Genetic or Autoimmune: POLG-Related Epilepsy Initially Treated as an Autoimmune Encephalitis, a Case Report

Abstract

Hospital neurologists participate at the forefront of managing fulminant acute and subacute onset epilepsy, frequently attributed to autoimmune encephalitis (AE). As the recognition of antibody-mediated AE grows, there is a growing number of patients who are treated as antibody-negative AE. While antibody-negative autoimmune processes should be considered in the setting of acute and subacute onset of fulminant epilepsy, other causes must be considered before subjecting patients to long-term immunomodulatory treatments and other potential therapeutic toxicities. We present the case of a previously healthy young man who presented with new-onset refractory seizures treated with escalating doses of anti-epileptic drugs as well as immunosuppression for presumed autoimmune epilepsy. He developed valproic acid induced hepatotoxicity requiring liver transplantation and was later found to have a POLG mutation. We discuss the presentation of POLG mutations as well as the diagnosis of seronegative autoimmune encephalitis. We highlight the need for a broad differential when evaluating new onset refractory seizures in an otherwise healthy person.

Introduction

Hospital neurologists are at the forefront of managing fulminant acute and subacute onset epilepsy, frequently attributed to autoimmune encephalitis (AE). As the recognition of antibody-mediated AE grows,^1-3 an increasing number of patients are treated as antibody-negative AE or are treated prior to receiving results from antibody testing.

There are obvious, and under appropriate circumstances acceptable, risks inherent in treating with immune therapies prior to a confirmed diagnosis. However, the increased awareness of autoimmune epilepsies warrants a discussion highlighting alternative etiologies as well. Herein, we present a case of refractory epilepsy initially presumed to be autoimmune that was later determined to be genetic in nature.

Case Presentation

A 20-year-old man presented to a local hospital after falling and hitting his head. Prior to this, he had not had any diagnosed medical conditions and denied feeling unwell. Due to high clinical concern for seizure, levetiracetam 500 mg every morning and 1000 mg at night was initiated. He returned 2 weeks later after a witnessed generalized tonic clonic seizure. An MRI brain demonstrated slightly expansile T2/FLAIR hyperintense lesions in the right frontoparietal lobe (Figure 1). He was ultimately transferred to a regional academic institution due to suspected persistent ictal activity. Upon arrival he had left focal myoclonic jerks. He was loaded with fosphenytoin 20 phenytoin equivalents(pe)/kg then started on maintenance of phenytoin 150 mg 3 times per day.

Figure 1.

Cortically-based slightly expansile T2/FLAIR hyperintense lesions in the right frontoparietal lobe.

His subsequent workup included MR venogram, CT angiogram, and conventional cerebral angiography without evidence of vasculitis. The PET CT brain demonstrated multiple areas of increased metabolic activity (right posterior parietal lobe, right thalamus, right frontal lobe, and left cerebellum). An electroencephalogram (EEG) obtained revealed findings considered consistent with myoclonic status epilepticus, and lumbar puncture (LP) revealed elevated protein with no other abnormalities (Table 1). He was then treated empirically with plasma exchange. Due to continued seizure activity, he then received IVIG 2 gram/kg over 5 days with limited clinical improvement. This was followed by methylprednisolone, cyclophosphamide, and rituximab. His seizure control improved, and it was concluded that cyclophosphamide and steroids were effective. He discharged with a working diagnosis of antibody-negative AE.

Table 1.

Lumbar Puncture Results.

	Glucose (mg/dL)	Protein (mg/dL)	Red blood cell count/UL	White blood cell count/UL	IgG index, CSF	Color	Special tests
1st lumbar puncture	48	131	0	1•		Colorless	**
2nd lumbar puncture	89	95	1	4•	0.74	Colorless
3rd lumbar puncture	48	242	1	1•			**

+Cell counts listed are from tube 4 unless otherwise stated.

•Lymphocytic predominance.

