A 6-Year-Old Girl With Fever, Weakness, and Ataxia

A previously healthy 6-year-old female presented to the ED in January 2021 with fever, slurred speech, and right-sided weakness and being off balance on her right side; these symptoms began the night before ED arrival. On the day of arrival, she had the interval development of headache and 1 episode of nonbloody, nonbilious emesis.

The patient was seen ∼10 days prior in an urgent care for left periorbital swelling and conjunctival redness, without painful extraocular movements or drainage from the eye. She was treated with a 7-day course of oral amoxicillin and clavulanate and a topical antibiotic ointment, with resolution of swelling and conjunctival redness.

There was no associated neck pain, dysphagia, or choking. There was no known injury, ingestion, recent travel, or arthropod exposure. The family recently added a new kitten to the household. The patient was fully immunized. Family history was noncontributory.

In the ED, she was afebrile and not hypoxemic; her heart rate was 116 per minute, respirations 24 per minute, and blood pressure was 104/55 mm Hg. She was mildly toxic appearing, fussy with examination, but alert and active with normal mentation. Neurologic examination was remarkable for mildly slurred speech, mild right facial droop in upper motor neuron pattern, mild tongue deviation to left, and flaccid tone with 4 of 5 strength testing in right upper and lower extremities. Left sided upper and lower extremity tone and strength were normal. Reflexes were 2+ and equal bilaterally in upper extremities. Left patellar reflex was 2+ and right patellar reflex was 3+. Right finger-to-nose dysmetria was present with an ataxic gait. She had mild left conjunctival injection without periorbital swelling. She did not have any adenopathy in preauricular or cervical areas. There were no signs of nuchal rigidity.

When seeing a patient in the ED with this presentation, what is your initial differential diagnoses and management strategy, Dr Little-Wienert?

In a patient with focal findings on neurologic examination, I become worried about a space-occupying lesion in the brain and/or stroke. Given the acute presentation, history of fever, and recent presentation that could have been consistent with an orbital cellulitis, my concern for an infectious process such as an intracranial abscess was higher than other diagnoses on my differential such as hemorrhagic or ischemic stroke, intracranial malignancy, aneurysm, or intracranial hemorrhage. For all of these diagnoses, neuroimaging should be obtained. As hemorrhagic strokes and intracranial hemorrhages can cause rapid decompensation and require timely intervention, it is important to first obtain the most readily available head imaging, which is the computed tomography (CT) scan at our institution. If the diagnosis is not discovered on head CT, then MRI of the head should be obtained to evaluate for the other diagnoses. In addition, a contrast-enhanced head CT is better for visualizing abscesses. Laboratory tests should be sent to evaluate for bleeding disorders, anemia, metabolic disorders, and inflammation or infection. Timely initiation of antibiotics to cover for intracranial pathogens is also important.

At this point, the differential diagnosis remained broad. Given fever after a recent periorbital cellulitis, infectious complications were considered: cerebral abscess, meningitis, encephalitis (herpes simplex virus), and empyema. For example, her constellation of symptoms with fever, headache, and focal neurologic deficits can be seen in complicated sinusitis without the characteristic preceding features of acute bacterial sinusitis.¹ Additionally, fungal central nervous system (CNS) infections and parasitic CNS infections are included on the differential.² 

Certainly “cannot miss” diagnoses such as stroke or neoplasm must be considered. The differential diagnosis of ischemic stroke in children is displayed in Table 1. One diagnosis in particular, diabetic ketoacidosis, can present with cerebral infarction and cerebral edema.³ We also consider stroke mimickers, primarily nonvascular etiologies that present similarly to a stroke. One study of pediatric patients in the ED identified the most common mimic diagnoses to be migraine, seizures, Bell’s palsy, conversion disorder, and syncope.⁴ Parinaud’s oculoglandular syndrome and Bartonella henselae have been reported to be associated with facial nerve paresis resulting from compression of the facial nerve by an enlarged preauricular lymph node or parotid gland.⁵ Other mimickers include hypoglycemia, postictal paresis, and acute disseminated encephalomyelitis.² In distinguishing these diagnoses, neuroimaging can be very helpful.

