodies were classified as IgG in six cases, IgM in four, IgA in four, both IgG and IgM in four, and both IgG and IgA in two. No patient had both IgM and IgA. All the antibodies in the normal controls and in four patients with other neurological diseases were of the IgM class (one multiple sclerosis and three ALS patients). Positive serology for H. influenzae infection was significantly more frequent in patients with Guillain–Barré syndrome than in the normal controls (P = 0.01) and patients with other neurological diseases (P = 0.009). In patients with Guillain–Barré syndrome, IgG anti-H. influenzae antibody was positive more frequently than in normal controls (P = 0.02) and patients with other neurological diseases (P = 0.02). There were no significant differences in the frequencies of elevated IgM and IgA anti-H. influenzae antibodies among the three groups. Figure 2 shows the serial titres of IgG anti-H. influenzae antibodies in H. influenzae-positive patients. Titres of all the IgG, IgM and IgA antibodies decreased gradually during the course of the illness.

Other infection serology was positive for C. jejuni [nine patients (20%)], cytomegalovirus [six patients (13%)] and M. pneumoniae [three patients (7%)] in the 46 Guillain–Barré syndrome patients.

Western blot analysis showed that the sera from all of six H. influenzae-positive Guillain–Barré syndrome patients recognized bands with a relative molecular mass of ~10 kDa on LPS blots from H. influenzae (Fig. 3) and that the sera from all CMV-positive (n = 3) and C. jejuni-positive (n = 3) Guillain–Barré syndrome patients did not recognize this band (Fig. 3). None of the sera from the six H. influenzae-positive Guillain–Barré syndrome patients had antibody against the LPS from E. coli.

Clinical characteristics of H. influenzae-positive Guillain–Barré syndrome patients

The clinical, electrophysiological and serological features of these six patients are given in Tables 1 and 2. Respiratory infections preceded the development of Guillain–Barré syndrome by 4–12 days in all the patients. Clinical features were characterized by less frequent cranial and sensory nerve involvement. Evolution was complete within 6 days. Clinical recovery was remarkably good, all six patients being able to walk unaided within 4 weeks and to run within 3 months.

Sputum samples were obtained from five of the H. influenzae-positive patients, and Haemophilus species were isolated from four of these specimens. Three were classified by species: two H. influenzae and one H. parainfluenzae. Both of the H. influenzae strains isolated from Guillain–Barré syndrome patients were serotyped as non-typable (non-encapsulated).

Electrodiagnostic studies showed the AMAN pattern in four (67%) of the H. influenzae-positive patients. The four patients had low compound muscle action potentials after distal stimulation (80% of normal lower limits) in multiple nerves, but no prolongation of distal latencies, decreases in conduction velocities, or temporal dispersion of responses. On electromyography 2 or 3 weeks after onset, there were ample fibrillation potentials in three of the four patients. In the other two patients who did not have the AMAN pattern, the absence of F waves was an isolated abnormality, and peripheral nerve conduction was entirely normal. Sensory nerve conduction was normal in all six patients. In all six patients, anti-ganglioside antibodies were present. Positivity for IgG anti-ganglioside antibodies was found for ganglioside GM1 (n = 5), GM1b (n = 3), GD1a (n = 4), GalNac-GD1a (n = 1), GD1b (n = 5) or GQ1b (n = 1), whereas positivity for IgM anti-ganglioside antibodies was found for ganglioside GM1 (n = 1), GM1b (n = 1), GD1b (n = 1) or GQ1b (n = 1). None of the H. influenzae-positive patients showed evidence of recent infection by C. jejuni, CMV, EBV or M. pneumoniae.

Relationship of H. influenzae-positive Guillain–Barré syndrome to axonal Guillain–Barré syndrome and to anti-ganglioside antibody

On the basis of their peripheral conduction abnormalities, 46 Guillain–Barré syndrome patients were classified as having AIDP (n = 18) or AMAN (n = 21), or were unclassified (n = 7). Of the 21 with AMAN, 19% (n = 4) were H. influenzae-positive.

