Acute Motor and Sensory Axonal Neuropathy

Whether acute motor and sensory axonal neuropathy is distinct from the more common AIDP or part of a continuum of disease remains a topic for on-going discussion. The initial clinical presentations of these disorders are virtually identical; however, children and adults with acute motor and sensory axonal neuropathy with a severe extent of axonal degeneration are more likely to have incomplete recovery. The electrophysiologic hallmark of this syndrome is the decrement or absence of sensory and motor action potential amplitudes, with only minimal slowing of nerve conduction velocities commensurate with the degree of axonal loss. In most instances, pathologic evaluation of the peripheral nervous system from patients within the spectrum of Guillain-Barré syndrome reveals evidence of endoneurial inflammation with primary demyelination and axonal degeneration [Berciano et al., 1993]. Either process can occur in an individual patient, within a particular peripheral nerve or nerve segment [Berciano et al., 1997]. The mechanisms through which the immune system targets specific constituents of the peripheral nervous system remain poorly understood. It is clear that the immune response can be remarkably specific in affecting epitopes associated with myelin, those associated with axolemma, or both [Griffin et al., 1996a, 1996b; Hughes et al., 1999]. There have been a number of studies attempting to associate the presence of circulating immunoglobulins directed against specific ganglioside moieties with specific Guillain-Barré syndrome-related entities. For the most part, there has been only limited correlation between the targeted antigens and the syndrome phenotype. The strongest associations have been reported in the Miller Fisher syndrome with an immune response to the ganglioside GQ1B [Hughes et al., 1999; Saida, 1996; Kusunoki, 2008; Nishimoto, 2008; Asbury, 2000; Hughes and Cornblath, 2005; Yuki, 2000; van Doorn, 2009; Willison et al., 2002, 2008].

Reisin et al. [1993, 1996] reported a series of 44 children with an acute axonal form of Guillain-Barré syndrome. These patients initially had severe reduction in the amplitude of the compound motor action potentials (less than 10 percent of lower limit of normal) and, after 2 weeks of illness, had diffuse, severe denervation on needle electrode examination. Compared to patients with compound motor action potentials that were greater than 10 percent of normal, they were more likely to require assisted ventilation (60 versus 6.2 percent), were more frequently quadriplegic at the peak of their disability (80 versus 18.7 percent), and required longer periods to improve one functional grade (mean, 63.6 days versus 16.6 days) and to become ambulatory (mean, 156 versus 17.6 days).

Acute Motor Axonal Neuropathy

Although first recognized among patients from southern Mexico about 50 years ago as a distinct variant of typical Guillain-Barré syndrome on the basis of disease pathology, the first clear and detailed description of acute motor axonal neuropathy resulted from studies of clustered outbreaks in rural regions of northern China. The clinical presentation is similar to AIDP or acute motor and sensory axonal neuropathy, but is distinguished by the discrete involvement of motor axons with sparing of sensory axons [Lu et al., 2000]. Electrophysiologic hallmarks of the disease are normal sensory nerve conduction velocity studies and diminished compound motor action potential amplitudes, with normal or near-normal motor conduction velocities [McKhann et al., 1991]. Diffuse denervation is present during needle electromyography. Histopathologic examination reveals normal-appearing sensory nerves with degeneration of motor axons within motor nerve roots and motor fibers in the peripheral nervous system [Griffin et al., 1995, 1996a, b; Ho et al., 1995, 1997a; Lu et al., 2000; McKhann et al., 1993]. Just as AIDP and acute motor and sensory axonal neuropathy occur after gastrointestinal infection with Campylobacter jejuni, the association with acute motor axonal neuropathy is even stronger [Hadan and Gregson, 2001; Islam et al., 2010; Kalra et al., 2009b; Kuwabara et al., 2004; Liu et al., 2003; Nachamkin et al., 2007; Nishimoto et al., 2008a; Ogawara et al., 2000, 2003]. Patients who develop acute motor axonal neuropathy in northern China come from rural areas where the ability to purify communal drinking water is often limited or nonexistent [Wu et al., 1999]. Similar findings have been noted in children with Guillain-Barré syndrome who were treated in Buenos Aires, Argentina [Paradiso et al., 1999]. The patients with acute motor axonal neuropathy constituted 30 percent of children with Guillain-Barré syndrome with a high incidence of evidence of prior infection with C. jejuni. Although some authors have suggested that axonal forms of Guillain-Barré syndrome have a more protracted and less complete recovery than in acute inflammatory demyelinating polyneuropathy, most reports have noted that the long-term outcomes are equivalent, or nearly so [Barzegar et al., 2008; Gupta et al., 2008; Kalra et al., 2009a; Korinthenberg et al., 2007; Ramachandran and Kuruvilla, 2004; Ramirez-Zamora et al., 2009; Wierzba et al., 2008].

