Adrenal lesion in multiple endocrine neoplasia type 1
Britt Skogseid, MD, Jonas Rastad, MD, Anders Gobl, PhD, Catharina Larsson, MD, Karin Backlin, MB, Claes Juhlin, MD, Göran Åkerström, MD, and Kjell Öberg, MD, Uppsala and Stockholm, Sweden
Background. Multiple endocrine neoplasia (MEN) type 1 is accompanied by adrenal involvement, but characteristics and clinical handling of this lesion have been insufficiently explored. Methods. Patients with MEN 1 (n = 43) were monitored (mean, 6.3 years) with annual biochemical and radiologic adrenal evaluation. Adrenal specimens were examined by in situ RNA-RNA hybridization for expression of the MENI candidate gene phospholipase C B3 (PLC (3) and immunostaining for insulin-like growth factor-1 receptor.
Results. Altogether 17 patients (40%) displayed adrenal enlargement, which was limited to the adrenal cortex and showed signs of progression, marked atypia, and cancer development in three of them. Only the carcinoma exhibited adrenocortical hormone excess. PLC B3 was expressed in the hyperplastic and adenomatous proliferation but not the carcinoma. Pancreatic endocrine tumors with insulin-proinsulin excess were overrepresented in the patients with adrenocortical involvement, but significant insulin-like growth factor-1 receptor immunoreactivity was restricted to the carcinoma. Conclusions. The prevalent adrenocortical lesion associated with MEN 1 requires regular attention because of malignant potential. It was unrelated to loss of constitution heterozygosity for the MENI locus (11q13) and PLC B3 expression, except for the cortical carcinoma exhibiting allelic losses involving also the Wiedemann-Beckwith gene at 11p15. Mechanisms for mitogenic relationships between the pancreatic and adrenal lesions of MEN 1 demand further clarification. (SURGERY 1995;118:1077-82.)
From the Departments of Internal Medicine and Surgery, University Hospital, Uppsala, and Department of Clinical Genetics, Karolinska Hospital, Stockholm, Sweden
CLASSIC LESIONS OF multiple endocrine neoplasia type 1 syndrome (MEN 1) encompass possibly monoclonal tu- mors of the parathyroids, endocrine pancreas, and an- terior pituitary gland.1,2 The MEN1 locus has been mapped to chromosome 11q13.3 Tumorigenesis in- volves inactivation of the wild-type MENI allele by vari- ably extensive deletions also in individual patients, which unmasks a constitutional mutation in a hitherto unidentified gene with tumor suppressor function.3, 4 A cDNA located within the 900 kb MENI locus has been found to represent phospholipase C (PLC) 33.3 This MEN1 candidate gene encodes a protein belonging to a family of signal transduction enzymes and exhibits re- duced expression in pancreatic endocrine tumors in- cluding those of MEN 1.
Supported by the Swedish Cancer Society, Lion’s Cancer Fund, and Torsten and Ragnar Söderbergs’s Fund.
Presented at the Sixteenth Annual Meeting of the American Associa- tion of Endocrine Surgeons, Philadelphia, Pa., April 23-25, 1995.
Reprint requests: Britt Skogseid, MD, Department of Internal Medi- cine, University Hospital, S-751 85 Uppsala, Sweden.
Copyright @ 1995 by Mosby-Year Book, Inc. 0039-6060/95/$5.00 + 0 11/6/67532
It has become increasingly evident that other neo- plasms are also overrepresented in patients with MEN 1.6 In a previous analysis adrenocortical proliferation was shown to be restricted to patients with MEN 1 with history of pancreatic endocrine tumors preferentially accompanied by insulin or proinsulin excess.7 In con- trast to the concept of tumorigenesis in hereditary neo- plasia,8 Southern blot analysis substantiated retained heterozygosity for the MENI locus in the benign adrenocortical lesions. A single adrenocortical carci- noma, however, showed loss of the MEN1 wild-type al- lele, which also encompassed losses of 11p alleles including the locus of hereditary adrenocortical carci- noma of Wiedemann-Beckwith syndrome.7,9 In this pa- per we report extended clinical and morphologic char- acteristics on the adrenocortical involvement of MEN 1, including PLC ß3 expression, as well as our current strategy for the handling of these patients.
