Structured Vascular Segment Analysis for Giant Cell Arteritis: A Proposed Scoring System for CT Angiography-Based...

Imaging for giant cell arteritis (GCA) has traditionally relied on ultrasound, MRI, and, more recently, nuclear medicine. Computed tomographic angiography (CTA) has not been systematically evaluated, despite near-universal availability, rapid acquisition, and comprehensive neck vessel assessment. We evaluated the ability of CTA to discriminate between biopsy-proven GCA patients and age- and sex-matched controls by assessing cervical arterial abnormalities across multiple cervical arterial segments.

This retrospective single-center study included 20 biopsy-proven GCA cases (with CTA performed within 4 weeks before or 2 weeks after biopsy) and 20 age- and sex-matched controls. Two neuroradiologists independently assessed 21 predefined arterial segments for abnormalities (absent/present). Segment-level positivity was defined by circumferential wall thickening, >50% stenosis/occlusion, or perivascular inflammatory fat-stranding. Vessel-group-level involvement was through consensus positivity (both readers positive; either side). Per-patient total scores were evaluated with ROC (DeLong AUC), prespecified thresholds with secondary exploratory analysis through McNemar testing, and inter-reader agreement by percent agreement and Cohen's κ.

Median age was 76 years (IQR, 70-84); 60% were female in both groups. Across 806 evaluable segments, involvement was most frequent along the maxillary (80%; 16/20), facial (75%; 15/20), and superficial temporal (70%; 14/20) arteries. Total-score discrimination was excellent (AUC 0.971 for both readers). At total score ≥3, sensitivity/specificity were 0.90/1.00 for both readers, with paired differences favoring cases (exact McNemar p<0.001). Pooled segment-level agreement was high (agreement 0.978; κ 0.924).

Structured multi-segment CTA scoring of cranio-cervical arterial segments showed excellent discrimination of biopsy-proven GCA from matched controls, supporting CTA as a useful diagnostic tool in GCA patients.