Dynamic pituitary MRI has high sensitivity and specificity for the diagnosis of mild Cushing's syndrome and should be part of the initial workup.

AIM: The diagnosis of mild or episodic Cushing's syndrome is difficult. The standard tests include 24-hour urinary free cortisol (UFC), night-time blood, or salivary cortisol measurements, and dexamethasone suppression tests. Imaging studies of the pituitary have not been recommended as part of the initial workup (only to help distinguish pituitary Cushing's disease from the ectopic ACTH syndrome) because of poor sensitivity and specificity. With the development of dynamic pituitary MRI which uses multiple coronal dynamic sequences following gadolinium intravenous contrast, we hypothesized that the sensitivity and specificity would be increased and MRI would provide useful information for the initial diagnosis of Cushing's syndrome.
METHODS: This was a retrospective chart review examining charts from 87 consecutive patients who were evaluated for Cushing's syndrome in a tertiary Endocrinology clinic over a one-year period. Most patients had mild and/or episodic hypercortisolism. Of these patients, 24 eventually were diagnosed with pituitary Cushing's syndrome by biochemical testing (24-h UFC and urinary 17-hydroxycorticosteroids, 11 PM salivary cortisol measurements, evening plasma cortisol), and 22 had the diagnosis of Cushing's syndrome excluded. Dynamic pituitary MRI (1.5 Tesla) was performed on all patients. The reader of the MRI was blind to the diagnosis.
RESULTS: Twenty-three of 24 patients had a MRI consistent with a pituitary lesion (21 with a microadenoma, two with pituitary asymmetry). In contrast, only 3 of 20 patients (2 patient did not have MRIs) in the Cushing's excluded group had a pituitary lesion on dynamic MRI. Dynamic pituitary MRI had the highest sensitivity and negative predictive value of any testing modalities and its specificity and positive predictive value were similar to that of other tests.
CONCLUSION: We conclude that almost all patients in this series with Cushing's syndrome have a lesion on dynamic pituitary MRI, a rate much higher than the 50-60% rate reported for non-dynamic MRIs. The false positive rate of 16% in our group of Cushing's excluded patients is similar to the literature value of 10% seen in normal volunteers and is acceptable since MRI is not used solely as a determinant for the diagnosis. While a negative MRI will miss those patients with adrenal or ectopic Cushing's syndrome, those patients can usually be diagnosed by other testing. Thus this preliminary study implies that dynamic pituitary MRI adds valuable information to assist in the diagnosis of Cushing's syndrome and should be ordered as part of the initial workup.