Original Article
Adrenocortical Cytopathology
HIRONOBU SASANO, MD, SOUICHIROU SHIZAWA, MD, AND HIROSHI NAGURA, MD
Cytopathologic smears and/or imprints of human adrenal cortex (9 cases) and its disorders were examined, including adrenocortical nodule (3 cases), adrenocortical adenoma (23 cases), carcinoma (8 cases), and renal cell carcinoma (6 cases). Immunocytochemistry directed against 38-hydroxysteroid dehydrogenase and adrenal-4-binding protein (Ad4BP), a transcription factor in steroidogenesis, was also performed. There were no cytologic differences between normal adrenal and adre- nocortical nodules. Large nuclei with prominent nucleoli were observed predominantly in adrenocortical neoplasms. Cellular atypia or pleomor- phism and the degree of cohesiveness were unreliable criteria in differ- entiating between adrenocortical adenoma and carcinoma. Mitosis and
necrotic materials were observed only in adrenocortical carcinoma. These cytologic findings were considered contributory, but not diagnos- tic when evaluating adrenocortical disorders because of marked intra- tumoral heterogeneity. There were no reliable cytologic criteria in differentiating adrenocortical and renal cell carcinoma. Immunocyto- chemistry of 38-hydroxysteroid dehydrogenase and especially Ad4BP was demonstrated to aid greatly in the differential diagnosis between these carcinomas by identifying adrenocortical parenchymal cells. (Key words: Cytology; Smear; Adrenal cortex; Immunocytochemistry) Am J Clin Pathol 1995; 104:161-166.
With the rapid and remarkable development of radio- logic techniques including computed tomography (CT), adrenal masses as small as 0.5 cm now can be clinically detected.1 Thus, an increased number of adrenal lesions have been detected in patients without clinical evidence of hormonal abnormalities. These lesions have been termed “incidentalomas,” and include adrenocortical nodules or neoplasms, ganglioneuroma and other adre- nal medullary lesions, vascular cysts, myelolipoma, fi- broma, metastatic malignant tumors, and other tumors of the upper abdomen including renal cell carcinoma.2 When evaluating patients with adrenal nodules/masses, it is important to determine whether these lesions are of adrenocortical origin or whether they are malignant.
Computed tomography-guided fine-needle aspiration can facilitate obtaining diagnostic material with minimal adverse effects. Therefore, this technique is being in- creasingly used in the diagnostic evaluation of intraab- dominal organs including the adrenal glands. The use- fulness of fine-needle aspiration in the diagnostic evaluation of adrenal masses has been demonstrated by various investigators.3-9 In addition to fine-needle aspi- ration cytology, the value of intraoperative imprints
and/or smear cytologic preparations has been re- ported,10.11 and the cytologic evaluation of surgically re- sected adrenal glands has become an important intraop- erative procedure. However, detailed studies on the cytologic appearance of the adrenal cortex and its disor- ders, including correlations with histologic findings and the application of immunocytochemistry, have not been reported.
In this study, we examined the cytologic appearance of benign and malignant adrenocortical lesions on smears and/or imprints from 43 patients to examine the poten- tial value of cytopathologic evaluation in human adre- nocortical disorders. We then compared the cytologic findings of adrenocortical carcinoma with those of renal cell carcinoma, which is one of the most important differential diagnoses of tumors in the upper abdomen. In addition, we performed immunocytochemistry for 36-hydroxysteroid dehydrogenase, which is involved in corticosteroidogenesis,12 and for adrenal-4-binding pro- tein (Ad4BP), a transcription factor of steroidogenesis13 in these cytologic smears to determine the potential value of these markers in differentiating between adreno- cortical and renal cell carcinoma.
MATERIALS AND METHODS
Preparation of Cell Smears
We studied specimens obtained from 9 patients with normal adrenal glands (those who underwent radical ne- phrectomy for renal cell carcinoma), 3 with adrenocorti- cal nodules, 23 with adrenocortical adenomas (5 Cush-
From the Department of Pathology, Tohoku University School of Medicine, Sendai, Japan.
Manuscript received January 25, 1995; revision accepted April 6, 1995.