In the ED, initial laboratory testing was significant for leukocytosis to 16 × 10*3/uL with 79% neutrophilic predominance and an absolute neutrophil count of 13.26 × 10*3/uL. Inflammatory markers were elevated with erythrocyte sedimentation rate at 49 mm per hr (normal <20) and C-reactive protein 4.9 mg/dL (normal <1.0). Blood cultures were collected. Neurology recommended the most readily available imaging of the head; computed tomography (CT), CT angiography, and computed tomography venography head and neck with and without contrast were ordered. These revealed no acute intracranial process and specifically no evidence of intracranial abscess. The CTs demonstrated questionable left periorbital soft tissue swelling but no evidence of orbital cellulitis or abscess.

The MRI performed with the stroke protocol revealed a 9 mm acute lacunar infarct in the left thalamus and posterior limb of internal capsule, 5 mm acute lacunar infarct in posterior limb of the right internal capsule, and 6 mm lesion in the right middle cerebellar peduncle, possibly representing a subacute infarct. Initially, the patient was started on vancomycin and ceftriaxone; acyclovir was added for empirical herpes simplex virus meningoencephalitis coverage. She was admitted to the ICU for further evaluation and management.

As a Neuroradiologist, Dr Kralik, what is your interpretation of the initial imaging findings obtained in the ED?

The head CT demonstrated no abnormalities, which is commonly the situation with acute stroke or meningitis. Head CT’s detection of acute intracranial infarction can include loss of cortical gray-white matter differentiation or obscured deep gray matter nuclei in the first 3 hours in ∼75% of patients and edema at 12 to 24 hours at the site of infarction. Small infarcts, less than 1 cm in size, can be difficult to appreciate, even past 24 hours on head CT. A head CT is useful for excluding the presence of acute hemorrhage or a mass. CT angiography demonstrated no arterial occlusion, stenosis, irregularity, or aneurysm. At this point, an MRI head would be useful to determine if stroke, meningitis, or some alternate process is present.

The initial MRI obtained was an MRI performed with the stroke protocol, designed to be a shorter MRI primarily focused on detecting a stroke. This MRI demonstrated: (1) A 0.9 cm focus of restricted diffusion on diffusion-weighted imaging (DWI) involving both the inferolateral left thalamus and posterior limb of the internal capsule with associated larger 2.5 cm region of fluid-attenuated inversion recovery (FLAIR) hyperintensity; (2) A 5 mm focus of restricted diffusion in the posterior limb of the right internal capsule with minimal corresponding T2/FLAIR hyperintensity; (3) A 0.6 cm focus of T2/FLAIR hyperintensity without definitive signal abnormality on DWI. Hyperacute infarcts can demonstrate diffusion restriction without FLAIR signal abnormality. Late acute and early subacute infarcts can demonstrate a combination of diffusion restriction with FLAIR hyperintensity, and subacute or chronic infarcts can demonstrate FLAIR hyperintensity without diffusion restriction. At this point, in conjunction with the clinical history, the findings would be most concerning for multiple small vessel infarcts, with some in the acute and subacute phases, which can occur with coagulopathy, diabetic ketoacidosis, vasculitis, and basilar meningitis. Less commonly, demyelinating lesions can have a mixed appearance on DWI and FLAIR that can mimic infarcts.

With these imaging findings, our differential expanded to consider causes of infarction: rheumatologic (autoimmune vasculitis), cardiac (endocarditis with septic embolization), and hematologic etiologies (hypercoagulable state). Infectious etiologies previously discussed remained on the differential including meningitis, with consideration of tuberculosis given possible basilar meningitis seen on MRI.

From your perspective as Pediatric Neurologists, Drs Yarimi and Shukla, how does a stroke typically present in the pediatric patient, and what are the typical etiologies?

A stroke, though rare in children, can present in a myriad of ways and presentation varies depending on age. In the neonatal population, seizure is the most common presenting symptom. For older children, seizure can also be a presenting sign, but focal neurologic deficits like hemiparesis, dysarthria, aphasia, numbness, ataxia, and cranial nerve dysfunction are more common. Nonspecific signs, including cardio-pulmonary dysfunction, headache, nausea, vomiting, and fever, can present with or without a focal neurologic finding.⁶ 

Etiologies for acute ischemic stroke include heart disease (congenital or acquired), vasculopathies (including dissections), postviral arteriopathies, congenital anomalies, and progressive arteriopathy. Other causes could be related to systemic vasculitis, hemoglobinopathies, coagulopathies, and metabolic disorders. Approximately 30% of children with a presentation of acute ischemic stroke may not have any identifiable risk factor at time of stroke. On the other hand, children may often have multiple etiologies related to acute ischemic stroke.⁷ 

Drs Yarimi and Shukla, what are the important next steps in the workup?

