Adrenocortical Carcinoma An Immunohistochemical Comparison With Renal Cell Carcinoma
MARK R. WICK, MD, DAVID L. CHERWITZ, MD, RONALD C. McGLENNEN, BS, MS, and LOUIS P. DEHNER, MD
From the Division of Surgical Pathology, Department of Laboratory Medicine and Pathology, University of Minnesota School of Medicine, Minneapolis, Minnesota
The diagnosis of adrenocortical carcinoma (ACC) is of- ten difficult, because this tumor may present with direct extension into adjacent renal parenchyma or with metastatic disease. Renal cell carcinoma and other histo- logically similar tumors are potentially confused with ACC by conventional light microscopy, and their sepa- ration from the latter is often impossible without the aid of additional studies. Furthermore, the distinction be- tween adrenal cortical adenoma and ACC may also be problematic. Because of these factors, the authors stud- ied 10 cases each of ACC, adrenocortical adenoma, and renal cell carcinoma (RCC) immunohistochemically, in an attempt to develop objective parameters which may aid in this differential diagnostic dilemma. Nontryp- sinized, formalin-fixed, paraffin-embedded specimens
were used in all cases, and tissue from the adrenocortical tumors was also studied for intermediate filament con- tent after protease digestion. All 20 nontrypsinized ad- renocortical neoplasms were positive for vimentin, but not for cytokeratin, epithelial membrane antigen, or blood group isoantigens. Conversely, each of 10 cases of RCC expressed epithelial membrane antigen, cytokera- tin, and blood group isoantigens, but none was im- munoreactive for vimentin. Two adenocortical carci- nomas and three adenomas manifested cytokeratin positivity after trypsin digestion. There were no significant differences between the immunostaining profiles of ACC and adrenocortical adenoma, which sug- gest that this distinction must still rely upon clinical and morphologic criteria. (Am J Pathol 1986, 122:343-352)
IN THE ABSENCE of clinical signs of adrenal corti- cal hyperfunction such as Cushing’s syndrome or inappropriate masculinization or feminization, the preoperative diagnosis of adrenocortical carcinoma (ACC) may present considerable difficulties. The spec- imen in question may represent a metastatic nodule or a lymph node containing trabecular profiles, nests, or formless sheets of large epithelioid cells, with granu- lar, eosinophilic, or clear cytoplasm. In these circum- stances, the differential diagnosis includes not only ACC but renal cell carcinoma (RCC), hepatocellular carci- noma, malignant melanoma, and anaplastic carcinoma of the lung and other sites.1 Because of the juxtaposi- tion of the adrenal gland to the kidney, it is not un- common for ACC to involve the renal parenchyma at diagnosis. Also, RCC may occasionally present as a metastasis to the contralateral adrenal gland.2 The con- sequences of the latter situation may include an inap- propriate adrenalectomy/nephrectomy for what is er- roneously regarded as an ACC.
Another problem with respect to adrenocortical ne- oplasms, even in the presence of endocrine dysfunc-
tion, is the pathologic distinction between benign and malignant tumors. Criteria derived from multivariate analysis of the gross and microscopic features of clini- cally verified adenomas and carcinomas3-8 still do not allow this separation to be made in all cases.
With these difficulties in mind, we compared the im- munohistochemical profiles of 10 cases of clinically confirmed ACC with those of RCC and adrenocorti- cal adenoma. In this report, the results of this analysis are presented.
