Adrenocortical carcinoma: a clinical and immunohistochemical study
MALIHA EDGREN1,2, BARBRO ERIKSSON1, ERIK WILANDER3, STEN NILSSON2, KJELL ÖBERG!
1Departments of Internal Medicine; 2Oncology; 3Pathology, University Hospital, Uppsala, Sweden Received September 30, 1994; Accepted February 2, 1995
Abstract. Twenty-one patients with adrenocortical carcinoma (ACC) were investigated. No correlation was found between size of the primary tumour and the occurrence of metastases, and no correlation was observed between tumour diameter and the functional status of the tumour or survival. Cytokeratins (CK) were expressed in 31% and vimentin (VIM) in 41% of cases. Epithelial membrane antigen (EMA) was detected in one patient with a non-functioning tumour, whereas carcinoembryonic antigen (CEA) was absent in all tumours. The antigens investigated showed no correlation with the morphology or functional status of the tumour. The role of immunohistochemical characterization of this heterogeneous malignancy is discussed.
Introduction
Adrenocortical carcinoma is a rare and highly malignant tumour. According to statistics from several large cancer registries, the age adjusted incidence has been estimated to 2 per million population per year (1,2). ACC can occur in all age groups but is more frequent at the average age of 40-50 years (2-4) Approximately 15 new cases occur annually in Sweden (5). The low incidence rate implies that the tumour is seldom encountered by the physician. Patients with ACC present with symptoms caused by hormone excess or symptoms related to an enlarged abdominal mass (7,8).
Most reports indicate that the majority of adrenocortical carcinomas arise in the left adrenal gland (for review see ref. 6). Bilateral adrenocortical carcinomas occur in a small percentage of patients. It may be difficult to determine whether they represent two primary tumours or if one is a metastasis from the contralateral gland. Cohn et al (7) have reported a series where adrenocortical carcinomas occurred with equal frequency on both sides.
The sex distribution of this malignant disease has been under debate. Several reports describe a higher incidence of this tumour in women, while others have found a male predominance and others equal sex distribution (for review see ref. 9).
The pathogenesis of ACC is unknown. Some reports suggest that ACC may arise in patients with congenital adrenal hyperplasia. Others describe the development of ACC in patients with abnormal adrenal activity (for review see ref 6). About half of the ACC patients have endocrine manifestations (3,10,16). Genetic deletion has been found on chromosome 11q and 17p involving the p53 locus (17).
The tumour may arise from any type of adrenocortical cell. Traditionally, ACC has been classified as functioning, i.e. producing abnormal amounts of adrenal cortical hormones, or non-functioning with little or no hormone production detectable with standard screening tests. More recently, different studies have revealed that most non- functioning tumours produce steroid precursors with little or no biological activity (7). Therefore, the term non- functioning implies that the hormone-related symptoms are absent in patients with these tumours (for review see ref. 8). Frequently, the tumours reach a conspicuous size without clinical signs of endocrine hyperactivity.
The histopathological diagnosis of ACC is enigmatic, because of the lack of distinct morphology in these tumours. They may, therefore, be misinterpreted as metastases from other primary sites. In addition, the immunohistochemical characteristics of ACC are variable, primarily due to an inconsistent expression of cytokeratins (CKs). To address this issue, we studied the histopathology and the immunostaining patterns of 22 ACCs with anticytokeratin antibodies as well as antibodies reactive with vimentin (VIM), eptihelial membrane antigen (EMA) and carcinoembryonic antigen (CEA). The results were compared with the morphology and the functional status of the tumour. Attempts were also made to see whether there existed any correlation between antigen expression and survival.