Address reprint requests to Dr. Sasano: Department of Pathology, Tohoku University School of Medicine, 2-1 Seiryou-machi, Sendai, Ja- pan 980.
| Cytologic Features | Adrenal | Renal Cell Carcinoma | |||
|---|---|---|---|---|---|
| Nor | Nod | Ade | Car | ||
| Nuclei | |||||
| Satterd/bare | ++/+++ | ++/+++ | ++/+++ | +/++ | +/++ |
| Hyperchromasia | -/+ | -/+ | ++/+++ | ++/+++ | ++/+++ |
| Nucleoli | -/+ | -/+ | ++/+++ | ++/+++ | ++/+++ |
| Atypia | - | - | - | +/++ | +/++ |
| Mitosis | - | - | - | +/++ | +/++ |
| Necrosis | - | - | - | +/++ | +/++ |
Nor = normal adrenal; Nod = adrenocortical nodules; Ade = adrenocortical adenoma; Car = adrenocortical carcinoma; (-) = absent; (+) = scant; (++) = moderate; (+++) = marked.
ing’s adenomas, 6 nonfunctioning adrenocortical adenomas, and 12 aldosteronomas), 8 with adrenocorti- cal carcinomas, and 6 with renal cell carcinomas. Adre- nocortical carcinoma was diagnosed based on the criteria of Weiss and colleagues.14,15 Cytology smears were made by imprinting the cut surface and/or gently scraping the cut surface with a clean blade. Smears were spread on standard glass slides for morphologic examination and Denhardt- or poly-l-lysine-coated slides for immuno- staining.
Fixation and Processing
In all the cases, smears were either immediately air- dried and stained using the May-Grünwald-Giemsa method or alcohol-fixed and stained by the Papanico- laou method. In 27 cases (including 5 normal adrenals, 2 nodules, 10 adenomas, 5 adrenocortical carcinomas, and 5 renal cell carcinomas), the smears were immediately placed in 4% or 8% paraformaldehyde in a phosphate- buffered saline (PBS), (pH 7.4), for 30 to 60 minutes at 4 ℃, and then transferred to cold absolute methanol at -20 ℃ for 5 minutes.16 These smears were stored at -20 ℃ in storage medium containing 50% glycerol.16,17 In the Papanicolaou-stained slides of these 27 cases, coverslips were removed and the slides were decolorized with xy- lene and ethanol for subsequent immunostaining.
Immunocytochemistry
The smear slides described above were washed five times with cold PBS for 5 minutes before the immuno- staining. The slides were then immersed in 0.01 mol/L PBS with 0.3% hydrogen peroxide for 15 minutes to block endogenous peroxidase. They were washed three times with 0.01 mol/L PBS for 5 minutes each, and then treated with 1% normal goat serum for 30 minutes at room temperature. After washing, the slides were incu- bated with the primary antibodies for 18 hours at 4 ℃ in
a moisture chamber. Details of immunostaining, as well as characteristics of the primary antibodies, have been previously described.12,13 The optimal dilution for im- munocytochemistry was 1:500 for 38-hydroxysteroid dehydrogenase and 1:300 for Ad4BP. The Histofine method (Nichirei, Tokyo, Japan) was employed for im- munostaining. After washing in 0.01 mol/L PBS, the slides were incubated for 30 minutes each at room tem- perature with biotinylated anti-rabbit immunoglobulin and peroxidase-conjugated streptavidin. The slides were washed with 0.01 mol/L PBS between each incubation. A final wash was followed by immersion of the slides for 5 to 10 minutes in a solution containing 0.06 mmol/L 3.3’-diaminobenzidine tetrahydrochloride and 2 mmol/ L hydrogen peroxide in 0.05% Tris-HCI buffered at pH 7.6. Specific staining was identified by the presence of a brown reaction product. The sections were mounted with a glycerol-gelatin water-soluble medium. Smears fixed in 4% paraformaldehyde were counterstained with 1% methyl green. Control sections were incubated with the antibody preincubated with an excessive amount of purified Ad4BP for 18 hours at 4 ℃ and with normal rabbit serum instead of primary antibodies.
RESULTS
Cytologic Features of Adrenocortical Disorders
Results were summarized in Table 1.
Normal Adrenal and Adrenocortical Nodule
Adrenocortical cells were characteristically polygonal with cytoplasm, but the majority of these cells demon- strated bare nuclei. The nuclei appeared round or oval without membrane irregularities with a fine chromatin pattern. Nucleoli were not prominent. Cytologic findings of the adrenocortical nodules were indistinguishable from those of the normal adrenal cortex.
