Statistics from Altmetric.com
- anti-GM1 antibody
- central nervous system
- chronic inflammatory demyelinating polyneuropathy
- immune deposit
- thyrotoxic autoimmune encephalopathy
Vasculitic neuropathy can occur in patients with connective tissue diseases. On the other hand, non-systemic vasculitic neuropathy has been established as an independent clinical entity, and the risks for systemic spread and death are small.1 In patients with this disorder, vasculitis is limited to the peripheral nervous system (PNS), and histological evaluation is essential for the definitive diagnosis. We encountered a patient with isolated nervous system vasculitis who developed lethal encephalopathy. He had a persistently high titre of anti-GM1 IgG antibody, which is occasionally detected in patients with chronic inflammatory demyelinating polyneuropathy (CIDP).
A 67 year old man had been under treatment for type 2 diabetes for 10 years. In September 1998, he was referred to our hospital because of weight loss and numbness of the lower limbs. He was mentally alert and had exophthalmos. Muscle weakness was prominent in the distal muscles of all four limbs. Sensation was disturbed with a “stocking and glove” distribution. Deep tendon reflexes were diminished in all four extremities.
Results of laboratory examination indicated diabetes mellitus and hyperthyroidism (haemoglobin A1c 6.8% (normal range 4.3–5.8); thyroid stimulating hormone <0.03 μIU/ml (normal range 0.2–3.2); free triiodothyronine 13.2 pg/ml (normal range 2.9–6.0); free thyroxine 7.65 ng/dl (normal range 0.78–2.10); antithyroglobulin antibody 3200 (normal <100); antithyroid microsomal antibody 26 800 (normal <100); antithyrotropin receptor antibody 16.1% (normal <10); thyroid stimulating antibody 207% (normal <150)). Serological examination for antiganglioside antibodies revealed anti-GM1 IgG, and the titre was 25 900 (normal <800) by enzyme linked immunosorbent assay. No other antiganglioside antibodies were detected by thin layer chromatography immunoblotting. Antimyelin associated glycoprotein and antinuclear antibody were not detected. Cerebrospinal fluid (CSF) cell count was 4/mm3 and protein level was 53 mg/dl (normal <45). Oligoclonal bands and myelin basic protein were not detected.
Motor nerve conduction velocities (MCV, m/sec) and distal compound muscle action potential amplitudes (CMAP, mV) were as follows: median nerve 52.5 (normal >49) and 0.4 (normal >5); ulnar nerve 45.5 (normal >48) and 2.7 (normal >4); tibial nerve 40.0 (normal >50) and 3.8 (normal >7); peroneal nerve 36.9 (normal >48) and 1.2 (normal >3), respectively. Conduction blocks were observed bilaterally in the ulnar and peroneal nerves at the common sites of entrapment. Sensory nerve conduction velocity was 37.8 m/sec (normal >48) in the median nerve and was not evoked in the sural nerve. The patient was diagnosed as having CIDP and Graves’ disease. After treatment with thiamazole, thyroid function and levels of thyroid related autoantibodies normalised but the peripheral neuropathy remained. Muscle weakness and numbness improved following treatment with prednisolone at 50 mg/day and pulse intravenous methyl prednisolone 1 g/day for three days in June 1999 but his symptoms exacerbated again after five months. Diabetes and hyperthyroidism were well controlled, but anti-GM1 IgG titre was elevated to between 11 900 and 40 700. Despite intravenous immunoglobulin (IVIg) therapy, he suddenly had convulsions and consciousness disturbance on 11 May 2000. CSF examination showed a normal cell count, but the protein level was increased to 121 mg/dl. Nerve conduction studies revealed further reduction in CMAP amplitude with conduction block and delayed MCVs. Brain magnetic resonance imaging (MRI) demonstrated slightly high intensity signals in frontal white matter on T2-weighted image (fig 1A, arrow). Intra-arterial angiography showed no evidence of a cerebrovascular accident. His consciousness disturbance responded partially to treatment with pulse methyl prednisolone, IVIg, and plasma exchange. However, a T2-weighted MRI in June 2000 showed large high intensity signals in the right frontal cortex and white matter (fig 1B). Some of the lesions were enhanced with Gd-DTPA on a T1-weighted image. In July 2000, the lesions were enlarged in the cerebral cortex and white matter (fig 1C). Despite repeated immunomodulating therapies, he died on 17 August 2000. An autopsy of the brain showed disseminated multiple plaques in the pons and bilaterally in the cerebral white matter (fig 1D), corresponding to demyelination and axonal loss (fig 1E(1)). Haemorrhagic transformations with mild infiltration of inflammatory cells in the vascular and perivascular regions were observed in the temporal cortex (fig 1E(2)) and cerebellar hemisphere. Peripheral nerve roots obtained from the lumbar plexus exhibited vasculitic occlusion of small epineurial and endoneurial vessels with inflammatory cell infiltration (fig 1E(3)) and demyelination and axonal degeneration (fig 1E(4)). Immunohistochemical study revealed intense signals for IgG, C3, and C4 in vessels from the temporal cortex, white matter, and peripheral nerve root (fig 1F), but the signals were unremarkable in specimens of unaffected regions. Vasculitis was not evident except for in the nervous system.
The present patient showed chronic sensorimotor polyneuropathy similar to CIDP. However, the most prominent feature was lethal encephalopathy and isolated vasculitis in the nervous system. The brain lesions mainly consisted of demyelinative changes in the white matter, but the lesions in the temporal cortex and peripheral nerve roots indicated vasculitis. Immune deposits of IgG and complements were detected in the vascular regions only in the affected regions of the nervous system. The immune deposits may be associated with vascular damage resulting in cortical haemorrhagic transformation. To the best of our knowledge, this is the first report of histopathological analysis of CNS involvement in vasculitic neuropathy with no evidence of systemic collagen disease.
Anti-GM1 antibody is occasionally detected in patients with CIDP or systemic collagen disease with neurological manifestations.2 However, it has not been studied in patients with non-systemic vasculitic neuropathy. Brain endothelial cells and endoneurial cells share GM1 ganglioside antigens with peripheral nerve tissues, and anti-GM1 antibody facilitates leakage in the blood–nerve barrier.3 These findings indicate that anti-GM1 antibody might have induced demyelinative change and vascular damage in both the PNS and CNS of the present patient. It is noteworthy that CNS white matter lesions have been detected in patients with CIDP.4 Interestingly, anti-asialo-GM1 antibody has been frequently detected in patients with Graves’ disease or Hashimoto’s thyroiditis.5 Although Graves’ disease may have contributed to the development of the encephalopathy, lethal encephalopathy is an extremely unusual outcome not only in non-systemic vasculitic neuropathy but also in thyrotoxic autoimmune encephalopathy.6
In conclusion, vasculitic neuropathy seemed to have resulted in the development of lethal encephalopathy in the present patient. Furthermore, a common autoimmune mechanism mediated by anti-GM1 antibody similar to that in CIDP may have been involved in the lesions in the CNS as well as PNS.
We would like to thank Drs Y Oshima and Y Sato for their clinical assistance, Dr H Yoshino, Kohnodai Hospital, National Center of Neurology and Psychiatry, for his extensive analysis of antiglycosphingolipid antibodies, and Dr T Komori, Tokyo Metropolitan Institute for Neuroscience, for his helpful comments.
Competing interests: none declared
If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.