Differences in membrane properties of axonal and

Differences in Membrane Properties of Axonal and Demyelinating Guillain-Barre´ Syndromes Satoshi Kuwabara, MD,1 Kazue Ogawara, MD,1 Jia-Ying Sung, MD,1 Masahiro Mori, MD,1 Kazuaki Kanai, MD,1 Takamichi Hattori, MD,1 Nobuhiro Yuki, MD,2 Cindy S.-Y. Lin, BE, MEngSc,3 David Burke, MD, DSc,3 and Hugh Bostock, PhD, FRS4

Guillain-Barre´ syndrome is classified into acute motor axonal neuropathy (AMAN) and acute inflammatory demyelinating polyneuropathy (AIDP) by electrodiagnostic and pathological criteria. In AMAN, the immune attack appears directed against the axolemma and nodes of Ranvier. Threshold tracking was used to measure indices of axonal excitability (refractoriness, supernormality, and threshold electrotonus) for median nerve axons at the wrist of patients with AMAN (n ⴝ 10) and AIDP (n ⴝ 8). Refractoriness (the increase in threshold current during the relative refractory period) was greatly increased in AMAN patients, but the abruptness of the threshold increases at short interstimulus intervals indicated conduction failure distal to the stimulation (ie, an increased refractory period of transmission). During the 4 week period from onset, the high refractoriness returned toward normal, and the amplitude of the compound muscle action potential increased, consistent with improvement in the safety margin for impulse transmission in the distal nerve. In contrast, refractoriness was normal in AIDP, even though there was marked prolongation of distal latencies. Supernormality and threshold electrotonus were normal in both groups of patients, suggesting that, at the wrist, membrane potential was normal and pathology was relatively minor. These results support the view that the predominantly distal targets of immune attack are different for AMAN and AIDP. Possible mechanisms for the reduced safety factor in AMAN are discussed. Ann Neurol 2002;52:180 –187

Guillain-Barre´ syndrome (GBS) is classified into demyelinating and axonal categories by clinical, electrophysiological, and pathological criteria.1– 4 In North America and Europe, the usual form of GBS is acute inflammatory demyelinating polyneuropathy (AIDP).5–7 In contrast, a considerable number of GBS patients have acute motor axonal neuropathy (AMAN) in China3 and Japan.8,9 Autopsy studies of AMAN patients have found extensive axonal degeneration of motor fibers,2 but most AMAN patients recover well10 or even faster than patients with AIDP.11 A likely interpretation for the quick recovery is immune-mediated reversible effects on the axolemma.8,10 In AMAN, previous electrodiagnostic studies have shown that both quick resolution of conduction block in the distal nerve terminals8 and at the common entrapment sites12 and the time course of this

recovery are different from those in AIDP patients. The mechanisms for conduction block in AMAN are unknown, but blockage of Na⫹ channels has been postulated as a possible pathophysiology in such disorders.10 –13 High titers of serum anti–GM1 antibodies are found in 10 to 42% of patients with GBS,3,9,14 –16 but whether this antibody plays a role in the pathophysiology of axonal dysfunction is a matter of controversy. Passive transfer of anti–GM1 antibodies to animal nerves has been shown to cause nerve conduction block in some studies,17 but not in others.18 Similarly, incubation of isolated nerve preparations in vitro with anti– GM1 antibodies has decreased Na⫹ currents or produced conduction block in some studies,19 –21 but not in others.22

From the 1Department of Neurology, Chiba University School of Medicine, Chiba; 2Department of Neurology, Dokkyo University School of Medicine, Tochigi, Japan; 3Prince of Wales Medical Research Institute, University of New South Wales and College of Health Sciences, University of Sydney, Australia; and 4Sobell Department of Neurophysiology, Institute of Neurology, Queen Square, London, United Kingdom.

Published online Jun 21, 2002, in Wiley InterScience (www.interscience.wiley.com). DOI: 10.1002/ana.10275

Received Mar 1, 2002, and in revised form Apr 5. Accepted for publication Apr 6.

180

© 2002 Wiley-Liss, Inc.

Address correspondence to Dr Kuwabara, Department of Neurology, Chiba University School of Medicine, 1-8-1 Inohana, Chuoku, Chiba 260-8670, Japan. E-mail: [email protected]

In the 1990s, the threshold tracking technique was developed to measure several indices of axonal excitability (such as refractoriness, supernormality, late subnormality, threshold electrotonus, and strength-duration properties), noninvasively in human subjects.23–27 These indices depend on the biophysical properties of the axonal membrane at the site of stimulation and can provide an indirect insight into Na⫹ or K⫹ channel function.23 We have used this technique in the hope that it might clarify the mechanism of conduction failure in AMAN, and the differences between AMAN and AIDP.

