ETIOLOGY AND PATHOGENESIS

Multiple factors such as infection (e.g., herpes simplex virus, varicella zoster virus, and human immunodeficiency virus), immune system dysregulation, and genetics have been implicated in its pathogenesis. Other associated conditions include lymphoma and β-amyloid deposition.

CLINICAL PRESENTATION

PACNS can affect patients at any age but is predominant between the fourth and sixth decades of life. Males are affected twice as often as females. The neurologic signs and symptoms are nonspecific and reflect the diffuse and often patchy nature of the cortical dysfunction. The course of the illness progresses over weeks or months, with insidious course between symptom onset and diagnosis of up to 6 months. However, a remitting-relapsing course has been reported with no new or worsening symptoms of disease at intervals of several months to years. The most common findings include chronic headaches (50%-69%), which are indolently progressive; thunderclap headaches should raise suspicion for RCVS. Cognitive impairment (30%-71%) is subtle, and acute change in mental status and consciousness are very uncommon. Persistent focal neurologic deficits or stroke can occur any time (13%-50%) but rarely at the onset of headache. Stroke secondary to PACNS usually involves multiple vessel territories. Transient ischemic attacks (16%-33%), paraparesis (3%-13%), seizures (7%-11%), myelopathy, or cranial nerve involvement can also occur. Constitutional signs and symptoms such as high-grade fever, weight loss, anorexia, weakness, or visceral target organ disease are atypical. PACNS should be highly considered in patients with the following features: (1) cerebral ischemia involving different vascular territories, distributed over time, accompanied by inflammatory changes in CSF analysis; (2) subacute or chronic headache with cognitive impairment or chronic aseptic meningitis; and (3) chronic meningitis after infectious and neoplastic disorders have been ruled out.

CLINICAL SUBTYPES

Granulomatous Angiitis of the CNS (GACNS). This subgroup predominantly affects males with an insidious headache accompanied by diffuse or focal neurologic deficits. Patients have strokelike symptoms without any other etiology of strokes. Signs and symptoms of systemic vasculitides are absent. CSF analysis reveals an aseptic meningitis. MRI typically shows multiple bilateral ischemic foci. The diagnosis is confirmed by finding small to medium-size vessel granulomatous angiitis with Langhans or foreign body giant cells on the brain biopsy. Treatment with a combination of cyclophosphamide and glucocorticoid is usually required.

PACNS with Lymphocytic Infiltrate by Brain Biopsy. This subclass of PACNS presents as clinical, radiologic, and CSF findings that are similar to those of GACNS. However, brain biopsy reveals the presence of lymphocytic angiitis versus the granulomatous lesions seen in GACNS.

Angiographically Defined PACNS. Medium-sized vessels are affected in this subclass of PACNS, which accounts for a higher frequency of abnormal findings of cerebral angiography. Abnormal CSF analysis is usually present. This is a poorly defined entity; thus, other cerebral arteriopathies such as RCVS, infections, and atherosclerosis should be investigated prior to diagnosis and treatment.

Mass-Lesion (ML-PACNS) Presentation. A rare subset of PACNS with a mass-like lesion has been described. The most common features of patients with ML-PACNS are headache (74%), focal neurologic deficit (64%), diffuse neurologic deficit (50%), seizures (47%), nausea and vomiting (21%), and constitutional symptoms (12%). MRI findings reveal a mass lesion. CSF analysis shows a distinctive aseptic meningitis picture in approximately two thirds of cases. Cerebral angiography can be abnormal in more than 50% of these patients with features of a mass effect but without signs of vasculitis. The diagnosis of CNS vasculitis is made by the findings of vasculitic changes on brain pathology after ruling out infections and neoplastic conditions.

Amyloid-β–Related Cerebral Angiitis (ABRA). Development of ABRA is attributed to the presence of amyloid β proteins. A higher percentage of ML-PACNS is seen in the amyloid-related angiitis subset and is associated with a poor outcome.

