Abstract
Objective: The authors evaluated the incidence and long-term prognostic factors of Guillain-Barré syndrome (GBS) in a prospective, population-based study.
Methods: Patients with GBS diagnosed according to National Institute of Neurological and Communicative Disorders and Stroke criteria in the 2-year period 1995 to 1996 in two Italian regions were prospectively followed up for 2 years after onset of GBS.
Results: A total of 120 patients were found, corresponding to a crude annual incidence rate of 1.36/100,000 population (95% CI, 1.13 to 1.63). A total of 7 (5.8%) patients, all but one with axonal or mixed EMG pattern, died acutely within 30 days from the onset of the disease. Acute mortality was due to respiratory involvement and intensive care unit complications. In multivariate analysis, a worse 2-year outcome (Hughes score ≥2) was related to a higher Hughes grade at nadir, axonal or mixed EMG, age ≥50 years, and absence of respiratory infections preceding GBS. The persistence of disability 2 years after the acute phase was related to axonal involvement and a worse status at nadir.
Conclusions: After adjustment to US population, the incidence rates for GBS from different countries showed no significant differences. Both acute mortality and long-term disability in GBS seem to be related to an axonal involvement and a Hughes grade ≥2 at nadir.
Additional material related to this article can be found on the Neurology Web site. Go to www.neurology.org and scroll down the Table of Contents for the April 8 issue to find the title link for this article.
Guillain-Barré syndrome (GBS) is an acute inflammatory demyelinating polyneuropathy of unknown origin. Although the pathogenesis of GBS remains unclear, there are increasing indications that it is an autoimmune disease, often triggered by a preceding infection.1 It usually requires hospitalization for early detection and treatment.2,3⇓ The reported incidence rates range from 0.6 to 4.0/100,000 population.1,4⇓ Despite effective therapies, such as IV immunoglobulins (IVIg) and plasma exchange (PE), acute mortality remains relatively high and about 20% of hospitalized patients may have a long-term disability.5 Prognostic factors, both for acute mortality and for residual disability, are not yet completely understood.
The Piemonte and Valle d’Aosta Register for Guillain-Barré Syndrome (PARGBS) was established in 1995 with the aim of evaluating the incidence, the clinical characteristics, and the prognostic factors of GBS in the two Italian regions using a prospective design.
Patients and methods.
Study area.
Piemonte and Valle d’Aosta are located in northwestern Italy; total area: 28,662 Km2; population: 4,418,503 inhabitants at the time of the 1991 population census.6
Patients.
The PARGBS is based on the collaboration of all 28 departments of neurology of Piemonte and Valle d’Aosta. Since January 1, 1995, all incident cases of GBS have been prospectively recorded; all the collaborating departments signaled the new cases of GBS every month to the coordinating center using a common form. In order to achieve a complete case ascertainment, secondary sources were used. A search was also performed in the Piedmont Central Regional Archive, where all discharges from public and private hospitals of Piedmont Region are recorded. The diagnoses are codified according to the International Classification of Diseases, 9th revision. The following codes were considered: 357.0 (acute infective polyneuritis), 357.8 (other polyneuritis), 357.9 (unspecified polyneuritis). The clinical records of the patients found through the Central Regional Archive were requested from the hospitals, and they were analyzed in order to verify if they met the diagnostic criteria. We also utilized the data from the Flaccid Palsy Surveillance Network, a collaborative study aimed at determining all cases of flaccid palsy in the pediatric age group (≤14 years) implemented by the Italian Health Ministry for the surveillance of acute anterior poliomyelitis and other acute flaccid palsies. This network was based on the active collaboration of all the pediatrics departments of the two regions.7
Investigators used a standard questionnaire to collect the main demographic variables, clinical history, neurologic and laboratory findings, and details of treatment. The time lag between symptom onset and peak of neurologic deficits (nadir) and between symptom onset and diagnosis was noted. Antecedent events were carefully recorded using a detailed checklist; common and unusual events and drug exposures in the 4 weeks preceding the onset of neurologic symptoms were included. EMG examination was performed according to standard procedures, which were indicated in detail in the questionnaire. Specific criteria for demyelination8 and primary axonopathy9 were used. Mixed neuropathy was diagnosed in patients with unexcitable or markedly reduced compound muscle action potentials (CMAP) in two or more nerves associated with the presence of at least two electrophysiologic findings suggesting demyelination. One or more lumbar punctures were performed in each case, and CSF was examined for cell count, protein, and glucose. For the analysis of prognostic factors, the first CSF and EMG examinations were considered. When possible, viral serology and anti-GM1 antibodies were assessed. Autonomic symptoms were noted, according to a checklist (hypertension, hypotension, orthostatic hypotension, cardiac arrhythmias, bladder and gastrointestinal dysfunction). As measures of autonomic involvement, continuous EKG recording (Holter) and the R-R interval duration were also employed. Respiratory impairment was evaluated by the presence of resting dyspnea, alteration of arterial blood gas levels, and the need to transfer the patient to the intensive care unit (ICU). The type of treatment (including dose and duration) was recorded. Each center was free to decide about patient treatment.
