Primary angiitis of the central nervous system

1. High-Yield Summary

Primary angiitis of the central nervous system (PACNS) is an exceedingly rare, severe, idiopathic inflammatory arteriopathy strictly confined to the parenchymal and leptomeningeal arterial vasculature of the brain and spinal cord, occurring in the absence of systemic vasculitis or secondary mimics.

• Primary Diagnostic Benchmark: Full-thickness brain and leptomeningeal biopsy remains the histopathological gold standard, demonstrating transmural granulomatous, lymphocytic, or necrotizing inflammatory vascular destruction. High-resolution vessel wall MRI (VWI) and catheter digital subtraction angiography (DSA) serve as pivotal vascular modalities, identifying smooth concentric mural enhancement and alternating segments of stenosis and ectasia.
• The Critical Diagnostic Move: Systematic exclusion of reversible cerebral vasoconstriction syndrome (RCVS), intracranial atherosclerosis, infectious endarteritis (notably varicella-zoster virus and neurosyphilis), and intravascular large B-cell lymphoma via comprehensive cerebrospinal fluid (CSF) analysis, high-resolution vessel wall imaging, and targeted biopsy before initiating cytotoxic immunosuppression.
• Standard Induction Regimen: Immediate combination therapy consisting of pulsed intravenous methylprednisolone (1,000 mg daily for 3 to 5 days) followed by high-dose oral prednisone (1 mg/kg/day) paired with intravenous or oral cyclophosphamide (or rituximab in selected subsets), supported by mandatory Pneumocystis jirovecii prophylaxis, transitioning to maintenance immunosuppression with azathioprine or mycophenolate mofetil for 18 to 24 months.
• Catastrophic Errors to Avoid: Omitting the leptomeninges during diagnostic surgical biopsy, which reduces diagnostic sensitivity from approximately 70% to under 30%; ruling out PACNS based on a normal catheter DSA, as angiography cannot resolve vessels smaller than 200 to 300 micrometers; and mistaking self-limiting RCVS for PACNS, exposing patients with non-inflammatory vasoconstriction to toxic immunosuppressive morbidity.

2. Anatomy

The clinical manifestations and surgical considerations of PACNS are governed by the microvascular architecture of the cerebral circulation, the structural components of the blood-brain barrier, and the anatomical corridors utilized for diagnostic tissue procurement.

Cerebrovascular Microcirculation and Circle of Willis Architecture

The arterial blood supply to the central nervous system is organized into macrovascular conducting conduits and microvascular parenchymal networks:

• Macrovascular Inflow Pathways: The paired internal carotid arteries (anterior circulation) and vertebral arteries uniting into the basilar artery (posterior circulation) converge at the base of the brain to form the Circle of Willis within the basal subarachnoid cisterns. These large-caliber elastic arteries (internal diameter 3 to 5 mm) give rise to the anterior, middle, and posterior cerebral arteries. Primary involvement of these large conducting trunks occurs in approximately 15% to 20% of PACNS patients (large-vessel subtype).
• Pial and Leptomeningeal Vascular Plexus: The major cerebral arteries ramify into medium and small muscular arteries (diameter 200 to 500 micrometers) that traverse the subarachnoid space over the cerebral and cerebellar convexities. These leptomeningeal vessels form an extensive collateral anastomotic meshwork before giving off terminal branches.
• Intraparenchymal Penetrating Arterioles: Small arterioles (diameter 50 to 200 micrometers) penetrate the cerebral cortex perpendicularly from the pial surface. These terminal microvessels lack functional collateral anastomoses, rendering the underlying cortex and subcortical white matter vulnerable to focal ischemic necrosis when transmural inflammatory occlusion develops.

Leptomeningeal Vasculature and Virchow-Robin Perivascular Spaces

The spatial distribution of vascular inflammation in PACNS reflects the anatomical interface between the subarachnoid space and the cerebral parenchyma:

• Leptomeningeal Compartment: The pia mater and arachnoid mater encase the cerebral convexities and enclose the subarachnoid space filled with cerebrospinal fluid. Small muscular arteries within the leptomeninges possess a distinct adventitia and medial muscular layer that are exposed to circulating immune mediators. PACNS displays a marked predilection for these leptomeningeal vessels.
• Virchow-Robin (Perivascular) Spaces: As pial vessels plunge into the cerebral cortex, they carry an invagination of the leptomeningeal sheath, creating a fluid-filled perivascular space. This channel is bounded internally by the vascular basement membrane and externally by the glia limitans superficialis. Inflammatory cuffs composed of lymphocytes and histiocytes accumulate within these spaces, producing the characteristic perivascular T2/FLAIR hyperintensity and contrast enhancement observed on neuroimaging.

Blood-Brain Barrier (BBB) Structural Integrity

The blood-brain barrier represents a specialized neurovascular unit regulating microvascular permeability. Its cellular architecture includes:

• Endothelial Tight Junction Complexes: Non-fenestrated brain capillary endothelial cells are joined by continuous belt-like tight junctions consisting of transmembrane proteins (claudin-5, occludin, junctional adhesion molecules) linked to actin filaments via cytoplasmic scaffolding proteins (zonula occludens ZO-1, ZO-2).
• Perivascular Cellular Sheaths: Pericytes embedded within the continuous capillary basement membrane maintain microvascular tone and endothelial stability. The abluminal boundary is formed by astrocytic end-feet (containing aquaporin-4 channels) that envelope more than 95% of the microvascular circumference.
• Pathophysiological Breakdown: Transmural inflammatory cell invasion disrupts tight junction proteins and degrades the vascular basement membrane through matrix metalloproteinase release. This loss of barrier integrity results in vasogenic edema, extravasation of serum proteins into the neuropil, and parenchymal microhemorrhages.

