Adult-onset hypogonadotropic hypogonadism caused by aberrant expression of aromatase in an adrenocortical adenocarcinoma

Andrew Advani1), 2), Sarah J. Johnson3), Moira R. Nicol4), Georgia Papacleovoulou4), Dean B. Evans5), Suresh Vaikkakara6), J. Ian Mason4)*, and Richard Quinton1),7)*

1)Department of Endocrinology, The Newcastle upon Tyne NHS University Hospitals Foundation Trust, Newcastle upon Tyne, UK

2) Keenan Research Centre in the Li Ka Shing Knowledge Institute, St. Michael’s Hospital and University of Toronto, Toronto, ON, Canada

3) Department of Cellular Pathology, The Newcastle upon Tyne NHS University Hospital Foundation Trust, Newcastle upon Tyne, UK

4) Reproductive & Developmental Sciences Division, Centre for Reproductive Biology, The Queen’s Medical Research Institute, University of Edinburgh, Edinburgh, UK

5) Novartis Institute for BioMedical Research, Basel, Switzerland

6) Sri Venkateswara Institute of Medical Sciences, Tirupati, Andhra Pradesh, India

7) Institute for Human Genetics, University of Newcastle upon Tyne, UK

Abstract. Estrogen-secreting adrenal cancers are extremely rare, with feminizing symptoms attributed to aromatase expression in the adrenal tumor. We describe a case of hypogonadotropic hypogonadism as a consequence of aberrant aromatase expression in a patient with adrenocortical adenocarcinoma. A 54 year-old man presented with a two month history of gynecomastia and reduced libido. Endocrine biochemistry at presentation showed hypogonadotropic hypogonadism (LH 2.4 U/L, FSH <1.0 IU/L, testosterone 2.8 nmol/L) with increased serum estrone (E1, 821 pmol/L) and estradiol (E2, 797 pmol/L) and subclinical ACTH-independent hypercortisolism (serum cortisol post 1mg overnight dexamethasone suppression test, 291 nmol/L). A right adrenal mass was identified on CT scanning and the patient underwent an open adrenalectomy. Post-operative evaluation showed normalization of serum levels of E1 (95 pmol/L), E2 (109 pmol/L), testosterone (11.4 nmol/L), LH (4.1 U/L) and FSH (5.9 IU/L), and of cortisol dynamics. Immunohistochemistry of the adrenal cancer confirmed aberrant expression of aromatase in most, although not all, carcinoma cells. Transcripts associated with utilization of promoters II, I.1 and I.3 were prominently represented in the tumor aromatase mRNA. This case highlights that clinical features of feminizing adrenocortical carcinomas can be secondary to estrogen production by aberrantly transcribed and translated aromatase within the tumor. Even in males, gonadotropin secretion is subject to predominantly estrogen-mediated feedback-inhibition. The diagnosis of adrenocortical adenocarcinoma should be considered in men presenting with low testosterone and gonadotropins, particularly in the presence of feminizing features.

Key words: Aromatase, Hypogonadotropic hypogonadism, Adrenocortical adenocarcinoma, Gynecomastia, Testosterone

ADRENOCORTICAL adenocarcinoma is a rare ma- lignancy with an incidence of approximately 1-2 per million population [1, 2]. Although clinical evidence of steroid excess occurs in approximately 60% of cas- es [1], feminizing symptoms secondary to estrogen-se- cretion are extremely uncommon [3, 4]. Microsomal

cytochrome P450 aromatase (aromatase) is the final and rate-limiting step in estrogen biosynthesis and is the product of the CYP19 gene. The enzyme catalyses the conversion of C19 steroids, androstenedione, tes- tosterone and 16a-hydroxyandrostenedione, to estrone (E1), estradiol (E2) and estriol respectively in a tissue- dependent manner, according to substrate availability. In adult humans, aromatase is expressed in ovaries, placenta and adipose tissue with low levels in testes and brain [5]. In the past, feminizing symptoms of ad- renocortical carcinomas had been attributed to periph- eral aromatization of adrenal androgens [6]. However,

Correspondence to: Dr. Richard Quinton, Department of Endocrinology, Royal Victoria Infirmary, Newcastle upon Tyne, NE1 4LP. UK. E-mail: richard.quinton@ncl.ac.uk

*These authors contributed equally to this work.

