Management of Adrenocortical Carcinoma

Bruno Allolio; Stefanie Hahner; Dirk Weismann; Martin Fassnacht


Clin Endocrinol. 2004;60(3) 

In This Article


Hormonal evaluation is mandatory in all patients with suspected ACC ( Table 1 ) and may be associated with improved survival (Icard et al., 1992). Unfortunately, hormone concentrations are usually of limited help in predicting malignancy. However, in the presence of an adrenal lesion, elevated serum dehydroepiandrosterone sulphate (DHEAS) levels suggest an ACC, as benign adrenocortical tumours often exhibit low DHEAS concentrations (Osella et al., 1994; Flecchia et al., 1995; Terzolo et al., 2000a). In addition, elevated serum 17β-oestradiol is a rare but rather typical marker of oestrogen-secreting ACC in men. Accordingly, in male patients with an adrenal tumour and elevated serum 17β-oestradiol, an ACC should be assumed until proven otherwise (Gabrilove et al., 1965). As cortisol hypersecretion is the most common hormone excess in ACC, evaluation for adrenal CS is essential including an overnight dexamethasone suppression test, assessment of urinary free cortisol excretion and determination of plasma ACTH. In case of subclinical CS, a corticotropin-releasing hormone test will predict the risk of adrenal insufficiency after complete tumour removal (Reincke et al., 1992).

Aldosterone-secreting ACCs are rare and usually present with hypokalaemia and very high serum aldosterone concentration. Aldosterone-seceting tumours smaller than 4 cm with only moderately elevated aldosterone levels are suggestive of a benign adenoma.

At presentation, additional steroids should be measured, as they may also serve as tumour markers during follow-up: urinary excretion of 17-ketosteroids, 17-α-hydroxyprogesterone, 11-deoxycortisol, deoxycorticosterone and in women with virilization also androstenedione and testosterone. In advanced ACC, serum LDH may serve as a marker of disease progression. Measurement of urinary catecholamine excretion or plasma metanephrines is mandatory to exclude phaeochromocytoma prior surgery.

The size of the adrenal mass, as measured by computed tomography (CT) or magnetic resonance imaging (MRI) remains the single best indicator of malignancy. In a recent series from France (Icard et al., 2001), mean tumour size at diagnosis was 12·0 ± 6·0 cm (n = 223) and mean tumour weight was 689 ± 822 g (n = 202). Similar results have been found in earlier series (Didolkar et al., 1981) and were confirmed recently (Vassilopoulou-Sellin & Schultz, 2001; Stojadinovic et al., 2002). The likelihood of ACC increases to 35-98% in patients with an adrenal mass > 6 cm (Ross & Aron, 1990). However, in recent years, additional imaging features (e.g. attenuation coefficients) have been used to discern malignancy in adrenal tumours.

A thin-collimation CT is the imaging method of choice for adrenal masses and for differentiation of benign from malignant lesions. Nonadenomatous lesions typically have higher CT density values due to their lower lipid content (Korobkin et al., 1996). ACCs are typically inhomogeneous, with irregular margins and irregular enhancement of solid components after intravenous (i.v.) contrast media. Calcifications are sometimes visible. Dependent on the threshold value of the Hounsfield units, sensitivity and specificity for characterization of an adrenal lesion as a benign adenoma in unenhanced CT ranged from 47% to 100% at a threshold of 2 HU, and from 88% to 84%, respectively, at a threshold of 20 HU (Boland et al., 1998). Recent studies suggest that delayed contrast-enhanced CT scans can be used to further characterize lesions with higher HU in unenhanced scans. As early as 3 min and up to 60 min after contrast enhancement, the mean CT attenuation value of adenomas is substantially lower than that of nonadenomas. Therefore, adrenal lesions with an attenuation value of more than 10 HU in unenhanced CT or an enhancement washout of less than 50% and a delayed attenuation of more than 35 HU (on 10-15 min delayed enhanced CT) are suspicious for malignancy (Lee et al., 1991; Korobkin et al., 1998; Szolar & Kammerhuber, 1998; Caoili et al., 2000; Pena et al., 2000). Local invasion or tumour extension into inferior vena cava as well as lymph node or other metastasis (lung and liver) is often found in advanced ACC.

MRI is equally as effective as CT in distinguishing malignant from benign lesions (Outwater et al., 1996; NIH state of science statement 2002). With the advent of dynamic gadolinium enhanced- and chemical shift-technique in the last decade, MR characterisation of adrenal masses has improved significantly. ACCs are typically isointense to liver on T1 and show intermediate to increased intensity on T2 (Fig. 1). The enhancement after gadolinium is distinct and the washout is usually slow. However, most MRI studies for adrenal lesions focused on differentiating adenoma from metastases, rather than from ACC. In these studies, the sensitivity of MRI for differentiation of benign and malignant adrenal masses ranged between 81% and 89%, with specificity between 92% and 99% (Bilbey et al., 1995; Korobkin et al., 1995; Heinz-Peer et al., 1999; Honigschnabl et al., 2002). Whether chemical-shift MRI can reliably differentiate adenoma from carcinoma has not yet been established (Dunnick & Korobkin, 2002). MRI is superior to CT in detecting tumour extension into the inferior vena cava (Goldfarb et al., 1990).