Materials and Methods
The files of the Division of Surgical Pathology at the University of Minnesota were surveyed for examples of ACC, seen between 1960 and 1984. Retrieved cases were
| Reagent | Source | Dilution |
|---|---|---|
| Anti-cytokeratins | ||
| 44, 46, 52, 54 kd-clone designation PKK1 | Lab Systems, Inc. | 1:160 |
| 54 kd-clone designation 35BH11 | Enzo Biochem, Inc. | 1:4000 |
| 49, 51, 57, 66 kd-clone designation 34BE12 | Enzo Biochem, Inc. | 1:4000 |
| 40, 50, 56.5, 58, 65-67 kd-clone | Hybritech, Inc. | 1:200 |
| designations AE1/AE3 | ||
| Anti-vimentin | ||
| Clone designation PK-V | Lab Systems, Inc. | 1:40 |
| Single immunoblot 58 kd | Dr. Susan Simonton | 1:80 |
| Clone designation BM52 | Department of Pathology University of Minnesota | |
| Anti-carcinoembryonic antigen | Hybritech, Inc. | 1:160 |
| Anti-epithelial membrane antigen | DakoPatts Co., Inc. | 1:160 |
| Anti-S100 protein | DakoPatts Co., Inc. | 1:1600 |
| Anti-alphafetoprotein | DakoPatts Co., Inc. | 1:700 |
| Anti-a, antitrypsin | DakoPatts Co., Inc. | 1:320 |
| Anti-ß2 microglobulin | Dr. Russell Curry, Department of Medicine, Univeristy of Minnesota | 1:300 |
| Anti-blood group isoantigens A, B, and H | DakoPatts, Inc. | 1:80 |
| Biotinylated peanut agglutinin | Vector Laboratories | 1:1600 |
| Biotinylated wheat germ agglutinin | Vector Laboratories | 1:3200 |
| Biotinylated soybean agglutinin | Vector Laboratories | 1:4000 |
| Biotinylated Dolichos biflorus agglutinin | Vector Laboratories | 1:3200 |
| Biotinylated Ulex europaeus lectin | Vector Laboratories | 1:3600 |
| Sheep antirabbit globulin | Antibodies, Inc. | 1:75 |
| Horse antimouse globulin (biotinylated) | Vector Laboratories | -* |
| Rabbit peroxidase-antiperoxidase complex | Sternberger Laboratories | 1:300 |
| Avidin-biotin-peroxidase complex | Vector Laboratories | -* |
* Vectastain Kit (Mouse).
reexamined histologically, and clinical data on them were reviewed. Only those neoplasms which behaved in a malignant fashion were retained in the study group (10 cases). For comparison, 10 cases of adrenocortical adenoma were selected for study, along with 10 ran- domly chosen examples of RCC.
Formalin-fixed, paraffin-embedded tissue from these 30 cases was stained with the use of the peroxidase- antiperoxidase (PAP) or avidin-biotin-peroxidase com- plex (ABC) techniques, as previously described.9,10 Sections were deparaffinized in xylene, incubated in methanolic hydrogen peroxide solution (0.6%) for 30 minutes, and rehydrated in graded alcohols and distilled water. After incubation in phosphate-buffered saline (PBS, pH 7.5) for 10 minutes, primary antibodies were applied, and sections were then incubated in moisture chambers for 18 hours at 4 C. Prior protease digestion was not employed in this portion of the study.
After rinsing with PBS, biotinylated horse anti- mouse or sheep anti-rabbit globulins were applied (for monoclonal and polyclonal primary antibodies, respec- tively), followed by incubation at room temperature for 1 hour. Sections were again rinsed in PBS and incubated with ABC or rabbit PAP complex for 1 hour at room temperature. Sections to which lectins had been applied were incubated with ABC for 1 hour at room tempera-
ture, after rinsing in PBS. Following a final rinse in phosphate buffer (pH 7.4), chromogenic development was accomplished by immersion of sections in phos- phate buffer-3,3’-diaminobenzidine solution (0.025-0.25 mg/ml) with 0.01% hydrogen peroxide, for 10 minutes. They were counterstained with Harris’ hematoxylin, de- hyrated, coverslipped with Permount, and examined by conventional light microscopy.
In addition, sections from each of the 20 adrenocor- tical tumors were exposed to 0.1% bovine trypsin (Sigma Chemical Co.) in 0.01% HC1/PBS, for 20 minutes at room temperature, before application of anti-cyto- keratin and anti-vimentin antibodies. The remainder of the immunohistochemical ABC technique was car- ried out as specified above.