Safe Neurosurgical Biopsy Corridors and Internervous Planes

Diagnostic open or stereotactic brain biopsy requires meticulous trajectory planning to maximize histopathological yield while safeguarding neurological function:

• Targeted vs. Non-Targeted Selection: When pre-procedural MRI reveals an accessible area of contrast enhancement, restricted diffusion, or cortical edema, targeted biopsy of the lesion yields diagnostic confirmation in 65% to 75% of cases. In patients with diffuse non-focal abnormalities or purely angiographic disease, non-targeted biopsy is directed to non-eloquent cortical regions.
• Non-Dominant Frontal Pole Corridor: The primary non-targeted biopsy site is the non-dominant (typically right) superior or middle frontal gyrus, positioned at least 3 cm anterior to the coronal suture. This corridor avoids the primary motor cortex (precentral gyrus) and supplementary motor areas.
• Non-Dominant Anterior Temporal Pole Corridor: The alternative corridor targets the anterior 2 to 3 cm of the non-dominant middle or inferior temporal gyrus, remaining anterior to the vein of Labbe and well clear of the auditory radiation and visual pathways (Meyer loop).
• Eloquent Danger Zones to Avoid: Surgical sampling is strictly contraindicated in the dominant perisylvian language cortex (Broca area in the pars opercularis/triangularis, Wernicke area in the posterior superior temporal gyrus), primary motor and somatosensory cortices, the visual striate cortex (calcarine sulcus), and the deep diencephalic or brainstem nuclei.

3. Pathogenesis

The pathogenesis of PACNS involves an aberrant, cell-mediated autoimmune cascade directed specifically against structural components of the cerebral and leptomeningeal vessel wall, leading to lumen thrombosis, downstream ischemia, and structural mural failure.

Autoimmune and Cell-Mediated Microvascular Cascade

Although the inciting autoantigens remain unidentified, the inflammatory cascade is orchestrated through defined immunological steps:

• Endothelial Activation and Adhesion: Localized inflammatory stimuli trigger endothelial activation, marked by the upregulation of cell adhesion molecules including vascular cell adhesion molecule-1 (VCAM-1), intercellular adhesion molecule-1 (ICAM-1), and E-selectin.
• T-Cell Extravasation and Infiltration: Circulating CD4+ T helper lymphocytes (predominantly Th1 and Th17 subsets) and CD8+ cytotoxic T lymphocytes adhere to activated endothelium and transmigrate across the vascular basement membrane into the tunica media and adventitia.
• Macrophage Activation and Cytokine Synthesis: Recruited macrophages and histiocytes produce potent pro-inflammatory cytokines including interferon-gamma (IFN-gamma), tumor necrosis factor-alpha (TNF-alpha), interleukin-1 beta (IL-1beta), and interleukin-6 (IL-6). These mediators perpetuate local vascular inflammation and induce smooth muscle cell apoptosis.
• Proteolytic Degradation: Activated macrophages and vascular smooth muscle cells secrete matrix metalloproteinases (notably MMP-2 and MMP-9), which cleave type IV collagen and elastin. This enzymatic action fragments the internal elastic lamina, undermining the structural stability of the arterial wall.

Histopathologic Patterns of Transmural Infiltration

Neuropathological evaluation identifies three distinct histopathological patterns of vessel wall infiltration in PACNS:

• Granulomatous Pattern (58% of Cases): The most common and classical histological presentation. Characterized by dense transmural mononuclear infiltration containing multinucleated giant cells (both Langhans and foreign-body types), epithelioid histiocytes, and activated lymphocytes. Granulomas are centered within the media and adventitia with extensive destruction of the internal elastic lamina.
• Lymphocytic Pattern (28% of Cases): Marked by intense, non-granulomatous, transmural infiltration composed predominantly of small, mature T-lymphocytes and plasma cells, accompanied by medial smooth muscle necrosis. Multinucleated giant cells are absent. This pattern requires immunohistochemical differentiation from intravascular large B-cell lymphoma and primary CNS lymphoma.
• Necrotizing Pattern (14% of Cases): The most acutely destructive variant, characterized by transmural fibrinoid necrosis of the arterial media resembling systemic polyarteritis nodosa. Neutrophilic infiltration with leukocytoclasia (nuclear debris) is prominent. This subtype is associated with acute clinical presentations, higher rates of intracranial hemorrhage, and vessel wall rupture.
• Amyloid-Beta-Related Angiitis (ABRA / CAA-ri Overlap): A specialized histopathologic pattern characterized by granulomatous or lymphocytic inflammation centered around arteries exhibiting heavy transmural deposition of beta-amyloid protein, representing an auto-inflammatory reaction to vascular amyloid.

