Fine-Needle Aspiration Cytology of a Case of Oncocytic Adrenocortical Carcinoma
Savitri Krishnamurthy, M.D.,1* Nelson G. Ordóñez, M.D.,1 Thomas O. Shelton, M.D.,2 Alberto G. Ayala, M.D.,1 and Nour Sneige, M.D.1
We report on the results of fine-needle aspiration cytology of a case of oncocytic adrenocortical carcinoma in a 39-yr-old man. The tumor invaded the inferior vena cava and extended up to the right atrium. Aspirate smears were very cellular and showed a monomorphic population of large polyhedral cells with abundant granular cytoplasm, predominantly distributed singly. Mitotic ac- tivity was inconspicuous, and there was no necrosis. Immunohis- tochemically, the tumor cells were positive for vimentin, cytoker- atin, and p53, and negative for synaptophysin, chromogranin, inhibin, and S-100. Ultrastructurally, the cytoplasm of the tumor cells was packed with mitochondria. The patient underwent left radical nephrectomy as well as a combined cardiopulmonary bypass, with atriotomy and resection of the tumor from the right atrium and inferior vena cava. Three months of postoperative fol- low-up were uneventful. Diagn. Cytopathol. 2000;22:299-303. @ 2000 Wiley-Liss, Inc.
Key Words: adrenal gland; oncocytic tumor; aspiration cytology; adrenocortical tumor
Oncocytic neoplasms rarely occur in the adrenal gland. By definition, these neoplasms are comprised exclusively of oncocytes, which by light microscopy are characterized by abundant granular eosinophilic cytoplasm.1 Ultrastructural examination demonstrates that the granularity seen in the cytoplasm is caused by a massive accumulation of mito- chondria, almost to the exclusion of all other organelles.2 Only 17 examples of oncocytic adrenocortical neoplasms have been reported so far in the English-language litera- ture.3-10 Fourteen were diagnosed as oncocytomas, two as oncocytic tumors of uncertain malignant potential, and one as oncocytic adrenocortical carcinoma. We herein report the
second example of an oncocytic adrenocortical carcinoma originating in the adrenal gland. The diagnosis was initially made by fine-needle aspiration cytology and was subse- quently confirmed by histopathologic examination of the resected tumor.
Case Report
A 39-yr-old black man with ascites, bilateral lower extrem- ity edema, and an abdominal mass was referred to the University of Texas M.D. Anderson Cancer Center for further evaluation. A CT scan of the abdomen and arterio- graphic examination revealed a large mass near the left kidney that invaded the inferior vena cava and extended up to the level of the right atrium. The ascites and bilateral leg edema were attributed to Budd-Chiari syndrome, which developed secondary to the occlusion of the hepatic venous outflow system, by a tumor thrombus. A thorough workup showed no metastatic disease. The patient was treated with diuretics to control his ascites. He also underwent two embolization procedures, which resulted in reduction in tumor size and markedly improved hepatic function. The patient subsequently underwent a radical nephrectomy on the left side as well as a combined cardiopulmonary bypass with atriotomy and resection of the tumor from the atrium and inferior vena cava. He did well postoperatively, except for a brief bout of pneumonia caused by Klebsiella pneu- moniae, which was treated with antibiotics. At the time of writing, the patient had received 3 months of uneventful follow-up care.
Materials and Methods
Fine-needle aspiration was performed with a 22-gauge nee- dle under CT guidance. Direct smears were prepared for Papanicolaou and Diff-Quik staining, and additional mate- rial was procured in RPMI for preparation of a cell block and in gluteraldehyde for ultrastructural examination.
For electron microscopy, portions of the aspirate material that were fixed in 2% glutaraldehyde were postfixed in 1%
1Department of Pathology, University of Texas M.D. Anderson Cancer Center, Houston, Texas
2 Department of Diagnostic Radiology, University of Texas M.D. Ander- son Cancer Center, Houston, Texas
*Correspondence to: Savitri Krishnamurthy, M.D., Division of Cytol- ogy, Box 53, University of Texas M.D. Anderson Cancer Center, 1515 Holcombe Blvd., Houston, TX 77030.
