Computed tomography versus magnetic resonance imaging for the evaluation of suspected pituitary adenomas

The purpose of this study was to evaluate and compare thin-section magnetic resonance imaging (MRI) and high-resolution computed tomography (CT) in patients with suspected pituitary adenomas. Twenty-two patients (19 women and three men) with hyperprolactinemia (N = 16), increased growth hormone secretion (N = 2), increased corticotropin secretion (N = 1), and nonsecreting adenomas (N = 3) were studied with both contrast-enhanced, high-resolution CT scanning and thin-section MRI. Contrast-enhanced examinations consisted of contiguous 1.5-mm coronal sections during contrast infusion. The MRI examinations consisted of spin-echo T1- and T2-weighted sequences with a 2.5-3.0-mm slice thickness on the coronal and sagittal planes. Fourteen women had similar findings on CT and MRI (four macroadenomas, six microadenomas, one wide stalk, two empty sellas, and one normal study). The remaining eight subjects had conflicting results: CT findings were compatible with a microadenoma in all eight patients, whereas MRI detected one enlarged pituitary, two empty sellas (one with prolapse of the optic chiasm) without evidence of adenoma, and five normal examinations. Thus, both studies detected macroadenomas accurately, but CT was frequently unable to diagnose correctly an empty sella. Because patients with possible microadenomas were not submitted to surgery, the accuracy of either radiologic method cannot be assessed at this time. However, we suggest that MRI is superior to CT because of its inherently greater soft-tissue contrast, which allows clear visualization of the optic chiasm, optic nerves, cavernous sinuses, and carotid arteries.