Adrenocortical Adenoma
The smears from adrenocortical adenomas varied widely between cases, but the majority of tumor cells demonstrated round or oval nuclei without membrane irregularities (Fig. 1). A number of tumor cells had foamy or compact cytoplasm. Smears occasionally con- tained cells with large nuclei and multiple prominent nucleoli. (Fig. 1) Mitotic figures were rare. Chromatin patterns varied from fine to coarse patterns. Some tumor cells demonstrated irregular nuclear morphology with hyperchromasia. Cytologic findings did not differ among aldosteronomas, Cushing’s adenomas, and nonfunction- ing adrenocortical adenomas.
Adrenocortical Carcinoma
Cytologic findings of adrenocortical carcinomas var- ied widely, not only among the cases examined, but also between different areas of the same smears. In some
smears, carcinoma cells formed sheets and had relatively regular round nuclei with fine chromatin patterns (Fig. 2). Even in these smears of cytologically bland polygonal cells, scattered mitoses were detected (Fig. 2). Some smears demonstrated atypical cytologic features includ- ing an increased nuclear to cytoplasmic ratio, bizarre and irregular nuclear morphology with hyperchromasia, and prominent nucleoli (Fig. 3). Bare nuclei were also ob- served in the smears. The nature of the cytoplasm varied among carcinoma cells, from abundant to none in amount and from clear or foamy to compact with lipo- fuscin granules in appearance. Necrotic material was also observed in some smears.
Renal Cell Carcinoma
Cytologic preparations of renal cell carcinoma of low grade demonstrated cohesive sheets of carcinoma cells (Fig. 4). The nuclear to cytoplasmic ratio did not mark- edly increase and nuclei were relatively round and regu-
| 38HSD | Ad4BP | |||
|---|---|---|---|---|
| PAP | PFA | PAP | PFA | |
| Nor | 5/5 | 5/5 | 0/5 | 5/5 |
| Nod | 2/2 | 2/2 | 0/2 | 2/2 |
| Ade | 10/10 | 10/10 | 0/10 | 10/10 |
| AC | 5/5 | 5/5 | 0/5 | 5/5 |
| RC | 2/5 | 2/5 | 0/5 | 0/5 |
PAP = slides decolorized from Papanicolaou-stained specimens; PFA = slides fixed in para- formaldehyde; Nor = normal adrenal; Nod = adrenocortical nodule; Ade = adrenocortical adenoma; AC = adrenocortical carcinoma; RC = renal cell carcinoma.
* The number represents the positive cases.
lar with hyperchromasia (Fig. 4). The carcinoma cells oc- casionally demonstrated foamy or prominently vacuolated cytoplasm (Fig. 4). Renal cell carcinomas of high grade contained single or small clusters of pleomor- phic or bizarre tumor cells. Foci of mitoses and necrosis were also observed in these smears and/or imprints.
Immunocytochemistry
Results of immunocytochemistry are summarized in Table 2.
Immunocytochemistry of 38-Hydroxysteroid Dehydrogenase
Both the paraformaldehyde-fixed slides and the slides decolorized from Papanicolaou stain of specimens from normal adrenal cortex, adrenocortical nodules, adeno- mas and carcinoma demonstrated marked immunoreac- tivity for 30-hydroxysteroid dehydrogenase in their cyto- plasm (Fig. 5). Heterogeneity of immunoreactivity was observed in these preparations, especially those from adrenocortical adenomas and carcinomas, but all cases demonstrated enzyme immunoreactivity in both para- formaldehyde-fixed slides and slides decolorized from the Papanicolaou stain. However, positive cells were fewer and relative immunointensity was weaker in de- colorized slides. Two cases of renal cell carcinoma were weakly positive for 36-hydroxysteroid dehydrogenase in the cytoplasm of the tumor cells in both preparations.
Immunocytochemistry of Ad4BP
On paraformaldehyde-fixed slides, Ad4BP immuno- reactivity was observed only in the nuclei of parenchy- mal cells from normal adrenal gland, adrenocortical nodules, adrenocortical adenomas, and carcinoma (Fig. 6), but not in any cells from renal cell carcinoma. Almost all adrenocortical parenchymal cells were positive for
Ad4BP although the degree of immunointensity varied. Slides decolorized from the Papanicolaou stain did not demonstrate any Ad4BP immunoreactivity in speci- mens of both normal and pathologic adrenals that were positive on paraformaldehyde-fixed specimens.