DIFFERENTIAL DIAGNOSIS

A diverse group of disorders has been described that possess clinical and angiographic features similar to PACNS. Atherosclerosis and thromboembolic disorders are likely if a patient has a history of traditional cardiovascular risk factors such as diabetes, hypertension, and hyperlipidemia. Stroke workup should be obtained. Presence of vessel calcifications in CT scans or carotid ultrasound images are suggestive of atherosclerotic disease. Fibromuscular dysplasia, moyamoya disease, and radiation vasculopathy can cause angiographic abnormalities that can be mistaken for PACNS. However, they have distinguishing extracranial angiographic features that differentiate them from PACNS.

It is important to differentiate PACNS from its close mimic RCVS. Differentiating these two conditions can be complex, yet is crucial because of contrasting therapy and prognosis. RCVS tends to occur in patients between the ages of 20 and 50 and females are affected more often than males. Presentation includes acute-onset severe headaches, often described as “thunderclap” headaches. Other common findings include focal neurologic symptoms (43%), generalized tonic-clonic seizures (17%), ischemic strokes (39%), convexity subarachnoid hemorrhage (34%), lobar hemorrhage (20%), and brain edema (38%). The presence of clinical cofactors (e.g., Valsalva maneuver, recent childbirth, intake of sympathomimetic drugs) can be identified and may be causal. Major ischemic or hemorrhagic stroke, progressive brain edema, and stroke-related death from severe, sustained cerebral vasoconstriction have also been described. The ESR is normal, and CSF analysis is benign. Angiographic finding for RVCS are those of “string of beads” in multiple vascular cerebral beds with reversibility of vascular abnormalities within 3 months. Brain imaging features include brain infarcts and hemorrhages that are typically located in hemispheric “watershed” regions. Convexity subarachnoid hemorrhage is usually minor and found to be more frequent in women and in patients with migraines.

Brain neoplasm (e.g., Hodgkin and non-Hodgkin lymphoma, leukemia, and lung cancers) and demyelinating diseases (e.g., multiple sclerosis and progressive leukoencephalopathy) may present as neuroimaging features similar to those of PACNS.

Systemic diseases and infections should be ruled out. If CNS infection is suspected, laboratory evaluation for specific pathogens should be guided by epidemiologic analysis and exposure risk factors. Etiologic possibilities include human immunodeficiency virus, syphilis, cytomegalovirus, varicella zoster, herpes simplex, hepatitis, tuberculosis, aspergillosis, histoplasmosis, and cysticercosis. Serologic and microbiologic studies should be requested.

Secondary and autoimmune-related vasculitides, such as rheumatologic diseases (e.g., SLE, Sjögren syndrome, scleroderma), systemic vasculitides (e.g., granulomatosis with polyangiitis, microscopic polyangiitis, polyarteritis nodosa, Behçet syndrome, Cogan syndrome), and other autoimmune disease (e.g., Crohn disease, sarcoidosis), can cause CNS involvement and should be ruled out.

DIAGNOSTIC APPROACH

There are no laboratory markers or radiologic features specific for PACNS. Establishing the diagnosis includes multiple laboratory tests to exclude other diagnoses, particularly infection and malignancy.

Laboratory studies should include serologic tests for rheumatologic disorders and other autoimmune conditions to rule out common mimics of PACNS. Elevation of acute phase markers likely indicates underlying infections or systemic vasculitides because acute phase reactants should be normal in PACNS. Blood cultures and molecular testing for pathogens should be requested when appropriate.

CSF analysis should be performed in all patients with PACNS. CSF analysis is abnormal in 80% to 90% of histologically proven cases of PACNS. CSF findings reflect aseptic meningitis, with modest lymphocytic pleocytosis (median CSF white blood cell count of 20 cells/mL), normal glucose, elevated protein concentration (median CSF protein concentration of 120 mg/dL), and occasionally the presence of oligoclonal bands and increased IgG synthesis. CSF testing including cultures, cell cytology, serologic analysis, and polymerase chain reaction can provide crucial information for detection for other PACNS mimics.