Subjects and Methods Subjects Eighteen consecutive GBS patients (15 men and 3 women) were studied (Table). Their condition fulfilled the clinical criteria for GBS,28 and their mean age was 42 years (range, 17–72 years). The first electrodiagnostic studies were performed within 3 weeks of the onset. Thirteen of the patients were treated with intravenous immunoglobulin infusions (n ⫽ 11) or plasmapheresis (n ⫽ 2), and pretreatment serum samples taken during the first 10 days after onset were stored. For threshold-tracking studies, control data were obtained from 37 healthy subjects with mean age of 42 years (range, 24 –72 years). Patients with chronic inflammatory demyelinating polyneuropathy (n ⫽ 15), diabetes mellitus (n ⫽ 23), and amyotrophic lateral sclerosis (n ⫽ 22) served as neurological controls. All subjects gave informed consent, and the study had the approval of the ethical committee of Chiba University School of Medicine.

Conventional Electrodiagnostic Studies Nerve conduction studies were performed using conventional procedures. Motor studies were made on the median, ulnar, tibial, and peroneal nerves. Sensory nerve conduction studies were performed to antidromic stimulation of the median nerve. Patients were classified as having AIDP or AMAN on the basis of the electrodiagnostic criteria of Ho and colleagues.3

Multiple Excitability Measures Using Threshold Tracking In the threshold-tracking studies, the current required to produce a compound muscle action potential (CMAP) that was 40% of maximum was determined with a computer program (QTRAC version 4.3 with multiple excitability protocol TRONDHM; Institute of Neurology, London) as described elsewhere.23–27 The current required to produce a specified CMAP size (40% of maximum) is referred to as the “threshold” for that CMAP size. The CMAP was recorded from the abductor pollicis brevis. For median nerve stimulation, the active electrode was placed over the nerve at the wrist, and the remote electrode was placed 10cm proximal over forearm muscle. Skin temperature near the stimulus site was maintained above 32.0°C. The stimulus-response curves were measured using test stimuli of duration 0.2 and 1.0 milliseconds. From these curves, strength-duration time constant (␶SD) was calculated using the following formula27,29:

␶ SD ⫽ 0.2共I 0.2 ⫺ I 1.0 兲/共I 1.0 ⫺ 0.2I 0.2 兲 where I0.2 and I1.0 are the threshold currents using test stimuli of 0.2- and 1.0-millisecond duration, respectively. From

Table. Clinical Profiles of Patients with Guillain-Barre´ Syndrome Patient No.

Age (yr)/Gender

Cranial Nerve Palsy

Acute motor axonal neuropathy 1 22/M No 2 17/M No 3 48/M No 4 35/M No 5 27/M No 6 32/F No 7 17/M Facial 8 26/M No 9 34/M No 10 24/M No Acute inflammatory demyelinating polyneuropathy 1 72/F Facial, bulbar 2 70/M Facial 3 56/F No 4 57/M No 5 71/M Facial 6 59/M Facial, bulbar 7 56/M No 8 36/M Facial a

Hughes Gradea

Antiganglioside IgG Antibody Against

No No No No No No No No No No

3 4 2 3 2 2 5 4 2 2

GM1b, GalNAc-GD1a GM1b, GalNAc-GD1a GM1, GalNAc-GD1a GalNAc-GD1a GM1b, GalNAc-GD1a GM1b GM1, GM1b, GalNAc-GD1a GM1, GM1b

Yes Yes Yes Yes Yes Yes Yes Yes

4 4 3 3 2 4 2 3

Sensory Loss

GM1, GM1b

At the peak. 2; able to walk 5 meters without aids; 3; able to walk 5m with aids; 4, unable to walk; 5, requiring assisted ventilation.