Primary antibodies utilized in these procedures in- cluded monoclonal anti-cytokeratin (hydridoma clones PKK1, AE1/AE3, 35BH11, and 34BE12),11-13 anti- vimentin (hybridoma clones PK-V and BM-52),14 anti- carcinomembryonic antigen (CEA), -epithelial mem- brane antigen (EMA), and anti-blood group isoantigens A, B, and H (BGI). Polyclonal rabbit antibodies to S100 protein, a-fetoprotein (AFP), a-antitrypsin (AAT), and ß2-microglobulin (B2M) were also employed. In addi- tion, biotinylated peanut agglutinin (PNA), wheat germ agglutinin (WGA), soybean agglutinin (SBA), Dolichos
| Case | Age/sex | Tumor size/weight | Adrenocortical Clinical endocrinopathy? | carcinoma Treatment | Outcome |
|---|---|---|---|---|---|
| 1 | 55/F | 10.5 cm GD /NR | 0 | None | Died 1 month after diagnosis with metastases to heart, lung, bone marrow, pancreas, and brain. |
| 2 | 80/F | Biopsy only | + : Cushing's | None | Died 1 month after diagnosis with metastases to liver |
| 3 | 67/F | Biopsy only | + : Cushing's | None | Died 2 weeks after diagnosis with metastases to liver and lungs |
| 4 | 29/F | 16 cm GD/ 675 g | +: Cushing's | S + C | Metastases to lungs 4 years after diagnosis, surgically resected; patient free of disease 8 years after diagnosis |
| 5 | 20/M | 14 cm GD/ 1200 g | 0* | S + C | Alive with metastases to lungs and liver, 18 months after diagnosis |
| 6 | 62/M | 12 cm GD/ NR | + : Conn's | S + C | Alive with metastases to lungs, 4 years after diagnosis |
| 7 | 36/M | 10 cm GD/ NR | 0 | S | Died 4 months after diagnosis with metastases to bones, lungs, and soft tissue |
| 8 | 55/M | NR/NR | 0 | S + R + C | Alive 10 years after diagnosis; local recurrence 8 years after diagnosis (resected): metastases to lungs 10 years after diagnosis |
| 9 | 66/M | 15 cm GD/ 500 g | 0 | S | Alive with metastases to lungs, 3 years after diagnosis |
| 10 | 30/M | NR/NR | 0 | S + C | Pulmonary metastasis at diagnosis (resected); alive 3 months after diagnosis |
| Case | Age/sex | Tumor size/weight | Adrenocortical Clinical endocrinopathy? | adenomas | Outcome |
|---|---|---|---|---|---|
| Treatment | |||||
| 11 | 9 mo/M | 7 cm GD/ 200 g | +: Cushing's | S | NED: 3 years |
| 12 | 71/F | 4 cm GD/ 50 g | 0 | S | NED: 3 years |
| 13 | 68/F | NR/NR | 0 | S | NED: 2.5 years |
| 14 | 19/F | 3 cm GD/ NR | + : Cushing's | S | NED: 2 years |
| 15 | 46/M | 1.8 cm GD/ NR | + : Conn's | S | NED: 4 years |
| 16 | 60/F | 2.5 cm GD/ NR | + : Cushing's | S | NED: 6 years |
| 17 | 49/F | 1.6 cm GD/ NR | + : Conn's | S | NED: 5 years |
| 18 | 3/F | 6.5 cm GD/ NR | + : Conn's | S | NED: 5 years |
| 19 | 27/F | 5 cm GD/ 30.5 g | + : Cushing's | S | NED: 5 years |
| 20 | 39/F | 2 cm GD/ 6 g | + : Conn's | S | NED: 2 years |
| Case | Age/sex | Clinical paraneoplasia? | Renal cell Treatment | carcinomas Metastases? | Outcome | |
|---|---|---|---|---|---|---|
| 21 | 73/M | 0 | S | 0 | DOC: 0.6 years | |
| 22 | 56/F | + : Hypertension | S | + : RLN | NED: 6.2 years | |
| 23 | 73/F | 0 | S | + : Lungs, bones | NED: 6 years | |
| 24 | 60/M | ? (Adrenal hyperplasia) | S | + : Lungs | DOD: 0.7 years | |
| 25 | 60/F | 0 | S | + : Lungs, bones | DOD: 0.7 years | |
| 26 | 61/F | 0 | S | + : Lungs, bone, brain | DOD: 0.2 years | |
| 27 | 70/M | 0 | S | ? | DUC: 0.9 years | |
| 28 | 57/M | 0 | S | 0 | NED: 5 years | |
| 29 | 46/F | 0 | S | + : RLN | NED: 5 years | |
| 30 | 57/M | 0 | S | 0 | NED: 4, 5 years |
GD, greatest dimension; NR, not recorded; S, surgery; R, radiotherapy; C, chemotherapy; NED, no evidence of disease; DOC, dead of other causes, DOD, dead of disease (tumor); DUC; dead of unknown cause; RLN, regional lymph nodes.