Of the 46 patients with Guillain–Barré syndrome, 35 had IgG or IgM anti-ganglioside antibodies. IgG anti-ganglioside antibodies were found for the gangliosides GM1 (n = 20), GM1b (n = 17), GD1a (n = 14), GalNac-GD1a (n = 8), GD1b (n = 12) and GQ1b (n = 5), and IgM anti-ganglioside antibodies were found for GM1 (n = 6), GM1b (n = 3), GM2 (n = 3), GD1b (n = 1) and GQ1b (n = 1). Anti-H. influenzae antibody was present in 17% of the patients carrying IgG and/or IgM anti-ganglioside antibodies, 25% of those with IgG anti-GM1 antibody, 42% of those with IgG anti-GD1b antibody and 29% of those with IgG anti-GD1a antibody.

Discussion

The aim of this retrospective study was to clarify the association between H. influenzae infection and Guillain–Barré syndrome using a case–control design, and to clarify the clinical and laboratory features of Guillain–Barré syndrome subsequent to H. influenzae infection. We found that a form of Guillain–Barré syndrome occurs after respiratory infection by H. influenzae in the Japanese population. Guillain–Barré syndrome patients with H. influenzae infection were characterized clinically as having had a preceding respiratory tract infection, less frequent cranial and sensory nerve involvement, and good recovery. Moreover, they frequently had primary axonal involvement of the motor fibres and carried anti-ganglioside antibodies. Most of these features agree with those of the Guillain–Barré syndrome associated with C. jejuni enteritis (Rees et al., 1995a; Visser et al., 1995; Jacobs et al., 1996). Our previous finding that a strain of H. influenzae has a GM1-like structure and elicits Guillain–Barré syndrome (Mori et al., 1999) is supported by the strong association between the anti-H. influenzae positivity and the anti-GM1 positivity and the results of Western blotting reported in this study.

Haemophilus influenzae is a Gram-negative human pathogen whose outer membrane contains LPS. There are two types of strain: encapsulated and non-encapsulated, non-typable (Apicella, 1994). The latter strain was used as the antigen in the ELISA carried out in this study. Previous reports suggest a causal relationship between Guillain–Barré syndrome and the H. influenzae type b (encapsulated type) vaccine, which is composed of the capsular antigen of the bacteria (D'Cruz et al., 1989; Gervaix et al., 1993), but subsequent evaluation has not confirmed this association (Gross and Hayes, 1991). The non-typable (non-encapsulated) H. influenzae, however, has been shown to be an important pathogenetic cause of community-acquired bacterial pneumonia in adults (Apicella, 1994), but had not been described as a pathogen of Guillain–Barré syndrome until our previous report (Mori et al., 1999).

Jacobs and colleagues reported that only 1% of 154 patients with Guillain–Barré syndrome had elevated anti-H. influenzae antibody and that there was no significant difference in the presence of the antibody when compared with the controls, including patients with other neurological diseases (Jacobs et al., 1998). There are a number of reasons for the different frequencies of Guillain–Barré syndrome patients with anti-H. influenzae antibody in their study and ours. Jacobs and colleagues did not give a detailed description of the H. influenzae used as antigen in the ELISA (they did not say whether the strain used was capsulated or was Guillain–Barré syndrome-related). Because the H. influenzae strain we isolated from Guillain–Barré syndrome patients was serotyped as non-typable and most pathogens that cause respiratory infections in adults are non-typable, we speculate that only non-typable H. influenzae causes Guillain–Barré syndrome. The use of our isolated strain as the antigen may have yielded a higher percentage of H. influenzae-positive Guillain–Barré syndrome cases. In addition, a difference in the populations of Guillain–Barré syndrome patients studied may have affected the frequency. Jacobs and colleagues studied only patients with moderate to severe Guillain–Barré syndrome who were unable to walk 10 m independently, whereas we included all consecutive Guillain–Barré syndrome patients regardless of disease severity. In fact, one of our H. influenzae-positive patients could walk unaided at the peak of illness. Finally, geographical differences may produce a difference in frequency. Different immunogenetic backgrounds, for example different human leukocyte antigen haplotypes, may produce different susceptibilities to H. influenzae, as they do to the C. jejuni (Koga et al., 1998a). Japanese Guillain–Barré syndrome patients have been classified much more frequently as having the AMAN pattern or being anti-GM1-positive (Kuwabara et al., 1998b) than patients in Western countries, where anti-GM1 antibody has been found in 10–25% of patients with Guillain–Barré syndrome (Vriesendorp et al., 1993; Rees et al., 1995a; Jacobs et al., 1996) and in only 7% of those with AMAN (Rees et al., 1995a).

