5.4 Adrenocortical cancer

Adrenocortical cancer (ACC) is among the most aggressive endocrine tumours with an overall very poor prognosis. Morbidity and mortality can be secondary to steroid hormone excess and/or tumour growth and metastases. This potentially poor outcome justifies the importance of considering malignancy in the man-agement of an adrenal mass. The diagnosis of malignancy in a patient with an adrenal tumour relies on careful investigations of clinical, biological, and imaging features before surgery and path-ological examination after tumour removal. Appropriate man-agement and follow-up by an expert multidisciplinary team is important to improve prognosis and to progress in these rare neoplasms.

ACC consists of monoclonal populations of cells, suggesting that tumour progression is the end result of an intrinsic genetic or epigenetic alteration. Monoclonal tumours result from alterations conferring a growth advantage to the cell initially affected.

These genetic events can be studied at the scale of the whole genome, as losses or gains of part or all of a chromosome. A large number of molecular techniques, such as comparative genomic hybridization and microsatellite analysis, have been used in genome-wide screens for such chromosomal alterations. It has been demonstrated by comparative genomic hybridization that chromosomal alterations are very frequent in ACC. Chromosomal losses were observed at 1p, 17p, 22p, 22q, 2q, and 11q in up to 62% of cases of ACC. Studies using microsatellite markers have demonstrated a high percentage of loss of heterozygosity or allelic imbalance at 11q13 (≥90%), 17p13 (≥85%), and 2p16 (92%) in ACC (1).

The genes involved in these molecular alterations could be classified as tumour suppressor genes on the one hand, and oncogenes on the other hand. Molecular alterations would lead to inactivation of the tumour suppressor genes and activation of the oncogenes. In various cancers the study of chromosomal rearrangement led to the identification of the oncogenes or tumour suppressor genes involved in their development. However, currently in ACC, such genes have been mostly identified by the study of familial diseases associated with adrenocortical tumours. Nevertheless, the loci of these genes are frequently altered in sporadic ACC, suggesting the importance of these loci and genes in the development of these tumours (Table 5.4.1; Fig. 5.4.1).

The tumour suppressor gene TP53 is located at 17p13 and its product is involved in the control of cell proliferation. Germline mutations in TP53 are responsible for the Li–Fraumeni syndrome. This syndrome displays dominant inheritance and confers susceptibility to breast cancer, soft tissue sarcoma, brain tumours, osteosarcoma, leukaemia, and ACC. Germline mutations in TP53 have been observed in 50–80% of children with apparently sporadic ACC in North America and Europe. In Southern Brazil, a specific germline mutation has been identified in exon 10 of the TP53 gene, R337H, which is observed in almost all paediatric cases (2). In sporadic ACC in adults, somatic mutations of TP53 are found in only 25–35% of the cases. Interestingly, loss of heterozygosity at 17p13 occurs in 85% of sporadic ACC (1).

The IGF2 gene, located at 11p15, encodes an important fetal growth factor which is maternally imprinted and expressed only from the paternal allele. Genetic or epigenetic changes in the imprinted 11p15 region, resulting in increased IGF2 expression, and mutations of the p57kip2 gene have been implicated in the Beckwith–Wiedemann syndrome. This overgrowth disorder is characterized by macrosomia, macroglossia, organomegaly, and developmental abnormalities (in particular abdominal wall defects with exomphalos), embryonal tumours (such as Wilms’ tumour), and ACC, neuroblastoma, and hepatoblastoma. Several studies have demonstrated that IGF2 is strongly overexpressed in approximately 90% of ACC (3).

Genetic alterations of the Wnt signalling pathway were initially identified in familial adenomatous polyposis coli and have been extended to a variety of cancers. Furthermore, familial adenomatous polyposis coli patients with germline mutations of the APC (adenomatous polyposis coli) gene, which lead to an activation of the Wnt signalling pathway, may develop adrenocortical tumour. The Wnt signalling pathway is normally activated during embryonic development. β-catenin is a key component of this signalling pathway. In ACC, β-catenin delocalization can be observed, consistent with an abnormal activation of the Wnt-signalling pathway. In a subset of adrenocortical tumours, this is explained by somatic activating mutations in the b-catenin gene (4).

