[Crohn's disease presenting with recurrent acute polyradiculoneuropathy]. / Polyradiculonévrite à rechutes révélant une maladie de Crohn.

INTRODUCTION: The neurological manifestations of Crohn's disease are rare, dominated by multiple mononeuropathies and the abnormalities of the white matter. Polyradiculoneurities remain exceptional. OBSERVATIONS: We report the case of a 33-year-old patient admitted for an ascending weakness of all four limbs. Eight years earlier he had presented a similar episode which had regressed spontaneously. The neurological examination revealed a tetraparesis with areflexia and hypotonia. These manifestations were concomitant with chronic diarrhea which had been neglected to date. The electrophysiological aspect was compatible with an acute polyradiculoneuritis. The analysis of the cerebrospinal fluid showed an albumino-cytological dissociation. The existence of the diarrhea directed the investigations towards an inflammatory enteropathy, which was attested later on by the endoscopic, radiologic and histological data leading to the diagnosis of active Crohn's disease. The diagnosis of a relapsing polyradiculoneuritis associated with Crohn's disease was retained. The patient was treated by salazopyrine-budesonide with improvement in the digestive and neurological manifestations after 3 years. CONCLUSION: The frequency of neurological features in Crohn's disease is not well documented. The incriminated mechanisms are either directly related to the disease (deficit in B12 vitamin or folic acid and/or by the means of an auto-immune vascularitis) or secondary to long-term treatment with metronidazole. The course of neurological manifestations is largely dependent on the course of the inflammatory disease.

Increased cerebrospinal fluid progranulin correlates with interleukin-6 in the acute phase of neuromyelitis optica spectrum disorder.

We examined progranulin (PGRN) levels in cerebrospinal fluid (CSF) samples during the acute phase in 15 patients with neuromyelitis optica spectrum disorders (NMOSD) and compared the results with those from 17 patients with multiple sclerosis (MS), 30 patients with other inflammatory neurological diseases (OIND), and 20 non-inflammatory controls (NIC). CSF PGRN levels of NMOSD patients were significantly higher than those of MS patients and NICs. These levels correlated with CSF interleukin-6 levels, CSF cell counts, CSF protein levels, improvements in the Expanded Disability Status Scale score, and affected total spinal cord lesion length in the NMOSD patients.

The Guillain-Barré syndrome: pulmonary-neurologic correlations.

In a retrospective analysis of 40 hospitalized patients with the Guillain-Barré syndrome (GBS), we related the use and outcome of assisted ventilation to specific quantitative details of the neurologic illness. Two patients had an unusually prolonged course: they were ventilated for 374 and 396 days before successful weaning. The other 38 patients were similar in most respects to those in previously reported series. Sixteen ventilated patients were hospitalized 56.6 +/- 10.6 (mean +/- S.E.M.) days, were ventilated 27.9 +/- 6.5 days, and had primarily pulmonary complications. There were 4 deaths during ventilation, and 9 of 13 survivors (69%) had a short-term excellent functional neurologic outcome. Attention to the neurologic details of the course of illness may spare some patients from tracheostomy. Twenty-two patients not requiring respirator support suffered distinctly less severe neuromuscular impairment with minimal cranial neuropathy, had no occurrence of pneumonia, and were discharged after 19.1 +/- 4.6 days. Eighty-one percent had an excellent functional outcome. The wide range of manifestations and severity of patients with GBS requires the attending physician to be flexible in dealing with each case and not make management decisions arbitrarily, by reference to a hypothetical "typical case."

Electrodiagnostic abnormalities in 113 consecutive patients with Guillain-Barré syndrome.

We performed electrodiagnostic tests on 113 consecutive patients with acute Guillain-Barré syndrome (103 within 3 weeks of onset). The most common motor conduction abnormalities were proximal conduction block alone (27%), proximal block associated with a distal lesion (27%), and generalized slowing (22%). Other combinations of abnormalities each occurred in fewer than 10% of patients. Thirty-seven percent of patients initially had normal sensory nerve conduction study results, most often in association with proximal conduction block. The characteristic early electrodiagnostic changes in Guillain-Barré syndrome were often present when cerebrospinal fluid protein concentration was still normal. Extensive early fibrillations occurred in 10 patients, 6 of whom recovered well. Patients with early generalized slowing of motor nerve conduction velocity, combined abnormalities, or low muscle action potential amplitudes in ulnar, median, and peroneal nerves generally, but not always, had poorer outcomes than patients with conduction block in one nerve segment. There was no consistent relationship between results of electrophysiologic studies and overall clinical grade or limb power, except that none of the patients with an isolated proximal block had virtual or complete paralysis in the same limb.