Diagnostic criteria.
The clinical records of each case were reviewed. The diagnosis was based on the National Institute of Neurological and Communicative Disorders and Stroke (NINCDS) diagnostic criteria.10 Exclusion criteria included the following: marked, persistent asymmetry of the neurologic signs; >5 mononuclear leukocytes in the spinal fluid; conditions such as diabetic or alcoholic neuropathy; neuropathies associated with exogenous toxic agents, metals, or drugs (specifically indicated in a checklist); poliomyelitis; or porphyria. Cases were classified in two groups: typical GBS (cases who met NINCDS criteria) and Miller-Fisher syndrome (MFS) (patients presenting with the triad of ataxia, ophthalmoplegia, and areflexia in the absence of relevant limb weakness).11 GBS variants (acute monophasic peripheral nervous system diseases with albumino-cytologic dissociation not fulfilling criteria for GBS or MFS) or other non-GBS polyneuropathies were excluded. Diagnoses made by local physicians were verified blindly by three members of the scientific committee, using data from the patients’ clinical records and the standard questionnaire filled out by the local physicians.
Follow-up.
Follow-up visits were performed at regular intervals (every day during hospitalization, at day 30, at discharge, and at 6 months, 1 year, and 2 years). Hughes scale12 was used to evaluate the clinical status of patients (grade 0 = healthy; grade 1 = minor signs or symptoms of neuropathy but capable of manual work; grade 2 = able to walk without support of a stick but incapable of manual work; grade 3 = able to walk with a stick, appliance, or support; grade 4 = confined to bed or chairbound; grade 5 = requiring assisted ventilation; grade 6 = dead).
Statistical analysis.
Incidence rates were adjusted to the 1990 US population13 and to the 1991 Italian population6 with the direct method of standardization. Ninety-five percent CI were calculated assuming a Poisson distribution.14 Standardized incidence ratios (SIR) were calculated for the nine provinces of the area, using the total population of the two regions as reference. Factors related to long-term prognosis were analyzed using a forward stepwise multivariate logistic regression, considering as outcome measure Hughes grade. Data were processed using SAS statistical package (version 6.12).15
Results.
Incidence rate.
During the period January 1, 1995, through December 31, 1996, a total of 120 patients (74 men, 46 women) were diagnosed with GBS. A total of 113 (94.2%) patients had a typical GBS and 7 (5.8%) a MFS. The mean annual crude incidence rate was 1.36/100,000 population (95% CI, 1.13 to 1.63): 1.78 (95% CI, 1.40 to 2.24) for men and 1.11 (95% CI, 0.81 to 1.48) for women. The mean annual incidence rate was 1.06 (95% CI, 0.78 to 1.41) in 1995 and 1.65 (95% CI, 1.29 to 2.08) in 1996. The mean annual incidence rate adjusted on the 1990 US population was 1.26 (95% CI, 1.05 to 1.51). The mean annual incidence rate adjusted on the 1991 Italian population was 1.32 (95% CI, 1.17 to 1.58). The annual incidence rates for the nine provinces of Piemonte and Valle d’Aosta ranged between 0.86 and 2.40/100,000 population, but the SIR showed no significant differences.
Clinical characteristics.
The mean age of the patients was 51.2 years (SD 21.5) (range 3 months to 86 years), higher among women (women, 57.7 ± 20.6; men, 47.1 ± 21.1; p = 0.009). The age-specific incidence rates showed a peak in the 60- to 69-year age range in men and in the 70- to 79-year age range in women (figure). The mean time from onset of symptoms to nadir was 9.7 (±6.9) days (median time, 7 days), with no significant difference between men (9.5 ± 6.6) and women (10.0 ± 7.5).
Figure. Age- and sex-specific incidence rates of Guillain-Barré syndrome in Piemonte and Valle d’Aosta, Italy, 1995–1996. Full line, men; dotted line, women.