Endothelial Disruption, In-Situ Thrombosis, and Ischemic-Hemorrhagic Necrosis

The functional consequences of vascular wall inflammation result from two primary mechanisms:

• Luminal Stenosis and In-Situ Thrombosis: Progressive transmural inflammatory infiltration and concentric intimomedial hyperplasia compromise luminal diameter. Concomitant endothelial desquamation exposes prothrombotic subendothelial collagen, triggering platelet aggregation, fibrin deposition, and in-situ microthrombosis. The resulting arterial occlusion produces downstream multifocal cortical and subcortical ischemic infarctions of variable size and age.
• Aneurysmal Ectasia and Hemorrhagic Transformation: Severe elastolysis and transmural necrosis weaken the structural integrity of muscular arteries. Hemodynamic stress induces focal ectasia, pseudoaneurysm formation, and microaneurysmal rupture. This pathology accounts for the intraparenchymal hematomas, sulcal subarachnoid hemorrhages, and cortical petechiae observed in 20% to 30% of patients.

4. Natural History

Without timely recognition and aggressive immunosuppressive intervention, PACNS follows an unrelenting, disabling course characterized by cumulative neurological injury.

Epidemiology and Demographic Predilection

PACNS is an ultra-rare disorder with an estimated annual incidence of 1.0 to 2.4 cases per 1,000,000 person-years across North American and European populations. Key demographic features include:

• Age Distribution: The peak onset occurs between 40 and 60 years of age, with a median age at diagnosis of 50 years. Although predominantly a disease of middle-aged adults, pediatric primary angiitis (cPACNS) and late-onset cases in patients older than 70 years are recognized.
• Sex Predilection: Cohort analyses demonstrate a slight male predilection, with male-to-female ratios ranging between 1.2:1 and 1.5:1.
• Absence of Systemic Features: By definition, PACNS lacks constitutional rheumatologic features; peripheral inflammatory signs such as cutaneous purpura, peripheral neuropathy, active glomerulonephritis, and severe systemic serositis are absent.

Disease Trajectory: Untreated Mortality vs Treated Remission

The natural history of PACNS has evolved dramatically with the advent of standardized immunosuppressive regimens:

• Untreated Prognosis: Historically, untreated PACNS resulted in progressive, stuttering neurological deterioration. Recurrent multifocal strokes, refractory status epilepticus, and rapidly progressive dementia led to severe functional dependency, coma, and mortality rates exceeding 80% within 12 to 24 months of symptom onset.
• Treated Trajectory: Modern induction therapy incorporating high-dose corticosteroids combined with cyclophosphamide or rituximab achieves clinical stabilization or favorable response in 70% to 80% of patients. Long-term mortality has dropped to between 8% and 15%.
• Relapse Kinetics: Despite achieving initial remission, PACNS carries a 25% to 35% risk of clinical or radiological relapse within 3 to 5 years. Relapse is most frequently triggered during rapid corticosteroid tapering or following the premature withdrawal of maintenance immunosuppression.
• Long-Term Morbidity: Approximately 35% to 45% of surviving patients maintain residual functional deficits (modified Rankin Scale score 2 or higher), including executive dysfunction, fatigue, chronic headache, and persistent motor weakness.

5. Clinical Presentation

The clinical presentation of PACNS is notoriously protean, reflecting the scattered anatomical distribution of microvascular ischemia across disparate cerebral territories.

History

Symptom onset is insidious and subacute in more than 75% of patients, evolving over weeks to several months. A minority present hyperacutely with sudden stroke-like deficits or seizures. Key clinical manifestations include:

• Headache (50% to 65% of Cases): The single most common presenting complaint. Typically described as dull, aching, diffuse, continuous, and refractory to conventional analgesics. Unlike the instantaneous thunderclap headache characteristic of RCVS, the headache of PACNS exhibits an indolent, progressive character.
• Cognitive Impairment and Behavioral Alterations (50% of Cases): Progressive memory loss, executive dysfunction, confusion, apathy, personality changes, or overt encephalopathy. Subcortical microvascular involvement frequently manifests as psychomotor slowing.
• Focal Neurological Deficits (50% to 60% of Cases): Multifocal transient ischemic attacks (TIAs) or completed ischemic/hemorrhagic strokes occurring in multiple vascular territories. Symptoms include hemiparesis, sensory loss, aphasia, neglect, and ataxia.
• Seizures (25% to 30% of Cases): Focal or secondarily generalized tonic-clonic seizures, secondary to cortical irritation from micro-infarctions or leptomeningeal inflammation.
• Cranial Neuropathies and Visual Disturbances (10% to 15% of Cases): Diplopia, ocular motor palsies (CN III, IV, VI), facial weakness, visual field deficits, or visual loss due to ischemic optic neuropathy or chiasmatic ischemia.
• Constitutional Symptoms: Low-grade fever, fatigue, and mild malaise occur in up to 15% of patients. High-spiking fevers, marked night sweats, weight loss, or joint pain are atypical and strongly point toward secondary systemic vasculitis, occult infection, or lymphoma.

