Management of Adrenocortical Carcinoma

Bruno Allolio; Stefanie Hahner; Dirk Weismann; Martin Fassnacht


Clin Endocrinol. 2004;60(3) 

In This Article

Therapy (Figure 2)

Proposed flow chart for adrenocortical carcinoma (ACC)*. (a) Follow-up intervals may increase with duration of remission, (b) drug monitoring is important. Aim at mitotane levels > 14 mg/l and < 20 mg/l, (c) see Table 6. *It is important to note that for all of the proposed therapeutic interventions, results from randomized phase III trials are lacking.

Complete surgical resection continues to be the treatment of choice for ACC (Dackiw et al., 2001). A margin-free resection (R0 resection) is a strong predictor of long-term survival (Khorram-Manesh et al., 1998; Icard et al., 2001; Kendrick et al., 2001). It is best performed by an experienced surgeon using a transabdominal or even a thoraco-abdominal approach (Dackiw et al., 2001; Icard et al., 2001). To avoid tumour spillage, the tumour capsule must remain intact. Invasion by or adherence of the carcinoma into adjacent organs often requires en bloc excision of the kidney, the spleen, partial hepatectomy or pancreatectomy (Icard et al., 2001). In addition, lymphadenomectomy has often to be included. The presence of a tumour thrombus in the renal vein or the inferior vena cava does not preclude a complete resection, although cardiac bypass technique may be necessary for successful removal of tumour tissue extending into the inferior vena cava or even the right atrium (Cheung & Thompson, 1989; Moul et al., 1991; Hedican & Marshall, 1997). The role of tumour debulking in the presence of metastatic disease is a matter of debate. Incomplete resection of the primary tumour or metastatic disease not amenable to surgery is associated with a particular poor prognosis. In most studies, the median survival is below 12 months (Crucitti et al., 1996; Icard et al., 1992; Zografos et al., 1994; Lee et al., 1995). However, tumour debulking may help to control hormone excess and may in individual cases facilitate other therapeutic options. Even if complete resection has been achieved, local recurrence and metastatic disease during follow-up is common. Risk factors include stage III tumours, a tumour diameter above 12 cm, a high mitotic index and intratumoural haemorrhage (Harrison et al., 1999; Stojadinovic et al., 2002).

Laparoscopic resection of benign adrenal tumours has been a major improvement in adrenal surgery. However, in our view, a laparascopic approach should not be used for a presumable ACC, because of the risk of tumour capsule violation, tumour fragmentation (Iino et al., 2000; Dackiw et al., 2001) and the potential difficulty to perform a definite margin-free R0 resection.

Surgical resection of recurrent disease is an important therapeutic option associated with prolonged survival (Bellantone et al., 1997; Jensen et al., 1991; Schulick & Brennan, 1999b), although cure is seldom achieved. Surgery for recurrent disease includes locoregional recurrence as well as isolated hepatic and pulmonary metastases. The most frequent indication for re-operation is locoregional disease (> 65%; Jensen et al., 1991; Favia et al., 2001). Recently, successful thermoablation for recurrent or metastatic ACC has also been reported (Wood et al., 2003).

Surgery-related mortality has improved but remains substantial (5%; Icard et al., 2001). It is particularly high for stage III disease with invasion of adjacent organs.

The role of radiotherapy in ACC has not been well defined and is usually regarded as of limited benefit (Schulick & Brennan, 1999a). However, palliative radiotherapy for metastatic disease was effective in a significant percentage of patients (Percarpio & Knowlton, 1976; Didolkar et al., 1981) and is the treatment choice for bone metastases (30-40 Gy). More importantly, radiotherapy may have a role as adjuvant postoperative radiation therapy in patients at high risk for local recurrence. Based on a small series of patients with stage III the survival was higher than expected form historic series (Markoe et al., 1991). Local tumour recurrence is common in ACC and the most frequent cause for re-operation (Jensen et al., 1991; Favia et al., 2001). Of note, in a series of children with ACC metastatic disease was invariably preceded by local recurrence of the disease (Ribeiro et al., 2000). Postoperative radiation of the tumour bed (50-60 Gy) may therefore improve the long-term outcome in stage III ACC or high-risk stage II patients (tumour diameter > 12 cm, high mitotic index, violation of the tumour capsule or frank evidence of tumour spillage during surgery). Modern treatment concepts with CT-planning, high-voltage radiation (4-6 MeV) and multiple fields are required for optimum results.

