Matthias Froehlich1, Thorsten A. Bley2, Marc Schmalzing3, Michael Gernert1, Rudolf A. Werner4, Jost Hillenkamp5, Karl Georg Haeusler6 and Konstanze V. Guggenberger2, 1University Hospital Wuerzburg, Dpt. of Internal Medicine II, Wuerzburg, Germany, 2University Hospital Wuerzburg, Dpt. of Radiology, Wuerzburg, Germany, 3University Hospital Wuerzburg, Wuerzburg, Germany, 4University Hospital Wuerzburg, Dpt. of Nuclear medicine, Wuerzburg, Germany, 5University Hospital Wuerzburg, Dpt. of Ophtalmology, Wuerzburg, Germany, 6University Hospital Wuerzburg, Dpt. of Neurology, Wuerzburg, Germany
Background/Purpose: Imaging techniques for the detection of mural inflammation are cornerstones in the diagnosis of giant cell arteritis (GCA). Magnetic resonance imaging (MRI) and ultrasound are central in this regard, but data comparing both methods are scarce.
Methods: Case-control study of 88 patients (mean age 70±9.6 years, 64% women), 44 with GCA and 44 with initial suspicion of GCA who finally did not have GCA. The standardized 3 Tesla MRI protocol included the temporal, occipital, vertebral, carotid, subclavian, and axillary arteries, the thoracic and abdominal aorta including the renal arteries. The standardized ultrasound protocol included the temporal, carotid, and axillary arteries using a linear array transducer at 18 MHz. Trained MRI and ultrasound diagnosticians blinded to other diagnostic findings or clinical data. Correctness of the GCA diagnosis was prospectively confirmed by the Rheumatology Department after 3 months by clinical examination.
Results: Overall, 44 patients with GCA (37 at initial diagnosis and 7 at GCA-relapse) and 44 controls were examined. MRI and ultrasound were performed on the same day in 81 patients. Time difference between MRI and ultrasound was median 0 (IQR 0) days. At the time of investigation, 21 of 44 GCA patients were therapy-naïve.
Ultrasound detected signs of vasculitis in 20 of 44 GCA patients (sensitivity 45%). In patients with supraaortic GCA, 20 of 34 (sensitivity 59%) patients were identified to have GCA. In patients on steroids (duration mean 60 days (IQR 67 days), ultrasound identified 11 of 23 patients (sensitivity 48%) to have GCA. The sensitivity of MRI was 73% (32 of 44 GCA patients), demonstrated occipital artery vasculitis in 22 of 44 patients. In supraaortic GCA, 27 of 34 (sensitivity 79% patients were identified by MRI.
In comparison, MRI diagnosed 13 additionally arteritis patients compared to ultrasound. Of those, arteritis was found in the thoracic aorta (n=3), the abdominal aorta (n=4), occipital artery (n=3), and vertebral artery (n=2) or combinations thereof. In 5 patients, ultrasound did not reveal vasculitis-typical findings, although they were detectable on MRI.
In the control group, all 44 patients were negative for vessel wall inflammation on both, MRI and ultrasound (specificity 100%). The main diagnoses in the control group were polymyalgia rheumatica (n=21), and non-arteritic anterior ischemic optic neuropathy (n=20).
Conclusion: MRI is superior to ultrasound in detecting inflammatory activity in GCA as more vessels are examined and evaluation of MRI is more sensitive. Involvement of occipital arteries is a common manifestation in GCA that is neither easily detectable by ultrasound nor readily amenable to biopsy but can be detected by MRI. Involvement of aorta and its branches is also common, which, with the exception of axillary arteries, cannot be recorded structurally in US.
Disclosures: M. Froehlich, None; T. Bley, None; M. Schmalzing, Novartis, AbbVie, AstraZeneca, Chugai/Roche, Janssen, Gilead, Boehringer/Ingelheim, Celgene, Medac, UCB, Sandoz; M. Gernert, None; R. Werner, None; J. Hillenkamp, None; K. Haeusler, None; K. Guggenberger, None.