Comparison Among Normal Adrenal, Adrenocortical Nodules, Adenoma and Carcinoma
There were no cytologic differences between tissue from normal adrenal cortex and adrenocortical nodules. Large nuclei with prominent nucleoli were observed pre- dominantly in the smears obtained from adrenocortical neoplasms. Variations in the size of the nuclei were much more pronounced in adrenocortical neoplasms than normal adrenal cortex or adrenocortical nodules. Mitotic figures and necrotic materials were observed only in adrenocortical carcinoma, but bizarre and atypi- cal nuclei were observed in both carcinomas and adeno- mas. No differences in the patterns or intensity of immu- noreactivity for both 36-hydroxysteroid dehydrogenase and Ad4BP were observed between adrenocortical ade- noma and carcinoma.
Comparison Between Renal Cell Carcinoma and Adrenocortical Carcinoma
In tumor cells with similar degrees of atypia (ie, equivalent nuclear to cytoplasmic ratios), renal cell carcinoma cells gen- erally exhibited more prominent hyperchromasia and nuclear irregularity than adrenocortical carcinoma cells. However, in Papanicolaou- and Giemsa-stained slides, there were no cytologic features that could differentiate renal cell carcinoma from adrenocortical carcinoma in the cases ex- amined. As described above, 36-hydroxysteroid dehydroge- nase and Ad4BP immunoreactivities were demonstrated in all of the cases of adrenocortical carcinoma. However, two cases of renal cell carcinoma were positive for 36-hydroxy- steroid dehydrogenase. There was marked intratumoral variability in the expression of the enzyme in adrenocortical carcinoma. Nuclear Ad4BP immunoreactivity was detected in all cases of adrenocortical carcinoma but in no renal cell carcinoma cases. Immunostaining could only be demon- strated in the paraformaldehyde-fixed specimens (Table 2).
DISCUSSION
With the advent of sophisticated radiologic techniques and the ability to accurately localize abdominal struc- tures and masses, the number of cytology specimens of adrenocortical disorders submitted to diagnostic pathol- ogy laboratories has increased. Most adrenal biopsies are performed by fine-needle aspiration. Therefore, it is an- ticipated that the need for cytologic diagnosis of lesions
in the adrenal glands will increase further.3 However, it is also true that cytologic features of the human adrenal cortex and its disorders have not been adequately de- scribed. Surgical pathologic diagnosis of adrenocortical disorders has been associated with a number of diagnos- tic difficulties. These difficulties include differentiation between adrenocortical nodules and neoplasms,18 adre- nocortical adenoma and carcinoma,14,15,19 and adreno- cortical carcinoma and other malignancies of the upper abdomen.18,20 Therefore, in this study, we examined cy- tologic features of normal human adrenal cortex and its lesions and renal cell carcinoma using immunostaining to determine whether cytologic examination can be of any value in the differential diagnoses described pre- viously.
Evaluation by Giemsa and Papanicolaou stains dem- onstrated the presence of large nuclei that may be of help in differentiating between benign adrenocortical nodules and neoplasms. Mitotic figures and necrosis are also sug- gestive of an adrenocortical malignancy. Katz and col- leagues have reported that the presence of necrosis, mi- tosis, and a loss of cohesion favor a malignant diagnosis based on the results of fine-needle aspiration cytology performed on three patients.3 However, we observed an absence of cohesion in cases of both adenoma and carci- noma. It is important to note that these cytologic features were not necessarily present in all of the cytologic prepa- rations obtained from adenoma and/or carcinoma. It is also important to recognize that nuclear and cellular atypia, which is very important in the diagnosis of a ma- lignancy in other tissues, is noncontributory in cases of adrenocortical disorders. Therefore, identification of these cytologic features is believed to contribute to the diagnosis when the whole adrenal specimen is available, but not in cases of fine-needle aspiration biopsy because of the intratumoral variation in the cytologic features. These difficulties in the cytologic diagnosis of human adrenocortical diseases exist, especially regarding malig- nancy, perhaps because there is no single set of histopath- ologic criteria and the diagnosis is generally based on the evaluation of multiple factors.14,15.19 Therefore, patholo- gists should not attempt to differentiate adrenocortical adenoma from carcinoma based only on cytologic find- ings.
