Article Text

Download PDFPDF

Paraneoplastic limbic encephalitis associated with potassium channel antibodies: value of anti-glial nuclear antibodies in identifying the tumour
  1. Luigi Zuliani1,
  2. Albert Saiz1,
  3. Bruno Tavolato2,
  4. Bruno Giometto3,
  5. Angela Vincent4,
  6. Francesc Graus5
  1. 1Service of Neurology, Hospital Clinic, Universitat de Barcelona and Institut d’Investigació Biomèdica August Pi i Sunyer (IDIBAPS), Barcelona, Spain
  2. 2Department of Neuroscience (2 Neurologic Clinic), University of Padua, Padua, Italy
  3. 3Service of Neurology, Treviso, Italy
  4. 4Neurosciences Group, Weatherall Institute of Molecular Medicine, University of Oxford, John Radcliffe Hospital, Oxford, UK
  5. 5Service of Neurology, Hospital Clinic, Universitat de Barcelona and Institut d’Investigació Biomèdica August Pi i Sunyer (IDIBAPS), Barcelona, Spain
  1. Correspondence to:
 Dr L Zuliani
 Department of Neuroscience (2 Neurologic Clinic), University of Padua, Ospedale San’Antonio Via Facciolati 71, Padova 35127, Italy; luigi.zuliani{at}unipd.it

Statistics from Altmetric.com

Limbic encephalitis is characterised by subacute development of short-term memory loss, seizures, confusion and psychiatric features.1 Paraneoplastic limbic encephalitis is generally associated with small cell lung cancer (SCLC), and around 50% of patients harbour Hu, CV2/CRMP5 or amphiphysin antibodies.1 Recently, antibodies against voltage-gated potassium channels (VGKC-abs) were identified as a new marker of a non-paraneoplastic, immunotherapy-responsive form of limbic encephalitis.2,3 However, the presence of VGKC-ab does not exclude a paraneoplastic cause of limbic encephalitis.1

We described previously a new antibody called anti-glial nuclear antibody (AGNA), a marker of SCLC shown to be closely associated with paraneoplastic Lambert–Eaton myasthenic syndrome (LEMS).4 Here we report on two patients with paraneoplastic limbic encephalitis associated with SCLC and VGKC-ab who also presented AGNA.

Methods

The presence of AGNA was assessed by immunohistochemistry on frozen sections of paraformaldehyde-fixed rat cerebellum in nine patients with limbic encephalitis and SCLC with (n = 2) or without (n = 7) VGKC-ab, and in seven patients with non-paraneoplastic limbic encephalitis and VGKC-ab. To confirm AGNA reactivity, an immunohistochemical competition assay was performed as described previously.4 VGKC-abs were detected by radioimmunoprecipitation (cutoff value: 100 pmol/l).2

Case reports

A middle-aged man was admitted to hospital with a 2-week history of behavioural changes and generalised seizures. A neurological examination showed short-term memory loss, confusion and aphasia. Electroencephalography showed slowing in the left temporal lobe. Magnetic resonance imaging (MRI) of the brain showed a hyperintense signal affecting the temporal lobes on T2-weighted images. Examination of the cerebrospinal fluid showed pleocytosis (20 lymphocytes/μl). A chest x ray showed a lung mass. Histological diagnosis of SCLC was made after biopsy of an enlarged groin adenopathy. The patient was treated with valproic acid, pulses of intravenous steroids and chemotherapy, with good control of seizures and remission of the behavioural problems and aphasia. After 6 months, in the setting of cancer progression, the patient had a relapse of limbic encephalitis (complex partial seizures and agitated behaviour) confirmed by a brain MRI. The patient was treated with repeated pulses of intravenous steroids, with good control of seizures and improvement of the agitation. He died 5 months later from cancer progression.

A middle-aged woman was admitted to hospital with a 2-month history of confusion, memory impairment and seizures. Neurological examination showed moderate short-term memory loss, mild confusion and abulia. Laboratory analysis showed hyponatraemia (118 mEq/l). Chest x ray showed a mediastinal mass confirmed by a computed tomography scan. Electroencephalography disclosed generalised slowing and epileptic activity in both temporal lobes. Cerebrospinal fluid examination and brain MRI were negative. A diagnosis of SCLC was made by transbronchial biopsy of the mediastinal lesion. Seizures did not disappear despite sodium level normalisation and treatment with multiple epilepsy drugs. Eventually the patient evolved to status epilepticus requiring intensive care and induction of barbituric coma. She died several months later from bronchopneumonia.

