Anti-&-Inhibin Marker of Choice for the Consistent Distinction between Adrenocortical Carcinoma and Renal Cell Carcinoma in Fine-Needle Aspiration
Patricia A. Fetsch, M.T.1 Celeste N. Powers, M.D.2 Maureen F. Zakowski, M.D.3 Andrea Abati, M.D.1
1 Cytopathology Section, Laboratory of Pathology, National Institutes of Health/National Cancer Insti- tute, Bethesda, Maryland.
2 Cytopathology Section, Virginia Commonwealth University-Medical College of Virginia, Richmond, Virginia.
3 Department of Pathology, Memorial Sloan-Ket- tering Cancer Center, New York, New York.
BACKGROUND. Anti-«-inhibin, an antibody directed against a peptide hormone, has been shown to be a useful diagnostic aid in surgical pathology material for the identification of sex cord-stromal neoplasms and recently has been described in adrenocortical carcinoma (ACC). The diagnosis of ACC versus renal cell carcinoma (RCC) may be difficult morphologically, particularly in fine-needle aspiration (FNA) material. To date, the immunohistochemical distinction of ACC from RCC is based on a panel of antibodies that include vimentin, cytokeratins, and epithelial membrane antigen. However, the reliability of this panel is weakened by incon- sistent staining patterns.
METHODS. Archival formalin fixed, paraffin embedded cell block sections from 45 FNAs of known primary and metastatic ACC and RCC as well as benign adreno- cortical nodules were stained with anti-«-inhibin using an avidin-biotin procedure. All samples were microwave pretreated and a biotin block was performed to reduce the background stain due to the high endogenous biotin often present in these types of samples.
RESULTS. All cases of ACC (n = 7; 100%) and benign adrenocortical cells (n = 15; 100%) were immunoreactive with the «-inhibin antibody, showing a diffuse cyto- plasmic and granular staining pattern. The staining intensity and number of immunoreactive cells varied within each sample, with the cases of ACC having the greatest proportion of immunoreactive cells and the strongest intensity. None of the cases of RCC (n = 23; 0%) were immunoreactive with anti-@-inhibin.
CONCLUSIONS. The morphologic distinction of ACC versus RCC in FNA material from renal, adrenal, and metastatic neoplasms is not always feasible based on cytology alone. However, due to the advent of the @-inhibin antibody, the reliable distinction of these entities now may be possible. The intense and specific immu- nostaining pattern for cells of adrenal origin, even in paucicellular samples, sug- gests potential for the widespread clinical utility of this marker by cytopathologists. Cancer (Cancer Cytopathol) 1999;87:168-72. @ 1999 American Cancer Society.
KEYWORDS: «-inhibin, immunocytochemistry, adrenocortical carcinoma, renal cell carcinoma, fine-needle aspiration.
I nhibin is a peptide hormone, normally produced by ovarian gran- ulosa cells, that inhibits the release of follicle-stimulating hormone from the pituitary gland.1 It is comprised of an alpha (a) subunit and a beta (B) subunit, and also has been shown to occur within a variety of normal tissues including the testes, placenta, pituitary gland, ad- renal cortex, renal tubules, and liver.1-8
Anti-a-inhibin, an antibody directed against the alpha subunit of this hormone, has been shown to be a useful diagnostic aid in surgical pathology material for the identification of sex cord-stromal neo- plasms and recently was described in adrenocortical carcinoma
Address for reprints: Andrea Abati, M.D., Cytopa- thology Section, Laboratory of Pathology, National Institutes of Health/National Cancer Institute, Bldg. 10, Room 2A19, 10 Center Drive MSC 1500, Be- thesda, MD 20892-1500.
Received October 30, 1998; revision received March 4, 1999; accepted March 10, 1999.
(ACC).1-4, 9-13 The morphologic distinction of ACC ver- sus renal cell carcinoma (RCC) may be difficult mor- phologically, particularly in fine-needle aspiration (FNA) material.14,15 To date, the immunohistochemi- cal differentiation of ACC from RCC is based on a panel of antibodies, including vimentin, cytokeratins, and epithelial membrane antigen (EMA).3,16,17 How- ever, the reliability of this panel is weakened by over- lapping and inconsistent staining patterns. In addi- tion, paucicellular specimens, common in FNA samples, may contribute to discrepancies between re- sults obtained in cytologic versus histologic samples. Positive immunoreactivity of even a few cells in a cytologic sample potentially may be of great signifi- cance.