False-positivity may occur in ELISA because H. influenzae is part of the normal flora of the upper respiratory tract and may occur together with other viral infections; moreover, there may be cross-reactions between H. influenzae and other bacterial species (Murphy and Apicella, 1987; Loosmore et al., 1997). Overestimation, however, is unlikely because our ELISA method for the detection of anti-H. influenzae antibody is specific when used for the aetiological diagnosis of pneumonia (Kerttula, 1987); H. influenzae was detected from purulent sputum specimens of two H. influenzae-positive Guillain–Barré syndrome patients; and there was no evidence of recent infection by any agent other than H. influenzae in any of the sputum and serum specimens tested.

All the H. influenza-positive Guillain–Barré syndrome patients had antibodies of the IgG class, but did not always have IgM or IgA antibodies. Because Guillain–Barré syndrome develops 1–3 weeks after an infection, the titre of IgM or IgA antibody may decrease until the onset of neuropathic symptoms. In fact, the IgG anti-C. jejuni antibody assay is more sensitive than the IgM assay for C. jejuni-related Guillain–Barré syndrome (Koga et al., 1998b). IgG class antibodies may function in the pathology of Guillain–Barré syndrome.

Structural studies of H. influenzae LPS show that the LPS molecules from some strains of H. influenzae have a ganglioside-like structure (Mandrell et al., 1992; Gibson et al., 1993; Risberg et al., 1997). Although GM1 was not found, the structure of the LPS was shown to differ among the strains. Elsewhere we have reported that a strain of H. influenzae may have a GM1-like structure. In addition, almost all the anti-H. influenzae-positive patients in this study had elevated IgG anti-GM1 antibodies. Moreover, Western blot analysis showed the presence of IgG antibodies against H. influenzae LPS in all the H. influenzae-positive patients, whereas these antibodies were not present in CMV-positive or C. jejuni-positive Guillain–Barré syndrome patients. We speculate (i) that there is an H. influenzae strain that causes Guillain–Barré syndrome, (ii) that this strain, like a certain C. jejuni strain, may carry a GM1-like epitope on its LPS, and (iii) that molecular mimicry between the ganglioside and the bacterial epitope may cause nerve damage because of an immune response. The similarities between Guillain–Barré syndrome subsequent to H. influenzae and C. jejuni infections may reflect the same pathophysiology. Despite the high incidence of H. influenzae infection in the overall population, Guillain–Barré syndrome is a rare disease, probably because only a particular strain causes Guillain–Barré syndrome or because persons who develop the disease differ in their responses to the infection.

In our study, all the H. influenzae-positive Guillain–Barré syndrome patients could walk unaided within 4 weeks of the nadir of the disease and could run within 3 months. Moreover, there seemed to be a better prognosis than that previously described for C. jejuni-related Guillain–Barré syndrome patients (Rees et al., 1995a; Visser et al., 1995; Jacobs et al., 1996). Some reports have denied that there is a relationship between anti-GM1 antibody and C. jejuni infection or between anti-GM1 antibody and a poor prognosis (Enders et al., 1993; Vriesendorp et al., 1993). We found no significant correlation between positivities for anti-GM1 and anti-H. influenzae antibodies. This may be because of the heterogeneity of the anti-GM1 antibody-positive patient group, which consisted of at least two subgroups of patients with Guillain–Barré syndrome subsequent to H. influenzae infection: those without anti-C. jejuni antibody who had a good recovery and those with Guillain–Barré syndrome subsequent to C. jejuni infection who had a relatively poor recovery (Kuwabara et al., 1998a).

In conclusion, H. influenzae infection elicits an axonal form of Guillain–Barré syndrome associated with anti-ganglioside antibodies similar to those present in C. jejuni-related Guillain–Barré syndrome, but there is better recovery from the H. influenzae-related form of the disease.