* Elevated level of urinary 17-ketosteroids.
biflorus lectin (DBL), and Ulex europaeus lectin (UEL) were used in histochemical analyses of the 30 tumors. The sources and working dilutions for these reagents are listed in Table 1.
Positive controls were represented by sections of stock tissues and neoplasms known to contain the antigens being studied. Negative control sections of the 30 ne- oplasms were obtained by substituting nonimmune rab-
bit serum or murine ascites fluid for primary antibod- ies, in the PAP and ABC methods, respectively.
Results
Clinical Findings
The clinical aspects of each case are summarized in Table 2.
Pathologic Findings
Macroscopic Features
The gross features in cases of resected ACC were typ- ical of this diagnosis.3-6,8 The native adrenal gland usu- ally was not apparent within such masses. The sizes of ACC averaged 13 cm in greatest dimension, and the mean weight in three examples was 792 g.
In contrast, adrenocortical adenomas were well- circumscribed and were bordered by easily identifiable adrenal tissue. They were brown or yellow, without hemorrhage or necrosis, and measured 3.1 cm in aver- age diameter; the largest weighed 200 g.
All 10 RCCs had a classic gross appearance. None involved the adrenal glands.
Microscopic Observations
Adrenocortical Carcinomas
Variable cytologic appearances characterized the cases of ACC in this series. Four tumors were composed of round to polygonal cells, with round or slightly ir- regular nuclei and indistinct nucleoli. The cytoplasm was abundantly vacuolated in these tumors, which gave them a “clear-cell” appearance (Figure 1). Another 4 cases had smaller cells with granular eosinophilic cytoplasm (Figure 2). Two further cases were composed of sheets of poorly cohesive, polygonal, or slightly spin- dled cells, with vesicular nuclei, prominent nucleoli, and scant amphophilic cytoplasm.
Additional features included extensive necrosis, oc-
casional “giant” tumor cells, and mitotic rates ranging from 1-3 per high-power (400x) field, with abnormal mitoses in 8 cases. Vascular invasion was observed in 4 cases, and capsular invasion was seen in 5.
Adrenocortical Adenomas
Nine adenomas were typified by nests and cords of cohesive, vacuolated tumor cells, with round to oval nuclei, abundant cytoplasm, and distinct borders. Mi- toses were lacking, but focal nuclear hyperchromasia and atypia were seen in 3 cases. None displayed nucleo- lar prominence, inflammation, necrosis, capsular per- meation, or vascular invasion.
Renal Cell Carcinomas
All 10 renal cell carcinomas included in this study had a clear-cell cytologic appearance (Figure 3). Five were Grade 2 neoplasms, and 5 were Grade 3, accord- ing to the criteria of Fuhrman et al.15 These tumors dis- played clustered or medullary cellular growth patterns, with focal necrosis and hemorrhage. Mitoses averaged one to two per high-power field.
Immunohistochemical Findings (Table 3)
All 20 adrenocortical tumors, whether benign or malignant, stained positively for vimentin content in a diffuse cytoplasmic pattern, both with and without prior trypsinization (Figure 4). Nontrypsinized speci- mens of these neoplasms were uniformly negative for cytokeratin, EMA, AAT, AFP, S100 protein, CEA, BGI,
and UEL and DBA binding. Two ACCs bound PNA, 3 bound SBA, and 1 bound WGA. Two displayed mem- brane positivity for B2M.
In addition to vimentin, B2M was the only other an- tigen expressed by nontrypsinized adrenocortical ade- nomas (3 cases) (Figure 5). The nonneoplastic adrenal cortical tissue included in specimens of adenoma showed identical antigenic profiles.
Two trypsinized adrenocortical carcinomas showed reactivity with anti-cytokeratins AE1/AE3 and PKKI; 3 similarly treated adrenocortical adenomas and 5 spec- imens of adenoma-related normal adrenal cortex showed comparable positivity. None reacted with 35BHIl or 34BE12.