The MEN1 gene, located at the 11q13 locus, is a tumour suppressor gene. A heterozygous inactivating germline mutation of MEN1 is found in about 90% of families affected by multiple endocrine neoplasia type 1 (MEN 1). The principal clinical features of this autosomal dominant syndrome include parathyroid (95%), endocrine pancreatic (45%), and pituitary (45%) tumours and thymic carcinoids. Adrenocortical tumours and/or hyperplasia are observed in 25–40% of MEN 1 patients. ACC has rarely been observed in MEN 1 patients. Somatic mutations in the MEN 1 gene are very rare in sporadic adrenocortical tumours. By contrast, loss of heterozygosity at 11q13 is observed in more than 90% of informative ACC and only 20% of adrenocortical adenomas. However, loss of heterozygosity in ACC involves almost all the 11q domain, suggesting that an, as yet unidentified, tumour suppressor gene located on the long arm of the chromosome is involved in ACC formation (5).

The use of large-scale analysis of gene expression, or transcriptome analysis, to study various cancers has been a source of important advances both in tumour classification and understanding of pathogenesis. This method has been applied recently to adrenocortical tumours and it appears that gene expression profiles of benign tumours differ markedly from that of ACC (6). A cluster of genes overexpressed in ACC are related to IGF-2 and other growth factors, and this has been termed the IGF-2 cluster. This cluster contains mainly growth factors and growth factor receptor genes. By contrast, a steroidogenic cluster of genes (such as CYP11A, CYP11B1, HSD3B, encoding steroidogenic enzymes) is expressed only at low level in ACC as compared to adrenocortical adenomas. As most of these genes are related to steroidogenesis, a dedifferentiation process might occur during malignant transformation (1). The observed differences in gene expression profile between benign and malignant ACC suggest that transcriptome analysis could potentially offer new diagnostic tools for the discrimination of benign from malignant tumours (7).

ACC is a rare tumour with an estimated incidence between 1 and 2 per million per year in adults, in North America and Europe. The prevalence has been estimated to be between 4 and 12 per million population (8). As observed in many rare tumours, the incidence is difficult to determine and the true numbers might be higher than the current estimations. For instance, the prevalence of adrenal incidentaloma range in the general population from 1% in subjects younger than 30 years to 7% in subjects older than 70 years. Among the group of adrenal incidentalomas selected for surgery, the frequency of ACC ranges between 3 and 10%.

In children, the incidence of ACC is considered as 10 times lower than in adults, except in South Brazil where there is a higher incidence of paediatric ACC due to the high prevalence of a specific germline TP53 mutation, as discussed above.

In some series there is a slightly increased female to male ratio (9), although not always reported. Among female patients with Cushing’s syndrome diagnosed during pregnancy, the frequency of ACC is higher than in nonpregnant female patients with Cushing’s syndrome (10).

Signs and symptoms leading to the diagnosis of ACC can be due to steroid excess, tumour mass, and effects of metastases (11). Although ACC is not the most frequent diagnosis in adrenal incidentalomas, nowadays the diagnosis of ACC is made with an increasing frequency during the diagnostic work-up of an incidentally discovered adrenal mass. This is important since it might be a way to diagnose an ACC at an earlier stage and to improve the prognosis by an early, complete surgical removal. This underlines the need for careful investigations of adrenal incidentalomas in order to decide whether to go for surgery if malignancy is suspected. Other specific features may be associated with rare genetic diseases such as the Li–Fraumeni and Beckwith–Wiedemann syndromes where ACC is part of a more complex syndrome, as discussed above.

Less than a third of ACCs are really ‘nonhypersecretory’ after careful hormonal investigations (11, 12). In these cases, one should be cautious not to overdiagnose a tumour of the adrenal area as an ACC. These nonhypersecretory ACCs can be diagnosed after investigation of adrenal incidentalomas or due to the consequences of local expansion of the tumour mass, e.g. local symptoms (pain, palpation of a tumour, venous thrombosis), or distant metastases (liver, lung, bones). Fever may occur, in some cases after tumour necrosis, but is a rather rare sign. Similarly, weight loss is rarely observed in ACC. Characteristically, the general health condition of patients affected by apparently endocrine inactive ACC is remarkably good, even during the early stages of metastasis. However, the general condition of the patient is most often preserved except at a very late stage when the tumour is nonsecreting. This explains why nonhypersecretory ACCs may be diagnosed only at a relatively late stage of the disease.