Physical Exam

A structured, multi-domain neurological physical examination is essential to localize multifocal central nervous system disease:

• Montreal Cognitive Assessment (MoCA): Administer 30-point standardized cognitive testing evaluating executive function, memory, attention, and language -> Positive Finding: Score under 26 points, with prominent deficits in delayed recall, clock drawing, and trail-making tasks [Sensitivity 82%, Specificity 76%].
• Pronator Drift Test: Patient maintains bilateral arms extended forward with palms supinated and eyes closed for 30 seconds -> Positive Finding: Slow pronation and downward drift of the affected forearm indicating subtle upper motor neuron pyramidal weakness [Sensitivity 74%, Specificity 88%].
• Confrontation Visual Field Testing: Test each quadrant monocularly across peripheral fields -> Positive Finding: Homonymous hemianopia, quadrantanopia, or scotoma reflecting cortical or optic radiation microvascular ischemia [Sensitivity 62%, Specificity 94%].
• Tandem Gait and Romberg Maneuver: Patient walks heel-to-toe in a straight line, followed by standing with feet together and eyes closed -> Positive Finding: Lateral stepping, truncal ataxia, or loss of balance indicating cerebellar hemispheric/vermal ischemia or proprioceptive pathway disruption [Sensitivity 68%, Specificity 84%].
• Nuchal Rigidity and Meningeal Signs (Kernig and Brudzinski): Passive flexion of the neck in supine position -> Positive Finding: Resistance to neck flexion or involuntary hip/knee flexion, indicating active leptomeningeal inflammation [Sensitivity 35%, Specificity 91%].

6. Imaging and Diagnostic Studies

The diagnosis of PACNS requires a multi-modality diagnostic hierarchy designed to demonstrate vascular wall inflammation, delineate parenchymal ischemia, and definitively exclude mimics.

Diagnostic Modality Hierarchy

The recommended diagnostic algorithm proceeds sequentially from non-invasive parenchymal and vessel imaging to invasive vascular and histopathological confirmation:

• 1. Brain MRI with DWI, FLAIR, SWI, and contrast administration (First-line structural assessment).
• 2. High-Resolution Vessel Wall MRI (VWI) (Non-invasive mural assessment of medium and large arteries).
• 3. Lumbar Puncture and Cerebrospinal Fluid Analysis (Mandatory inflammatory verification and infectious/neoplastic exclusion).
• 4. Catheter Digital Subtraction Angiography (DSA) (Catheter-based luminal mapping).
• 5. Open or Stereotactic Brain and Leptomeningeal Biopsy (Gold standard histopathologic verification).

Conventional Brain MRI and MRA

Magnetic resonance imaging is abnormal in more than 90% to 95% of patients with PACNS. While non-specific, a completely normal MRI carries a high negative predictive value:

• Diffusion-Weighted Imaging (DWI): Identifies multifocal, bilateral areas of restricted diffusion (high DWI signal with corresponding low apparent diffusion coefficient [ADC] values) representing acute and subacute ischemic infarctions scattered across multiple non-contiguous vascular distributions.
• T2-Weighted and FLAIR Sequences: Reveal confluent subcortical white matter hyperintensities, periventricular lesions, and cortical swelling. In 10% to 15% of cases, prominent leptomeningeal hyperintensity is evident.
• Susceptibility-Weighted Imaging (SWI): Detects microhemorrhages, cortical petechiae, or sulcal superficial siderosis in 30% to 50% of patients, reflecting microvascular transmural necrosis.
• T1 Post-Contrast Imaging: Demonstrates nodular, linear, or diffuse leptomeningeal enhancement in 10% to 15% of cases. In approximately 5% of patients, PACNS presents as a solitary or multifocal enhancing mass lesion (pseudotumoral PACNS) mimicking high-grade glioma or cerebral metastasis.

High-Resolution Vessel Wall MRI (VWI)

High-resolution 3-Tesla vessel wall MRI using 3D black-blood T1 post-contrast sequences directly images the arterial wall rather than the hollow lumen:

• Diagnostic Hallmark: Identifies smooth, concentric, circumferential vessel wall thickening with vivid, homogeneous contrast enhancement within the walls of intracranial arteries (sensitivity 80% to 90%, specificity 85% to 95% for medium and large vessels).
• Differential Specificity: Differentiates PACNS from intracranial atherosclerosis (which exhibits eccentric, focal, heterogeneous wall thickening and enhancement centered on plaque cores) and RCVS (which displays smooth luminal narrowing with minimal or absent vessel wall thickening and no significant contrast enhancement).
• Therapeutic Monitoring: Longitudinal VWI allows objective tracking of biological treatment response; resolution of vessel wall enhancement correlates with histological remission.

Catheter Digital Subtraction Angiography (DSA)

Biplane catheter DSA evaluates medium-sized muscular arteries (caliber 500 micrometers to 2 mm) with superior spatial resolution compared to MRA and CTA:

• Angiographic Hallmarks: Alternating segments of focal luminal stenosis and post-stenotic dilatation (the classic “string of beads” or “sausage-string” sign), smooth concentric arterial tapering, abrupt vascular cut-offs, delayed territorial transit times, and micro-collateral formation across multiple vascular distributions.
• Sensitivity Limitations (50% to 60%): Catheter DSA cannot resolve vessels smaller than 200 to 300 micrometers. Consequently, patients with purely small-vessel PACNS involving penetrating and leptomeningeal arterioles frequently display completely normal conventional angiograms.
• Specificity Limitations (30% to 40%): Luminal beading is not pathognomonic for vasculitis; identical appearances occur in RCVS, intracranial atherosclerosis, fibromuscular dysplasia, radiation vasculopathy, and infectious endarteritis.

