Metastatic disease is definitive of malignancy [23]. However, several imaging features should increase the suspicion of ACC within an adrenal mass [24]: tumor diameter >4 cm, irregular tumor margins, central intratumoral necrosis or hemorrhage, heterogeneous enhancement, invasion into adjacent structures, venous extension (renal vein or inferior vena cava (IVC)), and calcification. ACCs are usually large at presentation, ranging from 2 cm to 25 cm (average size ≈9 cm). Approximately 70% of ACCs are larger than 6 cm. They are bilateral in 2–10% of cases and are slightly more common on the left than on the right. Tumors are frequently hemorrhagic and necrotic, and may contain small areas of intracytoplasmic lipid or fatty regions. Using a logistic regression model, Hussain et al. found tumor size >4 cm and heterogeneous enhancement to be the most important discriminators of malignancy [25]. The existence of intracytoplasmic fat in ACCs has been attributed to cortisol and related fatty precursors in hormonally active tumors. On occasion, pockets of fat may be seen within the mass, indicating coexistent myelolipomatous tissue. Invasion into the IVC has been reported in 9–19% of ACC cases at presentation.

The typical appearance of ACC on unenhanced CT is of a large, inhomogeneous (but well-defined) suprarenal mass that displaces adjacent structures as it grows [26]. Regions of low attenuation correspond to necrosis pathologically. After the intravenous administration of contrast material, there is inhomoge- neous enhancement of the tumor, typically with greater enhancement seen peripherally and relatively little enhancement seen centrally, because of central necrosis [27]. Measurement of the attenuation of adrenal lesions on unenhanced CT is of great value in distinguishing between benign and malignant masses. Cumulative data obtained for the identification of adrenal adenomas indicate that ACCs rarely have an attenuation <10 Hounsfield units (HU). The specificity of this threshold for the identification of benign adenomas is ≈98%. Equally, ACCs retain intravenous contrast material and have absolute and relative percentage washout of <60% and <40%, respectively, 15 min after contrast administration or <50% and <40%, respectively, at 10 min [28]. Calcification (micro or coarse) is seen on CT in ≈30% of patients with ACC and is usually centrally located (Fig. 10.3). Calcification is rare in adenomas, but is present in ≈10% of pheochromocytomas [29]. Tumor thrombus extending into the IVC at presentation is not rare and is more frequently seen in right-sided tumors. A tumor thrombus within a vein is usually well encapsulated and can often be withdrawn intact from the vein [30]. The cephalad extent of the tumor thrombus can be identified on contrast-enhanced CT or magnetic resonance imaging (MRI).

CT is also of value in showing the local and distant spread of an ACC. Preservation of fat planes around the tumor indicates that there is no local invasion. Where there is a paucity of retroperitoneal fat, it may not be possible to determine if the tumor has invaded adjacent organs. Metastases are frequently found at presentation: regional and para-aortic lymph nodes (25–46%), lungs (45–97%), liver (48–96%), and bone (11–33%) are the common sites. Hepatic metastases tend to be hyper-vascular and are best seen on arterial-phase imaging after intravenous administration of contrast.

ACCs are typically heterogeneous in signal intensity on MRI because of hemorrhage and/or necrosis [31]. On T1-weighted imaging, ACCs are typically isointense or slightly hypointense to normal liver parenchyma. However, high T1 signal intensity is often seen because of hemorrhage. On T2-weighted imaging, ACCs are usually hyperintense to liver parenchyma and have a heterogeneous texture because of intratumoral cystic regions and hemorrhage [32]. A functioning ACC can contain small regions of intracytoplasmic lipid, resulting in small non-uniform areas of loss of signal on chemical shift imaging (<30% of the lesion) [33]. Although similar small non-uniform loss can occur in lipid-poor adenomas, the significant uniform signal loss seen in lipidrich adenomas does not occur.

The results of early studies suggested that proton MR spectroscopy may be useful in differentiating adrenal adenomas and pheochromocytomas from adrenal metastases and ACCs. Faria et al. looked at the spectral traces obtained from 60 patients with adrenal masses. Adenomas and pheochromocytomas could be differentiated from ACCs and metastases using choline:creatine ratios of >1.20 (92% sensitivity and 96% specificity) and choline:lipid ratios of >0.38 (92% sensitivity and 90% specificity). ACCs and pheochromocytomas could be differentiated from adenomas and metastases using a 4.0–4.3 ppm/creatine ratio >1.50 (87% sensitivity and 98% specificity). By combining these two spectral analyses, they could divide adrenal mass lesions into one of four distinct groups: adenoma, pheochromocytoma, ACC, or metastasis [34]. Although there were some criticisms of this study, the method appears to offer potential in helping to distinguish among adrenal mass lesions.

Fluorodeoxyglucose-positron emission tomography (FDG-PET) can be used to identify certain malignant adrenal masses by virtue of their increased metabolic activity; however, if FDG uptake is only modest, the likelihood of benign vs malignant is approximately equal [35]. FDG-PET combined with contrastenhanced CT has a sensitivity of 100% and specificity of 87–97% for identifying malignant adrenal masses. The lower specificity is because a small number of adenomas and other benign lesions mimic malignancy [36]. The novel PET tracer 11C metomidate (a marker of 11β-hydroxylase) is used as a tracer for adrenocortical tissue and is taken up by adenomas and ACCs. This marker differentiates adrenal cortical lesions from pheochromocytomas and metastases, which are uptake-negative [37]. However, the most valuable aspect of PET is its ability to detect distant metastases (one-third of patients with ACCs will have metastatic disease at presentation).