Most patients will present with signs of steroid excess. Cushing’s syndrome associated with signs of androgens excess progressing for a few months is the most characteristic presentation (Fig. 5.4.2). Signs of mineralocorticoid or oestrogen excess are less frequent but highly suggestive of the diagnosis of ACC in a patient with an adrenal mass above 3 cm.

More than three-quarters of ACC patients suffer from steroid-secreting tumours when based on biochemical diagnosis following careful hormonal investigations of plasma and urine. In the near future, it is expected that mass spectrometry urine analysis will allow detection of alterations of adrenal steroid profile in all patient with ACC (12). In the absence of steroid excess, one should be cautious before diagnosing a mass of the adrenal area, which does not appear as a benign adrenal adenoma on imaging, as an adrenocortical tumour. Hormonal investigations are important for the diagnosis of the nature of an adrenal mass. Evidence of steroid excess can link an adrenal mass to its cortical origin. Some patterns of steroid oversecretion are very suggestive of malignancy. Hormonal investigations also give important information for patient management and may serve as a tumour marker during postoperative follow-up and treatment of the metastatic stage of the disease. In 2006, the ACC working group of the European Network for the Study of Adrenal Tumours (ENSAT) recommended a minimal hormonal work-up in patient with ACC (12, 13) (Box 5.4.1).

In contrast to benign adrenocortical tumours, which characteristically secrete a single class of steroid, usually either cortisol or aldosterone, ACC often secrete several types of steroids. Cosecretion of androgens and cortisol is the most frequent and is highly suggestive of a malignant adrenocortical tumour. Cortisol oversecretion will induce centripetal obesity, protein wasting with skin thinning, and striae, muscle atrophy (myopathy), and osteoporosis. Cortisol excess can also cause impaired defence against infection, diabetes, hypertension, psychiatric disturbances, and gonadal dysfunction in men and women. Androgen oversecretion may induce various manifestations in women: hirsutism, menstrual abnormalities, infertility, and eventually frank virilization (alopecia, deepening of the voice, clitoris hypertrophy). ACC can also secrete mineralocorticoids and steroids precursors. Oversecretion of oestrogens can be observed in rare cases and is very suggestive of ACC in a male patient with an adrenal tumour, where it often results in gynaecomastia.

ACTH-independent cortisol oversecretion is easily demonstrable by increased urinary cortisol excretion, cortisol secretion that is not suppressible with high doses of dexamethasone, and associated undetectable plasma ACTH levels. Plasma 17-hydroxyprogesterone is often elevated as well as the specific adrenal androgen DHEA-S, which leads to increased plasma testosterone in females. Other steroids, such as compound S, 11-deoxycorticosterone (DOC), δ4-androstenedione, and oestradiol, can be overproduced by the tumour. Secretion of aldosterone by ACC is not frequent and can be detected by plasma aldosterone and renin assays.

Imaging is an essential diagnostic step for ACC, especially in cases of adrenal incidentaloma. It is important both for the diagnosis of malignancy of an adrenal mass but also for the extension work-up. Adrenal CT scan is a very informative imaging procedure for adrenocortical tumours (14) (see also Chapter 5.1). In ACC, it shows a unilateral mass, which is most often large (above 5–6 cm, and typically 10 cm and above), lowering the kidney. Apart from the size of the tumour, the features suggestive of malignancy are: the lack of homogeneity with foci of necrosis and irregular margins; and a high spontaneous density observed before contrast media injection during CT scan (above 10 HU), indicating a low fat content in contrast to a usually characteristic high fat content observed in adrenocortical adenomas (Fig. 5.4.3). Dynamic measurement of contrast-enhanced densities may provide a more sensitive way to distinguish between benign and malignant lesions. A CT scan also contribute to the detection of local invasion, and distant metastases (liver, lung). This emphasizes the need to perform a CT scan of the abdomen and the chest prior to any surgery of a suspected ACC. Locoregional vessel invasion through the renal veins and the inferior vena cava can extend up to the right atrium and may result in metastatic lung embolism (15). MRI can be used, and might be as effective as CT scan when dynamic-gadolinium enhanced and chemical shift are used to characterize an adrenal mass. MRI can also participate to the detection of liver metastasis and venous invasions.