MATERIAL AND METHODS
Patients and tissues. The study comprises 43 patients with MEN 1 (22 women and 21 men) from 11 MEN 1 kindreds, who had mean age of 54 years (range, 27 to 80 years) at study entry. The criteria for inclusion
encompassed indisputable MEN 1 diagnosis,6 possibili- ties for repeated radiologic and biochemical follow-up, and a computed tomography scan with 4 mm thick, contiguous slices of the adrenal area. Adrenal enlarge- ment was diagnosed when width of the adrenal wings exceeded twice the normal size. All recruited patients were normocalcemic after parathyroid surgery. Sixty- seven percent had pancreatic endocrine tumors, which were diagnosed at mean 43 years of age and recognized biochemically during 7.4 years. Pituitary enlargement encompassed 44% of the patients, who exhibited histo- ries of prolactin (n = 15) or growth hormone (n = 3) excess during mean 7.8 years before successful therapy. Another two of these patients had pituitary null cell ade- nomas. The patients were monitored 6.3 years (range, 1 to 11 years) after inclusion, and all of them were subjected to annual biochemical and radiologic reassessments.
Patients with adrenal involvement were monitored mean 4.7 years (range, 0.6 to 11 years) after recognition. Seven of them underwent unilateral adrenalectomy, and an additional enlarged gland was subjected to per- cutaneous core biopsy. All specimens were investigated by routine microscopy, which included staining with a polyclonal chromogranin antiserum.10 Tissue pieces were immediately frozen and stored at -70° C.
Biochemistry. Patients with adrenal enlargements were investigated as described.7 This examination in- cluded measurements of corticotrophin, proopiomel- anocortin, adrenocorticotropic hormone (ACTH), neu- ropeptide Y, dehydroepiandrosterone sulfate, aldoster- one, testosterone, and diurnal cortisol values in blood, urinary cortisol and catecholamines, and a urinary ste- roid profile. A 12-hour dexamethasone (1 mg) suppres- sion test was performed, whereby at least 50% reduction of serum cortisol values was considered normal. A cor- ticotrophin stimulation test evaluated plasma ACTH and serum cortisol responses during 90 minutes after 100 mg corticotrophin releasing factor administered intravenously (Corticoliberin human; Bissendorf Pep- tide, Hannover, Germany).
In situ hybridization. In situ RNA-RNA hybridization for PLC 33 was performed on enlarged portions of four MEN 1 adrenals, two of which have previously been in- vestigated for allelic loss,7 and two macroscopically and microscopically normal cortical portions of these spec- imens. Antisense and sense cRNA probes were made from a plasmid containing human PLC ß3 cDNA.5 Cry- osections of 4 um were incubated in 0.1 mol/L trieth- anolamine (pH 8.0) and 0.25% acetic anhydride for 10 minutes at room temperature. Approximately 5 x 107 cpm/ml of 35S-probe was applied in ×2 standard saline citrate (SSC) 50% formamide solution (×1 SSC is 0.15 mol/L sodium chloride and 0.015 mol/L sodium citrate; pH 7.0) containing 10% dextran sulfate (Kabi- Pharmacia, Uppsala, Sweden), 1 mg/ml tRNA (Boe-
hringer Mannheim, Mannheim, Germany), 1 mg/ml salmon sperm DNA (Sigma Chemical Company, St. Louis, Mo.), 2 mg/ml bovine serum albumin (Sigma Chemical Company) and 100 mmol/L dithiothreitol. After hybridization in a humidified chamber at 50° Cfor 4 hours or longer, the sections were washed in x2 SSC- 50% formamide at 56° C and treated with RNAse A (Boehringer Mannheim) for 15 minutes at 37° C. Auto- radiographic exposure encompassed 2 weeks at 4° C. RNA integrity was ascertained by ß-actin cRNA hybrid- ization. Antisense grain counts exceeding approxi- mately twice the background (sense) levels were con- sidered positive.
