Primary Angiitis of the Central Nervous System with Pathological Findings of Eosinophilic Granulomatosis with Polyangiitis

Abstract

A 79-year-old woman presented with difficulty walking and disturbance of consciousness. Magnetic resonance imaging revealed diffuse white matter lesions and abnormal signals along the surface of the brain and sulci. A brain biopsy revealed granulomatous vasculitis with eosinophil infiltration. There was no peripheral blood eosinophilia or evidence of angiitis in other organs, and primary angiitis of the central nervous system (PACNS) with pathological findings of eosinophilic granulomatosis with polyangiitis (EGPA) was diagnosed. Steroids and other immunosuppressant therapies showed only limited effects. PACNS with pathological findings of EGPA is extremely rare, and a prompt brain biopsy is necessary for a diagnosis.

Introduction

Primary angiitis of the central nervous system (PACNS) is an immune-mediated vasculitis localized in the brain and spinal cord arteries (1-3). Clinical manifestations are variable and nonspecific (e.g., headache, altered cognition, focal neurological deficits). Histologically, the inflammation affects both the large and small blood vessels in the central nervous system (CNS). As a result, vascular stenosis leads to ischemic lesions, and vessel wall weakening leads to vascular disruption and intracranial hemorrhaging (2). The etiology and pathogenesis of PACNS remain unclear.

Eosinophilic granulomatosis with polyangiitis (EGPA) is a disease that causes vasculitis in the CNS. EGPA is an anti-neutrophil cytoplasmic antibody (ANCA)-associated angiitis that is characteristically associated with bronchial asthma and eosinophilia (4,5). CNS involvement in EGPA is rare, appearing in only approximately 5% of all patients (6,7).

We herein report a rare case of PACNS with pathological findings of EGPA.

Case Report

A 79-year-old woman was admitted to our hospital with difficulty walking and disturbance of consciousness. Seventeen months before admission, the patient had developed a right frontal subcortical hemorrhaging. She had been treated conservatively, and mild left hemiparesis remained as a sequela. Seven months later, she experienced generalized tonic seizures, and antiepileptic drug administration was initiated. Subsequently, the cognitive impairment and gait disturbance gradually worsened. She had a history of hypertension but had no history of bronchial asthma or allergic rhinitis.

A neurological examination revealed mild disturbance of consciousness (Glasgow Coma Scale 13, E4V3M6), left-sided hemispatial neglect, mild hemiparesis, and frontal signs. On a general physical examination, no fever, skin lesions, weight loss, or gastrointestinal symptoms were noted.

Brain magnetic resonance imaging (MRI) showed hyperintense signals along the brain surface, sulci, and left occipital lobe on fluid-attenuated inversion recovery (FLAIR) imaging (Fig. 1a, b). Diffusion-weighted imaging (DWI) revealed multifocal high-intensity signals on the brain surface with mild contrast enhancement (Fig. 1c, d). The apparent diffusion coefficient map showed hypoattenuation corresponding to high-intensity signals on DWI. Magnetic resonance angiography revealed no findings suggestive of angiitis.

Figure 1.

(a-d) Brain magnetic resonance imaging (MRI) findings on admission. Fluid-attenuated inversion recovery (FLAIR) imaging revealed hyperintense signals along the brain surface, sulci, and left occipital lobe (a, b) (arrows). Diffusion-weighted imaging (DWI) showed multifocal high-intensity signals on the brain surface (c) (arrows), with mild contrast enhancement (d) (arrows). (e) Brain MRI at the onset of subcortical hemorrhaging (17 months before admission). FLAIR imaging showed hyperintense signals on the brain surface and sulci, probably due to hematoma (arrows). (f) Brain MRI at 10 months before admission. FLAIR imaging showed hyperintense signals on the surfaces of the bilateral occipital lobes (arrows).

On previous imaging tests, brain MRI at the onset of subcortical hemorrhaging (17 months before admission) had shown high-intensity signals on the brain surface and sulci on FLAIR, probably due to hematoma (Fig. 1e). T2* weighted imaging had shown low-intensity signals corresponding to high-intensity signals on FLAIR but no low-intensity signals within the brain parenchyma, suggesting microbleeds. Brain MRI at the onset of seizures (10 months before admission) had shown high-intensity signals on the surfaces of the bilateral occipital lobes on FLAIR (Fig. 1f).