Medical therapy aims at the control of hormone hypersecretion and - more importantly - partial or complete remission of tumour spread.

More than 40 years ago, Bergenstal et al. (1959) reported the first successful use of o,p'-DDD (mitotane) in patients with metastatic ACC. Mitotane (1,1 dichloro-2(o-chlorophenyl)-2-(p-chloro-phenyl) ethane) is an isomer of the insecticide p,p'-DDD and a chemical congener of the insecticide DDT. It is an adrenolytic compound with specific activity on the adrenal cortex (Schteingart, 2000). Its therapeutic effects depend on intradrenal metabolic transformation. Mitotane is hydroxylated in the mitochondria at the β-carbon and further transformed into an acyl-chloride. It has been reported that the active metabolites cause toxicity by oxygen activation with superoxide formation or by covalent binding to specific proteins (Schteingart, 2000).

The clinical efficacy of mitotane remains disputed. Hutter & Kayhoe (1966) collected a series of 138 patients and reported that 34% of 59 patients with evaluable measurable disease had objective tumour regression. Wooten & King (1993) have reviewed 551 cases in the English literature and reported a response rate of 35% with mostly partial and transient responses and only an occasional complete remission. More recent series have reported lower response rates (Khorram-Manesh et al., 1998; Baudin et al., 2001). In our own experience, only a minority of patients also exhibit objective tumour regression. However, in two patients with documented metastatic disease, we observed a lasting complete remission, which in both cases persists for now more than 7 years after mitotane withdrawal. Similar cases have been described in the literature (Remond et al., 1992).

The role of mitotane as adjuvant therapy after complete surgical removal of ACC remains a matter of debate (Venkatesh et al., 1989; Icard et al., 1992; Khorram-Manesh et al., 1998; Barzon et al., 1999; Kasperlik-Zaluska, 2000). Due to the high rate of locoregional or metastatic recurrence after seemingly curative resection adjuvant treatment options are clearly needed. However, in some series, adjuvant mitotane was associated with a poorer outcome (Vassilopoulou-Sellin et al., 1993). It is important to note that differences or similarities in survival of patients with ACC cannot be assessed without prospective randomized trials with a sufficient number of patients and therefore the value of adjuvant mitotane remains uncertain (Ahlman et al., 2001).

Mitotane is either given as tablets (Lysodren®, Bristol Myers Squibb, Princeton, USA) in doses > 3 g/day or as capsules of micronized mitotane mixed with cellulose acetylphthalate, with a lower absorption rate, but, possibly, a better gastrointestinal tolerance (usually higher doses up to 12 g/day; Luton et al., 1990; Baudin et al., 2001). Drug monitoring is important. It has been found that drug levels > 14 mg/l are required to induce tumour regression. An objective response in metastatic disease was found in 31% (Baudin et al., 2001), 55% (Haak et al., 1994) or > 80% (van Slooten et al., 1984) of patients achieving this level, whereas no response was seen in patients with a lower serum concentration. As side-effects are more frequent with drug levels > 20 mg/l (van Slooten et al., 1984) and drug concentrations in serum are not closely related to drug dose (Terzolo et al., 2000b), drug monitoring may also improve quality of life during mitotane treatment by avoiding over-treatment. Due to the long half-life of o,p'DDD, the highest trough levels are achieved only after several months of therapy (Baudin et al., 2001). Accordingly, in our experience, mitotane side-effects may become more pronounced with ongoing treatment despite a constant mitotane dose as drug levels gradually increase. Side-effects of mitotane occur frequently and are often dose-limiting ( Table 5 ). These effects are mainly gastrointestinal (diarrhoea, nausea, anorexia) or concern the CNS (lethargy, somnolence, ataxia, dizziness, confusion; Hutter & Kayhoe, 1966; Schteingart et al., 1982). Patients rarely tolerate doses > 6 g/day for long-term therapy.

