Two patterns of clinical recovery in Guillain-Barré syndrome with IgG anti-GM1 antibody

Guillain-Barré syndrome (GBS) consists of several subtypes based on electrophysiologic and pathologic findings.^1-3 Acute motor axonal neuropathy (AMAN), a pure motor axonal variant of GBS, has been recognized in northern China² and occurs in other countries.^4-5 Between 20 and 30% of GBS patients have high-titer anti-ganglioside GM1 antibody,^6-9 and the AMAN pattern of GBS in particular is associated with this antibody.^5,10 Whether anti-GM1 antibody is a marker for poor prognosis due to axonal degeneration is still a matter of controversy. Some studies suggest that anti-GM1 antibody is related to extensive axonal loss and poor outcome,^4,7,11 whereas others show no correlation between anti-GM1 antibodies and the severity, type (axonal versus demyelinating), or outcome of GBS.^8,9,12

Autopsy studies of Chinese AMAN patients with anti-GM1 antibodies found Wallerian-like degeneration in a large percentage of the motor fibers in the ventral roots,³ but most AMAN patients are reported to recover well,¹³ indicating that extensive axonal degeneration is not always the pathophysiologic base of this disorder. We investigated whether anti-GM1 antibody has prognostic value by following consecutive GBS patients admitted to a university hospital in Japan and by studying the correlation between clinical recovery and anti-GM1 antibody.

Methods. Patients. A total of 41 GBS patients seen at Chiba University Hospital between 1992 and 1997 were studied prospectively. They fulfilled the clinical criteria for GBS,¹⁴ and their disabilities were evaluated on the Hughes functional grading scale (grade 5, requires assisted respiration; grade 4, bed bound; grade 3, able to walk 5 m with aids; grade 2, ambulates independently; grade 1, minimal signs and symptoms, able to run). Patients were followed for 6 months from the onset of neurologic symptoms, every week for the first 6 weeks, and then at 3 and 6 months. Most of the patients received plasmapheresis (n = 20, 50 mL/kg plasma three to eight times; mean, 5.2 times) or IV immunoglobulin (Ig) treatment (n = 12, 400 mg/kg/day for 5 days).

Anti-ganglioside antibody assays. Patients' sera obtained during the first 4 weeks were stored at -80 °C until used. IgG and IgM class antibodies to ganglioside GM1 were measured using ELISA, as described elsewhere,¹⁵ by one of the authors (M.K.) who was blind to the clinical and electrophysiologic data. Antibody titer (1: x) was the highest serum dilution at which the optical density at 492 nm was 0.1 or greater. Serum was considered positive when the titer was 1:500 or more.

Electrophysiology. Nerve conduction studies were performed using conventional procedures, and the first studies were made within 3 weeks of onset by one of the investigators (S.K.). The classification of AMAN or acute inflammatory demyelinating polyneuropathy (AIDP) was based on the electrodiagnostic criteria of Ho et al.¹⁰ When the AMAN pattern and an amplitude reduction < 50% of the normal limits of the sensory nerve action potentials were present, patients were classified as having acute motor and sensory axonal neuropathy (AMSAN).⁹

Statistical analysis. Differences in percentages were tested with the chi-square or Fisher's exact test, and differences in medians with the Mann-Whitney U test using Statistica for Windows 4.3 software (StatSoft; Tulsa, OK) by one of the investigators (M.A.). The main outcome measures were improvement by ≥2 grades within the first 4 weeks, and whether the patient was able to walk independently at 6 months. p Values were two tailed. Kaplan-Meier curves were used to analyze the time taken to reach Hughes grade 2 (independent locomotion).

Results. Of the 41 GBS patients, 21 had anti-GM1 antibodies, 17 had only IgG antibodies, two had only IgM, and two had both classes. Data were compared between the 19 patients with and 22 patients without IgG anti-GM1 antibodies because IgG class antibodies were strongly associated with electrodiagnosis but IgM was not. The clinical features of the patients are summarized in table 1. IgG anti-GM1-positive patients had significantly frequent occurrence of preceding diarrhea and less frequent cranial nerve palsy sensory loss.

Clinical recovery in IgG anti-GM1-positive and -negative patients. The milestones of recovery in IgG anti-GM1-positive and -negative patient groups are shown in table 2. The mean Hughes grade scores in the two patient groups did not differ significantly at peak, and 1, 3, and 6 months after the onset of neurologic symptoms. However, the anti-GM1-positive group had significantly higher percentages of patients with poor outcome (unable to walk independently at 6 months) and those with rapid recovery (improvement by two Hughes grades or more during the first 4 weeks), although both groups of patients received similar treatments. We also analyzed the time until improvement by one functional grade to evaluate the speed of recovery (figure 1). The histogram in figure 1 shows the time periods for the two subgroups of anti-GM1-positive patients. Nine patients showed improvement by one grade within 14 days, whereas for the others improvement was delayed over months. In contrast, the time required for one grade of improvement for the anti-GM1-negative patients showed a distribution between those of the two subgroups of anti-GM1-positive patients and was relatively uniform. Consequently, there were dissimilar Kaplan-Meier curves of the probabilities of reaching Hughes grade 2 for IgG anti-GM1-positive and -negative groups (figure 2). Figure 3 shows the recovery of each patient and also indicates slower or faster recovery in IgG anti-GM1-positive patients, whereas IgG anti-GM1-negative patients had a uniform recovery speed.

