Clinical Findings

In 50–70% of cases, GBS develops 2–4 weeks after a prodromal gastroenteritic or respiratory illness, such as those caused by Epstein-Barr virus, cytomegalovirus, Mycoplasma pneumoniae , hepatitis, varicella, and other herpesviruses. Campylobacter jejuni is the most common infectious agent associated with axonal forms of GBS while cytomegalovirus infection causes a form of GBS with prominent sensory symptoms and involvement of the cranial nerves. The incidence of GBS after C. jejuni infection, however, is only 0.25–0.65/1000 cases. Occasional cases follow immunization or surgery, or develop in association with other illnesses, but childhood GBS has not been proven to result from vaccination against poliomyelitis, tetanus, or measles. Uncommonly, GBS may be the presenting manifestation of primary central nervous system or systemic lymphoma.

GBS can occur at any age. Occasional congenital and neonatal cases have been described, with vertical transmission between an affected mother and fetus also having been reported on several occasions. As a treatable cause of neonatal hypotonia, GBS should therefore be included in the differential diagnosis of the “floppy infant.” Approximately one-third of childhood cases occur before the age of 3 years.

Infants with GBS present with hypotonia and decreased movement, and sometimes with respiratory insufficiency. Irritability may be prominent. Diagnosis of infantile GBS is often delayed.

Older children usually complain of weakness and fatigue, paresthesias and pain, and have difficulty walking, rising from the floor, and climbing stairs. Weakness is generally symmetrical, starting in the lower extremities and then ascending into the upper extremities over days to weeks, and remaining distally predominant throughout the course of the illness. In a significant proportion (15–20%) of cases, however, weakness is primarily proximal. Cranial nerve involvement is more common in pediatric than adult GBS. Facial weakness and ophthalmoplegia are seen in up to 45% of affected children. Internal ophthalmoplegia is rare in GBS and should prompt consideration of tick paralysis, diphtheria, and botulism.

Neck, back, buttock, or leg pain, which is presumed to result from nerve root and peripheral nerve inflammation, is the initial manifestation in as many as 50% of children with GBS. This pain is often poorly localized and may cause marked irritability. Recognition of this symptom is especially important in young children who may be unable to describe weakness, and in whom pain and refusal to weight-bear is often initially thought to result from orthopedic or rheumatologic causes. The severity of pain tends to broadly correspond to the degree of weakness.

As many as 50% of children with GBS have ataxia resulting from weakness and sensory loss (“deafferentation”) rather than from cerebellar involvement. Rare cases of pediatric GBS present as the Miller Fisher variant of ataxia, ophthalmoplegia, and areflexia without peripheral weakness.

Respiratory muscle weakness in GBS is usually slowly progressive and tends to correlate with the degree of limb muscle weakness. Very rare cases of pediatric GBS present with acute-onset respiratory failure, which is life-threatening and may be associated with complete flaccid paralysis. Such severe and acute cases should prompt consideration of an underlying genetic neuropathy.

Meningism, headache, drowsiness, and irritability are seen occasionally in children presenting with a “pseudo-encephalopathic” form of GBS that mimics acute meningitis or meningoencephalitis. There are several reports of locked-in syndrome in children with severe GBS presenting with quadriparesis, areflexia, respiratory failure, and cranial nerve dysfunction. Neurophysiologic studies show inexcitable peripheral nerves. The prognosis of such cases is poor.

By definition GBS has a rapid course, reaching a maximal deficit within 4 weeks, although as many as 80% of children reach their clinical nadir within 2 weeks of onset. At their nadir 60% of children cannot walk, and 24% lose functional movement of their arms.

Physical examination reveals weakness, which ascends from the distal lower extremities over days–weeks, and loss of the muscle stretch reflexes. These changes may be quite patchy and can be asymmetric, especially at onset. Weakness usually begins in the legs and progresses rostrally, but can be proximally predominant at onset. The deep tendon reflexes are usually lost within the first week of the illness, but are occasionally preserved throughout its course. Rare cases of hyperreflexia in GBS are usually associated with the acute motor axonal variant of this condition.

