Acute cerebral atrophy in autoimmune encephalitis complicated by haemophagocytic lymphohistiocytosis

Abstract

Autoimmune encephalitis is a disease characterised by neural-specific antibodies. This case report presents a 20-year-old young man with a recent history of suspected viral encephalitis who presented with recurrent fevers and episodes of confusion. He was found to have anti-N-methyl-D-aspartate receptor (NMDAR) and α-amino-3-hydroxy-5-methyl-4-isoxazolepropionic acid 1 receptor (AMPAR1) positive autoantibodies and was diagnosed with autoimmune encephalitis. He subsequently developed global cerebral atrophy and was found to meet diagnostic criteria for haemophagocytic lymphohistiocytosis (HLH). This patient’s presentation was consistent with existing literature showing that autoimmune encephalitis may develop after an initial viral meningoencephalitis. However, concurrent anti-NMDAR and anti-AMPAR1 positive autoimmune encephalitis has not been reported in literature to date, and this case report represents one instance of its presentation. We speculate that multiple antibodies against neural surface antigens may increase the risk for systemic immune activation leading to HLH and acute cerebral atrophy.

Background

Autoimmune encephalitis is an inflammatory brain disorder associated with neural-specific autoantibodies. The first case of autoimmune encephalitis was described in the 60s,¹ but it was only within the past two decades that the disease became increasingly recognised as a clinical entity.² It is now known that several subtypes of neural-specific autoantibodies exist that target either intracellular or extracellular neural antigens. The most common of these is anti-N-methyl-D-aspartate receptor (NMDAR),² which affects the mesial temporal lobes and results in confusion, memory loss, psychosis and seizure after a viral-like prodrome.³ MRI findings are usually unremarkable.⁴

Here, we present a case of a young man diagnosed with NMDAR and α-amino-3-hydroxy-5-methyl-4-isoxazolepropionic acid 1 receptor (AMPA1R) encephalitis who developed cerebral atrophy over a span of 1 month and was found on additional diagnostic testing to meet criteria for haemophagocytic lymphohistiocytosis (HLH).

Case presentation

A 20-year-old young man in China presented with 2 months of fever and 3 days of altered mental status, headache, photophobia and urinary incontinence.

Two months prior, he was admitted to a regional hospital for fever and headache. During that admission, MRI showed diffuse meningeal thickening with bilateral swelling of the caudate and globus pallidus along with hyperintensity in the right basal ganglia (figure 1A). Cerebrospinal fluid (CSF) showed mildly elevated protein and glucose. CSF viral PCR studies for herpes simplex virus (HSV), Epstein-Bar virus, cytomegalovirus and HIV were negative and serum and CSF neural-specific antibodies were not detected. Given his MRI findings and CSF studies, he was empirically treated with acyclovir for presumed viral encephalitis and discharged.

Figure 1.

Fluid-attenuated inversion MRI from the patient (A) at the previous hospitalisation 2 months prior showing bilateral swelling of the caudate and globus pallidus along with hyperintensity in the right basal ganglia, (B) on hospital day 1 showing attenuation of the previous swelling and hyperintensitiy, and (C) on hospital day 36 showing diffuse cerebral atrophy.

Since discharge, he had intermittent fevers that remitted with antipyretics. Three days prior to the current admission, he was found by family confused and speaking nonsensical words. A temperature taken at home showed 38.3 °C and he was brought to the hospital. On admission, he was confused and disoriented and his speech was garbled and incomprehensible. His temperature was 39.0 °C. The rest of his vitals were within normal limits. On review of systems, his family endorsed that he has had poor energy and reduced appetite leading to weight loss of 15 kg over 2 months. He also recently developed constipation (last bowel movement was 4 days prior) and urinary incontinence. His medical, family and social histories were non-contributory. A full neurological examination could not be performed due to lack of corporation.

Investigations

Complete blood count (CBC) and complete metabolic panel showed leucocytosis (13.86×10⁹/L) and mild transaminitis but were otherwise normal. CSF showed mildly elevated protein (0.54 g/L) with normal glucose and no cells. A serum and CSF viral PCR and parasite study were negative. CSF neural-specific antibodies detection via cell-based assay found antibodies against NMDAR (titre, 1:10) and AMPA1R (titre, 1:3.2); serum also showed presence of anti-NMDAR (1:32) but anti-AMPA1R was undetected. MRI showed attenuation of the meningeal thickening and caudate and globus pallidus swelling that was seen in the previous MRI (figure 1B).

