Background:Ultrasound of the temporal ± axillary arteries showing a non-compressible halo sign is recommended for diagnosing patients with giant cell arteritis (GCA); however, its value for monitoring disease activity is still poorly understood.Objectives:To assess the sensitivity to change of ultrasound halo features and their association with disease activity and glucocorticoid (GC) treatment in patients with newly diagnosed GCA.Methods:Two centre prospective study of new patients with ultrasound confirmed-GCA who underwent serial ultrasound assessments of the temporal (TA) and axillary (AX) arteries at fixed time-points. The number of arterial segments with halo and the maximum halo intima-media thickness (IMT) per segment was recorded at each visit. Only time-points in which >80% of patients were assessed were considered for analysis. Sensitivity to change of the halo sign was calculated as standardised mean difference (SMD) for each time-point separately. Correlation between ultrasound findings and erythrocyte sedimentation rate (ESR), C-reactive protein (CRP), Birmingham Vasculitis Activity Score (BVAS), and GC treatment was performed using Spearman’s correlation coefficient. Logistic regression was used to determine the probability of being in remission for each unit increase (standardised) of the halo feature of interest. Remission was defined as the absence of relapse with a prednisolone dose <30 mg/day, and relapse as the recurrence of GCA-related symptoms or rise of CRP/ESR values not otherwise explained and requiring GC increase. Halo features at disease onset and first relapse were compared using Wilcoxon test.Results:A total of 49 patients (73.5% females) with a mean age of 78.2 ± 7.4 years at baseline were assessed at 354 visits. TA involvement was reported in 47 (95.9%) patients and AX involvement in 11 (22.4%); 220 arterial segments with halo were recorded (201 TA, 19 AX). Halo sensitivity to change was calculated for weeks 1, 3, 6, 12 and 24, and showed a significant SMD between all time-points and baseline for the TA halo features, but only after week 6 for the AX halo features (Table and Figure). The number of TA segments with halo, sum of TA halo IMT and maximum TA halo IMT showed a significant correlation with ESR (0.41; 0.44; 0.48), CRP (0.34; 0.39; 0.41), BVAS (0.29; 0.36; 0.35) and GC cumulative dose (-0.34; -0.37; -0.32). The likelihood of achieving disease remission was lower in patients with a higher number of TA segments with halo (OR 0.39, p<0.05) and increased values of sum and maximum TA halo IMT (OR 0.34, p<0.05). By contrast, AX halo features showed no correlation with disease activity, nor any association with attaining clinical remission. During the study period, 32 relapses were observed (mean time for first relapse of 31.8 weeks ± 18.5 days). Halo sign was present in 16/17 (94.1%) cases of first disease relapse, all showing an increased halo IMT (sum and maximum) in relation to the previous ultrasound assessment performed. When compared to disease onset, a lower mean number of segments with halo and mean sum of halo IMT was reported for patients presenting with their first clinical relapse (2.93 ± 1.59 vs. 4.85 ± 1.51, p=0.0012; 2.01 ± 1.13 vs. 4.49 ± 1.95 mm, p=0.0012).Table 1.SMD of halo features between baseline and different time-pointsWeek 1Week 3Week 6Week 12Week 24N of arterial segmentsSum of all segments with halo (n=49)-0.51-0.78-1.13-1.69-1.52Sum of TA segments with halo (n=47)-0.49-0.78-1.18-1.87-1.69Sum of AX segments with halo (n=11)-0.35--0.62-0.73-0.91Halo thickness (mm)Sum of all halo IMT (n=49)-0.98-1.44-1.37-1.60-1.48Sum of TA halo IMT (n=47)-1.01-1.55-1.54-1.81-1.69Sum of AX halo IMT (n=11)-0.15-0.45-0.81-0.84-0.98Max. TA halo IMT (n=47)-1.07-1.32-1.47-1.91-2.19Max. AX halo IMT (n=11)-0.04-0.29-0.94-1.13-1.01In bold p<0.05; n=n at baselineConclusion:Ultrasound is a reliable imaging tool to assess disease activity and response to treatment in patients with GCA. Future clinical trials in GCA should evaluate direct treatment effect on halo features as an outcome measure of interest.Acknowledgements:The first two authors contributed equally to this workDisclosure of Interests:None declared