Immunohistochemistry. Immunohistochemical in- vestigation for insulin-like growth factor-1 (IGF-1) receptor involved 6 um cryosections from two micro- scopically normal and four diseased portions of the MEN 1-associated adrenal lesions and four biochemi- cally nonhyperfunctioning cortical adenomas of pa- tients without MEN 1. After normal rabbit serum and 0.3% H2O2 sections were exposed for 30 minutes to a murine monoclonal antibody (20 µg/ml; Genzyme, Boston, Mass.), which immunoprecipitates the & and B subunits of the human IGF-1 receptor but not the insu-
lin and IGF-2 receptors,11 the immunoreaction was visualized with a peroxidase antiperoxidase technique and ethylcarbazole and counterstaining with Mayer’s hematoxylin. Antibodies were diluted in phosphate- buffered saline solution containing 1% bovine serum albumin and applied at room temperature. The mono- clonal antiparathyroid antibody E11 (10 µg/ml12) and nonimmune rabbit serum comprised negative controls failing to stain the cortical specimens.
Dispersed cells were generated from the adrenocor- tical cancer as described.13 Briefly, the tissue was cut fine and digested for 1 hour at 37° C in collagenase (2 mg/ ml; Boehringer Mannheim) and DNase (0.05 mg/ml; Sigma Chemical Company). After exposure to a calci- um-deficient medium and filtering through nylon meshes, the suspensions were purified on 25% isotonic Percoll (Kabi-Pharmacia). Cells with greater than 95% ability to exclude trypan blue were incubated at 4° C with the IGF-1 receptor antibody (20 µg/ml) followed by fluorescein isothiocyanate-coupled anti-mouse im- munoglobulin G and analyzed by epifluorescence mi- croscopy.
RESULTS
Enlarged adrenals were radiologically detectable in 17 (40%) of the patients. The adrenal lesions were bi- laterally distributed in seven of them (41%), and they averaged 2.4 cm in diameter (range, ~1 to 10 cm). Routine histologic examination of adrenalectomy and biopsy specimens substantiated diffuse or nodular hy- perplasia including a degenerative cyst (n = 4), differ-
A
B
entiated cortical adenoma (n=2), and single cases of adenoma with prominent atypia and cortical carci- noma. Size of the adrenal lesions remained unchanged during the observation period in 15 of the patients, whereas growth was observed in two patients. One of them displayed bilateral, minor (~1 cm), and partly calcified expansions during 4 years of observation. Dur- ing 13 months between reassessments this patient had a unilateral, 10 cm sized adrenocortical carcinoma de- velop associated with testosterone and estradiol excess and a pathologic urine steroid profile. After adrenalec- tomy she has undergone excision of a feminizing local recurrence. After 6 years of annual follow-up for a uni- lateral adrenal enlargement (~1 cm), another patient currently displays progression to 4 cm sized lesion with- out signs of endocrine excess. The measurements of hy- pothalamic, pituitary, and adrenal hormones showed no disturbances, except in a patient with empty-sella syndrome and the single case of adrenocortical carci- noma.
Comparison of the patients with MEN 1 with and without enlarged adrenals failed to show significant (p> 0.05) differences in mean ages at inclusion (50 ver- sus 56 years) and MEN 1 diagnosis (41 versus 42 years) and recognized durations of MEN 1 syndrome (9.4 ver- sus 14 years) and the pituitary hormone excess (8.4 ver- sus 7.3 years) before therapeutic normalization (Stu- dent’s unpaired t tests), as well as with respect to the proportion of pituitary tumor involvement (53% versus 38%, chi-squared test). Concomitant pancreatic endo- crine tumors, however, were significantly overrepre- sented (p<0.05, chi-squared test) in the patients exhib-
iting adrenal involvement on biochemical or radiologic examinations or pancreatic surgery. In fact, all patients with adrenal lesions had pancreatic endocrine lesions, and these were demonstrable in 46% of those without recognized adrenal proliferation. Furthermore, there was an overrepresentation of normoglycemic patients with insulin or proinsulin excess among those with ad- renal affection (82% versus 33%).
