Immunocytochemistry in adrenocortical tumours: a clinicomorphological study of 72 neoplasms *
Sören Schröder 1, Barbara-Christina Padberg 1, Eike Achilles 1, Konstanze Holl 1, Henning Dralle 3, and Günter Klöppel 2
1 Institute of Pathology, University of Hamburg, Hamburg, Federal Republic of Germany,
2 Institute of Pathology, Free University of Brussels, Brussels, Belgium
3 Department of Surgery, Hannover School of Medicine, Hannover, Federal Republic of Germany
Received July 18, 1991 / Accepted August 22, 1991
Summary. Surgical specimens of 72 adrenocortical tu- mours (ACTs) were investigated. Histologically, 57 tu- mours were classified as adenomas and 15 as carcino- mas. In 9 of the latter cases, distant metastases and/or lethal outcome of disease was recorded. Immunocyto- chemistry showed only 2 ACTs to be positive for cytok- eratin and 6 for vimentin. None of the 72 tumours showed argyrophilia or immunoreactivity for epithelial membrane antigen (EMA), S-100 protein, chromogranin A, Leu 7 or Leu-M1, while 31 cases exhibited positivity on immunostaining with a polyclonal antiserum against synaptophysin. All 72 ACTs were immunoreactive with the recently described antibody D11. Thus the panel of antibodies described here could not discriminate be- tween adenomas and carcinomas or between carcinomas with aggressive and indolent behaviour. Immunostaining with D11 and for EMA and Leu-M1 may help to distin- guish ACTs from phenotypically similar lesions of differ- ent histogenesis.
Key words: Adrenal gland - Adrenocortical tumours - Immunocytochemistry - Differential diagnosis
Introduction
Endocrine tumours are a favoured subject for immuno- cytochemical investigations. Until now, however, few studies - with differing results - have been published on adrenocortical tumours (ACTs). Nevertheless the di- lemma of correct histogenetic typing (adrenocortical ver- sus medullary versus extra-adrenal-metastatic) exists especially in non-functioning ACTs and the problem
cannot be solved by conventional histology. Neoplasms with histologically or clinically evident adrenocortical derivation, however, frequently present problems in the assessment of their biological potential (adenoma versus carcinoma). The aim of the present study was to deter- mine whether immunocytochemistry might be a helpful adjunct to solving these questions.
Materials and methods
Formalin-fixed, paraffin-embedded material obtained from surgi- cal specimens of 72 primary ACTs was analysed using conventional histology, histochemistry (Grimelius’ silver stain) (Grimelius 1968) and immunocytochemistry. As previously described in detail (Pad- berg et al. 1991), all neoplasms were reclassified according to the criteria laid down by Hough et al. (1979), Weiss (1984) and van Slooten et al. (1985). For tumours thus classified as carcinomas, the mitotic activity was assessed and the lesions were designated high grade versus low grade according to the criteria of Weiss et al. (1989) (>20 versus ≤20 mitoses per 50 high power fields). Utilizing the ABC method (Hsu et al. 1981), antibodies against cytokeratin (CK), vimentin (VIM), epithelial membrane antigen (EMA), S-100 protein (S-100), neuron-specific enolase (NSE), chromogranin A (Chr A), synaptophysin (SYN), Leu 7, Leu-M1 and the monoclonal antibody D11 were applied. Specifications as to the source and dilution of the antisera used are listed in Table 1. The medical records were reviewed, and all patients were moni- tored until spring 1990 (mean follow-up period 25 months’ range 1-88 months).
Results
On reclassification in line with the cited criteria 57 tu- mours were typed as adenomas and 15 as carcinomas. This classification was repeatable. The mean tumour weights were 65 g (8-1080 g) for adenomas and 645 g (34-3100 g) for carcinomas; 9 adenomas weighed more, 2 carcinomas less than 50 g.