Sensory loss is rarely prominent in childhood GBS, but can be identified on detailed examination in about 40% of cases. Identification of a sensory level is not compatible with GBS and should prompt immediate consideration of spinal cord pathology.

Involvement of the autonomic nervous system is seen in 25–50% of children with GBS, usually manifesting as blood pressure instability, sinus tachycardia, pupillary abnormalities, and abnormalities of sweating. Autonomic instability and urinary incontinence are more common in atypical or focal forms of pediatric GBS. Recent reports of hypertensive encephalopathy and posterior reversible encephalopathy as initial manifestations of GBS highlight the importance of considering this condition as part of the differential diagnosis of any acute dysautonomia.

Sphincter dysfunction (urinary retention, urinary and fecal incontinence) is occasionally seen in children with GBS but is far more common in transverse myelitis, tumors, and other spinal cord lesions. Although urinary retention occurs in 10–15% of children early in the course of GBS, persistent bladder or bowel involvement are indications for spinal imaging.

Clinical Variants of Guillain-Barré Syndrome

The various clinical subtypes of GBS are defined by their differential involvement of motor and sensory axons of peripheral nerves ( Table 20.1 ). In most populations, the most common form of GBS is acute inflammatory demyelinating polyneuropathy, which in most series accounts for about 75% of cases, while acute motor axonal neuropathy most often occurs in outbreaks associated with Campylobacter jejuni infection.

Pathophysiology

GBS is likely an autoimmune disorder resulting from T-cell activation and production of antibodies directed at antigenic proteins of the peripheral nerves. Infectious agents such as Epstein-Barr virus, cytomegalovirus, Mycoplasma pneumoniae , and Campylobacter jejuni , or stressors such as immunizations, surgery, or parturition may trigger formation of these antibodies. Although most such antibodies are directed at myelin proteins, in some cases axonal moieties may be the primary target of immune-mediated nerve injury. Thus, while GBS has traditionally also been known as acute inflammatory demyelinating polyneuropathy (AIDP), the primary pathologic process of this condition is in some cases axonal injury, as in the GBS variants acute motor axonal neuropathy (AMAN) and acute motor and sensory axonal neuropathy (AMSAN).

The immunologic mechanisms of nerve injury underlying GBS have been extensively studied but remain incompletely understood. The peripheral nervous system is usually protected from circulating macromolecules by a blood-nerve barrier, maintained at the level of endothelial cell tight junctions within capillary beds. This barrier blocks access of proteins and lymphocytes to the brain and peripheral nervous system, and stops antibody binding to peripheral nerve antigens.

The first step in the process underlying GBS is breakdown of this immunologic tolerance, which may occur after an infection or other immune-activating event as a result of molecular mimicry, superantigen mechanisms, or cytokine activation. The blood-nerve barrier is relatively deficient at the level of the nerve roots and distal motor nerve terminals, rendering these sites particularly vulnerable to immune-mediated peripheral nerve injury.

The next step involves antigen recognition by T cell receptors, and antigen processing through major histocompatibility complexes. Activated T cells gain access to their target autoantigens by crossing the blood-nerve barrier and migrating into the endoneurium via adhesion molecules on endoneurial endothelial cell walls. In concert with macrophages, these activated T cells react against autoantigens on Schwann cell or axonal constituents. Simultaneous activation of complement is triggered by antibody binding to Schwann cells and deposition of activated complement components.

In AIDP, these inflammatory infiltrates trigger segmental demyelination of motor and sensory nerve fibers, often in association with secondary axonal degeneration, which is generally first apparent at the level of the spinal roots and distal nerve endings where the blood-nerve barrier is relatively deficient. The target antigens on Schwann cells in AIDP are essentially unknown. Vesicular myelin degeneration in GBS may be discrete or diffuse, and may affect the peripheral nerves at any point from their origin in the spinal cord to the neuromuscular junction. The weakness of GBS results from conduction block and concomitant or primary axonal injury in affected motor nerves. Pain and paresthesiae are the clinical correlates of sensory nerve involvement. Panel A in Figure 20.1 shows a schema of the possible pathogenesis of antibody-mediated Guillian-Barré syndrome.