Nontrypsinized renal cell carcinomas uniformly ex- pressed EMA (Figure 6) and cytokeratin. All displayed
| Antigen/lectin | ACC | ACA | RCC |
|---|---|---|---|
| Cytokeratins | |||
| AE1/AE3 | 0/10++ | 0/10# | 10/10 |
| PKK1 | 0/10# | 0/10# | 7/10 (Cases 21, 22, 23, 26, 27, 28, 29) |
| 35BH11 | 0/10 | 0/10 | 2/10 (Cases 26, 28) |
| 34BE12 | 0/10 | 0/10 | 2/10 (Cases 26, 28) |
| Vimentin | |||
| PK-V | 10/10 | 10/10 | 0/10 |
| BM-52 | 10/10 | 10/10 | 0/10 |
| Carcinoembryonic antigen | 0/10 | 0/10 | 0/10 |
| Epithelial membrane antigen- | 0/10 | 0/10 | 10/10 |
| S100 Protein | 0/10 | 0/10 | 0/10 |
| a-fetoprotein | 0/10 | 0/10 | 0/10 |
| a,-antitrypsin | 0/10 | 0/10 | 0/10 |
| ß2-microglobulin | 2/10 (Cases 2, 8) | 3/10 (Cases 12, 15, 17) | 7/10 (Cases 22-25, 28-30) |
| Blood group isoantigens | 0/10 | 0/10 | 10/10 |
| Peanut agglutinin | 2/10 (Cases 1.5) | 0/10 | 7/10 (Cases 21-26, 30) |
| Wheat germ agglutinin | 1/10 (Case 1) | 0/10 | 6/10 (Cases 21, 22, 24-27) |
| Soybean agglutinin | 3/10 (Cases 1, 4, 5) | 0/10 | 2/10 (Cases 21, 26) |
| Dolichos biflorus lectin | 0/10 | 0/10 | 5/10 (Cases 21-23, 28, 29) |
| Ulex europaeus lectin | 0/10 | 0/10 | 0/10 |
ACC, adrenocortical carcinoma; ACA, adrenocortical adenoma; RCC, renal cell carcinoma.
* All results refer to exclusive use of nontrypsinized specimens.
t Cases positive/cases studied.
# Cases 2, 4, 15, 17, and 18 were positive with these antisera, after trypsin digestion.
reactivity with AE1/AE3, 7 were positive with PKKI, and 2 reacted with 35BHIl and 34BE12. Seven tumors manifested B2M positivity. Nine bound PNA, 2 bound SBA, 6 bound WGA, and 5 bound DBA. None dis- played binding of UEL. All 10 renal cell carcinomas expressed immunoreactivity for BGI (Figure 7), in a mixed membranous and cytoplasmic pattern. In con- trast, all failed to stain for CEA, AFP, AAT, or S100 protein content.
Positive and negative controls stained appropriately.
Discussion
Adrenocortical carcinoma is an uncommon neo- plasm, with a reported incidence of 2 cases per million individuals per year.16 Hence, when ACC presents with metastasis, pathologists are confronted with a relatively unfamiliar tumor in terms of its morphologic recogni- tion. This problem is compounded by the overlapping histologic features of ACC, RCC, malignant mela- noma, hepatocellular carcinoma, and anaplastic carci-
noma of the lung, as mentioned previously. These neo- plasms may have metastasized to the adrenal glands at diagnosis, and further confusion therefore results as to which organ is the primary site.
Of these differential diagnostic possibilities, RCC is usually the most troublesome entity to exclude, by con- ventional histologic study. If electron-microscopic spec- imens are available, ultrastructural study can contrib- ute to the separation of this tumor from ACC in some cases,1 but we elected to focus our attention on an im-
munohistochemical comparison between RCC and ACC, because this modality of study is more widely available and less expensive. The results of this analy- sis indicate that these two tumors are indeed separable by such means. All 10 ACCs were vimentin-positive and EMA- and BGI-negative, while the converse of these findings was obtained in an equal number of renal cell carcinomas. In addition, all cases of RCC contained cytokeratin, whereas none of the adrenocortical carci- nomas manifested reactivity for this intermediate fila-
ment with one of the methods we employed (lack of trypsin digestion).