Materials and methods
Patients. During 1984-1993, 27 patients (10 men and 17 women) with ACC were referred and admitted to the Department of Internal Medicine. In this retrospective study we were able to get access to tumour specimens from 21 of the patients. The diagnosis was confirmed preoperatively by
Correspondence to: Dr Maliha Edgren, Department of Internal Medicine, University Hospital, S-751 85 Uppsala, Sweden
Key words: adrenocortical adenocarcinoma, immunohistochemistry, cytokeratin, epithelial membrane antigen, vimentin, CEA
| Case | Age/Sex | Adrenalectomized | Clinical symptoms | Cushing | Laboratory results | Urine steriod profile |
|---|---|---|---|---|---|---|
| 1ª | 20/M | Right side nephrectomy | Feminizing T with gynecomastia | - | Elev. S testosteron and E2 | ND |
| 2 | 20/F | Left side | Virilizing T | + | Elev. S testosteron, DHEA-SO4 and androstendion | Elev. DHEA and other androgens |
| 3ª | 60/F | Right side | Feminizing, post menopause bleeding | + | Elev. S progesteron, testosoteron and U cortisol | ND |
| 4 | 44/F | Left side | Virilizing T | + | Elev. S testosteron, DHEA-SO4 and U cortisol | Elev. THS, DHEA, ANDRO and pregnantriol |
| 5 | 53/F | Left side | Cushing's syndrome | + | Elev. S testosteron, E2 and U cortisol | ND |
| 6ª | 25/F | Right side, nephrectomy | Virilizing T | + | Elev. S progesteron, E2 and U cortisol | ND |
| 7 | 24/F | Left side, nephrectomy splenectomy | Virilizing T | + | Elev. S testosteron, androstendion and U cortisol | ND |
| 8 | 36/F | Right side | Virilizing T | + | Elev. S Testosteron, DHEA-SO4, androstendion and U cortisol | ND |
| 9ª | 71/F | Left side | Feminizing T vaginal bleeding | + | Elev. S testosteron, E2, progesteron and U cortisol | Elev. DHEA, THS, ANDRO, pregnentriol and pregnantriol |
| 10ª | 59/F | Right side + enucliation liver metastasis | Virilizing T | + | Elev. S testosteron, DHEA-SO4, | Elev. THS E2 and U aldosteron |
| 11 | 29/F | Left side | Virilising T | + | Elev. S testosteron, androstendion and U cortisol | Elev. THS ANDRO and pregnentriol |
ªDeceased; S = serum; U = urinary; E2 = ostradiol; DHEA = dehydroepiandrosterone; THS = tetrahydro-11-deoxycortisol; ANDRO = androsterone
laboratory investigations in addition to CT and/or ultrasonography and chest X-ray. Angiography and fine needle aspiration cytology were also performed in some of the patients prior to surgery. All the patients underwent resection of the primary tumour and the diagnosis was confirmed histopathologically. Surgery was perfomed on 12 patients at the Department of Surgery, University Hospital, Uppsala. The remaining 9 patients were referred to the Department of Internal Medicine after diagnosis had been obtained and surgery performed at the patients’ local hospitals. The median age was 46 for the men (range 20-72 years) and 44 years for the women (range 20-71 years). The right adrenal was the primary site of the tumour in 12 patients (57%), the left adrenal in 10 patients (48%). In one patient, a tumour involving the right adrenal gland was initially disclosed; two years later a second tumour appeared in the left adrenal gland.
Presenting symptoms in patients with hormonally functioning ACC. Ten of the 11 patients (91%) with hormonally functioning tumours were women. In this report, a hormonally functioning tumour is defined as a tumour
associated with clinical features of Cushing’s syndrome, virilization, feminization, hyperaldostronism or a combination of any of these syndromes. All patients with hormonally functioning tumours presented initially with symptoms and signs related to excessive hormone production. Among this group of patients (10 females, 1 male), one patient exhibited Cushing’s syndrome. Nine patients showed some features of Cushing’s syndrome, which was associated with virilism in seven cases and hyperaldosteronism in one case. Hypercortisolism was present in 8 of the patients (73%). This was associated with increased testosteron serum levels in 6 patients (86%). One patient with hyperaldosteronism also had increased testosteron levels. Analysis of adrenal steroid metabolites in urine was performed in 5 of the patients. The results showed abnormal steroid profiles in all. Elevated dehydroepiandro- sterone (DHEA) was observed in 3 cases whereas tetrahydro- 11-deoxycortisol (THS) was elevated in 3 cases. The results obtained from these analyses are presented in Table I. The interval between onset of symptoms and subsequent surgery varied from 4 to 24 months (median 11 months).