E-mail: skrishna@notes.mdacc.tmc.edu
Received 2 August 1999; Accepted 1 November 1999
A
C
osmium tetroxide and embedded in Epon epoxy resin. Semithin sections of the specimens were stained with methenamine blue and evaluated by light microscopy. After sections were selected from the Epon-embedded blocks, ultrathin sections were cut and stained with uranyl acetate and lead citrate.
Immunohistochemical studies were performed on forma- lin-fixed, paraffin-embedded tissue sections using the avi- din-biotin-peroxidase complex (ABC) method in a Dako Auto Stainer (Carpinteria, CA).11 The primary antibodies used were a “cocktail” of three anticytokeratin mouse monoclonal antibodies (Mabs): AE1 and AE3 (Boehringer- Mannheim, Indianapolis, IN; 1:500 dilution) and CAM 5.2 (Becton-Dickinson, Mountain View, CA; 1:50 dilution) that recognize a wide range of high and low molecular weight cytokeratin peptides, V9 antivimentin MAb (Dako; 1:500), LK2H10 antichromogranin A MAb (Boehringer-Mann- heim; 1:2,000), SY38 antisynaptophysin MAb, R1 anti- inhibin MAb (Serotec, Inc., Raleigh, NC; 1:60), MIB-1 anti-Ki-67 MAb (Immunotech, Westbrook, ME; 1:120), D07 anti-p53 MAb (Dako; 1:75), and a polyclonal antibody raised in rabbit to S-100 protein (Dako; 1:1,200). The im- munoperoxidase staining was done using the LSAB2 per- oxidase kit (Dako). To enhance immunostaining, a heat epitope retrieval procedure was performed using a steamer,
B
as previously described.11 The immunostaining was devel- oped using either 3-amino-9-ethylcarbazole or 3,3’-diamino- benzidine tetrahydrochloride as the chromogen. Slides were counterstained with Mayer’s hematoxylin. To evaluate the specificity of the antibodies, known positive and negative tissues were used as controls.
Results
Cytology Findings
The aspirate smears were markedly cellular and showed a monomorphic population of large cells that were arranged both singly and clustered around blood vessels. The cells were polyhedral in shape and had abundant granular cyto- plasm and well-defined cell borders that were better appre- ciated in the Diff-Quik-stained preparations than in the smears using Papanicolaou stain (Fig. 1A,B). There was mild to moderate nuclear pleomorphism. Many of the tumor cells had an eccentrically located nucleus, and some of them contained a centrally located prominent nucleolus. The nu- clear chromatin was fine and evenly distributed throughout the nucleus. Mitotic figures were rare, and there was no evidence of necrosis. The H&E-stained cell block sections revealed tumor cells arranged in a predominantly trabecular array (Fig. 1C). Similar to the smears, the neoplastic cells in the cell block sections also exhibited abundant granular
eosinophilic cytoplasm, and some of them had a prominent central nucleolus. Mitotic activity was inconspicuous. There was no evidence of tumor necrosis. Immunohistochemical staining of the cell block sections demonstrated that the tumor cells were positive for vimentin and cytokeratin and negative for synaptophysin, chromogranin-A, inhibin, and S-100 protein. Ultrastructural examination showed the cy- toplasm of the tumor cells to be packed with mitochondria (Fig. 2). Only occasional mitochondria contained tubular cristae. Based on the cytological, immunohistochemical, and ultrastructural features of the tumor, the diagnosis of oncocytic adrenocortical neoplasm was made.
Gross Findings
The left radical nephrectomy specimen showed a rather well-circumscribed and partially encapsulated tumor mea- suring 14.0 × 12.0 × 11.0 cm, located above the upper pole of the left kidney in the region of the adrenal gland. No evidence of residual adrenal gland was seen. The tumor was separate from the kidney and invaded the left adrenal vein and the left renal vein. The cut surface of the tumor was variegated, with tan to pink homogeneous areas that alter- nated with extensive large yellowish areas of necrosis and hemorrhage.
