Correlation of vascular enhancement on MRI and clinical manifestations in patients with GCA.

GCA has diverse cranial manifestations, and cranial vessel MRI (cvMRI) can assess multiple arteries simultaneously. We aimed to determine whether vessel-specific enhancement on cvMRI correlates with clinical manifestations and diagnosis in GCA. We conducted a retrospective cohort study of patients who underwent cvMRI for suspected GCA at a single academic centre. We assessed associations between vessel wall enhancement on cvMRI and anatomically expected GCA symptoms. We used univariate and multivariable logistic regression adjusting for age, sex and glucocorticoid exposure to assess the relationship between vessel abnormalities and a diagnosis of GCA. Among 439 patients, 153 (34.9%) were diagnosed with GCA. Median time (quartile 1 [Q1]-quartile 3 [Q3]) from glucocorticoid initiation to cvMRI was 11.5 days (Q1-Q3: 5.8-207.2). Maxillary and ophthalmic artery involvement were specific but insensitive for jaw claudication and vision loss, respectively (specificity 87.1%, 95% CI, 80.3-93.9 and 72.6%, 95% CI, 61.5-83.7; sensitivity 40.0%, 95% CI, 26.5-51.4 and 43.3%, 95% CI, 33.1-53.6, respectively). Headache showed no association with specific vessels. Any arterial enhancement was associated with GCA in univariate analysis, but only temporal artery involvement was predictive in multivariable analysis (odds ratio [OR] 5.20, 95% CI, 1.65-16.88), particularly when bilateral (OR 9.82, 95% CI, 2.82-38.32). Temporal artery enhancement remains the strongest imaging correlate for a diagnosis of GCA. Vessel-specific enhancement on cvMRI correlates with vision loss and jaw claudication but not headache. cvMRI may aid diagnosis in diagnostically uncertain cases, even after glucocorticoid initiation.