PMCCPMCCPMCC

Search tips
Search criteria 

Advanced

 
Logo of jnnpsycJournal of Neurology, Neurosurgery and PsychiatryVisit this articleSubmit a manuscriptReceive email alertsContact usBMJ
 
J Neurol Neurosurg Psychiatry. 2007 November; 78(11): 1281–1290.
PMCID: PMC2117596

Immunotherapy responsive startle with antibodies to voltage gated potassium channels

Antibodies to potassium channels (VGKC‐Ab) were first associated with acquired neuromyotonia and its variant with CNS involvement, Morvan's syndrome.1 Recently, VGKC‐Ab were found in patients with non‐paraneoplastic limbic encephalitis (LE), characterised by personality changes, seizures and memory impairment.2 These patients may respond to immunotherapies. Thus the association of VGKC‐Ab and non‐paraneoplastic LE established the concept of a potentially reversible autoimmune encephalopathy.2 We describe a patient with startle syndrome and VGKC‐Ab, without neuromyotonia or LE, who responded dramatically to plasma exchange (PE) and immunosuppression, adding to the spectrum of disorders associated with VGKC‐Ab.

The patient was a 54‐year‐old female. Hypothyroidism was diagnosed at the age of 30 years. At the age of 52 years, sudden falls without loss of consciousness occurred. Two months later, massive muscle jerks appeared, spontaneously or elicited by auditory and somatosensory stimuli. The frequency of jerks increased during the ensuing months, together with mild gait impairment and scanning speech. Brain MRI was normal; CSF was normal except for the presence of oligoclonal bands. Prednisone (50 mg/day), gabapentin and levetiracetam were prescribed without improvement. Low doses of clonazepam were poorly tolerated and not effective. She was admitted to our institution in December 2003. Physical examination showed spontaneous and stimuli induced massive limb jerking, very mild limb ataxia and scanning speech, increased deep tendon reflexes and mild gait disturbance. Neuropsychological tests were normal. Brain MRI showed a faint non‐enhancing signal hyperintensity in the white matter of the middle cerebellar peduncles. No thymoma was found. CSF examination confirmed oligoclonal bands. Neoplastic markers (CA 15.3, CA 19.9, CA125, CEA, α‐fetoprotein) were normal. Antithyroid peroxidase and antithyroglobulin‐Ab were 278 (normal value <50 UI/ml) and 3096 (normal value <325 UI/ml), respectively. GAD and voltage gated calcium channel antibodies were absent; VGKC antibodies were 562 pM in serum and 46 pM in CSF (normal values <100 pM and <2pM, respectively).

Electroneuronography, long latency responses and EMG were normal. During video EEG, several episodes of sudden massive jerking occurred, spontaneously or elicited by unexpected auditory and somatosensory stimuli (see supplementary data, video1, 2, 3.avi; the supplementary data can be viewed on the J Neurol Neurosurg Psychiatry website at http://www.jnnpjnl.com/supplemental). Jerks correlated with EMG bursts lasting more than 100 ms and did not show habituation even when stimuli were repeated at intervals lower than 30 s. A high amplitude slow EEG transient with a phase reversal with respect to Cz consistently followed the onset of EMG bursts (fig A1). EMG recordings to evaluate the sequence of activation of different muscles in response to sudden stimuli showed an auditory blink on the orbicularis oculi (fig 1B1B,, arrow), followed by a generalised reflex response with earliest activation of the sternocleidomastoid, followed by masseter, trunk and limb muscles. The latency between the onset of EMG bursts recorded on intrinsic hand muscles and those recorded on biceps brachii slightly exceeded 20 ms (fig 1B1B).

figure jn98301.f1
Figure 1 (A1) EEG polygraphic recording performed before plasmapheresis; EMG electrodes recorded repeated spontaneous jerks involving the proximal and distal muscles of the upper right limb; bilateral and diffuse EEG slow transients followed the ...

The patient was treated with PE with a rapid and dramatic improvement in startle, time related with the first set of four exchanges. Thereafter, she received periodic PE for 1 year (two sessions every 2–3 months) and regular doses of oral cyclophosphamide with stable improvement. Prednisone was slowly tapered from 50 mg/day until withdrawal after 2 years and antiepileptic drugs considerably simplified. Video EEG performed after 1 year of PE and immunosuppression showed rare spontaneous jerks, clinically undetectable, appearing as occasional brief EMG bursts, followed by an EEG transient similar to that observed before immunotherapy (fig A2).

