# A corona-like distribution and patchy pattern of cerebellar infarcts identify patients with giant cell arteritis

## TL;DR

This study identifies unique brain infarct patterns in patients with giant cell arteritis, which can help improve diagnosis and treatment timing.

## Contribution

The study introduces a novel imaging-based diagnostic signature for intracranial giant cell arteritis using cerebellar infarct patterns.

## Key findings

- A corona-like infarct pattern showed 79% sensitivity and 64% specificity for GCA.
- A patchy infarct pattern was present in 53% of GCA cases with 93% specificity.
- Combining both patterns increased specificity to 98% but reduced sensitivity to 47%.

## Abstract

Cerebrovascular events are a potentially serious complication of giant cell arteritis (GCA) with intracranial involvement. However, diagnosing GCA in this context remains challenging, as classical clinical features may be absent.

To identify characteristic cerebellar infarct patterns associated with intracranial GCA and to differentiate them from other common causes of posterior circulation stroke.

Multicenter retrospective study.

A total of 125 patients with cerebellar infarctions of various etiologies were included. Among these, 19 patients had confirmed intracranial GCA. Infarct patterns were compared to those seen in strokes of cardioembolic origin (n = 42), arterio-arterial embolism from proximal vertebral artery atherosclerosis (n = 13), local atherosclerotic stenosis of the V4 segment (n = 21), and vertebral artery dissection (n = 30). Infarct topography was assessed using acute-phase diffusion-weighted magnetic resonance imaging. Sensitivity and specificity were calculated for individual imaging features.

Distinct imaging signatures were observed in patients with GCA. A “corona-like” infarct pattern, defined by sparing of the medial branch of the proximal posterior inferior cerebellar artery (PICA), demonstrated a sensitivity of 79% and a specificity of 64%. A patchy infarct pattern, characterized by scattered non-confluent lesions, was present in 53% of GCA cases and showed high specificity (93%). When both features were present, specificity increased to 98% and sensitivity was reduced to 47%.

Our findings reveal a distinct cerebellar infarct pattern associated with intracranial GCA, characterized by a corona-like configuration and patchy lesions predominantly involving the lateral PICA territory. Recognition of this imaging phenotype may enhance diagnostic accuracy in challenging cases and facilitate the timely initiation of immunosuppressive therapy.