More recently studies have demonstrated that ACCs almost always have a high uptake of [¹⁸F]2-fluoro-2-deoxy-d-glucose (FDG). Thus FDG positron emission tomography (FDG-PET) appears to distinguish between benign and malignant adrenal tumours (16). This simple, noninvasive imaging procedure is part of the extended work-up in ACC (17) (Fig. 5.4.4). It is especially informative when it is combined with a CT scan (PET/CT). Currently FDG-PET is used as a very sensitive method prior to surgery of an adrenal tumour considered as an ACC; it is also used to exclude metastasis and during follow-up of ACC, in particular if recurrence is biochemically suspected but not visualized by CT.

Adrenal scintigraphy with iodocholesterol is not routinely needed but can help in some situations. Bone scintigraphy may help evaluate bone metastases. However, in patients with Cushing’s syndrome bone remodelling and fractures can lead to false-positive results of bone scintigraphy and the wider use of FDG-PET might in the future replace it. New adrenal cortex specific scintigraphy imaging, using radiolabelled tracers such as metomidate, are under investigation and might be a promising tool in the not too distant future (18).

As discussed above, clinical, hormonal, and imaging investigations can be very suggestive of an ACC. Large adrenocortical tumours (>6 cm) are more likely to be malignant (Fig. 5.4.5), but tumour size is clearly not a valid criterion to diagnose or exclude malignancy. On the other hand, evidence of a metastatic adrenal mass with ACTH-independent steroid excess is almost diagnostic for ACC, with a few exception. However, histopathological diagnosis is always a very important step. In the case of nonhypersecreting and/or localized tumours, pathology is key to diagnose both the adrenocortical origin and the malignant nature of the mass. The adrenocortical origin of the tumour is based on the histological analysis, but also immunohistochemistry. Immunohistochemical markers are especially used to exclude other types of tumour, for instance a phaeochromocytoma will stained with a chromogranin A antibody but an ACC will not. The immunostains that can be positive in an adrenocortical tumour are either not specific (such as Melan A) or not used on a routine basis (such as SF-1 or steroidogenic enzyme).

As often is the case with endocrine tumours, the diagnosis of malignancy in adrenocortical lesions can be difficult for the pathologist. There is not a single pathological feature that allows the conclusive diagnosis of a malignant adrenal cortical tumour based on the adrenal mass histology alone. Combinations of various histological parameters that allow establishment of a ‘score’ for a given tumour have been developed. The most widely used is the Weiss score, featuring nine different items (19) (Box 5.4.2). Each item is given a value of one when it is present and zero when it is absent. The total score is obtained by adding up the values of each individual item. It is assumed that a score of 3 or above is most probably associated with a malignant tumour. Other approaches based on microscopic feature analysis have been developed but have been less widely used and are therefore less validated than the Weiss score. However, all these approaches suffer limitations and are dependent on the experience of the pathologist. Therefore, efforts towards developing informative molecular markers of malignancy in ACC are under way. As described previously, IGF-2 overexpression and allelic losses at 17p13 have been suggested as potential molecular markers (3). Immunohistochemistry of cyclin E or Ki-67, which are higher in malignant adrenocortical tumours, has also been suggested as potentially useful diagnostic tools (12, 13). More recently, large-scale transcriptome analysis using DNA chips has been used to develop molecular markers based on the expression level of two genes for the diagnosis of malignancy (7). In the near future, such research efforts into translation are likely to have an important impact on our ability to accurately diagnose and classify adrenocortical tumours.

Tumour staging is the most important prognostic factor in the diagnosis of ACC. The McFarlane staging, as modified by Sullivan, is the most commonly used and relies on surgical finding and extension work-up (20). It has been followed by the UICC/WHO TNM classification of ACC in 2004 (12). Four stages are differentiated with this score. Stage 1 and Stage 2 tumours are localized to the adrenal cortex and present a maximum diameter below or above 5 cm, respectively. Stage 3 tumours present with local infiltration reaching the surrounding adipose tissue or lymph node. Stage 4 tumours are associated with infiltration of the surrounding tissue and lymph nodes, invasion into adjacent organs, or distant metastases. The prognosis of Stage 1 and 2 tumours is better than that of Stage 3 or 4 tumours (11, 15) (Fig. 5.4.6). A score with slight modification offering a better discrimination of the survival between the Stage 2 and 3 tumours has recently been defined by ENSAT (21).

The metastatic spread of ACC involves mostly liver and lung, observed in about 35 to 50% of patients for each organ. Bone metastases are only diagnosed in 10 to 15% of cases. Metastatic spread to other organs is rare (11, 12).