Miller Fisher Syndrome

Ophthalmoplegia, ataxia, and areflexia characterize this Guillain-Barré syndrome variant, described originally in adults in 1956 [Arakawa et al., 1993; Bradshaw and Jones, 1992; Eggenberger et al., 1993; Jones, 2000; Sladky, 2004; Wong, 1997]. The incidence of Miller Fisher syndrome is probably 2–4 percent in children with Guillain-Barré syndrome-like illnesses (see Table 90-5) [Bradshaw and Jones, 1992; Jones, 2000; Sladky, 2004; Jones, 2000; Korinthenberg et al., 2007; Ryan, 2005]. Cerebrospinal fluid protein elevation is seen in virtually all patients, and in most patients, complete recovery occurs. Brainstem auditory-evoked potentials have demonstrated peripheral and central auditory conduction defects in children [Wong, 1997]. Of interest are recent investigations demonstrating antibodies against the ganglioside GQ1b that recognize similar epitopes from specific C. jejuni strains [Chaudhry et al., 2006; Jacobs et al., 1997a; Jones, 2000, Korinthenberg et al., 2007; Ryan, 2005; Takahashi et al., 2005] and after Hemophilus influenzae infection [Jacobs et al., 2008; Koga et al., 2001; Nagashima, 2007; Nishimoto et al., 2004; Nishimoto, 2008a]. Cross-reactivity with the GQ1b ganglioside, which is present in cranial and other peripheral nerves, may account for the restricted involvement seen in patients with Miller Fisher syndrome [Willison et al., 1993; Willison and O’Hanlon, 1999; Willison 2005a, b, 2007]. This is in contrast to the classic form of Guillain-Barré syndrome, in which molecular mimicry has been described between anti-GM1 antibodies and C. jejuni [Ang et al., 2001a; Carpo et al., 1998; Kuroki et al., 2001; Hughes and Cornblath, 2005, Jacobs et al., 2008; Meyer Zu Horste et al., 2010; Nachamkin et al., 2007]. Successful treatment with intravenous immune globulin has been reported in a 3-year-old boy with Miller Fisher syndrome [Arakawa et al., 1993].

Polyneuritis Cranialis

Polyneuritis cranialis can occur in children, and manifests with the acute onset of multiple cranial nerve palsies (usually bilateral VII and sparing of II), with elevated cerebrospinal fluid protein, slowing of motor conduction velocities, and good recovery [Lyu and Chen, 2004; Polo et al., 1992, 2002]. Bilateral facial weakness, dysphonia, and dysphagia are typical symptoms. Evidence suggests some association with cytomegalovirus infection. One study of 20 cytomegalovirus-associated Guillain-Barré syndrome patients compared the findings with earlier established data of C. jejuni-related Guillain-Barré syndrome patients and found that the patients were significantly younger, initially had a severe course, indicated by a high frequency of respiratory insufficiency, and often developed cranial nerve involvement and severe sensory loss [Visser et al., 1996]. These observations were in contrast to patients with C. jejuni infection, which was more commonly associated with an AIDP phenotype [Hughes and Cornblath, 2005; Jacobs et al., 2008; Meyer Zu Horste et al., 2010; Nachamkin et al., 2007; Nagashima et al., 2004; Nagashima, 2007; Nishimoto et al., 2008a]. Enhancement of multiple cranial nerves seen with postcontrast magnetic resonance imaging (MRI) has also been documented, but clinical symptoms and electrodiagnostic studies reflected abnormalities caused by some, but not all, of the enhancing cranial nerves [Fulbright et al., 1995; Morosini et al., 2003].