Cerebrospinal fluid interleukins, immunoglobulins, and fibronectin in neuroborreliosis.

Intrathecal synthesis of IgM and IgG, oligoclonal immunoglobulin bands, and the levels of fibronectin, soluble interleukin 2 receptor, interleukin 6, and tumor necrosis factor alpha were investigated with the use of enzyme-linked immunosorbent assay in 46 paired cerebrospinal fluid and serum samples from 32 patients with meningopolyradiculoneuritis due to Borrelia burgdorferi (Lyme borreliosis stage 2). Cerebrospinal fluid and serum interleukin 6, although not specific for neuroborreliosis, were good indicators of disease activity, while the serum soluble interleukin 2 receptor level was only mildly elevated. Tumor necrosis factor alpha was never detected in cerebrospinal fluid or serum specimens, and cerebrospinal fluid IgM, IgM index, and cerebrospinal fluid IgM/cerebrospinal fluid IgG ratios were significantly higher than in all other neuroimmunologic disorders evaluated and may be valuable diagnostic indicators for neuroborreliosis. The estimation of intrathecally synthesized IgG and IgM fractions for the differential diagnosis of neuroimmunologic disorders did not add to IgG and IgM index calculations.

Predominance of polymorphonuclear leukocytes in cerebrospinal fluid of AIDS patients with cytomegalovirus polyradiculomyelitis.

Cytomegalovirus (CMV) polyradiculomyelitis was diagnosed in 4 of 241 consecutive neurologically assessed human immunodeficiency virus type (HIV-1) seropositive patients. CMV-related neurologic disease was suspected on clinical grounds and was subsequently confirmed by CMV culture from cerebrospinal fluid (CSF) and/or CMV in situ hybridization on specific specimens. All four patients showed CSF pleocytosis with predominance of polymorphonuclear leukocytes (PMNs). Retrospective analysis of the results of CSF examination, performed in 143 of 241 patients with neurologic symptoms, showed pleocytosis in 58 of 143 patients. Predominance of PMNs was found in seven patients, including the four with CMV polyradiculomyelitis. It is concluded that in HIV-1 seropositive patients with a clinical diagnosis of polyradiculomyelitis, a predominance of PMNs in CSF could be an indication that the condition is CMV related. This should lead to early diagnosis and institution of specific antiviral therapy.

Guillain-Barré syndrome: activated complement components C3a and C5a in CSF.

Plasma and spinal fluid levels of complement activation products C3a and C5a were quantitated by radioimmunoassay in a group of 16 patients suffering from acute monophasic Guillain-Barré syndrome (GBS). Median CSF levels of C3a (118 ng/ml) and of C5a (9.6 ng/ml) were significantly elevated when compared with samples from a control group of patients with noninflammatory neurologic diseases. Plasma concentrations of these anaphylotoxic peptides were not significantly different between the two populations. Our findings indicate that the complement system is activated in the CSF of patients with acute GBS. Complement activation products may contribute to the inflammatory changes observed in this disorder.

Extra- and intrathecal production of antinerve and antibrain antibodies in Guillain-Barré syndrome: evaluation by an antibody index.

In serum and CSF samples from 19 patients with the Guillain-Barré syndrome (GBS), we assayed complement-fixing autoantibodies that react with peripheral and central nervous tissue. About two thirds of the serum and CSF samples reacted with preparations of one or both tissues. The distribution of antibodies, evaluated by the antibody index (CSF/serum titer divided by CSF/serum albumin), suggested predominantly extrathecal production in most patients, but an intrathecal contribution was also seen. Thus, the possible pathogenic effects of antineural antibodies in GBS may be exerted on both sides of the blood-brain barrier.

Serum and CSF levels of soluble intercellular adhesion molecule-1 (ICAM-1) in inflammatory neurologic diseases.

Intercellular adhesion molecule-1 (ICAM-1), a cell surface receptor important for cellular interactions in immune responses, especially leukocyte trafficking into inflamed tissue, is released in a soluble form (sICAM-1) into the extracellular space. In this study, we measured sICAM-1 in paired serum and CSF samples from patients with inflammatory diseases of the nervous system (IND) and calculated a sICAM-1 index as a measure of the intrathecal release of ICAM-1. In comparison with noninflammatory neurologic disease (NIND) controls, we found increased sICAM-1 index levels in viral meningoencephalitis, bacterial meningitis and, to a lesser degree, multiple sclerosis but not in Guillain-Barré syndrome. Serial examination of viral meningoencephalitis patients in most cases showed a decrease of sICAM-1 index in parallel with falling cell counts and clinical improvement. Except for those in bacterial meningitis, sICAM-1 serum levels of IND patients were not significantly different from those of NIND controls. The increased intrathecal release of sICAM-1 in viral meningoencephalitis and bacterial meningitis most likely reflects activation of macrophages and lymphocytes and provides evidence for a strong local immune response that itself, in addition to the infectious agent, may damage nervous tissue.

