Atypical Presentation of Seronegative Autoimmune Encephalitis as Refractory Status Epilepticus: A Case Report

Introduction

Autoimmune encephalitis (AIE) is an inflammatory brain disorder [1]. Recognition of AIE has significantly increased over the past decade, particularly in patients with unexplained neuropsychiatric presentation, and comprises almost 20% of all cases of encephalitis [2]. There are many etiologies of AIE, including idiopathic, post-infectious reactions, paraneoplastic, immunodeficiency, and genetic factors [3]. Clinical features of autoimmune encephalitis vary widely and can include behavioral changes, cognitive deficits, memory loss, and alterations in mental status. In severe cases, it can lead to life-threatening complications such as encephalopathy, status epilepticus, and coma [4]. Encephalitis can also involve fever, seizures, and sudden onset of focal neurological deficits [1].

AIE has recently been found to be closely associated with antibodies targeting neuronal cell surface proteins and ion channels within neurological tissues [4]. Detecting neuronal autoantibodies in cerebrospinal fluid (CSF) and serum is crucial for diagnosing AIE [1]. However, many AIE cases have negative neuronal autoantibody results, even when there is clear evidence of an immune-mediated disorder. Consequently, antibody testing is not definitive for diagnosing AIE [5]. The presence of neuronal autoantibodies categorizes AIE into 2 categories: seronegative and seropositive. Seropositive AIE includes common types such as N-methyl-D-aspartate receptor (NMDAR) encephalitis and leucine-rich glioma-inactivated 1 (LGI1) encephalitis [4].

Encephalitis is frequently assumed to be of infectious origin, as it is the most common cause. There are similarities in the clinical, laboratory, and imaging findings of all causes of encephalitis, AIE has variable clinical presentations based on causative antibodies, and patients can be autoantibody-seronegative, leading to missed or delayed diagnosis of AIE [1,4]. Therefore, diagnosing AIE can be challenging, necessitating heightened clinical suspicion to make the correct diagnosis [4]. A case report from Matera et al presented the case of a 16-year-old girl with 3 episodes of tonic-clonic seizures with loss of consciousness. She was initially diagnosed as having epilepsy. Nevertheless, with symptoms progression and no improvement, further investigations were pursued. Her brain MRI was unremarkable, but cerebrospinal fluid analysis and serum revealed positive anti-NMDAr antibody. After receiving intravenous immunoglobulin, steroids and undergoing plasma exchange, her symptoms improved markedly [6], showing the importance of detecting encephalitis early, as it can initially mimic primary epilepsy.

While Matera et al’s case presented seropositive encephalitis with anti-NMDAR, the present case report focuses on the diagnostic and therapeutic challenges posed by seronegative autoimmune encephalitis, showing the importance of early clinical suspicion and treatment of seronegative autoimmune encephalitis in the setting of absent or pending neuronal autoantibodies. Marked clinical improvement was seen when our patient received corticosteroids and plasmapheresis, after receiving a variety of antibiotics with no improvement. Seronegative autoimmune encephalitis diagnosis is often delayed or missed due to overlap with infectious encephalitis, different symptoms and clinical presentation, and unavailability of or delay in autoantibody testing. Our report shows the importance of clinical diagnostic criteria ing guiding management, especially in the absence of autoantibodies. We believe that our report makes a significant contribution to the literature because it adds to the existing evidence that treatment of seronegative autoimmune encephalitis should be considered, especially when infectious causes are excluded. The use of autoimmune encephalitis criteria as a practical tool can lead to early detection of the disease and early initiation of empirical immunotherapy, even without a confirmatory autoantibody test. The diagnostic criteria proposed by Graus et al (2016) feature 3 different diagnostic levels: possible, probable, and definite AIE. Of particular importance, the “possible AIE” category considers the patient’s clinical presentation and neuroimaging and CSF findings in the initiation of immunotherapy in seronegative cases [5].