A cerebrospinal fluid (CSF) examination revealed an elevated protein concentration (240.3 mg/dL), a cell count of 3 /μL (100% lymphocytes), and normal glucose (56 mg/dL). Bacterial and mycobacterial cultures were all negative. CSF cytology revealed a few lymphocytes and eosinophils, and no malignant cells were detected.

Blood tests revealed a normal cell count, including 2.7% eosinophils. The C-reactive protein level was slightly increased (0.76 mg/dL; normal range 0.00-0.14 mg/dL). The total immunoglobulin (Ig) E level was normal (155 IU/mL; normal range 0.0-232.0 IU/mL). The serum IgG4 level was normal (61.8 IU/mL; normal range 11.0-121.0 IU/mL). Rheumatological studies were all negative, including antinuclear antibody, ANCA, rheumatoid factor, anti-cyclic citrullinated peptide antibody, and anti-DNA antibody tests.

Chest and abdominal computed tomography revealed no abnormalities. The fecal occult blood test results were negative. Nerve conduction studies were unremarkable, and no findings suggested systemic vasculitis.

A brain biopsy was performed on the right frontal lobe. The biopsy showed predominantly eosinophilic inflammatory cell infiltration of the small cortical vessels, a small number of multinucleated giant cells, mild fibrinoid necrosis, and vascular stenosis (Fig. 2). Direct fast scarlet staining did not reveal any amyloid deposits. Although the brain biopsy showed findings consistent with EGPA, this case had no evidence of systemic vasculitis and did not meet the criteria for EGPA. Therefore, this case was diagnosed as PACNS based on the pathological findings of EGPA.

Figure 2.

Histological findings from a brain biopsy (Hematoxylin and Eosin staining). The brain biopsy showed granulomatous changes (a) (arrow), predominantly eosinophilic inflammatory cell infiltration of the small cortical vessels (b) (arrowheads), and a small number of multinucleated giant cells (arrow).

After admission, the patient's consciousness gradually deteriorated, and MRI revealed aggravation, such as the expansion of the T2 high-intensity area in the brain parenchyma. After a brain biopsy, steroid pulse therapy (methylprednisolone 1,000 mg/day for 3 days) followed by oral prednisolone (25 mg/day) was started. After the start of treatment, MRI showed a reduction in high-intensity signals on the brain surface and sulci on FLAIR; however, the patient's symptom improvement remained mild. Intravenous immunoglobulin therapy (20 g/day for 5 days) and high-dose intravenous cyclophosphamide therapy (500 mg monthly for 6 months) were administered for further improvement, but no changes were observed. The patient was transferred to a long-term care hospital approximately eight months after admission.

Discussion

PACNS is a rare inflammatory disease that affects only the CNS. The major symptoms are headache, altered cognition, and focal neurologic deficits; however, the clinical symptoms are highly variable and nonspecific (2,3). Diagnostic criteria for PACNS include the following:

1. the presence of an unexplained neurologic deficit after thorough clinical and laboratory evaluation,

2. documentation by cerebral angiography and/or a tissue examination of an arteritic process within the central nervous system, and

3. no evidence of systemic vasculitides or any other condition to which the angiographic or pathologic features could be secondary (1).

There are no characteristic findings, and the diagnosis is mainly made by finding evidence of vasculitis on a biopsy or angiography and excluding secondary CNS vasculitis (3).

EGPA, formerly known as Churg-Strauss syndrome, is a systemic vasculitis that involves small- to medium-sized arteries. The diagnosis of EGPA is based on a combination of clinical features (obstructive airway disease, nasal polyps, mononeuritis multiplex), eosinophilia, and extra-vascular eosinophilic-predominant inflammation on a biopsy (5). ANCA is associated with approximately 30-60% of cases diagnosed with EGPA (8,9). CNS involvement is recognized in approximately 5% of all cases (6,7).