On hospital day 3, the patient became comatose (Glasgow Coma Scale, GCS E4/V2/M4) with involuntary eye movements, increased muscular tone and respiratory depression. Arterial blood gas showed an oxygen saturation of 70 mm Hg. He was intubated for airway protection and given diazepam, midazolam and sodium valproate. An electroencephalography obtained hours later showed no seizure activity. He was extubated 2 weeks later, at which point his level of consciousness had improved (GCS E4/V3/M6). Repeat CBC at this point (hospital day 17) showed severe leucopenia (1.6 10³/μL with 52.5% neutrophils) and anaemia (haemoglobin 7.4 mg/L). A bone marrow biopsy was obtained which showed evidence of haemophagocytosis (figure 2). He was further found to have hyperferritinaemia (1671.9 ng/mL), hypertriglyceridaemia (422 mg/dL, fasting) and hypofibrinogenaemia (46.6 mg/dL). He was diagnosed with HLH based on fulfilling 5 out of 8 diagnostic criteria for HLH and a calculated H-score of 186, suggesting a 70%–80% chance of having HLH.⁵ Natural killer (NK) cell activity and soluble interleukin 2 receptor were not available at our institution. An MRI obtained on hospital day 36 showed severe brain atrophy including widening of sulci, bilateral temporal pole atrophy and ventricular enlargement (figure 1C).

Figure 2.

Histology of bone marrow showed haemophagocytosis (phagocytosis by a histiocyte of an erythrocyte), one of the diagnostic criteria for haemophagocytic lymphohistiocytosis.

Differential diagnosis

Our primary diagnosis was autoimmune encephalitis complicated by HLH leading to global cerebral atrophy. As further discussed in the Discussion section, detection of anti-NMDAR in both serum and CSF and anti-AMPAR in the CSF by cell-based assay strongly pointed in this direction. Our differential also included viral encephalitis; however, given the high sensitivity of DNA PCR of CSF samples for HSV,⁶ this was lower on the differential. Acute cerebral atrophy may further be explained by prolonged hypoxia, but the patient’s oxygen saturation was consistently greater than 90%, and the arterial blood gas showed only mild decrease in oxygen partial pressure, making this less likely the case.

Treatment

He was administered with two rounds of intravenous immunoglobulins (500 mg/kg/day for 5 days) and 5 days of high-dose methylprednisolone to treat both HLH and his underlying autoimmune encephalitis. He was also given seizure prophylaxis with antiseizure medications.

Outcome and follow-up

By the time of discharge 2 months after admission, his white cell count had improved (14.75 10³/μL with 72.9% neutrophils) and his haemoglobin was trending up (8.4 mg/L). On examination, he had regained the ability to follow simple commands but was still confused and at times inattentive. His cranial nerves were intact. Sensory examination was normal. Motor examination was remarkable for diffuse weakness and muscle atrophy. At 6-month follow-up, his MRI (not shown) did not show any improvement and he remained cognitively impaired, scoring 8/30 on the Montreal Cognitive Assessment with difficulties in orientation, spatial reasoning and executive function, short-term recall, attention, language comprehension and abstract reasoning. One year later, he remained clinically unchanged.

Discussion

This case report describes a young male patient with anti-NMDAR and anti-AMPAR1 positive autoimmune encephalitis, detected 2 months after an initial presentation that resembled viral encephalitis but with negative viral PCR. The finding of anti-NMDAR in both the serum and CSF in our patient is highly specific for neural antibody-mediated encephalitis.⁷ Anti-NMDAR encephalitis may develop following a viral encephalitis,⁸ although the exact mechanism—if it is due to molecular mimicry or a more general immune activation leading to autoimmunity—is unknown.

Anti-AMPAR encephalitis is poorly characterised due to its rarity and diverse clinical presentation. Available evidence suggest that it is associated with thymomas and malignancies of the lung, breast and thymus.⁹ No evidence of cancer was found in our patient. Interestingly, our patient’s CSF had a relatively high anti-AMPAR1 titre, which was undetected in the serum. An estimated 6% of healthy and disease controls may have a positive serum anti-AMPAR1 using cell-based assay.⁷ Because cell-based assay of CSF is thought to have a higher sensitivity than that of serum,⁹ it is unlikely that this patient’s anti-AMPAR1 level was a false positive and suggests intrathecal production of autoantibodies.

It is difficult to attribute brain atrophy to a single disease process in this patient. Global cerebral atrophy associated with anti-NMDAR encephalitis has been reported in the literature; however, in these cases, atrophy developed over a period of 8–12 months,¹⁰ not 1 month. Acute cerebral atrophy has also been reported in anti-AMPAR encephalitis.^{11 12} Lastly, HLH, a disorder of systemic macrophage and T-cell activation with multisystem involvement including the nervous system, can also result in irreversible neuronal damage.¹³ As far as we know, patients with both anti-NMDAR and anti-AMPAR1 have not been reported in the literature. We may speculate that the presence of antibodies against two neural surface antigens in this adult patient increased the risk of systemic immune activation, but more systematic studies should be conducted to draw stronger conclusions.

Learning points.

• Autoimmune encephalitis may develop after viral meningoencephalitis. Consider this alternative diagnosis in a patient who does not improve with antivirals.

• Although rare, acute cerebral atrophy is possible in autoimmune encephalitis. Consider pursuing repeat imaging on patients with a prolonged course.

• Haemophagocytic lymphohistiocytosis, a condition of systemic immune activation, may be a rare complication of autoimmune encephalitis. We speculate that this may be more common with patients with multiple antibodies against neural surface antigens, such as anti-NMDAR and anti-AMPAR1.