Given the MRI findings of multiple lacunar infarcts, evaluation for underlying risk factors for strokes would be the next step. Dedicated vessel imaging of the head and neck with either magnetic resonance angiography or CT angiography would be indicated to look for any vascular abnormalities. Magnetic resonance venography is indicated to evaluate for venous thrombosis. An echocardiogram is indicated to assess for cardiac risk factors, including intracardiac shunts, other structural heart disease, and presence of thrombus. Peripheral ultrasounds to evaluate for thrombi or clots should also be performed. A hemoglobin analysis, hypercoagulability workup, systemic infectious work-up, and metabolic work-up should all be obtained.

This patient’s critical condition necessitated a prompt and decisive diagnostic assessment. Priority lies in infectious and hematologic studies with vessel and cardiac imaging for potential urgent interventions. Ancillary studies follow based on ischemic patterns, vasculopathy, or cardiac anomalies. Thus, a negative D-Dimer without an intracardiac shunt would deprioritize peripheral ultrasounds.

During admission, the patient completed a full infectious workup, including a lumbar puncture. Cerebrospinal fluid (CSF) analysis was remarkable only for mildly elevated white blood cells 25, red blood cells 2, glucose 61, and protein 23. The CSF Gram stain was negative for organisms and the CSF culture remained negative. Viral studies, including herpes simplex virus, enterovirus, Epstein-Barr virus, cytomegalovirus, adenovirus, varicella zoster virus, human herpesvirus type 6, and arbovirus panels were all negative. Bartonella titers were negative. Tuberculosis testing with a QuantiFERON gold test was negative. Blood cultures also remained negative. Ultimately, an infectious etiology was thought to be less likely. Oligoclonal bands and ancillary autoimmune testing were not sent for at the time of the initial lumbar puncture.

In cases of pediatric stroke, Dr Lee-Kim, what are the common causes of hypercoagulability?

In pediatric arterial ischemic stroke, it is important to determine whether there is a cardioembolic source. Inherited hypercoagulable conditions, in order of decreasing prevalence, include Factor V Leiden mutation, Factor II G20210A mutation, deficiency of protein C or protein S, and antithrombin deficiency. Other thrombophilias include antiphospholipid antibodies, hyperhomocysteinemia, and elevated lipoprotein(a). Patients with 1 or more of these hypercoagulable conditions may have an increased risk for thrombosis, including stroke.^8,9 Additionally, patients with sickle cell disease or other inflammatory disorders may have increased coagulation factor levels, leading to a hypercoagulable state.

To evaluate for emboli as an etiology, an echocardiogram was performed, which revealed a patent foramen ovale with left to right shunting. No structural heart disease was detected, and biventricular systolic function was normal. No vegetations were visualized. There was no evidence of deep venous thrombosis on the Doppler ultrasounds of all extremities. The patient also had a complete dilated ophthalmologic exam performed by Ophthalmology, which was unremarkable and did not show any evidence of retinal emboli or vasculitis.

Careful consideration was given for the patent foramen ovale on echocardiogram. The shunting, as mentioned, was left to right across the atrial septum. However, as her brain lesions appeared symmetric and there were no peripheral or cardiac sources of thrombi identified, an embolic stroke was thought to be less likely. Similarly, with negative blood cultures and no evidence of endocarditis on echocardiogram, there was low suspicion for infarction caused by septic emboli.

Magnetic resonance venography head was obtained and negative for cerebral venous sinus thrombosis. One of her laboratories, partial thromboplastin time, was prolonged at 36.0 seconds. Additional testing revealed mildly low protein C, positive lupus anticoagulant, and elevated anticardiolipin antibody. She was negative for factor V Leiden and prothrombin gene mutations. She had normal homocysteine, antithrombin, and protein S levels. She had normal levels of factors 8, 9, and 11.

What is your impression of these results, Dr Lee-Kim? Could these results explain the patient’s presentation?