Table 1 Endocrine investigations pre- and post-operatively.
Pre-operativePost-operativeReference range
Estrone (E1) (pmol/L)821950-330
Estradiol (E2) (pmol/L)7971090-180
Testosterone (nmol/L)2.811.49-25
SHBG (nmol/L)552013-71
Calculated free testosterone (pmol/L)37297215-760
LH (U/L)2.44.13.0-13.0
FSH (IU/L)<1.05.91.3-9.2
11-Deoxycortisol (nmol/L)19.87.85.0-12.1
Cortisol (nmol/L) (post-1mg overnight dexamethasone)291<24<50
Urinary free cortisol (nmol/24h)269 & 46663(0-320)
ACTH (9am) (ng/L)7--(0-47)
DHEA sulfate (umol/L)1.9--(1.1-10.9)
Androstenedione (44) (nmol/L)5.7--(<13)
17a-hydroxyprogesterone (nmol/L)3.3--(0-10)
Aldosterone (pmol/L)169--(100-450)

recent case reports have demonstrated local aromatase mRNA and activity in the tumorous adrenal tissue [7].

We present the case of a man with a short his- tory of gynecomastia and loss of libido secondary to an estrogen-producing adrenocortical carcinoma. Immunostaining of the adrenal tumor revealed the presence of aromatase protein in cytoplasmic aggre- gates suggesting abnormal production and processing of the enzyme.

Case report

A 54 year-old man presented with a 2 month history of bilateral gynecomastia and reduced libido. There was no history of exposure to exogenous estrogens, androgen antagonists or of excess alcohol consump- tion. Physical examination revealed symmetrical 4cm x 4cm gynecomastia and diminished testicular volume (15 and 12 mL). He was found to have profound bio- chemical hypogonadotropic hypogonadism, associat- ed with markedly elevated estrogen levels and abnor- mal cortisol dynamics (Table 1). Contrast-enhanced computerized tomographic scanning identified an ap- proximately 6-6.5cm relatively homogenous enhanc- ing right adrenal mass which demonstrated areas of necrosis on magnetic resonance imaging. Plasma me- tanephrines were normal [normetadrenaline (pmol/L)

246 and 220 (reference range, 0-1000), metadrenaline (pmol/L) 167 and 155 (reference range, 0-600)].

The patient underwent open adrenalectomy and an enlarged and encapsulated 145g (70x70x65mm) right adrenal gland was removed without complica- tion. Intra-operatively, there was no evidence of met- astatic disease. Histological examination revealed that the gland was mostly occupied by a nodule com- posed of predominantly eosinophilic cells with dif- fuse architecture, focal confluent necrosis, moderate nuclear pleomorphism, abnormal mitoses, a raised mi- totic count (at least 11 per 50 high power fields) and broad fibrous bands (Fig. 1A-C). The Weiss score was therefore 6 out of 9, thereby securing the diagno- sis of adrenal cortical carcinoma [8, 9]. The Ki-67 in- dex was 44.1%. There was no capsular, sinusoidal or vascular invasion. The findings were those of an ad- renal cortical adenocarcinoma (pT2 NX MX) [10, 11]. Immunohistochemistry with a mouse monoclonal anti- body against human aromatase (Novartis clone #677) [12] showed most, although not all, carcinoma cells to stain positively. Within the positively staining carci- noma cells, the labeling was diffuse or finely granular within the cytoplasm (Fig. 1D). Of particular note were densely staining cytoplasmic aggregates or “inclusions” of various sizes and shapes observed within many of the positive carcinoma cells (Fig. 1D and E). Aromatase