In the present case, findings consistent with EGPA were found only on a brain biopsy, and EGPA was not diagnosed. Amyloid-β-related angiitis can cause eosinophilic vasculitis that is localized in the central nervous system (10). However, a brain biopsy in our case showed no amyloid deposits on the blood vessel walls, and brain MRI showed no microbleeds suggestive of cerebral amyloid angiopathy. Parasitic infection and drug-induced vasculitis can also cause eosinophilic vasculitis in the central nervous system; however, these diseases usually involve other organs. Therefore, PACNS with pathological findings of EGPA was diagnosed based on the clinical symptoms and pathological findings.

Eosinophilic vasculitis confined to the CNS is extremely rare. Indeed, to our knowledge, only three cases have been reported (Table) (11-13). As with other subtypes of PACNS, the clinical symptoms of eosinophilic vasculitis are variable and nonspecific, as are the lesion sites seen on MRI. A CSF analysis revealed abnormally, elevated protein levels in all cases, and mild pleocytosis was observed in one case (12). A brain biopsy revealed eosinophilic vasculitis in all cases, but a granulomatous changes were found in only one case (13). One of the three cases showed a good course with high-dose steroids alone (11), but the other two cases also required immunoglobulin, biologics, immunosuppressants, and steroids (12,13). In the present case, high-dose steroids, immunoglobulin, and cyclophosphamide were administered, but the response was poor. One reason for the poor treatment response may be the extended period from the onset to the start of treatment.

Table.

Summary of Previous Reports.

Reference	Age	Sex	Clinical symptoms	Lesion site on MRI	ANCA	CSF analysis	Brain biopsy	Treatment	Response
(11)	39	F	Confusional state headache left hemiparesis	Cerebral cortex leptomeninges	-	Normal cell count Elevated protein level	Eosinophilic vasculitis without granulomas	CCSs	Good
(12)	51	M	Memory decline	Cerebral cortex basal ganglia leptomeninges	-	Mild pleocytosis Elevated protein level	Eosinophilic vasculitis without granulomas	CCSs, IVIG, IVCY, RTX, TOC, MPZ	Fair
(13)	73	M	Conscious disturbance seizure	White matter	-	Normal cell count Elevated protein level Elevated IgG index	Eosinophilic vasculitis granuloma	CCSs, IVCY, IVIG, AZT	Good
This case	79	F	Conscious disturbance Gait disturbance	Cerebral cortex Leptomeninges White matter	-	Normal cell count Elevated protein level	Eosinophilic vasculitis granuloma	CCSs, IVIG, IVCY	Poor

MRI: magnetic resonance imaging, ANCA: anti-neutrophil cytoplasmic antibody, CSF: cerebrospinal fluid, IgG: immunoglobulin G, CCSs: corticosteroids, IVIG: intravenous immunoglobulin, IVCY: intravenous cyclophosphamide, RTX: rituximab, TOC: tocilizumab, MPZ: mepolizumab, AZT: azathioprine

In the present case, the patient had a history of subcortical hemorrhaging 17 months before admission. The major causes of subcortical hemorrhaging in the elderly are hypertension and amyloid angiopathy (14). Although this patient had a history of hypertension, it was well controlled, and a brain biopsy showed no amyloid deposits. Cerebral arteriovenous malformations (AVMs) can also cause subcortical hemorrhaging, but they rarely occur in elderly individuals (15). In addition, the imaging study showed no vascular abnormalities suggestive of an AVM. In retrospect, high-intensity FLAIR signals along the brain surfaces were seen seven months after the onset of the subcortical hemorrhaging. These findings suggests that the subcortical hemorrhaging in this case was caused by PACNS. Such high-intensity signals on FLAIR are difficult to detect because of hematoma immediately after the onset of subcortical hemorrhaging. Therefore, as in this case, patients with subcortical hemorrhaging of unknown cause may require multiple MRI follow-up examinations for the early diagnosis and treatment of the causative disease.

PACNS with pathological findings of EGPA is extremely rare and difficult to diagnose without evidence of vasculitis on imaging studies, as there are no specific clinical manifestations or biomarkers. Treatment based on systemic EGPA may be effective and should be initiated as soon as possible. A brain biopsy should be encouraged for the early diagnosis and treatment of similar cases. Therefore, it is necessary to accumulate similar cases to clarify the pathological mechanisms and develop effective treatments.

Informed consent was obtained from the guardian.

The authors state that they have no Conflict of Interest (COI).