Procedural dissimilarities account for the differences between some intermediate filament reactivities seen in our study and those observed by Miettinen et al17 in adrenal tissues and tumors. These authors noted the presence of cytokeratin in most of the adrenocortical tumors which were analyzed, but they uniformly em- ployed trypsin digestion of formalin-fixed tissue, as well as specimens fixed in absolute alcohol, and frozen sec- tions. These factors are known to enhance cytokeratin reactivity, compared with that seen in unmodified formalin-fixed tissue, as also suggested by our results in trypsinized tumors.18 Indeed, it is likely that they are necessary to demonstrate keratin proteins in formalin- fixed adrenal tumors, perhaps because of low antigen density in such lesions. On the other hand, RCC ap- pears to express cytokeratin positivity even in nontryp- sinized specimens, especially with selected monoclonal
antibodies (AE1/AE3 and PKK1). Hence, even though it does not maximize cytokeratin detection, the omis- sion of protease digestion of formalin-fixed tissue con- tributes to a diagnostically useful means of separating ACC and RCC, in concert with immunostaining for EMA and BGI. Sun et al19 and Nagle and colleagues20 have also observed a lack of cytokeratin reactivity in adrenocortical neoplasms, using nontrypsinized tissue.
The absence of EMA observed in adrenal cortex and adrenocortical tumors is in accord with the findings of Sloane and Ormerod.21 Although the adrenal is con- ceptualized as an “epithelial” structure, paraffinized specimens of this gland and its tumors appear to lack this membrane antigen, which is expressed by most other epithelia.
The observation of vimentin in ACC also has liter- ary precedent. Miettinen et al17 found this intermedi- ate filament in 8 of 21 formalin-fixed adrenocortical tumors. Although we did not detect vimentin positiv-
| Tumor | CKER | EMA | VIM | AFP/AAT | CEA | S100 | BGI |
|---|---|---|---|---|---|---|---|
| Adrenocortical carcinoma | 0 | 0 | + | 0 | 0 | 0 | 0 |
| Renal cell carcinoma | + | + | + | 0 | 0 | 0 | + |
| Hepatocellular carcinoma | + | 0 | 0 | + (75-92%) | + (30%) | 0 | NS |
| Anaplastic carcinomas of lung | + | + | 0 | NS | + | 0 | ± |
| Malignant melanoma | 0 | 0 | + | NS | 0 | + | 0 |
CKER, cytokeratin; EMA, epithelial membrane antigen; AFP/AAT, a-fetoprotein/a,-antitrypsin; CEA, carcinoembryonic antigen; S100, S100 protein; BGI, blood group isoantigens A, B, H; NS, not studied.
* Predicted results are based on the authors’ experience and published data on respective tumors (see text); all results refer to use of nontrypsinized specimens.
ity in any of 10 renal cell carcinomas, Herman et al have shown that such tumors may indeed contain this protein. 22
Results of lectin binding studies in cases of ACC and RCC showed no significant differences, negating their value in differential diagnosis of these two tumors. Simi- larly, further immunohistochemical studies of RCC and ACC which we have performed with monoclonal anti- bodies to urinary tract antigens (URO-2,3,4, Ortho Di- agnostics Systems, Inc.) have shown a lack of discrimi- nation between the two neoplasms (unpublished data). Others have made comparable observations on the specificity of these reagents.23
The immunohistologic features of ACC were virtu- ally identical to those of adrenocortical adenoma and normal adrenal cortex. A loss of cell membrane-asso- ciated B2M and BGI has been linked to malignant trans- formation in neoplasms of the skin and the urinary tract, respectively,24.25 but a trend of this type was not apparent in our cases of ACC and adrenocortical ade- noma. Therefore, one must still rely upon the use of microscopic criteria such as those outlined by Weiss6 and van Slooten et al8 to discern the biologic potential of adrenocortical neoplasms. These include nuclear atypia, mitotic activity, the presence of atypical mitoses, loss of cytoplasmic clarity, a diffuse growth pattern, necrosis, and vascular or capsular invasion, as indica- tors of carcinomatous change. Additional factors such as overall tumor size and weight are also helpful in the evaluation of the likely clinical behavior of any given adrenocortical tumor.26,27