NEUROIMAGING MODALITIES

The most frequently used imaging techniques in the evaluation of PACNS include CT, MRI, MRA, and conventional angiography. However, it is important to remember that none of these neuroimaging studies is specific for PACNS. MRI is abnormal in 90% to 100% cases of PACNS. Common locations of brain lesions are located in subcortical white matter, deep gray matter, deep white matter, and the cerebral cortex. Abnormalities include infarctions, which are detected in up to 53% of patients and can involve variable-size vascular territories. Nonspecific lesions in white matter identified with high-intensity T2-weighted MRI with a fluid-attenuated inversion-recovery sequence (FLAIR) are common. Other causes of T2-hypertintense foci should be ruled out (e.g., hypoxic-ischemic changes, age-related changes, migraine, multiple sclerosis, central pontine myelinolysis, metastasis, metabolic changes, eclampsia, and chemoirradiation).

Visualization of cerebral vessels can be obtained indirectly by MRA or directly by conventional angiography. The typical angiographic abnormalities of cerebral angiography in PACNS include alternating areas of irregular “beadings” (stenosis and dilatation), circumferential or eccentric luminal narrowing, and occlusions of one or more arteries and/or vascular mass effect that can involve single or multiple vessels (see Plate 5-60). Cerebral angiography is limited by its ability to demonstrate vasculitic changes in the small vessels (<500 µm in diameter), which contributes to the low sensitivity. Both sensitivity and specificity of cerebral angiography are poor; a negative angiogram does not exclude diagnosis of PACNS, nor is a positive angiogram diagnostic for PACNS. Incorporating the clinical information, CSF findings, and pathologic findings in diagnosing PACNS is important.

BRAIN BIOPSY

Brain biopsy is the gold standard for diagnosis of CNS vasculitis. Although this is an invasive procedure, morbidity has been shown to be relatively low. The classic histologic findings typically involve small to medium-sized arteries with segmental granulomatous lesions composed of multinucleated giant cells, predominantly histiocytic cells with variable number of plasma cells, histiocytes, neutrophils, and eosinophils (see Plate 5-60). False-negative biopsies occur in up to 25% of cases, owing to irregular involvement of brain parenchyma and inaccessibility of the lesion. Open-wedge tissue sampling of a radiologically identified area provides the highest yield. If the lesion is not accessible, it is advisable to sample the tip of the nondominant temporal lobe with overlying leptomeninges and underlying cortex. Brain biopsy is helpful in ruling out PACNS mimics, particularly infections and malignancy.

MANAGEMENT AND TREATMENT

Therapeutic strategies are based on retrospective observations and expert opinions. There are no randomized control studies to support the standard therapies.

Cyclophosphamide, in addition to glucocorticoids, has been the mainstay in the treatment of GACNS. The treatment recommendations for PACNS have been adapted from the treatment algorithm used for systemic small vessel vasculitides. A decision on the use of single or combination therapy is usually based on the acuity and severity of neurologic symptoms. Patients with mild neurologic symptoms can be treated with a single regimen of glucocorticoids. Glucocorticoid treatment is typically initiated to achieve rapid suppression of inflammation. Typical regimens include high oral doses of glucocorticoids (1 mg/kg/day) or intravenous pulse methylprednisolone, 1 g/day for 3 consecutive days followed by prednisone 60 mg/day. High-dose prednisone should be continued for 4 weeks until inflammation is well controlled and then followed by a slow tapering regimen. When combination therapy is required, cyclophosphamide is started at 2 mg/kg/day for 3 to 6 months. Patients should be monitored particularly for leukopenia, infections, and bladder toxicity. After achieving remission, alternative immunosuppressive agents such as azathioprine or mycophenolate mofetil should be used as maintenance therapy to limit the toxicity associated with long-term cyclophos-phamide therapy. Patients should also be given appropriate preventive measures, including prophylaxis for osteoporosis and Pneumocystis jiroveci pneumonia.