Lumbar Puncture and Cerebrospinal Fluid (CSF) Analysis

Lumbar puncture is mandatory in all patients with suspected PACNS unless contraindicated by significant space-occupying lesions or impending cerebral herniation:

• Typical Inflammatory Profile: The CSF is abnormal in 80% to 90% of biopsy-proven cases, exhibiting an aseptic inflammatory profile characterized by mild-to-moderate lymphocytic pleocytosis (typically 10 to 100 white blood cells/microliter, with more than 85% mononuclear cells), elevated total protein concentration (60 to 250 mg/dL), normal glucose concentration (CSF-to-serum glucose ratio greater than 0.5), and normal or mildly elevated opening pressure.
• High Negative Predictive Value: A completely normal CSF profile (0 to 1 WBC/microliter, normal protein, normal opening pressure) carries a negative predictive value of approximately 95%. When CSF parameters are normal, alternative diagnoses such as RCVS, CADASIL, or non-inflammatory atherothrombosis should be prioritized.
• Mandatory Infectious and Neoplastic Exclusion Panel: The diagnostic utility of CSF lies equally in excluding primary mimics. Analysis must include CSF Gram stain, bacterial/fungal/mycobacterial cultures, cryptococcal antigen, venereal disease research laboratory (VDRL) test, varicella-zoster virus (VZV) PCR, herpes simplex virus (HSV-1/2) PCR, enterovirus PCR, cytology, and flow cytometry (to rule out intravascular B-cell lymphoma).

Brain and Leptomeningeal Biopsy (Gold Standard)

Biopsy remains the definitive diagnostic arbiter, providing tissue confirmation before exposing patients to prolonged cytotoxic immunosuppression:

• Diagnostic Yield and False-Negative Rate: Targeted open biopsy directed to an abnormal, contrast-enhancing or T2-hyperintense cortical lesion yields diagnostic confirmation in 65% to 75% of cases. Non-targeted biopsy of the non-dominant frontal or temporal pole yields confirmation in 50% to 60%. The false-negative rate ranges from 25% to 35% due to the segmental, focal nature of vascular inflammation (skip lesions).
• The Leptomeningeal Requirement: Diagnostic sensitivity drops below 30% if the biopsy specimen contains only cerebral cortex without overlying leptomeninges. Intact pia-arachnoid tissue containing small surface arteries must be harvested.
• Alternative Pathological Diagnoses: In up to 40% of biopsies performed for suspected PACNS, the procedure identifies an unsuspected alternative diagnosis that redirects treatment, including intravascular lymphoma, neurosarcoidosis, cerebral amyloid angiopathy, fungal/mycobacterial vasculitis, or high-grade astrocytoma.

7. Classification

Standardized diagnostic criteria, clinicoradiological subtype categories, and neuropathological classifications guide clinical staging and therapeutic stratification in PACNS.

8. Treatment

The therapeutic strategy for PACNS is divided into an acute immunosuppressive induction phase to arrest active vascular inflammation, followed by a maintenance phase to prevent disease relapse while tapering glucocorticoid toxicity.

Medical / Nonoperative

Medical management constitutes the primary definitive treatment for PACNS, stratified by disease severity and subtype:

• High-Dose Corticosteroid Induction: Intravenous methylprednisolone pulse therapy (1,000 mg daily for 3 to 5 consecutive days). Following pulse therapy, transition to oral prednisone at 1 mg/kg/day (maximum 60 to 80 mg/day) as a single morning dose for 4 weeks. If clinical and radiographic stabilization is achieved, initiate a structured, slow oral taper, reducing by 5 to 10 mg every 2 to 4 weeks until reaching 15 mg/day, then tapering by 2.5 mg every 2 to 4 weeks toward complete discontinuation over 6 to 12 months.
• Cyclophosphamide Induction (First-Line Alkylating Agent): Concomitant administration of cyclophosphamide is standard for definite biopsy-proven PACNS, progressive large-vessel disease, or necrotizing histological subtypes. Delivered either as monthly intravenous pulses (500 to 750 mg/m² adjusted for renal function and nadir white blood cell counts, co-administered with mesna and vigorous intravenous hydration) for 3 to 6 months, or as daily oral therapy (1.5 to 2.0 mg/kg/day). Absolute neutrophil count (ANC) must be monitored every 2 weeks, maintaining ANC above 1,500 cells/microliter.
• Rituximab Alternative Induction: In patients with small-vessel PACNS, childbearing potential (to avoid cyclophosphamide gonadotoxicity), previous high cumulative cyclophosphamide exposure, or pre-existing urothelial disease, rituximab (375 mg/m² weekly for 4 weeks, or two 1,000 mg infusions separated by 14 days) serves as an effective alternative induction agent, demonstrating comparable remission rates.
• Mandatory Antimicrobial Prophylaxis: All patients receiving cyclophosphamide or combination high-dose corticosteroid therapy must receive prophylaxis against Pneumocystis jirovecii pneumonia using trimethoprim-sulfamethoxazole (TMP-SMX 800/160 mg single tablet three times weekly, or 400/80 mg single-strength daily). In sulfa-allergic patients, aerosolized pentamidine (300 mg monthly) or oral dapsone (100 mg daily, following G6PD screening) is administered.
• Maintenance Immunosuppression: Initiated after 3 to 6 months of successful induction therapy upon achieving clinical and radiological remission. Cyclophosphamide is discontinued and replaced by either:
  • Azathioprine: 2.0 to 2.5 mg/kg/day orally (following mandatory pre-treatment screening for thiopurine methyltransferase [TPMT] and NUDT15 polymorphisms).
  • Mycophenolate Mofetil (MMF): 1,000 to 1,500 mg orally twice daily.
  Maintenance therapy is maintained for a minimum of 18 to 24 months, with gradual taper over an additional 6 to 12 months if repeat neuroimaging and CSF remain quiescent.
• Supportive Prophylaxis: Proton pump inhibitors for gastroprotection; calcium carbonate (1,200 mg daily) and vitamin D3 (1,000 IU daily) combined with bisphosphonate therapy for bone mineral density preservation; and tight glycemic monitoring.