Histopathologic Examination
The tumor was partially encapsulated, and the neoplastic cells were arranged in a solid and trabecular pattern. There was mild to moderate nuclear pleomorphism. The most striking feature was the abundant brightly eosinophilic gran- ular cytoplasm of the tumor cells. The nuclei were vesicular, and some contained a prominent, centrally placed nucleolus. Mitotic activity was low, with less than one mitosis per 10 high-power fields. Large areas of necrosis were noted in the tumor. Tumor invasion into the left adrenal vein and left renal vein were confirmed. Immunohistochemical staining and ultrastructural examination were repeated, and the find-
ings were essentially similar to those observed initially at the time of fine-needle aspiration biopsy. In addition, the tumor was immunostained for Ki-67 and p53. The tumor proliferative fraction (TPF), as ascertained by the number of Ki-67-positive nuclei per 1,000 cells, was 84.12 The tumor cells were positive for p53.
Discussion
Oncocytic neoplasms occur most commonly in organs such as the kidney, salivary glands, and thyroid gland.1 However, they have been described in a variety of other sites, includ- ing the parathyroid glands, pituitary gland, larynx, and lung. Oncocytic neoplasms originating in the adrenal gland are rare.13 There have been only 17 oncocytic adrenocortical neoplasms reported in the English-language literature.3-10 The majority of those tumors (14 of the 17) were well- circumscribed, ranging in size from 3.0-15.0 cm. There was no capsular or vascular invasion and no necrosis; the tumors exhibited low mitotic activity. None of these lesions re- curred or metastasized during a follow-up period ranging from 8-54 mo. Because of this, these tumors were consid- ered to be benign and were diagnosed as oncocytomas. In the largest series of oncocytic adrenocortical tumors that were studied by Lin et al.,3 2 of 7 cases exhibited foci of necrosis and increased mitotic activity. There was, however, no evidence of metastasis or recurrence for a limited fol- low-up of 19 and 12 mo. Since the significance of increased mitotic activity and the presence of necrosis were not ap- parent in predicting the biologic behavior of these lesions, they were designated oncocytic adrenocortical tumors of uncertain malignant potential. In the only case of a malig- nant oncocytic adrenocortical carcinoma that has been re- ported in the literature, the disease locally invaded the liver and inferior vena cava.4 This tumor was shown to have mild nuclear pleomorphism. However, no mention was made of the degree of mitotic activity or the presence of necrosis. Our case of oncocytic adrenocortical carcinoma exhibited mild to moderate nuclear pleomorphism and low mitotic activity. Since necrosis was not noted in any of the smears obtained by aspiration from different areas of the tumor prior to embolization, it is possible that the extensive ne- crosis noted in the resected tumor may have been the result of embolization. Even though our case exhibited only mild to moderate pleomorphism without significant mitotic ac- tivity or necrosis, it was undoubtedly malignant in view of the extensive vascular invasion into the inferior vena cava and up to the right atrium. Thus, we believe that morpho- logical features cannot be relied upon to predict the biologic behavior of oncocytic adrenocortical tumors. The presence of capsular invasion, vascular invasion, or both, or of me- tastasis, is considered to be the only criterion of malignancy in an oncocytic adrenocortical carcinoma.
The appearance of oncocytic cells with abundant granular eosinophilic cytoplasm is distinctive enough to be recog-
| SY | CK | VIM | EMA | Inhibin | |
|---|---|---|---|---|---|
| Adrenal oncocytic neoplasm | + | + | + | - | - |
| Renal oncocytoma | - | + | - | + | - |
| Chromophobe renal-cell carcinoma | - | + | - | + | - |
| Renal-cell carcinoma, conventional with a predominance of granular cells | - | + | + | + | - |
| Adrenocortical carcinoma, conventional | + | + | + | - | + |
| Pheochromocytoma | + | - | + | - | - |
ªSY, synaptophysin; CK, cytokeratin; VIM, vimentin; EMA, epithelial membrane antigen.