IGF-1 receptor immunostainings showed weak, gran- ularly cytoplasmic reactivity in the parenchyma of the normal adrenal cortices and an adenoma associated with MEN 1, as well as in two of the nonhyperfunction- ing sporadic adenomas. The cortical parenchymas of the other specimens were unreactive except for the car- cinoma. This patient exhibited IGF-1 receptor staining, which encompassed 50% to 75% of the parenchymal cells and principally seemed to outline their periphery (Fig. 1, A). The surface immunoreactivity was confirmed by examination of dispersed cells (Fig. 1, B). All the in- vestigated tissues showed stromal IGF-1 receptor reac- tivity.
In situ antisense probe hybridization of normal adre- nal tissue and one adenoma and two hyperplasias from patients with MEN 1 displayed similar PLC 33 mRNA expression, which was evenly distributed over the cortex (Fig. 2, A). This level of expression equalled findings in normal pancreatic specimens from patients with MEN 15 and clearly exceeded the parallel sections hybridized with the sense probe (Fig. 2, B). In the adrenocortical carcinoma, however, the PLC 33 gene showed only background levels of expression (Fig. 2, D and E). The benign MEN 1-associated cortical lesions previously
A
B
C
D
E
F
have displayed retained constitutional heterozygosity for markers flanking the MEN1 locus, whereas the car- cinoma was characterized by loss of the MEN1 wild-type allele and alleles on 11p, 13q, and 17p .?
DISCUSSION
Consensus has not been reached in management of the MEN 1-associated adrenocortical proliferation de- spite the fact that this lesion is clearly overrepresented on clinical and autopsy analyses of this syndrome.7,14 The current prospective study includes 17 affected pa-
tients monitored for up to 11 years and consequently could serve as a primary basis for such an attempt. All patients with MEN 1 with adrenal lesions in this report and, to our knowledge, also other published materials have had cortical involvement combined with history of pancreatic endocrine tumors.6 This relationship com- prises an indirect but clinically strong argument for coupling between the adrenocortical and pancreatic proliferations of MEN 1. The pituitary, however, failed to display such coupling with respect to both its fre- quency and duration of hormone hypersecretion. Our
material of adrenal lesions included one patient with an empty sella demanding full substitution before devel- opment of detectable adrenal involvement.
Most of the current adrenal enlargements exhibited clinically and morphologically rather indolent charac- teristics. In three patients, however, this lesion displayed clinically more aggressive course with rapid growth, pronounced atypia, and cortical cancer development. Moreover, no clinical characteristic could predict this development, although these patients exhibited pan- creatic lesions accompanied by insulin or proinsulin excess. These circumstances underline that the MEN 1 adrenocortical involvement requires clinical attention and that adrenal radiology should involve all patients with MEN 1 exhibiting signs of pancreatic tumor on thorough biochemical evaluation. Known adrenal en- largement preferentially should be followed biochemi- cally and radiologically every year to every second year. A urinary steroid profile possibly constitutes the sole bio- chemical measure of highest diagnostic yield with respect to both the cortical dysfunction and malignant transfor- mation.15 Patients with symptoms or biochemical signs of cortical hormone excess in the absence of Morbus Cush- ing should be suspected for adrenocortical carcinoma.
The overtly normal PLC ß3 mRNA expression in the benign adrenocortical lesions strengthens the hypothe- sis generated by studies of allelic losses in the MEN 1 adrenal, parathyroid, and pancreatic lesions.7 The be- nign adrenocortical proliferation of this syndrome con- sequently seems to represent a MENI gene indepen- dent phenomenon, which retains PLC 33 expression and the wild-type MEN1 allele. These findings raise the question about mechanisms contributing to develop- ment of the adrenocortical involvement, which at least initially may comprise a polyclonal proliferation. The overrepresentation of pancreatic lesions causing in- sulin-proinsulin excess in these patients implies roles of adrenal IGF-1 and 2 receptors and relevance of endo- crine or even autocrine loops. Activation of these potentially mitogenic receptors in the adrenal cortex may be attained by insulin and proinsulin.16, 17 On the other hand, our clinical routine presents no evidence that sporadic insulinomas are accompanied by overrep- resentation of radiologically detectable adrenocortical enlargement, which at least partly coincides with the demonstration of meager IGF-1 receptor expression in the normal cortical parenchyma. Moreover, the current findings suggested surface IGF-1 receptor expression merely in the MEN 1-associated cortical carcinoma, which is a common finding in sporadic adrenocortical carcinomas causing androgen excess.17 The role of IGF receptors and effects of possibly other circulating mito- gens in patients with MEN 118, 19 consequently require further evaluation.
