doi: 10.1053/ejpn.2002.0557 available online at http://www.idealibrary.

com on European Journal of Paediatric Neurology 2002; 6: 115119

CASE STUDY

Congenital GuillainBarre syndrome associated with maternal inflammatory bowel disease is responsive to intravenous immunoglobulin
NIGEL S BAMFORD,1,2 WERNER TROJABORG,1 ARIEL A SHERBANY,3 DARRYL C DE VIVO1,2
1

Department of Neurology, 2Department of Pediatrics, Columbia University College of Physicians and Surgeons, Neurological Institute and New York Presbyterian Hospital; 3Nyack Hospital, Nyack, New York, USA

A 34-week floppy preterm infant born to a mother with acute ulcerative colitis presented with a progressive reduction in spontaneous limb movements, severe generalized hypotonia, areflexia, autonomic dysfunction and respiratory failure. Electromyography revealed pronounced denervation activity and markedly slow nerve conduction velocity (3 m/s) with evidence of conduction block. These findings indicated demyelination with additional axonal features. The infant was diagnosed with congenital GuillainBarre syndrome, was treated with intravenous immunoglobulin and showed clinical improvement within 48 hours of treatment. The relationship between inflammatory bowel syndrome and inflammatory demyelinating polyneuropathy is discussed.
Keywords: Neonate. GuillainBarre syndrome. Floppy infant syndrome. Ulcerative colitis. Inflammatory bowel disease. Intravenous immunoglobulin.

Introduction
GuillainBarre syndrome is an acute inflammatory demyelinating polyneuropathy that is rarely diagnosed during the neonatal period. The diagnosis of GuillainBarre is suspected when motor conduction velocities are slowed in a floppy infant. We report a case of intravenous immunoglobulin responsive congenital GuillainBarre syndrome associated with maternal ulcerative colitis and discuss the possible relationship between the two diseases.

Case study
A boy was born preterm at 34 weeks' gestation to a 34-year old gravida 5, para 1 mother with active ulcerative colitis for which she received steroids

during pregnancy. The mother reported reduced fetal movements 2 weeks before delivery. History was significant for three lost pregnancies of unknown cause. The family history was otherwise unremarkable. Labour and delivery were uncomplicated. Apgar scores were 7 and 8 at 1 and 5 minutes respectively. Birth weight was 2020 grams, length 42 cm and head circumference 30.5 cm. The newborn baby was hypoactive, hypotonic and mildly arthogrypotic with elbow and knee contractures, and showed decreased grasp and Moro reflexes, poor suck, and stretch reflexes at the patelia and biceps. Brief ventilatory assistance was necessary. A head sonogram, computed tomography (CT) scan, urinary amino and organic acid concentrations were normal. At 19 days of age, supplemental oxygen was needed for episodes of desaturation. The head circumference was 31 cm. The pupillary responses were normal and symmetric. Eye grounds were

Received 01.08.2001. Revised 12.11.2001. Accepted 20.11.2001. Correspondence: Dr Nigel S Bamford, Neurological Institute, Box 189, 710 West 168th Street, New York, NY 10032, USA. Tel: 1-212-305-1247; Fax: 1-212305-5450; e-mail: nsb4@columbia.edu

