Diagnostic utility of the HAS-GCA score versus other Southend Giant Cell Arteritis Probability Score-ultrasound prediction models in giant cell arteritis.

Abstract

Objective: The aim of this study was to validate the HAS-GCA score, a prediction model combining the Southend Giant Cell Arteritis Probability Score (SGCAPS) with ultrasonography (US) halo count, and compare its diagnostic utility with other prediction models combining SGCAPS with US scores for diagnosing giant cell arteritis (GCA).

Method: : Treatment-naïve patients with suspected GCA were prospectively included in an observational cohort and underwent clinical assessment, including temporal artery biopsy (TAB) and pretreatment US and [¹⁸F]fluorodeoxyglucose-positron emission tomography/computed tomography. The clinical diagnosis was confirmed after 24 weeks. Baseline data were assessed to calculate SGCAPS. Three prediction models combining SGCAPS with US scores - halo score, OMERACT GCA ultrasound score (OGUS), and halo count - were applied to classify patients into low-, intermediate-, and high-probability categories of GCA. The ability of the prediction models to correctly classify patients into low- and high-probability categories, and the proportion of patients with intermediate probability requiring additional tests, were compared.

Results: : GCA diagnosis was confirmed in 60/99 patients (61%). Prediction models based on halo count (HAS-GCA score), halo score, and OGUS had similar misclassification rates of 3%; HAS-GCA score had the lowest proportion of patients with intermediate probability (39% vs 47% and 59%, respectively).

Conclusion: : The HAS-GCA prediction model reliably stratified patients with suspected GCA into low- and high-probability categories, enabling confidence in ruling in or ruling out the GCA diagnosis. However, almost half of the patients required further evaluation to establish a correct diagnosis.