nized easily in fine-needle aspirate smears. The immunohis- tochemical profile of oncocytic adrenocortical tumors was established by Lin et al.3 All seven oncocytic tumors in that study showed positive staining for cytokeratins, vimentin, neuron-specific enolase, and antimitochondrial antibody; 70% of the cases expressed synaptophysin, but all of the tumor samples were negative for chromogranin and S-100 protein. A similar immunohistochemical pattern was seen in our case: the tumor was positive for cytokeratin and vimen- tin, and negative for chromogranin and S-100 protein. How- ever, our case was negative for synaptophysin, unlike the majority of the oncocytic tumors studied by Lin et al.3 Immunostaining for inhibin, which was recently shown to stain the majority of adrenocortical tumors, was also nega- tive in our case. 14
The differential diagnoses of oncocytic cells in fine- needle aspiration smears from a tumor located in the region of the upper pole of the kidney include adrenal oncocytic tumors, conventional adrenocortical carcinoma, the eosino- philic variant of pheochromocytoma, renal oncocytomas, the eosinophilic-cell variant of chromophobe renal-cell car- cinoma, and conventional renal-cell carcinoma with pre- dominant granular cells. Ancillary studies such as immuno- histochemical staining and ultrastructural examination can be very helpful in arriving at the correct diagnosis. A radiological examination, especially an arteriogram, can assist in localizing the tumor either to the adrenal gland or to the kidney. Although renal oncocytic neoplasms can be morphologically very similar to adrenal oncocytic neo- plasms, radiological correlation and pertinent immunohis- tochemical staining can help in the distinction. Renal onco- cytomas are usually positive only for cytokeratin and epithelial membrane antigen (EMA), unlike adrenal onco- cytic tumors, which are positive for cytokeratin, vimentin, and often for synaptophysin.15 Chromophobic renal-cell carcinomas stain positively with Hale’s colloidal iron stain and for EMA and cytokeratin, but are negative for vimentin and neuroendocrine markers.15 Conventional renal-cell car- cinomas with predominant granular cells are positive for EMA, cytokeratin, and vimentin, but negative for any of the neuroendocrine markers. Conventional adrenocortical tu- mors can show small foci of oncocytic cells, and immuno- histochemically, they may be very similar to oncocytic adrenocortical tumors. However, one would expect onco-
cytic tumor cells to be admixed with nononcocytic neoplas- tic cells. Pheochromocytomas are negative for cytokeratin and positive for vimentin, S-100 protein, and neuroendo- crine markers, including not only NSE and synaptophysin but also chromogranin.13 Table I is a synopsis of the im- munohistochemical findings of the above-discussed differ- ential diagnosis. Ultrastructural examination can also be helpful either in confirming the diagnosis of oncocytic tu- mors or in excluding other tumors with which they may be confused in light microscopy studies.
With the availability of advanced radiological techniques for the localization of the tumor and other ancillary proce- dures such as immunohistochemistry and electron micros- copy, a definite diagnosis of adrenal oncocytic tumor can be made on material obtained by fine-needle aspiration biopsy. However, the morphologic features and ancillary studies of fine-needle aspiration cytology material do not help in pre- dicting the biological behavior of these tumors. The signif- icance of mitotic activity or necrosis in these tumors is not clear, and the fact that clearly malignant tumors, such as the one we are describing, did not show high mitotic activity or necrosis is in itself an indication that cytologic features cannot be relied upon to distinguish a benign from a ma- lignant tumor. Attempts have been made to use the tumor proliferative fraction (TPF) to predict the behavior of these tumors. The TPF, as revealed by MIB-1 labeling, was found to be increased to levels of 85.0 and 95.0 in tumors that were designated as oncocytic tumors of uncertain malignant potential in the series of Lin et al.,3 unlike the tumors that were considered benign, which had TPFs of 1.0-65.0. The TPF of our case was 84, which is similar to those of the oncocytic adrenocortical tumors of uncertain malignant po- tential. Therefore, TPF alone cannot be reliably used to predict biologic behavior. Our case of malignant oncocytic adrenocortical tumor was immunohistochemically positive for wild-type p53, in contrast to the benign and borderline oncocytic adrenocortical tumors in the study of Lin et al.3 This suggests that the p53 tumor suppressor gene may play an important role in the biology of these tumors, and that increased expression of p53 may be predictive of malignant behavior. However, one cannot heavily depend upon p53 immunoexpression to be a definitive indicator of malignant behavior based on the expression pattern of a single case of malignant adrenocortical tumor.