In this study we characterize the immunostaining patterns against the a-inhibin subunit in ACC, benign adrenal cortical cells, and RCC and address the use- fulness of inhibin as a marker in the differential diag- nosis of ACC in cytologic material.
MATERIALS AND METHODS
Archival formalin fixed, paraffin embedded cell block sections from 45 FNAs of known primary and meta- static ACC (n = 7) and RCC (n = 23), as well as benign adrenocortical nodules (n = 15), were retrieved from the files of the National Institutes of Health/National Cancer Institute, Virginia Commonwealth University- Medical College of Virginia, and Memorial Sloan-Ket- tering Cancer Center. Samples of RCC included cases of documented granular, clear cell, sarcomatoid, chro- mophil, and mixed subtypes. All specimens were stained with anti-«-inhibin (1:10 dilution; Serotec, Ra- leigh, NC) using a modified avidin-biotin procedure (Vector Laboratories, Burlingame, CA) with 3,3’-dia- minobenzidine as the chromogen (Sigma Chemical Co., St. Louis, MO). A negative control (purified my- eloma protein, mouse immunoglobulin G1 kappa; Or- ganon Teknika Corp., Durham, NC) was run on each sample to assess background staining. Positive con- trols (normal testes) were assayed in parallel at the time of testing. All samples were microwave pre- treated, and a biotin block (Vector Laboratories) was performed to diminish background staining.18,19 The percentage and staining intensity of immunoreactive cells was evaluated blindly by one pathologist (A.A.).
RESULTS
All cases of ACC tested (n = 7; 100%) were immuno- reactive with the a-inhibin antibody, showing a dif- fuse cytoplasmic granular staining pattern. In 6 of the 7 cases of ACC tested (86%), immunostaining was present in > 75% of cells. In 1 case staining was present in < 25% of cells; however, this specimen was
sparsely cellular. The staining intensity was variable from cell to cell within each sample of ACC but gen- erally was moderate to strong (Fig. 1).
All cases of benign adrenal cortical lesions tested (n = 15; 100%) were immunoreactive with anti-a- inhibin. In the majority of these cases (11 of 15; 73%), positive immunoreactivity was evident in < 50% of cells in a given sample. In 4 of 15 cases (27%) staining was present in > 75% of cells. The staining intensity varied from cell to cell within each benign sample but, in contrast that of ACC, was of a weak-to-moderate intensity (Fig. 2).
All cases of RCC (n = 23) were completely non- immunoreactive (Fig. 3).
DISCUSSION
ACC is a rare malignant neoplasm with a poor prog- nosis. Histologic recognition of ACC generally is straightforward when there is a history of hormone production; however, ACC often is hormonally inac- tive. Due to its similarities to other neoplasms, most particularly RCC, the positive identification of ACC
can create a diagnostic dilemma on radiology and histology. Because FNA often is performed to diagnose retroperitoneal masses, pathologic interpretation by the cytology laboratory is routine. Comparison studies of ACC and RCC on FNA material have shown cyto- logic features characteristic to each entity.14 However, these criteria have not been proven reliable for differ- entiation in all cases.15 In addition, ACC and RCC may be difficult to identify by ultrastructural features.20
Currently, a panel of markers comprised of cyto- keratins, vimentin, and EMA provides the most valu- able means of discriminating ACC from RCC.3,16,17 Unfortunately, in many cases the immunohistochem- ical features overlap. The majority of RCCs show a simple cytokeratin pattern comprised of cytokeratins 8, 18, and 19. A similar spectrum of reactivity also can occur in ACC.17,21 In a study conducted by Miettinen, 50% of cases of ACC tested were positive for cytoker- atin.21 Similar studies noted cytokeratin immunoreac- tivity rates of 36% in ACC and up to 100% in RCC.22,23 Conversely, a study performed by Wick et al. found that cytokeratin analysis performed without protease
pretreatment consistently showed positive staining with RCC and nonimmunoreactivity with ACC.17 Thus, cytokeratin staining of these entities is not a reliable determinant. In addition, it has been shown that ad- enomas more frequently are immunoreactive with cy- tokeratins than with vimentin, contributing to the di- agnostic dilemma.16
On histologic samples it generally is accepted that ACC display immunoreactivity for vimentin but not for EMA whereas RCC most often are nonimmunore- active with vimentin but immunoreactive with EMA.17,21 However, these results also are variable. Studies of ACC by Tartour et al. showed an immuno- reactivity rate of 23% with vimentin and focal EMA immunoreactivity was detected in 2 of 12 cases test- ed.22 Thus the demand for sensitive and specific mark- ers of ACC is apparent.