In both groups, a clear predominance of female pa- tients was found (M:F adenomas 1:4.1; carcinomas 1:2.8). Pre-operatively, in the adenoma/carcinoma pa- tients the following hormonal symptoms had been docu- mented: hyperaldosteronism (Conn’s syndrome) 22/0;
* Dedicated to Prof. G. Seifert, former director of the Institute of Pathology, University of Hamburg, on the occasion of his 70th birthday
Offprint requests to : S. Schröder, Institute of Pathology, University of Hamburg (UKE), Martinistrasse 52, W-2000 Hamburg 20, Fed- eral Republic of Germany
| Antibody | Species | Source | Dilution | Reference |
|---|---|---|---|---|
| Anti-CK | ||||
| KL 1 | Mouse | Immunotech | 1:1000 | Viac et al. (1983) |
| MA-902 | Mouse | Enzo Diagnostics | 1:500 | Gown and Vogel (1984) |
| Anti-VIM | Mouse | Boehringer Mannheim | 1:20 | Osborn et al. (1984) |
| Anti-EMA | Goat | Sera-Lab | 1:100 | |
| Anti-S-100 | Rabbit | Histoprime | 1:20 | |
| Anti-NSE | Rabbit | Dakopatts | 1:3000 | |
| Anti-Chr A | Mouse | Hybritech | 1:500 | |
| Anti-SYN | Rabbit | Biometra | 1:500 | Jahn et al. (1985) |
| Anti-Leu 7 | Mouse | Becton Dickinson | 1:50 | |
| Anti-Leu-M1 | Mouse | Becton Dickinson | 1:30 | |
| D11 | Mouse | Own source | 1:5 | Schröder et al. (1990) |
hypercortisolism (Cushing’s syndrome) 13/9; virilization 2/5 (1/4 patients were additionally afflicted by Cushing’s syndrome), feminization 1/0; 20 and 5 patients respec- tively in each group had not shown endocrine abnormal- ities (non-functioning ACTs). At the end of the observa-
tion period, 55 adenoma patients showed continuous symptom-free survival; 2 patients had previously died from unrelated causes. Of the carcinoma patients, 7 had died from the tumour, 1 from other causes; 2 patients are alive with persistent tumour manifestations, while
a
b.
C
d
| Adenomas | Carcinomas | |||||||||
|---|---|---|---|---|---|---|---|---|---|---|
| CK | VIM | S-100 | SYN | EMA | CK | VIM | S-100 | SYN | EMA | |
| Wick et al. (1986) | 0/10 | 10/10 | 0/10 | 0/10 | 0/10 | 10/10 | 0/10 | 0/10 | ||
| Cote et al. (1989, 1990) | 6/6 | 4/6 | 2/3 | 0/3 | 0/13 | 13/13 | 8/12 | 3/12 | ||
| Miettinen et al. (1985) | 1/2 | 0/2 | 3/7 | 3/7 | ||||||
| Henzen- | 4/4ª | 4/4 | 2/4ª | 4/4 | ||||||
| Logmans et al. (1988) | ||||||||||
| Present study | 1/57 | 2/57 | 0/57 | 21/57 | 0/57 | 1/15 | 4/15 | 0/15 | 10/15 | 0/15 |
a Positivity only seen in frozen sections
5 continue to live symptom-free 25-65 months after sur- gery. Of 9 adrenocortical carcinomas associated with ag- gressive behaviour, 4 fell into the low grade and 5 into the high grade category, as defined by mitotic activity. For the remaining 6 malignancies with clinically benign behaviour, the respective figures were 4 low grade and 2 high grade cases.