Fig. 1 Histochemical and immunohistochemical staining of the estrogen-secreting adrenal adenocarcinoma. Figure 1A H&E stained section of the carcinoma showing diffusely arranged large eosinophilic epithelial cells with an area of confluent necrosis (original magnification x200). Figure 1B H&E stained section of the carcinoma demonstrating large eosinophilic epithelial cells with pleomorphic nuclei and an abnormal mitotic figure (arrow) (original magnification x400). Figure 1C H&E stained section showing the carcinoma (left), its capsule (centrally) and adjacent normal adrenal cortex (right) (original magnification x200). Figure 1D Immunohistochemistry for aromatase showing positive staining in carcinoma cells, but not in the adjacent normal adrenal (original magnification x400). Figure 1E Immunohistochemistry for aromatase within the cytoplasm of positively staining carcinoma cells (original magnification x400). Figure 1F Negative (isotype IgG) control (original magnification x400). Figure 1G PCR amplification of aromatase (CYP19) transcripts in the adrenal carcinoma. After reverse transcription complementary DNA samples were amplified using primers specific for the various CYP19 gene transcripts. M=size markers, size of fragments is shown in bp; C=coding region; PII=promoter II specific region; I.1=exon I.I; I.4=exon I.4 and I.3=exon I.3. The -ve was a PCR control.

A

B

C

D

E

F

G

1000

737

341

258

M

C

PII

1.1

1.4

1.3

-ve

Table 2 Urine steroid profiles pre- and post-operatively.
Steroid (ug/24h)Pre-operativePost-operativeNormal adult male (n=20)
MeanSD
5a-series
Androsterone20613561526520
11ß-OH-Androsterone434734855244
Allo-Tetrahydrocortisol130313811205509
5ß-series
Aetiocholanolone117310331308565
11ß-OH-Aetiocholanolone662435545496
Tetrahydrocortisol27959001275481

immunoreactivity was not observed in stromal cells of the adrenal gland. The aromatase transcripts associat- ed with utilization of the gonadal-associated aromatase promoter II were prominently represented in the tumor aromatase mRNA. However, significant amounts of transcripts associated with both promoters I.1 and I.3 were observed, while there was no evidence for pro- moter I.4-associated activity (Fig. 1G).

Urine steroid profiles pre-operatively showed an in- crease in 5ß- relative to 5a-reduced metabolites and of pregnanediol, pregnenetriol and a number of unusu- al additional metabolites suggestive of steroid produc- tion by the adrenal mass (Table 2). Post-operatively, the relative excesses previously noted were no longer present and the ratio of 5- to 5a-reduced steroids was reversed with relatively higher levels of 5a-reduced steroids, consistent with the increased levels normally found in adult males compared to females [13].

Post-operatively, gynecomastia regressed and libido improved, accompanied by normalization of all bio- chemical parameters (Table 1). Three years later there is no clinical, radiological or endocrine evidence of re- current disease.

Discussion

Estrogen-secreting adrenal tumors are an uncom- mon cause of gynecomastia and are almost invariably malignant [14]. Feminizing symptoms are attributed to the in situ generation of E, through aberrant aroma- tisation of adrenal androgens with trivial or no estro- gen secretion occurring in normal adrenals [15]. Our patient’s pre-operative endocrine profile was charac-

teristic of that of an estrogen-secreting tumor with in- creased circulatory levels of E1 and E2, low serum testosterone and gonadotropins and a moderately el- evated serum 11-deoxycortisol. Suppressed gonado- tropin secretion is a recognized consequence of aro- matization of testosterone to E2 [16], demonstrating that even in men gonadotropin secretion is predomi- nantly under estrogen-mediated feedback inhibition. Increased serum 11-deoxycortisol suggests impaired 11ß-hydroxylase activity within the tumor. Although decreased 11ß-hydroxylase activity is a characteristic feature of feminizing adrenocortical adenocarcinomas [4, 15, 17], it is unclear whether this is a consequence of estrogen excess or a direct effect of the neoplastic process.

Even though the patient was normotensive without Cushingoid features, the abnormal cortisol dynamics illustrate the importance of actively seeking evidence for autonomous cortisol secretion in all patients pre- senting with feminizing adrenocortical carcinomas. Patients with tumorous adrenal cortisol hypersecretion can develop cardiovascular collapse post-adrenalecto- my, due to “suppression” of the contralateral normal adrenal gland, unless peri- and post-operative gluco- corticoids are administered. In the setting of a low se- rum testosterone, the absence of supranormal levels of LH and FSH was suggestive of hypogonadotropic hy- pogonadism, although unequivocal diagnosis would have required a GnRH stimulation test.