Operative: Targeted Cortical and Leptomeningeal Biopsy Technique

The primary surgical intervention in PACNS is diagnostic open brain and leptomeningeal biopsy, executed using microsurgical technique to maximize diagnostic yield and avoid specimen crush artifacts:

• Preoperative Image Coregistration: High-resolution 3D volumetric T1 post-contrast and T2/FLAIR MRI sequences are uploaded to an optical or electromagnetic stereotactic neuronavigation system. The entry site is centered precisely over the superficial cortex displaying the most intense leptomeningeal enhancement or subcortical edema. If non-targeted, the non-dominant frontal pole (superior/middle frontal gyrus anterior to coronal suture) is mapped.
• Patient Positioning and Craniotomy: The patient is positioned supine with the head fixed in a three-pin Mayfield clamp, rotated 15 to 30 degrees away from the surgical side. A 4 to 5 cm curvilinear or straight skin incision is marked within the hairline. A mini-craniotomy (2.5 to 3.0 cm diameter) is performed using a high-speed electric drill and craniotome, elevating the bone flap cleanly from the underlying dura.
• Dural Opening and Leptomeningeal Preservation: Under operative microscope illumination, the dura is incised in a cruciate or U-shaped fashion and reflected. Meticulous care is taken not to strip or tear the underlying arachnoid mater. An intact 1 x 1 cm swath of vascularized leptomeninges containing visible small surface vessels is micro-dissected from the underlying cerebral cortex using micro-scissors and micro-forceps, then sharply excised.
• Cortical Wedge Resection: A corresponding 1 x 1 x 1 cm en-bloc wedge of cerebral cortex, incorporating both gray matter and subcortical white matter, is excised from the crest of the exposed gyrus. Electrocautery must be avoided across the margins of the tissue specimen to prevent thermal coagulative artifact; gentle micro-scissors and cold scalpel dissection are utilized.
• Specimen Handling and Distribution: The biopsy specimen is immediately divided in the operating theater:
  • Formalin-fixed aliquot: Light microscopy, hematoxylin and eosin (H&E), elastic-van Gieson (EVG) for internal elastic lamina integrity, and immunohistochemistry (CD3, CD20, CD68, beta-amyloid).
  • Glutaraldehyde aliquot: Electron microscopy for endothelial ultrastructure.
  • Fresh-frozen, non-fixed aliquot: Microbial PCR, fungal/mycobacterial culture, and molecular clonality analysis.
• Hemostasis and Cranial Reconstruction: Hemostasis within the cortical bed is achieved using low-power bipolar micro-coagulation and topical absorbable hemostats (oxidized regenerated cellulose, flowable gelatin-thrombin matrix). The dura is closed in a watertight fashion with 4-0 braided or monofilament suture. The bone flap is secured with titanium microplates, and the scalp is closed in layers.

Indications & Contraindications

Precise clinical thresholds govern the decision for surgical biopsy and the safe administration of cytotoxic immunosuppression:

• Indications for Brain Biopsy:
  • Unexplained progressive cognitive decline, multifocal ischemic strokes, or focal neurological deficits with abnormal inflammatory CSF and non-diagnostic or negative vascular imaging.
  • Presence of a solitary or multifocal enhancing mass lesion (pseudotumoral PACNS) where high-grade glioma, lymphoma, or infection cannot be differentiated by neuroimaging.
  • Clinical deterioration or therapeutic failure during empiric corticosteroid monotherapy, prior to escalating to high-dose cytotoxic cyclophosphamide.
• Absolute Contraindications:
  • Uncorrected systemic coagulopathy (platelet count < 50,000/microliter, International Normalized Ratio [INR] > 1.4, or active therapeutic anticoagulation that cannot be safely reversed).
  • Target pathology situated exclusively within deep brainstem, diencephalic, or primary motor/language cortex where wedge resection would induce catastrophic, permanent deficit.
  • Histopathological confirmation of systemic vasculitis on accessible extracranial tissue (e.g., temporal artery biopsy in giant cell arteritis, sural nerve/muscle biopsy in polyarteritis nodosa).

9. Outcome

Modern multimodal management has transformed PACNS from a universally fatal condition into a treatable, chronic vascular disease with quantifiable functional outcomes.