Soluble complement receptor type 1 in serum and cerebrospinal fluid of patients with Guillain-Barré syndrome and multiple sclerosis.

Activation of complement is critically involved in inflammatory reactions in both Guillain-Barré syndrome (GBS) and multiple sclerosis (MS). Soluble human complement receptor 1 (sCR1) blocks complement activation by both classical and alternative pathways. We studied serum and cerebrospinal fluid (CSF) concentrations of sCR1 in 23 patients with GBS, 27 patients with MS and 30 controls. No significant differences were found between patients and controls. Transient liver affection probably caused high serum sCR1 levels in two patients with GBS. The serum and CSF sCR1 levels were not correlated to the disease activity of GBS and MS, nor to the relapsing-remitting or chronic-progressive forms of MS. In GBS the CSF sCR1 levels correlated with the CSF total protein concentrations (r = 0.9, P < 0.01), suggesting that sCR1 leaks from serum into CSF via a damaged blood-nerve barrier. The serum sCR1 levels in GBS were slightly higher than in MS (P < 0.05). Whether this reflects changes in the release or consumption of sCR1 in these patients is at present unknown.

Study of IgA in the cerebrospinal fluid of neurological patients with special reference to size, subclass and local production.

IgA was assayed by particle counting immunoassay in cerebrospinal fluid (CSF) from non-neurological and neurological patients. Reference values had a logarithmic normal distribution with a mean of 1.54 mg/l and an upper limit of 5 mg/l. To estimate the possible intra-blood-brain barrier (BBB) production of IgA we have calculated an IgA index: CSF-IgA/serum-IgA: CSF-albumin/serum-albumin. Values higher than the upper reference limit of 0.41 were found in 12 out of 67 patients with multiple sclerosis (18%), in 5 out of 11 with aseptic meningitis, in 7 out of 8 with herpetic encephalitis, in 1 out of 8 with Guillain-Barré syndrome and in 2 cases of tuberculous meningitis. However, this index does not take into account the relative proportions of monomeric and polymeric IgA in CSF and serum. We therefore ultracentrifuged 17 paired CSF and serum samples and determined the relative proportions of monomeric and dimeric IgA and calculated the indices for monomeric and dimeric IgA. In controls the proportion of dimeric IgA in CSF was below 5% of total IgA whereas this proportion was increased up to 53.9% in the case of intra-BBB production of IgA, which is thus characterized by a very high dimeric IgA index. In all cases IgA1 remained the predominant subclass. These results had to be compared with those observed in cultures of peripheral blood lymphocytes, which secrete about equal proportions of monomeric and polymeric IgA pertaining to the IgA1 subclass.

Little change in cerebrospinal fluid amino acids in subtypes of multiple sclerosis compared with acute polyradiculoneuropathy.

Levels of free amino acids were determined in randomised, blinded samples of cerebrospinal fluid (CSF) from patients with relapsing-remitting or chronic progressive multiple sclerosis (MS), all in the active phase of disease. The levels were compared with amino acid amounts in patients with an acute polyradiculoneuropathy (Guillain-Barré syndrome (GBS)) and a control population of patients with no known neurological disease or deficit. The data did not indicate any significant changes in amino acid levels between MS subgroups. The only significant differences between MS patients and controls were a modest reduction in glutamate and a slight increase in taurine, but the changes were so small that the biological relevance is dubious. These results contrasted with the marked increases for many amino acids in CSF from patients with acute polyradiculoneuropathy compared with controls. The amino acid profile in cerebrospinal fluid (CSF) does not appear to provide evidence of differential pathology in multiple sclerosis (MS). The increase in hydrophobic amino acids and lysine in CSF from patients with acute polyradiculoneuropathy is consistent with transudation over the blood-CSF barrier following an infection. The increases in glutamine and alanine may reflect increased nitrogen removal from brain.

Cerebrospinal fluid (CSF) neuropeptide Y- and somatostatin-like immunoreactivities in man.

