Biopsy-proven PACNS: results from the large, multicentre cohort of cerebral vasculitis patients

Abstract

Background: Reports of primary angiitis of the central nervous system (PACNS) are mainly restricted to clinically suspected cases, but biopsy-proven ones are rare. Here, we present results from a large multicentre cohort of patients with biopsy-proven PACNS (BP-PACNS). In particular, we provide insights into characteristics and treatment responses of PACNS subtypes.

Methods: BP-PACNS patients treated between 1999 and 2021 were analysed. The outcome was assessed by the modified Rankin Scale (mRS). Between-group comparisons were performed by Kruskal-Wallis, χ² or Fisher's exact tests.

Results: In total, 57 patients were analysed (52% male). Of these, n=37 (65%) had a lymphocytic (L-PACNS), n=9 (16%) an amyloid-beta-related angiitis (ABRA), n=8 (14%) a granulomatous (G-PACNS) and n=3 (5%) a necrotising (N-PACNS) PACNS subtype. At the time of diagnosis, age differed significantly between groups (median age (years) L-PACNS 47, ABRA 64.5, G-PACNS 37, N-PACNS 65; p=0.008). The clinical course was mostly monophasic in L-PACNS and ABRA (65% and 75%, respectively), while relapsing-remitting in G-PACNS (63%). Median mRS at last follow up was 2 (IQR 1.25-4) in the study group. Worst outcome (median mRS 4) and highest mortality (25%) were seen in G-PACNS. Good induction treatment response was achieved in 77% of all BP-PACNS patients but was low in those with G-PACNS (29%).

Conclusions: In this large, multicentre series of BP-PACNS patients, G-PACNS had a worse functional outcome, a predominant relapsing-remitting disease and a poorer response to the induction therapy. An optimal first-line treatment regimen for G-PACNS patients should be further examined and established in larger studies to improve the outcome of G-PACNS patients.

Keywords: CEREBROVASCULAR DISEASE; STROKE; VASCULITIS.

Figure 1

Distribution of the histological patterns within the study cohort. The most prevalent histological pattern in the entire study group was L-PACNS (65%), followed by ABRA (16%), G-PACNS (14%) and N-PACNS (5%). ABRA, amyloid-beta related angiitis; G-PACNS, granulomatous pattern; L-PACNS, lymphocytic pattern; N-PACNS, necrotising pattern; PACNS, primary angiitis of the central nervous system.

Figure 2

Age at time of diagnosis. Age differed significantly between groups (p=0.008). While L-PACNS patients had a median age of 47 years (IQR 35–53.5), ABRA patients were 64.5 years (IQR 56.25–78), G-PACNS patients 37 years (IQR 25–42) and N-PACNS subjects 65 years of age. ABRA, amyloid-beta related angiitis; G-PACNS, granulomatous pattern; L-PACNS, lymphocytic pattern; N-PACNS, necrotising pattern; PACNS, primary angiitis of the central nervous system.

Figure 3

Treatment response to the induction therapy. Induction treatment comprised steroids and cyclophosphamide in the majority of the cases. The worst treatment response was observed in G-PACNS patients, while most of the other patients improved under the therapy. ABRA, amyloid-beta related angiitis; G-PACNS, granulomatous pattern; L-PACNS, lymphocytic pattern; N-PACNS, necrotising pattern; PACNS, primary angiitis of the central nervous system.

Figure 4

Outcome of the BP-PACNS study group. (a) mRS at last follow-up. At the last follow-up, G-PACNS patients showed the worst outcome assessed by the mRS. Median mRS at last follow-up was 4 (IQR 2–5.75) in G-PACNS patients, 3 (IQR 1–4.5) in L-PACNS patients and 2 (IQR 1.25–2.75) in ABRA and N-PACNS subjects. (b) Mortality in the study population. The numerically highest mortality (25%) was seen in the G-PACNS subgroup. L-PACNS patients died in 12%, ABRA in 13% of the cases and none of the N-PACNS subjects. ABRA, amyloid-beta related angiitis; BP-PACNS, biopsy-proven PACNS; G-PACNS, granulomatous pattern; L-PACNS, lymphocytic pattern; mRS, modified Rankin Scale; N-PACNS, necrotising pattern; PACNS, primary angiitis of the central nervous system.