Functional Outcomes and Modified Rankin Scale (mRS) Metrics

Longitudinal registry data provide clear benchmarks for functional recovery following contemporary induction and maintenance regimens:

• Favorable Functional Recovery (mRS 0 to 2): Approximately 55% to 75% of patients achieve favorable functional outcomes (modified Rankin Scale score 0, 1, or 2, indicating functional independence) at 12 to 24 months post-induction. Early initiation of combination corticosteroids and cyclophosphamide before the accumulation of widespread cerebral infarctions is the single strongest predictor of functional independence.
• Severe Functional Dependency (mRS 3 to 5): Observed in 20% to 30% of patients, driven by extensive baseline stroke burden, diagnostic delay exceeding 3 months, large-vessel stenosing subtype, or intracranial hemorrhage at presentation.
• Subtype Disparity in Recovery: Patients with biopsy-proven small-vessel PACNS exhibit significantly higher rates of cognitive and physical functional recovery compared to patients with large-vessel stenotic PACNS. Large-vessel disease carries a higher frequency of recurrent hemodynamic territorial infarctions and irreversible motor deficits.

Relapse Kinetics and Long-Term Survivorship

Long-term survival and relapse profiles demand extended clinical surveillance:

• Five-Year Overall Survivorship: Modern cohorts demonstrate 5-year survival rates ranging between 85% and 92%. Mortality in the contemporary era is driven primarily by opportunistic systemic infections and iatrogenic immunosuppressive complications rather than acute cerebral vasculitic progression.
• Cumulative Relapse Incidence: Relapse occurs in 25% to 35% of patients over a 3- to 5-year follow-up window. More than 70% of relapses develop within the first 18 months, coinciding with glucocorticoid dose reduction below 10 to 15 mg/day or during the transition from induction to maintenance therapy.
• Surveillance Strategy: Serial clinical evaluations paired with contrast-enhanced brain MRI (incorporating DWI and vessel wall sequences) are recommended every 3 months for the first year, every 6 months during the second year, and annually thereafter for a minimum of 5 years.

10. Complications

Complications in PACNS arise both from primary cerebrovascular tissue destruction and from the systemic toxicities of prolonged immunosuppressive therapy.

11. Pearls and Pitfalls

Navigating the complex diagnostic and therapeutic landscape of PACNS requires keen recognition of clinical mimics and surgical nuances:

• The RCVS Differentiation Pearl: Reversible cerebral vasoconstriction syndrome (RCVS) is the most frequent mimic of PACNS. RCVS is characterized by hyperacute thunderclap headaches, normal CSF parameters (white blood cell count under 5 cells/microliter, normal protein), and diffuse, smooth luminal vasoconstriction that spontaneously resolves within 12 weeks. High-resolution vessel wall MRI confirms the absence of mural thickening or contrast enhancement in RCVS. Never administer cytotoxic cyclophosphamide for RCVS.
• The Mandatory Leptomeningeal Sampling Rule: In performing diagnostic brain biopsy, never submit cerebral cortex alone. PACNS has a powerful predilection for small leptomeningeal arteries; sampling parenchymal cortex without overlying pia-arachnoid tissue causes false-negative rates to surge above 70%. Ensure an intact 1 x 1 cm vascularized leptomeningeal patch is secured.
• The Negative Angiography Trap: Catheter DSA cannot visualize vessels smaller than 200 to 300 micrometers. A completely normal catheter cerebral angiogram does not exclude PACNS, as more than 50% of biopsy-proven cases involve small intraparenchymal and leptomeningeal arterioles beneath angiographic spatial resolution.
• The CSF Normalcy Red Flag: Truly normal cerebrospinal fluid (0 to 1 white blood cell, normal protein, normal pressure) carries a 95% negative predictive value for PACNS. If the CSF is entirely normal, thoroughly re-evaluate for alternative etiologies such as non-inflammatory atherothrombosis, CADASIL, or RCVS before proceeding with invasive biopsy or cytotoxic therapy.
• The VZV Vasculopathy Rule-Out: Varicella-zoster virus (VZV) vasculopathy is the leading infectious mimic of PACNS, producing multifocal stenoses, ischemic strokes, and CSF lymphocytic pleocytosis often without a prior dermatomal rash. Always verify negative CSF VZV PCR and anti-VZV IgG antibody titers before initiating aggressive immunosuppression.
• The Corticosteroid Biopsy Window: Do not delay diagnostic brain biopsy once high-dose corticosteroids are initiated. Transmural inflammatory infiltrates can clear within 7 to 14 days of corticosteroid therapy, yielding non-diagnostic, indeterminate histology and leaving the clinician without tissue confirmation.
• Mesna and Hydration Co-Prescription: Never administer intravenous cyclophosphamide without concurrent sodium 2-mercaptoethanesulfonate (Mesna) and vigorous intravenous saline hydration. Mesna binds free acrolein within the urinary bladder, preventing hemorrhagic cystitis and reducing long-term urothelial malignancy risk.

12. Controversy

Significant debate persists across multiple diagnostic and therapeutic facets of PACNS management:

High-Resolution Vessel Wall MRI vs. Brain Biopsy as Definitive Arbiter

A central controversy centers on whether advanced 3T high-resolution vessel wall MRI (VWI) can supplant invasive brain and leptomeningeal biopsy as the primary diagnostic arbiter. Proponents of VWI emphasize that concentric, circumferential vessel wall enhancement provides direct non-invasive visualization of vascular wall inflammation throughout the cerebral arterial tree, avoids the 2% to 4% morbidity of open craniotomy, and circumvents the 25% to 35% false-negative sampling error inherent to biopsy skip lesions. Conversely, surgical and pathology consensus maintains that VWI cannot resolve small arterioles under 200 micrometers (where small-vessel PACNS predominates) and lacks histopathological specificity. VWI cannot definitively distinguish PACNS from intracranial atherosclerosis, neurosarcoidosis, VZV endarteritis, or intravascular lymphoma, conditions that require diametrically opposed clinical treatments.