Inhibin is a glycoprotein comprised of a and B subunits.5,6 It is secreted by several sex cord-stromal gonadal derivatives, with extragonadal expression oc- curring in the placenta, pituitary gland, adrenal gland, and liver.4,7,8,24 Laboratory quantitation of serum in- hibin levels has been used as an early marker for
tumor growth and recurrent disease in patients with granulosa cell tumors.7
Immunostaining with monoclonal antibodies to «-inhibin has been observed in the granulosa and theca cells of the ovary, Sertoli/Leydig cells within the testes, renal tubules, adrenal cortex, and placenta.1-3 The primary use of this antibody thus far has been as a marker of sex cord-stromal neoplasms.2,11,12 Granu- losa and Sertoli/Leydig cell tumors have shown im- munoreactivity with this antibody; other ovarian neo- plasms, germ cell tumors, soft tissue tumors, melanomas, lymphomas, and metastatic carcinomas previously tested have been found to be nonimmuno- reactive1,4,9,12 However, the interpretation of inhibin positive cells requires caution due to significant num- bers of positive staining, luteinized cells in a variety of primary and metastatic ovarian tumors.13
Immunoreactivity has been observed consistently in the normal adrenal cortex, with positivity noted most intensely in the inner cell layers representing the zona reticularis.24 Studies conducted with anti-@-in- hibin in surgical pathology material yielded promising results; positive immunostaining has been shown in the majority of cases of ACC and adenomas, but to our knowledge no staining has been observed in cases of RCC.3 The staining pattern has been described as cy- toplasmic but variable within the sample.3,10 Both cor- tisol- and aldosterone-producing and hormonally nonfunctional lesions of ACC have shown reactivity.24 It is interesting to note that tumors associated with the increased production of cortisol, androgen, or their precursors have demonstrated stronger immunoreac- tivity than aldosterone-secreting and nonfunctioning tumors.3 Similarly, in normal adrenal tissue, the most intense positive staining is found in the zona fascicu- lata (cortisol) and zona reticularis (androgen). Normal adrenal medulla is negative.3,10,24
The morphologic distinction of ACC versus RCC in FNA material of renal/adrenal and metastatic neo- plasms may not be possible based on cytology alone. In this study, all cases of ACC and benign adrenal lesions were immunoreactive with anti-«-inhibin, and all cases of RCC were nonimmunoreactive. The per- formance of the biotin block step was invaluable in interpretation because it greatly diminished the false- positivity associated with background staining due to endogenous biotin, which often is present in these types of tissues.19 The intense and specific immuno- staining pattern for cells of adrenal origin, even in paucicellular samples, suggests potential for the wide- spread clinical utility of this marker. Due to the advent of the a-inhibin antibody, the reliable distinction of these entities may now be possible for cytopatholo- gists.
REFERENCES
1. McCluggage W, Maxwell P, Sloan J. Immunohistochemical staining of ovarian granulosa cell tumors with monoclonal antibody against inhibin. Hum Pathol 1997;28(9):1034-8.
2. Pelkey T, Frierson H, Mills S, Stoler M. The diagnostic utility of inhibin staining in ovarian neoplasms. Int J Gynecol Pathol 1998;17(2):97-105.
3. Pelkey T, Frierson H, Mills S, Stoler M. The a subunit of inhibin in adrenal cortical neoplasia. Mod Pathol 1998;11(6): 516-23.
4. Flemming P, Wellmann A, Maschek H, Lang H, Georgii A. Monoclonal antibodies against inhibin represent key mark- ers of adult granulosa cell tumors of the ovary even in their metastases. Am J Surg Pathol 1995;19(8):927-33.