Kuwabara et al: Axonal/Demyelinating GBS

181

the stimulus-response curves, the currents required to produce CMAPs of 10 to 90% of the maximal CMAP were used to calculate ␶SD for CMAPs of different sizes.24 To measure the recovery of axonal excitability after a single supramaximal stimulus (ie, the “recovery cycle”), we delivered test stimuli at different intervals after the conditioning stimulus. The conditioning stimulus was supramaximal, and the test stimulus tracked the threshold for a 40% CMAP. Conditioning-test intervals were systematically changed from 200 to 2 milliseconds. In threshold electrotonus studies, membrane potential was altered using subthreshold polarizing currents which were 40% of the unconditioned threshold. Depolarizing and hyperpolarizing currents were used, each lasting 100 milliseconds, and their effects on the threshold for the test CMAP were measured before, during, and after the 100-millisecond current. For statistical analysis, differences in medians were tested with the Mann– Whitney U test, and a cross-correlation was tested with analysis of variance, using Statistica for Windows 98 software.

Antiganglioside Antibody Assays Sera from the patients were tested for the presence of IgM and IgG antibodies to GM1, GM1b, GD1a, GalNAc-GD1a, and GQ1b by enzyme-linked immunosorbent assay as described elsewhere.30,31 Serum was considered positive when the titer was a ratio of 1 to 500 or more.

Results Clinical Features and Electrodiagnosis The table shows clinical profiles of patients with GBS. Based on electrodiagnostic criteria, a diagnosis of AMAN (n ⫽ 10) or AIDP (n ⫽ 8) was made for 18 patients. Mean age was 28 years (range, 17– 48 years) for the AMAN group and 59 years (36 –72 years) for the AIDP group ( p ⬍ 0.01). Only one AMAN patient had cranial and sensory nerve involvement, whereas all AIDP patients had sensory disturbances, and five had facial palsy. Clinical disabilities evaluated by the Hughes grading scale were similar for the AMAN group (median, 3.0; range, 2.0 –5.0) and AIDP group (median, 3.0; range, 2.0 – 4.0). In median motor conduction studies, the mean distal latency was 4.3 milliseconds (range, 3.8 – 4.8 milliseconds) in AMAN patients, which was slightly longer than that of normal subjects (mean, 3.4 milliseconds; range, 3.1– 4.2 milliseconds; p ⬍ 0.05). AIDP patients had a much longer distal latency (mean, 8.7 milliseconds; range, 5.9 –16.3 milliseconds) than AMAN patients and normal subjects ( p ⬍ 0.0001). CMAP amplitudes were significantly lower in patients with AMAN or AIDP than in normal subjects ( p ⫽ 0.001). The mean amplitude of the negative peak of the distal CMAP was 3.9mV (range, 2.1–5.7mV) in AMAN patients and 2.5mV (range, 0.2–5.5mV) in AIDP patients ( p ⫽ 0.13). The mean motor conduction velocity was 53m/sec (range, 40 – 64m/sec) in AMAN

182

Annals of Neurology

Vol 52

No 2

August 2002

patients, and 40m/sec (32–50m/sec) in AIDP patients ( p ⫽ 0.001). Median sensory nerve conduction studies were normal in AMAN patients, whereas AIDP patients had absent (n ⫽ 3) or slowed (n ⫽ 7) sensory potentials. Nine of the 10 AMAN patients had IgG antibodies against ganglioside GM1, GM1b, GD1a, or GalNAc-GD1a, whereas none of the AIDP patients had any of the tested antibodies (Table). Multiple Excitability Measurements in Normal Subjects Because AMAN patients were significantly younger that AIDP patients, we looked for age-related changes in normal subjects. The 37 subjects were divided into two groups, young (aged 20 – 44 years; n ⫽ 20) and old (aged 45– 80 years; n ⫽ 17). Comparison of the findings between the two groups showed less supernormality for older subjects (mean ⫾ standard deviation, ⫺18.9 ⫾ 8.9%) than for younger subjects (⫺27.9 ⫾ 6.1%; p ⫽ 0.04), but the other indices were similar. Therefore in the analyses of supernormality, AMAN patients were compared with younger controls, and AIDP patients were compared with older controls. Other parameters were compared between the patients and all 37 normal subjects. Stimulus-response Curves and Strengthduration Properties In the stimulus-response curves, threshold currents were larger in the patients with AMAN or AIDP than in the normal subjects (Fig 1; p ⬍ 0.01 for 50% CMAP), and the spread of thresholds was greater in the patients than in controls. ␶SD for 50% CMAP was slightly greater for AIDP patients than for AMAN patients and normal subjects, but the differences were not statistically significant (see Fig 1). Recovery Cycle and Threshold Electrotonus The pattern of the recovery cycle was similar for normal controls and patients with AMAN or AIDP, with relative refractoriness less than 4 milliseconds, supernormality maximal at approximately the 7-millisecond conditioning-test interval, and late subnormality maximal at approximately 40 milliseconds (Fig 2). However, refractoriness, defined as the extent of the threshold increases during the relatively refractory period (eg, conditioning-test intervals of 2 and 2.5 milliseconds), was significantly greater in AMAN patients than in normal subjects (see Fig 2A; p ⬍ 0.0001). In 7 of the 10 AMAN patients, refractoriness at the 2-millisecond interval was higher than three standard deviations above the mean value for normal subjects (Fig 3A). When compared with patients with AIDP, chronic inflammatory demyelinating polyneuropathy, diabetes