Antiphospholipid antibodies (eg, lupus anticoagulant, anticardiolipin antibody, β-2 glycoprotein antibodies) in children can be transient, usually associated with inflammation and may resolve spontaneously. If an antiphospholipid antibody is detected, it is recommended that patients have testing repeated a minimum of 12 weeks later to document persistence of the antiphospholipid antibodies to meet criteria for antiphospholipid antibody syndrome. If the rest of the clinical picture supports a prothrombotic state, the presence of antiphospholipid antibodies in the context of an acute thrombosis should raise concern for catastrophic antiphospholipid antibody syndrome, for which antithrombotic therapy could be lifesaving. A low protein C level can be inherited or acquired secondary to acute thrombosis, liver disease, disseminated intravascular coagulation, sepsis, or vitamin K deficiency.

Although antiphospholipid antibody syndrome or protein C deficiency can cause a hypercoagulable state, further evaluation and future repeat testing may be necessary to determine whether these contributed to or resulted from the illness and thrombotic event.

Serum and CSF lactate testing were within normal limits, reassuring against a metabolic or mitochondrial disorder. What other rare etiologies should be considered in the case of pediatric stroke, Dr Chun? Could this be caused by vasculitis?

Pediatric stroke is such a rare entity, that it always grabs our attention and warrants a rheumatology consultation if no hypercoagulable risk factor, infectious etiology, connective tissue disorder, or metabolic process is identified.

An estimated 40% to 60% of pediatric arterial ischemic strokes are thought to be related to CNS vasculitis.¹⁰ From a rheumatologic standpoint, several primary and secondary vasculitides can present as a stroke. When it is isolated to the brain, childhood primary angiitis of the CNS can involve any size cerebral vessel, with large and medium-sized vessels often identified on imaging (MRI or magnetic resonance angiography, angiography), but small vessel involvement often requires a biopsy to confirm. CNS vasculitis can also be a manifestation of a systemic vasculitis, such as Takayasu’s arteritis, Polyarteritis Nodosa, and antineutrophil cytoplasmic antibodies (ANCA)-associated vasculitis (granulomatous polyangiitis, microscopic polyangiitis). However, these patients often have more systemic features, such as fever, nonblanching rash, arthritis, upper or lower airway disease, hypertension, kidney involvement, peripheral neuropathies, and weight loss, so it is important to take a good history, including a thorough family history.

Venous and arterial thromboses can be secondary to antiphospholipid antibodies, which can be a manifestation of lupus. Other rare systemic autoimmune diseases that can present with cerebrovascular insults include Behcet’s, which classically is associated with sinus venous thrombosis, and sarcoidosis, which can rarely present with cerebrovascular manifestations.

Lastly, a very young presentation or a history of recurrent strokes should prompt genetic evaluation for deficiency of ADA2, which is a recently described monogenic autoinflammatory disease associated with vasculopathy, intermittent fevers, and rash.¹⁰ 

This patient did not have any red flags on her history or physical examination to suggest multiorgan involvement. However, her young age at presentation warranted thorough laboratory evaluation.

Additional laboratories sent evaluated for underlying vasculitis: myeloperoxidase antibody, serine proteinase 3 antibody, and ANCA were all negative. There was no deficiency in plasma ADA2 activity. A serum test evaluating for causes of autoimmune CNS disorders, “pediatric autoimmune encephalopathy and CNS disorder evaluation,” was recommended by Neurology once it was recognized that this presentation was not typical of infectious meningoencephalitis. This was sent to an outside hospital and remained pending during admission with an 8 to 12 day turn-around.

During the admission, the patient had interval improvement in her strength; however, repeat head MRI was concerning for worsening of her acute process. This imaging was included with a sedated MRI of the spine to probe for any changes that might offer diagnostic clues. Imaging revealed a new T2/FLAIR hyperintense lesion at the left middle cerebellar peduncle. She subsequently had an angiogram performed with interventional radiology to evaluate for vasculitis; no signs of vasculitis were detected. Ultimately, given continued suspicions for a CNS autoimmune etiology, the decision was made by Neurology to start steroid treatment with a taper over 6 weeks.

The patient’s strength continued to improve over admission. She was discharged from the hospital to continue the steroid taper and daily aspirin. Although no final diagnosis was known at the time of discharge, several laboratories were pending. After discharge, the serum PCDES test resulted with a serum myelin oligodendrocyte glycoprotein (MOG) IgG positive, with a titer of 1:1000 (reference <1:20).