Because capsular invasion and vascular invasion, features that cannot be ascertained in cytology studies, are by far the most definitive indicators of malignant behavior, it is pru- dent to designate these tumors as oncocytic adrenocortical tumors on cytology. Molecular markers may be anticipated in the future to help in predicting the biological behavior of these tumors.
References
1. Chang A, Harawi SJ. Oncocytes, oncocytosis and oncocytic tumors. Pathol Annu 1992;27:263-304.
2. Ghadially FN. Diagnostic electron microscopy of tumours, 2rd ed. London: Butterworths, 1985.
3. Lin BTY, Bonsib SM, Mierau GW, Weiss LM, Medieros LJ . Onco- cytic adrenocortical neoplasms. A report of seven cases and review of literature. Am J Surg Pathol 1998;22:603-614.
4. El-Naggar AK, Evans DB, Mackay B. Oncocytic adrenocortical car- cinoma. Ultrastruct Pathol 1991;15:549-556.
5. Begin LR. Adrenocortical oncocytoma: case report with immunocyto- chemical and ultrastructural study. Virchows Arch A Pathol Anat His- topathol 1992;421:533-537.
6. Gandras EJ, Schwartz LH, Panicek DM, Levi G. Adrenocortical on- cocytoma: CT and MRI findings. J Comput Assist Tomogr 1996;20: 407-409.
FNAC OF ONCOCYTIC ADRENOCORTICAL CARCINOMA
7. Kakimoto S, Yushita Y, Sanefuji T, Kondo A, Fujishima N, Kishikawa M, Matsumoto K. Nonhormonal adrenocortical adenoma with onco- cytoma like appearances. Hinyokika Kiyo [Acta Urol Jpn] 1986;32: 757-763.
8. Sasano H, Suzuki T, Sano T, Kameya T, Sasano N, Nagura H. Adrenocortical oncocytoma. A true nonfunctioning adrenocortical tu- mor. Am J Surg Pathol 1991;15:949-956.
9. Waters PR, Haselhuhn GD, Gunning WT, Phillips ER, Selman SH. Adrenocortical oncocytoma: two case reports and review of literature. Urology 1997;49:624-628.
10. Erlandson RA, Reuter VE. Oncocytic adrenocortical adenoma. Ultra- struct Pathol 1991;15:539-547.
11. Miller K, Auld J, Jessup E, Rhodes A, Ashton-Key M. Antigen unmasking in formalin-fixed routinely processed paraffin wax-embed- ded sections by pressure cooking: a comparison with microwave oven heating and traditional methods. Adv Anat Pathol 1995;2:60-64.
12. Goldblum JR, Shannon R, Kaldjian EP, Thiny M, Davenport R, Thomp- son N, Lloyd RV. Immunohistochemical assessment of proliferative ac- tivity in adrenocortical neoplasms. Mod Pathol 1993;6:663-668.
13. Lack EE. Tumors of the adrenal gland and extra-adrenal paraganglia. In: Rosai J, ed. Atlas of tumor pathology. Series 3, fascicle 19. Washington, DC: Armed Forces Institute of Pathology; 1995. p 118-119.
14. Chivite A, Matias-Giuiu X, Pons C, Algaba F, Prat T. Inhibin A expres- sion in adrenal neoplasms. Appl Immunohistochem 1998;6:42-49.
15. Taki A, Nakatani Y, Misugi K, Yao M, Nagashina Y. Chromophobe renal cell carcinoma: an immunohistochemical study of 21 Japanese cases. Mod Pathol 1999;12:310-317.