Table 1

Clinical features of H. influenzae-positive Guillain–Barré syndrome patients

 Patient 
 
*0 = healthy; 1 = minor signs and symptoms; 2 = able to walk 5 m independently; 3 = able to walk 5 m with aid; 4 = bedbound; 5 = assisted respiration required; 6 = dead. IVIg = intravenous immunoglobulin therapy; PP = plasmapheresis. 
Age/sex 23/F 33/F 41/F 30/M 25/F 80/F 
Prodromal symptoms Cough, sputum Cough, sputum Cough, sore throat Fever, sputum, sore throat Fever, cough, sputum, sore throat, rhinorrhoea Cough, sputum, rhinorrhoea 
Cranial nerve involvement No No No No Right abducens Bulbar 
Sensory loss No No No No All modalities No 
Treatment IVIg PP PP None IVIg PP 
Disability (Hughes grade*)       
At peak 
At 4 weeks 
At 3 months 
Time until nadir (days) 
 Patient 
 
*0 = healthy; 1 = minor signs and symptoms; 2 = able to walk 5 m independently; 3 = able to walk 5 m with aid; 4 = bedbound; 5 = assisted respiration required; 6 = dead. IVIg = intravenous immunoglobulin therapy; PP = plasmapheresis. 
Age/sex 23/F 33/F 41/F 30/M 25/F 80/F 
Prodromal symptoms Cough, sputum Cough, sputum Cough, sore throat Fever, sputum, sore throat Fever, cough, sputum, sore throat, rhinorrhoea Cough, sputum, rhinorrhoea 
Cranial nerve involvement No No No No Right abducens Bulbar 
Sensory loss No No No No All modalities No 
Treatment IVIg PP PP None IVIg PP 
Disability (Hughes grade*)       
At peak 
At 4 weeks 
At 3 months 
Time until nadir (days) 
Table 2

Laboratory features of H. influenzae-positive Guillain–Barré syndrome patients

 Patient 
 
*Unclassified Haemophilus species. 
Sputum culture Not done H. parainfluenzae HaemophilusH. influenzae Negative H. influenzae 
IgG/IgM/IgA       
Anti-H. influenzae antibody titre 32 000/2000/(–) 32 000/2000/(–) 64 000/(–)/1000 64 000/2000/(–) 8000/2000/(–) 128 000/(–)/1000 
Electrophysiological diagnosis AMAN AMAN Isolated F wave    
Absence AMAN Isolated F wave     
Absence AMAN      
Anti-ganglioside antibody (titre)       
IgG GM1 (2000) GM1 (4000) GD1a (500) GM1 (8000) GM1 (500) GM1 (32 000) 
  GM1b (1000) GD1b (500) GM1b (32 000) GD1a (500) GM1b (512 000) 
  GD1a (2000)  GD1b (1000) GD1b (2000) GD1a (128 000) 
  GalNAc-GD1a (4000)   GQ1b (64 000) GD1b (8000) 
  GD1b (4000)     
IgM    GM1b (500) GM1 (1000)  
GD1b (500)       
GQ1b (4000)       
 Patient 
 
*Unclassified Haemophilus species. 
Sputum culture Not done H. parainfluenzae HaemophilusH. influenzae Negative H. influenzae 
IgG/IgM/IgA       
Anti-H. influenzae antibody titre 32 000/2000/(–) 32 000/2000/(–) 64 000/(–)/1000 64 000/2000/(–) 8000/2000/(–) 128 000/(–)/1000 
Electrophysiological diagnosis AMAN AMAN Isolated F wave    
Absence AMAN Isolated F wave     
Absence AMAN      
Anti-ganglioside antibody (titre)       
IgG GM1 (2000) GM1 (4000) GD1a (500) GM1 (8000) GM1 (500) GM1 (32 000) 
  GM1b (1000) GD1b (500) GM1b (32 000) GD1a (500) GM1b (512 000) 
  GD1a (2000)  GD1b (1000) GD1b (2000) GD1a (128 000) 
  GalNAc-GD1a (4000)   GQ1b (64 000) GD1b (8000) 
  GD1b (4000)     
IgM    GM1b (500) GM1 (1000)  
GD1b (500)       
GQ1b (4000)       
Fig. 1

ELISA values for IgG anti-H. influenzae antibodies (A), IgM anti-H. influenzae antibodies (B) and IgA anti-H. influenzae antibodies (C). GBS = Guillain–Barré syndrome; NC = normal controls; OND = other neurological diseases (multiple sclerosis and ALS). Dashed lines show cut-off values (3 SD above the mean value for normal controls).