Case Report

A 27-year-old man with no significant medical history presented with an abrupt onset of intermittent fever exceeding 38°C and severe generalized headache lasting for 1 week. The fever had progressively worsened, with episodes partially alleviated by acetaminophen, associated with chills, increased sleepiness, and loss of consciousness for a few minutes occurring only once. He subsequently experienced a generalized tonic-clonic seizure characterized by upward eye movement and tongue biting. Since the seizure, he had remained drowsy, although the exact duration of the seizure was unknown. There were no associated bladder or bowel disturbances. He had no history of epilepsy, head trauma, substance use, contact with sick people, significant weight changes, or night sweats. Other systemic reviews were unremarkable. He was admitted to a local hospital and was electively intubated because he developed status epilepticus, with a Glasgow Coma Scale (GCS) of 3 out of 15. He was sedated with midazolam and propofol infusion. He was started on antiepileptic therapy with levetiracetam, lacosamide, and valproic acid. He was suspected to have meningoencephalitis, for which intravenous vancomycin, meropenem, acyclovir, and dexamethasone were given.

After 10 days in the hospital without improvement, and challenges with extubation and tapering off anti-seizure medications, he was referred to our Intensive Care Unit at Prince Mohammed Bin Abdulaziz National Guard Hospital in Al Madinah for further evaluation and management. Upon admission, he continued treatment with valproic acid, levetiracetam, lacosamide, acyclovir, meropenem, and vancomycin. However, due to the onset of renal failure, vancomycin was changed to linezolid. During the neurological examination, the power, reflexes, and sensation could not be assessed as he was sedated, with a GCS score of 3 out of 15. The fundoscopic examination showed a healthy optic disc with a cup-to-disc ratio (CDR) of 0.5, no evidence of papilledema, a flat retina, a dry macula, and normal retinal vasculature. The remainder of the physical examination revealed no significant findings.

A comprehensive diagnostic laboratory investigation was initiated, which demonstrated mild elevation of white blood cells WBC (11.8×109/L), normocytic hypochromic anemia with low hemoglobin level (10.5 g/L), and normal platelet count. In addition, the inflammatory markers erythrocyte sedimentation rate (ESR) and C-reactive protein (CRP) were elevated (93 mm/h and 27 mg/L, respectively). Other initial laboratory revealed electrolyte disturbance, with low potassium, magnesium, and phosphorus levels. The kidney function test, blood gases, and coagulation profile were within normal limits. However, there was a mild elevation of aspartate aminotransferase (AST) and alanine aminotransferase (ALT), mostly secondary to the aggressive anti-seizure medications, which normalized before the next repeated tests. Blood, urine, and respiratory samples were collected to identify potential infectious agents, but all cultures showed no growth. Additionally, a metabolic panel of thyroid-stimulating hormone (TSH), B12 levels, glucose, and a urine drug screen returned normal results. A brain CT scan was performed and revealed ill-defined bilateral cerebral and brain stem hypodensities involving the grey and white matter. These findings were suggestive of cerebritis, and encephalitis accompanied by cerebral edema.

Throughout his hospital stay, and involving the neurology team, the patient had a lumbar puncture with cerebrospinal fluid (CSF) analyses, electroencephalography (EEG), and brain magnetic resonance imaging (MRI). The initial CSF analysis showed colorless fluid, with mild elevation in glucose at 4.55 mmol/L and neutrophil cell count 30, while protein levels remained normal at 0.33 g/L (Table 1). Microbiological studies of cerebrospinal fluid and serum were conducted, yielding unremarkable results, except for the detection of IgG antibodies for herpes simplex virus (HSV) I/II and Varicella zoster virus (HZV) (Table 2). EEG revealed non-specific encephalopathy with evidence of epileptogenic potentials. The brain MRI showed asymmetrical bilateral cortical areas of abnormal hyperintensities on T2-weighted images involving the temporal and frontal lobes (Figure 1), with cerebellar hemispheres involvement showing bilateral hyperintense signals (Figure 2). Diffusion-weighted imaging (DWI) MRI also showed bilateral areas of restricted diffusion in the temporal and frontal cortex (Figure 3).

The patient was initially treated empirically with antivirals and antibiotics, as he was assumed to have meningoencephalitis. After receiving 18 days of a total 21 planned days of acyclovir and completing the course of antibiotics for a total of 14 days, the patient was still developing fever and super-refractory status epilepticus, which is defined as status epilepticus that continues or recurs 24 hours or more after the onset of anesthetic therapy. Despite aggressive treatment for this patient with 4 anti-epileptics and an anesthesia-induced coma, he had an immediate recurrence of seizure activity with each attempt to reduce sedation. Given the lack of clinical improvement despite appropriate antiviral and antibiotic therapy, along with the absence of a clear cause for viral and bacterial meningoencephalitis, AIE was considered in the differential diagnosis by applying the possible AIE criteria.