The background levels of PLC ß3 mRNA expression
in the adrenocortical carcinoma was expected, because this specimen previously has shown loss of the wild-type allele for the MEN1 locus. These findings nevertheless strengthen the suggestion that PLC ß3 is a good candi- date for the MEN1 gene,5 because the allele harboring the germline mutation was affected to such an extent that no message seemed to be translated in absence of the wild-type allele. The tumor did, however, also show loss of heterozygosity for 11p markers, including the Wiedenmann-Beckwith locus.9 This circumstance could be the ultimate cause of tumorigenesis in the adrenal cortex, whereby loss of the MEN1 locus in the present carcinoma might be fortuitous.
Colleges at the Departments of Radiology (Per Gunnar Lindgren, Håkan Ahlström), Clinical Chemistry (Leif Wide), and Pathology (Erik Wilander) of the Uppsala University Hos- pital are gratefully acknowledged for study contributions.
REFERENCES
1. Vasen HFA, Lamers CBHW, Lips CJM. Screening for the multi- ple endocrine neoplasia syndrome type 1: a study of 11 kindreds in the Netherlands. Arch Intern Med 1989;149:2717-22.
2. Skogseid B, Eriksson B, Lundquist G, et al. Multiple endocrine neoplasia type 1: a ten year prospective screening study in four kindreds. J Clin Endocrinol Metab 1991;73:281-7.
3. Larsson C, Skogseid B, Öberg K, Nakamura Y, Nordenskjöld M. Multiple endocrine neoplasia type 1 maps to chromosome 11 and is lost in insulinoma. Nature 1988;332:85-7.
4. Beckers A, Abs R, Reyniers E, et al. Variable regions of chromo- some 11 loss in different pathological tissues of a patient with the multiple endocrine neoplasia type I syndrome. J Clin Endocrinol Metab 1994;79:1498-502.
5. Weber G, Friedman E, Grimmond S, et al. The phospholipase C B3 gene located in the MEN1 region show loss of expression in endocrine tumours. Hum Mol Genet 1994;3:1775-81.
6. Skogseid B, Rastad J, Öberg K. Multiple endocrine neoplasia type 1: clinical features and screening. Endocrinol Metab Clin North Am 1994;23:1-18.
7. Skogseid B, Larsson C, Lindgren P-G, et al. Clinical and genetic features of adrenocortical lesions in multiple endocrine neopla- sia type 1. J Clin Endocrinol Metab 1992;75:76-81.
8. Knudson AG. Mutation and cancer: statistical study of retino- blastoma. Proc Natl Acad Sci USA 1971;68:820-3.
9. Koufos A, Grundy P, Morgan K, et al. Familial Wiedemann-Beck- with syndrome and a second Wilms’ tumor locus both map to 11p 15.5 Am J Hum Genet 1989;44:711-9.
10. Eriksson B, Arnberg H, Öberg K, et al. A polyclonal antiserum against chromogranin A and B-a new sensitive marker for neu- roendocrine tumors. Acta Endocrinol 1990;122:145-55.
11. Kull FC, Jacobs S, Su YF, Svobuda ME, Van Wyk JJ, Cuatrecasas P. Monoclonal antibodies to receptor for insulin and somatome- din C. J Biol Chem 1983;258:6561-6.
12. Juhlin C, Holmdahl R, Johansson H,-Rastad J, Åkerström G, Klareskog L. Monoclonal antibodies with exclusive reactivity against parathyroid cells and tubule cells of the kidney. Proc Natl Acad Sci USA 1987;84:2990-4.