Seventy patients (58.3%) reported an infective/clinical event in the 4 weeks preceding the onset of acute paralysis; the most frequent previous events were upper respiratory infections and influenza (40.8%) and gastroenteritis (12.8%). In 4 cases (3.3%), a malignant tumor was diagnosed in the month preceding or following the onset of GBS; all these cases met the NINCDS criteria for the diagnosis of GBS. In 3 cases (2.5%), all presenting in December 1995, symptom onset was preceded by influenza immunization.
The electrophysiologic studies showed a pattern of demyelination in 62 patients (51.6%), axonal degeneration in 29 (24.2%), and a mixed pattern with both demyelination and axonal degeneration in 29 (24.2%). No correlation was found between previous gastrointestinal infections and axonal or demyelinating EMG pattern (Fisher exact test).
CSF examination was performed in all cases. Most patients underwent only one lumbar puncture. Elevated CSF protein levels (greater than 45 mg/dL) were noted in 100 cases (83.3%). The mean level was 105.5 mg/dL (SD 68.5).
Autonomic dysfunction was recorded in 27 (22.5%) cases and predominantly affected the cardiovascular system. Sphincter muscles were involved in 16 cases (13.3%). Respiratory symptoms were present in 24 cases (20%), 18 (15%) of whom had to be admitted to ICU.
Clinical status, according to Hughes scale, at nadir, at 30 days after onset, and at 2 years after onset is reported in table 1. In univariate analysis, a worse Hughes score at nadir was related to the presence of axonal or mixed nerve damage (p = 0.004), autonomic nervous system involvement (p = 0.009), and age ≥50 years (p = 0.01). Using forward stepwise multivariate logistic regression with Hughes score as outcome measure, the same variables remained significant (table 2).
Table 1 Clinical status (Hughes’ scale) at nadir, at 30 days, and at 2 years after GBS onset*
Table 2 Factors significantly related to a worse clinical status (Hughes’ scale ≥3) at nadir in multivariate analysis
Fatal cases.
A total of 7 (5.8%) patients died acutely within 30 days from the onset of the disease. Comparing deceased and surviving patients, patients with fatal cases were slightly older (mean age, 58 years [SD 21.0] for surviving cases vs 50 years [SD 18.4] for fatal cases; p = NS) and reported a similar frequency of previous infective events (influenza, upper respiratory infection, or gastrointestinal infections) (57.1% vs 63.7%; p = NS). Fatal cases more frequently had sphincter abnormalities (57% vs 10.6%; p = 0.016), whereas the frequency of autonomic symptoms was similar in fatal and nonfatal cases (28.6% vs 18.6%; p = NS). All but one of the fatal cases (85.7%) had an axonal or mixed EMG pattern vs 46% of nonfatal cases (p = NS). No fatal case had a MFS. Six fatal cases (85.7%) vs nine nonfatal cases (8.0%) (p < 0.0001) were admitted to ICU due to respiratory failure: in all of them the direct cause of death was a ventilator-associated pneumonia. The last patient had a sudden death that was attributed to a pulmonary embolus because she had a deep venous thrombosis. She was not autopsied.
Long-term outcome.
The outcome analysis was performed on 108 patients, excluding seven patients deceased within 30 days from onset and five patients lost to follow-up. At the end of the 2-year period, 86 patients (79.6%) had a good recovery (Hughes grade < 2). In univariate analysis the variables significantly related to a poor recovery (Hughes’ grade ≥2) were Hughes grade at nadir ≥3, respiratory impairment, age ≥50 years, axonal or mixed EMG, no infective antecedents, and autonomic involvement. Two different multivariate models were evaluated (table 3). The first model, which included Hughes grade at nadir, retained only Hughes grade and age. The second model, which did not include Hughes grade at nadir, retained 3 variables: axonal or mixed EMG, age ≥50 years, and absence of respiratory infections preceding the onset of GBS.
Table 3 Variables significantly related to a worse clinical status (Hughes’ scale ≥2) at 2 years after onset in multivariate analysis
No case of GBS recurred during the 2 years of prospective follow-up, nor in the following 2 years.
Treatments and prognosis.
The most frequently employed treatment was PE (43%), followed by IVIg (22%) and steroids (14%). A total of 14 (11.7%) patients received no treatment. No difference in the 30-day and 2-year outcome was found when comparing main treatments (PE vs IVIg), either including or not Hughes grade at nadir as a covariate.
Discussion.