Acute Pandysautonomia

Acute pandysautonomia is a relatively rare disorder, even within the group of diseases falling under the Guillain-Barré syndrome rubric. There are no large series of well-studied patients, and the most extensive reports cobble together a handful of subjects that have often been seen at institutions over a protracted time period. The collection of reports available in the literature, as might be anticipated, represents a heterogeneous patient population. The patients, by definition, have in common symptoms of autonomic failure manifestations that have a limited range of expression. Hence, dysautonomia provides the common thread that binds this group into a single entity without connoting a homogeneous underlying pathophysiology. Typically, such patients present with orthostatic hypotension, intestinal dysmotility, bladder dysfunction, syncope, vomiting, and difficulties with urination [Low, 1994]. They may also exhibit hypo- or hyperhidrosis, pupillary abnormalities, sensory deficits, and loss of reflexes. In addition, some patients have been described with symptoms or signs reflecting neurological abnormalities outside the confines of the autonomic nervous system and should probably be excluded from this diagnostic category. Even within this relatively homogeneous clinical group there is considerable heterogeneity in terms of its severity and outcome. In rare circumstances when the disorder has proven fatal, pathological data indicate that some patients have abnormalities quite distinctive from Guillain-Barré syndrome. There are a number of reports of patients who have developed an acute sensory neuronopathy or ganglionopathy that is usually associated with a poor recovery. Fortunately, most children who present with acute autonomic failure recover spontaneously, or respond to treatment with corticosteroids or intravenous immune globulin [Mericle and Triggs, 1997; Ramirez et al., 2004]. Acute dysautonomia most often is observed in the context of associated features that are typical of Guillain-Barré syndrome-like illnesses, making the diagnosis more straightforward. When children present with isolated acute autonomic failure, there is virtually no way to distinguish a Guillain-Barré syndrome variant from a ganglioneuronopathy early in the course of the illness. Empiric treatment is probably of limited risk and may be warranted when the dysautonomia is incapacitating. Most patients will improve relatively rapidly and fully recover.

Guillain-Barré Syndrome with Central Nervous System Manifestations

Although rare, Guillain-Barré syndrome with central nervous system (CNS) manifestations is reasonably well described in adults and children [Bradshaw and Jones, 2001; Maier et al., 1997; Nadkarni and Lisak, 1993; Okumura et al., 2002]. However, the incidence, severity, neuroimaging, and pathologic characterization of peripheral and central nervous system autoimmune disease in children remain elusive. One study reported postmortem findings in 13 adult patients with Guillain-Barré syndrome whose death ranged from 1 day to 12 months from the onset of neurologic symptoms [Maier et al., 1997]. CNS pathologic findings were predominantly secondary to injury to spinal nerve roots and cranial nerves. Evidence of primary inflammation was present, especially within the spinal cord, medulla, and pons, with clusters of mononuclear cells around small vessels. Although features of primary demyelination were identified in the peripheral nervous system in 12 of 13 patients, they were not detected in the CNS in any patients. Nadkarni and Lisak [1993] reported a 28-year-old patient with acute Guillain-Barré syndrome who developed bilateral optic neuritis and extensive CNS white-matter lesions on MRI. This illness was associated with Mycoplasma pneumoniae infection and suggested some degree of overlap with acute disseminated encephalomyelitis. It was hypothesized that certain infections might have a shared epitope that targets the peripheral and central nervous system. The clinical manifestations of CNS involvement in patients with Guillain-Barré syndrome-like illnesses range from nonspecific encephalopathy to more typical features of acute disseminated encephalomyelitis, transverse myelitis, optic neuritis, or brainstem/cerebellar syndromes [Bernard et al., 2008; Martens-Le Bouar and Korinthenberg, 2002]. The CNS component may be obscured by the peripheral nerve disease. The combination of mixed central and peripheral nervous system syndromes in children is rare, although not negligible. In one series, nearly 15 percent of children were thought to have evidence of central and peripheral nervous system disease; however, the actual frequency of Guillain-Barré syndrome-plus syndromes is probably lower. It is likely that such involvement in young children may be clinically difficult to detect, particularly if there is severe generalized weakness with or without respiratory failure.