Additional antisera which were employed in this se- ries were chosen to provide comparative, methodology- related, immunohistochemical data on ACC, with ref- erence to hepatocellular carcinoma, malignant mela- noma, and anaplastic non-small-cell lung cancer. Al- though a systematic intralaboratory comparison of these tumors with ACC was not performed, published ac- counts and our own informal observations of their im- munostaining characteristics indicate that ACC is dis- similar to each of them, if one utilizes nontrypsinized tissue. Most hepatocellular carcinomas are cytokeratin-, AFP-, and AAT-positive28,29 (11 of 12 cases studied in our laboratory); whereas ACC appears to lack all three of these substances. Malignant melanomas, like ACC, are reactive for vimentin but lack EMA and cytokera- tin.20,21,30 However, the overwhelming majority of mela- nomas contain S100 protein, in contrast to ACC.31 Pul- monary large-cell carcinomas and poorly differentiated adenocarcinomas have uniformly expressed cytokera- tin, CEA, and EMA, in our experience (15 cases) and that of others.21,32 Therefore, it would appear that im- munohistochemical assessment (with the methodologic stipulations discussed above) is valuable in the differen-
tial diagnosis of carcinomas arising in the adrenal cor- tices, kidneys, liver, and lungs, and in their separation from amelanotic malignant melanoma (Table 4).
Because of the relatively small number of cases in this series, our results should not be considered abso- lutely definitive. Nevertheless, we believe that they will provide surgical pathologists with a useful adjunct to conventional microscopy in the diagnosis of adrenocor- tical carcinoma.
References
1. Silva EG, Mackay B, Samaan NA, Hickey RC: Adrenocor- tical carcinomas: an ultrastructural study of 22 cases. Ul- trastructural Pathol 1982, 3:1-7
2. Foucar E, Dehner LP: Renal cell carcinoma occurring with contralateral adrenal metastasis: A clinical and pathologic trap. Arch Surg 1979, 114:955-963
3. Tang CK, Gray GF: Adrenocortical neoplasms: Progno- sis and morphology. Urology 1975, 5:691-695
4. Hough AJ, Hollifield JW, Page DL, Hartmann WH: Prognostic factors in adrenal cortical tumors. Am J Clin Pathol 1979, 72:390-399
5. Hogan TF, Gilchrist KW, Westring DW, Citrin DL: A clinical and pathological study of adrenocortical carci- noma: Therapeutic implications. Cancer 1980, 45: 2880-2883
6. Weiss LM: Comparative histologic study of 43 metastasiz- ing and nonmetastasizing adrenocortical tumors. Am J Surg Pathol 1984, 8:163-169
7. Schteingart DE, Oberman HA, Friedman BA, Conn JW: Adrenal cortical neoplasms producing Cushing’s syn- drome: A clinicopathologic study. Cancer 1968, 22: 1005-1013
8. van Slooten H, Schaberg A, Smeenk D, Moolenaar AJ: Morphologic characteristics of benign and malignant adrenocortical tumors. Cancer 1985, 55:766-773
9. Hsu S-M, Raine L, Fanger H: Use of avidin-biotin- peroxidase complex (ABC) in immunoperoxidase tech- niques: A comparison between ABC and unlabeled an- tibody (PAP) procedures. J Histochem Cytochem 1981, 29:577-580
10. Sternberger LA, Hardy PH Jr, Cuculis JJ, Meyer HG: The unlabeled antibody enzyme method of immuno- histochemistry: preparation and properties of soluble antigen-antibody complex (horseradish peroxidase- antihorseradish peroxidase) and its use in identification of spirochetes. J Histochem Cytochem 1970, 18:315-333
11. Holthofer H, Miettinen M, Lehto V-P, Lehtonen E, Vir- tanen I: Expression of vimentin and cytokeratin types of intermediate filament proteins in developing and adult human kidneys. Lab Invest 1984, 50:522-529