Induction Immunosuppression: Rituximab vs. Cyclophosphamide

The selection of primary induction therapy remains hotly contested. Historically, cyclophosphamide combined with corticosteroids established the standard of care based on landmark cohort data. However, cyclophosphamide carries substantial risks of hemorrhagic cystitis, late bladder carcinoma, opportunistic myelosuppression, and permanent sterility. Emerging multicenter registry evidence indicates that rituximab (anti-CD20 monoclonal antibody) achieves comparable rates of clinical remission and functional stabilization with a superior safety profile, particularly in small-vessel and lymphocytic subtypes. Nonetheless, skeptics argue that randomized controlled trial data specifically validating rituximab in PACNS are lacking, and cyclophosphamide remains preferred by many centers for severe, rapidly progressive, necrotizing or large-vessel disease.

Cerebral Amyloid Angiopathy-Related Inflammation (CAA-ri) Overlap

Whether amyloid-beta-related angiitis (ABRA) and cerebral amyloid angiopathy-related inflammation (CAA-ri) represent a distinct clinicopathological entity or a subtype of PACNS continues to spark academic debate. CAA-ri patients are typically older (over 65 years), carry a high frequency of the apolipoprotein E (APOE) epsilon4/4 genotype, and exhibit dramatic, asymmetric T2/FLAIR white matter hyperintensities with prominent cortical microbleeds. While the vascular wall exhibits transmural granulomatous inflammation indistinguishable from PACNS, the inflammation is driven by an auto-reactive immune response against vascular amyloid-beta deposits. Unlike classical PACNS, CAA-ri frequently responds completely to corticosteroid monotherapy without requiring alkylating agents, leading many authorities to classify it as a unique autoimmune encephalopathy.

13. Evidence-Based Synthesis

Modern clinical decision-making is grounded in high-quality randomized controlled trials and prospective registry analyses, corroborated by seminal biomechanical and anatomical descriptions.

14. References

• Beuker C, Strunk D, Rawal R et al. Primary Angiitis of the CNS: A Systematic Review and Meta-analysis. Neurol Neuroimmunol Neuroinflamm. 2021. doi:10.1212/NXI.0000000000001093 [Level I: Systematic Review / Meta-Analysis]
• Fisse AL, Bonberg N, Beuker C et al. Disease Characteristics and Treatments Associated with Outcome in Primary Angiitis of the Central Nervous System-A Multicenter Cohort Study in 163 Patients. Ann Neurol. 2025. doi:10.1002/ana.27295 [Level II: Multicenter Comparative Cohort Study]
• Deb-Chatterji M, Aghel P, Pinnschmidt H et al. Biopsy-proven PACNS: results from the large, multicentre cohort of cerebral vasculitis patients. J Neurol Neurosurg Psychiatry. 2025. doi:10.1136/jnnp-2025-335764 [Level II: Multicenter Prospective Cohort Study]
• Salvarani C, Hunder GG, Brown RD Jr. Primary Central Nervous System Vasculitis. N Engl J Med. 2024. doi:10.1056/NEJMra2314942 [Level V (Expert Clinical Review): Seminal State-of-the-Art Review]
• Salvarani C, Brown RD Jr, Calamia KT et al. Primary central nervous system vasculitis: analysis of 101 patients. Ann Neurol. 2007. doi:10.1002/ana.21226 [Level IV: Landmark Retrospective Cohort Study]
• Calabrese LH, Mallek JA. Primary angiitis of the central nervous system: report of 8 new cases, review of the literature, and proposal for diagnostic criteria. Medicine (Baltimore). 1988. doi:10.1097/00005792-198801000-00002 [Level V (Expert Consensus / Criteria): Seminal Diagnostic Criteria Formulation]
• Giannini C, Salvarani C, Hunder G, Brown RD. Primary central nervous system vasculitis: pathology and mechanisms. Acta Neuropathol. 2012. doi:10.1007/s00401-012-0973-9 [Level V (Histopathologic Consensus): Seminal Neuropathologic Case Series]

15. Master Class Pro-Tip

When executing a diagnostic cortical and leptomeningeal biopsy for suspected PACNS, never allow the neurosurgical team to take a simple cortical punch or needle core biopsy. The diagnostic hallmark of PACNS resides in the small muscular arteries of the leptomeninges; taking cortex without overlying pia-arachnoid causes false-negative rates to soar above 70%. Ensure the craniotomy is positioned over a gyrus displaying prominent sulcal enhancement on 3T vessel wall MRI, micro-dissect an intact 1 x 1 cm patch of vascularized leptomeninges prior to cortical incision, and excise a full 1 cm³ cortical wedge without electrocautery across specimen margins. Immediate intraoperative division into formalin for light microscopy, glutaraldehyde for electron microscopy, and fresh-frozen tissue for microbial PCR prevents non-diagnostic investigations and protects against catastrophic immunosuppression misdirection.

Master Class Surgical Pearl: Targeted Cortical-Leptomeningeal Biopsy Technique and Diagnostic Triage in PACNS