Due to its adrenolytic activity long-term mitotane treatment induces adrenal insufficiency. As its action is more pronounced in fasciculata cells, glucocorticoid deficiency precedes mineralocortioid deficiency. Inadequately treated adrenal insufficiency enhances mitotante-induced side-effects and reduces mitotane tolerance (Kasperlik-Zaluska, 2000). Because an increased metabolic clearance of glucocorticoids (e.g. dexamethasone) has been reported (Robinson et al., 1987) high-dose glucocorticoid replacement is needed. Hydrocortisone is the treatment of choice (Robinson et al., 1987; Kasperlik-Zaluska, 2000) and the glucocorticoid replacement is monitored best with careful clinical assessment and measurements of plasma ACTH levels, which should not be elevated. A daily dose of 50 mg hydrocortisone (20 mg–20 mg–10 mg) and more may be needed. Fludrocortisone may be added depending on blood pressure, serum potassium levels and plasma renin activity.

Increases in hepatic gamma glutamyl transaminase levels are frequent (Luton et al., 1990; Neuman et al., 2001) and in most cases do not require withdrawal of the drug. However, serious hepatotoxicity has also been described. Mitotane increases serum cholesterol mainly by increasing low-density lipoprotein (LDL) cholesterol. This increase may be amenable to statin therapy (Maher et al., 1992). In addition, mitotane can prolong the bleeding time by changing platelet aggregation response (Haak et al., 1991).

Of particular importance are mitotane-induced endocrine abnormalities. Mitotane strongly increases hormone-binding globulins (e.g. cortisol-binding globulin, sex hormone-binding globulin). Thus, measurement of total hormone concentrations may give normal results in the presence of clearly impaired bioavailability of free hormones (van Seters & Moolenaar, 1991). Additionally, total thyroxine levels may be reduced as mitotane competes with endogenous thyroxin for thyroxine-binding globulin-binding sites (Marshall & Tompkins, 1968). In some patients, free thyroid hormone concentrations decrease and thyroxin replacement may also become necessary.

For management of nausea, 5-hydroxytryptamine (5-HT) blockers may be useful. In case of significant neuropsychiatric side-effects, drug treatment is interrupted for a minimum of 1 week and restarted with a lower dose. Due to the long half-life, significant serum concentrations may persist for weeks to months after cessation of therapy.

In patients with advanced local or metastatic disease, not amenable to surgical resection cytotoxic chemotherapy has been investigated (see Table 6 ).

In ACC, strong expression of the multidrug-resistance gene mdr-1 has been observed (Abraham et al., 2002) resulting in high levels of p-glycoprotein, which acts as a drug efflux pump and may cause chemotherapy failure (Ahlman et al., 2001). In vitro studies have shown that mitotane may partially reverse multidrug resistance by inhibiting drug efflux (Bates et al., 1991). This observation has led to protocols combining mitotane with cytotoxic chemotherapy, although a recent study casts doubts on the efficacy of mitotane to act as an effective p-glycoprotein antagonist in vivo (Abraham et al., 2002). Several cytotoxic agents have been used as single drugs or in combination to treat patients with advanced ACC (see Table 6 ) including cisplatin, doxorubicin, etoposide, vincristine, 5-fluorouracil and streptozocin (Ahlman et al., 2001). Although the results are variable, there is evidence that cisplatin alone or in combination with etoposide has some activity in advanced ACC (Bukowski et al., 1993; Burgess et al., 1993; Berruti et al., 1998; Bonacci et al., 1998; Williamson et al., 2000). Bonacci et al. (1998) treated 18 patients with etoposide (100 mg/m2/day on days 1-3) and cisplatin (100 mg/m2/day on day 1) every 4 weeks maintaining mitotane therapy with an overall response of 33%. Similarly, Burgess et al. (1993), using the same drugs without mitotane, reported a response rate of 46%. However, results of a more recent study in 45 patients with nonresectable or metastatic carcinoma using etoposide (100 mg/m2/day on days 1-3) and cisplatin (50 mg/m2/day on days 1 and 2) were clearly inferior with objective response in only 11% of patients (Williamson et al., 2000). In this study, mitotane was withheld or interrupted during cytotoxic chemotherapy. The highest response rate so far has been observed in a phase II multicentre trial from Italy using the combination of etoposide (100 mg/m2/day on days 5-7), doxorubincin (20 mg/m2/day on days 1 and 8) and cisplatin (40 mg/m2/day on days 1 and 9) every 4 weeks (3-8 cycles) given together with continuous mitotane (planned dose 4 g/day). According to WHO criteria, an overall response rate of 53·5% was achieved (two complete and 13 partial responses in 28 patients). Due to mitotane side-effects, a reduced mitotane dose (2-3 g/day) was given in the majority of these patients. Recently, Khan et al. (2000) have evaluated the efficacy of streptozocin plus mitotane in ACC. Oral mitotane (1-4 g/day) was given together with intravenous streptozocin (1 g/day for 5 days, thereafter 2 g once every 3 weeks). Complete or partial responses were obtained in 36·4% (eight out of 22) of patients with measurable disease. Importantly, in this paper Khan et al. (2000) provide evidence of a possible efficacy of this regime in an adjuvant setting after surgery. In a nonrandomised study, streptozocin plus mitotane significantly increased survival compared to patients who did not receive treatment after complete tumour resection. Again, such finding needs confirmation in a prospective randomized trial, as selection bias is likely.