10903798/02/06/0115+5 $35.00

& 2002 European Paediatric Neurology Society

116 normal. Ocular movements were full and no ptosis was present. Facial strength was normal. There were no tongue fasciculations and lingual papillae were present. Suck and swallow coordination were poor. The withdrawal of limbs to light touch and painful stimuli was appropriate and there was no tremor. Serial neurological examinations documented ascending paralysis culminating in absent spontaneous limb movements, severe generalized axial and appendicular hypotonia, and areflexia. Autonomic instability, marked by unstable heart rate and increasing apnoea eventually required intubation. Electromyography (EMG) revealed abnormal spontaneous activity in the right brachial biceps, quadriceps and anterior tibial muscles, which suggested denervation (Table 1). Giant muscle units were absent. Nerve conduction studies, using both needle and surface electrodes for stimulation and recording, failed to elicit motor or sensory responses. Interpretation of this study suggested denervation activity in all four limbs, inexcitable nerves or a technical failure. Serum creatine kinase was 206 U/litre. Arterial blood gas, haemogram, liver function, and serum chemistry were normal. Serum very-long chain fatty acids, mevalonic acid, and chromosomes were normal. DNA studies for spinal muscular atrophy gene deletion and myotonic dystrophy triplicate repeats were negative. Electroencephalography revealed excessive sharp frontal transients and trains of delta activity. The background activity was abnormal with a lack of complexity, consistent with mild diffuse cerebral dysfunction. A cranial magnetic resonance imaging study was normal. Cerebrospinal fluid analysis on day 58 revealed protein and glucose values of 62 mg/dl (normal:
Table 1

Case study: N S Bamford et al. 20100) and 80 mg/dl respectively. There were no white blood cells/mm3 and 1 red blood cell/mm3. Additional cerebrospinal fluid sampled on day 69 showed a protein rise to 72 mg/dl, glucose of 60 mg/dl and 1 white blood cell/mm3. A musclenerve biopsy revealed an excessive variation in fibre type. No nerve tissue was identified. However, the clinical picture of ascending weakness with areflexia, in the context of an essentially normal muscle biopsy, suggested a demyelinating process and prompted treatment. Intravenous immunoglobin at 400 mg/kg/day was given over 5 consecutive days. Improvement was seen within 48 hours with rapid resolution of autonomic instability and a gradual descending recovery of strength. A repeat EMG at age 77 days again revealed pronounced abnormal spontaneous activity with prominent sharp waves and fibrillations in the extensor digitorum brevis, abductor hallucis muscles, and less so in the anterior tibialis muscle (Table 1). Motor conduction studies were performed in the left peroneal and tibial nerves with concentric needle recordings from the respective small foot and anterior tibialis muscles. Distal motor latencies were prolonged, more so to the foot muscles than to the anterior tibial muscle. Fwaves were markedly prolonged. The motor conduction velocity was strikingly slow (3 m/s) to the foot muscles with evidence of conduction block (Fig. 1). The compound muscle action potential amplitude was reduced in all the muscles tested to < 1 mV. At age 5 months, he had recovered completely with normal tendon reflexes and tone. The child currently is 4 years old, cognitively intact with normal motor development and no neurological deficits.

Electromyography (EMG) and nerve conduction study results Age 19 days 0 mV 0 mV 0 ms Absent Absent 0 1+ None 77 days 0.88 mV 0.35 mV 5.9 ms 3 m/s Prolonged 0.39 2+ None Normal values > 1.6 mV < 3 ms > 30 m/s

Nerve conduction study and EMG parameter CMAP amplitude Distal Proximal

Distal latency Conduction velocity F-wave latency Proximal/distal amplitude ratio Abnormal spontaneous activity fibrillations/positive sharp waves Giant muscle unit potentials

The initial EMG with nerve conduction study on day 19, revealed absent motor and sensory responses in the right median, musculocutaneous and peroneal nerves. At 77 days, compound muscle action potentials (CMAP) could be evoked in the anterior tibialis muscle with low amplitude and slow conduction velocity. Distal and F-wave latencies are prolonged and abnormal spontaneous activity, suggestive of denervation, is present in both studies. Normal newborn values are provided.30,31 Spontaneous activity: 1+=two or more sites; 2+=more than three sites. mV, millivolts; ms, milliseconds; m/s, metres per second.