Cerebrospinal Fluid

A characteristic laboratory finding supporting the diagnosis of Guillain-Barré syndrome is albuminocytologic dissociation or a disproportionate elevation of cerebrospinal fluid protein in the absence of significant evidence of inflammation (i.e., >10 mononuclear cells/mm³ of cerebrospinal fluid) after the first week of symptoms of the disease, essentially regardless of the subtype. The presence of significant cerebrospinal fluid pleocytosis (>50 mononuclear cells/mm³) or the presence of polymorphonuclear leukocytes in the cerebrospinal fluid should raise doubt about the diagnosis [Asbury and Cornblath, 1990; Jones, 1996]. The elevated cerebrospinal fluid protein includes both albumin and immunoglobulins. This feature of the illness is due to breakdown of the blood–nerve barrier within the subarachnoid space surrounding spinal nerve roots. Cerebrospinal fluid protein elevation may not be detectable early in the course in about 20 percent of children. If sites of inflammation within the peripheral nervous system are distal to the intraspinal roots, protein elevation within cerebrospinal fluid may be minimal. In most reported pediatric series, the cerebrospinal fluid protein levels have been in the range of 80–200 mg/dL (normal being less than 40 mg/dL) [Briscoe et al., 1987; Epstein and Sladky, 1990; Jones, 1996, 2000; Rantala et al., 1991; Ryan, 2005; Sladky, 2004].

Electrophysiologic Testing

Electrophysiologic testing can help to confirm the diagnosis and also provide useful adjunctive information, including characterizing the pathology and disease subtype more precisely, and potentially contributing information concerning prognosis. The pathologic hallmark of AIDP is multifocal, immune-mediated demyelination within spinal roots and peripheral nerves [Asbury et al., 1969]. The electrophysiologic correlate is the presence of multifocal slowing of nerve conduction or conduction block [Alma et al., 1998; Cros and Trigges, 1996; Hadden et al., 1998; Vriesendorp et al., 1995; Hughes and Cornblath, 2005; van Doorn, 2005; van Doorn et al., 2008; Vucic et al., 2009]. In the case of acute inflammatory demyelinating polyneuropathy, the task of the clinical electrophysiologist is to demonstrate electrophysiologic evidence of multifocal demyelination in sensory and motor peripheral nerves indicative of an acquired inflammatory disorder. This task requires demonstrating asymmetric slowing of nerve conduction along the course of peripheral nerves, disparate conduction velocities among comparable nerve segments, or the presence of conduction block. Asymmetric slowing along the course of a peripheral nerve might be manifest as disproportionate prolongation of the distal motor latency compared with the degree of conduction slowing in a more proximal nerve segment. Alternatively, markedly prolonged F-wave or H-reflex latencies in the face of reasonably normal distal motor latency and nerve conduction velocity in the limb are indicative of slowing of conduction in proximal nerve segments or spinal roots. Conduction block and temporal dispersion of proximally evoked compound motor action potentials are also features of multifocal segmental demyelination in peripheral nerves and are indicative of an acquired inflammatory demyelinating neuropathy. Box 90-2 includes the somewhat dated but useful consensus criteria of electrodiagnostic features considered supportive of the diagnosis of Guillain-Barré syndrome [Asbury and Cornblath, 1990]. Application of a modified version of these criteria resulted in 60 percent of adult patients with Guillain-Barré syndrome fulfilling the criteria for polyneuropathy at the first examination (mean time interval, 6 days since disease onset) [Meulstee and van der Meche, 1995].

Magnetic Resonance Imaging

Because the differential diagnosis includes transverse myelitis, a possibility that is underscored when weakness is confined to the lower extremities and the child is complaining of back pain, spinal MRI studies occasionally may need to be acquired. Several observers have noted marked gadolinium enhancement of the cauda equina and lumbar nerves that is indicative of an inflammatory process confined to dorsal and ventral spinal nerve roots in individuals with Guillain-Barré syndrome [Baran et al., 1993; Bertorini et al., 1995; Crino et al., 1994]. Conspicuous nerve root enhancement in some studies appeared to correlate with pain, disability grade, and duration of recovery [Gorson et al., 1996]. Nerve root enhancement was also 83 percent sensitive in the diagnosis of acute Guillain-Barré syndrome and was present in 95 percent of typical cases; it is potentially useful in patients in whom the electrophysiologic abnormalities are equivocal.