Several other agents have been used in the treatment of advanced ACC. Suramin, an antitrypanosomal agent, may induce transient remission in occasional patients (Allolio et al., 1989) but its use is limited by significant toxicity (Arlt et al., 1994). Gossypol, a plant toxin from cotton seed oil, induced a partial remission in three out of 18 patients (17%) with metastatic ACC. However, three patients died of their disease without achieving the intended drug levels and had been eliminated from the analysis (Flack et al., 1993).

Because hormone excess (in particular hypercortisolism) is associated with a decreased of quality of life and an increased risk of complications, it is essential that patients do not suffer from CS. Adrenostatic drugs other than mitotane may be needed to control endocrine activity (Luton et al., 1990). Metyrapone, ketoconazole, etomidate and aminoglutethimide inhibit P450 steroidogenic enzymes like 11β-hydroxylase and side-chain cleavage enzyme (Feldman, 1986). Aminoglutethimide is also an inhibitor of aromatase activity. Ketoconazole (400-1200 mg/day) is most frequently used and may even possess antiproliferative activity in some patients with ACC (Contreras et al., 1985). Adrenal insufficiency requiring hormone replacement and hepatotoxicity are the most frequent side-effects. It is important to note that ketoconazole may impair the adrenolytic action of mitotane (Schteingart, 2000).

Intravenous etomidate is the most potent adrenostatic drug available (Allolio et al., 1988) and is probably the treatment of choice to rapidly control severe life-threatening hypercortisolism also in ACC. Again, there is evidence that etomidate and also aminoglutethemide possess some antiproliferative activity in adrenocortical tumour cells (Fassnacht et al., 2000).

The use of adrenostatic drugs - including mitotane - always requires supervision by an experienced endocrinologist.

Close follow-up is of vital importance in ACC to detect recurrence at a time when surgical intervention is still possible. In our experience, this aspect is often neglected after complete tumour removal in stage I and II patients as cure is prematurely assumed. Unfortunately, recurrence is common and most patients eventually succumb to their disease (Wajchenberg et al., 2000; Icard et al., 2001; Vassilopoulou-Sellin & Schultz, 2001). Staging using CT should be performed every 3-4 months during the first 2 years after complete tumour removal. Intervals may then increase with disease-free time from surgery. In functioning tumours, hormonal marker (e.g. DHEAS) may rise again after surgery long before tumour tissue becomes detectable by imaging techniques. The duration of follow-up has not been standardised but should probably be indefinite.

There is some evidence that in the last two decades earlier diagnosis and improved surgical management have both led to a significantly better outcome (Icard et al., 2001; Vassilopoulou-Sellin & Schultz, 2001). It is also important to bear in mind that ACC is a heterogenous disease with some patients surviving for more than 10 years despite metastatic disease, whereas others die within a few months from a rapidly progressive disease not responding to any available therapy.