Fig. 1

ELISA values for IgG anti-H. influenzae antibodies (A), IgM anti-H. influenzae antibodies (B) and IgA anti-H. influenzae antibodies (C). GBS = Guillain–Barré syndrome; NC = normal controls; OND = other neurological diseases (multiple sclerosis and ALS). Dashed lines show cut-off values (3 SD above the mean value for normal controls).

Fig. 2

Longitudinal study of IgG anti-H. influenzae antibody titre in the anti-H. influenzae-positive Guillain–Barré syndrome patients. Data for five of the six patients are shown (the serum of the sixth patient was not sampled sequentially after 2 weeks).

Fig. 2

Longitudinal study of IgG anti-H. influenzae antibody titre in the anti-H. influenzae-positive Guillain–Barré syndrome patients. Data for five of the six patients are shown (the serum of the sixth patient was not sampled sequentially after 2 weeks).

Fig. 3

Reactivity of serum antibodies against the purified H. influenzae LPS in H. influenzae-positive Guillain–Barré syndrome patients, one CMV-positive Guillain–Barré syndrome patient [CMV(+)] and one C. jejuni-positive Guillain–Barré syndrome patient [C. jejuni(+)].

Fig. 3

Reactivity of serum antibodies against the purified H. influenzae LPS in H. influenzae-positive Guillain–Barré syndrome patients, one CMV-positive Guillain–Barré syndrome patient [CMV(+)] and one C. jejuni-positive Guillain–Barré syndrome patient [C. jejuni(+)].

We wish to thank Michiaki Koga and Nobuhiro Yuki (Department of Neurology, Dokkyo University School of Medicine, Japan) for measuring the anti-ganglioside and anti-C. jejuni antibodies and Kenji Takagi (Department of Neurology, Matsudo Municipal Hospital) for providing the clinical information. We also wish to thank Ms Setsuko Kubo, Mr Masaharu Watanabe, Ms Naoko Ishiyama and Mr Shouta Murata (Biological Laboratory Department, Chiba University Hospital) and Ms Emi Utsuno (Second Department of Microbiology, Chiba University School of Medicine) for technical assistance.

References

Apicella MA. Infections due to Haemophilus influenzae, HACEK group, Capnocytophaga, and H. ducreyi. In: Isselbacher KJ, Braunwald E, Wilson JD, Martin JB, Fauci AS, Kasper DL, editors. Harrison's principles of internal medicine. 13th ed. New York: McGraw-Hill; 1994. p. 652–4.
Asbury AK, Cornblath DR. Assessment of current diagnostic criteria for Guillain–Barré syndrome. [Review].
Ann Neurol
 
1990
;
27 Suppl
:
S21
–4.
Boucquey D, Sindic CJ, Lamy M, Delmée M, Tomasi JP, Laterre EC. Clinical and serological studies in a series of 45 patients with Guillain–Barré syndrome.
J Neurol Sci
 
1991
;
104
:
56
–63.
Burman LÅ, Leinonen M, Trollfors B. Use of serology to diagnose pneumonia caused by nonencapsulated Haemophilus influenzae and Moraxella catarrhalis.
J Infect Dis
 
1994
;
170
:
220
–2.
D'Cruz OF, Shapiro ED, Spiegelman KN, Leicher CR, Breningstall GN, Khatri BO, et al. Acute inflammatory demyelinating polyradiculoneuropathy (Guillain–Barré syndrome) after immunization with Haemophilus influenzae type b conjugate vaccine.
J Pediatr
 
1989
;
115
:
743
–6.
Enders U, Karch H, Toyka KV, Michels M, Zielasek J, Pette M, et al. The spectrum of immune responses to Campylobacter jejuni and glycoconjugates in Guillain–Barré syndrome and in other neuroimmunological disorders.
Ann Neurol
 
1993
;
34
:
136
–44.
Gervaix A, Caflisch M, Suter S, Haenggeli C-A. Guillain–Barré syndrome following immunisation with Haemophilus influenzae type b conjugate vaccine.
Eur J Pediatr
 