** Negative studies included the Mayo Epilepsy-autoimmune evaluation panel.

He returned to the same institution 1 week later with recurrent focal seizures. A repeat lumbar puncture revealed isolated protein elevation (Table 1). He received another course of high-dose methylprednisolone followed by a second dose of rituximab. His seizures were felt to respond, and at discharge he was prescribed levetiracetam 2000 mg twice a day, fosphenytoin 150 mg 3 times a day, lacosamide 100 mg twice a day, mycophenolate mofetil 500 mg twice a day, and prednisone 60 mg daily.

One month later, his fosphenytoin was replaced with valproic acid and levocarnitine by his local neurologist. Approximately 3 months after starting the valproic acid, he presented to the regional academic center with 2 weeks of nausea and vomiting, jaundice, and mild asterixis. His labs on presentation included INR 3.1, ALT 269, AST 144, total bilirubin of 6, and normal alkaline phosphatase. Potentially hepatotoxic agents, including mycophenolate mofetil and valproic acid, were discontinued. He was placed on prolonged EEG and remained seizure-free on clonazepam 0.5 mg twice a day in addition to levetiracetam and lacosamide.

Workup for acute liver injury, including a comprehensive infectious workup, were negative aside from a transjugular liver biopsy demonstrated microvesicular fatty change most consistent with drug-induced liver injury (DILI) attributed to valproic acid. This medication was discontinued at admission, and his liver transaminases trended down during the hospital course. At discharge, the mycophenolate mofetil was increased to 1000 mg twice a day, and his antiepileptic regimen was adjusted to twice daily dosing of levetiracetam 2000 mg, lacosamide 100 mg, and topiramate 150 mg. He was readmitted 2 days later with worsening transaminase elevation and encephalopathy. A subsequent liver ultrasound demonstrated moderate steatosis and slow flow in the portal vein. At this point, the patient was transferred to a tertiary center for liver transplant evaluation. His lacosamide was changed to pregabalin, along with levetiracetam and topiramate. A third LP and an MRI were obtained. The latter demonstrated resolution of the prior T2 signal changes.

Given his age, MRI abnormalities, seizures, and hepatopathy, a medical genetics consultation was recommended to evaluate for the possibility of mitochondrial or metabolic disorders. A three-generation pedigree was obtained and the family history was deemed negative for pertinent heritable disorders.

The patient underwent an orthotopic liver transplantation and his post-transplant course was complicated by focal seizures, ultimately controlled after resuming therapy with lacosamide 100 mg twice a day. Genetic testing demonstrated a POLG mutation, and prednisone and mycophenolate mofetil were discontinued. His seizures remain well controlled on the current regimen.

Discussion

The differential for new onset refractory seizures in a young, previously healthy male is broad and includes infectious, autoimmune, and genetic/inherited etiologies.

There are several infectious etiologies that can present with seizures. Up to 50% of patients with HSV encephalitis and 10-74% of patients with bacterial infections have symptomatic seizures as part of their clinical course.⁴ As our patient did not present with fever or mental status changes, an infectious etiology was considered unlikely. New-onset refractory status epilepticus (NORSE), febrile infection-related epilepsy syndrome (FIRES), and acute encephalitis with refractory repetitive partial seizures (AERRPS) are syndromes associated with a postinfectious presentation. They commonly have associated MRI or CSF changes.⁵ While our patient did demonstrate signal changes on T2/FLAIR MRI sequences, his CSF studies were only notable for elevated protein.

Seizures can also be a manifestation of AE. These encephalitides are classically accompanied by identifiable and etiologic autoantibodies identifiable in CSF and/or serum. Features that suggest a diagnosis of autoimmune epilepsy include: onset with status epilepticus, early pharmacoresistance, onset after 30 years of age, cognitive impairment or rapid decline, known associated autoimmune disorder, and mesial temporal T2-signal abnormalities on brain MRI, especially if these are bilateral.^5,6

POLG Mutation

The POLG gene is responsible for encoding mitochondrial DNA polymerase. With more than 150 identified pathogenic POLG mutations, a variety of phenotypic syndromes may occur, including Alpers-Huttenlocher syndrome (seizures, cortical blindness, and developmental regression in childhood),⁷ progressive external ophthalmoplegia (PEO), SANDO (sensory ataxic neuropathy with dysarthria and ophthalmoparesis), mitochondrial epilepsy with ragged red fibers,⁸ and mitochondrial spinocerebellar ataxia-epilepsy syndrome.⁹