In general, the prognosis for ACC is still grim. MacFarlane (1958) has reported that patients with untreated ACC have a median survival of 3 months only. In treated ACC, overall 5-year survival ranged between 23% and 60% in different series (Nader et al., 1983; Venkatesh et al., 1989; Haak et al., 1990, 1993; Luton et al., 1990; Icard et al., 1992, 2001; Vassilopoulou-Sellin & Schultz, 2001). Patients with stage I and II have a similar prognosis which is significantly better than that for stage III and IV patients (Wajchenberg et al., 2000). In a recent series including 253 patients from France, the 5-year actuarial survival rates were 60% for stage I, 58% for stage II, 24% for stage III and 0% for stage IV (Icard et al., 1992). The overall rate was 38% with a rate of 50% in patients who underwent resection for cure. At present, the most important prognostic factors remain, therefore early stage and complete tumour removal aiming at cure (Pommier & Brennan, 1992; Soreide et al., 1992; Zografos et al., 1994; Schulick & Brennan, 1999a, 1999b). Accordingly, in a recent series reported by Vassilopoulou-Sellin & Schultz (2001), long-term survivors (> 5 years) had significantly less extensive disease at diagnosis (P < 0·001) in comparison to patients with the shortest survival (< 11 months). In contrast, they found no differences in age, gender and functionality. Tumour size may be important, as patients with completely resected large (> 12 cm) tumours had significantly reduced survival (Harrison et al., 1999). Recently, Stojadinovic et al. (2002) have analysed additional parameters in a series of 124 patients using molecular expression profiles and morphologic patterns in tissue specimens. In their analysis tumour necrosis (P < 0·01), a mitotic rate of more than five of 50 high-power fields (P = 0·004) and atypic mitotic figures (P = 0·008) were associated with reduced disease-free survival. In addition, high proliferative activity as assessed by Ki-67 staining and evidence for mutated p53 are associated with advanced stage ACC and poor prognosis (Stojadinovic et al., 2002). Endocrine activity of ACC has no general influence on prognosis as compared to nonfunctioning ACC (Favia et al., 2001). However, there is some evidence that tumours secreting androgens or steroid precursors only have a better prognosis than cortisol-secreting ACC or tumours secreting both cortisol and androgens (Icard et al., 1992; Ribeiro et al., 2000).

Progress in the management of ACC is hampered by its low incidence and heterogeneity. Not a single prospective randomized study (phase III trial) has been performed to directly compare different treatment modalities. Accordingly, claims of improved survival for certain treatment options (e.g. mitotane for stage IV disease; Icard et al., 1992) are not well founded, as selection bias is likely. Only multicentre and probably multinational (e.g. European) efforts will change this picture. To this end, a number of consecutive steps need to be taken: patients with ACC should be treated in a few specialized national centres to provide the necessary optimum interdisciplinary care. These centres should then build a national ACC registry to further enhance patient recruitment and to standardize patient care. Some registries have already been established in France and Italy, and a German registry is underway. These national registries should be harmonized and could serve as a multinational nucleus for prospective trials of sufficient size. In our view, two important areas should be the focus of such trials. The first concerns adjuvant therapy after surgical resection for cure, as the majority of these patients will develop local recurrence or metastatic disease. Possible options are mitotane with or without streptozocin and/or adjuvant radiotherapy of the tumour bed. An untreated control group is needed, because for none of these treatment options has a beneficial effect been established. The other area is stage IV ACC. Here it is important to compare mitotane alone with a combination of mitotane plus cytotoxic drugs.

Undoubtedly, innovative treatment options need to be developed based on a better understanding of the molecular pathogenesis of ACC. At present, inhibition of IGF-II signalling (e.g. by small molecule IGF-I receptor antagonists) seems to be a promising approach. In addition, antiangiogenic drugs and immunotherapy should be investigated in patients with progressive disease.

In conclusion, at this time it is our responsibility not only to provide the best possible care for individual patients with ACC but also to set up structures that will allow us to make systematic progress in the management of this dreadful disease.

CLICK HERE for subscription information about this journal.