Diagnostic Value of Intra-orbital Oculomotor Nerve Short Tau Inversion Recovery Hyperintensity in Oculomotor Nerve Palsy.

BACKGROUND AND PURPOSE: Oculomotor nerve palsy has diverse etiologies, with MRI, including fat-suppressed contrast-enhanced T1WI, widely used for evaluation. Although intra-orbital oculomotor nerve STIR hyperintensity has been reported in diabetic oculomotor nerve palsy, its consistency across diverse oculomotor nerve palsy etiologies remains unclear. We evaluated this and determined whether orbital STIR provides complementary imaging information in oculomotor nerve palsy evaluation by comparing STIR with a GRE-based contrast-enhanced T1WI protocol. MATERIALS AND METHODS: This retrospective, single-center study included 31 consecutive patients with oculomotor nerve palsy and 43 normal controls who underwent orbital 2D STIR imaging. The index test was intra-orbital oculomotor nerve STIR hyperintensity; the reference standard was clinical diagnosis. Contrast-enhanced T1WI was available in 25 patients and 30 normal controls. Two neuroradiologists independently evaluated the oculomotor nerve, with emphasis on the intra-orbital segment and supplementary assessment of the cisternal and cavernous segments. Diagnostic performance was assessed using lesion- and patient-based analyses. Sequences were compared using the McNemar test. RESULTS: A total of 74 participants (31 oculomotor nerve palsy, 43 normal controls; 40 men, 34 women; median age 68 years [interquartile range, 52-76]) were evaluated. In the lesion-based analysis, intra-orbital STIR hyperintensity showed the highest sensitivity and accuracy among the three STIR-evaluated segments as an imaging correlate of clinically diagnosed oculomotor nerve palsy (sensitivity, 90.3%; specificity, 100%; accuracy, 95.9%). In the same segment, GRE-based contrast-enhanced T1WI showed lower sensitivity and accuracy (sensitivity, 12.0%; specificity, 96.7%; accuracy, 58.2%). In the patient-based analysis (n = 55), STIR showed higher accuracy (94.5% vs 78.2%; P = .004) and sensitivity (92.0% vs 64.0%; P = .016), with comparable specificity (96.7% vs 90.0%; P = .50). Among 25 patients with oculomotor nerve palsy evaluated using both sequences, STIR demonstrated imaging abnormalities corresponding to the clinically affected oculomotor nerve in seven patients (28.0%) in whom no corresponding abnormality was identified on contrast-enhanced T1WI. CONCLUSIONS: Intra-orbital oculomotor nerve STIR hyperintensity was frequently observed across diverse etiologies as an imaging correlate of clinically diagnosed oculomotor nerve palsy. Orbital STIR imaging may serve as a useful non-contrast complement to conventional MRI protocols in the evaluation of oculomotor nerve palsy.