Occasionally, the differential diagnosis between adre- nocortical carcinoma and other tumors involving the adrenal or upper abdomen (especially clear cell renal cell carcinoma), is very difficult.2º Nguen has reported that adrenocortical carcinoma tends to demonstrate more pleomorphism than renal cell carcinoma in cytology preparations,4 but we could not confirm this conclusion in the present study. We could not find any reliable cyto-
logic criteria for differentiating between adrenocortical and renal cell carcinoma based solely on Giemsa- and Papanicolaou-stained preparations.
Immunohistochemistry has provided important infor- mation in differentiating between various neoplasms. Immunostaining techniques themselves are potentially easily adaptable for use with a variety of cytologic prepa- rations.21 In this study, we employed antibodies against 36-hydroxysteroid dehydrogenase and Ad4BP for the immunocytochemical differentiation between adreno- cortical and renal cell carcinoma. The former enzyme is involved in progesterone biosynthesis12 and is immuno- localized in adrenocortical carcinoma.22 Ad4BP is a zinc finger DNA-binding protein and a transcription factor that regulates the expression of steroidogenic enzyme genes13,25 and has been expressed primarily in steroido- genic tissues, including adrenocortical parenchymal cells.13 36-hydroxysteroid dehydrogenase immunoreac- tivity was observed in all of the specimens from the adre- nal cortex. Denaturation of the enzyme immunoreactiv- ity appeared to occur through the processes of staining or decolorization but immunoreactivity was detected in Papanicolaou-stained specimens following decoloriza- tion. However, marked intratumoral variability of en- zyme expression among adrenocortical carcinoma cells was observed, as has been previously reported.22 Two cases of renal cell carcinoma also showed immunoreac- tivity, which is consistent with peripheral steroid metab- olism and the corresponding expression of this enzyme in nonsteroidogenic tissues and their neoplasms.24 Therefore, caution should be exercised in applying 38- hydroxysteroid dehydrogenase immunocytochemistry for differentiating between renal cell carcinoma and adrenocortical carcinoma. However, Ad4BP immunore- activity was exclusively observed in adrenocortical pa- renchymal cells. Almost all adrenocortical carcinoma cells, regardless of degree of differentiation, were immu- nocytochemically positive for Ad4BP, whereas none of renal cell carcinoma cells were immunoreactive. There- fore, Ad4BP appears to be useful in diagnosing adreno- cortical carcinoma, but its practical value may be limited since immunostaining cannot be done in routinely pro- cessed cytologic preparations. This is probably due to the denaturation of the nuclear antigen during Papanicolaou preparation and/or subsequent decolorization. How- ever, Ad4BP immunocytochemistry can contribute greatly to the differential cytologic diagnosis between adrenocortical and renal cell carcinoma when specimens that are appropriately processed for immunostaining are available.
In summary, it is difficult to differentiate between adrenocortical disorders, especially between adenoma
Downloaded from http://ajcp.oxfordjournals.org/ by guest on June 5, 2016
and carcinoma, based on cytopathologic findings alone. Differentiating between adrenocortical carcinoma and renal cell carcinoma is also difficult, but immunostain- ing of 30-hydroxysteroid dehydrogenase and/or Ad4BP, especially the latter, has the potential to contribute to the differential diagnosis.
REFERENCES
1. Huiras CM, Pehling GB, Caplan RH. Adrenal insufficiency after operative removal of apparently non-functioning adrenal ade- nomas. JAMA 1989;261:894-898.
2. Sasano H, Suzuki T, Sano T, et al. Adrenocortical oncocytoma: A true nonfunctioning adrenocortical tumor. Am J Surg Pathol 1991; 15:949-956.
3. Katz RL, Pathel S, Mackay B, Zornoza J. Fine-needle aspiration cytology of the adrenal gland. Acta Cytol 1984; 28:269-282.
4. Nguyen GK. Percutaneous fine-needle aspiration biopsy cytology of the kidney and adrenal. Pathol Annu 1987; 22:163-191.