Possible immunopathogenesis of the Guillain-Barré syndrome. Panel A shows the putative immunopathogenesis of acute inflammatory demyelinating polyneuropathy. Autoantibodies may bind to myelin antigens, activating complement and triggering formation of the membrane-attack complex (MAC) on the outer surface of Schwann cells. Vesicular myelin degeneration is followed by removal of myelin debris by macrophages. Panel B shows the proposed immunopathogenesis of acute motor axonal neuropathy (AMAN). Myelinated axons are divided into four functional regions: the nodes of Ranvier, paranodes, juxtaparanodes, and internodes. Myelin gangliosides GM1 and GD1a are strongly expressed at the nodes of Ranvier, where the voltage-gated sodium (Nav) channels are localized. Contactin-associated protein (Caspr) and voltage-gated potassium (Kv) channels are sited at the paranodes and juxtaparanodes. In AMAN, IgG anti-GM1 or anti-GD1a autoantibodies bind to the nodal axolemma, leading to MAC formation, complement deposition, loss of Nav clusters and detachment of paranodal myelin. This leads to failure of nerve conduction and axonal degeneration.

From Yuki and Hartung with permission.

In AMAN, IgG antibodies and activated complement bind to the motor nerve axolemma at the nodes of Ranvier, with subsequent formation of the membrane attack complex, axonal injury and degeneration, without focal demyelination or lymphocytic infiltrates (Panel B, Figure 20.1 ). This may impair nerve conduction by blocking sodium axonal channels or by triggering complement-mediated disruption of nodal and paranodal sodium channel clusters. Similar changes are seen in both motor and sensor nerve fibers in AMSAN.

Many investigators have tried to correlate specific antibody-ganglioside responses to symptoms and signs in discrete GBS variants ( Table 20.1 ). Gangliosides (ceramide-sugar complexes containing sialic acid and an oligosaccharide core) are common constituents in and on the surface of neuronal and axolemmal membranes, but they are relatively rare on the surface of Schwann cells and myelin. Not surprisingly, antibody responses against specific gangliosides appear most robust in those GBS variants that particularly target peripheral nerve axons. In most series, the most commonly identified autoantibodies are those against GM1 gangliosides, which are often seen in patients with AMAN and AMSAN. AMAN has also been linked with an antibody response to ganglioside GD1a, and less commonly to other ganglioside species. The Miller Fisher syndrome is associated with IgG1 antibodies to the ganglioside GQ1b. Antibody responses are not specific to the various GBS clinical and neurophysiologic sub-types.

Molecular mimicry has been postulated to underlie the link between antecedent infections and other stressors, and the development of specific antibody responses in GBS. Specific surface components of infectious agents, such as lipo-oligosaccharide in the outer membrane of C. jejuni , resemble specific gangliosides, GM1 and GD1a. A cross-reacting immune response may underlie some cases of GBS, with regional patterns of neurologic deficit reflecting varying ganglioside expression in different neural tissues.

B-cell mediated antibody responses are not the sole basis of GBS. Experimental allergic neuritis (EAN), a murine model of GBS, can be induced by immunization using peripheral myelin, whole constituent myelin proteins, or even protein fragments, or by passive transfer of activated cells sensitized against these proteins. Segmental demyelination in EAN is more severe when it occurs in response to both activated T cells and targeting antibodies, suggesting that both T cell- and B cell-mediated immunity are important in this disease model. Pro-inflammatory cytokine levels are also often diffusely increased in GBS, but their exact role in disease pathogenesis remains unclear.

