According to the American Journal of Roentgenology (AJR), adrenal-protocol CT has limited utility in the evaluation of heterogeneous adrenal nodules.
“Adrenal-protocol CT has poor diagnostic performance for distinguishing adenomas from non-adenomas among heterogeneous adrenal nodules, regardless of the method used for region of interest (ROI) placement,” writes Michael T. Corwin, MD, an abdominal radiologist at the University of California Davis Medical Center in Sacramento.
Corwin and his colleagues’ AJR-accepted manuscript included 164 patients (mean age, 59.1 years; 61 men, 103 women) from seven institutions, with a total of 164 heterogeneous adrenal nodules evaluated by adrenal-protocol CT (all nodules had an available pathologic reference standard).
At each institution, a single investigator evaluated CT images, and ROIs were placed on portal venous phase images via four ROI methods: standard, single large in nodule center; high, single on nodule’s highest-attenuation area; low, single on nodule’s lowest-attenuation; average: mean of three ROIs on nodule’s superior, middle, and inferior thirds using the standard approach. After placing ROIs in identical locations on unenhanced and delayed phases, absolute washout was determined for all methods.
Ultimately, in the AJR authors’ seven-institution study, adrenal-protocol CT had poor diagnostic performance (sensitivity, 57%-68%; specificity, 51%-85%) for distinguishing adenomas from non-adenomas among heterogeneous adrenal nodules, regardless of the method used for ROI placement.
Featured image: Axial CT images in unenhanced (A), portal venous (B), and delayed (C) phases show 5.9-cm heterogeneous left adrenal mass. Absolute washout was less than 60% for all four ROI methods, indicating diagnosis of non-adenoma: standard ROI (black circle), -43%; high ROI (dashed white circle), 42%; low ROI (solid white circle), 26%; average ROI, XX%. Surgical resection yielded pathologic diagnosis of adenoma. Thus, absolute washout yielded false-negative result for all four ROI methods.