5. Min K-W, Song J, Boesenberg M, Acebey J. Adrenal cortical nod- ule mimicking small round cell malignancy on fine needle aspi- ration. Acta Cyto/ 1988; 32:543-546.
6. Cochand-Priollet B, Jacquenod P, Warnet A, et al. Adrenal cortical carcinoma: A case diagnosed by fine-needle aspiration cytology. Acta Cyto/ 1988; 32:128-130.
7. Kenady DE, Sloan DA, Schwartz RW. Diagnosis and treatment of thyroid, adrenal, and thymic tumors. Curr Opin Oncol 1992;4: 89-98.
8. Ettinghausen SE, Burt ME. Prospective evaluation of unilateral adrenal masses in patients with operable non-small-cell lung cancer. J Clin Oncol 1991;9:1462-1466.
9. Wadin GE, Nance KV, Silverman JF. Fine-needle aspiration cy- tology of the adrenal grand: Fifty biopsies in 48 patients. Arch Pathol Lab Med 1992; 116:841-846.
10. Nochomavitz L, Sidaway M, Silverberg SG, et al. Intraoperative consultation: A guide to smears, imprints, and frozen sections. Chicago: ASCP Press, 1989.
11. Sasano H, Geelhoed GW, Silverberg SG. Intraoperative cytologic evaluation of lipid in the diagnosis of parathyroid adenoma. Am J Surg Pathol 1988; 12:282-286.
12. Sasano H, Mason JI, Sasano N, Nagura H. Immunolocalization of 36-hydroxysteroid dehydrogenase in human adrenal cortex and in its disorders. Endocrine Pathology 1990; 1:94-101.
13. Morohashi K, Iida Hm, Nomura M, et al. Functional difference between Ad4BP and ELP and their distributions in steroido- genic tissues. Mol Endocrinol 1994;644:643-653.
14. Weiss LM. Comparative histologic study of 43 metastasizing and nonmetastasizing adrenocortical tumors. Am J Surg Pathol 1984;8:163-169.
15. Weiss LM, Medeiros LJ, Vickery AL. Pathologic features of prog- nostic significance in adrenocortical carcinoma. Am J Surg Pa- thol 1989; 13:202-206.
16. Sasano H, Miyazaki S, Nishihira T, et al. The proliferative cell frac- tion in cytology specimens: A study of human esophageal carci- noma. Am J Clin Pathol 1992;98:161-166.
17. Deligeorgi-Politi H, Bartziota EV, Mavrogeorgis A, Sekeris EE. Determination of estrogen receptors in cytologic imprints of breast cancers using immunocytochemical methods. Acta Cytol 1990; 34:881-884.
18. Sasano N, Sasano H. The adrenal cortex. In: Kovacs K, Asa SL, eds. Functional Endocrine Pathology. London: Blackwell, 1990, pp 548-584.
19. Hough AJ, Hollifield JW, Page DL, Hartman WH. Prognostic fac- tors in adrenocortical tumors: A mathematical analysis of clini- cal and morphologie data. Am J Clin Pathol 1979; 72:390-399.
20. Lack. EE, Travis WD, Oertel JE. Adrenal cortical neoplasms. In: Lack EE, ed. Pathology of the Adrenal Glands. New York: Churchill Livingstone, 1990, pp 115-172.
21. Redard M, Vassilakos P, Weintraub J. A simple method for estro- gen receptor antigen preservation in cytologic specimens con- taining breast carcinoma cells. Diagn Cytopathol 1989;5:188- 193.
22. Sasano H, Suzuki T, Nagura H, Nishikawa T. Steroidogenesis in human adrenocortical carcinoma: Biochemical activities, im- munohistochemistry, and in situ hybridization of steroidogenic enzymes and histopathologic study in nine cases. Hum Pathol 1993; 24:397-404.
23. Morohashi K, Honda S, Inomata Y, et al. A common transacting factor, Ad4-binding protein, to the promoters of steroidogenic P-450s. J Biol Chem 1992; 267:17913-17919.
24. Sasano H, Nagura H, Harada N, Goukon Y, Kimura M. Immuno- localization of aromatase and other steroidogenic enzymes in human breast disorders. Hum Pathol 1994; 25:530-535.
Downloaded from http://ajcp.oxfordjournals.org/ by guest on June 5, 2016