We have measured cerebrospinal fluid (CSF) neuropeptide Y-like immunoreactivity (NPY-LI) and somatostatin-like immunoreactivity (SLI) in control subjects and in patients with various neurologic disorders. We observed a significant reduction in CSF SLI in control subjects over 60 years of age, compared with the younger controls. CSF SLI was significantly decreased in multiple sclerosis (MS), or Guillain-Barre syndrome, compared with that of age-matched control subjects. A reduced concentration of NPY-LI was found in CSF of patients with MS. We have also examined the molecular heterogeneity of peptide-LI in CSF. Gel chromatography, not high performance liquid chromatography (HPLC), suggested two NPY immunoreactive materials in CSF. Gel chromatography and HPLC revealed three SLI components in CSF: somatostatin 14, somatostatin 28 and a higher molecular weight precursor. Our results suggest that 1) there may be more than one form of NPY in human CSF, and 2) somatostatin neurons might be more susceptible to alteration than NPY neurons in various pathological conditions and aging.

Demonstration of oligoclonal immunoglobulin G in Guillain-Barré syndrome and lymphocytic meningoradiculitis by isoelectric focusing.

Isoelectric focusing (IEF) of serum and CSF revealed oligoclonal IgG in 13 out of 16 patients with the Guillain-Barré syndrome (GBS), in 10 of them in serum only and in 3 in serum and CSF. Seventeen out of 19 patients with lymphocytic meningoradiculitis (LMR) showed oligoclonal IgG, 12 of them in CSF only. These findings, together with additional results, mean that in GBS oligoclonal IgG is synthesized mainly outside the CNS and in LMR within the CNS. Follow-up studies revealed changes in the oligoclonal IgG during the course of GBS and LMR. After treatment by plasma exchange the disappearance of oligoclonal IgG bands was followed by an improvement of GBS symptoms. The oligoclonal IgG bands returned in correlation with worsening of the disease. We were not able to elucidate the antibody character of oligoclonal IgG in GBS or LMR. No antibodies against the myelin basic protein (fragment 89-169) were detectable in the sera and CSF in any of the patients tested.

T and B lymphocytes in the cerebrospinal fluid of various neurological diseases.

Cerebrospinal fluids (CSF) from 66 patients with a variety of neurological disorders were studied for total protein content, absolute amount of albumin, IgA, IgG and IgM, as well as their quotients (fraction to total protein ratio), cell numbers and B cell and T cell levels. In addition, the percentage of B cells and T cells in the blood was determined in 34 patients and serum immunoglobulin levels were estimated in 51 patients. In noninflammatory diseases of the CNS, the percentage of B cells was slightly higher and T cell levels were lower in the CSF in comparison to corresponding blood values. The B cell to T cell ratio in viral meningitis was altered in the CSF. An apparent increase in the T cell level led to a decrease of B cell values. Similar changes were also found in optic neuritis. The percentage of T cells was higher in relapsing multiple sclerosis than in the chronic progressive form. There were less striking changes in the B cell to T cell ratios in the CSF of other inflammatory diseases of the CNS.

The soluble 60-kDa tumour necrosis factor receptor: no difference found between patients with relapsing-remitting multiple sclerosis and controls: increasing levels are associated with the recovery from Guillain-Barré syndrome.

We determined serum and cerebrospinal fluid (CSF) levels of the soluble 60-kDa tumour necrosis factor (TNF) receptor (sTNF-R p60) in 50 patients with relapsing-remitting multiple sclerosis (MS) and in 18 patients with Guillain-Barré syndrome (GBS). Neither in serum nor in CSF samples was there a statistically significant difference between mean receptor concentrations of patients with MS (serum: 1064, SD 262 pg/ml; CSF: 555, SD 130 pg/ml), with other noninflammatory neurological diseases (serum: 1008, SD 248 pg/ml; CSF: 530, SD 112 pg/ml) and with healthy control subjects (serum: 918, SD 180 pg/ml). In order to determine disease activity, magnetic resonance imaging (MRI) of the brain was performed in all MS patients. The mean sTNF-R p60 levels of patients who showed gadolinium DTPA enhancement on MRI were not different from those without enhancement (1034, SD 274 pg/ml vs 1099, SD 248 pg/ml in serum samples and 546, SD 109 pg/ml vs 565, SD 152 pg/ml in CSF samples). In GBS, the sTNF-R p60 levels of serum and CSF samples were significantly higher than in MS and all control groups except for the group with viral meningitis (VM) (GBS: 1544, SD 834 pg/ml in serum, 882, SD 147 pg/ml in CSF; VM: 1518, SD 375 pg/ml in serum, 1131, SD 611 pg/ml in CSF; P < 0.001 for serum samples and P < 0.005 for CSF samples). Serial serum sTNF-R p60 measurements in 13 patients with GBS showed an increase in receptor levels parallel with the recovery from the disease (1276, SD 374 pg/ml at the time of disease onset, 1554, SD 482 pg/ml 14-24 days later and 1787, SD 525 pg/ml after 28-32 days). From our results and the conflicting data of previous studies, we could not agree with the suggestion that the assessment of sTNF-R p60 in MS patients is a useful marker for disease activity. In GBS, subsequently increasing sTNF-R p60 levels are associated with recovery from the disease. It remains to be shown whether they might represent a relevant pathogenetic factor during this stage of GBS.