The current evidence favors a pathogenic mechanism for GBS in which antibodies and activated T cells act in concert to coordinate a macrophage-mediated attack on specific components of the peripheral nerve that may include myelin, Schwann cells, motor and/or sensory axons. Nerve injury may be increased or lessened by local and circulating pro- or anti-inflammatory cytokines. Immune targeting in this process can be very focal, as in specific pure sensory, motor, or dysautonomic forms of GBS, or may result in diffuse injury to sensory and motor axons and their accompanying Schwann cells and myelin.

Investigations

There is no single investigation that can confirm or disprove the diagnosis of GBS, particularly early in its course. The diagnosis is contingent on supporting evidence from the clinical examination, lumbar puncture, neurophysiologic, radiologic and/or ancillary investigations, with exclusion of alternative diagnoses by the same means.

Lumbar puncture is generally undertaken in cases of suspected GBS, to rule out infectious conditions and identify the typical findings of GBS. These include elevation of the CSF protein (reflecting nerve root demyelination) without evidence of active infection (lack of CSF pleocytosis). This finding, known as cytoalbuminologic dissociation , is not always present in GBS; an increase in the CSF protein may not be seen in the first 48 hours of symptoms, and the CSF protein sometimes remains normal for as long as a week. The CSF protein level generally correlates poorly with disease severity. CSF microscopy in GBS generally reveals fewer than 5 leukocytes/mm ³ , but occasional patients have 10–50 mononuclear cells/mm ³ . Significant CSF pleocytosis (>50 leukocytes/mm ³ ) is very atypical of childhood GBS, and should prompt consideration of other disorders such as poliomyelitis, enterovirus 71, and HIV infection. As CSF pleocytosis is not invariable in myelitis, or even encephalitis, the CSF cell count may not reliably differentiate between GBS and other conditions. The CSF is usually normal in tick paralysis and botulism.

Neurophysiologic studies (nerve conduction studies and electromyography) confirm the clinical diagnosis of GBS and characterize the underlying process as primarily demyelinating or axonal. Peripheral nerve demyelination is reflected in slowing of nerve conduction, which tends to be more profound in younger children. Because these changes are patchy in GBS, testing of multiple nerves characteristically reveals variable increases in distal latencies, slowing of nerve conduction, conduction block, and dispersion of motor responses. The late responses (F and H waves), which test both distal and more proximal segments of the peripheral nerves, are often more sensitive in identifying mild slowing of nerve conduction. Both severe demyelination and axonal injury cause loss of amplitude of motor and sensory responses. In the pure axonal forms of GBS, conduction velocities are preserved despite loss of amplitude of motor responses. Electromyography reveals a loss of motor unit action potentials (decreased recruitment) early in the illness. Changes of acute denervation (fibrillation potentials and positive sharp waves) are seen only in cases with significant axonal injury (whether primary or secondary), and are not usually apparent until 7–10 days into the disease course, although they may be seen earlier in very small children.

Electrodiagnostic finding are, to some extent, predictive of outcome in pediatric GBS. In primarily axonal forms of GBS, a severe reduction in CMAP amplitudes generally predicts a prolonged, and potentially incomplete, recovery. If the underlying process is acute peripheral nerve demyelination, this often primarily affects proximal nerve roots and the terminal segments of motor nerves, and can be accompanied by conduction block, predominantly as the nerve reaches its termination in the intramuscular regions. In some cases this results in unrecordable motor responses, and it may not be possible, on neurophysiologic criteria, to distinguish this situation from extensive axonal degeneration. The prognostic implications of severely reduced CMAP amplitudes must therefore be interpreted with caution. When peripheral responses are unrecordable, a search for more proximal responses (such as from stimulation of the phrenic or axillary nerves) may enable identification of a primarily demyelinating process. Electromyographic evidence of acute denervation is always suggestive of severe axonal injury and a less good prognosis.

Performance of a detailed neurophysiologic study enables diagnosis of GBS in as many as 90% of cases during the first week of symptoms. Changes are virtually universal by the second week of illness, by which time a definitive diagnosis can almost always be made. Electrodiagnostic studies are uncomfortable, however, and can be technically challenging in small children; therefore, they should be performed only by individuals with appropriate pediatric expertise.