(a) T2-weighted SE (spin-echo)-sequence of a 1-year-old boy with adrenal carcinoma: well defined inhomogenous tumor with cystic areas (arrows) representing necrosis. (b) Corresponding T1-weighted SE image after contrast administration: inhomogenous contrast enhancement with tumor necrosis (arrows).

Adrenal scintigraphy (NP-59) is not widely available, is time-consuming (3-5 days) and the diagnostic value beyond CT and MRI is controversial. Therefore, we do not recommend scintigraphy in patients with presumed ACC. In contrast, 18F-fluoride-oxyglucose positron emission tomography (18F-FDG-PET) has demonstrated good performance in differentiating malignant form benign adrenal lesion in retrospective studies (Boland et al., 1995; Maurea et al., 1999; Becherer et al., 2001; Yun et al., 2001). Moreover, FDG-PET can be used to detect metastatic disease. Prospective studies are needed to further validate the role of FDG-PET. 11C-metomidate PET has been successfully used for imaging of non-necrotic ACC (Khan et al., 2003). Major disadvantages are the limited availability and high costs of PET methods.

Fine-needle aspiration (FNA)/cut biopsy is not recommended to establish the diagnosis of ACC due to the risk of complications (up to 12%; Kloos et al., 1995), in particular needle tract metastases (Mody et al., 1995), and its controversial diagnostic value. However, in a recent prospective study adrenal cut biopsy was investigated in an 'ex vivo' approach in 220 consecutive adrenal lesions after surgical removal (Saeger et al., 2003). The overall sensitivity and specificity were 94·6 and 95·3%, respectively, suggesting significant diagnostic potential. However, despite ideal conditions for biopsy, in 10 cases the material was insufficient or not representative. Moreover, these data arose from an ex vivo approach with no risk of complications (e.g. tumour spillage). Thus an in vivo study would be needed to evaluate whether similar results can be obtained in a clinical setting.

For staging of established ACC, we recommend a high resolution CT of thorax and abdomen. FDG-PET may occasionally be helpful in differentiating metastasis from benign lesions. At the time of diagnosis and in case of bone pain, a bone scintigraphy with consecutive conventional X-ray studies of regions with an increased uptake is performed. Hormone measurements are also occasionally important: after presumed complete tumor removal in patients with ACC and CS, postoperative endogenous cortisol should be subnormal, otherwise stage IV should be assumed even if no metastases in imaging are detected in MRI and/or CT.

Even after surgical removal of the adrenal tumour, the diagnosis may remain difficult. As with tumour size in adrenal imaging, tumour weight is important, as most adenomas weigh between 20 and 50 g, while most malignant cortical tumours weigh more than 100 g (Saeger, 2000). For diagnosis of ACC, different diagnostic scores (Hough et al., 1979; Weiss, 1984; van Slooten et al., 1985; Weiss et al., 1989; Table 2 ) have been developed. Typical histopathological markers of malignancy are a high number of mitoses, atypical mitoses, vessel or capsule invasion and necroses. Molecular markers have been widely studied in recent years (Wachenfeld et al., 2001). However, no single marker is diagnostic of ACC. A Ki-67 staining index of more than 5% in adrenocortical tumours is suggestive of an ACC. To differentiate metastases from ACC or atypical pheochromocytomas, immunostaining and the comparison with the primary extraadrenal tumour is often necessary (Saeger, 2000). The marker D11 is useful, as it is positive in almost all cortical but negative in medullary adrenal tumors. To identify a pheochromocytoma or a neuroendocrine carcinoma, chromogranin A is the best marker. Keratin filaments are usually demonstrable in metastases from carcinomas.

For staging of ACC, the system of MacFarlane (1958) modified by Sullivan et al. (1978) is most frequently used and predicts the prognosis (Soreide et al., 1992; Barzon et al., 1997; Luton et al., 2000; Wajchenberg et al., 2000; Icard et al., 2001; Kendrick et al., 2001). However, modifications proposed by Lee et al. (1995) and Icard et al. (1992) are plausible, as they may better reflect the natural history of the disease and correlate more closely with other staging sytems used for solid tumours (Dackiw et al., 2001; see Table 3 ). In the majority of patients with stage I to III, complete tumour removal may be achievable, whereas this is highly unlikely in the presence of distant metastases (stage IV). In this revised staging system, stage IV is defined by the presence of distant metastasis.

While in older series (Wooten & King, 1993), most patients were diagnosed in advanced disease (stage IV), some more recent studies have reported the highest percentage of patients in stage II (Icard et al., 2001; Kendrick et al., 2001), probably reflecting improved and more widely available imaging technology.

Distant metastases affect most often liver and lung (see Table 4 ).