| Case | Age/Sex | Adrenalectomy | Signs and symptoms | Laboratory resultsª | Urinary steroid profile |
|---|---|---|---|---|---|
| 12 | 65/F | Right side | Pain | No abnormality | Normal |
| 67 | Left side | Computor tomographyb | No abnormality | Normal | |
| 13 | 59/F | Right side | Fatigue, weight loss, pain, fever, acute abdomen | No abnormality | Normal |
| 14 | 46/M | Left side | Fatigue, weight loss, pain | Elev. urinary cortisol | Elev DHEA, pregnentriol and pregnendiol |
| 15 | 60/M | Left side, splenectomy and nephrectomy | Pain, fever | Elev. urinary cortisol | Elev. androsterone, pregnendiol, pregnentriol |
| 16€ | 42/F | Left side | Fatigue, pain, fever | No abnormality | ND |
| 17c | 72/M | Left side | Fatigue, weight loss | No abnormality | ND |
| 18c | 58/M | Right side and nephrectomy | Weight loss, pain | No abnormality | ND |
| 19 | 60/F | Left side and splenectomy | Weight loss, pain, fever | No abnormality | Elev. THS, pregnendiol pregnentriol |
| 20c | 25/M | Right side and nephrectomy | Fatigue | No abnormality | ND |
| 21 | 33/M | Right side | Accidentally by ultrasound | Elev. urinary cortisol | Elev. THS, pregnentriol 17a-OH-pregnenolone |
“Urinary cortisol, normal value 60-250 nmol/d. “The tumour was discovered in a routine CT check up following the first tumour. “Deceased. THS = Tetrahydro-11-deoxycortisol; DHEA = Dehydroepiandrosterone; ND = not done.
Presenting symptoms in patients with hormonally non- functioning ACC. Ten of the 21 patients (48%) had hormonally non-functioning tumours, i.e. tumours not causing any clinically recognizable endocrine syndrome. The most common presenting symptoms in this group (6 males, 4 females) were abdominal pain in 7/10 (70%) patients and weight loss in 5/10 (50%) patients. In this group of patients, 3 were observed to have a mild increase in urinary cortisol levels, which varied from 300 to 500 nmol/d (normal 60-250 nmol/d). Analysis of adrenal steroid metabolites in urine was performed in 6 of the patients, of which 4 exhibited pathological profiles. Elevated THS was observed in two cases, whereas DHEA was elevated in one case. The results are presented in more detail in Table II. In one patient with bilateral ACC, preoperative urine samples showed normal urinary steroid profiles on both occasions. The interval between the onset of symptoms and subsequent surgery varied from 2 to 12 months (median 6 months.)
Findings at surgery. At surgery, the tumour was confined to the adrenal gland in 13/21 (62%) patients. The median size of the tumour was 12 cm (range 4.7-22 cm). Of the patients with extra-adrenal extension of the disease, 2 were males (25%) with hormonally non-functioning tumours and the remaining 6 patients (75%) were female with hormonally functioning tumours.
Sites of metastases. The most common sites of metastastic deposits were regional lymph nodes (62%), the liver (37%) and the lungs (25%). In patients with metastatic disease, the
diameter of the primary tumour ranged from 7 to 22 cm (median of 12.5 cm).
Histopathology and immunohistochemistry. Formalin-fixed and paraffin-embedded tissue specimens from the 22 tumours were re-examined with light microscopy and the diagnosis ACC was verified according to the criteria of Weiss et al (10) and Slooten et al (11). In two patients, both primary tumours and lymph node metastases were examined. In one patient, the specimen was obtained from the first relapse. Utilizing the ABC method (13), antibodies reactive with cytokeratins (CKs), vimentin (VIM), epithelial membrane antigen (EMA) and carcinoembryonic antigen (CEA) were applied. Information regarding the source and dilution of the antibodies used is listed in Table III.
