How can reversible cerebral vasoconstriction syndrome be differentiated from primary central nervous system vasculitis on contrast‑enhanced MRI (including vessel‑wall imaging)?

Differentiating RCVS from CNS Vasculitis on Contrast-Enhanced MRI

Contrast-enhanced vessel wall imaging is the key discriminator: CNS vasculitis demonstrates arterial wall enhancement while RCVS does not. 1

Primary Imaging Distinction: Vessel Wall Enhancement Pattern

The single most important MRI feature is the presence or absence of vessel wall enhancement on contrast-enhanced sequences:

• CNS vasculitis shows arterial wall enhancement (typically smooth and concentric in 69% of cases, or smooth and eccentric in 23%) with persistent wall thickening 2
• RCVS shows negligible-to-mild or absent wall enhancement despite diffuse uniform wall thickening in 77% of cases 2
• This distinction increases diagnostic accuracy from approximately 36% with luminal imaging alone to 89% when vessel wall enhancement is evaluated 3

Optimal MRI Protocol

Use contrast-enhanced MRI brain with dedicated vessel wall imaging sequences plus time-of-flight MRA 4:

• Black-blood contrast-enhanced T1-weighted sequences with fat suppression are essential 2
• Standard brain sequences (T2/FLAIR, DWI) to evaluate parenchymal changes 4
• Time-of-flight MRA without contrast for luminal assessment 4

Temporal Evolution on Follow-Up Imaging

Serial imaging within 3-4 weeks is diagnostically critical when initial vessel wall imaging is equivocal:

• RCVS demonstrates complete resolution of arterial narrowing and wall thickening in 89% of cases within 3 months (typically 1-4 weeks) 5, 6
• CNS vasculitis shows persistent or progressive arterial narrowing at median 17-month follow-up, with stable concentric enhancement and thickening in 67% of cases 5, 2

Parenchymal MRI Patterns That Support Each Diagnosis

Findings More Suggestive of CNS Vasculitis:

• Multiple infarcts of variable ages (present in 50% of cases) 3
• Multiple small deep infarcts 7
• Extensive deep white matter lesions 7
• Tumor-like mass lesions (5% of cases) 3, 7
• Meningeal enhancement (8% of cases) 3
• Multiple gadolinium-enhanced lesions 7
• Multiple microhemorrhages 8

Findings More Suggestive of RCVS:

• Subarachnoid hemorrhage (more common than in vasculitis) 7
• Vasogenic edema (predominates in RCVS) 7
• Normal brain imaging at admission (69% of RCVS cases vs. 0% of vasculitis) 7
• Cervical artery dissection (found only in RCVS, not vasculitis) 7

Luminal Imaging Patterns (Less Specific)

Both conditions show multifocal segmental arterial narrowing and dilatation, creating substantial diagnostic overlap 8, 1:

• Time-of-flight MRA is abnormal in 81% of vasculitis cases 3
• Both demonstrate "beading" pattern with alternating stenosis and dilatation 8
• Luminal imaging alone cannot reliably distinguish between the two entities 8, 1

Critical Diagnostic Algorithm

1. Obtain contrast-enhanced MRI with vessel wall imaging sequences (not just standard brain MRI) 4
2. Assess for arterial wall enhancement:
   • Present = strongly favors CNS vasculitis 1, 5
   • Absent = strongly favors RCVS 1, 5
3. If wall enhancement is equivocal or mild, obtain follow-up imaging at 2-4 weeks:
   • Resolution of vasoconstriction = confirms RCVS 5, 6
   • Persistent/progressive narrowing = confirms vasculitis 5, 2
4. Integrate parenchymal findings to support the diagnosis 7

Common Pitfalls to Avoid

• Do not rely on conventional angiography (DSA) or MRA/CTA alone – these have low specificity with significant overlap between RCVS and vasculitis 8, 1
• Do not use standard brain MRI without dedicated vessel wall imaging sequences – this misses the critical enhancement pattern 3
• Do not skip follow-up imaging when initial vessel wall imaging is inconclusive – temporal evolution is diagnostically definitive 5, 6
• Do not assume normal CSF excludes vasculitis – ESR and acute-phase reactants are characteristically normal in primary CNS vasculitis 3
• Misdiagnosis as vasculitis is common and leads to inappropriate immunosuppression in a self-limited condition 9, 6