The overall survival of patients with ACC is poor, with a 5-year survival rate below 35% in most series. However, this depends on tumour stage. It is likely that progress in medical management of adrenal incidentalomas and more sensitive investigations of modest signs of steroid excess will increase the detection of localized ACC (stages 1 or 2). This should improve the overall survival rate. A better survival has been reported in younger patients, but this is not a constant finding (11). Cortisol-secreting tumours might be associated with a worse prognosis (11, 22). This could be due to the morbidity associated with Cushing’s syndrome or to differences in tumour progression. Some pathological features, such as a high mitotic rate or atypical mitotic figures as well as a high Ki-67 labelling, have been shown to be associated with a poor prognosis (12, 23, 24). This suggests that tumour biology plays a role in the prognosis. Here again, gene profiling has been used for tumour classification in research programmes to define molecular markers that might be useful in the near future for prognostication and therefore patient management (7).

Treatment aims at correcting both steroid oversecretion and its clinical consequences in cases of secreting tumours and to eradicate the tumour in all cases. The best way to achieve both goals is the complete removal of the tumour whenever it is possible, depending on tumour stage and the patient’s condition (Fig. 5.4.7).

Steroid oversecretion when clinically significant and not curable by tumour removal requires anticortisolic and/or symptomatic treatment. In this indication, mitotane is the drug most often used because it also has a cytotoxic effect on the adrenocortical cells, as discussed below. In some situation of severe Cushing’s syndrome requiring rapid control of steroid excess, other drugs can also be used, eventually as combined therapy (e.g. ketoconazol, metyrapone, etomidate).

Surgery of the adrenal tumour is the major treatment of stage 1–3 ACC. It can also be discussed in Stage 4 patients. The initial surgery is a crucial therapeutic step in the management of ACC. It should therefore be performed by trained surgeons, with experience of the management of adrenal tumours, to achieve complete tumour removal and avoid tumour spillage. Complete tumour removal and avoidance of violation of the tumour capsule is very important to increase the probability of long-term remission (15, 25). Open adrenalectomy is currently recommended as laparoscopic removal of malignant adrenocortical tumours could be associated with a high risk of peritoneal dissemination (26). Glucocorticoid replacement therapy should be started at the time of surgery of cortisol-secreting tumours to avoid adrenal deficiency resulting from long-term ACTH suppression by the ACC-associated cortisol oversecretion and thus functional suppression of the contralateral adrenal gland.

In Stage 4 patients with distant metastasis, tumour debulking with removal of the adrenal tumour can be discussed in order to reduce steroid excess and sometimes also tumour bulk, and requires multidisciplinary consideration. Surgery is discussed depending of the tumour bulk and spread as well as the growth velocity of the tumour. However, it is important to weigh the postoperative recovery period and the expected residual tumour mass and the systemic options for treatment in the discussion. When the number of metastasis is limited, their surgical removal can also be discussed.

In cases of local recurrence, surgery represents the preferred therapeutic option. If the patient had a long disease-free interval prior to development of the local recurrence, surgery can offer a good probability of long-lasting disease-free survival.

Radiofrequency thermal ablation of liver and lung metastasis below 4 to 5 cm maximum diameter can be an alternative to surgical removal (13). Bone metastasis, e.g. in cases of spinal compression, can be operated to improve neurological impairment, but is also responsive in many cases to radiotherapy. Radiation therapy in ACC has often been considered as not very effective to control tumour growth. However, it has been recently suggested that it could help to prevent local recurrence, if not to prolong survival (12). Whether the tumour bed should be irradiated following initial, presumed curative surgery is widely debated and currently not established.

