For the full versions of these articles see bmj.com CLINICAL REVIEW Guillain-Barré syndrome John B Winer University Hospital Birmingham, Guillain-Barré syndrome is a peripheral neuropathy é Queen Elizabeth Hospital, that causes acute neuromuscular failure. Misdiagnosis Box 1 Differential diagnosis of Guillain-Barr syndrome Edgbaston, Birmingham B15 2TH is common and can be fatal because of the high Hypokalaemia [email protected] frequency of respiratory failure, which contributes to Polymyositis Cite this as: BMJ 2008;337:a671 the 10% mortality seen in prospective studies.1 Our Lead poisoning doi:10.1136/bmj.a671 understanding of the wide spectrum of the disease and Porphyria its pathogenesis has increased enormously in recent Transverse myelitis and neuromyelitis optica years. Several high quality randomised controlled trials have established the effectiveness of early treatment. What are the clinical features? All types of Guillain-Barré syndrome present with What is the spectrum of Guillain-Barre´ syndrome? acute neuropathy, defined as progressive onset of limb é The clinical spectrum of Guillain-Barr syndrome is weakness that reaches its worst within four weeks. — varied at least three different types have been Limb weakness is usually global14—both proximal and identified. In Europe and North America about 95% distal—unlike that of dying back axonopathy, such as of cases are acute inflammatory demyelinating poly- neuropathy associated with drug toxins or alcohol, radiculoneuropathy and the other 5% are acute axonal which is usually distal. Sensory loss is variable in acute motor disorder and acute sensory and motor axonal inflammatory demyelinating polyradiculoneuropathy. neuropathy.2 The frequency of these axonal neuropa- Typically there are sensory symptoms but few sensory thies varies throughout the world, and in Asia and signs.14 Reflexes are usually lost early in the illness, South America they make up about 30% of the although acute motor axonal neuropathy can be syndrome.3 The closely relatedMiller Fisher syndrome associated with retained reflexes or even brisk is thought to be an inflammatory neuropathy that reflexes.15 Autonomic signs such as tachycardia, affects the cranial nerves to the eye muscles in hypertension, or lack of sinus arrhythmia are particular, and it is characterised by ophthalmoplegia, common.16 The respiratory system is affected in a accompanied by areflexia and ataxia but not third of cases, but this may not be associated with clear weakness.4 Some cases of acute inflammatory demye- dyspnoea, which makes it more difficult to assess.17 It is linating polyradiculoneuropathy have features of the essential to measure vital capacity in such cases to SOURCES AND Miller Fisher syndrome, but with associated weakness. anticipate failing respiratory effort.18 A falling vital SELECTION CRITERIA The incidence of Guillain-Barré syndrome varied capacity is a more useful warning sign of incipient I prepared this review by from 1.2 per 100 000 to 1.6 per 100 000 in the most respiratory arrest than blood gases or oxygen satura- searching Cochrane recent and carefully conducted European studies.56 tion, which often remain normal until breathing stops reviews, Medline, The incidence rises with age but is bimodal in some altogether. The cranial nerves are often affected, with PubMed, and my 7 personal archive of studies, with a minor peak in young adults, and is facial weakness and bulbar palsy the most common 8 references. I downloaded slightly more common in men. Twelve cases of problems, followed by an eye movement disorder. é é and assessed all familial Guillain-Barr syndrome have been Guillain-Barr syndrome can be confusedwith diseases 910 11 references that dealt with described, but there is no strong HLA link, Guillain-Barré syndrome although one study suggested a link with a CD1 poly and its subtypes—acute morphism.12 Recurrence of pure Guillain-Barré syn- Box 2 Infections that have been linked to Guillain-Barré inflammatory drome is rare, and patients with a more chronic disease syndrome demyelinating that resembles Guillain-Barré syndrome but takes Campylobacter jejuni polyradiculoneuropathy, longer than four weeks to reach its nadir are classified Epstein Barr virus acute motor axonal as having subacute or chronic inflammatory demyeli- Cytomegalovirus neuropathy, and acute nating polyneuropathy. These patients behave differ- motor and sensory Mycoplasma ently to those with Guillain-Barré syndrome and often axonal neuropathy. Human immunodeficiency virus respond to steroids.13 BMJ | 26 JULY 2008 | VOLUME 337 227 CLINICAL REVIEW causes of acute neuropathy not caused by Guillain- Nerve cell membrane Barré syndrome (box 1). It may be helpful to measure Ganglioside GM1 Out In