• Open Access

Suspected Limbic Encephalitis and Seizure in Cats Associated with Voltage-Gated Potassium Channel (VGKC) Complex Antibody


Corresponding author: Akos Pakozdy, Clinical Department for Companion Animals and Horses, Clinic for Small Animals Internal Medicine, University of Veterinary Medicine, Veterinaerplatz 1, Vienna 1210, Austria; e-mail: akos.pakozdy@vetmeduni.ac.at.



Treatment-resistant complex partial seizures (CPS) with orofacial involvement recently were reported in cats in association with hippocampal pathology. The features had some similarity to those described in humans with limbic encephalitis and voltage-gated potassium channel (VGKC) complex antibody.


The purpose of this pilot study was to evaluate cats with CPS and orofacial involvement for the presence of VGKC-complex antibody.


Client-owned cats with acute orofacial CPS and control cats were investigated.


Prospective study. Serum was collected from 14 cats in the acute stage of the disease and compared with 19 controls. VGKC-complex antibodies were determined by routine immunoprecipitation and by binding to leucine-rich glioma inactivated 1 (LGI1) and contactin-associated protein-like 2 (CASPR2), the 2 main targets of VGKC-complex antibodies in humans.


Five of the 14 affected cats, but none of the 19 controls, had VGKC-complex antibody concentrations above the cut-off concentration (>100 pmol/L) based on control samples and similar to those found in humans. Antibodies in 4 cats were directed against LGI1, and none were directed against CASPR2. Follow-up sera were available for 5 cats in remission and all antibody concentrations were within the reference range.

Conclusion and Clinical Importance

Our study suggests that an autoimmune limbic encephalitis exists in cats and that VGKC-complex/LGI1 antibodies may play a role in this disorder, as they are thought to in humans.


contactin-associated protein-like 2


complex partial seizures




feline CPS with orofacial involvement


glutamate decarboxylase


limbic encephalitis


leucine-rich glioma inactivated 1


magnetic resonance imaging


voltage-gated potassium channel

Limbic encephalitis (LE) associated with serum antibodies against voltage-gated potassium channel complexes (VGKC-complexes) is a relative newly recognized acute epileptic condition in humans that usually occurs in the absence of a neoplastic disorder.[1-3] The antibodies are directed against different proteins of VGKC-complexes, principally leucine-rich glioma inactivated 1 (LGI1) and contactin-associated protein-like 2 (CASPR2).[4, 5] The clinical presentation frequently is acute and includes confusion and cluster seizures involving the facial muscles. Magnetic resonance imaging (MRI) frequently shows high signal intensity in the hippocampal region. Seizures often respond poorly to antiepileptic drugs, but well to immunotherapy.[1, 4-6]

We recently reported an acute seizure disorder with orofacial involvement in cats (feline CPS with orofacial automatism, FEPSO).[7] The clinical presentation included distinctive complex partial cluster seizures, consisting of facial twitching, salivation, motionless staring (motor arrest), lip smacking, chewing, licking, swallowing, mydriasis, and vocalization. Such episodes lasted usually for a few seconds to 1 minute (see supplemental video). Frequent postictal signs included behavioral changes and aggression. Brain MRIs showed bilateral hippocampal T1 hypo- and isointensity and T2 hyperintensity. Initial resistance to antiepileptic treatment was observed. The histopathologic changes were found mainly in the hippocampus and included degeneration and inflammation. We suggested that a primary immune-mediated process may cause limbic encephalitis and that hippocampal neuronal loss may be secondary to inflammation, as also described in humans with LE.[6] The main aim of this study was to determine whether affected cats had increased serum VGKC-complex antibody concentration during the acute stage of the disease, whether the antibodies bound LGI1/CASPR2, and to compare results with follow-up and control samples.

Materials and Methods

Cats from our clinic were prospectively selected for acute cluster onset of CPS with orofacial involvement. Patients were included in the acute stage of the disease within 10 days of the first clinical signs (Group A: FEPSO acute) and after immunosuppressive treatment in partial or full remission (Group B: FEPSO in remission). Group C included adult cats from our clinic with different problems without CPS (Control). Sera were collected from all included cats and stored at −76°C. Statistical analysis was performed by Median test. A P value < .05 was considered significant.

The VGKC-complex antibodies were determined by routine immunoprecipitation with rabbit brain extracts as previously described,[1] and used routinely by clinical laboratories for the diagnosis of LE in humans. Whole rabbit brain membranes were solubilized in digitonin (2%) and VGKC-complexes were labeled with 125 I-dendrotoxin,1 which binds to Kv1 subtypes of VGKCs. Aliquots were incubated at 4°C with 5 μL of the feline serum overnight. The antibody-bound-125I-DTX-VGKC complexes were precipitated by addition of a secondary anticat antibody (DAKO), and the precipitates pelleted, washed, and counted on a gamma counter. The results were expressed as picomoles VGKC-complex/L (pM) after subtraction of the mean results from 3 healthy human controls. The control feline sera precipitated a mean of 1.1 ± 31.1 pM (SD). The cut-off of 100 pM, used also in assays for human, was chosen to represent> mean + 3 SDs of the control feline sera results. VGKC-complex antibody-positive sera also were tested for binding to LGI1 and CASPR2 on cell-based assays[4] in which cells expressing the antigens on their surface are tested for binding of patients’ sera by indirect immunofluorescence. Glutamate decarboxylase (GAD) antibodies were measured by routine radioimmunoprecipitation with a commercial assay.2