Despite numerous studies, there is still some uncertainty concerning GBS outcome. This could be due to the fact that most studies are based on small populations that are analyzed retrospectively or focus on patients admitted to clinical trials, which usually include more severely affected individuals. In both cases, the results of the studies may have relevant biases related either to the selection of patients or to the difficulty in correctly determining retrospectively the risk factors and the progression of the disease. Moreover, one reason explaining the discrepancies in the epidemiologic results is the difference in diagnostic criteria used for defining GBS, a problem that can be overcome by considering only the studies based on NINCDS criteria. The difference in the population structure can be eliminated with standardization to a common reference population.
The annual incidence rate of GBS observed in Piemonte and Valle d’Aosta in the period 1995 to 1996 is in the range of the values reported in the other studies based on NINCDS diagnostic criteria (see supplementary table e1 at www.neurology.org). This finding supports the idea that the disease is evenly distributed in the world, independently from the ethnic, socioeconomic, and environmental characteristics of the studied areas. A possible underascertainment of less severe cases may be considered in our as in most other studies.19 However, in our series 7 patients (5.8%) with a mild symptomatology were not admitted to hospitals and were ascertained only on an outpatient basis. The examined period is too short to determine the presence of a seasonal trend, although an increased incidence was observed during winter and spring. The frequency of Fisher variant is similar to that reported in previous studies.3,16-20⇓⇓⇓⇓⇓
A higher incidence of GBS in males has been reported in most studies, with a male to female ratio ranging from 1.1:1 to 2:1. This finding is similar to what has been observed in two other autoimmune disorders of the peripheral nervous system: multifocal motor neuropathy with conduction blocks21 and chronic inflammatory demyelinating neuropathy.22,23⇓ The incidence rates in our series increased with age, with a peak in the 60- to 69-year age group. We did not find a bimodal shape of age-specific curve observed in other studies.2,24,25⇓⇓ However, the decrease of incidence in the oldest age classes could be due to an underascertainment of cases in very old patients.19 The higher incidence in the elderly has been related to the supposed increased susceptibility to autoimmune disorders in old age.1
Infective antecedent events were found in 58% of patients, a value in the range of studies performed using the same diagnostic criteria (see supplementary table e2 at www.neurology.org). However, in our series the frequency of gastrointestinal events was lower than in other surveys.
The percentage of cases with axonal impairment in our series (about 25%), according to electrophysiologic data, was slightly higher than that reported in other epidemiologic and clinical series.17,26⇓ This could be related to the fact that in our prospective study an electrophysiologic protocol for defining the type of nerve involvement was established, allowing a more uniform evaluation of patients. Conversely, retrospective studies are usually based on standard electrophysiologic examinations. Moreover, it has been shown recently that the criteria used for defining demyelinating and axonal damage greatly influence the relative frequency of the various forms in GBS.27 We did not observe any significant correlation between axonal damage and previous gastrointestinal infections that have been related to Campylobacter jejuni infection.1
Early mortality rate due to GBS has a remarkable variation across studies (range, 3.2 to 8%). This observation may be related to differences in ability to obtain emergency interventions; further studies specifically directed toward the determination of the factors related to early mortality are necessary in order to plan more effective therapeutic interventions. In our survey, death generally occurred in older patients with an axonal or mixed EMG pattern, and in all but one case it was related to ventilator-associated pneumonia. Confirming a previous observation,28 no death was related to autonomic disturbances or therapy complications (PE or IVIg).
Long-term prognostic factors.
According to previous reports, the overall recovery (i.e., Hughes grade <2) of patients with GBS ranges from 54 to 64% at 6 months,17,24⇓ 48 to 90% at 1 year,5,24,29,30⇓⇓⇓ and 60 to 88% at 2 years.5,29⇓ The remarkably high ranges reported in literature are likely to be due to the different design of the studies: in general, studies from single specialized centers and series based on pharmacologic trials show a worse outcome, likely due to the inclusion of patients with more serious disease. In our survey, 80% of patients had a good recovery at 2 years. Several factors related to a worse recovery have been described, including severity at admission,5 axonal EMG,5 reduction of CMAP amplitude,17,26,30,31⇓⇓⇓ severity at nadir,5,17,29,32⇓⇓⇓ latency to nadir,5 age over 40 or 50,5,16,17,26,30,33,34⇓⇓⇓⇓⇓⇓ longer duration of the plateau phase,29,32⇓ and previous gastroenteritis.5,35⇓ We found that Hughes grade at nadir ≥3 and age ≥50 years were significantly related to a worse outcome at 2 years. However, in the multivariate analysis not including Hughes grade at nadir, axonal or mixed EMG and previous respiratory infections, other then age ≥50, were also selected by the model.