Results
Clinical characteristics. The clinical and laboratory data at time of diagnosis are summarized in Tables I, II and IV. The presenting features were hormonal excess in 11/21 (52%), fatigue in 5/21 (24%), weight loss in 5/21 (24%) and abdominal pain in 7/21 (33%) patients. Distant metastases were present at the time of diagnosis in 8/21 (38%) patients. No correlation was found between size of the primary tumour and the the occurrence of metastases and no correlation was observed between tumour diameter and the functional status of the tumour. Nor did we observe any correlation between tumour diameter and survival.
| Antibody | Code / clone | Protein concentration | Dilution | Final concentration | Source |
|---|---|---|---|---|---|
| CK* | MN F116 | 92 µg/ml | 1/50 | 1.84 µg/ml | DAKO |
| Chrom A | LK2H10 | 1/1000 | Boehringer Mannheim | ||
| Vimentin | VIM 3B4 | 1 mg/ml | 1/200 | 5 µg/ml | Proteas Progen |
| EMA | E29 | 13.1 g/l | 1/100 | 131 µg/ml | DAKO |
| CEA | MAK431/26 | 1 mg/ml | 1/500 | 2 µg/ml | Behring |
| CK 7 | 1/100 | ||||
| CK 8 | 6D7 | 5.7 mg/ml | 1/4 | 1.4 mg/ml | IDL |
| CK 13 | 1/100 | ||||
| CK 18 | 3F3 | 1.6 mg/ml | 1/2 | 0.8 mg/ml | IDL |
| CK 19 | 1/25 |
*Indicates pretreatment of section with protease before antibody application.
a
b
Histopathology. The growth pattern seen with light microscopy was either solid or alveolar, with the former type predominating. In most tumours, both features were seen in different areas of the same tumour. The tumour cells commonly exhibited a slight to moderate atypia. However, the tumour cells sometimes disclosed an abundance of cytoplasm and areas with marked nuclear polymorphism, including giant and polynuclear cells. These resembled to some degree the histology of pheochromocytomas. A negative argyrophil reaction and lack of chromogranin immunoreactivity were of discriminatory aid in these cases. In four tumours, the dominating cell population was of the clear cell type, consistently exhibited an alveolar growth pattern and could, thus, easily be misinterpreted as renal cell
adenocarcinomas (hypernephromas). The results are presented in Table IV.
Immunohistochemistry. Seven of the 22 tumours (32%) stained positively with the anticytokeratin antibody. Five (71%) of these tumours were hormonally non-functioning. In two tumours (cases no. 2 and 18), the majority of the cells stained positively with the pan-cytokeratin antibody. The 7 pan-cytokeratin positive tumours were also tested with antibodies reactive with various subtypes of cytokeratin i.e. CK7, 8, 18, 19 and 13. All 7 tumours were positive for CK7 and negative for CK13. The results are summarized in Table V.