Electrophysiologic subtypes. The 41 patients were classified as having AMAN (14 patients), AMSAN (1 patient), or AIDP (20 patients) or were unclassified (six patients). Among the unclassified patients, 5 had absent F waves as an isolated abnormality, and the remaining patient had mild prolongation of distal latencies alone. None had inexcitable nerves. There were strong associations between the positivity of the IgG anti-GM1 antibodies and the AMAN/AMSAN pattern and between negativity and the AIDP pattern (table 3). IgM class antibodies had no correlation with the electrodiagnosis, and 2 patients with IgM antibodies alone had the AIDP pattern.

Prognostic factors among the anti-GM1-positive patients. Because the anti-GM1-positive group showed two different recovery periods, the prognostic factors were studied further. Between the patients who improved by two or more Hughes grades within a month (n = 9) and those who did not (n = 10), the factors that did not differ significantly were age, sex, preceding diarrhea, maximum Hughes grade, time from onset of neurologic symptoms to initiation of treatment, IgG anti-GM1 antibody titer, electrophysiologic subtype, and distal amplitude of the compound muscle action potential of the abductor pollicis brevis in the initial study. However, the number of patients who received IVIg therapy was significantly higher in the rapid recovery subgroup (5 of 9 versus 1 of 10, p = 0.05). Table 4 shows the association between clinical recovery and treatments. In the anti-GM1-positive group, rapid recovery (an improvement of two or more Hughes grades within 4 weeks) appeared to be associated with IVIg treatment, whereas there was no difference in clinical recovery between patients treated with IVIg and plasmapheresis in the anti-GM1-negative group.

Discussion. Our findings show that anti-GM1 antibody is not always a marker of poor outcome. Although the anti-GM1-positive group included a significantly higher number of patients for whom the prognosis was poor, probably because of extensive axonal degeneration, the mean disabilities throughout the course of the disease were similar to those for the patients without anti-GM1 antibodies when compared as patient groups. This is because the rest of the anti-GM1-positive patients recovered more quickly than the anti-GM1-negative patients, most of whom have AIDP. Therefore, there are two patterns of clinical recovery in the anti-GM1-positive group of patients: slower and faster than the recovery for the anti-GM1-negative patients.

The subgroup of patients with rapid recovery was indicated from findings for AMAN patients whose sera often had anti-GM1 antibodies. Ho et al.¹³ reported that the recovery times of AMAN and AIDP patients were similar and that most AMAN patients improved quickly. They also reported that a few AMAN patients experienced prolonged recoveries. AMAN in northern China occurs seasonally in rural areas and mainly affects children, but our findings show similar rapid improvement in the AMAN pattern of GBS that occurs in adults. The time course of recovery for Chinese AMAN or AIDP patients¹³ probably reflects the natural course of the disease because none received plasma exchange or IVIg treatment. Conversely, most of our patients were treated with plasmapheresis or IVIg. Although similar numbers of patients were treated with these immunotherapies in the anti-GM1-positive and anti-GM1-negative groups, there was marked rapid improvement only in the former group. We therefore speculate that conduction failure in anti-GM1-associated GBS potentially is early reversible and may respond better than conduction failure in AIDP after adequate treatment. Furthermore, because the anti-GM1-positive patients who recovered quickly had more often been treated with IVIg, the therapy may be more effective than plasmapheresis in treating anti-GM1-positive GBS patients. This speculation is consistent with the results of Visser et al.,⁵ who showed that the responses to IVIg treatment of "motor" GBS patients who often had anti-GM1 antibodies were better than those to plasma exchange. Whether the recovery of anti-GM1-positive GBS patients is markedly affected by the treatment modality may be of clinical relevance.

The rapid recovery of AMAN patients may be explained by quickly reversible immune-mediated changes at the nodes of Ranvier in the motor fibers, or by degeneration and regeneration of intramuscular motor nerves, as suggested previously.¹³ Demyelination is considered less likely because of the absence of demyelinative conduction slowing.^2,10,13 The different recovery patterns for the anti-GM1-positive and -negative patients also support the unlikeliness of demyelination as the pathophysiologic base of anti-GM1-positive GBS or AMAN. We think that the marked rapid improvement seen within days in some of our patients (see figure 2) could have been caused by the early resolution of physiologic conduction failure at the nodes of Ranvier. The time courses could hardly be explained by regeneration or remyelination. This speculation is supported by several lines of evidence. Anti-GM1 antibodies are reported to affect the sodium and potassium currents,¹⁶ and autopsy studies of some AMAN patients with severe paralysis show minimal pathologic change in the motor nerves³ but exhibit immunoglobulin and complement deposition at the nodes of Ranvier.¹⁷ The possibility of distal degeneration and regeneration cannot, however, be excluded as a recovery mechanism that occurs somewhat later.

Although some studies suggest that Campylobacter jejuni infection rather than anti-GM1 antibody is a marker of poor prognosis,^9,12 the faster or slower recovery was not associated with preceding diarrhea in this study. We did not assay for C. jejuni and could not evaluate its prognostic value.

The strong association between the positivity of IgG anti-GM1 antibody and the AMAN/AMSAN pattern suggests that this antibody is associated closely with the pathogenesis of axonal GBS, whereas the IgM class antibody appears to have no association. Extensive axonal degeneration is not, however, a regular feature. Moreover, some IgG anti-GM1-positive patients recover more quickly than IgG anti-GM1-negative AIDP patients. Early reversible effects, other than demyelination, may be responsible for this pattern of recovery. Anti-GM1 antibody assays are useful in detecting the subgroup of patients with recovery patterns different from the pattern in classic demyelinating GBS patients. Additional studies are needed to clarify the pathogenesis of anti-GM1 antibody and the mechanism of rapid recovery, which may have implications for the choice of therapy in GBS.