1993
;
152
:
613
–4.
Gibson BW, Melaugh W, Phillips NJ, Apicella MA, Campagnari AA, Griffiss JM. Investigation of the structural heterogeneity of lipooligosaccharides from pathogenic Haemophilus and Neisseria species and of R-type lipopolysaccharides from Salmonella typhimurium by electrospray mass spectrometry.
J Bacteriol
 
1993
;
175
:
2702
–12.
Gross TP, Hayes SW. Haemophilus conjugate vaccine and Guillain–Barré syndrome [letter].
J Pediatr
 
1991
;
118
:
161
.
Hao Q, Saida T, Kuroki S, Nishimura M, Nukina M, Obayashi H, et al. Antibodies to gangliosides and galactocerebroside in patients with Guillain–Barré syndrome with preceding Campylobacter jejuni and other identified infections.
J Neuroimmunol
 
1998
;
81
:
116
–26.
Ho TW, Mishu B, Li CY, Gao CY, Cornblath DR, Griffin JW, et al. Guillain–Barré syndrome in northern China. Relationship to Campylobacter jejuni infection and anti-glycolipid antibodies.
Brain
 
1995
;
118
:
597
–605.
Hughes RA, Newsom-Davis JM, Perkin GD, Pierce JM. Controlled trial of prednisolone in acute polyneuropathy.
Lancet
 
1978
;
2
:
750
–3.
Jacobs BC, van Doorn PA, Schmitz PI, Tio-Gillen AP, Herbrink P, Visser LH, et al. Campylobacter jejuni infections and anti-GM1 antibodies in Guillain–Barré syndrome.
Ann Neurol
 
1996
;
40
:
181
–7.
Jacobs BC, Rothbarth PH, van der Meché FG, Herbrink P, Schmitz PI, de Klerk MA, et al. The spectrum of antecedent infections in Guillain–Barré syndrome. A case–control study.
Neurology
 
1998
;
51
:
1110
–5.
Kaldor J, Speed BR. Guillain–Barré syndrome and Campylobacter jejuni: a serological study.
Br Med J
 
1984
;
288
:
1867
–70.
Kerttula Y, Leinonen M, Koskela M, Mäkelä PH. The aetiology of pneumonia. Application of bacterial serology and basic laboratory methods.
J Infect
 
1987
;
14
:
1421
–30.
Koga M, Yuki N, Kashiwase K, Tadokoro K, Juji T, Hirata K. Guillain–Barré and Fisher's syndromes subsequent to Campylobacter jejuni enteritis are associated with HLA-B54 and Cw1 independent of anti-ganglioside antibodies.
J Neuroimmunol
 
1998
;
88
:
62
–6.
Koga M, Yuki N, Takahashi M, Saito K, Hirata K. Close association of IgA anti-ganglioside antibodies with antecedent Campylobacter jejuni infection in Guillain–Barré and Fisher's syndromes.
J Neuroimmunol
 
1998
;
81
:
138
–43.
Kurtti P, Isoaho R, von Hertzen L, Keistinen T, Kivela SL, Leinonen M. Influence of age, gender and smoking on Streptococcus pneumoniae, Haemophilus influenzae and Moraxella (Branhamella) catarrhalis antibody titres in an elderly population.
Scand J Infect Dis
 
1997
;
29
:
485
–9.
Kuwabara S, Asahina M, Koga M, Mori M, Yuki N, Hattori T. Two patterns of clinical recovery in Guillain–Barré syndrome with IgG anti-GM1 antibody.
Neurology
 
1998
;
51
:
1656
–60.
Kuwabara S, Yuki N, Koga M, Hattori T, Matsuura D, Miyake M, et al. IgG anti-GM1 antibody is associated with reversible conduction failure and axonal degeneration in Guillain–Barré syndrome.
Ann Neurol
 
1998
;
44
:
202
–8.
Laemmli UK. Cleavage of the structural proteins during the assembly of the head of bacteriophage T4.
Nature
 
1970
;
227
:
680
–5.
Loosmore SM, Yang YP, Coleman DC, Shortreed JM, England DM, Klein MH. Outer membrane protein D15 is conserved among Haemophilus influenzae species and may represent a universal protective antigen against invasive disease.
Infect Immun
 
1997
;
65
:
3702
–7.
Mandrell RE, McLaughlin R, Aba Kwaik Y, Lesse A, Yamasaki R, Gibson B, et al. Lipooligosaccharides (LOS) of some Haemophilus species mimic human glycosphingolipids, and some LOS are sialylated.
Infect Immun
 