Vimentin was expressed in 9/21 (41%) tumours. Five of these (56%) were hormonally functioning whereas 4 (44%)
| Patient number | Age | Sex | f/nf | Tumour diameter in cm | Metastasis at operation | Cytokeratin | CEA | EMA | Vimentin | Morphology |
|---|---|---|---|---|---|---|---|---|---|---|
| 1 | 20 | M | f | 18.5 | 0 | + | neg | neg | neg | large cell alveolar |
| 2 | 20 | F | f | 12 | 0 | +++ | neg | neg | neg | large cell solid |
| 3 | 60 | F | f | 13 | + | neg | neg | neg | + | large cell alveolar |
| 4 | 44 | F | f | 9 | 0 | neg | neg | neg | ++ | cytoplasm rich solid |
| 5 | 53 | F | f | 11 | + | neg | neg | neg | neg | cytoplasm rich solid |
| 6 | 25 | F | f | 18 | + | neg | neg | neg | neg | large cell solid |
| 7 | 24 | F | f | 22 | + | neg | neg | neg | ++ | cytoplasm deficient solid |
| 8 | 36 | F | f | 12 | 0 | neg | neg | neg | neg | cytoplasm rich alveolar |
| 8 | 36 | F | f | lymph node | 0 | neg | ND | ND | neg | ND |
| 9 | 71 | F | f | 12 | ·+ | neg | neg | neg | + | cytoplasm rich solid |
| 10 | 59 | F | f | 7 | + | neg | neg | neg | ++ | polymorph cell solid |
| 11 | 29 | F | f | 15 | 0 | neg | ND | neg | neg | ND |
| 12 | 65 | F | nf | 8 | 0 | ++ | neg | neg | ++ | clear cell alveolar |
| 12 | 67 | F | nf | 4.7 | 0 | ++ | neg | neg | ++ | clear cell alveolar |
| 13 | 59 | F | nf | 6 | 0 | ++ | neg | neg | neg | large cell solid |
| 14 | 46 | M | nf | 14 | + | + | neg | neg | neg | polymorph cell alveolar |
| 14 | 46 | M | nf | lymph node | ++ | neg | neg | neg | polymorph cell solid | |
| 15 | 60 | M | nf | 17 | 0 | neg | neg | neg | neg | large cell solid |
| 16 | 42 | F | nf | unknown | unknown | neg | neg | neg | neg | cytoplasm deficient solid |
| 17 | 72 | M | nf | 14 | 0 | neg | neg | neg | neg | cytoplasm deficient solid |
| 18 | 58 | M | nf | 7 | + | +++ | neg | ++ | neg | large cell alveolar |
| 19 | 60 | F | nf | 17 | 0 | neg | neg | neg | neg | polymorph cell solid |
| 20 | 25 | M | nf | 10 | 0 | neg | neg | neg | + | polymorph alveolar |
| 21 | 33 | M | nf | 7 | 0 | neg | neg | neg | ++ | clear cell alveolar |
f = functioning; nf = non functioning; neg=negative
| Case | F/NF® | CK | CK7 | CK8 | CK18 | CK19/13 |
|---|---|---|---|---|---|---|
| 1 | F | + | ++ | - | - | -/- |
| 2 | F | +++ | ++ | +++ | +++ | -/- |
| 12 | NF | 1++ | ++ ++ | - | - | -/- |
| NF | 2+ | - | - | -/- | ||
| 13 | NF | ++ | ++ | ++ | ++ | -/- |
| 14 | NF | + | ++ | ++ | ++ | -/- |
| 14 | *LN | ++ | ++ | ++ | ++ | -/- |
| 18 | NF | +++ | ++ | ++ | ++ | -/- |
Case 12 had bilateral ACCs discovered two years apart. F = functioning; NF = non-functioning; “LN = lymph node found at operation.
were non-functioning. In the vimentin-positive tumours, the stained cells were always in the minority. Seven of the 9 immunostained tumours were cytokeratin-negative while the 2 tumours from the patient with bilateral ACC (no 12) displayed positive immunostaining for both vimentin and cytokeratin. Metastases were present at the time of diagnosis in 4 vimentin-positive tumours, compared with 2 cytokeratin- positive tumours. No correlation was seen between cytokeratin and/or vimentin expression and survival.
All tumours lacked CEA immunoreactivity whereas one showed EMA expression (case no. 18). No correlation was found between the immunoreactivity and the morphological picture of the tumours in routinely haematoxylin-eosin stained sections. The immunohistochemical results are presented in Table IV.
Discussion
ACC can occur in all age groups but more is common between the ages of 40 and 50 years (2-4). The median age in our series was 46 for men and 44 years for women. In agreement with several reports (3,7,9), the majority of tumours occurred in women, with a female/male ratio of 2:1.
The results obtained in the present study show that 50% of the patients had endocrine manifestations. This is in accordance with previously reported data (3,16). It is also well known that malignant cells are highly inefficient in their synthesis of steroid hormones and that a large tumour mass is required to give elevated steroid levels (18). This could explain the larger tumour mass median (12 cm) in the patients with functioning tumours compared with patients with the somewhat smaller (median 10 cm) non-functioning tumours. One would expect that a large tumour mass would easily invade adjacent organs and give rise to metastases. Metastases were observed in 8 cases, of which 6 had hormonally functioning tumours.