Neuroimaging has become a valuable adjunct to neurophysiologic studies in suspected GBS. Post-gadolinium enhancement of the peripheral nerve roots and cauda equina may be seen on spinal magnetic resonance imaging in children with GBS. Similar findings are often also seen in the cranial nerves of patients with clinical evidence of cranial neuropathy. These findings are seen in as many as 95% of patients with GBS, but they are not specific to this disorder. Similar changes can be seen after lumbar puncture and in chronic genetic and inflammatory neuropathies. Spinal magnetic resonance imaging (MRI) should be undertaken as a matter of urgency in children presenting with severe back pain, a sensory level, or prominent sphincter dysfunction, in order to exclude transverse myelitis or spinal cord compression.

Nerve biopsy is virtually never required for diagnosis of pediatric GBS but, where performed, typically shows mononuclear endoneurial inflammatory infiltrates, variable degrees of Wallerian degeneration, and macrophage-mediated segmental demyelination.

Antiganglioside antibodies are identifiable in about 50% of patients with GBS. There is a degree of ganglioside specificity to GBS subtype, but there is overlap between these subtypes and the ganglioside antibody subtypes identifiable in individual subjects ( Table 20.1 ). Where available, testing for these antibodies can be very useful in confirming the clinical diagnosis, particularly in patients with atypical presentations or formes fruste of GBS.

Differential Diagnosis

The differential diagnosis of childhood GBS includes disorders involving the central nervous system, peripheral nerves, neuromuscular junction, nerve, and muscle ( Box 20.2 ). See Case Example 20.1 .

As many as 50% of children with GBS have significant ataxia, which may relate to proximal weakness causing clumsiness and frequent falls, or to sensory ataxia from decreased afferent input. These findings raise the possibility of acute cerebellar ataxia, which can be distinguished from GBS in that it is not associated with weakness or loss of the muscle stretch reflexes.

Transverse myelitis and acute disseminated encephalomyelitis (ADEM) are syndromes of central nervous system demyelination that share with GBS an autoimmune pathogenesis and tendency to develop after a viral illness or other prodrome. Occasional cases of concurrent ADEM and GBS or GBS and transverse myelitis are reported, in which instance there is a case for combined treatment with corticosteroids and IVIg or plasma exchange. There can be considerable overlap in the clinical presentation of acute myelopathy and early GBS. Both can be associated with back and/or leg pain, and weakness initially confined to the lower extremities. A sensory level may be difficult to identify in the young child, and reflexes may be depressed early in the course of the acute spinal cord syndrome. Sphincter dysfunction is common with spinal cord lesions, but relatively rare (and usually transient) in GBS. Transverse myelitis may be associated with marked elevation of the CSF protein and significant CSF pleocytosis. Neurophysiologic studies are usually normal in transverse myelitis, but there may be loss of late responses (F and H waves) corresponding with the affected spinal level(s).

Poliomyelitis has been largely eradicated in countries with widespread immunization programs, but continues to surface in nonimmunized groups and immunocompromised individuals. Where the attenuated oral vaccine is employed, cases of vaccine-associated paralytic polio can follow mutation of the attenuated live virus to a wild-type phenotype causing paralytic disease. The risk of paralytic poliomyelitis after immunization with the oral live-attenuated Sabin vaccine is 1 in 2.5 million vaccinations. This disorder will hopefully no longer be seen now that most countries have transitioned to immunization with the inactivated Salk vaccine. Other enteroviruses can also cause acute focal, asymmetric limb weakness, usually in association with fever and pain, which can mimic GBS. Probably the most common single species is enterovirus 71, which has been responsible for epidemic outbreaks with a relatively high incidence of neurologic complications among infected children. In addition to direct viral invasion of the central nervous system causing meningoencephalitis or poliomyelitis, this strain of enterovirus appears to be highly immunogenic, resulting in postinfectious disorders such as acute cerebellar ataxia and GBS. CSF analysis shows a polymorphonuclear pleocytosis, while nerve conduction studies reflect acute denervation without demyelination. Neuroimaging may show striking preferential involvement of the anterior horn. A definitive diagnosis is achieved by serologic testing.