With both antibodies applied, CK immunoreactivity was only demonstrated in 1 adenoma and 1 carcinoma case (Fig. 1a). Two adenomas and 4 carcinomas were positive for VIM (Fig. 1b). None of the 72 ACTs was decorated by antibodies against EMA, S-100, Chr A, Leu 7 and Leu-M1 or showed argyrophilia. Upon NSE immunostaining, 12 adenomas and 8 carcinomas exhib- ited a doubtful (not clearly negative) result. In 21 ad- enomas and 10 carcinomas, an unequivocal cytoplasmic positivity was recorded upon immunostaining with the polyclonal SYN antiserum (Fig. 2a, b). All 72 ACTs were immunoreactive for D11; in addition to general nuclear staining, all 25 tumours associated with Cush- ing’s syndrome, feminization or virilization and 21 non- functioning neoplasms showed diffuse cytoplasmic stain- ing (Fig. 2c, d).
Discussion
Our series underlines the relevance of the three common- ly used classification schemes for the histological dis- crimination of benign and malignant ACTs (Hough et al. 1979; van Slooten et al. 1985; Weiss 1984). All neoplasms typed in this way as adenomas followed an uneventful postoperative course, although 9 of these weighed more than 50 g (some considerably more) - a point to remember in the differential diagnosis from ad- renocortical carcinoma (Schteingart et al. 1968), as re- cently re-emphasized by Rosai (1989). Surprisingly, how- ever, 5 of our 15 ACTs typed as carcinomas have re- mained recurrence-free. Hence, the malignancies in our material exhibited an altogether more favourable behav- iour than such neoplasms in several other reports in the literature.
In our study, mitotic activity was of limited value in predicting prognosis, since a benign course was seen for 2 of 7 high grade tumours, while 4 of 8 low grade malignancies showed aggressive behaviour. As we re- cently demonstrated in a cytophotometric study of 62 ACTs (Padberg et al. 1991), static DNA measurements do not enable us to make an adequate assessment of the malignant potential of individual adrenocortical car- cinomas cases or a precise differentiation between adren- ocortical adenomas and carcinomas. According to our findings, immunocytochemistry cannot be regarded as a helpful adjunct to solving these problems. Rather, its relevance lies in the correct histogenetic typing of ACTs - regardless of their biological potential.
Our observation of ACTs with phenotypic similarity to adrenomedullary tumours (and vice versa) (Schröder et al. 1990) emphasizes the well-known likelihood of confusion between the two tumour forms, as has been described authoritatively in monographs (Page et al. 1986) as well as individual case reports (Ramsay et al. 1987). The possibility of differentiating between primary and secondary adrenal neoplasms would appear to be even more important than the exclusion of phaeochro- mocytoma. In a literature survey recently performed by Lemmers et al. (1989), 14 cases of renal cell carcinomas (RCCs) with solitary metastasis to the contralateral ad- renal gland have been described, and two additional cases from our files could be added to this series (Schröder et al. 1990). Because of the juxtaposition of the adrenal gland to the kidney, it is also not uncommon for adrenocortical carcinoma to involve the renal paren- chyma at diagnosis. Hence, several authors have asked whether immunocytochemistry might be a helpful ad- junct to discriminating between ACTs and RCCs.
The immunocytochemical data published so far are restricted to relatively small case numbers and mostly to the analysis of few antigens, whereby - as illustrated in Table 2 - considerable discrepancies exist between different studies. Thus, CK has been reported to be al- ways negative in ACTs (Wick et al. 1986), or to be posi- tive in adenomas, while negative in carcinomas (Cote et al. 1989, 1990), or to be detectable in some, but not
all adenomas and carcinomas (Miettinen et al. 1985), and this only upon using frozen sections, while being consistently negative upon use of paraffin sections (Hen- zen-Logmanns et al. 1988). In our study, identical results were obtained with the broad-spectrum cytokeratin anti- body KL 1 (Viac et al. 1983) and the antibody MA-902 recognizing low-molecular-weight cytokeratin 8 (Gown and Vogel 1984), both probes showing positivity with each one of 57 adenomas and 15 carcinomas. We would thus conclude that CK positivity, although rare, cannot exclude the diagnosis of primary ACT.