Clinically, a virilizing ACC in a man and a feminizing tumour in a woman could easily be missed. This may explain the delay in obtaining the diagnosis in patients with clinical endocrinopathy (median 11 months) compared with patients without clinical endocrinopathy (median 6 months) and the frequency of metastases (75 and 25%, respectively). An explanation could also be that well-differentiated and steroid- hormone-producing tumours grow more slowly than less well-differentiated, non-functioning tumours (18).
Generally, functioning tumours are more common in female ACC patients (3,7,9). This finding was verified in our study, since the female/male ratio was 10/1. We also found that this tumour type was frequent, particularly in the younger age group (32 years; range 20-72).
Non-functioning tumours tend to predominate in males and older patients (for review see ref. 9). We also observed a slight predominance of males (6/10) in this group of patients. The median age of these patients was 59 years (range 25-72).
The diagnosis of ACC is enigmatic in that the histopathological appearance is non-discriminatory. Furthermore, when applying various immunocytochemical tumour markers, inconclusive results are often obtained with respect to the phenotypic expression of epithelial and mesenchymal intermediate filament proteins and, thus insufficient for a differential diagnosis of both metastatic epithelial and mesenchymal malignant tumours. Histopathologically, two main features are seen in the adrenocortical carcinomas; one type had a solid alveolar growth pattern and one a clear cell pattern. The former type may be misinterpreted as a metastasis of an undifferentiated carcinoma from other primary sites and the clear cell type is structurally quite similar or even identical to the clear cell carcinoma (hypernephroma) of renal cell origin. Because of this, the diagnosis of an adrenal cortical carcinoma is best verified after evaluation of both clinical and pathological features, including symptoms, presence of signs of increased hormonal activity, absence of malignant tumour of other origin and evaluation of macro- and microscopic pattern of the adrenal cortical mass. In the present study, all these phenomena were evaluated before the final diagnosis of primary tumours of adrenocortical origin was made.
One explanation for the lack of cytokeratin positivity in some ACCs in our study could be that the epitopes are destroyed or masked by the formalin fixation and possibly also by the paraffin-embedding procedure. This has been observed previously by other authors (13). The specimens were enzyme pretreated in order to unmask the antigens.
However, another explanation may be that cytokeratin expression is lost on adrenocortical neoplastic transformation in certain ACCs or is of too low a density to be detected immunohistochemically. In our series of 21 ACC patients, we noted the presence of cytokeratin in the tumour tissue in 7 cases. These tumours were also tested against CK 7, 8, 13 and 19, respectively. We observed the absence of CKs 19 and 13 and the presence of CK 7 in all these tumours. The inability to demonstrate all CK subtypes predicted to be present by immunohistochemistry has been described by other authors (for review see ref. 13). Cytokeratins were expressed in 5 of the hormonally non-functioning tumours. However, we do not know whether the expression of cytokeratin is, in some way, related to the functional status of the tumour. We also observed that the expression of CK had no correlation to the patient survival.
Vimentin, which is an intermediate filament present in normal as well as malignant mesenchymal tissues, was observed in 9 tumours. This result is in agreement with those reported by other authors (14). Vimentin expression in neoplastic adrenal cortex seems to have an inverse relationship to cytokeratin expression (10). Seven vimentin- positive tumours in our series lacked cytokeratin expression. The presence of VIM in 9/22 ACC cases is in accordance with previous reports (13,14). Interestingly, both tumours from the patient with bilateral, non-functioning ACC expressed both intermediate filament types, i.e. cytokeratin as well as vimentin. This could, possibly, reflect a different biological behaviour for bilaterally occurring tumours. Vimentin was expressed almost equally in functioning (5/11) and non-functioning (4/10) tumours. We observed no correlation between the expression of VIM and/or cytokeratin levels and survival.