5. Chen CLC. Inhibin and activin as paracrine/autocrine fac- tors. Endocrinology 1993;132:4-5.
6. Jaatinen T, Penttila T, Kaipia A, Ekfors T, Parvinen M, Top- pari J. Expression of inhibin alpha, betaA, and betaB mes- senger ribonucleic acids in the normal human ovary and in polycystic ovarian syndrome. J Endocrinol 1994;143:127-37.
7. Lappohn R, Burger H, Bouma J, Bangah M, Krans M, Brujin H. Inhibin as a marker for granulosa-cell tumors. N Engl J Med 1989;321:790-3.
8. Spencer S, Rabinovici J, Mesiano S, Goldsmith P, Jaffe R. Activin and inhibin in the human adrenal gland. Regulation and differential effects in fetal and adult cells. J Clin Invest 1992;90:142-9.
9. McCluggage W, Shanks J, Whiteside C, Maxwell P, Banerjee S, Biggart J. Immunohistochemical study of testicular sex cord-stromal tumors, including staining with anti-inhibin antibody. Am J Surg Pathol 1998;22(5):615-9.
10. Chivite A, Matias-Guiu X, Pons C, Algaba F, Prat J. Inhibin A expression in adrenal neoplasms. Appl Immunohistochem 1998;6(1):42-9.
11. Kommoss F, Oliva E, Bhan A, Young R, Scully R. Inhibin expression in ovarian tumors and tumor-like lesions: an immunohistochemical study. Mod Pathol 1998;11(7):656- 64.
12. Costa M, Ames P, Walls J, Roth L. Inhibin immunohisto- chemistry applied to ovarian neoplasms: a novel, effective, diagnostic tool. Hum Pathol 1997;28(11):1247-54.
13. Hildebrandt R, Rouse R, Longacre T. Value of inhibin in the identification of granulosa cell tumors of the ovary. Hum Pathol 1997;28(12):1387-95.
14. Sharma S, Singh R, Verma K. Cytomorphology of adreno- cortical carcinoma and comparison with renal cell carci- noma. Acta Cytol 1997;41(2):385-92.
15. Sasano H, Shizawa S, Nagura H. Adrenocortical cytopathol- ogy. Am J Clin Pathol 1995;104(2):161-6.
16. Gaffey M, Traweek S, Mills S, Travis W, Lack E, Medeiros L, et al. Cytokeratin expression in adrenocortical neoplasia: an immunohistochemical and biochemical study with implica- tions for the differential diagnosis of adrenocortical, hepa- tocellular, and renal cell carcinoma. Hum Pathol 1992;23: 144-53.
17. Wick M, Cherwitz D, McGlennen R, Dehner L. Adrenocorti- cal carcinoma. Am J Pathol 1986;122(2):343-52.
18. Fetsch P, Cormier J, Hijazi Y. Immunocytochemical detec- tion of MART-1 in fresh and paraffin embedded malignant melanomas. J Immunother 1997;20(1):60-4.
19. Rodriguez-Soto J, Warnke R, Rouse R. Endogenous avidin- binding activity in paraffin-embedded tissue revealed after microwave treatment. Appl Immunohistochem 1997;5(1): 59-62.
20. Mackay B, Ordonez N, Khoursand J. The ultrastructure and immunocytochemistry of renal cell carcinoma. Ultrastruct Pathol 1987;11:483-502.
21. Miettinen M. Neuroendocrine differentiation in adrenocor- tical carcinoma: new immunohistochemical findings sup- ported by electron microscopy. Lab Invest 1992;66(2):169- 74.
22. Tartour E, Caillou B, Tenenbaum F, Schroder S, Luciani S, Talbot M, et al. Immunohistochemical study of adrenocor-
tical carcinoma: predictive value of the D11 monoclonal antibody. Cancer 1993;72(11):3296-303.
23. DeLong W, Grignon D, Eberwein P, Shum D, Wyatt J. Sar- comatoid renal cell carcinoma. Arch Pathol Lab Med 1993; 117:636-40.
24. McCluggage W, Burton J, Maxwell P, Sloan J. Immunohis- tochemical staining of normal, hyperplastic, and neoplastic adrenal cortex with a monoclonal antibody against «-in- hibin. J Clin Pathol 1998;51:114-6.