1992
;
60
:
1322
–8.
Mishu B, Blaser MJ. Role of infection due to Campylobacter jejuni in the initiation of Guillain–Barré syndrome. [Review].
Clin Infect Dis
 
1993
;
17
:
104
–8.
Mori M, Kuwabara S, Miyake M, Dezawa M, Adachi-Usami E, Kuroki H, et al. Haemophilus influenzae has a GM1 ganglioside-like structure and elicits Guillain–Barré syndrome.
Neurology
 
1999
;
52
:
1282
–4.
Murphy TF, Apicella MA. Nontypable Haemophilus influenzae: a review of clinical aspects, surface antigens, and the human immune response to infection. [Review].
Rev Infect Dis
 
1987
;
9
:
1
–15.
Oomes PG, Jacobs BC, Hazenberg MP, Bänffer JR, van der Meché FG. Anti-GM1 IgG antibodies and Campylobacter bacteria in Guillain–Barré syndrome: evidence of molecular mimicry.
Ann Neurol
 
1995
;
38
:
170
–5.
Rees JH, Gregson NA, Hughes RA. Anti-ganglioside GM1 antibodies in Guillain–Barré syndrome and their relationship to Campylobacter jejuni infection.
Ann Neurol
 
1995
;
38
:
809
–16.
Rees JH, Soudain SE, Gregson NA, Hughes RA. Campylobacter jejuni infection and Guillain–Barré syndrome.
N Engl J Med
 
1995
;
333
:
1374
–9.
Risberg A, Schweda EK, Jansson PE. Structural studies of the cell-envelope oligosaccharide from the lipopolysaccharide of Haemophilus influenzae strain RM.118–28.
Eur J Biochem
 
1997
;
243
:
701
–7.
Ropper AH, Wijdicks EFM, Truax BT. Guillain–Barré syndrome. Philadelphia: F.A. Davis; 1991. p. 57–72.
Visser LH, van der Meché FG, van Doorn PA, Meulstee J, Jacobs BC, Oomes PG, et al. Guillain–Barré syndrome without sensory loss (acute motor neuropathy). A subgroup with specific clinical, electrodiagnostic and laboratory features. Dutch Guillain–Barré Study Group.
Brain
 
1995
;
118
:
841
–7.
Vriesendorp FJ, Mishu B, Blaser MJ, Koski CL. Serum antibodies to GM1, GD1b, peripheral nerve myelin, and Campylobacter jejuni in patients with Guillain–Barré syndrome and controls: correlation and prognosis.
Ann Neurol
 
1993
;
34
:
130
–5.
Westphal O, Jann K. Bacterial lipopolysaccharides. Extraction with phenol–water and further application of the procedure.
Methods Carbohydrate Chem
 
1965
;
5
:
83
–91.
Winer JB, Hughes RA, Anderson MJ, Jones DM, Kangro H, Watkins RP. A prospective study of acute idiopathic neuropathy. II. Antecedent events.
J Neurol Neurosurg Psychiatry
 
1988
;
51
:
613
–8.
Yuki N, Yoshino H, Sato S, Miyatake T. Acute axonal polyneuropathy associated with anti-GM1 antibodies following Campylobacter enteritis.
Neurology
 
1990
;
40
:
1900
–2.
Yuki N, Handa N, Taki T, Kasama T, Takahashi M, Saito K, et al. Cross-reactive antigen between nervous tissue and a bacterium elicits Guillain–Barré syndrome: molecular mimicry between ganglioside GM1 and lipopolysaccharide from Penner's serotype 19 of Campylobacter jejuni.
Biomed Res
 
1992
;
13
:
451
–3.
Yuki N, Yoshino H, Sato S, Shinozawa K, Miyatake T. Severe acute axonal form of Guillain–Barré syndrome associated with IgG anti-GD1a antibodies.
Muscle Nerve
 
1992
;
15
:
899
–903.
Yuki N, Taki T, Inagaki F, Kasama T, Takahashi M, Saito K, et al. A bacterium lipopolysaccharide that elicits Guillain–Barré syndrome has a GM1 ganglioside-like structure.
J Exp Med
 
1993
;
178
:
1771
–5.