Graves' ophthalmopathy: eye muscle involvement in patients with diplopia.

BACKGROUND: Diplopia identifies patients with eye muscle involvement in Graves' ophthalmopathy (GO).
OBJECTIVE: To identify clinical parameters that could eliminate the need for magnetic resonance imaging (MRI) to assess the activity of inflammation in the eye muscles of GO patients with diplopia.
METHODS: In 43 patients with GO with recently developed diplopia, orbital ultrasound and MRI were performed. Muscle diameters and MRI T2 relaxation times were measured, and the amount of orbital connective tissue was calculated from MRI scans and compared with ultrasound readings, diplopia grades, degree of protrusion, ocular pressure, tear production, antibody levels and hormonal parameters of thyroid function.
RESULTS: No correlation was found between diameters of 233 extraocular muscles measured by MRI and by ultrasound. For each of the four muscles, there was a diameter above which ultrasound was always unreliable. MRI data were used in further analysis. Of the muscles examined, the inferior rectuses were the most frequently enlarged - at least one, in 93% of cases. Medial, lateral and superior rectuses were enlarged in 59%, 37% and 34% of the orbits respectively. The pattern of muscle involvement of the two orbits tended to be symmetric (r=0.49, P=0.003), particularly for the medial rectuses (r=0.90, P=0.000). Proptosis correlated with the sum of the muscle diameters for a given eye (right eye: r=0.54, P=0.003; left eye: r=0.57, P=0.001), but it failed to correlate with the amount of orbital connective tissue. In 53% of the patients, normal T2 relaxation times were found in all eight muscles. There was only a weak correlation between muscle thickness and T2 relaxation time (r=0.49, P=0.003), indicating that muscle enlargement alone is not a sign of disease activity. The severity of diplopia was independent of T2 relaxation time. The amount of orbital connective tissue showed a negative correlation with the greatest T2 relaxation time for a given eye (r= -0.52, P=0.004); this suggests that disease types exist that have predominant muscle involvement and predominant connective tissue expansion. No correlation between connective tissue expansion and proptosis, diplopia grade, muscle thickness or disease duration was found - that is, connective tissue expansion is not a major factor in diplopia. Both muscle and connective tissue findings were independent of thyroid function.
CONCLUSION: Ultrasound and MRI eye muscle diameter readings do not correlate, because of the inherent inaccuracy of orbital ultrasound. Muscle enlargement alone does not mean oedematous swelling and active disease. Neither ultrasound, nor any combination of 11 clinical and laboratory parameters provided the degree of information on muscles and connective tissue that was obtainable by MRI. In unclear cases of recently developed diplopia, before orbital decompression surgery, in the case of treatment failure or if, for any other reason, imaging is needed in GO, MRI is the method of choice.