Magnetic resonance imaging and tendon anomaly associated with congenital superior oblique palsy

PURPOSE: To assess the relationship between tendon anomalies and the volume of the superior oblique muscle in patients with congenital unilateral superior oblique palsy. METHODS: Thirty-three patients with unilateral con genital superior oblique palsy were enrolled. Coronal, cross-sectional magnetic resonance imaging scans of the superior oblique muscle were obtained, and the volume of the paretic superior oblique muscle belly was calculated as a percentage of the superior oblique muscle belly on the normal side. The percentage volume of the affected superior oblique muscle was compared retrospectively with the angle of the vertical deviation in the primary position, the tendon looseness determined by a traction test, and other intraoperative findings. RESULTS: When a tendon was loose, the volume of its muscle belly was significantly smaller than the belly of muscles with a normal taut tendon (Mann-Whitney U test, P =.0005). The average vertical deviation of patients assessed to have loose tendons was 4.80 prism diopters, and the deviation in patients with normal tendons was 9.90 prism diopters, The mean vertical deviation of patients with atrophic muscle belly on magnetic resonance imaging was 18.1 prism diopters, and that with normal muscle structure was 10.1 prism diopters, The cases with loose tendon as determined by the traction test after administration of general anesthesia and the cases with atrophic muscle belly had significantly larger vertical deviation in the primary position than the cases with normal tendons and muscles. (Mann-Whitney U test, P = .01 and .0196, respectively). CONCLUSIONS: The traction test is sensitive enough to detect anomalies of the superior oblique tendons. Anomalous superior oblique tendons are nearly always associated with attenuated superior oblique muscle and this information provides us with an explanation for the phenomenon of laxity of the superior oblique tendon, (Am J Ophthalmol 1999;127:379-387. (C) 1999 by Elsevier Science Inc, All rights reserved.).