It has previously been described that EMA and CEA are not expressed in normal or neoplastic adrenocortical cells (10). Our data support this finding concerning CEA. However, EMA expression was seen in 1/22 tumours (patient no. 18). This patient had a non-functioning ACC with a large cell alveolar appearance.
In conclusion, our analyses of cytokeratins and vimentin in ACC are in agreement with previous results; i.e. one third of the tumours expressed cytokeratins and about one third expressed vimentin but only rarely did the expression of both antigens occur in the same tumour. We observed no correlation between immunoreactivity and the histo- pathological pattern of the tumours. Nor did we observe any correlation between antigen expression and survival. Thus, the usefulness of these stainings in the differential diagnosis of ACC versus other tumours and as prognostic markers appears to be limited, as is the use of EMA and CEA. The need for better, clinically useful, makers in the management of this disease is, thus, urgent.
References
1. Karakousis CP, Rao U, Moore R: Adrenal Adenocarcinomas: histologic grading and survival. J Surg Oncol 29: 105-111, 1985.
2. Söreide JA, Barbarnd K, Thoresen S: Adrenal cortical carcinoma in Norway 1970-1984. World J Surg 16: 663-668, 1992.
3. Henly DJ, Van Heerden JA, Grand CS, Carney JA, Carpenter PC: Adrenal cortical carcinoma: A continuing challenge. Surgery 94: 926, 1983.
4. Brennan MF: Adrenocortical carcinoma. Cancer 37: 348-365, 1987.
5. National Board of Health and Welfare: The Cancer Registry, Cancer Incidence in Sweden 1984. Stockholm, 1987.
6. Joshua I, Barzilay S, Artemis G, Pazianos: Adrenocortical carcinoma, urologic clinic of North America 3: 457-468, 1987.
7. Cohn K, Gottesman L, Brennan M: Adrenocortical carcinoma. Sugery 100: 1177-1179, 1986.
8. Pommier RF, Brennan MF: Management of adrenal neoplams. Current Problems. Surgery 28: 659, 1991.
9. Decker RA, Keuhner ME: Adrenocortical carcinoma. Amer Surgeon 57: 502-513, 1991.
10. Weiss LM, Medeiros LJ, Vickery AL: Pathologic features of prognostic significance in adrenocortical carcinoma. Am J Surg Pathol 13: 202-206, 1989.
11. Slooten H van, Schaberg A, Smeenk D, Moolenaar J: Morphologic characteristics of benign and malignant adrenocortical tumor. Cancer 55: 766-773, 1985.
12. Hsu SM, Raine L, Fanger H: Use of avidin-biotin peroxidase complex (ABC) in immunoperoxidase technique: A comparison between ABC and unlableled antibody (PAP). J Histochem Cytochem 29: 577-580, 1981.
13. Gaffey NJ, Traweek T, Mills SE, Travis WD, Lack EE, Medeiros JM, Weiss LM: Cytokeratin expression in adrenocortical neoplasia: an immunohistochemical and biochemical study with implications for the differential diagnosis of adrenocortical hepatocellular and renal cell carcinoma. Human Pathol 23: 144-153, 1992.
14. Markuu M: Neuroendocrine differentiation in adrenocartical carcinoma: new immunohistochemical findings supported by electron microscopy. Laboratory Invest 66: 169-174, 1992.
15. Lippset MB, Hertz R, Ross GT: Clinical and pathological aspects of adrenocortical carcinoma. Am J Med 35: 374, 1963.
16. Thompson NW: Adrenocortical carcinoma in endocrine surgery update. NW Thompson, AI Vinik (eds). New York, London, Grune and Stratton, pp119-128, 1983.
17. Ohgaki H, Kleihues P, Heitz PU: p53 mutation in sporadic adrecortical tumors. Int J Cancer 54: 408-410, 1993.
18. Sasano H, Suzuki T, Nagura H, Nishikawa T: Steriodogenesis in human adrenocortical carcinoma: Biochemical activities immunohistochemistry and in situ hybridization of steroidogenic enzymes and histopathologic study in nine cases. Human Pathol 24: 397-404, 1993.