 Case study: Congenital GuillainBarre syndrome

117 latency and prolonged F-waves. In addition, low amplitude compound muscle action potentials and denervation activity indicates axonal degeneration, a frequent effect of primary inflammatory demyelination.12 It appears that the inability to evoke motor or sensory responses during the first electrophysiological examination reflected nerve inexcitability, also reported in axonal disease.13 Taken together, these data suggest a demyelinating disorder with additional axonal features. Although progressive clinical deterioration favours the diagnosis of chronic inflammatory demyelination polyneuropathy,14 the duration of symptoms10 and autonomic instability11 are consistent with Guil lainBarre syndrome. The pathophysiology which links demyelinating polyneuropathy and maternal inflammatory bowel disease is not well understood. However, a relationship between inflammatory bowel disease and neurological dysfunction in adults has been suggested. Two adults with Crohn's disease devel oped GuillainBarre syndrome with axonal features during the course of their illness and treatment with steroids in one produced a favourable response in motor strength.15 The development of GuillainBarre syndrome has also been described in two adult patients with ulcerative colitis in remission.16 These authors suggest that peripheral neuropathies could be regarded as an `extraintestinal manifestation' of inflammatory bowel disease.15,16 A more comprehensive review of neurological disorders associated with ulcerative colitis and Crohn's disease was published in 1995.17 The results of this survey revealed evidence of neurological involvement in 3% of patients with inflammatory bowel disease and were associated with both ulcerative colitis (47%) and Crohn's disease (53%). The onset of neurological disease usually precedes intestinal symptoms (ranging from 3 months to 12 years) but also occurs during exacerbation of inflammatory bowel disease symptoms. Neurological disease in these patients included peripheral neuropathy (31%), Guillain Barre syndrome (16%), myelopathy (26%), and less frequently, dermatomyositis, myasthenia gravis and various stroke syndromes. In addition, epilepsy has been found to constitute 1550% of neurological involvement in inflammatory bowel disease.18,19 Epilepsy and associated neurological symptoms also occur in patients with gluten sensitivity.20,21 Hadjivassiliou and colleagues have speculated that neurotoxic gliadin antibodies are involved in the neuropathological process.21 These antibodies demonstrate autoimmune activity with crossreactivity to epitopes on cerebellar Purkinje cells.22

Fig. 1. Evoked compound muscle action potentials (CMAP) in the abudctor hallucis muscle at age 77 days. Motor responses were recorded with a concentric needle electrode after stimulation of the anterior tibial nerve at the ankle and popliteal fossa. The conduction velocity is 3 m/s with a distal latency of 5.9 ms (35 mm). Conduction block is suggested by a 61% reduction of the CMAP amplitude with proximal stimulation compared with distal stimulation. The CMAP amplitude is < 1 mV and the proximal/ distal CMAP amplitude ratio is 0.39.

Discussion
The acute15 and chronic6,7 forms of inflammatory demyelinating polyneuropathy have been reported only rarely in the neonate. In all cases, hypotonia is apparent at birth with weakness and hyporeflexia. Where tested, motor nerve conduction velocities, evoked response amplitudes, and sensory nerve action potentials were reduced or absent with evidence of condution block. In congenital acute inflammatory demyelinating polyneuropathy or GuillainBarre syndrome, early improvement occurs within the first few months with complete recovery by 1 year. Newborns with chronic inflammatory demyelinating polyneuropathy present with an indolent or relapsing course and may respond to corticosteroid therapy. In our patient, and one other previously reported case of con genital GuillainBarre syndrome,5 the infant was born to a mother with active ulcerative colitis. Our patient was symptomatic at birth with diffuse hypotonicity and arthrogryposis, followed by ascending flaccid quadroparasis, areflexia, autonomic instability and pulmonary insufficiency. The differential diagnosis of such floppy infants is broad.8 However, consistent with demyelinating disease,911 our electrophysiology studies revealed slow peripheral nerve conduction velocities, motor nerve conduction block, prolonged distal motor