Regarding antibodies against VIM, two (Henzen- Logmanns et al. 1988; Wick et al. 1986) of the four stu- dies concur in that adrenocortical adenomas and carci- nomas are invariably positively stained. This observation is, however, not shared by other authors (Cote et al. 1990; Miettinen et al. 1985) and contrasts with our re- sults using the monoclonal antibody V9. We found VIM immunoreactivity in 6 of 72 ACTs, somewhat more fre- quently among carcinomas as compared with adenomas, yet we cannot attribute any differential diagnostic value to this finding, since many RCCs also exhibit VIM posit- ivity (Cote et al. 1990; Waldherr and Schwechheimer 1985).
Immunostaining for EMA and Leu-M1 appears to be more promising in the discrimination of primary and secondary adrenal neoplasms. Hence, a total of 98 ACTs seen be Wick et al. (1986), Sloane and Ormerod (1981) and in our material were EMA negative. Positivity for this antigen was recorded in all 20 RCCs in the two previous series. All our 72 ACTs, in addition, were Leu- M1 negative, while we observed Leu-M1 positivity in 40 of 62 primary RCCs investigated with the same probe (unpublished data). Diagnostic relevance had also origi- nally been ascribed to the detection of blood group isoantigens A, B and H, reported to be present in all 10 RCCs, but in none of 20 primary ACTs (Wick et al. 1986). Recently, however, negativity for blood group an- tigens A, B, H and Lewis Y in both ACTs and RCCs has been communicated in all cases, while Lewis a, b or X were positive in many, though not all, primary RCCs (13/17), but negative among all 15 ACTs investi- gated (Cote et al. 1990). The diagnostic usefulness of Lewis blood group antigens is yet further hampered by the fact that metastatic RCCs express these markers even less frequently (Cordon-Cardo et al. 1989). It should also be mentioned in this context that lectin binding studies do not offer a practical decisive factor in the differential diagnosis of RCC, adrenocortical adenoma and carcinoma, since such investigations did not show significant discrepancies between these three (Sasano et al. 1988; Wick et al. 1986).
According to Wick et al. (1986), and in contrast to the findings of Cote et al. (1990) we did not see S-100 positivity among our ACTs. Confirming the further re- sults of the latter authors (Cote et al. 1990), all our 72 cases were negative upon Chr A immunostaining and all neoplasms were devoid of Leu 7 reactivity or argyro- philia. They thus lacked evidence of neuroendocrine dif- ferentiation. In our own and others’ experience (Heitz 1987; Nemeth et al. 1987) the diagnostic value of NSE
immunostaining is limited by an occasionally inconclu- sive result. In 20 lesions which were unequivocally relat- ed to the adrenal cortex, we observed an NSE reaction which could not clearly be defined as negative. However, the recently described diagnostic value of the antibody D11 (Schröder et al. 1990), which decorates only normal and neoplastic cortical cells in the adrenal, was conclusi- vely confirmed in this more extensive series of ACTs.
Without evident association with any other morpho- logical or functional feature, 31 ACTs showed reactivity upon immunostaining with a polyclonal antiserum against SYN (Jahn et al. 1985). Such positivity, which we never saw with normal adrenocortical cells, is analo- gous to SYN positivity (in the absence of Chr A staining) described by Cote et al. (1989) in 3 out of 6 metastatic adrenocortical carcinomas. Though these authors did not state the source of the antisera used, their results have most probably been obtained by the monoclonal SYN antibody originally described by Wiedenmann and Franke (1985) and since employed in several large stu- dies on the detection of SYN reactivity in a variety of neural and neuroendocrine tissues and tumours (Wie- denmann and Huttner 1989). Since no plausible explana- tion can be found for the occurrence of p38/SYN, known to be present in presynaptic (Jahn et al. 1985; Wiedenmann and Franke 1985) and chromaffin secreto- ry vesicles (Lowe et al. 1988), also in ACTs, “SYN-like immunoreactivity” appears to be an appropriate term to designate our positive immunocytochemical results among such lesions. This phenomenon presents a prob- lem regarding the immunocytochemical differential diag- nosis of adrenal tumours, since 3 of our 64 phaeochro- mocytomas investigated for a variety of neuroendocrine markers showed positivity only for SYN, while being negative for NSE, Chr A and Grimelius’ silver stain (Padberg et al. 1990). Our findings thus not only demand additional molecular biological analyses to clarify the SYN-like immunoreactivity of neoplastic adrenocortical cells, but should provoke a broader immunocytochemi- cal evaluation of the pattern of different antibodies against SYN in other non-neuroendocrine tissues and tumours.