Fig 1. Stimulus-response curves using stimuli of 1.0millisecond duration, and strength-duration time constant for the 50% compound muscle action potential (CMAP) in normal subjects (n ⫽ 37), and in patients with acute motor axonal neuropathy (AMAN; n ⫽ 10) or acute inflammatory demyelinating polyneuropathy (AIDP; n ⫽ 8). Error bars indicate standard error.

mellitus, or amyotrophic lateral sclerosis, patients with AMAN had significantly greater refractoriness at the 2-millisecond interval (Fig 4; p ⬍ 0.01). At the same stimulus sites (median nerve at wrist), refractoriness of sensory axons was similar for AMAN patients and normal subjects (see Fig 2C). The recovery cycles were almost identical for normal subjects and AIDP patients (see Figs 2B and 3). In threshold electrotonus, the threshold changes produced by subthreshold depolarizing and hyperpolarizing currents were similar for AMAN patients and normal subjects. AIDP patients tended to have the smaller slow phase of depolarizing threshold change to depolarizing current than normal subjects, but the difference was not statistically significant ( p ⫽ 0.09).

Fig 2. Recovery cycle of axonal excitability in normal subjects and (A) patients with acute motor axonal neuropathy (AMAN; n ⫽ 10) or (B) acute inflammatory demyelinating polyneuropathy (AIDP; n ⫽ 8). Data for each patient group are compared with those for age-matched normal subjects (“normal-young,” 20 – 44 years; n ⫽ 20; “normal-old,” 45– 80 years; n ⫽ 17). (C) The increase in “refractoriness” of motor axons in AMAN (A) did not involve sensory axons. Data are given as mean ⫾ standard error.


183

who underwent serial studies. Refractoriness decreased or returned to the normal range within 30 days of the onset of neurological symptoms in most of the patients. There was an inverse relationship between the change in CMAP amplitude and the change in refractoriness when the data were normalized to allow comparison across subjects ( p ⫽ 0.025). The Effects of Local Cooling of the Muscle on Recovery Cycle in a Normal Subject To study changes in the recovery cycle due to changes in distal refractoriness only, we applied local cooling to the motor point of the abductor pollicis brevis muscle with an ice pack in a normal subject. Figure 3C shows the recovery cycle curves before and after cooling. Temperature over the muscle was 33.6°C before and 29.0°C after cooling, whereas temperature at the wrist was maintained at 33.9°C. Refractoriness at the 2.0and 2.5-millisecond intervals increased abruptly during cooling, mimicking the pattern observed in AMAN patients. Stimulus-response curves, strength-duration time constant and threshold electrotonus did not change significantly with distal cooling, presumably because they reflect properties of the axonal membrane at the stimulation site. Discussion Our results show several differences in axonal excitability properties for motor axons of patients with AMAN and those with AIDP in vivo. The main findings were

Fig 4. Refractoriness at the conditioning-test interval of 2 milliseconds in normal subjects, and patients with acute motor axonal neuropathy (AMAN), acute inflammatory demyelinating polyneuropathy (AIDP), chronic inflammatory demyelinating polyneuropathy (CIDP), diabetic neuropathy (DM), and amyotrophic lateral sclerosis (ALS). Error bars indicate standard errors. (asterisk) p ⬍ 0.05, compared with the other groups. Fig 3. Superimposed recovery cycle curves of patients with (A) acute motor axonal neuropathy (AMAN; n ⫽ 10) or (B) acute inflammatory demyelinating polyneuropathy (AIDP; n ⫽ 8). Dotted lines indicate 95% confidence intervals for agematched normal subjects. (C) Data for a single normal subjects in whom recordings were made before (33.6°C) and after (29.0°C) local cooling applied to the motor point of the abductor pollicis brevis.