Other acute neuropathies can be distinguished from GBS based on the clinical setting or associated findings. Children with mitochondrial diseases such as pyruvate dehydrogenase deficiency can present with an acute illness resembling GBS. In such cases an inborn error of metabolism may be suggested by pre-existing developmental delay or epilepsy, the presence of systemic lactic acidosis, or a suggestive family history. Nerve conduction studies in such cases may be normal or may show a demyelinating or axonal neuropathy.

Guillain-Barré syndrome can occur in subjects with underlying inherited neuropathies such as Charcot-Marie-Tooth disease. In such cases the diagnosis may be known prior to development of an acute deterioration raising the question of a superimposed inflammatory process, or may be suspected when a child with GBS recovers less quickly and completely than would be expected. An underlying inherited neuropathy should be suspected in very severe cases of GBS, or where multiple family members develop an apparently acquired neuropathy.

Neuromuscular transmission disorders are sometimes confused with GBS. In small children infantile botulism should be considered, particularly where there is early constipation, involvement of the bulbar muscles, and internal and/or external ophthalmoplegia. Infant botulism is generally seen in the first 36 weeks of life. The CSF examination is normal in botulism, while the neurophysiologic examination reveals incremental responses on rapid repetitive nerve stimulation, with florid electromyographic changes (abundant fibrillation potentials and low-amplitude polyphasic motor unit action potentials). Myasthenia gravis tends to have a slower time course than GBS, and can generally be distinguished from a neuropathy because of the findings of ptosis, fluctuating ophthalmoparesis, and prominent proximal weakness. Tick paralysis typically affects young children between 2 and 5 years of age, and causes an ascending quadriparesis with internal and external ophthalmoplegia and early bulbar involvement (see below). When appendicular muscles are involved, children are frequently areflexic. On nerve conduction studies, tick paralysis causes loss of amplitude of the motor responses. The CSF is usually normal in this condition.

Acute myopathies are uncommon in childhood. Children with inflammatory myopathies may complain of myalgia and weakness, with muscle tenderness on palpation. The serum creatine kinase is usually markedly elevated in children with significant weakness due to an acute myositis, and mildly elevated in dermatomyositis, which tends to have a more indolent course and is often associated with skin changes. Periodic paralysis may present in childhood with weakness, hypo- or areflexia with normal level of consciousness, and a normal sensory examination. Bouts of paralysis are usually relatively brief and self-limited, resolving over hours, and may be associated with unusually high or low serum potassium values.

Management

On diagnosis, all children with GBS should be admitted to hospital for monitoring. Although weakness and hypotonia may be relatively mild at onset, the potential for sudden, sometimes fatal respiratory or autonomic compromise should always be anticipated in this condition. Children with mild GBS who are able to ambulate unassisted are usually treated expectantly. Those with rapid clinical progression, loss of the ability to walk, or significant bulbar or respiratory compromise should receive specific immunomodulatory treatment, with the aims of amelioration of disease severity and acceleration of recovery. Many pediatric clinicians would have a lower threshold for treating infants with immunosuppressive therapies because of the greater difficulty in assessing clinical progression in very small children.

Vigilant supportive therapy is vital in GBS and includes monitoring for respiratory and autonomic complications of this disorder, in addition to pain management and prevention of complications of immobility (constipation, pressure areas, contractures, and renal calculi). The vital signs and respiratory capacity should be closely monitored. Children should be admitted to a pediatric intensive care unit if they have a rapidly progressive course, inability to flex the neck or arms, flaccid limb weakness, or significant bulbar, respiratory, or autonomic dysfunction. Intubation and mechanical ventilation should be considered when vital capacity falls below 15 mL/kg of body weight, arterial pO ₂ falls below 70 mmHg, or there is significant fatigue. As many as 15–20% of children require ventilatory support for acute GBS. The need for mechanical ventilation relates to the severity of the limb and axial weakness, and is also predicted by severe hypotension.