Acknowledgements. Supported by the Deutsche Forschungsge- meinschaft (Schr 274/4-1) and the Hamburger Krebsgesellschaft (Nr. 384).
References
Cordon-Cardo C, Reuter VE, Finstand CL, Sheinfeld J, Lloyd KO, Fair WR, Melamed MR (1989) Blood group-related anti- gens in human kidney: modulation of Lewis determinants in renal cell carcinoma. Cancer Res 49:212-218
Cote R, Cordon-Cardo C, Reuter V, Rosen P (1989) Monoclonal antibodies in the diagnosis of adrenal cortical carcinoma (ab- stract 117). Mod Pathol 2:20 A
Cote RJ, Cordon-Cardo C, Reuter VE, Rosen PP (1990) Immuno- pathology of adrenal and renal cortical tumors. Coordinated change in antigen expression is associated with neoplastic con- version in the adrenal cortex. Am J Pathol 136:1077-1084
Gown A, Vogel AM (1984) Monoclonal antibodies to human inter- mediate filament proteins. II. Distribution of filament proteins in normal human tissues. Am J Pathol 114:309-321
Grimelius L (1968) A silver nitrate stain for a2 cells in human pancreatic islets. Acta Soc Med Ups 73:243-270
Heitz PU (1987) Neuroendocrine tumor markers. In: Seifert G (ed) Morphological tumor markers. Springer, Berlin Heidelberg New York, pp 279-306
Henzen-Logmanns SC, Stel HV, Muijen GNP van, Mullink HH, Meijer CJLM (1988) Expression of intermediate filament pro- teins in adrenal cortex and related tumours. Histopathology 12:359-372
Hough AJ, Hollifield JW, Page DL, Hartmann WH (1979) Prog- nostic factors in adrenal cortical tumors. A mathematical analy- sis of clinical and morphological data. Am J Clin Pathol 72:390-399
Hsu SM, Raine L, Fanger H (1981) Use of avidin-biotin-peroxidase complex (ABC) in immunoperoxidase techniques: a compari- son between ABC and unlabeled antibody (PAP) procedures. J Histochem Cytochem 29:577-580
Jahn R, Schiebler W, Ouimet C, Greengard P (1985) A 38,000- dalton membrane protein (p38) present in synaptic vesicles. Proc Natl Acad Sci USA 82:4137-4141
Lemmers M, Ward K, Hatch T, Stenzel P (1989) Renal adenocarci- noma with solitary metastasis to the contralateral adrenal gland: report of 2 cases and review of the literature. J Urol 141:1177-1180
Lowe AW, Madeddu L, Kelly RB (1988) Endocrine secretory gran- ules and neuronal synaptic vesicles have three integral mem- brane proteins in common. J Cell Biol 106:51-59
Miettinen M, Lehto VP, Virtanen I (1985) Immunofluorescence microscopic evaluation of the intermediate filament expression of the adrenal cortex and medulla and their tumors. Am J Pathol 118:360-366
Nemeth J, Galian A, Mikol J, Cochand-Priollet B, Wassef M, La- vergne A (1987) Neuron-specific enolase and malignant lym- phomas (23 cases). Virchows Arch [A] 412:89-93
Osborn M, Debus E, Weber K (1984) Monoclonal antibodies spe- cific for vimentin. Eur J Cell Biol 34:137-143