Serial Studies of Refractoriness and Compound Muscle Action Potentials Figure 5 shows serial changes in refractoriness at the 2-millisecond interval and the amplitude of distal CMAPs of the median nerve in seven AMAN patients

184

Annals of Neurology

Vol 52

No 2

August 2002

Fig 5. Refractoriness (conditioning-test interval, 2 milliseconds) and amplitude of compound muscle action potentials (CMAP) after median nerve stimulation at the wrist in seven patients with acute motor axonal neuropathy (AMAN), who underwent sequential studies. Open symbols on the right indicate mean (⫾SE) of data for normal subjects for refractoriness (n ⫽ 37) and CMAP amplitude (n ⫽ 101). There was an inverse relationship between the change in CMAP amplitude and the change in refractoriness (p ⫽ 0.025).

markedly greater refractoriness for AMAN patients and its rapid normalization, associated with a recovery in amplitude of CMAPs. Excitability indices did not show significant changes in the median nerve at the wrist of AIDP patients. These findings suggest that pathology is more prominent in the distal nerve segments than at the wrist in both AMAN and AIDP, and that mechanisms of conduction failure are different in the two subtypes of GBS. The changes in excitability properties in AMAN

were characterized by increased refractoriness. Greater refractoriness was not seen in patients with amyotrophic lateral sclerosis or diabetic neuropathy (see Fig 4), suggesting that it was not the result merely of axonal degeneration. Threshold tracking provides reliable data about the excitability properties at the point of stimulation, but does so only when impulse transmission between the stimulus site and the muscle is secure. An increase in refractoriness in our recordings therefore may reflect either a true increase in refractoriness at the wrist or an impaired refractory period of transmission distal to the wrist, producing transmission failure of the second of a pair of closely spaced impulses.32 The abnormal recovery cycles recorded from individual AMAN patients (see Fig 3A) were characterized by abrupt departures from the normal range at the short interstimulus interval. The curves differ in appearance from those previously recorded in conditions where refractoriness at the wrist was deliberately prolonged by ischemia, by depolarization by applied currents,25 or by cooling,26 in all of which the recovery cycles had smooth curves. We therefore interpret the increased refractoriness in AMAN patients as being caused by an impaired refractory period of transmission distal to the wrist, probably in the distal nerve terminals. This interpretation was supported by the findings shown in Figure 3C, in which a similar abrupt deviation from a normal recovery cycle was produced by local cooling at the motor point. Our recovery cycle data therefore provide evidence for a critically reduced safety factor for impulse conduction in the distal nerve terminals of AMAN patients.8,10 Because the site of conduction failure was remote from the stimulus site, our data provide no direct evidence on the biophysical basis for the reduced safety factor, but some speculations are in order. One hypothesis, as described in the introduction, is a reduction in the number of functioning Na⫹ channels. Blockade of Na⫹ channels can cause either conduction slowing or conduction block and would account for their rapid reversal.8 This is seen in human poisoning by tetrodotoxin and saxitoxin, which specifically block voltage-dependent Na⫹ channels: the conduction slowing and decreased CMAP amplitudes return to normal within days.33,34 Na⫹ channel function is altered by tissue temperature,26 and the similar patterns of change in refractoriness after local cooling at the motor point support the possibility of altered Na⫹ channel function in AMAN. In normal subjects, voluntary contraction impairs the refractory period of transmission of impulses 2 to 3 milliseconds apart, probably in the distal nerve terminals of the motor axons.35 This normal physiological limitation could become clinically relevant if pathology de-


