Characterization of adrenal masses using unenhanced CT: an analysis of the CT literature.

TLDR

Unenhanced CT can reliably characterize incidentally detected adrenal masses using density measurements, but the optimal density threshold to differentiate benign from malignant lesions remains controversial. This study aims to establish a consensus on the optimal density threshold by performing a pooled analysis of CT literature. The authors analyzed ten CT reports, extracting density measurements for 495 adrenal lesions (272 benign, 223 malignant). Threshold analysis revealed sensitivities ranging from 47 % to 88 % and specificities from 100 % to 84 % across thresholds, indicating that the chosen threshold depends on patient population and cost‑benefit considerations, and that striving for absolute certainty of benignity can result in unacceptably low sensitivity.

Abstract

Unenhanced CT scanning can reliably characterize incidentally detected adrenal masses when observers use density measurements of the adrenal gland. However, controversy exists as to the optimal density threshold required to differentiate benign from malignant lesions. This study attempts to establish a consensus by performing a pooled analysis of data found in the CT literature. Ten CT reports were analyzed, from which individual adrenal lesion density measurements were obtained for 495 adrenal lesions (272 benign lesions and 223 malignant lesions). Threshold analysis generated a range of sensitivities and specificities for lesion characterization at different density thresholds. Sensitivity for characterizing a lesion as benign ranged from 47% at a threshold of 2 H to 88% at a threshold of 20 H. Similarly, specificity varied from 100% at a threshold of 2 H to 84% at a threshold of 20 H. The attempt to be absolutely certain that an adrenal lesion is benign may lead to an unacceptably low sensitivity for lesion characterization. The threshold chosen will depend on the patient population and the cost-benefit approach to patient care.