Padberg BC, Achilles E, Garbe E, Dralle H, Klöppel G, Schröder S (1990) Histologie, Immuncytochemie und DNA-Cytophoto- metrie des adrenalen Phäochromocytoms (PCC) - eine morpho- logisch-klinische Untersuchung an 64 Tumoren. Verh Dtsch Ges Pathol 74:289-294
Padberg BC, Lauritzen I, Achilles E, Holl K, Bressel M, Klöppel G, Dralle H, Schröder S (1991) DNA cytophotometry in adren- ocortical tumours - a clinicomorphological study of 66 cases. Virchows Arch [A] 419:167-170
Page DL, DeLellis RA, Hough AJ (1986) Tumors of the adrenal. Atlas of tumor pathology, series 2, fascicle 23. Armed Forces Institute of Pathology, Washington, DC
Ramsay JA, Asa SL, Nostrand AWP van, Hassaram ST, Harven EP de (1987) Lipid degeneration in pheochromocytomas mim- icking adrenal cortical tumors. Am J Surg Pathol 11:480-486 Rosai J (1989) Adrenal gland and other paragangliomas. In: Acker- man’s surgical pathology, 7th edn, vol 1. Mosby, St. Louis, pp 789-818
Sasano H, Nose M, Sasano N (1988) Lectin histochemistry in ad- renocortical hyperplasia and neoplasms with emphasis on carci- noma. Arch Pathol Lab Med 113:68-72
Schröder S, Niendorf A, Achilles E, Dietel M, Padberg BC, Beisie- gel U, Dralle H, Bressel M, Klöppel G (1990) Immunocyto- chemical differential diagnosis of adrenocortical neoplasms us- ing the monoclonal antibody D11. Virchows Arch [A] 417: 89- 96
Schteingart DE, Oberman HA, Friedman BA, Conn JW (1968) Adrenal cortical neoplasms producing Cushing’s syndrome. A clinicopathologic study. Cancer 22:1005-1013
Sloane JP, Ormerod MG (1981) Distribution of epithelial mem- brane antigen in normal and neoplastic tissues and its value in diagnostic tumor pathology. Cancer 47:1786-1795
Slooten H van, Schaberg A, Smeenk D, Moolenaar J (1985) Mor- phologic characteristics of benign and malignant adrenocortical tumors. Cancer 55:766-773
Viac J, Reano A, Brochier J, Staquet MJ, Thivolet J (1983) Reactiv- ity pattern of a monoclonal anti-keratin antibody (KL 1). J Invest Dermatol 81: 351-354
Waldherr R, Schwechheimer K (1985) Co-expression of cytokeratin and vimentin intermediate-sized filaments in renal cell carcino- mas. Virchows Arch [A] 408:15-27
Weiss LM (1984) Comparative histologic study of 43 metastasizing and non-metastasizing adrenocortical tumors. Am J Surg Pa- thol 8:163-169
Weiss LM, Medeiros LJ, Vickery AL (1989) Pathologic features of prognostic significance in adrenocortical carcinoma. Am J Surg Pathol 13:202-206
Wick MR, Cherwitz DL, McGlennen RC, Dehner LP (1986) Ad- renocortical carcinoma. An immunohistochemical comparison with renal cell carcinoma. Am J Pathol 122: 343-352
Wiedenmann B, Franke WW (1985) Identification and localization of synaptophysin, an integral membrane glycoprotein of Mr 38,000 characteristic of presynaptic vesicles. Cell 41:1017- 1028
Wiedenmann B, Huttner WB (1989) Synaptophysin and chromo- granins/secretogranins - widespread constituents of distinct types of neuroendocrine vesicles and new tools in tumor diagno- sis (review). Virchows Arch [B] 58:95-121