POLG-related pathophysiology typically presents at approximately 18 years of (range: infancy to 36 years), often with refractory focal clonic motor seizures, myoclonus, epilepsia partialis continua, or status epilepticus. Additional comorbid features include cognitive impairment, ataxia, headache, or peripheral neuropathy. Our patient did not have these additional features as part of his presentation. Seizures among this cohort have a propensity to originate from the occipital or temporo-occipital region, commonly from the right hemisphere with left-sided clinical seizures.¹⁰ One proposed mechanism of neurological injury is “mitochondrial activity depletion,” caused by a mismatch in neuronal energy demand and mitochondrial ATP synthesis. The occipital cortex is among the most [metabolically] active cerebral regions and is vulnerable to this type of injury. MRI may demonstrate T2/FLAIR hyperintensities in the thalamic, occipital, and cerebellar regions without a clear ictal association. EEG findings are variable--showing occipital or posterior interictal activity, diffuse slowing, normal background, or a generalized epileptiform pattern.

Importantly, POLG mutations are associated with rapidly progressive valproate-induced hepatotoxicity with subsequent liver failure within months. Pathologic findings of valproic acid-induced hepatotoxicity include micronodular cirrhosis, decreased hepatocytes, microvesicular fat, bile ductular proliferation, and pleomorphic mitochondria on pathology. Screening for POLG mutations is recommended in children or adolescents with intractable seizures before starting valproic acid.¹¹

The more challenging scenario is a patient with suspected AE without any identifiable antibody, as was initially presumed with this case. For “autoantibody-negative but probable autoimmune encephalitis,” diagnosis requires all 4 of the following criteria¹²

1. Rapid progression (less than 3 months) of working memory deficits (short-term memory loss), altered mental status, or psychiatric symptoms.

2. Exclusion of well-defined syndromes of autoimmune encephalitis (eg, typical limbic encephalitis, Bickerstaff’s brainstem encephalitis, acute disseminated encephalomyelitis).

3. Absence of well-characterized autoantibodies in serum and CSF, and at least 2 of the following criteria: MRI abnormalities suggestive of autoimmune encephalitis; pleocytosis, CSF-specific oligoclonal bands or elevated CSF IgG index, or both; brain biopsy showing inflammatory infiltrates and excluding other disorders (eg, tumor).

4. Reasonable exclusion of alternative causes.

On reflection, our patient did not have a clear encephalitis syndrome as he presented with new onset seizures in the absence of mental status changes, cognitive decline, movement disorders, or psychiatric symptoms. His initial imaging (MRI, PET CT) changes could be attributed to ictal findings and his CSF did not reveal underlying inflammation: it was lacking pleocytosis and demonstrated no unique oligoclonal bands. Furthermore, initial AEDs as well as aggressive immunomodulatory treatments were ineffective; in the setting of an autoimmune process, one would expect that with immunosuppression his seizure activity would improve.

In the adult population, genetic or congenital causes of epilepsy are perceived as uncommon. This sets up potential bias in the initial evaluation of new onset medically resistant epilepsy. In the absence of clear findings to suggest definite autoimmune mediated cause, we suggest broadening the differential to include genetic causes, especially when evaluating young adults. Genetic causes of epilepsy can present from infancy to late adulthood, though it is becoming increasingly clear that the genetics are quite complex.¹³ Though POLG mutations are just one of many, mitochondrial disorders, adrenoleukodystrophies, and progressive myoclonic epilepsies can also present in the same age group.¹⁴

Though important to identify autoimmune epilepsy syndromes early, we suggest maintaining a careful differential and considering the aforementioned criteria for diagnosing antibody-negative disease in young-adults with new onset refractory epilepsy. With mounting evidence for early and aggressive immune therapy,¹⁵ there can be a pressure to diagnose (and treat) quickly. In the absence of a well-defined clinical syndrome or clear markers suggesting CNS inflammation, we recommend considering genetic epilepsy syndromes on the differential patient patients with sudden onset refractory seizures.