When completed tumour removal is not possible or in cases of recurrence, medical treatment with o,p′DDD (ortho, para′, dichlorodiphenyldichloroethane, or mitotane) is recommended (21). Mitotane has a specific cytotoxic effect on adrenal cortical cells; it also inhibits steroid synthesis by an action on steroidogenic enzymes. Interestingly, mitotane is usually effective to control steroid excess in patients with secreting ACC. Most series reported in the literature on the efficacy of o,p′DDD in ACC are retrospective analyses with variable results regarding tumour progression. An objective tumour regression has been observed in about 25% of the cases (13). Patients with a cortisol-secreting ACC might have a better survival when treated with mitotane started in the 3 months following surgery of the adrenal tumour (11). A mitotane blood level of at least 14 mg/l seems to improve the tumour response (12, 13). However, side effects (mainly gastrointestinal and, at higher mitotane levels, neurological) often limit the ability to reach this suggested therapeutic plasma level. The daily mitotane dose required to achieve 14 mg/l varies from patients to patients. Therefore close monitoring of mitotane blood level is very helpful to remain in the narrow range between 14 and 20 mg/l, considered by most authors as the therapeutic range of mitotane in ACC. Since o,p′DDD invariably induces adrenal insufficiency, glucocorticoid replacement has to be initiated concurrently with mitotane and should be administered at increased doses (e.g. 40 to 80 mg per day) due to induction of Cortisol Binding Globulin (CBG) and cortisol metabolism by mitotane. The benefit of mitotane treatment as an adjuvant medical treatment after ‘complete’ surgical removal of a Stage 1 or 2 ACC remains to be conclusively demonstrated; however, at present, based on data from retrospective series, adjuvant mitotane is recommended for patients with a high risk of tumour recurrence (large tumour, potential capsule violation, high Ki-67) (27). Randomized international trials are expected to clarify whether mitotane should be recommended postoperatively in all patients with ACC.

Several cytotoxic chemotherapy regimens have been used in ACC. They are usually considered in patients with tumour progression under mitotane therapy. Various drugs have been used and the experience is still limited. It is currently accepted, since the Ann Arbor international conference on adrenocortical cancer (25), that combined treatment with cisplatin, etoposide, and doxorubicin together with mitotane or streptozotocin plus mitotane are the better regimens. The first phase III trial in ACC, the international FIRM-ACT trial, comparing these two regimens is currently in its final phase and will inform future management (12).

Considering the rarity of ACC, significant advances have been made in the last decade in the understanding of the pathophysiology. The advances have also been important for a better diagnosis and might ultimately lead to a better assessment of prognosis. However, much more progress needs to be achieved, especially to improve therapeutic efficiency. Due to the rarity of ACC, collaborative work performed in national and international networks dedicated to adrenocortical tumours will be key for ensuring the development of better diagnostic and therapeutic tools. In Europe this is the goal of ENSAT, which has been developed in the background of several national networks (in France, Italy, Germany, and UK) already working successfully in this field.

The pathological diagnosis of malignancy of an adrenocortical tumour can be difficult in some cases. Although careful analysis by an expert pathologist solves most cases, there are still some suspicious tumours with a borderline Weiss score (3) that are difficult to classify. The prognosis of a tumour diagnosed as malignant, especially after complete surgical resection of a Stage 1 or 2 ACC, is heterogeneous and still difficult to predict.

The surgical procedure has not been defined in a homogeneous way. The benefit of large en bloc aggressive surgical resection, which could lead to kidney ablation, and the strategy for lymph node removal need to be discussed. The possibility, by expert surgeons, to use laparoscopic resection of small ACC restricted to the adrenal without increasing the risk of local recurrence or peritoneal metastasis needs to be determined.

The benefit of radiotherapy has been suggested recently in retrospective studies, while ACC has usually been considered to be nonsensitive to radiotherapy. The place for radiotherapy as adjuvant or curative therapy will have to be established. The benefit of mitotane as adjuvant therapy is most often accepted but needs to be demonstrated in prospective trials.

The development of new immunohistochemical markers should improve the pathological diagnosis and prognostication of ACC. Genomic studies will allow a better classification of adrenocortical tumours leading to the development of molecular markers for the classification and prognostication of adrenocortical tumours. These studies are also giving new insights on the pathophysiology of ACC and this should help to define new targeted therapies. The use of gas chromatography/mass spectrometry assays of urinary steroid currently investigated will help to define steroid profile for the diagnosis and follow-up of ACC. Similar proteomic approaches on urine and plasma are also expected. The development of new specific scintigraphies (such as ¹²³I-iodometomidate or ¹¹C-metomidate) is in progress and should improve tumour diagnosis and follow-up. Radiolabelled tracers could also be used for metabolic radiotherapy. The results of the FIRM-ACT study will determine the respective role of the two cytotoxic chemotherapies currently considered as the best options (cisplatin, etoposide, doxorubicin or streptozotocin). New targeted therapies are currently in preclinical and clinical studies. Among these, inhibitors of the IGF receptors are very attractive in view of the strong evidence for a major role of IGF-2 overexpression in the pathogenesis of ACC.

Dr Frédérique Tissier (Service d’Anatomopathologie, Hôpital Cochin) for help with the figures for this chapter.