Immunocytochemistry on rat brain slices with patient's IgG and CSF showed staining of cortical neurons (layer V) and apical dendrites with a pattern similar to that of commercial antibody to Kv1.63 (see supplementary fig 2A–C; supplementary fig 2A–C can be viewed on the J Neurol Neurosurg Psychiatry website at http://www.jnnpjnl.com/supplemental).

Hyperekplexia or exaggerated startle can be distinguished from the normal startle response by its lower threshold, greater extent and resistance to habituation, criteria that were fulfilled by the features of the reflex movement disorder of our patient. The sequence of muscle activation and delay between EMG bursts recorded on proximal and distal limb muscles indicated a slow conduction time compatible with the activation pattern found in pathological startle reactions.4

An exaggerated motor startle reflex is one of the main features of hereditary hyperekplexia, but it has also been described secondary to other disorders involving the brainstem. In genetically transmitted startle disease, hyperexcitability probably results from a dysfunction of glycine mediated inhibition. In our patient, the excessive startle, which improved rapidly after PE, was associated with raised VGKC‐Ab in serum and CSF, suggesting that an insufficient hyperpolarising effect might play a pathogenic role. To our knowledge, this is the first report on the association between startle and VGKC‐Ab. The lack of neuromyotonia and of MRI changes of LE cannot be explained at present. The faint MRI changes observed in our patient may reflect the different distribution of VGKC within the CNS. The VGKC‐Ab in our patients were mainly directed against the Kv1.6 subunit and these antibodies are less frequently observed in patients with LE.5

Jerks improved rapidly and dramatically after the first set of four PE. Improvement was then maintained with repeated apheresis and oral cyclophosphamide, with progressive tapering of prednisone up to withdrawal. The rapid response to PE suggests a role for humoral immunity in our patient.

Thyroid autoimmunity has been reported in patients with neuromyotonia or LE with VGKC antibodies.1,2,3 Our patient suffered hypothyroidism, and had positive TPO antibodies. However, the clinical features did not fulfil the diagnostic criteria for Hashimoto encephalopathy as they were not compatible with either the vasculitic or diffuse–progressive type presentation of Hashimoto encephalopathy. Moreover, there is still no evidence for the pathogenic role of thyroid antibodies.

The association of startle, thyroid antibodies, oligloclonal bands and response to plasmapheresis indicates an autoimmune pathogenesis in our patient and suggests that VGKC antibodies should be assayed in patients with CNS hyperexcitability.

Supplementary data (video1, 2, 3.avi and fig 2A–C) can be viewed on the J Neurol Neurosurg Psychiatry website at http://www.jnnpjnl.com/supplemental.

Supplementary Material

[web only media]

Footnotes

Competing interests: Professor A Vincent's department receives payments and royalties for VGKC antibody assays. The other authors have not received any financial support from manufacturers of drugs or products cited in this manuscript.

Supplementary data (video1, 2, 3.avi and fig 2A–C) can be viewed on the J Neurol Neurosurg Psychiatry website at http://www.jnnpjnl.com/supplemental.

References

1. Liguori R, Vincent A, Clover L. et al Morvan's syndrome: peripheral and central nervous system and cardiac involvement with antibodies to voltage‐gated potassium channels. Brain 2001. 1242417–2426.2426 [PubMed]
2. Vincent A, Buckley C, Schott J M. et al Potassium channel antibody‐associated encephalopathy: a potentially immunotherapy responsive form of limbic encephalitis. Brain 2004. 1271–13.13
3. Antozzi C, Frassoni C, Vincent A. et al Sequential antibodies to potassium channels and glutamic acid decarboxylase in neuromyotonia. Neurology 2005. 641290–1293.1293 [PubMed]
4. Brown P. Neurophysiology of the startle syndrome and hyperekplexia. Adv Neurol 2002. 59153–159.159 [PubMed]
5. Kleopa K A, Elman L B, Lang B. et al Neuromyotonia and limbic encephalitis sera target mature Shaker‐type K+ channels: subunit specificity correlates with clinical manifestations. Brain 2006. 1291571–1584.1584

Articles from Journal of Neurology, Neurosurgery, and Psychiatry are provided here courtesy of BMJ Publishing Group