In both patients, the search for antineuronal antibodies (Hu, Yo, Ri, Ma1 and 2, CV2/CRMP-5, Tr, GAD, and amphiphysin) was negative. VGKC-ab levels were 170 and 220 pmol/l. On immunohistochemistry, serum from both patients showed intense labelling of the nuclei of Bergmann glia in the Purkinje cell layer of the cerebellum (fig 1). The presence of AGNA in our patients’ sera was confirmed by the immunocompetition assay.

Figure 1

 Frozen section of paraformaldehyde-fixed rat cerebellum immunoreacted with the serum of patient 1. There is intense labelling of the nuclei of Bergmann glia in the Purkinje cell layer typical of anti-glial nuclear antibody. Magnification ×200.

AGNA immunoreactivity was not detected in the serum of seven patients with limbic encephalitis and SCLC without VGKC-ab and seven patients with non-paraneoplastic limbic encephalitis and VGKC-ab.

Discussion

The diagnosis of classical limbic encephalitis is made in the setting of subacute onset of short-term memory loss, seizures and behavioural changes, imaging or pathological evidence of involvement of the limbic system, and exclusion of other potentially confusing diagnoses.1 Patient 1 fulfilled the criteria for classic limbic encephalitis of definite paraneoplastic origin.1 The clinical syndrome of patient 2 was initially attributed to a severe hyponatraemia, a common finding in limbic encephalitis associated with VGKC-ab.2,3 However, when the sodium level was normalised, the clinical syndrome did not improve and the brain MRI ruled out permanent hyponatraemia-related brain damage. A normal MRI does not rule out the possibility of limbic encephalitis. Normal brain MRI, including fluid-attenuated inversion-recovery and diffusion-weighted imaging, may repeatedly be seen in patients with autopsy-proved paraneoplastic limbic encephalitis (F Graus, unpublished data).

The clinical significance of the VGKC-ab levels in our patients is not clear. Some authors consider the cut-off value in the diagnosis of VGKC-ab-associated limbic encephalitis to be 400 pmol/l,2 but reports of patients with epilepsy and encephalopathy with lower titres5 suggest that the clinical relevance of VGKC-ab titres has to be better defined.

Our patients’ sera presented AGNA defined by robust immunoreactivity with the nuclei of Bergmann glia, an astrocyte population localised in the cerebellar cortex around the Purkinje cells.4 We identified AGNA as a marker of SCLC rather than paraneoplastic neurological syndromes. However, we observed a tight association between AGNA and voltage-gated calcium channel antibodies in the setting of SCLC. We found that among 35 patients with SCLC and voltage-gated calcium channel antibodies (LEMS 30, paraneoplastic cerebellar degeneration 5), 18 (51%) had AGNA whereas this antibody was not detected in a series of 19 patients with non-paraneoplastic LEMS, nor in 100 of 113 patients without paraneoplastic neurological disorders and SCLC.4

In this study, we found AGNA in two patients with limbic encephalitis, SCLC and VGKC-ab. No AGNA reactivity was found in patients with limbic encephalitis, SCLC without VGKC-ab, or in those with non-paraneoplastic limbic encephalitis and high titres of VGKC-ab. The characteristic immunoreactivity of AGNA was not reported in any patient from the two series of patients with VGKC-associated non-paraneoplastic limbic encephalitis,2,3 although both were published before our description of AGNA. These data suggest that, in the setting of SCLC, the immune mechanisms that trigger AGNA are closely associated with those responsible for the generation of antibodies against neuronal membrane channels.

In conclusion, in patients with a clinical diagnosis of limbic encephalitis and positive VGKC antibodies, irrespective of the titre, the presence of AGNA should be evaluated because its detection must prompt the search for lung cancer. Future studies will confirm the association between AGNA and channel antibodies in patients with paraneoplastic neurological syndromes.

Acknowledgments

We thank Raffaele Pezzani and Mercè Bonastre for their excellent technical assistance, and Dr M D Bonifati (Trento, Italy) and all neurologists of the Second Neurologic Clinic (Padua, Italy) for providing clinical data.

References

View Abstract

Footnotes

  • Funding: L Z is a recipient of a fellowship grant of the European Neurological Society. This study was supported in part by grant PI030028 from the Fondo de Investigaciones Sanitarias, Madrid, Spain (to FG).

  • Competing interests: None.

Request Permissions

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.