Can Imaging Predict Subclinical Cortisol Secretion in Patients With Adrenal Adenomas? A CT Predictive Score

The aim of this study is to determine whether any correlation between CT findings and functional parameters exists to predict subclinical glucocorticoid secretion.

This is a retrospective database study of 55 patients with incidentally discovered adenomas, investigated through CT with an adrenal protocol, assessing diameters and attenuation values on the unenhanced and contrast-enhanced phases. Patients underwent blood cortisol and corticotropin evaluation and overnight dexamethasone suppression test (DST), in accordance with clinical recommendations. Cortisol levels higher than 50 nmol/L after DST identified subclinical cortisol secretion. We identified 28 subjects with lipid-rich nonsecreting adenomas, nine with lipid-rich secreting adenomas, 11 with lipid-poor nonsecreting adenomas, and seven with lipid-poor secreting adenoma.

Cortisol levels after DST were significantly and positively related to mass diameters. At univariate analysis, maximum and minimum diameters and attenuation in the delayed phase were significantly related to the presence of secreting or nonsecreting adenoma; at multivariate analysis, only the minimum diameter and the attenuation in the venous phase entered the stepwise logistic regression. Similarly, minimum diameter and attenuation in the venous phase emerged also at the multivariate stepwise regression between radiologic parameters and cortisol levels after DST. The formula of the radiologic score computed by using the coefficients of the multivariate regression was as follows: (0.1914 × minimum diameter) + (0.0308 × enhanced attenuation). The diagnostic accuracy of this discriminatory score in differentiating secreting from nonsecreting adenomas was 84.9%, the sensitivity was 81.3%, and the specificity was 87.2%. Adenomas with scores greater than 7.59 were considered as secreting adenomas, and adenomas with scores less than 7.36 were considered as nonsecreting adenomas.

This study shows that imaging parameters can predict subclinical cortisol hypersecretion in patients with adrenal adenomas.