185

creased the safety margin for impulse conduction further. An alternative explanation for the reduced safety factor in AMAN is suggested by autopsy studies,36 which have shown that the first visible signs of immune attack are the presence of macrophages overlaying the nodal gap, followed by insinuation of macrophage processes under the myelin terminal loops into the periaxonal spaces. The first step could reduce the safety factor by increasing the resistance of nodal currents, and the second by short-circuiting them. Disruption of the axo–glial junction by macrophage processes would produce a type of demyelination (“myelin detachment”)37 that can block conduction with only mild slowing, because the primary electrical consequence is a decrease in the effective nodal leak resistance. This mechanism could account for the increased refractoriness and its rapid reversal seen in our AMAN patients, whereas the subsequent invasion of the periaxonal space by macrophages could lead to irreversible degeneration.2,36 Whatever its mechanism, the reduced safety factor in the distal nerve in AMAN can account for both conduction block in some fibers and the prolonged refractory period of transmission in others. Similarly, the rapid recovery of the safety factor accounts for the parallel recovery of CMAP amplitude and reduction in refractoriness documented in Figure 5. Unexpectedly, the AIDP patients showed no significant changes in excitability of median nerve axons at the wrist, despite profound prolongation of distal latencies. Exposure of paradodal or internodal axolemma by demyelination should affect ␶SD, supernormality and threshold electrotonus.23 For example, paranodal fast K⫹ channels limit the size of supernormality: when fast K⫹ channels are exposed by demyelination, supernormality decreases significantly.23 The negative results in this study suggest that demyelination is more severe distally in the nerve terminals largely sparing the wrist segment. In conclusion, the differences in membrane properties suggest that the predominantly distal targets of the immune attack are different for AMAN and AIDP,2,38 and that the mechanism of conduction failure is different. Our data indicate that in the acute phase of AMAN the safety factor for impulse transmission is critically reduced in distal nerve segments. Because of inaccessibility of the nerve terminals to excitability testing, further studies are required to determine the mechanisms of conduction block in AMAN.

This study was supported in part by a grant for Neuroimmunological Diseases (GBS1-4, T.H. and S.K.) from the Ministry of Health and Welfare of Japan.

186

Annals of Neurology

Vol 52

No 2

August 2002

References 1. Feasby TE, Gilbert JJ, Brown WF, et al. An acute axonal form of Guillain-Barre´ polyneuropathy. Brain 1986;109:1115–1126. 2. Griffin JW, Li CY, Ho TW, et al. Guillain-Barre´ syndrome in northern China. The spectrum of neuropathological changes in clinically defined cases. Brain 1995;118:575–595. 3. Ho TW, Mishu B, Li CY, et al. Guillain-Barre´ syndrome in northern China. Relationship to Campylobacter jejuni infection and anti-glycolipid antibodies. Brain 1995;118:597– 605. 4. Choudhury D, Arora A. Axonal Guillain-Barre´ syndrome: a critical review. Acta Neurol Scand 2001;103:267–277. 5. Asbury AK, Anarson BG, Adams RD. The inflammatory lesion in idiopathic polyneuritis. Its role in pathogenesis. Medicine 1969;48:173–215. 6. Prineas JW. Pathology of the Guillain-Barre´ syndrome. Ann Neurol 1981;9:6 –19. 7. Haden RDM, Cornblath DR, Hughes RAC, et al. Electrophysiological classification of Guillain-Barre´ syndrome: clinical associations and outcome. Ann Neurol 1998;44:780 –788. 8. Kuwabara S, Yuki N, Koga M, et al. IgG anti-GM1 antibody is associated with reversible conduction failure and axonal degeneration in Guillain-Barre´ syndrome. Ann Neurol 1998;44: 202–208. 9. Ogawara K, Kuwabara S, Mori M, et al. Axonal Guillain-Barre´ syndrome: relation to anti-ganglioside antibodies and Campylobacter jejuni infection in Japan. Ann Neurol 2000;48: 624 – 631. 10. Ho TW, Lin CY, Cornblath DR, et al. Patterns of recovery in the Guillain-Barre´ syndromes. Neurology 1997;48:695– 700. 11. Kuwabara S, Asahina M, Koga M, et al. Two patterns of clinical recovery in Guillain-Barre´ syndrome with IgG anti-GM1 antibody. Neurology 1998;51:1656 –1660. 12. Kuwabara S, Mori M, Ogawara K, et al. Axonal involvement at the common entrapment sites in Guillain-Barre´ syndrome with IgG anti-GM1 antibody. Muscle Nerve 1999;22: 840 – 845. 13. Waxman SG. Sodium channel blockade by antibodies: a new mechanism of neurological disease. Ann Neurol 1995;37: 421– 423. 14. Enders U, Karch H, Toyka KV, et al. The spectrum of immune responses to Campylobacter jejuni and glycoconjugates in Guillain-Barre´ syndrome or in other neuroimmunological disorders. Ann Neurol 1993;34:136 –144. 15. Rees JH, Gregson NA, Hughes RAC. Anti-GM1 antibodies in Guillain-Barre´ syndrome and their relationship to Campylobacter jejuni infection. Ann Neurol 1995;38:809 – 816. 16. Vriesendorp FJ, Mishu B, Blaser MJ, Koski CL. Serum antibodies to GM1, GD1b, peripheral nerve myelin, and Campylobacter jejuni in patients with Guillain-Barre´ syndrome and controls: correlation and prognosis. Ann Neurol 1993;34: 130 –135. 17. Santro M, Uncini A, Corpo M, et al. Experimental conduction block induced by serum from a patient with anti-GM1 antibodies. Ann Neurol 1992;31:385–390. 18. Harvey GK, Toyka KV, Zielasek J, et al. Failure of anti-GM1 IgG or IgM to induce conduction block following intraneural transfer. Muscle Nerve 1995;18:338 –394. 19. Arasaki K, Kusunoki S, Kudo N, Kanazawa I. Acute conduction block in vitro following exposure to antiganglioside sera. Muscle Nerve 1993;16:587–593. 20. Takigawa T, Yasuda H, Kikkawa R, et al. Antibodies against GM1 ganglioside affect K⫹ and Na⫹ currents in isolated rat myelinated nerve fibers. Ann Neurol 1995;37:436 – 442.