Aromatase mRNA has been previously demonstrat- ed in feminizing adrenal tumors [7, 17]. However, the presence of biologically active translated protein has typically been assumed, based on the detection of aro-

matase activity by the tritiated water technique, even though aromatase activity in adrenocortical adenocar- cinoma is approximately 50-fold lower than that found in human placenta [4, 17]. In the present report, we demonstrated the distribution of the translated aro- matase protein within an adrenal tumor, the feminizing features and endocrinological presentation of the pa- tient serving as an indicator that the immunoreactive aromatase was biologically functional. Under normal conditions, aromatase is located in the endoplasmic reticulum of estrogen producing cells [18]. In our pa- tient, cellular immunostaining was heterogeneous. In positive cells, immunostaining was diffuse or finely granular within the cytoplasm, with additional immu- nopositive inclusions or aggregates.

The site-specific expression of aromatase is regulat- ed by tissue-specific promoters: proximal promoter II in ovary, distal promoter I.1 in placenta and promoters I.3 and I.4 in adipose tissue [5]. Promoter II is proxi- mal to the translation start site and is considered to be the primitive promoter regulating estrogen expression. As has been shown in adrenocortical tumors [19], and others, promoter II was prominent for the pathological aromatase expression in our patient. However, tran- scripts derived from promoters I.1 and I.3 were also detectable. While promoter I.3 transcripts have been previously reported in adrenal tumors, to our knowl- edge this is the first report of adrenal tumor aromatase expression utilizing the so-called ‘placental’ promot- er, which occurred in the absence of placental tissue within the tumor.

Although serum estrogens were elevated, the E1:E2 ratio was approximately one. In the context of low pre-operative peripheral testosterone levels, it has pre- viously been suggested that E2 production originates

from peripheral E, conversion via 17-hydroxysteroid dehydrogenase type 1 (17HSD1) [7, 17]. However, we recently showed that adrenal cancer cells express the reductive type 5 17ß-hydroxysteroid dehydroge- nase (AKR1C3) which is capable of converting E, to E2 as well as androstenedione to testosterone, indi- cating the capacity for direct E2 production by the tu- mor [12]. Locally synthesized estrogens may promote cellular proliferation in aromatase expressing female breast carcinomas [20, 21]. Although we did not mea- sure receptor activity, estrogen receptor ß expression has previously been reported in a steroid-producing adrenocortical cell line [22]. Accordingly, it is possi- ble that estrogens produced by the adrenal carcinoma cells may have contributed to the high rate of cellular proliferation observed in our patient.

In summary, although estrogen-secreting adreno- cortical carcinomas are extremely rare, there are now sufficient data to indicate how aberrant aromatase expression can result in a characteristic clinical and biochemical presentation. The diagnosis should be considered in all patients presenting with hypogonad- otropic hypogonadism, particularly if there is also gy- necomastia.

Acknowledgements

The authors gratefully acknowledge the as- sistance of the nursing staff of the Programmed Investigations Unit and the laboratory staff of the Clinical Biochemistry Department at The Newcastle upon Tyne NHS University Hospital Foundation Trust, Newcastle upon Tyne, UK and the Newcastle Supra Regional Assay Service (SAS) for Endocrinology.

References

1. Allolio B, Fassnacht M (2006) Clinical review: Adrenocortical carcinoma: clinical update. J Clin Endocrinol Metab 91:2027-2037.

2. Advani A, Vaikkakara S, Gill MS, Arun CS, Pearce SH, Ball SG, James RA, Lennard TW, Bliss RD, Quinton R, Johnson SJ (2008) Impact of standardised reporting in adrenocortical carcinoma: a single centre clinicopatho- logical review. J Clin Pathol 61:939-944.

3. Gabrilove JL, Sharma DC, Wotiz HH, Dorfman RI (1965) Feminizing Adrenocortical Tumors in the Male. A Review of 52 Cases Including a Case Report.

Medicine (Baltimore) 44:37-79.

4. Kimura M, Itoh N, Tsukamoto T, Kumamoto Y, Takagi Y, Mori Y (1995) Aromatase activity in an estrogen- producing adrenocortical carcinoma in a young man. J Urol 153:1039-1040.