Although there have been no prospective placebo-controlled randomized trials of these therapies in pediatric GBS, a number of prospective and retrospective studies have suggested that treatment with plasma exchange (PE) or intravenous immunoglobulin (IVIg) hastens recovery of independent ambulation in GBS, and shortens requirement for hospital admission. Both therapies appear most effective when given within 2 weeks of onset of symptoms. Neither therapy has been proven to improve long-term outcome in this disorder, although both improve outcome at 12 months after presentation. Both are thought to act by neutralizing antibody-mediated peripheral nerve dysfunction, and are probably most effective when administered within 7 days of symptom onset. Specifically, intravenous human immunoglobulin may act by binding pathogenic antibodies, downregulation of B cell-mediated antibody production and accelerated antibody catabolism, and complement inhibition. Other possible mechanisms of action include blockade of activated T cell receptors, augmentation of suppressor T cell activity, or inhibition of lymphocyte proliferation. Plasmapheresis reduces levels of circulating auto-antibodies, and may reduce levels of circulating pro-inflammatory cytokines or cell adhesion molecules.

Intravenous immunoglobulin is generally preferred for treatment of childhood GBS because of administrative ease and a preferable side-effect profile. A total dose of 2g/kg IVIg is given over 2 to 5 days and is generally well tolerated at all ages. Most centers give IVIg over 2 days, although two small studies have suggested a slightly greater risk of relapse with this regimen as opposed to a 5-day course. Possible side-effects of immunoglobulin include fever, headache, nausea and vomiting, myalgia, allergic reactions, hypercoagulability, and transmission of bloodborne infections. Comparison of therapeutic regimens has been limited by differences between treatment groups, including disease severity and time to treatment, but a meta-analysis of such trials that have been conducted to date suggests that IVIg is as effective as plasma exchange in accelerating recovery from GBS. A clinical response to IVIg is usually apparent within 3 to 7 days. Intravenous immunoglobulin is the agent of choice in patients with proven antibody-mediated GBS.

Plasma exchange (PE), or plasmapheresis, can be a safe and effective treatment of GBS for children weighing more than 10 kg, and is usually performed 4 to 6 times on an alternate-day schedule, to a total of 250 ml/kg (or triple-volume) exchange. Adverse effects of plasma exchange may include hypotension, infection, and hypercoagulability. A recent meta-analysis of treatment trials with plasmapheresis for GBS suggested that, overall, adults treated for GBS with PE have a slightly higher risk of relapse in the first year after treatment than those who are untreated, but 12 months after presentation are more likely to have recovered fully. PE has been shown in several small studies of pediatric GBS to decrease time to independent walking. There are few comparative trials of PE and IVIg in childhood GBS. A recent Egyptian study reported that PE might be superior to IVIg in very severe pediatric GBS necessitating ventilatory support, although this benefit was limited to time required for mechanical ventilation, and not to length of intensive care unit stay or short-term neurological outcome.

The effects of treatment with both PE and IVIg are commonly delayed for several days, during which interval clinical progression may occur. In severe cases, or patients whose condition deteriorates a few weeks after treatment (“treatment-related fluctuations”), repeated courses of plasmapheresis or IVIg may be beneficial. Sequential or combined administration of these treatments has not been studied in childhood but does not show additional benefit in adults with severe GBS. A current focus of research interest is individual variability in immunoglobulin pharmacokinetics, which may result in very variable IgG levels after immunoglobulin infusions and hence variable responsiveness to treatment. An ongoing double-blind randomized trial is examining whether there is a subpopulation of GBS patients with low IgG increments after infusions who would benefit from a second course of therapy.

Alternative techniques of immunoglobulin ultrafiltration, such as CSF filtration and immunoabsorption, have been reported in small adult series but not yet trialed in children with GBS.