118 These findings underscore the relationship between inflammatory bowel disease and neurolo gical disease. In congenital GuillainBarre syndrome, involvement of the neonatal peripheral nervous system appears to occur when mothers demonstrate active ulcerative colitis during the gestational period. The pathogenesis of inflammatory bowel disease remains unclear but evidence suggests that an autoimmune, antibody mediated mechanism is involved.23 Likewise, a self antibodyantigen interaction has also been implicated in GuillainBarre syndrome and chronic inflammatory demyelinating polyneuropathy. Infection by Campylobacter jejuni has been linked to Guillain Barre syndrome and to exacerbations of inflammatory bowel disease.24,25 The immunopathogenesis of C. jejuni in GuillainBarre syndrome provides indirect support for the `shared epitope' model of this disease. Furthermore, autoantibodies to peptides spanning two extracellular domains of peripheral myelin protein-22 were recently described in 58% of adults with GuillainBarre syndrome and 41% of adults with chronic inflammatory demyelinating polyneuropathy.26 The transfer of maternally derived antibodies to the fetus from mothers with immune-mediated neurological disease occurs in transient neonatal myasthenia gravis. In this syndrome, transplacental antibody transfer occurs during the third trimester and newborns are affected to varying degrees proportional to the persistence of neonatal serum antibody concentration.27,28 Blocking antibodies specifically directed at epitopes of the mature but not the fetal neuromuscular junction have been described and, as seen in our patient, may account for a delay in the onset of symptoms following birth.28 In addition, antibodies directed against the fetal neuromuscular junction have been observed in healthy mothers with an obstetric history of arthrogryposis multiplex congenita.29 Of note, our patient had elbow and knee contractures consistent with arthrogryposis multiplex congenita which, together with decreased prenatal fetal movements, suggests an in utero onset of disease. The relationship between maternal ulcerative colitis and congenital GuillainBarre syndrome also suggests an immune-mediated process. It is possible that circulating maternal autoantibodies to specific structural proteins in ulcerative colitis cross the placenta and affect the infant while sparing the mother.11 Direct neonatal antibody synthesis by the neonate against a maternally derived antigen is also a consideration. As congenital GuillainBarre syndrome has been associated with maternal ulcerative colitis, it appears likely that additional cases may exist in

Case study: N S Bamford et al. which infants of mothers with ulcerative colitis are neurologically impaired but to a lesser degree. These babies may be diagnosed with `transient neonatal hypotonia' or would not readily come to medical attention because of the minor nature of the symptoms. Conversely, these infants may be severely affected with resulting miscarriage, early infant death, or arthrogryposis multiplex congenita. Future studies to examine the epidemiology, pathogenesis and outcome of infants born to mothers with inflammatory bowel syndrome would be instructive and help to identify the mechanism and spectrum of disease that may exist.

Acknowledgements
The authors thank Dr Arthur P Hays for reviewing the biopsy muscle tissue and the Colleen Giblin Charitable Foundation for Pediatric Neurology. Dr Bamford is a Neurological Sciences Academic Development Awardee (K12-NS01698).

References
1 Chambers R, MacDermot V. Polyneuritis as a cause of `amyotonia congenita'. Lancet 1957; 1: 397401. 2 Gilmartin RC, Ch'ien LT. GuillainBarre syndrome with hydrocephalus in early infancy. Arch Neurol 1977; 34: 567569. 3 Carroll JE, Jedziniak M, Guggenheim MA. Guillain Barre syndrome: another cause of the `floppy infant'. Am J Dis Child 1977; 131: 699700. 4 Al-Qudah AA, Shahar E, Logan WJ, Murphy EG. Neonatal GuillainBarre syndrome. Pediatr Neurol 1988; 4: 255256. 5 Jackson AH, Baquis GD, Shah BL. Congenital Guil lain-Barre syndrome. J Child Neurol 1996; 11: 407 410. 6 Pasternak JF, Fulling K, Nelson J, Prensky AL. An infant with chronic, relapsing polyneuropathy responsive to steroids. Dev Med Child Neurol 1982; 24: 504510. 7 Sladky JT, Brown MJ, Berman PH. Chronic inflammatory demyelinating polyneuropathy of infancy: a corticosteroid-responsive disorder. Ann Neurol 1986; 20: 7681. 8 Mason TBA, De Vivo DC. The hypotonic infant. In The Motor Disorders Textbook. Younger, DS ed. Philadelphia: Lippincott-Raven, 1998: 103109. 9 Meulstee J, van der Meche FG. Electrodiagnostic criteria for polyneuropathy and demyelination: application in 135 patients with GuillainBarre syndrome. Dutch GuillainBarre Study Group. J Neurol Neurosurg Psychiatry 1995; 59: 482486. 10 Research criteria for diagnosis of chronic inflammatory demyelinating polyneuropathy (CIDP). Report