21. Weber F, Ru¨ del R, Aulkemeyer O, Brinkmeier H. Anti-GM1 antibodies can block neuronal voltage-gated sodium channels. Muscle Nerve 2000;23:1414 –1420. 22. Hirota N, Kaji R, Bostock H, et al. The physiological effect of anti-GM1 antibodies on saltatory conduction and transmembrane currents in single motor axons. Brain 1997;120: 2159 –2169. 23. Bostock H, Cikurel K, Burke D. Threshold tracking techniques in the study of human peripheral nerve. Muscle Nerve 1998; 21:137–158. 24. Kiernan M, Burke D, Andersen K, Bostock H. Multiple measures of axonal excitability: a new approach in clinical testing. Muscle Nerve 2000;23:399 – 409. 25. Kiernan M, Bostock H. Effects of membrane polarization and ischaemia on the excitability properties of human motor axons. Brain 2000;123:2542–2551. 26. Kiernan MC, Cikurel K, Bostock H. Effects of temperature on the excitability properties of human motor axons. Brain 2001; 124:816 – 825. 27. Kuwabara S, Cappelen-Smith C, Lin C, et al. Excitability properties of median and peroneal motor axons. Muscle Nerve 2000;23:1365–1373. 28. Asbury AK, Cornblath DR. Assessment of current diagnostic criteria for Guillain-Barre´ syndrome. Ann Neurol 1990; 27(suppl):S21–S24. 29. Weiss G. Sur la possibilite´ de rendre comparables entre eux les appareils servant l’excitation e´lectrique. Arch Ital Biol 1902;35: 413– 446.

30. Yuki N, Taki T, Handa S. Antibodies to GalNAc-GD1a and GalNAc-GD1b in Guillain-Barre´ syndrome subsequent to Campylobacter jejuni enteritis. J Neuroimmunol 1996;71: 155–161. 31. Yuki N, Ang CW, Koga M, et al. Clinical features and response to treatment in Guillain-Barre´ syndrome associated with antibodies to GM1b ganglioside. Ann Neurol 2000;47:314 –321. 32. McDonald WI, Sears TA. The effects of experimental demyelination on conduction in the central nervous system. Brain 1970;93:583–598. 33. Oda K, Araki K, Totoki T, Shibasaki H. Nerve conduction study of human tetrodotoxication. Neurology 1989;39: 743–745. 34. Long RR, Sargent JC, Hammer K. Paralytic shellfish poisoning: a case report and serial electrophysiologic observations. Neurology 1990;40:1310 –1312. 35. Kuwabara S, Lin C S, Mogyoros I, et al. Voluntary contraction impairs the refractory period of transmission in healthy human axons. J Physiol (Lond) 2001;531:265–275. 36. Griffin JW, Li CY, Macko C, et al. Early nodal changes in the acute motor axonal neuropathy pattern of Guillain-Barre´ syndrome. J Neurocytol 1996;25:33–51. 37. Bostock H. The pathophysiology of demyelination. In: Herndon RM, Seil FJ, eds. Multiple sclerosis. Current status of research and treatment. New York: Demos Publications, 1994: 89 –112. 38. Hafer-Macko C, Hsieh S-T, Li CY, et al. Acute motor axonal neuropathy: an antibody-mediated attack on axolemma. Ann Neurol 1996;40:635– 644.


187