12. Gown AM, Vogel AM: Monoclonal antibodies to inter- mediate filament proteins of human cells: Unique and cross-reacting antibodies. J Cell Biol 1982, 95:414-424
13. Woodcock-Mitchell J, Eichner C, Nelson WG, Sun T-T: Immunolocalization of keratin polypeptides in human epidermis using monoclonal antibodies. J Cell Biol 1982, 95:580-588
14. Lehtonen E, Lehto V-P, Paasivuo R, Virtanen I: Parietal and visceral entoderm differ in their expression of inter- mediate filaments. EMBO J 1983, 2:1023-1028
15. Fuhrman SA, Lasky LC, Limas C: Prognostic significance of morphologic parameters in renal cell carcinoma. Am J Surg Pathol 1982, 6:655-663
16. Boven E, Vermorken JB, van Slooten H, Pinedo HM: Complete response of metastasized adrenocortical car-
cinoma with o,p’-DDD: Case report and literature review. Cancer 1984, 53:26-29
17. Miettinen M, Lehto V-P, Virtanen I: Immunofluorescence microscopic evaluation of the intermediate filament ex- pression of the adrenal cortex and medulla and their tumors. Am J Pathol 1985, 118:360-366
18. Sheibani K, Tubbs RR: Enzyme immunohistochemistry: Technical aspects. Semin Diagn Pathol 1984, 1:235-250
19. Sun T-T, Shih C, Green H: Keratin cytoskeletons in epi- thelial cells of internal organs. Proc Natl Acad Sci USA 1979, 76:2813-2817
20. Nagle RB, McDaniel KM, Clark VA, Payne CM: The use of antikeratin antibodies in the diagnosis of human ne- oplasms. Am J Clin Pathol 1983, 79:458-466
21. Sloane JP, Ormerod MG: Distribution of epithelial mem- brane antigen in normal and neoplastic tissues and its value in diagnostic tumor pathology. Cancer 1981, 47:1786-1795
22. Herman CJ, Moesker O, Kant A, Vooijs GP, Ramaekers FCS: Is renal cell (Grawitz) tumor a carcinosarcoma? Evi- dence from analysis of intermediate filament types. Vir- chows Arch Cell Pathol 1983, 44:73-83
23. Tomaszewski JE, Kornstein MJ: Sensitivity and specificity of URO antibodies in diagnostic surgical pathology (Abstr). Lab Invest 1985, 52:68A
24. Holden CA, Shaw M, McKee PH, Sanderson AR, Mac- Donald DM: Loss of membrane B2 microglobulin in ec- crine porocarcinoma: Its association with the histopath- ologic and clinical criteria of malignancy. Arch Dermatol 1984, 120:732-735
25. Limas C, Lange P: Altered reactivity for A, B, and H antigens in transitional cell carcinoma of the urinary blad- der. Cancer 1980, 46:1366-1373
26. Didolkar MS, Bescher RA, Elias EG, Moore RH: Natu- ral history of adrenocortical carcinoma: a clinicopatho- logic study of 42 patients. Cancer 1981, 47:2153-2161
27. King DR, Lack EE: Adrenocortical carcinoma: A clini- cal and pathologic study of 49 cases. Cancer 1979, 44:239-244
28. Thung SN, Gerber MA, Sarno E, Popper H: Distribu- tion of five antigens in hepatocellular carcinoma. Lab In- vest 1979, 41:101-105
29. Battifora H: Recent progress in the immunohistochemis- try of solid tumors. Sem Diag Pathol 1984, 1:251-271
30. Miettinen M, Lehto V-P, Virtanen I: Presence of fibroblast-type intermediate filaments (vimentin) and ab- sence of neurofilaments in pigmented nevi and malignant melanomas. J Cutan Pathol 1983, 10:188-192
31. Springall DR, Gu J, Cocchia D, Michetti F, Levene A, Levene MM, Marangos PJ, Bloom SR, Polak JM: The value of S100 immunostaining as a diagnostic tool in hu- man malignant melanomas. Virchows Arch Pathol Anat 1983, 400:331-344
32. Said JW, Nash G, Tepper G, Banks-Schlegel S: Keratin proteins and carcinoembryonic antigen in lung carci- nomas: An immunoperoxidase study of 54 cases with ul- trastructural correlations. Hum Pathol 1983, 14:70-76
Acknowledgments
The authors are grateful to Drs. Stanley Sprei, Ft. Myers, Florida, Stephen Ewing, and Charles R. Chedister, Min- neapolis, Minnesota, and James Elrod, San Antonio, Texas for allowing us to include their cases in this study and for providing clinical follow-up information.