5. Simpson ER, Mahendroo MS, Means GD, Kilgore MW, Hinshelwood MM, Graham-Lorence S, Amarneh B, Ito Y, Fisher CR, Michael MD, et al. (1994) Aromatase cytochrome P450, the enzyme responsible for estrogen biosynthesis. Endocr Rev 15:342-355.

6. Hemsell DL, Edman CD, Marks JF, Siiteri PK,

MacDonald PC (1977) Massive extranglandular arom- atization of plasma androstenedione resulting in femi- nization of a prepubertal boy. J Clin Invest 60:455-464.

7. Phornphutkul C, Okubo T, Wu K, Harel Z, Tracy TF, Jr., Pinar H, Chen S, Gruppuso PA, Goodwin G (2001) Aromatase p450 expression in a feminizing adrenal ad- enoma presenting as isosexual precocious puberty. J Clin Endocrinol Metab 86:649-652.

8. Weiss LM (1984) Comparative histologic study of 43 metastasizing and nonmetastasizing adrenocortical tu- mors. Am J Surg Pathol 8:163-169.

9. Weiss LM, Medeiros LJ, Vickery AL, Jr. (1989) Pathologic features of prognostic significance in adre- nocortical carcinoma. Am J Surg Pathol 13:202-206.

10. Macfarlane DA (1958) Cancer of the adrenal cortex; the natural history, prognosis and treatment in a study of fifty-five cases. Ann R Coll Surg Engl 23:155-186.

11. Sullivan M, Boileau M, Hodges CV (1978) Adrenal cortical carcinoma. J Urol 120:660-665.

12. Nicol MR, Papacleovoulou G, Evans DB, Penning TM, Strachan MW, Advani A, Johnson SJ, Quinton R, Mason JI (2009) Estrogen biosynthesis in human H295 adrenocortical carcinoma cells. Mol Cell Endocrinol 300:115-120.

13. Raven PW, Taylor NF (1996) Sex differences in the hu- man metabolism of cortisol. Endocr Res 22:751-755.

14. Fassnacht M, Kenn W, Allolio B (2004) Adrenal tu- mors: how to establish malignancy? J Endocrinol Invest 27:387-399.

15. McKenna TJ, O’Connell Y, Cunningham S, McCabe M, Culliton M (1990) Steroidogenesis in an estrogen- producing adrenal tumor in a young woman: compari- son with steroid profiles associated with cortisol- and

androgen-producing tumors. J Clin Endocrinol Metab 70:28-34.

16. Santen RJ (1975) Is aromatization of testosterone to es- tradiol required for inhibition of luteinizing hormone secretion in men? J Clin Invest 56:1555-1563.

17. Young J, Bulun SE, Agarwal V, Couzinet B, Mendelson CR, Simpson ER, Schaison G (1996) Aromatase ex- pression in a feminizing adrenocortical tumor. J Clin Endocrinol Metab 81:3173-3176.

18. Ma CX, Adjei AA, Salavaggione OE, Coronel J, Pelleymounter L, Wang L, Eckloff BW, Schaid D, Wieben ED, Weinshilboum RM (2005) Human aro- matase: gene resequencing and functional genomics. Cancer Res 65:11071-11082.

19. Bouraima H, Lireux B, Mittre H, Benhaim A, Herrou M, Mahoudeau J, Guillon-Metz F, Kottler ML, Reznik Y (2003) Major hyperestrogenism in a feminizing adreno- cortical adenoma despite a moderate overexpression of the aromatase enzyme. Eur J Endocrinol 148:457-461.

20. Sasano H, Harada N (1998) Intratumoral aromatase in human breast, endometrial, and ovarian malignancies. Endocr Rev 19:593-607.

21. Moreau F, Mittre H, Benhaim A, Bois C, Bertherat J, Carreau S, Reznik Y (2009) Aromatase expression in the normal human adult adrenal and in adrenocortical tumors: biochemical, immunohistochemical, and mo- lecular studies. Eur J Endocrinol 160:93-99.

22. Montanaro D, Maggiolini M, Recchia AG, Sirianni R, Aquila S, Barzon L, Fallo F, Ando S, Pezzi V (2005) Antiestrogens upregulate estrogen receptor beta ex- pression and inhibit adrenocortical H295R cell prolif- eration. J Mol Endocrinol 35:245-256.