To exclude paraneoplastic causes of AIE, a thoracoabdominal pelvic CT scan was performed, revealing no significant abnormalities. Additional workup including autoimmune serology was requested, which returned unremarkable results (Table 3). A lumbar puncture was repeated (Table 1), and neuronal autoantibodies were sent for analysis before initiating treatment of AIE (Table 3).

Because the result of the neuronal autoantibodies might take weeks to return and because negative results do not exclude AIE, he was started on the first-line AIE therapy, which included corticosteroids and plasmapheresis. Our reasoning for therapy initiation was solidified after applying the possible AIE criteria. We found that our patient met multiple points from the criteria: new focal CNS findings, seizure unexplained by a previously known seizure disorder, and MRI findings suggestive of encephalitis. Following the first plasmapheresis session, the patient experienced a fever reduction, and seizure activity was successfully controlled. He was extubated after the second session of plasmapheresis, with the GCS improving to 13 out of 15. After completion of a total of 3 plasmapheresis sessions, he was transferred to the general ward for rehabilitation, speech therapy, and a swallowing assessment.

A brain MRI was repeated 4 weeks after the last plasmapheresis session and showed a significant decrease in previously noted regression of the previous hyperintensities depicted on FLAIR and T2-weighted images (Figure 4), with resolution of the previously noted cerebellar involvement (Figure 5). Brain MRI DWI showed near-complete resolution of the previously noted bilateral frontotemporal diffusion restriction (Figure 6). At 3 weeks after the repeated lumber puncture, the results for autoantibodies returned unremarkable (Table 3). After that, the patient was referred to an outside rehabilitation center and discharged on the anti-seizure medications levetiracetam 1500 mg every12 hours, lacosamide 100 mg every12 hours, and oxcarbazepine 600 mg every 12 hours.

Discussion

AIE is a severe inflammatory brain disorder and is potentially fatal [7], but it is considered one of the most common treatable causes of encephalitis [8]. Seronegative AIE is considered one of the major subtypes of AIE [9]. Existing criteria for AIE depend on the presence of antibody testing, which can delay the diagnosis. To overcome this, Graus et al (2016) concluded that it is not realistic to have antibody testing as a part of early diagnostic criteria for evaluating AIE, for 3 reasons: antibody testing can take many weeks to come back, depending on the hospital’s capabilities; negative antibody results do not exclude the diagnosis of AIE; and a positive result does not confirm the diagnosis [5].

The diagnostic criteria of AIE and its subtypes were proposed by Graus et al in 2016, which focused on clinical neurological assessment and diagnostic tests by brain MRI, EEG, and CSF analysis, and were not dependent on autoantibody status. This resulted in establishing 3 levels of clinical evidence for autoimmune encephalitis: possible and probable, for which the autoantibody status is not needed in most cases; and definite, for which the autoantibody status is often needed [5].

Early consideration of AIE in the differential diagnosis is important for all patients who present with clinical and MRI findings suggestive of encephalitis, regardless of the detection antibodies, because early recognition and initiation of the proper treatment are crucial for good prognosis, reducing the burden of immune-mediated deficits, preventing relapses and persistent disparities in AIE [4].

In the present case, AIE was diagnosed based on the presence of criteria for possible autoimmune encephalitis and antibody-negative probable autoimmune encephalitis criteria, as reported by Graus et al [5]. Our patient was investigated for antibodies to anti-NMDA, anti-RI, and anti-Yo, which all came back negative. Other possible causes were excluded, including infections, hypothyroidism, drug toxicity, and metabolic causes such as hyponatremia. In addition, paraneoplastic syndrome was ruled out through tumor markers and whole-body CT scans.

