ABSTRACT
INTRODUCTION: This study seeks to describe the predictive value of the signal intensity ratio (SIR) in magnetic resonance imaging-turbo inversion recovery magnitude (MRI-TIRM) in patients with Graves' orbitopathy (GO) with regard to predictability of therapy response. METHODS: Included in this prospective pilot study were 36 consecutive patients with GO and 25 control subjects. Patients were clinically assessed according to the European Group on Graves' Orbitopathy recommendations with active GO defined by a clinical activity score (CAS) > or = 3. On magnetic resonance (MR) imaging, muscle inflammation was measured with a region of interest set within the brightest extra-ocular muscle both on coronal turbo inversion recovery magnitude (TIRM) and on fat suppressed gadolinium-enhanced T1-weighted sequences. To calculate the SIR, the measured signal intensity was set in proportion to that of the ipsilateral temporalis muscle. RESULTS: Signal intensity ratio in coronal T2-weighted TIRM sequences in either group ranged from 1.22 to 4.92 (mean 2.04) in patients with GO and from 1.18 to 2.4 (mean 1.63) in controls without GO. The observed differences were significant on the TIRM sequences (right eye p = 0.023; left eye p = 0.022), whereas, no significant differences could be detected on the T1-weighted sequences (right eye p = 0.396; left eye p = 0.498). A cut off value of SIR > 2.5 for a CAS > or = 4 to discriminate active from inactive patients was statistically calculated. CONCLUSION: T2 relaxation time is a reliable tool in detecting active GO. The difference in T2-SIR versus T1-SIR is helpful to distinguish inflammatory oedema of the extra ocular muscles from intra-orbital congestion due to reduced venous outflow.