Among cats with CPS and orofacial involvement, 14 were sampled in acute stage of the disease (A group), and 5 cats were in remission (B group). In the control group (C) 19 cats were included. Five of the 14 diseased cats in group A had VGKC antibody concentrations above the cut-off concentration for positivity (100 pmol/L), whereas none of the control cats (group C) showed increased VGKC antibody concentrations (Fig 1). The median VGKC concentration was 60 pmol/L (range, 1–344) in group A and 13 pmol/L (range, 1–94) in group C, which differed significantly (Median test, P = .001). Four of the 5 VGKC-complex antibody-positive cats had LGI1 antibodies and none had CASPR2 antibodies. No difference could be found in clinical signs associated with the seizures between VGKC-complex antibody-positive and -negative cats. Three cats were euthanized during the acute stage of the disease because of the owners’ request. After antiepileptic, supportive, and corticosteroid treatment of the remaining cats, 10 experienced full remission (seizure-free) and 1 cat had partial remission (occasionally recurring seizures). Histopathologic examination was available in only 1 cat and changes were consistent with hippocampal necrosis.[7, 8] Serum samples at follow-up (group B, 5 cats) were within the reference range (median concentration, 51 pmol/L; range, 21–92) 1–12 months after the initial measurement. Antibodies to GAD were not detected in any sera, test, or control.

Figure 1.

Voltage-gated potassium channel-complex (VGKC-complex) antibodies in cats with complex partial seizures and orofacial involvement (FEPSO) in the acute stage of the disease (14), in remission (5) and in control cats (19).


In our prospective study of cats with naturally occurring acute onset of CPS with orofacial involvement (FEPSO), 5 of 14 (36%) showed VGKC-complex antibody concentrations that were greater than the reference value for human (>100 pmol/L), whereas results in the control cats were all below this value. Importantly, 4 of the 5 cats had antibodies directed against LGI1, which is the main identified component of the VGKC-complex antibodies. Thus, this condition in many respects appears to be similar to LE in humans with VGKC-complex/LGI1 antibodies.

The main clinical features of LE in humans are subacute short-term memory loss, complex partial seizures, and psychiatric clinical signs. However, such clinical signs are less well established in cats, although the clinical presentation has obvious similarities: acute or subacute signs indicating limbic system disease including behavioral changes (fear and aggression) as well as CPS with orofacial involvement. An association between facial seizures and presence of VGKC-complex/LGI1 antibody recently was reported in adult human patients with facial and ipsilateral arm seizures.[6] These cats could have had a similar syndrome with an interspecies difference in clinical manifestation as seizures of the ipsilateral front limbs were not observed. Unfortunately, the association between ictal clinical signs and limbic localization could not be confirmed by electroencephalography (EEG) as EEG was not performed, but,[9] examining similar complex partial orofacial seizures in cats, localized the epileptic activity to the temporal-parietal region by interictal EEG[8].

In our previous report, we concluded that complex partial cluster seizures with orofacial involvement in cats often are associated with hippocampal pathology without other structural brain disease.[7] We also demonstrated that MRI is useful in the antemortem diagnosis and that T2-weighted images may reveal hyperintense lesions restricted to the hippocampus and other extrahippocampal regions, such as the piriform lobe. Human patients commonly exhibit increased MRI signal intensity in the mesial temporal lobes with VGKC-complex antibody-associated limbic encephalitis, particularly on T2 or fluid-attenuated inversion recovery (FLAIR) sequences. MRI abnormalities in the temporal lobes can be seen in human patients and in dogs with seizures, regardless of the cause, suggesting that they can be the consequence of recurrent seizures. Our results indicating an association between hippocampal inflammatory and degenerative changes and VGKC-complex/LGI1 antibodies support an underlying autoimmune reaction in both humans and cats.

Recent investigations have shown that most of the antibodies thought previously to be directed against VGKCs in humans are, in fact, binding to other cell membrane proteins, particularly LGI1, CASPR2, and contactin 2 that are components of VGKC complexes.[10] The presence of LGI1 antibodies in the VGKC-complex antibody-positive cats, therefore, strongly supports a similar etiology as observed in the human disease. However, 2 important questions must be considered. First, Is the presence of antibodies specific to the neurologic disorder? and secondly, Are these antibodies pathogenic? Antibodies to VGKC-complexes were not detected in any of the control cats including 3 with brain tumors. The presence of autoantibodies does not appear to be a nonspecific inflammatory response because other antibodies such as those against GAD, an intracellular protein also associated with LE in humans,[10] were not detected in these cats. Although very few studies have been reported, the major pathophysiologic mechanism is likely to be interference with the expression or function of VGKCs as a result of antibodies binding to LGI1 with which they are closely associated in the neuronal membrane. This would lead to increased neurotransmitter release with resulting neurologic deficits and seizures.[10]

The most important finding of our study is that autoimmune limbic encephalitis likely exists in cats and that the VGKC-complex antibody may play an important etiologic role in seizure diseases in this species. We found that 36% of cats with acute onset CPS with orofacial involvement showed increased concentrations of IgG against VGKC-complexes in the acute stage of their disease, and the clinical course and follow-up examination strongly suggest that immune-mediated limbic encephalitis has a fair prognosis in cats. Further investigations with larger case recruitment and complete diagnostic evaluation including brain high-field MRI, CSF analysis, EEG, histology, and follow-up examinations including cats with different CNS disorders and different seizure types are necessary to validate our conclusion and to explore further similarities with the human disease.


This study was not supported by a grant.

Conflict of Interest: Authors disclose no conflict of interest.


  1. 1

    Perkin-Elmer, Boston, MA

  2. 2

    RSR Ltd, Cardiff, UK