The human endogenous local anesthetic-like factor (ELLF) is functionally neutralized by serum albumin.

The cerebrospinal fluid (CSF) of patients with multiple sclerosis or Guillain-Barré syndrome contains a factor that inhibits excitation of nerve and muscle cells like local anesthetics. CSF samples containing the endogenous local anesthetic-like factor (ELLF) were analyzed by gel filtration chromatography and ultraviolet (UV) absorption at 210 nm. The active component was in a single peak corresponding to a molecular weight of 600-800 Da. This peak was decreased and the Na+ channel blocking activity was neutralized by the addition of 40 g/l human serum albumin to the CSF. When the albumin was separated from the CSF/albumin mixture by acetonitrile treatment, the Na+ channel blocking activity reappeared. The ELLF and its neutralization may be of relevance for the clinical fluctuations known with these diseases.

Beta-trace protein concentration in cerebrospinal fluid is decreased in patients with bacterial meningitis.

Although meninges represent a major site of biosynthesis, beta-trace protein (beta-trace) has not been studied in the cerebrospinal fluid (CSF) of meningitis patients. We measured beta-trace in lumbar CSF of normal controls (n = 27) and in patients with various neurological diseases (n = 92) by an immunonephelometric assay. The mean concentration of beta-trace in CSF of control patients was 16.6+/-3.6 mg/l. In bacterial meningitis (n = 41), CSF beta-trace was significantly decreased (8.7+/-3.9 mg/l; P< 0.001), whereas in spinal canal stenosis it was elevated (29.2+/-10.3 mg/l; P= 0.002). In viral meningoencephalitis (n = 12), beta-trace CSF concentrations were normal. Beta-trace concentrations remained below the normal range even after curing of bacterial meningitis, and normalisation of CSF leucocytes and blood-CSF barrier function. Beta-trace may be a useful tool for studying the pathophysiology of bacterial meningitis.

Cytokine pattern in the cerebrospinal fluid from patients with GBS and CIDP.

Macrophage-colony stimulating factor (M-CSF) and, less frequently, IL-1 beta and IL-6 were detected in the cerebrospinal fluid (SF) from Guillain-Barré syndrome (GBS) patients. IL-1 alpha, IL-2, IL-10, TNF alpha, and IFN gamma were not found. Detectable cytokine levels were not observed in chronic inflammatory demyelinating polyneuropathy (CIDP) SF nor in any of the sera studied. These findings suggest a prominent intrathecal activation of cells of the monocyte/macrophage lineage (Mø) in GBS, and further support the hypothesis of a crucial role for Mø in GBS immunopathology.

Isoelectric focusing of CSF proteins in known or probable infectious neurological diseases and the Guillain-Barré syndrome.

The CSF and serum proteins of 120 patients with known or probable infectious neurological diseases or the Guillain-Barré syndrome were examined with thin-layer IEF. All but two of these patients exhibited one or combinations of different CSF-protein aberrations in the acidic and alkaline range. Aberrant non-Ig fractions (including transferrin, the tau-fraction and gamma-trace protein) were found in frequencies varying between 4 and 48%. CSF Ig components of restricted heterogeneity, i.e. oligoclonal bands and/or regional increases of gamma-globulins, were more frequent in patients with (meningo-)encephalitic or (meningo-)-myelitis/radiculitic disorders (respectively 69 and 48%) than in subjects with meningitis or Guillain-Barré syndrome (17%). The occurrence of such Ig abnormalities was higher in subacute or chronic than in acute disease and in subjects examined greater than 4 weeks after the onset rather than earlier. Ig-band spectra with marked anodal extension were found predominantly in (meningo-)encephalitic disorders with infratentorial symptoms. Age and sex were not found to influence the occurrence of abnormal Ig fractions. Such components could be detected in spite of pronounced blood-CSF barrier defects.