Besides excluding alternative etiologies, most patients with encephalitis undergo brain MRI with contrast. The findings could be normal or non-specific, but sometimes can suggest an autoimmune cause [5]. Independent of the underlying cause, limbic system involvement is related to immune-mediated encephalitis shown by T2/FLAIR. The cortical and subcortical areas will show hyperintensity in the temporal lobe, with most cases being bilateral and asymmetrical. With immune-mediated encephalitis, most often involving the basal ganglia, contrary to herpes simplex encephalitis [10]. In nuclear medicine, the presence of bilateral temporal hypermetabolism is considered one of the characteristic findings of limbic encephalitis, and may even be evident when brain MRI imaging appears normal [11]. This is supported by one of the largest multimodal neuroimaging studies, by Baumgartner et al, which demonstrated that 18F-FDG-PET imaging has a diagnostic role in detecting autoantibody-negative limbic encephalitis, suggesting unidentified antibodies involving the limbic system [12]. Another common pattern is bilateral occipitoparietal hypometabolism in NMDAR antibody encephalitis [13]. All other MRI patterns (cortical/subcortical, striatal, diencephalic, brainstem, encephalomyelitis, and meningoencephalitis) can support possible or probable AIE unless the autoantibodies panel is positive for a clinically relevant antibody. Diffuse or patchy contrast enhancement suggestive of inflammation was seen in a few patients [5]. Importantly, brain MRI can also help exclude alternative diagnoses such as acute stroke or neoplasm [13].

Our patient’s brain MRI showed bilateral hyperintensity with diffusion restriction involving the temporal, frontal, and parietal lobes, which can be consistent with herpes encephalitis. However, the CSF viral studies were negative. These brain MRI findings suggested encephalitis of immune-mediated origin.

Intensive care unit (ICU) admission is indicated in AIE with respiratory compromise, refractory status epilepticus, and severe dysautonomia. AIE is commonly treated with immunosuppressive therapy. Several retrospective studies have shown the importance of starting immunotherapy once AIE is highly suspected, and infectious etiologies are excluded based on CSF results [13]. This case also demonstrates the challenges in managing super-refractory status epilepticus, as standard antiepileptic therapy and anesthesia-induced coma were insufficient. A case report by Okamoto et al (2025) described a case of dextromethorphan overdose leading to the development of refractory status epilepticus. This patient required intubation with continuous infusion of propofol and midazolam due to the ineffectiveness of standard antiepileptic therapy. The patient’s mental status gradually improved, leading to extubation in day 4 and full recovery by day 9 [14]. Although the etiologies were distinct, both cases highlight the importance of management escalation with intensive care support, anesthesia-induced coma, and or immunotherapy in patients unresponsive to standard antiepileptic treatments.

First-line therapy of AIE includes intravenous immunoglobulins, corticosteroids, or plasmapheresis [3–5,15]. There is no strong evidence of a difference in efficacy between IVIG and plasmapheresis [4]. In the absence of response, second-line therapy should be started with rituximab or cyclophosphamide [3,4,15]. Our patient was initially started empirically on acyclovir with presumed herpes simplex encephalitis as the diagnosis, with poor response to the treatment. Considering AIE in the differential, the patient was started on first-line AIE therapy, including methylprednisolone for 5 days and plasmapheresis for 4 sessions.

This case report has limitations as we could not test for some antibodies because specific essays were not available in our center, such as anti-AMPA, anti-contactin-associated protein-like 2 (CASPR2), anti-glutamic acid decarboxylase (GAD), and anti-GABAa-R and anti-GABAb-R. We did not sample brain tissue, which is considered a criterion for diagnosis of seronegative AIE. More studies of autoimmune encephalitis in our population should be considered to evaluate the incidence of this disorder and improve future diagnosis, treatment, and long-term follow-up.

Conclusions

Patients with autoimmune encephalitis can develop complex neuropsychiatric symptoms. Making the diagnosis of this disorder is challenging due to its complexity, and rarity, as well as the variety of clinical presentations, and autoimmune antibodies causing it. Autoimmune encephalitis should be included in the differential diagnosis of all patients with unclear origin of a rapidly progressive encephalopathy, particularly in the absence of infectious findings on cerebrospinal fluid (CSF) analysis. The diagnosis should not be delayed until receiving the autoantibody panel, as early consideration of and optimizing appropriate immunotherapy can reduce the mortality rate and long-term sequelae and allow a speedy recovery.