13. Backlin C, Juhlin C, Grimelius L, et al. Monoclonal antibodies recognising normal and neoplastic human adrenal cortex. Endocr Pathol 1995;6:21-34.
14. Ballard HS, Frame B, Hartsock RJ. Familial multiple endocrine adenoma-peptic ulcer complex. Medicine 1964;43:481-516.
15. Gröndal S, Eriksson B, Hagnäs L, Werner S, Curstedt T. Steroid
profile in urine: a useful tool in the diagnosis and follow up of adrenocortical carcinoma. Acta Endocrinol 1990;122:656-63.
16. La Rocca R, Danesi R, Stein C, et al. Insulin like growth factor II mRNA expression in human adrenocortical carcinoma [Ab- stract]. Proc Annu Meet Am Assoc Cancer Res 1989;30:A246.
17. Backlin K, Juhlin C, Hellman P, Åkerström G, Rastad J. Immuno- histochemical detection of insulin-like growth factor (IGF) 1 and its receptor in human adrenocortical tumors [Abstract]. Pro- ceedings of the Fourth International Conference on Anticancer Research. Crete, 1992:30.
18. Zimering MB, Brandi ML, deGrange DA, et al. Circulating fibro- blast growth factor-like substance in familial multiple endocrine neoplasia type 1. J Clin Endocrinol Metab 1990;70:149-54.
19. Zimering MB, Riley DJ, Thakker-Varia S, et al. Circulating fibro- blast growth factor-like autoantibodies in two patients with mul- tiple endocrine neoplasia type 1 and prolactinoma. J Clin Endo- crinol Metab 1994;79:1546-52.
DISCUSSION
Dr. Norman Thompson (Ann Arbor, Mich.). Thanks for bringing this to our attention. I think you have one of the larg- est series with lesions in the adrenal, and they are a dilemma. Did you say only one of your tumors was functional? This is the kind of dilemma I have run into on several occasions. This was a patient with MEN 1 who presented with both Zollinger-Elli- son syndrome and Cushing’s syndrome, and one goes through a differential diagnosis of Cushing’s syndrome in such a patient. This patient had a low ACTH level, and it was obvious that there was a primary adrenal cause for Cushing’s syn- drome.
The next slide shows the cut specimen of the left adrenal gland, which shows the combination of a tumor and hyperpla- sia. The other adrenal was hyperplastic as well. We frequently see bilateral hyperplasia in patients with MEN I, often without functional abnormalities.
I excised this adrenal and left the other adrenal intact. When they are smaller than this and nonfunctional, we are leaving them. You raise the possibility of malignancy, but there are only about five cases in the world literature that I can find. When it is an incidental adenoma, should we just leave it there? How often should we do follow-up studies in these patients? What, in your opinion, is the likelihood of malignancy devel- oping in the long term in individual MEN 1 cases?
Dr. Rastad. There are rare descriptions of functional corti- cal lesions in patients with MEN 1. There is one from the Netherlands who was associated with an aldosterone excess. In your case, which we have discussed earlier, we don’t know whether the gastrinoma produced corticotrophin or ACTH. In the event of normal to suppressed ACTH levels, your patient might represent one of the few benign functional lesions in the adrenal cortex of patients with MEN 1. We would suggest that functional tumors should be removed, particularly when symptoms and signs of adrenocortical excess with minor adre- nal involvement develop during follow-up in these patients.
All the functional tumors with MEN 1 have been enlarged, which supports that unilateral adrenalectomy is sufficient in these cases. We would follow up these patients annually, and we would perform operation if the lesion progresses in terms of radiologic size or develops biochemical signs of steroid ex- cess. We would leave nonhyperfunctioning lesions measuring more than about 2 cm, particularly if they exhibit the previously mentioned signs of progression. We don’t have a lot of experience, however, because this is uncommon.
Dr. Clark. Do you use similar criteria for incidental ade- nomas, in general?
Dr. Rastad. That would be quite similar, yes. We know, however, that the risk progression is higher in patients with MEN 1 than in an ordinary incidental adenoma, and we would follow them more regularly.