Corticosteroid therapy was previously a mainstay of treatment for GBS, but has now been shown to be ineffective when used alone or in combination with IVIg. Steroid therapy may in fact slow recovery from GBS, and may be complicated by glucose intolerance and hypertension. Corticosteroids may, however, sometimes have a place in the treatment of severe radicular or neuropathic pain in GBS.

Blood pressure lability and arrhythmias can be life-threatening in very unwell children with GBS. Hypertension should be treated only when symptomatic or extreme, as there may be marked sensitivity to antihypertensive agents. Pacing devices are rarely required for children with severe bradycardia. Occasionally patients with severe GBS require treatment for gastroparesis, ileus, and urinary retention.

Pain is commonly underrecognized and undertreated in childhood GBS. Nonsteroidal anti-inflammatory medications and agents used to treat neuropathic pain (gabapentin, pregabalin, and amitriptyline) are often ineffective early in the disease course, but should be considered early, with rapid dose increases as tolerated. Corticosteroids may be effective in alleviating pain, although they do not otherwise ameliorate the course of GBS. Opioid analgesics are often required for adequate analgesia. Pain control is a paramount consideration even though narcotic analgesia may contribute to respiratory depression and constipation.

Physiotherapy should be initiated immediately upon control of pain in childhood GBS, and continued throughout the recovery period, to maintain skin integrity and joint range of motion, and prevent contractures. Careful attention should also be paid to nutrition, as the consequences of inadequate caloric intake include accelerated muscle catabolism.

Immobilization hypercalcemia is occasionally seen in children with severe GBS, and when severe may require treatment with calcitonin and bisphosphonates.

Outcome

Children with GBS typically have a shorter clinical course and more complete recovery than adults with this disorder. As many as 60% of children become nonambulant during their illness. Fifteen to 20% require ventilatory support, but respiratory failure does not predict a persisting deficit. Most children reach their clinical nadir within 2 weeks, and recovery begins soon thereafter. In most cases there is minimal residual impairment by 1 to 4 months from onset. Ultimately more than 90% of children recover fully, while a small minority have mild weakness (most commonly of the ankle dorsiflexors) but are able to walk unaided. Children with AIDP generally recover more quickly than children affected by AMAN or AMSAN.

Electrodiagnostic markers of severe axonal injury do not invariably predict poor outcome in pediatric GBS. Similarly, development of antibodies to GM1, GD1b, GD1a, and/or GT1a is not invariably associated with poor prognosis. Outcomes are less good in very young children and children who are very weak at presentation, quadriparetic on day 10, require ventilator support, or have inexcitable motor nerves.

Mortality in childhood GBS is low—of the order of 1–2%—and generally results from respiratory failure. The importance of zealous respiratory, cardiovascular, and autonomic monitoring in all cases of GBS cannot be overemphasized. Many deaths are due to potentially preventable respiratory complications. Autonomic instability is a predictor of fatal cardiac arrhythmias in GBS. Relapses occur in about 4% of cases of childhood GBS, and are generally responsive to immunomodulatory treatment. Occasionally children go on to have recurrent weakness and are ultimately diagnosed with chronic inflammatory demyelinating polyneuropathy.

Future Directions

Various aspects of the natural history and response to therapy in GBS are poorly understood. Individual susceptibility to development of forms of GBS (e.g. to AMAN after C. jejuni infection) is likely underpinned by as yet undefined genetic and other host factors, including immune-response polymorphisms. These are an active area of research interest.

An international multicenter collaborative study, the IGOS 1000 trial, is currently underway. This study aims to better characterize clinical, neurophysiologic, therapeutic, pharmacokinetic, immunological, and genetic features of children and adults with GBS, with a goal to identify predictors of course, prognosis, and response to therapy. An important aspect of such studies is both better definition of the natural history of this clinically heterogeneous condition and the establishment of outcome measures by which to assess the efficacy of new treatments. Research into pediatric GBS is particularly challenging in that it must address these issues in the context of the maturational and growth-related changes in peripheral nerve function in childhood.