Appendix
Piemonte and Valle d’Aosta Register for GBS (PARGBS): Coordinating center: 2nd Division of Neurology, Department of Neuroscience, University of Torino, Italy. Project coordinator: A. Chiò, MD. Study monitors: A. Calvo, MD, N. Di Vito, MD, M. Vercellino, MD. Scientific Committee: A. Bertolotto, MD, E. Bottacchi, MD, A. Chiò, MD, D. Cocito, MD, M.T. Giordana, MD, M. Leone, MD, L. Mazzini, MD, G. Mora, MD. Collaborating centers: A. Chiò, MD, A.A. Terreni, MD, D. Schiffer, MD, R. Mutani, MD, D. Cocito, MD, B. Bergamasco, MD, I. Rainero, MD (Department of Neuroscience, Section of Neurology, University of Torino, and Azienda Ospedaliera San Giovanni Battista, Torino); A. Bertolotto, MD, A. Tribolo, MD, R. Sciolla, MD, F. Mondino, MD, M.T. Giordana, MD (Department of Neuroscience, Section of Neurology, University of Torino, and Azienda Ospedaliera San Luigi Gonzaga, Orbassano); M. Leone, MD, P. Gaviani, MD, F. Monaco, MD (Department of Neurology, “Amedeo Avogadro” University, Novara); M. De Mattei, MD, E. Morgando, MD (Department of Neurology, Azienda Ospedaliera San Giovanni Battista, Torino); L. Sosso, MD, M. Gionco, MD (Department of Neurology, Ospedale Mauriziano, Torino); U. Morino, MD, M. Nobili, MD (Department of Neurology, Ospedale Martini, Torino); L. Appendino, MD (Department of Neurology, Ospedale Maria Vittoria, Torino); L. Mazzini, MD, D. Piazza, MD (Department of Neurology, Ospedale S. Giovanni Bosco, Torino); E. Oddenino, MD, W. Liboni, MD (Department of Neurology, Ospedale Gradenigo, Torino); G. Vaula, MD, G. Ferrari, MD (Department of Neurology, Ivrea); M. Favero, MD, C. Doriguzzi Bozzo, MD (Department of Neurology, Pinerolo); P. Santamaria, MD (Department of Neurology, Vercelli); U. Massazza, MD, E. Bollani, MD (Department of Neurology, Biella); A. Villani, MD, R. Conti, MD (Department of Neurology, Domodossola); G. Mora, MD, C. Balzarini, MD (Department of Neurological Rehabilitation, Fondazione S. Maugeri, Clinica del Lavoro e della Riabilitazione, IRCCS, Scientific Institute of Veruno); M. Palermo, MD (Department of Neurology, Alessandria); F. Vergnano, MD (Department of Neurology, Casale Monferrato); S. Cordera, MD, C. Buffa, MD (Department of Neurology, Novi Ligure); M.T. Penza, MD (Department of Neurology, Tortona); F. Fassio, MD (Department of Neurology, Asti); P. Meineri, MD (Department of Neurology, Azienda Ospedaliera Santa Croce e Carle, Cuneo); A. Cognazzo, MD, C. Mocellini, MD, A. Dutto, MD, A. Cucatto, MD (Department of Neurology, Savigliano); C. Cavestro, MD, W. Troni, MD (Department of Neurology, Alba); G. Corso, MD, E. Bottacchi, MD (Department of Neurology, Aosta).
Acknowledgments
Acknowledgment
The authors thank Prof. A. Moiraghi Ruggenini, MD, for data from the Piemonte and Valle d’Aosta network used for surveillance of pediatric acute flaccid palsies.
- Received September 8, 2002.
- Accepted December 18, 2002.
References
Letters: Rapid online correspondence
REQUIREMENTS
If you are uploading a letter concerning an article:
You must have updated your disclosures within six months: http://submit.neurology.org
Your co-authors must send a completed Publishing Agreement Form to Neurology Staff (not necessary for the lead/corresponding author as the form below will suffice) before you upload your comment.
If you are responding to a comment that was written about an article you originally authored:
You (and co-authors) do not need to fill out forms or check disclosures as author forms are still valid
and apply to letter.
Submission specifications:
- Submissions must be < 200 words with < 5 references. Reference 1 must be the article on which you are commenting.
- Submissions should not have more than 5 authors. (Exception: original author replies can include all original authors of the article)
- Submit only on articles published within 6 months of issue date.
- Do not be redundant. Read any comments already posted on the article prior to submission.
- Submitted comments are subject to editing and editor review prior to posting.