MOG is a glycoprotein produced by the oligodendrocyte and is found on the oligodendrocyte surface membrane. It functions in the formation, maintenance, and disintegration of the myelin sheaths.¹¹ Detection of this autoantibody suggests a steroid-responsive inflammatory CNS demyelinating disorder that is primarily targeting the myelin of the oligodendrocyte. The presentation of MOG antibody-associated disease is variable based on age. In patients less than 11 years of age, it is more likely to manifest as an acute disseminated encephalomyelitis (ADEM)-like phenotype. This may present clinically as an encephalopathy, with multifocal neurologic deficits and bilateral supratentorial cerebral lesions. In adolescents and adults, focal syndromes such as optic neuritis are more commonly seen.^11,12 

Drs. Yarimi and Shukla, would you consider this patient’s clinical presentation to be within the spectrum of MOG antibody-associated disease? Does the diagnosis explain the infarcts seen on MRI? What is the expected course and prognosis?

MOG antibody-associated demyelinating disease is a newly described autoimmune disorder that can present with both monophasic and multiphasic demyelinating syndromes in both adults and children. Typical presentations include acute disseminated encephalomyelitis (ADEM), optic neuritis, ADEM-optic neuritis, relapsing optic neuritis, multiphasic ADEM, and neuromyelitis optica spectrum disorder. More recently, MOG antibodies have also been described in patients who present with a clinical picture of autoimmune encephalitis and, in rarer cases, with leukodystrophy-like presentations. MOG antibodies have also been found to rarely present with a CNS vasculitis-like picture,¹² as in our patient. Testing for MOG antibodies should be performed in all patients who present with neurologic changes and abnormal MRI given the wide spectrum of presentations.¹³ MOG antibody-associated demyelinating disease is typically treated with high-dose steroids and intravenous immunoglobulin in the acute period (with more severe cases requiring use of plasmapheresis). Typically, patients do not need further immunotherapy outside of the acute period; however, there is a small percentage that do, and this group may require more than 1 repeat course. The use of intravenous (IV) methylprednisolone is usually effective.¹⁴ In patients with multiphasic disease, the use of additional immunomodulatory treatments can be considered.¹⁵ 

When testing for MOG antibody-associated disease, it is important to be aware of the caveats in interpreting results. First, testing with MOG-IgG is the gold standard test. Additionally, use of cell-based assays is optimal, as the protein-based enzyme-linked immunosorbent assay testing tends to have unreliable results. Each patient’s clinical context must be taken into consideration when receiving a positive result.¹⁶ 

Our patient’s testing was performed with a fluorescence activated cell sorting assay (FACS) and identified the IgG antibody with a titer of 1:1000. Studies have shown high titers to be more predictive of MOG antibody-associated disease. One study in particular of 1260 clinical samples with 92 positive MOG-IgG results showed no false positives in patients with titers ≥1:1000.^16,17 

Our patient continued follow up with Neurology and Hematology in the outpatient clinics after discharge. Follow up testing in Hematology revealed normalized levels of protein C and antiphospholipid antibodies (lupus anticoagulant and cardiolipin antibody). As mentioned earlier, antiphospholipid antibodies may be positive after viral infections or an acute stroke,¹⁸ so repeat testing 12 weeks after a positive test is advised. Therefore, in our case, no hypercoagulable condition was identified as a potential contributing factor.

The Neurologist recommended repeat imaging with an MRI head with and without contrast 2 months after discharge. It showed interval decrease in the lesions and no evidence of new lesions. Our patient’s weakness is now resolved after the interventions, which is consistent with the typical course seen in children. The prognosis is good, and the majority of children (80%) experience a monophasic disease.^11,12 At this point in the knowledge of this autoimmune disorder, in most cases, there does not appear to be permanent deficits from the illness.¹³ 

ADEM

acute disseminated encephalomyelitis

ANCA

anti-neutrophil cytoplasmic antibodies

CNS

central nervous system

CSF

cerebrospinal fluid

CT

computed tomography

DWI

diffusion-weighted imaging

ED

emergency department

FLAIR

Fluid-attenuated inversion recovery

MRI

magnetic resonance imaging

MOG

myelin oligodendrocyte glycoprotein

MOGAD

MOG antibody-associated demyelinating disease