 Case study: Congenital GuillainBarre syndrome

from an Ad Hoc Subcommittee of the American Academy of Neurology AIDS Task Force. Neurology 1991; 41: 617618. Trojaborg W. Acute and chronic neuropathies: new aspects of GuillainBarre syndrome and chronic inflammatory demyelinating polyneuropathy, an overview and an update. Electroencephalogr Clin Neurophysiol 1998; 107: 303316. Cros D, Triggs WJ. GuillainBarre syndrome: clinical neurophysiologic studies. Rev Neurol (Paris) 1996; 152: 339343. Gupta SK, Taly AB, Suresh TG, Rao S, Nagaraja D. Acute idiopathic axonal neuropathy (AIAN): a clinical and electrophysiological observation. Acta Neurol Scand 1994; 89: 220224. Connolly AM. Chronic inflammatory demyelinating polyneuropathy in childhood. Pediatr Neurol 2001; 24: 177183. Moormann B, Herath H, Mann O, Ferbert A. Involvement of the peripheral nervous system in Crohn's disease. Nervenarzt 1999; 70: 1107111. Zimmerman J, Steiner I, Garish D, Argor Z. GuillainBarre syndrome: a possible extraintestinal manifestation of ulcerative colitis. J Clin Gastroenterol 1985; 7: 301303. Lossos A, Yriver Y, Eliakim A, Steiner I. Neurologic aspects of inflammatory bowel disease. Neurology 1995; 45: 416421. Gendelman S, Present D, Janowitz HD. Neurological complications of inflammatory bowel disease (IBD), abstract. Gastroenterology 1982; 82: 1065. Elsbety AE, Bertorini TE. Neurological complications in Crohn's disease, abstract. Ann Neurol 1991; 30: 271. Gordon, N. Cerebellar ataxia and gluten sensitivity: a rare but possible cause of ataxia, even in childhood. Dev Med Child Neurol 2000; 42: 283286.

119
21 Hadjivassiliou H, Gibson A, Davies-Jones GAB et al. Does cryptic gluten sensitivity play a part in neurological illness? Lancet 1996; 347: 369371. 22 Hadjivassiliou H, Grunewald RA, Davies-Jones GAB et al. Humoral response in the pathogenesis of gluten ataxia, abstract. Ann Neurol 2001; 50: S37. 23 Podolsky DK. Inflammatory bowel disease (first of two parts). N Engl J Med 1991; 325: 928934. 24 Griffin JW, Ho TW. The GuillainBarre syndrome at 75: the Campylobacter connection. Ann Neurol 1993; 34: 125127. 25 Podolsky DK. Inflammatory bowel disease (second of two parts). N Engl J Med 1991; 325: 10081016. 26 Gabriel CM, Gregson NA, Hughes RA. Anti-PMP22 antibodies in patients with inflammatory neuropathy. J Neuroimmunol 2000; 104: 139146. 27 Lefvert AK, Osterman PO. Newborn infants to myasthenic mothers: a clinical study and an investigation of acetylcholine receptor antibodies in 17 children. Neurology 1983; 33: 133138. 28 Bushwald B, de Baets M, Luijckx GJ, Toyka KV. Neonatal GuillainBarre syndrome: blocking antibodies transmitted from mother to child. Neurology 1999; 53: 12461253. 29 Riemersma S, Vincent A, Beeson D et al. Association of arthrogryposis multiplex congenita with maternal antibodies inhibiting fetal acetylcholine receptor function. J Clin Invest 1996; 98: 23582363. 30 Dunn HG, Buckler W, Morrison GCE, Emery AW. Conduction velocity of motor nerves in infants and children. Pediatrics 1964; 34: 708727. 31 Blom S, Fibbstrom O. Motor conduction velocities in newborn infants of various gestational ages. Acta Paediatr Scand 1968; 57: 377384.

11

12 13

14 15 16

17 18 19 20