By Ken Uchino, MD, Leonard Calabrese, DO, and Rula Hajj-Ali, MD

Arriving at a diagnosis of central nervous system vasculitis (CNS-V) is fraught with challenges. Clinical presentation can be quite variable, and there is no classic presentation. Further complicating matters, the condition has several mimics, brain tissue is inaccessible and there is no disease-specific test. However, advances in neuroimaging and next-generation sequencing — along with the involvement of a multidisciplinary clinical team — have added formidably to our knowledge of CNS-V.

Recently, two different disease subtypes have been postulated: the small vessel variant (SVV) and large-medium vessel variant (LMVV).¹ We sought to clarify the clinical characteristics, brain MRI findings and diagnostic accuracy of high-resolution vessel wall image (HR-VWI) and brain biopsy according to the affected vessel size in CNS-V patients. We discuss the results of this study in a poster presented at ACR Convergence 2020.²

Distinguishing between LMVV and SVV

Using Cleveland Clinic’s CNS vasculopathy registry, we identified 34 CNS-V patients who underwent digital subtraction angiography (DSA) or magnetic resonance angiography (MRA) at our institution from 2012-2019. For this study, patients with cerebral vasculature indicating vasculitis in proximal or middle arterial segments had LMVV, and patients with only smaller distal branch vessel involvements or normal angiography had SVV.

Based on angiographical findings, 11 (32.4%) of the patients in our cohort were LMVV and 23 (67.6%) were SVV. Patients with LMVV had more infarcts and were more likely to have concentric vessel wall enhancement (VWE) on HR-VWI, as shown in Table 1. Meningeal/parenchymal contrast enhancement of lesion was more frequently observed in the SVV group, and the majority of patients in this group were diagnosed by brain biopsy, as shown in Table 2. Brian biopsy was positive in 100% of patients with SVV compared with only 57.1% in patients with LMVV.

Table 1. Brain MRI, cerebral angiogram and HR-VWI findings in CNS-V patients (LMVV vs. SVV)

	LMVV (n = 11)	SVV (N = 23)	P Value
MRI findings, n (%)
Brain infarcts	11 (100.0)	14 (60.9)	0.017
Parenchymal hemorrhage	1 (9.1)	5 (21.7)	0.638
Subarachnoid hemorrhage	0 (0.0)	4 (17.4)	0.280
White matter lesion	2 (18.2)	12 (52.2)	0.076
Tumor-like lesion	0 (0.0)	8 (34.8)	0.034
Contract enhancement lesion	5 (45.5)	20 (87.0)	0.033
Assessment of the cerebral vascular image, n (%)
Both MRA and DSA	11 (100.0)	13 (56.5)	0.014
Only MRA	0 (0.0)	8 (34.8)	0.034
Only DSA	0 (0.0)	2 (8.7)	1.000
Arterial stenosis, n (%)
Proximal segment	7 (63.6)	0 (0.0)	< 0.001
Middle segment	10 (90.9)	0 (0.0)	< 0.001
Smaller distal branch	10 (90.0)	8 (34.8)	0.003
None	0 (0.0)	15 (62.5)	< 0.001
HR-VWI findings, n (%)
VWE	10 (100.0) [n = 10]	2 (14.2) [n = 14]	< 0.001
Concentric VWE	9 (90.0) [n = 10]	1 (7.1) [n = 14]	< 0.001
Eccentric VWE	1 (10.0)[n = 10]	1 (7.1) [n = 14]	1.000

Table 2. Brain biopsy and CSF findings in CNS-V patients (LMVV vs. SVV)

	LMVV (n = 11)	SVV ( n = 23)	P Value
Biopsy proven CNS-V, n (%)	4 (36.4)	18 (78.3)	0.026
Biopsy positive, n (%)	4 (57.1) [n = 7]	18 (100.0) [n = 18]	0.015
Pathological patterns, n (%)
Granulomatous vasculitis	1 (14.3) [n = 7]	1 (11.1) [n = 18]	0.490
Lymphocytic vasculitis	3 (42.9) [n = 7]	12 (66.7) [n = 18]	0.378
Necrotizing vasculitis	0 (0.0) [n = 7]	5 (27.8) [n = 18]	0.274
CSF findings, n (%)
Abnormal*	11 (100.0) [n = 11]	16 (80.0) [n = 20]	0.269
Pleocytosis*	10 (90.9) [ n = 11]	12 (60.0) [n = 20]	0.106
Elevated protein*	8 (72.7) [n = 11]	13 (65.0) [n = 20]	1.000

Arriving at a diagnosis of CNS-V

Diagnostic approach for CNS-V depends on the affected vessel size. In LMVV patients, the diagnosis of CNS-V is mainly based on evidence of concentric VWE on HR-VWI. By contrast, brain biopsy should be considered in patients with suspected SVV CNS-V if parenchymal enhancement of lesion is detected. In patients without VWE on HR-VWI and targeting contrast enhancement lesion, careful assessments including CSF abnormalities, DSA findings and clinical pictures are needed for the diagnosis of the CNS-V. 

We take a systematic approach to the work-up of any patient suspected to have CNS-V. This approach includes a general history and physical exam, with a thorough review of symptoms associated with systemic autoimmune disease. Clinicians should look for infectious and/or malignant conditions that might be associated with many of these nonspecific symptoms, especially fever, malaise, joint pains and weight loss. When CNS-V is suspected, brain biopsy is of high yield.

References

1. Thaler C, Kaufmann-Bühler AK, Gansukh T, et al. Neuroradiologic Characteristics of Primary Angiitis of the Central Nervous System According to the Affected Vessel Size. Clin Neuroradiol. 2019;29(1):37-44. 
2. Shimoyama T, Uchino K, Calabrese L, Hajj-ali R. Clinical Characteristics, Brain MRI Findings, and Diagnostic Approach of the Central Nervous System Vasculitis by Affected Vessel Size [abstract]. Arthritis Rheumatol. 2020; 72 (suppl 10).