antiganglioside antibodies, as well as antibodies to Campylobacter jejuni. Ceramide What causes Guillain-Barre´ syndrome? Around 75% of patients have a history of preceding infection, usually of the respiratory and gastrointestinal Campylobacter 22 GM1-like Campylobacter cell wall tract. A large number of infections have been linked lipopolysaccharide Out In to the onset of the syndrome, but only a few associations have been established (box 2). How are nerves damaged? The syndrome is triggered by infection in three Core oligosaccharide Lipid A quarters of patients; a third have serological evidence Polysaccharide Peptidoglycan of C jejuni infection and a few continue to excrete C (O-chain) jejuni in faeces.23 This association with preceding infection suggested that the altered immunity in the Fig 1 | Structural similarities between ganglioside GM1 in nerve cell membranes and a syndrome may result from the infectious organism Campylobacter jejuni 27 lipopolysaccharide. Adapted, with permission, from a review by Ang sharing epitopes with an antigen in peripheral nerve tissue. It has now been established that C jejuni of the spinal cord, brainstem, or muscle (box 1). About lipopolysaccharide shares epitopes with certain gang- 20% of patients are still ambulatory at the time of liosides (fig 1). 24 The closest association between diagnosis but some will deteriorate to become bed antibodies and the neurological disease is seen with bound.19 Occasionally patients with mild disease Miller Fisher syndrome, where more than 90% of develop mild distal weakness only. patients have antibodies against the ganglioside GQ1b, 25 although only a small proportion of these How can you confirm the clinical diagnosis? patients have evidence of a preceding C jejuni infection. Nerve conduction studies are the most useful con- Thus, several different organisms may cross react with firmatory test and are abnormal in 85% of patients, peripheral nerve antigens. Evidence that these even early on in the disease.2 They should be repeated after two weeks if they are normal initially. Typically these show signs of conduction block, prolonged distal latencies, delayed F waves, and sometimes the paradox of a small median sensory action potential with retained sural responses.20 Motor conduction velocities are usually normal initially but may slow later. Guillain- Barré and Strohl documented the increase in cerebro- spinal fluid protein, which is helpful diagnostically but is not specific to Guillain-Barré syndrome.21 Finding more than 50×106 cells/l in cerebrospinal fluid casts extreme doubt on the diagnosis. Some patients produce inappropriate amounts of antidiuretic hor- mone, and it is good practice to check electrolytes. In appropriate circumstances, measuring concentrations of porphyrins or lead may help diagnose unusual UNANSWERED QUESTIONS AND FUTURE RESEARCH What is the minimum amount of intravenous immunoglobulin needed to accelerate recovery and exactly how does it work? What is the value of a second course of intravenous immunoglobulin in patients who do not respond to the first course? How useful are prophylactic antibiotics against asymptomatic Campylobacter jejuni infection? Fig 2 | Electron microscopy of a nerve biopsy specimen from a Does physiotherapy speed up recovery? patient with Guillain-Barré syndrome associated with HIV Trials are needed of novel treatments such as antiganglioside columns and complement infection showing a macrophage apparently stripping myelin inhibitors from a denuded axon. Reproduced, with permission, from the book by Hughes14 228 BMJ | 26 JULY 2008 | VOLUME 337 CLINICAL REVIEW Study Treatment ControlOdds ratio Weight Odds ratio (fixed) (95% CI) (%) (fixed) (95% CI) TIPS FOR THE NON-SPECIALIST Greenwood 1984 5/14 3/15 4.0 2.22 (0.42 to 11.83) Guillain-Barré syndrome should be considered in any McKhann 1985 64/122 47/123 47.9 1.78 (1.07 to 2.97) patient developing rapidly progressive limb weakness Osterman 1984 14/18 6/20 2.7 8.17 (1.89 to 35.38) Absent reflexes are a “red flag” for Guillain-Barré Raphaël 1987 67/109 41/111 33.7 2.72 (1.58 to 4.70) syndrome in patients with rapidly progressive Raphaël 1997 26/45 13/46 11.7 3.47 (1.45 to 8.32) weakness Patients with suspected Guillain-Barré syndrome Total (95% CI) 176/308 110/315 100.0 2.49 (1.80 to 3.44) should be referred to hospital as an emergency 0.1 0.2 0 5 10 A history of weakness preceded by respiratory or Favours Favours gastrointestinal tract infection suggests Guillain-Barré control treatment syndrome Fig 3 | Forest plot of plasma exchange compared with supportive treatment in Guillain-Barré syndrome. Proportion of patients who improved one grade at 4 weeks. Adapted, with permission, from a Cochrane review34 relatively few affected patients