[Prediction for effectiveness of steroid pulse therapy by MRI in Graves' ophthalmopathy]

Fifteen patients with Graves' ophthalmopathy (GO) were treated with intravenous methylprednisolone (steroid pulse therapy, 1g daily for 3 days a week, 2-4 times) and followed up by ophthalmological assessment and magnetic resonance imaging (MRI). The signal intensity of enlarged eye muscle and retrobulbar fat was examined with MRI at 0.5T with short inversion time inversion recovery (STIR) sequences. The signal intensity of eye muscle and retrobulbar fat tissue in STIR was evaluated as the ratio to cerebral substantia alba (signal intensity ratio). The thickness of enlarged eye muscle was measured by T1-weighted coronal images. The signal intensity ratios of enlarged eye muscle of GO patients were significantly higher than those of eight normal subjects. Although the signal intensity ratios of muscle and retrobulbar fat before therapy were not related to the severity of clinical findings of GO assessed by ophthalmopathy index, the initial signal intensity ratios of eye muscle and retrobulbar fat of ten patients with improved clinical findings of GO after steroid pulse therapy tended to be higher than those of five patients without improvement by the therapy. After the therapy the signal intensity ratios of muscle and retrobulbar fat were significantly decreased in ten patients with favorable response. Our data suggested that high signal intensity in STIR may reflect edema caused by acute inflammation associated with GO. In conclusion, MRI may be a useful tool for determining the indication and prognosis of steroid pulse therapy. We strongly recommend measuring the signal intensity of eye muscle as well as muscle thickness in MRI to evaluate the activity of GO.